Thèse Coût Énergétique de la Plasticité Synaptique H/F - Doctorat.Gouv.Fr

Établissement : École supérieure de physique et de chimie industrielles de la Ville de Paris - PSL (ESPCI Paris - PSL) École doctorale : Sciences du Vivant Laboratoire de recherche : Plasticité du Cerveau Direction de la thèse : Pierre-Yves PLACAIS ORCID 0000000184264465 Début de la thèse : 2026-10-01 Date limite de candidature : 2026-08-31T23:59:59 Le « sauvetage énergétique cérébral » a récemment émergé comme un concept thérapeutique prometteur pour les troubles neurodégénératifs liés au vieillissement. Cela nécessite une compréhension approfondie des besoins énergétiques qui sous-tendent le fonctionnement du cerveau. Si les études sur la consommation énergétique cérébrale ont révélé le coût élevé de l'activité électrochimique des neurones, dû aux potentiels d'action et à la neurotransmission, les mécanismes de plasticité synaptique sont restés un angle mort du coût énergétique global du calcul cérébral. Le cerveau de la drosophile offre une possibilité unique d'extraire la relation entre performance et consommation d'énergie dans un circuit neuronal bien caractérisé, capable de plasticité, à la base de la mémoire associative olfactive à court terme (STM).Le conditionnement aversif classique conduit à la formation d'une mémoire (évitement appris d'une odeur), dont la composante la plus immédiatement formée est la STM, qui décroît en moins de deux heures. La STM est encodée dans les corps pédonculés (mushroom bodies, MBs), une structure cérébrale réunissant les cellules de Kenyon (KCs), qui reçoivent les informations du stimulus olfactif ; les neurones de sortie des corps pédonculés (MBONs), partenaires post-synaptiques des KCs déclenchant des comportements d'évitement ou d'approche ; et les neurones dopaminergiques (DANs), neurones qui véhiculent les signaux de renforcement et modulent les synapses KC-MBON. Les MB sont organisés en modules anatomiques fonctionnellement indépendants, caractérisés par un système synaptique tripartite local, mettant en jeu une dépression des synapses KC-MBON induite par la dopamine, mécanisme central de plasticité synaptique encodant la STM. De plus, la STM est soumise à un mécanisme dit « d'oubli actif » (active forgetting), médié par l'activité tonique post-apprentissage des mêmes DANs impliqués dans l'induction de la plasticité pendant l'apprentissage.Grâce à des tests comportementaux, à l'imagerie in vivo en microscopie biphotonique avec conditionnement in situ, couplée à l'utilisation de senseurs métaboliques de pointe, nos objectifs sont les suivants :- évaluer le coût énergétique de la plasticité lors de la formation d'une STM aversive dans les KCs et identifier sa source métabolique ;- étudier le coût énergétique de l'activité bimodale des neurones MP1 et estimer le ratio de consommation énergétique entre la persistance de la mémoire et l'oubli.Ce projet apportera des éléments pionniers concernant la quantification du coût énergétique de la plasticité synaptique, nécessaire pour une meilleure estimation de la consommation énergétique cérébrale, désormais considérée comme un aspect clé dans les recherches fondamentales et médicales. Introduction and state of the art: Neuronal computation involves energy consumption, under metabolic limitations which have shaped, throughout evolution, the cognitive mechanisms of living organisms1. Decline in cognitive function, e.g. occurring in neurodegenerative disorders or with aging, have been associated with bioenergetic deficits2. Unfortunately, while studies of brain budget have revealed a high cost of neurons electrochemical activity, linked to firing and neurotransmission3, mechanisms of synaptic plasticity have remained a dead angle of the overall energetic cost of brain computation. Drosophila melanogaster brain offers a unique possibility to extract relationship between performance and energy consumption in a well-characterised and genetically tractable neural circuit capable of plasticity. Here, we use the olfactory associative short-term memory (STM) paradigm to investigate the energetic cost of synaptic plasticity.In Drosophila, classical aversive conditioning4 (pairing the delivery of an odorant to electric shocks) leads to memory formation manifesting as learned odorant avoidance, among which the most immediately formed and best-characterized component is STM, decaying within two hours. The mushroom body (MB) is the center for olfactive memory processing. It reunites three major cell types: the Kenyon cells (KCs), cholinergic neurons receiving the olfactory stimulus information; the mushroom body output neurons (MBONs), post-synaptic partners of KCs which encode valence and triggers avoidance or approach behaviour5; and dopaminergic neurons (DANs), input neurons which convey reinforcement signals and locally modulate KC-MBON synapses (Fig.1). DANs and MBONs are anatomically paired, tiling KC axon bundles into parallel and functionally independent anatomical modules6 characterised by a local tripartite synaptic system. In particular, aversive STM is formed in the 1 compartment of the lobe7,5 (Fig.1), innervated by the DANs MP18, through an heterosynaptic plasticity mechanism leading to the depression of the KC-MBON synapse5,8 (Fig.1). Interestingly, aversive STM in KCs is also subjected to active forgetting, mediated by the tonic activity of the same MP1 neurons after learning9,10,11. Heterosynaptic plasticity, notably cAMP production