Asthma Airway Inflammation in a Mouse Model of Contributes to ...

1 downloads 0 Views 217KB Size Report
Allergic asthma is an inflammatory disease of the airways characterized by eosinophilic inflammation and airway hyper-reactivity. Cytokines and chemokines ...
The Journal of Immunology

IFN-␥-Inducible Protein 10 (CXCL10) Contributes to Airway Hyperreactivity and Airway Inflammation in a Mouse Model of Asthma1 Benjamin D. Medoff,2*† Alain Sauty,2§ Andrew M. Tager,*† James A. Maclean,* R. Neal Smith,‡ Anuja Mathew,* Jennifer H. Dufour,* and Andrew D. Luster3* Allergic asthma is an inflammatory disease of the airways characterized by eosinophilic inflammation and airway hyper-reactivity. Cytokines and chemokines specific for Th2-type inflammation predominate in asthma and in animal models of this disease. The role of Th1-type inflammatory mediators in asthma remains controversial. IFN-␥-inducible protein 10 (IP-10; CXCL10) is an IFN-␥-inducible chemokine that preferentially attracts activated Th1 lymphocytes. IP-10 is up-regulated in the airways of asthmatics, but its function in asthma is unclear. To investigate the role of IP-10 in allergic airway disease, we examined the expression of IP-10 in a murine model of asthma and the effects of overexpression and deletion of IP-10 in this model using IP-10-transgenic and IP-10-deficient mice. Our experiments demonstrate that IP-10 is up-regulated in the lung after allergen challenge. Mice that overexpress IP-10 in the lung exhibited significantly increased airway hyperreactivity, eosinophilia, IL-4 levels, and CD8ⴙ lymphocyte recruitment compared with wild-type controls. In addition, there was an increase in the percentage of IL-4-secreting T lymphocytes in the lungs of IP-10-transgenic mice. In contrast, mice deficient in IP-10 demonstrated the opposite results compared with wild-type controls, with a significant reduction in these measures of Th2-type allergic airway inflammation. Our results demonstrate that IP-10, a Th1-type chemokine, is up-regulated in allergic pulmonary inflammation and that this contributes to the airway hyperreactivity and Th2-type inflammation seen in this model of asthma. The Journal of Immunology, 2002, 168: 5278 –5286.

A

llergic asthma is an inflammatory disease of the airways characterized by eosinophilic infiltration of the lung and airway hyperreactivity (AHR).4 The inflammatory response in asthma has been characterized as a Th2 lymphocytedriven process based on the cytokine expression profiles of T lymphocytes recruited into the lung (1). However, the inflammatory process in asthma is exceedingly complex, involving multiple cell types and numerous mediators, including those characteristic of Th1-type inflammation (2). The functional role of Th1-type responses in asthma is controversial (2– 4). Some studies have shown that augmenting the levels of Th1 lymphocytes or cytokines can down-regulate the Th2 inflammatory response (5–9), whereas other studies have demonstrated that *Center for Immunology and Inflammatory Diseases, Division of Rheumatology, Allergy, and Immunology, †Pulmonary and Critical Care Unit, and ‡Department of Pathology, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114; and §Division of Allergology and Clinical Immunology, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland Received for publication November 6, 2001. Accepted for publication March 6, 2002. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by a GlaxoWellcome Pulmonary Fellowship award (to B.D.M.) and National Institutes of Health Grants F32AI50399 (to B.D.M.), KO8HL04087 (to A.M.T.), F32CA88721 (to J.H.D.), F32HL10375 (to A.M.), and RO1CA69212/RO1AI40618 (to A.D.L.). 2

B.D.M. and A.S. contributed equally to this work.

3

Address correspondence and reprint requests to Dr. Andrew D. Luster, Center for Immunology and Inflammatory Diseases, Division of Rheumatology, Allergy, and Immunology, Massachusetts General Hospital, Building 149-8301, 13th Street, Charlestown, MA 02129. E-mail address: [email protected] 4 Abbreviations used in this paper: AHR, airway hyperreactivity; BAL, bronchoalveolar lavage; BK5, bovine keratin 5; IP-10 (CXCL10), IFN-␥-inducible protein 10; I-TAC, IFN-inducible T cell ␣-chemoattractant; Mig (CXCL9), monokine induced by IFN-␥; Penh (CXCL11), enhanced pause.

Copyright © 2002 by The American Association of Immunologists

these interventions can increase inflammation (10 –13). Few studies have examined the role of Th1-type chemokines in the inflammatory process in asthma. A number of chemokines have been shown to be up-regulated in allergic asthma, and they are believed to be important in the pathogenesis of this disease (2). Studies in murine models of allergic airways disease have shown that chemokines are essential for the initiation and progression of airways inflammation (14). Chemokines expressed in these models are predominantly those that attract eosinophils and Th2-type lymphocytes (14). Th2 lymphocytes are preferentially recruited into the lung in murine models of asthma and are crucial for the development of airway inflammation (4). The role of chemokines preferentially active on Th1 lymphocytes has not been thoroughly examined in these models. IFN-␥inducible protein 10 (IP-10; CXCL10) is a chemokine that preferentially attracts Th1 lymphocytes through its receptor CXCR3, which is expressed at high levels on these cells (15–18). IP-10 is induced in a variety of cells in response to the Th1 cytokine IFN-␥ (19, 20). IP-10 expression is most often associated with Th1-type inflammatory diseases, where it is thought to play an important role in the recruitment of Th1 lymphocytes into tissues. However, IP-10 expression has also been shown to be increased in the airways of asthmatics (21), but its role in Th2-type inflammatory diseases such as asthma remains unknown. In our study we used a murine model of allergic airway inflammation to evaluate the role of IP-10 in a Th2-type inflammatory response characteristic of asthma. The experiments performed demonstrate that IP-10 is up-regulated during the inflammatory response seen in this model. We used genetically modified mice that overexpress IP-10 in lung epithelial cells or mice that are deficient in IP-10 expression to show that IP-10 contributes to the inflammatory response in the airways as well as to AHR. 0022-1767/02/$02.00

The Journal of Immunology

Materials and Methods Mice Wild-type FVB mice used in these experiments were purchased from Charles River Breeding Laboratories (Wilmington, MA). IFN-␥⫺/⫺ mice (on the BALB/c background) and BALB/c control mice were purchased from The Jackson Laboratory (Bar Harbor, ME). IP-10-transgenic mice (in FVB background) and IP-10⫺/⫺ mice (in Sv129 background) were generated in our laboratory (22–24). Sv129 wild-type mice were used from our own colony of mice and were the same Sv129 strain used to generate the IP-10⫺/⫺ mice. Mice were used at 6 – 8 wk of age and were housed in a pathogen-free animal facility. All animals were fed sterile food and autoclaved water ad libitum. For each experiment, age- and sex-matched groups of mice were used.

