Tumor necrosis factor alpha alters the cytotoxic effect of hydrogen peroxide in cultured hepatocytes.

We examined whether tumor necrosis factor-alpha (TNF) affects the cytotoxic capacity of reactive oxygen species on rat hepatocytes in culture. Both TNF and reactive oxygen species are involved in many inflammatory events including hepatic ischemia/reperfusion injury and endotoxic shock. Synchronous treatment of hepatocytes with both TNF and H2O2 demonstrated that TNF (2000 ng/ml) enhanced the cytotoxic effect of H2O2 (500 microM). By contrast, pretreatment with TNF (2000 ng/ml) for 24 h followed by exposure to H2O2 (1000 microM) reduced the reactive oxygen-induced cytotoxicity. We conclude that TNF increases the effects of reactive oxygen-induced cytotoxicity when exposed synchronously, whereas TNF pretreatment induces a cytoprotective effect to reactive oxygen species, presumably by up-regulation of the reduced form of glutathione levels in hepatocytes.

Role of epinephrine in TNF and IL-6 production from isolated perfused rat liver.

A bidirectional communication exists between the nervous system and the immune system. Evidence has accumulated suggesting that cytokines-immune peptides influence sympathetic neuronal survival and that cytokines can promote the secretion of catecholamines. Using an isolated perfused rat liver (IPRL) preparation, we have shown that the liver is an important source of circulating cytokines in response to lipopolysaccharide (LPS) and that corticosterone dose dependently influenced LPS-induced production of tumor necrosis factor (TNF) and interleukin-6 (IL-6). In this study, we investigated the direct effect of epinephrine (another stress hormone) on the production of TNF and IL-6 in liver. We demonstrated that epinephrine (1 microM/ml) alone did not induce TNF bioactivity but significantly increased IL-6 bioactivity from IPRL effluent. When the IPRL was infused with LPS, epinephrine significantly decreased TNF bioactivity. Epinephrine in LPS-treated livers also significantly increased IL-6 bioactivity. Both responses were totally inhibited by the beta-blocker propranolol (10 microM/ml). Anisomycin, a protein synthesis inhibitor, infused into the IPRL completely blocked the rise in TNF and IL-6 concentrations in the effluent leaving the IPRL, supporting the hypothesis that the synthesis (or release) of these cytokines was dependent on protein synthesis. We then attempted to determine whether epinephrine exerts similar effects in vitro. Using isolated Kupffer cells and hepatocytes, we found that epinephrine alone had no effect on TNF and IL-6 production in Kupffer cells and hepatocytes but significantly decreased LPS-induced TNF bioactivity and increased LPS-induced IL-6 bioactivity in Kupffer cells. Our data support the hypothesis that epinephrine can promote IL-6 secretion from IPRL.(ABSTRACT TRUNCATED AT 250 WORDS)

Role of corticosterone in TNF and IL-6 production in isolated perfused rat liver.

Using an isolated perfused rat liver (IPRL) preparation, we assessed whether corticosterone may contribute to the rise in tumor necrosis factor (TNF) and interleukin-6 (IL-6) in rats after injection with lipopolysaccharide (LPS) or exposure to psychological stress. Intravenous infusion of LPS into the IPRL led to dose-dependent increases in TNF and IL-6 concentrations in the effluent. Anisomycin, a protein synthesis inhibitor, completely blocked the rise in TNF and IL-6 concentration in the IPRL effluent, supporting the hypothesis that the synthesis (or release) of these cytokines was dependent on protein synthesis. Intravenous infusion of corticosterone at nonstressed (35 ng/ml) and stressed levels (350 ng/ml) increased TNF and/or IL-6 release. However, when LPS was combined with corticosterone, the lower dose of corticosterone facilitated the release of cytokines, whereas the higher dose of corticosterone suppressed the release of cytokines. We then showed that isolated Kupffer cells were capable of significant TNF and IL-6 production and that corticosterone decreased LPS-induced cytokine production in these cells. Our data support the hypothesis that the liver is an important source of circulating cytokines in response to LPS. In addition, although in vitro data generally support the hypothesis that corticosterone suppresses the production of cytokines, our in situ data support the hypothesis that physiological levels of corticosterone cause an increase in TNF and IL-6.

