Overexpression of HOXA9 upregulates NF-κB signaling to promote human hematopoiesis and alter the hematopoietic differentiation potentials

The HOX genes are master regulators of embryogenesis that are also involved in hematopoiesis. HOXA9 belongs to a cluster of HOX genes that play extensively studied roles in hematopoiesis and leukemogenesis. We established HOXA9-inducible human embryonic stem cells (HOXA9/hESCs) with normal pluripotency and potential for hematopoiesis, which could be used to analyze gene function with high accuracy. HOXA9/hESCs co-cultured with aorta–gonad–mesonephros-derived stromal cells (AGM-S3) were induced to overexpress HOXA9 with doxycycline (DOX) at various times after hematopoiesis started and then subjected to flow cytometry. Induction of HOXA9 from Day 4 (D4) or later notably promoted hematopoiesis and also increased the production of CD34+ cells and derived populations. The potential for myelogenesis was significantly elevated while the potential for erythrogenesis was significantly reduced. At D14, a significant promotion of S phase was observed in green fluorescent protein positive (GFP+) cells overexpressing HOXA9. NF-κB signaling was also up-regulated at D14 following induction of HOXA9 on D4. All of these effects could be counteracted by addition of an NF-κB inhibitor or siRNA against NFKB1 along with DOX. Overexpression of HOXA9 starting at D4 or later during hematopoiesis significantly promoted hematopoiesis and the production of myeloid progenitors while reduced the production of erythroid progenitors, indicating that HOXA9 plays a key role in hematopoiesis and differentiation of hematopoietic lineages.


Background
The Class I homeobox (HOX) genes, which encode a family of homeodomain-containing transcription factors, are master regulators of trunk and tail development during embryogenesis that were initially characterized in Drosophila (Sukumar and Shah 2010). In mammals, 39 genes have arisen from the original HOX gene cluster through several rounds of duplication, resulting in the formation of four clusters (A, B, C, and D) (Amores et al. 1998).These clusters, which exhibit high levels of conservation and redundancy of function (He et al. 2011), are distributed on human chromosomes as follows: HOXA, 7p15; HOXB, 17q21; HOXC, 12q13; and HOXD, 2q31 (Rice and Licht 2007).
HOXA9, a homeodomain transcription factor that plays a key role in hematopoiesis, is the gene that is most differentially expressed between CD34+ cells derived from human cord blood (hCB) and human embryonic stem cells (hESCs) ; in particular, the gene is much more strongly expressed in the former lineage (Wang et al. 2005;Domingo-Reines et al. 2017). HOXA9 is highly expressed in human hematopoiesis stem/progenitor cells (HSPCs) and then downregulated during differentiation (Sauvageau et al. 1994;Zeng et al. 2019). Hoxa9-knockout mice do not have a lethal phenotype, but have a much lower hematopoiesis efficiency (Lawrence et al. 1997). Deficiencies in Hoxa9 gene lead to myeloid and lymphoid defects and abnormalities in HSC function (Lawrence et al. 1997). Hoxa9-/-HSCs cannot rebuild the blood system (Lawrence et al. 2005); conversely, enforced expression of Hoxa9 in bone marrow HSPCs improves the production of human hematopoietic stem cells (hHSCs) and myeloid precursors (Thorsteinsdottir et al. 2002), and HOXA9 overexpression in hESC improves hematopoiesis in vitro (Ramos-Mejia et al. 2014). These studies indicate that HOXA9 plays key roles in hematopoiesis. Overexpression of HOXA9 in murine marrow cells leads to expansion of HSCs and committed progenitors that likely undergo transformation to acute myeloid leukemia (AML) with coexpression of Meis1 and Pbx3, emphasizing the gene's key role in blood physiology and pathology (Dassé et al. 2012) In order to further elucidate the functions of key genes in hematopoiesis, including HOXA9 , our group developed the Tet-on system based on piggyBac with a green fluorescent protein (GFP) tag, PB-Tet-on-OE, and successfully used it to establish transgenic hESC lines (Chen et al. 2017;Zhou et al. 2019). The result is a reliable gene delivery system that can manipulate and track gene expression during early hematopoiesis originating from hESCs with much higher resolution and lower background interference than traditional lentiviral systems (Chen et al. 2017;Zhou et al. 2019;Sun et al. 2019). The aorta-gonad-mesonephros-derived stromal cell (AGM-S3) co-culture system is based on AGM-S3 cells line derived from the stromal cells of AGM region of mouse embryo, which could induce hESC to produce the HSPCs (hematopoietic stem/progenitor cells) and many types of blood cells in vitro (Ma et al. 2008a;Ma et al. 2008b;Ma et al. 2007) , which is a suitable in vitro system for observing the function of HOXA9 during hematopoiesis. Our results provide clear evidence that overexpression of HOXA9 promotes hematopoiesis with elevated myeloid potential and weakened erythroid potential, and suggest a molecular/cellular mechanism for its role in hematopoiesis.

