Effect of dexamethasone on carrageenin-induced inflammation in the lung

To study the anti-inflammatory mechanisms of glucocorticoids, we have compared the effects of intratracheal carrageenin (2.5 mg) on control rats and those in which inflammation was subdued by prior dexamethasone treatment (10 mg/l in drinking water). Inflammation was maximal 48 h post-carrageenin. After dexamethasone, carrageenin caused tittle inflammation or oedema (wet lung (mg), n = 6, mean ± S.E.M.; control, 995 ± 51; carrageenin + dexamethasone, 1144 ± 83; compared with carrageenin alone, 1881 ± 198), but rats had more lung lavage neutrophils than those given carrageenin alone (PMN × 106 /lung, mean ± S.E.M.; control, 0.055 ± 0.003; carrageenin + dexamethasone, 8.54 ± 1.52; compared with carrageenin alone, 6.30 ± 1.71). Proteolysis and partial inactivation of the anti-inflammatory mediator, lipocortin 1 (Lcl), in carrageenin-instilled rats was offset in those also given dexamethasone, by increased Lc1 levels (intact Lc1 ng/ml lavage fluid, n = 4, mean ± S.E.M.; control 24 ± 6; carrageenin 15 ± 4; carrageenin + dexamethasone, 40 ± 15). Maintenance of sufficient intact (fully active) extracellular Lc1 may contribute to the actions of glucocorticoids.


Introduction
for example by altering the expression of their receptors. 5 Inflammation is a complex response to tissue Lipocortin 1 (Lcl), a member of the annexin damage, often due to microbial invasion, in-family, 6 is glucorticoid-inducible in vivo -9 and voMng the release and interactions of many has anti-inflammatory activity, 1 expression of mediators including cYlokines proteinases, which is thought to require binding of Lcl to eicosanoids and oxidants from resident and specific sites on cell surfaces. 11 It is normally newly infiltrating cells. It is a particularly im-present in the lung, both intracellularly, in a portant defensive mechanism in the airways variety of different cell types 2 and extracellularly, 13 14 and lungs, which are constantly exposed to in the epithelial lining fluid.' Glucocorticoids inhaled biological, chemical and physical increase the concentrations of Lcl in a variety of insults, lung cells, including alveolar epithelial cells 15 and 16 The endogenous mechanisms regulating alveolar macrophages, and may promote its inflammation in the lung and in other tissues release onto the cell surface and/or into epitheare incompletely understood. However, studies lial lining fluid 13'17' where it would be available on rodents have shown that prior surgical or for receptor binding. Thus, Lcl could play a chemical adrenalectomy exacerbates the inflam-pivotal role in the anti-inflammatory actions of matory response to any given stimulus, 2'3 an glucocorticoids in the lung, but currently its coneffect which can be ameliorated by glucotfibutionis unclear. corticoid replacement therapy, indicating that In this paper, we describe a model of carraphysiological levels of endogenous circulating geenin-induced pneumonitis which we have charglucocorticoid normally regulate inflammatory acterized in control animals and in those in processes. Glucocorticoids operate via specific which the inflammatory response was subdued receptors at the transcriptional level to alter the by prior and concurrent treatment with exogenexpression and activities of specific gene proous glucocorticoid. We have also quantified ducts (reviewed in Reference 4). They may also extracellular Lcl to assess its putative contribuexert indirect effects on inflammatory mediators, tion to the actions of dexamethasone.

