Search tips
Search criteria 


Logo of ccLink to Publisher's site
Cell Cycle. 2012 May 15; 11(10): 1918–1928.
PMCID: PMC3359121

The DNA damage checkpoint protein ATM promotes hepatocellular apoptosis and fibrosis in a mouse model of non-alcoholic fatty liver disease


Steatoapoptosis is a hallmark of non-alcoholic fatty liver disease (NAFLD) and is an important factor in liver disease progression. We hypothesized that increased reactive oxygen species resulting from excess dietary fat contribute to liver disease by causing DNA damage and apoptotic cell death, and tested this by investigating the effects of feeding mice high fat or standard diets for 8 weeks. High fat diet feeding resulted in increased hepatic H2O2, superoxide production, and expression of oxidative stress response genes, confirming that the high fat diet induced hepatic oxidative stress. High fat diet feeding also increased hepatic steatosis, hepatitis and DNA damage as exemplified by an increase in the percentage of 8-hydroxyguanosine (8-OHG) positive hepatocytes in high fat diet fed mice. Consistent with reports that the DNA damage checkpoint kinase Ataxia Telangiectasia Mutated (ATM) is activated by oxidative stress, ATM phosphorylation was induced in the livers of wild type mice following high fat diet feeding. We therefore examined the effects of high fat diet feeding in Atm-deficient mice. The prevalence of apoptosis and expression of the pro-apoptotic factor PUMA were significantly reduced in Atm-deficient mice fed the high fat diet when compared with wild type controls. Furthermore, high fat diet fed Atm−/− mice had significantly less hepatic fibrosis than Atm+/+ or Atm+/− mice fed the same diet. Together, these data demonstrate a prominent role for the ATM pathway in the response to hepatic fat accumulation and link ATM activation to fatty liver-induced steatoapoptosis and fibrosis, key features of NAFLD progression.

Keywords: ATM, hepatic fibrosis, nonalcoholic fatty liver disease, oxidative stress, steatoapoptosis


Nonalcoholic fatty liver disease (NAFLD) is an emerging health epidemic with a poorly understood pathogenesis.1 NAFLD comprises a spectrum of hepatic lesions ranging from steatosis with deposition of triglycerides within hepatocytes to the more aggressive nonalcoholic steatohepatitis (NASH), which includes hepatitis and fibrosis.1,2 Fatty liver disease is highly prevalent in the US, with an estimated 30% of the population possessing NAFLD and 6% suffering from NASH. These numbers rise to 90% and 30% respectively when obese populations are examined.3 NASH-induced complications are severe, and approximately one-third of NASH patients develop cirrhosis and may progress to hepatocellular carcinoma (HCC) or liver failure.1,2,4 There is a clear link between dietary fat consumption and fatty liver disease; however, the factors that promote the progression of NAFLD to NASH and its associated complications currently are not well understood.2,5-7

During NAFLD/NASH, an excessive amount of nonesterified free fatty acids (FFAs) overwhelm the ability of the liver to convert FFAs to triglycerides, resulting in apoptotic cell death, a process known as steato- or lipoapoptosis.8,9 In both humans and animal models, hepatocyte apoptosis correlates with NASH severity and fibrosis stage, suggesting that steatoapoptosis plays a causative role in disease progression.9-13 The exact mechanisms and cellular pathways linking hepatic lipid accumulation to steatoapoptosis and the progression of NAFLD/NASH remain largely unknown. However, oxidative stress may have a central role, as oxidative metabolism is increased in NAFLD patients,14 and recent studies indicate that saturated FFAs induce apoptosis through the generation of reactive oxygen species (ROS).13

Eukaryotic cells protect genomic integrity in part through checkpoint mechanisms, including a pathway centered on the protein kinase ataxia telangiectasia mutated (ATM), which is activated by double-strand DNA breaks (DSB). Once activated, ATM phosphorylates intermediary protein substrates, including CHK2, p53 and H2AX, leading to cell cycle arrest and DNA repair or apoptosis.15 Failure to activate ATM after DNA damage leads to defective cell cycle control and impaired DNA repair. In humans, ATM deficiency results in the genomic instability syndrome ataxia telangiectasia (A-T). A-T patients cannot respond to DSB properly, resulting in a variety of deleterious phenotypes, including cerebellar ataxia, immunodeficiency, premature aging, sterility, chromosomal instability, shortened lifespan and cancer predisposition.15 ATM is also implicated in metabolic disease, characterized by a variety of conditions, including insulin resistance, glucose intolerance, elevated serum cholesterol and lipid levels and atherosclerosis.16,17 Importantly, ATM can be directly activated by ROS independent of DSBs, and ROS-induced ATM signaling is associated with phosphorylation of only a subset of ATM targets, including p53 and CHK2 but not H2AX.17,18 Further, cultured Atm-deficient cells are more sensitive to oxidative stress.19 Hepatocytes with excess fat are exceptionally vulnerable to oxidative stress and DNA damage,4,14,20 but the potential role of ATM in this process is unknown.

Due to the increasing prevalence of NAFLD/NASH and the clear connection to complications including HCC, the overall objective of this study was to establish and characterize an inducible animal model capable of mechanistically evaluating the interaction between hepatic fat accumulation, DNA damage, oxidative stress and NAFLD/NASH disease progression. Our results demonstrate that lipid-laden hepatocytes experience increased oxidative stress and undergo ATM-dependent steatoapoptosis, which occurs independently of detectable H2AX phosphorylation. These findings provide strong evidence that an ATM-mediated DNA damage response contributes to NAFLD progression by promoting apoptosis and fibrosis in the liver.


Hepatic oxidative stress is promoted by high-fat diet feeding.

