Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Allergic airway inflammation is associated with a pronounced accumulation of eosinophils in the lung tissue and bronchial lumen, and this is likely to be a consequence of the coordinated expression of cytokines, chemokines, and adhesion molecules [see (1, 2)]. The cytokines interleukin (IL)-1 and tumor necrosis factor (TNF)- have been detected in the airways of asthmatics (3, 4). In a guinea pig model of allergic airway disease we demonstrated inhibitory effects of an IL-1 receptor antagonist on airway eosinophilia and hyperreactivity, as well as the generation of TNF-, following antigen challenge (5). Neutralization of TNF- inhibits antigen-induced lung eosinophil recruitment in the mouse (6) and guinea pig (our unpublished observations), and we have shown that intratracheal instillation of TNF- causes an eosinophilic lung inflammation in the guinea pig (7).

IL-13 and the related cytokine IL-4 are strongly implicated as key players in allergic disease. Although regarded principally as products of T lymphocytes (8), these cytokines are also produced by mast cells and basophils (9, 10), and their expression has been detected in bronchoalveolar lavage fluid (BALF) from antigen-challenged asthmatics (11, 12). Both cytokines induce immunoglobulin (Ig) class switching to IgE (13) and upregulate vascular cell adhesion molecule (VCAM)-1 expression (14). However, actions of IL-13 and IL-4 consistent with an anti-inflammatory role have been indicated by their activity in several systems. Both cytokines inhibit IL-1, TNF-, and nitric oxide synthesis but increase IL-1 receptor antagonist production (15). IL-13 inhibits chemokine synthesis by vascular endothelial and airway smooth-muscle cells (20), although it can enhance chemokine synthesis in some systems (23). In vivo, IL-13 has been shown to inhibit TNF- release, neutrophil accumulation, and TNF- production in rat lung immune complex injury (26), and endotoxin-induced lethality and increased serum TNF- in the mouse (27).

In this study we have examined the effect of human recombinant IL-13 on airway inflammation induced by guinea pig TNF- (gpTNF-) or antigen in the guinea pig. This study demonstrates that IL-13 exhibits anti-inflammatory actions in these systems that are not shared with human recombinant IL-4, and that the action of IL-13 may be related to its ability to suppress the release of eosinophil-stimulating activity in the airways.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Recombinant gpTNF- (specific activity of 160 U/ng as assessed by WEHI cytolytic assay) was produced and purified as described previously (7). Human recombinant IL-13 was from Sanofi (Labege, France) and human recombinant IL-4 was purchased from R&D Systems (Oxford, UK). Endotoxin levels for these cytokines, as determined by the suppliers, was less than 0.1 ng/mg protein. Unless otherwise stated, all other reagents were purchased from Sigma Chemical Co. (Poole, UK).

Animals

Animal care and all experimental procedures were carried out under UK Home Office Licence approval. Dunkin-Hartley guinea pigs were raised and housed in the University of Bath facilities. Animals (weighing 300 to 500 g at the beginning of procedures) of either sex were used throughout the study.

Intratracheal Instillation

For tracheal instillation, animals were sedated by intramuscular injection of 40 mg/kg ketamine (Vetalar; Parke-Davis Co., Pontypool, Gwent, UK) and 5 mg/kg xylazine (Rompun; Bayer Plc, Bury St. Edmunds, Suffolk, UK). Test agents or cytokine vehicle (0.1% low endotoxin bovine serum albumin [BSA] in pyrogen-free saline [Steripack, Runcorn, UK]) were instilled into the trachea by the orolaryngeal route with the aid of a laryngoscope. When combinations of agents were administered, these were given as a mixture in a total volume of 50 µl. Animals were observed until recovery from sedation and allowed free access to food and water until killing.

Antigen Sensitization and Challenge

Animals were sensitized by intraperitoneal injection of ovalbumin (OA, 10 µg) with aluminum hydroxide gel (2 mg) as adjuvant on Days 1 and 14. Tracheal instillation of cytokines or vehicle was performed between Days 28 and 42 as described, and animals were allowed to recover from sedation for 20 min before challenge with OA aerosol (100 mg in 10 ml saline generated using compressed air jet nebulizer at 7 liters/min) while protected from anaphylaxis with mepyramine (10 mg/kg intraperitoneally). Control (naive) animals were instilled with cytokines and exposed to OA aerosol without prior sensitization.

