Fig 2: Chromatin Signatures of Nr2F1 and Hoxa5 in F9 WT and RAR knockout cells. (A) Schematic of Nr2F1 IEE and promoter regions. Nr2F1 putative TSSs are indicated by angled arrows (P1, P2 and P3). The proximal nucleotide sequences are shown for the Nr2F1 promoter and IEE region, and the putative enhancer element is underlined. The GC content of the evaluated regions is illustrated in the background graphs, with CpG islands marked in gray. Note the local depletion of CpGs proximal to the Nr2F1 RefSeq promoter (P3). White arrows indicate genomic regions included in the Nr2F1 reporter constructs (Fig. 2B, 0.5 and 1.0 kb, respectively). The gray bars specify primary transcripts from the indicated TSSs. (B) ChIP analysis shows that Polycomb repressive marks (Suz12, H3K27me3 and Ring1B) decrease at the Hoxa5 promoter and RARE, but increase at the Nr2F1 promoter and IEE in F9 WT cells in response to a 24 h RA treatment (left, boxed). Similar patterns of Nr2F1 epigenetic changes were observed in RAR knockout cells (KO, right). (C) ChIP analysis shows that transcriptional permissive marks (H3K9/K14ac, H3K4me3 and polII-CTD) increase at the Hoxa5 promoter and RARE, and at the Nr2F1 promoter and IEE in response to a 24 h RA treatment (left). Similar patterns of Nr2F1 epigenetic changes were observed in RARa knockout cells, whereas H3K9/14ac levels did not increase in RARß2 and RAR? knockout cells (right). (D) ChIP analysis shows Histone 3 occupancy (H3) and the non-specific background signal (IgG). The ChIP signals are depicted relative to the total chromatin input in each ChIP. Statistical significance (P < 0.05) is indicated by asterisks for the effect of RA, (n.s.: non-significant).

Fig 3: Summary model of Nr2F1 and Hoxa5 epigenetic signatures in response to RA. Nr2F1 and Hoxa5 display different epigenetic signatures on RA treatment of stem cells. Nr2F1 is characterized by increased levels of PRC (Suz12 and Ring1B) and the associated H3K27me3 histone mark in response to RA. In contrast, the Hoxa5 epigenetic signature is characterized by dissociation of PRC (Suz12 and Ring1B) and reduction of the H3K27me3 histone mark. Note that the epigenetic signatures of Hoxa5 are similar in presence and absence of Suz12 after RA addition. In contrast, the epigenetic signature of Nr2F1 on RA treatment differs in the presence and absence of Suz12, thus potentially explaining the increased transcriptional activity on Suz12 knockdown. For both Nr2F1 and Hoxa5, the transcriptional induction is marked by increased levels of H3K4me3, H3K9ac and H3K14ac permissive histone marks. The different marks are depicted as shapes whose sizes reflect the relative abundance in the specified condition. Histone H3 is depicted as a gray circle. PRC1 and PRC2 are represented by Ring1B and Suz12, respectively.

Fig 4: shRNA-mediated Suz12 knockdown increases Nr2F1 and Nr2F2 transcript levels. In F9 cells, Suz12 protein and transcript levels (upper left and upper right, respectively) are reduced below detection and to 10%, respectively, by expression of a Suz12-specific shRNA. The Suz12 knockdown eliminates the PRC2-specific histone mark (H3K27me3), but has only a limited effect on the permissive H3K4me3 mark (western blot, upper left). The Suz12 knockdown increases Nr2F1 (middle left) and Nr2F2 (bottom left) transcript levels (24 h RA treatment) by quantitative RT-PCR, but has no effect on Hoxa5 (middle right) and Hoxa1 (bottom right) transcript levels. Statistical significance is indicated by double asterisk for the effect of RA (relative to vehicle-treated cells), and by asterisk for the effect of Suz12 knockdown in RA-treated conditions (48 h, relative to shLuc control cells).

Fig 5: The response to Suz12 depletion distinguishes two classes of RA-inducible genes. (A) In ES cells, Suz12 depletion increases Nr2F1, Nr2F2, Meis1, Sox9 and BMP2 transcript levels (upper panel), but has no effect on Hoxa5, Hoxa1, Cyp26a1, Cyp26b1 and RARß2 transcript levels (lower panel) on RA treatment. Transcripts were measured by quantitative RT-PCR. Statistical significance is indicated by double asterisk for the effect of RA (relative to vehicle-treated cells), and by asterisk for the effect of Suz12 knockdown in RA-treated cells (48 h, relative to shLuc control cells). (B) Suz12 transcript levels are diminished by expression of a Suz12-specific shRNA, whereas transcript levels of the 36B4 reference gene are unaffected. (C) The levels of Nr2F1 protein are elevated on Suz12 depletion and increase in an RA-dependent manner by western blotting. The levels of Polycomb proteins Suz12 and Ezh2 are reduced on depletion of Suz12, whereas actin levels remain unchanged by western blotting. Western blots were performed at least three times with different cell extracts; a representative blot is shown.