Cloning of T Lymphocytes from Bronchoalveolar Lavage Fluid Bernard E. A. HoI, Frans H. KrouweIs, Ben Bruinier, Richard M. R. Reijneke, Hein J. J. Mengelers, Leo Koenderman, Henk M. Jansen, and Theo A. Out Department of Pulmonology and Clinical Immunology Laboratory, Academic Medical Centre, and Laboratory for Experimental and Clinical Immunology, CLB, Amsterdam, and Department of Pulmonology, University Hospital Utrecht, Utrecht, The Netherlands
We have prepared T-cell clones from bronchoalveolar lavage fluid (BALF) from four healthy, nonsmoking persons and from four patients with allergic asthma. T cells were cloned by direct limiting dilution and with the use of a fluorescent activated cell sorter with an automated cell deposition unit. T-cell clones from the blood (PB) were prepared as well. The cloning efficiencies of T cells from BALF ranged from 3 to 40 % and were lower than those obtained from PB T cells (18 to 72 %). The cloning conditions generated CD4+ as well as CD8+ clones. The very late antigen-4, VLA-4, was more frequently expressed on CD4+ T-cell clones from BALF than from the blood (P < 0.05). CD8+ clones from BALF were more frequently VLA-I + than those from blood (P < < 0.01). Mitogen- and monoclonal antibody-driven proliferation of CD4 + clones showed that BALF clones were well responsive to proliferation stimuli similar to those from the blood. Analysis of interleukin-4 production by 10 BALF and 10 PB clones showed large variations between individual CD4 + clones (BALF: range, < 100 to 700 pg/ml; PB: range, < 100 to 1,100 pg/ml), indicating the generation of different types of clones, which was also clear from analysis of interferon-v production. The analysis of properties of BALF T-cell clones and their regulation will improve insight into immunologic reactions in the lungs.
Lung T lymphocytes increase in number in many lung diseases and are supposed to play an important role in immunologically mediated lung diseases (1). Analysis of these cells has been mainly restricted to enumeration of changes in subpopulations in relation to the inflammatory response (2). Thus, it was found that there is selective recruitment of T cells to the lungs (3-6), where they may undergo typical 10cal differentiation (5), making them distinct from from T cells in the peripheral blood (PB). T lymphocytes on the epithelial surface of the lower respiratory tract exchange slowly with those in the blood (7). They have been studied after having been obtained by bronchoalveolar lavage (BAL). The surface marker profile differs from that in the PB (8, 9), in that bronchoalveolar lavage fluid
(Received in original form December 6,1991 and in revisedform May 21,
1992) Address correspondence to: Prof. Dr. H. M. Jansen, Department of Pulmonology, F4-206, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. Abbreviations: automated cell deposition unit, ACDU; bronchoalveolar lavage, BAL; bronchoalveolar lavage fluid, BALF; concanavalin A, Con A; fluorescein isothiocyanate, FITC; interferon-y IFN-/,; interleukin, IL; Iscove's modified Dulbecco's medium supplemented with pooled human serum, IMDMS; monoclonal antibody, mAb; mean fluorescence intensity, MFI; Mann-Whitney U test, MWU; peripheral blood, PB; peripheral blood mononuclear cells, PBMC; phycoerythrin, PE; phytohemagglutinin, PHA; phorbol myristate acetate, PMA. Am. J. Respir. Cell Mol. BioI. Vol. 7. pp. 523-530, 1992
(BALF) T cells belong mainly to the memory (CD45RO+) subtype and frequently express the VLA-l antigen that is associated with a late-activation stage of the cell (3, 7). Information on functional properties of BALF T cells is difficult to obtain because of the low number of cells recovered by BAL and because of the presence of alveolar macrophages in BALF which influence T-cell functions (10). Although this may be overcome in part by analysis of mRNA expression at the single-cell level, full information on the functions of local T cells may be obtained by studying T-cell clones. Apart from studying properties of BALF-derived T-cell clones from healthy individuals, it is of interest to study those derived from single T-cell progenitors in the BALF from patients with inflammatory lung diseases, e.g., asthma. The aim of the present study was to obtain procedures for cloning and propagation of T lymphocytes from BALF from healthy persons and from patients with allergic asthma. In patients with asthma, the BALF was obtained after bronchoprovocation with allergen as part of a study analyzing the various inflammatory cells in BALF after allergen provocation. This is the first report on establishing T-cell clones directly out of T cells in the BALF from healthy individuals and patients with asthma. We obtained CD4 + as well as CD8+ clones from this compartment in both groups. The expression of several cell surface molecules on the clones from BALF differed from those from blood that were obtained from the same persons as control. The clones were well responsive to proliferation-inducing stimuli and showed a
524
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 71992
broad range of interleukin-4 (IL-4) and interferon-v (IFN --y) production.
