Previous Article | Next Article ![]()
Journal of Clinical Microbiology, May 2005, p. 2456-2461, Vol. 43, No. 5
0095-1137/05/$08.00+0 doi:10.1128/JCM.43.5.2456-2461.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Department of Molecular Biotechnology, Ghent University, Ghent, Belgium,1 Laboratory of Physiology and Immunology of Domestic Animals, Catholic University of Leuven, Leuven, Belgium,2 Institut für Geflügelkrankheiten, Justus-Liebig-Universität, Giessen, Germany,3 Istituto Zooprofilattico Sperimentale della Lombardia e dell' Emilia Romagna, Sezione di Pavia, National Reference Laboratory for Animal Chlamydioses, Pavia, Italy,4 United States Department of Agriculture, National Animal Disease Center, Ames, Iowa,5 Institute for Environmental Health, Seattle, Washington6
Received 31 August 2004/ Returned for modification 13 November 2004/ Accepted 20 December 2004
|
|
|---|
|
|
|---|
Serotyping is performed with a panel of serovar-specific monoclonal antibodies (MAbs) in an immunofluorescence test. MAb-secreting hybridomas were previously screened by immunofluorescence staining using whole organisms (1, 2) and, as for Chlamydia trachomatis, were believed to recognize the major outer membrane protein (MOMP) (4). Only the serovar A- and D-specific MAbs were characterized using Western blotting and a radioimmunoprecipitation assay (RIPA), and those MAbs indeed recognized MOMP (22). However, the other serovar-specific MAbs still needed to be characterized. Consequently, the first objective of the present study was the characterization of the complete panel of serovar-specific MAbs. The second objective was to compare MOMP serotyping, ompA restriction fragment length polymorphism (RFLP), and ompA sequencing for characterizing C. psittaci strains from Psittaciformes, Columbiformes, Anseriformes, and Galliformes. MOMP serotyping and ompA RFLP have been compared previously (15, 18, 21) and sometimes gave incongruent results. Therefore, ompA sequencing was included as the reference typing method.
Twenty-one C. psittaci isolates from Belgium (n = 7), Germany (n = 6), Italy (n = 7), and The Netherlands (n = 1) were characterized (Table 1). Additionally, serovars A to F reference strains VS1, CP3, GD, NJ1, MN, and VS225 were used (2) (Table 2). Bacteria were grown in Buffalo Green Monkey (BGM) cells as described previously (21, 23).
|
View this table: [in a new window] |
TABLE 1. European C. psittaci isolates
|
|
View this table: [in a new window] |
TABLE 2. Serotyping and genotyping results for laboratory reference strains and European field strains
|
Genomic DNA for ompA sequencing was prepared as described by Wilson et al. (25). ompA was PCR amplified as described previously (22), resulting in a fragment of 1,065 to 1,098 bp depending on the strain. PCR products were purified using QIAGEN spin columns (Westburg, Leusden, The Netherlands) and cloned into pGemT (Promega, Madison, WI) following the manufacturer's protocol. Heat shock-transformed (30 s, 42°C) JM109-competent cells (Promega) were grown on Luria-Bertani (LB) plates in the presence of carbenicillin (30 µg/ml; Duchefa, Haarlem, The Netherlands). Five clones were selected and subsequently grown in liquid LB medium for plasmid purification (QIAGEN Tip 100 Westburg). ompA sequence analysis was performed by the VIB Genetic Service Facility (University of Antwerp, Belgium) and the Laboratory of Physiology and Immunology of Domestic Animals (KULeuven, Belgium) using vector associated T7 and SP6 priming sites.
Obtained ompA sequences were submitted to GenBank, aligned with previously published ompA sequences using ClustalX software (16), and subsequently imported into TreeconW (17) using the Jukes and Cantor algorithm and the Neighbor-Joining method (12) for designing the ompA phylogenetic tree (Fig. 1). Two thousand bootstraps were calculated (9). The tree was rooted with the Chlamydophila caviae (GPIC) ompA sequence (26) (GenBank AF269282).
