Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2024 Nov 15;38(22):e70181. doi: 10.1096/fj.202401526RR

TAD‐dependent sub‐TAD is required for enhancer–promoter interaction enabling the β‐globin transcription

Dasoul Lee 1, Jin Kang 1, AeRi Kim 1,✉
PMCID: PMC11698014  PMID: 39545685

Abstract

Topologically associating domains (TADs) are chromatin domains in the eukaryotic genome. TADs often comprise several sub‐TADs. The boundaries of TADs and sub‐TADs are enriched in CTCF, an architectural protein. Deletion of CTCF‐binding motifs at one boundary disrupts the domains, often resulting in a transcriptional decrease in genes inside the domains. However, it is not clear how TAD and sub‐TAD affect each other in the domain formation. Unaffected gene transcription was observed in the β‐globin locus when one boundary of TAD or sub‐TAD was destroyed. Here, we disrupted β‐globin TAD and sub‐TAD by deleting CTCF motifs at both boundaries in MEL/ch11 cells. Disruption of TAD impaired sub‐TAD, but sub‐TAD disruption did not affect TAD. Both TAD and sub‐TAD disruption compromised the β‐globin transcription, accompanied by the loss of enhancer–promoter interactions. However, histone H3 occupancy and H3K27ac were largely maintained across the β‐globin locus. Genome‐wide analysis showed that putative enhancer–promoter interactions and gene transcription were decreased by the disruption of CTCF‐mediated topological domains in neural progenitor cells. Collectively, our results indicate that there is unequal relationship between TAD and sub‐TAD formation. TAD is likely not sufficient for gene transcription, and, therefore, sub‐TAD appears to be required. TAD‐dependently formed sub‐TADs are considered to provide chromatin environments for enhancer–promoter interactions enabling gene transcription.

Keywords: enhancer, sub‐TAD, TAD, transcription, β‐globin locus

To explore how TADs and sub‐TADs affect each other in their domain formation and contribute to gene transcription, the β‐globin TAD and sub‐TAD were disrupted by deleting CTCF motifs at both boundaries in MEL/ch11 cells. Disruption of TAD impaired sub‐TAD, but sub‐TAD disruption did not affect TAD. Both TAD and sub‐TAD disruption compromised the β‐globin transcription, accompanied by the disruption of enhancer–promoter interactions. However, histone H3K27ac was largely maintained across the β‐globin locus.

graphic file with name FSB2-38-e70181-g005.jpg

Abbreviations

3C

chromosome conformation capture

APA

aggregate peak analysis

ChIP

chromatin immunoprecipitation

CTCF

CCCTC‐binding factor

E‐P

enhancer–promoter

HMBA

hexamethylene bisacetamide

HSs

hypersensitive sites

LCR

locus control region

MEL/ch11 cells

murine erythroleukemia cells containing a human chromosome 11

NPCs

neural progenitor cells

sgRNAs

single‐guide RNAs

sumCC

sum of contact counts

TADs

topologically associating domains

TPM

transcripts per million

1. INTRODUCTION

The eukaryotic genome is systematically organized into the three‐dimensional space of the nucleus. Genome‐wide studies based on chromosome conformation capture (3C) assay, which analyze the proximity of DNA in a chromatin context by measuring ligation frequency between DNA fragments, present a series of topological domains across the mammalian genome. 1 , 2 These domains are referred to as topologically associating domains (TADs) and span from several hundred kilobases to one megabase. The boundaries of TADs are enriched in the CCCTC‐binding factor (CTCF) protein, which directly binds to DNA and acts as an insulating factor. 3 Depletion of the CTCF protein disrupts the interaction between TAD boundaries and results in loss of insulation for neighboring TADs. 4 , 5 , 6 , 7 Deletion of CTCF motifs at one boundary of TADs disrupts the organization of the chromatin domain, resulting in the mis‐regulation of gene transcription. 8 , 9 , 10 These findings suggest that TADs mediated by the CTCF protein provide a chromatin context for the transcriptional regulation of genes.

TADs comprise several sub‐TADs, which are small sub‐domains formed through chromatin interactions. 11 Although TADs are conserved across cell types and species, 1 , 2 sub‐TADs are tissue‐specific and dynamically organized during cell differentiation. 12 , 13 CTCF plays an architectural role in the formation of sub‐TADs by anchoring chromatin loops extruded by the cohesin complex. 14 , 15 Deletion of CTCF motifs at one boundary of sub‐TADs weakens chromatin interactions inside sub‐TADs and disturbs the insulation between sub‐TADs. 16 , 17 This leads to transcriptional induction of proto‐oncogenes in nonmalignant cells, 17 transcriptional suppression of oncogenes in tumor cells, 18 transcriptional delay of development‐regulatory genes, 19 and failure of gene rearrangement in immune cells. 16 Chromatin interactions between enhancers and promoters appear to be affected by the loss of sub‐TAD. 20 However, the disruption of TAD or sub‐TAD did not disturb gene transcription in all instances. Unaffected gene transcription was observed in several loci including the Runx1 and Csn loci when the domains containing these loci were disrupted by the loss of their boundaries at one side. 12 , 13 , 21 , 22 , 23

Human β‐like globin genes are transcriptionally activated by the locus control region (LCR). 24 The LCR is an erythroid‐specific enhancer consisting of 4 DNase I hypersensitive sites (HSs) and is located upstream of β‐like globin genes. The β‐globin locus including the LCR is present in a sub‐TAD that is flanked by a set of convergent CTCF motifs (Figure 1A). 25 , 26 This sub‐TAD is formed tissue‐specifically in human fetal and adult erythroblasts and erythroid K562 cells, and neighbors on another sub‐TADs in both sides. 25 , 27 Five CTCF sites are present at the boundaries of TAD and sub‐TAD containing the β‐globin locus. In our previous study, concomitant destruction of CTCF motifs at the boundaries of TAD and sub‐TAD disrupted chromatin interaction between the boundaries, resulting in transcriptional inactivation of the β‐globin gene. 28 However, the destruction of a CTCF motif at one boundary of TAD or sub‐TAD did not affect β‐globin transcription even with the disrupted interaction of the boundary. These results raised the question of how TADs and sub‐TADs affect gene transcription.

FIGURE 1.

