ChIP-seq of 3xFLAG-AHL15 in native and overexpression conditions.

A: Heatmap and profile plots of AHL15 ChIP-seq peaks scaled over genes showing that AHL15 is enriched near the TSS and TES, and is largely absent over gene bodies. Shoots of ten day-old pAHL15:3xFLAG-AHL15 were used for ChIP for the “native” condition, and 31 day-old p35S:3xFLAG-AHL15 shoots were used for the “overexpression” condition. Profile plots show the log2 of ChIP reads over input reads. B: Annotation of 3xFLAG-AHL15 ChIP-seq peaks in A over genomic features of the Araport11 annotation of the Arabidopsis genome. TES: region up to 1 kb upstream of the TES. Promoter: region up to 3 kb downstream of the TSS. TES / promoter: regions where promoters overlap with the TES region of neighbouring genes. Intergenic: regions not covered by any of the before features. C: Venn diagram showing the overlap in peak annotations between the five ChIP-seq samples in A. 1948 peaks were identified in all five datasets (shared), 2906 peaks were shared between the three overexpression conditions and the second native ChIP (semi-shared), and 4999 peaks were identified in all three overexpression (OX) conditions (OX only). D: Histogram of the peak score of significant ChIP-seq peaks in p35S:3xFLAG-AHL15 sample 2 (OX 2). Peaks identified in all five conditions have the highest peak score, followed by semi-shared peaks. OX-only peaks have a lower peak score than shared or semi-shared peaks, and peaks shared in less than 3 conditions have the lowest overall peak score. Reads were aligned with Bowtie2 and peak calling was performed with MACS2.

AHL15 binds to AT-rich DNA but does not recognize a specific DNA binding motif.

A: Table showing the top 8 enriched motifs identified by HOMER2 within the set of shared peaks, with the p-value of motif enrichment, log10 of the p-value of motif enrichment, the percent of AHL15 peaks (targets) containing the motif, and the percent of background sequences containing the motif. B: Graph showing the AHL15 targets over background, meaning the number of times each motif (48 in total) is identified by HOMER2 in shared AHL15 peaks divided by its number over the entire genome (expressed as %; 100% means the motif is almost only present in AHL15 peaks). C: Graph showing the frequency with which each motif in B identified by HOMER2 occurs in shared AHL15 peaks, expressed as the log2 value of the % of shared AHL15 targets containing the specific motif (occurrence of more than 1% gives a positive value, 1% gives a value of 0, and less than 1% results in a negative value). Motifs in figures B and C are arranged by descending target : background value.

AHL15 binding at DNA regions near TES and TSS is enhanced for genes differentially expressed by AHL15-GR activation.

A: Volcano plot showing differentially expressed genes in 10-day-old p35S:AHL15-GR seedlings after 8 hours treatment with DEX compared to the three control conditions (p35S:AHL15-GR seedlings treated with mock solution and Col-0 seedlings treated with either DEX or mock solution). Blue dots indicate genes that are significantly more than 2-fold downregulated, red dots indicate genes that are significantly more than 2-fold upregulated, light grey dots show genes that are significantly differentially expressed with a fold change of less than 2, and dark grey dots (black when overlapping) represent genes whose expression was not significantly altered. B-D: Venn diagrams showing overlap of differentially expressed genes with shared ChIP-seq peaks (B), semi-shared peaks (C), or overexpression-only peaks (D). E: Table showing a selection of genes shared between ChIP-seq and RNA-seq datasets identified in B-D. Log2FC: log2 fold change in gene expression. F: Profile plots of AHL15 ChIP-seq peaks over AHL15-bound differentially expressed genes vs all genes. AHL15 enrichment is stronger near the TSS of AHL15-bound differentially expressed genes and especially upregulated genes compared to the average of all genes. Near the TES, AHL15 enrichment is highest for downregulated AHL15-bound genes.

AHL15 delays plant developmental phase transitions in a chromatin accessibility-independent manner.

