Jérôme Bürgi
Capillary Morphogenesis Gene 2 (CMG2) is a 55kDa single-pass transmembrane protein best described as the main Anthrax Toxin Receptor. However, CMG2 physiological role remains elusive and understanding better its in vivo function was the main objective of my thesis. My first aim was to investigate the prevalence of intrinsically disordered domains in the human transmembrane proteome. Using prediction tools, I found that 50% of transmembrane proteins had at least one domain of minimum 30 amino-acids predicted as intrinsically disordered (IDD), these proteins being mainly involved in cell signaling and adhesion. The majority of the IDDs were localized in the cytoplasm, and the intrinsic disorder containing proteins were significantly more phosphorylated and interacted on average with more proteins than fully ordered proteins, corroborating their role in signaling. CMG2 loss-of-function mutations is the known cause of a rare inherited genetic disorder called Hyaline Fibromatosis Syndrome (HFS). HFS patients develop subcutaneous nodules, gingival hypertrophy, contracture of the large joints, and in the most severe cases diarrhea and premature death. In a second part of my thesis, I was involved in understanding the molecular defects underlying missense mutations localized in the ectodomain and a highly conserved cytoplasmic domain of CMG2. We found that CMG2 was able to bind the actin cytoskeleton through Talin and Vinculin when free of its extracellular ligand. Interestingly, the HFS mutation p.C218R, altered the cytoskeleton release and led to a CMG2 protein able to bind both the extracellular ligand and the actin cytoskeleton. Strikingly, we observed that all but one cytoplasmic missense mutation hindered actin binding, a defect potentially relevant to HFS pathogenesis. Furthermore, we showed that the Src-like kinase family are involved in the release of Talin upon ligand binding and that CMG2 binding to its extracellular ligand is critical for the correct orientation of the mitotic spindle of zebrafish epiblast cells, which allows cell division along the Animal/Vegetal axis. Through the analysis of cmg2 knock-out mice, along with the study of HFS patient samples, we aimed at better defining CMG2 function. We showed that type VI Collagen was strongly accumulating in the cmg2 KO mice uterus tissue leading to failure of parturition, an accumulation that was also observed in HFS patient nodules. In addition, we observed that CMG2 was able to interact and potentiate MT1-MMP, and that type VI Collagen is a bona fide ligand for CMG2, its binding leading to the phosphorylation of the receptor. With the scope to understand the HFS patient subcutaneous nodules pathogenesis, we investigated the gene expression profile of fibroblasts derived from non-affected and affected tissues by RNA sequencing. Fibroblasts derived from affected tissues appeared to differentially express genes involved in the TGFBeta pathway and cytoskeleton, such as alpha smooth muscle actin, a marker of myofibroblast activation. We observed the presence of myofibroblasts in the patient nodules, and showed that CMG2 was limiting their activation process, loss-of-function mutations in CMG2 leading to an exacerbated differentiation. As we observed CMG2 interaction with the TGFBeta receptor 2, we postulated that CMG2 influence the TGFBeta pathway, a process that is altered in HFS patient and potentially lead to the formation of nodules.
EPFL2015
