IL-6 protects enterocytes from hypoxia-induced apoptosis by induction of bcl-2 mRNA and reduction of fas mRNA.

Interleukin-6 (IL-6) has been shown to rescue enterocytes from hypoxia-induced apoptosis when given orally following hemorrhagic shock. In vitro models using an intestinal epithelial cell line (IEC-6) cultured with lipopolysaccharide (LPS) under low O2 conditions, to mimic intestinal conditions, show that these cells also undergo apoptosis, which can be reduced by subsequent culture with IL-6. To examine further the mechanisms of rescue, we cultured normal rat intestinal epithelial cells (IEC-6) under both normoxic and hypoxic conditions and analyzed their responses to LPS and IL-6. We showed that IEC-6 expressed IL-6 receptor on its surface. Further, IEC-6 cells could be rescued from hypoxia-induced apoptosis by co-culture with IL-6. RNase protection assay (RPA) examination revealed that under hypoxic conditions, IEC-6 cells that were resistant to apoptosis showed reduced fas expression and increased bcl-2 expression after co-culture with LPS+IL-6.

Orally administered IL-6 induces elevated intestinal GM-CSF gene expression and splenic CFU-GM.

Orally administered interleukin (IL)-6 has been shown to be of benefit in eliminating Campylobacter infection and in preventing sepsis following hemorrhage. In related experiments, it was seen that proliferating cells were found in the spleens of untreated mice given IL-6 by oral gavage. Injection of the DNA label, BrdU, showed that significant proliferation began at 4 h and peaked at 24 h in the splenic red pulp of animals given oral IL-6. Mice given saline showed no increase in splenic BrdU uptake. Histological analysis suggested a hematopoietic lineage for these cells. Clonogenic assays performed on spleen cells taken from mice given oral IL-6 revealed that increased granulocyte-macrophage colony forming units (GM-CFU) were present at 24 h post-IL-6 administration. No increase in GM colonies occurred in mice fed IL-3, granulocyte-colony stimulating factor (G-CSF) or granulocyte-macrophage (GM)-CSF. RT-PCR analysis of intestinal mRNA from treated mice revealed that GM-CSF mRNA was elevated at 4 h after oral IL-6 administration, but not in mice fed other cytokines. It is suggested that oral administration of IL-6 induces both proliferation and a brief elevation of GM-CFU in the hematopoietic spleens of mice. This increase appears to be the result of increased GM-CSF mRNA production in the intestines of mice fed IL-6.

Disulfiram inhibits TNF-alpha-induced cell death.

Disulfiram, a clinically employed alcohol deterrent, was recently discovered to inhibit caspase-3 and DNA fragmentation. Using LLC-PK1 cells and murine liver as models, we examined if the drug inhibited TNF-alpha-induced cell death. Disulfiram produced dose-dependent inhibition of TNF-alpha-induced cell death as well as caspase-3-like activity. Disulfiram retained 80% of its effect when added 4 h after TNF-alpha. Disulfiram protected the cells from cytokine-induced death for at least 6 days. The cells rescued by the drug preserved the ability to proliferate. The cells died spontaneously after exposure to TNF-alpha for just 70 min. Co-administration of 15 microM disulfiram and TNF-alpha for 70 min prior to their removal abolished TNF-alpha-induced killing, and this was associated with restoration of mitochondrial membrane potential and suppression of reactive oxygen species. Treatment of mice with TNF-alpha and D-galactosamine for 5 h markedly increased hepatic DNA fragmentation and caspase-3-like activity. Disulfiram at 0.6 mmol/kg abolished these effects. We conclude that disulfiram is a potent inhibitor of TNF-alpha-induced cell death in vitro. The underlying mechanisms include stabilization of mitochondrial membrane potential, suppression of reactive oxygen species, and inhibition of caspase-3-like activity. We further conclude that disulfiram inhibits DNA fragmentation in vivo in association with the blockade of caspase-3-like activity.

Microvascular effects of oral interleukin-6 on ischemia/reperfusion in the murine small intestine.

