Document E3pN00XmR8R5qYyyND79Xv0N

Early resistance to cell death and to onset of the mitochondrial permeability transition during hepatocarcinogenesis with 2-acetylaminofluorene Peter-Christian Klhn***, Maria Eugenia Soriano*1) William Irwin*, Daniele Renzo*, Luca Scorrano*5, Annette Bitsch15, Hans-Gnter Neumann11, and Paolo Bernards*^ Department o f Biom edical Sciences, University o f Padua, V iale Giuseppe Colombo 3, 1-35121 Padua, Italy; -Venetian institute o f M olecular Medicine, V ia Orus 2, 1-35123 Padua, itaiy; and 1lD epartm ent o f Toxicology, Universitt W rzburg, Versbacher Strasse 9, D -97078 W rzburg, Germ any Com m unicated by Douglas C. W allace, University o f California College o f M edicine, Irvine, CA, June 12, 2003 (received fo r review November 26, 2002) A hallmark of tumorigenesis is resistance to apoptosis. To explore whether resistance to cell death precedes tumor form ation, we have studied the short-term effects of the hepatocardnogen 2-acetylaminofluorene (AAF) on liver mitochondria, on hpato cytes, and on the response to bacterial endotoxin lipopolysaccha ride (LPS) in albino Wistar rats. We show that after as early as two weeks of AAF feeding liver mitochondria developed an increased resistance to opening of the permeability transition pore (PTP), an inner membrane channel that is involved in various forms of cell death. Consistent with a mitochondrial adaptive response in vivo, (i) AAF feeding increased the expression of BCL-2 in mitochondria, and (/V) hepatocytes isolated from AAF-fed rats became resistant to PTP-dependent depolarization, cytochrome c release, and cell death, which were instead observed in hepatocytes from rats fed a control diet. AAF-fed rats were fully protected from the hepatotoxic effects of the injection of 20-30 fjug of IP S plus 700 mg of D~galactosamine (o-GalN) x kg"*1of body weight, a treatment that in control rats readily caused a large increase of terminal de~ oxynucleotidyltransferase-mediated dUTP nick end labeling-posi tive cells in liver cryosections and release of alanine and aspartate aminotransferase into the bloodstream. Treatment with LPS and D-GaIN triggered cleavage of BID, a BCL-2 family member, in the livers of both control- and AAF-fed animals, whereas caspase 3 was cleaved only in control-fed animals, indicating that the mitochon drial proapoptotic pathway had been selectively suppressed dur ing AAF feeding. Phenotypic reversion was observed after stop ping the carcinogenic diet. These results underscore a key role of mitochondria in apoptosis and demonstrate that regulation of the mitochondrial PTP is altered early during AAF carcinogenesis, which matches, and possibly causes, the increased resistance of hepatocytes to death stimuli in vivo. Both events precede tumor formation, suggesting that suppression of apoptosis may contrib ute to the selection of a resistant phenotype, eventually increasing the probability of cell progression to the transformed state. T he role of mitochondria in cell death is being increasingly recognized (1). Mitochondrial dysfunction due to a perme ability transition can precipitate a bioenergetic crisis with ATP depletion and Ca2+ dysrgulation (2-4). On the other hand, mitochondria can release proteins that cause cell death through both caspase-dependent and caspase-independent mechanisms (5-9). Release of these apoptogenic factors is modulated by members of the BCL-2 family of proteins in a way that is consistent with their role in apoptosis. Antiapoptotic members inhibit (8, 10, 11) whereas proapoptotic members favor the release (12-16) in a process that is initiated by insertion of truncated BID in the outer mitochondrial membrane (17-19) and/or by the mitochondrial permeability transition (8, 16, 20-22). Defects of apoptosis are critical to both tumorigenesis and drug resistance (23). An involvement of mitochondria in the resistance of cancer cells to apoptosis is supported by many in vitro studies (24), but whether this holds true in vivo remains to be established. The early formation of drug-resistant hepatocytes during feeding with the hepatocardnogen 2-acetylaminofluorene (AAF) is well documented (25), but the basis for resistance and its potential role in tumorigenesis remain obscure. Metabolic activation of AAF is a prerequisite for the manifestation of both genotoxic and nongenotoxic effects (26). Previous work has shown that 2-nitrosofluorene, a metabolite of AAF, undergoes redox cycling after reduction by the mitochondrial respiratory chain (27) and is able to trigger opening of the permeability transition pore (PTP) (28). Yet, liver mitochondria isolated from rats fed with AAF were strikingly resistant to PTP opening well before the animals developed liver cancer, leading Neumann and Coworkers to propose that mitochondrial adaptation may play a role in the selection of resistant hepatocytes (28). The present study was designed to investigate whether