Abstract
Although human pluripotent stem cells (hPSCs) can theoretically be differentiated into any cell type, their ability to generate hematopoietic cells shows a major variability from one cell line to another. The reasons of this variable differentiation potential, which is constant and reproducible in a given hPSC line, are not clearly established. In order to study this phenomenon, we comparatively studied 4 human embryonic stem cell lines (hESC) and 11 human induced pluripotent stem cell (hiPSC) lines using transcriptome assays. These cell lines exhibited a significant variability to generate in vitro hematopoiesis as evaluated by day-16 embryoid body (EB) formation followed by clonogenic (CFC) assays. Four out of 11 iPSC lines (PB6, PB9, PB12.1, and PB14.3) were found to lack any hematopoietic differentiation ability whereas 7 cell lines showed variable hematopoietic potential. Among hESC lines, H9 and CL0 had low H1 and SA01 exhibited high hematopoietic potential using the above assays. Among hESC and hIPSC displaying hematopoietic potential, two sub-groups were further defined based on their hematopoietic CFC efficiency: a group of poor (generation of less than 100 CFC/105 cells, PB4 / PB10 /H9 /CL01), and high hematopoietic competency (more than 120 CFC/105 cells, PB3/ PB6.1 /PB7 /PB13 /PB17 /SA01/H1). Using global miRNome analysis performed at the pluripotency stage, the expression of 754 individual miRNAs was analyzed from 15 hPSC lines in order to explore a potential predictive marker between both sub-groups of pluripotent cells according to their hematopoietic potency. Using this approach, 27 miRNAs out of 754 appeared differentially expressed allowing the identification of a miRNA signature associated with hematopoietic-competency. The hematopoietic competency was associated with down-regulation of miR-206, miR-135b, miR-105, miR-492, miR-622 and upregulation of miR-520a, miR-296, miR-122, miR-515, miR-335. Amongst these, miR-206 harbored the most significant variation (0.04-Fold change). To explore the role of miRNA-206 in this phenomenon, we have generated a miR-206-eFGP-Puro lentiviral vector which was transfected in hESC line H1 followed by puromycin selection. As a control, H1 cell line was transfected with a Arabidopsis thaliana microRNA sequence (ath-miR-159a), which has no specific targets in mammalian cells. The correct expression of the transgenes were evaluated by flow cytometry (using GFP) and q-RT-PCR for miR-206 expression. The hematopoietic potential of H1 cell line and its miR-206-overexpressing counterpart was then tested using standard in vitro assays via d16-EB generation. We found that both CFC numbers and percentage of CD34+ were significantly lower in H1-mir-206-derived day-16 EB cells than in H1-ath- derived day-16 EB cells (p < 0.05). Thus, over-expression of miR-206 in this blood-competent hESC appeared to repress its hematopoietic potential at very early stage, since a similar lower CFC efficiency was observed in day-3 EB cells derived from miR-206 overexpressing H1 cell line. We then conducted an integrative bioinformatics analysis on miR-206 predicted target genes. To this end, 773 mRNA target transcripts of the broadly conserved (across vertebrates) miR-1-3p/206 family were identified in the TargetScan database and were integrated into the global transcriptomic analysis performed by microarray on day-16 EB cells. Using supervised ranking product analysis, 62 predicted gene targets of the miR-1-3p/206 family were found to be significantly up-regulated in hematopoietic-competent EB samples including the transcription factors RUNX1 and TAL1. Hierarchical unsupervised clustering, based on this subset of 62 predicted mir-206 target genes, fully discriminated hematopoietic-deficient from hematopoietic-competent cells. In conclusion, miRNA profiling performed at pluripotency stage could be useful to predict the ability to human iPSC to give rise to blood cell progenitors. This work emphasizes for the first time the critical role of the muscle-specific miR-206 in hematopoietic differentiation. Finally, these results suggest that genetic manipulation of hESC/iPSC could be used to enhance their hematopoietic potential and to design protocols for generation of hPSC-derived hematopoietic stem cells with long-term reconstitution ability.
Disclosures
No relevant conflicts of interest to declare.
Generation and characterization of stable pig pregastrulation epiblast stem cell lines
