The oxidation states of DJ-1 dictate the cell fate in response to oxidative stress triggered by 4-hpr: autophagy or apoptosis?

Abstract

Aim: Chemotherapy-induced reactive oxygen species (ROS) not only contribute to apoptosis, but also trigger autophagy. Since autophagy is reported to protect cancer cells from apoptosis, this weakens the therapeutic effect of chemotherapy. This study aimed at identifying the key molecules that determine the cellular response to ROS and, therefore, provide better strategies to increase chemotherapeutic efficiency.

Results: Increasing concentrations of N-(4-hydroxyphenyl) retinamide (4-HPR)-treatment pushed autophagy down to apoptosis in a dose-dependent manner, and 4-HPR-induced ROS contribute to this process. Since we found that ASK1-regulated JNK1 and p38 are responsible for 4-HPR-induced autophagy and apoptosis, respectively, we further utilized co-immunoprecipitation followed by liquid chromatography-tandem mass spectrometry analysis to identify proteins that specifically bind to ASK1 under different oxidative states. Of note, DJ-1, a crucial antioxidant protein, was identified. Interestingly, DJ-1 functions as a redox sensor that senses ROS levels and determines the cellular response to 4-HPR: Under mild oxidative stress, moderate oxidation of DJ-1 is recruited to inhibit the activity of ASK1 and maintain cell viability by activating autophagy; under a lethal level of oxidative stress, excessive oxidized DJ-1 dissociates from ASK1 and activates it, thereby initiating p38 activation and enabling the cells to commit to apoptosis. Moreover, the depletion of DJ-1 increases the sensitivity of tumor cells to 4-HPR both in vitro and in vivo.

Innovation: Our results reveal that the different oxidation states of DJ-1 function as a cellular redox sensor of ROS caused by 4-HPR and determine the cell fate of autophagy or apoptosis. Moreover, the results suggest that DJ-1 might be a potent therapeutic target for cancer treatment.

Conclusion: ROS-mediated changes in the oxidation state of DJ-1 are involved in 4-HPR's effect on pushing autophagy down to apoptosis. Consequently, this change mediates ASK1 activation by regulating DJ-1-ASK1 complex formation and determines the cell fate of autophagy or apoptosis.

FIG. 1.

4-HPR pushes autophagy down to apoptosis in a concentration threshold-dependent manner in vitro and in vivo. (A–C) HeLa cells were treated with serial concentrations of 4-HPR for 24 h and evaluated to determine autophagy and apoptosis. (A) Total cell extracts were probed with antibodies against LC-3B, PARP, and β-Actin. The density of the immunoreactive bands was calculated by Quantity One software, and the data are shown as the ratio versus β-Actin. The data are expressed as the mean±standard deviation (SD) (n=3, **p<0.01, ***p<0.001 vs. control). CF, cleaved fragment of PAPP; I, LC3-I; II, LC3-II. (B) Acridine orange (AO) staining, MDC staining, GFP-LC3 transfection assays, and electron microscopy were performed. White arrow: appearance of autophagosomes; yellow arrow: autophagosomes; red arrow: condensation of chromatin. (C) Fluorescence activated cell sorting (FACS) quantification of acidic vesicular organelles (AVOs) or apoptosis with acridine orange or annexin V/PI in HeLa cells exposed to 4-HPR. (D–G) The mice transplanted with human xenografts (HeLa) were randomly divided into three groups and given injections of serial doses of 4-HPR (20 or 40 mg/kg/day, i.v.) for a period of 7 days (n=at least 4). The tumor volume was recorded daily, and the relative tumor volume (RTV) was calculated. At the end of the experiment, the tumor was weighed and analyzed with TUNEL staining and western blotting. (D) RTV was expressed as the mean± standard error (SE). *p<0.05, **p<0.01, 40 mg/kg versus control; ^#p<0.05, ^##p<0.01, 40 mg/kg versus 20 mg/kg. (E) Representative images of tumors in the different groups are shown along with the tumor weight (mean±SE), inhibition rate, and T/C value (RTV of treatment/RTV of control). (F) The TUNEL-positive rate of each tumor was analyzed (top panel), and representative merged images of TUNEL staining of tumor tissues in the different groups (bottom panel) are shown. (G) Expression of LC-3B and β-Actin in tumor tissues from the different groups was detected by western blotting. To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

FIG. 2.

