- Email: [email protected]
Cytoprotection by the osmolytes betaine and taurine in ischemia- reoxygenation injury in the perfused rat liver
Cytoprotection by the Osmolytes Betaine and Taurine in IschemiaReoxygenation Injury in the Perfused Rat Liver MATTHIAS WETTSTEIN
Medium osmolarity sensitively regulates Kupffer cell functions like phagocytosis and prostaglandin (PG) and cytokine production. Betaine and taurine, recently identified as osmolytes in liver cells, interfere with these effects. Because Kupffer cell activation is an important pathogenic mechanism in ischemia-reoxygenation injury, the influence of osmolarity and osmolytes was investigated in a rat liver perfusion model of warm ischemia. Livers were perfused with different medium osmolarities for 60 to 90 minutes in the absence of oxygen, followed by another 90 minutes of reoxygenation. Lactate dehydrogenase (LDH) leakage into the effluent perfusate during the hypoxic and the reoxygenation period was eight- to 10-fold higher with a medium osmolarity of 385 mosmol/L than in normo-osmolarity, and further decreased with hypoosmolar perfusion buffer. Betaine and taurine addition to the perfusate in near physiological concentrations decreased hypoxia-reoxygenation–induced LDH leakage, aspartate transaminase (AST) leakage, and perfusion pressure increase in hyperosmolar and normo-osmolar perfusions. Stimulation of PGD2 , PGE2 , thromboxane B2 (TXB2 ), and tumor necrosis factor a (TNF-a) release, as well as induction of carbon uptake by the liver during reoxygenation, were suppressed by betaine and taurine, pointing to an interference of these osmolytes with Kupffer cell function. In contrast, endothelial cell function as assessed by hyaluronic acid (HA) uptake was not influenced. It is concluded that warm ischemia-reoxygenation injury in rat liver is aggravated by hyperosmolarity and attenuated by hypo-osmolarity. The osmolytes betaine and taurine have a protective effect, presumably by inhibition of Kupffer cell activation. (HEPATOLOGY 1997;26:1560-1566.) Ischemic liver injury is of major clinical relevance during organ transplantation and in shock syndromes. Loss of cell function and cell injury during the ischemic period are thought to be caused by adenosine triphosphate depletion, activation of nonlysosomal proteases, and glucose depletion.1-3 Although the exact pathophysiological mechanisms
Abbreviations: TNF-a, tumor necrosis factor a; PG, prostaglandin; TXB2 , thromboxane B2 ; LDH, lactate dehydrogenase; AST, aspartate transaminase; HA, hyaluronic acid. From the Clinic for Gastroenterology, Hepatology, and Infectiology, HeinrichHeine-University, Du¨sseldorf, Germany. Received January 16, 1997; accepted July 24, 1997. Supported by the Deutsche Forschungsgemeinschaft (grants We 1936/1-1, Ha 1160/ 5-1, Ha 1160/6-1), and the Leibniz Program. Address reprint requests to: Prof. Dr. D. Ha¨ussinger, Medizinische Einrichtungen der Heinrich-Heine-Universita¨t, Klinik fu¨r Gastroenterologie, Hepatologie und Infektiologie, Moorenstrasse 5, 40225 Du¨sseldorf, Germany. Fax: 49-211-8118752. Copyright q 1997 by the American Association for the Study of Liver Diseases. 0270-9139/97/2606-0026$3.00/0
AND
DIETER HA¨ USSINGER
are not fully understood, activation of Kupffer cells has been identified as an important mechanism promoting further liver injury during early reoxygenation in several experimental models following warm4-6 and cold ischemia.7,8 Kupffer cell activation results in increased formation of reactive oxygen species, release of cytokines such as interleukin-1, tumor necrosis factor a (TNF-a), and platelet activating factor, as well as eicosanoids like prostaglandins (PG) and thromboxane.4,5,7-10 These mediators act on other cells in the liver, e.g., PGs and thromboxane disturb microcirculation by vasoconstriction; reactive oxygen species may inactivate antiproteases; and cytokines induce the expression of adhesion molecules promoting the infiltration of neutrophils.11-13 Recently, activation of mitogen-activated protein kinases during reoxygenation following cold ischemia was demonstrated in transplanted rat livers that may be perpetuated by TNF-a.14 The importance of Kupffer cells is further substantiated by the protective effect of Kupffer cell inactivation by gadolinium chloride or methyl palmitate treatment in ischemia-reperfusion injury,5,15 although a recent study questioned these results.16 In rat livers, Kupffer cell function is regulated by changes of ambient osmolarity: endotoxin-induced PGE2 , PGD2 , and thromboxane B2 (TXB2 ) formation and cyclooxygenase-2expression are stimulated seven- to 10-fold when ambient osmolarity increases from 300 to 350 mosmol/L.17 TNF-a production and phagocytosis by Kupffer cells are also sensitive to osmolarity changes.18,19 Recently, betaine and taurine were identified as osmolytes in macrophages including Kupffer cells.20-22 Organic osmolytes are compounds that are accumulated or released by the cells in response to hyperosmotic cell shrinkage or hypo-osmotic cell swelling, respectively, to maintain cell volume homeostasis. Osmolytes must be nonperturbing solutes that do not interfere with protein function even at high intracellular concentrations.23-25 Therefore, only a few classes of organic compounds, i.e., polyols (inositol, sorbitol), methylamines (betaine, a-glycero-phosphocholine), and certain amino acids such as taurine, have evolved as osmolytes in living cells. In mammals, osmolytes have been identified in astrocytes, renal medullary cells, and lens epithelia.26-28 Hyperosmotic exposure induces betaine transporter BGT1 and taurine transporter messenger RNA expression within 3 to 6 hours in a macrophage cell line and Kupffer cells, leading to an intracellular accumulation of betaine and taurine.20-22 In contrast, betaine and taurine are released when ambient osmolarity decreases. Osmolarity effects on prostanoid synthesis, cyclooxygenase-2-expression, phagocytosis, and TNF-a production can be suppressed by betaine.20-22 In view of this recently recognized role of osmolytes as regulators of Kupffer cell function, the current
1560
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
hepa
WBS: Hepatology
WETTSTEIN AND HA¨USSINGER
HEPATOLOGY Vol. 26, No. 6, 1997
1561
study was undertaken to investigate the effect of betaine and taurine in ischemia-reoxygenation injury in the perfused rat liver. MATERIALS AND METHODS Liver Perfusions. Male Wistar rats (range, 120-180 g body weight)
with free access to a stock diet were raised in the local institute for laboratory animals and held according to the local ethical guidelines. Livers were isolated and perfused as described previously29 in a blood-free non-recirculating system with bicarbonate-buffered Krebs-Henseleit saline supplemented with 2.1 mmol/L lactate and 0.3 mmol/L pyruvate. The influent K/-concentration was 5.9 mmol/ L. Betaine and taurine (Sigma, Deisenhofen, Germany) were dissolved in the perfusion buffer. Perfusate flow was 3.5 to 4 mL/g liver/min. The perfusate was equilibrated with O2 /CO2 (95/5 by volume) yielding a PO2 of 523 { 22 mm hg (n Å 4) in the influent (normoxia) as determined with a blood gas analyzer. During the hypoxic perfusion period, perfusate was equilibrated with N2 /CO2 (95/5 by volume), and PO2 was 23 { 4 mm hg (15 minutes after change to N2 /CO2 ). The temperature was 377C. Osmolarity of normotonic perfusion fluid was 305 mosmol/L. In hypo-osmotic perfusions, NaCl concentration in the perfusion buffer was reduced by 40 mmol/L, yielding an osmolarity of 225 mosmol/L. For hyperosmotic perfusions, NaCl was added, yielding the osmolarities indicated. In all experiments, livers were initially perfused with normo-osmolar buffer; aniso-osmolar perfusion was started 15 minutes before hypoxia. In respective experiments, betaine and taurine were added 30 minutes before hypoxia and present throughout the experiments. K/ concentration in the effluent perfusate was monitored with a K/-sensitive electrode (Radiometer, Willich, Germany), and portal pressure was measured continuously with a pressure transducer. Assays. Lactate dehydrogenase (LDH) and aspartate transaminase (AST) in the effluent perfusate were determined in kinetic photometric assays as described.30 TXB2 and PGD2 in the effluent perfusate were measured with radioimmunoassays obtained from Amersham (Braunschweig, Germany), and PGE2 with a radioimmunoassay using [3H]labeled PGE2 (Amersham) and a specific antiserum to PGE2 (Sigma, Deisenhofen, Germany) as described.17 TNF-a was measured with an enzyme-linked immunosorbent assay (Laboserv, Staufenberg, Germany). For determination of endothelial cell function, hyaluronic acid (HA) (200 ng/mL) was infused close to the liver for 20 minutes before hypoxia, during the last 20 minutes of the 90-minute hypoxic period, and at the end of the reoxygenation period. Triplicate samples of influent perfusion buffer and of effluent perfusate taken during the last 10 minutes of each HA infusion were assayed for HA with a radiometric test based on the binding to HA binding proteins (Pharmacia, Freiburg, Germany). HA uptake was calculated from the difference between influent and effluent concentrations. Carbon uptake by the perfused liver was determined as described.31 In brief, Pelikan black ink no. 17 was dialyzed against distilled water for 48 hours and added to the influent perfusate, yielding an absorbance at 578 nm of approximately 1.2, corresponding to a carbon concentration of approximately 0.5 mg/mL. Carbon uptake was calculated from DA578 between influent and effluent perfusate. Steady-state uptake rates were reached within 5 to 10 minutes. Statistics. Data are expressed as means { SEM, with the number of independent experiments as indicated. Statistical analyses were performed using a general linear models procedure for repeatedmeasures ANOVA and the Student’s t test as indicated. P õ .05 was considered significant. RESULTS Osmolarity Dependence of Ischemia-Reoxygenation Injury in the Perfused Rat Liver. Cell volume has been shown to be a major
determinant of cell function.32,33 Cell volume can be modulated by changes in ambient osmolarity or cumulative amino acid uptake. In rat livers, hypo-osmolar perfusion (225 mos-
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
FIG. 1. Osmolarity dependence of anoxia-reoxygenation–induced LDH release in the perfused rat liver. Livers were first perfused with buffer saturated with O2 /CO2 (95/5 vol/vol), then for 60 minutes with N2 /CO2 -saturated buffer (hypoxia), followed by a 90-minute period of reoxygenation. Different osmolarities were obtained by variation of the NaCl concentration in the perfusion buffer. Data represent means { SEM (seven independent experiments with 305 mosmol/L, six with 385 mosmol/L, and four with each of the other conditions). The curves for 345 mosmol/L and 385 mosmol/L were significantly different from 305 mosmol/L (P õ .01, repeated-measures ANOVA). (●) 385 mosmol/L; (l) 345 mosmol/L; (j) 305 mosmol/L; and (m) 225 mosmol/L.
