Escherichia coli hemolysin (HlyA) is a prototype of a large family of pore-forming proteinaceous exotoxins that have been implicated in the pathogenetic sequelae of severe infection and sepsis, including development of acute lung injury. In the present study in rabbit alveolar macrophages (AMs), subcytolytic concentrations of purified HlyA evoked rapid synthesis of platelet-activating factor, with quantities approaching those in response to maximum calcium ionophore challenge. In parallel, large quantities of leukotriene (LT) B4 and 5-, 8-, 9-, 12-, and 15-hydroxyeicosatetraenoic acid (HETE) were liberated from HlyA-exposed AMs depending on exogenous arachidonic acid (AA) supply. Coadministration of eicosapentaenoic acid (EPA) dose dependently suppressed generation of the proinflammatory lipoxygenase products LTB4 and 5-, 8-, 9-, and 12-HETE in parallel with the appearance of the corresponding EPA-derived metabolites LTB5 and 5-, 8-, 9-, and 12–hydroxyeicosapentaenoic acid (HEPE). At equimolar concentrations, EPA turned out to be the preferred substrate over AA for these AM lipoxygenase pathways, with the sum of LTB5and 5-, 8-, 9-, and 12-HEPE surpassing the sum of LTB4 and 5-, 8-, 9-, and 12-HETE by >80-fold. In contrast, coadminstration of EPA did not significantly reduce HlyA-elicited generation of the anti-inflammatory AA lipoxygenase product 15-HETE. We conclude that AMs are sensitive target cells for HlyA attack, resulting in marked proinflammatory lipid mediator synthesis. In the presence of EPA, lipoxygenase product formation is shifted from a pro- to an anti-inflammatory profile.
- Escherichia coli hemolysin
- n-3 fatty acid
alveolar macrophages (AMs) represent the resident phagocytic effector cells at the interface between external environments and the lung parenchyma, playing a pivotal role for pulmonary host defense competence (35, 40). The capacity of these cells to generate large quantities of lipid mediators, in particular the 5-lipoxygenase products arachidonic acid (AA) and platelet-activating factor (PAF), is essential for lung antimicrobial defense but may also largely contribute to pathogenic sequelae in acute and chronic inflammatory lung diseases (30,32, 35, 42). The functional differentiation from peripheral blood monocytes to AMs is accompanied by a relative enrichment of membrane phospholipid pools and changes in the machinery of lipid mediator synthesis. In AMs, 5-lipoxygenase is localized in a cytoplasmic pool and an intranuclear pool, both of which move to the nuclear envelope after cell activation (36,37). Free AA for the lipoxygenase step may then be provided by endogenous phospholipolytic activities or by supply of the AMs with free exogenous AA, with a 5-lipoxygenase activating protein serving as an AA transfer protein (35). The arising products, leukotriene (LT) B4 and 5-hydroxyeicosatetraenoic acid (HETE), both possessing strong neutrophil chemotactic activity, are major contributors to the recruitment of these leukocytes into the alveolar compartment under inflammatory conditions (10,21, 29). In addition, AMs have been shown to generate 5-, 8-, 9-, and 12-HETE when appropriately stimulated, which are also implicated in leukocyte chemotaxis, activation, and aggregation (20, 21, 44).
The n-6 fatty acid AA represents the predominant polyunsaturated fatty acid in the common Western diet and current nutritional formulations. In contrast, n-3 fatty acids such as eicosapentaenoic acid (EPA) and docosahexaenoic acid make up an appreciable part of the fat in cold-water fish and seal meat and may serve as alternative lipid precursors for various lipoxygenase pathways, with the formation of 5-series LTs instead of the 4-series LTs derived from AA and hydroxyeicosapentaenoic acids (HEPEs) instead of HETEs (14, 27, 31, 34). Interestingly, many of the n-3 fatty acid-derived metabolites, including LTB5, possess markedly reduced inflammatory and vasomotor potencies compared with the AA-derived lipid mediators and may even exert antagonistic functions. Against this background, oral and intravenous supplementation with EPA-rich fish oil and n-3 fatty acid-enriched lipid infusions, intending to shift the AA-to-EPA ratio toward predominance of the latter lipid mediator precursor, is under investigation as a therapeutic regimen in various inflammatory diseases (13-15, 19, 22,23, 41). An impact of n-3 fatty acids on AM lipid mediator synthesis has previously been suggested (25, 26).
