Endotoxin in human disease. Part 1: Biochemistry, assay, and possible role in diverse disease states

Effects of bactericidal/permeability-increasing protein on endotoxin-induced fever and Escherichia coli-induced shock in rabbits

Binding of bactericidal/permeability-increasing protein (BPI) to endotoxin inhibits endotoxin-triggered responses. We investigated the effects of BPI on endotoxin fever and E. coli-induced septic shock in rabbits. Pre-incubation of endotoxin with BPI blocked fever compared to control rabbits (n = 6). A marked reduction in fever was also observed when BPI was injected before endotoxin. E. coli-challenge resulted in 66% mortality (n = 6); pre-treatment with BPI resulted in survival of all animals (n = 3). Mean arterial blood pressure was higher in BPI-treated compared to control rabbits. Comparable leukopenia and thrombocytopenia was observed with either BPI or vehicle treatment. Tumor necrosis factor (TNF) and interleukin-1 receptor antagonist were similarly elevated in both BPI- and saline-treated rabbits. However, in BPI treated rabbits, peak TNF levels were 34 % lower compared to saline controls ( P < 0.05). Further studies are warranted to assess whether BPI may have therapeutic potential for the treatment of septic shock.

Delayed therapy with a polymyxin B-dextran conjugate (PMX-622) improves survival in rabbits with Gram-negative peritonitis

Polymyxin B (PMB) is an amphipathic nephrotoxic antibiotic, which has been shown to neutralize the effects of endotoxin both in vitro and in vivo. PMB-D70 (PMX-622), a covalent conjugate of PMB with dextran 70 (D70), is less nephrotoxic than the parent compound. We sought to determine whether therapy with PMB-D70, in addition to conventional antimicrobial chemotherapy, could improve survival in a model of Gram-negative peritonitis. At T = 0 h, New Zealand white rabbits were implanted intraperitoneally with 10 ml of a suspension containing hemoglobin (40 μg/ml), mucin (150 μg/ml), and 1.0 ± 0.2 x 104 cfu/kg of viable Escherichia coli (O18:K1). Beginning at T = 4 h, the rabbits were treated with gentamicin (5 mg/kg every 12 h) for five doses or until death, and infused for 24 h or until death with either D70 or PMB-D70. Two pairs of groups were studied (doses indicate cumulative amounts infused over 24 h). The PMB-D70 (low dose) group received PMB-D70 (5 mg/kg of the PMB component) and the D70 (low dose) group received an equivalent dose of D70. The PMB-D70 (high dose) group received PMB-D70 (10 mg/kg of the PMB component) and the D70 (high dose) group received an equivalent dose of D70. Results for the two PMB-D70 groups, on the one hand, and the two D70 group, on the other hand, were statistically indistiguishable and, accordingly, were pooled for all analyses. Survival at 7 days was 11/25 (44%) for rabbits treated with PMB-D70 as compared to 2/23 (9%) for animals treated with D70 ( P = 0.007). We conclude that adjuvant treatment with PMB-D70 improves survival in a clinically relevant model of Gram-negative sepsis.

CRTH2 is a critical regulator of neutrophil migration and resistance to polymicrobial sepsis

Although arachidonic acid cascade has been shown to be involved in sepsis, little is known about the role of PGD(2) and its newly found receptor, chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2), on the septic response. Severe sepsis is associated with the failure of neutrophil migration. To investigate whether CRTH2 influences neutrophil recruitment and the lethality during sepsis, sepsis was induced by cecal ligation and puncture (CLP) surgery in mice. CRTH2 knockout (CRTH2(-/-)) mice were highly resistant to CLP-induced sepsis, which was associated with lower bacterial load and lower production of TNF-α, IL-6, and CCL3. IL-10, an anti-inflammatory cytokine, was higher in CRTH2(-/-) mice, blunting CLP-induced lethality in CRTH2(-/-) mice. Neutrophil accumulation in the peritoneum was more pronounced after CLP in CRTH2(-/-) mice, which was associated with higher CXCR2 levels in circulating neutrophils. Furthermore, sepsis caused a decrease in the level of acetylation of histone H3, an activation mark, at the CXCR2 promoter in wild-type neutrophils, suggesting that CXCR2 expression levels are epigenetically regulated. Finally, both pharmacological depletion of neutrophils and inhibition of CXCR2 abrogated the survival benefit in CRTH2(-/-) mice. These results demonstrate that genetic ablation of CRTH2 improved impaired neutrophil migration and survival during severe sepsis, which was mechanistically associated with epigenetic-mediated CXCR2 expression. Thus, CRTH2 is a potential therapeutic target for polymicrobial sepsis.

