Decreased albumin and increased fibrinogen secretion by single hepatocytes from rats with acute inflammatory reaction

Effects of hepatic zonal oxygen levels on hepatocyte stress responses

BACKGROUND

Hepatocytes spend their lifetimes in a gradient of oxygen, hormones, and enzymes. We used a three-dimensional Matrigel model to determine whether hepatocytes cultured at perivenous (zone 3) oxygen levels differed in susceptibility to anoxia-induced cell injury compared with hepatocytes cultured at periportal (zone 1) oxygen levels.

MATERIALS AND METHODS

Hepatocytes were harvested from Sprague Dawley rats and cultured at 9% oxygen (hepatic zone 1) or 5% oxygen (hepatic zone 3) and stressed at 0% oxygen. Microscopy, real-time reverse transcriptase-polymerase chain reaction, and enzyme-linked immunosorbent assay were used to assess cell viability, mitochondrial potential, acute phase responses, and membrane blebbing.

RESULTS

Hepatocytes cultured in Matrigel with HepatoZyme medium at zone 1 and zone 3 oxygen conditions were viable for 1 wk and showed acute phase responses as measured by interleukin-6-induced fibrinogen production. In response to 3 h anoxia, cells maintained at the perivenous oxygen level showed increased membrane blebbing and increased loss of mitochondrial membrane potential in comparison to the periportal oxygen cultured cells. Cells at perivenous oxygen also showed a reduced ability to recover following reoxygenation.

CONCLUSIONS

Hepatocytes can remain viable and functional for extended periods in culture at low oxygen levels that mimic the hepatic perivenous environment, yet these cells are more susceptible to anoxia-induced damage than hepatocytes cultured at the periportal oxygen level. The small population of perivenous hepatocytes may be critical in determining the fate of the liver during ischemia/reperfusion since hepatocytes cultured at that concentration appear to be more labile in response to anoxia.

Effects of tumor necrosis factor-alpha on glucose and albumin production in primary cultures of rat hepatocytes

Tumor necrosis factor-alpha (TNF) is known to alter significantly in vivo hepatic glucose and albumin metabolism. However, it remains unclear whether the observed effects represent direct actions of this factor or secondary responses due to the recruitment of other mediator systems. The present study was designed to investigate direct actions of TNF on glucose and albumin production in primary cultures of rat hepatocytes. Addition of TNF to the culture medium resulted in a 45% to 50% reduction in glucose production from a control level of 239 +/- 15 nmol/plate.h. This effect was reversed by addition of anti-TNF monoclonal antibody. In glycogen-depleted cells, short-term (5-hour) incubation with TNF did not affect hepatocyte albumin secretion, which was 8.13 +/- 0.29 microgram/plate.h. However, in cells exposed to insulin or in non-glycogen-depleted cells, addition of TNF resulted in a 10% to 25% reduction in albumin production. These findings indicate that TNF exerts direct inhibitory effects on hepatocyte glucose and albumin production, but the effects on the latter process are modest. A notable aspect of the findings is that the albumin effects are insulin or glucose substrate-dependent, which may have implications regarding liver function during nutritional support in critical illness.

Decreased toxicity of polymorphonuclear neutrophils toward hepatocytes isolated from rats with acute inflammatory reaction

We have recently demonstrated that polymorphonuclear neutrophils were toxic to hepatocytes through a protease-mediated mechanism. Since synthesis of antiproteases is markedly increased during acute inflammatory reaction, the aim of this work was to investigate the toxicity of neutrophils against normal vs. inflammatory rat hepatocytes. Acute inflammatory reaction was induced by subcutaneous injection of turpentine 24 hr before the experiments. Hepatocytes from normal and turpentine-treated rats were isolated by collagenase digestion. They were incubated with human neutrophils stimulated by 1 mg/ml opsonized zymosan. Cytotoxicity was quantified by the percentage of alanine aminotransferase activity released by hepatocytes in culture medium after an 18-hr incubation period. By comparison to normal hepatocytes, inflammatory hepatocytes were more resistant to the toxicity of neutrophils. At a neutrophil/hepatocyte ratio of 20:1, the alanine aminotransferase activity releases were 53.7% +/- 5.4% (mean +/- 1 S.E.) and 27.4% +/- 4.8% for normal and inflammatory hepatocytes, respectively. Similarly, inflammatory hepatocytes were found to be less sensitive than normal hepatocytes to the toxic effect of purified neutrophil cathepsin G. In contrast, both types of hepatocytes exhibited the same sensitivity to H2O2 generated by a system consisting of glucose and glucose oxidase. Two arguments suggested that the resistance of inflammatory hepatocytes to protease toxicity was explained by an increased production of antiproteases by these cells: (a) when tested against cathepsin G and porcine pancreatic elastase activities, the protease inhibitory capacity of conditioned medium from inflammatory hepatocytes was higher than that of conditioned medium from normal hepatocytes; (b) conditioned medium from inflammatory hepatocytes markedly reduced the toxicity of stimulated neutrophils as that of cathepsin G.(ABSTRACT TRUNCATED AT 250 WORDS)

Correlation between plasma membrane surface area and transferrin secretion rate in isolated hepatocytes