Feeding mice a high-fat diet with cholesterol and cholic acid leads to increased expression of oxidative stress-related genes.21 We directly measured oxidative stress associated with excessive fat consumption by harvesting hepatocytes from mice fed either a standard diet (SD) or a high-fat diet (HFD) for 8 weeks and performing Amplex Red staining to detect H2O2 and dihydrorhodamine (DHR) staining to detect superoxide species. Hepatocytes from mice fed the HFD (n = 4) showed significantly elevated levels of H2O2 compared with SD-fed mice (n = 3) (Fig. 1A; *p < 0.05). Superoxide levels, as measured by DHR staining, were also significantly increased in hepatocytes from HFD-fed mice as compared with those from SD-fed animals (Fig. 1B and C; *p = 0.01). Quantitative PCR analysis of the NADPH oxidase p22phox gene, which plays a central role in superoxide production,21 additionally was performed. Consistent with the H2O2 and superoxide quantification, there were significant increases in p22phox expression in livers from HFD-fed mice compared with those from mice fed SD (Fig. 1D; *p = 0.002). These data indicate that HFD feeding causes increased oxidative stress in hepatocytes.

figure cc-11-1918-g1
Figure 1. High fat diet feeding induces hepatic oxidative stress. Mice were fed SD or HFD for 8 weeks. Hepatocytes were then harvested and analyzed for (A) hydrogen peroxide production (Amplex-Red assay) or (B, C) the presence of superoxide (dihydrorhodamine ...

High-fat diet feeding induces hepatomegaly, steatohepatitis and ATM activation and promotes hepatic fibrosis in an Atm-dependent manner.

ATM activation plays an important role in the cellular response to ROS, and A-T patients display increased sensitivity to oxidative stress.18,19,22 We therefore assessed ATM pathway activation in mice following HFD feeding via immunofluorescent staining of liver sections using an antibody that detects phosphorylated, active ATM (pS1987 of mouse ATM, corresponding to pS1981 of human ATM).23 Hepatocytes from Atm+ (Atm+/+ and Atm+/-) HFD-fed mice displayed both nuclear and cytoplasmic foci containing p-ATM, while hepatocytes from Atm+ SD-fed mice did not, confirming hepatic ATM activation in the Atm+ HFD-fed group (Fig. 2A). The majority of the foci appeared to be localized to the periphery of the nucleus. The detection of HFD-induced ATM phosphorylation was specific, as no staining was observed in Atm-/- hepatocytes from mice fed either diet. We next tested the ability of Atm-deficient mice to respond to HFD feeding. Livers were examined grossly and microscopically after 8 weeks of feeding SD or HFD to determine if HFD feeding induced hepatic changes consistent with NAFLD. For all measures described below, we observed no significant differences between Atm+/+ and Atm+/- mice and, therefore, in most cases present combined data for these genotypes under the label Atm+. Irrespective of Atm status, there was 100% prevalence of fatty liver development in mice fed the HFD (n = 22), whereas all mice fed the SD (n = 23) had grossly normal livers (Fig. 2B). Mice fed the HFD also had marked increases in liver weight (Fig. 2C; *p < 0.0001) but not in total body weight (Fig. 2C and D) compared with mice fed the SD, indicating that the HFD induced hepatic fat accumulation independent of obesity. Consistent with previous reports in references 2426, Atm-/- mice had slightly lower body weights compared with Atm+ mice regardless of diet (Fig. 2D). Upon both H&E and Oil-red-O staining, hepatocytes from HFD-fed Atm+/+ and Atm-/- mice displayed marked cellular hypertrophy and both microvesicular and macrovesicular steatosis (Fig. 3A). In contrast, there was no abnormal hepatic fat accumulation in mice fed the SD. Further, mice fed the HFD had significantly higher fatty liver scores compared with SD-fed controls irrespective of Atm genotype (Fig. 3B,*p < 0.0001). These data demonstrate that 8 weeks of HFD feeding produced marked hepatic steatosis consistent with NAFLD in both Atm-expressing and Atm-deficient mice.

figure cc-11-1918-g2
Figure 2. High fat diet feeding induces grossly evident fatty liver without fat accumulation in other tissues. Atm+ and Atm−/− mice were fed SD or HFD for 8 weeks. Liver weight, body weight, and liver appearance were evaluated ...
figure cc-11-1918-g3
Figure 3. High fat diet feeding induces hepatic steatosis and hepatitis regardless of Atm status, but leads to more fibrosis in Atm+ livers. H&E (A, first row, scale bars represent 20 μm), Sirius red (A, second row, scale bars ...

Additional hepatic phenotypes resulting from 8 weeks of HFD feeding were evaluated based on H&E staining and pico-sirius red staining for collagen. Livers from both Atm+ and Atm-/- mice fed the HFD demonstrated moderate, predominantly peri-portal, mixed inflammatory cell infiltrates (Fig. 3A). In contrast, there were no hepatic inflammatory lesions in mice fed the SD. Further, all mice fed the HFD, regardless of genotype, had significantly higher inflammation (Fig. 3C, *p < 0.05) and fibrosis (Fig. 3D, **p < 0.05) scores when compared with mice fed the SD. Notably, Atm+ HFD-fed mice developed significantly more fibrosis than Atm-/- mice (Fig. 3D, ***p < 0.05). Differences in fibrosis development were highlighted by fewer and smaller pico-sirius red-positive foci in the livers of Atm-/- mice as compared with those of Atm+ mice (Fig. 3A). These data demonstrate that 8 weeks of HFD feeding produces marked steatohepatitis regardless of Atm status, and that subsequent fibrosis is dependent, at least in part, on ATM.

High-fat diet feeding induces oxidative nucleic acid damage but limited histone H2AX phosphorylation.

