INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A wide variety of conditions including sepsis, trauma, hypovolemic shock, and reperfusion injury have been associated with lung injury (1). More than one-third of patients with sepsis also have mild to moderate lung injury, and one-fourth have established acute respiratory distress syndrome (ARDS). Endotoxin, a lipopolysaccharide (LPS) component of the cell wall of gram-negative bacteria is an important mediator in the pathogenesis of sepsis and ARDS. LPS leads to the production and release of a wide array of mediators including cytokines such as tumor necrosis factor- (TNF-) and interleukin-1 (IL-1), chemokines such as interleukin-8 (IL-8), and oxygen free radicals from neutrophils, macrophages, endothelium, and epithelium. Of these mediators, TNF- is the best characterized and plays an important role in causing systemic organ damage either directly or indirectly by the stimulation of other mediators.

The lung has a plethora of cells that could be a source of inflammatory mediators contributing to an overall systemic host response. In addition to the large number of inflammatory cells (e.g., neutrophils and macrophages) that are present in the lung, many of the structural cells such as epithelial cells, endothelial cells, and interstitial cells can produce a host of proinflammatory mediators in response to a variety of stimuli. For example, alveolar type II cells can produce TNF- in response to LPS (2) as well as in response to mechanical ventilation (3), although their regulatory mechanisms appear to be different from that reported for inflammatory cells such as macrophages. Rat alveolar type II cells (2) and bronchial epithelial cells (4) can also produce macrophage inflammatory protein-2 (MIP-2), originally found in a cell line derived form mouse peritoneal macrophages (5). MIP-2 is a chemokine for neutrophil recruitment and activation (5), and an important mediator leading to acute lung injury.

In addition to LPS-induced changes in airway and vascular mechanics, a role for endogenous catecholamines in the cellular response to endotoxin is apparent. Several reports (6, 7) have demonstrated that there may be alterations of adrenergic receptor response associated with endotoxemia or sepsis in cultured rat hepatocytes. In isolated monocytes/macrophages or whole blood culture, the administration of -adrenoreceptor agonists can inhibit LPS-induced TNF- production (8). Both animal and human alveolar type II cells are rich in -adrenergic receptors (8, 16), but it is unclear whether the -adrenergic receptor reactivity is altered in acute lung injury. Furthermore, it is unknown if the specific subtype of -adrenergic receptor can attenuate LPS-induced inflammatory response in the absence of hemodynamic influences after acute lung injury.

Studies to address questions such as these have often made use of cell culture system containing one cell type, or coculture of a few cell types (17); however, the lung is a complex structure with many cell types communicating with each other at many levels. At the other extreme, investigators have used in vivo models, but it is often difficult to isolate specific mechanisms because of the contribution of mediators and cells from the circulation, nervous innervation of the lung, and hormonal mechanisms. Thus, there is a need for a model in which lung cells are kept intact more closely simulating in vivo conditions, but in which the effect of the circulation is absent. The purpose of this study was threefold: first, to investigate whether lung explants could serve as this ex vivo acute inflammatory model as reflected by release of cytokines and lipid peroxidation after LPS administration; second, if so, to determine whether catecholamines can influence the appearance of cytokines and lipid peroxidation in LPS-stimulated lung tissues; and third, to clarify the mechanisms by which catecholamines regulate the LPS-induced inflammatory response in lung explants.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung Explant Preparation

Adult male Sprague-Dawley rats (Charles River Laboratories, St. Constant, PQ, Canada) weighing 400 to 600 g were used in accordance with he institutional animal welfare guidelines of the University of Toronto. All animals were anesthetized with pentobarbital sodium (40 mg/kg, intraperitoneally) and intubated through a tracheostomy with a sterile angiocatheter (14-gauge, Angiocath; Becton Dickinson Infusion Therapy Systems, Sandy, UT). The surface of the anterior chest wall and upper abdomen were sterilized with 70% ethanol. Through a midline abdominal incision, the animal was exsanguinated by dissecting the abdominal aorta. After opening the chest, the right ventricle was punctured, and a needle-guided vascular catheter (16-gauge, Angiocath; Becton Dickinson Infusion Therapy Systems) was advanced into the main pulmonary artery. The pulmonary vessels were perfused in situ with 30 ml of normal saline. Using aseptic technique the trachea, lungs, and the heart were dissected en bloc from the animals.

