Altered Adhesion Molecule Expression and Endothelial Cell Activation Accompany the Recruitment of Human Granulocytes to the Lung after Segmental Antigen Challenge Steve N. Georas, Mark C. Liu, Walter Newman, L. Dawson Beall, Becky A. Stealey, and Bruce S. Bochner Divisions of Clinical Immunology and Pulmonary and Critical Care Medicine, Johns Hopkins Asthma and Allergy Center, Baltimore, and Otsuka America Pharmaceutical, Inc., Rockville, Maryland
Mounting evidence suggests that inflammatory cells recruited to the lung can contribute to the pathogenesis of asthma. The factors governing the activation and recruitment of circulating cells to the lung remain unknown, but an early step in this process is the interaction of adhesion molecules on circulating cells with those on endothelial cells. We used a segmental antigen challenge model followed 18 h later by bronchoalveolar lavage (BAL) to study granulocyte recruitment to the lung in 14 allergic subjects. Using immunofluorescence and flow cytometry, we determined the expression of the adhesion molecules CDllb, L-selectin (LECAM-l), and VLA-4 on BAL and peripheral blood granulocytes. Total cell count and percentages of recovered eosinophils and basophils were significantly increased in BAL fluids from antigenchallenged segments. Compared with their peripheral blood counterparts, CDllb expression was increased 2- to 3-fold on BAL eosinophils, basophils, and neutrophils (n = 9, P < 0.05). In contrast, L-selectin expression was significantly decreased on BAL cells (n = 3 to 4, P < 0.05). Similar phenotypic changes were observed on all three cell types, and on neutrophils recovered from saline-challenged control lung segments. In two subjects, VLA-4 IX (CD49d) expression on BAL eosinophils was 78 ± 5% of that seen on peripheral blood eosinophils. Because ELAM-l (endothelial leukocyte adhesion molecule-l, E-selectin) expression occurs during allergic inflammation and is shed after endothelial activation, we used a sensitive enzyme-linked immunosorbent assay to analyze BAL supernatants for a soluble form of this molecule (sELAM-l). sELAM-l was detected in lOx concentrated BAL fluid only after antigen challenge (1.01 ± 0.05 versus 0.06 ± 0.03 nglrnl, antigen versus saline challenge, n = 12, P< 0.05). These data show that altered adhesion molecule expression and endothelial activation accompany cell recruitment to the lung during allergic inflammation. Because similar phenotypic changes were observed on BAL neutrophils recovered after saline challenge, these events appear to be associated with leukocyte recruitment regardless of stimulus. The precise mechanism responsible for the enhanced recruitment of eosinophils and basophils after antigen challenge remains to be elucidated.
In recent years, increasing emphasis has been placed on the role of airway inflammation in the pathogenesis of asthma. Numerous studies using bronchoscopy and bronchoalveolar (Received in original form September 9, 1991 and in revised form March 2, 1992) Address correspondence to: Bruce S. Bochner, M.D., Assistant Professor of Medicine, Johns Hopkins Asthma and Allergy Center, 5501 Bayview Circle, Baltimore, MD 21224. Abbreviations: bronchoalveolar lavage, BAL; bovine serum albumin, BSA; endothelial leukocyte adhesion molecule-l, ELAM-I; enzyme-linked immunosorbent assay, ELISA; fluorescein isothiocyanate, FITC; interleukin, IL; monoclonal antibody, mAb; mean fluorescence intensity, MFI; phosphate-buffered saline, PBS; phosphate-buffered saline containing 0.1% bovine serum albumin, PBS-BSA; phycoerythrin, PE; protein nitrogen unit, PNU; soluble endothelial leukocyte adhesion molecule-l, sELAM-I; tumor necrosis factor, TNF. Am. J. Respir. CeU Mol. BioI. Vol. 7. pp. 261-269, 1992
lavage (BAL) have documented evidence of airway inflammation in even mild asthmatics (reviewed in reference 1). Inflammatory cells have also been detected infiltrating peribronchial tissue in biopsies obtained from asthmatic subjects (2). Both resident and recruited inflammatory cells are felt to contribute to the pathophysiology of asthma by secreting a variety ofbioactive substances, including chemical and lipid mediators and cytokines (1, 3). Many of these substances have been implicated in causing the bronchoconstriction, hyperreactivity, airway edema, and epithelial desquamation that are characteristic of asthma (1). The importance of inflammatory cell recruitment in the pathogenesis of asthma and other allergic diseases is further supported by the observation that drugs effective in the treatment of these diseases (e.g., corticosteroids) prevent the antigen-induced accumulation of inflammatory cells (4). Despite their obvious presence, the mechanisms by which
262
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
