The apparent loss of PRC2 chromatin occupancy as an artefact of RNA depletion

Abstract

RNA has been implicated in the recruitment of chromatin modifiers, and previous studies have provided evidence in favour and against this idea. RNase treatment of chromatin is a prevalent tool for the study of RNA-mediated regulation of chromatin modifiers, but the limitations of this approach remain unclear. RNase A treatment during chromatin immunoprecipitation (RNase-ChIP or rChIP) reduces chromatin occupancy of the H3K27me3 methyltransferase PRC2. This led to suggestions of an "RNA bridge" between PRC2 and chromatin. Here we show that RNase A treatment during chromatin immunoprecipitation leads to the apparent loss of all facultative heterochromatin, including both PRC2 and H3K27me3 genome wide. This phenomenon persists in mouse embryonic stem cells, human cancer cells and human-induced pluripotent stem cells. We track this observation to a gain of DNA from non-targeted chromatin, sequenced at the expense of DNA from facultative heterochromatin, which reduces ChIP signals. Our results point to substantial limitations in using RNase A treatment for mapping RNA-dependent chromatin occupancy and invalidate conclusions that were previously established for PRC2 based on this assay. HighlightsO_LIRNA degradation during ChIP-seq is insufficient to displace PRC2 from chromatin. C_LIO_LIRNA degradation led to the artificial depletion of ChIP-seq signals in multiple cell lines. C_LIO_LIArtificially reduced ChIP-seq signals are explained by a gain of non-targeted DNA. C_LIO_LIRNA is critical in maintaining the solubility of chromatin during experimentation. C_LI