Liver Biology and Pathobiology
Oxidative mitochondrial DNA damage and deletion in hepatocytes of rejecting liver allografts in rats: Role of TNF-α†
Article first published online: 16 JUN 2005
Copyright © 2005 American Association for the Study of Liver Diseases
Volume 42, Issue 1, pages 208–215, July 2005
How to Cite
Nagakawa, Y., Williams, G. M., Zheng, Q., Tsuchida, A., Aoki, T., Montgomery, R. A., Klein, A. S. and Sun, Z. (2005), Oxidative mitochondrial DNA damage and deletion in hepatocytes of rejecting liver allografts in rats: Role of TNF-α. Hepatology, 42: 208–215. doi: 10.1002/hep.20755
Potential conflict of interest: Nothing to report.
- Issue published online: 16 JUN 2005
- Article first published online: 16 JUN 2005
- Manuscript Accepted: 26 APR 2005
- Manuscript Received: 24 NOV 2004
An orthotopic liver transplant model in the rat was used to evaluate the role of tumor necrosis factor alpha (TNF-α) in liver transplant rejection. There were significantly increased levels of TNF-α mRNA and parallel increases in 8-hydroxy-2′ deoxyguanosine (8-OHdG) indicative of oxidative DNA damage present 7 to 12 days after transplantation. Cells staining positively for 8-OHdG were localized to the cytoplasm of hepatocytes adjacent to the TNF-α expressing inflammatory cells in the portal areas or in patches surrounded by inflammatory cells in the hepatic sinusoids. Significantly more cells staining for 8-OHdG were found in the allogeneic grafts that were strongly rejected than in the syngeneic controls or in the grafts placed in species that accepted the allograft permanently after a rejection episode. TUNEL reactivity lagged 2 days behind peak reactivity for 8-OHdG. On day 12 after transplantation, many cells stained for both 8-OHdG and TUNEL, indicating that the cells suffering oxidative DNA injury were undergoing apoptosis or death. Oxidative injury resulted in mtDNA deletion consisting of 4,834 base-pairs. Studies of hepatocytes cultured from normal rats displayed dose-dependent relationships between TNF-α concentration and 8-OHdG and mtDNA mutation. Repetitive intraperitoneal injection of Enbrel, a TNF receptor blocker, significantly decreased hepatocyte 8-OHdG levels and the frequency of deleted mtDNA while greatly extending graft survival time. In conclusion, the data presented implicate TNF-α as being capable of causing oxidative DNA damage and mtDNA mutation in hepatocytes. (HEPATOLOGY 2005;42:208–215.)
Acute rejection of the liver is characterized by inflammatory cell infiltration and increased cytokine production. Tumor necrosis factor alpha (TNF-α) serum levels are significantly elevated in liver transplant patients experiencing acute rejection,1, 2 but this is nonspecific, as it is elevated in other forms of liver injury and liver failure. TNF-α has been shown to increase the production of reactive oxygen species (ROS) in cultured cardiac myocytes, vascular endothelial cells, tumor cells, and hepatocytes.3–5 Excessive amounts increase ROS production by mitochondria, which are likely to damage cellular macromolecules, including DNA. Mitochondrial DNA (mtDNA) is 10 to 20 times more vulnerable to oxidative damage and subsequent mutations than nuclear DNA.6–8 Thus, ROS produced in mitochondria are likely to cause mtDNA damage, leading to the cascade of events culminating in apoptosis or cell death. The oxidation of deoxyguanosine to 8-hydroxy-2′-deoxyguanosine (8-OHdG) and subsequent mutations of mtDNA are reported to be involved in the pathogenesis of many chronic conditions.9–14
We hypothesized that severe oxidative mtDNA damage in hepatocytes occurs acutely by the release of inflammatory cytokines such as TNF-α during acute rejection, and that this damage is not repaired completely, leading to deletions that contribute to liver failure after transplantation. To test this hypothesis, we first measured the levels of 8-OHdG in mtDNA, the frequency of 4,834-bp deletions of mtDNA, and the levels of TNF-α in rats receiving syngeneic liver grafts, grafts destined to rejected, and grafts accepted despite a major histocompatibility barrier. Second, we examined the effects of TNF-α alone on ROS production and mtDNA deletion in primary cultures of normal hepatocytes. Finally, we evaluated the effect of anti–TNF-α Enbrel (etanercept; Immunex Corporation, Thousand Oaks, CA) treatment on 8-OHdG levels in mtDNA, the frequency of mtDNA with a 4,834-bp deletion, and graft and animal survival times.
