Glucocorticoid regulation of hepatic TNF production following cecal ligation and puncture sepsis.

Tumor necrosis factor-alpha (TNF) is a critical early mediator in the genesis of a systemic inflammatory response during a septic insult. Many of the harmful effects evident during sepsis are ascribed to excessive endogenous TNF production. Because the liver is an important source of circulating TNF during endotoxicosis, and because glucocorticoids are believed to have a regulatory role in suppressing endogenous TNF production, we evaluated the effect of adrenalectomy on the hepatic production of TNF in an isolated perfused liver model after cecal ligation and puncture (CLP) sepsis. Fasted, male Holtzman rats (n = 4/group) underwent CLP alone, adrenalectomy (ADREX) alone, or CLP plus ADREX (CLP/ADREX). Two hours after the operation, the rat livers were explanted and perfused in an isolated recirculating model. Serum TNF levels were greater in CLP/ADREX rats than in both other groups. TNF production in the perfused liver was greater in the CLP/ADREX rats when compared with either CLP alone or ADREX alone. A separate mortality study was performed (N = 35) that demonstrated a CLP induced mortality of 45%, and a CLP/ADREX mortality of 100%. Thus, adrenalectomy increased circulating TNF and hepatic TNF production as well as mortality in CLP sepsis. These findings suggest an important role for endogenous glucocorticoids in modulating hepatic TNF production during CLP-induced sepsis.

Molsidomine increases endotoxic survival and decreases cytokine production.

We hypothesized that nitric oxide (NO) may exert feedback regulatory control over cytokine production and improve survival in endotoxin (ETX) shock. To test this hypothesis, we evaluated the pre-endotoxin effect of the NO donor molsidomine (MOL) on circulating tumor necrosis factor (TNF), interleukin (IL)-1, and IL-6 levels, the production of these cytokines in the perfused liver, and endotoxic lethality in mice. Male BDF mice weighing 28-32 g were administered either 100 mg/kg MOL or saline (SAL) i.p. Thirty minutes later, the mice received either 50 mg/kg Salmonella enteriditis ETX or SAL i.p. Mice were killed at 90 min after ETX or SAL, for either the determination of plasma cytokine levels by enzyme-linked immunosorbent assays or for use in the perfused liver assessment of cytokine production. MOL treatment significantly reduced the post-ETX circulating levels of TNF to 84%, IL-1 to 65%, and IL-6 to 56%, as compared with SAL-treated ETX controls. Endotoxic livers from MOL-pretreated mice produced 82% less TNF, 88% less IL-1, and 54% less IL-6 over a 2 h perfusion, as compared with SAL-treated ETX controls. MOL pretreatment also decreased lethality in ETX shock from 90 to 50% (p < .05). Therefore, NO may provide a beneficial effect during sepsis by inhibiting hepatic cytokine production, and thus providing survival benefits.

Partial hepatectomy reduces the endotoxin-induced peak circulating level of tumor necrosis factor in rats.

The pro-inflammatory cytokine tumor necrosis factor (TNF) is dramatically and transiently elevated in the circulation during endotoxic and septic shock and is a primary mediator in the pathogenesis of the sepsis syndrome. TNF peaks in the circulation 90 min after endotoxin administration with little variation, even among species. The specific cells, tissues, or organs that produce the high circulating level of TNF in septic shock remain unknown. The most likely sources are macrophage-laden tissues such as the liver and the spleen and circulating blood leukocytes. This study evaluated whether the liver is an important source producing the TNF spike 90 min after endotoxin. To test this hypothesis, we measured the peak circulating level of TNF following an endotoxin injection in rats subjected to a two-thirds hepatectomy versus sham-operated controls. Hepatechectomized rats produced 64% less TNF after endotoxin than controls (857 + or - 143 pg/mL plasma vs. 2410 + or - 491, respectively; p < .01). In contrast, splenectomy did not significantly after peak TNF levels versus sham-operated controls following an endotoxin injection (1380 + or - 148 pg/mL plasma vs. 1710 + or - 291). Furthermore, incubation of rat blood with endotoxin for 90 min did not significantly increase TNF above controls. These experiments demonstrate an important role for the liver in producing the high circulating levels of TNF after an endotoxin injection and suggest that hepatic-specific cytokine modulation deserves study for a therapeutic benefit in septic shock.

