Primary Immunization in the Canine Lung Soluble Antigen Induces a Localized Response 1 - 4
DAVID N. WEISSMAN, DAVID E. BICE, BRUCE A. MUGGENBURG, PATRICK J. HALEY, GEORGE M. SHOPP, and MARK R. SCHUYLER
Introduction
Instillation of neoantigen into the distal respiratory tract has been an important model for elucidating mechanisms of primary humoral immunity in the lung (1,2). In general, a particulate neoantigen, the sheep red blood cell (SRBC), has been utilized for instillation, with antibody-forming cell(AFC) responses quantitated by plaque-forming assay (1,2). In all species studied, the primary site at which an immune response was induced after immunizing the lung with SRBC has been the draining lung-associated lymph node, with large numbers of AFC found at this site (1-6). There may be species differences in the numbers of AFC released from lungassociated lymph nodes after primary intrapulmonary immunization (1). In dogs and nonhuman primates, large numbers of specific AFC are released from lungassociated lymph nodes into the circulation, and functioning AFC reach the pulmonary parenchyma from the blood (1-3, 7, 8). In these large species, far greater numbers of AFC are found in inflamed, immunized lobes than in saline-instilled, control lobes (1, 2, 7, 8). In comparison with dogs and nonhuman primates, after a primary immunization, rodents release relatively fewer AFC from lungassociated lymph nodes into the circulation, and the penetration of the primary AFC response into lung parenchyma is correspondingly decreased (1). However, a primary intratracheal immunization with large doses of SRBC is still capable of recruiting functioning AFC to the lungs of mice (1, 9). Compared to what is known about the primary response to intrapulmonary instillation of particulate antigen, far less is known for soluble antigen. In previous studies with rodents, the specific humoral response induced by immunizing the murine lung with soluble neoantigen was similar to that induced by particu6
SUMMARY Primary immunization of the dog by intralobar Instillation of particulate antigen induces an Intense, localized pulmonary antibody response. In contrast, although soluble antigen can also Induce local antibody responses after repeated deposition in the canine respiratory tract, its ability to induce local responses after primary immunization has not been well characterized. To document such responses, we Immunized five beagle dogs using a bronchoscope to instill 10 mg keyhole limpet hamocyanln (KLH) Into a single lung lobe (Immunized) and saline Into a contralaterallung lobe (control). Over the next 3 wk, we monitored specific Immune responses In blood and bronchoalveolar lavage (BAL) fluids obtained from the Immunized and control lung lobes. Primary Intrapulmonary Immunization of dogs with KLH resulted In antl-KLH antibody responses both in blood and in immunized and control BAL fluids. However, immunoglobulin class-specific expression of response differed between the Immunized and control lung lobes. Specific IgM and IgA responses were significantly greater In the Immunized lobes. In contrast, specific IgG, and cells producing specific IgG, were quantitatively similar in lavage fluids derived from Immunized and control lung lobes. These studies demonstrate that primary Immunization of the dog by Intralobar Instillation of soluble antigen stimulates a locallgM and IgA response and an IgG response that distributes to both immunized and unlmmunlzed lung. This pattern of Immunoglobulin class-specific pulmonary antibody response has the potential to Importantly influence regional responses to intrapulmonary antigen. AM REV RESPIR DIS 1992; 145:6-12
late SRBC (10). Specifically, after primary immunization, functioning AFC were present in draining paratracheal nodes but could not be demonstrated in pulmonary parenchyma (10). These studies did not address potential species differences in responses to pulmonary immunization, for example the possibility that soluble antigen could recruit functioning AFC to the lungs of a larger animal species, such as the dog. In the present study, we used a canine model to investigate the expression of specific humoral immunity after primary intrapulmonary immunization with a well-characterized, highly immunogenic, soluble antigen, keyhole limpet hemocyanin (KLH). Wehypothesized that primary immunization of dogs by intrapulmonary instillation of KLH would recruit antigen-specific AFC to immunized lung lobes, in a fashion similar to that previously noted for SRBC. These studies demonstrate that, as with SRBC, primary intrapulmonary immunization with soluble KLH recruits functioning AFC to the canine lung. However, KLHimmunized and contralateral control
lung lobes manifest important immunoglobulin class-specific differences in response that were not apparent in previous studies following intrapulmonary immunization with particulate SRBC. Methods
Reagents RPMI consisted of RPMI 1640tissue culture medium, supplemented with 25 mM HEPES, (Received in originalform December 18, 1990and in revised form June 5, 1991) 1 From the Albuquerque Veterans Administration Medical Center, the Lovelace Medical Foundation, Research Division, and the Inhalation Toxicology Research Institute, Albuquerque, New Mexico. 2 Supported in part by the VeteransAdministration, U.S. Department of Energy Office of Health and Environmental Research Contract No. DEAC04-76EVOI013, National Institutes of Health BiomedicalResearch Support Grant Program Grant No. BRSG S07 RR-05583-22, and the American Lung Association. 3 Presented in part at the 1989 American Thoracic Society Meeting, Cincinnati, OH. 4 Correspondence and requests for reprints should be addressed to David N. Weissman, M.D., Albuquerque VAMC,Pulmonary/lilA, 2100Ridgecrest Drive SE, Albuquerque, NM 87108.
