The p66Shc protein controls redox signaling and oxidation-dependent DNA damage in human liver cells
Abstract
The p66Shc protein mediates oxidative stress-related injury in multiple tissues. Steatohepatitis is characterized by enhanced oxidative stress-mediated cell damage. The role of p66Shc in redox signaling was investigated in human liver cells and alcoholic steatohepatitis. HepG2 cells with overexpression of wild-type or mutant p66Shc, with Ser36 replacement by Ala, were obtained through infection with recombinant adenoviruses. Reactive oxygen species and oxidation-dependent DNA damage were assessed by measuring dihydroethidium oxidation and 8-hydroxy-2′-deoxyguanosine accumulation into DNA, respectively. mRNA and protein levels of signaling intermediates were evaluated in HepG2 cells and liver biopsies from control and alcoholic steatohepatitis subjects. Exposure to H2O2 increased reactive oxygen species and phosphorylation of p66Shc on Ser36 in HepG2 cells. Overexpression of p66Shc promoted reactive oxygen species synthesis and oxidation-dependent DNA damage, which were further enhanced by H2O2. p66Shc activation also resulted in increased Erk-1/2, Akt, and FoxO3a phosphorylation. Blocking of Erk-1/2 activation inhibited p66Shc phosphorylation on Ser36. Increased p66Shc expression was associated with reduced mRNA levels of antioxidant molecules, such as NF-E2-related factor 2 and its target genes. In contrast, overexpression of the phosphorylation defective p66Shc Ala36 mutant inhibited p66Shc signaling, enhanced antioxidant genes, and suppressed reactive oxygen species and oxidation-dependent DNA damage. Increased p66Shc protein levels and Akt phosphorylation were observed in liver biopsies from alcoholic steatohepatitis compared with control subjects. In human alcoholic steatohepatitis, increased hepatocyte p66Shc protein levels may enhance susceptibility to DNA damage by oxidative stress by promoting reactive oxygen species synthesis and repressing antioxidant pathways.
aberrant production of reactive oxygen species (ROS) has been recognized as a major determinant of DNA damage, leading to disruption of tissue homeostasis, organ dysfunction, and onset of chronic degenerative disorders (24, 27, 30). p66Shc has recently emerged as a master regulator of ROS production and a critical intracellular switch conveying oxidative stress signals to DNA damage in multiple cells and tissues, including the vascular wall and heart (11), kidney (29), osteoblasts (1), lymphocytes (25), and hepatocytes (10).
Rodents with genetic deletion of p66Shc demonstrate a prolonged life span due to significant resistance to oxidative stress (3, 21, 32) p66Shc−/− mice are also protected against experimental diabetic glomerulopathy (19), diabetic cardiomyopathy, and hyperglycemia-induced endothelial dysfunction and atherogenesis (36), confirming that p66Shc mediates oxidative stress-dependent tissue damage. Furthermore, phosphorylation of p66Shc on Ser36 has been identified as the key signaling event mediating p66Shc activation and promotion of its downstream cellular effects (21). In the liver, the levels of total and Ser36-phosphorylated p66Shc protein were found to be significantly augmented in the mouse nonalcoholic steatohepatitis (NASH) model (31). Conversely, ethanol-induced oxidative stress was found to be attenuated in the liver of p66Shc−/− mice (12), suggesting that p66Shc may be involved in the hepatocyte damage in response to metabolic injuries. In addition, ablation of p66shc gene in mouse hepatocytes suppressed cellular apoptosis and ROS production after hypoxia/reoxygenation through upregulation of Mn superoxide dismutase (SOD) and redox factor-1 (13).
Normally, cells adapt to increased ROS levels by upregulating antioxidant genes (24, 30) and neutralizing ROS through the low-molecular weight antioxidant and phase II detoxifying enzymes (2, 35). The NF-E2-related factor 2 (Nrf2) is a master gene involved in the regulation of phase II and antioxidant enzymes [e.g., glutathione S-transferase alpha 5 (GSTA5), glutathione S-transferase muscle 2 (GSTM2), and MnSOD] (2, 15). Reduced expression of cardiac Nrf2 was indeed associated with significant increase in nitrosative DNA damage (5). In hepatocytes, Nrf2 was shown to be required for cell survival during liver development, its deficiency resulting in enhanced oxidative stress both in the normal and injured liver (4). While the detoxifying and ROS-scavenging role of Nrf2 has been recognized in multiple cytoprotective activities (4, 15, 35), the relationship between p66Shc and Nrf2 has not been explored.
In this study, we show that p66Shc protein expression is increased in human alcoholic steatohepatitis (ASH) and that in human liver cells p66Shc controls intracellular ROS levels, the antioxidant Nrf2 and Forkhead box protein O3a (FoxO3a) pathways, and the extent of oxidative DNA damage.
MATERIALS AND METHODS
Antibodies and reagents.
Anti-Shc monoclonal antibody was from BD Transduction Laboratories (Lexington, KY). Anti-Shc/p66 (pSer36) antibody was from Calbiochem (Darmstadt, Germany). Anti-MAP kinase (ERK-1/2) antibodies were obtained from Zymed Laboratories (San Francisco, CA). Anti-GAPDH antibody (FL-335) was from Santa Cruz Biotechnology (Santa Cruz, CA). Phospho-Akt (Ser473), total Akt, phospho-p42/p44 MAP kinase (Thr202/Tyr204), phospho-FoxO1a(Thr24)/FoxO3a(Thr32), total FoxO3a, phosphorylated Thr183/Tyr185-SAPK/JNK, total SAPK/JNK, phosphorylated Thr180/Tyr182-p38 MAPK, and total p38 MAPK antibodies were purchased from Cell Signaling Technology (Beverly, MA). The MEK inhibitor U0126 was obtained from Calbiochem (La Jolla, CA). Anti-8-oxoguanine monoclonal antibody was purchased from Millipore (MAB3560; Millipore, Billerica, MA). Alexa Fluor546 anti-rabbit antibody and the fluorescent dye dihydroethidium (DHE) were obtained from Invitrogen (Invitrogen, Carlsbad, CA). H2O2 was from Sigma Aldrich (St. Louis, MO).
Cell cultures.
