INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Nuclear factor (NF)-B is a ubiquitous transcription factor complex that directs high-level transcription of many cytokines, adhesion molecules, and proinflammatory genes. In unstimulated cells, NF-B is present in an inactive form in the cytoplasm, where it is associated with one of several inhibitory molecules (IBs). Upon cell stimulation, this IB is phosphorylated and degraded by the 26S proteasome. Degradation of the IB unit liberates NF-B and allows its translocation to the nucleus and activation of target genes. Activation of NF-B appears to be a critical and proximal step in the initiation of neutrophilic inflammation (1, 2).

In order to develop a convenient, quantitative method for examining NF-B activation in vivo, we have engineered a line of transgenic mice (referred to as HLL [HIV-LTR/ Luciferase] mice) that carries the proximal 5' human immunodeficiency virus (HIV-1) long terminal repeat (LTR), driving the expression of Photinus luciferase complementary DNA (cDNA) (3). The proximal HIV-LTR is a well-characterized NF-B-responsive promoter (4), containing a TATA box, an enhancer region between nucleotides 82 and 103 with two NF-B motifs, and three Sp1 boxes from nucleotides 46 to 78. In primary cell culture, NF-B activation is absolutely required for transcriptional activity of the proximal HIV-LTR (7, 8). We have shown that luciferase activity in these transgenic mice reflects NF-B activation over time and is useful in evaluating in vivo regulation of inflammation by NF-B (3).

Glucocorticoids are among the most frequently used immunosuppressive and antiinflammatory drugs. Despite their wide use, the mechanism of antiinflammatory action of glucocorticoids is not completely defined. Glucocorticoid effects are mediated through an intracellular receptor, the glucocorticoid receptor, a member of the steroid hormone receptor superfamily (9, 10). Upon hormone binding, the cytoplasmic glucocorticoid receptor can enter the nucleus, dimerize, and bind to specific DNA sequences located in the 5' promoter region of many genes, leading to modulation of gene transcription (11). The activated glucocorticoid receptor may also interact with other transcription factors, such as NF-B and activating protein-1, to regulate gene transcription (12). In vitro studies have shown that some of the antiinflammatory actions of glucocorticoids are due to inhibition of NF-B (13). Although glucocorticoids are useful in the treatment of some inflammatory diseases, they have not proven beneficial in diseases that are characterized by acute neutrophilic inflammation, such as the systemic inflammatory response syndrome and the adult respiratory distress syndrome (ARDS) (16). Because previous studies with glucocorticoids in animal models of lung inflammation/injury have shown variable results, we investigated the ability of glucocorticoids to modulate NF-B and neutrophilic alveolitis in a mouse model of LPS-induced lung inflammation.

Bioluminescence imaging techniques have been used to study light emission from internal organs of small animals, using state-of-the art technology (20, 21). We adopted this technology as a noninvasive means of measuring luciferase production in the lungs of our NF-B reporter-transgenic (HLL) mice. After intraperitoneal (IP) injection of luciferin, the substrate for luciferase, we can detect light emission from mice with an intensified charge-coupled device (ICCD) camera, and can quantify this emission by photon counting over a standardized area of the lungs. In these studies we validated the use of in vivo bioluminescence detection to quantify luciferase expression in the lungs, and used this technique to examine the effects of glucocorticoids on NF-B activation and neutrophilic lung inflammation induced by aerosolized LPS in HLL mice. We correlated NF-B-dependent luciferase activity with total and differential cell counts in lung lavage fluid and with concentration of the NF-B dependent chemokine macrophage inflammatory protein (MIP)-2 in lavage fluid. These studies showed that glucocorticoids do not inhibit NF-B activation, MIP-2 production, or neutrophilic alveolitis in mice after inhalation of LPS, and that at higher doses, glucocorticoids enhance NF-B activation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In Vitro Experiments

