Detection of Legionella pneumophila by Real-Time PCR for the mip Gene

Abstract

A real-time PCR assay for the mip gene of Legionella pneumophila was tested with 27 isolates of L. pneumophila, 20 isolates of 14 other Legionella species, and 103 non-Legionella bacteria. Eight culture-positive and 40 culture-negative clinical specimens were tested. This assay was 100% sensitive and 100% specific for L. pneumophila.

Legionella pneumophila is the most common pathogenic species of the 42 recognized Legionella species (3, 4, 29). Significant mortality rates among the elderly and patients with severe underlying disease may occur as a result of infection with this pathogen (5). Diagnostic delay may also result in increased mortality (15). Therefore, rapid tests, such as direct fluorescent-antibody stains and urinary antigen assays, have been developed (10, 18). Although useful, these assays have sensitivities less than 100% (9, 10, 11, 14). Nucleic acid amplification assays have been shown to be useful for the detection of Legionella (1, 8, 14, 17, 19, 20, 24, 27, 28). The genes that encode the 5S and 16S ribosomal subunits have been shown to contain signature sequences that are useful for the identification of L. pneumophila (8, 14, 15, 17, 24, 28) and a variety of other organisms. More recently, target sequences on these genes have been used in conjunction with real-time PCR for the detection of the Legionella genus, as well as the species L. pneumophila (14, 27).

The macrophage infectivity potentiator gene, which encodes a 24-kDa protein virulence factor that facilitates the entry of legionellae into amoebae and macrophages, has sufficient sequence variability between the Legionella species to also afford the specific detection of L. pneumophila by PCR (2, 6, 7, 12, 13, 14, 16, 21, 22, 23, 25). Although two groups have described real-time PCR assays for the detection of L. pneumophila via detection of the mip gene in water samples, to date only one group has evaluated a real-time PCR for this genetic target for the detection of L. pneumophila in clinical specimens (1, 14, 28). Therefore, we have attempted to confirm the utility of this gene as a target for the detection of L. pneumophila by real-time PCR, using a set of primers and hybridization probes that were different from those previously described (14).

We describe a sensitive and specific hybridization probe-based real-time PCR assay for the detection of L. pneumophila through the detection of the mip gene (GenBank accession number AF095230). This assay was used in conjunction with the LightCycler System (Roche Molecular Biochemicals, Indianapolis, Ind.) with fluorescent resonance energy transfer technology. Primers and fluorescently labeled hybridization probes were designed by Brian Caplin, formerly of Idaho Technologies, Salt Lake City, Utah. The sequences of the primers were as follows: forward primer (LpmipFp), 5′-GCAATGTCAACAGCAA 3′; reverse primer (LpmipRp), 5′-CATAGCGTCTTGCATG 3′. The 3′ end of the first hybridization probe (LpmipHP-1) was labeled with fluorescein (fam); the sequence of this probe was 5′-CAACTTATCCTTGTCTGTAGCT-[fam]-3′. The 5′ end of the second hybridization probe (LpmipHP-2) was labeled with 640-N-hydroxysuccinimide ester (Red 640), and the 3′ end of the probe was phosphorylated (p) to prevent probe extension during PCR; the sequence of this probe was 5′-[Red 640]-TGATGTGGCATCGGTTG-p-3′. The amplicon was 159 nucleotides in length. A BLAST search of the GenBank database demonstrates a high predicted specificity, with the only cross-reacting bacteria being Legionella worsleiensis (GenBank accession number LWU60164) and Legionella fairfieldensis (GenBank accession number LFU60163).

The LightCycler FastStart DNA Master Hybridization Probe Kit (Roche) was used with a final volume of 20 μl, which consisted of 18 μl of master mix and 2 μl of nucleic acid extract. The master mix was prepared with the following final concentrations per capillary tube: 0.5 μM (both) primers, 0.2 μM LpmipHP-1, 0.4 μM LpmipHP-2, and 3.0 mM MgCl_2. The experimental LightCycler protocol consisted of 10 min at 95°C for Taq polymerase activation, 45 cycles of PCR amplification (95°C for 0 s, 54°C for 10 s, and 72°C for 6 s), melting (40 to 95°C at 0.1°C/s), and a cooling step (40°C for 2 min). To determine the analytical sensitivity of this assay, bacterial DNA for testing was prepared by using the High Pure PCR Template Preparation Kit (Roche Diagnostics) according to the manufacturer's instructions. The DNA concentration was determined for pooled DNA extracts from the same isolate of L. pneumophila by measuring the optical density at 260 nm with the GeneQuant spectrophotometer (Pharmacia, Piscataway, N.J.). The DNA concentration was 2.1 (± 0.2) mg/ml. Serial dilutions of the extract were tested, and the described assay was able to detect 10 fg of DNA, which is equivalent to approximately 10 organisms.

