Induction of Cyclooxygenase-2 in Alveolar Macrophages after Acid Aspiration : Selective Cyclooxygenase-2 Blockade Reduces Interleukin-6 Production

All animal experiments were performed in compliance with the guidelines given by the Committee on Animal Research of the University of California, San Francisco.

Eight- to twelve-week-old male C57BL/6 mice (body weights, 25–30 g) were purchased from Simonsen Laboratories (Gilroy, CA). The mice were fed regular rodent chow and water ad libitum. Mice were anesthetized briefly with inhaled methoxyflurane (Metofane; Pitman-Moore, Inc., Mundelein, IL). The acid group was instilled with 75 micro liter hydrochloric acid solution (pH 1.15, 0.1 N, endotoxin-free; Sigma Chemical Co., St. Louis, MO) mixed with normal saline (normal saline:hydrochloric acid ratio was 1:2). The saline group was instilled with 75 micro liter of one third normal saline solution (normal saline: H²O ratio was 1:2) into one lung via a blunt end feeding needle (24 G; Popper and Sons, Inc., New Hyde Park, NY) inserted orally. [18]Mice were allowed to recover from anesthesia and breathe room air spontaneously for the experimental period. The treatment group received a specific cyclooxygenase-2 blocker, N-2(-cyclohexyloxy-4-nitrophenyl) methane-sulphonamide (NS-398; Biomol, Plymouth Meeting, PA); each mouse was given 250 micro gram intraperitoneally, 2.5 mg/ml suspended in dimethyl sulfoxide, 1 h before acid instillation. [12,19]The nontreatment group received sham treatments of 100 micro liter dimethyl sulfoxide by intraperitoneal injection.

Mice were reanesthetized 2 and 4 h after acid instillation with 50 mg/kg of pentobarbital sodium. Their tracheas were exposed, and 20-gauge stub adapters were inserted and secured. Lungs were lavaged with cold 0.1% ethylenediaminetetraacetic acid/phosphate-buffered saline (PBS; 5 ml). Lavage fluid from two or three mice was pooled to obtain enough alveolar macrophages for analysis. Lavage fluid was centrifuged for 10 min at 4 [degree sign] Celsius at 1,000 x g to separate the cells from supernatant. Cells were washed and resuspended in PBS (5 ml). Alveolar macrophage numbers and viability were quantified by hemacytometer with trypan blue staining (> 90%). Macrophage concentration was adjusted to x 10⁶cells/ml in RPMI1640 supplemented with 10% heat-inactivated fetal bovine serum and antibiotic agents. Macrophages were purified by allowing them to adhere to the bottom of the wells by incubation for 1 h in multiwell tissue culture plates. Nonadherent cells were removed by replacing the RPMI medium.

Total ribonucleic acid (RNA) was extracted from 5 x 10⁵freshly harvested alveolar macrophages with TRIzol reagent (Gibco BRL, Gaithersburg, MD). RNA concentrations were measured carefully by a microspectrophotometer (Spectramax; Molecular Devices, Menlo Park, CA), and 2.0 micro gram total RNA was reverse transcribed using reverse transcriptase, M-MLV (Promega, Madison, WI), and oligo-dT¹²-18 primers (Gibco BRL). The following specific primers for cyclooxygenase-1 and cyclooxygenase-2 were made for polymerase chain reaction (PCR):

cyclooxygenase-1

5'-primer: 5'-GGT TGA GGC ACT GGT GGA TGC CTT C-3'

3'-primer: 5'-AGA CAG ACC CGT CAT CTC CAG GGT A-3'

cyclooxygenase-2

5'-primer: 5'-ACT CAC TCA GTT TGT TGA GTC ATT C-3'

3'-primer: 5'-TTT GAT TAG TAC TGT AGG GTT AAT G-3'.