Transgenic hESCs exhibit inducible expression and normal pluripotency
We constructed PB-Tet-on-GFP-T2A-hHOXA9 and established the corresponding H1-derived inducible HOXA9/hESC line (Fig. 1a). HOXA9/hESCs treated with or without DOX for 48 h were monitored by fluorescence microscopy (Fig. 1b), as well as by an qRT-PCR and western blotting ( Fig. 1c-d), which revealed highly stringent and efficient induction of HOXA9. Western blot analysis showed that stemness-specific markers, including OCT4, SOX2, and NANOG, were expressed at normal levels in HOXA9/hESCs (Fig. 1d), demonstrating that these cells have a normal pluripotency potential.
HOXA9 overexpression from D4 or later promotes hESCderived hematopoiesis Making use of the Tet-on inducible system, we induced HOXA9 overexpression during the various stages of hematopoiesis in the AGM-S3 co-culture system and then performed flow cytometry detection at D4, D8, and D14, which results were represented in Fig. 2. When HOXA9 was overexpressed starting from D0 to D14, the KDR+ cells were not influenced (Fig. 2a, Fig. S2A) while the most hematopoiesisrelated populations at D8 and D14 were significantly reduced (Fig. 2b, c, Fig. S2B, S2C), indicating that overexpression of HOXA9 from the earliest stage did not hinder the induction of mesoderm, but severely blocked human hematopoiesis. But when HOXA9 was overexpressed from D4 or later it promoted the production of CD34 + CD43−, CD34 + CD43+, and CD34 − CD43+ populations at D8 (Fig. 2b and Fig.  S2B), and of CD34 + CD43+, CD34 − CD43+, CD34 + CD45+, especially CD34 − CD45+ populations at D14 except that GPA + CD71+, GPA + CD71− populations at D14 reduced in the most case ( Fig. 2c and Fig.  S2C). If overexpressed starting from D2 or D4 to D14, the production of CD34 + CD43− and CD34 + CD45− populations was also elevated at D14. These observations indicated that hematopoiesis was significantly enhanced by overexpression of HOXA9 starting at D4 except for erythrogenesis.
HOXA9 overexpression has different effects on hematopoietic potential at different stages of mesoderm development To elucidate the function of HOXA9 at different stages of mesoderm development, the key populations with hematopoietic potential, e.g., the KDR+ populations at D1 or D4, were isolated from the corresponding co-cultures, and cultured in the AGM-S3 co-culture system. Overexpression of HOXA9 blocked the production of CD34+ cells originating from KDR+ cells isolated from co-cultures at D1 (Fig. 3a), whereas for KDR+ cells isolated from cocultures at D4, HOXA9 overexpression dramatically increased the production of hematopoiesis-related populations, including CD34 + CD43−, CD34+CD43+, and CD34 − CD43+, and even CD34 + CD45−, CD34 + CD45+, and CD34-CD45+ (Fig. 3b).