circulating cells are recruited to the lung in asthma remain unknown. A critical initial step in cell recruitment is the interaction of adhesion molecules on circulating cells with those on vascular endothelium adjacent to the inflammatory site (5). Two important families of adhesion molecules are the integrins and the selectins (6-9). The integrins are heterodimers that mediate both cell-cell and cell-substratum interactions. One integrin, termed CDllb/CD18 (Mol, Mac-l, or CR3), has an important role in chemotaxis and is expressed by many cell types including granulocytes (6). Another integrin, termed VLA-4 (CD49d/CD29, a4~1), is not found on neutrophils but is expressed on the surface of lymphocytes, eosinophils, and basophils, where it functions as a counter-receptor for vascular cell adhesion molecule-l (VCAM-1), a cytokine-inducible adhesion molecule belonging to the immunoglobulin supergene family (10-15). Monoclonal antibodies (mAbs) directed against integrins prevent leukocyte accumulation in many animal models of inflammation (6). One notable exception is a rabbit model of lung injury in which anti-CD18 mAbs are ineffective in attenuating the response to certain inhaled toxins (16). The selectins are a more recently described family that share structural similarities including an N-terminallectinlike domain (8,9). One member of this family, termed L-selectin (LECAM-1, Leu-8, or MEL-14), is found on a number of circulating cells (17-19). In vitro studies have shown that leukocyte stimulation (which leads to a significant increase in CD1lb expression) causes shedding of L-selectin from the cell surface (18-21). In animal studies, L-selectin shedding also accompanies transendothelial migration in vivo, and neutrophils lacking L-selectin cannot emigrate out of the circulation (20, 21). These and other studies have led to the hypothesis that L-selectin binding (to an as yet unknown carbohydrate ligand) is important for initial cell attachment to the endothelium, but that it must be shed during diapedesis (21). E-selectin (formerly termed ELAM-1, endothelialleukocyte adhesion molecule-l), another member of the selectin family, is expressed by cytokine-stimulated endothelial cells in vitro and can support the binding of numerous cell types including neutrophils, eosinophils, and basophils (12, 22, 23). One ligand for E-selectin is the tetrasaccharide sialylLewis X (sl.e-) (24-26). After cell activation in vitro, E-selectin expression peaks after 4 to 6 h, then declines toward resting levels by 24 h (22). Although the fate of E-selectin in this system is unknown, a soluble form of the molecule (termed soluble ELAM-1 or sELAM-1) has been detected in supernatants of cytokine-activated endothelial cells (W. Newman, L. D. Beall, C. W. Carson et al. , manuscript submitted). E-selectin expression has been shown to accompany allergic inflammation in human skin (23, 27, 28). In studies oflung injury, E-selectin has been detected in the pulmonary vasculature due to antigen (29), immune-complex (30), or endotoxin (31) mediated inflammation. In two of these models, anti-E-selectin mAbs were effective at attenuating the pulmonary inflammation by blocking influx of neutrophils into the airway (29, 30). Local challenge with antigen has become a frequently employed technique to study the allergic inflammation that occurs in the airways within hours of challenge. For example, we and others have previously shown that segmental antigen challenge of appropriate allergic subjects leads to
significant local mediator release and BAL leukocytosis 18 to 48 h after challenge (32-34). In the present study, we used this challenge model to study cell recruitment to the airway in allergic subjects. Because most of the in vivo studies of inflammatory cell recruitment to date have been performed using animal models and nonspecific inflammatory stimuli, we wondered if neutrophils, eosinophils, and basophils recruited to the human lung during allergic inflammation would express more CDllb and VLA-4, and less L-selectin, than their circulating counterparts. Because eosinophils and basophils appear to be preferentially recruited to the lung in this model (33), we hypothesized that a different pattern of adhesion molecule expression might accompany the influx of these cells. Using immunofluorescence and flow cytometry, we analyzed the cell surface expression of CD11b and L-selectin on eosinophils, basophils, and neutrophils isolated from BAL fluids obtained 18 h after challenge with either antigen or saline; VLA-4 expression was also measured on eosinophils. These values were compared with those obtained from granulocytes in peripheral blood. In addition, to test the hypothesis that endothelial cell activation occurs in the lung after antigen challenge, we used a sensitive doublesandwich enzyme-linked immunosorbent assay (ELISA) to determine whether a soluble form of ELAM-I is released in the lung at sites of antigen challenge.