Bone marrow-derived macrophages. HLL mice were killed by asphyxiation with CO₂. Cellular material was aspirated from femurs and spun at 400 × g at 4° C for 5 min. Cells were then resuspended in Dulbecco's modified Eagle's medium with 10% fetal bovine serum and 10% L929 cell-conditioned medium (LCM). The cells were allowed to mature into phenotypic macrophages by incubation in the presence of LCM for 5 d before the experiments were done. Total cell counts were determined with a grid hemocytometer, and 2 × 10⁶ cells were plated per well for experiments. For experiments, cells were placed in serum-free medium for 8 h, with the subsequent addition of dexamethasone (or vehicle) for 24 h. LPS was then added and the cells were incubated for 6 h and harvested.

Peritoneal macrophages. Cells were obtained by lavaging the peritoneum of HLL mice at 72 h after instillation of 500 µl of thioglycollate broth. The lavage fluid was then centrifuged at 400 × g, the cell pellet was suspended in 1 ml of RPMI culture medium, and 1.5 × 10⁶ cells per well were plated for experiments. Differential cell counts indicated that 95% of the cells were macrophages. For experiments, cells were placed in serum-free medium for 8 h, with the subsequent addition of dexamethasone (or vehicle) for 24 h. LPS was then added and the cells were incubated for 6 h and harvested.

In Vivo Experiments

Animal model. Transgenic mice weighing 20 to 30 g and expressing photinus luciferase cDNA under the 5' HIV-LTR (C57B6/DBA background) were used for all experiments. Aerosolized Escherichia coli LPS (serotype 055; B5; Sigma Chemical Co., St. Louis, MO) was given by ultrasonic nebulization in a closed chamber for 30 min. Either dexamethasone (Sigma) (0.3 µg/g, 1 µg/g, or 10 µg/g in sterile phosphate buffered saline [PBS]) or vehicle (PBS) was given in divided doses by two intraperitoneal injections, at 24 h and 1 h, respectively, before LPS. After LPS treatment, mice were asphyxiated with CO₂, the tracheas were cannulated, and the lungs were lavaged in situ with sterile, pyrogen-free physiologic saline. Saline was instilled in four 1-ml aliquots and gently withdrawn with a 1-ml tuberculin syringe. Lungs were then removed. One lung was ground in 1 ml of reporter lysis buffer (Promega, Madison, WI) and stored at 20° C for luciferase assays, and the other lung was frozen at 70° C for total RNA extractions and other measurements.

In vivo bioluminescence. Mice were anesthetized with ketamine/ xylazine before imaging, in order to immobilize them for the duration of the integration time of photon counting (3 min). Luciferin (0.75 g/mouse in 200 µl isotonic saline) was injected by IP injection, and mice were imaged with an ICCD camera (C2400-32; Hamamatsu, Bridgewater, NJ). For the duration of photon counting, mice were placed inside a light-tight box. Light emission from the mouse was detected as photon counts with the ICCD camera and image-processing hardware and software (Hamamatsu). A digital false-color photon emission image of the mouse was generated, and photon counts were quantitated over a standard area corresponding to the region of the thorax overlying the midlung zone. Baseline photon counts were obtained before challenge with LPS so that each mouse could be used as its own control.

Lung luciferase measurement. Luciferase measurements on postmortem tissue samples were made by adding 100 µl of freshly reconstituted luciferase assay buffer to 20 µl of the lung homogenate ground in reporter lysis buffer (both buffers from Promega). Luciferase activity was then quantified as relative light units, using a standard luminometer. Luciferase activity was normalized for protein content, which was measured with the Bradford assay (22).

Lung lavage fluid total and differential cell counts. Lung lavage fluid was centrifuged at 400 × g for 10 min to separate cells from supernatant. Supernatant was saved separately and frozen, and pelleted cells were suspended in serum-free RPMI culture medium. Total cell counts were determined on a grid hemocytometer. Differential cell counts were made by staining cytocentrifuge slides with a modified Wright's stain (Diff-Quik; American Scientific Products, San Diego, CA) and counting 400 to 600 cells in cross section.

