pubmed.ncbi.nlm.nih.gov

Activation of the alpha-globin gene expression correlates with dramatic upregulation of nearby non-globin genes and changes in local and large-scale chromatin spatial structure - PubMed

  • ️Sun Jan 01 2017

Activation of the alpha-globin gene expression correlates with dramatic upregulation of nearby non-globin genes and changes in local and large-scale chromatin spatial structure

Sergey V Ulianov et al. Epigenetics Chromatin. 2017.

Abstract

Background: In homeotherms, the alpha-globin gene clusters are located within permanently open genome regions enriched in housekeeping genes. Terminal erythroid differentiation results in dramatic upregulation of alpha-globin genes making their expression comparable to the rRNA transcriptional output. Little is known about the influence of the erythroid-specific alpha-globin gene transcription outburst on adjacent, widely expressed genes and large-scale chromatin organization. Here, we have analyzed the total transcription output, the overall chromatin contact profile, and CTCF binding within the 2.7 Mb segment of chicken chromosome 14 harboring the alpha-globin gene cluster in cultured lymphoid cells and cultured erythroid cells before and after induction of terminal erythroid differentiation.

Results: We found that, similarly to mammalian genome, the chicken genomes is organized in TADs and compartments. Full activation of the alpha-globin gene transcription in differentiated erythroid cells is correlated with upregulation of several adjacent housekeeping genes and the emergence of abundant intergenic transcription. An extended chromosome region encompassing the alpha-globin cluster becomes significantly decompacted in differentiated erythroid cells, and depleted in CTCF binding and CTCF-anchored chromatin loops, while the sub-TAD harboring alpha-globin gene cluster and the upstream major regulatory element (MRE) becomes highly enriched with chromatin interactions as compared to lymphoid and proliferating erythroid cells. The alpha-globin gene domain and the neighboring loci reside within the A-like chromatin compartment in both lymphoid and erythroid cells and become further segregated from the upstream gene desert upon terminal erythroid differentiation.

Conclusions: Our findings demonstrate that the effects of tissue-specific transcription activation are not restricted to the host genomic locus but affect the overall chromatin structure and transcriptional output of the encompassing topologically associating domain.

Keywords: Alpha-globin genes; CTCF; Chromatin compartment; Chromatin spatial structure; TAD; Transcription.

PubMed Disclaimer

Figures

Fig. 1
Fig. 1

Analysis of transcriptomes of the studied cell types. a Genome-wide variation of gene expression in the studied cell types (proliferating and differentiated HD3 cells are designated by HD3pr and HD3dif, respectively). Numbers of upregulated, downregulated, and ubiquitously transcribed genes for each pair of the cell types are shown at the upper left corner of the plots. The following genes are highlighted in the plots: (1) erythroid transcription factors GATA2, SCL (Tal1), FOG1, LMO2, NF-E2, Ldb1 and KLF1 (EKLF); (2) enzymes involved in heme synthesis including FECH and CPOX; (3) transferrin receptor (TFRC); and (4) the alpha-globin gene π (HBZ) and beta-globin gene (HBG2). BACH2 and EBF1 are lymphoid transcription factors. b Normalized profiles of total rRNA-depleted RNA-seq within the studied region. The functional AgGD from the 3′-end of the NPRL3 gene to the 3′-end of the TMEM8A gene is highlighted in pink. Alpha-globin genes are highlighted in red, and non-globin genes located within the AgGD are highlighted in brown. c Transcription level changes between cell types for all genes within the studied region. d The RNA-seq profile of the AgGD and closest neighbors. Intergenic transcription profiles are highlighted in black, and genic transcription profiles are highlighted in gray. Positions of genes from the Ensembl database are highlighted in red

Fig. 2
Fig. 2

Analysis of A/B-like chromatin compartment structure of the studied genomic region. a Heatmap of the differentiated HD3 cells demonstrating an increased interaction frequency between gene-rich zones I and III. A-like and B-like chromatin compartments are outlined using black rectangle and black stipple triangle, respectively. The first principal component value is shown below the heatmap. Structural zones I (5′-terminal gene-rich area), II (gene desert), and III (3′-terminal gene-rich area) were selected manually based on the density of CpG-islands and annotated genes. b The averaged interaction profiles of 18 genomic bins #74-91 from zone III (virtual “anchor”) with all bins from the remaining part of the studied region. Circles show average values of 5C counts between genomic bins #74-91 with each of the other bins in the studied region; standard deviation is shown. c 5C counts corresponding to interactions of the AgGD and flanking regions from the TAD T3 with the gene desert. Thick black lines represent median values. Two asterisks represent a significant difference with a P-value <0.01 calculated using a one-tailed Wilcoxon’s signed-rank test. Three asterisks represent a significant difference with a P-value <0.001. The same notations are in panels (d) and (e). d The distributions of 5C counts within the A-like chromatin compartment. e 5C counts between genomic bins #74-91 and the remaining part of zone III and the gene desert. f 5C counts between four groups of bins from zones I and III (differentiated HD3 cells); bins were divided into groups according to their transcription level. g 5C counts between four groups of bins from zones I and III; bins were divided into groups according to their degree of the transcription level increase in the HD3dif cells as compared to the HD3pr cells

