Superiority of Molecular Techniques for Identification of Gram-Negative, Oxidase-Positive Rods, Including Morphologically Nontypical Pseudom
http://www.100md.com
微生物临床杂志 2005年第8期
Department of Medical Microbiology and Hygiene, University of Ulm, Ulm, Germany
ABSTRACT
Phenotypic identification of gram-negative bacteria from Cystic Fibrosis (CF) patients carries a high risk of misidentification. Therefore, we compared the results of biochemical identification by API 20NE with 16S rRNA gene sequencing in 88 gram-negative, oxidase-positive rods, other than morphologically and biochemically typical P. aeruginosa, from respiratory secretions of CF patients. The API 20NE allowed correct identification of the bacterial species in 15 out of 88 (17%) isolates investigated. Agreement between the API and the 16S rRNA gene sequencing results was high only in isolates with an API result classified as "excellent identification." Even API results classified as "very good identification" or "good identification" showed a high rate of misidentification (67% and 84%). Fifty-two isolates of morphological and biochemical nontypical Pseudomonas aeruginosa, representing 59% of all isolates investigated, were not identifiable or misidentified in the API 20NE. Therefore, rapid molecular diagnostic techniques like real-time PCR and fluorescence in situ hybridization (FISH) were evaluated in this particular group of bacteria for identification of the clinically most relevant pathogen, P. aeruginosa. The LightCycler PCR assay with a P. aeruginosa-specific probe showed a sensitivity and specificity of 98.1% and 100%, respectively. For FISH analysis, a newly designed P. aeruginosa-specific probe had a sensitivity and specificity of 100%. In conclusion, molecular methods are superior over biochemical tests for identification of gram-negative, oxidase-positive rods in CF patients. In addition, real-time PCR and FISH allowed identification of morphologically nontypical isolates of P. aeruginosa within a few hours.
INTRODUCTION
Chronic bacterial colonization of the respiratory tract, leading to exacerbations of pulmonary infection, is the major cause of disease and death in Cystic Fibrosis (CF) patients. The most common pathogen in respiratory secretions of CF patients is Pseudomonas aeruginosa, and Staphylococcus aureus, Haemophilus influenzae, and members of the Burkholderia cepacia complex also play an important role in CF lung disease (13, 14, 17). Other gram-negative glucose nonfermenters such as Achromobacter xylosoxidans, Ralstonia pickettii, and Stenotrophomonas maltophilia are also occasionally recovered from CF respiratory samples, but their pathogenic significance remains to be fully clarified (4, 14, 17). Recent studies that applied molecular approaches for the identification of unusual pathogens in CF patients revealed the presence of various rarely or even newly described species belonging to the genera Bordetella, Comamonas, Inquilinus, Pandoraea, Ralstonia, etc. (2-5, 16, 22, 23). Determination of the clinical relevance of gram-negative bacteria other than P. aeruginosa in CF patients is, however, hampered by the difficult identification of these pathogens by conventional laboratory techniques.
Phenotypic identification of bacteria grown from CF patients carries a high risk of misidentification regarding both manual methods and commercial identification systems (12, 16, 18, 19, 21). In a study comparing various commercial systems for the identification of B. cepacia and P. aeruginosa isolated from CF patients, the API 20NE system was superior to other systems but misidentification or failure of identification of Burkholderia gladioli and Ralstonia pickettii occurred in a significant percentage of isolates (21).
In our experience, species like Agrobacterium radiobacter, Comamonas testosteroni, Pseudomonas alcaligenes, Ochrobactrum anthropi, and Photobacterium damsela are at times identified by the API 20NE system in respiratory specimens from CF patients. Detection of these species often leads to uncertainties concerning indication and choice of therapy. In addition, the prevalence of these species may be overestimated due to misidentification by biochemical identification systems.
Since correct identification is the prerequisite for assessing the clinical significance of rarely encountered species and for reliable detection of important pathogens like P. aeruginosa, we compared the biochemical identification results (API 20NE) of 88 gram-negative, oxidase-positive rods from respiratory secretions of CF patients with the results of 16S rRNA gene sequencing. We could demonstrate that identification by API 20NE yielded correct results in a minority of cases only. Therefore, rapid molecular diagnostic tests, i.e., real-time PCR and fluorescence in situ hybridization (FISH), were evaluated for identification of the most frequent and clinically relevant pathogen, P. aeruginosa.
MATERIALS AND METHODS
Clinical isolates and specimen processing. All respiratory secretions of CF patients (n = 82 patients) obtained during 2003 were included in the study. The specimens were plated on sheep blood agar, chocolate agar, MacConkey agar, and Burkholderia cepacia selective agar (BC agar, containing 100 mg/liter ticarcillin and 300,000 IU/liter polymyxin B; MAST Diagnostica) and were incubated for 48 h at 36°C. The BC agar plates were incubated another 5 days at room temperature. All gram-negative, oxidase-positive rods except those that were phenotypically identified as P. aeruginosa (green pigmentation or mucoidity, growth at 42°C, growth on MacConkey agar, and colistin susceptibility or at least good identification in the API 20NE; see also Fig. 1) were included in the study. Isolates that did not fulfill the morphological criteria for P. aeruginosa or mucoid P. aeruginosa as outlined in Fig. 1 and isolates that were identified in the API 20NE as a species other than P. aeruginosa or as P. aeruginosa with low reliability were subjected to 16S rRNA gene sequencing (see Fig. 1 and below) and partly represented "morphologically nontypical P. aeruginosa."
Phenotypic tests. Identification of gram-negative, oxidase-positive rods was performed according to the algorithm shown in Fig. 1. Oxidase activity was checked with 1% tetramethyl p-phenylenediamine dihydrochloride. API 20NE (BioMerieux) was performed according to the instructions of the manufacturer. The API stripes were incubated for 48 h at 30°C under ambient air. The results were interpreted with the APILAB PLUS software (version 3.3.3). Susceptibility testing against colistin was performed with 10-μg disks (BD) on Mueller-Hinton agar (Heipha). Isolates with zone of inhibition of <11 mm after incubation at 36°C for 16 to 24 h were regarded as resistant.
16S rRNA gene sequencing. DNA was prepared by using the QiAamp DNA Mini kit according to the instructions of the manufacturer. Eluted DNA was stored at –20°C until further use. The complete 16S rRNA gene was sequenced using the primers 16Sfor and 16Srev described previously (7). Sequence analysis was performed on a 310 Genetic Analyser (ABI Prism Biosystems) using a Dye Terminator Cycle Sequencing Ready Reaction kit (ABI Prism). The sequences were compared to sequences available in the GenBank database by using the BLAST algorithm. In addition, all sequences were compared to the sequences of the type strain of the respective species, including the following isolates (strain and GenBank accession numbers are in parentheses): Achromobacter xylosoxidans subsp. xylosoxidans (ATCC 9220, AF411021), Agrobacterium radiobacter (ATCC 19358, AJ389904), B. multivorans (LMG13010, Y18703), Ochrobactrum anthropi (LMG 5140, AJ242580), P. aeruginosa (ATCC 10145, AF094713), P. alcaligenes (ATCC 14909, AF094721), Pseudomonas brassicacearum (CFBP 11706, AF100321), Pseudomonas synxantha (IAM 12356, D84025), and Sphingomonas paucimobilis (ATCC 29837, U37337). Species identification was assumed when the base homologies between the clinical isolate and the type strain sequence exceeded 99.0%.
