Teses / dissertações sobre o tema "Rare genetic disease"

Coeliac disease is a common (1% prevalence) inflammatory disease of the small intestine, involving the role of tissue transglutaminase and HLA-­‐DQ binding immuno-­‐dominant wheat peptides. The disease is highly heritable, however, at most only 40% of this heritability is explained by HLA-­‐DQ and risk variants from genome wide association and fine mapping studies. The hypothesis of the research in this thesis is that rare (minor allele frequency <0.5%) mutations of large effect size (odds ratios ~2 – 5) exist, especially in multiply affected pedigrees, which account for the missing heritability of disease. NimbleGen exome capture and Illumina GAIIx high throughput sequencing was performed in 75 coeliac disease individuals from 55 multiply affected families. Candidate genes were chosen from various analytical strategies: linkage, shared variants between multiple related subjects and gene burden tests for multiple potentially causal variants. Highly multiplexed amplicon sequencing, using Fluidigm technology, of all RefSeq exons from 24 candidate genes in 2,304 coeliac cases and 2,304 controls was performed to locate further rare variation. Gene burden tests on a highly stringent post quality control dataset identified no significant associations (P<1x10-­‐3) at the resequenced candidate genes. The strategy of sequencing multiply affected families, and deep follow up of candidate genes, has not identified new disease risk mutations. Common variants (and other factors, e.g. environmental) may instead account for familial clustering in this common autoimmune disease.

This dissertation aims at investigating the association between genotypes and phenotypes in human. Both common and rare regulatory variants have been studied. The phenotypes include disease risk, clinical traits and gene expression levels. This dissertation describes three different types of association study. The first study investigated the relationship between common variants and three sub-clinical traits as well as three complex diseases in the Center for Health Discovery and Well Being study (CHDWB). The second study is GWAS analysis of TNF-α and BMI/CRP conducted as a contribution to meta-GWAS analyses of these traits with investigators at the University of Groningen in the Netherlands, and the 1000 Genomes Consortium. The third study was the most original contribution of my thesis as it assessed the association between rare regulatory variants in promoter regions and gene expression levels. The results clearly show an enrichment of rare variants at both extremes of gene expression. This dissertation provides insight into how common and rare variants associate with broadly-defined quantitative phenotypes. The demonstration that rare regulatory variants make a substantial contribution to gene expression variation has important implications for personalized medicine as it implies that de novo and other rare alleles need to be considered as candidate effectors of rare disease risk.

Studies of large families with inherited single gene disorders identified a role of rare genetic variation as a cause of disease and enabled gene-based diagnosis. The increasing availability of population-scale genomic sequencing implies the potential to extend gene-based diagnosis from individuals with monogenic disease to the prediction of disease risk in the general population. Cardiovascular disease (CVD), as a highly heritable condition with significant public health burden, represents an excellent place to consider the promise and limitations of extending our understanding of rare variation in single gene disorders to the general population.

High-throughput sequencing technologies have changed the way we identify, study and understand the role of rare variation in Mendelian diseases. Sequencing in complex diseases have proven to be more challenging to interpret, but methods and approaches are being developed to aid in our understanding of variation in these diseases.

Lung Function is a physiological measurement used for monitoring respiratory health and in the diagnosis of chronic obstructive pulmonary disease (COPD), a leading cause of morbidity and mortality worldwide. Lung function and COPD are influenced by a combination of environmental and genetic factors. This thesis aims to investigate the genetic basis of these traits, with a particular focus on the effect of low frequency and rare genetic variants, so far largely overlooked in genome-wide association studies (GWAS). An analysis of exome array data and COPD identifies novel associations between COPD risk and low frequency single nucleotide polymorphisms (SNPs) in MOCS3 and IFIT3 and between a rare SNP in SERPINA12 and percent predicted forced expiratory volume in one second (FEV1) in COPD cases. Recently developed methods for the meta-analysis of gene-based tests are empirically evaluated and shown to be approximately equivalent to a mega-analysis using individual level data for a quantitative trait. These methods are then applied in a meta-analysis of exome array data and quantitative lung function measures. This meta-analysis identifies no gene-based associations; however genome-wide significant (P < 5x10⁻⁸) single variant associations are identified in two novel regions: a SNP near LY86 associated with the ratio of FEV1 to forced vital capacity (FVC) and a SNP near FGF10 associated with FVC in ever smokers. Finally the largest GWAS to date of two lung function flow measures (peak expiratory flow [PEF] and forced expiratory flow between 25% and 75% of FVC [FEF25-75]) is described. The overlap in variants associated with PEF and FEF25-75 and volumetric measures of lung function (FEV1, FVC and FEV1/FVC) is examined, and 10 SNPs are identified as showing association with PEF (P < 5x10⁻⁸), but no other lung function trait with P < 5x10⁻⁵. These findings have the potential to provide insight into the biological mechanisms underlying lung health and disease.

Dementia is a clinical state caused by neurodegeneration and characterized by a loss of function in cognitive domains and behavior. Alzheimer’s disease (AD) is the most common form of dementia. Although the amyloid β (Aβ) protein and hyperphosphorylated tau aggregates in the brain are considered to be the key pathological hallmarks of AD, the exact cause of AD is yet to be identified. In addition, clinical diagnoses of AD can be error prone. Many previous studies have compared the clinical diagnosis of AD against the gold standard of autopsy confirmation and shown substantial AD misdiagnosis Hippocampal sclerosis of aging (HS-Aging) is one type of dementia that is often clinically misdiagnosed as AD. AD and HS-Aging are controlled by different genetic architectures. Familial AD, which often occurs early in life, is linked to mainly mutations in three genes: APP, PSEN1, and PSEN2. Late-onset AD (LOAD) is strongly associated with the ε4 allele of apolipoprotein E (APOE) gene. In addition to the APOE gene, genome-wide association studies (GWAS) have identified several single nucleotide polymorphisms (SNPs) in or close to some genes associated with LOAD. On the other hand, GRN, TMEM106B, ABCC9, and KCNMB2 have been reported to harbor risk alleles associated with HS-Aging pathology. Although GWAS have succeeded in revealing numerous susceptibility variants for dementias, it is an ongoing challenge to identify functional loci and to understand how they contribute to dementia pathogenesis. Until recently, rare variants were not investigated comprehensively. GWAS rely on genotype imputation which is not reliable for rare variants. Therefore, imputed rare variants are typically removed from GWAS analysis. Recent advances in sequencing technologies enable accurate genotyping of rare variants, thus potentially improving our understanding the role of rare variants on disease. There are significant computational and statistical challenges for these sequencing studies. Traditional single variant-based association tests are underpowered to detect rare variant associations. Instead, more powerful and computationally efficient approaches for aggregating the effects of rare variants have become a standard approach for association testing. The sequence-kernel association test (SKAT) is one of the most powerful rare variant analysis methods. A recently-proposed scan-statistic-based test is another approach to detect the location of rare variant clusters influencing disease. In the first study, we examined the gene-based associations of the four putative risk genes, GRN, TMEM106B, ABCC9, and KCNMB2 with HS-aging pathology. We analyzed haplotype associations of a targeted ABCC9 region with HS-Aging pathology and with ABCC9 gene expression. In the second study, we elucidated the role of the non-coding SNPs identified in the International Genomics of Alzheimer’s Project (IGAP) consortium GWAS within a systems genetics framework to understand the flow of biological information underlying AD. In the last study, we identified genetic regions which contain rare variants associated with AD using a scan-statistic-based approach.

Rare diseases, when considered as a whole, affect up to 7% of the population, which would represent 3.5 million individuals in the United Kingdom alone. However, while 'personalised medicine' is now yielding remarkable results using recent sequencing technologies in terms of diagnosing genetic conditions, we have made much less headway in translating this patient information into therapies and effective treatments. Even with recent calls for greater research into personalised treatments for those affected by a rare disease, progress in this area is still severely lacking, in part due to the astronomical cost and time involved in bringing treatments to the clinic. Gene correction using the recently-described genome editing technology CRISPR/Cas9, which allows precise editing of DNA, offers an exciting new avenue of treatment, if not cure, for rare diseases; up to 80% of which have a genetic component. This system allows the researcher to target any locus in the genome for cleavage with a short guide-RNA, as long as it precedes a highly ubiquitous NGG sequence motif. If a repair sequence is then also provided, such as a wild-type copy of the mutated gene, it can be incorporated by homology-directed repair (HDR), leading to gene correction. As both guide-RNA and repair template are easily generated, whilst the machinery for editing and delivery remain the same, this system could usher in the era of 'personalised medicine' and offer hope to those with rare genetic diseases. However, currently it is difficult to test the efficacy of CRISPR/Cas9 for gene correction, especially in vivo. Therefore, in my PhD I have developed a novel fluorescent reporter system which provides a rapid, visual read-out of both non-homologous end joining (NHEJ) and homology-directed repair (HDR) driven by CRISPR/Cas9. This system is built upon a cassette which is stably and heterozygously integrated into a ubiquitously expressed locus in the mouse genome. This cassette contains a strong hybrid promoter driving expression of membrane-tagged tdTomato, followed by a strong stop sequence, and then membrane-tagged EGFP. Unedited, this system drives strong expression of membrane-tdTomato in all cell types in the embryo and adult mouse. However, following the addition of CRISPR/Cas9 components, and upon cleavage, the tdTomato is rapidly excised, resulting via NHEJ either in cells without fluorescence (due to imperfect deletions) or with membrane-EGFP. If a repair template containing nuclear tagged-EGFP is also supplied, the editing machinery may then use the precise HDR pathway, which results in a rapid transition from membrane-tdTomato to nuclear- EGFP. Thereby this system allows the kinetics of editing to be visualised in real time and allows simple scoring of the proportion of cells which have been edited by NHEJ or corrected by HDR. It therefore provides a simple, fast and scalable manner to optimise reagents and protocols for gene correction by CRISPR/Cas9, especially compared to sequencing approaches, and will prove broadly useful to many researchers in the field. Further to this, I have shown that methods which lead to gene correction in our reporter system are also able to partially repair mutations found in the disease-causing gene, Zmynd10; which is implicated in the respiratory disorder primary ciliary dyskinesia (PCD), for which there is no effective treatment. PCD is an autosomal-recessive rare disorder affecting motile cilia (MIM:244400), which results in impaired mucociliary clearance leading to neonatal respiratory distress and recurrent airway infections, often progressing to lung failure. Clinically, PCD is a chronic airway disease, similar to CF, with progressive deterioration of lung function and lower airway bacterial colonization. However, unlike CF which is monogenic, over 40 genes are known to cause PCD. The high genetic heterogeneity of this rare disease makes it well suited to such a genome editing strategy, which can be tailored for the correction of any mutated locus.

The aetiology of common diseases is shaped by the effects of genetic and environmental factors. Big efforts have been devoted to unravel the genetic basis of disease with the hope that it will help to develop new therapeutic treatments and to achieve personalized medicine. With the development of high-throughput genotyping technologies, hundreds of association studies have described many loci associated to disease. However, the depiction of disease architecture remains incomplete. The aim of this work is to perform exhaustive comparisons across human populations to evaluate pressing questions. Our results provide new insights in the allele frequency of risk variants, their sharing across populations and the likely architecture of disease
La etiología de las enfermedades comunes está formada por factores genéticos y ambientales. Se ha puesto mucho empeño en describir sus bases genéticas. Este conocimiento será útil para desarrollar nuevas terapias y la medicina personalizada. Gracias a las técnicas de genotipado masivo, centenares de estudios de asociación han descrito una infinidad de genes asociados a enfermedad. Pese a ello, la arquitectura genética de las enfermedades no ha sido totalmente descrita. Esta tesis pretende llevar a cabo exhaustivas comparaciones entre poblaciones para responder diversas preguntas candentes. Nuestros resultados dan pistas sobre la frecuencia de los alelos de riesgo, su presencia entre poblaciones y la probable arquitectura de las enfermedades.

Nonalcoholic fatty liver disease (NAFLD) affects roughly 30% of the general population and its prevalence is increasing worldwide, in particular in Western countries where it is projected to become the main cause of hepatocellular carcinoma (HCC) within 2025. Although the majority of NAFLD progressing towards HCC are individuals with advanced fibrosis, NAFLD-HCC frequently develops even in patients without cirrhosis. Family history and genetic factors play an important role in the pathogenesis of progressive NAFLD and of HCC. Our hypothesis is that both common and rare variants may have a role in influencing HCC predisposition. Aim of this study was (1) to evaluate whether the rs641738 C>T common genetic variant in the MBOAT7/TMC4 locus, the last genetic variant recently associated with NAFLD development and progression towards early phases of damage, also predisposes to HCC in patients stratified by the presence of severe fibrosis, and (2) to assess whether telomere attrition and inherited rare hTERT mutations associates with NAFLD-HCC development, since mutations in Telomerase gene have been linked to familial liver diseases. In the first part of the study, we found that the rs641738 T allele is associated with NAFLD-HCC independently of fibrosis severity and of several confounders (allelic OR 1.65, 95% confidence interval 1.08-2.55; n=765), particularly in patients without advanced fibrosis (p<0.001). The rs641738 risk T allele is associated with reduced MBOAT7 expression, evaluated by qRT-PCR on liver biopsy, specifically in patients without severe fibrosis, and was more strongly associated with HCC development in this specific subgroup (p=0.02). Moreover, the number of PNPLA3, TM6SF2, and MBOAT7 risk alleles is associated with NAFLD-HCC development independently of clinical factors (p<0.001), but it did not significantly improve their predictive accuracy. The independent association between the 5 MBOAT7 risk variant and increased HCC risk has also been confirmed in a combined cohort of chronic hepatitis C or alcoholic liver disease patients without cirrhosis (n=1121). In the second part of the study, we found that telomere length, evaluated by qRT-PCR, is progressively reduced in 40 NAFLD-HCC vs. 45 NAFLD-cirrhosis (p=0.048) and 64 healthy controls (p=0.0006), independently of age and sex. We detected an enrichment of rare germline coding mutations in hTERT in NAFLD-HCC patients vs. controls, even after considering 78 European primary liver cancers. No mutations were found in NAFLD-cirrhosis and local controls, and only one in 503 Europeans from the 1000Genomes Project. Thus, overall mutations frequency was 0.025 in cases vs. <0.001 in controls (p=0.0005 at Burden test). hTERT mutations occurred predominantly in females (p=0.03) and were predominantly located in the N-terminal template-binding domain of the gene (p=0.037 for specific enrichment). The frameshift Glu113Argfs*79 and the missense Glu668Asp damaging mutations co-segregated with liver disease in families. Besides Glu668Asp, the Ala67Val variant resulted in reduced intracellular protein levels in HEK-293 cells. These data suggest that telomere attrition and certain hTERT mutations are involved in the pathogenesis of NAFLD-HCC. In conclusion, the identification of new common prognostic marker, such as the MBOAT7 rs641738 variant, in NAFLD patients and the evaluation of hTERT rare mutations in NAFLD-HCC may improve HCC screening strategies, by assisting in the identification of subjects at risk even in the absence of severe fibrosis and of family members at high risk of progressive liver disease, respectively. In future, functional studies are needed in order to understand the exact role and the interplay of the several common and rare genetic variants which contribute to the molecular mechanisms that underlie the pathogenesis of NAFLD-HCC.

