Abstract
Chlamydia pneumoniae, an important respiratory pathogen, is difficult to culture, and detection rates by conventional PCRs vary considerably. A new quantitative ompA-based real-time PCR assay based on TaqMan technology for detection of C. pneumoniae in respiratory samples is described, and its performance in terms of sensitivity and reproducibility is compared with those of four published conventional PCRs (one single-step PCR targeting a cloned PstI fragment; two nested PCRs, one targeting the 16S rRNA gene followed by hybridization and the other targeting the ompA gene; and a touchdown enzyme time-release [TETR] PCR also targeting the 16S rRNA gene). Both ompA-based PCRs showed the best analytical sensitivity. All five assays could detect even lower target levels from spiked sputum, with the 16S rRNA assays performing better than the ompA-based nested PCR (10−6 inclusion-forming units [IFU] were detected in four of four and two of four replicates by the 16S rRNA TETR PCR and the 16S rRNA nested PCR, respectively). In general, the ompA-based real-time protocol produced the most consistent positive results for all replicates tested down to 10−6 IFU. Eight of 45 patient sputum specimens (18%) were C. pneumoniae DNA positive in at least one of four replicates tested by at least one assay. Without taking into consideration the analytical sensitivity or the reproducibility of the test results, the numbers of C. pneumoniae DNA-positive sputum specimens (n = 8) were four, three, two, two, and one for the 16S rRNA TETR assay, the PstI-based single-step PCR, the ompA-based real-time PCR, the ompA-based nested touchdown PCR, and the 16S rRNA-based nested PCR, respectively. However, the overall rate of concordance of positive results was low. Only one cell culture-positive sputum specimen was positive by four of five assays (14 of 16 replicates; mean cycle threshold value, 25; 108 particles/ml of sputum). Thirty-seven specimens were C. pneumoniae negative by all five assays for all replicates tested, as were all negative controls (n = 65 to 100 per testing panel). No PCR inhibitors were detected by real-time PCR or by the 16S rRNA-based nested assay. We confirm that the analytical sensitivity of an assay for the detection of C. pneumoniae does not necessarily predict its ability to detect its target in sputum. A quantitative, fast, and easy-to-handle diagnostic approach such as the ompA-based real-time TaqMan PCR described here might improve the detection of C. pneumoniae in respiratory samples.
Chlamydia pneumoniae has emerged as an important and common cause of respiratory tract infection in humans (11, 12, 15), but isolation of the organism proves difficult. On the other hand, serology, at present the test most often used for routine diagnosis, often leads microbiologists and clinicians to face a diagnostic dilemma. If at all, serology only allows a retrospective diagnosis, with detection of a significant increase in antibody titers perhaps taking weeks (13, 15), and it is not uncommon for apparently healthy individuals to fulfill the serologic criteria for acute infection (13). In many cases, it is nearly impossible to determine if a patient's antibody titer is due to acute, previous, or chronic C. pneumoniae infection, even if the clinical picture is taken into account.
Therefore, it seems reasonable to apply nucleic acid amplification techniques (NAAs) for the detection of a respiratory pathogen which is rather difficult to find by other methods. PCR offers a rapid and sensitive diagnostic approach, and until now, numerous amplification techniques have been described for the detection of C. pneumoniae. However, conventional PCRs have limitations, and a recently published review article thoroughly discusses not only the principles but also the shortcomings of the molecular biology-based amplification methods available for the detection of C. pneumoniae (6). Most of these PCR assays or techniques have also already been used in a number of studies for the detection of C. pneumoniae in clinical samples (2, 4-7). However, no standardized assays are available, and positivity rates for clinical samples have varied considerably by protocol, number of replicates analyzed, and testing center (1, 6, 18, 22). Several studies have documented the use of various conventional PCRs with respiratory specimens (6), but in particular, no studies have systematically evaluated or compared these assays in terms of their sensitivities or the reproducibilities of the test results.
Real-time PCR technology offers a new diagnostic approach which also allows amplicon quantification. This new technique combines amplification and quantitative product detection via specific hybridization in one step. This eliminates the need to open tubes that may contain amplicons for nested PCR, Southern blotting, or gel electrophoresis, thus essentially minimizing the risk of cross-contamination.
The aim of this study was to design and establish a quantitative real-time PCR assay for routine diagnosis of C. pneumoniae infection from respiratory samples of patients suffering from respiratory tract infection. In order to meet the criteria proposed for a validated assay, our new real-time PCR assay was compared with four conventional PCRs (8, 17, 20, 24). Those PCRs were included in the recommendations of a consensus paper published by the Centers for Disease Control and Prevention (Atlanta, Ga.) and the Laboratory Centre for Disease Control (Ottawa, Ontario, Canada) (10).
For this purpose, all five assays were evaluated by systematically comparing them in terms of their analytical sensitivities and sensitivities when they were applied to sputum controls spiked to mimic real-life respiratory specimens as much as possible. Finally, specimens from patients suffering from acute respiratory infection were analyzed. Since the reproducibility of conventional C. pneumoniae PCR results has presented a problem in recent studies, all analyses performed by the various assays with the various testing panels in this study were repeated four times.
MATERIALS AND METHODS
Propagation and quantification of C. pneumoniae.
All chlamydial strains propagated in this study were cultured as described earlier by Roblin et al. (21). Briefly, for preparation of the purified C. pneumoniae DNA and the C. pneumoniae-infected HEp-2 cells used to determine the analytical sensitivity and to spike sputum, respectively, serial dilutions of a frozen chlamydial stock (a local respiratory C. pneumoniae isolate) were cultured in triplicate onto HEp-2 cells (CCL-23; American Type Culture Collection [ATCC], Manassas, Va.), and infectivity titers were determined 72 h after infection and expressed as the number of inclusion-forming units (IFU) per milliliter. To obtain purified C. pneumoniae elementary bodies (EBs) for DNA preparation, stock chlamydiae were purified by differential centrifugation over a sucrose-Urografin gradient. The corresponding chlamydial particles were determined by immunofluorescence microscopy (magnification, ×630; Axioplan 2; Zeiss, Göttingen, Germany) as follows: aliquots of chlamydial dilutions were centrifuged onto lysin-coated coverslips and stained with a direct, genus-specific immunofluorescence antibody (IMAGEN-Chlamydia; DAKO, Hamburg, Germany).
