Hlf Expression Marks Early Emergence of Hematopoietic Stem Cell Precursors With Adult Repopulating Potential and Fate

In the aorta-gonad-mesonephros (AGM) region of mouse embryos, pre-hematopoietic stem cells (pre-HSCs) are generated from rare and specialized hemogenic endothelial cells (HECs) via endothelial-to-hematopoietic transition, followed by maturation into bona fide hematopoietic stem cells (HSCs). As HECs also generate a lot of hematopoietic progenitors not fated to HSCs, powerful tools that are pre-HSC/HSC-specific become urgently critical. Here, using the gene knockin strategy, we firstly developed an Hlf-tdTomato reporter mouse model and detected Hlf-tdTomato expression exclusively in the hematopoietic cells including part of the immunophenotypic CD45– and CD45+ pre-HSCs in the embryonic day (E) 10.5 AGM region. By in vitro co-culture together with long-term transplantation assay stringent for HSC precursor identification, we further revealed that unlike the CD45– counterpart in which both Hlf-tdTomato-positive and negative sub-populations harbored HSC competence, the CD45+ E10.5 pre-HSCs existed exclusively in Hlf-tdTomato-positive cells. The result indicates that the cells should gain the expression of Hlf prior to or together with CD45 to give rise to functional HSCs. Furthermore, we constructed a novel Hlf-CreER mouse model and performed time-restricted genetic lineage tracing by a single dose induction at E9.5. We observed the labeling in E11.5 AGM precursors and their contribution to the immunophenotypic HSCs in fetal liver (FL). Importantly, these Hlf-labeled early cells contributed to and retained the size of the HSC pool in the bone marrow (BM), which continuously differentiated to maintain a balanced and long-term multi-lineage hematopoiesis in the adult. Therefore, we provided another valuable mouse model to specifically trace the fate of emerging HSCs during development.

In the aorta-gonad-mesonephros (AGM) region of mouse embryos, pre-hematopoietic stem cells (pre-HSCs) are generated from rare and specialized hemogenic endothelial cells (HECs) via endothelial-to-hematopoietic transition, followed by maturation into bona fide hematopoietic stem cells (HSCs). As HECs also generate a lot of hematopoietic progenitors not fated to HSCs, powerful tools that are pre-HSC/HSCspecific become urgently critical. Here, using the gene knockin strategy, we firstly developed an Hlf-tdTomato reporter mouse model and detected Hlf-tdTomato expression exclusively in the hematopoietic cells including part of the immunophenotypic CD45 − and CD45 + pre-HSCs in the embryonic day (E) 10.5 AGM region. By in vitro co-culture together with long-term transplantation assay stringent for HSC precursor identification, we further revealed that unlike the CD45 − counterpart in which both Hlf-tdTomato-positive and negative sub-populations harbored HSC competence, the CD45 + E10.5 pre-HSCs existed exclusively in Hlf-tdTomato-positive cells. The result indicates that the cells should gain the expression of Hlf prior to or together with CD45 to give rise to functional HSCs. Furthermore, we constructed a novel Hlf-CreER mouse model and performed time-restricted genetic lineage tracing by a single dose induction at E9.5. We observed the labeling in E11.5 AGM precursors and their contribution to the immunophenotypic HSCs in fetal liver (FL). Importantly, these Hlf-labeled early cells contributed to and retained the size of the HSC pool in the bone marrow (BM), which continuously differentiated to maintain a balanced and long-term multi-lineage hematopoiesis in the adult. Therefore, we provided another valuable mouse model to specifically trace the fate of emerging HSCs during development.