To obtain thin and reproducibly sized pieces of lung tissue, the lungs were prepared for slicing by adapting an agarose inflation technique (18). Low-melting-point agarose (1%, wt/vol; agarose type VII: low gelling temperature, A-4018; Sigma Chemical Co., St. Louis, MO) was dissolved in bicarbonate-buffered culture medium (minimum essential medium, sodium bicarbonate, amino acid supplement, sodium pyruvate, vitamin supplement, 50 mg/ml gentamicin and insulin), heated in a microwave oven and cooled to 37° C before use. The agarose was then instilled as a liquid into the trachea, inflating the lung to approximate total lung capacity (4.8 ml/100 g body weight). The lungs were placed in a sterile container at 4° C for 30 min, solidifying the agarose. The lungs were then separated from the heart. Four percent agarose dissolved in bicarbonate-buffered culture medium heated to 37° C was poured into a sterile open-mouthed 60-ml syringe (B-D; Becton Dickinson, Franklin Lakes, NJ) until it was two-thirds full. The left and right lungs were then separately placed in syringes and embedded by filling the syringe with 4% agarose. A rubber stopper was used to close the mouth of the syringe. The syringe containing the lung was cooled at 4° C for 30 min, solidifying the embedding agarose. The resulting agarose-filled and -embedded lungs were then sliced on a microtome (PanaVise Products, Inc., Reno, NV) into slices 0.5 to 1.0 mm thick.

The lung explants were incubated overnight in a culture plate (90 × 15 mm Petri Dish, Nunclon; Nalge Nunc International, Copenhagen, Denmark) with 20 ml of bicarbonate-buffered culture medium at 37° C in a humidified chamber with 5% CO₂.

After 20 h in culture, the lung explants were washed extensively with fresh bicarbonate-buffered culture medium, and then transferred to a preweighed dish (35 × 10 mm Petri dish, Falcon; Becton Dickinson Labware, Lincoln Park, NJ) with each slice in 2 ml of bicarbonate-buffered culture medium. The 2-ml medium yielded a fluid layer that was > 2 mm in thickness and thus covered the 0.5 to 1.0-mm-thick lung slices. After gently shaking the culture dish to mix the drugs in the medium, the weight of each explant was determined.

Experimental Protocol

Series 1. The lung explants were prepared, and in this series of experiments the viability of the lung explants over the experimental period was assessed using the following assays:

1. 1. Lactate dehydrogenase (LDH) release. Cell injury in the explants was monitored by LDH release over a period of 7 d with or without LPS challenge (n = 6 each). To validate the LDH technique, a "positive" control group (n = 6) was examined in which the culture medium that contained the gentamicin was not changed.
2. 2. Light microscopy and methacholine test. To examine the contractile response, a methacholine (Pharmacy, The Toronto Hospital, Toronto, Ontario) challenge in the control explants was performed (n = 6). After generation of baseline images of the airways under light microscopy, 10¹⁰ M methacholine was added to the surface of the airways of the explants, and an on-line video camera was used to record the airway constriction.
3. 3. Morphology of lung explants. All lung explants used for LDH experiments were fixed by immersion in 10% buffered neutral formalin (BDH Inc., Toronto, Ontario), and processed using standard histological techniques. A pathologist read the morphological slides in a blinded fashion.