FIGURE 1

Open in a new tab

Deletion of CTCF‐binding motifs at both boundaries of the human β‐globin TAD and sub‐TAD using the CRISPR/Cas9 system in MEL/ch11 cells. (A) Hi‐C map and CTCF ChIP‐seq track show the human β‐globin TAD comprising three sub‐TADs with CTCF binding at their boundaries in human erythroblast. In the β‐globin locus, the β‐like globin genes and LCR are represented as black and yellow rectangles, respectively. The CTCF motifs are represented as triangles with motif orientation in the β‐globin TAD. Blue and red triangles indicate the CTCF motifs at the boundaries of the TAD and sub‐TADs, respectively. (B) Edited sequences at TAD boundaries (ΔTAD; ΔC3,7) and sub‐TAD boundaries (Δsub‐TAD; ΔC4,5) are listed in tables for two clonal cells (#1 and #2). Bold‐faced and dashed bases represent intact CTCF motifs and deleted sequences, respectively. Occupancy of CTCF (C) and SMC3 (D) was analyzed at CTCF motif‐deleted sites in the control (Con), ΔTAD (ΔT), and Δsub‐TAD (ΔsT) cells using ChIP‐qPCR. Normal rabbit IgG (Ig) and mouse Actin genes served as negative controls. (E) At the top, vertical gray bars indicate Hind III cut sites in the β‐globin TAD and horizonal arrows mark the location of PCR primers for the 3C assay. Black and gray arrows locate in DNA fragments containing CTCF sites and lacking them, respectively. Relative ligation frequencies between DNA fragments were analyzed using the 3C assay. Fragments containing C3 and C5 served as anchors indicating TAD and sub‐TAD boundaries, respectively. DNA fragments lacking CTCF sites were represented with asterisks on the X‐axis. Data are represented as the means ± SEM of two to four ChIP and five 3C biological repeats. p‐Values were calculated using the two‐tailed Student's t‐test. *p < .05, **p < .01.

In the present study, to explore how TADs and sub‐TADs affect each other in their domain formation and contribute to gene transcription inside the domains, we generated two types of clonal cells by deleting CTCF motifs at both boundaries of the β‐globin TAD and sub‐TAD containing the β‐globin locus, and comparatively analyzed the CTCF motif‐deleted cells. Our analysis showed that the β‐globin TAD and sub‐TAD affect each other unequally and TAD‐dependently formed sub‐TAD is required for the transcription of β‐globin gene.

2. MATERIALS AND METHODS

2.1. Cell culture

Murine erythroleukemia cells containing a human chromosome 11 (MEL/ch11 cells) 29 were cultured in DMEM (high Glucose, pyruvate; Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (P/S) (Gibco). 293FT cells derived from human embryonic kidney were grown in DMEM supplemented with 10% FBS, 1% P/S, and 1% non‐essential amino acids. To induce transcription of the β‐like globin genes, MEL/ch11 cells (1.5 × 105 cells/mL) were treated with 5 mM of hexamethylene bisacetamide (HMBA) for 72 h.

2.2. Deletion of CTCF motifs using the CRISPR/Cas9 system

The CRISPR/Cas9 technique was used to delete CTCF‐binding motifs, as described previously. 30 Single‐guide RNAs (sgRNAs) were designed for the four CTCF motifs at the boundaries of the human β‐globin TAD (C3 and C7) and sub‐TAD (C4 and C5) using a software tool (http://crispr.mit.edu). The oligonucleotide sequences of the sgRNAs are listed in Table S1. Oligonucleotides for sgRNA were phosphorylated, annealed, and cloned into the lentiCRISPRv2 (Addgene, #52961) or pLH‐spsgRNA2 vectors (#64114). The oligo‐inserted vectors were transfected into 293FT cells (1 × 106) with a ViraPower packaging mix using Lipofectamine 2000 (Invitrogen) to produce lentiviruses. MEL/ch11 cells (1 × 106) were transduced using 2.5–3.0 mL of 293FT cell supernatant with 6 μg/mL polybrene for 24 h, and then selected with 2 μg/mL puromycin or 500 μg/mL hygromycin. Clonal cells were seeded into 96‐well plates. The deleted sequences of the CTCF motifs were analyzed using Sanger sequencing.

2.3. Chromatin immunoprecipitation (ChIP)

ChIP assays were performed as described previously. 31 Briefly, MEL/ch11 cells (2 × 107) were cross‐linked with 1% formaldehyde. Chromatin was fragmented using 100 U Micrococcal nuclease (MNase) into mainly mono‐ and di‐nucleosomes, reacted with 2–2.5 μg of antibodies for 3 h at 4°C, and pulled down using protein A or G agarose beads (Millipore). Immunoprecipitated DNA was purified using phenol extraction and ethanol precipitation and quantitatively analyzed by comparison with the input DNA using qPCR. Mouse Actin was used as the internal control for qPCR. The antibodies used in this study were CTCF (07‐729), H3K27me3 (07‐449), normal rabbit IgG (12‐370), and normal goat IgG (NI02‐100UG) from Millipore; SMC3 (ab9263), H3 (ab1791), and H3K27ac (ab4729) from Abcam; and GATA1 (sc‐1234), TAL1 (sc‐12984), and LDB1 (sc‐11198) from Santa Cruz Biotechnology. The primers used for the ChIP assay are listed in Table S2.

2.4. Reverse transcription‐quantitative PCR (RT‐qPCR)

Total RNA was extracted from MEL/ch11 (2 × 106) cells using QIAzol (Qiagen) and reverse‐transcribed into cDNA using a GoScript kit (Promega). The cDNA was analyzed using qPCR and the comparative Ct method. The amount of cDNA for the β‐like globin gene was normalized to that of cDNA for the mouse Actin gene. The primers used for RT‐qPCR are listed in Table S3.

2.5. Chromosome conformation capture (3C)

The 3C assay was performed as described previously. 28 MEL/ch11 cells (2 × 106) were cross‐linked, lysed, and digested with 600 U Hind III at 37°C with shaking at 900 rpm overnight. The digested chromatin was ligated with 400 U T4 DNA ligase at 16°C with shaking at 600 rpm for 4 h. The ligated DNA was purified using phenol extraction and ethanol precipitation. To correct primer efficiency in qPCR, control DNA was prepared by mixing MEL/ch11 genomic DNA and an equal quantity of bacterial artificial chromosome (BAC) clone DNA covering the β‐globin TAD. Relative ligation frequency was calculated by comparing the quantity of ligated fragments with that of control DNA and then normalizing to the ligation frequency between two fragments in the ILK locus on human chromosome 11. The primers used for 3C are listed in Table S4.

