Effects of Diazepam and Ketamine on Pilocarpine-Induced Status Epilepticus in Mice

NEUROSCIENCE RESEARCH ARTICLE S. Wang et al. / Neuroscience 421 (2019) 112–122

Effects of Diazepam and Ketamine on Pilocarpine-Induced Status Epilepticus in Mice Siyan Wang, Maxime Le´vesque and Massimo Avoli * Montreal Neurological Institute and Departments of Neurology & Neurosurgery, and of Physiology, McGill University, 3801 University Street, Montre´al, Qc H3A 2B4, Canada

Abstract—Status epilepticus (SE) is a life-threatening condition needing immediate care to prevent brain damage. SE with electrographic and behavioral features similar to those seen in humans is reproduced in rodents by i.p. pilocarpine injection, and can be terminated by diazepam and ketamine treatment but only behaviourally, not electrographically. Little is known on the behavioral and EEG effects induced by a delayed administration of ketamine (25 mg/kg) after diazepam (10 mg/kg) or vice versa. Therefore, we analysed behavior and EEG activity recorded from the mouse hippocampal CA3 region before, during SE and after anticonvulsant treatments. In the first group (n = 4), diazepam was administered one hour before ketamine whereas in the second group (n = 4) ketamine was administered one hour before diazepam. The EEG SE did not disappear after each of the two treatments but progressed within 4 h to a pattern of interictal discharges. However, diazepam administration before ketamine significantly shortened the time of behavioral recovery compared to when ketamine was administered before diazepam (p < 0.05). The two protocols were also associated to distinct EEG changes in gamma and high frequency oscillations. In conclusion, although diazepam and ketamine are not effective in stopping EEG SE, diazepam administration one hour before ketamine shortens behavioral recovery in pilocarpine-treated mice. Ó 2019 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: status epilepticus, pilocarpine, diazepam, ketamine, CA3 hippocampal region, mesial temporal lobe epilepsy.

2018) or unintentional ingestion of chemoconvulsants. Such an event occurred in Canada, in 1987, when some individuals were poisoned with domoic acid from diatoms in mussels. These patients experienced SE (Perl et al., 1990), that was followed 1 year later by temporal lobe refractory seizures and hippocampal sclerosis (Cendes et al., 1995). The condition of a SE that is followed, but a few days later, by spontaneous refractory seizures is reproduced in animals by employing systemic injection of pilocarpine (Curia et al., 2008) or kainic acid (Le´vesque and Avoli, 2013). In rodents, behavioral convulsions associated to pilocarpine-induced SE are stopped by the systemic administration of diazepam and ketamine (Martin and Kapur, 2008; Bortel et al., 2010; Le´vesque et al., 2011, 2012; Salami et al., 2014; Behr et al., 2017; Niquet et al., 2017). However, we have recently found in mice that although the simultaneous injection of ketamine and diazepam terminated behavioral SE, the electrographic SE persisted for hours and evolved into continuous interictal spike activity (Wang et al., 2019). Therefore, in the present study we aimed at establishing whether the sequence of injection of diazepam and ketamine could exert a significant impact on the suppression of behavioral and electrographic SE. In addition, we analyzed the effect of these drugs on EEG oscillations in the beta and gamma

INTRODUCTION According to the International League Against Epilepsy, Status epilepticus (SE) is defined as a tonic-clonic or focal seizure with impaired consciousness lasting more than 5 min (Trinka et al., 2015). SE is considered as a serious neurological emergency and life-threatening condition if immediate care is not taken to stop it (Chin et al., 2004). Benzodiazepines are usually administered as initial therapy, then followed by valproic acid, levetiracetam or fosphenytoin if the first-line benzodiazepine treatment does not halt SE (Glauser et al., 2016). When first and second-line treatment are unsuccessful, anesthetic agents such as propofol are used (Rai and Drislane, 2018). SE can lead to the development of mesial temporal lobe epilepsy (MTLE), in which refractory focal seizures initiate from the hippocampus, rhinal cortices or amygdala (Engel, 1996). SE can occur following head trauma (Ding et al., 2016), febrile seizures (Leung et al., *Corresponding author. Address: Montreal Neurological Institute, 3801 University Street, Montre´al, H3A 2B4 Qc, Canada. Fax: + 1514-398-8106. E-mail address: [email protected] (M. Avoli). Abbreviations: SE, Status epilepticus; MTLE, mesial temporal lobe epilepsy. https://doi.org/10.1016/j.neuroscience.2019.10.009 0306-4522/Ó 2019 IBRO. Published by Elsevier Ltd. All rights reserved. 112

S. Wang et al. / Neuroscience 421 (2019) 112–122

113

All procedures were performed according to protocols and guidelines of the Canadian Council on Animal Care and were approved by the Animal Care Committee of McGill University. C57BL/6 mice (The Jackson Laboratory) were bred and maintained in-house. Animals were housed under a controlled environment (22 ± 2 °C, 12 h light/dark schedule). Food and water were provided ad libitum.

activity on the EEG without a return to baseline activity (Mu¨ller et al., 2009). Diazepam (10 mg/kg, s.c.; CDMV, Qc, Canada) or ketamine (25 mg/kg, s.c.; CDMV, Qc, Canada) were injected after 60 min to stop SE. Mice were separated into two groups with different injection protocols. The first group was administered diazepam one hour before ketamine (n = 4) (Fig. 1A) and the second group was administered ketamine one hour before diazepam (n = 4) (Fig. 2A). We did not evaluate the effect of a single administration of either diazepam or ketamine on behavior and EEG since it is known that such procedure does not stop SE in rodents (Walton and Treiman, 1988; Niquet et al., 2016). The administration of the first drug (either diazepam or ketamine) was used as the reference point (time = 0) for the analysis of EEG activity after pharmacological intervention.

