Structural and functional integrity of rat liver perfused in backward and forward directions

Cholestasis, metabolism and biliary lipid secretion during perfusion of rat liver with different bile salts

The biological effects of bile acids depend largely upon their molecular structure. When bile acid uptake exceeds the maximal biliary secretory rate (SRm) cholestasis occurs. In order to characterize the influence of bile acid structure on its cholestatic potency we systematically studied SRm, maximal bile flow, maximal and cumulative phospholipid and cholesterol secretion with different taurine-conjugated tri-, di- and keto bile acids (Table I) in the isolated perfused rat liver. Bile acids with a high critical micellar concentration (CMC) promoted the greatest bile flow; a positive non-linear correlation between CMC and maximal bile flow was found. 3 alpha-Hydroxylated bile acids with a hydroxyl group in 6 alpha and/or 7 beta position and lacking a 12 alpha hydroxy group had a high SRm. SRm was not related to CMC or maximal bile flow, respectively. Phospholipids and cholesterol were secreted in a nearly fixed ratio of 12:1; a strong linear relationship could be observed. Cumulative phospholipid secretion over 48 min was significantly lower for non and poor micelle forming bile acids (TDHC and TUC) than for those with comparatively low CMC values (TUDC, TC, THC, THDC, TCDC) (70-140 vs. 210-450 nmol/g liver). At SRm all bile acids with good micelle forming properties showed a similar cumulative biliary lipid output. However, when biliary lipid output was related to 1 mumol bile acid secreted bile acids with a low SRm induced the highest lipid secretion (TCDC, TC). These data (1) demonstrate that a 6 alpha and/or a 7 beta hydroxy group on the steroid nucleus reduce cholestatic potency if the 12 alpha hydroxy group is absent, (2) suggest that in the case of micelle forming bile acids the total amount of phospholipids secreted in bile (depletion of cellular phospholipids) is associated with the occurrence of cholestasis whereby bile acids with a low SRm deplete the cellular phospholipid content at much lower bile acid concentrations than those with a higher SRm and (3) imply that bile acids with non and poor micelle forming properties (TDHC, TUC) presumably do not cause cholestasis (solely) by depletion of cellular phospholipids.

Use of an artificial oxygen carrier in isolated rat liver perfusion: first demonstration of net glucose uptake at physiological portal glucose concentrations using a hemoglobin-free perfusate

A defect in isolated perfused rat-liver (IPRL) preparations has been proposed to explain discrepancies between in vivo and in vitro findings regarding hepatic glucose metabolism. The aim of the present study was to investigate whether a preparation of IPRL using a synthetic hemoglobin-free perfusate was capable of net glucose uptake and glycogen deposition at physiological portal substrate concentrations. Livers from fed anaesthetized rats were perfused in a recirculating system using a fluorocarbon emulsion as artificial oxygen carrier. Depending on the prevailing glucose concentration, livers exhibited net glucose uptake or release with a threshold value of 5.5-6.0 mM glucose. Net glucose uptake was associated with net glycogen deposition (+0.23 to +0.59 mumol C6 min-1 g-1). From 5.8 mM (n = 3) and 10.0 mM (n = 8), initial concentration glucose levels fell to 5.3 +/- 0.2 mM after 210 min (n = 3) and 6.3 +/- 0.9 mM after 120 min (n = 8), respectively. This was equivalent to a net glucose uptake of -0.16 and -0.45 mumol min-1 g-1. Anoxia reversibly switched hepatic glucose balance from net uptake (-0.42 mumol min-1 g-1) to release (+0.69 mumol min-1 g-1) followed by net uptake (-0.50 mumol min-1 g-1) after reinstitution of aerobic conditions. We conclude that the composition of perfusion media might play a pivotal role for studies of glucose metabolism in the isolated perfused rat liver. In our experimental model, using a hemoglobin-free synthetic medium, net glucose uptake was readily demonstrated at physiological portal substrate concentrations similar to the in vivo situation.

Taurohyocholate, taurocholate, and tauroursodeoxycholate but not tauroursocholate and taurodehydrocholate counteract effects of taurolithocholate in rat liver

The infusion of taurolithocholate (TLC) in vivo or in the isolated perfused liver of the rat causes cholestasis and cellular necrosis. In order to analyze the protective effect of bile salts differing in number and steric position of their hydroxy groups against TLC-induced cholestasis, isolated rat livers were perfused with taurocholate (TC), taurohyocholate (THC), tauroursocholate (TUC), taurodehydrocholate (TDHC), and tauroursodeoxycholate (TUDC) (16 and 32 mumol/l) with or without TLC (8 and 16 mumol/l). Bile flow, bile salt secretion, and the hydroxylation pattern of the bile salts secreted were analyzed. TLC caused complete cholestasis after 15 min of perfusion. All bile salts studied had a protective effect. THC, TC, and TUDC completely abolished the cholestasis induced by TLC while TUC did so only for the first 10 min. TDHC was protective only as long as it was biotransformed into hydroxyoxo bile salts. Coinfusion of bile salts did not influence uptake of TLC (greater than or equal to 93% of dose). Differences were found regarding the amount of TLC biotransformed (% of uptake): TC 50%; THC 32%; TUDC 36%; TUC 20%. Light microscopy revealed cellular necrosis, and dilated canaliculi were found in livers perfused with TLC only or in combination with TUC or TDHC, while the other bile salts prevented these changes. We conclude that bile salts with low micelle-forming capacity have little protective effect against TLC-induced cholestasis. These bile salts induce less biotransformation of TLC than TC, THC, and TUDC. The protective effect is not dependent on the hydrocholeretic effect of the added bile salt and is not due to an uptake inhibition.

