Ceramide-induced integrated stress response overcomes Bcl-2 inhibitor resistance in acute myeloid leukemia

Abstract

Inducing cell death by the sphingolipid ceramide is a potential anticancer strategy, but the underlying mechanisms remain poorly defined. In this study, triggering an accumulation of ceramide in acute myeloid leukemia (AML) cells by inhibition of sphingosine kinase induced an apoptotic integrated stress response (ISR) through protein kinase R-mediated activation of the master transcription factor ATF4. This effect led to transcription of the BH3-only protein Noxa and degradation of the prosurvival Mcl-1 protein on which AML cells are highly dependent for survival. Targeting this novel ISR pathway, in combination with the Bcl-2 inhibitor venetoclax, synergistically killed primary AML blasts, including those with venetoclax-resistant mutations, as well as immunophenotypic leukemic stem cells, and reduced leukemic engraftment in patient-derived AML xenografts. Collectively, these findings provide mechanistic insight into the anticancer effects of ceramide and preclinical evidence for new approaches to augment Bcl-2 inhibition in the therapy of AML and other cancers with high Mcl-1 dependency.

Graphical abstract

Figure 1.

The BH3-only proteins Noxa and Bim are essential for MP-A08–induced cell death. (A) Mcl-1 immunoprecipitation of MV411 cells treated with MP-A08 for 6 hours and subjected to immunoblot analysis with the indicated antibodies. (B) Quantitative polymerase chain reaction analysis of MV411 cells treated with MP-A08 (20 μM) over 6 hours. (C) Immunoblot analysis of MV411 cells treated with MP-A08 in the presence of cycloheximide for 6 hours. MV411 (D) or OCI-AML3 (E) were treated with increasing concentrations of MP-A08 for 6 hours and subjected to immunoblot analysis with the indicated antibodies. MV411 cells were lentivirally transduced with shRNAs targeting Noxa (F), Bim (G), or Bid (H) and treated with doxycycline for 48 hours and MP-A08 (15 μM) for either 6 hours for immunoblot analysis or 24 hours for cell viability using Annexin V staining. All qualitative data are representative of at least 3 independent experiments, and all quantitative data are the mean ± standard error of the mean from 3 independent experiments. Statistical significance was assessed by Student t test. *P < .05; **P < .005; ***P < .0001. N.S., not significant.

Figure 2.

MP-A08 induces ATF4 dependent Noxa transcription. (A) MV411 cells were treated with MP-A08 (20 μM), daunorubicin (DNR) (1 μM), or cytarabine (Ara-C; 1 μM) for 6 hours; lysed; and subjected to immunoblot analysis with the indicated antibodies. (B) The UPR. (C) MV411 cells were treated with MP-A08 (20 μM) over a 6-hour period, lysed, and subjected to immunoblot analysis with the indicated antibodies. (D) MV411 cells were stably transduced with 2 different CRISPR guide sequences targeting SPHK1 (g1 and g2), lysed, and subjected to immunoblot analysis with the indicated antibodies. The efficiency of the SPHK1 knockout was confirmed via SPHK1 activity assays of lysates from those cells in assay conditions largely selective for SPHK1 over SPHK2 (supplemental Figure 5C). (E) MV411 cells were treated with MP-A08 (20 μM), alone or in combination with the eIF2b agonist, ISRIB (200 nM) over 6 hours for quantitative polymerase chain reaction analysis of Noxa messenger RNA levels and immunoblot analysis with the indicated antibodies. Statistical significance was assessed by Student t test. *P < .05 (n = 3). (F) MV411 cells were stably transduced with a doxycycline-inducible shRNA targeting ATF4. Cells were treated with 1 μg/mL doxycycline for 48 hours and MP-A08 (20 μM) for 6 hours before cell lysis for immunoblot analysis. ns, a nonspecific band. (G) Chromatin immunoprecipitation analysis of the Noxa promoter in response to MP-A08 treatment (20 μM) of MV411 cells for 6 hours. Statistical significance was assessed by Student t test. *P < .05 (n = 4). (H) Primary AML samples were treated with increasing concentrations of MP-A08 and subjected to immunoblot analysis with the indicated antibodies. All qualitative data are representative of at least 3 independent experiments, and all quantitative data represent the mean ± standard error of the mean of at least 3 independent experiments.