Assessment of Airway Inflammation

Airway inflammation was assessed by BAL 20 to 24 h after tracheal instillation. Animals were killed by an overdose of pentobarbitone and a midline incision made in the neck to allow insertion of a cannula into the trachea. Lungs were lavaged with 4 × 10 ml phosphate-buffered saline (PBS) pH 7.4 containing 1 mM ethylenediaminetetraacetic acid and 0.1% (wt/vol) BSA, and fractions were centrifuged to pellet BALF cells. The supernatant from the first two lavages was pooled and stored in aliquots at 20°C before bio- or immunoassay. The cell pellets from all four lavages were then combined, and cell infiltration was assessed by total counts using a hemocytometer and differential counts of cytospins stained with hematoxylin and eosin. In some experiments, lungs were dissected, chopped into pieces (approximately 5 mm²), rinsed in PBS, and stored at 70°C prior to eosinophil peroxidase (EPO) assay.

EPO Measurement

EPO was extracted from frozen lung pieces by grinding in a liquid nitrogen-cooled pestle and mortar followed by homogenization in PBS/0.5% hexadecyltrimethyl ammonium bromide. EPO was extracted by sonication and ultracentrifugation, and the sample was heated for 2 h at 60°C. Samples were then cooled and EPO activity was measured spectrophotometrically against a horseradish peroxidase standard using o-phenylenediamine substrate (30).

Eosinophil Activation Assay

The ability of BALF samples to elevate intracellular free calcium concentration ([Ca²⁺]_i) in guinea pig eosinophils loaded with fura-2 was assessed as previously described (31). Briefly, guinea pig eosinophils were elicited by repeated injection of horse serum into the peritoneal cavity and purified on a density gradient. Eosinophils of 95% purity were loaded by incubation with 2 µM fura-2 acetoxymethyl ester (Molecular Probes, Eugene, OR) for 30 min. After washing, the ability of agents to cause increases in eosinophil [Ca²⁺]_i was assessed by monitoring the fluorescence ratio (excitation 340 and 380 nm, emission 510 nm) of stirred eosinophil aliquots at 37°C using a fluorimeter linked to an on-line computerized data analysis system (PTI Inc., Surrey, UK) to convert fluorescence signals to nanomolar [Ca²⁺]_i.

Eotaxin Immunoassay

Guinea pig eotaxin was measured by enzyme-linked immunosorbent assay using a murine monoclonal antibody as capture and a rabbit polyclonal antibody (32) as detector. Synthetic guinea pig eotaxin (from Drs. G. Andrews and H. Showell, Pfizer Central Research, Groton, CT) was used as the standard over a range of 10 to 320 pM. After preliminary assays to determine optimum dilution, all BALF samples were diluted 40-fold with assay buffer (PBS/0.1% BSA/0.05% thimerosal) and assayed on the same day. The intra-assay coefficient of variation was 3.2%.

Statistical Analysis

Results are presented as mean ± SEM for n animals per group. BALF cell data were found to approximate a log normal distribution, and were log₁₀ transformed before analysis. Statistical significance for the differences between treatment groups was determined using one-way analysis of variance followed by the Student-Newman-Keuls method for multiple comparisons between groups (33).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Inhibition of gpTNF--Induced Airway Inflammation by IL-13

We have previously characterized the accumulation of eosinophils in BALF in response to gpTNF- (7); a 24-h time point and gpTNF- dose of 50 ng was used in this study to elicit a reproducible airway inflammation. Tracheal instillation of gpTNF- (50 ng) elicited a pronounced increase in BALF eosinophils and monocyte/ macrophages, with a smaller recruitment of neutrophils (see Figure 1). Total lymphocyte numbers in BALF were very low in vehicle-instilled animals (0.05 ± 0.002 million, mean ± SEM, n = 8), and numbers were slightly increased to 0.14 ± 0.02 million (n = 8, P < 0.05) by gpTNF- treatment. When coinstilled with gpTNF-, IL-13 (1, 10, or 100 ng) was found to inhibit markedly BALF leukocyte infiltration (Figure 1). The inhibitory effect of IL-13 was dose-related, with significant suppression of both eosinophil and monocyte/macrophage accumulation at 10 ng IL-13 and almost complete suppression at the highest IL-13 dose (100 ng) tested. The relatively small neutrophil accumulation induced by TNF- was not changed significantly by any of the IL-13 doses tested. Furthermore, IL-13 alone did not induce significant changes in the numbers of any leukocyte type compared with control, vehicle-instilled animals (Figure 1).

View larger version (39K): [in this window] [in a new window]   Figure 1.   Inhibition of gpTNF--induced lung inflammation by IL-13. Animals were instilled with cytokine vehicle (0.1% BSA/ saline), IL-13 alone (100 ng), gpTNF- (50 ng), or gpTNF- (50 ng) plus IL-13 (1 to 100 ng). Bars represent the mean ± SEM cell numbers for n = 6 to 8 animals per group. *P < 0.01, **P < 0.001 compared with vehicle-treated group; +P < 0.05, ++P < 0.01, +++P < 0.001 compared with animals treated with gpTNF- alone.