Immunization and challenge protocol Mice were injected i.p. with 10 ␮g OVA (Sigma-Aldrich, St. Louis, MO) and 1 mg aluminum hydroxide suspended in 0.5 ml normal saline on days 0 and 7. Sham-immunized mice received aluminum hydroxide alone. Mice underwent aerosol challenge with OVA (50 mg/ml in normal saline) or normal saline alone on days 14, 14 –17, or 14 –18. OVA challenge was performed by placing mice in a Plexiglas box (22 ⫻ 23 ⫻ 14 cm) and aerosolizing OVA using a nebulizer (DeVilbiss, Somerset, PA), driven by compressed air for 20 min. Mice were sacrificed 18 –24 h after the last aerosol challenge, except in the kinetics studies where they were sacrificed at the indicated times postchallenge.

Airway responsiveness AHR was measured noninvasively using a whole-body plethysmograph (Buxco, Sharon, CT). AHR was expressed as the enhanced pause (Penh), a calculated number based on inspiratory and expiratory times and pressures. Penh has been shown to correlate with measurements of airway resistance (25). The average Penh over 3 min was determined after 2-min exposure to aerosolized normal saline as a baseline. The average Penh over 3 min was then determined after exposing the mice for 2 min to aerosolized methacholine (Sigma-Aldrich) at increasing concentrations (3.125, 6.25, 12.5, and 25 mg/ml in normal saline) and was expressed as the percentage of change from the baseline.

Bronchoalveolar lavage Bronchoalveolar lavage (BAL) was performed at 18 –24 h after the last aerosol challenge. Mice were anesthetized with chloral hydrate (400 ␮g/g). The trachea was exposed and cannulated with polyethylene tubing. The lungs were lavaged with six 0.5-ml aliquots of PBS (Mediatech, Herndon, VA) containing 0.6 mM EDTA. Lavage fluid recovered from the first 1 ml of instilled PBS/EDTA was collected separately from the rest of the BAL. Both BAL fractions were centrifuged at 540 ⫻ g at 4°C, and the pelleted cells from both fractions were pooled for analysis. The supernatant of the BAL recovered from the first 1 ml instilled was kept frozen at ⫺80°C for subsequent analysis. The cells from both BAL fractions were exposed for 30 s to Tris (0.014 M)/NH4Cl (0.14 M) to lyse RBC, and the remaining live cells, as determined by trypan blue exclusion, were washed in PBS and enumerated in a hemocytometer. Cell differential counts were determined by enumerating macrophages, neutrophils, eosinophils, and lymphocytes on cytocentrifuge preparations of the cells stained with a combination of Wright stain (EM Sciences, Gibbstown, NJ) and Diff-Quick (Dade Behring, Newark, DE).

Flow cytometry Cells recovered from the BAL were resuspended in PBS with 1% BSA (Intergen, Purchase, NY) and incubated for 30 min with 25 ␮g/ml 2.4G2 anti-Fc␥RIII/II (CD16/CD32) at 4°C. Samples of ⬃106 cells were then stained with FITC-conjugated anti-murine CD3 mAb and PE-conjugated anti-murine CD19 mAb in one reaction and with FITC-conjugated antimurine CD8 mAb, PE-conjugated anti-murine CD4 mAb, and allophycocyanin-conjugated anti-murine CD25 mAb in a separate reaction. All Abs were obtained from BD PharMingen (San Diego, CA). Cells were washed with PBS and then fixed in 2% paraformaldehyde. Flow cytometry was performed after gating on the lymphocyte population using a FACSCalibur analytical flow cytometer (BD Biosciences, Mountain View, CA) and were analyzed using CellQuest software (BD Biosciences).

Limulus amebocyte lysate assay for LPS Five percent OVA in normal saline was serially diluted in water. The samples were added to the tubes provided in the Limulus assay kit (Charles

5279 River Endosafe, Charleston, SC). The solution was mixed and incubated at 37°C for 1 h and then assessed for coagulation.

Cytokine assays Levels of IL-4, IL-5, and IFN-␥ in BAL were measured by ELISA (Endogen, Woburn, MA).

IP-10 ELISA The level of IP-10 in the serum of IP-10-transgenic and control animals was determined by ELISA. Serial dilutions of a known quantity of IP-10 (standard), serum samples, and serum samples with added IP-10 (to a final concentration of 10 ng/ml) were coated onto 96-well plates overnight. After washing and blocking the plate with 1% BSA, 0.3 ␮g/ml biotin-labeled Ab to murine IP-10 (Cedarlane Laboratories, Hornby, Canada) was added to each sample. After a 2-h incubation, avidin (Sigma-Aldrich) was added to the samples for 30 min. Finally, ABTS (Sigma-Aldrich) was added, and the samples were incubated for another 20 min. The absorbance was measured at 405 nm, and the concentration of IP-10 in the samples was determined by comparison to the standards. As a positive control for this assay, we were able to detect 10 ng/ml recombinant IP-10 added to serum samples.

Histopathologic examination Lungs were flushed free of blood by slowly injecting 10 ml PBS into the right ventricle before excision for all studies. The left lung was inflated with 10% buffered formalin to 25 cm H2O pressure and transferred into vials containing 10% buffered formalin. Multiple paraffin-embedded, 5-␮m sections of the entire mouse lung were prepared and stained with H&E and periodic acid-Schiff. The slides were evaluated by light microscopy, and the amounts of inflammation and mucous were assessed by a pathologist blinded to the genotype of the animals.

Whole-lung digests and flow cytometry on lung lymphocytes The right lung was excised and minced into small pieces with a scissors. The pieces were digested for 1 h in a lysis solution containing PBS, 10% FCS, 150 U/ml collagenase III (Worthington Biochemical, Lakewood, NJ), and 850 U/ml hyaluronidase (Sigma-Aldrich). The digested lungs were then extruded through a mesh strainer, and the collected cells were washed once with PBS. Live cells were enumerated by a hemocytometer, as determined by trypan blue exclusion. Cells were then resuspended in RPMI with 10% FCS to a concentration of 1 ⫻ 106 cells/ml. Five 1-ml aliquots per lung were put into 24-well plates and incubated for 1 h at 37°C after adding PMA (50 ng/ml; Sigma-Aldrich) and ionomycin (250 ng/ml; Sigma-Aldrich). Monensin (1 ␮g; Golgi stop; BD PharMingen) was added to three of the five wells per lung (Th2 samples), and 1 ␮g brefeldin A (Golgi plug; BD PharMingen) was added to the other two wells (Th1 samples). The cells were incubated for 4 h at 37°C. The cells were then placed at 4°C overnight. The cells were pelleted and washed with PBS/1% BSA. The cells were incubated for 30 min with 2.4G2 anti-Fc␥RIII/II at 4°C. The Th2 samples were stained with FITC-conjugated anti-murine CD3 mAb, and the Th1 samples were stained with PE-conjugated CD4 mAb for 20 min at 4°C. The samples were washed with PBS, and 100 ␮l reagent A (Fix and Perm kit; Caltag Laboratories, Burlingame, CA) was added to each sample. The cells were incubated for 20 min at room temperature and were washed twice with PBS. Reagent B (100 ␮l; Fix and Perm kit; Caltag Laboratories) was added to each sample. The Th2 samples were then stained with PEconjugated IgG1, PE-conjugated anti-IL-4 mAb, or PE-conjugated antiIL-5 mAb. The Th1 samples were stained with either FITC-conjugated IgG1 or FITC-conjugated anti-IFN-␥ mAb. All Abs were obtained from BD PharMingen. Samples were incubated for 20 min at 4°C and washed twice with PBS/1% BSA. The samples were resuspended in 200 ␮l PBS/1% BSA. Flow cytometry was performed after gating on the lymphocyte population using a FACSCalibur analytical flow cytometer and analyzed using CellQuest software.