La méthylation de l'ADN est un processus vital pour le développement des mammifères. Sa distribution anormale,notamment au niveau des régions répétées du génome, est une signature pathologique. La découverte de maladies héréditaires touchant la stabilité du génome a permis des avancées considérables dans l'identification des acteurs et des mécanismes. Nous avons choisi d'étudier le syndrome ICF (Immunodéficience, instabilité Centromérique et anomalies Faciales), première maladie génétique identifiée avec des défauts de la méthylation de l’ADN, liés à une instabilité chromosomique. Lorsque j'ai commencé ma thèse, des mutations dans les gènes DNMT3B et ZBTB24 avaient été décrites comme causes génétiques du syndrome. Cependant, d'autres causes génétiques restaient inconnues. Nos travaux ont permis d'identifier deux nouveaux gènes, CDCA7 et HELLS, dont les mutations sont responsables du syndrome. J'ai montré que leur perte de fonction dans les cellules somatiques entraîne un défaut de méthylation des répétitions centromériques, suggérant leur rôle dans le maintien de la méthylation de l'ADN. Par conséquent, l'étude de l'étiologie d'une maladie génétique rare a permis d'identifier de nouveaux « gardiens » de la stabilité du génome, avec des fonctions jusqu'alors insoupçonnées dans les processus de méthylation de l'ADN et dans le développement. Au cours de mon doctorat, j'ai établi des cartes de méthylation des cellules de patients ICF afin d'identifier les cibles communes et distinctes de ces facteurs, ainsi que leurs caractéristiques génomiques et épigénomiques. Contrairement aux mutations de DNMT3B,celles de ZBTB24, CDCA7 et HELLS affectent la méthylation dans des régions pauvres en CpG, dans des régions intergéniques et dans des répétitions d'ADN intercalées. Plus généralement, ce sont les régions d'hétérochromatine qui sont les plus touchées et en particulier des clusters des gènes codants et non codants, dont certains sont exprimés de manière monoallélique. Pour mieux caractériser le rôle de ZBTB24 dans le développement et la méthylation de l'ADN,nous avons généré un modèle murin mutant qui nous a permis de monter que ZBTB24 était essentielle pour le développement embryonnaire précoce. De plus, ZBTB24 jouerait un rôle dans l'établissement de la méthylation des séquences répétées de l'ADN, à la fois en tandem ou intercalé. Fait intéressant, ZBTB24 semble être également impliqué dans l'établissement de la marque répressive H3K9me3, suggérant un rôle de la protéine dans le "dialogue" entre la méthylation de l'ADN et celle des histones. Dans l'ensemble, mon travail met l'accent sur la façon dont la méthylation de l'ADN et les marques d'hétérochromatine sont établies et maintenues à des gènes uniques et des répétitions de l'ADN, et fournit de nouveaux acteurs et mécanismes à considérer dans les études sur le maintien de la stabilité du génome
DNA methylation is an essential process for the development of mammals. Its abnormal distribution, particularly at the level of the repeated regions of the genome, is a pathological signature. The discovery of hereditary diseases affecting DNA methylation and the stability of the genome allowed a considerable progress in the identification of their actors and mechanisms. We chose to study the ICF (Immunodeficiency, Centromeric Instability and Facial Abnormalities) syndrome, the first genetic disorder identified with defects in the distribution of DNA methylation, linked to chromosomal instability. When I started my PhD, mutations in two genes had been described to cause the ICF syndrome: DNMT3B and ZBTB24. However, the genetic origin of a subset of ICF patients remained unknown. We identified mutations in CDCA7 and HELLS as causative of the ICF syndrome. I showed that their loss of function in somatic cells results in the loss of DNA methylation at centromeric repeats, strongly suggestive of a role DNA methylation maintenance. Hence, the study of the aetiology of a genetic disease provided new candidate “guardians” of DNA repeats and genome stability, with virtually unknown functions but with exciting potential roles in the DNA methylation machinery and in development. During my PhD, I established methylation maps in ICF patients cells to identify common and distinct targets of these factors, as well as their genomic and epigenomic characteristics. In contrast to DNMT3B mutations, those in ZBTB24, CDCA7 and HELLS affect methylation at CpG-poor regions in intergenic genomic locations and at interspersed DNA repeats, and more generally, at genomic locations with heterochromatic features. Their integrity is required for the methylated status of coding and non-coding clusters of genes, some of which are expressed in a monoallelic manner. To better characterize the role of ZBTB24 in development and DNA methylation pathways, we generated a mouse model carrying mutations in ZBTB24. We showed that ZBTB24 is essential for early development, while it seemed to be dispensable for in vitro differentiation of murine ES cells. We implicated ZBTB24 in the establishment of DNA methylation at DNA repeats, both in tandem or interspersed, in differentiating ES cells. Interestingly, ZBTB24 seems to be also implicated in the establishment of the repressive mark H3K9me3 suggesting that ZBTB24 may indirectly control DNA methylation through an interplay with histone marks. As a whole, our work sheds light on how DNA methylation and heterochromatin marks are established and maintained at unique genes and DNA repeats, and provides new actors and mechanisms to consider in studies of the maintenance of genome stability

Background: Genetic variants that elicit aberrant splicing of pre-messenger RNA (pre-mRNA) are recognised as causative variants in ~30-50% of genetic disorders. However, it is still not possible to predict reliably if and how a variant will impact splicing, limiting the application of in silico splice prediction tools in variant interpretation. Most splicing variants fall outside the essential splice site and, in the absence of RNA testing, remain classed variants of uncertain significance (VUS) according to ACMG-AMP (American College of Medical Genetics and Genomics and Association for Molecular Pathology) guidelines. Sequence analysis of spliced messenger RNA (mRNA) is the only definitive means to determine the precise nature of variant associated mis-splicing. Tissues with limited accessibility, for instance vital organs, present a challenge for RNA testing of genes with tissue specific expression. Fortunately, clinically accessible tissues such as blood and fibroblasts can be used to infer variant associated mis-splicing outcomes in the manifesting tissue. A further challenge arises from the lack of guidance on how functional evidence (PS3/BS3 criteria) from RNA studies should be applied to variant interpretation within the current ACMG-AMP framework. There is an urgent need to establish ACMG-AMP aligned quality standards and guidelines for complex RNA assay data for accurate and consistent variant interpretation between clinical laboratories. Methods: Families were recruited from local area health districts across Australia and New Zealand using inclusion criteria to ascertain putative splicing variants with high clinical suspicion of causality. More than 120 families with diverse monogenic conditions were triaged into PCR-based RNA testing, with comparative RNA-sequencing for 38 cases. Consensus ascertainment criteria, standard practices for PCR-based RNA testing, and RNA assay interpretation rubric were devised through consultation with the clinical and molecular genetics community via surveys, live polls and SpliceACORD consortium (Australasian Consortium for RNA Diagnostics) meetings. Results: Informative RNA assay data was obtained for 96% cases, enabling variant re-classification for 75% of variants. RNA testing reports were used to guide clinical care and genetic counselling, and 75% of diagnosis were clinician-reported to have a positive impact for the family. PCR-based RNA diagnostics has the capacity to analyse 81.3% of clinically significant genes and to allow phasing of RNA splicing events. Variant associated mis-splicing was highly reproducible between affected individuals and heterozygotes, and between different biospecimens. Discussion: We provide a standardised protocol for PCR-based RNA testing and ACMG-AMP aligned recommendations for the interpretation of RNA assay data. Our study demonstrates the significant diagnostic and health benefits of RNA analysis as adjunct testing to extend diagnostic yield from genomic testing.

AGC1 deficiency is a rare demyelinating disease caused by mutations in the SLC25A12 gene, which encodes for the mitochondrial glutamate-aspartate carrier 1 (AGC1/Alarar), highly expressed in the central nervous system. In neurons, impairment in AGC1 activity leads to reduction in N-acetyl-aspartate, the main lipid precursor for myelin synthesis (Profilo et al., 2017); in oligodendrocytes progenitors cells, AGC1 down regulation has been related to early arrest proliferation and premature differentiation (Petralla et al., 2019). Additionally, in vivo AGC1 deficiency models i.e., heterozygous mice for AGC1 knock-out and neurospheres from their subventricular zone, respectively, showed a global decrease in cells proliferation and a switch in neural stem cells (NSCs) commitment, with specific reduction in OPCs number and increase in neural and astrocytic pools (Petralla et al., 2019). Therefore, the present study aims to investigate the transcriptional and epigenetic regulation underlying the alterations observed in OPCs and NSCs biological mechanisms, in either AGC1 deficiency models of Oli-neu cells (murine immortalized oligodendrocytes precursors cells), partially silenced by a shRNA for SLC25A12 gene, and SVZ-derived neurospheres from AGC1+/- mice. Western blot and immunofluorescence analysis revealed significant variations in the expression of transcription factors involved in brain cells’ proliferation and differentiation, in association with altered histone post-translational modifications, as well as histone acetylases (HATs) and deacetylases (HDACs) activity/expression, suggesting an improper transcriptional and epigenetic regulation affecting both AGC1 deficiency in vitro models. Furthermore, given the large role of acetylation in controlling in specific time-windows OPC maturation (Hernandez and Casaccia; 2015), pharmacological HATs/HDACs inhibitions were performed, confirming the involvement of chromatin remodelling enzymes in the altered proliferation and early differentiation observed in the AGC1 deficiency models of siAGC1 Oli-neu cells and AGC1+/- mice-derived neurospheres.

AGC1-deficiency is a rare genetic neurodegenerative disease caused by defects in the SLC25A12 gene encoding for a mitochondrial aspartate-glutamate carrier, important in the metabolism of brain amino acids and myelin synthesis. AGC1 mutations lead to drastic reduction of carrier activity, which results in hypotonia, developmental delay, intractable epilepsy and cortical atrophy with altered myelin formation in the CNS, most likely due to a dramatic reduction of N-acetyl aspartate levels in the brain. The aim of this thesis was to study the molecular mechanisms underlying AGC1-deficiency in appropriate in vitro (oligodendrocyte precursor cells) and in vivo (SLC25A12 heterozygous knockout mice) disease models in particular by focusing on oligodendrocyte precursor alterations to better define pathogenetic mechanisms that could potentially lead to identify new interesting therapeutic targets. Furthermore, our study focused on the evaluation of Neural Stem Cells proliferation and differentiation in neuronal and glia cells, both astrocytes and oligodendrocytes, in neurospheres derived from the subventricular zone of our mouse model. Lastly, preliminary experiments have been performed on NSCs derived from induced Pluripotent Stem cells from AGC1-deficiency patients and healthy controls, in order to further validate our data in human cells. All results showed a proliferation deficit of OPCs that was not due to mitochondrial biochemical alterations, but rather associated with an alteration of trophic factors essential for maintaining the balance between oligodendrocyte proliferation and differentiation, mainly PDGFα and TGFβ. These results supported that alterations induced by AGC1 reduced activity could impair the physiological cross-talk mediated by growth factors between neurons and OPCs necessary for OPCs proliferation and neuronal survival. The importance of this study lies also on the fact that mitochondrial dysfunction is at the basis not only of AGC1-deficiency, but also of other neurodegenerative and demyelinating diseases, some of which are rare, while others are widely spread, such as multiple sclerosis.

L’albinisme est un groupe de pathologies génétiques hétérogènes sur le plan clinique. On retrouve de manière constante une atteinte ophtalmologique pouvant associer un nystagmus, une hypoplasie fovéolaire et des anomalies de décussation des voies optiques. L’atteinte cutanéophanérienne est plus variable avec une hypopigmentation de la peau, des cheveux et de l’iris. L’hypopigmentation est la conséquence directe et visible d’un défaut de synthèse de mélanine au sein des mélanocytes cutanés chez les patients atteints d’albinisme. La pathologie ophtalmologique résulte quant à elle d’un défaut de signalisation entre l’épithélium pigmentaire de la rétine et la rétine neurosensorielle au cours du développement. Sur le plan génétique, vingt gènes ont été identifiés à ce jour définissant vingt types d’albinisme classés en trois formes : oculocutanée, oculaire, ou syndromique. Le séquençage de ces vingt gènes permet de confirmer le diagnostic clinique dans 70% des cas laissant un nombre significatif de patients sans caractérisation génétique. Parmi ces derniers, on trouve des sujets chez lesquels on aura identifié uniquement un ou deux variants de signification indéterminée (VSI).Cette étude porte sur 5 gènes majeurs dont les altérations provoquent un albinisme oculocutané à transmission autosomique récessive : TYR, TYRP1, DCT, OCA2 et SLC45A2. Les cinq protéines correspondantes jouent un rôle spécifique dans la synthèse de mélanine soit en tant qu’enzyme de la mélanogénèse (TYR, TYRP1 et DCT), soit en tant que régulateur du pH mélanosomal (protéine P, SLC45A2). Leurs fonctions précises dans un contexte donné (espèce, type cellulaire, stade développemental) restent cependant peu documentées.Le premier objectif de ce travail est d’améliorer le diagnostic des patients atteints d’albinisme par plusieurs approches d’analyses de cohortes, de tests fonctionnels et de test de sauvetage de phénotype par des VSI. Le second objectif est la création de modèles cellulaires mélanocytaires inactivés pour les 5 gènes afin d’étudier leur phénotype moléculaire et pigmentaire.Tout d’abord, nous rapportons l’analyse statistique d’une cohorte de patients porteurs de deux variants faux-sens fréquents dans le gène TYR classés VSI. Cette étude montre que la combinaison de ces deux variants faux-sens est pathogène dans le contexte d’un haplotype comportant un variant du promoteur ce qui a d’importantes applications sur la stratégie diagnostique.Puis nous présentons des stratégies robustes pour caractériser des variants synonymes et faux-sens du gène OCA2 classés VSI. Ces travaux ont permis de prouver l’effet pathogène de ces variants par augmentation du saut de l’exon 10 du gène OCA2. Nous montrons également que la recherche d’anomalies d’épissage peut être réalisée à partir d’un simple prélèvement sanguin.Afin de compléter les connaissances sur le rôle de chacune des protéines codées par les cinq gènes d’intérêt (TYR, OCA2, TYRP1, SLC45A2 et DCT), nous avons créé des modèles de cellules mélanocytaires MNT1 inactivées pour chacun des 5 gènes grâce au système CRISPR-Cas9. Nous comparons ces 5 modèles par différentes analyses permettant d’évaluer leur capacité de production de mélanine et leurs caractéristiques moléculaires. Une analyse comparative par protéomique nous permet de mettre en évidence les régulations communes et spécifiques dépendantes de ces gènes.Enfin le modèle inactivé pour TYR est exploité dans le cadre d’un test de rescue pour évaluer la pathogénicité de VSI de TYR. Ce test nous montre que des variants faux-sens peuvent être reclassés bénins ou pathogènes en fonction de leur capacité à restaurer le phénotype pigmentaire dans ce modèle.L’analyse de ces modèles cellulaires permet d’apporter de nouvelles connaissances sur la physiologie de la biosynthèse de mélanine. Les approches développées pour l’analyse des variants seront transférées dans le cadre hospitalier et permettront d’améliorer significativement le diagnostic
Albinism is a group of clinically heterogeneous genetic disorders. Ophthalmologic features are constant including nystagmus, foveal hypoplasia and chiasmal misrouting. Cutaneous features are more variable with hypopigmentation of the skin, hair, and iris. In affected patients, hypopigmentation results directly from a defect in melanin synthesis in skin melanocytes. Ophthalmologic features are due to a signaling defect between the retinal pigmented epithelium and the neuronal retina during development. To date, twenty genes have been identified, defining twenty types of albinism, classified into three forms: oculocutaneous, ocular, or syndromic. Sequencing of these twenty genes confirms the diagnosis in 70% of cases, leaving a significant number of patients without genetic determinism. Among these, patients may bear one or two variants of unknown significance (VUS). This study focuses on 5 major genes which cause autosomal recessive oculocutaneous albinism: TYR, TYRP1, DCT, OCA2 and SLC45A2. The five corresponding proteins have a specific role in melanin synthesis either as a melanogenic enzyme (TYR, TYRP1 and DCT) or as a melanosomal pH regulator (OCA2, SLC45A2). Their precise functions in a given context (species, cell type, developmental stage) remain poorly documented.The first objective of this work is to improve the diagnosis of patients with albinism by multiple approaches of cohort analysis, functional tests and rescue test of VUS. The second objective is the design of melanocyte cell models knock-out (KO) for the 5 genes in order to study their molecular and pigmentary phenotype.First, we report the statistical analysis of a cohort of patients carrying two frequent missense variants in TYR gene classified as VUS. This study shows that the combination of these two missense variants is pathogenic if included in a haplotype with a promoter variant which has important implications in diagnosis strategies.Then we present solid strategies to characterize synonymous and missense variants of OCA2 gene classified as VUS. These analyses prove the pathogenic effect of these variants by increasing skipping of exon 10 of OCA2 gene. We also show that splicing anomalies can be detected from a simple blood sample.In order to increase knowledge on the role of each of the proteins encoded by the five genes of interest (TYR, OCA2, TYRP1, SLC45A2 and DCT), we designed MNT1 melanocyte cell models that are each KO for one of the 5 genes using the CRISPR-Cas9 system. We compare these 5 models by different analyses to assess their melanin production capacity and their molecular characteristics. A comparative proteomic assay allows us to reveal the common and specific regulations between these genes.Finally, the KO model for TYR is used in a rescue test to assess pathogenicity of TYR VUS. This test shows us that missense variants can be reclassified as benign or pathogenic depending on their ability to restore the pigmentary phenotype in this model.Analysis of these cellular models provides new knowledge on melanin biosynthesis physiology. The techniques developed for the analysis of variants will be transferred in a hospital setting to significantly improve diagnosis

