Hif-2α is not essential for cell-autonomous hematopoietic stem cell maintenance

Hematopoietic stem cells (HSCs) reside in the hypoxic bone marrow (BM) microenvironment, which modulates metabolism of HSCs and is ultimately important for HSC maintenance.^1,2  Hypoxia-inducible factor-1 (Hif-1) and Hif-2 are key mediators of cellular responses to hypoxia and regulate gene expression to facilitate adaptation to hypoxia. Oxygen-regulated α-subunits of Hif-1 and Hif-2, namely Hif-1α and Hif-2α, are paralogues that have common and also distinct functions during responses to hypoxia.³  Both Hif-1α and Hif-2α transcripts are present at higher levels in HSCs compared with primitive progenitor cell populations.⁴  Although several studies established the essential role for Hif-1α in HSC functions,^4-7  the cell-autonomous role for Hif-2α in HSC maintenance remains unclear. Here, we demonstrate that HSCs do not require Hif-2α to self-renew and sustain long-term multilineage hematopoiesis and that additional deletion of Hif-1α has no impact on Hif-2α–deficient HSC functions.

Hif-1α^fl/fl, Hif-2α^fl/fl, Mx1-Cre, and Vav-iCre mice have been described previously.^8-11  Sex-matched mice, 9 to 12 weeks old, were used for all analyses of steady-state hematopoiesis. Animal experiments were authorized by UK Home Office.

All BM samples were stained and analyzed as described previously.^9,10 

All donor mice were CD45.2⁺. For HSC transplantations, 100 Lin⁻c-Kit⁺Sca-1⁺ (LSK)CD48⁻CD150⁺ HSCs sorted from donor BM were transplanted together with 200 000 CD45.1⁺ BM cells. For secondary transplantations, we injected 2000 CD45.2⁺ LSK cells (sorted from BM of primary recipients) together with 200 000 CD45.1⁺ BM cells. To delete floxed alleles, the recipients received 6 doses of poly(I)-poly(C) (pIpC; 0.3 mg per dose) 8 weeks after transplantation.⁹ 

Statistical significance was determined using 2-tailed Student t tests assuming unequal variance.

Although Hif-2α is required for hematopoiesis in a noncell autonomous manner,^12,13  its role in cell-autonomous HSC maintenance remains unknown. To address this, we deleted Hif-2α specifically in the hematopoietic system using Vav-iCre. Vav-iCre mice¹¹  constitutively express the codon-improved Cre (iCre)¹⁴  driven by the Vav regulatory elements,¹⁵  resulting in hematopoietic-specific gene deletion shortly after the emergence of definitive HSCs¹⁶  and ensuring recombination in all HSCs.¹⁷  We bred Hif-2α^fl/fl mice (in which exon 2, encoding the basic helix-loop-helix domain, is flanked by loxP sites)^8,12,18  to Vav-iCre mice.^10,11  We found that Hif-2α^fl/fl;Vav-iCre mice (which lack wild-type Hif-2α transcript in their hematopoietic cells; supplemental Figure 1) were born at normal Mendelian ratios and matured to adulthood without obvious phenotypes (results not shown). Immunophenotypic analyses of Hif-2α^fl/fl;Vav-iCre mice revealed normal numbers of B-lymphoid, myeloid, and erythroid cells in the BM and unaffected numbers of thymic T cells (Figure 1A). BM cells from Hif-2α^fl/fl;Vav-iCre mice and control mice efficiently generated colonies in methylcellulose and had comparable lineage differentiation potential (Figure 1B). Concordantly, the 2 cohorts of mice had similar numbers of Lin⁻c-Kit⁺Sca-1⁻ (LK) myeloid progenitor cells and LSK stem and primitive progenitor cells (Figure 1C-D). The analysis of the BM LSK compartment of Hif-2α^fl/fl;Vav-iCre mice and control mice using CD150 and CD48 markers revealed no statistically significant difference in total numbers (Figure 1E) and frequencies (Figure 1F) of LSKCD48⁻CD150⁺ HSCs and primitive progenitors (LSKCD48⁺CD150⁺ and LSKCD48⁺CD150⁻ cells) between the 2 groups of mice. Subfractionation of the LSK compartment on the basis of CD34 expression confirmed that frequencies and total numbers of LSKCD34⁻ HSCs and LSKCD34⁺ cells (containing short-term HSCs and multipotent progenitors¹⁹ ) were not significantly different between Hif-2α^fl/fl;Vav-iCre mice and control mice (supplemental Figure 2A-B). These experiments revealed that Hif-2α is dispensable for steady-state HSC maintenance and multilineage hematopoiesis.

To investigate the self-renewal capacity of Hif-2α–deficient HSCs, we conducted serial transplantation experiments. We transplanted CD45.2⁺LSKCD48⁻CD150⁺ HSCs sorted from Hif-2α^fl/fl;Vav-iCre mice and control mice together with 200 000 CD45.1⁺ wild-type BM cells into primary CD45.1⁺ recipients (Figure 1G). HSCs of both genotypes efficiently reconstituted long-term hematopoiesis (Figure 1H; supplemental Figure 3) and equally contributed to BM HSC and primitive cell compartments of the recipient mice (Figure 1I-J). Having confirmed the efficient Hif-2α excision in donor-derived Hif-2α^fl/fl;Vav-iCre cells in primary recipients (Figure 1K), we sorted LSKs from primary recipients and transplanted them into secondary recipients. Hif-2α–deficient LSKs sustained long-term reconstitution in secondary recipients comparably to control LSKs (Figure 1L). To test the role for Hif-2α in HSC stress responses, we treated adult Hif-2α^fl/fl;Vav-iCre mice and control mice with 5-fluorouracil (5-FU). We observed a transient reduction of HSC numbers in Hif-2α^fl/fl;Vav-iCre mice and control mice followed by efficient recovery of HSCs of both genotypes (Figure 1M). Therefore, Hif-2α is dispensable for posttransplantation HSC self-renewal and their ability to respond to hematopoietic stress.

