Evaluation of Chlamydia pneumoniae 43- and 53-Kilodalton Recombinant Proteins for Serodiagnosis by Western Blot

Abstract

Chlamydia pneumoniae is a common cause of respiratory infection. It has also been shown to be associated with coronary heart disease. Two proteins that have been reported to be recognized frequently during human infection are proteins having molecular masses of 43 and 53 kDa. In order to develop a useful alternative serological test to the microimmunofluorescence (micro-IF) assay, recombinant 43-kDa and 53-kDa chlamydia-specific proteins were evaluated in dot blot and/or for comparison to the standard micro-IF test. Primers for amplification were derived from genome sequence information for two C. pneumoniae genes (CPn0809 and CPn0980) encoding 53-kDa proteins and four C. pneumoniae genes (CPn0562, CPn0927, CPn0928, and Cpn0929) encoding 43-kDa proteins of unknown function, which were Chlamydia specific and not found in Chlamydia trachomatis. The 53-kDa protein product of CPn0809 or the N-terminal 18-kDa portion had better specificity than any of the 43-kDa recombinants but was much less sensitive than micro-IF. In contrast, the 53-kDa protein encoded by CPn0980 was recognized by 11 of 12 (92%) acute-phase sera, 35 of 46 (76%) chronic sera, 0 of 12 micro-IF-negative sera (C. pneumoniae and C. trachomatis negative), and 1 of 12 (8%) C. pneumoniae negative, C. trachomatis positive sera. Thus, it appears that the 53-kDa protein encoded by CPn0980 has potential use for serodiagnosis of C. pneumoniae infection.

Chlamydia pneumoniae is a human respiratory pathogen causing acute respiratory disease (4). Significant attention has focused on the association of this organism with cardiovascular disease. Investigators worldwide have demonstrated an association with atherosclerosis by seroepidemiological studies and the presence of the organisms within atherosclerotic lesions (1). Because of the potential impact that C. pneumoniae infection could have on public health practices if the organism is found to play a role in atherosclerosis, rapid diagnosis of infection is critical.

The current “gold standard” for serodiagnosis of C. pneumoniae infection is the microimmunofluorescence (micro-IF) test (10). Although this test remains the only specific and sensitive test for C. pneumoniae serodiagnosis, it is not readily adaptable for routine use in diagnostic clinical laboratories due to the requirement for highly trained personnel. Attempts to identify immunodominant C. pneumoniae antigens that are recognized during human infection have yielded variable results with respect to the frequency and pattern of recognition (2, 3, 6, 8, 11, 12, 16). Two C. pneumoniae antigens which appear to be frequently recognized in immunoblotting studies are 43-kDa and 53-kDa proteins.

In the present study, we investigated the diagnostic usefulness of recombinant antigens of 43 and 53 kDa by Western blot. If such antigens are found, it should facilitate the development of an alternative serological test for diagnosis of C. pneumoniae infection.

MATERIALS AND METHODS

Serodiagnosis.

Serum antibodies against Chlamydia trachomatis and C. pneumoniae were determined by the micro-IF test using formalin-fixed whole elementary bodies (EBs) (15). Antibody titers were interpreted according to diagnostic criteria described by Wang et al. (15). Acute C. pneumoniae infection was determined by a fourfold increase in antibody titers when acute- and convalescent-phase sera were available. If only a single serum sample was available, an immunoglobulin G (IgG) titer of ≥512 or IgM titer of ≥16 was the diagnostic criteria for acute infection. An IgG titer of ≥8 indicated past or chronic infection. At Focus Technologies, Inc., kits that are developed at the company and available for research and investigational purposes permit semiquantitation of micro-IF IgG and IgM antibodies in the micro-IF test (reference numbers IF1250G and IF1250M; Focus Technologies, Inc., Cypress, Calif.). This test was standardized against the classic micro-IF test.