Burkholderia recA gene sequencing. Sequencing of the recA gene was done in all isolates that were identified as members of the Burkholderia cepacia complex by 16S rRNA gene sequencing. The complete recA gene (1,043 bp) was amplified with the primers BCR1 and BCR2 (15), and sequencing was performed with the primers BCR2 and BCR4 (15) as described above.
Pulsed-field gel electrophoresis (PFGE) of P. aeruginosa. In patients with detection of P. aeruginosa at more than one visit to the CF clinic, PFGE was performed for all isolates that were determined by 16S rRNA gene sequencing as P. aeruginosa. PFGE was done with the contour-clamped homogeneous electric field DRIII equipment (Bio-Rad) in 1% agarose at 14°C and a voltage of 200 V with a pulse-rate of 5 to 60 s, using the restriction enzyme SpeI. For interpretation of banding pattern the recommendations of Tenover et al. were followed (20).
Real-time LightCycler PCR for detection of P. aeruginosa. Approximately 5 to 10 colonies of an overnight culture of the strain were suspended in 500 μl of lysis buffer (1% Triton X-100, 0.5% Tween 20, 10 mM Tris-HCl, pH 8.0, 1 mM EDTA; all from Sigma) and boiled for 10 min at 95°C. After a centrifugation step of 5 min at 20,000 x g, the supernatant was used for PCR. The PCR was performed on a second-generation LightCycler instrument (Roche Diagnostics) as described previously, using the primers RWO1 and DG74 and the TaqMan probe Pseu aeru (24).
Fluorescence in situ hybridization (FISH). Some colonies of an overnight culture of the strain were taken with a cotton swab premoistened in 0.9% NaCl and transferred to a glass slide. Slides were fixed in 70% methanol for 10 min. For FISH analysis, two different oligonucleotide probes were used: the formerly published P. aeruginosa-specific probe Psae (5'-TCT CGG CCT TGA AAC CCC) (10), binding to the 23S rRNA and labeled with fluorescein isothiocyanate (and also with Cy3 in separate experiments) as fluorescent dye, and a newly designed P. aeruginosa-specific probe Psae16S-182 (5'-CCA CTT TCT CCC TCA GGA CG), binding to the 16S rRNA and labeled with Cy3. The new probe was designed by using the ARB program package, available at http://www.arb-home.de/. FISH was performed at 46°C using 40% formamide in the hybridization buffer and the corresponding salt concentrations in washing buffer as described elsewhere (1). The new probe was validated with the reference strain P. aeruginosa ATCC 27853 as well as with several clinical isolates of P. aeruginosa that were confirmed by 16S rRNA gene sequencing. Furthermore, other closely related species were tested, including Pseudomonas putida (clinical isolate), Pseudomonas stutzeri (ATCC 17588), Acinetobacter baumannii (ATCC 19606), Acinetobacter junii (ATCC 17908), and Sphingomonas paucimobilis (clinical isolate), and no cross-reactions were observed. In addition, a eubacterial EUB probe (EUB 388 5'-GCT GCC TCC CGT AGG AGT) that detects all bacterial species (1) was used as a control. The stained samples were rapidly air dried, embedded, and examined with a Zeiss Axioplan 2 microscope. Samples with a fluorescence intensity clearly exceeding the background were regarded as positive. A 4',6'-diamidino-2-phenylindole stain (Hoechst) was done in order to ensure the presence of bacteria on the slide and was positive in all samples.
RESULTS
Identification by API 20NE. During the study period, 405 isolates of gram-negative, oxidase-positive rods were isolated from 82 CF patients. Two-hundred ninety-nine of these isolates were identified as P. aeruginosa (mucoid or nonmucoid) based on typical growth characteristics (11) and the diagnostic algorithm outlined in Fig. 1, and were identified as P. aeruginosa by API 20NE. These included 2 isolates with "excellent" API result, 12 isolates with "very good" API result, and 4 isolates with "good" API result. Eighty-eight isolates (22%) were subjected to API 20NE and 16S rRNA gene sequencing according to the algorithm applied (Fig. 1). Of these 88 isolates, the API result showed excellent, very good, or good identification of one the following species or genera in only 41 isolates (47%): Agrobacterium radiobacter, Achromobacter xylosoxidans, Burkholderia cepacia, Comamonas testosteroni or P. alcaligenes complex, Pasteurella spp., Photobacterium damsela, Pseudomonas spp., P. fluorescens, Psychrobacter phenylpyruvicus, and Ralstonia pickettii (Table 1). In 11 isolates, the API indicated acceptable identification of a species (Table 1). In 36 isolates (41%), the API result was unacceptable (n = 4), unreliable (n = 4), or ambiguous (n = 3) or had low selectivity (n = 25).
16S rRNA gene sequencing. 16S rRNA gene sequencing allowed identification to the species level in 87 out of 88 isolates and to the genus level (Chryseobacterium spp.) in one isolate. Of the seven species that were excellently identified by API, five were confirmed by 16S rRNA gene sequencing. However, both isolates identified as P. fluorescens in the API (Table 1) were identified as P. synxantha and P. brassicacearum by sequencing, both representing species that are not included in the API database. Concerning those isolates showing very good identification by API, only 5 of the 12 isolates that were identified as Burkholderia cepacia could be confirmed by 16S rDNA sequencing. Eight isolates of Achromobacter xylosoxidans were misidentified in the API as either B. cepacia or P. fluorescens (Table 1). Concerning those 19 isolates showing good identification by API, only two isolates of Achromobacter xylosoxidans and one isolate of Ochrobactrum anthropi were correctly identified. Fifteen isolates of P. aeruginosa were misidentified as either Comamonas testosteroni or P. alcaligenes complex, Photobacterium damsela, P. fluorescens, Psychrobacter phenylpyruvicus, or Ralstonia pickettii (Table 1). Regarding those isolates with acceptable identification by API, a single isolate of Burkholderia cepacia complex was identified correctly (Table 1). Concerning those isolates with low selectivity, unreliable, ambiguous, or unacceptable identification (n = 36), 16S rRNA gene sequencing confirmed the species that was proposed by the API software as the closest match in two cases only (Table 1), i.e., one isolate each of Burkholderia cepacia complex and Sphingomonas paucimobilis. Twenty-seven of the 36 isolates (75%) were identified as P. aeruginosa by 16S rRNA gene sequencing, and the other isolates included Achromobacter xylosoxidans (n = 4), Burkholderia cepacia complex (n = 1), Chryseobacterium spp. (n = 1), and Inquilinus limosus (n = 1). Altogether, 52 of the 88 isolates investigated in this study (59%) were identified as P. aeruginosa by 16S rRNA sequencing. Sequencing of the Burkholderia recA gene revealed that all isolates of B. cepacia complex found in this study belonged to the species B. multivorans.