mito chondrial resistance is matched by an early inhibition of cell death regulation ex vivo and in vivo. We confirmed that within 2 weeks of feeding with a diet containing 0.01-0.04% (wt/wt) AAF, liver mitochondria developed an increased resistance to opening of the PTP (28). Consistent with a mitochondrial adaptive response in vivo, hepatocytes isolated from AAF-fed rats became resistant to PTP-dependent depolarization, cyto chrome c release, and cell death. AAF-fed rats were fully protected from the hepatotoxic effects of the injection of Escherichia coli lipopolysaccharide (LPS) plus D-galactosamine (D-GalN) (29), a treatment that in control rats readily caused a large increase of terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL)-positive cells in liver cryo sections and release of alanine and aspartate aminotransferase into the bloodstream. Treatment with LPS plus D-GalN caused BID cleavage irrespective of AAF feeding, whereas caspase 3 was cleaved only in animals fed a control diet. These results demonstrate that mitochondrial adaptation is an early effect of AAF feeding, which is likely to cause the increased resistance of hepatocytes to death stimuli in vivo. Materials and Methods Chemicals and Antibodies. LPS from E. coli serotype 011LB4, D-GalN hydrochloride, collagnase type IV, collagen type I, valinomycin, and AAF were purchased from Sigma. Calciumgreen-5N and tetramethylrhodamine methyl ester (TMRM) Abbreviations: AAF, 2-acetyiaminofluorene; D-GaiN, D-galactosarnine; LPS, lipopclysaccharide of Escherichia coii; PTP, permeability transition pore; TMRM, tetrarriethyirhodarriine methyl ester; IN F, tumor necrosis factor; TUNEL, terminai deoxynucieotidyltransferasemediated dUTP nick end-labeling, 1P.-C.K. and M.E.5. contributed equally to this work. ` Present address: MRC Prion Unit. Department of Neurodegenerative Diseases, Institute of Neurology, University College London, Queen Square, London WC1N 3EG, United Kingdom. Il'To whom correspondence should be addressed at: Dipartimento di Scienze Biomediche Sperimentali, Viale Giuseppe Colombo 3, 1-35121 Padua, Italy. E-mail: bernardi blo.unlpd.it. 10014-10019 | PNAS | August 19,2003 | vol. 100 j no. 17 www.pnas.org/cgi/doi/10.1073/pnas.1633614100 Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039347-00001 $ were purchased from Molecular Probes. The mouse IgGl anti body directed against BCL-2 was purchased from Transduction Laboratories (Lexington, KY). A rabbit polyclonal antiserum raised against the core subunit 2 (Core-2) of the mitochondrial bci complex was kindly provided by Hermann Schgger. Anti bodies against cytochrome c (clone 6H2.B4) and BID were from PharMingen, the antibody against cleaved caspase 3 was from Cell Signaling (CELBIO, Milan), and the antibody against actin was from Sigma. The rabbit polyclonal antibody raised against the rat bc\ complex and the GD3 gangiioside were generous gifts of Roberto Bisson and Fidia Research Laboratories (Abano Terme, Italy), respectively. B Ca*2* GD3 [A A :0.05 ,,1 min:, C ar;- [uM] VaS Val AAQ5 ^ 3 \ b 1 min Animals. Male albino Wistar rats (180-200 g) had free access to a standard diet (Altromin 1324, Altrogge, Lage/Lippe, Ger many) or standard diet supplemented with 0.01-0.04% AAF. In vivo treatment was essential because concentrations up to 1 mM AAF had no effects on the FTP and on respiration when added directly to isolated mitochondria (results not shown). Animals were kept under controlled conditions of temperature and humidity on a 12-h light/12-h dark cycle. With the exception of Fig. ID, experiments on mitochondria, hepatocytes, or living animals were carried out after 3-4 weeks of feeding with AAF or control diet. At this time point, mitochondrial PTP resistance was already significant (see Fig. ID), whereas toxic effects of AAF-feeding were not apparent as assessed by monitoring the body weight, which did not show a significant difference relative to control animals (30). All preparations and treatments were always carried out in parallel for the two sets of feeding conditions. Assays on Isolated Mitochondria. Rat liver mitochondria were isoiated as described (27), Mitochondrial volume changes were determined from absorbance changes at 540 nm with an Aminco DW2000 spectrophotometer (SLM, Urbana, IL). The Ca2+ retention capacity of mitochondrial preparations was assessed fluorimetrically in the presence of the Ca2H indicator calciumgreen-SN with a Perkin-Elmer LS50B spectrofluorimeter ex actly as described (31). All instruments were equipped with magnetic stirring and thermostatic control. The incubation conditions are specified in the figure legends. Fsg. 1. Effect of A A F feeding on m itochondria! Ca2i loading capacity and on PTP-dependent and valinomycin-dependent swelling. The incubation me dium contained 200 mM sucrose, 10 mM Tris-M ops, 5mM succinate/Tris, 1 mM Pj/Tris, 10 M EGTA/Tris, and 2 /xM rotenone. Final volum e w as 2 ml, pH 