Thèse Stress Mécanique et Reprogrammation Transcriptionnelle Rôle de la Relocalisation des Septines dans les Maladies Hépatiques avec une Surcharge Lipidique H/F - Doctorat.Gouv.Fr

Établissement : Université Paris-Saclay GS Santé et médicaments École doctorale : Innovation thérapeutique : du fondamental à l'appliqué Laboratoire de recherche : Physiopathogenèse et traitement des maladies du foie Direction de la thèse : Christian POUS ORCID 0000000225027854 Début de la thèse : 2026-10-01 Date limite de candidature : 2026-04-22T23:59:59 Laccumulation de gouttelettes lipidiques dans le foie provoque une stéatose susceptible d'évoluer vers une stéatohépatite avec installation de fibrose, puis vers une cirrhose et, in fine, vers un carcinome hépatocellulaire. Ces pathologies hétérogènes ont un fort impact en matière de santé publique, mais leurs mécanismes cellulaires et moléculaires sont encore imparfaitement compris. Par ailleurs, une surcharge en gouttelettes lipidiques est observée dans divers cancers agressifs, où elle favorise l'apport énergétique, l'adaptation au stress et/ou la chimiorésistance. Nous disposons de données suggérant que les gouttelettes lipidiques dans le cytoplasme perturbent l'organisation du cytosquelette et altèrent la mécanotransduction. Ce processus pourrait constituer une réponse précoce à divers stress, induisant une reprogrammation transcriptionnelle, des changements d'état/d'identité cellulaire ainsi que des modifications du microenvironnement, favorisant la progression et l'agressivité pathologique.
Plus précisément, nous avons révélé des liens entre la perte de mécanotransduction, la relocalisation des septines de l'actine vers les microtubules induisant alors la perte des fibres de stress et l'émergence d'une résistance aux taxanes dans des hépatocytes. Nous proposons d'explorer ces acteurs, ainsi que leurs effets transcriptionnels dans un contexte de surcharge lipidique, en utilisant des lignées de cellules hépatiques humaines et des modèles murins de stéatose. Nous étudierons aussi ce mécanisme dans des hépatocytes soumis à une compression externe mimant la fibrose. Réciproquement, nous évaluerons l'impact des remaniements du cytosquelette et de la mécanostransduction sur la taille et le devenir des gouttelettes lipidiques, ainsi que sur les propriétés de la matrice extracellulaire et des fibroblastes voisins.
Ces études fondamentales visent à identifier de nouvelles cibles thérapeutiques potentielles et les résultats obtenus pourraient être transposables à d'autres pathologies, notamment certains cancers agressifs présentant aussi une accumulation intracellulaire de lipides. Our laboratory is located at the Faculty of Pharmacy of the Paris-Saclay University (Moissan building) and is part of the Inserm UMR-S-1193 unit (Pathophysiology and Treatment of Liver Diseases, HEPAREG). We are interested in fundamental aspects of cellular adaptation to stress involving the cytoskeleton (microtubules, septins, actin microfilaments and intermediate filaments) and/or the dynamics of membrane organelles (mitochondria, Golgi apparatus, autophagosomes) (1-4). In connection with liver diseases, our group studies the adaptative mechanisms involved during the hypothermic preservation of liver grafts, the stress response involving all components of the cytoskeleton.Among the components of the cytoskeleton, this project focuses more specifically on septins. These are filament-forming GTPases that delineate territories of precise curvature on cell membranes but also associate with certain cytoskeleton elements (actin stress fibers, and in some cases microtubules). There, they may recruit specific effectors, acting as scaffolds or may prevent molecule diffusion, functioning as diffusion barriers. The laboratory has shown that taxane resistance in cancer cells can be induced by a rapid and drastic cytoskeletal rearrangement in which septins relocalize from actin to microtubules (5-7). Remarkably, this remodeling is accompanied by a weakening of actin stress fibers and their redistribution to the cell periphery, along with reduced intracellular tension (as measured by traction force microscopy, TFM) and increased elasticity of the cell cortex (as measured by atomic force microscopy, AFM) (unpublished data). Our preliminary data suggest that this rearrangement, along with the observed concomitant reduction of nuclear translocation of the transcription co-activator YAP, also arises from a decrease in the stiffness of the extracellular matrix (ECM) or from the accumulation of oleic acid lipid droplets. In the latter case, we believe that lipid droplets lead to a cytoplasmic crowding-related mechanical stress that may act as Taxol® stress or ECM stiffness reduction, parallel to their role in cell energy supply or in the sequestration of lipophilic chemotherapies.

