Stem Cell Reports Article GATA2 Promotes Hematopoietic Development and Represses Cardiac Differentiation of Human Mesoderm Julio Castaño,1,2 Sergi Aranda,3 Clara Bueno,4 Fernando J. Calero-Nieto,5 Eva Mejia-Ramirez,1 Jose Luis Mosquera,6 Enrique Blanco,3 Xiaonan Wang,5 Cristina Prieto,4 Lorea Zabaleta,7 Elisabetta Mereu,8 Meritxell Rovira,1 Senda Jiménez-Delgado,1 Daniel R. Matson,9 Holger Heyn,8,10 Emery H. Bresnick,9 Berthold Göttgens,4 Luciano Di Croce,3,11 Pablo Menendez,4,11,12 Angel Raya,1,2,11 and Alessandra Giorgetti1,* 1Center of Regenerative Medicine in Barcelona (CMRB), Hospital Duran i Reynals, Gran Via de L’Hospitalet, 199-203, Hospitalet de Llobregat, Barcelona 08908, Spain 2Center for Networked Biomedical Research on Bioengineering, Biomaterials, and Nanomedicine (CIBER-BBN), Madrid 28029, Spain 3Center for Genomic Regulation (CRG), The Barcelona Institute of Science and Technology, Universitat Pompeu Fabra, Barcelona 08003, Spain 4Josep Carreras Leukemia Research Institute and Department of Biomedicine, School of Medicine, University of Barcelona, Barcelona 08036, Spain 5Department of Hematology, Wellcome and MRC Cambridge Stem Cell Institute and Cambridge Institute for Medical Research, University of Cambridge, Cambridge, UK 6Bioinformatics Unit, Bellvitge Biomedical Research Institute (IDIBELL), Barcelona, 08908 Spain 7Laboratory of Hematological Diseases, Fundación Inbiomed, San Sebastian, 20009, Spain 8CNAG-CRG, Center for Genomic Regulation (CRG), Barcelona Institute of Science and Technology (BIST), Barcelona, Spain 9Department of Cell and Regenerative Biology, UW-Madison Blood Research Program, Carbone Cancer Center, University of Wisconsin School of Medicine and Public Health, Madison, WI 53705, USA 10Universitat Pompeu Fabra, Barcelona, Spain 11Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona 08010, Spain 12Centro de Investigación Biomedica en Red en Cancer (CIBERONIC) ISCIII, Barcelona, Spain *Correspondence: agiorgetti@cmrb.eu https://doi.org/10.1016/j.stemcr.2019.07.009 SUMMARY In vertebrates, GATA2 is a master regulator of hematopoiesis and is expressed throughout embryo development and in adult life. Although the essential role of GATA2 in mouse hematopoiesis is well established, its involvement during early human hematopoietic development is not clear. By combining time-controlled overexpression of GATA2 with genetic knockout experiments, we found that GATA2, at the mesoderm specification stage, promotes the generation of hemogenic endothelial progenitors and their further differen- tiation to hematopoietic progenitor cells, and negatively regulates cardiac differentiation. Surprisingly, genome-wide transcriptional and chromatin immunoprecipitation analysis showed that GATA2 bound to regulatory regions, and repressed the expression of cardiac devel- opment-related genes. Moreover, genes important for hematopoietic differentiation were upregulated by GATA2 in a mostly indirect manner. Collectively, our data reveal a hitherto unrecognized role of GATA2 as a repressor of cardiac fates, and highlight the importance of coordinating the specification and repression of alternative cell fates. INTRODUCTION gorized as ‘‘hematopoietic’’ GATA factors and regulate the development of diverse hematopoietic lineages (Bresnick During embryonic development, hematopoietic stem cells et al., 2012; Johnson et al., 2012; Katsumura et al., 2017; (HSCs) emerge from hemogenic endothelium in the Orkin, 1992). The importance of GATA2 in HSC specifica- ventral wall of the aorta-gonad-mesonephros (AGM) re- tion was first highlighted by gene targeting studies, because gion (Dzierzak and Speck, 2008; Ivanovs et al., 2011, ablation of Gata2 is embryonic lethal at embryonic day (E) 2014), and their specification is tightly orchestrated by 10.5 due to the collapse of primitive and definitive hemato- temporal changes in the expression of master regulators poiesis (Gao et al., 2013; Ling et al., 2004; Tsai and Orkin, during endothelial-to-hematopoietic transition (EHT). 1997). Notably, analysis of chimeric embryos generated While animal models have been crucial in identifying with Gata2-null embryonic stem cells (ESCs) indicated several master regulators, such as GATA2, RUNX1, and that these cells failed to contribute to any hematopoietic TAL1 (Robert-Moreno et al., 2005; Wilson et al., 2010a, lineage (Tsai et al., 1994). Likewise, mouse Gata2-null 2010b), how these factors drive human HSC emergence endothelial cells failed to produce HSCs because of during EHT remains poorly understood. impaired EHT (de Pater et al., 2013; Gao et al., 2013; John- GATA2 belongs to an evolutionarily conserved family of son et al., 2012; Lim et al., 2012). A primary role of GATA2 zinc finger transcription factors comprising six members: in promoting EHT has been recently demonstrated in hu- GATA1 to GATA6 (Merika and Orkin, 1993; Molkentin, mans (Gomes et al., 2018; Kang et al., 2018; Zhou et al., 2000). GATA2, together with GATA1 and GATA3, are cate- 2019). Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 j ª 2019 The Author(s). 515 This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). In adult hematopoiesis, GATA2 is expressed at high levels We next established transgenic hiPSCs in which the in HSCs, early hematopoietic progenitors, and in expression of transgenic GATA2 could be temporally erythroid/megakaryocyte lineages (Vicente et al., 2012). controlled by doxycycline (Dox) administration (hereafter Recent studies showed that GATA2 haploinsufficiency is termed iGATA2-hiPSCs) (Figure S1A). Robust transgenic associated with some familial cases of myelodysplastic syn- overexpression of GATA2 was confirmed in four clones drome, bone marrow failure, immunodeficiency, and (CL6, CL9, CL201, CL204) derived from two independent MonoMAc syndrome (Dickinson et al., 2011; Hahn et al., iGATA2-hiPSC lines by western blotting after 2 days of 2011; Wlodarski et al., 2016), further supporting its impor- Dox treatment (Figure S1B). qRT-PCR analysis and in vivo tant role in HSCs. Conversely, enforced expression of functional assays showed that iGATA2-hiPSCs retained GATA2 in cord blood-derived HSCs confers increased quies- the expression of pluripotency markers and also the capac- cence, an important hallmark of HSCs (Tipping et al., ity to generate teratomas (Figure S1C). 