Of the 18 isolates identified as P. aeruginosa by API 20NE (Fig. 1), 8 isolates were available for 16S rRNA gene sequencing and the sequencing result confirmed the API result. The eight isolates included one isolate with "excellent" API result, five isolates with "very good" API result, and two isolates with "good" API result.
Identification of persistent isolates of P. aeruginosa. In six patients, P. aeruginosa was detected by 16S rRNA gene sequencing on more than one occasion during the study period. PFGE of all isolates revealed persistent infection in five of these patients, and each patient carried a unique, persistent strain (data not shown). Interestingly, results of API 20NE differed considerably within the same strain. In Table 2, an overview of the interpretable results of the API 20NE in persistent strains is given. Ambiguous, unreliable, or unacceptable results as well as identifications with low selectivity were omitted.
Specific identification of P. aeruginosa by real-time PCR and FISH. Due to the predominance of P. aeruginosa among the isolates investigated in this study and due to the importance of correct identification of this pathogen in CF respiratory specimens, we tried to specifically detect P. aeruginosa by rapid molecular techniques, i.e., real-time PCR and FISH.
For real-time PCR, a formerly developed P. aeruginosa-specific TaqMan probe assay was run on the LightCycler (24). By using this PCR assay, 51 out of 52 isolates of P. aeruginosa were identified correctly, and none of the other 36 isolates was false positive, corresponding to a sensitivity of 98.1% and a specificity of 100%. In the one false-negative isolate, the primers produced the correct PCR product and the binding site of the probe was not mutated, as determined by sequencing of the amplicon. However, for generating sufficient amounts of PCR products for sequencing, Q-solution (Roche) that improves PCR results in GC-rich isolates had to be used in this sample.
For FISH, two different P. aeruginosa-specific probes were used. The formerly published probe Psae correctly identified 51 out of 52 isolates of P. aeruginosa and none of the other isolates, thus resulting in a sensitivity of 98.1% and a specificity of 100%. The one false-negative isolate in the FISH was different from the false-negative isolate in the PCR described above. The newly designed probe Psae16S-182 even detected all 52 isolates of P. aeruginosa and none of the other isolates, corresponding to a sensitivity and specificity of 100%.
DISCUSSION
Apart from P. aeruginosa and members of the B. cepacia complex, other gram-negative, oxidase-positive rods may probably also play roles in progression of CF lung disease. Identification of these bacterial species with biochemical methods is difficult due to low metabolic activity and morphological and biochemical variability of the isolates. Thus, in this study, molecular techniques were used to verify the results of the widely distributed biochemical identification system API 20NE for identification of unusual gram-negative, oxidase-positive rods isolated from CF patients. Isolates clearly identified as P. aeruginosa by commonly accepted morphological growth characteristics (n = 299) were not included in the study. In the study population of unusual gram-negative, oxidase-positive rods not including morphologically typical P. aeruginosa, the API 20NE allowed correct identification of the bacterial species in 15 of 88 (17%) isolates investigated (Table 1). Agreement between the API and the 16S rRNA gene sequencing result was high only in isolates with an API result classified as "excellent identification." Even API results classified as "very good identification" or "good identification" showed a high rate of misidentification (67% and 84%, respectively). The main restrictions of the API 20NE were misidentification of Achromobacter xylosoxidans as Burkholderia cepacia, Ochrobactrum anthropi, or Pseudomonas fluorescens and misidentification of P. aeruginosa as various species, including predominantly Comamonas testosteroni or P. alcaligenes complex, P. fluorescens, and Photobacterium damsela (Table 1). Difficulties of the API 20NE system to identify members of the Burkholderia cepacia complex have been described previously (6, 21). Misidentification of Achromobacter xylosoxidans is considerably worrying, since the relevance of this species in progression of CF lung disease is not yet fully understood and correct identification of the species is a prerequisite for assessing its clinical significance. Relying on API 20NE results when species other than P. aeruginosa are identified may lead to overestimation of the prevalence of unusual species and to uncertainties regarding therapy of such organisms. This applies also to Comamonas testosteroni and P. alcaligenes. Interestingly, all but one of the isolates identified by API 20NE as Comamonas testosteroni or P. alcaligenes complex, which represented 11% of all isolates investigated in the study, belonged to the species P. aeruginosa. Altogether, P. aeruginosa was identified by 16S rRNA sequencing in approximately 60% of isolates in this study. In patients persistently infected with P. aeruginosa, the API 20NE results varied markedly even within the same strain.
Isolates clearly identified as P. aeruginosa by the API 20NE were only partly (8 out of 18 isolates) available for 16S rRNA gene sequencing. Although sequencing confirmed the API results in these isolates, the specificity of the API 20NE regarding identification of P. aeruginosa cannot be determined from this study due to the small sample size and incomplete data. In conclusion, results of the API 20NE must be interpreted with great care in this particular population of nonfermentative, oxidase-positive rods obtained from CF patients.
Misidentification of P. aeruginosa has remarkable consequences for the patients and should be avoided by all means. Despite its exactness, DNA sequencing is not suitable for routine identification of P. aeruginosa in the diagnostic microbiology laboratory due to its high costs and work load. In addition, public DNA databases that do not involve quality control measures for the deposited sequences also contain faulty or incorrect assigned sequences, thus requiring an experienced user for interpretation of DNA sequencing data. Furthermore, identification of bacterial species by 16S rRNA gene sequencing is a quite novel technique, and the degree of 16S rRNA gene homology that is necessary to assign an isolate to a certain species has not been defined clearly yet and may vary considerably between different bacterial families and genera. Therefore, we evaluated the application of two rapid molecular methods, real-time LightCycler PCR and FISH, for identification of P. aeruginosa within this group of unusual gram-negative rods. For both methods, only some colonies of the bacterial strain are needed, and the methods can be performed very quickly within 90 min. PCR has already earlier been used successfully for identification of P. aeruginosa in CF patients (9, 18, 19). The LightCycler PCR assay showed very high sensitivity and specificity and allowed detection of all but one isolate of P. aeruginosa. The one false-negative result may be explained by increased secondary structures of the DNA preventing the probe from binding, since the primers produced the correct PCR product and the binding site of the probe was not mutated, as determined by sequencing of the amplicon. However, for generating sufficient amounts of PCR products for sequencing, Q-solution (Roche) that improves PCR results in GC-rich isolates had to be used in this sample. The FISH technique has already been used earlier to detect pathogenic bacteria directly in respiratory samples from CF patients (8). Using the newly designed P. aeruginosa-specific probe Psae16S-182, FISH allowed correct identification of all isolates of P. aeruginosa included in this study. The formerly published probe Psae (10), however, was false negative in one isolate of P. aeruginosa that was different from the isolate that was false negative in the PCR. Both probes had a specificity of 100%.