Electrode implantation

EEG analysis

Eight (n = 8) mice of age 58–96 days were anaesthetized with 3% isoflurane in 100% O2. Mice were placed in a stereotaxic frame. Three holes were drilled in the bone to allow the implantation of one anchor screw (2.4 mm length), a bipolar electrode and a reference. The bipolar electrode (20–35 kO) was implanted into the CA3 subfield of the hippocampus (ML: 3, AP: 2.85, DV: 4). The reference electrode (5–10 kO) was bent at 90° angle to overlie the cortex on the contralateral hemisphere. The screw, bipolar electrode and reference were fixed to the skull using dental acrylic cement. After surgery, for 72 h post-op, chloramphenicol (Erfa, Qc, Canada) and lidocaine (5%; Odan) were applied around the surgery site. Carprofen (20 mg/kg s.c.; Merail, Qc, Canada) and enrofloxacin (5 mg/kg, s.c., CDMV, Qc, Canada) were injected every 24 h while buprenorphine (0.1 mg/kg s.c., CDMV, Qc, Canada) was injected every 8 h. One ml of 0.9% saline was also injected s.c. every 24 h.

EEG activity during SE was classified into stages according to Treiman et al. (1990). Stage 1 was characterised by isolated seizures and interictal spikes, stage 2 by waxing and waning ictal discharges, stage 3 by continuous ictal discharges, stage 4 by continuous ictal discharges with flattening of the EEG and stage 5 with periodic ictal discharges. The end of SE was defined as the absence of continuous spiking activity. Behavioral activity was monitored for a time-period of 24 h following the onset of SE. The delay between the injection of diazepam or ketamine and a return to normal behavior – consisting of walking, grooming, eating and exploratory behavior – was also calculated. Spectral power during SE was analyzed using pwelch function in Matlab 7.11.0 (The Mathworks, MA, USA) and the area under the curve for alpha (3–15 Hz), beta (18– 25 Hz) and gamma (40–80 Hz) frequency bands were calculated, for bins of 30 s. Since stages of SE were different in duration between animals, we converted each stage into a scale from 0 (start) to 100% (end). The manipulation of the animal for the injection of diazepam and ketamine induced artifacts on the EEG that were removed from further analysis. Therefore, for easier comparison with stages 2 and 3 - which were converted into a scale from 0 to 100% - we also converted the EEG recordings after diazepam and ketamine into the same scale. The time-period that corresponded to the last 20% of stage 3 was compared to the first 20% after the injection of the first drug (diazepam or ketamine). We did not compare the effect of the administration of the second drug in order to avoid introducing any confounding results caused by the combination of ketamine and diazepam.

frequency ranges as well as on high-frequency ranges (HFOs, ripples: 80–200 Hz, fast ripples: 250–500 Hz); HFOs are considered as biomarkers of pathological network activity (Jefferys et al., 2012; Le´vesque et al., 2018).

MATERIAL AND METHODS Animals

EEG recordings Mice were then housed in custom-made boxes (30  30  40 cm) and allowed one week to recover from surgery. Electrodes were connected to a multichannel cable and electrical swivel (Commutator SL 18C, HRS Scientific, Qc, Canada). Continuous recordings were performed from one day before to one day after pilocarpine treatment. EEG signals were amplified by an interface kit (Mobile 36ch LTM ProAmp, Stellate, Qc, Canada) and low-pass filtered at 500 Hz with 2000 Hz sampling rate per channel. Data collection was performed using a monitoring software (Harmonie, Stellate, Qc, Canada).

HFO analysis Induction of status epilepticus Mice were first injected with scopolamine methylnitrate (1 mg/kg i.p.; Sigma-Aldrich, Qc, Canada) and 30 min later with pilocarpine (200 mg/kg, i.p.; Sigma-Aldrich, Qc, Canada). Doses of 100 mg/kg of pilocarpine were then administered every 30 min until the occurrence of SE, which was characterized by continuous spiking

Methods used to detect and analyze HFOs were previously published by Le´vesque et al. (2012). Briefly, raw EEG signals were bandpass filtered between 80 and 200 Hz (ripples) and 250–500 Hz (fast ripples). Filtered signals were then normalized to a 10 min reference time-period before the injection of scopolamine. After filtering and normalization, oscillations with 4 consecutive

114

S. Wang et al. / Neuroscience 421 (2019) 112–122

Fig. 1. A: Schematic diagram showing the injection protocol in the first group of mice in which diazepam was administered one hour before ketamine. Multiple injections of pilocarpine (100 mg/kg i.p.) were in some cases needed to induce SE (red arrows). B: Representative EEG recordings from the hippocampus of a pilocarpine-treated mouse before SE (B1), during stage 2 (B2a) and stage 3 SE (B2b), and after the administration of diazepam (B3) and ketamine (B4). Power spectral analyses are also shown. Note that stage 2 SE was characterised by rhythmic bursts punctuated by episodes of flattening on the EEG (B2a) whereas stage 3 was characterised by continuous and constant ictal spiking (B2b). The administration of diazepam induced rhythmic ictal bursts (B3a) that persisted after the administration of ketamine (B4a).

peaks with amplitude higher than 3 SD of the mean were considered as HFOs. Then, the time period was converted to a scale from 0% to 100% for comparison. The number of HFOs in each 1% bin was calculated and averaged across animals. Statistical analysis Non-parametric Mann-Whitney U and Wilcoxon tests were used since values were not normally distributed. The level of significance was set at p < 0.05.