2009). Then, considering the expression of endogenous GATA2, We sought to explore the role of GATA2 during human we induced GATA2 expression from day 2 to 7 in EBs hematopoietic development by inducing GATA2 expres- generated from iGATA2-hiPSCs (Figures 1A and S1D– sion in differentiating human induced pluripotent stem S1G). Flow cytometry analysis showed that enforced cells (hiPSCs) (Takahashi et al., 2007). We show that expression of GATA2 significantly enhanced the produc- GATA2 induction during mesoderm patterning robustly tion of HEPs (2.5-fold increase of CD31+CD34+CD45 promotes the generation of hemogenic endothelial pro- cells and 2-fold increase of CD34+CD43–CD45– cells) in genitors (HEPs), and their further differentiation into he- EBs at day 10 (Figures 1D and 1E), and promoted the gener- matopoietic progenitor cells (HPCs). Global transcriptome ation of HPCs (5-fold increase of CD34+CD43+CD45+ analysis and chromatin immunoprecipitation sequencing cells) at day 15 (Figures 1D and 1E). (ChIP-seq) combined with DNA massive sequencing re- We used colony-forming unit (CFU) assays to confirm vealed that GATA2 directly represses genes that promote that GATA2 overexpression promotes hematopoiesis from cardiac cell fate differentiation and activates master he- iGATA2-hiPSCs. Dox treatment (days 2–7) significantly matopoietic regulators via direct and indirect mechanisms. increased the total number of hematopoietic CFCs in day Remarkably, GATA2 knockout impaired hematopoietic 10 EBs (Figure 2A). Notably, CFU scoring revealed an development and enhanced cardiac potential of meso- enhancement in all types of hematopoietic colonies (Fig- dermal progenitors. ure 2A), suggesting that GATA2 expression promotes he- matopoietic commitment by inducing mesodermal specifi- cation to HEPs at very early stages. RESULTS To better understand the role of GATA2 in early hemato- poiesis, we treated iGATA2-hiPSCs with Dox at distinct GATA2 Promotes Robust Hematopoietic stages of EB development (days 2–7, 7–15, and 10–15). In- Differentiation duction of GATA2 at days 2–7 had the greatest effect on To analyze the impact of GATA2 in early human hemato- HEP and HPC generation (Figure 2B), while treatment at poiesis, we first examined endogenous GATA2 expression days 7–15 had no significant effects on the HEP popula- in hiPSCs induced to form embryoid bodies (EBs) in tion, and only a small effect was observed for early HPCs. serum-free medium with the successive addition of BMP4 By contrast, GATA2 overexpression at days 10–15 led to a (days 0–3), CHIR92001 (days 2–3), and hematopoietic cyto- significant decrease in HPC numbers (Figure 2B), consis- kines (days 3–15) (Figure 1A). This protocol promotes tent with previous findings showing that high GATA2 mesoderm induction (days 2–3), specification of meso- expression in HSCs blocks normal hematopoiesis (Persons dermal cells to bipotential hemato-endothelial progenitors et al., 1999; Tipping et al., 2009). (CD31+CD34+CD43-CD45; days 3–10) that can originate To address whether the higher hematopoietic output both endothelial and hematopoietic cells and could be could be a consequence of a higher HEP generation, considered equivalent to HEPs (Ayllon et al., 2015), CD34+ cells were purified from day 10 EBs with or without and further commitment of HEPs to definitive HPCs Dox treatment, and were cultured on OP9 stromal cells (CD34+CD43+CD45+; days 10–15) (Giorgetti et al., 2017; (Choi et al., 2009; Ramos-Mejia et al., 2014) (Figure 2C, up- Sturgeon et al., 2014). GATA2 was initially expressed at per panel). After 4 days of coculture, both CD34+CD45+ and day 2 (Figure 1B), at the onset of mesoderm formation total CD45+ cell subpopulations increased significantly in marked by the expression of T and MIXL1 (Figure 1C). Its the cocultures derived from Dox-treated cells (3- and expression then progressively increased along with the 4-fold, respectively), whereas the number of HEPs decreased emergence of HEPs and HPCs, in parallel with the master slightly (Figure 2C, lower panel). To further characterize hemogenic regulators RUNX1 and SCL (Figure 1B). the effect of GATA2 during mesoderm patterning, we 516 Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 Figure 1. Early GATA2 Induction Enhances Hematopoietic Development from hiPSCs (A) hiPSC hematopoietic differentiation based on EB generation. (B) Time course of endogenous GATA2, SCL/TAL1, and RUNX1 expression during EB development, normalized to GAPDH. (C) Time course of endogenous mesodermal marker expression (BRACHYURY and MIXL1) during EB development, normalized to GAPDH. In (B) and (C), data represent the mean ± SD of three independent experiments. (D) Representative flow cytometry analysis of HEPs (CD31+CD34+CD45– and CD34+CD43–CD45–) and HPCs (CD34+CD45+ and CD43+CD45+) in EBs at days 10 and 15 in control and Dox-treated cells. (E) Quantitative summary of HEP and HPC analysis at days 10 and 15 of EB differentiation in control and Dox-treated cells. Data represent the mean ± SD of 10 independent experiments. *p < 0.05, ***p < 0.001. performed single-cell cloning assays of mesodermal cells ment slightly increased the hematopoietic/endothelial (KDR+CD34–CD31–) using OP9 stromal cells and conditions ratio over control (no-Dox) (Figure S2A, upper table). We that support hemato-endothelial differentiation. Dox treat- repeated single-cell clonal analysis of CD31+CD34+CD43– Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 517 (legend on next page) 518 Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 cells purified from day 7 EBs with or without Dox, finding region and its zebrafish equivalent (Espin-Palazon et al., that Dox-treated cells showed an increased number of he- 2014; He et al., 2015; Sawamiphak et al., 2014). In accord matopoietic colonies but a smaller proportion of endothe- with this, gene ontology (GO) and gene set enrichment lial growth (Figure S2A, lower table), overall increasing the analysis showed that GATA2-overexpressing HEPs were hematopoietic/endothelial ratio by >4-fold. highly enriched for genes associated with immune Because GATA2 has been associated with adult HSC sur- response (IRF7, IFI27, IFIT1, TMEM173, IFI6, IFITM1, vival (Tipping et al., 2009), we questioned whether the in- TRIM6, TRIM14, and TRIM25) (Figure S3A). Surprisingly, crease in hematopoiesis was the consequence of GATA2- several highly significant GO categories of downregulated mediated proliferation/survival of emerging HEPs and genes were related to heart development and cardiogenesis HPCs. We analyzed apoptosis and cell-cycle distribution (Figure S3B), including transcription factors such as TBX3, after GATA2 induction in HEPs and HPCs at days 10 and MYOCD, PTX2, NR2F2, and FOXC2, and structural genes 15 of EB development, respectively, finding that GATA2 including TNNC1, RYR3, SPNS2, DVL3, SMO, NEBL, induction did not affect survival or proliferation of differ- HEG1, and CCM2L (Figure S3B). entiating cells (Figures 2D, 2E, and S2B). Furthermore, expression of genes related to angiogenesis Overall, our data are consistent with recent reports (Kang and endothelial cell differentiation (JAG-1, KDR, SOX17, et al., 2018; Zhou et al., 2019) showing that, rather than in- PCDH12, TEK, ESM1, and SCUBE1) were also found to be duction of HEP specification or the selective proliferation/ downregulated (Figure S3B). These data strongly suggest survival of HEPs/HPCs, GATA2 induces hematopoietic that GATA2 has a dual activity during mesodermal development by promoting EHT. patterning. GATA2 Activates the Hematopoietic Program and GATA2 Directly Binds Cardiac Genes Inhibits