Enzyme-linked immunosorbent assay for MIP-2. MIP-2 levels were measured with a sandwich-type enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's instructions (R&D Systems, Minneapolis, MN).

Reverse transcription-polymerase chain reaction. Total RNA was purified from lungs by a modification of the method of Chirgwin and associates (23). Extracted RNA was reverse-transcribed at 42° C for 60 min. A polymerase chain reaction (PCR) (30 cycles) was performed in a 25-µl reaction volume with 1× Taq polymerase buffer, 0.2 mM deoxynucleotide triphosphate, 1.5 mM MgCl₂, 0.5 U Taq polymerase, 0.5 µM oligonucleotide primers, and reverse transcription (RT) products. The primers for luciferase were 5'-CTGCCACGCCCGCGTC-3' and 5'-ACGCTTCCATCTTCCAGCGA-3', and the primers for -actin were 5'-CCAGGTCATCACCATTGGCAA-3' and 5'-GAAGCATTT GCGGTGGACCAT-3'.

Statistical Analysis

For comparison among groups, a one-way analysis of variance (ANOVA) was used, with the Tukey-Kramer multiple comparisons test (values of p < 0.05 were considered significant).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Dexamethasone Suppresses LPS-Induced NF-B Activation in Peritoneal and Bone Marrow Macrophages

To examine the effects of dexamethasone treatment on LPS-induced NF-B activation in vitro, we obtained peritoneal and bone marrow-derived macrophages from three HLL transgenic mice. These cells were treated with dexamethasone (1 mM) or vehicle (PBS) for 24 h before being treated with LPS (5 µg/ ml). Cells were harvested at 6 h after LPS treatment, and luciferase measurements were made on cell lysates. LPS treatment significantly increased luciferase expression in both macrophage cell types (Figure 1). Luciferase activity in untreated control peritoneal macrophages was 309 ± 24 (mean ± SEM) relative light units (RLU)/10 µg protein (n = 3), and in bone marrow macrophages was 426 ± 23 RLU/10 µg protein (n = 3). LPS treatment increased luciferase activity in peritoneal macrophages to 570 ± 39 RLU/10 µg protein, and in bone marrow macrophages to 708 ± 39 RLU/10 µg protein (p < 0.05). Treatment with dexamethasone had no effect on luciferase activity in either cell type. However, dexamethasone completely blocked the increase in luciferase activity induced by LPS in both peritoneal and bone marrow macrophages (223 ± 26 RLU/10 µg protein and 248 ± 28 RLU/10 µg protein, respectively) (p < 0.01 versus the corresponding LPS-treated sample).

View larger version (22K): [in this window] [in a new window]   Figure 1.   Luciferase activity in cell lysates of peritoneal macrophages and bone marrow-derived macrophages from HLL mice measured with the luminometer. Cells were treated in culture with dexamethasone (dex) (1 mM) or vehicle (PBS) for 24 h, followed by 5 µg/ml of endotoxin (LPS) or vehicle. Cells were then incubated for 6 h. Four treatment groups were studied: vehicle alone (contol), dexamethasone alone (dex), vehicle followed by LPS (LPS), and dexamethasone followed by LPS (dex + LPS). Values are mean ± SEM of three separate experiments. (*p < 0.05 versus with other groups).

Development of a Model for Lung Inflammation in HLL Transgenic Reporter Mice Following Inhalation of LPS

To develop a model of neutrophilic lung inflammation following inhalation of LPS, we initially performed time-course and dose-response experiments (Figure 2). Figure 2A shows that neutrophil accumulation in the airspaces began to occur by 3 h after inhalation of LPS (30 min of exposure to a 1 mg/ml solution). The threshold LPS concentration for induction of significant neutrophilic influx into the lungs at 4 h was 1 mg/ml, as shown in dose-response studies (Figure 2B). For subsequent studies, we exposed mice to a 1 mg/ml LPS solution aerosolized over a period of 30 min, and killed the mice at 4 h after inhalation of LPS for counting of lung lavage neutrophils.

