Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Filters applied. Clear all
. 2016 Mar 7;36(5):572-87.
doi: 10.1016/j.devcel.2016.01.024. Epub 2016 Feb 25.

Dynamic Gene Regulatory Networks Drive Hematopoietic Specification and Differentiation

Affiliations
Free PMC article

Dynamic Gene Regulatory Networks Drive Hematopoietic Specification and Differentiation

Debbie K Goode et al. Dev Cell. .
Free PMC article

Abstract

Metazoan development involves the successive activation and silencing of specific gene expression programs and is driven by tissue-specific transcription factors programming the chromatin landscape. To understand how this process executes an entire developmental pathway, we generated global gene expression, chromatin accessibility, histone modification, and transcription factor binding data from purified embryonic stem cell-derived cells representing six sequential stages of hematopoietic specification and differentiation. Our data reveal the nature of regulatory elements driving differential gene expression and inform how transcription factor binding impacts on promoter activity. We present a dynamic core regulatory network model for hematopoietic specification and demonstrate its utility for the design of reprogramming experiments. Functional studies motivated by our genome-wide data uncovered a stage-specific role for TEAD/YAP factors in mammalian hematopoietic specification. Our study presents a powerful resource for studying hematopoiesis and demonstrates how such data advance our understanding of mammalian development.

Figures

None
Figure 1
Figure 1
Integrated Global Data over a Whole Developmental Pathway (A) UCSC browser screenshot depicting the Tal1 locus aligning RNA-seq, DNaseI-seq, and ChIP-seq data from the six stages of development depicted in the left-hand flow chart. The stage-specific color scheme is used in all subsequent figures. Panels display ChIP-seq data for four histone modifications (left) and 16 different TFs (right) plus DHS data. The grayed-out regions indicate known regulatory regions: from left to right, promoters 1a and 1b, enhancers +19 and +40. (B–D) Hierarchical clustering of cell populations based on the normalized expression values of the genes (B), normalized correlation among the DHS sites (C), and correlation among the TF sites (D). The correlations were normalized between −1 and +1 to preserve the color scale. ESC, embryonic stem cell; HB, hemangioblast; HE, hemogenic endothelium; HP, hematopoietic progenitors, MES, mesoderm. (E) Functional enrichment for genes that are differentially regulated during developmental transitions (T1–T5) in the progression of hematopoietic commitment. (F) The expression dynamics of the differentially expressed genes in the pathway given in (A) that are clustered into 31 patterns. The standardized expression values (zij) of the differentially regulated genes in the developmental pathway (Figure S1E) were clustered into 31 expression patterns, and the plot shows the expression profiles of these patterns. The methodology is detailed in Supplemental Experimental Procedures.
Figure 2
Figure 2
Chromatin Programming during Progressive Lineage Commitment (A) Schematic diagram of the method used to coarse grain the 23-state chromatin model to four potential chromatin states. (B) Clustering of promoters (1Kb up or downstream of the transcription start site, TSS) based on their chromatin state patterns (left) and the clustering of the expression pattern of genes that are constitutively expressed (right). (C and D) Integration of DHS pattern and TF binding. (C) Methodology of the integrative analysis of chromatin dynamics and TF binding events. (D) TF binding events and the p values denoting the significance of overlap are depicted as gray-scale density plots, shown as dots. Integration of DHS (rows) and TF binding (columns) patterns across the six stages with the population size of each DHS pattern given on the right-hand side. For significance calculations only DHS patterns with a population size >100 were considered.
Figure 3
Figure 3
Integration of Chromatin Dynamics, TF Binding Events, and Gene Expression during Hematopoietic Specification (Left) Flow diagram of data integration. The average expression values (log10(FPKM)) of genes in expression patterns E1 to E31 were calculated for each developmental stage. The significance (p < 0.0001) of the overlap between the genes in each expression pattern and a given TF ChIP-seq peak set was obtained using gene-set control analysis. Z scores were obtained from the mean enrichment of H3K27ac in TF binding sites at these loci using bootstrapping. (Right) Average expression levels for each pattern (rows) are shown as a red-blue heatmap (see key), with columns for each cell type labeled at the top. The columns are further divided into TF ChIP-seq experiments, and the significant overlap between TF binding events and gene sets belonging to each expression pattern are depicted by gray-scale density plots shown as dots (see key). Significant overlap of these binding events and H3K27ac sites are also shown as a density plot depicted by yellow-green boxes (see key).
Figure 4
Figure 4
Dynamic Gene Regulatory Network Driving Hematopoietic Specification For each developmental stage the 16 TFs used in ChIP-seq experiments are shown as nodes in a GRN. The color of each node corresponds to the level of gene expression (see key). The chromatin accessibility at each promoter is shown as open/circular (DHS presence) or closed/octagonal (DHS absence), and the border color of the node corresponds to the promoter state according to the coarse-grain four-state model mentioned in Figure 2 (see key). Arrows indicate binding events of a TF (source) at loci encoding all TFs (target). The arrow color relates to the promoter state of the target TF encoding gene. No emanating arrow from a node indicates absence of ChIP data. For information about which ChIP experiments were conducted in which cell type, see Figure S1B. Note that we did not include C/EBPβ binding in the ES cell GRN, since the publicly available datasets did not contain the respective data.
Figure 5
Figure 5
Gene Regulatory Network Analysis Is Informative for Reprogramming Success (A) Schematic of reprogramming experiments. (B) Bar chart showing the number of hematopoietic colonies generated from fibroblast cells after overexpression of different combinations of TAL1, LMO2, FLI1 and GATA2 as indicated in the table below. Grayed areas in this table highlight successful production of hematopoietic colonies. The number of hematopoietic colonies generated from fibroblast cells after overexpression of Tal1, Lmo2, Fli1, Gata2, and the indicated combinations of these four TF encoding genes. Data presented are mean ± SEM of individual experiments (n = 7). (C) Correlation coefficient analyses of gene expression profiles generated by RNA-seq from hematopoietic cells generated by reprogramming from MEFs at day 12 and day 21 (D12 and D21) of the experiment with gene expression patterns generated from in vitro differentiated cells. The heatmap shows the correlation between expression data from each stage (columns) and each experiment (rows). All red/pink rectangles show correlation values that are significantly different from all blue rectangles (Z transform test, p < 0.01). (D) Model depicting the hierarchy of transcriptional regulation during hematopoietic differentiation. Target TF genes are listed in the arrow, with those that are upregulated after binding events listed on the left and those that are downregulated listed on the right.
Figure 6
Figure 6
Chromatin and Transcription Factor Binding Dynamics and Identification of Regulators (A) Schematic representation of the methodology for pairwise motif clustering at distal DHS (for used motifs see Table S6). (B) Relative motif enrichment (RE) scores for the motifs in DHS unique to a given cell type compared with each of the other five cell types were clustered. RUNX and TEAD motifs are indicated by blue and red arrows, respectively. The significance of the RE scores were computed using the bootstrapping method (Table S7). (C) Significance of co-localization of TEAD motifs in TF ChIP peaks. Red stars indicate significant co-localization.
Figure 7
Figure 7
A Role for TEAD Factors in Early Hematopoietic Specification (A) TEAD and YAP localize to the nucleus of a subset of TIE2+ endothelium within yolk sac blood island of E7.5 embryos. E7.5 embryo sections were stained as indicated. Asterisks mark TIE2+ cells on the outer edge of the developing blood island, which show nuclear localization for both TEAD and YAP. Cells within the blood island are maturing primitive erythrocytes and do not show nuclear localization for either TEAD or YAP. (B and C) TEAD activity is required during the early phase of hematopoietic commitment. The TEAD-YAP inhibitor verteporfin (9.6 μM) or DMSO vehicle was added on day 1, 2, 3, or 4 of EB culture. Day-1 EB corresponds to mesoderm (MES) commitment, day 2–3 to hemangioblast (HB), commitment and day 4–5 to HE and HP specification. The frequency of CD41+ hematopoietic cells was determined on day 7. (B) Representative FACS plots. (C) Quantification of the percentage of CD41+ cells from n ≥ 3 independent experiments. Data presented are mean ± SEM, paired t test. (D) Genome browser screenshot showing TEAD4 binding to the Tal1 locus in HB together with other chromatin and binding features. (E) Genomic distribution of TEAD4 peaks together with TF binding motifs enriched in distal (left) and proximal (right) peaks.

