Aplicação clínica do sequenciamento e análise bioinformática de exomas

Abstract

Approximately 7,000 Mendelian diseases are described, but less than half of these disorders were associated with mutations in specific genes. In this work we performed bioinformatics analysis of exomes from patients with rare Mendelian diseases, treated at the Hospital das Clínicas da UFMG or Laboratório Gene Núcleo de Genética Médica, aiming to find variants that may be causally associated with the diseases. We conducted a study grounded in exome analysis, which has focused in two distinct clinical applications: (1) to improve the genotype-phenotype correlation in chromosomal disorders, by studying a patient with chromosome 1p13.2 microdeletion; (2) to study Mendelian diseases with undefined cause; we analyzed two families with infantile myofibromatosis and two patients with brain calcifications and seizures. In the first case we evaluated the phenotype of patients with 1p13.2 microdeletions and we found that they had many features of Noonan Syndrome. The overlap deletion region of all the patients contained the NRAS gene, gain-of-function mutations of which have been described as causing Noonan Syndrome, Type 6. We performed thus exome sequencing analysis in our patient in order to examine the possibility of a recessive pathogenic variant unmasked by his hemizygous deletion and also to verify if he had pathogenic variants in any of the Noonan Syndrome genes. We concluded that their phenotype may be causally related to deletion of NRAS. Thus, we have the paradox that NRAS haploinsufficiency appears to be able to cause similar characteristics to gain-of-function mutations. Similar situation have been described for Noonan syndrome with gain-of-function mutations and haploinsufficiency of the SHOC2 gene and also for the related Cardio-facio-cutaneous syndrome, in which signs of the disease can appear due to both MAP2K2 deletions and gain-of-function mutations. Consequently we concluded that dysregulation of the Ras/MAPK pathway caused either by gain-of-function mutations or haploinsufficiency may result in similar phenotypes. We also performed exome analysis to study Mendelian diseases with poorly defined cause, for this we analyzed two families with infantile myofibromatosis (MI), which is a rare disorder, with molecular pathogenesis poorly understood, characterized by the development of benign tumors in the skin, muscle, bones and viscera. In this study, we analyzed firstly the exomes of two brothers diagnosed with visceral multicentric IM, and of its healthy and consanguineous parents. We identified a homozygous variant in the gene NDRG4 in the two brothers. Consistent with the IM phenotype, NDRG4 is a tumor-related gene. Thus, we suggested that the variant in NDRG4 may be the causal variant in the studied family with autosomal recessive IM. To confirm our findings, we surveyed the literature for IM families with inheritance possibly autosomal recessive. Then we studied a French family with two affected siblings of healthy and not consanguineous parents. However, we identified in the two brothers a heterozygous variant in the gene PDGFRB. Variants in this gene have been associated with autosomal dominant form of IM. To our surprise, the mother of the patients also had the same variant in PDGFRB. We also found that both brothers inherited paternally a heterozygous variant in PTPRG, an enzyme known to dephosphorylate PDGFRB, reducing its activity. Then we suggested that variants in PTPRG may explain the full phenotypic penetrance in the affected siblings. Therefore, we propose that the variant in NDRG4 should be investigated in other families with autosomal recessive IM, and variants in the gene PDGFRB are associated only with the autosomal dominant form of the disease. In addition, we studied two patients with microcephaly, seizures and calcifications in pontine tegmentum of the brain, which is a rare disorder with poorly understood etiology. We suspected of probably pathogenic variants in four candidate genes: CCBE1, ARL5B, RSU1 and CRLS1. However, by Sanger sequencing technique, we found that such variants were exome sequencing errors. We performed then new analysis of the exomes, looking for genes that are known to cause epilepsy. Doing this analysis we found candidate variants in SCN1A and CHD2. Heterozygous mutations in these genes have been causally related to Dravet syndrome, characterized by seizures with onset in the first year of life. Nevertheless, these variants were also not confirmed by Sanger sequencing. Future analyzes will then be required to identify novel candidate genes. In conclusion, with this study we verified that exome analysis can improve the quality of genotype-phenotype correlations of syndromes caused by chromosomal abnormalities and can assist in the detection of genes that, when mutated, may cause genetic diseases. Our results contributed to a better understanding of the studied genetic diseases, supporting, thus, the diagnosis of future patients and contributing to the monitoring and genetic counseling of patients and their families.