Cardiac Genes To determine the direct effects of GATA2 on gene transcrip- To gain mechanistic insight into how GATA2 promotes he- tion, we performed ChIP-seq analysis for GATA2 occu- matopoietic development, we performed RNA sequencing pancy on FACS-sorted GFP+ cells at day 7 of EB develop- (RNA-seq) of fluorescence-activated cell sorting (FACS)- ment, coinciding with the maximum expression of the sorted control or GATA2-overexpressing HEPs from day 2 GATA2 transgene (Figure S1E). We identified 2,097 to 7 of differentiation. We used the criteria of >1.5-fold GATA2 binding-associated genes (Figures 3A, S3C, and change and adjusted p value < 0.05 to identify differentially S3D). GATA2 was found to be enriched around the tran- expressed genes (DEGs) in the two treatment groups. scriptional start site of many genes harboring a GATA2 Among 1,127 genes significantly deregulated by binding motif (Figures 3B–3D and S3E), indicating that GATA2 induction, 700 were downregulated and 427 GATA2 does not occupy unscheduled genomic binding were upregulated (Table S2). Consistent with the in vitro sites in iGATA2-hiPSCs. Integrated analysis with RNA-seq results, GATA2 activated a broad spectrum of genes data indicated that only 8% of upregulated genes were regulating HSC/HPC development (e.g., RUNX1, MYB, decorated with GATA2 (35 genes in total), whereas up to STAT1, ITGA2B/CD41, SPN/CD43, SPI1/PU.1, ZBT3, and 20% of downregulated genes were occupied by GATA2 ALDH1A1), as well as of genes of myeloid (CD33, CD53, (143) (Figure 3E). Probabilistic analysis indicated that CD48, CSFR1, and MPO) and erythroid (NFE2, GATA1, GATA2 occupancy on repressed genes was significantly KLF1, HBZ, HBE1, HBA1, and HBG2) lineages (Figure S3A). higher than would be expected by random chance Proinflammatory cytokines have been proposed as positive (p < 1016), suggesting a predominant function of GATA2 regulators of definitive hematopoiesis in the mouse AGM repressing gene transcription during mesodermal lineage Figure 2. GATA2 Induction Promotes Hemogenic Endothelium Transition (A) CFU potential of day 10 EB progenitors in control and Dox-treated cells. Colonies were counted from each group after 2 weeks of culture and scored for the following morphological subsets: burst-forming unit-erythroid (E); CFU-granulocyte, macrophage (GM); CFU-gran- ulocyte, erythroid, macrophage, megakaryocyte (GEMM); CFU-granulocyte (CFU-G); and CFU-macrophage (CFU-M). Data represent the mean ± SD of the total number of colonies per 50,000 cells seeded of 6 independent experiments. (B) Quantitative summary of HEP and HPC analysis at days 10 and 15 of EB differentiation following stepwise treatment with Dox. Data represent the mean ± SD of 6 independent experiments. (C) Schematic of sorted HEP cell differentiation using OP9 coculture (upper panel). GATA2-induced HEPs showed a higher capacity to differentiate into CD34+/CD45+ and CD45+ hematopoietic cells (lower panel). Data represent the mean ± SD of 3 independent experiments. (D) Representative dot plots and bar graphs show cell-cycle analysis using EdU and DAPI staining on HEPs and HPCs treated or not with Dox. Data represent the mean ± SD of 3 independent experiments. (E) Apoptosis analysis in HEPs and HPCs in control and Dox-treated cells. Data represent the mean ± SD of 3 independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001. Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 519 (legend on next page) 520 Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 differentiation (Figure 3E). Among the upregulated GATA2 distinct cell clusters (Figure 4A). The t-distributed stochas- target genes, we identified determinants for hematopoietic tic neighbor embedding (t-SNE) projection visualized that development such as RUNX1 and NFE2 (Figures 3F and 3G). day 2 EB-derived cells formed one expression signature Of note, several of the downregulated GATA2 target genes cluster, whereas mesodermal (day 5) and hemato-endothe- were associated with heart development, such as TBX3 lial (day 7) cells showed the presence of two subpopula- and PITX2 (Figures 3F and 3G), confirming the RNA-seq tions in each cluster (Figure 4B). We assigned biological data. identities to each cluster based on the expression of key The specific binding of GATA2 to proximal gene regions marker genes (adjusted p value < 0.05) (Table S3). Cluster of cardiac regulators strongly suggests that GATA2 directly 1 (EBs day 2) showed high expression of pluripotency genes controls the repression of genes involved in cardiac (NANOG, OCT4/POU5f1, ZFP42/REX1, DPPA4, and SALL2), development. To test this, we used a luciferase reporter indicating maintenance of undifferentiated pluripotent assay in COS7 cells for three cardiac genes (TBX3, PITX2, cells. Cluster 2 (mesodermal S1) showed a multi-lineage and ISL-1), according to the ChIP-seq data. Luciferase mesodermal identity (BMP4, MSX2, PDGFRA, LGR5, activity driven by the 30 UTR of the TBX3 promoter, CDX2, FRZB, HOXA1, BMP5, FTH1, WNT5A, WNT5B, the 1,960/+369 region of the ISL-1 promoter, and HAND1, TNNT1, and TBX3). Cluster 3 (mesodermal S2) the 4,350/3,192 region of the PITX2 promoter, was lacked expression of mesodermal markers, but co-ex- significantly reduced by GATA2 coexpression (Figure 3H). pressed ETV2, its target SCL/TAL1, and EGFL7, which are Thus, in addition to validating GATA2 as a pro-hematopoi- responsible for restricting mesoderm specification to endo- etic fate regulator, these data point to an additional and un- thelium fate (Wareing et al., 2012), and genes suggestive of described role for GATA2 in repressing cardiac regulation. a mesenchymal phenotype (COL1A1, COL6A3, COL3A1, COL6A2, ACTA2, and LUM). Cluster 4 (hemato-endothelial Single-Cell RNA-Seq of Mesodermal and Hemato- S1) and cluster 5 (hemato-endothelial S2) shared the Endothelial Progenitors expression of several key hemato-endothelial markers To better characterize the effect of GATA2 overexpression (CDH5, PECAM1, ICAM2, CD40, ESAM, FLI1, ERG, ETS1, on mesodermal diversification, we performed single-cell and HHEX); however, hemato-endothelial S2 cells showed RNA-seq (scRNA-seq) on FACS-purified mesodermal a more restrictive endothelial identity (expressing selec- (KDR+CD34–CD31–; EBs d5) and hemato-endothelial pro- tively NOTCH4, SOX7, NRP2, TEK, and NOS3). Also, we genitor (KDR+CD34+CD43–; EBs d7) cells with or without noted that hemato-endothelial S1 was enriched for genes Dox (Figure S4A). EB-derived cells from day 2 were profiled that regulate cell cycle (AURKB, TOP2A, CDK1, MKI67, as the starting population. BRCA2, CASC5, and CDCA5). In unsupervised clustering based on the Seurat method Next, we retrospectively colored each cluster on the basis (Butler et al., 2018), we identified five transcriptionally of Dox and no-Dox conditions (Figure 4C). For each cluster, Figure 3. Identification of GATA2 Target Genes during Mesoderm Specification (A) Venn diagram indicating the overlap between the number of genes found