Other bacterial isolates.
To test for the specificity and cross-reactivity of the real-time PCR assay, 37 microorganisms from the American Type Culture Collection, the Culture Collection of the University of Goeteborg (CCUG), and the Quality Assurance Laboratory of the Central Public Health Laboratory at Colindale (London, United Kingdom) were tested and are listed in Table 1.
TABLE 1.
Microorganismsa used to test the specificity and cross-reactivity of the real-time PCR assay designed to detect the ompA gene of C. pneumoniae
| Species | Source (strain or reference) |
|---|---|
| C. pneumoniae | TW-183; ATCC (VR-2282) |
| C. pneumoniae | AR-39; ATCC (53592) |
| C. pneumoniae | MUL-1; M. Maass |
| C. pneumoniae | WIEN I (3) |
| C. pneumoniae | WIEN II (3) |
| C. pneumoniae | WIEN III (3) |
| C. pneumoniae | Local respiratory isolate |
| C. trachomatis | ATCC (VR-902B) |
| Staphylococcus aureus | ATCC (25923) |
| Staphylococcus lugdunensis | ATCC (700328) |
| Staphylococcus epidermidis | QCb |
| Streptococcus pneumoniae | ATCC (6303) |
| Streptococcus pyogenes | ATCC (19615) |
| Streptococcus mitis | QC |
| Streptococcus sanguis | QC |
| Streptococcus uberis | QC |
| Streptococcus milleri | QC |
| Enterococcus faecalis | QC |
| Haemophilus influenzae | ATCC (10211) |
| Haemophilus parainfluenzae | QC |
| Escherichia coli | ATCC (11229) |
| Mycoplasma pneumoniae | CCUG (10119) |
| Neisseria perflava | CCUG (17915) |
| Neisseria subflava | CCUG (23930) |
| Corynebacterium diphtheriae | QC |
| Bordetella pertussis | QC |
| Moraxella catarrhalis | QC |
| Legionella pneumophila | QC |
| Klebsiella pneumoniae | QC |
| Eikenella corrodens | QC |
| Pseudomonas aeruginosa | QC |
| Stenotrophomonas maltophilia | QC |
| Bacteroides fragilis | QC |
| Peptostreptococcus anaerobius | QC |
| Lactobacillus casei | QC |
| Prevotella corporis | QC |
| Porphyromonas asaccharolyticus | QC |
A total of 37 isolates were tested.
QC, quality control strains; the isolates were obtained from the Quality Assurance Laboratory of the Central Public Health Laboratory at Colindale, London, United Kingdom, for the purpose of quality control.
DNA extraction.
DNA from C. pneumoniae-purified EBs, the spiked sputum panel, and 45 frozen respiratory specimens was extracted by use of the QIAamp DNA mini kit (Qiagen Inc., Valencia, Calif.), and the protocol for body fluids outlined in the instructions of the manufacturer was followed. The DNA elution volumes for the different panels were as described below, and the amounts of template DNA used for the different PCR assays or the individual testing series are displayed in Table 2. To test for cross-reactivity, bacterial DNA was extracted from freshly cultured organisms by the protocol for bacterial plate cultures of the QIAamp DNA mini kit (Qiagen), eluted in 50 μl of elution buffer AE, and stored at −20°C prior to real-time PCR analysis (5 μl/25 μl of the PCR mixture; the tests were done in triplicate).
TABLE 2.
Protocols for PCR assays evaluated
| Reference | Target gene | Format | Amplicon size (bp) | Amt of input DNA per PCR (μl)a
|
|
|---|---|---|---|---|---|
| Purified DNA and mock-infected sputum | Clinical respiratory specimens | ||||
| Campbell et al. (8) | Cloned PstI | Single step | 437 | 10/50 | 5/50 |
| Madico et al. (17) | 16S rRNA | TETR | 197 | 10/50 | 5/50 |
| Nadrchal et al. (20) | 16S rRNA | Nested followed by dot blotting | 489/304b | 10/50 | 5/50 |
| 1/50c | 1/50c | ||||
| Tong and Sillis (24) | ompA | Nested touchdown | 333/207 | 10/50 | 5/50 |
| 1/50c | 1/50c | ||||
| This study | ompA | Real-time TaqMan system | 85 | 10/50 | 5/25 |
Input DNA (μl) per PCR for each replicate (n = 4) of each dilution of the C. pneumoniae DNA dilution series or mock-infected sputum tested as well as from clinical specimens (a fifth replicate in which 30 copies of C. pneumoniae DNA was added as an inhibition control was tested by a 16S rRNA-based PCR [20] and the ompA-based real-time PCR).
The amplicon sizes are for the first PCR/nested PCR products.
Nested PCR run.
During each DNA extraction procedure extraction-negative controls (ENCs) containing salmon sperm (20 μg/extraction) in double-distilled water were processed at every second position. Quantification and determination of the quality of the DNA stocks were performed by spectrophotometric analysis (VersaFluor; Bio-Rad, London, England).
Analytical sensitivity.
A C. pneumoniae DNA dilution series was prepared to study the analytical sensitivities of all five assays. For this purpose, purified EBs of a chlamydial stock solution (a local respiratory C. pneumoniae isolate) of known infectivity titer (2 × 107 IFU/ml) were used to prepare the DNA (without the addition of carrier DNA); the titer was calculated such that the solution contained DNA corresponding to 1 × 105 IFU/μl of eluate. Subsequently, 10-fold serial DNA dilutions (n = 20) were prepared and tested in preliminary experiments by a conventional 16S rRNA-based PCR assay (20). Eight DNA concentrations (range, 103 to 10−4 IFU/10 μl of DNA) were chosen for further experiments and stored as small aliquots at −80°C.
Mock-infected controls.
