The identity crisis of Hif-1α in HSC biology

In this issue of Blood, Vukovic et al provide compelling data that hematopoietic stem cells (HSCs) do not require the transcription factor hypoxia-inducible factor 1α (Hif-1α) to duplicate themselves (ie, self-renew), reconstitute long-term hematopoiesis, or sustain hematopoiesis following injury.¹  These observations affect hematology because they challenge previous notions regarding the role of Hif-1α in HSC biology.² 

Adult HSCs are multipotent progenitors that can sustain themselves as well as give rise to the various blood cell lineages for the lifetime of an organism.³  Mammalian HSCs reside in a specialized microenvironment within the bone marrow (BM) called the HSC niche. HSC behaviors such as self-renewal, differentiation, or quiescence are dictated by the integration of external cues from the HSC niche through cell-intrinsic signals.⁴  Though many HSC-regulating niche factors have been identified, the role of oxygen as a critical physiological factor governing HSC biology has been a recent topic of debate.

High-resolution imaging studies have shown that the BM is a low-oxygen organ⁵  and that quiescent HSCs display molecular characteristics associated with hypoxia (a state of insufficient oxygen availability).⁶  However, others have shown that HSCs display a molecular signature of hypoxia regardless of their surrounding oxygen availability.⁷  Additional studies have shown that the BM is highly vascularized and that HSCs do not preferentially localize to low-oxygen niches⁵  but rather reside close to blood vessels.⁴ 

Hif-1α is a transcription factor that is activated in hypoxic cells and drives transcriptional programs (eg, metabolic adaptation, angiogenesis, proliferation) that allow cells to negotiate hypoxic environments. Hif-1α is activated in mouse HSCs^2,7  and genetic ablation of Hif-1α diminishes the hematopoietic reconstitution abilities of mouse HSCs, suggesting that Hif-1α is an intrinsic regulator of HSC behavior.²  Conversely, deletion of another hypoxia-inducible transcription factor, Hif-2α, either alone or in combination with Hif-1α, does not alter the ability of HSCs to reconstitute hematopoiesis.⁸ 

Using the same genetically engineered strain of Hif-1α floxed mice used in the Takubo et al²  studies, Vukovic et al have found that deletion of Hif-1α in the hematopoietic system does not affect HSC survival or the ability of HSCs to reconstitute long-term, multilineage hematopoiesis. Using serial transplantation assays, Vukovic et al also observed that the absence of Hif-1α does not affect HSC self-renewal and that the loss of Hif-1α expression does not significantly affect how HSCs or other hematopoietic progenitors respond to hematopoietic injury mediated by the myeloablative agent, 5-fluorouracil. The reports put forth by Vukovic et al and Guitart et al⁸  establish a convincing argument that both Hif-1α and Hif-2α are not essential cell-intrinsic regulators of HSC function. However, the reasons the results of these studies conflict with those of Takubo et al²  may potentially be due to subtle differences in study design between the 2 reports (see table).

Although both groups used the Mx1-Cre transgene to induce Hif-1α gene ablation (Hif-1α^fl/fl;Mx1-Cre mice) in the hematopoietic system, the time and specific location of Hif-1α deletion significantly varied between the 2 studies. Although Mx1-Cre efficiently recombines floxed alleles in the hematopoietic system, it is well-documented that it is expressed in components of the BM microenvironment⁹  and sites of extramedullary hematopoiesis.¹  To circumvent the nonhematopoietic effects of the Mx1-Cre transgene, Vukovic et al transplanted whole BM cells from Hif-1α^fl/fl;Mx1-Cre mice into wild-type recipients and then induced Hif-1α deletion, thus selectively deleting Hif-1α in the hematopoietic system. In contrast, the vast majority of the HSC analyses carried out by Takubo et al²  were performed in a context in which Hif-1α deletion was not restricted to the hematopoietic system.

In the subset of experiments in which Takubo et al²  did assess the HSC-autonomous role of Hif-1α, they observed that the loss of Hif-1α expression enhanced multilineage reconstitution at 4 months postinduction of gene deletion. However, follow-up analysis at 11 months post-Hif-1α deletion (15 months posttransplant) revealed that the absence of Hif-1α significantly reduced multilineage reconstitution. These timelines differ from the HSC-autonomous studies conducted by Vukovic et al, who assessed multilineage reconstitution at 2 and 32 weeks following Hif-1α deletion (10 and 40 weeks posttransplant). Interestingly, studies evaluating the impact of an aging niche on HSC behavior^3,10  have shown that the HSC function is differentially affected by young and old niches, raising the possibility that Hif-1α may influence the behavior of HSCs exposed to aged niches.

Last, although Takubo et al²  observed that unrestricted deletion of Hif-1α in both the hematopoietic system and the BM microenvironment reduces the frequency of functional HSCs in the BM, they also observed that the spleens of Hif-1α–deleted mice harbored significantly more functional HSCs than control mice. These observations are of particular interest when considering the well-recognized phenomena that extramedullary hematopoiesis often occurs in the presence of faulty BM microenvironment support.

The results of Vukovic et al have convincingly shown that Hif-1α is not an essential cell intrinsic regulator of relatively young HSC function. However, further studies are needed to determine whether Hif-1α deletion in nonhematopoietic BM microenvironment cell populations influences HSC behavior as well as if intrinsic Hif-1α expression is needed when HSCs are exposed to an aged niche. Determining how HSCs interact with their microenvironment and regulate their cell intrinsic status is critical to understanding the etiology of blood malignancies as well as to maximizing the application of HSCs in regenerative and transplantation medicine.