HepG2 human hepatoma cells were from American Type Culture Collection (Rockville, MD) and were cultured in MEM supplemented with 10% FCS (both from GIBCO, Invitrogen, Paisley, UK), 100 U/ml penicillin, 100 mg/ml streptomycin (Lonza, Iquique, Chile), and nonessential amino acids (NEA; GIBCO, Invitrogen).
Adenoviral transfection studies.
The recombinant adenoviruses were generated by cloning either the wild-type p66Shc-encoding cDNA or the Ala36 p66Shc mutant into the shuttle vector pAdTrack-CMV, containing a green fluorescent protein epitope. Adenovirus production and cells infection were performed as previously described (18, 23).
Immunoblotting analysis.
Cell lysate preparation and immunoblotting analysis were performed as previously described (22, 23). Briefly, HepG2 cells mechanically detached in ice-cold lysis buffer, containing 50 mM HEPES pH 7.5, 150 mM NaCl, 1 mM MgCl2, 1 mM CaCl2, 4 mM EDTA, 1% Triton X-100, 10% glycerol, 50 mM NaF, and 10 mM NaPP, supplemented with 100 μM PMSF, 5 ng/ml leupeptin, 1 μg/ml aprotinin, and 2 mM Na3VO4. Cell lysates were cleared by centrifugation. Protein concentration was determined by the Bradford assay (Bio-Rad, Hercules, CA), and equal protein samples (60 μg) were separated on 7%−10% SDS-PAGE gels, as appropriate, and electrotransferred onto Hybond-P polyvinylidene difluoride filters (Amersham Life Science, Arlington Heights, IL). The filters were then probed with the specific primary antibodies, and the immunoreactive bands were visualized with horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit IgG (H+L) (Bio-Rad), as appropriate, using an ECL Plus Immunoblotting Detection System (Amersham Life Science), and quantified by densitometric analysis using the Versadoc imaging system (Bio-Rad).
Immunofluorescence analysis of FoxO3a.
To visualize FoxO3a translocation, immunofluorescence analysis was performed, as previously described (7). Briefly, HepG2 cells were grown on coverslips in complete medium for the indicated times, then fixed with 3.7% formaldehyde at room temperature for 45 min, and permeabilized at room temperature with 0.1% Triton X-100. Subsequently, coverslips were incubated with primary antibodies (1:250 dilution) in PBS containing 2% BSA (16 h at 4 C), followed by 1 h of incubation with secondary Alexa546 Fluor anti-mouse goat antibody (1:500; Molecular Probes, Eugene, OR) or Alexa488 Fluor anti-rabbit goat antibody (1:500; Molecular Probes). Coverslips were mounted on glass slides with Vectashield (Vector Laboratories, Burlingame, CA). Images were acquired on a Leica DM IRE2 confocal microscope or on Leica fluorescence microscope DM RXA2 (Leica Microsystems, Heerbrugg, Switzerland), as appropriate.
Measurement of ROS.
Intracellular ROS production was assessed through the evaluation of dihydroethidium oxidation using a Jasco FP6200 spectrofluorimeter (Jasco, Easton, MD) (7). Cells were incubated with 20 mM dihydroethidium for 0.5 h at 37°C in a serum-free medium in the dark, then washed with PBS, collected, and resuspended in assay buffer (100 mM potassium phosphate pH 7.4 and 2 mM MgCl2), using an aliquot for protein determination. The fluorescence increase (480-nm excitation and 567-nm emission wavelengths) caused by the ROS-dependent oxidation of dihydroethidium was expressed as arbitrary units normalized by cell protein content.
Gene expression analysis.
RNA was extracted using RNeasy minikit (Qiagen, Hilden, Germany), according to the manufacturer's instructions, as described previously (18). After total RNA was isolated from HepG2 cells, genomic DNA contamination was eliminated by DNase digestion (Qiagen), and cDNA was obtained using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Weiterstadt, Germany). Oligonucleotide primers used for quantitative (q)RT-PCR was as follows: human β-glucoronidase: forward CTCATTTGGAATTTTGCCGATT, reverse CCGAGTGAAGATCC; human rRNA 18s: forward CGAACGTCTGCCCTATCAACTT, reverse ACCCGTGGTCACCATGGTA; human Nrf2: forward AAACCAGTGGATCTGCCAAC, reverse GACCGGGAATATCAGGAACA; human CYP1A1: forward GCTGACTTCATCCCTATTCTTCG, reverse TTTTGTAGTGCTCCTTGACCATCT; human GSTA5: forward CATTCACCTGGTGGAACTTTTCTA, reverse CTGCCAGGCTGCAGAAACTT; human GSTM2: forward CCGATTTGAGGGCTTGGA, reverse CCATCTTTGTGAACACAGGTCTTG; human SOD2: forward GTTGGCTTGGTTTCAATAAGGAA, reverse TCCCCAGCAGTGGAATAAGG; and human catalase: forward TTCGATCTCACCAAGGTTTGG, reverse TGGATTCCGGTTTAAGACCAGTT.
The PCR reactions were carried out in an ABI PRISM 7500 System (Applied Biosystems, Weiterstadt, Germany). The PCR reactions were carried out under the following conditions: 50°C for 2 min, 95°C for 10 min, 40 cycles at 95°C for 15 s, and 60°C for 1 min. Relative gene expression levels were determined by analyzing the changes in SYBR green fluorescence during qRT-PCR using the ΔΔCt method. To confirm amplification of specific transcripts, melting curve profiles were produced at the end of each reaction. The mRNA level of each gene was normalized using β-actin as internal control.
Assessment of oxidative DNA damage.
Oxidative DNA damage in the HepG2 cells was estimated by measuring the levels of 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodG) in DNA using the method of Polytarchou et al. (26). Briefly, HepG2 cells were fixed in 4% formaldehyde, either before or 30 min after treatment with H2O2, and then stained with an anti-8-hydroxyguanine antibody. Coverslips were mounted on glass slides, with Vectashield mounting medium containing 4′,6-diamidino-2-phenylindole (DAPI; Vector Laboratories). Images were obtained using a Nikon Eclipse 80i microscope with a ×10 objective and a Spot charge-coupled-device camera. Images were quantified as red/blue ratios by using Adobe Photoshop (Adobe Systems).
Ex vivo studies in liver biopsies.