To understand the basis for the effects of Atm deficiency on HFD-induced liver pathology, we evaluated whether Atm loss affects ROS levels in HFD-fed mice. To a similar extent as was observed in Atm+ mice (Fig. 1), HFD feeding of Atm-/- mice caused increased H2O2 and superoxide accumulation relative to SD-fed mice (Fig. S1). Levels of 8-hydroxyguanosine (8-OHG), an oxidized nucleotide and marker of oxidative nucleic acid damage were then quantified to determine if the elevated oxidative stress in fatty livers resulted in hepatocyte DNA damage. There was minimal staining for 8-OHG in the SD group, while there were multiple 8-OHG-positive aggregates within the nuclei and cytoplasm of hepatocytes from HFD-fed mice (Fig. 4A). In some nuclei, 8-OHG appeared to be localized to the periphery of the nucleus. Quantification of 8-OHG-positive cells showed that there were significantly more positive cells in the Atm+ and Atm-/- groups fed the HFD than in mice fed the SD (Fig. 4B; *p = 0.001). There was no significant difference between Atm+ and Atm-/- HFD-fed groups. These findings indicate that the HFD causes oxidative nucleic acid damage in both Atm-expressing and Atm-deficient mice.

figure cc-11-1918-g4
Figure 4. High fat diet feeding increases hepatic nucleic acid damage but does not induce detectable phosphorylation of histone H2AX. (A) Frozen sections of livers from mice fed SD (left) or the HFD (right) were labeled with anti- 8-hydroxyguanosine ...

To determine the nature and extent of the DNA damage response induced by HFD feeding, we performed immunohistochemistry on liver sections from HFD- and SD-fed groups and quantified the levels of ©H2AX, the phosphorylated form of histone H2AX which occurs in response to DNA DSBs.27 In contrast to our earlier observations of HFD feeding-induced ATM phosphorylation (Fig. 2A), there was minimal positive staining for H2AX phosphorylation in hepatic nuclei from mice of all experimental groups (Fig. 4C; Fig. S2; *p < 0.0001). Importantly, significant increases in ©H2AX staining were detected in liver sections from control mice treated with ionizing radiation. Overall, these data demonstrate that there is DNA damage in HFD fed hepatocytes without substantial phosphorylation of H2AX.

High-fat diet feeding increases Atm-dependent steatoapoptosis.

One of the outcomes of ATM pathway activation is apoptosis,15 and steatoapoptosis is important in fatty liver progression.9,12,13 The induction of apoptosis by HFD feeding was evaluated by terminal uridine deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining of liver sections (n = 5–10/group). TUNEL staining detects DNA fragmentation, a characteristic feature of apoptotic cells. HFD groups contained significantly more TUNEL-positive hepatocytes compared with SD groups, which exhibited minimal positive staining (Fig. 5A and Fig. S3; *p ≤ 0.0001). Notably, among the HFD-fed mice, Atm+ hepatocytes underwent significantly more apoptosis than those from Atm-/- mice (Fig. 5A; **p < 0.001). To dissect the molecular basis for this ATM-dependent apoptotic response to HFD feeding, we used quantitative PCR to measure the expression of the DNA damage-responsive, pro-apoptotic gene Puma.28 HFD-fed Atm+ mice (n = 12) had increased Puma expression when compared with all other groups, including HFD-fed Atm-deficient mice (n = 10–12/group), indicating that PUMA is induced after HFD feeding through a mechanism that requires ATM (Fig. 5B; *p < 0.01). HFD feeding similarly resulted in increased PUMA protein levels as measured by immunoblotting of livers from Atm+ HFD-fed mice, and this induction of PUMA was greatly reduced in samples from Atm-deficient mice (Fig. 5C). Together, these data indicate that HFD feeding induces ATM-dependent upregulation of PUMA and hepatocyte apoptosis.

figure cc-11-1918-g5
Figure 5. High fat diet feeding induces Atm-dependent steatoapoptosis. (A) TUNEL staining to identify apoptotic cells in liver sections from mice of the indicated group (n = 5–10/group) was quantified and revealed that HFD groups, regardless ...


NAFLD represents an array of hepatic lesions ranging from benign steatosis to hepatic inflammation and cirrhosis. Because of the increasing prevalence of this disease, a more thorough understanding of the factors involved in NAFLD progression is warranted. This study investigated the roles of ATM, a protein kinase involved in maintaining genomic stability, in a diet-induced model of NAFLD/NASH. Other models of NASH use genetic defects, chemical agents or nutrient depletion to induce steatohepatitis and hepatic fibrosis. Nutritional models, such as sucrose- and fructose-rich diets, cause only minimal hepatic steatosis and inflammation.21 Conversely, the diet used in this study closely recreated the NAFLD phenotype of steatohepatitis and fibrosis by coupling high fat with cholesterol and cholic acid. Fat accumulation elsewhere in the body was not observed, presumably due to the fat source and the relatively short time period mice were fed the HFD.

Eight weeks of HFD feeding caused hepatic phenotypes in mice, consistent with NAFLD, including hepatic steatosis and oxidative stress indicated by elevated levels of ROS and increased expression of p22phox, a gene encoding a subunit of NADPH oxidase. Hepatic lipid accumulation is considered an initiating factor in NAFLD pathogenesis and renders hepatocytes susceptible to oxidative stress. High levels of oxidative stress correlate with increasing severity of liver disease in humans.4 In this study, we showed that HFD-induced steatohepatitis and hepatic oxidative stress were present in both Atm+ and Atm-/- mice, which recapitulates essential components of NAFLD/NASH and creates a hepatic environment ideal to study the effects of Atm loss on NAFLD progression.