Series 2. Additional lung explants were exposed for 4 h to various doses of LPS (Salmonella typhosa LPS, L-6386; Sigma Chemical Co.) at 1, 10, 100, and 1,000 ng/ml (n = 6 each) to determine the dose dependence of LPS on TNF- production. A dose of 100 ng/ml was then chosen to study the time-dependent response of LPS on TNF- production. Based on these data, a dose of 100 ng/ml of LPS and 4- to 8-h study periods were used in the subsequent experiments. The dose (0.0001, 0.001, 0.01, 0.1, 1, 10, 100 ng/ml) dependence of isoproterenol [()-isoproterenol hydrochloride, I-6504; Sigma] and phenylephrine (L-phenylephrine hydrochloride, P-6126; Sigma) on LPS-induced TNF- production was also examined (n = 6 each). A final concentration of 10 ng/ml for phenylephrine and 1 ng/ml for isoproterenol was chosen for the subsequent experiments.

Series 3. The lung explants were randomly divided into four groups and treated as follows: (1) Controlmedium alone; (2) Phenylephrine (an -adrenergic agonist); (3) Isoproterenol (a -adrenergic agonist); and (4) Forskolin (an adenylate cyclase activator)20 µM final concentration (Forskolin; Sigma). In each group, the lung explants were incubated for 4 h (n = 9) or 8 h (n = 9) at 37° C in 5% CO₂, in the absence or presence of LPS. At the end of the experiments, the tissues were immediately removed from the culture medium, frozen in liquid nitrogen, and the medium supernatants were collected. The tissues and supernatants were stored at 70° C for subsequent measurements of cytokines, cyclic adenosine monophosphate (cAMP), and lipid peroxidation.

Assay for LDH

A colorimetric assay of the Cytotoxicity Detection Kit (LDH) (Boehringer Mannheim GmbH, Mannheim, Germany) was used. LDH is a stable cytoplasmic enzyme present in all cells. It is rapidly released into the cell culture supernatant upon damage of the plasma membrane. LDH activity was determined using an enzymatic test. Briefly, nicotinamide adenine dinucleotide (NAD⁺) was reduced to NADH/ H⁺ by the LDH-catalyzed conversion of lactate and pyruvate. The catalyst (diaphorase) then transferred H/H⁺ from NADH/H⁺ to the tetrazolium salt 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-phenylerazolium chloride) which was reduced to formazan. The color of the formazan dye formed was proportional to the amount of LDH released from cells and was assessed by spectrometry at 500 nm.

Assay for cAMP

A cAMP Enzyme Immunoassay Kit (Cayman Chemical Co., Ann Arbor, MI) was used to determine the cAMP concentrations of the lung homogenates. The tissues were placed into a preweighed Eppendorf vial to determine tissue weight. The tissues were then homogenized in the 50 mM phosphate buffer and 5% trichloroacetic acid (TCA) solution using a sonicator. The homogenates were centrifuged at 1,500 × g for 10 min at 4° C. The pellet was discarded and the supernatant was placed into a clean test tube. The TCA was extracted with water-saturated dimethyl ether. The residual ether was removed by heating the samples to 70° C for 5 min. The measurements were carried out following the instructions in the kit. The cAMP assay is based on the competition between free cAMP and a cAMP tracer (cAMP linked to an acetylcholinesterase molecule) for a limited number of cAMP-specific binding sites of rabbit antiserum. The product of the enzymatic reaction shows a distinct yellow color and absorbs strongly at 412 nm. The intensity of this color, determined spectrophotometrically, was inversely proportional to the amount of free cAMP present in the sample.

Assay for Cytokines

Commercial sandwich enzyme-linked immunosorbent assay (ELISA) kits for rat TNF-, MIP-2, and interleukin-10 (IL-10) (BioSource International, Inc., Camarillo, CA) were used to determine the concentrations of TNF- (detection limit: 15 pg/ml), MIP-2 (detection limit: 10 pg/ml), and the anti-inflammatory cytokine IL-10 (detection limit: 15 pg/ml) in culture medium. The assays recognize natural and recombinant rat TNF-, MIP-2, and IL-10. Cytokine concentrations were normalized per gram of tissue weight.