Exposure to bacterial toxins is one of the most potent challenges AMs may undergo, with sepsis and severe pneumonia representing prototype diseases. Besides lipopolysaccharide (LPS), pore-forming proteinaceous exotoxins are of major interest because they represent a large family of toxins originating from both gram-positive and gram-negative bacteria known to exert profound effects on various target cell types (1). In retrospective studies and animal models,Escherichia coli has been identified as first, the most frequent organism isolated in patients with bacteremia due to gram-negative organisms and second, a leading cause of acute respiratory distress syndrome and pneumonia in E. coli-induced sepsis syndrome (11, 27,48). Interestingly, ∼50% of Escherichia coli isolates causing extraintestinal infections in humans produce a proteinaceous hemolysin (HlyA), which creates transmembrane lesions of 1–2 nm as the basic toxin mechanism and has been implicated as a significant virulence factor in human diseases (11).
Employed in subcytolytic concentrations, HlyA was shown to represent the most potent stimulus of the phosphatidylinositol hydrolysis-related signal transduction pathway in human neutrophils known to date (12), linked with granule exocytosis, respiratory burst, and lipid mediator generation. Correspondingly, stimulation of phosphoinositide hydrolysis was recently demonstrated for human endothelial cells (11). Moreover, in perfused lungs, this toxin was found to provoke pulmonary hypertension, severe vascular leakage, and marked lipid mediator generation (8,14).
In the present study, we demonstrate that AMs are highly sensitive target cells for HlyA attack. Subcytolytic concentrations of the toxin provoked pronounced PAF synthesis, concomitant with the liberation of large quantities of LTB4 and proinflammatory HETEs, the latter being critically dependent on the exogenous precursor fatty acid supply. Interestingly, EPA turned out to be the preferred substrate over AA, suppressing proinflammatory lipid mediator generation concomitant with the appearance of large amounts of LTB5and HEPEs. Both the responsiveness of AMs to the bacterial exotoxin and the impact of n-3 versus w-6 fatty acid supply may be of interest for the understanding and treatment of lung injury in severe infection.
20-COOH-LTB4, 20-OH-LTB4, LTC4, LTE4, 6-trans-LTB4, 6-trans-12-epi-LTB4, LTB4, LTD4, 5S,6R-dihydroxyeicosatetraenoic acid (diHETE), 5S,6S-diHETE, 15-HETE, 11-HETE, 8-HETE, 12-HETE, 9-HETE, 5-HETE, and prostaglandin (PG) B1were purchased from BIOMOL (Hamburg, Germany). LTB4δ-lactone, 5,6-dihydroxyeicosatrienoic acid δ-lactone, and the epoxyeicosatrienoic acid (EET) standards 14,15-EET, 11,12-EET, 8,9-EET, and 5,6-EET were supplied by Cascade Biochem (Reading, UK). All eicosanoids were checked for purity and quantified spectrophotometrically before use. Butylated hydroxytoluene, pentanesulfonic acid, and triethylamine were obtained from Sigma (Munich, Germany). EDTA and formic acid were purchased from Merck (Darmstadt, Germany). Water was purified with a Milli-Q system (Millipore, Eschborn, Germany). Methanol was delivered by Burdick & Jackson (Muskegon, MI), acetonitrile was purchased from J. T. Baker (Deventer, The Netherlands), and isopropyl alcohol was obtained from Fluka (Buchs, Switzerland). All solvents used were HPLC grade or better. The prepared solutions were passed through a 0.2-μm filter (Millipore) and degassed under vacuum. Before use, the solvent mixtures and the mobile phase were additionally degassed by sonication. Chromatographic supplies included 5-μm silica gel column packing (Shandon, Astmoor, UK), Chromabond C-18ec cartridges (Machery-Nagel, Düren, Germany) and HPLC-grade solvents distilled in glass (Fluka, Heidelberg, Germany). Medium 199, FCS, HEPES, Hanks' balanced salt solution, phosphate-buffered saline, and antibiotics were obtained from GIBCO (Karlsruhe, Germany). PAF (1-O-hexadecyl-2-acetyl-sn-glycero-3-phosphocholine) and A-23187 were obtained from Sigma (Deisenhofen, Germany). The PAF-receptor antagonist WEB-2086 was generously supplied by Boehringer Ingelheim (Ingelheim, Germany). [3H]PAF and [3H]serotonin were obtained from Amersham (Dreieich, Germany). All other biochemicals were obtained from Merck (Munich, Germany).