RESULTS EEG patterns during status epilepticus A schematic of the injection protocol for the first group of mice is shown in Fig. 1A. Fig. 1B1 illustrates baseline EEG activity recorded from the CA3 region before the induction of SE. As reported by Treiman et al. (1990), different stages of SE were observed, based on the patterns of EEG activity. However, stage 1, characterised by discrete seizures with slow interictal spiking (Treiman

et al., 1990), was observed in only 37.5% of mice (3/8) in our study. We therefore considered the onset of SE as the onset of stage 2, which was observed in all mice. On average, in both groups, the onset of stage 2 SE occurred 16.2 min (±9.6 min) after the last injection of pilocarpine. Stage 2 was characterised by rhythmic bursts of spikes that were punctuated by episodes of flattening on the EEG (mean duration of stage 2 = 15.8 (±13.3) min) (Fig. 1B2a but see also Fig. 2B2a). Stage 3 also occurred in all animals and was associated with continuous ictal spiking (Fig. 1B2b but see also Fig. 2B2b). Since diazepam or ketamine were administered within one hour after SE onset, we could not observe stages 4 and 5, as reported by Treiman et al. (1990). Fig. 1B3 and B4 show representative EEG recordings after the administration of diazepam, first, and later of ketamine to stop SE (n = 4). The sole injection of diazepam changed the epileptiform activity associated with SE to rhythmic ictal bursts (Fig. 1B3), a pattern that did not appear to be altered by ketamine injection one hour later (Fig. 1B4). The continuous spiking activity associated to SE disappeared 291 (IQR = 219.5–337)

S. Wang et al. / Neuroscience 421 (2019) 112–122

115

Fig. 2. A: Schematic diagram showing the injection protocol in the second group of mice in which ketamine was administered one hour before diazepam. Multiple injections of pilocarpine (100 mg/kg i.p.) were in some cases needed to reach SE (red arrows). B: Representative EEG recordings from a pilocarpine-treated mouse in which ketamine was administered before diazepam. EEG activity before SE (B1), during stage 2 (B2a) and stage 3 SE (B2b), as well as after the administration of ketamine (B3) and diazepam (B4) is shown. Ketamine administered alone to terminate SE induced prolonged spindle-like bursts, that were punctuated by episodes of flattening on the EEG (B3a). The administration of diazepam abolished these bursts and induced continuous ictal spiking (B4a). Power spectral analyses are also shown.

minutes after the first injection of diazepam (with ketamine injected 1 h after diazepam), and it was replaced by irregular interictal spikes (Fig. 3C, 7 h). Although behavioral convulsions stopped after the administration of diazepam, a return to normal behavior characterised by walking, grooming, eating and exploratory behavior was observed only 794 (IQR = 478–1007.5) minutes after diazepam injection (with ketamine injected 1 h after diazepam) (n = 4) (table 1). None of the mice died following the administration of diazepam and ketamine.

A schematic of the injection protocol used in the second group of mice is shown in Fig. 2A. Baseline EEG recordings (Fig. 2 B1) and SE patterns associated to stages 2 (Fig. 2 B2a) and 3 (Fig. 2 B2b) from an animal in the second group (n = 4), in which ketamine was administered one hour before diazepam, are shown in Fig. 2B. Ketamine administration per se induced phases of prolonged spiking that were punctuated by episodes of flattening on the EEG, similar to what was reported by Martin and Kapur (2008) (Fig. 2B3). The subsequent administration of diazepam in these mice abol-

" Fig. 3. A: Bar graph showing the average EEG recovery time in both groups. No significant differences were observed (U4,4 = 22, p = 0.31). B: Bar graph showing the average behavioral recovery time in both groups. The administration of diazepam one hour before ketamine significantly shortened behavioral recovery time compared to when ketamine was administered one hour before diazepam (U4,4 = 10, p < 0.05). C: Representative EEG recordings from a pilocarpine-treated mouse in which diazepam was administered before ketamine to stop SE. EEG recordings were extracted at 3, 7, 11, 15 and 18 h after the administration of diazepam. Note that epileptiform activity associated to SE still persisted 3 h after the administration of diazepam. A transition to interictal spikes (insets) was then observed. D-E: Line graphs showing the dynamics of spiking activity associated to SE after the administration of diazepam one hour before ketamine (D) and after the administration of ketamine one hour before diazepam (E). Note that the end of SE on the EEG was associated to a significant decrease of ictal spiking in both groups and that in the second group, behavioral recovery time was significantly delayed.

116

S. Wang et al. / Neuroscience 421 (2019) 112–122

117

S. Wang et al. / Neuroscience 421 (2019) 112–122 Table 1. Recovery to baseline EEG and normal behavior during the two pharmacological procedures used to stop SE. Injection of diazepam before ketamine

Injection of ketamine before diazepam

No

Stage 2 SE duration (min)

Recovery to baseline EEG (min)

Recovery to normal behavior (min)

No

Stage 2 SE duration (min)

Recovery to baseline EEG (min)

Recovery to normal behavior (min)

1 2 3 4

8 13 7 20

194 337 337 245

580 376 1007 1008

5 6 7 8

23 6 27 22

196 328 157 238

1440 1440 1267 1276

ished these bursts and induced continuous ictal spiking (Fig. 2B4). EEG activity returned to baseline 217 (IQR = 176.5–283) minutes after injection of ketamine (with diazepam injected 1 h after ketamine). Only 50% (2/4) of mice returned to normal behaviour (walking, grooming, eating and exploratory behavior) within 24 h. The average recovery time in these mice was 1358 (IQR = 1271.5–1440) minutes. None of the mice died after the administration of ketamine followed by diazepam. When comparing the delay between the administration of diazepam or ketamine and the termination of continuous ictal spiking associated to SE, we found no significant differences between groups (U4,4 = 22, p = 0.31) (Fig. 3A). However, the administration of diazepam before ketamine significantly improved the behavioural recovery time compared to when ketamine was injected before diazepam (U4,4 = 10, p < 0.05) (Fig. 3B). These findings therefore indicate that although convulsive activity was terminated by diazepam and ketamine, none of the two methods was more effective in shortening EEG activity associated to SE. This is further illustrated in Fig. 3C, which shows EEG recordings from an animal in which diazepam was administered before ketamine. Three hours after the administration of diazepam, EEG epileptiform activity associated to SE was still observed; this SE pattern stopped approximately 245 min after injection of diazepam and was replaced by continuous and irregular interictal spikes (Fig. 3C) (cf., Wang et al., 2019). Fig. 3D and E show the dynamics of the spiking activity associated to SE after the administration of either diazepam followed by ketamine or ketamine followed by diazepam. In both groups, the high frequency spiking associated to SE was eventually replaced by a pattern of irregular interictal spikes that we considered as the end of SE. Changes in spectral frequency Next, we analysed the EEG activity in different frequency bands in the two experimental groups. To quantify the progression of ictal spiking frequency over time, we filtered the signal between 3 and 15 Hz, which comprised the average frequency of ictal spiking during stage 2 (6.2 ± 2.3 Hz) and stage 3 (8.8 Hz ± 1.72 Hz). The analysis of beta (18–25 Hz) and gamma (40–80 Hz) frequency bands was also performed since it has been reported that these frequency bands are significantly modulated by diazepam in both mice and rats (van Lier