HOXA9 overexpression promoted the hematopoietic population at different stages
To elucidate the function of HOXA9 on different stages of hematopoiesis, CD34 + CD43+ populations at D8 or D14 were isolated from the corresponding co-cultures and subjected to further culture. The CD34 + CD43+ population sorted from the D8 co-culture were resuspended in fulllineage hematopoietic differentiation medium. Overexpression of HOXA9 starting on the first day of further culture significantly promoted production of CD45+ cells (Fig. 4). The CD34 + CD43+ population sorted from D14 co-culture was plated in myeloid or erythroid expansion medium. Overexpression of HOXA9 significantly promoted production of CD45+ cells, but decreased the production of GPA + CD71+ and GPA + CD71− cells (Fig. 5), indicating that HOXA9 overexpression could markedly promote hematopoiesis, especially myelogenesis, but blocked erythrogenesis.

HOXA9 promotes myelogenesis, but impairs erythrogenesis
The non-induced HOXA9/hESC co-culture cells and the GFP+ cell fractions of HOXA9/hESC or GFP/hESC co- TRE, tet-on regulation element; CMV mini, cytomegalovirus minimum promoter; T2A, Thosea asigna virus 2A peptide. b After HOXA9/hESCs were induced with DOX for 48 h, the cells were imaged by fluorescence microscopy, allowing observation of co-expressed GFP. c qRT-PCR was used to confirm that inducible expression of HOXA9 was highly stringent and efficient at the transcriptional level. d WB detection was used to confirm HOXA9 overexpression at the protein level, and the pluripotency of HOXA9/hESC using ant-SOX2, OCT4, and NANOG antibodies culture cells induced from D4, were cultured for colony formation. The number of CFU-E of cells derived from GFP+ cells was less than that of CFU-E of cells derived from non-induced cells, whereas the number of CFU-GM derived from GFP+ cells was greater than that of CFU-GM derived from non-induced cells (Fig. 6Ai). Similar results were obtained for the CD34 + CD43+ population in GFP+ or non-induced cells (Fig. 6Aii). The GFP+ cells or GFP + CD34 + CD43+ cells sorted from D4-induced GFP/hESC co-cultures at D14 obtained similar colony numbers to the corresponding cells sorted from non-induced HOXA9/hESC co-cultures. These observations indicate that overexpression of HOXA9 from D4 can promote the production of myeloid progenitor cells while decreasing the production of erythroid progenitor cells, which effects were not caused by the overexpression of GFP. The morphologies of hematopoietic colonies were Fig. 2 Overexpression of HOXA9 from D4 promotes hematopoiesis in co-culture with AGM-S3 cells. HOXA9/hESCs were treated with DOX starting from D0, D2, D4, D6, D8, D10, or D12 during co-culture with AGM-S3 cells until D14, and the co-culture cells were subjected to flow cytometry analysis using 7-AAD and the following combinations of antibodies: a KDR (at D4), b CD34/CD43 (at D8), and c GPA/CD71, CD34/ CD43, or CD34/CD45 (at D14). Results were compared between noninduced co-cultures and the GFP+ fraction of co-cultures treated with DOX starting on different days. When HOXA9 was induced from the earliest stage, especially D0, CD34 + CD43−, CD34 − CD43+, CD34 + CD43+ populations at D8, and the most hematopoietic populations at D14 were significantly decreased. On the contrary, most above populations at D8 and D14 were dramatically increased by induction of HOXA9 from D4, especially the CD34 − CD45+ population (but GPA + CD71− and GPA + CD71+ populations were decreased in the most case), indicating that HOXA9 strongly promotes hematopoiesis except for the early stage. The experiments were repeated three times, and P < 0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) Fig. 3 Different effects of HOXA9 overexpression on the hematopoietic potential of the mesoderm population at different stages. KDR+ cells were sorted from non-induced HOXA9/hESC co-cultured with AGM-S3 at D1 (a) or D4 (b). About 5 × 10 3 sorted cells were re-plated on irradiated AGM-S3 in 24-well plates, treated with or without DOX, and flow cytometry analysis was performed after 10 days using anti-CD34/ CD43/CD45 antibodies. The results revealed that the production of CD34+ cells and their subsequent population could be significantly decreased by overexpression of HOXA9 from D0 and continued until D14, whereas their production was significantly increased when overexpression of HOXA9 began on D4 and continued until D14. The experiments were repeated three times, and P < 0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) confirmed by phase contrast microscopy ( Fig HOXA9 overexpression up-regulates NF-κB signaling, and promotion of hematopoiesis by HOXA9 is abolished by inhibition of NF-κB signaling Co-cultures of HOXA9/hESCs with AGM-S3 were induced starting at D4, and qRT-PCR detection at D14 revealed that the expression level of the NFKB1 and NFKB2 genes was significantly enhanced when HOXA9 was induced, which indicated that NF-κB signaling pathway was up-regulated ( Fig. 7a). When QNZ (10 nM), an inhibitor of the NF-κB signaling pathway, or siRNA against KFKB1 was added along with DOX starting at D4, flow cytometry analysis at D14 showed that the promotion of hematopoiesis (especially myelogenesis) and the impair of erythrogenesis by HOXA9 overexpression were partially or completely abolished (Fig. 7b, Fig. S3). These observations indicated that the effects of HOXA9 overexpression on hematopoiesis during the late stage were closely related to NF-κB signaling.
The effects of HOXA9 overexpression on hematopoiesis were related to a change in cell cycle status, which can be reversed by inhibition of NF-κB signaling Cell cycle assays revealed a significant increase in the proportion of cells in S phase in co-cultures induced starting at D4, whereas addition of QNZ or siRNA against NFKB1 at the same time abolished this increase. Similar effects were observed for all cells, as well as for the CD45+ population. This observation indicated that the increase in proliferation is the main cellular mechanisms responsible for the significant expansion of the CD45+ population, which is closely related to NF-κB signaling. When NF-κB signaling was blocked by a specific inhibitor or siRNA against NFKB1, all of these effects disappeared (Fig. 7c). The promotion effects of HOXA9 overexpressed on CD34 + CD43+ cells at D8. CD34 + CD43+ cells were sorted from the non-induced HOXA9/hESCs co-cultured with AGM-S3 at D8. About 5 × 10 3 sorted cells were re-plated in full-lineage hematopoietic differentiation medium (48well), treated with or without DOX, and flow cytometry analysis was performed after 12 days using 7-AAD and either combination of anti-CD34/ CD43, anti-CD34/CD45 or anti-CD43/CD45 antibodies. All these assays revealed a significant increase in the CD34 − CD43+, CD34 − CD45+, and CD43 + CD45+ populations when HOXA9 was overexpressed during further culture. The experiments were repeated three times, and P < 0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) Changes in gene expression underlie HOXA9-promoted hematopoiesis via up-regulation of NF-κB signaling and the cell cycle To elucidate the mechanism by which HOXA9 promoted myeloid progenitors, GFP+ cells of corresponding induced co-cultures were subjected to gene analysis controlled by the non-induced co-cultures. The surface markers of hematopoiesis (Kennedy et al. 2012), such as CD43, CD45, CD37, CD33, and other genes that promote the development of myeloid progenitor cells and hematopoietic differentiation, such as HOXC4, RUNX1, and MYB, were up-regulated, whereas genes related to erythroblast development, such as GATA1, EPOR, and GATA3, were down-regulated. Genes that promote the G1/S phase transition, such as E2F2, RBL1, and CDK6, were up-regulated, and cyclin-dependent kinase inhibitors, such as P21, P27, P57, were down-regulated. Genes related to the NF-κB signaling pathway, such as NFKB1, BAFF, and BAFFR, were up-regulated. Genes related to myelogenesis, such as PU.1, EGR1, and GFI1 were upregulated. The rich factor plot of KEGG pathway enrichment analysis of RNA-seq results for signaling pathway also indicated the up-regulation of the NF-κB signaling pathway. These data indicate that HOXA9 promoted hematopoiesis, and especially myelogenesis while impaired the erythrogenesis by up-regulating NF-κB signaling and the cell cycle (Fig. 8).