Bronchoalveolar lavage fluid endotoxin elevation in human single lung transplant recipients during rejection.

Endotoxin has been implicated as an aetiological factor in liver transplant rejection and acute graft-versus-host disease. To investigate the role of endotoxin in human single lung transplant rejection we measured the level of endotoxin in bronchoalveolar lavage (BAL) fluid from six subjects at baseline and during rejection, which was defined histologically from transbronchial biopsy. Differential cell counts of BAL fluid cells, the levels of protein and albumin in BAL fluid, and serum albumin levels were also examined. BAL fluid albumin to serum ratio was calculated to evaluate alveolar-capillary leakage. A significant elevation of BAL fluid endotoxin with rejection compared with baseline was observed. Standardizing endotoxin levels to BAL fluid volume, protein, or albumin were all of similar significance. Examination of BAL fluid cell population revealed a significant elevation in the percentage of lymphocytes with rejection. No significant difference between BAL fluid protein levels, BAL fluid albumin levels. BAL fluid albumin to protein ratio, serum albumin levels, or BAL fluid albumin to serum albumin ratio was seen with rejection. We conclude that BAL fluid endotoxin levels increase during lung rejection, endotoxin levels can be accurately standardized to millilitres of BAL fluid, and abnormal alveolar-capillary leakage does not appear to account for the increased BAL fluid endotoxin.

Glucose transporters (GLUT1, 2, & 4) in fat, muscle and liver in a rat model of endotoxic shock.

To explore the mechanisms for hypoglycemia in our rat model of septic shock, we examined whether changes occur in glucose transporter isoform protein level. Total membrane protein fractions were collected from tissues 6 to 8 hours after endotoxin injection at which time animals exhibited hypoglycemia (7.2 +/- 0.5 vs. 2.6 +/- 1.2mM) and lactacidemia (1.0 +/- 1.0 vs. 5.1 +/- 1.8mM/L) as compared to saline-treated controls. The protein level of glucose transporter isoforms GLUT1 and 4 in fat did not significantly change in septic shock when compared to control animals (126 +/- 22% and 114 +/- 79%, respectively). Likewise, no change was seen in GLUT1 or 4 in muscle (124 +/- 52% and 101 +/- 28%, respectively). The protein abundance of isoforms GLUT1 and 2 in liver were not significantly altered (123 +/- 35% and 101 +/- 23%, respectively). Septic shock induced hypoglycemia cannot be directly explained by changes in total glucose transporter protein levels.

Peritoneal dialysis using conventional, lactate--containing solution sterilized by ultrafiltration.

Peritoneal dialysis was performed on 7 end-stage renal failure patients using an I-lactate-containing solution sterilized by ultrafiltration through polyamide filters. The patients tolerated the procedure well while their metabolic acidosis and azotemia improved.

Polymyxin-B suppresses endotoxin-induced insulin hypersecretion in pancreatic islets.

Endotoxin induces insulin hypersecretion in vivo, which results in hyperinsulinemia and glucose dyshomeostasis. Polymyxin-B (PMX-B), an inhibitor of protein kinase C (PKC), has been shown to ameliorate the consequences of endotoxin-induced hyperinsulinemia in vivo. To explore the mechanism for this effect in vitro, this study determined whether PMX-B could alter endotoxin-induced insulin hypersecretion in isolated pancreatic islets of Langerhans. Pancreases were obtained from fasted, male, Sprague-Dawley rats treated with either saline (control) or endotoxin (S. enteritidis B, 16.7 mg/kg, i.v.). Three hours after the experimental treatment, islets were isolated by collagenase digestion and then incubated for 1 hr in Krebs Ringer bicarbonate buffer containing 0.5% bovine albumin, 10 mM HEPES, 300 mg/dl D-glucose, phorbol 12-myristate 13-acetate (PMA, 1 microM when present), and PMX-B (1 or 10 mM when present). In the absence of PMA and PMX-B, "endotoxic" islets hypersecreted immunoreactive insulin (IRI) relative to control islets. PMA, the prototypical PKC activator, significantly increased IRI secretion from both control and "endotoxic" islets. The additional inclusion of PMX-B (1 mM or 10 mM) in the incubation media significantly reduced insulin secretion from both control and "endotoxic" islets and suppressed the insulin hypersecretion observed in "endotoxic" islets. Since insulin secretion occurs at least partially through mechanisms dependent on PKC activation, the ability of PMX-B to suppress insulin hypersecretion in "endotoxic" islets suggests that activation of PKC within pancreatic beta-cells may play a role in the excess insulin secretion and hyperinsulinemia associated with endotoxicosis.