Insulin suppresses endotoxin-induced oxidative, nitrosative, and inflammatory stress in humans

OBJECTIVE

To investigate whether insulin reduces the magnitude of oxidative, nitrosative, and inflammatory stress and tissue damage responses induced by endotoxin (lipopolysaccharide [LPS]).

RESEARCH DESIGN AND METHODS

Nine normal subjects were injected intravenously with 2 ng/kg LPS prepared from Escherichia coli. Ten others were infused with insulin (2 units/h) for 6 h in addition to the LPS injection along with 100 ml/h of 5% dextrose to maintain normoglycemia.

RESULTS

LPS injection induced a rapid increase in plasma concentrations of nitric oxide metabolites, nitrite and nitrate (NOM), and thiobarbituric acid-reacting substances (TBARS), an increase in reactive oxygen species (ROS) generation by polymorphonuclear leukocytes (PMNLs), and marked increases in plasma free fatty acids, tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), monocyte chemoattractant protein-1 (MCP-1), macrophage migration inhibition factor (MIF), C-reactive protein, resistin, visfatin, lipopolysaccharide binding protein (LBP), high mobility group-B1 (HMG-B1), and myoglobin concentrations. The coinfusion of insulin led to a total elimination of the increase in NOM, free fatty acids, and TBARS and a significant reduction in ROS generation by PMNLs and plasma MIF, visfatin, and myoglobin concentrations. Insulin did not affect TNF-α, MCP-1, IL-6, LBP, resistin, and HMG-B1 increases induced by the LPS.

CONCLUSIONS

Insulin reduces significantly several key mediators of oxidative, nitrosative, and inflammatory stress and tissue damage induced by LPS. These effects of insulin require further investigation for its potential use as anti-inflammatory therapy for endotoxemia.

S100A12 mediates aortic wall remodeling and aortic aneurysm

RATIONALE

S100A12 is a small calcium binding protein that is a ligand of RAGE (receptor for advanced glycation end products). RAGE has been extensively implicated in inflammatory states such as atherosclerosis, but the role of S100A12 as its ligand is less clear.

OBJECTIVE

To test the role of S100A12 in vascular inflammation, we generated and analyzed mice expressing human S100A12 in vascular smooth muscle under control of the smooth muscle 22alpha promoter because S100A12 is not present in mice.

METHODS AND RESULTS

Transgenic mice displayed pathological vascular remodeling with aberrant thickening of the aortic media, disarray of elastic fibers, and increased collagen deposition, together with increased latent matrix metalloproteinase-2 protein and reduction in smooth muscle stress fibers leading to a progressive dilatation of the aorta. In primary aortic smooth muscle cell cultures, we found that S100A12 mediates increased interleukin-6 production, activation of transforming growth factor beta pathways and increased metabolic activity with enhanced oxidative stress. To correlate our findings to human aortic aneurysmal disease, we examined S100A12 expression in aortic tissue from patients with thoracic aortic aneurysm and found increased S100A12 expression in vascular smooth muscle cells.

CONCLUSIONS

S100A12 expression is sufficient to activate pathogenic pathways through the modulation of oxidative stress, inflammation and vascular remodeling in vivo.

Cellular mechanisms in sepsis

Mortality remains very high among septic patients despite the advanced treatments rendered in intensive care units. The development of septic shock is multifactorial. Tissue damage and organ dysfunction may be caused not only by the microorganisms but also by the inflammatory mediators released in response to the infection. Cytokines (tumor necrosis factor, interleukin-1, interleukin-6, interleukin-8, high-mobility group box-1 protein, macrophage migratory inhibitory factor) and noncytokines (nitric oxide, platelet-activating factor, complements, and eicosonoids) may inflict tissue injury and contribute to multiple organ dysfunction and cell death (or apoptosis). Gram-negative bacteria are the most common organisms identified in septic patients. The pathological effects of gram-negative bacteria are conveyed through lipopolysaccharide derived from the bacterial cell membrane. Lipopolysaccharide activates the nuclear factor kappa B, which triggers the release of inflammatory mediators. Protein components from gram-positive bacteria, fungi, or viruses may evoke the activation of nuclear factor kappa B in a similar fashion as lipopolysaccharide. Endogenous anti-inflammatory mediators are released in response to the infection and act to control the overwhelming systemic inflammatory response. The fragile balance between negative and positive feedback on the inflammatory mediators is the key factor that modulates the cellular damage and influences the clinical outcome.