Serum IgE and OVA-specific IgE determination Blood was obtained by right ventricular puncture. The blood was allowed to clot, and serum was removed after centrifuging the sample at the highest speed in a microcentrifuge for 5 min. Total IgE levels were measured by ELISA (BD PharMingen). OVA-specific IgE levels were measured by plating 100 ␮l 10 ␮g/ml OVA onto 96-well plates. Samples were incubated on the plates for 2 h, and bound IgE Abs were determined using an ELISA.

5280

IP-10 CONTRIBUTES TO ALLERGIC AIRWAY INFLAMMATION

Northern blot analysis Total cellular RNA was isolated from the lungs by homogenizing the tissue with a Polytron (Brinkmann Instruments, West Orange, NY) in 4 M guanidine hydrochloride and pelleting the RNA through a 5.7 M CsCl2 cushion. Northern blotting was performed by fractionating 10 ␮g total RNA/ lane on a 1.2% agarose gel containing 0.7% formaldehyde, transferring the RNA to GeneScreen (DuPont, Wilmington, DE), and then hybridizing the membrane with [32P]dCTP Klenow-labeled random-primed mouse cDNA probes. The following fragments were used as probes: a cDNA from the murine IP-10 gene (26), a cDNA from the murine monokine induced by IFN-␥ (Mig; CXCL9) gene (provided by J. Farber, National Institutes of Health, Bethesda, MD), a cDNA from the murine IFN-inducible T cell ␣-chemoattractant (I-TAC; CXCL11) gene (provided by G. Werner-Felmayer, University of Innsbruck, Innsbruck, Austria), a cDNA from the murine eotaxin gene (27), and a 2-kb ClaI-KpmI fragment of the bovine keratin 5 (BK5) gene that contains the 3⬘ intron, splice site, and polyadenylation signal (BK5-I/pa) (22). ␤-Actin cDNA was used as a control for RNA loading. Signal quantitation was determined using a phosphor imager (Molecular Imager System; Bio-Rad, Hercules, CA).

Western blot analysis Total cellular protein was isolated from the lungs by homogenizing the tissue with a Polytron in lysis buffer consisting of 50 mM Tris (pH 7.5), 1 mM EDTA, 1 mM DTT, 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholic acid, 0.1% SDS, and 1% Triton X-100 supplemented with antiproteases (0.1 mg/ml pepstatin A, 0.03 mM leupeptin, 145 mM benzamidine, 0.37 mg/ml aprotinin, and 1 mM PMSF). Cell lysates were incubated on ice for 30 min, and residual tissue was removed by two centrifugations at 16,000 ⫻ g at 4°C for 15 min. Total protein concentrations of the samples were determined by bicinchoninic acid assay (Endogen). Two hundred micrograms of total protein was taken, and the final volume was adjusted to 200 ␮l with 10 mM Tris. The samples were pelleted with a 1-h centrifugation at 55,000 rpm at 4°C. The pellets were resuspended in PBS and sample buffer and then boiled for 5 min. The samples were fractionated on a 12.5% Tris-Tricene gel. Proteins were transferred onto a polyvinylidene difluoride membrane (NEN, Boston, MA) and blocked in PBS containing 0.1% Tween 20 and 5% nonfat dry milk. IP-10 was identified by sequentially incubating the membrane at room temperature with a 1/2,000 dilution of an affinity-purified polyclonal rabbit anti-murine IP-10 Ab (28), followed by a 1/3,000 dilution of an HRP-conjugated goat anti-rabbit Ig Ab (Bio-Rad). The membrane was washed repeatedly in PBS containing 0.1% Tween 20 between incubations and was developed using an ECL kit (Amersham Pharmacia Biotech, Piscataway, NJ).

Data analysis Data are expressed as the mean ⫾ SEM, unless otherwise indicated. Results were interpreted using two-tailed Student’s t test or two-way ANOVA. High and low data values were discarded from each group in the BAL cell counts and flow cytometric analysis from IP-10⫺/⫺ mice and their wild-type controls to eliminate the effects of extreme outliers. Differences were considered statistically significant at p ⬍ 0.05.

Results IP-10, Mig, I-TAC, and IFN-␥ are induced in the lung following OVA sensitization and challenge We determined the expression of IP-10, Mig, I-TAC, IFN-␥, and TNF-␣ mRNA in the lung at baseline and 3, 6, 12, 24, and 48 h after a single OVA challenge in BALB/c mice (Fig. 1A). IP-10 expression was induced at 3, 6, and 12 h after OVA challenge in both mice immunized with OVA and sham-immunized mice. This induction was greater and more sustained in OVA-immunized and -challenged mice. Mig, I-TAC, and IFN-␥ were induced to a greater extent in OVA-immunized and -challenged mice, but to a lesser degree than IP-10 (almost 10-fold less expression). TNF-␣ was induced to the same extent after OVA challenge in both shamand OVA-immunized animals. The induction of TNF-␣, IP-10, Mig, I-TAC, and IFN-␥ in sham-immunized, OVA-challenged animals is believed to be secondary to LPS contamination of the OVA solution. A representative Northern blot of lung RNA harvested 24 h after the last of four OVA challenges demonstrates