Hepatic ischemia/reperfusion injury: importance of oxidant/tumor necrosis factor interactions.

Previous studies have demonstrated a role for both tumor necrosis factor (TNF) and reactive oxygen intermediates (ROI) in hepatic ischemia/reperfusion (I/R) injury. Biologically active TNF was present in liver homogenates in ischemic and nonischemic lobes after 2 h of ischemia but without reperfusion. Using an in situ liver perfusion model, we measured ROI, TNF, and hepatic enzymes in the effluent after 2 h of ischemia. Increased reduction of ferricytochrome C was observed in the hepatic effluent, indicative of the formation of ROI. Treatment of animals with TNF neutralizing antisera significantly reduced both ROI and aspartate aminotransferase (AST). Animals treated with superoxide dismutase (SOD), or SOD + catalase (CAT) had greater TNF in the hepatic effluent compared with I/R alone; however, SOD or SOD + CAT did not cause additional release of AST.SOD + CAT plus anti-TNF serum resulted in significant protection compared with SOD + CAT plus control serum. Reperfusion of ischemic liver with 4 mM H2O2 increased both TNF and AST. Optimal protection of hepatocellular injury from reperfusion injury is achieved with a combination of antioxidants and inhibition of TNF.

Tumor necrosis factor, interleukin 6, and the acute phase response following hepatic ischemia/reperfusion.

We report here the production of systemic levels of tumor necrosis factor (TNF) and interleukin 6 (IL-6) and associated changes in serum macroglobulin to albumin ratios in a nonlethal rat model of partial hepatic ischemia/reperfusion (I/R). Plasma IL-6 was detectable and elevated at 1 hr of reperfusion as compared to sham-operated controls (I/R rats = 12,100 +/- 3860 pg/ml; sham rats = 5260 +/- 842 pg/ml; IL-6 values = means +/- SEM) and reached maximal levels at 6 hr of reperfusion (I/R rats = 47,400 +/- 25,700 pg/ml; sham rats = 3370 +/- 394 pg/ml), in contrast to maximal TNF levels at 30 min of reperfusion (I/R rats = 72 +/- 15 pg/ml; sham rats = 2 +/- 2 pg/ml; TNF values = means +/- SEM). Pretreatment with neutralizing TNF antisera prior to ischemia resulted in a reduction of IL-6 at 1 hr of reperfusion (anti-TNF = 3870 +/- 2550 pg/ml; control serum = 7650 +/- 1670 pg/ml), but was without effect on IL-6 levels at subsequent time points over the 24 hr of reperfusion. Electrophoretic determination of macroglobulin (alpha 1 + alpha 2) and albumin concentrations in sham-operated and ischemia/reperfusion animals demonstrated an elevation in the macroglobulin/albumin ratio in both groups over time, suggestive of an acute phase response, and the ratio was unchanged by immunoneutralization of TNF prior to ischemia/reperfusion. We conclude that this model of hepatic ischemia/reperfusion results in temporally distinct systemic elevations in IL-6 and TNF; however, the induction of IL-6 and the associated changes in serum macroglobulin concentration are independent of TNF.

Interleukin-6 expression in immunologically elicited murine macrophages.

We report the expression of both interleukin-6 (IL6) messenger RNA and biological activity in complete Freund's adjuvant-elicited peritoneal macrophages (CFA-M phi). IL6 mRNA expression peaked between 4 and 8 h of lipopolysaccharide (LPS) stimulation; biological activity was maximal at approximately 18 h of stimulation. LPS-induced IL6 mRNA was inhibited by treatment with cycloheximide (5 micrograms/ml), implicating the participation of a secondary protein mediator in the induction process, or the dependence upon protein synthesis for receptor ligand interactions. Comparison of CFA-M phi with resident peritoneal macrophages suggests that the elicited cell population makes more IL6 in response to LPS than the resident population on a per cell basis.