Materials and Methods
Male Lewis (RT11) and DA (RT1Aa,b) rats were purchased from Harlan Sprague-Dawley (Indianapolis, IN) and used at 7 to 8 weeks of age. Animals were maintained in pathogen-free facility of Johns Hopkins Medical Institutions. Animals were cared for according to NIH guidelines and under a protocol approved by the Johns Hopkins University Animal Care Committee.
Orthotopic Liver Transplantation.
Orthotopic liver transplantation was performed according to a method previously described.15 The liver grafts were preserved in cold (4°C) saline (0.9%) for 1 hour before reperfusion. The hepatic artery was not reconstructed. Three combinations were selected: (1) a model of syngenic liver transplantation (Lewis into Lewis); (2) a model of chronic allograft acceptance (Lewis into DA); and (3) a model of acute allograft rejection and death in 10 to 12 days (DA into Lewis). Allograft survival was determined by recipient survival, and rejection was confirmed histologically.
Administration of Enbrel.
Enbrel is a dimeric fusion protein consisting of the extracellular ligand-binding portion of the human 75-kDa TNF receptor linked to the Fc portion of human IgG1. Inactivation of TNF-α was performed by repetitive intraperitoneal injection of 10 mg/kg Enbrel every 2 days (days 0, 2, 4, 6, 8) postorthotopic liver transplantation. Control animals received saline at the same times. In selected experiments, hepatocytes were isolated and liver tissues were harvested from Enbrel-treated rats on day 10 after transplantation.
Hepatocyte Isolation and Culture.
Hepatocytes were isolated from transplanted livers or nontransplanted Lewis rats using a 2-step collagenase perfusion according to the method described by Seglen.16 The viability of the initial cell suspension of hepatocytes was typically between 80% and 90% (trypan blue). Isolated hepatocytes were used for either mitochondria isolation or in vitro TNF-α treatment. For in vitro assay, isolated hepatocytes were inoculated in Dulbecco modified Eagle medium supplemented with 10% fetal bovine serum, 50 nmol/L dexamethasone (Sigma, St Louis, MO), 20 mmol/L HEPES, 0.5mg/L insulin (Sigma), 1 mmol/L ascorbic acid 2-phophate, and penicillin/streptomycin at a density of 1.5 × 105cells/cm2. All cells were maintained in a humidified incubator at 37°C/5% CO2 for 3 days, and then the medium were completely replaced by serum-free and dexamethasone-free Dulbecco modified Eagle medium 24 hours before stimulation. For stimulation, hepatocytes were exposed to recombinant TNF-α (BioSource International, Camarillo, CA) at various concentrations and maintained in culture for an additional 16 hours.
ELISA Assay for Measurement of 8-OHdG Levels in mtDNA in Hepatocytes.
The levels of 8-OHdG in mtDNA were measured by ELISA. Briefly, hepatocytes were homogenized in 5 mL 50mmol/L Tris-HCl (pH 7.4) with 50 strokes of Dounce homogenizer. The homogenates were centrifuged at 800g for 10 minutes to precipitate nuclear fraction. The supernatant was again centrifuged at 7,000g for 10 minutes to yield mitochondrial fraction. The mtDNA was extracted using the DNeasy tissue kit (Qiagen, Santa Clara, CA) according to the manufacturer's instructions. Briefly, 5 μg nuclease P1 (Roche, Indianapolis, IN) was added to 20 μg (for cultured hepatocytes) or 50 μg (for hepatocytes from liver grafts) isolated mtDNA samples. After purging with a nitrogen steam to prevent the artificial formation of 8-OHdG., the mixtures were incubated at 37°C for 1 hour to digest the DNA to nucleotides. Then, 5 μL 500 mmol/L Tris-HCl (pH 8.0), 10 mmol/L MgCl2, and 0.6 units alkaline phosphatase (New England Biolabs, Beverly, MA) were added to the samples. After purging with a nitrogen steam, the mixtures were incubated at 37°C for 1 hour to hydrolyze the nucleotides to nucleosides. The nucleoside samples were used for the determination of 8-OHdG by a competitive ELISA kit (8-OHdG check, Japan Institute for the Control of Aging, Shizuoka, Japan). The determination range was 0.125 to 10 ng/mL or 0.5 to 200 ng/mL. The levels of 8-OHdG were expressed as amounts of 8-OHdG (ng) per milligram mtDNA.