To mimic clinical respiratory specimens as much as possible, mock-infected sputum controls were prepared as follows: C. pneumoniae-negative (as determined by 16S rRNA-based PCR [20]) sputum samples from our routine microbiology laboratory were pooled, centrifuged at 20,000 × g (30 min, 10°C), mechanically homogenized in TE (Tris-EDTA) buffer (pH 7.5), and divided into small aliquots. Prior to DNA extraction, each aliquot (n = 14) of sputum was spiked with dilutions of C. pneumoniae-infected HEp-2 cells (prepared from the same stock described above; titer, 2 × 107 IFU/ml) so that the sputum contained a constant DNA background (1 μg/10 μl of eluate) plus DNA derived from C. pneumoniae-infected HEp-2 cells (corresponding to a range between 1 × 105 and 1 × 10−6 IFU/10 μl of eluate). DNA from 10 spiked sputum samples (range, 1 × 103 to 1 × 10−6 IFU/10 μl of DNA extract) was chosen in a preliminary analysis (by 16S rRNA-based PCR) for further experiments and stored as small aliquots at −80°C.
Patient respiratory samples.
Frozen (−80°C) respiratory specimens (total, n = 45; sputum specimens, n = 39; nasopharyngeal swabs, n = 3; bronchoalveolar lavage [BAL] fluids, n = 3) collected during a study conducted to evaluate a blood-based C. pneumoniae PCR for patients suffering from acute respiratory tract infection (2) were used for DNA extraction. Samples were processed as follows: sputum and BAL fluids were centrifuged at 20,000 × g for 30 min at 4°C prior to division and homogenization of one half in 1 ml of TE buffer with sterile glass grinders. Nasopharyngeal swabs were squeezed in 1 ml of TE buffer. One hundred microliters of either sputum, BAL fluid, or swab samples was used for DNA extraction in this study, and DNA was eluted in 150 μl of AE buffer and stored as small aliquots at −80°C prior to analyses by the five PCR assays. Cell culture and serology results (microimmunofluorescence test; Labsystems, Helsinki, Finland) were available for most patients. Quantification of DNA in all extracts was performed by spectrophotometry (VersaFluor; Bio-Rad).
Quantitative ompA-based real-time PCR.
The C. pneumoniae-specific sequences of the PCR primers and probe were selected from the ompA gene encoding the major outer membrane protein (MOMP) of C. pneumoniae (GenBank accession number AF131889) with Primer Express Software (Applied Biosystems, Foster City, Calif.) and synthesized by Applied Biosystems. The PCR product generated (QM85) was 85 bp; and the sequences of the primers and the TaqMan probe were as follows: forward primer QMOMP1, 5′-GATCCGCTGCTGCAAACTATACT-3′; reverse primer QMOMP2, 5′-GTGAACCACTCTGCATCGTGTAA-3′; and TaqMan probe QMOMPS, 5′-TAGGCCGGGTTAGGTCTATCTACGGCAGT-3′. The QMOMPS TaqMan probe was fluorescence labeled at the 5′ end with 6-carboxyfluorescein as the reporter dye and at the 3′ end with 6-carboxytetramethylrhodamine as the quencher. A search was performed with the BLAST program to check the specificities of the primers and probe. In addition, published sequences encoding the MOMP of C. pneumoniae (n = 11), C. psittaci (n = 2), C. trachomatis (n = 2), and six other chlamydial species were aligned to the primers and probe in order to verify the species specificity of the assay by using ClustalW multiple-sequence alignments software (version 1.82) (Fig. 1).
FIG. 1.
ClustalW multiple-sequence alignments of the 85-bp PCR product (QM85), primers (forward primer, QMOMP1; reverse primer, QMOMP2) and probe (QMOMPS) of the real-time PCR assay for the detection of the ompA genes of C. pneumoniae (row 1, MOMP [AF131889]; row 2, MOMP [AF131230]; row 3, MOMP [AF131229]; row 4, IOL-207, CHTMOMPEB [M64064]; row 5, J138, ompA [NC_002491]; row 6, CWL-029, ompA [AE001363]; row 7, AR39, MOMP [AE002167]; row 8, CHTMOMPP [M69230]; row 9, koala type I, CHTOMPAAI [M73038]; row 10, CHTMOMPEQ [L04982]), C. suis (row 11), PCLH197, ompA [AJ440241]), C. psittaci (row 12, strain 84-55, omp1 [Y1656]; row 13, 6BC, MOMP [56980]), C. trachomatis (row 14, C/TW3/OT, omp1 [AF352789]), C. trachomatis (row 15, strain D/Ep6, omp1 [X77364]), C. felis (row 16, Fpn/pring, CPFPNMOMP [X61096]), C. muridarum (row 17, SFPD, CHTMOMPZ [L19221]), C. caviae (row 18, ATCC VR813, ompA [AF269282]), C. pecorum (row 19, LW613, ompA [AJ440240]), and C. abortus (row 20, LLG, omp1 [AF272945]) (the designations in brackets are GenBank accession numbers). Dots and empty spaces indicate identities and gaps compared with the target sequence of the C. pneumoniae strain with GenBank accession no. AF131889, respectively.
Optimization of the real-time PCR conditions was done as suggested in the manual provided by Applied Biosystems. All instructions were followed in order to achieve a highly efficient amplification reaction by using an ABI Prism 7700 sequence detector. The forward and reverse primers were tested in each possible combination between 50 and 600 nM/PCR mixture with purified C. pneumoniae DNA, and the concentrations of those combinations that gave the highest yield of amplification product without any nonspecific signals were chosen. Further titration experiments with various probe concentrations (50 to 250 nM) were done to increase the yield of the specific PCR product, but 200 nM probe/PCR mixture proved optimal. No inhibitory effects were observed when the assay was applied to spiked sputum in a number of experiments, irrespective of whether the sputum was spiked with purified C. pneumoniae DNA, EBs, or C. pneumoniae-infected HEp-2 cells. DNA for these experiments was extracted by Qiagen protocols. For all optimization experiments, the cycling conditions were those suggested by the manufacturer, and all default program settings were used.
