Stiripentol exhibits higher anticonvulsant properties in the immature than in the mature rat brain
U676, Inserm, Paris, France
Pediatric Neurology Service, APHP, Robert-Debré Hospital, Paris, France
Sorbonne Paris Cité, Univsity Paris Diderot, INSERM UMR676, Paris, France
Address correspondence to Stéphane Auvin, Service de Neurologie Pédiatrique et des Maladies Métaboliques, CHU Hôpital Robert Debré, 48, boulevard Sérurier, 75935 Paris Cedex 19, France. E-mail: email@example.com
After the first positive experimental data in rodents in the early 1970s demonstrating the anticonvulsant effect of stiripentol (STP), in vitro studies showed that STP acts directly on γ-aminobutyric acid A (GABAA) receptors. Chloride influx is higher when these receptors contain an α3 subunit, leading to the hypothesis that STP might exhibit higher efficacy in the immature brain.
We explored this issue by studying the efficacy of STP in P21 and P75 rats using the pentylenetetrazol model of acute seizures or the lithium-pilocarpine status epilepticus model. P21 and adult rats received vehicle, 150, 250, or 350 mg/kg of STP, i.p., 1 h before evaluating the anticonvulsant. We also studied the blood and brain levels of STP as well as the expression and the messenger RNA (mRNA) levels of the α3 subunit of the GABAA receptors at both ages.
STP exhibited anticonvulsant properties in both models at both ages, but STP was more effective in P21 than in P75 rats. This was shown by the significant suppression of seizure or status epilepticus occurrence in P21 with 350 mg/kg STP, whereas the same dose had no significant effect at P75. The blood level, brain level, and blood/brain ratio of STP did not explain these differences between the two age groups. Moreover, the higher anticonvulsant properties in the immature brain were not explained by the mRNA level or protein expression of the GABAA α3 subunit at either age.
Stiripentol exhibits higher anticonvulsant properties in the immature than in the mature brain. These findings require further investigation because it might lead to new clinical developments.
Stiripentol (STP) is an antiepileptic drug (AED) belonging to the aromatic allylic alcohols (Trojnar et al., 2005). Clinical trials in adults have been suspended owing to the absence of efficacy in association with carbamazepine (Martinezlage et al., 1984; Loiseau et al., 1988). In contrast, pediatric studies were conclusive. Two randomized trials demonstrated the efficacy of STP in combination with clobazam (CLB) and valproate (VPA) in children with Dravet syndrome (Chiron et al., 2000).
STP is an inhibitor of the cytochrome P450 (CYP) enzymes leading to the belief that its anticonvulsant actions arise from its ability to increase the concentrations of other AEDs (Tran et al., 1996; Giraud et al., 2006). However, experimental studies in adult rodents conducted in the early 1970s showed that STP exhibits anticonvulsant activity (Poisson et al., 1984; Shen et al., 1990, 1992). This effect seems to be related to a direct neuronal mechanism of action (Quilichini et al., 2006; Fisher, 2009). In hippocampal slices, STP directly modulates γ-aminobutyric acid (GABA)–mediated postsynaptic currents (Quilichini et al., 2006).
The expression levels of the GABA-R subunits change during the development or in brain disorders such as epilepsy (Laurie et al., 1992; Sperk et al., 2004; Zhang et al., 2007). Because of the higher chloride influx observed with STP through GABA receptor containing an α3 subunit (Fisher, 2009), it might be suggested that STP may be more effective in the immature brain. To test this hypothesis, we decide to explore the efficacy of STP in the immature (21st postnatal day [P21]) brain compared to the mature brain (P75) in experimental animal models of seizure. We also evaluated the plasma and brain concentrations of STP after drug treatments at both ages as well as the brain expression of the GABA-R α3 subunit.
The experiments were performed on P21 and P75 male Wistar rats (Charles River, L'arbresle, France). The animals were housed three per cage with an alternating 12 h light/dark cycle. All procedures were approved by the Animal Research Committee of Bichat-Debré University and were performed in agreement with local, institutional, national, and international laws and regulations. P21 rats were used to compare to adult rats because of the similarity of the epileptic phenomenology in the pentylenetetrazol (PTZ) model and in the lithium-pilocarpine model (Auvin et al., 2006; Mares, 2012).
