SCL/TAL1-mediated transcriptional network enhances megakaryocytic specification of human embryonic stem cells

Abstract

Human embryonic stem cells (hESCs) are a unique in vitro model for studying human developmental biology and represent a potential source for cell replacement strategies. Platelets can be generated from cord blood progenitors and hESCs; however, the molecular mechanisms and determinants controlling the in vitro megakaryocytic specification of hESCs remain elusive. We have recently shown that stem cell leukemia (SCL) overexpression accelerates the emergence of hemato-endothelial progenitors from hESCs and promotes their subsequent differentiation into blood cells with higher clonogenic potential. Given that SCL participates in megakaryocytic commitment, we hypothesized that it may potentiate megakaryopoiesis from hESCs. We show that ectopic SCL expression enhances the emergence of megakaryocytic precursors, mature megakaryocytes (MKs), and platelets in vitro. SCL-overexpressing MKs and platelets respond to different activating stimuli similar to their control counterparts. Gene expression profiling of megakaryocytic precursors shows that SCL overexpression renders a megakaryopoietic molecular signature. Connectivity Map analysis reveals that trichostatin A (TSA) and suberoylanilide hydroxamic acid (SAHA), both histone deacetylase (HDAC) inhibitors, functionally mimic SCL-induced effects. Finally, we confirm that both TSA and SAHA treatment promote the emergence of CD34(+) progenitors, whereas valproic acid, another HDAC inhibitor, potentiates MK and platelet production. We demonstrate that SCL and HDAC inhibitors are megakaryopoiesis regulators in hESCs.

Figure 1

SCL overexpression increases megakaryocyte and platelet production from hESCs. (a) Schematic of the megakaryocytic differentiation of hESCs. This differentiation protocol is divided in two stages: EBs differentiation stage (from day 0 to 15) in the presence of hematopoietic cytokines (bFGF, BMP4 VEGF, FLT3L, SCF, and THPO) and OP9 coculture differentiation stage (from day 15 through day 32) where hESC-derivatives are plated on OP9 stromal cells supplemented with heparin, THPO, and SCF. (b) Representative flow cytometry showing how megakaryocytic cells (MK) and platelets are identified and analyzed throughout the OP9 coculture. Peripheral blood platelets were used to establish flow cytometry settings for analysis. (c) Kinetics of megakaryocytes and platelets emergence from two different hESCs lines (AND1 and HS181) transduced with the empty vector (EV) or SCL-expressing vector (SCL). Absolute number of MKs (upper panels) and platelets (bottom panels) were obtained using flow cytometry counting beads. (d) Accumulative kinetics of megakaryocytes and platelets emergence from EV-hESCs and SCL-hESCs at days 22 and 26 of megakaryocytic development. Data represents mean ± SEM for independent experiments.

Figure 2

SCL overexpression does not affect morphological, molecular, and functional propierties of MKs and platelets derived from hESCs. (a) Papanicolaou staining of cells purified from the supernatants of EV and SCL hESCs differentiated into OP9 cocultures for 20 days. Scale bar represents 100 µm. On the right panel, cell diameter ranges of the MKs differentiated from EV or SCL hESCs. (b) Representative confocal microscopy images of 26-day-old EV or SCL hESC-derived MKs activated by thrombin for 2 hours over fibrinogen-coated plates. Actin distribution was detected by rhodamine-phalloidin staining. CD61 staining and multinucleated cells confirm megakaryocytic commitment. DAPI was used for nuclear counterstain. Bar = 50 µm. Additionally, flow cytometry analysis of megakaryocytic control (C) or ADP-activated cells (+ADP) is shown at day 26 of megakaryocytic differentiation. PAC1 antibody was used to detect the activated form of the αIIbβ3 integrin after ADP treatment. Data represents mean ± SEM for independent experiments. (c) Confocal microscopy images of 26-day-old EV and SCL hESCs derived platelets or peripheral blood platelets (PB platelets) from a donor activated with thrombin for 2 hours over fibrinogen-coated plates. CD61 and phalloidin staining were used to evaluate platelets morphology. Bar = 10 µm. On the right panels, platelet activation after ADP stimulation was measured by flow cytometry using PAC1 antibody. Data represents mean ± SEM for independent experiments. (d) Quantitative RT-PCR analysis of several MKs transcription factors (FLI-1, NF-E2, GATA-1, and SCL) and surface markers (CD41 and CD61) in cells from supernatants of EV and SCL hESCs. Relative expression is shown normalized to EV cells. Data represents mean ± SEM for independent experiments.

