There are different conformational states of proteins characterized by different Gibbs free energy levels, manifested in folding-unfolding dynamics, for example. Recently, a set of protein states, which require relatively small amount of folding energies, emerged as subjects of intensive research, and proteins or regions characterized by the presence of these states have been termed as ‘Intrinsically Disordered Proteins’ (IDP) and ‘Intrinsically Disordered Protein Regions’ (IDPR), respectively. Predisposition for intrinsic disorder of a query protein is encoded in its amino acid sequence and composition, and can be rather accurately predicted using several intrinsic disorder algorithms. Since pathology of many human diseases can be driven by proteins characterized by high intrinsic disorder scores, research on various disease-associated proteins is often started with the analysis of their intrinsic disorder propensities. In this work, I utilized computational approaches based on the concept of intrinsic disorder to address three health-related issues. To this end, I developed a novel computational platform for disorder-based drug discovery and applied this tool for finding inhibitors of the cancer-related MBD2-NuRD complex, utilized molecular dynamic simulations to explain the effects of mutations on the functionality of the X-linked protoporphyria-related protein ALAS, and used bioinformatics tools to examine the effects ofcardiomyopathy-related mutations in cardiac troponin. Since the complex between the Methyl-CpG-binding domain protein 2 (MBD2) and the Nucleosome Remodeling Deacetylase complex (NuRD) specifically binds to the mCpG-island and blocks tumor suppressor gene expression, finding an inhibitor of this MBD2-NuRD complex is hypothesized to be important for the development of novel anti-cancer drugs. I found that the site, which is responsible for the MBD2 interaction with thetranscriptional repressor p66-α (p66α, which is a part of the NuRD complex), is characterized by a specific disorder-to-order transition pattern, this pattern showed a remarkable similarity to the disorder-to-order pattern of the Myc transcription factor binding site for the Max transcription factor. Importantly, several inhibitors of the Myc-Max interaction targeting the disorder-to-order transition site of Myc were previously described. By applying molecular docking at the disorder-to-order transition site of MBD2, two compounds were identified and further evaluated through molecular dynamics simulations. Anti-leukemia and anti-metastasis effectiveness of these compounds was demonstrated in dedicated in vitro and in vivo experiments conducted by our collaborators. In relation to the defective protein associated with the X-linked protoporphyria (XLPP), the hepta-variant of mouse erythroid 5-aminolevulinate synthase (mALAS2), previously shown to be characterized by a remarkable acceleration of the reaction rate, was investigated through molecular dynamics simulations. In this study, a loop to β-strand transition was observed, and this observation was crucial for a better understanding of the previously described rate-enhancing effects of seven simultaneous variations in the active loop site of this protein. Finally, a wide spectrum of bioinformatics tools was applied to carefully analyze a potential role of intrinsic disorder in a set of cardiomyopathy-related mutations in the components of human cardiac troponin. This analysis revealed that, in comparison with the wild type troponin, chains containing the disease-associated mutations were typically characterized by a local decrease in intrinsic disorder propensity. These mutations affected some disorder-based protein-protein interaction sites and caused remarkable rearrangements of the complex pattern of post-translational modifications. Therefore, this work illustrates that inclusion of the protein intrinsic disorder analysis into the arsenal of techniques used by the biomedical researchers represents an important and promising approach that provides novel inputs for the better understanding of protein behavior in relation to human disease at the molecular level. Techniques and methods developed and utilized in this study will significantly contribute to future biomedical research.

Les protéines sont des acteurs majeurs dans les processus moléculaires et cellulaires d’un organisme. La remise en question des modalités associées aux fonctions de ces macromolécules a récemment été apportée par le concept de désordre intrinsèque. Celui-ci définit l’absence (transitoire ou permanente) de structure tridimensionnelle de certaines protéines ou régions protéiques comme étant directement liée à leurs fonctions. Chez les virus à ARN, les propriétés des protéines ou régions désordonnées semblent associées aux capacités de ces micro-organismes à détourner la machinerie cellulaire de l’hôte en interagissant avec de multiples partenaires, et à s’adapter aux nombreuses contraintes auxquelles ils doivent faire face en tant que parasites obligatoires. Ce travail porte sur les potyvirus, figurant parmi les pathogènes de plantes les plus dommageables étudiés à ce jour. L'objectif de cette thèse a été d’explorer les fonctions associées au désordre intrinsèque dans le cycle infectieux des potyvirus ainsi que dans le processus d’adaptation. Notre approche a ainsi démontré que : i) le désordre est ubiquitaire chez le genre Potyvirus ; ii) les régions de désordre conservées chez plusieurs protéines de potyvirus semblent être associées à leur(s) fonction(s) pendant l'infection ; iii) les régions désordonnées sont généralement associées à moins de contraintes évolutives, suggérant ainsi leur implication dans les processus adaptatifs des potyvirus ; iv) les régions prédites comme désordonnées semblent privilégier l’apparition de mutations et donc la capacité d’un virus à accumuler de la diversité génétique au cours de l'évolution sur son hôte naturel ; v) ce travail a permis de corréler le taux en désordre de la protéine viral genome-linked (VPg) du Potato virus Y à sa capacité à s’adapter à la résistance récessive pvr23 du piment. / Proteins are essential actors involved in a majority of molecular and cellular processes. The features associated with the functions of these macromolecules have been recently questioned with the emergence of the intrinsic disorder concept. It defines the transitory or permanent lack of 3D structure in some proteins or regions as directly related to their functions. Among RNA viruses, the properties of disordered proteins may be linked to the ability of these microorganisms to hijack the host machinery by interacting with multiple partners, as well as to adapt to the multiple constraints they must face as obligatory parasites. This work focuses on the Potyvirus genus, which includes some of the most damaging plant pathogens studied to date. The goal of this thesis was to explore the functions associated with intrinsic disorder in the infectious cycle of this viral genus as well as in its process of adaptation. Our studies have shown that i) intrinsic disorder is ubiquitous in potyviruses; ii) intrinsically disordered regions (IDR) of some of potyviral proteins are likely to be associated with important functions for the viral cycle ; iii) IDR are generally less evolutionary constrained, suggesting an adaptive potential of these regions ; iv) predicted IDR seem to favor the appearance of mutations and therefore virus ability to accumulate genetic diversity during its evolution in natural host ; v) an experimental disorder modulation within the Viral genome-linked (VPg) protein has been demonstrated as positively correlated with the adaptive ability of the Potato virus Y to overcome the pvr23 recessive resistance in pepper.