mol/L) increases intracellular water space by approximately 17%, whereas it is decreased by 16% in hyperosmolarity (385 mosmol/L).34,35 To evaluate the effect of osmolarity on the susceptibility to ischemia-reoxygenation injury, rat livers were perfused with buffer of different osmolarity. After 45 minutes of perfusion in the presence of oxygen, the medium was gassed with N2 /CO2 for 60 minutes. LDH release into the effluent perfusate continuously increased during the hypoxic period (Fig. 1). After reoxygenation, there was an initial decrease of LDH leakage, followed again by an increase until the end of the experiment, representing reoxygenation injury. There was a marked osmolarity dependence of liver injury: with a medium osmolarity of 385 mosmol/L, maximal LDH leakage during hypoxia was 1,592 { 823 mU/g liver/ min (mean { SEM) and 1,475 { 172 at the end of reoxygenation (Fig. 1). In normo-osmolarity, LDH release was only 167 { 51 and 198 { 28 mU/g liver/min, respectively, and even somewhat lower with hypo-osmolar buffer. In control experiments without a hypoxic period, maximal LDH release after 3 hours of hyperosmolar perfusion (385 mosmol/L) was 176 { 46 mU/g liver/min (Table 1), indicating that hyperosmolar exposure per se had only a minor effect on LDH leakage, in line with data from a previous report.36 AST release, a more specific marker of parenchymal cell injury, was also higher with hyperosmolar buffer than in normo-osmolarity as determined in an additional series of experiments: after 90 minutes of hypoxia and 90 minutes of reoxygenation, AST release was 53.7 { 3.9 mU/g liver/min at 385 mosmol/ L compared with 25.9 { 4.0 mU/g liver/min at 305 mosmol/ L (n Å 4 each). The data show that liver injury is aggravated
hepa
WBS: Hepatology
1562 WETTSTEIN AND HA¨USSINGER
HEPATOLOGY December 1997
TABLE 1. Release of LDH, AST, PGD2 , PGE2 , TXB2 , and TNF-a and Carbon Uptake Following 90 Minutes of Hypoxia and 90 Minutes of Reoxygenation in Hyperosmolar Perfused Rat Liver
LDH in effluent (mU/g/min) AST in effluent (mU/g/min) PGD2 release (pg/g/min) PGE2 release (pg/g/min) TXB2 release (pg/g/min) TNF-a release (pg/g/min) Carbon uptake (mg/g/min)
Control (Normoxia)
Hypoxia
Hypoxia / 1 mmol/L Taurine / 1 mmol/L Betaine
176 { 46
461 { 18*
166 { 42†
19.3 { 2.8
53.7 { 3.9*
25.8 { 2.9†
ND
6,245 { 634
850 { 63†
ND
1,146 { 691
268 { 31†
ND
836 { 332
526 { 113
135 { 25
234 { 30*
148 { 26†
466 { 19
778 { 74*
471 { 6†
NOTE. Livers were perfused for 30 minutes with normo-osmolar buffer (305 mosmol/L), then with hyperosmolar buffer (385 mosmol/L). After 15 minutes of hyperosmolarity, gassing of the perfusate was changed from O2/ CO2 (95/5 vol/vol) to N2/CO2 for 90 minutes (hypoxia), followed by a 90minute period of reoxygenation. In control experiments (normoxia), buffer was saturated with O2/CO2 throughout the experiment. Values were determined at the end of the reoxygenation period. Basal values before the hypoxic period were below detection limits of the respective tests for all parameters. Data represent means { SEM of four independent experiments for each condition. Abbreviation: ND, not detectable; detection limits for the eicosanoids were approximately 200 pg/g/min. * Significantly different from normoxia experiments (t test, P õ .05). † Significantly different from hypoxia experiments without betaine and taurine (t test, P õ .05).
by hyperosmolar cell shrinkage, but attenuated by hypo-osmolar cell swelling. Protective Effect of Taurine and Betaine. The aggravation of LDH leakage in hyperosmolar perfusions was fully reversed by the addition of betaine or taurine to the perfusion buffer (Fig. 2). With betaine (1 mmol/L) or betaine and taurine (100 mmol/L each), LDH in the effluent perfusate was not higher than in normo-osmolar perfusions. In an additional series of experiments, it was shown that hypoxia and reoxygenation resulted in a significant increase of LDH and AST leakage from the liver compared with normoxic controls (Table 1). This was accompanied by a stimulation of PGD2 , PGE2 , TXB2 , and TNF-a release from the liver. Carbon uptake rate, which is seen as a parameter for Kupffer cell activation,6,31 was also stimulated (Table 1). In the presence of betaine and taurine, the hypoxia-reoxygenation–induced enzyme leakage, as well as eicosanoid production, TNF-a, release and carbon uptake, were significantly lower (Table 1). The volume regulatory K/-uptake induced by changing from normo-osmolar perfusion buffer (305 mosmol/L) to hyperosmolar perfusion (385 mosmol/L) was not significantly different in the presence and absence of osmolytes: net K/-uptake was 15.0 { 2.7 mmol/g liver (n Å 4) without betaine and taurine and 16.8 { 0.8 mmol/g liver (n Å 4) in the presence of betaine and taurine (100 mmol/L each). This suggests that betaine and taurine under the experimental conditions used
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
did not interfere with the ionic mechanism of cell volume regulation in hepatic parenchymal cells, which make up the largest part of the liver cell mass. Betaine and taurine were also protective in normo-osmolar perfusions (305 mosmol/L). Betaine (100 mmol/L) decreased both LDH leakage during hypoxia (maximum 1,442 { 562 mU/g liver/min compared with 2,901 { 481 in controls) and reoxygenation (251 { 46 and 756 { 114 mU/g/min, respectively) (Fig. 3). The effects of taurine and betaine were additive (not shown). AST leakage after 90 minutes of hypoxia and 90 minutes of reoxygenation was 16.7 { 3.8 mU/ g liver/min in the presence of betaine and taurine (100 mmol/ L each) compared with 25.9 { 4.0 mU/g liver/min in the absence of the osmolytes. The increase in vascular resistance during the reoxygenation period was 57% lower in the presence of betaine, demonstrating an attenuation of microcirculatory disturbances (Table 2). The microcirculatory disturbances are probably caused by the induction of eicosanoid formation in Kupffer cells leading to intrahepatic vasoconstriction and local ischemia. In line with this, the cyclooxygenase inhibitor ibuprofen also decreased hypoxia-reoxygenation–induced increase of vascular resistance and LDH leakage (Fig. 4; Table 2). In normo-osmolar perfusions, taurine also decreased PGD2 release during reoxygenation (Fig. 5). Carbon uptake at the end of the reoxygenation period was 507 { 52 mg/g liver/ min in the presence of betaine and taurine (100 mmol/L each) compared with 748 { 77 mg/g liver/min in hypoxiareoxygenation experiments without osmolytes (n Å 4 for each condition). In contrast to hyperosmolarity, in normoosmolar perfusions, PGE2 , TXB2 , and TNF-a release were
FIG. 2. Effect of betaine and taurine on hypoxia-reoxygenation–induced LDH release in hyperosmolar liver perfusion. Data represent means { SEM (six independent control experiments with 385 mosmol/L, four with each of the other conditions). The curves with 100 mmol/L betaine / 100 mmol/L taurine and 1 mmol/L betaine were significantly different from control without osmolyte (P õ .01, repeated-measures ANOVA). (●) 385 mosmol/L; (j) 385 mosmol/L / 100 mmol/L betaine; (m) 385 mosmol/L / 100 mmol/L betaine / 100 mmol/L taurine; and (l) 385 mosmol/L / 1 mmol/L betaine.
hepa
WBS: Hepatology
WETTSTEIN AND HA¨USSINGER
HEPATOLOGY Vol. 26, No. 6, 1997
FIG. 3. Effect of betaine and taurine on hypoxia-reoxygenation–induced LDH release in normo-osmolar liver perfusions. The hypoxic period was 90 minutes. Data represent means { SEM (six independent control experiments without osmolytes, four independent experiments with betaine and taurine, each). The curve for taurine was significantly different from control (P õ .05, repeated-measures ANOVA). (●) 305 mosmol/L; (l) 305 mosmol/L / 100 mmol/L betaine; and (j) 305 mosmol/L / 100 mmol/L taurine.
below the detection limits of the respective tests (not shown). This does not necessarily mean that these compounds are not involved in the pathogenesis of reoxygenation injury in the experimental model used, as the perivenous scavenger cells very effectively eliminate mediators released in the liver sinusoids before reaching the systemic circulation.37 In summary, the decrease of hypoxia-reoxygenation–induced eicosanoid production, cytokine release, and carbon uptake point to an inhibition of Kupffer cell activation during reoxygenation by betaine and taurine, thereby attenuating liver injury.
1563
FIG. 4. Effect of ibuprofen on hypoxia-reoxygenation–induced LDH release in normo-osmolar liver perfusion. The hypoxic period was 90 minutes. Data represent means { SEM (six independent control experiments without ibuprofen, four independent experiments with ibuprofen). The curves were significantly different (P õ .01, repeated-measures ANOVA). (●) 305 mosmol/L; and (m) 305 mosmol/L / 50 mmol/L ibuprofen.
Influence of Anoxia-Reoxygenation and Osmolytes on Endothelial Cell Function. HA uptake and its degradation to lactate and
acetate is regarded as a sensitive and specific parameter of endothelial cell function.38-41 In the perfused rat liver, HA
TABLE 2. Effect of Betaine and Ibuprofen on Increase of Hepatic Vascular Resistance in Hypoxia-Reoxygenation Injury in Normo-osmolar Perfused Rat Liver Basal Vascular Resistance (cm H2O/mL/min)
305 mosmol/L, 90-min hypoxia (control) 305 mosmol/L, 90-min hypoxia / betaine 100 mmol/L 305 mosmol/L, 90-min hypoxia / ibuprofen 50 mmol/L
Increase of Vascular Resistance (cm H2O/mL/min)
0.19 { 0.02 (8)
0.21 { 0.06
0.22 { 0.02 (5)
0.09 { 0.03*
0.15 { 0.03 (5)
0.11 { 0.02*
NOTE. Livers were perfused for approximately 45 minutes with buffer saturated with O2/CO2 (95/5 vol/vol), then for 90 minutes with N2/CO2saturated buffer (hypoxia), followed by a 90-minute period of reoxygenation. Basal vascular resistance (VR) represents perfusion pressure divided by perfusate flow before hypoxia. Increase in VR represents VR at the end of the reoxygenation period minus basal VR. Medium osmolarity was 305 mosmol/L. Data represent means { SEM (number of experiments). * Significantly different from control experiments (t test, P õ .05).