Bacterial isolate DNA extracts used for assay validation were obtained from the Institute of Medical Microbiology and Hygiene, University of Regensburg, Regensburg, Germany, and the Cleveland Clinic Foundation, Cleveland, Ohio. The Institute of Medical Microbiology and Hygiene provided DNA extracts from 30 of the Legionella isolates, which included one representative isolate of each of the L. pneumophila serogroups 1 to 14 and 16 isolates of 14 Legionella species other than L. pneumophila (Table 1). These DNA extracts were prepared from cultured organisms by using the MagNA Pure LC DNA Isolation Kit I (Roche) according to the manufacturer's instructions. The Cleveland Clinic Foundation provided DNA extracts from 17 Legionella isolates (13 L. pneumophila isolates and 4 isolates of Legionella species other than L. pneumophila) and 103 non-Legionella bacteria representing 76 species (Tables 1 and 2). The DNA from the Legionella isolates was obtained by using the QIAamp DNA Mini Kit (Qiagen, Valencia, Calif.) according to the manufacturer's instructions. The nucleic acid from the non-Legionella bacteria was obtained from cultured organisms by boiling a loopful of organisms in a simple lysis buffer, which has been previously described (26). This assay was positive for all of the L. pneumophila isolates and negative for all of the Legionella species other than L. pneumophila and for the non-Legionella bacteria tested (Fig. 1). The sensitivity and specificity of this assay for confirmation of cultured isolates tested were 100 and 100%, respectively. We consider that the true specificity of the assay may be slightly less than 100%, since L. worsleiensis and L. fairfieldensis would be expected to produce a positive result with this assay based on the results of the BLAST search. These organisms, however, were not available for testing.

TABLE 1.

Isolates of Legionella species tested

Source	Organism(s)	No. of isolates tested	LightCycler results
University of Regensburg	L. pneumophila serogroups 1-14 (one isolate of each serogroup)	14	All positive
	L. bozemaniae and L. longbeachae	2	Both negative
	L. micdadei, L. gormanii, L. dumoffii, L. jordanis, L. hackeliae, L. oakridgensis (ATCC 33761), L. israelensis (ATCC 43119), L. wadsworthii (ATCC 33877), L. tucsonensis (ATTCC 49180), L. feeleii (ATCC 35849), L.sainthelensi (ATCC 35248), and L. birminghamensis (ATCC 43702)	1 each	All negative
Cleveland Clinic Foundation	L. pneumophila (five of serogroup 1, one of serogroup 6, and seven that were not serotyped)	13	All positive
	L. bozemanae	1	Negative
	L. jordanis	1	Negative
	L. micdadei	2	Both negative

TABLE 2.

Isolates of non-Legionella bacteria tested

Type	Organism(s)	No. of isolates tested	LightCycler results
Gram negative	Salmonella spp., Shigella spp. Yersinia sp., Citrobacter sp., Escherichia coli, Proteus spp., Klebsiella spp., Serratia sp., Pseudomonas sp., Bacteroides sp., Moraxella sp., Haemophilus sp., Acinetobacter sp., and Enterobacter spp.	3 each	All negative
	Providencia sp., Bartonella spp., and Burkholderia cepacia	2 each	All negative
	Neisseria spp.	9	All negative
	Vibrio cholerae, Eikenella corrodens, Afipia felis, Campylobacter jejuni, Rhizobium sp., and Mesorhizobium sp.	1 each	All negative
Gram positive	Staphylococcus aureus (including ATCC 25923)	5	All negative
	Staphylococcus epidermidis	3	All negative
	Staphylococcus saprophyticus, Micrococcus sp., Stomatococcus sp., and Lactobacillus sp.	2 each	All negative
	Enterococcus sp., viridans group streptococcus, Streptococcus pneumoniae, group A Streptococcus, group B Streptococcus, Aerococcus sp., Bacillus spp., and Corynebacterium spp.	3 each	All negative

FIG. 1.

Real-time PCR results for Legionella isolates.