While lipid droplets appear capable of recruiting septins to their cytoplasmic face during HCV viral infection (8), their intrahepatocyte accumulation induces a loss of actomyosin stress fibers, cytoplasmic retention of YAP, and chromatin condensation (9-11 and our unpublished data), suggesting an alteration in mechanotransduction. We will test whether, like Taxol®-induced stress, lipid droplet overload stress alters intracellular tension, cell cortex elasticity and induces a shift in mechanotransduction toward stemness, mesenchymal, quiescence or senescence states rather than normal epithelial traits (Axis 1.1), potentially contributing to disease progression.
Furthermore, it has been proposed that a reduction in intracellular tension in cancer cells can result from compression exerted by a capsule of contractile cancer-associated fibroblasts (CAFs) (12) through an actomyosin- and septin-dependent mechanism (13). We hypothesize that in the liver, by analogy, the activation of portal fibroblasts and/or stellate cells (pericytes) into myofibroblasts (a type of mesenchymal stem cell involved in liver fibrosis) could lead to compression of hepatocytes, thereby triggering a reduction in their internal tension (similar to that observed in cells grown on substrates with low ECM stiffness). We will therefore explore whether cocultures (myofibroblasts-hepatocytes) could mimic this phenomenom and trigger septin relocalization to microtubules, further amplifying the progression of steatosis into hepatic fibrosis by altering mechanotransduction and the associated transcriptional response (Axis 1.2). Conversely, we will also explore how this cytoskeletal rearrangement of lipid droplet-filled hepatocytes influences the neighbor cells (transdifferentiation of hepatic stellate cells into a pro-contractile and pro-fibrotic phenotype) and the organization of the ECM (collagen fiber alignment, level of expression of main matrix proteins) (Axis 1.3).

Septins are required for lipid droplet biogenesis, and the overexpression of the Sept9i1 isoform increases their size and perinuclear localization (8, 14-15). Actin filaments and microtubules participate in droplet dissociation and transport, respectively (16-17). We will therefore test whether the relocalization of septins from actin stress fibers onto microtubules affects lipid droplet transport, metabolism, and/or fusion/fission dynamics (Axis 2).

We will also explore molecular and pharmacological approaches to maintain septins associated with actin stress fibers in stressed epithelial cells (Axis 3). We have previously demonstrated that septins co-align with actin stress fibers through local activation of Cdc42 and its downstream effectors Borg2/3, thereby contributing to stress fiber stabilization (6). To prevent septin subcellular relocalization under lipid overload condition, we will perform transient transfection of plasmids encoding constitutively active Cdc42 or Borg2(6). As observed for Taxol®-treated cells, this should restore nuclear YAP accumulation. Jasplakinolide will be used to stabilize actin filaments, while forchlorfenuron (FCF) will modulate septin filamentation.

Finally the liver organ is organized into lobules consisting of portal triads, hepatocytes arranged in linear cords between a capillary network, and a central vein. This zonation corresponds to different hepatocyte functions, but also to distinct mechanical properties. Indeed, in the periportal zone (available O2 and nutrients, lipid metabolism), YAP is nuclear, whereas in the centrilobular zone (scarce O2 and nutrients, favored glycolysis, xenobiotic catabolism), YAP is cytoplasmic. This cytoplasmic retention of YAP is due to the destabilization of actin filaments by the regulatory protein CapZ in this zone (18). We will therefore investigate whether our molecular players respond differently in these two zones using control and steatotic mouse models (Axis 4). Steatosis, also known as MASLD (Metabolic dysfunction-Associated Steatotic Liver Disease), refers to the accumulation of triglycerides in the liver. If left untreated, it can evolve into a more severe fibrotic and inflammatory condition called MASH (Metabolic dysfunction-Associated SteatoHepatitis) and may eventually lead to cirrhosis and even to cancer. Steatosis affects 16% of the french population and 30-40% of the world population. Lipid droplets accumulated in the cytoplasm of hepatocytes are sources of biochemical but also of mechanical stress. We will focus on the latter aspect where lipid droplet-mediated cytoplasmic crowding leads to a profound reorganization of the cytoskeleton. Indeed, we have preliminary data showing that upon lipid overload, septin filaments very early relocalize from actin stress fibers to microtubules, leading to microtubule stabilization and thinning of stress fibers that accumulate at the cell cortex. Actin stress fibers are essential for mechanotransduction, the process by which extracellular and intracellular mechanical signals are integrated and transmitted to elicit an adapted transcriptional response. Such transduction is mediated in part by the transcriptional coactivators YAP/TAZ and the transcription factor MRTF. Interestingly, this early septin relocalization paralleled by stress fiber alteration was also observed upon Taxol® chemotherapy treatment or growth on soft extracellular matrix (ECM). Across all three conditions, the same cytoskeleton rearrangement was also associated with decreased nuclear translocation of YAP, indicating defects in mechanotransduction. Moreover, we identified a set of eight genes that are early upregulated both under Taxol® treatment and upon cell culture on a soft ECM. These genes may therefore represent a transcriptomic signature of septin relocalization-mediated alterations in mechanotransduction, whenever cells are challenged by internal or external stressors, including by therapeutic drugs.

