Digitonin-collagenase perfusion for efficient separation of periportal or perivenous hepatocytes.

Intact rat liver cells from the perivenous region were isolated by collagenase perfusion after first destroying the periportal region by a brief portal infusion of digitonin. Periportal cells were isolated after retrograde digitonin infusion. Significantly higher alanine aminotransferase, gamma-glutamyltransferase and lactate dehydrogenase activities and lower glutamate dehydrogenase and pyruvate kinase activities in periportal than in perivenous cells demonstrate marked separation. The high yield allows further characterization in vitro of the cell populations.     

Allyl alcohol cytotoxicity and glutathione depletion in isolated periportal and perivenous rat hepatocytes

The mechanism of the periportal (p.p.) toxicity of allyl alcohol (AlOH) was investigated in p.p. and perivenous (p.v.) hepatocytes isolated by digitonin-collagenase perfusion. The distinct origin of the cell preparations was confirmed by the p.p./p.v. ratios of alanine aminotransferase (p.p./p.v. = 1.8), lactate dehydrogenase (1.3) and glutamine synthetase (0.10). The activity of alcohol dehydrogenase (ADH) was not markedly different in p.p. and p.v. cells. Both types of cells oxidized AlOH at a high but equal rate of about 3 mumol/(min.g cells). Concomitantly with rapid oxidation of 0.7 mM AlOH, glutathione (GSH) was depleted by about 95% and its secretion was completely inhibited in both cell types. Although the GSH content was partially restored during a subsequent 3-h incubation, cellular ATP and K+ content gradually decreased and the leakage of lactate dehydrogenase increased in both types of cells. However, the p.p. cells tended to resist AlOH in vitro better, probably due to their 26% higher GSH content after preincubation with L-methionine. Altering the partial pressure of oxygen in physiological range had no effect on the toxicity of AlOH. The results are contrary to the suggestions that the p.p. location of AlOH liver injury is caused by higher ADH activity or higher oxygen tension in the p.p. zone. Rather, the regiospecificity of the injury may be due to rapid uptake and oxidation of AlOH in the p.p. region.     

A high-sucrose diet increases gluconeogenic capacity in isolated periportal and perivenous rat hepatocytes

A high-sucrose (SU) diet increases gluconeogenesis (GNG) in the liver. The present study was conducted to determine the contribution of periportal (PP) and perivenous (PV) cell populations to this SU-induced increase in GNG. Male Sprague-Dawley rats were fed an SU (68% sucrose) or starch (ST, 68% starch) diet for 1 wk, and hepatocytes were isolated from the PP or PV region of the liver acinus. Hepatocytes were incubated for 1 h in the presence of various gluconeogenic substrates, and glucose release into the medium was used to estimate GNG. When incubated in the presence of 5 mM lactate, which enters GNG at the level of pyruvate, glucose release (nmol x h(-1) x mg(-1)) was significantly increased by the SU diet in both PP (84.8 +/- 3.4 vs. 70.4 +/- 2.6) and PV (64.3 +/- 2.5 vs. 38.2 +/- 2.1) cells. Addition of palmitate (0.5 mM) increased glucose release from lactate in PP cells by 11.6 +/- 0.5 and 20.6 +/- 1.5% and in PV cells by 11.0 +/- 4.4 and 51.1 +/- 9.1% in SU and ST, respectively. When cells were incubated with 5 mM dihydroxyacetone (DHA), which enters GNG at the triosephosphate level, glucose release was significantly increased by the SU diet in both cell types. In contrast, glucose release from fructose (0.5 mM) was significantly increased by the SU diet in PV cells only. These changes in glucose release were accompanied by significant increases in the maximal specific activities of glucose-6-phosphatase (G-6-Pase) and phosphoenolpyruvate carboxykinase (PEPCK) in both PP and PV cells. These data suggest that the SU diet influences GNG in both PP and PV cell populations. It appears that SU feeding produces changes in GNG via alterations in at least two critical enzymes, G-6-Pase and PEPCK.     

