Phosphorylated AKT inhibits the apoptosis induced by DRAM-mediated mitophagy in hepatocellular carcinoma by preventing the translocation of DRAM to mitochondria

Abstract

Increasing autophagy is beneficial for curing hepatocellular carcinoma (HCC). Damage-regulated autophagy modulator (DRAM) was recently reported to induce apoptosis by mediating autophagy. However, the effects of DRAM-mediated autophagy on apoptosis in HCC cells remain unclear. In this study, normal hepatocytes (7702) and HCC cell lines (HepG2, Hep3B and Huh7) were starved for 48 h. Starvation induced apoptosis and autophagy in all cell lines. We determined that starvation also induced DRAM expression and DRAM-mediated autophagy in both normal hepatocytes and HCC cells. However, DRAM-mediated autophagy was involved in apoptosis in normal hepatocytes but not in HCC cells, suggesting that DRAM-mediated autophagy fails to induce apoptosis in hepatoma in response to starvation. Immunoblot and immunofluorescence assays demonstrated that DRAM translocated to mitochondria and induced mitophagy, which led to apoptosis in 7702 cells. In HCC cells, starvation also activated the phosphatidylinositol 3-kinase (PI3K)/AKT pathway, which blocks the translocation of DRAM to mitochondria through the binding of p-AKT to DRAM in the cytoplasm. Inactivation of the PI3K/AKT pathway rescued DRAM translocation to mitochondria; subsequently, mitochondrial DRAM induced apoptosis in HCC cells by mediating mitophagy. Our findings open new avenues for the investigation of the mechanisms of DRAM-mediated autophagy and suggest that promoting DRAM-mediated autophagy together with PI3K/AKT inhibition might be more effective for autophagy-based therapy in hepatoma.

Figure 1

Starvation-induced autophagy involves apoptosis in 7702 cells but not in three HCC cell lines. The 7702 (wild-type p53) and HCC cell lines (HepG2 (wild-type p53), Hep3B (p53 null) and Huh7 (p53 ^A220G)) were starved (sta) for 48 h with or without pre-treatment with BafA 1 to inhibit autophagy. (a) M30 immunoreactivity and calcein AM/PI assays were used to detect apoptosis in all cell lines in response to starvation. M30 and PI staining (red) indicated early and late apoptotic cell death, respectively. Nuclei were stained with DAPI. Representative images of cells were obtained with a fluorescence microscope at × 40 magnification. (b) Quantitation of M30-positive and PI-positive cells observed in (a). (c) Autophagy and apoptosis were detected in cells in response to starvation for 48 h via immunoblot analysis with the indicated antibodies (anti-LC3 and p62 antibodies for detecting autophagy; anti-PARP antibody for detecting apoptosis). (d and e) An immunoblot assay was used to detect the effect of autophagy inhibition via BafA 1 on starvation-induced apoptosis in cells. (f and g) Quantification of M30-positive and PI-positive cells starved for 48 h with or without autophagy inhibition using BafA 1. The data are presented as the mean±S.D. of three independent experiments in (b), (f) and (g)

Figure 2

Starvation induces DRAM-mediated autophagy in 7702, HepG2, Hep3B and Huh7 cells. (a and b) The 7702, HepG2, Hep3B and Huh7 cell lines were starved for 48 h, and real-time PCR (a) and immunoblot (b) assays were used to detect DRAM expression. The immunoblot assay was also used to detect the expression of p53, p73 and LC3I/II (b). (c) The 7702, HepG2, Hep3B and Huh7 cell lines were starved (sta) for 48 h with or without pre-transfection with DRAM siRNA (DRAM si) or control siRNA (control si). Confocal microscopy was used to detect the formation of GFP-LC3 puncta. Magnification, × 40. (d) Quantification of cells with >5 GFP-LC3 puncta. (a and d) The data are presented as the mean±S.D. of three independent experiments. (e) The effect of DRAM siRNA on DRAM expression in cells with or without starvation for 48 h was determined using an immunoblot assay

Figure 3

DRAM-mediated autophagy in response to starvation induces apoptosis only in 7702 cells and not in three HCC cell lines. The 7702, HepG2, Hep3B and Huh7 cell lines were transfected with DRAM siRNA (DRAM si) and were then starved (sta) for 48 h. (a) An immunoblot assay was used to detect LC3 I/II and cleaved PARP fragment (p85). (b and c) Quantification of starvation-induced M30-positive and PI-positive cells with or without DRAM knockdown using DRAM si. The data are presented as the mean±S.D. of three independent experiments

Figure 4

p53 is critical for the induction of DRAM-mediated autophagy in 7702 and HepG2 cells expressing wild-type p53. (a) The 7702 and HepG2 cell lines were transfected with p53 siRNA (p53 si) and were then starved (sta) for 48 h. An immunoblot assay was used to detect the effect of p53 knockdown by p53 si on the expression of DRAM, LC3 I/II, p62 and the cleaved PARP fragment (p85). (b) The 7702 and HepG2 cell lines were transfected with DRAM siRNA (DRAM si) or p53 si, or co-transfected with DRAM si and p53 si and were then starved for 48 h. M30 immunoreactivity (red) was used to detect the effect of siRNA-induced DRAM or p53 knockdown or co-knockdown of DRAM and p53 on apoptosis. Representative immunofluorescence images of cells were obtained with a fluorescence microscope at × 40 magnification. Nuclei were stained with DAPI