Figure 3.

Ceramides drive a PERK-independent ISR. (A-C) MV411 cells were treated with vehicle control (0.1% dimethyl sulfoxide [DMSO]; blue bars) or 20 μM MP-A08 (red bars) for 6 hours and analyzed by liquid chromatography-mass spectrometry. (A) Quantitation of individual ceramide (Cer) and dihydroceramide (dhCer) species. (B) Quantitation of sphingosine (Sph) and dihydrosphingosine (dhSph) species. (C) Quantitation of S1P and dihydrosphingosine 1-phosphate (dhS1P). All data are presented as picomoles per million cells; mean ± standard deviation of 4 independent experiments. Statistical significance was assessed by Student t test. *P < .05; **P < .01; ***P < .001. (D) MV411 cells were treated with MP-A08 (20 μM), PF-543 (1 μM), and SK1-I (10 μM) for 6 hours; lysed; and subjected to immunoblot analysis. (E) MV411 cells were treated with ceranib-2 for 16 hours and assessed for cell viability by Annexin V/propidium iodide staining. Data are the mean ± standard deviation of 2 independent experiments. MV411 cells were treated with ceranib-2 (10 μM) for 6 hours and subjected to immunoblot analysis with the indicated antibodies. (F) MV411 cells were treated with MP-A08 (20 μM), C2-ceramide (10 μM), C6-ceramide (10 μM), C2-dhCeramide (10 μM), and C6-dhCeramide (10 μM) (all introduced from 2.5 mM stock solutions in dimethyl sulfoxide) for 6 hours; lysed; and subjected to immunoblot analysis with the indicated antibodies. (G) HAP1 wild-type and PERK^−/− cells were treated with MP-A08 (20 μM) for 6 hours, lysed, and subjected to immunoblot analysis with the indicated antibodies. (H-I) MV411 cells were stably transduced with a doxycycline-inducible shRNA targeting PERK. Cells were treated with 1 μg/mL doxycycline for 48 hours and MP-A08 (20 μM) for 6 hours before (H) quantitative PCR analysis of PERK messenger RNA levels and (I) immunoblot analysis with the indicated antibodies. Data are the mean ± standard error of the mean of 3 independent experiments. Statistical significance was assessed by Student t test. **P < .01.

Figure 4.

Ceramides drive a PKR-dependent integrated stress response. (A) The ISR. (B-C) MV411, OCI-AML3, MOLM13, HL-60, and THP-1 cells were treated with MP-A08 alone (20 μM) or in combination with the GCN2 inhibitor A-92 (5 μM), PKR inhibitors C16 (5 μM) or 2-AP (0.1-10 mM) or PERK inhibitor AMG-44 (5 μM) for 6 hours before immunoblot analysis with the indicated antibodies. (D) MV411 cells were treated with MP-A08 (10 μM) and 2-AP (5 mM) for 16 hours and assessed for cell viability by Annexin V/propidium iodide staining. Data are the mean ± standard error of the mean (SEM) of 3 independent experiments. Statistical significance was assessed by Student t test. (E) Wild-type (WT) or PKR knockout MV411 cells were treated with MP-A08 (15 μM) for 16 hours and assessed for cell viability by Annexin V/propidium iodide staining. Data are the mean ± SEM of 4 independent experiments. Statistical significance was assessed by Student t test. Immunoblot analysis of WT or PKR knockout MV411 cells with the indicated antibodies. (F) PKR-HA was immunoprecipitated from transiently transfected HEK 293T cells, incubated with exogenous C6-ceramide (10 μM) for 30 minutes and subjected to a PKR activity assay, using autophosphorylation as the readout. Data are the mean ± SEM of 4 independent experiments. (G) Lysates from HEK293T cells transfected with pcDNA3/PKR-HA was incubated with ceramide conjugated to agarose beads or control beads overnight at 4°C, washed, and resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and the associated PKR was detected by immunoblot analysis with anti-HA antibodies on an Odyssey imaging system.

