Sage Journals: Discover world-class research

Abstract

The acute and delayed effects of anoxia on synaptic transmission and long-term potentiation (LTP) were examined in the CA1 region of rat hippocampal slices. Oxygen deprivation for 20 minutes completely but reversibly depressed excitatory postsynaptic potentials mediated by both N-methyl-d-aspartate receptors (NMDAR) and non-NMDAR. Although LTP was reliably produced by a single tetanus delivered 30 minutes after reoxygenation, LTP could not be induced when a tetanus was delivered 70 to 100 minutes after reoxygenation. A tetanus delivered 100 minutes after reoxygenation produced lasting synaptic enhancement when 100 μmol/L d,l-amino-phosphonovaleric acid (APV), a competitive NMDAR antagonist, was administered during the period of oxygen deprivation. The delayed effects of oxygen deprivation were not blocked when APV was administered after oxygen deprivation. Similarly, the delayed effects on LTP induction were overcome by inhibitors of nitric oxide synthase when the nitric oxide synthase inhibitors were administered during anoxia, but not when administered after oxygen deprivation. These results suggest that untimely activation of NMDAR and nitric oxide release during anoxia produce delayed inhibition of LTP induction and may be involved in the memory defects that occur subsequent to cerebral hypoxia.

Keywords

Anoxia Hypoxia Long-term potentiation Nitric oxide Hippocampal slices

Excessive or sustained activation of N-methyl-d-aspartate receptors (NMDAR) produces a delayed form of neuronal injury (Rothman, 1983) that is believed to participate in the neurodegeneration accompanying several neurologic disorders, including stroke (Rothman and Olney, 1995). During the development of NMDAR-mediated neuronal damage, nitric oxide synthase (NOS) appears to be activated and nitric oxide (NO) is released (Dawson et al., 1991). In agreement with this, we have observed that NOS inhibitors attenuate delayed NMDA-mediated neuronal damage in rat hippocampal slices (Izumi et al., 1992a). Furthermore, evidence suggests the participation of NO in several forms of delayed neuronal damage, including ischemic and hypoxic brain injury (Akira et al., 1994; Hamada et al., 1994; Izumi et al., 1996; Maiese et al., 1994; Wallis et al., 1992; Zhu and Erdemli, 1995).

The NMDAR are involved not only in delayed neuronal toxicity, but also in the induction of long-term potentiation (LTP) in the CA1 hippocampal region (Collingridge et al., 1983). However, untimely activation of NMDAR either by perfusion of low-magnesium solutions (Coan et al., 1989) or weak tetanic stimuli (Huang et al., 1992) diminishes the ability of a strong tetanus to evoke lasting synaptic enhancement. Additionally, low concentrations of NMDA or agonists at NMDAR inhibit LTP (Izumi et al., 1992b). The inhibitory effects of NMDA on LTP are blocked by NOS inhibitors and hemoglobin (Izumi et al., 1992c), suggesting that NO released by low-level NMDAR activation plays a key role in the inhibition (Zorumski and Izumi 1993). Consistent with this, administration of NO releasers inhibits LTP induction when a tetanus is delivered more than 1 hour after drug washout (Izumi and Zorumski, 1993).

These observations suggest that NMDAR-mediated neurotoxicity and NMDAR-mediated LTP inhibition share common features, and that there may be pathologic conditions in which activation of NMDAR impairs neuronal function in the absence of neurodegeneration. During hypoxia or ischemia, endogenous excitatory amino acids (EAA) are released extracellularly (Hegstad et al., 1996; Lobner and Lipton, 1993; Rothman, 1983). Ultimately, these EAA participate in a cascade of events leading to neuronal death. However, brief bouts of hypoxia may produce changes in EAA levels or EAA receptor function that could alter mechanisms involved in synaptic plasticity, either before or in the absence of substantial neuronal damage. Consistent with this view, brief episodes of brain hypoxia and ischemia have been reported to impair learning and memory in the absence of significant neuronal loss (Ando et al., 1987). This raises the possibility that the increases in extracellular EAA interfere with processes required for memory formation, perhaps through altered NMDAR function. Because LTP is a cellular mechanism believed to be involved in certain types of memory, we examine the effects of oxygen deprivation on synaptic transmission and LTP generation in the CA1 region of rat hippocampal slices.

MATERIALS AND METHODS

Slices from 76 male albino rats (postnatal day 30 ± 2, body weight 99.5 ± 22.4 g) were prepared from the septal half of the hippocampus using standard techniques (Zorumski et al., 1996). Rats were anesthetized with halothane and decapitated. The hippocampus was quickly dissected into gassed (95% oxygen–5% carbon dioxide) artificial cerebrospinal fluid (ACSF) containing (in millimolars): 124 NaCl, 5 KCl, 2 CaCl₂, 22 NaHCO₃, 1.25 NaH₂PO₄, 2 MgSO₄, and 10 glucose at a cold temperature (4° to 6°C). Transverse slices (500-μm thick) were cut with a Campden vibratome (Campden Instruments Ltd, Sileby, Loughborough, U.K.). After recovery for at least 1 hour at 30°C in an incubation chamber, each slice was submerged in a constant flow (2 mL/min) recording chamber and maintained at 30°C during the course of the experiment. During periods of oxygen deprivation, the bath solution was changed to one of the same composition as the control ACSF but gassed with 95% nitrogen and 5% carbon dioxide.

Extracellular recordings were obtained from the apical dendritic layer of the CA1 region using 5 to 10 mΩ glass electrodes filled with 2 mol/L NaCl. Evoked synaptic responses were elicited with 0.1- to 0.2-millisecond constant current pulses through a bipolar electrode placed in the Schaffer collateral–commissural pathway. Excitatory postsynaptic potentials (EPSP) were monitored by applying single test pulses every minute at an intensity sufficient to evoke 50% to 60% of the maximal EPSP amplitude based on control input–output curves. The EPSP monitoring was interrupted 20 and 90 minutes after reoxygenation, and 20 and 60 minutes after tetanic stimulation to obtain complete input–output curves. The protocol for LTP induction consisted of a single electrical tetanus for 1 second at 100 Hz with the same intensity of stimulus used to monitor EPSP. The degree of synaptic enhancement reported in the text was measured as the change in the EPSP slope at the 50% point on the EPSP input–output curve 60 minutes after tetanic stimulation compared with the value obtained 10 to 20 minutes before the tetanus. Values in the text represent means ± SD. Only a single slice from each hippocampus was used for each group of experiments. The Mann-Whitney U test and Student's t test were used for comparisons between groups.