Stiripentol powder (Biocodex, Gentilly, France) was diluted in Tween 80, and then diluted in saline (5% in saline) (Luszczki et al., 2010). Four treatments were studied at both ages (P21 and P75 rats): vehicle, 150, 250, or 350 mg/kg of STP, i.p., 1 h before each experiment (Shen et al., 1990; Luszczki et al., 2010). The number of rats included in this study is reported in Table 1.
At P20 or P74, we implanted, under isoflurane anesthesia, a cortical bipolar electrode (Plastics One Inc, Roanoke, VA, U.S.A.) on the dura on each side of the brain (P20: AP, 2.5 mm; ML, 3 or −3 mm; V, 0 mm to the bregma and P74: AP, 3 mm; ML, 4 or −4 mm; V, 0 mm to the bregma). The animals were connected to an MP100/EEG100B acquisition system (BIOPAC, Santa Barbara, CA, U.S.A.). EEG was acquired using ACQKNOWLEDGE 4.1 software (BIOPAC) along with simultaneous digital video. Both EEG and behavioral responses were analyzed off-line in a blinded fashion.
The convulsant drug (PTZ) produces a behavioral seizure. Pentylenetetrazol (Sigma-Aldrich, St Louis, MO, U.S.A.) was dissolved in 0.9% saline and was administered subcutaneously (s.c.) at a dose of 100 mg/kg (in a volume of 5 ml/kg). We started video–electroencephalography (EEG) recording 1 h before the injection and recorded each animal for 3 h. Using EEG recordings we were able to identify the delay between PTZ injection and the various seizure types induced by this drug (absence, myoclonus, and tonic–clonic seizures) (Auvin et al., 2006). Using video-EEG analysis, we also measured the duration of the tonic–clonic seizure (TCSz).
Lithium-pilocarpine status epilepticus (SE) model
At P20 or P74, animals were injected subcutaneously with 3 mEq/kg lithium chloride (Sigma-Aldrich). After 16–18 h, SE was induced by i.p. injection of pilocarpine (PC) (Sigma-Aldrich) at a dose of 30 mg/kg. We started video-EEG recording 1 h before the injection of pilocarpine. After 2 h of SE, we injected 2.5 mg/kg of diazepam (Roche, Boulogne-Billancourt, France) in order to reduce mortality (Francois et al., 2006). We stopped the recording 1 h before perfusion for histology.
Rats received an injection of pentobarbital (100 mg/kg, i.p.) and underwent transcardiac perfusion-fixation with 0.9% saline followed by 4% paraformaldehyde 24 h after the start of lithium-pilocarpine-induced SE. We studied the vehicle groups and the groups that received 250 mg/kg STP, i.p., at both ages. Brains were processed for inclusion in paraffin, cut into 10-μm-thick coronal sections, and stained with hematoxylin & eosin (H&E; Sigma, Saint-Quentin Fallavier, France). Injured cells, identified by their eosinophilic (acidophilic) cytoplasm and pyknotic nuclei, were counted bilaterally in the CA1, CA3, hilus, and dentate gyrus in three sections separated by 100 μm starting with the first section in the hippocampus approximating −3.6 mm posterior to the bregma.
Plasma and brain concentration of stiripentol
Separate groups of animals (Table 1) were injected, i.p., with STP at different doses. Blood was collected 60 min after STP injection by intracardiac puncture under anesthesia with isoflurane using propylene tubes containing lithium heparinate. The animals were then perfused with saline solution to clear the vessels (250 ml/animals). The brain and the cerebellum were removed and stored at −20°C. Blood samples were centrifuged at 3,500 g/min at 4°C. Plasma samples were stored at −20°C. The concentrations of STP were measured by high-performance liquid chromatography (HPLC) with ultraviolet detection. The samples were prepared as follows: (1) 50 μl of plasma sample were included in 100 μl internal standard (2 μg of AB1191 per ml of 50/50 methanol/H2O) and 50 μl of 50/50 methanol/H2O with 2 ml of tert butyl methylether; (2) The weighed brain sample were mixed with 100 μl internal standard (50 μg of AB1191 per ml of 50/50 methanol/H2O) and 50 μl of 50/50 methanol/H2O with 2 ml of H2O. A 0.5 ml volume of homogenate was extracted with 3 ml of pentane. Following agitation for 15 min, decantation and freezing in a dry ice/acetone bath, the organic layer was transferred and evaporated to dryness under nitrogen at 45°C. Residues of the samples were reconstituted with 200 μl of mobile phase (acetonitrile/H2O; 60/40) and used for HPLC analysis. The concentration of STP in plasma (μg/ml) and brain (μg/g) was expressed as total drug concentration. The coefficients of variation for intraassays and interassays were roughly 0.2–13% and 3–14%, respectively.
