Cell surface changes and enzyme release during hypoxia and reoxygenation in the isolated, perfused rat liver.

We examined the effects of hypoxia and reoxygenation in isolated, perfused rat livers. Hypoxia induced by a low rate of perfusion led to near anoxia confined to centrilobular regions of the liver lobule. Periportal regions remained normoxic. Within 15 min, anoxic centrilobular hepatocytes developed surface blebs that projected into sinusoids through endothelial fenestrations. Periportal hepatocytes were unaffected. Both scanning and transmission electron microscopy suggested that blebs developed by transformation of preexisting microvilli. Upon reoxygenation by restoration of a high rate of perfusion, blebs disappeared. Other changes included marked shrinkage of hepatocytes, enlargement of sinusoids, and dilation of sinusoidal fenestrations. There was also an abrupt increase in the release of lactate dehydrogenase and protein after reoxygenation, and cytoplasmic fragments corresponding in size and shape to blebs were recovered by filtration of the effluent perfusate. We also studied phalloidin and cytochalasin D, agents that disrupt the cytoskeleton. Both substances at micromolar concentrations caused rapid and profound alterations of cell surface topography. We conclude that hepatic tissue is quite vulnerable to hypoxic injury. The morphological expression of hypoxic injury seems mediated by changes in the cortical cytoskeleton. Reoxygenation causes disappearance of blebs and paradoxically causes disruption of cellular volume control and release of blebs as cytoplasmic fragments. Such cytoplasmic shedding provides a mechanism for selective release of hepatic enzymes by injured liver tissue.

Centrilobular injury following hypoxia in isolated, perfused rat liver.

Hypoxia was produced in isolated, hemoglobin-free, perfused rat liver by reducing the flow rate of oxygen-carrying fluid entering the organ. The procedure caused anoxia in centrilobular regions. In these anoxic areas, structural derangements developed rapidly, characterized by bleb-like protrusions of hepatocyte plasma membrane through fenestrations in the sinusoidal endothelium. Periportal tissue remained normoxic and was completely spared. Cellular injury resulting from localized anoxia may play an important role in the pathogenesis of centrilobular liver disease.

Relationship between alveolar PO2 and the rate of p-nitroanisole O-demethylation by the cytochrome P-450 pathway in isolated rabbit lungs.

The relationship between alveolar PO2 and the rate of O-demethylation of p-nitroanisole, a model substrate for cytochrome P-450 -linked mixed-function oxidation, was evaluated in the isolated rabbit lung perfused with Krebs-Ringer bicarbonate buffer. The appearance of the product, p-nitrophenol, in the pulmonary perfusate was measured spectrophotometrically, The PO2 of the ventilating gas was varied with an accurate gas mixing pump and measured with an electrochemical O2 analyzer. In control lungs ventilated with 5% CO2 in air, the rate of p-nitrophenol production was approximately equal to 3.1 +/- 0.04 (mean +/- SE; n = 9) mumol/h per g dry wt. p-Nitrophenol production was unaltered when O2 in the ventilating gas was decreased to 1%, but it was depressed reversibly when alveolar O2 WAS 0.1% OR LESS AND WAS ABOLISHED DURING VENTILATION WITH 0.005% O2. The rate of the reaction was inhibited by 50% when alveolar PO2 was 0.3 mm Hg representing and intracellular [O2] OF approximately equal to muM. In the presence of metyrapone (0.1--1 mM), an inhibitor of cytochrome P-450-dependent reactions, p-nitrophenol production was 0.07--0.17 mumol/h per g dry wt. Ventilation of lungs with varying CO concentration in 20% O2 resulted in 50% inhibition of p-nitrophenol production when CO concentration was 10% (CO/O2 = 0.5). These results indicated that O-demethylation of p-nitroanisole by the lung is a cytochrome P-450-dependent reaction and that its rate is not affected until alveolar PO2 is less than 1 mm Hg.

NADPH oxidase-derived free radicals are key oxidants in alcohol-induced liver disease.