Materials and Methods Subjects The study protocol was approved by the Institutional Review Board, and all subjects gave written consent. Subject characteristics and experimental design are shown in Table 1. Fourteen healthy, nonsmoking adult volunteers with either allergic rhinitis (n =: 6), asthma (n = 2), or both (n = 6), diagnosed according to previously described clinical criteria (33), were recruited for study. There were seven men and seven women, 27 ± 2 yr of age (mean ± SEM). All subjects were asymptomatic at the time of study. One subject was using an intranasal steroid spray once daily, which was discontinued 24 h before challenge, and three subjects were on oral contraceptives; the remaining subjects were not taking regular medications. Because one male subject was studied on two occasions 4 mo apart, a total of 15 challenges were performed (see Table 1). Antigen Selection All subjects underwent skin prick testing to a battery of common aeroallergens. Subsequently, intradermal skin prick testing with serial dilutions of antigen was performed as described (33) to determine the optimal antigen and dose for peripheral airway challenges. The following antigens were used for segmental challenge (see Table 1): short ragweed antigen (Ambrosia elatior; Greer Laboratories, Lenoir, NC; n = 9); dust mite (Dermatophagoides farinae; HollisterStier, Spokane, WA; n = 3); and timothy grass (Phleum pratense, Greer Laboratories; n = 2). Peripheral Airway Challenges and Bronchoalveolar Lavage The initial bronchoscopic challenge was performed in midafternoon. After premedication with 0.6 mg atropine, given
Georas, Liu, Newman et al.: Adhesion Molecules in Allergic Pulmonary Inflammation
TABLE 1
Study design and subject characteristics Experiments I Subject Age No.* Sex (yr) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
M M M F F M F F M M M F M F F
29 26 24 43 22 22 21 22 23 23 20 38 35 27 27
Ag Dose Dxt
Antigens
R A A,R
Timothy Ragweed Ragweed Ragweed D. farinae Ragweed Ragweed D. farinae Timothy Timothy Ragweed Ragweed D. farinae Ragweed Ragweed
R R R A,R R A,R A,R A,R R A,R A,R R
(PNU)§
50 500 500 500 500 500 500 500 500 500 500 500 500 500 500
Phenotyping Blood, Blood, Blood, Blood, Blood, Blood, Blood, Blood, Blood,
BAL BAL BAL BAL BAL BAL BAL BAL BAL
sELAM No No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
263
(Sigma Chemical Co., St. Louis, MD) or aldan blue as described (33). Total cell count for a given BAL was determined by multiplying the cell count per ml by the total volume returned. Numbers of a given cell type were then calculated by multiplying the percentage of that cell by total cell count. After determining cell counts and viability, the remaining BAL fluid was centrifuged at 1,500 rpm for 10 min at 4 0 C. Supernatants were removed, frozen, and stored at -80 0 C until further analyzed. The cell pellet was resuspended in Hanks' balanced salt solution (GIBCD); approximately 5 X IQ6 cells were removed, washed twice in PAG buffer (25 mM Pipes, 110 mM NaCl, 5 mM KCl containing 0.003% human serum albumin, and 0.1% D-glucose [Sigmaj), and resuspended in phosphate-buffered saline containing 0.1% bovine serum albumin (PBS-BSA) (Sigma). These cells (n = 9; see Table 1)were then labeled and analyzed for expression of various surface markers (see below).
