ARTICLE IN PRESS Clinical Nutrition (2005) 24, 375–384
http://intl.elsevierhealth.com/journals/clnu
ORIGINAL ARTICLE
Does dietary ornithine a-ketoglutarate supplementation protect the liver against ischemia–reperfusion injury? Heidi Schustera, Marie-Ce ´line Blanca,c, Carine Genthona, b `s Le Tourneaud, Patrice The ´rond , Dominique Bonnefont-Rousselotb, Agne Jean-Pascal De Bandta,c,, Luc Cynobera,c a
Laboratoire de Biologie de la Nutrition, EA2498, Faculte´ de Pharmacie, Universite´ Paris 5, 4 avenue de l’Observatoire, 75270 Paris Cedex 06, France b Laboratoire de Biochimie Me´tabolique et Clinique, EA3617, Faculte´ de Pharmacie, Universite ´ Paris 5, France c Service de Biochimie A, Ho ˆpital Ho ˆtel-Dieu, AP-HP, Paris, France d Service d’Anatomie Pathologie, Ho ˆpital Ho ˆtel-Dieu, AP-HP, Paris, France Received 26 May 2004; accepted 7 December 2004
KEYWORDS Glutamine; Arginine; Oxidative stress; Nitric oxide
Summary Nutritional supplementation with glutamine, arginine and their precursors has been proposed to contribute to the protection against ischemiareperfusion-related injuries. The aim of this study was to evaluate in an isolated perfused rat liver model the preventive effect of a 4-day oral ornithine aketoglutarate (OKG) supplementation against warm ischemia–reperfusion (I–R) injury, and the involvement of nitric oxide synthesis. Rats were fed a controlled regimen supplemented with either OKG (5 g kg1; n ¼ 15) or an isonitrogenous mixture of non-essential amino acids (Control; n ¼ 6) for 4 days. Livers were subsequently prepared for isolated perfusion experiments, including a 45 min no-flow ischemic period. The OKG-treated group was divided into two groups according to the absence (OKG; n ¼ 8) or presence of a NO-synthase inhibitor, L-Nonitro-arginine methyl ester (OKG L-NAME; n ¼ 7) during liver perfusion. Liver cytolysis after ischemia was demonstrated by an elevated alanine aminotransferase release during the last 15 min of reperfusion that was significantly higher in the OKG-L-NAME group. Tumor necrosis factor a (TNFa) production was transiently increased only in the control group just after ischemia. At the end of the reperfusion period, liver superoxide dismutase activity was significantly lower in the OKG-L-NAME group compared to control animals.
Corresponding author. Laboratoire de Biologie de la Nutrition, EA2498, Faculte ´ de Pharmacie, Universite´ Paris 5, 4 avenue de
l’Observatoire, 75270, Paris Cedex 06, France. Tel.: +33 1 53 73 99 45; fax: +33 1 53 73 99 52. E-mail addresses: [emailprotected] (H. Schuster), [emailprotected] (J.-P. De Bandt). 0261-5614/$ - see front matter & 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.clnu.2004.12.002
ARTICLE IN PRESS 376
H. Schuster et al. Dietary OKG administration had only a limited effect in this model of mild hepatic I–R, leading mainly to reduced TNFa production. As the content of lipid peroxidation products was not modified, it seems that OKG acts on the inflammatory response rather than on oxidative reactions. This action can tentatively be attributed to the role of OKG as a glutamine precursor rather than to the synthesis of arginine and nitric oxide. & 2004 Elsevier Ltd. All rights reserved.