In North America, liver disease due to alcohol consumption is an important cause of death in adults, although its pathogenesis remains obscure. Despite the fact that resident hepatic macrophages are known to contribute to early alcohol-induced liver injury via oxidative stress, the exact source of free radicals has remained a mystery. To test the hypothesis that NADPH oxidase is the major source of oxidants due to ethanol, we used p47(phox) knockout mice, which lack a critical subunit of this major source of reactive oxygen species in activated phagocytes. Mice were treated with ethanol chronically, using a Tsukamoto-French protocol, for 4 weeks. In wild-type mice, ethanol caused severe liver injury via a mechanism involving gut-derived endotoxin, CD14 receptor, production of electron spin resonance-detectable free radicals, activation of the transcription factor NF-kappaB, and release of cytotoxic TNF-alpha from activated Kupffer cells. In NADPH oxidase-deficient mice, neither an increase in free radical production, activation of NF-kappaB, an increase in TNF-alpha mRNA, nor liver pathology was observed. These data strongly support the hypothesis that free radicals from NADPH oxidase in hepatic Kupffer cells play a predominant role in the pathogenesis of early alcohol-induced hepatitis by activating NF-kappaB, which activates production of cytotoxic TNF-alpha.

Peroxisome proliferators activate Kupffer cells in vivo.

The mechanism by which peroxisome proliferators increase cell replication and cause liver tumors in rodents remains unknown. When activated, Kupffer cells, the resident hepatic macrophages, release a variety of mitogenic stimuli that could theoretically increase cell proliferation in nearby hepatocytes. Therefore, in the present study we evaluated the effect of two potent peroxisome proliferators, nafenopin and WY-14,643, on Kupffer cell activation in vivo. Kupffer cell phagocytosis was determined continuously by monitoring rates of colloidal carbon uptake in the isolated, perfused liver after drug treatment in vivo. In the absence of peroxisome proliferators, colloidal carbon increased rates of oxygen uptake from 88 +/- 10 to 110 +/- 11 mumol/g/h. Livers from rats treated with either nafenopin (2-24 h) or WY-14,643 (24 h) were perfused for approximately 15 min with Krebs-Henseleit buffer and then with buffer containing colloidal carbon (2 mg/ml). Five h after nafenopin treatment (100 mg/kg i.g.), basal rates of colloidal carbon uptake of 136 +/- 12 mg/g/h were increased to 188 +/- 12 and remained elevated after 24 h (203 +/- 3 mg/g/h). Nafenopin also increased rates in a dose-dependent manner (one-half-maximal response, approximately 75 mg/kg). Similarly, WY-14,643 elevated rates of colloidal carbon uptake 1.8-fold over controls. Functional parameters of Kupffer cells were also affected. For example, WY-14,643 increased plasma nitrite significantly. This study demonstrates clearly that nafenopin and WY-14,643 activate Kupffer cell phagocytosis, suggesting a role for cell-to-cell communication in the stimulation of cell replication by peroxisome proliferators.

Ethanol potentiates doxorubicin-induced inhibition of cell survival in cultured Chinese hamster ovary cells.

Doxorubicin is an antineoplastic drug which undergoes oxidation-reduction cycling and produces toxicity to some cancer cell lines. Since oxidation-reduction cycling requires reducing equivalents and because ethanol metabolism via alcohol dehydrogenase (ADH) increases NADH, the effect of ethanol on doxorubicin toxicity was examined in cultured cells. Since some cells exhibit resistance to anthracyclines such as doxorubicin, two different Chinese hamster ovary cell lines were used, one sensitive (AUX B1) and one resistant (CHRC5) to doxorubicin. Studies were designed to determine if ethanol could decrease resistance to doxorubicin. Cells were treated for 24 h with doxorubicin in the presence or absence of ethanol, and the number of live cells was estimated spectrophotometrically. Ethanol (60-150 mM) potentiated the doxorubicin-induced decrease in cell number in both cell lines. In AUX B1 cells the concentration of doxorubicin required for half-maximal inhibition of cell survival was reduced 20-fold by ethanol, and a completely nontoxic concentration of doxorubicin decreased the number of surviving cells to 30% in the presence of ethanol. Addition of ethanol to the medium also increased doxorubicin-induced inhibition of cell survival in CHRC5 cells, but the effect was less dramatic than in AUX B1 cells. The effect of ethanol on cell number was concentration related; the half-maximal response was observed with about 1 mM ethanol. The hypothesis that ethanol potentiates doxorubicin toxicity by generation of NADH during metabolism by ADH was strengthened by the observations that both cell lines possess ADH activity (30-400 units/10(12) cells) and that ethanol (0.1-0.5 mM) increased NADH fluorescence 15-80% over basal values in cultured cells. Further, the effect of doxorubicin on cell number was also potentiated by another substrate for ADH, 2-ethylhexanol. Desferrioxamine, an iron chelator, increased survival in cells treated with doxorubicin plus ethanol by up to 60% (half-maximal effect, 1 mM), and (+)-catechin, a radical scavenger, abolished the decrease in cell number due to doxorubicin plus ethanol at concentrations greater than 0.1 mM. Allopurinol, an inhibitor of xanthine oxidase with radical scavenging properties, diminished the effect of doxorubicin plus ethanol on cell number by 60% (P less than 0.05). Taken together, these data are consistent with the hypothesis that ethanol potentiates toxicity due to doxorubicin by providing reducing equivalents for oxidation-reduction cycling which produce toxic reduced oxygen species.

