Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Lipopolysaccharide (LPS) is a major pathogenic factor in gram-negative sepsis, which is characterized by shock, coagulopathy, and multiorgan dysfunction. In response to systemic LPS exposure, proinflammatory cytokines such as tumor necrosis factor (TNF)-, interleukin (IL)-1, and interferon- are produced by the host, which have been shown to either directly or indirectly mediate many of the hemodynamic and inflammatory changes and organ damage in sepsis. Animal models of septic shock indicated that apoptosis, an active cellular process of cell death under genetic control, contributed to primary organ damage. Systemic LPS administration in mice resulted in apoptotic cell death in the endothelium of several tissues, including intestine and lung (1). In vivo blocking of TNF- by anti-TNF antibody (1) or TNF-binding proteins (2) attenuated LPS-induced apoptotic rates, suggesting that TNF- is the primary mediator in endothelial apoptotic cell death induced by systemic administered LPS.

The respiratory system is continuously exposed to low levels of LPS, which is ubiquitously present as a contaminant on airborne particles, including air pollution (4), organic dusts (5), and cigarette smoke (6). Exposure to high LPS levelsfor example, agricultural workers in contact with organic dusts, or heavy smokersis known to provoke acute lung inflammation, partly initiated via the early endogenous induction of IL-1 and TNF- in the lung. These cytokines are thought to contribute to the pathogenesis of acute inflammation by inducing the expression of endothelial leukocyte adhesion molecules and chemokines, consequently leading to recruitment of neutrophils into alveoli. Neutrophils play a prominent role in the host defense against pathogens, but are also considered to be responsible for pulmonary injury, manifested by increased lung vascular permeability, edema, and cell death (7). Neutrophil presence was suggested to induce apoptotic cell death in primary human bronchial epithelial cells (8). In vitro studies directed at resolving tissue injury caused by LPS have indicated that LPS can directly trigger pulmonary cells to undergo apoptosis. Bingisser and coworkers showed that LPS induced apoptosis in human alveolar macrophages dose-dependently (9). In addition, LPS was reported to cause apoptotic cell death in primary cultures of isolated bovine and sheep pulmonary artery endothelial cells, and in a bovine pulmonary artery endothelial cell line (10, 11).

In view of the continuous exposure of the lungs to LPS, we studied whether local exposure to LPS in vivo results in apoptotic cell death in lungs. To this end, mice were instilled intratracheally with LPS, and apoptotic cell death in lung tissue was assessed via ligation-mediated (LM) polymerase chain reaction (PCR) and terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end label staining. In this study we show for the first time that local LPS exposure results in apoptotic cell death in bronchial epithelial cells at early time points after exposure. In addition, we investigated the role of infiltrating neutrophils and TNF- in LPS-induced apoptosis and demonstrate that neither neutrophils nor TNF- mediate apoptotic cell death in lungs at early time points.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

Male Swiss mice (30 to 40 g) were obtained from Charles River Breeding Laboratories (Heidelberg, Germany). Animals were housed individually in standard laboratory cages and allowed food and water ad libitum throughout the experiments. The studies were carried out under a protocol approved by the Institutional Animal Care Committee of Maastricht University.

Male TNF/lymphotoxin (LT)- double knockout (/) and wild type (+/+) mice were obtained from Dr. H. P. Eugster (University Hospital Zurich, Department of Internal Medicine, Zurich, Switzerland [12]). These mutant mice were bred at the Laboratory of Molecular Biology, University of Gent (Gent, Belgium) according to the Belgian and European Union guidelines for the use and care of laboratory animals. The experiments were performed in Gent under a protocol approved by the Institutional Animal Care Committee of the University of Gent.