Assay for Lipid Peroxidation

The commercial colorimetric assay LPO-586 (Bioxytech; OXIS International, Portland, OR) was used for determination of lipid peroxidation by measuring malonaldehyde (MDA) and 4-hydroxyalkenal (4-HNE) in the lung homogenates. The assay is based on the reaction of a chromogenic reagent; the absorbance was measured spectrophotometrically at 586 nm. To set up the LPO-586 assay on lung explant samples, the kinetics of the color development was followed in comparison with that of the MDA and 4-HNE standards provided in the kit, according to the manufacturer's instruction. The absorbance at 586 nm (A₅₈₆) of the samples reached a plateau and remained stable suggesting an absence of non-MDA or non-4-HNE reactivity in the samples.

Statistical Analysis

A one-way analysis of variance (ANOVA) followed by Newman-Keuls test was used for statistical analysis of the data. Differences were considered statistically significant at p < 0.05. Data are presented as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung Explant Viability

The data from several types of experiments demonstrated that the control lung explants remained viable over the study period. First, there was no significant cell injury in explants as monitored by LDH release, which markedly increased only in the explants in which culture medium was unchanged for 4 d (Figure 1). In supporting the sensitivity of LDH measurement with respect to cellular injury, Martin and coworkers (19) demonstrated a direct correlation of the cytotoxic index with LDH release in rat lung explants exposed to hydrogen peroxide. Second, TNF- concentrations in response to LPS increased in a dose- and time-dependent manner (Figure 2), indicating that the cellular machinery necessary for protein synthesis was intact over the study period. Third, when methacholine was added to the surface of the explants, significant airway concentration was observed under light microscopy. Fourth, histological examination demonstrated that at the end of a 96-h culture period, the lung explants had well-preserved alveolar structures with well-maintained cellularity in the alveolar septa, clear cellular borders, and a lack of cellular debris within the alveolar space. Conversely, this integra was significantly altered in explants with bacterial contamination (Figure 3).

View larger version (51K): [in this window] [in a new window]   Figure 1.   Cell injury in the lung explants was monitored by LDH release during a period of 7 d with or without LPS (100 ng/ml) incubation. Contamination = A positive control group was designed by not changing antibiotic-containing culture medium, and bacterial contamination was found 4 d later. *p < 0.05 versus any other groups under the same conditions.

View larger version (15K): [in this window] [in a new window]   Figure 2.   Dose dependence of LPS on TNF- production (left panel ). The lung explants were exposed for 4 h to various doses of LPS. Time-dependent effect of LPS (100 ng/ml) on TNF- production (right panel ).

View larger version (47K): [in this window] [in a new window]   Figure 3.   Rat lung explants cultured in a humidified atmosphere with 5% CO2. (A) Cultured for 96 h (×100). (B) Bacterial contaminated lung explants as a "positive" control, in which culture medium was not changed for 96 h (×100).

TNF-, MIP-2, IL-10, and Lipid Peroxidation in Lung Explants

Under control conditions, the basal TNF- concentration was 0.18 ± 0.04 ng/ml/g tissue at 4 h, and remained consistently low over an 8-h period of study (Figure 4). After LPS stimulation, the concentrations of TNF- increased to 4.13 ± 0.23 ng/ ml/g tissue at 4 h and to 12.12 ± 0.55 ng/ml/g tissue at 8 h (both p < 0.05 versus control, Figure 4). The basal concentration of MIP-2 was 60.0 ± 7.4 ng/ml/g tissue at 4 h, and remained stable at 8 h (Figure 5). After LPS stimulation, MIP-2 concentrations increased in a time-dependent manner to 99.4 ± 16.5 ng/ ml/g tissue (p = NS) at 4 h, and to 165.6 ± 10.3 ng/ml/g tissue (p < 0.05) at 8 h (Figure 5). IL-10 was undetectable, and remained undetectable at 4 and 8 h after LPS challenge. MDA concentrations were 27.6 ± 2.5 µM/g tissue, and 4-HNE concentrations were 34.0 ± 3.0 µM/g tissue at 4 h, and both variables remained stable over an 8-h study period (Figure 6). After LPS stimulation, MDA and 4-HNE did not significantly change at 4 h but dramatically increased to 48.4 ± 17.5 µM/g tissue and 59.7 ± 18.8 µM/g tissue (both p < 0.05) at 8 h, respectively (Figure 6).