Tauroursodeoxycholate prevents taurolithocholate-induced cholestasis and toxicity in rat liver

Ursodeoxycholate has been advocated for the treatment of cholestatic liver diseases. The coinfusion of tauroursodeoxycholate with taurolithocholate in the perfused rat liver completely prevented the decrease of bile flow and the increase of oxygen uptake found with taurolithocholate only. Bile flow and bile salt secretion were increased with the coinfusion of both bile acids as compared with the infusion of tauroursodeoxycholate only (+4.30 microliters/g liver per 30 min) with 16 and 32 mumol/l tauroursodeoxycholate (+1.55 microliters/g liver per 30 min with 80 and 160 mumol/l). Morphological examination revealed a 50% decrease of the number of necrotic cells in the periportal area. Tauroursodeoxycholate did not inhibit the uptake of taurolithocholate, but increased its transcellular passage and biotransformation. Thus, tauroursodeoxycholate prevents taurolithocholate-induced cholestasis and liver cell toxicity probably by an intracellular mechanism.

The multiple-indicator dilution technique for characterization of normal and retrograde flow in once-through rat liver perfusions

The technique of normal and retrograde rat liver perfusion has been widely used to probe zonal differences in drug-metabolizing activities. The validity of this approach mandates the same tissue spaces being accessed by substrates during both normal and retrograde perfusions. Using the multiple-indicator dilution technique, we presently examine the extent to which retrograde perfusion alters the spaces accessible to noneliminated references. A bolus dose of 51Cr-labeled red blood cells, 125I-albumin, 14C-sucrose and 3H2O was injected into the portal (normal) or hepatic (retrograde) vein of rat livers perfused at 10 ml per min per liver. The outflow perfusate was serially collected over 220 sec to characterize the transit times and the distribution spaces of the labels. During retrograde perfusion, red blood cells, albumin and sucrose profiles peaked later and lower than during normal perfusion, whereas the water curves were similar. The transit times of red blood cells, albumin and sucrose were longer (p less than 0.005), whereas those for water did not change. Consequently, retrograde flow resulted in significantly larger sinusoidal blood volumes (45%), albumin Disse space (42%) and sucrose Disse space (25%) than during normal flow, whereas the distribution spaces for total and intracellular water remained unaltered. The distension of the vascular tree was confirmed by electron microscopy, by which occasional isolated foci of widened intercellular recesses and spaces of Disse were observed. Cellular ultrastructure was otherwise unchanged, and there was no difference found between normal and retrograde perfusion for bile flow rates, AST release, perfusion pressure, oxygen consumption and metabolic removal of ethanol, a substrate with flow-limited distribution, which equilibrates rapidly with cell water (hepatic extraction ratios were virtually identical: normal vs. retrograde, 0.50 vs. 0.48 at 6 to 7.4 mM input concentration). These findings suggest that the functional and metabolic capacities of the liver remain unperturbed during retrograde perfusion, rendering the technique suitable for the investigation of zonal differences in drug-metabolizing enzymes.

Competing pathways in drug metabolism. II. An identical, anterior enzymic distribution for 2- and 5-sulfoconjugation and a posterior localization for 5-glucuronidation of gentisamide in the rat liver

Gentisamide (2,5-dihydroxybenzamide, GAM), a substrate that forms two monosulfates at the 2 and 5 positions (GAM-2S and GAM-5S) and a monoglucuronide at the 5 position (GAM-5G), was delivered at 8 or 80 microM by normal (N) and retrograde (R) flows to the once-through rat liver preparation. At the lower (8 microM) input concentration, ratios of conjugate formation rate, GAM-5S/GAM-5G and GAM-2S/GAM-5G, were decreased significantly (4.01 +/- 1.42 to 2.93 +/- 0.99, and 1.13 +/- 0.65 to 0.66 +/- 0.41, respectively) whereas a small but significant increase in the steady-state extraction ratio, E (0.89 +/- 0.029 to 0.94 +/- 0.016), was observed upon changing the flow direction from N to R. At the higher input GAM concentration (80 microM), conjugate formation rate ratios were relatively constant for GAM-5S/GAM-5G (1.18 +/- 0.08 to 1.11 +/- 0.12) and GAM-2S/GAM-5G (0.33 +/- 0.05 to 0.31 +/- 0.05) upon changing flow direction from N to R, despite a slight increase in E from 0.87 +/- 0.023 to 0.91 +/- 0.016 was observed. These results suggest that the sulfotransferase activities responsible for 2- and 5-sulfoconjugations are identically distributed and localized anterior to 5-glucuronidation activities during a normal flow of substrate into the rat liver (entering the portal vein and exiting the hepatic vein), and that the presence of uneven distribution of conjugation activities is discriminated only at the lower input drug concentration. At high concentration (greater than Km for all systems), saturation of all pathways occurs, and other anteriorly/identically distributed competing pathways would fail to perturb downstream intrahepatic drug concentrations and the resultant conjugation rates. The lack of change in metabolic profiles renders the condition unsuitable for examination of uneven distribution of enzymes in the liver. These observations were generally predicted by theoretical enzymic models of consistent distribution patterns. Because 2- and 5-sulfation were mediated by systems of similar Km but different Vmax values, two possibilities, the same isozyme of sulfotransferase being involved in the formation of two enzyme-substrate complexes to form two distinctly different products or two isozymes of sulfotransferases of identical distribution, were discussed.