Experimental Protocol

Intratracheal (IT) instillation technique was performed according to Starcher and Williams (13). A control experiment was performed in which bromothymol blue dissolved in 50 µl 0.9% NaCl was instilled in mice to check distribution of solution in the lung. Macroscopic and microscopic analysis demonstrated that blue marker dye had spread throughout the whole lung. Mice (n = 6 per group) were anesthetized by intraperitoneal (IP) injection of 3 mg/kg xylazine (Sedamun, Auv Cuijk, The Netherlands) and 75 mg/kg ketamin (Nimatek, Auv Cuijk, The Netherlands). LPS (Escherichia coli, serotype O55:B5; Sigma, St. Louis, MO) dissolved in 50 µl sterile 0.9% NaCl was instilled intratracheally via a canule, followed by 0.15 ml of air. The dose of LPS used was 5 µg/ mouse, which corresponds to the LPS dose delivered to the human lung by smoking approximately 25 cigarettes (6). No signs of an overall toxic effect of the dose of LPS used were observed in the trachea, airways, and lungs, either in this study or in those by others (14, 15). Sham mice were instilled intratracheally with 50 µl LPS-free sterile 0.9% NaCl, whereas control mice received no treatment. After IT treatment, the mice were kept in an upright position for 10 min to allow the fluid to spread throughout the lungs. Mice were killed at 2, 4, 8, 24, or 72 h after instillation, and blood was collected by heart puncture. After thoracotomy, lungs were prepared for bronchoalveolar lavage (BAL) or DNA extraction, RNA extraction, myeloperoxidase (MPO) analysis, and light microscopy.

TNF- neutralization in vivo was accomplished by using antimurine TNF- monoclonal antibody (mAb) TN3, a complementarity-determining regions-grafted murine immunoglobulin (Ig) G2a (a generous gift from Celltech, Slough, UK), which was shown to have neutralizing capacities in vivo (16). IgG class-matched control antibody (IgG2a mAb 5D7 [17]) was used as control reagent. Administration of these reagents to mice not subjected to IT LPS instillation did not influence any of the parameters investigated (data not shown). Mice treated with anti-TNF- mAb TN3 and control IgG2a mAb 5D7 received a single IP injection of 1 mg antibody in 1 ml LPS-free sterile 0.9% NaCl at 2 h before IT LPS instillation. At 4 or 24 h after LPS, mice were killed and the lungs were prepared for DNA extraction and histology.

Analysis of DNA Fragmentation

Genomic DNA was isolated from snap-frozen lung tissue of the right lung using a DNA purification kit (Wizard; Promega, Madison, WI). DNA concentration and purity were ascertained by electrophoresis on an ethidium bromide-stained 0.8% agarose gel followed by ultraviolet (UV) illumination and UV spectrophotometric analysis at wavelengths of 260 and 280 nm. DNA fragmentation in lung tissue was investigated with a commercially available LM-PCR assay kit (Apoalert; Clontech, Palo Alto, CA), enabling sensitive and semiquantitative measurement of the extent of apoptosis. Dephosphorylated adaptors (12- and 24-mer) were ligated to 500 ng of DNA with T4 DNA ligase for 16 h at 16°C. The 24-mer adaptor also served as primer in the LM-PCR, in which 25 ng of ligated DNA was amplified under the following conditions: hot start (72°C for 8 min) with Taq polymerase (Perkin-Elmer/Cetus, Emeryville, CA) added after 3 min, 22 cycles (94°C for 15 s, 72°C for 3 min), and postcycling (72°C for 15 min). Amplified DNA was separated by electrophoresis on a 1.2% agarose gel containing ethidium bromide and visualized by UV illumination. Intensity of the 360-base pair (bp) band was digitally analyzed (SigmaGel 1.0; SPSS, Chicago, IL) and expressed in arbitrary units.

Histology

After thoracotomy, the left lung was inflated with 10% phosphate-buffered formalin (pH 7.4) at a pressure of 20 cm H₂O through the trachea for 15 min and subsequently fixed in 10% phosphate-buffered formalin for 24 h. After paraffin embedding, 4 µm sections were cut and stained with hematoxylin and eosin (H&E) for histologic analysis.

