BCR-ABL suppresses autophagy through ATF5-mediated regulation of mTOR transcription

Abstract

The oncoprotein BCR-ABL transforms myeloid progenitor cells and is responsible for the development of chronic myeloid leukemia (CML). In transformed cells, BCR-ABL suppresses apoptosis as well as autophagy, a catabolic process in which cellular components are degraded by the lysosomal machinery. The mechanism by which BCR-ABL suppresses autophagy is not known. Here we report that in both mouse and human BCR-ABL-transformed cells, activating transcription factor 5 (ATF5), a prosurvival factor, suppresses autophagy but does not affect apoptosis. We find that BCR-ABL, through PI3K/AKT/FOXO4 signaling, transcriptionally up-regulates ATF5 expression and that ATF5, in turn, stimulates transcription of mammalian target of rapamycin (mTOR; also called mechanistic target of rapamycin), a well-established master negative-regulator of autophagy. Previous studies have shown that the BCR-ABL inhibitor imatinib mesylate induces both apoptosis and autophagy, and that the resultant autophagy modulates the efficiency by which imatinib kills BCR-ABL-transformed cells. We demonstrate that imatinib-induced autophagy is because of inhibition of the BCR-ABL/PI3K/AKT/FOXO4/ATF5/mTOR pathway that we have identified in this study.

Figure 1

ATF5 suppresses cell death in normal but not BCR-ABL–transformed cells. (A) Cell viability analysis in 32D or 32D/BCR-ABL cells stably expressing a nonsilencing (NS) or ATF5 shRNA. Error bars represent SD. For each cell line, values were normalized to those observed at day 0. (B) 32D or 32D/BCR-ABL cells were treated with a luciferase or ATF5 siRNA and monitored for cell death by annexin V–PE staining. Error bars represent SD. *P < .05; **P > .05.

Figure 2

ATF5 inhibits autophagy in BCR-ABL–transformed cells. (A) 32D/BCR-ABL or K562 cells treated with either a NS or ATF5 shRNA were monitored for expression of ATF5 and LC3B by immunoblot analysis. β-actin (ACTB) was monitored as a loading control. The fold change in LC3B-II levels on ATF5 knockdown was quantified by measuring the intensity of the LC3B-II signal in the immunoblots and normalizing to ACTB levels. (B) 32D/BCR-ABL or K562 cells ectopically expressing a GFP-LC3B fusion protein were treated with either a NS or ATF5 shRNA and analyzed by fluorescence microscopy. Representative images are shown. (C) Immunoblot analysis of p62 levels in K562 cells expressing a NS or ATF5 shRNA. (D) K562 cells expressing a NS or ATF5 shRNA were stained with acridine orange in the absence or presence of bafilomycin A1 (Baf A1). (E) Immunoblot analysis of ATF5 and LC3B levels in K562 cells stably expressing FLAG or FLAG-ATF5, and treated in the absence or presence of imatinib (IM; 10μM for 48 hours).

Figure 3

BCR-ABL regulates ATF5 expression independent of IL-3. (A) qRT-PCR analysis monitoring Atf5 expression in 32D or 32D/BCR-ABL cells cultured in the presence or absence of IL-3. Error bars represent SD. (B) Immunoblot analysis of ATF5 levels in 32D or 32D/BCR-ABL cells cultured in the presence or absence of IL-3. (C) qRT-PCR monitoring Atf5 expression in 32D/BCR-ABL or 32D/BCR-ABL(T315I) cells treated in the absence or presence of imatinib (10μM for 16 hours). Error bars represent SD. (D) Immunoblot analysis of ATF5 levels in 32D/BCR-ABL or 32D/BCR-ABL(T315I) cells in the absence or presence of imatinib (10μM for 16 hours). (E) qRT-PCR analysis monitoring ATF5 expression in K562 cells treated with or without imatinib (10μM for 48 hours). Error bars represent SD. (F) Immunoblot analysis of ATF5 levels in K562 cells treated with or without imatinib (10μM for 48 hours). (G) qRT-PCR monitoring of ATF5 expression in human peripheral blood cells isolated from chronic-phase CML patients (n = 3). Cells were treated in the absence or presence of imatinib (10μM for 16 hours). Error bars represent SE. *P < .05; **P > .05.