In conclusion, both rapid molecular techniques, real-time PCR and FISH, are highly useful for the rapid and specific identification of P. aeruginosa from CF patients. By implementing real-time PCR or FISH in the diagnostic work-up of unusual gram-negative rods isolated from CF patients, misidentification of P. aeruginosa can be avoided and rapid identification of the species within the same day is achievable. In comparison to real-time PCR, FISH is a very cheap method that does not require costly technical equipment and is, thus, especially suited for smaller or non-university-based laboratories. Apart from P. aeruginosa, rapid molecular diagnostic techniques may also be useful for the identification of other pathogenic bacteria in CF patients, like B. cepacia complex. A respective FISH probe has been published earlier and showed promising results (8).
REFERENCES
Amann, R. I., B. J. Binder, R. J. Olson, S. W. Chisholm, R. Devereux, and D. A. Stahl. 1990. Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl. Environ. Microbiol. 56:1919-1925.
Coenye, T., J. Goris, T. Spilker, P. Vandamme, and J. J. LiPuma. 2002. Characterization of unusual bacteria isolated from respiratory secretions of cystic fibrosis patients and description of Inquilinus limosus gen. nov., sp. nov. J. Clin. Microbiol. 40:2062-2069.
Coenye, T., L. Liu, P. Vandamme, and J. J. LiPuma. 2001. Identification of Pandoraea species by 16S ribosomal DNA-based PCR assays. J. Clin. Microbiol. 39:4452-4455.
Coenye, T., P. Vandamme, and J. J. LiPuma. 2002. Infection by Ralstonia species in cystic fibrosis patients: identification of R. pickettii and R. mannitolilytica by polymerase chain reaction. Emerg. Infect. Dis. 8:692-696.
Coenye, T., P. Vandamme, and J. J. LiPuma. 2003. Ralstonia respiraculi sp. nov., isolated from the respiratory tract of cystic fibrosis patients. Int. J. Syst. Evol. Microbiol. 53:1339-1342.
Ferroni, A., I. Sermet-Gaudelus, E. Abachin, G. Quesne, G. Lenoir, P. Berche, and J. L. Gaillard. 2002. Use of 16S rRNA gene sequencing for identification of nonfermenting gram-negative bacilli recovered from patients attending a single cystic fibrosis center. J. Clin. Microbiol. 40:3793-3797.
Hiraishi, A. 1992. Direct automated sequencing of 16S rDNA amplified by polymerase chain reaction from bacterial cultures without DNA purification. Lett. Appl. Microbiol. 15:210-213.
Hogardt, M., K. Trebesius, A. M. Geiger, M. Hornef, J. Rosenecker, and J. Heesemann. 2000. Specific and rapid detection by fluorescent in situ hybridization of bacteria in clinical samples obtained from cystic fibrosis patients. J. Clin. Microbiol. 38:818-825.
Karpati, F., and J. Jonasson. 1996. Polymerase chain reaction for the detection of Pseudomonas aeruginosa, Stenotrophomonas maltophilia and Burkholderia cepacia in sputum of patients with cystic fibrosis. Mol. Cell Probes 10:397-403.
Kempf, V. A., K. Trebesius, and I. B. Autenrieth. 2000. Fluorescent in situ hybridization allows rapid identification of microorganisms in blood cultures. J. Clin. Microbiol. 38:830-838.
Kiska, D. L., and P. H. Gilligan. 2003. Pseudomonas, p. 719-728. In P. R. Murray, E. J. Baron, M. A. Pfaller, F. C. Tenover, and R. H. Yolken (ed.), Manual of clinical microbiology, 8th ed. American Society for Microbiology, Washington, D.C.
Kiska, D. L., A. Kerr, M. C. Jones, J. A. Caracciolo, B. Eskridge, M. Jordan, S. Miller, D. Hughes, N. King, and P. H. Gilligan. 1996. Accuracy of four commercial systems for identification of Burkholderia cepacia and other gram-negative nonfermenting bacilli recovered from patients with cystic fibrosis. J. Clin. Microbiol. 34:886-891.
Klinger, J. D., and M. J. Thomassen. 1985. Occurrence and antimicrobial susceptibility of gram-negative nonfermentative bacilli in cystic fibrosis patients. Diagn. Microbiol. Infect. Dis. 3:149-158.
Lyczak, J. B., C. L. Cannon, and G. B. Pier. 2002. Lung infections associated with cystic fibrosis. Clin. Microbiol. Rev. 15:194-222.
Mahenthiralingam, E., J. Bischof, S. K. Byrne, C. Radomski, J. E. Davies, Y. Av-Gay, and P. Vandamme. 2000. DNA-based diagnostic approaches for identification of Burkholderia cepacia complex, Burkholderia vietnamiensis, Burkholderia multivorans, Burkholderia stabilis, and Burkholderia cepacia genomovars I and III. J. Clin. Microbiol. 38:3165-3173.
McMenamin, J. D., T. M. Zaccone, T. Coenye, P. Vandamme, and J. J. LiPuma. 2000. Misidentification of Burkholderia cepacia in US cystic fibrosis treatment centers: an analysis of 1,051 recent sputum isolates. Chest 117:1661-1665.
Miller, M. B., and P. H. Gilligan. 2003. Laboratory aspects of management of chronic pulmonary infections in patients with cystic fibrosis. J. Clin. Microbiol. 41:4009-4015.
Qin, X., J. Emerson, J. Stapp, L. Stapp, P. Abe, and J. L. Burns. 2003. Use of real-time PCR with multiple targets to identify Pseudomonas aeruginosa and other nonfermenting gram-negative bacilli from patients with cystic fibrosis. J. Clin. Microbiol. 41:4312-4317.
Spilker, T., T. Coenye, P. Vandamme, and J. J. LiPuma. 2004. PCR-based assay for differentiation of Pseudomonas aeruginosa from other Pseudomonas species recovered from cystic fibrosis patients. J. Clin. Microbiol. 42:2074-2079.
Tenover, F. C., R. D. Arbeit, R. V. Goering, P. A. Mickelsen, B. E. Murray, D. H. Persing, and B. Swaminathan. 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol. 33:2233-2239.
van Pelt, C., C. M. Verduin, W. H. Goessens, M. C. Vos, B. Tummler, C. Segonds, F. Reubsaet, H. Verbrugh, and A. van Belkum. 1999. Identification of Burkholderia spp. in the clinical microbiology laboratory: comparison of conventional and molecular methods. J. Clin. Microbiol. 37:2158-2164.
Wallet, F., T. Perez, S. Armand, B. Wallaert, and R. J. Courcol. 2002. Pneumonia due to Bordetella bronchiseptica in a cystic fibrosis patient: 16S rRNA sequencing for diagnosis confirmation. J. Clin. Microbiol. 40:2300-2301.
Wellinghausen, N., A. Essig, and O. Sommerburg. 2005. Inquilinus limosus in patients with Cystic Fibrosis, Germany. Emerg. Infect. Dis. 11:457-459.