* Subjects 9 and 10 are the same individual who was studied twice, 2 mo apart. t Diagnoses (Ox) of asthma (A) and allergic rhinitis (R) were made according to standard clinical criteria (26). Antigens were selected as described in MATERIALS AND METHODS. D. farinae denotes Dermatophagiodes farinae. § The total antigen dose used for bronchoscopic challenge is shown in protein nitrogen units (PNU). In subject I, the dose was reduced because of extreme skin test reactivity. II Studies performed after each antigen challenge are indicated. Phenotyping denotes immunofluorescent labeling as described in MATERIALS AND METHODS. Values on BAL cells were always compared with those obtained from analyses of simultaneously obtained blood leukocytes. For subjects 6 through 9, phenotyping was also performed on blood cells obtained immediately prior to antigen challenge (i.e., 18 h before BAL). sELAM denotes soluble ELAM assay on serum and BAL supernatants (see MATERIALS AND METHODS).
Peripheral Blood Leukocyte Analysis Venous blood was obtained from each subject immediately before each bronchoscopy. Complete blood counts and differentials were determined using an automated Coulter Counter S+4 (Coulter Electronics, Hialeah, FL) and examination of Wright's stained slides. Leukocytes were isolated from EDTA-anticoagulated blood after sedimentation (90 min, room temperature) in 6% dextran (Kendell McGaw Laboratories, Irvine, CA; n = 9) for immunofluorescent analysis. The leukocyte-rich fraction was withdrawn, washed twice in PAGbuffer, then resuspended in 1 ml of PBS-BSA. Viability as determined by erythrosin B exclusion routinely exceeded 95 %.
intramuscularly, and nebulized 4 % lidocaine, the bronchoscope was introduced and local challenges with both saline and antigen were carried out as described previously (33). Briefly, the tip of the bronchoscope was wedged into a subsegmental bronchus and 5 ml of prewarmed normal saline was instilled ("saline-challenge"). The bronchoscope was withdrawn and then wedged into a contralateral subsegmental bronchus: 5 ml of antigen diluted in normal saline was instilled ("antigen-challenge"). Typically, the lingula and right middle lobe bronchus were used. An antigen concentration of 100 protein nitrogen units (PNU)/ml was selected, based on safety and efficacy as previously shown (33); in one subject, the concentration was reduced to 10 PNU/ml because of extreme skin test sensitivity (see Table 1). Instilled fluids were allowed to remain in the airways; BAL was not performed at this time. The following morning, approximately 18 h after challenge (17.7 ± 0.2 h), a second bronchoscopy was performed. Each of the previously challenged subsegments was lavaged with five 20-ml aliquots of prewarmed normal saline, and fluids recovered from both sites were separately pooled. The volume recovered was determined, and the fluid filtered through two layers of gauze. Cell counts and viability were determined by light microscopy using a hemocytometer and trypan blue exclusion (GIBCD, Grand Island, NY). To determine cell type, cell differentials were performed on separate cytocentrifuge preparations stained with Diff-Quik
Immunofluorescence and Flow Cytometry Peripheral blood and BAL leukocytes were analyzed for expression of CD1Ib and L-selectin after labeling with saturating concentrations of phycoerythrin (PE)-conjugated mAbs (PE-anti-Leu-I5 and PE-anti-Leu-8, respectively; Becton Dickinson, Mountain View, CA). To avoid further cell purification techniques that could alter expression of both of these molecules (20), eosinophils and neutrophils were distinguished by flow cytometry after dual labeling with fluorescein isothiocyanate (FITC)-conjugated anti-CD16 (Becton Dickinson). A combination of light scatter and green fluorescence was then used to distinguish eosinophils (CD16negative) from neutrophils (CD16-positive); the expression of CD1Ib and L-selectin on each cell type could then be determined by quantifying red fluorescence (see Figure 1). In separate cell-sorting experiments (FACStar; Becton Dickinson), using this method, > 80% of CD16-negative cells were found to be eosinophils (data not shown). In separate experiments (n = 2), peripheral blood and BAL cells were also stained with an mAb recognizing the a subunit of the integrin VLA-4 (HP2/I [AMAC, Inc., Westbrook, ME] or L25 [a generous gift of Dr. David Buck, Becton Dickinson Immunocytometry Systems, San Jose, CAD. Adhesion molecule expression on basophils was analyzed after dual labeling with FITC-conjugated polyclonal goat anti-human IgE (Kirkegaard & Perry Labs, Gaithersburg, MD) and PE-conjugated anti-Leu-IS (CDllb) or anti-Leu-8.
