VILLARD Laurent

Epilepsy is a frequent (around 1%) and highly heterogeneous neurological disorder. Among the various forms of childhood epilepsy, our team is interested in a group of rare genetic epilepsies: developmental and epileptic encephalopathies (DEE). These diseases are characterized by 1- early onset, from the very first weeks of life; 2- clinical manifestations of erratic myoclonus, spasms and/or quasi-continuous partial seizures; 3- a highly abnormal interictal EEG; 4- a very severe course, with high mortality during the first years of life; 5- absence of effective treatment, particularly conventional anti-epileptics. Because the vast majority of cases are sporadic, the only way to intervene effectively will be to develop new therapeutic approaches. This is our final aim.

> Diagnostic activity

During the last 15 years, Laurent Villard has supervised the molecular diagnosis of genetic epilepsies at the Assistance Publique Hôpitaux de Marseille. This activity enables a diagnosis to be made in 40% of cases when seizures are present from the first week of life, and strengthens the team's clinical and molecular knowledge of these diseases, as well as its expertise in the field. It also makes it possible to increase the size of the cohorts, and to develop translational research projects (see below). With our colleagues at the Lyon, La Pitié Salpétrière and Strasbourg university hospitals, we set up in 2015 a national network for the diagnosis of genetic epilepsies (EPIGENE network; see PMID 35091117). Our hospital cohort has now exceeded 3,700 patients.

> Identification of new genes

In collaboration with Kerstin Kutsche's team in Hamburg, we have identified two new neonatal epilepsy genes (PMID 20890276). These are two NMDA receptor subunits encoded by the GRIN2B and GRIN2A genes. In collaboration with other teams, we have also demonstrated that composite heterozygous mutations in the TBCD124 gene are frequent and cause several types of epilepsy (PMID 23526554; PMID 25769375; PMID 26207815).

We participated in the identification of a new epileptic encephalopathy gene with the genetics team at Dijon University Hospital (PMID 24995870). More recently, we have identified several genes thanks to exome sequencing: UBR5 studied with Philippe Campeau in Canada (article in preparation), PCDHGC4 (PMID 34244665), KMT2E (PMID 31079897) or SYT1 currently studied in collaboration with James Rothman's team at Yale (USA). We also participated in the development of a predictive system for X-linked disease genes (PMID 36323681).

Our team has been awarded H2020 funding as part of the Twinning 2019 call. This program paired us with Tunisia to better understand the molecular landscape of recessive epilepsies (Service de Neurologie Pédiatrique, Hôpital Hedi Chaker de Sfax). Genetic studies using NGS will now be carried out in Sfax (PMID 34273994) and eventually shared with our team for variant analysis and functional studies.

> Pathophysiology of DEE

The KCNQ2 gene is the major gene in neonatal DEE. This gene encodes a subunit of a potassium channel producing the M current, responsible for the hyperpolarization phase following an action potential. Through the team's diagnostic activity, we have identified several dozen de novo pathogenic variants in this gene. Historically, pathogenic KCNQ2 variants were known in patients with benign familial neonatal epilepsy (BFNS). In BFNS patients, the neurological prognosis is good and development is generally normal. These forms are usually familial. The fact that pathogenic variants in the same gene cause such different (and severe) clinical pictures led us to hypothesize that the potassium channel containing KCNQ2 might be differentially affected in the two patient populations. We therefore built a research program on the pathophysiology of KCNQ2-DEE in collaboration with INMED (Inserm U1249), INT (CNRS UMR7289) and i-Stem (Centre d'Etude des Cellules Souches, Evry) (ANR 2014 and ANR 2019 funding, L. Villard coordinator).

In order to have cellular models adapted to our projects, we have produced induced pluripotent cell lines (iPS). Thanks to our collaboration with i-Stem, our team has learned how to produce mature human cortical neurons. We have also produced the first KCNQ2-DEE mouse model in the form of a knock-in mouse containing a recurrent variant associated with a severe and typical phenotype (for a review of the models see PMID 36047730). Our recent work shows that this model perfectly reproduces the expected pathology (PMID 32239694). This model has attracted the interest of several pharmaceutical companies, and we have initiated pharmacological trials for some of these companies' candidate molecules (NDA and non-exclusive licenses in place).

Our research into KCNQ2-DEE has already led to some original results concerning the mechanisms of epilepsy. We have demonstrated that certain mutations are capable of inducing delocalization of KCNQ2 channels, which constitutes a new mechanism for this disease (PMID 26007637). Another study showed that certain gain-of-function mutations have a similar effect to loss-of-function mutations (PMID 27030113). More recently, we have demonstrated the role of Kcnq2 in the genesis of motor rhythms and characterized cortical activity in the team's knock-in mice (PMID 33186352; PMID 35389519). We have also developed a seizure triggering system in this model and characterized the brain regions activated by seizures (PMID 37187037). More recently, we showed that a chronic treatment with carbamazepine in the mouse is able to reduce the number of seizure and, more importantly, to improve the neurodevelopmental outcomes (PMID 40632471). Evaluation of a new treatment strategy is currently performed in Marseille University Hospital (Pr Mathieu Milh).

