Intrahepatic nuclear factor-kappa B activity and alpha 1-acid glycoprotein transcription do not predict outcome after cecal ligation and puncture in the rat.

OBJECTIVE: Sepsis is the leading cause of death in critically ill surgical patients. Septic hepatic dysfunction, an important determinant of outcome, is poorly understood but includes inappropriate transcriptional down-regulation. This may be modulated by proinflammatory cytokines. We hypothesized that intrahepatic changes in tumor necrosis factor (TNF)/interleukin (IL)-1-linked processes, such as the activation of the p50 homodimeric and the p65/p50 heterodimeric isoforms of the transcription factor nuclear factor (NF)-kappaB or transcription of the acute phase reactant alpha1-acid glycoprotein (AGP), would correlate with recovery from sepsis. DESIGN: Prospective experimental comparison of sham operation and nonlethal and lethal sepsis in male Sprague-Dawley rats. INTERVENTIONS: Nonlethal sepsis was induced by using single-puncture cecal ligation and puncture (CLP). Lethal sepsis was induced via double-puncture CLP. NF-kappaB DNA binding activity was determined by using electrophoretic mobility shift analysis with differentiation of p50/p50 and p50/p65 isoforms by using appropriate antibodies. AGP transcription was assessed with transcription elongation analysis, intrahepatic IL-1beta, and TNF-alpha abundance by using immunohistochemistry, and serum IL-1beta was assessed by using ELISA. MAIN RESULTS: Overall NF-kappaB activity increased equivalently over time after both single- and double-puncture CLP, with a peak occurring 3 hrs after intervention. In single-puncture CLP, there was an increase in the binding of the p50 homodimer form over time. After double-puncture CLP, no such change was observed. AGP transcription was increased equivalently in both models. Intrahepatic IL-1beta was detected 16 and 24 hrs after single-puncture CLP and 6 hrs after double-puncture CLP. After double-puncture CLP, intrahepatic TNF-alpha was detected at 6, 16, and 24 hrs. Serum IL-1beta was undetectable after both single- and double-puncture CLP. CONCLUSIONS: Although AGP transcription was similar in mild and fulminant sepsis, double-puncture CLP increased the binding activity of the p50 homodimer relative to binding of the p50/p65 NF-kappaB heterodimer. These results imply that transcriptional activity not linked to acute phase responses is an important determinant of outcome in sepsis.

Compensatory hepatic regeneration after mild, but not fulminant, intraperitoneal sepsis in rats.

Sepsis is the leading cause of death in surgical intensive care units. Although both mild sepsis secondary to cecal ligation and single puncture (CLP) and fulminant, double puncture CLP (2CLP) may provoke hepatocyte death, we hypothesize that regeneration compensates for cell death after CLP but not 2CLP. In male Sprague-Dawley rats, hepatic necrosis, as determined by serum alpha-glutathione S-transferase (alpha-GST) levels, was significantly but equally elevated over time after both CLP and 2CLP. Apoptosis, evaluated using both terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling and morphological examination, was minimal after both CLP and 2CLP. Regeneration, assayed by staining tissue for incorporation of exogenously administered bromodeoxyuridine, was present after CLP but not after 2CLP. To further substantiate impaired regeneration, steady-state levels of mRNAs encoding JunB, LRF-1, and cyclin D1 were determined. After 2CLP, the absence of JunB, LRF-1, and cyclin D1 mRNAs confirmed failed activation of the mitogen-activated protein kinase-linked proliferative pathway and progression through the cell cycle. Therefore, failed hepatocyte regeneration may be a manifestation of hepatic dysfunction in fulminant sepsis.

Intraabdominal sepsis down-regulates transcription of sodium taurocholate cotransporter and multidrug resistance-associated protein in rats.

Hepatic dysfunction in sepsis is characterized by hyperbilirubinemia and intrahepatic cholestasis. We hypothesize that sepsis causes decreased hepatic transcription of the bile acid transporter sodium taurocholate cotransporter (Ntcp) and the organic anion transporter multidrug resistance-associated protein (Mrp2) and that interleukin (IL)-6 is important in the down-regulation of Ntcp and Mrp2 expression. Male Sprague-Dawley rats underwent induction of mild, nonlethal sepsis by cecal ligation and single puncture (CLP) or fulminant sepsis by cecal ligation and double puncture (2CLP). Hepatic transcription of Ntcp and Mrp2 rapidly decreased after CLP or 2CLP. Seventy-two hours later, transcription was 60% of baseline in CLP and 14% of baseline in 2CLP. Serum bilirubin was elevated from 24 h onward and cholestasis was observed on fixed liver specimens at 24, 48, and 72 h after 2CLP but not after CLP. Steady-state Ntcp and Mrp2 mRNA was decreased in IL-6-treated cultured hepatocytes and in normal rats given 1 mg/kg intravenous IL-6. We conclude that 1) Ntcp and Mrp2 transcription is down-regulated transiently after CLP and persistently after 2CLP; 2) 2CLP results in hyperbilirubinemia and cholestasis, in part due to persistently decreased transcription of Ntcp and Mrp2; and 3) altered Ntcp and Mrp2 transcription is mediated in part by IL-6.

