MDM2 antagonists induce p53-dependent apoptosis in AML: implications for leukemia therapy

Abstract

Although TP53 mutations are rare in acute myeloid leukemia (AML), inactivation of wild-type p53 protein frequently occurs through overexpression of its negative regulator MDM2 (murine double minute 2). Recently, small-molecule antagonists of MDM2, Nutlins, have been developed that inhibit the p53-MDM2 interaction and activate p53 signaling. Here, we study the effects of p53 activation by Nutlin-3 in AML cells. Treatment with MDM2 inhibitor triggered several molecular events consistent with induction of apoptosis: loss of mitochondrial membrane potential, caspase activation, phosphatidylserine externalization, and DNA fragmentation. There was a positive correlation in primary AML samples with wild-type p53 between baseline MDM2 protein levels and apoptosis induced by MDM2 inhibition. No induction of apoptosis was observed in AML samples harboring mutant p53. Colony formation of AML progenitors was inhibited in a dose-dependent fashion, whereas normal CD34+ progenitor cells were less affected. Mechanistic studies suggested that Nutlin-induced apoptosis was mediated by both transcriptional activation of proapoptotic Bcl-2 family proteins, and transcription-independent mitochondrial permeabilization resulting from mitochondrial p53 translocation. MDM2 inhibition synergistically enhanced cytotoxicity of cytosine arabinoside and doxorubicin in AML blasts but not in normal hematopoietic progenitor cells. p53 activation by targeting the p53-MDM2 interaction might offer a novel therapeutic strategy for AML that retain wild-type p53.

Figure 1.

Blockade of p53-MDM2 interaction inhibits the growth of cell lines with wild-type p53 through cell-cycle arrest and apoptosis induction. (A) Time course of effects induced in OCI-AML-3 and MOLM-13 cells by Nutlin-3a (active enantiomer) and -3b (less active enantiomer) on viable cell number. The cells were incubated with a range of concentrations of Nutlin-3a (1 μM, ▪; 5 μM, •; 10 μM, ▴) or -3b (1 μM, □; 5 μM, ○; 10 μM, ▵), and the cell viability was determined by trypan blue exclusion method. Results are expressed as the percentage of the viable cell number in an untreated group, and represent the average of triplicate cultures. (B) AML cell lines with wild-type p53 (OCI-AML-3 and MOLM-13) or mutant p53 (HL-60 and NB4) cells were incubated with the indicated concentrations of Nutlin-3a or -3b for 72 hours and the cell viability was determined by trypan blue exclusion method. MDM2 inhibitor showed significant cytotoxic activity in OCI-AML-3 and MOLM-13 cells. Results are expressed as the mean plus or minus the standard deviation (SD). *P < .05. (C) Nutlin-3a causes cell-cycle arrest in leukemia cells with wild-type p53. OCI-AML-3 (○), MOLM-13 (□), HL-60 (•), and NB4 (▪) cells were cultured for 12 hours in the presence of Nutlin-3a at the indicated concentrations, and stained for DNA content. Cell-cycle distribution was analyzed using ModFit LT software. Results are expressed as percentage of S-phase cells in DMSO-treated group. Nutlin-3a at 2.5 μM induced almost maximal cell-cycle arrest in both OCI-AML-3 and MOLM-13 cells. Results are representative of 3 independent experiments. (D) Cells were incubated with the indicated concentrations of Nutlin-3a or -3b for 48 hours (24 hours for MOLM-13 cells), and the annexin V–positive fractions were measured by flow cytometry. □ represents untreated controls. Results are expressed as mean ± SD.

Figure 2.

Blockade of p53-MDM2 binding induces apoptosis in primary AML cells. (A) Primary AML cells from patients 1 to 5 were incubated with the indicated concentrations of Nutlin-3a (active enantiomer, ▪) or -3b (less active enantiomer, ) for 72 hours, and the annexin V–positive fractions were measured by flow cytometry. □ represents DMSO-treated controls. (B) Primary AML cells from patient 8 were treated with the indicated concentrations of Nutlin-3a, and flow cytometry analysis using CD34-PE and annexin V–FITC antibodies was performed at the indicated times.

Figure 3.