targeted by GATA2 in each replicate of the ChIP-seq of GATA2 in EBs (plus Dox). (B) MEME-ChIP motif analysis on the sequence of the GATA2 peaks in two independent ChIP-seq replicates showing an enrichment of GATA2 motifs at the center of the ChIP-seq peaks, as the top-ranked motif. (C) Genomic distribution of ChIP-seq peaks of GATA2 compared with the whole genome in two independent clones. The pie chart represents the distribution of GATA2 peaks corrected by the genome-wide distribution of each gene feature (background circle distribution). The pie chart indicates that GATA2 preferentially occupies TSS neighborhood regions, including 50 UTR and PROXIMAL regions. The DISTAL region is the region within 2.5 and 0.5 kbp upstream of the TSS. The PROXIMAL region is the region within 0.5 kbp of the TSS. CDS is the protein CoDing Sequence. INTRONS are intronic regions. INTERGENIC is the remainder of the genome. TSS is the transcription start site. (D) Meta-gene plot showing the GATA2 ChIP-seq profile occupancy in iPSCs (CL6 no-Dox) and EBs (CL6 plus Dox) and IgG from 5 kb of the TSS until +5 kb of transcription end site (TES). (E) Histogram of the proportion of GATA2 target genes deregulated after induction of GATA (iGATA2) (up or down, 1.5-fold change, false discovery rate < 0.05) during EB development (p value was calculated using chi-square test). (F) Functional analysis on the mammalian phenotype enrichment ontology for the deregulated GATA2 target genes (MGI database). (G) UCSC genome browser screenshots of several developmental hematopoietic and cardiac genes targeted by GATA2. (H) Relative luciferase activity of COS7 cells transfected with reporter plasmids for ISL1 (1,960/+369), PITX2 (4,350/3,192), or TBX3 30 UTR genes and pWPI-GATA2 plasmid. Relative luciferase activity is shown as the ratio of luciferase activity to that in cells cotransfected with mock vector. A representative result of three independent experiments performed in triplicate is shown as the mean ± SD. Differences were determined using Student’s t test; *p < 0.05, **p < 0.01. Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 521 (legend on next page) 522 Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 the number of cells per condition was calculated and target exon 2 of GATA2 and generate knockout hiPSC- each Dox condition was compared with that of the corre- GATA2 clones (hiPSC-GATA2KO). After expanding individ- sponding no-Dox condition (Table S3). A chi-square test ual clones, we selected two targeted clones with biallelic showed that treatment and cluster populations were not mutations (CL14 and CL19; Figure 5A). Immunocyto- independent (p = 0.0002687). Considering an adjusted chemistry and qRT-PCR analysis confirmed the absence p value < 0.05, the post hoc comparisons showed that, at of GATA2 expression during hiPSC-GATA2KO differentia- day 5 of EB development, the proportion of mesodermal tion (Figures S5A and S5B). No predicted CRISPR off-targets S1 cells increased significantly in Dox conditions were detected by genomic sequencing of hiPSC-GATA2KO (chi-square = 6.816191, adjusted p value = 0.018067005) clones (Figure S5C). The hiPSC-GATA2KO clones retained (Figure 4D). At day 7, GATA2 overexpression induced a normal karyotype and maintained the expression of plu- an enrichment of the hemato-endothelial S1 cluster ripotency markers (Figures S5D and S5E), confirming that (chi-square = 3.082502, adjusted p value = 0.079138704) the GATA2 knockout is compatible with human stem cell (Figure 4D). pluripotency. Based on these data, we speculate that GATA2 enhances We differentiated wild-type and hiPSC-GATA2KO clones the proportion of cells with a multi-lineage mesodermal into hematopoietic cells as described earlier (see Figure 1A). signature and further increases the probability, at the sin- As expected, GATA2 knockout significantly affected gle-cell level, that a mesodermal progenitor acquired a HEP generation (CD31+CD34+CD45–) in day 10 EBs (Fig- hemato-endothelial transcriptional profile. These data are ure 5B) and markedly decreased the number of HPCs consistent with our FACS analysis (Figure S4A) and are in (CD34+CD43+CD45+) in day 15 EBs (Figure 5B). Moreover, line with recent findings showing that GATA2 overexpres- we consistently failed to detect hematopoietic CFCs at day sion enhances the generation of mesodermal cells and 10 of EB development (data not shown). further promotes EHT (Zhou et al., 2019). In line with the transcription and ChIP-seq data, qRT- Finally, we performed a new clustering analysis of meso- PCR analysis revealed that the cardiac regulators TBX3, dermal (day 5) and hemato-endothelial single cells (day 7) NKX2.5, and MYOCD were progressively upregulated dur- based on DEGs (1,127 genes) identified in our bulk RNA- ing hiPSC-GATA2KO EB development, whereas the hemato- seq analysis (HEPs day 10) (Table S4). Using t-SNE and hier- poietic transcription factors SCL, GATA1, and PU.1 were archical clustering to visualize the data, five main clusters markedly suppressed (Figure 5C). were found (Figures S4B and S4C). GO enrichment analysis As these observations strongly suggest that interfering and over-representation analysis (Boyle et al., 2004) re- with GATA2 expression might be an effective strategy to vealed that cluster 0 and cluster 4 were associated with generate cardiomyocytes ex vivo, we differentiated hiPSC- hemato-endothelial progenitors (ESCR, CDH5, ERG, GATA2KO and iGATA2-hiPSC lines (with or without Dox SOX17, and HOXA9), cluster 3 mesoderm (TBX3 and administration, days 2–7) to cardiomyocytes using a MSX1), cluster 1 mesenchymal (FN1, LAMA1, and well-characterized protocol (Lian et al., 2013) (Figure 6A), COL5A1), and cluster 2 epiblast (EPCAM and CDH1) (Fig- and measured the expression of the cardiac structural ure S4D). These data are in line with the scRNA-seq analysis proteins troponin I (cTnI) and myosin heavy chain and suggest that most of the DEGs (1,056 genes) of day 10 (MHC) after 20 days. Remarkably, we found a 3-fold in- were already differentially expressed in our single cells at crease in the number of cTnI+ MHC+ cells in differentiated d5 and d7. hiPSC-GATA2KO cells compared with iGATA2-iPSC cells in the absence of Dox (Figures 6B and 6C). Conversely, Dox GATA2 Knockout Inhibits Hematopoietic administration led to a significant decrease in the number Development and Favors Cardiomyogenesis of cTnI+ MHC+ cells in the iGATA2-iPSC line (Figures 6B Our results so far strongly suggest that transient GATA2 and 6C). As a functional readout of cardiomyocyte genera- expression promotes hematopoietic differentiation and tion, we monitored for the appearance of beating cells. represses alternative mesodermal fates during HEP specifi- Whereas the majority of hiPSC-GATA2KO cells started cation. To address whether GATA2 is necessary for specifi- beating at days 8–10 of differentiation and persisted cation of HEPs, we used CRISPR/Cas9 gene editing to throughout the 20-day experiment (Videos S1 and S2), Figure 4. Single-Cell RNA-Seq Analysis Reveals GATA2 as a Driver of Hemato-Endothelial Specification (A) Heatmap showing the expression pattern of top 10 gene markers distinguishing 5 clusters from scRNA-seq analysis. Colored top bar indicates assigned cluster. Red indicates the highest scaled expression and blue the lowest. (B) t-SNE visualization plot of day 2 (d2) EBs, mesodermal d5 and hemato-endothelial d7 cells according to clusters in (A). (C) t-SNE visualization plot of single cells of each cluster retrospectively colored by Dox (red) and no-Dox treatment (gray). (D) Stacked bar diagram showing the percentage of cells in each cluster with respect to treatment conditions. Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 523 Figure 5. GATA2 Is Essential for HPC Development (A) Representative sequencing of targeted homozygous hiPSC clones at the GATA2 locus using sgRNA1 and sgRNA2; PAM sequences are labeled in yellow. (B) Quantitative summary of HEP and HPC FACS analysis at d10 and d15 of EB development in wild-type (WT) and hiPSC-GATA2KO clones. Data represent the mean ± SD fold change relative to WT iPSC lines. (C) qRT-PCR analysis of hematopoietic and cardiac regulators in WT and hiPSC-GATAKO clones during EB development. Data represent the mean ± SD of 3 independent experiments. **p < 0.01, ***p < 0.001. 524 Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 Figure 6. GATA2 Knockout Promotes Car- diomyocyte Differentiation (A) Schematic of the cardiomyocyte differ- entiation protocol. (B) Representative FACS analysis of the percentage of cTnI+ MHC+ cells in hiPSC- GATA2KO and iGATA2-iPSC lines (with and without Dox treatment) at day 20 of differentiation. (C) Bar graph showing the mean percentage of cTnI+ MHC+ cells in (B). Data represent the mean ± SD of 3 independent experi- ments. *p < 0.05, ***p < 0.001. very few iGATA2-iPSC cells treated with Dox were beating et al., 2017; Kouskoff et al., 2005; Liu et al., 2013; Schoene- (Video S3). Taken together, these results suggest that beck et al., 2007; Van Handel et al., 2012). For example, GATA2 is both an activator of hematopoiesis and a overexpression of Scl in mesodermal cells promotes he- repressor of cardiac cell fate. matopoietic development at the expense of cardiomyogen- esis in differentiated mouse ESCs (Ismailoglu et al., 2008), whereas Scl-deficient mice show ectopic cardiomyocytes DISCUSSION in yolk salc endothelium and die at E9.5 due to the com- plete absence of hematopoiesis (Shivdasani et al., 1995; GATA2 has long been implicated as a master regulator of Van Handel et al., 2012). Similarly, Etv2/Er71 deficiency murine hematopoiesis (Tsai et al., 1994; Tsai and Orkin, leads to a complete block of hemato-endothelial develop- 1997), and its dysregulation is associated with human ment and a concomitant expansion of the cardiac lineage immunodeficiency syndromes (Spinner et al., 2014). How- in mutant embryos (Koyano-Nakagawa and Garry, 2017; ever, its role in early human hematopoiesis is less clear. Hu- Lee et al., 2008; Liu et al., 2012; Rasmussen et al., 2011), man iPSCs are a useful model to interrogate the molecular whereas Etv2 overexpression in differentiating mouse mechanisms driving early hematopoietic development. ESCs shows the opposite phenotype (Liu et al., 2012). To our knowledge, this is the first study showing a dual To date, a direct role for GATA2 in cardiac development activity of GATA2 in human early hematopoiesis. We has not been demonstrated. Gata2-null mice have no show that GATA2 drives mesoderm progenitors to differen- apparent cardiovascular phenotype, but to our knowledge tiate into blood cells and represses cardiac fates. Indeed, our studies specifically addressing heart malformation in GATA2 knockout model confirmed that loss of GATA2 ac- Gata2 mutants have not been performed. Whether tivity impairs hematopoietic development and enhances GATA2 overexpression in vivo enhances hematopoietic cardiomyocyte differentiation. cell specification while concurrently retarding cardiac Earlier studies both in vivo and in stem cell models sug- development should be addressed in the future. gested that hematopoietic and cardiac lineages develop in GATA2 is known to cooperate with ER71/ETV2 and close proximity and are mutually antagonistic (Bussmann SCL/TAL1 to regulate endothelial and hematopoietic pro- et al., 2007; Chagraoui et al., 2018; Chan et al., 2013; Freire grams in stem cells (Elcheva et al., 2014; Shi et al., 2014). Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 525 The recent observation that coexpression of ER71/ETV2, improve the yield of target cells (blood or cardiomyocytes) SCL/TAL1, and GATA2 during mouse ESC differentiation from hPSCs for drug screening and disease modeling. enhances FLK-1+ hemangioblast production while block- In summary, we establish a novel role for GATA2 during ing cardiac output (Liu et al., 2013) suggests a critical regu- mesodermal lineage specification and provide new insights latory relationship between these factors during mesoderm into the complex regulatory network that controls human diversification. Nevertheless, the gene regulatory network early hematopoietic development. governing hematopoietic and cardiac development is poorly understood. Our bulk RNA-seq data revealed that GATA2 overexpression failed to upregulate ER71/ETV2 EXPERIMENTAL PROCEDURES expression in our HEP model, which is consistent with Human iPSC Culture the finding that ER71 is expressed before GATA2 during Human GATA2-hiPSC and iPSC-GATA2KO lines were maintained both early mouse and ESC development (Liu et al., 2013). on Matrigel-coated 60-mm plates in mTESR1 medium (Giorgetti Based on recent studies (Chagraoui et al., 2018; Org et al., et al., 2017). Culture medium was changed daily and cells were 2015), it is reasonable to presume that GATA2 and SCL passaged weekly by EDTA dissociation. collaborate to promote blood specification at the expense of cardiac fates. Yet, our deep DNA motif enrichment anal- EB-Based Hematopoietic Differentiation ysis of GATA2 targets demonstrate the absence of SCL bind- Human iPSCs were differentiated as described (Giorgetti et al., ing sites (CANNTG) in both GATA2-bound activated and 2017). In brief, EBs were treated with the GSK3 inhibitor repressed genes, which suggests that these two factors—at CHIR99021 (3 mM) from day 2 to 3 of culture (Sturgeon et al., least in our experimental model—act independently. Of 2014). From day 3, EBs were cultured in differentiation medium note, recent analysis at the single-cell level revealed that until day 15. To induce transgene GATA2 expression Dox only a subset of mouse E8.5 Scl/ endothelial cells upregu- (1 mg/mL) was added from day 2 to 7 of differentiation. lated expression of a few cardiac-related genes. However, those cells did not display a full cardiomyocyte transcrip- Cardiac Differentiation tional program and continued to express key endothelial Human iPSCs were differentiated in monolayer