DNA Nick-End Labeling of Tissue Sections

Histologic aspects of apoptosis were studied by terminal deoxynucleotidyl transferase (TdT)-mediated deoxyuridine triphosphate (dUTP) nick-end labeling (TUNEL), performed according to the protocol described by Gavrieli and colleagues (18) with slight modifications. Briefly, 4-µm paraffin sections were deparaffinized and pretreated with 20 µg/ml Proteinase K (Sigma). Endogenous peroxidase was quenched with 0.5% hydrogen peroxide (H₂O₂). Sections were incubated for 1 h at 37°C in a solution consisting of 25 mM Tris, 200 mM sodium cacodylate, 0.25 mg/ml bovine serum albumin (Sigma), 1.5 mM cobalt chloride, 40 µM digoxigenin-11-dUTP (Boehringer Mannheim, Mannheim, Germany), and 30 U/ml TdT (Boehringer Mannheim), pH 6.6. The labeling reaction was terminated by transferring the sections into 300 mM sodium chloride/30 mM sodium citrate for 30 min at 37°C. Incorporated digoxigenin-11-dUTP was demonstrated with peroxidase-conjugated sheep antidigoxigenin antibody (Boehringer Mannheim). The labeled antibody was visualized with diaminobenzidine (Boehringer Ingelheim, Heidelberg, Germany). Sections were lightly counterstained with hematoxylin and mounted. Negative controls included TdT-free labeling mixture. Thymus sections from the same animal served as positive controls.

Determination of MPO

MPO was isolated from snap-frozen lung tissue of the right lung as described by Kuebler and coworkers (19). Enzymatic detection of MPO was performed in a 96-well plate (Greiner, Nurtingen, Germany) as previously described (20). Briefly, assay mixtures consisted of 40 µl 0.75 mM H₂O₂ in 80 mM phosphate-buffered saline (PBS) (pH 5.4) and 40 µl sample diluted in 50 mM PBS (pH 6.0) and 0.5% hexa-1,6-bis-decyltrimethylammonium bromide (Sigma). The reaction was initiated by adding 20 µl of 8 mM 3,3',5,5'-tetramethylbenzidine (TMB) (Boehringer Mannheim) in dimethyl sulfoxide (Sigma) and stopped after 15 min by adding 100 µl/well 1 M H₂SO₄. Subsequently, optical density was determined at 450 nm. All samples were assayed in triplicate. MPO activity was calculated per milligram of lung tissue and corrected for wet/dry ratios. A titration curve of horseradish peroxidase was used for the calculation of MPO activity, which is expressed in arbitrary units.

Reverse Transcription/PCR

Total RNA was isolated from snap-frozen lung tissue of the right lung using a commercially available kit (SV Total RNA Isolation System; Promega). Total RNA concentration and purity were ascertained by electrophoresis on an ethidium bromide-stained 0.8% agarose gel followed by UV illumination and UV spectrophotometric analysis at wavelengths of 260 and 280 nm. The quantity of 5 µg of total RNA was reverse transcribed in a 20-µl volume using oligo(dT) primers and Moloney murine leukemia virus reverse transcriptase (RT) (Life Technologies, Paisley, UK) according to the supplier's recommendations. PCR for murine TNF- and -actin was performed in a 25-µl reaction volume containing 100 µM of each deoxynucleotide triphosphate, 200 nM sequence-specific primers, and 0.5 U Taq DNA polymerase (Perkin-Elmer/Cetus) during 35 cycles under the following conditions: 95°C for 30 s, 60°C for 45 s, and 72°C for 30 s. PCR primers used in RT-PCR for TNF- (307 bp) and -actin (348 bp) were designed as previously described (20). Amplified PCR products were analyzed on a 1.2% agarose gel containing ethidium bromide and visualized by UV illumination. A mock PCR (without complementary DNA [cDNA]) was included to exclude contamination.

