p53 loss promotes acute myeloid leukemia by enabling aberrant self-renewal

Abstract

The p53 tumor suppressor limits proliferation in response to cellular stress through several mechanisms. Here, we test whether the recently described ability of p53 to limit stem cell self-renewal suppresses tumorigenesis in acute myeloid leukemia (AML), an aggressive cancer in which p53 mutations are associated with drug resistance and adverse outcome. Our approach combined mosaic mouse models, Cre-lox technology, and in vivo RNAi to disable p53 and simultaneously activate endogenous Kras(G12D)-a common AML lesion that promotes proliferation but not self-renewal. We show that p53 inactivation strongly cooperates with oncogenic Kras(G12D) to induce aggressive AML, while both lesions on their own induce T-cell malignancies with long latency. This synergy is based on a pivotal role of p53 in limiting aberrant self-renewal of myeloid progenitor cells, such that loss of p53 counters the deleterious effects of oncogenic Kras on these cells and enables them to self-renew indefinitely. Consequently, myeloid progenitor cells expressing oncogenic Kras and lacking p53 become leukemia-initiating cells, resembling cancer stem cells capable of maintaining AML in vivo. Our results establish an efficient new strategy for interrogating oncogene cooperation, and provide strong evidence that the ability of p53 to limit aberrant self-renewal contributes to its tumor suppressor activity.

Figures

Figure 1.

Activation of endogenous KrasG12D and shRNA-mediated suppression of p53 cooperate to induce AML. (A) Schematic of the experimental strategy. Retroviral LGmCreER is used to cotransduce GFP, a miR30-based shRNA, and 4-OHT-inducible self-excising CreERT2 into HSPCs isolated from embryonic day 13.5–15.5 fetal livers of LSL-KrasG12D embryos. Cre-mediated recombination is induced by treatment with 4-OHT in vitro or in vivo. (B) Representative spleen sonogram and three-dimensional reconstruction volumetry in recipient mice of HSPCs of indicated genotypes 6 wk after transplantation. Numbers indicate average spleen volume and standard deviation in six mice for each genotype. Only KrasG12D-shp53 recipient mice show severe splenomegaly at this preleukemic stage. (C) Whole-body imaging of representative mice reconstituted with KrasG12D, shp53, and KrasG12D-shp53 HSPCs. Only KrasG12D-shp53 mice show strong accumulation of GFP-positive cells in bones (b), spleen (s), and liver (l). (D) Kaplan-Meier curve showing the survival of mice reconstituted with KrasG12D-shp53 HSPCs (n = 57) and controls. (**) P = 0.0022 and (***) P < 0.0001 (log-rank test) indicate highly significant survival reduction compared with recipients of shp53 (n = 12) and KrasG12D (n = 20) control HSPCs, respectively. Secondary transplantation (2nd tx) curve indicating survival of sublethally irradiated recipient mice transplanted with KrasG12D-shp53 leukemias (n = 16). (E) Representative immunophenotyping of KrasG12D-shp53 leukemic bone marrow showing strong infiltration of GFP-positive cells that are positive for Mac1 and Gr1. (F) Representative histopathology of KrasG12D-shp53-induced AMLs, including MPD-like leukemias (panels A–D) and acute myelomonocytic leukemia (panel E) and AML without maturation (panel F). (Panel A) Peripheral blood containing elevated numbers of neutrophilic and monocytic cells. Arrowhead denotes Howell-Jolly body. (Panel B) Sternal bone marrow filled with maturing myeloid cells. (Panel C) Spleen with vastly expanded red pulp filled with erythroid cells (top left) and maturing myeloid elements (bottom right). (Panel D) Liver with extensive perivascular and sinusoidal hematopoiesis. (Panel E) Bone marrow cytospin showing blasts and monocytic cells. (Panel F) Bone marrow cytospin showing immature blast cells. Blood smear and bone marrow cytospin: Wright-Giemsa; bar, 20 μm. Liver and spleen: hematoxylin and eosin; bar, 50 μm.

Figure 2.

Molecular characterization of KrasG12D-shp53-induced AML. (A) PCR analysis to verify Cre-mediated recombination of the LSL cassette. (Top panel) Two-step PCR in peripheral blood of a representative KrasG12D-shp53 HSPC recipient mouse before and 5 d after 4-OHT treatment in vivo and after leukemia onset in bone marrow (bm) and spleen (spl). Recombination is indicated by the appearance of a 749-bp band after the first (I), and a 249-bp band after the second (II) PCR. (Bottom panel) LSL-KrasG12D recombination in leukemic mice (1–5) analyzed in bone marrow and spleen (left and right lane in each sample, respectively). (nc) Negative control. (B) Western blotting of p53 in wild-type bone marrow, Arf−/−, and p53−/− MLL/ENL-induced control leukemias and three independent KrasG12D-shp53 AMLs (#1, #2, and #3) without (−) or with (+) adriamycin (Adr) treatment in vitro. (C) Flow cytometry analysis of intracellular phosphorylated Mapk1/Mapk3 (pErk) and Rps6 (pS6) in bulk and core (Lin−cKit+) populations of wild-type bone marrow and KrasG12D-shp53 leukemia. (D) Elevated basal levels of intracellular phosphorylated Mapk1/Mapk3 (pErk) and Rps6 (pS6) in multiple independent KrasG12D-shp53 AMLs.