cultures with mod- markers (Pijuan-Sala et al., 2019; Scialdone et al., 2016). ulators of canonical Wnt signaling (Lian et al., 2013). Contracting Therefore, the role of SCL in the specification of hemato- cardiomyocytes could be observed between day 8 and 10 of differ- poietic fate and as a cadiac repressor needs to be entiation. Differentiated cells were disaggregated at day 20 with 0.25% trypsin-EDTA for 5–8 min at 37 C for FACS analysis. determined. ChIP-seq indicated that GATA2 acts more as a repressor than an activator during mesodermal diversification. Spe- ChIP-Seq Combined with DNA Massive Sequencing cifically, GATA2 binds directly to cardiac regulator pro- GATA2-iPSCs (CL6 and CL9) were differentiated in the presence of moters, leading to their downregulation. Although previ- Dox (from day 2 to 7). ChIP experiments were performed with 1.5 3 106 cells using the ChIP-IT High Sensitivity Kit from Active ous studies in HPCs and mature blood cells have shown Motif (no. 53040), and specific antibodies for GATA2 (Santa Cruz; that endogenous GATA2 preferentially occupies sites sc-9008) or rabbit IgG as isotype control (Abcam; 172730), distant to promoters (Calero-Nieto et al., 2014; Fujiwara Libraries were prepared according to Illumina instructions and et al., 2009; Huang et al., 2016; Wilson et al., 2015), our sequenced using the HiSeq 2000 platform (Illumina). analysis revealed GATA2 binding promoter regions. Whether this apparent discrepancy is a consequence of Statistical Analysis cell-type-specific differences in GATA2 occupancy, the All data are expressed as mean ± SD. Statistical comparisons were fact that GATA2 is overexpressed in our studies, or both, performed using Student’s t test (95% confidence interval). Statis- is unclear. tical significance was defined as p < 0.05. Enforced expression of instructive factors is an accepted strategy to guide lineage fate commitment (Doulatov ACCESSION NUMBERS et al., 2013; Elcheva et al., 2014; Nakajima-Takagi et al., 2013; Navarro-Montero et al., 2017; Ramos-Mejia et al., The accession number for the ChiP-seq data reported in this paper 2014; Sugimura et al., 2017), and offers the possibility to is GEO: GSE107639; the accession number for the single-cell RNA- seq data is GEO: GSE133996; and the accession number for the generate any differentiated cell type from hPSCs. Our study RNA-seq data is GEO: GSE118980. provides cellular and molecular evidence that GATA2 induction promotes an enhancement of mesodermal cells with hemato-endothelial potential, and decreases the SUPPLEMENTAL INFORMATION probability that alternative mesodermal fates occur. Supplemental Information can be found online at https://doi.org/ Accordingly, GATA2 could be a target for manipulation to 10.1016/j.stemcr.2019.07.009. 526 Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 AUTHOR CONTRIBUTIONS Butler, A., Hoffman, P., Smibert, P., Papalexi, E., and Satija, R. (2018). Integrating single-cell transcriptomic data across different Contribution, J.C. conceived the study, designed and performed conditions, technologies, and species. Nat. Biotechnol. 36, the experiments, analyzed the data, and wrote the manuscript. 411–420. S.A., F.J.C.-N., J.L.M., E.M.-R., C.B., E.B., X.W., C.P., L.Z., S.J.-D., Calero-Nieto, F.J., Ng, F.S., Wilson, N.K., Hannah, R., Moignard, V., E.M., D.R.M., and M.R. performed the experiments and analyzed Leal-Cervantes, A.I., Jimenez-Madrid, I., Diamanti, E., Wernisch, the data. H.H., E.H.B., B.G., L.d.C., P.M., and A.R. analyzed the L., and Gottgens, B. (2014). Key regulators control distinct tran- data and interpreted the results. A.G. conceived the study, de- scriptional programmes in blood progenitor and mast cells. signed and performed experiments, analyzed data, and wrote the EMBO J. 33, 1212–1226. manuscript. Chagraoui, H., Kristiansen, M.S., Ruiz, J.P., Serra-Barros, A., Richter, ACKNOWLEDGMENTS J., Hall-Ponsele, E., Gray, N., Waithe, D., Clark, K., Hublitz, P., et al. (2018). SCL/TAL1 cooperates with Polycomb RYBP-PRC1 to sup- The authors thank Dr Catherine Porcher (University of Oxford) for press alternative lineages in blood-fated cells. Nat. Commun. 9, the SCL antibody, Olalla Iglesias, Anna Duarri, and Yvonne Ri- 5375. chaud for help with cell differentiation, and core of CMRB for tech- Chan, S.S., Shi, X., Toyama, A., Arpke, R.W., Dandapat, A., Iaco- nical assistance. A.G. is supported by Ramón y Cajal Program vino, M., Kang, J., Le, G., Hagen, H.R., Garry, D.J., et al. (2013). (RyC-2013-13221) and MINECO (SAF2016-80205-R). C.B. is sup- Mesp1 patterns mesoderm into cardiac, hematopoietic, or skeletal ported by the FIS/ISCIII (PI14/01191), Spanish Cancer Association myogenic progenitors in a context-dependent manner. Cell Stem (CI15152720BUEN). J.L.M. is supported by ISCIII (CA18/00045) Cell 12, 587–601. and by the European Social Fund (ESF). M.R. is supported by MINECO (SAF2015-73226-JIN). P.M. is funded by the PERIS Pro- Choi, K.D., Yu, J., Smuga-Otto, K., Salvagiotto, G., Rehrauer, W., gram of the Generalitat de Catalunya and by the Obra Social la Vodyanik, M., Thomson, J., and Slukvin, I. (2009). Hematopoietic Caixa-Fundacion Josep Carreras. H.H. is supported by Miguel Ser- and endothelial differentiation of human induced pluripotent vet (CP14/00229). CNAG-CRG Core funding is from the ISCIII, stem cells. Stem Cells 27, 559–567. and the Generalitat de Catalunya. Gottgens laboratory is sup- de Pater, E., Kaimakis, P., Vink, C.S., Yokomizo, T., Yamada-Ina- ported by The Wellcome Trust, MRC, CRUK, NIH-NIDDK, and gawa, T., van der Linden, R., Kartalaei, P.S., Camper, S.A., Speck, core support grants by The Wellcome Trust to the Wellcome and N., and Dzierzak, E. (2013). Gata2 is required for HSC generation MRC Cambridge Stem Cell Institute. Additional funding came and survival. J. Exp. Med. 210, 2843–2850. from Generalitat de Catalunya (2014-SGR-1460 and PERIS Dickinson, R.E., Griffin, H., Bigley, V., Reynard, L.N., Hussain, R., SLT002/16/00234); MINECO (SAF2015-69706-R); ISCIII/FEDER Haniffa, M., Lakey, J.H., Rahman, T., Wang, X.N., McGovern, N., (RD16/0011/0024); ACCIÓ/FEDER (AdvanceCat); Fundació la et al. (2011). Exome sequencing identifies GATA-2 mutation as Marató de TV3 (201534-30); and CERCA Program/Generalitat de the cause of dendritic cell, monocyte, B and NK lymphoid defi- Catalunya. ciency. Blood 118, 2656–2658. Doulatov, S., Vo, L.T., Chou, S.S., Kim, P.G., Arora, N., Li, H., Had- Received: May 3, 2018 land, B.K., Bernstein, I.D., Collins, J.J., Zon, L.I., et al. (2013). In- Revised: July 12, 2019 duction of multipotential hematopoietic progenitors from human Accepted: July 15, 2019 pluripotent stem cells via respecification of lineage-restricted pre- Published: August 8, 2019 cursors. Cell Stem Cell 13, 459–470. Dzierzak, E., and Speck, N.A. (2008). Of lineage and legacy: the REFERENCES development of mammalian hematopoietic stem cells. Nat. Immu- Ayllon, V., Bueno, C., Ramos-Mejia, V., Navarro-Montero, O., nol. 9, 129–136. Prieto, C., Real, P.J., Romero, T., Garcia-Leon, M.J., Toribio, M.L., Elcheva, I., Brok-Volchanskaya, V., Kumar, A., Liu, P., Lee, J.H., Bigas, A., et al. (2015). The Notch ligand DLL4 specifically marks Tong, L., Vodyanik, M., Swanson, S., Stewart, R., Kyba, M., et al. human hematoendothelial progenitors and regulates their he- (2014). Direct induction of haematoendothelial programs in hu- matopoietic fate. Leukemia 29, 1741–1753. man pluripotent stem cells by transcriptional regulators. Nat. Boyle, E.I., Weng, S., Gollub, J., Jin, H., Botstein, D., Cherry, J.M., Commun. 