The size of the PCR products from cyclooxygenase-1 and cyclooxygenase-2 are 524 and 583 base pairs, respectively. The specificity of the PCR reaction for cyclooxygenase-1 and cyclooxygenase-2 complementary deoxyribonucleic acid (cDNA) was confirmed by digestion of PCR products by sequence-specific restriction enzymes (cyclooxygenase-1: Sac I and Sma I; cyclooxygenase-2: Pst I and Nco I). For quantitation of cyclooxygenase-1 and cyclooxygenase-2 mRNA levels, the Escherichia coli plasmid pT2M containing specific internal standards for cyclooxygenase-1 and cyclooxygenase-2 was constructed by internal deletion of sequences: The truncated fragments (320 base pairs) from cyclooxygenase-1 and cyclooxygenase-2 PCR products were incorporated into the E. coli cloning plasmid pBSII SK(+)(Stratagene, La Jolla, CA) at the EcoR I and Kpn I restriction sites and multiplied as plasmids in an E. coli strain Dh5 alpha. Plasmid solutions were adjusted to 1 x 10⁴molecules/micro liter after double digestion by restriction enzymes (EcoR I and Kpn I) and used as internal standards for competitive PCR. For IL-1beta, IL-6, tumor necrosis factor (TNF)-alpha, inducible nitric oxide synthase (iNOS), and hypoxanthine-guanine phospho-ribosyltransferase (HPRT), a multiple cytokine competitor construct pLOC (provided by David B. Corry, University of California, San Francisco, San Francisco, CA) was used with specific cytokine primers as described. [20,21]We obtained confirmation that the cDNA samples contained the same amount of DNA by measuring the constitutively expressed HPRT gene first. The amounts of cDNA for cyclooxygenase-1, cyclooxygenase-2, and the various cytokine genes were measured based on this standardization, and the ratio of wild-type genes to competitor genes was amplified by PCR. For HPRT measurement, 1 x 10⁵competitor molecules were added. Polymerase chain reaction was performed for 35 cycles with the following conditions: 30 s at 94 [degree sign] Celsius for denaturing DNA, 20 s at 55–60 [degree sign] Celsius annealing temperature, and 40 s at 72 [degree sign] Celsius for extension of the reaction. Products of PCR were separated by electrophoresis on 1.5% agarose gels and visualized by ethidium bromide staining.

Alveolar macrophages were harvested 4 h after instillation. Cells were counted with a hemacytometer, and cell viability was calculated. After washing the cells once with PBS, they were resuspended in PBS, and 1 x 10⁵cells were attached to glass slides by cytospin preparation. Slides were air-dried and stored at -80 [degree sign] Celsius until staining. Immunocytochemistry was performed using the avidin-biotinylated alkaline phosphatase complex method (Vectastain ABC-AP complex kit; Vector, Burlingame, CA). Frozen slides were brought to room temperature and fixed in 2% paraformaldehyde. Cells were permeabilized with 0.1% triton-X/PBS. All washing procedures were performed with high-salt PBS (2.7% NaCl/PBS). Blocking of nonspecific reactions to the secondary antibody was performed with 5% normal goat serum plus 10% normal horse serum in PBS. Slides were incubated with the primary antibody (polyclonal anti-PGHS-2, polyclonal rabbit immunoglobulin G; or control, preimmune rabbit serum; Oxford Biomedical Res., Inc., Oxford, MI) at 1/50 dilution in the blocking buffer (overnight at 4 [degree sign] Celsius). After the secondary antibody (goat anti-rabbit immunoglobulin G, biotinylated anti-rabbit immunoglobulin G; Vector) was applied (1 micro gram/ml in blocking buffer for 30 min at room temperature), slides were incubated with the avidin-biotinylated alkaline phosphatase complex and developed with the substrate (Fast Red TR/Naphthol AS-MX; Sigma). After mounting, slides were viewed at high magnification and photographed. Cells were counted at low magnification (100 cells per visual area), and the ratio of positively stained cells was calculated.