In detail, PCR was performed in 96-well MicroAmp optical plates (Applied Biosystems), with the reaction mixtures consisting of 25 μl of the TaqMan Universal Master mix including dUTP and uracyl N-glycosylase (AmpErase UNG; Applied Biosystems), each of the primers at a concentration of 300 nM, and 200 nM TaqMan probe in a total reaction volume of 50 μl for the purified C. pneumoniae DNA series and the C. pneumoniae-spiked sputum samples (mock-infected controls). Since the amount of DNA from each respiratory sample to be tested by five assays for five replicates (four replicates with unspiked samples and one replicate spiked with the amount of DNA corresponding to 30 particles of C. pneumoniae) was limited, the reaction volumes were changed from 50 to 25 μl (preliminary experiments revealed no differences in terms of the sensitivities and the reproducibilities of the results compared to those achieved with 50-μl reaction volumes). The amounts of DNA used as templates for the individual testing series and assays are displayed in Table 2. Four replicates of each dilution of purified C. pneumoniae DNA, the mock-infected sputum specimen, the clinical respiratory samples, and the cross-reactivity panel were tested. The setup for a typical run (maximum of 96 reactions) including clinical specimens of unknown C. pneumoniae content was as follows: triplicates of five C. pneumoniae DNA dilutions (1 to 10−4 IFU/PCR mixture, which was the same amount used to test the analytical sensitivity; the dilutions were stored as small aliquots at −80°C) as standards for quantification and unknown samples (tested four times each) matched with one no-template control and one ENC for each clinical sample tested. Thus, specimens from 13 patients (including 13 no-template controls and 13 ENCs) could be tested in one quantitative run. Amplification and detection of the PCR product were performed with an ABI Prism 7700 sequence detection instrument (Applied Biosystems), as suggested by the manufacturer, by use of all default program settings. Briefly, cycling conditions were as follows: after 2 min at 50°C and 10 min at 95°C, the samples were submitted to 40 cycles, each consisting of a step at 95°C for 15 s, followed by a step at 60°C for 1 min. The PCR product was detected as an increase in fluorescence during the PCR extension phase when the probe was cleaved by the 5′ exonuclease activity of the Taq DNA polymerase. This cleavage interrupts the fluorescence resonance energy transfer and the reporter dye starts to fluoresce in proportion to the level of PCR product generated. The cycle threshold (CT) values, defined as the number of cycles at which the fluorescence of the reporter dye first exceeds the calculated background level, were automatically estimated by the instrument for each reaction. Standard graphs of CT values obtained from serial dilutions of purified C. pneumoniae DNA and sputum samples spiked with C. pneumoniae-infected HEp-2 cells were constructed. CT values for unknown, clinical respiratory samples were plotted against the standard graphs for purified C. pneumoniae DNA. The amount of C. pneumoniae present in the clinical samples was calculated with ABI PRISM sequence detection software (version 1.6). In addition, DNA from each clinical sample (a fifth replicate) was tested in the presence of C. pneumoniae DNA (30 particles) to check for PCR inhibitors by comparing the amplification plots for the spiked samples to that obtained by using an amount of DNA corresponding to 30 particles as a template alone. PCR analyses were considered negative for C. pneumoniae DNA if the CT values exceeded 40 cycles.
Conventional PCRs.
Table 2 summarizes the protocols for the four conventional PCR assays applied in this study. Basically, one assay was a single-step amplification protocol that targets a cloned PstI fragment (8). Two protocols, one that targets the 16S rRNA gene (20) and another that targets the ompA gene (24) of C. pneumoniae, were nested assays. To increase the sensitivities and specificities, the targets were amplified by a two-step procedure with two different primer pairs. To achieve a higher stringency of binding of primers during the first amplification steps, the ompA-based nested PCR was additionally based on the touchdown technique: a high annealing temperature was used during the initial cycles to increase specificity, followed by a decrease in the annealing temperatures, i.e., every second cycle, for efficient amplicon amplification. The characteristics of the touchdown enzyme time-release (TETR) assay, which also targeted part of the 16S rRNA gene of C. pneumoniae, were use of a hot-start polymerase to avoid artifacts before amplification, a touchdown protocol for annealing temperatures to reduce the level of background amplification and thus to increase the specificity, and an enzyme time-release protocol to allow 60 cycles of amplification for improved analytical sensitivity (17).
If not stated otherwise in Table 2, the reaction conditions as well as product detection were as published in the indicated references. Negative controls were matched at every third position at each of the first and nested PCR levels (in addition to the ENCs, which were processed at every second position during DNA extraction). To check for potential inhibitors in the 45 respiratory samples by a conventional assay as well, a fifth replicate was spiked with C. pneumoniae DNA corresponding to 30 particles, and amplification was done by the 16S rRNA-based nested PCR (20). Exhaustive care was taken with all conventional PCRs to avoid amplicon carryover.
RESULTS
Analytical sensitivity.
The results obtained with the C. pneumoniae DNA dilution series are shown in Fig. 2. Four assays identified C. pneumoniae DNA down to a level of 10−1 IFU or 35 chlamydial particles in at least one of four analyses. In detail, both ompA-based PCRs showed the best analytical sensitivities, with 10−1 IFU in three of four replicates and 10−2 IFU in two of four replicates by the real-time protocol (mean CT value, 38.5) and the nested PCR of Tong and Sillis (24), respectively. The least sensitive assay was the single-step PstI-based PCR, which detected one of four replicates to a dilution of 10 IFU or 3,500 particles.
FIG. 2.
PCR results obtained with a purified C. pneumoniae DNA dilution series. One IFU per PCR mixture corresponds to 35 chlamydial particles/10 μl, which was used as a template in a total PCR volume of 50 μl. •, ompA real-time PCR (this study); ▴, 16S rRNA-based TETR assay (17); ⧫, PstI-based single-step assay (8); ▪, 16S rRNA-based nested PCR (20); ∗, ompA-based nested touchdown (24).
Mock-infected sputum.
The results obtained with sputum spiked with C. pneumoniae-infected HEp-2 cells are shown in Fig. 3. All five PCRs could detect target levels from spiked sputum even lower than those detected from purified C. pneumoniae DNA. The least sensitive assay was again the single-step protocol, which, however, also detected as little as 10−3 IFU in three of four replicates tested. Moreover, the results obtained by this assay were more reproducible than those obtained by the ompA-based nested touchdown PCR of Tong and Sillis (24) in this test series. The real-time protocol produced the most consistent, reproducible positive results for all replicates tested down to a dilution of 10−6 IFU.
FIG. 3.
PCR results obtained with sputum spiked with C. pneumoniae-infected HEp-2 cells. •, ompA-based real-time PCR (this study); ▴, 16S rRNA-based TETR assay (17); ⧫, PstI-based single-round assay (8); ▪, 16S rRNA-based nested PCR (20); ∗, ompA-based nested touchdown (23).