Chronic liver injury results in fibrosis.29,30 After 8 weeks of HFD feeding, Atm+ mice exhibited more hepatic fibrosis than Atm-/- mice. Interestingly, the same trend was observed for hepatic apoptosis, with Atm+ mice exhibiting more hepatocyte apoptosis than Atm-/- mice after chronic HFD feeding. This is consistent with data indicating that apoptosis contributes to fibrosis in NAFLD/NASH progression as well as with in vitro studies that suggest a direct link between hepatocyte apoptosis and fibrosis.10,12,13,31 Although the precise mechanistic details have not been fully resolved, available data indicate that apoptosis and cell clearance stimulate quiescent hepatic stellate cells, triggering their transformation to a myofibroblast phenotype and subsequent extracellular matrix deposition.11,32

Differences in fibrosis and steatoapoptosis between HFD-fed Atm+ and Atm-/- mice correlated with expression levels of the pro-apoptotic gene Puma at both the mRNA and protein levels. Excessive ROS production, as seen in this model, may induce apoptosis via various pathways, including the intrinsic pathway of apoptosis, which involves PUMA signaling.11,33 Recent studies identify PUMA as an important factor in hepatocyte steatoapoptosis, and PUMA inhibition or deficiency leads to decreased hepatic apoptosis.33-36 Moreover, hepatic Puma expression is increased in NASH patients as compared with expression levels in NAFLD or healthy patients, implicating PUMA in fatty liver disease progression, likely through functions related to steatoapoptosis and fibrosis formation.9,34 Puma activation relies primarily on the DNA damage-inducible transcription factor p53.33 ATM is an upstream activator of p53, linking ATM to PUMA-mediated apoptosis in a manner consistent with the ATM-dependent upregulation of Puma expression in this model. We also evaluated components of the extrinsic pathway of apoptosis, including Fas and Fas receptor, via quantitative PCR. There were appreciable increases in Fas receptor expression between SD and HFD groups, but the levels were nearly identical between Atm+ and Atm-/- HFD groups (Fig. S4). Further, there was no detectable expression of Fas ligand in any treatment group (data not shown). Although we cannot entirely rule out involvement of the extrinsic pathway of apoptosis in high-fat-mediated apoptotic cell death, it does not appear to account for differences observed in hepatocyte death between HFD fed Atm+ and Atm-/- mice.

One of the most common causes of hepatic genomic instability is ROS-induced DNA damage.37,38 In this study, chronic HFD feeding caused oxidative nucleic acid damage regardless of Atm status as evidenced by the similar percent of 8-OHG-positive cells in HFD-fed Atm+ and Atm-/- mice. Interestingly, phosphorylation of histone H2AX, a marker of DNA DSBs, was not detectably induced in the HFD groups. ©H2AX-positive hepatocytes were present in IR-treated control samples, demonstrating that the immunohistochemical assay for ©H2AX was functional. Thus, hepatocytes are capable of phosphorylating H2AX but do not do so to a detectable level in response to HFD feeding. Studies show that ROS-induced ATM activation can occur in the absence of DNA DSBs, and this is associated with phosphorylation of only a subset of ATM targets that does not include H2AX.17,18 Taken together, these data suggest that ATM-mediated fatty liver phenotypes may be the result of an oxidative stress response rather than the classical ©H2AX associated signaling pathway. The observation of phospho-ATM in the cytoplasm of fat-laden hepatocytes provides further support for a ROS-mediated mechanism of ATM activation.17 Within the nucleus, both 8-OHG and p-ATM seemed to be more abundant at the nuclear periphery. One hypothesis to explain the observed localization of 8-OHG is that the low permeability of reactive oxygen species limits their mobility through the nuclear membrane, thus creating the most extensive oxidative damage at the nuclear periphery.39 The observed staining patterns also may relate to previous in vitro studies, demonstrating that oxidative damage preferentially occurs in heterochromatin, which is predominantly located in the nuclear periphery, and that ATM is critical in repairing this damage.40

NASH carries a risk of HCC formation. It is well established that apoptosis promotes hepatic fibrosis, and both are key risk factors for HCC development, with hepatic fibrosis and cirrhosis being present in 80–90% of patients with HCC.30,41,42 In this study, steatoapoptosis and fibrosis were observed at highest levels in ATM-expressing HFD-fed mice, indicating that the ATM pathway promotes both during hepatic lipid accumulation. Although apoptosis is generally regarded as a tumor-suppressing mechanism that clears damaged cells at risk of undergoing malignant transformation, there also are several examples of tumor-promoting effects of apoptosis, including in HCC.43 Here, the ultimate effect of reduced fibrosis and apoptosis on tumor development could not be determined, because Atm-/- mice develop thymic lymphoma at 3–4 mo of age,24-26 preventing long-term evaluation of liver carcinogenesis. However, based upon the recognized influences of apoptosis and fibrosis during HCC development, it is reasonable to hypothesize that the observed Atm-dependent increases in these processes could promote hepatic carcinogenesis. Therefore, pharmacological inhibition of ATM may be beneficial to limit fibrosis formation and NAFLD progression.

In summary, our findings suggest that hepatic fat accumulation triggers oxidative stress and activation of an ATM-mediated DNA damage response, resulting in steatoapoptosis and hepatic fibrosis. Human trials have been inconclusive regarding effective therapy for NAFLD, including antioxidants, insulin sensitizers, weight loss or a combination of these modalities.44,45 However, our data suggest that modulation of ROS levels and ATM activation may provide novel therapeutic targets to prevent the progression of NAFLD. The mouse model described here provides a system for further experimental evaluation of how candidate therapeutics, like the antioxidants NAC and Tempol, which suppress the ROS to which ATM responds,46-48 impact fatty liver disease.

Materials and Methods

Animal husbandry.

Atm-null mice were used on a FVB/N strain background.25Atm+/- mice were intercrossed, and progeny were divided into four study groups. In the first and second group, Atm+ (Atm+/+ and Atm+/-) and Atm-/- mice were fed a standard diet (SD) (<0.05% cholesterol, 0% cholic acid, 5% triglyceride; 7012 Harlan Teklad LM-485 Mouse/Rat Sterilizable Diet), and in the third and fourth groups, Atm+ and Atm-/- mice were fed a high-fat diet (HFD) (1.0% cholesterol, 0.5% cholic acid, 18% triglyceride; TD.880511, Harlan, Teklad Lab Animal Diets; “Paigen diet”) for eight weeks. Initially, both sexes were assessed separately; however, no differences were seen between the sexes, and male and female data were combined (data not shown). Likewise, data from Atm+/- and Atm+/+ mice were initially evaluated independently and combined when data indicated there were no differences between these groups. A subset of mice was whole-body irradiated with 10 Gy irradiation 8 h prior to tissue collection. Mice were housed under specific pathogen-free conditions in an Association for the Assessment and Accreditation of Laboratory Animal Care International accredited facility. Mice received food and water ad libitum. All protocols were approved by the Cornell University Institutional Animal Care and Use Committee.