Tuberous sclerosis complex (TSC) is an autosomal dominant multisystem disorder characterized by development of hamartomas, intellectual disability, seizures and autism. TSC is caused by inactivating mutations in either the TSC1 or the TSC2 genes. A pathogenic variant is not identified in up to 10% of the patients with a clinical diagnosis of TSC despite full molecular assessment. These individuals are referred to as NMI (No Mutation Identified), and it is not clear if they should be monitored and/or treated in similar fashion to those with known etiology of TSC. To identify the genetic cause of TSC in these patients, we selected ten individuals with a definite clinical diagnosis of TSC and NMI and performed a pilot study. Three different technologies were used and their results were compared: chromosomal microarray, trio whole genome sequencing, and targeted deep-coverage next generation sequencing of TSC1 and TSC2. We identified mosaic variants in TSC1/TSC2 in six patients. No variants in other genes were detected in the remaining individuals. Based on the results of the pilot study and on recent literature, we then performed targeted deep-coverage TSC1/TSC2 sequencing on 200 affected individuals using peripheral blood DNA and saliva or other tissue samples where available. We identified 24 patients with mosaic pathogenic variants in TSC1 (n=2) or TSC2 (n=22), defining a rate of mosaicism of 12%. Mosaic variant allele frequency (VAF) ranged from 2% to 32% in blood and from 2% to 35% in saliva. Most affected individuals had low-level mosaicism (VAF ≤10%). We performed an extensive analysis of the phenotype, and show that individuals with a mosaic variant in TSC1/TSC2 often display normal cognitive functioning, although other TSC-associated neuropsychiatric disorders (TAND) are seen in 62% of the patients. Cortical tubers are invariably present and seizures are diagnosed in 54% of the cohort, but infantile spasms are rare. The number of cutaneous manifestations in these patients is often insufficient to meet diagnostic criteria, except for facial angiofibromas. We observed a high frequency of pulmonary and renal manifestations in our mosaic cohort, which are as severe as those seen in the general TSC population. None of the patients who have reproduced transmitted the variant to their offspring. In conclusion, our study shows that genome sequencing fails to identify rare variants in new genes related to TSC, and a third gene is therefore unlikely to exist. We demonstrated that at least one out of 10 patients with a clinical diagnosis of TSC carries a mosaic pathogenic variant in TSC1 or TSC2. We also showed that individuals with mosaic variants have a distinctive phenotypic severity, with important implications for surveillance.

Thanks to the affordability of DNA sequencing, hundreds of thousands of individuals with rare disorders are undergoing whole-genome sequencing in an effort to reveal novel disease aetiologies, increase our understanding of biological processes and improve patient care. However, the power to discover the genetic causes of many unexplained rare diseases is hindered by a paucity of cases with a shared molecular aetiology. This thesis presents research into statistical and computational methods for determining the genetic causes of rare diseases. Methods described herein treat important aspects of the nature of rare diseases, including genetic and phenotypic heterogeneity, phenotypes involving multiple organ systems, Mendelian modes of inheritance and the incorporation of complex prior information such as model organism phenotypes and evolutionary conservation. The complex nature of rare disease phenotypes and the need to aggregate patient data across many centres has led to the adoption of the Human Phenotype Ontology (HPO) as a means of coding patient phenotypes. The HPO provides a standardised vocabulary and captures relationships between disease features. I developed a suite of software packages dubbed 'ontologyX' in order to simplify analysis and visualisation of such ontologically encoded data, and enable them to be incorporated into complex analysis methods. An important aspect of the analysis of ontological data is quantifying the semantic similarity between ontologically annotated entities, which is implemented in the ontologyX software. We employed this functionality in a phenotypic similarity regression framework, 'SimReg', which models the relationship between ontologically encoded patient phenotypes of individuals and rare variation in a given genomic locus. It does so by evaluating support for a model under which the probability that a person carries rare alleles in a locus depends on the similarity between the person's ontologically encoded phenotype and a latent characteristic phenotype which can be inferred from data. A probability of association is computed by comparison of the two models, allowing prioritisation of candidate loci for involvement in disease with respect to a heterogeneous collection of disease phenotypes. SimReg includes a sophisticated treatment of HPO-coded phenotypic data but dichotomises the genetic data at a locus. Therefore, we developed an additional method, 'BeviMed', standing for Bayesian Evaluation of Variant Involvement in Mendelian Disease, which evaluates the evidence of association between allele configurations across rare variants within a genomic locus and a case/control label. It is capable of inferring the probability of association, and conditional on association, the probability of each mode of inheritance and probability of involvement of each variant. Inference is performed through a Bayesian comparison of multiple models: under a baseline model disease risk is independent of allele configuration at the given rare variant sites and under an alternate model disease risk depends on the configuration of alleles, a latent partition of variants into pathogenic and non-pathogenic groups and a mode of inheritance. The method can be used to analyse a dataset comprising thousands of individuals genotyped at hundreds of rare variant sites in a fraction of a second, making it much faster than competing methods and facilitating genome-wide application.

Infectious diseases are a major cause of morbidity and mortality worldwide. Streptococcus pneumoniae and Neisseria meningitidis are major causes of severe bacterial disease which can manifest as invasive disease such as bacteraemia and meningitis. Exposure to these pathogens is relatively widespread, yet only a minority of individuals develop invasive disease. A host genetic component to infectious disease susceptibility has been implied from twin and adoptee studies. A role for rare large effect genetic variants in predisposition to infection has been demonstrated through the study of individuals with primary immunodeficiencies. However, a majority of these studies have been undertaken in individuals with a history of recurrent disease or in multi-case families. The relative role of rare genetic variants of moderate to large effect at the population level has not been widely explored. This thesis presents effort made using next generation sequencing methods to identify rare genetic variants that lead to increased susceptibility to bacterial disease focussing on meningococcal disease, pleural infection(empyema), pneumococcal disease and sepsis phenotypes. Using an exome sequencing approach in 13 cases with invasive meningococcal disease, a novel mutation leading to a complement deficiency and increased risk of meningococcal infection was identified and functionally validated in one individual. This mutation in the CFP gene was demonstrated as leading to impaired properdin secretion. Further analysis implicated loss of function mutations in CD4 and ZAP70 as novel loci for meningococcal disease susceptibility. A case control association analysis for sepsis susceptibility highlighted the possible role for small Rho GTPases in sepsis pathology. By aggregating all rare predicted deleterious mutations in a gene, four genes in this pathway, (ROCK2, ARHGAP18, FYN and CDC42BPG) were implicated as having an excess of rare deleterious variants in sepsis samples compared to population controls. A similar approach identified low frequency genetic variants in the CD109 gene as predisposing to empyema susceptibility in children. Finally, preliminary evidence from adult individuals with invasive pneumococcal disease points to a potential role of the RNASE7 gene in invasive pneumococcal disease susceptibility. This association was primarily due to a predicted deleterious missense mutation present in cases and absent in controls. Taken together, these results have identified a number of potential loci with rare variants associated with susceptibility to severe phenotypes of bacterial diseases.

Advances in high-throughput genomic technologies have facilitated the collection of DNA information for thousands of individuals, providing unprecedented opportunities to explore the genetic architecture of complex disease. One important finding has been that the majority of variants in the human genome are low in frequency or rare. It has been hypothesised that recent explosive growth of the human population afforded unexpectedly large amounts of rare variants with potentially deleterious effects, suggesting that rare variants may play a role in disease predisposition. But, importantly, rare variants embody a source of information through which we may learn more about our recent evolutionary history. In this thesis, I developed several statistical and computational methods to address problems associated with the analysis of rare variants and, foremost, to leverage the genealogical information they encode. First, one constraint in genome-wide association studies is that lower-frequency variants are not well captured by genotyping methods, but instead are predicted through imputation from a reference dataset. I developed the meta-imputation method to improve imputation accuracy by integrating genotype data from multiple, independent reference panels, which outperformed imputations from separate references in almost all comparisons (mean correlation with masked genotypes r²>0.9). I further demonstrated in simulated case-control studies that meta-imputation increased the statistical power to identify low-frequency variants of intermediate or high penetrance by 2.2-3.6%. Second, rare variants are likely to have originated recently through mutation and thereby sit on relatively long haplotype regions identical by descent (IBD). I developed a method that exploits rare variants as identifiers for shared haplotype segments around which the breakpoints of recombination are detected using non-probabilistic approaches. In coalescent simulations, I show that such breakpoints can be inferred with high accuracy (r²>0.99) around rare variants at frequencies <0.05%, using either haplotype or genotype data. Third, I show that technical error poses a major problem for the analysis of whole-genome sequencing or genotyping data, particularly for alleles below 0.05% frequency (false positive rate, FPR=0.1). I therefore propose a novel approach to infer IBD segments using a Hidden Markov Model (HMM) which operates on genotype data alone. I incorporated an empirical error model constructed from error rates I estimated in publicly available sequencing and genotyping datasets. The HMM was robust in presence of error in simulated data (r²>0.98) while nonprobabilistic methods failed (r²<0.02). Lastly, the age of an allele (the time since its creation through mutation) may provide clues about demographic processes that resulted in its observed frequency. I present a novel method to estimate (rare) allele age based on the inferred shared haplotype structure of the sample. The method operates in a Bayesian framework to infer pairwise coalescent times from which the age is estimated using a composite posterior approach. I show in simulated data that coalescent time can be inferred with high accuracy (rank correlation >0.91) which resulted in a likewise high accuracy for estimated age (>0.94). When applied to data from the 1000 Genomes Project, I show that estimated age distributions were overall conform with frequency-dependent expectations under neutrality, but where patterns of low frequency and old age may hint at signatures of selection at certain sites. Thus, this method may prove useful in the analysis of large cohorts when linked to biomedical phenotype data.

In the last decades, several studies have explored the human microbiota in different body niches, notably in the gastrointestinal tract. The gut microbiota has been recognized as an additional organ that co-evolves with humans and interacts with host physiology. The gut microbiota plays a key role in the host metabolism and the extraction of energy from food. Diet is the strongest factor shaping the gut microbiota: differences in macronutrient intake may select the growth of specific microbial taxa. Indeed, the diet provides different substrates such as undigested carbohydrates for fermentation, affecting the composition of the gut microbiota and, consequently, the production of microbial metabolites. Short chain fatty acids (SCFAs), mainly acetate, propionate, and butyrate, are the major end products of anaerobic fermentation by gut microbes. Modifications in the proportions of the gut microbes and, consequently, the production of SCFAs have been suggested to play a role in several pathological conditions. If the diet is modifiable to stimulate the growth of beneficial microbes, it could represent a target for prevention and treatment of a possible dysbiosis. In congenital metabolic disorders such as phenylketonuria (PKU) and glycogen storage disease (GSD), diet is considered the mainstay for the treatment of the pathology. PKU diet is based on the restriction of phenylalanine intake and the supplementation of essential amino acids and micronutrients. PKU diet is similar to a vegan alimentation since all the animal products are excluded. GSD dietary management provides the intake of slow-release carbohydrates to prevent hypoglycaemia and limits the assumption of simple sugars. In this research, we aimed to analyze the impact of dietary therapy on the composition of the gut microbiota and the production of SCFAs in two congenital metabolic disorders (PKU and GSD). PKU and GSD diseases showed an alteration of the gut microbiome. If diet represents the only treatment of the disease and is not modifiable, a personalized intervention to restore a healthy gut microbiome is to consider. Nowadays, the attention is focused on the supplementation of probiotics and prebiotics.

Epidermolysis Bullosa simplex (EBS) is a rare genetic disorder that is typically inherited in an autosomal dominant fashion and affects approximately 1 out of 20 000 individuals. This disease is caused by mutations in either the KRT14, KRT5 or PLEC genes. These genes code for proteins involved in the formation of the cytoskeleton in basal keratinocytes, which form the basal layer of the epidermis. The cytoskeleton provides structural support to the basal keratinocytes and mutations in these genes cause cytoskeletal malformations, making these cells more susceptible to physical stress. This results in the cells undergoing lysis under trivial mechanical stress and causing the epidermis to detach from the dermis, the layer immediately below the epidermis. This leads to the primary symptom of EBS: the formation of blisters. The goal of this project is to recreate EBS in zebrafish using transgenesis and to create stable mutant transgenic line. In the future, high throughput drug screening will be done on mutant zebrafish embryos to find potential drug candidates that can alleviate the symptoms of EBS. To accomplish this, missense and deletion mutations in zebrafish krt5 cDNA using site-directed mutagenesis were performed. It was previously shown that mice models for this disease die shortly after birth and thus no stable mutant lines were able to be created. To ensure embryo survival and avoid a similar fate, mutant krt5 cDNA was expressed in non-essential tissue, such as the embryonic fin fold using a fin epithelial-specific enhancer named epi. These constructs were injected into one-cell stage zebrafish embryos, which were raised and screened for integration of the construct in their germ cells. While results from injected embryos were promising, mutant transgenic zebrafish did not demonstrate any blistering. In an attempt to induce blistering, mutant zebrafish embryos were placed under various environmental stressors known to worsen the symptoms of EBS. This was not successful. Expression of mutant keratin 5 in the basal epidermis of the entire embryo using the 2.3kb upstream region of the zebrafish krt5 gene to drive expression also did not yield any results. More investigations are needed to determine if it will be possible to use the zebrafish to model EBS.