Respiratory specimens.
Forty-five respiratory samples (86% of which were sputum samples) were analyzed by five PCR assays (with four replicates of each assay). No inhibitors were detected in any of the 45 samples when they were spiked with 30 chlamydial particles and analyzed by the real-time assay or the 16S rRNA-based nested PCR (20). For eight sputum samples (18%), at least one of four replicates was C. pneumoniae DNA positive by at least one assay. The overall concordance of positive results is shown in Fig. 4. Table 3 summarizes the available clinical information and the cell culture results as well as the serology results for the eight patients from whom these samples were obtained. One sputum sample (sample L5) was positive by four of five assays (14 of 16 replicates). The PstI-based PCR did not detect C. pneumoniae in this specimen in any of the four replicates. Specimen L5 was the only specimen from which C. pneumoniae could also be isolated by cell culture. The serum sample from this patient revealed C. pneumoniae-specific immunoglobulin G (IgG) titers of ≥1:512 and IgA titers of ≥1:64. A second sputum specimen (specimen L25) was positive for C. pneumoniae DNA only by the real-time PCR in two of four replicates (mean CT value, 38.5). C. pneumoniae could not be isolated from the sputum of patient L25, although C. pneumoniae-specific antibody titers of 1:16 and 1:128 for IgM and IgG, respectively, could be detected in the corresponding serum sample. When the amount of C. pneumoniae present in the real-time PCRs for specimen L5 was plotted against the standard graph of the CT values obtained with serial dilutions of purified C. pneumoniae DNA, the amount of C. pneumoniae present in the real-time PCRs for that specimen (5 μl of DNA analyzed/replicate) was calculated to correspond to 1 × 102 to 1 × 103 IFU/5 μl of DNA extract, 3 × 103 to 3 × 104 IFU/100 μl of sputum, or 1 × 108 particles/ml of sputum. Thirty-seven specimens were C. pneumoniae negative by all five assays for all replicates tested. All negative controls (65 to 100 controls/testing panel; ENCs as well as controls at the PCR level) were negative by all five PCR assays in all repetitive testing experiments.
FIG. 4.
Overall concordance of positive PCR results obtained with sputa from eight patients (patients L5, L7, L17, L25, B10, W32, W35, and W39). At least one of four replicates was positive for C. pneumoniae DNA by at least one assay. , sputum from which C. pneumoniae was isolated by cell culture. •, ompA-based real-time PCR (this study); ▴, 16S rRNA-based TETR assay (17); ⧫, PstI-based single-step assay (8); ▪, 16S rRNA-based nested PCR (20); ∗, ompA-based nested touchdown (24).
TABLE 3.
Clinical characteristics and C. pneumoniae infection status of eight patients with symptoms of acute respiratory tract infection and detectable C. pneumoniae DNA in respiratory samples in at least one of four replicates tested by at least one assay
| Patient no. | Sexa | Age (yr) | Specimen type | Finding on chest X ray | Clinical information |
C. pneumoniae status
|
|||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Culture resultb | PCR
|
Reciprocal MIFc titer
|
|||||||||
| No. of positive replicates/ no. of replicates analyzed for each assay | Assayd | IgM | IgG | IgA | |||||||
| L5 | F | 65 | Sputum | Stripy infiltrate dorsobasically in right lobe | Dry cough, dyspnea | + | 4/4 | • | —f | ≥512 | ≥64 |
| 4/4 | ▪ | ||||||||||
| 4/4 | ▴ | ||||||||||
| 2/4 | ∗ | ||||||||||
| L7 | M | 69 | Sputum | Pneumonia verified | Productive cough; dyspnea | − | 4/4 | ⧫ | — | 32 | — |
| L17 | M | 61 | Sputum | No abnormal findings | Dry cough for 3 mo | − | 4/4 | ⧫ | — | 128 | 32 |
| L25 | F | 31 | Sputum | Pneumonia verified | Dry cough; fever | − | 2/4 | • | 16 | 128 | — |
| B10 | M | 73 | BAL fluid | Pneumonia verified | Productive cough | − | 1/4 | ∗ | — | 128 | 32 |
| W32 | F | 61 | Sputum | NDe | Dyspnea; cough | − | 4/4 | ⧫ | — | ||
| 3/4 | ▴ | ||||||||||
| W35 | F | 78 | Sputum | ND | Dry cough | − | 1/4 | ▴ | — | 64 | 8 |
| W39 | F | 35 | Sputum | ND | Chronic sinusitis maxillaris; atypical respiratory symptoms for 3 mo | − | 1/4 | ▴ | — | 128 | — |
F, female; M, male.
+, positive; −, negative.
MIF, microimmunofluorescence test.
•, ompA-based real-time PCR (this study); ▴, 16S rRNA based TETR assay (17); ⧫, PstI-based single-step assay (8); ▪, 16S rRNA-based nested PCR (20); ∗, ompA based nested touchdown (24).
ND, not done.
—, no antibodies of the respective class detectable.
DISCUSSION
Due to a lack of consensus on NAAs for the detection of C. pneumoniae, the decisions related to the PCR protocol that should be followed for various clinical specimens in the daily clinical routine and the best means of controlling the methodology in order to ensure reliable results are difficult. The main problems involve sensitivity and, in particular, specificity as well as the reproducibility of the results (1, 6, 18, 22). However, a few conventional PCR assays were suggested to meet at least some of the criteria necessary to create highly accurate results (10), but as far as we know, these PCRs have not been systematically compared with each other and have not been evaluated with respiratory specimens from symptomatic patients.
Besides the various conventional assays, the real-time PCR technology is an additional diagnostic tool that offers numerous advantages such as an automated, nonnested format including template inactivation and the combination of amplification, hybridization with a highly specific probe, and product detection in a single step. Hence, once a PCR is set up, the manipulation of tubes which potentially contain highly positive liquid becomes unnecessary. Furthermore, these advantages are offered without a loss of sensitivity or specificity and offer the additional option of quantifying the PCR product and detecting inhibition by comparison of the amplification plots for samples spiked with C. pneumoniae and unspiked samples.
