|
|
Blood, Vol. 112, Issue 9, 3889-3899, November 1, 2008
The role of the polycomb complex in silencing  -globin gene expression in nonerythroid cells
Blood Garrick et al.
112: 3889
Supplemental materials for: Garrick et al
Files in this Data Supplement:
- Document 1. Supplemental materials and methods (PDF, 81.3 KB)
- Figure S1 (JPG, 78.7 KB)
-
(A) Quantitative ChIP experiments in EBV-lymphoblastoid cells using antibodies against H3K9me3 or H3K27me3. Immunoprecipitated material was analysed by real time PCR at the amplicons indicated surrounding the α-globin locus (shown in Fig. 1) or an amplicon within exon 1 of the adult β-globin gene (HBB ex1) or a control amplicon within the 18S rRNA gene. Grey dashed lines demark probes around the adult HBA genes themselves. (B) Quantitative ChIP experiments were carried out in i) primary neutrophils, ii) HeLa cells or iii) K562 cells, using an antibody against H3K27me3 and analysed by real time PCR at the amplicons indicated. H3K27me3 was enriched at the α-globin genes in neutrophils and HeLa cells (non-erythroid) but was absent in K562 cells (erythroid). It is interesting that H3K27me3 (and SUZ12 and EZH2, see Fig.S3) is not present at the active α-globin genes in K562 cells but as noted by others,8 PcG is specifically recruited to the abnormally silenced β-globin gene in these cells. The profile in HeLa cells was somewhat atypical, in that greatest enrichment for H3K27me3 was observed at the 3′IZHVR probe, although significant enrichment was still observed at the adult α-globin genes themselves (probes 5′HBA, HBA ex1-2, HBA ex3). H3K27me3 was also found to be significantly enriched above background at the control 18S amplicon in primary neutrophils. For both A and B, data indicate the percentage of input material precipitated at each amplicon. Shown are the mean values obtained from replicate experiments with error bars indicating the range of values.
- Figure S2. Chromatin was immunoprecipitated from primary T lymphocytes or proerythroblasts using an antibody against unmodified histone H3 and was applied to a custom tiled microarray as described in Fig. 2 (JPG, 58 KB)
-
Shown is a 250 kb genomic segment containing the α-globin locus (delineated by grey dashed lines) with the fold enrichments at each probe plotted against chromosomal position. The locus is annotated as described in Fig. 1.
- Figure S3 (JPG, 66.3 KB)
-
(A) ChIP experiments were carried out on EBV-lymphoblastoid or K562 cells using antibodies against SUZ12 or EZH2 and analysed by real time PCR at the amplicons indicated surrounding the α-globin locus (shown in Fig. 1) or a control amplicon within the 18S rRNA gene. Data indicate the percentage of input material precipitated at each amplicon. Shown are the mean values obtained from replicate experiments with error bars indicating the range of values. For both SUZ12 and EZH2, the distribution in non-expressing EBV-lymphoblasts was very similar to that observed for the H3K27me3 modification (Fig. S1A), enrichment being highest in regions within or adjacent to the large CpG islands (H and I) associated with the adult α-globin genes (marked by vertical lines). (B) ChIP was carried out in primary T lymphocytes using an antibody against SUZ12, and immunoprecipitated material was analysed by real time PCR at amplicons around the adult α-globin gene (5′HBA and HBA ex1-2), around the adult β-globin gene (HBB ex1, 3′HBB) or within HS5 of the βLCR (βLCR HS5) or an amplicon within the 18S ribosomal RNA gene. Data indicate the percentage of input material precipitated at each amplicon.
- Figure S4 (JPG, 37.3 KB)
-
(A) Cells treated with siRNAs as shown were stained with H3K27me3 and analysed by quantitative confocal microscopy. A minimum of 150 cells was analysed in each population. Raw signal in each cell was normalised to the mean signal in the control siRNA-treated population. Data from each population after normalisation are presented as box and whisker plots. Outliers (points outside the 25th or 75th percentile by more than 1.5 times the interquartile range) are plotted separately as black dots. The mean of each population is shown as a white dot. The fluorescence signal observed in the two populations was statistically different (p<0.001; two-tailed student’s t-test) at each time point.
- Figure S5. The α-globin CpG island is unmethylated in non-expressing tissues (JPG, 63.6 KB)
-
(A) Shown is a schematic of the 1.51 kb PstI (P) fragment containing the α-globin CpG island (grey bar) and genes (block arrow; note this assay does not distinguish the duplicated HBA2 and HBA1 genes). Below is shown sites within this fragment for the methylation-sensitive enzymes EagI (Ea) or HpaII (H) (which are also recognized by its methylation-insensitive isoschizomer MspI (M)). (B) Southern blot analysis of DNA from the tissues indicated, after digestion with either PstI alone or with both PstI and EagI as shown. The entire 1.5 kb PstI fragment was used as a probe. Black arrows indicate bands resulting from digestion with PstI alone and grey arrows indicate bands resulting from further digestion with the methylation-sensitive EagI. Complete digestion by EagI was observed in all tissues. (C) Southern blot analysis of DNA from the tissues indicated, after digestion with either PstI alone or with both PstI and HpaII as shown. The entire 1.5 kb PstI fragment was used as a probe. Black arrows indicate bands resulting from digestion with PstI alone and grey arrows indicate bands resulting from further digestion with the methylation-sensitive HpaII. The pattern expected for complete digestion by HpaII (as observed in the last lane where DNA was digested with PstI and the methylation-insensitive isoschizomer MspI) was observed in all tissues. Previous bisulphite sequencing analysis also revealed that 98% (46 of 47) of tested CpG dinucleotides within the α-globin promoter (nucleotides 162529-162887 of chromosome 16) were unmethyated in DNA from normal peripheral blood (which is derived from non-erythroid cells).4
- Figure S6. EBV-lymphoblastoid cells were either mock-treated, or treated with 0.2 µM TSA for 4 or 8 h as indicated and used in ChIP experiments (JPG, 72.4 KB)
-
Immunoprecipitated material was analysed by real time PCR at probes throughout the adult α-globin genes (Fig. 1) and at an amplicon in the 18S ribosomal RNA gene, (representing background enrichment). Data are presented as fold change relative to enrichment observed in the mock-treated culture. Data show the mean values obtained from replicate timecourses with error bars indicating the range of values. (A) ChIP experiments were carried out using an antibody against RNA pol II. (B) ChIP experiments were carried out using an antibody against SUZ12. At probes 5′HBA, HBA prom and HBA ex1/2 (where SUZ12 is usually recruited in EBV-lympoblasts) (Fig.S3), a statistically significant decrease in SUZ12 signal (relative to the 0 h sample of each time course) was observed after 8 hours with TSA (p < 0.05, student’s t-test). A similar depletion in SUZ12 was observed at HBA ex3, but the change was not significant (p=0.084). At 18S, where SUZ12 enrichment is only background (Fig.S3), no significant change was observed (p=0.495). (C) ChIP experiments were carried out using antibodies against H3K27me3.
- Figure S7. Promoter CpG islands which were either bound (n = 72) or unbound (n = 93) by SUZ12 in human ES cells were randomly selected and scored for size (bp) and total number of CpG dinucleotides of the annotated island (annotations from the UCSC genome browser http://genome.ucsc.edu) (JPG, 36.1 KB)
-
Data are presented as box and whisker plots. Outliers (points outside the 25th or 75th percentile by more than 1.5 times the interquartile range) are plotted separately as black dots. The mean of each population is shown as a white dot. Statistical differences were determined using a two-tailed student’s t-test.
|
|