View larger version (20K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Figure 2.   Time course (A) and dose-response (B) curves for total cells and total neutrophils (PMN) in lung lavage fluid after treatment with inhaled LPS. (A) Total cells and total PMN in lung lavage fluid at baseline and up to 4 h after inhalation of a 1 mg/ml solution of LPS. In untreated mice, > 98% of lavage fluid cells are macrophages. Standard lung lavage was performed to a total volume of 3 ml; each time point represents the mean of two mice. (B) Dose- response curves for total cells and total PMN influx into the airways 4 h after inhalation of a 0.1 to 3 mg/ml solution of LPS. Values are mean ± SEM; n = 3 mice at each dose, 12 mice in total.

To examine expression of luciferase messenger RNA (mRNA) in lungs after inhalation of LPS, we performed RT- PCR, using RNA extracted from whole-lung homogenates of HLL mice. Luciferase mRNA expression was not identified at 0 to 30 min after exposure to aerosolized LPS (Figure 3); however, luciferase mRNA expression was prominently upregulated by 2.5 to 3 h after inhalation of LPS. Therefore, the timing of expression of luciferase mRNA correlated with the onset of neutrophilic alveolitis in this model.

View larger version (37K): [in this window] [in a new window]   Figure 3.   RT-PCR of whole-lung RNA showing the time course for luciferase mRNA expression after inhalation of LPS (1 mg/ml solution) in HLL mice (top band). -Actin mRNA detection is shown as a control.

Figure 4 shows data from experiments done with in vivo bioluminescence detection to quantitate lung luciferase activity after inhalation of LPS. Initially, basal luciferase activity was measured in untreated mice after IP injection of luciferin (0.75 mg). Mice were imaged immediately after injection and every 30 min thereafter until return of photon counts to baseline (Figure 4A). Mice were then treated with aerosolized LPS (1 mg/ml solution for 30 min). Three hours after inhalation of LPS, mice were reinjected with luciferin and again imaged immediately after injection and every 30 min thereafter. Photon counts over a standardized region of the chest in untreated mice peaked at 36,675 (± 4,316) at 1 h after luciferin injection. Following treatment with inhaled LPS, photon counts increased to a mean of 97,894 (± 7,552) (p < 0.001) (Figure 4A) at 1 h after luciferin injection. Photon emission from the lungs returned to baseline by 120 min after luciferin injection. This primarily represents a lack of bioavailable substrate, since reinjecting luciferin resulted in another peak of detected luciferase activity in mice treated with LPS (not shown). After bioluminescence imaging, mouse lungs were harvested and luciferase activity was measured in lung homogenates with standard luciferase activity assays. These values were closely correlated with photon emission from the lungs as determined by bioluminescence in each animal (R² = 0.84) (Figure 4B). These studies indicate that bioluminescence is a powerful methodology for detecting luciferase activity in vivo.

View larger version (20K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Figure 4.   Bioluminescence of HLL mice before and after inhalation of LPS (1 mg/ml solution nebulized over a period of 30 min). (A) Graph showing the time course for detection of bioluminescence from lungs after IP injection of luciferin (0.75 mg). Photon emission was quantified by computerized counting of detected photon emission in a 3-min scan over a standardized area of the thorax. Mice were scanned every 30 min after IP injection of luciferin until return to baseline (circles). Subsequently, these mice were treated with inhaled LPS. Three hours later, luciferin was reinjected and mice were scanned every 30 min until return to baseline (squares), values are mean ± SEM, (n = 6 mice, p < 0.05 for comparison between baseline and post-LPS at 30 and 60 min after luciferin). (B) Correlation between peak lung photon counts as detected by bioluminescence and luciferase activity in lung homogenates of HLL mice after inhalation of LPS. Lungs were harvested after bioluminescence imaging and luciferase assays were done on lung homogenates. (R2 = .84).

