| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 88 No. 1 July 2002, pp. 236-248
Copyright ©2002 by the American Physiological Society
Traumatic Brain Injury Laboratory, Cara Phelan Centre for Trauma Research and Department of Anaesthesia, St. Michael's Hospital, University of Toronto, Toronto, Ontario M5B 1W8 Canada
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Tian, Guo-Feng and Andrew J. Baker. Protective Effect of High Glucose Against Ischemia-Induced Synaptic Transmission Damage in Rat Hippocampal Slices. J. Neurophysiol. 88: 236-248, 2002. Cerebral ischemic damage is an important cause of morbidity and mortality. However, there is conflicting evidence regarding the effect of the extracellular glucose concentration in focal and global ischemic injury. This study was designed to investigate this effect in ischemia-induced synaptic transmission damage in rat hippocampal slices. Slices were superfused with artificial cerebrospinal fluid (ACSF) containing various concentrations of glucose before and after ischemia. The evoked somatic postsynaptic population spike (PS) and dendritic field excitatory postsynaptic potential (fEPSP) were extracellularly recorded in the CA1 stratum pyramidal cell layer and s. radiatum after stimulation of the Schaeffer collaterals, respectively. The glucose concentration in ACSF before and after ischemia determined the duration of ischemia tolerated by synaptic transmission as demonstrated by complete recovery of the somatic PS and dendritic fEPSP. Specifically, the somatic PS and dendritic fEPSP completely recovered following 3, 4, and 5 min of ischemia only when slices were superfused with ACSF containing 4, 10, and 20 mM glucose before and after ischemia, respectively. The latencies of the somatic and dendritic ischemic depolarization (ID) occurrence in the CA1 s. pyramidal cell layer and s. radiatum were significantly longer with 10 than 4 mM glucose in ACSF before ischemia and significantly longer with 20 than 10 mM glucose in ACSF before ischemia. Regardless of the glucose concentration in ACSF before and after ischemia, the somatic PS and dendritic fEPSP only partially recovered when ischemia was terminated at the occurrence of ID. These results indicate that high glucose in ACSF during the period before and after ischemia significantly protects CA1 synaptic transmission against in vitro ischemia-induced damage through postponing the occurrence of ID.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
With few exceptions,
glucose is the obligatory energy substrate for the brain, and it is
almost entirely oxidized to carbon dioxide and water in the brain
(Magistretti 1999
). Glucose and oxygen are two essential
energy supplies for the brain. Therefore deficiencies in glucose and/or
oxygen that result from hypoglycemia, hypoxia, and ischemia can cause
morphological and functional injuries in the brain. Indeed, cerebral
ischemic damage is an important cause of morbidity and mortality, and
such ischemic insults may be global or result from a focal interruption
of blood flow (Wass and Lanier 1996). However, there is
conflicting evidence regarding the effects of the extracellular glucose
concentration in focal and global cerebral ischemic injury. For
example, certain studies have suggested that hyperglycemia exacerbates
cerebral ischemic damage and that this may occur by the attendant
increased lactic acid production and the generation of reactive oxygen
species (Gisselsson et al. 1999; Huang et al.
1996; Katsura et al. 1994; Kondo et al.
2000; Li and Siesjö 1997; Li et al.
1999; Lundgren et al. 1992; Rehncrona et
al. 1981; Smith et al. 1986). Other reports
suggest that an elevated supply of glucose before hypoxic-ischemia reduces brain damage (Ginsberg et al. 1987; Kraft
et al. 1990; Roos 1999; Vannucci et al.
1996; Zasslow et al. 1989). The release of
glutamate during brain ischemia may play a critical role in ischemic
damages (Choi and Rothman 1990). Although one study has shown that hyperglycemia enhanced the accumulation of extracellular glutamate in the cerebral cortex (not in hippocampus) of rats subjected
to forebrain ischemia (Li et al. 1999), in contrast a
number of studies have shown that hyperglycemia can lower the glutamate
concentration during the cerebral ischemic period (Choi et al.
1994; Guyot et al. 2000; Kanthan et al.
1996; Phillis et al. 1999). These observations
are all from in vivo experiments, and the distribution and level of
extracellular glucose in the vulnerable and less vulnerable regions are uncertain.
Neuronal functional response to ischemia, as measured by
electrophysiology, can be influenced by extracellular glucose
concentrations. Elevated glucose in artificial cerebrospinal fluid
(ACSF) improves the recovery of neuronal function from anoxic
challenges in hippocampal slices (Grigg and Anderson
1989; Roberts 1993; Roberts and Sick 1992; Schurr et al. 1987; Wang et al.