Sequencing exomes—the 1% of the genome that codes for proteins—has increased the rate at which the genetic basis of a patient’s disease is determined. Unfortunately, when a patient does not carry a well-established pathogenic variant, it is extremely challenging to establish which of the tens of thousands of variants identified in that individual is contributing to their disease. In these situations, variants must be prioritized to make further investigation more manageable. In this thesis, we have focused on creating statistical frameworks and models to aid in the interpretation of rare variants and towards establishing gene-level metrics for variant prioritization. We developed a sensitive and specific workflow to detect newly arising (de novo) variants from exome sequencing data of parent-child trios, and created a sequence-context based mutational. This mutational model was the basis of a rigorous statistical framework to evaluate the significance of de novo variant burden not only globally, but also per gene. When we applied this framework to de novo variants identified in patients with an autism spectrum disorder, we found a global excess of de novo loss-of-function variants as well as two genes that harbored significantly more de novo loss-of-function variants than expected. We also used the mutational model to predict the expected number of rare (minor allele frequency < 0.1%) variants in exome sequencing datasets of reference individuals. We found a significant depletion of missense and loss-of-function variants in a subset of genes, indicating that these genes are under strong evolutionary constraint. Specifically, we identified 3,230 genes that are intolerant of loss-of-function variation and that set of genes is enriched for established dominant and haploinsufficient disease genes. Similarly, we searched for regions within genes that were intolerant of missense variation. The most missense depleted 15% of the exome contains 83% of reported pathogenic variants found in haploinsufficient disease genes that cause severe disease. Additionally, both gene-level and region-level constraint metrics highlight a set of de novo variants from patients with a neurodevelopmental disorder that are more likely to be pathogenic, supporting the utility of these metrics when interpreting rare variants within the context of disease.
Medical Sciences

Dissertation (PhD(Agric))--University of Stellenbosch, 2004.
ENGLISH ABSTRACT: This study was carried out to investigate whether environmental and/or genetic factors had an effect on the incidence of epistaxis related to exercise-induced pulmonary haemorrhage among racehorses in Southern Africa. A further aim was to estimate the heritability of liability to epistaxis in the Southern African Thoroughbred population. For the purpose of the environmental study, the data covering the period 1986-2001 and involving a total of778 532-race runs, were analysed. This included the following race start information: date of race (day/month/year), age, sex, breeder, trainer, distance, jockey, state of going, weight carried, centre of racing and altitude. The genetic part of the data was two-fold in nature and included firstly the analysis of all horses that suffered epistaxis whilst racing in Southern Africa from 1986 to 2001 and involving 1118 individual bleeders. The second genetic analysis included the same Southern African population plus those Southern African horses exported to Mauritius and then being recorded as bleeders in that country (1252 bleeders in total). Pedigree data covering the period 1960-1986 was used as required to calculate the incidence of bleeding amongst ancestors of the post 1986 era. Only pedigrees of horses that raced were included in this study as it was not possible to predict whether non-runners would have bled had they raced. Consequently all non-runners and also those that raced overseas in countries where bleeding occurrence was not recorded were excluded. Veterinarians employed by the Jockey Club suspended officially recorded horses that showed epistaxis as demonstrated by frank bleeding from the nostrils after racing. Oncourse endoscopy is not employed as a routine on any of the Southern African racetracks. In the environmental study epistaxis was identified in 1 287 race starts (0.165%). Epistaxis related to exercise-induced pulmonary haemorrhage was significantly (p<0.001) associated with altitude, age, race year, month and the day of racing. More horses demonstrated epistaxis at sea level than at altitude, between the months of May - October than the rest of the year, in older horses than in horses less than three-years old, after 1995 than between the years 1986 and 1995, and on Fridays and Sundays than on any other week day. No association could be established between epistaxis and breeder, trainer, distance, jockey, state of going, sex and weight carried. The heritability of liability method as described by Falconer (1989) was used to estimate the relative importance of heredity and environment. For the period investigated, the population incidence for epistaxis in Southern African horses was 2.1%. The estimation of heritability ofliability showed that first-degree relatives had a figure of 55.4%. The heritability of second- and third degree relatives were 41.3% and 30.4% respectively. The data investigated depicts horses that bled almost exclusively on race days as only a small percentage (- 5%) was reported as having bled during exercise. Accordingly, the full extent of epistaxis amongst racing Thoroughbreds in Southern Africa is difficult to gauge. Pedigree and race run data from Thoroughbreds racing in Southern Africa, covering the period 1986-2002 (63 146) horses in pedigree data-set and 778 532 race runs, were further analysed in order to study genetic and environmental factors affecting the incidence of epistaxis as associated with EIPH (exercise-induced pulmonary haemorrhage). As fixed effects for the model, variables that were tested significantly in a preliminary data analyses, were included. Various combinations of such variables namely age, weight, altitude, sex, month and going were tested. Fixed effects that were included in the fmal model were gender, going and altitude. The heritability estimates from a logit transformed analysis for epistaxis fitting both the animal and sire generalized mixed models were 0.23 and 0.40 respectively, which indicated that epistaxis as associated with EIPH in the Southern African Thoroughbred sires has a strong genetic basis. Genetic trends indicating an increase in epistaxis were also found. It is concluded that the frequency of epistaxis related to pulmonary haemorrhage is associated with altitude, winter and spring months and the horse's age. It is suggested that racing at a lower altitude may increase the probability of exercise-induced pulmonary haemorrhage. It is clear that epistaxis in the racing Thoroughbred has a strong genetic basis. It is further suggested that horses showing frank bleeding from the nostrils after racing or exercise, be suspended and not used for breeding purposes. This would result in relatively fast progress being made towards eradicating this costly scourge of the modem Thoroughbred racehorse. Affected stallions and those racing whilst being treated with furosemide, should be barred from breeding and not be considered as future sires. Estimated breeding values for epistaxis should be used as a tool for selecting against it and be considered in breeding programmes to decrease the incidence thereof.
AFRIKAANSE OPSOMMING: 'n GENETIESE ANALISE VAN DIE VOORKOMS VAN LONGBLOEDING IN DIE SUID-AFRIKAANSE RENPERD: Die doel met hierdie studie was om vas te stelof omgewings- of genetiese faktore enige invloed op die voorkoms van longbloeding in die Suid-Afrikaanse renperd het. 'n Verdere doelstelling was om die oorerflikheid op die onderliggende verspreiding van longbloeding in die Suid-Afrikaanse Volbloedpopulasie te bepaal. Vir die omgewingstudie is data wat oor die periode 1986-2001 strek en wat 'n totaal van 778 532 wedren-deelnames ingesluit het, statisties ondersoek. Die data het die volgende inligting ingesluit: datum van deelname, ouderdom, geslag, teler, afrigter, afstand van wedren, jokkie, toestand van baanoppervlakte, gewig gedra, sentrum waar deelname plaasgevind het en die hoogte bo seespieël van die sentrum. Die studie van die genetiese aspekte het eerstens 'n analise van al die perde wat longbloeding tydens 'n wedren in Suider-Afrika gedurende die jare 1986-2002 ondervind het (I118 perde), en tweedens dieselfde populasie perde, plus die Suiderlike-Afrikaanse perde wat uitgevoer is na Mauritius en bloeding daar ondervind het, (1252 perde), ingesluit. Ter aanvulling is uitgebreide stamboomdata van voorouers gedurende 1960-1986 gebruik om die voorkoms van longbloeding tydens die post 1986 tydvak te bepaal. Slegs stambome van renperde wat aktief aan renne deelgeneem het, is in die data ingesluit aangesien dit nie moontlik was om te voorspel of 'n perd wat nooit aan wedrenne deelgeneem het nie, longbloeding sou ondervind indien dit wel deelgeneem het. Dus is alle renperde wat nooit aan wedrenne deelgeneem het, asook daardie perde wat in die buiteland deelgeneem en waar longbloeding nie aangeteken word nie, uitgesluit. Alle perde wat bloeding van die neus na wedrenne getoon het, is deur veeartse in diens van die Jokkie Klub van Suid-Afrika ondersoek, as 'n bloeier aangeteken en van verdere deelname aan wedrenne geskors. Endoskopie word op geen van die Suid- Afrikaanse renbane as 'n standaard praktyk na wedrenne uitgevoer nie. Longbloeding het in 1 287 perde of gedurende 0.165% van alle wedrenne plaasgevind. Longbloeding soos geassossieer met EIPH, (exercise-induced pulmonary haemorrhage), is betekenisvol (p<0.001) met hoogte bo seespieël, ouderdom, dag van deelname, maand, en jaar verbind. Meer perde het longbloeding by seevlak in vergelyking met hoër vlakke bo seespieël ondervind, tussen die maande Mei-Oktober as die res van die jaar, in perde ouer as drie-jaar, na 1995 as tussen die jare 1986-1995, op Vrydae en Sondae as enige ander dag van die week. en meer by reuns as by merries of hingste. Geen verwantskap kon tussen bloeding en teler, afrigter, afstand, jokkie, toestand van baan, geslag en gewig gedra, gevind word nie. Die oorerflikheid op die onderliggende verspreiding vir longbloeding soos omskryf deur Falconer (1989), is gebruik om die relatiewe belangrikheid van oorerflikheid en omgewing te bepaal. Vir die periode bestudeer, was die voorkoms van longbloeding in die Suid-Afrikaanse renperd 2.1%. Die oorerflikheid van longbloeding was 55.4% vir eerste-graadse verwantes. By tweede-graadse verwantes was die ooretlikheid 41.3% en by derde-graadse verwantes 30.4%. Die data wat ondersoek is, was bykans uitsluitlik die van perde wat tydens wedrenne gebloei het en slegs 'n baie klein persentasie (~ 5%) was aangeteken as perde wat tydens oefening gebloei het. Dus is die volle omvang van longbloeding in Suider-Afrikaanse Volbloedperde moeilik om akkuraat te bepaal. Die stamboom- en wedrendata van Suid-Afrikaanse Volbloedperde is verder ontleed in 'n poging om die genetiese en omgewingsfaktore se invloed op die voorkoms van longbloeding, soos geassosieer met EIPH te bepaal. As vaste effekte vir die model is veranderlikes wat betekenisvol gevind was, ingesluit. Verskeie kombinasies van hierdie veranderlikes soos ouderdom, gewig, hoogte bo seespieël, geslag, maand en toestand van die baan is ingesluit. Die vaste effekte wat in die finale model ingesluit is, was geslag, toestand van die baan en hoogte bo seespieël. Die beraamde oorerflikheid verkry vanaf 'n "logit" getransformeerde analise vir longbloeding wat beide die diere- en vader- gemengde model gepas het, was onderskeidelik 0.23 en 0.40, wat 'n aanduidending is dat longbloeding, soos geassosieer met ElPH, 'n sterk genetiese grondslag het. Genetiese tendense het ook gedui op 'n toename in die voorkoms van longbloeding, veraloor die laaste vyf jaar van die studie. Samevattend is die bevinding dat die frekwensie van longbloeding 'n betekenisvolle verwantskap toon met hoogte bo seespieël, winter en lente maande en die perd se ouderdom. Dit word voorgestel dat renperde wat deelneem aan wedrenne by laer vlakke van hoogte bo seespieël, meer onderhewig aan longbloeding sal wees. Uit die resultate verkry is dit duidelik dat longbloeding 'n genetiese grondslag het. Dit word voorgestel dat perde wat fisiese simptome van neusbloeding na of gedurende wedrenne toon, geskors word van verdere deelname en ook nie toegelaat word om mee te teel nie. Hierdie maatreëls behoort aanleiding te gee dat relatief vinnige vordering gemaak sal word in die strewe om hierdie ongewenste sindroom in die moderne Volbloed te verminder. Aangetaste hingste, asook die wat aan wedrenne deelgeneem het terwyl hul behandeling ontvang met furosemide, moet nie toegelaat word om te teel en nie as toekomstige teelhingste oorweeg word nie. Die waarde van voorspelde teelwaardes vir longbloeding moet nie onderskat word in seleksie daarteen nie en moet in teelprogamme om die voorkoms daarvan te verminder, oorweeg word.

Cette thèse porte sur le développement d'une méthode pour modéliser des systèmes complexes en utilisant des graphes de connaissances et des algorithmes de raisonnement automatisés. La méthode de modélisation a été appliquée aux maladies rares pour prédire leurs causes, depuis les niveaux génétique, cellulaire et physiologique jusqu'au niveau de l'organisme entier. Pour la création du graphe de connaissances, deux ontologies, GO et HPO, ont été utilisées. Étant donné qu'il n'existait pas de bases de données avec des relations entre ces ontologies, une méthode d'apprentissage automatique a été développée pour inférer des relations et appliquée aux ontologies GO et HPO. La thèse est complétée par une méthode d'apprentissage automatique pour inférer des effets délétères après une variation génétique appelée INDEL. Dans l'ensemble, le travail d'intelligence artificielle présenté dans cette thèse de doctorat aide les chercheurs à comprendre ce qui se passe dans le corps humain à différents niveaux d'abstraction, depuis l'apparition d'une variation génétique jusqu'au développement d'une maladie rare
This thesis is about the development of a method for modeling complex systems using knowledge graphs and automated reasoning algorithms. The modeling method was applied to rare diseases to predict their causes from the genetic to the cellular, physiological, and whole organism levels. For the creation of the knowledge graph, two ontologies, GO and HPO, were used. Since there were no databases with relationships between these ontologies, a machine learning method was developed to infer relationships and applied to both GO and HPO ontologies. The thesis is completed by a machine learning method to infer deleterious effects after a genetic variation called INDEL. Altogether, the artificial intelligence work presented in this doctoral thesis assists rare disease researchers in understanding what happens in the human body at various levels of abstraction, from the occurrence of a genetic variation to the development of a rare disease

Le malattie mitocondriali sono disturbi genetici caratterizzati da difetti di fosforilazione ossidativa causati da mutazioni nel DNA mitocondriale, o in geni nucleari i cui prodotti sono legati alla fosforilazione ossidativa o alla biologia mitocondriale. La prima parte del progetto è stata focalizzata sulla generazione e la caratterizzazione di un modello murino di malattia mitocondriale, Ttc19ko. I pazienti con mutazioni in TTC19 sviluppano danni neurologici e deficit di complesso III. Ttc19 è una proteina mitocondriale che sembra essere coinvolta nell’assemblaggio e/o stabilità del complesso III. Abbiamo dimostrato che il modello murino Ttc19ko ha sintomi neurologici, debolezza muscolare e riduzione dell’attività locomotoria spontanea, in analogia con la malattia umana. L’analisi istologica ha rivelato alcune anomalie neurologiche con presenza di accumuli di ubiquitina e GFAP. I topi Ttc19ko hanno una riduzione complessiva del metabolismo energetico, una diminuzione del consumo di O2 e di produzione di CO2. L’attività enzimatica del complesso III è significativamente ridotta nei tessuti e ciò è legato ad un aumento della produzione di ROS. L’analisi BNGE ha mostrato una riduzione della incorporazione della subunità catalitica Rieske nel complesso assemblato. L’immunoprecipitazione di TTC19-Flag in colture cellulari trattate con amminoacidi marcati ha rivelato una maggiore interazione tra Ttc19 e le subunità del pre-complesso III, e una minore interazione con proteine Rieske e Uqcrh, entrambe assemblate tardivamente. Abbiamo inoltre dimostrato che Ttc19 rimane associata al complesso III assemblato. Nell’insieme, questi risultati suggeriscono che Ttc19 è un fattore intrinseco di assemblaggio del complesso III che interagisce con il pre-complesso III facilitando così l'incorporazione della proteina Rieske. La seconda parte del progetto è stata focalizzata sulla messa a punto di una terapia genica su un altro modello murino di malattia mitocondriale, MPV17ko. Mutazioni in MPV17 causano una sindrome epatocerebrale con deplezione del mtDNA, insufficienza epatica a esordio precoce, gravi crisi ipoglicemiche e morte. Il trapianto di fegato e l'alimentazione frequente a base di carboidrati a lento rilascio sono le uniche terapie disponibili, anche se in seguito si sviluppano danni neurologici. Il ruolo fisiologico di MPV17 non è chiaro. Abbiamo dimostrato che MPV17 fa parte di un complesso ad alto peso molecolare a composizione sconosciuta, che è essenziale per il mantenimento del mtDNA nel fegato. In dieta standard, il topo MPV17-/- non mostra quasi alcun sintomo di disfunzione epatica, ma una dieta chetogenica porta questi animali a sviluppare cirrosi e insufficienza epatica grave. Tuttavia, quando l'espressione di MPV17 è ripristinata dalla somministrazione di virus adeno-associato, si assiste ad un ricostituzione del complesso supramolecolare contenente Mpv17, ad un ripristino completo del numero di copie di mtDNA, ed alla prevenzione dell’insufficienza epatica indotta dal dieta chetogenica. Questi risultati aprono nuove prospettive terapeutiche per il trattamento delle sindromi da deplezione del mtDNA indotte da mutazioni nel gene MPV17.
Mitochondrial diseases are genetic disorders characterized by defects in oxidative phosphorylation caused by mutations in mitochondrial DNA, or in nuclear genes whose products are related to oxidative phosphorylation or mitochondrial biology. The first part of the project was focused on the generation and characterization of a mouse model of mitochondrial disease, Ttc19ko. Patients with mutations in TTC19 were characterized by neurological impairments and mitochondrial respiratory complex III deficiency. Ttc19 is a mitochondrial protein that seems to be associated to complex III assembly and/or stability. We showed that Ttc19ko mice have neurological symptoms, muscular weakness and reduction in spontaneous locomotors activity, clearly resembling the human disease. Brain also had neurological abnormalities with presence of ubiquitin and GFAP positive staining. Comprehensive lab animals monitoring system revealed a reduction in O2 consumption, CO2 production and energy expenditure in Ttc19ko mice, indicating an overall reduction of energy metabolism. Complex III activity was significantly reduced in tissues and this was linked to an increased ROS production. BNGE analysis of mitochondrial complex III showed a substantial reduction of the incorporation of the catalytic Rieske iron-sulfur protein into the fully assembled complex. A stable isotope labelling by amino acids in cell culture (SILAC) expressing TTC19-Flag followed by immunoprecipitation and mass spec analysis revealed a higher scored interaction between Ttc19 and the subunits of the pre-complexIII, and a lower scored interaction with Rieske protein and Uqcrh, both of them are late assembled subunits. We also demonstrated that Ttc19 is associated to the fully assembled complex III. Taken together, these results suggests that Ttc19 is an intrinsic assembly factor of complex III that interacts with the pre-complex III thus facilitating the incorporation of the late assembled Rieske protein. The second part of the project was focused on a gene therapy approach on a second mouse model of mitochondrial disease, MPv17ko. Mutations in hMPV17 cause a hepatocerebral form of mtDNA depletion syndrome hallmarked by early-onset liver failure, leading to premature death. Liver transplantation and frequent feeding using slow-release carbohydrates are the only available therapies, although surviving patients develop slowly progressive neuropathy. The physiological role of Mpv17 is still unclear. We showed that Mpv17 is part of a high molecular weight complex of unknown composition, which is essential for mtDNA maintenance in liver. On a standard diet, Mpv17ko mouse shows hardly any symptom of liver dysfunction, but a ketogenic diet leads these animals to liver cirrhosis and failure. However, when expression of human MPV17 is carried out by adeno-associated virus mediated gene replacement, the Mpv17ko mice are able to reconstitute the Mpv17-containing supramolecular complex, restore liver mtDNA copy number and oxidative phosphorylation proficiency and prevent liver failure induced by the KD. These results open new therapeutic perspectives for the treatment of MPV17-related liver-specific MDS.

