INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Two ₄ integrins have been characterized to date, namely ₄₁ (very late antigen-4 [VLA-4], CD49d/CD29) and ₄₇ (LPAM-1, CD41/CD61). VLA-4 is an adhesion receptor of particular pertinence for allergic inflammation and asthma. It is expressed on all mononuclear leukocytes, eosinophils, basophils, and mast cells (1, 2). It is also expressed at low levels on rat blood (3) and human blood neutrophils (4). The interaction of VLA-4 with its counterreceptors is implicated in the transendothelial and interstitial migration of monocytes and lymphocytes (5, 6). The interaction of VLA-4 with vascular cell adhesion molecule-1 (VCAM-1) is sufficient in itself to facilitate lymphocyte migration across the endothelium at inflammatory sites, in the absence of leukocyte function-associated antigen-1 (LFA-1) (7), and also mediates eosinophil recruitment in allergic reactions (8, 9). The ₄₇ integrin is expressed on most lymph node T and B cells, natural killer (NK) cells and also eosinophils. It binds MAdCAM-1 (mucosal addressin, gut homing receptor), VCAM-1, and fibronectin, and plays a role as a homing receptor of T cells to mucosal tissues (Peyer's patches) (10). The involvement of VLA-4 in lung pathology has been extensively characterized whereas the relative importance of VLA-4 and ₄₇ remains to be resolved. Available data, however, indicate that probably both ₄ integrins have important roles (11).

The blockade of leukocyte adhesion by monoclonal antibodies (mAb) directed to the ₄ integrin subunit (CD49d) decreases the late airway responses (LR) after allergen challenge in the sheep (12) and the guinea pig (13). In the Brown Norway (BN) rat model of allergic bronchoconstriction, the blockade of either ₄ integrins or ₂ integrins (LFA-1 and Mac-1) inhibits both the early responses (ER) and LR in sensitized animals (14). The reduction of the ER caused by ₄ blockade in the rat was confirmed by a similar study in the sheep using a peptide which was a CS-1 ligand mimic, the binding site of VLA-4 on fibronectin (15). The reduction of the ER in the rat was associated with a decrease in histamine in the bronchoalveolar lavage (BAL) and cysteinyl-leukotriene (cys-LT) excretion in the bile, consistent with mast cell inhibition (16). Adhesion receptors have trans-membrane signaling properties that mediate cell activation (2, 17), supporting the plausibility of interference with cell activation by mAb as the mechanism of inhibition of airway responses. Such mechanisms may be different from actions on cell recruitment. The mechanisms of the reduction of the LR by anti-₄-integrin pretreatment is uncertain. The possibility that it is a secondary event caused by inhibition of the ER seems unlikely because anti-₄ integrin treatment administered 2 h after the ER was also effective in the sheep (12). Several cell types other than mast cells express ₄-integrins and are potential sites of action for such treatments. Anti-₄-integrin treatment inhibits eosinophil activation as indicated by reduction in eosinophil peroxidase (EPO) (12). However, the T cell also expresses ₄-integrins (1, 2, 11) and this cell is of importance in LR.

To test the importance of interference with T-cell activation in the mediation of the effects of anti-₄-integrin treatment in the rat, we employed the technique of adoptive transfer to generate LR (18). CD4⁺ T cells harvested from sensitized donor rats transfer allergen sensitivity to naive recipients, which develop LR after challenge. These LR appear to be similar from the standpoint of the pattern of inflammation, T-cell cytokine expression (20) and the predominant biochemical mediators, cys-LT (21), to those that develop after challenge of actively sensitized rats, but are not associated with antigen-specific IgE and typical ER (19). The effects of anti-₄-integrin treatment with a well-characterized mAb, TA-2 (22), on the T-cell driven LR, the number of T cells in the airways as well as the number of cells expressing T helper cell, type 2 (Th2) cytokines in the BAL were examined. We reasoned that interference with T-cell migration to the airways would be reflected in a reduction in CD3⁺ cells whereas effects on activation would result in a disproportionate reduction of T-cell cytokine expression compared with cell numbers.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals and Sensitization

Inbred male BN rats (8 ± 1 wk old, 189 ± 32 g) were sensitized to ovalbumin (OVA) by subcutaneous injection of a suspension containing 1 mg of OVA (grade V; Sigma Chemical Co., St. Louis, MO) and 4.3 mg of aluminum hydroxide (EM Industries Inc., Gibbstown, NJ) in normal saline. An intraperitoneal injection of Bordetella pertussis vaccine (IAF; Laval-des-Rapides, QC, Canada) containing 5 × 10⁹ heat-killed bacilli was administered as an adjuvant. The rats were purchased from Harlan Olac (Bicester, Oxon, UK) and maintained in a conventional animal facility at McGill University before study. The research protocol was in compliance with the guidelines of the Canadian Council on Animal Care and was approved by the animal care committee of McGill University.