Materials and Methods
Experimental model of inflammation: Animals. Groups of mature CFY rats, bred inhouse and weighing between 180 and 220g at the beginning of the experiment, were housed in pairs and given unlimited access to food and (d) Dexamethasone (DEX). These animals were used to determine the effect of dexamethasone on the normal lung over the entire time course of the experiment. They received the same dose of oral dexamethasone for the same length of time (96 h) as the CARR + DEX treated rats, but were not exposed to carrageenin. water or water containing dexamethasone (see (e) Baseline dexamethasone (BASE DEX). These animals were needed to establish the effect of below). At the end of the appropriate experimental period, all animals were anaesthetized dexamethasone pretreatment on the lungs of rats with halothane then sacrificed by sodium thio-in the CARR+ DEX experimental group prior to pentone overdose, carrageenin administration. Thus, they were dosed with dexamethasone for 48 h then sacri-Carrageenin treatment. Animals undergoing car-riced as described above. rageenin instillation were lightly anaesthetized with halothane and 2.5mg carrageenin/0.5ml Pilot studies. As well as the carrageenin dosing sterile 0.15M NaCl (lambda-carrageenin (Type regimen and the experimental time course IV), Sigma Chemical Co., Poole, UK)was instilled described above, we performed pilot studies to into the lung via the trachea using a previously assess the degree of pneumonitis in animals described method, 9 except that the suspension killed (a) 48 h after a high intratracheal (i.t.) dose was instilled through a flexible 3FG catheter of carrageenin (5mg/0.5ml NaCl i.t.); (b)4h (Portex Ltd.) which was introduced through the after the standard dose of carrageenin (2.5  Laboratories, Warwick, UK) was administered in Assessment of the inflammatory response to cardrinking water at a dose of 10 mg/1. Average conrageenin with and without dexamethasone: sumption of fluid was approximately 10% of body weight per day. This represented a dose of Light microscopy. At least two animals were approximately l mg dexamethasone/kg body assigned to each of the experimental groups weight/day, described above. Following treatment and sacrifice according to the protocols described above, Experimental design. The various experimental the lungs were excised and fully inflated under groups are described in more detail below, constant pressure (25 cm) with formaldehyde (a) Controls (CON). Control animals were (4% v/v in phosphate buffered saline) which was required to illustrate and define the physiological instilled via the trachea. Following fixation, the norm and were, therefore, entirely untreated, lungs were prepared using standard histological They were not sham-instilled because the main procedures and stained with haematoxylin and purpose of the experiment was to establish the eosin. They were examined blind. effect of dexamethasone on the inflamed lung, not to establish the contribution of saline to the Bronchoalveolar lavage and venous blood. Four inflammatory response provoked by carrageenin rats were assigned to each group. Following instillation, death, the pulmonary circulation was perfused in (b) Carrageenin (CARR). These animals were situ with 0.15 M NaC1 via the right ventricle. used to determine the effect of carrageenin on When the lungs were free of blood, they were the lung. Animals were instilled with carrageenin removed from the thoracic cavity and lavaged via as described above and sacrificed 48 h later, the trachea with 6 x 8 ml portions of 0.15 M (c) Carrageenin and dexamethasone NaCl. 19 The recovered lavage fluid was pooled (CARR+ DEX). This group of rats was used to and the volume recorded, a sample removed for examine the effect of dexamethasone treatment total cell count and differential cell profile by on the carrageenin-inflamed lung. The animals Wright-Giemsa staining, then the remaining fluid received dexamethasone in drinking water for was processed as described previously, i 48h prior to carrageenin exposure. They were Venous blood samples (0.2ml) were taken then exposed to intratracheal carrageenin and from each animal for differential white blood cell continued to receive dexamethasone for a further analysis before the beginning of the experiment, 48 h. At this time, they were sacrificed as describat the time of carrageenin instillation and at the ed above for comparison with rats exposed to time of sacrifice. Smears were made of each carrageenin alone, blood sample, allowed to air dry and stained with Eect of dexamethasone in pneumonitis Wright-Geimsa stain. The proportion of poly-Analysis of extracellular lipocortin 1 in bronchomorphonuclear neutrophils (PMN) and mono-alveolar lavage fluid. The acellular bronchonuclear cells was determined by light microscopy, alveolar lavage fluid (BALF) samples analysed in Approximately 150 cells were counted per smear this part of the experiment were those collected except in dexamethasone-treated animals in as described in a previous section. The cellular which circulating leukocytes were sparse.
