| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Respiratory and Systemic Infection Laboratory (RSIL), Health Protection Agency (HPA) Central Public Health Laboratory (CPHL), 61 Colindale Avenue, London NW9 5HT, UK 2Vaccine Evaluation Unit, Microbiology Department, Gloucestershire Royal Hospital, Gloucester GL1 3NN, UK
Correspondence Carmen L. Sheppard carmen.sheppard{at}hpa.org.uk
Received September 8, 2003
Accepted December 10, 2003
The development and clinical evaluation of a LightCycler PCR assay, including an internal process control (IPC), to detect the Streptococcus pneumoniae autolysin gene in clinical samples is reported. The assay was developed to provide a second target for use in conjunction with existing pneumolysin PCR assays to increase the reliability of non-culture PCR diagnosis of pneumococcal infection. Primers amplify a 173 bp fragment of the autolysin gene (lytA), which is detected by fluorescence-labelled hybridization probes. An IPC was designed to check for the presence of PCR inhibitors and loss of assay sensitivity. The IPC product was amplified by the lytA primers and detected by a second set of hybridization probes. The analytical specificity of the autolysin PCR assay was 100 % against 39 other bacterial species tested; these included related streptococci and other organisms. The assay, which could reliably detect 50 fg purified pneumococcal DNA per reaction, was capable of distinguishing between S. pneumoniae and atypical Streptococcus mitis and Streptococcus oralis strains known to contain the lytA gene. Using DNA extracts from a panel of EDTA bloods from patients with blood-culture-confirmed pneumococcal infection, the autolysin PCR had a sensitivity of 42.9 %, which was similar to a previously reported TaqMan pneumolysin PCR (43.8 %) run in parallel. Total agreement was shown between the autolysin assay and the pneumolysin TaqMan assay when used to test 23 culture-negative clinical samples, of which eight were positive by PCR, adding valuable clinical information. A specific autolysin-based LightCycler assay has been developed to complement pneumolysin PCR for the detection of S. pneumoniae in clinical samples. This should be a particularly useful tool for the rapid and sensitive diagnosis of pneumococcal meningitis, even after an antibiotic has been administered. However, poor sensitivity on blood samples limits its usefulness in other bacteraemic infections.
Abbreviations: CSF, cerebrospinal fluid; IPC, internal process control.
The sequence of the IPC PCR product and a comparison of autolysin crossing points and pneumolysin crossing thresholds are available as supplementary data in JMM Online.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTS AND DISCUSSIONConclusionACKNOWLEDGEMENTSREFERENCES |
|---|
PCR offers the potential for highly sensitive diagnosis of pneumococcal disease from virtually any sample type. PCR detection has been used in a number of previous studies (Corless et al., 2001; Lorente et al., 2000; Toikka et al., 1999). In these assays, the pneumolysin gene (ply) was the pneumococcal target used. However, when a positive result is obtained in samples considered negative by the established ‘gold standard’ (culture from blood), it is difficult to confirm that this is due to improved sensitivity rather than poor specificity. One solution is to test samples using a second PCR targeting an unrelated gene. Positive results from both targets in the same sample would enhance confidence in the validity of a positive result.
We report the development of a real-time PCR assay targeting the autolysin gene (lytA) and including an internal process control (IPC), an important development for clinical diagnosis. The assay was designed to run in parallel with an existing pneumolysin PCR assay, with the aim of increasing the reliability of non-culture pneumococcal diagnosis by PCR.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONConclusionACKNOWLEDGEMENTSREFERENCES |
|---|
A specificity panel consisting of 59 isolates, including 18 species of streptococci and examples of 21 other closely related species or examples of common pathogens, was assembled from the NCTC collection and field strains taken from the Health Protection Agency, Respiratory and Systemic Infection Laboratory (RSIL) culture collection (Table 1).
|
Study EDTA blood samples.
A total of 198 whole EDTA blood samples were obtained from microbiology laboratories in south-west England (the former Public Health Laboratory Service South West Group). Samples were taken from hospitalized patients with well-documented bacteraemic infections caused by S. pneumoniae (105) or other bacteria (93) (Table 2). The samples were taken immediately before or within 24 h of starting appropriate antibiotic treatment and stored frozen at -80 °C for up to several years until DNA extraction. All samples were tested blind and disease categories were not revealed until testing had been completed.
|
Culture-confirmed CSF samples.