Oral administration of interleukin-6 (IL-6) has been shown to reduce hemorrhage-induced bacterial translocation from the gut in mice and rats. To examine the intestinal microvasculature, mice were given the electron-dense tracer horseradish peroxidase (HRP) after hemorrhage and IL-6 or vehicle administration. In normal mice and in those hemorrhaged and given IL-6, the electron-dense marker, administered intravenously, could be found in intestinal capillaries and between mucosal epithelial cells, suggesting that the microvasculature was patent. In mice given saline after shock, however, no marker was present in the gut, suggesting that the intestinal microvasculature was unable to deliver the marker to the epithelia. When mice were given HRP intralumenally (il) the tracer was able to penetrate between intestinal epithelial cells only in mice given vehicle after hemorrhage. This finding suggests that hemorrhaged mice were susceptible to sepsis and endotoxic shock from the leaky gut. In normal and IL-6-treated mice, the tracer was unable to pass from the lumen between mucosal epithelial cells, because the presence of an intact zonula occludens prevented passage. Functional studies supported the electron microscopy findings. Bacteria were cultured from the livers of mice fed vehicle after hemorrhage, but not from those fed IL-6. These data support the conclusions that parts of the intestinal microvasculature remain diminished after hemorrhage and resuscitation and that oral IL-6 restores this circulation.

Differentially increased IL-6 mRNA expression in liver and spleen following injection of liposome-encapsulated haemoglobin.

Injection of the red cell substitute liposome-encapsulated haemoglobin (LEH) induces increased serum interleukin (IL)-6 in the absence of other inflammatory cytokines. In vitro studies found that IL-6 mRNA was increased in Mphi and endothelial cell lines by co-culture with LEH. In the present study, cytokine mRNA expression in extracts of livers, spleens, lungs and kidneys after LEH injection was determined by semi-quantitative RT-PCR. The distribution of cells expressing IL-6 mRNA in livers and spleens was visualized by in situ hydridization; extracts of kidney and lung did not show increased IL-6 mRNA and were not studied further. IL-6 mRNA accumulation in livers and spleens was increased at 4 h following LEH injection and had declined by 24 h. In the liver, cells expressing IL-6 mRNA were located in endothelia of hepatic and portal veins, and hepatic sinuses, Kupffer cells and epithelial cells of bile ducts. Endothelium of hepatic arteries did not express IL-6 mRNA. Lymphocytes, haematopoietic cells and macrophages expressed IL-6 mRNA in spleens. The data suggest that cells of the reticuloendothelial system (RES) might be a significant source of increased plasma IL-6 in vivo after LEH administration.

IL-6 rescues enterocytes from hemorrhage induced apoptosis in vivo and in vitro by a bcl-2 mediated mechanism.

Following a hemorrhagic event, damage to the highly metabolic intestinal tissue induces loss of barrier function leading to bacterial escape and LPS contamination of the host. Orally administered IL-6 restores intestinal barrier function following hemorrhage in both rat and mouse models. IL-6 prevents apoptosis in a variety of lymphoid cells and lines, through the activation of the proto-oncogene bcl-2. This communication elucidates the role of the IL-6-bcl-2 interaction in intestinal apoptosis following hemorrhagic shock. Terminal deoxynucleotidyl-transferase-mediated dUTP nick end labeling (TUNEL) and p53 immunohistochemical staining were used to examine intestines from mice hemorrhaged and fed saline or IL-6 and enterocytes (IEC-6) exposed to hypoxia and LPS alone or LPS and IL-6 in vitro. In situ hybridization for bcl-2 expression was performed on intestines or enterocytes. Intestinal sections from mice hemorrhaged and fed IL-6 showed reduction in apoptosis and increases in bcl-2 gene expression relative to sections taken from mice hemorrhaged and fed saline. IEC-6 cells exposed to hypoxia and LPS had high numbers of TUNEL staining cells. Subsequent exposure to IL-6 after hypoxia and LPS reduced apoptotic cell numbers and increased bcl-2 gene expression. The data show that exposure of intestinal epithelial cells to IL-6 either by oral administration in hemorrhaged mice or by coculture following hypoxia and LPS treatment results in increased bcl-2 gene expression and reduced damage from apoptosis.