Dexamethasone Does Not Suppress in vivo NF-B Activation, MIP-2 Production, or Neutrophilic Alveolitis in Mice Treated with Aerosolized LPS

To study the in vivo effects of glucocorticoids on neutrophilic lung inflammation after LPS treatment, we treated HLL mice with IP injections of dexamethasone (0.3, 1, or 10 µg/g) or vehicle in two divided doses before exposure to LPS (24 h and 1 h respectively, before LPS). We utilized large doses of dexamethasone to ensure maximal glucocorticosteroid receptor binding and steroid effect. Each mouse was used as its own control, and baseline photon emission from the lungs was assessed before LPS administration. In these experiments, luciferin was injected and a single scan was performed 30 min after its injection. Subsequent to the baseline scan, the second dose of dexamethasone was administered and LPS aerosolization was begun 1 h later. Three hours after inhalation of LPS, a second injection of luciferin was given and the mice were scanned 30 min later. After scanning, mice were killed, and their lungs were lavaged in situ, removed, and homogenized for luciferase activity measurements. Figure 5A shows that treatment with 0.3 µg/g and 1.0 µg/g of dexamethasone failed to inhibit luciferase expression in lungs of these mice as measured with bioluminescence. At the highest dose of dexamethasone (10 µg/g), lung luciferase activity was significantly increased over that of mice treated with vehicle followed by LPS. This increase in bioluminescence is easily seen on the computer-enhanced image, which uses an artificial color scale to identify the intensity of detected photon emission (white is the highest intensity) (Figure 5B). The mouse to the left in Figure 5B shows baseline luciferase activity in the lungs. The mouse was imaged 30 min after IP luciferin and before LPS inhalation. The middle picture is of the same mouse imaged 4 h after inhalation of LPS (30 min after IP luciferin). The mouse on the right was treated with inhaled LPS after dexamethasone injection (10 µg/g) and was imaged 4 h after inhalation of LPS. Increased luciferase activity in mice treated with the highest doses of dexamethasone was further confirmed by measuring luciferase activity in lung homogenates (Figure 6). By this assay, treatment with dexamethasone at doses of 1 µg/g and 10 µg/g significantly increased lung luciferase activity as compared with that of mice treated with LPS alone (p < 0.05). These studies show that systemic treatment with dexamethasone does not inhibit LPS-induced luciferase expression in the lungs. Surprisingly, at very high doses of dexamethasone, LPS-induced luciferase activity is augmented.

View larger version (15K): [in this window] [in a new window]   View larger version (71K): [in this window] [in a new window]   Figure 5.   (A) Lung bioluminescence in HLL mice treated with IP dexamethasone (0.3 to 10 µg/g) or vehicle before inhalation of LPS (1 mg/ ml). Luciferin was injected before and 3 h after inhalation of LPS. Lung bioluminescence was measured 30 min after IP luciferin injection by measuring photon counts over a standardized area of the thorax. Photon counts are reported as peak-to-baseline differences, so that each mouse served as its own control. Baseline bioluminescence (prior to LPS treatment) was similar in mice treated with dexamethasone or vehicle. Lung bioluminescence after LPS treatment was increased at the highest dexamethasone dose as compared with LPS alone (n = 5 in each group, *p < 0.05 compared with LPS alone). (B) Photon emission from the lungs after IP luciferin at baseline (left panel ), in the same mouse 3.5 h after treatment with LPS (LPS, middle panel ), and in a mouse treated with dexamethasone followed by LPS treatment (Dex [10 µg/g] + LPS, right panel).

