Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cysteinyl leukotrienes (cys LTs) play an important role in the airway responses of a number of species, including human subjects, to allergen challenge (Foster et al., 1988; Manning et al., 1990; Sapienza et al., 1990; Taylor et al., 1991; Wegner et al., 1993). Early responses are generally thought to reflect mast cell activation and degranulation through IgE cross-bridging (Ishizaka et al., 1980). The airway responses appear to be mediated by both mast cell granule products, including histamine (Ishizaka et al., 1983) or serotonin in the rat (Nagase et al., 1995), and newly formed metabolites of arachidonic acid such as prostaglandin D₂ and cys LTs (Schleimer et al., 1986). LTD₄ may have an even more important role in the late allergic response. Blockade of the cys LT1 receptor in humans (Taylor et al., 1991), sheep (Wegner et al., 1993), and rat (Sapienza et al., 1990; Martin et al., 1993) markedly attenuates the late airway response (LAR). Indeed, in the sheep (Wegner et al., 1993) and in the rat (Sapienza et al., 1990), selective LTD₄ antagonists completely abolish the LAR, suggesting that LTD₄ is the single most important mediator of this reaction. The cellular source of cys LTs has not been as yet determined although the eosinophil is a plausible candidate in humans.

In an attempt to understand better the potential role of the T cell in allergen-induced late responses, we developed a T cell-dependent model of allergic responses that does not require IgE (Watanabe et al., 1995a, 1997) and therefore does not involve the participation of cells that are activated through either the high- or low-affinity receptors for IgE, FcR 1 or 2. CD4⁺ T cells harvested from allergen-primed donors transfer the ability to develop allergen-induced late responses and airway eosinophilia to recipient unsensitized rats but without the typical early responses (Watanabe et al., 1995a). It would not be too surprising if the principal biochemical mediators of T cell-mediated LAR were different from those seen in actively sensitized animals. The purpose of the present study was to determine therefore whether cys LTs are important mediators of CD4⁺-driven LAR, which, if true, would suggest that this LAR has a similar biochemical mechanism to the LAR in actively sensitized animals. To do this, we harvested T cells from the cervical lymph nodes of sensitized donor animals 2 weeks after sensitization in the s.c. tissues of the neck. The CD4⁺ cells were purified and administered i.p. to naive recipients that then underwent allergen challenge 2 days later as previously described (Watanabe et al., 1995a). Pulmonary resistance (R_L) measurements were made for 8 h after challenge for the determination of the LAR and bronchoalveolar lavage (BAL) was performed to determine the magnitude of the inflammatory response to challenge. The role of cys LTs was examined by testing the effects of pranlukast, a potent and selective cys LT1 receptor antagonist (Nakai et al., 1988) on changes in R_L and BAL fluid cells. We also examined the effects of cys LT antagonism on airway eosinophilia and interleukin (IL)-5, one of the principal cytokines responsible for eosinophil recruitment to the airways.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and Sensitization. Male Brown Norway (BN) rats (SSn substrain), ranging from 6 to 8 weeks in age and 190 to 240 g in weight, were purchased from Harlan Sprague-Dawley UK Inc. (Blackthorn, England) and maintained in conventional animal facilities at McGill University (Montreal, Quebec, Canada). Donor rats were actively sensitized with an s.c. injection of a mixture of 1 mg of ovalbumin (OVA) (Sigma Chemical Co., St. Louis, MO) and 4.28 mg of aluminum hydroxide gels (Anachemia Chemicals, Montreal, Quebec, Canada). Simultaneous injection of 0.5 ml of Bordetella pertussis vaccine containing 6 × 10⁹ heat-killed bacilli (Armand-Frappier Institute, Laval-Des-Rapides, Quebec, Canada) was performed i.p. as an adjuvant. These sensitized rats were used as donors of antigen-primed T cells, whereas naive BN rats were used as recipients for primed T cells.

