Modulation of GABAergic dysfunction due to SCN1A mutation linked to Hippocampal Sclerosis

Introduction

Mesial temporal lobe epilepsy (MTLE) with hippocampal sclerosis (Hs) represents the most common type of focal drug-resistant epilepsy.¹ Although MTLE is now routinely treated with neurosurgery, over a third of MTLE patients do not achieve seizure freedom,² and surgery can have important adverse consequences. Better treatment options, or even prevention, of MTLE and Hs are therefore needed, but rational therapy remains elusive because their causes remain unclear. A large body of evidence indicates the biologic processes that are relevant in the pathogenesis of Hs include glial activation, immune response, synaptic transmission, signal transduction, ions transport, and synaptic plasticity.^{3, 4} In the hippocampus, GABAergic inhibitory dysfunction mostly contributes to hyperexcitability in MTLE with Hs.^5-7

Of interest, recent studies illustrate SCN1A involvement in the epileptogenic neuronal network underlying MTLE associated with Hs.⁸ Accordingly, we already illustrated that MTLE with febrile seizures may be part of the epileptic phenotype encountered in a large family carrying a SCN1A mutation.^{9, 10} This M145T mutation of SCN1A cosegregated in all affected individuals of a large family affected by simple febrile seizures (FS), three of whom later developed MTLE.¹⁰ Most important, the M145T mutation is located in a highly conserved amino acid in the first transmembrane segment of domain I of the SCN1A, and functional studies in mammalian cells demonstrated that the M145T mutation causes a 60% reduction of current density and a 10 mV positive shift of the activation curve.⁹ More recently, one family member with refractory MTLE and Hs has benefited from surgery, and now we wish to report the extensive neurophysiological study from the resected hippocampal tissue of this patient, in comparison with seven sporadic patients with refractory MTLE and Hs. We believe that this study may provide important new directions to the physiopathological comprehension of MTLE with Hs and FS, and ultimately may pave the road to develop new therapeutic approaches in MTLE.

Patients and Methods

Results

Patch-clamp experiments in human slices

To highlight the functional implications of mNa_V 1.1 expression, patch-clamp experiments were performed on pyramidal cells and interneurons of the resected hippocampus of the mNa_V1.1 patient, comparing the results with those obtained from sporadic MTLE patients (patients, # 4-8, Table S1). Current steps (100–150 pA, 500 ms duration) applied to the cells were able to elicit one or more action potentials (APs), with pyramidal neurons from all patients exhibiting APs with similar characteristics, in particular with no change in AP threshold (MTLE: −53 ± 3 mV; n = 10; # 4-8, Table S1, mNa_V 1.1 patient: −53 ± 3 mV; n = 5; P = 0.64; Fig. 1A,B). By contrast, three interneurons from the mNa_V1.1 patient showed a significantly depolarized AP threshold (−40 ± 3 mV) in comparison to interneurons from MTLE patients (−58 ± 12 mV, n = 5; # 4-8, Table S1; P = 0.04; Fig. 1A,B, see Table S2 for electrophysiological parameters). These data suggest that mNa_V1.1 expressing interneurons may have a reduced intrinsic excitability as expected from M145T mutation of SCN1A.⁹

GABA currents rundown

To investigate how GABA currents respond in mNa_V1.1 patient’s hippocampus, we performed microtransplantation experiments injecting oocytes with hippocampal tissues from sporadic MTLE patients (without known genetic mutations and with Hs, Fig. 2A), one control individual (Fig. 2B) and mNa_V 1.1 patient (Fig. 2C-E). The results confirmed that a strong, pathological GABA-current rundown is present in drug-resistant MTLE hippocampi (49.0 ± 9.1%; n = 33; patients #1,2,4,5 Table S1; Fig. 2A), whereas the same phenomenon was absent in the control (75.8 ± 7.0 %; n = 10; # 3, Table S1; Fig. 2B) as previously shown.^5-7

Interestingly, the sclerotic hippocampus of mNa_V 1.1 patient also showed a GABAergic rundown (43.2 ± 3.1 %; n = 19; Fig. 2C) similar to sporadic MTLE patients. This current rundown was mitigated by pretreatment of 2 hours with endogenous factors, as BDNF, or exogenous agents, as the cannabis derivative cannabidivarine (CBDV, Fig. 2D; Table S3) showing again a strong similarity with data previously published for MTLE patients.^{5, 11}