The effect of 4-HPR on pushing autophagy down to apoptosis is dependent on the levels of 4-HPR-induced ROS. (A) HeLa cells were exposed to 5 or 10 μM 4-HPR for 1 h, and the mean level of ROS was determined using carboxy-DCFDA. (B) HeLa cells were pretreated with NAC (2 or 4 mM) for 0.5 h and exposed to 4-HPR for the next 24 h. Total cell extracts were probed with antibodies against LC-3B, PARP, and β-Actin. The density of immunoreactive bands was calculated by Quantity One software, and the data are shown as the ratio versus β-Actin. The data are expressed as the mean±SD (n=3, *p<0.001 vs. control). CF, cleaved fragment of PAPP; I, LC3-I; II, LC3-II. (C) HeLa cells were pretreated with NAC (2 or 4 mM) for 0.5 h and exposed to 4-HPR for the next 1 h. ROS levels were measured by FACS using carboxy-DCFDA. (D, E) Both the intracellular glutathione (D) and the percentage of free sulfhydryl groups in intracellular proteins (E) were determined after the cells were treated with 5 or 10 μM 4-HPR for the indicated time periods. Bars represent the mean±SD of three independent experiments. *4-HPR treatment versus control (*p<0.05; **p<0.01; and ***p<0.001); ^#10 μM 4-HPR versus 5 μM 4-HPR (^#p<0.05). NAC, N-acetyl-L-cysteine; ROS, reactive oxygen species. To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

FIG. 3.

ASK1-regulated JNK1 and p38 signaling pathways are responsible for 4-HPR-induced autophagy and apoptosis, respectively. (A) HeLa cells were treated with 5 or 10 μM 4-HPR for the indicated times, and the total cell extracts were probed with antibodies against LC-3B, PARP, JNK1/2, p-JNK1/2 (Thr183/Tyr185), p38, p-p38 (Thr180/Tyr182), and β-Actin. 1, JNK1 or p-JNK1; 2, JNK2 or p-JNK2; CF, cleaved fragment of PAPP; I, LC3-I; II, LC3-II. (B) ASK1 knockdown abolished the activation of JNK1 and p38 signaling induced by 4-HPR in HeLa cells. Cells were transfected with either 100 nM scrambled siRNA or 100 nM ASK1 siRNA for 24 h before exposure to 4-HPR for an additional 3 h. (C–F) HeLa cells transfected with siRNA against JNK1 or p38 were treated with 4-HPR for an additional 24 h and evaluated for autophagy and apoptosis. (C, E) Total cell extracts were probed with antibodies against JNK1, p38, LC-3B, PARP, and β-Actin. (D, F) Quantification of apoptosis in siRNA-transfected HeLa cells treated with 4-HPR using annexin V/PI followed by FACS analysis. All the immunoblots shown are from one representative experiment (out of three that gave similar results). β-Actin was used as the loading control. ASK1, apoptosis signal-regulating kinase 1; JNK, c-Jun N-terminal kinase; siRNA, small interfering RNA.

FIG. 4.

ASK1 forms a complex with DJ-1 in the presence of mild levels of ROS but not in the presence of excessive levels of ROS. (A, B) HeLa cells were treated with 5 or 10 μM 4-HPR for 3 h followed by co-immunoprecipitation with anti-ASK1 antibody. (A) The immunoprecipitated proteins were separated by SDS-PAGE followed by silver staining (left), and LC-MS/MS analysis was conducted to identify proteins that differently bind to ASK1 under varied oxidative states. Five unique bands were identified, and their associated information is listed in the table (right). (B) An immunoprecipitation assay was performed to validate the interaction between DJ-1 and five unique proteins identified in the LC-MS/MS analysis. The cells were treated with 5 or 10 μM 4-HPR for 3 h, and co-immunoprecipitation was subsequently performed with an anti-ASK1 antibody. Trx was used as a positive control. Total cell extracts were also probed with antibodies against ASK1, HSPA1B, HNRNPAB, RPS3A, NUDT21, DJ-1, and Trx. *Nonspecific bands. (C) A reverse immunoprecipitation assay was performed to validate the interaction between DJ-1 and ASK1. HeLa cells were treated with 5 or 10 μM 4-HPR for 3 h, and co-immunoprecipitation was subsequently performed with an anti-DJ-1 antibody. The total cell extracts were also probed with antibodies against ASK1 or DJ-1. (D) Immunofluorescence analysis was performed to validate the interaction between DJ-1 and ASK1. HeLa cells treated with 5 or 10 μM 4-HPR for 3 h were probed with antibodies against DJ-1 (red) or ASK1 (green), and their immunofluorescence was analyzed. Enlarged views of each representative image (boxed) were shown on the right. White arrows: co-localization of ASK1 and DJ-1. LC-MS/MS, liquid chromatography-tandem mass spectrometry; Trx, thioredoxin. To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

FIG. 5.