For the reasons mentioned above, the real-time PCR technology for the detection of C. pneumoniae has entered use in microbiology laboratories. However, only a few centers have published data on how protocols with various patient specimens were evaluated in terms of specificity, sensitivity, and the reproducibility of the test results (16, 23). The use of various patient specimens seems to be most important, as it has been shown that the analytical sensitivity of an assay for detection of C. pneumoniae does not necessarily predict the ability to detect its target in clinical specimens (6, 18).
Because the ompA gene of C. pneumoniae is highly conserved within the species (14), it was chosen as the target for the real-time PCR presented in this study. In order to ensure the specificity of our target sequence (QM85), the DNAs of eight chlamydial strains as well as numerous representative organisms likely to be present in the respiratory tract, either as pathogens or as part of the normal flora, were repeatedly tested by the new protocol, but only C. pneumoniae strains could be amplified. In addition, alignments done with sequences in public databases revealed no matches other than those for sequences of the gene encoding the C. pneumoniae MOMP (Fig. 1). The ompA real-time PCR presented in this study produced highly sensitive and quantitative results that could be reproduced for all panels tested.
By interassay comparison, each of the five assays detected purified C. pneumoniae DNA down to very low levels and detected DNA in the mock-infected sputum, and depending on the target level, the results could be reproduced in most cases. In general, all PCRs detected target levels from spiked controls lower than those detected from purified DNA. A likely explanation could be that the DNA used to test the analytical sensitivity was prepared from purified EBs. In contrast and to mimic specimens from real patients as much as possible, mock-infected sputum was prepared by spiking it with whole C. pneumoniae-infected HEp-2 cells instead of purified DNA only. Thus, for the DNA panel, because the EBs were purified, extracted without the addition of carrier DNA, and then diluted, particles as well as target DNA might have been lost during purification and DNA extraction, respectively. In contrast, DNA extracts derived from the mock-infected sputum panel contained DNA from the sputum specimen itself plus HEp-2 cell DNA (in addition to the chlamydial targets), which probably acted as a carrier. In addition, the presence of free DNA or nonviable organisms could have played a role and could have resulted in different detection limits.
For the purified DNA panel, the ompA-based assays produced the most reproducible sensitive results, with a detection limit of three particles (mean CT value, 38.5) for the real-time assay. These findings are in concordance with those reported by Mahony et al. (18), who compared the performances of various conventional PCRs for the detection of C. pneumoniae and found that the PCR of Tong and Sillis (24) (ompA-based nested touchdown PCR) produced the most sensitive results with purified DNA and peripheral blood mononuclear cells from patients with coronary heart disease. In contrast, Tondella et al. (23) found that their 16S rRNA-based nested PCR (19) as well as their newly described ompA-based VD4 real-time assay were superior to the nested PCR protocol of Tong and Sillis (24) when the assays were applied to C. pneumoniae or peripheral blood mononuclear cell specimens from stroke patients. Our data obtained with C. pneumoniae-spiked sputum are in agreement with those of Tondella et al. (23): both 16S rRNA-based conventional PCRs (17, 20) and the ompA-based real-time assay were superior in terms of sensitivity and the reproducibility of the results.
Chernesky et al. (9) recently described the analytical performance of a new industry-derived C. pneumoniae PCR research kit (the LCx assay; Abbott Laboratories). They found that it was more sensitive and reproducible than all of the in-house NAAs. These were basically conventional PCRs, the performances of which were compared with each other by using numerous C. pneumoniae strains in a multicenter setting. Interestingly, the investigators noted that the overall sensitivities of the conventional NAAs within the participating centers varied markedly, although extraction was standardized and three of six testing sites used identical primers. Interassay and intercenter variations must be considered and could partly explain the differences in sensitivity observed in our study for assays targeting different templates when they were applied to different test panels. For the various test series, it remains to be clarified to what extent the target gene or, rather, the format of an assay influences test performance. Even without taking into account the possibility of calculating chlamydial loads, assay comparison studies involving real-time PCR protocols (16, 23) have shown the overall superiority of this technique over conventional PCRs. Nevertheless, as suggested by Chernesky et al. (9), having access to an industry-produced highly sensitive assay in kit form such as the LCx C. pneumoniae RUO PCR from Abbott Laboratories should ensure consistent performance. However, an evaluation of this promising NAA with large numbers of clinical specimens is still missing. It would be of interest and importance to directly compare this LCx test with PCR assays based on real-time technology, preferably in a multicenter comparison trial analyzing large numbers of patient specimens.
In our hands, the least sensitive assay was the single-step protocol (8). Interestingly, it also detected as few as 10−3 IFU from spiked sputum in three of four replicates tested, which might be more than sufficient in the case of a respiratory tract infection. Compared to the results of the PCR of Tong and Sillis (24) with the spiked sputum panel, which demonstrated greater variability, with the number of positive replicates going up as the target concentration decreased, better reproducibility was achieved with the single-step PstI assay (Fig. 3). The presence of inhibitors might partly explain this phenomenon; however, it is unlikely that inhibition would have influenced the sensitivity as well as the reproducibility of the results of the PCR of Tong and Sillis (24) exclusively. Amplicon carryover, on the other hand, is a known insidious problem connected with the application of conventional PCRs. However, excessive care was taken to avoid contamination, which, nevertheless, would not have explained the poor performance of the PCR assay of Tong and Sillis (24) with dilutions containing higher target levels (e.g., the assay did not detect 1 and 10−1 IFU in four of four replicates tested; Fig. 3).
Finally, the performances of the four conventional PCRs and the ompA-based real-time PCR assay were assessed with respiratory specimens from 45 patients with acute respiratory tract infection. For these patients, cell culture and serological data were also available. As shown in Fig. 4, the results obtained with these unknown clinical samples were less concordant. The discrepant results of the various PCR assays were disappointing and similar to the results reported for endarterectomy samples (1). We were, however, not able to calculate the sensitivities and specificities of these five assays because of a lack of a “gold standard” or because of the availability of only a poor gold standard (cell culture) and because only 45 specimens were analyzed. In addition, neither the clinical picture, serology, nor other investigational results were useful in the context of defining a patient's respiratory tract infection as being truly due to C. pneumoniae.
