Identification and Validation of a Putative Polycomb Responsive Element in the Human Genome

Abstract

Epigenetic cellular memory mechanisms that involve polycomb and trithorax group of proteins are well conserved across metazoans. The cis-acting elements interacting with these proteins, however, are poorly understood in mammals. In a directed search we identified a potential polycomb responsive element with 25 repeats of YY1 binding motifthatwe designate PRE-PIK3C2B as it occurs in the first intron of human PIK3C2B gene. It down regulates reporter gene expression in HEK cells and the repression is dependent on polycomb group of proteins (PcG). We demonstrate that PRE-PIK3C2B interacts directly with YY1 in vitro and recruits PRC2 complex in vivo. The localization of PcG proteins including YY1 to PRE-PIK3C2B in HEK cells is decreased on knock-down of either YY1 or SUZ12. Endogenous PRE-PIK3C2B shows bivalent marking having H3K27me3 and H3K4me3 for repressed and active state respectively. In transgenic Drosophila, PRE-PIK3C2B down regulates mini-white expression, exhibits variegation and pairing sensitive silencing (PSS), which has not been previously demonstrated for mammalian PRE. Taken together, our results strongly suggest that PRE-PIK3C2B functions as a site of interaction for polycomb proteins.

Figure 1. Effect of PRE-PIK3C2B on transcription.

A. Part of the vectors used in transfection assays: CMV represents the promoter, GFP is the reporter present in all the constructs. pcDNA3.1-GFP is control plasmid without PRE-PIK3C2B, pc(1 kb)UP is pPRE-PIK3C2B/UP-GFP and pc(1 kb)DN is pPRE-PIK3C2B/DN-GFP with PRE-PIK3C2B (inverted triangle) cloned upstream and downstream of the reporter respectively. B. GFP expression indicated as ratio relative to that from pcDNA3.1GFP. pc(25 mer)11UP and pc(25 mer)11DN are 25 mer oligo corresponding to repeating unit of PRE-PIK3C2B cloned upstream and downstream of the GFP reporter respectively. pcΔYY1UP and pcΔYY1DN are repeating unit without YY1 motif cloned upstream and downstream of the GFP reporter respectively. Error bars, S.E.M of assay triplicate is shown. (**) p-value<0.005, n = 3

Figure 2. Directinteraction of YY1 with PRE-PIK3C2B.

Labeled 25 merOligo corresponding to repeating unit of PRE-PIK3C2B (5′AGTGAAGCCATCATGTGAGAATACC3′) was used as the probe in gel mobility shift assay with purified His-YY1. Cold probe 1 is competing unlabeled cognate (Oligo25 mer), 2- (OligoΔYY1) repeating unit without YY1 recognition sequence (5′AGTGAAGTGAGAATACC3′). Each competing oligo was used at a concentration 100× higher than the labeled probe. The super shift with anti-YY1 is shown in the last lane (arrow).

Figure 3. Interaction of PcG proteins with PRE-PIK3C2B in HEK cells.

The experiments were carried out in HEK293 ells. A-Diagrammatic representation of the endogenous PRE-PIK3C2B region with primers (arrows) used for PCR and the 25 mer repeating unit (filled box). B and C- ChIP with antibodies indicated. Input is 20% of sonicated chromatin. D-Quantitative PCR for PRE-PIK3C2B following ChIP in control cells and cells transfected with siRNAYY1 and siRNA SUZ12. E-Interaction of EZH2 with PRE-PIK3C2B estimated by qPCR in absence and presence of YY1siRNA: F-Localization of EED and the H3K4me3 at PRE-PIK3C2B in absence and presence of siRNAYY1. Error bars: S.E.M of assay in triplicate, * p<0.05, **p-value<0.005,n = 3. The qPCR profile for positive and negative controls are shown in Figs. S3 and S4.

Figure 4. YY1 and SUZ12 dependent transcriptional repression by PRE-PIK3C2B.

A-Expression of endogenous PIK3C2B in HEK cells analysed by qPCR. The expression is increased in presence of siRNAYY1 and siRNASUZ12. Expression in untransfected HEK cells is taken as control. B- Effect on reporter gene expression. The nomenclature for vectors is same as in Figure 1. Ratio of GFP expression in pcDNA-GFP, pc(1 kb)UP or pc(1 kb)DN with and without co-transfection with siRNA YY1 and siRNA SUZ12 is shown on the Y-axis. Error bars, S.E.M of assay in triplicate is shown * p-value<0.05, ** p-value<0.005, n = 3.

Figure 5. Pairing sensitive silencing (PSS) of miniwhite expression.

Eye pigmentation due to miniwhite expression in transgenic flies of PI-17 line are shown. I- P/+ and P/P, heterozygous and homozygous transgenic lines, homozygous flies show PSS; II- ΔP/+ and ΔP/ΔP, are transgenic flies where PRE-PIK3C2B is flipped-out by crossing with Cre flies. The frames a-d show rescue of PSS in different genetic backgrounds as indicated. Arrow in frame a marks PSS in homozygous flies, arrow in frame c and d mark the rescue of PSS.

Figure 6. Genetic interactions of PRE-PIK3C2B with PRC genes in transgenic flies.

Effect of PRE-PIK3C2B in the background of different PcG mutations on rescue of eye pigmentation is shown: cases of rescue of eye colour are marked with an arrow; trans-heterozygotes show better recovery of eye pigmentation. We have used same conditions of image processing in all the cases.

Figure 7. Effect of Pho mutation on PRE-PIK3C2B mediated repression.

A- Expression of miniwhite mRNA by qPCR. The genetic background is indicated below the histograms.* p<0.05, **p<0.005. B- comparison of eye pigmentation between transgenic flies with and without PHO mutation. ΔPI indicates the lines where PRE-PIK3C2B is flipped out.

Figure 8. Genetic interactions of PRE-PIK3C2B with Trx genes in transgenic flies.

Effect of PRE-PIK3C2B in the background of different TrxG mutations on eye colour is shown: single mutants zeste(A), brm (B) and double mutant zeste and brm(C). The left panel shows the transgenic flies heterozygous for PRE-PIK3C2B and the right panel are transgenic flies deleted for PRE-PIK3C2B in similar background. We have used same conditions of image processing in all the cases.