The oxidation states of DJ-1 mediate the activation of ASK1 by regulating DJ-1-ASK1 complex formation and dictate the cellular response to 4-HPR-induced ROS. (A) HeLa cells were incubated with 5 or 10 μM 4-HPR for 3 h. The oxidation of DJ-1 was analyzed using isoelectric focusing electrophoresis followed by western blotting with an anti-DJ-1 antibody. (B) HeLa cells were incubated with 5 or 10 μM 4-HPR for the indicated times and subsequently treated with 0.5 mM DSS or DMSO for 2 h, and extracted proteins were analyzed by western blotting with an anti-DJ-1 antibody. (C, D) HeLa cells were pretreated with 10 μM compound 23 for 6 h, followed by 4-HPR exposure for an additional 3 h. Total cell extracts were analyzed by co-immunoprecipitation (C) or western blotting (D) with the indicated antibodies. (E, F) HeLa cells were pretreated with 10 μM compound 23 for 6 h, followed by 4-HPR exposure for an additional 3 h. The cells were then evaluated for autophagy and apoptosis. (E) Total cell extracts were probed with antibodies against LC-3B, PARP, and β-Actin. (F) Quantification of apoptosis with annexin V/PI plus FACS analysis. (G, H) HeLa cells were transfected with the indicated artificial DJ-1 mutants, followed by treatment with 5 or 10 μM 4-HPR for an additional 24 h. Total cell extracts were utilized to detect PARP cleavage and LC3-II accumulation. All the immunoblots shown are from one representative experiment (out of three that gave similar results). β-Actin was used as the loading control. 1, JNK1 or p-JNK1; 2, JNK2 or p-JNK2; CF, cleaved fragment of PAPP; I, LC3-I; II, LC3-II; DSS, disuccinimidyl suberate.

FIG. 6.

DJ-1 depletion enhances the sensitivity of tumor cells to 4-HPR in vitro and in vivo. (A, B) DJ-1 knockdown enhanced the induction of apoptosis caused by 4-HPR (5 and 10 μM, 24 h). Cells were transfected with either 100 nM mock or DJ-1 siRNA for 24 h and subsequently exposed to 4-HPR treatment (5 and 10 μM, 6 or 24 h). The cells were analyzed by western blotting (A) or annexin V/PI staining plus FACS (B). 1, JNK1 or p-JNK1; 2, JNK2 or p-JNK2; CF, cleaved fragment of PAPP; I, LC3-I; II, LC3-II. (C–G) DJ-1 depletion enhanced the sensitivity of tumor cells to 4-HPR in vivo in a xenograft nude mouse model generated using HeLa-shRNA-con or HeLa-shRNA-DJ-1 cells. (C) The HeLa cells were transduced with lentiviral pGFP-V-RS and pGFP-V-RS-DJ-1 as described in “Methods” section, and the knockdown efficiency was validated by western blotting. (D–G) The mice transplanted with HeLa-shRNA-con or HeLa-shRNA-DJ-1 xenografts were randomly divided into three groups and given injections of serial doses of 4-HPR (20 or 40 mg/kg/day, i.v.) for a period of 13 days (n=6). The tumor volume was recorded daily, and the RTV was calculated. (D) RTV was expressed as the mean±SE. *40 mg/kg versus vehicle (*p<0.05; **p<0.01); ^#40 mg/kg versus 20 mg/kg (^#p<0.05; ^###p<0.001). (E, F) Representative images of tumors (black arrow) in the different groups are shown. (G) At the end of the experiment, the tumors were weighed, and the inhibition rate and T/C value were calculated. To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

FIG. 7.

DJ-1 depletion enhances the apoptosis-induction ability of 4-HPR by activating p38 signaling in vivo. (A) The TUNEL-positive rate of each tumor was analyzed (right panel), and representative merged images of TUNEL staining in the tumor tissues in different groups (left panel) are shown. (B) Expression of LC-3B and β-Actin in each tumor tissue from mice administered 4-HPR was detected by western blotting. I, LC3-I; II, LC3-II. (C) Tumors from each group were analyzed by immunofluorescence with antibodies against DJ-1 (green), p-p38 (red), or p-JNK1 (green). To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

FIG. 8.

Scheme for the effect of 4-HPR on pushing autophagy down to apoptosis. To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars