Structure-Based Approaches to Classify the Functional Impact of ZBTB18 Missense Variants in Health and Disease

Abstract

The Cys2His2 type zinc finger is a motif found in many eukaryotic transcription factor proteins that facilitates binding to genomic DNA so as to influence cellular gene expression. One such transcription factor is ZBTB18, characterized as a repressor that orchestrates the development of mammalian tissues including skeletal muscle and brain during embryogenesis. In humans, it has been recognized that disease-associated ZBTB18 missense variants mapping to the coding sequence of the zinc finger domain influence sequence-specific DNA binding, disrupt transcriptional regulation, and impair neural circuit formation in the brain. Furthermore, general population ZBTB18 missense variants that influence DNA binding and transcriptional regulation have also been documented within this domain; however, the molecular traits that explain why some variants cause disease while others do not are poorly understood. Here, we have applied five structure-based approaches to evaluate their ability to discriminate between disease-associated and general population ZBTB18 missense variants. We found that thermodynamic integration and Residue Scanning in the Schrodinger Biologics Suite were the best approaches for distinguishing disease-associated variants from general population variants. Our results demonstrate the effectiveness of structure-based approaches for the functional characterization of missense alleles to DNA binding, zinc finger transcription factor protein-coding genes that underlie human health and disease.

Understanding the impact of ZBTB18 missense variation on transcription factor function in neurodevelopment and disease

“…Consistent with studies of the Rnd2 3′ enhancer, we found that wildtype ZBTB18 bound native sequences Id2 ‐bs1 and Id2 ‐bs2 with high affinity, while the N461S variant bound more strongly than wildtype ZBTB18 yet it does not form stable complexes with Id2 ‐bs2 (Blake et al, 2021). On the other hand, the R495G variant does not form stable complexes with Id2 ‐bs1 and binds Id2 ‐bs2 weakly (Blake et al, 2021). Therefore, disease‐associated ZBTB18 missense variants N461S and R495G disrupt sequence‐specific DNA binding that is essential for regulating the expression of downstream target genes Rnd2 and Id2 .…”

“…luciferase reporter assays, chromatin immunoprecipitation (ChIP), electrophoretic mobility shift assays (EMSAs), in utero electroporation (IUE) studies; as well as as well as more recently reported screening methods (e.g. multi‐plex reporter assays (MPRAs) (Mulvey, Lagunas, & Dougherty, 2021), mammalian targeted damID (MaTaDa) (Cheetham et al, 2018), Cut&Run (Meers, Bryson, Henikoff, & Henikoff, 2019)) and, where feasible, binding free energy calculations (Blake et al, 2021; Hemming et al, 2019; Hemming et al, 2020)) in order to study the DNA‐binding, protein–protein interaction and transcriptional regulatory signalling properties of TFs as well as their query variants (Table 3). It is noteworthy that MPRAs, MaTaDa and Cut&Run leverage advances of recent years in massively parallel sequencing technologies and robotic screening platforms to make them exciting new functional screening approaches to map causal missense TF variants in health and disease.…”

“…Feasibility studies using calculation of relative binding free energy against homology models of ZBTB18-DNA complexes have validated such approaches for classification of pathogenicity of ZBTB18 missense variants (Blake et al, 2021). Combining in silico approaches with assays to describe the landscape of DNA motif binding in vivo, in living neurons by TFs and their missense variants will be highly informative for quantifying genotype-phenotype relationships in a tissuespecific context.…”

“…To further support this finding, we conducted molecular modelling studies between ZBTB18 and several bona fide regulatory enhancer motifs (named Id2 ‐bs1 and Id2 ‐bs2) within the Id2 gene (Blake, Hemming, Heng, & Agostino, 2021), an essential Zbtb18 downstream target gene for neurodevelopment and skeletal muscle formation (Cargnin et al, 2018; Hirai et al, 2012; Yokoyama et al, 2009). Consistent with studies of the Rnd2 3′ enhancer, we found that wildtype ZBTB18 bound native sequences Id2 ‐bs1 and Id2 ‐bs2 with high affinity, while the N461S variant bound more strongly than wildtype ZBTB18 yet it does not form stable complexes with Id2 ‐bs2 (Blake et al, 2021). On the other hand, the R495G variant does not form stable complexes with Id2 ‐bs1 and binds Id2 ‐bs2 weakly (Blake et al, 2021).…”

Sequencing Directly from Clinical Specimens Reveals Genetic Variations in HCMV-Encoded Chemokine Receptor US28 That May Influence Antibody Levels and Interactions with Human Chemokines

“…The residue scanning/affinity maturation tool of Schrodinger Biologics Suite ( 58 ) was used to generate variants in US28 Toledo and assess their impact on binding to chemokines and the G protein assembly (i.e., all of the αβγ subunits), adapting our previous work ( 59 ). Calculations assessed changes in chemokine binding to the US28-G protein assembly (i.e., chemokine treated as ligand, US28-G protein assembly treated as receptor) and changes in G protein assembly binding to the chemokine-US28 assembly (i.e., αβγ assembly treated as ligand, chemokine-US28 assembly treated as receptor).…”