7
PRIMARY LUNG IMMUNIZATION
2 mM L-glutamine (GIBCO, Grand Island, NY), and 50 ug/rnl of gentamycin (Sigma Chemical Company, St. Louis, MO). RPMII FCS consisted of RPMI supplemented with 10070 heat-inactivated fetal calf serum (HyClone Laboratories, Inc., Logan, UT). Ficoll'" 400-sodium diatrizoate solution (Isolymphs) was obtained from Gallard-Schlesinger Chemical Manufacturing Corp. (Carle Place NY). Goat antidog IgM, IgG, and IgA antibodies and sheep antirabbit IgG-FITC conjugate were purchased from Organon Teknika-Cappel (Malvern, PA). Rabbit antidog Ig was purchased from Immunovisions, Inc. (Springdale, AR). Rabbit antigoat, alkaline phosphatase-conjugated antibody; p-nitrophenylphosphate; and Tween-® 20 were obtained from Sigma Chemical Company (St. Louis, MO).
Keyhole Limpet Hemocyanin Keyhole limpets were obtained from Pacific Biomarine (Venice, CA). Because of the nonspecific inflammatory properties of many commercially available KLH preparations, the KLH used in these studies was prepared according to the procedure of Vandenbark and colleagues (11). Briefly,each limpet was rinsed with sterile saline and its foot was rubbed briskly with a gloved hand. Hemolymph oozing from the limpets was collected overnight at 4° C. KLH was prepared from crude hemolymph by differential ultracentrifugation (11). The pelleted KLH was dissolved in normal saline, dialyzed at 4° C against normal saline, and filter sterilized through a OA5-llm filter. KLH prepared in this manner did not cause nonspecific skin reactions in guinea pigs, dogs, or humans (data not shown). Endotoxin levels, as determined by the limulus amebocyte lysate assay (E-toxate®; Sigma Chemical Company, S1. Louis, MO), were less than 0.1 ng/rng of KLH.
Experimental Animals Five female beagle dogs, 2.5 to 3 yr old, were used in these studies. None had previously been exposed to KLH. Before the pulmonary immunizations, all dogs were shown to be healthy by physical examination, chest roentgenography, routine blood chemistries, and complete blood counts. Dogs were fed once a day with Wayne dog food (Wayne Pet Food Division, Continental Grain Company, Chicago, IL); water was available ad libitum. The dogs were housed in indoor-outdoor facilities at the Inhalation Toxicology Research Institute.
Animal Procedures Transbronchoscopic instillation of KLH into selected lung lobes was performed in a manner similar to that previously described (1, 4, 7, 8). Briefly, each dog was anesthetized with 5010 halothane in oxygen and intubated and anesthesia maintained with 2.5% halothane in oxygen. A fiberoptic bronchoscope was then directed through the endotracheal tube and into the desired bronchial orifice. A poly-
ethylene catheter was passed though the suction channel of the bronchoscope into the airway and gently advanced as distally as possible. For each dog, we instilled 1 ml sterile normal saline through a catheter placed in the left cardiac lobe and then instilled 10 mg KLH in 1 ml sterile normal saline through a catheter placed in the right cardiac lobe. Bronchoalveolar lavages were performed in intubated, anesthetized dogs by passing a fiberoptic bronchoscope into immunized or saline-instilled bronchial sites and wedging it into place. Lavages were performed using five lO-ml aliquots of normal saline, each injected through the bronchoscope and then recovered by applying gentle suction with a syringe. For each experimental animal, we collected the following samples: blood by jugular venous puncture before intrapulmonary immunization and on Days 3, 5, 7, 10, 12, 14, 17,19, and 21 after immunization, and BAL obtained from immunized right cardiac lung lobes and saline control left cardiac lung lobes at the same intervals as noted for venous blood.