![]() View larger version (15K): [in a new window] |
FIG. 1. ompA phylogenetic tree. An asterisk indicates that an additional sequence was found in the isolate. Bootstrap values are shown if the reliability is equal to or greater than 85%.
|
Serotyping was performed by use of an indirect immunofluorescence staining on C. psittaci-infected BGM monolayers grown on 13-mm-diameter coverslips on the bottom of Chlamydia Trac Bottles (CTB; International Medical, Brussels, Belgium). Each isolate was inoculated in 8 CTBs, and 72 h postinfection (hpi) monolayers were fixed in methanol (10 min, 20°C), where after nonspecific binding sites were blocked with PBS (pH 7.3) plus 5% FCS (1 h, room temperature). All dilutions were prepared in PBS plus 1% bovine serum albumin (BSA), and all incubations were performed in a moist chamber at 37°C for 1 h. Following blocking, 25 µl of either PBS, the family-specific control MAb 7B6III, or the differentiating dilution of the six serovar-specific MAbs A to F was added. The differentiating dilution of the serovar-specific MAbs was designated as the dilution at which the serovar-specific MAb reacted only with the corresponding reference strain. Differentiating dilutions were 1/5,000 for all MAbs, except for the serovar D-specific MAb, where a dilution of 1/500 had to be used. The monolayers were incubated and washed three times (5 min) in PBS plus 1% BSA, and subsequently 25 µl of fluorescein isothiocyanate-conjugated rabbit anti-mouse immunoglobulin (IgG H+L; DakoCytomation; diluted 1/200) was added. CTBs were incubated and washed two times (5 min) in PBS followed by two times (30 s) in bidest. Coverslips were removed, mounted on microscopic glass slides, and examined with a confocal laser scanning microscope (x600; Radiance 2000; Bio-Rad).
None of the MAbs, except for the family-specific control MAb 7B6III, recognized MOMP in immunoblotting, whether reducing or nonreducing conditions were used. However, all MAbs recognized MOMP in the RIPA, indicating they recognize conformational epitopes, which is rational since 90% of antigen epitopes are predicted to be discontinuous and highly conformational (19). B-cell epitopes are considered to contain 15 to 20 residues derived from two to five peptide segments, occupying a surface of 70 to 90 nm (7). Strangely, the serovar A- and D-specific MAbs recognized MOMP in former immunoblottings and RIPAs and now failed to react in immunoblotting (22). Looking for an explanation, we found out that present and formerly used serovar A- and D-specific MAbs originated from different clones (A. A. Andersen, personal communication). Recognition by the formerly used MAbs was probably less highly conformation dependent, or MAbs showed a higher affinity for the serovar-specific epitope. However, maybe the C. psittaci MOMP, like in C. trachomatis, can carry more than one serovar-specific epitope (4), and present and formerly used MAbs were against different serovar-specific epitopes on MOMP. Epitope mapping could provide an answer. Isotyping revealed the following results: IgG1 (7B6III), IgG2b (VS1/7A7/E8, MN/4A10/A4, VS225/D6/36), and IgG3 (CP3/4C6/G2, GD/2H9, NJ1/2GB/C2).
ompA sequencing of 21 isolates revealed eight genotypes A, five genotypes B, two genotypes D, and one genotype E and F (Table 2). Remarkably, in nine birds, mostly ducks, genotyping discovered a new ompA sequence, designated E/B. The E/B ompA showed high amino acid homology with the ompA of genotypes A, B, and E (97.3%, 98.6%, and 99.5%, respectively) and was characterized by a unique combination of a guanosine on positions 1,006 and 1,021 and a cytidine on position 1,022 resulting in an A instead of S on position 341 in the variable segment 4 of MOMP. In 94% of the cases, bootstrap values in the ompA phylogenetic tree supported the creation of an E/B branch starting from the genotype E group (Fig. 1). The latter justified the proposition of the new genotype E/B, closely related to genotypes A, B, and E. Interestingly, BLAST search revealed one ompA sequence (N352) (GenBank L04980) 100% identical to our E/B sequence (5). N352 was isolated from an English duck and classified as serovar E using ompA RFLP (14). E/B strains were isolated in three different countries (Italy, Germany, and recently also Belgium) and in different hosts (urban pigeons, commercial fattening Pekin ducks, and turkeys). Genotype E/B was highly prevalent, occurring mostly in ducks, while the closely related genotype E was only discovered in one pigeon. RFLP cannot distinguish between genotype E/B and E. Perhaps the 12 duck strains (including N352) formerly classified by Sayada et al. (14) using ompA RFLP were actually genotype E/B instead of genotype E, leading to the conclusion that we may have underestimated the occurrence of genotype E/B in the past. Five out of 21 birds (3 urban pigeons, a lovebird, and a turkey) were simultaneously infected with two different genotypes, which is in accordance with previous studies showing the presence of mixed infections (20, 21).