et al., 2004; Scheffzu¨k et al., 2013). Gamma oscillations are also known to emerge during epileptiform activity in kainic acid-treated rats (Medvedev et al., 2000; Le´vesque et al., 2009). Fig. 4A and D show the distribution of spectral power over the normalized time for each frequency band, in each group. When comparing the last 20% of stage 3 (n = 20 data points representing the average of 4 animals) to the first 20% of the time-period after diazepam (n = 20 data points, n = 4 animals), no significant changes in power were observed in the 3–15 Hz frequency band (Z = 0.30, p = 0.76), but a significant increase in power between 18 and 25 Hz (Z = 3.91, p < 0.05) and a significant decrease in power between 40 and 80 Hz (Z = 3.91, p < 0.05) occurred (Fig. 4B). When comparing the last 20% of the time-period of stage 3 to the last 20% of the time-period after diazepam, a significant increase in power was observed between 3 and 15 Hz (Z = 3.92, p < 0.05) and between 18 and 25 Hz (Z = 3.92, p < 0.05), while the power in the 40–80 Hz frequency range significantly decreased (Z = 3.92, p < 0.05) (Fig. 4C). In the second group, in which ketamine was administered one hour before diazepam, ketamine induced a significant decrease in power between 3 and 15 Hz (Z = 3.92, p < 0.05) but an increase between 18 and 25 Hz (Z = 3.81, p < 0.05) and 40–80 Hz (Z = 3.92, p < 0.05), when comparing the last 20% of the time-period of stage 3 to the first 20% of the time-period after ketamine (Fig. 4E). When comparing the last 20% of the time-period of stage 3 to the last 20% of the time-period after ketamine, we observed a decrease in power between 3 and 15 Hz (Z = 3.92, p < 0.05) and between 40 and 80 Hz (Z = 3.51, p < 0.05), but a significant increase between 18 and 25 Hz (Z = 3.92, p < 0.05) (Fig. 4F). Altogether, these findings indicate that spectral power between 3 and 15 Hz, which corresponds to the average frequency of ictal spiking, significantly increases when diazepam is administered one hour before ketamine. Power in the beta range (18–25 Hz) was not differentially modulated, since it significantly increased after both the administration of either diazepam or ketamine. Finally, both ketamine and diazepam induced a significant decrease of gamma activity (40–80 Hz). Analysis of high-frequency oscillations Representative examples of a ripple and a fast ripple occurring during SE are shown in Fig. 5A and B, respectively. Ripples and fast ripples were observed during stage 2 and 3 of SE, as well as after the

118

S. Wang et al. / Neuroscience 421 (2019) 112–122

Fig. 4. Power analysis of stage 2 SE, stage 3 SE and after the administration of diazepam or ketamine to stop SE. A: Average power between 3 and 15 Hz (ictal spiking), 18–25 Hz (beta) and 40–80 Hz (gamma) from animals (n = 4) first treated with diazepam to stop SE. The grey areas represent the first and last 20% of normalised time-periods used for analysis. B: In the first group, diazepam induced a significant increase in beta (18–25 Hz) (Z = 3.92, p < 0.05) power but a significant decrease in gamma (40–80 Hz) activity (Z = 3.92, p < 0.05) when comparing the last 20% of stage 3 to the first 20% of the normalised time-period after diazepam. C: Similar results were observed in the beta (Z = 3.92, p < 0.05) and gamma (Z = 3.92, p < 0.05) frequency ranges when comparing the last 20% of stage 3 to the last 20% of the normalised time-period after diazepam. An increase in ictal spiking was however observed in the last 20% of the normalised time-period after diazepam (Z = 3.92, p < 0.05). D: Average power in each frequency band in animals (n = 4) in which ketamine was administered before diazepam. E: Ictal spiking significantly decreased in the first 20% after ketamine compared to the last 20% of stage 3 (Z = 3.92, p < 0.05). An increase in power in the beta (Z = 3.81, p < 0.05) and gamma frequency range was also observed (Z = 3.92, p < 0.05). Similar results were observed when comparing the last 20% of stage 3 to the last 20% of the normalised time-period after SE (F) except for gamma activity which decreased during the last 20% after ketamine compared to the last 20% of stage 3 SE (Z = 3.51, p < 0.05).

