Tale S1. Total and mapped reads produced for each ChIP-seq sample Sample Total Reads Mapped Reads Col- H3K27me3 rep1 125662 1334323 (85.76%) Col- H3K27me3 rep2 9176437 7986731 (87.4%) atmi1a//c H3K27m3 rep1 188786 7986731 (87.4%) atmi1a//c H3K27m3 rep2 11372 832393 (83.9%) clf28/swn7 H3K27me3 rep1 9831362 519596 (52.85%) clf28/swn7 H3K27me3 rep2 15239663 856182 (55.82%) Col- H2AK121u rep1 17823921 15895954 (89.18%) Col- H2AK121u rep2 1721488 15531618 (9.25%) atmi1a//c H2AK121u rep1 147283 8355733 (56.84%) atmi1a//c H2AK121u rep2 17587127 11734768 (66.72%) Col- H2AK121u rep1 12319965 1756227 (87.31%) Col- H2AK121u rep2 1268263 9791317 (81.13%) clf28/swn7 H2AK121u rep1 1369852 117815 (85.47%) clf28/swn7 H2AK121u rep2 133154 121139 (9.2%) Col- H2AK121u rep1 21853396 15824831 (72.41%) Col- H2AK121u rep2 19184893 16335871 (85.15%) lhp1 H2AK121u rep1 25534779 228281 (79.43%) lhp1 H2AK121u rep2 2734198 1666391 (8.35%) Input 43489511 31275 (71.74%)
Tale S2. Primers used for ChIP qpcr ABCG35 Fw ABCG35 Rv TOUCH4 Fw TOUCH4 Rv FBA2 Fw FBA2 Rv FAD3 Fw FAD3 Rv AIL5 Fw AIL5 Rv FLC Fw FLC Rv GA3OX1 Fw GA3OX1 Rv KNU Fw KNU Rv SPL8 Fw SPL8 Rv DPA4 Fw DPA4 Rv SUP Fw SUP Rv CAL Fw CAL Rv SEP3 Fw SEP3 Rv MGP Fw MGP Rv WUS Fw WUS Rv WOX12 Fw WOX12 Rv ChIP primers (5-3 ) GATGGATTACGATCCAGCTCAT CTGCTTGCTTTGCTTACACTAC ACCACAAACCAATCTAACTCAATAC CAGAGGAGGTGATGATAAGAGAAA TTCCTCCTACGCCGATGA AAACACGTGGTAGAGAAGAGA GGAGACCGGAAGAAAGAAGAAA CTTAACCCAACAGTGCTTAGGA GTTGTCTTCTCTTCCGACCTAC CTAGGAAGCTAGCGGCTATTG CGTACTTATCGCCGGAGGAG CATCGAGAAAGCTCGTCAGC GGTGCCTTCCAAATCTCAAAC GAAGAGACTACCGGTGAGAAAC AGTTCTACACATCTCAAGCTCTC GCAAGGGTAAGGACGAAGAA CCAACAGCAACAACCAGTTTC GGATGGTAGATGGGATCGGA GACACGTCATCTTCGGAGAAA TTTGACTGCTGTTCCAGTAAGA CTTTCACAGTCTTAACACCAAAGAG GAGAGAGAGAGAGAGATATAGAGATGAG GCACTCTTTCACATTACCATCATTAG CCTCTTCAATTCAACCCTACCC CCTATGAGGGTCTTTGGTACAC ACATGTCAGCTTCATTACTTGA CTACCGACCAATCAGCGTATC CGATCTCGTGAAGACCGTTATTA CCACAGCATCAGCATCATCA GACACGTGTAACCACCAGAG TGGATCGTCGTCATCTCAAATC GATTCTCCATACCATTGTTGTTCTC
H2AK121u 152 126 WT Rep2 FPKM 1488 Rep2 Rep1 WT Rep1 FPKM H3K27me3 414 26 WT Rep2 FPKM 6843 Rep1 Rep2 WT Rep1 FPKM Figure S1. Comparison of H2AK121u and H3K27me3 ChIP-seq replicates in WT Araidopsis seedlings at 7 DAG. (a) Scatterplot of pairwise comparison etween H2AK121u ChIP-seq replicates (left panel) and Venn diagram showing overlapping H2AK121u marked genes etween replicates (right panel). () Scatterplot of pairwise comparison etween H3K27me3 ChIP-seq replicates (left panel) and Venn diagram showing overlapping H3K27me3 marked genes etween replicates (right panel).