Tissue harvest and histopathology.

After eight weeks of HFD or SD feeding, mice were euthanized with CO2. The liver was removed aseptically and weighed. Approximately 30 mg of liver was flash frozen in both liquid N2 and in Tissue-Tek O.C.T. (Sakura Finetek) and stored at -80°C. The remainder of the liver was fixed in 10% neutral buffered formalin. Tissues were processed, paraffin embedded, cut into 4–5 µ sections and stained with hematoxylin and eosin. Additional liver sections were stained with Oil Red-O or pico- Sirius Red. Tissues were evaluated by a pathologist blinded to sample identity and study design. Hepatic fat accumulation, inflammation and fibrosis were scored on an ascending 0–4 scale.49

Quantitative PCR.

RNA was extracted from liver samples using the E.Z.N.A Total RNA Kit (Omega), and residual genomic DNA was removed with DNase treatment (Omega), using the manufacturer’s recommendations. Total RNA (1–2 µg) was used for synthesis of complimentary DNA (qScript cDNA Synthesis Kit; Quanta Biosciences). PerfeCTa SYBR Green FastMix, Low ROX quantitative polymerase chain reaction (PCR) Master Mix (Quanta Biosciences) was used for the quantitative PCR reaction utilizing Applied Biosystems 7500. Gene expression levels were normalized to Gapdh expression, and quantification was determined via the ΔΔCT method using the Atm+ SD group as the comparative delta value (expression of 1). Error bars represent relative log conversion of standard error as described previously in reference 50. Primer sets were acquired (Integrated DNA Technologies) as follows: glyceraldehyde-3-phosphate dehydrogenase (Gapdh), p22phox, Puma, Fas and Fas Receptor (Table S1).


Hepatocyte apoptosis was quantified by terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling per manufacturers recommendation (TUNEL; ApopTag Peroxidase In Situ Apoptosis Detection Kit; Millipore). Positive staining was detected with 3,3'-diaminobenzidine tetrahydrochloride (DAB) (Invitrogen) and sections were counterstained with hematoxylin. Liver sections from 5 animals per group were TUNEL stained. For each section, four 20x fields were chosen randomly and the number of TUNEL-positive and TUNEL-negative hepatocytes were quantified.

Immunohistologic detection of phosphorylated H2AX was determined on formalin-fixed, paraffin embedded liver sections. Briefly, sections were deparaffinized and rehydrated. Antigen retrieval was performed by heating the samples to 95°C in 0.25 mM EDTA, pH 8, over 50 min. Sections were treated with 3% H2O2 for 10 min and blocked with 4% bovine serum albumin and 20% Tween in 1x TBS at room temperature and then incubated overnight with primary antibody (1:200 dilution; mouse monoclonal anti-©H2AX antibody; Millipore) at 4°C. Slides were treated with biotinylated secondary antibody for 20 min and detected with Histostain Kit (Invitrogen). Quantification of positive ©H2AX hepatocytes was performed in the same manner described for TUNEL staining.

Hydrogen peroxide, 8-hydroxyguanosine, phospho-ATM and dihydrorhodamine imaging and quantification.

Hepatic H2O2 levels were detected using the Amplex Red Hydrogen Peroxide/Peroxidase Assay Kit (Molecular Probes). Single-cell hepatocyte suspensions were obtained from fresh liver samples, and red blood cells were removed using a percol gradient. Reactions containing 50 [proportional, variant]M Amplex Red reagent, 0.1 U/ mL horse radish peroxidase and 1 x 105 hepatocytes in 50 mM sodium phosphate buffer, pH 7.4, were incubated for 30 min at 37°C. Fluorescence was measured with a microplate reader using excitation at 530 nm and fluorescence detection at 590 nm. Absolute H2O2 levels were determined in relationship to a standard curve.

Hepatic superoxide was visualized using dihydrorhodamine 123 (DHR) (Molecular Probes/Invitrogen) fluorescent imaging. Hepatocytes were harvested similarly to the H2O2 assay. Hepatocytes were washed in 1x PBS, resuspended in media and incubated with 100 [proportional, variant]L of 25 [proportional, variant]g/mL solution of DHR in the dark at 37°C for 20 min. Hepatocytes were counted, and cytospins were performed with 1 x 106 cells. Images were obtained via confocal microscopy using excitation 596 nm, detection at 600–650 nm and magnification 63x.51 Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich). Mean red fluorescent values were determined by averaging the fluorescence within a set number of pixels placed over five areas within a 63x field, three fields per sample. Results were expressed on a scale of 0–250.

DNA damage in hepatocytes was detected using an anti-8-hydroxyguanosine (8-OHG) antibody (ab10802, Abcam), and the presence of phosphorylated ATM in hepatocytes was detected using an anti-phospho-Serine1981-ATM (p-ATM) antibody (#200-301-400, Rockland) on frozen liver sections. Liver samples were embedded in Tissue-Tek O.C.T., flash frozen in liquid N2 and sectioned at 8–10 [proportional, variant]M. Sections were fixed in a mixture of three parts acetone, one part pure ethanol and washed in 1x PBS. Sections were blocked with casein/10% donkey serum for 20 min at room temperature, washed in 1x PBS and then incubated with goat anti-8-OHG antibody (1:100 dilution) for 1 h at 37°C or mouse anti-p-ATM antibody (1:200 dilution) overnight at 4°C in a humidifying chamber. Sections were washed in 1x PBS and treated with secondary antibodies labeled with Texas Red (1:500 dilution; Jackson ImmunoResearch). Nuclei were labeled with 4',6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich). For 8-OHG quantification, the percentage of positive 8-OHG cells was calculated over three 80x fields from three different mice per group.

