Thrombin stimulates albumin transcytosis in lung microvascular endothelial cells via activation of acid sphingomyelinase.

Transcellular albumin transport occurs via caveolae that are abundant in lung microvascular endothelial cells. Stimulation of albumin transcytosis by proinflammatory mediators may contribute to alveolar protein leak in lung injury, yet the regulation of albumin transport and its underlying molecular mechanisms are so far incompletely understood. Here we tested the hypothesis that thrombin may stimulate transcellular albumin transport across lung microvascular endothelial cells in an acid-sphingomyelinase dependent manner. Thrombin increased the transport of fluorescently labeled albumin across confluent human lung microvascular endothelial cell (HMVEC-L) monolayers to an extent that markedly exceeds the rate of passive diffusion. Thrombin activated acid sphingomyelinase (ASM) and increased ceramide production in HMVEC-L, but not in bovine pulmonary artery cells, which showed little albumin transport in response to thrombin. Thrombin increased total caveolin-1 (cav-1) content in both whole cell lysates and lipid rafts from HMVEC-L, and this effect was blocked by inhibition of ASM or de novo protein biosynthesis. Thrombin-induced uptake of albumin into lung microvascular endothelial cells was confirmed in isolated-perfused lungs by real-time fluorescence imaging and electron microscopy of gold-labeled albumin. Inhibition of ASM attenuated thrombin-induced albumin transport both in confluent HMVEC-L and in intact lungs, whereas HMVEC-L treatment with exogenous ASM increased albumin transport and enriched lipid rafts in cav-1. Our findings indicate that thrombin stimulates transcellular albumin transport in an acid sphingomyelinase-dependent manner by inducing de novo synthesis of cav-1 and its recruitment to membrane lipid rafts.

Ex vivo perfusion of the isolated rat small intestine as a novel model of Salmonella enteritis.

Using an ex vivo perfused rat small intestinal model, we examined pathological changes to the tissue, inflammation induction, as well as dynamic changes to smooth muscle activity, metabolic competence, and luminal fluid accumulation during short-term infection with the enteropathogenic bacteria Salmonella enterica serovar Typhimurium and Yersinia enterocolitica. Although few effects were seen upon Yersinia infection, this system accurately modeled key aspects associated with Salmonella enteritis. Our results confirmed the importance of the Salmonella Pathogenicity Island 1 (SPI1)-encoded type 3 secretion system (T3SS) in pathology, tissue invasion, inflammation induction, and fluid secretion. Novel physiological consequences of Salmonella infection of the small intestine were also identified, namely, SPI-1-dependent vasoconstriction and SPI-1-independent reduction in the digestive and absorptive functions of the epithelium. Importantly, this is the first small animal model that allows for the study of Salmonella-induced fluid secretion. Another major advantage of this model is that one can specifically determine the contribution of resident cell populations. Accordingly, we can conclude that recruited cell populations were not involved in the pathological damage, inflammation induction, fluid accumulation, nutrient absorption deficiency, and vasoconstriction observed. Although fluid loss induced by Salmonella infection is hypothesized to be due to damage caused by recruited neutrophils, our data suggest that bacterial invasion and inflammation induction in resident cell populations are sufficient for fluid loss into the lumen. In summary, this model is a novel and useful tool that allows for detailed examination of the early physiopathological effects of Salmonella infection on the small intestine.

Correction: Hydroxyethyl Starch (HES 130/0.4) Impairs Intestinal Barrier Integrity and Metabolic Function: Findings from a Mouse Model of the Isolated Perfused Small Intestine.

