Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Cyclooxygenase, which catalyzes the initial metabolic step in the formation of prostaglandins, has traditionally been regarded as an enzyme that is passively present in cells reacting to the varying supply of the precursor arachidonic acid. This feature appears to hold true in many cells for the isoform cyclooxygenase 1 (Cox-1), which is regulated in the course of developmental processes, but seems to be constantly expressed under most (patho)physiological conditions (1, 2). In contrast, the second isoform, cyclooxygenase 2 (Cox-2), has emerged as a dynamically regulated enzyme, readily inducible by inflammatory and endocrine factors such as lipopolysaccharides, tumor necrosis factor, or interleukin 1 in cell culture studies (2). Thus, tentatively, Cox-1 appears to be predominantly involved in physiological and regulatory processes, whereas Cox-2 is centrally implicated in inflammatory processes (6). Discrimination between the roles of Cox-1 and Cox-2 is also of major importance for the further development of antiinflammatory strategies, as current nonsteroidal antiinflammatory drugs (NSAIDs) cause a nonselective inhibition of both isoenzymes with unwanted side effects due to intervention with regulatory processes, but tools for selective inhibition of Cox-2 are now becoming available (10).

Cyclooxygenase-dependent pathways are centrally involved in a large array of physiological and pathophysiological processes in the lung. These include the regulation of lung vascular tone including perfusion-ventilation matching (13), vascular and interstitial tissue remodeling (16, 17), the regulation of capillary endothelial and alveolar epithelial permeability (18, 19), surfactant homeostasis (20), macrophage-related inflammatory and host defense events in the alveolar compartment (21, 22), the control of bronchial mucous secretion and transport (23, 24), and last (but not least), the regulation of bronchial tone including the pathogenesis of such socioeconomically important diseases as bronchial asthma and chronic obstructive lung disease (25, 26). Against this background, it is not surprising that a study addressing mRNA expression for Cox-1 and Cox-2 in the homogenates of different human tissues found the lung, together with prostate, to be the organ with the highest levels of cyclooxygenase (27). Interestingly, however, the levels of Cox-2 surpassed even those of Cox-1, although the homogenates originated from trauma victims with nondisturbed lung function, thus questioning the preceeding assumption that Cox-2 is implicated predominantly in inflammatory processes.

Because lung tissue is composed of a large number of different cell types, (at least 40), analysis of whole tissue homogenate will of course not enable researchers to ascribe Cox-1 and Cox-2 expression to specific cellular and functional activities. Information from lung cell culturing in vitro has hitherto been limited. Low Cox-2 expression was found in human airway epithelial cells, which markedly increased in response to in vitro lipopolysaccharide (LPS) stimulation (28). In alveolar macrophages, obtained by bronchoalveolar lavage, baseline Cox-2 activity was missed, but again became detectable in response to in vitro LPS challenge (5, 22, 29, 30). Fibroblasts isolated and cultured from patients undergoing resectional lung surgery generated prostaglandins in vitro after stimulation by cytokines, presumably via induction of Cox-2 (4). To our knowledge, no previous studies address the cellular localization sites of Cox-1 and Cox-2 in the intact lung. The current study uses immunohistochemistry with immunogold-silver staining and quantitative image analysis to localize the expression of both Cox-1 and Cox-2 in the lung tissue of normal rats. The quantitative assessment allowed the performance of an objective, observer-independent evaluation of the staining intensity of the stained structures for each antibody. The study is limited in that, because of the unknown antigen-binding capacity of the polyclonal antibodies used, a direct comparison between the staining intensity of Cox-1 and Cox-2 for the different cell types is not possible. The technique does, however, allow differences within the Cox-1 and Cox-2 immunoreactivities to be addressed, and visualization of the pattern of distribution of these isoenzymes.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Reagents