RNA extraction and relative quantification of Gaba3a gene expression by real-time PCR
Quantification of the relative expression of the GABA3A gene was achieved using glyceraldehyde phosphate dehydrogenase (GAPDH) as a standard for quantification. Samples were isolated from P21 (n = 6) and P75 (n = 6) rats. Hippocampus and cortices were dissected, and total RNA was extracted with the RNeasy mini kit, including the DNase step, according to the manufacturer's instructions (Qiagen, Courtaboeuf, France). RNA quality and concentration were assessed by spectrophotometry with the Nanodrop apparatus (Thermoscientific, Wilmington, DE, U.S.A.). Total RNA (1 ng) was subjected to reverse transcription using the Iscript kit from Biorad (Marnes la Coquette, France).
All the samples were set up in duplicate using Sybr Green Supermix (Bio-Rad) on the CFX96 apparatus (Bio-Rad). The protocol was 40 cycles with two-step program (5 s of denaturation at 96°C and 10 s of annealing at 60°C). Amplification specificity was assessed with a melting curve analysis. Primers were designed using PRIMER 3 (http://simgene.com/) software (for primer sequences see Figure 7).
Quantification of the relative expression of the Gaba3a gene was performed using Gapdh mean value from each sample, with the Bio-Rad CFX MANAGER 2.1 software (Bio-Rad, Hercules, CA, U.S.A.).
Freshly excised hippocampus and cortices from P21 (n = 6) and P75 (n = 6) rats were soaked in cell lysis buffer supplemented with protease inhibitors. Total proteins (60 μg/lane) were loaded onto 12% sodium dodecyl sulfate (SDS)-polyacrylamide gels and transferred to nitrocellulose membranes (GE Healthcare, Uppsala, Sweden). Membranes were probed with rat anti-actin (1/5,000; Sigma), anti-GABA3A (1/500; Sigma) antibodies. Antibodies were revealed by horseradish peroxidase–coupled conjugates, and the staining intensity was analyzed using the IMAGEJ program (NIH Image, http://rsbweb.nih.gov/ij/).
Our primary end point was to assess the number of animals that did not have TCSz using the PTZ model and the number of animals that did not start status epilepticus using the lithium-pilocarpine model. Categorical variables were analyzed using the Fisher's exact test. Data are expressed as mean ± standard error of the mean. Statistical analysis for continuous variables was performed using the Kruskal-Wallis test and Dunn post hoc analysis in PRISM 5.02 (Graphpad Software, San Diego, CA, U.S.A.).
Stiripentol prevents TCSz or SE occurrence only in P21 rats
Using the PTZ model, 350 mg/kg of STP, i.p., significantly decreased the number of P21 rats that experienced a TCSz but had no significant effect on P75 rats (Fig. 1). Similarly, the number of P21 rats that experienced SE induced by the lithium-pilocarpine model was significantly decreased following treatment with 350 mg/kg of STP, i.p., whereas there was no significant decrease in the occurrence of SE at P75 (Fig. 1).
Effect of STP on the latency to the various seizure types induced by PTZ
Stiripentol did not have any effect on the occurrence of absence seizure at both ages. In P21 rats, we found that 250 mg/kg of STP increased the delay from the injection of PTZ to the myoclonus and to the TCSz (Fig. 2). STP had no effect on the TCSz duration in P21 rats (Fig. 2). In P75 rats, 250 mg/kg of STP delayed the occurrence of the TCSz (Fig. 3). The duration of the TCSz was longer following treatment with 150 mg/kg of STP compared to both vehicle and 250 mg/kg of STP (Fig. 3).