Characterization of mutagenic glucuronide formation from benzo(a)pyrene in the nonrecirculating perfused rat liver.

Excretion of mutagenic metabolites of benzo(a)pyrene into bile from livers of corn oil- or 3-methylcholanthrene-treated Sprague-Dawley rats perfused with a nonrecirculating perfusion system was quantitated. Mutagenic benzo(a)pyrene metabolites were detected using Salmonella typhimurium (strain TA 98) grown in the presence of limiting amounts of histidine. Microsomes were not included in the bacterial assay since metabolic activation was carried out by the perfused liver. Mutagenic activity was detected only if beta-glucuronidase was added to the assay mixture or if bile was treated with acid to hydrolyze glucuronides prior to assay. When livers were perfused with 20 microM benzo(a)pyrene, stable, mutagenic glucuronides were exported from corn oil-treated livers at maximal rates of 149 +/- 24 (S.E.) revertants/g/hr and at rates of 225 +/- 22 revertants/g/hr in livers from 3-methylcholanthrene-treated rats. Chromatography of bile by high-performance liquid chromatography demonstrated that two peak areas contained phenolic glucuronides which were hydrolyzed by beta-glucuronidase. These two peaks, one which cochromatographed with authentic 3-benzo(a)pyrenyl-beta-D-glucuronide, accounted for all of the mutagenic activity in bile from livers perfused with benzo(a)pyrene. A good correlation (r = 0.86) between rates of mutagen production and rates of formation of phenolic glucuronides was observed under a variety of experimental conditions. The mutagenic activity observed with pure 3-benzo(a)pyrenyl-beta-D-glucuronide exposed to beta-glucuronidase was 4 revertants/nmol. When the rate of mutagen production was divided by the rate of production of 3-benzo(a)pyrenyl-beta-D-glucuronide by the perfused liver, a value of 4 revertants/nmol was also obtained. Therefore, it is concluded that mutagens exported in bile from livers perfused with benzo(a)pyrene can be accounted for predominantly by hydrolysis products of phenolic glucuronides.

Inhibition of glucuronidation of benzo(a)pyrene phenols by long-chain fatty acids.

