Search

Epigenetic profile of two classes of CTCF binding regions in BORIS-positive cells (K562). a Heatmaps demonstrate the association of CTCF&BORIS bound regions with active promoters (H3K4me3 and RNAPII, ENCODE) and enhancers (H3K27ac, ENCODE) mapped in K562 cells. The tag density was subjected to k-means ranked clustering with five clusters expected. b Genome browser view of two examples of Super Enhancers mapped in K562 cells. H3K27ac and p300 tracks were adopted from ENCODE. c Scatter plots show the overlapping of CTCF-only and CTCF&BORIS bound regions (left panel) and BORIS-only and CTCF&BORIS bound regions (right panel) with ChIP-seq data available for K562 cells by ENCODE. The dots represent the percentage of CTCF-only, BORIS-only (x-axis), and CTCF&BORIS (y-axis) bound regions overlapping with each factor mapped by ChIP-seq in K562 cells. The red dots are labeled with the name of the factor mapped by ChIP-seq. d The percentage of BORIS-only, CTCF&BORIS and CTCF-only bound regions (y-axis) overlapping with each factor (x-axis). e Heatmap demonstrates the enrichment of RNAPII, CAGEs, H3K4me3, H2AZ, H3K27ac, ZNF143, and SMC3 ChIP-seq data (ENCODE data) at CTCF&BORIS bound regions in contrast to CTCF-only bound regions. The tag density of ChIP-seq data was collected within a 10-kb window around the summit of CTCF peaks mapped in K562 cells. The collected data were subjected to k-means clustering using linear normalization. f Average tag density (tags/10 million) of multiple factors mapped by ChIP-seq in K562 cells (ENCODE data) across BORIS-only (blue), CTCF-only (red) and CTCF&BORIS (purple) bound regions mapped in K562 cells. The names of factors used in ChIP-seq are labeled on the top of the each plot. g Histogram shows the genomic distribution of CTCF and BORIS occupancy at promoters plus the 5’ UTR (3 kb up- and downstream of TSSs), gene bodies, and intergenic regions. (PPTX 680 kb)

Comparison of DNaseI digital genomic footprinting (DGF) and phastConss46 conservation score at 2xCTSes and 1xCTSes. a Heatmaps show DNaseI cleavage density mapped in K562 and NHDF cells at 2xCTSes. 2xCTSes (CTCF&BORIS bound regions in K562 cells) were sorted into 12 different groups (1000 genomic regions in each group) depending on the distance between two CTCF motifs. The distance is shown by double arrows and the number of base pairs between two CTCF motifs. The same genomic coordinates were used to plot DNaseI cleavage density mapped in K562 and NHDF cells and for the analysis of average phastCons46 conservation score (shown in the middle). The schematic presentation of 2xCTSes with two CTCF motifs (red boxes) under one ChIP-seq peak and the proposed occupancy of 2xCTSes by CTCF (red circle) and BORIS (blue circle) in K562 and NHDF cells are shown at the top of panels. The two CTCF motifs were not sorted based on plus or minus strand. The tag density data were collected within a 400-bp window around the first (left) 20-bp CTCF core motifs (0) under a single CTCF ChIP-seq peak. b Heatmap shows DNaseI cleavage density mapped in K562 cells at 1xCTSes with CTCF motif on minus (upper panel) and plus (lower panel) strands. The average phastCons conservation score at the same regions as in DNAseI panels is shown below. c, d Heatmaps show DNaseI cleavage density mapped in K562 cells at 2xCTSes. 2xCTSes were separated into two groups with both CTCF motifs located on the plus strand (c) or minus strand (d). The distance between two CTCF motifs varies from 10 (bottom) to 100 bp (top). (PPTX 1337 kb)

CTCF and BORIS bind to the same genomic regions simultaneously. a The distribution of peak intensity for CTCF (AllCTCF) and BORIS (AllBORIS) bound regions in the K562 cell line. The peak intensity (tag density) is shown for CTCF peaks that coincide with BORIS peaks (CTCF&BORIS), for BORIS peaks that coincide with CTCF peaks (BORIS&CTCF), and for CTCF-only and BORIS-only bound regions. Data are shown in logarithmic scale; ***p < 0.01. b Genome browser view of CTCF and BORIS occupancy in K562 and Delta47 cells in combination with ChIP-Re-ChIP-seq data for K562 and Delta47 cells. The gray frames highlight CTCF-only, CTCF&BORIS, and BORIS-only bound regions. Only CTCF&BORIS bound regions were detected in ChIP-Re-ChIP data. (PPTX 121 kb)

