INFECTION AND IMMUNITY, Feb. 1979, p. 412 417 0019-9567/79/02-0412/06$02.00/0
Vol. 23, No. 2
Human Cell-Mediated Immune Responses to Chlamydial Antigens LAVELLE HANNA,l* LINDA SCHMIDT,' MARGARET SHARP,' DANIEL P. STITES AND ERNEST JAWETZ' Departments of Microbiology' and Laboratory Medicine and Medicine,2 University of California Medical Center, San Francisco, California 94143 ,
Received for publication 11 October 1978
A reproducible method was developed to determine the ability of chlamydial antigens to stimulate lymphocytes from volunteers. In tests repeated 4 to 14 times, the cells from a given volunteer gave a relatively narrow range of responses, but there were great differences in the mean response of different volunteers. In the entire group of 52 volunteers, lymphocyte stimulation was significantly associated with the presence of antibody, but in a given individual results of one test did not aid in predicting the results of the other. A majority of persons with either antichlamydial antibody or elevated lymphocyte stimulation, or both, did not have a history of signs or symptoms within a spectrum of chlamydial diseases. This may reflect the great frequency of asymptomatic infection with these organisms. The lymphocytes of some individuals were stimulated to a significantly greater degree by antigens of one chlamydial species (Chlamydia trachomatis or C. psittaci) than by the other. These and other cell-mediated reactions in human chlamydial infections, and their possible medical significance, are under continued study.
Chlamydia trachomatis is an intracellular antibody responses to chlamydiae in humans parasite which greatly resembles gram-negative (5), but, apart from skin test results, virtually no bacteria and which causes important human data are available on human CMI reactions. A disease, particularly of the eye and the respira- few studies on delayed-type hypersensitivity and tory and genital tracts, involving millions of per- other types of CMI responses have been resons. Some characteristics of chlamydial infec- ported from other animal species in which the tion are the often mild or subclinical but exceed- natural course of chlamydial infection appears ingly chronic course, the ability of tetracycline to be different from that encountered in humans to suppress or eradicate acute infection, and the (7, 8, 14). Therefore, we undertook limited studapparent lack of solid immunity to reinfection in ies on CMI in humans with or without known humans (10). exposure to chlamydiae. The first reaction studIn the course of human chlamydial infections, ied was lymphocyte stimulation by chlamydial a variety of antibodies develop, directed against antigens as measured by increased deoxyribogroup-reactive and immunotype-specific anti- nucleic acid synthesis. The purpose of this work gens. However, substantial titers of such anti- was to establish the reproducibility of this test bodies (whether in serum or secretions) seem to with chlamydial antigens, its relation to the presplay virtually no role in ridding the infected ence of antichlamydial antibodies, and its specperson of parasites or in protecting a "cured" ificity with regard to the two chlamydial species. individual against reinfection. It is tempting to MATERIALS AND METHODS believe that in this-as in other-intracellular infection, cell-mediated immune (CMI) reacStudy population. Fifty-two persons (aged 23 to tions are prominent and perhaps medically im- 63 years) were tested 1 to 14 times each. All volunteers portant. It has been suggested that delayed-type were acquainted with the purpose of the investigation hypersensitivity may play a role in aggravating and signed informed consent documents (University of California, San Francisco, Protocol no. 280201; chlamydial disease of the eye by favoring fibro- guidelines human experimentation of the U.S. Deblastic proliferation, which may lead to loss of partment for of Health, Education, and Welfare were vision (3). Conversely, active cell-mediated re- followed). Some volunteers had a history of probable actions may contribute to resistance to chlam- exposure to chlamydial infection; others did not. A ydial infection. majority were unaware of past symptoms or signs There is a large body of information regarding fitting into the known spectrum of clinical chlamydial 412