For histologic assessment, some slices were fixed in 1% paraformaldehyde and 1.5% glutaraldehyde overnight, embedded in araldite, cut into 1-μm sections, and stained with methylene blue and azure II (Izumi et al., 1996).

6,7-Dinitroquinoxaline-2,3-dione (DNQX) and 2-amino-5,6-dihydro-6-methyl-4H-1,3-thiazine were purchased from Tocris Cookson (St. Louis, MO, U.S.A.), and 7-nitroindazole (7-NIA) was obtained from Research Biochemicals International, Inc. (Natick, MA, U.S.A.). Other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Nifedipine and DNQX were dissolved in ethanol (20 mmol/L) and NaOH (30 mmol/L), respectively.

RESULTS

The EPSP mediated by non-NMDAR were depressed completely by 20 minutes' substitution of nitrogen for oxygen in ACSF. However, during the 20-minute period of oxygen deprivation, hypoxic injury potentials (Fairchild et al., 1988, Rader and Lanthorn, 1989; Wallis et al., 1992) were not observed and reoxygenation was accompanied by full restoration of EPSP to baseline levels within 20 minutes. After reoxygenation, EPSP remained stable for at least 100 minutes (EPSP changes after 90 minutes of reoxygenation: −0.6 ± 7.8%, N = 5). The failure to observe hypoxic depolarization and the full recovery of EPSP after reoxygenation indicate that the initial synaptic depression did not result from irreversible neuronal damage. Furthermore, 20 minutes of oxygen deprivation did not produce either acute (N = 3) or delayed histologic neurodegeneration 90 minutes after the insult (N = 4) in the CA1 region of hippocampal slices (Fig. 1).

FIG. 1.

Histologic appearance of hippocampal slices during and after oxygen deprivation. The photomicrographs depict the appearance of the CA1 region of hippocampal slices during (A) and 90 minutes after 20 minutes of oxygen deprivation (B). In neither case is significant neuronal damage apparent. Magnification ×360.

To examine the effects of hypoxia on LTP, nitrogen was substituted for oxygen for 20 minutes at various times before or after the delivery of tetanic stimulation. As reported previously (Arai et al. 1990), oxygen deprivation immediately after tetanic stimulation inhibits LTP (Fig. 2; change in EPSP slope 60 minutes after tetanus: +4.0 ± 3.2%, N = 4). However, oxygen deprivation for 20 minutes starting 30 minutes after tetanic stimulation did not alter previously established LTP (Fig. 3; change in EPSP slope before and 60 minutes after oxygen deprivation: +26.1 ± 7.8 %, +26.1 ± 3.9%, respectively, N = 4). When administered 20 minutes before the delivery of tetanic stimulation, oxygen deprivation failed to block LTP induction completely (Fig. 4; changes in EPSP slope 60 minutes after tetanus: +22.9 ± 4.9%, N = 7), although the degree of potentiation was less than control values obtained in slices not exposed to oxygen deprivation (control LTP: +44.9 ± 28.0% change in EPSP slope, N = 8, P < 0.05). In contrast, LTP was inhibited completely when tetanic stimulation was delivered 100 minutes after reoxygenation (Fig. 5; changes in EPSP slope: +5.1 ± 9.8%, N = 5).

FIG. 2.

Long-term potentiation (LTP) inhibition by oxygen deprivation immediately after tetanic stimulation. The left graph shows the time course of change (± SD) in excitatory postsynaptic potential (EPSP) slope in slices treated with oxygen deprivation for 20 minutes (denoted as N₂, filled bar) immediately after tetanic stimulation. Tetanic stimulation (100 Hz × 1 second) was delivered in all figures at time 0 indicated by an arrow (↑). The traces below the graph show representative EPSP obtained before (left) and 20 minutes (middle) and 60 minutes (right) after tetanic stimulation. Numbers on each trace denote the time that representative EPSP were obtained. Dashed traces in all figures are the initial baseline responses for comparison. Calibration bar: 5 milliseconds, 1 mV. The right graph shows the input–output curves for EPSP slopes obtained before (open circles) and 60 minutes after tetanic stimulation (filled circles) from a single slice.

FIG. 3.

Effects of oxygen deprivation on the maintenance of LTP. The left graph shows the time course of EPSP slope change in slices treated with 20 minutes of oxygen deprivation (filled bar) administered 30 minutes after tetanic stimulation (↑). The traces below the graph show representative EPSP obtained before (left) and 20 minutes (middle) and 120 minutes (right) after tetanic stimulation. A dotted trace in the middle depicts EPSP during oxygen deprivation. Numbers on each trace denote the time that EPSP were obtained. Dashed traces show the initial baseline responses for comparison. Calibration bar: 5 milliseconds, 1 mV. The right graph shows the input–output curves for EPSP slopes obtained before (open circles) and 20 minutes (triangles) and 120 minutes after tetanic stimulation (filled circles) from a single slice.

FIG. 4.

Effects of oxygen deprivation before tetanic stimulation on LTP induction. The left graph shows the time course of changes in EPSP slope in slices treated with 20 minutes of oxygen deprivation (filled bar) 30 minutes before tetanic stimulation (↑). The traces below the graph show representative EPSP obtained before (left) and during (second) oxygen deprivation and before (third) and 60 minutes (right) after tetanic stimulation. Numbers on each trace denote the time that EPSP were obtained. Dashed traces show the initial baseline responses for comparison. Calibration bar: 5 milliseconds, 1 mV. The right graph shows the input–output curves for EPSP slopes obtained before oxygen deprivation (open circles) and 60 minutes after tetanic stimulation (filled circles) from a single slice.

FIG. 5.

