- Email: [email protected]
Detection of serum antibodies against Chlamydia pneumoniae by ELISA
ELSEVIER
IMMUNOLOGY AND MEDICAL MICROBIOLOGY FEMS Immunology
and Medical Microbiology
14 (1996) 179-183
Detection of serum antibodies against Chkzrnydia pneumoniae by ELISA Kei Numazaki
*, Tadashi Ikebe, Shunzo Chiba
Department of P,ediatn’cs, Sapporo Medical University School of Medicine, XI W.16 Chuo-Ku, Sapporo 060, Japan Received 8 February 1996; revised 19 March 1996; accepted 19 March 1996
Abstract Chlamydia pneumoniae causes pneumonia and other respiratory infections in children, adolescents and adults. We tried to evaluate the diagnostic value of detection of serum antibodies by ELISA for C. pneumoniae infections in Japanese children. Serum IgG, IgA and IgM antibodies to C. pneumoniae were determined by the microimmunofluorescence (MIF) test. Serum IgG and IgA antibodies were also determined by ELISA test kits. Results obtained by ELISA were compared with those obtained by MIF test. IgG antibody to C. pneumoniae was detected in 135 (39.5%) by ELISA and in 125 (36.5%) by MIF out of 342 sera from Japanese infants and children without respiratory infections (aged from 2 months old to 15 years old). IgA antibody to C. pneumoniae was detected in 129 (37.7%) by ELISA and in 117 (34.2%) by MIF out of 342 sera tested. Of 342 specimens 113 were IgG-positive by ELISA and MIF (sensitivity: 90.4%, specificity: 89.9%, r = 0.853). Of 342 sera 28 had IgG antibody titers of 1:256 and none had titers 1:512 or higher by MIF. Of 28 infants and children a total of nine were less than 4 years of age. On the other hand, of 342 specimens 99 were IgA-positive by ELISA and MIF (sensitivity: 84.6%, specificity: 86.7%, r = 0.769). Of 342 sera 16 had IgA antibody titers of 1:256 or higher by MIF. Of 16 infants and children, ten were less than 4 years of age. ELISA had excellent sensitivity and specificity relative to MIF test for detection of IgC and IgA antibodies to C. pneumoniae. It was suggested that C. pneumoniae infection in Japanese infants and children under 4 years of age was not infrequent. Keywords: Chlamydia pneumoniae; Serum antibody;
Respiratory-tract
1. Introduction
ELISA
niae.
Chlamydia pneumoniae is a common etiological agent in acute respiratory infections, including pneumonia, bronchitis, pharyngitis, and sinusitis [l]. A total of between 5% to 10% of all cases of adult pneumonia and bronchitis are caused by C. pneumo-
* Corresponding author. Fax: +81 (11) 611 0352.
infection;
Tel:
0928-8244/96/$15.00 Copyright PII $0928-8244(96)000:!9-6
+81
(11) 611 2111 ext. 3413;
0 1996 Federation
of European
More than half of the adults in the United States and in many other countries including Japan have IgG antibodies to C. pneumoniae. Although the elevated serum antibodies to C. pneumoniae and the presence of circulating Chlamydia-specific immune complexes have been found in several C. pneumoniae infections, the pathogenesis of C. pneumoniae infections in childhood has yet to be clarified. Seroepidemiologic studies in children have shown an increasing prevalence of antichlamydial antibody with age, indicating the presence of active C. pneuMicrobiological
Societies. Published
by Elsevier Science B.V.
180
Ke Numazaki et al. / FEMS Immunology and Medical Microbiology 14 (1996) 179-183
moniae infections in childhood [l]. In our previous
study, 11 of 223 (4.9%) Japanese infants and children with pneumonia had serum IgM antibodies to C. pneumoniae determined by using the microimmunofluorescence (MIF) test using TW-183 strain as antigen [2]. As the prevalence of serum antibodies against C. pneumoniae is relatively high, a serological test for the diagnosis of infection was thought to be of limited availability. However, in respiratory chlamydial infections in childhood the diagnosis can be performed by serological examinations. Recently, enzyme immunoassays have been reported to measure antibodies against defined epitopes of the chlamydial major outer membrane protein (MOMP) [31. In this study we tried to evaluate the diagnostic value of detection of serum antibodies by ELISA for C. pneumoniae infections in Japanese childhood. Results obtained by ELISA were compared with those obtained by MIF.
