Divergent effects of supraphysiologic Notch signals on leukemia stem cells and hematopoietic stem cells

Abstract

The leukemia stem cell (LSC) hypothesis proposes that a subset of cells in the bulk leukemia population propagates the leukemia.We tested the LSC hypothesis in a mouse model of Notch-induced T-cell acute lymphoblastic leukemia (T-ALL) in which the tumor cells were largely CD4+ CD8+ T cells. LSC activity was enriched but rare in the CD8+ CD4 HSA(hi) immature single-positive T-cell subset. Although our murine T-ALL model relies on transduction of HSCs, we were unable to isolate Notch-activated HSCs to test for LSC activity. Further analysis showed that Notch activation in HSCs caused an initial expansion of hematopoietic and T-cell progenitors and loss of stem cell quiescence, which was followed by progressive loss of long-term HSCs and T-cell production over several weeks. Similar results were obtained in a conditional transgenic model in which Notch activation is induced in HSCs by Cre recombinase. We conclude that although supraphysiologic Notch signaling in HSCs promotes LSC activity in T-cell progenitors, it extinguishes self-renewal of LT-HSCs. These results provide further evidence for therapeutically targeting T-cell progenitors in T-ALL while also underscoring the need to tightly regulate Notch signaling to expand normal HSC populations for clinical applications.

Figure 1

Weak, nonleukemogenic Notch alleles and strong, leukemogenic Notch alleles progressively deplete HSC activity and numbers over time. (A) Lethally irradiated mice were reconstituted with 5-FU–treated donor C57BL/6 BM cells transduced with MigR1 control or activated Notch1 (ΔEGFΔLNRΔP). BM cells at 6 weeks after transplantation were analyzed for donor-derived hematopoietic progenitor cells (GFP⁺Lineage⁻Sca-1⁺c-Kit⁺). Representative flow cytometry plots are shown. (B-C) Lethally irradiated CD45.1⁺ congenic mice were reconstituted with 5-FU–treated donor Rag1^−/− CD45.2⁺ BM cells transduced with MigR1 control or activated Notch1 alleles (ΔEGFΔLNR, ΔEGFΔLNRΔP, and ICN). At 6 weeks (B) and 13 weeks (C) after transplantation, cells were gated on the HSC compartment (Lineage⁻Sca-1⁺c-Kit⁺Flt3⁻) and analyzed for donor reconstitution (GFP⁺). Experiments were performed twice.

Figure 2

Notch progressively depletes HSC activity and numbers over time. (A) Lethally irradiated CD45.1⁺ congenic mice were reconstituted with 5-FU–treated donor Rag1^−/−CD45.2⁺ BM cells transduced with MigR1 control or activated Notch1 (ΔEGFΔLNRΔP). BM DN T-cell production (Thy-1⁺CD25⁺) was measured at 6 weeks and 13 weeks after transplantation. Cells are gated on the CD45.2⁺GFP⁺Lineage⁻ cells. (B) Scatter plot analysis of DN T-cell production by MigR1 controls and Notch mice at 6 and 13 weeks after transplantation. (C) The LT-HSC compartment (Lineage⁻Sca-1⁺c-Kit⁺CD34⁻Flt3⁻) was measured for donor reconstitution by GFP⁺ MigR1 control and Notch-activated HSCs at 6 and 13 weeks after transplantation. (D) Scatter plot analysis of donor-derived (GFP⁺) HSCs (Lineage⁻Sca-1⁺c-Kit⁺CD34⁻Flt3⁻) at 6 and 13 weeks after transplantation. Experiments were performed 3 times.

Figure 3

HSPCs are transiently driven into cell cycle by Notch signaling. (A) Lethally irradiated mice were reconstituted with 5-FU–treated donor Rag1^−/− BM cells transduced with MigR1 control or activated Notch1. At 2 weeks and 6 weeks after transplantation, the mice were injected with BrdU and then killed 16 hours later. (A) The donor-derived BM HSPC compartment (GFP⁺Lineage⁻Sca-1⁺c-Kit⁺) was stained with 7-AAD and intracellular antibody against BrdU. Cells entering cell cycle are BrDU⁺ with 2N DNA content by 7-AAD stain. Apoptotic cells are BrdU⁻ and found in the sub2N region of the 7-AAD stain. (B) Bar graph analysis of BrdU⁺ MigR1 control and Notch HSPCs entering cell cycle at 2 and 6 weeks after transplantation. Three mice per condition.

Figure 4

Noncompetitive transplantation of Notch-activated HSPCs leads to hematopoietic failure. (A) Lethally irradiated CD45.1⁺ congenic mice were reconstituted with 5-FU–treated donor CD45.2⁺Rag1^−/− BM cells transduced with MigR1 control or activated Notch1. At 6 weeks after transplantation, CD45.2⁺GFP⁺Lineage⁻Sca-1⁺c-Kit⁺ cells (HSPCs) were sorted and transplanted into CD45.1⁺ hosts at 500 cells per mouse. (B) Kaplan-Meier graph showing the proportion of mice alive after transplantation. (C) Scatter plot analysis of moribund mice at 11 days after transplantation showing peripheral blood hemoglobin concentration; absolute numbers of total donor-derived (CD45.2⁺GFP⁺) splenocytes, splenic ter119⁺ erythroid cells, splenic CD11b⁺ myeloid cells, Thy-1⁺ T cells, splenic Lineage⁻Sca-1⁺c-Kit⁺ HSPCs, and splenic Lineage⁻Sca-1⁻c-Kit⁺CD34⁺CD16/62^hi granulocyte-macrophage progenitors (GMPs); and day 11 CFU-S colonies. The experiment was performed twice; 4 mice for MigR1 and 9 mice for activated Notch1.