Western blotting.

Cells were harvested and solubilized in RIPA buffer as described previously in reference 52, followed by sonication. Total protein was quantified by Bradford assay, resolved on 6% to 14% SDS-PAGE gels, and subjected to immunoblotting using antibodies against PUMA (Abcam, #ab9643) and anti-® actin (Sigma, #A5441). Western blot imaging and quantification was performed using a Versa Doc Imaging System (Bio-Rad Laboratories).

Statistical analyses.

Parametric data were analyzed utilizing one-way analysis of variance followed by Tukey’s post-test or by t-test. Nonparametric data were analyzed by the Mann-Whitney-U test followed by Dunn’s post test. p-values of ≤ 0.05 were considered statistically significant. Data are presented as mean ± standard error of the mean unless otherwise specified. All analyses were made with GraphPad Prism version 5.0 for Windows (GraphPad Software).

Supplementary Material

Additional material



nonalcoholic fatty liver disease
nonalcoholic steatohepatitis
hepatocellular carcinoma
free fatty acids
reactive oxygen species
ataxia telangiectasia mutated
double strand DNA breaks
ataxia telangiectasia
standard diet
high fat diet
ionizing radiation
p53 upregulated modulator of apoptosis
terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling
dihydrorhodamine 123

Disclosure of Potential Conflicts of Interest

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.



The authors thank Eric Denkers and his lab for valuable suggestions and the use of substrates; Luce Guanzini for reading and editing the manuscript; and Drew Kirby for technical support. This work was supported by the Cornell University College of Veterinary Medicine Dean’s Fund for Clinical Excellence (to E.K.D.); the Cornell University College of Veterinary Medicine Clinical Fellowship Program (to E.K.D.); a Cornell University College of Veterinary Medicine Graduate Research Assistantship (to G.B.); and NIH grants R01 CA108773 (to R.S.W.), R03 HD058220 (to R.S.W.), and K08 DK077728 (to K.J.M.). Irradiations were performed with shared instrumentation supported by NCRR grant S10RR023781.



