Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Alveolar epithelial cells, situated at the boundary between the alveolar air space and the interstitium, are important components in host defense. Type I cells function as a barrier to prevent the invasion of pathogens, and type II cells produce surfactant that may modulate inflammatory reactions of alveolar macrophages (1). More recently, alveolar epithelial cells have been recognized to be involved in the recruitment and activation of inflammatory cells through the production of cytokines and chemokines, in response to various stimuli from the alveolar space (2, 3). Overexpression of cytokines and chemokines also contributes to acute lung injury induced by severe sepsis and other injurious conditions (2, 3).

Recently, we have found that isolated rat alveolar epithelial cells release tumor necrosis factor (TNF)- in response to endotoxin lipopolysaccharide (LPS) challenge (4). Compared with alveolar macrophages (5), the dose of LPS that evokes TNF- release from alveolar epithelial cells was higher, whereas the amount of TNF- produced was less (4). Because TNF- is an early response cytokine that can further induce other cytokines and chemokines from a variety of cell types, we anticipated that LPS-induced TNF- production by alveolar epithelial cells might be an intermediary for the production of chemokines from the same cell type.

Chemokines are chemotactic cytokines for leukocyte and monocyte recruitment and activation at the sites of infection or tissue injury (6). Alveolar epithelial cells can secrete a variety of chemokines, including interleukin (IL)-8 (7), monocyte chemoattractant protein (MCP)-1 (8, 9), and regulated on activation, normal T cell expressed and secreted (RANTES) (10). Polymorphonuclear leukocytes (PMN) infiltration in the alveolar space is a main feature of acute lung injury, which is mainly mediated by C-X-C chemokines such as IL-8 (7) and its rodent homologue, macrophage inflammatory protein (MIP)-2 (11, 12). However, whether alveolar epithelial cells can produce MIP-2 is unknown.

MIP-2 was initially purified from a mouse macrophage cell line (RAW264.7) stimulated with endotoxin (13). Rat MIP-2 was recently cloned and expressed as a 7.9-kD peptide (14) that showed dose-dependent chemotactic activity for PMN (14). This activity of MIP-2 has been further demonstrated in the lung from several animal models with a variety of pathogens. For example, increased MIP-2 messenger RNA (mRNA) and/or protein was observed in the lung, in response to the intratracheal instillation of Klebsiella pneumoniae (17), Pseudomonas aeruginosa (18), LPS (14, 19), -quartz (20), and other dust particles (21). MIP-2 was also involved in immunoglobulin (Ig) G immune complex-induced lung injury (22). In addition, intraperitoneal administration of LPS resulted in an increase in neutrophil influx into the lung, which was at least partly due to the increased MIP-2 (23, 24). To further determine the direct function of MIP-2 in the lung, a replication-defective adenoviral vector was used to deliver MIP-2 gene by intratracheal instillation. Significant increase in neutrophils and alveolar macrophages was found from the lung lavage fluid (25). When recombinant MIP-2 was delivered into the alveolar space of rats through a catheter wedged into a bronchus, profound neutrophil localization was observed in both the vascular and alveolar spaces (24). Increased MIP-2 was associated with accumulation of inflammatory cells, especially PMN, and acute lung injury. Intraperitoneal instillation of anti-MIP-2 antiserum (17) or intrapulmonary instillation of anti-MIP-2 antibody (14, 22) decreased the influx of neutrophils in the lung and attenuated lung injury.

In the present study we demonstrate that primary cultured alveolar epithelial cells were able to produce MIP-2 in response to LPS stimulation, which was regulated at both the transcriptional and post-transcriptional levels. An antisense oligonucleotide (ON) against TNF- inhibited LPS-induced production of TNF- from alveolar epithelial cells in a dose-dependent and sequence-specific fashion. This treatment also inhibited LPS-induced MIP-2 production. Further, polyclonal and monoclonal antibodies against TNF- attenuated LPS-induced MIP-2. These results suggested an autoregulatory mechanism in alveolar epithelial cells in response to endotoxin stimulation, which may play an important role in host defense as well as inflammatory reactions related to acute lung injury.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Rat Alveolar Epithelial Cell Isolation and Culture

Alveolar type II cells were obtained using the method of Dobbs (26), which has been previously described (27). Briefly, male adult Sprague-Dawley rats (Harlan, Indianapolis, IN) weighing approximately 250 g were anesthetized by intraperitoneal injection of sodium pentobarbital (100 mg/kg body weight; MTC Pharmaceuticals, Cambridge, ON, Canada) and killed by transection of the descending aorta and inferior vena cava. Alveolar epithelial cells were separated from the alveolar basement membrane by incubation of the isolated lung tissue with porcine pancreatic elastase (Worthington Biochemical Corp., Freehold, NJ). Contaminating alveolar macrophages were removed by differential adherence to petri dishes precoated with rat IgG (Sigma Chemical Co., St. Louis, MO). The number and viability of fresh cell suspensions were counted after staining with crystal violet and trypan-blue exclusion. The viabilities of fresh alveolar epithelial cell suspensions were greater than 95%.

Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% (vol/vol) fetal bovine serum (FBS) and 12.5 µg/ml gentamicin. To reduce the contamination of alveolar macrophages in the primary culture, culture media were changed daily for 2 d before LPS treatment. As we have reported recently (4), this maneuver reduced the number of macrophages to undetectable levels by cell-surface ectoenzyme-alkaline phosphatase staining (28) or by immunofluorescent staining with a monoclonal antibody for CD45 (Bio-Can, Mississauga, ON, Canada), a surface marker for macrophages and leukocytes. The purity of alveolar epithelial cells in the culture system was confirmed with phase-contrast microscopy and immunofluorescent staining with anticytokeratin and anti-pro-surfactant protein (SP)-C antibodies (specific markers for epithelial cells and type II pneumocytes, respectively).

LPS and TNF- Stimulation and Inhibitor Treatments

Freshly isolated rat lung cells (10⁶ cells/ml) were plated on 24-well culture plates (1 ml/well; Corning Glass Works, Corning, NY) in DMEM plus 10% FBS, and maintained at 37°C with 5% CO₂. At 2 d after inoculation, cells were treated with various concentrations of LPS from Escherichia coli (Sigma) or recombinant rat TNF- (Biosource, Camarillo, CA) in DMEM plus 10% FBS. To study the regulatory mechanisms of LPS-induced MIP-2 release, cells were pretreated with various concentrations of cycloheximide or actinomycin D for 2 h, and then stimulated with or without LPS for another 4 h. Cell culture media were collected and analyzed for MIP-2 by enzyme-linked immunosorbent assay (ELISA). No cytotoxic effect on rat alveolar epithelial cells was found for these compounds with the concentrations used in the present study (see subsequent discussion).

Cytokine Measurements

MIP-2 and TNF- concentrations in the culture media were measured in duplicate or triplicate using ELISA kits (Biosource), following the manufacturer's instructions. According to the manufacturer, the kit for rat MIP-2 has no cross-activity to many human and mouse cytokines; and no cross-activity with rat interferon- and MCP-1. The kit for TNF- has 0.15% cross-activity to human TNF- and 100% cross-activity to mouse TNF-. The optical density (OD) of each well was read at 450 nm with an NM600 microplate reader (Dynatech Laboratories, Chantilly, VA). The final concentration was calculated by converting the OD readings against a standard curve.

Immunofluorescent Staining and Confocal Microscopy

To study the distribution of MIP-2-producing cells, freshly isolated lung cells were seeded on glass coverslips in six-well culture plates at a density of 0.5 × 10⁶ cells/ml. Cells were cultured in DMEM plus 10% FBS. Medium was changed every day. At 48 h after cell isolation, cells were washed twice with DMEM and treated with or without LPS (10 µg/ml) in DMEM plus 10% FBS for 2 h. The cells on coverslips were washed with phosphate-buffered saline (PBS) and fixed with 3.7% formaldehyde for 10 min. After rinsing with PBS, cells were permeabilized for 10 min in methanol, slides were washed, and nonspecific absorbency was blocked with 10% normal goat serum in PBS for 60 min at room temperature. After rinsing with PBS, cells were incubated at 37°C for 60 min with a rabbit polyclonal antirat MIP-2 antibody (Cedarlane Laboratories, Hornby, ON, Canada) diluted 1:40 in PBS containing 3% normal goat serum. Coverslips were washed three times for 5 min with PBS, and then incubated for 60 min at 37°C in the dark with fluorescein isothiocyanate (FITC)-conjugated goat antirabbit IgG (Jackson Immunoresearch Laboratories, West Grove, PA) diluted 1:60 in PBS with 3% normal goat serum. After three washes with PBS, coverlips were rinsed briefly in distilled water and mounted with an antifading reagent (SlowFade; Molecular Probes, Eugene, OR). Confocal laser scanning was performed using a Bio-Rad MRC-600 confocal microscope (Bio-Rad, Mississauga, ON, Canada), equipped with a krypton/argon laser. To determine the specificity of staining, the first antibody was replaced with nonspecific rabbit IgG (Sigma) or omitted from the staining procedure.

To compare the intracellular concentration and localization of MIP-2 in the presence or absence of LPS stimulation, some slides were double stained with tetramethylrhodamine isothiocyanate (TRITC)-labeled concanavalin A (Con A) (Sigma), to visualize the endoplasmic reticulum (ER) arrangement (29). After the second antibody incubation, TRITC-labeled Con A, diluted (1:50) from the stock solution (20 mg/ml) with distilled water, was added to slides and incubated at 37°C for 30 min in the dark. Other steps remained the same as described previously.