2. Materials and methods 2.1. Study populations Serum specimens were obtained from 342 infants and children (aged between 2 months and 15 years) without respiratory tract infections between January 1990 to January 1995. Over the same period serum specimens of acute phase were obtained from 151 patients with pneumonia, 28 with bronchitis, and from 10 suffering from attacks of bronchial asthma (aged 0 months to 15 years). 2.2. Serological study Serum IgG, IgA and IgM antibodies to C. pneuwere determined by the microimmunofluorescence (MIF) test using formalinized elementary bodies (EBs) of C. pneumoniae TW-183, TWAR strain as antigen (provided from Washington Research Foundation, Seattle, WA). Chlamydial IgG, IgA, and IgM antibodies were measured by using goat fluorescein isothiocyanate-conjugated antihuman IgG, IgA and IgM (Cappel Laboratories, Malvem, PA). Diagnosis of acute C. pneumoniae moniae
infection was obtained serologically based on the criteria described previously [1,2]. Serum antibody titers of 1: 16 or higher with MIF test were considered as positive in this study. Serum IgG and IgA antibodies to C. pneumoniae were also determined by ELISA test kits (CPlOOOChlamydia 9504A) provided by Hitachi Chemical Co., Ltd., Tokyo, Japan. Briefly, chlamydial outer membrane complex (COMC) was isolated from purified elementary bodies (EBs) of C. pneumoniae YK-41 strain [4], originally isolated by Dr. Yasuo Kanamoto, Hiroshima Prefectual Institute of Public Health, Hiroshima, Japan. C. pneumoniae YK-41 strain was grown in cycloheximide-treated HL cells. Infected cells were harvested 96 h after inoculation. Harvested cell suspensions were pooled and disrupted by sonication. Elementary bodies (EBs) were separated and purified by two-layer cushioning centrifugation (30% Urografin and 50% sucrose), followed by continuous Urografin gradient centrifugation (35-50%). COMC was isolated from purified EBs, which were treated with 10 mM sodium phosphate (Sarkosyl buffer) containing 2% Sarkosyl (Nlauroylsarcosine sodium salt), 1.5 n-&l EDTA and 0.14 mM NaCl as described previously [5]. After ultrasonic disruption, centrifugation and washing with PBS, the pellet was resuspended in 20 mM sodium phosphate buffer containing 10 mM MgCl, with 2.5 mg/ml DNase and RNase. After centrifugation, residual DNase and RNase were eliminated from the pellet and COMC antigen was obtained as the Sarkosyl-insoluble fraction. Wells of microtiter plates were coated with 0.05 mg/well COMC antigens, which had been diluted in 10 mM sodium phosphate, pH 6.4. Fifty ~1 of diluted (1:20) serum was reacted with COMC, followed by reaction with alkaline phosphatase-labelled antihuman IgG and IgA antibodies. The substrate ( p-nitrophenyl phosphate) was then added to each well and the OD was measured at 405 nm. In this way, it was possible to determine the presence or absence of anti-chlamydial antibodies in the test of screened subjects. A 50 ~1 aliquot of a teat test specimen was placed in a test tube and diluted with 1000 ~1 of diluent; this was employed as the IgA assay. Next, 100 ~1 of this IgA sample was placed in a separate test tube and 900 ~1 of the diluent was added; this was employed as the IgG assay sample. A negative control, a positive control
Ke Numazaki et al. / FEMS Immunology and Medical Microbiology I4 (1996) 179-I 83
and the assay sample, each in a volume of 100 ~1, and negative controls were used to calculate the cut-off values as reported previously [6] (cut-off value was determined as 0.200 OD>. In this system no cross-reactivities were proved between antibodies to C. trachomatis or C. psittaci.
3. Results Of 342 sera from infants and children without respiratory tract infections, 153 (44.7%) were IgG and/or IgA antibody-positive by ELBA. IgG antibody to C. pneumonicre was detected in 135 (39.5%) by ELBA and in 12:5 (36.5%) by MIP out of 342 sera tested (Table 1). The IgG antibody detection rate increased in subjlects between ages of 7 and 15 years. IgA antibody to C. pneumoniae was detected in 129 (37.7%) by E.LISA and in 117 (34.2%) by MIP out of 342 sera tested (Table 2). Of 176 infants and children less than 4 years of age, 63 (35.8%) were IgG and/or IgA and 51 (29.0%) were IgG antibody-positive by ELISA. Of 151 infants and children with pneumonia 30 (19.9%) were diagnosed
Table 1 Prevalence of IgG antibody to C. pneumoniae in Japanese infants and children without respiratory tract infections by MIF and by ELBA Age O-5 months 6-11 months 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years 10 years 11 years 12 years 13 years 14 years 15 years Total
No. of subjects
27 22 59 44 24 25 15 13 11 9 14 13 10 16 14 10 16 342
No. of positive by MIF (o/o) 15 2 9 14 7 8 5 4 5 4 7 7 5 9 9 6 9
(55.6) (9.1) (15.2) (31.7) (29.1) (32.0) (33.3) (30.7) (45.5) (44.4) (50.0) (53.8) (50.0) (56.3) (64.3) (60.01 (56.3)
125 (36.5)
181
Table 2 Prevalence of IgA antibody to C. pneunwnfae in Japanese infants and children without respiratory tract infections by MIF and by ELISA No. of subjects