Introduction Warm ischemia–reperfusion (I–R)-induced hepatic injuries are of major concern during liver surgery or in shock situations.1 Inhibition of mitochondrial respiration, and therefore depletion of hepatic ATP content during ischemia, affects cell membrane transport systems resulting in severe deterioration of cellular ion homeostasis. The resulting change in hepatocyte hydration status leads to cell shrinkage and acts as a strong inducer of proteolysis. The activation of protein degradation has been demonstrated as a predictive factor of graft failure.2 Indeed, I–R leads to the activation of proteases such as calpain,3 which are inducers of apoptotic cell death and are involved in the proteolytic conversion of xanthine dehydrogenase to xanthine oxidase. Hypoxanthine accumulation due to ATP catabolism and activation of xanthine oxidase lead to the generation of damaging reactive oxygen species (ROS) on reperfusion.4 Kupffer cells contribute to post-ischemic oxidative stress by releasing pro-inflammatory cytokines (tumor necrosis factor a (TNFa), IL-1, IL-6).1,5 In turn, inducible nitric oxide synthase (iNOS) from hepatocytes and Kupffer cells is activated to produce nitric oxide (dNO). Both cytotoxic and cytoprotective effects of dNO have been demonstrated, related, respectively, to its ability to restore microcirculation via its vasodilatory properties5 and to the inhibitory effect of large amounts of dNO on energy metabolism leading to cell death. During I–R, oxidative and nitrosative stresses may overwhelm physiological scavengers, i.e., the three main hepatic detoxifying enzymes: superoxide dismutase (SOD), catalase and glutathione peroxidase (GPx), and reduced glutathione (GSH). Thus, energy depletion and proteolytic mechanisms in combination with inflammatory and oxidative phenomena contribute to I–R-induced hepatic injuries. Among the various nutritional strategies to counteract these damaging processes, the use of glutamine6 or arginine,7,8 two amino acids (AAs) involved in the regulation of protein metabolism and inflammation or their precursors,9,10 has received considerable attention in recent years. Besides its regulatory activities in autophagic proteolysis and nitrogen metabolism, and its role
as a substrate for rapidly proliferating cells, the utility of glutamine comes from its possible role as a glutathione precursor. On the other hand, arginine is the physiological precursor for dNO synthesis. Ornithine a-ketoglutarate (OKG), which is a precursor for both AAs,11,12 may be of interest in this setting. It has been shown to display anabolic and anticatabolic properties and to be a potent modulator of immune cell functions.11–13 A protective effect of OKG against I–R injury has been demonstrated in the intestine:10 ornithine and arginine partially prevent ischemia-related lesions of the intestinal mucosa, most probably via the production of dNO and its ability to quench superoxide anion. Nutritional status has been shown to influence I–R injury in the liver.14 The question we address is the utility of OKG supplementation as a preventive nutritional treatment in the context of programmed elective surgery. The aim of the present study was to evaluate whether a short-term (4 days) oral supplementation with OKG could exert protective effects against I–R in the liver. An ex vivo model of warm ischemia in the isolated perfused rat liver was chosen because it provided a mechanistic approach in well-controlled conditions.15,16 Protective effects were evaluated in terms of oxidative stress, inflammatory response and metabolic function. Because of its role as an arginine precursor and the eventual associated induction of dNO synthesis, we expected a possible double effect of OKG on reperfusion injury. For this reason, a group of the OKG-supplemented rats were additionally treated with L-No-nitro-arginine methyl ester (L-NAME), an NOS inhibitor, during the perfusion in order to distinguish between the glutamine/glutathione and the arginine/dNO pathways.
Materials and methods Reagents OKG (Cetornans) was kindly provided by Chiesi S.A. (Courbevoie, France). Bovine serum albumin (Fraction V) was obtained from Calbiochem (VWR, Fontenay-sous-Bois, France); AAs, L-NAME and
ARTICLE IN PRESS OKG and ischemia–reperfusion injury other reagents were from Sigma-Aldrich (St-Quentin-Fallavier, France).
Animals Male Sprague–Dawley rats (Charles River, Lyon, France) weighing 175–200 g were housed individually in standard cages in a controlled environment with normal 12-h light/dark cycle and at a constant temperature of +21 1C. The animals were acclimatized during a 4-day period at our animal facility with free access to water and a standard chow diet (UAR A04, Usine d’Alimentation Rationnelle, Villemoisson-sur-Orge, France) ad libitum. The research protocol complied with the guidelines for animal care of our institution. One of the authors (L.C.) is officially authorized (75–461) by the French Ministry of Agriculture and Forestry to perform experiments in rodents.
Experimental design and diets On the first day of the study, rats were randomly assigned to receive a standard chow for 4 days supplemented with either OKG (5 g kg1 BW; n ¼ 15)17 or an isonitrogenous (total nitrogen supply: 2.9 g kg1 BW) mixture of non-essential amino acids (NEAAs) supplying alanine, serine, glycine, histidine, proline and asparagine (0.46, 0.54, 0.39, 1.08, 0.59 and 0.68 g kg1 BW, respectively; control group; n ¼ 6). None of these NEAAs in such low amounts is expected to exert any pharmacological effect. Diets were given at 90% of spontaneous intake (as recorded during the acclimatization period) to ensure complete consumption.12,17 Diets supplied approximately 350 kcal kg1 BW d1. At the end of the nutrition period, rats were weighed and prepared for liver isolation and perfusion.