Lipopolysaccharide down-regulates Sp1 binding activity by promoting Sp1 protein dephosphorylation and degradation

We examined the in vivo effect of lipopolysaccharide (LPS) on Sp1 (promoter-selective transcription factor 1) DNA binding activity and studied the mechanisms involved in mouse lungs. The Sp1 DNA complex displayed a major band composed of Sp1, Sp2, and Sp3 trimer and a minor band composed of Sp3 homodimer. Compared with control, nuclear proteins from lungs challenged with LPS for 60, 90, 120, 150, 180, and 240 min, respectively, showed a markedly reduced Sp1 binding activity. Down-regulation of Sp1 binding activity was accompanied by a reduced expression of two Sp1-dependent genes (endothelial nitric oxide synthase and cyclooxygenase-1). Immunoprecipitation-Western blot experiments demonstrated that LPS dephosphorylated Sp1 protein at serine and threonine residues but not at the tyrosine residue. Dephosphorylation of Sp1 protein in vitro significantly reduced Sp1 DNA binding activity. Deglycosylation of Sp1 protein also reduced Sp1 binding activity. However, LPS did not cause Sp1 deglycosylation. LPS markedly reduced nuclear Sp1 protein level but had no significant effect on Sp1 mRNA abundance and on Sp1 protein nuclear translocation. Both Sp1 protein dephosphorylation and Sp1 protein degradation are temporally correlated to the reduced Sp1 binding activity. Our results demonstrate that challenge of mice with LPS in vivo down-regulates Sp1 DNA binding activity through promoting Sp1 protein dephosphorylation and degradation.

Effect of endotoxin on fibronectin synthesis of rat primary cultured hepatocytes

AIM: To investigate the effect of endotoxin on liver fibrosis and further define the role of hepatocytes in production of fibronectin in primary livercell culture by endotoxin.

METHODS: After isolation and seeding of hepatocytes, the obtained cells were added to various doses (0, 5, 10, 15 and 20 mg/L) of LPS treated culture media. The cells were collected and counted at various periods (0, 12, 24, 48, 72, 96, 120 h). The concentrations of fibronectin were tested by electrophoresis.

RESULTS: The fibronectin levels tended to increase with prolongation of culture time. There was a sharp increase after 72 h in 10 or 15 LPS treated group. The peak level of fibronectin was above 20 mg/L. However, cell proliferation was inhibited during the course. Cell number of untreated control group (4.6 ± 0.1 × 10⁶) was about three fold that of 20 LPS treated group (1.6 ± 0. 2 × 10⁶) at 120 h.

CONCLUSION: Hepatocytes have a potent ability to produce fibronectin stimulated by endotoxin, suggesting that hepatocytes might participate in the process of liver fibrosis.

Endotoxin-induced ileal Vo2-Do2 alterations do not correlate with the severity of ileal injury

PURPOSE

Altered Vo2-Do2 relationships are most often noted to occur in the setting of sepsis or endotoxin (LPS)-induced systemic organ microvascular injury and are generally thought to be causally linked to that injury. However, we have recently shown that ileal microvascular injury is not associated with altered ileal Vo2-Do2, relationships. Thus, we hypothesized that the severity of LPS-induced systemic organ microvascular injury would not correlate with the development of systemic organ Vo2-Do2 alterations.

MATERIALS AND METHODS

To test this hypothesis, we used the in situ autoperfused feline ileal preparation to simultaneously examine microvascular permeability, reflected as the ileal lymph to plasma protein concentration ratio (CL/CP), and ileal Vo2-Do2 relationships 2 hours after intravenous LPS (0.75-2.0 mg/kg; n = 9) and in matching controls (n = 5).