Figure 5

Both starvation-induced p73 and rAd-p53-induced p53 overexpression can induce DRAM-mediated autophagy in Hep3B and Huh7 cells. (a) Hep3B and Huh7 cells were transfected with p73 siRNA (p73 si) and were then starved (sta) for 48 h. An immunoblot assay was used to detect the effect of p73 knockdown by siRNA on the expression of DRAM, LC3 I/II, p62 and cleaved PARP fragment (p85). (b) Hep3B and Huh7 cells were transfected with DRAM siRNA (DRAM si) or p73 si, or co-transfected with DRAM si and p73 si; the cells were then starved for 48 h. Starvation-induced M30-positive cells were quantified. The data are presented as the mean±S.D. of three independent experiments. (c) Hep3B and Huh7 cells were infected with rAd-p53 and were then starved for 48 h. An immunoblot assay was used to detect the effect of p53 overexpression on the expression of p73, DRAM, LC3 I/II and cleaved PARP fragment (p85). (d) Hep3B and Huh7 cells were infected with rAd-p53 with or without pre-treatment with DRAM siRNA and subsequently starved for 48 h. An immunoblot assay was used to detect the effect of DRAM knockdown via siRNA on autophagy. (e) rAd-p53-infected Hep3B and Huh7 cells were pre-treated with DRAM siRNA and were then starved for 48 h. M30 immunoreactivity (red) was used to detect the effect of siRNA-mediated DRAM knockdown on p53 overexpression-induced apoptosis. Nuclei were stained with DAPI. Representative immunofluorescence images of cells were obtained with a fluorescence microscope at × 40 magnification

Figure 6

Activation of the PI3K/AKT signaling pathway inhibits the effect of DRAM-mediated autophagy on apoptosis in HCC cell lines. (a) An immunoblot assay was used to detect the activation of the PI3K/AKT pathway in 7702, HepG2, Hep3B and Huh7 cells starved (sta) for 48 h. (b) Cells were starved for 48 h with or without pre-treatment by transfection with PI3K siRNA (PI3K si). The ratio of apoptotic cells was determined by quantification of M30-positive cells. (c) An immunoblot assay was used to detect the effect of siRNA-induced PI3K knockdown on the expression of p53, p73, DRAM and LC3 I/II. (d) HepG2, Hep3B and Huh7 cells were transfected with DRAM siRNA (DRAM si) or co-transfected with DRAM and PI3K siRNAs; the cells were then starved for 48 h. The ratio of apoptotic cells was determined by quantification of M30-positive cells. (e) Cells were starved for 48 h with or without pre-transfection with PI3K si. The ratio of apoptotic cells was determined by quantification of M30-positive cells. The data are presented as the mean±S.D. of three independent experiments in (b), (d) and (e)

Figure 7

Phosphorylated AKT binds DRAM and prevents it from translocating to mitochondria in three HCC cell lines. (a) HepG2, Hep3B and Huh7 cells were starved (sta) for 48 h with or without pre-transfection with PI3K siRNA (PI3K si). The cytoplasm (C) and mitochondria (M) were extracted using Percoll gradient centrifugation. The levels of DRAM, LC3 I/II and p-AKT were evaluated by immunoblot assay. (b) 7702 cells were starved for 48 h, and an immunoblot assay was used to evaluate the levels of DRAM and LC3 I/II in the extracted cytoplasm and mitochondria. Anti-VDAC and anti-β-actin antibodies were used as controls for the extracted mitochondria and cytoplasm in (a) and (b). (c and d) p-AKT binds to DRAM in the cytoplasm but not in the mitochondria of HCC cells after starvation treatment for 48 h. HepG2, Hep3B (c) and Huh7 (d) cells were starved for 48 h, and the mitochondria and cytoplasm were extracted using Percoll gradient centrifugation. An anti-DRAM antibody was used to immunoprecipitate DRAM in the extracted mitochondria and cytoplasm. Total cell lysates were used as input or were immunoprecipitated with control IgG as indicated. DRAM and p-AKT were then analyzed by immunoblot assay using anti-DRAM and p-AKT antibodies. (e) Confocal microscopy was used to detect DRAM (green) and p-AKT (red) colocalization, as shown by the yellow sites in the cytoplasm of HepG2, Hep3B and Huh7 cells, in response to starvation for 48 h. DAPI was used to stain nuclei. Magnification, × 1000. (f) The diagram summarizes the interactions between DRAM-mediated mitophagy and the PI3K/AKT pathway in normal hepatocytes (blue) and HCC cells (red) (details are provided in the Results section)