HOXA9 mRNA expression was up-regulated at D6 and gradually decreased at the late stage of hematopoietic differentiation in the AGM-S3 co-culture system qRT-PCR analysis at various time points showed that HOXA9 expression at the mRNA level was up-regulated at D6 dramatically and gradually decreased at the late stage of hematopoietic development for non-transgenic H1 hESCs in the AGM-S3 co-culture system: HOXA9 had much lower expression during mesoderm induction (D0-D4) than the one during hematopoietic differentiation (D6-D14), and a small increase could be observed at D14 (Fig. S4).

Discussion
HOXA9 plays a crucial role in blood cell development. Defects in HOXA9 lead to impaired hematopoiesis and abolish the ability of HSCs to repopulate irradiated recipients (Lawrence et al. 2005). Overexpression of HOXA9 Fig. 5 The positive effects of HOXA9 overexpression on CD34 + CD43+ cells at D14. CD34 + CD43+ cells were sorted from the non-induced HOXA9/hESCs co-cultured with AGM-S3 cells at D14. About 5 × 10 3 sorted cells were resuspended in 250 μl myeloid or erythroid expansion medium and seeded into each well of a 48-well plate, and then treated with or without DOX. flow cytometry analysis was performed after 8 days using 7-AAD and combination of anti-CD34/CD45 or anti-GPA/CD71 antibodies. All of these assays revealed a significant increase in the CD34 − CD45+ population meanwhile decrease in GPA + CD71+ population when HOXA9 was overexpressed during further culture. The experiments were repeated three times, and P < 0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) dramatically increases the efficiency of hematopoiesis (Ramos-Mejia et al. 2014;Ohno et al. 2013). However, the role of HOXA9 during human hematopoiesis, and its potential to increase the production of hematopoietic cells and cause abnormal hematopoiesis, need to been further studied in detail with higher accuracy.
To address this problem, we used a novel piggyBac Tet-on system, developed previously by us, to detailly detect the effects of enforced expression of HOXA9 during different co-culture stages of hematopoiesis with much higher accuracy. AGM-S3 in vitro co-culture system had also been used to reveal the function of HOXA9 in hematopoietic differentiation. In such systems, induction of HOXA9 at various times revealed distinctive, primarily positive, functions of HOXA9 at different stages of mesodermal development and hematopoiesis, suggesting that the gene has the potential to improve human hematopoiesis.
Induction of HOXA9 from D0 did not significantly influence production of KDR+ cells that marked mesoderm (Vodyanik et al. 2006) (Fig. 2a, Fig. S2A) while it caused that the CD34+ cells and derived populations had significantly decreased at D8 and D14 (Fig. 2b, Fig.  S2B). KDR+ cells isolated from the D1 co-culture did not produce CD34+ cells upon DOX induction (Fig. 3a). The number of CFU-GM colonies with myelogenic potential increased significantly, but the number of CFU-E colonies with erythrogenic potential decreased, which were not caused by overexpression of GFP. The experiments were repeated three times, and P < 0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) Fig. 7 The effects of HOXA9 overexpression from D4 on hematopoiesis are closely related to NF-κB signaling and might be caused by a change in the cell cycle state. a qRT-PCR detection at D14 revealed that expression levels of NFKB1 and NFKB2 were significantly enhanced in D4-induced HOXA9/hESCs co-cultured with AGM-S3 cells, which indicated that at the late stage NF-κB signaling was up-regulated by HOXA9 overexpression. b D4-induced HOXA9/ hESCs co-cultured with AGM-S3 cells were treated with 10 nM QNZ (an NF-κB signaling pathway inhibitor) or 20 nM siRNA against NFKB1, and flow cytometry analysis at D14 revealed that QNZ or siRNA against NFKB1 attenuated the positive effects of D4-induced HOXA9/hESCs on CD34 − CD43+, CD34 + CD43+, CD34 − CD45+, CD34 + CD45+ populations and the inhibitory effects on CD71 + GPA+ population. c Analyses of cell cycle state revealed that HOXA9 induction from D4 increased the proportion of cells in S phase in co-cultured cells or CD45+ cells expressing HOXA9 (the GFP+ fraction); this effect was eliminated by treatment with QNZ or siRNA against NFKB1. The experiments were repeated three times, and P < 0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) These observations demonstrate that induction of HOXA9 from the earliest stage (D0) can block the mesoderm-hemogenesis transition without influencing the induction of the mesoderm. This is similar to what has been observed before for other hematopoiesis-related genes, such as RUNX1 (Chen et al. 2017), indicating that the overexpression of these genes at the earliest stage of co-culture might be harmful to the mesoderm-hemogenesis transition.