INTRODUCTION
Hematopoietic stem cells (HSCs) are a unique cell population which can differentiate into nearly all kinds of blood lineage and maintain its quantity by self-renewal throughout the lifetime (Dzierzak and Bigas, 2018). At embryonic day (E) 10.5, the first functional HSCs are detected in the aorta-gonad-mesonephros (AGM) region (Muller et al., 1994;Medvinsky and Dzierzak, 1996;de Bruijn et al., 2000de Bruijn et al., , 2002North et al., 2002). At around E11.5, AGM HSCs start to migrate and colonize the fetal liver (FL) where they rapidly expand (Sánchez et al., 1996). During the perinatal period, HSCs begin to migrate to the bone marrow (BM) where they permanently reside throughout adult life (Morrison et al., 1995;Christensen et al., 2004).
To isolate and decipher the embryonic hematopoietic populations, several researches have made great efforts in developing novel enrichment strategies or mouse models over the past few decades. With the help of the genetic lineage tracing mouse model VE-Cadherin-Cre, HSCs are proven to be derived from endothelial cells (Zovein et al., 2008;Chen et al., 2009). The progenies of endothelial cells contribute to a fraction of hematopoietic cells in adult BM, thymus, and spleen when labeled at E9.5 (Zovein et al., 2008;Chen et al., 2009). Additionally, time-lapse confocal imaging of Ly6A-GFP embryo slices confirms that aortic endothelial cells expressing GFP can generate Ly6A-GFP + Kit + CD41 + hematopoietic stem progenitor cells (HSPCs) at E10.5 (Boisset et al., 2010). The iterations of index-sorting analyses on Gata2IRESVenus reporter mice suggest that the functional HSCs emerge in the E11 IAHCs with a quantity of one to two cells (Vink et al., 2020). Recently, HECs are proven to be highly enriched by a cocktail of surface markers named PK44 (CD41 − CD43 − CD45 − CD31 + CD201 + Kit + CD44 + ) or the newly established reporter Neurl3-EGFP (Hou et al., 2020).
Hlf is a transcription factor that belongs to the proline and acidic amino acid-rich (PAR) basic leucine zipper (bZip) family. Hlf is initially recognized in leukemia, and recent studies have identified its high expression in BM HSCs and essential role in maintaining hematopoiesis in a quiescent state (Gazit et al., 2013;Komorowska et al., 2017;Wahlestedt et al., 2017). In embryos, Hlf is specially expressed in IAHCs and FL HSCs without marking yolk sac erythro-myeloid progenitors, and its expression level is up-regulated along the maturation of HSCs during development (Yokomizo et al., 2019).
Recently, we identified Hlf as a signature gene for pre-HSCs in addition to HSCs (Zhou et al., 2016), suggesting that Hlf is a suitable molecule for studying the emergence of early pre-HSCs and their contribution to adult hematopoiesis. Therefore, we newly constructed Hlf-tdTomato and Hlf-CreER mouse models to investigate the dynamic marker changes of pre-HSCs and the adult fate of pre-HSCs labeled by Hlf-CreER.

Mice
Mice were fed at the Laboratory Animal Center of Academy of Military Medical Sciences in accordance with institutional guidelines. Mouse operations were approved by the Animal Care and Use Committee of the institute. The Hlf tdTomato/+ reporter mouse line and the Hlf CreER/+ lineage tracing mouse line were generated with the CRISPR/Cas9-mediated gene knockin technique by Beijing Biocytogen and Shanghai Model Organisms Center, respectively. The ROSA ZsGreen/ZsGreen reporter mice were described previously (Madisen et al., 2010). CD45.1/1 mice were purchased from Jackson Laboratory. CD45.2/2 mice were purchased from SPF (Beijing) Biotechnology Co., Ltd. All mice were maintained on C57BL/6 background. Embryos were staged by the number of somite pair (sp): E10.5, 36-40 sp; E11.0, 41-45 sp; and E11.5, 46-50 sp. The AGM region was dissected as previously described (Li et al., 2012).