7.4, 25C. Ail experiments w ere started by the addition (not shown) of 0.4 mg x m i-1 liver m itochondria prepared from rats fed either a control diet (traces a in all paneis) or a diet containing 0.04% A A F (traces b in all panels) fo r 3-4 w eeks. (A) The medium w as supplem ented w ith 1 M calcium-green-SN, and the indicated concentrations o f Ca2; w e re added at the tim e poi nts marked by arrow s. (8) W here indicated 50 /xM Ca2f and 40 /xM GD3 gangiioside w ere added. ( 0 The medium w as supplem ented w ith 1 mM K G , and 1 /xM valinomycin (Val) was added w here indicated by the arrows. These experiments are representative of 10-45 replicates each. FTP adaptation (i.e ., increased resis tance to FTP opening relative to controls in protocols simi iar to those o f B) was observed in m itochondria isoiated from 45 o f 50 AAF-fed anim als. (D) Male W istar rats w ere fed a diet containing 0.04% (a ), 0.02% ( ), or 0.01% (#) AAF. A t the indicated tim e points, rats w ere killed and liver mitochondria w ere isoiated. Osmotic swelling o f m itochondria (1 mg x m l-1) was determ ined Western Blot Analysis. For the experiments o f Fig. 2, rat liver mitochondria were lysed in a buffer containing 50 mM NaF, 40 mM sodium pyrophosphate, 20 mM Tris-HCl (pH 8.0), 10 mM EGTA, 5 mM MgC!2, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM phenyhnethylsulfonyl fluoride, 20 fig X mi 1aprotinin, and 20 jxg X m i-1 ieupeptin for 1 h at 4C, and the lysates were cleared by centrifugation at 14,000 X g for 10 min at 4C. For the experiment of Fig. 4E, liver homogenates were extracted with 0.3 M mannitol/5 mM Tris-Mops (pH 7.4)/4 mM KH2PO4/I mM EGTA/2 mM phenylmethylsulfonyl as shown in 8, except th at the concentration o f Pi/Tris was 15 mM and th e C a 21 addition, w hich was sufficient to trigg er mitochondrial swelling in th e control m itochondria, was 80 fiM . Data are expressed as inhibition o f swelling rates, w h ere 0 is the rate o f control m itochondria (corresponding to A A /A t values of 249 9, 201 34, 210 82, and 232 64 x 10 3 X min 1fo r m itochondria from w eeks 2 ,4 ,8 , and 16, respectively) and 1 is the inhibi tion observed in the presence o f 120 nM cyclosporin A (corresponding to A A /A t values o f 4 2 ,3 2, 1 x 1, and 10 4 X 10"3 X m in -1 fo r m itochondria from w eeks 2 ,4 , 8, and 16, respectively). The rate o f swelling w as determ ined by linear regression of the Asm versus tim e diagram. Data represent mean values of four animals, and curves w ere fitted by interpolation. Standard deviations were omitted for clarity and did not exceed 0.1. Swelling rates fo r mi tochondria from AAF-fed animals were significantly d ifferent from control values for ail tim e points (P C 0.001, Student's ( test). Kihn et ai. PNAS i August 19,2003 | voi. 100 | no, 17 | 10015 Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039347-00002 A Control Bd2 AAF AP feeding [weeks] 2 4 8 18 .......... f >:............................ i !........................................t r------ Core-2 B 16 streptomycin, 200 nM dexamethasone, 1 nM insulin, 12.5 mM Hepes, and 25 mM bicarbonate (final pH 7.2). After 16 h the culture medium was replaced by serum-free M199 medium supplemented with 100 units X ml'"1 penicillin, 100 xg X ml'"1 streptomycin, 12.5 mM Hepes, and 25 mM bicarbonate (pH 7.2), and the experiments were carried out as further described in the following paragraph. Experiments with isolated Hepatocytes. The mitochondrial mem brane potential was assessed by epifluorescence microscopy based on the mitochondrial accumulation of TMRM. Briefly, hepatocytes were preincubated with 20 nM TMRM and 1.8 lM cyclosporin 11in Hanks' balanced salt solution (HBSS) contain ing Ca2+ and Mg2+ (pH 7.4) for 20 min at 37C, and changes of mitochondrial TMRM fluorescence were monitored with the Olympus 1MT-2 inverted epifluorescence microscope equipped with a cooled charge-coupled device (CCD) camera exactly as described (33). The distribution of cytochrome c was determined exactly as described in ref. 34. Cell death was determined based on staining with propidium iodide. Cells were washed three times with HBSS, incubated with 2 aM propidium iodide in HBSS for 20 min, rinsed, and immediately analyzed with the Olympus IMT-2 epifluorescence microscope. PB Control A A F feeding [w eeks] Fig. 2. Effect o f AAF feeding on m itochondria! BCL-2 expression. (A) The levels o f expression o f BCL-2 and o f com plex !ll subunit Core-2 w ere deter mined by Western blotting of isolated mitochondrial proteins w ith antibodies against BCL-2 and Core-2. Anim als w e re fed either a control diet or a diet containing 0.02% AAF for the indicated length of tim e, and each lane was loaded w ith mitochondria! proteins from one Individual anim al. (8) Protein expression levels w e re quantified as described in M aterials an d M e th o d s, and the levels o f expression w ere norm alized to control values. Data represent