RNA Extraction and Semiquantitative Reverse Transcriptase/Polymerase Chain Reaction

Cells were cultured in six-well plates (4 × 10⁶ cells/well). At 48 h after cell isolation, cells were treated with LPS (10 µg/ml) in DMEM plus 10% FBS for various periods. Media were removed and cells were washed twice with cold PBS. Total RNA was extracted using an RNeasy total RNA extraction kit (Qiagen, Chatsworth, CA). Briefly, cells were lysed with 0.25 ml of lysis buffer with 1% of freshly added -mercaptoethanol (10 µl/ml), and homogenized with a QIAshredder column (Qiagen). The total RNA was extracted following the manufacturer's instructions. RNA concentration was measured with a spectrophotometer (Beckman DU 640B; Beckman, Fullerton, CA). Total RNA (5 µg) was used for reverse transcription reaction, using a Superscript II kit (GIBCO BRL, Mississauga, ON, Canada). Reverse transcriptase (RT) product from 0.5 µg RNA was used for polymerase chain reaction (PCR). The forward and reverse primers for -actin were designed within exons 1 and 3 of complementary DNA (cDNA) of rat cytoplasmic -actin, respectively (30), which were synthesized by ACGT Corporation (Toronto, ON, Canada). The primers for MIP-2 recognize cDNA of rat MIP-2 between the 39th and 258th nucleotides in the coding region (16), which were purchased from Biosource. The sequences of primers for -actin and MIP-2 are listed in Table 1. The PCR mixture was set up in a total volume of 30 µl, containing 3 µl 10× PCR buffer (200 mM Tris-Cl, pH 8.4, and 500 mM KCl), 1 µl 50 mM MgCl₂, 0.5 µl 10 mM deoxynucleotide triphosphate mix, 0.5 µl of each PCR primer (10 µM), and 0.3 µl Taq polymerase (GIBCO BRL). PCR was performed with a programmable thermal cycler (PTC-100; MJ Research, Watertown, MA). The optimized PCR conditions are described in Table 1. A total of 10 µl of PCR product was electrophoresed on 1% agarose gel with ethidium bromide staining for visualization, and the gels were photographed and quantified with a gel documentation system (Gel Doc 1000; Bio-Rad). To ensure the compatibility, RT-PCR was performed simultaneously on all samples collected from each experiment. PCR products were analyzed on the same gel. The OD of PCR-product bands was quantified with integrated image analysis software (Molecular Analyst, Version 1.5; Bio-Rad, Hercules, CA). With optimized PCR conditions, all data were collected without saturation or missing bands. The background of OD reading for each band was subtracted locally. Each experiment was conducted at least twice to ensure the reproducibility. To exclude possible DNA contamination during the PCR, RNA samples were amplified by PCR without reverse transcription. No band was observed, suggesting that there was no DNA contamination in the RNA preparation procedure (data not shown).

Treatment of Cells with Antisense Agents

To specifically inhibit LPS-induced TNF- synthesis from alveolar epithelial cells, nuclease-resistant phosphorothioate ON target to the initiation site of TNF- mRNA (5' TGT GCT CAT GGT GTC TTT 3') and its inverted sequence (5' TTT CTG TGG TAC TCG TGT 3') were synthesized (Oligo Etc., Wilsonville, OR). These ONs were first described and used by Rojanasakul and coworkers (31). The antisense ON effectively inhibited silica-induced TNF- production in alveolar macrophages (31). Antisense and inverted ONs were delivered to alveolar epithelial cells using Lipofectin (GIBCO BRL) as a carrier, which has been used to deliver antisense against TNF- into macrophages (32). The surface of the cationic liposomes is positively charged, and thus is attracted to the phosphate backbone of negatively charged ONs to form DNA-lipid complexes. A good charge ratio suggested by the manufacturer is 1:1. Our ONs were 18-mer, therefore, 1 µM of ON was mixed with 18 µM of Lipofectin. ONs and Lipofectin were diluted with serum- and antibiotic-free DMEM, incubated at room temperature for 15 min, and then mixed together and allowed to stand at room temperature for another 45 min. The DNA-liposome mixture was diluted in series with antibiotic-free DMEM to various concentrations, added to cells, and incubated at 37°C for 4 h. After the preincubation, culture media were replaced with regular DMEM plus 10% FBS, and cells were incubated with or without LPS (10 µg/ml) for an additional 4 h. Cell culture media were collected, and TNF- and MIP-2 were analyzed by ELISA.

Neutralization of TNF- with Antibodies

To inhibit secreted TNF- from alveolar epithelial cells, anti-TNF- neutralizing antibodies were used. At 48 h after cell isolation, cells were washed twice with DMEM and pretreated with various concentrations of rabbit antimouse TNF- polyclonal antibody (Genzyme, Cambridge, MA) or rat antimouse/rat TNF- monoclonal antibody (PharMingen, San Diego, CA) for 1 h. As negative controls, cells were pretreated with either nonimmune rabbit IgG (Sigma) or rat IgG (Sigma). Cells were then stimulated with LPS (10 µg/ml) for an additional 4 h. Cell culture media were collected, and MIP-2 was analyzed by ELISA.