BAL

The trachea was exposed with a midline incision and cannulated with a steel catheter. The lungs were lavaged six times with sterile 0.9% NaCl at a volume of 1.5 ml/wash. The average fluid recovery was greater than 90%. The BAL fluid (BALF) was centrifuged at 1,500 rpm for 10 min at 4°C and the supernatants were stored at 20°C until analysis.

Enzyme-Linked Immunosorbent Assay for Murine TNF-

TNF- concentrations in BALF and plasma were determined using a specific enzyme-linked immunosorbent assay (ELISA) as described by Dentener and colleagues (21). Briefly, 96-well immuno maxisorp plates (Nunc, Roskilde, Denmark) were coated with 5 µg/ml hamster antimurine TNF- mAb (TN3; kindly provided by Celltech). Murine recombinant (r) TNF- (kindly provided by Genentech, San Francisco, CA) was used for standard titration curves. Polyclonal rabbit antimurine TNF- (Genzyme, Cambridge, MA) was followed by peroxidase-conjugated goat antirabbit IgG (Jackson, West Grove, PA) and TMB was used as substrate. The ELISA had a lower detection limit of 50 pg/ml.

Statistical Analysis

Data are expressed as means ± standard error of the mean (SEM). Statistical analysis was performed by means of Mann-Whitney U test and probability values below 0.05 were considered statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Induction of Apoptosis in Lungs after Intratracheal Instillation of LPS

In the present study, apoptotic cell death after IT LPS instillation was determined by ligation-mediated PCR on DNA isolated from lung tissue. As shown in Figure 1, local LPS administration resulted in induction of DNA fragmentation ladders of approximately 180-bp multimers, which are considered to be characteristic of apoptotic cell death. DNA laddering was already evident 2 h after LPS treatment and peaked at 24 h (intensity: 360-bp band 2.7 U [2 h] versus 24.7 U [24 h]). At 72 h after exposure, DNA laddering was strongly diminished. Little or no DNA laddering was detected in lung tissue from saline-treated and control mice. These results indicate that IT LPS instillation induced apoptosis in the lung, which was already evident at early time points.

View larger version (57K): [in this window] [in a new window]   Figure 1.   Induction of DNA laddering in lung tissue by IT LPS instillation. Mice received 5 µg of LPS intratracheally and were killed at 2, 4, 8, 24, and 72 h after exposure. Control mice (C) and saline-treated mice (S) were killed after 24 h. DNA isolated from snap-frozen lung samples was amplified by LM-PCR and separated by electrophoresis on a 1.2% agarose gel containing ethidium bromide. Intensity of the 360-bp band was digitally analyzed to quantify apoptosis (expressed in arbitrary units). M, molecular weight markers (range: 100 to 2,000 bp). One representative experiment from a series of three is shown.

Involvement of Neutrophils in LPS-Induced Early Apoptosis in Lung Tissue

Several studies have reported that infiltrating neutrophils, recruited to the lungs after IT LPS instillation, are eliminated by apoptosis during the resolution of acute pulmonary inflammation (22, 23). Further, a recent paper suggested that neutrophil presence could induce apoptotic cell death in human bronchial epithelial cells (8). Therefore, we examined the role of neutrophils in the observed LPS-induced early apoptosis in the lung by comparing the kinetics of pulmonary apoptosis and pulmonary neutrophil influx after IT LPS instillation. Histologic assessment revealed that local LPS challenge resulted in a time-dependent neutrophil accumulation (Figure 2). Neutrophils were absent at 2 and 4 h after exposure, but presence of neutrophils in the alveolar spaces was evident from 8 h after LPS treatment and peaked at 24 h. At 72 h after exposure, neutrophil accumulation was diminished. In saline-treated and control mice, neither changes in lung histology nor a significant neutrophil influx were observed.

View larger version (72K): [in this window] [in a new window]   Figure 2.   Neutrophil infiltration in lung tissue after IT LPS instillation. Lung tissue of LPS-treated and saline-treated mice was fixed in 10% formalin, processed according to standard procedure, and stained with H&E. (A) Saline-treated, 24 h. (B) LPS-treated, 4 h. (C) LPS-treated, 24 h. Bar = 60 µm. Note the accumulation of neutrophils in the alveolar spaces at 24 h after exposure, which is absent at 4 h after LPS instillation and in saline-treated controls.