m ean values SD o f three to five anim als fo r each condition. fluoride/10 fig X m l-1 aprotinin/1% Nonidet P-40 for 30 min at 4C, and the lysates were centrifuged at 90,000 X g in a Beckman L7-55 ultracentrifuge. The extracts were diluted with 2X Laemmli gel sample buffer and boiled for 3 min. Identical protein amounts were separated by SDS/PAGE, electroblotted onto Immobilon-P poly(vinylidene difluoride) membranes (Millipore), and incubated with the antibodies as specified in the figure legends. For the experiments of Fig. 2, blots were scanned with the TLC Scanner 3 (Camag, Berlin) and quantified with the manufacturer's software, c a t s 4.04. Preparation of Hpatocytes. Hepatocytes were isolated by colla gnase perfusion essentially as described by Berry and Friend (32). Hepatocytes were purified through a 40% (vol/vol) Percoll/Krebs-Henseleit Hepes buffer solution and rinsed free of Percoll with M199 medium supplemented with 24.7 mM Hepes and 25 mM bicarbonate (pH 7.2). Hepatocyte integrity was assessed by trypan blue exclusion and found to be always >85%. Cells were plated in six-well tissue culture plates coated with 0.01% (wt/vol) collagen type I (5 X 10s cells per well). The culture medium was M199 supplemented with 5% heatinactivated FBS, 100 units X ml-1 penicillin, 100 eg X ml-1 Treatment with IPS and D-GalN. Rats were treated with a single i.p . injection of 20-30 xg of LPS plus 700 mg of D-GalN X kg'"1of body weight. After 8 h blood samples were withdrawn from the tail vein, animals were killed by cervical dislocation, and livers were isolated for TUNEL. The TUNEL reaction was carried out on 10 /mi-thick liver cryosections with an in situ cell death detection kit (Roche Molecular Biochemicals, Milan). Sections preincubated for 30 min with 5 eg X ml-1 DNase I (GIBCO/ BRL) served as positive controls. An aliquot of serum was used to determine the concentration of tumor necrosis factor (TNF) -a, whereas the remainder was stored at --80C until determination of serum enzymes was performed. TNF-a was determined by using a rat TNF-a immunoassay (Quantikine M, R & D Systems) according to the manufacturer's instructions. Serum levels of alanine aminotransferase and aspartate amino transferase were determined according to standard procedures. Statistical Procedures. Paired, two-tailed Student's l test was used to determine the statistical significance of differences between sample mean. P < 0.05 and P < 0.01 are denoted with one or two asterisks, respectively. Results In the experiments of Fig. L4, we tested the mitochondrial Ca2H-retention capacity, a sensitive measure of the propensity of mitochondria to open the PTP after Ca2^ uptake. Mitochondria from animals fed a control diet accumulated "125 nmol of Ca2+ per mg of protein before PTP opening, which can be identified by release of the accumulated Ca2+ (trace a). Mitochondria isolated from rats fed a diet containing 0.04% AAF for 3 weeks were more resistant, in that nearly twice as much Ca2+ was needed to trigger opening of the PTP (trace b). Mitochondria from AAF-fed rats became more resistant to other PTPinducing stimuli. Fig. IB shows that after the uptake of 100 nmol of Ca2+ per mg of protein (a load that was not sufficient to cause PTP openingperse), addition of GD3 ganglioside readily caused swelling of mitochondria from control rats (trace a). As expected of a PTP-dependent event, swelling was fully inhibited by cyclosporin A (results not shown, but see ref. 35). No swelling was observed after addition of GD3 ganglioside to mitochondria from AAF-fed rats (trace b). Similar results were obtained when mitochondria were treated with other PTP-inducing agents (acetoacetate, arachidonic acid, diamide, menadione, 2-nitrosofluorene, or uncoupler; results not shown). In contrast, 10016 | vwvw.pnas.org/cgi/doi/10.1073/pn3s.1633614100 Kihn eta!. Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039347-00003 PTP-independejnt mitochondrial swelling by the addition of the K + ionophore valinomycin (Fig. 1C) caused an identical absor bance decrease in mitochondria from control and AAF-fed animals (traces a and h, respectively). Half-maximal FTP inhi bition was already detectable after 2 weeks of treatment with the lowest AAF dose (0.01%) and reached a peak between 4 and 8 weeks (Fig. ID). The inhibitory effect was transient, a decrease being detected with all AAF doses between 8 and 16 weeks of treatment, i.e., a time period that is characterized by a marked increase of focal lesions in vivo (30, 36). FTP resistance was matched by expression of mitochondrial BCL-2, w'hieh increased with the same time course as FTP adaptation (Fig. 2 A and B). We next studied the responses of the FTP in situ in primary cultures of hepatocytes prepared from control and AAF-fed rats. In the experiments of Fig. 3A, hepatocytes were challenged with 200 jxM arachidonic acid, a cytotoxic concentration (21, 37) that readily caused