BMP4 and perivascular cells promote hematopoietic differentiation of human pluripotent stem cells in a differentiation stage-specific manner

Abstract

The efficient and reproducible derivation and maturation of multipotent hematopoietic progenitors from human pluripotent stem cells (hPSCs) requires the recapitulation of appropriate developmental stages and the microenvironment. Here, using serum-, xeno-, and feeder-free stepwise hematopoietic induction protocols, we showed that short-term and high-concentration treatment of hPSCs with bone morphogenetic protein 4 (BMP4) strongly promoted early mesoderm induction followed by increased hematopoietic commitment. This method reduced variations in hematopoietic differentiation among hPSC lines maintained under chemically defined Essential 8 medium compared to those maintained under less-defined mTeSR medium. We also found that perivascular niche cells (PVCs) significantly augmented the production of hematopoietic cells via paracrine signaling mechanisms only when they were present during the hematopoietic commitment phase. A protein array revealed 86 differentially expressed (>1.5-fold) secretion factors in PVC-conditioned medium compared with serum-free control medium, of which the transforming growth factor-β inducible gene H3 significantly increased the number of hematopoietic colony-forming colonies. Our data suggest that BMP4 and PVCs promote the hematopoietic differentiation of hPSCs in a differentiation stage-specific manner. This will increase our understanding of hematopoietic development and expedite the development of hPSC-derived blood products for therapeutic use.

Fig. 1. Comparison of the effects of low- and high-dose BMP4 on the hematopoietic differentiation of hPSCs.

a A schematic overview of the stepwise hematopoietic induction protocol for hESCs (CHA15) and hiPSCs (hEF12-2 and iPS-NT4-S1). b Representative bright-field images of colonies at the early (day 2) and late (day 17) stages of hematopoietic differentiation with BMP4 (20 vs 100 ng/mL). Scale bar, 200 μm. c, d Effects of low- and high-dose BMP4 on the induction efficiency (c) and the yield (d) of hematopoietic lineage cells from hPSCs were analyzed by flow cytometry. The bars indicate the mean ± SD; *p < 0.05, **p < 0.01.

Fig. 2. Comparison of hPSC culture conditions for the optimization of hematopoietic differentiation from hPSCs.

a Representative bright-field images of colonies during hematopoietic differentiation of hiPSC lines (CMC003, CMC009 and CMC011) maintained in two different culture conditions (E8+Vit vs mTeSR+Mat). b Temporal expression patterns of hematopoietic lineage markers (CD45−CD31+, hemogenic precursors; CD34+CD45+ and CD34+CD43+, hematopoietic progenitors; CD34−CD45+ and CD34−CD43+, mature blood cells) during hematopoietic development from hiPSC lines were analyzed by flow cytometry. The bars indicate the mean ± SD.

Fig. 3. Comparison of hematopoietic progenitor capacity to generate colonies.

a Assessment of hematopoietic progenitor capacity of hiPSCs maintained under E8 + Vit vs mTeSR + Mat culture conditions. b Distribution of CFU subtypes. c qPCR analysis for Brachyury and MIXL1 transcripts in undifferentiated hiPSC cultures. The bars indicate the mean ± SD. CFU-E, erythrocyte; CFU-G, granulocyte; CFU-M, macrophage.

Fig. 4. PVCs enhance the hematopoietic differentiation of hPSCs via paracrine action.

a Experimental scheme used to determine the paracrine effects of PVCs on the hematopoietic differentiation of hPSCs (CHA15 and iPS-NT4-S1). b Representative bright-field images of colonies on day 17 of hematopoietic differentiation. Scale bar, 50 μm. c Effects of PVC-CM on the production of hematopoietic lineage cells for two different exposure times (days 0–9 and days 9–17). d The total number of CFUs was counted by plating 1 × 10⁴ cells in methylcellulose. e Distribution of CFU subtypes. The bars indicate the mean ± SD; *p < 0.05, **p < 0.01 (control vs. CM 0–9 or CM 9–17). CFU-GM, granulocyte-macrophage.

Fig. 5. Protein array analysis reveals the secretion of ECM proteins, including BIGH3.

a Heatmaps for upregulated and downregulated proteins in PVC-CM compared to the control. The color spectrum from blue to red indicates low to high expression. b Molecular and cellular functions associated with PVC-secreted proteins. c Flow cytometry (upper) and qRT-PCR (lower) analysis revealed that supplementation with BIGH3 during the hematopoietic commitment phase (days 9–17) did not enhance the production of hematopoietic or endothelial lineage cells. d Cultures differentiated with BIGH3 produced more CFU-E and CFU-M colonies than the control cultures. The bars indicate the mean ± SD; *p < 0.05.