Six recently collected CSF samples, from patients with suspected pneumococcal meningitis who had a blood culture isolate of S. pneumoniae, were made available specifically for evaluation of the assays. One of the CSF samples was taken after administration of cefotaxime; however, no data were available on antibiotic administration for the other patients.
Culture-negative clinical samples.
During the period of assay development and evaluation, 18 CSF samples, two samples of post-mortem lung, one bronchial biopsy, one EDTA blood and one sputum sample were received by RSIL for investigation of clinical illness (meningitis or pneumonia). All of these samples were negative by conventional culture methods.
Extraction of DNA from bacterial cultures.
The positive control strain was extracted using a Nucleon DNA extraction kit (Promega), according to the manufacturer's instructions. The DNA pellets were resuspended in 50 µl TE buffer (Sigma-Aldrich).
The specificity and serotype panel strains were extracted using the Roche MagNAPure robot using DNA Isolation kit III (Roche Diagnostics) according to the manufacturer's instructions.
Extraction of DNA from clinical samples.
EDTA blood samples were extracted using a DNA blood mini kit (Qiagen) following the manufacturer's instructions using the double volume protocol; samples (400 µl) were extracted using double volumes of lysis buffer, Qiagen protease and ethanol. The sample was eluted in 200 µl Qiagen elution buffer.
The sputum sample was treated with Sputasol (Oxoid) for 30 min at room temperature. Tissue samples (approx. 1 mm cubes) were ground in PBS (200 µl) prior to extraction. Treated sputum (200 µl) and tissue samples were extracted using the DNA blood mini kit (Qiagen) following the manufacturer's protocol for tissue samples.
CSF samples were extracted using the GenSpin DNA extraction kit (Whatman Biosciences) according to the manufacturer's instructions, except that the DNA was eluted in 100 µl water instead of 200 µl. This extraction method had previously been found to be effective and convenient for use on CSF samples (unpublished data).
Autolysin primers and probe design.
Primers were designed to amplify a 173 bp region of the autolysin gene (lytA) (GenBank accession no. AF345846). The primer sites chosen showed complete agreement with the 20 S. pneumoniae strain sequences that were available from GenBank, but had differences from sequences from Streptococcus mitis and Streptococcus oralis strains previously reported to contain the autolysin gene (Whatmore et al., 2000). The probe-binding sites also showed complete identity among the 20 S. pneumoniae strain sequences but significant differences from other streptococci. The sequences of the primers and probes are shown in Table 3.
|
Autolysin IPC.
Primers for the construction of the IPC contained the autolysin primer sequences external to sequences specific for a 264 bp portion of bacteriophage lambda DNA (Table 3). This gave a 302 bp fragment that could be amplified by the lytA primers, a similar construct to that used in previous studies (Povlsen et al., 1998; Sachadyn & Kur, 1998). The IPC fragment was cloned into plasmids using the TOPO TA cloning kit (Invitrogen/Life Technologies), which were then linearized by restriction with XbaI (Roche Diagnostics) and subsequently spiked into the master mix of all LightCycler reactions at a level of approximately 22 copies per reaction. The IPC was designed to be longer than the target product in order to favour target amplification.
The IPC product was detected by a second set of hybridization probes specific to the lambda sequence contained within the fragment (Table 3). The sequence of the final IPC product amplified by the autolysin primers is available as supplementary data in JMM Online.
LightCycler Autolysin PCR.
After optimization experiments, the following PCR protocol was adopted. The reaction mixture consisted of 1x FastStart hybridization probe reaction mixture (Roche Diagnostics), which contains dUTP instead of TTP, allowing for PCR product contamination removal (Rys & Persing, 1993; Udaykumar et al., 1993), 3 mM MgCl2 (Roche Diagnostics), LytA-F and LytA-R primers (0.5 µM each), LytA-DNR probe (0.2 µM), LytA-ACR probe (0.35 µM), IPC-DNR probe (0.25 µM), IPC-ACR probe (0.25 µM), lytA IPC (0.1 fg, approx. 22 copies) and heat-labile uracil DNA-glycosylase (1 U) (Roche Diagnostics). An aliquot (2 µl) of the extracted DNA sample was added to the reaction capillary (Roche Diagnostics) containing 18 µl reaction mixture. Later experiments confirmed that 5 µl sample in 15 µl mastermix could be used without increasing inhibition.