The Cox-2 antibody (PG26, polyclonal rabbit IgG) and preimmune serum (PG26c) were obtained from Oxford Biomedical Research (Oxford, MI), and the silver enhancement solution was acquired from Aurion (Wageningen, The Netherlands). The Cox-2 antibody is directed against a peptide corresponding to the amino acid sequence 581-598 of the carboxy-terminal region of Cox-2 (5). The Cox-1 antibody (affinity-purified goat polyclonal IgG) was obtained from Santa Cruz Biotechnology (Cox-1, M-20; Santa Cruz Biotechnology, Santa Cruz, CA). The Cox-1 antibody was prepared against a peptide corresponding to the amino acid sequence 583-602, mapping at the carboxy terminus of Cox-1. Both antibodies bind specifically to one of the isoenzymes and do not cross-react. All other biochemicals were obtained from E. Merck AG (Darmstadt, Germany).

Animals

CD rats (Sprague-Dawley) were obtained from Charles River, Laboratories (Sulzfeld, Germany). All experimental procedures were performed in conformity with the guidelines of the U.S. National Institutes of Health (31).

Lung Isolation and Perfusion

The rats (male, body weight 350-400 g) were deeply anaesthetized with pentobarbital sodium (100 mg/kg body weight, intraperitoneal). Following local anesthesia with 2% xylocaine and median incision, the trachea was dissected and a tracheal cannula was immediately inserted. A median laparatomy was performed and subsequently the rats were then anticoagulated with 1,000 units of heparin. After midsternal thoracotomy, the right ventricle was incised, a cannula was fixed in the pulmonary artery, and the apex of the heart was cut off to allow pulmonary venous outflow. Simultaneously pulsatile perfusion with buffer solution was started. The buffer contained 2.4 mM CaCl₂, 1.3 mM MgCl₂, 4.3 mM KCl, 1.1 mM KH₂PO₄, 125.0 mM NaCl, 25 mM NaHCO₃, as well as 13.32 mM glucose (pH ranged between pH 7.35 and 7.40). The lungs were continuously perfused for about 5 min for washout of blood.

Tissue Preparation for Immunohistochemistry and Reverse Transcription-Polymerase Chain Reaction

A phosphate-buffered saline (PBS: 0.1, pH 7.4)-buffered 3% paraformaldehyde solution was instilled into the trachea for fixation at a pressure of 20 cm H₂O; subsequently, the lungs were immersed in the same fixation solution for 1 h (n = 5). In addition, one lung was snap frozen in liquid nitrogen for cryomicrotomy. Another five lungs were snap frozen in liquid nitrogen and stored at 80°C until they were prepared for application of the polymerase chain reaction (PCR) technique.

For paraffin embedding the whole lungs were dissected in tissue blocks from all lobes. The tissue blocks were embedded in low-temperature paraffin with a melting temperature of 40-42°C. Sectioning (10 µm) was performed on all paraffin-embedded and frozen tissue blocks.

Immunohistochemistry

The paraffin sections were dewaxed, rehydrated, and washed three times (5 min each) in PBS (0.01 M phosphate, 150 mM NaCl, pH 7.6). They were treated for 15 min with a 1% Triton solution. The sections were preincubated in PBS containing 5% goat serum, 0.025% bovine serum albumin type C (BSA-C), 0.05% Tween 20, and 0.02 M glycine to block unspecific binding. Overnight incubation with the polyclonal primary antibody rabbit-anti PGHS-2 (PG26, polyclonal; Oxford Biomedical Research) diluted 1:50 in PBS containing 0.025% BSA-C-0.05% Tween 20, was carried out at 4°C. Cox-1 was correspondingly addressed with the polyclonal primary antibody goat anti-Cox-1 (Cox-1, M-20; Santa Cruz Biotechnology) diluted 1:200 in PBS dilution buffer. The sections were then washed in PBS and incubated overnight at 4°C for Cox-2 with goat anti-rabbit F(ab)₂ gold conjugate (Aurion) diluted 1:400, and for Cox-1 with rabbit anti-goat F(ab)₂ gold conjugate (Nanoprobes, Stony Brook, NY; Biotrend, Cologne, Germany) diluted 1:400. The sections were then washed in PBS again and fixed for 5 min in 2% phosphate-buffered glutardialdehyde. After several washes in glass-double-distilled water the sections were incubated in Silver Enhancer solution (Aurion) for 50 min. Counterstaining of the sections was performed with nuclear fast red. Control staining was performed by omission of the primary antibody or by substitution with unspecific preimmune serum at the same dilution. Counterstaining with alcian blue was performed with some of the Cox-2 sections for identification of a special type of macrophage-like cell, which was intensely stained by anti-Cox-2 staining. After silver enhancement these sections were washed for 3 min in 3% acetic acid and afterward stained for 30 min with 1% alcian blue solution. Counterstaining was performed with nuclear fast red as described previously.