Differential localization of microsomal mixed function oxidase activity in periportal and perivenous hepatocytes isolated from rat liver

1. The activity per mg of microsomal protein of aminopyrine N-demethylase was higher in perivenous (PV) than in periportal (PP) hepatocytes of rat, but when it was expressed per cytochrome P-450 content the difference in the activity was not significant. 2. The activity of 7-ethoxycoumarin O-deethylase, when expressed per mg protein and per P-450 content, was significantly higher in PV than in PP cells. 3. The activities of dimethylnitrosamine(DMNA) N-demethylase and aniline p-hydroxylase were not significantly different between two subpopulations of isolated hepatocytes when either expressed per mg protein or per P-450 content.     

Urea synthesis in freshly isolated and in cultured periportal and perivenous hepatocytes.

Periportal hepatocytes isolated by digitonin/collagenase perfusion produced urea faster than did similarly prepared perivenous hepatocytes, in both the presence and the absence of amino acids and various urea precursors. There was no difference between the two cell types in rates of intracellular proteolysis. The initial difference in urea synthesis persisted for 5 days during primary culture, but then gradually disappeared. Our results demonstrate that the periportal dominance of urea formation is unrelated to the currently existing acinar microenvironment in the intact liver, but probably reflects differences in acinar key enzyme activities only slowly converging during culture.     

Characterization of isolated Fe-loaded rat hepatocytes prepared by collagenase perfusion

Summary Hepatocytes from livers of rats loaded by Fe-dextran treatment were isolated by an in situ collagenase perfusion technique and evaluated for their biochemical, cytochemical, and morphological characteristics in cell culture. Iron loads 15 times higher than in normal rat liver cells isolated in the same way were retained in the preparations with 40% present as hemosiderin. A simple centrifugation-mathematical approach is described for the calculation of Fe content in the hepatocyte (95%) and reticuloendothelial (5%) fractions in the isolates. The cells were cultured for 22 h without loss of protein synthesis capability or significant changes in cell count, viability, endogenous glutamate-oxaloacetate transaminase (GOT) or Fe and were morphologically similar in most respects to unloaded (normal) hepatocytes similarly cultured. Studies are in progress to assess the utility of these preparations as a model for Fe mobilization from Fe-loaded animals. This work is supported by National Institutes of Health Grants AM 25647-03 (M. Dawson, Principal Investigator) and GM 28158-01 (C. Tyson, Principal Investigator). The technical assistance of Mr. Jack E. Dabbs, Mr. Charles Hart, and Mr. Randy Douglas is acknowledged.     

Characterization of isolated Fe-loaded rat hepatocytes prepared by collagenase perfusion

Hepatocytes from livers of rats loaded by Fe-dextran treatment were isolated by an in situ collagenase perfusion technique and evaluated for their biochemical, cytochemical, and morphological characteristics in cell culture. Iron loads 15 times higher than in normal rat liver cells isolated in the same way were retained in the preparations with 40% present as hemosiderin. A simple centrifugation-mathematical approach is described for the calculation of Fe content in the hepatocyte (95%) and reticuloendothelial (5%) fractions in the isolates. The cells were cultured for 22 h without loss of protein synthesis capability or significant changes in cell count, viability, endogenous glutamate-oxaloacetate transaminase (GOT) or Fe and were morphologically similar in most respects to unloaded (normal) hepatocytes similarly cultured. Studies are in progress to assess the utility of these preparations as a model for Fe mobilization from Fe-loaded animals.     

Glutathione replenishment capacity is lower in isolated perivenous than in periportal hepatocytes.

The zonal distribution of GSH metabolism was investigated by comparing hepatocytes obtained from the periportal (zone 1) or perivenous (zone 3) region by digitonin/collagenase perfusion. Freshly isolated periportal and perivenous cells had similar viability (dye exclusion, lactate dehydrogenase leakage and ATP content) and GSH content (2.4 and 2.7 mumol/g respectively). During incubation, periportal cells slowly accumulated GSH (0.35 mumol/h per g), whereas in perivenous cells a decrease occurred (-0.14 mumol/h per g). Also, in the presence of either L-methionine or L-cysteine (0.5 mM) periportal hepatocytes accumulated GSH much faster (3.5 mumol/h per g) than did perivenous cells (1.9 mumol/h per g). These periportal-perivenous differences were also found in cells from fasted rats. Efflux of GSH was faster from perivenous cells than from periportal cells, but this difference only explained 10-20% of the periportal-perivenous difference in accumulation. Furthermore, periportal cells accumulated GSH to a plateau 26-40% higher than in perivenous cells. There was no significant difference in gamma-glutamylcysteine synthetase or glutathione synthetase activity between the periportal and perivenous cell preparations. The periportal-perivenous difference in GSH accumulation was unaffected by inhibition of gamma-glutamyl transpeptidase or by 5 mM-glutamate or -glutamine, but was slightly diminished by 2 mM-L-methionine. This suggests differences between periportal and perivenous cells in their metabolism and/or transport of (sulphur) amino acids. Our results suggest that a lower GSH replenishment capacity of the hepatocytes from the perivenous region may contribute to the greater vulnerability of this region to xenobiotic damage.     