Age
No. of positives by MIF (%)
O-5 months 6-l 1 months 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years 10 years 11 years 12 years 13 years 14 years 15 years Total
No. of positive by ELISA (46)
27 22 59 44 24 25 15 13 11 9 14 13 10 16 14 10 16
1 (3.7) 5 (22.7) 11 (18.6) 22 (50.0) 7 (29.1) 8 (32.0) 2 (13.3) 4 (30.8) 5 (45.5) 2 (22.2) 7 (50.0) 5 (38.5) 5 (50.0) 9 (56.3) 9 (64.3) 6 (60.0) 9 (56.3)
2 (7.4) 5 (22.7) 14 (23.7) 23 (52.3) 6 (25.0) ll(44.0) 2 (13.3) 4 (30.8) 6 (54.5) 2 (22.2) 7 (50.0) 6 (46.2) 6 (60.0) 12 (75.0) 8 (57.1) 6 (60.0) 9 (56.3)
342
117 (34.2)
129 (37.7)
as C. pneumoniae pneumonia serologically with MIP test. Pneumonia due to C. pneumoniae in 151 infants and children aged 6 months to 15 years are summarized in Table 3. Of 30 infants and children with pneumonia, 9 were less than 4 years of age. Of 151
No. of positive by ELBA (%) 16 2 10 15 8 8 5 5 5 5 8 7 6 10 9 6 10
(59.3) (9.1) (16.9) (34.1) (33.3) (32.0) (33.3) (38.5) (45.5) (55.5) (57.1) (53.8) (60.0) (62.5) (64.3) (60.0) (62.5)
135 (39.5)
Table 3 Incidence children
of C. pneumoniae pneumonia
Age O-5 months 6- 11 months 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years 10 years Total
No. of subjects
24 11 25 6 14 16 10 16 6 7 4 12 151
in Japanese
infants and
No. of C. pneumoniae pneumonia 0
30 (19.9%)
182
Ke Numazaki et al. / FEMS Immunology and Medical Microbiology
Table 4 Detection of IgG and IgA antibodies Respiratory disorders
tract
4. Discussion to C. pneumoniae by ELISA
No. of tested
No. of positive with IgG and IgA antibodies (%)
Pneumonia Bronchitis Bronchial asthma
1.51 28 10
32 (21.2) 1 (3.6) 6 (60.0)
Control
342
110 (32.2)
sera from patients with pneumonia 32 (21.2%) were both IgG and IgA antibody-positive by ELBA (Table 4). Of 32 infants and children with IgG and IgA-positive by ELBA, 28 were diagnosed as acute C. pneumoniae infection by MIF test. Of 10 patients with attack of bronchial asthma, 6 had IgG and IgA antibodies to C. pneumoniae. One patient with bronchitis had IgG and IgA antibodies. Of 342 sera from infants and children without respiratory tract infections (considered as control), 110 (32.2%) were IgG and IgA antibody-positive. The results of detection IgG and IgA antibodies detection by ELBA and MIF were compared each other in 342 sera from infants and children without respiratory tract infections. Of 342 specimens 113 were IgG-positive by ELISA and MIF (Table 5, r = 0.853). Of 342 sera 28 had IgG antibody titers of 1:256 and none had titers 1:512 or higher by MIF. Of 28 infants and children 9 were less than 4 years of age. On the other hand, of 342 specimens 99 were IgA-positive by ELISA and MIF (Table 6, r = 0.769). Of 342 sera 16 had IgA antibody titers of 1:256 or higher by MIF. Of 16 infants and children 10 were less than 4 years of age.
Table 5 Comparison of ELISA with MIF for the detection antibody to C. pneumoniae
Total Sensitivity:
Positive Negative
of serum IgG
C. pneumoniae usually causes pneumonia and other respiratory infections in children, adolescents and adults [7]. Pneumonia due to C. pneumoniue frequently occurs in children more than 4 years of age. Chlamydial infections induce inflammatory changes that might cause modulation of secretion of some cytokines [S]. As cultures of lung biopsies from infants with chlamydial pneumonia have frequently failed to yield the organism, immunological reactions of the host to these agents appear to be more important than the direct effects of C. pneumoniae in the pathogenesis of respiratory infections [9-121. It is also likely that specific cellular interactions as well as secretion of IFN-7 and other cytokines are important in the pathogenesis of C. pneumoniae pneumonia [13]. Difference of clinical features of C. pneumoniae infections in infants and children by age may be also due to these differences of immunological reactions to the organism. It was suggested that C. pneumoniae infection was infrequent in pre-school-age children, was frequent in school-age children, and continued to be detectable throughout adult life. Grayston [14] reported that acute C. pneumoniae infection occurred in less than 1% of children aged from 7 months to 4 years of age and in 6% of 5- to lCyear-olds. From the results of the present study of MIF and ELISA it was also suggested that C. pneumoniae infection in Japanese infants and children less than 4 years of age was not infrequent. Although pneumonia and bronchitis are the most frequently recognized illness, asymptomatic or unrecognized illness are the most common result of C. pneumoniae infection as shown in this study. C. pneumoniae infection in childhood induces a time-limited antibody response and most Table 6 Comparison of ELISA with MIF for the detection antibody to C. pneumoniae
Total
MIF
ELISA
MIF
Positive
Negative
113 12
22 195
135 207
ELISA
125
217
342
Total
90.4%, Specificity:
14 (1996) 179-183
89.9%.
Total
Positive
Sensitivity:
Positive Negative
of serum IgA
Negative
99 18
30 195
129 213
117
225
342
84.6%, Specificity:
86.7%.