When HOXA9 was induced from D4, the CD34 − CD45+ population dramatically increased at D14 by Fig. 8 The bioinformation analysis of HOXA9 overexpression. a Heatmaps of RNA-seq results of GFP+ cells sorted from D4-induced HOXA9/hESCs co-cultures or control cells without induction at D14. Pairwise comparison of transcription levels of selected genes. Heatmaps of selected genes associated with cell cycle, hematopoietic potential, and the NF-κB signaling pathway. In each row, red, black, and green reflect high, medium, and low expression, respectively. b The rich factor plot of KEGG pathway enrichment analysis of RNA-seq results for signaling pathway, in which item the degree of color stands for the P value and the size of node stands for the gene count about 4 to 5-fold relative to the non-induced controls (Fig. 2c). qRT-PCR analysis of non-trangenic H1 hESCs co-cultured with AGM-S3 revealed that mRNA expression of HOXA9 significantly increased at D6 and gradually dropped thereafter (Fig. S4). Therefore, overexpression of HOXA9 from D4 dramatically promoted hematopoiesis is reasonable for D4 is very near to the windows of HOXA9 function. This might explain why D4 is the key period during which HOXA9 promotes hematopoiesis.
To elucidate the detailed mechanism by which HOXA9 induction starting after D4 promotes hematopoiesis, we isolated the CD34 + CD43+ population by sorting HOXA9/hESCs co-cultured with AGM-S3 at D8 or D14, and then subjected them to further hematopoietic culture to test their potential for hematopoiesis (Vodyanik et al. 2011).
DOX induction during subsequent culture of the CD34 + CD43+ population isolated from the D8 coculture dramatically increased the production of CD34 − CD43+ and CD34 − CD45+ populations, which overlapped with the CD43 + CD45+ population (Yang et al. 2008;Vodyanik et al. 2011). When DOX induction was performed from the starting day of suspension culture, the production of CD34 − CD43+ and CD34 − CD45+ populations, was significantly higher than in the absence of DOX induction (Fig. 4). It is clear that induction of HOXA9 from D8 promoted their hematopoietic potential, as revealed by the increase in the production of CD34 + CD43+, CD34 − CD43+, CD34 + CD45+, CD34 − CD45+ cells detected at D14 (Fig. 2c).