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
FIG. 5. PGD2 release in normo-osmolar liver perfusion. Means { SEM of four experiments each. In the presence of taurine, the values after reoxygenation were significantly lower than controls (P õ .05, repeated-measures ANOVA). (●) 305 mosmol/L; and (m) 305 mosmol/L / 100 mmol/L taurine.
hepa
WBS: Hepatology
1564 WETTSTEIN AND HA¨USSINGER
HEPATOLOGY December 1997
FIG. 6. HA uptake of the perfused rat liver. Livers were perfused for approximately 50 minutes with buffer saturated with O2 /CO2 (95/5 vol/vol), then for 90 minutes with N2 /CO2 -saturated buffer (hypoxia), followed by a 90-minute period of reoxygenation. During the last 20 minutes of each period, HA (200 ng/mL) was infused and HA uptake was determined as described in Materials and Methods. Medium osmolarity was 305 mosmol/ L. Data represent means { SEM of four independent experiments for each condition. *Significantly different from uptake before hypoxia (P õ .01, t test). (j) control; and ( ) 100 mmol/L betaine / 100 mmol/L taurine.
uptake is saturated with concentrations above 150 ng/mL in the influent perfusate.38 To determine the importance of endothelial cell injury and the protective effect of betaine and taurine, HA (200 ng/mL) was infused for 20-minute periods before hypoxia, during the last 20 minutes of the 90minute hypoxic period, and during the last 20 minutes of the reoxygenation period. HA uptake was calculated from the difference between influent and effluent concentrations. Basal HA uptake was 277 { 51 ng/g liver/min (mean { SEM, n Å 4) in control livers without osmolytes and 302 { 30 ng/ g liver/min in the presence of betaine and taurine (100 mmol/ L each) (Fig. 6). At the end of the hypoxic perfusion, HA uptake decreased to 174 { 57 and 164 { 23 ng/g liver/min, respectively, and, after 90 minutes of reoxygenation, it was 193 { 29 and 210 { 76 ng/g liver/min, respectively. There were no significant differences between controls and experiments in the presence of betaine and taurine, suggesting that the protective effect of these osmolytes is probably not caused by attenuation of endothelial cell injury. DISCUSSION
Cell volume is an important determinant of liver cell function: cell swelling is an anabolic and proliferative signal, whereas cell shrinkage is catabolic and antiproliferative.32,33 The data presented here show an osmolarity dependence of cell injury in a rat liver model of ischemia-reoxygenation: LDH leakage was approximately 10 times higher with hyperosmolar perfusion buffer than with normo-osmolar medium and further decreased with hypo-osmolar buffer. This is in line with a previous report that t-butylhydroperoxide–induced cell injury is also aggravated by ambient hyperosmolarity.36 Several factors may contribute to osmolarity dependence of cell injury. TNF-a production, cyclooxygenase-2 expression, and PG synthesis are increased by hyperosmolar-
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
ity and decreased in hypo-osmolarity.17,18 Hypo-osmolarity stimulates bile secretion and is accompanied by the activation of mitogen-activated protein kinases in liver, which is also seen as a protective mechanism.42 In addition, the canalicular elimination of mediators of inflammation such as cysteinyl leukotrienes is decreased by hyperosmolarity and increased in hypo-osmolarity.43 The osmolarity dependence of cell injury may be of importance in view of the finding that severe illness and catabolic states are accompanied by a decrease of cellular water content.33 Cell shrinkage may therefore not only promote catabolism, but also lead to an increased susceptibility toward ischemic injury. The beneficial effect of cell volume increase induced by hypo-osmolar solutions in warm ischemia-reoxygenation or t-butylhydroperoxide–induced cell injury and the aggravation of liver injury by hyperosmolarity must be separated from cell swelling caused by loss of cell function. The situation in warm ischemia may also be different from cold ischemia, because solutions containing impermeant agents such as lactobionate, raffinose, and hydroxyethyl starch (e.g., UW solution) are used to prevent hypoxic swelling during cold organ storage.44 Several compounds have been shown to attenuate liver injury in experimental models of warm ischemia and reoxygenation, including L-arginine, calcium channel blockers, anti–intercellular adhesion molecule-1 antibodies, and PGE1 .45-48 In accordance with the potential importance of Kupffer cell activation for reoxygenation injury, inactivation of Kupffer cells by methyl palmitate or gadolinium chloride pretreatment was also protective.5 The data presented here show a protective effect of betaine and taurine in warm ischemia-reoxygenation injury. The presence of betaine or taurine in the perfusion medium in near-physiological concentrations decreases LDH and AST leakage and attenuates microcirculatory disturbances. Several mechanisms may be involved in this protective effect. In hyperosmolar perfusions, betaine and taurine may be accumulated intracellularly, thereby compensating for the hyperosmolarity-induced cell shrinkage. The physiological concentration of betaine in blood is 20 to 60 mmol/L, and it can be accumulated in Kupffer cells to concentrations of 60 to 70 mmol/L.21 However, in hepatic parenchymal cells, betaine is of minor importance as an osmolyte, because betaine uptake rates are low and betaine transporter BGT1 is not expressed.22 In addition, hyperosmolarity-induced volumeregulatory K/-release was not different in the presence and absence of betaine and taurine (100 mmol/L each). This makes an interference of these osmolytes with cell volume regulation in parenchymal cells unlikely, which make up the largest part of the liver cell mass, but does not exclude such a mechanism in a smaller compartment as Kupffer cells. For taurine, a role as an antioxidant, in conjugation reactions and as a membrane stabilisator, has been described.49 The antioxidative potential of taurine is based on the formation of chloramines.50 Taurine protects isolated hepatocytes against toxic compounds including carbon tetrachloride, although the mechanisms of this effect remain unclear.51 Osmolytes were shown to stabilize protein structures by prevention of denaturation.52 However, the protective effect in the experimental model used here seems to be caused by the modulation of Kupffer cell function. Betaine and taurine were recently identified as osmolytes in a macrophage cell line and rat Kupffer cells.20-22 Kupffer cell functions such as cyclooxygenase-2 expression,
hepa
WBS: Hepatology
WETTSTEIN AND HA¨USSINGER
HEPATOLOGY Vol. 26, No. 6, 1997
phagocytosis, and PG and TNF-a production are modulated by ambient osmolarity and osmolytes.17-19 In the study presented here, hypoxia-reoxygenation increased the release of PGs, thromboxane, and TNF-a and stimulated carbon uptake, indicating activation of Kupffer cells. These effects were more pronounced in hyperosmolar perfusions than under normo-osmolar conditions. The Kupffer cell activation was suppressed in the presence of betaine and taurine. Although the importance of the various cytokines, eicosanoids, and reactive oxygen species released during reoxygenation is not clear and may vary between different experimental models, activation of arachidonic acid metabolism by lipid peroxidation is seen as an important mechanism.53 Hydroperoxide was shown to induce the release of eicosanoids from the perfused rat liver.54 Some of the eicosanoids released by nonparenchymal cells such as PGE2 and prostacyclin are potentially protective, whereas others such as thromboxane, PGD2 and 6-keto-PGF2a are considered to promote liver injury by local vasoconstriction.46,55 In our study, hypoxiareoxygenation also stimulated the formation of the potentially protective PGE2 , but the other mediators released seem to override this effect. In addition, recent experiments in isolated hepatocytes showed an increased susceptibility toward oxidative stress in the presence of PGE2 (vom Dahl S, Kubitz R, Ha¨ussinger D, unpublished data, December 1996). The importance of eicosanoid formation for the reoxygenation injury is substantiated by the protective effect of the cyclo-oxygenase inhibitor, ibuprofen, in line with previous studies showing attenuation of ischemic injury by indomethacin or prostacyclin infusion.56,57 HA uptake was decreased by approximately 35% to 40% during anoxia and reoxygenation. HA is a high-molecularweight polysaccharide that is removed from circulation specifically by liver endothelial cells via a receptor-mediated endocytotic process.40,41 However, there was no significant difference with and without osmolytes and no correlation to LDH leakage. Therefore, in this model of ischemia-reoxygenation injury, endothelial cell damage appears to be less important. This is in line with previous findings that sinusoidal cells are more susceptible to cold ischemia than hepatic parenchymal cells, whereas warm ischemia induces parenchymal cell injury.7,39,58 The protective effect of betaine and taurine in ischemiareoxygenation may gain clinical importance in organ preservation for transplantation. A hepatoprotective role of betaine was also shown in a recent study demonstrating that supplemental dietary betaine prevented ethanol-induced fatty infiltration and increased S-adenosylmethionine levels in rat livers.59 The potential of osmolytes to modify Kupffer cell function provides new perspectives in many pathophysiological conditions involving nonparenchymal cells, because other known modulators of Kupffer cell function are toxic, which excludes their application in humans.
4. 5. 6.
7. 8. 9. 10.
11.
12. 13.
14.
15.
16. 17.
18. 19. 20. 21. 22. 23.
Acknowledgment: The expert technical assistance of Claudia Holneicher is gratefully acknowledged.
24.
REFERENCES
25.
1. Jaeschke H. Preservation injury: mechanism, prevention and consequences. J Hepatol 1996;25:774-780. 2. Bronk SF, Gores GJ. pH-dependent nonlysosomal proteolysis contributes to lethal anoxic injury of rat hepatocytes. Am J Physiol 1993;264: G744-G751. 3. Calmus Y, Cynoper L, Dousset B, Lim SK, Soubrane O, Conti F, Houssin
26.
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
27.