LTP inhibition by oxygen deprivation applied 100 minutes before tetanic stimulation. The left graph shows the time course of change in EPSP slope in slices treated with 20 minutes of oxygen deprivation (filled bar). Tetanic stimulation (↑) delivered 100 minutes after oxygen deprivation failed to induce LTP. The traces below the graph show representative EPSP obtained before (left) and during (second) oxygen deprivation and before (third) and 60 minutes (right) after tetanic stimulation. Numbers on each trace denote the time that EPSP were obtained. Dashed traces show the initial baseline responses for comparison. Calibration bar: 5 milliseconds, 1 mV. The right graph shows the input–output curves for EPSP slopes obtained before (open circles), 20 minutes after oxygen deprivation (open triangles), 10 minutes before (filled triangles) and 60 minutes after tetanic stimulation (filled circles) from a single slice. Notice that the maximal responses, as well as the threshold, are not altered.

These results suggest that oxygen deprivation alters LTP induction in at least two ways: (1) an acute inhibition occurring during the period immediately after the tetanus, and (2) a delayed inhibition occurring more than 30 minutes after the end of oxygen deprivation. Previously, Arai and others (1990) found that the inhibition of LTP by oxygen deprivation immediately after theta-burst stimulation was blocked by 1,3-dipropyl-8-cyclopentylxanthine (DPCPX), an adenosine A1 receptor antagonist. However, we observed that delayed LTP inhibition was not blocked by DPCPX. In the presence of 200 nmol/L DPCPX, oxygen deprivation for 20 minutes produced only partial depression of EPSP, suggesting that adenosine release during hypoxic conditions contributes to acute synaptic depression (Fig. 6). Despite the preservation of EPSP during oxygen deprivation, EPSP did not recover fully after reoxygenation (EPSP changes 90 minutes after reoxygenation compared with initial values: −35.5 ± 11.3%, N = 4, P < 0.001 compared with control recovery). Furthermore, LTP could not be induced by tetanic stimulation delivered 100 minutes after reoxygenation (Fig. 6; EPSP changes: −2.5 ± 16.8%, N = 4). Thus, it appears that delayed hypoxic LTP inhibition is not mediated by activation of adenosine A1 receptors, and differs from acute hypoxic LTP inhibition (Arai et al., 1990).

Because activation of NMDAR is critical for LTP induction in the CA1 region (Collingridge et al., 1983), we examined whether modulation of NMDAR by oxygen deprivation participates in delayed LTP inhibition. The NMDAR-mediated EPSP (NMDAR-EPSP) were monitored in the presence of 30 μmol/L DNQX, an antagonist of non-NMDAR, and 0.1 mmol/L magnesium. Similar to non–NMDAR-mediated EPSP, NMDAR-EPSP were depressed completely by oxygen deprivation. During reoxygenation, the recovery of NMDAR-EPSP was promptly followed by a transient rebound potentiation lasting 20 minutes or less. Subsequently, NMDAR-EPSP returned to control levels (Fig. 7A; NMDAR-EPSP changes 60 minutes after oxygen deprivation: +4.2 ± 13.8 %, N = 7).

FIG. 6.

The delayed inhibition of LTP by oxygen deprivation is not mediated by activation of adenosine A1 receptors. In the presence of 200 nmol/L DPCPX (1,3-dipropyl-8-cyclopentylxanthine), a potent antagonist of adenosine A1 receptors (open thin bar), oxygen deprivation for 20 minutes (filled bar) only partially depressed EPSP. Despite this preservation of EPSP during oxygen deprivation, the recovered EPSP showed gradual fading after reoxygenation. Furthermore, LTP was not induced by tetanic stimulation delivered 100 minutes after reoxygenation. The traces depictEPSP obtained at the times indicated on each wave. The dashed traces are the initial baseline responses. Calibration bar: 5 milliseconds, 1 mV.

FIG. 7.

Changes in N-methyl-d-aspartate receptor–mediated EPSP (NMDAR-EPSP) during and after oxygen deprivation. (A) The graph shows the time course of change in EPSP slope (± SD) for responses mediated by NMDA receptors (NMDAR-EPSP) during and after 20 minutes of oxygen deprivation (filled bar). The NMDAR-EPSP were evoked during continuous administration of 30 μmol/L DNQX and 0.1 mmol/L magnesium. (B) Comparison of NMDAR-EPSP before and after oxygen deprivation in standard ACSF. The NMDAR-EPSP were evoked during a 20-minute administration of 30 μmol/L DNQX and 0.1 mmol/L magnesium (open bars). Filled circles show rescaled values of NMDAR-EPSP adjusting the initial NMDAR-EPSP before oxygen deprivation to 100%. After full recovery of EPSP mediated by non-NMDA receptors, oxygen was removed from the medium. The NMDAR-EPSP after oxygen deprivation were not significantly changed in comparison with those recorded before the insult.

Because oxygen deprivation in the presence of DNQX and low magnesium may produce changes that differ from hypoxic exposure in control media, a second set of experiments was performed to examine whether oxygen deprivation in the standard control media containing 2 mmol/L magnesium alters NMDAR-EPSP (Fig. 7B). After recording stable non–NMDAR-EPSP in control ACSF, the bath solution was switched to one containing 30 μmol/L DNQX and 0.1 mmol/L magnesium to allow study of isolated NMDAR-EPSP. The bath then was switched back to control solution until non–NMDAR-EPSP recovered to baseline. Slices then were oxygen-deprived for 20 minutes in control ACSF followed by reoxygenation for 20 minutes. The ACSF was changed back to the solution with DNQX and a low magnesium content to reexamine NMDAR-EPSP. In these experiments, no significant change in NMDAR-EPSP was detected (Fig. 7B; changes in non–NMDAR-EPSP: −1.0 ± 6.7%; changes in NMDAR-EPSP: −12.7 ± 10.6%, N = 5). It thus seems unlikely that the delayed inhibition of LTP results from sustained suppression of NMDAR.