View larger version (21K): [in this window] [in a new window]   Figure 4.   (Left panel ) effects of isoproterenol, phenylephrine, and forskolin on the production of TNF- in lung explants without LPS incubation *p < 0.05 versus all other groups at the same time point. (Right panel ) effects of isoproterenol, phenylephrine, and forskolin on the production of TNF- in lung explants with LPS (100 ng/ml) incubation. n = 9 for each group. *p < 0.05 versus LPS alone at the same time point. p < 0.05 versus same treatment at 4 h.

View larger version (61K): [in this window] [in a new window]   Figure 5.   Effects of isoproterenol, phenylephrine, and forskolin on the production of MIP-2 lung explants with LPS incubation at 4 and 8 h. *p < 0.05 versus control at the same time point.

View larger version (28K): [in this window] [in a new window]   Figure 6.   Effects of isoproterenol, phenylephrine, and forskolin on the production of MDA (left panel ) and 4-HNE (right panel ) in lung explants with LPS incubation. *p < 0.05 versus control at 8 h. p < 0.05 versus LPS and phenylephrine + LPS at 8 h, respectively.

Effects of Isoproterenol and Phenylephrine on the Production of TNF- and MIP-2, and Lipid Peroxidation in Lung Explants

Under control conditions, neither isoproterenol nor phenylephrine administration significantly influenced basal levels of TNF- (Figure 4), MIP-2, IL-10, or lipid peroxidation (data not shown) over the 8-h study period.

After administration of LPS, isoproterenol treatment markedly decreased the production of TNF- both at 4 h and 8 h (p < 0.05, Figure 4). The TNF- attenuating effect of isoproterenol was associated with a time-dependent increase in intracellular cAMP concentration at 4 h (2.5 ± 0.5 versus 1.2 ± 0.4 nM/g tissue, p = NS) and at 8 h (4.7 ± 0.9 versus 1.4 ± 0.4 nM/g tissue, p < 0.05) after LPS administration (Figure 7). However, isoproterenol did not influence the LPS-induced evaluation of MIP-2 (Figure 5). Isoproterenol tended to decrease LPS-induced MDA and 4-HNE concentrations although the difference did not reach statistical significance (Figure 6).

View larger version (23K): [in this window] [in a new window]   Figure 7.   Effects of isoproterenol, phenylephrine, and forskolin on the production of cAMP in lung explants without (left panel ) or with LPS (right panel ) incubation. *p < 0.05 versus control at the same time point.

The phenylephrine treatment tended to attenuate the LPS-induced production of TNF- at 8 h but this difference was not statistically significant (9.5 ± 2.0 versus 12.1 ± 0.5 ng/ml/g tissue, p = not significant [NS]) (Figure 4). Phenylephrine did not significantly affect MIP-2 production or the levels of intracellular cAMP and lipid peroxidation (Figures 5-7).