Next, MPO activity in lung homogenates was measured to quantify the relative neutrophil accumulation in the lung. In line with our histologic assessment of neutrophil infiltration, MPO activity was not detected in lung homogenates from LPS-treated mice at 2 and 4 h after exposure, but MPO activity increased to 17, 44, and 26 units at 8, 24, and 72 h after LPS instillation, respectively (Figure 3). MPO activity was not demonstrated in lung tissue of saline-treated or control mice.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Quantification of pulmonary neutrophil infiltration after IT LPS instillation. Mice received 5 µg of LPS intratracheally and were killed 2, 4, 8, 24, and 72 h after instillation. Saline-treated mice (S) were killed 24 h after instillation. MPO activity was measured in lung homogenates to quantify the amount of infiltrated neutrophils. Data are expressed in arbitrary units (means ± SEM). *P < 0.05 versus saline.

Comparison of the kinetics of apoptosis (Figure 1) and neutrophil influx (Figure 3) in the lung revealed that apoptosis was evident from 2 h after instillation, whereas neutrophil infiltration did not occur until 8 h after LPS challenge, demonstrating that apoptosis clearly preceded the influx of neutrophils. These data show that neutrophils did not contribute to apoptotic cell death at early time points.

Apoptotic Cell Death in Bronchial Epithelial Cells after IT LPS Instillation

Next, we performed TUNEL on lung sections to determine which pulmonary cells undergo apoptosis after IT LPS treatment. As shown in Figure 4, TUNEL demonstrated intense brown nuclear staining in part of the bronchial epithelial cells from 4 h after LPS exposure, indicating apoptotic cell death. Morphologic examination demonstrated characteristic condensed chromatin in the nuclei of positive-stained bronchial epithelial cells, confirming their apoptotic state. In contrast, other pulmonary cells, e.g., alveolar type II and I cells, smooth-muscle cells, endothelial cells, and alveolar macrophages, did not display positive staining at any time point after local LPS challenge. From 24 h after exposure onward, mainly infiltrated neutrophils displayed positive staining and only few bronchial epithelial cells did so, whereas at 72 h after exposure positive nuclear staining was demonstrated only in infiltrated neutrophils. These results demonstrated early apoptosis in bronchial epithelial cells after IT LPS exposure followed by apoptosis in infiltrated neutrophils at later time points.

View larger version (70K): [in this window] [in a new window]   Figure 4.   Bronchial epithelial cells undergo early apoptosis in response to local LPS administration. Lung tissue of LPS- and saline-treated mice was fixed in 10% formalin, processed according to standard procedure, and stained by the TUNEL method. (A) Saline-treated, 4 h. (B) LPS-treated, 4 h. (C) LPS-treated, 24 h. Bar = 60 µm. Note the brown nuclear staining in bronchial epithelial cells (arrows) at 4 and 24 h after LPS exposure, which is absent in other pulmonary cell types. Infiltrated neutrophils displayed positive staining at 24 h after exposure.

LPS-Induced Early Apoptosis in Lung Tissue Is Not Mediated by TNF-

Because the proinflammatory cytokine TNF- is suggested to be the primary mediator in endothelial apoptotic cell death after systemic administration of LPS, we investigated the role of TNF- in the observed early apoptosis after IT LPS challenge. To confirm the presence of TNF- in lungs after local LPS administration, we examined TNF- protein production by specific ELISA. As shown in Table 1, IT LPS instillation resulted in elevated levels of TNF- protein in the BALF, peaking at 4 h. TNF- protein was not detectable in plasma, indicating local production of TNF- in the lung. To further establish that TNF- was lung-derived, expression of messenger RNA (mRNA) for TNF- was assessed by RT-PCR. Figure 5 shows enhanced pulmonary TNF- mRNA expression at 2 h after LPS challenge, as compared with saline-treated mice.