Fig. 3
Fig. 3

The studied genomic region is partitioned into TADs largely conserved between lymphoid and erythroid cells. The heatmaps show 5C data normalized by the total number of sequencing reads in the 5C dataset, binned at a 30 Kb resolution, iteratively corrected and smoothed. Histograms of the 5C counts are shown to the right of the heatmaps. Gray rectangles below the heatmaps show TADs that were annotated using the Lavaburst package. TADs T1 and T2 (recognized as fused into one domain T3 in erythroid cells) harbor the alpha-globin gene domain and flanking regions. The graphs demonstrating CTCF binding in the three cell types are based on previously published ChIP-seq data [39]. The direction of forward- and reverse-oriented CTCF binding motifs within the ChIP-seq peaks is shown below the ChIP-seq peaks using red and blue tailless arrows, respectively

Fig. 4
Fig. 4

High-resolution C-TALE-identified TAD profile around AgGD. a A schematic representation of the C-TALE procedure. For the experimental details, see “Methods” and Additional file 8: Supplementary Methods. b Heatmaps showing C-TALE data normalized by the total number of sequencing reads in the C-TALE dataset, binned at a 5 Kb resolution and iteratively corrected. Dashed lines show the positions of chromatin contact domain boundaries identified using the Lavaburst package. Gray triangles in the left panel represent 5C-identified TADs in DT40 cell line. Gene positions are shown on the diagonal of each panel, and a detailed map of the studied region is presented below the heatmaps. Dark triangles inside the gene boxes show the direction of transcription

Fig. 5
Fig. 5

Activation of alpha-globin gene transcription is accompanied by local changes in chromatin interaction profile. a C-TALE heatmaps showing local chromatin interaction profile in a close vicinity to AgGD (region chr14:11957600-12236390). Black star denotes the boundary between T1 and T2 TADs. All other notations are as in Fig. 4b. b C-TALE heatmap showing interactions enriched in differentiated HD3 cells as compared to DT40 cells (right part of the map) and to proliferating HD3 cells (left part of the map). The positions of FISH probes are shown with magenta and green rectangles. Regions of the map corresponding to interactions between FISH probes are encircled with dashed squares. c Schematic representation C-TALE-identified sub-TADs within T1 and T2 TADs. d The distributions of C-TALE counts corresponding to chromatin interactions inside the sub-TAD harboring alpha-globin gene cluster. Thick black lines represent median values. Asterisks represent a significant difference with a P-value <10−4 (one-tailed Wilcoxon’s signed-rank test). e The distributions of C-TALE counts corresponding to chromatin interactions between genomic bins, separated by the distance more than 60 Kb throughout the shown region without including the sub-TAD harboring alpha-globin gene cluster. All notations are as in panel (d). f Scale-plots showing the dependency of contact probability on genomic distance within the shown region. g Violin plots showing the distributions of spatial distance between FISH probes. White dots represent the medians. Asterisks represent a significant difference with a P-value <0.01 (two asterisks), and with a P-value <0.001 (three asterisks) (Mann–Whitney rank-sum test). Representative examples of FISH images are shown on the right. Scale bar 1 µm

Fig. 6
Fig. 6

DpnII 3C-analysis of chromatin contacts of the alpha-globin MRE and nearby CTCF-occupied region −3.5 CBSs. a Map showing the positions of genes and CTCF ChIP-seq peaks. Test and anchor regions are outlined with gray and pink vertical rectangles, respectively. CBSs are designated according to their distance (in kilobases) from the MRE. b 3C interaction profile of the −3.5 CBSs. Ligation frequencies averaged between biological replicates are shown. Error bars represent SEM. Anchor position relative to positions of test regions is outlined with vertical dashed line. Horizontal gray line represent a relative noise level ±SEM (see “Methods”). A closer view of the CTCF ChIP-seq profile of −3.5 CBS is shown above the diagram. c 3C interaction profile of the MRE. All notations are as in panel (b). d 3C interaction profile of the +46 CBS. All notations are as in panel (b). e Schematic representation of the CTCF-anchored loops observed around AgGD

Similar articles

Cited by

References

    1. Ulianov SV, Gavrilov AA, Razin SV. Nuclear compartments, genome folding, and enhancer–promoter communication. Int Rev Cell Mol Biol. 2015;315:183–244. doi: 10.1016/bs.ircmb.2014.11.004. - DOI - PubMed
    1. Vernimmen D, Bickmore WA. The hierarchy of transcriptional activation: from enhancer to promoter. Trends Genet. 2015;31:696–708. doi: 10.1016/j.tig.2015.10.004. - DOI - PubMed
    1. Dekker J, Mirny L. The 3D genome as moderator of chromosomal communication. Cell. 2016;164:1110–1121. doi: 10.1016/j.cell.2016.02.007. - DOI - PMC - PubMed
    1. Rao SS, Huntley MH, Durand NC, Stamenova EK, Bochkov ID, Robinson JT, et al. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell. 2014;159:1665–1680. doi: 10.1016/j.cell.2014.11.021. - DOI - PMC - PubMed
    1. Mifsud B, Tavares-Cadete F, Young AN, Sugar R, Schoenfelder S, Ferreira L, et al. Mapping long-range promoter contacts in human cells with high-resolution capture Hi-C. Nat Genet. 2015;47:598–606. doi: 10.1038/ng.3286. - DOI - PubMed

Publication types

MeSH terms

Substances