Temporal changes in PTEN and mTORC2 regulation of hematopoietic stem cell self-renewal and leukemia suppression

Abstract

Pten deletion from adult mouse hematopoietic cells activates the PI3-kinase pathway, inducing hematopoietic stem cell (HSC) proliferation, HSC depletion, and leukemogenesis. Pten is also mutated in human leukemias, but rarely in early childhood leukemias. We hypothesized that this reflects developmental changes in PI3-kinase pathway regulation. Here we show that Rictor deletion prevents leukemogenesis and HSC depletion after Pten deletion in adult mice, implicating mTORC2 activation in these processes. However, Rictor deletion had little effect on the function of normal HSCs. Moreover, Pten deletion from neonatal HSCs did not activate the PI3-kinase pathway or promote HSC proliferation, HSC depletion, or leukemogenesis. Pten is therefore required in adult, but not neonatal, HSCs to negatively regulate mTORC2 signaling. This demonstrates that some critical tumor suppressor mechanisms in adult cells are not required by neonatal cells. Developmental changes in key signaling pathways therefore confer temporal changes upon stem cell self-renewal and tumor suppressor mechanisms.

Figure 1. HSCs that proliferate under physiological conditions exhibit little or no increase in AKT phosphorylation

(A) Representative flow cytometry plots for HSC cell cycle analysis. (B) The percentage of CD150⁺CD48⁻LSK HSCs in S/G2/M phases of the cell cycle from E14.5 fetal (n=3), P14 neonatal (n=3), 8 week old adult (n=6), 8 week old pIpC treated Mx1-Cre; Pten^fl/fl mice (n=3) and cyclophosphamide/G-CSF treated adult mice (n=3). All comparisons are relative to 8 week-old adult mice (*, p< 0.05; **, p< 0.01; ***, p< 0.001). Error bars reflect standard deviation and p-values were calculated by two-tail students t-test. (C) CD48⁻LSK HSCs/MPPs have a similar cell cycle distribution as HSCs. (D) PI3-kinase pathway components in HSCs/MPPs (30,000 CD48⁻LSK cells per lane). (E) AKT phosphorylation in fetal HSCs in G0/G1 versus S/G2/M phases of the cell cycle (10,000 cells per lane). (F) Adult GMP myeloid restricted progenitors, but not HSCs/MPPs, exhibited increased AKT and S6 phosphorylation after cyclophosphamide/2 days G-CSF treatment (20,000 cells per lane). (G) Even after cyclophosphamide and 4 days of G-CSF, mobilized spleen HSCs/MPPs (20,000 cells per lane) did not exhibit increased AKT phosphorylation relative to normal adult HSCs/MPPs.

Figure 2. Pten is not required to regulate HSC maintenance, proliferation, or function in neonatal mice

(A, B) BrdU incorporation in bone marrow cells and HSCs (24 hour pulse; 3 to 4 independent experiments; *, p<0.05). (C) The numbers of HSCs in the spleens of mice that received pIpC 2 days after birth then were analyzed at 2 or 4 weeks after birth, or that received pIpC at 6 weeks after birth and were analyzed 8 weeks after birth (n=4-10 mice per genotype, *, p<0.05; ***, p<0.0001). (D) Splenocyte BrdU incorporation (24 hour BrdU pulse; n=3-4 mice per time point and genotype; **, p<0.01). (E-H) Long-term competitive repopulation assays with 10 wild-type or Pten deficient (Pten Δ/Δ) adult or neonatal HSCs (n=7-10 recipients/age/genotype; *, p<0.05; **, p<0.01; ***, p<0.001). The threshold at which donor cells could be detected above background (in negative control mice, 0.5%) is indicated by the black line. (I, J) Spleen cellularity (I) and mass (J) for Pten deleted adult and neonatal Mx1-Cre; Pten^fl/fl mice as well as littermate controls two weeks after pIpC treatment (n=10-13 mice/age/genotype; *, p<0.05; ***, p<1 × 10⁻¹⁰). (K-N) Representative low power (40x) spleen sections stained with hematoxylin and eosin from wild-type adult (K), Pten deleted adult (L), wild-type neonatal (M) and Pten deleted neonatal (N) mice. For all panels, error bars reflect standard deviation and p-values were calculated by two-tailed students t-test.