Figure 4

BCR-ABL stimulates ATF5 transcription through PI3K/AKT/FOXO4 signaling. (A) qRT-PCR monitoring ATF5 expression in 32D/BCR-ABL or K562 cells treated in the absence of presence of the PI3K inhibitor LY294002 (20μM for 48-72 hours). Error bars represent SD. (B) qRT-PCR monitoring ATF5 expression in 32D/BCR-ABL or K562 cells expressing either empty vector or a vector encoding PIK3CA(E545K). Error bars represent SD. (C) Luciferase reporter assays. A 1548-bp Atf5 promoter fragment, or as a control empty vector, was tested for its ability to drive a heterologous firefly luciferase gene after transient transfection in 32D/BCR-ABL cells. Error bars represent SD. (D) Luciferase reporter assays. 32D/BCR-ABL cells transfected with the reporter construct containing 1548 bp of the Atf5 promoter were treated with either DMSO, imatinib (5μM for 24 hours), LY294002 (20μM for 48 hours), PHT-427 (20μM for 24 hours), or JAK inhibitor I (10μM for 24 hours). Error bars represent SD. (E) qRT-PCR monitoring Atf5 expression in 32D/BCR-ABL cells stably expressing either a NS or FOXO4 shRNA and treated in the absence or presence of LY294002 (20μM for 18 hours). Error bars represent SD. (F) Luciferase reporter assay. 32D/BCR-ABL cells transfected with the reporter construct containing 1548 bp of the Atf5 promoter and expressing either an NS or FOXO4 shRNA were treated in the absence of presence of LY294002 (20μM for 18 hours). Error bars represent SD. *P < .05; **P > .05.

Figure 5

Signaling through PI3K/AKT inhibits autophagy in BCR-ABL–transformed cells. (A) Immunoblot analysis monitoring the levels of phosphorylated AKT, ATF5, LC3B, and ACTB in 32D/BCR-ABL cells treated in the absence or presence of LY294002 (20μM for 48 hours). (B) Immunoblot analysis monitoring the levels of phosphorylated AKT, ATF5, LC3B, and ACTB in 32D/BCR-ABL cells expressing either empty vector or a vector encoding PIK3CA(E545K). (C) Immunoblot analysis monitoring phosphorylated AKT, ATF5, and LC3B levels in 32D/BCR-ABL cells treated with imatinib (5μM for 24 hours). (D) Immunoblot analysis monitoring phosphorylated AKT, ATF5, and LC3B levels in imatinib-treated 32D/BCR-ABL cells (5μM for 24 hours) expressing empty vector or PIK3CA(E545K).

Figure 6

mTOR, a master negative regulator of autophagy, is an ATF5 target gene. (A) qRT-PCR analysis monitoring expression of 62 autophagy-related genes in 32D/BCR-ABL cells after treatment with an ATF5 siRNA. Expression of Atf5 is shown as a positive control. The expression of each gene was normalized to that obtained after treatment of cells with a luciferase siRNA. The red line indicates 2-fold decrease in expression. Error bars represent SD. (B) 32D/BCR-ABL cells treated with a NS or ATF5 shRNA were monitored for mTOR and phosphorylated mTOR levels by immunoblot analysis. (C) qRT-PCR analysis (left) or immunoblot analysis (right) monitoring mTOR expression in K562 cells treated with either a NS or ATF5 shRNA. Error bars represent SD. (D) ChIP analyses monitoring binding of ATF5 to the mTOR promoter in 32D/BCR-ABL cells in the presence or absence of imatinib (5μM for 24 hours). (E) Luciferase reporter assay. Fragments containing 539, 2010, 2017, or 3007 bp of the mTOR promoter, or as a control empty vector, were tested for their ability to drive heterologous firefly luciferase gene after transient transfection in 32D/BCR-ABL cells. Error bars represent SD. (F) 32D/BCR-ABL cells transfected with the reporter construct containing 3007 bp of the mTOR promoter were treated with either DMSO, imatinib (5μM for 24 hours), LY294002 (20μM for 24 hours), PHT-427 (20μM for 24 hours) or JAK inhibitor I (10μM for 24 hours). (G) qRT-PCR monitoring mTOR expression in 32D/BCR-ABL cells stably expressing either a NS or FOXO4 shRNA and treated in the absence or presence of LY294002 (20μM for 18 hours). Error bars represent SD. *P < .05; **P > .05.

Figure 7

ATF5 inhibits autophagy in BCR-ABL–transformed cells by transcriptional stimulation of mTOR expression. (A) 32D/BCR-ABL or K562 cells treated with either DMSO or rapamycin were monitored for ATF5 and LC3B levels by immunoblot analysis. (B) 32D/BCR-ABL or K562 cells ectopically expressing a GFP-LC3B fusion protein were treated with DMSO, rapamycin, or PP242 and analyzed by fluorescence microscopy. Representative images are shown. (C) qRT-PCR monitoring mTOR mRNA levels in imatinib-treated 32D/BCR-ABL (5μM for 24 hours) and K562 cells (10μM for 48 hours). (D) qRT-PCR monitoring of mTOR expression in human peripheral blood cells isolated from chronic-phase CML patients (n = 3). Cells were treated in the absence or presence of imatinib (10μM for 16 hours). Error bars represent SE. *P < .05. (E) Immunoblot analysis of mTOR levels in K562 cells stably expressing FLAG or FLAG-ATF5, and treated in the absence or presence of imatinib (10μM for 48 hours). Tubulin was monitored as a loading control. (F) Schematic representation of the BCR-ABL/PI3K/AKT/FOXO4/ATF5/mTOR-mediated autophagy-inhibition pathway. Fading of a protein indicates loss of function.