OP9-DL1 Co-culture and Transplantation Assay
The OP9-DL1 stromal cells were thawed 4 days before coculture, and 3 × 10 4 OP9-DL1 stromal cells (passage 3-7) were seeded into a 24-well plate the day before co-culture without irradiation. A total of 15 AGM regions were dissected and digested into single-cell suspension. The cells from AGM regions were stained with antibodies and sorted by flow cytometry with the gating strategies for indicated populations. Then, the sorted cells were seeded into a 24-well plate containing OP9-DL1 feeder cells with 3 ee (embryo equivalents) per well and incubated in α-MEM (Gibco) with 10% fetal bovine serum (Hyclone) and cytokines (100 ng/ml SCF, 100 ng/ml IL-3, and 100 ng/ml Flt3 ligand, PeproTech) for 6 days. On the fourth day, a half amount of cell culture medium mentioned above was added to the wells. Six days later, the co-cultured cells in each well were detached by 0.25% trypsin (Beyotime Biotechnology, C0203) and harvested independently for flow cytometry analysis or transplantation assay. For transplantation, male CD45.1/1 were mated with female Hlf tdTomato/+ (CD45.2/2) to obtain CD45.1/2 embryos. Cell suspension harvested after 6 days of co-culture were mixed with 2 × 10 4 nucleated BM helper cells (CD45.2/2). The mixture was injected via the tail vein into 8-12week-old female recipients (CD45.2/2) which had been exposed to a split dose of 9 Gy γ-irradiation ( 60 Co). The recipients demonstrating ≥5% donor-derived chimerism in peripheral blood were considered as successfully reconstituted. Donor contribution to each blood lineage is calculated as: (percentage of donor cells of a given lineage in CD45 + cells/total percentage of a given lineage in CD45 + cells) × 100%; Donor contribution to LSK (Lin − Sca1 + Kit + ) cells is calculated as: percentage of donor cells in LSK cells × 100% (Benz et al., 2012;Li et al., 2012;Ye et al., 2017).

Immunofluorescence
Embryos were isolated and fixed with 4% paraformaldehyde for 2-4 h at 4 • C, embedded in paraffin and sectioned into 5-µm slices using Leica RM2235. Sections were placed in an oven at 60 • C for 15 min, deparaffinized with ethanol of gradient concentration, and subsequently placed in a microwave for 20 min at 95 • C within citrate buffer (pH 6.0). The sections were cooled to room temperature and then placed in 3% H 2 O 2 for 20 min to remove endogenous peroxidase. Sections were blocked in blocking solution (1:1, Cat#ZLI-9056, Zhongshan golden bridge) for 30 min at room temperature, and then incubated with primary antibodies overnight at 4 • C. After being washed with PBS three times, sections were incubated with corresponding secondary antibodies (Zhongshan golden bridge) for 30 min at room temperature. After being washed with PBS, sections were stained with DendronFluor TSA (Histova, NEON 4-color IHC Kit for FFPE, NEFP450, 1:100, 20-60 s). After the staining of the first antigen was completed, we thoroughly eluted the primary and secondary antibodies by re-heating the slides in a microwave with citrate buffer (pH 6.0) for 20 min at 95 • C, and repeated the above steps from the microwave repair step to the DendronFluor TSA staining step, so that each antigen was labeled by distinct fluorophores finally. After all the antibodies were stained, the slices were stained with DAPI (4' ,6-diamidino-2phenylindole). Images were collected by confocal microscope (Nikon Ti-E A1/ZEISS LSM 880). The primary antibodies and their clone number, dilution ratio, catalog number, and company are listed as follows: Rabbit polyclonal anti-RFP (1:1,500, Cat#600-401-379, Rockland); Rabbit monoclonal anti-Estrogen Receptor alpha (SP1, 1:200, Cat#ab16660, Abcam), Rabbit monoclonal anti-RUNX1/AML1 + RUNX3 + RUNX2 (EPR3099, 1:150, Cat#ab92336, Abcam), and Rabbit monoclonal anti-CD31 (EPR17259, 1:1,000, Cat#ab182981, Abcam). The second antibody was HRP-labeled Goat anti-Rabbit IgG polymer (1:1, Cat#PV6001, Zhongshan golden bridge).

Inducible Genetic Lineage Tracing
Tamoxifen free base (T5648-5G, Sigma) 200 mg was dissolved in 10 ml sunflower-seed oil (S5007; Sigma) and shaken overnight at the room temperature. For lineage tracing, Hlf CreER/+ mice were firstly bred with ROSA ZsGreen/ZsGreen mice to generate Hlf CreER/+ ; ROSA ZsGreen/+ (referred as Hlf-CreER;ZsGreen), then male Hlf-CreER;ZsGreen mice were crossed with female C57BL/6 mice. The pregnant C57BL/6 mice were administered with a single dose of tamoxifen by gavage (3 mg per 30 g body weight) at E9.5. At indicated embryonic stages, the pregnant C57BL/6 mice were sacrificed, the embryos were isolated with subsequent genotyping, and the Hlf-CreER;ZsGreen embryos were used for further flow cytometry analysis, with the ROSA ZsGreen/+ embryos from the same litter serving as controls. For adult mice analysis, Hlf-CreER;ZsGreen offspring with a single dose of tamoxifen induction at E9.5 as described above were used for flow cytometry analysis.

