Multimodal Neuroimaging of Gene-Brain Relationships in Williams Syndrome

National Institute Of Mental Health

Investigators

Linked publications & trials

Abstract

The Clinical and Translational Neuroscience Branch continues to work toward discovery of novel genetic contributions to brain structure, function, and clinically relevant behavior and cognition through a series of ongoing multimodal neuroimaging studies of individuals with copy number variation in the 7q11.23 Williams Syndrome (WS) genomic region (hemizygous microdeletion or duplication of a contiguous segment of DNA at this locus). These studies have been responsible for seminal advances in elaborating the neural underpinnings of both visuospatial and socio-emotional aspects of the 7q11.23 phenotype. Via multiple neuroimaging techniques, including voxel- and surface-based cortical morphometry, diffusion tensor imaging, functional MRI, we have established that the visuospatial construction deficits in WS are linked to convergent intraparietal sulcus alterations. Specifically, in this brain region, we have shown that individuals with WS harbor disrupted neural integrity, altered activation during spatial judgments, gray matter volume and sulcal depth reductions, and associated neural fiber tract anomalies. Similarly, in pursuit of systems-level correlates of the hypersociability and non-social anxiety observed in WS, we have found decreased amygdala activation evoked by viewing pictures of faces with fear-inducing content and, conversely, increased amygdala response to non-social frightening pictures, abnormalities that were linked to altered prefrontal regulation in structural equation models. We have also identified convergent alterations in anterior insula structure, function, and inter-regional connectivity that predict the characteristic WS personality. Efforts this year have focused on data collection of these same structural and functional measurements of visuospatial and socio-emotional systems integrity, with additional in vivo neuroimaging measurements of myelination, in a growing cohort of children with and without WS critical region copy number variation (i.e., individuals with one in WS, two in typically developing TD, or three in Dup7 copies of affected genes) as part of our longitudinal WS neurodevelopmental initiative. In proof-of-concept work aimed at establishing neurostructural gene-dosage effects, we have found increasing overall brain size (Dup7>TD>WS) but decreasing relative cerebellar size (WS>TD>Dup7) with copy number of affected genes. Interestingly, both of these Dup7 phenotypes (larger brain size and relatively smaller cerebellum) have been described in the autism literature, particularly in boys, although these findings are not without controversy. Following this work, we are undertaking similar gene-dosage analyses of more localized morphometry throughout the brain, as well as local gyrification index and resting-state whole-brain connectivity, the latter using a connectome-wide association study approach as well as independent component and dual-regression analyses. In pursuit of understanding the heterogeneity across individuals with copy number variation in the Williams Syndrome genomic region, we have embarked on studies of the effects of single nucleotide polymorphisms and genetic haplotypes in the remaining (for WS) or duplicated (Dup7) strand of the chromosomal region. We have developed novel methods to achieve specialized genotyping from SNP-chip data and applied these methods in proof-of-concept work testing the hypothesis that common variation in the ELN gene (and not other 7q11.23 genes) would predict clinically meaningful abnormalities of aortic structure. We were able to generate haploid and triploid genotype calls across the affected region and identified a single nucleotide polymorphism associated with aortic stenosis in WS participants and protection from aortic dilation in Dup7 participants. Ongoing work will focus on understanding how sequence variation within the WS region predisposes to variability in neural phenotypes, such as above-mentioned macrostructural characteristics that we have observed to be associated with 7q11.23 copy number variation in a gene-dose dependent manner. Preliminary data from our WS developmental cohort has already demonstrated parietal hypofunction during visuospatial challenge along with altered social network activation during processing of socially salient stimuli, consistent with the hypothesis that both visuospatial and social neurobiological differences in WS are rooted in early life. Recently in this cohort, we have uncovered atypical patterns of intraparietal sulcus functional connectivity in WS, which feature diminished cooperativity with visual networks but, in contrast, enhanced social brain network linkage. This work offers a neural circuit-based view of how these diverse visuospatial and social circuits integrate in the context of 7q11 copy variation and in the context of the behavioral characteristics of this population. We have also formalized our longitudinal analysis methods and have made available to the neuroimaging community a software tool to address some of the challenges of analyzing longitudinal neuroimaging data (Chen et al., 2021). In addition, we have established processing pipelines for analyzing the diverse array of neuroimaging data types being collected. Overall, this project seeks not only to expand knowledge of the WS-related brain systems in childhood, but also to identify developmental trajectory (throughout childhood) and gene dose-response characteristics of neural abnormalities underlying visuospatial and socio-emotional alterations in this syndrome using a longitudinal, repeated measures design. Preliminary proof-of-concept analyses in this vein have already been successful. Though data accrual will require years of careful and concerted effort to complete, the potential for these studies to shed unprecedented light on genetic contributions to brain development is enormous. This work includes the following studies: NCT01132885, NCT00004571, NCT00001258