Survivin prevents the Polycomb Repressor Complex 2 from methylating Histone 3 lysine 27

Abstract

Survivin is a small protein that belongs to the inhibitor of apoptosis protein family and participates in cell division and apoptosis. It was actively studied in human cancers, inflammatory diseases and in autoimmune diseases. Here, we reveal that survivin takes part in epigenetic gene silencing by interaction with the polycomb repressive complex 2 (PRC2). PRC2 silences gene expression through tri-methylation of lysine 27 on histone 3 (H3K27). We detected differential expression of PRC2 core subunits in CD4+T cells with different survivin expression. ChIP-seq experiments indicated that survivin binds chromatin that overlap with the regions occupied by PRC2. ChIP-seq of H3K27 in CD4+T cells indicate that inhibition of survivin leads to a substantial increase in H3K27 tri-methylation by PRC2 in contrast to other histone modifications, which lends support to that survivin prevents PRC2 from functioning. Survivin binds peptides derived from PRC2 subunits Ezh2, EED, Suz12 and Jarid2 in a peptide microarray that cover intersubunit interfaces, catalytic residues, and present binding sites for substrates, DNA, and regulatory proteins. Amino acid composition of the peptides has substantial predictive power for survivin interaction in the peptide microarray as determined by multilayer perceptron classification analysis. NMR experiments with 15N labelled survivin indicate that peptide colocalization does not entirely depend on binding mediated by short range interactions. These results indicate that survivin interacts with PRC2, preventing the methylation of H3K27 and specific gene silencing.

Femtosecond X-ray snapshots reveal correlated displacements of specific distal atoms in a protein crystal

“…Concerning the biological function of these groups, there has been speculation about which properties of peptides are crucial in determining their interaction with promiscuous proteins such as survivin. In this example, it appears that the sequence does not provide substantially more deterministic power compared to composition 62,63 . However, for an effective model, the composition must include a detailed description of functional groups in the amino acids, rather than simply considering the number of hydrophobic and hydrophilic amino acids, net charge, and charge density.…”