Image Analysis

By means of an image analysis system, composed of an 80486 host computer with an MFG-grabber board (ITI, Bedford, MA) and a Sony 3-CCD (DXC-930P) color video camera mounted on a Leitz Ortholux II (Leitz, Wetzlar, Germany), images of the immunohistochemically stained sections were digitized. Microscope settings were kept constant throughout all measurements (objective: Leitz NPL Fluotar 340/1.30 oil, Leica IGS epipolarization filter block; illumination: 12-V tungsten-halogen lamp, 100 W, stabilized DC-transformer [Statron, Furstenwalde, Germany]; voltage setting: 10 V/8 A). Adjustment of all microscope settings and calibration of the measurement system with a reference slide were done before measurement. Epipolarization depiction of the silver-enhanced colloidal gold particles created a complete segmentation between positively stained and nonstained tissue. Gray-scale images were digitized to 8-bit accuracy, resulting in an intensity scale ranging from 0 to 255. Image analysis was performed by means of the image analysis program OPTIMAS (Optimas, Bothell, WA). Three randomly selected complete cross-sections from different lung lobes were measured. Structures and cell types were classified in the bright-field image and the corresponding epipolarization image. A measurement of the mean gray values of the selected structures was performed and subsequently the data were transferred into the spreadsheet program Excel (Microsoft). For direct visualization of staining intensity a pseudocolor scale with 11 colors was chosen, each representing an equal sector of the intensity scale, and applied to the images. Figure 1 shows the different steps of image acquisition and subsequent analysis.

View larger version (25K): [in this window] [in a new window]   Figure 1.   (a) Cox-2 immunogold-silver staining of vascular smooth muscle cells of partially muscular vessels. Scale bar: 20 µm. (b) Epipolarization image of (a). Note the intensive brightness of silver-enhanced immunogold staining. (c) Pseudocolor image of (b). A scale ranging from 0 to 255 with 11 colors was applied to the epipolarization image representing the gray values (see color bar).

Background measurement was performed to evaluate the influence of unspecific antibody binding.

Statistical Analysis

The statistical significance of differences between means was determined by analysis of variance (ANOVA). P < 0.05 was considered significant. The Bonferroni's multiple comparison test was used to determine which groups were significantly different from each other.

Detection of Cyclooxygenase mRNA: Reverse Transcription-Polymerase Chain Reaction

Tissue samples from lungs of individual animals were homogenized in a guanidine thiocyanate solution in an Ultraturrax (IKA, Stanfen, Germany). Total RNA was isolated by the guanidinium thiocyanate method. RNA yield and integrity were assessed by ultraviolet (UV) spectrophotometry and ethidium bromide staining. RNA (0.3 µg) was reverse transcribed with oligo(dT) primers as described (32). Primers specific for Cox-1 and Cox-2 were designed similar to those described by Lee and coworkers (5): Cox-1 (5' forward primer, GCA TGT GGC TGT GGA TGT CAT C; 3' reverse primer, CAC TAA GAC AGA CCC GTC ATC TCC) and Cox-2 (5' forward primer, CTC ACT CAG TTT GTT GAG TC; and 3' reverse primer, GAT TAG TAC TGT AGG GTT ATT G). These primer pairs were found to yield PCR products of 450 base pairs (bp) and 583 bp for Cox-1 and Cox-2, respectively. GAPDH primers were used as described (32). Amplified cDNA bands were detected by ethidium bromide staining and the volumes evaluated by densitometry. The yield of the amplified product was tested and found to be linear for the amount of input RNA and PCR cycle number. Twenty-seven to 29 cycles were used for Cox-1 and Cox-2 and 18 to 20 cycles were used for amplification of GAPDH.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Immunohistochemistry