Here, focusing on lipid overload in hepatocytes, we aim to investigate whether this early subcellular relocalization of septin filaments, associated with mechanotransduction alteration, could trigger the specific transcriptional signature activation observed following Taxol® treatment and growth on soft ECM. We further propose that this response may represent an early adaptive event contributing to disease progression and therapy resistance through the acquisition of new cellular traits and alterations of the cellular microenvironment.

In this purpose, we will investigate how the lipid overload-mediated changes in the interplay between cytoskeletal components influence:
1) Cell behavior and fate from both biological and mechanical perspectives.
1.1) Does septin relocalization drive a change in cell state/identity toward fibrosis, aggressive and chemoresistant traits through a transcriptional signature common to what is observed upon Taxol® treatment and cell culture on a soft matrix?
1.2) How does septin relocalization change force transmission along the ECM-cytoskeleton-nucleus axis?
1.3) What is in return the impact on the ECM and on neighbor fibroblast/pericyte cells?
2) How does this cytoskeleton remodeling impact the characteristics of lipid droplets?
3) Can septin binding to stress fibers be kept to restrain mechanotransduction defects and deleterious cell adaptation to inner mechanical stress (lipid overload)?
4) Are the alterations of the septin-mechanotransduction axis recovered in vivo in mouse models of steatosis and steatohepatitis? This project relies on classical cell biology (immunofluorescence, video microscopy, high-resolution confocal microscopy [INFINITY STED-3D], light-sheet microscopy), biochemistry (Western blot) and biophysical (traction force microscopy (TFM), atomic force microscopy (AFM)) techniques that we will perform on the immortalized hepatocyte cell line HHL16 (+/- oleate). We have the molecular tools to visualize and modify the cytoskeleton and mechanotransduction.
In collaboration with the Moissan imaging platform (MIPSIT, Dr Séverine DOMENICHINI), we will perform high-resolution imaging of the subcellular organization (3D superresolution microscopy using adaptive optics / STED microscopy; Organ clearing followed by light sheet microscopy) and ECM network organization (reflection microscopy). With Drs Clément CAMPILLO and Guillaume LAMOUR (University of Evry Paris-Saclay, LAMBE, UMR8587) we will perform cell biomechanical measurements (AFM microscopy).
The 2D cocultures (epithelial cells - fibroblasts/CAFs or hepatic stellate cells HSCs/myofibroblasts...) will be generated with the expertise and material of Drs Daniela MATIC VIGNJEVIC (Curie Institute) and Carlos PEREZ-GONZALEZ (starting his group at Gustave Roussy from September 2026), who both developed this technique (12) and agreed to support us in its implementation in our lab for this project.
For pathophysiological studies involving primary liver cells and steatotic murine livers (induced by a High-Fat Diet HFD or a Methionine Choline deficient Diet MCD), we will collaborate with our team leader, Dr Thierry TORDJMANN, and our colleague, Dr Gregory MERLET (Moissan building, INSERM 1193, HEPAREG, team joined in January 2026). Liver tissues will be analyzed by histology, immunofluorescence, and immuno-histochemistry to examine lipid accumulation, and cytoskeleton organization, nuclear morphology, and YAP localization. Expression of the pertinent signature genes will be evaluated by RT qPCR and, when feasible, by Western blot or tissue staining. Spatial analysis across liver architecture as well as tissue clearing approaches to perform in-depth imaging of thick samples will also be considered.

La candidate, Ziyana BHARWANI, est inscrite au Master international D2HP de Paris-Saclay et effectue son stage de M2 dans notre laboratoire. Elle est motivée, dynamique, rigoureuse, avide d'apprendre et s'est intégrée rapidement dans notre équipe de recherche. Son niveau d'anglais et son expérience internationale sont un plus pour le projet.