Positive correlation between MDM2 levels and apoptosis and effect of MDM2 inhibitor on AML progenitors in primary AML samples. (A) Correlation of MDM2 protein levels relative to normal bone marrow cells with degree of apoptosis by MDM2 inhibition in 16 primary AML cells with wild-type p53. MDM2 protein expression levels relative to an internal control, β-actin, were determined in each sample, and then compared with normal bone marrow cells. There was a significant positive correlation between relative MDM2 levels and the percentage of specific annexin V induction (r = 0.805, P < .01). (B) Effect of Nutlin-3a on AML and normal clonogenic progenitors. Data represent average results from 5 different AML samples (white bar) and 5 magnetically separated normal CD34⁺ cells (gray bars). Results are expressed as the mean plus or minus the standard error of the mean (SEM) of the number of colonies in the presence of 1 μM Nutlin-3a compared with the number in control cells. *P < .05.

Figure 4.

Transcriptional activation of proapoptotic Bcl-2 family proteins and transcription-independent apoptosis mediate Nutlin-induced apoptosis. (A) Expression of apoptosis- and cell-cycle–associated proteins in OCI-AML-3 cells, which were treated with 5 μM Nutlin-3a for the indicated times. Nutlin-3a induced increased protein expression of p53, MDM2, and p21 in OCI-AML-3 cells in a time-dependent fashion. Nutlin-3a induced Noxa, a BH3-only member of the Bcl-2 family, followed by caspase activation. β-actin was used to confirm equal loading of proteins. Arrowheads indicate cleared caspases. (B) Expression of proapoptotic Bcl-2 family proteins in primary AML cells, which were treated with 5 μM Nutlin-3a for the indicated times. Nutlin-3a induced at least 1 proapoptotic Bcl-2 family member protein in cells from 3 primary AML samples (nos. 6, 14, and 18 in Table 1) examined. Nutlin-3a induced Noxa, Puma, and Bax up-regulation in case 14 (top panel), Puma in case 18 (bottom panel), and Noxa in case 6 (not shown). β-actin was used to confirm equal loading of proteins. (C) OCI-AML-3 cells or MOLM-13 cells at a starting concentration of 2 × 10⁵ cells/mL were cultured for 24 hours in the presence of DMSO (□), 3.5 μM cycloheximide (), 10 μM Nutlin-3a (), or a combination of cycloheximide and Nutlin-3a (▪). Δψ_m was assessed by flow cytometry. Results are expressed as mean ± SD of triplicate measurements. Comparable results were obtained in 2 other independent experiments. *P < .05.

Figure 5.

p53 localization and its interaction with Bcl-X_L. (A) Localization of p53 in OCI-AML-3 and MOLM-13 cells. Cells were treated with 10 μM Nutlin-3a for 3 hours. Cells were fixed, stained for p53 (green) and mitochondrial marker protein cytochrome c oxidase IV (red), and visualized by confocal microscopy. Localization of p53 to mitochondria is indicated by the yellow-orange color in the merged images. (B) OCI-AML-3 and MOLM-13 cells were incubated with 10 μM Nutlin-3a for 6 hours, and untreated (–) or treated (+) cells were immunoprecipitated with anti-p53 or anti–Bcl-X_L antibodies and immunoblotted for Bcl-X_L or p53. Immunoprecipitation with anti–Bcl-X_L followed by p53 blotting showed a specific p53/Bcl-X_L complex in both cells. The complex was not detectable in reverse, suggesting that a minor fraction of increased p53 protein might bind to Bcl-X_L protein.

Figure 6.

Representative p53 localization patterns in primary AML cells from case 14, which were treated with 10 μM Nutlin-3a for 3 hours. Cells were stained for p53 (green) and mitochondrial marker protein cytochrome c oxidase IV (red) and visualized by confocal microscopy. Nuclei were counterstained with DAPI (blue). Localization of p53 to mitochondria is indicated by the yellow-orange color in the merged images. (A) Untreated cells showed low levels of diffusely distributed p53. (B) After treatment, individual cells showed either cytoplasmic, cytoplasmic and nuclear, or nuclear accumulation of p53. In this case, a majority (75%) of cells showed cytoplasmic accumulation of p53, and the inhibitory effect of cycloheximide on loss of Δψ_m compared with Nutlin-3a alone was 26%. A preferential translocation of cytoplasmic p53 to mitochondria suggests that cytoplasmic p53 mediates apoptosis mainly at the level of mitochondria.