Figure 5.

MP-A08 and venetoclax induces potent synergistic activity in AML cell lines. (A) MV411 cells were treated for up to 6 hours with 20 μM MP-A08 (blue bars), 10 nM venetoclax (red bars), alone or in combination (green bars). Cell viability was analyzed every 2 hours by Annexin V/propidium iodide staining. Data are means ± standard error of the mean (SEM; n = 4). Statistical significance was assessed by Student t test. ****P < .0001. Drug synergy was assessed with the Chou-Talay combination index (CI) method whereby CI values <1 are classified as synergy. (B) Factor-dependent myeloid wild-type and Bax/Bak^−/− cells treated with MP-A08 (20 μM) and/or venetoclax (10 nM). Data are the mean ± SEM (n = 4). Statistical significance was assessed by Student t test. ****P < .0001. MV411 (C-D), HL-60 (E), MOLM13 (F), and OCI-AML3 (G). (H) MV411 cells were stably transduced with a doxycycline-inducible shRNA targeting Bcl-2 and treated with doxycycline (1 μg/mL) for 48 hours and MP-A08 (10 μM) for 24 hours. Cell viability was assessed by Annexin V staining. All data are the mean ± SEM of 3 independent experiments. Statistical significance was assessed by Student t test. ***P < .0001. (I) UT-7 cells were treated with MP-A08 and venetoclax for 24 hours and assessed for cell viability by Annexin V/propidium iodide staining. Drug synergy was assessed using the Chou-Talay CI method. (J) UT-7 cells were treated with increasing concentrations of MP-A08 for 6 hours and subjected to immunoblot analysis. MV411 (K) and OCI-AML3 (L) cells were treated with MP-A08, venetoclax, or in combination for 6 hours and subjected to immunoblot analysis with the indicated antibodies. All qualitative data are representative of at least 3 independent experiments, and all quantitative data are the mean ± SEM of at least 3 independent experiments.

Figure 6.

MP-A08 and venetoclax treatment exhibits antileukemic activity in primary AML samples. (A) Primary AML cells were treated with increasing concentrations of MP-A08 for 6 hours, lysed, and subjected to immunoblot analysis with indicated antibodies. (B-C) Primary AML samples were treated with MP-A08 and venetoclax for 6 hours and assessed for cell viability by Annexin V staining. Data are displayed as the mean ± range of duplicate technical replicates. Statistical significance was assessed by Student t test. *P < .05; ***P < .0005. Synergy was determined by the CI method. Fluorescence-activated cell sorting–purified iLSCs were seeded alone (D) or on an MSC coculture layer (E), treated with MP-A08 and venetoclax for 24 hours, and assessed for cell viability by Annexin V staining. Data are displayed as the mean ± range of duplicate technical replicates. Statistical significance was assessed by Student t test. *P < .05. Synergy was determined by the Webb fractional product method. (F) Normal bone marrow–derived CD34⁺ cells were treated with MP-A08 and venetoclax for 24 hours and assessed for cell viability by Annexin V staining. Data are displayed as the mean ± range of duplicate technical replicates. (G) Representative immunohistochemistry staining using human specific mitochondrial antibody (MTC02) of an NSG mouse sternum engrafted with primary AML cells. Bar represents 100 μm. (H-I) NSG mice were engrafted with primary AML blasts and bled weekly to confirm disease engraftment (>1% hCD45⁺ in peripheral blood). Mice received vehicle, MP-A08 (100 mg/kg intraperitoneally), venetoclax (75 mg/kg orally), or both daily for 2 weeks. Engraftment was quantified by assessing the percentage of human CD45⁺ cells in the bone marrow of recipient mice. Each symbol represents the percentage of CD45⁺ cells observed in a separate mouse. Significance was assessed by Student t test. (J) Mutational analysis of AML patient samples treated with venetoclax from the Beat AML Project. The average area under the curve (AUC) is a measure of drug sensitivity (the higher the AUC, the more resistant) derived from ex vivo drug sensitivity assays. Statistical significance was assessed by Student t test with Welch’s correction. **P < .01; ***P < .0001. (K) Primary AML samples identified by whole-exome sequencing containing PTPN11, TP53 (L), or K-Ras (M) mutations were treated with MP-A08 and venetoclax for 6 hours and assessed for cell viability by Annexin V staining. Data are displayed as the mean ± range of duplicate technical replicates. Statistical significance was assessed by Student t test. **P < .01.