FIGURE 1. Expression of IP-10, Mig, I-TAC, IFN-␥, and TNF-␣ mRNA following a single OVA challenge in OVA-immunized or shamimmunized mice. A, Quantitative phosphor imager analysis of Northern blots of 10 ␮g total RNA isolated from the lungs of BALB/c mice at the indicated times following a single OVA challenge in OVA-immunized (OVA/OVA) or sham-immunized (SI/OVA) mice. Data are presented as the mean level of signal intensity relative to expression at baseline (n ⫽ 2 mice/group for SI/OVA and n ⫽ 4 mice/group for OVA/OVA at each time point). B, Representative Northern blot analysis of IP-10, Mig, and I-TAC expression 24 h following four OVA challenges in OVA-immunized (OVA/OVA) or sham-immunized (SI/OVA) mice. The blot was sequentially hybridized with IP-10, Mig, I-TAC, and ␤-actin probes and exposed to autoradiograms for the following times: 5 days, 11 days, 21 days, and 5 h, respectively. Each lane represents RNA from a single mouse lung. Note the difference in the y-axis value on the IP-10, Mig, and I-TAC graphs, representing the dramatic increase in IP-10 relative to Mig and I-TAC.

increased expression of IP-10, Mig, and I-TAC in mice immunized with OVA compared with sham-immunized animals (Fig. 1B). IP-10 expression is partially independent of IFN-␥ Northern blot analysis of lung mRNA from wild-type and IFN␥⫺/⫺ mice after four OVA challenges demonstrates attenuated upregulation of IP-10 expression in OVA-immunized IFN-␥⫺/⫺ compared with wild-type animals (Fig. 2). Mig expression was undetectable in IFN-␥⫺/⫺ mice compared with wild-type mice 24 h after OVA challenge. IP-10, Mig, and eotaxin expression in IP-10-transgenic mice Transgenic mice were engineered by our laboratory to overexpress IP-10 in epithelial cells, including cells in the lung, using the BK5 promoter (22). IP-10 protein was not detected in the serum of transgenic and wild-type mice using an ELISA that was sensitive to 3 ng/ml. However, Western blot analysis revealed the presence of IP-10 in the serum at ⬃1 ng/ml in the transgenic mice. Northern

The Journal of Immunology

5281

FIGURE 2. Expression of IP-10 and Mig following OVA challenges in OVA-immunized or sham-immunized BALB/c wild-type and IFN-␥⫺/⫺ mice. Northern blot analysis of 10 ␮g total RNA isolated from the lungs of BALB/c wild-type (WT) or IFN-␥⫺/⫺ mice, either OVA-immunized and OVAchallenged (OVA/OVA) or sham-immunized and OVA-challenged (SI/OVA). Lungs were harvested 18 –24 h after the fourth OVA challenge. The Northern blot was sequentially hybridized with IP-10, Mig, and ␤-actin cDNAs and exposed to autoradiographs for the following times: 7 days, 12 days, and 5 h, respectively. Each lane represents RNA from a single mouse lung.

blot analysis was performed on lung mRNA from wild-type and IP-10-transgenic mice after four OVA challenges (Fig. 3A). IP-10 expression was up-regulated in the OVA-immunized wild-type mice as well as the sham- and OVA-immunized IP-10-transgenic mice. All IP-10-transgenic mice had greater overall expression of IP-10 than wild-type mice. In contrast, Mig and eotaxin expression was up-regulated only in OVA-immunized mice. Expression of the IP-10 transgene, which contains a portion of the BK5 gene, was detected with a BK5-specific probe only in IP-10-transgenic mice. Western blot analysis was performed on lung lysates isolated from wild-type and IP-10-transgenic mice after four OVA chal-

lenges (Fig. 3B). IP-10 expression was seen in wild-type OVAimmunized mice and in OVA-immunized IP-10-transgenic mice, as well as IP-10-transgenic mice that were not exposed to OVA (naive). IP-10 protein expression was greater in IP-10-transgenic mice than in wild-type mice. AHR is enhanced by IP-10 expression AHR is a prominent component of this murine model of asthma. OVA-immunized and -challenged IP-10-transgenic mice had increased AHR compared with OVA-immunized and -challenged wild-type mice, as measured by Penh, a noninvasive correlate of airway resistance (Fig. 4A). All OVA-immunized and -challenged mice were more reactive than OVA-immunized and saline-challenged mice. Two-way ANOVA demonstrated a significant positive interaction between genetic strain (IP-10 transgenic vs wild type) and dose of methacholine. Thus, the rate of change in Penh with increasing dose of methacholine was significantly greater for the IP-10-transgenic mice than for wild-type animals. This implies significantly greater AHR in the OVA-immunized and -challenged IP-10-transgenic animals. To further explore the role of IP-10 in AHR, we used our recently generated IP-10⫺/⫺ mice (23, 24). Measurement of Penh in IP-10⫺/⫺ mice showed no statistically significant differences in AHR compared with sex- and age-matched wild-type control mice (Fig. 4B) as determined by two-way ANOVA. IP-10 expression modulates BAL total cell counts and eosinophils

FIGURE 3. Expression of IP-10, Mig, and eotaxin in OVA-immunized and OVA-challenged or sham-immunized and OVA-challenged wild-type and IP-10-transgenic mice. A, Northern blot analysis of 10 ␮g total RNA isolated from the lungs of wild-type (WT) or IP-10-transgenic (IP-10 TG) mice. Mice were either OVA-immunized and OVA-challenged (OVA/ OVA) or sham-immunized and OVA-challenged (SI/OVA). Lungs were harvested 18 –24 h after the fourth OVA challenge. The Northern blot was sequentially hybridized with IP-10, Mig, and eotaxin cDNA probes, and probes specific for the BK5-IP-10 transgene (BK5-I/pA), and ␤-actin. The blots were exposed to autoradiographs for the following times: 3 days, 8 days, 2 days, 4 days, and 5 h, respectively. Each lane represents RNA from a single mouse lung. B, Western blot analysis of 200 ␮g total protein isolated from the lungs of wild-type or IP-10-transgenic mice. Each lane represents protein from a single mouse lung. Lungs were harvested 18 –24 h after the fourth OVA challenge. Five, 10, and 50 pg recombinant murine IP-10 (rmIP-10) were used as a positive control.