Peritoneal dialysis using bicarbonate-containing solution sterilized by ultrafiltration.

We performed acute peritoneal dialysis on eight end-stage renal disease patients using a bicarbonate-containing solution sterilized by ultrafiltration through polyamide filters. The patients tolerated the procedure well; their azotemia and metabolic acidosis improved.

Absence of the Staub-Traugott effect in endotoxicosis.

Intravenous glucose tolerance tests (IVGTTs) using 1.2 g/kg of D-glucose in fasted, male rats were repeated twice, with an interval of 75 min, during early (2.5 hr) and late (26.5 hr) nonlethal endotoxicosis induced by Salmonella enteritidis endotoxin (ETX) at 10 micrograms/kg i.v. In comparison with saline controls, glucose tolerance was dramatically increased in early endotoxicosis; however, it was restored to near normal in late endotoxicosis. Augmentation of tolerance by repeated IVGTTs i.e. the Staub-Traugott effect, occurred in the saline control groups; however, it was absent in both early and late endotoxicosis. The elevations of plasma insulin (IRI) in response to the glucose loads were exaggerated vs. controls in both early and late endotoxicosis; however, the hyperinsulinemias were not significantly diminished by repeated IVGTTs in any group. These results indicate loss of the Staub-Traugott effect in endotoxicosis that is 1) independent of augmented insulin secretory responses to a repeated glucose challenge and 2) dependent on some adaptive cellular mechanisms to repeated glucose loads which were impaired in nonlethal endotoxicosis.

Antagonism of endotoxic glucose dyshomeostasis by protein kinase C inhibitors.

Activation of protein kinase C (PKC) by bacterial lipopolysaccharide had recently been implicated in the pathogenetic sequence of gram-negative sepsis, endotoxicosis, hyperinsulinism, and the alterations in glucoregulation that eventuate in glucose dyshomeostasis. This study used the peptide antibiotic polymyxin B (PMX-B) and H-7, an isoquinoline sulfonamide, as inhibitors of PKC activation to evaluate responses to provocative insulin and glucose tolerance tests in control vs. endotoxic rats. Fed male rats were treated with either Salmonella enteritidis endotoxin (ETX; 0.33 mg/kg iv) or saline 120 min before intravenous insulin tolerance testing (IVITT) with human insulin (1 U/kg) or intravenous glucose tolerance testing (IVGTT) with D-glucose (1.2 g/kg). H-7 in dimethyl sulfoxide at 25 mg/kg, PMX-B in saline at 0.25 mg/kg, or the respective vehicles were administered 5 min before the tolerance tests. Neither H-7 nor PMX-B had any significant acute effects on basal plasma glucose or lactate values. The decline in plasma with IVITT was augmented by ETX; however, concomitant H-7 or PMX-B attenuated the insulin hypoglycemia. The computed half-life of glucose in the IVGTT was decreased by ETX; however, concomitant H-7 or PMX-B decreased the tolerance alteration. In addition, both H-7 and PMX-B attenuated the rise in insulin induced by the IVGTT. Thus the hyperinsulinism and the glucoregulatory disturbances in endotoxicosis may be mediated by PKC activation and ameliorated by PKC inhibition.

Altered glucose transporter mRNA abundance in a rat model of endotoxic shock.

To better understand molecular mechanisms of glucose transport in shock, we studied glucose transporter isoform mRNA abundance after injection of S. enteritidis endotoxin (40 mg/kg) or saline. Six to 8 hours after injection, endotoxin-treated animals compared to controls became hypoglycemic (44 +/- 6 vs. 111 +/- 4 mg/dl) and lactacidemic (5.9 +/- 0.5 vs. 1.3 +/- 0.1). At such times, tissue RNA was isolated and hybridized to Riboprobes for GLUT1 (erythrocyte), GLUT2 (liver), and GLUT4 (muscle/fat) glucose transporter isoforms and expressed as percent of control. GLUT1 mRNA abundance was increased in fat (660%, p less than .05), soleus muscle (314%, p less than .05), and liver (871%, p less than .001) of endotoxin-treated rats. Soleus muscle GLUT4 mRNA levels were increased (+33%, p less than .02), while liver GLUT2 mRNA levels were markedly decreased (-58%, p less than .01). The overall increase in GLUT1 mRNA abundance accompanied by lowered liver GLUT2 mRNA levels may either cause or reflect profoundly altered glucose transport.

