Functional modeling of genomic variation
Over the last years, gene discovery studies in ASD have identified mutations in genes with regulatory functions as well as in genes involved in synaptic processes. The natural spatiotemporal expression patterns of many of these ASD implicated genes coalesce on subnetworks with distinct developmental trajectories. The next challenge lies in translating this data to biologically relevant molecular pathways to understand the underlying ASD disease mechanisms. In our laboratory we use iPSC-derived neuronal cells which enable us to measure the effects of early neurodevelopmental events on different types of functional neural cells in time. We use CRISPR genome engineering against two different isogenic backgrounds to disrupt synaptic and chromatin-associated genes commonly hit by rare de novo mutations in ASD, e.g. CHD8, SCN2A, AUTS2, FOXP1, POGZ. Characterization of the derivative neuronal models includes transcriptional and epigenetic profiling, as well as synaptic phenotyping, to identify points of molecular and functional convergence associated with these perturbations. We hypothesize that there is a convergent transcriptional signature in neuronal models of perturbations of functionally different ASD genes.
Paternally inherited deletions of 15q11.2-q13 cause Prader-Willi syndrome (PWS), a distinct genomic disorder for which the causal genetic lesion has long-been established yet the mechanism that causes the disorder remains unclear. PWS involves the deletion of a sizeable segment on chromosome 15, which is uniquely vulnerable NAHR due to the presence of multiple SDs, leading to two major classes of deletion: Type I deletion (6 Mb), and Type II deletion (5.3 Mb). The critical feature contributing to PWS has multiple genomic components with varied degrees of association to phenotype; including SNORD116 snoRNA gene cluster, the SNURF-SNRPN transcriptional complex and no other genomic features within the 15p11-q13 region, and the ipw region. To explore the causative genetic driver(s) of PWS, we are generating microdeletions in isogenic iPSC lines and reprogramming them into relevant cell lineages (e.g., hypothalamic neurons). We will define both the molecular and cellular consequences of PWS deletions and determine if they are attributable to an individual driver gene or the contribution of multiple genes through targeted CRISPR inactivation of candidate drivers. These genomic studies will lay the groundwork for testing whether targeted demethylation can ameliorate the aberrant gene expression and cellular phenotypes in this PWS model.
Reciprocal copy number variant (rCNV) of a segment of chromosome 16p11.2 (OMIM #611913) is one of the most significant recurrent genomic disorders. It has been associated with neurodevelopmental disorders (NDDs), including intellectual disability, autism spectrum disorder (ASD), schizophrenia, and obesity. The mechanism of NAHR-mediated CNV formation involves the mispairing of the flanking segmental duplications (SDs), resulting in either the loss or gain of the unique 593 kb genic segment, containing 25 protein-coding genes. However, the pathogenic mechanism, the functional relevance of individual genes within 16p11.2 RGD, and the combined contributions of multiple genes are unknown. 16p11.2 RGD is highly penetrant for NDDs, delineating the critical drivers of 16p11.2 RGD and the highly interconnected hub genes will shed light on disease pathogenesis and therapeutics.
X-linked Dystonia Parkinsonism (XDP)
Dystonia-Parkinsonism is a unique X-linked neurodegenerative disorder that is fully penetrant and lethal. The disorder is indigenous to the Philippines and represents a type of familial scourge, particularly on the Island of Panay. The onset of XDP occurs in males around the age 40 years and symptoms are temporally separated by dystonic and Parkinsonian phenotypes. While the causal haplotype was linked to a segment of the X-chromosome in a series of studies over 25 years ago, the causal variant and pathogenic mechanism was unknown. Talkowski Lab developed novel assembly-based functional genomics methods, integrated with induced-pluripotent stem cell (iPSC) derived neuronal models to explore the genomes and in vitro transcriptomes of XDP patients. These studies discovered the causal variant in XDP to be a novel noncoding sine-VNTR-alu (SVA) retrotransposition that inserted into intron 32 of TAF1, a critical gene involved in the TIID transcriptional complex. Transcriptome assembly revealed that this SVA resulted in aberrant splicing of exon 32-33 and anomalous intron retention (IR), with concomitant reduction of TAF1 expression. Remarkably, we were able to ameliorate this deficit by excising the SVA using CRISPR/Cas9 (Aneichyk et al., 2018, Cell). In a related study in Dr. Ozelius’s lab, we discovered that the hexameric repeat (CCCTCT)n of the SVA was unstable, and that the size of this repeat was strongly correlated with age-of-onset (AOO) of the disorder, reminiscent of the trinucleotide repeat expansion in Huntington’s disease (Bragg et al., 2017, PNAS). Thus, our collaborative team has likely solved the question of the causal variant in XDP and is currently exploring the critical questions of functional mechanism, in vivo alterations in the XDP brain, and the potential for precision therapeutics. We expand this research endeavor to interrogate the in vivo molecular characteristics of XDP, including the hexameric repeat expansion and transcriptional changes associated with XDP using our recently established XDP post-mortem brain tissue collection in the Philippines. In collaboration with the industrial partner, we are also screening an anti-sense oligonucleotide (ASO) library to rescue the aberrant splicing observed in XDP.