VOL. 23, 1979
HUMAN CMI RESPONSES TO CHLAMYDIAL ANTIGENS
disease (3, 11). A minority had experienced ocular, pulmonary, or genital tract symptoms compatible with chlamydial infection. In five volunteers there was laboratory evidence (isolation of agent or antibody titer rise) of chlamydial infection 2 to 12 years before the present study. Preparation of antigens. Strain LB-1 (LGV-2) was used as a prototype of C. trachomatis and strain 6BC was used as a prototype of C. psittaci. Both of these had been passed in eggs many times, but particles from infected yolk sac maintained their antigenic identity with type-specific sera tested by immunofluorescence (4). In addition, some tests included two recent isolates of C. psittaci, 562F and 343, both obtained through the courtesy of J. Schachter. Infected yolk sac preparations containing about 108 50% egg lethal doses (106 inclusion-forming units) were diluted 1/50 in Eagle minimum essential medium with 10% fetal calf serum, 50 tig of gentamicin per ml, and 2.5 tig of amphotericin B (Fungizone) per ml. Ten milliliters of this material was placed on confluent sheets of Vero cells in 150-cm2 plastic flasks and adsorbed for 2.5 h at room temperature. The inoculum was then removed, the monolayers were washed twice with Hanks balanced salt solution, and 25 ml of the growth medium containing an additional 2% glucose was added to each flask. The flasks were incubated in a 5% CO2 incubator at 350C for 3 to 4 days. At the end of this incubation period, cells were scraped off with a rubber policeman, pooled, and shaken vigorously with glass beads. The resulting material was diluted onehalf, and 10-ml volumes were used to inoculate monolayers for further Vero cell passage. After two such passages, the disintegrated cells were spun at 2,000 rpm for 5 min at 80C. The sediment was discarded, and the supernatant was spun at 10,000 rpm in a refrigerated RC-5 Sorvall centrifuge for 5 h. The pellets were resuspended in sterile phosphate-buffered saline (pH 7.5) and pooled, and antibiotics were added (penicillin, 300 ug/ml; streptomycin, 500 jig/ml), blended in a Vortex mixer, and stored at 4°C. If bacteriological cultures of this material remained sterile, particles were counted in Giemsa-stained smears (9) and adjusted to 109 to 101'(/ml. These constituted the final antigen preparations, which were stored for up to 15 months at 4°C. Identical preparations were made from normal, uninfected yolk sac materials, subjected to the same number of cell culture passages and to equivalent dilutions, to serve as antigen controls. Antibody titration. The titer of immunotype-specific antibodies was determined by means of an indirect microimmunofluorescence test. The details of the test have been published (4). Antigens of the nine common immunotypes of C. trachomatis were included, as well as 6BC antigen to represent C. psittaci. Lymphocyte stimulation tests. Heparinized
413
Lymphocyte stimulation tests were carried out in 96-well, flat-bottomed microculture plates (Linbro) as previously described (2). To each well was added, separately, 0.1 ml of a lymphocyte suspension of 5 x 106 cells per ml and 0.1 ml of antigen dilution, using a Hamilton repeating dispenser and a gas-tight syringe, rinsed thoroughly between different antigen dilutions. Each test or control antigen was used in triplicate, and each test included a diluted commercial candida antigen control (Allergenic Extract Dermatophytin "O." 1/40; Hollister-Stier Laboratories, Spokane, Wash.) and a medium control. Plates were incubated in a 100% humidified incubator with an atmosphere of 5% CO2 in air at 37°C. After 5 days of incubation, 1 ,uCi in 0.05 ml of [3H]thymidine (specific activity, 6.7 Ci/mmol) was added to each well. After an additional 6 h of incubation, the cultures were harvested onto glass filter paper strips with a MASH II harvester (Microbiological Associates, Bethesda, Md.). The paper was dried overnight, and the disks were punched out, placed in plastic scintillation vials with 2 ml of Liquifluor-toluene (New England Nuclear, Boston, Mass.) added, and counted for 2 min in a Delta 300 liquid scintillation counter (Searle Analytic, Inc.). Calculations. The mean counts per minute were determined for each triplicate of a given chlamydial or control antigen