Transglutaminase differentially regulates growth signalling in rat perivenous and periportal hepatocytes

The influence of transglutaminase 2 (TG2) activity on the proliferative effect of epidermal growth factor (EGF) and on EGF receptor affinity in periportal hepatocytes (PPH) and perivenous hepatocytes (PVH) has been investigated using a primary culture system. PPH and PVH subpopulations have been isolated using the digitonin/collagenase perfusion technique. DNA synthesis was assessed by [3H] thymidine incorporation into hepatocytes. The assay for binding of [125I] EGF to cultured hepatocytes was analysed by Scatchard plot analysis. Pretreatment with the TG2 inhibitor monodansylcadaverine (MDC) greatly increased EGF-induced DNA synthesis in both PPH and PVH. Furthermore, [125I] EGF binding studies in PVH treated with MDC indicated that high-affinity EGF receptor expression was markedly up-regulated, whereas in PPH, there was no significant effect. Treatment with retinoic acid (RA), an inducer of TG2 expression, significantly decreased EGF-induced DNA synthesis in both PPH and PVH. Binding studies in the presence of RA revealed that the high-affinity EGF receptor was down-regulated and completely absent in both PPH and PVH. These results suggest that TG2 was involved in the differential growth capacities of PPH and PVH through down-regulation of high-affinity EGF receptors.     

Distribution of hepatic glutaminase activity and mRNA in perivenous and periportal rat hepatocytes.

Perivenous and periportal hepatocytes were isolated by the digitonin/collagenase perfusion technique. The specific activity of phosphate-activated glutaminase was 2.33-fold higher in periportal cells than in perivenous cells. Similarly, the relative abundance of glutaminase mRNA was 2.6-fold higher in samples from periportal cells. The distribution of glutaminase activity and mRNA was compared with those for glutamine synthetase (predominantly perivenous) and phosphoenolpyruvate carboxykinase (predominantly periportal). The results suggest that phosphate-activated glutaminase is predominantly expressed in the periportal zone of the liver acinus.     

Different proliferative responses of periportal and perivenous hepatocytes to EGF.

Stimulation of DNA synthesis by EGF was compared in cultured periportal and perivenous hepatocyte populations. Periportal hepatocytes responded to EGF more sensitive (IC50-values 20 vs 75 ng/ml) and with a higher maximal stimulation (420 vs 290%) than perivenous hepatocytes with respect to both [3H]thymidine incorporation and labeling index. The glutamine synthetase-positive hepatocytes responded much less to EGF than did the perivenous cells in general. The simultaneous presence of insulin increased the sensitivity for EGF predominantly in the periportal hepatocytes. These inherent differences in the growth potential of hepatocytes from different acinar localizations may contribute to different growth patterns across the lobules in normal and regenerating liver.     

Regulation of glycogen synthesis from glucose and gluconeogenic precursors by insulin in periportal and perivenous rat hepatocytes.