Samples and Cell Populations Serum was prepared from clotted blood and stored frozen at - 20° C. Peripheral blood mononuclear cells (PBMC) were obtained by centrifugaiton of heparinized venous blood over Ficoll-Hypaque", The PBMC recovered from the interface of the plasma-FicollHypaque gradient werewashed twice in RPMI and the resuspended in RPMIIFCS. Bronchoalveolar lavage specimens were separated into bronchoalveolar lavage fluid (BALF) supernatants and bronchoalveolar lavage cell (BLC) populations by centrifugation (500 x g for 10min at 4° C). BALF samples were decanted and stored at - 20° C. BLC pellets were then washed twice in RPMI and resuspended in RPMIIFCS. In all cases, cell populations obtained were enumerated using a Coulter counter (Coulter Electronics, Inc., Hialeah, FL). Cell differentials were performed by counting cells on cytocentrifuge preparations stained with DiffQuik®(American ScientificProducts, McGaw Park, IL). Cell viabilities were determined by trypan blue dye exclusion. BALF total protein concentrations weremeasured by a modified Lowry technique using a commercially available kit (Sigma Diagnostics, S1. Louis, MO).
Measurement of In Vitro Anti-KLH Antibody Production The ability of cells from BAL or blood to produce anti-KLH antibody in vitro was assessed by culturing the cells at a concentration of 2 x 106 cells per ml in sterile 12 x 75 mm polypropylene test tubes (16h at 37° C in 5% CO 2 and RPMI 1640 media supplemented with 10% FCS, 25 mM HEPES buffer, and 50 ug/rnl of gentamycin). To control for passive release of antibody from cells (12), parallel cultures were performed either with or
without 50 ug/ml of cycloheximide. Conditioned culture media supernatants were obtained by centrifugation of crude culture fluids at 500 x g to pellet cells and debris and were stored at - 20° C. Anti-KLH antibody content in conditioned culture media supernatants was measured using an enzymelinked immunosorbent assay (ELISA), as described subsequently. Studies wereperformed simultaneously on stored culture media supernatants from all time points. In vitro antiKLH antibody production was calculated as the difference between the anti-KLH antibody content of supernatants from untreated cell cultures and the antibody content of supernatants from cultures of cells treated with cycloheximide.These studies therefore measured true production, rather than passive release, of antibody from cells (12-14).
Measurement of Specific Anti-KLH Antibody Anti-KLH antibody was detected using a modfication of a previously described ELISA (10, 15). Briefly, assays were carried out in 96-well, flat-bottomed, polyvinyl microtiter plates (Falcon, Oxnard, CA) that had been coated with 1 ug/ml of KLH in 0.1 M carbonate buffer, pH 9.6, at 4° C, overnight. Plates were washed three times with 0.05% Tween® 20 (Sigma Chemical Co., St. Louis, MO) in phosphate-buffered saline (PBS) before use.Todetect specificanti-KLH antibody, samples were diluted into the linear range of the ELISA with PBS and added to wells, as follows: serum, 1:100 for IgG or IgA and 1:500 for IgM; BALF, 1:10; and conditioned culture media supernatants, 1:4. Plates were incubated overnight at 4° C and washed again in PBS/1\veen. Finally, plates were sequentially incubated with the following solutions, with washing steps between each incubation: goat antidog IgM, IgO, or IgA antibody; alkaline phosphatase-conjugated antigoat second antibody; and the chromogenic alkaline phosphatase substrate, p-nitrophenylphosphate. Color development was detected by measuring optical density at 410 nM (OD 4 I o ) with an automated ELISA plate reader. Specific antibody levels were expressed as color development (OD 4IO) relative to anti-KLH antibody standards by reading plates when wells containing the antibody standards reached an OD 4 1o of 1.0. Standards were chosen so that, for an OD 4 1o of 1.0 or less, the relationship between sample color development and sample anti-KLH antibody content was linear.
Statistics Comparisons between data obtained from immunized and control lung lobes were made using one-tailed paired Student's t tests (16). Multiple comparisons between data collected at various time intervals after intrapulmonary immunization were accomplished by analysis of variance and Bonferroni's multiple-range test (RS/l statistical software; BBN Software Products Corp., Cambridge, MA).
8
WEISSMAN, BICE, MUGGENBURG, HALEY, SHOPP, AND SCHUYLER
In all cases, significance is reported at the p 0.05 level.