Epigenetic profile of two classes of CTCF binding regions in BORIS-negative cells (GM12878). a Average tag density (tags/10 million) of CTCF, RNAPII, DNaseI digestion, H3K4me3, H3K27ac, and RAD21 (mapped by ChIP-seq in GM12878 cells, ENCODE data) at 2xCTSes (blue) and 1xCTSes (red). The data were normalized to the number of mapped reads and binding regions. b Upper panel: scatter plot shows overlapping of 2xCTSes (y-axis) and 1xCTSes (x-axis) with multiple ENCODE data available for GM12878. Similar to CTCF&BORIS and CTCF-only bound regions in K562 (Fig. S7c, left panel in Additional file 7), no correlation (R2 = 0.41) was found between 2xCTSes and 1xCTSes with respect to co-occupancy with the majority of transcription factors, histone modifications, and chromatin remodeling factors. Middle panel: scatter plot shows an overlapping of 1xCTSes mapped in K562 (y-axis) and in GM12878 cells (x-axis) with ENCODE data for K562 and GM12878 cells, respectively. 1xCTSes of GM12878 and K562 cells were correlated with each other (R2 = 0.98) with respect to co-occupancy with the majority of ENCODE data common for the two cell lines. Lower panel: scatter plot shows the overlapping of 2xCTSes in K562 (y-axis) and in GM12878 (x-axis) with ENCODE data for K562 and GM12878 cells (R2 = 0.93), respectively. (PPTX 220 kb)

CTCF&BORIS bound regions enclose at least two CTCF binding sites. a Genome browser view of CTCF and BORIS occupancy at seven CTCF&BORIS (blue bracket) and seven CTCF-only (red bracket) bound regions in K562 and Delta47 cells. Top panel: at the top of the genome browser view, the names of sequences are highlighted by black or red colors, which represent at least two or one CTCF motifs under the peaks, respectively. Lower panel: EMSA with the corresponding CTCF&BORIS and CTCF-only binding regions. The ~200-bp 32P-labeled probes were incubated with either in vitro translated luciferase (−) or with 11 ZF CTCF domain (11ZFs). The slower (shown by arrow with two red dots) and faster (arrow with one red dot) migrating shifted bands correspond to double and single occupancy, respectively. Of note, the CTCF&BORIS bound regions enclosing one CTCF motif (gene/chromosome name in red) are showed double occupancy, similar to CTCF&BORIS bound regions enclosing two CTCF motifs (black), while all CTCF-only bound regions showed single occupancy. b Three CTCF&BORIS bound regions (TP53, IRF2BP1 and chr17 intergenic regions), each enclosing two CTCF motifs, were split into two fragments and subjected to EMSA. The separation of CTCF motifs resulted in single occupancy of each fragment with one CTCF motif, confirming the presence of two CTCF binding sites in each sequence. c 200-bp 32P-labeled probes representing KDM3B, FOXA3, and TP53 promoters were incubated with nuclear extracts from either K562 or NHDF cells. All lanes contained the indicated nuclear extract, except the first lane for K562, where only free probe is present (F). Nuclear extracts and probe were also incubated with either control mouse IgG (−) or antibodies against CTCF or BORIS. The red and blue arrows point to the super-shift bands corresponding to CTCF–DNA and BORIS–DNA complexes, respectively. (PPTX 215 kb)