Cytotoxicity

Cytotoxic effects of LPS, TNF-, chemical inhibitors, ONs, and Lipofectin were examined by simultaneous double staining with fluorescein diacetate (FDA) and propidium iodide (PI), a rapid, convenient, and reliable method of determining cell viability (33). Cells were incubated in 96-well plates. After treatment with various agents mentioned above, culture media were removed, and cells were washed twice with Dulbecco's PBS (DPBS). The stock solutions of FDA (5 mg/ml in acetone) and PI (0.02 mg/ml in DPBS) were stored at 4°C in the dark. The working solution was freshly diluted and mixed with DPBS. The final solution, containing 2 µg of FDA and 0.6 µg of PI, was added to each well for 3 min at room temperature. The viability of cells was examined with a fluorescent microscope with a 520- and 590-nm filter set. Viable cells fluoresced bright green; nonviable cells were bright red. The viability of cells in all groups in this study was found to be comparable to that of the control group without LPS or drugs.

Statistical Analysis

All experiments were carried out with materials collected from at least two to three separate cell cultures in duplicate or triplicate. Data are expressed as means ± standard error (SE) from separated experiments, or from measurements of representative experiments. Data were analyzed by one-way analysis of variance (ANOVA), followed by Student-Newman-Keuls (SNK) test (34) with significance defined as P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

LPS Induces MIP-2 Release from Alveolar Epithelial Cells

When cells were incubated with various concentrations of LPS for 4 h, there was a dose-dependent release of MIP-2 into the culture medium (Figure 1A). Maximal stimulation of MIP-2 release was observed when cells were treated with 10 µg/ml of LPS, which was the same as the dose that induced the maximal release of TNF- from the same cell type (4). This dose was then used in all subsequent experiments. To study the time course of LPS-induced MIP-2 gene expression and protein production, cells were cultured on six-well plates. The absolute concentrations of MIP-2 in the culture media were different from those seen in LPS dose-response studies. However, in several separate experiments, LPS induced a rapid release of MIP-2 within the first 4 to 8 h, and a slower increase over the next 16 to 20 h (Figure 1B). Morphology of LPS-treated and control cells appeared identical based on phase-contrast microscopy and no cell injury was observed at any given concentration of LPS over a 24-h culture, as determined by FDA/PI double staining in the present study (data not shown).

View larger version (13K): [in this window] [in a new window]   Figure 1.   LPS induced dose- and time-dependent release of MIP-2 from primary cultured alveolar epithelial cells. (A) Cells were cultured in 24-well plates with DMEM supplemented with 10% FBS, and treated with various concentrations of LPS for 4 h. Cell culture media were collected, and MIP-2 was analyzed with ELISA. Data are means ± SE (n = 3 wells) from a representative experiment. P < 0.0001 as assessed by one-way ANOVA. *P < 0.05 compared with groups treated with 105 to 104 µg/ml of LPS, as assessed by SNK test. (B) Cells were cultured in six-well plates with DMEM supplemented with 10% FBS, and treated with LPS (10 µg/ml) for various periods. Data are means ± SE from three separate experiments in duplicate. P < 0.0001 as assessed by one-way ANOVA. *P < 0.05 compared with groups treated with LPS for 0 to 2 h, as assessed by SNK test.

By extending culture to 48 h with medium changed daily, the number of alveolar macrophages was decreased to almost zero, as measured with immunofluorescent staining with an antibody for CD45, a surface marker of macrophages and leukocytes (data not shown). The purity of alveolar epithelial cells was also confirmed with phase-contrast microscopy and immunofluorescent staining for cytokeratin (marker for epithelial cells) and pro-SP-C (marker for type II pneumocytes) (data not shown). To confirm that MIP-2 was produced by alveolar epithelial cells and not due to the trace number of contaminating alveolar macrophages, we used confocal microscopy to examine the distribution and localization of MIP-2. Cells were treated with LPS (10 µg/ml) for 2 h. Immunofluorescent staining using a polyclonal anti-MIP-2 antibody demonstrated a positive reaction throughout the cell layer (Figure 2), with MIP-2 localized to the cytoplasm (Figure 2). As a negative control, staining of the same batch of cells was performed, omitting primary antibody or replacing it with nonimmune rabbit IgG. In both cases, no staining was detected (data not shown).

View larger version (130K): [in this window] [in a new window]   Figure 2.   Intracellular localization of LPS-induced MIP-2. Cells were cultured on coverslips with DMEM supplemented with 10% FBS. After stimulation with LPS (10 µg/ml) for 2 h, cells were fixed and stained with an anti-MIP-2 polyclonal antibody and examined with confocal microscopy. Positive staining was observed throughout the cell layer and localized in the cytoplasm. No staining was observed from nonimmune rabbit IgG-stained cells, or when the primary antibody was omitted from the staining protocol (data not shown). The figure is an example from three separate experiments.