Augmentation of endotoxic lethality and glucose dyshomeostasis by phorbol ester.

To investigate the role of protein kinase C (PKC) activation in the pathogenesis of endotoxin (ETX) shock, the in vivo effects of phorbol 12-myristate 13-acetate (PMA) on ETX-induced lethality and glucose dyshomeostasis were determined. Fed rats (300-400 g) were treated intravenously with incremental doses of Salmonella enteritidis ETX and either the vehicle, 110 mg/kg ip dimethyl sulfoxide (DMSO), or 0.5 mg/kg ip PMA dissolved in DMSO. PMA significantly increased ETX-induced lethality to doses of 1.0-20 mg/kg. PMA augmented the initial hyperglycemia, late hypoglycemia, and hyperlactacidemia after 1 mg/kg iv ETX to rats anesthetized with pentobarbital sodium. In contrast, 4 alpha-phorbol, a phorbol derivative that does not activate PKC, had no effect on either lethality or the glucose and lactate responses. Hyperinsulinemia after 1 mg/kg iv ETX was prolonged by PMA but not by 4 alpha-phorbol. Insulin tolerance testing (0.5 U/kg iv) produced an exaggerated hypoglycemic response in PMA-treated endotoxic (0.33 mg/kg) rats. Glucose tolerance to 1.2 g/kg iv was increased by ETX and PMA attenuated the increased tolerance. Thus PKC activation may be involved in the pathogenesis of lethal endotoxicosis and associated glucose dyshomeostasis.

Comparative and interactive in vivo effects of tumor necrosis factor alpha and endotoxin.

A comparative and interactive analysis of the effects of tumor necrosis factor alpha (TNF) and endotoxin (ETX) on selected hemodynamic and glucoregulatory alterations was performed in conscious, unrestrained, adult male Holtzman rats. Rats with indwelling carotid artery and jugular vein cannulae were administered intravenous (i.v.) bolus injections of either (1) ETX at 1.55 or 30 mg/kg; (2) TNF at 0.1, 0.5, or 1.0 mg/kg; or (3) TNF plus ETX as a low dose co-treatment at 0.1 mg/kg plus 1.55 mg/kg, respectively. Control groups received either saline or heat-inactivated TNF. TNF induced a lethal response such that 1.0 mg/kg resulted in five out of six deaths within 6 hr. Elevated pulse rates, early hyperglycemia, late hypoglycemia, hyperlactacidemia, hypoinsulinemia, and elevated catecholamine concentrations were evident after injection of 1.0 mg/kg TNF. The pathophysiological alterations observed after 1.0 mg/kg TNF were comparable to the changes observed after the administration of a highly lethal, 30 mg/kg dose of ETX (six out of six deaths within 24 hr). Co-treatment with low doses of TNF plus ETX resulted in the rapid demise of the rats, resulting in six out of six deaths within 4 hr. The resultant shocklike state was accompanied by significant hypotension, hyperglycemia and hyperlactacidemia, similar to the changes induced by highly lethal doses of TNF or ETX alone. This study supports the involvement of TNF in the pathogenesis of gram-negative septic shock and documents the hemodynamic and glucoregulatory alterations which accompany the exacerbated shocklike state induced after co-treatment with separately, minimally lethal doses of TNF plus ETX.

Exogenous ATP and hepatic hemodynamics in the perfused rat liver.

The actions of exogenous adenosine triphosphate (ATP) on hepatic hemodynamics were analyzed in the isolated perfused rat liver. The effects of ATP (10, 40, and 160 microM) on hepatic circulatory resistances were determined from changes in portal vein flow, portal vein perfusion pressure, hepatic arterial perfusion pressure, and hepatic capacitance. ATP decreased portal vein flow under conditions of constant perfusion pressure, but it increased portal vein perfusion pressure under constant flow perfusion. ATP increased the resistance of the presinusoidal regions and reduced hepatic capacitance. Adenine derivatives including adenosine, adenosine monophosphate (AMP), adenosine diphosphate (ADP), and ATP also decreased portal vein flow; the order of vascular activity of the adenine compounds was ATP greater than ADP greater than AMP = adenosine. The shock-protective actions of ATP may reflect hepatic hemodynamic adjustments.