Interestingly, we found that acute co-application of cannabidiol (CBD, 5 µM, GABA 50 µM), which has been reported to be clinically effective in Dravet syndrome,¹² potentiated both GABA-current amplitude from mNa_V 1.1 patient (I_GABA; +29.8 ± 4.1 %; n = 13; Fig. 2E, Table S3) and from sporadic MTLE patients (I_GABA; +27.5 ± 8.8 %; n = 22; #1,2, Table S3).¹³

Discussion

The novelty of our study is that we performed electrophysiological experiments from an exclusive hippocampal tissue obtained from a patient with MTLE and Hs, who also carried the M145T SCN1A loss-of-function mutation. Compared to hippocampal tissue from MTLE patients, there was a more severe impairment of GABAergic transmission, due to the lower GABAergic activity related to the Na_V1.1 loss-of-function, in addition to the characteristic GABA-current rundown, a hallmark of MTLE.^{5-7, 11}

We recorded valuable APs on both pyramidal neurons and interneurons from mNa_V 1.1 and five (of seven) sporadic MTLE patients. In all of them, the histopathological study showed the features of Hs related to MTLE, with the characteristic pattern of neuronal loss in CA1 (see Table S1).^{3, 4} Nonetheless, only in hippocampal mNa_V 1.1 slices, AP threshold was more depolarized in interneurons than in pyramidal neurons, reflecting a reduced excitability of GABAergic interneurons related to SCN1A mutation.^{14, 15} In spite of the limited number of recordings on mNav 1.1 interneurons, this type of AP responses paralleled the functional data obtained from cells transfected with M145T mutated cDNA.⁹

Because of the limited quantity of fresh tissue from the patient and technical difficulty of recordings from acute hippocampal slices, part of the tissue was snap-frozen and used for recordings in microtransplanted oocytes.^{6, 16} Notably, we found a GABA current rundown that was very similar to the one recorded from hippocampi of sporadic MTLE patients.^{5-7, 11} The GABA current rundown is considered an electrophysiological dysfunction of refractory human MTLE,^5-7 which may explain, at least in part, the hyperexcitability underlying the ictogenesis in these patients. Interestingly, we previously showed in epileptic rats that GABA current rundown arises from the hippocampal region (during status epilepticus) then spreading to temporal cortex in the chronic phase of seizures.^{7, 17}

We already emphasized that the pharmacological modulation of the GABAergic impairment (i.e., GABA current rundown) underlying MTLE may represent an interesting therapeutic approach in pharmacoresistant MTLE.¹⁷ In this way, it is not surprising that BDNF could ameliorate the GABA current rundown both in the mNa_V 1.1 and sporadic MTLE patients, as previously reported for both refractory human MTLE⁵ and preclinical animal models of MTLE.⁷ Notably, CBDV continues to show its potential in recovering the mNav 1.1 GABAergic rundown as previously shown for MTLE.¹¹ Furthermore, CBD was capable to increase GABA current amplitude in mNa_V 1.1 and MTLE patients, as reported in Dravet syndrome and tuberous sclerosis complex.^{12, 13, 18} Altogether our results emphasize the role of cannabis derivatives as treatment for refractory epilepsies.

In our patient, the overall clinical, pharmacological, and neuropathological findings were consistent with those observed in sporadic MTLE with Hs, including the excellent outcome from anterior temporal lobectomy. In this way, our findings reinforce the belief that known genetic defect, which is present diffusely in the brain, do not necessarily preclude a good prognosis following epilepsy surgery, if surgery is a reasonable option based on the concordance of other data during presurgical evaluation. It cannot be excluded, however, that the milder loss-of-function impairment of SCN1A related to the M145T mutation could also influence the better postsurgical outcome.

Important, as previously reported in Dravet syndrome and tuberous sclerosis complex,^{13, 18} CBD was capable to increase GABA current amplitude in mNa_V 1.1 and MTLE patients, that was recently approved as treatment for Dravet and Lennox–Gastaut syndromes.¹² Thus, in accordance with the “interneuron hypothesis,”¹⁴ our findings support the assumption that seizures in mNa_V 1.1 patient could arise from a minor GABA release from interneurons together with a reduction of GABAergic postsynaptic efficacy after repetitive stimulation (i.e., rundown). Major therapeutic implication of these findings is that by rescuing the GABAergic inhibitory activity it is possible to improve the clinical outcome of this kind of patients. In this way, this study ultimately may pave the road to develop new antiepileptic approaches in epileptic patients with Hs.

Conflicts of Interest

None of the authors has any conflict of interest to disclose.