Following OVA immunization and four OVA challenges, total cell counts in the BAL were significantly increased in IP-10-transgenic mice compared with wild-type mice. Differential counts of the BAL cells demonstrated significantly increased eosinophils and macrophages in the IP-10-transgenic mice (Fig. 5A) compared with wild-type controls. Neutrophil counts were not significantly different (data not shown). BAL cellular analysis in IP-10⫺/⫺ mice demonstrated the opposite findings, with significantly fewer total cells and eosinophils compared with wild-type mice (Fig. 5B). Again, there was no difference in neutrophil levels (data not shown). IP-10 expression modulates lymphocyte trafficking into the airways Analysis of lymphocyte subsets by flow cytometry demonstrated a significant increase in the number of CD8⫹ lymphocytes in BAL in IP-10-transgenic mice compared with wild-type mice. In addition, there was a nonsignificant trend toward increased CD3⫹ lymphocytes and CD4⫹ lymphocytes in BAL of IP-10-transgenic mice compared with wild-type mice (Fig. 6A). Similar analysis in

5282

IP-10 CONTRIBUTES TO ALLERGIC AIRWAY INFLAMMATION

FIGURE 4. AHR in wild-type, IP-10-transgenic, or IP-10⫺/⫺ mice immunized and challenged with OVA. A, AHR in normal saline-challenged wild-type and IP-10-transgenic mice (control; pooled from three experiments, n ⫽ 6 mice each), OVA-challenged wild-type mice (pooled from two experiments, n ⫽ 10 mice), and OVA-challenged IP-10-transgenic mice (pooled from two experiments, n ⫽ 11 mice). AHR was measured after the fourth OVA challenge. AHR is expressed as the percentage of change in the average Penh over baseline in the 3 min following a 2-min exposure to increasing concentrations of aerosolized methacholine (MC) in normal saline. ⴱ, Two-way ANOVA demonstrates a significant interaction between the genetic strain of the mice and the dose of MC given (p ⫽ 0.0275). B, AHR in normal saline-challenged wild-type and IP-10⫺/⫺ mice (control; pooled from three experiments, n ⫽ 6 mice), OVA-challenged wild-type mice (pooled from three experiments, n ⫽ 15 mice), and OVA-challenged IP-10⫺/⫺ mice (pooled from three experiments, n ⫽ 15 mice). AHR was measured after the fourth OVA challenge. AHR is expressed as the percentage of change in the average Penh over baseline in the 3 min following a 2-min exposure to increasing concentrations of aerosolized methacholine (MC) in normal saline. Two-way ANOVA demonstrated no significant difference between the groups.

IP-10⫺/⫺ mice revealed the opposite finding, with significantly reduced T cells, B cells (CD19⫹ lymphocytes), CD4⫹ lymphocytes, and CD8⫹ lymphocytes compared with wild-type controls (Fig. 6B). IP-10 expression modulates cytokine production in the airways Analysis of cytokine levels in BAL fluid was performed by ELISA following OVA immunization and five OVA challenges. IP-10transgenic animals had significantly increased IL-4 levels and a nonsignificant increase in IL-5 compared with wild-type mice (Fig. 7A). In contrast, IP-10⫺/⫺ mice had significantly decreased IL-4 and IL-5 levels compared with wild-type controls (Fig. 7B). IFN-␥ levels were not different in wild-type and IP-10-transgenic or IP10⫺/⫺ mice. The differences in the levels of IL-4/IL-5 and IFN-␥ in the two wild-type groups probably relate to the different strains used for the experiments (FVB strain for the IP-10-transgenics and Sv129 for the IP-10⫺/⫺ mice). Serum IgE levels are unaffected by IP-10 expression Serum levels of total IgE and OVA-specific IgE were determined in wild-type mice and IP-10-transgenic mice after five OVA challenges, and no difference was seen (data not shown). Serum total IgE and OVA-specific IgE were also not different in wild-type and IP-10⫺/⫺ mice after five OVA challenges (data not shown).

Qualitative analysis of inflammation in the lung by histopathology is unaffected by changes in IP-10 expression The histopathology of lungs from IP-10-transgenic mice, IP-10⫺/⫺ mice, and their controls after OVA immunization and five OVA challenges was reviewed by a pathologist blinded to the origin of the tissue. The pathologist assessed the tissue for peribronchial inflammation and mucous gland hypertrophy. There were no apparent differences in the degree of peribronchial inflammation or in the number of mucous-secreting cells (data not shown).

IP-10 expression modulates lymphocyte trafficking into the lung To better quantitate the effect of IP-10 expression on type 2 and type 1 lymphocyte recruitment into the lung, we isolated lymphocytes from the lungs of OVA-immunized and OVA-challenged animals and analyzed them for surface marker expression as well as intracellular cytokine expression by flow cytometry. IP-10transgenic mice had a greater percentage of IL-4- and IL-5-positive T cells (CD3⫹) compared with wild-type mice (Fig. 8A). Only the percentage of IL-4-positive T cells was significantly different. In contrast, IP-10⫺/⫺ mice had a significantly lower percentage of IL-4-positive T cells compared with wild-type animals (Fig. 8B). The percentage of IFN-␥-positive T cells recruited into the lung

The Journal of Immunology

5283

FIGURE 5. BAL cell counts and differentials following OVA immunization and challenge in wild-type, IP-10-transgenic, and IP-10⫺/⫺ mice. A, BAL total cell counts and numbers of macrophages and eosinophils are shown for wild-type and IP-10-transgenic animals after OVA immunization and daily OVA challenges for 4 days (one experiment, n ⫽ 6 mice/group). ⴱ, p ⬍ 0.05 for wild-type vs IP-10-transgenic mice. B, BAL total cell counts and numbers of macrophages, neutrophils, lymphocytes, and eosinophils were determined in wild-type and IP-10⫺/⫺ animals after OVA immunization and daily OVA challenges for 5 days (pooled from two experiments, n ⫽ 8 mice/group). ⴱ, p ⬍ 0.05 for wild-type vs IP-10⫺/⫺ mice.

was not statistically different between the genetically modified mice and their wild-type controls.

Discussion The role of Th1-type responses in asthma is of great interest, as there are several proposed therapies or preventive measures that attempt to reduce allergic airway inflammation by enhancing Th1 inflammatory responses (2). There is evidence that Th1-type inflammation is up-regulated in human asthma (29) and in the OVAinduced model of asthma (12), and there are conflicting data sug-

gesting that Th1 inflammation can both augment (10 –13) and attenuate (5–9) allergic inflammation. In this study we have demonstrated that the Th1-type chemokine IP-10 is induced in the lung in the OVA model of allergic airway disease. Using genetically modified mice we have shown that IP-10 overexpression in the lung augments AHR as well as several measures of Th2-type inflammation in the airways, namely, eosinophil recruitment, type 2 lymphocyte (IL-4⫹CD3⫹) recruitment, and IL-4 levels. In addition, we have found a significant increase in the number of CD8⫹ lymphocytes in the airways. Similar experiments in IP-10⫺/⫺ mice

FIGURE 6. BAL lymphocyte subset counts following OVA immunization and challenge in wild-type, IP-10-transgenic, and IP-10⫺/⫺ mice. A, Total numbers of BAL CD3⫹, CD19⫹, CD4⫹, CD8⫹, and CD4⫹CD25⫹ lymphocytes were determined in wild-type and IP-10-transgenic animals after OVA immunization and daily OVA challenges for 5 days (pooled from three experiments, n ⫽ 17 mice/group). ⴱ, p ⬍ 0.05 for wild-type vs IP-10-transgenic mice. B, Total numbers of BAL CD3⫹, CD19⫹, CD4⫹, CD8⫹, and CD4⫹CD25⫹ lymphocytes were determined in wild-type and IP-10⫺/⫺ animals after OVA immunization and daily OVA challenges for 5 days (pooled from two experiments, n ⫽ 12 mice/group). ⴱ, p ⬍ 0.05 for wild type vs IP-10⫺/⫺ mice.