2000; Zhu and Krnjevi
1999). More
recently, we have systematically examined the impact of extracellular
glucose on the anoxic synaptic transmission in hippocampal slices. Our
study showed that the levels of extracellular glucose is an important
factor in determining anoxic synaptic transmission damage. A lack of
glucose is the major cause of anoxic synaptic transmission damage and
high levels of extracellular glucose can prevent this damage
(Tian and Baker 2000). Another recent study showed that
high glucose administrated before the ischemic insult has a
neuroprotective effect against ischemia-induced damage to cultured
cortical neurons (Seo et al. 1999).
The in vitro ischemia model provides a convenient way to examine the
relation between extracellular glucose and neuronal function during
ischemia-like insults. Although the effect of glucose on functional
recovery in hippocampal slice is well established, this study was
designed to evaluate the neuronal functional impact of varying glucose
supply during the period before and after ischemia with emphasis on the
effect of high glucose on the ischemic depolarization (ID) and how this
correlates with functional recovery following ischemia. Lack of
recovery of synaptic transmission was used as the index of
ischemia-induced neuronal functional injury. We superfused hippocampal
slices with ACSF containing different concentrations of glucose before
and after ischemia and then examined the recovery of synaptic
transmission following different durations of ischemia. Because the ID
has been suggested as a critical factor for the ischemia-induced
synaptic transmission damage (Mayevsky 1990; Obeidat and Andrew 1998; Watson and Lanthorn
1995), in one set of experiments we fixed the duration of the
ischemic insult and in another set of experiments we terminated the
ischemic insult at the occurrence of ID.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Hippocampal slices from rat brains were prepared as described
previously (Tian and Baker 2000). Briefly, male
Sprague-Dawley rats (39-40 days old, 170-190 g) were anesthetized
with 2.0-2.5% halothane in oxygen and then decapitated. The brain was
immediately removed and maintained in an ice-cold ACSF for 3-5 min
before slicing. The brain was mounted on an aluminum block and
transversely sliced (~400 µm) in ice-cold (<3°C) ACSF using a
vibratome (series 1000, Technical Products International, St. Louis,
MO). The slices were then kept in oxygenated ACSF at room temperature
(22-23°C) for 
1 h before the experiment. The composition of the
standard ACSF was (in mM) 126 NaCl, 3 KCl, 1.4 KH2PO4, 2.4 CaCl2, 1.3 MgSO4, 26 NaCO3, and 10 glucose. The ACSF was equilibrated
and continuously bubbled with 95% O2-5%
CO2, pH 7.4 at 36.5 ± 0.5°C.
For electrophysiological recording, the slice was placed in a closed,
box-like superfusion chamber made from removable plates; a small slit
allowed access to the electrodes (Tian and Baker 2000).
The slice was fully submerged in the superfusion chamber and
continuously superfused (7-8 ml/min) with ACSF equilibrated and
continuously bubbled with 95% O2-5%
CO2. Humidified, warmed 95%
O2-5% CO2 was blown over
the chamber to ensure a warm oxygenated local environment. All
recordings were made at slice temperatures between 36 and 37°C. To
achieve a stable experimental temperature, the ACSF was warmed before
superfusing the slice, using a water bath controlled by a temperature
controller. The temperature of the ACSF in the superfusion chamber was
continuously monitored using a YSI 4600 series precision thermometer
with a micro YSI 451 temperature sensor (YSI, Yellow Springs, OH). The
slice was made "ischemic" by superfusing it with ischemic ACSF
substituted equimolar mannitol for glucose and preequilibrated and
continuously bubbled with 95% N2-5%
CO2, and humidified, warmed 95%
N2-5% CO2 was blown over
the chamber to ensure a warm oxygen-free local environment.
Field potentials were recorded extracellularly through glass pipettes
filled with 150 mM NaCl (tip resistance of 2-5 M
). The electrode
was placed in the CA1 stratum pyramidal or s. radiatum to monitor
somatic population spikes (PS) or dendritic field excitatory postsynaptic potentials (fEPSP), respectively (Zhang et al.
1999). Signals were recorded using an Axopatch 200B amplifier
(Axon Instrument, Foster City, CA), and data were stored and analyzed
with pCLAMP software (version 6.0.4, Axon Instrument).
Schaeffer collaterals were electrically stimulated using a bipolar tungsten electrode placed in the s. radiatum of CA1. Stimulation pulses of constant current (0.1 ms, 0.4-0.9 mA for evoking maximal somatic PS, but only 0.02-0.08 mA to evoke maximal dendritic fEPSP without contamination by the PS) were generated by a Grass S88 stimulator (Grass Instrument, West Warwick, RI) and delivered through an isolation unit (PSIU6) every 30 s.