from ATP, involves energy consumption. Whether specific ATP-producing metabolic process(es) sustains it remains unknown. The argininePhosphate-arginine (argP-arg) phosphagen buffer, homologous to the creatineP-creatine pathway in Vertebrates, is a promising candidate. It enables fast creation of ATP, mediated by the enzyme arginine-kinase (argK), consistent with a rapid need for energy during memory formation. Moreover, it was shown that genetic knock-down (KD) of argK in KCs reduced STM in a courtship conditioning assay13. The active erasure mechanism of learning-induced plasticity urges to investigate also energy consumption associated with activity in MP1 neurons as part of the overall energetic cost of STM. We will therefore characterise and quantify the cost of active forgetting. Objectives of the thesis: Through behavioural assays and in vivo subcellular metabolism imaging, the thesis objectives are thus:- to assess the energetic cost of plasticity upon aversive STM formation in KCs and identify its metabolic source - to investigate the energetic cost of MP1 neurons bimodal activity and estimate the energy consumption ratio of memory persistence and forgettingObjective 1: The energetic cost of synaptic plasticity underlying STM in the KCsA. We will assess energy consumption during STM formation in KCs by imaging flies in vivo under a two-photon (2P) microscope with in situ conditioning.1) The set up will be crafted with the odor delivery system14 already developed in the lab and the addition of a shock grid platform. The proper delivery of shocks and odorants will be confirmed by recording Ca2+ response of MP1 neurons (for shocks) and KCs (for odors) with GCaMP6s.2) To track ATP consumption of KCs over time, we will express in these neurons the most performant ATP sensor iATPsnfr215. To isolate the energy that is used for learning-induced synaptic plasticity specifically from the one used for sensory responses, we will compare the recordings upon paired to unpaired conditioning (where the same odor and electric shocks are delivered, but separate in time).3) To measure the variation law of energy consumption upon plasticity, we will perform progressively strengthened conditioning in situ under the microscope and image ATP consumption in KCs during STM formation. Similar variation curves will be obtained for cAMP responses in KCs as a function of conditioning intensity, with the cAMP sensor TEpacVV 16. Comparing ATP and cAMP variation laws will allow determining if cAMP synthesis is a major energy expense in plasticity.B. We want to identify the main source of ATP used by KCs during STM formation.1) We will test the effect of argK KD in aversive STM behavioural assays, using two RNAis expressed at adulthood only (using the inducible expression system Target like in previous work from the host lab17,18). If we do not observe any STM defect with argK KD, we will test non-canonical metabolic pathways, e.g. -oxidation, and simultaneous KD of prominent ones (achieved by generating lines carrying multiple RNAis against the relevant genes) to impair STM.2) We will confirm that the relevant KD condition impairs the learning-induced ATP variations in KCs.Objective 2: The energetic cost of MP1 neurons' activity mediating forgettingA. Two activity modes of MP1 neurons are relevant for STM: phasic (sensory response to punishment stimuli mediating learning) and tonic (ongoing activity without stimulus, mediating forgetting after learning). No study has addressed the bioenergetics of DANs activity so far.1) We will simultaneously image in vivo by dual-color 2P microscopy Ca2+ (using jRCaMP19) and ATP in MP1 neurons, during spontaneous activity and during conditioning.2) We will KD the distinct metabolic pathways in MP1 neurons and measure whether it affects either of the two modes of the neurons' activity. For this, we will co-express GCaMP6s and inducible RNAis targeting key proteins of each tested pathway.3) We will test if the KD of the metabolic pathways that have been identified as necessary for MP1 tonic and/or phasic activity also leads to corresponding aversive STM phenotypes in behavourial assays. B. We want to address if maintaining STM in KCs has a quantifiable energetic cost, and if it does, how it compares with the cost of forgetting, mainly mediated by tonic activity of MP1 neurons.1) We will estimate the cost of MP1-mediated active forgetting by comparing ATP evolution in those neurons after conditioning between controls and flies with disrupted MP1 neurons tonic activity. To inhibit MP1 neurons active forgetting, we will use the results from above, if we find a metabolic pathway fueling specifically MP1 neurons tonic activity, or perform gradual starvation of the flies, which has been shown to proportionally inhibit tonic activity of the neurons20. 2) To address if there is energy allocated in KCs to maintain synaptic plasticity, we will compare their ATP consumption immediately after learning between intact- or impaired-forgetting conditions (using the same methodology as above). Results from both sub-aims will allow estimating the cost ratio between maintenance of STM and forgetting.Overall, this project will be pioneer in characterising, both qualitatively and quantitively, the energetic cost of synaptic plasticity. This is necessary for a better estimation of brain energy consumption, which is now considered a key aspect in fundamental and medical studies

Une candidate est déjà pressentie