Systemic sepsis following hemorrhagic shock: alleviation with oral interleukin-6.

Gut-origin sepsis is a serious medical complication of military injuries following hemorrhage. Splanchnic ischemia induces intestinal necrosis leading to systemic bacteremia. Rat and mouse models of hemorrhagic shock were used to investigate bacterial translocation from the gut. Orally administered ameliorative treatments using the cytokine interleukin-6 (IL-6) were able to reduce or eliminate sepsis following hemorrhage. To mimic battlefield wounds and hemorrhage, anesthetized mice were bled from the femoral artery, held at a mean arterial blood pressure of 35 mm Hg for 1 hour, and then resuscitated with shed blood and 2-fold volume lactated Ringer's solution. Anesthetized rats were bled from the carotid artery at a rate of 15 ml/kg at 1 ml/minute. Bacteriological cultures of livers and mesenteric lymph nodes from hemorrhaged animals given recombinant IL-6 had significantly fewer colonies per gram of tissue than saline-fed controls. 125I-labeled IL-6 remained in the gut for up to 6 hours giving regional protection, whereas labeled interleukin-2 was disseminated throughout the body in the same time. In vivo and vitro studies of IL-6 showed that long incubations with high doses of trypsin, chymotrypsin, or intestinal contents were necessary to inactivate the bioactivity of this cytokine. Electron microscopy showed that epithelial cells from hemorrhaged mice fed saline had sparse or missing villi and vacuolated cytoplasm. Epithelial cells from control mice or mice hemorrhaged and fed cytokine appeared completely normal. Oral administration of IL-6 on the battlefield may be an important treatment for the prevention of sepsis following hemorrhage.

Cellular responses to surgical trauma, hemorrhage, and resuscitation with diaspirin cross-linked hemoglobin in rats.

BACKGROUND: Resuscitation with acellular oxygen carrier solutions offers the potential advantage of improved oxygen delivery compared with crystalloid solutions, but the detailed consequences of improved resuscitation have not been fully evaluated. This study evaluated local and systemic cellular effects of trauma, hemorrhage, and resuscitation in a model of hemorrhage and surgical trauma. METHODS: Rats with a 10 cm full-thickness incisional wound and a 15 mL/kg hemorrhage were either not resuscitated or resuscitated with blood or diaspirin cross-linked hemoglobin (DCLHb). Cellular proliferative responses were evaluated at 1.5, 6, 24, and 48 hours after wounding by labeling in vivo with 5-bromo-2'-deoxyuridine. Plasma levels of interleukin-6, tumor necrosis factor-alpha, and interferon-gamma were measured by bioassay or enzyme-linked immunosorbent assay (ELISA). Bacterial translocation was measured by culturing liver homogenates. RESULTS: Trauma inhibited keratinocyte and hepatocyte proliferation at 1.5 and 6 hours, and stimulated subsequent proliferation of keratinocytes and liver nonparenchymal cells. DCLHb stimulated wound keratinocyte proliferation, attenuated the inhibition of hepatocyte proliferation, eliminated bacterial translocation to the liver, protected the intestine from ischemic damage, and induced a rapid increase of interleukin-6 during the early phase of injury. CONCLUSIONS: Surgical trauma alone, or in combination with hemorrhage, modulated cell proliferation both in the wound and in the remote organs of intestine and liver. DCLHb enhanced wound healing and cell proliferation as well as, or better than, freshly drawn blood, which may be beneficial for trauma care.

Kinetics of cytokine gene expression in macrophage and endothelial cell lines following liposome encapsulated haemoglobin (LEH) treatment in vitro.