mitochondrial depolarization as measured by the release of intramitochondrial TMRM in hepatocytes from rats fed a control diet (b, compare with a). A striking resistance to mitochondrial depolarization was observed in hepatocytes pre pared from AAF-fed rats (d, compare with c). The vast majority, but not all, hepatocytes from AAF-fed rats were resistant to arachidonic acid-induced mitochondrial depolarization. Indeed, the hepatocyte marked by the arrow' in c and d underwent a mitochondrial depolarization that was indistinguishable from that of hepatocytes prepared from control animals. In Fig. 3B, the time course of the fluorescence changes induced by arachi donic acid in the two populations of hepatocytes is compared, confirming resistance to FTP opening in situ in hepatocytes prepared from AAF-fed rats. Control hepatocytes treated with arachidonic acid readily released mitochondrial cytochrome c (Fig. 3C) and became positive for nuclear staining with pro- pidium iodide (Fig. 3D), whereas hepatocytes from AAF-fed rats retained cytochrome c in the mitochondrial compartment and remained impermeable to propidium iodide (Fig. 3 C and D, respectively). To test whether AAF-feeding also conferred resistance to liver cytotoxic stimuli in vivo, we injected rats with a single bolus of LPS plus D-GalN, a well established protocol to induce acute hepatotoxicity through TNF-a release (29, 38). After 8 h of treatment with 20 ;xg of LPS plus 700 mg of D-GalN X kg-1 of body weight, liver cryosections revealed a sizeable number of TUNEL-positive hepatocytes in control but not in AAF-fed animals (Fig. 4x1 b and e, respectively). TUNEL positivity in control rats correlated with a significant number of apoptotic bodies in hematoxylin/eosin-stained slices, which were rarely detectable in slices from AAF-fed animals (data not shown). Inflammation, a frequent event in control rats that is indicated by lymphocyte infiltration of liver parenchyma, was absent in Fig, 3. Effect of AAF feeding on PTP-dependent mitochondrial depolariza tion, cytochrome c release, and cell survival after treatm ent of primary cul tures o f hepatocytes w ith arachidonic acid. Hepatocytes w e re isolated by collagenase perfusion as described in M aterials an d M e th o d s from rats fed either a control diet or a diet containing 0.04% AAF for 4 weeks, and kept 24 h in the CO2 incubator. (A) Hepatocytes w e re loaded fo r 20 min w ith 20 nM TMRM and transferred to the microscope stage. Shown are fluorescence im ages (m agnification ><400} o f hepatocytes from control (a and b ) and AF-fed anim ais (c and d) before (a and c) and 15 min afte r the addition of 200 pM arachidonic acid (b and d). (B) Time course of the mitochondrial fluorescence changes afte r the addition of 200 ,xM arachidonic acid (Ara) in hepatocytes from control () and AAF-fed (O) rats. Values {norm alized to the initial fluorescence) report th e mean fluorescence SD as determ ined in four independent determinations on the same preparations of hepatocytes and are representative o f four separate experim ents. W here indicated, 1 /.M carbonyl cyanide p-trifluorom ethoxyphenylhydrazone (FCCP) was added. (C and D) Two hundred micromolar arachidonic acid or vehicle was added to the hepatocytes, and the incubation was continued fo r a fu rth er 30 min la tim e point at w hich the cytochrom e c distribution was determ ined (C)] or 60 min [a time point at which nuclear staining w ith propidium iodide was determined (D)]. For both C and D, addition o f vehicle or 200 /xM arachidonic acid is denoted by open and filled bars, respectively. A lower localization index corresponds to a more di f fuse distribution o f cytochrom e c relative to the bc\ complex; see ref. 34 fo r details. Values refer to the m ean SD o f four independent determinations on the same preparations of hepatocytes and are representative o f th ree (cytochrome c) or fo ur {propidium iodide) separate exp erim ents. Kiohn et al. PNAS i August 19,2003 | vol. 100 | no, 17 | 10017 Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039347-00004  5 y * ;v > - * 2JG ' . . . .. ^ #' <** B D Control diet 1 3 BID "w AAF diet tSlD Ht:...... Casp3 Aetin liiB ilii liiii| M p H W i LPS)-6alN ++ - + Fig, 4, Effect of A A F feeding on liver TUN EL staining, serum enzym e release, and B!D and caspase 3 cleavage afte r treatm ent w ith IPS plus o-GalN in vivo. Rats fed a control diet o ra diet containing 0.04% AAF for 3 weeks w ere treated w ith a single i.p. injection o f vehicle (isotonic saiine) or w ith LPS plus D-GaiN. A fte r 8 h bjood w as w ith d raw n from th e tail vein to determ ine plasma levels of alanine am inotransferase (ALT) and aspartate am inotransferase (AST). Extracts from liver homogenates and liver cryosections were prepared and analyzed by W estern blotting and TUN EL staining, respectively (see M aterials and M ethods for details). (A) M agnifications of X50 of TUNEL-stained cryo sections