Dendritic Cells Are Required for the Development of Chronic Eosinophilic Airway Inflammation in Response to Inhaled Antigen in Sensitized Mice1 Bart N. Lambrecht,2* Beno�� ��t Salomon,�� David Klatzmann,�� and Romain A. Pauwels* Asthma is characterized by chronic eosinophilic inflammation of the airways, and allergen-specific Th2 lymphocytes are thought to play a major role in the development and maintenance of this type of inflammation in allergic asthma. It is generally accepted that airway dendritic cells (DC) are essential for stimulating naive T cells in a primary immune response to inhaled Ag and for the development of allergic sensitization. We have examined the role of airway DC in stimulating memory T cells in a secondary response to inhaled Ag and the subsequent development of chronic airway inflammation. In our mouse model of asthma, OVA aerosol challenge in OVA-sensitized mice leads to CD4-dependent peribronchial and perivascular eosinophilic inflammation, lung Th2 cytokine production, and systemic IgE production. We have used conditional depletion of airway DC by treatment of thymidine kinase-transgenic mice with the antiviral drug ganciclovir to deplete DC during the secondary exposure to OVA. In sensitized thymidine kinase-transgenic mice, a significant decrease in the number of bronchoalveolar CD4 and CD8 T lymphocytes and B lymphocytes was seen after ganciclovir treatment. In addition, Th2 cytokine-associated eosinophilic airway inflammation was almost completely suppressed. These studies demonstrate for the first time that the DC is essential for presenting inhaled Ag to previously primed Th2 cells in the lung, leading to chronic eosinophilic airway inflammation. Altering the function of airway DC may therefore be an important target for new anti-asthma therapy. The Journal of Immunology, 1998, 160: 4090 ��� 4097. T lymphocytes, secreting Th2 cytokines such as IL-4 and IL-5 in response to inhaled Ag, play a major role in the pathogenesis of allergic bronchial asthma (1). A critical step in the induction of a T cell immune response is the uptake, processing, and presentation of Ag by professional APCs. These cells present immunogenic epitopes of the Ag on the surface of MHC molecules and provide essential costimulatory signals for the induction and differentiation of a primary T cell response (2, 3). Recent studies have emphasized the importance of the dendritic cell (DC)3 as the most potent APC for the induction of a primary immune response to exogenous Ag (4 ��� 6). The ability of naive cells to respond vigorously to DC is largely attributable to the high expression of costimulatory ligands B7-1, B7-2, and ICAM-1 on the surface of DC (7). In contrast, in vitro experiments have shown that previously activated T cells are less dependent on costimula- tory signals for optimal responses to rechallenge with Ag and can respond to a variety of APC, including resting B cells and macrophages (8, 9). In the lung, the network of airway DC is particularly well devel- oped to capture inhaled Ag (10���12). Its location above the basement membrane of the airway epithelium ensures accessibility to inhaled Ag. Upon encounter with inhaled Ag, airway DC migrate to the drain- ing lymph nodes of the lung, up-regulate expression of costimulatory ligands, and interact with naive T lymphocytes, inducing a primary immune response (13���15). However, no data exist on whether DC are critical in the presentation of inhaled Ag to previously activated or memory T cells. We have addressed this question in a T cell-driven mouse model of asthma. We have previously reported that inhala- tional challenge with OVA aerosol in OVA-sensitized mice leads to Th2 cytokine-dependent eosinophilic airway inflammation, bronchial hyperreactivity, and IgE production, findings highly characteristic of human allergic asthma (16���18). In this study, we have used trans- genic (TG) mice expressing the herpes simplex virus type 1 thymidine kinase (TK) suicide gene in the cells of the myeloid DC lineage, allowing the conditional ablation of DC by treatment with the nucle- oside analogue ganciclovir (GCV) (19). This suicide gene technique allows a well-controlled time window of DC depletion during which experiments can be performed in a GCV-free environment. By de- pleting DC before challenge with inhaled Ag, we show that DC are essential for the presentation of Ag to previously activated T cells and are critical for the subsequent development of chronic allergic airway inflammation. These findings have important implications for under- standing the APC requirements for reactivation of memory T cells. The requirement for functional DC for the development of eosino- philic airway inflammation in sensitized mice suggests that DC may provide a useful target for future anti-asthma therapy. Materials and Methods Animals TG mice carrying the TK transgene under the HIV-LTR promoter (19) were from the 12th backcross to C57BL/6 mice. As these mice show an *Department of Respiratory Diseases, University Hospital, B-9000 Ghent, Belgium and �� Laboratoire de Biologie et The ��rapeutique de Pathologies Immunitaires, Centre National de la Recherche Scientifique ERS 107, Ho ��pital de la Pitie ��-Salpe ��trie `re, 75651 Paris Cedex 13, France Received for publication September 17, 1997. Accepted for publication December 22, 1997. 