Skip to main content
. 2019 Aug 16;8:e47835. doi: 10.7554/eLife.47835

Figure 2. A repressive chromatin environment at EIN2 down-regulates its expression.

(A) Genome browser screenshot visualizes the levels of the depicted chromatin features at the EIN2 gene in untreated 3-day-old etiolated Ler, ein6-1, een-1 and ein6-1 een-1 seedlings. Occupancy of H3K27me3, H2A.Z, H3K4me3 and H2Aub was determined with ChIP-seq, mRNA expression was measured with RNA-seq and levels of methylated cytosines (CG in yellow, CHG in blue, CGG in pink) were determined with MethylC-seq. To ensure an accurate comparison of individual chromatin features between genotypes, the tracks were normalized to the respective sequencing depth. Normalization was separately done for each chromatin feature. Biological replicate 1 of the H3K27me3 and H2A.Z ChIP-seq datasets is shown. (B) Venn diagram illustrates the overlap between genes that show a significant increase of H3K27me3 (2-fold enrichment over ein6-1), H2AZ (2-fold enrichment over een-1) and H2Aub (2-fold enrichment over Ler) in ein6-1 een-1 mutants and also a significant decrease of H3K4me3 (1.5-fold enrichment in Ler over ein6-1 een-1). In addition, genes that contain differentially methylated regions (DMRs) with ten or more methylated cytosines in at least one genotype were included as well. (C) Graphical illustration of H3K27me3 and EIN6 occupancy at the EIN2 gene determined with ChIP-seq. Sequencing reads were merged between biological replicates for the H3K27me3 ChIP-seq using untreated 3-day-old etiolated Ler (gray) and ein6-1 een-1 (blue) seedlings (two replicates each) and the EIN6 ChIP-seq (red) using Ler 35S:EIN6-FLAG seedlings (three replicates). The occupancy was calculated as the ratio between the respective merged ChIP and the merged Ler IgG control in 100 bp bins from 2.4 kb upstream to 7.7 kb downstream of the transcriptional start site (TSS) of EIN2 and is shown as log2 fold change. Negative values which reflect lower occupancy in the ChIP sample compared to the IgG control sample were set to zero.

Figure 2.

Figure 2—figure supplement 1. Quantitative description of the chromatin environment at EIN2 in ein6-1 een-1 mutants.

Figure 2—figure supplement 1.

(A) Protein domain structure of EEN as well as a multiple amino acid sequence alignment of EEN, IES6 and human IES6 (hIES6). The YL1-C domain is indicated in blue. (B) Aggregated H3K27me3 profile of 2369 Group I genes shows H3K27me3 occupancy from 1 kb upstream to 2 kb downstream of the TSS in Ler, ein6-1, een-1 and ein6-1 een-1 seedlings. (C) Genome browser screenshot shows differential enrichment of H3K27me3 at an example Group I gene and Group II gene. To ensure an accurate comparison of individual chromatin features between genotypes, the tracks were normalized to the respective sequencing depth. (D) H3K27me3 occupancy of 54 Group II genes in Ler, ein6-1, een-1 and ein6-1 een-1 seedlings is shown as an aggregated H3K27me3 profile from 1 kb upstream to 2 kb downstream of the TSS. The H3K27me3 occupancy was calculated as the ratio between the two respective merged ChIP replicates and the two merged Ler IgG control replicates. (E) Quantification of H3K27me3 levels in the 5’UTR intron, gene body and 3’UTR of the EIN2 gene are shown. The H3K27me3 occupancy in these regions was calculated as the ratio between the respective merged ChIP-seq samples and merged Ler IgG control samples. (F) Spearman’s correlation plot shows correlation of read coverages between the antibody validation H2A.Z datasets from this study (Col-0, pie1-1, Col-0 HTA11:HTA11-GFP (αH2A.Z) and Col-0 HTA11:HTA11-GFP (αGFP)) and three publicly available H2A.Z ChIP-seq datasets (Carter et al., 2018; Wollmann et al., 2017; Coleman-Derr and Zilberman, 2012). Clustering is determined by the degree of correlation. (G) Heatmap shows the H2A.Z occupancy at all Arabidopsis genes in the indicated genotypes. Levels of H2A.Z from 1 kb upstream to 2 kb downstream of the TSS are shown. (H) Genome browser screenshot shows differential enrichment of H2A.Z in Col-0, pie1-1 and Col-0 HTA11:HTA11-GFP seedlings. Moreover, H2A.Z enrichment is also shown for three publicly available H2A.Z ChIP-seq datasets (Carter et al., 2018; Wollmann et al., 2017; Coleman-Derr and Zilberman, 2012). Genetic background and used antibodies are indicated. The Col-0 IgG track serves as a control and the Ler 5mC track indicates methylated cytosines (CG in yellow, CHG in blue, CGG in pink). The shape difference of H2A.Z domains in the Coleman-Derr and Zilberman, 2012 dataset can be explained by the MNase treatment of the chromatin. (I) Levels of H2A.Z in the 5’UTR intron, gene body and 3’UTR of EIN2 in Ler, ein6-1, een-1 and ein6-1 een-1 seedlings are shown. The H2A.Z occupancy in these regions was calculated as the ratio between the respective merged ChIP-seq samples and merged Ler IgG control samples.
Figure 2—figure supplement 2. EIN2 displays an aberrant chromatin signature ein6-1 een-1 mutants and is a direct target of EIN6.

Figure 2—figure supplement 2.

(A), (B) Quantification of H3K4me3 (A) and H2Aub (B) levels in the 5’UTR intron, gene body and 3’UTR of EIN2 are shown. The occupancy in these regions was calculated as the ratio between the respective ChIP-seq samples and the Ler IgG control sample. (C) Levels of DNA methylation were detected with MethylC-seq and the three DMRs within the gene body of EIN2 are shown. Exact localization of this three DMR’s (DMR1: chr5, 799482–790672; DMR2: chr5, 791018–791663; DMR3; chr5, 791980–792344) is indicated in a genome browser screenshot showing levels of CG, CHG and CHH at EIN2 in Ler seedlings. (D) Schematic representation indicates the localization of two CTCTGYTY-motifs in the 5’UTR region of EIN2. (E) (F) The CTCTGYTY-motif was discovered as the top-ranked motif in EIN6 ChIP-seq data (E) as well as in EIN6-ZNF DAP-seq data (F) determined by MEME analysis using the 1000 top-ranked peaks that were identified by GEM. (G) Genome browser screenshot visualizes the in vivo binding of full length EIN6 and the in vitro binding of its C-terminal zinc fingers (EIN6-ZNF) to EIN2 and TOUCH4 (TCH4). EIN6-FLAG ChIP-seq and EIN6-ZNF-HALO DAP-seq was employed to monitor in vivo and in vitro binding, respectively. ChIP-seq data from a previous study (Cui et al., 2016) showing, that REF6 but not a truncated REF6 version (REF6ΔZNF) can bind to EIN2 and TCH4, is shown as well. (H) Triple response phenotype of 3-day-old etiolated seedlings of Ler, ein6-1 een-1 and ein6-1 een-1 overexpressing either a wild-type EIN6 version or a truncated version without the zinc fingers (EIN6ΔZNF). Seedlings were grown on control LS media or on LS media supplemented with 10 µM ACC.