Abolishing spontaneous epileptiform activity in human brain tissue through AMPA receptor inhibition

Abstract Objective The amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid receptor (AMPAR) is increasingly recognized as a therapeutic target in drug‐refractory pediatric epilepsy. Perampanel (PER) is a non‐competitive AMPAR antagonist, and pre‐clinical studies have shown the AMPAR‐mediated anticonvulsant effects of decanoic acid (DEC), a major medium‐chain fatty acid provided in the medium‐chain triglyceride ketogenic diet. Methods Using brain tissue resected from children with intractable epilepsy, we recorded the effects of PER and DEC in vitro. Results We found resected pediatric epilepsy tissue exhibits spontaneous epileptic activity in vitro, and showed that DEC and PER inhibit this epileptiform activity in local field potential recordings as well as excitatory synaptic transmission. Interpretation This study confirms AMPAR antagonists inhibit epileptiform discharges in brain tissue resected in a wide range of pediatric epilepsies.


Introduction
Molecular targets for pediatric epilepsy treatment have largely centered on voltage-gated sodium and calcium channels. The recent introduction of Perampanel (PER), a non-competitive amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR) antagonist, has introduced the concept of AMPAR modulation as an effective anticonvulsant strategy. 1,2 Similarly, recent studies in epilepsy rodent models have demonstrated the anti-epileptic action of decanoic acid (DEC), a major medium-chain fatty acid provided in the medium-chain triglyceride (MCT) ketogenic diet (KD), is mediated through direct and selective inhibition of AMPARs. 3,4 Human tissue samples provide a valuable tool for preclinical drug screening and mechanism-of-action epilepsy studies. 5 This is particularly useful in pediatrics, where few anticonvulsant drug trials include children and adult study results are often extrapolated to children with pediatricians prescribing "off-label". 6 In most human brain tissue preparations, epileptiform activity is induced through manipulation of artificial cerebrospinal fluid (aCSF) magnesium or potassium or by adding pro-convulsant agents (e.g. 4-aminopyridine). [7][8][9][10] We have found pediatric human brain tissue to be hyperexcitable and have recorded spontaneous epileptic activity. We show that antagonism of AMPAR has a profound anticonvulsant effect in tissue derived from a spectrum of difficultto-treat seizure syndromes.

Local field potential (LFP) recordings
About 450 µm thick brain slices were prepared and stored and recordings made as previously reported. 12 Epileptiform events were classified as activity displaying an amplitude fourfold greater than the root mean square baseline amplitude, providing the event count, while the time difference between these events provided the interevent interval (IEI). Statistical analysis was conducted using Prism 8. Measurements expressed as median (M), interquartile range (Q1-Q3) and min-max values.
Drugs DEC (Sigma, Dorset, UK) and PER (Eisai, Hatfield, UK) were prepared as 1 M stock using dimethyl sulfoxide.
Since inhibitory interneurons are subject to glutamatergic drive, any change in AMPAR activity is likely to depress gamma-aminobutyric acid (GABA) release and, indeed, we have seen this in recordings where PER is applied. It seems unlikely, however, that DEC would have any direct effect on GABA release. To investigate the actions of DEC on isolated inhibitory synaptic currents, we used whole-cell patch electrodes filled with the ionotropic glutamate receptor channel blocker, IEM-1460, at 1.5 mmol/L. Under these conditions, in which AMPARs were blocked from inside the recorded cell, we recorded miniature IPSCs (mIPSCs) from principal neurons in the presence of TTX 1 µmol/L (Fig. 1K). Analysis of pooled data revealed no significant change in amplitude, (28.9 AE 2.5 vs. 30.1 AE 2.2 pA, n = 5, P = ns), IEI of recorded mIPSCS after addition of 1 mmol/L DEC ( Fig. 1L; 129