Long-chain fatty acids inhibit glucuronidation of benzo(a)pyrene phenols in perfused liver; therefore, this study was designed to investigate interactions of fatty acids with beta-glucuronidase, glucuronosyl transferase, and energy supply. In beta-glucuronidase-deficient C3H/He mice, infusion of oleate (250 microM) increased the release of free benzo(a)pyrene phenols from 14 to 33 nmol/g/h and decreased release of glucuronides into the perfusate from 25 to 17 nmol/g/h. Rates of accumulation of glucuronides in the liver were also diminished from 11 to 4 nmol/g/h after infusion of oleate (250 microM). Fatty acids did not affect the release of benzo(a)pyrene metabolites into bile, and the ratio of free phenol to glucuronide production was increased from 0.57 to 1.30. A similar trend was observed in livers from DBA/2 mice that have beta-glucuronidase. Rates of hydrolysis of benzo(a)pyrene-O-glucuronide were not altered in isolated microsomes by addition of oleoyl coenzyme A (CoA) or octanoyl CoA (10- approximately 100 microM). Thus, we conclude that fatty acids do not alter glucuronidation by acting on beta-glucuronidase. The concentration of cofactors (UDP-glucuronic acid, UDP-glucose, and adenine nucleotides) involved in hepatic conjugation was not altered by infusion of concentrations of oleate (300 microM) that inhibited glucuronidation in perfused livers. When oleate concentrations were increased to 600 microM, UDP-glucuronic acid and UDP-glucose decreased 44 and 49%, respectively, and the ATP:ADP ratio declined concomitantly. Oleoyl CoA inhibited UDP-glucuronosyl transferase noncompetitively (half-maximal inhibition, 10 microM) in microsomes with 3-hydroxy-benzo(a)pyrene or p-nitrophenol as substrate. In contrast, octanoyl CoA was a very poor inhibitor of transferase activity. Inhibition of the transferase by oleoyl CoA was increased markedly by treatment with detergents (Triton X-100), i.e., half-inhibition of glucuronosyl transferase was obtained with about 2 microM oleoyl CoA. Inhibition of UDP-glucuronosyl transferase by oleoyl CoA was also increased in a dose-dependent manner by albumin, possibly due to increasing access of the CoA derivative to the enzyme. Collectively, these data indicate that fatty acids diminish glucuronidation via the formation of acyl CoA compounds that inhibit UDP-glucuronosyl transferase noncompetitively.

Microcirculation of hepatic nodules from diethylnitrosamine-treated rats.

The microcirculation of nodules (0.5 to 10 mm in diameter) from diethylnitrosamine-treated rats was studied in perfused livers. Microlight guides were placed on nodules and surrounding tissue on the capsular surface of the liver to measure fluorescence due to fluorescein-dextran (12 microM), a dye confined to the vascular space, infused via the hepatic artery and portal vein separately or simultaneously. The fluorescence increase due to fluorescein-dextran infusion via the artery and vein simultaneously was used to compare vascular space in nodules with that of surrounding tissue. The vascular space of nodules less than 1 mm in diameter was only about one-half as large as that of surrounding tissue. In contrast, in nodules 1 to 2 mm in diameter, the vascular space was similar to values from surrounding tissue. This was largely due to an increase in the fluid entering via the artery. As nodules grew from 2 to 10 mm in diameter, the vascular space decreased as a function of nodule size to 40% of surrounding tissue. The sum of fluorescence increases due to fluorescein-dextran infused via the artery and vein separately always equalled values obtained from simultaneous infusions. From these measurements, the fraction of vascular fluid observed by the microlight guide that entered the liver via the artery was calculated. In tissue surrounding nodules, fluid entering from the artery was 19% of the total, a value approximating the fraction of fluid pumped into the liver via the artery (25%). The percentage of fluid in the nodule that entered the liver via the hepatic artery increased progressively to 100% of the total as nodules grew from 2 to 10 mm in diameter. Thus, nodules become increasingly dependent on the hepatic artery and less dependent on blood supply via the portal vein as they grow.

Effect of bile salts on rates of formation, accumulation, and export of mutagenic metabolites from benzo(a)pyrene produced by the perfused rat liver.