Liver perfusion Non-fasted rats underwent pentobarbital anesthesia (60 mg kg1 i.p.) and liver perfusion was performed using a very well-standardized and well-validated procedure.18 Briefly, after cannulation of the bile duct, 1 ml saline containing 1000 U heparin was injected into the hepatic vein, and the portal vein was ligatured and then cannulated. Livers were perfused under constant physiological pressure (12 cm H2O) via the portal vein in a recirculating system in a thermostatically controlled (37 1C) cabinet. The perfusion medium (300 ml) was a Krebs–Henseleit buffer supplemented with albumin 30 g l1, glucose 8.5 mM and a
377 mixture of AAs (alanine, leucine, glutamine, proline, phenylalanine, histidine, tryptophan, methionine at twice their physiologic plasma concentrations) as previously described.15 Because of the antiproteolytic properties of the AAs added,19 these conditions prevent the acute proteolysis observed during AA-free perfusion.15 The perfusate was maintained at 37 1C, pH 7.4 and oxygenated with an O2/CO2 mixture (95/5 v/v). At the end of a 30 min equilibration period (t30) a balanced AA mixture (without glutamine) at twice plasma concentrations was added to ensure reliable AA exchanges.15 Arginine concentration in the perfusate at t30 was 166 mmol l1. Hepatic I–R was initiated in all experiments 10 min (t40) later by clamping the portal vein catheter for 45 min (t85); reperfusion was then performed for 60 min (t145). Throughout the perfusion, pH, portal flow and bile flow were constantly monitored. Bile was collected constantly into tubes and flow was estimated gravimetrically, assuming a specific mass of 1 g ml1. Samples of perfusate were taken at times t35, t85, t95, t100, t115, t130 and t145 to assay AAs, enzymes (AST, ALT and LDH), glucose, urea, NOd and TNFa. The OKG-treated animals were divided (randomization at the start of the in vivo experiment) into two groups according to the presence (OKG-L-NAME group: n ¼ 8) or absence (OKG group: n ¼ 7) of 1 mmol l1 L-NAME during liver perfusion. Results were compared with those obtained in the control group.
Sample management Before liver isolation, blood samples (500 ml) were drawn on heparin from the hepatic vein. After centrifugation, plasma was deproteinized with sulfosalicylic acid (30%) for AA measurement. Samples of perfusate taken at various times (see above) were stored at 80 1C until analysis. At the end of the experiment, livers were cleaned and weighed on an electronic balance; samples of approximately 500 mg each were either immediately frozen in liquid nitrogen and then stored at 80 1C or treated for histological examination.
Analytical methods Enzyme activities (AST, ALT and LDH) and glucose and urea concentrations in the perfusate were measured by standard methods on an OLYMPUS AU600 analyzer with dedicated reagents (OLYMPUS, Rungis, France). For AA analysis, perfusate samples were immediately deproteinized with sulfosalicylic acid
ARTICLE IN PRESS 378 (30 mg ml1). AAs were assayed by ion exchange chromatography and spectrometric detection after ninhydrin reaction on a Jeol Aminotac JLC-500 V analyzer (Jeol, Croissy-sur-Seine, France). d NO production was measured as the sum of nitrite (NO 2 ) and nitrate (NO3 ) in perfusate using a colorimetric assay (R&D Systems, Lille, France) based on the Griess reaction after conversion of nitrate to nitrite by nitrate reductase,20 with a detection limit of 1.35 mmol l1. TNFa was measured in perfusate by a quantitative sandwich enzyme immunoassay technique with microplates pre-coated with a monoclonal antibody specific for rat TNFa (R&D Systems, Lille, France), with a detection limit of 5 pg ml1. For measurement of the hepatic hydration status after reperfusion, liver samples were weighed before and after drying for 48 h at 90 1C. For the determination of hepatic AA and protein contents, tissue samples were homogenized in 10 volumes of ice-cold 10% trichloroacetic acid containing 0.5 mM EDTA and centrifuged. AA concentrations were measured in the supernatant as described above. Arginine levels in the liver are not reported since they were very low, and under the detection limit. After delipidation, the liver protein content was measured in the precipitate using Gornal’s method. To determine liver malondialdehyde (MDA) content and SOD activity, samples of liver tissue were homogenized in ice-cold potassium phosphate buffer; after centrifugation the supernatant was immediately frozen at 80 1C. Liver MDA content was evaluated by the measurement of TBARS assayed spectrofluorometrically,21 after acidic condensation at 95 1C between MDA and thiobarbituric acid. The condensation product was assessed by fluorometry with an excitation wavelength of 515 nm and an emission wavelength of 548 nm. Malondialdehydebis-(diethylacetal) was used as standard. The determination of liver SOD activity was based on the production of Od anions by the 2 xanthine/xanthine oxidase system, and the reaction of these free radicals with 2-(4-iodophenyl)3(4-nitrophenol)-5-phenyltetrazolium chloride to produce red formazan (absorption wavelength ¼ 505 nm). SOD activity was assessed by its ability to inhibit the latter reaction. A commercially available kit (Randox, Roissy, France)22 adapted on a Konelabs analyzer (Kone, Espoo, Finland) was used for this determination. Glutathione in its reduced form (GSH) was determined by reversed-phase high-performance liquid chromatography (HPLC) coupled with coulometric electrochemical detection.23 Two to five hundred milligrams of hepatic tissue were
H. Schuster et al. homogenized for 30 s with a Dounce homogenizer in 1 ml of cold perchloric acid 0.3 M. After centrifugation, the supernatant was diluted 40 times in mobile phase (50 mmol l1 KH2PO4 adjusted to pH 2 with 85% H3PO4). Fifty microliters of the diluted supernatant were then analyzed by HPLC. A Spherisorb C-18 reversed-phase ODS-2 column (5 mm; 150 4.6 mm) was used. GSH eluted from the HPLC column was detected by a Coulochems detector equipped with a model 5010s dual analytical cell and a model 5020s guard cell (ESA, Bedford, MA).23
Histological examination At the end of perfusion, 2 mm-thick liver samples were cut from the middle of the median lobe and immediately fixed in 10% formalin. They were then embedded in paraffin; 3 mm slices were cut and then stained with hematein–eosin–saffron (HES) for light microscope analysis. Histological data were compared in a blinded way with those obtained for livers perfused for 145 min in similar conditions but without ischemia.