Effects of Adenylyl Cyclase Activation on the Production of TNF-, MIP-2, and Lipid Peroxidation in Lung Explants

Treatment with the adenylyl cyclase activator forskolin tended to decrease the levels of TNF- at 4 h (3.0 ± 1.0 versus 4.2 ± 0.2 ng/ml/g tissue, p = NS), but this decrease did not reach significance until 8 h (4.0 ± 0.7 versus 12.1 ± 1.3 ng/ml/g tissue, p < 0.05) after LPS challenge (Figure 4). Forskolin also tended to decrease MIP-2 production at 4 h (74.7 ± 4.7 versus 99.4 ± 16.5 ng/ml/g tissue) and at 8 h (139.0 ± 5.4 versus 165.6 ± 10.3 ng/ml/g tissue, Figure 5), but these difference did not reach statistical significance. Both MDA and 4-HNE remained at control levels throughout the study period after forskolin treatment (Figure 6). These effects were associated with a tendency to a time-dependent increase in intracellular cAMP concentration at 4 h (3.3 ± 1.2 versus 1.2 ± 0.4 nM/g tissue, p = NS) and a significant increase at 8 h (7.1 ± 2.0 versus 1.4 ± 0.4 nM/g tissue, p < 0.05) after LPS administration (Figure 7).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung Explants as a Model to Evaluate Lung Inflammatory Response

One of the most difficult clinical problems for the intensivist is the management of multiple organ dysfunction syndrome. The lung is often one of the first organs to fail and to express an inflammatory response after an initial insult such as sepsis. The lung may then act as a nidus of inflammation, leading to the systemic release of cytokines and oxygen free radicals, which may eventually result in the development of multiple organ dysfunction. To investigate the mechanisms of acute inflammation in the lung, a number of models including in vivo animals and in vitro cell cultures have been employed. In vivo models have the advantage of more directly mimicking the clinical situation, but determination of basic mechanisms is difficult. In vitro models such as cell culture allow one to focus on specific cells but lack consideration of the lung as a complex organ. Obviously, both types of models are necessary depending upon the hypotheses being examined. However, a model that approximates the in vivo conditions of the lung, and that is able to react to exogenous challenges without influence from other systems and/or organs would be very useful for mechanistic studies. We therefore investigated whether lung explants consisting of the full complement of lung cells in their normal ratios and spatial configuration could serve as an ex vivo lung model to examine the inflammatory response.

The lung explant model was established to quantitatively evaluate whether lung tissues could be directly activated by exposure to LPS, an agent known to cause acute lung injury in humans. Our results demonstrate that LPS induced a local inflammatory response, reflected by a significant increase in TNF-, MIP-2, and lipid peroxidation, in an environment where recruitment of blood-derived inflammatory cells was minimized by perfusing the pulmonary vessels with normal saline before lung excision. Several studies have suggested that the LPS binding protein (LBP), normally present in serum, is required for cells to respond to LPS. Although we did not add serum to the lung explant culture medium, we found that the lung cells exhibited a profound inflammatory response after LPS stimulation, suggesting that some lung cells may also be able to express LBP. This is in accord with data obtained recently by Wong and colleagues (20) who demonstrated that rat pulmonary artery smooth muscle cells stimulated with interleukin-1 could produce LBP.

We did not investigate the specific cell types that produced cytokines and lipid peroxidation in the present study. However, Isowa and associates (2) reported that in response to LPS, rat alveolar epithelial cells produced TNF- through different mechanisms than macrophages. The interaction between alveolar macrophages and other parenchymal cells may also contribute to the production of cytokines and lipid peroxidation observed in the present study.

The major advantage of the present lung explant model is the ability to study cell-cell and cell-matrix interactions under conditions simulating those seen in the intact lung, and also the ability to test multiple interventions from the same preparation under very well-controlled conditions. The application of additional techniques such as in situ hybridization may further enhance the value of using the lung explant model to identify the precise source of any inflammatory mediators generated in this fashion. Pulmonary vessel culture may be another useful model to study the source of the mediators generated in the lung vasculature (21).