2.6. ChIP‐seq library preparation

ChIPed DNA was processed using the NEBNext Ultra II DNA Library Prep Kit (NEB), according to the manufacturer's instructions (NEB). Briefly, ChIPed DNA (10 ng for H3K27ac and 1 ng for H3K27me3) was repaired at the ends by NEBNext Ultra II End Prep enzyme mix, ligated with NEBNext adaptors, selected at 175–200 bp in size using NEBNext Sample Purification Beads (E7103S), and amplified using NEBNext Multiplex Oligonucleotides from Illumina (96 Unique Dual Index, E6440S). Following purification, the final libraries were quantified and sequenced in 100 bp paired‐end reads using the Illumina NovaSeq 6000 system.

2.7. ChIP‐seq data analysis

ChIP‐seq raw reads were qualified by removing input reads with poor‐quality values (quality score, 20; minimum read length, 36 bp) using the Trimmomatic tool. 32 Qualified reads were aligned to the human chromosome 11 reference genome in the BAM file format using Bowtie2. 33 Aligned reads were filtered using a minimum MAPQ score of 20 and sorted based on chromosomal coordinates. Potential PCR duplicates were eliminated using RmDup tool. 34 The filtered BAM files were converted into BigWig files using the S3V2 normalization method. 35 H3K27ac and H3K27me3 around the β‐globin TAD were visualized using an Integrated Genome Browser (IGB). 36

2.8. Public data analysis

ATAC‐seq and CTCF, H3K4me1, and H3K27ac ChIP‐seq data in mouse neural progenitor cells (NPCs) were analyzed in similar ways with MEL/ch11 ChIP‐seq data. Qualified ATAC‐seq and ChIP‐seq reads were aligned to the mm10 genome using Bowtie2. Aligned reads were filtered with a MAPQ score of 20 and sorted based on chromosomal coordinates. Potential PCR duplicates were removed from the ChIP‐seq BAM files using RmDup tool. To visualize the H3K27ac state, H3K27ac BAM files were converted to BigWig files using deeptools bamCoverage 37 with normalization to the effective mm10 genome size. RNA‐seq reads in mouse NPCs were trimmed and filtered using the Trimmomatic tool. Qualified reads were aligned to the mm10 genome using STAR 38 with an exon annotation GTF file obtained from the UCSC database. The aligned reads were counted for all genes using featureCounts. 39 Gene transcription was quantified by calculating transcripts per million (TPM) and was visualized using BigWig files generated by deep tools bamCoverage. Putative enhancers were identified by overlapping control ATAC‐seq peaks (summits ±1 Kb) with control H3K4me1 and H3K27ac peaks. TSSs (TSS + 1 bp forward gene body) and CTCF peaks were removed from the putative enhancers (Data S1 and S2).

Mouse NPC H3K4me3 PLAC‐seq raw reads were processed using HiC‐Pro tool. 40 Paired‐end reads were separately aligned to the mm10 genome and assigned to the DpnII restriction fragments. The assigned read pairs were filtered to detect valid interactions by removing singletons, dangling‐ends, self‐cycles, and dumped pairs. The validated pairs were transformed into a hic file format using the hicpro2juicebox function with Knight‐Ruiz normalization. H3K4me3‐related loops were identified in both control and CTCF‐depleted cells using FitHiChIP with a minimum distance threshold of 10 Kb and FDR threshold of 0.05 (Data S3). 41 Sum of contact counts (sumCC) from FitHiChIP output were compared between control and CTCF‐depleted NPCs to determine the change of interaction frequency. To calculate the aggregate peak analysis (APA) score, hic files were converted to cool files with 5 Kb resolution using the hic2cool tool. The APA score for enhancer–promoter loops was determined by the coolpup.py with parameter—mindist 0 and visualized using the plotpup.py with parameters—center 1—ignore_central 1. 42 Chromatin interactions, H3K27ac, and gene transcription in mouse NPCs were visualized using pyGenomeTracks. 43

2.9. Statistical analysis

p‐Values for ChIP, 3C, and RT‐qPCR data were calculated using the two‐tailed Student's t‐test. Gene transcription and sumCC for enhancer–promoter (E–P) loops in control and CTCF‐depleted mouse NPCs were statistically analyzed using the Wilcoxon rank‐sum test. The p‐value less than .05 means statistical significance (*p < .05, **p < .01). the normality of RNA‐seq and PLAC‐seq data was checked using Kolmogorov–Smirnov test.

3. RESULTS

3.1. Deletion of CTCF‐binding motifs at both boundaries of β‐globin TAD and sub‐TAD disrupts the topological structure

The human β‐globin locus is surrounded by five CTCF‐binding sites that are referred to as C3–C7 in this study (Figure 1A). The CTCF sites correspond to the boundaries of the β‐globin TAD and sub‐TADs. The binding motifs of the five CTCF sites were convergent in the direction but divergent to the neighboring motifs C2 and C8. To disrupt TAD and a sub‐TAD containing the human β‐globin locus, we deleted two sets of CTCF‐binding motifs using the CRISPR/Cas9 system in MEL/ch11 cells. 29 One set contained C3 and C7 present at the TAD boundaries and the other contained C4 and C5 at the sub‐TAD boundaries. Two clonal cells were obtained for each set of deletions in which 11–41 bp, including the CTCF motif, were deleted (Figure 1B). Deletion of the CTCF motif abolished CTCF binding, as shown using the ChIP assay (Figure 1C). The binding of SMC3, a subunit of the cohesin complex, was significantly compromised by the motif deletion (Figure 1D). A 3C assay was performed to analyze the chromatin interactions. It was shown that the deletion of C3 and C7 (ΔTAD, ΔT) decreased interaction between DNA fragments corresponding to the TAD boundaries (Figure 1E left graph). Interaction between DNA fragments corresponding to the sub‐TAD boundaries was abolished by the deletion of C4 and C5 (Δsub‐TAD, ΔsT) (Figure 1E right graph). The CTCF motif‐deleted boundaries lost interaction with the unmodified boundaries of TAD or sub‐TADs. In addition, reduced interaction was observed between the C3 TAD boundary and non‐boundary regions lacking CTCF sites (marked as an asterisk in the X‐axis of the graph) in ΔT clonal cells (Figure 1E left graph). These results indicate that the two sets of CTCF motif deletion disrupted β‐globin TAD and sub‐TAD, respectively, by leading to failure of chromatin interaction between their boundaries.