CD4⁺ Cell Purification and Transfer

Fourteen days after sensitization, a purified suspension of CD4⁺ cells was prepared from the cervical lymph nodes of sensitized rats by immunomagnetic cell separation (magnetic activated cell sorting [MACS]). In each experiment, 5 to 7 cervical lymph nodes per animal were dissected from two sensitized donors, and 2 × 10⁶ CD4⁺ cells from the final mixed suspension were injected intraperitoneally into each of 2 or 3 unsensitized and syngeneic recipient rats. The donor lymph nodes were dissected, isolated from the surrounding connective tissue, and transferred to a phosphate-buffered saline (PBS) medium containing 0.5% bovine serum albumin (BSA) (Fraction V; Sigma Chemical Co., St. Louis, MO). Cells were harvested by mincing and passing the tissue through a stainless steel sieve. The cell suspension was twice filtered through a nylon mesh and centrifuged at 274 × g for 5 min. The pellet was incubated with 20 µl per 10⁶ cells of a 5 µg/ml dilution of the following primary antibodies, purchased from Cedarlane Laboratories (Hornby, ON, Canada): ED9 (IgG₁ mouse anti-rat Myeloid Differentiation Antigen), OX33 (IgG₁ mouse anti-rat B cell), and OX8 (IgG₁ mouse anti-rat CD8). The cells were then washed and incubated with magnetic microbeads conjugated to monoclonal rat anti-mouse IgG₁ (Miltenyi Biotec, Bergisch Gladbach, Germany). Cells were resuspended in 0.5 ml of 0.5% BSA/PBS and passed through a depletion column across a magnetic field provided by magnetic cell separator (Miltenyi Biotec).

Phenotyping of Purified CD4⁺ T Cells

The purity of the CD4⁺ lymphocyte suspensions and the expression of VLA-4 were assessed by flow cytometry. Aliquots containing approximately 10⁶ cells were respectively incubated with ED9, OX33, OX8, W3/25 (IgG₁ mouse anti-rat CD4⁺; Cedarlane) and TA-2 (IgG₁ mouse mAb to rat ₄-integrin subunit) (22) primary antibodies. BALB/ C mouse control ascites fluid was used as a negative control. A fluorescein isothiocyanate (FITC)-labeled goat anti-mouse IgG (Life Technologies Inc., Gaithersburg, MD) was the secondary antibody that was used for detection. The TA-2 mAb was provided by Dr. T. B. Issekutz (Dalhousie University, Halifax, NS, Canada).

Study Protocol

Two days after the CD4⁺ cell transfer, the recipient rats were randomly assigned to one of three groups. In one group (TA-2 group, n = 8) the rats received an intraperitoneal injection containing 7 mg/kg of TA-2 and 1 h later they were challenged for 5 min with 5% aerosolized OVA using a Hudson nebulizer (Hudson Respiratory Care, Inc., Temecula, CA). This dose of TA-2 inhibited both ER and LR in actively sensitized animals (14). An intraperitoneal dose of 6 mg/kg was reported to produce plasma concentrations that are sufficient to inhibit adhesion of leukocytes to the endothelium (23).

The rats assigned to the second group (control-Ab group, n = 10) were injected intraperitoneally with an isotype-matched irrelevant antibody (3H11-B9; Dr. T. B. Issekutz, Dalhousie University), and were also challenged with 5% aerosolized OVA. The CD4⁺ cell recipients assigned to the third group were sham-challenged with 5% aerosolized BSA (sham group, n = 8), without being submitted to any treatment.