Human sera used in this study were from our serum banks from previous studies conducted at the University of Washington and from Focus Technologies, Inc., and were collected over a 3-month period for routine diagnostic testing performed by the company and kept frozen since December 1999. The studies at the University of Washington included a multicenter study on antimicrobial therapy of patients with acute respiratory disease collected in 1991 to 1993 and a study on patients with ectopic pregnancy and their case-matched controls conducted at the Group Health Corporative in 1981 to 1986. For the latter study, micro-IF titers for both C. pneumoniae and C. trachomatis were available. Upon collection, sera were refrigerated for testing by micro-IF and subsequently stored at −20°C. Sera used in this study had been frozen and thawed one to four times.

Recombinant proteins.

Iijima et al. reported isolation of a monoclonal antibody (MAb) that reacted specifically with an immunodominant C. pneumoniae 53-kDa protein recognized during human infection (5). Subsequently, the gene encoding the protein recognized by this MAb was isolated, and the DNA sequence was used to develop a C. pneumoniae-specific PCR (7). This gene is designated CPn0809 and has a C. trachomatis homolog (CT578), with which it shows 45% amino acid sequence homology. Another gene encoding a C. pneumoniae protein of similar molecular mass (CPn0980) is listed as similar to a Saccharomyces cerevisiae 52.9-kDa protein based on DNA sequence homology, but no homologue is found in C. trachomatis. Whether there is any antigenic cross-reactivity of the C. pneumoniae and S. cerevisiae proteins is unknown. Thus, the two genes encoding 53-kDa proteins (CPn0809 and CPn0980) and four C. pneumoniae genes of unknown function encoding 43-kDa proteins (CPn0562, CPn0927, CPn0928, and Cpn0929), which were Chlamydia specific and not found in C. trachomatis, were selected from genome sequence information (7; http://chlamydia-www.berkeley.edu:4231/index.html).

The forward (F) and reverse (R) primers are summarized in Table 1. Primers to amplify the CPn0809 fragments included EcoRI recognition sequences for directional cloning into pGEX-4T3 (Pharmacia, Amersham, N.J.) by digestion with EcoRI in order to obtain in-frame cloning. Because the expressed protein was lethal (CPn0980), the recombinant plasmid containing the 1.45-kb insert was digested with EcoRI and SmaI to obtain the N-terminal fragment of 514 bp, which was subcloned into pGEX-4T3 to generate the correct orientation. In addition, the 1.45-kb fragment was subcloned into the pBAD/TOPOThio vector, which is designed to permit expression of lethal proteins by growth in RM medium, which contains glucose, which represses pBAD expression (Invitrogen, Carlsbad, Calif.). All of the remaining amplified products (CPn0562, CPn0927, CPn0928, CPn0929, and CPn0980) were ligated directly into the pBAD/Topo/Thio vector by TA cloning and expressed by the Topo/Thio fusion expression system (Invitrogen) according to the directions of the manufacturer.

TABLE 1.

Primers used for amplification of 43- and 53-kDa proteins

Locus	Forward primer (5′→3′)	Reverse primer (5′→3′)
CPn0562	TATAAGGTAATAATGTCAATAGCT	ATTATCGAAATGTCTTTGAATATG
CPn0927	GACTCTTGTATGATCCCAT	AGGTTGCTGCTTTCTG
CPn0928	AAGGATTTCGCATTCAC	GAGACTCTCAAGGATTGC
CPn0929	AAGTTTCATGGCTCCAATTCAC	GAGCTTTTTCAAAATTGGTGT
CPn0809	GGGAATTCGTCTATTTCATCTTC	GGGAATTCTGATTGCGGCATACG
CPn0980	ATGCTTAATCATGCTAAAAAG	CTCTTTTATTTTAGGAAGCTT