CD34 + CD43+ population isolated from D14 cocultures was further cultured in myeloid expansion medium, which yielded a population consisting almost entirely of 100% CD34 − CD45+ cells after DOX induction. Induction of HOXA9 from starting day of suspension culture produced much greater CD34 − CD45+ cells than 53.4%, the value without any induction (Fig. 5). The colony culture assay also revealed that GFP+ cells, from either whole cocultures or the CD34 + CD43+ population that overexpress HOXA9, can produce a higher number of CFU-GM colonies than their non-induced counterparts (Fig. 6A). On the contrary, if these cells were further cultured in erythroid expansion medium, it only yielded much less GPA + CD71+ populations (0.2%) after DOX induction compared with non-induced control cells (5.4%) (Fig. 5). The corresponding colony culture assay also showed that the number of CFU-E colonies derived from the induced co-culture cells was much lower than their non-induced counterparts, which were not caused by the overexpression of GFP (Fig. 6A). Taken together, above results indicated the overexpression of HOXA9 (not GFP) from D4 lead to a significant increase in the myelogenic potential of the CD34 + CD43+ population isolated from D14 co-cultures while a decrease in their erythrogenic potential. To elucidate the molecular mechanism of HOXA9 in hematopoiesis, we investigated the relevant signaling pathways. The function of HOXA9 is closely relative to NF-κB signaling (Trivedi et al. 2008;Han et al. 2019). The research of Menendez's group also speculated that NF-κB signaling might be a potential candidate to participate in the promotion effects of HOXA9 on human hematopoiesis according bioinformation analysis (Ramos-Mejia et al. 2014) while no further research was reported, so we focused on this signaling and explored its relationship to HOXA9. The results showed that NF-κB signaling was significantly up-regulated when HOXA9 was induced from D4 (Fig. 7a). QNZ, an inhibitor of NF-κB signaling, efficiently blocked the promotion of myelogenesis and impairment of erythrogenesis caused by HOXA9 overexpression when it was added along with DOX starting at D4. The siRNA against NFKB1 had similar effects with QNZ (Fig. 7b). These observations indicated that NF-κB signaling was highly relevant to these effects, consistent with a hypothesis proposed in a previous study (Trivedi et al. 2008;Chen et al. 2007;Zhong et al. 2018;Trivedi et al. 2007;Ledoux and Perkins 2014;Yu et al. 2016;Yin et al. n.d.;Panepucci et al. 2010;Kuo et al. 2013). In some reports, continued expression of HOXA9 was shown to inhibit the endothelial cell (EC) activation pathway by interfering with the transcriptional activity of NF-κB (Trivedi et al. 2008;Trivedi et al. 2007), while in our co-culture system HOXA9 induction from D4 or later increased NF-κB transcriptional activity, which might involve a different molecular mechanism.
Cell cycle regulation is extremely important for hematopoiesis, which is always associated with upregulation of the cell cycle (Hao et al. 2016;Palis 2016) . Cell cycle analysis at D14 revealed that, after induction of HOXA9 from D4, a greater proportion of GFP+ cells, including GFP + CD45+ cells, were in S-phase, and that this was inhibited by QNZ or siRNA against NFKB1 (Fig. 7c). This observation suggests that HOXA9 might be able to alter cell cycle status, with the participation of NF-κB signaling (Panepucci et al. 2010;Kuo et al. 2013;Poulos et al. 2016). This was further confirmed by the corresponding heatmaps and the rich factor plot of KEGG pathway enrichment analysis of of total RNA Sequencing data (Fig. 8).

Conclusions
In this study, we analyzed the molecular/cellular mechanism of HOXA9 involvement in blood physiology using a unique AGM-S3 co-culture system and a Tet-on inducible system based on piggyBac. The results indicated that induction of HOXA9 from D4 or later to D14 during co-culture with AGM-S3 efficiently promoted hematopoiesis and myelogenesis while obviously impaired erythrogenesis, probably via changes in NF-κB signaling and cell cycle status. The detailed molecular/ cellular mechanism of HOXA9 in hematopoietic differentiation should be further explored in the future.

Establishment of HOXA9 inducible transgenic hESC lines
The coding region of HOXA9 was inserted between the SwaI and EcoRI sites of PB-Tet-on-OE to construct PB-Tet-on-GFP-T2A-hHOXA9, which was co-transfected in H1 hESCs along with the helper vector PB200PA-1 using Lipofectamine 3000 (Invitrogen). Positive colonies were selected with 1 μg/ml puromycin, and then passaged using ReleSR (STEM CELL Technologies) to establish an inducible hESC line (HOXA9/hESC) that coexpressed GFP. Induction by DOX was confirmed by quantitative reverse-transcription PCR (qRT-PCR) and western blot analyses, and pluripotency was confirmed by western blot analyses of SOX2, OCT4, and NANOG. The method is described in detail in Supplementary Materials and qPCR primer pairs (HOXA9-2 and GAPDH-2) were listed in Table S1.