Detection of mtDNA Deletion by Polymerase Chain Reaction.
Total mtDNA was extracted from hepatocyte-derived mitochondria as previously described. The primer sets for amplification of common mtDNA deletion of 4,834-bp, which was reported to be one of the most frequent deletions,17, 18 were 5′-TTT CTT CCC AAA CCT TTC CT-3′ (7,837 to 7,856-bp) and 5′-AAG CCT GCT AGG ATG CTT C-3′ (13,108-13,126 bp). The primer sets for control amplification of wild-type mtDNA were 5′-GGT TCT TAC TTC AGG GGC CAT C-3′ (15,782-15,892 bp) and 5′-GTG GAA TTT TCT GAG GGT AGG C-3′ (16,279-16,300 bp).13 Sequence and numbering are based on the rat complete mitochondrial genomes (GenBank accession number AJ 428514). Polymerase chain reaction (PCR) contained 0.2 mmol/L deoxyribonucleotide triphosphate, 0.2 μmol/L of each primer, 1.0 unit Platinum Taq DNA polymerase (Invitrogen, Carlsbad, CA), and 1 μg total DNA as template in a 50-μL reaction solution. The thermal cycling condition was started with one cycle at 94°C for 3 minutes, and 6 cycles at 94°C for 1 minute, 64°C for 1 minute (−1°C/cycle), 72°C for 1minute 30 seconds. This was followed by 34 cycles at 94°C for 1 minute, 60°C for 1 minute, 72°C for 1 minute 30 seconds, and 72°C for final extension for 5 minutes. PCR products were electrophoresed on 1.5% agarose gels and visualized with ethidium bromide staining. The identity of the amplified PCR product was confirmed by sequencing using the AB 3730 DNA Analyzer (Applied Biosystems, Foster City, CA).
Reverse Transcription-PCR Analysis for TNF-α mRNA in Liver Grafts.
Whole liver specimens were kept frozen at −80°C until homogenized for RNA extraction using a QIAamp Kit (Qiagen). TNF-α mRNA expression was analyzed by reverse transcription PCR as previously described.19
Measurement of ROS Production.
A fluorescent probe, 2′, 7′-dichlorofluorescin diacetate (DCFH-DA; Sigma), was used for the assessment of intracellular ROS formation in cultured hepatocytes. This assay is a reliable method for the measurement of intracellular ROS such as hydrogen peroxide (H2O2), hydroxyl radical (OH−), and hydroperoxides.20–22 DCFH-DA was dissolved in absolute ethanol at a concentration of 5 mmol/L. Hepatocytes were grown on collagen-coated glass coverslips in 6-well culture plates. On culture day 4, TNF-α at various concentrations or media (control) was administered simultaneously with DCFH-DA (5 μmol/L) in culture medium. After incubation at 37°C for 2 hours, hepatocytes were washed with phosphate-buffered saline. Fluorescence images were acquired by microscopy.
Cut sections of 4 μm were prepared from formalin-fixed paraffin-embedded tissues for 8-OHdG staining or frozen tissue for TNF-α staining. Each representative section was stained with hematoxylin-eosin (H&E), and immunohistochemical stains were performed with the avidin-biotin-peroxidase complex method, using Vectastain ABC kits (Vector Laboratories, Burlingame, CA). The following antibodies were used: monoclonal anti–8-OHdG antibody (Institute for the Control of Aging, Japan) (1:100); monoclonal anti–TNF-α (R&D Systems Inc., Minneapolis, MN) at the concentration of 5 μg/mL. The antibodies were incubated at 4°C overnight. Antigen retrieval of paraffin section was achieved by a microwave. Double staining of 8-OHdG and terminal deoxynucleotidyltransferase-mediated nick-end labeling (TUNEL) staining were also performed by immunofluorescent stains using frozen sections. TUNEL staining was carried out using dUTP-FITC, according to the instructions of the manufacturer (In Situ Cell Death Detection Kit, Roche). After TUNEL staining, the sections were incubated with monoclonal anti–8-OHdG antibody (Japan Institute for the Control of Aging) and goat cyanine 2-conjugated anti-Goat IgG (1:500) (Jackson ImmunoResearch, West Grove, PA).
