Epileptic seizures are characterized by hyperexcitability and hypersynchronism of specific neuronal populations (1). The use of animal models of epilepsy may provide insights into the neural networks that are affected. The systemic pilocarpine (PILO) model of epilepsy (2,3) is useful because it resembles temporal lobe epilepsy (TLE) in humans in its electrographic, behavioral, and morphologic alterations. Croiset and De Wied (4,5) studied adrenocroticotropic hormone (ACTH) and vasopressin effects in seizures evoked by i.c.v. microinjections of PILO. Additionally, Millan et al. (6) characterized, with hippocampal microdialysis probes, the amino acid release in either freely moving animals treated with i.p. PILO or anesthetized rats treated with intrahippocampal (HIP)-PILO. More recently, Smolders et al. (7) and Lindekens et al. (8) checked hippocampal amino acid levels and the effectiveness of some antiepileptic drugs (AEDs) in this model. However, in none of these articles was a detailed behavioral, electrophysiologic, and morphologic analysis of the model made. In the current work, we used a more specific paradigm of TLE, microinjecting PILO in the hilus of dentate gyrus (HIP-PILO), and we asked the following questions: (a) Does SE occur in animals injected with HIP-PILO? (b) What are the associations between behavior and EEG? (c) Do these animals develop spontaneous recurrent seizures (SRSs)? And (d) What morphologic alterations occur after periods of 15 and 30 days after HIP-PILO? The main findings of our work were that animals treated with HIP-PILO displayed SE, SRSs, neo-Timm–positive mossy fiber sprouting, and limbic epileptiform activity, without animal loss after SE, thus supporting this as a selective and efficient model of TLE.
Summary: Purpose: We studied, by means of video-EEG and neo-Timm histochemistry, the behavioral, electrophysiologic, and structural characteristics of seizures induced by intrahippocampal microinjection of pilocarpine (HIP-PILO), a selective model of temporal lobe epilepsy (TLE).
Methods: We investigated the behavioral and electrophysiologic (hippocampus and amygdala EEG) evaluation of status epilepticus (SE) induced by HIP-PILO and the consequent spontaneous recurrent seizures (SRSs). We evaluated hippocampal structural rearrangements after SE by means of neo-Timm staining.
Results: HIP-PILO induced SE in 17 (71%) of 24 animals. Although three animals displayed spontaneous remission of SE (not used in analysis) before the established SE duration (90 min), none of those undergoing SE died. Of SE animals, 10 (71%) of 14 had SRSs. All animals with SE had clear-cut mossy fiber sprouting (MFS) in the inner molecular layer of the dentate gyrus and epileptiform activity in hippocampus and amygdala.
Conclusions: HIP-PILO rats displayed SE, SRS, MFS, and limbic epileptiform activity, without animal loss after SE. Thus, our data support this as a selective and efficient model of TLE.
Wistar rats (n = 34; 200–250 g, b.w.) were cannulated in the dentate hilus by means of stereotaxic surgery (9). The experimental group (n = 24) was injected with PILO (2.4 mg/μl; 1 μl), and the control group (n=10) was injected with saline (0.9%; 1 μl). All animals that had SE were injected either with diazepam (DZP; 5 mg/kg) or thionembutal (25 mg/kg) after 90 min of SE. Another experimental group (n = 7) was implanted with electrodes in the amygdala (AP, 2.1 mm; L, 4.7 mm; V, 7.1 mm) and with chemitrodes in the hippocampus (AP, 6.30 mm; L, 4.5 mm; V, 4.5 mm) for further EEG recordings. In all cases, control EEG (baseline) was recorded after 0.9% saline injection, during 30 min before PILO application, during the 90 min of SE duration, and during the blockade of SE by DZP or thionembutal. We monitored the behavior of all animals subjected to SE for 15 (n = 7) and 30 days (n=7), 8 h/day, using a camcorder (Gradiente GCP-195). These animals were perfused for neo-Timm (0.03% sodium sulfide) staining at the end of the observation period, and their brains subsequently frozen for histochemistry. All EEG recordings were made using a Cyberamp (Axon Instruments) and a Biopac MP100 amplifier and A/D converter, respectively. A video card All-in-wonder PRO (ATI) was used to synchronize the video signal with the EEG recordings.
