Polycomb proteins targeted by a short repeat RNA to the mouse X chromosome

Abstract

To equalize X-chromosome dosages between the sexes, the female mammal inactivates one of her two X chromosomes. X-chromosome inactivation (XCI) is initiated by expression of Xist, a 17-kb noncoding RNA (ncRNA) that accumulates on the X in cis. Because interacting factors have not been isolated, the mechanism by which Xist induces silencing remains unknown. We discovered a 1.6-kilobase ncRNA (RepA) within Xist and identified the Polycomb complex, PRC2, as its direct target. PRC2 is initially recruited to the X by RepA RNA, with Ezh2 serving as the RNA binding subunit. The antisense Tsix RNA inhibits this interaction. RepA depletion abolishes full-length Xist induction and trimethylation on lysine 27 of histone H3 of the X. Likewise, PRC2 deficiency compromises Xist up-regulation. Therefore, RepA, together with PRC2, is required for the initiation and spread of XCI. We conclude that a ncRNA cofactor recruits Polycomb complexes to their target locus.

Figure 1. The PRG2 complex contains Xist

A. Map of Xist. B,C,D. RIP in indicated cells, α, antibodies. E.. Effects of RNase pretreatment on RIP signals. F. Strand-specific RIP at R1 by realtime PCR, normalized to input RNA. Error bar, 1 standard deviation (SD). G. Quantitative RIP by realtime PCR at positions R1–R5. H. DNA ChIP using indicated antibodies, shown as a fraction of input and standardized to normal IgG ChIP.

Figure 2. A small RNA within Xist

A. Map of RepA and the 5′ end of Xist. B,C. Strand-specific realtime PCR quantitates RNA copies at R6–R9 in ES cells (B) or MEFs (C), normalized to standard curve. D. RNA FISH using RepA probe. E. Northern analysis of RepA and Tsix (5′ and 3′ positions). F. 3′ RACE of RepA. G. Transient transfection of luciferase reporter constructs comparing RepA (bp79–320) versus Xist P1 promoters, each normalized to vector control. P, student t-tests in indicated pairwise comparisons. H. DNA FISH of RepA transgenic female ES cells. Xist P1 promoter is not in transgenes. Arrows, transgene. Tsix detected by pSx7. I. Quantitative RIP in representative clones B5 and C5 +/− Doxycycline induction. No-antibody controls yielded no detectable RNA.

Figure 3. RepA RNA directly binds PRC2 in vitro.

A. One Repeat A unit. WT, wildtype sense. Mut, mutated. AS, antisense. DsI and DsII, randomized Xist sequences. B. EMSA using female ES nuclear extract (NE). Comp, competitors at 500X molar excess. Arrow, sense shift. Arrowhead, AS shift. C. Antisense binding competed by sense but not nonspecific RNAs. D. EMSA supershifts (*) with α-Ezh2. E. EMSA using recombinant hPRC2 (sub)complexes. fEzh2, Drosophila Ezh2. F. hPRC2 bound by antisense but not by randomized probes.

Figure 4. RepA/PRC2 knockdowns compromise XCI initiation

A. Xist RNA-H3K27me3 immunoFISH in knockdown clones. shRNA: RA, RepA. X1, Xist exon 1. Scr, scrambled control. B. Xist and Tsix levels at indicated positions in knockdown clones. C. ImmunoFISH: Frequency of Xist upregulation (Xist⁺) and H3-K27me3 foci. P, pairwise comparison against Scr-12 control for Xist⁺ frequencies. D. EB growth in shRNA clones. E. Xist-H3K27me3 immunoFISH after Eed, Ezh2, or control knockdown. F. Eed and Ezh2 mRNA levels after knockdown in Tsix^Δ/+ES cells. P, student t-test. G. qRT-PCR of Xist RNA after indicated knockdowns. P, student t-test. H. ImmunoFISH: Frequency of Xist upregulation and H3-K27 trimethylation after indicated knockdowns. P compares Xist foci numbers in controls (‘expected’) versus Eed/Ezh2 knockdowns.

Figure 5. PRC2 and Xi associate in the perinucleolar compartment after XCI

A. Immunostain: Ezh2 and Suz12 concentrate around the nucleolus (B23+). B. Ezh2 and Suz12, but not H3K9me3, showed perinucleolar enrichment. C. Xist RNA-Ezh2 immunoFISH. D. Xist DNA-Ezh2 immunoFISH in Xa^WTXi^ΔXist MEFs (6). E. Summary and model.