dilution. Inspection of the control counts per minute indicated frequently two- or threefold variations in counts per minute obtained with adjoining dilutions. The lowest control antigen dilutions (1:8 to 1:16) often had the lowest counts per minute, perhaps because these control antigen concentrations had an inhibitory effect on thymidine incorporation in deoxyribonucleic acid synthesis. To minimize capricious, arbitrary values for the "stimulation index" (SI) (i.e., a given high count per minute value with a certain chlamydial antigen dilution, divided by that same dilution of control antigen counts per minute), we considered it more meaningful to take as the denominator of the fraction the mean counts per minute of the adjoining three highest control antigen values (dilutions of 1:8 to 1:2,048). Similarly, we felt that the stimulating properties of chlamydial antigens (between 1:8 and 1:524,288) would be represented better by not using as numerator a single high counts per minute value, but rather the mean counts per minute of the adjoining three highest values of chlamydial antigens. Based on these considerations, the following formula for calculating the SI was used: SI = mean counts per minute of triplicate cultures with the three chlamydial antigen dilutions (adjoining) giving the highest values/mean counts per minute of triplicate cultures with the three control antigen dilutions (adjoining) giving the highest values. Standard error was calculated for all means derived from multiple tests on the same individual. The sigblood withdrawn from volunteers was diluted with nificance of the difference between means was assessed phosphate-buffered saline (pH 7.2) to make a total of by Student's t test (13). 2.5 to 3.5 times the original volume of blood. Lymphocytes were separated from this diluted blood by the RESULTS Ficoll-Hypaque density gradient separation procedure of antibody and Relation antichlamydial (1). After repeated centrifugation and washing, a final yield of 1 X 106 to 2 x 106 lymphocytes per ml of CMI. Antibody to a chlamydial antigen in a serum dilution of 1:10 or higher was detected in original donor blood was obtained.
414
HANNA ET AL.
INFECT. IMMUN.
30 (57.7%) of the 52 volunteers. This probably reflects the prevalence of asymptomatic chlamydial infection, since only 11 of these 30 volunteers gave a history of signs or symptoms in the past which fitted into the spectrum of clinically manifest chlamydial infections (10, 11). Six individuals with elevated lymphocyte stimulation responses (SI, >3) in repeated tests had no measurable antibody. Four of the six had been exposed to chlamydiae during laboratory work or had experienced symptoms compatible with chlamydial infection. On the other hand, eight volunteers with significant antibody levels had no significant lymphocyte stimulation (SI,
Vol. 23, No. 2
Human Cell-Mediated Immune Responses to Chlamydial Antigens LAVELLE HANNA,l* LINDA SCHMIDT,' MARGARET SHARP,' DANIEL P. STITES AND ERNEST JAWETZ' Departments of Microbiology' and Laboratory Medicine and Medicine,2 University of California Medical Center, San Francisco, California 94143 ,
Received for publication 11 October 1978
A reproducible method was developed to determine the ability of chlamydial antigens to stimulate lymphocytes from volunteers. In tests repeated 4 to 14 times, the cells from a given volunteer gave a relatively narrow range of responses, but there were great differences in the mean response of different volunteers. In the entire group of 52 volunteers, lymphocyte stimulation was significantly associated with the presence of antibody, but in a given individual results of one test did not aid in predicting the results of the other. A majority of persons with either antichlamydial antibody or elevated lymphocyte stimulation, or both, did not have a history of signs or symptoms within a spectrum of chlamydial diseases. This may reflect the great frequency of asymptomatic infection with these organisms. The lymphocytes of some individuals were stimulated to a significantly greater degree by antigens of one chlamydial species (Chlamydia trachomatis or C. psittaci) than by the other. These and other cell-mediated reactions in human chlamydial infections, and their possible medical significance, are under continued study.
