The mammalian cerebral cortex is a complex laminar structure with a variety of neuronal and non-neuronal cell types that develop in a finely orchestrated and stereotypic manner. Final laminar position and synaptic specificity of most cortical cell types are well described. Strikingly, any alteration in the developmental unfolding of one of these processes, even for a single cell type among tens, can be sufficient to generate neurodevelopmental disorders. However, how the establishment of this precise cellular architecture is regulated at the molecular level remains largely unknown. Several lines of evidence suggest a role of cell-cell communications via ligand-receptor (LR) interactions. Using a single cell RNA-seq (scRNA-seq) approach in mice, we have generated a bioinformatic atlas that infers LR based cellular communications across all cell types over somatosensory (SS) cortex development. Querying our atlas for known LR interactions has demonstrated its validity, but new LR-mediated cell-cell interactions remain to be discovered to interrogate its power as a hypothesis generator. In parallel, a technique called Multiplexed-Error Robust Fluorescence In Situ Hybridization (MERFISH) has been recently developed and implemented by us, which images single cell transcriptomes in situ and thereby adds precious information about spatial expression. Here, we will: (i) test some LR interactions predicted by our scRNA-seq atlas for a role in SS microcircuit development, (ii) build on the MERFISH technique to complement our atlas with spatial information and to characterize SS cortex cellular development with unprecedented resolution and (iii) use MERFISH to interrogate altered developmental processes in the SS cortex of a mouse model of neurodevelopmental disorder, the Neurod2 KO mouse.

Fragile X syndrome (FXS) is a monogenic pathology responsible of the main cause of inherited intellectual disability (ID) and autism. FXS affects 1/4000 men and 1/8000 women. There is no treatment yet validated. FXS is due to the absence or the loss of function of the FMRP (Fragile X Mental Retardation Protein). The Fmr1 knock-out mouse model (Fmr1-KO) recapitulates the symptoms of FXS and demonstrates that the lack of FMRP leads to alterations of synaptic plasticity underlying neuronal troubles of FXS. FMRP is an RNA binding protein whose absence causes an excessive translation of hundreds of proteins in neurons of several brain regions. This neuronal protein synthesis excess leads to the alteration of several forms of translation-dependent synaptic plasticity. Thus, FMRP appears as a key regulator of the mechanisms underlying the inter-neuronal communication. The precise molecular function of FMRP in this process is however still not fully understood. One outstanding question that remains unresolved despite extensive research efforts is how quasi-ubiquitous FMRP controls the translation of hundreds of mRNAs specifically in neurons. In this context, understand how the absence of FMRP leads to synaptic alterations remains a major goal to define the molecular basis of FXS and identify a treatment. Towards this goal, this project is based on our recent discovery that the loss of FMRP, besides leading to protein translation excess in neurons, is also leading to diacylglycerol (DAG) and phosphatidic acid (PA) lipid signaling deregulation. In neurons, FMRP is mostly associated with diacylglycerol kinase kappa (DGKk) mRNA and positively controls its translation. DGK enzymes are the master regulators of the switch between DAG- and PA- signaling pathways that control protein translation and actin filament stability, respectively, and that are proposed to orchestrate synaptic plasticity. The loss of DGKk is sufficient to reproduce FXS associated symptoms in the mouse. These data lead to a change of paradigm in the pathological mechanism of FXS and the function of FMRP: DGKk is a primary target of FMRP in neurons and the excess of DAG and a lack of PA signaling consecutive to DGKk deregulation contributes predominantly to the pathology. These data open new avenues of research towards understanding of FMRP function, and suggests novel therapeutic means. The scientific program, based on our newly identified pathomechanism, aims at understanding the molecular basis of Fragile X syndrome by following two main axes: understand the molecular function of the FMRP protein and identify a novel way of intervention. The program is organized to achieve four distinct goals: 1) define the molecular mechanism of FMRP translation control of Dgkk in neurons, 2) determine the functional consequences of Dgkk deregulation in the mouse and humans, or its absence (new Dgkk-KO mouse model) and demonstrate that its deregulation is critical for FXS condition, 3) validate DGKk as a novel therapeutic target for FXS in the Fmr1-KO mouse model. The program will be performed by a consortium that combines a panel of expertise spanning RNA/protein interactions, lipid signaling, neuronal electrophysiology, and animal behavior. The main expected outcome will be a detailed molecular description of the molecular mechanism by which FMRP contributes to the control of local protein translation within neurons, the deregulation of which is causing the well-defined neurological alterations of the Fragile X syndrome. A second main expected outcome is the validation of a proof of concept for a novel therapeutic mean in the FXS mouse model.