Only animals that had the electrode and cannula position confirmed by histology were used. Data showed that 17 (71%) of 24 injected animals had SE. Three animals showed spontaneous remission of SE before the limit of 90 min. Although they were not included in SRS and EEG analysis, they displayed neo-Timm–positive sprouting (not shown). Ten (71%) of 14 SE animals had SRSs between 2 and 30 days after HIP-PILO. All animals that went into SE survived. The latency for SE was variable (29.86 ± 5.03 min, SEM; range, 6–72 min). Including animals from both groups (15 and 30 days after HIP-PILO), we found SRS with 16% class 2, 6% class 3, 50% class 4, and 28% class 5 (10) in a period of 30 days. The cumulative frequency of SRSs increased with time (Fig. 1A). The distribution of different severity indexes during the 30-day period is illustrated in Fig. 1B. Video-EEG recordings were conducted before and during SE (four of seven animals) and in animals that did not go into SE (three of seven). These recordings allowed behavioral–EEG correlations (Fig. 1C and D). The first epileptiform EEG discharges, generally associated with orofacial automatisms, occurred in the hippocampus (197.25 ± 23.56 s, SEM) and then sequentially in the amygdala (257.256 ± 18.51 s, SEM). Subsequent epileptiform EEG discharges, during SE, occurred during head and forelimb clonus, rearing, and falling. The amygdala signal had higher amplitude when compared with hippocampal EEG (Fig. 1C and D). In contrast to control rats (Fig. 1E), all SE groups had positive neo-Timm staining in the internal molecular layer of the dentate gyrus (Fig. 1F). Additionally, qualitative evaluation showed that some animals had stronger staining in the ipsilateral side from the PILO injection and in the ventral hippocampus (data not shown).
The main findings of this work support HIP-PILO as a selective and efficient model of TLE. We detected a high number of animals developing SE after HIP-PILO without associated death and with clear-cut incidence of SRSs over a period of between 15 and 30 days. Additionally they displayed SE, SRSs, neo-Timm–positive MFS, and hippocampal and amygdala epileptiform activity. Behavioral analysis showed that the number of seizures increased with time, with oscillations of the Racine (10) behavioral score, but with the presence of intense although variable seizure activity. Because some animals (12.5%) had spontaneous remission of SE, we suggest that endogenous mechanisms could block seizures.
Although several publications addressed the effect of intracerebral PILO microinjection, none of them evaluated the correlation between behavioral expression, deep EEG in limbic areas, and structural rearrangements. For example, Millan et al. (6) described the increase of hippocampal glutamate and aspartate, besides the animals being freely moving (i.p. PILO) or anesthetized (HIP-PILO). Croiset and De Wied (4,5) studied ACTH and arginine vasopressin anticonvulsant or convulsant effects, respectively, associated with i.c.v. PILO. Smolders et al. (7) used microdialysis probes to inject PILO and to sample γ-aminobutyric acid (GABA), glutamate, and dopamine after systemic (30 mg/kg) or local (10 mM) PILO application. Finally, Lindekens et al. (8) also evaluated valproate and its derivatives effects against HIP-PILO injection. Their electrophysiogic studies were done by electrocorticogram (EcoG) and not by deep recordings. None of these works addressed the morphologic alterations and the presence of SRSs.
In the current work all animals subjected to SE showed sprouting in the inner molecular layer of the dentate gyrus, independent of both time after SE (15 and 30 days) and occurrence of SRSs, suggesting that only the first injury is necessary for this morphologic alteration (3). Although we cannot blame a single structure for the behavioral and EEG disturbances found in this model, the typical synaptic alterations in the hippocampus of animals subjected to SE resemble those described in the human TLE.
We also observed that EEG patterns during SE could be clearly associated with behavior. Although the seizure initiates in the hippocampus, it spreads to amygdala, suggesting that other structures could be involved in the generation and maintenance of TLE. After high-dose injections of PILO, Turski et al. (2) detected amygdala electrographic discharges. These findings suggest that a disturbance in limbic circuitry, rather than in a single structure, could be responsible for TLE.
In conclusion, we showed that HIP-PILO is a reliable and efficient model to study TLE, because it induces SE, SRSs, EEG limbic alterations, and MFS.
Acknowledgment: We thank the Brazilian Foundations FAPESP (grants 92/4464-3; 93/2023-2, and 99/06756-0), CNPq (grants 521596/94 and 521697/96-4), and PRONEX (grant 357/96) for financial support. N.G-C., J.A.C.O., and M.A.F. were recipients of CNPq and CAPES-Brazil fellowships.