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using a 12% discontinuous gel or a gradient preformed gel (4 to15%; Bio-Rad, Hercules, Calif.) was performed according to Laemmli (9). After SDS-PAGE, proteins were transferred electrophoretically to nitrocellulose sheets for immunoblotting according to Towbin et al. (13). Following transfer, nitrocellulose sheets were immersed in 3% bovine serum albumin (BSA) in Tris-buffered saline (TBS)–0.1% Tween 20 at room temperature for 1 h to block excess protein-binding sites. The nitrocellulose sheets were then cut in small strips. Each strip was reacted with a human serum diluted 1:50 in phosphate-buffered saline (PBS) containing 0.5% Tween 20 at room temperature for 10 to 12 h and washed six times for 5 min each in TBS–0.1% Tween 20. Immune reactions were identified with goat anti-human immunoglobulins conjugated to horseradish peroxidase (ICN Cappel, Costa Mesa, Calif.). Color development was observed on addition of 4-chloro-1-naphthol (Sigma Chemical Co., St. Louis, Mo.) after rinsing nitrocellulose twice each with 1.5% BSA in PBS containing 0.5% Tween 20 and PBS only. Sensitivity and specificity of Western blot results in comparison to micro-IF were determined according to Vecchio (14).

RESULTS

Rationale and immunoblot analyses using recombinant 43-kDa proteins.

Although 43- and 53-kDa proteins have been recognized frequently by immunoblot analyses (2, 5, 11, 12), the identities of the specific proteins recognized have remained unknown. Several genes encoding proteins of similar molecular weight have been identified by the C. pneumoniae genome sequencing project. The strategy of this study was to use sequence information from the C. pneumoniae genome sequencing project (http://chlamydia www.berkeley.edu:4231/index.html) to identify Chlamydia-specific proteins with these molecular masses (6). Four genes encoding 43-kDa proteins (CPn0562, CPn0927, CPn0928, and Cpn0929), which are not found in C. trachomatis, and two 53-kDa proteins (CPn0809 and CPn0980), all of unknown function, were targeted. Screening of the 43-kDa proteins by immunoblot analyses using 15 micro-IF-positive sera and 15 micro-IF-negative sera did not demonstrate any differences in reactivity between micro-IF-positive or micro-IF-negative sera. Specifically, CPn0562 and CPn0927 gave weak reactions with 7 of 15 (47%) and 5 of 15 (33%) micro-IF-positive sera but reacted with equal frequency to micro-IF-negative sera. Cpn0928 and CPn0929 were not recognized by any of the micro-IF-positive sera tested. Thus, these four proteins were not explored any further.

Immunoblot analyses of recombinant 53-kDa proteins. (i) Evaluation of C5T78 hypothetical protein (CPn0809).

Two different approaches were used in assessing the usefulness of the CPn0809 recombinant protein. Western blot analyses using the entire recombinant protein found that 23 of 35 (66%) sera from patients with acute C. pneumoniae respiratory infection and 7 of 35 (20%) from patients with micro-IF titers indicating chronic or past infection were reactive and 6 of 30 (20%) micro-IF-negative sera had apparent reactivity. Anti-C. trachomatis sera did not react. Because of the insoluble nature of the entire protein and difficulty in purification, fragments encoding the 18-kDa N-terminal portion of this 53-kDa protein were tested for their recognition by nine antisera that were reactive with the recombinant 53-kDa. All sera tested reacted with the 18-kDa protein. Of the sera from patients with antibody titers indicative of acute C. pneumoniae infection, 5 of 12 (42%) were positive by Western blot (Table 2), while 16 of 49 (33%) patients with antibody titers indicating chronic or past infection were positive. Because there is a 45% shared sequence homology of the C. pneumoniae protein with the C. trachomatis homologue, a panel of sera with known antibody titers to both C. trachomatis and C. pneumoniae was tested to determine specificity. Of the 12 sera that were negative for C. pneumoniae and C. trachomatis micro-IF antibody, 1 (8%) recognized the 18-kDa portion and 2 (17%) sera that were micro-IF positive for C. trachomatis reacted by Western blot. In comparison to the micro-IF test, recognition of the 18-kDa N-terminal portion of the 53-kDa protein had a sensitivity of 33% and a specificity of 87.5%.