View larger version (14K): [in this window] [in a new window]   Figure 6.   Lung luciferase activity as measured in lung homogenates from HLL mice treated with IP dexamethasone (0.3 to 10 µg/g) or vehicle before inhalation of LPS (same mice as reported in Figure 5). Lungs were harvested after bioluminescence imaging and luciferase activity assays were performed on lung homogenates (n = 5 in each group, *p < 0.05 compared with LPS alone).

Dexamethasone treatment did not suppress MIP-2 levels or inhibit neutrophilic alveolitis in lung lavage fluid at 4 h after inhalation of LPS (Figure 7). No differences were found in MIP-2 levels as measured by ELISA or in neutrophils as measured by total and differential cell counts for mice treated with dexamethasone prior to LPS as compared with mice treated with LPS alone. Substantial levels of MIP-2 and large numbers of neutrophils were found in all groups, indicating that inhaled LPS induced intense neutrophilic inflammation.

View larger version (17K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Figure 7.   (A) MIP-2 levels and (B) neutrophil counts in lung lavage fluid of mice treated with IP dexamethasone (0.3 to 10 µg/g) or vehicle before inhalation of LPS (1 mg/ml). There were no significant intergroup differences in MIP-2 levels or neutrophil counts for any of the treatment groups (n = 5 in each group).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our studies show that glucocorticoids inhibit NF-B activation in vitro, and are consistent with several previous reports that glucocorticoids inhibit NF-B activation in cultured cells (24). In contrast, we showed, with bioluminescence as a surrogate marker of NF-B activation in a transgenic model of lung inflammation induced by inhaled LPS, that glucocorticoids did not suppress activation of NF-B in vivo, and at very high doses, glucocorticoids upregulated activation of NF-B. We have previously shown that our HLL transgenic reporter mice represent a good model for investigating NF-B-dependent responses by demonstrating that: (1) luciferase activity in these mice correlates with activation of NF-B as detected by gel mobility shift assays; (2) luciferase activity correlates with mRNA expression of NF-B-dependent cytokines; and (3) specifically blocking NF-B activation in these mice, by expressing a trans-dominant inhibitor of NF-B in the liver, blocks LPS-induced luciferase activity (3). These studies emphasize the importance of animal model research in the study of inflammation, as opposed to extrapolation from in vitro studies with cultured cells.

Bioluminescence refers to the process of visible light emission in living organisms mediated by an enzyme catalyst. Luciferase catalyzes a chemiluminescent reaction that produces light without the need for optical excitation. Bioluminescence at the emission wavelength of luciferin activated by firefly luciferase (560 nm) can be imaged as deep as several centimeters within tissue, which allows at least organ-level resolution. Bioluminescence imaging techniques are being developed to noninvasively study disease processes such as infection and tumor growth in intact animals (27). A mouse model of Salmonella infection, using a strain of S. typhimurium with a plasmid conferring constitutive expression of bacterial luciferase, has been developed to study the progression of Salmonella infection in the gastrointestinal tract and its response to antibiotic drugs (20, 21). In the present study, we used bioluminescence detection to quantify NF-B activation over time in lungs of intact HLL mice challenged with aerosolized LPS. After LPS treatment, we were able to demonstrate an increase in photon emission over the lung region, as measured by bioluminescence, that correlated with increased luciferase activity as measured with standard luciferase assays done on lung homogenate. This suggests that bioluminescence is a powerful methodology for detecting NF-B-dependent gene expression in reporter mice. This technique can reduce the number of animals required for experimentation, since each animal can be used as its own control, which minimizes the effects of biologic variation. Furthermore, multiple measurements of NF-B activation can be made in the same animal over time. We believe that this is a useful experimental methodology to investigate potential antiinflammatory treatments in vivo.