Navigation using an internal representation of the environment or a cognitive map requires allothetic information about the spatial arrangement of the landmarks and idiothetic (self-motion) information. Grid cells in the medial entorhinal cortex (MEC) exhibit a striking grid-like firing pattern that tesselate the environment and are supposed to be mainly driven by idiothetic cues in order to compute distances travelled in the environment. However, this point of view has been challenged by several studies showing that grid cells are highly sensitive to allothetic (visual) cues suggesting that these cues might be important in the establishment of grid cells firing patterns. Finally, the grid cells population might be functionally more heterogeneous than previously thought as they are present in two populations of projecting cells in layer 2 of the MEC that are characterized by different morpho-functional properties: stellate and pyramidal cells. Our general goal is to clarify grid cells functions by analyzing how they encode distance travelled and how they react to local contextual changes in a familiar environment (to update the cognitive map). To do so we will take advantage of the recent development of virtual reality: an efficient tool to modify environments instantaneously and in a very controlled and reliable way. In our set up, a head-fixed mouse is running back and forth on a virtual linear track in one out of four conditions that differ 1) in the richness of wall patterns (poor or rich linear track) and 2) in the availability of virtual 3D-objects inside the track (with or without object). Grid cells will be recorded while the animal is running in its familiar track before being exposed to a sudden change in the availability of the 3D-objects. We have already started to validate this protocol on hippocampal place cells. Preliminary results indicate that depending on the availability of wall patterns and 3D objects, place cells showed distance (firing at the same distance from start in both directions) or position coding (firing at the same position relative to the external visual cues). Interestingly, the paucity of wall patterns and 3D objects favored distance coding while the presence of 3D-objects favored position coding. Finally, place cells were very sensitive to local contextual changes of the environment as they ‘remapped’ when the availability of 3D-objects was modified. Thus this protocol enables us to study grid cells activity in conditions where distance or position coding is favored and to determine how they react to contextual changes in the environment We would like to answer four questions. First, are grid cells exhibiting distance and/or position coding? Second how are grid cells influenced by local contextual changes in the environment? Third, are pyramidal and stellate grid cells in layer 2 functionally heterogeneous? To differentiate stellate and pyramidal grid cells we will express Channerhodopsin-2 in one of these two populations of MEC layer 2 neurons in order to identify them by their excitatory response to light. Last, are grid cells necessary for distance/position coding and/or the update of cognitive maps in hippocampal place cells? To answer this question we will test the effects of an optical inactivation of the medial septum, a manipulation known to affect the spatial periodicity of grid cells, on place cells recorded in the virtual linear track. We believe that answering our four questions will be essential to understand the function of grid cells in spatial cognition. As grid cells are present in layer 2 of the MEC, one of the first region to be affected in Alzheimer’s disease, we think that a clear functional characterization of this layer is a prerequisite to develop new therapeutics aimed at ameliorating memory deficits seen in this illness.