Immunomagnetic Separation of CD4⁺ T Cell Subsets and Adoptive Transfer. Fourteen days after sensitization, T cells were isolated and CD4⁺ cells were purified with previously described techniques (Watanabe et al., 1995a). First, mononuclear cells were obtained from either two or three cervical lymph nodes of donor rats by mincing of tissue and subsequent passage through a stainless steel sieve. Enriched cells were then washed and passed through a nylon mesh to remove debris. Next, CD4⁺ T cells were obtained by negative selection with a magnetic cell sorter (MACS; Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). Cells in the cell suspension that we wished to remove were labeled with a mixture of primary specific monoclonal antibodies (mAbs), washed, and subsequently labeled with MACS rat anti-mouse IgG1 microbeads. The antibodies used were OX-8 (MCA P48, mouse anti-rat CD-8 mAb, IgG1; Cedarlane Laboratories, Hornby, Ontario, Canada) in combination with OX-33 (MCA P49, mouse anti-rat k chain mAb, IgG1; Prince Laboratories Inc., Toronto, Ontario, Canada) and ED9 (MCA P340, mouse anti-rat myeloid differentiation antigen, IgG1; Prince Laboratories Inc.) in a dilution of 1:100. The labeled cells were then washed and passed through the MACS column and the effluent containing the desired cells was collected. ED9, which recognizes a membrane antigen on rat macrophages, monocytes, dendritic cells, and granulocytes, was used to remove nonlymphocytic cells in the cell suspension, whereas OX-33 was used to remove B cells. Flow cytometry was performed on a sample of cell isolates with W3/25 (MCA P55, mouse anti-rat CD4 mAb, IgG1; Cedarlane Laboratories) to establish the purity of fractions of the CD4⁺ cells after purification by MACS.

Measurement of Physiological Changes of R_L. After transfer of cells, 2 days were allowed to elapse before antigen challenge. This time point was confirmed to result in adequate transfer of antigen sensitivity to airway challenge in the BN rat (Watanabe et al., 1995a). The recipients of 5 million CD4⁺ T cells were divided into three groups and were treated as follows: 1) one group (n = 9) was challenged with OVA 1 h after the administration by gavage of sodium carboxymethylcellulose solution (0.5% w/v in distilled water), the vehicle for pranlukast; 2) a second group (n = 8) was challenged with OVA 1 h after gavage with 3 mg/kg pranlukast in sodium carboxymethylcellulose solution (0.5% w/v in distilled water); and 3) the third group (n = 8) was challenged with BSA having received vehicle as described above.

BAL and Immunostaining for Major Basic Protein. BAL was performed 8 h after the challenge by instilling 25 ml of normal saline via the tracheal tube, and cell number was counted on a fresh specimen with a hemacytometer. With this procedure, ~90% of the instilled fluid is recovered and recovery is fairly consistent from rat to rat. Cytospin slides were prepared with a cytospin (model II; Shandon, Pittsburgh, PA) and air-dried for 5 min. The numbers of each cell were assessed by May-Grünwald-Giemsa staining on 200 cells. After acetone-methanol fixation, BAL cells were immunostained with an antibody to major basic protein (MBP) BMK 13 (kindly provided by Dr. Redwan Moqbel, University of Alberta, Edmonton, Alberta, Canada), and MBP-positive cells were quantified by the alkaline phosphatase anti-alkaline phosphatase method by an investigator blinded to group status. A minimum of 500 BAL cells was counted and the percentage of cells expressing MBP immunoreactivity was evaluated.

IL-5 Detection by In Situ Hybridization. To detect bronchoalveolar cells that were expressing IL-5 mRNA, the technique of in situ hybridization with a digoxigenin-labeled cRNA probe was used. The probe was generated from cDNA that was inserted into a pGEM vector and linearized with appropriate enzymes. Transcription was performed in the presence of SP6 or T7 RNA polymerases and labeling was done with digoxigenin-11-UTP so as to generate antisense and sense probes, respectively. The cytospins were permeabilized with proteinase K and then prehybridized with 50% formamide and 2× standard saline citrate. After application of the probe, the sections were hybridized at 42°C overnight. Nonspecific binding was removed by posthybridization washing under high-stringency conditions and subsequent treatment with RNase. The hybridization signal was visualized by incubating the cytospins overnight with sheep polyclonal anti-digoxigenin antibodies conjugated with alkaline phosphatase. Color development was achieved by adding the freshly prepared substrate (X-phosphate-5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium). To ensure the specificity of our signal, we performed the in situ hybridization with the sense probe and following pretreatment of the tissues with RNase. The slides were coded and positive cells were counted by an observer blinded to group status. The results were expressed as the mean numbers of positive cells per 1000 BAL cells.

