Identifying novel disease genes in genetically undiagnosed individuals with Rett syndrome and related neurodevelopmental disorders

© 2020 Simranpreet KaurRett Syndrome (RTT) is a severe neurodevelopmental disorder (NDD) resulting in severe cognitive and physical impairments. Despite being predominantly caused by pathogenic variants in the methyl-CpG-binding (MECP2) gene, between 3 – 15% of classic and atypical RTT individuals do not have a genetic diagnosis. Classic RTT individuals exhibit an apparently normal development until 6 to 18 months of age after which developmental regression occurs. Atypical RTT individuals have many features of classic RTT but do not meet all the specific diagnostic criteria. Recently, the classification of RTT has been expanded to include individuals with clinical features overlapping RTT and other NDDs, often referred to as RTT-like individuals. The clinical and genetic diagnosis for RTT-like individuals is further complicated due to the complexity of NDDs and there is an unmet need to provide a precise genetic diagnosis for these individuals. Next generation sequencing (NGS) studies are continuously identifying hidden genetic links between relevant molecular pathways and RTT. Through whole genome sequencing (WGS), our lab had identified a de novo heterozygous missense variant [c.744C>A; (p.Asp248Glu)] in the motor domain of kinesin-3 family member 1A (KIF1A) in one classic RTT female. Single nucleotide heterozygous variants in KIF1A have been implicated in a number of severe neurological disorders, collectively known as KIF1A-associated neurological disorders (KANDs). KIF1A encodes a neuron-specific kinesin molecular motor protein essential for ATP-dependent anterograde axonal transport of synaptic cargos along microtubules. In order to determine additional RTT/RTT-like individuals with KIF1A variations, we collected further clinical and genetic information from our collaborators of 30 individuals with 18 different missense variants affecting the critical motor domain of KIF1A. After careful clinical assessment, we identified three additional individuals with different novel missense variants exhibiting overlapping RTT-like and KAND clinical features. In silico tools predicted all variants to affect proper protein folding and were predicted to be likely disease causing. In addition, in vitro functional analysis using the highly specific neurite tip accumulation and microtubule gliding assays, demonstrated all variants to have reduced microtubule based movement, suggesting these variants are indeed significantly pathogenic. Comparison of the clinical features of the remaining 27 KAND individuals with 16 variants in the KIF1A motor domain suggested that specific clinical features and phenotypic severity was largely dependent upon the location of the variant. Using an NGS approach, we identified pathogenic MECP2 variants, previously missed by mainstream genetic testing, in seven RTT individuals including a case of a mosaic male. In addition, we found variants in two genes that are known to be associated with NDDs and RTT-like syndromes; Structural Maintenance of Chromosomes 1A [SMC1A; c.3576delA; p.(Val1193Serfs*2)] and SH3 and multiple ankyrin repeat domains 3 (SHANK3; c.2265+1G>A). This work continues to expand the genetic basis of RTT. Through whole exome sequencing (WES), we have also identified an atypical RTT female with a heterozygous nonsense variant [c.3385C>T; p.(Arg1129*)] in Lysine Acetyltransferase 6A (KAT6A) that encodes a chromatin remodelling protein. Heterozygous protein truncating variants in this gene have been associated with KAT6A-related intellectual disability. Through various collaborations, we identified a further four individuals with KAT6A variants [c.3820G>T; p.(Glu1274*), c.3399_3400dup; p.(Lys1134Argfs*14), c.3377delC; p.(Ser1126Phefs*8) and c.3631_3632delGT; p.(Val1211*)] who had clinical symptoms overlapping with RTT/RTT-like individuals. Through systematic re-analysis of a reported cohort of 76 individuals with KAT6A-related intellectual disability we recognized two additional individuals exhibiting RTT-like clinical features with KAT6A variants [c.4254_4257del; p.(Glu1419Trpfs*12) and c.3661G>T; p.(Glu1221*)] . All the identified variants in the seven RTT-like individuals were clustered in the last exon of KAT6A and in silico analysis predicted the variants to cause a dominant-negative effect. These seven individuals exhibited clinical features overlapping between RTT and KAT6A-related intellectual disability that was previously unrecognized. Using singleton WGS, a novel heterozygous nonsense variant in another chromatin regulator gene, Chromodomain helicase DNA-binding protein 8 [CHD8; c.5017C>T; p.(Arg1673*)] was identified in an atypical RTT individual. Heterozygous protein truncating variants in CHD8 have been implicated in NDDs including Autism Spectrum Disorders (ASDs). Functional analysis confirmed reduction in CHD8 protein levels in her fibroblasts, confirming the pathogenicity of the identified variant. In another RTT-like female, a de novo heterozygous missense variant [c.271G>A; p.(Asp91Asn)] in the Eukaryotic Translation Elongation Factor 1 Alpha 2 (EEF1A2), involved in protein translation, was identified through singleton WES. This variant has been previously reported in a female with NDD. Interestingly, a single case with the same EEF1A2 variant [c.274G>A, p.(Ala92Thr)] has also been reported in a RTT-like female. Thus, our findings further established the casual association between EEF1A2 and a RTT-like phenotype. In a RTT-like individual, a de novo large deletion at chromosome 9q34.11 (hg19:131,455,942-131,743,585) resulted in the loss of 13 genes, including the 3’ end of the coding sequence of SET Nuclear Proto-Oncogene (SET) and the 5’ end of the coding sequence of Nucleoporin-188 kDa (NUP188). The truncation of SET resulted in the loss of a highly conserved critical nucleosome assembly protein (NAP) domain crucial for assembly of nucleosomes and chromatin fluidity. Subsequent WES in the same individual identified a missense variant in NUP188 [c.3922C>T; p.(Arg1308Cys)] on the other allele. Preliminary functional studies in individual’s fibroblasts showed reduced NUP188 protein levels and enlarged nuclei, suggestive of perturbed NUP188 function. In this individual two genes, SET and NUP188, may be contributing to the affected individual’s phenotype. Interestingly, a homozygous variant in a novel candidate disease gene, Potassium Channel Tetramerization Domain Containing 16 [KCTD16; c.937T>A; p.(Ser313Thr)], was also revealed in a classic RTT female. KCTD16 encodes an auxiliary subunit that associates with GABA-B receptor and regulates receptor response in an agonist concentration dependent manner. Although variants in the KCTD family of proteins have previously been reported in individuals with NDDs, defects specifically in KCTD16 are yet to be linked with any human disease. Our preliminary electrophysiological studies in Xenopus oocytes investigating perturbed GABA-B receptor kinetics in response to the KCTD16 variant [p.(Ser313Thr)] did not reveal any significant differences when compared to wild type, as well as a variant commonly found in the healthy population [p.(Asp160Asn)]. Despite this, we plan to continue these investigations in a mammalian Chinese Hamster Ovary (CHO) cell-based model to further evaluate the variant’s pathogenicity. Overall, this study has provided functional evidence of variations affecting the motor domain of KIF1A in the pathophysiology of RTT/RTT-like disorders. In addition to identifying pathogenic variations in four known RTT-related genes (MECP2, SHANK3, SMC1A and EEF1A2), in this project we have further expanded the genetic landscape of RTT/RTT-like disorders to include variations in five additional genes (KAT6A, CHD8, SET, NUP188 and KCTD16). We recommend that the testing of these genes should be considered routinely while analysing NGS data in mutation negative RTT/ RTT-like individuals. The identification of additional members of key molecular pathways perturbed in RTT has further expanded our understanding of the underlying biology behind RTT, and this may pave the way for future targeted therapeutic options for RTT