IL-1 and its role in rat carrageenan pleurisy

The carrageenan pleurisy model, which is characterized by cellular influx and oedema, has been used to examine the effects of anti-inflammatory compounds such as naproxen. Interleukin-1α and β (IL-1) are known to be pro-inflammatory mediators, and their roles in this model are unknown. Intrapleural injection of 1% viscarin carrageenan or saline was administered to male Lewis rats. Four to 24 h later, cell counts, fluid volumes and IL-1β levels (measured by ELISA) were determined in the pleural cavity. Serum corticosterone levels were measured only at 4 h. Significant increases in IL-1β levels precede cell influx suggesting IL-1β plays a role in the maintenance of cell accumulation in the pleural cavity. None of the drugs tested, including the IL-1 receptor antagonist, maintained pleural cell influx and IL-1β levels at control levels. When human IL-1α or β or rat IL-1β were injected individually into the pleural cavity, none of these cytokines were pro-inflammatory, as measured by increased cell influx and fluid extravasation. These results suggest that although IL-1β levels increase in the pleural cavity in response to carrageenan, IL-1 per se is not the initiator of the pro-inflammatory events of cell influx and oedema in this model.


Introduction
The components of inflammation include leukocyte infiltration and fluid accumulation which accompany the cardinal signs of inflammation such as heat, swelling, redness and pain. Many mediators, including lipids, proteinases, biogenic amines, and peptides have been implicated in this process. In order to investigate the roles of such mediators, many animal models have been developed. One such model is carrageenan induced pleurisy in the rat, in which the phlogistic agent is usually viscarin carrageenan. This model has been extensively studied, has the advantage of being versatile, and has been used to evaluate standard anti-inflam- [1][2][3][4][5] matory agents.
Using the carrageenan pleurisy model the potential involvement of interleukin-l(IL-1), a cytokine which is a known pro-inflammagen, 6 has been investigated. Among its many activities IL-1 can stimulate cells to induce synthesis of other cytokines, v increase activity of enzymes such as phospholipase A2, 8'9 and is responsible for the accumulation of neutrophils at the sites of inflammation. 1 The exact role of IL-1 in the inflammatory process induced by carrageenan is still unclear. The purpose of this study was to investigate the role of IL-1 and the IL-1 receptor antagonist in this model. (C) 1993 Rapid Communications of Oxford ktd Materials and Methods Reagents: Human and rat IL-1 0 and /g, human 1L-1 receptor antagonist (rhIL-lra), and rabbit polyclonal anti-mouse IL-1 antibodies were prepared at The Du Pont Merck Pharmaceutical Co. All other reagents were purchased from Sigma Chemical Company, St. Louis, MO, or through VWR, and were reagent grade.
Pleurisy: Male rats (Lewis rats, LEW/Crl BR Charles River Laboratories, Wilmington, MA) weighing 200-250 g were divided into groups of ten. The procedures have been described previously. [1][2][3][4] Pleurisy tests were carried out using either rats fasted for 16 h prior to the study or rats fed ad libitum. Animals were treated orally with appropriate drugs or subcutaneously with the rhIL-lra 1 h prior to the introduction of the inflammagen.
Animals were anaesthetized with ether and placed left side down and injected intrapleurally with 0.2 ml of 1% viscarin carrageenan (Sigma Chemical Co.) or rat IL-1//, human IL-10 or fl at doses described. Four to 24 h after injection animals were euthanized using COg. An intercostal incision was made in the pleural cavity, the cavity lavaged with 6 ml of 0.9% NaC1, exudate volumes recorded, and fluids centrifuged at 658 x g" (Sorvall RC5C, Du Pont, Newtown, CT) for 10 min at 5C. Samples overtly bloody were discarded. A 1 ml sample of supernatant was saved and stored at 4C for determination of IL-lfl levels by ELISA. The remaining supernatant was decanted and cells resuspended in 2 ml of 0.9% NaC1 containing 2% heparin. Aliquots were used to determine total cell (Coulter Counter, Coulter Corp, Hialeah, FL) and differential cell counts using Wright's stain. Changes in fluid volumes were calculated as the difference between the initial 6 ml used for the lavage and the final volume recovered from the pleural cavity.