View larger version (56K): [in this window] [in a new window]   Figure 5.   Transcription of TNF- and -actin in murine lung tissue as determined by RT-PCR. Mice received 5 µg of LPS or saline (S) intratracheally and were killed 2 h after instillation. Total RNA was isolated from lung tissue and analyzed for TNF- and -actin mRNA expression using RT-PCR. Equal intensity of -actin mRNA bands confirmed equal loading of cDNA in RT-PCR. One representative experiment from a series of three is shown.

The putative role of TNF- in LPS-induced apoptosis at early time points was determined using three experimental approaches. First, TNF- was neutralized using an anti- TNF- antibody (TN3). Mice were pretreated with 1 mg/ mouse TN3 2 h before IT LPS instillation and killed 4 h after instillation. As shown in Figure 6A, pretreatment with TN3 did not reduce the extent of DNA laddering in lung tissue induced by IT LPS as compared with LPS-treated controls (intensity: 360-bp band 7.0 U [LPS + TN3] versus 7.5 U [LPS]). Pretreatment with an IgG class-matched control antibody (5D7) also did not affect DNA laddering at 4 h after LPS challenge (data not shown). The effectiveness of inhibition of TNF- by the pretreatment with TN3 was concluded from experiments in which mice were pretreated with TN3 followed by IT LPS instillation and killed 24 h after instillation. Histologic assessment showed that LPS-induced pulmonary neutrophil accumulation was strongly diminished by TN3 (data not shown).

View larger version (51K): [in this window] [in a new window]   Figure 6.   TNF- is not involved in LPS-induced early apoptosis in lung tissue. (A) TNF- neutralization or IT TNF- instillation. Swiss mice received saline (lane 1) or 5 µg of LPS intratracheally (lane 2), or mice were pretreated with an anti-TNF- antibody (TN3, 1 mg/mouse) 2 h before LPS instillation (lane 3). Mice were killed 4 h after IT treatment. An additional set of mice received 5 µg of IT murine TNF- (lane 4) and were killed after 2 h. DNA fragmentation in lung tissue was detected by LM-PCR, and intensity of the 360-bp band was analyzed (expressed in arbitrary units). M, molecular weight markers (range: 100 to 2,000 bp). Note that anti-TNF- administration did not reduce DNA laddering induced by IT LPS, whereas IT TNF- instillation did not result in DNA laddering. (B) LPS-induced early apoptosis in lung tissue of TNF/LT- knockout mice. TNF/LT- knockout (/) and wild-type (+/+) mice received IT saline (S) or 5 µg of LPS (LPS), and were killed 4 h after exposure. M, molecular weight markers. Note that IT LPS exposure in both TNF/LT- / and +/+ mice resulted in increased DNA laddering. One representative experiment from a series of three is shown.

Next, mice were instilled intratracheally with a high dose of murine rTNF- (5 µg) to investigate the properties of TNF- to induce apoptosis in lungs. Because LPS-induced TNF- production was evident from 2 h after instillation, mice were killed at 2 h after exposure, and the extent of DNA laddering in the lung was determined by LM-PCR. Figure 6A shows that local TNF- instillation did not induce DNA laddering (intensity: 360-bp band 0.2 U [TNF-] versus 7.5 U [LPS]).