*
264
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
,,"r-""T"""- - - - -- - ----,
BLOOD
BLOOD
Mounting evidence suggests that inflammatory cells recruited to the lung can contribute to the pathogenesis of asthma. The factors governing the activation and recruitment of circulating cells to the lung remain unknown, but an early step in this process is the interaction of adhesion molecules on circulating cells with those on endothelial cells. We used a segmental antigen challenge model followed 18 h later by bronchoalveolar lavage (BAL) to study granulocyte recruitment to the lung in 14 allergic subjects. Using immunofluorescence and flow cytometry, we determined the expression of the adhesion molecules CDllb, L-selectin (LECAM-l), and VLA-4 on BAL and peripheral blood granulocytes. Total cell count and percentages of recovered eosinophils and basophils were significantly increased in BAL fluids from antigenchallenged segments. Compared with their peripheral blood counterparts, CDllb expression was increased 2- to 3-fold on BAL eosinophils, basophils, and neutrophils (n = 9, P < 0.05). In contrast, L-selectin expression was significantly decreased on BAL cells (n = 3 to 4, P < 0.05). Similar phenotypic changes were observed on all three cell types, and on neutrophils recovered from saline-challenged control lung segments. In two subjects, VLA-4 IX (CD49d) expression on BAL eosinophils was 78 ± 5% of that seen on peripheral blood eosinophils. Because ELAM-l (endothelial leukocyte adhesion molecule-l, E-selectin) expression occurs during allergic inflammation and is shed after endothelial activation, we used a sensitive enzyme-linked immunosorbent assay to analyze BAL supernatants for a soluble form of this molecule (sELAM-l). sELAM-l was detected in lOx concentrated BAL fluid only after antigen challenge (1.01 ± 0.05 versus 0.06 ± 0.03 nglrnl, antigen versus saline challenge, n = 12, P< 0.05). These data show that altered adhesion molecule expression and endothelial activation accompany cell recruitment to the lung during allergic inflammation. Because similar phenotypic changes were observed on BAL neutrophils recovered after saline challenge, these events appear to be associated with leukocyte recruitment regardless of stimulus. The precise mechanism responsible for the enhanced recruitment of eosinophils and basophils after antigen challenge remains to be elucidated.
In recent years, increasing emphasis has been placed on the role of airway inflammation in the pathogenesis of asthma. Numerous studies using bronchoscopy and bronchoalveolar (Received in original form September 9, 1991 and in revised form March 2, 1992) Address correspondence to: Bruce S. Bochner, M.D., Assistant Professor of Medicine, Johns Hopkins Asthma and Allergy Center, 5501 Bayview Circle, Baltimore, MD 21224. Abbreviations: bronchoalveolar lavage, BAL; bovine serum albumin, BSA; endothelial leukocyte adhesion molecule-l, ELAM-I; enzyme-linked immunosorbent assay, ELISA; fluorescein isothiocyanate, FITC; interleukin, IL; monoclonal antibody, mAb; mean fluorescence intensity, MFI; phosphate-buffered saline, PBS; phosphate-buffered saline containing 0.1% bovine serum albumin, PBS-BSA; phycoerythrin, PE; protein nitrogen unit, PNU; soluble endothelial leukocyte adhesion molecule-l, sELAM-I; tumor necrosis factor, TNF. Am. J. Respir. CeU Mol. BioI. Vol. 7. pp. 261-269, 1992
lavage (BAL) have documented evidence of airway inflammation in even mild asthmatics (reviewed in reference 1). Inflammatory cells have also been detected infiltrating peribronchial tissue in biopsies obtained from asthmatic subjects (2). Both resident and recruited inflammatory cells are felt to contribute to the pathophysiology of asthma by secreting a variety ofbioactive substances, including chemical and lipid mediators and cytokines (1, 3). Many of these substances have been implicated in causing the bronchoconstriction, hyperreactivity, airway edema, and epithelial desquamation that are characteristic of asthma (1). The importance of inflammatory cell recruitment in the pathogenesis of asthma and other allergic diseases is further supported by the observation that drugs effective in the treatment of these diseases (e.g., corticosteroids) prevent the antigen-induced accumulation of inflammatory cells (4). Despite their obvious presence, the mechanisms by which
262
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