Materials and Methods Study Subjects The study included eight nonsmokers (four healthy, nonatopic, and four atopic). The healthy subjects (mean age, 36 yr; range, 31 to 41) had no history of pulmonary disorders. All atopic persons (mean age, 23.5 yr; range, 20 to 28) had a history of asthma and were in a steady state of their disease. They were allergic to house dust mite allergen and at least to one other aero-allergen, as determined by positive (>+ +) skin prick tests and RAST. The patients had a bronchial hyperreactivity for histamine: the preantigen provocative concentration of histamine required to cause a 20 % decrease in FEV! (PC 20 histamine) was < 2 mg/ml. The FEV! and the vital capacity were higher than 60% of the predicted values. Patients with a diminished FEV! showed an improvement to at least 90% of the predicted FEV! after bronchodilatory therapy. Medication was stopped 14 days before the BAL, except for {3-sympathicomimetic inhalation, which was stopped 8 h before the start of the study. None of the study subjects had signs of respiratory tract infection for at least 2 mo before the BAL. The study, performed after obtaining informed consent from all subjects, was approved by the Internal Review Boards. Bronchoprovocation with Allergen The bronchial provocation with allergen was performed by 2-min inhalations (tidal breathing) with lO-min intervals. Aerosols were inhaled through a nebulizer (model 646; DeVilbiss Co., Somerset, PA). The allergen provocation was preceded by an initial inhalation of an aerosol of buffer solution (Vivodiagnost; Laboratory Diephuis, Groningen, The Netherlands) to which none of the patients showed a fall in FEV! or peak expiratory flow. The allergen used for inhalation provocation was Dermatophagoides pteronyssinus (10,000 biological U/rnl, SQ503; Laboratory Diephuis). The concentration used for the first inhalation was calculated from the skin-prick titration test and PC 20 for histamine according to Cockcroft and co-workers (11). Inhalations were performed with aerosols containing stepwise doubling doses of antigen until the FEV! fell at least 20% from baseline values. BAL BAL of healthy subjects (experiments 1 to 4) was performed as described previously (12). Eight 20-rnl aliquots of sterile NaCI (154 mM) warmed to 3rC were instilled sequentially, and each aliquot was immediately aspirated by gentle suction into a sterile siliconized trap. The first seven aliquots were used for the analysis of proteins and cells in BALE The eighth aliquot was collected in 20 rnl of Iscove's modified Dulbecco's medium supplemented with pooled human serum (IMDMS) (10%, vol/vol; heat-inactivated). The specimens contained < 5 % of bronchial epithelial cells and erythrocytes. BAL of the atopic asthmatics (experiments 5 to 8) was performed 5 to 6 h after the inhalation challenge. This is the time point at which possibly occurring late-phase allergic reactions are starting to become manifest. In asthma pa-
tients, four sequential 50-ml aliquots of NaCI were used for lavage. Each aliquot was aspirated separately in siliconized tubes. Eight milliliters of the second aliquot of BALF were immediately withdrawn, added to 20 ml of IMDMS, filtered, and used for the experiments. Preparation of Cells from BALF The cells were separated by centrifugation for 10 min at 500 x g and resuspended in IMDMS. Cell viability was determined by trypan blue dye exclusion and was invariably> 90%. Cytospin preparations (Shandon cytospin 2; Cheshire, UK) were made for analysis of cell differentials. In experiments 1 to 4, the alveolar macrophages were partially removed using anyone of three different methods: experiment 1, Ficoll-paque density gradient sedimentation (s.g. 1.062; Pharmacia, Uppsala, Sweden); experiments 2 and 3, adherence to plastic culture flasks (45 min, 37°C; Falcon, Lincoln Park, NJ); and experiment 4, density sedimentation over Percoll (s.g. 1.063; Pharmacia). The mononuclear cells in the pellet (experiments 1 and 4) and the nonadherent fraction (experiments 2 and 3) were counted in a Burker hemocytometer, and the percentage of lymphocytes was determined on cytospin preparations stained with Jenner-Giemsa, The percentage of T lymphocytes among the lymphocytes was assumed to be 75% (2). The cells were resuspended in culture medium 1: RPMI 1640 supplemented with 2 mM t-glutamine, penicillin (100 U/rnl; Gist-brocades, Delft, The Netherlands), streptomycin (100 ILg/rnl; GIBCO, Grand Island, NY), fungizone (0.2 ILg/rnl; Squibb & Sons, Princeton, NJ), 10% (vol/vol) supernatant of concanavalin A (Con A)-activated peripheral blood mononuclear cells (PBMC) (13), and 10% (vol/vol)-pooled heatinactivated normal human serum (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, CLB, Amsterdam, The Netherlands). In experiments 