BACKGROUND: The application of hydroxyethyl starch (HES) for volume resuscitation is controversially discussed and clinical studies have suggested adverse effects of HES substitution, leading to increased patient mortality. Although, the intestine is of high clinical relevance and plays a crucial role in sepsis and inflammation, information about the effects of HES on intestinal function and barrier integrity is very scarce. We therefore evaluated the effects of clinically relevant concentrations of HES on intestinal function and barrier integrity employing an isolated perfused model of the mouse small intestine. METHODS: An isolated perfused model of the mouse small intestine was established and intestines were vascularly perfused with a modified Krebs-Henseleit buffer containing 3% Albumin (N=7) or 3% HES (130/0.4; N=7). Intestinal metabolic function (galactose uptake, lactate-topyruvate ratio), edema formation (wet-to-dry weight ratio), morphology (histological and electron microscopical analysis), fluid shifts within the vascular, lymphatic and luminal compartments, as well as endothelial and epithelial barrier permeability (FITC-dextran translocation) were evaluated in both groups. RESULTS: Compared to the Albumin group, HES perfusion did not significantly change the wet-to-dry weight ratio and lactate-to-pyruvate ratio. However, perfusing the small intestine with 3% HES resulted in a significant loss of vascular fluid (p<0.01), an increased fluid accumulation in the intestinal lumen (p<0.001), an enhanced translocation of FITC-dextran from the vascular to the luminal compartment (p<0.001) and a significantly impaired intestinal galactose uptake (p<0.001). Morphologically, these findings were associated with an aggregation of intracellular vacuoles within the intestinal epithelial cells and enlarged intercellular spaces. CONCLUSION: A vascular perfusion with 3% HES impairs the endothelial and epithelial barrier integrity as well as metabolic function of the small intestine.

Hydroxyethyl starch (HES 130/0.4) impairs intestinal barrier integrity and metabolic function: findings from a mouse model of the isolated perfused small intestine.

BACKGROUND: The application of hydroxyethyl starch (HES) for volume resuscitation is controversially discussed and clinical studies have suggested adverse effects of HES substitution, leading to increased patient mortality. Although, the intestine is of high clinical relevance and plays a crucial role in sepsis and inflammation, information about the effects of HES on intestinal function and barrier integrity is very scarce. We therefore evaluated the effects of clinically relevant concentrations of HES on intestinal function and barrier integrity employing an isolated perfused model of the mouse small intestine. METHODS: An isolated perfused model of the mouse small intestine was established and intestines were vascularly perfused with a modified Krebs-Henseleit buffer containing 3% Albumin (N=7) or 3% HES (130/0.4; N=7). Intestinal metabolic function (galactose uptake, lactate-to-pyruvate ratio), edema formation (wet-to-dry weight ratio), morphology (histological and electron microscopical analysis), fluid shifts within the vascular, lymphatic and luminal compartments, as well as endothelial and epithelial barrier permeability (FITC-dextran translocation) were evaluated in both groups. RESULTS: Compared to the Albumin group, HES perfusion did not significantly change the wet-to-dry weight ratio and lactate-to-pyruvate ratio. However, perfusing the small intestine with 3% HES resulted in a significant loss of vascular fluid (p<0.01), an increased fluid accumulation in the intestinal lumen (p<0.001), an enhanced translocation of FITC-dextran from the vascular to the luminal compartment (p<0.001) and a significantly impaired intestinal galactose uptake (p<0.001). Morphologically, these findings were associated with an aggregation of intracellular vacuoles within the intestinal epithelial cells and enlarged intercellular spaces. CONCLUSION: A vascular perfusion with 3% HES impairs the endothelial and epithelial barrier integrity as well as metabolic function of the small intestine.

Quinidine, but not eicosanoid antagonists or dexamethasone, protect the gut from platelet activating factor-induced vasoconstriction, edema and paralysis.