administration of diazepam followed by ketamine (Fig. 5C) or ketamine followed by diazepam (Fig. 5D). Further analysis revealed that when comparing the last 20% of stage 2 and stage 3 SE, ripples occurred at similar rates during stage 2 (n = 20 data points) and stage 3 (n = 20 data points) (Z = 1.57, p = 0.12) (n = 20 data points) (Fig. 5E), while fast ripples occurred at significantly higher rates during stage 3 (n = 20 data points) compared to stage 2 (n = 20 data points) (Z = 3.5295, p < 0.05) (Fig. 5F). We then compared rates of HFOs during the last 20% (n = 20 data points) of the normalised time-period of stage 3 to the first and to the last 20% (n = 20 data points) of the normalised time-period after the administration of diazepam. We found that diazepam induced a significant increase of ripples (Z = 3.85, p < 0.05) and fast ripples (Z = 3.92, p < 0.05) during the first 20%, but a significant decrease of ripples (Z = 3.93, p < 0.05) and fast ripples (Z = 3.68, p < 0.05) during the last 20% of the normalised time-

period (Fig. 6A and B). Similar results were observed for ripples (Z = 3.92, p < 0.05) and fast ripples (Z = 3.92, p < 0.05) during the first 20% after the administration of ketamine as well as ripples (Z = 3.00, p < 0.05) and fast ripples (Z = 3.41, p < 0.05) during the last 20% after the administration of ketamine (Fig. 6C and D). Finally, we investigated whether ripples and fast ripples were differentially modulated by diazepam and ketamine, during the last 20% of the normalised time-period after the administration of either drug to stop SE. We found that the decreases in ripple (Z = 3.77, p < 0.05) (Fig. 6E) and fast ripple (Z = 3.37, p < 0.05) activity (Fig. 6F) after diazepam were significantly larger compared to those observed after ketamine. Altogether, these findings indicate that both diazepam and ketamine induce similar results on ripples and fast ripples when compared to rates of HFOs during stage 3 SE. However, ripples and fast ripples occur at significantly lower rates after diazepam compared to ketamine.

S. Wang et al. / Neuroscience 421 (2019) 112–122

119

Fig. 5. A, B: Representative EEG recordings from a pilocarpine-treated mouse showing a ripple (Aa) and a fast ripple (Ba). C, D: Distribution of ripples and fast ripples during stage 2 SE, stage 3 SE, after the administration of diazepam or ketamine. Note that ripples and fast ripples were observed during SE and after the administration of diazepam or ketamine to stop SE. E: Bar graph showing the average rate of ripples during stage 2 and stage 3 SE. No significant differences were observed (Z = 1.57, p = 0.12). F: Bar graph showing the average rate of fast ripples during stage 2 and stage 3 SE. Fast ripples occurred at significantly higher rates during stage 3 compared to stage 2 (Z = 3.53, p < 0.05).

DISCUSSION The main findings of our study can be summarised as follows. First, multiple stages of SE are observed in pilocarpine-treated mice, as it was previously described in pilocarpine-treated rats and epileptic patients (Treiman et al., 1990). Second, although none of the two pharmacological methods is more effective in stopping EEG epileptiform activity associated to SE, the administration of diazepam before ketamine significantly shortened the behavioral recovery period compared to when ketamine was administered one hour before diazepam. Third, stage 3 SE showed similar rates of ripples compared to stage 2 SE but higher rates of fast ripples. Fourth, both injection protocols induced a significant decrease of power in the gamma frequency range (40– 80 Hz). Finally, rates of ripples and fast ripples were significantly lower when diazepam was administered before ketamine compared to when ketamine was administered before diazepam. As reported by Treiman et al. (1990) in pilocarpineand kainic acid-treated rats as well as in epileptic patients, we have observed that SE in pilocarpine-treated mice was characterised by multiple stages that were associated to different activity patterns on the EEG. However, although

stage 2 SE was observed in all animals, we found that only 37.5% of animals showed stage 1 SE, characterised by discrete seizures and interictal spikes. This discrepancy is probably due to the high sensitivity of mice to pilocarpine (Zablocka and Esplin, 1963; Curia et al., 2008; Buckmaster and Haney, 2012), which would make them more likely to enter directly into stage 2 SE after the injection of this drug. In addition, we did not observe stages 4 and 5 SE, presumably because we administered diazepam or ketamine approximately one hour after pilocarpine injection; accordingly, Treiman et al., (1990) have reported that stage 4 SE occurs approximately 2 h after the administration of pilocarpine or kainic acid. In line with previous studies (Walton and Treiman, 1988; Niquet et al., 2016), we found that stage 3 SE is resistant to both diazepam or ketamine. Diazepam administered alone to pilocarpine-treated rats was indeed not effective in terminating EEG seizures, which could be indicative of a decreased efficacy of GABAA receptormediated inhibition (Gao et al., 2007), due to the internalisation of benzodiazepine-sensitive GABAA receptors; this process should make them functionally inactive during the transition from isolated seizures (stage 1 SE) to continuous ictal activity (stage 2–5) (Goodkin, 2005; Naylor, 2005; Goodkin et al., 2008; Terunuma et al.,