5 to 8 with the cells of the atopic asthmatics, the BAL cells were labeled with anti-leufib"" (antiCD2; Becton-Dickinson, Mountain View, CA) and isolated with the use of a FACStarP!" cell-sorter (Becton-Dickinson). FLl high-positive cells (negative for FL2) within the lymphogate were used for the cloning experiments. The cell viability of the cells in this preparation was determined using trypan blue dye exclusion and was > 95 %. Isolation of PBMC Heparinized PB was obtained from all eight subjects about 0.5 to 1 h before lavage, and mononuclear cells were isolated on Ficoll-paque (s.g. 1.077) gradients. After three washes in Earle's balanced salt solution (Flow Laboratories, Irvine, UK) containing 10% (vol/vol) pasteurized human plasma (GPO; Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, CLB), the cells were resuspended in culture medium 1. The percentage T cells in the PBMC was determined with anti-Ieu4 (Becton-Dickinson), either on cytospins (14) or by flow cytometry using FACScan (Becton-Dickinson), Immunofluorescent Staining and Flow Cytometry After isolation of the PBMC, the cells were stained with the following monoclonal antibodies (mAbs): anti-Ieu5b (antiCD2), anti-Ieu4 (anti-CD3), anti-Ieu3 (anti-CD4), anti-Ieu2a (anti-CD8), anti-human leukocyte (anti-CD45), anti-IeuM3
HoI, Krouwels, Bruinier et al.: Cloning of T Lymphocytes from BALF
(anti-CDl4) (all anti-leu mAbs were from Becton-Dickinson), all labeled with fluorescein isothiocyanate (FITC) or phycoerythrin (PE). Cells were incubated for 30 min at 4°C with saturating amounts of mAb, washed twice in phosphatebuffered saline containing 10% (vol/vol) GPO, and analyzed by flow cytometry on a FACScan. Lymphocytes were gated on the basis of their forward- and right-angle light scatter. Less than 1% of the cells in the lymphogate thus obtained stained positively with the anti-leu-M3 mAb. Only cells in the lymphogate population were analyzed for fluorescence. When indicated, an additional gate was set on the FLl histogram to discriminate between CD2 F1TC or CD3 F1TC positive and negative cells. A minimum of 1,000 cells was accumulated for each analysis. Cells in BALF were analyzed in a similar way. Alveolar macrophages, which are autofluorescent, were gated out on the basis of their fluorescence in the FL3 channel before staining with any mAb. Thus, almost all autofluorescent signals from macrophages in the FLl and FL2 channels were gated out, as well as the large cells in the scatter plot. In addition to the gates mentioned earlier for PBMC, therefore, an additional gate was set for BALF cells with the use of the FL3 histogram. FITC- and PE-labeled mAbs ofthe same isotype but irrelevant specificity were used as controls to exclude nonspecific binding. The percentage of positively stained cells was computed relative to the control threshold set with the control mAb. The percentage of immunofluorescence-positive cells in the lymphogate was corrected for any positive fluorescence with an isotypematched irrelevant mAb « 2%). Cloning of T Cells Cloning was carried out immediately after preparation of the cells. In experiments 1 to 4, the cloning procedure used was essentially as described by Rotteveel and associates (15), adapted from Moretta and colleagues (16). BAL and PB lymphocytes were seeded into Terasaki wells (Greiner, Frickenhausen, Germany) at a density of 0.3 lymphocyte/well in a final volume of 20 ILl culture medium 1, also containing 2 X 10" feeder cells and phytohemagglutinin (PHA)-P (2 ILg/ ml, HA-16; Wellcome, Dartford, UK). The feeder cells consisted of irradiated (4,000 rad) autologous PBMC. Control wells contained feeder cells only. The plates were incubated at 37°C in a humidified atmosphere of 5% CO2 in air. Cultures were examined for clonal growth after 14 days (BALF and PB T cells) and 21 days (BALF T cells) by visual inspection with an inverted-stage microscope. Positive wells were reseeded in 96-well round-bottom plates and then in 24-well plates after 1 wk (both from Costar, Cambridge, MA). The clones were passed by culturing clone cells together with irradiated heterologous PBMC and PHA (2 ILg/ml) in culture medium 1. Two days after restimulation, culture medium 1 was added to the wells. Clone supernatants were periodically tested for mycoplasmal and other microorganism contamination. After 4 wk in culture, clones were frozen using a computer-controlled freezing device (Kryo 10; Cryotech, Schagen, The Netherlands) and cryopreserved in liquid nitrogen. The cloning efficiency was determined by dividing the number of growth-positive wells by the number of predicted positive wells, assuming that all T cells seeded were induced to clonal proliferation. The number of predicted positive wells was calculated according to Lefkovits and Waldmann (17).