Lcl content was not determined in this study. Extracellular Lcl levels were measured using a Electron microscoly. The experiment was perpreviously described 2 competitive enzyme-linked formed on two rats per treatment group as immunosorbent assay (ELISA) modified as described for the light microscopy study, except follows. Gelatin (2% w/v) was used as a blocking that the lungs were inflated with 2% paraagent, wells were coated with 100 l.tl Lcl (1 lg/ formaldehyde and 2.5% glutaraldehyde in 0.1 M ml) and the polyclonal anti-Lcl antibody was sodium cacodylate buffer, pH 7.4. After 24-72 h, used at a dilution of 1:30000 in phosphate bufthis fixative was replaced with 50 mM sodium fered saline containing 0.5% Tween. cacodylate buffer, pH 7.4, containing 7.5% Samples of BALF were concentrated 7.5-fold sucrose. The lungs were encoded and stored at by centrifugation in Centricon units with a 4C until prepared for electron microscopy using nominal molecular weight cut-off of 2 000 kDa. standard techniques. They were examined blind.
Concentrates were separated by SDS-polyacrylamide gel electrophoresis, followed by Assessment of oedema by wet and dry lung Western blotting and staining for Lcl as describweight measurements. Animals (six or seven per ed previously. 4 The relative proportions of intact group) were treated as described above. Follow-and proteolysed Lcl were determined bYe4 scaning sacrifice, the lungs were excised and lavaged ning densitometry as described previously. as described above. 9 The major airways were removed, then the parenchymal tissue was Statistics: Where appropriate, statistical analysis blotted to remove excess moisture and weighed was performed using a Wilcoxon rank sum test (wet weight). The weighed tissue was then for unpaired data. The level of significance was minced and homogenized in a Waring blender p < 0.05.
with an equal mass of distilled water to give a 50% (w/w) homogenate which was divided into Rut two equal portions. One half was placed in a pre-weighed tube and the dry weight was deter-Macroscopic appearance: Lungs from rats given mined by heating to 100C and re-weighing daily carrageenin alone (CARR) appeared mottled with until no further weight loss occurred. The wet hard, lumpy areas. The lungs of all other animals weight of I ml of autologous blood was mea-were macroscopically normal. The pleura sured, then its dry weight was determined by the appeared normal in all animals. same method. The remaining half of each lung homogenate Histopathology: In pilot studies, no changes were was centrifuged at 30000 x g at 4C for I h to seen at the light microscopic level 4 h after carraremove debris. The resulting supernatant was geenin instillation (Fig. lb). The peak effect of assayed in duplicate for haemoglobin by the carrageenin (CARR) was observed 48h after cyanmethaemoglobin method. 21 The haemoglobinstillation and consisted of large areas of interin content of a sample of autologous whole stitial and intra-alveolar pneumonitis, with alveoblood from each rat was also quantified by the lar wall thickening and a prominent cellular cyanmethaemoglobin method. The haematocrit infiltration (Fig. lc). By 96h after instillation, was measured by centrifuging duplicate capillary there were fewer cells in airspaces and intertubes of venous blood in a Hawksley microstitium, and the alveolar walls were less swollen; centrifuge for 2 min and reading the percentage the inflammation appeared to be resolving of cells on a calibrated scale. (Fig. ld). In pilot studies (data not shown) it was The contribution of blood to the wet and dry found that a dose of 5 mg carrageenin provoked weights of each lung could then be calculated as the same inflammatory response at 48 h postdescribed previously, 22 except that haemoglobin instillation as 2.5 mg which we, therefore, believe was used as a marker of blood. It was then pos-produces a maximal effect. Dexamethasone sible to determine the blood-free lung water and alone (DEX: BASE DEX) had no effect on pulwet and dry lung weights. The haemoglobin monary histopathology at any of the time points content of bronchoalveolar lavage, both fluid and examined (Fig. lf). In animals given both glucocells, was also assessed but was below the limits corticoid and carrageenin (CARR+ DEX), there of detection in all samples, was little alveolar wall thickening; however, the   Data expressed as means _ _ _ standard error of mean; n 4. Where no error bar is visible, the standard error is within the diameter of the symbol.