5284

IP-10 CONTRIBUTES TO ALLERGIC AIRWAY INFLAMMATION

FIGURE 7. BAL cytokine levels determined by ELISA following OVA immunization and challenge in wild-type, IP-10-transgenic, and IP-10⫺/⫺ mice. A, Concentrations of IL-4, IL-5, and IFN-␥ determined by ELISA on BALs from wild-type and IP-10-transgenic mice. Data are the average value from two pooled experiments (n ⫽ 10 animals/group). ⴱ, p ⫽ 0.01 for IL-4 concentration in wild-type vs IP-10-transgenic mice. B, Concentrations of IL-4, IL-5, and IFN-␥ determined by ELISA on BALs from wild-type and IP-10⫺/⫺ mice. Data are the average value from two pooled experiments (n ⫽ 10 animals/group in the IL-5 and IFN-␥ reactions and n ⫽ 8 animals/ group in the IL-4 reactions). ⴱ, p ⬍ 0.05 for IL-4 and IL-5 concentrations in wild-type vs IP-10⫺/⫺ mice.

confirm a pathogenic role for IP-10 in this model by demonstrating reduced airway eosinophilia, type 2 lymphocyte recruitment, and IL-4 levels in the absence of IP-10. Changes in the level of IP-10 expressed in this model of allergic disease did not affect IgE Ab production or measures of Th1-type inflammation (IFN-␥ expression or Th1 lymphocyte recruitment). These data suggest that IP-10 may directly influence Th2-type inflammatory responses. Few studies to date have examined the effect of alterations in the expression of Th1-type chemokines, such as IP-10, in animal mod-

els of asthma. IP-10 is a chemokine that has been associated with Th1-type responses based on its induction by IFN-␥ expression (30) and the ability of IP-10 to preferentially attract Th1 lymphocytes (15–18). In this work we show that IP-10 expression is enhanced in a murine model of asthma. This is consistent with data demonstrating elevated levels of IP-10 protein in the BAL fluid of asthma patients compared with healthy controls (21). We found that the increase in IP-10 expression in the OVA model of asthma is only partially dependent on IFN-␥ expression, suggesting that

FIGURE 8. Intracellular cytokine staining in lymphocytes isolated from the lungs of wild-type, IP-10-transgenic, and IP-10⫺/⫺ mice following OVA immunization and challenge. A, Percentage of CD3⫹ lymphocytes staining positively for IL-4, IL-5, or IFN-␥. Lymphocytes were isolated from whole-lung digests of wild-type or IP-10-transgenic mice. Data are the average value from one experiment (n ⫽ 5 animals/group). ⴱ, p ⫽ 0.05 for IL-4 in wild-type vs IP-10-transgenic mice. B, Percentage of CD3⫹ lymphocytes staining positively for IL-4, IL-5, or IFN-␥. Lymphocytes were isolated from whole-lung digests of wild-type or IP-10⫺/⫺ mice. Data are the average value from one experiment (n ⫽ 5 animals/group). ⴱ, p ⫽ 0.008 for IL-4 in wild-type vs IP-10⫺/⫺ mice.

The Journal of Immunology additional pathways contribute to IP-10 up-regulation in this model. In contrast, the expression of another CXCR3 chemokine ligand, Mig, was undetectable in the absence of IFN-␥. The role of IP-10 in allergic inflammation has not been well defined. In vitro studies suggest that IP-10 could antagonize allergic inflammation by blocking the chemokine receptor CCR3 and inhibiting eosinophil and Th2 lymphocyte recruitment (31). The data we present here suggest that this effect is not significant in vivo, at least in the robust model of Th2-type inflammation used in our study. Another recent study has examined the consequences of overexpression of IP-10 in a murine model of asthma using an adenovirally mediated gene delivery system (32). This study demonstrated that increased IP-10 expression reduced airway eosinophilia and IL-4 levels and increased IFN-␥ levels. Our results differ from these findings in that we show that IP-10 directly enhances Th2-type inflammation. The divergent results seen between these two studies may be related to the different models used. The previous study used an adenovirus to deliver the GM-CSF gene to the airways during OVA challenge. This technique creates a Th2-type inflammatory response in the lung without prior OVA immunization. They then coexpressed IP-10 with GM-CSF in the airways, again using adenoviral vectors, and this reduced allergic airway inflammation compared with a control viral infection. The model in this study is clearly different from the model used in our study. The response to IP-10 in the airways may be modified by GM-CSF expression and the lack of an immunization step before OVA challenge. In addition, coexistent viral infection in the setting of allergen challenge may affect the response to IP-10. The data presented in our paper suggest a pathogenic role for IP-10 in a model of allergic airways disease. There are several potential mechanisms for the effects of IP-10 on allergic inflammation. Some prior studies have suggested that augmentation of Th1-type inflammatory disease can increase allergic airway disease (3). However, the fact that IFN-␥ expression or Th1 lymphocyte recruitment was not altered when the OVA model of allergic airways disease was induced in IP-10-transgenic or IP-10-deficient mice suggests that IP-10 contributes to Th2-type inflammation directly rather than by increasing Th1-type inflammation. In our study we demonstrated an increase in type 2 lymphocyte recruitment into the lungs. Because we used CD3 as a surface marker to identify T lymphocytes for the intracellular cytokine staining experiments, we are unable to distinguish whether the IL-4⫹ T cells recruited into the lung are Th2 lymphocytes and/or Tc2 lymphocytes (type 2 cytotoxic cells; IL-4⫹CD8⫹ lymphocytes). Flow cytometry of lymphocytes from the BAL revealed that both CD4⫹ and CD8⫹ lymphocyte recruitment is affected by alterations in IP-10 expression, suggesting that IP-10 may be directly increasing both Th2 and Tc2 recruitment. The receptor for IP-10, CXCR3, has been reported to be expressed on Th2 lymphocytes in humans and mice, albeit at lower levels than on Th1 lymphocytes (15, 18). Other studies have suggested that CXCR3 is found on both Th1 and Th2 lymphocytes recruited into inflammatory sites in vivo (33, 34). Thus, IP-10 could enhance allergic inflammation in the OVA model of asthma by direct recruitment of Th2 lymphocytes into the lung. While increased Th2 lymphocyte recruitment could clearly increase allergic inflammation, the potential implications of alterations in the number of Tc2 lymphocytes are unclear. Some studies have suggested that CD8⫹ lymphocytes have minimal effects on OVA-induced AHR and inflammation (4, 35), while others have indicated that CD8⫹ lymphocytes are necessary for the allergic response in the airways (36) and may be a marker of more severe disease in humans (37). Taken together, the data presented above suggest that IP-10 can increase allergic inflammation by enhancing type 2 lymphocyte recruitment into the lung.