It is generally considered that in exocytosis the size of the secreting cells does not increase when the membranes of exocytosis vesicles fuse with the plasma membrane. As the factors involved in the regulation of this phenomenon are poorly understood, we thought it worthwhile to investigate the relationship between the plasma membrane surface area and secretory activity. Isolated rat hepatocytes were prepared by liver collagenase perfusion. Secretion of the plasma protein, transferrin (Tf) was detected at the single cell level with specific anti-rat transferrin antibodies using the reverse hemolytic plaque test. Hepatocyte surface and hemolytic ring surface areas were calculated from diameters of hepatocyte and hemolytic plaque measured after 5h of incubation. A highly significant correlation was established between the plaque-forming hepatocyte surface areas and the corresponding hemolytic surface areas. This result was confirmed using an automatic image analysis method. Two-month-old rats were compared to 4-month-old rats. We observed that the ratio of the quantity of transferrin secreted by hepatocytes to the hepatocyte surface area was constant for a given incubation time, whatever the size of the hepatocytes. These results suggest that the plasma membrane surface area of hepatocytes may constitute a limiting factor in Tf secretion.

Inhibitory effect of the acute inflammatory reaction on liver regeneration after partial hepatectomy in the rat

We evaluated the influence of an acute inflammatory reaction that triggers the synthesis of exportable proteins by hepatocytes on liver regeneration after partial hepatectomy, which induces the synthesis of proteins necessary for liver cell proliferation. In hepatectomized rats with a turpentine-induced acute inflammatory reaction, the first peaks of hepatic DNA synthesis and mitosis were significantly inhibited compared with pair-fed controls subjected to partial hepatectomy only, and the liver DNA concentration at various times after partial hepatectomy was significantly lower in the former than in the latter. Inhibition was not obtained when the acute inflammatory reaction was induced 12 h or more before partial hepatectomy, suggesting that the inhibitory effect of turpentine administration depended on early events in the acute inflammatory reaction. These data suggest that one possible mechanism responsible for inhibition of regeneration might be competition at the transcriptional or the translational level between liver syntheses of various proteins, and that, under certain conditions, liver-specific functions might take precedence over regenerative functions.

Effect of ethanol on turpentine-induced acute phase response in rats

Turpentine-induced acute-phase response and its modulation by ethanol in rats at 48 hr has been studied. There was more than 2.3-5.1 fold increase in fibrinogen and seromucoids concentrations in plasma, accompanied by 28% decline in albumin concentration in turpentine-stimulated rats. The fractional synthesis rate of these two acute-phase proteins was increased by 4.1-6.4 fold, while that of albumin (non acute-phase protein) was reduced by 32.6%. Ethanol inhibited this induction of acute-phase protein synthesis at 48 hr. The inhibition of acute-phase response by ethanol was significantly more pronounced for seromucoids than for fibrinogen and appeared to be dependent on the carbohydrate content of the acute-phase glycoprotein.

A simple bioassay for monocyte-derived hepatocyte stimulating factor: increased synthesis of alpha 2-macroglobulin and reduced synthesis of albumin by cultured rat hepatocytes

Cytokines released from monocytes upon stimulation by lipopolysaccharide cause a number of cells to undergo proliferative and synthetic changes. At least one of these cytokines affects hepatocytes in vivo causing increased synthesis of a series of acute-phase proteins. We have established an in vitro micro-assay for hepatocyte stimulating factor (HSF) using primary cultures of normal rat hepatocytes. Measurement of increased synthesis of alpha 2-macroglobulin and decreased synthesis of albumin caused by exogenously added factor constitute a sensitive parameter for quantifying HSF. For comparing various cytokines preparations, we have defined a unit of HSF activity in terms of a stimulation index. We have used this assay to follow some preliminary attempts to isolate the factors responsible for stimulation of synthesis of acute-phase reactant by the liver.

The acute-phase response of cultured rat hepatocytes. System characterization and the effect of human cytokines

Hepatocytes were isolated from adult livers and cultured for periods of up to 5 days as monolayers at an initial density of 10(6) cells/10cm2 in Williams E medium containing insulin, dexamethasone and 5% foetal-calf serum. The daily production of 11 plasma proteins was measured by electroimmunoassay and compared with the concentrations of the same proteins in the plasma of normal rats and of those with experimental inflammation. Hepatocytes from normal rats synthesized proteins in relative amounts which were similar to the relative proportions of the same proteins in the plasma of turpentine-injected animals. The pattern changed only slowly during 5 days in culture, but it did so profoundly either when the medium was devoid of dexamethasone or when human cytokines (from endotoxin-stimulated monocytes or unstimulated human squamous-carcinoma cell line COLO-16) were added. The cytokines consistently increased the synthesis of alpha 2-macroglobulin and fibrinogen and depressed that of albumin; variable increases in the synthesis of alpha 1-acute-phase globulin, alpha 1-acid glycoprotein, haptoglobin and alpha 1-proteinase inhibitor, and variable decreases in transferrin synthesis, were seen, whereas the synthesis of antithrombin III, alpha 1-macroglobulin and prothrombin remained virtually unaffected. The cytokine effects on protein synthesis required the presence of dexamethasone. The hepatocyte-stimulating activity derived from monocytes chromatographed on Sephadex G-100 corresponding to 30 000 Da, as opposed to the lymphocyte-activating factor, which was eluted as a molecule of approx. 15 000 Da. This suggests that both activities probably reside with distinct molecular species in the preparations of human cytokines.