LPS-Induced MIP-2 Production Is Regulated at Both Transcriptional and Post-transcriptional Levels

The regulatory mechanisms of MIP-2 production in alveolar epithelial cells were studied by treating cells with inhibitors of protein and RNA synthesis. Cycloheximide, an inhibitor of translation, blocked LPS-induced MIP-2 release in a dose-dependent manner (Figure 3), suggesting that most of the MIP-2 released into the culture media was newly synthesized protein. To confirm this speculation, cells were treated with or without LPS, double stained with the antibody against rat MIP-2 and TRITC-labeled Con A for ER, and examined with confocal microscopy. ER was equally stained in control and LPS-treated cells (Figure 4). In contrast, the staining for MIP-2 in the control cells was almost not detectable, but it was clearly increased in cells treated with LPS (10 µg/ml) for 2 h (Figure 4). These results provide further evidence that MIP-2 released into the culture medium was newly synthesized in the alveolar epithelial cells.

View larger version (16K): [in this window] [in a new window]   Figure 3.   LPS-induced MIP-2 release was blocked by cycloheximide, a translational inhibitor. Cells were cultured with DMEM supplemented with 10% FBS and pretreated with various concentrations of cycloheximide for 2 h, then incubated with LPS (10 µg/ml) for another 4 h. Cell culture media were collected and MIP-2 was analyzed with ELISA. Data are expressed as percent of LPS stimulation (means ± SE, n = 3 wells from a representative experiment). P < 0.0001 as assessed by one-way ANOVA. *P < 0.05 compared with the LPS-treated group in the absence of cycloheximide, as assessed by SNK test.

View larger version (93K): [in this window] [in a new window]   Figure 4.   LPS stimulation increases intracellular MIP-2 content. Cells were cultured on coverslips with DMEM supplemented with 10% FBS. After incubation with or without LPS (10 µg/ml) for 2 h, cells were fixed and stained with an anti-MIP-2 polyclonal antibody followed by FITC-conjugated secondary antibody and TRITC-labeled Con A to visualize the ER, and examined with confocal microscopy. Cells were simultaneously imaged with two channels: one for FITC (left side) and the other for TRITC (right side). The staining for ER was not affected by LPS treatment. In contrast, staining for intracellular MIP-2 was clearly increased after LPS stimulation. Figures are examples from three separate experiments.

To determine whether the synthesis of MIP-2 requires new transcripts of the gene, cells were treated with LPS (10 µg/ml) for various time intervals, total RNA was extracted, and mRNA levels of MIP-2 were determined by semiquantitative RT-PCR. A basal level of MIP-2 mRNA was detected in untreated cells. The MIP-2 mRNA was rapidly increased within 2 h of LPS stimulation and then the expression was decreased slowly (Figure 5). In addition, actinomycin D, an inhibitor of gene transcription, blocked LPS-induced MIP-2 release from alveolar epithelial cells in a dose-dependent manner (Figure 6). Therefore, MIP-2 synthesis is effectively regulated at both the transcriptional and post-transcriptional levels in alveolar epithelial cells.

View larger version (15K): [in this window] [in a new window]   Figure 5.   LPS-induced MIP-2 gene expression. Cells were cultured in six-well plates and treated with LPS (10 µg/ml) for various periods. RNA was extracted, reverse transcribed to cDNA, and amplified by PCR. PCR products were resolved by gel electrophoresis and analyzed by a gel documentation system. Because -actin mRNA level was not affected by LPS treatment, it was used as an internal control. Arbitrary units of densitometry analysis are expressed as the ratio between MIP-2 and -actin from the same sample. All values are means ± SE from three experiments (P < 0.001 as assessed by one-way ANOVA). The level of MIP-2 mRNA at 2 h of LPS treatment was significantly higher than that at 0 or 24 h (P < 0.05 as assessed by SNK test). The level of MIP-2 mRNA at 24 h of LPS treatment was also significantly lower than that at 1 or 4 h (P < 0.05 as assessed by SNK test).

View larger version (17K): [in this window] [in a new window]   Figure 6.   LPS-induced MIP-2 release was blocked by actinomycin D, a transcriptional inhibitor. Cells were cultured with DMEM supplemented with 10% FBS and pretreated with various concentrations of actinomycin D for 2 h, then incubated with LPS (10 µg/ml) for another 4 h. Cell culture media were collected and MIP-2 was analyzed with ELISA. Data are expressed as percent of LPS stimulation (means ± SE, n = 3 wells from a representative experiment). P < 0.0001 as assessed by one-way ANOVA. *P < 0.05 compared with the LPS-treated group in the absence of actinomycin D, as assessed by SNK test.