Liposome encapsulated haemoglobin (LEH), a blood substitute, has been shown to increase serum IL-6 levels in mice, but the cellular source of this cytokine is unknown. These experiments were conducted to determine cytokine gene expression by macrophages (P388D1) and endothelial cells (EOMA) after LEH treatment in vitro. P388D1 and EOMA cells were incubated with LEH, final concentration of 10%, for 0, 0.5, 1, 2, 4, 8 and 24 h. Total RNA was extracted at the above times and semi-quantitative RT-PCR was carried out to detect cytokine mRNA expression in the presence of competitive internal standards specific for mouse IL-1 beta, IL-6, TGF-beta, TNF-alpha and GM-CSF. The results demonstrated that although constitutive cytokine gene expression exists in both cell lines, the level of IL-6 mRNA was prominently elevated by incubation with LEH, rapidly reaching a peak at about 4 h and gradually declining over the next 20 h after LEH treatment. In contrast, there was no significant change in mRNA accumulation of IL-1 beta, TGF-beta, TNF-alpha or GM-CSF at all times. These in vitro findings suggest that macrophages and endothelial cells may be the cellular sources of the elevated levels of IL-6 found in serum after in vivo LEH administration.

Systemic bacteraemia following haemorrhagic shock in mice: alleviation with oral Interleukin 6.

A murine model of haemorrhagic shock was used to investigate bacterial translocation from the gut and subsequent systemic immunoreduction. Anaesthetized mice were bled from the femoral artery, and held at a mean arterial blood pressure of 35 mm Hg for one hour then resuscitated with shed blood and two-fold volume lactated Ringer's solution. Upon awakening, they were given cytokines or control media orally. Bacteriological cultures of livers, spleens and mesenteric lymph nodes from haemorrhaged mice given cytokine had significantly fewer bacteria/gm of tissue than those given media. Recombinant IL-6 mimicked the effects seen with crude cytokines. Reduction of proliferation among spleen cells from haemorrhaged mice was observed and could be partially returned to normal by cytokine feeding. Mixing experiments in which cells from haemorrhaged mice were added to those of normal mice in an MLR showed no suppressor activity. Flow cytometry analysis revealed a reduction in CD 3+ cells at 16 hours post-haemorrhage in mice fed control media or cytokines, suggesting that reduced proliferative capacity may be due to loss of function rather than active suppression. Histological examination of the intestines of haemorrhaged mice fed cytokines or media revealed restoration of intestinal mucosal integrity by cytokine administration. These results suggest that oral administration of IL-6 may be an important treatment for the prevention of systemic sepsis following haemorrhage.

Multiple responses to administration of liposome-encapsulated hemoglobin (LEH): Effects on hematopoiesis and serum IL-6 levels.

Liposome-encapsulated hemoglobin (LEH) has been tested in animals as an oxygen-carrying red cell substitute and has been shown to be beneficial in the treatment of hemorrhagic shock. The effects of LEH on immune responses have not been studied thoroughly in any well-controlled model. Using a murine model, we evaluated nephrotoxicity and hepatotoxicity as well as immune function parameters following LEH administration. Following intravenous administration of LEH, 1) a serum spike of interleukin-6 (IL-6) occurred in mice at 4-8 hours, with no elevation of IL-1, tumor necrosis factor (TNF), or interferon-gamma (IFN-gamma); 2) the serum liver function enzymes SGOT (AST, aspartate aminotransferase) and SGPT (ALT, alanine aminotransferase) were elevated at 48 hours; 3) only a slight increase in serum antibody to bovine hemoglobin was observed; and 4) increased hematopoietic activity was observed in the spleen and bone marrow. The finding that only IL-6 but not the associated TNF, IL-1, or IFN-gamma is secreted in vivo following LEH administration is novel and may have significance in defining the mechanisms underlying specific adverse responses observed with LEH administration in animals.

Controlled release of transforming growth factor-beta from lipid-based microcylinders.

The release of transforming growth factor-beta (TGF-beta) from a lipid microstructure has been demonstrated. Lipid microcylinders, with dimensions of 100 x 0.5 microns and composed of a diacetylenic lipid, have been loaded with 25 ng TGF-beta/mg lipid. Physical and bioactive release characteristics of TGF-beta from these microcylinders and from microcylinders embedded in an agarose hydrogel are reported. Release of TGF-beta from lipid microcylinders follows typical diffusion-limited characteristics, where 10-12% of the TGF is released in the first 10 h at 37 degrees C. The release rate is shown to be temperature controlled and dependent on the integrity of the lipid microcylinder. Immobilization of the lipid microcylinder in a hydrogel matrix composed of agarose and gelatin does not impair the diffusion of TGF-beta from the lipid microcylinders. The utilization of microcylinders as release vehicles in wound repair is discussed.