The only exception was patient L5, from whom C. pneumoniae could be isolated by cell culture. The ompA-based real-time PCR as well as both of the 16S rRNA-based conventional assays detected C. pneumoniae DNA in sputum in all four replicates of each of the assays, which was in concordance with the findings reported for mock-infected sputa. Two of the four replicates of the PCR of Tong and Sillis (24) were positive. Hence, for sputum specimen L5 the overall concordance of positive results was quite good. Only the single-step PCR did not detect its target in specimen L5 in any of the four replicates tested. The selection of the target fragment as well as primers for amplification by this assay has been based on isolation of a 474-bp PstI fragment from several clones based on a search of a C. pneumoniae genomic library database conducted 10 years ago. Compared to the degrees of conservation of the 16S rRNA and the ompA gene, there are few data on the degree of conservation of this region within the species. The presence of inhibitors in these replicates tested by the PstI-based assay is not very likely, since identical aliquots were tested by real-time PCR and no inhibitors were found to be present.
As shown by the real-time protocol, the amount of chlamydiae in specimen L5 was calculated to be about 1 × 102 to 1 × 103 IFU/5 μl of DNA extract or 3 × 103 to 3 × 104 IFU/100 μl of sputum. The high chlamydial load in this specimen is in agreement with data recently reported by Kuoppa et al. (16), who compared their ompA-based real-time PCR with the PCR assay of Tong and Sillis (24). They confirmed that sputum (if it can be produced) is the material of choice for analysis in the case of C. pneumoniae respiratory tract infection, with chlamydial loads being an average of 106 copies/ml of sputum. Calculated as the number of particles, our culture-confirmed C. pneumoniae-positive sputum specimen (specimen L5) contained about 108 particles/ml. Therefore, in terms of sensitivity, the PstI-based single-step PCR should have detected it, too.
A second sputum sample (sample L25) tested positive in two of four replicates of the real-time assay. However, the CT values were within a range corresponding to those for the negative controls of the DNA panel. Calculations based on the standard graphs revealed that less than 10−6 IFU (or less than one particle per ml of sputum) was present, which might not be a relevant finding at all, at least in the case of a respiratory tract infection. On the other hand, the patient from whom this specimen was obtained, a 31-year-old woman presenting with symptoms of atypical pneumonia, had a detectable C. pneumoniae-specific IgM antibody titer of 1:16 (Table 3). If the results of all testing attempts as well as the clinical picture are taken into account, the findings for patient L25 might reflect early, acute C. pneumoniae (primary) infection. On the other hand, this result for sputum sample L25 could not be confirmed by any of the other assays. Melting curve analysis, which can be done with hybridization probes or molecular beacons (used instead of TaqMan probes), could have added additional information in the context of the specificity of the PCR product for sputum sample L25.
In summary, 8 of 45 (18%) patients had at least one positive result by at least one assay. However, as shown in Fig. 4 and also in Table 3, the PCR results did not correlate well and the patients' histories as well as the serology results could not corroborate the positive PCR results in most cases; e.g., the results of four of four replicate tests were positive for one patient (patient W32) by the least sensitive assay, the single-step PstI-based assay, and these results were confirmed by three of four replicates of the 16S rRNA TETR assay, but the patient had no detectable C. pneumoniae-specific antibodies. For two further patients (patients L7 and L17) with low IgG antibody titers, the single-round PstI-based PCR was repeatedly positive (four of four replicates), but this was the only assay by which they were positive. Unfortunately, a second serum sample was not available, but considering the fact that both patients had suffered from symptoms for months, seroconversion should have occurred by the time that the patients presented.
All negative controls (65 to 100 controls/panel; ENCs and controls at various levels in the PCR mixture) were correctly identified in the present study, and the same experienced technologist performed all experiments done in this study. Thus, interperson variations should not be a reason for the differences in the results. One additional point must be considered, however: the majority of specimens analyzed in this study were sputum specimens. Further studies should evaluate whether the overall detection rate could be improved and whether a better concordance of positive findings between assays could be achieved by using throat swab specimens instead of sputum specimens.
In conclusion, the ompA-based real-time PCR presented here seems to be a promising new NAA for the quantitative detection of C. pneumoniae, but further studies evaluating this assay with clinical specimens are necessary and should, preferably, be done in multicenter settings under standardized conditions with identical standards for quantification. In addition to methodological standardization efforts, further extensive investigations are required in order to learn the meaning of CT values and how these quantitative results should be interpreted for the various clinical conditions that C. pneumoniae may cause.