Figure S2. Identification of H2AK121u peaks. Two different detection methods, MACS and SICER, were coincident in 93.79 % of the identified H2AK121u peaks
H2AK121u Length H3K27me3 18 18 H2AK121u H3K27me3 18 18 Figure S3. Genome wide localization of H2AK121u and H3K27me3 marks in Araidopsis WT seedlings at 7 DAG. (a) Boxplots representing length distriution of H2AK121u and H3K27me3 peaks. H3K27me3 peaks have an average length of 1.7 K, while H2AK121u peaks were sharper with an average length of.6 K. This difference is significant according to a p-value of 2.2x1-16 otained using the Wilcoxon test. () Representative genome rowser views of H2AK121u and H3K27me3 peaks.
Browser view only-h2ak121u marked active genes 2. ChIP qpcr ABCG35; FPKM 15.2 1.6 WT mock WT IP TOUCH4; FPKM 357.13 % INPUT 1.2.8.4 FBA2; FPKM 1817.2 FAD3; FPKM 93.58 only-h3k27me3 marked genes Browser view 2. 1.6 WT mock WT IP ChIP qpcr CAL %INPUT 1.2.8 SUP.4 SEP3 Figure S4. ChIP-qPCR validations of H2AK121u levels in WT at 7 DAG. (a) H2AK121u levels at only-h2ak121u marked active genes. The expression levels of the genes are indicated in FPKM. () H2AK121u levels at only-h3k27me3 marked genes. Structure of the genes and location of the region amplified y qpcr (green line) are shown in the Browser views. For each locus, the amount of immunoprecipitated DNA using H2AK121u antiody (IP) or no antiody (mock) is indicated as % of input. Error ars represent SD etween replicates.
12 only-h2ak121u H2AK121u/H3K27me3 12 only-h2ak121u repressed only-h2ak121u active 1 1 8 8 RPKM 6 4 6 4 2 2-2 TSS 33% 66% TES 2-2 TSS 33% 66% TES 2 RPKM Figure S5. Genomic distriution of H2AK121u marks at different categories of marked genes. (a) Metagene plot showing the distriution of H2AK121u marks at only-h2ak121u and H2AK121u/H3K27me3 marked genes. () Metagene plot showing the distriution of H2AK121u marks at only-h2ak121u active and repressed genes.
Sample Total Numer of Reads Concurrent Pair Alignment Rate WT rep1 57837872 51421443 (88.7%) WT rep2 49749961 4318555 (86.8%) atmi1a//c rep1 54627448 48458316 (88.6%) atmi1a//c rep2 46742567 4695811 (87.%) Figure S6. RNA-seq analysis of WT and atmi1a//c at 7 DAG. (a) Numer of reads and concurrent pair alignment rate per sequencing sample. The numers indicate a high read sequencing quality and the lack of sample contamination. () Scatterplots of pairwise comparison etween RNA-seq replicates of WT (Col-, left panel) and atmi1a//c (right panel).
H2AK121u/H3K27me3 only-h3k27me3 -log(p-value) Figure S7. Different families of Transcription Factors (TFs) are enriched in H2AK121u/H3K27me3 and only-h3k27me3 marked genes. While H2AK121u/H3K27me3 marked genes showed enrichment for different TF families, only-h3k27me3 marked genes were significantly enriched for MADSox transcription factors of the MIKC type, which are mostly involved in the control of flowering time, flower, seed, and fruit development. A p-value of.5 was used as a cut-off to determine significance (indicated as dashed line). X- axis of the plot represents the significance level using -log(p-value). The p-value was computing using Fisher s Exact test.
clf28/swn7 Rep2 FPKM WT Rep1 FPKM clf28/swn 7Rep1 FPKM WT Rep2 FPKM lhp1 Rep2 FPKM WT Rep2 FPKM WT Rep1 FPKM lhp1 Rep1 FPKM Figure S8. Comparison of clf28/swn7 and lhp1 H2AK121u ChIP-seq replicates at 7 DAG. (a) Scatterplots of pairwise comparison etween H2AK121u ChIP-seq replicates of WT (left panel) and clf28/swn7 (right panel). () Scatterplots of pairwise comparison etween H2AK121u ChIP-seq replicates of WT (left panel) and lhp1 (right panel).