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 in part by the Levenslijn project of the Fonds Weten- schappely ��k Onderzoek Vlaanderen and by a research grant from Zeneca Belgium. B.N.L. is a recipient of a scholarship of the Fund for Scientific Research Vlaanderen. B.S. was supported by La Fondation pour la Recherche Me ��dicale of France. E-mail address: bart.lambrecht@rug.ac.be 2 Address correspondence and reprint requests to Dr. Bart N. Lambrecht, Department of Respiratory Diseases, University Hospital Ghent, De Pintelaan 185, B-9000 Ghent, Belgium. 3 Abbreviations used in this paper: DC, dendritic cell BAL, bronchoalveolar lavage GCV, ganciclovir N-TG, nontransgenic TK, thymidine kinase TG, transgenic WT, wild-type LTR, long terminal repeat SPF, specific pathogen free. Copyright �� 1998 by The American Association of Immunologists 0022-1767/98/$02.00

early wasting syndrome after treatment with GCV, most probably due to ectopic transgene expression in cells of the intestinal epithelium, we per- formed all experiments using irradiated (13 Gy) C57BL/6 hosts reconsti- tuted with TG bone marrow, referred to as TK-TG (20). Wild-type (WT) C57BL/6 mice and C57BL/6 mice that had received bone marrow from nontransgenic (N-TG) littermates served as controls. All animals were housed in sterile microisolator units under specific pathogen free (SPF) conditions. Sensitization and challenge with OVA On day 0, mice (n 5 8 ���10 per group) were actively immunized with OVA (grade III, Sigma Chemical, St. Louis, MO) by the i.p. injection of 10 mg OVA adsorbed to 100 mg of alum adjuvant. Sham-sensitized animals re- ceived i.p. PBS. From day 16 to day 20 after immunization, animals were exposed daily for 30 min to an aerosol of 1% (w/v) OVA in PBS. Groups of 10 mice were placed in an exposure chamber connected to the outlet of an ultrasonic nebulizer that delivers an aerosol of particles with a mean diameter of 3.5 mm (Microvernebler R80, Microvernebler, Zurich, Switzerland). Depletion of dendritic cells by treatment with GCV On day 9, mice were anesthetized by the i.p. injection of avertin (2.5% in PBS) (Sigma). Miniosmotic pumps (ALZET model 2001, Alza, Palo Alto, CA), delivering a dose of 50 mg/kg/day of GCV (Cymevene, Syntex, Brus- sels, Belgium) over a 7-day time period, were implanted s.c. In control experiments, PBS-filled pumps were implanted. On day 15 of the experi- ment, the miniosmotic pumps were removed. To verify the depletion of airway DC in these mice, we modified a protocol to visualize the rat airway DC network (21). In brief, animals were anesthetized and perfused with 1% paraformaldehyde fixative in PBS (pH 7.4) for 2 min through the ascending aorta. Tracheas were removed, opened by a midline incision and pinned flat on silicone-coated petri dishes (Syl- gard 184, Dow Corning, Midland, MI). After permeabilization in PBS containing 0.3% Triton X-100 (Sigma), tissues were preblocked with rabbit serum, incubated with M5/114 mAb (rat IgG2, anti-I-A, and I-Eb,d,k, Boehringer Mannheim, Indianapolis, IN) for 36 h at room temperature, followed by incubation with rabbit anti-rat F(ab9)2 fragments conjugated to horseradish peroxidase (Serotec, Oxford, U.K.) for 24 h. Signal was de- veloped by incubation for 20 min in 0.05% diaminobenzidine (DAB) in Tris-buffered saline (pH 7.6). Tissues were dehydrated in serial alcohol steps and cleared in toluene. The entire trachea was mounted in D.P.X. mounting medium (Nustain, University Hospital, Nottingham, U.K.). The cellular density of the airway DC network (cells/mm2) was calculated us- ing the Impact image analysis system (Alcatel, TITN Answare, Oberkochen, Germany). Bronchoalveolar lavage (BAL) On day 21 of the experiment, 24 h after the last aerosol challenge, mice were sacrificed by sodium pentobarbital overdose (60 mg/kg body weight). BAL was performed with 2 3 1 ml of Ca21- and Mg21-free HBSS sup- plemented with 0.05 mM sodium EDTA, as described (17). The BAL fluid was centrifuged (10 min, 4��C, 700 3 g), and supernatant was collected and stored at 280��C until analysis of cytokine content. After resuspension in HBSS, cells were counted in a hemocytometer (Coulter Counter, Hertford- shire, U.K.). Differential cell counts were performed on cytospin prepara- tions (Cytospin 2, Shandon, Cheshire, U.K.) stained with May-Gru��nwald- Giemsa by classification of 300 cells on standard morphologic criteria. The cytokine content in unconcentrated BAL fluid was determined using com- mercially available cytokine ELISA kits. The ELISA test