Lye Numazaki et al./ FEMS Immunology and Medical Microbiology
people are reinfected throughout life. Patients with respiratory infections who had antibodies against C. pneumoniae exhibited a change of titers between acute and convalescent sera. Determination of serum antibodies by ELBA is suitable for identifying C. pneumoniae as a cause of respiratory tract infections. In general IgG antibody titers of 1: 16 or higher were regarded as evidence of past infection. Kanamoto et al. [15] previously reported the prevalence of IgG antibodies to C. pneumoniae of 1:32 or higher by MIF in Japanese individuals. However, Grayston et al. [14] suggested that IgG antibody titers of 1:8 to 1:256 by MIF also indicated past infection. The difference in the detection rate of IgG and IgM antibodies to C. pneumoniae in Japanese infants and children might be as a result of these backgrounds. A slight difference in peptide profile was also seen between C. pneumoniae YK-41 and TW-183, TWAR strains. The difference in antibody titer between these two strains had not been studied using enough serum samples in a previous study [4]. Cross-reactivity as antigens been these two strains was not examined. A:ntigenic variations between C. pneumoniae strains isolated in worldwide base also need be studied. ELBA had excellent sensitivity and specificity relative to MIF test for detection of IgG and IgA antibodies against C. pneumoniae as shown this study. However, combination of IgG and IgA antibody detection by ELBA was not possible in applying the diagnosis of chlamydial respiratory-tract infections as shown in this study. Further studies are needed to refine ELISA assay for detection of specific IgM antibody to C. pneumoniae.
Acknowledgements We thank Kiyotaka Kawagoe and Toshihide Morikawa, Hitachi Chemical Co. Ltd., for technical assistance and for providing test kits.
References [l] Grayston, J.T. (1992) Infections caused by Chlamydia pneumoniae strain TWAR. Clin. Infect. Dis. 15, 757-763.
14 (1996) 179-183
183
[2] Numazaki, K., Chiba, S. and Umetsu, M. (1992) Detection of IgM antibodies to Chlamydia trachomaris, Chlamydiu pneumonia, and Chlamydia psittaci from Japanese infants and children with pneumonia. In vivo 6, 601-604. [3] Jones, H.M., Schachter, J. and Stephens, R.S. (1992) Evaluation of the humoral immune response in trachoma to Chlamydia trachomatis major outer membrane proteins by sequence-defined immunoassay. J. Infect. Dis. 166,915-919. [4] Kanamoto, Y., Iijima, Y., Miyashita, N., Matsumoto, A. and Sakano, T. (1993) Antigenic characterization of ChZamydia pneumoniae isolated in Hiroshima, Japan. Microbial. Immunol. 37, 495-498. [S] Caldwell, H.D., Kromhout, J. and Schachter, J. (1981) F’urification and partial characterization of the major outer membrane protein of Chlamydia rrachomads. Infect. Immun. 3 1, 1161-l 176. [6] Brade, L., Brunnemann, H., Ernst, M., Fu, Y., Holst, O., Korsma, P., NZher, H., Persson, K. and Brade, H. (1994) Occurrence of antibodies against chlamydial lipopolysaccharide in human sera as measured by ELISA using an artificial glycoconjugate antigen. FEMS Immunol. Med. Microbial. 8, 27-42. [71 Kuo, C.-C., Jackson, L.A., Campbell, L.A. and Grayston, J.T. (1995) Chlamydia pneumoniae (TWAR). Clin. Microbio. Rev. 8, 451-461. Bl Rothermel, C.D., Schachter, J., Lavrich, P., Lipsits, E.C. and Francus, T. (1989) Chlamydia trachornatis-induced production of interleukin-1 by human monocytes. Infect. Immun. 57, 2705-2711. t91 Surcel, H.M., Syrjala, H., Leinonen, M., Saikku, P. and Herva, E. (1993) Cell-mediated immunity to Chlamydia pneumoniae measured as lymphocyte blast transformation in vitro. Infect. Immun. 61, 2196-2199. M.R., Smith, P. [lOI Nakajo, M.N., Roblin, P.M., H-erschlag, and Nowakowski, M. (1990) Chlamydicidal activity of human alveolar macrophages. Infect. Immun. 58, 3640-3644. [ill Numazaki, K., Suzuki, K. and Chiba, S. (1995) Replication of Chlamydia trachomaris and C. pneumoniae in the human monocytic cell line U-937. J. Med. Microbial. 42, 191-195. WI Numazaki, K., Wainberg, M.A. and McDonald, J. (1989) Chlamydia trachomatis infections in infants. Can. Med. Assoc. J. 140, 615-622. [131 Numazaki, K. and Chiba, S. (1996) Serum g-a-interferon in patients with penumonia caused by Chlumydia pneumoniae. Pediatr. Infect. Dis. J. 15, 174-175. [141 Grayston, J.T. (1994) Chlamydia pneumoniae (TWAR) infection in children. Pediatr. Infect. Dis. J. 13, 675-685. [15] Kanamoto, Y., Ouchi, K., Mizui, M., Ushio, M. and Usui, T. (1991) Prevalence of antibodies to Chlamydia pneumoniae TWAR in Japan. J. Clin. Microbial 29, 816-818.