Liver biopsy specimens were obtained from patients admitted to the Liver Unit (Clinical Division “A. Murri,” Azienda Ospedaliero-Universitaria Policlinico) with clinical and analytic features of ASH, including alcohol intake >80 g/day, increased aminotransferase and γ-glutamyl transpeptidase levels, and no other identifiable cause of liver disease (9). Histologic grading was as follows: 1) degree of hepatocellular damage/ballooning (0, none; 1, mild; 2, severe) and presence of Mallory bodies, mega-mitochondria, and cholestasis (0, no; 1, yes); 2) degree of lymphocytic infiltration (0, none; 1, mild; 2, moderate; 3, severe); 3) degree of polymorphonuclear infiltration (0, none; 1, mild; 2, moderate; 3, severe); 4) degree of steatosis (G0: <10%; G1 10–33%; G2, 33–66%; G3 ≥66%); 5) degree of lobular fibrosis (0, none; 1, mild; 2, moderate; 3, severe); and 6) fibrosis stage (0, no fibrosis; 1, portal; 2, portal fibrosis and few septa; 3, septal fibrosis without cirrhosis; 4, cirrhosis) (9, 20). The protocol was approved by the Independent Ethics Committee at the Azienda Ospedaliero-Universitaria Policlinico Consorziale, Bari, Italy, and all patients gave their written informed consent.
Statistical analysis.
Data are presented as means ± SE. Normal distribution of data was assessed by the Kolmogorov-Smirnov test (P > 0.05). Statistical analysis was performed by the Student's t-test or the one-way ANOVA with Tukey's multiple comparison test, as appropriate, using Minitab 15.1. Significance was assumed at P < 0.05.
RESULTS
p66Shc is activated by oxidative stress and promotes ROS synthesis in HepG2 cells.
To understand the relationship between oxidative stress and p66Shc, p66shc phosphorylation was examined in HepG2 cells exposed to H2O2. Although the endogenous p66Shc levels were relatively low in wild-type HepG2 cells, phosphorylation of p66Shc on Ser36 could be detected in a dose-dependent manner upon exposure to H2O2 (Fig. 1A), and this was associated with increased intracellular ROS levels (Fig. 1D). To investigate the effects of increased p66Shc protein levels in liver cells, HepG2 cells with selective overexpression of p66Shc (HepG2/p66Shc) were obtained by infection with a recombinant adenovirus encoding p66Shc (Fig. 2). In the absence of H2O2, phosphorylation of p66Shc on Ser36 was found to be increased severalfold in HepG2/p66Shc compared with control HepG2/mock cells (P < 0.0001; Fig. 1, B and C), and it was further enhanced in a dose-dependent manner upon H2O2 exposure (Fig. 1B), peaking at 15 min (P < 0.001 vs. basal; Fig. 1C). ROS levels were increased approximately threefold in HepG2/p66Shc compared with control cells under basal conditions (P < 0.001 vs. wild-type HepG2 and HepG2/mock; Fig. 1D). In addition, exposure of HepG2/p66Shc cells to H2O2 led to further increase in ROS synthesis (P < 0.005 vs. wild-type HepG2 and HepG2/mock; Fig. 1D). Thus p66Shc conveys extracellular oxidative stress signals to increase ROS synthesis in liver cells.
Fig. 1.p66Shc phosphorylation on Ser36 and reactive oxygen species (ROS) synthesis in HepG2 cells. A: dose-response of p66Shc phosphorylation on Ser36 in wild-type HepG2 cells exposed to H2O2 for 15 min. Representative immunoblots of p66Shc phosphorylation (top) and protein content (bottom) are shown. B: dose response of p66Shc phosphorylation on Ser36 in HepG2/p66Shc and HepG2/mock cells exposed to H2O2 for 15 min. C: time course of p66Shc phosphorylation in HepG2/p66Shc and HepG2/mock cells exposed to 0.5 mM H2O2. Representative immunoblots of p66Shc phosphorylation on Ser36 (top) and Shc protein content (bottom; all 3 Shc protein isoforms are shown). The quantitation of phosphorylated p66Shc on Ser36 in multiple experiments is also shown (⧫: HepG2/p66Shc; ■: HepG2/mock). D: quantification of ROS levels in wild-type HepG2, HepG2/p66Shc, and HepG2/mock cells stimulated with H2O2 for 15 min (black bars) or left untreated (white bars). Results in A–D represent the means ± SE of at least n = 5 independent experiments. A.U., arbitrary units. *P < 0.05 vs. basal; #P < 0.05 vs. wild-type HepG2 and HepG2/mock.
Fig. 2.p66Shc overexpression in HepG2 cells. HepG2 cells were transduced with different doses of Ad/p66Shc, Ad/mock, and Ad/p66ShcAla36 adenoviral constructs, expressing a green fluorescence protein, at 90% confluence, as described in materials and methods. A: morphology of confluent wild-type HepG2, HepG2/p66Shc, HepG2/mock, and Ad/p66ShcAla36 cells incubated with the indicated doses of adenovirus under light microscopy (top) and corresponding fluorescent signal assessed under a fluorescent microscope (bottom). PFU, plaque-forming units. Magnification: ×10. B: representative immunoblot of all 3 Shc isoforms in HepG2/p66Shc, HepG2/mock, and Ad/p66ShcAla36 cells incubated with the indicated doses of adenovirus or left untreated. C: quantitation of p66Shc in HepG2/p66Shc incubated with the indicated doses of adenovirus or left untreated. *P < 0.05 vs. control.
p66Shc activates the Erk and Akt/FoxO3a pathways in HepG2 cells.
The activation of Erk and of Akt/FoxO3a pathways, which have been shown to be involved in p66Shc signaling and oxidative stress responses (5, 11), was assessed next. Phosphorylation levels of Erk-1/2 (Fig. 3A) and Akt (Fig. 3B) were found to be significantly increased in response to H2O2 treatment in both HepG2/p66Shc and HepG2/mock cells (P < 0.001 vs. basal; Fig. 3, A and B). However, Erk-1/2 and Akt phosphorylation showed higher levels and Erk-1/2 also an earlier 15-min peak after H2O2 challenge in HepG2/p66Shc than in control cells (P < 0.001 vs. HepG2/mock; Fig. 3, A and B). Both JNK-1/2 and p38 MAPK were also found to be activated upon H2O2 treatment in both HepG2/p66Shc and control cells; however, phosphorylation levels of these kinases were similar in HepG2/p66Shc and control cells (data not shown).
