Nitric oxide (NO) was first identified as an intercellular messenger molecule in the brain a little more than a decade ago. Since then, its role has been widely recognized as an unconventional neurotransmitter/neuromodulator that is biosynthesized on demand, is quickly diffusible, and is able to activate various signaling pathways in adjoining cells (1, 2).
NO formation occurs via NO synthase, a Ca2+/calmodulin-dependent enzyme that transforms L-arginine into NO and citrulline (3, 4). The synthase can be activated via enhanced glutamatergic transmission, mediated especially by N-methyl-D-aspartate (NMDA) receptors. The released NO subsequently induces soluble guanyl cyclase, thereby increasing intracellular cyclic GMP levels (5–8).
Since there is much evidence that NO can modulate pathological and physiological processes, it can be speculated that NO is also a factor in epilepsy. Although the role of NO in epileptogenesis has been examined in a number of studies (9–15), the results to date are contradictory. Several studies reported experiments involving pretreatment with NOS inhibitors for both chemically- and electrically-induced seizures in animals (10, 14), but the results were inconsistent. Hence the role of NO in seizure events is still unclear. Some studies showed that NOS inhibitors delayed the onset time of seizures (9–11); however, others reported that the severity of seizures was increased by NOS inhibitors (12–14), and another study (15) found that they showed no effect (15). These inconsistencies may be partly the result of different experimental conditions (such as animal species and age). In many of the studies referenced above, NO involvement in seizures was surmised only from behavioral changes in animals that had been preadministered with NOS inhibitors.
The present studies were designed to examine the involvement of NO during convulsions by both behavioral measurements and direct observation of NO by in vivo/ex vivo EPR. The NO levels in the brain of PTZ-doped mice and rats were observed by in vivo and ex vivo EPR spectroscopy during convulsions, and the NO content in various isolated brain tissues was quantitated at the X-band as well.
MATERIALS AND METHODS
Sodium diethyldithiocarbamate (DETC), NG-nitro-L-arginine (L-NNA), NG-monomethyl-L-arginine (L-NMMA), and pentylenetetrazole (PTZ) were from Sigma Chemical Co. Sodium citrate, phenytoin, dimethyl sulfoxide (DMSO), and FeSO4 · 7H2O were obtained from Wako Chemical Co. 3-Bromo-7-nitroindazole was from Dojindo Laboratories. Pentobarbital was from Dainippon Pharmaceutical Co., Ltd., and diazepam was obtained from Yamanouchi Pharmaceutical Co., Ltd.
The animal protocols were institutionally approved according to the National Institute of Health Animal Care and Use protocol. Male Wistar rats (120–150 g, 8 weeks old) and male ICR mice (15–20 g, 4 weeks old) were maintained on laboratory chow and water ad libitum on a 12-hr light/dark cycle.
Administration of PTZ
The animals were placed in a plastic chamber (30 × 30 × 40 cm) and observed before and after PTZ administration. After the rats displayed a resting posture, they were injected intraperitoneally (i.p.) with PTZ (in saline solution) at doses of 10–100 mg/kg. Their behavior was classified and scored as follows: 1 = no convulsion behavior, 2 = head twitching, 3 = clonic convulsion, 4 = kangaroo posture (or violent convulsion), and 5 = falling back (according to the criteria of Ito et al. (16)). Consequently, the definitions used in this study were: clonic convulsion, score 3 behavior; and tonic convulsion, score 4 or 5 behavior. For studies of anticonvulsants, diazepam (5 mg/kg) or phenytoin (60 mg/kg) was injected 30 min before PTZ administration.