The effect of sodium taurocholate on the biliary export of stable mutagenic phenolic glucuronide metabolites of benzo(a)pyrene from livers of corn oil- or 3-methylcholanthrene-treated rats was studied using a nonrecirculating perfusion system. Sterile bile samples were collected every 4 min and assayed for mutagens using the Ames Salmonella (Ta 98) test without addition of microsomes but containing beta-glucuronidase. Rates of export of mutagens produced from benzo(a)pyrene (20 microM) into the bile were stimulated 5-fold by the bile salt sodium taurocholate, concomitant with a 2- to 3-fold increase in bile flow. Steady-state rates of 60 and 90 revertants/g/h were observed in bile when 20 microM benzo(a)pyrene was infused into livers from corn oil or 3-methylcholanthrene-treated rats, respectively. These rates of efflux were increased to 250 and 550 revertants/g/h by the addition of taurocholate. Rates of production of mutagenic phenolic metabolites which account for the mutagenic activity were determined by adding rates of efflux into bile and effluent perfusate with rates of accumulation of metabolites in the cell. In livers from 3-methylcholanthrene-treated rats, rates (8 min) of benzo(a)pyrene phenol formation averaged 300 nmol/g/h during the initial 20 min of perfusion but increased to 450 nmol/g/h after 1 h. The addition of taurocholate increased maximal rates of phenol efflux in the bile from 6 to 148 nmol/g/h and decreased rates of phenol accumulation in intracellular stores from 342 to 220. Rates of efflux into the vena cava effluent averaged 120 nmol/g/h and were not affected by taurocholate. Infusion of dehydrotaurocholate increased the appearance of metabolites of benzo(a)pyrene in the effluent perfusate but did not change rates of efflux into bile. Taurocholate doubled rates of output of phenolic metabolites into the effluent perfusate when bile flow was arrested by perfusion with calcium-free buffer. Thus, mutagenic glucuronides from benzo(a)pyrene phenols accumulated in hepatocytes much faster than rates at which they were exported. Total rates of production of phenolic glucuronides by the liver were not affected by bile salts; however, taurocholate stimulated their export into bile, while dehydrotaurocholate increased their concentration in the effluent perfusate. Both salts probably act by displacing metabolites from intracellular binding sites.

Effect of size on portal circulation of hepatic nodules from carcinogen-treated rats.

The portal circulation of diethylnitrosamine-initiated nodules (0.5 to 7 mm in diameter) was studied in rat livers perfused exclusively via the portal vein. Microlight guides were placed on normal and nodular tissue on the capsular surface of the liver to measure pyridine nucleotide fluorescence (366 leads to 450 nm). When oxygen tension of the inflow perfusate was lowered, fluorescence in both normal tissue and small nodules (less than 2 mm in diameter) increased sharply due to the reduction of pyridine nucleotides, indicating previous normoxia. In contrast, similar manipulations did not increase fluorescence in nodules greater than 2 mm in diameter, demonstrating that nicotinamide adenine dinucleotide was reduced maximally previously; i.e., the nodules were anoxic. Direct measurements of nodule oxygen concentrations with a miniature oxygen electrode confirmed these results. 7-Hydroxycoumarin or fluorescein could be detected with micro-light guides in normal tissue and nodules less than 2 mm in diameter but not in nodules greater than 2 mm in diameter. Furthermore, fluorescent microscopy indicated an absence of fluorescein in nodules greater than 2 mm in diameter. Therefore, with four independent optical and polarographic techniques, we have demonstrated reduced portal circulation in nodules greater than 2 mm in diameter; however, smaller nodules could not be differentiated from normal tissue.

Effect of fatty acids on formation, distribution, storage, and release of benzo(a)pyrene phenols and glucuronides in the isolated perfused rat liver.

The hydroxylation of benzo(a)pyrene and conjugation, storage, and release of benzo(a)pyrene phenols and glucuronides by the perfused rat liver were studied in the presence and absence of acute addition of physiological concentrations of common dietary fatty acids. The actions of fatty acids on the oxidation and conjugation of benzo(a)pyrene in the intact liver were compared with their actions on microsomes isolated from rat liver. Rats were treated with beta-naphthoflavone to stimulate polycyclic aromatic hydrocarbon metabolism. Long-chain fatty acyl CoA compounds (palmitoyl CoA, oleoyl CoA, linolenoyl CoA; 50 microM) inhibited hydroxylation of benzo(a)pyrene by isolated microsomes by about 45%; however, long-chain fatty acids did not affect overall rates of hydroxylation of benzo(a)pyrene by the perfused liver at concentrations ranging up to 300 microM. The medium-chain acyl CoA compound, octanoyl CoA, also did not affect benzo(a)pyrene hydroxylation in microsomes or liver. Although fatty acids did not alter rates of hydroxylation, the ratio of free benzo(a)pyrene phenols to glucuronides (F/G ratio) increased about 60% (P less than 0.05) in livers perfused with long-chain fatty acids (palmitate, oleate, linolenate). Inhibition of glucuronidation was not observed with the medium-chain fatty acid, octanoate. Benzo(a)pyrene phenols and glucuronides accumulated linearly in the liver at rates of approximately 40 nmol/g/h. A second action of both long- and medium-chain length fatty acids was to increase rates of release of benzopyrene phenols into the perfusate by 50 to 80%. Fatty acids did not effect release of benzo(a)pyrene phenols and glucuronides into bile. Taken together, these data support the hypothesis that fatty acids displace carcinogenic metabolites of benzo(a)pyrene from binding sites in the liver which enter the circulation and travel to target tissues.