Fig. 3.Activation of Erk-1/2 and Akt/FoxO3a in HepG2/p66Shc and HepG2/mock cells. HepG2/p66Shc and HepG2/mock cells were incubated with 0.5 mM H2O2 for the indicated times or left untreated. Representative immunoblots of Erk-1/2 (A), Akt (B), and FoxO3a (C) total protein content and phosphorylation (left), and the quantitation of results from multiple experiments (right: ⧫: HepG2/p66Shc; ■: HepG2/mock). Results represent the mean ± SE of at least n = 5 independent experiments. *P < 0.05 vs. basal; #P < 0.05 vs. HepG2/mock.
Akt-mediated FoxO3a phosphorylation on Thr32 promotes both its inactivation and translocation from the nucleus to the cytoplasm (8). In both HepG2/mock and HepG2/p66Shc cells, the levels of Thr32 phosphorylation of FoxO3a were increased following challenge with H2O2 (P < 0.005 vs. basal; Fig. 3C) and augmented in p66Shc overexpressing vs. control cells (P < 0.005 in HepG2/p66Shc vs. HepG2/mock; Fig. 3C), paralleling Akt phosphorylation (Fig. 3B). The subcellular distribution of FoxO3a in the cytoplasmic and nuclear compartment was then investigated. In control HepG2/mock cells, endogenous FoxO3a could be detected almost exclusively in the nucleus in the basal state, whereas it was relocated predominantly in the cytoplasm following 15 min of H2O2 stimulation (Fig. 4A). By contrast, in HepG2/p66Shc cells, FoxO3a showed reduced nuclear staining and increased cytoplasmic localization already in the basal state, with minimal changes induced by exposure to H2O2 (Fig. 4B). Thus H2O2-mediated p66Shc activation is followed by activation of the Erk-1/2 and Akt/FoxO3a pathways and translocation of FoxO3a from the nuclear to the cytoplasmic compartment. These responses are enhanced, and already under basal conditions, when p66Shc is overexpressed.
Fig. 4.FoxO3a localization in HepG2/mock and HepG2/p66Shc cells. HepG2/mock (A) and HepG2/p66Shc (B) cells were incubated with 0.5 mM H2O2 for 15 min or left untreated. The magnified images show FoxO3a localization in representative HepG2/mock and HepG2/p66Shc cells. Adenovirus-infected cells are shown in green according to green fluorescence protein expression (green, top). FoxO3a was visualized with a rabbit polyclonal antibody followed by the addition of ALEX488 (red) labeled anti-rabbit antisera. TOPRO (blue) was used to visualize the nuclei. In control HepG2/mock cells, endogenous FoxO3a could be detected almost exclusively in the cell nucleus in the absence of H2O2 (A, white arrows in images on the left), whereas it was relocated predominantly in the cytoplasm after exiting the nucleus following H2O2 stimulation for 15 min (A, white arrows in images on the right). In HepG2/p66Shc cells, FoxO3a showed predominant cytoplasmic localization and reduced nuclear staining already in the absence of H2O2 (B, white arrows in images on the left), and this was not significantly modified by exposure to H2O2 (B, white arrows in images on the right). Images are representative of 4 independent experiments.
Role of ERK in Ser36 phosphorylation of p66Shc and Akt/FoxO3a phosphorylation.
The role of Erk-1/2 activation in p66Shc phosphorylation was investigated next by using the MEK inhibitor U0126. As expected, pretreatment with U0126 completely abrogated Erk-1/2 phosphorylation in both HepG2/mock and HepG2/p66Shc cells, both under basal conditions and after H2O2 stimulation (P < 0.0001 vs. cells not exposed to U0126; Fig. 5A and data not shown). This was associated with a significant decrease in p66Shc phosphorylation on Ser36, both in the absence and presence of H2O2 (Fig. 5B and data not shown). Similar results were obtained using PD098059, another inhibitor of MEK (data not shown). Furthermore, inhibiting the Erk-1/2 pathway with U0126 also significantly reduced the phosphorylation of Akt and FoxO3a following H2O2 challenge (P < 0.05 vs. cells not exposed to the MEK inhibitor; Fig. 5, C and D). Finally, treatment with U0126 markedly reduced H2O2-induced ROS production in control and HepG2/p66Shc cells, respectively (P < 0.01 vs. cells not exposed to U0126; Fig. 5E). Altogether, these results suggest that activation of Erk-1/2 contributes to p66Shc phosphorylation and its downstream signaling.
Fig. 5.Effects of the MEK inhibitor on p66Shc phosphorylation, Akt/FoxO3a signaling, and ROS production. Representative immunoblots (left) assessing total protein content and phosphorylation of Erk-1/2 (A), p66Shc (B), Akt (C), and FoxO3a (D) in HepG2/p66Shc cells. The quantitation of results from multiple experiments is also shown (right). HepG2/p66Shc were pretreated with 20 μM U0126 for 2 h or left untreated before exposure to 0.5 mM H2O2 for 15 min. E: quantitation of ROS levels in HepG2/mock and HepG2/p66Shc cells stimulated with 0.5 mM H2O2 for 15 min (black bars) or left untreated (white bars). Results represent the means ± SE of at least n = 5 independent experiments. *P < 0.05 vs. basal; #P < 0.05 vs. H2O2-stimulated cells treated with U0126; §P < 0.05 vs. unstimulated HepG2/mock treated with U0126; †P < 0.05 vs. HepG2/mock.
Role of p66Shc Ser36 phosphorylation in Erk and Akt/FoxO3a signaling and ROS production.
To assess whether phosphorylation of p66Shc on Ser36 is necessary for efficient signal propagation, a phosphorylation-defective p66Shc protein, in which Ser36 was mutated to Ala, was overexpressed in HepG2 cells (HepG2/p66Shc-Ala36). No significant differences in p66Shc protein levels were observed between HepG2/p66Shc and HepG2/p66Shc-Ala36 cells (Fig. 6A), while Ser36 phosphorylation of p66Shc was undetectable in HepG2/p66Shc-Ala36 cells both under basal conditions and after H2O2 stimulation (Fig. 6, A and B). Interestingly, Erk-1/2 activation was significantly decreased and Akt/FoxO3a phosphorylation was completely blunted in HepG2/p66Shc-Ala36 compared with HepG2/p66Shc cells, both in the absence and presence of H2O2 stimulation (Fig. 7, A–C; P < 0.01 vs. HepG2/p66Shc). In addition, H2O2 exposure failed to increase ROS production in HepG2/p66Shc-Ala36, differently than in control cells (Fig. 7D), indicating that Ser36 phosphorylation of p66Shc is critical for ROS production in response to oxidative stress. Thus Ser36 phosphorylation of p66Shc plays an important role in activation of the Erk and Akt/FoxO3a signaling pathways and ROS synthesis in HepG2 cells.