Calculations and statistics Substrate exchanges (F) were calculated as F¼
Ct1 V 1 Ct2 V 2 ; ðt1 t2 ÞW
where Ct1 and Ct2 represent concentrations of the substrate at time t1 and t2 in two consecutive samples of perfusate, V 1 and V 2 represent the perfusate volume at the same times taking into account sampling and evaporation (0.2 ml min1) and W represents the liver wet weight. Exchanges for longer periods of time were calculated as the means of successive values corrected for time. Positive values denote substrate release and negative values uptake by the liver. Data are expressed as mean7SEM. Statistical significance of the differences was determined by ANOVA plus the Fisher test using StatViews (SAS Institutes Inc., Cary, HNC). A value of Po0:05 was considered as significant.
Results Animal characteristics During the feeding period, the animals consumed their daily food allowance completely. At the end of this feeding period animal weights were similar
ARTICLE IN PRESS OKG and ischemia–reperfusion injury
379
Portal flow (% of preischemic value)
Control
OKG
OKG L-NAME
a a a 100 b
80
b b,c
c c c
d d d
d d d
115
130
d
d d
60 40 20 0 35
95
100
145
T i me Before ischemia
ischemia
reperfusion
Figure 1 Effects of OKG and I–R on liver portal flow. Rats were fed for 4 days on either a control or an OKGsupplemented diet (5 g kg1 BW). For isolated liver perfusion, half of the OKG-treated livers was perfused in the presence of L-NAME. All livers were perfused for 145 min with a period of warm ischemia between t40 and t85 min. Data (mean7SEM) are presented as a percentage of pre-ischemic portal flow. Bars with different letters are significantly different at Po0:05:
in the control and OKG groups (animal weight: Control: 23972; OKG: 24273; OKG L-NAME: 23973 g, ns). At the end of the feeding period, plasma ornithine levels were significantly higher in OKG groups (Control: 43711; OKG: 95731 mmol l1, Po0:05) while other AAs were not affected. Liver weights and liver hydration at the end of the experiment were similar in all groups (liver weight after I–R: Control: 9.170.2; OKG: 9.570.2; OKG L-NAME: 9.670.1 g, ns; liver hydration after I–R: Control: 74.870.5%; OKG: 74.770.6%; OKG LNAME: 76.170.6%, ns).
Hepatic function Bile flow, similar in all three groups before ischemia (T 0 2T 35 : 0.4670.04 ml min1 g1), was markedly decreased on reperfusion in all experiments (T 85 2T 115 : 0.0770.01 ml min1 g1; Po0:05 vs. T 0 2T 35 ); a significant improvement was noted during the second half of the reperfusion period (T115–T145: 0.1570.02 ml min1 g1; Po0:05 vs. T 85 2T 115 ) with no difference between groups. In all three groups, portal flow initially decreased in the immediate reperfusion period, and progressively increased to reach 86–90% of pre-ischemia level at the end of the experiments (Fig. 1). There was no significant ALT release before ischemia (Table 1). After ischemia, the rate of
ALT release was significantly increased in all groups in the first minutes of reperfusion; it returned to pre-ischemia level before increasing again until the end of the experiments. In the last 15 min of the reperfusion period, ALT release was significantly higher in the OKG-L-NAME group. Kinetics of AST and LDH release followed a similar pattern but differences between groups did not reach statistical significance (data not shown).
Hepatic metabolism A significant glucose production was observed in all groups before ischemia and in the control group just after ischemia. On reperfusion, glucose production significantly decreased and was abolished at the end of the study period in all groups (Fig. 2). Neither AA exchange (data not shown) nor urea production (Control: 0.1970.02; OKG: 0.1370.06; OKG L-NAME: 0.2070.02 mmol min1 g1, ns) was affected by I–R, and similar substrate exchanges were observed in the three groups. Cumulated total or branched-chain AA exchanges were similar in all groups. No difference was found in liver AAs and protein content (Control: 133713; OKG: 107712; OKG LNAME 109713 mg g1 liver, ns) at the end of reperfusion.
ARTICLE IN PRESS 380 Table 1
H. Schuster et al. Effect of I/R and oral OKG administration on liver cytolysis: ALT release. Control
OKG a
0–35 min 85–95 min 95–115 min 115–130 min 130–145 min
OKG L-NAME a
3.971.4a 24.579.4a,b 17.976.8a 46.8715.3b,c 178.2770.1d
2.371.0 26.375.9a,b 13.375.9a 42.4716.8b,c 91.4737.7c
1.970.6 40.077.7b,c 11.372.4a 31.474.7b,c 81.5715.8c
Note: Samples of perfusate were taken at the different time points indicated for the measurement of ALT activity. Results (mean7SEM) represent ALT release (IU min1 g1) calculated between two consecutive points. Values with different superscripts are significantly different at Po0:05:
5.0
Control
OKG
OKG L-NAME
Glucose flux (µmol min-1 g-1)
4.0 a
a b
a
3.0 a
b
2.0
b
b 1.0
b b b
b b
b b
0.0
Time
-1.0 85-95
0-35 Before ischemia
ischemia
95-115
115-130
130-145
reperfusion
Figure 2 Effects of OKG and I–R on hepatic glucose fluxes. Rats were fed for 4 days on either a control or an OKGsupplemented diet (5 g kg1 BW). For isolated liver perfusion, half of the OKG-treated livers was perfused in the presence of L-NAME. All livers were perfused for 145 min with a period of warm ischemia between t40 and t85 minutes. Data are presented as mean7SEM and expressed in mmol min1 g1. Bars with different letters are significantly different at Po0:05:
Inflammation and oxidative stress parameters TNFa release increased transiently only during the first 10 min of reperfusion. It reached statistical significance in controls (Po0:05; Table 2), but not in the OKG-treated groups. d NO production remained very limited throughout the experiment without any significant difference between the three groups (Control: 0.0470.04; OKG: 0.0170.03; OKG L-NAME: 0.0770.03 nmol min1 g1, ns). At the end of the reperfusion period, liver SOD content tended to decrease in the OKG groups, significantly so in the OKG-L-NAME group compared with control animals (Control: 1469783; OKG: 14077193; OKG L-NAME: 12747139 IU g1;
Po0:05 OKG L-NAME vs. Control). Reduced glutathione content was not different in the three groups (Control: 3.270.3; OKG: 2.970.4; OKG LNAME: 2.670.2 mmol g1 liver, ns). We did not observe any difference in the MDA content of the liver between the three groups (Control: 341779; OKG: 280734; OKG L-NAME: 324750 mmol g1 liver, ns).
Histological examination All livers displayed a well-preserved architecture with the exception of a slight sinusoidal extension. Ischemia-induced morphological changes resulted in different degrees of vacuolization, not observed in livers that underwent perfusion without ischemia (Fig. 3a). Three levels of vacuolization were
ARTICLE IN PRESS OKG and ischemia–reperfusion injury Table 2
381
I/R-induced TNFa release. Control
0–35 min 85–95 min 95–145 min
OKG a
11.372.2 44.2714.4b 12.571.9a
OKG L-NAME a
12.773.3 39.5710.7a,b 15.971.7a
12.573.4a 34.6710.5a,b 13.172.2a
Note: Pre- and post-ischemic tumor necrosis factor a (TNFa) productions (mean7SEM) are expressed as pg min1 g1. Values with different superscripts are significantly different at Po0:05:
Figure 3 Light micrographs of liver slices (HES, G 20) obtained at t145 after 40 min perfusion, 45 min ischemia and 60 min reperfusion time for the I–R livers or after 145 min of perfusion for the non-ischemic perfused livers. All livers maintained almost normal hepatic architecture. There were no differences between the three I–R groups, although different degrees of lesion severity could be discerned: (a) non-ischemic perfused liver with sinusoidal extension, (b) I–R liver with centrolobular vacuolization, (c) I–R liver with centro-mediolobular vacuolization and (d) I–R liver presenting diffuse vacuolization.
observed depending on location and extent: in the centrolobular region (Fig. 3b); in centro- and mediolobular regions (Fig. 3c); diffused vacuolization affecting the whole liver lobule (Fig. 3d). These three degrees of vacuolization were similarly observed in the three groups of livers subjected to I–R.
