Constitutively active AKT depletes hematopoietic stem cells and induces leukemia in mice

Abstract

Human cancers, including acute myeloid leukemia (AML), commonly display constitutive phosphoinositide 3-kinase (PI3K) AKT signaling. However, the exact role of AKT activation in leukemia and its effects on hematopoietic stem cells (HSCs) are poorly understood. Several members of the PI3K pathway, phosphatase and tensin homolog (Pten), the forkhead box, subgroup O (FOXO) transcription factors, and TSC1, have demonstrated functions in normal and leukemic stem cells but are rarely mutated in leukemia. We developed an activated allele of AKT1 that models increased signaling in normal and leukemic stem cells. In our murine bone marrow transplantation model using a myristoylated AKT1 (myr-AKT), recipients develop myeloproliferative disease, T-cell lymphoma, or AML. Analysis of the HSCs in myr-AKT mice reveals transient expansion and increased cycling, associated with impaired engraftment. myr-AKT-expressing bone marrow cells are unable to form cobblestones in long-term cocultures. Rapamycin, an inhibitor of the mammalian target of rapamycin (mTOR) rescues cobblestone formation in myr-AKT-expressing bone marrow cells and increases the survival of myr-AKT mice. This study demonstrates that enhanced AKT activation is an important mechanism of transformation in AML and that HSCs are highly sensitive to excess AKT/mTOR signaling.

Figure 1

Myeloid expansion in myr-AKT mice. (A) Kaplan-Meier survival curve of 3 separate BMT experiments using myr-AKT retrovirus (n = 30) or MIG empty vector control (n = 5; P < .001). (B) Hematoxylin and eosin–stained histopathology sections of a representative liver and spleen of myr-AKT recipients, revealing infiltration with immature myeloid and erythroid cells (arrows). Details on image acquisition can be found in supplemental Methods. To the left is a photograph of spleens from myr-AKT and MIG control mice. (C) Flow cytometric analysis of the myeloid, lymphoid, and erythroid lineages of BM and spleen (SP) single-cell suspensions from myr-AKT or MIG transplantation recipients. All were gated for GFP⁺ cells. Values are the mean ± SEM. *P < .05.

Figure 2

T-cell lymphoma in myr-AKT mice. (A) Photograph of the thymus from a myr-AKT mouse and a MIG control mouse. (B) Flow cytometric analysis of thymus, BM, and spleen single-cell suspensions from a myr-AKT mouse with T-cell lymphoma. Representative plots are shown. All were gated for GFP⁺ cells. (C) Hematoxylin and eosin–stained histopathology sections of thymus, lung, heart, BM, and muscle sections of a myr-AKT mouse with T-cell lymphoma. Details on image acquisition can be found in supplemental Methods.

Figure 3

AML in myr-AKT mice. (A) Hematoxylin and eosin–stained histopathology sections of spleen, liver, and BM from a representative myr-AKT mouse with AML. Details on image acquisition can be found in supplemental Methods. (B) Flow cytometric analysis of BM and spleen cells from a representative myr-AKT mouse with AML. All were gated for GFP⁺ cells. (C) AML and T-cell lymphoma are transplantable, whereas MPD is not. Kaplan-Meier survival curves representing secondary transplantations of splenocytes from myr-AKT mice with MPD, thymocytes from myr-AKT mice with T-cell lymphoma, or splenocytes from myr-AKT mice with AML injected into the tail veins of sublethally irradiated C57 Bl/6 mice. MPD secondary transplantation mice were followed for 120 days with no evidence of disease. (D) Western blot of representative splenocyte and thymocyte lysates from diseased myr-AKT mice and a WT age-matched control mouse. (E) Intracellular flow cytometry on GFP-gated thymocytes and GFP-gated CD71^hic-Kit^hi splenocytes from a myr-AKT mouse and a WT age-matched control mouse.