Fig. 6.p66Shc phosphorylation on Ser36 in HepG2 cells overexpressing the mutant p66ShcAla36 protein. A: representative immunoblots of Shc protein content (top) and p66Shc phosphorylation on Ser36 (bottom). B: densitometric analysis of 5 independent experiments (white bars: untreated cells; black bars: H2O2-stimulated cells). Wild-type HepG2, HepG2/p66Shc, HepG2/p66Shc-Ala36, and HepG2/mock cells were incubated with 0.5 mM H2O2 for 15 min or left untreated. Data represent the quantitation of at least n = 5 independent experiments. *P < 0.05 vs. basal; #P < 0.05 vs. wild-type HepG2, HepG2/mock, and HepG2/p66Shc-Ala36.
Fig. 7.Activation of Erk and Akt/FoxO3a pathways and ROS synthesis in HepG2 cells overexpressing the phosphorylation defective p66Shc Ala36 mutant. HepG2/p66Shc and HepG2/p66Shc-Ala36 cells were incubated with 0.5 mM H2O2 for the indicated times or left untreated. Representative immunoblots (left) of Erk-1/2 (A), Akt (B), and FoxO3a (C) total protein content and phosphorylation, respectively. The quantitation of results from multiple experiments is also shown (right: ⧫: HepG2/p66Shc; ■: HepG2/p66Shc-Ala36). Results represent the means ± SE of at least n = 5 independent experiments. *P < 0.05 vs. basal; #P < 0.05 vs. HepG2/p66Shc-Ala36. D: quantification of ROS levels in wild-type HepG2, HepG2/p66Shc, HepG2/p66Shc-Ala36, and HepG2/mock cells stimulated with 0.5 mM H2O2 for 15 min (black bars) or left untreated (white bars). Data represent the quantitation of at least n = 5 independent experiments. *P < 0.05 vs. basal; #P < 0.05 vs. wild-type HepG2, HepG2/mock, and HepG2/p66Shc-Ala36.
Nrf2 and Nrf2 target genes in HepG2 cells overexpressing p66Shc.
In the light of the protective role of Nrf2 against oxidative stress (17–19), whether Nrf2 and the Nrf2-induced antioxidant response element (ARE) target genes would be affected by changes in p66Shc protein levels in HepG2 cells was examined next. Interestingly, Nrf2 mRNA levels were significantly lower in HepG2/p66Shc than in control HepG2/mock cells in the absence of H2O2 (P < 0.05 vs. basal control cells; Fig. 8) and remained significantly reduced by >50% at multiple times following induction of oxidative stress with H2O2 up to 240 min (P < 0.05 vs. control cells at 15–60 min; Fig. 8). Changes in Nrf2 mRNA levels were paralleled by similar changes in gene expression of major Nrf2 target genes. Indeed, mRNA levels of cytochrome P450 (CYP)1A1, GSTM2, and GSTA5 were significantly reduced in HepG2/p66Shc compared with HepG2/mock cells both under basal conditions and at multiple time-points following exposure to H2O2 (P < 0.05 vs. HepG2/mock; Fig. 8). Furthermore, Nrf2 and Nrf2-induced ARE genes showed markedly reduced mRNA levels in HepG2/p66Shc compared with HepG2/p66Shc-Ala36 cells (P < 0.05 vs. HepG2/p66Shc-Ala36; Fig. 8). Finally, HepG2/p66Shc-Ala36 cells exhibited higher mRNA levels of CYP1A1, GSTM2, GSTA5, and SOD2 than control HepG2/mock cells (P < 0.05 vs. HepG2/mock; Fig. 8), consistent with the mutant p66Shc Ala36 variant acting as a dominant-negative protein.
Fig. 8.mRNA expression levels of NF-E2-related factor 2 (Nrf2) and Nrf2-target genes in HepG2/mock, HepG2/p66Shc, and HepG2/p66Shc-Ala36 cells. Nrf2, CYP1A1, GSTM2, GSTA5, and SOD2 mRNA expression levels were measured by quantitative RT-PCR in HepG2/mock (grey squares), HepG2/p66Shc (black squares), and HepG2/p66Shc-Ala36 (white triangles) following exposure to 0.5 mM H2O2 for the indicated times. Data represent the means ± SE of at least n = 5 independent experiments. *P < 0.05 HepG2/p66Shc vs. HepG2/mock; #P < 0.05 HepG2/p66Shc vs. HepG2/p66Shc-Ala36; §P < 0.05 HepG2/p66Shc-Ala36 vs. HepG2/mock.
Overexpression of p66Shc promotes oxidative DNA damage.
8-OxodG is a sensitive marker of ROS-induced DNA damage (27). To investigate the possibility that forced expression/activation of p66Shc may lead to DNA damage via increased ROS synthesis, the extent of 8-oxodG accumulation was assessed in individual cells infected with the distinct adenoviral constructs by immunofluorescence. Expression of green fluorescent protein by the adenovirus allowed identification of infected cells. With this method, 8-oxodG staining was found to be increased in HepG2/p66Shc compared with HepG2/mock and HepG2/p66Shc-Ala36 cells in the absence of H2O2 (Fig. 9, A–C). Treatment of HepG2/p66Shc with H2O2 for 60 min further increased 8-oxodG accumulation at 15 min (data not shown) and 30 min (Fig. 9B). By contrast, 8-oxodG staining was almost undetectable in H2O2-treated HepG2/p66Shc-Ala36 compared with both HepG2/p66Shc and HepG2/mock cells (Fig. 9, A–C). Overall, these data demonstrate that p66Shc promotes ROS-induced DNA damage in liver cells and that this is inhibited by the phosphorylation-defective p66Shc protein.
