This "COSMIT" project focuses on a rare disease, the Costello syndrome (CS), the underlying mechanisms of which are not understood and for which no treatment is available. This rare disease is revealed in the first months of life and is characterized by postnatal growth retardation, thick lines, intellectual deficit and skin abnormalities, with muscle and heart alterations. Our aim is to elucidate the biological mechanisms responsible for this syndrome caused by a mutation (G12S) in the HRAS gene. Other less frequent mutations in this gene are the cause of rare diseases grouped under the term "rasopathies" as well as many cancers (HRASG12V). The fact that the French Association of Costello Syndrome and CFC based in Bordeaux has funded a novel mouse model of that disease (HRASG12S generated by the Mouse Clinical Institute) makes the context of this project particularly interesting, competitive and innovative. Indeed, we propose here to explore several facets of the pathophysiology of CS using this mice, but also skin fibroblasts obtained from patients, induced-pluripotent stem cells (iPSCs) and neural-crest derivatives. While current effort on the study of CS pathophysiology focus on the HRAS-related ERK signaling pathway, we discovered that HRAS constitutive activation in human fibroblasts inhibits the activity of both LKB1-kinase and its main target, the AMP-activated protein kinase (AMPK). This triggers both a metabolic reprogramming and an alteration of focal adhesion and cytoskeleton potientally involved in CS patophysiology notably during neural crest differentiation. Interestingly, the AMPK inhibition induced by HRAS can be reversed by using a pharmacological activator of AMPK, suggesting that a therapeutic approach could be derived from our observations. In addition, we explored the mice and found an alteration of muscle strength, locomotor activity, heart function, macrocephalia and hypertension. Further imaging analysis that are still ongoing confirmed the existence of potential cranio-facial changes in the mice that need to be further characterized. In addition, a transgenic mouse model of the cancer-associated HRASG12V mutation unexpectedly triggered the CS in mice, raising a fundamental question on the possible dose-effect of HRAS constitutive activation in Rasopathies, since the G12V mutation triggers a stronger biochemical activity of HRAS as compared to the G12S mutation. Therefore, we will compare the effect of those two mutations of HRAS on cell models, looking more closely at the impact on LKB1-AMPK and ERK signaling. We will perform an indepth analysis of HRASG12S mice through metabolic, cardiac and cognitive phenotyping and embryogenesis, with an emphasis on neural crest cells development. Since drugs are available to stimulate AMPK in human and mice, as resveratrol or AICAR, we will test their therapeutic effect on the mice model of CS. Inhibitors of RAS constitutive activation will also be tested as the Zopra cocktail developped by team 2. Lastly, we will evaluate the alteration of senescence in those mice and cell models of CS, since AMPK could regulate progerin levels via SRSF1 and because alteration in working memory was found in the HRASG12S mice and CS patients, a feature previously reported in patients with accelerated aging. Our project thus fits perfectly in the challenge "Health and Wellbeing" because we are studying the mechanisms of a rare disease on an original mouse model and patient-derived cell models, and we propose to evaluate two therapeutic approaches. We regrouped a consortium of internationally recognized experts in energy metabolism (R. Rossignol), mice generation and phenotypic characterization (Y. Herault) and rare diseases pathophysiology and therapy (N. Levy), along with a patient association to achieve the general aim of this project and disseminate our findings.

The intestine is a critical interface with the environment. By feeding, animals get exposed to toxins or toxicants contaminating nutrients. We have discovered a novel stress response of Drosophila enterocytes: pore-forming toxin or xenobiotics such as caffeine, metallic ions, or paraquat trigger a limited extrusion of apical cytoplasm and damaged organelles, possibly along with the toxicant, thus yielding a homogenously thin intestinal epithelium. There is no increased enterocyte cell death in the process. The enterocytes then recover their original shape and volume in 6-9 hours by a noncell-autonomous process. This response protects enterocytes from occasional intoxications from microbial or environmental origin present in contaminated food. This protective response thus involves two distinct phases. First occurs a limited extrusion of cytoplasm that does not lyse it, the purge; next, a recovery phase takes place allowing the intestinal epithelium to regain its original thickness and morphology. We have identified the induction of several secreted peptide-coding genes named "what else" by a pore-forming toxin, hemolysin, secreted by the entomopathogenic bacterium Serratia marcescens. Their expression requires a cyclin of rather undefined function, which does not involve a cell cycle role in mature enterocytes. Interestingly, what else genes ectopic expression compensates the cyclin mutant phenotype, i.e., an intestinal epithelium that remains thin for lack of recovery phase. This ectopic expression of what else genes also recapitulates another property of the system: a prior exposure to hemolysin or to some xenobiotics primes the intestinal epithelium against a secondary exposure to hemolysin, a Sm virulence factor. Thus, What else factors account for the noncell-autonomous properties of the recovery phase. We have identified through combined genetic and transcriptomics approaches about 100 genes potentially required for the recovery phase. A first aim of our project shall consist in studying how what else genes expression is regulated. Thus, we shall study on the one hand a putative transcription factor that binds to the cyclin and on the other hand a signal transduction pathway that may be initiated by mitochondria, which are strongly affected prior to the cytoplasmic purge. We shall determine in a second objective how a developmental regulator makes enterocytes competent for the recovery phase induced by hemolysin or xenobiotics, either to initiate what else genes expression or further downstream in the enterocytes induced by these secreted peptides. A third goal will be to understand how recovery occurs from a metabolic standpoint. We hypothesize that there is an inversion of normal metabolic fluxes, from organs and tissues back to the intestine. Once this point will have been demonstrated, we shall focus on an amino-acid transporter that may also regulate a major cellular growth pathway. Our preliminary data show that human enterocyte cell cultures that form epithelia, as well as the mouse intestinal epithelium in vivo, also undergoes epithelial thinning and subsequent rapid recovery upon exposure to Sm hemolysin. Thus, enterocyte purge coupled to fast recovery may be evolutionarily conserved, a hypothesis we shall attempt to validate, first in cultured epithelia and then in vivo. Next, we shall test whether exposure to xenobiotics also triggers the purge of mammalian enterocytes in culture and in vivo.

