Retinoic Acid Is a Negative Regulator In Hematopoietic Development and Definitive Hematopoietic Progenitor Maintenance

Retinoic acid (RA) is a key morphogen involved in the establishment of various cell-fates in the developing embryo. The establishment of certain tissues, including mesodermal derivates such as the somites and heart depends on restricted exposure to RA during embryonic development, however, in regard to the possible involvement of RA for the emergence of blood in the human embryo very little is known. By using an inhibitor of cellular RA synthesis (DEAB) together with our human iPS-to-blood differentiation system we report enhanced generation of human hematopoietic progenitors cells (CD43/45+,CD34+), and early progenitor cells with an adult hematopoietic stem cell phenotype (CD43/45+,CD34+,CD38,CD90+,CD45RA-) possessing myeloid and lymphoid differentiation capacity. RA inhibition increased the output of these early progenitor cells by 2.7-fold (p-value: 0,006, n=6) compared to DMSO control, and increased the number of clonogenic progenitors (CFUs) by 2-fold (p-value: 0,022, n=6). This improvement is consistent when using both a human ES line (Hues3) and a human iPS-line (RB9-CB1, generated from cord blood). RA inhibition also increased the number of iPS-derived blood progenitors capable of differentiating into the lymphoid lineage, generating phenotypic T-lineage and natural killer cells after sub-culture on OP9-DL1 stroma. Conversely, and in support of our findings, directly adding RA was found to severely decrease the blood generation efficiency, with the early progenitor fraction significantly decreased even at low levels of RA exposure. In order to assess how the RA inhibition led to the increase in early progenitors, whether by directing pluripotent stem cell differentiation of mesodermal precursors towards blood, or by inferring an early progenitor cell maintenance and/or expansion, we performed the following analyses. Firstly, the loss of clonogenic progenitors was reduced 1.6 fold with extended culture in the presence of the inhibitor. Secondly, within the iPS derived blood (CD43/45+) fraction of the cultures, RA inhibition increased the frequency of the CD34+,CD38-,CD90+,CD45RA- compartment by 2.1-fold (p-value: 0,021, n=6). Similarly when comparing the CFU-forming capacity from equal numbers of sorted hematopoietic progenitors (CD43/45+,CD34+), cells generated with the inhibitor demonstrated 2-fold higher CFU frequency suggesting increased preservation of CFU-forming cells within the CD34+ progenitor fraction. Together this suggests that progenitor maintenance is improved with the inhibitor, however, the contribution to the total fold expansion seen with RA inhibition cannot be attributed solely to the increased maintenance of early progenitor cells in the culture. Using multi-timepoint RT-qPCR of samples obtained during the differentiation protocol, we analysed expression of key developmental genes and saw improved commitment of the differentiating culture towards mesoderm (BRACHYURY), decreased amounts of anterior lateral plate/cardiogenic mesoderm (NKX2.5), and enhanced expression of genes known to be relevant for definitive hematopoiesis and hemogenic endothelium (RUNX1, FLK1, APLNR, PDGFRa). In addition, preliminary data show that RA inhibition increases the frequency of the CD34hi,CD90hi,CD43neg population observed at day 8, recently described as the precursor population of definitive hematopoiesis, indicating that RA inhibition confers an early positive effect on the developmental commitment towards the hematopoietic mesoderm. Together, our data indicate that modulation of retinoic acid signaling improves blood generation from human pluripotent stem cell lines both by improving the development towards the hematopoietic mesoderm by favoring mesodermal blood precursor commitment, and by preserving the immature progenitor population from differentiation. Therefore, we propose that RA is a negative regulator of mesodermal formation and the specification towards hematopoietic precursors, and that it promotes the exhaustion of definitive hematopoietic progenitors. This suggests the possibility that blood evolutionarily formed in a region of the mesoderm where RA exposure from neighboring tissues, such as RA producing ectoderm, was restricted. Therefore, using small molecules to inhibit RA signaling is a novel approach to improve the generation of human blood in vitro.