1565
D, et al. Evidence for the detrimental role of proteolysis during liver preservation in humans. Gastroenterology 1995;108:1510-1516. Jaeschke H, Bautista AP, Spolarics Z, Spitzer JJ. Superoxide generation by Kupffer cells and priming of neutrophils during reperfusion after hepatic ischemia. Free Radic Res Commun 1991;15:277-284. Jaeschke H, Farhood A. Neutrophil and Kupffer cell–induced oxidant stress and ischemia-reperfusion injury in rat liver in vivo. Am J Physiol 1991;260:G355-G362. Lindert KA, Caldwell-Kenkel JC, Nukina S, Lemasters JJ, Thurman RG. Activation of Kupffer cells on reperfusion following hypoxia: particle phagocytosis in a low-flow, reflow model. Am J Physiol 1992;262:G345G350. Caldwell-Kenkel JC, Currin RT, Tanaka Y, Thurman RG, Lemasters JJ. Kupffer cell activation and endothelial cell damage after storage of rat livers: effects of reperfusion. HEPATOLOGY 1991;13:83-95. Rymsa B, Wang JF, de Groot H. O2 -release by activated Kupffer cells upon hypoxia-reoxygenation. Am J Physiol 1991;261:G602-G607. Zhou W, McCollum MO, Levine B, Olson MS. Inflammation and platelet-activating factor production during hepatic ischemia-reperfusion. HEPATOLOGY 1992;16:1236-1240. Colletti LM, Remick DG, Burtch GD, Kunkel SL, Strieter RM, Campbell DA. Role of tumor necrosis factor-a in the pathophysiologic alterations after hepatic ischemia-reperfusion injury in the rat. J Clin Invest 1990; 85:1936-1943. Scoazec JY, Durand F, Degott C, Delautier D, Bernuau J, Belghiti J, Benhamou JP, et al. Expression of cytokine-dependent adhesion molecules in postreperfusion biopsy specimens of liver allograft. Gastroenterology 1994;107:1094-1102. Vollmar B, Glasz J, Leiderer R, Post S, Menger MD. Hepatic microcirculatory failure is a determinant of liver dysfunction in warm ischemiareperfusion. Am J Pathol 1994;145:1421-1431. Hanazaki K, Kuroda T, Haba Y, Shiohara E, Kajikawa S, Iida F. Protective effect of the thromboxane A2 receptor antagonist ONO 3708 on ischemia-reperfusion injury in the dog liver. Surgery 1994;116:896903. Bradham CA, Stachlewitz RF, Gao W, Qian T, Jayadev S, Jenkins G, Yusuf H, et al. Reperfusion after liver transplantation in rats differentially activates the mitogen-activated protein kinases. HEPATOLOGY 1997;25:1128-1135. Marzi I, Cowper K, Takei Y, Lindert K, Lemasters JJ, Thurman RG. Methyl palmitate prevents Kupffer cell activation and improves survival after orthotopic liver transplantation in the rat. Transplant Int 1991;4: 215-220. Imamura H, Sutto F, Brault A, Huet P-M. Role of Kupffer cells in cold ischemia-reperfusion injury of rat liver. Gastroenterology 1995;109: 189-197. Zhang F, Warskulat U, Wettstein M, Schreiber R, Henninger HP, Decker K, Ha¨ussinger D. Hyperosmolarity stimulates prostaglandin synthesis and cyclooxygenase-2 expression in activated rat liver macrophages. Biochem J 1995;312:135-143. Zhang F, Warskulat U, Ha¨ussinger D. Modulation of tumor necrosis factor-a release by anisoosmolarity and betaine in rat liver macrophages (Kupffer cells). FEBS Lett 1996;391:293-296. Warskulat U, Zhang F, Ha¨ussinger D. Modulation of phagocytosis by anisoomolarity and betaine in rat liver macrophages (Kupffer cells) and RAW 264.7 mouse macrophages. FEBS Lett 1996;391:287-292. Warskulat U, Wettstein M, Ha¨ussinger D. Betaine is an osmolyte in RAW 264.7 mouse macrophages. FEBS Lett 1995;377:47-50. Zhang F, Warskulat U, Wettstein M, Ha¨ussinger D. Identification of betaine as an osmolyte in rat liver macrophages (Kupffer cells). Gastroenterology 1996;110:1543-1552. Warskulat U, Wettstein M, Ha¨ussinger D. Osmoregulated taurine transport in H4IIE hepatoma cells and perfused rat liver. Biochem J 1997; 321:683-690. Yancey PH, Clark ME, Hand SC, Bowlus RD, Somero GN. Living with water stress: evolution of osmolyte systems. Science 1982;217:12141222. Handler JS, Kwon HM. Regulation of renal cell organic osmolyte transport by tonicity. Am J Physiol 1993;265:C1449-C1455. Burg MB. Molecular basis of osmotic regulation. Am J Physiol 1995; 268:F983-F996. Nakanishi, T, Turner RJ, Burg MB. Osmoregulation of betaine transport in mammalian renal medullary cells. Am J Physiol 1990;258:F1061F1067. Zhou C, Agarwal N, Cammata PR. Osmoregulatory alterations in myoinositol uptake by bovine lens epithelial cells. Invest Ophthalmol Vis Sci 1994;35:1236-1242.
hepa
WBS: Hepatology
1566 WETTSTEIN AND HA¨USSINGER
HEPATOLOGY December 1997
28. Paredes A, McManus M, Kwon HM, Strange, K. Osmoregulation of the Na/-inositol cotransporter activity and mRNA levels in brain glial cells. Am J Physiol 1993;263:C1282-C1288. 29. Sies H. The use of perfusion of liver and other organs for the study of microsomal electron transport and cytochrom P-450 systems. Methods Enzymol 1978;52:48-59. 30. Bergmeyer HU, ed. Methods of Enzymatic Analysis. 3rd ed. Weinheim: VCH Verlagsgesellschaft, 1983. 31. Cowper KB, Currin RT, Dawson TL, Lindert KA, Lemasters JJ, Thurman RG. A new method to monitor Kupffer-cell function continuously in the perfused rat liver. Biochem J 1990;266:141-147. 32. Ha¨ussinger D, Gerok W, Lang F. Regulation of cell function by the cellular hydration state. Am J Physiol 1994;267:E343-E355. 33. Ha¨ussinger D, Roth E, Lang F, Gerok W. Cellular hydration state: an important determinant of protein catabolism in health and disease. Lancet 1993;341:1330-1332. 34. Vom Dahl S, Hallbrucker C, Lang F, Gerok W, Ha¨ussinger D. Regulation of liver cell volume and proteolysis by glucagon and insulin. Biochem J 1991;278:771-777. 35. Vom Dahl S, Ha¨ussinger D. Nutritional state and the swelling-induced inhibition of proteolysis in perfused rat liver. J Nutr 1996;126:395-402. 36. Saha N, Stoll B, Lang F, Ha¨ussinger D. Effect of anisotonic cell-volume modulation on glutathione-S-conjugate release, t-butylhydroperoxide metabolism and the pentose-phosphate shunt in perfused rat liver. Eur J Biochem 1992;209:437-444. 37. Ha¨ussinger D, Stehle T. Hepatocyte heterogeneity in response to icosanoids: the perivenous scavenger cell hypothesis. Eur J Biochem 1988; 175:395-403. 38. Deaciuc IV, Bagby GJ, Lang CH, Spitzer JJ. Hyaluronic acid uptake by the isolated, perfused rat liver: an index of hepatic sinusoidal endothelial cell function. HEPATOLOGY 1993;17:266-272. 39. Sutto F, Brault A, Lepage R, Huet P-M. Metabolism of hyaluronic acid by liver endothelial cells: effect of ischemia-reperfusion in the isolated perfused rat liver. J Hepatol 1994;20:611-616. 40. Smedsrod B, Pertoft H, Eriksson S, Fraser RE, Laurent TC. Studies in vitro on the uptake and degradation of sodium hyaluronate in rat liver endothelial cells. Biochem J 1984;223:617-626. 41. McGary CT, Raja RH, Weigel PH. Endocytosis of hyaluronic acid by rat liver endothelial cells. Biochem J 1989;257:875-884. 42. Noe´ B, Schliess F, Wettstein M, Heinrich S, Ha¨ussinger D. Regulation of taurocholate excretion by a hypo-osmolarity–activated signal transduction pathway in rat liver. Gastroenterology 1996;110:858-865. 43. Wettstein M, Noe´ B, Ha¨ussinger D. Metabolism of cysteinyl leukotrienes in the perfused rat liver: the influence of endotoxin pretreatment and the cellular hydration state. HEPATOLOGY 1995;22:235-240. 44. Southard JH, Belzer FO. Organ preservation. Annu Rev Med 1995;46: 235-247.
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
45. Jones SM, Thurman RG. L-Arginine minimizes reperfusion injury in a low-flow, reflow model of liver perfusion. HEPATOLOGY 1996;24:163168. 46. Grafa F, Gugenheim J, Saint-Paul M-C, Cavanel C, Lapalus F, Ouzan D, Militerno G, et al. Protective effect of prostaglandin E1 on normothermic liver ischemia. Eur Surg Res 1991;23:278-284. 47. Nakano H, Kuzume M, Namatame K, Yamaguchi M, Kumada K. Efficacy of intraportal injection of anti ICAM-1 monoclonal antibody against liver cell injury following warm ischemia in the rat. Am J Surg 1995; 170:64-66. 48. Nauta RJ, Tsimoyiannis E, Uribe M, Walsh DB, Miller D, Butterfield A. The role of calcium ions and calcium channel blockers in experimental ischemia-reperfusion–induced liver injury. Ann Surg 1991;213:137142. 49. Huxtable RJ. Physiological actions of taurine. Physiol Rev 1992;72:101163. 50. Weiss SJ, Klein R, Slivka A, Wie M. Chlorination of taurine by human neutrophils—evidence for hypochlorous acid generation. J Clin Invest 1982;70:598-607. 51. Timbrell JA, Seabra V, Waterfield CJ. The in vivo and in vitro protective properties of taurine. Gen Pharmacol 1995;26:453-462. 52. Liu Y, Bolen DW. The peptide backbone plays a dominant role in protein stabilization by naturally occurring osmolytes. Biochemistry 1995;34: 12884-12891. 53. Mathews WR, Guido DM, Fisher MA, Jaeschke H. Lipid peroxidation as molecular mechanism of liver cell injury during reperfusion after ischemia. Free Radic Biol Med 1994;16:763-770. 54. Hallbrucker C, Ritter M, Lang F, Gerok W, Ha¨ussinger D. Hydroperoxide metabolism in rat liver: K/ channel activation, cell volume changes and eicosanoid formation. Eur J Biochem 1993;211:449-458. 55. Post S, Goerig M, Otto G, Manner M, Senninger N, Kommerell B, Herfarth C. Prostanoid release in experimental liver transplantation. Transplantation 1990;49:490-494. 56. Briegel J, Haller M, Zulke C. Primary graft nonfunction following orthotopic liver transplantation: treatment with prostacyclin. Transplant Proc 1992;24:2693-2670. 57. Pittet JF, Morel DR, Mentha G, LeCoultre C, Suter PM, Rohner A. Protective effect of indomethacin in the development of the postreperfusion syndrome during liver transplantation in pigs. Transplant Proc 1991;23:2290-2296. 58. McKeown CMB, Edwards V, Phillips MJ, Harvey PRC, Petrunka CN, Strasberg SM. Sinusoidal lining damage: the critical injury in cold preservation of allografts in the rat. Transplantation 1988;46:178-191. 59. Barak AJ, Beckenhauer HC, Tuma DJ. S-Adenosylmethionine generation and prevention of alcoholic fatty liver by betaine. Alcohol 1994;11:501503.