Expression of LPS-induced Inflammatory Response in Lung Explants

The expression of cytokines and lipid peroxidation was different in terms of the concentrations produced and the timing of the response. The basal TNF- levels secreted from lung tissues were very low, and near the detection limit of our assay. Interestingly, lung MIP-2 concentrations were approximately 50-fold higher than TNF- in the normal lung explants. Several investigators have recently reported that rat primary alveolar type II cells (2) and bronchial epithelial cells (4) exhibit a high level of MIP-2 expression. Taken together, these data suggest that lung tissues are an important source of MIP-2. After LPS stimulation, both TNF- and MIP-2 concentrations increased dramatically. We sought to investigate the expression in the lung explants of IL-10, a potent anti-inflammatory cytokine that inhibits the synthesis of proinflammatory cytokines such as TNF- and IL-8 by monocytes/macrophages (22), polymorphonuclear leukocytes (PMN) (23), and eosinophils (24). However, IL-10 concentrations were not detectable in the lung explants despite TNF- and MIP-2 levels that were markedly elevated after LPS challenge. This suggests that for a given stimulation, the inflammatory response in lung cells may be different from other host cells such as neutrophils. Whether IL-10 would have been detectable if the pulmonary vasculature had not been flushed with normal saline remains to be tested.

Lipid peroxidation is a well-established mechanism of cellular injury, and is used as an indicator of oxidative stress in cells and tissues. Polyunsaturated fatty acid peroxides generate MDA and 4-HNE upon decomposition. Measurement of MDA and 4-HNE has been used as an indicator of lipid peroxidation (25). We found that both MDA and 4-HNE significantly increased in lung explants 8 h after LPS stimulation. Although lung capillary endothelium appears to be the predominant cellular source of oxidants with lung injury (26), other types of cells, such as the alveolar type II cells (27) and interstitial tissues (28), may contribute to the lung oxidant stress as well. Obviously, small numbers of PMN and macrophages could also contribute to O₂-derived free radical production through the reduced nicotinamide adenine dinucleotide phosphate (NADPH) or other pathways.

Our results indicate that during pulmonary sepsis, inflammatory mediators can be released from the lung tissues without the need to invoke the involvement of de novo systemic host effector cells.

Isoproterenol-inhibited Cytokine Release in LPS-stimulated Lung Explants

Previous studies have established that stimulation of -adrenergic receptors inhibits LPS-induced TNF- production by host effector cells such as monocytes (9). Furthermore, epinephrine and the -receptor agonist salmeterol have been reported to attenuate the release of TNF- after administration of LPS to mice (10), dogs (29), and humans (14). Epinephrine inhibited LPS-induced TNF- production in an isolated perfused rat liver model (30). However, the effects of adrenergic agents on cytokines and lipid peroxidation in the lung have not been well documented. Also, systemic hemodynamic changes induced by the adrenoreceptors in in vivo models could have significantly influenced the results. The present study is the first to examine the role of the adrenoreceptor- cAMP pathway on production of cytokines and lipid peroxidation of lung tissues in culture.

We found that isoproterenol attenuated the LPS-induced production of TNF- without significantly affecting MIP-2. Isoproterenol is classified as a ₁- and ₂-agonist. The mechanism of action of -agonists has been studied in considerable detail. Activation of ₁- or ₂-receptors results in a highly specific activation of adenylate cyclase and an increase in the conversion of ATP to cAMP, mediated by a stimulatory guanine nucleotide-dependent coupling protein analogous to the inhibitory "N" protein (31). cAMP is the major "second messenger"of -receptor activation. We believe that the attenuating effect of isoproterenol on TNF- release is mainly due to ₂-adrenoreceptor action for at least two reasons: first, a number of studies have reported anti-inflammatory properties of ₂-adrenoreceptors under conditions of stress (32), and second, it is known that similar to human alveolar cells (16), -adrenergic receptors are primarily of the ₂-subtype in rat alveolar type II cells (8), a cell type that is responsible for significant TNF- production in lung cells (2).