Figure 3

SCL overexpression accelerates megakaryocytic progenitor emergence from hESCs. (a) Schematic of the megakaryocytic progenitors analysis from hESCs. This differentiation protocol is reduced to the EB differentiation stage in presence of hematopoietic cytokines as described. Flow cytometry analysis was performed at days 10, 14, and 17 of megakaryocytic differentiation. (b) Representative flow cytometry showing how megakaryocytic progenitors (CD34⁺ and CD34⁺CD41⁺) are identified and analyzed throughout the differentiation. (c) Kinetics of megakaryocytic progenitors (CD34⁺ and CD34⁺CD41⁺) emergence from EV-hESCs or SCL-hESCs. Data represents mean ± SEM for two independent experiments. (d) CFU-Mega from EV and SCL-EBs. (e) Representative images of EV and SCL CFU-Mega colonies. (f) Quantitative RT-PCR analysis of several transcription factors associated to megakaryocytic development (FLI-1, NF-E2, GATA-1, and SCL) and surface markers (CD41 and CD61) expressed in EBs of EV and SCL hESCs cultures at days 10, 14, and 17 of megakaryocytic differentiation. Relative expression is shown normalized to EBs from EV-hESCs at day 10 of differentiation. Data represents mean ± SEM for independent experiments.

Figure 4

SCL overexpression triggers a megakaryocytic transcriptional network. (a) Five significant biological pathways from Reactome database obtained by comparing different gene expression profiles using GSEA software: (i) differentiated EV (14-day EV) versus undifferentiated EV (0-day EV); (ii) differentiated SCL (14-day SCL) versus undifferentiated SCL (0-day SCL); (iii) differentiated SCL (14-day SCL) versus differentiated EV (14-day EV). Biological pathways are ranked by P value from the original GSEA output. False discovery rate (FDR) values are also included. Statistically significant pathways are considered when P value is less than 0.05 and FDR less than 25% (0.25). N/A means not available. (b) Top 20 biological functions of genes differentially expressed in SCL 14 days EBs compared with EV-hESCs 14 days EBs, ranked by P value using IPA software. (c) Predicted activated biofuntions included within the category “Hematological System Development and Function” in SCL 14 days EBs (14-day SCL) compared with EV 14-day EBs (14-day EV) using IPA software. Z-score values are represented as black bars using the left X axis. –log (P value) is represented as filled red circles connected by red lines on the right X axis. (d) Transcriptional regulatory network controlling MPL expression in Differentiated versus undifferentiated EV and SCL-hESCs. Orange arrows show direct predicted regulation. Relative expression levels are color coded at the bottom scale bar.

Figure 5

SCL overexpression potentiates megakaryocytic regulatory network. (a) GSEA shows highly significant enrichment for gene expression data from megakaryocytic progenitors and transcription factor promoter occupancy in human megakaryocytes (Tijssen et al). Statistical significant overlaps are considered when P value is less than 0.05 and FDR less than 25% (0.25). (b) Activation z-score values from the six transcription factors (FLI-1, GATA1, GATA2, NF-E2, RUNX1, and TAL1/SCL) obtained by IPA analysis. Z-score value greater than 2 can be considered statistically significant. (c,d) Regulatory network models for gene expression in megakaryocytic progenitors from (c) EV-hESCs and (d) SCL-hESCs. The links to downstream target genes are shown on the seven transcription factors remarked (FLI-1, GATA1, GATA2, NF-E2, RUNX1, TAL1/SCL, and ZFPM1/FOG1). MPL gene is also remarked. Relative expression levels are color coded at the bottom. Predicted activated molecules appear in orange.

Figure 6

HDAC inhibitors modulate megakaryopoesis from hESCs. (a) Schematic of the megakaryocytic progenitors analysis from hESCs in the presence or absence of TSA, SAHA, or VPA. (b) Analysis of relative viability (7-ADD negative cells) and normalized CD34⁺ progenitors emergence from hESCs treated with three different HDAC inhibitors at increasing concentrations using control (C) untreated cells as a reference. (c) Normalized total number of megakaryocytes (MKs) and platelets (PLTs) throughout megakaryocytic differentiation in the presence of the three different HDAC inhibitors at increasing concentrations. Data represented as mean ± SEM for independent experiments.