Mammalian Proteome Profiling Reveals Readers and Antireaders of Strand-Symmetric and -Asymmetric 5-Hydroxymethylcytosine-Modifications in DNA

Abstract

The cytosine (C) modifications 5-methylcytosine (mC) and 5-hydroxymethylcytosine (hmC) are central regulatory elements of mammalian genomes. Both marks occur in dinucleotide contexts of double stranded DNA in either strand-symmetric or -asymmetric fashion, but it is still poorly understood how this symmetry information is selectively read out by the nuclear proteome as basis of potential symmetry-dependent regulation mechanisms. We report comparative enrichment/proteomics studies with promoter probes being strand-symmetrically or asymmetrically modified with C, mC and hmC, enabling a direct assessment of their reader profiles in the same sequence, tissue, and experimental contexts. We identify in human and mouse nuclear lysates a high number of tissue-specific readers for hmC-modified sequences that fall into distinct, probe-specific sub-groups, including members of important transcription factor classes and chromatin regulators. Among them, we discover the master regulators MYC and MAX that play central roles in cell (de)differentiation and cancer progression to read hmC in a sequence-dependent manner. We also find RFX5, a transcription factor critically involved in primary MHC class II deficiency to discriminate between specific hmC symmetries in CpG dyads. Our findings provide further support for the hypothesis that hmC symmetry information can provide distinct regulatory outputs, and provide a resource for studying the molecular mechanisms triggered by symmetric and asymmetric hmC modifications in chromatin regulation during development and disease.