New genes and pathways implicated in Rett syndrome: considerations and future applications

Abstract

STUDY I

HYPOTHESIS Genetic alterations have already been recognized as major etiological factors for ASD and ID, of which RTT is an example. In addition to the contribution of polymorphic variants that confer low or moderate risk of appearance of these neurodevelopmental defects, de novo mutations affecting genes in a number of cellular pathways have been reported to be a cause of ASD, ID and associated NDDs (Vissers et al., 2010; Ronemus et al., 2014; Gilissen et al., 2014; Deciphering Developmental Disorders Study, 2015). Although the strong correlation between NDDs and genetic factors has been long established, the exact genetic background of ASD and ID remains unclear because of the strong heterogeneity of these disabilities. Genetic research has focused on the use of unbiased genome-wide approaches, including genomic microarrays and, more recently, NGS technology with the use of extensive gene panels, the exome or the whole genome. Consequently, new ASD and ID genes are now being identified in rapid succession. Among these new candidate genes, supportive role for the Jumonji Domain Containing 1C (JMJD1C) histone demethylase in ASD and ID has been fostered by both positional cloning strategies (Castermans et al., 2007) and exome-sequencing studies (Neale et al., 2012; Iossifov et al., 2014). Thus, based on the strong genetic architecture underlying NDDs and the emerging evidences for a role of JMJD1C as potential candidate genes, in the present PhD thesis we proposed to study the involvement of JMJD1C alterations in ASD, ID and RTT and their effects on the protein function.

OBJECTIVES The specific objectives of study I are as follows:

1. To investigate the occurrence of JMJD1C mutations in the mentioned disabilities, a comprehensive mutational analysis was performed in samples

from ASD, ID and RTT, searching for SNVs and indels, as well as larger genetic defects. 2. To address the functional consequences of the identified de novo JMJD1C- Pro163Leu mutation, the intracellular localization of the wildtype and mutated JMJD1C protein, as well as the efficiency in demethylating a non-histone target of JMJD1C, MDC1 (mediator of DNA-damage checkpoint 1), was studied in immunofluorescence, fractionation and immunoprecipitation experiments in HEK293 cells. 3. To direct a molecular explanation for the implication of JMJD1C in Rett Syndrome, the interaction between MeCP2 and JMJD1C, in the wildtype and mutated form, was assessed in an immunoprecipitation assay. 4. To study the cellular effects of the disruption of wildtype JMJD1C on a neuronal system, we analyzed the existence of changes in dendritic branching of primary neuron cultures from neonatal mouse hippocampus.

STUDY II

HYPOTHESIS Mutations in MECP2 cause most of the classical or typical forms of RTT (Chahrour and Zoghbi, 2007). Approximately 8 % of classic RTT and 42 % of variant RTT patients are MECP2 mutation-negative (Monrós et al. 2001; Percy, 2008). Some of the latter group have mutations in other genes, such as that of CDKL5, which is described in individuals with an early seizure onset variant of RTT (Kalscheuer et al., 2003) or FOXG1, which is responsible for the congenital variant of RTT (Ariani et al., 2008). However, there remains a subset of patients with a clinical diagnosis of RTT who are mutation-negative for all the aforementioned genes. In the present PhD thesis, we proposed to identify new candidate genes that could explain the RTT-like phenotype of several clinical cases without mutations in MECP2, CDKL5 and FOXG1, using NGS, with the purpose of expanding the knowledge of the impaired biological pathways in RTT.

OBJECTIVES The specific objectives of study II are as follows:

1. To identify previously undescribed variants potentially implicated in RTT-like phenotype, WES was performed on a cohort of 19 Spanish parent–child trios and one family with two affected daughters presenting features associated with RTT. A bioinformatics process of WES data was realized to filter and select putative pathogenic de novo variants absent or present with very low frequency in the control population, and putatively dangerous for protein function. 2. To realize a differential diagnosis among diverse neurodevelopmental disorders with overlapping phenotype, a gene-association analysis was carried out to define a list of variants previously associated with neurodevelopmental disorders and another one with undescribed new variants not previously associated with. 3. To demonstrate a neurological implication for a loss of function of detected candidate genes not previously associated with neurodevelopmental disorders, the model organism C. elegans was used to confirm a genotype-phenotype correlation by performing locomotion assays in worm mutants that carry deleterious mutations in the orthologous genes to those human genes with potentially pathogenic mutations in the patients.