Inhibition of TNF- with Antisense ON Suppresses MIP-2 Production

We noted that the LPS-induced MIP-2 requires the same dosage of LPS to reach the maximal response as that observed for TNF- from alveolar epithelial cells (4). Compared with TNF-, LPS-induced MIP-2 release and increase of steady-state mRNA levels of MIP-2 occurred later. It has been shown that stimulation of primary cultured rat type II cells with TNF- induced MIP-2 gene expression (20). TNF- was reported to induce production of MIP-2 protein from the RAW264.7 macrophage cell line (35). To determine whether TNF- can stimulate MIP-2 production from alveolar epithelial cells, we incubated cells with recombinant rat TNF- for 4 h; we found a dose-dependent release of MIP-2 (Table 2).

We speculated that LPS-induced TNF- release might be related to the MIP-2 production from the same cell type. Preincubation of alveolar epithelial cells with antisense ON against TNF- delivered with liposome decreased TNF- in a dose-dependent manner (Figure 7A). To exclude nonspecific effects of ON or Lipofectin, cells were preincubated with an ON with inverted sequence from the antisense ON in the presence of Lipofectin, or with Lipofectin alone. None of these treatments affected TNF- production (Figure 7B). Interestingly, in the presence of antisense ON against TNF-, the LPS-induced MIP-2 production was also inhibited in a similar dose-dependent manner (Figure 8A). This effect was specific to antisense ON, because neither the inverted sequence ON nor the liposome alone had such effect (Figure 8B). We combined data from three experiments, in which cells were treated with various concentrations of antisense ON against TNF-, and plotted the percent inhibition of MIP-2 by antisense ON against TNF- treatment from individual culture well against the percent inhibition on TNF-. Both inhibitory effects are highly correlated (P < 0.001) (Figure 9). The inhibitory effect was not due to cytotoxic effect of antisense ON treatment, because the viability of cells remained no different between control and LPS-treated cells in the presence or absence of antisense, inverted ONs, or liposome, as determined by FDA/PI double staining (data not shown). To exclude the possibility of cross-reaction of antisense ON against TNF- with the MIP-2 gene sequence, a homology comparison analysis was performed using the DNASIS computer program (Hitachi Software Engineering Co., Ltd., South San Francisco, CA). Neither the antisense nor inverted sequences is related to rat MIP-2.

View larger version (16K): [in this window] [in a new window]   Figure 7.   Antisense against TNF- inhibited LPS-induced TNF- production from alveolar epithelial cells in a dose-dependent and sequence-specific manner. (A) Cells were pretreated with various concentrations of antisense for 4 h using Lipofectin as a carrier, then challenged with LPS (10 µg/ml) for another 4 h. The concentrations of TNF- in the medium were measured. Data are means ± SE combined from three experiments. P < 0.01 as assessed by one-way ANOVA. *P < 0.05 versus cells treated without antisense, as assessed by SNK test. (B) To test the specificity of the antisense, cells were pre-incubated with Lipofectin alone (LIPF), or Lipofectin plus antisense (AS) or inverted (INV) ONs (1 µM) for 4 h, and then stimulated with LPS (10 µg/ml) for 4 h. Data are means ± SE (n = 3 wells) from a representative experiment. P < 0.01 as assessed by one-way ANOVA. *P < 0.05 versus cells treated with LPS only, as assessed by SNK test.

View larger version (16K): [in this window] [in a new window]   Figure 8.   Antisense against TNF- inhibited LPS-induced MIP-2 production from alveolar epithelial cells. MIP-2 concentrations were measured from the samples as described in Figure 7. (A) Data are means ± SE combined from three experiments. P < 0.01 as assessed by one-way ANOVA. *P < 0.05 versus cells treated without antisense, as assessed by SNK test. (B) Data are means ± SE (n = 3 wells) from a representative experiment. P < 0.01 as assessed by one-way ANOVA. *P < 0.05 versus cells treated with LPS only, as assessed by SNK test.

View larger version (21K): [in this window] [in a new window]   Figure 9.   The inhibitory effects of antisense against TNF- on LPS-induced TNF- and MIP-2 production are highly correlated. Cells were preincubated with various concentrations of antisense TNF- ON for 4 h with Lipofectin, then challenged with LPS (10 µg/ml) for another 4 h. The percentage of inhibition of antisense TNF- on LPS-induced MIP-2 production is plotted against that on TNF- production. Data are combined from three experiments.

Neutralization of Secreted TNF- with Antibodies Suppresses MIP-2 Production

To further confirm the role of TNF- in LPS-induced MIP-2 production, we treated cells with neutralizing antibodies for TNF- before and during LPS stimulation. Cells were incubated with 1:1,000 dilution of anti-TNF- polyclonal antibody (36) (equivalent to about 10 µg/ml of IgG) or 1 µg/ml of anti-TNF- monoclonal antibody. Both polyclonal and monoclonal antibodies significantly suppressed MIP-2 production from alveolar epithelial cells (P < 0.05) (Figure 10). When the concentrations of polyclonal or monoclonal antibody were increased to 1:30 dilution or 10 µg/ml, respectively, inhibitory effects remained at the same levels (data not shown). Nonspecific IgG corresponding to each antibody had no inhibitory effect (Figure 10).