hepa
WBS: Hepatology
Medium osmolarity sensitively regulates Kupffer cell functions like phagocytosis and prostaglandin (PG) and cytokine production. Betaine and taurine, recently identified as osmolytes in liver cells, interfere with these effects. Because Kupffer cell activation is an important pathogenic mechanism in ischemia-reoxygenation injury, the influence of osmolarity and osmolytes was investigated in a rat liver perfusion model of warm ischemia. Livers were perfused with different medium osmolarities for 60 to 90 minutes in the absence of oxygen, followed by another 90 minutes of reoxygenation. Lactate dehydrogenase (LDH) leakage into the effluent perfusate during the hypoxic and the reoxygenation period was eight- to 10-fold higher with a medium osmolarity of 385 mosmol/L than in normo-osmolarity, and further decreased with hypoosmolar perfusion buffer. Betaine and taurine addition to the perfusate in near physiological concentrations decreased hypoxia-reoxygenation–induced LDH leakage, aspartate transaminase (AST) leakage, and perfusion pressure increase in hyperosmolar and normo-osmolar perfusions. Stimulation of PGD2 , PGE2 , thromboxane B2 (TXB2 ), and tumor necrosis factor a (TNF-a) release, as well as induction of carbon uptake by the liver during reoxygenation, were suppressed by betaine and taurine, pointing to an interference of these osmolytes with Kupffer cell function. In contrast, endothelial cell function as assessed by hyaluronic acid (HA) uptake was not influenced. It is concluded that warm ischemia-reoxygenation injury in rat liver is aggravated by hyperosmolarity and attenuated by hypo-osmolarity. The osmolytes betaine and taurine have a protective effect, presumably by inhibition of Kupffer cell activation. (HEPATOLOGY 1997;26:1560-1566.) Ischemic liver injury is of major clinical relevance during organ transplantation and in shock syndromes. Loss of cell function and cell injury during the ischemic period are thought to be caused by adenosine triphosphate depletion, activation of nonlysosomal proteases, and glucose depletion.1-3 Although the exact pathophysiological mechanisms
Abbreviations: TNF-a, tumor necrosis factor a; PG, prostaglandin; TXB2 , thromboxane B2 ; LDH, lactate dehydrogenase; AST, aspartate transaminase; HA, hyaluronic acid. From the Clinic for Gastroenterology, Hepatology, and Infectiology, HeinrichHeine-University, Du¨sseldorf, Germany. Received January 16, 1997; accepted July 24, 1997. Supported by the Deutsche Forschungsgemeinschaft (grants We 1936/1-1, Ha 1160/ 5-1, Ha 1160/6-1), and the Leibniz Program. Address reprint requests to: Prof. Dr. D. Ha¨ussinger, Medizinische Einrichtungen der Heinrich-Heine-Universita¨t, Klinik fu¨r Gastroenterologie, Hepatologie und Infektiologie, Moorenstrasse 5, 40225 Du¨sseldorf, Germany. Fax: 49-211-8118752. Copyright q 1997 by the American Association for the Study of Liver Diseases. 0270-9139/97/2606-0026$3.00/0
AND
DIETER HA¨ USSINGER
are not fully understood, activation of Kupffer cells has been identified as an important mechanism promoting further liver injury during early reoxygenation in several experimental models following warm4-6 and cold ischemia.7,8 Kupffer cell activation results in increased formation of reactive oxygen species, release of cytokines such as interleukin-1, tumor necrosis factor a (TNF-a), and platelet activating factor, as well as eicosanoids like prostaglandins (PG) and thromboxane.4,5,7-10 These mediators act on other cells in the liver, e.g., PGs and thromboxane disturb microcirculation by vasoconstriction; reactive oxygen species may inactivate antiproteases; and cytokines induce the expression of adhesion molecules promoting the infiltration of neutrophils.11-13 Recently, activation of mitogen-activated protein kinases during reoxygenation following cold ischemia was demonstrated in transplanted rat livers that may be perpetuated by TNF-a.14 The importance of Kupffer cells is further substantiated by the protective effect of Kupffer cell inactivation by gadolinium chloride or methyl palmitate treatment in ischemia-reperfusion injury,5,15 although a recent study questioned these results.16 In rat livers, Kupffer cell function is regulated by changes of ambient osmolarity: endotoxin-induced PGE2 , PGD2 , and thromboxane B2 (TXB2 ) formation and cyclooxygenase-2expression are stimulated seven- to 10-fold when ambient osmolarity increases from 300 to 350 mosmol/L.17 TNF-a production and phagocytosis by Kupffer cells are also sensitive to osmolarity changes.18,19 Recently, betaine and taurine were identified as osmolytes in macrophages including Kupffer cells.20-22 Organic osmolytes are compounds that are accumulated or released by the cells in response to hyperosmotic cell shrinkage or hypo-osmotic cell swelling, respectively, to maintain cell volume homeostasis. Osmolytes must be nonperturbing solutes that do not interfere with protein function even at high intracellular concentrations.23-25 Therefore, only a few classes of organic compounds, i.e., polyols (inositol, sorbitol), methylamines (betaine, a-glycero-phosphocholine), and certain amino acids such as taurine, have evolved as osmolytes in living cells. In mammals, osmolytes have been identified in astrocytes, renal medullary cells, and lens epithelia.26-28 Hyperosmotic exposure induces betaine transporter BGT1 and taurine transporter messenger RNA expression within 3 to 6 hours in a macrophage cell line and Kupffer cells, leading to an intracellular accumulation of betaine and taurine.20-22 In contrast, betaine and taurine are released when ambient osmolarity decreases. Osmolarity effects on prostanoid synthesis, cyclooxygenase-2-expression, phagocytosis, and TNF-a production can be suppressed by betaine.20-22 In view of this recently recognized role of osmolytes as regulators of Kupffer cell function, the current
1560
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
hepa
WBS: Hepatology
WETTSTEIN AND HA¨USSINGER
HEPATOLOGY Vol. 26, No. 6, 1997
1561
study was undertaken to investigate the effect of betaine and taurine in ischemia-reoxygenation injury in the perfused rat liver. MATERIALS AND METHODS Liver Perfusions. Male Wistar rats (range, 120-180 g body weight)
with free access to a stock diet were raised in the local institute for laboratory animals and held according to the local ethical guidelines. Livers were isolated and perfused as described previously29 in a blood-free non-recirculating system with bicarbonate-buffered Krebs-Henseleit saline supplemented with 2.1 mmol/L lactate and 0.3 mmol/L pyruvate. The influent K/-concentration was 5.9 mmol/ L. Betaine and taurine (Sigma, Deisenhofen, Germany) were dissolved in the perfusion buffer. Perfusate flow was 3.5 to 4 mL/g liver/min. The perfusate was equilibrated with O2 /CO2 (95/5 by volume) yielding a PO2 of 523 { 22 mm hg (n Å 4) in the influent (normoxia) as determined with a blood gas analyzer. During the hypoxic perfusion period, perfusate was equilibrated with N2 /CO2 (95/5 by volume), and PO2 was 23 { 4 mm hg (15 minutes after change to N2 /CO2 ). The temperature was 377C. Osmolarity of normotonic perfusion fluid was 305 mosmol/L. In hypo-osmotic perfusions, NaCl concentration in the perfusion buffer was reduced by 40 mmol/L, yielding an osmolarity of 225 mosmol/L. For hyperosmotic perfusions, NaCl was added, yielding the osmolarities indicated. In all experiments, livers were initially perfused with normo-osmolar buffer; aniso-osmolar perfusion was started 15 minutes before hypoxia. In respective experiments, betaine and taurine were added 30 minutes before hypoxia and present throughout the experiments. K/ concentration in the effluent perfusate was monitored with a K/-sensitive electrode (Radiometer, Willich, Germany), and portal pressure was measured continuously with a pressure transducer. Assays. Lactate dehydrogenase (LDH) and aspartate transaminase (AST) in the effluent perfusate were determined in kinetic photometric assays as described.30 TXB2 and PGD2 in the effluent perfusate were measured with radioimmunoassays obtained from Amersham (Braunschweig, Germany), and PGE2 with a radioimmunoassay using [3H]labeled PGE2 (Amersham) and a specific antiserum to PGE2 (Sigma, Deisenhofen, Germany) as described.17 TNF-a was measured with an enzyme-linked immunosorbent assay (Laboserv, Staufenberg, Germany). For determination of endothelial cell function, hyaluronic acid (HA) (200 ng/mL) was infused close to the liver for 20 minutes before hypoxia, during the last 20 minutes of the 90-minute hypoxic period, and at the end of the reoxygenation period. Triplicate samples of influent perfusion buffer and of effluent perfusate taken during the last 10 minutes of each HA infusion were assayed for HA with a radiometric test based on the binding to HA binding proteins (Pharmacia, Freiburg, Germany). HA uptake was calculated from the difference between influent and effluent concentrations. Carbon uptake by the perfused liver was determined as described.31 In brief, Pelikan black ink no. 17 was dialyzed against distilled water for 48 hours and added to the influent perfusate, yielding an absorbance at 578 nm of approximately 1.2, corresponding to a carbon concentration of approximately 0.5 mg/mL. Carbon uptake was calculated from DA578 between influent and effluent perfusate. Steady-state uptake rates were reached within 5 to 10 minutes. Statistics. Data are expressed as means { SEM, with the number of independent experiments as indicated. Statistical analyses were performed using a general linear models procedure for repeatedmeasures ANOVA and the Student’s t test as indicated. P õ .05 was considered significant. RESULTS Osmolarity Dependence of Ischemia-Reoxygenation Injury in the Perfused Rat Liver. Cell volume has been shown to be a major
determinant of cell function.32,33 Cell volume can be modulated by changes in ambient osmolarity or cumulative amino acid uptake. In rat livers, hypo-osmolar perfusion (225 mos-
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
FIG. 1. Osmolarity dependence of anoxia-reoxygenation–induced LDH release in the perfused rat liver. Livers were first perfused with buffer saturated with O2 /CO2 (95/5 vol/vol), then for 60 minutes with N2 /CO2 -saturated buffer (hypoxia), followed by a 90-minute period of reoxygenation. Different osmolarities were obtained by variation of the NaCl concentration in the perfusion buffer. Data represent means { SEM (seven independent experiments with 305 mosmol/L, six with 385 mosmol/L, and four with each of the other conditions). The curves for 345 mosmol/L and 385 mosmol/L were significantly different from 305 mosmol/L (P õ .01, repeated-measures ANOVA). (●) 385 mosmol/L; (l) 345 mosmol/L; (j) 305 mosmol/L; and (m) 225 mosmol/L.