The transient increase in NMDAR-EPSP during reoxygenation suggests that excessive or untimely activation of NMDAR might contribute to the delayed inhibition of LTP. Administration of 100 μmol/L d,l-amino-phosphonovaleric acid (APV), a competitive NMDA antagonist, during oxygen deprivation did not significantly alter the time course of EPSP changes during and after oxygen deprivation (Fig. 8). However, tetanic stimulation delivered 100 minutes after the hypoxic exposure reliably produced LTP in these slices (changes in EPSP slope: +41.9 ± 14.6%, N = 5, P < 0.001 compared with anoxic controls). This suggests that activation of NMDAR during the hypoxic insult results in the failure of LTP induction. Because the LTP inhibition produced by untimely NMDAR activation is calcium-dependent (Izumi et al., 1992b), we also examined whether activation of low and high voltage–activated calcium channels contributes to delayed hypoxic LTP inhibition using 50 μmol/L nickel and 20 μmol/L nifedipine. When administered during oxygen deprivation, neither nickel nor nifedipine reversed delayed LTP inhibition (data not shown, changes in EPSP slope 60 minutes after tetanic stimulation: +4.4 ± 6.6% and −9.3 ± 6.1% for 50 μmol/L nickel and 20 μmol/L nifedipine, respectively, N = 4 each).

FIG. 8.

Effects of APV on delayed hypoxic LTP inhibition. The graph shows the time course of change in EPSP (mean ± SD) and effects of d,l-amino-phosphonovaleric acid (APV), a competitive NMDAR antagonist, on delayed LTP inhibition by oxygen deprivation (filled bar). A dose of 100 μmol/L APV was administered during oxygen deprivation (open bar) or during reoxygenation (cross-hatched bar). In either case, APV had no effect on the depression or recovery of EPSP. Tetanic stimulation delivered 100 minutes after oxygen deprivation produced LTP when APV was administered during oxygen deprivation (open circles) but not when APV was administered only during reoxygenation (filled circles). The traces depict EPSP obtained at the times indicated on each wave in slices treated with APV during oxygen deprivation (left rows) and during reoxygenation (right row). Calibration bar: 5 milliseconds, 1 mV.

Since LTP induction was not blocked completely when tetanic stimulation was delivered within 30 minutes after reoxygenation, the inhibitory effect of NMDAR activation appears to require time to mature after oxygen deprivation. To examine the possibility that the harmful effects of NMDAR activation occur after reoxygenation, APV was administered immediately after, rather than during, oxygen deprivation. When APV was administered only during the recovery period, tetanic stimulation induced a short-term potentiation that faded to baseline within 60 minutes (Fig. 8; changes in EPSP slope: +4.3 ± 6.7%, N = 4). The failure to observe LTP when APV was administered only during reoxygenation did not result from failure to remove the drug because control slices exposed to APV for a similar period exhibited robust synaptic enhancement (+43.7 ± 7.6% change in EPSP slope, N = 3). These results indicate that the untimely NMDAR activation responsible for delayed LTP inhibition occurs during the period of hypoxia, although the subsequent biological events appear to occur more slowly.

In previous studies, we found that NOS inhibitors prevent NMDAR-mediated LTP inhibition (Izumi et al., 1992c). To examine the involvement of NO in hypoxic LTP inhibition, 100 μmol/L l-N^G-monomethylarginine (l-NMMA), a competitive and broad-spectrum NOS inhibitor, was administered during or after oxygen deprivation. Unlike APV, l-NMMA did not depress NMDAR-EPSP (changes in EPSP slope: −1.2 ± 22.5 %, N = 6). Administration of l-NMMA during oxygen deprivation did not significantly alter the suppression or recovery of EPSP (changes in EPSP slope 90 minutes after reoxygenation: −5.3 ± 26.2%, N = 5). However, LTP was successfully induced by tetanic stimulation delivered 100 minutes after oxygen deprivation (Fig. 9; changes in EPSP slope: +26.6 ± 13.6%, N = 5, P < 0.05 compared with anoxic controls). The effects of l-NMMA on hypoxic LTP inhibition were reversed by coadministration of 1 mmol/L l-arginine, a substrate for NOS (Fig. 9; changes in EPSP slope: +6.1 ± 7.8%, N = 5). Moreover, 3 μmol/L l-nitroarginine-methyl-ester (NAME), another broad-spectrum NOS inhibitor, overcame the effects of hypoxic exposure when administered during oxygen deprivation (Fig. 10; changes in EPSP slope: +31.7 ± 18.6%, N = 5, P < 0.05 compared with anoxic controls). 7-Nitroindazole (100 μmol/L), a more selective inhibitor of neuronal NOS (Babbedge et al., 1992) that inhibits ischemic neuronal damage (Izumi et al., 1996; Yoshida et al. 1994), overcame the LTP inhibiting effects of oxygen deprivation and allowed LTP induction by tetanic stimulation delivered 70 minutes after reoxygenation (Fig. 11; +31.0 ± 10.1% change in EPSP slope, N = 5, P < 0.05 compared with anoxic controls: +0.1 ± 12.7% change, N = 4). In contrast, 2-amino-5,6-dihydro-6-methyl-4H-1,3-thiazine (100 nmol/L), a potent inhibitor of inducible NOS (iNOS) (IC₅₀ = 3.6 nmol/L, Nakane et al., 1995), did not reverse the effects of oxygen deprivation (changes in EPSP slope: +4.1 ± 3.3%, N = 4, data not shown). These results suggest that the delayed inhibition of LTP produced by oxygen deprivation is most likely mediated by NO release after the activation of NMDAR and neuronal NOS.

FIG. 9.

Effects of l-NMMA on delayed hypoxic LTP inhibition. The delayed inhibition of LTP was overcome by 100 μmol/L l-N^G-monomethylarginine (l-NMMA; open circles) administered during oxygen deprivation. The l-NMMA effects were reversed by 1 mmol/L l-arginine, a natural substrate of NOS, administered simultaneously (filled circles). Traces in the left row depict EPSP in a slice treated with l-NMMA alone. Traces in the right row are EPSP from a slice treated with l-arginine and l-NMMA. Traces were obtained at the times indicated on each wave. Calibration bar: 5 milliseconds, 1 mV.

FIG. 10.

Effects of l-nitroarginine-methyl-ester (NAME) on delayed hypoxic LTP inhibition. The delayed inhibition of LTP was overcome by 3 μmol/L NAME administered during oxygen deprivation. Traces depict EPSP at times indicated on each wave. Calibration bar: 5 milliseconds, 1 mV.