View larger version (17K): [in this window] [in a new window]   Figure 10.   Antibodies against TNF- inhibited LPS-induced MIP-2 production from alveolar epithelial cells. Cells were pretreated with polyclonal (1:1,000 dilution, equivalent to about 10 µg/ml of IgG concentration) or monoclonal (1 µg/ml) antibodies against TNF- for 1 h, then challenged with LPS (10 µg/ml) for another 4 h with antibodies. To exclude possible nonspecific effects of antibodies, cells were pretreated with either nonimmune rabbit IgG (10 µg/ml) or rat IgG (1 µg/ml) as negative controls. The concentrations of MIP-2 in the medium were measured by ELISA. Data are presented as percent inhibition ratio to LPS- induced MIP-2 (means ± SE, n = 3). P < 0.01 as assessed by one-way ANOVA. *P < 0.05 versus cells treated with LPS only, as assessed by SNK test.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The vigorous recruitment and activation of inflammatory cells in the alveolar space characterize effective host defense against invasion of bacteria and other pathogens. Paradoxically, however, overreaction of the host defense can contribute to development of acute lung injury. A better understanding of mechanisms that control the cytokine network in the lung will provide clues for better management and prevention of acute lung injury. It has been recognized that cytokines can be released and sensed by many cell types in the body. For example, intestinal epithelial cells have been suggested as sensors of pathogens and local regulators in host defense (37). Similar mechanisms may also exist in the lung.

Alveolar Epithelial Cells as a Source of MIP-2

In the present study we demonstrated that alveolar epithelial cells can produce MIP-2 in vitro in response to LPS or TNF- stimulation. Although rat alveolar macrophages can express MIP-2 mRNA (38), for a number of reasons we do not think that macrophages are the source of the MIP-2 in our study. With the cell culture conditions we used, the number of macrophages in the culture at the time of LPS stimulation was almost undetectable. The purity of alveolar epithelial cells at the time of LPS stimulation was confirmed with phase-contrast microscopy and immunofluorescent staining for cytokeratin, an intermediate filament of cytoskeleton specifically expressed in cells of epithelial origin, and pro-SP-C, an intracellular precursor of SP-C. In addition, MIP-2 protein was identified intracellularly throughout the cell layer and localized intracellularly, as determined by confocal microscopy. The optimal concentration of LPS that stimulates MIP-2 release from the murine macrophage RAW264.7 cell line was 1 µg/ ml with 6 h of incubation (35), which was very similar to what we found from primary cultured alveolar epithelial cells. The maximal release of MIP-2 from alveolar epithelial cells was also comparable to that of RAW264.7 cells. These results further excluded the possibility that MIP-2 released into culture medium was from the trace amount of contaminated alveolar macrophages. The similarity of capacity and kinetics of MIP-2 production between alveolar epithelial cells and macrophages suggests that alveolar epithelial cells could be one of the major sources of MIP-2 in the lung upon appropriate stimulation.

These observations are in accord with other data in the literature. For example, using an in situ hybridization technique, Driscoll and coworkers found that exposure of rats to -quartz induced expression of MIP-2 mRNA in the epithelial cells lining the terminal bronchioles and alveolar ducts as well as macrophages and alveolar type II cells in the more distal lung (20). They also showed that LPS and TNF- stimulation increased mRNA of MIP-2 in primary cultured type II cells from rat lung and a rat alveolar type II cell line, RLE-6TN (20). Although protein levels of MIP-2 were not measured in these studies, it is possible that type II cells are a source of MIP-2 both in vitro and in vivo.

Our results show that the basal levels of MIP-2 mRNA expression and protein in the cell and released into the culture medium were very low, were increased after LPS challenge, and were blocked significantly by inhibitors for either transcription or translation. The increase of MIP-2 protein content was much more significant than that of the MIP-2 mRNA level, which further supports the presence of a regulatory mechanism of MIP-2 expression at the post-transcriptional level. This de novo synthesis of MIP-2 in alveolar epithelial cells is likely true in vivo, because no expression of MIP-2 mRNA was found by in situ hybridization from control rats, whereas -quartz challenge rapidly induced MIP-2 expression in lung epithelial cells (20).