2 2 2 2 H2AK121u WT clf28/swn7 WT H2AK121u lhp1-2 -1 TSS 1 2-2 -1 TSS 1 2-2 -1 TSS 1 2-2 -1 TSS 1 2 WT clf28/swn7 1.5.25 1.5.25 @H2AK121u @H3 Figure S9. The gloal levels of H2AK121u marks are increased in clf28/swn7 mutants. (a) ChIP-seq density heatmaps of H2AK121u marks in WT and clf28/swn7 (left panel) and WT and lhp1 (right panel) mutants at genomic regions surrounding the TSS of target genes. () Western lot (WB) analysis showing gloal H2AK121u levels in WT and clf28/swn7 mutants at 7 DAG. A two fold dilution series of extracted chromatin efore immunoprecipitation was proed with the indicated antiodies to ensure quantitative results.
Rep 1 Rep 2 Rep 1 Rep 2 Figure S1. H2AK121u levels at H2AK121u peaks in clf28/swn7 and lhp1 mutants compared to WT after normalization with a modified MAnorm protocol. (a,) Fold-change (M) and average intensity (A) for common H2AK121u peaks efore normalization and for all peaks after normalization in the comparison etween (a) clf28/swn7 and WT and () lhp1 and WT. Common peaks used to uild the rescaling model are highlighted in lack. Peaks with increased levels of H2AK121u were indicated in red and with decreased levels in lue.
1,5,8,4 % INPUT,6,4,2 WT mock WT IP clf28/swn7 mock clf28/swn7 IP % INPUT,3,2,1 WUS WOX12 2,5 1,6 2 1,4 1,2 % INPUT 1,5 1,5 % INPUT 1,8,6,4,2 AIL5 FLC Figure S11. ChIP-qPCR validations of H2AK121u levels in clf28/swn7 at 7 DAG. (a) Genes showing unaltered or increased levels of H2AK121u marks in clf28/swn7. () Genes showing decreased levels of H2AK121u marks in clf28/swn7. Structure of the genes and location of the region amplified y qpcr (green line) are shown in the Browser views. For each locus, the amount of immunoprecipitated DNA using H2AK121u antiody (IP) or no antiody (mock) is indicated as % of input. Error ars represent SD etween replicates.
WT Rep2 FPKM atmi1a//c Rep2 FPKM WT Rep1 FPKM atmi1a//c Rep1 FPKM WT Rep2 FPKM atmi1a//c Rep2 FPKM WT Rep1 FPKM atmi1a//c Rep1 FPKM Figure S12. Comparison of atmi1a//c ChIP-seq replicates at 7 DAG. (a) Scatterplots of pairwise comparison etween H2AK121u ChIP-seq replicates of WT (left panel) and atmi1a//c (right panel). () Scatterplots of pairwise comparison etween H3K27me3 ChIP-seq replicates of WT (left panel) and atmi1a//c (right panel).
WT atmi1a//c WT atmi1a//c c WT atmi1a//c 1.5.25 1.5.25 @H2Au @H3 Figure S13. The gloal levels of H2AK121u marks are reduced in atmi1a//c mutants. (a, ) ChIP-seq density heatmaps of H2AK121u marks in WT and atmi1a//c mutants at genomic regions surrounding the TSS of target genes. (a) reads per kiloase and million mapped reads (RPKM) and () total lirary size (reads per million reads sequenced (RPM)) normalizations produced similar results with a sharper apparent decrease in the case of total lirary size normalization when comparing atmi1a//c to WT. (c) WB analysis showing gloal levels of H2AK121u in atmi1a//c mutants compared to WT. A two fold dilution series of extracted chromatin efore immunoprecipitation was proed with the indicated antiodies.
DPA4 SPL8 GA3OX1 3 2,5 % INPUT 2 1,5 1 WT mock WT IP atmi1a//c mock atmi1a//c IP,5 DPA4 SPL8 GA3OX1 Figure S14. H2AK121u levels at selected genes in WT and atmi1a//c. ChIP qpcr analysis of H2AK121u levels at genes in WT and atmi1a//c mutants. Structure of the genes and location of the region amplified y qpcr (green line) are shown in the Browser views. For each locus, the amount of immunoprecipitated DNA using H2AK121u antiody (IP) or no antiody (mock) is indicated as % of input. Error ars represent SD etween replicates.