Fig. 9.8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodG) accumulation following p66Shc overexpression in HepG2 cells. HepG2/mock (A), HepG2/p66Shc (B), and HepG2/p66Shc-Ala36 (C) cells were treated with 0.5 mM H2O2 or left untreated and then analyzed 30 min later by immunofluorescence, evaluating accumulation of 8-oxodG, a sensitive marker of DNA damage. For each condition, from left to right, the 1st column shows DAPI nuclei staining, the 2nd column shows the green fluorescence protein (GFP) in infected cells, the 3rd column shows 8-oxodG staining, and the 4th column displays the merged staining (A–C). Bar = 50 μm. Results are representative of n = 4 independent experiments.
p66Shc in human liver biopsies from subjects with ASH.
Finally, protein expression and phosphorylation levels of p66Shc were examined in liver biopsies of subjects with ASH. The severity of liver injury was graded according to the presence and extent of steatosis, fibrosis, and lymphocytic/polymorphonuclear infiltration (9, 20), ranging from G0/F0 to G3/F4. The protein levels of p66Shc were found to be increased in the human liver biopsies in parallel with the severity of liver injury (Fig. 10A), in the absence of changes in p52Shc and p46Shc protein abundance (Fig. 10A). Furthermore, p66Shc expression was found to be significantly increased in the group with higher extent of steatosis (>33%) and presence of fibrosis and lymphocytic/polymorphonuclear infiltration (P < 0.05, G2-3/F1-4 vs. G0-1/F0; Fig. 10A). Furthermore, increased p66Shc protein levels were associated with significant augmentation of Akt phosphorylation, in the absence of significant changes in Akt protein levels (P < 0.05, G2-3/F1-4 vs. G0-1/F0; Fig. 10B). Other signaling reactions, including p66Shc phosphorylation on Ser36 and FoxO3a phosphorylation on Thr32, could not be assessed since they were below the sensitivity of the immunoblotting technique (data not shown).
Fig. 10.Protein levels of p66Shc in liver from ASH subjects. A: protein levels of Shc isoforms in human liver biopsies. Representative immunoblots of Shc protein isoforms in liver biopsies from individual subjects with various degrees of steatosis, fibrosis, and lymphocytic/polymorphonuclear infiltration are shown (top). GAPDH protein content was used as loading control. The quantitation of p66Shc vs. p52-p46Shc ratio in liver samples considered individually (bottom left) and grouped as G0-1/F0 (steatosis <33% and absence of fibrosis and lymphocytic/polymorphonuclear infiltration) and as G2-3/F1-4 (steatosis >33% and presence of fibrosis and lymphocytic/polymorphonuclear infiltration; bottom right) is also shown. B: Akt in human liver biopsies. Representative immunoblots of Akt phosphorylation and total protein content in liver biopsies from individual subjects with various degrees of steatosis, fibrosis and lymphocytic/polymorphonuclear infiltration are shown (top). The quantitation of phospho-Akt vs. total Akt liver biopsies considered individually (bottom left) and grouped as G0-1/F0 and G2-3/F1-4 (bottom right) is also shown. *P < 0.05 vs. G0-1/F0.
DISCUSSION
The adapter protein p66Shc has been shown to mediate oxidative stress-related injury in multiple cell types and under a variety of pathophysiological conditions, including obesity, diabetes, and steatohepatitis (1, 10–13, 19, 21, 24, 25, 27–32, 36). Specifically, p66Shc has been involved in hepatocyte lipid accumulation and cytotoxicity in both experimental ethanol intoxication and NASH (17, 31). The results from this study demonstrate for the first time that p66Shc is both a target and an enhancer of oxidative stress in liver cells and that p66Shc protein levels are increased in the liver of individuals with ASH. In addition, forced expression of p66Shc promotes ROS synthesis and reduces expression of Nrf2 and Nrf2-induced ARE genes, increasing incorporation of 8-oxodG into cellular DNA, a marker of oxidative stress-related DNA damage. All of these effects require phosphorylation of p66Shc on Ser36.
In control HepG2 cells the levels of p66Shc were relatively low, yet phosphorylation of the protein could be induced by exposure to H2O2. However, when p66Shc was overexpressed by adenoviral-mediated gene transfer its phosphorylation on Ser36 was increased (Fig. 1C) and this was associated with increased phosphorylation of Erk-1/2 and Akt/FoxO3a (Fig. 3, A–C) and ROS synthesis (Fig. 1D).
While overexpression of p66Shc was associated with increased Erk-1/2 phosphorylation, which was reduced in HepG2 overexpressing a defective p66Shc-Ala36 (Fig. 7A), inhibition of the Erk pathway using various MEK inhibitors significantly reduced Ser36 phosphorylation of p66Shc both in the absence and presence of H2O2 stimulation. Altogether, these results suggest that p66Shc and Erk-1/2 are involved in a reciprocal “regulatory loop”: increased p66Shc expression/signaling results in Erk-1/2 activation that in turn promotes Ser36-phosphorylation of p66Shc (Fig. 11). These data are in line with previous observations in breast and prostate cancer cells, in which increased p66Shc protein levels were positively correlated with Erk-1/2 phosphorylation (16, 34), and p66Shc knockdown led to reduced Erk-1/2 activation (34), but not with other studies showing inhibition of Erk signaling by p66Shc (32). Conversely, Erk-dependent phosphorylation of p66Shc on Ser36, which is then responsible for p66Shc-dependent phosphorylation of FoxO3a, as observed in this study (Fig. 5), was previously shown in mouse embryo fibroblasts (32). However, the phosphorylation levels of p66Shc on Ser36 were not completely abolished when Erk-1/2 was fully inhibited (Fig. 5), suggesting that additional protein kinases other than Erk-1/2 may be involved in Ser36-phosphorylation of p66Shc. Indeed, depending on the cell type and stimulus (e.g., H2O2, TGF-β, various cytokines), protein kinase C (PKC)-β and -δ and β1Pix (Pak-interacting exchange factor) have also been shown to be involved in p66Shc phosphorylation on Ser36 (6, 22, 31). Furthermore, we have recently reported that TNF-α promotes p66Shc phosphorylation on Ser36 via the stress-kinase JNK in human endothelial cells (23). However, both JNK-1/2 and p38 MAPK activities were not affected by p66Shc overexpression in HepG2 cells (data not shown), suggesting that they are not part of the same regulatory loop as Erk-1/2.