Measurement of Airway Responses to Challenge

Airway responses were measured during an 8-h period after challenge (24). The animals were anesthetized with 1.5 g/kg of urethane intraperitoneally (Sigma Chemical Co.) and intubated with a 6-cm-long endotracheal polyethylene tube (PE 240; Becton-Dickinson, Sparks, MD). They were then placed on a heating pad, and rectal temperature was continuously monitored. The free end of the endotracheal tube was plugged into a Plexiglas chamber, attached at its opposite end to a pneumotachograph coupled to a differential pressure transducer. A PE-160 polyethylene catheter, filled with normal saline and connected to a second differential pressure transducer, was placed in the esophagus. The other port of this transducer was referenced to the Plexiglas chamber. The chamber was flushed with a continuous flow of 1.5 L/min of oxygen-enriched air (47% oxygen) between neubulizations or measurements.

Transpulmonary pressure and flow were measured every 15 min during spontaneous tidal breathing. The signals from both transducers were recorded for 10 s, amplified, and passed through an analog to digital converter to a personal computer. Pulmonary resistance (RL, cm H₂O · ml¹ · s) was calculated with a data analysis software package (Anadat; RHT Info Dat Inc., Montreal, QC, Canada). The late response was quantitated as the area between the curve of RL and the baseline RL during the 3- to 8-h interval after challenge.

BAL and Tissue Preparation

At the end of the in vivo recording of RL, the animals received an overdose of anesthesia followed by BAL. The lavage was performed with 25 ml of saline, and total cell counts were done in a hemacytometer. BAL fluid was then centrifuged and resuspended in Hanks' balanced salt solution (Life Technologies Inc., Grand Island, NY). Slides for microscopy were prepared by cytocentrifugation (Cytospin II; Shandon, Cheshire, UK). Differential counts for macrophages, neutrophils, and lymphocytes were based on May-Grünwald-Giemsa staining and standard morphologic criteria. Eosinophils were immunostained for major basic protein (MBP) on acetone/methanol-fixed specimens. A set of slides was fixed in 4% paraformaldehyde and stored at 80° C until being further analyzed by in situ hybridization.

After BAL, the lungs and tracheobronchial tree were dissected, and the lungs were inflated and fixed for 4 h in 4% paraformaldehyde. The lungs were then midsagittally sectioned, and tissue blocks were sampled, washed in 15% sucrose-PBS and stored at 80° C for further analysis.

Immunocytochemistry and In Situ Hybridization

The MBP immunostaining to detect eosinophils in BAL was performed using an alkaline phosphatase antialkaline phosphatase (APAAP) method. Cytocentrifuged specimens were incubated with a murine monoclonal IgG₁ antibody against the eosinophilic granule protein MBP (clone BMK-13, a gift of Dr. R. Moqbel, University of Alberta), and with an unconjugated rabbit anti-mouse immunoglobulin secondary antibody (DAKO Corporation, Carpinteria, CA), followed by incubation with APAAP mouse monoclonal complex (DAKO Corp.). The slides were developed with alkaline phosphatase substrate (naphthol phosphate, Sigma Chemical Co.), added to Fast Red (Sigma Chemical Co.) as a chromogen, and counterstained with hematoxylin.

On tissue sections, immunostaining for CD3⁺ cells was done using a labeled streptavidin-biotin peroxidase technique. Incubation with a primary mouse anti-rat CD3 mAb (Pharmingen, San Diego, CA) was followed by a detection system consisting of a biotinylated secondary antibody, a peroxidase-conjugated streptavidin complex, and a diaminobenzidine-based substrate-chromogen (reagents from DAKO Corp.). A purified mouse IgG₁ isotype standard (Pharmingen) was used as primary antibody for negative control.

Cells expressing messenger RNA (mRNA) for interleukin-5 (IL-5) or interferon gamma (IFN-) were detected in BAL cytospin slides by in situ hybridization. For this purpose, radiolabeled antisense and sense riboprobes were prepared from complementary DNA (cDNA) coding for the corresponding cytokines. The cDNAs were inserted into pGEM vectors, grown in Escherichia coli and linearized. The radiolabeled RNA probes were obtained by in vitro transcription in the presence of ³⁵S-uridine triphosphate, using the T7 or SP6 RNA polymerases. The specimens were permeabilized with 1 µg/ml proteinase K, and underwent a prehybridization treatment to prevent nonspecific binding of the RNA probes and optimize the efficiency of the hybridization. Prehybridization consisted of a 30-min incubation at 37° C in 10 mM N-ethyl maleimide and 10 mM iodoacetamide, followed by a 10-min incubation in 0.5% acetic anhydride and 0.1 M triethanolamide, and a 15-min exposure to 50% formamide and standard saline citrate (SSC) at 40° C. Hybridization was performed with the radiolabeled RNA probes, diluted in a hybridization buffer containing 100 mM dithiothreitol, to block nonspecific binding of the ³⁵S-labeled probes. The specimens underwent posthybridization washings through decreasing concentrations of SSC at 40° C. Unhybridized single-stranded complementary RNA (cRNA) probes were removed by incubation in a 20 µg/ml solution of ribonuclease A (RNase A). After dehydration in ethanol, the slides were exposed to autoradiography emulsion (Amersham LM-2; Nycomed Amersham Ltd, Oakville, ON, Canada) for 10 d, developed (Kodak D19 developer; Eastman Kodak, Rochester, NY), and counterstained with hematoxylin. Negative control experiments using sense probes and RNase pretreatment of sections before antisense probe application, were carried out to confirm probe specificity.