Discussion In this study, the influence of OKG administration on I–R-related hepatic injury was investigated in an isolated perfused rat liver model. Our data do not support a major role of a supplementation of the diet with OKG as a means for preventing
ARTICLE IN PRESS 382 hepatic I–R injury, at least in previously healthy animals. As in other studies of normothermic ischemia,24 reperfusion-induced vacuolization of hepatocytes, although this was not associated with severe morphological damage noted on histological examination. Apoptotic hepatocytes were not quantified, but in relatively short-term ischemia other authors using TUNEL assay did not find any evidence of increased apoptosis after the ischemic insult.24 The limited effect of OKG, at variance with the proposed protective role of glutamine and its precursors,6 might be related to this relative resistance of the liver to the ischemic insult in this model. This is supported by the absence of severe morphological damage noted on histological examination. When assessing the consequence of I–R on liver function, a number of factors have to be considered. As non-fasted rats were used in our experiments, the progressive decrease in glucose release with time is probably the consequence of glycogen utilization and depletion rather than being related to the ischemic episode. Fed rats have replenished hepatic glycogen stores and therefore glycogen breakdown enables glucose release during the first part of the perfusion. On the other hand, at the end of the experiments glucose release was no longer observed, which is consistent with the observation that glycogen stores are depleted after 1 h of reperfusion.25 In the absence of any significant modification in AA exchanges or in urea production when comparing pre-ischemic and post-ischemic data, modification in AA utilization for gluconeogenesis can therefore probably be ruled out. On the other hand, there are clear indications of deterioration in hepatic function following ischemia and reperfusion. After ischemia, liver reperfusion is constantly associated with diminished bile flow, as observed in our experiments. As liver perfusions were performed in the absence of bile acids, bile flow in our model represents bile-acidindependent bile production.18 It has been demonstrated that this component of bile flow is closely related to glutathione release into bile.26 It can thus be proposed that the decrease in bile flow is related to glutathione consumption due to an oxidative stress. In addition, ischemia leads to an energy loss and therefore to a decrease in ATP availability for energy-dependent transport of reduced glutathione into the bile.26 Ischemia/reperfusion is also associated with the release of enzymes, e.g. ALT, AST, LDH markers of cytolysis. Interestingly, cytolysis, which was highest immediately on reperfusion, decreased thereafter and then increased again until the end of the
H. Schuster et al. experiments. This biphasic course has already been described by other authors,27 who observed a peak at the onset of reperfusion and a second progressive linear increase 15 min later; such a time course has also been demonstrated for the production of ROS in a similar model.27 During I–R of the liver in vivo, early damage is believed to be due to Kupffer cell activation, while neutrophils seem more widely implicated in the damage beyond 6 h.1 As our experiments were performed with a cell-free perfusate, it could be assumed that only Kupffer cell-related alterations were involved in the deterioration of hepatic function observed here. Activation of Kupffer cells in our experiments is in accordance with the observed release of the proinflammatory cytokine TNFa, TNFa production being highest just after ischemia with a decrease later on. Through its ability to induce apoptosis and necrosis in liver transplants, this cytokine plays a prominent role in the impairment of liver function, as suggested by a demonstration that pre-treatment with anti-TNFa antibodies protected liver from apoptosis and significantly reduced postischemic hepatic injury.28 Although a possible preventive effect of OKG administration against I–R injury has been evoked, no major effect of this therapy was noted here. As stated above, OKG is a precursor of arginine and glutamine11 which are presumed to potently interfere with I–R-induced processes. Concerning the ability of OKG to promote arginine and dNO synthesis, no significant dNO production was observed in our experiments in either the presence or absence of OKG. However, the addition of the NOS inhibitor L-NAME affected liver function with higher cytolysis in the late reperfusion time. This suggests that either only minimal amounts of dNO, undetectable under our experimental conditions, were produced, probably by constitutive dNO-synthase, or that dNO remained trapped in the liver to exert paracrine effects. This dNO may have a marked influence on vascular microcirculation. This also suggests that iNOS, which could produce higher amounts of dNO, was not activated here, at least during the time span of the reperfusion period investigated. This is in contrast with the observation of a marked dNO production at the onset of reperfusion29 and then after 1 h of reperfusion30 described by other authors using in vivo models. The absence of blood cells in our isolated perfused rat liver model certainly contributed to these findings. OKG is a precursor of glutamine, which has been suggested to promote the synthesis of glutathione. It has been shown that glutamine supplementation attenuates GSH depletion or promotes its synthesis