In vivo biologic and immunohistochemical analysis of interleukin-1 alpha, beta and tumor necrosis factor during experimental endotoxemia. Kinetics, Kupffer cell expression, and glucocorticoid effects.

Using a model of sepsis induced by parenteral challenge of mice with bacterial lipopolysaccharide (LPS), the authors analyzed the in vivo expression of interleukin-1 (IL-1) alpha,beta and tumor necrosis factor (TNF). Both TNF and IL-1 alpha,beta were detected in hepatic sinusoidal macrophages (Kupffer cells), immunohistochemically. Kinetic analysis showed a clear sequence of synthesis. Tumor necrosis factor was produced first, reaching maximal expression at 1 hour after LPS challenge, then rapidly disappeared. IL-1 beta followed, reaching maximal expression at 2 to 3 hours, then dropped off by 6 hours. Interleukin-1 alpha expression reached a peak at 6 hours and had disappeared by 18 hours. Analysis of serum bioactivity also revealed sequential expression that correlated with immunohistochemical findings. Tumor necrosis factor was maximal at 1 hour and IL-1 at 6 hours. The IL-1 bioactivity was not due to interleukin-6 (IL-6), as this was depleted from specimens by immunoabsorption. Also IL-6 bioactivity reached maximal levels at 3 hours, earlier than IL-1. Pretreatment with 4 mg/kg dexamethasone significantly decreased Kupffer cell expression of TNF and IL-1 alpha (about 80% and 60% suppression, respectively) but had less effect on IL-1 beta expression (about 30% suppression). Accordingly, serum levels of TNF were suppressed by 75% while serum IL-1 was decreased by 39%, indicating differential sensitivity of these cytokines to glucocorticoids. Endogenous corticosteroid levels increased as TNF levels decreased, supporting the contention that glucocorticoids regulate TNF synthesis. In contrast, IL-1 levels rose concurrently with corticosterone. These data indicate a sequential activation of cytokine gene expression in vivo, which may be critical to the cascade of events leading to septic shock, and provide evidence that Kupffer cells are a major source of cytokines in endotoxemia. Finally, the differential sensitivity of cytokine expression to glucocorticoids may in part explain the inadequacy of the latter in the treatment of sepsis.

Regulation of monokine gene expression: prostaglandin E2 suppresses tumor necrosis factor but not interleukin-1 alpha or beta-mRNA and cell-associated bioactivity.

Prostaglandin E2 (PGE2)-mediated suppression of macrophage interleukin-1 alpha,beta and tumor necrosis factor-alpha synthesis was examined at the cellular and molecular levels. Treatment of lipopolysaccharide (LPS)-stimulated adjuvant-elicited murine macrophages with 5 x 10(-7) M PGE2 caused a 70% reduction in cell-associated TNF but had no suppressive effect on cell-associated interleukin-1 (IL-1) activity. Consistent with this result, Northern blot and nuclear transcription analyses demonstrated suppression of TNF mRNA but PGE2 had no effect on IL-1 alpha and IL-1 beta mRNA accumulation, as compared to LPS controls. Immunoperoxidase staining for cell-associated TNF alpha, IL-1 alpha, and IL-1 beta demonstrated that PGE2 suppressed TNF, but not IL-1 alpha or -beta expression, supporting the bioassay data. These results imply that PGE2-mediated regulation of IL-1 alpha,beta and TNF alpha is quite distinct. Synthesis of TNF appears to be regulated at least at the level of transcription, whereas that for IL-1 alpha and -beta is regulated post-transcriptionally.

In situ hybridization analysis of macrophage-derived tumor necrosis factor and interleukin-1 mRNA.