Wellinghausen, N., B. Wirths, A. R. Franz, L. Karolyi, R. Marre, and U. Reischl. 2004. Algorithm for the identification of bacterial pathogens in positive blood cultures by real-time LightCycler polymerase chain reaction (PCR) with sequence-specific probes. Diagn. Microbiol. Infect. Dis. 48:229-241.(Nele Wellinghausen, Julia)
ABSTRACT
Phenotypic identification of gram-negative bacteria from Cystic Fibrosis (CF) patients carries a high risk of misidentification. Therefore, we compared the results of biochemical identification by API 20NE with 16S rRNA gene sequencing in 88 gram-negative, oxidase-positive rods, other than morphologically and biochemically typical P. aeruginosa, from respiratory secretions of CF patients. The API 20NE allowed correct identification of the bacterial species in 15 out of 88 (17%) isolates investigated. Agreement between the API and the 16S rRNA gene sequencing results was high only in isolates with an API result classified as "excellent identification." Even API results classified as "very good identification" or "good identification" showed a high rate of misidentification (67% and 84%). Fifty-two isolates of morphological and biochemical nontypical Pseudomonas aeruginosa, representing 59% of all isolates investigated, were not identifiable or misidentified in the API 20NE. Therefore, rapid molecular diagnostic techniques like real-time PCR and fluorescence in situ hybridization (FISH) were evaluated in this particular group of bacteria for identification of the clinically most relevant pathogen, P. aeruginosa. The LightCycler PCR assay with a P. aeruginosa-specific probe showed a sensitivity and specificity of 98.1% and 100%, respectively. For FISH analysis, a newly designed P. aeruginosa-specific probe had a sensitivity and specificity of 100%. In conclusion, molecular methods are superior over biochemical tests for identification of gram-negative, oxidase-positive rods in CF patients. In addition, real-time PCR and FISH allowed identification of morphologically nontypical isolates of P. aeruginosa within a few hours.
INTRODUCTION
Chronic bacterial colonization of the respiratory tract, leading to exacerbations of pulmonary infection, is the major cause of disease and death in Cystic Fibrosis (CF) patients. The most common pathogen in respiratory secretions of CF patients is Pseudomonas aeruginosa, and Staphylococcus aureus, Haemophilus influenzae, and members of the Burkholderia cepacia complex also play an important role in CF lung disease (13, 14, 17). Other gram-negative glucose nonfermenters such as Achromobacter xylosoxidans, Ralstonia pickettii, and Stenotrophomonas maltophilia are also occasionally recovered from CF respiratory samples, but their pathogenic significance remains to be fully clarified (4, 14, 17). Recent studies that applied molecular approaches for the identification of unusual pathogens in CF patients revealed the presence of various rarely or even newly described species belonging to the genera Bordetella, Comamonas, Inquilinus, Pandoraea, Ralstonia, etc. (2-5, 16, 22, 23). Determination of the clinical relevance of gram-negative bacteria other than P. aeruginosa in CF patients is, however, hampered by the difficult identification of these pathogens by conventional laboratory techniques.
Phenotypic identification of bacteria grown from CF patients carries a high risk of misidentification regarding both manual methods and commercial identification systems (12, 16, 18, 19, 21). In a study comparing various commercial systems for the identification of B. cepacia and P. aeruginosa isolated from CF patients, the API 20NE system was superior to other systems but misidentification or failure of identification of Burkholderia gladioli and Ralstonia pickettii occurred in a significant percentage of isolates (21).
In our experience, species like Agrobacterium radiobacter, Comamonas testosteroni, Pseudomonas alcaligenes, Ochrobactrum anthropi, and Photobacterium damsela are at times identified by the API 20NE system in respiratory specimens from CF patients. Detection of these species often leads to uncertainties concerning indication and choice of therapy. In addition, the prevalence of these species may be overestimated due to misidentification by biochemical identification systems.
Since correct identification is the prerequisite for assessing the clinical significance of rarely encountered species and for reliable detection of important pathogens like P. aeruginosa, we compared the biochemical identification results (API 20NE) of 88 gram-negative, oxidase-positive rods from respiratory secretions of CF patients with the results of 16S rRNA gene sequencing. We could demonstrate that identification by API 20NE yielded correct results in a minority of cases only. Therefore, rapid molecular diagnostic tests, i.e., real-time PCR and fluorescence in situ hybridization (FISH), were evaluated for identification of the most frequent and clinically relevant pathogen, P. aeruginosa.
MATERIALS AND METHODS
Clinical isolates and specimen processing. All respiratory secretions of CF patients (n = 82 patients) obtained during 2003 were included in the study. The specimens were plated on sheep blood agar, chocolate agar, MacConkey agar, and Burkholderia cepacia selective agar (BC agar, containing 100 mg/liter ticarcillin and 300,000 IU/liter polymyxin B; MAST Diagnostica) and were incubated for 48 h at 36°C. The BC agar plates were incubated another 5 days at room temperature. All gram-negative, oxidase-positive rods except those that were phenotypically identified as P. aeruginosa (green pigmentation or mucoidity, growth at 42°C, growth on MacConkey agar, and colistin susceptibility or at least good identification in the API 20NE; see also Fig. 1) were included in the study. Isolates that did not fulfill the morphological criteria for P. aeruginosa or mucoid P. aeruginosa as outlined in Fig. 1 and isolates that were identified in the API 20NE as a species other than P. aeruginosa or as P. aeruginosa with low reliability were subjected to 16S rRNA gene sequencing (see Fig. 1 and below) and partly represented "morphologically nontypical P. aeruginosa."
Phenotypic tests. Identification of gram-negative, oxidase-positive rods was performed according to the algorithm shown in Fig. 1. Oxidase activity was checked with 1% tetramethyl p-phenylenediamine dihydrochloride. API 20NE (BioMerieux) was performed according to the instructions of the manufacturer. The API stripes were incubated for 48 h at 30°C under ambient air. The results were interpreted with the APILAB PLUS software (version 3.3.3). Susceptibility testing against colistin was performed with 10-μg disks (BD) on Mueller-Hinton agar (Heipha). Isolates with zone of inhibition of <11 mm after incubation at 36°C for 16 to 24 h were regarded as resistant.
16S rRNA gene sequencing. DNA was prepared by using the QiAamp DNA Mini kit according to the instructions of the manufacturer. Eluted DNA was stored at –20°C until further use. The complete 16S rRNA gene was sequenced using the primers 16Sfor and 16Srev described previously (7). Sequence analysis was performed on a 310 Genetic Analyser (ABI Prism Biosystems) using a Dye Terminator Cycle Sequencing Ready Reaction kit (ABI Prism). The sequences were compared to sequences available in the GenBank database by using the BLAST algorithm. In addition, all sequences were compared to the sequences of the type strain of the respective species, including the following isolates (strain and GenBank accession numbers are in parentheses): Achromobacter xylosoxidans subsp. xylosoxidans (ATCC 9220, AF411021), Agrobacterium radiobacter (ATCC 19358, AJ389904), B. multivorans (LMG13010, Y18703), Ochrobactrum anthropi (LMG 5140, AJ242580), P. aeruginosa (ATCC 10145, AF094713), P. alcaligenes (ATCC 14909, AF094721), Pseudomonas brassicacearum (CFBP 11706, AF100321), Pseudomonas synxantha (IAM 12356, D84025), and Sphingomonas paucimobilis (ATCC 29837, U37337). Species identification was assumed when the base homologies between the clinical isolate and the type strain sequence exceeded 99.0%.