from control animals (a-c) and AAF-fed animals (d - f ) treated w ith saline (a and d), or w ith 20 /xg o f LPS (b and e) or 30 /xg o f LPS plus 700 mg of D-GaiN x kg ' 1 o f body w eig h t (c and f). (B) Percentage of TUNEL-positive hepatocytes in control (filled bars) or AAF-fed (open bars) anim als afte r treatm en t w ith saiine ( -) or 20 jag o f LPS plus 700 mg of D-GaiN x kg ' 1of body w eight (+). For each condition nuclear labeling was determined from 20 high-power fields, which corresponds to a total of 1,200-1,500 hepatocytes. Error bars denote the SD of six (treatm ent w ith LPS plus D-GaiN) or four (treatm ent w ith saiine) independent experim ents. (C and D) Plasma levels of A LT and A5T, respectively, in anim als treated w ith a control diet (filled bars) or A A F diet (open bars) afte r injection of saiine ( -) or 20 jag of LPS pius 700 mg o f D-GalN x kg -1 o f body w eig h t (%). Error bars denote the SD o f five independent experim ents. () W estern biot analysis o f BID, truncated BID (tBiD), cleaved caspase (Gasp) 3, and actin in liver extracts from control-fed (lanes 1-3) and AAF-fed (lanes 4 - 6 ) anim ais; samples in ianes 1 and 4 w ere from anim als treated w ith saiine, w hereas samples in lanes 2, 3, 5, and 6 w ere prepared 8 h a fter treatm en t w ith 20 /xg o f LPS plus 700 mg o f D-GaiN x kg ' 1 of body weight. The same biot was sequentially probed w ith all antibodies after stripping, and exposure times were adjusted to the intensity of the signals. AAF-fed rats (data not shown). We also treated one pair of rats with 30 /xg of LPS plus 700 mg of D-GaiN X kg 1of body weight, a dose that caused diffuse hepatonecrosis in the control (Fig. 4Ac) but not in the AAF-fed animal (Fig. 4Af). The graph of Fig, 4B documents the protection obtained in AAF-fed rats chal lenged with 20 /xg of LPS plus 700 mg of D-GalN X kg-1 of body weight. We measured the changes of TNF-a after the addition of the same amounts of LPS plus D-GalN in five AAF-fed and five control-fed rats. The peak TNF-a serum levels displayed the following changes (fold increase over the basal level): 9.1 4.8 for the group of control-fed animals; and 9.1 2.9 for the AAF-fed animals. The results obtained with TUNEL staining were consistent with serum alanine and aspartate aminotrans ferase levels measured 8 h after administration of LPS plus D-GalN (Fig. 4 C and D). Apical caspases were activated in both control and AAF-fed rats, as shown by cleavage of BID to truncated BID (tBID), whereas caspase 3 was cleaved only in control-fed animals (Fig. 4E). Note that a small amount of tBID could be detected in AAF-fed animals even in the absence of LPS plus D-GalN. Thus, resistance to LPS cytotoxicity is down stream of BID cleavage. To investigate whether discontinuation of the AAF diet would reverse the inhibition of the permeability transition, a group of four animals was fed a diet containing 0.04% AAF for 4 weeks, followed by control diet for a further 12 weeks. Measurements of FTP opening at this time indicated that inhibition of the permeability transition had been relieved by 80% in protocols identical to those described in Fig. ID, and the animals regained sensitivity to hepatotoxicity after injection of LPS plus D-GalN (data not shown). Discussion An involvement of the mitochondrial permeability transition in the regulation of cell death has been demonstrated in several in vitro systems (39), but remains to be established in vivo, where evidence is limited to animal models of stroke (40, 41). In the present study we addressed the question of whether an increased resistance to PTP opening plays a role in carcinogenesis in a rat liver model. We confirmed that mitochondria isolated from livers of AAF-fed rats are strikingly resistant to Ca2+-dependent PTP opening well before the clonal expansion of focal lesions (28). Adaptation appears to be specific for the PTP, because the response of liver mitochondria from AAF-adapted rats is indis tinguishable from that of control mitochondria when valinomycin-dependent K + uptake rather than PTP opening is used to induce swelling. These findings demonstrate that inhibition of swelling does not depend on mitochondrial structural constraints or on differences in formation of the proton electrochemical gradient, issues that had not been addressed in previous studies. Increased resistance to PTP opening was matched, and pos sibly caused, by an increased mitochondrial expression of BCL-2. This is at apparent variance with a previous study on mitochon dria from transgenic mice overexpressing BCL-2 in the liver, where no difference in PTP opening could be detected despite in vivo resistance to hepatotoxicity by anti-Fas antibodies (42). We note that the increased PTP resistance to Ca2+ is not an all-or-nothing event (Fig. 1) that can therefore be missed in protocols based on a single, large Ca2+ load (42), and that our data are in keeping with results