6/ A fantastic collaboration between the @DavidovichLab and the @Eddy_McGlinn labs, led by a brilliant PhD student, @SamAgius0, who is now looking for a postdoc position!

5/ We propose that H3K27me3 mimicry has repeatedly emerged during evolution as a mechanism to antagonise PRC2 and restrict the spread of Polycomb domains.

4/ In mouse ESCs, the same mutations lead to increased H3K27me3 levels and enhanced spreading of Polycomb domains.

3/ In mouse embryos, we show that mutations that disrupt H3K27me3 mimicry delay the activation of Hox genes and cause homeotic transformations characteristic of Polycomb gain of function.

2/ Some accessory subunits of PRC2 mimic the repressive mark H3K27me3. This was proposed to “jump-start” PRC2 during de novo H3K27me3 deposition, but the physiological relevance remained unknown.

8/ We propose that H3K27me3 mimicry has repeatedly emerged through evolution to restrict the spread of Polycomb domains by breaking the PRC2-H3K27me3 positive feedback loop.

7/ We find that the allosteric regulatory activities of JARID2 and PALI1 synergise to antagonise PRC2 during mouse development and restrict the spread of Polycomb domains in mECS.

6/ Unexpectedly, the H3K27me3 mimicry-defective JARID2 K116R exhibited a homeotic transformation characteristics of Polycomb gain of function! This seemingly goes against the prevalent paradigm where JARID2 K116me3 is believed to positively regulate PRC2.

5/ To determine what is the function of H3K27me3 mimicry, we used CRISPR/Cas to generate 3 knock-in mouse lines: each included a single K-to-R mutation that specifically prevents JARID2, PALI1 or PALI2 from mimicking H3K27me3.

4/ Why do JARID2 and PALI1/2 mimic H3K27me3? They are thought to "jump start" PRC2 in the absence of H3K27me3, to initiate H3K27me3 de novo. While this model agrees with biochemical and KO+rescue experiments in cells, it was never tested in vivo.