Figure 3.

p53 suppression and KrasG12D induce self-renewal of myeloid progenitor cells in vitro. (A) BrdU incorporation in transduced mature myeloid cells (MM; Gr1+), myeloid progenitors (MP; Lin−cKit+Sca1−IL7Rα−), and MPP and stem cells (KSL; Lin−cKit+Sca1+) in indicated genotypes 4 d following 4-OHT-induced activation of KrasG12D. (B) Frequency of GFP+ HSCs (Lin−cKit+Sca1+CD150+CD48−) and myeloid progenitors (MP; Lin−cKit+Sca1−IL7Rα−) in all transduced (GFP+) HSPCs of indicated genotypes (color coding as in A) 4 d after activation of KrasG12D. (C) Methylcellulose colony formation and serial replating assay of transduced 4-OHT-treated HSPCs of indicated genotypes. Representative bright-field and GFP fluorescent images showing colonies formed after the fourth round of replating. (D) Quantification of GFP+ colonies of indicated genotypes over six rounds of replating (color coding as in A). (E) Percentage of GFP+ cells of indicated genotypes at the end of each round of replating (color coding as in A).

Figure 4.

p53 suppression expands myeloid progenitors in vivo and, in concert with endogenous, KrasG12D promotes leukemogenesis. (A) Representative bone marrow flow cytometry profiles of stem and myeloid progenitor compartments of mice reconstituted with HSPCs of indicated genotypes, including conventional Lin/cKit/Sca1 staining (MP; Lin−cKit+Sca1−IL7Rα−), KSL (Lin−cKit+Sca1+), SLAM markers (HSC; Lin−cKit+Sca1+CD150+CD48−), and histograms indicating the percentage of GFP+ cells in myeloid progenitors. (B) Frequency of GFP+ cells in total bone marrow (all), myeloid progenitors (MP), and HSCs of mice reconstituted with HSPCs of indicated genotypes 3 wk after in vivo 4-OHT treatment to induce KrasG12D (n = 4–5 in each group). (C) Quantification of GFP+ myeloid progenitors within GFP+ bone marrow cells of mice reconstituted with HSPCs of indicated genotypes (n = 4–5 in each group). (D) Quantification of GFP+ HSCs within GFP+ bone marrow cells of mice reconstituted with HSPCs of indicated genotypes (n = 4–5 in each group).

Figure 5.

KrasG12D-shp53 myeloid progenitor cells induce AML. (A) Representative bone marrow flow cytometry profiles of stem and progenitor compartments in KrasG12D-shp53 leukemic mice, including conventional Lin/cKit/Sca1 staining (left two panels), SLAM markers (middle panel), and a histogram indicating the percentage of GFP+ in HSCs and myeloid progenitors (right two panels). (B) Schematic overview of generation, sorting, and secondary transplantation of KrasG12D-shp53 myeloid progenitor cells. (C) Kaplan-Meier curve of lethally irradiated recipient mice transplanted with 2000–5000 doubly sorted myeloid progenitor cells isolated from six independent primary recipients of KrasG12D-shp53 HSPCs (n = 18) and KrasG12D HSPCs (n = 3). (D) Representative bone marrow flow cytometry analysis of leukemia induced by transplantation of KrasG12D-shp53 myeloid progenitors. (Left panel) The bone marrow is dominated by GFP+ cells, which are predominantly expressing Mac1 and Gr1. (E) Histopathology analysis of leukemias induced by transplantation of KrasG12D-shp53 myeloid progenitors. (Panel A) Sternal bone marrow densely packed with blasts and few maturing myeloid cells (100×). (Panel B) Spleen with areas predominated by blasts and maturing neutrophils (100×). (Panel C) liver with invasion of leukemic cells with differentiation (40×). All hematoxylin and eosin-stained. Bars: 40× images, 50 μm; 100× images, 20 μm.

Figure 6.

Conditional p53 deletion recapitulates the effects of RNAi-mediated p53 suppression. (A) Methylcellulose cultures (bright-field images) of bone marrow cells of indicated genotypes after the third round of replating. Robust CFU-M-, CFU-G-, and CFU-GM-like colonies are only formed by KrasG12D;p53Δ2-10 bone marrow cells. (B) Quantification of colony formation in bone marrow cells of indicated genotypes over four rounds of replating. (C) Frequency of mature myeloid cells (MM; Mac1+Gr1+), myeloid progenitors (MP; Lin−cKit+Sca1−IL7Rα−), and HSCs (Lin−cKit+Sca1+CD150+CD48−) in bone marrow of mice with indicated genotypes (color coding as in B) 7–10 d after pIpC treatment (n = 3–5 for each genotype). (D) Kaplan-Meier curve of lethally irradiated recipient mice transplanted with 2000–5000 myeloid progenitor cells doubly sorted from primary Mx1-Cre;LSL-KrasG12D;p53loxp/loxp (n = 9) and Mx1-Cre;LSL-KrasG12D (n = 3) mice. (E) Histopathology analysis of donor and recipient mice of Mx1-Cre;LSL-KrasG12D;p53loxp/loxp myeloid progenitors. (Panel A) Bone marrow (100×) of a donor mouse at the time of myeloid progenitor isolation (10 d after pIpC induction of Cre) showing normal mixed hematopoeisis. (Panels B–D) Bone marrow (100×), spleen (40×), and liver sections (40×), respectively, of a myeloid progenitor recipient mouse showing leukemia infiltrates. All hematoxylin and eosin-stained. Bars: 40× images, 50 μm; 100× images, 20 μm.