Benzo(a)pyrene phenol production by perfused rat liver and its inhibition by ethanol.

A rapid and inexpensive method has been developed to estimate rates of benzo(a)pyrene phenol production by perfused rat liver. This method is based on the measurement of benzo(a)pyrene phenols utilizing a simple fluorometric procedure. Within 2 to 3 min after infusion of benzo(a)pyrene bound to serum albumin, phenols are excreted into the perfusate, primarily as glucuronide and sulfate conjugates. Maximal rates of phenol release were 8 to 10 nmol/g/hr in livers from control rats and 40 to 42 nmol/g/hr in livers from 3-methylcholanthrene-treated rats. Fasting of 3-methylcholanthrene-treated rats for 24 hr prior to perfusion experiments did not affect either the rate of phenol production or the extent of their conjugation. Ethanol (20 mM) inhibited rates of phenol formation by 50% in livers from fasted, 3-methylcholanthrene-treated rats but had no effect on benzo(a)pyrene hydroxylase activity in isolated hepatic microsomes. These data indicate that ethanol inhibits phenol formation from benzo(a)pyrene in intact liver, probably by diminishing the supply of the cofactor reduced nicotinamide adenine dinucleotide phosphate.

Conjugation of benzo(a)pyrene 7,8-dihydrodiol-9,10-epoxide in infant Swiss-Webster mice.

Benzo(a)pyrene 7,8-dihydrodiol-9,10-epoxide (BPDE), accepted as the ultimate carcinogen of benzo(a)pyrene, has a very short half-life in aqueous solutions yet induces lung tumors when injected into infant mice. To evaluate the possibility that metabolites of BPDE, principally in the form of stable conjugates, contribute to binding to DNA in peripheral tissues, infant mice were injected i.p. with 39 nmol (+/- ) anti-BPDE. One h after injection, 5% of the dose was recovered in serum and appeared mostly as conjugated metabolites (54% as glucuronides and 16% as glutathione conjugates). Amounts of direct acting electrophiles in serum estimated by trapping with DNA comprised less than 0.02% of the injected dose. No more than 10% of the radioactivity in extracts of liver, lung, and kidney was recovered as BPDE. Glutathione conjugates predominated in the liver and lung, whereas glucuronides were the major metabolites in kidney. Radioactivity bound to DNA in liver, lung, and kidney was 21.5, 42.7, and 7.8 pmol/mg, respectively. Despite the rapid conversion of BPDE to stable conjugates, 32P-postlabeling profiles of DNA adducts in lung closely resembled that noted after addition of BPDE directly to lung homogenate. Thus, the reactive intermediate as well as stable conjugates of BPDE may be transported to target tissues where they initiate tumors.

Adenine nucleotides and carbohydrates in subpopulations of hepatic nodules with normal and compromised microcirculation.

Fluorescein-isothiocyanate dextran (FITC-dextran), a dye confined to the vascular space, was infused via the hepatic artery and portal vein into perfused livers from fed rats treated with diethylnitrosamine for 4 to 5 months. Fluorescence due to FITC-dextran was detected with fiberoptic microlight guides placed on surface nodules of about 5 mm in diameter. Nodules were categorized into groups with normal and compromised microcirculation based on their fluorescence following infusion of FITC-dextran. Similar results were obtained when nodules were classified based on reflectance of trypan blue. Despite compromised microcirculation, ATP and ADP levels as well as ATP/ADP ratios were comparable in both groups of nodules; however, AMP was elevated in FITC-dextran-negative nodules (i.e., those with compromised microcirculation). Nodules with compromised microcirculation also contained higher glucose and lactate levels than nodules that were well perfused; however, glycogen was five times lower than in FITC-dextran-positive nodules. Fasting reduced ATP/ADP ratios in poorly perfused nodules in comparison to well-perfused nodules. In perfused livers from fed rats where glycogen was high, however, ATP/ADP ratios and rates of ATP depletion during ischemia were the same in well-perfused and poorly perfused nodules. Products of glycogen breakdown (e.g., glucose and lactate) were elevated in nodules from livers of fed but not fasted rats. The results indicate that alteration of perfusion of hepatic nodules does not change ATP levels nor the capacity of nodules to utilize high energy phosphate during anoxia. Thus, near normal energy status is maintained from glycogen metabolism in poorly perfused nodules via glycolysis. Since basal ATP content and utilization is comparable in well and poorly perfused nodules, compromised energy status is unlikely to explain selection of nodules that regress to near normal hepatocytes.