La forte hétérogénéité génétique et les modes de transmission complexes des maladies rares posent le défi d'identifier le variant causal si un seul patient le porte, en utilisant des données de séquençage et des méthodes d'analyse standard. Pour aborder ce problème, la méthode PSAP utilise des distributions nulles par gène de scores de pathogénicité CADD pour évaluer la probabilité d'observer un génotype donné dans la population générale. L'objectif de ce travail était de répondre au manque de diagnostic des maladies rares grâce à des méthodes statistiques. Nous proposons PSAP-genomic-regions, une extension de la méthode PSAP au génome non codant, en utilisant comme unités de test des régions prédéfinies reflétant la contrainte fonctionnelle à l'échelle du génome entier. Nous avons implémenté PSAP-genomic regions et sa version initiale PSAP-genes dans Easy-PSAP, un workflow Snakemake intuitif et adaptable, accessible aussi bien aux chercheurs qu'aux cliniciens. Appliqué à des familles touchées par de l'infertilité masculine, Easy-PSAP a permis la priorisation de variants candidats pertinents dans des gènes connus et nouveaux. Nous nous sommes ensuite concentrés sur le digénisme, le mode le plus simple de transmission complexe, qui implique l'altération simultanée de deux gènes pour développer une maladie. Nous avons décrit et évalué les méthodes actuelles publiées dans la littérature pour détecter le digénisme et proposé de nouvelles stratégies pour améliorer le diagnostic de ce mode de transmission complexe
High genetic heterogeneity and complex modes of inheritance in rare diseases pose the challenge of identifying an n-of-one sequencing data and standard analysis methods. To tackle this issue, the PSAP method uses gene-specific null distributions of CADD pathogenicity scores to assess the probability of observing a given genotype in a healthy population. The goal of this work was to address rare disease lack of diagnosis through statistical strategies. We propose PSAP-genomic-regions an extension of the PSAP method to the non-coding genome, using as testing units predefined regions reflecting functional constraint at the scale of the whole genome.We implemented PSAP-genomic-regions and the initial PSAP-genes in Easy-PSAP a user-friendly and versatile Snakemake workflow, accessible to both researchers and clinicians. When applied to families affected by male infertility, Easy-PSAP allowed the prioritization of relevant candidate variants in known and novel genes. We then focused on digenism, the most simple mode of complex inheritance, which implicates the simultaneous alteration of two genes to develop a disease. We reviewed and benchmarked current methods in the literature to detect digenism and put forward new strategies to improve the diagnostic of this complex mode of inheritance

Internationally there is no single definition of a rare disease: in European countries and beyond is determined on the basis of different thresholds of prevalence, in accordance with regulatory requirements: in the European Union (EU), a disease is rare if it has a population prevalence of less than or equal to 1/2000 inhabitants (0.05 %). Rare diseases, from a clinical and epidemiological point of view-, are a very heterogeneous group, with about 5000-8000 entities, mostly with genetic origin, often severe, chronically debilitating, progressive and potentially fatal, representing a major public health problem; nevertheless, in the past, have received little scientific interest, are hardly included in international classifications and barely visible at the level of health information, for many of them also epidemiologic data are not yet available. The main reason seems to be the absence of a system of correct coding and traceability in health information systems: the existing classification systems for collecting phenotypic and genotypic data, are widely heterogeneous, lacking a standard terminology, slightly inter-operable, solely dedicated to the classification of the phenotype or supplied with database features of variants and related phenotypes profiles. Some of these databases are already operating, other developing, along with the implementation of NGS technologies, which guaranteed the medical scientific community, compared to traditional techniques, the study of a wide range of genetic diseases, with the collapse of cost and time of data analysis. Worldwide were established centers and research projects in highly specialized techniques of genome sequencing and development of databases of genomic variants, that have made available a large variety and quantity of new information, often difficult to manage and interpret. The commitment of the scientific community is therefore the systematic organization of that huge amount of data, potentially useful to the correct interpretation of the pathogenic significance of genomic variants in relation to the clinical phenotype. Based on emerging data from the literature, the records of patients and pathologies constitute interesting tools for the development of knowledge useful for the clinical and epidemiological aspect and in the field of rare diseases. Many existing registers must address issues of quality, accessibility and standardization of data, optimization of resources, and interoperability, related to fragmentation of coding systems of diseases. The situation on European and international scene seems to favor the establishment of global registers or common platforms for coordination of all existing and future registers, following a uniform approach, coordinated and of the highest quality. The implementation of this model is currently a strategic objective of the European community within the project EPIRARE and of the US community in the project on Global Rare Diseases Registry (GRDR). In the evolutionary path that supports the implementation of the public health service to improve the exercise of the right to healthcare of citizens, the current development of innovative information technologies, which make available comprehensive information, shared and able to guarantee citizens the best continuity of care, is critical. In Italy, the Electronic Health Record (FSE) is the computerized collection of data of the clinical history of an individual, designed as part of the National Health Service (SSN) according to the National Guidelines issued by the Ministry of Health in November 2010. The process is slow, but with the introduction of the SFE is to be hoped that in the near future, we witness a revolutionary change in the relationship between health infrastructure and citizen, so far characterized by communication barriers, rigid and cumbersome bureaucratic procedures and paperwork with improvement of the flow of data and the diagnostic process and healthcare. Also in the international context is documented the usefulness of the so-called "Electronic Health Record" (EHR) or "Personal Health Record" (PHR) to assist this process in healthcare, as well as for measuring and monitoring the quality of care and for translational research, but the construction and development of the FSE are still at a preliminary stage and the challenges of making the application infrastructure are numerous, especially for developing countries. Another major challenge on many fronts to health systems in the field of rare diseases is that represented by the heterogeneous group of patients without diagnosis, individuals with known diseases incorrectly recognized, experiencing delays and diagnostic errors, or individuals with unknown diseases; these patients over-use inadequate diagnostic pathways and their care needs are not satisfied, with high human and social costs. To date the problem of patients without diagnosis was universally recognized worldwide as central, to the planning of health care networks for patients with a rare disease, but its still unresolved: the attempts made in the USA with Undiagnosed Diseases Program (UDP) of National Institute of Health (NIH) clinical Center have shown promising results, but appeared poorly sustainable in clinical practice and still have limited epidemiological value. Nationally some Accredited Centres for excellence in the diagnosis and treatment of specific rare diseases or groups of related rare diseases, have established projects targeted on finding solutions to methodological collection of clinical phenotype-genotype correspondence, supporting the diagnostic process of patients with rare diseases, with special attention to patients without a diagnosis or with a generic presumptive diagnosis: Region-University research Project "Next Generation Sequencing and Gene Therapy to diagnose and Cure rare Diseases in Emilia Romagna (rarer)" and Ministerial Project "A multicenter collaborative research network for the identification and study of rare undiagnosed patients: the impact on the rare disease National Health Service network (UnRareNet)". In the broader context of these research projects present PhD project designed to create a unique collection of analytical data, describing phenotype and genotype features of patients with known or suspected diagnosis, between groups of selected rare diseases, based on a computerized platform, with implementation of the model of the Regional Registry for Rare Diseases, already used in the Veneto Region. A further objective was the creation of a specific product useful for clinical management in the field of rare diseases, especially with respect to efficiency, speed and accuracy of diagnostic assessment, through the transfer of the instrument for collection of clinical phenotype-genotype data in patients with diseases or groups of selected rare diseases in the form of generalized Computerized Medical Record (CCI). The form CCI should have features which cut across rare diseases registry and the hospital medical records, dedicated to a single episode of care, and should structur on the emerging pattern of computerized personal health record, such as in Italy, the "Fascicolo Sanitario Elettronico". The added value is the integration of the relational database research potential with the clinical applications of CCI in the practical management and care of patients with rare diseases. It was also planned the networking of CCI as part of a shared use of clinical, genetic and healthcare data in real contexts of clinical excellence, such as those offered by the RARER and UnRareNet projects. The project is structured in three different phases: 1. identification and sharing, within the common RARER protocol, of a suitable methodology for the construction of 8 different disease registries, with implementation of the clinical database and structuring a form of collection of genotypic data; 2. realization and network sharing of module CCI such as a sustainable product in a clinical setting, containing clinical information in a relational database; 3. implementation of the acquisition data system and module CCI, with the expansion of the patient population in the context of the project UnRareNet, data analysis, identifying the correlation between genotype and phenotype pictures, and design of an expert system able to manage complex procedures for data processing. The research project proposed the implementation of a relational database with the collection through CCI of clinical data in patients with rare diseases with known or suspected diagnosis, in the context of some rare diseases of specific interest. The clinical entities and major and minor entities, describing phenotype and genotype features are classified in the database according to a classificatory nodular and multihierarchical like a tree method, which integrates uni and multidimensional internationally recognized classification systems and, at the same timem respects the anatomical structure and function of the human organism. The database includes 72,360 records, distributed between the main and secondary entities (signs, symptoms, comorbidity, impairments, diagnostic, attributes) and relationships between entities, 117 different clinical entities, 63 different genes and 1,186 total patients. The results obtained during the development and implementation of the system of collection of clinical information, with the implementation of the module CCI in the context of the project RARER, have proved the usability of the product in a preliminary context; the expansion of the clinical application in the context of the project UnRareNet, has produced promising results and the system showed a good performance in terms of similar and dissimilar phenotypes, managing fully clinical information regarding total number of different rare diseases and proving thus be generalized flexible and interoperable. The data are still limited at the time, but having analytical characteristics pre-coded, are immediately usable for future complex processing. The collection of genetic information and integration with clinical information appeared feasible with promising results, although the project UnRareNet this is still absolutely at a preliminary stage; patients without diagnosis is about 10% of the total, but more than half of them do not yet have genetic investigation, which proved to be useful to confirm the diagnosis in 53% of cases with known disease. On the total sample of patients, the use of NGS technologies seems to be limited at this stage, but the results highlight the potential of the modern techniques of genome sequencing in the implementation of the diagnostic process; expanding the population of patients enrolled and the targeted application of these new technologies could allow wider knowledge on the phenotypic and genotypic profile of the considered diseases. Having a broader patient population, the creation of the expert system, of which the foundations were laid in the context of this project, will be possible through the collection of theoretical frequencies of clinical and diagnostic findings of some groups of rare diseases related (metabolic, mitochondrial), arising from the international literature and to be included in the system for processing of clinical scenarios. The expert system will allow the validation of diagnoses simulated by collecting new cases with known diagnosis, and the formulation of new diagnoses of known diseases or identification of new clinical entities, even with use of neural networks or Kohonen networks and Fuzzy methods. The methodology for collecting phenotypic and genotypic information in patients with rare diseases, proposed in this PhD project, proved sufficient to satisfy the requirements of applicability and suggested a number of future applications, which aim the expansion of knowledge on genomic variability of rare diseases and to reduce the number of patients without a diagnosis.
A livello internazionale non esiste una definizione univoca di malattia rara: paesi europei ed extraeuropei la stabiliscono in base a differenti soglie di prevalenza in accordo alle legislazioni vigenti: nell'Unione Europea (UE), una patologia è rara se presenta una prevalenza di popolazione uguale o inferiore a 1/2000 abitanti (0,05%). Le malattie rare sono, sotto il profilo clinico-epidemiologico, un gruppo molto eterogeneo, con circa 5000-8000 entità per lo più ad origine genetica, spesso gravi, cronicamente invalidanti, progressive e potenzialmente fatali, che rappresentano un grave problema di salute pubblica; nonostante ciò, sono state in passato di scarso interesse scientifico, sono poco incluse nelle classificazioni internazionali e poco visibili a livello d'informazione sanitaria, per molte di esse inoltre non sono ancora disponibili dati epidemiologici. La ragione principale sembra essere l'assenza di un sistema di corretta codifica e rintracciabilità nei sistemi informativi sanitari: gli attuali sistemi classificativi e di raccolta dell'informazione fenotipica e genotipica sono ampiamente eterogenei, privi di terminologia standard, scarsamente inter-operabili, unicamente dedicati alla classificazione del fenotipo, o con caratteristiche di database genomici con profili fenotipi variante-correlati. Alcuni di questi database, sono già operativi, altri in via di sviluppo, di pari passo con l'implementazione delle tecnologie NGS, che hanno garantito alla comunità medico scientifica, rispetto alle tecniche tradizionali, lo studio di una vastissima gamma di malattie genetiche, con crollo dei costi e dei tempi d'analisi dei dati. A livello mondiale sono nati centri e progetti di ricerca altamente specializzati in tecniche di sequenziamento genomico e nello sviluppo di database di varianti genomiche, che hanno reso disponibile una grande varietà e quantità di nuove informazioni, spesso di difficile gestione ed interpretazione. L'imperativo della comunità scientifica è pertanto l'organizzazione sistematica di questa enorme quantità di dati, potenzialmente utili alla corretta interpretazione del significato patogeno delle varianti genomiche in relazione al fenotipo clinico. In base a dati emergenti dalla letteratura, i registri di pazienti e di patologia costituiscono strumenti interessanti per lo sviluppo di conoscenze utili sotto il profilo epidemiologico-clinico in ambito di malattie rare. Molti registri esistenti devono affrontare problematiche di qualità, standardizzazione e accessibilità dei dati, di ottimizzazione delle risorse e di interoperabilità legate alla frammentazione dei sistemi di codifica delle malattie. La situazione a livello europeo ed extra-europeo sembra favorire la costituzione di registri globali o di piattaforme comuni di coordinamento di tutti i registri esistenti e futuri, seguendo un approccio uniforme, coordinato e di alta qualità. La realizzazione di questo modello è attualmente un obiettivo strategico della comunità europea nell'ambito del progetto EPIRARE e della comunità statunitense nell'ambito del progetto sul Global Rare Diseases Registry (GRDR). Nel percorso evolutivo che sostiene l'implementazione del servizio sanitario pubblico per migliorare l'esercizio del diritto alla salute del cittadino è fondamentale l'attuale sviluppo di tecnologie informatiche innovative che rendano disponibile informazioni complete, condivise ed idonee a garantire al cittadino la migliore continuità assistenziale. In Italia Fascicolo Sanitario Elettronico ('FSE') è la raccolta informatizzata dei dati della storia clinica di un individuo, progettata nell'ambito del Servizio Sanitario Nazionale (SSN) secondo le Linee Guida Nazionali emanate dal Ministero della Salute nel novembre 2010. Il processo è lento, ma con l'introduzione del FSE è auspicabile che nel prossimo futuro, si assista ad un rivoluzionario cambiamento dei rapporti tra infrastruttura sanitaria e cittadino, con miglioramento del flusso di dati e del processo diagnostico-assistenziale. Anche nel contesto internazionale è documentata l'utilità del cosiddetto 'Electronic Health Record' (EHR) o 'Personal Health Record' (PHR) a supporto di tale processo in ambito sanitario, oltre che per la misurazione ed il monitoraggio della qualità delle cure e per la ricerca scientifica traslazionale, ma realizzazione e lo sviluppo del FSE sono ancora in fase preliminare e le sfide da affrontare per rendere applicativa tale infrastruttura sono numerose, soprattutto per i paesi in via di sviluppo. In ambito di malattie rare altra sfida importante su molti fronti per i sistemi sanitari è quella rappresentata dal gruppo eterogeneo dei pazienti senza diagnosi, ovvero gli individui con malattie note non correttamente riconosciute, che subiscono ritardi ed errori diagnostici, o gli individui con patologie sconosciute; questi pazienti sovra- utilizzano percorsi diagnostico-terapeutici inadeguati e presentano bisogni assistenziali non soddisfatti, con elevati costi umani e sociali. Ad oggi il problema dei pazienti senza diagnosi è stato universalmente riconosciuto a livello mondiale come centrale nella programmazione sanitaria della rete di assistenza per malati rari, ma appare ancora irrisolto: i tentativi messi in atto in USA con Undiagnosed Diseases Program (UDP) del National Institute of Health (NIH) Clinical Center hanno dato risultati promettenti, ma sono apparsi ancora scarsamente sostenibili nella pratica clinica e di ancora limitato valore epidemiologico. A livello nazionale alcuni Centri Accreditati per eccellenza nella diagnosi e cura di specifiche malattie rare o gruppi di malattie rare correlate, hanno istituito progetti mirati alla ricerca di soluzioni metodologiche per raccolta dell'informazione clinica nella sua corrispondenza fenotipo-genotipica, a sostegno del processo diagnostico- assistenziale dei malati rari, con particolare riguardo ai pazienti senza diagnosi o con un generico sospetto diagnostico: Progetto di ricerca Regione-Università 'Next Generation Sequencing and Gene Therapy to Diagnose and Cure Rare Diseases in Regione Emilia Romagna (RARER)' e Progetto Ministeriale 'A multicenter collaborative research network for the identification and study of rare undiagnosed patients: the impact on the rare disease National Health Service network (UnRareNet)'. Nel più ampio contesto dei suddetti progetti di ricerca il presente progetto di dottorato si è posto l'obiettivo primario di realizzare uno strumento unico di raccolta di dati analitici descriventi il fenotipo e il genotipo di pazienti con diagnosi nota o sospetta, all'interno di gruppi di malattie rare selezionate, basato su una piattaforma informatizzata, con implementazione del modello del Registro Regionale per le Malattie Rare già in uso in Regione Veneto. Ulteriore obbiettivo è stata la creazione di un prodotto specifico utile alla gestione clinica in ambito di malattie rare, soprattutto rispetto ad efficienza, rapidità e correttezza d'inquadramento diagnostico, attraverso il trasferimento dello strumento descrittivo di raccolta dell'informazione clinica fenotipo-genotipica di pazienti con malattie o gruppi di malattie rare selezionate, nel modulo generalizzato di Cartella Clinica Informatizzata (CCI). Il modulo CCI deve possedere caratteristiche trasversali rispetto ai registri di malattia ed alle cartelle informatizzate ospedaliere, dedicate ad un singolo episodio di cura e strutturarsi sul modello emergente dei record sanitari individuali informatizzati, quali ad esempio in Italia, il Fascicolo Sanitario Elettronico individuale. Il valore aggiunto è l'integrazione delle potenzialità di ricerca, offerte dalla struttura del database relazionale, con le applicazioni cliniche del modulo CCI nella gestione pratica del percorso assistenziale e di cura dei malati rari. Si è programmato inoltre l'inserimento in rete del modulo CCI, nell'ambito di un utilizzo condiviso dell'informazione clinica, genetica ed assistenziale in contesti reali d'assistenza clinica d'eccellenza, quali quelli offerti dai progetti RARER e UnRareNet. Il progetto si è strutturato in 3 diverse fasi: 1. identificazione e condivisione, all'interno del protocollo comune RARER, della metodologia idonea alla creazione di 8 registri di malattie rare, con implementazione delle banca dati clinica e strutturazione di un modulo di raccolta del dato genotipico; 2. realizzazione e condivisione in rete del modulo CCI, come prodotto sostenibile in ambito clinico, contenente l'informazione clinica fenotipica e genotipica descritta all'interno di un database relazionale; 3. implementazione sistema d'acquisizione dati e del modulo CCI con l'ampliamento della popolazione nel contesto del progetto UnRareNet, analisi dei dati, con identificazione della correlazione tra quadri fenotipici e genotipo e progettazione di un sistema esperto in grado di gestire procedure complesse di elaborazione dati. Il progetto di ricerca ha previsto la realizzazione del database relazionale con raccolta di dati clinici di malati rari con diagnosi nota o sospetta nell'ambito di alcune patologie rare di specifico interesse attraverso il modulo CCI. Le entità nosologiche e le entità principali e secondarie del database, descriventi il fenotipo ed il genotipico del paziente sono classificate e quindi organizzate secondo una logica classificatoria nodulare e multigerarchica ad albero, che integra i sistemi classificativi uni e multidimensionali riconosciuti a livello internazionale e contemporaneamente rispetta la suddivisone anatomica e funzionale dell'organismo umano. Il database contiene 72360 record, distribuiti tra entità principali e secondarie (segni, sintomi, comorbidità, menomazioni, indagini diagnostiche, attributi) e relazioni tra entità, 117 differenti entità nosologiche, 63 differenti geni e 1186 pazienti totali. I risultati ottenuti nella fase di creazione ed implementazione del sistema di raccolta dell'informazione clinica, con la messa in atto del modulo CCI nel contesto del progetto RARER, hanno dimostrato l'usabilità del prodotto in un ambito preliminare; l'ampliamento dell'applicazione clinica nel contesto del progetto UnRareNet, ha prodotto risultati promettenti ed il sistema ha mostrato una buona tenuta in condizioni di fenotipi simili e dissimili, gestendo in modo completo informazioni cliniche riguardanti complessivamente numerose differenti malattie rare e dimostrandosi quindi generalizzabile, flessibile ed inter-operabile. I dati contenuti sono al momento ancora limitati, ma possedendo caratteristiche analitiche e precodificate appaiono subito utilizzabili per future elaborazioni complesse. La raccolta dell'informazione genetica e l'integrazione con l'informazione clinica sono apparse anch'esse fattibili, con risultati promettenti, seppur nell'ambito del progetto UnRareNet ancora in fase assolutamente preliminare; i pazienti senza diagnosi sono circa il 10% circa del totale, ma di questi più della metà non dispone ancora d'indagine genetica, che si è rivelata utile alla conferma diagnostica nel 53% dei casi con malattia nota. Sul totale dei pazienti del campione l'utilizzo di tecnologie NGS appare in questa fase limitato, ma i risultati ottenuti sottolineano le potenzialità delle moderne tecniche di sequenziamento genomico nell'implemetazione del processo diagnostico; l'ampliamento della popolazione di pazienti arruolati e l'applicazione mirata di queste tecnologie potrebbe consentire nuova più ampia conoscenza sul profilo fenotipico e genotipico delle malattie considerate. Disponendo di una più ampia popolazione di pazienti sarà possibile l'istruzione del sistema esperto, di cui si sono gettate le basi nel contesto del presente progetto, attraverso la raccolta di frequenze teoriche di segni clinici e reperti diagnostici di alcuni gruppi di malattie rare correlate (metaboliche, mitocondriali), derivanti dalla letteratura internazionale e destinate al sistema per elaborazione di casi clinici simulati. Il sistema esperto potrà consentire la validazione di diagnosi simulate attraverso la raccolta di nuovi casi con diagnosi nota e la formulazione di nuove diagnosi di patologie note o l'identificazione di nuove entità nosologiche, anche con utilizzo di reti neurali o reti di kohonen e metodologie Fuzzy. La metodologia di raccolta dell'informazione fenotipica e genotipica del paziente con malattia rara, proposta nel presente progetto di dottorato si è dimostrata idonea a soddisfare i requisiti d'applicabilità richiesti ed ha suggerito una serie di applicazioni future che mirano all'ampliamento delle conoscenze sulla variabilità genomica della malattie rare ed alla riduzione del numero dei malti senza diagnosi.