FIG. 11.

Involvement of neuronal nitric oxide synthase in delayed hypoxic LTP inhibition. Oxygen deprivation (filled bar) 70 minutes before tetanic stimulation resulted in the failure of LTP induction (filled circles). This LTP inhibition was reversed by 100 μmol/L 7-nitro-indazole (7-NIA), a specific inhibitor of brain NOS, administered during oxygen deprivation (open circles). Traces on the right depict EPSP from a slice treated with 7-NIA before (upper trace) and 60 minutes after oxygen deprivation (middle), and 60 minutes after tetanic stimulation (lower). The traces on the left show EPSP from a slice not treated with 7-NIA. Dashed traces are control EPSP obtained before oxygen deprivation. Calibration bar: 5 milliseconds, 1 mV.

To determine whether NO is a delayed messenger mediating hypoxic inhibition of LTP, 100 μmol/L l-NMMA was administered after oxygen deprivation. As we found previously (Izumi et al., 1992c), 100 μmol/L l-NMMA by itself did not block LTP induction (changes in EPSP slope: +40.6 ± 38.1% and +38.3 ± 18.9 % when l-NMMA was administered for 10 or 90 minutes before tetanic stimulation, N = 5 and 6, respectively). Also, l-NMMA was not effective in reversing hypoxic LTP inhibition when administered during the period of reoxygenation before tetanic stimulation (Fig. 12; changes in EPSP slope: +2.5 ± 17.9%, N = 4). Likewise, l-NMMA administered 10 minutes before tetanic stimulation did not reverse the hypoxic inhibition of LTP (changes in EPSP slope: −0.9 ± 5.3%, N = 4). Thus, we conclude that the activation of NOS that produces LTP inhibition occurs before reoxygenation.

FIG. 12.

During reoxygenation, l-NMMA fails to overcome delayed hypoxic LTP inhibition. Administration of 100 μmol/L l-NMMA after oxygen deprivation (open bar) failed to alter delayed LTP inhibition (filled circles). This failure of l-NMMA to reverse hypoxic LTP inhibition did not result from inhibition of LTP by l-NMMA because administration of 100 μmol/L l-NMMA for 100 minutes in control slices failed to alter LTP (open circles). Traces depict EPSP obtained at the times indicated on each wave in a control slice treated with l-NMMA for 100 minutes before tetanic stimulation (right traces) and from a slice treated with l-NMMA for 100 minutes during reoxygenation (left traces). Dashed traces are control EPSP. Calibration bar: 5 milliseconds, 1 mV.

The delayed effects of NO on LTP induction were studied further using sodium nitroprusside (SNP), a NO releaser. We have previously shown that administration of 10 μmol/L SNP for 10 minutes blocks LTP induction by tetanic stimulation delivered during SNP administration (Izumi et al., 1992c). In the current study, tetanic stimulation was delivered at two times after 20 minutes' exposure to 10 μmol/L SNP (Fig. 13). Although LTP was successfully induced by tetanic stimulation delivered 30 minutes after washout of SNP (changes in EPSP slope: +35.8 ± 6.0%, N = 4), LTP was induced only in one of five slices when tetanic stimulation was delivered 100 minutes after the washout of SNP (+4.2 ± 18.3%, N = 5, P < 0.05).

FIG. 13.

Delayed inhibition of LTP by 10 μmol/L sodium nitroprusside (SNP). The graph shows the effect of 10 μmol/L SNP administered for 20 minutes beginning 30 minutes (open bar) or 100 minutes (filled bar) before the delivery of tetanic stimulation (arrow). The LTP was successfully induced in slices treated with SNP 30 minutes before tetanic stimulation (open circles), whereas LTP induction was blocked by pretreatment of slices with SNP 100 minutes before tetanic stimulation (filled circles). Traces in the left row depict EPSP obtained 10 minutes before (upper) and after tetanic stimulation (lower) in a slice treated with 10 μmol/L SNP 30 minutes before tetanic stimulation. The traces on the right are EPSP 10 minutes before (upper) and 60 minutes after (lower) tetanic stimulation in a slice treated with 10 μmol/L SNP 100 minutes before tetanic stimulation. Dashed traces are control EPSP obtained before SNP administration. Calibration bar: 5 milliseconds, 1 mV.

DISCUSSION

Mental deterioration and amnesia are important clinical sequelae in survivors of cerebral hypoxia. After severe episodes of hypoxia, neuronal loss clearly contributes to changes in cognitive function. However, milder insults that do not produce significant neuronal damage also can result in memory problems. In some cases, particularly after carbon monoxide exposure, the memory deficits become manifest over time (Grinker et al., 1926, Deckel 1994). The mechanisms underlying the delayed neurologic effects of hypoxia are not clear (Hardy and Thom, 1994). Because LTP is a leading candidate to be a mechanism involved in certain forms of memory, this process may be useful for determining how hypoxia impairs mnemonic function.

To investigate the effects of hypoxia on LTP, we used 20 minutes of oxygen deprivation in rat hippocampal slices. Synaptic responses were depressed completely during the anoxic period but recovered fully after reoxygenation. Whereas 20 minutes oxygen deprivation with glucose reduction produces severe neuronal damage (Izumi et al., 1996), oxygen deprivation alone did not result in histologic deterioration (Fig. 1). Prior studies using rodent hippocampal slices show variable degrees of recovery of synaptic transmission after relatively brief anoxic episodes, ranging from little recovery (Opitz and Reymann 1991; Schurr et al., 1987; Wallis et al., 1992) to complete restoration (Canhao et al., 1994; Chen et al., 1996; Yoneda and Okada, 1989). To produce irreversible loss of population spikes in the CA1 region during 40 minutes of hypoxia, Clark and Rothman (1987) found it necessary to increase the bath temperature from 30° to 34°C and lower the glucose concentration from 10 to 7 mmol/L. Similar adverse effects of low glucose and higher temperature have been described by others (Taylor and Weber, 1993; Zhu and Erdemli, 1995). In our experiments performed at 30°C, it took 10 to 15 minutes for oxygen deprivation to depress EPSP completely. During the period of oxygen deprivation, we did not observe hypoxic depolarization, a predictor of irreversible cell damage after hypoxia (Somjen et al., 1993), and EPSP were restored completely to control levels 20 minutes after reoxygenation. Recovery of synaptic transmission during reoxygenation is reported to be accelerated by NMDAR antagonists and NOS inhibitors (Chen et al., 1996; Opitz and Reymann, 1991; Zhu and Erdemli, 1995). Perhaps because our conditions for hypoxia are mild and perhaps because of the young age of the rats used in the current study, we did not observe significant effects of APV or NOS inhibitors either during oxygen deprivation or during the reoxygenation period.