(CARR) was associated with significant increases in the total cell number; the majority of cells in BALF were AM, although there was also a significan increase in the number of PMN (Fig. 2). Dexamethasone treatment for 48 h (BASE DEX) had no effect on cell number or type (Fig. 2). However, after 4 days of dexamethasone (DEX), the total number of cells in BALF was reduced compared with the untreated (CON) animals (Fig. 2). Interestingly, whilst the total number of cells in BALF from animals given carrageenin plus dexamethasone (CARR+DEX) was intermediate between those given either stimulus alone, there were more PMN in their BAI than in that of any other treatment group (Fig. 2), indicating that the dexamethasone had selectively suppressed the AM infiltration. The mean number of AM/g wet tissue following various treatments was as follows: CON, 4.02 x 106/g; BASE DEX, 2.58 x 106/g; DEX, 1.58 x 106/g; CARR,, 11.35 x 106/g; CARR + DEX, 2.87 x 106/g.
Orculating white blood cells: The blood smears indicated that carrageenin had no effect on the differential white blood cell ratios at any time point (CON 9___ 2% polymorphs, CARR 16 _+ 3% polymorphs, mean + S.E.M., n 6 per group, p > 0.05). In rats treated with dexamethasone, or carrageenin and dexamethasone, most of the circulating leucocytes were polymorphs (DEX 74 +_ 2%; CARR+ DEX 83 +_ 3%). This change in differential count was due to a substantial reduction in the number of circulating mononuclear cells; the number of circulating PMN fell by a lesser amount in response to exogenous glucocorticoid.
PMN were also observed. In addition, the alveolar spaces contained many round bodies, usually devoid of organelles and thus, appearing comparatively electron-lucent (Fig. 3a). Similar structures were associated with some Type II alveolar epithelial cells and high magnification revealed continuity between the organelle-rich cell cytoplasm and the electron-lucent bulge (Fig. 3b), indicating that the bodies probably originated from the Type II cells. Type II cells were present in increased numbers, often several in close proximity (Fig. 3a). Damage to alveolar Type I cells was common (Fig. 3b). Occasionally, lengths of basement membrane denuded of epithelium were seen; these were associated with areas of interstitial oedema (Figs 3c and 3d). There was also evidence of endothelial damage, with shrinking and increased cell density (Fig. 3d). An increase in interstitial collagen was observed.
By comparison, the changes were minor in the lungs of rats given carrageenin plus dexamethasone (CARR+DEX). There was a slight increase in the number of AM compared with the control groups, and occasional swelling of the Type I epithelial cells was seen ( Fig. 1; EM not shown). No difference was seen between groups given dexamethasone alone (DEX) and untreated controls (CON), all of which appeared normal ( Fig. 1; EM not shown). Assessment of oedema by wet and dry lung weights: In untreated rats, approximately 90% of the total lung weight (excluding blood) was water. Carrageenin treatment alone (CARR) resulted in significant increases in lung water and dry lung weight compared with the control and those given glucocorticoid alone (DEX and BASE DEX; Table 1). The total amount of blood in carrageenin-treated lungs (CARR) was unchanged compared with the control (CON; Table 1), but because of the increased lung weight, blood represented only 6% of the total wet weight rather than 12% as in the controls (Table 1). 3. Electron micrographs of lung sections from rats treated with carrageenin alone (CARR) and sacrificed 48 h after instillation. In each figure, the bar represents 10 microns unless otherwise stated. (a) Many foamy macrophages (m) and an electron lucent round body (rb) in the alveolar space; also in the section are two Type II epithelial cells (11) in close proximity (magnification x 2 000). (b) A Type II epithelial cell from which protrudes a bulbous extension; high magnification (x 20000) confirms cytoplasmic continuity with the Type II cell. The section also shows a damaged, vacuolated Type (I) cell (magnification of main picture x 5 000). (c) Interstitial oedema (o) in close proximity to a region of denuded epithelial basement membrane (arrows), part of which (A) is shown at higher magnification in Fig. 3(d) (magnification x 2 700). (d) Fragments of epithelium associated with a length of denuded basement membrane (arrow). Note also abnormal capillary endothelium (en) with increased density and cell shrinkage (magnification x 28000; the bar represents micron). Data represent mean -tstandard error of the mean; < CON; > DEX; > all other groups; < DEX; p < 0.05 in a Wilcoxon Rank Sum Test.

