Conformational and Thermodynamic Differences Underlying Wild-Type and Mutant Eleven-Nineteen-Leukemia YEATS Domain Specificity for Epigenetic Marks

Abstract

Histone post-translational modifications (PTMs) are interpreted by multiple reader domains and proteins to regulate gene expression. The eleven-nineteen-leukemia (ENL) YEATS domain is a prototypical PTM reader that recognizes multiple lysine acetylation marks on the histone H3 tails as a way of recruiting chromatin remodellers. Two ENL YEATS mutations have been identified which have been linked with leukemia, Wilms tumor, and other forms of cancer and result in either an insertion or deletion of residues in the loop connecting beta sheets distant from the protein active site. In vitro experiments have shown that these mutations modulate the selectivities of YEATS domains for various lysine acetylation marks, although different experiments have provided contrasting views on the abilities of the insertion and deletion mutants to discern specific PTMs. Here, we have performed multiple molecular dynamics simulations of wild-type and insertion and deletion mutant YEATS domains free from and in complex with two PTM peptides: one that is acetylated at K9 of H3 and the other that is acetylated at residue K27 of H3. Results show that these two peptides have distinct flexibilities and binding energetics when bound to YEATS domains and that these properties are affected by interactions with residues within and outside of the peptide consensus motif. Furthermore, these properties are modulated by the YEATS insertion and deletion mutants, which results in disparate binding effects in these systems. Together, these results suggest that only the partial exposure of histone tails is sufficient in the context of nucleosomes for YEATS-mediated recognition of acetylation marks on histone tails. They also caution against the overinterpretation of results obtained from experiments on reader domain-histone peptide binding in isolation and not in the full-length nucleosome context.

The kth nearest neighbor method for estimation of entropy changes from molecular ensembles

“…The approach has been used also grouping together variables and considering the mutual information between pairs of groups of variables. 53 The approach has been used to assess the effect of correlation on conformational entropy for a host-guest system, 84 to find the entropy of unfolded protein chains, 87 the change of entropy upon folding 87 and binding, [88][89][90][91] the change of entropy upon chemical modifications, 92 mutations 93 and to assess the relationship of the conformational entropy of proteins with NMR measured HN order parameter, 87 which has been proposed as an experimental "proxy" for conformational entropy. 94,95 In all these works the periodic Euclidean metric was used, but other metrics could be used as well as detailed by Hnizdo et al 96 The kNN method for conformational entropy is currently implemented in a few software programs.…”