The evoked extracellular responses in the CA1 s. pyramidal cell layer
after stimulation of the Schaeffer collaterals started with a small
downward stroke (Fig. 1A,
), the presynaptic volley (PV) immediately following the stimulus
artifact (
in figures). The PV was followed by an upward waveform,
the fEPSP. During the fEPSP, there was a sharp downward stroke, the
postsynaptic population spike (PS) (Fig. 1A, *). The somatic
PV amplitude was evaluated by calculating the voltage difference
between the baseline and its negative peak. The somatic PS amplitude
was evaluated by calculating the voltage difference between its
positive peak and its negative peak (Tian and Baker
2000). The typical evoked responses in the stratum radiatum
started also with a small downward stroke (Fig. 2A, ), the presynaptic
volley (PV) immediately following the stimulus artifact ( in
figures). The PV was followed by an downward waveform, the dendritic
fEPSP (Fig. 2A, *). The dendritic fEPSP amplitude was
evaluated by calculating the voltage difference between its negative
peak and the baseline. The dendritic PV was usually not measurable
because it was always immersed in fEPSP due to the low intensity of
stimulation.
|
|
To examine the effect of glucose on ischemia-induced synaptic transmission damage, slices were superfused with ACSF containing different concentrations of glucose before and after ischemia. In one set of experiments, glucose concentrations were 4, 10 and 20 mM, and in another series, glucose was used in combination with mannitol (4 mM glucose plus 6 mM mannitol and 10 mM glucose plus 10 mM mannitol). The slices were superfused with ACSFs with different concentrations of glucose for 30 min before ischemia and 60 min restitution after ischemia, and during ischemia with glucose-free ACSF. The recovery of synaptic transmission following the ischemic insult was assessed by expressing the somatic PS amplitude or dendritic fEPSP amplitude at the end of 60-min restitution following ischemia as a percentage of control (before ischemia) amplitude.
All experimental data in different groups are expressed as group
means ± SD. Student's t-test (for 2 groups) and ANOVA
followed by Dunnett's test (for 3 groups) were used to determine the
statistical significance of any differences, and significance was
defined at P < 0.05.
All chemicals that composed the ACSF were purchased from Fisher Scientific (Fair Lawn, NJ).
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Extracellular recordings of evoked responses
Figure 1A (left) and Fig. 2A
(left) show the evoked responses extracellularly recorded
from the CA1 s. pyramidal cell layer and s. radiatum after stimulation
of the Schaeffer collaterals, respectively, when the slice was
superfused with the ACSF containing 4 mM glucose. The typical evoked
responses in the CA1 s. pyramidal cell layer consisted of three parts:
the PV, the fEPSP, and the PS as described in METHODS. The
extracellular recordings could be stable for 
3 h when the slice was
superfused with ACSF containing 4, 10, and 20 mM glucose in normal
conditions (Tian and Baker 2000). The maximal PV and PS
amplitudes were 0.9 ± 0.3 and 7.0 ± 1.2 mV
(n = 59), respectively. The typical evoked responses in
the stratum radiatum consisted of two parts: the PV and the fEPSP as
well as described in METHODS. The maximal amplitude of dendritic fEPSP without population spike contamination was 2.1 ± 0.3 mV (n = 51), and the PV was usually not measurable
because it was always immersed in fEPSP due to the low intensity of
stimulation. There were no significant differences in somatic PV,
somatic PS and dendritic fEPSP when recorded in different extracellular
glucose concentration, indicating that the alterations in extracellular glucose concentration did not affect the CA1 synaptic physiology in the
"normal" condition.
Effects of ischemia on evoked responses
When slices were superfused with ACSF containing 4 mM glucose (approximating extracellular glucose concentration in vivo) before ischemia, the PV persisted until the end of a 3-min ischemic challenge (Figs. 1A and 2A) but disappeared at the end of a 4-min ischemic insult (Figs. 1C and 2C). However, the PV persisted until the end of a 4-min ischemic challenge when the slices were superfused with ACSF containing 10 mM glucose before ischemia (Figs. 3A and 4A). Furthermore, the PV persisted until the end of a 5-min ischemic challenge when the slices were superfused with ACSF containing 20 mM glucose before ischemia (Figs. 3C and 4C). Thus, while a 4-min ischemic insult impaired the excitation and conduction of Schaeffer collaterals when the slices were superfused with ACSF containing 4 mM glucose before ischemia, this impairment was prevented when, before ischemia, the slices were superfused with ACSF containing 10 and 20 mM glucose for the 4- and 5-min ischemic challenges, respectively. These results indicate high glucose in ACSF prior to ischemia provided a protective effect to the axonal terminals of the Schaeffer collaterals during the ischemic challenge.