Similar articles

See all similar articles

Cited by 57 articles

See all "Cited by" articles

References

    1. Barozzi I., Simonatto M., Bonifacio S., Yang L., Rohs R., Ghisletti S., Natoli G. Coregulation of transcription factor binding and nucleosome occupancy through DNA features of mammalian enhancers. Mol. Cell. 2014;54:844–857. - PMC - PubMed
    1. Batta K., Florkowska M., Kouskoff V., Lacaud G. Direct reprogramming of murine fibroblasts to hematopoietic progenitor cells. Cell Rep. 2014;9:1871–1884. - PMC - PubMed
    1. Beyer T.A., Weiss A., Khomchuk Y., Huang K., Ogunjimi A.A., Varelas X., Wrana J.L. Switch enhancers interpret TGF-beta and Hippo signaling to control cell fate in human embryonic stem cells. Cell Rep. 2013;5:1611–1624. - PubMed
    1. Bonifer C., Hoogenkamp M., Krysinska H., Tagoh H. How transcription factors program chromatin—lessons from studies of the regulation of myeloid-specific genes. Semin. Immunol. 2008;20:257–263. - PubMed
    1. Chen X., Xu H., Yuan P., Fang F., Huss M., Vega V.B., Wong E., Orlov Y.L., Zhang W., Jiang J. Integration of external signaling pathways with the core transcriptional network in embryonic stem cells. Cell. 2008;133:1106–1117. - PubMed

Publication types

MeSH terms

Feedback