Single Cell RNA-Seq Dataset Source
The scRNA-seq data are from our previous study (Zhou et al., 2016) and downloaded from the GEO database (GSE67120).

Establishment of an Hlf-tdTomato Reporter Mouse Model
By analyzing the transcriptomic dataset we previously constructed (Zhou et al., 2016), we found that Hlf was expressed in almost all of the HSC-competent cells from embryo to adult but not in endothelial cells of the mid-gestational AGM region, with a gradual up-regulation from pre-HSCs to adult HSCs ( Figure 1A). Therefore, for a better understanding about the stepwise specification of nascent HSCs, we generated an Hlf-tdTomato reporter mouse model using the gene knockin strategy to insert a P2A-tdTomato cassette between exon 4 and stop codon of Hlf gene locus without disrupting its expression ( Figure 1B).
We firstly evaluated the expression pattern of Hlf-tdTomato in hematopoietic populations of the E11.0 AGM region, E15.5 FL, and adult BM and peripheral blood by fluorescence activated cell sorting (FACS) analyses. With the use of the highly functionally enriched markers of pre-HSCs (Zhou et al., 2016), we showed that about half of the immunophenotypic T1 pre-HSCs and most if not all of T2 pre-HSCs at E11.0 were positive for Hlf-tdTomato (Figures 1C,D). Furthermore, Hlf-tdTomato was expressed in more than 80% of immunophenotypic HSCs and multi-potent progenitors, much higher than in common myeloid progenitors, but was seldom expressed in other committed progenitors and mature blood cells in E15.5 FL ( Figure 1C and Supplementary Figure 1A). In adult BM, Hlf-tdTomato was expressed in almost all the immunophenotypic long-term HSCs and short-term HSCs, irrespective of the marker combinations used, and also in a predominant portion of multi-potent progenitors and in about half of common myeloid progenitors, but hardly in other progenitors and mature blood cells (Figure 1C and Supplementary  Figures 1B,C). Analysis of peripheral blood showed that mature blood lineages, including granulocytes/monocytes, B cells, and T cells, lacked Hlf-tdTomato expression ( Figure 1C and Supplementary Figure 1D). The expression pattern of Hlf-tdTomato in various hematopoietic populations at different developmental stages was not only basically in accordance with our transcriptomic data ( Figure 1A) but also principally in line with a previous report (Yokomizo et al., 2019) validating the successful construction of the Hlf-tdTomato reporter mouse model. We next investigated the expression of Hlf-tdTomato at E10.5 by immunostaining and flow cytometry. Immunostaining on sections of the E10.5 AGM region showed that Hlf-tdTomato was expressed in the IAHC cells, co-expressing hematopoietic transcription factor Runx1 and endothelial marker CD31, in line with a previous report (Yokomizo et al., 2019). On the other hand, Hlf-tdTomato expression was not detected in the endothelial layer of the dorsal aorta ( Figure 1E). Consistently, FACS analyses indicated that Hlf-tdTomato expression was not detected in the immunophenotypic non-hematopoietic cells (CD41 − CD43 − CD45 − ) from multiple intra-embryonic sites, including the AGM region, head, trunk, and limb. In contrast, its expression was confined to the immunophenotypic hematopoietic cells (CD41 + /CD43 + /CD45 + ) ( Figure 1F and Supplementary Figures 2A,B). We also noticed that only 0.069 ± 0.013% cells in the E10.5 AGM region expressed Hlf-tdTomato, in which 17.0 ± 3.4% cells were CD45-negative (Figures 1G,H). This expression pattern suggested that Hlf-tdTomato expression may initiate from CD45 − pre-HSCs along HSC specification.
For in vivo functional validation, Hlf-tdTomato + and Hlf-tdTomato − cells within CD31 + Kit + CD45 − population from the E10.5 AGM region were, respectively, co-cultured with OP9-DL1 for 6 days, then the culture products were transplanted into lethally irradiated adult mice. At 16 weeks post-transplantation, multi-lineage repopulation was observed in peripheral blood and multiple organs (BM, spleen, and thymus) of the recipients of both Hlf-tdTomato + (2/4) and Hlf-tdTomato − (2/5) groups. Moreover, the contribution of the donor cells to the immunophenotypic HSCs and hematopoietic progenitors in BM was also observed, validating the bona fide HSC competence (Figures 2E-J and Supplementary  Figures 3A,B). These data indicated that precursors of HSCs were distributed in both Hlf-tdTomato + and Hlf-tdTomato − subpopulations of CD45 − cells in the E10.5 AGM region. Considering the restricted expression of Hlf-tdTomato in the hematopoietic cells (Figure 1F), the CD45 − Hlf-tdTomato + population with the HSC-competence should belong to T1 pre-HSCs. On the other hand, as HSC-competent HECs in the AGM region represented by PK44 did not express Hlf-tdTomato (Supplementary Figure 2B), the repopulating capacity of the progenies of CD31 + Kit + CD45 − Hlf-tdTomato − population should be partially due to the existence of HSC-primed HECs therein, although the existence of Hlf-tdTomato − pre-HSCs could not be completely excluded.
Next, Hlf-tdTomato + and Hlf-tdTomato − cells within the CD31 + Kit + CD45 + population from the E10.5 AGM region were, respectively, isolated and co-cultured with OP9-DL1 for 6 days, then transplantation assay was performed with the culture products. Of note, multi-lineage repopulation in peripheral blood and multiple organs (BM, spleen, and thymus) at 16 weeks post-transplantation was only detected in the recipients of Hlf-tdTomato + group, consistent with the detection of the chimerism in the immunophenotypic HSCs and hematopoietic progenitors in BM (Figures 3F-K and Supplementary Figures 3C,D). Therefore, Hlf-tdTomato expression further enriched the HSCcompetence in E10.5 CD45 + population. The finding also indicated that if a cell expresses CD45 prior to Hlf at this early stage, it has lost the possibility of specification toward HSCs.

