Thèse Signalisation Stat3 Non-Canonique au Niveau de l'Unité Glio-Vasculaire en Physiopathologie H/F

Université Paris-Saclay GS Life Sciences and Health

Les astrocytes remplissent des fonctions clés dans le cerveau grâce à des domaines subcellulaires spécialisés, tels que les prolongements périvasculaires (PvAP). Ces domaines pourraient agir comme capteurs et effecteurs précoces au cours des maladies, mais leur régulation précise reste mal connue. Les astrocytes subissent d'important changements en conditions physiopathologiques, sous l'influence notamment de la voie JAK-STAT3, qui module leur profil transcriptionnel. Or, STAT3 peut aussi activer des voies non canoniques, indépendantes de la transcription, permettant une signalisation locale et rapide dans les PvAP.
Nos données préliminaires montrent que l'ARNm et la protéine de STAT3 sont présents dans les PvAP, à des niveaux variables selon la région, l'âge ou l'activité. STAT3 présente aussi des modifications post-traductionnelles (PTM) non canoniques dans les PvAP. Nous émettons l'hypothèse que STAT3 agit dans les PvAP comme un capteur spatialement restreint et efficace de signaux physiopathologiques et pourrait alors moduler les fonctions gliovasculaires via une signalisation potentiellement non canonique.
Nous visons 1) à caractériser la localisation subcellulaire de STAT3 et ses potentielles voies de signalisation non canoniques dans les pieds astrocytaires, en utilisant des immunomarquages fluorescents sur des tranches de cerveau de souris dans un modèles de neuronflammation vasculaire aigu. Ensuite, grâce à une nouvelle souris knock-in spécifique aux astrocytes exprimant STAT3-GFP qui sera disponible à l'automne 2026, nous : 2) suivrons la localisation et la dynamique de STAT3 dans les PvAP par imagerie biphotonique chez des souris éveillées en physiopathologie ; 3) identifierons les partenaires d'interaction de STAT3 et ses PTM dans les PvAP par protéomique astrocyte-spécifique; 4) perturberons la signalisation de STAT3 dans les PvAP et évaluerons son impact sur le répertoire moléculaire, la structure et la fonction de l'unité gliovasculaire.
Ce projet vise à définir un rôle nouveau, spatialement localisé et potentiellement non canonique de STAT3 dans les astrocytes, ouvrant des perspectives pour cibler les altérations pathologiques de l'unité gliovasculaire.