Although 20 minutes of oxygen deprivation did not reverse previously established LTP, hypoxia impaired LTP induction in two ways. First, LTP was inhibited when tetanic stimulation was followed immediately by 20 minutes of oxygen deprivation (Fig. 2). Second, LTP could not be induced when tetanic stimulation was delivered 70 to 100 minutes after oxygen deprivation (Fig. 5). Because LTP can be induced by tetanic stimulation delivered 30 minutes after hypoxic treatment (Fig. 4), it appears that the mechanisms underlying the delayed LTP inhibition differ from those involved in the immediate posttetanic inhibition of LTP. Indeed, delayed LTP inhibition was not altered by DPCPX, an adenosine receptor antagonist that blocks the effects of oxygen deprivation immediately after theta-burst stimulation (Arai et al., 1990). During oxygen deprivation, DPCPX did reduce the depression of EPSP (Canhao et al., 1994). However, the subsequent recovery of synaptic responses was only partial, suggesting that preservation of synaptic function during oxygen deprivation does not necessarily result in better recovery after the insult. As suggested by Boissard and coworkers (1992), suppression of synaptic transmission by adenosine may promote recovery of neuronal function after oxygen deprivation.

We initially suspected that delayed hypoxic LTP inhibition may result from persistent depression of NMDAR after oxygen deprivation. Prior studies report that NMDAR-EPSP are more sensitive to hypoxic insults and recover more slowly after reoxygenation than non–NMDAR-EPSP (Canhao et al., 1994). However, we found that NMDAR-EPSP recovered completely after 20 minutes of hypoxia and exhibited a transient facilitation during the recovery period. This makes it unlikely that delayed hypoxic LTP inhibition results from a decrement in NMDAR function. We next considered whether excessive or untimely activation of NMDAR during or after hypoxic insults inhibits LTP induction, since prior studies have shown that excessive activation of NMDAR results in a failure in LTP production. For example, tetanic stimulation delivered in the presence of low magnesium fails to induce LTP (Coan et al., 1989; Izumi et al., 1992c), and perfusion of low concentrations of NMDA or agonists at NMDAR blocks LTP induction (Izumi et al., 1992b). We previously reported that the inhibitory effects of NMDA are mediated by NO release (Izumi et al., 1992c). Consistent with this, we found that delayed hypoxic LTP inhibition is blocked by APV and NOS inhibitors.

Since NMDAR-EPSP are transiently facilitated after hypoxic exposure (Gozlan et al., 1994), tetanic stimulation after oxygen deprivation could result in excessive activation of NMDAR and NO release. In the current study, however, we failed to observe persistent anoxia-mediated changes in NMDAR-EPSP with our 20-minute hypoxic paradigm. Although the use of APV during reoxygenation was limited by its known inhibitory effects on LTP induction, we observed that administration of APV during, but not after, oxygen deprivation prevented delayed hypoxic inhibition of LTP. This suggests that the activation of NMDAR during oxygen deprivation is critical for the delayed effects on LTP induction.

Several lines of evidence suggest that activation of NMDAR is involved in the neurologic sequelae that follow anoxic insults (Patel et al., 1992; Rothman et al., 1987; Schurr et al., 1995; Zhu and Erdemli, 1995). Furthermore, the cerebral edema due to carbon monoxide poisoning is ameliorated by ketamine, an antagonist of NMDAR (Penney and Chen, 1996). The activation of NMDAR during carbon monoxide poisoning is followed by excessive NOS activity (Ischiropoulos et al., 1996), which may result in neuronal damage. Additionally, l-NMMA, a NOS inhibitor, improves recovery of neuronal function after anoxic insults (Wallis et al., 1992). Taken together, these studies support the idea that a cascade involving NMDAR and NOS is involved in the delayed effects of oxygen deprivation.

Although NO has been proposed to be a retrograde messenger for LTP induction (Schuman and Madison, 1993), conflicting results have been reported by investigators when experiments are conducted at temperatures above room temperature (Williams et al., 1993; Bannerman et al. 1994). Using slices kept at 30°C, we previously found that l-NMMA and l-nitroarginine did not inhibit LTP induction (Izumi et al., 1992b). In the current study, administration of 100 μmol/L l-NMMA during reoxygenation failed to overcome the inhibitory effects of oxygen deprivation on LTP induction, although l-NMMA administered during oxygen deprivation blocked the delayed inhibition of LTP. Because l-NMMA does not inhibit NMDAR (Schuman and Madison 1993, current study), it seems unlikely that an effect on NMDAR contributes to the mechanism by which l-NMMA prevents delayed LTP inhibition. Rather, our results suggest that the release of NO triggered by the activation of NMDAR during hypoxia leads to the delayed inhibition of LTP induction. Furthermore, the reversal of LTP inhibition by 7-NIA suggests that neuronal NOS is a key contributor. Although iNOS may contribute to CNS damage after ischemia, significant iNOS activity is not detected until 2 days after ischemia (Yoshida et al., 1994). Furthermore, NOS activity peaks between 15 minutes and 2 hours after ischemia and is sensitive to 7-NIA (Strosznajder and Chalimoniuk, 1996). Coupled with the fact that an iNOS inhibitor failed to reverse hypoxic LTP inhibition, these observations make it unlikely that iNOS contributes significantly to the effects reported here.

