The Polycomb repressive complex 2 (PRC2) confers transcriptional repression through histone H3 lysine 27 trimethylation (H3K27me3). Here, we examined how PRC2 is modulated by histone modifications associated with transcriptionally active chromatin. We provide the molecular basis of histone H3 N terminus recognition by the PRC2 Nurf55-Su(z)12 submodule. Binding of H3 is lost if lysine 4 in H3 is trimethylated. We find that H3K4me3 inhibits PRC2 activity in an allosteric fashion assisted by the Su(z)12 C terminus. In addition to H3K4me3, PRC2 is inhibited by H3K36me2/3 (i.e., both H3K36me2 and H3K36me3). Direct PRC2 inhibition by H3K4me3 and H3K36me2/3 active marks is conserved in humans, mouse, and fly, rendering transcriptionally active chromatin refractory to PRC2 H3K27 trimethylation. While inhibition is present in plant PRC2, it can be modulated through exchange of the Su(z)12 subunit. Inhibition by active chromatin marks, coupled to stimulation by transcriptionally repressive H3K27me3, enables PRC2 to autonomously template repressive H3K27me3 without overwriting active chromatin domains.

Cys2-His2 zinc finger (C2H2-ZF) proteins represent the largest class of putative human transcription factors. However, for most C2H2-ZF proteins it is unknown whether they even bind DNA or, if they do, to which sequences. Here, by combining data from a modified bacterial one-hybrid system with protein-binding microarray and chromatin immunoprecipitation analyses, we show that natural C2H2-ZFs encoded in the human genome bind DNA both in vitro and in vivo, and we infer the DNA recognition code using DNA-binding data for thousands of natural C2H2-ZF domains. In vivo binding data are generally consistent with our recognition code and indicate that C2H2-ZF proteins recognize more motifs than all other human transcription factors combined. We provide direct evidence that most KRAB-containing C2H2-ZF proteins bind specific endogenous retroelements (EREs), ranging from currently active to ancient families. The majority of C2H2-ZF proteins, including KRAB proteins, also show widespread binding to regulatory regions, indicating that the human genome contains an extensive and largely unstudied adaptive C2H2-ZF regulatory network that targets a diverse range of genes and pathways.

C2H2 zinc finger proteins represent the largest and most enigmatic class of human transcription factors. Their C2H2-ZF arrays are highly variable, indicating that most will have unique DNA binding motifs. However, most of the binding motifs have not been directly determined. In addition, little is known about whether or how these proteins regulate transcription. Most of the ∼700 human C2H2-ZF proteins also contain at least one KRAB, SCAN, BTB, or SET domain, suggesting that they may have common interacting partners and/or effector functions. Here, we report a multifaceted functional analysis of 131 human C2H2-ZF proteins, encompassing DNA binding sites, interacting proteins, and transcriptional response to genetic perturbation. We confirm the expected diversity in DNA binding motifs and genomic binding sites, and provide motif models for 78 previously uncharacterized C2H2-ZF proteins, most of which are unique. Surprisingly, the diversity in protein–protein interactions is nearly as high as diversity in DNA binding motifs: Most C2H2-ZF proteins interact with a unique spectrum of co-activators and co-repressors. Thus, multiparameter diversification likely underlies the evolutionary success of this large class of human proteins.

The carboxy-terminal domain (CTD) of the RNA polymerase II (RNAP II) subunit POLR2A is a platform for modifications specifying the recruitment of factors that regulate transcription, mRNA processing, and chromatin remodelling. Here we show that a CTD arginine residue (R1810 in human) that is conserved across vertebrates is symmetrically dimethylated (me2s). This R1810me2s modification requires protein arginine methyltransferase 5 (PRMT5) and recruits the Tudor domain of the survival of motor neuron (SMN, also known as GEMIN1) protein, which is mutated in spinal muscular atrophy. SMN interacts with senataxin, which is sometimes mutated in ataxia oculomotor apraxia type 2 and amyotrophic lateral sclerosis. Because POLR2A R1810me2s and SMN, like senataxin, are required for resolving RNA-DNA hybrids created by RNA polymerase II that form R-loops in transcription termination regions, we propose that R1810me2s, SMN, and senataxin are components of an R-loop resolution pathway. Defects in this pathway can influence transcription termination and may contribute to neurodegenerative disorders.