Cox-1. A marked staining was observed in the bronchial epithelial cells of all bronchi, but was more intensive in the large bronchi of the first and second generation (Figure 2) than in the following branching generations (Table 1). Quantitative image analysis of the immunogold-silver-staining intensity demonstrated mean gray values of bronchial epithelial cell staining intensity to be twice as high in the large bronchi of the first and second generation than of the third generation and bronchioles. The data for bronchial epithelial cells of large bronchi were significantly different from all other stained structures except the vascular smooth muscle cells of the large hilum veins (P < 0.05). Bronchial smooth muscle cells exhibited a focal staining of low intensity. In addition, vascular smooth muscle cells of the large veins located at the hilum exhibited a strong staining reaction (Figure 3). Mean gray values of vascular smooth muscle cells of the large hilum veins even surpassed the values of bronchial epithelial cells. These muscle cells are structurally related to cardiac muscle cells and accompany the pulmonary veins of rats into the lung parenchyma (33, 34).

View larger version (48K): [in this window] [in a new window]   Figure 2.   (a) Cox-1 immunogold-silver staining of bronchial epithelial cells located in first- or second-generation bronchi. Scale bar: 20 µm. (b) Pseudocolor image of the accompanying epipolarization image. Note the exclusive staining of the epithelium.Figure 3. (a) Cox-1 immunogold-silver staining of vascular smooth muscle cells of a large vein at the lung hilum. The muscle cells structurally resemble cardiac myocytes. Scale bar: 20 µm. (b) Pseudocolor image of the corresponding epipolarization image.Figure 4. (a) Cox-1 immunogold-silver staining of an alveolar macrophage. Note the granular pattern of the immunostaining. Scale bar: 20 µm. (b) Pseudocolor conversion of the accompanying epipolarization image shows low gray values. The granular staining pattern is clearly visible.

Cox-2. Gray-scale measurements of the Cox-2 immunostaining are given in Table 2. Strong staining was noticed in the vascular smooth muscle cells mainly of partially muscular vessels (Table 2, Figures 1 and 6), which may be pre- or postcapillary (35). In addition, vascular smooth muscle cells of the large extrapulmonary hilum veins showed extensive staining. In contrast, smooth muscle cells of arteries at the hilum displayed only some focal staining with weak intensity.

View larger version (52K): [in this window] [in a new window]   Figure 5.   (a) Cox-2 immunogold-silver staining of bronchial epithelial of first- or second-generation bronchi. Note the macrophage-like cell located in the connective tissue below the bronchial wall. Scale bar: 20 µm. (b) Pseudocolor image of the accompanying epipolarization image.Figure 6. (a) Cox-2 immunogold-silver staining of vascular smooth muscle cells of partially muscular vessels. The partially muscular vessel shown here is likely a vein. Note the smooth muscle cells bulging into the vascular lumen. Scale bar: 20 µm. (b) Pseudocolor conversion of the accompanying epipolarization image shows high staining intensities.Figure 7. (a) Cox-2 immunogold-silver staining of macrophage-like cells located in the connective tissue surrounding bronchi or vessels. Note the intensive partly granular staining. Scale bar: 20 µm. (b) The pseudocolor image of the accompanying epipolarization image reveals intensive staining of macrophage-like cells.

Reverse Transcription-Polymerase Chain Reaction

The PCR products of Cox-1 and Cox-2 mRNA were detected in all parts of lung tissue following amplification for 27-29 cycles. The detected mRNAs were found evenly expressed in all probed lung lobes (Figure 8).