The results were expressed as mean values ± SEM of n-independent experiments. Analysis was performed by ANOVA with P less than .05 considered significant.
Accumulation of 8-OHdG in Hepatocytes of Acute Rejecting Liver Allografts Associated With Inflammatory Cells Infiltration.
8-OHdG, which is a marker of oxidative DNA damage,13 was detected at low levels in mtDNA of hepatocytes recovered from nontransplanted DA and Lewis livers by using a competitive ELISA method (Fig. 1). In both syngenic and allogeneic liver transplantation models, the levels of 8-OHdG in mtDNA were slightly increased on day 3 and day 5 posttransplantation compared with nontransplanted livers, and there was no significant difference in 8-OHdG levels between syngenic and allogeneic models. Thereafter, the 8-OHdG levels were significantly higher in allogeneic models. The level of 8-OHdG in mtDNA increased in hepatocytes of allografts destined for acute rejection, which was significantly higher than that found in the chronically accepted model at 10 and 12 days after liver transplantation.
To localize the expression of 8-OHdG in transplanted liver grafts, liver tissues were analyzed by histology and immunostaining. In acutely rejecting liver allografts 10 days after transplantation, heavy mononuclear cellular infiltration was present in the expanded portal areas with disruption of liver architecture. The pattern of cellular infiltration was similar in the transplanted livers destined to be accepted, but the number of infiltrating cells were much less. There were few infiltrating inflammatory cells in portal areas in the syngenic model at 10 days after transplantation (Fig. 2A).
Figure 2B shows a representative result in transplanted livers of immunohistochemical staining for 8-OHdG. Few 8-OHdG–positive cells were recognized in syngenic liver grafts on day 10 after transplantation. However, 8-OHdG–positive hepatocytes were recognized in acutely rejecting liver allografts on days 7, 10, and 12 after transplantation, especially in the area immediately around infiltrating inflammatory cells in the expanded portal areas (left panel in Fig. 2B). Figure 2C shows that the number of 8-OHdG–positive hepatocytes increased with time in rejecting liver allografts, and began to appear in the hepatic sinusoids where there was also inflammatory cell infiltration. In fact, the 8-OHdG–positive hepatocytes were invariably surrounded by or adjacent to infiltrating inflammatory cells (left panel in Fig. 2C). By contrast, there were only a few 8-OHdG–positive cells in accepting liver allografts on day 10 after transplantation (middle panel in Fig. 2B), and the number of 8-OHdG–positive cells remained the same on day 12 after transplantation (data not shown). Although the method for 8-OHdG detection stains nuclear as well as mtDNA, the staining was localized in the cytoplasm, indicating that this oxidative adduct was present mainly in the mitochondria.
To determine whether 8-OHdG–positive hepatocytes were undergoing apoptosis/necrosis, immunofluorescence staining for 8-OHdG and TUNEL were performed. Figure 2D shows that the TUNEL-positive cells and 8-OHdG–positive cells were recognized on days 7, 10, and 12 after transplantation in rejecting liver allografts, and the number of positive cells increased with time. TUNEL and 8-OHdG–double-positive cells were not present in rejecting liver allografts at day 7 and day 10 after transplantation (left and middle panels in Fig. 2D). However, double-positive cells were recognized in rejecting liver allografts on day 12 after transplantation (right panel in Fig. 2D). These results suggest that some of 8-OHdG–positive cells were the ones undergoing apoptosis/necrosis in a delayed fashion.
Accumulation of 8-OHdG in Hepatocytes of Acute Rejecting Liver Allografts Paralleled the Increased TNF-α Expression.