obtained in neural ceil lines where BCL-2 overexpression promoted an increased mitochon drial Ca2+ retention capacity (43). PTP adaptation can also be detected in primary cultures of hepatocytes challenged with arachidonic acid, a potent inducer of cell death through the PTP (21). We assume that resistance to pore opening is an adaptation to reactive species generated by AAF metabolism in the liver. These species include 2-nitrosofluorene, which drains electrons from the respiratory chain at the level of the bc\ complex, undergoes redox cycling with produc tion of superoxide anion, and is a potent inducer of PTP opening (27, 28, 44). We think that up-regulation of BCL-2 may be a response to the increased oxidative stress (45), which would cause inhibition of the PTP and result in a decreased propensity to the release of cytochrome c and other apoptogenic factors (8). 10018 I www.pnas.org/cgi/doi/10.1073/pnas.1633614100 Kihn eta!. Sierra Club v. EPA 18cv3472 NDCA Attachments Prod 1 ED 002061 00039347-00005 It is remarkable that the opposite phenotype, i.e., an increased sensitivity to FTP opening ex vivo and increased liver apoptoses were observed in the Sod2+/ mice, which have a decreased ability to cope with oxidative stress (46). Our demonstration that AAF feeding induces an early resistance to the otherwise hepatotoxic treatment with LPS plus D-GalN clearly indicates that an anti-apoptotic response is initiated by AAF in vivo. Feeding with AAF did not affect the TNF-a response to LPS plus D-GalN and the formation of tBID, which depends on activation of apical caspases, but it did completely prevent the activation of caspase 3. These data indicate that in this model of hepatotoxicity cell death strictly depends on the mitochondrial pathway, and that the latter is selectively suppressed by AAF feeding. The ability to evade apoptosis is an essential hallmark of both tumorigenesis and drug resistance (23), and the early appearance of hepatocytes resistant to apoptosis during A AF carcinogenesis may substantially promote the accumulation of mutations. The inhibition of cell demise would in turn interfere with the DNA damage response pathways, leading to the accumulation of unrepaired DNA lesions, genomic instability, and malignant progression (47, 48). In other words, the selection of resistant cells would represent a tumor-promoting property of AAF, and our data indicate that resistance is established through an epigenetic rather than a mutagenic process. First, resistance was observed before the clonal expansion of preneoplastic cells (30), suggesting that cells acquire resistance to cell death stimuli before the selection of critical mutations in genes of growth regulating proteins, and before their transformation into auton omously growing cancer cells (49). Second, increased resistance was fully reversible after returning to the control diet, a feature that is characteristic of epigenetic processes, whereas a pheno type generated by mutations is presumed to be irreversible (50, 51). In summary, our results support a sequence of events in which the chronic challenge to the liver by AAF metabolites triggers epigenetic responses that include mitochondrial FTP adapta tion. The benefit of resistance to AAF toxicity would be out weighed by the persistence of mutated cells, eventually resulting in the formation of tumors. It will be interesting to identify the relevant changes by comparing the protein expression profiles of liver mitochondria from control and AAF-fed rats. We thank Dr. Martina Zaninotto (Department of Laboratory Medicine, Padua University Hospital) for the determination of serum enzymes, Prof. Hermann Schgger for generously providing Core-2 antibody, and Ray Young for expert help with graphics. This work is in partial fulfillment of the requirements for the Ph.D. degree of M.E.S. It was supported in part by grants from the Ministero per l'Universit e la Ricerca Scientifica e Tecnologica "1 mitocondri nella fisiopatologia cellulare: meccanismi patogenetici e sintesi chimica di nuovi farmaci," the Associazione Italiana per la Ricerca sul Cancro, and the ArmeniseHarvard Foundation (to P.B.). W.I. was supported by a Fellowship from Teethon-Itay (Grant 1141 to P.B.). P.-C.K. gratefully acknowledges the Deutscher kademiseher Austauschdienst (DAAD), the Forschungszentrum Karlsruhe, and the Armenise-Harvard foundation for support during his visits to Padua. 1. Bernard!, P., Petronilli, V., D i lis a , F. St Forte, M. (2001) Trends Biochem. Set. 26, 1.12-117. 2. Im berti, R., Nieininen, A. L,. H erm an, B. & Lem asters, J. J. (1993) J. Phar macol. Exp. Ther. 265, 392-400. 3. D uchen, M. R., M cG uinness, O., Brown, L. A, & C rom pton, M. (199.3) Cardiovasc. Res. 27, 1790-1794. 4. Pastorino, J. G., Snyder, I. W., Serroni, A., H oek, I. B. & Farber, I. L. (1993) /. Biot. Chem. 268, 13791-13798. 5. Liu, X., Kim. C, N., Yang. J., Jemmerson, R. & Wang, X. (1996) Cell 86,147-157. 