Burkholderia recA gene sequencing. Sequencing of the recA gene was done in all isolates that were identified as members of the Burkholderia cepacia complex by 16S rRNA gene sequencing. The complete recA gene (1,043 bp) was amplified with the primers BCR1 and BCR2 (15), and sequencing was performed with the primers BCR2 and BCR4 (15) as described above.
Pulsed-field gel electrophoresis (PFGE) of P. aeruginosa. In patients with detection of P. aeruginosa at more than one visit to the CF clinic, PFGE was performed for all isolates that were determined by 16S rRNA gene sequencing as P. aeruginosa. PFGE was done with the contour-clamped homogeneous electric field DRIII equipment (Bio-Rad) in 1% agarose at 14°C and a voltage of 200 V with a pulse-rate of 5 to 60 s, using the restriction enzyme SpeI. For interpretation of banding pattern the recommendations of Tenover et al. were followed (20).
Real-time LightCycler PCR for detection of P. aeruginosa. Approximately 5 to 10 colonies of an overnight culture of the strain were suspended in 500 μl of lysis buffer (1% Triton X-100, 0.5% Tween 20, 10 mM Tris-HCl, pH 8.0, 1 mM EDTA; all from Sigma) and boiled for 10 min at 95°C. After a centrifugation step of 5 min at 20,000 x g, the supernatant was used for PCR. The PCR was performed on a second-generation LightCycler instrument (Roche Diagnostics) as described previously, using the primers RWO1 and DG74 and the TaqMan probe Pseu aeru (24).
Fluorescence in situ hybridization (FISH). Some colonies of an overnight culture of the strain were taken with a cotton swab premoistened in 0.9% NaCl and transferred to a glass slide. Slides were fixed in 70% methanol for 10 min. For FISH analysis, two different oligonucleotide probes were used: the formerly published P. aeruginosa-specific probe Psae (5'-TCT CGG CCT TGA AAC CCC) (10), binding to the 23S rRNA and labeled with fluorescein isothiocyanate (and also with Cy3 in separate experiments) as fluorescent dye, and a newly designed P. aeruginosa-specific probe Psae16S-182 (5'-CCA CTT TCT CCC TCA GGA CG), binding to the 16S rRNA and labeled with Cy3. The new probe was designed by using the ARB program package, available at http://www.arb-home.de/. FISH was performed at 46°C using 40% formamide in the hybridization buffer and the corresponding salt concentrations in washing buffer as described elsewhere (1). The new probe was validated with the reference strain P. aeruginosa ATCC 27853 as well as with several clinical isolates of P. aeruginosa that were confirmed by 16S rRNA gene sequencing. Furthermore, other closely related species were tested, including Pseudomonas putida (clinical isolate), Pseudomonas stutzeri (ATCC 17588), Acinetobacter baumannii (ATCC 19606), Acinetobacter junii (ATCC 17908), and Sphingomonas paucimobilis (clinical isolate), and no cross-reactions were observed. In addition, a eubacterial EUB probe (EUB 388 5'-GCT GCC TCC CGT AGG AGT) that detects all bacterial species (1) was used as a control. The stained samples were rapidly air dried, embedded, and examined with a Zeiss Axioplan 2 microscope. Samples with a fluorescence intensity clearly exceeding the background were regarded as positive. A 4',6'-diamidino-2-phenylindole stain (Hoechst) was done in order to ensure the presence of bacteria on the slide and was positive in all samples.
RESULTS
Identification by API 20NE. During the study period, 405 isolates of gram-negative, oxidase-positive rods were isolated from 82 CF patients. Two-hundred ninety-nine of these isolates were identified as P. aeruginosa (mucoid or nonmucoid) based on typical growth characteristics (11) and the diagnostic algorithm outlined in Fig. 1, and were identified as P. aeruginosa by API 20NE. These included 2 isolates with "excellent" API result, 12 isolates with "very good" API result, and 4 isolates with "good" API result. Eighty-eight isolates (22%) were subjected to API 20NE and 16S rRNA gene sequencing according to the algorithm applied (Fig. 1). Of these 88 isolates, the API result showed excellent, very good, or good identification of one the following species or genera in only 41 isolates (47%): Agrobacterium radiobacter, Achromobacter xylosoxidans, Burkholderia cepacia, Comamonas testosteroni or P. alcaligenes complex, Pasteurella spp., Photobacterium damsela, Pseudomonas spp., P. fluorescens, Psychrobacter phenylpyruvicus, and Ralstonia pickettii (Table 1). In 11 isolates, the API indicated acceptable identification of a species (Table 1). In 36 isolates (41%), the API result was unacceptable (n = 4), unreliable (n = 4), or ambiguous (n = 3) or had low selectivity (n = 25).
16S rRNA gene sequencing. 16S rRNA gene sequencing allowed identification to the species level in 87 out of 88 isolates and to the genus level (Chryseobacterium spp.) in one isolate. Of the seven species that were excellently identified by API, five were confirmed by 16S rRNA gene sequencing. However, both isolates identified as P. fluorescens in the API (Table 1) were identified as P. synxantha and P. brassicacearum by sequencing, both representing species that are not included in the API database. Concerning those isolates showing very good identification by API, only 5 of the 12 isolates that were identified as Burkholderia cepacia could be confirmed by 16S rDNA sequencing. Eight isolates of Achromobacter xylosoxidans were misidentified in the API as either B. cepacia or P. fluorescens (Table 1). Concerning those 19 isolates showing good identification by API, only two isolates of Achromobacter xylosoxidans and one isolate of Ochrobactrum anthropi were correctly identified. Fifteen isolates of P. aeruginosa were misidentified as either Comamonas testosteroni or P. alcaligenes complex, Photobacterium damsela, P. fluorescens, Psychrobacter phenylpyruvicus, or Ralstonia pickettii (Table 1). Regarding those isolates with acceptable identification by API, a single isolate of Burkholderia cepacia complex was identified correctly (Table 1). Concerning those isolates with low selectivity, unreliable, ambiguous, or unacceptable identification (n = 36), 16S rRNA gene sequencing confirmed the species that was proposed by the API software as the closest match in two cases only (Table 1), i.e., one isolate each of Burkholderia cepacia complex and Sphingomonas paucimobilis. Twenty-seven of the 36 isolates (75%) were identified as P. aeruginosa by 16S rRNA gene sequencing, and the other isolates included Achromobacter xylosoxidans (n = 4), Burkholderia cepacia complex (n = 1), Chryseobacterium spp. (n = 1), and Inquilinus limosus (n = 1). Altogether, 52 of the 88 isolates investigated in this study (59%) were identified as P. aeruginosa by 16S rRNA sequencing. Sequencing of the Burkholderia recA gene revealed that all isolates of B. cepacia complex found in this study belonged to the species B. multivorans.