The consortium funded by the European Joint Program on Rare Diseases 2020 (France, Belgium, Germany, Italy - TreatKCNQ project) to which we belong aimed to develop and test safer and more effective derivatives of retigabine, a molecule active on KCNQ potassium channels but which has been withdrawn from the market due to unexpected side effects. The project also aimed to identify new activators and blockers of KCNQ channels which will be tested in our models (human neurons and knock-in mice). We have identified several candidate molecules. We are also studying the potential of antisense oligonucleotides for allele-specific inactivation.

In parallel, we are studying the axon initial segment (AIS) in human neuronal models and in knock-in mice, in order to determine whether compensatory mechanisms linked to the structure of the AIS come into play in KCNQ2-DEE. We have begun "omics" studies (RNA-seq and proteomics), the results of which will be completed and integrated over the coming months (with the help of the systems biology team of MMG lead by Anaïs Baudot). We plan to study neurochemical changes in the brains of our animal models. We have a microelectrode array recording system which we will used to study the electrophysiological signature of human neurons from patients.

Intellectual Disability (ID) is the most common major handicap in children and young adults and affects nearly 3% of the general population. The causes of these disorders are very diverse (environmental or genetic) and, despite advances in cytogenetics and molecular biology, nearly 50% of cases remain unexplained, mainly for non-syndromic ID. The importance of these diseases in terms of public health and the complexity of the biological processes involved make understanding the pathophysiological basis of ID one of the major scientific and medical challenges over the coming years. The project presented here aims to improve diagnosis for patients presenting ID, to better understand the pathophysiology of these conditions and therefore the brain development.

• New genes identification

The Human Neurogenetic team works in close collaboration with the clinicians of the Timone Hospital’s medical genetics and pediatric neurology departments. Taking advantages of this fruitful collaboration, we have begun to set up a cohort of patients with ID associated or not with other clinical signs. This cohort currently includes more than twenty families. Thanks to Whole Exome sequencing (WES), we were able to describe two genes involved in new neurodevelopmental disorders: 1) PCDHGC4 involved in a syndromic ID associated with progressive microcephaly, seizures and joint anomalies (PMID: 34244665) and 2) A specific developmental and epileptic encephalopathy (DEE) linked to NAPB (PMID: 37014259). We also have identified new variants in known ID genes as TRAPPC2L (PMID: 36849228), DYNC1H1, SNX14 or NDST1 (under review).

However, despite the power of WES, about 25% of patients in our cohort are still undiagnosed. We want to improve bioinformatics tools to better identify CNV (Copy Number Variation) or transposable elements that could explain the pathology in a part of these patients. To this aim, we are part of a pilot study led by Dr. S. Gorokhova in collaboration with GBiM platform (Genomic Bioinformatic Marseille), which obtained the support of the French Foundation for Rare Diseases.

In parallel, we want to combine WES data and transcriptome analyses (performed by RNA sequencing). This technology can be used to identify abnormal events, such as alternative splicing, novel transcripts, and fusion genes, as well as characterize intronic variant of uncertain significant. Therefore, we propose to study the transcriptome of blood, fibroblasts or iPSC-derived neurons in order to study specific neuronal transcripts.

• NDST1 functional analysis

When variants are identified in a gene, it is essential to perform an in-depth functional study to confirm their involvement in the patients’ pathology. Recently, we identified the same homozygous NDST1 (N-deacetylase/N-sulfotransferase member 1) variant, in two independent families. This gene encodes a bifunctional enzyme involved in the heparan sulfate (HS) biosynthetic pathway. Recessive missense pathogenic variants in NDST1 have been reported in 18 ID patients but no functional studies were carried out. In order to discriminate between a rare polymorphism, which would not explain the disease, and a pathogenic variant, we established a collaboration with Dr. Lena Kjellén (Uppsala University, Sweden) to evaluate both enzymatic activities of the mutant protein (under review).

To complement these results, we would like to study the different variants identified to confirm that both domains of this enzyme are indeed involved in the pathology.

• Autophagy and intellectual disability

Autophagy is a self-degradative process that plays an essential physiological role as housekeeper in removing non-functional proteins, damaged organelles and eliminating intracellular pathogens. This pathway is even more important in neurons since they are post-mitotic cells. Recently, we identified a single homozygous variant, in two brothers with a syndromic ID disorder, in a newly characterized gene involved in the autophagy pathway: RMC1 (Regulator of MON1-CCZ1). The Mon1-Ccz1 complex is involved in the recruitment of Rab7 during the endosome and autophagosome maturation and their fusion with the lysosome resulting in the digestion of enclosed material. To the best of our knowledge, this gene has never been described in any human pathology. The functional analysis of RMC1 variant is essential to confirm its involvement in the patients’pathology.

Our preliminary results are promising and show that the Autophagy seems to be disturbed in patients’fibroblasts. To confirm these results, we will continue to analyze this pathway in fibroblasts. We would like also to develop iPSC-derived neurons in order to study axonal growth, dendrites development and their electrophysiological properties with a multi-electrode array system.