Sepsis and endotoxemia are associated with hepatocyte injury, which has also been shown to greatly alter the number of adrenergic receptors and to depress the responses to adrenoreceptor agonists (6, 7). We did not measure the number of -adrenergic receptors in the present study. However, the response to isoproterenol in the lung explants was sustained for up to 8 h. We therefore believe that -adrenergic receptors in the lung tissues were not significantly altered by endotoxin during the study period.

Phenylephrine has been classified as a relative ₁-selective adrenoreceptor agonist. The -adrenoreceptor effect does not appear to involve a change of adenylate cyclase activity or cAMP concentration in cells, so that we did not expect levels of TNF- to be affected by phenylephrine. However, phenylephrine at the dose used somewhat lowered the concentrations of TNF- and showed a higher cAMP concentration compared with the group treated with LPS alone, although these differences were not statistically significant. This effect may be caused by the minor -adrenergic properties of phenylephrine (31).

Whereas the LPS-induced TNF- production was markedly attenuated by isoproterenol in the lung explants, both isoproterenol and phenylephrine did not significantly influence MIP-2 concentrations. This suggests that although LPS increased production of both TNF- and MIP-2, the regulatory mechanisms are different. Unlike TNF-, MIP-2 production is not significantly regulated by adrenergic-receptors and intracellular cAMP in lung explants.

Mechanisms of Isoproterenol Inhibition of TNF- and Lipid Peroxidation

Okonogi and coworkers (33) reported inhibition of cAMP accumulation in response to LPS in murine peritoneal macrophages. Our results differthere was an inverse correlation between intracellular cAMP and TNF- level (12, 14, 15). In the present study, isoproterenol attenuated LPS-induced TNF- production associated with an increase in intracellular cAMP levels. Forskolin activates adenylate cyclase in all tissues tested to date. Forskolin caused a 10-fold increase in the activity of adenylate cyclase in rat cerebral cortical membranes with ATP as substrate, and a maximal 35-fold increase in the accumulation of radioactive cAMP in adenine-labeled rat cerebral cortical slices (34). The unique action of forskolin on adenylate cyclase provides a powerful tool for investigating this complex enzyme. In the present study, administration of forskolin increased intracellular cAMP synthesis and also inhibited TNF- production. A large body of evidence suggests that -adrenoreceptor agonists decrease TNF- release through a mechanism mediated by cAMP (12, 26, 29). Severn (12) suggested that a -adrenoreceptor agonist-cAMP pathway may act at a posttranscriptional level, because they found that although epinephrine inhibited TNF- production, similar amounts of TNF- messenger RNA (mRNA) were detected in the presence of epinephrine, in a cell line (THP-1) derived from human leukemia monocytes.

We found that isoproterenol tended to blunt, and forskolin significantly attenuated the LPS-induced lipid peroxidation in lung explants. This phenomenon may be related to the direct action of cAMP, an indirect effect of decreased TNF-, or a direct influence on lipid peroxidation, or a combination of all three. In the present study, we were unable to address the exact mechanisms by which isoproterenol or forskolin attenuated lipid peroxidation. However, it is known that there is a close association between oxidative stress and cytokine production under pathological conditions (35). The attenuating effect of forskolin on lipid peroxidation is most likely caused by the decreased production of TNF- because a significant increase in TNF- production occurred earlier than that of lipid peroxidation. As a consequence, forskolin-cAMP may break a vicious circle in that TNF- can increase the production of lipid peroxidation, which in turn may activate inflammatory cells and induce further release of TNF-.

The present study also has potential clinical implications to diseases other then sepsis. For example, inflammation, increased cytokine production, and oxidant stress are characteristics of acute lung inflammation and asthma. The present study supports the hypothesis that the administration of -adrenergic agonists may not only dilate airway smooth muscle, but may also directly protect lung tissues from inflammation. Early inhibition of TNF- production and lipid peroxidation by these agents may modify the course of various illnesses.