ELISA: A sandwich ELISA, 11 using a polyclonal rabbit antiserum obtained by immunization with murine recombinant IL-lfi, was used to measure the IL-lfl levels from the rat pleural cavities. The microtitre wells were coated with a protein A purified rabbit-anti-mlL-lfl lgG (20#g/ml in phosphate buffered saline) at room temperature for 1 h. The wells were washed with Dulbecco's phosphate buffered saline. Volumes (100 #1) of sample or known recombinant IL-lfl dilutions were added to the wells, followed by a 2 h incubation at room temperature. The washing step was repeated, and 100#1 of the biotinylated affinity-purified rabbit-anti-mIL-lfi antibody was added and allowed to incubate at room temperature for 1 h. The optimal dilution of this secondary antibody was determined by titre prior to the assay. The assay was developed using avidin coupled to alkaline phosphatase followed by p-nitrophenyl phosphate substrate. Incubation was continued at room temperature for 90 min, and the plates were read at 405 nm (UV Kinetics microplate reader, Molecular Devices Corp., Menlo Park, CA). The concentration of IL-lfl in the samples was calculated from the IL-lfl standard curve. This assay was specific for IL-lfl; IL-10 and the rhIL-lra did not cross-react. The lower level of detection for IL-lfl was 0.1 ng/ml. Recombinant human and rat IL-lo and fl and recombinant human IL-1 receptor antagonist: All IL-1 proteins 12 and the rhIL-lra were produced in Escherichia coli and purified by standard column chromatography methods. Samples were tested for endotoxin using the LAL assay (Whittaker M.A. Bioproducts) with all preparations shown to contain less than 10 EU/mg of protein. The characteristics of the rhlL-lra were identical to those published previously. 3 Corticosterone levels: Blood was collected into vacutainers containing sodium heparin from anaesthetized animals via heart puncture. Plasma samples were obtained by centrifugation for 10 min at 658 x g'. Samples were assayed using an RIA for corticosterone, the major glucocorticoid in rat plasma (MetPath Inc., Rockville, MD). The normal range for this assay is 10-60 ng/ml. 4 Statistical analysis" Data are expressed as mean-kstandard error of the mean, determined by analysis of variance followed by Duncan's multiple range test (SAS Institute, Cary, NC). Statistical significance was considered to be p < 0.05.

Results
Time course: Figure 1A shows the response to carrageenan over a 24 h time course. Cell influx increased 13-fold from 1 x 107 cells/ml (30min) to 1.3 x 108 cells/ml (4 h) and remained elevated through 24 h (1.4 108 cells/ml) (p < 0.05). At 4 h the cells present were >90% PMN's and by 24 h > 50% mononuclear cells were found. These percentages are consistent with previous findings. 2 Fluid extravasation significantly increased from 0.2 ml in saline treated animals to a maximum of 2.5 ml in carrageenan treated animals after 16 h (p < 0.05). Cell influx remained constant through 24 h, while fluid volume changes significantly decreased from 16 through 24 h (p < 0.05). Drugs: The anti-inflammatory drugs dexamethasone, piroxicam, and naproxen were tested in this model at 4 h, a time at which cells, fluid volume and IL-lfl levels were significantly elevated in the carrageenan treated rats. None of these drugs maintained cell influx and fluid volumes at control levels. At all doses tested, cell influx was approximately midway between the saline and the carrageenan treated animals ( Fig. 2A). (Dexamethasone was dosed from 0.01-10 mg/kg; piroxicam was dosed from 1-10mg/kg; naproxen was dosed at 40 and 100 mg/kg.) Fluid extravasation was significantly reduced in all drug doses by 15%, which was equivalent to 0.7 ml (data not shown).