Preparation of AMs.
For the isolation of AMs, the tracheae of 2.5- to 3.5-kg rabbits were cannulated, and lung lavage with 30 ml of saline containing 3 mM EDTA was performed three times as previously described (17). The alveolar cells were recovered from the pooled lavage by centrifugation at 300 g for 15 min at room temperature and kept for 1 h in FCS (37°C, CO2-equilibrated incubator). Immediately before experimental use, the cells were washed twice and suspended in Hanks' balanced salt solution-HEPES buffer. The yield of AMs from each rabbit was in the range of 0.32–1.0 × 108 cells. AMs were identified by morphological criteria (air-dried and Giemsa-stained smears) and by esterase staining in each experiment as well as by random electron-microscopic examinations. The percentage of macrophages within the total number of lavaged cells consistently was >96%; contaminating polymorphonuclear neutrophils were <1.5% in all preparations. Cell viability in the presence of stimulus application was assessed by trypan blue exclusion and lactate dehydrogenase (LDH) release. Under all experimental conditions, viability of AMs was >95% (trypan blue exclusion) and LDH release was consistently <3%.
Preparation of E. coli HlyA.
HlyA was prepared as previously described (1). The endotoxin content of the preparation was reduced to ∼3 ng LPS/μg protein. The hemolytic titer was assessed directly before use and is expressed in hemolytic units (HU) per milliliter; 1 HU/ml corresponded to ∼100 ng protein/ml. “Aging” of HlyA was provoked by preincubation of the toxin at 37°C for 1 h.
Solid-phase extraction and HPLC equipment and stationary phases.
Solid-phase extraction (SPE) was performed with octadecylsilyl (ODS) cartridges (Chromabond C18ec) mounted in a Chromabond vacuum manifold (all from Macherey-Nagel). The cell supernatants were extracted on 200 mg/3 ml cartridges (sorbent mass, 200 mg; reservoir volume, 1 ml). The HPLC equipment comprised a Gynkotek model GINA 160 autosampler (Gynkotek, Munich, Germany) equipped with a 500-μl sample loop and a 250-μl injection syringe, a Gynkotek model 480 HPLC pump, a Gynkotek model 320S photodiode array detector, a Waters model 990 photodiode array detector (Waters/Millipore), a Shimadzu model SPD-6A variable wavelength ultraviolet detector (Shimadzu, Kyoto, Japan) and two in-line filters (0.2-μm stainless steel frit) from Latek (Eppelheim, Germany). All separations were carried out on an analytic HPLC column (length 250 mm × ID 4 mm) filled with ODS-Hypersil (particle size, 3 μm; pore size, 100 Å) obtained from Shandon. In parallel with the processing of genuine samples, calibration mixtures of authentic standards were analyzed and used for absolute quantitation of the actual set of samples. Moreover, quality control samples containing known amounts of standards dissolved in the eicosanoid-free perfusate were processed with each sample batch under the same conditions, allowing for determination of the actual recovery for each compound.
SPE and HPLC procedures.
AA metabolites released in macrophages were analyzed as previously described by this group (24) with a method combining1) SPE, 2) isocratic reverse-phase HPLC separation, and 3) online photodiode array detection and spectrum analysis for identification and measurement of all LTs, HETEs, and EETs within one run. After collection, all samples were supplemented with constant amounts of PGB1 as an internal standard, centrifuged for 10 min at 1,500 g, and immediately subjected to SPE or stored at −20°C until analyzed. Before SPE, the samples were equilibrated with EDTA for 15 min, treated with isopropanol, then vortexed for 5 min, applied to the conditioned extraction columns mounted in a vacuum manifold without vacuum suction, and subjected to the extraction procedure. Eluates were collected in 1.1-ml vials (Chromacol, Welwyn Garden City, UK) and dried gently under a stream of N2. Dried samples were extracted with 60 μl of mobile phase, vortexed, sonicated, centrifuged briefly, and injected into the HPLC system. The mobile phase consisted of a 54:8:38:0.001 (vol/vol/vol/vol) mixture of acetonitrile, methanol, water, and formic acid supplemented with 6 mM pentanesulfonic acid and 20 μM EDTA (pH 4.38 adjusted with 10% aqueous triethylamine). This technique achieved baseline separation of all eicosanoids within 43 min (Fig.1) except for the coeluting 20-COOH-LTB4 and 20-OH-LTB4 and 6-trans-LTB4 and 6-trans-12-epi-LTB4 as well as for 8-HETE and 12-HETE. The eluting analytes were monitored with a photodiode array detector, which provided full ultraviolet spectra (190–340 nm) of eluting compounds and allowed checking for peak purity and subtraction of possible coeluting material. The eicosanoids were identified and quantified at 270 (LTs), 237 (HETEs), and 204 (EETs) nm by spectra plot on the peak maximum, including the use of peak purity control method, coelution with commercial standards, and an internal standard method employing PGB1.