Astrocytes perform a wide range of essential functions, including roles in blood-brain barrier maintenance, synaptogenesis, ion homeostasis, neurotransmitter buffering, and the secretion of neuroactive factors1. They have specialized compartments for specific and necessary functions. Perisynaptic astrocytic processes (PAPs) play a crucial role in the formation, maturation, and activity of synapses through various mechanisms such as gliotransmission, neurotransmitter reuptake, and potassium buffering2,3. In addition, astrocytic perivascular processes (PvAPs), or endfeet, are key components of the neurovascular unit. Indeed, endfeet play a crucial role in blood-brain barrier integrity and brain homeostasis4,5. In the endfeet, a specific molecular repertoire that allows regulation of the blood-brain barrier (BBB), endothelial transport, the recruitment of immune cells by endothelial cells, perivascular homeostasis, blood flow, and metabolite transfer to neurons. The specialization of the distal processes and their unique molecular repertoire is defined as the polarity of astrocytes6, which is disrupted in several neurological diseases such as Alzheimer's disease, leukodystrophies, and epilepsy.
In pathological conditions, astrocytes become reactive in a process called astrogliosis. This process is characterized by several biological changes, including upregulation of genes involved in inflammation with augmented secretion of cytokines, morphological changes such as hypertrophy of cell bodies and processes, and altered metabolic activity7. Moreover, we showed that the activation of the Janus Kinase 2-Signal Transducer and Activator of Transcription 3 (JAK2-STAT3) pathway is a master regulator of reactivity in astrocytes8. JAK-STAT3 is a ubiquitous signaling cascade, whose activation by cytokines, hormones or growth factors results in Tyrosine 705 (Tyr705) phosphorylation of STAT3, dimerization, nuclear translocation and transcriptional activation of target genes. But STAT3 can also have other cellular roles, through yet unclear non-canonical signaling modalities, which could contribute to the large diversity of astrocyte reactive state7. For example, in various cells outside the brain, STAT3 is known to interact with microtubules, mitochondria or autophagosomes and regulate cytoskeleton function, energy metabolism and autophagy respectively, independently of its transcriptional activity9,10. Non-canonical pathways, defined here as any STAT3 signaling action that does not involve transcriptional activation by Tyr705-phospho-STAT3 proteins, are totally overlooked for astrocyte physiology and reactive responses. In particular, whether STAT3 can locally transduce signals and regulate astrocyte functions within PvAP is unknown. Our preliminary results show that i) STAT3 polysomal mRNA and protein are present in PvAP along with some STAT3 regulators such as phosphatases, ii) STAT3 relative abundance in PvAP changes with brain region, aging or activity, iii) STAT3 presents post-translational modifications (PTM) typical of non-canonical signaling in gliovascular units.
Based on these promising data, we hypothesize that i) local STAT3 signaling in PvAP is a local detector of physiopathological signals, which may then trigger specific responses at the vascular interface, further impacting vascular, astrocyte and neuronal function. ii) STAT3 signaling in PvAP could operate through non-canonical signaling independently of nuclear transcription, allowing local and fast responses finely tuned to the stimulus.

We hypothesize that i) local STAT3 signaling in PvAP is a local detector of physiopathological signals, which may then trigger specific responses at the vascular interface, further impacting vascular, astrocyte and neuronal function. ii) STAT3 signaling in PvAP could operate through non-canonical signaling independently of nuclear transcription, allowing local and fast responses finely tuned to the stimulus.

We aim 1) to characterize STAT3 subcellular localization and potential non-canonical signaling in astrocytic endfeet using fluorescent immunostainings on mouse brain in an acute model of vascular inflammation. Then, thanks to a novel astrocyte-specific STAT3-GFP knock-in mouse which will be available at the fall 2026, we will: 2) track STAT3 localization and dynamics in PvAP with 2-photon imaging in awake mice in physiopathology ; 3) identify STAT3 interactors and PTMs in PvAP via astrocyte-specific proteomics; 4) interfere with PvAP-specific STAT3 signaling and assess its impact on the molecular repertoire, structure and function of the gliovascular unit.

Aim 1: We will assess the subcellular localization of STAT3 in astrocytes in normal and pathological conditions. We will use the well characterized model of acute intraperitoneal (ip) injection of lipopolysaccharide (LPS, 5 mg/kg, at 1h to 72h post injection), which mimics sepsis and induces significant changes in endothelial cells and astrocytes in the prefrontal cortex (PFC)11, 12. On immunostained mouse brain sections, we will then quantify total STAT3 abundance in PvAP and the nucleus/endfeet ratio of STAT3 to assess its subcellular localization in the prefrontal cortex, hippocampus and striatum, as astrocytes are known to display significant region specificity. We will also measure Tyr705-P-STAT3 as a direct measure of STAT3 canonical signaling, and other PTM-specific STAT3 (Serine-727-Phosphorylated-STAT3 and Lysine-685-Acetylated-STAT3) as indicators of non-canonical signaling. This quantitative analysis will assess how a peripheral inflammatory stimulus impacts STAT3 signaling at the gliovascular interface.