Measurement of NO Content by EPR
The rats were injected i.p. with 500 mg/kg DETC (in saline). Subsequently they were injected subcutaneously with iron-citrate complex (50 mg/kg FeSO4 · 7H2O + 250 mg/kg sodium citrate). Both injections were administered 30 min prior to PTZ injection. For the NOS inhibitor experiments, 50–100 mg/kg L-NNA or 50–250 mg/kg L-NMMA, or 1–20 mg/kg 3Br-7NI were injected i.p. 30 min prior to PTZ injection. All solutions were prepared in saline solution, except for 3Br-7NI, which was dissolved in DMSO. After PTZ injection, rats that had also received the NO-trapping agents (DETC and iron-citrate complex), were observed and scored for behavioral changes. The rats were then killed and their brains were immediately excised. The olfactory bulb (OB), hippocampus (HI), cerebellum (CL), cerebral cortex (CR), and hypothalamus (HT) were immediately dissected and kept frozen on dry ice.
L-Band In Vivo EPR
After scoring behavioral changes in the PTZ-doped mice (15–20 g), the animals were anesthetized with pentobarbital (35 mg/kg) and placed in a modified bridged loop-gap resonator in an L-band microwave bridge assembly, as described previously (17). All L-band EPR measurements were completed within 10 min after the mice were anesthetized. Typical spectrometer conditions were: frequency = 1.25 GHz; applied magnetic field = 425 gauss; incident microwave power = 50 mW; 20 KHz modulation amplitude = 1.0 gauss; sweep rate = 50 gauss/min; and response time = 0.3 s.
Ex Vivo EPR
Excised, sectioned brain tissues were weighed and subsequently measured in quartz tissue flat cells at ambient temperature in a JEOL FES-FE X-band EPR spectrometer. All EPR measurements were completed within 10 min after the tissue samples were prepared. The amount of trapped NO was determined from the intensity of the low-field EPR peak and compared with standards. The NO levels were calculated by subtracting the basal NO levels obtained from normal controls from the total NO levels detected in animals with convulsions. Typical instrumental settings were: field scan = 100 gauss; sweep time = 4 min; time constant = 0.3 s; modulation amplitude = 2.0–3.2 gauss at 100 KHz; and microwave power = 20–30 mW.
NO Synthase Assay
NO synthesis, catalyzed by NOS, was measured by the method of Feelisch and Noack (18), in which the oxidation of oxyhemoglobin to methemoglobin by NO is monitored spectrophotometrically. The absorption difference between 401 and 411 nm was measured continuously with a Hitachi dual-wavelength spectrophotometer (model 556). The brain homogenates from both control and convulsive rats were obtained according to published methods (19). The incubations contained 1.5 μM oxyhemoglobin, 1 mM MgCl2, 50 mM phosphate (pH 7.2), and 10–20% (V/V) brain extract. NO synthesis was initiated by addition of 1 mM L-arginine and 1 mM NADPH.
Statistical significance was assessed by one-factor analysis of variance (ANOVA), followed by Dunnett's test. For nonparametric data, the Mann-Whitney U-test was used.
NO Generation in PTZ-Doped Rats
Injection of PTZ evoked dose-dependent clonic or tonic convulsions in all 26 rats used in these experiments. At relatively high doses of PTZ (60 mg/kg), tonic convulsion was induced in 16 of 18 rats. The threshold dose (40 mg/kg PTZ) induced clonic convulsion in eight of eight rats. In the tonic convulsion experiments, the first convulsive twitch appeared about 1.5 min after PTZ administration, and the violent convulsion (score 4) or falling-back (score 5) appeared within 5 min (N = 8).
NO generation was measured in excised brain tissues of PTZ-doped rats showing tonic convulsions (scores 4 and 5) by X-band EPR spectroscopy at ambient temperature. Figure 1 depicts EPR spectra of tissue samples from the CR and CL. The three-line EPR spectra were identical to a control spectrum of (DETC)2-Fe-NO standard (aN = 12.8 gauss, g = 2.04). The EPR signal was suppressed by the preadministration of NOS inhibitors (L-NMMA or L-NNA) 30 min before PTZ injection (spectrum not shown). Hence, the results show that NO is enzymatically generated by NOS in the brain. In control normal rats doped with saline instead of PTZ, a similar EPR spectrum (Fig. 1d), with much less intensity than that of Fig. 1b or c, was detected in the CL. This spectrum was also suppressed with L-NMMA (data not shown), suggesting that NO detected in normal control rats was produced by the constitutive NOS.