3.2. Disruption of the interaction between β‐globin TAD boundaries prevents sub‐TAD formation, but the interaction between sub‐TAD boundaries does not affect TAD formation

To examine how TAD and sub‐TAD affect each other, chromatin interaction between their boundaries was analyzed in control, ΔT, and ΔsT cells. DNA fragments corresponding to one boundary of β‐globin sub‐TADs (C5) were used as an anchor to analyze chromatin interaction with the other sub‐TAD boundary (C4). This interaction was significantly reduced in ΔT clonal cells (Figure 2A left graph), indicating the failure of sub‐TAD formation owing to the disruption of TAD. As expected, the interaction of the sub‐TAD boundary with TAD boundaries (C3 and C7) was reduced. In contrast, the interaction between TAD boundaries was maintained in two ΔsT clonal cells (Figure 2A right graph), even though DNA fragments containing C6, which are close to the C7 TAD boundary, lost interaction with the C3 TAD boundary. The unaffected interaction between fragments containing C3 and C7 is considered to show the maintenance of TAD. To examine whether these results were related to CTCF occupancy in the unmodified CTCF motifs, we analyzed the CTCF ChIPed DNA. CTCF was detected in unmodified motifs without notable changes (Figure 2B). Taken together, it appears that TAD is required for sub‐TAD formation but sub‐TAD is dispensable for TAD formation. The CTCF binding itself is not likely to be sufficient for sub‐TAD formation.

FIGURE 2.

FIGURE 2

Open in a new tab

Chromatin interaction between β‐globin TAD and sub‐TAD boundaries in ΔTAD and Δsub‐TAD MEL/ch11 cells. (A) Relative ligation frequencies between fragments containing CTCF motifs were analyzed using the 3C assay in control (Con), ΔTAD (ΔT), and Δsub‐TAD (ΔsT) MEL/ch11 cells. Fragments containing C5 and C3 were used as anchors to measure ligation frequencies between boundaries of sub‐TAD (blue shadow) and TAD (pink shadow), respectively. Asterisks on the X‐axis represent negative fragments. (B) Occupancy of CTCF was analyzed at unmodified CTCF sites using ChIP‐qPCR. Normal rabbit IgG (Ig) and mouse Actin genes served as negative controls. (C) Hi‐C map and CTCF ChIP‐seq track show the β‐globin TAD and two adjacent TADs in human erythroblast. Three pink shadows flanked by blue triangles represent TADs. Under the diagram, vertical gray bars indicate Hind III sites for DNA fragments containing CTCF sites. Horizontal arrows mark the location of primers used in the 3C assay. Relative ligation frequencies in ΔT cells were analyzed between fragments containing the C1 and other CTCF sites by comparing them to ligation frequency in control cells. Data are presented as the means ± SEM of five 3C and two to four ChIP biological repeats. p‐Values were calculated using the two‐tailed Student's t‐test. *p < .05, **p < .01.

The β‐globin TAD neighbors on other TADs, which contain CTCF sites at their boundaries (Figure 2C). To extend the analysis of topological structure into neighboring TADs, we performed a 3C assay using the boundary of adjacent TAD (C1) as an anchor. The C1 boundary interacted with C2, C8, and C9 boundaries at a similar level in control and ΔT#1 cells, supporting the unaffected formation of neighboring TADs (Figure 2C bottom graph). However, the interaction of the C1 boundary with the β‐globin TAD boundaries (C3 and C7) was decreased in ΔT cells. Increased interaction was observed between C1 boundary and fragments containing C5 and C6. These results could be interpreted that a part of β‐globin TAD merged into adjacent TAD by the disruption of TAD structure, even though this change of topological structure did not prevent the β‐globin sub‐TAD disruption.

3.3. Disruption of TAD and sub‐TAD compromises transcription of the β‐globin gene

Unequal relationship between the β‐globin TAD and sub‐TAD in their organization generated two types of topological situations, namely, the loss of both TAD and sub‐TAD and the maintenance of TAD without sub‐TAD. We analyzed transcription of the human β‐globin gene using RT‐qPCR in ΔT and ΔsT cells after differentiation for 72 h. Although MEL/ch11 control and Δ clonal cells were similarly changed into red color after differentiation (Figure 3A), the transcription of human β‐globin was significantly decreased by ~80% in both ΔT and ΔsT cells compared with control cells (Figure 3B). This is accompanied by weakened interaction of the β‐globin gene with LCR hypersensitive sites (HSs) (Figure 3C). The chromatin looping factors, GATA1, TAL1, and LDB1, were also less frequently detected in LCR HSs in ΔT#1 and ΔsT#1 cells (Figure 3D). These results indicate that both TAD and sub‐TAD are required for β‐globin transcription. It is likely that TAD itself is not sufficient for chromatin interaction between enhancers and promoters, and that sub‐TAD is necessary.

FIGURE 3.

FIGURE 3

Open in a new tab

β‐Globin transcription and enhancer–promoter interaction in ΔTAD and Δsub‐TAD MEL/ch11 cells. (A) Control (Con) and Δ clonal MEL/ch11 cells were differentiated using HMBA treatment to activate the β‐globin gene. (B) β‐globin gene transcription was measured using RT‐qPCR with two primer sets spanning two exons and was normalized with cDNA of mouse Actin gene. (C) The human β‐globin locus is depicted in detail. Four HSs are shown with yellow rectangles in the LCR. Vertical gray bars indicate Hind III cut sites and horizontal arrows mark the location of 3C primers in the β‐globin locus (black indicates positive fragments and gray indicates negative fragments). Relative ligation frequencies were analyzed between fragments containing the β‐globin gene and LCR HSs using the 3C assay. The fragment containing the β‐globin gene was used as an anchor. DNA fragments lacking CTCF sites were represented with asterisks on the X‐axis. (D) Occupancy of chromatin looping factors was analyzed at the LCR HSs using ChIP‐qPCR. Immunoprecipitated DNA was normalized by input DNA. The mouse Actin gene served as a negative control. Data are presented as the means ± SEM of three to four RT‐qPCR, five 3C, and two to three ChIP biological repeats. p‐Values were calculated using a two‐tailed Student's t‐test. *p < .05, **p < .01.