Post-HPLC PAF bioassay.
PAF production by AMs was quantified by HPLC separation and induction of [3H]serotonin release from prelabeled rabbit platelets. After termination of macrophage incubation, the total cell-associated and extracellular PAF content was lipid extracted and subjected to straight-phase HPLC separation. The column (25 × 0.46 cm) was packed with 5-μm silica gel particles and eluted isocratically with acetonitrile-methanol-phosphoric acid (95.5:3.5:1) at a flow rate of 1.8 ml/min. Eluate fractions were collected at the appropriate PAF retention time, again lipid extracted for removal of phosphoric acid present in the mobile phase, evaporated to dryness, and redissolved in 50 μl of assay buffer for induction of platelet serotonin release. Preparation of platelets and the protocol for the bioassay were essentially as published by Pinckard et al. (39). The serotonin secretion into the platelet supernatant was determined by liquid scintillation counting and related to baseline levels in buffer-incubated platelets. Maximum serotonin release was assessed in parallel by cell lysis with Triton X-100 (final concentration 0.83% wt/vol). Aliquots of each sample were used to ascertain the specifity of platelet secretion by the inhibitory effect of the PAF-receptor antagonist WEB-2086 (1 μM).
Measurement of PAF by bioincorporation of radiolabel.
PAF identification was further verified by incorporation of labeled acetate according to Tessner et al. (43). Macrophages were stimulated in the presence of 50 μCi of [3H]acetate (7.75 Ci/mmol) in a total volume of 1 ml. Reactions were stopped by addition of three volumes of chloroform-methanol (2:1 vol/vol), and extraction was performed according to Bligh and Dyer (3). The entire lipid extract was evaporated to dryness, redissolved in 60 μl of mobile phase, and subjected to straight-phase HPLC separation as described inPost-HPLC PAF bioassay. Eluate fractions corresponding to appropriate standard retention times were collected and assayed for radioactivity by liquid scintillation counting.
Measurement of LDH.
As a marker for overt cytotoxicity, LDH was quantified by a standard colorimetric technique. Enzyme release is expressed as a percentage of total enzyme activity liberated in the presence of 100 μg/ml of mellitin.
For all experiments of the measurement of PAF (extraction of cells and cell supernatant) and LTs (measured in extracted cell supernatant), 1 × 107 cells were used. According to preceding pilot experiments, the following concentrations were chosen for the different agents: 1 HU/ml of HlyA and 10 μM A-23187. The maximum dose of HlyA was chosen to stay below the threshold of overt cell lysis (seeControl experiments for data on LDH release).
The data are means ± SE. Analysis of variance was used to test for differences between the different groups; a P value of <0.05 was considered to indicate significance.
PAF generation in AMs.
Analysis of liberated and cell-bound PAF by post-HPLC bioassay revealed a rapid four- to fivefold increase in the formation of this lipid mediator in response to HlyA challenge (Fig.2). Maximum amounts of PAF were elicited at 1 HU/ml of HlyA and nearly approached those in response to 5–10 μM A-23187. Analysis of kinetics displayed plateauing of the PAF release reaction in response to both agents within 5–10 min. These data from post-HPLC bioassay were further corroborated by measurement of [3H]acetate bioincorporation (data not shown).
Eicosanoid production in AMs.
Incubation of isolated AMs with HlyA (1 HU/ml) in the presence of 10 μM free AA provoked the synthesis of large amounts of LTB4 (maximal concentration ≈ 3 ng/1 × 107 cells; Fig. 3). Product release into the supernatant plateaued within 5–10 min, demonstrated dose dependency on HlyA (maximum efficacy at ≈1 HU/ml), and was strictly dependent on exogenous precursor fatty acid supply, peaking at 20 μM AA (Fig. 4). In the absence of exogenous AA, no significant LTB4 release was provoked by HlyA.