Generation of an Inducible Hlf-CreER Mouse Model
To determine the physiological contribution of the Hlfexpressing early cells, including pre-HSCs and nascent HSCs, to FL and adult hematopoiesis, we newly generated an Hlf-CreER mouse model by CRISPR/Cas9-mediated gene knockin strategy, with an inducible CreERT2 cassette inserted between exon 4 and 3 UTR of the Hlf gene locus (Figure 4A). Immunostaining of ER in the AGM region showed the expression of the inducible Cre recombinase specifically in the IAHC cells from E9.5 to E10.5 (Figure 4B), consistent with the expression pattern of Hlf-tdTomato ( Figure 1E).
Next, Hlf-CreER;ROSA-LSL-ZsGreen (referred to as Hlf-CreER;ZsGreen) mice were generated for lineage tracing. We performed the single-dose tamoxifen induction at E9.5, 24 h before the first HSCs appear at E10.5 (Figure 4C), and the labeling would involve the pre-HSCs and emerging HSCs marked by Hlf expression mainly within the next 48 h from injection (Zovein et al., 2008). We first assessed the labeling proportions of several populations related to HSC ontogeny by FACS analyses, including endothelial cells, IAHC cells, as well as T1 and T2 pre-HSCs in the E11.5 AGM region. Endothelial cells were not labeled as expected (Figure 4D and Supplementary Figure 4A). In contrast, the labeling was found in 3.9 ± 0.7% of the immunophenotypic IAHC cells (CD31 + Kit hi ) ( Figure 4D and Supplementary Figure 4A). Compared with that in T1 pre-HSCs (CD31 + CD45 − CD41 lo Kit + CD201 hi , 1.2 ± 0.8%), the average constitution of labeled cells in T2 pre-HSCs (CD31 + CD45 + Kit + CD201 hi ) was higher (5.1 ± 2.5%) ( Figure 4D and Supplementary Figure 4A), in line with the higher expression of Hlf-tdTomato in T2 pre-HSCs than in T1 pre-HSCs ( Figure 1D). Interestingly, the labeling of pre-HSCs in individual AGM regions was concentrated in either T1 or T2 pre-HSCs (Supplementary Figure 4B), suggesting the transient dynamics of pre-HSC maturation.
We then evaluated the contribution of these Hlf-labeled early cells to the immunophenotypic HSCs in FL. Of note, the average constitution of lineage-labeled cells in FL HSCs (Lin − Sca-1 + Mac-1 lo CD201 + ) (Zhou et al., 2016) was 13.6 ± 1.8% at E13.5 and 9.4 ± 1.6% at E15.5, higher than that in pre-HSCs in the E11.5 AGM region (Figures 4D,E). When using another surface marker combination ESLAM (CD45 + CD201 + CD150 + CD48 − ) (Benz et al., 2012) to check E15.5 FL HSCs, we came to a similar conclusion (Supplementary Figure 4C). These results indicated that Hlf-CreER;ZsGreen was able to trace part of AGM pre-HSCs and/or nascent HSCs without labeling endothelial cells when induced at E9.5, and the progenies of the labeled cells could migrate and colonize FL, possibly with an expansion to some extent.