3.4. Histone H3 occupancy and H3K27ac are largely maintained without TAD and sub‐TAD in the β‐globin locus

An active chromatin structure is required for the binding of transcription factors mediating enhancer–promoter interactions. To investigate how the chromatin structure was affected by the disruption of TAD and sub‐TAD, we analyzed the occupancy of histone H3 in ΔT#1 and ΔsT#1 cells. Histone H3 occupancy was increased at C3 and C7 by the CTCF motif deletion, even though this increase was not notable in the motif‐deleted C4 and C5 (Figure 4A). The β‐globin locus did not exhibit a notable change of H3 occupancy in ΔT#1 and ΔsT#1 cells. In particular, H3 eviction (low level of H3 occupancy) at the LCH HSs was sustained, despite the disruption of TAD or sub‐TAD. Next, acetylation at lysine (K) 27 of histone H3 (H3K27ac), an active histone modification, was analyzed across the β‐globin TAD using the ChIP‐seq assay. The loss of CTCF motifs in ΔT#1 and ΔsT#1 cells decreased H3K27ac at the boundaries of TAD and sub‐TAD, respectively (Figure 4B). However, H3K27ac enrichment across the β‐globin locus was largely maintained in ΔT#1 and ΔsT#1 cells. These results indicate that the active chromatin structure of the β‐globin locus is formed independently from specific topological organization.

FIGURE 4.

FIGURE 4

Open in a new tab

Histone occupancy and H3K27ac of the human β‐globin locus in ΔTAD and Δsub‐TAD MEL/ch11 cells. (A) Occupancy of histone H3 in the β‐globin locus was analyzed using ChIP‐qPCR. Immunoprecipitated DNA was compared with input DNA and normalized to the mouse Actin gene, which served as an internal control. Normal rabbit IgG (Ig) served as a negative control. Data present the means ± SEM of three to four biological repeats. (B) Histone H3K27ac was analyzed using ChIP‐seq in and around the β‐globin TAD. Histone H3K27me3 was presented for control cells. At the bottom of the tracks, CTCF sites and the human β‐globin locus are depicted by triangles and rectangles, respectively. p‐Values were calculated using the two‐tailed Student's t‐test. *p < .05.

3.5. Disruption of topological domains impairs enhancer–promoter interactions across the genome

TAD and sub‐TAD were required for the interaction of the β‐globin promoter with LCR HSs. To explore whether TAD and sub‐TAD affect enhancer–promoter interactions across the genome, we analyzed H3K4me3 PLAC‐seq (also known as HiChIP), ChIP‐seq, and RNA‐seq data in mouse control and CTCF‐depleted NPCs, which are publicly available. 44 It was reported that CTCF‐mediated TADs were genome‐widely disrupted in the CTCF‐depleted NPCs. 44 To identify genome‐wide enhancer–promoter interactions, the anchors of H3K4me3‐related loops were overlapped with putative enhancers and transcription start sites (TSSs) of highly transcribed genes (transcripts per million (TPM) ≥ 50), because robust gene transcription may be promoted by enhancers (Figure 5A). H3K4me3 loops containing enhancers at one anchor and TSSs at the other anchor were defined as enhancer–promoter (E–P) loops. More than half of E–P loops were disappeared in the CTCF‐depleted cells (Figure 5B). The interaction of remaining E–P loops (n = 802) weakened as shown in the APA analysis and comparative analysis of interaction frequency (Figure 5C,D). This result was accompanied by a transcriptional reduction in genes interacting with enhancers (wEnh) (Figure 5E). However, H3K27ac was largely maintained in the enhancers interacting with promoters in spite of the disruption of topological domains (Figure 5F). Two representative gene loci, the Cd9 and Gdi1, showed a decrease of enhancer–promoter interaction and gene transcription in the CTCF‐depleted NPCs where topological domains surrounding the gene loci were disrupted (Figure 5G). These results are compatible with our findings in the β‐globin locus that resulted from the disruption of TAD and sub‐TAD by CTCF motif deletion at their boundaries. Thus, these genome‐wide results are thought to support the idea that TADs and sub‐TADs are required for enhancer–promoter interactions enabling gene transcription.

FIGURE 5.

FIGURE 5

Open in a new tab

Genome‐wide enhancer–promoter interactions and gene transcription in CTCF‐depleted NPCs. (A) Enhancer–promoter (E–P) loops (n = 1839) were identified using public H3K4me3 PLAC‐seq, ATAC‐seq, ChIP‐seq, and RNA‐seq data in mouse NPCs by overlapping the anchors of H3K4me3 loops (n = 71 368) with putative enhancers (Enh; n = 4497) and TSSs of highly transcribed genes (TPM ≥ 50, n = 4836). (B) E–P loops were counted in control (Con) and CTCF‐depleted (dCTCF) NPCs. (C) The APA plot represents the average interaction score for the remaining E–P loops (n = 802) using H3K4me3 PLAC‐seq data. (D) Violin plots represent log2‐transformed contact counts of the remaining E–P loops. (E) Transcription was measured in genes interacting with enhancers (wEnh; n = 1257) and not interacting with enhancers (nwEnh; n = 2397) using RNA‐seq data. Violin plots represent log2‐transformed TPM values. (F) Histone H3K27ac was analyzed in the enhancers (n = 1478) located at the anchors of the E–P loops. The enhancers were sorted by Con H3K27ac signals in heatmaps. (G) Con and dCTCF Hi‐C, H3K4me3 PLAC‐seq, CTCF ChIP‐seq, RNA‐seq, and H3K27ac ChIP‐seq data were visualized at two representative loci, the Cd9 and Gdi1 genes, using pyGenomeTracks. The location and orientation of annotated genes are depicted as black rectangles with arrowheads at the bottom of the tracks. Red and black dashed circles indicate interaction between CTCF sites and between putative enhancers and the TSSs, respectively. The enhancers and genes are highlighted by yellow and gray shadows, respectively. p‐Values were calculated using the Wilcoxon rank‐sum test. *p < .05, **p < .01, NS, not significant.