Fig. 11.Hypothetical model of p66Shc-dependent redox signaling in human liver cells. P, phosphorylation.
The involvement of the Nrf2 signaling pathway in eliciting cell survival and resistance to oxidative stress has been recently reported (2, 5, 35). Nrf2 plays a central role in cytoprotection, by detoxifying and eliminating ROS, xenobiotics and electrophilic carcinogens, as well as by removing damaged proteins and organelles (35). Compared with control cells, Nrf2 knockout cardiomyocytes showed significantly higher ROS levels under basal conditions, which were further enhanced upon exposure to high glucose concentrations (14). Similarly, in primary mouse hepatocytes Nrf2 gene ablation resulted in enhanced oxidative stress, impaired activation of the MAPK pathway, and reduced mRNA expression of ROS-detoxifying enzymes (4). In line with these findings, our results show for the first time a link between the redox protein p66Shc and the Nrf2 pathway. Liver cells overexpressing p66Shc showed reduced mRNA levels of Nrf2 and of its downstream detoxifying target genes, such as CYP1A1, GSTM2, and GSTA5, in association with enhanced ROS synthesis and increased oxidative DNA damage. Conversely, cells overexpressing the phosphorylation-defective p66Shc mutant displayed augmentation of gene expression of Nrf2 and its downstream target genes, reduced ROS levels, and minimal 8-oxodG accumulation. Thus, p66Shc appears to foster cellular oxidative stress responses by suppressing Nrf2 expression (Fig. 11). The Akt/FoxO3a signaling pathway is also involved in preventing accumulation of ROS and consequent cell damage by upregulating antioxidant enzymes (33). Since forced activation of p66Shc resulted in increased FoxO3a phosphorylation and nuclear exclusion (Figs. 3 and 4), and this was not observed following overexpression of p66ShcAla36 (Fig. 7), enhanced p66Shc signaling may potentially promote oxidative DNA damage by both repressing Nrf2 and inactivating FoxO3a (Fig. 11).
Significant elevations of p66Shc mRNA and protein levels were recently reported in liver biopsies from individuals with nonalcoholic fatty liver disease (NAFLD) and NASH, compared with normal liver samples (31). In line with these findings, we found increased p66Shc protein levels in liver biopsies of subjects with ASH, a well-characterized condition of oxidative stress-induced cellular damage (10, 29). Moreover, p66Shc protein abundance correlated with the degree of histological abnormalities and disease severity, being increased to a greater extent in subjects with higher degree of fibrosis and steatosis (Fig. 10), in line with the results in NAFLD/NASH (14). Increased protein expression of p66Shc in liver biopsies with more severe grading was associated with augmented Akt phosphorylation, suggesting the functional relevance of these findings (Fig. 10). Inhibition of hepatic p66Shc signaling may thus represent an attractive strategy to counteract progression of hepatocyte damage in both alcoholic and nonalcoholic fatty liver disease.
In conclusion, overexpression of p66Shc in human hepatocytes promotes ROS accumulation and increases susceptibility to H2O2-induced oxidative stress, leading to reduced levels of cytoprotective genes and consequently increased DNA damage. Modulation of the redox homeostasis by limiting p66Shc expression and/or activity in human hepatocytes may open novel therapeutic approaches for oxidative stress-associated liver diseases.
GRANTS
This work was supported by
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the author(s).
AUTHOR CONTRIBUTIONS
Author contributions: S. Perrini, L.L., and F.G. conception and design of research; S. Perrini, F.T., A.N., A.C., A.L., R.F., T.M.C., and F.G. analyzed data; S. Perrini, F.T., A.N., A.L., T.M.C., L.L., G.P., and F.G. interpreted results of experiments; S. Perrini, F.T., and S. Porro prepared figures; S. Perrini drafted manuscript; S. Perrini, F.T., A.N., C.P., A.C., V.O.P., C.C., F.D.S., S. Porro, A.L., R.F., M.D.F., T.M.C., F.P., L.L., G.P., and F.G. approved final version of manuscript; F.T., C.P., V.O.P., C.C., F.D.S., S. Porro, R.F., M.D.F., and F.P. performed experiments; A.N., A.L., L.L., G.P., and F.G. edited and revised manuscript.
ACKNOWLEDGMENTS
We thank Dr. Alessandro Peschechera for technical assistance and Dr. Antonia Gentile for performing the histologic grading of liver biopsies.
REFERENCES
- 1. . Oxidative stress stimulates apoptosis and activates NF-kappaB in osteoblastic cells via a PKCbeta/p66shc signaling cascade: counter regulation by estrogens or androgens. Mol Endocrinol 24: 2030–2037, 2010.
Crossref | PubMed | Google Scholar - 2. . Nrf2: a potential target for new therapeutics in liver disease. Clin Pharmacol Ther 92: 340–348, 2012.
Crossref | PubMed | ISI | Google Scholar - 3. . Deletion of p66Shc in mice increases the frequency of size-change mutations in the lacZ transgene. Aging Cell 12: 177–183, 2013.
Crossref | PubMed | ISI | Google Scholar - 4. . Impaired liver regeneration in Nrf2 knockout mice: role of ROS-mediated insulin/IGF-1 resistance. EMBO J 27: 212–223, 2008.
Crossref | PubMed | ISI | Google Scholar - 5. . The heme oxygenase 1 inducer (CoPP) protects human cardiac stem cells against apoptosis through activation of the extracellular signal-regulated kinase (ERK)/NRF2 signaling pathway and cytokine release. J Biol Chem 287: 33720–33732, 2012.
Crossref | PubMed | ISI | Google Scholar - 6. . Endothelin-1 induces p66Shc activation through EGF receptor transactivation: Role of beta(1)Pix/Galpha(i3) interaction. Cell Signal 22: 325–329, 2010.
Crossref | PubMed | ISI | Google Scholar - 7. . Role of UBC9 in the regulation of the adipogenic program in 3T3–L1 adipocytes. Endocrinology 151: 5255–5266, 2010.
Crossref | PubMed | ISI | Google Scholar - 8. . Regulation of the intracellular localization of Foxo3a by stress-activated protein kinase signaling pathways in skeletal muscle cells. Mol Cell Biol 30: 470–480, 2010.