5, 4372. and Sherlock, G. (2004). GO::TermFinder–open source software Espin-Palazon, R., Stachura, D.L., Campbell, C.A., Garcia-Moreno, for accessing Gene Ontology information and finding significantly D., Del Cid, N., Kim, A.D., Candel, S., Meseguer, J., Mulero, V., and enriched Gene Ontology terms associated with a list of genes. Bio- Traver, D. (2014). Proinflammatory signaling regulates hematopoi- informatics 20, 3710–3715. etic stem cell emergence. Cell 159, 1070–1085. Bresnick, E.H., Katsumura, K.R., Lee, H.Y., Johnson, K.D., and Per- Freire, A.G., Waghray, A., Soares-da-Silva, F., Resende, T.P., Lee, D.F., kins, A.S. (2012). Master regulatory GATA transcription factors: Pereira, C.F., Nascimento, D.S., Lemischka, I.R., and Pinto-do, O.P. mechanistic principles and emerging links to hematologic malig- (2017). Transient HES5 activity instructs mesodermal cells toward nancies. Nucleic Acids Res. 40, 5819–5831. a cardiac fate. Stem Cell Reports 9, 136–148. Bussmann, J., Bakkers, J., and Schulte-Merker, S. (2007). Early Fujiwara, T., O’Geen, H., Keles, S., Blahnik, K., Linnemann, A.K., endocardial morphogenesis requires Scl/Tal1. PLoS Genet. 3, e140. Kang, Y.A., Choi, K., Farnham, P.J., and Bresnick, E.H. (2009). Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 527 Discovering hematopoietic mechanisms through genome-wide Koyano-Nakagawa, N., and Garry, D.J. (2017). Etv2 as an essential analysis of GATA factor chromatin occupancy. Mol. Cell 36, regulator of mesodermal lineage development. Cardiovasc. Res. 667–681. 113, 1294–1306. Gao, X., Johnson, K.D., Chang, Y.I., Boyer, M.E., Dewey, C.N., Lee, D., Park, C., Lee, H., Lugus, J.J., Kim, S.H., Arentson, E., Chung, Zhang, J., and Bresnick, E.H. (2013). Gata2 cis-element is required Y.S., Gomez, G., Kyba, M., Lin, S., et al. (2008). ER71 acts down- for hematopoietic stem cell generation in the mammalian embryo. stream of BMP, Notch, and Wnt signaling in blood and vessel pro- J. Exp. Med. 210, 2833–2842. genitor specification. Cell Stem Cell 2, 497–507. Giorgetti, A., Castano, J., Bueno, C., Diaz de la Guardia, R., Lian, X., Zhang, J., Azarin, S.M., Zhu, K., Hazeltine, L.B., Bao, X., Delgado, M., Bigas, A., Espinosa, L., and Menendez, P. (2017). Hsiao, C., Kamp, T.J., and Palecek, S.P. (2013). Directed cardiomyo- Proinflammatory signals are insufficient to drive definitive he- cyte differentiation from human pluripotent stem cells by modu- matopoietic specification of human HSCs in vitro. Exp. Hematol. lating Wnt/beta-catenin signaling under fully defined conditions. 45, 85–93.e82. Nat. Protoc. 8, 162–175. Gomes, A.M., Kurochkin, I., Chang, B., Daniel, M., Law, K., Satija, Lim, K.C., Hosoya, T., Brandt, W., Ku, C.J., Hosoya-Ohmura, S., N., Lachmann, A., Wang, Z., Ferreira, L., Ma’ayan, A., et al. (2018). Camper, S.A., Yamamoto, M., and Engel, J.D. (2012). Conditional Cooperative transcription factor induction mediates hemogenic Gata2 inactivation results in HSC loss and lymphatic mispattern- reprogramming. Cell Rep. 25, 2821–2835.e7. ing. J. Clin. Invest. 122, 3705–3717. Hahn, C.N., Chong, C.E., Carmichael, C.L., Wilkins, E.J., Brauti- Ling, K.W., Ottersbach, K., van Hamburg, J.P., Oziemlak, A., Tsai, gan, P.J., Li, X.C., Babic, M., Lin, M., Carmagnac, A., Lee, Y.K., F.Y., Orkin, S.H., Ploemacher, R., Hendriks, R.W., and Dzierzak, E. et al. (2011). Heritable GATA2 mutations associated with familial (2004). GATA-2 plays two functionally distinct roles during the myelodysplastic syndrome and acute myeloid leukemia. Nat. ontogeny of hematopoietic stem cells. J. Exp. Med. 200, 871–882. Genet. 43, 1012–1017. Liu, F., Bhang, S.H., Arentson, E., Sawada, A., Kim, C.K., Kang, I., Yu, J., Sakurai, N., Kim, S.H., Yoo, J.J., et al. (2013). Enhanced he- He, Q., Zhang, C., Wang, L., Zhang, P., Ma, D., Lv, J., and Liu, F. mangioblast generation and improved vascular repair and regener- (2015). Inflammatory signaling regulates hematopoietic stem ation from embryonic stem cells by defined transcription factors. and progenitor cell emergence in vertebrates. Blood 125, 1098– Stem Cell Reports 1, 166–182. 1106. Liu, F., Kang, I., Park, C., Chang, L.W., Wang, W., Lee, D., Lim, D.S., Huang, J., Liu, X., Li, D., Shao, Z., Cao, H., Zhang, Y., Trompouki, Vittet, D., Nerbonne, J.M., and Choi, K. (2012). ER71 specifies E., Bowman, T.V., Zon, L.I., Yuan, G.C., et al. (2016). Dynamic con- Flk-1+ hemangiogenic mesoderm by inhibiting cardiac mesoderm trol of enhancer repertoires drives lineage and stage-specific tran- and Wnt signaling. Blood 119, 3295–3305. scription during hematopoiesis. Dev. Cell 36, 9–23. Merika, M., and Orkin, S.H. (1993). DNA-binding specificity of Ismailoglu, I., Yeamans, G., Daley, G.Q., Perlingeiro, R.C., and GATA family transcription factors. Mol. Cell. Biol. 13, 3999–4010. Kyba, M. (2008). Mesodermal patterning activity of SCL. Exp. Hematol. 36, 1593–1603. Molkentin, J.D. (2000). The zinc finger-containing transcription factors GATA-4, -5, and -6. Ubiquitously expressed regulators of tis- Ivanovs, A., Rybtsov, S., Anderson, R.A., Turner, M.L., and Medvin- sue-specific gene expression. J. Biol. Chem. 275, 38949–38952. sky, A. (2014). Identification of the niche and phenotype of the Nakajima-Takagi, Y., Osawa, M., Oshima, M., Takagi, H., Miyagi, S., first human hematopoietic stem cells. Stem Cell Reports 2, Endoh, M., Endo, T.A., Takayama, N., Eto, K., Toyoda, T., et al. 449–456. (2013). Role of SOX17 in hematopoietic development from hu- Ivanovs, A., Rybtsov, S., Welch, L., Anderson, R.A., Turner, M.L., man embryonic stem cells. Blood 121, 447–458. and Medvinsky, A. (2011). Highly potent human hematopoietic Navarro-Montero, O., Ayllon, V., Lamolda, M., Lopez-Onieva, L., stem cells first emerge in the intraembryonic aorta-gonad-meso- Montes, R., Bueno, C., Ng, E., Guerrero-Carreno, X., Romero, T., nephros region. J. Exp. Med. 208, 2417–2427. Romero-Moya, D., et al. (2017). RUNX1c regulates hematopoietic Johnson, K.D., Hsu, A.P., Ryu, M.J., Wang, J., Gao, X., Boyer, M.E., differentiation of human pluripotent stem cells possibly in Liu, Y., Lee, Y., Calvo, K.R., Keles, S., et al. (2012). Cis-element cooperation with proinflammatory signaling. Stem Cells 35, mutated in GATA2-dependent immunodeficiency governs hema- 2253–2266. topoiesis and vascular integrity. J. Clin. Invest. 