To determine whether the accumulation of 8-OHdG–positive hepatocytes in rejecting liver allografts was associated with increased TNF-α expression, TNF-α mRNA was semiquantitatively analyzed in liver grafts after liver transplantation by reverse transcription PCR. Figure 3A shows that TNF-α mRNA was detected at low levels in nontransplanted DA or Lewis livers. In syngeneic liver grafts, TNF-α mRNA was increased compared with nontransplanted livers and peaked at 3 days after transplantation. The levels of TNF-α mRNA were significantly higher in rejecting liver allografts compared with accepted liver allografts on days 10 and 12 after transplantation, whereas there was no significant difference between rejecting and acceptance models at 3, 5, and 7 days after transplantation (Fig. 3A). Notably, the change of TNF-α mRNA levels paralleled the increased levels of 8-OHdG (Fig. 1) in the mtDNA of the hepatocytes after transplantation.
The increased TNF-α expression in rejecting liver allografts was confirmed by immunohistochemistry staining. Figure 3B shows that there are numerous TNF-positive infiltrating inflammatory cells in expanded portal and hepatic sinusoidal areas in acute rejecting liver allografts on day 10 after transplantation (right panel). In contrast, fewer TNF-α positive cells were recognized in syngeneic (left panel) and acceptance models (data not shown).
Multiple mtDNA Deletions in Acute Rejecting Liver Allografts.
Deletions in mtDNA are the probable consequences of oxidative injury. Template mtDNA from rejecting liver allografts was amplified in a condition of 40 cycles using the primer pair between 7,835 and 13,129 bp. PCR amplification showed that multiple mtDNA deletions, including the common mtDNA deletion of 4,834 bp, were found in hepatocytes recovered from the rejected liver at days 10 and 12 after transplantation (Fig. 4A). The mtDNA recovered from rejecting liver allografts exhibited higher levels of deletion products compared with syngeneic liver grafts. In addition to the common mtDNA deletion of 4,834 bp, the representative PCR products were further analyzed by auto-sequencer. The 226-bp fragment was generated from mtDNA with the 5,074-bp deletion, which is located between 8,018 and 13,091 bp and is flanked by a 9-nucleotide direct repeat 5′-AAACGCCTA-3′ between 8,018 and 8,026 bp or between 13,092 and 13,100 bp. The 719-bp fragment was generated from mtDNA with the 4,571-bp deletion, which is located between 8,358 and 12,927 bp and is flanked by an 18-nucleotide direct repeat 5′-TCATCATCGAAACTATCA-3′ between 8,340 and 8,357 bp or between 12,910 and 12,927 bp (Fig. 4B).
TNF-α–Induced ROS Production and Oxidative mtDNA Damage in Hepatocytes In Vitro.
Primary cultures of normal hepatocytes were stimulated by rat recombinant TNF-α. The cell viability judged by trypan blue exclusion was unchanged after exposure to 100 ng/mL TNF-α for 24 hours (data not shown). However, as shown in Fig. 5A, an increase occurred in DCFH fluorescence within hepatocytes 2 hours after exposure to TNF-α at a concentration of 10 ng/mL or 100 ng/mL. Furthermore, immunofluorescence staining showed that the number of 8-OHdG–positive cells in hepatocytes co-cultured with TNF-α was significantly increased in a TNF-α dose-dependent fashion (Fig. 5B). Using the ELISA method (Fig. 5C), the level of 8-OHdG in mtDNA of hepatocytes stimulated with TNF-α at concentration of 10 ng/mL or 100 ng/mL was significantly increased compared with control hepatocytes without TNF-α or hepatocytes exposed to TNF-α at 1 ng/mL.
Effect of Intervention by Inactivation of TNF-α on mtDNA Damage and Deletion in Hepatocytes of Acute Rejecting Liver Allografts.
Inactivation of TNF-α was attempted by intraperitoneal injection of Enbrel (10 mg/kg) at days 0, 2, 4, 6, and 8 posttransplantation. The control transplanted rats were injected with the same volume of saline. Enbrel treatment significant decreased the 8-OHdG levels in mtDNA of hepatocytes recovered from acute rejecting liver allografts. Similarly, liver recipients in the rejecting model treated with Enbrel had low levels of the mtDNA deletion products at 10 days after transplantation (Fig. 6B). Fifty percent survival time was extended from 14 to 26 days (Fig. 6C).