Phagocytes and monokines: perspectives on metabolic regulation in sepsis.

The importance of monocytes and macrophages in antimicrobial host defense physiology has long been recognized. The mononuclear phagocyte system includes the precursor cells in bone marrow, or promonocytes, circulating peripheral blood monocytes, and the resident tissue macrophages of both fixed and free types. This review focuses on the pathophysiological role of the monokines, especially interleukin-1 in metabolic regulation during sepsis.

Exogenous ATP and carbohydrate metabolism in the rat liver.

The present study analyzed the actions of exogenous ATP on liver carbohydrate metabolism by using both isolated perfused rat livers and hepatocytes as model systems. The metabolic parameters measured in the perfused livers were glucose production, lactate production, oxygen consumption, and K+ movement. All parameters were monitored continuously during the perfusions in order to assess transient effects. ATP increased both glucose and lactate production in the isolated rat liver when perfused with either constant pressure or constant flow. ATP increased oxygen consumption in the perfused liver at a low concentration (10 microM) but was inhibitory at high concentrations (40 and 160 microM). ATP also increased K+ efflux in the isolated perfused liver. ATP did not alter basal glucose production in isolated hepatocytes; however, it did inhibit glucagon-stimulated glucose output by hepatocytes. ATP inhibited lactate production in the isolated hepatocytes. Extracellular ATP has the potential to modify basal as well as hormonal modulation of hepatic carbohydrate metabolism during shock.

Dexamethasone alters glucose, lactate, and insulin dyshomeostasis during endotoxicosis in the rat.

This study examined the effects of dexamethasone on the glucose, lactate, and insulin dyshomeostasis of endotoxicosis in the rat. To assess the effects of dexamethasone on carbohydrate dyshomeostasis, sequential measurements of plasma glucose and lactate were made in vivo. The effects of dexamethasone on endotoxin-induced portal and systemic hyperinsulinemia also were evaluated in vivo. The ability of dexamethasone to alter the insulin hypersecretory state of the endotoxic pancreas was evaluated by using the in vitro perfused rat pancreas. Endotoxicosis caused severe hypoglycemia and hyperlactacidemia in the pentobarbital-anesthetized rat. Dexamethasone prevented the hypoglycemia, significantly reduced the hyperlactacidemia, and prevented endotoxin-induced hyperinsulinemia in vivo. Dexamethasone reduced endotoxin-induced pancreatic hypersecretion of insulin when administered as a pretreatment 1 h prior to endotoxin treatment; however, when administered as a cotreatment with endotoxin, no significant reduction of insulin hypersecretion occurred. The results of this study suggest that the ameliorating effects of dexamethasone on glucose and lactate dyshomeostasis occur not only at the peripheral tissue level, but also through effects on insulin secretion.

Glucose dyshomeostasis and cardiovascular failure in endotoxic dogs.

Fasted mongrel dogs were anesthetized with pentobarbital sodium and instrumented for the on-line measurement of blood glucose (BG), a lead II electrocardiogram, and pressures in the left ventricle and pulmonary and systemic artery concomitant with on-line monitoring of BG. Serum insulin was measured by radioimmunoassay, and cardiac output (CO) was determined by thermodilution. Stroke work (SW) and pulmonary and systemic resistances were calculated. After a 30-min control period dogs were treated with Escherichia coli endotoxin (E) or normal saline (S) and then observed for 10 h or until death. Preinjection control BG was maintained in S dogs, and early hyperglycemia (H) was observed in six dogs; in contrast 10 E dogs showed no hyperglycemia (NH). During the late stages all E dogs were markedly hypoglycemic. In both groups of E dogs an early hyperinsulinemia occurred. CO and SW were depressed in both groups of E dogs. These variables were significantly lower in NH than in H dogs. Pulmonary and systemic resistance progressively increased in NH dogs after endotoxin administration. The results suggest that the ability to increase blood glucose levels after endotoxin injection is important for the maintenance of cardiovascular function. Glucose dyshomeostasis leading to hypoglycemia, however, may be a factor in the development of endotoxic cardiovascular failure.