CORNELIA DE LANGE SYNDROME AND RELATED DISORDERS: NEW INSIGHTS INTO GLOBAL TRANSCRIPTIONAL DISTURBANCES DUE TO MUTATIONS IN CHROMATIN-ASSOCIATED FACTORS

Cornelia de Lange syndrome (CdLS) is a rare disorder characterized by an extensive clinical heterogeneity. The main features of the syndrome are characteristic facial dysmorphisms and a variable level of intellectual disability, growth retardation and developmental delay. Though, the number and severity of the clinical signs vary among patients. An extensive genetic heterogeneity partially accounts for the reported clinical variability. Mutations in different cohesin-associated proteins are in fact responsible for the onset of the syndrome. The known CdLS-genes include NIPBL, SMC1A, SMC3, RAD21 and HDAC8. Alterations in the cohesin loader NIPBL are found in more than half of CdLS cases and are associated with a classical phenotype and with a high frequency of limb malformations. In addition, mosaicism has been proven to play an important role in association with NIPBL. Mutations in the structural elements SMC1A, SMC3 and RAD21 and in the regulator HDAC8 account for about 10-15% of CdLS cases. The phenotype of those patients who harbor mutations in these genes is usually milder or atypical. The five genes all together, also taking into account the role of mosaicism, can explain about 70% of CdLS cases. In an internationally assembled cohort of patients we were able to identify 109 mutations in NIPBL, 8 mutations in SMC1A, 15 mutations in SMC3 and 11 mutations in HDAC8, thus increasing the total number of mutations so far described for CdLS. In addition, by the use of next generation sequencing techniques we were able to identify mutations in five genes different from cohesin in six unrelated patients with a clinical diagnosis of CdLS. The five genes include those encoding for different subunits of the chromatin remodeling complex named SWI/SNF and for the transcriptional repressor ANKRD11. Mutations in these genes have been so far associated to Coffin-Siris syndrome and KBG syndrome, respectively. Protein-protein interaction experiments also showed a direct interaction of the SWI/SNF subunit SMARCB1 with the cohesin-related proteins NIPBL and SMC3. These direct link between cohesin and SWI/SNF subunits indicate that mutations affecting the two protein complexes might determine the deregulation of overlapping sets of genes. Our newly identified variants contribute to a better understanding of the correlation between genotype and phenotype in the presence of mutations in the known-CdLS genes. Notwithstanding, different pehotypes have been observed in patients carrying the same DNA alteration, hence suggesting that environmental factors may play an important role in the delineation of the observed clinical features. Additionally, the identification of mutations in chromatin-associated factors responsible for syndromes different from CdLS indicate the existence of a broad pleiotropy that should be taken into account while assessing the clinical and molecular diagnosis. Furthermore, we investigated the molecular mechanisms underlying the syndrome in the presence of missense substitutions or small in-frame deletions in SMC1A, a X-linked gene that localizes in a region of the X-chromosome that partially escapes X-inactivation in humans. Our expression analysis revealed that the transcript is expressed at higher levels in females as compared to males, and that there are no differences in the expression of the SMC1A protein between healthy and affected females. In addition, pyrosequencing analysis showed that CdLS female patients harboring mutations in SMC1A tend to express the wild type allele at higher levels as compared to the mutant allele. All together, these data suggests that the pathogenesis of the syndrome in the presence of mutations affecting SMC1A which do not disrupt the reading frame might be linked to a dominant negative effect exerted by the mutant protein on the wild type.