Recruitment of tissue resident cells to hydrogel composites: in vivo response to implant materials.

A model of local cellular recruitment was established using hydrogel matrices composed of alginate implanted subcutaneously into mice. Cells which trafficked to the matrix blocks were recovered and characterized for surface phenotype using fluorescently labelled antibodies and flow cytometry (fluorescence activated cell sorting). Temporal information of the differential recruitment of cells was determined. The basic pattern of recruitment in response to the hydrogels was established and mimicked that seen in a local inflammatory response. Neutrophils (PMN) were rapidly recruited (1 d) followed by macrophages and lymphocytes (1-3 d). Cell surface phenotype studies included the determination of CD3+, CD4+ and CD8+ cells, Mac-1+ cells, and immunoglobulin bearing cells. Microscopic analysis revealed numerous activated PMNs and monocyte derived foamy macrophages. Fluorescence immunocytochemistry of frozen sections of the block revealed that macrophages, CD3+ and natural killer cells were all recruited to the interior of the block. Ultrastructural analysis (transmission electron microscopy) showed highly activated macrophages, with abundant rough endoplasmic reticulum and secretory vesicles. Cells which remained on the surface of the matrix block were CD44 positive migratory cells. Electron microscopic evidence showed foamy macrophages with a varying degree of involvement with the hydrogel material. Surface scanning electron microscopy revealed numerous fibroblast-like cells coating the surface of the block. We suggest that these methods may be used to address the inflammatory response elicited with a a variety of implanted materials such as hydrogels, silicones, ceramics and metals. Furthermore, this model has been useful in determining cellular responses to cytokines and growth factors under similar conditions.

An improved model for the examination of biological effects of locally administered cytokines.

Cytokines incorporated into agarose blocks and implanted subcutaneously into mice establish an in vivo gradient which can be used to mimic a local inflammatory process. We have developed a model in which cellular influx into cytokine impregnated blocks parallels the normal cellular reaction to infections or wounds. Agarose blocks containing supernatants from ConA activated rat spleen cells attracted neutrophils within 4 h. These cells were followed by lymphocytes and macrophages in 24 h. Flow cytometry analysis of lymphoid cells on day 1 revealed that 38% were Ig+ (B cell marker), 60% MAC-2,3+ and 20% Thy 1.2+ of which only a small fraction were expressing CD4 on their surface. These numbers changed with time following implantation of the blocks. Initially, isolated adherent cells (macrophages) were resting, with low phagocytic activity. Cells isolated from blocks at later time points were activated, as evidenced by their increased ability to ingest fluoresceinated beads. The secretion patterns of cells trafficking to murine rIL-1 containing agarose blocks were examined. TNF, IL-6 and antibody secreting cells were found. No IL-2 was detected at any time. We believe that this model will be of value in studies of local actions of cytokines.

Modulation of mucosal immunity against Campylobacter jejuni by orally administered cytokines.

The effect of oral recombinant interleukin (rIL) treatment on the course of Campylobacter jejuni infection and the development of mucosal immunity in mice was investigated. rIL-2, rIL-5, and rIL-6 were administered to mice at 24 and 6 h before infection and at 0, 24, and 48 h after infection with C. jejuni HC, and the subsequent development of an immune response and intestinal colonization resistance were determined. In this model, orally administered cytokines retained their biological activities with no apparent side effects. Following infection, initial bacterial counts in fecal samples collected from cytokine-treated and untreated mice were similar. However, within 48 h of infection a greater than 3-log-unit reduction in the number of C. jejuni shed in the feces was found for rIL-6-treated animals. Colonization levels were similarly reduced in rIL-5-treated mice, although the rate of clearance was somewhat slower. In contrast, rIL-2 treatment had no significant effect on colonization levels compared with that in controls. Oral rIL-6 treatment was also associated with enhanced intestinal and systemic Campylobacter-specific immunoglobulin A responses compared with those observed in either rIL-5- or rIL-2-treated animals. Upon rechallenge, initial colonization in all cytokine-treated groups was approximately 2 log units lower than that in controls. However, local infection was controlled only in rIL-2-treated mice over time. rIL-5 and rIL-6 treatment had only a marginal effect on colonization resistance following rechallenge. On the basis of these results, it appears that rIL-5 or rIL-6 may function to modulate the induction and/or expression of anti-C. jejuni immunity through different mechanisms.