Erect of dexamethasone in pneumonitis
duced the lung water within 48h (BASE DEX; Table 1) and the dry lung weight by 96 h (DEX).
There was less blood in glucocorticoid-treated lungs (DEX) compared with the controls (CON; Table 1), but as a percentage of the total wet lung weight, it was unchanged (11-12%). The lungs of animals given both glucocorticoid and carrageenin (CARR + DEX) were intermediate between those given either stimulus alone, in terms of dry lung weight and lung water, and did not differ from controls (CON; Table 1). Likewise, the amount of blood in the lungs of rats given both carrageenin and dexamethasone (CARR+ DEX) was not significantly different to any other group, but as a percentage of total lung weight, it was intermediate between carrageenin-treated animals (CARR) and those in the other treatment groups.
Extracellular lipocortin 1 in bronchoalveolar lavage fluid.. Lcl was detected by ELISA in BALF from all the animals investigated. Approximately 90% of the Lcl in BAI from control (CON) rats was intact, as determined by Western blotting. Treatment with dexamethasone for 48h (BASE DEX) or 96h (DEX) was associated with a doubling of total Lcl in BALF, which again, was over 90% intact. Carrageenin-treated rats (CARR) also had twice as much Lcl in their BALF as controls, but over half of it was in the proteolysed form (Fig. 4). When both glucocorticoid and irritant were administered to the same animals (CARR+DEX), the total Lcl concentration was approximately 10 times the control levels and was significantly greater than all the other treatments. However, as only 15% of this Lcl was in its intact form, the concentration of intact Lcl did not differ significantly from any other treatment group (Fig. 4). None of these changes could be explained by changes in BAL volume.

Discussion
This study describes some of the morphological, cellular and biochemical changes occurring in the rodent lung following the intratracheal instillation of carrageenin. We observed an inflammatory response which was maximal at 48 h post-instillation and which was characterized by interstitial oedema and an influx of macrophages and PMN. Prior administration of an exogenous glucocorticoid greatly reduced the inflammation, confirming atotal Lcl/ml > CON; bintact Lcl/ml > CON; in each case, p < 0.05 in a Wilcoxon rank sum test for unpaired data. Data expressed as means -t-_ standard error of mean; n 4. Where no error bar is shown, the error is within the diameter of the symbol. the hypothesis that carrageenin is essentially inert that a major effect of the glucocorticoids is to if the inflammatory response is subdued. Further-down-regulate the actions of these cells. more, dexamethasone was associated with an One of the most intriguing features of the increase in extracellular levels of Lcl, an anti-current model is that whilst the epithelial damage inflammatory protein which may contribute to the and interstitial oedema caused by carrageenin actions of the glucocorticoid, was clearly prevented by dexamethasone admin-Historically, carrageenin has been used to istration, this effect was accompanied by sigprovoke an inflammatory response in a wide nificant increases in both the number and range of tissues, including the lungs of rabbits, percentage of lung PMN. This was observed cats and rodents. [24][25][26] The effects of carrageenin despite the reduced number of circulating leukoon lung morphology in this study are similar to cytes and suggests that either the cells were those reported previously, 25'27-3 namely, an early actively recruited into the lung (by changes in infiltration of PMN and macrophages, many of chemotaxins or adhesion molecules) or were not which phagocytosed carrageenin particles, plus cleared, perhaps because the PMN did not epithelial cell damage and basement membrane become apoptotic or because the reduced denudation, followed by Type II cell hyperplasia, number of macrophages prevented removal of We have not found specific reference to the carapoptotic PMN. The data further suggest that rageenin-induced emergence of electron-lucent once inflammation is established, the PMN is less 'blebs' from Type II cells described in the deleterious than initially. There is some evidence current investigation. However, such a phenomto support this view, suggesting that PMN may enon has been described in a clinical context in play a subtle role in modulation of the inflammaa case of Type II cell adenocarcinoma and thus, tory and immune responses, and may even is not necessarily a specific response to carrageepromote mechanisms of repair (reviewed in nin, 3 but may be a nonspecific effect associated Reference 1). In addition, there is evidence that with epithelial cell damage and hyperplasia. In PMN are activated by cytokines and growth addition to confirming the morphological obser-factors released from monocytes and macrovations of previous authors, we have demon-phages such as GM-CSF and TNF. Hence, glucostrated that the response to carrageenin can be corticoids may reduce PMN activity in the lung suppressed by pretreatment with dexamethasone, directly (for .example, as indicated in Reference In this study we found no change in the 34) or indirectly by reducing both the number of profile of circulating white cells after intratracheal macrophages (Fig. 2) and their ability to release carrageenin. This is in contrast to intraperitoneal mediators which stimulate PMN. 35 administration in which there is a marked We hypothesize that the protective effect of increase in the number of circulating PMN prob-the glucocorticoids may be due in part to the ably due to increased release from the bone increase in extracellular Lcl observed in this marrow, perhaps combined with the lack of study. Lcl probably exerts its anti-inflammatory alternative routes of cell clearance from the peri-activity by binding to cell surface receptors. 1 toneal cavity 32 and a decline in monocytes.