Figure 3. Loss of Pten and p53 leads to the development of T-ALL in adult but not in neonatal mice

(A-D) Kaplan-Meier survival curves for Mx1-Cre; Pten^fl/fl; p53^−/− (A and C; n=6-8 mice/treatment) and Mx1-Cre; Pten^fl/fl; p53^+/− mice (B and D; n=7-9 mice/treatment) following pIpC treatment at 2 days (neonatal) or 6 weeks (adult) after birth. Survival times are shown either as the time following pIpC treatment (A, B) or as time after birth (C, D). Pten mutant mice that were treated neonatally with pIpC survived significantly longer after pIpC treatment than mice treated with pIpC as adults (***, p<0.001 by the log rank test). (E-H) Spleen and thymus sections from adult mice that were treated with pIpC at 6 weeks after birth then sacrificed for analysis 17 days later (40x, 400x inset). (I-L) Spleen and thymus sections from neonatal mice treated with pIpC 2 days after birth then sacrificed for analysis 17 days later (40x). (M-P) Spleen and thymus sections from mice that received pIpC 2 days after birth and developed T-ALL as adults (40x, inset 400x).

Figure 4. Neonatal HSCs do not require PTEN to negatively regulate AKT phosphorylation by mTORC2 in vivo

(A) Western blots were performed on 30,000 wild-type or Pten deficient CD48⁻LSK cells isolated from adult and neonatal mice. After Pten deletion AKT and S6 phosphorylation increased in adult HSCs/MPPs but little increase was observed in neonatal HSCs/MPPs. (B) AKT and GSKβ phosphorylation increased in adult but not in neonatal HSCs/MPPs after Pten deletion. (C) AKT phosphorylation in 20,000 HSCs/MPPs from 3 and 4-week old control and Pten deficient mice. (D) HSCs/MPPs from adult and neonatal mice were incubated in Iscove’s medium with 2% fetal bovine serum for 30 minutes. (E) The expression levels of several PI3-kinase pathway components were similar in adult and neonatal HSCs/MPPs (30,000 cells per lane). (F) Phosphorylation of mTOR at Ser2481 was not different between adult and neonatal HSCs/MPPs. (G) AKT phosphorylation in 20,000 wild-type or Pten deficient CD48⁻LSK cells (HSCs/MPPs), CD48⁺LSK cells (restricted progenitors), or bone marrow cells from adult or neonatal mice. (H, I) AKT and S6 phosphorylation in neonatal and adult, control and Pten deficient LMPPs, GMPs (H) and CD4+CD8+ thymocytes (I). (J, K) In both adult and neonatal mice, Pten deletion significantly reduced B-cell frequencies and significantly increased myeloid cell frequencies (n=6 to 7 mice/age/genotype; **, p< 0.01; ***, p<0.001). Error bars always represent standard deviation. (L) pIpC treatment of Mx1-Cre; Rictor^fl/fl mice eliminated RICTOR expression by LSK cells. (M, N) Rictor deletion reduced AKT Ser473 phosphorylation in CD48⁻LSK cells (M), and had similar effects as Torin1 treatment with respect to AKT phosphorylation at Ser473 but not with respect to S6 phosphorylation (N).

Figure 5. Rictor deletion had only modest effects on HSC frequency and function

(A) Vav-Cre mediated Rictor deletion in fetal HSCs slightly but significantly increased HSC numbers in P14 mice (n=5-7 mice per genotype, *p<0.05). Mx1-Cre mediated Rictor deletion at P2 did not significantly affect HSC numbers at P14. (B, C) Vav-Cre or neonatal Mx1-Cre mediated Rictor deletion did not affect cell cycle distribution (B) or BrdU incorporation (C) in HSCs at P14. (D-F) Deletion of Rictor in 6 week old adult Mx1-Cre;Rictor^fl/fl mice did not significantly affect bone marrow cellularity (D), HSC frequency (E) or absolute HSC number (F) when analyzed 18-24 weeks after pIpC treatment. However, MPP frequency (E) and MPP numbers (F) were reduced in Rictor deleted mice relative to control mice (n=5-7 mice per genotype, ***, p<0.001). (G-K) When 300,000 control or Rictor-deficient CD45.2 bone marrow cells were transplanted along with 300,000 CD45.1 wild-type bone marrow cells into irradiated mice, recipients of Rictor-deficient bone marrow were always long-term multilineage reconstituted by donor cells but levels of donor cell reconstitution were sometimes significantly lower than in recipients of control cells (n=11-14 recipients per genotype, *, p<0.05; **, p<0.01; ***, p<0.001). (L) The frequencies of donor HSCs and MPPs were similar in primary recipients of control and Rictor deficient bone marrow cells (n=11-14 per genotype). (M, N) Secondary recipients of Rictor deficient and control bone marrow cells exhibited similar levels of long-term multilineage reconstitution (n=15 recipients per genotype). Error bars reflect standard deviation and p-values were calculated by two-tailed student’s t-tests.