Effect of STP on the latency to the start of status epilepticus
We found that STP delayed the start of the SE at a dose of 250 mg/kg in P75 rats and 350 mg/kg in P21 rats (Fig. 4).
Stiripentol at 250 mg/kg, i.p., exhibited a neuroprotective effect in Li-PC-induced status epilepticus model
Because a significant number of P21 rats did not experience SE, we decided to evaluate the neuroprotective properties of STP with 250 mg/kg at both ages. As described previously, the level of cell injury in the vehicle groups was higher in the CA1 in P21 rats, whereas the hilus and the dentate gyrus showed a higher level of cell injury in P75 rats (Sankar et al., 1998). At P21, we observed a decrease in cell injury in the CA1 area and in the hilus in the group that received STP 250 mg/kg, i.p., whereas in adult rats a significant reduction in the level of cell injury was observed in the CA1 and the dentate gyrus (Fig. 5).
Plasma, brain, and brain/plasma ratio of STP are similar in P21 and P75 rats that received 350 mg/kg
We compared the STP levels in plasma and in brain after an injection of 350 mg/kg, i.p., because this dose produces the strongest anticonvulsant effect (decrease of the number of rats that experienced TCSz or SE) (Fig. 6). There were no significant differences in plasma level, brain level, or in the brain/plasma ratio of STP between P21 or P75 rat groups.
mRNA and protein levels of α3 subunit GABA-R at both ages
We evaluate the α3 subunit of the GABA-R at both ages because it has been previously suggested that the higher chloride influx through the GABA receptor containing an α3 subunit might be responsible for a higher efficacy of STP in the developing brain (Fisher, 2009) (Fig. 7). We found an increase in the α3 subunit of the GABA-R mRNA in the cortices and in the hippocampus of P75 rats compared to P21 rats (Fig. 7). We observed a trend for an increase in expression of the α3 subunit of the GABA-R in P75 compared to P21 rats (p = 0.06).
Stiripentol exhibits higher anticonvulsant properties in P21 rats than in P75 rats, as shown by a decrease in the occurrence of seizure or status epilepticus in the former group. This effect was not related to any difference in the STP concentrations between the two age groups. Although, it is now established that STP acts though a direct neuronal mechanism (Quilichini et al., 2006; Fisher, 2009), we found no evidence of a higher level of α3 subunit of the GABA-R in the P21 rats compared to the P75 rats.
The anticonvulsant properties of STP have been established in several animal models. An elevation of seizure threshold in the PTZ model has been reported using an intraperitoneal injection of STP (300 mg/kg resulting in a plasma concentrations above 35 mg/L). Maximal anticonvulsive response was observed with an injection of 450 mg/kg or plasma concentration at or above 120 mg/L (Shen et al., 1990). In our study, we observed an anticonvulsant effect of STP in both age groups. This was shown by a delay in the occurrence of the generalized clonic seizure with PTZ or in the occurrence of the induced SE. This effect was observed at a blood level of STP above 40 mg/L (Fig. 6). In mice, other authors showed using PTZ that the median effective dose (ED50) for STP was 200 mg/kg, i.p. At the same dose, STP protected about 40% of mice in the bicuculline seizure model and 20% in the strychnine seizure model (Poisson et al., 1984). STP was also effective in maximal electroshock seizure model in mice (ED50 = 240 mg/kg, i.p.) (Poisson et al., 1984). STP also showed anticonvulsant properties in alumina-gel Rhesus monkeys with seizure induced by 4-deoxypyridoxine. Acute administration of STP (150 mg/kg, i.p.) was compared with standard AED (Lockard et al., 1985). STP delays the onset of seizures similarly to what occurs in valproate, but did not suppress them (Lockard et al., 1985). All of these observations demonstrate a direct antiepileptic effect of STP, and led to its clinical development without identifying its mechanism of action.
