Hypoxia-responsive genes are up-regulated in LSCs. BCR-ABL–expressing LSK cells and control LSK cells were sorted from mice receiving BCR-ABL or empty vector-transduced BM cells for DNA microarray analysis. (A) The heatmap compares the activation of the HIF1α signaling pathway in BCR-ABL–expressing LSK cells and control LSK cells. (B) Gene set enrichment analysis (GSEA) displays the expression profiling of HIF1α targets in BCR-ABL–expressing LSK cells (P < .001). (C) Microarray data showed higher expression of HIF1α, vascular endothelial growth factor (VEGF), glucose transporter type 1 (GLUT1), and transforming growth factor α (TGFα) in BCR-ABL–expressing LSK cells comparing to normal LSK cells. (D) Real-time RT-PCR was performed with primers specific for HIF1α, VEGF, GLUT1, TGFα, and PGK1. The results were normalized using actin as a control, and shown as mean ± SD. *P < .05; **P < .01.

HIF1α is essential for CML development. (A) Deletion of HIF1α in hematopoietic cells. Mice carrying the floxed HIF1α allele were crossed with Vav-Cre transgenic mice in which the Vav promoter drives the expression of the Cre recombinase. RT-PCR analysis showed that HIF1α was undetectable in sorted HSCs (Lin⁻Sca-1⁺c-Kit⁺) from HIF1α^flox/flox-Vav-Cre mice. (B) Kaplan-Meier survival curves for primary and secondary recipients of empty vector or BCR-ABL–transduced BM cells from WT or HIF1α^−/− donor mice. For secondary BM transplantation, BM cells from primary control and CML recipient mice which received empty vector or BCR-ABL–transduced WT and HIF1α^−/− BM cells were analyzed by FACS, and BM cells containing equal number of WT or HIF1α^−/− GFP⁺Lin⁻Sca-1⁺c-Kit⁺ cells along with 2 × 10⁵ WT BM cells (CD45.1) were transplanted into each lethally irradiated secondary recipient mouse. (C) FACS analysis of GFP⁺Gr-1⁺ cells in PB of primary and secondary recipients of empty vector or BCR-ABL–transduced BM cells from WT or HIF1α^−/− donor mice (n = 5). Mean values (± SD) are shown. (D) The total numbers of GFP⁺Gr-1⁺ cells in PB of primary and secondary recipients of empty vector or BCR-ABL–transduced BM cells from WT or HIF1α^−/− donor mice (n = 5). Mean values (± SD) are shown. *P < .05; **P < .01; ***P < .001. (E) The spleen weight of primary and secondary recipients of empty vector or BCR-ABL–transduced BM cells from WT or HIF1α^−/− donor mice (n = 12). Mean values (± SD) are shown; ***P < .001. (F) Gross appearance of the lungs and spleens showed severe lung hemorrhages and splenomegaly in secondary recipients of WT LSCs but not HIF1α^−/− LSCs at day 14 after BMT, compared with those from recipients of WT and HIF1α^−/− HSCs. (G) H&E staining of tissue sections from lung and spleen of secondary recipients. The scale bar represents 50 μm. (H) Kaplan-Meier survival curves for the secondary BM transplantation. BM cells from WT or HIF1α^flox/flox mice were transduced with BCR-ABL-iCre-GFP retrovirus, and transplanted into lethally irradiated recipient mice. CML cells without HIF1α failed to induce CML disease in the secondary BMT. (I) The percentages of leukemia cells in PB at day 14 in WT and HIF1α^−/− secondary CML mice.

Effects of HIF1α deletion on retroviral transduction efficiency and the homing ability of normal and BCR-ABL–transduced total BM cells or LSK cells. (A) HIF1α deletion does not affect retroviral transduction efficiency. WT and HIF1α^−/− BM cells were transduced with empty vector or BCR-ABL-GFP retrovirus, and 2 days later, the percentages of GFP⁺ cells were determined by FACS. (B) FACS analysis showed the similar homing ability of WT and HIF1α^−/− BM cells. A total of 3 × 10⁶ BM cells from WT or HIF1α^−/− mice (CD45.2) were transplanted into lethally irradiated recipients (CD45.1). The donor-derived BM cells (CD45.2) were detected by FACS in 3 hours after BMT (n = 4). Mean values (± SD) are shown. NS indicates no significance. (C) HIF1α deletion does not cause a reduction of the homing ability of BCR-ABL–transduced BM cells. A total of 2 × 10⁶ leukemia cells (GFP⁺) from CML mice transplanted with BCR-ABL–transduced WT or HIF1α^−/− BM cells were transplanted into lethally irradiated recipients (CD45.1). Three hours later, donor-derived leukemia cells (GFP⁺) in the BM were detected by FACS (n = 4 for each group). Mean values (± SD) are shown; **P < .01. (D) HIF1α deletion does not cause a reduction of the homing ability of control or BCR-ABL–transduced stem cells. BM cells were transduced with GFP vector or BCR-ABL-GFP, and 2 × 10⁴ sorted normal LSK cells and BCR-ABL–expressing LSK cells (GFP⁺Lin⁻Sca-1⁺c-Kit⁺) were transplanted into lethally irradiated recipients (CD45.1). Three hours later, donor-derived leukemia cells (GFP⁺) in the BM were detected by FACS (n = 3 for each group). Mean values (± SD) are shown.