A, B: ATAC-seq on 10 day-old p35S:AHL15-GR seedlings that were mock- or DEX treated for 30 minutes, 1 hour or 8 hours. A: Principal Component Analysis (PCA) of ATAC-seq samples, showing that most samples cluster together and that differences between replicates of the same condition (indicated by 1 and 2 for the same colour) are similar or larger than differences between conditions. B: ATAC-seq profiles (purple and green) and AHL15 ChIP-seq profiles (blue) at the FT locus, a gene that was strongly repressed 8 hours after DEX treatment. Each track shows the overlay of bigwig files from two (ATAC-seq and AHL15 native ChIP-seq) or three (AHL15 OX ChIP-seq) biological replicates. Arrows in the intron of the FT gene model indicate transcriptional direction. C: Phenotype of mock-treated and DEX-treated p35S:AHL15-GR plants at two weeks after DEX treatment (24 day-old plants). D: Bolting time of p35S:AHL15-GR plants following mock or DEX treatment, as shown in C. Seedlings were transferred to soil two days after DEX treatment and monitored for flowering time (n = 25; Wilcoxon test, p = 3.71·10-9). Plants were grown in long day conditions (16 hours light : 8 hours dark).

AHL15 binds DNA regions characterised by reduced chromatin accessibility and a depletion of epigenetic marks.

A-D: ChIP-seq profiles of pAHL15:3xFLAG-AHL15 (AHL15 native) and p35S:3xFLAG-AHL15 (AHL15 OX) scaled over all Arabidopsis genes and compared to the ATAC-seq profile of p35S:AHL15-GR plants treated with DEX or mock solution for 1 hour (A), or to the ChIP-seq profiles of Histone 1 (H1) (B), Histone 2 variants H2A, H2A.X and H2A.Z (C), or the Histone 3 methylation marks H3K9me1, H3K9me2, and H3K27me1 (D). E: Alignment of the AHL15 native (AHL15_native) and overexpression (AHL15_OX) ChIP-seq profiles (light and dark blue, respectively) at the TCP16 and EULS3 loci with the averaged ATAC-seq profiles of ten day-old p35S:AHL15-GR seedlings that were mock- (green) or DEX (orange) treated for 30 minutes, 1 hour or 8 hours. Below the AHL15 ChIP-seq profiles, the ChIP-seq profiles of the epigenetic marks mentioned in B-D are shown.

AHL15 shows overlapping DNA binding with GH1-HMGA2/HON5, suggesting a role in gene looping.

A: Profile plot of GH1-HMGA2/HON5 ChIP-seq reads from (Zhao et al., 2021) scaled over Arabidopsis genes, showing a similar enrichment pattern as AHL15. B: GH1-HMGA2/HON5 ChIP-seq reads are were plotted over three classes of AHL15 ChIP-seq peaks presented in Figure 1C shared (identified in native and overexpression conditions, n = 1948), semi-shared (identified in one native sample and in all overexpression samples, n = 2906), and OX-only (identified in all three overexpression samples, n = 4999). C: AHL15 and GH1-HMGA2/HON5 ChIP-seq peaks around the FLC and RCD1 loci. The RNA-seq tracks below indicate mapped reads per strand; the FLC gene is located on the reverse strand (R) and RCD1 is located on the forward (F) strand. The ChIP-seq and RNA-seq tracks show an average of biological replicates of the same condition. D: Graph showing the percentages of self-looping genes in the individual native or overexpression (OX) ChIP-seq samples not bound by AHL15, bound by AHL15 and upregulated upon DEX treatment, bound by AHL15 and downregulated upon DEX treatment, and bound by AHL15 but not differentially expressed (d.e.). (**** indicates significant difference in the χ2 test, p < 0.0001). E< 0.0001). E: Table showing the number of self-looping genes identified in each ChIP-seq sample and sub-group shown in D.

Phenotype of 31 day-old p35S:3xFLAG-AHL15 plants right before harvesting for ChIP.

White arrowheads mark plants with strong AHL15-induced delay of development that were used for 3xFLAG-AHL15 overexpression ChIP. Plants were grown in long day condition (16h light / 8h dark), and the scale on the bottom of the image is in cm.

HOMER de novo motif finding results for overexpression-only peaks.