circulating cells are recruited to the lung in asthma remain unknown. A critical initial step in cell recruitment is the interaction of adhesion molecules on circulating cells with those on vascular endothelium adjacent to the inflammatory site (5). Two important families of adhesion molecules are the integrins and the selectins (6-9). The integrins are heterodimers that mediate both cell-cell and cell-substratum interactions. One integrin, termed CDllb/CD18 (Mol, Mac-l, or CR3), has an important role in chemotaxis and is expressed by many cell types including granulocytes (6). Another integrin, termed VLA-4 (CD49d/CD29, a4~1), is not found on neutrophils but is expressed on the surface of lymphocytes, eosinophils, and basophils, where it functions as a counter-receptor for vascular cell adhesion molecule-l (VCAM-1), a cytokine-inducible adhesion molecule belonging to the immunoglobulin supergene family (10-15). Monoclonal antibodies (mAbs) directed against integrins prevent leukocyte accumulation in many animal models of inflammation (6). One notable exception is a rabbit model of lung injury in which anti-CD18 mAbs are ineffective in attenuating the response to certain inhaled toxins (16). The selectins are a more recently described family that share structural similarities including an N-terminallectinlike domain (8,9). One member of this family, termed L-selectin (LECAM-1, Leu-8, or MEL-14), is found on a number of circulating cells (17-19). In vitro studies have shown that leukocyte stimulation (which leads to a significant increase in CD1lb expression) causes shedding of L-selectin from the cell surface (18-21). In animal studies, L-selectin shedding also accompanies transendothelial migration in vivo, and neutrophils lacking L-selectin cannot emigrate out of the circulation (20, 21). These and other studies have led to the hypothesis that L-selectin binding (to an as yet unknown carbohydrate ligand) is important for initial cell attachment to the endothelium, but that it must be shed during diapedesis (21). E-selectin (formerly termed ELAM-1, endothelialleukocyte adhesion molecule-l), another member of the selectin family, is expressed by cytokine-stimulated endothelial cells in vitro and can support the binding of numerous cell types including neutrophils, eosinophils, and basophils (12, 22, 23). One ligand for E-selectin is the tetrasaccharide sialylLewis X (sl.e-) (24-26). After cell activation in vitro, E-selectin expression peaks after 4 to 6 h, then declines toward resting levels by 24 h (22). Although the fate of E-selectin in this system is unknown, a soluble form of the molecule (termed soluble ELAM-1 or sELAM-1) has been detected in supernatants of cytokine-activated endothelial cells (W. Newman, L. D. Beall, C. W. Carson et al. , manuscript submitted). E-selectin expression has been shown to accompany allergic inflammation in human skin (23, 27, 28). In studies oflung injury, E-selectin has been detected in the pulmonary vasculature due to antigen (29), immune-complex (30), or endotoxin (31) mediated inflammation. In two of these models, anti-E-selectin mAbs were effective at attenuating the pulmonary inflammation by blocking influx of neutrophils into the airway (29, 30). Local challenge with antigen has become a frequently employed technique to study the allergic inflammation that occurs in the airways within hours of challenge. For example, we and others have previously shown that segmental antigen challenge of appropriate allergic subjects leads to
significant local mediator release and BAL leukocytosis 18 to 48 h after challenge (32-34). In the present study, we used this challenge model to study cell recruitment to the airway in allergic subjects. Because most of the in vivo studies of inflammatory cell recruitment to date have been performed using animal models and nonspecific inflammatory stimuli, we wondered if neutrophils, eosinophils, and basophils recruited to the human lung during allergic inflammation would express more CDllb and VLA-4, and less L-selectin, than their circulating counterparts. Because eosinophils and basophils appear to be preferentially recruited to the lung in this model (33), we hypothesized that a different pattern of adhesion molecule expression might accompany the influx of these cells. Using immunofluorescence and flow cytometry, we analyzed the cell surface expression of CD11b and L-selectin on eosinophils, basophils, and neutrophils isolated from BAL fluids obtained 18 h after challenge with either antigen or saline; VLA-4 expression was also measured on eosinophils. These values were compared with those obtained from granulocytes in peripheral blood. In addition, to test the hypothesis that endothelial cell activation occurs in the lung after antigen challenge, we used a sensitive doublesandwich enzyme-linked immunosorbent assay (ELISA) to determine whether a soluble form of ELAM-I is released in the lung at sites of antigen challenge.