120

S. Wang et al. / Neuroscience 421 (2019) 112–122

1997; Nelson et al., 2002), which could potentiate the effect of pilocarpine, which is a cholinergic agonist. As reported by Martin and Kapur (2008) and by Fritsch et al. (2010), the combined administration of diazepam and ketamine was also ineffective in terminating SE on the EEG. In our study, epileptiform activity associated to SE was still observed up to 3 h after, suggesting that although convulsive behavioral symptoms are reduced, epileptiform network activity still occurs in temporal lobe regions independent of whether diazepam or ketamine is administered first. We then observed a rapid transition to irregular interictal spike activity as shown in our previous study (Wang et al., 2019). Pilocarpine could therefore have immediate neurotoxic and inflammatory effects that favor neuronal hyperexcitability (Pitsch et al., 2017); such mechanisms could explain the occurrence of irregular interictal spikes early after SE. These findings therefore challenge the notion of a latent period characterized by normalization of the brain activity in rodent models of MTLE, and support the hypothesis that epileptogenesis is a continuous process during which epileptiform activity is generated early after the initial brain insult. Fig. 6. A-B: Bar graph showing the average rates of HFOs during the last 20% of the normalised We have also observed that the time-period of stage 3, the first and last 20% after the administration of either diazepam or administration of both diazepam and ketamine to stop SE. Diazepam induced a significant increase of ripples (Z = 3.85, p < 0.05) ketamine induced a significant during the first 20% of the normalised time-period after the administration of diazepam compared decrease in power in the gamma to stage 3. A significant decrease of ripples (Z = 3.92, p < 0.05) was however observed during the last 20% after the administration of diazepam compared to stage 3 (A). Similar results were frequency range (40–80 Hz). observed for fast ripples (B). C-D: Ketamine induced similar effects on HFOs since we observed a Gamma oscillations are known to be significant increase of ripples (Z = 3.92, p < 0.05) during the first 20% but a significant decrease associated to the activity of epileptic (Z = 3.00, p < 0.05) during the last 20% compared to stage 3 (C). Similar results were observed networks, both in patients (Fisher for fast ripples (D). E: Bar graph showing the average rate of ripples during the last 20% of the et al., 1992; Perucca et al., 2013; normalised time-period after the administration of diazepam compared to the same time-period after the administration of ketamine. Ripple rates were significantly lower under diazepam Ren et al., 2015) and in animal modcompared to ketamine (U20,20 = 257, p < 0.05). F: Bar graph showing the average rate of fast els (Medvedev and Willoughby, ripples during the last 20% of the normalised time-period after the administration of diazepam 1999; Medvedev et al., 2000; compared to the same time-period after the administration of ketamine. Fast ripple rates were also Medvedev, 2002; Gnatkovsky et al., significantly lower under diazepam compared to ketamine (U20,20 = 279.5, p < 0.05). 2008; Le´vesque et al., 2009). It is also been proposed that gamma oscillations are involved in epileptogenesis 2008; Goodkin and Kapur, 2009). The decreased efficacy (Medvedev et al., 2011). of GABAergic signaling could also be attributed to the Analysing the HFOs during SE induced by 4depolarizing shift of GABAA receptors due to the aminopyridine in rats, Salami et al. (2016) have reported decreased efficacy and internalization of the KCC2 that seizures leading to SE are associated to higher rates cotransporter during SE (Ellender et al., 2014; Silayeva of fast ripples compared to what is occurring in animals et al., 2015). The inefficacy of ketamine (or other NMDA that generate only isolated seizures. It was hypothesized receptor antagonists) to stop SE was also previously that SE depends on the preponderant activation of princiestablished (Fariello et al., 1989; Bertram and Lothman, pal (glutamatergic) cell networks, since these networks 1990; Martin and Kapur, 2008) and was attributed to the are also involved in the generation of fast ripples ketamine-induced release of acetylcholine in cortical and (Jefferys et al., 2012). Accordingly, inhibition of principal temporal lobe regions (Sato et al., 1996; Kikuchi et al.,

S. Wang et al. / Neuroscience 421 (2019) 112–122

cells delays the onset of SE in the lithium-pilocarpine model (Sukhotinsky et al., 2013) while perforant path stimulation-induced SE depends on NMDA receptor activation (Wasterlain et al., 2000). In our study, we found that stage 3 SE is associated to higher rates of fast ripples compared to stage 2, indicating that this stage could rest on the synchronous firing of principal cells. Activation of NMDA and AMPA-mediated transmission is indeed increased during SE (Naylor, 2005; Goodkin et al., 2008; Goodkin and Kapur, 2009; Naylor et al., 2013; Rajasekaran et al., 2013). Therefore, although epileptiform activity is not suppressed on the EEG, the increase in GABAergic inhibition induced with diazepam may inhibit excitatory glutamatergic activity thus explaining the significantly low rates of fast ripples in the hippocampus compared to animals treated with ketamine. In conclusion, our findings confirm that in pilocarpinetreated mice the administration of diazepam and ketamine is effective in stopping behavioral convulsive seizures associated to SE. However, EEG epileptiform activity persists for more than 3 h after the administration of these two anticonvulsant agents and progresses to a pattern of interictal spiking without any recovery to the control condition. We have also found that the administration of diazepam one hour before ketamine significantly shortens behavioral recovery and decreases gamma, ripple and fast ripple activity associated to SE, which are believed to mirror pathological network activity associated to epileptogenesis (Medvedev et al., 2011; Jefferys et al., 2012). These findings therefore suggest that this method results in a network change that favours recovery. Further studies should investigate whether this method is associated to a longer latent period or to lower rates of spontaneous seizures.

CONFLICTS OF INTEREST None of the authors has any conflict of interest to disclose.

ACKNOWLEDGEMENTS This study was supported by the Canadian Institutes of Health Research (grants MOP-130328 and PJT-153310) and the Savoy Epilepsy Foundation.

REFERENCES Behr C, Le´vesque M, Stroh T, Avoli M (2017) Time-dependent evolution of seizures in a model of mesial temporal lobe epilepsy. Neurobiol Disease 106:205–213. Bertram EH, Lothman EW (1990) NMDA receptor antagonists and limbic status epilepticus: a comparison with standard anticonvulsants. Epilepsy Res 5:177–184. Bortel A, Le´vesque M, Biagini G, Gotman J, Avoli M (2010) Convulsive status epilepticus duration as determinant for epileptogenesis and interictal discharge generation in the rat limbic system. Neurobiol Dis 40:478–489. Buckmaster PS, Haney MM (2012) Factors affecting outcomes of pilocarpine treatment in a mouse model of temporal lobe epilepsy. Epilepsy Res 102:153–159. Cendes F, Andermann F, Carpenter S, Zatorre RJ, Cashman NR (1995) Temporal lobe epilepsy caused by domoic acid