1. Cohen JC, Horton JD, Hobbs HH. Human fatty liver disease: old questions and new insights. Science. 2011;332:1519–23. doi: 10.1126/science.1204265. [PMC free article] [PubMed] [Cross Ref]
2. Starley BQ, Calcagno CJ, Harrison SA. Nonalcoholic fatty liver disease and hepatocellular carcinoma: a weighty connection. Hepatology. 2010;51:1820–32. doi: 10.1002/hep.23594. [PubMed] [Cross Ref]
3. Williams CD, Stengel J, Asike MI, Torres DM, Shaw J, Contreras M, et al. Prevalence of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis among a largely middle-aged population utilizing ultrasound and liver biopsy: a prospective study. Gastroenterology. 2011;140:124–31. doi: 10.1053/j.gastro.2010.09.038. [PubMed] [Cross Ref]
4. Hardwick RN, Fisher CD, Canet MJ, Lake AD, Cherrington NJ. Diversity in antioxidant response enzymes in progressive stages of human nonalcoholic fatty liver disease. Drug Metab Dispos. 2010;38:2293–301. doi: 10.1124/dmd.110.035006. [PubMed] [Cross Ref]
5. Caldwell S, Park SH. The epidemiology of hepatocellular cancer: from the perspectives of public health problem to tumor biology. J Gastroenterol. 2009;44(Suppl 19):96–101. doi: 10.1007/s00535-008-2258-6. [PubMed] [Cross Ref]
6. Mehta K, Van Thiel DH, Shah N, Mobarhan S. Nonalcoholic fatty liver disease: pathogenesis and the role of antioxidants. Nutr Rev. 2002;60:289–93. doi: 10.1301/002966402320387224. [PubMed] [Cross Ref]
7. Ong J, Younossi ZM, Reddy V, Price LL, Gramlich T, Mayes J, et al. Cryptogenic cirrhosis and posttransplantation nonalcoholic fatty liver disease. Liver Transpl. 2001;7:797–801. doi: 10.1053/jlts.2001.24644. [PubMed] [Cross Ref]
8. Kusminski CM, Shetty S, Orci L, Unger RH, Scherer PE. Diabetes and apoptosis: lipotoxicity. Apoptosis. 2009;14:1484–95. doi: 10.1007/s10495-009-0352-8. [PubMed] [Cross Ref]
9. Cazanave SC, Elmi NA, Akazawa Y, Bronk SF, Mott JL, Gores GJ. CHOP and AP-1 cooperatively mediate PUMA expression during lipoapoptosis. Am J Physiol Gastrointest Liver Physiol. 2010;299:G236–43. doi: 10.1152/ajpgi.00091.2010. [PubMed] [Cross Ref]
10. Feldstein A, Gores GJ. Steatohepatitis and apoptosis: therapeutic implications. Am J Gastroenterol. 2004;99:1718–9. doi: 10.1111/j.1572-0241.2004.40573.x. [PubMed] [Cross Ref]
11. Feldstein AE, Gores GJ. Apoptosis in alcoholic and nonalcoholic steatohepatitis. Front Biosci. 2005;10:3093–9. doi: 10.2741/1765. [PubMed] [Cross Ref]
12. Machado MV, Cortez-Pinto H. Cell death and nonalcoholic steatohepatitis: where is ballooning relevant? Expert Rev Gastroenterol Hepatol. 2011;5:213–22. doi: 10.1586/egh.11.16. [PubMed] [Cross Ref]
13. Noguchi Y, Young JD, Aleman JO, Hansen ME, Kelleher JK, Stephanopoulos G. Tracking cellular metabolomics in lipoapoptosis- and steatosis-developing liver cells. Mol Biosyst. 2011;7:1409–19. doi: 10.1039/c0mb00309c. [PubMed] [Cross Ref]
14. Sunny NE, Parks EJ, Browning JD, Burgess SC. Excessive hepatic mitochondrial TCA cycle and gluconeogenesis in humans with nonalcoholic fatty liver disease. Cell Metab. 2011;14:804–10. doi: 10.1016/j.cmet.2011.11.004. [PubMed] [Cross Ref]
15. McKinnon PJ. ATM and the molecular pathogenesis of ataxia telangiectasia. Annu Rev Pathol. 2012;7:303–21. doi: 10.1146/annurev-pathol-011811-132509. [PubMed] [Cross Ref]
16. Armata HL, Golebiowski D, Jung DY, Ko HJ, Kim JK, Sluss HK. Requirement of the ATM/p53 tumor suppressor pathway for glucose homeostasis. Mol Cell Biol. 2010;30:5787–94. doi: 10.1128/MCB.00347-10. [PMC free article] [PubMed] [Cross Ref]
17. Ditch S, Paull TT. The ATM protein kinase and cellular redox signaling: beyond the DNA damage response. Trends Biochem Sci. 2012;37:15–22. doi: 10.1016/j.tibs.2011.10.002. [PMC free article] [PubMed] [Cross Ref]
18. Guo Z, Kozlov S, Lavin MF, Person MD, Paull TT. ATM activation by oxidative stress. Science. 2010;330:517–21. doi: 10.1126/science.1192912. [PubMed] [Cross Ref]
19. Barzilai A, Yamamoto K. DNA damage responses to oxidative stress. DNA Repair (Amst) 2004;3:1109–15. doi: 10.1016/j.dnarep.2004.03.002. [PubMed] [Cross Ref]
20. Gao D, Wei C, Chen L, Huang J, Yang S, Diehl AM. Oxidative DNA damage and DNA repair enzyme expression are inversely related in murine models of fatty liver disease. Am J Physiol Gastrointest Liver Physiol. 2004;287:G1070–7. doi: 10.1152/ajpgi.00228.2004. [PubMed] [Cross Ref]
21. Matsuzawa N, Takamura T, Kurita S, Misu H, Ota T, Ando H, et al. Lipid-induced oxidative stress causes steatohepatitis in mice fed an atherogenic diet. Hepatology. 2007;46:1392–403. doi: 10.1002/hep.21874. [PubMed] [Cross Ref]
22. Barzilai A, Rotman G, Shiloh Y. ATM deficiency and oxidative stress: a new dimension of defective response to DNA damage. DNA Repair (Amst) 2002;1:3–25. doi: 10.1016/S1568-7864(01)00007-6. [PubMed] [Cross Ref]
23. Bakkenist CJ, Kastan MB. DNA damage activates ATM through intermolecular autophosphorylation and dimer dissociation. Nature. 2003;421:499–506. doi: 10.1038/nature01368. [PubMed] [Cross Ref]
24. Barlow C, Hirotsune S, Paylor R, Liyanage M, Eckhaus M, Collins F, et al. Atm-deficient mice: a paradigm of ataxia telangiectasia. Cell. 1996;86:159–71. doi: 10.1016/S0092-8674(00)80086-0. [PubMed] [Cross Ref]
25. Elson A, Wang Y, Daugherty CJ, Morton CC, Zhou F, Campos-Torres J, et al. Pleiotropic defects in ataxia-telangiectasia protein-deficient mice. Proc Natl Acad Sci U S A. 1996;93:13084–9. doi: 10.1073/pnas.93.23.13084. [PubMed] [Cross Ref]
26. Xu Y, Ashley T, Brainerd EE, Bronson RT, Meyn MS, Baltimore D. Targeted disruption of ATM leads to growth retardation, chromosomal fragmentation during meiosis, immune defects, and thymic lymphoma. Genes Dev. 1996;10:2411–22. doi: 10.1101/gad.10.19.2411. [PubMed] [Cross Ref]