mol/L) increases intracellular water space by approximately 17%, whereas it is decreased by 16% in hyperosmolarity (385 mosmol/L).34,35 To evaluate the effect of osmolarity on the susceptibility to ischemia-reoxygenation injury, rat livers were perfused with buffer of different osmolarity. After 45 minutes of perfusion in the presence of oxygen, the medium was gassed with N2 /CO2 for 60 minutes. LDH release into the effluent perfusate continuously increased during the hypoxic period (Fig. 1). After reoxygenation, there was an initial decrease of LDH leakage, followed again by an increase until the end of the experiment, representing reoxygenation injury. There was a marked osmolarity dependence of liver injury: with a medium osmolarity of 385 mosmol/L, maximal LDH leakage during hypoxia was 1,592 { 823 mU/g liver/ min (mean { SEM) and 1,475 { 172 at the end of reoxygenation (Fig. 1). In normo-osmolarity, LDH release was only 167 { 51 and 198 { 28 mU/g liver/min, respectively, and even somewhat lower with hypo-osmolar buffer. In control experiments without a hypoxic period, maximal LDH release after 3 hours of hyperosmolar perfusion (385 mosmol/L) was 176 { 46 mU/g liver/min (Table 1), indicating that hyperosmolar exposure per se had only a minor effect on LDH leakage, in line with data from a previous report.36 AST release, a more specific marker of parenchymal cell injury, was also higher with hyperosmolar buffer than in normo-osmolarity as determined in an additional series of experiments: after 90 minutes of hypoxia and 90 minutes of reoxygenation, AST release was 53.7 { 3.9 mU/g liver/min at 385 mosmol/ L compared with 25.9 { 4.0 mU/g liver/min at 305 mosmol/ L (n Å 4 each). The data show that liver injury is aggravated
hepa
WBS: Hepatology
1562 WETTSTEIN AND HA¨USSINGER
HEPATOLOGY December 1997
TABLE 1. Release of LDH, AST, PGD2 , PGE2 , TXB2 , and TNF-a and Carbon Uptake Following 90 Minutes of Hypoxia and 90 Minutes of Reoxygenation in Hyperosmolar Perfused Rat Liver
LDH in effluent (mU/g/min) AST in effluent (mU/g/min) PGD2 release (pg/g/min) PGE2 release (pg/g/min) TXB2 release (pg/g/min) TNF-a release (pg/g/min) Carbon uptake (mg/g/min)
Control (Normoxia)
Hypoxia
Hypoxia / 1 mmol/L Taurine / 1 mmol/L Betaine
176 { 46
461 { 18*
166 { 42†
19.3 { 2.8
53.7 { 3.9*
25.8 { 2.9†
ND
6,245 { 634
850 { 63†
ND
1,146 { 691
268 { 31†
ND
836 { 332
526 { 113
135 { 25
234 { 30*
148 { 26†
466 { 19
778 { 74*
471 { 6†
NOTE. Livers were perfused for 30 minutes with normo-osmolar buffer (305 mosmol/L), then with hyperosmolar buffer (385 mosmol/L). After 15 minutes of hyperosmolarity, gassing of the perfusate was changed from O2/ CO2 (95/5 vol/vol) to N2/CO2 for 90 minutes (hypoxia), followed by a 90minute period of reoxygenation. In control experiments (normoxia), buffer was saturated with O2/CO2 throughout the experiment. Values were determined at the end of the reoxygenation period. Basal values before the hypoxic period were below detection limits of the respective tests for all parameters. Data represent means { SEM of four independent experiments for each condition. Abbreviation: ND, not detectable; detection limits for the eicosanoids were approximately 200 pg/g/min. * Significantly different from normoxia experiments (t test, P õ .05). † Significantly different from hypoxia experiments without betaine and taurine (t test, P õ .05).
by hyperosmolar cell shrinkage, but attenuated by hypo-osmolar cell swelling. Protective Effect of Taurine and Betaine. The aggravation of LDH leakage in hyperosmolar perfusions was fully reversed by the addition of betaine or taurine to the perfusion buffer (Fig. 2). With betaine (1 mmol/L) or betaine and taurine (100 mmol/L each), LDH in the effluent perfusate was not higher than in normo-osmolar perfusions. In an additional series of experiments, it was shown that hypoxia and reoxygenation resulted in a significant increase of LDH and AST leakage from the liver compared with normoxic controls (Table 1). This was accompanied by a stimulation of PGD2 , PGE2 , TXB2 , and TNF-a release from the liver. Carbon uptake rate, which is seen as a parameter for Kupffer cell activation,6,31 was also stimulated (Table 1). In the presence of betaine and taurine, the hypoxia-reoxygenation–induced enzyme leakage, as well as eicosanoid production, TNF-a, release and carbon uptake, were significantly lower (Table 1). The volume regulatory K/-uptake induced by changing from normo-osmolar perfusion buffer (305 mosmol/L) to hyperosmolar perfusion (385 mosmol/L) was not significantly different in the presence and absence of osmolytes: net K/-uptake was 15.0 { 2.7 mmol/g liver (n Å 4) without betaine and taurine and 16.8 { 0.8 mmol/g liver (n Å 4) in the presence of betaine and taurine (100 mmol/L each). This suggests that betaine and taurine under the experimental conditions used
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
did not interfere with the ionic mechanism of cell volume regulation in hepatic parenchymal cells, which make up the largest part of the liver cell mass. Betaine and taurine were also protective in normo-osmolar perfusions (305 mosmol/L). Betaine (100 mmol/L) decreased both LDH leakage during hypoxia (maximum 1,442 { 562 mU/g liver/min compared with 2,901 { 481 in controls) and reoxygenation (251 { 46 and 756 { 114 mU/g/min, respectively) (Fig. 3). The effects of taurine and betaine were additive (not shown). AST leakage after 90 minutes of hypoxia and 90 minutes of reoxygenation was 16.7 { 3.8 mU/ g liver/min in the presence of betaine and taurine (100 mmol/ L each) compared with 25.9 { 4.0 mU/g liver/min in the absence of the osmolytes. The increase in vascular resistance during the reoxygenation period was 57% lower in the presence of betaine, demonstrating an attenuation of microcirculatory disturbances (Table 2). The microcirculatory disturbances are probably caused by the induction of eicosanoid formation in Kupffer cells leading to intrahepatic vasoconstriction and local ischemia. In line with this, the cyclooxygenase inhibitor ibuprofen also decreased hypoxia-reoxygenation–induced increase of vascular resistance and LDH leakage (Fig. 4; Table 2). In normo-osmolar perfusions, taurine also decreased PGD2 release during reoxygenation (Fig. 5). Carbon uptake at the end of the reoxygenation period was 507 { 52 mg/g liver/ min in the presence of betaine and taurine (100 mmol/L each) compared with 748 { 77 mg/g liver/min in hypoxiareoxygenation experiments without osmolytes (n Å 4 for each condition). In contrast to hyperosmolarity, in normoosmolar perfusions, PGE2 , TXB2 , and TNF-a release were
FIG. 2. Effect of betaine and taurine on hypoxia-reoxygenation–induced LDH release in hyperosmolar liver perfusion. Data represent means { SEM (six independent control experiments with 385 mosmol/L, four with each of the other conditions). The curves with 100 mmol/L betaine / 100 mmol/L taurine and 1 mmol/L betaine were significantly different from control without osmolyte (P õ .01, repeated-measures ANOVA). (●) 385 mosmol/L; (j) 385 mosmol/L / 100 mmol/L betaine; (m) 385 mosmol/L / 100 mmol/L betaine / 100 mmol/L taurine; and (l) 385 mosmol/L / 1 mmol/L betaine.
hepa
WBS: Hepatology
WETTSTEIN AND HA¨USSINGER
HEPATOLOGY Vol. 26, No. 6, 1997
FIG. 3. Effect of betaine and taurine on hypoxia-reoxygenation–induced LDH release in normo-osmolar liver perfusions. The hypoxic period was 90 minutes. Data represent means { SEM (six independent control experiments without osmolytes, four independent experiments with betaine and taurine, each). The curve for taurine was significantly different from control (P õ .05, repeated-measures ANOVA). (●) 305 mosmol/L; (l) 305 mosmol/L / 100 mmol/L betaine; and (j) 305 mosmol/L / 100 mmol/L taurine.
below the detection limits of the respective tests (not shown). This does not necessarily mean that these compounds are not involved in the pathogenesis of reoxygenation injury in the experimental model used, as the perivenous scavenger cells very effectively eliminate mediators released in the liver sinusoids before reaching the systemic circulation.37 In summary, the decrease of hypoxia-reoxygenation–induced eicosanoid production, cytokine release, and carbon uptake point to an inhibition of Kupffer cell activation during reoxygenation by betaine and taurine, thereby attenuating liver injury.
1563
FIG. 4. Effect of ibuprofen on hypoxia-reoxygenation–induced LDH release in normo-osmolar liver perfusion. The hypoxic period was 90 minutes. Data represent means { SEM (six independent control experiments without ibuprofen, four independent experiments with ibuprofen). The curves were significantly different (P õ .01, repeated-measures ANOVA). (●) 305 mosmol/L; and (m) 305 mosmol/L / 50 mmol/L ibuprofen.
Influence of Anoxia-Reoxygenation and Osmolytes on Endothelial Cell Function. HA uptake and its degradation to lactate and
acetate is regarded as a sensitive and specific parameter of endothelial cell function.38-41 In the perfused rat liver, HA
TABLE 2. Effect of Betaine and Ibuprofen on Increase of Hepatic Vascular Resistance in Hypoxia-Reoxygenation Injury in Normo-osmolar Perfused Rat Liver Basal Vascular Resistance (cm H2O/mL/min)
305 mosmol/L, 90-min hypoxia (control) 305 mosmol/L, 90-min hypoxia / betaine 100 mmol/L 305 mosmol/L, 90-min hypoxia / ibuprofen 50 mmol/L
Increase of Vascular Resistance (cm H2O/mL/min)
0.19 { 0.02 (8)
0.21 { 0.06
0.22 { 0.02 (5)
0.09 { 0.03*
0.15 { 0.03 (5)
0.11 { 0.02*
NOTE. Livers were perfused for approximately 45 minutes with buffer saturated with O2/CO2 (95/5 vol/vol), then for 90 minutes with N2/CO2saturated buffer (hypoxia), followed by a 90-minute period of reoxygenation. Basal vascular resistance (VR) represents perfusion pressure divided by perfusate flow before hypoxia. Increase in VR represents VR at the end of the reoxygenation period minus basal VR. Medium osmolarity was 305 mosmol/L. Data represent means { SEM (number of experiments). * Significantly different from control experiments (t test, P õ .05).