ARTICLE IN PRESS OKG and ischemia–reperfusion injury in various cells, including hepatocytes.31–33 However, in our experiments we did not observe any difference in glutathione content of the livers or in glutamine exchange in the OKG-treated rats. One possible explanation for this finding is that OKG was administered to healthy rats and the animals were studied in a fed state before the oxidative stress at a time when, taking into account the tight regulation of hepatic oxidative status, there was no reason for elevated production of GSH. It might have been more appropriate to give the OKG concomitantly with the ischemic insult at a time when the substrate requirement for antioxidant defense was increased. Finally, while glutamine is a strong regulator of proteolysis via cell swelling and activation of p38MAPK,34 there is no evidence of an OKG effect on protein degradation. Taking into account the mild severity of the I–R insult, this is not surprising since we did not observe any I–R-related increase in branched-chain AA release, which is a marker of hepatic proteolysis.19 However, two elements suggest a positive effect of OKG supplementation against oxidative stress. There was no significant increase in TNFa production in the OKG-treated groups just after ischemia compared with pre-ischemic levels, suggesting a lesser inflammatory response, while intrahepatic activity of the antioxidant enzyme SOD tended to be lower. As oxidative stress through the activation of transcription factors such as NFkB leads to the activation of the expression of both inflammatory factors, such as TNFa, and enzymes involved in the protection against oxidative processes, such as SOD, it may be that the lower expression of TNFa and SOD activity in the liver signals a lower oxidative insult or a higher regulation of the inflammatory process in OKG pre-treated rats. As TBARS, which are considered as markers of lipid peroxidation,35 were not different in the three groups, OKG may act on inflammatory response rather than on oxidative reactions. This could probably be related to the known immunoregulatory properties of glutamine, which has been demonstrated to reduce the production of proinflammatory cytokines, for example in the intestinal immune cells36 or in blood mononuclear cells of pancreatitis patients.37 In conclusion, in this model of mild hepatic ischemic insult, supplementation of the oral diet with OKG has no beneficial effect on the liver. However, a lower production of TNFa associated with a lower intrahepatic SOD activity suggests that OKG may exert some protective effects via its role as a glutamine precursor and the immunoregulatory properties of this AA. Further experiments are
383 required to assess whether OKG administration concomitant to ischemic insult improves liver function more effectively. Also, preoperative nutritional status should be considered when assessing OKG effects in malnourished rats.
Acknowledgements Part of this study was financed by a grant from the DAAD (German academic exchange service).
References 1. Lentsch AB, Kato A, Yoshidome H, McMasters KM, Edwards MJ. Inflammatory mechanisms and therapeutic strategies for warm hepatic ischemia/reperfusion injury. Hepatology 2000;32:169–73. 2. Calmus Y, Cynober L, Dousset B, et al. Evidence for the detrimental role of proteolysis during liver preservation in humans. Gastroenterology 1995;108:1510–6. 3. Kohli V, Madden JF, Bentley RC, Clavien PA. Calpain mediates ischemic injury of the liver through modulation of apoptosis and necrosis. Gastroenterology 1999;116:168–78. 4. Rosser BG, Gores GJ. Liver cell necrosis: cellular mechanisms and clinical implications. Gastroenterology 1995;108:252–75. 5. Alexander B. The role of nitric oxide in hepatic metabolism. Nutrition 1998;14:376–90. 6. Furst P, Kuhn KS. Amino-acid substrates in new bottles: implications for clinical nutrition in the 21st century. Nutrition 2000;16:603–6. 7. Bedirli A, Sozuer EM, Muhtaroglu S, Alper M. The role of oxygen free radicals and nitric oxide in organ injury following hemorrhagic shock and reinfusion. Int J Surg Invest 2000;2:275–84. 8. Burra P, Chirizzi L, Cardin R, et al. Warm hepatic ischemia in pigs: effects of L-arginine and oligotide treatment. J Invest Surg 2001;14:303–12. 9. Tazuke Y, Wasa M, Shimizu Y, Wang HS, Okada A. Alanylglutamine-supplemented parenteral nutrition prevents intestinal ischemia–reperfusion injury in rats. J Parenter Enteral Nutr 2003;27:110–5. 10. Duranton B, Schleiffer R, Gosse F, Raul F. Preventive administration of ornithine alpha-ketoglutarate improves intestinal mucosal repair after transient ischemia in rats. Crit Care Med 1998;26:120–5. 11. Cynober L. Ornithine a-ketoglutarate as a potent precursor of arginine and nitric oxide: a new job for an old friend. J Nutr 2004;134:2858S–62S. 12. Pernet P, Coudray-Lucas C, Schneid C, Jardel A, Cynober L. Dose dependency of the effect of ornithine a-ketoglutarate on tissue glutamine concentrations and hypercatabolic response. Br J Nutr 2004;92:627–34. 13. Moinard C, Chauveau B, Walrand S, et al. Phagocyte functions in stressed rats: comparison of modulation by glutamine, arginine and ornithine 2-oxoglutarate. Clin Sci (Lond) 1999;97:59–65. 14. Domenicali M, Caraceni P, Vendemiale G, et al. Food deprivation exacerbates mitochondrial oxidative stress in rat liver exposed to ischemia–reperfusion injury. J Nutr 2001;131:105–10.