Just as the delayed hypoxic inhibition of LTP is distinct from acute hypoxic LTP inhibition, the delayed inhibition of LTP by SNP (Fig. 9) should be differentiated from its acute effects (Izumi et al., 1992c). We previously observed that long-term depression of EPSP by 100 μmol/L SNP was followed by a failure of LTP induction (Izumi and Zorumski, 1993). Importantly, 100 μmol/L active SNP does not depress NMDAR-EPSP (Izumi et al., 1993), suggesting that the delayed inhibition of LTP by 10 μmol/L SNP does not result from inactivation of NMDAR. Furthermore, NO appears to hamper LTP induction slowly after release. Thus, NOS inhibitors may have variable effects when administered after an episode of acute cerebral hypoxia, depending on the timing of NOS activation. This is of practical importance because therapeutic intervention is possible only after the insult in most cases of hypoxia–ischemia. Although NMDAR antagonists appear to be of limited value when administered after ischemic insults (Nellgård et al., 1991; Dezsi et al., 1994), a neuronal NOS inhibitor has been shown to reduce infarct volume after transient ischemia if injected after reperfusion (Zhang et al., 1996). Additionally, administration of NAME after reperfusion diminishes memory deficits after transient ischemia (Ohno et al., 1994). In the current study, however, a NOS inhibitor administered after hypoxic exposure failed to restore LTP induction, suggesting that delayed administration of NOS inhibitors may not be effective in preventing the memory impairment that follows brief or milder bouts of hypoxia. Nevertheless, better understanding of the biological targets of NO may lead to therapeutic strategies that are effective in the posthypoxic period.

Acknowledgment: The authors thank Jianxin Que for technical assistance.

Footnotes

Abbreviations used

References

Akira

Henry

Baldwin

Wasterlain

(1994) Nitric oxide participates in excitotoxic mechanisms induced by chemical hypoxia. Brain Res 645:285–290

Ando

Kametani

Osada

Iwamoto

Kimura

(1987) Delayed memory dysfunction by transient hypoxia and its prevention with forskolin. Brain Res 405:371–374

Arai

Larson

Lynch

(1990) Anoxia reveals a vulnerable period in the development of long-term potentiation. Brain Res 511:353–357

Babbedge

Bland-Ward

Hart

Moore

(1992) Inhibition of rat cerebellar nitric oxide synthase by 7-nitro indazole and related substituted indazoles. Br J Pharmacol 110:225–228

Bannerman

Chapman

Kelly

PAT

Butcher

Morris

RGM

(1994) Inhibition of nitric oxide synthase does not prevent the induction of long-term potentiation in vivo. J Neurosci 14:7415–7425

Boissard

Lindner

Gribkoff

(1992) Hypoxia produces cell death in the rat hippocampus in the presence of an A1 adenosine receptor antagonist: an anatomical and behavioral study. Neuroscience 48:807–812

Canhao

de Mendonȱa

Ribeiro

(1994) 1,3-Dipropyl-8-cyclopentylxanthine attenuates the NMDA response to hypoxia in the rat hippocampus. Brain Res 661:265–273

Chen

Z-F

Schottler

Arlinghaus

Kassell

Lee

(1996) Hypoxic neuronal damage in the absence of hypoxic depolarization in rat hippocampal slices: the role of glutamate receptors. Brain Res 708:82–92

Clark

Rothman

(1987) Blockade of excitatory amino acid receptors protects anoxic hippocampal slices. Neuroscience 21:665–671

10.

Coan

Irving

Collingridge

(1989) Low-frequency activation of the NMDA receptor system can prevent the induction of LTP. Neurosci Lett 105:205–210

11.

Collingridge

Kehl

McLennan

(1983) Excitatory amino acids in synaptic transmission in the Schaffer collateral–commissural pathway of the rat hippocampus. J Physiol (Lond) 334:33–46

12.

Dawson

London

Bredt

Snyder

(1991) Nitric oxide mediates glutamate neurotoxicity in primary cultures. Proc Natl Acad Sci USA 88:6368–6371

13.

Deckel

(1994) Carbon monoxide poisoning and frontal lobe pathology: two case reports and a discussion of the literature. Brain Inj 8:345–356

14.

Dezsi

Greenberg

Sladky

Araki

Hamar

Reivich

(1994) Prolonged effects of MK-801 in the cat during focal cerebral ischemia and recovery: survival, EEG activity and histopathology. J Neurol Sci 121:110–120

15.

Fairchild

Parsons

Wasterlain

Rinaldi

Wallis

(1988) A hypoxic injury potential in the hippocampal slices. Neurosci Lett 453:357–361

16.

Gozlan

Diabira

Chinestra

Ben-Ari

(1994) Anoxic LTP is mediated by the redox modulatory site of the NMDA receptor. J Neurophysiol 72:3017–3022

17.

Grinker

(1926) Parkinsonism following carbon monoxide poisoning. J Nerv Ment Dis 64:18–29

18.

Hamada

Hayakawa

Hattori

Mikawa

(1994) Inhibitor of nitric oxide synthesis reduces hypoxic–ischemic brain damage in the neonatal rat. Pediatr Res 35:10–14

19.

Hardy

Thom

(1994) Pathophysiology and treatment of carbon monoxide poisoning. J Toxicol Clin Toxicol 32:613–629

20.

Hegstad

Berg-Johnsen

Haugstad

Hauglie-Hanssen

Langmoen

(1996) Amino acid release from human cerebral cortex during simulated ischaemia in vitro. Acta Neurochir 138:234–241

21.

Hershkowitz

Katchman

Veregge

(1993) Site of synaptic depression during hypoxia: a patch-clamp analysis. J Neurophysiol 69:432–441

22.

Huang

Y-Y

Colino

Selig

Malenka

(1992) The influence of prior synaptic activity of the induction of long-term potentiation. Science 255:730–733

23.

Ischiropoulos

Beers

Ohnishi

Fisher

Gamer

Thom

(1996) Nitric oxide production and perivascular nitration in brain after carbon monoxide poisoning in the rat J Clin Invest 97:2260–2267

24.

Izumi

Benz

Clifford

Zorumski

(1992a) Nitric oxide inhibitors attenuate N-methyl-d-aspartate excitotoxicity in rat hippocampal slices. Neurosci Lett 135:227–230

25.

Izumi

Benz

Clifford

Zorumski

(1993) Neurotoxic effects of sodium nitroprusside in rat hippocampal slices. Exp Neurol 121:14–23

26.