Figure 5

Competitive transplantation of Notch-activated HSPCs leads to progressive exhaustion of HSC activity and numbers. A similar experiment was performed as in Figure 4A but with the addition of 200 000 whole CD45.1⁺ BM cells to prevent mortality from hematopoietic failure. (A) BM DN T-cell production (Thy-1⁺CD25⁺) in MigR1 control and Notch mice was measured in the BM at 2 and 16 weeks after transplantation. (B) Scatter plot analysis of DN T-cell production by MigR1 controls and Notch mice at 2 and 16 weeks after transplantation. (C) The LT-HSC compartment (Lineage⁻Sca-1⁺c-Kit⁺CD34⁻Flt3⁻) was measured for donor reconstitution by GFP⁺ MigR1 control and Notch-activated LT-HSCs at 2 and 16 weeks after transplantation. (D) Scatter plot analysis of donor-derived (GFP⁺) LT-HSCs (Lineage⁻Sca-1⁺c-Kit⁺CD34⁻Flt3⁻) at 2 and 16 weeks after transplantation. Experiment was performed 3 times. MigR1, 6 mice at 2 weeks; 5 mice at 16 weeks. Notch, 3 mice at 2 weeks; 6 mice at 16 weeks.

Figure 6

Notch progressively depletes LT-HSC activity and numbers over time in nontransplanted Rosa26-Flox-STOP-Flox-Notch-GFP mice. Mx-cre^tgRosa26-Flox-STOP-Flox-YFP control mice (“YFP control”) and Mx-cre^tgRosa26-Flox-STOP-Flox-ICN-GFP (“Notch”) mice were injected with pI-pC to activate YFP control and ICN-GFP, respectively in HSCs. (A) BM DP T-cell production (CD4⁺CD8⁺) by the HSC compartment was measured at 4 and 16 weeks after injection in YFP control and Notch mice. (B) Scatter plot analysis of BM DP T-cell production by YFP control and Notch mice at 4 and 16 weeks after injection. (C) The LT-HSC compartment (Lineage⁻Sca-1⁺c-Kit⁺CD150⁺CD48⁻) was measured for the donor-derived cells (YFP⁺ or GFP⁺) in YFP control and Notch mice at 4 and 16 weeks after injection. (D) Scatter plot analysis of donor-derived (YFP⁺ or GFP⁺) LT-HSCs (Lineage⁻Sca-1⁺c-Kit⁺CD150⁺CD48⁻) in YFP control and Notch mice at 4 and 16 weeks after injection. One experiment was performed; 3 mice per condition at 4 weeks and 5 mice per condition at 6 weeks.

Figure 7

Competitive transplantation of Notch-activated LT-HSCs from Rosa26-Flox-STOP-Flox-ICN-GFP mice leads to cell autonomous exhaustion of HSC activity and numbers. (A) Mx-cre^tgRosa26-Flox-STOP-Flox-YFP (“YFP control”) mice and Mx-cre^tgRosa26-Flox-STOP-Flox-ICN-GFP (“Notch”) mice were injected with pI-pC to activate YFP and ICN-GFP, respectively in HSCs. Then, 4 weeks later YFP⁺ or GFP⁺ Lineage⁻Sca-1⁺c-Kit⁺CD150⁺CD48⁻ BM cells (HSCs) were sorted and transplanted into syngeneic hosts at 500 cells per mouse along with 200 000 whole syngeneic BM cells. (B) BM DP T-cell production (CD4⁺CD8⁺) in YFP control and Notch mice was measured in the bone marrow at 4 and 16 weeks after transplantation. (C) Scatter plot analysis of DP T-cell production by YFP control and Notch mice at 4 and 16 weeks after transplantation. (D) The LT-HSC compartment (Lineage⁻Sca-1⁺c-Kit⁺CD150⁺CD48⁻) was measured for donor reconstitution by YFP⁺ control and Notch HSCs at 4 and 16 weeks after transplantation. Scatter plot analysis of donor-derived (YFP⁺ or Notch) LT-HSCs (Lineage⁻Sca-1⁺c-Kit⁺CD150⁺CD48⁻) at 4 and 16 weeks after transplantation. (E) Cell-cycle analysis was performed in sorted YFP control and Notch LT-HSCs (YFP⁺ or GFP⁺ Lineage⁻Sca-1⁺c-Kit⁺CD150⁺CD48⁻ BM cells). Propidium iodide⁺ cells were excluded. Pyronin⁺/Hoechst 33 342⁺ cells (top left quadrant) are in G₁. Pyronin⁻/Hoescht⁻ cells (bottom left quadrant) are in G₀. Pyronin⁺/Hoechst 33 342⁺ cells (top right quadrant) are in S/G₂/M.