Because Vav-iCre recombines during embryogenesis, HSCs in Hif-2α^fl/fl;Vav-iCre mice may activate compensatory mechanisms that bypass Hif-2α deficiency. To rule this out, we examined HSC maintenance following acute Hif-2α deletion. We generated Hif-2α^fl/fl;Mx1-Cre mice in which efficient recombination is induced by treatment with pIpC.^9,20  We mixed CD45.2⁺ BM cells from untreated Hif-2α^fl/fl;Mx1-Cre or control mice with CD45.1⁺ BM cells and transplanted them into recipients (Figure 2A). At 8 weeks after transplantation, the mice received 6 doses of pIpC resulting in efficient Hif-2α deletion in donor-derived CD45.2⁺ cells (Figure 2B). The contribution of CD45.2⁺ cells to the stem and progenitor cell compartments of the recipients was measured 2, 4, and 36 weeks after pIpC treatment (Figure 2C). The percentage of CD45.2⁺ cells in total BM, LSK cell compartments, and in HSCs of the recipients was similar regardless of the genotype of transplanted CD45.2⁺ cells.

Next, Hif-2α–deficient and control LSK cells from primary recipients were transplanted to secondary recipients. Chimerism analysis at 24 weeks after transplantation showed that cells of both genotypes efficiently repopulated secondary recipients (results not shown). Therefore, acute deletion of Hif-2α has no impact on the survival of HSCs and their long-term functions.

Finally, we asked whether maintenance of Hif-2α–deficient HSCs can be compromised by additional Hif-1α deletion. We generated adult Hif-1α^fl/fl;Hif-2α^fl/fl;Vav-iCre mice and control mice and found that they had comparable numbers of BM LSKCD48⁻CD150⁺ HSCs, LSKCD48⁺CD150⁺, and LSKCD48⁺CD150⁻ cells (Figure 2D), and mature hematopoietic cells (Figure 2E), indicating that Hif-1α/Hif-2α–deficient HSCs survive and sustain steady-state hematopoiesis. To test the long-term reconstitution capacity of Hif-1α/Hif-2α–deficient HSCs, we transplanted BM cells from Hif-1α^fl/fl;Hif-2α^fl/fl;Vav-iCre mice and control mice into recipients (Figure 2F) and found that Hif-1α and Hif-2α deficiency (Figure 2G) had no major impact on short-term or long-term reconstitution of recipients (Figure 2H; supplemental Figure 4). Therefore, the ability of Hif-2α–deficient HSCs to sustain steady-state hematopoiesis and repopulate primary recipients is not compromised by additional Hif-1α deletion, indicating that compensatory Hif-1α expression is not predominantly responsible for unperturbed hematopoiesis in the absence of Hif-2α.

Hypoxia within the BM microenvironment is thought to play important roles in the maintenance of the metabolic state of HSCs and hypoxia, and its signaling pathways are implicated in HSC functions.^1,2 Hif-1α deficiency does not severely perturb steady-state hematopoiesis and the ability of HSCs to reconstitute multilineage hematopoiesis in primary recipients, but this deficiency results in the loss of HSC activity after cumulative stress, such as serial transplantation.⁴  Here, we demonstrated that Hif-2α–deficient HSCs sustain steady-state hematopoiesis and efficiently reconstitute multilineage hematopoiesis in serial transplantation assays. We also showed that additional deletion of Hif-1α does not affect steady-state maintenance of Hif-2α–deficient HSCs and their ability to reconstitute hematopoiesis in primary recipients. Further understanding of the detailed functional relationship between these paralogous Hif-α isoforms will be essential to provide important insights into the complexity of Hif-dependent signaling in HSC functions.

The authors thank Celeste Simon and Randall Johnson for providing Hif-2α^fl/fland Hif-1α^fl/fl mice, respectively.

K.R.K. is a Cancer Research UK Senior Cancer Research Fellow. This work was supported by a Senior Cancer Research Fellowship from Cancer Research UK (C29967/A14633) (K.R.K.), grants from Leukaemia & Lymphoma Research (08037, 10030, 11047, 11041) (K.R.K.), Kay Kendall Leukaemia Fund (KKL506) (K.R.K.), the Royal Society (2010R1) (K.R.K.), the National Health Service Greater Glasgow and Clyde Endowment Fund (K.R.K.), and the Cancer Research UK programme grant (A11008) (T.L.H.).

Contribution: K.R.K. designed the experiments and wrote the manuscript; T.L.H., T.E., P.J.R., and P.P. helped with experimental designs; A.V.G. and C. Subramani performed FACS, transplantations, and data analysis; and A.A.-D., M.V., G.S., D.G., C. Sepulveda, and K.D. assisted with mouse husbandry, genotyping and FACS.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Kamil R. Kranc, MRC Centre for Regenerative Medicine, 5 Little France Dr, University of Edinburgh, Edinburgh EH16 4UU, United Kingdom; e-mail: kamil.kranc@ed.ac.uk.