To identify the proportions of the cells expressing IL-5 mRNA in BAL that corresponded to T cells or eosinophils, colocalization experiments by means of simultaneous immunocytochemistry and in situ hybridization were done. The BAL cytocentrifuged specimens were first immunostained for CD3 or MBP, and then processed for in situ hybridization to detect IL-5 mRNA.

Morphometry on Airway Sections

The density of CD3⁺ cells related to surface unit of airway sections was measured. For this purpose, positive cells were counted between the inner limit of the epithelium and the outer limit of the adventitia, and referenced to the examined area. Various intrapulmonary airways (11 ± 3) were averaged per animal. The measurement of the airway wall section area was computer-assisted, by superimposing the image of the airway onto a calibrated digitizing tablet.

Data Analysis

Data are expressed as mean ± SEM, or as median and interquartile range in the case of those distributions with significant skewness. Cellular counts in BAL (leukocyte subsets and positive cells for IL-5 or IFN- mRNA) are given in both relative (cells percent of per thousand) and absolute terms (referenced to the total cell counts, × 10³ cells/ml). CD3⁺ cell counts in tissues (airway sections) are expressed as cells/mm². All specimens for microscopic analysis were coded and examined in a blinded fashion.

Comparisons for quantitative variables among independent groups were evaluated with one-way analysis of variance (ANOVA), and Fisher's least significant difference test for post hoc multiple comparisons. The 95% confidence interval (95% CI) of the mean difference was calculated to estimate the size of effects, where appropriate. Associations between quantitative variables were assessed with Pearson's coefficient of correlation. A p value less than 0.05 is considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of ₄-integrin Blockage on Adoptively Transferred LR

The baseline RL was 0.139 ± 0.006 cm H₂O · ml¹ · s (mean ± SEM) for the overall study population, and there were no significant differences among the groups (p = 0.80). The average RL increased after the third hour after antigen challenge in the control-Ab-treated group, whereas it remained unchanged in the TA-2-treated and the BSA-challenged groups (Figure 1A). The values of LR are detailed in Table 1. The TA-2-treated groups had significantly reduced LR compared with the control-Ab groups, which had larger responses than the sham-challenged group (Figure 1B). TA-2 reduced the median LR to 13.6% of its value in the control-Ab group.

View larger version (31K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Figure 1.   Late airway responses. (A) Mean RL against time. Error bars represent SEM. (B) LR distributions. LR was calculated for each animal, as the area between the curve of RL and the RL baseline value from the third to the eighth hours after challenge. The boxes represent median and interquartile range (Q25 to Q75), and the percentiles P10 and P90 (bars). Circles represent data for the individual animals. *p < 0.05.

BAL Leukocyte Counts

The values for total and differential cell counts in BAL are shown in Table 2. No significant differences were observed for the counts of total leukocytes, and for the differential counts of macrophages, neutrophils, or lymphocytes, based on May-Grünwald-Giemsa staining of cytocentrifuged specimens and standard morphologic criteria (Figure 2A). The LR correlated moderately with the neutrophils expressed as a percent (r = 0.621; p = 0.001) or as absolute counts (r = 0.508; p = 0.026) in the study population, regardless of the fact that the differences in the neutrophil counts among groups were not significant. This overall correlation was driven by a strong partial correlation within the control-Ab group (r = 0.92, p = 0.001 for the relative counts; r = 0.92, p = 0.026 for the absolute counts (Figure 2B). No correlation between the LR and BAL neutrophils was observed in either the sham-challenged or the TA-2 groups.