ENCODE 3 (2012-2017) expanded production and added new types of assays 8 (Fig. 1, Extended Data Fig. 1), which revealed landscapes of RNA binding and the 3D organization of chromatin via methods such as chromatin interaction analysis by paired-end tagging (ChIA-PET) and Hi-C chromosome conformation capture. Phases 2 and 3 delivered 9,239 experiments (7,495 in human and 1,744 in mouse) in more than 500 cell types and tissues, including mapping of transcribed regions and transcript isoforms, regions of transcripts recognized by RNA-binding proteins, transcription factor binding regions, and regions that harbour specific histone modifications, open chromatin, and 3D chromatin interactions. The results of all of these experiments are available at the ENCODE portal (http://www.encodeproject.org). These efforts, combined with those of related projects and many other laboratories, have produced a greatly enhanced view of the human genome (Fig. 2), identifying 20,225 protein-coding and 37,595 noncoding genes

Development of an accurate protein–DNA recognition code that can predict DNA specificity from protein sequence is a central problem in biology. C2H2 zinc fingers constitute by far the largest family of DNA binding domains and their binding specificity has been studied intensively. However, despite decades of research, accurate prediction of DNA specificity remains elusive. A major obstacle is thought to be the inability of current methods to account for the influence of neighbouring domains. Here we show that this problem can be addressed using a structural approach: we build structural models for all C2H2-ZF–DNA complexes with known binding motifs and find six distinct binding modes. Each mode changes the orientation of specificity residues with respect to the DNA, thereby modulating base preference. Most importantly, the structural analysis shows that residues at the domain interface strongly and predictably influence the binding mode, and hence specificity. Accounting for predicted binding mode significantly improves prediction accuracy of predicted motifs. This new insight into the fundamental behaviour of C2H2-ZFs has implications for both improving the prediction of natural zinc finger-binding sites, and for prioritizing further experiments to complete the code. It also provides a new design feature for zinc finger engineering.

Highlights d Tight and selective UbVs target USP15 catalytic and adaptor domains d UbV inhibitors lock the USP15 active site in an inactive conformation d A strand-swapped UbV dimer binds two DUSP domains simultaneously d Linear UbV dimers are potent and specific USP15 inhibitors in cells SUMMARYThe multi-domain deubiquitinase USP15 regulates diverse eukaryotic processes and has been implicated in numerous diseases. We developed ubiquitin variants (UbVs) that targeted either the catalytic domain or each of three adaptor domains in USP15, including the N-terminal DUSP domain. We also designed a linear dimer (diUbV), which targeted the DUSP and catalytic domains, and exhibited enhanced specificity and more potent inhibition of catalytic activity than either UbV alone. In cells, the UbVs inhibited the deubiquitination of two USP15 substrates, SMURF2 and TRIM25, and the diUbV inhibited the effects of USP15 on the transforming growth factor b pathway. Structural analyses revealed that three distinct UbVs bound to the catalytic domain and locked the active site in a closed, inactive conformation, and one UbV formed an unusual strand-swapped dimer and bound two DUSP domains simultaneously. These inhibitors will enable the study of USP15 function in oncology, neurology, immunology, and inflammation.

Small-molecule inhibitors of DNA methyltransferases such as RG108 represent promising candidates for cancer drug development. We report the synthesis and in vitro analysis of a biotinylated RG108 conjugate, 2-(1,3-dioxo-1,3-dihydro-isoindol-2-yl)-3-(5-[3-[5-(2-oxo-hexahydro-thieno[3,4-d]imidazol-4-yl)pentanoylamino]propoxy]-1H-indol-3-yl)propionic acid (bio-RG108), for the evaluation of interactions with DNA methyltransferase enzymes. The structural design of the chemically modified inhibitor was aided by molecular modeling, which suggested the possibility for extensive chemical modifications at the 5-position of the tryptophan moiety in RG108. The inhibitory activity of the corresponding derivative was confirmed in a cell-free biochemical assay, where bio-RG108 showed an undiminished inhibition of DNA methyltransferase activity (IC50 = 40 nM). Bio-RG108 therefore represents a suitable bioconjugate for the elucidation of inhibitory mechanisms and for the affinity purification of RG108-associated proteins.