IL-lfl levels decreased in the drug treated rats, but not in a dose dependent manner. Dexamethasone decreased IL-lfl levels by 64%, from 0.62 ng/ml at 0.03 mg/kg to 0.21 ng/ml at 1 mg/kg < 0.05). Piroxicam did not significantly reduce IL-1 levels (0.51 ng/ml). Naproxen, tested at a single dose of 40 mg/kg, reduced IL-lfl levels by 35%, from 0.51 ng/ml (controls) to 0.33 ng/ml at 40 mg/kg (Fig. 2B). rhIL-lra: The rhIL-lra also maintained cell influx approximately midway between the saline and carrageenan controls (Fig. 2A). The maximum reduction in cell influx was observed at the 1 mg/kg dose, decreasing the cells from 9.5 x 107 cells/ml to 7.3 x 107 cells/ml (23%; NS). The rhlL-lra significantly (p < 0.05) increased IL-lfi levels from 0.50 ng/ml to 0.78 ng/ml at 40 mg/kg in a non-dose dependent manner (Fig. 2B). fluid volumes with any of the dosing regimens in the carrageenan treated fasted or fed rats compared with saline treated controls (data not shown). IL-1/ levels in lavaged pleural cavities from fed animals dosed with rhIL-lra were significantly increased compared with saline controls, returning toward controls with 80 mg/kg (Table 1). Fasted animals had IL-lfl levels ranging from 0.40-+-0.04 ng/ml (carrageenan treated animals) to 0.79 -+-0.14 ng/ml in animals dosed with 20 mg/kg rhIL-lra. All IL-lfl levels were significantly increased compared with saline controls (0.25 _+ 0.01 ng/ml) ( Table 1).
Corticosterone levels" Plasma levels were measured in the fed and fasted rats as an indication of stress (Table 2). Normal rat corticosterone levels range from 10 to 60 ng/ml. 14 The rhIL-lra had no effect on plasma corticosterone levels in fasted rats. The fed carrageenan treated animals had corticosterone levels of 485-+-21 ng/ml. In fed animals, as the doses of rhIL-lra increased from 3 to 80 mg/kg, corticosterone levels decreased from 502 + 35 ng/ml to 42 +_ 7 ng/ml in a dose dependent manner (Table 1;p < 0.05), returning to the normal range.
In comparison, in fasted animals corticosterone levels did not decrease in a dose dependent manner ( Table 2); nor did the corticosterone levels return to the normal range (10-60ng/ml), even at 80mg/kg. At 40 and 80mg/kg rhIL-lra, the corticosterone levels were significantly increased in the fasted animals compared with the fed ones. These data demonstrate that carrageenan admin- istration significantly increases plasma corticosterone and the stress of fasting increases the variability of the plasma corticosterone levels.
Intrapleural injection of human IL-lfl or 0 or rat IL-lfl: IL-lfl (human and rat) and IL-1 (human) were injected intrapleurally to determine if any of these cytokines are pro-inflammagens by themselves. There were no significant changes in cell influx or fluid volumes from 30 min to 24 h with any of the cytokines tested in doses ranging from 0.001-100 /2g/rat for human and rat IL-lfl and 50-200 #g/rat for human IL-10 (Fig. 4).

Discussion
Carrageenan is an irritant which sets up a cascade of events that leads to cell and fluid influx in the pleural cavity, as described previously. 1-s Small blood vessels in the subpleural tissues become more permeable, allowing serum proteins, proteinases such as complement, and cells to move into the pleural cavity. The serum complement components may be responsible for the cell influx, since many are chemotactic. Complement inhibitors are active in this model within the first hour after the administration of the carrageenan. The purpose of the present study was to further characterize the model by examining the possible role of IL-1.
The total number of cells recoverable from the pleural cavity at the end of 4h was 1.2 108 cells/ml, which is consistent with the cell influx reported 6 h after carrageenan administration. 2 IL-lfl levels were maximum at 1 h, preceding the maximal influx of cells. However, cells remained elevated for 24 h, by which time 1L-1 levels had returned to control. This observation suggests that IL-1 may play a role in attracting cells into the pleural cavity. Since IL-1 was present in the pleural cavity, the possibility that the IL-1 receptor antagonist affected  The ED30 values for cellular infiltration with orally administered dexamethasone is 0.05 mg/kg, naproxen is 1.4mg/kg and piroxicam is 0.7 mg/kg, indicating that these drugs are active in this model, is Fluid volumes significantly decreased by approximately 15% with IL-1/ levels remaining constant.