Chlamydia trachomatis is an intracellular antibody responses to chlamydiae in humans parasite which greatly resembles gram-negative (5), but, apart from skin test results, virtually no bacteria and which causes important human data are available on human CMI reactions. A disease, particularly of the eye and the respira- few studies on delayed-type hypersensitivity and tory and genital tracts, involving millions of per- other types of CMI responses have been resons. Some characteristics of chlamydial infec- ported from other animal species in which the tion are the often mild or subclinical but exceed- natural course of chlamydial infection appears ingly chronic course, the ability of tetracycline to be different from that encountered in humans to suppress or eradicate acute infection, and the (7, 8, 14). Therefore, we undertook limited studapparent lack of solid immunity to reinfection in ies on CMI in humans with or without known humans (10). exposure to chlamydiae. The first reaction studIn the course of human chlamydial infections, ied was lymphocyte stimulation by chlamydial a variety of antibodies develop, directed against antigens as measured by increased deoxyribogroup-reactive and immunotype-specific anti- nucleic acid synthesis. The purpose of this work gens. However, substantial titers of such anti- was to establish the reproducibility of this test bodies (whether in serum or secretions) seem to with chlamydial antigens, its relation to the presplay virtually no role in ridding the infected ence of antichlamydial antibodies, and its specperson of parasites or in protecting a "cured" ificity with regard to the two chlamydial species. individual against reinfection. It is tempting to MATERIALS AND METHODS believe that in this-as in other-intracellular infection, cell-mediated immune (CMI) reacStudy population. Fifty-two persons (aged 23 to tions are prominent and perhaps medically im- 63 years) were tested 1 to 14 times each. All volunteers portant. It has been suggested that delayed-type were acquainted with the purpose of the investigation hypersensitivity may play a role in aggravating and signed informed consent documents (University of California, San Francisco, Protocol no. 280201; chlamydial disease of the eye by favoring fibro- guidelines human experimentation of the U.S. Deblastic proliferation, which may lead to loss of partment for of Health, Education, and Welfare were vision (3). Conversely, active cell-mediated re- followed). Some volunteers had a history of probable actions may contribute to resistance to chlam- exposure to chlamydial infection; others did not. A ydial infection. majority were unaware of past symptoms or signs There is a large body of information regarding fitting into the known spectrum of clinical chlamydial 412
VOL. 23, 1979
HUMAN CMI RESPONSES TO CHLAMYDIAL ANTIGENS
disease (3, 11). A minority had experienced ocular, pulmonary, or genital tract symptoms compatible with chlamydial infection. In five volunteers there was laboratory evidence (isolation of agent or antibody titer rise) of chlamydial infection 2 to 12 years before the present study. Preparation of antigens. Strain LB-1 (LGV-2) was used as a prototype of C. trachomatis and strain 6BC was used as a prototype of C. psittaci. Both of these had been passed in eggs many times, but particles from infected yolk sac maintained their antigenic identity with type-specific sera tested by immunofluorescence (4). In addition, some tests included two recent isolates of C. psittaci, 562F and 343, both obtained through the courtesy of J. Schachter. Infected yolk sac preparations containing about 108 50% egg lethal doses (106 inclusion-forming units) were diluted 1/50 in Eagle minimum essential medium with 10% fetal calf serum, 50 tig of gentamicin per ml, and 2.5 tig of amphotericin B (Fungizone) per ml. Ten milliliters of this material was placed on confluent sheets of Vero cells in 150-cm2 plastic flasks and adsorbed for 2.5 h at room temperature. The inoculum was then removed, the monolayers were washed twice with Hanks balanced salt solution, and 25 ml of the growth medium containing an additional 2% glucose was added to each flask. The flasks were incubated in a 5% CO2 incubator at 350C for 3 to 4 days. At the end of this incubation period, cells were scraped off with a rubber policeman, pooled, and shaken vigorously with glass beads. The resulting material was diluted onehalf, and 10-ml volumes were used to inoculate monolayers for further Vero cell passage. After two such passages, the disintegrated cells were spun at 2,000 rpm for 5 min at 80C. The sediment was discarded, and the supernatant was spun at 10,000 rpm in a refrigerated RC-5 Sorvall centrifuge for 5 h. The pellets were resuspended in sterile phosphate-buffered saline (pH 7.5) and pooled, and antibiotics were added (penicillin, 300 ug/ml; streptomycin, 500 jig/ml), blended in a Vortex mixer, and stored at 4°C. If bacteriological cultures of