Le syndrome de Marfan (MFS) est une maladie génétique rare. Les signes cliniques sont principalement squelettiques, oculaires et cardiovasculaires. Le premier gène identifié est FBN1. Une LSDB UMD créée en 1995 a permis de colliger les mutations identifiées chez les patients. D’autres gènes ont été identifiés comme causant des syndromes apparentés : FBN2, TGFBR1, TGFBR2, ACTA2, SMAD3, MYH11 et MYLK. Ma thèse avait pour but de mettre à jour ces différentes bases et de créer celles des nouveaux gènes. Les techniques de séquençage nouvelle génération dans la pratique clinique amènent des médecins non spécialistes à rapporter des variations secondaires dans ces gènes « actionable ». Il s’agit de la ressource la plus complète pour les cliniciens et les généticiens pour interpréter les variants associés au MFS et ses pathologies associées pour les variants primaires et secondaires.La bicuspidie de la valve aortique (2 feuillets au lieu de 3) est la malformation cardiovasculaire la plus fréquente touchant 0,6 à 2 % de la population. Ma thèse avait pour but d’identifier de nouveaux gènes impliqués dans la BAV. Grâce à une cohorte de 200 patients, le gène HOXA1, un facteur de transcription, a pu être examiné de manière plus approfondie. Alors qu’il contient une répétition de 10 histidines, les individus hétérozygotes mutés en présentent 11. Mes études ont permis de démontrer l’imputabilité de cette mutation. Dans un second temps, un séquençage d’exomes entiers a permis de mettre en évidence de nouveaux variants prédits pathogènes. Ces données permettront de mieux comprendre le rôle physiologique d’HOXA1 et des nouveaux gènes candidats dans la pathogénèse de la BAV chez l’humain. / Marfan Syndrome is a rare genetic disorder. Clinical signs are mainly skeletal, ocular and cardiovascular. The first gene identified is FBN1. A LSDB UMD was created in 1995 to colligate all the mutations identified in. Thereafter, several others genes were identified involved in related disorders : FBN2, TGFBR1 and TGFBR2, ACTA2, SMAD3, MYH11 and MYLK. The objectives of my PhD were to update these databases and to create new ones for genes. Next generation sequencing in clinical practice leads non-specialized doctors to report pathogenic secondary variants in “actionable” genes. Our databases are the only resources providing access to the full spectrum of known pathogenic mutations with checked and interpreted data from many reference diagnostic laboratories and research centers worldwide, and the most comprehensive resources for clinicians and geneticists to interpret variations linked to Marfan syndrome and related disorders not only primary but also secondary variants.Bicuspid Aortic Valve is the most common cardiovascular malformation affecting 0.6-2% of the population (2 leaflets instead of 3). Thanks to a cohort of 200 BAV patients, HOXA1 gene, a transcription factor, has been screened. While HOXA1 contains a string of 10 histidine repeats, these individuals are heterozygous for an 11 histidine repeat variant. My studies showed the imputability of this mutation. In a second phase, whole exome sequencing allow us to highlight new variants predicted pathogenic. Studies are still ongoing to confirm their imputability. These data contribute to our better understanding of the physiological of HOXA1 and new candidate genes in the pathogenesis of BAV disease in humans.

A Síndrome de Marfan (SMF) é a enfermidade hereditária mais comum dentre as que afetam o sistema conjuntivo, causada por mutações da glicoproteína fibrilina-1, o principal componente estrutural das microfibrilas elásticas da matriz extracelular. As manifestações fenotípicas da SMF são sistêmicas e acometem tipicamente os sistemas ocular, esquelético e cardiovascular, este uma importante causa de morbi-mortalidade. Entretanto, não está claro como a mutação induz a doença. Estudos anteriores sugerem anomalias morfológicas do retículo endoplasmático (RE) ou retenção intracelular da fibrilina-1 nos estágios avançados da SMF. Entretanto, a contribuição do enovelamento da fibrilina-1 mutada e do estresse do RE na fisiopatologia celular da SMF não é conhecida. Proteínas mal-enoveladas podem levar à retenção intracelular e/ou aumento da degradação através da via de degradação associada ao RE (ERAD), além da indução da resposta a proteínas mal-enoveladas (UPR), ambas com potencial contribuição à fisiopatologia de doenças, incluindo a SMF. Assim, estudamos em fibroblastos embrionários isolados de camundongos (MEFs) com SMF se a fibrilina-1 mutada é reconhecida pelo controle de qualidade do RE pelo seu mal- enovelamento e induz estresse do RE por sua retenção intracelular. Demonstramos que a mutação na fibrilina-1 per se não promoveu chaperonas marcadoras de UPR ou geração de oxidantes. Além disso, não levou a uma maior sensibilização das células à indução exógena de estresse do RE, nem promoveu maior morte celular após inibição do proteassoma. Além disso, não foi observada retenção intracelular da fibrilina-1 nas células SMF, e mesmo após inibição da via secretora ou indução de estresse do RE, a inibição da secreção da fibrilina-1 foi similar nos MEFs SMF e wild-type (WT). A dissulfeto isomerase proteica (PDI), uma importante chaperona redox do RE, interage com fibrilina-1, e seu silenciamento levou a um aumento na secreção da fibrilina-1 pelos MEFs WT, mas não SMF. Além disso, o silenciamento da PDI promoveu a desorganização da matriz extracelular depositada de fibrilina-1 nos MEFs WT, enquanto nos MEFs SMF, a desorganização basal da matriz não foi adicionalmente alterada. Em paralelo, investigações in vivo mostraram que o estresse do RE não é induzido em camundongos SMF