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
FIG. 5. PGD2 release in normo-osmolar liver perfusion. Means { SEM of four experiments each. In the presence of taurine, the values after reoxygenation were significantly lower than controls (P õ .05, repeated-measures ANOVA). (●) 305 mosmol/L; and (m) 305 mosmol/L / 100 mmol/L taurine.
hepa
WBS: Hepatology
1564 WETTSTEIN AND HA¨USSINGER
HEPATOLOGY December 1997
FIG. 6. HA uptake of the perfused rat liver. Livers were perfused for approximately 50 minutes with buffer saturated with O2 /CO2 (95/5 vol/vol), then for 90 minutes with N2 /CO2 -saturated buffer (hypoxia), followed by a 90-minute period of reoxygenation. During the last 20 minutes of each period, HA (200 ng/mL) was infused and HA uptake was determined as described in Materials and Methods. Medium osmolarity was 305 mosmol/ L. Data represent means { SEM of four independent experiments for each condition. *Significantly different from uptake before hypoxia (P õ .01, t test). (j) control; and ( ) 100 mmol/L betaine / 100 mmol/L taurine.
uptake is saturated with concentrations above 150 ng/mL in the influent perfusate.38 To determine the importance of endothelial cell injury and the protective effect of betaine and taurine, HA (200 ng/mL) was infused for 20-minute periods before hypoxia, during the last 20 minutes of the 90minute hypoxic period, and during the last 20 minutes of the reoxygenation period. HA uptake was calculated from the difference between influent and effluent concentrations. Basal HA uptake was 277 { 51 ng/g liver/min (mean { SEM, n Å 4) in control livers without osmolytes and 302 { 30 ng/ g liver/min in the presence of betaine and taurine (100 mmol/ L each) (Fig. 6). At the end of the hypoxic perfusion, HA uptake decreased to 174 { 57 and 164 { 23 ng/g liver/min, respectively, and, after 90 minutes of reoxygenation, it was 193 { 29 and 210 { 76 ng/g liver/min, respectively. There were no significant differences between controls and experiments in the presence of betaine and taurine, suggesting that the protective effect of these osmolytes is probably not caused by attenuation of endothelial cell injury. DISCUSSION
Cell volume is an important determinant of liver cell function: cell swelling is an anabolic and proliferative signal, whereas cell shrinkage is catabolic and antiproliferative.32,33 The data presented here show an osmolarity dependence of cell injury in a rat liver model of ischemia-reoxygenation: LDH leakage was approximately 10 times higher with hyperosmolar perfusion buffer than with normo-osmolar medium and further decreased with hypo-osmolar buffer. This is in line with a previous report that t-butylhydroperoxide–induced cell injury is also aggravated by ambient hyperosmolarity.36 Several factors may contribute to osmolarity dependence of cell injury. TNF-a production, cyclooxygenase-2 expression, and PG synthesis are increased by hyperosmolar-
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
ity and decreased in hypo-osmolarity.17,18 Hypo-osmolarity stimulates bile secretion and is accompanied by the activation of mitogen-activated protein kinases in liver, which is also seen as a protective mechanism.42 In addition, the canalicular elimination of mediators of inflammation such as cysteinyl leukotrienes is decreased by hyperosmolarity and increased in hypo-osmolarity.43 The osmolarity dependence of cell injury may be of importance in view of the finding that severe illness and catabolic states are accompanied by a decrease of cellular water content.33 Cell shrinkage may therefore not only promote catabolism, but also lead to an increased susceptibility toward ischemic injury. The beneficial effect of cell volume increase induced by hypo-osmolar solutions in warm ischemia-reoxygenation or t-butylhydroperoxide–induced cell injury and the aggravation of liver injury by hyperosmolarity must be separated from cell swelling caused by loss of cell function. The situation in warm ischemia may also be different from cold ischemia, because solutions containing impermeant agents such as lactobionate, raffinose, and hydroxyethyl starch (e.g., UW solution) are used to prevent hypoxic swelling during cold organ storage.44 Several compounds have been shown to attenuate liver injury in experimental models of warm ischemia and reoxygenation, including L-arginine, calcium channel blockers, anti–intercellular adhesion molecule-1 antibodies, and PGE1 .45-48 In accordance with the potential importance of Kupffer cell activation for reoxygenation injury, inactivation of Kupffer cells by methyl palmitate or gadolinium chloride pretreatment was also protective.5 The data presented here show a protective effect of betaine and taurine in warm ischemia-reoxygenation injury. The presence of betaine or taurine in the perfusion medium in near-physiological concentrations decreases LDH and AST leakage and attenuates microcirculatory disturbances. Several mechanisms may be involved in this protective effect. In hyperosmolar perfusions, betaine and taurine may be accumulated intracellularly, thereby compensating for the hyperosmolarity-induced cell shrinkage. The physiological concentration of betaine in blood is 20 to 60 mmol/L, and it can be accumulated in Kupffer cells to concentrations of 60 to 70 mmol/L.21 However, in hepatic parenchymal cells, betaine is of minor importance as an osmolyte, because betaine uptake rates are low and betaine transporter BGT1 is not expressed.22 In addition, hyperosmolarity-induced volumeregulatory K/-release was not different in the presence and absence of betaine and taurine (100 mmol/L each). This makes an interference of these osmolytes with cell volume regulation in parenchymal cells unlikely, which make up the largest part of the liver cell mass, but does not exclude such a mechanism in a smaller compartment as Kupffer cells. For taurine, a role as an antioxidant, in conjugation reactions and as a membrane stabilisator, has been described.49 The antioxidative potential of taurine is based on the formation of chloramines.50 Taurine protects isolated hepatocytes against toxic compounds including carbon tetrachloride, although the mechanisms of this effect remain unclear.51 Osmolytes were shown to stabilize protein structures by prevention of denaturation.52 However, the protective effect in the experimental model used here seems to be caused by the modulation of Kupffer cell function. Betaine and taurine were recently identified as osmolytes in a macrophage cell line and rat Kupffer cells.20-22 Kupffer cell functions such as cyclooxygenase-2 expression,
hepa
WBS: Hepatology
WETTSTEIN AND HA¨USSINGER
HEPATOLOGY Vol. 26, No. 6, 1997
phagocytosis, and PG and TNF-a production are modulated by ambient osmolarity and osmolytes.17-19 In the study presented here, hypoxia-reoxygenation increased the release of PGs, thromboxane, and TNF-a and stimulated carbon uptake, indicating activation of Kupffer cells. These effects were more pronounced in hyperosmolar perfusions than under normo-osmolar conditions. The Kupffer cell activation was suppressed in the presence of betaine and taurine. Although the importance of the various cytokines, eicosanoids, and reactive oxygen species released during reoxygenation is not clear and may vary between different experimental models, activation of arachidonic acid metabolism by lipid peroxidation is seen as an important mechanism.53 Hydroperoxide was shown to induce the release of eicosanoids from the perfused rat liver.54 Some of the eicosanoids released by nonparenchymal cells such as PGE2 and prostacyclin are potentially protective, whereas others such as thromboxane, PGD2 and 6-keto-PGF2a are considered to promote liver injury by local vasoconstriction.46,55 In our study, hypoxiareoxygenation also stimulated the formation of the potentially protective PGE2 , but the other mediators released seem to override this effect. In addition, recent experiments in isolated hepatocytes showed an increased susceptibility toward oxidative stress in the presence of PGE2 (vom Dahl S, Kubitz R, Ha¨ussinger D, unpublished data, December 1996). The importance of eicosanoid formation for the reoxygenation injury is substantiated by the protective effect of the cyclo-oxygenase inhibitor, ibuprofen, in line with previous studies showing attenuation of ischemic injury by indomethacin or prostacyclin infusion.56,57 HA uptake was decreased by approximately 35% to 40% during anoxia and reoxygenation. HA is a high-molecularweight polysaccharide that is removed from circulation specifically by liver endothelial cells via a receptor-mediated endocytotic process.40,41 However, there was no significant difference with and without osmolytes and no correlation to LDH leakage. Therefore, in this model of ischemia-reoxygenation injury, endothelial cell damage appears to be less important. This is in line with previous findings that sinusoidal cells are more susceptible to cold ischemia than hepatic parenchymal cells, whereas warm ischemia induces parenchymal cell injury.7,39,58 The protective effect of betaine and taurine in ischemiareoxygenation may gain clinical importance in organ preservation for transplantation. A hepatoprotective role of betaine was also shown in a recent study demonstrating that supplemental dietary betaine prevented ethanol-induced fatty infiltration and increased S-adenosylmethionine levels in rat livers.59 The potential of osmolytes to modify Kupffer cell function provides new perspectives in many pathophysiological conditions involving nonparenchymal cells, because other known modulators of Kupffer cell function are toxic, which excludes their application in humans.
4. 5. 6.
7. 8. 9. 10.
11.
12. 13.
14.
15.
16. 17.
18. 19. 20. 21. 22. 23.
Acknowledgment: The expert technical assistance of Claudia Holneicher is gratefully acknowledged.
24.
REFERENCES
25.
1. Jaeschke H. Preservation injury: mechanism, prevention and consequences. J Hepatol 1996;25:774-780. 2. Bronk SF, Gores GJ. pH-dependent nonlysosomal proteolysis contributes to lethal anoxic injury of rat hepatocytes. Am J Physiol 1993;264: G744-G751. 3. Calmus Y, Cynoper L, Dousset B, Lim SK, Soubrane O, Conti F, Houssin
26.
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
27.