RESULTS

As expected, all LPS-treated animals were found to have extensive ileal histological damage and marked increases in the CL/CP compared with controls (0.308 +/- 0.019 v 0.097 +/- 0.009; P < .001). In addition, although the critical Do2 (Do2c) was elevated in the LPS-treated animals relative to controls (34.2 +/- 5.0 v 16.7 +/- 1.4 mL/min/kg; P < .03), there was no correlation between the Do2c and the CL/CP in the LPS-treated animals. Finally, ileal wet to dry weight ratios after LPS did not differ from controls.

CONCLUSION

Taken together, these data suggest that factors other than organ injury, as assessed by morphological and permeability alterations, are important in the pathogenesis of altered systemic organ Vo2-Do2 relationships after LPS.

Early signaling events by endotoxin in PC12 cells: involvement of tyrosine kinase, constitutive nitric oxide synthase, cGMP-dependent protein kinase, and Ca2+ channels

We studied the effects of endotoxin from Escherichia coli (E. coli) on Ca2+ channel activity in PC12 cells using the cell-attached patch clamp technique. Endotoxin (1-100 ng/ml) decreased channel availability (n x Po) to about one third of control values, an effect that required 3.5 +/- 1 min (mean +/- SD; n = 13) to reach steady state. The biophysical properties of the channel, including slope conductance (22 pS; 40 mM Ba2+), voltage dependence of n x Po, and open times (tau 1 = 0.78 ms, tau 2 = 8.9 ms) for the two open states at 0 mV, were not altered. The effect of endotoxin was blocked by polymyxin-B, indicating involvement of the lipid-A moiety of lipopolysaccharide, and by the tyrosine kinase (tk) inhibitor, tyrphostin. The effect of endotoxin was mimicked by 8-bromo-cGMP (100 microM), and was blocked by the inhibitor of cGMP-dependent protein kinase (PKG), H-8, suggesting involvement of the cGMP/PKG pathway. The effect of endotoxin also was blocked by the nitric oxide (NO) synthase inhibitor, NG-monomethyl-L-arginine monoacetate, suggesting involvement of nitric oxide synthase (NOS). The rapidity of the effect of endotoxin on Ca2+ channel activity suggested that constitutive NOS (cNOS) was involved, in accordance with our finding that endotoxin-induced transcriptional induction of NOS, as measured by nitrite production, required > 6 hr. We conclude that early signaling events by endotoxin in PC12 cells involve tk, cNOS, cGMP/PKG, and Ca2+ channels.

Effects of dexamethasone and cyclosporin A on the accumulation of eosinophils in acute cutaneous inflammation in the guinea-pig

1. Eosinophils are thought to play an important role in the pathophysiology of allergic diseases and pharmacological suppression of their recruitment is considered to be of therapeutic benefit. In the present study we have assessed and compared the effects of treatment with dexamethasone and cyclosporin A on the accumulation of 111In-labelled eosinophils and local oedema formation in sites of acute inflammation in guinea-pig skin. 2. When injected locally 150 min prior to i.d. administration of antigen in a passive cutaneous anaphylactic (PCA) reaction, dexamethasone (10(-9) to 3 x 10(-7) mol per site) dose-dependently inhibited oedema formation by up to 50%. Similarly, oedema formation induced by PAF and lipopolysaccharide (LPS), but not by zymosan-activated plasma (ZAP), was significantly inhibited by dexamethasone. In contrast, 111In-eosinophil accumulation measured in response to i.d. injection of PAF, LPS and ZAP or in the PCA reaction was not altered. 3. Systemic treatment with dexamethasone (4 mg kg-1, i.v., 150 min pretreatment period) inhibited both oedema formation and 111In-eosinophil accumulation induced by PAF, ZAP, LPS and in the PCA reaction. 4. The effects of i.d. injection of cyclohexamide (2 x 10(-7) mol per site) on 111In-eosinophil accumulation and oedema formation induced by PAF, ZAP or in a PCA reaction were evaluated in order to assess the dependency of these responses on protein synthesis. Cycloheximide had no effect on the responses measured. In contrast, 111In-eosinophil accumulation, but oedema formation, induced by LPS was inhibited by 30%. 5. Acute (10 mg kg-1, i.v., 15 min pretreatment) or prolonged (10 mg kg-1, s.c. daily for 3 days) systemic treatment with cyclosporin A had no effect on 111In-eosinophil accumulation or oedema formation induced by PAF, ZAP, LPS or in the PCA reaction. 6. In conclusion, we demonstrate preferential inhibitory effects of dexamethasone on 111In-eosinophil accumulation according to its site of administration. In addition we show that dexamethasone inhibits protein synthesis-independent acute inflammation in guinea-pig skin. Finally, our results do not support the concept that eosinophils are an important cellular site of action for the inhibitory effects of cyclosporin A in a guinea-pig model of allergic inflammation.