The increase in extracellular Lcl observed in Although an increase in oedema has been response to dexamethasone ma have been observed within 2-3 h after carrageenin in the caused by de novo Lcl synthesis, translocation rat paw, 1 in the lung we saw no morphological to the cell surface 36 or a combination of both, evidence of oedema at 4h. However, by 48 h and may have resulted in enhanced Lcl activity interstitial oedema was visible by EM (Fig. 3) and within the lung. Interestingly, carrageenin adminlung water was markedly increased (Table 1). istration was also associated with an increase in Dexamethasone administration prevented the Lcl in the epithelial lining fluid. We cannot interstitial oedema and the accompanying epithel-exclude the possibility that carrageenin induced ial damage and was also associated with a decline Lcl indirectly, by stimulating a rise in circulating in macrophage numbers in the lung. This is in corticosterone. However, inflammatory stimuli keeping with the findings of Heluy-Neto et al. 33 such as paraffin oil 8 and endotoxin 3v have been who studied the effects of glucocorticoid deple-shown to cause increases in Lcl in glucocortition (by adrenalectomy) on carrageenin-induced coid-depleted (adrenalectomised) rats, so it is inflammation of the pleura, and reported that possible that the increase was a direct effect of both the number of monocytes and the inflam-the inflammogen on resident or inflammatory matory exudate in the pleural cavity increased cells. In carrageenin-treated animals, the premore following an adrenalectomy than a sham dominant form of Lcl was a 34kDa species operation. Thus, together, these studies suggest which lacks the N-terminal believed to confer that the activity of activated monocytes may cause functional specificity 6 and which was probably. oedema and exacerbate the inflammation, and clipped by the action of neutrophil elastase. Reduction of intact Lcl by proteolysis may feedback to stimulate further Lcl synthesis and/or release, and this may explain the apparent synergy between carrageenin and dexamethasone with regard to extracellular Lcl levels; the two stimuli may have increased extracellular Lcl by different mechanisms. Although functional Lcl was markedly reduced in the epithelial lining fluid of rats given carrageenin only, prior and concurrent administration of dexamethasone maintained intact (functional) extracellular Lcl at the control level. The very high level of total Lcl in animals given glucocorticoid and inflammogen suggests that release of this protein in its functional form earlier in the response may have contributed to the protective effect of dexamethasone. The anti-inflammatory mechanisms of tcl are complex and incompletely understood, but may include a dose-dependent reduction in interactions between IgG and Fc, receptors on PMN and monocytes, an effect which may reduce the oxidant-producing capacity and other pro-inflammatory functions of these cells.
In summary, in this study we have used a model of carrageenin-induced pulmonary inflammation to investigate the possible anti-inflammatory mechanisms of action of dexamethasone. We conclude that an increase in extracellular Lcl may be one mechanism by which this synthetic glucocorticoid controls pulmonary inflammation.