To assess the effect of cytokines on the trapping or accumulation of eosinophils in the lung interstitium, lung samples were taken from animals subsequent to BAL and EPO were measured. IL-13 alone had no effect on tissue EPO levels, but IL-13 significantly reduced the EPO levels in guinea pig lung when coinstilled with 50 ng TNF- (Table 1). Hence, IL-13 reduced both tissue and BALF eosinophilia in TNF--treated guinea pigs.

Effect of IL-4 on Lung Inflammation

The effect of tracheal instillation of IL-4 was assessed in a more limited study. In comparison with the effects of IL-13, IL-4 (100 ng) by itself increased monocyte/macrophage numbers in BALF from 9.1 ± 1.1 million (n = 8) to 17.6 ± 3.6 (n = 3), which was significant (P < 0.05). No significant change in the numbers of eosinophils, neutrophils, or lymphocytes was observed. Furthermore, cell counts in animals instilled with IL-4 (100 ng) plus gpTNF- (50 ng) were not significantly changed, compared with the response to gpTNF- alone. Cell counts (millions) for the gpTNF- alone group (n = 8) compared with the IL-4 plus gpTNF- (n = 3) group were 23.2 ± 1.4 versus 17.5 ± 3.3 for monocyte/macrophages, 21.6 ± 3.9 versus 20.1 ± 1.8 for eosinophils, 4.8 ± 1.6 versus 6.5 ± 2.9 for neutrophils, and 0.14 ± 0.02 compared with 0.12 ± 0.05 for lymphocytes.

Inhibition of Antigen-Induced Airway Inflammation by IL-13

The effect of intratracheal IL-13 instillation was also assessed in antigen-challenged animals. Aerosol antigen challenge of sensitized, but not naive, animals resulted in large increases in monocyte/macrophage, eosinophil, and neutrophil numbers in BALF (Figure 2). When animals were instilled with IL-13 (100 ng) prior to antigen challenge, eosinophil accumulation was inhibited by approximately 60%, and 1 and 10 ng of IL-13 were also able to reduce eosinophil numbers in BALF. The small reduction in antigen-induced monocyte/macrophage and neutrophil numbers did not achieve statistical significance. Lymphocyte numbers were increased in antigen-challenged, sensitized animals (0.9 ± 0.2 million cells [mean ± SEM, n = 6]) compared with 0.1 ± 0.02 million in challenged, naive animals, P < 0.01) and pretreatment of sensitized animals with IL-13 (10 ng) reduced this accumulation to 0.2 ± 0.03 million lymphocytes (P < 0.01).

View larger version (36K): [in this window] [in a new window]   Figure 2.   Inhibition of antigen-induced lung inflammation by IL-13. Naive or sensitized (sens) animals were instilled with cytokine vehicle or IL-13 (1 to 100 ng) 20 min prior to aerosol OA challenge. Bars represent the mean ± SEM cell numbers in BALF 24 h after OA challenge, n = 5 to 6 animals per dose. **P < 0.001 compared with antigen-challenged naive animals; ++P < 0.01, +++P < 0.001 compared with vehicle-pretreated sensitized animals.

Eotaxin Immunoreactivity and Eosinophil-Stimulating Activity in BALF

BALF eotaxin levels from gpTNF--treated animals were not changed compared with control animals, and these BALF samples did not elicit significant eosinophil calcium responses (data not shown). Exposure of sensitized animals to an aerosol of antigen resulted in significantly increased levels of immunoreactive eotaxin in BALF at 24 h when compared with eotaxin levels in naive animals 24 h after exposure to antigen (Table 2). This increase was reduced by over 50% by instillation with IL-13 (100 ng) prior to antigen challenge, but this change was not statistically significant, possibly due to the large variation in eotaxin levels within the groups. IL-13 instillation did not change eotaxin levels in BALF from naive, antigen-challenged animals.

Eosinophil-stimulating activity was assessed by the ability of BALF samples to elevate [Ca²⁺]_i in fura-2-loaded guinea pig eosinophils. As shown in Figure 3, BALF from sensitized, antigen-challenged animals stimulated rapid and transient increases in [Ca²⁺]_i, whereas BALF from naive, challenged animals did not stimulate large eosinophil calcium responses. Furthermore, treatment of animals with intratracheal IL-13 (100 ng) markedly reduced the ability of antigen to release eosinophil-stimulating activity into BALF.