Data Analysis. The early airway response (EAR) was calculated from the maximal R_L within 30 min after challenge as a percentage of the baseline value of R_L. The LAR was calculated as the area under the curve of R_L against time from 3 to 8 h after challenge, after subtraction of the baseline values of R_L.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of Pranlukast on EARs and LARs. The baseline R_L was not significantly different among the three groups (OVA-challenged group: 0.165 ± 0.012 cm H₂O/ml/s; BSA-challenged group: 0.152 ± 0.007; pranlukast-treated and OVA-challenged group: 0.164 ± 0.011; P > .05 by ANOVA). Rats undergoing OVA challenge demonstrated a small but significant increase in R_L within 30 min after OVA challenge (121.9 ± 4.2% baseline R_L; P < .01; Fig. 1). The R_L in the BSA-challenged group did not change significantly after BSA challenge (105.9 ± 1.8%; P > .05). Pretreatment with pranlukast 1 h before challenge failed to suppress this early rise in R_L after exposure to aerosolized OVA (115.9 ± 6.0%; P = NS; Fig. 2).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Changes in RL (percentage of baseline value) are shown as a function of time after OVA challenge in the vehicle-treated group (n = 9) and the pranlukast-treated group (n = 8) and after BSA challenge in the control group (n = 8). The symbols represent the mean data and the vertical lines are 1 S.E.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   EARs were calculated as the peak pulmonary resistance in the first 30 min after challenge are shown for the vehicle-treated group (n = 9), pranlukast-treated animals (n = 8), and the BSA-challenged group (n = 8). The horizontal lines represent the means of the data. An ANOVA and Tukey test were used to compare means. NS, not significant (P > .05).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   LARs, calculated as the area under the curve of RL against time from 3 to 8 h after OVA challenge, are shown for the vehicle-treated group (n = 9), pranlukast-treated animals (n = 8), and the BSA-challenged group (n = 8). The horizontal lines represent the mean of the data. An ANOVA and Tukey test were used to compare means. #P < .05, *P < .01.

Leukocyte Profile in BAL Fluid. There was a significant difference among total cell counts performed on BAL fluid (P < .05; ANOVA; Fig. 4). The difference was attributable to the OVA-challenged group, which had 3.55 ± 0.41 × 10⁶ cells compared with the BSA-challenged group (2.39 ± 0.18 × 10⁶ cells). Treatment with pranlukast did not modify the total cell counts (2.65 ± 0.45 × 10⁶ cells). Of the cells in BAL, there were significant differences observed both in the number of lymphocytes and eosinophils in the pranlukast-treated group. The lymphocytes in the OVA-challenged group were significantly higher (0.12 ± 0.02 × 10⁶ cells) compared with pranlukast-treated animals(0.02 ± 0.01 × 10⁶ cells; P < .05), whereas the eosinophils, as determined by positive immunostaining for MBP (Fig. 5), were significantly reduced after pranlukast (0.10 ± 0.01 × 10⁶ cells) compared with the OVA-challenged group (0.22 ± 0.03 × 10⁶ cells; P < .05).

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Cell counts on BAL fluid at 8 h after challenge are shown. Columns represent the mean values and the vertical bars are 1 S.E. The cells were identified on light microscopy with a modified Giemsa stain. Two hundred cells were counted. Comparison of means was made by ANOVA and Tukey tests. #P < .05, *P < .01.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Eosinophils in BAL fluid were identified by immunocytochemical staining with BMK13, an mAb against MBP. The columns indicate the mean values and the vertical bars represent 1 S.E. Comparison among means was done with an ANOVA and Tukey tests. *P < .01.