View larger version (47K): [in this window] [in a new window]   Figure 8.   Cyclooxygenase mRNA expression. Total RNA was extracted from total left lung (ll), middle lobe (m), lower lobe (l), and upper lobe (u) of the right lung. mRNA expression of Cox-1, Cox-2, and GAPDH was detected by RT-PCR as described in MATERIALS AND METHODS.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Employing a reverse transcription-polymerase chain reaction (RT-PCR) and immunohistochemistry, both Cox-1 and Cox-2 were readily detectable in the lungs of normal rats. Cellular localization studies ascribed Cox-1 predominantly to bronchial epithelial cells, smooth muscle cells of the large hilum veins, and (with lower expression) alveolar macrophages and endothelial cells of the large pulmonary arteries at the hilum. In contrast, the pattern of Cox-2 distribution was characterized by its localization in macrophage- and mast cell-like cells in perivascular, peribronchial, and subpleural tissues, smooth muscle cells of partially muscular small vessels and veins; and to some lesser extent in bronchial epithelial cells. Differential regulatory functions must be expected to be linked to this differential cellular distribution of Cox-1 and Cox-2.

The RT-PCR analysis performed in homogenate fractions originating from upper, middle, and lower lobes of rat lungs confirmed the observation in human lung tissue (27), that both Cox-1 and Cox-2 are easily detected in this organ. The former study was performed in trauma victims, in whom rapid induction of inflammatory events due to the catastrophic event might not be fully excluded. Currently, however, lung tissue was obtained under strictly physiological conditions, thus excluding any doubt that both Cox-1 and Cox-2 are expressed in the lung under physiological conditions. To the best knowledge of the authors, this study is the first to characterize the cellular localization sites of Cox-1 and Cox-2 in the intact lung.

Quantification of immunohistochemical staining intensity by a combination of epipolarization microscopy and image analysis revealed marked differences between different cell types. A precise classification of the stained structures was achieved by superimposing the epipolarization image on the conventional bright-field image. Pseudocolor images were used for direct visualization of gray values. Gray-scale values of the Cox-1 and Cox-2 immunostaining cannot, of course, be compared directly. It is likely that the antigen-binding capacity of the polyclonal primary antibodies used might be different. However, quantitative comparison can be performed on data derived from experiments using the same antibody. Image analysis of epipolarization images offers an objective, reproducible, and observer-independent evaluation compared with a subjective judgment of staining intensity.

The pattern of Cox-1 distribution was predominated by its localization in bronchial epithelial cells. This was to be anticipated from studies in cultured tracheobronchial epithelial cells in vitro (28, 38). Cox-1-dependent prostanoid generation has been implicated in the regulation of bronchial tone (39). In addition, marked Cox-1 immunostaining was observed in smooth muscle cells of large veins at the hilum. These cells are developmentally closely related to left atrial muscle cells (33, 34), and it may be speculated that the generation of prostanoids, if vasodilatory in type, might be implicated in the establishment of patency of the venous lung outflow. Moreover, some positive Cox-1 staining of alveolar macrophages was found throughout. This observation corresponds to studies in alveolar macrophages obtained by bronchoalveolar lavage, in which Cox-1-related limited baseline synthesis of PGE₂ was noted (5, 22, 29). Cox-1 immunostaining was found to be positive in endothelial cells of large pulmonary arteries in the intact rat lung. This observation is supported by cell culture studies that found endothelial cells to possess Cox-1 mRNA and protein, but no Cox-2 mRNA, or Cox-2 protein, under basal conditions (40, 41).

The pattern of Cox-2 distribution in lungs fixed under physiological conditions was markedly different from that of Cox-1. The strongest staining was noted in macrophage-like cells in the bronchial wall or adjacent to the peribronchial tissue and in the connective tissue around pulmonary vessels. The identity of these cells is not absolutely clear, but we assume them to be cells of the mononuclear phagocyte system. A further substantial fraction could be mast cells because of their size and granular appearance. In addition, counterstaining with alcian blue showed a double-staining pattern with the Cox-2 staining in this fraction (42). Mast cells are known to synthesize membrane-derived lipid mediators via arachidonic acid metabolism along the cyclooxygenase or lipoxygenase pathways (43). In particular, rat connective tissue mast cells have been reported to generate predominantly prostaglandins (44). Moreover, the presence of a cyclooxygenase in the lipid bodies of human lung mast cells was proved by ultrastructural immunogold and autoradiographic localization (45, 46). These cells were not stained by Cox-1 immunostaining. Although the macrophage-like cells were measured to show the highest staining intensity for Cox-2, they probably contribute to a minor extent to the overall expression of Cox-2 in the rat lung because of their low total cell number. The different staining intensities detected for different cell types may indicate local differences in the production of arachidonic acid metabolites, dependent on the cellular equipment with downstream enzymes. These microenvironmental concentrations of such mediators may be of more importance for regulatory mechanisms in distinct compartments within the lung than the overall production rate.

