|Home | About | Journals | Submit | Contact Us | Français|
A rapid two-step scheme based on PCR amplification and enzymatic digestion analysis of a 226-bp fragment of the 16S rRNA gene was developed to identify the Legionella genus by PCR amplification and to differentiate the Legionella pneumophila and non-Legionella pneumophila species by enzymatic digestion analysis. Among 42 ATCC strains (16 strains of L. pneumophila and 26 strains of non-L. pneumophila) and 200 Legionella isolates from environmental water samples, including pools, rivers, lakes, and cooling towers in Guangdong province, 99.59% of L. pneumophila and non-L. pneumophila strains were correctly identified and differentiated by this scheme. The procedure of this two-step identification and differentiation scheme is simple and takes only about 4 h. These results suggest that this two-step scheme provides a simple and convenient method for the rapid identification and differentiation of L. pneumophila and non-L. pneumophila species.
Legionella species, which are fastidious and ubiquitous worldwide in natural water environment such as rivers, lakes, and artificial water systems, are the causative agent of Legionnaire's disease (7, 23). L. pneumophila is the most common pathogenic species within the genus Legionella and is the main cause of Legionnaire's disease, which appears as a mild respiratory illness, an acute life-threatening pneumonia, or Pontiac fever (25). Many Legionella species have been recognized as human illness agents (2). In addition to L. pneumophila, 20 Legionella species have been documented as human pathogens on the basis of their isolation from clinical material (4), but they occur at very low frequencies (12). Non-L. pneumophila species also have been reported to be infectious (9). The majority of the confirmed infections involving non-L. pneumophila were from immunosuppressed patients (4).
Clinical manifestations caused by Legionella infection usually are indistinguishable from the pneumonia of other bacterial etiologies. Since the symptoms are atypical, it is difficult to clinically identify the actual causative agent (4). Therefore, the identification of Legionella species and the differentiation of L. pneumophila and non-L. pneumophila species have been of increasing importance (2).
Current methods for the detection of Legionella species are based on culture techniques, which take at least 3 to 10 days. The long turnaround time (TAT) limits these methods for their clinical utility. Additional problems with culture detection include low sensitivity, the requirement of special media, adequate specimen processing, the need for technical experts, and microbial contamination inhibiting Legionella growth. Contamination by some other microorganism that is viable but nonculturable can appear after the 3 to 10 days of culture for Legionella because of the fastidious nature of the these microorganism and the requirement of prolonged incubation periods for their growth. These problems make Legionella isolation and identification challenging (3, 6, 15).
Methodologies exploited for the identification of Legionella isolates include direct fluorescent antibody (DFA), urine antigen detection (5, 11), and sequence-based genotypic classification schemes such as PCR and real-time PCR (2, 10, 19, 21). Urine antigen detection is the most widely used and is considered to be specific for L. pneumophila serogroup 1, but it misses about 40% of legionellosis cases (5, 11). DFA has a low sensitivity for the diagnosis of respiratory samples. These two methods cannot detect non-L. pneumophila spp. Although L. pneumophila is the most frequent cause of legionellosis, non-L. pneumophila species also may cause serious or fatal disease (16, 24). Non-L. pneumophila species in respiratory specimens are not detectable by the L. pneumophila direct fluorescent antigen test. Similarly, the Legionella urinary antigen test will be negative for urine specimens from patients with infections caused by these bacteria, and fluorescent antibody stains are not commercially available for the identification of non-L. pneumophila isolates (22). Therefore, the rapid identification and differentiation of L. pneumophila and non-L. pneumophila is of critical importance for the diagnosis.
Several sequence-based genotypic classification schemes have been reported to identify and differentiate Legionella species in laboratories. A multiplex PCR assay was used to detect and differentiate L. pneumophila and non-L. pneumophila species by targeting a 386-bp fragment of the 16S rRNA gene and the macrophage infectivity potentiator (mip) gene fragment. This method had low specificity because the mip special primers for L. pneumophila species were not specific (19).
Real-time PCR has some benefits compared to routine diagnosis, as it can minimize the manual time for the PCR and make the use of post-PCR analysis convenient. These diagnostic PCR assays targeted specific regions, including the 16S rRNA gene (10, 18, 23), the 23S-5S spacer region (8), the 5S rRNA gene (13, 14), and the mip gene (3, 17). They can detect only the genus Legionella and cannot differentiate L. pneumophila from non-L. pneumophila species.