4. DISCUSSION

The present study using the destruction of both boundaries shows the unequal relationship between TAD and sub‐TAD in their domain formation and the requirement of TAD‐dependently formed sub‐TAD in gene transcription. The β‐globin TAD was required for sub‐TAD containing the β‐globin locus, but not vice versa (Figure 6). This unequal relationship indicates that TAD is insufficient for gene transcription and enhancer–promoter interaction in the β‐globin locus, and sub‐TAD is required. However, an active chromatin structure of the β‐globin LCR was established in TAD‐ and sub‐TAD‐independent manners. Our genome‐wide analysis supports the notion that the topological domains are required for the interaction between enhancers and promoters rather than the active histone modification in enhancers. Thus, TAD‐dependently formed sub‐TADs are considered to contribute enhancer–promoter interactions that enable gene transcription.

FIGURE 6.

FIGURE 6

Open in a new tab

Effect of disruption of TAD and sub‐TAD in the human β‐globin locus. Results were combined into a diagram. Both TAD and sub‐TAD were disrupted in ΔTAD cells, but TAD was maintained irrespective of disruption of sub‐TADs in Δsub‐TAD cells. Chromatin interaction between LCR HSs and the β‐globin gene, binding of looping factors to the LCR HSs, and transcription of the β‐globin gene decreased in both ΔTAD and Δsub‐TAD cells, whereas histone H3 eviction and H3K27ac at LCR HSs were largely maintained in the β‐globin locus.

Our reciprocal analysis showed that β‐globin TAD was required for sub‐TAD formation, but that sub‐TAD was not required for TAD formation. This unequal relationship is considered to reveal the topological hierarchy of chromatin organization. This is compatible with previous studies showing invariant TAD across cell types and differentiation‐related sub‐TAD formation. 1 , 12 , 13 , 45 Chromatin interaction between β‐globin TAD boundaries is unaffected by reduced/enforced H3K27ac and elimination of enhancer 46 , 47 but the β‐globin sub‐TAD is formed in erythroid cell specifically and erythroid factor GATA‐1 dependently. 27 This kind of unequal relationship is likely to be present between sub‐TAD and enhancer–promoter interaction. The β‐globin sub‐TAD was established when the enhancer–promoter interaction of the β‐globin locus was impaired by the depletion of looping factor TAL1. 48 Sub‐TAD‐dependent enhancer–promoter interactions were observed in other loci, such as the HoxD and Thy1 loci. 19 , 20 Therefore, these findings suggest that hierarchy is present in the formation of the topological structures of chromatin. TADs may provide a special environment for sub‐TADs, and sub‐TADs may provide this for enhancer–promoter interactions.

TAD and sub‐TAD appear to be required for gene transcription in the β‐globin locus. Even the destruction of one boundary for TAD or sub‐TAD did not affect the β‐globin transcription in our previous study, 28 it was likely due to complementary domains that were newly formed by the remaining boundaries. These complementary domains were likely formed for other loci, such as the Runx1 locus and Csn locus that showed unaffected gene transcription with the loss of one boundary for TAD or sub‐TAD. 12 , 13 , 21 , 22 , 23 The new TADs and sub‐TADs may be formed by moving a TAD boundary to neighboring sub‐TAD boundary and merging two sub‐TADs into one domain, respectively, and may functionally substitute for the original domains. However, the complete disruption of domains owing to the loss of both boundaries compromised the β‐globin gene transcription in the present study and was not recovered functionally by the newly formed interaction with a neighboring TAD. Thus, the TADs and sub‐TADs are considered to be essential for genes inside the domains to be transcribed in a chromatin context.

Enhancers have active chromatin structures that are hypersensitive to DNase I, less occupied by histones, and enriched with histone acetylation such as H3K27ac. 49 , 50 This structure allows eRNA transcription and transcription factor binding, consequently mediating the interaction between enhancers and target genes. 51 , 52 , 53 The results of the present study showed that an active histone modification H3K27ac and histone H3 eviction can occur in the β‐globin enhancer LCR HSs independently of TAD or sub‐TAD formation. Unaffected H3K27ac was observed in the Tcra‐Tcrd locus where one boundary of sub‐TAD flanking the locus was removed, even the rearrangement of gene segments was impaired in the Tcra locus by the disruption of sub‐TAD. 16 In contrast, the interaction of enhancers with target genes appears to require compact domains such as the sub‐TAD as shown by the complete disruption of the β‐globin TAD and sub‐TAD in the present study. These findings are supported by our genome‐wide analysis. A recent study showed that the enhancer–promoter interactions of the Trim25 and Irank2 loci disappeared due to the depletion of CTCF in mouse dendritic cells. 54 Therefore, TAD‐dependent sub‐TADs may play an essential role in gene transcription by providing a three‐dimensional environment for enhancers–promoter interaction.

AUTHOR CONTRIBUTIONS

AeRi Kim designed the research; Dasoul Lee conducted the experiment and acquired the data; Dasoul Lee and Jin Kang analyzed and interpreted the data; Dasoul Lee, Jin Kang, and AeRi Kim wrote the original draft. AeRi Kim revised the manuscript. All authors read and approved the final manuscript.

DISCLOSURES

The authors declare no conflicts of interest.

Supporting information

Data S1. Putative enhancers identified using public ATAC‐seq and ChIP‐seq data in mouse NPCs.

FSB2-38-e70181-s001.xlsx (328.9KB, xlsx)

Data S2. TSSs of the highly transcribed genes (TPM ≥ 50).

FSB2-38-e70181-s004.xlsx (12.1MB, xlsx)

Data S3. Loops identified using public H3K4me3 PLAC‐seq data in mouse NPCs.

FSB2-38-e70181-s002.xlsx (161.1KB, xlsx)

Table S1. The sequences for sgRNA oligos, primers, and TaqMan probes used for this study.

FSB2-38-e70181-s003.docx (24.1KB, docx)

ACKNOWLEDGMENTS

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF‐2020R1I1A3054808) and by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. 2022R1C1C2006355).

Lee D, Kang J, Kim A. TAD‐dependent sub‐TAD is required for enhancer–promoter interaction enabling the β‐globin transcription. The FASEB Journal. 2024;38:e70181. doi: 10.1096/fj.202401526RR

DATA AVAILABILITY STATEMENT

MEL/ch11 control, ΔTAD, and Δsub‐TAD H3K27ac, and control H3K27me3 ChIP‐seq data are available in the Gene Expression Omnibus (GEO) database under GEO accession number GSE269934. The GEO accession numbers for public NGS data are GSE102197 for Hi‐C and GSE102184 for CTCF ChIP‐seq in human adult erythroblast; GSE241376 for ATAC‐seq in wildtype mouse NPCs; and GSE94452 for CTCF, H3K4me1, H3K27ac ChIP‐seq, RNA‐seq, Hi‐C, and H3K4me3 PLAC‐seq in control and CTCF‐depleted NPCs.