|
|
In every case, the fEPSP completely disappeared at the end of a 3-min ischemic challenge when the slices were superfused with ACSF containing 4 mM glucose before ischemia (Figs. 1A and 2A). However, the fEPSP usually persisted until the end of a 4-min ischemic challenge (Figs. 3A and 4A), when the slices were superfused with ACSF containing 10 mM glucose before ischemia. Furthermore, the fEPSP usually persisted until the end of even a 5-min ischemic challenge (Figs. 3C and 4C), when the slices were superfused with ACSF containing 20 mM glucose before ischemia. This retention of the fEPSP until the end of a 4- and even a 5-min ischemic challenge when the slices were superfused with ACSF containing 10 and 20 mM glucose before ischemia, respectively, indicated that synaptic terminals could still be excited, and retained their ability to release neurotransmitters, and that postsynaptic membrane receptors retained their ability to respond to released neurotransmitters. In contrast, those abilities were abolished at the end of a 3-min ischemic challenge when the slices were superfused with ACSF containing 4 mM glucose before ischemia.
In all cases, PS completely disappeared within 3 min of the introduction of ischemia regardless of the glucose concentrations in ACSF before ischemia (Figs. 1 and 3).
Effect of glucose concentration on the recovery of synaptic transmission following ischemia
Table 1 summarizes the somatic PS and dendritic fEPSP recoveries following different durations of ischemia in the presence of different concentrations of glucose in ACSF before and after ischemia.
|
When slices were superfused with ACSF containing 4 mM glucose before and after ischemia, following a 3-min ischemic challenge, recoveries of somatic PV, somatic PS, and dendritic fEPSP amplitudes at the end of 60-min restitution were 95 ± 4% (n = 7), 103 ± 5% (n = 7), and 101 ± 6% (n = 7) (Figs. 1, A and B, and 2, A and B, and Table 1), respectively. However, although somatic and dendritic PV always recovered following a 4-min ischemic insult, the somatic fEPSP, the somatic PS, and the dendritic fEPSP displayed no such recovery after 60-min restitution (Figs. 1, C and D, and 2, C and D, and Table 1). These results indicated that when the slices were superfused with ACSF containing 4 mM glucose before and after ischemia, although the synaptic transmission could completely recover following a 3-min ischemic challenge, synaptic transmission was irreversibly damaged following a 4-min ischemic insult.
When the slices were superfused with ACSF containing 10 mM glucose before and after ischemia, following a 4-min ischemic challenge, recoveries of somatic PV, somatic PS, and dendritic fEPSP amplitudes at the end of 60-min restitution were 101 ± 13% (n = 6), 104 ± 10% (n = 6), and 99 ± 4% (n = 6) (Figs. 3, A and B, and 4, A and B, and Table 1). With 20 mM glucose in ACSF before and after ischemia, following a 5-min ischemic challenge, recoveries of somatic PV, somatic PS, and dendritic fEPSP amplitudes at the end of 60-min restitution were 98 ± 11% (n = 7), 107 ± 6% (n = 7), and 99 ± 7% (n = 7; Figs. 3, C and D, and 4, C and D, and Table 1). These results indicate that, when the slices were superfused with ACSF containing 10 and 20 mM glucose before and after ischemia, synaptic transmission could completely recover following the 4- and even 5-min ischemic challenges, respectively; thus the 4- and even 5-min ischemic challenges could not damage synaptic transmission under these conditions.
Effect of glucose on the latency and amplitude of ischemic depolarization
When slices were superfused with ACSF containing 4 mM glucose
before ischemia, the DC field potential always shifted rapidly to
negative value during the 4-min ischemic insults. This rapid negative-going DC shift was an ischemia-induced spreading
depression-like depolarization, the ischemic depolarization (ID)
(Watson and Lanthorn 1995). However, such ID did not
occur during the 4- and even 5-min ischemic challenges with 10 and 20 mM glucose in ACSF before ischemia, respectively. Because the ID may
play a critical role in determining the ischemia-induced synaptic
transmission damage (Mayevsky 1990; Obeidat
and Andrew 1998; Watson and Lanthorn 1995), we
examined effects of glucose on the latency of the ID occurrence and on the recovery of synaptic transmission following ischemia terminated at
the occurrence of ID.