These data substantiate the involvement of several mechanisms in the inflammatory process induced by carrageenan, since individually none of the drugs prevented the pro-inflammatory changes in the pleural cavity. The IL-1 receptor antagonist inhibits IL-1 by binding to its receptors and has no apparent agonist activity. 3 It has been shown to be active in several septic shock models, 16'7 increasing animal survival. Administration of this protein does not completely prevent IL-1/ influx or cell accumulation in the pleural cavity in response to carrageenan. Meyers et al. TM have reported that in rats fed ad libitum, the rhIL-lra decreased total cell influx in a dose dependent manner from 28% to 74% at doses ranging from 0.3-10 mg/kg. In our model using rats fed ad libitum, the rhlL-lra did decrease cell influx, but by no more than 45%. The response in the fasted animals was shown to be more variable, not dose dependent and the maximum inhibition of 38 Mediators of Inflammation. Vol 2.1993 cell influx was approximately 35%. In general, the IL-lfl levels in the pleural cavity were higher in the fasted animals than in the fed ones. Therefore, fasting and carrageenan induced inflammation together lead to a response that requires more rhlL-lra to down-regulate it.
IL-1 is one of the key mediators of the brain-endocrine immune response to stress. 9 IL-1 activates the hypothalamic-pituitary-adrenal axis; its effect on this axis appears to be mediated by activation of secretion of corticotropin releasing factor, leading to the release of ACTH and corticosterone. 2'2 In stress there are elevations in plasma corticosterone and IL-1 levels in rats. TM We stressed our animals by fasting them for 16 h prior to administration of the carrageenan. We hypothesized that this stress should increase plasma corticosterone levels and may increase IL-1 levels in the pleural cavity resulting in higher doses of the rhIL-lra to down-regulate this response. The rhlL-lra should lead to a reduction in the IL-1 induction of plasma corticosterone levels. Carrageenan does induce increased plasma corticosterone levels compared with controls whether the animals are fed or fasted. We observed more variability in the response of fasted animals to the rhIL-lra. In the rats fed ad libitum, plasma corticosterone levels decreased in a dose dependent manner as the dose of the rhIL-lra increased, suggesting that IL-1induced corticosterone secretion was effectively inhibited. Corticosterone levels in the food deprived animals did not return to normal. IL-1 levels in the pleural cavity only returned toward the saline controls at the highest dose of the receptor antagonist. These findings demonstrate that 16 h of food deprivation does play a role in the response of the animal to carrageenan, both in its plasma corticosterone and pleural cavity IL-lfl levels, and suggest that IL-1 may mediate this response.
In order to investigate the role of IL-1 as a pro-inflammatory agent in the pleurisy model, we injected human and rat IL-1/ and human IL-I directly into the pleural cavity. When the human and rat IL-10 and fl were injected across a wide range, there was no inflammatory response in the pleural cavity, as determined by changes in cell influx or fluid volume. These data suggest that IL-1 is an important component in the amplification of inflammation, but that it is not the initiator of the inflammatory events. In this model several inflammatory mediators including histamine, bradykinin and thromboxane have been detected in the pleural exudates of rats. The role these mediators play in exudate formation is unclear, since the injection of any one of these mediators alone does not result in a significant fluid or cellular response.
Thus, the observation that IL-1 per se is not a pro-inflammagen is consistent with the findings for other mediators. The combination of one or more of such mediators may be responsible for the inflammation.
The present work demonstrates that levels of IL-lfl are found in the pleural cavity of the carrageenan injected rats in association with increased cellular influx and oedema. The stress of 16h of food deprivation also influences the response of the animals to the carrageenan induced injury, indicated by increased plasma corticosterone and pleural IL-lfl levels. However, human IL-10 and fl and rat IL-lfi are not pro-inflammatory agents when injected individually into the rat pleural cavity under the conditions we used, suggesting that IL-1 per se is not the initiator of cell influx and fluid extravasation in this model.