Previous studies involving animal models of lung inflammation induced by LPS and ozone have shown variable responses to dexamethasone treatment. In a rat model of pulmonary inflammation induced by intratracheal LPS (1 mg/kg), dexamethasone treatment (2 to 4 mg/kg given by IP injection) resulted in a 74% reduction in lung lavage fluid levels of tumor necrosis factor (TNF)-, yet MIP-2 accumulation was unchanged (30). In another study, dexamethasone was given in doses of up to 30 mg/kg in a rat model of neutrophilic lung inflammation induced by intratracheal LPS (1 mg/kg). Three hours after LPS administration, dexamethasone had reduced the neutrophil influx, but MIP-2 levels in lung lavage fluid were unaffected (31). In contrast, dexamethasone significantly suppressed MIP-2 synthesis by LPS-stimulated rat alveolar macrophages in vitro (31). In models of systemic endotoxemia, treatment with dexamethasone has resulted in different responses in mice and guinea pigs (32). Treatment of mice with IP injection of dexamethasone at 3 mg/kg, 10 mg/kg, and 25 mg/kg at 1 h before LPS injection (20 mg/kg) reduced serum TNF- levels by more than 80%. In contrast, treatment of guinea pigs with dexamethasone at up to 25 mg/kg did not result in significant reduction in TNF- levels. Interestingly, in this study, dexamethasone treatment did not block the accumulation of neutrophils in lung lavage fluid in either species at 6 h (32). On the basis of these studies, it is clear that glucocorticoids do not consistently inhibit parameters of inflammation induced by bacterial LPS in animals, possibly because NF-B is activated in these animals despite glucocorticoid treatment. Dexamethasone treatment may suppress NF-B activation induced by stimuli other than LPS. Rats exposed to ozone (3 ppm for 6 h) had decreased expression of the NF-B-dependent chemokine CINC and decreased lung NF-B activation after IP injection of dexamethasone at a dose of 3 mg/kg (33). Dexamethasone also diminished the number of neutrophils recovered in the lung lavage fluid in this model.

In a human study, 23 normal subjects were given LPS alone or after infusion of 500 mg of hydrocortisone at various time points before administration of LPS (34). As compared with subjects treated with LPS alone, hydrocortisone treatment 6 h before LPS or simultaneously with LPS resulted in suppression of serum TNF- levels, but levels of interleukin (IL)-6 were unaffected. Subjects who received LPS 12 h or 144 h after hydrocortisone infusion generated significantly greater circulating levels of both IL-6 and TNF- than did the group treated with LPS alone. Clinical trials in which methylprednisolone has been used in the treatment of ARDS or septic shock have yielded disappointing results (16). In a randomized, prospective, double-blind, placebo-controlled trial of high-dose methylprednisolone given over a period of 24 h to patients early in the course of established ARDS, treatment with this glucocorticoid had no effect on survival or reversal of respiratory failure (16, 17). Similarly, well-designed studies of high-dose glucocorticoids have shown no benefit in the treatment of severe sepsis and septic shock (18, 19).

Since glucocorticosteroids have potent antiinflammatory activity in a variety of settings, the explanation for the lack of their efficacy in suppressing LPS-induced inflammation in vivo is uncertain. One possibility is that glucocorticoids can induce the production of macrophage inhibitory factor, which can overcome glucocorticoid inhibition of cytokine production by macrophages, and has been shown to block the protective effects of glucocorticoids on lethality in a mouse model of endotoxic shock (35, 36). Another potential effect of glucocorticoids that may explain the finding that steroids are ineffective in neutrophilic inflammation is their ability to prolong neutrophil survival. In an in vitro study, dexamethasone was shown to be a potent inhibitor of neutrophil apoptosis (37).

Our studies indicate that the effects of glucocorticoids in intact mice are different from those observed in cultured cells. The actions of glucocorticoids on inflammatory processes are complex and certainly involve factors other than NF-B; however, the lack of effectiveness of glucocorticoids in inhibiting NF-B and blocking LPS-induced neutrophil recruitment to the lungs may explain the lack of efficacy of these drugs in disease processes that involve neutrophil-mediated tissue injury. This study points out that since it is not always possible to extrapolate in vitro data to an in vivo setting, it is necessary to study the regulation of inflammation in intact animals.