The latencies of the somatic ID occurrence were 3.60 ± 0.15 min (n = 6), 4.61 ± 0.14 min (n = 7), and 6.42 ± 0.23 min (n = 7), and the somatic ID amplitudes were 4.72 ± 0.76 mV (n = 6), 4.82 ± 0.78 mV (n = 7), and 4.63 ± 0.99 mV (n = 7; Figs. 5, 7, and 9 and Table 2) when slices were superfused with ACSF containing 4, 10, and 20 mM glucose before ischemia, respectively. The latencies of the dendritic ID occurrence were 3.55 ± 0.16 min (n = 8), 4.76 ± 0.14 min (n = 7), and 6.45 ± 0.32 min (n = 8) and the dendritic ID amplitudes were 8.51 ± 1.23 mV (n = 8), 8.08 ± 0.95 mV (n = 7), and 6.76 ± 0.79 mV (n = 8; Figs. 6, 8, and 10 and Table 3) when slices were superfused with ACSF containing 4, 10, and 20 mM glucose before ischemia, respectively. The latencies of the somatic and dendritic ID occurrence were significantly longer with 10 mM glucose in ACSF before ischemia than with 4 mM glucose in ACSF before ischemia and significantly longer with 20 mM glucose in ACSF before ischemia than with 10 mM glucose in ACSF before ischemia (Tables 2 and 3). Although the differences of the somatic ID amplitudes with 4, 10, and 20 mM glucose in ACSF were not statistically significance (Table 2), the differences of the dendritic ID amplitudes with 20 mM glucose in ACSF were significantly larger than with 4 and 10 mM glucose in ACSF before ischemia (Table 3). These results indicate that elevated glucose concentrations in ACSF before ischemia postponed the occurrence of ID.
|
|
|
|
Effect of glucose on the recovery of synaptic transmission following ischemia terminated at the occurrence of ischemic depolarization
When slices were superfused with ACSF containing 4, 10, and 20 mM glucose before and after ischemia, following ischemia that was terminated at the occurrence of somatic ID, the somatic PV always completely recovered. However, the recoveries of somatic PS amplitudes at the end of 60 min restitution were 14 ± 5% (n = 6), 25 ± 17% (n = 7), and 20 ± 11% (n = 7), respectively (Figs. 5, 7, and 9 and Table 2). When slices were superfused with ACSF containing 4, 10, and 20 mM glucose before and after ischemia, following ischemia that was terminated at the occurrence of dendritic ID, recoveries of dendritic fEPSP amplitudes at the end of 60-min restitution were 21 ± 11% (n = 8), 22 ± 6% (n = 7), and 19 ± 5% (n = 8), respectively (Figs. 6, 8, and 10 and Table 3). The recoveries of the somatic PS and dendritic fEPSP amplitudes were not significantly different between groups (Tables 2 and 3). These results indicate that synaptic transmission could only partially recover following ischemia terminated at the occurrence of ID, regardless of the glucose concentration in ACSF before and after ischemia.
|
|
|
|
Effect of osmolality on the latency and amplitude of ischemic depolarization
Because the ACSFs contained different concentrations of glucose before and after ischemia, their osmolalities were different, and the protective effects of higher glucose in ACSF before and after ischemia might be due to the higher osmolality. Therefore we examined the effect of osmolality on the latency of the ID occurrence and on the recovery of synaptic transmission following ischemia terminated at the occurrence of ID.
To evaluate the effects of higher osmolality in the ACSF, 6 mM mannitol was added into ACSF containing 4 mM glucose to yield an ACSF with an osmolality equal to that of ACSF containing 10 mM glucose, and 10 mM mannitol was added into ACSF containing 10 mM glucose to yield an ACSF with an osmolality equal to that of ACSF containing 20 mM glucose. When slices were superfused with ACSF containing 4 mM glucose plus 6 mM mannitol before ischemia, the somatic ID occurred at 3.77 ± 0.15 min (n = 6) after the introduction of ischemia, and the amplitude of somatic ID was 4.45 ± 0.75 mV (n = 6; Fig. 11, A and B, and Table 2). With 10 mM glucose plus 10 mM mannitol in ACSF before ischemia, the somatic ID occurred at 4.68 ± 0.19 min (n = 7) after the introduction of ischemia, and the amplitude of somatic ID was 4.54 ± 0.76 mV (n = 7; Fig. 12, A and B, and Table 2). The latencies of the somatic ID occurrence in ACSF with 4 mM glucose plus 6 mM mannitol before ischemia were not significantly different from those in ACSF with 4 mM glucose without mannitol before ischemia (Table 2). The latencies of the ID occurrence in ACSF with 10 mM glucose plus 10 mM mannitol before ischemia were also not significantly different from those in ACSF with 10 mM glucose without mannitol before ischemia (Table 2). The differences of the ID amplitudes were not statistically significant between slices exposed to 4 mM glucose and 4 mM glucose plus 6 mM mannitol in ACSF before ischemia nor between slices exposed to 10 mM glucose and 10 mM glucose plus 10 mM mannitol in ACSF before ischemia (Table 2). These results indicate that the delay in the occurrence of ID with higher glucose in ACSF before ischemia was not due to the higher osmolality of the ACSF.