Hepatic gene expression and cytokine responses to sterile inflammation: comparison with cecal ligation and puncture sepsis in the rat.

Inflammatory stimulation of hepatic acute phase protein expression is, in part, modulated by tumor necrosis factor-alpha (TNFalpha), interleukin-1beta (IL-beta), and IL-6. These cytokines also may mediate some aspects of the persistent inflammation and metabolic dysregulation of sepsis. Cecal ligation and puncture (CLP) sepsis in male Sprague-Dawley rats inappropriately decreases hepatocellular transcription of phosphoenolpyruvate carboxykinase (PEPCK), glucose-6-phosphatase (G6Pase), carnitine palmitoyltransferase II (CPTII), acetyl CoA acyltransferase (ACA), and ornithine transcarbamylase (OTC). We hypothesize that 1) transcriptional reprogramming does not occur after simple inflammation induced by subcutaneous turpentine injection, 2) the pattern of acute phase gene expression after CLP differs from that following turpentine injection, and 3) the different responses reflect differences in the intrahepatic activity of TNFalpha/IL-1beta or IL-6. Gene expression, transcription factor activity, and cytokine abundance were determined after either a subcutaneous injection of turpentine or CLP. After turpentine injection, PEPCK, G6Pase, CPTII, ACA, and OTC expression were unchanged, different from previously reported data following CLP. Both turpentine injection and CLP increased expression of TNFalpha/IL-1beta-regulated alpha1-acid glycoprotein, and IL-6-regulated alpha2-macroglobulin and decreased expression of transthyretin (a negative acute phase protein). However, the magnitude and temporal pattern of expression differed. Turpentine injection increased the activity of the TNFalpha/IL-1beta-linked transcription factor NF-kappaB and the intrahepatic abundance of TNFalpha in a manner similar to that observed after CLP but only slightly altered the activity of the IL-6-linked transcription factor Stat-3 and intrahepatic IL-6 abundance. This differed significantly from observations after CLP. We conclude that CLP-induced alterations in hepatic gene expression may reflect differences in IL-6 activity.

Intrahepatic STAT-3 activation and acute phase gene expression predict outcome after CLP sepsis in the rat.

Interleukin-6 (IL-6) regulates hepatic acute phase responses by activating the transcription factor signal transducer and activator of transcription (STAT)-3. IL-6 also may modulate septic pathophysiology. We hypothesize that 1) STAT-3 activation and transcription of alpha2-macroglobulin (A2M) correlate with recovery from sepsis and 2) STAT-3 activation and A2M transcription reflect intrahepatic and not serum IL-6. Nonlethal sepsis was induced in rats by single puncture cecal ligation and puncture (CLP) and lethal sepsis via double-puncture CLP. STAT-3 activation and A2M transcription were detected at 3-72 h and intrahepatic IL-6 at 24-72 h following single-puncture CLP. All were detected only at 3-16 h following double-puncture CLP and at lower levels than following single-puncture CLP. Loss of serum and intrahepatic IL-6 activity after double-puncture CLP correlated with mortality. Neither intrahepatic nor serum IL-6 levels correlated with intrahepatic IL-6 activity. STAT-3 activation following single-puncture CLP inversely correlated with altered transcription of gluconeogenic, ketogenic, and ureagenic genes. IL-6 may have both beneficial and detrimental effects in sepsis. Fulminant sepsis may decrease the ability of hepatocytes to respond to IL-6.

Sepsis-induced depression of rat glucose-6-phosphatase gene expression and activity.

Sepsis in rats decreases the hepatic expression of the gluconeogenic enzyme glucose-6-phosphatase (G6Pase). The aim of this study was to investigate the relationship among G6Pase transcription, mRNA, enzymatic activity, and serum glucose levels at different intervals during mild or fulminant sepsis. Both fulminant and mild sepsis immediately decreased hepatic G6Pase mRNA levels. In mild sepsis, levels began to recover late in the time course. Serum glucose levels were maintained in mild sepsis but decreased markedly in fulminant sepsis. G6Pase transcription after fulminant sepsis decreased and never recovered. A similar transcriptional decrease was noted in mild sepsis, but some recovery occurred in this state. Histochemistry after mild sepsis revealed a decrease in G6Pase protein and enzymatic activity that paralleled transcription. These studies suggest that changes in G6Pase transcription and activity are early markers for sepsis-induced alterations in hepatic function. Mechanisms other than gene expression and enzymatic activity serve to maintain glucose levels in mild sepsis, but in the fulminant disorder, compensatory mechanisms fail and hypoglycemia develops.

Altered hepatic gene expression in fecal peritonitis: changes in transcription of gluconeogenic, beta-oxidative, and ureagenic genes.