Figure 6. Rictor is necessary for increased HSC proliferation, HSC mobilization and HSC depletion following Pten deletion

(A, B) Rapamycin did not inhibit AKT phosphorylation after 7 days in vivo (A) or 30 minutes in vitro (B). (C) The percentage of HSCs in S/G2/M phases of the cell cycle in control, Mx1-Cre; Rictor^fl/fl (RictorΔ/Δ), Mx1-Cre; Pten^fl/fl (PtenΔ/Δ) and Mx1-Cre; Pten^fl/fl; Rictor^fl/fl (PtenΔ/Δ; RictorΔ/Δ) mice, with or without 7 days of rapamycin treatment, 2 weeks after pIpC treatment (n=3-10 mice per genotype and treatment). (D) Rictor deletion, but not rapamycin treatment, significantly reduced HSC mobilization to the spleen following Pten deletion (n=8-9 mice/genotype). (E, F) Rapamycin treatment, but not Rictor deletion, increased bone marrow HSC numbers (E) and reduced HSC mobilization to the spleen following cyclophosphamide/G-CSF treatment (F; n=5 mice/treatment). (G-K) 10 CD45.2+ donor HSCs from Pten–deficient mice failed to give long-term multilineage reconstitution in irradiated mice. Rictor–deficient HSCs gave long-term multilineage reconstitution in 3 of 12 recipient mice, but levels of donor cell reconstitution were generally lower than observed from control HSCs. Pten; Rictor compound deficient HSCs were able to give long-term multilineage reconstitution in nearly all irradiated mice at levels similar to control HSCs, with the exception of the B-cell lineage (I). (n=12-16 mice/genotype). (L-P) Following secondary transplantation, Rictor–deficient and Pten; Rictor compound deficient donor cells gave long-term multilineage reconstitution of nearly all secondary recipient mice (N, O) (n=9-22 mice/genotype). Error bars represent standard deviation and p-values were calculated by two-tailed student’s t-tests: *, p<0.05; **, p<0.01; ***, p<0.001; N.S., not significant (p>0.05).

Figure 7. Rictor deficiency reduced the severity of myeloproliferative disorder and the incidence of leukemia after Pten deletion

(A) Two weeks after pIpC treatment in adult mice, spleen cellularity increased significantly in Mx1-Cre; Pten^fl/fl (PtenΔ/Δ mice due to myeloproliferative disorder and this increase was significantly attenuated by Rictor deletion and rapamycin treatment (n=5-10 mice/genotype, *, p<0.05; **, p<0.01; ***, p<0.001; N.S., not significant (p>0.05)). (B) Kaplan-Meier survival curves after pIpC treatment showed that Rictor deletion significantly increased the survival of Pten deleted mice (p<0.001 relative to PtenΔ/Δ; Rictor+/+ mice by log-rank test). (C-E) Spleen and thymus sizes of mice from the same experiment 16 weeks after pIpC treatment (E; n=5-7 mice/genotype, ***, p<0.001). (F-M) A myeloproliferative disorder was evident in the spleens of PtenΔ/Δ; RictorΔ/Δ mice (I) but not control or PtenΔ/+; RictorΔ/Δ mice (F, H). T-ALL was observed in the spleens of PtenΔ/Δ; Rictor+/+ mice (G, arrows) but not in the spleens of other genotypes. Thymuses of control (J), Pten+/+; RictorΔ/Δ (L), and PtenΔ/Δ; RictorΔ/Δ (M) mice did not have T-ALL, in contrast to PtenΔ/Δ; Rictor+/+ thymuses (K).