To confirm these results, TNF/LT- double-knockout mice (/) and wild-type (+/+) control mice were used. Mice received LPS or saline intratracheally, and were killed at 4 h after instillation. As shown in Figure 6B, basic levels of DNA fragmentation ladders differed between TNF/LT- / and +/+ mice (intensity: 360-bp band 1.0 U [S /] and 0.1 U [S +/+]). In spite of this difference, local LPS exposure resulted in increased DNA laddering in both TNF/LT- / and +/+ mice (intensity: 360-bp band 5.7 U [LPS /] and 2.1 U [LPS +/+]) compared with saline-treated controls, thereby demonstrating that in the absence of TNF-, DNA laddering was induced by IT LPS instillation. Together, these data clearly demonstrate that TNF- is not the primary mediator in LPS-induced early apoptotic cell death in the lung.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

LPS is ubiquitously present as a contaminant on airborne particles, including air pollution (4), organic dusts (5), and cigarette smoke (6). Local exposure to LPS is known to be associated with pulmonary cytokine production resulting in infiltration of neutrophils and pulmonary injury. However, little is known about the direct effects of LPS in pulmonary injury, such as apoptotic cell death. In this study, apoptosis in lung tissue after local (IT) exposure to LPS of 5 µg/mouse was investigated. This amount of LPS was considered relevant, taking into account that smoking of one cigarette delivers approximately 0.2 µg LPS to the lung (6) and the reported 2 to 3 log less sensitivity of mice to LPS compared with humans (24, 25). Moreover, occupational exposure to (organic dust containing) LPS is known to deliver an even higher LPS dose (30 to 60 µg) to the lung over an 8-h work shift (26). To identify apoptosis in lung tissue, we used LM-PCR on genomic DNA, a very sensitive technique that enables detection of apoptosis in small percentages of cells (27). Interestingly, our results show that IT LPS instillation resulted in induction of DNA fragmentation ladders in lung tissue, which was already evident 2 h after LPS treatment, thereby indicating that IT LPS instillation induces apoptosis in the lung at early time points. Histologic analysis by TUNEL using the sensitive digoxygenin (DIG)-anti-DIG labeling system, which reduces the chance of detecting secondary necrotic cells (22), demonstrated nuclear staining in bronchial epithelial cells from 4 h after exposure. Additional morphological examination demonstrated characteristic condensed chromatin in the nuclei of positive-stained bronchial epithelial cells, thereby confirming their apoptotic state in response to IT LPS instillation.

Neutrophils play a prominent role in the host defense against pathogens, but are also considered to be responsible for pulmonary injury by the release of toxic contents, such as proteases and oxygen radicals. Disruption of bronchial epithelial cell interactions with neighbor cells or extracellular matrix is known to enhance apoptosis (10, 28), which could be one of the mechanisms for neutrophils to induce pulmonary injury. Indeed, a recent study showed that neutrophil presence induced apoptotic cell death in bronchial epithelial cells in vitro (8). Our finding that apoptosis in the LPS-challenged lung clearly preceded pulmonary neutrophil infiltration excludes the possibility that neutrophils are involved in bronchial epithelial cell apoptosis at early time points. These observations are in line with a previous study, demonstrating that LPS-induced endothelial cell apoptosis in vivo also occurred in absence of an inflammatory response (2). Interestingly, TUNEL revealed that infiltrated neutrophils displayed positive staining from 24 h after LPS exposure, thereby demonstrating that apoptotic cell death in infiltrating neutrophils contributes to increased DNA laddering seen in lungs at later time points after exposure. In line with our observations in mice, infiltrated neutrophils are reported to be eliminated from sites of acute inflammation by apoptosis in rat models of acute pulmonary inflammation, induced by ozone (22) or IT LPS (23) exposure. In these rats, apoptotic neutrophils become quickly engulfed and degraded by alveolar macrophages, which may explain the quick decrease in DNA laddering in lungs observed in our murine model at 72 h after LPS exposure.