IL-5 Expression in BAL Cells. There was a markedly increased number of IL-5 mRNA-positive cells in the BAL fluid of OVA-challenged animals compared with the BSA-challenged group (Fig. 6). The pranlukast-treated animals had a significant reduction in IL-5-positive cells although the number of cells positive for IL-5 mRNA was still substantially elevated compared with that of the BSA-challenged animals.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Number of IL-5-expressing cells were counted on cytospins of BAL fluid and expressed per thousand cells. A digoxigenin-labeled cRNA probe was used to detect IL-5 by in situ hybridization. An ANOVA and Tukey test were used to compare differences among means. *P < .01.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have demonstrated in the current study that airway responsiveness to OVA challenge can be transferred by CD4⁺ T cells from sensitized donors to naive recipients and that such recipients develop late allergic responses after aerosol challenge. These results confirm our previous findings that LAR can be adoptively transferred (Watanabe et al., 1995a,b). The extent to which CD4⁺ T cell-driven late responses are similar in mechanisms to the LAR in actively sensitized animals is of particular interest. If indeed they were one and the same phenomenon, this would provide clear evidence of a dissociation between the IgE-mediated early responses and the LAR because OVA-specific IgE appears to be absent in recipients of CD4⁺ T cells despite the development of LAR (Watanabe et al., 1995a,b). Recent evidence supports the concept of independence of a number of allergen-triggered processes within the airways and allergen-specific IgE. Mehlhop et al. (1997) have demonstrated that eosinophilic bronchial inflammation and hyperresponsiveness to injected methacholine occur after allergen challenge of IgE-deficient mice and of comparable degree to those observed in wild-type mice. Likewise, OVA-induced hyperresponsiveness in sensitized mice is independent of IL-4 and allergen specific immunoglobulins (Hogan et al., 1997). The present study confirms the similarity of T cell-driven LAR to the LAR in actively sensitized animals, not only from the standpoint of temporal changes in pulmonary resistance, airway eosinophilia, and T cell cytokine expression (Watanabe et al., 1997) but also in terms of the important role of cys LTs in both responses.

The mechanism by which CD4⁺ T cells transfer sensitivity to OVA challenge to naive recipients has not yet been established. However, the responses occur after challenge with the sensitizing antigen only. Small but rapid changes in pulmonary resistance were observed after OVA challenge that were not significantly inhibited by pranlukast. Given the previously demonstrated lack of IgE, which is responsible for triggering usual early responses, the nature of these early responses is unknown. They do imply proximity of the effector cells to the initial sites of allergen deposition within the lungs. We speculate that the LAR results from antigen presentation by airway dendritic cells or other antigen-presenting cells that activate OVA-specific CD4⁺ cells that have migrated to the airways after their i.p. administration. Even if the inhibition by pranlukast of the CD4⁺ T cell-driven LAR is evidence that the cys LTs are important mediators of the LAR, the link between the CD4⁺ T cells and the LAR has not been made. Triggering of the LAR by T cells is presumably through the agency of other cells that synthesize cys LTs because T cells do not (Poubelle et al., 1987). However, the enzymatic machinery for cys LT synthesis (5-lipoxygenase and its activating protein FLAP, LTC₄ synthase) is regulated by the T cell cytokines, granulocyte macrophage colony-stimulating factor, and IL-3 (Pouliot et al., 1994; Ring et al., 1996) and IL-5 (Cowburn et al., 1999). The cells within the lung that are potential sites of cys LT production are the monocytes, macrophages, and mast cells. Eosinophils, which are present in considerable numbers after allergen challenge, can produce LTC₄ when harvested from humans but not from guinea pigs or rats (Sun et al., 1989; Hirata et al., 1990; Yu et al., 1995), thus their potential role in the LAR is more difficult to understand in guinea pigs and rats. However, another mechanism by which cells lacking LTC₄ synthase may contribute to cys LT production is through the export of LTA₄ for conversion to LTC₄ by transcellular metabolism by other cells within the lungs such as the neutrophil (Palmantier et al., 1998). But whether such a mechanism applies to the eosinophil does not appear to have been addressed.

OVA challenge of CD4⁺ recipients causes an increased expression of IL-4 and IL-5 in cells in the BAL fluid that has been interpreted as indicating the activation and recruitment of OVA-specific T cells of the Th2 phenotype, which are associated with allergic inflammation (Watanabe et al., 1997). Our findings are consistent in showing substantial numbers of IL-5 mRNA-expressing cells in the BAL fluid. Because the number of BAL lymphocytes also increased in response to allergen challenge, the increased number of IL-5-positive cells is probably also attributable to the T cells recruited into the airway lumen. The respective kinetics for the increased expression of IL-5 in resident and recruited T cells in response to allergen is not presently known in this animal model. However, we postulate that these two sources of T cells account for the increased number of BAL cells expressing IL-5, and that they both participate in the recruitment of eosinophils in the airways. Eosinophilia and IL-5 expression have been closely correlated in our previous studies (Watanabe et al., 1997) and in murine and guinea pig models of allergic inflammation this cytokine appears to account in large part for infiltration of the airways with eosinophils (van Oosterhout et al., 1993; Kung et al., 1995).