525
In experiments 5 to 8, the cloning procedure was further improved as follows: the BAL and PB cells isolated by the FACSstarpl" were seeded at 1 cell/well by the automated cell deposition unit (ACDU) (18) in 96-well round-bottom trays (Costar) that were coated with immobilized anti-CD3 (CLBT3/3, 10 ILg/rnl; CLB) in a final volume of 100 ILl culture medium 2, which also contained 1 X 105 feeder cells. Culture medium 2 consisted of IMDMS, 2 X 10-5 M mercaptoethanol (Merck, Munchen, Germany), 2 mM Napyruvate (Merck), and antibiotics as described for culture medium 1. The further procedure was essentially as described for experiments 1 to 4. The cloning efficiency was determined by dividing the number of growth-positive wells by the total number of CD3+ cells seeded. To study the influence of alveolar macrophages on the clonal outgrowth of T cells, alveolar macrophages were selected by the FACSstarplus on the basis of their lightscatter pattern. They were seeded as described above at 5 cells/well into wells (100 wells) already containing 1 PB CD2+ cell. Analysis of T-Cell Clone Phenotype The cell surface markers of the clones were analyzed on cytocentrifuge slides (experiments 1 to 3) or by flow cytometry (experiments 4 to 8) after 4 wk in culture. The presence of the CD3, CD4, and CD8 antigens on the cells on the cytospins was determined with anti-leu4, anti-leu3a, and anti-leu2a, respectively, followed by incubation with polyclonal goat anti-mouse immunoglobulin labeled with gold particles (5 ILm, Auroprobel''Llvl; Janssen, Olen, Belgium) and silver enhancement (Intense'r'M; Janssen) (19). The mAbs used in the flow cytometry were identical to those used to characterize the PB and BAL cells. In addition, the mAbs anti-TCR-a(J-1 (Becton-Dickinson), UCHLl (antiCD45RO; DAKO, Santa Barbara, CA), anti-VLA-1 (T Cell Sciences, Cambridge, MA), anti-VLA-4 and anti-VLA-5 (kindly provided by Dr. C. Figdor, The Netherlands Cancer Institute, Amsterdam, The Netherlands), and anti-CD28 (CLB-CD28/1; CLB) were applied. Proliferation Assay Clones were tested at the end of a 1-wkgrowth cycle, essentially as described by Rotteveel and associates (15). Cloned T cells (4 X 104) were cultured in medium 2 in 96-well round- or flat-bottom culture plates (Costar). Culture conditions and concentrations of stimuli were determined in preliminary experiments (data not shown). The cells were stimulated in two ways: with PHA and with mouse mAb IgG2a anti-CD3 (CLBT3/3; immobilized on microtiter plates, coated at 0.5 ILg/rnl). Cultures were performed in a final volume of 150 ILl for 3 days at 37°C in a humidified atmosphere of 5% cO2 in air. Four hours before the end of the incubation period, DNA synthesis was assessed by labeling the cells with 0.2 ILCi ['H]thymidine/rnl (specific activity, 2,000 mCi/mmol; NEN Dupont, Boston, MA). The cells were harvested onto glass filter strips with a semiautomatic cell harvester (Skatron, Lier, Norway). The filter discs were counted in a (J-scintillation counter (Packard 2000CA; Packard Instruments, Meriden, CT). Results are expressed as mean cpm (± 1 SD) of quintuple cultures.
526
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
Analysis of IL-4 and IFN--y Production Clones were incubated (2 x 10' cells/well, 200 JlI) at 37°C in microtiter plates (96 wells, round-bottom) containing culture medium 2. The cells were stimulated by immobilized anti-CD3 and/or a combination of a pair of anti-CD2 mAbs (CLB-Tll.l/1 and CLB-Tl1.2/1, 1 Jlglml), anti-CD28 (CLBCD28/1, 1 Jlglml) together with phorbol myristate acetate (1 nglml; Sigma Chemical Co., St. Louis, MO) (20). After 42 h, the plates were centrifuged and the supernatants frozen until determination of IL-4 and IFN-'¥. IL-4 was assayed by the method of Van der Pouw-Kraan and co-workers (21). Amplification in the assay was achieved using catalyzed reporter deposition (CARD) (22). IFN--ywasmeasured as described by Van der Meide and colleagues (23). Statistics The Mann-Whitney U (MWU) test was applied for the analysis of differences between groups. Probabilities below 0.05 were considered significant.
Results Cells in BALF Table 1 shows the cells in BALE The percentage of lymphocytes in the BALF samples from healthy subjects and patients with asthma ranged from 1 to 12%, with no difference between the two groups. The percentage of eosinophils in the samples from the asthmatics ranged from 4 to 6 % compared with < 1% in control subjects. T-Iymphocyte subpopulations in BALF and PB were determined in experiments 4 to 8 (Table 1). The CD4/CD8 ratio ranged from 0.65 to 1.6 and was similar to that reported in other studies (2). Preparatioft of Cells In experiments 1 to 4, several approaches for the removal of alveolar macrophages were applied. The percentages oflympbocytes in the original BALF and after partial removal of the macrophages were 1% and 4% (Ficoll-paque), 5% and 29% (plastic adherence), 7% and 32% (plastic adherence), and 11% and 76% (Percoll), respectively. Recovery of lymphocytes was> 75 % in all experiments. As there were still too many macrophages left in the lymphocyte preparations, the T cells were positively selected from the BALF cell TABLE 1
Cells in BALF* Subject No.