View larger version (38K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Figure 2.   Leukocyte counts in BAL. (A) Differential counts for macrophages, neutrophils, and lymphocytes, based on May-Grünwald- Giemsa stain. Macrophages are referenced to the left ordinate, and neutrophils and lymphocytes to the right ordinate. Bars represent mean and SEM. The eosinophil counts were based on MBP immunostaining, and are shown in Figure 3. (B) Relationship between the LR and the BAL neutrophils and eosinophils (MBP+ cells). Boxes: Partial correlation LR versus neutrophils in the antigen-challenged, control (cAb) group (r = 0.92, p = 0.001). Circles: Correlation LR versus MBP+ cells for the study population (r = 0.46, p = 0.019).

View larger version (17K): [in this window] [in a new window]   View larger version (29K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 3.   Phenotype of transferred lymphocytes, airway T cells, and cytokine expression. (A) Analysis of the transferred cell populations by flow cytometry. The cells were harvested from the cervical lymph nodes of OVA- sensitized donors, and purified for CD4. The plot shows the overlaid histograms for the IgG1 isotype control immunoglobulin (1), TA-2 (anti-4) labeled cells (2), and CD4-labeled cell (3). The 4-integrin distribution, after control peak subtraction, accounts for 93.2% of the CD4+ cells. (B) Distributions of CD3+ cells in airway sections (cells/mm2). The bars represent mean and standard error. The antigen challenge in the control-Ab group and the TA-2 treatment did not result in significant differences. (C) IL-5 mRNA+ cells in BAL. 0/00: Cells per thousand. *p < 0.05. Boxes represent median and interquartile range (Q25 to Q75), and bars represent percentiles P10 and P90. (D) IFN- mRNA+ cells in BAL (cells per thousand). *p < 0.05. Boxes and bars as defined in C.

Cellular Phenotypes and Cytokine Expression

The purity of the CD4⁺ cell suspensions harvested from the sensitized donors and purified by MACS averaged 97.5% ± 0.7% in independent determinations from three experiments. The ₄-integrin expression was detectable by a TA-2 dilution of 2.5 µg/ml in 93.2% of the transferred CD4⁺ cells (Figure 3A).

No significant differences were found among the groups for the counts of immunostained CD3⁺ cells per unit area of airway wall (Figure 3B). There were 116.1 ± 13.4 CD3⁺ cells/mm² in the sham-challenged group; 110.2 ± 17.4 CD3⁺ cells/mm² in the control-Ab group; and 136.3 ± 16.5 CD3⁺ cells/mm² in the TA-2 group (mean ± SEM; p = 0.506).

The counts of positive cells for IL-5 or IFN- mRNA in BAL are given in Table 3. IL-5 mRNA⁺ cells in BAL were increased in the control-Ab-treated rats, compared with the sham-challenged group, both for the relative (p < 0.001) and absolute (p = 0.004) counts. The TA-2-treated rats had a significant reduction of the IL-5 mRNA⁺ cells, compared with the control-Ab-treated group (p < 0.001 and p = 0.004, for relative and absolute counts respectively) (Figure 3C). The LR correlated weakly but significantly with the proportion of BAL cells expressing IL-5 mRNA in the overall study population (r = 0.413; p = 0.040). The proportion of cells expressing IFN- mRNA was significantly decreased in both OVA-challenged groups (p < 0.05), compared with the sham-challenged animals (Figure 3D).

The rats in the control-Ab group had greater counts for immunostained (MBP⁺) eosinophils in BAL than the sham-challenged animals, in relative (p < 0.001) or absolute (p = 0.015) terms. The anti-₄-integrin treatment (TA-2 group) decreased significantly the MBP⁺ cell counts, with respect to the control-Ab group (relative: p < 0.001; absolute: p = 0.008) (Figure 4A). Neither the counts of BAL cells expressing IL-5 or IFN- mRNA, nor the MBP⁺ cells, correlated significantly with the differential counts of lymphocytes, neutrophils, or macrophages. The intensity of the LR correlated moderately, for the overall study population, with the MBP⁺ relative counts in BAL (r = 0.46, p = 0.019) and was comparably correlated with the absolute counts (Spearman's rho = 0.45, p = 0.052) (Figure 2B).

