Single-cell transcriptomics reveals a new dynamical function of transcription factors during embryonic hematopoiesis

  1. Isabelle Bergiers  Is a corresponding author
  2. Tallulah Andrews
  3. Özge Vargel Bölükbaşı
  4. Andreas Buness  Is a corresponding author
  5. Ewa Janosz
  6. Natalia Lopez-Anguita  Is a corresponding author
  7. Kerstin Ganter  Is a corresponding author
  8. Kinga Kosim
  9. Cemre Celen
  10. Gülce Itır Perçin
  11. Paul Collier  Is a corresponding author
  12. Bianka Baying  Is a corresponding author
  13. Vladimir Benes  Is a corresponding author
  14. Martin Hemberg  Is a corresponding author
  15. Christophe Lancrin  Is a corresponding author
  1. European Molecular Biology Laboratory, EMBL Rome, Italy
  2. Wellcome Trust Sanger Institute, United Kingdom
  3. European Molecular Biology Laboratory, Germany
15 figures and 12 additional files

Figures

Figure 1 with 6 supplements
Analysis of in vivo and in vitro EHT by single-cell qRT-PCR identified the expression of key TFs at the Pre-HSPC stage.

(A) Principal component analysis (PCA) plot showing the cells isolated at E9, E10.5 and E11 from AGM and YS according to the indicated sample clusters (SC). For each tissue, each time point …

https://doi.org/10.7554/eLife.29312.002
Figure 1—figure supplement 1
Hierarchical clustering analysis of AGM VE-Cad+ cells.

(A) FACS plots of VE-Cad and CD41 expression in the AGM region at E9, E10.5 and E11. Several embryos (seven for E9, 13 for E10.5 and 9 for E11) were pooled for each time point and three …

https://doi.org/10.7554/eLife.29312.003
Figure 1—figure supplement 2
Hierarchical clustering analysis of YS VE-Cad+ cells.

(A) FACS plots of VE-Cad and CD41 expression in the YS region at E9, E10.5 and E11. Several embryos (seven for E9, 13 for E10.5 and 9 for E11) were pooled for each time point and three …

https://doi.org/10.7554/eLife.29312.004
Figure 1—figure supplement 3
Time-course analysis of EHT in the AGM and YS.

Scatter plots showing the average expression of endothelial genes versus the average expression of hematopoietic genes at the indicated developmental time points.

See also Supplementary file 11.

https://doi.org/10.7554/eLife.29312.005
Figure 1—figure supplement 4
Comparison of EHT subpopulations from AGM and YS.

(A) Correlation study performed on the average gene expression for the groups highlighted in Figure 1A, Figure 1—figure supplements 1B and 2B. (B) Hierarchical clustering analysis done with …

https://doi.org/10.7554/eLife.29312.006
Figure 1—figure supplement 5
Hierarchical clustering analysis of YS Gfi1-/- Gfi1b-/- cells.

(A) FACS plot of GFP expression in the Gfi1–/– Gfi1b–/– yolk sac region at E9.5. Single cells expressing highly GFP were isolated and studied with sc-q-RT-PCR. (B) Hierarchical clustering analysis …

https://doi.org/10.7554/eLife.29312.007
Figure 1—figure supplement 6
Hierarchical clustering analysis of in vitro differentiated ESCs.

(A) Hierarchical clustering analysis done with sc-q-RT-PCR data from in vitro differentiated ESCs. Four main clusters were defined. (B) Violin plots showing the expression of the 24 genes shown in Fi…

https://doi.org/10.7554/eLife.29312.008
Generation of a doxycycline-inducible ESC line for the simultaneous expression of the eight transcription factors.

(A) Scheme showing the generation of the i8TFs ESC line. (B) Western blot showing the protein expression of the eight TFs after doxycycline (dox) treatment in the Empty line and in the i8TFs ESC …

https://doi.org/10.7554/eLife.29312.009
Figure 3 with 5 supplements
Hematopoietic differentiation of the i8TFs ESC line highlights the role of the eight transcription factors in the generation of Pre-HSPCs.

(A) Representative FACS plots of VE-Cad, cKit and CD41 expression after 3 days of BL-CFC culture of the indicated ESC lines (n = 3). (B) Graphs showing the average numbers of round cells counted per …

https://doi.org/10.7554/eLife.29312.010
Figure 3—figure supplement 1
Microscopy pictures of the i8TFs BL-CFC culture after one, two and three days of BL-CFC culture in -dox and +dox conditions.

The scale bar corresponds to 200 μm. See also Supplementary file 11.

https://doi.org/10.7554/eLife.29312.011
Figure 3—figure supplement 2
Comparison of eVSM and Endo populations response to the induction of the 8TFs.

(A) FACS plot of VE-Cad and CD41 expression at day 1 of BL-CFC culture. The rectangles highlight the enriched vascular smooth muscle (eVSM) and endothelial (Endo) populations. (B) Hierarchical …

https://doi.org/10.7554/eLife.29312.012
Figure 3—figure supplement 3
PCA plots of microarrays data.