The recruitment of eosinophils into the airways after OVA challenge is probably complex, potentially involving several cytokines, chemokines, and lipid chemoattractant substances (Bell et al., 1997). The current study also revealed evidence of proinflammatory effects of cys LTs after allergen challenge. Pranlukast caused a reduction in both lymphocyte and eosinophil numbers in BAL fluid sampled at 8 h after challenge. Inhibitors of 5-lipoxygenase have been shown to reduce allergen-induced eosinophilia in the BN rat (Bell et al., 1997) and their efficacy has been usually attributed to LTB₄ (Richards et al., 1991). Cys LTs are not usually considered as chemotactic factors for inflammatory cells. The instillation of LTD₄ intratracheally in BN rats does not cause eosinophilia, arguing against a direct effect of cys-LTs on eosinophils (Stamatiou et al., 1998). However, several studies have shown chemotactic activity of cys-LTs for eosinophils both in vivo and in vitro (Foster and Chan, 1991; Laitinen et al., 1993; Spada et al., 1994) and cys LT1 antagonists have been described to reduce sputum and blood eosinophilia in human asthmatic subjects (Pizzichini et al., 1999). A recent study demonstrated prolonged eosinophilia in guinea pig airways after aerosolization of LTD₄ (Underwood et al., 1996). This effect was blocked by both pranlukast and an anti-IL5 antibody TRFK-5, again suggesting an indirect mechanism of action of cys LTs in the recruitment of eosinophils. The reduction in IL-5 in pranlukast-treated animals in the current study is also consistent with a role for this cytokine in the BAL eosinophilia. It is interesting in this regard that lymphocyte numbers also were reduced by pranlukast, which may account for the observed reduction in IL-5 expression. We cannot exclude the possibility that the reduction in IL-5-positive cells is in part a reflection of the diminution in eosinophil numbers because these cells also express IL-5 (Desreumaux et al., 1993). However, recent experiments have shown that IL-5 expression in BAL cells obtained from CD4⁺ T cell-transferred and OVA-challenged rats is largely associated with CD3⁺ cells (D. Ramos, Q. Hamid, and J. G. Martin, unpublished data). The C-C chemokine eotaxin is also important for the development of airway eosinophilia (Jose et al., 1994). When exogenous eotaxin is administered to mice it causes eosinophilia and airway hyperresponsiveness to infused acetylcholine (Hisada et al., 1999). Both phenomena are mediated in part by cys LTs; pranlukast attenuated both the degree of eosinophilia and airway hyperresponsiveness (Hisada et al., 1999). The above-mentioned results suggest that the well characterized dependence of allergen-induced eosinophilia on IL-5 and eotaxin involves an important interaction with cys LTs.

Although pranlukast inhibited the LAR completely, it reduced only partly the recruitment of eosinophils in OVA-challenged CD4⁺ recipients rats. Recently, pranlukast has been shown to decrease the production of the Th2 cytokines IL-4 and IL-5 but not the Th1 cytokine IL-2 by peripheral blood mononuclear cells of patients with asthma stimulated with specific antigens (Tohda et al., 1999). Interestingly, IL-2 is also a potent chemoattractant for eosinophils (Weller, 1992), and in the sensitized BN rat model, the administration of IL-2 has been demonstrated to enhance allergic responses by a mechanism other than an increase in cys LT production (Renzi et al., 1999). Collectively, these data suggest that the persistence of airway eosinophilia in the pranlukast-treated rats might be attributable to the chemotactic activity of IL-2. Furthermore, our results suggest that allergen-induced LAR and airway eosinophilia are two dissociable events. Although, the possibility that a specific eosinophil subpopulation is responsible for the development of the LAR and that the recruitment of these eosinophils in the airways was blocked by pranlukast is not excluded. In summary, CD4⁺ T cell-driven late allergic responses are mediated by LTD₄ and in this respect are similar to the late responses in actively sensitized animals. We postulate that the activation of 5-lipoxygenase and the synthesis of its products are under T cell control. Inhibition of LTD₄ by pranlukast, a potent and selective LTD₄ antagonist was effective in preventing not only increases in pulmonary resistance after challenge but also in reducing lymphocyte and eosinophil numbers in BAL fluid. There was also a reduction in IL-5-expressing cells in the airways. The mechanism of the effects on cellular influx into the airways is not known but based on current literature it seems likely that a reduction in IL-5 within the airways may be responsible for the alterations in eosinophil numbers. This animal model of allergic airway responses supports the efficacy of cys LT1 receptor antagonists as inhibitors of both bronchoconstriction and airway inflammation in asthma.