NO generation was also measured in the brain tissue of rats undergoing clonic convulsions. The results are as summarized in Fig. 2, together with the tonic convulsion data. Rats that had a clonic convulsion survived for more than 10 min, whereas most of the rats that underwent a tonic convulsion died in less than 10 min. In order to compare NO generation quantitatively in both cases (tonic and clonic convulsions), the levels calculated from (DETC)2-Fe-NO EPR spectra were extrapolated to 10 min (including tonic convulsions). Hence, Fig. 2 shows that more NO would be generated in a tonic vs. a clonic convulsion in all tissues in the brain (OB, HI, CL, CR, and HT). Furthermore, note that in each type of convulsion, NO generation appeared to be constant in all tissues examined, indicating that NO was apparently found uniformly in all tissues, or that locally generated NO diffused rapidly across these tissues.
In Vivo L-Band EPR Spectroscopy on PTZ-Doped Mice
After a tonic convulsion was observed for 3 min and confirmed, mice doped with high (80 mg/kg) doses of PTZ were immediately anesthetized and placed in the L-band EPR system. Distinct (albeit weak) three-line EPR spectra were recorded in the brain (Fig. 3a). Similar EPR spectra were observed in the mice with clonic convulsions, although the spectral intensity was even lower (spectrum b in Fig. 3). In control mice doped with only saline, no signal was observed. Tissues from the CR and CL were removed immediately after in vivo EPR observation, and measured by X-band EPR. A typical EPR spectrum obtained from the CR is shown in Fig. 3c. The (DETC)2-Fe-NO signal was identical to the spectra obtained from rats as shown in Fig. 1b and c. Preadministration of the brain-specific NOS inhibitor, 3Br-7NI, caused the obliteration of the in vivo (DETC)2-Fe-NO EPR signal, as depicted in Fig. 3d for mice showing tonic convulsions.
Effect of NOS Inhibitors on NO Generation and Behavioral Changes in PTZ-Doped Rats
As mentioned above, the effects of NOS inhibitors were examined in rats by ex vivo EPR at X-band. Figure 4a shows NO generation in several brain tissues (the HI, CR, and CL) of PTZ-doped rats undergoing tonic convulsion. L-NNA reduced NO levels to <30%, which was similar to the levels in the clonic convulsions (cf. Fig. 2). Remarkably, 3Br-7NI inhibited NO generation in tonic convulsions significantly, i.e., to more than 98% in all of the brain tissues examined. As to behavioral changes, the results showed that both L-NNA and 3Br-7NI made the convulsive seizures less strong, suppressing the transition to a tonic convulsion (Fig. 4b). The combined results of Fig. 4a and b suggest that at least the onset of the convulsive seizure is not induced directly by NO, i.e., NO does not directly induce clonic seizures in PTZ-doped rats, but NO may be generated as a result of the convulsion.
Effect of Anticonvulsant Pretreatment on NO Generation in PTZ-Treated Rats
The effects of the anticonvulsants diazepam and phenytoin on NO production were examined. Behavioral changes were observed after PTZ administration in rats that had been pretreated 30 min earlier with diazepam (5 mg/kg) or phenytoin (60 mg/kg). At PTZ doses up to 80 mg/kg, no tonic convulsions were observed in a group of four rats. The average behavioral score during each convulsion was <3 in each case. The EPR results show that pretreatment with either anticonvulsant suppressed NO production in the CL and CR of PTZ-doped rats, down to levels found in clonic convulsion (Fig. 5).
As noted in the Introduction, the role of NO in epileptogenesis is unclear. NO has been suggested to be both an anticonvulsant and a proconvulsant in animals (9, 10, 12–16). These inconsistencies may be due to the fact that although these studies measured the effects of NOS inhibitors, they did not actually measure NO production. The advantage of the current L-band EPR study is that we can make direct NO measurements by detecting it as the (DETC)2-Fe-NO complex in the brain of living mice. We were able to confirm that NO levels are elevated in tonic and clonic convulsions.