TABLE 2.

Detection of serum antibodies against recombinant C. pneumoniae 53-kDa proteins by Western blot

Serum samplea	No. of samples positive/ no. tested (% positive)
	CPn 0809		CPn 0980
	Whole antigen	18-kDa antigen
C. pneumoniae antibody positive
Acute	23/35 (66)	5/12 (42)	11/12 (92)
Chronic	7/35 (20)	16/49 (33)	35/46 (76)
Total	30/70 (43)	21/61 (39)	46/58 (79)
C. pneumoniae antibody negative
Cpn−/Ctr−	ND	1/12 (8)	0/12 (0)
Cpn−/Ctr+	ND	2/12 (17)	1/12 (8)
Total	6/30 (20)b	3/24 (12)	1/24 (4)

^aAbbreviations: Cpn, C. pneumoniae; Ctr, C. trachomatis; ND, not determined. 

^bIn this panel of sera, only C. pneumoniae antibodies were determined. 

(ii) Evaluation of C. pneumoniae protein similar to 52.9-kDa S. cerevisiae CPn0980.

Using the same panels from the University of Washington studies with the CPn0980 53-kDa protein, 11 of 12 (92%) sera from persons with acute infection were reactive with the recombinant protein in Western blot while 35 of 46 (76%) patients with micro-IF antibody indicating chronic infection were positive. Of the 12 micro-IF sera negative for both C. pneumoniae and C. trachomatis, 0 were positive and only 1 sample that was micro-IF negative for C. pneumoniae and positive for C. trachomatis was positive. Overall, only 1 of 24 (4%) sera that were negative for C. pneumoniae antibody was reactive in Western blot with the CPn0980 protein. Recognition of the CPn0980 protein by Western blot had a sensitivity of 79% and a specificity of 95.8% to the micro-IF test.

DISCUSSION

Although the micro-IF test remains the gold standard for serodiagnosis and the tool for seroepidemiologic studies of C. pneumoniae infection, difficulties in interpretation of the test resulting in interlaboratory variations in reporting of micro-IF titers underscore the need for development of alternative methods to permit routine and standardized serodiagnostic testing for laboratory diagnosis (10). Many studies have been done to identify C. pneumoniae-specific antigens that are recognized during human infection to facilitate development of such serodiagnosis. Although no clear pattern of reactivity indicative of C. pneumoniae infection has emerged, several C. pneumoniae-specific and immmunodominant antigens have been described. This study focused on two immunodominant proteins of 43 and 53 kDa that were reported to be recognized frequently during human infection and suggested as putative candidates for development of serodiagnostic testing (2, 5, 11, 12). In a previous report by Freidank et al., 93% of sera positive for C. pneumoniae by micro-IF were positive for a 54-kDa protein of similar molecular mass, and this reactivity was specific with only 2 and 3% of C. pneumoniae micro-IF-negative sera that were either simultaneously negative for C. trachomatis or positive for C. trachomatis antibody, respectively (3). A similar frequent recognition pattern of a 53-kDa protein was noted by Iijima et al. (5). Although specific recognition of a 53-kDa C. pneumoniae protein has been reported, as has a C. pneumoniae MAb with reactivity against the C. pneumoniae protein designated as hypothetical protein CT578 (CPn0809), the identity of the antigen recognized in immunoblots has not been determined. For the sera used in our study, the frequency of detection of the N-terminal portion of the 53-kDa protein was much less sensitive than micro-IF serology and thus would not be useful as a sole component of a serodiagnostic test. However, combined use of this antigen with others may prove useful. In contrast, the frequent recognition of the 53-kDa protein designated as similar to a 52.9-kDa S. cerevisiae protein (CPn0980) suggests that this protein is the immunodominant protein that was recognized during human infection in the earlier studies (3, 5) and the best candidate for serodiagnosis of those evaluated. Investigations are under way to evaluate use of this recombinant in an enzyme-linked immunosorbent assay-based format.