In addition to LTB4, HlyA (1 HU/ml) evoked large quantities of the lipoxygenase products 5-, 8-, 9-, 12-, and 15-HETE, with a predominance of 5- and 15-HETE (Fig. 5).
Combined application of HlyA and both precursor fatty acids, AA (fixed concentration of 10 μM) and EPA (increasing concentrations of 2.5–20 μM), resulted in the release of both 4-series and 5-series LTs, HETEs, and HEPEs (Fig. 6). Within the EPA-derived products, the rank order was 5-HEPE > 15-HEPE ≈ 8- and 12-HEPE > LTB5 ≈ 9-HEPE. The overall sum of EPA-derived lipoxygenase products detected in the presence of both 10 μM EPA and 10 μM AA surpassed that of AA-derived products under these conditions. In particular, generation of the 5-lipoxygenase products of AA (LTB4 and 5-HETE) as well as of the 8-, 9-, and 12-lipoxygenase products of AA was suppressed to near baseline in the presence of increasing concentrations of EPA. In contrast, 15-HETE liberation declined only very moderately in the presence of free EPA, with 15-HETE levels still ranging at ≈13 ng/ml even in the presence of 20 μM free EPA.
In contrast to the pronounced LTB4/5 and HET(P)E generation in HlyA-challenged AMs, no EET release and no release of nonenzymatic hydrolysis products of LTA4(6-trans-LTB4 and 6-trans-12-epi-LTB4) were noted in the presence of both free AA and free EPA (example shown in Fig. 1).
Aged HlyA used at a protein concentration corresponding to 1 HU/ml did not induce PAF generation or eicosanoid production in isolated alveolar macrophages (n = 4 experiments each). Similarly, LPS (used up to 100 ng/ml) was ineffective in provoking the acute release of these lipid mediators (2-h incubation; n = 4 experiments each). In the concentration range used, HlyA incubation evoked only very moderate LDH release as a marker of overt cell lysis: ∼3–5% of total cellular LDH was liberated within the current 20-min observation period in response to 1 HU/ml of HlyA. Even a twofold higher toxin dose, surpassing the range of HlyA presently employed, provoked the liberation of only 12–16% of total cellular LDH.
In the present study, incubation of AMs with subcytolytic concentrations of HlyA provoked dose-dependent lipid mediator generation, with quantities approaching those in response to calcium ionophore challenge. PAF liberation was accompanied by the release of various lipoxygenase products of AA (LTB4 and 5-, 8-, 9-, 12-, and 15-HETE), the latter being critically dependent on exogenous precursor fatty acid supply. Coadministration of AA and EPA demonstrated that the n-3 fatty acid is by far the preferred substrate for the lipoxygenase pathways, resulting in near-complete suppression of LTB4 and 5-, 8-, 9-, and 12-HETE, with concomitant appearance of large quantities of LTB5 and 5-, 8-, 9-, and 12-HEPE. In contrast, the anti-inflammatory agent 15-HETE was only marginally reduced in the presence of EPA.
There is compelling evidence that the HlyA-elicited activation of AMs was evoked by the bacterial exotoxin itself and not by contaminating LPS. 1) A toxin preparation with markedly depleted LPS content was used (residual contamination ∼ 3 ng LPS/μg protein, corresponding to ∼300 pg/ml of LPS in the experiments with the optimal HlyA dose of 1 HU/ml). 2) Rapid loss of toxin activity occurred on aging of the preparation at 37°C, a feature repeatedly described for hemolysin but not compatible with established endotoxin characteristics (11). 3) The experiments were performed in the absence of serum components, i.e., LPS binding proteins, which exert an important function in communicating LPS effects. And 4), control studies with LPS concentrations up to 3 μg/ml, i.e., four orders of magnitude above the current maximum LPS contamination, did not provoke any significant liberation of PAF and AA lipoxygenase products under the given experimental conditions and within the observation period of 10 min. These findings do not deny the fact that responsiveness to LPS is an important feature of AMs, but they exclude that the low LPS contamination of the HlyA preparation might be responsible for the currently described rapid lipid mediator synthesis in AMs.