Both the in vivo and ex vivo EPR measurements confirmed that more NO was produced in the tonic than in the clonic convulsions, perhaps indicating that NO levels were proportional to the intensity of the convulsion. These results suggest that NO works as a proconvulsant. However, in the presence of NOS inhibitors, both L-NNA and 3Br-7NI suppressed the transition from clonic to tonic convulsion at high PTZ doses, while the behavioral change score dropped from 5 to 3. 3Br-7NI completely inhibited NO generation, but L-NNA was effective to only 70%. The results with the anticonvulsants diazepam and phenytoin, which are used to cure epilepsy, were similar to the results with the NOS inhibitors (NO levels were not completely suppressed, although the tonic convulsion was completely attenuated). Overall, these results indicate that NO does not directly induce clonic convulsion in PTZ-doped animals. The different inhibitory action observed with the two NOS inhibitors may reflect differences in both tissue accessibility and specificity. 3Br-7NI can penetrate the blood brain barrier, its inhibitory effect is higher than that of L-NNA, and it acts at a different site in the brain. These considerations support the hypothesis that NO is produced as a consequence of convulsive seizure. One possible mechanism involves GABA-activated channels: PTZ-induced inhibition of GABA-activated channels causes excess glutamate release from presynapses. The voltage-dependent NMDA receptor (NMDAR) binds this excess glutamate, followed by an influx of large amounts of Ca ions into postsynapses that bind to calmodulin and activate nNOS, which is anchored to the NMDAR by the adaptor protein PSD95, followed by the elevation of NO. With high doses of PTZ, this phenomenon is not transient and continues for a relatively long time. On the other hand, in the case of low doses of PTZ, the amount of released glutamate is not enough to activate NMDAR, and NO production is not enhanced in this situation. Therefore, high levels of NO generated by nNOS play an important role in inducing a tonic convulsion. Another possibility could involve NO reacting with regulatory proteins by nitrosylating proteins directly by the reaction with free cysteins. These modified proteins might then induce further membrane polarization in neurons, as manifested by subsequent tonic behavior in PTZ-doped animals. All of the NOS inhibitors examined in this study inhibited tonic convulsion and NO production to a lesser or greater extent. While we favor the hypothesis that NO is generated as a consequence of onset of seizure, further studies may confirm the possibility that the excess NO molecules induce tonic convulsion.
It is important to note that Bashkatova et al. (11) trapped NO produced during convulsions by iron-DETC complex and quantitated NO from EPR spectra of iron-DETC-NO complex at 77 K. Although EPR sensitivity is higher at 77 K than at ambient temperature, the EPR spectra are more complicated due to a superposition of some intrinsic (DETC)2-Cu(II) complexes, which are effectively “EPR silent” at ambient temperature. Room-temperature EPR measurements have the advantage of receiving a simple triplet EPR signal from (DETC)2-Fe-NO, with no background EPR signal from intrinsically derived copper complexes.
Several groups have succeeded in directly detecting NO in septic-shock animals by in vivo and ex vivo EPR spectroscopy (20–22). We previously confirmed NO generation in septic animals in vivo, and quantitated NO in the brain tissues of rats doped with endotoxin (23, 24). The NO content in brain tissue was estimated at 10–20 nmol/g of tissue in the CR and CL (24), which was in the same range as in the tonic convulsions measured in this study. Thus, it is worthwhile to compare the NO levels generated in both septic and seizure mouse brain by in vivo EPR spectra. The detected NO level in brain tissues by ex vivo EPR was the same in both septic and seizure mice, but the signal-to-noise ratio (SNR) of the L-band EPR spectrum obtained from the head region of seizure animals was much lower than that in septic animals, indicating that the total amount of NO was smaller and that the NO-generating site was restricted to the brain in the seizure mice. Future studies will involve EPR imaging experiments in order to visualize the site of NO production.
This work was supported by a Grant-in-Aid for Scientific Research from the J.S.P.S., Japan.