A: Table of the top 8 enriched motifs identified by HOMER2 within the set of overexpression-only peaks, with the p-value of motif enrichment, log10 of the p-value of motif enrichment, the percent of AHL15 peaks (targets) containing the motif, and the percent of background sequences containing the motif. B: The number of times each motif (61 in total) is identified by HOMER2 in overexpression-only AHL15 peaks divided by its number over the entire genome (expressed as %: 100% means the motif is almost only present in AHL15 peaks). C: The frequency with which each motif identified by HOMER2 occurs in overexpression-only AHL15 peaks, expressed as the Log2 value of the % of shared AHL15 targets containing the specific motif (occurrence of more than 1% gives a positive value, 1% gives a value of 0, and less than 1% results in a negative value). Motifs in figures B and C are arranged by descending target : background value.

HOMER de novo motif finding results for semi-shared peaks.

A: Table of the top 11 enriched motifs identified by HOMER2 within the set of semi-shared peaks, with the p-value of motif enrichment, log10 of the p-value of motif enrichment, the percent of AHL15 peaks (targets) containing the motif, and the percent of background sequences containing the motif. B: The number of times each motif (57 in total) is identified by HOMER2 in semi-shared AHL15 peaks divided by its number over the entire genome (expressed as %: 100% means the motif is almost only present in AHL15 peaks). C: The frequency with which each motif identified by HOMER2 occurs in semi-shared AHL15 peaks, expressed as the Log2 value of the % of shared AHL15 targets containing the specific motif (occurrence of more than 1% gives a positive value, 1% gives a value of 0, and less than 1% results in a negative value). Motifs in figures B and C are arranged by descending target : background value.

HOMER de novo motif finding results calculated based on all AHL15 peaks (shared, semi-shared, and overexpression-only).

A: Table of the top 20 enriched motifs identified by HOMER2 within all AHL15 peaks, with the p-value of motif enrichment, log10 of the p-value of motif enrichment, the percent of AHL15 peaks (targets) containing the motif, and the percent of background sequences containing the motif. B: The number of times each motif (75 in total) is identified by HOMER2 in all AHL15 peaks divided by its number over the entire genome (expressed as %: 100% means the motif is almost only present in AHL15 peaks). C: The frequency with which each motif identified by HOMER2 occurs in all AHL15 peaks, expressed as the Log2 value of the % of shared AHL15 targets containing the specific motif (occurrence of more than 1% gives a positive value, 1% gives a value of 0, and less than 1% results in a negative value). Motifs in figures B and C are arranged by descending target : background value.

Principal component analysis of RNA-seq datasets.

RNA-seq was performed on RNA isolated from 10 day-old wild-type (Col-0) or p35S:AHL15-GR seedlings that mock or DEX treated for 15 minutes. Per replicate (n=3) shoot tissue of five plants was harvested at 8 hours after treatment for RNA isolation and sequencing.

Enrichment of ATAC-seq peaks over 3xFLAG-AHL15 ChIP-seq peaks showing that AHL15 binds DNA regions with reduced chromatin accessibility.

ATAC-seq reads from p35S:AHL15-GR plants treated with DEX or mock for 30 minutes, 1 hour or 8 hours plotted over three classes of AHL15 ChIP-seq peaks presented in Figure 1C: shared (identified in native and overexpression conditions, n = 1948), semi-shared (identified in one native sample and in all overexpression samples, n = 2906), and OX-only (identified in all three overexpression samples, n = 4999). Random genomic coordinates with similar size distribution as the ChIP-seq peaks were used as a negative control (random, n = 15221 regions).

Enrichment of H1, H2A, H2Z.X, H2A.Z, H3K9me1, H3K9me2, and H3K927me1 over 3xFLAG-AHL15 ChIP-seq peaks shows that AHL15 binding sites are characterized by a depletion of epigenetic marks.

A-G: ChIP-seq profiles for the epigenetic marks H2A (A), H2A.X (B), H2A.Z (C), H3K27me1 (D), H3K9me1 (E), H3K9me2 (F), and H1 (G) plotted over three classes of AHL15 ChIP-seq peaks presented in Figure 1C: shared (identified in native and overexpression conditions, n = 1948), semi-shared (identified in one native sample and in all overexpression samples, n = 2906), and OX-only (identified in all three overexpression samples, n = 4999). Random genomic coordinates with similar size distribution as the ChIP-seq peaks were used as a negative control (random, n = 15221 regions).