Although the molecular mechanisms responsible for hematopoietic stem cell (HSC) mobilization in response to G-CSF or chemotherapy have not been fully elucidated, various murine models demonstrate that both the expansion of the neutrophil pool within the bone marrow (BM) and the down-regulation of the chemokine CXCL12/SDF-1 play a critical role. The main function of neutrophils is to destroy pathogens and necrotic tissues during inflammation. Inflamed and damaged tissues are often very hypoxic due to local disruption of blood supply. Although most cells are stressed in hypoxic conditions, neutrophils and macrophages are attracted and activated by hypoxia, a property dependent on the stabilization of hypoxia-inducible transcription factor-1α (HIF-1α) which is absolutely required for neutrophil and macrophage function. Mice with a conditional deletion of the HIF-1α gene specifically in myeloid cells cannot mount an inflammatory response with neutrophils and macrophage displaying impaired motility, invasiveness and bacterial killing. We therefore hypothesized that hypoxia may play a major role in HSC mobilization. Considering that granulocyte/monocyte progenitors consume 30 times more O2 than their fully differentiated progeny and that mobilized BM is enriched in those progenitors, we calculated that mobilized BM should consume twice more O2 than steady-state BM (3.40μmole O2/femur/hour vs 1.64μmole O2/femur/hour in the mouse). According to the Krogh’s model that predicts O2 distribution in solid tissues, this increased O2 consumption should lead to O2 exhaustion resulting in hypoxia within mobilized BM. We demonstrated this in vivo by injecting the hypoxia-specific probe pimonidazole hydrochloride into mobilized and non-mobilized BALB/c mice 3 hours prior to sacrifice. Pimonidazole covalently binds to tissue proteins when O2 tension is below 10mmHg (O2 tension is 95mmHg in arterial blood, 40mmHg in venous and capillary blood and 150mmHg in the atmosphere at sea level). Staining of femoral BM with a monoclonal antibody (mAb) specific for pimonidazole showed that the hypoxic area (≤10mmHg or <1.3% O2) was restricted to the endosteal region of non-mobilized BM, but extended throughout the BM at the peak of G-CSF-induced (day 4) and CY-induced (day 7) mobilization. Western-blots with an anti-HIF-1α mAb showed a 20–30 fold increase in HIF-1α protein in mobilized BM cell lysates compared to steady-state BM cell lysates, a result confirmed by immunohistochemistry on femur sections. As one of the genes activated by HIF-1 transcription factor is VEGF-A and administration of VEGF-A causes HSC mobilization, VEGF-A transcripts were quantified by qRT-PCR and found to be significantly increased in mobilized BM. Taken together these data demonstrate that the increase in granulocyte progenitors resulting from the administration of G-CSF or CY, depletes O2 from the BM hematopoietic microenvironement. As a result, most of the BM extravascular compartment becomes markedly hypoxic with stabilization of HIF-1α transcription factor in BM cells and increased transcription of VEGF, a potent mobilizing cytokine that increases blood vessel permeability. Furthermore, as hypoxia has been shown to inhibit osteoblast proliferation and function in vitro, increased hypoxia may also be in part responsible for the strong inhibition of osteoblast function, bone formation and CXCL12 transcription observed during HSC mobilization.

Disclosures: Antisense Therapeutics Pty Ltd.

Author notes

*

Corresponding author

Sign in via your Institution