Only 19 out of 21 isolates could be serotyped (Table 2), revealing six mixed infections. The two remaining isolates only reacted with the family-specific MAb 7B6III; RFLP was successful in all cases, and like sequencing, revealed five mixed infections. Differentiating restriction patterns were obtained for genotypes A to F when performing an in silico AluI digest (Fig. 2), but the new genotype E/B could not be distinguished from genotype E. Genotypes A and B generated characteristic bands of 110 and 117 bp, respectively. Genotype E could be distinguished from genotypes A and B because the restriction pattern lacked the 110 and 117 bands. Genotype C generated a specific band of 438 bp. Genotypes D and F were characterized by 471-bp and 222-bp bands, respectively (Fig. 2). Due to inefficient restriction digests, the gel-determined and theoretically calculated RFLP restriction patterns sometimes differed, although all isolates could be genotyped by looking at characteristic bands.
![]() View larger version (26K): [in a new window] |
FIG. 2. Gel-based and in silico-based ompA RFLP using AluI. Thin lines represent the in silico-calculated lengths of the restriction fragments.
|
Serotyping cannot always be performed directly on clinical samples, as there may be insufficient antigen, requiring culture. However, culture is expensive and labor-intensive, requires a special biosafety level, and postpones serotyping for at least 3 to 6 days. Interpretation of the immunofluorescence staining is time-consuming and needs the interpreter's keen eye. At present, serotyping cannot conclusively detect genotype E/B. Theoretically, ompA RFLP analysis is more sensitive and specific, less expensive, and much faster. However, amplification of the full-length ompA gene directly from clinical samples is sometimes impossible without former culture. RFLP produces distinct restriction patterns, leading to conclusive typing results when looking at the presence of known specific bands except for the new genotype E/B, which generates pattern E. Like for RFLP analysis, ompA sequencing sometimes requires former culture to obtain the full-length PCR product for subsequent cloning. Sequencing is more expensive and more time-consuming than RFLP. However, at present, ompA sequencing is the only typing method capable to identify all known genotypes, including genotype E/B.
As formerly observed, certain genotypes were preferentially, though not strictly, associated with a specific bird species (21). Genotype A occurred in all psittacines, genotype B occurred in three of five pigeons, and genotype E/B was present in all ducks. Interestingly, so far, genotype F was only found in an American parakeet (VS225) and a German parrot (84/2334) (2, 18). Presently, we isolated an additional genotype F strain on a Belgian turkey farm dealing with respiratory disease.
Serotyping was historically important but has been replaced by ompA RFLP. However, RFLP cannot distinguish genotype E from the newly discovered genotype E/B. ompA sequencing is at present the only way to discover all known avian ompA genotypes and additional new genotypes. We clearly showed that new genotypes cannot always be discovered by RFLP nor by serotyping, as single-nucleotide polymorphisms did not generate another restriction pattern (unless they appear at the restriction site) and base mutations did not influence the conformation of the serovar-specific epitope(s). The present study emphasizes the need for a discriminatory, more rapid genotyping technique for examining both birds and humans given the zoonotic risk of C. psittaci infections. For this purpose, a combinatory species- and genotype-specific real-time PCR is currently being evaluated.
Nucleotide sequence accession numbers. The ompA gene sequences for the A, B, D, E, F, and E/B genotypes were submitted to GenBank and assigned accession numbers AY762608, AY762609, AY762610, AY762611, AY762612, and AY762613, respectively.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»