Clinical specimens were obtained from the clinical microbiology laboratory at the Cleveland Clinic Foundation. Eight clinical specimens were identified that were culture positive for Legionella (7 L. pneumophila isolates and 1 Legionella bozemanae isolate). These were stored at −20°C. Twenty bronchoalveolar lavage (BAL) and 20 sputum specimens that were culture negative for L. pneumophila were also tested. All the clinical specimens were tested with the MONOFLUO Legionella pneumophila IFA (immunofluorescent-antibody) Test Kit (direct fluorescent-antibody assay [DFA]; Bio-Rad Laboratories, Redmond, Wash.), which was performed according to the manufacturer's guidelines, and cultured for Legionella by using buffered charcoal-yeast extract agar (BCYE agar; BD Diagnostic Systems, Sparks, Md.) with 10 days of incubation. DNA extracts from the clinical specimens were prepared by using the Qiagen DNA Mini Kit (Qiagen) according to the manufacturer's guidelines. Extracts were suspended in a final volume of 100 μl. The seven specimens that were culture positive for L. pneumophila were PCR positive with the described assay, while the specimen that contained L. bozemanae and the 40 culture-negative BAL and sputum specimens were negative (Fig. 2 and Table 3). Only five of the seven specimens that were culture and PCR positive for L. pneumophila were positive by DFA for L. pneumophila. In this limited study of clinical specimens, the results of real-time PCR were equivalent to those of culture (seven of seven positive) for the detection of L. pneumophila and better than those of the DFA (five of seven positive).

FIG. 2.

Real-time PCR results for L. pneumophila in clinical specimens.

TABLE 3.

Clinical specimens tested for L. pneumophila by real-time PCR detection of the mip gene

Patient no.	Specimen sourcea	Culture resultsb	DFA resultsc	LightCycler results
1	BAL fluid	L. pneumophila serogroup 1	+ (TP)	+ (TP)
2	BAL fluid	L. pneumophila serogroup 1	+ (TP)	+ (TP)
3	ET tube	L. pneumophila serogroup 1	+ (TP)	+ (TP)
4	LUL tissue	L. pneumophila serogroup 1	− (FN)	+ (TP)
5	BAL fluid	L. pneumophila (not serotyped)	+ (TP)	+ (TP)
6	BAL fluid	L. pneumophila serogroup 6	+ (TP)	+ (TP)
7	Sputum	L. pneumophila serogroup 1d	− (FN)	+ (TP)
8	BAL fluid	L. bozemanae serogroup 1	− (TN)	− (TN)
9-28	Sputum	Negative	− (TN)	− (TN)
29-48	BAL fluid	Negative	− (TN)	− (TN)

^aET tube, endotracheal; LUL, left upper lobe of the lung.

^bResults as reported by the Ohio State Department of Health and/or the Centers for Disease Control.

^cTP, true positive; TN, true negative; FN, false negative.

^dSubsequent specimens from the same patient not included in the analysis were negative by DFA and culture but remained positive by PCR.

PCR has been shown to be a useful diagnostic tool for the clinical microbiologist for infections caused by a variety of human pathogens, including L. pneumophila. Real-time PCR is often favored over traditional PCR, since the results are available in a closed system, which diminishes the likelihood of laboratory contamination, and is more timely since detection and amplification occur simultaneously. Two real-time PCR assays have been described for the detection of the genus Legionella and the species L. pneumophila. One of these assays detects the Legionella genus and differentiates L. pneumophila by using two sets of hybridization probes that target signature sequences in the 16S ribosomal subunit gene (27). The other assay detects the Legionella genus by using a PCR with the 5S ribosomal rNA gene and detects L. pneumophila by targeting the mip gene (14). We have confirmed the utility of the mip gene as a target for the detection of L. pneumophila by real-time PCR by using an assay different than that previously reported. This assay could potentially also be used as a follow-up to a pan-Legionella assay to help determine whether the positive result was due to L. pneumophila, as previously described (14). Alternatively, we plan to use this assay as a replacement for DFA and as a confirmatory test for cultured isolates. We have opted to retain the BCYE culture medium, since occasionally we have recovered unexpected pathogens, such as Nocardia species and Blastomyces dermatitidis, from this medium. The retention of the BCYE medium will also serve to recover the non-L. pneumophila isolates, which overall are much less frequently encountered than the L. pneumophila isolates (3). In addition, retention of this medium affords the recovery of the L. pneumophila isolates for submission to the public health laboratories and for epidemiologic purposes.