Intestinal circulatory disturbances, atony, edema and swelling are of great clinical relevance, but the related mechanisms and possible therapeutic options are poorly characterized, in part because of the difficulties to comprehensively analyze these conditions. To overcome these limitations we have developed a model of the isolated perfused rat small intestine where all of these symptoms can be studied simultaneously. Here we used this model to study the role of eicosanoids, steroids and quinidine in platelet-activating factor (PAF)-induced intestinal disorders. A vascular bolus of PAF (0.5 nmol) triggered release of thromboxane and peptidoleukotrienes into the vascular bed (peak concentration 35 nM and 0.8 nM) and reproduced all symptoms of intestinal failure: mesenteric vasoconstriction, translocation of fluid and macromolecules from the vasculature to the lumen and lymphatics, intestinal edema formation, loss of intestinal peristalsis and decreased galactose uptake. All effects of PAF were abolished by the PAF-receptor antagonist ABT491 (2.5 µM). The COX and LOX inhibitors ASA and AA861 (500 µM, 10 µM) did not exhibit barrier-protective effects and the eicosanoid antagonists SQ29548 and MK571 (10 µM, each) only moderately attenuated the loss of vascular fluid, the redistribution to the lumen and the transfer of FITC dextran to the lumen. The steroid dexamethasone (10 µM) showed no barrier-protective properties and failed to prevent edema formation. Quinidine (100 µM) inhibited the increase in arterial pressure, stabilized all the intestinal barriers, and reduced lymph production and the transfer of FITC dextran to the lymph. While quinidine by itself reduced peristalsis, it also obviated paralysis, preserved intestinal functions and prevented edema formation. We conclude that quinidine exerts multiple protective effects against vasoconstriction, edema formation and paralysis in the intestine. The therapeutic use of quinidine for intestinal ailments deserves further study.

Ingestion of (n-3) fatty acids augments basal and platelet activating factor-induced permeability to dextran in the rat mesenteric vascular bed.

Loss of intestinal barrier function and subsequent edema formation remains a serious clinical problem leading to hypoperfusion, anastomotic leakage, bacterial translocation, and inflammatory mediator liberation. The inflammatory mediator platelet activating factor (PAF) promotes eicosanoid-mediated edema formation and vasoconstriction. Fish oil-derived (n-3) fatty acids (FA) favor the production of less injurious eicosanoids but may also increase intestinal paracellular permeability. We hypothesized that dietary (n-3) FA would ameliorate PAF-induced vasoconstriction and enhance vascular leakage of dextran tracers. Rats were fed either an (n-3) FA-rich diet (EPA-rich diet; 4.0 g/kg EPA, 2.8 g/kg DHA) or a control diet (CON diet; 0.0 g/kg EPA and DHA) for 3 wk. Subsequently, isolated and perfused small intestines were stimulated with PAF and arterial pressure and the translocation of fluid and macromolecules from the vasculature to lumen and lymphatics were analyzed. In intestines of rats fed the EPA-rich diet, intestinal phospholipids contained up to 470% more EPA and DHA at the expense of arachidonic acid (AA). The PAF-induced increase in arterial pressure was not affected by the EPA-rich diet. However, PAF-induced fluid loss from the vascular perfusate was higher in intestines of rats fed the EPA-rich diet. This was accompanied by a greater basal loss of dextran from the vascular perfusate and a higher PAF-induced transfer of dextran from the vasculature to the lumen (P = 0.058) and lymphatics. Our data suggest that augmented intestinal barrier permeability to fluid and macromolecules is a possible side effect of (n-3) FA-rich diet supplementation.

A model of the isolated perfused rat small intestine.

Intestinal edema remains a serious clinical problem, and novel approaches to study its pathophysiology are needed. It was our aim to develop a long-term stable isolated perfused rat small bowel preparation permitting analysis of vascular, luminal, interstitial, and lymphatic compartments and to demonstrate the utility of this model by studying the effects of the proinflammatory mediator platelet-activating factor (PAF). A temperature-controlled chamber with an integrated balance was designed to perfuse isolated intestines through the mesenteric artery and the gut lumen. Steroids or oxygen carriers were not needed. Functional and morphological integrity of the tissue was preserved for several hours as confirmed by oxygen consumption, venous lactate-to-pyruvate ratio, arterial and venous pH, lactose digestion and galactose uptake, intravascular and luminal pressures, maintained fluid homeostasis, gut motility, and quantitative light microscopic analysis. Administration of PAF caused typical effects such as vasoconstriction, gut atony, and loss of galactose uptake. PAF also elicited a transient loss of 20% of the perfusate liquid from the mesenteric vascular bed, two-thirds of which were transferred to the lumen. All these responses were entirely reversible. This new model provides detailed insights into the physiology of the small intestine and will allow to study fundamental processes such as fluid homeostasis, barrier functions, transport mechanisms, and immune responses in this organ. Using this model, here we show a dramatic and yet reversible response of the rat small bowel to PAF, suggesting luminal water clearance as a novel safety factor in the intestine that may be of clinical relevance.