View larger version (28K): [in this window] [in a new window]   Figure 3.   Elevation of [Ca2+]i in fura-2-loaded eosinophils by BAL supernatants taken 24 h after OA challenge. (A) Responses to BAL from vehicle-treated naive animals, vehicle-treated sensitized animals, or IL-13 (100 ng)-treated sensitized animals. Bars indicate mean ± SEM elevation of [Ca2+]i (nM) by 1/20 diluted BALF samples from 5 to 6 animals per group. **P < 0.001 compared with vehicle-treated naive group; ++P < 0.01 compared with vehicle-treated sensitized group. B, C, and D show sample calcium traces from eosinophils treated with 1/20 diluted BALF from vehicle-treated naive, vehicle-treated sensitized, or IL-13- treated sensitized animals, respectively. The arrows indicate the time of BALF addition to the eosinophils.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study reports, for the first time, potent anti-inflammatory effects of human recombinant IL-13 in guinea pig lung inflammation. These findings are in contrast with the well-established roles of Th₂ cytokines in the development of immediate hypersensitivity (34), and suggest that a selective promotion of certain Th₂ cytokine actions may be beneficial in allergic disease.

Intratracheal administration of IL-13 reduced both gpTNF- and antigen-induced airway inflammation. It is noteworthy that IL-13 reduced gpTNF- induced accumulation of eosinophils and monocyte/macrophages to baseline levels, whereas antigen-induced eosinophil accumulation was only partially suppressed, and the accumulation of other cell types in response to antigen was not significantly reduced. This may have been due to the stronger inflammatory signal promoted by antigen, which probably involves the sustained synthesis and release of a variety of inflammatory mediators. The early phase of leukocyte recruitment to antigen-challenged guinea pig lung probably involves the rapid release of lipid-derived chemoattractants, including eicosanoids and platelet-activating factor (35, 36). In addition, a delayed increase in the expression of eotaxin, monocyte chemotactic protein (MCP)-1, and macrophage inflammatory protein-1 is found in the airways of antigen-challenged animals and asthmatics (37- 40). Important roles for these chemokines in antigen-elicited leukocyte recruitment to the mouse lung have been demonstrated using neutralizing antibody or gene therapy techniques (37, 39, 41, 42).

In vitro, gpTNF- has only weak direct chemotactic activity (7), and the in vivo leukocyte attractant action of TNF- is likely to involve, at least in part, the synthesis of chemokines and increased adhesion molecule expression by lung cells (43). We therefore postulate that IL-13 suppresses airway inflammation by inhibiting the expression of such molecules. Bioassay and immunoassay of BALF indicated increased eosinophil-stimulating activity and elevated eotaxin levels following antigen challenge of sensitized animals. IL-13 treatment reduced both of these by over 50%, although the change in eotaxin concentration did not reach statistical significance. Eotaxin is a potent activator of eosinophil calcium responses (48) although, in a guinea pig model similar to that used in the present study, levels were maximal 3 to 6 h after challenge and were reduced (but still above control) at 24 h (32). Regulated on activation, normal T cells expressed and secreted is not an eosinophil attractant in the guinea pig (31) but may contribute to the monocyte/macrophage recruitment in our experiments.

In addition to suppressing leukocyte accumulation in BALF, IL-13 also inhibited gpTNF--induced increases in lung tissue EPO. These measurements were made subsequent to BAL, and so reflect the total numbers of tissue-associated eosinophils, but not those cells that have transmigrated into the bronchial lumen. Hence, the effect of IL-13 was not a consequence of increased eosinophil adherence to the lung vascular endothelium (47) or entrapment of cells in the interstitium without proceeding to transmigration into the bronchial lumen, but was a result of decreased eosinophil accumulation in all lung compartments.

The anti-inflammatory activity of IL-13 was in contrast with the weakly proinflammatory action of IL-4. Human IL-4 does not bind or activate murine IL-4 receptors (49, 50), although the present study suggests some activity of human IL-4 in the guinea pig. IL-13 does not have absolute species specificity (see Reference 8, and this work), although differences in potency may exist. To date, guinea pig IL-13 and IL-4 have not been identified, so we are unable to determine whether the effects reported here with human cytokines would also be true if syngenic cytokines were used. When studied in the same species, IL-13 generally exhibits a subset of the activities of IL-4 (8), and this is largely accounted for by a model by which IL-4 binds both IL-13 or IL-4 receptor subunits, but IL-13 does not bind the IL-4 receptor (see 51). However, an alternative IL-13 receptor  subunit has been identified and this does not interact with IL-4 (52). Although the significance of this subunit in the cellular response to cytokines remains to be established, it raises the possibility that, in some cell types, IL-13 may exhibit actions that are not shared with IL-4. Furthermore, recent studies (53) in human lung fibroblasts indicate that IL-4 and IL-13 activate different cell subsets and that IL-4 exhibits more pronounced inflammatory activity by increasing MCP-1 and VCAM-1 expression.

The results presented here indicate that it may be possible to dissociate the pro- and anti-inflammatory activities of Th₂ cytokines. Although the proallergic actions of IL-13 make the cytokine itself an unattractive therapeutic, selective promotion of the anti-inflammatory actions, via the IL-13 receptor, may provide a novel strategy in the therapy of eosinophilic lung inflammation.