IMMUNOLOGY AND MEDICAL MICROBIOLOGY FEMS Immunology
and Medical Microbiology
14 (1996) 179-183
Detection of serum antibodies against Chkzrnydia pneumoniae by ELISA Kei Numazaki
*, Tadashi Ikebe, Shunzo Chiba
Department of P,ediatn’cs, Sapporo Medical University School of Medicine, XI W.16 Chuo-Ku, Sapporo 060, Japan Received 8 February 1996; revised 19 March 1996; accepted 19 March 1996
Abstract Chlamydia pneumoniae causes pneumonia and other respiratory infections in children, adolescents and adults. We tried to evaluate the diagnostic value of detection of serum antibodies by ELISA for C. pneumoniae infections in Japanese children. Serum IgG, IgA and IgM antibodies to C. pneumoniae were determined by the microimmunofluorescence (MIF) test. Serum IgG and IgA antibodies were also determined by ELISA test kits. Results obtained by ELISA were compared with those obtained by MIF test. IgG antibody to C. pneumoniae was detected in 135 (39.5%) by ELISA and in 125 (36.5%) by MIF out of 342 sera from Japanese infants and children without respiratory infections (aged from 2 months old to 15 years old). IgA antibody to C. pneumoniae was detected in 129 (37.7%) by ELISA and in 117 (34.2%) by MIF out of 342 sera tested. Of 342 specimens 113 were IgG-positive by ELISA and MIF (sensitivity: 90.4%, specificity: 89.9%, r = 0.853). Of 342 sera 28 had IgG antibody titers of 1:256 and none had titers 1:512 or higher by MIF. Of 28 infants and children a total of nine were less than 4 years of age. On the other hand, of 342 specimens 99 were IgA-positive by ELISA and MIF (sensitivity: 84.6%, specificity: 86.7%, r = 0.769). Of 342 sera 16 had IgA antibody titers of 1:256 or higher by MIF. Of 16 infants and children, ten were less than 4 years of age. ELISA had excellent sensitivity and specificity relative to MIF test for detection of IgC and IgA antibodies to C. pneumoniae. It was suggested that C. pneumoniae infection in Japanese infants and children under 4 years of age was not infrequent. Keywords: Chlamydia pneumoniae; Serum antibody;
Respiratory-tract
1. Introduction
ELISA
niae.
Chlamydia pneumoniae is a common etiological agent in acute respiratory infections, including pneumonia, bronchitis, pharyngitis, and sinusitis [l]. A total of between 5% to 10% of all cases of adult pneumonia and bronchitis are caused by C. pneumo-
* Corresponding author. Fax: +81 (11) 611 0352.
infection;
Tel:
0928-8244/96/$15.00 Copyright PII $0928-8244(96)000:!9-6
+81
(11) 611 2111 ext. 3413;
0 1996 Federation
of European
More than half of the adults in the United States and in many other countries including Japan have IgG antibodies to C. pneumoniae. Although the elevated serum antibodies to C. pneumoniae and the presence of circulating Chlamydia-specific immune complexes have been found in several C. pneumoniae infections, the pathogenesis of C. pneumoniae infections in childhood has yet to be clarified. Seroepidemiologic studies in children have shown an increasing prevalence of antichlamydial antibody with age, indicating the presence of active C. pneuMicrobiological
Societies. Published
by Elsevier Science B.V.
180
Ke Numazaki et al. / FEMS Immunology and Medical Microbiology 14 (1996) 179-183
moniae infections in childhood [l]. In our previous
study, 11 of 223 (4.9%) Japanese infants and children with pneumonia had serum IgM antibodies to C. pneumoniae determined by using the microimmunofluorescence (MIF) test using TW-183 strain as antigen [2]. As the prevalence of serum antibodies against C. pneumoniae is relatively high, a serological test for the diagnosis of infection was thought to be of limited availability. However, in respiratory chlamydial infections in childhood the diagnosis can be performed by serological examinations. Recently, enzyme immunoassays have been reported to measure antibodies against defined epitopes of the chlamydial major outer membrane protein (MOMP) [31. In this study we tried to evaluate the diagnostic value of detection of serum antibodies by ELISA for C. pneumoniae infections in Japanese childhood. Results obtained by ELISA were compared with those obtained by MIF.