1 2 3 4 5 6 7 8
Cells in BALF (x 100/ml)
0.9 1.5
1.2 1.6 1.5
2 2.4 1.7
M
L
E
(%)
(%)
(%)
98 94 92 88 79
1 5 7
We have prepared T-cell clones from bronchoalveolar lavage fluid (BALF) from four healthy, nonsmoking persons and from four patients with allergic asthma. T cells were cloned by direct limiting dilution and with the use of a fluorescent activated cell sorter with an automated cell deposition unit. T-cell clones from the blood (PB) were prepared as well. The cloning efficiencies of T cells from BALF ranged from 3 to 40 % and were lower than those obtained from PB T cells (18 to 72 %). The cloning conditions generated CD4+ as well as CD8+ clones. The very late antigen-4, VLA-4, was more frequently expressed on CD4+ T-cell clones from BALF than from the blood (P < 0.05). CD8+ clones from BALF were more frequently VLA-I + than those from blood (P < < 0.01). Mitogen- and monoclonal antibody-driven proliferation of CD4 + clones showed that BALF clones were well responsive to proliferation stimuli similar to those from the blood. Analysis of interleukin-4 production by 10 BALF and 10 PB clones showed large variations between individual CD4 + clones (BALF: range, < 100 to 700 pg/ml; PB: range, < 100 to 1,100 pg/ml), indicating the generation of different types of clones, which was also clear from analysis of interferon-v production. The analysis of properties of BALF T-cell clones and their regulation will improve insight into immunologic reactions in the lungs.
Lung T lymphocytes increase in number in many lung diseases and are supposed to play an important role in immunologically mediated lung diseases (1). Analysis of these cells has been mainly restricted to enumeration of changes in subpopulations in relation to the inflammatory response (2). Thus, it was found that there is selective recruitment of T cells to the lungs (3-6), where they may undergo typical 10cal differentiation (5), making them distinct from from T cells in the peripheral blood (PB). T lymphocytes on the epithelial surface of the lower respiratory tract exchange slowly with those in the blood (7). They have been studied after having been obtained by bronchoalveolar lavage (BAL). The surface marker profile differs from that in the PB (8, 9), in that bronchoalveolar lavage fluid
(Received in original form December 6,1991 and in revisedform May 21,
1992) Address correspondence to: Prof. Dr. H. M. Jansen, Department of Pulmonology, F4-206, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. Abbreviations: automated cell deposition unit, ACDU; bronchoalveolar lavage, BAL; bronchoalveolar lavage fluid, BALF; concanavalin A, Con A; fluorescein isothiocyanate, FITC; interferon-y IFN-/,; interleukin, IL; Iscove's modified Dulbecco's medium supplemented with pooled human serum, IMDMS; monoclonal antibody, mAb; mean fluorescence intensity, MFI; Mann-Whitney U test, MWU; peripheral blood, PB; peripheral blood mononuclear cells, PBMC; phycoerythrin, PE; phytohemagglutinin, PHA; phorbol myristate acetate, PMA. Am. J. Respir. Cell Mol. BioI. Vol. 7. pp. 523-530, 1992
(BALF) T cells belong mainly to the memory (CD45RO+) subtype and frequently express the VLA-l antigen that is associated with a late-activation stage of the cell (3, 7). Information on functional properties of BALF T cells is difficult to obtain because of the low number of cells recovered by BAL and because of the presence of alveolar macrophages in BALF which influence T-cell functions (10). Although this may be overcome in part by analysis of mRNA expression at the single-cell level, full information on the functions of local T cells may be obtained by studying T-cell clones. Apart from studying properties of BALF-derived T-cell clones from healthy individuals, it is of interest to study those derived from single T-cell progenitors in the BALF from patients with inflammatory lung diseases, e.g., asthma. The aim of the present study was to obtain procedures for cloning and propagation of T lymphocytes from BALF from healthy persons and from patients with allergic asthma. In patients with asthma, the BALF was obtained after bronchoprovocation with allergen as part of a study analyzing the various inflammatory cells in BALF after allergen provocation. This is the first report on establishing T-cell clones directly out of T cells in the BALF from healthy individuals and patients with asthma. We obtained CD4 + as well as CD8+ clones from this compartment in both groups. The expression of several cell surface molecules on the clones from BALF differed from those from blood that were obtained from the same persons as control. The clones were well responsive to proliferation-inducing stimuli and showed a
524
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 71992
broad range of interleukin-4 (IL-4) and interferon-v (IFN --y) production.