|
|
Effect of osmolality on the recovery of synaptic transmission following ischemia terminated at the occurrence of ischemic depolarization
When slices were superfused with ACSF containing 4 mM glucose plus 6 mM mannitol before and after ischemia, following ischemia that was terminated at the occurrence of somatic ID, although the somatic PV always completely recovered, recovery of somatic PS amplitudes at the end of 60-min restitution was 15 ± 6% (n = 6; Fig. 11, C and D, and Table 2). With 10 mM glucose plus 10 mM mannitol in ACSF before and after ischemia, following ischemia that was terminated at the occurrence of somatic ID, although the somatic PV always completely recovered, recovery of somatic PS amplitudes at the end of 60 min restitution was 19 ± 9% (n = 7; Fig. 12, C and D, and Table 2). Recovery of somatic PS amplitudes following ischemia terminated at the occurrence of somatic ID in ACSF with 4 mM glucose plus 6 mM mannitol before and after ischemia was not significantly different from that in ACSF with 4 mM glucose without mannitol before and after ischemia (Table 2). Recovery of somatic PS amplitudes following ischemia terminated at the occurrence of somatic ID in ACSF with 10 mM glucose plus 10 mM mannitol before and after ischemia was not significantly different from that in ACSF with 10 mM glucose without mannitol before and after ischemia (Table 2). These results indicate that the protective effects of higher glucose in ACSF before and after ischemia against ischemia-induced synaptic transmission damage were not due to higher osmolality of the ACSF.
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Neurons are generally believed to be dependent on a continuous supply of oxygen and glucose. Ischemic injury is presumed to occur through the rapid depletion of energy stores and the attendant depletion of ATP. It has been suggested that ID plays a critical role in ischemia-induced brain damage. We examined the hypothesis that if an increased supply of glucose were provided prior to in vitro ischemia brain tissue could withstand longer ischemic challenges through postponing the occurrence of ID. The implications include the ability of tissue to use increased stores of glucose by anaerobic metabolism and delaying the depletion of ATP, or the increased storage of aerobically produced energy containing products. A number of studies have led to the suggestion that the detrimental effects of the products of increased anaerobic metabolism from an increased supply of glucose outweigh any potential benefits of that increased energy supply. In contrast, we have demonstrated that an increased glucose concentration in ACSF during the before and after ischemic period protects synaptic transmission against ischemia-induced damage in rat hippocampal slices, and this occurs through the postponement of the occurrence of ID.
Protective effect of high glucose against ischemia-induced synaptic transmission damage
With 4 mM glucose in ACSF before and after ischemia, although
synaptic transmission could completely recover following a 3 min of
ischemic challenge, it was irreversibly abolished following a 4 min of
ischemia. However, in slices superfused with ACSF containing 10 mM
glucose before and after ischemia, the synaptic transmission was not
damaged by 4 min of ischemic challenge; even 5 min of ischemic
challenge did not impair the synaptic transmission when the slices were
superfused with ACSF containing 20 mM glucose before and after
ischemia. These results demonstrate that elevated glucose concentration
in ACSF before and after ischemia protects the synaptic transmission
against the ischemia-induced damage. In addition to such a protective
effect in superfused hippocampal slices, high glucose in culture media
has also been shown to provide neuroprotective effects against
ischemia-induced damage to cultured cortical neurons (Seo et al.
1999).
Exposure to high glucose before ischemia may increase energy stores in
the slices (Siemkowicz and Hansen 1978) and thereby postpone the energy depletion during ischemia. This concept is supported by evidence that hyperglycemia improves the energy state in
ischemic neocortex, in both the penumbra and the focus regions by
increasing tissue ATP values (Folbergrová et al.
1992; Vannucci et al. 1996; Wagner and
Lanier 1994). In addition, exposure to high glucose before
ischemia enhances mitochondrial potentials, reducing
Ca2+ release from the mitochondria and lowering
intracellular Ca2+ concentration
(Grøndahl et al. 1998; Lobner and Lipton
1993; Nabetani et al. 1997; Zhang and
Lipton 1999) and glutamate levels (Choi et al.
1994; Guyot et al. 2000; Kanthan et al.
1996; Phillis et al. 1999; Seo et al.
1999). This evidence suggests the mechanism by which elevated
glucose exposure attenuates the effect of ischemia in vitro. It also
forms the rationale for its possible application to in vivo ischemia.
In vitro and in vivo differences and their relevance
It is important to recognize that the impact of glucose on
ischemic neurologic injury remains unclear. There are many studies that
demonstrate a detrimental effect on the outcome of ischemic neurologic
injury of elevated glucose exposure using in vivo models (Gisselsson et al. 1999; Huang et al.
1996; Katsura et al. 1994; Kondo et al.
2000; Li and Siesjö 1997; Li et al.
1999; Lundgren et al. 1992; Rehncrona et
al. 1981; Smith et al. 1986). This study and
others would suggest a beneficial impact but do so with the use of in
vitro models (Grigg and Anderson 1989; Roberts
1993; Roberts and Sick 1992; Schurr et
al. 1987; Seo et al. 1999; Wang et al.