1565
D, et al. Evidence for the detrimental role of proteolysis during liver preservation in humans. Gastroenterology 1995;108:1510-1516. Jaeschke H, Bautista AP, Spolarics Z, Spitzer JJ. Superoxide generation by Kupffer cells and priming of neutrophils during reperfusion after hepatic ischemia. Free Radic Res Commun 1991;15:277-284. Jaeschke H, Farhood A. Neutrophil and Kupffer cell–induced oxidant stress and ischemia-reperfusion injury in rat liver in vivo. Am J Physiol 1991;260:G355-G362. Lindert KA, Caldwell-Kenkel JC, Nukina S, Lemasters JJ, Thurman RG. Activation of Kupffer cells on reperfusion following hypoxia: particle phagocytosis in a low-flow, reflow model. Am J Physiol 1992;262:G345G350. Caldwell-Kenkel JC, Currin RT, Tanaka Y, Thurman RG, Lemasters JJ. Kupffer cell activation and endothelial cell damage after storage of rat livers: effects of reperfusion. HEPATOLOGY 1991;13:83-95. Rymsa B, Wang JF, de Groot H. O2 -release by activated Kupffer cells upon hypoxia-reoxygenation. Am J Physiol 1991;261:G602-G607. Zhou W, McCollum MO, Levine B, Olson MS. Inflammation and platelet-activating factor production during hepatic ischemia-reperfusion. HEPATOLOGY 1992;16:1236-1240. Colletti LM, Remick DG, Burtch GD, Kunkel SL, Strieter RM, Campbell DA. Role of tumor necrosis factor-a in the pathophysiologic alterations after hepatic ischemia-reperfusion injury in the rat. J Clin Invest 1990; 85:1936-1943. Scoazec JY, Durand F, Degott C, Delautier D, Bernuau J, Belghiti J, Benhamou JP, et al. Expression of cytokine-dependent adhesion molecules in postreperfusion biopsy specimens of liver allograft. Gastroenterology 1994;107:1094-1102. Vollmar B, Glasz J, Leiderer R, Post S, Menger MD. Hepatic microcirculatory failure is a determinant of liver dysfunction in warm ischemiareperfusion. Am J Pathol 1994;145:1421-1431. Hanazaki K, Kuroda T, Haba Y, Shiohara E, Kajikawa S, Iida F. Protective effect of the thromboxane A2 receptor antagonist ONO 3708 on ischemia-reperfusion injury in the dog liver. Surgery 1994;116:896903. Bradham CA, Stachlewitz RF, Gao W, Qian T, Jayadev S, Jenkins G, Yusuf H, et al. Reperfusion after liver transplantation in rats differentially activates the mitogen-activated protein kinases. HEPATOLOGY 1997;25:1128-1135. Marzi I, Cowper K, Takei Y, Lindert K, Lemasters JJ, Thurman RG. Methyl palmitate prevents Kupffer cell activation and improves survival after orthotopic liver transplantation in the rat. Transplant Int 1991;4: 215-220. Imamura H, Sutto F, Brault A, Huet P-M. Role of Kupffer cells in cold ischemia-reperfusion injury of rat liver. Gastroenterology 1995;109: 189-197. Zhang F, Warskulat U, Wettstein M, Schreiber R, Henninger HP, Decker K, Ha¨ussinger D. Hyperosmolarity stimulates prostaglandin synthesis and cyclooxygenase-2 expression in activated rat liver macrophages. Biochem J 1995;312:135-143. Zhang F, Warskulat U, Ha¨ussinger D. Modulation of tumor necrosis factor-a release by anisoosmolarity and betaine in rat liver macrophages (Kupffer cells). FEBS Lett 1996;391:293-296. Warskulat U, Zhang F, Ha¨ussinger D. Modulation of phagocytosis by anisoomolarity and betaine in rat liver macrophages (Kupffer cells) and RAW 264.7 mouse macrophages. FEBS Lett 1996;391:287-292. Warskulat U, Wettstein M, Ha¨ussinger D. Betaine is an osmolyte in RAW 264.7 mouse macrophages. FEBS Lett 1995;377:47-50. Zhang F, Warskulat U, Wettstein M, Ha¨ussinger D. Identification of betaine as an osmolyte in rat liver macrophages (Kupffer cells). Gastroenterology 1996;110:1543-1552. Warskulat U, Wettstein M, Ha¨ussinger D. Osmoregulated taurine transport in H4IIE hepatoma cells and perfused rat liver. Biochem J 1997; 321:683-690. Yancey PH, Clark ME, Hand SC, Bowlus RD, Somero GN. Living with water stress: evolution of osmolyte systems. Science 1982;217:12141222. Handler JS, Kwon HM. Regulation of renal cell organic osmolyte transport by tonicity. Am J Physiol 1993;265:C1449-C1455. Burg MB. Molecular basis of osmotic regulation. Am J Physiol 1995; 268:F983-F996. Nakanishi, T, Turner RJ, Burg MB. Osmoregulation of betaine transport in mammalian renal medullary cells. Am J Physiol 1990;258:F1061F1067. Zhou C, Agarwal N, Cammata PR. Osmoregulatory alterations in myoinositol uptake by bovine lens epithelial cells. Invest Ophthalmol Vis Sci 1994;35:1236-1242.
hepa
WBS: Hepatology
1566 WETTSTEIN AND HA¨USSINGER
HEPATOLOGY December 1997
28. Paredes A, McManus M, Kwon HM, Strange, K. Osmoregulation of the Na/-inositol cotransporter activity and mRNA levels in brain glial cells. Am J Physiol 1993;263:C1282-C1288. 29. Sies H. The use of perfusion of liver and other organs for the study of microsomal electron transport and cytochrom P-450 systems. Methods Enzymol 1978;52:48-59. 30. Bergmeyer HU, ed. Methods of Enzymatic Analysis. 3rd ed. Weinheim: VCH Verlagsgesellschaft, 1983. 31. Cowper KB, Currin RT, Dawson TL, Lindert KA, Lemasters JJ, Thurman RG. A new method to monitor Kupffer-cell function continuously in the perfused rat liver. Biochem J 1990;266:141-147. 32. Ha¨ussinger D, Gerok W, Lang F. Regulation of cell function by the cellular hydration state. Am J Physiol 1994;267:E343-E355. 33. Ha¨ussinger D, Roth E, Lang F, Gerok W. Cellular hydration state: an important determinant of protein catabolism in health and disease. Lancet 1993;341:1330-1332. 34. Vom Dahl S, Hallbrucker C, Lang F, Gerok W, Ha¨ussinger D. Regulation of liver cell volume and proteolysis by glucagon and insulin. Biochem J 1991;278:771-777. 35. Vom Dahl S, Ha¨ussinger D. Nutritional state and the swelling-induced inhibition of proteolysis in perfused rat liver. J Nutr 1996;126:395-402. 36. Saha N, Stoll B, Lang F, Ha¨ussinger D. Effect of anisotonic cell-volume modulation on glutathione-S-conjugate release, t-butylhydroperoxide metabolism and the pentose-phosphate shunt in perfused rat liver. Eur J Biochem 1992;209:437-444. 37. Ha¨ussinger D, Stehle T. Hepatocyte heterogeneity in response to icosanoids: the perivenous scavenger cell hypothesis. Eur J Biochem 1988; 175:395-403. 38. Deaciuc IV, Bagby GJ, Lang CH, Spitzer JJ. Hyaluronic acid uptake by the isolated, perfused rat liver: an index of hepatic sinusoidal endothelial cell function. HEPATOLOGY 1993;17:266-272. 39. Sutto F, Brault A, Lepage R, Huet P-M. Metabolism of hyaluronic acid by liver endothelial cells: effect of ischemia-reperfusion in the isolated perfused rat liver. J Hepatol 1994;20:611-616. 40. Smedsrod B, Pertoft H, Eriksson S, Fraser RE, Laurent TC. Studies in vitro on the uptake and degradation of sodium hyaluronate in rat liver endothelial cells. Biochem J 1984;223:617-626. 41. McGary CT, Raja RH, Weigel PH. Endocytosis of hyaluronic acid by rat liver endothelial cells. Biochem J 1989;257:875-884. 42. Noe´ B, Schliess F, Wettstein M, Heinrich S, Ha¨ussinger D. Regulation of taurocholate excretion by a hypo-osmolarity–activated signal transduction pathway in rat liver. Gastroenterology 1996;110:858-865. 43. Wettstein M, Noe´ B, Ha¨ussinger D. Metabolism of cysteinyl leukotrienes in the perfused rat liver: the influence of endotoxin pretreatment and the cellular hydration state. HEPATOLOGY 1995;22:235-240. 44. Southard JH, Belzer FO. Organ preservation. Annu Rev Med 1995;46: 235-247.
AID
Hepa 0018
/
5p2a$$$341
11-12-97 09:34:06
45. Jones SM, Thurman RG. L-Arginine minimizes reperfusion injury in a low-flow, reflow model of liver perfusion. HEPATOLOGY 1996;24:163168. 46. Grafa F, Gugenheim J, Saint-Paul M-C, Cavanel C, Lapalus F, Ouzan D, Militerno G, et al. Protective effect of prostaglandin E1 on normothermic liver ischemia. Eur Surg Res 1991;23:278-284. 47. Nakano H, Kuzume M, Namatame K, Yamaguchi M, Kumada K. Efficacy of intraportal injection of anti ICAM-1 monoclonal antibody against liver cell injury following warm ischemia in the rat. Am J Surg 1995; 170:64-66. 48. Nauta RJ, Tsimoyiannis E, Uribe M, Walsh DB, Miller D, Butterfield A. The role of calcium ions and calcium channel blockers in experimental ischemia-reperfusion–induced liver injury. Ann Surg 1991;213:137142. 49. Huxtable RJ. Physiological actions of taurine. Physiol Rev 1992;72:101163. 50. Weiss SJ, Klein R, Slivka A, Wie M. Chlorination of taurine by human neutrophils—evidence for hypochlorous acid generation. J Clin Invest 1982;70:598-607. 51. Timbrell JA, Seabra V, Waterfield CJ. The in vivo and in vitro protective properties of taurine. Gen Pharmacol 1995;26:453-462. 52. Liu Y, Bolen DW. The peptide backbone plays a dominant role in protein stabilization by naturally occurring osmolytes. Biochemistry 1995;34: 12884-12891. 53. Mathews WR, Guido DM, Fisher MA, Jaeschke H. Lipid peroxidation as molecular mechanism of liver cell injury during reperfusion after ischemia. Free Radic Biol Med 1994;16:763-770. 54. Hallbrucker C, Ritter M, Lang F, Gerok W, Ha¨ussinger D. Hydroperoxide metabolism in rat liver: K/ channel activation, cell volume changes and eicosanoid formation. Eur J Biochem 1993;211:449-458. 55. Post S, Goerig M, Otto G, Manner M, Senninger N, Kommerell B, Herfarth C. Prostanoid release in experimental liver transplantation. Transplantation 1990;49:490-494. 56. Briegel J, Haller M, Zulke C. Primary graft nonfunction following orthotopic liver transplantation: treatment with prostacyclin. Transplant Proc 1992;24:2693-2670. 57. Pittet JF, Morel DR, Mentha G, LeCoultre C, Suter PM, Rohner A. Protective effect of indomethacin in the development of the postreperfusion syndrome during liver transplantation in pigs. Transplant Proc 1991;23:2290-2296. 58. McKeown CMB, Edwards V, Phillips MJ, Harvey PRC, Petrunka CN, Strasberg SM. Sinusoidal lining damage: the critical injury in cold preservation of allografts in the rat. Transplantation 1988;46:178-191. 59. Barak AJ, Beckenhauer HC, Tuma DJ. S-Adenosylmethionine generation and prevention of alcoholic fatty liver by betaine. Alcohol 1994;11:501503.
hepa
WBS: Hepatology