Aim 2: We will use a new knock-in mouse line, bearing one allele of the murine Stat3 gene fused in C-terminal with Gfp, expressed in a Cre-dependent manner. After crossing with Aldh1L1-CreERT2 mice and 5d tamoxifen treatment, mice with a STAT3-GFP fusion protein only present in astrocytes (STAT3-GFPastro mice) will be generated. This STAT3-GFP fusion protein was shown to be fully functional13. This will allow direct live monitoring of STAT3 levels, localization and trafficking in astrocytes by 2 photon (2P) imaging (Aim 2) but also purification of astrocytic STAT3 by its GFP tag (Aim 3). This mouse is being generated at Phenomin-ICS and will be available by the end of 2026.
To monitor STAT3 trafficking in PvAP in pathological conditions, we will perform 2P imaging on the dorsal PFC of awake STAT-GFPastro mice implanted with a cranial window. We will quantify STAT3 total levels, PvAP/nucleus ratio, and trafficking in nucleus or PvAP assessed via recovery after photobleaching under basal conditions and following acute ip LPS injection. Mice will be monitored between 1h to 72h post-injection to cover early signaling events and longer lasting changes.

Aim 3: We will use STAT3-GFPastro mouse cortex to purify STAT3 and associated proteins by co-immunoprecipitation with a GFP antibody from 3 types of samples: whole astrocytes, astrocyte nuclei and gliovascular units (composed of PvAP attached to isolated brain vessels by the basal lamina, as developed by our collaborator M. Cohen-Salmon14). Samples will be collected in basal conditions and following LPS injection, at the most relevant time point identified in Aim 2. Proteomics analysis will identify STAT3 binding partners and PTM in whole astrocytes, nucleus or PvAP; in basal or LPS conditions. Identified partners will be compared with available datasets: a list of translated mRNA in cortical PvAP (M. Cohen-Salmon, unpublished) and lists from B. Diaz-Castro (UKDRI)12 and B. Khakh (UCLA)15 of PvAP proteins and receptor-ligand pairs with endothelial cells, including on LPS-treated mice. The 3-5 most interesting STAT3 partners in PvAPs (based on astrocyte expression, data mining validation, known function, antibody availability) will be validated by immunofluorescent confocal and super-resolution microscopy on brain sections from wild type (WT) and STAT3-GFPastro mice in basal and LPS conditions. We will use specific markers to identify preferential association with capillaries or large vessels. Depending on available antibodies, interactions between STAT3 and a candidate partner will be validated in situ by proximity ligation assays.

Aim 4: We will finally attempt to interfere with STAT3 PvAP signaling. If Aim3 identifies PvAP-specific STAT3 partners, regulators or PTM, we will target them to inhibit STAT3 local signaling. Specifically, we will knock-down an identified PvAP partner, by leveraging an astrocyte-specific CRISPR- and AAV-based method16. To target a PvAP-specific PTM, we will use the same approach to generate a knock-in point-mutated STAT3 on the amino acid subjected to this PTM (e.g. S>A mutation to prevent serine-phosphorylation). We will then assess how STAT3 signaling inhibition in PvAP impacts the i) PvAP molecular repertoire using translating ribosome affinity purification of isolated gliovascular units coupled to RNAseq17, ii) structural organization of the full gliovascular unit, with super-resolution and electron microscopy (EM)18, including the cytoskeleton, as STAT3 can bind to it19, iii) astrocyte metabolism, since PvAP are crucial sites of metabolic regulation. We will measure oxidation of 14C-labelled metabolic substrates into 14C02 by isolated astrocytes20 and assess respiration on isolated mitochondria with Seahorse. This will be done in WT mice in basal conditions and after LPS exposure. If Aim 3 fails to identify a PvAP-specific strategy, we will use a more global approach. Using astrocyte-specific AAV gene transfer, we will compare i) mice with AAV-shRNA-mediated STAT3 knock-down in astrocytes, ii) mice expressing a STAT3 dominant-negative mutant without its DNA binding domain (STAT3V463)21. This STAT3 mutant cannot activate nuclear transcription, resulting in blunted astrocyte STAT3 canonical signaling, and iii) WT mice with both canonical and non-canonical signaling. By comparing these 3 groups using the same metrics, we will identify processes at the glio-vascular unit that are regulated by STAT3 through canonical or non-canonical signaling.