2. Materials and methods 2.1. Study populations Serum specimens were obtained from 342 infants and children (aged between 2 months and 15 years) without respiratory tract infections between January 1990 to January 1995. Over the same period serum specimens of acute phase were obtained from 151 patients with pneumonia, 28 with bronchitis, and from 10 suffering from attacks of bronchial asthma (aged 0 months to 15 years). 2.2. Serological study Serum IgG, IgA and IgM antibodies to C. pneuwere determined by the microimmunofluorescence (MIF) test using formalinized elementary bodies (EBs) of C. pneumoniae TW-183, TWAR strain as antigen (provided from Washington Research Foundation, Seattle, WA). Chlamydial IgG, IgA, and IgM antibodies were measured by using goat fluorescein isothiocyanate-conjugated antihuman IgG, IgA and IgM (Cappel Laboratories, Malvem, PA). Diagnosis of acute C. pneumoniae moniae
infection was obtained serologically based on the criteria described previously [1,2]. Serum antibody titers of 1: 16 or higher with MIF test were considered as positive in this study. Serum IgG and IgA antibodies to C. pneumoniae were also determined by ELISA test kits (CPlOOOChlamydia 9504A) provided by Hitachi Chemical Co., Ltd., Tokyo, Japan. Briefly, chlamydial outer membrane complex (COMC) was isolated from purified elementary bodies (EBs) of C. pneumoniae YK-41 strain [4], originally isolated by Dr. Yasuo Kanamoto, Hiroshima Prefectual Institute of Public Health, Hiroshima, Japan. C. pneumoniae YK-41 strain was grown in cycloheximide-treated HL cells. Infected cells were harvested 96 h after inoculation. Harvested cell suspensions were pooled and disrupted by sonication. Elementary bodies (EBs) were separated and purified by two-layer cushioning centrifugation (30% Urografin and 50% sucrose), followed by continuous Urografin gradient centrifugation (35-50%). COMC was isolated from purified EBs, which were treated with 10 mM sodium phosphate (Sarkosyl buffer) containing 2% Sarkosyl (Nlauroylsarcosine sodium salt), 1.5 n-&l EDTA and 0.14 mM NaCl as described previously [5]. After ultrasonic disruption, centrifugation and washing with PBS, the pellet was resuspended in 20 mM sodium phosphate buffer containing 10 mM MgCl, with 2.5 mg/ml DNase and RNase. After centrifugation, residual DNase and RNase were eliminated from the pellet and COMC antigen was obtained as the Sarkosyl-insoluble fraction. Wells of microtiter plates were coated with 0.05 mg/well COMC antigens, which had been diluted in 10 mM sodium phosphate, pH 6.4. Fifty ~1 of diluted (1:20) serum was reacted with COMC, followed by reaction with alkaline phosphatase-labelled antihuman IgG and IgA antibodies. The substrate ( p-nitrophenyl phosphate) was then added to each well and the OD was measured at 405 nm. In this way, it was possible to determine the presence or absence of anti-chlamydial antibodies in the test of screened subjects. A 50 ~1 aliquot of a teat test specimen was placed in a test tube and diluted with 1000 ~1 of diluent; this was employed as the IgA assay. Next, 100 ~1 of this IgA sample was placed in a separate test tube and 900 ~1 of the diluent was added; this was employed as the IgG assay sample. A negative control, a positive control
Ke Numazaki et al. / FEMS Immunology and Medical Microbiology I4 (1996) 179-I 83
and the assay sample, each in a volume of 100 ~1, and negative controls were used to calculate the cut-off values as reported previously [6] (cut-off value was determined as 0.200 OD>. In this system no cross-reactivities were proved between antibodies to C. trachomatis or C. psittaci.
3. Results Of 342 sera from infants and children without respiratory tract infections, 153 (44.7%) were IgG and/or IgA antibody-positive by ELBA. IgG antibody to C. pneumonicre was detected in 135 (39.5%) by ELBA and in 12:5 (36.5%) by MIP out of 342 sera tested (Table 1). The IgG antibody detection rate increased in subjlects between ages of 7 and 15 years. IgA antibody to C. pneumoniae was detected in 129 (37.7%) by E.LISA and in 117 (34.2%) by MIP out of 342 sera tested (Table 2). Of 176 infants and children less than 4 years of age, 63 (35.8%) were IgG and/or IgA and 51 (29.0%) were IgG antibody-positive by ELISA. Of 151 infants and children with pneumonia 30 (19.9%) were diagnosed
Table 1 Prevalence of IgG antibody to C. pneumoniae in Japanese infants and children without respiratory tract infections by MIF and by ELBA Age O-5 months 6-11 months 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years 10 years 11 years 12 years 13 years 14 years 15 years Total
No. of subjects
27 22 59 44 24 25 15 13 11 9 14 13 10 16 14 10 16 342
No. of positive by MIF (o/o) 15 2 9 14 7 8 5 4 5 4 7 7 5 9 9 6 9
(55.6) (9.1) (15.2) (31.7) (29.1) (32.0) (33.3) (30.7) (45.5) (44.4) (50.0) (53.8) (50.0) (56.3) (64.3) (60.01 (56.3)
125 (36.5)
181
Table 2 Prevalence of IgA antibody to C. pneunwnfae in Japanese infants and children without respiratory tract infections by MIF and by ELISA No. of subjects