5285 We also observed that changes in the level of IP-10 could modulate eosinophil recruitment to the airways in the OVA model of allergic pulmonary inflammation used in our studies. This effect on eosinophil recruitment could be secondary to the effects of IP-10 on the modulation of type 2 lymphocyte recruitment into the lung or could be a direct effect of IP-10 on eosinophils. In this regard there is a single study that reported CXCR3 expression on eosinophils (38). In that paper IP-10 was shown to mediate eosinophil chemotaxis and activation (38). If similar findings can be demonstrated in vivo, this would suggest that IP-10 could be modulating the eosinophilic inflammation seen in the OVA model of asthma by directly affecting the recruitment of eosinophils into the airways. AHR is a primary feature of asthma and the OVA-induced model of allergic airway disease. In our study we found that transgenic mice (on an FVB background) that overexpress IP-10 in the lung had significantly increased AHR, as assessed by Penh. AHR appeared to be unchanged in IP-10⫺/⫺ mice; however, these mice were on an Sv129 background, which develops less AHR in the OVA model of allergic airway disease and thus may not demonstrate subtle differences. The change in AHR seen with increased IP-10 could occur indirectly from the effects on type 2 lymphocyte and eosinophil recruitment seen in our experiments. Alternatively, IP-10 could enhance AHR by directly modulating smooth muscle cell function. There is a single study demonstrating that IP-10 directly stimulates vascular smooth muscle cells to proliferate and migrate in vitro (39). If this activity is relevant in the OVA model of allergic pulmonary inflammation, then IP-10 could enhance AHR by directly modulating the density of muscle cells around the airway and/or their contractile properties. In our murine model of asthma we found that IP-10 modulates Th2-type airway inflammation. We hypothesize that IP-10 contributes to AHR and airway inflammation by increasing the recruitment of Th2 lymphocytes and eosinophils. The fact that the receptor for IP-10 is expressed on Th2 lymphocytes and possibly on eosinophils and smooth muscle cells provides a potential mechanism for this effect. In broader terms, our study suggests that chemokines and complex inflammatory diseases, such as asthma, cannot be completely compartmentalized into Th1-type or Th2-type inflammation. Factors from both types of inflammation may indeed be involved in establishing the allergic inflammatory state in the airways. Furthermore, attempts to modify the polarization of the inflammatory response by altering chemokine expression may prove complicated. Indeed, asthma therapies that up-regulate IP-10 (such as IFN-␥) may actually lead to enhancement of airway inflammation.

Acknowledgments We thank Scott D. Bercury, Andrew Carfone, and Josephine Leung for technical assistance, and Dr. Myles Wolf for assistance with statistical analysis.

References 1. Romagnani, S. 2000. The role of lymphocytes in allergic disease. J. Allergy Clin. Immunol. 105:399. 2. Busse, W. W., and R. F. Lemanske. 2001. Asthma. N. Engl. J. Med. 344:350. 3. Castro, M., D. D. Chaplin, M. J. Walte, and M. J. Holtzman. 2000. Could asthma be worsened by stimulating the T-helper type 1 immune response? Am. J. Respir. Cell Mol. Biol. 22:143. 4. Lee, N. A., E. W. Gelfand, and J. J. Lee. 2001. Pulmonary T cells and eosinophils: coconspirators or independent triggers of allergic respiratory pathology? J. Allergy Clin. Immunol. 107:945. 5. Gavett, S. H., O. H. DJ, X. Li, S. K. Huang, F. D. Finkelman, and M. Wills-Karp. 1995. Interleukin 12 inhibits antigen-induced airway hyperresponsiveness, inflammation, and Th2 cytokine expression in mice. J. Exp. Med. 182:1527. 6. Lack, G., K. L. Bradley, E. Hammelmann, H. Renz, J. Loader, D. Y. Leung, G. Larsen, and E. W. Gelfand. 1996. Nebulized IFN-␥ inhibits the development of secondary allergic responses in mice. J. Immunol. 157:1432. 7. Pauwels, R. A., G. J. Brusselle, and J. C. Kips. 1997. Cytokine manupulation in animal models of asthma. Am. J. Respir. Crit. Care Med. 156:S78.

5286

IP-10 CONTRIBUTES TO ALLERGIC AIRWAY INFLAMMATION

8. Cohn, L., R. J. Homer, N. Niu, and K. Bottomly. 1999. T helper 1 cells and interferon ␥ regulate allergic airway inflammation and mucus production. J. Exp. Med. 190:1309. 9. Huang, T. J., P. A. MacAry, P. Eynott, A. Moussavi, K. C. Daniel, P. W. Askenase, D. M. Kemeny, and K. F. Chung. 2001. Allergen-specific Th1 cells counteract efferent Th2 cell-dependent bronchial hyperresponsiveness and eosinophilic inflammation partly via IFN-␥. J. Immunol. 166:207. 10. Li, L., Y. Xia, A. Nguyen, L. Feng, and D. Lo. 1998. Th2-induced eotaxin expression and eosinophilia coexist with Th1 responses at the effector stage of lung inflammation. J. Immunol. 161:3128. 11. Hansen, G., G. Berry, R. H. DeKruyff, and D. T. Umetsu. 1999. Allergen-specific Th1 cells fail to counterbalance Th2 cell-induced airway hyperreactivity but cause severe airway inflammation. J. Clin. Invest. 103:175. 12. Randolph, D. A., C. J. Carruthers, S. J. Szabo, K. M. Murphy, and D. D. Chaplin. 1999. Modulation of airway inflammation by passive transfer of allergen-specific Th1 and Th2 cells in a mouse model of asthma. J. Immunol. 162:2375. 13. Randolph, D. A., R. Stephens, C. J. Carruthers, and D. D. Chaplin. 1999. Cooperation between Th1 and Th2 cells in a murine model of eosinophilic airway inflammation. J. Clin. Invest. 104:1021. 14. Lukacs, N. W., S. H. P. Oliveira, and C. M. Hogaboam. 1999. Chemokines and asthma: redundancy of function or a cooordinated effort? J. Clin. Invest. 104:995. 15. Bonecchi, R., G. Bianchi, P. P. Bordignon, D. D’Ambrosio, R. Lang, A. Borsatti, S. Sozzani, P. Allavena, P. A. Gray, A. Mantovani, et al. 1998. Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (Th1s) and Th2s. J. Exp. Med. 187:129. 16. Siveke, J. T., and A. Hamann. 1998. T helper 1 and T helper 2 cells respond differentially to chemokines. J. Immunol. 160:550. 17. Loetscher, M., B. Gerber, P. Loetscher, S. A. Jones, L. Piali, I. Clark-Lewis, M. Baggiolini, and B. Moser. 1996. Chemokine receptor specific for IP10 and mig: structure, function, and expression in activated T-lymphocytes. J. Exp. Med. 184:963. 18. Sallusto, F., D. Lenig, C. R. Mackay, and A. Lanzavecchia. 1998. Flexible programs of chemokine receptor expression on human polarized T helper 1 and 2 lymphocytes. J. Exp. Med. 187:875. 19. Luster, A. D., J. C. Unkeless, and J. V. Ravetch. 1985. ␥-Interferon transcriptionally regulates an early-response gene containing homology to platelet proteins. Nature 315:672. 20. Luster, A. D., and J. V. Ravetch. 1987. Biochemical characterization of a ␥ interferon-inducible cytokine (IP-10). J. Exp. Med. 166:1084. 21. Miotto, D., P. Christodoulopoulos, R. Olivenstein, R. Taha, L. Cameron, A. Tsicopoulos, B. Tonnel, O. Fahy, J. J. Lafitte, A. D. Luster, et al. 2001. Expression of IFN-␥-inducible protein; monocyte chemotactic proteins 1, 3, and 4; and eotaxin in Th1- and Th2-mediated lung diseases. J. Allergy Clin. Immunol. 107:664. 22. Luster, A. D., R. D. Cardiff, J. A. MacLean, K. Crowe, and R. D. Granstein. 1998. Delayed wound healing and disorganized neovascularization in transgenic mice expressing the IP-10 chemokine. Proc. Assoc. Am. Physicians 110:183. 23. Hancock, W. W., W. Gao, V. Csizmadia, K. L. Faia, N. Shemmeri, and A. D. Luster. 2001. Donor-derived IP-10 initiates development of acute allograft rejection. J. Exp. Med. 193:975. 24. Dufour, J. H., M. Dziejman, M. T. Liu, J. H. Leung, T. E. Lane, and A. D. Luster. 2002. IFN-␥-inducible protein 10 (IP-10, CXCL10)-deficient mice reveal a role for IP-10 in effector T cell generation and trafficking. J. Immunol. 168:3195.