Glycogen synthesis in hepatocyte cultures is dependent on: (1) the nutritional state of the donor rat, (2) the acinar origin of the hepatocytes, (3) the concentrations of glucose and gluconeogenic precursors, and (4) insulin. High concentrations of glucose (15-25 mM) and gluconeogenic precursors (10 mM-lactate and 1 mM-pyruvate) had a synergistic effect on glycogen deposition in both periportal and perivenous hepatocytes. When hepatocytes were challenged with glucose, lactate and pyruvate in the absence of insulin, glycogen was deposited at a linear rate for 2 h and then reached a plateau. However, in the presence of insulin, the initial rate of glycogen deposition was increased (20-40%) and glycogen deposition continued for more than 4 h. Consequently, insulin had a more marked effect on the glycogen accumulated in the cell after 4 h (100-200% increase) than on the initial rate of glycogen deposition. Glycogen accumulation in hepatocyte cultures prepared from rats that were fasted for 24 h and then re-fed for 3 h before liver perfusion was 2-fold higher than in hepatocytes from rats fed ad libitum and 4-fold higher than in hepatocytes from fasted rats. The incorporation of [14C]lactate into glycogen was 2-4-fold higher in periportal than in perivenous hepatocytes in both the absence and the presence of insulin, whereas the incorporation of [14C]glucose into glycogen was similar in periportal and perivenous hepatocytes in the absence of insulin, but higher in perivenous hepatocytes in the presence of insulin. Rates of glycogen deposition in the combined presence of glucose and gluconeogenic precursors were similar in periportal and perivenous hepatocytes, whereas in the presence of glucose alone, rates of glycogen deposition paralleled the incorporation of [14C]glucose into glycogen and were higher in perivenous hepatocytes in the presence of insulin. It is concluded that periportal and perivenous hepatocytes utilize different substrates for glycogen synthesis, but differences between the two cell populations in the relative utilization of glucose and gluconeogenic precursors are dependent on the presence of insulin and on the nutritional state of the rat.     

Dual-digitonin-pulse perfusion. Concurrent sampling of periportal and perivenous cytosol of rat liver for determination of metabolites and enzyme activities.

A previously described digitonin-perfusion technique [Quistorff, Grunnet & Cornell (1985) Biochem. J. 226, 289-297], by which intracellular material of rat liver could be liberated, has been refined, now allowing release of cytosol of high purity from both periportal and perivenous parts of the same liver. The cytosolic fractions are obtained by perfusing the liver for short intervals (10-20 s) with digitonin (4-5 mg/ml), first in the normal perfusion direction and then, after an interval of 1-2 min, in the retrograde direction, the eluate being collected during and after both intervals. The technique is termed 'dual-digitonin-pulse perfusion'. The eluate fractions showed a peak specific activity of the cytosolic enzymes alanine aminotransferase (ALAT), lactate dehydrogenase (LDH) and pyruvate kinase (PK) of 3-5-fold higher than obtained in a biopsy from the same liver. For glutamine synthetase (GS) a 10-fold higher specific activity was obtained. Zonation, defined as the ratio of the specific activities in periportal and perivenous eluates, of ALAT, LDH and PK was 10, 1.7 and 0.70 respectively. Zonation of GS was less than 0.01. These factors may be modified by a slight zonation of cytosolic protein of 1.2-1.3. Peak concentrations in the eluate of ATP, ADP, Pi, NAD+ and glycerol 3-phosphate were 32.5 +/- 11.4, 19.9 +/- 4.3, 71.9 +/- 25.4, 2.41 +/- 0.83 and 6.84 +/- 2.74 nmol/mg of protein for periportal eluates. There was no difference between periportal and perivenous eluates except for glycerol 3-phosphate, which was significantly higher in perivenous eluates, 12.8 +/- 4.5 nmol/mg of protein.     

Transglutaminase down-regulates the dimerization of epidermal growth factor receptor in rat perivenous and periportal hepatocytes

Objective:  Recently, we found that transglutaminase 2 (TG2) might be involved in the difference in proliferative capacities between periportal hepatocytes (PPH) and perivenous hepatocytes (PVH) through down-regulation of high-affinity epidermal growth factor receptor (EGFR). However, it is uncertain whether this high-affinity EGFR contributes to the hepatocyte growth signalling pathway. Here, we have investigated the influence of TG2 on EGF-induced EGFR dimerization and its phosphorylation, which are important steps in the hepatocyte proliferative/growth signalling pathway, in PPH and PVH.
Materials and methods:  PPH and PVH were isolated using the digitonin/collagenase perfusion technique. Amounts of TG2, EGFR dimerization and its phosphorylation were determined by Western blot analysis.
Results:  Pretreatment with monodansylcadaverine, an inhibitor of TG2, greatly increased EGF-induced EGFR dimerization and its phosphorylation in PVH compared with PPH. Conversely, treatment with retinoic acid, an inducer of TG2, significantly decreased EGF-induced EGFR dimerization and its phosphorylation with a significant increase in TG2 expression and its catalysed products, isopeptide bonds, in both subpopulations. It was found that EGFR served as a substrate for TG2.
Conclusion:  The present data showed good correlation with our previous data on EGF-induced DNA synthesis and EGFR-binding affinity to EGF. These results suggest that zonal difference in cell growth between PPH and PVH may be caused by down-regulation of EGFR dimerization and subsequent autophosphorylation through TG2-mediated cross-linking of EGFR.     