TNF-, a proinflammatory cytokine with pleiotropic effects, is known to be rapidly produced in the lung during acute pulmonary inflammation after local LPS exposure (29, 30) as was also demonstrated at both mRNA and protein levels in this study. TNF- is also known to induce apoptosis in various cell types both in vitro and in vivo. Murine models of endotoxic shock have provided abundant evidence that TNF- acts as the primary mediator in LPS-induced endothelial cell apoptosis in lung (1, 3), with ceramide as the intracellular effector molecule (2). In contrast with the role of TNF- in LPS-induced endothelial cell apoptosis, our findings exclude TNF- as a primary mediator in bronchial epithelial cell apoptosis induced by IT LPS instillation. We showed that anti-TNF- treatment did not decrease the extent of DNA laddering in lung tissue at 4 h after exposure as compared with LPS-treated controls, whereas pulmonary neutrophil infiltration was reduced 24 h after LPS exposure, thereby confirming that biologic activity of TNF- was inhibited. These results were confirmed by experiments performed in TNF/LT- double-knockout mice, which demonstrated that DNA fragmentation ladders in lung tissue increased 4 h after LPS exposure in both TNF/LT- knockout and wild-type mice. Moreover, exposure of mice to a high dose of murine rTNF- intratracheally did not increase DNA laddering in lung tissue. Mallampalli and associates (31) recently reported that IT TNF- instillation in rats did not increase apoptosis in another type of pulmonary epithelial cells (alveolar type II cells) at 4 to 24 h after exposure. These studies suggest that TNF- is not essential for epithelial cell apoptosis in the lung.

The present observations suggest a direct role for LPS in bronchial epithelial cell apoptosis. To our knowledge, only few studies have investigated the mechanisms underlying LPS-induced apoptosis in vitro and in vivo. Haimovitz-Friedman and coworkers (2) demonstrated that endothelial cell apoptosis induced by systemic LPS exposure is mediated by sequentially TNF- and ceramide, which is known to act as a second messenger in pleiotropic cellular functions, including proliferation, differentiation, and apoptosis (32). A possible mechanism for LPS-mediated apoptosis in epithelial cells may involve Toll-like receptor (TLR)-4, which has been identified as transmembrane coreceptor in the LPS binding protein (LBP)/CD14-dependent LPS signaling across the cell membrane (33). Recently, upregulation of TLR-4 mRNA was demonstrated in heart and lung in response to LPS administration (34). The observation that LPS-induced DNA laddering in endothelial cells was inhibited by addition of an anti-CD14 antibody (11) supports the involvement of the LBP/CD14/TLR-4 pathway in LPS-mediated apoptotic cell death. Another mechanism by which LPS could trigger apoptosis may be via directly mimicking the second messenger function of ceramide in apoptotic cell death. Lipid A is known to have strong structural similarities with ceramide (35). In addition, in vitro studies also suggested functional similarities between LPS and ceramide. Joseph and colleagues (35) demonstrated that lipid A and native LPS directly stimulated ceramide-activated protein kinase (CAPK) activity, a direct intracellular target for ceramide in the process of apoptosis, in human leukemia (HL-60) cells and isolated human neutrophils without generating ceramide. This CAPK activity was markedly enhanced by LBP and required CD14. The observation that LPS can be transferred into phospholipid bilayers by CD14 and LBP (36) supports the idea that LPS can directly interact with ceramide-responsive enzymes in the plasma membrane. Whether CD14, TLRs, and ceramide are involved in bronchial epithelial cell apoptosis observed after IT LPS exposure needs to be further elucidated.

The respiratory system is challenged continuously by LPS via inhaled air containing airborne particles and pathogens. In this study we have demonstrated that IT exposure of a relevant dose of LPS induces a mild transient acute inflammation preceded by apoptosis in part of the bronchial epithelial cells throughout the bronchial tree. Whether this early induced apoptotic cell death contributes either to tissue damage, due to disturbance of the main barrier between the environment and the respiratory system, or to innate immunity is not yet understood. However, apoptosis is increasingly being identified as a protective response by the host to pathogens, with the suicide of individual cells enhancing the survival of the multicellular organism as a whole (37). We consider the demonstrated apoptosis important, because clearance of harmed bronchial epithelial cells by the process of apoptosis would prevent disruption of the epithelial layer (as necrosis would do), thereby preventing the spread of pathogens. Further research needs to be performed to elucidate the role of LPS-mediated bronchial epithelial cell apoptosis in airway injury.