REFERENCES
- 1.Apfalter, P., F. Blasi, J. Boman, C. A. Gaydos, M. Kundi, M. Maass, A. Makristathis, A. Meijer, R. Nadrchal, K. Persson, M. L. Rotter, C. Y. W. Tong, G. Stanek, and A. M. Hirschl. 2001. Multicenter comparison trial of DNA extraction methods and PCR assays for detection of Chlamydia pneumoniae in endarterectomy specimens. J. Clin. Microbiol. 39:519-524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Apfalter, P., J. Boman, M. Loidl, H. Hienerth, A. Makristathis, J. Pauer, F. Thalhammer, B. Willinger, M. L. Rotter, and A. M. Hirschl. 2001. Application of blood-based polymerase chain reaction for detection of Chlamydia pneumoniae in acute respiratory tract infections. Eur. J. Clin. Microbiol. Infect. Dis. 20:584-586. [DOI] [PubMed] [Google Scholar]
- 3.Apfalter, P., M. Loidl, R. Nadrchal, A. Makristathis, M. Rotter, M. Bergmann, P. Polterauer, and A. M. Hirschl. 2000. Isolation and continuous growth of Chlamydia pneumoniae from arterectomy specimens. Eur. J. Clin. Microbiol. Infect. Dis. 19:305-308. [DOI] [PubMed] [Google Scholar]
- 4.Balin, B. J., H. C. Gerard, E. J. Arking, D. M. Appelt, P. J. Branigan, J. T. Abrams, J. A. Whittum-Hudson, and A. P. Hudson. 1998. Identification and localization of Chlamydia pneumoniae in the Alzheimer's brain. Med. Microbiol. Immunol. 187:23-42. [DOI] [PubMed] [Google Scholar]
- 5.Boman, J., and C. A. Gaydos. 2000. Polymerase chain reaction detection of Chlamydia pneumoniae in circulating white blood cells. J. Infect. Dis. 181(Suppl. 3):S452-S454. [DOI] [PubMed] [Google Scholar]
- 6.Boman, J., C. A. Gaydos, and T. C. Quinn. 1999. Molecular diagnosis of Chlamydia pneumoniae infection. J. Clin. Microbiol. 37:3791-3799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Boman, J., and M. Hammerschlag. 2002. Chlamydia pneumoniae and atherosclerosis: critical assessment of diagnostic methods and relevance to treatment studies. Clin. Microbiol. Rev. 15:1-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Campbell, L. A., M. P. Melgosa, D. J. Hamilton, C. C. Kuo, and J. T. Grayston. 1992. Detection of Chlamydia pneumoniae by polymerase chain reaction. J. Clin. Microbiol. 30:434-439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chernesky, M., M. Smieja, J. Schachter, J. Summergill, L. Schindler, N. Solomon, K. Campbell, L. A. Campbell, A. Cappuccio, C. Gaydos, S. Chong, J. Moncada, J. Phillips, D. Jang, B. J. Wood, A. Petrich, M. Hammerschlag, M. Cerney, and J. Mahony. 2002. Comparison of an industry-derived LCx Chlamydia pneumoniae PCR research kit to in-house assays performed in five laboratories. J. Clin. Microbiol. 40:2357-2362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dowell, S. F., R. W. Peeling, J. Boman, G. M. Carlone, B. S. Fields, J. Guarner, M. R. Hammerschlag, L. A. Jackson, C. C. Kuo, M. Maass, T. O. Messmer, D. F. Talkington, M. L. Tondella, S. R. Zaki, and the Chlamydia pneumoniae Workshop Participants. 2001. Standardizing Chlamydia pneumoniae assays: Recommendations from the Centers for Disease Control and Prevention (USA) and the Laboratory Centre for Disease Control (Canada). Clin. Infect. Dis. 33:492-503. [DOI] [PubMed] [Google Scholar]
- 11.Grayston, J. T., M. B. Aldous, A. Easton, S. P. Wang, C. C. Kuo, L. A. Campbell, and J. Altman. 1993. Evidence that Chlamydia pneumoniae causes pneumonia and bronchitis. J. Infect. Dis. 168:1231-1235. [DOI] [PubMed] [Google Scholar]
- 12.Hammerschlag, M. R., K. Chirgwin, P. M. Roblin, M. Gelling, W. Dumornay, L. Mandel, P. Smith, and J. Schachter. 1992. Persistent infection with Chlamydia pneumoniae following acute respiratory illness. Clin. Infect. Dis. 14:178-182. [DOI] [PubMed] [Google Scholar]
- 13.Hyman, C. L., P. M. Roblin, C. A. Gaydos, T. C. Quinn, J. Schachter, and M. R. Hammerschlag. 1995. Prevalence of asymptomatic carriage of Chlamydia pneumoniae in subjectively healthy adults: assessment by polymerase chain reaction-enzyme immunoassay and culture. Clin. Infect. Dis. 20:1174-1178. [DOI] [PubMed] [Google Scholar]
- 14.Jantos, C. A., S. Heck, R. Roggendorf, M. Sen-Gupta, and J. Hegemann. 1997. Antigenic and molecular analyses of different Chlamydia pneumoniae strains. J. Clin. Microbiol. 35:620-623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kuo, C. C., L. A. Jackson, and J. T. Grayston. 1995. Chlamydia pneumoniae (TWAR). Clin. Microbiol. Rev. 8:451-461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kuoppa, Y., J. Boman, L. Scott, U. Kumlin, I. Eriksson, and A. Allard. 2002. Quantitative detection of respiratory Chlamydia pneumoniae infection by real-time PCR. J. Clin. Microbiol. 40:2273-2274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Madico, G., T. C. Quinn, J. Boman, and C. A. Gaydos. 2000. Touchdown enzyme time release-PCR for detection of Chlamydia trachomatis, C. pneumoniae, and C. psittaci using the 16S and 16S-23S spacer genes. J. Clin. Microbiol. 38:1085-1093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mahony, J. B., S. Chong, B. K. Coombes, M. Smieja, and A. Petrich. 2000. Analytical sensitivity, reproducibility of results, and clinical performance of five PCR assays for detecting Chlamydia pneumoniae DNA in peripheral blood mononuclear cells. J. Clin. Microbiol. 38:2622-2627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Messmer, T. O., S. K. Skelton, J. F. Moronay, H. Daugharty, and B. S. Fields. 1997. Application of a nested, multiplex PCR to psittacosis outbreaks. J. Clin. Microbiol. 35:2043-2046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Nadrchal, R., A. Makristathis, P. Apfalter, M. Rotter, W. Trubel, I. Huk, P. Polterauer, and A. M. Hirschl. 1999. Detection of Chlamydia pneumoniae DNA in atheromatous tissues by polymerase chain reaction. Wien. Klin. Wochenschr. 111:153-156. [PubMed] [Google Scholar]
- 21.Roblin, M. R., W. Dumornay, and M. R. Hammerschlag. 1992. Use of HEp-2 cells for improved isolation and passage of Chlamydia pneumoniae. J. Clin. Microbiol. 30:1968-1971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Smieja, M., J. B. Mahony, C. H. Goldsmith, S. Chong, A. Petrich, and M. Chernesky. 2001. Replicate PCR testing and probit analysis for detection and quantitation of Chlamydia pneumoniae in clinical specimens. J. Clin. Microbiol. 29:1796-1801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Tondella, M. L. C., D. F. Talkington, B. P. Holloway, S. F. Dowell, K. Cowley, M. Soriano-Gabarro, M. S. Elkind, and B. S. Fields. 2002. Development and evaluation of real-time PCR-based fluorescence assays for detection of Chlamydia pneumoniae. J. Clin. Microbiol. 40:575-583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Tong, C. Y. W., and M. Sillis. 1993. Detection of Chlamydia pneumoniae and Chlamydia psittaci in sputum samples by PCR. J. Clin. Pathol. 46:313-317. [DOI] [PMC free article] [PubMed] [Google Scholar]