Flow cytometry and cell sorting
Co-cultured cells were dissociated with 0.25% trypsin-EDTA solution (Invitrogen), filtered through a 40 μm nylon mesh to obtain a single-cell suspension, stained with corresponding antibodies (Table S2), and subjected to flow cytometry analysis using a FACSCanto II system, or cell sorting using a FACSJazz™ Sorter (BD Biosciences). All flow cytometry data were analyzed using FlowJo 10. The corresponding populations were isolated at different stages and performed the consequent hematopoietic culture as Fig. S1 presented.

Differentiation of myeloid and erythroid with suspension culture
Non-induced HOXA9/hESC co-cultures at D14 were dissociated by treatment with 0.25% trypsin solution and stained with 7-AAD and the combination of anti-CD34/ CD43 antibodies. CD34 + CD43+ cells were sorted from the corresponding co-cultured cells, and about 5 × 10 3 sorted cells of each type were resuspended in 250 μl myeloid expansion medium (Cat# 02693; STEMCELL Technologies) or erythroid expansion medium (Stem-Span™ SFEM II (Catalog#09605, STEMCELL Technologies) containing 10 − 6 M Dexamethasone, 4 U/ml EPO, 100 ng/ml IL-6, and 100 ng/ml SCF), then seeded into 48-well plates to be further cultured with or without DOX induction for 8 days, with a medium exchange every other day. At the end of 8 days, the cells were subjected to flow cytometry.
Testing the activation of NF-κB signaling and detecting the antagonistic effects of NF-κB signaling inhibition on HOXA9 overexpression at the late stage of hematopoiesis D4-induced HOXA9/hESCs co-cultured with AGM-S3 cells were subjected to qRT-PCR detection at D14 using NFKB1 and NFKB2 primer pairs. Order QNZ (an inhibitor of the NF-κB signaling pathway) from Selleck inc. and dissolve it using DMSO. Order siRNAs against NFKB1 from Sangon Biotech Inc. D4-induced HOXA9/ hESCs co-cultured with AGM-S3 cells were treated from D4 with 10 nM QNZ (Selleck) or the equal volume of DMSO as control, or treated from D4 with 20 nM siRNA against NFKB1 or the equal concentration of control siRNA, and then subjected to qRT-PCR, flow cytometry, and cell cycle analysis at D14. Untreated coculture cells were used as a negative control. The method is described in detail in Supplementary Materials and qPCR primer pairs were listed in Table S1.

RNA-Seq
Non-induced or D4-induced HOXA9/hESC co-cultures at D14 were dissociated by treatment with 0.25% trypsin solution. Non-induced co-culture cells and GFP+ fractions of D4-induced co-culture cells were sorted, and about 1 × 10 6 cells were treated with 1 ml TRIzol (Life Technologies). RNA was extracted and isolated, and quantified. Total RNA Sequencing was performed by BGI Tech. Analysis of gene functions was performed using the Dr. Tom Multi-Group Data Mining System (http://report.bgi.com). The experiments were repeated three times.
Detection of HOXA9 mRNA expression during co-culture of non-transgenic H1 hESCs with AGM-S3 To explore the molecular mechanism of HOXA9 in hematopoietic development, HOXA9 mRNA levels in co-culture cells of non-transgenic H1 hESC were monitored by qRT-PCR every other day, including on D0, D2, D4, D6, D8, D10, D12, and D14. The method is described in detail in Supplementary Materials and qPCR primer pairs (HOXA9-1 and GAPDH-1) were listed in Table S1.

Statistical analysis
All data are presented as means ± SD; statistical analyses were performed using Student's t test. P < 0.05 was considered significant.