Focal cortical Dysplasia (FCD) is one of the major causes of focal onset seizures refractory to antiepileptic drugs (AEDs). Only one-third of these patients are eligible for epilepsy surgery (i.e. the removal of the epileptogenic area), which offers the chance of seizure remission for 30%-40% of patients. There is a clear need to develop new and more efficient therapy. Gene therapy constitutes a promising alternative to conventional therapies as it can be applied focally using appropriate viral vectors and is less invasive than surgical resection, likely reducing undesirable changes in brain function. Various ongoing preclinical projects currently investigate the benefits of therapies aimed at increasing the expression of inhibitory peptides or ion channels or using chemogenetic compounds for the treatment of epilepsy. In this project we propose a different strategy, aimed at reducing the expression of kainate receptors (GluK2-containing KARs), a subclass of glutamate receptors involved in epileptogenesis, through local viral transduction of selective miRNAs. One of the main advantages of miRNAs is that it locally inhibits specific neuronal proteins without altering the genetic code. In a preclinical project, we recently proved that this strategy was efficient in reducing epileptiform events in temporal lobe epilepsy without affecting normal synaptic transmission (Boileau et al, under review). The aim of our project, supported by preliminary data, is to evaluate if the GluK2-containing KAR miRNA strategy can be proposed for treating focal cortical epilepsies and, more specifically, type II FCD.

The polarity of GABA signaling in the brain relies on transmembrane chloride gradients which are themselves regulated by cation/chloride cotransporters (CCCs). Intraneuronal chloride buildup promotes pathological activities in the human epileptic cortex and animal models of epilepsy. It is therefore crucial to develop novel approaches to prevent intraneuronal chloride accumulation in the pathological brain. We have recently identified an unsuspected role of membrane gangliosides in the control of CCC function in neurons. Interestingly, gangliosides, including monosialotetrahexosylganglioisde (GM1), are largely represented in neuronal lipid rafts and regulate several cellular processes including the membrane trafficking of neuronal proteins. Reduced GM1 levels are observed in mouse epilepsy models as well as in peritumoral tissue from human brain resections and participate in the development of seizures. Moreover, ganglioside metabolism deregulation is associated with epilepsy in animals and humans. Conversely, increasing ganglioside levels reduces brain trauma and SE-related damages. We propose to dissect the molecular mechanisms underlying the modulation of CCCs function by GM1 and to evaluate its therapeutic potential. Our preliminary results show a specific interaction of GM1 with both KCC2 and NKCC1 and regulations of their membrane mobility. In addition, we have identified a point mutation in KCC2 that abolishes GM1-KCC2 interaction. Importantly, this point-mutation in the KCC2 GM1-binding domain is associated with human epilepsy. The GABGANG consortium gathers leaders in their respective fields to address three main aims via a multidisciplinary, 4-year program. Our first aim is to elucidate the role of gangliosides in the regulation of KCC2 and NKCC1 membrane diffusion, nanoscale organization, membrane stability and function, and its consequences on chloride homeostasis and GABA signaling. Our second aim is to evaluate the contribution of GM1-CCC interaction in an animal model of temporal lobe epilepsy (TLE). In particular, we will evaluate the impact of status epilepticus on gangliosides and CCC expression. Then, we will study the impact of altered interaction with GM1 on CCC membrane diffusion and hyperexcitability in the epileptic hippocampus. If such effects are demonstrated, we will then test whether increasing GM1 levels can rescue CCC function and prevent interictal/ictal activity in the pilocarpine rodent TLE model. Our last aim is to address the impact of GM1 manipulations on GABA signaling and network activity in human postoperative cortical tissue. To achieve this, we will combine several state-of-the-art approaches including advanced single particle tracking, live 2-photon chloride imaging, mass-spectroscopy imaging, in vivo intrahippocampal multichannel electrophysiological recordings, GRIN-lens based endoscopic calcium imaging combined with behavior and multielectrode-array recordings from human organotypic cultures. This work is bound to disclose entirely novel mechanisms regulating neuronal chloride homeostasis and GABA signaling and identify novel therapeutically relevant targets that will be useful for intractable epilepsy as well as other neurological and psychiatric conditions associated with altered neuronal chloride transport.