27. Kinner A, Wu W, Staudt C, Iliakis G. Gamma-H2AX in recognition and signaling of DNA double-strand breaks in the context of chromatin. Nucleic Acids Res. 2008;36:5678–94. doi: 10.1093/nar/gkn550. [PMC free article] [PubMed] [Cross Ref]
28. Yu J, Zhang L. PUMA, a potent killer with or without p53. Oncogene. 2008;27(Suppl 1):S71–83. doi: 10.1038/onc.2009.45. [PMC free article] [PubMed] [Cross Ref]
29. Krizhanovsky V, Yon M, Dickins RA, Hearn S, Simon J, Miething C, et al. Senescence of activated stellate cells limits liver fibrosis. Cell. 2008;134:657–67. doi: 10.1016/j.cell.2008.06.049. [PMC free article] [PubMed] [Cross Ref]
30. Povero D, Busletta C, Novo E, di Bonzo LV, Cannito S, Paternostro C, et al. Liver fibrosis: a dynamic and potentially reversible process. Histol Histopathol. 2010;25:1075–91. [PubMed]
31. Zhan SS, Jiang JX, Wu J, Halsted C, Friedman SL, Zern MA, et al. Phagocytosis of apoptotic bodies by hepatic stellate cells induces NADPH oxidase and is associated with liver fibrosis in vivo. Hepatology. 2006;43:435–43. doi: 10.1002/hep.21093. [PubMed] [Cross Ref]
32. Canbay A, Friedman S, Gores GJ. Apoptosis: the nexus of liver injury and fibrosis. Hepatology. 2004;39:273–8. doi: 10.1002/hep.20051. [PubMed] [Cross Ref]
33. Villunger A, Michalak EM, Coultas L, Müllauer F, Böck G, Ausserlechner MJ, et al. p53- and drug-induced apoptotic responses mediated by BH3-only proteins puma and noxa. Science. 2003;302:1036–8. doi: 10.1126/science.1090072. [PubMed] [Cross Ref]
34. Cazanave SC, Mott JL, Elmi NA, Bronk SF, Werneburg NW, Akazawa Y, et al. JNK1-dependent PUMA expression contributes to hepatocyte lipoapoptosis. J Biol Chem. 2009;284:26591–602. doi: 10.1074/jbc.M109.022491. [PMC free article] [PubMed] [Cross Ref]
35. Michalak EM, Jansen ES, Happo L, Cragg MS, Tai L, Smyth GK, et al. Puma and to a lesser extent Noxa are suppressors of Myc-induced lymphomagenesis. Cell Death Differ. 2009;16:684–96. doi: 10.1038/cdd.2008.195. [PMC free article] [PubMed] [Cross Ref]
36. Michalak EM, Villunger A, Adams JM, Strasser A. In several cell types tumour suppressor p53 induces apoptosis largely via Puma but Noxa can contribute. Cell Death Differ. 2008;15:1019–29. doi: 10.1038/cdd.2008.16. [PMC free article] [PubMed] [Cross Ref]
37. Farinati F, Cardin R, Bortolami M, Grottola A, Manno M, Colantoni A, et al. Estrogens receptors and oxidative damage in the liver. Mol Cell Endocrinol. 2002;193:85–8. doi: 10.1016/S0303-7207(02)00100-4. [PubMed] [Cross Ref]
38. Lu S, Shen KC, Wang Y, Brooks SC, Wang YA. Impaired hepatocyte survival and liver regeneration in Atm-deficient mice. Hum Mol Genet. 2005;14:3019–25. doi: 10.1093/hmg/ddi333. [PubMed] [Cross Ref]
39. Nunomura A, Perry G, Pappolla MA, Wade R, Hirai K, Chiba S, et al. RNA oxidation is a prominent feature of vulnerable neurons in Alzheimer’s disease. J Neurosci. 1999;19:1959–64. [PubMed]
40. Woodbine L, Brunton H, Goodarzi AA, Shibata A, Jeggo PA. Endogenously induced DNA double strand breaks arise in heterochromatic DNA regions and require ataxia telangiectasia mutated and Artemis for their repair. Nucleic Acids Res. 2011;39:6986–97. doi: 10.1093/nar/gkr331. [PMC free article] [PubMed] [Cross Ref]
41. El-Serag HB. Hepatocellular carcinoma. N Engl J Med. 2011;365:1118–27. doi: 10.1056/NEJMra1001683. [PubMed] [Cross Ref]
42. Yasui K, Hashimoto E, Komorizono Y, Koike K, Arii S, Imai Y, et al. Japan NASH Study Group, Ministry of Health, Labour, and Welfare of Japan Characteristics of patients with nonalcoholic steatohepatitis who develop hepatocellular carcinoma. Clin Gastroenterol Hepatol. 2011;9:428–33, quiz e50. doi: 10.1016/j.cgh.2011.01.023. [PubMed] [Cross Ref]
43. Tang D, Lotze MT, Kang R, Zeh HJ. Apoptosis promotes early tumorigenesis. Oncogene. 2011;30:1851–4. doi: 10.1038/onc.2010.573. [PubMed] [Cross Ref]
44. Dowman JK, Armstrong MJ, Tomlinson JW, Newsome PN. Current therapeutic strategies in non-alcoholic fatty liver disease. Diabetes Obes Metab. 2011;13:692–702. doi: 10.1111/j.1463-1326.2011.01403.x. [PubMed] [Cross Ref]
45. Foster T, Budoff MJ, Saab S, Ahmadi N, Gordon C, Guerci AD. Atorvastatin and antioxidants for the treatment of nonalcoholic fatty liver disease: the St Francis Heart Study randomized clinical trial. Am J Gastroenterol. 2011;106:71–7. doi: 10.1038/ajg.2010.299. [PubMed] [Cross Ref]
46. Reliene R, Schiestl RH. Antioxidants suppress lymphoma and increase longevity in Atm-deficient mice. J Nutr. 2007;137(Suppl):229S–32S. [PubMed]
47. Reliene R, Schiestl RH. Experimental antioxidant therapy in ataxia telangiectasia. Clin Med Oncol. 2008;2:431–6. [PMC free article] [PubMed]
48. Schubert R, Erker L, Barlow C, Yakushiji H, Larson D, Russo A, et al. Cancer chemoprevention by the antioxidant tempol in Atm-deficient mice. Hum Mol Genet. 2004;13:1793–802. doi: 10.1093/hmg/ddh189. [PubMed] [Cross Ref]
49. Rogers AB, Boutin SR, Whary MT, Sundina N, Ge Z, Cormier K, et al. Progression of chronic hepatitis and preneoplasia in Helicobacter hepaticus-infected A/JCr mice. Toxicol Pathol. 2004;32:668–77. doi: 10.1080/01926230490524247. [PubMed] [Cross Ref]
50. Maurer KJ, Rao VP, Ge Z, Rogers AB, Oura TJ, Carey MC, et al. T-cell function is critical for murine cholesterol gallstone formation. Gastroenterology. 2007;133:1304–15. doi: 10.1053/j.gastro.2007.07.005. [PMC free article] [PubMed] [Cross Ref]
51. Sobreira C, Davidson M, King MP, Miranda AF. Dihydrorhodamine 123 identifies impaired mitochondrial respiratory chain function in cultured cells harboring mitochondrial DNA mutations. J Histochem Cytochem. 1996;44:571–9. doi: 10.1177/44.6.8666742. [PubMed] [Cross Ref]
52. Harlow E, Lane D. Lysing Tissue-Culture Cells for Immunoprecipitation. Cold Spring Harbor Protocols. 2006;2006:pdb.prot4531. doi: 10.1101/pdb.prot4531. [PubMed] [Cross Ref]

Articles from Cell Cycle are provided here courtesy of Taylor & Francis