Oxidants from nicotinamide adenine dinucleotide phosphate oxidase are involved in triggering cell proliferation in the liver due to peroxisome proliferators.

It was shown that 4-chloro-6-(2,3-xylidino)-2-pyrimidinylthio acetic acid (Wy-14,643), a potent peroxisome proliferator, caused rapid oxidant-dependent activation of nuclear factor kappaB (NF-kappaB) in Kupffer cells in vivo and activated superoxide production by isolated Kupffer cells. Here, we tested the hypothesis that NADPH oxidase (NADPH OX) is the source of oxidants increased by Wy-14,643. Indeed, both activation of NF-kappaB and increases in cell proliferation due to a single dose of Wy-14,643 (100 mg/kg) were prevented completely when rats were pretreated with diphenyleneiodonium (1 mg/kg), an inhibitor of NADPH OX. p47phox is a critical subunit of NADPH OX; therefore, p47phox knockout mice were used to specifically address the hypothesis of NADPH OX involvement. In livers of wild-type mice, Wy-14,643 activated NF-kappaB, followed by an increase in mRNA for tumor necrosis factor a. Importantly, these changes did not occur in p47phox knockouts. Moreover, when Kupffer cells were treated with Wy-14,643 in vitro, superoxide production was increased in cells from wild-type but not p47phox-null mice. Finally, when mice were fed a Wy-14,643-containing (0.1%) diet for 7 days, the increase in liver weight and cell proliferation caused by Wy-14,643 in wild-type mice was blocked in p47phox-null mice. Combined, these results are consistent with the hypothesis that Wy-14,643 activates NADPH OX, which leads to NF-kappaB-mediated production of mitogens that causes hepatocellular proliferation characteristic of this class of nongenotoxic carcinogens.

Several nongenotoxic carcinogens uncouple mitochondrial oxidative phosphorylation.

A number of plasticizers and lipid-lowering drugs induce peroxisomes and cause hepatocellular carcinoma in rodents by mechanisms which remain unknown. In this study, seven structurally dissimilar peroxisome proliferating agents were shown to uncouple oxidative phosphorylation in isolated rat liver mitochondria. For example, perfluorooctanoate (0.5 mM) increased succinate-induced (state 4) mitochondrial respiration by over 50% while stimulation of state 3 respiration with ADP was minimal (i.e., uncoupling occurred). Interestingly, compounds which are potent carcinogens in vivo (e.g., Wy-14,643 and perfluorooctanoate) were more powerful uncouplers of oxidative phosphorylation in vitro than weak tumor-causing agents (e.g., valproate). Uncoupling also occurred in vivo. Basal rates of oxygen uptake in perfused livers from chronically treated rats were increased from 137 +/- 7 mumol g-1/h in pair-fed controls to 153 +/- 5 mumol g-1/h after 2.5 months of feeding Wy-14,643 (0.1% w/v in diet). Concomitantly, rates of urea synthesis from ammonia, a process highly dependent on ATP supply, were reduced almost completely from 104 +/- 10 mumol g-1/h to 13 +/- 6 mumol g-1/h. Bile flow, another energy-dependent process, was also reduced significantly by treatment with Wy-14,643 in vivo for 24 h. Taken together, these data indicate that energy supply for cellular processes such as urea synthesis and bile flow was disrupted in vivo due to uncoupling of oxidative phosphorylation by Wy-14,643. It is proposed that peroxisomal proliferators accumulate in the liver where they uncouple mitochondrial oxidative phosphorylation and interfere with cellular energetics.

Regulation of urea synthesis in sublobular regions of the liver lobule by oxygen.