Conserved responses to trichostatin A in rodent lungs exposed to endotoxin or stretch.

Histone deacetylase (HDAC) isoenzymes have been suggested as possible drug targets in pulmonary cancer and in inflammatory lung diseases such as asthma and COPD. Whether HDAC inhibition is pro- or anti-inflammatory is under debate. To further examine this clinically relevant paradigm, we analyzed 8 genes that are upregulated by two pro-inflammatory stimuli, i.e. endotoxin and mechanical stress (overventilation), in isolated rat and mouse lungs, respectively. We studied the effect of the HDAC inhibitor trichostatin A (TSA) under control conditions, in response to endotoxin and overventilation, and on the effects of the steroid dexamethasone. TSA affected gene expression largely independent of the stimulus (endotoxin, overventilation) and the species (rat, mouse) leading to upregulation of some genes (Tnf, Cxcl2) and downregulation of others (Cxcl10, Timp1, Selp, Il6). At the protein level, TSA reduced the stimulated release of TNF, MIP-2alpha and IL-6, indicating that TSA may affect protein translation independent from gene transcription. In general, the anti-inflammatory effects of TSA on gene expression and protein release were additive to that of dexamethasone, suggesting that both drugs employ different mechanisms. We conclude that pro-inflammatory stimuli induce distinct sets of genes that are regulated by HDAC in a diverse, but consistent manner across two rodent species. The present findings together with previous in vivo studies suggest that the effect of HDAC inhibition in the intact lung is in part anti-inflammatory.

Pulmonary responses to overventilation in late multiple organ failure.

BACKGROUND: Patients with multiple organ failure (MOF) require mechanical ventilation for several days. The enormous significance of the ventilation strategy for the outcome of these patients is well appreciated. However, most studies have focused on the onset and the early phase of MOF. It was the aim of the current study to investigate the effect of ventilation in the course of MOF. METHODS: Using a model where mice develop MOF 7-14 days after intraperitoneal injection of zymosan, the authors analyzed lung functions, signaling pathways, and mediator release in response to protective ventilation (end-expiratory pressure -3 cm H2O; end-inspiratory pressure -10 cm H2O) and overventilation (-22.5 cm H2O) in isolated lungs ex vivo. RESULTS: On day 7, pulmonary compliance, pulmonary resistance, and tidal volume were normal, but vascular resistance was elevated compared with untreated animals. During ex vivo ventilation, these lungs showed enhanced nuclear factor-kappaB activation, Akt kinase phosphorylation, and cytokine release, and this was further aggravated by overventilation. After 14 days, zymosan-treated animals were characterized by pulmonary hypertension, reduced tidal volume, elevated pulmonary resistance, and increased mediator production. However, in these lungs, neither nuclear factor-kappaB activation nor cytokine production where enhanced by overventilation. CONCLUSIONS: The zymosan model is characterized by pulmonary inflammation, diminished lung functions, and chronic hypertension. Mechanical ventilation with high distending pressures further augmented cytokine production in this chronic model of MOF only if it significantly augmented tidal volume. The authors speculate that these findings may be explained on the basis of different degrees of lung stretch.

Molecular and functional changes of pulmonary surfactant in response to hyperoxia.