The clinical spectrum of cognitive disorders (CD) varies widely from Intellectual Disability (ID) to Autism Spectrum Disorder (ASD) and is estimated to affect 1-3% of the population. Genetic evidence indicates that one major functional group of CD-related proteins corresponds to proteins that are enriched at synaptic compartments, defining the concept of synaptopathies. While many studies have focused on the involvement of neurons in the pathology, the multi-partite synapse questions the contribution of the glia in CD, thus urging the study of the astrocyte-to-neuron bidirectional communication in cellular and animal models. This project is based on the functional characterization of the interactions between astroglia and neurons in the context of synaptopathies. We will focus our studies on Oligophrenin1 gene, which mutations are associated with ID, ASD or schizophrenia. The Ophn1 gene encodes a RhoGAP protein that is expressed not only in neurons, but also in astrocytes during development and in adult. At the synapse, its neuronal functions have been largely reported, but its roles in astroglial cells are still unexplored. We have shown in a mouse model of the pathology that ophn1 loss is responsible for endocytosis defects in astrocytes together with hyperactivation of RhoA/ROCK and PKA pathways. Preliminary data suggest that loss of ophn1 function in glia also contributes to the dendritic spine phenotype observed in neurons. The cellular and molecular defects in astrocytes deserve a full characterization of the consequences of ophn1 inactivation in these cells. In this context, we will address the two following scientific questions: 1. What are the cell autonomous and non-cell autonomous consequences of oligophrenin1 loss of function in astrocytes? (i) To decipher its function in astrocytes, we will establish primary astroglial cultures from constitutive KO brain and monitor their morphology, adhesion capacities and vesicular trafficking. Given the well-known function of ophn1 on synapses, we will investigate the synaptic densities on co-culture systems of WT neurons on KO astrocytes and vice versa. Since we have access to iPS cells developed from cutaneous fibroblasts of OPHN1 mutated ID patients, we will differentiate them into glia or neurons to allow us to extend and confirm our results to humans. (ii) Beside these in vitro studies, conditional floxed alleles of ophn1 will be crossed to neuronal or astroglial Cre driver lines. Double transgenic animals (Flox;Cre) will be assessed in behavioral tests and we will compare their phenotypes to previously reported constitutive models. (iii) In parallel, we will perform stereotaxic injections of Cre expressing lentiviral vectors in floxed animals in order to study in only few cells the consequences of ophn1 inactivation. We will analyze astroglia cells at the molecular level (markers), cellular level (morphology) and functional level (electrophysiological recording and Ca2+ imaging). 2. Does restoring or improving astroglial functions can rescue some of the ID-linked phenotype observed in mice? We will tackle this question using two different approaches either genetic or pharmacologic. Since Ophn1 has been shown to negatively regulate the RhoA/ROCK signaling pathway, we will test whether reducing RhoA levels can restore ophn1 function in astrocytes from the conditional mouse model. Alternatively, we will treat conditional KO mice with drugs previously identified as beneficial in in vitro or in vivo systems. Our consortium aims to decipher the astroglial function of ophn1 a molecule involved in synaptopathy and to highlight the astrocyte contribution to the pathology. Our goal is ambitious because we have a comprehensive and multi-disciplinary strategy to explore Ophn1 mouse ID models. Through this comprehensive study from cell to the animal, we also hope to uncover new pharmacological approaches targeting the glial cells to improve the learning capacities of ID patients.