DAVID N. WEISSMAN, DAVID E. BICE, BRUCE A. MUGGENBURG, PATRICK J. HALEY, GEORGE M. SHOPP, and MARK R. SCHUYLER
Introduction
Instillation of neoantigen into the distal respiratory tract has been an important model for elucidating mechanisms of primary humoral immunity in the lung (1,2). In general, a particulate neoantigen, the sheep red blood cell (SRBC), has been utilized for instillation, with antibody-forming cell(AFC) responses quantitated by plaque-forming assay (1,2). In all species studied, the primary site at which an immune response was induced after immunizing the lung with SRBC has been the draining lung-associated lymph node, with large numbers of AFC found at this site (1-6). There may be species differences in the numbers of AFC released from lungassociated lymph nodes after primary intrapulmonary immunization (1). In dogs and nonhuman primates, large numbers of specific AFC are released from lungassociated lymph nodes into the circulation, and functioning AFC reach the pulmonary parenchyma from the blood (1-3, 7, 8). In these large species, far greater numbers of AFC are found in inflamed, immunized lobes than in saline-instilled, control lobes (1, 2, 7, 8). In comparison with dogs and nonhuman primates, after a primary immunization, rodents release relatively fewer AFC from lungassociated lymph nodes into the circulation, and the penetration of the primary AFC response into lung parenchyma is correspondingly decreased (1). However, a primary intratracheal immunization with large doses of SRBC is still capable of recruiting functioning AFC to the lungs of mice (1, 9). Compared to what is known about the primary response to intrapulmonary instillation of particulate antigen, far less is known for soluble antigen. In previous studies with rodents, the specific humoral response induced by immunizing the murine lung with soluble neoantigen was similar to that induced by particu6
SUMMARY Primary immunization of the dog by intralobar Instillation of particulate antigen induces an Intense, localized pulmonary antibody response. In contrast, although soluble antigen can also Induce local antibody responses after repeated deposition in the canine respiratory tract, its ability to induce local responses after primary immunization has not been well characterized. To document such responses, we Immunized five beagle dogs using a bronchoscope to instill 10 mg keyhole limpet hamocyanln (KLH) Into a single lung lobe (Immunized) and saline Into a contralaterallung lobe (control). Over the next 3 wk, we monitored specific Immune responses In blood and bronchoalveolar lavage (BAL) fluids obtained from the Immunized and control lung lobes. Primary Intrapulmonary Immunization of dogs with KLH resulted In antl-KLH antibody responses both in blood and in immunized and control BAL fluids. However, immunoglobulin class-specific expression of response differed between the Immunized and control lung lobes. Specific IgM and IgA responses were significantly greater In the Immunized lobes. In contrast, specific IgG, and cells producing specific IgG, were quantitatively similar in lavage fluids derived from Immunized and control lung lobes. These studies demonstrate that primary Immunization of the dog by Intralobar Instillation of soluble antigen stimulates a locallgM and IgA response and an IgG response that distributes to both immunized and unlmmunlzed lung. This pattern of Immunoglobulin class-specific pulmonary antibody response has the potential to Importantly influence regional responses to intrapulmonary antigen. AM REV RESPIR DIS 1992; 145:6-12
late SRBC (10). Specifically, after primary immunization, functioning AFC were present in draining paratracheal nodes but could not be demonstrated in pulmonary parenchyma (10). These studies did not address potential species differences in responses to pulmonary immunization, for example the possibility that soluble antigen could recruit functioning AFC to the lungs of a larger animal species, such as the dog. In the present study, we used a canine model to investigate the expression of specific humoral immunity after primary intrapulmonary immunization with a well-characterized, highly immunogenic, soluble antigen, keyhole limpet hemocyanin (KLH). Wehypothesized that primary immunization of dogs by intrapulmonary instillation of KLH would recruit antigen-specific AFC to immunized lung lobes, in a fashion similar to that previously noted for SRBC. These studies demonstrate that, as with SRBC, primary intrapulmonary immunization with soluble KLH recruits functioning AFC to the canine lung. However, KLHimmunized and contralateral control
lung lobes manifest important immunoglobulin class-specific differences in response that were not apparent in previous studies following intrapulmonary immunization with particulate SRBC. Methods
Reagents RPMI consisted of RPMI 1640tissue culture medium, supplemented with 25 mM HEPES, (Received in originalform December 18, 1990and in revised form June 5, 1991) 1 From the Albuquerque Veterans Administration Medical Center, the Lovelace Medical Foundation, Research Division, and the Inhalation Toxicology Research Institute, Albuquerque, New Mexico. 2 Supported in part by the VeteransAdministration, U.S. Department of Energy Office of Health and Environmental Research Contract No. DEAC04-76EVOI013, National Institutes of Health BiomedicalResearch Support Grant Program Grant No. BRSG S07 RR-05583-22, and the American Lung Association. 3 Presented in part at the 1989 American Thoracic Society Meeting, Cincinnati, OH. 4 Correspondence and requests for reprints should be addressed to David N. Weissman, M.D., Albuquerque VAMC,Pulmonary/lilA, 2100Ridgecrest Drive SE, Albuquerque, NM 87108.