(A) PCA plot showing the microarray analysis results from the eVSM after 24 hr of dox treatment on the Empty cell line. Ellipses indicate the different groups of samples. (B) PCA plot showing the …

https://doi.org/10.7554/eLife.29312.013
Figure 3—figure supplement 4
Analysis of microarrays data.

(A) GO term enrichment analysis of the genes differentially expressed between –dox and +dox conditions for each population. (B) Venn diagrams showing the genes differentially expressed for both …

https://doi.org/10.7554/eLife.29312.014
Figure 3—figure supplement 5
Analysis of the i1TF ESC lines.

(A) Scheme showing the eight constructs used to make the i1TF ESC lines. (B) Western blots showing the protein expression of the eight transcription factors after doxycycline treatment for each of …

https://doi.org/10.7554/eLife.29312.015
Figure 4 with 2 supplements
Single-cell transcriptomics analysis of i8TFs and i1TF ESC lines.

(A) Hierarchical clustering showing the sc-q-RT-PCR results for 95 genes on the ten inducible ESC lines after three days of BL-CFC culture. The clusters were defined according to the intersection of …

https://doi.org/10.7554/eLife.29312.016
Figure 4—figure supplement 1
Six scatter plots are displayed corresponding to the six populations defined in Figure 4.

The data set is identical to the one in Figure 4. Here, fractions of cells are shown instead of absolute cell numbers. For the –dox condition, a box plot has been added. Significant differences …

https://doi.org/10.7554/eLife.29312.017
Figure 4—figure supplement 2
Comparison of Empty and i8TFs day 3 BL-CFC cultures by single-cell RNA sequencing.

(A) Experimental outline followed to compare the Empty and i8TFs ESCs with single-cell RNA sequencing. (B) PCA plot showing sc-RNA-seq analysis of the Empty and i8TFs cell lines after three days of …

https://doi.org/10.7554/eLife.29312.018
Figure 5 with 2 supplements
The eight TFs induction gives rise to a cell population similar to Pre-HSCs type I.

(A) Representative FACS plots of VE-Cad, CD45, CD43 and CD41 expression after 3 days of BL-CFC culture of the i8TFs ESC line in the presence or absence of dox (n = 3). (B) Bar graphs showing the …

https://doi.org/10.7554/eLife.29312.019
Figure 5—figure supplement 1
Hierarchical clustering analysis of Pro-HSCs and Pre-HSCs type I.

(A) FACS plots of VE-Cad, CD45, CD43 and CD41 expression in the AGM region at E10 (31–32 somite pairs). Single cells from Pro-HSCs (VE-Cad+ CD41+ CD45- CD43-) and Pre-HSCs type I (VE-Cad+ CD41+ CD45-

https://doi.org/10.7554/eLife.29312.020
Figure 5—figure supplement 2
PCA plots comparing the single-cell-q-RT-PCR data of Figures 1, 4 and 5.

(A) PCA plot displaying the single-cell qPCR results of all cells from Figures 1, 4 and 5 (a total of 1,660 cells). (B) Same PCA plot as (A) but highlighting the populations from Figure 1B. (C) Same …

https://doi.org/10.7554/eLife.29312.021
Figure 6 with 1 supplement
Single-cell transcriptomics analysis reveals the transcriptional networks involved in Pre-HSPCs generation.

(A) Network built from gene correlations. Gpr56 is highlighted with a red asterisk. (B) Centrality values for all the genes containing a heptad peak within 1 kb of the TSS that are significantly …

https://doi.org/10.7554/eLife.29312.022
Figure 6—figure supplement 1
Single-cell transcriptome analysis following over-expression of the 8TFs in eVSM cells.

(A) Experimental workflow used to differentiate in vitro eVSM cells from the i8TFs ESC line and to analyze them by sc-RNA-seq after 24 hr of treatment. (B) Scatter plot showing the expression of the …

https://doi.org/10.7554/eLife.29312.023
Figure 7 with 1 supplement
Single-cell transcriptome analysis suggests that Runx1 and Fli1 have opposite functions during EHT.

(A) Network built from gene correlations found in the Peirera dataset (Pereira et al., 2016). (B) Heatmap displaying Spearman correlations between the 210 genes found to be correlated to Runx1 and …

https://doi.org/10.7554/eLife.29312.024
Figure 7—figure supplement 1
Bioinformatic analysis of the Pereira dataset.

(A) GO term enrichment analysis of the genes correlated with Fli1 and Runx1. The color code matches the gene clusters from Figure 7B. (B) Heatmap displaying hierarchical clustering of the 210 genes …

https://doi.org/10.7554/eLife.29312.025
Figure 8 with 1 supplement
The Erg/Fli1 and Runx1/Gata2 groups of transcription factors have opposite functions during EHT.