Crossref | PubMed | ISI | Google Scholar - 9. . Hepatic expression of candidate genes in patients with alcoholic hepatitis: correlation with disease severity. Gastroenterology 132: 687–697, 2007.
Crossref | PubMed | ISI | Google Scholar - 10. . Electron transfer between cytochrome c and p66Shc generates reactive oxygen species that trigger mitochondrial apoptosis. Cell 122: 221–233, 2005.
Crossref | PubMed | ISI | Google Scholar - 11. . p66Shc links alpha1-adrenergic receptors to a reactive oxygen species-dependent AKT-FOXO3A phosphorylation pathway in cardiomyocytes. Circ Res 104: 660–669, 2009.
Crossref | PubMed | ISI | Google Scholar - 12. . p66(Shc) has a pivotal function in impaired liver regeneration in aged mice by a redox-dependent mechanism. Lab Invest 90: 1718–1726, 2010.
Crossref | PubMed | ISI | Google Scholar - 13. . Preventing hypoxia/reoxygenation damage to hepatocytes by p66(shc) ablation: up-regulation of anti-oxidant and anti-apoptotic proteins. J Hepatol 48: 422–432, 2008.
Crossref | PubMed | ISI | Google Scholar - 14. . Nrf2 is critical in defense against high glucose-induced oxidative damage in cardiomyocytes. J Mol Cell Cardiol 46: 47–58, 2009.
Crossref | PubMed | ISI | Google Scholar - 15. . Nrf2-dependent upregulation of antioxidative enzymes: a novel pathway for hypoxic preconditioning-mediated delayed cardioprotection. Mol Cell Biochem 385: 33–41, 2014.
Crossref | PubMed | ISI | Google Scholar - 16. . Elevated levels of p66 Shc are found in breast cancer cell lines and primary tumors with high metastatic potential. Clin Cancer Res 6: 1135–1139, 2000.
PubMed | ISI | Google Scholar - 17. . Role of the life span determinant P66(shcA) in ethanol-induced liver damage. Lab Invest 88: 750–760, 2008.
Crossref | PubMed | ISI | Google Scholar - 18. . TNFα signals via p66(Shc) to induce E-selectin, promote leukocyte transmigration and enhance permeability in human endothelial cells. PLoS One 8: e81930, 2013.
Crossref | PubMed | ISI | Google Scholar - 19. . Deletion of p66Shc longevity gene protects against experimental diabetic glomerulopathy by preventing diabetes-induced oxidative stress. Diabetes 55: 1642–1650, 2006.
Crossref | PubMed | ISI | Google Scholar - 20. . Respective roles of porto-septal fibrosis and centrilobular fibrosis in alcoholic liver disease. J Pathol 201: 55–62, 2003.
Crossref | PubMed | ISI | Google Scholar - 21. . The p66shc adaptor protein controls oxidative stress response and life span in mammals. Nature 402: 309–313, 1999.
Crossref | PubMed | ISI | Google Scholar - 22. . Epidermal growth factor receptor phosphorylates protein kinase C-delta at Tyr332 to form a trimeric complex with p66Shc in the H2O2-stimulated cells. J Biochem 143: 31–38, 2008.
Crossref | PubMed | ISI | Google Scholar - 23. . Involvement of the p66Shc protein in glucose transport regulation in skeletal muscle myoblasts. Am J Physiol Endocrinol Metab 296: E228–E237, 2009.
Link | ISI | Google Scholar - 24. . Mitochondrial injury, oxidative stress, and antioxidant gene expression are induced by hepatitis C virus core protein. Gastroenterology 122: 366–375, 2002.
Crossref | PubMed | ISI | Google Scholar - 25. . Diabetes induces p66shc gene expression in human peripheral blood mononuclear cells: relationship to oxidative stress. J Clin Endocrinol Metab 90: 1130–1136, 2005.
Crossref | PubMed | ISI | Google Scholar - 26. . The JmjC domain histone demethylase Ndy1 regulates redox homeostasis and protects cells from oxidative stress. Mol Cell Biol 28: 7451–7464, 2008.
Crossref | PubMed | ISI | Google Scholar - 27. . Nonalcoholic steatohepatitis. Gastroenterology 121: 710–723, 2001.
Crossref | PubMed | ISI | Google Scholar - 28. . Diabetes promotes cardiac stem cell aging and heart failure, which are prevented by deletion of the p66shc gene. Circ Res 99: 42–52, 2006.
Crossref | PubMed | ISI | Google Scholar - 29. . p66Shc mediates high-glucose and angiotensin II-induced oxidative stress renal tubular injury via mitochondrial-dependent apoptotic pathway. Am J Physiol Renal Physiol 299: F1014–F1025, 2010.
Link | ISI | Google Scholar - 30. . Reactive oxygen species in cell signaling. Am J Physiol Lung Cell Mol Physiol 279: L1005–L1028, 2000.
Link | ISI | Google Scholar - 31. . p53/p66Shc-mediated signaling contributes to the progression of non-alcoholic steatohepatitis in humans and mice. J Hepatol 57: 837–843, 2012.
Crossref | PubMed | ISI | Google Scholar - 32. . A p53-p66Shc signalling pathway controls intracellular redox status, levels of oxidation-damaged DNA and oxidative stress-induced apoptosis. Oncogene 21: 3872–3878, 2002.
Crossref | PubMed | ISI | Google Scholar - 33. . Hydrogen peroxide sensing and signaling. Mol Cell 26: 1–14, 2007.
Crossref | PubMed | ISI | Google Scholar - 34. . Expression of p66(Shc) protein correlates with proliferation of human prostate cancer cells. Oncogene 24: 7203–7212, 2005.
Crossref | PubMed | ISI | Google Scholar - 35. . Nuclear factor (erythroid-derived 2)-like 2 activation-induced hepatic very-low-density lipoprotein receptor overexpression in response to oxidative stress contributes to alcoholic liver disease in mice. Hepatology 59: 1381–1392, 2014.
Crossref | PubMed | ISI | Google Scholar - 36. . Repression of P66Shc expression by SIRT1 contributes to the prevention of hyperglycemia-induced endothelial dysfunction. Circ Res 109: 639–648, 2011.
Crossref | PubMed | ISI | Google Scholar