for determination of murine IFN-g and IL-2 had a sensitivity of 2 pg/ml and 3 pg/ml, re- spectively (Quantikine M, R & D Systems, Abingdon, U.K.). The ELISA tests for determination of murine IL-4 and IL-5 had a sensitivity of 5 pg/ml (Biotrak, Life Science, Amersham, U.K.). Flow cytometric analysis of BAL fluid cells Monoclonal antibodies conjugated to phycoerythrin or FITC were pur- chased from PharMingen (San Diego, CA). All reactions were performed on ice in staining buffer (PBS, 1% BSA, and 0.02% sodium azide). Ali- quots of 2 3 105 BAL cells were incubated for 15 min with 2.4G2 mAb at 5 mg/ml to reduce nonspecific binding via the FcgRII(CD32). After wash- ing in staining buffer, cells were stained with anti-CD3-phycoerythrin (clone 145-2C11) and anti-CD4-FITC (clone H129.19), anti-CD8a-FITC (clone 53-6.7) or anti-B cell mAb B-220-FITC (CD45R, clone RA3-6B2) at 1 mg of mAb/106 cells for 30 min. Controls were generated by staining cells under identical conditions with fluorochrome-conjugated irrelevant myeloma Igs. Cells were fixed in 1% paraformaldehyde in PBS and ana- lyzed on a FACS-Vantage flow cytometer (Becton Dickinson Immunocy- tometry Systems, San Jose, CA). Data were acquired in list mode on 1 3 104 cells and analyzed using Cell Quest software (Becton Dickinson). Airway histology After BAL was performed, fixative (4% paraformaldehyde in PBS) was gently infused through the lavage catheter by a continuous release pump under pressure- and volume-controlled conditions. The lungs were resected and fixed for an additional 4 h. After routine paraffin embedding, 4-mm sections were stained with May-Gru ��nwald-Giemsa and hematoxylin-eosin and examined by light microscopy for histologic changes. Measurement of serum IgE Blood was drawn by cardiac puncture for measurement of serum IgE by isotype-specific ELISA as described previously (18). Briefly, total serum IgE was determined by coating 96-well plates overnight with monoclonal rat anti-mouse IgE LO-ME-3 (H. Bazin, Experimental Immunology Unit, Brussels, Belgium). After blocking with 1% BSA, sample dilutions and known IgE standard were incubated, followed by biotin-conjugated mono- clonal rat anti-mouse IgE LO-ME-2 (H. Bazin) and peroxidase-streptavidin conjugate. The detection limit of the assay was 0.05 mg/ml. OVA-specific IgE was determined by coating plates overnight with OVA grade V (Sigma). Serial dilutions of serum were applied, followed by biotin-conjugated anti-mouse IgE. A serum pool of OVA-sensitized mice was used as internal laboratory standard. A 1:100 dilution of this pool was chosen as arbitrary unit. The lower detection limit of this assay is 0.00165 U/ml. Statistical analysis Total and and OVA-specific IgE levels were logarithmically transformed before calculation of the mean 6 SEM. Comparison of means between different groups was performed with the Mann-Whitney U test for unpaired data using the Spreadware Statistics (Spreadware, Palm Desert, CA) sta- tistical package (22). Differences were considered significant if p , 0.05. Results Conditional depletion of DC from the airways To determine whether DC could be depleted from the airways of TK-TG mice, animals were treated with a continuous release in- fusion of GCV delivered via miniosmotic pumps, as previously described (19). The cellular density of the network lining the air- ways was visualized in tracheal whole mounts. The pattern of MHC class II staining in the airways of WT mice revealed a net- work of strongly positive cells with a dendritic morphology in the upper layers of the respiratory epithelium (Fig. 1, A and B). The density of the network decreased from the upper to the lower tra- chea (Fig. 1D). Administration of GCV for 6 days to TK-TG mice led to a disappearance of more than 95% of airway DC (Fig. 1, C and D), a reduction maintained for at least 4 or 5 days after dis- continuation of GCV treatment (not shown). After the same treat- ment regimen of 6 days, the cellular composition of BAL fluid of unimmunized TK-TG mice was identical in GCV- and PBS- treated animals (Table I). Effect on cellular composition of BAL fluid of DC depletion during secondary challenge with OVA In initial experiments, WT C57BL/6 mice were sensitized to OVA by i.p. injection of OVA in alum and challenged 2 wk later with OVA-aerosol. The cellular composition of BAL fluid was mea- sured as an indicator of airway inflammation. The total number of BAL fluid eosinophils was significantly higher in actively sensi- tized animals compared with sham-sensitized animals ( p , 0.05) (Fig. 2A). The absolute numbers of BAL CD31, CD41, and CD81 T lymphocytes and B lymphocytes were significantly higher in actively sensitized compared with sham-sensitized animals ( p , 0.05) (Fig. 2, B and C). The Ag-induced changes in BAL fluid were not affected by treating WT animals with GCV (Fig. 2, A to C). Moreover, the cellular changes observed in BAL fluid were 4091 The Journal of Immunology