The use of the sequence of the mip gene was reported to be able to accurately discriminate among 39 Legionella species, because the mip gene was specific to most Legionella members and the subsequent interspecies sequence variation was sufficient to discriminate clearly between species, and it enabled the species-specific identification of most Legionella species implicated in human disease (17). A sequence-based classification scheme based on the smaller 5S rRNA gene (104 bp) and partial 16S rRNA gene sequencing was less discriminatory than that's with the mip gene (14, 17). However, both methods were complex and time-consuming.
The use of partial 16S rRNA gene sequencing for the identification of L. pneumophila and non-L. pneumophila species was another reported method to differentiate the Legionella species. It was based on gene sequencing and alignment. This method also was a time-consuming method, and its specificity has been questioned (22).
For all of the reasons mentioned above, a rapid, sensitive, and accurate method for the identification of Legionella species, and the differentiation of L. pneumophila and non-L. pneumophila species, becomes extremely important. Here, we pursue a rapid and accurate two-step identification and differentiation scheme to identify and differentiate L. pneumophila and non-L. pneumophila species. The scheme uses an initial PCR amplification technique for the 16S rRNA gene to identify the Legionella species, followed by an enzymatic digestion analysis of the 226-bp fragment of the 16S rRNA gene to differentiate between L. pneumophila and non-L. pneumophila.
To design a valid PCR, we analyzed the fragments of 16S rRNA genes from 82 Legionella strains and performed a theoretical analysis. Based on the results from this analysis, we developed a PCR amplification method to detect all species of the genus Legionella.
The differentiation of L. pneumophila and non-L. pneumophila was achieved by the enzymatic digestion analysis of the 226-bp fragment of the 16S rRNA gene from the PCR assay. This method was based on a finding from our bioinformatics analysis of a 226-bp fragment in the Legionella 16S rRNA gene. The bioinformatics data showed that the ACNGT (N = A, G, C, or T) base sequence, identifiable and digestible by HpyCH4III endonuclease, had a remarkably consistent pattern: it appeared at bp 178 to 182 of L. pneumophila while non-L. pneumophila strains had a variable sequence in this site. If the 226-bp fragment is digested with HpyCH4III endonuclease, the 226-bp PCR products from L. pneumophila species will have two fragments with sizes of 180 and 46 bp, while non-L. pneumophila isolates will not have them. By running digested samples on agarose electrophoresis, L. pneumophila and non-L. pneumophila species can be correctly differentiated.
In summary, this two-step assay system is used to detect the genus Legionella by PCR amplification and then to differentiate L. pneumophila and non-L. pneumophila by enzymatic digestion analysis. The performance of this scheme was evaluated with a panel of well-characterized Legionella strains obtained from the ATCC and with isolated strains in our laboratory from environmental water samples, such as lakes, pools, rivers, and cooling towers in Guangdong province. The test results by this scheme for the strains mentioned above had remarkable sensitivity and specificity.
Forty-two Legionella strains and 12 non-Legionella strains were supplied by the American Type Tissue Culture Collection (ATCC) (Table (Table1).1). Two hundred Legionella strains were from our laboratory. They were isolated from environmental water samples in Guangdong province, and their identities were confirmed by fatty acid analysis, sequence typing analysis, and biochemical analysis. All Legionella strains tested were grown on buffered charcoal yeast extract (BCYE) agar and incubated at 37°C in 5% CO2. The environmental water isolates also were tested by amplified fragment length polymorphism (AFLP) typing or serogroup typing to understand their differences. The serogroup types of the 146 environmental L. pneumophila isolates were serogroup 1, serogroup 2, and some undefined serogroups. AFLP types of non-L. pneumophila isolates were EWGLI AFLP 013 London, EWGLI AFLP 012 Rome, EWGLI AFLP 014 London, EWGLI AFLP 001 Lugano, EWGLI AFLP 009 London, EWGLI AFLP 019 Dresden, EWGLI AFLP 015 Dresden EUL131, etc. The 54 non-L. pneumophila isolates were L. longbeachae, L. sainthelensi, L. feeleii, L. oakridgensis, L. gormanii, etc.