In addition, vascular smooth muscle cells could be identified to display marked Cox-2 expression under baseline conditions. The muscle cells were particularly found in the partially muscular vessels and in the large hilum veins. Interestingly, the smooth muscle cells of large pulmonary arteries and bronchial airways were not that intensely stained. The differentiation between arteries and veins of the lung vasculature is less clear than in the systemic circulation, where pressure and resistance are high and the morphologic differences between thick-walled arteries and thin-walled veins are obvious. Smaller pulmonary vessels are particularly difficult to classify (35, 36). For this reason, we cannot decide whether the partially muscular vessels are of precapillary or postcapillary origin, or both. As intraacinar, small lung vessels are intimately involved in the regulation of lung vascular tone and perfusion distribution, most evident during hypoxic vasoconstriction (47, 48), the Cox-2 expression in small lung vessels may be related to physiological pulmonary vasomotor activity.

Clearly lower, but still obvious, Cox-2 immunostaining was noted in bronchial epithelial cells, particularly in the large bronchi at the hilum. This observation corroborates cell culture studies, which demonstrated low levels of Cox-2 in untreated human airway epithelial cells (28). In vitro studies also showed that cultured human bronchial epithelial cells express Cox-2 constitutively (49, 50). Whereas these studies found little or no Cox-1 in their cell lines in vitro, another group, which examined human bronchial biopsies, found that both cyclooxygenase isoenzymes are expressed in normal human respiratory epithelium (51).

Almost no staining of either Cox-1 or Cox-2 could be detected in the alveolar walls. A few single cells in the alveolar septa were noticed, which showed specific Cox-2 staining. Those may be interstitial fibroblasts, macrophages, or intracapillary leukocytes (52, 53), which all have been reported to express Cox-2 in vitro (1, 3, 4, 29, 54). Alveolar macrophages were scarcely immunohistochemically stained for Cox-2, except for single cells of this macrophage type (about two or three per cross-section of a complete rat lobe), located mainly beneath the pleura. As already discussed, unstimulated alveolar macrophages obtained by lavage have been shown to contain Cox-1 and synthesize PGE₂, but do not express Cox-2 (5, 22, 29). In accordance with in vitro studies, we did not detect Cox-2 expression in pulmonary vascular endothelial cells.

In conclusion, the current study provides strong evidence that in addition to Cox-1, the isoenzyme Cox-2 is constitutively expressed in lung tissue, but with a different pattern of cellular localization. This view is supported by studies in other organ systems, which also raised evidence that Cox-2 may be the constitutive isoform of the cyclooxygenase in certain cell types. Immunocytochemical localization of Cox-1 and Cox-2 in the rat stomach ascribed these isoforms to different types of mucous cells (55). In the brain and spinal cord of rats, Cox-2 has been identified as the major isoform of the cyclooxygenase enzyme, expressed under basal conditions (9, 32, 56).

Physiological and pathophysiological aspects of a constitutive expression of Cox-2 in lung tissue deserve further investigation. Prostanoid generation via this enzyme may be involved not only in bronchomotor and vasomotor activities, but also in the regulation of lung barrier properties, pro- and antiinflammatory pathways, and interstitial matrix deposition, as suggested by the finding of diminished Cox-2 expression and PGE₂ generation in lung fibroblasts cultured from patients with idiopathic pulmonary fibrosis (17).