121

intoxication: Evidence for glutamate receptor-mediated excitotoxicity in humans. Ann Neurol 37:123–126. Chin RFM, Neville BGR, Scott RC (2004) A systematic review of the epidemiology of status epilepticus. Eur J Neurol 11:800–810. Curia G, Longo D, Biagini G, Jones RSG, Avoli M (2008) The pilocarpine model of temporal lobe epilepsy. J Neurosci Methods 172:143–157. Ding K, Gupta PK, Diaz-Arrastia R (2016) Epilepsy after Traumatic Brain Injury. In: Translational Research in Traumatic Brain Injury (Laskowitz D, Grant G, eds) Frontiers in Neuroscience. Boca Raton (FL): CRC Press/Taylor and Francis Group. Available at: http://www.ncbi.nlm.nih.gov/books/NBK326716/ [Accessed March 20, 2019]. Ellender TJ, Raimondo JV, Irkle A, Lamsa KP, Akerman CJ (2014) Excitatory effects of parvalbumin-expressing interneurons maintain hippocampal epileptiform activity via synchronous afterdischarges. J Neurosci 34:15208–15222. Engel J (1996) Introduction to temporal lobe epilepsy. Epilepsy Res 26:141–150. Fariello RG, Golden GT, Smith GG, Reyes PF (1989) Potentiation of kainic acid epileptogenicity and sparing from neuronal damage by an NMDA receptor antagonist. Epilepsy Res 3:206–213. Fisher RS, Webber WR, Lesser RP, Arroyo S, Uematsu S (1992) High-frequency EEG activity at the start of seizures. J Clin Neurophysiol 9:441–448. Fritsch B, Stott JJ, Joelle Donofrio J, Rogawski MA (2010) Treatment of early and late kainic acid-induced status epilepticus with the noncompetitive AMPA receptor antagonist GYKI 52466. Epilepsia 51:108–117. Gao XG, Liu Y, Liu XZ (2007) Treatment of late lithium-pilocarpineinduced status epilepticus with diazepam. Epilepsy Res 74:126–130. Glauser T, Shinnar S, Gloss D, Alldredge B, Arya R, Bainbridge J, Bare M, Bleck T, Dodson WE, Garrity L, Jagoda A, Lowenstein D, Pellock J, Riviello J, Sloan E, Treiman DM (2016) Evidence-based guideline: treatment of convulsive status epilepticus in children and adults: report of the guideline committee of the American epilepsy society. Epilepsy Curr 16:48–61. Gnatkovsky V, Librizzi L, Trombin F, de Curtis M (2008) Fast activity at seizure onset is mediated by inhibitory circuits in the entorhinal cortex in vitro. Ann Neurol 64:674–686. Goodkin HP (2005) Status epilepticus increases the intracellular accumulation of GABAA receptors. J Neurosci 25:5511–5520. Goodkin HP, Joshi S, Mtchedlishvili Z, Brar J, Kapur J (2008) Subunit-specific trafficking of GABAA receptors during status epilepticus. J Neurosci 28:2527–2538. Goodkin HP, Kapur J (2009) The impact of diazepam’s discovery on the treatment and understanding of status epilepticus. Epilepsia 50:2011–2018. Jefferys JGR, Menendez de la Prida L, Wendling F, Bragin A, Avoli M, Timofeev I, Lopes da Silva FH (2012) Mechanisms of physiological and epileptic HFO generation. Prog Neurobiol 98:250–264. Kikuchi T, Wang Y, Shinbori H, Sato K, Okumura F (1997) Effects of ketamine and pentobarbitone on acetylcholine release from the rat frontal cortex in vivo. Br J Anaesth 79:128–130. Leung AK, Hon KL, Leung TN (2018) Febrile seizures: an overview. Drugs in Context 7:1–12. Le´vesque M, Bortel A, Gotman J, Avoli M (2011) High-frequency (80– 500 Hz) oscillations and epileptogenesis in temporal lobe epilepsy. Neurobiol Disease 42:231–241. Le´vesque M, Langlois JMP, Lema P, Courtemanche R, Bilodeau GA, Carmant L (2009) Synchronized gamma oscillations (30–50 Hz) in the amygdalo-hippocampal network in relation with seizure propagation and severity. Neurobiol Disease 35:209–218. Le´vesque M, Salami P, Gotman J, Avoli M (2012) Two seizure-onset types reveal specific patterns of high-frequency oscillations in a model of temporal lobe epilepsy. J Neurosci 32:13264–13272. Le´vesque M, Avoli M (2013) The kainic acid model of temporal lobe epilepsy. Neurosci Biobehav Res 37:2887–2899. Le´vesque M, Salami P, Shiri Z, Avoli M (2018) Interictal oscillations and focal epileptic disorders. Eur J Neurosci 48:2915–2927.