Killed Campylobacter elicits immune response and protection when administered with an oral adjuvant.

The heat-labile toxin (HLT) of enterotoxigenic Escherichia coli (ETEC) is a potent oral adjuvant. We determined whether the ETEC HLT could be mixed with killed campylobacter to induce an immune response protective upon subsequent challenge with live pathogens. Mice were immunized orally three times with 10(9) sonicated campylobacter with or without 25 micrograms of ETEC HLT, and humoral immune responses in intestinal lavage fluids measured by ELISA. Whereas 10(9) live bacteria induced strong intestinal IgA responses, killed bacteria did not unless ETEC HLT was also added. The magnitude of the antibody response was dependent on the amount of antigen given. The ETEC HLT given with bacteria also induced a potent cross-reaction with cholera toxin. The latter had an adjuvant effect in mice similar to that of ETEC HLT. Protection against colonization was studied in mice and rabbits. In contrast to non-immune animals, those given live organisms or sonicated cells mixed with ETEC HLT quickly cleared homologous, but not heterologous, Lior serotypes of Campylobacter upon challenge. These data show for the first time that ETEC HLT can potentiate an immune response to killed campylobacter that promotes a rapid clearance of live pathogens from the intestine.

Natural killer cell cytotoxicity and T-cell proliferation is enhanced by avoidance behavior.

Natural killer (NK) cell activity and mitogen-stimulated spleenocyte proliferation were measured in rats exposed to stress in the form of avoidable and unavoidable shock. Rats that could avoid shock exhibited higher NK activity than either unshocked controls or rats that could not avoid shock. The latter were yoked to the avoidance rats and thus received the same number and frequency of shocks as did the avoidance group. The increased NK activity in the avoidance group appears due to a higher number of NK cells in this group as compared with those in the control or unavoidable shock groups. Additionally, NK activity was found to be proportional to avoidance response rate, with a majority of animals exceeding the minimal temporal avoidance requirement. Mitogen-stimulated proliferation of spleenocytes was also increased several fold in the group that could avoid shock as compared with that which could not and controls. The difference in NK activity and mitogen-stimulated proliferation could not be ascribed to differences in cortisol levels. The results indicate that behavior which results in the avoidance of aversive stimuli can lead to significant enhancement of immune system competence.

Identification of Campylobacter jejuni and Campylobacter coli antigens with mucosal and systemic antibodies.

The development of a rapid and specific diagnostic assay for Campylobacter infections is important in determining the etiology of acute diarrhea in humans. Studies have shown that sonicated whole bacteria or partially purified antigens cross-reacted with antibodies against other closely related bacteria. To solve the problems of specificity, we identified specific antigens of Campylobacter jejuni and Campylobacter coli for use in diagnostic assays. We investigated the responses of serum, urine, and intestinal lavage antibodies in infected (fed live bacteria) and parenterally immunized (intraperitoneal injection of sonicated whole bacteria with adjuvant) mice directed against C. jejuni or C. coli by Western blot (immunoblot) analysis. Antibody responses were examined weekly for up to 28 days. Fewer antigens were detected by urinary and intestinal lavage fluid immunoglobulin A (IgA) than serum IgG and IgM for both parenterally immunized and infected mice. Serum from parenterally immunized mice detected more antigens than that from infected mice. Two high-molecular-weight antigens (62,000 and 43,000) were predominantly detected by serum, urine, and intestinal lavage fluids of both parenterally immunized and infected mice. Serum antibodies from 28-day parenterally immunized mice detected one antigen specific to C. coli with a molecular weight of 38,000 and one antigen specific to C. jejuni with a molecular weight of 27,000. An immunodominant protein with a molecular weight of 31,000 common to both C. jejuni and C. coli was also recognized by serum antibodies from parenterally immunized mice.