Each LightCycler run contained five standard serial dilutions of purified pneumococcal DNA ranging in concentration from 500 pg (approx. 20 000 genome equivalents) to 50 fg (approx. 20 genomes) per reaction and a ‘no template’ control.
The LightCycler PCR parameters were: initial denaturation at 95 °C for 10 min and 45 cycles of denaturation at 95 °C for 10 s, annealing at 60 °C for 9 s with single fluorescence acquisition and elongation at 72 °C for 8 s, all with a temperature ramp rate of 20 °C s-1. A melting curve program was used to check the specificity of the probe binding. This consisted of heating to 95 °C for 0 s and then 40 °C for 2 s followed by slow ramping (0.1 °C s-1) with continuous fluorescence acquisition up to 70 °C. The mean melting temperature of the lytA probes was 61.7 °C.
The data were analysed using Roche LightCycler software version 3.5. Samples were recorded as positive where a crossing point (Cp) of less than 45 cycles was noted and the melting curve analysis revealed the probe melting temperature peak to be above 60 °C. Any samples that did not give a positive result and in which the IPC did not amplify were recorded as inhibitory and were repeated undiluted (to check for technical errors) and diluted 1/5 and 1/10 in nuclease-free water (Promega).
Pneumolysin TaqMan assay.
The pneumolysin TaqMan assay was adapted from a previously reported multiplex assay (Corless et al., 2001) as described elsewhere (Sheppard et al., 2003). The plate was cycled on an Applied Biosystems 7700 sequence detection system according to the reported conditions.
|
RESULTS AND DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONConclusionACKNOWLEDGEMENTSREFERENCES |
|---|
Analytical sensitivity
In multiplex with the IPC, the assay reported here could reliably detect 50 fg positive-control DNA (equivalent to approx. 20 genome copies) per reaction. Smaller amounts (down to a single genome equivalent) were detected in some experiments.
Titration of the IPC showed that it gave reliable results at 22 copies per reaction (crossing point 
38 cycles), whilst still allowing detection beyond the lowest standard target concentration. All strains of S. pneumoniae from the serotype panel gave positive amplification signals, although one strain of serotype 8 showed an 1 °C decrease compared with the mean probe melting Tm, which was due to a single base change in the binding sequence.
Analytical specificity
Theoretical specificity was ascertained by a BLAST (Altschul et al., 1990) search of the probe and primer sequences in which no exact match was found with any non-pneumococcal sequence.
DNA extracts (
10 ng µl-1) from the organisms in the specificity panel (Table 1) were tested using the autolysin assay. No positive results were obtained from any of the organisms, including S. mitis and S. oralis strains carrying a lytA gene.
Study EDTA blood samples
Overall PCR sensitivity of the LightCycler autolysin assay on the stored blood samples was very similar to that of the pneumolysin TaqMan assay, but, for both assays, the sensitivity was disappointing. Of 105 pneumococcal blood-culture-positive samples, 45 were positive in the autolysin LightCycler assay (42.9 %) and 46 positive in the pneumolysin TaqMan assay (43.8 %). Of these positive samples, only 35 samples (33.3 %) were positive by both autolysin and pneumolysin.
A comparison of autolysin crossing points (Cp) and pneumolysin crossing thresholds (Ct) for the EDTA blood samples is available as supplementary data in JMM Online. The crossing points from EDTA blood extracts showed good agreement at numbers of cycles below the 38 cycle Cp of the 50 fg standard. However, as the concentration of target DNA decreased, the degree of disparity between the two assays increased. As this is likely to happen in samples from which cultures are likely to be negative and a dual-target confirmed positive result would be most useful, a strategy of testing involving replication of samples and repeated testing of discrepant results might have to be employed. This is a previously reported method of increasing PCR sensitivity from samples where the starting DNA concentration is very low (Smieja et al., 2001).