Whole exome sequencing (WES) is a powerful tool to identify new molecules involved in skeletal homeostasis. In particular we used WES to establish the molecular diagnosis of two particular skeletal diseases: osteopetrosis and the acrofrontofacionasal dysostosis 1 (AFFND1). The osteopetroses are a group of rare bone diseases characterized by increased bone density due to the failure in bone resorption. Due to their genetic heterogeneity, WES represents a valuable strategy to identify the genetic defect. We analyzed osteopetrotic patients with autosomal dominant osteopetrosis (ADO) and autosomal recessive osteopetrosis (ARO), which is the most severe form. In our cohort we performed molecular diagnosis of 4 ADOI or ADOII patients that carried mutations in the LRP5 and CLCN7 genes, respectively. The analysis of ARO patients confirmed TCIRG1 as the most frequently mutated gene, identified mutations in the other known ARO genes and in genes very rarely associated with osteopetrosis, namely FERMT3 and USB1. Of note, we demonstrated the causative role of four deep intronic mutations in TCIRG1 gene and two different synonymous changes in the TCIRG1 and CLCN7 genes in the pathogenesis of the disease. In addition, WES helped in the differential diagnosis in a patient who was found to bear a mutation in the FAM20C gene, known to cause Raine syndrome. Regarding AFFND1, this is an extremely rare syndrome, comprising facial and skeletal abnormalities, short stature and intellectual disability. WES found a novel truncating mutation in the neuroblastoma-amplified sequence (NBAS) gene in two Indian patients (c.6237-3C>G). This mutation impaired NBAS functions in HEK293T cells overexpressing the truncated NBAS protein. Furthermore, we demonstrated that NBAS expression in mouse embryos was compatible with a role in bone and brain development and that the depletion of endogenous z-nbas in fish embryos resulted in defective morphogenesis of chondrogenic cranial skeletal elements. Overall, we provided evidence supporting the hypothesis of a causative role of the mutated NBAS gene in the pathogenesis of AFFND1. In conclusion, we effectively exploited WES in the genetic diagnosis of rare skeletal diseases. We also highlighted potential limitations of this approach, specifically with respect to deep intronic mutations and synonymous changes, and underlined the importance to complement WES with analysis at the transcript level and functional validation, when possible.

La découverte du défaut moléculaire en cause dans les maladies rares est une étape importante en vue d'un traitement spécifique ainsi que d'un meilleur diagnostic, ce qui permet de réduire le délai diagnostique, de mieux connaître l'histoire naturelle de la maladie, et ainsi d'améliorer les traitements symptomatiques et la prévention secondaire. Les gènes de plus de 1500 maladies rares monogéniques restent à découvrir. Beaucoup de maladies rares qui frappent les enfants de parents en bonne santé correspondent à des maladies génétiques récessives autosomiques. Certaines paraissent extrêmement rares mais, une fois le gène identifié dans une famille princeps, beaucoup d'autres cas s'avèrent dus à des défauts du même gène. Les patients que nous étudions sont issus de familles consanguines comptant souvent de nombreux sujets atteints. Une seule famille de ce type peut permettre l'identification du gène par cartographie d'homozygotie et clonage positionnel.

Nous avons recruté dans ce travail des cas familiaux ou sporadiques de six maladies autosomiques récessives rares de gène inconnu.

La stratégie de cartographie par homozygotie nous a permis de mettre en évidence de nouveaux loci morbides dans quatre de ces maladies (épilepsie myoclonique progressive EPM3 ;syndrome marfanoïde avec microsphérophakie ;atrophie optique isolée ;et syndrome de microcéphalie et diabète précoce) ou de réduire la taille de loci déjà connus (microcéphalies primaires MCPH2 et MCPH4 ;et syndrome de Harboyan CDPD1). Nous avons pu caractériser de nouvelles mutations dans les gènes déjà connus ASPM (microcéphalie primaire MCPH5) et SLC4A11 (syndrome de Harboyan) et corréler celles-ci aux données cliniques. Enfin nous avons identifié les gènes KCTD7 et LTBP2 comme responsables respectivement des maladies EPM3 et syndrome marfanoïde avec microsphérophakie, en y découvrant des mutations chez les malades.

Doctorat en Sciences médicales
info:eu-repo/semantics/nonPublished

L’avènement du séquençage haut débit d’exome (SHD-E) en diagnostic et en recherche ces dernières années a conduit à l’identification des bases génétiques de nombreuses pathologies mendéliennes, permettant de résoudre de nombreuses situations d’errance diagnostique. Néanmoins, l’analyse des données de SHD-E permet uniquement d’identifier des variations pathogènes ou probablement pathogènes dans 30 à 45 % des situations sans diagnostic. En effet, certaines limites existent, tant au niveau clinique, moléculaire et bioinformatique. L’évolution constante des connaissances cliniques, du nombre de nouveaux gènes impliqués en pathologie humaine, et des corrélations clinico-biologique a un impact important sur l’analyse des données, entraînant une amélioration progressive de la recherche diagnostique. Des limites techniques inhérentes à la technologie, avec en particulier des régions non couvertes, existent, mais se sont également significativement réduites ces dernières années. Enfin, au-delà de l’analyse de SNV et de CNV, d’autres anomalies génétiques peuvent être responsables de maladies rares, nécessitant un développement bioinformatique pour optimiser les résultats. Bien que le séquençage à haut débit du génome permette de résoudre des observations, en particulier en cas de variations dans les régions non codantes ou les variants de structure, il existe encore de nombreuses informations à extraire et à exploiter à partir des données de SHD-E.L’objectif de cette thèse a donc été de participer à l’amélioration des approches bioinformatiques d’analyse de données de SHD-E pour l’identification de nouveaux gènes ou mécanismes moléculaires impliqués dans des maladies génétiques rares afin de réduire l’errance diagnostique des patients.Plusieurs stratégies ont ainsi été mises en place. La première stratégie a consisté en une réanalyse recherche de données de 80 patients ayant bénéficié d’un SHD-E au laboratoire CERBA (thèse CIFRE) dont la lecture diagnostique était négative. Elle a conduit à la mise en évidence deux nouveaux gènes candidats dans la déficience intellectuelle syndromique, dont le gène OTUD7A (article 1). La deuxième stratégie a consisté en la mise au point d’un pipeline bioinformatique pour extraire les données du génome mitochondrial à partir des données de SHD-E. L’ADN mitochondrial n’est pas ciblé par les kits de capture d’exome mais peut être extrait des données capturées indirectement rendant son analyse possible à partir de données de SHD-E préexistantes. A partir de la collection GAD d’exomes de patients sans diagnostic, deux variations causales ont été identifiées chez deux individus atteints de troubles neuro-développementaux sur 928 personnes étudiées, et ainsi résoudre une errance diagnostique dans 0,2 % des patients sans diagnostic (article 2). La troisième stratégie a consisté en la mise en place d’un pipeline bioinformatique d’identification des éléments mobiles au sein des données d’exome, étant attendu qu’environ 0,3 % des variations pathogènes du génome humain ont pour origine l’insertion de novo d’un élément mobile. A partir de la collection GAD d’exomes de 3322 patients sans diagnostic, cette étape a permis d’identifier deux cas en lien avec l’insertion d’un élément Alu au sein d’un exon du gène FERMT1 et du gène GRIN2B (article 3 en cours d’écriture).Cette thèse a permis de repousser certaines limites de la technologie d’exome. D’autres perspectives existent, et sont explorées par l’équipe, en lien avec le projet Européen Solve-RD
In the last years, the advent of exome sequencing (ES) in diagnosis and in research led to the identification of the genetic bases of many Mendelian disorders, allowing many diagnostic wavering cases to be solved. Nevertheless, ES data analysis only leads to the identification of pathogenic or likely pathogenic variants in 30 to 45 % of the undiagnosed cases. Indeed, some limits exist, both at clinical, molecular and bioinformatic levels. The constant evolution of the clinical knowledge, of the number of genes involved in human diseases, and of the clinical-biological correlations, has a significant impact on data analysis, leading to a progressive improvement in diagnostic research. Limits of the current technologies, especially not covered regions, exist, but have been significantly reduced in the recent years. Although genome sequencing will solve some undiagnosed cases, especially in case of non-coding or structural variants, there is still a lot of information to be extracted and analyzed from ES data. Finally, beyond SNV and CNV analyzes, other genetic events can be involved in rare disorders, requiring a bioinformatic development to optimize results.The aim of the project was therefore to improve bioinformatic approaches of ES data analysis in order to identify new molecular mechanisms involved in rare genetic disorders and reduce diagnostic wavering.Several strategies were established. The first one consisted in reanalysing ES data from 80 undiagnosed patients, who were sequenced by the Laboratoire CERBA (CIFRE thesis). It led to the identification of 2 new candidate genes involved in ID, especially OTUD7A gene (article 1). The second strategy was the development of a bioinformatic pipeline in order to extract mitochondrial DNA data from ES data. The mitochondrial genome is not targeted by exome capture kits but can be extracted from off-target data, giving the opportunity to analyze it from preexisting ES data. From the GAD exomes cohort of undiagnosed patients, 2 causal variations were identified in 2 individuals out of 928, affected with neuro-developmental disorder. It thus solved the diagnostic wavering in 0.2 % of patients without diagnosis (article 2). The third strategy consisted in the development of a bioinformatic pipeline to identify mobile elements insertion within ES data, with the expectation that about 0.03 % of the pathogenic variants originate from de novo mobile element insertion. From the GAD exomes cohort of 3322 undiagnosed patients, this step led to the identification of two Alu element insertions in FERMT1 and GRIN2B gene exons (article 3, in process).This PhD permitted to push out some ES limits. Other perspectives exist, and are explored by the GAD team, in connection with the European Solve-RD project