REFERENCES

  • 1. Dixon JR, Selvaraj S, Yue F, et al. Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature. 2012;485:376‐380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Nora EP, Lajoie BR, Schulz EG, et al. Spatial partitioning of the regulatory landscape of the X‐inactivation centre. Nature. 2012;485:381‐385. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Hou C, Zhao H, Tanimoto K, Dean A. CTCF‐dependent enhancer‐blocking by alternative chromatin loop formation. Proc Natl Acad Sci U S A. 2008;105:20398‐20403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Nora EP, Goloborodko A, Valton AL, et al. Targeted degradation of CTCF decouples local insulation of chromosome domains from genomic compartmentalization. Cell. 2017;169:930‐944. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Zuin J, Dixon JR, van der Reijden MI, et al. Cohesin and CTCF differentially affect chromatin architecture and gene expression in human cells. Proc Natl Acad Sci U S A. 2014;111:996‐1001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Hyle J, Zhang Y, Wright S, et al. Acute depletion of CTCF directly affects MYC regulation through loss of enhancer‐promoter looping. Nucleic Acids Res. 2019;47:6699‐6713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Lee R, Kang MK, Kim YJ, et al. CTCF‐mediated chromatin looping provides a topological framework for the formation of phase‐separated transcriptional condensates. Nucleic Acids Res. 2022;50:207‐226. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Anania C, Acemel RD, Jedamzick J, et al. In vivo dissection of a clustered‐CTCF domain boundary reveals developmental principles of regulatory insulation. Nat Genet. 2022;54:1026‐1036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Li Y, Liao Z, Luo H, et al. Alteration of CTCF‐associated chromatin neighborhood inhibits TAL1‐driven oncogenic transcription program and leukemogenesis. Nucleic Acids Res. 2020;48:3119‐3133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Luo H, Wang F, Zha J, et al. CTCF boundary remodels chromatin domain and drives aberrant HOX gene transcription in acute myeloid leukemia. Blood. 2018;132:837‐848. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Phillips‐Cremins JE, Sauria ME, Sanyal A, et al. Architectural protein subclasses shape 3D organization of genomes during lineage commitment. Cell. 2013;153:1281‐1295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Owens DDG, Anselmi G, Oudelaar AM, et al. Dynamic Runx1 chromatin boundaries affect gene expression in hematopoietic development. Nat Commun. 2022;13:773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Hanssen LLP, Kassouf MT, Oudelaar AM, et al. Tissue‐specific CTCF‐cohesin‐mediated chromatin architecture delimits enhancer interactions and function in vivo. Nat Cell Biol. 2017;19:952‐961. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Fudenberg G, Imakaev M, Lu C, Goloborodko A, Abdennur N, Mirny LA. Formation of chromosomal domains by loop extrusion. Cell Rep. 2016;15:2038‐2049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Sanborn AL, Rao SS, Huang SC, et al. Chromatin extrusion explains key features of loop and domain formation in wild‐type and engineered genomes. Proc Natl Acad Sci U S A. 2015;112:E6456‐E6465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Zhao H, Li Z, Zhu Y, et al. A role of the CTCF binding site at enhancer Ealpha in the dynamic chromatin organization of the Tcra‐Tcrd locus. Nucleic Acids Res. 2020;48:9621‐9636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Hnisz D, Weintraub AS, Day DS, et al. Activation of proto‐oncogenes by disruption of chromosome neighborhoods. Science. 2016;351:1454‐1458. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Sreenivas P, Wang L, Wang M, et al. A SNAI2/CTCF interaction is required for NOTCH1 expression in rhabdomyosarcoma. Mol Cell Biol. 2023;43:547‐565. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Rodriguez‐Carballo E, Lopez‐Delisle L, Willemin A, et al. Chromatin topology and the timing of enhancer function at the HoxD locus. Proc Natl Acad Sci U S A. 2020;117:31231‐31241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Ren G, Jin W, Cui K, et al. CTCF‐mediated enhancer‐promoter interaction is a critical regulator of cell‐to‐cell variation of gene expression. Mol Cell. 2017;67:1049‐1058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Cao Y, Liu S, Cui K, Tang Q, Zhao K. Hi‐TrAC detects active sub‐TADs and reveals internal organizations of super‐enhancers. Nucleic Acids Res. 2023;51:6172‐6189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Lee HK, Willi M, Wang C, et al. Functional assessment of CTCF sites at cytokine‐sensing mammary enhancers using CRISPR/Cas9 gene editing in mice. Nucleic Acids Res. 2017;45:4606‐4618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Willi M, Yoo KH, Reinisch F, et al. Facultative CTCF sites moderate mammary super‐enhancer activity and regulate juxtaposed gene in non‐mammary cells. Nat Commun. 2017;8:16069. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Krivega I, Dean A. Chromatin looping as a target for altering erythroid gene expression. Ann N Y Acad Sci. 2016;1368:31‐39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Huang P, Keller CA, Giardine B, et al. Comparative analysis of three‐dimensional chromosomal architecture identifies a novel fetal hemoglobin regulatory element. Genes Dev. 2017;31:1704‐1713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Rao SS, Huntley MH, Durand NC, et al. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell. 2014;159:1665‐1680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Kang Y, Kim YW, Kang J, Yun WJ, Kim A. Erythroid specific activator GATA‐1‐dependent interactions between CTCF sites around the β‐globin locus. Biochim Biophys Acta Gene Regul Mech. 2017;1860:416‐426. [DOI] [PubMed] [Google Scholar]
  • 28. Kang J, Kim YW, Park S, Kang Y, Kim A. Multiple CTCF sites cooperate with each other to maintain a TAD for enhancer‐promoter interaction in the β‐globin locus. FASEB J. 2021;35:e21768. [DOI] [PubMed] [Google Scholar]
  • 29. Kim S, Kim YW, Shim SH, Kim CG, Kim A. Chromatin structure of the LCR in the human β‐globin locus transcribing the adult δ‐ and β‐globin genes. Int J Biochem Cell Biol. 2012;44:505‐513. [DOI] [PubMed] [Google Scholar]
  • 30. Kim YW, Kim A. Deletion of transcription factor binding motifs using the CRISPR/spCas9 system in the β‐globin LCR. Biosci Rep. 2017;37:BSR20170976. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Cho Y, Song SH, Lee JJ, et al. The role of transcriptional activator GATA‐1 at human β‐globin HS2. Nucleic Acids Res. 2008;36:4521‐4528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30:2114‐2120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Langmead B, Salzberg SL. Fast gapped‐read alignment with Bowtie 2. Nat Methods. 2012;9:357‐359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Li H, Handsaker B, Wysoker A, et al. The sequence alignment/map format and SAMtools. Bioinformatics. 2009;25:2078‐2079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Xiang G, Giardine BM, Mahony S, Zhang Y, Hardison RC. S3V2‐IDEAS: a package for normalizing, denoising and integrating epigenomic datasets across different cell types. Bioinformatics. 2021;37:3011‐3013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Freese NH, Norris DC, Loraine AE. Integrated genome browser: visual analytics platform for genomics. Bioinformatics. 2016;32:2089‐2095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Ramirez F, Ryan DP, Gruning B, et al. deepTools2: a next generation web server for deep‐sequencing data analysis. Nucleic Acids Res. 2016;44:W160‐W165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Dobin A, Davis CA, Schlesinger F, et al. STAR: ultrafast universal RNA‐seq aligner. Bioinformatics. 2013;29:15‐21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30:923‐930. [DOI] [PubMed] [Google Scholar]
  • 40. Servant N, Varoquaux N, Lajoie BR, et al. HiC‐Pro: an optimized and flexible pipeline for Hi‐C data processing. Genome Biol. 2015;16:259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Bhattacharyya S, Chandra V, Vijayanand P, Ay F. Identification of significant chromatin contacts from HiChIP data by FitHiChIP. Nat Commun. 2019;10:4221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Flyamer IM, Illingworth RS, Bickmore WA. Coolpup.py: versatile pile‐up analysis of Hi‐C data. Bioinformatics. 2020;36:2980‐2985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Lopez‐Delisle L, Rabbani L, Wolff J, et al. pyGenomeTracks: reproducible plots for multivariate genomic datasets. Bioinformatics. 2021;37:422‐423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Kubo N, Ishii H, Xiong X, et al. Promoter‐proximal CTCF binding promotes distal enhancer‐dependent gene activation. Nat Struct Mol Biol. 2021;28:152‐161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Smith EM, Lajoie BR, Jain G, Dekker J. Invariant TAD boundaries constrain cell‐type‐specific looping interactions between promoters and distal elements around the CFTR locus. Am J Hum Genet. 2016;98:185‐201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Kim YW, Kang Y, Kang J, Kim A. GATA‐1‐dependent histone H3K27 acetylation mediates erythroid cell‐specific chromatin interaction between CTCF sites. FASEB J. 2020;34:14736‐14749. [DOI] [PubMed] [Google Scholar]
  • 47. Kim J, Kang J, Kim YW, Kim A. The human β‐globin enhancer LCR HS2 plays a role in forming a TAD by activating chromatin structure at neighboring CTCF sites. FASEB J. 2021;35:e21669. [DOI] [PubMed] [Google Scholar]
  • 48. Yun WJ, Kim YW, Kang Y, Lee J, Dean A, Kim A. The hematopoietic regulator TAL1 is required for chromatin looping between the β‐globin LCR and human γ‐globin genes to activate transcription. Nucleic Acids Res. 2014;42:4283‐4293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Ong CT, Corces VG. Enhancer function: new insights into the regulation of tissue‐specific gene expression. Nat Rev Genet. 2011;12:283‐293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. Tafessu A, Banaszynski LA. Establishment and function of chromatin modification at enhancers. Open Biol. 2020;10:200255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51. Kim YW, Kim A. Histone acetylation contributes to chromatin looping between the locus control region and globin gene by influencing hypersensitive site formation. Biochim Biophys Acta Gene Regul Mech. 2013;1829:963‐969. [DOI] [PubMed] [Google Scholar]
  • 52. Kang Y, Kim YW, Kang J, Kim A. Histone H3K4me1 and H3K27ac play roles in nucleosome eviction and eRNA transcription, respectively, at enhancers. FASEB J. 2021;35:e21781. [DOI] [PubMed] [Google Scholar]
  • 53. Zhu Y, Sun L, Chen Z, Whitaker JW, Wang T, Wang W. Predicting enhancer transcription and activity from chromatin modifications. Nucleic Acids Res. 2013;41:10032‐10043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54. Yang B, Kim S, Jung WJ, et al. CTCF controls three‐dimensional enhancer network underlying the inflammatory response of bone marrow‐derived dendritic cells. Nat Commun. 2023;14:1277. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1. Putative enhancers identified using public ATAC‐seq and ChIP‐seq data in mouse NPCs.