2000; Zhu and Krnjevi 1999). The key
difference may be the clearance of substances such as extracellular
potassium ions, lactate, and free radicals formed during ischemic
insults (Croning and Haddad 1998; Li and
Siesjö 1997), by continuous superfusion with fresh ACSF
employed in the in vitro models, which may not be cleared in the in
vivo situation. Furthermore it has been suggested that it is the excess
production and accumulation of these substances, which are indeed the
primary mechanism by which elevated glucose exposure confers its
detrimental effect. Thus this raises the question of the significance
and relevance of our, and others', in vitro results.
First, the in vitro results presented in this study elucidate features
of the pathophysiology of ischemic injury. For example, the importance
of the occurrence of the ID and, in turn, the dependence of the latency
of ID on immediately prior glucose exposure is highlighted by these
experiments. Furthermore, these in vitro results add further evidence
to the purported notion that elevated glucose exposure, while expected
to protect against ischemia, is detrimental in vivo due to the
secondary effects of excess accumulation of toxic products
(Croning and Haddad 1998; Li and Siesjö
1997) .
While it is generally acknowledged that elevated glucose exposure
should be avoid in situations of brain ischemia, the evidence has yet
to be widely conclusive (Wass and Lanier 1996).
Importantly, brain ischemia occurs in multiple ways, each with
differing characteristics. Most intriguing is the situation where
manipulation of glucose exposure prior to predictable ischemia remains
possible. Coronary bypass surgery, a common surgical operation in North
America, often involves the use of cardio-pulmonary bypass.
Neurological deficits are common and large numbers micro-emboli have
been documented and implicated (Baker et al. 1995;
Brown et al. 2000; Taylor 1998; Taylor et al. 1999). The degree of elevated glucose
exposure, which is common physiologic response to cardio-pulmonary
bypass, has not been conclusively correlated to the degree of
neurologic injury. The hypothesis advanced by the findings of our
study, that elevated glucose exposure would actually be beneficial in brain ischemia were it not for the accumulation of toxic byproducts, may indeed be relevant in ischemia induced by micro-emboli rather that
large territory strokes. We conjecture that the micro-environment may
at least contribute to buffering or handling the toxic products of
ischemia and accommodate the potentially beneficial effects of elevated
glucose exposure similar to in vitro studies presented here. This
conjecture has yet to be tested.
Role of the occurrence of ischemic depolarization in ischemia-induced synaptic transmission damage
When slices were superfused with ACSF containing 4 mM glucose
before ischemia, ID always occurred during a 4-min ischemic insults.
However, ID never occurred during the 4- and even 5-min ischemic
challenges when the slices were superfused with ACSF containing 10 and
20 mM glucose before ischemia, respectively. In slices superfused with
ACSF containing 4 mM glucose before and after ischemia, the CA1
synaptic transmission was irreversibly impaired by the 4-min ischemic
insult, but the CA1 synaptic transmission could endure the 4- and even
5-min ischemic challenges when the slices were superfused with ACSF
containing 10 and 20 mM glucose before and after ischemia,
respectively. Therefore the occurrence of ID may play a critical role
in determining the ischemia-induced damage to synaptic transmission
(Mayevsky 1990; Obeidat and Andrew 1998;
Watson and Lanthorn 1995).
The latencies of ID occurrence with 10 mM glucose in ACSF before
ischemia were significantly longer than those with 4 mM glucose in ACSF
before ischemia. The latencies of ID occurrence with 20 mM glucose in
ACSF before ischemia were significantly longer than those with 10 mM
glucose in ACSF before ischemia. These results demonstrate that
elevated glucose concentrations in ACSF before ischemia postpone the
occurrence of ID and are consistent with those showing that
hyperglycemia delays the occurrence of ID for in vivo global ischemia
(de Crespigny et al. 1999; Els et al. 1997). The ID may be primarily the result of inhibition of Na, K-ATPase activities in hippocampal neurons (Tanaka et al.
1997). The postponement of ID occurrence of high glucose before
ischemia might be due to delaying the inactivation of Na, K-ATPase by
increasing energy stores in the slices (Siemkowicz and Hansen
1978), thereby postponing the energy depletion during ischemia
(Folbergrová et al. 1992; Vannucci et al.
1996; Wagner and Lanier 1994). Glutamate accumulation in the interstitial space resulted from the reverse operation of neuronal glutamate transporters may accelerate the generation of ID (Jabaudon et al. 2000; Madl and
Burgesser 1993; Rossi et al. 2000;
Szatkowski and Attwell 1994; Tanaka et al. 1997). The postponement of ID occurrence by high glucose before ischemia might be also due to lowering the glutamate accumulation during ischemia (Choi et al. 1994; Guyot et al.
2000; Kanthan et al. 1996; Phillis et al.