TNF- as a Regulator for Chemokine Production in Alveolar Epithelial Cells

The most intriguing observation in the present study was that downregulation of TNF- with the antisense ON targeting TNF- inhibited LPS-induced MIP-2 production. Neutralization of TNF- with antibodies also suppressed MIP-2 production. The direct stimulatory effect of TNF- and the indirect inhibitory effect of antisense and antibodies against TNF- on MIP-2 production suggest that LPS-induced MIP-2 production be mediated in part by TNF- synthesized from the same cell type. The exogenous TNF- required to stimulate significant MIP-2 production was 1 to 5 ng/ml (Table 2), which was higher than that of secreted TNF- (less than 0.5 ng/ml; Figure 7B) from these cells upon LPS stimulation. Therefore, in addition to the TNF--dependent mechanism, LPS may induce MIP-2 through other pathways. The inhibitory effect of antisense directly against TNF- on MIP-2 production was as high as 60%, suggesting the importance of TNF- to LPS- induced MIP-2 production. The inhibitory effect of anti- TNF- antibodies on MIP-2 production was less than that of antisense ON. One possibility is that TNF- produced by alveolar epithelial cells may stimulate MIP-2 production intracellularly. Microinjection of TNF- to normal macrophages or the J774 macrophage-like cell line induced rapid cell death (39), suggesting that TNF- may have an intracellular activity. The role of TNF- as an intermediary regulating MIP-2 production has been implicated on the basis of several observations in vitro and in vivo. For example, injection of TNF- into subcutaneous air pouches of mice, a model to study acute inflammation, resulted in an increase in MIP-2 and several other chemokines (40). Respiratory syncytial virus infection in macrophage RAW264.7 cells induced MIP-2 production in a biphasic pattern. The presence of anti-TNF- antibody partially inhibited the late phase (41). Incubation with anti-IL-1 and anti-TNF- antibodies together reduced acute ozone exposure-induced MIP-2 gene expression from alveolar macrophages (38).

Recruitment and activation of mononuclear phagocytes and neutrophils are critical regulatory events for control of pulmonary inflammation. In addition to MIP-2, there is indirect evidence that TNF- may also mediate LPS-induced production of other chemokines from type II cells. The increase of TNF- in the lung preceded that of MCP-1 (42) and RANTES (43) after LPS challenge. When animals were treated with soluble TNF receptor/IgG construct (as a TNF- antagonist) before LPS stimulation, there was a 60% reduction in steady-state levels of either MCP-1 (44) or RANTES (45) mRNA in lung homogenates. A human lung epithelial cell line (A549) has been used as a model to study functions of alveolar epithelial cells. The behavior of this cell line, as well as other cell lines, varies from primary cultured cells. For example, LPS induced MCP-1 from primary cultured rat type II cells (7) but not from A549 cells (9). Similarly, LPS induced MIP-2 from primary cultured rat lung alveolar epithelial cells, but not IL-8, a human C-X-C chemokine, from A549 cells (46). When A549 cells were stimulated with TNF-, both MCP-1 (9) and IL-8 (46) could be induced. We have shown that in response to LPS stimulation primary cultured alveolar epithelial cells produced TNF- (4), but A549 cells did not (our unpublished observation). It seems that LPS-induced chemokine production from these two cell types is correlated to their ability to produce TNF- upon LPS stimulation.

Together, these results suggest there is an autoregulatory mechanism by which bacteria, endotoxin, virus, and other pathogens may activate the production of chemokines with TNF- as an intermediary. We propose that TNF- produced in LPS-stimulated alveolar epithelial cells may mediate the production of chemokines such as MIP-2 to recruit and activate inflammatory cells to the site of infection. Endotoxin-induced TNF- production in alveolar epithelial cells may therefore be considered as an alert signal in the host defense in the lung (Figure 11). The amount of chemokines released, such as MIP-2, from alveolar epithelial cells is much greater than TNF-. Thus, TNF- produced by alveolar epithelial cells could be amplified, leading to a greater release of chemokines to recruit inflammatory cells from the circulation. Under certain pathologic circumstances, the autoregulatory mechanism may be exaggerated, and may contribute to the acute inflammatory injury.

View larger version (23K): [in this window] [in a new window]   Figure 11.   An autoregulatory mechanism of alveolar epithelial cells in host defense. In response to endotoxin (LPS) released from bacteria (1), alveolar epithelial cells, especially type II pneumocytes, may produce small amounts of TNF- (2), which subsequently participate in the activation of production of chemokines (3) such as MIP-2, from the same cell type (4). Chemokines will recruit and activate monocytes and neutrophils (5) to the alveolar space (6). During acute inflammatory reaction, this mechanism may be exaggerated and may contribute to acute lung injury.

The Potential of Using Antisense TNF- as a Therapeutic Approach

Blockade of LPS-induced TNF- release from pulmonary epithelial cells with antisense ONs may be a potential therapeutic approach to protect the lung from acute injury. Antisense ONs have been developed as a potent approach to selectively block the synthesis of proteins of interest. A short single strand of DNA ONs, usually 15 to 20 nucleotides in length, can be delivered into cells and bind to the specific RNA or DNA to block its function (44). Although at present the use of antisense technology is primarily as a tool for dissecting the functions of proteins in vitro, extensive effort is under way to develop and apply antisense ONs as therapy for a number of diseases (45, 47). Therapeutic application of antisense technology has been used to inhibit inflammatory cytokines both in vitro and in vivo (48, 49). Because LPS-induced TNF- release seems to play a central role in regulating the production of other cytokines and chemokines, the antisense ON targeting TNF- may represent a novel therapeutic approach in a number of inflammation disease states.