From our results, blood samples appear to contain very small numbers of organisms. Disappointing PCR sensitivity when testing EDTA blood samples may therefore be due to inefficiencies of DNA extraction methods in combination with small sampling volumes and length of storage. Spiking the samples with the IPC at extraction would give an indication of the extraction efficiency, but this application of the IPC has not yet been investigated. A previous PCR study of whole blood samples that were stabilized with lysis buffer containing guanidinium within 3 h of collection and prior to freezing reported higher sensitivity than we have achieved (Michelow et al., 2002). This indicated that greater PCR sensitivity might be possible if samples were processed promptly after collection, as would be the case in routine use.
Eight of 93 non-pneumococcal, blood-culture-positive samples gave a positive result in the autolysin assay on blind testing (three with Escherichia coli, one group B streptococcus, one group G streptococcus, two Staphylococcus aureus and one with Enterococcus faecalis isolated). Given the theoretical specificity of this assay and the fact that all of the bacterial species had representatives in the specificity panel, it seems unlikely that the false-positive results were due to specific but erroneous amplification and detection by the lytA primers and probes. Alternative and more likely explanations include dual infection or contamination with S. pneumoniae DNA during the collection and processing of the samples. In support of this, one of the two non-pneumococcal, blood-culture positive samples that were positive by the pneumolysin TaqMan assay was also positive in the autolysin assay. This dual-target positive sample remained positive in both assays on repeat testing, giving crossing values of around 35 cycles in both. The sample was taken from a patient with Staphylococcus aureus septicaemia. Staphylococcus aureus is very likely to have out-competed any pneumococcal growth and this may therefore have been a case of dual infection. Dual infections have been suggested in previous pneumococcal PCR evaluations as a possible source of apparent false-positive results (Michelow et al., 2002).
The DNA extracts from the ‘false-positive’ samples were repeat-tested in both assays and a positive result was recorded for only one of the eight samples in the autolysin assay; however, both of the two pneumolysin-positive samples were again positive on repeat.
Assay sensitivity on CSF samples
The LightCycler PCR can provide results in approximately 2 h from sample receipt. This means it could be ideal as an urgent diagnostic test for CSF samples from patients with suspected pneumococcal meningitis. The six CSF samples from patients with blood culture isolates of pneumococci all gave strong positive results (Cp/Ct values less than 35 cycles) in both autolysin and pneumolysin assays, even though antibiotics may have been administered.
The Binax NOW pneumococcal antigen test can offer a result on CSF samples in approximately 15 min; however, confirmation of the result by PCR would greatly increase confidence in the diagnosis, as it has not been proven definitively that high-level carriage in children does not affect the results of the Binax test when used on CSF, as it does in urine (Dowell et al., 2001).
Culture-negative clinical samples
Autolysin and pneumolysin assay results showed complete agreement when used to test the culture-negative clinical samples (Table 4). Of the 18 clinical CSF samples, five were positive by both assays and the remainder were negative by both assays. Both of the post-mortem lung samples gave positive results in both assays, as did the sputum sample. The bronchial biopsy and EDTA blood sample were negative by both assays.
|
IPC indication of PCR inhibition
Clinical samples (especially whole blood) may contain substances that inhibit PCR (Al-Soud & Rådström, 2001; Al-Soud et al., 2000). The DNA extraction procedure may not remove these substances completely from the sample and, therefore, inhibition of PCR may give false-negative results. As the IPC in this assay is run at close to the detection limit of the system, general non-amplification may also indicate that the sensitivity of the PCR assay has been affected by suboptimal reaction mixture components or conditions.
Non-amplification of the IPC identified five study blood samples that were inhibitory to PCR and for which a negative result could not be confidently scored. The samples were retested at 1/5 and 1/10 dilutions, and all but one then gave a normal IPC amplification curve; the remaining sample remained inhibitory even at 1/10 dilution. None of the samples gave a positive result for autolysin after dilution and none was positive in the pneumolysin TaqMan assay.
Atypical S. mitis and S. oralis strains
The specificity panel included examples of atypical S. mitis and S. oralis strains (Whatmore et al., 2000) containing lytA and ply genes. No positive signal was obtained from these organisms in the autolysin assay, but they were positive in the pneumolysin assay. When the autolysin reaction products were analysed by agarose gel electrophoresis, bands of the size corresponding to the autolysin gene fragment were seen (data not shown). This was found to be due to several base pair differences in the probe-binding region, which meant the detection probes could not bind to the product.