7
PRIMARY LUNG IMMUNIZATION
2 mM L-glutamine (GIBCO, Grand Island, NY), and 50 ug/rnl of gentamycin (Sigma Chemical Company, St. Louis, MO). RPMII FCS consisted of RPMI supplemented with 10070 heat-inactivated fetal calf serum (HyClone Laboratories, Inc., Logan, UT). Ficoll'" 400-sodium diatrizoate solution (Isolymphs) was obtained from Gallard-Schlesinger Chemical Manufacturing Corp. (Carle Place NY). Goat antidog IgM, IgG, and IgA antibodies and sheep antirabbit IgG-FITC conjugate were purchased from Organon Teknika-Cappel (Malvern, PA). Rabbit antidog Ig was purchased from Immunovisions, Inc. (Springdale, AR). Rabbit antigoat, alkaline phosphatase-conjugated antibody; p-nitrophenylphosphate; and Tween-® 20 were obtained from Sigma Chemical Company (St. Louis, MO).
Keyhole Limpet Hemocyanin Keyhole limpets were obtained from Pacific Biomarine (Venice, CA). Because of the nonspecific inflammatory properties of many commercially available KLH preparations, the KLH used in these studies was prepared according to the procedure of Vandenbark and colleagues (11). Briefly,each limpet was rinsed with sterile saline and its foot was rubbed briskly with a gloved hand. Hemolymph oozing from the limpets was collected overnight at 4° C. KLH was prepared from crude hemolymph by differential ultracentrifugation (11). The pelleted KLH was dissolved in normal saline, dialyzed at 4° C against normal saline, and filter sterilized through a OA5-llm filter. KLH prepared in this manner did not cause nonspecific skin reactions in guinea pigs, dogs, or humans (data not shown). Endotoxin levels, as determined by the limulus amebocyte lysate assay (E-toxate®; Sigma Chemical Company, S1. Louis, MO), were less than 0.1 ng/rng of KLH.
Experimental Animals Five female beagle dogs, 2.5 to 3 yr old, were used in these studies. None had previously been exposed to KLH. Before the pulmonary immunizations, all dogs were shown to be healthy by physical examination, chest roentgenography, routine blood chemistries, and complete blood counts. Dogs were fed once a day with Wayne dog food (Wayne Pet Food Division, Continental Grain Company, Chicago, IL); water was available ad libitum. The dogs were housed in indoor-outdoor facilities at the Inhalation Toxicology Research Institute.
Animal Procedures Transbronchoscopic instillation of KLH into selected lung lobes was performed in a manner similar to that previously described (1, 4, 7, 8). Briefly, each dog was anesthetized with 5010 halothane in oxygen and intubated and anesthesia maintained with 2.5% halothane in oxygen. A fiberoptic bronchoscope was then directed through the endotracheal tube and into the desired bronchial orifice. A poly-
ethylene catheter was passed though the suction channel of the bronchoscope into the airway and gently advanced as distally as possible. For each dog, we instilled 1 ml sterile normal saline through a catheter placed in the left cardiac lobe and then instilled 10 mg KLH in 1 ml sterile normal saline through a catheter placed in the right cardiac lobe. Bronchoalveolar lavages were performed in intubated, anesthetized dogs by passing a fiberoptic bronchoscope into immunized or saline-instilled bronchial sites and wedging it into place. Lavages were performed using five lO-ml aliquots of normal saline, each injected through the bronchoscope and then recovered by applying gentle suction with a syringe. For each experimental animal, we collected the following samples: blood by jugular venous puncture before intrapulmonary immunization and on Days 3, 5, 7, 10, 12, 14, 17,19, and 21 after immunization, and BAL obtained from immunized right cardiac lung lobes and saline control left cardiac lung lobes at the same intervals as noted for venous blood.