Materials and Methods Study Subjects The study included eight nonsmokers (four healthy, nonatopic, and four atopic). The healthy subjects (mean age, 36 yr; range, 31 to 41) had no history of pulmonary disorders. All atopic persons (mean age, 23.5 yr; range, 20 to 28) had a history of asthma and were in a steady state of their disease. They were allergic to house dust mite allergen and at least to one other aero-allergen, as determined by positive (>+ +) skin prick tests and RAST. The patients had a bronchial hyperreactivity for histamine: the preantigen provocative concentration of histamine required to cause a 20 % decrease in FEV! (PC 20 histamine) was < 2 mg/ml. The FEV! and the vital capacity were higher than 60% of the predicted values. Patients with a diminished FEV! showed an improvement to at least 90% of the predicted FEV! after bronchodilatory therapy. Medication was stopped 14 days before the BAL, except for {3-sympathicomimetic inhalation, which was stopped 8 h before the start of the study. None of the study subjects had signs of respiratory tract infection for at least 2 mo before the BAL. The study, performed after obtaining informed consent from all subjects, was approved by the Internal Review Boards. Bronchoprovocation with Allergen The bronchial provocation with allergen was performed by 2-min inhalations (tidal breathing) with lO-min intervals. Aerosols were inhaled through a nebulizer (model 646; DeVilbiss Co., Somerset, PA). The allergen provocation was preceded by an initial inhalation of an aerosol of buffer solution (Vivodiagnost; Laboratory Diephuis, Groningen, The Netherlands) to which none of the patients showed a fall in FEV! or peak expiratory flow. The allergen used for inhalation provocation was Dermatophagoides pteronyssinus (10,000 biological U/rnl, SQ503; Laboratory Diephuis). The concentration used for the first inhalation was calculated from the skin-prick titration test and PC 20 for histamine according to Cockcroft and co-workers (11). Inhalations were performed with aerosols containing stepwise doubling doses of antigen until the FEV! fell at least 20% from baseline values. BAL BAL of healthy subjects (experiments 1 to 4) was performed as described previously (12). Eight 20-rnl aliquots of sterile NaCI (154 mM) warmed to 3rC were instilled sequentially, and each aliquot was immediately aspirated by gentle suction into a sterile siliconized trap. The first seven aliquots were used for the analysis of proteins and cells in BALE The eighth aliquot was collected in 20 rnl of Iscove's modified Dulbecco's medium supplemented with pooled human serum (IMDMS) (10%, vol/vol; heat-inactivated). The specimens contained < 5 % of bronchial epithelial cells and erythrocytes. BAL of the atopic asthmatics (experiments 5 to 8) was performed 5 to 6 h after the inhalation challenge. This is the time point at which possibly occurring late-phase allergic reactions are starting to become manifest. In asthma pa-
tients, four sequential 50-ml aliquots of NaCI were used for lavage. Each aliquot was aspirated separately in siliconized tubes. Eight milliliters of the second aliquot of BALF were immediately withdrawn, added to 20 ml of IMDMS, filtered, and used for the experiments. Preparation of Cells from BALF The cells were separated by centrifugation for 10 min at 500 x g and resuspended in IMDMS. Cell viability was determined by trypan blue dye exclusion and was invariably> 90%. Cytospin preparations (Shandon cytospin 2; Cheshire, UK) were made for analysis of cell differentials. In experiments 1 to 4, the alveolar macrophages were partially removed using anyone of three different methods: experiment 1, Ficoll-paque density gradient sedimentation (s.g. 1.062; Pharmacia, Uppsala, Sweden); experiments 2 and 3, adherence to plastic culture flasks (45 min, 37°C; Falcon, Lincoln Park, NJ); and experiment 4, density sedimentation over Percoll (s.g. 1.063; Pharmacia). The mononuclear cells in the pellet (experiments 1 and 4) and the nonadherent fraction (experiments 2 and 3) were counted in a Burker hemocytometer, and the percentage of lymphocytes was determined on cytospin preparations stained with Jenner-Giemsa, The percentage of T lymphocytes among the lymphocytes was assumed to be 75% (2). The cells were resuspended in culture medium 1: RPMI 1640 supplemented with 2 mM t-glutamine, penicillin (100 U/rnl; Gist-brocades, Delft, The Netherlands), streptomycin (100 ILg/rnl; GIBCO, Grand Island, NY), fungizone (0.2 ILg/rnl; Squibb & Sons, Princeton, NJ), 10% (vol/vol) supernatant of concanavalin A (Con A)-activated peripheral blood mononuclear cells (PBMC) (13), and 10% (vol/vol)-pooled heatinactivated normal human serum (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, CLB, Amsterdam, The Netherlands). In experiments 