122, 3692–3704. Org, T., Duan, D., Ferrari, R., Montel-Hagen, A., Van Handel, B., Kang, H., Mesquitta, W.T., Jung, H.S., Moskvin, O.V., Thomson, Kerenyi, M.A., Sasidharan, R., Rubbi, L., Fujiwara, Y., Pellegrini, J.A., and Slukvin, I.I. (2018). GATA2 is dispensable for specification M., et al. (2015). Scl binds to primed enhancers in mesoderm to of hemogenic endothelium but promotes endothelial-to-hemato- regulate hematopoietic and cardiac fate divergence. EMBO J. 34, poietic transition. Stem Cell Reports 11, 197–211. 759–777. Katsumura, K.R., Bresnick, E.H., and Group, G.F.M. (2017). The Orkin, S.H. (1992). GATA-binding transcription factors in hemato- GATA factor revolution in hematology. Blood 129, 2092–2102. poietic cells. Blood 80, 575–581. Kouskoff, V., Lacaud, G., Schwantz, S., Fehling, H.J., and Keller, G. Persons, D.A., Allay, J.A., Allay, E.R., Ashmun, R.A., Orlic, D., Jane, (2005). Sequential development of hematopoietic and cardiac S.M., Cunningham, J.M., and Nienhuis, A.W. (1999). Enforced mesoderm during embryonic stem cell differentiation. Proc. Natl. expression of the GATA-2 transcription factor blocks normal he- Acad. Sci. U S A 102, 13170–13175. matopoiesis. Blood 93, 488–499. 528 Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 Pijuan-Sala, B., Griffiths, J.A., Guibentif, C., Hiscock, T.W., Jawaid, Tipping, A.J., Pina, C., Castor, A., Hong, D., Rodrigues, N.P., W., Calero-Nieto, F.J., Mulas, C., Ibarra-Soria, X., Tyser, R.C.V., Ho, Lazzari, L., May, G.E., Jacobsen, S.E., and Enver, T. (2009). High D.L.L., et al. (2019). A single-cell molecular map of mouse gastrula- GATA-2 expression inhibits human hematopoietic stem and pro- tion and early organogenesis. Nature 566, 490–495. genitor cell function by effects on cell cycle. Blood 113, 2661– Ramos-Mejia, V., Navarro-Montero, O., Ayllon, V., Bueno, C., Ro- 2672. mero, T., Real, P.J., and Menendez, P. (2014). HOXA9 promotes he- Tsai, F.Y., Keller, G., Kuo, F.C., Weiss, M., Chen, J., Rosenblatt, M., matopoietic commitment of human embryonic stem cells. Blood Alt, F.W., and Orkin, S.H. (1994). An early haematopoietic defect 124, 3065–3075. in mice lacking the transcription factor GATA-2. Nature 371, Rasmussen, T.L., Kweon, J., Diekmann, M.A., Belema-Bedada, F., 221–226. Song, Q., Bowlin, K., Shi, X., Ferdous, A., Li, T., Kyba, M., et al. Tsai, F.Y., and Orkin, S.H. (1997). Transcription factor GATA-2 is (2011). ER71 directs mesodermal fate decisions during embryogen- required for proliferation/survival of early hematopoietic cells esis. Development 138, 4801–4812. and mast cell formation, but not for erythroid and myeloid termi- Robert-Moreno, A., Espinosa, L., de la Pompa, J.L., and Bigas, A. nal differentiation. Blood 89, 3636–3643. (2005). RBPjkappa-dependent Notch function regulates Gata2 Van Handel, B., Montel-Hagen, A., Sasidharan, R., Nakano, H., and is essential for the formation of intra-embryonic hematopoiet- Ferrari, R., Boogerd, C.J., Schredelseker, J., Wang, Y., Hunter, S., ic cells. Development 132, 1117–1126. Org, T., et al. (2012). Scl represses cardiomyogenesis in prospective Sawamiphak, S., Kontarakis, Z., and Stainier, D.Y. (2014). Inter- hemogenic endothelium and endocardium. Cell 150, 590–605. feron gamma signaling positively regulates hematopoietic stem Vicente, C., Conchillo, A., Garcia-Sanchez, M.A., and Odero, M.D. cell emergence. Dev. Cell 31, 640–653. (2012). The role of the GATA2 transcription factor in normal and Schoenebeck, J.J., Keegan, B.R., and Yelon, D. (2007). Vessel and malignant hematopoiesis. Crit. Rev. Oncol. Hematol. 82, 1–17. blood specification override cardiac potential in anterior meso- Wareing, S., Eliades, A., Lacaud, G., and Kouskoff, V. (2012). ETV2 derm. Dev. Cell 13, 254–267. expression marks blood and endothelium precursors, including Scialdone, A., Tanaka, Y., Jawaid, W., Moignard, V., Wilson, N.K., hemogenic endothelium, at the onset of blood development. Macaulay, I.C., Marioni, J.C., and Gottgens, B. (2016). Resolving Dev. Dyn. 241, 1454–1464. early mesoderm diversification through single-cell expression Wilson, N.K., Foster, S.D., Wang, X., Knezevic, K., Schutte, J., profiling. Nature 535, 289–293. Kaimakis, P., Chilarska, P.M., Kinston, S., Ouwehand, W.H., Dzier- Shi, X., Richard, J., Zirbes, K.M., Gong, W., Lin, G., Kyba, M., zak, E., et al. (2010a). Combinatorial transcriptional control in Thomson, J.A., Koyano-Nakagawa, N., and Garry, D.J. (2014). blood stem/progenitor cells: genome-wide analysis of ten major Cooperative interaction of Etv2 and Gata2 regulates the develop- transcriptional regulators. Cell Stem Cell 7, 532–544. ment of endothelial and hematopoietic lineages. Dev. Biol. 389, Wilson, N.K., Kent, D.G., Buettner, F., Shehata, M., Macaulay, I.C., 208–218. Calero-Nieto, F.J., Sanchez Castillo, M., Oedekoven, C.A., Shivdasani, R.A., Mayer, E.L., and Orkin, S.H. (1995). Absence of Diamanti, E., Schulte, R., et al. (2015). Combined single-cell func- blood formation in mice lacking the T-cell leukaemia oncoprotein tional and gene expression analysis resolves heterogeneity within tal-1/SCL. Nature 373, 432–434. stem cell populations. Cell Stem Cell 16, 712–724. Spinner, M.A., Sanchez, L.A., Hsu, A.P., Shaw, P.A., Zerbe, C.S., Wilson, N.K., Timms, R.T., Kinston, S.J., Cheng, Y.H., Oram, S.H., Calvo, K.R., Arthur, D.C., Gu, W., Gould, C.M., Brewer, C.C., Landry, J.R., Mullender, J., Ottersbach, K., and Gottgens, B. et al. (2014). GATA2 deficiency: a protean disorder of hematopoie- (2010b). Gfi1 expression is controlled by five distinct regulatory re- sis, lymphatics, and immunity. Blood 123, 809–821. gions spread over 100 kilobases, with Scl/Tal1, Gata2, PU.1, Erg, Sturgeon, C.M., Ditadi, A., Awong, G., Kennedy, M., and Keller, G. Meis1, and Runx1 acting as upstream regulators in early hemato- (2014). Wnt signaling controls the specification of definitive and poietic cells. Mol. Cell Biol. 30, 3853–3863. primitive hematopoiesis from human pluripotent stem cells. Nat. Wlodarski, M.W., Hirabayashi, S., Pastor, V., Stary, J., Hasle, H., Biotechnol. 32, 554–561. Masetti, R., Dworzak, M., Schmugge, M., van den Heuvel-Eibrink, Sugimura, R., Jha, D.K., Han, A., Soria-Valles, C., da Rocha, E.L., Lu, M., Ussowicz, M., et al. (2016). Prevalence, clinical characteristics, Y.F., Goettel, J.A., Serrao, E., Rowe, R.G., Malleshaiah, M., et al. and prognosis of GATA2-related myelodysplastic syndromes in (2017). Haematopoietic stem and progenitor cells from human children and adolescents. Blood 127, 1387–1397, quiz 1518. pluripotent stem cells. Nature 545, 432–438. Zhou, Y., Zhang, Y., Chen, B., Dong, Y., Zhang, Y., Mao, B., Pan, X., Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Lai, M., Chen, Y., Bian, G., et al. (2019). Overexpression of GATA2 Tomoda, K., and Yamanaka, S. (2007). Induction of pluripotent enhances development and maintenance of human embryonic stem cells from adult human fibroblasts by defined factors. Cell stem cell-derived hematopoietic stem cell-like progenitors. Stem 131, 861–872. Cell Reports 13, 31–47. Stem Cell Reports j Vol. 13 j 515–529 j September 10, 2019 529