1999; Seo et al. 1999). An excessive Ca2+ concentration increase caused by the
Ca2+ influx via glutamate receptors and the
Ca2+ release from store sites is associated with
the ID generation. The postponement of ID occurrence of high glucose
before ischemia might be due to decreasing Ca2+
release from the mitochondria and lowering intracellular
Ca2+ concentration as well (Grøndahl et
al. 1998; Lobner and Lipton 1993;
Nabetani et al. 1997; Zhang and Lipton
1999).
Although elevated glucose concentrations in ACSF before ischemia
postponed the occurrence of ID, elevated glucose concentrations in ACSF
before and after ischemia did not improve PS recoveries when ischemia
was terminated at the occurrence of ID. These results indicate that
high glucose in ACSF before and after ischemia does not provide further
protective effects against ischemia-induced damage to synaptic
transmission when ischemia was terminated at the occurrence of ID and
are consistent with those suggesting that membrane function can be
damaged if ischemia is terminated after the occurrence of ID
(Fung et al. 1999; Tanaka et al. 1997, 1999; Uchikado et al. 2000).
The amplitudes of dendritic ID were significantly larger than the
amplitudes of somatic ID, consistent with the finding that swelling is more pronounced in neuronal dendrites than in neuronal somata (Obeidat and Andrew 1998). Although the
differences of the somatic and dendritic ID amplitudes were
statistically significance, the latencies of the somatic and dendritic
ID occurrence and the recoveries of synaptic transmission were not
statistically different with the same concentration glucose in ACSF. In
addition, synaptic transmission could only partially recover following
ischemia terminated at the occurrence of ID, regardless of the glucose
concentration in ACSF before and after ischemia. Therefore the ID
occurrence and not the amplitude of ID was the determinant of the
ischemia-induced synaptic transmission damage.
Therefore these results demonstrate that high glucose in ACSF before and after ischemia protects synaptic transmission against ischemia-induced damage by postponing the occurrence of ID in rat hippocampal slices.
Protective effect of high osmolality against ischemia-induced synaptic transmission damage
Because a hyperosmolality environment improves functional recovery
from anoxic insults (Ballyk et al. 1991; Huang et
al. 1996; Payne et al. 1996; Schurr et
al. 1987), it may have contributed to the protective effects
against ischemia-induced damage to synaptic transmission in our
experiments. However, the present study indicates that the protective
effects of higher glucose in ACSF before and after ischemia against
ischemia-induced synaptic transmission damage are not due to the higher
osmolality of the ACSF and are consistent with those showing that
mannitol treatment does not provide protective effects against focal
cerebral ischemia induced by permanent middle cerebral artery occlusion
in rabbits (Öktem et al. 2000).
In summary, with 4 mM glucose in ACSF before and after ischemia, CA1 synaptic transmission was only able to endure 3 min of ischemic challenge and was irreversibly impaired by 4 min of ischemic insult. However, with 10 mM glucose in ACSF before and after ischemia, CA1 synaptic transmission was able to resist the 4-min ischemic challenges and with 20 mM glucose in ACSF before and after ischemia even the 5-min ischemic challenges did not impair CA1 synaptic transmission. The latencies of ID occurrence with 10 mM glucose in ACSF before ischemia were significantly longer than those with 4 mM glucose in ACSF before ischemia, and the latencies of ID occurrence with 20 mM glucose in ACSF before ischemia were significantly longer than those with 10 mM glucose in ACSF before ischemia. Nevertheless, elevated glucose concentrations did not provide any further protective effects for CA1 synaptic transmission when ischemia was terminated at the occurrence of ID. Therefore the occurrence of ID plays a critical role in determining the ischemia-induced damage to the synaptic transmission. These results indicate that elevated glucose concentrations in ACSF before and after ischemia enhance the resistance of synaptic transmission to ischemia-induced damage and postpone the occurrence of ID.
We conclude that high glucose in ACSF during the period before and after ischemia significantly protects CA1 synaptic transmission against in vitro ischemia-induced damage through postponing the occurrence of ID.
| |
ACKNOWLEDGMENTS |
|---|
The authors thank Drs. J. Duffin and L. Zhang for critical comments on manuscript and Dr. M. Zhao for excellent assistance.
This work was supported by the Ontario Neurotrauma Foundation (ONAO-00180).
| |
FOOTNOTES |
|---|
Address for reprint requests: G.-F. Tian, St. Michael's Hospital, Room 7080, Bond Wing, 30 Bond St., Toronto, Ontario, M5B 1W8 Canada (E-mail: tiang{at}smh.toronto.on.ca).
Received 12 July 2001; accepted in final form 7 March 2002.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
K.
Persistent block of CA1 synaptic function by prolonged hypoxia.
Neuroscience
90:
759-770, 1999[Web of Science][Medline].
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
