Sudden cardiac death (SCD) is a major cause of adult mortality, accounting for about 350 000 deaths in Europe annually. SCDs in patients without structural heart disease are caused by ventricular arrhythmias generated from areas with abnormal electrical properties (‘substrate’) present within the heart. Paradoxically, despite the existence of efficient preventive therapies, most patients at risk cannot be identified pre-emptively to avoid sudden death. The major problems are 1) our incomplete mechanistic understanding in the role of electrical substrates in arrhythmia induction, 2) the information derived from the current benchmark noninvasive data is insufficient. Real-time, noninvasive imaging of cardiac electrical activity can be obtained through electrocardiographic imaging (ECGI) that could be used to gain insight into the mechanisms of sudden cardiac death. My aim is to develop a novel non-invasive electrical mapping tool, which will allow detection of key electrical signatures related directly to the substrates responsible for lethal arrhythmias, in order to understand the mechanisms underlying SCD in patients with structurally normal hearts. The unique approach proposed to achieve this objective will consist of: (1) defining the key electrical signatures specific to arrhythmogenic substrates using animal models designed to reflect previously identified clinical phenotypes and human donor hearts from SCD survivors; (2) developing and validating ECGI and post-processing techniques to identify these critical signatures using both experimental and clinical data; (3) Apply the noninvasive tools developed in a clinical setting to identify the mechanisms underlying arrhythmogenic risk in SCD survivors This project will constitute a new paradigm in clinical cardiac investigations and could allow a major breakthrough in the prevention of arrhythmic deaths in the world.

Schizophrenia is a severe psychiatric disease, that affects more than 23 million persons in the world and close to 600 000 in France. That represents around 1% of the global population, with men being more affected than women. Schizophrenia is characterized by psychopathological, cognitive, and neurobiological abnormalities, and deficiency in perception, emotion and social behaviour. The frequency, intensity and type of symptoms are highly variable among individuals, resulting in great heterogeneity in clinical presentation. This pathology is also associated to metabolic disturbances, including abnormalities in lipid levels both in the brain and peripheral tissues. Most schizophrenia treatments employ the use of antipsychotics, whose main therapeutic target is the D2R, a G-protein coupled receptor (GPCR). Drug efficacy is still highly variable between patients, due in part to the fact that antipsychotics also target other GPCRs both in monomeric and heteromeric forms. Due to recent data on the potential allosteric modulation of the D2R and other GPCRs activity by lipids, and their impact in receptor heteromerization, the “membrane hypothesis of schizophrenia” may be explained by changes in intracellular signalling as a consequence of lipid-triggered D2R heterodimerization. Some studies suggest an impact of polyunsaturated fatty acids (PUFA) and cholesterol on the heterodimerization of D2R with adenosine 2A receptor. These two receptors are at their highest expression level in the brain structures involved in schizophrenia physiopathology. Plus, a decrease in the formation of these heteromers has been observed post-mortem in schizophrenic patients, and in animal schizophrenia model, compared to healthy subjects. This heteromer reduction may be correlated with the altered lipid profiles also observed post-mortem in schizophrenic patients, but this hypothesis has never been validated. SCHIZOLIP aims to investigate the role of membrane composition on D2R heteromerization and its impact on antipsychotic activity. To do so, studies on HEK293 cells and reconstituted model membranes will be performed. Various biophysical methods will be carried out: Forster Resonance Energy Transfer (FRET), Proximity Ligation Assay (PLA), Plasmon Waveguide Resonance (PWR), Microscale Thermophoresis (MST). In fine, this project will allow a better understanding of the mechanisms of antipsychotics, namely the parameters that modulate their pharmacological activity, therefore opening new perspectives for the development of innovative therapeutic strategies.