PER inhibits excitatory but not inhibitory postsynaptic currents
To gain a better understanding of PER's ability to abolish spontaneous epileptiform activity, recordings of neurotransmitter release were conducted (Fig. 2E-J). Recordings revealed that principal neurons received GABAergic IPSCs with a mean median IEI of 96.99 AE 23.23 msec (Fig. 2F) and a mean median amplitude of  (Fig. 2H), IEI was significantly increased from 173 AE 217 to 2361 AE 873 msec (P = 0.03) in the presence of PER (Fig. 2I). The effect of PER on amplitude did not reach statistical significance (19.87

Discussion
This study confirms that AMPAR inhibition, either by DEC or PER, is effective in abolishing spontaneous epileptiform activity in human tissue from drug-resistant pediatric epilepsy patients through a direct reduction in excitatory neurotransmission.
DEC, a major constituent of the MCT KD, has previously been shown to have an anticonvulsant action in acute in vitro rat hippocampal slice models of epileptiform activity, acting through modulation of excitatory neurotransmission. [3][4][5] In our human tissue LFP recordings, the anticonvulsant effects of DEC were clearly demonstrated at the 300 µmol/L concentration, consistent with reported therapeutic pediatric plasma concentrations. 14,15 In WCPC experiments, the anticonvulsant mechanism was shown to be likely through the reduction of post-synaptic excitatory neurotransmission via AMPARs.
Similar effects were seen with PER, via a reduction in the frequency of EPSCs and overall charge transfer. These effects were observed in tissue from a wide array of epilepsy syndromes, and underline the importance of the AMPAR in understanding the development and treatment of the epilepsies. Like many aspects of epilepsy, it is not a simple case of 'too much' AMPAR activity leading to seizures, for example, in anti-AMPAR autoimmune encephalitis, antibodies generated against AMPAR epitopes lead to AMPAR internalization, thereby reducing excitatory drive while still causing temporal lobe seizures in man. 16,17 Similarly, recent work in a rat model of chronic TLE in our laboratory suggests that seizures induce a profound loss of AMPAR expression in vulnerable networks. 12 Hence, AMPARs would appear to play both causative and compensatory roles in seizures and epileptogenesis.
Currently, the management for children with drug-resistant epilepsy includes referral to an epilepsy surgery unit for assessment of suitability for resective surgery. Researchers have attempted to explore the role of the KD in improving seizure outcomes in epilepsy surgery patients with Focal Cortical Dysplasia Type II (the EDIBLE research study: www.edible.org.uk). While studies in human brain tissue do not provide a substitute for clinical trials in children, this preclinical study suggests that the MCT KD or PER may be effective in reducing seizure burden pre-operatively for these patients. Indeed, a synergistic therapeutic effect of DEC and PER was recently demonstrated in a study using in vitro epilepsy animal models and adult human tissue from brain tumor patients. 4 One criticism of human epileptic tissue research is the lack of "control" tissue. 18,19 For obvious ethical reasons, we cannot and will never have the opportunity to obtain pediatric "normal" brain tissue for the ideal control. However, this should not prevent the use of human tissue as an invaluable resource to study the true pathophysiological changes in pediatric drug-resistant epilepsy. Recent studies in human brain tissue have demonstrated the presence of extensive species-specific differences in neuron types and properties as compared to the rodent brain. 20,21 A clear example of this is the expression of 5HT3 receptors, which are present on excitatory cells throughout the human forebrain, but expressed only on inhibitory GABA cells in rodent brain. 22 Therefore, the use of human epileptic brain tissue could provide novel insights to complement preclinical animal model studies. This is of the particular importance of the testing of treatments in medically refractory epilepsy, as few animal models recapitulate fully the diversity of seizure types, localization and etiology of the patients seen in the clinic. In pediatric epilepsy research, slice preparations from resected tissue are extremely valuable and rare, and provide a unique substrate for testing novel and approved compounds in pharmacoresistant epilepsy.
In summary, this study shows the potential application of human tissue samples for epilepsy research and drug development. The heterogeneity of the human tissue samples is a true reflection of the variable etiology of refractory epilepsy and should be considered a strength of the human tissue approach to pediatric epilepsy research.