(A) Scheme showing the i5TFs and i6TFs constructs used to generate the two inducible ESC lines missing Runx1, Gata2 and Cbfb, or Fli1 and Erg (respectively). (B) Representative FACS plots of VE-Cad, …

https://doi.org/10.7554/eLife.29312.026
Figure 8—figure supplement 1
Study of the role of Erg/Fli1 and Runx1/Gata2 groups of transcription factors in EHT.

(A) Graphs showing the average numbers of round cells counted per frame (n = 3) in a 48 hr time course of HE culture for the indicated cell lines. Error bars represent standard deviations. (B) …

https://doi.org/10.7554/eLife.29312.027
Model for the dynamical function of transcription factors during the formation of blood stem and progenitor cells.

Proposed model for the dynamical function of transcription factors during the endothelial to hematopoietic transition. The upper blue arrow section shows the balance between transcription factors …

https://doi.org/10.7554/eLife.29312.028
Author response image 4
(Left) 8TFs network in our sc-RNA-Seq data.

(Right) 8TFs network in Pereira sc-RNA-Seq data.

https://doi.org/10.7554/eLife.29312.047

Additional files

Supplementary file 1

(Related to Figures 1, 3, 4, 5 and 8.) Description of the genes used for single-cell quantitative RT-PCR.

The genes are classified by categories. A gene can belong to several categories.

https://doi.org/10.7554/eLife.29312.031
Supplementary file 2

(Related to Figure 1.) Single-cell quantitative RT-PCR data shown in Figure 1A and B.

The first worksheet contains Ct values, the second log2 expression data and the third the metadata relative to the cells shown in Figure 1A and B.

https://doi.org/10.7554/eLife.29312.032
Supplementary file 3

(Related to Figure 1D.) Single-cell quantitative RT-PCR data shown in Figure 1D.

The first worksheet contains Ct values, the second log2 expression data and the third the metadata relative to the cells shown in Figure 1D.

https://doi.org/10.7554/eLife.29312.033
Supplementary file 4

(Related to Figure 3—figure supplement 2B.) Single-cell quantitative RT-PCR data shown in Figure 3—figure supplement 2B.

The first worksheet contains Ct values, the second log2 expression data and the third the metadata relative to the cells shown in the Figure 3—figure supplement 2.

https://doi.org/10.7554/eLife.29312.034
Supplementary file 5

(Related to Figure 3—figure supplements 3B and 4A,) Microarray data results shown in Figure 3—figure supplement 3B.

The first worksheet contains metadata, the nine other sheets contain the results of expression contrast between two populations indicated in the name of the corresponding worksheet. The worksheets 10–14 contain information for the GO term analysis presented in Figure 3—figure supplement 4A.

https://doi.org/10.7554/eLife.29312.035
Supplementary file 6

(Related to Figure 4.) Single-cell quantitative RT-PCR data shown in Figure 4A and B.

The first worksheet contains Ct values, the second log2 expression data, the third the metadata relative to the cells shown in Figure 4A and B, the fourth the matrix for Figure 4B and the fifth the p-values for Figure 4B.

https://doi.org/10.7554/eLife.29312.036
Supplementary file 7

(Related to Figure 5.) Single-cell quantitative RT-PCR data shown in Figure 5C.

The first worksheet contains Ct values, the second log2 expression data and the third the metadata relative to the cells shown in Figure 5C.

https://doi.org/10.7554/eLife.29312.037
Supplementary file 8

(Related to Figure 5—figure supplement 2.) Single-cell quantitative RT-PCR data shown in Figure 5—figure supplement 2.

The first worksheet contains Ct values, the second log2 expression data and the third the metadata relative to the cells shown in Figure 5—figure supplement 2.

https://doi.org/10.7554/eLife.29312.038
Supplementary file 9

(Related to Figure 7.) Results of the Spearman correlation analysis shown in Figure 7B.

The first worksheet contains all correlation values shown in the heatmap of Figure 7B. The second contains the list of Fli1 target genes defined by Wilson et al. (2010). The third contains the list of Runx1 target genes defined by Wilson et al. (2010). The fourth contains the name of Fli1 targets among the 210 genes list and their corresponding Spearman correlation values. The fifth contains the name of Runx1 targets among the 210 genes list and their corresponding Spearman correlation values. The sixth worksheet contains the name of both Fli1 and Runx1 targets among the 210 genes list and their corresponding Spearman correlation values.

https://doi.org/10.7554/eLife.29312.039
Supplementary file 10

(Related to Figure 8 and Figure 8—figure supplement 1.) Single-cell quantitative RT-PCR data shown in Figure 8E and Figure 8—figure supplement 1B.

The first worksheet contains Ct values, the second log2 expression data and the third the metadata relative to the cells shown in Figure 8E.

https://doi.org/10.7554/eLife.29312.040
Supplementary file 11

(Related to all Figures except Figure 7 and Figure 7—figure supplement 1.) Table summarizing the experiments done in the manuscript.

For each figure panel, the type of experiments is described as well as the number of times they were carried out.

https://doi.org/10.7554/eLife.29312.041
Transparent reporting form
https://doi.org/10.7554/eLife.29312.042

Download links