The current studies emphasize the important role of TNF-α in liver transplant rejection in this rat model. TNF-α mRNA was highly expressed in livers destined to be rejected (DA to Lewis) compared with the syngeneic (Lewis to Lewis) or the allogeneic accepted (Lewis to DA) transplants. Early parenchymal injury manifested by oxidative DNA damage and mtDNA deletions appeared adjacent to the portal zones having high concentrations TNF-α protein and inflammatory cells. Of greatest importance, there was attenuation of oxidative injury and marked prolongation of allograft function and animal survival when the activity of TNF-α was blocked by Enbrel.
That TNF-α causes oxidative injury to mitochondria is well established. In fact, Schulze-Osthoff et al.23 showed in vitro that TNF-α interfered with mitochondrial electron transport detected after a 1-hour exposure. Cell death occurred at 3 to 6 hours. Most of the ROS were formed at the unisemiquinone site. The injury was mitochondria specific and could be prevented by blocking complex one or by provision of anti-oxidants. It is also established that TNF-α combines with the TNF-R, resulting, after complex reactions, in the activation of caspase 8 and the extrinsic apoptotic pathway. TNF-α, via caspase 8 cleavage of BID, also causes mitochondrial permeability transition, leading to cytochrome c release and activation of the intrinsic apoptotic pathway.24 Whereas caspase activation explains most of the activity of TNF, caspase blockade with a broad-spectrum inhibiter zVAD-fmk results in enhanced TNF toxicity in vivo.25 Mice given zVAD-fmk and TNF succumb from an oxidative injury that is not blocked by superoxide dismutase or catalase, but by antioxidants that enter cells. Blockade of phospholipase A2, a source of intracellular ROS, also prevented this oxidative injury. Caspase blockade is explained by caspase cleavage of phospholipase A2, providing a brake on this severe manifestation of TNF toxicity.
Our results do not provide direct evidence that TNF-α–induced mtDNA damage is the cause of acute graft failure. In vitro exposure to concentrations of TNF-α sufficient to cause oxidative injury and mtDNA mutations, did not result in cell death in vitro. Whereas higher concentrations of TNF and longer incubation may have resulted in cell death, we chose to direct our focus to the nature of the oxidative injury. Whereas others have described excessive ROS production and morphological injury in mitochondria after TNF-α exposure, this study demonstrates that TNF-α causes mtDNA damage and mutation in hepatocytes.
It is clear from our immunohistological studies that cells in the portal areas adjacent to the inflammatory infiltrate are exposed to higher concentrations of TNF-α than cells closer to the central vein. This creates an injury gradient in vivo. At one extreme, the TNF-α concentration may overwhelm the caspase “feedback loop” and cause necrosis via the phospholipase A2 mechanism. At the other, cells survive with intact respiratory function. In the middle are cells surviving with faulty repair, leaving large deletions in mtDNA similar to the large-scale deletions found in aged rats,18, 26, 27 kidneys of diabetic rats,13 and in the muscles of humans with progressive external ophthalmoplegia.28–31 The deletions of mtDNA in rats and humans are flanked by two homologous repeats that span a region that encodes five kinds of tRNA. Included in the omissions are respiratory enzyme sub-units for complexes I, IV, and V. As a consequence, the injury to mtDNA caused by TNF-α and acute rejection is likely to be progressive. Hepatocytes surviving with impaired mitochondrial respiratory function are more likely to continue to suffer from faulty reduction of O−2, sustaining ever worsening oxidative damage, and are less likely to meet the metabolic needs of the normal liver. Fortunately, the liver has immense regenerative capacity provided at least in part by extrahepatic cells. This finding may explain why the liver transplant is less apt to undergo chronic failure compared with the kidney. The kidney depends on hypertrophy rather than regeneration to compensate for loss of mass, and we speculate that mtDNA deletions may explain the progressive tubular atrophy seen in all types of chronic renal failure, including allograft rejection. In the case of the liver, the target of chronic rejection is the biliary ductules, and it is interesting that these structures are in the midst of the inflammatory infiltrate, and, therefore, subjected to the greatest challenge to their mitochondria.