Materials and Methods Subjects The study protocol was approved by the Institutional Review Board, and all subjects gave written consent. Subject characteristics and experimental design are shown in Table 1. Fourteen healthy, nonsmoking adult volunteers with either allergic rhinitis (n =: 6), asthma (n = 2), or both (n = 6), diagnosed according to previously described clinical criteria (33), were recruited for study. There were seven men and seven women, 27 ± 2 yr of age (mean ± SEM). All subjects were asymptomatic at the time of study. One subject was using an intranasal steroid spray once daily, which was discontinued 24 h before challenge, and three subjects were on oral contraceptives; the remaining subjects were not taking regular medications. Because one male subject was studied on two occasions 4 mo apart, a total of 15 challenges were performed (see Table 1). Antigen Selection All subjects underwent skin prick testing to a battery of common aeroallergens. Subsequently, intradermal skin prick testing with serial dilutions of antigen was performed as described (33) to determine the optimal antigen and dose for peripheral airway challenges. The following antigens were used for segmental challenge (see Table 1): short ragweed antigen (Ambrosia elatior; Greer Laboratories, Lenoir, NC; n = 9); dust mite (Dermatophagoides farinae; HollisterStier, Spokane, WA; n = 3); and timothy grass (Phleum pratense, Greer Laboratories; n = 2). Peripheral Airway Challenges and Bronchoalveolar Lavage The initial bronchoscopic challenge was performed in midafternoon. After premedication with 0.6 mg atropine, given
Georas, Liu, Newman et al.: Adhesion Molecules in Allergic Pulmonary Inflammation
TABLE 1
Study design and subject characteristics Experiments I Subject Age No.* Sex (yr) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
M M M F F M F F M M M F M F F
29 26 24 43 22 22 21 22 23 23 20 38 35 27 27
Ag Dose Dxt
Antigens
R A A,R
Timothy Ragweed Ragweed Ragweed D. farinae Ragweed Ragweed D. farinae Timothy Timothy Ragweed Ragweed D. farinae Ragweed Ragweed
R R R A,R R A,R A,R A,R R A,R A,R R
(PNU)§
50 500 500 500 500 500 500 500 500 500 500 500 500 500 500
Phenotyping Blood, Blood, Blood, Blood, Blood, Blood, Blood, Blood, Blood,
BAL BAL BAL BAL BAL BAL BAL BAL BAL
sELAM No No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
263
(Sigma Chemical Co., St. Louis, MD) or aldan blue as described (33). Total cell count for a given BAL was determined by multiplying the cell count per ml by the total volume returned. Numbers of a given cell type were then calculated by multiplying the percentage of that cell by total cell count. After determining cell counts and viability, the remaining BAL fluid was centrifuged at 1,500 rpm for 10 min at 4 0 C. Supernatants were removed, frozen, and stored at -80 0 C until further analyzed. The cell pellet was resuspended in Hanks' balanced salt solution (GIBCD); approximately 5 X IQ6 cells were removed, washed twice in PAG buffer (25 mM Pipes, 110 mM NaCl, 5 mM KCl containing 0.003% human serum albumin, and 0.1% D-glucose [Sigmaj), and resuspended in phosphate-buffered saline containing 0.1% bovine serum albumin (PBS-BSA) (Sigma). These cells (n = 9; see Table 1)were then labeled and analyzed for expression of various surface markers (see below).