Hepatic zonation of drug metabolizing enzymes. Studies on hepatocytes isolated from the periportal or perivenous region of the liver acinus

Models of hepatic intraacinar zonation have been proposed previously; in most models, direct visualization of the acinar destruction is not possible while intact hepatocyte recovery-viability often presents a problem for subsequent metabolic studies. In the present studies, the liver is isolated in situ and perfused with Krebs-Henseleit buffer, pH 7.4. A 1.5-mL intrahepatic volume of a 7 mM digitonin solution is then injected at a flow rate of 6 mL/min for 15 s via the portal vein or via the vena cava for selective destruction of the periportal (PP) or perivenous (PV) region of the acinus. To avoid diffusion of the detergent throughout the acinus, the liver is then immediately perfused with oxygenated Hanks buffer in the direction opposite to that of digitonin injection. The preparation can then be used for histological evaluation, for studies on isolated-perfused liver, or for isolation of hepatocytes. Direct visualization of the acinar destruction can be achieved by coloring the permeabilized cells with 0.2 mM trypan blue; the liver is then fixed in situ by a 10-min perfusion with paraformaldehyde and histological evaluation is achieved by eosine staining of liver slices. Following isolation of hepatocytes by collagenase perfusion, a highly significant PV localization was found for the synthesis of glutamine, the N-demethylation of aminopyrine, and the glucuronidation of p-nitrophenol, whereas a highly significant PP zonation was found for alanine aminotransferase. By contrast, no specific acinar zonation was found for the enzymes 7-ethoxycoumarin O-deethylase and aniline p-hydroxylase. Total cytochrome P-450 was 0.42 +/- 0.006 and 0.4 +/- 0.03 nmol/10(6) hepatocytes in PV and PP, respectively (nonsignificant).(ABSTRACT TRUNCATED AT 250 WORDS)     

Quantitative histochemical study of glucose-6-phosphatase in periportal and perivenous human hepatocytes.

The glucose-6-phosphatase activity of periportal and perivenous human hepatocytes was studied with a quantitative method. The results obtained with histogram of light intensity distributions indicate that the enzyme reaction was 1.3 to 2.5 fold higher in periportal zone than in perivenous zone. The profiles of light intensity along portal----hepatic venous distances show a progressive decrease of enzyme activity with highest values in periportal hepatocytes.     

Coexpression of periportal and perivenous enzymes in rat hepatocytes after experimental bile duct ligation: Comparison with intrasplenically transplanted hepatocytes

The coexpression of normally periportal and perivenous markers has been described in heterotopically transplanted hepatocytes. To determine whether such a coexpression might also occur in hepatocytes retaining their original intrahepatic location, we compared in bileduct-ligated livers and intrasplenically transplanted hepatocytes, the expression and distribution of the predominantly periportal glucose-6-phosphatase, succinate dehydrogenase, and lactate dehydrogenase, the predominantly perivenous glutamate dehydrogenase, NADPH-dehydrogenase, and -hydroxybutyrate dehydrogenase, and the strictly perivenous glutamine synthetase. The coexpression of high levels of the two periportal markers glucose-6-phosphatase and lactate dehydrogenase and of the perivenous marker NADPH dehydrogenase was observed in two situations: in clusters of hepatocytes isolated within the ductular proliferation in bile-duct-ligated livers and the majority of intrasplenically transplanted hepatocytes. The expression of glutamine synthetase was different according to the site. The protein was observed in certain intrasplenically transplanted hepatocytes bordering the splenic vessels but was never detected in hepatocyte clusters found in bile-duct-ligated livers. Our study therefore suggests that the coexpression of periportal and perivenous markers in the same hepatocytes is likely to be a non-specific consequence of the loss of the normal connections of hepatocytes with the normal liver microcirculation.