Figure 4

Sustained AKT signaling in myr-AKT mice causes depletion of LSK cells and progenitors and increased apoptosis. (A) Left bar graph: Percentage GFP of BM transduced with myr-AKT-GFP or MIG control retrovirus on the day of tail vein injection. Results are the mean of data from 3 independent transplantation experiments. Right bar graphs: Percentage GFP in recipient BM and spleen (Spln) in MIG or myr-AKT mice killed at 6 to 8 weeks after transplantation. Error bars represent SEM. (B) Decrease in percentage GFP⁺ cells in the LSK and progenitor compartments of myr-AKT BM. BM of control MIG transplantation mice or diseased myr-AKT transplantation mice killed at 6 to 8 weeks after transplantation was stained with a lineage antibody cocktail, goat anti–rat antibody, and then Sca1 and c-Kit to distinguish the LSK and progenitor populations. The percentage of GFP⁺ cells in each population is shown. (C) Competitive engraftment of BM transduced with myr-AKT or MIG control retrovirus. BM was transduced with myr-AKT or MIG retrovirus and then injected into mice as previously described. Recipient mice were killed at 2 weeks after transplantation, and percentage GFP in the BM was quantified by flow cytometry. The fold change in percentage GFP in the BM was determined as: percentage GFP in recipient BM/percentage GFP in donor BM. The values for MIG controls were normalized to 1 for the analysis. (D) Apoptosis analysis of the LSK and progenitor compartments in BM from MIG or diseased myr-AKT mice killed at 6 to 8 weeks after transplantation. Apoptosis analysis of the LSK and mixed progenitor populations from myr-AKT mice. BM cells were gated on GFP⁺ LSK and Lin⁻c-kit⁺Sca1⁻ (Prog) cells. Representative annexin V vs propidium iodide plots are shown for these populations. (E) Methylcellulose plating assays of BM cells from myr-AKT and MIG control mice, killed at 6 to 8 weeks after transplantation. Bar graphs reveal the total number of colonies seen at each round of replating every 7 days. (F) Methylcellulose plating assays of spleen cells from myr-AKT and MIG control mice, killed at 6 to 8 weeks after transplantation. Bar graphs represent the total number of colonies seen at each round of replating every 7 days.

Figure 5

Short-term induction of AKT signaling causes expansion of the LSK compartment and increased LSK cycling. BMT assay with myr-AKT-ER or vector control retrovirus, followed by tamoxifen induction after engraftment for 3 days. (A) Multiparameter flow cytometric analysis of the Thy1.1-gated LSK and progenitor BM populations. The experiment was repeated 3 times, with 3 or 4 mice per group in each experiment. (B) Analysis of the Thy1.1-gated CD34⁻LSK BM subpopulation of myr-AKT-ER mice and vector control mice. The Student paired 2-tailed t test was used to compare the percentage Thy1.1⁺ CD34⁻ LSK cells between samples. (C) Cell-cycle analysis of Thy1.1-gated BM LSKs from myr-AKT-ER mice or vector control mice. Hoechst and pyronin Y were used to resolve the G₀, G₁, and S/G₂/M stages of the cell cycle. The experiment was repeated 3 times, with 2 to 4 mice per group in each experiment.

Figure 6

HSC depletion of myr-AKT–transduced BM is dependent on mTOR signaling but not on increased ROS. (A) Rapamycin but not N-acetylcysteine (NAC) rescues cobblestone-forming activity of myr-AKT–transduced BM in the LTC-IC assay. A total of 500 000 retrovirally transduced GFP-sorted BM cells/well were cocultured for 4 weeks with 300 000 OP9 stromal cells. After 4 weeks in coculture, cells were trypsinized and plated into M3434 methylcellulose media, and colonies were scored after 7 days. Rapamycin (Rap) or NAC was added at the time of initial plating, and treatment was continued for 4 weeks. Each drug experiment was performed with at least 5 replicates. (B) ROS levels in the LSK compartment after short-term induction of AKT signaling in vivo. Induction of myr-AKT expression using the myr-AKT-ER BMT system was performed as previously described. Thy1.1⁺ LSK and progenitor cells were sorted from the BM, and freshly sorted cells were stained with DCF-DA to determine relative levels of ROS. The experiment was performed twice, with 3 or 4 mice per group in each experiment. Right bar graphs: Fold change in DCF-DA peak intensity of sorted LSK and progenitor cells. The values for vector samples were normalized to 1 for the analysis.

Figure 7

Rapamycin treatment reduces the incidence of T-cell lymphoma and leads to increased survival in myr-AKT mice. (A) Kaplan-Meier survival curve of myr-AKT mice given daily intraperitoneal injections of vehicle (n = 6) or 4 mg/kg rapamycin (n = 7), starting at 4 weeks after transplantation. The curve on the left represents overall survival, whereas the curve on the right shows the incidence of T-cell lymphoma, here defined as an enlarged thymus weighing more than 200 mg. (B) Organ weights of vehicle- and rapamycin-treated myr-AKT mice at the time of death. (C) Flow cytometric analysis of the myeloid lineages of BM and spleen (SP) single-cell suspensions from representative vehicle- and rapamycin-treated myr-AKT mice. All were gated for GFP⁺ cells. Values are mean ± SEM. *P < .05. (D) Percentage of GFP⁺ cells in the BM and spleen of vehicle- and rapamycin-treated myr-AKT mice.