The recruitment of chromatin modifiers by long noncoding RNAs: lessons from PRC2

Abstract

Polycomb repressive complex-2 (PRC2) is a histone methyltransferase required for epigenetic silencing during development and cancer. Among chromatin modifying factors shown to be recruited and regulated by long noncoding RNAs (lncRNAs), PRC2 is one of the most studied. Mammalian PRC2 binds thousands of RNAs in vivo, and it is becoming a model system for the recruitment of chromatin modifying factors by RNA. Yet, well-defined PRC2-binding motifs within target RNAs have been elusive. From the protein side, PRC2 RNA-binding subunits contain no known RNA-binding domains, complicating functional studies. Here we provide a critical review of existing models for the recruitment of PRC2 to chromatin by RNAs. This discussion may also serve researchers who are studying the recruitment of other chromatin modifiers by lncRNAs.

Keywords: PRC2; RNA–protein interaction; epigenetic silencing; histone modification; long noncoding RNAs.

FIGURE 1.

Multiple factors involved in the recruitment of PRC2 to chromatin. PRC2 is a multisubunit histone methyltransferase complex that is regulated and recruited to its target genes through interactions with various factors. These include direct and indirect interactions with nucleosomes, multiple proteins, RNA (which also inhibits the histone methyltransferase activity of PRC2), and possibly DNA. The dashed lines are not meant to indicate which subunit of PRC2 is involved in which potential interaction.

FIGURE 2.

Previously proposed models for the recruitment of PRC2 to chromatin by lncRNA, and RNA in general, compared with RNA-independent recruitment mechanisms. (A) Eviction (“Junk Mail Model”): Promiscuous RNA binding to nascent transcripts leads to eviction of PRC2 from highly active genes (Davidovich et al. 2013; Kaneko et al. 2013) and inhibits its HMTase activity (Cifuentes-Rojas et al. 2014; Herzog et al. 2014; Kaneko et al. 2014b). Simultaneously, H3K4me3 and H3K36me3 active chromatin marks prevent deposition of PRC2 to nucleosomes and inhibit its HMTase activity (Schmitges et al. 2011; Yuan et al. 2011). (B) RNA-independent recruitment: PRC2 is recruited to chromatin through interactions with protein-binding factors, nucleosomes, and/or DNA (for reviews, see Ringrose and Paro 2007; Margueron and Reinberg 2011; Di Croce and Helin 2013; Simon and Kingston 2013; Comet and Helin 2014). (C) Direct and specific interactions with lncRNAs: Site-specific recruitment of PRC2 could occur in cis or in trans (references in text). (D) Bridging and remodeling: Recruitment of PRC2 by RNA can be mediated through protein bridging factors, such as JARID2 (da Rocha et al. 2014; Kaneko et al. 2014a,b), or through RNA structure remodeling, as suggested for ATRX (Sarma et al. 2014). (E) Masking: PRC2 is masked from binding to certain RNA transcripts that are already bound by other factors, thus providing a binding preference (Herzog et al. 2014). (F) Scanning (“Junk Mail Model”): PRC2 interacts with nascent RNA transcripts promiscuously (Davidovich et al. 2013; Kaneko et al. 2013) and scans for repressive epigenetic marks or recruiting factors. Unless deposition to nucleosomes takes place, PRC2 is poised and in check (Kaneko et al. 2014b) while its HMTase activity is inhibited by the RNA. (G) Maintenance: After repression is achieved, PRC2 maintains the repressed epigenetic state through direct binding to nucleosomes carrying H3K27me3 marks (Hansen et al. 2008), which also stimulate its HMTase activity (Margueron et al. 2009). (H) PRC2-independent repression: Transcription shutoff of a Polycomb target gene can take place in a PRC2-independent manner and can lead to subsequent recruitment of PRC2 (Riising et al. 2014). Importantly, most of these models are not mutually exclusive.

FIGURE 3.

Alternative explanations for observations that were previously provided to support the recruitment of PRC2 for epigenetic repression through specific interactions with lncRNAs. (A) In undifferentiated cells the differentially regulated gene is active and the lncRNA is repressed or lowly expressed. The cascade of events that could take place through differentiation is illustrated by three models, all of which are compatible with current literature. Model 1 (top, purple arrows): A given lncRNA is differentially expressed and binds PRC2 by specific protein–RNA interactions, either directly or through a bridging factor. This leads to the recruitment of PRC2 to the regulated gene (dashed arrow). Next, PRC2 introduces the H3K27me3 mark. Model 2 (middle, red arrows): The lncRNA is differentially expressed and also binds PRC2, possibly through promiscuous protein–RNA interactions. Yet, the association of PRC2 with the lncRNA does not directly cause its deposition to chromatin. Instead, recruitment of PRC2 to the regulated gene takes place independently by other factors (for reviews, see Ringrose and Paro 2007; Margueron and Reinberg 2011; Di Croce and Helin 2013; Simon and Kingston 2013; Comet and Helin 2014). Model 3 (bottom, red arrows): The differentially regulated gene is repressed in a PRC2-independent manner. The lncRNA could function in this process, but it is not the driving force for the recruitment of PRC2, which only interacts with the RNA through promiscuous interactions. Next, transcription shutoff leads to the recruitment of PRC2 to chromatin (Riising et al. 2014), where it maintains the repressed epigenetic state. In all three models, the end point is identical: In differentiated cells the lncRNA is expressed while the differentially regulated gene is epigenetically repressed, with PRC2 maintaining the repressed epigenetic state. (B–D) Hypothetical in vivo data commonly provided to support the recruitment of PRC2 by specific interactions with lncRNAs (see text for details). (B) Expression analysis supports all three models, but cannot exclude any of them. (C) Chromatin immunoprecipitation (ChIP) for the differentially expressed gene identifies the recruitment of PRC2 to chromatin during differentiation (α-EZH2), deposition of its repressive epigenetic mark (α-H3K27me3), and reduction in RNA polymerase occupancy over the regulated gene (α-Pol II), yet these observations agree with all three models and are therefore not definitive. (D) RNA immunoprecipitation confirms that PRC2 is associated with the lncRNA to a greater extent, compared with a negative control RNA, to a statistically significant degree. Although this observation is consistent with Model 1, it is not definitive and can also be explained by an undetected bridging factor or simply by a higher degree of promiscuous interactions, as previously observed between PRC2 and various nonrelevant RNAs in vivo (Kaneko et al. 2013) and suggested to be the result of competition between PRC2 and other RNA-binding proteins (Herzog et al. 2014).