FSB2-38-e70181-s001.xlsx (328.9KB, xlsx)

Data S2. TSSs of the highly transcribed genes (TPM ≥ 50).

FSB2-38-e70181-s004.xlsx (12.1MB, xlsx)

Data S3. Loops identified using public H3K4me3 PLAC‐seq data in mouse NPCs.

FSB2-38-e70181-s002.xlsx (161.1KB, xlsx)

Table S1. The sequences for sgRNA oligos, primers, and TaqMan probes used for this study.

FSB2-38-e70181-s003.docx (24.1KB, docx)

Data Availability Statement

MEL/ch11 control, ΔTAD, and Δsub‐TAD H3K27ac, and control H3K27me3 ChIP‐seq data are available in the Gene Expression Omnibus (GEO) database under GEO accession number GSE269934. The GEO accession numbers for public NGS data are GSE102197 for Hi‐C and GSE102184 for CTCF ChIP‐seq in human adult erythroblast; GSE241376 for ATAC‐seq in wildtype mouse NPCs; and GSE94452 for CTCF, H3K4me1, H3K27ac ChIP‐seq, RNA‐seq, Hi‐C, and H3K4me3 PLAC‐seq in control and CTCF‐depleted NPCs.

Articles from The FASEB Journal are provided here courtesy of Wiley

RESOURCES