AM HlyA challenge provoked rapid, pronounced PAF synthesis, with maximum levels of this lipid mediator being elicited by 1 HU/ml of HlyA, which corresponds to a subcytolytic concentration range for the macrophages as evident from the very moderate LDH liberation. As anticipated, the presence of exogenous AA was not a prerequisite for the exotoxin-induced PAF generation. The potency of HlyA to elicit AM PAF synthesis is of interest against the background that this lipid mediator has been implicated in the pathogenetic sequelae of a variety of inflammatory lung diseases (4, 31,33). PAF was suggested to contribute to lung parenchyma recruitment and activation of various leukocyte types, including neutrophils, eosinophils, mononuclear phagocytes, and lymphocytes; it may be involved in the transendothelial and transepithelial passage of these leukocytes, and it was found to be a “priming” agent of inflammatory cells.
Some of the pulmonary PAF effects have been ascribed to secondary induction of LT synthesis (6, 7), and a strong LTB4 release reaction was indeed noted to accompany the enhanced PAF synthesis in HlyA-stimulated lungs. Overall, LT quantities were again found to approach those in response to A-23187 challenge. However, a more detailed analysis did clearly demonstrate that the exotoxin-elicited lipoxygenase product formation is an independent event and is not mediated via an autocrine PAF loop. First, LTB4 generation occurred even more rapidly than PAF synthesis, with values plateauing within 1 min, whereas PAF plateauing was observed within 5 min. And second, the employment of a PAF-receptor blocker did not interfere with the HlyA-elicited LT synthesis (data not shown). Interestingly, the formation of all lipoxygenase products in HlyA-challenged AMs was critically dependent on exogenous AA supply, with only very minor amounts of eicosanoids being released in the absence of exogenous precursor fatty acids. This finding is reminiscent of N-formyl-methionyl-leucyl-phenylalanine- and HlyA-evoked LT generation in human neutrophils where both HlyA and the peptide ligand caused PAF liberation but nevertheless demanded exogenous AA to induce the parallel activation of the 5-lipoxygenase pathway (12). A most likely explanation for such a finding may be the limited amount or accessibility of hemolysin-induced AA release to the 5-lipoxygenase enzyme. Exogenous AA may then compensate for the restricted availability of the endogenous precursor to be readily metabolized to the hydroperoxy intermediate 5-hydroperoxyeicosatetraenoic acid and subsequently to the unstable 5,6-epoxide LTA4, the latter being converted to LTB4 as the major enzymatic secretion product. This finding may be relevant in view of the substantial levels of free extracellular AA arising at the sites of inflammatory events, estimated to surpass 10 μM (18, 43). Moreover, plasma free AA concentrations between 25 and 100 μM were recently detected in septic patients (28). The current employment of 10 μM AA, which per se did not evoke any macrophage LT synthesis, unmasked HlyA as a most potent activator of AM 5-lipoxygenase activity. No studies with preincubation of macrophages with AA or EPA were performed, known from a previous study (5) to result in progressive incorporation of these lipid mediator precursors into different macrophage phospholipid pools.
In parallel with LTB4, marked HETE generation was noted in HlyA-challenged AMs, with total quantities surpassing those of the LTs severalfold. Next to LTB4, the chemotactic and proinflammatory potencies of which are well documented for the alveolar compartment (10, 29), 5-, 8-, 9-, and 12-HETE, which may be synthesized via either the 12-lipoxygenase or the cytochrome P-450 pathway, are well-established chemotactic agents (14, 29, 44,47). Moreover, human neutrophils and monocytes may convert 5-HETE to 5-oxo-ETE that again is very effective in stimulating migration of human eosinophils (38). In contrast, 15-HETE, which represents the predominant HETE in the HlyA-challenged AMs, is known as an anti-inflammatory agent because it inhibits the proinflammatory 5-lipoxygenase pathway and the production of tumor necrosis factor-α as well as the generation of eicosanoids via the 12-lipoxygenase pathway (9, 45,46). At the same time, it has a potent autostimulatory effect on its own biosynthesis (44, 47). Cysteinyl-LT generation was not detected in HlyA-exposed AMs, in contrast to a previous report (27) on the synthesis of these LTs in AMs during phagocytotic events. Similarly, no EET formation was noted in HlyA-challenged AMs. This finding is remarkable in view of the fact that bacterial exotoxins such as HlyA were recently noted to provoke a marked release of 11,12-EET > 8,9-EET > 14,15 EET > 5,6-EET into the intravascular space in perfused rabbit and human lungs (24), suggesting that the cellular sources of pulmonary EET generation are different from AMs.