Alveolar macrophages harvested from mice as described in the harvesting method were cultured in RPMI1640 at 37 [degree sign] Celsius in a humidified atmosphere containing 5% CO²for 24 h at a concentration of 1 x 10⁶cells/ml, 200 micro liter/well, in 96-well tissue culture plates. When indicated, NS-398, a specific cyclooxygenase-2 blocker, was added to the culture medium at 100-micro Meter concentration at the beginning of the incubation. Culture media was collected after 24 h of incubation and stored at -80 [degree sign] Celsius until assayed for PGE², cytokines, or nitrite. Cell viability was 90–95% at the end of the 24-h incubation period, in media with or without NS-398.

Prostaglandin E²was measured in the supernatant of the primary culture medium of alveolar macrophages by PGE²enzyme immunoassay kit (Cayman Chemical, Ann Arbor, MI). Each sample was measured in triplicate. The specificity of this assay for PGE²is 100%. The lower detection limit was 5 pg/ml.

Enzyme-linked immunosorbent assays were performed for IL-1beta, IL-6 and TNF-alpha as previously described. [18]The supernatant of the alveolar macrophage culture was used for measurements. For IL-6, microtiter plates (enzyme-linked immunosorbent assay plate, Easy wash; Corning, Cambridge, MA) were coated overnight at 4 [degree sign] Celsius with capture antibody (rat anti-mouse IL-6 monoclonal antibody, MP5–20F3, rat immunoglobulin G¹; PharMingen, San Diego, CA) and blocked with 3% bovine serum albumin/PBS for 2 h at room temperature. Recombinant standards and samples were then added to plates and incubated for 4 h at room temperature. After incubation with biotinylated developing antibody (rat anti-mouse IL-6 monoclonal antibody, MP5–32C11, rat immunoglobulin G^2a; PharMingen) for 45 min at room temperature and alkaline phosphatase-conjugated streptavidin (Jackson ImmunoResearch Laboratories, West Grove, PA) for 45 min at 37 [degree sign] Celsius, plates were developed with p-nitrophenyl-phosphate (Sigma 104) dissolved in 0.1 M alkaline buffer solution (2-amino-2-methyl-1-pro-panol buffer, 1.5 M, pH 10.3, Sigma 221) and read at an OD of 405 nm using a microplate reader. Plates were washed two to six times with 0.05% Tween/PBS between each procedure. Sample concentrations were calculated by comparison with a standard curve of mouse recombinant IL-6 (PharMingen). Interleukin-1beta and TNF-alpha were measured with the same methods using the following reagents: rabbit anti-mouse IL-1beta polyclonal antibody, biotinylated anti-mouse IL-1beta monoclonal antibody (1400.24.17, mouse immunoglobulin G¹), recombinant mouse IL-1beta (all from Endogen, Cambridge, MA), rat anti-mouse TNF-alpha monoclonal antibody (MP6-XT22, rat immunoglobulin G¹), biotinylated rabbit anti-mouse TNF-alpha polyclonal antibody, and recombinant mouse TNF-alpha (all from PharMingen). Each sample was measured in triplicate. Lower detection limits for IL-1beta, IL-6, and TNF-alpha were 50.0, 1.0, and 5.0 pg/ml, respectively.

Nitrite concentrations in the supernatant of the alveolar macrophage culture were measured by Griess reaction. The reagents were prepared immediately before assay. One hundred microliters of supernatant was mixed with 100 micro liter Griess reagent (final concentration, 1% sulfanilamide, 0.1% n-(1-naphthyl)-ethylene-diamine, 1.7% H³PO⁴) and incubated for 30 min in the dark. Absorbance was measured at 550 nm using a microplate reader. Concentrations were determined using a NaNO²standard (0.3–100.0 micro Meter).

Results are expressed as mean +/- SD. Comparisons between groups were performed with a one-way analysis of variance, and a Bonferroni/Dunn test was used for multiple comparison. A probability value < 0.05 was considered statistically significant and is identified with an asterisk in the figures.