* Subjects 9 and 10 are the same individual who was studied twice, 2 mo apart. t Diagnoses (Ox) of asthma (A) and allergic rhinitis (R) were made according to standard clinical criteria (26). Antigens were selected as described in MATERIALS AND METHODS. D. farinae denotes Dermatophagiodes farinae. § The total antigen dose used for bronchoscopic challenge is shown in protein nitrogen units (PNU). In subject I, the dose was reduced because of extreme skin test reactivity. II Studies performed after each antigen challenge are indicated. Phenotyping denotes immunofluorescent labeling as described in MATERIALS AND METHODS. Values on BAL cells were always compared with those obtained from analyses of simultaneously obtained blood leukocytes. For subjects 6 through 9, phenotyping was also performed on blood cells obtained immediately prior to antigen challenge (i.e., 18 h before BAL). sELAM denotes soluble ELAM assay on serum and BAL supernatants (see MATERIALS AND METHODS).
Peripheral Blood Leukocyte Analysis Venous blood was obtained from each subject immediately before each bronchoscopy. Complete blood counts and differentials were determined using an automated Coulter Counter S+4 (Coulter Electronics, Hialeah, FL) and examination of Wright's stained slides. Leukocytes were isolated from EDTA-anticoagulated blood after sedimentation (90 min, room temperature) in 6% dextran (Kendell McGaw Laboratories, Irvine, CA; n = 9) for immunofluorescent analysis. The leukocyte-rich fraction was withdrawn, washed twice in PAGbuffer, then resuspended in 1 ml of PBS-BSA. Viability as determined by erythrosin B exclusion routinely exceeded 95 %.
intramuscularly, and nebulized 4 % lidocaine, the bronchoscope was introduced and local challenges with both saline and antigen were carried out as described previously (33). Briefly, the tip of the bronchoscope was wedged into a subsegmental bronchus and 5 ml of prewarmed normal saline was instilled ("saline-challenge"). The bronchoscope was withdrawn and then wedged into a contralateral subsegmental bronchus: 5 ml of antigen diluted in normal saline was instilled ("antigen-challenge"). Typically, the lingula and right middle lobe bronchus were used. An antigen concentration of 100 protein nitrogen units (PNU)/ml was selected, based on safety and efficacy as previously shown (33); in one subject, the concentration was reduced to 10 PNU/ml because of extreme skin test sensitivity (see Table 1). Instilled fluids were allowed to remain in the airways; BAL was not performed at this time. The following morning, approximately 18 h after challenge (17.7 ± 0.2 h), a second bronchoscopy was performed. Each of the previously challenged subsegments was lavaged with five 20-ml aliquots of prewarmed normal saline, and fluids recovered from both sites were separately pooled. The volume recovered was determined, and the fluid filtered through two layers of gauze. Cell counts and viability were determined by light microscopy using a hemocytometer and trypan blue exclusion (GIBCD, Grand Island, NY). To determine cell type, cell differentials were performed on separate cytocentrifuge preparations stained with Diff-Quik
Immunofluorescence and Flow Cytometry Peripheral blood and BAL leukocytes were analyzed for expression of CD1Ib and L-selectin after labeling with saturating concentrations of phycoerythrin (PE)-conjugated mAbs (PE-anti-Leu-I5 and PE-anti-Leu-8, respectively; Becton Dickinson, Mountain View, CA). To avoid further cell purification techniques that could alter expression of both of these molecules (20), eosinophils and neutrophils were distinguished by flow cytometry after dual labeling with fluorescein isothiocyanate (FITC)-conjugated anti-CD16 (Becton Dickinson). A combination of light scatter and green fluorescence was then used to distinguish eosinophils (CD16negative) from neutrophils (CD16-positive); the expression of CD1Ib and L-selectin on each cell type could then be determined by quantifying red fluorescence (see Figure 1). In separate cell-sorting experiments (FACStar; Becton Dickinson), using this method, > 80% of CD16-negative cells were found to be eosinophils (data not shown). In separate experiments (n = 2), peripheral blood and BAL cells were also stained with an mAb recognizing the a subunit of the integrin VLA-4 (HP2/I [AMAC, Inc., Westbrook, ME] or L25 [a generous gift of Dr. David Buck, Becton Dickinson Immunocytometry Systems, San Jose, CAD. Adhesion molecule expression on basophils was analyzed after dual labeling with FITC-conjugated polyclonal goat anti-human IgE (Kirkegaard & Perry Labs, Gaithersburg, MD) and PE-conjugated anti-Leu-IS (CDllb) or anti-Leu-8.
*
264
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
,,"r-""T"""- - - - -- - ----,
BLOOD
BLOOD