Of the 18 isolates identified as P. aeruginosa by API 20NE (Fig. 1), 8 isolates were available for 16S rRNA gene sequencing and the sequencing result confirmed the API result. The eight isolates included one isolate with "excellent" API result, five isolates with "very good" API result, and two isolates with "good" API result.
Identification of persistent isolates of P. aeruginosa. In six patients, P. aeruginosa was detected by 16S rRNA gene sequencing on more than one occasion during the study period. PFGE of all isolates revealed persistent infection in five of these patients, and each patient carried a unique, persistent strain (data not shown). Interestingly, results of API 20NE differed considerably within the same strain. In Table 2, an overview of the interpretable results of the API 20NE in persistent strains is given. Ambiguous, unreliable, or unacceptable results as well as identifications with low selectivity were omitted.
Specific identification of P. aeruginosa by real-time PCR and FISH. Due to the predominance of P. aeruginosa among the isolates investigated in this study and due to the importance of correct identification of this pathogen in CF respiratory specimens, we tried to specifically detect P. aeruginosa by rapid molecular techniques, i.e., real-time PCR and FISH.
For real-time PCR, a formerly developed P. aeruginosa-specific TaqMan probe assay was run on the LightCycler (24). By using this PCR assay, 51 out of 52 isolates of P. aeruginosa were identified correctly, and none of the other 36 isolates was false positive, corresponding to a sensitivity of 98.1% and a specificity of 100%. In the one false-negative isolate, the primers produced the correct PCR product and the binding site of the probe was not mutated, as determined by sequencing of the amplicon. However, for generating sufficient amounts of PCR products for sequencing, Q-solution (Roche) that improves PCR results in GC-rich isolates had to be used in this sample.
For FISH, two different P. aeruginosa-specific probes were used. The formerly published probe Psae correctly identified 51 out of 52 isolates of P. aeruginosa and none of the other isolates, thus resulting in a sensitivity of 98.1% and a specificity of 100%. The one false-negative isolate in the FISH was different from the false-negative isolate in the PCR described above. The newly designed probe Psae16S-182 even detected all 52 isolates of P. aeruginosa and none of the other isolates, corresponding to a sensitivity and specificity of 100%.
DISCUSSION
Apart from P. aeruginosa and members of the B. cepacia complex, other gram-negative, oxidase-positive rods may probably also play roles in progression of CF lung disease. Identification of these bacterial species with biochemical methods is difficult due to low metabolic activity and morphological and biochemical variability of the isolates. Thus, in this study, molecular techniques were used to verify the results of the widely distributed biochemical identification system API 20NE for identification of unusual gram-negative, oxidase-positive rods isolated from CF patients. Isolates clearly identified as P. aeruginosa by commonly accepted morphological growth characteristics (n = 299) were not included in the study. In the study population of unusual gram-negative, oxidase-positive rods not including morphologically typical P. aeruginosa, the API 20NE allowed correct identification of the bacterial species in 15 of 88 (17%) isolates investigated (Table 1). Agreement between the API and the 16S rRNA gene sequencing result was high only in isolates with an API result classified as "excellent identification." Even API results classified as "very good identification" or "good identification" showed a high rate of misidentification (67% and 84%, respectively). The main restrictions of the API 20NE were misidentification of Achromobacter xylosoxidans as Burkholderia cepacia, Ochrobactrum anthropi, or Pseudomonas fluorescens and misidentification of P. aeruginosa as various species, including predominantly Comamonas testosteroni or P. alcaligenes complex, P. fluorescens, and Photobacterium damsela (Table 1). Difficulties of the API 20NE system to identify members of the Burkholderia cepacia complex have been described previously (6, 21). Misidentification of Achromobacter xylosoxidans is considerably worrying, since the relevance of this species in progression of CF lung disease is not yet fully understood and correct identification of the species is a prerequisite for assessing its clinical significance. Relying on API 20NE results when species other than P. aeruginosa are identified may lead to overestimation of the prevalence of unusual species and to uncertainties regarding therapy of such organisms. This applies also to Comamonas testosteroni and P. alcaligenes. Interestingly, all but one of the isolates identified by API 20NE as Comamonas testosteroni or P. alcaligenes complex, which represented 11% of all isolates investigated in the study, belonged to the species P. aeruginosa. Altogether, P. aeruginosa was identified by 16S rRNA sequencing in approximately 60% of isolates in this study. In patients persistently infected with P. aeruginosa, the API 20NE results varied markedly even within the same strain.
Isolates clearly identified as P. aeruginosa by the API 20NE were only partly (8 out of 18 isolates) available for 16S rRNA gene sequencing. Although sequencing confirmed the API results in these isolates, the specificity of the API 20NE regarding identification of P. aeruginosa cannot be determined from this study due to the small sample size and incomplete data. In conclusion, results of the API 20NE must be interpreted with great care in this particular population of nonfermentative, oxidase-positive rods obtained from CF patients.
Misidentification of P. aeruginosa has remarkable consequences for the patients and should be avoided by all means. Despite its exactness, DNA sequencing is not suitable for routine identification of P. aeruginosa in the diagnostic microbiology laboratory due to its high costs and work load. In addition, public DNA databases that do not involve quality control measures for the deposited sequences also contain faulty or incorrect assigned sequences, thus requiring an experienced user for interpretation of DNA sequencing data. Furthermore, identification of bacterial species by 16S rRNA gene sequencing is a quite novel technique, and the degree of 16S rRNA gene homology that is necessary to assign an isolate to a certain species has not been defined clearly yet and may vary considerably between different bacterial families and genera. Therefore, we evaluated the application of two rapid molecular methods, real-time LightCycler PCR and FISH, for identification of P. aeruginosa within this group of unusual gram-negative rods. For both methods, only some colonies of the bacterial strain are needed, and the methods can be performed very quickly within 90 min. PCR has already earlier been used successfully for identification of P. aeruginosa in CF patients (9, 18, 19). The LightCycler PCR assay showed very high sensitivity and specificity and allowed detection of all but one isolate of P. aeruginosa. The one false-negative result may be explained by increased secondary structures of the DNA preventing the probe from binding, since the primers produced the correct PCR product and the binding site of the probe was not mutated, as determined by sequencing of the amplicon. However, for generating sufficient amounts of PCR products for sequencing, Q-solution (Roche) that improves PCR results in GC-rich isolates had to be used in this sample. The FISH technique has already been used earlier to detect pathogenic bacteria directly in respiratory samples from CF patients (8). Using the newly designed P. aeruginosa-specific probe Psae16S-182, FISH allowed correct identification of all isolates of P. aeruginosa included in this study. The formerly published probe Psae (10), however, was false negative in one isolate of P. aeruginosa that was different from the isolate that was false negative in the PCR. Both probes had a specificity of 100%.
