Geneotype-to-Phenotype Landscape of the Methyl CpG Binding Domain Highlights a Critical and Novel Role in Neuron Enhancer Regulation and ASD

Background: The methyl CpG binding proteins including MBD2, MBD3, and MeCP2 have long been thought of as genome regulators for methylation. Throughout their evolution, particularly into invertebrates, the factors have lost their ability to bind methylated DNA, yet maintain their GATAD2A interactions for the Nucleosome Remodeling Deacetylase (NuRD) complex. Recent ENCODE work in HepG2 cells has suggested an MBD3 based GATAD2A role in enhancer biology not driven by DNA methylation, which has not been explored to date in most cell types.

Objectives: In this work we track the inter and intra species evolution, biochemistry, and expression of MBD containing proteins and GATAD2A based NuRD components to determine if they play a role in enhancer biology of neurons.

Methods: Each of 11 MBD containing genes were taken through sequence (1,229 open reading frames), structure, dynamic (120 simulations), human variant (>120,000 human exome/genomes), expression (FANTOM, HPA, GTEx, Allen Brain Atlas) and function analysis tools. Both iPSC and NPC cell lines had GATAD2A endogenously FLAG tagged using CRISPR/Cas9 followed by differentiation and ChIP-Seq in hepatocytes (translate HepG2 ENCODE transcription factor maps) and neurons.

Results: We mapped three critical amino acids (2 Arginine and 1 Tyrosine) to methylation coordination within MBD as confirmed by evolutionary selection, biophysics and molecular dynamic simulations (mds). The movement of these amino acids in mds revealed MBD2, MBD3 and MECP2 to coordinate methyl and hydroxymethyl DNA. Analysis of invertebrates resolved novel functional MBDs in a wide array of species including lancelet, tick, scorpion, spider, cockroach, octopus, and sea cucumber. Moreover, analysis of expression of species that do not possess methyl or DNA binding MBDs, but still coordinate GATAD2A and the NuRD complex, suggest brain and peripheral nerve function. Translation of GATAD2A data from ENCODE HepG2 using ChIP-Seq from iPSC derived hepatocytes and neurons, elucidated a critical role of GATAD2A regulating enhancers of multiple autism associated genes in non-proliferating cells (AUTS2, PCDH7, TBL1XR1, TENM2, ADGRL2, PTPRD, SPRY2, CSMD3, FSTL5, TENM4, CNTN5, and UNC5D). Further dissection of human expression databases revealed a shared MBD expression profile between brain and liver relative to hundreds of tissue types, where MBD2 expression is low and MBD3 and MECP2 highly expressed, suggesting a shift in methyl binding to MECP2 and GATAD2A localization to enhancers of ASD associated genes. Single cell analysis of the Middle Temporal Gyrus revealed only 41% of cells expressing a functional MBD protein (16% MECP2, 11% MBD2, 7% MBD3, 3% MECP2/MBD2, 2% MECP2/MBD3, 2% MBD2/MBD3, 0.4% MECP2/MBD2/MBD3) and 29% expressing GATAD2A/B, with only 13% of cells overlapping (7% MECP2, 5% MBD2, 4% MBD3).

Conclusions: The regulation of methylation binding in neurons and hepatocytes by MBD proteins seems to suggest a more diverse role than previous thought, with GATAD2A driven NuRD complex serving a primary role in enhancer regulation of ASD genes and MECP2 present in the most cells for methylation binding, suggesting a model for Rett syndrome genotype-to-phenotype MECP2 mutations.