Bacterial strains were purified, and nucleic acid extraction was performed with the TIANamp bacterial DNA kit (Tiangen Biotech Co., Ltd., Beijing, China) according to the manufacturer's instructions. Each extraction included a negative control sterilized to eliminate the potential contamination of Legionella DNA and other environmental contaminants. The extracted sample, ~200 μl in volume containing the genome DNA, was frozen in microcentrifuge tubes at −20°C before analysis.
After searching all of the sequence information for the 386 fragments of 16S rRNA genes of genus Legionella from the NCBI (http://www.ncbi.nlm.nih.gov/) nucleotide database, we found that the sequences of the 386 fragments for genus Legionella were different from other non-Legionella species. At the same time, the sequences within the genus Legionella had few variations. We used the fragments from 82 strains, including 8 strains of L. pneumophila and 74 strains of non-L. pneumophila, for analysis by ClustalX (version 2.0) (1, 20), a bioinformatics software package used to align gene sequences or protein sequences.
PCR primer sequences were designed for targeting the 226-bp fragment of the 16S rRNA gene for the genus Legionella. These 226 bp form the upstream fragment of the total 386 bp. The design mutated site 195 of the 226-bp fragment from T to G. The sequence of the forward primer was 5′-AGGGTTGATAGGTTAAGAGC-3′. The reverse primer was 5′-ATTCCACTACCCTCTCCCATACTCGAGTCAACC-3′.
Each PCR included 2× Taq PCR MasterMix including polymerase, reaction buffer, deoxynucleoside in a ready-to-use formulation, 3 mM MgCl2, 20 mM Tris-HCl (pH 8.3), 100 mM KCl (Tiangen Biotech Co., Ltd). Primers, double-distilled water (ddH2O), and 4 μl of template were added for a total volume of 50 μl. A GeneAmp PCR system 9700 with a 96-well sample block module (Applied Biosystems) was used for the reaction. The thermal cycling profile consisted of an initial incubation at 94°C for 3 min, followed by 35 cycles of 94°C for 30 s, 57°C for 30 s, and 72°C for 30 s, and finally 72°C for 5 min. PCR products were purified by a TIANquick Midi Purification kit (Tiangen Biotech Co., Ltd., Beijing, China) according to the manufacturer's instructions.
The endonuclease of HpyCH4III utilized in the enzymatic digestion analysis of purified PCR products was from New England Biolabs (Neb-China, Beijing, China). Each enzymatic digestion system included 10× NEB buffer 4, ddH2O, HpyCH4III endonuclease, and 8 μl of purified PCR product for a total volume of 12 μl. Each digestion reaction was carried out at 37°C for 45 min. After the enzymatic digestion, 6 μl of each reaction product was used for agarose electrophoresis analysis.
The primer's specificity was analyzed by using the following species as controls: Salmonella typhi (ATCC 14028), Enterobacter cloacae (ATCC 13047), Klebsiella pneumoniae (ATCC 35657), Haemophilus influenzae (ATCC 10211), Proteus mirabilis (ATCC 35659), Enterococcus faecalis (ATCC 29212), Neisseria meningitides (WSB 0702), Pseudomonas aeruginosa (ATCC 27853), Yersinia enterocolitica (WSB 6612), Stenotrophomonas maltophilia (ATCC 51331), Escherichia coli (ATCC 25922), and Staphylococcus aureus (ATCC 6358). These bacterial strains included five bacillus bacterial strains. They were Salmonella typhi, Enterobacter cloacae, Klebsiella pneumoniae, Haemophilus influenzae, and Proteus mirabilis. The five bacillus bacterial strains utilized in the specificity test of primers were more significant than other bacterial strains, because bacillus strains were closer to genus Legionella in evolution than the other bacterial strains. The nucleic acid extraction of these strains was performed under the same conditions as those mentioned above for Legionella.
The bioinformatics analysis found that the site at bp 178 to 182 of the 386-bp fragment from the 16S rRNA gene for L. pneumophila showed a remarkably consistent pattern: all strains had the base sequence of ACNGT (N = A, G, C, or T), while non-L. pneumophila strains had a variable base sequence in this region (Fig. (Fig.1).1). Our bioinformatics analysis also found that some of the non-L. pneumophila strains possessed the same base sequence of ACNGT, but it occurred in at bp 151 to 155 or bp 191 to 195, not at bp 178 to 182. The non-L. pneumophila strains with ACNGT in the bp 151 to 155 site included L. birminghamiensis, L. brunensis, L. donaldsonii, L. lansingensis, L. quinlivanii, and L. spiritensis. Only a few non-L. pneumophila strains had the same base sequence in the bp 191 to 195 site (Table (Table22).