ARTICLE IN PRESS 384 15. De Bandt JP, Cynober L, Ballet F, Coudray-Lucas C, Rey C, Giboudeau J. Amino acid metabolism in isolated perfused rat liver. J Surg Res 1990;49:8–13. 16. Blonde-Cynober F, Plassart F, de Bandt JP, et al. Metabolism of alpha-ketoisocaproic acid in isolated perfused liver of cirrhotic rats. Am J Physiol 1995;268:E298–304. 17. Le Boucher J, Obled C, Farges MC, Cynober L. Ornithine alpha-ketoglutarate modulates tissue protein metabolism in burn-injured rats. Am J Physiol 1997; 273:E557–63. 18. De Bandt JP, Lasnier E, Rey C, et al. Effects of amino acids on bile acid-dependent and independent bile flow in the isolated perfused rat liver. J Hepatol 1999;30:843–9. 19. Poso AR, Wert Jr JJ, Mortimore GE. Multifunctional control of amino acids of deprivation-induced proteolysis in liver. Role of leucine. J Biol Chem 1982;257:12114–20. 20. Nims RW, Cook JC, Krishna MC, et al. Colorimetric assays for nitric oxide and nitrogen oxide species formed from nitric oxide stock solutions and donor compounds. Methods Enzymol 1996;268:93–105. 21. Yagi K. A simple fluorometric assay for lipoperoxide in blood plasma. Biochem Med 1976;15:212–6. 22. de Haan JB, Newman JD, Kola I. Cu/Zn superoxide dismutase mRNA and enzyme activity, and susceptibility to lipid peroxidation, increases with aging in murine brains. Brain Res Mol Brain Res 1992;13:179–87. 23. Benrahmoune M, Therond P, Abedinzadeh Z. The reaction of superoxide radical with N-acetylcysteine. Free Radical Biol Med 2000;29:775–82. 24. Schauer RJ, Gerbes AL, Vonier D, et al. Glutathione protects the rat liver against reperfusion injury after prolonged warm ischemia. Ann Surg 2004;239:220–31. 25. Ishikawa T, Yagi T, Sadamori H, et al. Kupffer cell activation in the survival discrepancy between liver grafts from enterally and parenterally fed donors. Transplant Int 1998;11(Suppl. 1):S410–6. 26. Koeppel TA, Trauner M, Mennone A, Arrese M, Rios-Velez L, Boyer JL. Role of glutathione in hepatic bile formation
H. Schuster et al.
27.
28.
29.
30. 31.
32.
33.
34.
35.
36.
37.
during reperfusion after cold ischemia of the rat liver. J Hepatol 1998;28:812–9. Okuda M, Lee HC, Chance B, Kumar C. Glutathione and ischemia–reperfusion injury in the perfused rat liver. Free Radical Biol Med 1992;12:271–9. Ben Ari Z, Hochhauser E, Burstein I, et al. Role of anti-tumor necrosis factor-alpha in ischemia/reperfusion injury in isolated rat liver in a blood-free environment. Transplantation 2002;73:1875–80. Cottart CH, Do L, Blanc MC, et al. Hepatoprotective effect of endogenous nitric oxide during ischemia–reperfusion in the rat. Hepatology 1999;29:809–13. Battista S, Mengozzi G, Bar F, et al. Nitric oxide level profile in human liver transplantation. Dig Dis Sci 2002;47:528–34. Flaring UB, Rooyackers OE, Wernerman J, Hammarqvist F. Glutamine attenuates post-traumatic glutathione depletion in human muscle. Clin Sci (Lond) 2003;104:275–82. Chang WK, Yang KD, Chuang H, Jan JT, Shaio MF. Glutamine protects activated human T cells from apoptosis by upregulating glutathione and Bcl-2 levels. Clin Immunol 2002;104:151–60. Yu JC, Jiang ZM, Li DM. Glutamine: a precursor of glutathione and its effect on liver. World J Gastroenterol 1999;5:143–6. vom Dahl S, Ha ¨ussinger D. The role of amino acids in the control of proteolysis. In: Cynober L, editor. Metabolic and therapeutic aspects of amino acids in clinical nutrition. Boca Raton, FL: CRC Press; 2004. p. 275–89. Therond P, Bonnefont-Rousselot D, Davit-Spraul A, Conti M, Legrand A. Biomarkers of oxidative stress: an analytical approach. Curr Opin Clin Nutr Metab Care 2000;3:373–84. Coeffier M, Miralles-Barrachina O, Le Pessot F, et al. Influence of glutamine on cytokine production by human gut in vitro. Cytokine 2001;13:148–54. de Beaux AC, O’Riordain MG, Ross JA, Jodozi L, Carter DC, Fearon KC. Glutamine-supplemented total parenteral nutrition reduces blood mononuclear cell interleukin-8 release in severe acute pancreatitis. Nutrition 1998;14:261–5.