122

S. Wang et al. / Neuroscience 421 (2019) 112–122

Martin BS, Kapur J (2008) A combination of ketamine and diazepam synergistically controls refractory status epilepticus induced by cholinergic stimulation. Epilepsia 49:248–255. Medvedev A, Mackenzie L, Hiscock JJ, Willoughby JO (2000) Kainic acid induces distinct types of epileptiform discharge with differential involvement of hippocampus and neocortex. Brain Res Bull 52:89–98. Medvedev A, Willoughby JO (1999) Autoregressive Modeling of the Eeg in Systemic Kainic Acid-Induced Epileptogenesis. Int J Neurosci 97:149–167. Medvedev AV (2002) Epileptiform spikes desynchronize and diminish fast (gamma) activity of the brain. Brain Res Bull 58:115–128. Medvedev AV, Murro AM, Meador KJ (2011) Abnormal interictal gamma activity may manifest a seizure onset zone in temporal lobe epilepsy. Int J Neural Syst 21:103–114. Mu¨ller CJ, Gro¨ticke I, Hoffmann K, Schughart K, Lo¨scher W (2009) Differences in sensitivity to the convulsant pilocarpine in substrains and sublines of C57BL/6 mice. Genes Brain Behav 8:481–492. Naylor DE (2005) Trafficking of GABAA receptors, loss of inhibition, and a mechanism for pharmacoresistance in status epilepticus. J Neurosci 25:7724–7733. Naylor DE, Liu H, Niquet J, Wasterlain CG (2013) Rapid surface accumulation of NMDA receptors increases glutamatergic excitation during status epilepticus. Neurobiol Disease 54:225–238. Nelson C, Burk J, Bruno J, Sarter M (2002) Effects of acute and repeated systemic administration of ketamine on prefrontal acetylcholine release and sustained attention performance in rats. Psychopharmacology 161:168–179. Niquet J, Baldwin R, Norman K, Suchomelova L, Lumley L, Wasterlain CG (2016) Midazolam-ketamine dual therapy stops cholinergic status epilepticus and reduces Morris water maze deficits. Epilepsia 57:1406–1415. Niquet J, Baldwin R, Suchomelova L, Lumley L, Eavey R, Wasterlain CG (2017) Treatment of experimental status epilepticus with synergistic drug combinations. Epilepsia 58:e49–e53. Perl TM, Be´dard L, Kosatsky T, Hockin JC, Todd ECD, Remis RS (1990) An outbreak of toxic encephalopathy caused by eating mussels contaminated with domoic acid. N Engl J Med 322:1775–1780. Perucca P, Dubeau F, Gotman J (2013) Widespread EEG changes precede focal seizures Charpier S, ed. PLoS One 8 e80972. Pitsch J, Becker AJ, Schoch S, Mu¨ller JA, de Curtis M, Gnatkovsky V (2017) Circadian clustering of spontaneous epileptic seizures emerges after pilocarpine-induced status epilepticus. Epilepsia 58:1159–1171. Rai S, Drislane FW (2018) Treatment of refractory and superrefractory status epilepticus. Neurotherapeutics 15:697–712. Rajasekaran K, Joshi S, Kozhemyakin M, Todorovic MS, Kowalski S, Balint C, Kapur J (2013) Receptor trafficking hypothesis revisited: plasticity of AMPA receptors during established status epilepticus. Epilepsia 54:14–16. Ren L, Kucewicz MT, Cimbalnik J, Matsumoto JY, Brinkmann BH, Hu W, Marsh WR, Meyer FB, Stead SM, Worrell GA (2015) Gamma

oscillations precede interictal epileptiform spikes in the seizure onset zone. Neurology 84:602–608. Salami P, Le´vesque M, Avoli M (2016) High frequency oscillations can pinpoint seizures progressing to status epilepticus. Exp Neurol 280:24–29. Salami P, Le´vesque M, Benini R, Behr C, Gotman J, Avoli M (2014) Dynamics of interictal spikes and high-frequency oscillations during epileptogenesis in temporal lobe epilepsy. Neurobiol Dis 67C:97–106. Sato K, Wu J, Kikuchi T, Wang Y, Watanabe I, Okumura F (1996) Differential effects of ketamine and pentobarbitone on acetylcholine release from the rat hippocampus and striatum. Br J Anaesth 77:381–384. Scheffzu¨k C, Kukushka VI, Vyssotski AL, Draguhn A, Tort ABL, Brankacˇk J (2013) Global slowing of network oscillations in mouse neocortex by diazepam. Neuropharmacology 65:123–133. Silayeva L, Deeb TZ, Hines RM, Kelley MR, Munoz MB, Lee HHC, Brandon NJ, Dunlop J, Maguire J, Davies PA, Moss SJ (2015) KCC2 activity is critical in limiting the onset and severity of status epilepticus. Proc Natl Acad Sci 112:3523–3528. Sukhotinsky I, Chan AM, Ahmed OJ, Rao VR, Gradinaru V, Ramakrishnan C, Deisseroth K, Majewska AK, Cash SS (2013) Optogenetic delay of status epilepticus onset in an in vivo rodent epilepsy model Bazhenov M, ed. PLoS One 8 e62013. Terunuma M, Xu J, Vithlani M, Sieghart W, Kittler J, Pangalos M, Haydon PG, Coulter DA, Moss SJ (2008) Deficits in phosphorylation of GABAA receptors by intimately associated protein Kinase C activity underlie compromised synaptic inhibition during status epilepticus. J Neurosci 28:376–384. Treiman DM, Walton NY, Kendrick C (1990) A progressive sequence of electroencephalographic changes during generalized convulsive status epilepticus. Epilepsy Res 5:49–60. Trinka E, Cock H, Hesdorffer D, Rossetti AO, Scheffer IE, Shinnar S, Shorvon S, Lowenstein DH (2015) A definition and classification of status epilepticus – report of the ILAE task force on classification of status epilepticus. Epilepsia 56:1515–1523. van Lier H, Drinkenburg WHIM, van Eeten YJW, Coenen AML (2004) Effects of diazepam and zolpidem on EEG beta frequencies are behavior-specific in rats. Neuropharmacology 47:163–174. Walton NY, Treiman DM (1988) Response of status epilepticus induced by lithium and pilocarpine to treatment with diazepam. Exp Neurol 101:267–275. Wang S, Le´vesque M, Avoli M (2019) Transition from status epilepticus to interictal spiking in a rodent model of mesial temporal epilepsy. Epilepsy Res 152:73–76. Wasterlain CG, Liu H, Mazarati AM, Baldwin RA, Shirasaka Y, Katsumori H, Thompson KW, Sankar R, Vasconselos AP, Nehlig A (2000) Self-sustaining status epilepticus: a condition maintained by potentiation of glutamate receptors and by plastic changes in substance P and other peptide neuromodulators. Epilepsia 41: S134–S143. Zablocka B, Esplin DW (1963) Central excitatory and depressant effects of pilocarpine in rats and mice. J Pharmacol Exp Ther 140:162–169.

(Received 15 May 2019, Accepted 3 October 2019) (Available online 6 November 2019)