5 to 8 with the cells of the atopic asthmatics, the BAL cells were labeled with anti-leufib"" (antiCD2; Becton-Dickinson, Mountain View, CA) and isolated with the use of a FACStarP!" cell-sorter (Becton-Dickinson). FLl high-positive cells (negative for FL2) within the lymphogate were used for the cloning experiments. The cell viability of the cells in this preparation was determined using trypan blue dye exclusion and was > 95 %. Isolation of PBMC Heparinized PB was obtained from all eight subjects about 0.5 to 1 h before lavage, and mononuclear cells were isolated on Ficoll-paque (s.g. 1.077) gradients. After three washes in Earle's balanced salt solution (Flow Laboratories, Irvine, UK) containing 10% (vol/vol) pasteurized human plasma (GPO; Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, CLB), the cells were resuspended in culture medium 1. The percentage T cells in the PBMC was determined with anti-Ieu4 (Becton-Dickinson), either on cytospins (14) or by flow cytometry using FACScan (Becton-Dickinson), Immunofluorescent Staining and Flow Cytometry After isolation of the PBMC, the cells were stained with the following monoclonal antibodies (mAbs): anti-Ieu5b (antiCD2), anti-Ieu4 (anti-CD3), anti-Ieu3 (anti-CD4), anti-Ieu2a (anti-CD8), anti-human leukocyte (anti-CD45), anti-IeuM3
HoI, Krouwels, Bruinier et al.: Cloning of T Lymphocytes from BALF
(anti-CDl4) (all anti-leu mAbs were from Becton-Dickinson), all labeled with fluorescein isothiocyanate (FITC) or phycoerythrin (PE). Cells were incubated for 30 min at 4°C with saturating amounts of mAb, washed twice in phosphatebuffered saline containing 10% (vol/vol) GPO, and analyzed by flow cytometry on a FACScan. Lymphocytes were gated on the basis of their forward- and right-angle light scatter. Less than 1% of the cells in the lymphogate thus obtained stained positively with the anti-leu-M3 mAb. Only cells in the lymphogate population were analyzed for fluorescence. When indicated, an additional gate was set on the FLl histogram to discriminate between CD2 F1TC or CD3 F1TC positive and negative cells. A minimum of 1,000 cells was accumulated for each analysis. Cells in BALF were analyzed in a similar way. Alveolar macrophages, which are autofluorescent, were gated out on the basis of their fluorescence in the FL3 channel before staining with any mAb. Thus, almost all autofluorescent signals from macrophages in the FLl and FL2 channels were gated out, as well as the large cells in the scatter plot. In addition to the gates mentioned earlier for PBMC, therefore, an additional gate was set for BALF cells with the use of the FL3 histogram. FITC- and PE-labeled mAbs ofthe same isotype but irrelevant specificity were used as controls to exclude nonspecific binding. The percentage of positively stained cells was computed relative to the control threshold set with the control mAb. The percentage of immunofluorescence-positive cells in the lymphogate was corrected for any positive fluorescence with an isotypematched irrelevant mAb « 2%). Cloning of T Cells Cloning was carried out immediately after preparation of the cells. In experiments 1 to 4, the cloning procedure used was essentially as described by Rotteveel and associates (15), adapted from Moretta and colleagues (16). BAL and PB lymphocytes were seeded into Terasaki wells (Greiner, Frickenhausen, Germany) at a density of 0.3 lymphocyte/well in a final volume of 20 ILl culture medium 1, also containing 2 X 10" feeder cells and phytohemagglutinin (PHA)-P (2 ILg/ ml, HA-16; Wellcome, Dartford, UK). The feeder cells consisted of irradiated (4,000 rad) autologous PBMC. Control wells contained feeder cells only. The plates were incubated at 37°C in a humidified atmosphere of 5% CO2 in air. Cultures were examined for clonal growth after 14 days (BALF and PB T cells) and 21 days (BALF T cells) by visual inspection with an inverted-stage microscope. Positive wells were reseeded in 96-well round-bottom plates and then in 24-well plates after 1 wk (both from Costar, Cambridge, MA). The clones were passed by culturing clone cells together with irradiated heterologous PBMC and PHA (2 ILg/ml) in culture medium 1. Two days after restimulation, culture medium 1 was added to the wells. Clone supernatants were periodically tested for mycoplasmal and other microorganism contamination. After 4 wk in culture, clones were frozen using a computer-controlled freezing device (Kryo 10; Cryotech, Schagen, The Netherlands) and cryopreserved in liquid nitrogen. The cloning efficiency was determined by dividing the number of growth-positive wells by the number of predicted positive wells, assuming that all T cells seeded were induced to clonal proliferation. The number of predicted positive wells was calculated according to Lefkovits and Waldmann (17).