In the control group, mRNA levels for the cyclooxygenase-1 and cyclooxygenase-2 genes in alveolar macrophages were lower than 10% of the HPRT mRNA (Figure 1(B), rows 1 and 2). Saline instillation did not affect these levels. Two hours after acid instillation, however, a larger amount of cyclooxygenase-2 mRNA was detected; this large amount was the same at 4 h (Figure 1(B), row 2). In contrast, the cyclooxygenase-1 mRNA was not increased after acid instillation (Figure 1(B), row 1).

We found more IL-1beta gene expression in alveolar macrophages from mice instilled with either saline or acid compared with the control mice (Figure 1(B), row 3). There was a slight increase in IL-6 gene expression in alveolar macrophages transiently 2 h after saline instillation, but the level was the same as that of controls at 4 h. In the acid group, however, there was more IL-6 gene expression at 2 h after acid instillation compared with the control group, and it remained at this level after 4 h (Figure 1(B), row 4). Messenger RNA levels of TNF-alpha and iNOS genes were not different among control, saline, and acid groups (Figure 1(B), rows 5 and 6).

Two different staining patterns for cyclooxygenase-2 in alveolar macrophages were detected from immunocytochemical analysis. One pattern was the localized staining of the perinuclear membrane, suggesting minimal expression of cyclooxygenase-2 (Figure 2(A)). The other pattern was strong staining not only in the perinuclear region but also in the cytoplasmic region, indicating high expression of cyclooxygenase-2 (Figure 2(B)). Almost all macrophages showed baseline regional perinuclear cyclooxygenase-2 expression. None of the alveolar macrophages from control mice showed cytoplasmic expression of cyclooxygenase-2. In the saline group, < 1% of the cells (0.8 +/- 1.5%) showed cytoplasmic expression. Significantly more alveolar macrophages (11.3 +/- 4.4%) harvested from mice instilled with acid showed the strong cytoplasmic cyclooxygenase-2 expression (P < 0.01).

In 24-h culture media, we detected a trend of release of more PGE²from alveolar macrophages harvested 1 h after acid instillation compared with saline instillation (Figure 3). A significantly higher level of IL-6 production also was detected in the alveolar macrophages harvested after acid instillation compared with the saline group (Figure 4). The concentrations of IL-1beta and TNF-alpha were lower than detectable levels in the culture media from all groups of macrophages (data not shown). Low concentrations of nitric oxide metabolites (NO²sup - < 3 micro Meter) were detected in all groups of the macrophage culture, and there was no significant difference between groups (data not shown).

Pretreatment of mice with the selective cyclooxygenase-2 blocker, NS-398, resulted in the PGE²production from alveolar macrophages harvested after acid instillation being at the same low level as that occurring with the addition of the cyclooxygenase-2 blocker to the macrophage culture medium (Figure 3). Interleukin-6 release from alveolar macrophages harvested from mice instilled with acid was significantly reduced by pretreatment with NS-398 (Figure 4). Although cyclooxygenase-2 blockade did not change mRNA levels of the cyclooxygenase-2 gene in the alveolar macrophages harvested from mice instilled with acid, decreases of IL-6 mRNA levels were detected at 2 and 4 h after acid instillation (Figure 1(B)).

The role of prostaglandins in acid aspiration remains controversial. [22]In canine acid aspiration models, nonsteroidal antiinflammatory drugs have been used successfully to reduce lung edema and improve pulmonary function, [7,23]whereas other investigators have found that nonsteroidal antiinflammatory drugs improved the physiologic status of the lung but did not change mortality in sepsis. [24] 

Nonspecific blockade of both cyclooxygenase-1 and -2 results in the well-known toxicities of gastric ulceration and renal and platelet dysfunction. [12]Since the cloning of cyclooxygenase-2, compounds have been developed that selectively block cyclooxygenase-2 while preserving the functions of cyclooxygenase-1. [12,19,25,26]The specific cyclooxygenase-2 inhibitor we used in our study, NS-398, is a highly selective inhibitor of cyclooxygenase-2 and forms an irreversibly inhibited complex. [19]In the acute phase of an inflammatory response (2–4 h after stimuli), the increase in prostaglandin production is entirely dependent on cyclooxygenase-2 protein synthesis. [16]In a carrageenin-induced subcutaneous air pouch model [20]and a pleurisy model in rats, [26]these specific cyclooxygenase-2 blockers were used successfully to reduce inflammation, suggesting the possibility of treating various inflammatory disorders with these new compounds without the hazards of conventional nonsteroidal antiinflammatory drugs.