Nineteen cultivars and fourteen breeding lines were evaluated for partial resistance to crown rust Puccinia coronata race 264. Multivariate statistical methods such as principal component and cluster analyses were employed to identify significant resistance parameters and to group oat genotypes with similar rust resistance characteristics. This involved two separate investigations consisting of two experiments each, the first experiment conducted under field conditions and the second conducted under growth bench conditions. From both of the investigations a group of oat genotypes with high partial resistance to P. coronata race 264 was obtained. These are OA 712-17, OA 712-33, Glen, Woodstock, QO 220.13, and QO 574.21. These oat genotypes are currently being used as parents in crosses in the Macdonald Campus of McGill University oat breeding program.

L’identification des mutations génétiques impliquées dans les maladies rares est un prérequis pour mieux les comprendre, les traiter et accompagner les patients. Pour se faire, des modèles animaux présentant des maladies spontanées homologues aux maladies humaines sont très prometteurs. Le chien développe spontanément des maladies génétiques, rares chez l’Homme, mais fréquentes dans certaines races de chiens, ce qui simplifie les analyses génétiques. Ma thèse a porté sur deux maladies neurologiques : l’épilepsie et la neuropathie. Pour l’épilepsie, l’objectif était d’identifier des variants génétiques à partir de données de génotypage et de séquençage de génomes complets de deux races canines prédisposées. Un locus lié à la maladie a été identifié dans une race et des variants ponctuels et structuraux candidats ont été identifiés dans les deux races et sont en cours de validation par séquençage ciblé. Pour la neuropathie, l’équipe avait identifié une mutation en amont du gène GDNF, responsable d’une neuropathie sensitive chez des chiens de chasse. J’ai participé à la validation fonctionnelle de cette mutation. De plus, GDNF étant un excellent gène candidat pour les neuropathies humaines, j’ai séquencé ce gène chez 111 patients et extrait les variants de GDNF d’une base de données d’exomes et de génomes de plus de 600 patients. J’ai ainsi identifié 21 variants rares ou inconnus et les ai priorisé selon leurs impacts prédits in silico. Ces deux projets, alliant analyses génétiques, génomiques et fonctionnelles, chez l’homme et le chien, montrent le potentiel du chien pour l’identification de gènes candidats dans des maladies rares et/ou complexes chez l’Homme
The identification of genetic mutations involved in rare diseases is a prerequisite for a better understanding, therapies and care to patients. To this aim, animal models declaring spontaneous diseases, homologous to human diseases are very promising. Dogs spontaneously develop genetic diseases, rare in humans, but frequent in some dog breeds, which simplifies the genetic analyzes. My thesis focused on two neurological diseases: epilepsy and neuropathy. For epilepsy, the goal was to identify genetic variants from genotyping data and sequencing of whole genome of dogs from two predisposed breeds. A disease-related locus has been identified in one breed and candidate point mutations and structural variants were identified in the two breeds and are being validated by targeted sequencing. For neuropathy, the team previously identified a mutation upstream of the GDNF gene, responsible for sensory neuropathy in hunting dogs. I participated to the functional validation of this mutation. In addition, GDNF being an excellent candidate gene for human neuropathies, I sequenced this gene in 111 patients and extracted GDNF variants from a database of exomes and genomes from more than 600 patients. I identified 21 rare or unknown variants and prioritized them according to their in silico predicted impacts. These two projects, combining genetics, genomics and functional analyses, in humans and dogs, show the dog's potential for identifying candidate genes in rare and / or complex diseases in humans

Ankylosing spondylitis (AS) is a common inflammatory arthritis of the spine and other affected joints, which is highly heritable, being strongly influenced by the HLA-B27 status, as well as hundreds of mostly unknown genetic variants of smaller effect. The aim of my research was to confirm some of the previously observed genetic associations and to identify new associations, many of which are in biological pathways relevant to AS pathogenesis, most notably the IL-23/T_H17 axis (IL23R) and antigen presentation (ERAP1 and ERAP2). Studies presented in this thesis include replication and refinement of several potential associations initially identified by earlier GWAS (WTCCC-TASC, 2007 and TASC, 2010). I conducted an extended study of IL23R association with AS and undertook a meta-analysis, confirming the association between AS and IL23R (non-synonymous SNP rs11209026, p=1.5 x 10-9, OR=0.61). An extensive re-sequencing and fine mapping project, including a meta-analysis, to replicate and refine the association of TNFRSF1A with AS was also undertaken; a novel variant in intron 6 was identified and a weak association with a low frequency variant, rs4149584 (p=0.01, OR=1.58), was detected. Somewhat stronger associations were seen with rs4149577 (p=0.002, OR=0.91) and rs4149578 (p=0.015, OR=1.14) in the meta-analysis. Associations at several additional loci had been identified by a more recent GWAS (WTCCC2-TASC, 2011). I used in silico techniques, including imputation using a denser panel of variants from the 1000 Genomes Project, conditional analysis and rare/low frequency variant analysis, to refine these associations. Imputation analysis (1782 cases/5167 controls) revealed novel associations with ERAP2 (rs4869313, p=7.3 x 10-8, OR=0.79) and several additional candidate loci including IL6R, UBE2L3 and 2p16.3. Ten SNPs were then directly typed in an independent sample (1804 cases/1848 controls) to replicate selected associations and to determine the imputation accuracy. I established that imputation using the 1000 Genomes Project pilot data was largely reliable, specifically for common variants (genotype concordence~97%). However, more accurate imputation of low frequency variants may require larger reference populations, like the most recent 1000 Genomes reference panels. The results of my research provide a better understanding of the complex genetics of AS, and help identify future targets for genetic and functional studies.

Background. DNA sequencing is a field of evolving molecular biology. In the last decade, new generation sequencing technologies - known as Next Generation Sequencing (NGS) - have become increasingly common in clinical practice. These new technologies, by their characteristics, are completely revolutionized to the analysis of the human genome. Aim of the thesis. The aim of the thesis is to verify the applicability of the new generation sequencing methods (NGS). In this regard, the Illumina and Agilent different massively-parallel sequencing strategies were compared, which were adopted for the study of patients analyzed at the Genetics Unit of the Hospital of Padua. The research activity is also focused on the development of models for the pathogenicity validation and the characterization of the genetic variants identified by NGS sequencing. A new variant in the COQ4 gene was identified in a patient with coenzyme Q deficiency syndrome and its potential ability to alter the splicing mechanism was studied by hybrid minigene system. A functional complementation in S. cerevisiae yeast was performed in order to examine and validate the pathogenicity of some missense mutations in the ASS1 gene, identified in patients with citrullinemia type 1, with the intent to extending the analysis to the potential new variants identifiable in the future. Results and conclusions. The comparative analysis of the NGS technologies TruSeq Custom Amplicon Low Input Library Prep Kit (Illumina) and HaloPlex HS Target Enrichment System (Agilent Technologies) showed how the second one is more efficient in diagnostic, considering the followed parameters: coverage, results variability, pseudogenes and false positive discrimination, number of amplicons generated for the analysis. For these reasons, from 2017 we use only the HaloPlex HS Target Enrichment System (Agilent Technologies). In the study, 85 patients positive for neonatal metabolic screening for Mucopolysaccharidosis type I (MPS I) or for Biotinidase Deficiency or Phenylketonuria (PKU), have been analyzed by Next Generation Sequencing: molecular analysis made it possible to distinguish true positive patients from false positives to neonatal metabolic screening (it would be important to review the cutoff of this screening), allowing to avoid unnecessary therapies and significantly reduce the cost of the investigation. The hybrid minigene system used for the validation of the new IVS4+1G>A variant in COQ4 gene was considered an effective and simple method to demonstrate the pathogenicity of the mutation: it behaves like a hypomorph allele causing the activation of a cryptic splicing site at exon 4 of the COQ4 gene producing both the wild-type transcript and a smaller one less of about 100 bp. Finally, the functional complementation in S. cerevisiae yeast system developed to test the pathogenicity of missense mutations in the ASS1 gene, allowed to classify the mutants in the two classes: class I mutations that completely abolish yeast growth, class II mutations that use a residual growth. This system was useful to validate the pathogenicity of the mutations studied but not to made correlation between the genotype and the phenotype of the patients.

Transaldolase-Mangel (TALDO) ist ein extrem seltener, angeborener Stoffwechseldefekt, von dem weltweit nur 34 Fälle bekannt sind. Der Defekt geht auf den Verlust des Enzyms Transaldolase 1 aus dem nicht-oxidativen Pentosephosphat-Weg (nicht-oxPPW) zurück und äußert sich in einem weiten Spektrum klinischer Symptome. Die schwerwiegendsten Folgen sind Leber- und Nierenmangelfunktionen, die zum sehr frühen Tod führen können. Desweiteren leiden 15 % der Patienten an wiederkehrenden Infektionen. Neutrophile Granulozyten (Neutrophile) sind die häufigsten weißen Blutkörperchen im Menschen und essentiell für die angeborene Immunantwort gegen Infektionserreger. Ich habe hier funktionale Aspekte von TALDO-Neutrophilen untersucht. Der oxidative Pentosephosphat-Weg (oxPPW) stellt das Reduktionsäquivalent NADPH bereit, welches indirekt für die Entstehung von reactive oxygen species (ROS)-abhängigen Neutrophil Extracellular Traps (NETs) verantwortlich ist. Der Beitrag des nicht-oxPPW zur ROS-abhängigen NET-Bildung ist bislang nicht bekannt. In dieser Arbeit konnte ich für Neutrophile aus drei TALDO-Patienten eine jeweils komplett abwesende Entstehung ROS-abhängiger NETs und einen deutlich verringerten oxidativen Burst nach PMA-Stimulation zeigen. Um diese Beobachtungen in einem unabhängigen Modelsystem zu bestätigen, habe ich mit Hilfe des CRISPR-Cas9-Systems, ‚knock-out‘ Mutanten von Transaldolase 1 und dessen Partnerenzym Transketolase in der Neutrophil-ähnlichen Zelllinie PLB-985 hergestellt. Die dergestalt genetisch manipulierten Zellen waren nicht mehr zu PMA-induziertem Zelltod in der Lage. Dies ist somit der erste genetische Beweis für die Abhängigkeit des oxidativen Burst und der Bildung von NETs vom nicht-oxPPW. Diese Erkenntnis trägt zum einen zum mechanistischen Verständnis der NET-Entstehung bei und liefert zum anderen eine potentielle Erklärung für einige der bei TALDO beobachteten Symptome. Desweiteren wurden einige der metabolischen Erfordernisse für die Bildung von NETs mit Hilfe von Inhibitoren untersucht. Die erhaltenen Erkenntnisse zeigen, dass das initiale Maximum des oxidativen Bursts für NET-Bildung unerheblich ist und vielmehr die ROS-Generierung nach ca. 50 Minuten entscheidende Bedeutung für diese hat.
Transdaldolase 1-deficiency (TALDO) is a rare genetic disease with only 34 described cases globally. Transaldolase 1 is part of the non-oxidative pentose phosphate pathway (PPP) and its deficiency results in many clinical symptoms including kidney and liver failure, which can lead to early child-mortality. Some of these patients suffer from recurrent infections, for example in the respiratory tract. Neutrophils are the most abundant white blood cells and essential for the innate immune defence against bacterial and fungal pathogens. The PPP generates reduced NADPH that is crucial for the generation of superoxide by the NADPH oxidase NOX2. In turn, NOX2 is essential for neutrophil extracellular trap (NET) formation. NETs occur through the neutrophil-specific cell death netosis and consist of chromatin decorated with granular proteins. Here I report that neutrophils of three TALDO patients did not make NETs. Deletion of transaldolase 1, and its partner enzyme transketolase, in the neutrophil-like PLB-985 cell line reduced ROS generation and cell death. This confirms that transaldolase 1 is required for NET formation. We present, to the best of our knowledge, the first genetic evidence that the non-oxidative PPP is required for ROS generation and NET formation. Furthermore, some of the metabolic requirements for NET formation were assessed. The obtained data indicate that the initial peak of the oxidative burst is irrelevant for NET formation but the ROS generation after 50 minutes on the contrary has crucial significance.