Samples and Cell Populations Serum was prepared from clotted blood and stored frozen at - 20° C. Peripheral blood mononuclear cells (PBMC) were obtained by centrifugaiton of heparinized venous blood over Ficoll-Hypaque", The PBMC recovered from the interface of the plasma-FicollHypaque gradient werewashed twice in RPMI and the resuspended in RPMIIFCS. Bronchoalveolar lavage specimens were separated into bronchoalveolar lavage fluid (BALF) supernatants and bronchoalveolar lavage cell (BLC) populations by centrifugation (500 x g for 10min at 4° C). BALF samples were decanted and stored at - 20° C. BLC pellets were then washed twice in RPMI and resuspended in RPMIIFCS. In all cases, cell populations obtained were enumerated using a Coulter counter (Coulter Electronics, Inc., Hialeah, FL). Cell differentials were performed by counting cells on cytocentrifuge preparations stained with DiffQuik®(American ScientificProducts, McGaw Park, IL). Cell viabilities were determined by trypan blue dye exclusion. BALF total protein concentrations weremeasured by a modified Lowry technique using a commercially available kit (Sigma Diagnostics, S1. Louis, MO).
Measurement of In Vitro Anti-KLH Antibody Production The ability of cells from BAL or blood to produce anti-KLH antibody in vitro was assessed by culturing the cells at a concentration of 2 x 106 cells per ml in sterile 12 x 75 mm polypropylene test tubes (16h at 37° C in 5% CO 2 and RPMI 1640 media supplemented with 10% FCS, 25 mM HEPES buffer, and 50 ug/rnl of gentamycin). To control for passive release of antibody from cells (12), parallel cultures were performed either with or
without 50 ug/ml of cycloheximide. Conditioned culture media supernatants were obtained by centrifugation of crude culture fluids at 500 x g to pellet cells and debris and were stored at - 20° C. Anti-KLH antibody content in conditioned culture media supernatants was measured using an enzymelinked immunosorbent assay (ELISA), as described subsequently. Studies wereperformed simultaneously on stored culture media supernatants from all time points. In vitro antiKLH antibody production was calculated as the difference between the anti-KLH antibody content of supernatants from untreated cell cultures and the antibody content of supernatants from cultures of cells treated with cycloheximide.These studies therefore measured true production, rather than passive release, of antibody from cells (12-14).
Measurement of Specific Anti-KLH Antibody Anti-KLH antibody was detected using a modfication of a previously described ELISA (10, 15). Briefly, assays were carried out in 96-well, flat-bottomed, polyvinyl microtiter plates (Falcon, Oxnard, CA) that had been coated with 1 ug/ml of KLH in 0.1 M carbonate buffer, pH 9.6, at 4° C, overnight. Plates were washed three times with 0.05% Tween® 20 (Sigma Chemical Co., St. Louis, MO) in phosphate-buffered saline (PBS) before use.Todetect specificanti-KLH antibody, samples were diluted into the linear range of the ELISA with PBS and added to wells, as follows: serum, 1:100 for IgG or IgA and 1:500 for IgM; BALF, 1:10; and conditioned culture media supernatants, 1:4. Plates were incubated overnight at 4° C and washed again in PBS/1\veen. Finally, plates were sequentially incubated with the following solutions, with washing steps between each incubation: goat antidog IgM, IgO, or IgA antibody; alkaline phosphatase-conjugated antigoat second antibody; and the chromogenic alkaline phosphatase substrate, p-nitrophenylphosphate. Color development was detected by measuring optical density at 410 nM (OD 4 I o ) with an automated ELISA plate reader. Specific antibody levels were expressed as color development (OD 4IO) relative to anti-KLH antibody standards by reading plates when wells containing the antibody standards reached an OD 4 1o of 1.0. Standards were chosen so that, for an OD 4 1o of 1.0 or less, the relationship between sample color development and sample anti-KLH antibody content was linear.
Statistics Comparisons between data obtained from immunized and control lung lobes were made using one-tailed paired Student's t tests (16). Multiple comparisons between data collected at various time intervals after intrapulmonary immunization were accomplished by analysis of variance and Bonferroni's multiple-range test (RS/l statistical software; BBN Software Products Corp., Cambridge, MA).
8
WEISSMAN, BICE, MUGGENBURG, HALEY, SHOPP, AND SCHUYLER
In all cases, significance is reported at the p 0.05 level.