Age
No. of positives by MIF (%)
O-5 months 6-l 1 months 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years 10 years 11 years 12 years 13 years 14 years 15 years Total
No. of positive by ELISA (46)
27 22 59 44 24 25 15 13 11 9 14 13 10 16 14 10 16
1 (3.7) 5 (22.7) 11 (18.6) 22 (50.0) 7 (29.1) 8 (32.0) 2 (13.3) 4 (30.8) 5 (45.5) 2 (22.2) 7 (50.0) 5 (38.5) 5 (50.0) 9 (56.3) 9 (64.3) 6 (60.0) 9 (56.3)
2 (7.4) 5 (22.7) 14 (23.7) 23 (52.3) 6 (25.0) ll(44.0) 2 (13.3) 4 (30.8) 6 (54.5) 2 (22.2) 7 (50.0) 6 (46.2) 6 (60.0) 12 (75.0) 8 (57.1) 6 (60.0) 9 (56.3)
342
117 (34.2)
129 (37.7)
as C. pneumoniae pneumonia serologically with MIP test. Pneumonia due to C. pneumoniae in 151 infants and children aged 6 months to 15 years are summarized in Table 3. Of 30 infants and children with pneumonia, 9 were less than 4 years of age. Of 151
No. of positive by ELBA (%) 16 2 10 15 8 8 5 5 5 5 8 7 6 10 9 6 10
(59.3) (9.1) (16.9) (34.1) (33.3) (32.0) (33.3) (38.5) (45.5) (55.5) (57.1) (53.8) (60.0) (62.5) (64.3) (60.0) (62.5)
135 (39.5)
Table 3 Incidence children
of C. pneumoniae pneumonia
Age O-5 months 6- 11 months 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years 10 years Total
No. of subjects
24 11 25 6 14 16 10 16 6 7 4 12 151
in Japanese
infants and
No. of C. pneumoniae pneumonia 0
30 (19.9%)
182
Ke Numazaki et al. / FEMS Immunology and Medical Microbiology
Table 4 Detection of IgG and IgA antibodies Respiratory disorders
tract
4. Discussion to C. pneumoniae by ELISA
No. of tested
No. of positive with IgG and IgA antibodies (%)
Pneumonia Bronchitis Bronchial asthma
1.51 28 10
32 (21.2) 1 (3.6) 6 (60.0)
Control
342
110 (32.2)
sera from patients with pneumonia 32 (21.2%) were both IgG and IgA antibody-positive by ELBA (Table 4). Of 32 infants and children with IgG and IgA-positive by ELBA, 28 were diagnosed as acute C. pneumoniae infection by MIF test. Of 10 patients with attack of bronchial asthma, 6 had IgG and IgA antibodies to C. pneumoniae. One patient with bronchitis had IgG and IgA antibodies. Of 342 sera from infants and children without respiratory tract infections (considered as control), 110 (32.2%) were IgG and IgA antibody-positive. The results of detection IgG and IgA antibodies detection by ELBA and MIF were compared each other in 342 sera from infants and children without respiratory tract infections. Of 342 specimens 113 were IgG-positive by ELISA and MIF (Table 5, r = 0.853). Of 342 sera 28 had IgG antibody titers of 1:256 and none had titers 1:512 or higher by MIF. Of 28 infants and children 9 were less than 4 years of age. On the other hand, of 342 specimens 99 were IgA-positive by ELISA and MIF (Table 6, r = 0.769). Of 342 sera 16 had IgA antibody titers of 1:256 or higher by MIF. Of 16 infants and children 10 were less than 4 years of age.
Table 5 Comparison of ELISA with MIF for the detection antibody to C. pneumoniae
Total Sensitivity:
Positive Negative
of serum IgG
C. pneumoniae usually causes pneumonia and other respiratory infections in children, adolescents and adults [7]. Pneumonia due to C. pneumoniue frequently occurs in children more than 4 years of age. Chlamydial infections induce inflammatory changes that might cause modulation of secretion of some cytokines [S]. As cultures of lung biopsies from infants with chlamydial pneumonia have frequently failed to yield the organism, immunological reactions of the host to these agents appear to be more important than the direct effects of C. pneumoniae in the pathogenesis of respiratory infections [9-121. It is also likely that specific cellular interactions as well as secretion of IFN-7 and other cytokines are important in the pathogenesis of C. pneumoniae pneumonia [13]. Difference of clinical features of C. pneumoniae infections in infants and children by age may be also due to these differences of immunological reactions to the organism. It was suggested that C. pneumoniae infection was infrequent in pre-school-age children, was frequent in school-age children, and continued to be detectable throughout adult life. Grayston [14] reported that acute C. pneumoniae infection occurred in less than 1% of children aged from 7 months to 4 years of age and in 6% of 5- to lCyear-olds. From the results of the present study of MIF and ELISA it was also suggested that C. pneumoniae infection in Japanese infants and children less than 4 years of age was not infrequent. Although pneumonia and bronchitis are the most frequently recognized illness, asymptomatic or unrecognized illness are the most common result of C. pneumoniae infection as shown in this study. C. pneumoniae infection in childhood induces a time-limited antibody response and most Table 6 Comparison of ELISA with MIF for the detection antibody to C. pneumoniae
Total
MIF
ELISA
MIF
Positive
Negative
113 12
22 195
135 207
ELISA
125
217
342
Total
90.4%, Specificity:
14 (1996) 179-183
89.9%.
Total
Positive
Sensitivity:
Positive Negative
of serum IgA
Negative
99 18
30 195
129 213
117
225
342
84.6%, Specificity:
86.7%.