Figure 6.

MP-A08 and venetoclax treatment exhibits antileukemic activity in primary AML samples. (A) Primary AML cells were treated with increasing concentrations of MP-A08 for 6 hours, lysed, and subjected to immunoblot analysis with indicated antibodies. (B-C) Primary AML samples were treated with MP-A08 and venetoclax for 6 hours and assessed for cell viability by Annexin V staining. Data are displayed as the mean ± range of duplicate technical replicates. Statistical significance was assessed by Student t test. *P < .05; ***P < .0005. Synergy was determined by the CI method. Fluorescence-activated cell sorting–purified iLSCs were seeded alone (D) or on an MSC coculture layer (E), treated with MP-A08 and venetoclax for 24 hours, and assessed for cell viability by Annexin V staining. Data are displayed as the mean ± range of duplicate technical replicates. Statistical significance was assessed by Student t test. *P < .05. Synergy was determined by the Webb fractional product method. (F) Normal bone marrow–derived CD34⁺ cells were treated with MP-A08 and venetoclax for 24 hours and assessed for cell viability by Annexin V staining. Data are displayed as the mean ± range of duplicate technical replicates. (G) Representative immunohistochemistry staining using human specific mitochondrial antibody (MTC02) of an NSG mouse sternum engrafted with primary AML cells. Bar represents 100 μm. (H-I) NSG mice were engrafted with primary AML blasts and bled weekly to confirm disease engraftment (>1% hCD45⁺ in peripheral blood). Mice received vehicle, MP-A08 (100 mg/kg intraperitoneally), venetoclax (75 mg/kg orally), or both daily for 2 weeks. Engraftment was quantified by assessing the percentage of human CD45⁺ cells in the bone marrow of recipient mice. Each symbol represents the percentage of CD45⁺ cells observed in a separate mouse. Significance was assessed by Student t test. (J) Mutational analysis of AML patient samples treated with venetoclax from the Beat AML Project. The average area under the curve (AUC) is a measure of drug sensitivity (the higher the AUC, the more resistant) derived from ex vivo drug sensitivity assays. Statistical significance was assessed by Student t test with Welch’s correction. **P < .01; ***P < .0001. (K) Primary AML samples identified by whole-exome sequencing containing PTPN11, TP53 (L), or K-Ras (M) mutations were treated with MP-A08 and venetoclax for 6 hours and assessed for cell viability by Annexin V staining. Data are displayed as the mean ± range of duplicate technical replicates. Statistical significance was assessed by Student t test. **P < .01.

Figure 7.

Ceramides induce ISR activation and sensitize cells to Bcl-2 inhibition. The accumulation of proapoptotic sphingolipids, such as ceramide, in response to SPHK inhibition is sensed by the eIF2a kinase PKR. PKR activation culminates in an apoptotic ISR mediated by master transcription factor ATF4. ATF4 promotes Bcl-2 dependency by the transcriptional upregulation of Noxa and the subsequent binding to and inactivation of Mcl-1.