View larger version (14K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Figure 4.   Relationship between BAL eosinophils and IL-5 expression. (A) Eosinophil counts in BAL, as detected by MBP immunostaining. 0/00: Cells per thousand. *p < 0.05. Boxes represent median and interquartile range (Q25 to Q75), and bars represent percentiles P10 and P90. (B) Correlation of eosinophils (MBP+) versus IL-5 mRNA+ cells in BAL (r = 0.96, p < 0.001). 0/00: Cells per thousand. The linear regression function for the study population is represented. The colocalization for IL-5 mRNA and CD3 or MBP demonstrated that the major source of IL-5 was T cells (77.5% of IL-5 mRNA+ cells).

A strong correlation was observed between IL-5 mRNA⁺ and MBP⁺ cells (Figure 4B) for the study population (r = 0.962, p < 0.001 as cells per thousand, and r = 0.975, p < 0.001 for absolute counts). The correlation in the control-Ab group was significant, for both the relative (r = 0.878, p = 0.002) and the absolute (r = 0.997, p < 0.001) counts.

Localization of IL-5 mRNA to T cells and Eosinophils

The MBP and CD3 phenotypes were analyzed in the BAL cells expressing IL-5 mRNA, by means of in situ hybridization for IL-5 mRNA in combination with immunocytochemistry for MBP or CD3. This colocalization was done to confirm that CD3⁺ cells were the main source of IL-5 mRNA, and to discount the possibility that IL-5 mRNA⁺ eosinophils accounted for the correlation between MBP⁺ and IL-5 mRNA⁺ cells. The CD3⁺ cells averaged 77.5% ± 3.05% of the IL-5 mRNA⁺ cells, and only 15.5% ± 2.16% of the IL-5 mRNA⁺ cells were MBP⁺ (eosinophils). The proportions of IL-5 mRNA positive cells that were CD3⁺ or MBP⁺ were not significantly different among the groups.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The adoptive transfer of LR, in the absence of ER, by means of antigen-primed CD4⁺ lymphocytes in the BN rat provides a means of testing if the inhibition of the LR by the blockade of ₄-integrins is the result of effects on the cells mediating the LR. Using such a model we found that the blockade of the ₄ integrin subunit with a mAb (TA-2) abolished the CD4⁺ T-cell-driven LR and decreased the numbers of eosinophils (MBP⁺ cells) and cells expressing IL-5 mRNA in BAL. The effect on eosinophil numbers was unaccompanied by any demonstrable changes in the BAL counts of total cells, macrophages, neutrophils, or lymphocytes, or in the density of CD3⁺ cells in airway sections.

Several studies have demonstrated effects of anti-₄ integrin treatments on allergen-induced airway responses. The blockade of ₄ integrins by mAb in the guinea pig abolished LR although it had variable effects on airway hyperresponsiveness (AHR) (13, 25, 26). Alpha-4-integrin blockade inhibits not only late onset events such as the LR but also the ER (12, 14) and the release of mast cell mediators (16). It is possible that the inhibition of mast cell activation could contribute to the inhibition of the LR. However, Abraham and coworkers (12) tested the effects of an anti-₄ integrin mAb (HP ¹/₂) given 2 h after allergen challenge to Ascaris suum-sensitized sheep and this postchallenge treatment was effective in blocking LR and AHR. In the present study the ₄-integrin blockage also resulted in an effective inhibition of the CD4⁺ T-cell-dependent LR. These observations support the hypothesis that the inhibition of the LR as a result of ₄-integrin blockade can be mediated independently of the effects on the mast cell.

The eosinophil has been postulated to be the cellular target for the observed effects of anti-₄-integrin treatment on LR in the sheep (12). The altered LR and AHR were attributed to effects on eosinophil activation because EPO was reduced by the treatment. Effects on both eosinophils and T cells have been reported also in the other species. In the guinea pig reduced numbers of lung T cells and eosinophils and a diminution of EPO activity (13, 25, 26) have been measured. In the mouse, the blockade of VLA-4 or its ligand VCAM-1 inhibited antigen-induced eosinophil and T-cell recruitment into the trachea (27). We reasoned also that the T cell was a plausible target for inhibitory effects on LR not only because it expresses ₄ integrins but because the eosinophil does not synthesize cys-LT in the rodent and is therefore unlikely to be directly involved in the LR (28, 29). The reduction in the T cell cytokine, IL-5, by TA-2 treatment in the current experiments supports the role of the T cell as the primary target for anti-₄-integrin treatment. The predominant localization of IL-5 to CD3⁺ cells indicates that changes in other cellular sources of IL-5 are not sufficient to account for the observed degree of inhibition of this cytokine.