Lye Numazaki et al./ FEMS Immunology and Medical Microbiology
people are reinfected throughout life. Patients with respiratory infections who had antibodies against C. pneumoniae exhibited a change of titers between acute and convalescent sera. Determination of serum antibodies by ELBA is suitable for identifying C. pneumoniae as a cause of respiratory tract infections. In general IgG antibody titers of 1: 16 or higher were regarded as evidence of past infection. Kanamoto et al. [15] previously reported the prevalence of IgG antibodies to C. pneumoniae of 1:32 or higher by MIF in Japanese individuals. However, Grayston et al. [14] suggested that IgG antibody titers of 1:8 to 1:256 by MIF also indicated past infection. The difference in the detection rate of IgG and IgM antibodies to C. pneumoniae in Japanese infants and children might be as a result of these backgrounds. A slight difference in peptide profile was also seen between C. pneumoniae YK-41 and TW-183, TWAR strains. The difference in antibody titer between these two strains had not been studied using enough serum samples in a previous study [4]. Cross-reactivity as antigens been these two strains was not examined. A:ntigenic variations between C. pneumoniae strains isolated in worldwide base also need be studied. ELBA had excellent sensitivity and specificity relative to MIF test for detection of IgG and IgA antibodies against C. pneumoniae as shown this study. However, combination of IgG and IgA antibody detection by ELBA was not possible in applying the diagnosis of chlamydial respiratory-tract infections as shown in this study. Further studies are needed to refine ELISA assay for detection of specific IgM antibody to C. pneumoniae.
Acknowledgements We thank Kiyotaka Kawagoe and Toshihide Morikawa, Hitachi Chemical Co. Ltd., for technical assistance and for providing test kits.
References [l] Grayston, J.T. (1992) Infections caused by Chlamydia pneumoniae strain TWAR. Clin. Infect. Dis. 15, 757-763.
14 (1996) 179-183
183
[2] Numazaki, K., Chiba, S. and Umetsu, M. (1992) Detection of IgM antibodies to Chlamydia trachomaris, Chlamydiu pneumonia, and Chlamydia psittaci from Japanese infants and children with pneumonia. In vivo 6, 601-604. [3] Jones, H.M., Schachter, J. and Stephens, R.S. (1992) Evaluation of the humoral immune response in trachoma to Chlamydia trachomatis major outer membrane proteins by sequence-defined immunoassay. J. Infect. Dis. 166,915-919. [4] Kanamoto, Y., Iijima, Y., Miyashita, N., Matsumoto, A. and Sakano, T. (1993) Antigenic characterization of ChZamydia pneumoniae isolated in Hiroshima, Japan. Microbial. Immunol. 37, 495-498. [S] Caldwell, H.D., Kromhout, J. and Schachter, J. (1981) F’urification and partial characterization of the major outer membrane protein of Chlamydia rrachomads. Infect. Immun. 3 1, 1161-l 176. [6] Brade, L., Brunnemann, H., Ernst, M., Fu, Y., Holst, O., Korsma, P., NZher, H., Persson, K. and Brade, H. (1994) Occurrence of antibodies against chlamydial lipopolysaccharide in human sera as measured by ELISA using an artificial glycoconjugate antigen. FEMS Immunol. Med. Microbial. 8, 27-42. [71 Kuo, C.-C., Jackson, L.A., Campbell, L.A. and Grayston, J.T. (1995) Chlamydia pneumoniae (TWAR). Clin. Microbio. Rev. 8, 451-461. Bl Rothermel, C.D., Schachter, J., Lavrich, P., Lipsits, E.C. and Francus, T. (1989) Chlamydia trachornatis-induced production of interleukin-1 by human monocytes. Infect. Immun. 57, 2705-2711. t91 Surcel, H.M., Syrjala, H., Leinonen, M., Saikku, P. and Herva, E. (1993) Cell-mediated immunity to Chlamydia pneumoniae measured as lymphocyte blast transformation in vitro. Infect. Immun. 61, 2196-2199. M.R., Smith, P. [lOI Nakajo, M.N., Roblin, P.M., H-erschlag, and Nowakowski, M. (1990) Chlamydicidal activity of human alveolar macrophages. Infect. Immun. 58, 3640-3644. [ill Numazaki, K., Suzuki, K. and Chiba, S. (1995) Replication of Chlamydia trachomaris and C. pneumoniae in the human monocytic cell line U-937. J. Med. Microbial. 42, 191-195. WI Numazaki, K., Wainberg, M.A. and McDonald, J. (1989) Chlamydia trachomatis infections in infants. Can. Med. Assoc. J. 140, 615-622. [131 Numazaki, K. and Chiba, S. (1996) Serum g-a-interferon in patients with penumonia caused by Chlumydia pneumoniae. Pediatr. Infect. Dis. J. 15, 174-175. [141 Grayston, J.T. (1994) Chlamydia pneumoniae (TWAR) infection in children. Pediatr. Infect. Dis. J. 13, 675-685. [15] Kanamoto, Y., Ouchi, K., Mizui, M., Ushio, M. and Usui, T. (1991) Prevalence of antibodies to Chlamydia pneumoniae TWAR in Japan. J. Clin. Microbial 29, 816-818.