In conclusion, both rapid molecular techniques, real-time PCR and FISH, are highly useful for the rapid and specific identification of P. aeruginosa from CF patients. By implementing real-time PCR or FISH in the diagnostic work-up of unusual gram-negative rods isolated from CF patients, misidentification of P. aeruginosa can be avoided and rapid identification of the species within the same day is achievable. In comparison to real-time PCR, FISH is a very cheap method that does not require costly technical equipment and is, thus, especially suited for smaller or non-university-based laboratories. Apart from P. aeruginosa, rapid molecular diagnostic techniques may also be useful for the identification of other pathogenic bacteria in CF patients, like B. cepacia complex. A respective FISH probe has been published earlier and showed promising results (8).
REFERENCES
Amann, R. I., B. J. Binder, R. J. Olson, S. W. Chisholm, R. Devereux, and D. A. Stahl. 1990. Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl. Environ. Microbiol. 56:1919-1925.
Coenye, T., J. Goris, T. Spilker, P. Vandamme, and J. J. LiPuma. 2002. Characterization of unusual bacteria isolated from respiratory secretions of cystic fibrosis patients and description of Inquilinus limosus gen. nov., sp. nov. J. Clin. Microbiol. 40:2062-2069.
Coenye, T., L. Liu, P. Vandamme, and J. J. LiPuma. 2001. Identification of Pandoraea species by 16S ribosomal DNA-based PCR assays. J. Clin. Microbiol. 39:4452-4455.
Coenye, T., P. Vandamme, and J. J. LiPuma. 2002. Infection by Ralstonia species in cystic fibrosis patients: identification of R. pickettii and R. mannitolilytica by polymerase chain reaction. Emerg. Infect. Dis. 8:692-696.
Coenye, T., P. Vandamme, and J. J. LiPuma. 2003. Ralstonia respiraculi sp. nov., isolated from the respiratory tract of cystic fibrosis patients. Int. J. Syst. Evol. Microbiol. 53:1339-1342.
Ferroni, A., I. Sermet-Gaudelus, E. Abachin, G. Quesne, G. Lenoir, P. Berche, and J. L. Gaillard. 2002. Use of 16S rRNA gene sequencing for identification of nonfermenting gram-negative bacilli recovered from patients attending a single cystic fibrosis center. J. Clin. Microbiol. 40:3793-3797.
Hiraishi, A. 1992. Direct automated sequencing of 16S rDNA amplified by polymerase chain reaction from bacterial cultures without DNA purification. Lett. Appl. Microbiol. 15:210-213.
Hogardt, M., K. Trebesius, A. M. Geiger, M. Hornef, J. Rosenecker, and J. Heesemann. 2000. Specific and rapid detection by fluorescent in situ hybridization of bacteria in clinical samples obtained from cystic fibrosis patients. J. Clin. Microbiol. 38:818-825.
Karpati, F., and J. Jonasson. 1996. Polymerase chain reaction for the detection of Pseudomonas aeruginosa, Stenotrophomonas maltophilia and Burkholderia cepacia in sputum of patients with cystic fibrosis. Mol. Cell Probes 10:397-403.
Kempf, V. A., K. Trebesius, and I. B. Autenrieth. 2000. Fluorescent in situ hybridization allows rapid identification of microorganisms in blood cultures. J. Clin. Microbiol. 38:830-838.
Kiska, D. L., and P. H. Gilligan. 2003. Pseudomonas, p. 719-728. In P. R. Murray, E. J. Baron, M. A. Pfaller, F. C. Tenover, and R. H. Yolken (ed.), Manual of clinical microbiology, 8th ed. American Society for Microbiology, Washington, D.C.
Kiska, D. L., A. Kerr, M. C. Jones, J. A. Caracciolo, B. Eskridge, M. Jordan, S. Miller, D. Hughes, N. King, and P. H. Gilligan. 1996. Accuracy of four commercial systems for identification of Burkholderia cepacia and other gram-negative nonfermenting bacilli recovered from patients with cystic fibrosis. J. Clin. Microbiol. 34:886-891.
Klinger, J. D., and M. J. Thomassen. 1985. Occurrence and antimicrobial susceptibility of gram-negative nonfermentative bacilli in cystic fibrosis patients. Diagn. Microbiol. Infect. Dis. 3:149-158.
Lyczak, J. B., C. L. Cannon, and G. B. Pier. 2002. Lung infections associated with cystic fibrosis. Clin. Microbiol. Rev. 15:194-222.
Mahenthiralingam, E., J. Bischof, S. K. Byrne, C. Radomski, J. E. Davies, Y. Av-Gay, and P. Vandamme. 2000. DNA-based diagnostic approaches for identification of Burkholderia cepacia complex, Burkholderia vietnamiensis, Burkholderia multivorans, Burkholderia stabilis, and Burkholderia cepacia genomovars I and III. J. Clin. Microbiol. 38:3165-3173.
McMenamin, J. D., T. M. Zaccone, T. Coenye, P. Vandamme, and J. J. LiPuma. 2000. Misidentification of Burkholderia cepacia in US cystic fibrosis treatment centers: an analysis of 1,051 recent sputum isolates. Chest 117:1661-1665.
Miller, M. B., and P. H. Gilligan. 2003. Laboratory aspects of management of chronic pulmonary infections in patients with cystic fibrosis. J. Clin. Microbiol. 41:4009-4015.
Qin, X., J. Emerson, J. Stapp, L. Stapp, P. Abe, and J. L. Burns. 2003. Use of real-time PCR with multiple targets to identify Pseudomonas aeruginosa and other nonfermenting gram-negative bacilli from patients with cystic fibrosis. J. Clin. Microbiol. 41:4312-4317.
Spilker, T., T. Coenye, P. Vandamme, and J. J. LiPuma. 2004. PCR-based assay for differentiation of Pseudomonas aeruginosa from other Pseudomonas species recovered from cystic fibrosis patients. J. Clin. Microbiol. 42:2074-2079.
Tenover, F. C., R. D. Arbeit, R. V. Goering, P. A. Mickelsen, B. E. Murray, D. H. Persing, and B. Swaminathan. 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol. 33:2233-2239.
van Pelt, C., C. M. Verduin, W. H. Goessens, M. C. Vos, B. Tummler, C. Segonds, F. Reubsaet, H. Verbrugh, and A. van Belkum. 1999. Identification of Burkholderia spp. in the clinical microbiology laboratory: comparison of conventional and molecular methods. J. Clin. Microbiol. 37:2158-2164.
Wallet, F., T. Perez, S. Armand, B. Wallaert, and R. J. Courcol. 2002. Pneumonia due to Bordetella bronchiseptica in a cystic fibrosis patient: 16S rRNA sequencing for diagnosis confirmation. J. Clin. Microbiol. 40:2300-2301.
Wellinghausen, N., A. Essig, and O. Sommerburg. 2005. Inquilinus limosus in patients with Cystic Fibrosis, Germany. Emerg. Infect. Dis. 11:457-459.
Wellinghausen, N., B. Wirths, A. R. Franz, L. Karolyi, R. Marre, and U. Reischl. 2004. Algorithm for the identification of bacterial pathogens in positive blood cultures by real-time LightCycler polymerase chain reaction (PCR) with sequence-specific probes. Diagn. Microbiol. Infect. Dis. 48:229-241.(Nele Wellinghausen, Julia)