Izumi

Benz

Clifford

Zorumski

(1996) Nitric oxide inhibitors attenuate ischemic degeneration in the CA1 rat hippocampal slices. Neurosci Lett 210:157–160

27.

Izumi

Clifford

Zorumski

(1992b) Low concentrations of N-methyl-d-aspartate inhibit the induction of long-term potentiation in rat hippocampal slices. Neurosci Lett 137:245–248

28.

Izumi

Clifford

Zorumski

(1992c) Inhibition of long-term potentiation by NMDA-mediated nitric oxide release. Science 257:1273–1276

29.

Izumi

Zorumski

(1993) Nitric oxide and long-term synaptic depression in the rat hippocampus. Neuroreport 4:1131–1134

30.

Lobner

Lipton

(1993) Intracellular calcium levels and calcium fluxes in the CA1 region of the rat hippocampal slice during in vitro ischemia: relationship to electrophysiological cell damage. J Neurosci 13:4861–4871

31.

Maiese

Wagner

Boccone

(1994) Nitric oxide: a downstream mediator of calcium toxicity in the ischemic cascade. Neurosci Lett 166:43–47

32.

Nakane

Klinghofer

Kuk

Donnelly

Budzik

Pollock

Basha

Carter

(1995) Novel potent and selective inhibitors of inducible nitric oxide synthase. Mol Pharmacol 47:831–834

33.

Nellgård

Wieloch

(1992) Postischemic blockade of AMPA but not NMDA receptors mitigates neuronal damage in the rat brain following transient severe cerebral ischemia. J Cereb Blood Flow Metab 12:2–11

34.

Ohno

Yamamoto

Watanabe

(1994) Intrahippocampal administration of the NO synthase inhibitor l-NAME prevents working memory deficits in rats exposed to transient cerebral ischemia. Brain Res 634:173–177

35.

Opitz

Reymann

(1991) Blockade of metabotropic glutamate receptors protects rat CA1 neurons from hypoxic injury. Neuroreport 2:455–457

36.

Patel

Yim

GKW

Isom

(1992) Blockade of N-methyl-d-aspartate receptors prevents cyanide-induced neuronal injury in primary hippocampal culture. Toxicol Appl Pharmacol 115:124–129

37.

Penney

Chen

(1996) NMDA receptor-blocker ketamine protects during acute carbon monoxide poisoning, while calcium channel-blocker verapamil does not. J Appl Toxicol 16:297–304

38.

Rader

Lanthorn

(1989) Experimental ischemia induces a persistent depolarization blocked by decreased calcium and NMDA antagonists. Neurosci Lett 99:125–130

39.

Rothman

(1983) Synaptic activity mediates death of hypoxic neurons. Science 220:536–537

40.

Rothman

Olney

(1995) Excitotoxicity and the NMDA receptor: still lethal after eight years. Trends Neurosci 18:57–58

41.

Rothman

Thurston

Hauhart

Clark

Solomon

(1987) Ketamine protects hippocampal neurons from anoxia in vitro. Neuroscience 21:673–678

42.

Schuman

Madison

(1993) Nitric oxide as an intercellular signal in long-term potentiation. Semin Neurosci 5:207–215

43.

Schurr

Payne

Rigor

(1995) Protection by MK-801 against hypoxia-, excitotoxin-, and depolarization-induced neuronal damage in vitro. Neurochem Int 26:519–525

44.

Schurr

West

Reid

Tseng

Reiss

Rigor

(1987) Increased glucose improves recovery of neuronal function after cerebral hypoxia in vitro. Brain Res 421:135–139

45.

Somjen

Aitken

Czeh

Jing

Young

(1993) Cellular physiology of hypoxia of the mammalian central system. Res Publ Assoc Res Nerv Ment Dis 71:51–65

46.

Strosznajder

Chalimoniuk

(1996) Biphasic enhancement of nitric oxide synthase activity and cGMP level following brain ischemia in gerbils. Acta Neurobiol Exp 56:71–81

47.

Taylor

Weber

(1993) Effect of temperature on synaptic function after reduced oxygen and glucose in hippocampal slices. Neuroscience 52:555–562

48.

Wallis

Panizzon

Wasterlain

(1992) Inhibition of nitric oxide synthase protects against hypoxic neuronal injury. Neuroreport 3:645–648

49.

Williams

Errington

Y-G

Lynch

Bliss

TVP

(1993) The search for retrograde messengers in long-term potentiation. Semin Neurosci 5:149–158

50.

Yoneda

Okada

(1989) Effects of anoxia and recovery on the neurotransmission and level of high-energy phosphates in thin hippocampal slices from the guinea-pig. Neuroscience 28:401–407

51.

Yoshida

Limmroth

Irikura

Moskowitz

(1994) The NOS inhibitor, 7-nitroindazole, decreases focal infarct volume but not the response to topical acetylcholine in pial vessels. J Cereb Blood Flow Metab 14:924–929

52.

Yoshida

Waeber

Huang

Moskowits

(1995) Induction of nitric oxide synthase activity in rodent brain following middle cerebral artery occlusion. Neurosci Lett 194:214–218.

53.

Zhang

Feif

Macdonald

Tang

Kamp

Gentile

Shkespeare

Murray

Chopp

(1996) ARL 17477, a potent and selective neuronal NOS inhibitor decreases infarct volume after transient middle cerebral artery occlusion in rats. J Cereb Blood Flow Metab 16:599–604

54.

Zhu

Erdemli

(1995) Nitric oxide may be a mediator of effects of prolonged but not brief anoxia in CA1 neurons in slices. Neuropharmacology 34:75–80

55.

Zorumski

Izumi

(1993) Nitric oxide and hippocampal synaptic plasticity. Biochem Pharmacol 46:777–785

56.

Zorumski

Mennerik

Izumi

(1996) Assessment of synaptic effects of nitric oxide in hippocampal neurons. Methods Neurosci 31:283–299

Oxygen Deprivation Produces Delayed Inhibition of Long-Term Potentiation by Activation of NMDA Receptors and Nitric Oxide Synthase

Abstract

Keywords

MATERIALS AND METHODS

RESULTS

DISCUSSION

Footnotes

Abbreviations used

References