Sepsis alters energy production and utilization by the liver. Changes in both metabolic pathways that produce substrate (gluconeogenesis or ketogenesis) for organism-wide consumption or provide an alternative source of fuel for the liver (beta-oxidation and amino acid metabolism) have been identified. In this study, we test the hypothesis that these changes occur via an alteration in the transcription of key enzymes within each pathway. Male Sprague-Dawley rats were made septic using cecal ligation and single puncture with sham operated animals serving as controls. Hepatic tissue was harvested at 0, 3, 6, 16, 24, 48, and 72 h had either total RNA or hepatic nuclei were isolated. Using Northern blot hybridization analysis, the steady-state levels of phosphoenolpyruvate carboxykinase, glucose-6-phosphatase, carnitine palmitoyl transferase II, acetyl coenzyme A-acyltransferase, and ornithine transcarbamylase mRNAs were determined. Using transcript elongation analysis, the rate of transcription of each gene was investigated. Relative to control, steady-state mRNA levels and rates of transcription for all five genes were decreased by ligation and single puncture. These decreases were persistent, with only partial recovery of either mRNA levels or transcription rates at 72 h. These findings could explain in part the long-term alterations in gluconeogenesis, beta-oxidation, and ureagenesis observed in sepsis. More importantly, decreased transcription of certain genes seems to be a characteristic of sepsis-induced changes in hepatic function. Understanding the mechanisms that decrease transcription also may explain other aspects of sepsis in the liver and other organ systems.

Acute-phase gene expression correlates with intrahepatic tumor necrosis factor-alpha abundance but not with plasma tumor necrosis factor concentrations during sepsis/systemic inflammatory response syndrome in the rat.

OBJECTIVES: To test the hypothesis that after cecal ligation and puncture in the rat, there is increased expression of the tumor necrosis factor (TNF)/interleukin-1-dependent, acute-phase reactant alpha 1-acid glycoprotein in the liver, and that this change correlates temporally with increased abundance of TNF-alpha in the hepatic parenchyma but not with circulating concentrations of TNF-alpha. DESIGN: Prospective, randomized, controlled study. SETTING: Research laboratory at the University of Pennsylvania School of Medicine. SUBJECTS: Male, adolescent Sprague-Dawley rats, weighing 200 to 300 g. INTERVENTIONS: The procedure of cecal ligation and single puncture with an 18-gauge needle was performed in one group of animals. Control animals underwent sham operation. At 0, 3, 6, 16, 24, 48, and 72 hrs after either procedure, blood was collected and the liver was isolated and perfusion-fixed with 2% paraformaldehyde. In a second group of animals, liver tissue was harvested for isolation of total hepatic RNA. MEASUREMENTS AND MAIN RESULTS: Northern blot hybridization analysis demonstrated an increase in steady-state concentrations of alpha 1-acid glycoprotein messenger RNA that peaked at 16 hrs after cecal ligation and puncture. The alpha 1-acid glycoprotein messenger RNA was not detected in control animals. TNF-alpha concentrations in the plasma, as determined by enzyme-linked immunosorbent assay, were detected 3 and 6 hrs after cecal ligation and puncture. However, TNF-alpha concentrations were undetectable in the plasma at other time points after cecal ligation and puncture and at all time points in the sham-operated animals. Immunohistochemical staining of 7-micron hepatic sections demonstrated a progressive increase in TNF-alpha abundance, with a peak at 16 hrs. Alterations in alpha 1-acid glycoprotein gene expression correlated in time with intrahepatic TNF-alpha abundance, but not with plasma TNF-alpha concentrations. CONCLUSIONS: The changes in TNF-alpha-dependent hepatic gene expression that accompany an animal model of the systemic inflammatory response syndrome correlate with intrahepatic, and not circulating, TNF-alpha concentrations and reflect paracrine, and not endocrine, activity. Therefore, plasma concentrations of TNF-alpha do not appropriately reflect hepatocellular responses during the systemic inflammatory response syndrome.

Increased expression of cytokine-induced neutrophil chemoattractant in septic rat liver.

Hepatocellular dysfunction in sepsis may be neutrophil mediated. We therefore tested the hypothesis that sepsis-induced neutrophil accumulation is associated with increased expression of the chemokine, cytokine-induced neutrophil chemoattractant (CINC). In Sprague-Dawley rats made septic by cecal ligation and puncture, we demonstrate a time-dependent increase in CINC mRNA, which returns to baseline by 48 h. By in situ hybridization, this mRNA is present in hepatocytes and nonparenchymal cells. CINC protein levels in septic animals parallel mRNA levels and resolve by 48 h. Because CINC expression is induced by cytokines including tumor necrosis factor-alpha (TNF- alpha), we show, by immunohistochemistry, that sepsis elevates intrahepatic TNF-alpha. Finally, because the CINC promoter is transactivated by the transcription factor, nuclear factor kappa B (NF-kappa B), we determined that hepatic NF-kappa B DNA binding increases dramatically, peaking 16 h after cecal ligation and puncture. Thus activated NF-kappa B may mediate CINC induction in sepsis. This constellation of findings suggests a mechanism by which sepsis may induce neutrophil accumulation in the liver and may have implications regarding sepsis-induced hepatic dysfunction.