Tumor necrosis factor (TNF) and interleukin-1 (IL-1) play an intimate role in the initiation and maintenance of inflammatory reactions due to their pluripotent activities. In this paper, we describe the use of an in situ hybridization analysis as an effective means to probe for TNF and IL-1 mRNA levels in primary macrophage cultures and macrophage cell lines. A significant increase in lipopolysaccharide (LPS)-induced TNF mRNA accumulation was demonstrated by in situ hybridization using either a 35S-labeled synthetic oligonucleotide (30-mer) complementary to TNF mRNA or a 35S-randomly primed labeled TNF DNA probe. An augmentation in TNF mRNA accumulation, as assessed by increasing grains/cell, was demonstrated over a wide concentration range of LPS. This accumulation was shown using both immunologically elicited primary macrophage cultures and the macrophage cell line RAW 264.7. Interestingly, the RAW 264.7 constitutively produced TNF in the absence of specific stimulus and this tonic production was observed at the molecular level via in situ hybridization analysis. Specificity of the in situ hybridization technique was shown by a complete loss in binding of 35S-probe after either RNase digestion or competition with "cold-labeled" probe. beta-actin served as a 35S-labeled control probe where the number of actin-specific grains/cell was not altered by stimulating macrophages with LPS. IL-1 alpha mRNA was also increased by LPS stimulation of macrophages as assessed by in situ hybridization. The LPS-dependent increase in macrophage mRNA for TNF and IL-1 alpha, as assessed by in situ hybridization, was confirmed by classical Northern blot analysis as well as the production of biologically-active protein.

Human circadian rhythms in temperature, trace metals, and blood variables.

The literature supports the concept that circadian changes in body temperature reflect changes in the thermoregulatory set point. We were interested in studying the relationship between the circadian rhythm in body temperature and 24-h variations in plasma concentrations of iron, zinc, circulating leukocyte counts, and plasma interleukin 1 (IL-1) activity. Eight healthy men were studied for two separate 48-h sessions. Rectal temperature, plasma iron and zinc concentrations, plasma IL-1 activity, circulating leukocyte counts, and several other blood variables were monitored. Circadian rhythms in temperature, trace metals, and various leukocyte populations were demonstrated. The 24-h pattern of changes in plasma concentrations of iron and zinc approximate an inverse relationship with rectal temperature. Although we were unable to detect any IL-1 activity in human plasma collected at 4-h intervals, the daily changes in plasma trace metal concentrations and the variations in leukocyte populations may provide indirect evidence for a daily variation in local (e.g., in liver) or central nervous system release of IL-1.

Effect of antipyretic drugs on circadian rhythm in body temperature of rats.

In many species, including the laboratory rat, body temperature varies on a circadian (24 h) basis. There is considerable evidence that the circadian rise in body temperature is attributable to an elevation in thermoregulatory set point. We hypothesized that this rise in set point may be mediated by prostaglandins. If this hypothesis is correct, then it should be possible to block or reduce the nighttime rise in body temperature by the administration of prostaglandin synthesis inhibitors. Rats were injected with the prostaglandin synthesis inhibitors sodium salicylate, acetylsalicylic acid, and indomethacin at 5:00 P.M. and at 9:00 A.M. Administration of these drugs had little effect on body temperature during the day but caused a significant fall in body temperature at night when temperature is normally in the rising or plateau phase of the cycle. We conclude that prostaglandin synthesis is an important component of the circadian rise in body temperature in the rat. In addition, evidence is presented that there exists a cryogenic factor that opposes the nighttime prostaglandin-mediated rise in body temperature.

Low-molecular-weight substance in peritoneal exudate is antibacterial at febrile temperature.

Peritoneal exudate was collected from rabbits 18 hr after ip injection of shellfish glycogen. The fluid was centrifuged and the peritoneal exudate supernatant (PES) retained for use in growth studies. The growth of Pasteurella multocida in PES was inhibited at 41 degrees C (febrile temperature for rabbits) as compared with 39 degrees C (afebrile temperature), suggesting the presence of an antibacterial agent active at febrile temperature. Further studies indicated that this antibacterial substance has a molecular weight less than 5000 Da. Heat treatment (70 degrees C, 1 hr) had no influence upon the activity of the inhibitory factor.