The finding described above was extremely important, because the base sequence of ACNGT can be digested by HpyCH4III endonuclease. With this in mind, we came up with the novel idea to use HpyCH4III endonuclease to digest the 226-bp fragment, the upstream segment of the 386-bp fragment from the 16S rRNA gene, and then to differentiate L. pneumophila and non-L. pneumophila by the digested fragment sizes using agarose electrophoresis. The only issue with this design initially was that some of the non-L. pneumophila strains with the base sequence of ACNGT in the bp 191 to 195 site would have digested fragments that were too similar to those of the L. pneumophila strains. It was difficult to separate them by agarose electrophoresis. To resolve this issue, a novel primer was designed to mutate the 195 site from T to G, which changed the bp 191 to 195 base sequence from ACTGT to ACTGG.
Based on this bioinformatics analysis for all of the Legionella species in the NCBI database, we concluded that L. pneumophila and non-L. pneumophila isolates could be detected and differentiated rapidly by the novel scheme of the PCR assay, followed by enzymatic digestion analysis. The PCR assay was used for identifying the genus Legionella. The PCR products of the 226-bp fragment, after it was reacted with HpyCH4III endonuclease, could lead to the specific differentiation of L. pneumophila and non-L. pneumophila. The key for this scheme was that the 226-bp fragment from L. pneumophila strains would be digested to 180- and 46-bp fragments by HpyCH4III, while the 226-bp fragment from a non-L. pneumophila strain either could not be digested or was digested with 153- and 73-bp fragments (digested at the bp 151 to 155 site).
Primer and template were selected to target the 226-bp fragment of the 16S RNA gene for the genus Legionella. This PCR assay was optimized to be simple and versatile. More than 240 genomic DNA samples of the genus Legionella, including L. pneumophila and non-L. pneuophila strains, were analyzed with the PCR. The results for the 226-bp fragments of these samples are shown in Fig. Fig.2a2a.
The goal of this study was to design a method for the rapid detection and identification of Legionella species. PCR is considered a rapid method with high sensitivity. However, the specificity of PCR is important in the assay. To determine the specificity of the primer, 12 different strains, including bacillus, Gram-negative, and Gram-positive strains, were used in the experiment. Genomic DNA extraction and PCR methods were performed with the same procedures as those for the Legionella strains. The PCR results showed the Legionella strains had the 226-bp fragments while all 12 of the other bacterial strains did not have them, even with the annealing temperature lowered to 50°C. (Fig. (Fig.33).
PCR products from Legionella strains were purified with a TIANquick Midi Purification kit and then digested by HpyCH4III endonuclease. The digested products were run on agarose electrophoresis to identify fragment sizes.
In this study, 161 strains of L. pneumophila and 81 strains of non-L. pneumophila were analyzed with the two-step scheme. Among them, 16 strains of L. pneumophila and 26 non-L. pneumophila species were ATCC reference strains. The other 200 isolated strains were environmental isolates that had been identified to be L. pneumophila or non-L. pneumophila by fatty acid analysis, biochemical analysis, and sequence typing analysis. The following method was used to differentiate L. pneumophila and non-L. pneumophila: a strain was L. pneumophila if its 226-bp fragment of the 16S rRNA gene was digested to two fragments with sizes of 46 and 180 bp; a strain was non-L. pneumophila if its 226-bp fragment either was not digested or was digested to two fragments with sizes of 153 and 73 bp. The experimental results from the reference strains are shown in Table Table1.1. One hundred sixty of 161 strains of L. pneumophila had the digested 46- and 180-bp fragments, with only one strain of L. pneumophila whose 226-bp fragment was not digested (Fig. (Fig.2b),2b), an accuracy of 99.38%. Of the 81 non-L. pneumophila species tested by this scheme, all were correctly identified, an accuracy of 100%. Among them, 58 had their 226-bp fragments undigested by HpyCH4III (Fig. (Fig.2b);2b); the other 23 strains had their 226-bp fragments digested to two fragments with sizes of 153 and 73 bp. Twenty of these 23 digested strains were L. feeleii. The other three were birminghamiensis, L. nautarum, and L. quinlivanii.