In this study, significant induction of cyclooxygenase-2 was detected in alveolar macrophages after acid aspiration. First, the marked upregulation of cyclooxygenase-2 was confirmed by measurement of mRNA levels using a highly sensitive method, competitive PCR. [27]Second, intense expression of the cyclooxygenase-2 protein in alveolar macrophages after acid aspiration was demonstrated by specific cyclooxygenase-2 immunocytochemical staining. Cells with less intense cytoplasmic staining were likely lavaged from the contralateral lung, which did not receive acid. The source of increased production of PGE sub 2 was identified as cyclooxygenase-2 by selective blockade of cyclooxygenase-2 with NS-398.

Recently, the interactions between prostaglandins produced by cyclooxygenase-2 and various inflammatory cytokines or nitric oxide have been examined in vivo and in vitro. [16,17,28–30]In this model, despite the fact that an increased mRNA level of IL-1beta was detected in alveolar macrophages after acid aspiration, the relation of IL-1beta released from alveolar macrophages to cyclooxygenase-2 upregulation was unclear, as similar upregulation also was observed in alveolar macrophages from mice instilled with saline. In addition, no detectable concentrations of IL-1beta were measured in the macrophage culture medium. We found a close relationship between cyclooxygenase-2 and IL-6, however; increased production of IL-6 in alveolar macrophages after acid aspiration was prevented by cyclooxygenase-2 blockade. The roles of TNF-alpha [1,31]and nitric oxide in acid aspiration, a chemical injury, remain controversial. [32]These mediators are implicated in lung injury from bacteria; however, we did not detect upregulation of TNF-alpha or iNOS even with our extremely sensitive competitive PCR method. We conclude that these mediators are not involved in the early phase of inflammation in our model of acid aspiration. Others have shown in a rabbit model of acid lung injury that IL-8 is a major contributor in the inflammatory process. [3]The homologous protein for IL-8 has not been identified in the mouse. Mediators of the chemokine family (KC, MIP-2, and others), however, may be released from alveolar macrophages to play a role in acute lung injury.

Using a murine model of acid aspiration, we have demonstrated that acid aspiration induces cyclooxygenase-2 in alveolar macrophages. We have demonstrated that IL-6 production from alveolar macrophages was significantly increased after acid aspiration. Using a specific cyclooxygenase-2 inhibitor, we also have shown that cyclooxygenase-2-derived prostanoids were responsible for induction of macrophage IL-6 production. It is well established that IL-6 plays a major role in the propagation of inflammatory responses, such as fever induction in the hypothalamus and induction of the acute phase response in the liver. [33,34]Interleukin-6-deficient mice were found to have impaired fever induction and acute phase response to inflammatory stimuli. [35]In addition, IL-6 production correlates with clinical outcome in the systemic inflammatory response syndrome. [36,37]In this model, the upregulation of IL-6 was reduced by selective cyclooxygenase-2 blockade. This result is consistent with previous reports using a rat arthritis model [17]and peritonitis model [13]in that induction of cyclooxygenase-2 is responsible for the increase in PGE²concentrations, which in turn induces IL-6.

We conclude that cyclooxygenase-2 is strongly induced in alveolar macrophages after acid aspiration, and specific blockade of cyclooxygenase-2 may be used in decreasing production of inflammatory mediators such as PGE²and IL-6, which may cause tissue injury.

The authors thank Richard Shanks for technical assistance.