Hlf-Labeled Early Cells Contribute to Long-Term and Multi-Lineage Adult Hematopoiesis
To evaluate the contribution of Hlf-labeled early cells to adult hematopoiesis, we administrated a single dose of tamoxifen at E9.5, and the contribution to multiple hematopoietic lineages in peripheral blood was chased up to 32 weeks after birth ( Figure 4C). The labeling in total CD45 + leukocytes was ranged from 5.2 to 7.7%, with little fluctuation from 4 to 32 weeks (Figures 5A,B). The constitutions of traced cells in different lineages were similar, including granulocytes/monocytes (CD45 + Gr1/Mac-1 + ), B lymphocytes (CD45 + B220 + ), T lymphocytes (CD45 + CD3 + ), erythrocytes (CD45 − Ter119 + ), and megakaryocytes (CD45 − CD41 + ), with on average of 5.2-8.1% for all the lineages and all the time points detected (Figures 5A,B). To further determine whether and to what extent the Hlflabeled early cells contributed to BM hematopoiesis in adult, we analyzed BM samples from three 8-9-month-old mice by flow cytometry. The constitutions of lineage-labeled cells in the immunophenotypically defined committed hematopoietic progenitors, including common myeloid progenitors, granulocyte-monocyte progenitors, and megakaryocyteerythroid progenitors, were similar to each other, and also to that in the upstream HSC-enriched LSK (Lin − Sca1 + Kit + ) cells as well as to those in the mature blood lineages in the peripheral blood (Figures 5C,D). Interestingly, when using different marker combinations to define the enriched long-term HSCs, short-term HSCs, and multi-potent progenitors, the results varied regarding the contribution of lineage-traced cells to each of these HSPC populations (Figures 5E,F), suggestive of the heterogeneity within these immunophenotypically defined HSPC populations.
Taken together, these data unambiguously indicated that the early Hlf-expressing cells in the embryo contribute to and retain the size of the HSC pool in the BM, which continuously differentiate to maintain a balanced and long-term multi-lineage hematopoiesis in the adult. Therefore, here we provided a valuable mouse model to specifically trace the fate of emerging HSCs in the embryo.