Le diagnostic des rétinopathies pigmentaires pose différents problèmes, au niveau clinique comme au niveau moléculaire. En premier lieu, il s'agit de maladies rares, la faible prévalence de chaque pathologie à l'échelle de la population mondiale rend difficile leur étude. En second lieu, la caractérisation phénotypique de ces maladies est délicate car les symptômes qui en découlent s'avèrent très similaires. De manière liée, l'œil et le processus de la vision s'avèrent complexes et impliquent les produits d'expression de nombreux gènes. Ainsi, bien que les rétinopathies soient majoritairement monogénique et respectent le modèle d'hérédité mendélienne, les causes génétiques des maladies sont variées. Sur la base de ce double constat, nous proposons deux approches méthodologiques complémentaires menant à une meilleure compréhension de ce groupe de pathologies. Une première approche a pour finalité l'acquisition du jeu exhaustif des gènes impliqués. Les travaux portent sur l'exploitation des puces de génotypage. Nous effectuons une étude de liaison génétique entre les variations ponctuelles et les pathologies. Une seconde approche porte sur la représentation des connaissances associées aux phénotypes cliniques. Un composant ontologique est construit afin d'expliciter les savoirs nécessaires au diagnostic. Les données collectées sur le long terme par les experts sont étiquetées au travers de termes organisés au sein d'un thésaurus dédié. Les profils cliniques des patients et des maladies sont manipulés sous forme de collections de caractéristiques et comparés au moyen d'une mesure de similarité adaptée. L'objectif est alors de proposer un système d'aide au diagnostic
Diagnosis of retinitis pigmentosa could be difficult regarding both to clinics or molecular issues. Firstly, there are rare diseases, so the prevalence of each pathology in the world population is very low. Secondly, the symptoms of diseases are very similar, so their phenotypic characterization is hard. Moreover, the eye and the visual process are complex and numerous genes' products are implicated. Although retinopathies are mainly monogenic and mendelian inherited diseases, the polymorphisms involved in these diseases are very diverse.These both observations lead us to develop two complementary methodological approaches in a view to better understand the retinopathies.The first approach aims to identify all the genes involved in the diseases using genotyping chips. For this purpose, we studied genetic linkage between single nucleotide variations and pathologies. The second approach leads to the representation of clinical knowledge. An ontological compound was built to make explicit the knowledge involved in the process of diagnosis. The data previously collected by experts were labeled by terms that were organized in a specific thesaurus. The clinic profiles of the patients and diseases were handled as features collections and were compared by similarity calculations. The goal of this work is to build a knowledge-based system for diagnosis

Introducción: El síndrome Cutaneo Esquelético Fosfatémico (SCEF) es una enfermedad mosaica ultrarara causada por mutaciones RAS postzygóticas que se define por la asociación de nevus epidérmicos y/o melanocíticos, sobreproducción de factor de crecimiento fibroblastico-23 (FGF23), y lesiones displásicas óseas. Dada su rareza, el SCEF no ha sido caracterizado adecuadamente. Asimismo, se desconoce el papel de las mutaciones RAS en la génesis de las lesiones óseas, así como la fisiopatología de la sobreproducción de FGF23 en el SCEF. Hipótesis: 1) El fenotipado de pacientes con SCEF así como el análisis de todos los casos de SCEF publicados proveerán información clínica relevante, 2) Las lesiones óseas displásicas son la fuente de sobreproducción de FGF23 en el SCEF, y 3) La creación de modelos experimentales adecuados informarán sobre la fisiopatología del SCEF. Objetivos: 1) Caracterizar el SCEF clínicamente mediante el estudio de una cohorte de sujetos y una revisión bibliográfica de todos los casos publicados, 2) Identificar el tejido responsable de la sobreproducción de FGF23 en el SCEF, y 3) Investigar los efectos de las mutaciones RAS en el desarrollo de lesiones óseas. Sujetos y métodos: 5 sujetos con SCEF fueron fenotipados de manera exhaustiva. Se identificaron 51 casos en la revisión bibliográfica compatibles con SCEF. Se recogieron y analizaron datos sobre la histología ósea, alteraciones esqueléticas, minerales, cutáneas y de otros tejidos, así como las respuestas a diferentes tratamientos para la hipofosfatemia en los sujetos de la cohorte y la literatura. Se evaluaron los efectos de las mutaciones RAS en la producción/expresión de FGF23 en células humanas mesenquimales de médula ósea (hBMSCs) y en células IDG-SW3 transducidas con adenovirus portadores de mutaciones RAS, en hBMSCs mutantes procedentes de un paciente con SCEF, y en células IDG-SW3 tratadas con medio de cultivo de hBMSCs mutantes. Se trasplantaron subcutáneamente hBMSCs mutantes en ratones inmunocomprometidos para generar osículos con características histológicas de SCEF. Se crearon 3 modelos de ratones transgénicos mediante la recombinación Cre-Lox que se basaban en la activación del transgén LoxP-STOP-LoxP-KrasG12D mediante los siguientes promotores de recombinasa Cre: la porción 3.6kb del colágeno 1 alpha-1 inducible por tamoxifeno, Prx1 inducible por tamoxifeno, y Prx1 no inducible. Resultados: La mayor parte de sujetos con SCEF presentaban fracturas, deformidades esqueléticas, y escoliosis. La hipofosfatemia no estaba presente en el nacimiento. Se detectó osteomalacia severa en la histología ósea sin otras alteraciones. Se identificaron lesiones diplásicas en todos los compartimentos esqueléticos. La suplementación con fosfato y calcitriol fue efectiva en la mayoría de sujetos para el raquitisimo/hipofosfatemia. La extirpación de los nevus no se asoció de manera clara con un incremento de la fosfatemia. Se observó una mejoría de las alteraciones minerales asociadas a la edad. Se identificaron en muchos sujetos varias manifestaciones extra-óseas/extra-cutáneas, incluyendo neoplasias, pero no se identificaron cánceres óseos. No se detectaron incrementos de FGF23 en ninguno de los experimentos in vitro. No se pudieron caracterizar adecuadamente los osículos dada una formación ósea insuficiente. No se identificaron anormalidades en los modelos de ratón inducibles. No se identificó ningún ratón doble transgénico para la Prx1 cre-recombinasa y LoxP-STOP-LoxP-KrasG12D. Conclusiones: el fenotipado de la cohorte y la revisión bibliográfica proporcionó información clínica relevante. No obstante, el tejido responsable de la sobreproducción de FGF23 en el SCEF sigue sin conocerse pero los resultados sugieren que los nevus no son la fuente. Otras conclusiones incluyen la falta de idoneidad de las hBMSCs para estudiar la producción FGF23, la escasa eficiencia/reproducibilidad de la formación ósea mediante la implantación de hBMSCs en ratones inmunocomprometidos, y la letalidad de la expresión generalizada de mutaciones Kras en células esqueléticas embrionarias murinas.
Introduction: Cutaneous Skeletal Hypophosphatemia Syndrome (CSHS) is an ultra-rare mosaic disorder caused by postzygotic RAS mutations that is defined by the association of epidermal and/or melanocytic nevi, fibroblast growth factor-23 (FGF23) excess, and a skeletal dysplasia. Because of its rarity, CSHS has been inadequately characterized. Further, unknown are the roles of RAS mutations in dysplastic bone formation and the source and physiopathology of FGF23 overproduction in CSHS. Thesis hypotheses: 1) A thorough phenotypic assessment of patients with CSHS, and the analysis of all potential reported CSHS cases will provide important clinical information, 2) Dysplastic bone is the source of FGF23 overproduction in CSHS and, 3) The generation of appropriate experimental models will inform of CSHS’s physiopathology. Objectives: 1) To clinically characterize CSHS through the study of a cohort of subjects and a review of all reported cases, 2) To identify the physiopathology of FGF23 overproduction in CSHS, 3) To investigate the effects of hyperactive Ras on dysplastic bone formation. Subjects and methods: Five CSHS subjects were phenotyped and underwent prospective data collection. A review of the literature identified 51 subjects in whom the findings were compatible with CSHS. Data on nevi, bone histology, mineral and skeletal alterations, abnormalities in other tissues, and response to treatments of hypophosphatemia were analyzed. To assess the effects of RAS mutations on FGF23 production in “bone-like” cells, FGF23 production/gene expression was measured in 1) human bone marrow stromal cells (hBMSCs) and 2) IDG-SW3 cells transduced with mutated adenoviral RAS constructs, in 3) mutant hBMSCs from a CSHS patient, and in 4) IDG-SW3 cells treated with media in which mutant CSHS hBMSCs had been cultured. To further characterize CSHS bone histology, an ossicle formation assay was performed by transplanting mutant CSHS hBMSCs subcutaneously into immunocompromised mice. Finally, 3 tissue-specific mutant Ras knock-in mouse models were designed through Cre-Lox recombination technology to investigate the effects of activating Ras mutations in bone. The models consisted in the activation of a LoxP-STOP-LoxP-KrasG12D (LSL-KrasG12D) transgene through 3 different Cre-recombinases driven by the following promoters: 1) tamoxifen-inducible 3.6 kb collagen type I α-1, 2) tamoxifen-inducible paired-related homeobox gene-1(Prx1), and 3) non-inducible Prx1. Results: From a clinical standpoint, fractures, limb deformities, and scoliosis affected most CSHS subjects. Hypophosphatemia was not present at birth. Bone histology only revealed severe osteomalacia without other obvious abnormalities. Skeletal dysplasia was reported in all anatomical compartments. Phosphate and calcitriol supplementation was effective in treating rickets in most patients. Convincing data that nevi removal improved blood phosphate levels was lacking. An age-dependent improvement in mineral abnormalities was observed. A spectrum of extra-osseous/extra-cutaneous manifestations that included both benign and malignant neoplasms was present in many subjects, although osteosarcoma remains unreported. From an experimental standpoint, increased FGF23 production/expression was not detected with any of the experimental conditions in the in vitro experiments. Poor bone formation with the ossicle assay hampered histological characterization of the ossicles. No abnormal findings were detected in the models featuring the tamoxifen-inducible promoters. No double transgenic offspring harboring both LSL-KrasG12D and the non-inducible Prx1-Cre recombinase were identified. Conclusions: The study of patients and reports of subjects with CSHS provided relevant and novel clinical information. Also, the analysis of affected patient tissues allowed further appreciation of the effects of hyperactive RAS in different organs. Still, the source of excess FGF23 in CSHS remains unknown, although data indicate it is probably not produced by the nevi. Other conclusions include the lack of suitability of BMSCs to study FGF23 production/expression, the poor reproducibility of the subcutaneous ossicle formation assay, and the lethality of widespread Kras mutations in early embyrionic skeletal cells in mice.

Il Registro Nazionale Italiano per la FSHD (INRF), raccoglie i dati dei pazienti affetti da Distrofia Muscolare facio-scapolo-omerale (FSHD) (MIM\#158900) una rara miopatia ereditaria. Una piattaforma per la gestione dei dati, la MOMIS FSHD Web Platform, ha permesso di studiare ed integrare le informazioni derivanti da diverse fonti. Fino al 2020, l'INRF ha raccolto dati molecolari su 7485 soggetti e identificato 3396 individui portatori di un D4Z4 Reduced Allele (DRA), il marcatore genetico associato alla FSHD. I dati clinici sono raccolti attraverso protocolli di valutazione standardizzati e validati. Dal 2009 sono state raccolte sistematicamente 3574 schede cliniche che raccolgono informazioni su dati demografici, storia clinica e valutazione neurologica e consentono la descrizione dei fenotipi e loro classificazione in categorie: fenotipo classico, incompleto, soggetti asintomatici/sani e fenotipo atipico/complesso. L'integrazione di tali informazioni ha dimostrato sia l'ampia variabilità fenotipica che la penetranza incompleta della FSHD sostenendo l'idea che l'espressione e l'evoluzione della malattia siano determinate da diversi fattori. In effetti, il marcatore della FSHD è un polimorfismo comune, rilevato nel 3\% della popolazione sana. Inoltre, il 23.4\% dei portatori di DRA della nostra coorte sono asintomatici. La presente tesi si propone di sottolineare l'importanza della raccolta sistematica di dati clinici standardizzati. Questo approccio ha migliorato la comprensione dell'espressione di questa malattia rara e ha reso esplicita la sua complessità. Il capitolo 1 fornisce un'ampia panoramica della FSHD, sui registri di malattia e sull’INRF. Il capitolo 2 descrive I metodi per la diagnosi molecolare della malattia e materiale e metodi degli studi svolti nel lavoro di tesi. Gli studi, i cui risultati sono descritti nel capitolo 4 sono i seguenti: 1) INRF: analisi dati clinici e molecolari della coorte del registro, descrizione della sua struttura, della sua organizzazione e dell’inserimento nel contesto Europeo; 2) Studio di correlazione genotipo/fenotipo dei portatori di DRA con 9-10 RU: il 46.0\% dei casi indice non mostra il fenotipo classico; il 10.0\% dei parenti mostra un fenotipo classico; il 70.9\% dei parenti portatori non mostra alcuna compromissione motoria; Caratterizzazione di 125 soggetti con fenotipo incompleto (assenza di debolezza facciale). Questo fenotipo risulta significativamente più lieve rispetto al fenotipo classico. Su 33 famiglie con un probando con fenotipo incompleto, in 18 (54.5\%) il probando era l'unico che esprimeva un profilo miopatico. Il 36\% di questi 125 soggetti, non era portatore di DRA, suggerendo che altri elementi possano essere alla base di questo fenotipo; 4) Studio di fattori ambientali che possono avere un ruolo su progressione o esordio di malattia. Lo studio retrospettivo su una coorte di portatori di DRA, conferma che la pratica regolare delle attività fisiche nei giovani non è dannosa ma, al contrario, è associata a una severità clinica più lieve in età adulta; 5-6) Analisi coppie in consulenza prenatale e valutazione del rischio genetico associato alla FSHD. Con l'obiettivo di migliorare la consulenza genetica fornendo un’approccio personalizzato sulla storia familiare dei consultandi, è stato sviluppato uno strumento innovativo (classificatore) basato sulla tecnologia del Machine Learning. Esso e’ in grado di prevedere la probabilità che un neonato sviluppi un fenotipo miopatico in base alla storia familiare e ai dati molecolari, ottenendo l’80\% di sensibilità e oltre 70\% di specificità. Questo software ha il potenziale di aumentare la qualità della consulenza e diventare un alleato affidabile nella valutazione del rischio nell’ambito della medicina personalizzata.
The Italian National Registry for FSHD (INRF), collects data from patients suffering from Facioscapulohumeral Muscular Dystrophy (FSHD) (MIM\#158900) a rare hereditary myopathy. A platform for data management, the MOMIS FSHD Web Platform, has made it possible to study and integrate information from different sources. Until 2020, INRF collected molecular data on 7485 subjects and identified 3396 individuals carrying a D4Z4 Reduced Allele (DRA), the genetic marker associated with FSHD. Clinical data are collected through standardized and validated evaluation protocols. Since 2009, 3574 clinical forms have been systematically collected including information on demographic data, clinical history and neurological evaluation and allow the description of phenotypes and their classification into categories: classical phenotype, incomplete, asymptomatic/healthy subjects and atypical/complex phenotype. The integration of such information demonstrated both the wide phenotypic variability and the incomplete penetrance of FSHD supporting the idea that the expression and evolution of the disease are determined by several factors. In fact, the marker of FSHD is a common polymorphism, detected in 3\% of the healthy population. In addition, 23.4% of DRA carriers in our cohort are asymptomatic. The present thesis aims to underline the importance of the systematic collection of standardized clinical data. This approach improved the understanding of the expression of this rare disease and made its complexity explicit. Chapter 1 provides a broad overview of the FSHD, disease registries and the INRF. Chapter 2 describes the methods for molecular diagnosis of the disease and material and methods of the studies carried out in this thesis work. The results of the studies are described in Chapter 4 as follows: 1) INRF: analysis of clinical and molecular data of the registry cohort, description of its structure, its organization and its contribution in the European context; 2) genotype/phenotype correlation study of DRA carriers with 9-10 RU: 46.0\% of index cases do not show the classic FSHD phenotype; 10.0\% of relatives show a classical phenotype; 70.9\% of carrier relatives show no motor impairment. 3)Characterization of 125 subjects with incomplete phenotype (absence of facial weakness). This phenotype is significantly milder than the classical phenotype. Out of 33 families with a proband with incomplete phenotype, in 18 (54.5\%) the proband was the only one that expressed a myopathic profile. 36% of these 125 subjects were not carriers of DRA, suggesting that other elements may underlie this phenotype. 4) Study of environmental factors that may play a role in disease progression or onset. The retrospective study on a cohort of DRA carriers confirms that the regular practice of physical activities in young people is not harmful but, on the contrary, is associated with milder clinical severity in adulthood; 5-6) Evaluation of couples in prenatal counselling and assessment of the genetic risk associated with FSHD. With the aim of improving genetic counseling by providing a personalized approach to the family history of the consultants, an innovative tool (classifier) based on Machine Learning technology has been developed. The classifier is able to predict the probability that a newborn will develop a myopathic phenotype based on family history and molecular data, achieving 80\% sensitivity and over 70\% specificity. This software has the potential to increase the quality of counselling and become a reliable ally in risk assessment in personalised medicine.