Previous studies (8, 11, 12,14) in human neutrophils and endothelial cells demonstrated that HlyA is a most potent inductor of the preformed phosphoinositide hydrolysis-related signal transduction pathway. Although not directly addressed in the present investigation, corresponding efficacy in the AMs might readily explain the activation of the 5-lipoxygenase and additional lipoxygenase pathways as “downstream” events. The molecular mechanisms by which HlyA triggers phosphoinositide hydrolysis and possibly additional signal transduction pathways remain to be elucidated.
An impressive finding of the present study is the profound impact of exogenous EPA on the profile of lipoxygenase products liberated from the HlyA-challenged AMs. Dose-dependent suppression of LTB4and 5-, 8-, 9-, and 12-HETE formation was noted in parallel with the appearance of major amounts of LTB5 and the corresponding HEPEs. At equimolar concentrations of free exogenous AA and EPA (10 μM each), the sum of 5-, 8-, 9-, and 12-HEPE and LTB5surpassed the sum of the proinflammatory agents 5-, 8-, 9-, and 12-HETE and LTB4 by more than nearly eight orders of magnitude. The most likely explanation for this dramatic impact of an altered state of precursor fatty acid supply on the lipid mediator composition is that1) exogenous EPA, similar to exogenous AA, has ready access to the activated lipoxygenases to serve as a substrate for the different oxygenation steps and 2) at equimolar concentrations, the maximum metabolizing velocity of EPA is even higher than that of AA for these lipoxygenase pathways. This has already been shown for AA and EPA when competing for the 5-lipoxygenase step in human neutrophils (16), in-line with the finding of a higher affinity of the purified neutrophil 5-lipoxygenase for EPA compared with that for AA; however, the dramatic preference of EPA over AA for the various lipoxygenase pathways in AMs has hitherto not been recognized. Most interestingly, this impact of EPA turned out to be different for the generation of the “anti-inflammatory” agent 15-HETE. Increasing concentrations of the n-3 lipid precursor resulted in the appearance of major quantities of 15-HEPE; however, the high level of 15-HETE formation was not significantly suppressed. The underlying reasons for this apparently differential regulation of 15-HETE synthesis compared with the other macrophage lipoxygenase steps are currently unknown but may be of interest in view of the role of these lipid mediators in controlling pro- and anti-inflammatory events in the alveolar compartment.
In conclusion, the present study presents two interesting observations that may have an impact on inflammatory events in the alveolar compartment, in particular under infectious conditions. First, the AMs turned out to be highly sensitive target cells for HlyA, resulting in strong lipid mediator synthesis including PAF, LTB4, and various HETEs. This is assumed to be relevant for infections withE. coli, given the fact that HlyA is a well-known pathogenicity factor of these bacteria (1), but may also be of interest against the background that related pore-forming proteinaceous exotoxins are generated by a large number of gram-negative (e.g., Pseudomonas aeruginosa) and gram-positive (e.g., Staphylococcus aureus andStreptococcus pneumoniae) bacteria, most relevant for sepsis and pneumonia (1, 2). And second, offering the alternative lipid mediator precursor EPA dramatically suppressed the generation of the proinflammatory agents LTB4 and 5-, 8-, 9-, and 12-HETE together with the appearance of LTB5 and HEPEs, which are assumed to be devoid of proinflammatory potencies but to possess antagonistic potencies. Of note, the synthesis of the anti-inflammatory AA lipoxygenase metabolite 15-HETE was not blocked in the presence of EPA. These in vitro observations fit very well with recent data from exotoxin-challenged perfused lungs in which coapplication of AA enhanced but coapplication of EPA dampened inflammatory events such as vasomotor changes and vascular leakage (14). The n-3 precursor fatty aid supply, whether administered via an intravenous route, e.g., by use of fish oil-derived lipid infusions, or whether applied via an inhalant route may thus offer to manipulate inflammatory lipid mediator synthesis in the alveolar compartment.
We thank Prof. Dr. Sucharit Bhakdi for generously providingEscherichia coli hemolysin.
This work was supported by Deutsche Forschungsgemeinschaft Grant SFB 547.
Address for reprint requests and other correspondence: U. Sibelius, Dept. of Internal Medicine, Justus-Liebig-Univ., 35385 Giessen, Germany (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society