25. Hamelmann, E., J. Schwarze, K. Takeda, A. Oshiba, G. L. Larsen, C. G. Irvin, and E. W. Gelfand. 1997. Noninvasive measurement of airway responsiveness in allergic mice using barometric plethysmography. Am. J. Respir. Crit. Care Med. 156:766. 26. Luster, A. D., and P. Leder. 1993. IP-10, a -C-X-C- chemokine, elicits a potent thymus-dependent antitumor response in vivo. J. Exp. Med. 178:1057. 27. Rothenberg, M. E., A. D. Luster, and P. Leder. 1995. Murine eotaxin: an eosinophil chemoattractant inducible in endothelial cells and in interleukin 4-induced tumor suppression. Proc. Natl. Acad. Sci. USA 92:8960. 28. Luster, A. D., S. M. Greenberg, and P. Leder. 1995. The IP-10 chemokine binds to a specific cell surface heparan sulfate site shared with platelet factor 4 and inhibits endothelial cell proliferation. J. Exp. Med. 182:219. 29. Krug, N., J. Madden, A. E. Redington, P. Lackie, R. Djukanovic, U. Schauer, S. T. Holgate, A. J. Frew, and P. H. Howarth. 1996. T-cell cytokine profile evaluated at the single-cell level in BAL and blood in allergic asthma. Am. J. Respir. Cell Mol. Biol. 14:319. 30. Sauty, A., M. Dziejman, R. A. Taha, A. D. Iarossi, K. Neote, E. A. Garcia-Zepeda, Q. Hamid, and A. D. Luster. 1999. The T cell-specific CXC chemokines IP-10, Mig, and I-TAC are expressed by activated bronchial epithelial cells. J. Immunol. 162:3549. 31. Loetscher, P., A. Pellegrino, J. H. Gong, I. Mattioli, M. Loetscher, G. Bardi, M. Baggiolini, and I. Clark-Lewis. 2001. The ligands of CXC chemokine receptor 3, I-TAC, Mig, and IP10, are natural antagonists for CCR3. J. Biol. Chem. 276:2986. 32. Wiley, R. E., K. Palmer, B. U. Gajewska, M. R. Stampfli, D. Alvarez, A. J. Coyle, J. C. Gutierrez-Ramos, and M. Jordana. 2001. Expression of the Th1 chemokine IFN-␥-inducible protein 10 in the airway alters mucosal allergic sensitization in mice. J. Immunol. 166:2750. 33. Nanki, T., and P. E. Lipsky. 2000. Lack of correlation between chemokine receptor and T(h)1/T(h)2 cytokine expression by individual memory T cells. Int. Immunol. 12:1659. 34. Annunziato, F., L. Cosmi, G. Galli, C. Beltrame, P. Romagnani, R. Manetti, S. Romagnani, and E. Maggi. 1999. Assessment of chemokine receptor expression by human Th1 and Th2 cells in vitro and in vivo. J. Leukocyte Biol. 65:691. 35. Gonzalo, J. A., C. M. Lloyd, L. Kremer, E. Finger, A. C. Martinez, M. H. Siegelman, M. Cybulsky, and J. C. Gutierrez-Ramos. 1996. Eosinophil recruitment to the lung in a murine model of allergic inflammation: the role of T cells, chemokines, and adhesion receptors. J. Clin. Invest. 98:2332. 36. Hamelmann, E., A. Oshiba, J. Paluh, K. Bradley, J. Loader, T. A. Potter, G. L. Larsen, and E. W. Gelfand. 1996. Requirement for CD8⫹ T cells in the development of airway hyperresponsiveness in a marine model of airway sensitization. J. Exp. Med. 183:1719. 37. O’Sullivan, S., L. Cormican, J. L. Faul, S. Ichinohe, S. L. Johnston, C. M. Burke, and L. W. Poulter. 2001. Activated, cytotoxic CD8⫹ T lymphocytes contribute to the pathology of asthma death. Am. J. Respir. Crit. Care Med. 164:560. 38. Jinquan, T., C. Jing, H. H. Jacobi, C. M. Reimert, A. Millner, S. Quan, J. B. Hansen, S. Dissing, H. J. Malling, P. S. Skov, et al. 2000. CXCR3 expression and activation of eosinophils: role of IFN-␥-inducible protein-10 and monokine induced by IFN-␥. J. Immunol. 165:1548. 39. Wang, X., T.-L. Yue, E. H. Ohlstein, C.-P. Sung, and G. Z. Feuerstein. 1996. Interferon-inducible protein-10 involves vascular smooth muscle cell migration, proliferation and inflammatory response. J. Biol. Chem. 271:24286.