Our results demonstrating a higher efficacy of STP in the immature than in the mature brain seem consistent with a recent study showing that STP alone has a protective effect against hyperthermic seizure in Scn1a RX/+ at 1 month but not at 5 months (Cao et al., 2012). It is not possible to exclude the potential role of a difference in the plasma level of STP between the two ages since, the plasma value of STP was reported only in the 5-month-old mice (Cao et al., 2012). The authors also speculated that this effect might be related to a developmental expression profile subunit of the GABA-R (Fisher, 2009; Cao et al., 2012). Indeed, in patch-clamp studies in CA3 pyramidal neurons of immature rats, STP directly enhanced central GABAergic transmission by increasing both the GABA release and the duration of the GABA-R activation in a concentration-dependent manner and through a barbiturate-like effect (Quilichini et al., 2006). Using patch-clamp recordings from transiently transfected mammalian cells, it has also been shown that, in addition to a direct effect of STP on the GABA-R, the effects of STP are modulated according to the subunit composition of the receptor, that is, a higher chloride influx when the GABA-R contains an α3 subunit (Fisher, 2009). In our study, we found a trend for higher expression of the α3 subunit in the hippocampus of adult Wistar rats compared to the P21 rats. This was consistent with a higher level of mRNA of the α3 subunit in adults compared to the P21 rats in both the cortex and the hippocampus. Laurie et al. (1992) have shown a decrease of mRNAs of the α3 subunit in the cortex during brain maturation and a mild increase in the hippocampus. Similar to our findings, Yu et al. (2006) reported an increase from P10 to P30 of the α3 subunit in the cortex and in the hippocampus by immunocytochemistry followed by a slow decrease from P30 to P90. If the increase of expression of the α3 subunit of the GABA-R during brain maturation cannot explain our findings, other GABA-R subunits might explain the higher anticonvulsant properties of the STP in young animals. Expression of the β1 subunit might be relevant, since it increases in the hippocampus and in the cortex during brain maturation (Laurie et al., 1992), whereas STP induced chloride influx is decreased when the GABA-R contained a β1 subunit (Fisher, 2009). Further investigations are needed to decipher the exact molecular mechanisms that could explain the higher effects of STP in the immature than in the mature brain.
Our study permits one to prove for the first time that STP reduces the level of SE-induced cell injury. Due to the developmental characteristics of the lithium-pilocarpine model of SE (maximal level of cell injury in CA1 in 2- and 3-week-old rat compared to older rat, cell injury in dentate gyrus starting to be observed at 3-week-olds and in older rats) (Sankar et al., 1998), it is not possible to compare the data from the two age groups studied. The neuroprotective properties of STP might be of interest in Dravet syndrome because this childhood epileptic encephalopathy is characterized by febrile and afebrile, generalized, and unilateral SE occurring in the first year. However, it remains unclear whether the repetitive occurrence of SE is involved in SE-induced cell injury in Dravet syndrome (Catarino et al., 2011). Even in imaging studies, changes have rarely been reported in these patients, even after status epilepticus (Sakakibara et al., 2009; Tang et al., 2011). It has also been hypothesized that SCN1A mutation may provide a protective effect on hippocampal neurons (Auvin et al., 2008), but more research is needed to determine the occurrence of cell injury and the effect of AEDs. A comparison between historical patients with Dravet syndrome and patients who have been treated with STP might help to further explore this hypothesis.
Further investigations to understand the mechanisms underlying the more pronounced effect of STP in the immature brain might open new perspectives for the use of this AED. There are, for example, accumulating data showing a benefit of STP in malignant migrating partial seizure in infancy (Coppola et al., 1995; Perez et al., 1999; Djuric et al., 2011; Merdariu et al., 2013). New experimental data investigating the anticonvulsant profile at early developmental ages might be of particular interest, even though the comparison with adult rats would not be possible due to the developmental changes in seizure pattern in animal models (Auvin et al., 2012; Mares, 2012).
The authors thank Biocodex for the gift of the stiripentol. Stéphane Auvin is partially supported by INSERM Grant (Contrat Interface INSERM 2010). Pascal Dournaud is partially supported by Université Denis Diderot-Paris 7, Assistance publique-Hôpitaux de Paris (AP- HP) (Contrat d'Interface). Pierre Gressens is partially supported by Assistance publique-Hôpitaux de Paris (AP- HP) (Contrat d'Interface).
None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
Stéphane Auvin is child epileptologist and is also conducting research works in Robert-Debŕ Hospital, Paris.