Endotoxin priming exacerbates acute reflux pancreatitis in the rat

Recently, endotoxaemia has been reported as a prognostic marker in acute pancreatitis. However, the role of endotoxin in inducing or aggravating acute pancreatitis is not fully understood. We administered endotoxin 400 micrograms/kg i.p. to rats 24 h before performing either a closed duodenal loop (group B) or a sham operation (group D). Pancreatic damage and overall survival were compared with the results obtained in rats not exposed to endotoxin undergoing either closed duodenal loop (group A) or sham treatment (group C). In a first set of experiments, 24 h after laparotomy blood samples were collected and the animals were sacrificed; survival up to 8 days was estimated in a second set of experiments. Group B had higher lipase concentrations and more severe tissue damage than group C (P < 0.05). A larger number of abscesses was observed in both group B and group D as compared to group C (P < 0.05). Survival was significantly shorter in group B (P < 0.0001). We conclude that priming with endotoxin worsens the extent of pancreatic damage induced by the closed duodenal loop procedure in the rat, possibly favouring selective homing of neutrophils to the site of inflammation, in similarity to what happens in the Shwartzman phenomenon.

Local and systemic effects of endotoxin in contracting skeletal muscle

PURPOSE

This study was performed to determine the direct effect of endotoxin on force generation, O2 consumption and vascular resistance in contracting skeletal muscle.

MATERIALS AND METHODS

We vascularly isolated the gastrocnemius muscle of dogs anesthetized with sodium pentobarbital and mechanically ventilated. The muscle was perfused from a proximal vessel or by a pump that withdrew blood from the contralateral leg. The nerve to the muscle was stimulated with supramaximal voltage 12 tr/min, 15-Hz impulses and a duty cycle of 0.4. Blood flow was measured with an electromagnetic flow probe, and oxygen consumption (VO2) was calculated from the flow and arterial-venous O2 content. In Protocol 1 (local infusion), contractions were stimulated for 30 minutes and endotoxin (n = 6) or saline (n = 6) was infused into the vasculature of the isolated gastrocnemius muscle after 10 minutes of contraction and continued for another 20 minutes of contraction. In Protocol 2 (systemic infusion), The normal tension and flow to the gastrocnemius were established and endotoxin (n = 5) or saline (n = 5) infused systemically. One hour later, the flow was set at the control level of contracting muscle, and contractions were stimulated for 30 minutes.

RESULTS

In both groups, endotoxin did not alter the tension VO2, arterial venous oxygen difference, or vascular resistance at the end of the stimulation period.

CONCLUSION

Endotoxin must affect muscle force production by acting through intermediates such as cytokines, and the effect is not apparent in the first 60 minutes.

Splanchnic ischaemia and its role in multiple organ failure

Multiple organ failure remains the leading cause of death in the intensive care unit. Increasing numbers of investigators have focused their attention on the role of gastrointestinal tract in the pathogenesis of this syndrome. Their data indicate that inadequate gut perfusion leads to a measurable imbalance between oxygen delivery and the needs of the tissues, i.e., ischaemia. Gut ischaemia of sufficient duration impairs gastrointestinal tract barrier function, facilitating the passage of enteric bacterial endotoxin into the circulation. It has been hypothesized that production of tumor necrosis factor alpha, and other biologic mediators by endotoxin-stimulated macrophages, triggers a generalized and uncontrolled inflammatory response that ultimately leads to multiple organ failure. Preliminary evidence suggests that survival can be improved significantly if gut ischaemia is promptly identified and aggressively treated by administration of fluids and inotropic drugs, using gastric intramucosal pH as the therapeutic endpoint. Future studies are needed to determine whether additional treatment modalities can improve outcome once the inflammatory response has fully developed.