Periportal and pericentral regions of the liver lobule were isolated from perfused rat liver using a micropunch and incubated in Krebs-Henseleit buffer (pH 7.6) containing 2% poly(ethylene glycol) in Eagle's basal medium, PMSF (50 micrograms/ml) and leupeptin (20 micrograms/ml) for 2 h at 25 degrees C under and O2/CO2 (95:5%) gas phase. Maximal rates of urea production from ammonium chloride were 96.4 +/- 8.7 and 32.8 +/- 5.4 mumol/g per h at 800 and 200 microM O2. Thus, urea synthesis was 2-3-times greater at high than low O2 tension in plugs from periportal and pericentral regions of the liver lobule.

Taurine blunts LPS-induced increases in intracellular calcium and TNF-alpha production by Kupffer cells.

Activation of Kupffer cells by lipopolysaccharide (LPS) plays a pivotal role in the onset of pathophysiological events that occur during endotoxemia and intracellular calcium ([Ca2+]i) is involved in LPS-stimulated cytokine production. Recently, it was shown that Kupffer cells contain a glycine-gated chloride channel. Because taurine, a ubiquitous sulfur-containing beta-amino acid, acts similarly to glycine in neurons by causing hyperpolarization, it was hypothesized that taurine would act via a similar mechanism, blunting the LPS-induced increase in [Ca2+]i in Kupffer cells. To test this hypothesis, Kupffer cells were isolated from female Sprague-Dawley rats and cultured for 24 h. LPS-induced changes in [Ca2+]i were monitored fluorometrically in single cells, whereas levels of tumor necrosis factor alpha (TNF-alpha) released by Kupffer cells after exposure to LPS were measured by enzyme-linked immunosorbent assay. Taurine significantly blunted the LPS-induced increase in [Ca2+]i in a dose-dependent manner (IC50, 0.1 mM). This effect was reversed by strychnine (1 microM) and was prevented when chloride was removed from the extracellular media. Moreover, taurine increased 36Cl- uptake by Kupffer cells in a dose-dependent manner (EC50, 0.2 mM). Furthermore, strychnine (1 microM) reversed the effect of taurine on 36Cl- uptake. These results indicate that taurine activates a glycine-gated chloride channel in Kupffer cells causing chloride influx. In addition, LPS-induced TNF-alpha production was reduced by more than 40% by taurine, an effect that was also reversed by strychnine. In conclusion, taurine blocks the increase in [Ca2+]i due to LPS and significantly reduces TNF-alpha production by mechanisms involving chloride influx into the Kupffer cell.

Adenoviral gene delivery can inactivate Kupffer cells: role of oxidants in NF-kappaB activation and cytokine production.

Kupffer cells play a significant role in the pathogenesis of several liver diseases; therefore, a potential therapeutic strategy would be to inactivate the Kupffer cell with a gene-delivery system. Although recombinant adenovirus provides robust, transgene expression in parenchymal cells, whether adenovirus transduces Kupffer cells is unclear. Thus, the purpose of this study was to evaluate this possibility. In animals infected with adenovirus, Kupffer cells were identified positively to express adenoviral transgenes by immunohistochemical techniques and Western blot analysis, indicating that Kupffer cells are transduced in vivo. Indeed, isolated Kupffer cells were transduced in vitro with recombinant adenovirus in a dose-dependent manner. Moreover, adenoviral transduction of Kupffer cells was blocked by inhibitors of alphaVbeta5 integrin, the co-receptor for adenovirus binding, supporting the hypothesis that adenovirus transduces Kupffer cells via an alphaVbeta5 integrin-dependent mechanism. Indeed, it is shown here that Kupffer cells express alphaVbeta5 integrins. In a functional assay, infection of isolated Kupffer cells with adenovirus containing superoxide dismutase or IkappaB alpha super-repressor blunted LPS-induced nuclear transcription factor kappa B (NF-kappaB) activation and tumor necrosis factor alpha (TNF-alpha) production but not IL-10 production. Moreover, superoxide production was blocked by expression of superoxide dismutase. These data support the hypothesis that LPS-induced NF-kappaB activation and TNF-alpha production in Kupffer cells are oxidant-dependent. These findings suggest that Kupffer cell-targeted approaches may be a potential therapeutic strategy against many inflammatory diseases including early alcohol-induced liver injury.