this material remained sterile, particles were counted in Giemsa-stained smears (9) and adjusted to 109 to 101'(/ml. These constituted the final antigen preparations, which were stored for up to 15 months at 4°C. Identical preparations were made from normal, uninfected yolk sac materials, subjected to the same number of cell culture passages and to equivalent dilutions, to serve as antigen controls. Antibody titration. The titer of immunotype-specific antibodies was determined by means of an indirect microimmunofluorescence test. The details of the test have been published (4). Antigens of the nine common immunotypes of C. trachomatis were included, as well as 6BC antigen to represent C. psittaci. Lymphocyte stimulation tests. Heparinized
413
Lymphocyte stimulation tests were carried out in 96-well, flat-bottomed microculture plates (Linbro) as previously described (2). To each well was added, separately, 0.1 ml of a lymphocyte suspension of 5 x 106 cells per ml and 0.1 ml of antigen dilution, using a Hamilton repeating dispenser and a gas-tight syringe, rinsed thoroughly between different antigen dilutions. Each test or control antigen was used in triplicate, and each test included a diluted commercial candida antigen control (Allergenic Extract Dermatophytin "O." 1/40; Hollister-Stier Laboratories, Spokane, Wash.) and a medium control. Plates were incubated in a 100% humidified incubator with an atmosphere of 5% CO2 in air at 37°C. After 5 days of incubation, 1 ,uCi in 0.05 ml of [3H]thymidine (specific activity, 6.7 Ci/mmol) was added to each well. After an additional 6 h of incubation, the cultures were harvested onto glass filter paper strips with a MASH II harvester (Microbiological Associates, Bethesda, Md.). The paper was dried overnight, and the disks were punched out, placed in plastic scintillation vials with 2 ml of Liquifluor-toluene (New England Nuclear, Boston, Mass.) added, and counted for 2 min in a Delta 300 liquid scintillation counter (Searle Analytic, Inc.). Calculations. The mean counts per minute were determined for each triplicate of a given chlamydial or control antigen dilution. Inspection of the control counts per minute indicated frequently two- or threefold variations in counts per minute obtained with adjoining dilutions. The lowest control antigen dilutions (1:8 to 1:16) often had the lowest counts per minute, perhaps because these control antigen concentrations had an inhibitory effect on thymidine incorporation in deoxyribonucleic acid synthesis. To minimize capricious, arbitrary values for the "stimulation index" (SI) (i.e., a given high count per minute value with a certain chlamydial antigen dilution, divided by that same dilution of control antigen counts per minute), we considered it more meaningful to take as the denominator of the fraction the mean counts per minute of the adjoining three highest control antigen values (dilutions of 1:8 to 1:2,048). Similarly, we felt that the stimulating properties of chlamydial antigens (between 1:8 and 1:524,288) would be represented better by not using as numerator a single high counts per minute value, but rather the mean counts per minute of the adjoining three highest values of chlamydial antigens. Based on these considerations, the following formula for calculating the SI was used: SI = mean counts per minute of triplicate cultures with the three chlamydial antigen dilutions (adjoining) giving the highest values/mean counts per minute of triplicate cultures with the three control antigen dilutions (adjoining) giving the highest values. Standard error was calculated for all means derived from multiple tests on the same individual. The sigblood withdrawn from volunteers was diluted with nificance of the difference between means was assessed phosphate-buffered saline (pH 7.2) to make a total of by Student's t test (13). 2.5 to 3.5 times the original volume of blood. Lymphocytes were separated from this diluted blood by the RESULTS Ficoll-Hypaque density gradient separation procedure of antibody and Relation antichlamydial (1). After repeated centrifugation and washing, a final yield of 1 X 106 to 2 x 106 lymphocytes per ml of CMI. Antibody to a chlamydial antigen in a serum dilution of 1:10 or higher was detected in original donor blood was obtained.
414
HANNA ET AL.
INFECT. IMMUN.
30 (57.7%) of the 52 volunteers. This probably reflects the prevalence of asymptomatic chlamydial infection, since only 11 of these 30 volunteers gave a history of signs or symptoms in the past which fitted into the spectrum of clinically manifest chlamydial infections (10, 11). Six individuals with elevated lymphocyte stimulation responses (SI, >3) in repeated tests had no measurable antibody. Four of the six had been exposed to chlamydiae during laboratory work or had experienced symptoms compatible with chlamydial infection. On the other hand, eight volunteers with significant antibody levels had no significant lymphocyte stimulation (SI,