com 1 ou 3 meses de idade, apesar de manifestações fenotípicas evidentes. Entretanto, concomitante à progressão da doença, detectamos a ocorrência de estresse do RE nas aortas ascendentes dos camundongos aos 6 meses. Esta detecção foi exclusiva desta região da aorta e não ocorreu em outros órgãos afetados ou não afetados pela SMF. Assim, a manifestação do fenótipo clássico da SMF não requer uma perda da homeostase do RE diretamente induzida pela fibrilina-1 mutada. Ao contrário, esta é capaz de evadir mecanismos de controle de qualidade mediados pela PDI, sendo secretada normalmente. Assim, esta evasão do controle de qualidade pela PDI é uma condição permissiva essencial para o fenótipo da SMF. Por outro lado, o estresse do RE é uma característica evolutiva do aneurisma da aorta ascendente na SMF concomitante ao agravamento do fenótipo neste tecido / Marfan syndrome (MFS) is the most common connective tissue hereditary disease, caused by mutations in the glycoprotein fibrillin-1, the main structural component of extracellular matrix elastic microfibrils. MFS phenotypic manifestations are systemic and typically involve the ocular, skeletal and cardiovascular systems, the latter a major cause of morbidity/mortality. However, how gene mutation induxes disease is yet unclear. Previous studies suggest endoplasmic reticulum (ER) morphological abnormalities or fibrillin-1 intracellular retention in advanced MFS stages. However, the contribution of mutated fibrillin-1 folding and ER stress to MFS cellular pathophysiology is unknown. Un/misfolded proteins may associate with their intracellular retention and/or increased degradation through ER-associated degradation (ERAD), in addition to inducing the unfolded protein response (UPR), both sharing potential contributions to disease pathophysiology, including MFS. Thus, we studied in embryonic fibroblasts (MEFs) isolated from WT and MFS mice, if mutated fibrillin-1 can be recognized by ER quality control as a misfolded protein, able to induce ER stress due to its intracellular retention. We showed that fibrillin-1 mutation by itself did not promote UPR chaperone markers or oxidant generation. Moreover, it did not sensitize cells to exogenous ER stress nor affected cell survival curves after proteasome inhibition. Furthermore, no intracellular retention of fibrillin-1 was observed in MFS cells, and even after secretory pathway inhibition or ER stress induction, fibrillin-1 secretion inhibition was similar in MFS and wild-type (WT) MEFs. Protein disulfide isomerase (PDI), an important ER redox chaperone, interacts with fibrillin-1 and its silencing induced an increased fibrillin-1 secretion in WT, but not MFS MEFs. Besides, PDI silencing promoted fibrillin-1 extracellular matrix disorganization in WT MEFs, whereas in MFS MEFs, the basal matrix disorganization was not further modified. Parallel in vivo evaluations demonstrated that ER stress is also not induced in 1 and 3 month-old mice MFS, despite evident phenotypical manifestations. However, concomitant to accelerated disease progression at 6 months, ER stress was detectable in ascendant aorta, but not in other disease-affected or unaffected organs. Thus, classic MFS phenotype manifestations do not require loss of ER homeostasis directly induced by mutated fibrillin-1. Contrarily, the latter can evade a PDI-mediated quality control mechanism to be normally secreted. Therefore, evading such PDI-mediated quality control is an essential permissive condition for enabling the MFS phenotype. On the other hand, ER stress is an evolutive feature of MFS ascendant aorta aneurysm concomitant to phenotype progression in this tissue

Many studies have demonstrated a connection between the fibrillin matrix and TGFβ signalling, but at present the mechanistic basis for this link is unclear. An interaction between the C-terminus of Latent TGFβ Binding Protein 1 (LTBP1) and the N-terminus of fibrillin1 has previously been identified, and may have the potential to directly link the fibrillin matrix to TGFβ signalling. To investigate the structural basis for this interaction, several multi-domain fragments of fibrillin1 and LTBP1 were expressed prokaryotically and refolded in vitro. After initial characterisation to confirm folding, the structure, dynamics, and interdomain interactions of these fragments were investigated in more detail using NMR techniques. Domains in both LTBP1 and fibrillin1 appear to demonstrate folds consistent with homologous structures, and while the LTBP1 C-terminal cbEGF14-TB3-EGF3-cbEGF15 region contains many flexible linkers and few interdomain interactions, the fibrillin1 EGF2-EGF3-hyb1-cbEGF1 region appears rigid, with interfaces forming between all domains present. SPR studies were used to demonstrate binding between distinct LTBP1 and fibrillin fragments, suggesting interactions between multiple domains are involved in the LTBP1-fibrillin1 interaction. The binding sites involved were then mapped to specific residues using HSQC titration studies, and structural models for the LTBP1-fibrillin1 interaction were generated based on these data. Predictions from these models were used to target residues for site-directed mutagenesis, based on their potential involvement in salt bridges, and when certain residues were replaced with those of opposite charge, reductions in binding could be seen in the SPR assay. These key residues were consistent with a particular model of the LTBP1-fibrillin1 interaction, as derived from the HSQC titration data. The conservation of potential binding site residues through deuterostome evolution also supports an important biological role for the LTBP-fibrillin interaction.