6. Du, C,, Fang, M , Li, Y , Li, L. & W ang, X. (2000) Ceil 102, 33-42. 7. Bkert, P. G , Siike, j,, Connolly, L, M,, Reid, G, E , M oritz, R. L. St. V atu , D, L. (2000) Ceil 102, 43-53. 8. Susin, S. A., Zamzami, N , Castedo, M , H irsch, T , M archetti, P,, M acho, A , Dartgas, E., Geuskens, M. St Kroem er, G. (1996)7. Exp. Med. 184, 1331-1341. 9. Li, L, Y., Luo. X. & W ang, X . (2001) Nature 412, 95-99. 10. Yang, .1., Lilt, X , Bhaila, K , Kirn, C. N., ibrado, A , M , Cai, J., Peng, 'P, L, Jottes, D. P. & Wang, X. (1997) Science 275, 1129-1132. 11. Kluck, R. M,, Bossy-W etzel, E , G reen, D . R . & Newmeyer, D . D . (1997) Science 275, 1132-1136. 12. Cosulich, S. C,, W orrali, V., Hedge, P. J., G reen, S. & Clarke, P, R, (1997) Cutr. Biol. 7, 913-920. 13. M anon, S., Chaudhuri, B. & G uerin, M. (1997) FE ES Lett. 415, 29-32. 14. Jurgensm eier, J. M., Xie, Z., Deveraux, Q., Ellerby, L , B redesen, D . & R eed, J. C. (1998) Proc. Nail. Acad. Sci, USA 95, 4997-5002. 15. Eskes, R., A ntonsson, B , O sen-Sand A , M ontessuit, S.. Richter. C , Sadoui, R , M azzei, G , Nichols, A. & M artlnou, J.-C. (998) J. Cel! Biol. 143, 217-224. 16. Pastorino, J. G ,, Chen, S. T., Tafani, M , Snyder, I. W . & Farber, J. L. (1998) J. Biol. Chem. 273, 7770-7775. 17. Li, H , Zhu, H , Xu, C. J. & Y uan, J. (1998) Cell 94, 491-501. 18. Luo, X.. Budihardjo, L, Zou, H., Slaughter, C. & Wang, X. (1998) Cell 94,481-490. 19. Gross. A.. Yin, X, M., W attg, K., W ei, M. C., Jo c k e l J., M ilhmatt, C., E rdjunient-B rom age, H ,, Tem pst, P. & K orsmeyer, S. J. (1999) J. Biol. Chem. 114, 1156-1163. 20. B radham , C. A., Q ian, T., Streetz. K., Trautwein, C., B renner, D . A . St Lem asters. J. J. (1998) Mol. Ceil. Biol. 18, 6353-6364. 21. Scorrano, L., Penzo, D,, Petrortiili. V., Pagano, F, & Berttardi, P. (2001)7. Biol. Chem. 276, 12035-12040. 22. Scorrano, I,., Ashiya, M., Buttle, K,, W eiler, S., O akes, S. A., M annella, C. A. & K orsmeyer, S. J. (2002) Dev. Ceil 2, 55-67. 23. Johnstone, R . W .. Ruefli, A . A. & Lowe. S. W . (2002) Cell 108, 153-164. 24. Igney, F. II. & Kram.rn.er. P. H. (2002) Nat. Rev. Cancer 2, 277-288. 25. C arr, B. I. (1987) Cancer Res. 47, 5577-5583. 26. Heflich, R. H. & Neft, R . E. (1994) Muted. Res. 318, 73-114. 27. Klohn, P. C., Massaih:: Fi. St. Neumann, IL G, (1995) Biochirn. Biophys. Acta 1229, 363-372. 28. Klohn, P. C.. Bitsch, A. ,7 N eum ann, H . G . (1998) Carcinogenesis 19, 1185-1190. 29. Leist, M., G antner, F., Bohiinger, L, 'Piega, G., G erm ana. P, G . &. W'endel, A. (1995) A m . J. Pathol. 146, 220-1234. 30. Bitsch, A., Hadjiolov, N., K lhn, P. C., Bergm ann, O., Zw irner-B aier, I. & N eum ann, H . G. (2000) Toxicol. Sci. 55, 44-51. 31. Fontaine, E., Eriksson, O., Ichas. F. St. B ernardi, P. (1998) 7. Biol. Chem. 273, 12662-12668. 32. Berry. M. N. & Friend, D, S. (1969) 7. Cell Bioi. 43, 506-520. 33. Petronilli, V., M iotto, G ,, C anton, M., C olonna, R., Bernardi, P. & D i U sa, F. (1999) Biophys. J. 76, 725-734. 34. Petroniii, V., Penzo. D., Scorrano, L,, Bernardi, P. & D i Lisa, F. (2001) 7. Biol. Chem. 276, 12030-12034. 35. Scorrano, E., Petronilli, V., Di Lisa, F. & Bernardi, P. (1999)7. Biol. Chem. 214, 22581-22585. 36. Bitsch, A., Klohn, P. C., H adj iolov, N,, Bergm ann, O. & N eum ann, H . G . ( 1999) Cancer Lett. 143, 223-227. 37. Gugliucci, A., R anzato, E., Scorrano. L,, C olonna. R., Petronilli, V,, Cusan, C., Prato, M , M ancini, Mi, Pagano, F. St. Bernardi, P. (2002) 7. B io l Chem. I l l , 31789-31795. 38. W ang, I. H., R edm ond, H . P., W'atson, R . W. St Bouchier-Hayes, D . (1995) A m . J. Physiol. 269, G297-G 304. 39. Bernardi, P.. Scorrano, L., Colonna, R.. Petronilli, V. & D i Lisa. F. (1999) Eur. J. Biochem. 264, 687-701. 40. L, P. A , U chino, H , Fim er, E. & Siesjo, B. K. (1997) Brain Res. 753, 133-140. 41. Friherg, H., Ferrand-D rake, M., Bengtsson, F., H alestrap, A. P. St W ieloch, T. (1998) 7. Neurosci. 18, 5151-5159. 42. Yang, J. C.. Kahn, A . St C ortopassi. G. (2000) Toxicology 151, 65-72. 43. M urphy, A. N., Bredesen, D, E., Corsopassi, G ,, W ang, E. St Fiskurn, G , (1996) Proc. Natl. Acad. Set. USA 93, 9893-9898. 44. Klohn, P. C., B randt, U . & N eum ann, H . G . (996) FE ES Lett. 389, 233-237. 45. H ockenbery, D. M.. Oitvai, Z. N., Vii:. X. M., M iiliman, C. L, St Korsmeyer. S. J. (1993) Cell 75, 241-251. 46. K okoszka, .1, E., Coskim, P., Esposito, E. A . & W allace, D, C. (2.001) Proc. Nail. Acad. Sci. USA 98, 2278-2283. 47. Zhou, B. B. & Eliedge, S. J. (2000) .Nature 408, 433-439. 48. Skorski. T. (2002) Nat. Rev. Cancer 2, 351-360. 49. Haim. W . C. & W einberg, R . A. (2002) Nat. Rev. Cancer 2, 331-341. 50. W eber, E, & Bannasch, P. (1994) Carcinogenesis 15, 1235-1242. 51. Z.erhan, H,, Radig, S,, K opp-Sdm eider, A . & Bannasch, P. (1994) Carcino genesis 15, 2467-2473. Klhn et ai. Sierra Club v. EPA 18cv3472 NDCA PNAS j August 19,2003 | voi. 100 j no, 17 j 10019 Attachments Prod 1 ED 002061 00039347-00006