BORIS is involved in transcriptional program of cancers. a The sequence recognized by ZFN and cleaved by Fok I is shown by black and red letters, respectively. b Surveyor assay (CEL-I) for ZFN-induced mutations in the BORIS gene. The proportions of wild-type and mutant alleles are shown as a table at the bottom of the gel. c Surveyor assay shows ten single-cell clones with the mutated BORIS locus. d Western blot shows the level of BORIS protein in wild type (wt) and ten mutant clones from panel (c). e K562 cells transfected with ZFN produced at least ten times less colonies in soft agar compared with control. f Western blot shows the level of BORIS protein in K562 before (0) and after phorbol 12-myristate 13-acetate (PMA) treatment. Wright–Giemsa staining analysis of K562 transfected with either control vector or ZFN in comparison with K562 treated with PMA for 3 days. Both transfection of K562 with ZFN and treatment with PMA showed morphologic changes related to megakaryocytic differentiation. h Upper panel: ChIP-seq and RNA-seq tracks show CTCF and BORIS occupancy at the WISP2 gene and dowregulation of WISP2 expression upon BORIS induction in MCF7 cells. Lower panel: RNA-seq RPKM values for the WISP2 gene expression. i Downregulation of inflammatory response pathway in response to stable BORIS expression in MCF7 cells. k ChIP-seq tracks show CTCF and BORIS occupancy at GAL3ST1, FOXA3, and PRAME loci in NHDF, OVCAR8, Delta47, and K562 cells. The tracks are labeled with the molecules against which antibodies were directed and cell lines used in ChIP-seq. RNA-Seq tracks demonstrate gene expression in BORIS-positive cells (K562) in contrast to BORIS-negative cells (NHEK and NHDF). l Quantitative PCR analysis of GAL3ST1, FOXA3, and PRAME gene expression in BORIS-positive cells (Testes, K562, OVCAR8, Delta47) and BORIS-negative cells (GM12878, NHDF, NHEK). (PPTX 1328 kb)

CTCF&BORIS bound regions enclose at least two CTCF binding motifs. a The motifs identified by MEME for CTCF-only, CTCF&BORIS and BORIS-only bound regions in K562 cells. Additionally, the presence of the CTCF motif was confirmed in 70 % of BORIS-only bound regions. b, c The alignment of human and mouse sequences representing five CTCF&BORIS bound regions (b) and two CTCF-only bound regions (c). The summits of CTCF and BORIS peaks are highlighted by red and blue nucleotides, respectively. The CTCF motif is shown in bold italics and underlined. The CTCF motif is indicated with black and blue colors, depending on sense and antisense orientation, respectively. d, e Number of CTCF motifs in the genomic regions, encircling 100 bp upstream and downstream of the summit of CTCF peaks at CTCF&BORIS bound regions (d) and from the summit of CTCF-only peaks (e). The histogram shows the percentage of CTCF binding regions (y-axis) containing 0 to 7 CTCF motifs (x-axis). f The distribution of distances between two CTCF motifs at 1-bp resolution (x-axis) at CTCF&BORIS bound regions (y-axis). The CTCF&BORIS bound regions that have only two CTCF motifs under the peaks were selected for the analysis. (PPTX 3766 kb)

EMSA demonstrates the presence of two CTCF binding sites inside BORIS bound regions (CTCF&BORIS and BORIS-only). a Top panels: ChIP-seq tracks show CTCF and BORIS occupancies mapped in K562, Delta47 and NHDF cells. From left to right, TP53 promoter represents CTCF&BORIS bound region; seven CTCF sites residing in the H19-IGF2 imprinting control region are examples of CTCF-only bound regions; the BMI promoter represents a BORIS-only bound region; the PROB1 exon illustrates that a BORIS-only bound region mapped in K562 cells is a CTCF&BORIS bound region in Delta47 cells. Bottom panels: binding of in vitro translated 11 ZF domain of CTCF (11ZF), full-length CTCF (CTCF) and full-length BORIS (BORIS) proteins to the TP53 promoter, one of seven CTCF sites in the H19-IGF2 imprinting control region (fourth CTCF binding site), BMI promoter, and PROB1 exon. In vitro translated luciferase protein was used as a negative control (Luc). The DNA complexes with 11 ZF domain, CTCF, and BORIS proteins are indicated by black, red and blue arrows, respectively. The single and double shifts with 11 ZF domain are shown by single and double black arrows, respectively. The genomic coordinates (hg19) of DNA probes are indicated at the bottom of the gels. b, c Five examples of CTCF-only (b) and BORIS-only (c) bound regions. Upper panel: ChIP-seq tracks show CTCF and BORIS occupancies at ten genomic regions mapped in K562, OVCAR8, and Delta47 cells. Lower panel: EMSA with the genomic sequences shown on the upper panel. In vitro translated 11ZF domain of CTCF (11ZF), full-length CTCF (CTCF) and BORIS (BORIS) proteins were incubated with the corresponding CTCF-only and BORIS-only bound regions. The single and double shifts with 11 ZF domain are shown by single and double black arrows, respectively. (PPTX 353 kb)

The primers used in the study for amplification of EMSA probes and for quantitative PCR. (DOCX 17Â kb)