AI by UBx is a proposal of the University of Bordeaux built conjointly with the center Inria Bordeaux-Aquitaine to accelerate the ongoing development of Bordeaux as an AI reference center thanks to a proactive policy to attract national and international talents in innovative Ph.D tracks. The overall objective of the University of Bordeaux is to double the number of Ph.D graduating in AI by 2025. This relies on a three-fold approach: 1) increasing Ph.D financing opportunities (including public and industrial funding), 2) attracting new profiles by granting them access to a high-end multidisciplinary environment of research, innovation and entrepreneurship, 3) strengthening international partnerships with leading universities sharing the same vision of AI for Good. Why Bordeaux? Though being relatively discrete, Bordeaux is today a key digital innovation hub in France and in Europe, and home of a vibrant digital ecosystem. Birthplace of a swarm of 29 strong innovative AI start-ups coordinated and propelled by the Smart4D and HealthTIC clusters, Bordeaux hosts a larger community of AI innovators than other 3IA hubs such as Toulouse, Grenoble and Nice, as revealed by a 2019 BPI France cartography of AI start-ups. A strong innovator itself, the University of Bordeaux is producing, together with national research institutions embedded on its campus (namely Inria, CNRS, Inserm, Inra or CEA) and industrial partners, original research underpinning innovation. The University of Bordeaux is ranked 24th in Reuters’ 100 most innovative European universities (2019), thus reaching 5th place among the leading French universities. From the launch of its AI strategy in 2017, 39 key industrial players have joined the dynamics of the University of Bordeaux, covering all the sectors targeted by the French Plan for AI (Plan Villani). This includes Huawei and Betclic, 16 partnerships for health, including Maincare, Roche, Sanofi, Canon, IPSEN, IQVIA, CEVA santé animale; 3 partnerships for transportation, including large manufacturers such as PSA and Michelin; 4 partnerships addressing security and defense needs with Thales, Dassault, Ariane Group and Safran Tech; 5 partnerships addressing environmental issues with mid-size companies and policy-makers such as Bordeaux Metropole. A regional research and innovation network is also currently on build under the Conseil Regional de Nouvelle-Aquitaine impulsion and Inria coordination. AI by UBx is supported by a group of 40+ local researchers involved in the dynamics and covering a large range of disciplines from computer sciences and mathematics, to physics, medical sciences, epidemiology, neurosciences, economics or political sciences. This proposal is presented by the University Bordeaux and will be operated by its Graduate Research School (Collège des Écoles Doctorales), that is a component entirely focusing on supervising doctoral studies and graduating an average of 500 Ph.D per year, over 8 specialty doctoral schools. This proposal is targeting the financing of 15 full Ph.D grants, which corresponds to a quarter of the set objective to increase the volume of Ph.D students in AI for the duration of the program (5 years). The University of Bordeaux is organizing the RoboCup 2020 (the major international competition in robotics) in June in Bordeaux, and will capitalize on its leverage effect to increase the impact of the program and reinforce the attraction of students.

The plasticity of excitatory transmission in the brain is the cellular basis of learning. Its dysfunction is at the origin of neurodevelopmental and neurodegenerative diseases. Despite decades of work on the molecular mechanism behind synaptic plasticity, many of its aspects remain unknown. A major mechanism for modifying the efficiency of excitatory transmission is the regulation of the number and traffic of the AMPA subtype of glutamate receptors (AMPAR). Our project aims to solve one of the main mysteries of synaptic plasticity mediated by these receptors, which lies in the respective roles of the major subclasses of these receptors, and in particular those that are either permeable (CP-AMPAR) or impermeable (CI-AMPAR) to calcium. The presence of either of the receptor subtypes has very important functional consequences. This will be achieved through the development of innovative labeling and imaging tools, specific to receptor subpopulations, and combined with physiology. This project brings together cell biologists, structural biologists and imagers. We will break one of the essential barriers to our understanding of the mechanisms of synaptic plasticity. Our objectives are: 1. Leverage our development of monovalent ligands specific to GluA1-3 subunits to measure their specific organization. 2. Implement a synthetic antibody discovery stream to develop ligands of different AMPAR populations. 3. Image nanoscale organization and surface traffic of endogenous AMPAR subunits and subpopulations in cultured neurons and brain slices. 4. Study the recruitment of different AMPAR populations during short- and long-term plasticity in brain slices. Derive multivalent ligands to specifically disrupt synaptic function. To achieve these goals, we will combine existing and novel ligands with dynamic super-resolution imaging methods and physiology techniques. The collaboration between a structural biologist with unique capabilities to develop new ligands specific to molecular assemblies - and thus to specific complexes of AMPA receptors, and a cellular neurobiologist with expertise in nanoscopic imaging of AMPA receptors and their physiology, provides an exceptional opportunity to address a fundamental question about the mechanism of glutamatergic synapse function that remains unanswered due to the lack of adequate tools. Through the use and development of specific ligands for the extracellular domains of calcium-permeable and non-permeable AMPA receptors, combined with the use of STORM imaging, STED, expansion microscopy and single molecule tracking, we will for the first time determine the dynamic nanoscopic organization of these different receptor subtypes around and within the synapse. We will use both dissociated culture models of mouse hippocampal neurons and organotypic brain slices. We will be able to directly visualize the dynamic reorganizations of the different receptor subcategories during synaptic plasticity processes in order to better understand their functional roles. Finally, by exploiting our recent developments in functional modification of multivalent ligands using optogenetics and suicide substrates, we will be able to modify with light both the trafficking and the existence of specific receptor subcategories in order to analyze their respective contributions to different stages of synaptic plasticity and neuronal physiology.

Based on the neuroprotective properties of the cannabinoid 1 receptor (CB1R) activation, a number of studies have explored the role of this receptor in several neurodegenerative diseases, including Alzheimer’s disease (AD). AD is a neurodegenerative pathology leading to memory loss and cognitive decline, and represents the most common cause of dementia worldwide. This pathology, among others, is associated with alterations in the brain lipid content. More specifically, AD is associated to a significant decrease in docosahexaenoic acid (DHA), the main ?3-PUFA in the brain, and alterations in the expression or functionality of CB1R have also been described both in AD patients’ brains and animal models. Correlative data from the literature and recent data in the team strongly indicate a link between membrane PUFAs and CB1R signaling. The goal of the project is to investigate the multiple ways of how PUFAs influence the function, conformational states and localization of CB1R by using multi-disciplinary approaches including membrane and cellular models investigated through biochemical, fluorescence spectroscopy and cutting edge microscopic methods.