525
In experiments 5 to 8, the cloning procedure was further improved as follows: the BAL and PB cells isolated by the FACSstarpl" were seeded at 1 cell/well by the automated cell deposition unit (ACDU) (18) in 96-well round-bottom trays (Costar) that were coated with immobilized anti-CD3 (CLBT3/3, 10 ILg/rnl; CLB) in a final volume of 100 ILl culture medium 2, which also contained 1 X 105 feeder cells. Culture medium 2 consisted of IMDMS, 2 X 10-5 M mercaptoethanol (Merck, Munchen, Germany), 2 mM Napyruvate (Merck), and antibiotics as described for culture medium 1. The further procedure was essentially as described for experiments 1 to 4. The cloning efficiency was determined by dividing the number of growth-positive wells by the total number of CD3+ cells seeded. To study the influence of alveolar macrophages on the clonal outgrowth of T cells, alveolar macrophages were selected by the FACSstarplus on the basis of their lightscatter pattern. They were seeded as described above at 5 cells/well into wells (100 wells) already containing 1 PB CD2+ cell. Analysis of T-Cell Clone Phenotype The cell surface markers of the clones were analyzed on cytocentrifuge slides (experiments 1 to 3) or by flow cytometry (experiments 4 to 8) after 4 wk in culture. The presence of the CD3, CD4, and CD8 antigens on the cells on the cytospins was determined with anti-leu4, anti-leu3a, and anti-leu2a, respectively, followed by incubation with polyclonal goat anti-mouse immunoglobulin labeled with gold particles (5 ILm, Auroprobel''Llvl; Janssen, Olen, Belgium) and silver enhancement (Intense'r'M; Janssen) (19). The mAbs used in the flow cytometry were identical to those used to characterize the PB and BAL cells. In addition, the mAbs anti-TCR-a(J-1 (Becton-Dickinson), UCHLl (antiCD45RO; DAKO, Santa Barbara, CA), anti-VLA-1 (T Cell Sciences, Cambridge, MA), anti-VLA-4 and anti-VLA-5 (kindly provided by Dr. C. Figdor, The Netherlands Cancer Institute, Amsterdam, The Netherlands), and anti-CD28 (CLB-CD28/1; CLB) were applied. Proliferation Assay Clones were tested at the end of a 1-wkgrowth cycle, essentially as described by Rotteveel and associates (15). Cloned T cells (4 X 104) were cultured in medium 2 in 96-well round- or flat-bottom culture plates (Costar). Culture conditions and concentrations of stimuli were determined in preliminary experiments (data not shown). The cells were stimulated in two ways: with PHA and with mouse mAb IgG2a anti-CD3 (CLBT3/3; immobilized on microtiter plates, coated at 0.5 ILg/rnl). Cultures were performed in a final volume of 150 ILl for 3 days at 37°C in a humidified atmosphere of 5% cO2 in air. Four hours before the end of the incubation period, DNA synthesis was assessed by labeling the cells with 0.2 ILCi ['H]thymidine/rnl (specific activity, 2,000 mCi/mmol; NEN Dupont, Boston, MA). The cells were harvested onto glass filter strips with a semiautomatic cell harvester (Skatron, Lier, Norway). The filter discs were counted in a (J-scintillation counter (Packard 2000CA; Packard Instruments, Meriden, CT). Results are expressed as mean cpm (± 1 SD) of quintuple cultures.
526
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992
Analysis of IL-4 and IFN--y Production Clones were incubated (2 x 10' cells/well, 200 JlI) at 37°C in microtiter plates (96 wells, round-bottom) containing culture medium 2. The cells were stimulated by immobilized anti-CD3 and/or a combination of a pair of anti-CD2 mAbs (CLB-Tll.l/1 and CLB-Tl1.2/1, 1 Jlglml), anti-CD28 (CLBCD28/1, 1 Jlglml) together with phorbol myristate acetate (1 nglml; Sigma Chemical Co., St. Louis, MO) (20). After 42 h, the plates were centrifuged and the supernatants frozen until determination of IL-4 and IFN-'¥. IL-4 was assayed by the method of Van der Pouw-Kraan and co-workers (21). Amplification in the assay was achieved using catalyzed reporter deposition (CARD) (22). IFN--ywasmeasured as described by Van der Meide and colleagues (23). Statistics The Mann-Whitney U (MWU) test was applied for the analysis of differences between groups. Probabilities below 0.05 were considered significant.
Results Cells in BALF Table 1 shows the cells in BALE The percentage of lymphocytes in the BALF samples from healthy subjects and patients with asthma ranged from 1 to 12%, with no difference between the two groups. The percentage of eosinophils in the samples from the asthmatics ranged from 4 to 6 % compared with < 1% in control subjects. T-Iymphocyte subpopulations in BALF and PB were determined in experiments 4 to 8 (Table 1). The CD4/CD8 ratio ranged from 0.65 to 1.6 and was similar to that reported in other studies (2). Preparatioft of Cells In experiments 1 to 4, several approaches for the removal of alveolar macrophages were applied. The percentages oflympbocytes in the original BALF and after partial removal of the macrophages were 1% and 4% (Ficoll-paque), 5% and 29% (plastic adherence), 7% and 32% (plastic adherence), and 11% and 76% (Percoll), respectively. Recovery of lymphocytes was> 75 % in all experiments. As there were still too many macrophages left in the lymphocyte preparations, the T cells were positively selected from the BALF cell TABLE 1
Cells in BALF* Subject No.
1 2 3 4 5 6 7 8
Cells in BALF (x 100/ml)
0.9 1.5
1.2 1.6 1.5
2 2.4 1.7
M
L
E
(%)
(%)
(%)
98 94 92 88 79
1 5 7