Rep 1 Rep 2 Figure S15. H2AK121u levels at H2AK121u peaks in atmi1a//c mutant compared to WT after normalization with a modified MAnorm protocol. Foldchange (M) and average intensity (A) for common H2AK121u peaks efore normalization and for all peaks after normalization in the comparison etween atmi1a//c and WT. Common peaks used to uild the rescaling model are highlighted in lack. Peaks with increased levels of H2AK121u were indicated in red and with decreased levels in lue.
WT atmi1a//c % INPUT 1,8,6,4,2 WUS % INPUT 1,2 1,8,6,4,2 MGP WT mock WT IP atmi1a//c mock atmi1a//c IP % INPUT,45,4,35,3,25,2,15,1,5 WOX12 % INPUT 5 4 3 2 1 KNU c Figure S16. Genes displaying reduced levels of H2AK121u marks were significantly enriched in genes that are transcriptionally upregulated in atmi1a//c mutants. (a) ChIP qpcr analysis of H2AK121u levels at selected genes in WT and atmi1a//c mutants. Structure of the genes and location of the region amplified y qpcr (green line) are shown in the Browser views. For each locus, the amount of immunoprecipitated DNA using H2AK121u antiody (IP) or no antiody (mock) is indicated as % of input. Error ars represent SD etween replicates. () Enrichment score of upregulated genes in the different gene categories according to their level of change in H2AK121u etween WT and atmi1a//c mutants. Genes displaying a reduction of H2AK121u levels equal or igger than 2% were significantly enriched in genes that ecame activated in atmi1a//c mutants, indicating that a reduction of 2% may already have an impact on gene expression. (c) Fraction of genes in the different categories according to their levels of H2AK121u marks in atmi1a//c mutants that were statistically significant (p<.5 computed y MAnorm ased on a Bayesian model [42]).
H2AK121u levels in atmi1a//c 41 2976 Reduced Levels (<8% of WT levels) 1171 No change (8-12% of WT levles) Increased levels (>12% of WT levels) Figure S17. Loss of AtBMI1 function impacts H2AK121u levels. Pie chart showing the total numer of genes with altered levels of H2AK121u in atmi1a//c mutants at 7 DAG. 78% of H2AK121u marked genes displayed reduced levels of these marks in atmi1a//c mutants, ranging from to 8% of WT levels.
WT atmi1a//c WT atmi1a//c c WT atmi1a//c @H3K27me3 1.5.25 1.5.25 @H3 Figure S18. ChIP-seq density heatmaps of H3K27me3 marks in WT and atmi1a//c mutants at genomic regions surrounding the TSS of target genes. (a) reads per kiloase and million mapped reads (RPKM) and () total lirary size (reads per million reads sequenced (RPM)) normalizations produced similar results with a sharper apparent decrease in the case of total lirary size normalization when comparing atmi1a//c to WT. (c) WB analysis showing gloal levels of H3K27me3 in atmi1a//c mutants compared to WT. A two fold dilution series of extracted chromatin efore immunoprecipitation was proed with the indicated antiodies.
Rep 1 Rep 2 Figure S19. H3K27me3 levels at H3K27me3 peaks in atmi1a//c mutant compared to WT after normalization with a modified MAnorm protocol. Foldchange (M) and average intensity (A) for common H3K27me3 peaks efore normalization and for all peaks after normalization in the comparison etween atmi1a//c and WT. Common peaks used to uild the rescaling model are highlighted in lack. Peaks with increased levels of H3K27me3 were indicated in red and with decreased levels in lue.
H3K27me3 levels in atmi1a//c 1255 2575 313 Reduced Levels (<8% of WT levels) No change (8-12% of WT levles) Increased levels (>12% of WT levles) Figure 2. Loss of AtBMI1 function impacts H3K27me3 levels. Pie chart showing the total numer of genes with altered levels of H3K27me3 in atmi1a//c mutants at 7 DAG. 44% of H2AK121u marked genes displayed reduced levels of these marks in atmi1a//c mutants, ranging from to 8% of WT levels, ut there were also a 18.3% of the genes that gained H3K27me3 marks.