DISCUSSION
Considering that Hlf is an HSC-related gene that is specifically expressed in both embryonic and adult HSCs but neither in primitive blood cells nor erythro-myeloid progenitors, which is conserved in both mouse and human (Shojaei et al., 2005;Magnusson et al., 2007;Yokomizo et al., 2019;Zeng et al., 2019), here we specifically chose it and successfully constructed two mouse models, Hlf-tdTomato reporter mice and Hlf-CreER lineage tracing mice, which would without doubt serve as valuable mouse models for studying HSC biology.
In addition to validating the expression pattern of Hlf-tdTomato in FL and BM, which was principally consistent with the findings in a recent study (Yokomizo et al., 2019), we made an effort in the present study to explore the expression pattern of Hlf-tdTomato in early pre-HSCs with functional evaluation. Previous study has reported the presence of CD45 − pre-HSCs in the E10.5 AGM region (Rybtsov et al., 2011); however, whether there exist CD45 + pre-HSCs at E10.5 has been controversial (Boisset et al., 2015). Based on the in vitro co-culture system we used, the progenies of CD45 + Hlf-tdTomato + but not CD45 + Hlf-tdTomato − cells in the E10.5 AGM region were capable of repopulating irradiated adult recipients. Thus, we demonstrated the existence of CD45 + pre-HSCs in E10.5 AGM, which express Hlf. Moreover, our finding suggested that Hlf expression prior to CD45 expression is a prerequisite for the cells to keep the HSC competence. Unlike the CD45 + cells, CD45 − cells showed repopulating activity in both Hlf-tdTomato + and Hlf-tdTomato − counterparts. However, presently, we could not figure out to what extent the HSCprimed HEC population contributed to the repopulating capacity of the Hlf-tdTomato − sub-population in the E10.5 AGM region. Further precise study designs are needed to address this question. It is conceivable that Hlf-tdTomato reporter can be used as an indicator of HSC emergence when carrying out reprogramming experiment or inducing ES cell differentiation in vitro.
Although a variety of mouse models have been used to trace the fate of HSCs, there is still room for improvement. Some mouse models have been used to mark adult HSCs, such as Pdzk1ip1-CreERT2, Fgd5-CreERT2, Vwf-CreERT2, Krt18-CreERT2, and Tie2-MerCreMer, with varied efficiencies to label adult HSCs when induced at adult stage Sawai et al., 2016;Carrelha et al., 2018;Chapple et al., 2018;Sawen et al., 2018). However, the expression of most of these molecules is not specific to HSCs during development. Some mouse models, including Cdh5-CreERT2, Tie2-MerCreMer, and Runx1-MerCreMer, have been used to label embryonic HSCs when induced at mid-gestation (Zovein et al., 2008;Gomez Perdiguero et al., 2015;Hoeffel et al., 2015). Since these molecules are also expressed in the cells of other waves of hematopoiesis in addition to HSC generation during embryogenesis, and the occurring period for different waves of hematopoiesis overlaps to a certain extent, lineage tracing models using HSC-specific markers are urgently needed to more specifically and clearly delineate the fate of embryonic HSCs exclusively. Together with a previous study showing that Hlf is not expressed in erythro-myeloid progenitors, a population belonging to HSCindependent hematopoiesis (Yokomizo et al., 2019), our present findings suggested Hlf-CreER mice as an unprecedented model for studying embryonic HSC-derived hematopoiesis.
Our previous studies have reported the expression of CD201 (known as EPCR, encoded by Procr) in HSC-primed HECs and pre-HSCs, which is a stemness molecule and also expressed in aortic endothelial cells, as well as the contribution of CD201labeled embryonic cells to fetal and adult blood lineages revealed by Procr-CreER-mediated lineage tracing when induced at E9.5 (Zhou et al., 2016;Lan, 2017;Zheng et al., 2019;Hou et al., 2020). Moreover, we have identified Neurl3 as a signature gene of HSC-primed HECs and pre-HSCs and achieved the enrichment of HSC-competent HECs by the Neurl3-EGFP reporter (Hou et al., 2020). In the future, by combining these mouse models, we will more precisely delineate the comprehensive roadmap for the diversified fate choice of different hematopoietic lineages from embryonic HSCs.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

ETHICS STATEMENT
The animal study was reviewed and approved by the Academy of Military Medical Sciences (Fifth Medical Center of Chinese PLA General Hospital).

AUTHOR CONTRIBUTIONS
YL and BL designed the study. WT performed the cell sorting, culture, immunostaining, and transplantation assays with help from JHe, TH, ZB, CW, HW, RY, YN, YY, YG, SH, JHo, JW, and JZ. WT, YL, and BL wrote the manuscript and revision. All authors reviewed the manuscript.