The results of our study raise the issue of the relative importance of the blockade of ₄ integrins on cell migration versus activation in the inhibition of the LR. Although LR and AHR are often linked to leukocyte infiltration, previous experiments in the BN rat (14, 30) and the sheep (12) suggest that impaired cell migration is unlikely to be the entire explanation of the observed effects. The attenuation of the ER which appears to be attributable to inhibition of mast cell activation is strong evidence for interference with cell activation, but the lack of consistent changes in inflammatory cell populations in BAL or airway tissues also requires explanation. Henderson and coworkers have deduced that interference with T-cell activation is an important part of the anti-₄-integrin effect in the inhibition of allergen-induced AHR in the mouse (31). The systemic administration of mAb to ₄-integrins abrogated eosinophil recruitment in BAL, but did not affect AHR or IL-4 and IL-5 synthesis, whereas the local blockade of ₄ integrins on intrapulmonary leukocytes by intranasally administered mAb prevented AHR and the synthesis of IL-4 and IL-5. This finding was interpreted as indicating an important local effect of the intranasally administered mAb on T-cell activation. In the current study the reduction in IL-5 in the absence of a change in the density of airway T cells is consistent also with a greater importance of activation over migration processes. VLA-4 has signal transduction properties and has been implicated as a costimulatory or "second signal" molecule in antigen presentation (32), and it is possible that it is interference with this event that accounts for our findings. The absence of an increase of T cells in the airways of the animals in the control-Ab group suggests that a rapid phase of T-cell recruitment is unlikely to occur upon antigen challenge.

The strong correlation between the IL-5 mRNA⁺ and MBP⁺ cells in BAL indicates the important relationship between IL-5, which in our experiments colocalized to predominantly CD3⁺ cells, and airway eosinophil infiltration (36). We speculate that a reduction of IL-5 secretion by airway T cells was likely responsible for the reduction in MBP⁺ cells. The mechanism of the reduction in IL-5 expression by anti-₄-integrin treatment is unclear. The transferred CD4⁺ cells, which come from sensitized rats and drive the LR in this system, have not been tracked and identified in the recipient rats. These transferred cells are critical for the LR, presumably because some of them home to the lung or associated secondary lymphoid structures of the recipients. This likely happens during the 48 h that elapse between the cell transfer and the allergen challenge. These cells (2 · 10⁶) represent a very small fraction of the total T-cell pool in the recipients and their presence is not readily detectable as a change in airway T-cell numbers, nor can they be visualized by labeling these cells fluorescently prior to transfer (unpublished observations). These cells are unlikely to be numerous enough to account for all of the observed IL-5 expression in the BAL. A previous study using antisense oligodeoxynucleotides (ODN) to target cytokine gene expression in the CD4⁺ transferred cell population showed that IL-5 synthesis by BAL cells in OVA-challenged recipients was only slightly reduced by pretreatment of the transferred cells with anti-IL-5 antisense ODN (37).

The OVA-challenged animals, whether treated with TA-2 or with control-Ab, had significantly reduced frequencies of IFN- mRNA⁺ cells in BAL, compared with the sham-challenged group. This result is suggestive of an IFN- downregulation in these animals, and it is consistent with a Th2 lymphocyte shift in the context of a response to allergen in this model. However, the anti-₄ TA-2 antibody did not have any effect per se on the expression of IFN- mRNA in this study, suggesting a selective effect of TA-2 on Th2 type cells.

In summary, the blockade of the ₄ integrins by a specific mAb abolished the transferred CD4⁺ cell-driven LR in the BN rat, and reduced the IL-5 mRNA⁺ cells and eosinophils (MBP⁺) in BAL. We speculate that the anti-₄-integrin treatment acts on CD4⁺ T-cell expression of IL-5 so as to contribute to the observed reductions in LR and BAL eosinophilia. A reduction of IFN- expression in BAL of allergen-challenged animals is consistent with Th2 lymphocyte activation, and reciprocal inhibition of Th1 cytokines. TA-2 mAb does not further reduce IFN- expression, suggesting a selective effect on Th2 type CD4⁺ T cells. Our data suggest that the ₄- integrin adhesion receptors may be involved in the pathobiology of asthma through effects on T-cell functions such as priming and activation within the lung parenchyma.