Overall, 241 of 242 strains of L. pneumophila (n = 161) and non-L. pneumophila (n = 81) were correctly identified by PCR and differentiated by the enzymatic digestion assay, an accuracy of 99.59%. Experiment results for all strains are shown in Tables Tables33 and and44.
Many methodologies have been reported to detect Legionella species. These methods, including culture (6), DFA (5, 11), specific PCR (2), and real-time PCR (21), all had their limitations. Another method using mip or 16S rRNA gene sequencing was promising because it offered sequence-based identification for L. pneumophila and non-L. pneumophila species (19, 22), but it was complex and time-consuming. Therefore, a rapid, accurate, and specific method for the identification and differentiation of L. pneumophila and non-L. pneumophila species is strongly desired.
This study demonstrated a two-step rapid identification scheme for the genus Legionella and the differentiation of Legionella pneumophila and non-Legionella pneumophila species. The scheme is based on a PCR amplification followed by an enzymatic digestion analysis of a 226-bp fragment of the 16S rRNA gene. The PCR product of the 226-bp fragment from the 16S rRNA gene was the key element in discriminating the Legionella species from other strains, while the HpyCH4III enzymatically digested fragments of the 226-bp fragment were used to differentiate L. pneumophila and non-L. pneumophila species. The PCR method was more sensitive and reliable for detecting the genus Legionella than DFA and culture (3, 5, 6, 11, 17). The PCR method in this study correctly identified all of the 242 strains of Legionella, which included 161 strains of L. pneumophila and 81 strains of non-L. pneumophila species. The other bacterial strains, including Salmonella typhi, Enterobacter cloacae, Klebsiella pneumoniae, etc., were successfully separated from Legionella by this method. The sensitivity of the PCR assay in this scheme was 102 CFU (data not shown). The base sequence ACNGT in the 16S rRNA gene of L. pneumophila, and the polymorphisms between L. pneumophila and non-L. pneumophila species, were the essential part of this scheme. This is the first time that this base sequence has been used to differentiate between L. pneumophila and non-L. pneumophila species. The methods of the sequencing of the mip and 16S rRNA genes also utilized gene polymorphisms and interspecies sequence variation to discriminate L. pneumophila and non-L. pneumophila, but they were time-consuming and thus not practical for clinical purposes.
Our bioinformatics study found the unique nature of polymorphisms for the 226-bp fragment of the 16S rRNA gene between L. pneumophila and non-L. pneumophila. The 226-bp fragments from the PCR assay for L. pneumophila could be digested by HpyCH4III to two fragments with sizes of 180 and 46 bp, while those from non-L. pneumophila either would not be digested or would be digested but with two different sizes of fragments (153 and 73 bp). One hundred sixty of 161 L. pneumophila strains had their PCR products digested into 180- and 46-bp fragments (Fig. (Fig.2b),2b), an accuracy of 99.38%. All 81 strains of non-L. pneumophila species were successfully discriminated from L. pneumophila because their 226-bp fragments either were not digested by HpyCH4III (n = 58) or were digested but with fragments of 153 and 73 bp (n = 23), an accuracy of 100%.
The specificity of primers for Legionella also was evaluated. Twelve different bacterial strains (non-Legionella species), including bacillus, Gram-negative, and Gram-positive strains, were used in the evaluation. Species of bacillus were more important controls for the study because they are closer to Legionella morphologically and genetically, as many have assumed. The targeted 226-bp fragments were from the 16S rRNA genes that were considered to be an important marker of evolution. The PCR amplification for all 12 of these bacterium strains (5 were bacillus) did not have the targeted 226-bp products, indicating the primer's specificity for Legionella.
In conclusion, we have developed a novel two-step scheme for the rapid identification of Legionella species and the differentiation of L. pneumophila and non-L. pneumophila species. This assay system is accurate and takes only about 4 h, whereas the traditional culture method takes 3 to 10 days. Experimental results showed that the two-step method achieved an accuracy of 99.59% (241 of 242). We hope that this convenient two-step scheme can be the choice for the rapid diagnosis of L. pneumophila and non-L. pneumophila infections, especially for the non-L. pneumophila clinical diagnosis.
The research was supported by the Ministry of Science and Technology of the People's Republic of China and by a National High Tech grant (project no. 2008ZX10004-006).
Published ahead of print on 9 December 2009.