Surfactant comprises phosphatidylcholine (PC) together with anionic phospholipids, neutral lipids, and surfactant proteins SP-A to-D. Its composition is highly specific, with dipalmitoyl-PC, palmitoyl-myristoyl-PC, and palmitoyl-palmitoleoyl-PC as its predominant PC species, but with low polyunsaturated phospholipids. Changes in pulmonary metabolism and function in response to injuries depend on their duration and whether adaptation can occur. We examined in rats prolonged (7 days) versus acute (2 days) exposure to non-lethal oxygen concentrations (85%) with respect to the composition and metabolism of individual lung phospholipid molecular species. Progressive inflammation, structural alteration, and involvement of type II pneumocytes were confirmed by augmented bromodeoxyuridine incorporation, broadening of alveolar septa, and increased granulocyte, macrophage, SP-A, and SP-D concentrations. Surfactant function was impaired after 2 days, but normalized with duration of hyperoxia, which was attributable to inhibition but not to alteration in SP-B/C concentrations. Phospholipid pool sizes and PC synthesis by lung tissue, as assessed by [methyl-(3)H]-choline incorporation, were unchanged after 2 days, although after 7 days they were elevated 1.7-fold. By contrast, incorporation of labeled PC into tissue pools of surfactant and lung lavage fluid decreased progressively. Moreover, concentrations of arachidonic acid containing phospholipids were augmented at the expense of saturated palmitoyl-myristoyl-PC and dipalmitoyl-PC. We conclude a persisting impairment in the intracellular trafficking and secretion of newly synthesized PC, accompanied by a progressive increase in alveolar arachidonic acid containing phospholipids in spite of recovery of acutely impaired surfactant function and adaptive increase of overall PC synthesis.

Phosphatidylcholine metabolism of rat trachea in relation to lung parenchyma and surfactant.

Pulmonary surfactant prevents alveolar collapse and contributes to airway patency by reducing surface tension. Although alveolar surfactant, consisting mainly of phospholipids (PL) together with neutral lipids and surfactant-specific proteins, originates from type II pneumocytes, the contribution of airway epithelia to the PL fraction of conductive airway surfactant is still debated. We, therefore, analyzed the composition, synthesis, and release of phosphatidylcholine (PC) molecular species as the main surfactant PL of the rat trachea compared with the lung. Analyses of individual PC molecular species with HPLC and electrospray ionization mass spectrometry revealed that the rat trachea contained and synthesized much more palmitoyloleoyl-PC, palmitoyllinoleoyl-PC, and palmitoylarachidonoyl-PC, together with increased amounts of alkylacyl-PC, and less surfactant-specific species such as dipalmitoyl-PC than the lung. Organ cultures with [methyl-3H]choline as precursor of PC revealed that, in the trachea, synthesized PC was retained in the tissue, rather than secreted. [Methyl-3H]choline-labeled dipalmitoyl-PC was a negligible component in the trachea, and, in contrast to the lungs, palmitoyloleoyl-PC was enriched in tracheal secretions. We conclude that the surfactant fraction in the airways does not originate from the airways but is produced in the alveolar space and transported upward.

Molecular species compositions of lung and pancreas phospholipids in the cftr(tm1HGU/tm1HGU) cystic fibrosis mouse.

Fatty acid analysis of phospholipid compositions of lung and pancreas cells from a cystic fibrosis transmembrane regulator (CFTR) negative mouse (cftr(-/-))suggested that a decreased concentration of docosahexaenoate (22:6(n-3)) and increased arachidonate (20:4(n-6)) may be related to the disease process in cystic fibrosis (CF). Consequently, we have determined compositions of the major phospholipids of lung, pancreas, liver, and plasma from a different mouse model of CF, the cftr(tm1HGU/tm1HGU) mouse, compared with ZTM:MF-1 control mice. Electrospray ionization mass spectrometry permitted the quantification of all of the individual molecular species of phosphatidylcholine (PtdCho), phosphatidylethanolamine (PtdEtn), phosphatidylglycerol (PtdGly), phosphatidylserine (PtdSer), and phosphatidylinositol (PtdIns). There was no deficiency of 22:6(n-3) in any phospholipid class from lung, pancreas, or liver from mice with the cftr(tm1HGU/tm1HGU). Instead, the concentration of 20:4(n-6) was significantly decreased in plasma PtdCho species and in pancreas and lung species of PtdEtn, PtdSer, and PtdIns. These results demonstrate the variability of membrane phospholipid compositions in different mouse models of CF and suggest that in cftr(tm1HGU/tm1HGU) mice, the apparent deficiency was of 20:4n-6- rather than of 22:6n-3-containing phospholipid species. They highlight a need for detailed phospholipid molecular species analysis of cells expressing mutant CFTR from children with CF before the therapeutic effects of administering high doses of 22:6(n-3)-containing oils to children with CF can be fully evaluated.