This observation may suggest an additional role for the autolysin assay in determining the identity of atypical oral streptococcal and pneumococcal isolates. Autolysin reaction products from isolates (or clinical samples) giving a non-resolvable pneumolysin-positive, autolysin-negative result would be analysed by agarose gel electrophoresis. The presence of bands of the correct size may indicate the detection of atypical isolates rather than typical pneumococci.
|
Conclusion |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONConclusionACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONConclusionACKNOWLEDGEMENTSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONConclusionACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. P. H. Peters, R. F. de Boer, T. Schuurman, S. Gierveld, M. Kooistra-Smid, M. A. van Agtmael, C. M. J. E. Vandenbroucke-Grauls, M. C. J. Persoons, and P. H. M. Savelkoul Streptococcus pneumoniae DNA Load in Blood as a Marker of Infection in Patients with Community-Acquired Pneumonia J. Clin. Microbiol., October 1, 2009; 47(10): 3308 - 3312. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. K. Fry, J. Duncan, K. Wagner, O. Tzivra, N. Doshi, D. J. Litt, N. Crowcroft, E. Miller, R. C. George, and T. G. Harrison Role of PCR in the diagnosis of pertussis infection in infants: 5 years' experience of provision of a same-day real-time PCR service in England and Wales from 2002 to 2007 J. Med. Microbiol., August 1, 2009; 58(8): 1023 - 1029. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Smith, C. L. Sheppard, A. Hogan, T. G. Harrison, D. A. B. Dance, P. Derrington, R. C. George, and on behalf of the South West Pneumococcus Study Gro Diagnosis of Streptococcus pneumoniae Infections in Adults with Bacteremia and Community-Acquired Pneumonia: Clinical Comparison of Pneumococcal PCR and Urinary Antigen Detection J. Clin. Microbiol., April 1, 2009; 47(4): 1046 - 1049. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. S. Kim, D.-M. Kim, G. P. Neupane, Y.-m. Lee, N.-W. Yang, S. J. Jang, S.-I. Jung, K.-H. Park, H.-R. Park, C. S. Lee, et al. Comparison of Conventional, Nested, and Real-Time PCR Assays for Rapid and Accurate Detection of Vibrio vulnificus J. Clin. Microbiol., September 1, 2008; 46(9): 2992 - 2998. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Y. Fukushima, K. Yanagihara, Y. Hirakata, K. Sugahara, Y. Morinaga, S. Kohno, and S. Kamihira Rapid Identification of Penicillin and Macrolide Resistance Genes and Simultaneous Quantification of Streptococcus pneumoniae in Purulent Sputum Samples by Use of a Novel Real-Time Multiplex PCR Assay J. Clin. Microbiol., July 1, 2008; 46(7): 2384 - 2388. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. d. G. S. Carvalho, M. L. Tondella, K. McCaustland, L. Weidlich, L. McGee, L. W. Mayer, A. Steigerwalt, M. Whaley, R. R. Facklam, B. Fields, et al. Evaluation and Improvement of Real-Time PCR Assays Targeting lytA, ply, and psaA Genes for Detection of Pneumococcal DNA J. Clin. Microbiol., August 1, 2007; 45(8): 2460 - 2466. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Y. Fukushima, Y. Hirakata, K. Sugahara, K. Yanagihara, A. Kondo, S. Kohno, and S. Kamihira Rapid Screening of Topoisomerase Gene Mutations by a Novel Melting Curve Analysis Method for Early Warning of Fluoroquinolone-Resistant Streptococcus pneumoniae Emergence J. Clin. Microbiol., December 1, 2006; 44(12): 4553 - 4558. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. Leeming, K. Cartwright, R. Morris, S. A. Martin, M. D. Smith, and on behalf of the South-West Pneumococcus Study Gro Diagnosis of Invasive Pneumococcal Infection by Serotype-Specific Urinary Antigen Detection J. Clin. Microbiol., October 1, 2005; 43(10): 4972 - 4976. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Yang, S. Lin, A. Khalil, C. Gaydos, E. Nuemberger, G. Juan, J. Hardick, J. G. Bartlett, P. G. Auwaerter, and R. E. Rothman Quantitative PCR Assay Using Sputum Samples for Rapid Diagnosis of Pneumococcal Pneumonia in Adult Emergency Department Patients J. Clin. Microbiol., July 1, 2005; 43(7): 3221 - 3226. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