HIF1α deletion suppresses LSCs. (A) FACS analysis of LSCs in BM of secondary recipients of WT or HIF1α^−/− BM cells from primary CML mice. (B-C) The percentages and total numbers of stem cells (GFP⁺Lin⁻Sca-1⁺c-Kit⁺ cells and GFP⁺CD48⁻CD150⁺Lin⁻Sca-1⁺c-Kit⁺ cells) in BM of secondary recipient mice receiving BM cells from WT or HIF1α^−/− control and CML mice were analyzed by FACS; *P < .05; **P < .01. (D) The loss of HIF1α causes a decrease in the colony-forming ability of BCR-ABL–expressing LSK cells. Sorted normal LSK cells and BCR-ABL–expressing LSK cells (GFP⁺Lin⁻Sca-1⁺c-Kit⁺ cells) were plated into methycellulose medium, colonies were counted, and cells were serially replated. Data show results from representative experiments. Mean values (± SEM) are shown; *P < .05; ***P < .001.

HIF1α is required for survival maintenance of LSCs. (A) The cell-cycle analysis of normal LSK cells and BCR-ABL–expressing LSK cells. The percentages of HIF1α^−/− BCR-ABL–expressing LSK cells were low in the S phase and higher in the G₀-G₁ phase, compared with WT BCR-ABL–expressing LSK cells (n = 6). However, normal WT and HIF1α^−/− LSK cells showed similar percentages of cells in G₀-G₁, S, and G₂-M phases. (B) The cell-cycle analysis of BCR-ABL–expressing LSK cells using Ki67 and Hoechst 33 342 showed an accumulation of G₁ phase of HIF1α^−/− BCR-ABL–expressing LSK cells. (C) RT-PCR analysis showed elevated expression of cell-cycle inhibitors p16^Ink4a, p19^Arf, and p57 in the sorted LSK cells and BCR-ABL–expressing LSK cells from HIF1α^−/− control and CML mice compared with the sorted LSK cells and BCR-ABL–expressing LSK cells from WT control and CML mice. BM cells from control and CML mice at day 14 post-BMT were collected, and LSK cells and BCR-ABL–expressing LSK cells were sorted for extracting total RNA. Mean values (± SD) are shown; **P < .01; ***P < .001. (D) The apoptosis of normal LSK cells and BCR-ABL–expressing LSK cells from BM of control and CML mice (n = 6). HIF1α deletion induced the apoptosis of BCR-ABL–expressing LSK cells, but did not affect the apoptosis of normal LSK cells. (E) FACS analysis showed a higher percentage of active caspase positive HIF1α^−/− leukemia progenitors. (F) RT-PCR analysis showed elevated expression of p53 in the sorted BCR-ABL–expressing LSK cells from HIF1α^−/− CML mice comparing to the sorted BCR-ABL–expressing LSK cells from control CML mice. Mean values (± SD) are shown; **P < .01. (G) Simultaneous knockdown of p16^Ink4a and p19^Arf rescued the colony-forming ability of HIF1α^−/− BCR-ABL–expressing LSK cells. Western blot analysis showed that protein levels of p16^Ink4a and p19^Arf in p53^−/− MEF and CML leukemia cells were dramatically reduced by shRNA no. 1 that targets both p16^Ink4a and p19^Arf. BM cells from WT and HIF1α^−/− normal or CML mice were infected by shRNA no. 1, and plated into methycellulose media after selection by puromycin. Colonies were counted and cells from the colonies were serially replated. Mean values (± SD) are shown; **P < .01; ***P < .001. (H) Loss of HIF1α did not affect the senescence of normal LSK cells and BCR-ABL–expressing LSK cells. Sorted GFP⁺Lin⁻Sca-1⁺c-Kit⁺ cells from the BM of WT or HIF1α^−/− control and CML mice were stained by X-Gal, and the number of β-galactosidase–expressing cells were counted.

Effects of HIF1α deficiency on normal HSCs. (A) RT-PCR analysis for the expression of HIF1α, VEGF, GLUT1, TGFα, and PGK1 in WT or HIF1α^−/− normal LSK cells and BCR-ABL–expressing LSK cells. (B) Gene set enrichment analysis of DNA microarray data displays gene expression profiling of HIF1α targets in WT or HIF1α^−/− normal LSK cells and BCR-ABL–expressing LSK cells. (C) FACS analysis of HSCs in BM cells from WT or HIF1α^−/− mice. The percentages of LSK cells (Lin⁻Sca-1⁺c-Kit⁺), LT-HSCs (Lin⁻Sca-1⁺c-Kit⁺CD34⁻CD135⁻), and ST-HSCs (Lin⁻Sca-1⁺c-Kit⁺CD34⁻CD135⁺) were similar in WT and HIF1α^−/− mice; however, HIF1α^−/− mice showed a higher percentage of MPP (Lin⁻Sca-1⁺c-Kit⁺CD34⁺CD135⁺) (n = 5). (D) Similar percentages of CD41⁻CD48⁻CD150⁺ Lin⁻Sca-1⁺c-Kit⁺ cells in BM cells of WT or HIF1α^−/− mice (n = 3).