The usual target for sequence-based identification of Leptospira species is the 16S rRNA gene. However, because the 16S rRNA gene is not polymorphic enough, it is necessary to sequence a 1500 bp segment of this gene for accurate identification. Based on the alignment of previously determined rpoB of three Leptospira strains, we designed and tested a primer pair that enabled us to amplify and sequence a 600 bp segment of Leptospira rpoB. This segment was species-specific for the 16 species tested, but was unable to separate Leptospira interrogans serovars accurately. For the 11 L. interrogans serovars tested, only seven genotypes could be determined. We thus think that analysis of partial rpoB may be useful as an initial screening test for the identification of a new isolate of Leptospira and detection or identification of Leptospira in clinical or environmental samples, but not for serovar determination.
The genus Leptospira is an incredibly varied group of organisms, containing hundreds of serovars and genetic types, which can occupy diverse environments, habitats and life cycles. Included within the genus are highly pathogenic host-specific strains and harmless free-living waterborne strains. Historically, this genus is divided into two species, namely the pathogenic (e.g., Leptospira interrogans) and the free-living saprophytic (e.g. Leptospira biflexa) species. The latter is indigenous to fresh surface water (Johnson & Faine, 1984). Of late, DNA–DNA hybridization techniques have been used to classify the genus Leptospira into 17 genomospecies (Yasuda et al., 1987; Brenner et al., 1999; Levett, 2001). According to this scheme, pathogenic Leptospira comprise eight species. These are L. interrogans sensu stricto, Leptospira weilii, Leptospira borgpetersenii, Leptospira noguchii, Leptospira santarosai, Leptospira alexanderi, Leptospira kirschneri (formerly Leptospira alstoni) and Leptospira genospecies 1. This species assignment is consistent with the results of the phylogenetic analyses based on the rrs gene, which codes for the 16S rRNA gene (Hookey et al., 1993; Postic et al., 2000). However, the discriminatory power of even complete rrs (size: c. 1500 bp) may be limited, due to the absence of a high degree of polymorphism in this gene.
Previously, we have demonstrated the usefulness of rpoB sequence analysis to help differentiate bacterial species, including spirochetes (Renesto et al., 2000; Khamis et al., 2004). This prompted us to design a pair of universal primers to amplify and sequence a small fragment of rpoB gene. The objective was to evaluate whether an such analysis would be useful for routine identification of Leptospira species.
Materials and methods
To begin with, we aligned the 3678 bp sequences of the rpoB of Leptospira available in the GenBank (L. biflexa Patoc, AF150880; L. interrogans Lai, NC-004342; L. interrogans Copenhageni, NC-005823). For this, we used a standard software (clustalw: http://npsa-pbil.ibcp.fr). Then, we used svarap software to determine the most variable area bordering the conserved regions (Khamis et al., 2004; Colson et al., 2005). This enabled us to design a pair of primers. The pair included Lept 1900f (5′-CCT-CAT-GGG-TTC-CAA-CAT-GCA-3′) and Lept 2500r (5′-CGC-ATC-CTC-RAA-GTT-GTA-WCC-TT-3′) bordering a hypervariable region between the position 1900 and 2500. Using this pair of primers, we have amplified and sequenced this area in 19 L. interrogans strains (Table 1). In this study, we have focused our attention on each species and several serovars of L. interrogans. We did this considering the fact that L. interrogans is the most common Leptospira spp. pathogenic to humans. Strains DNA was extracted using the QIAamp DNA extraction kit (Qiagen, Hilden, Germany) according to the manufacturer's recommendations. All PCR mixtures contained 2.5 × 10−2 U of Taq polymerase µL−1, 1 × Taq buffer, 1.8 mM MgCl2 (Gibco BRL, Life Technologies, Cergy Pontoise, France), 200 µM concentrations of dATP, dCTP, dTTP and dGTP (Boehringer Manheim GmbH, Hilden, Germany) and 0.2 µM concentrations of each primer (Eurogentec, Seraing, Belgium). PCR mixtures were subjected to 35 cycles of denaturation at 94°C for 30 s, primer annealing at 51°C for 30 s and extension at 72°C for 2 min. Every amplification program began with a denaturation step of 95°C for 2 min and ended with a final elongation step of 72°C for 10 min. Amplicons were purified for sequencing using a QIAquick spin PCR purification kit (Qiagen) following the protocol of the supplier. Sequencing reactions were carried out with the same pair of primers and the reagents of ABI Prism 3100 DNA sequencer (dRhod. Terminator RR Mix, Perkin Elmer Applied Biosystems) following the standard automated sequencer protocol. Multiple sequence alignments and percentages of similarity between the different species with rpoB and 16S rRNA gene were made using clustalw on the EMBL-EBI web server (http://www.ebi.ac.uk/clustalw/). Phylogenetic trees were obtained from DNA sequences by the neighbour-joining method using the mega 3.1 software (http://www.megasoftware.net).
Only sequences in bold type were determined and deposited in Genbank in the present study.
Results and discussion
The sequence similarities in the 600 bp segment of rpoB among the species tested varied between 60.5% and 92.3% (Table 2). In case of the 1500 bp segment of rrs, such figures varied between 88% and 99.7% (Table 3). For three major groups of species, the topology of the tree constructed using partial rpoB (Fig. 1) was identical to the tree we constructed using rrs (Fig. 2). However, the intragroup topology was different among these three major groups of species. For example, the branches were more widely separated in the rpoB tree. However, the bootstrap value obtained using partial rpoB and rrs was not significantly different (considering nodes with bootstrap value ≥75%, 6/13 vs. 9/14, respectively, P=0.3 by a χ2 test). Using a partial rpoB sequence, we were able to distinguish accurately 10 serovars of L. interrogans from other species. This demonstrates the usefulness of partial rpoB sequence in the identification of Leptospira down to the species level.
Dendrogram representing phylogenetic relationships among Leptospira species. This dendrogram was constructed using the neighbour-joining method. The tree was derived from alignment of partial rpoB sequences. The support of each branch, as determined from 1000 bootstrap samples, is indicated by the value at each node (in %). IH means Icterohaemorrhagiae. The sequences used are those listed in Table 1.
Dendrogram representing phylogenetic relationships among Leptospira species. This dendrogram was constructed using the neighbour-joining method. The tree was derived from alignment of 16S rRNA gene sequences. The support of each branch, as determined from 1000 bootstrap samples, is indicated by the value at each node (in %). The sequences used are those listed in Table 1.
Our data demonstrate that rpoB is a good reflection of the gene contents of Leptospira. Apart from what we observed in Leptospira genomospecies 3, there was a good correlation (Fig. 3) between the G+C content of the genome and partial rpoB [coefficient of determination (R2) > 0.7]. Our findings were similar to those obtained by others (Fournier et al., 2006). In Leptospira genomospecies 3, there was no good correlation between the G+C content of the genome and partial rpoB. The reason for this discrepancy remains unclear to us. Future studies should address this issue, as due to technical limitations the determination of G+C content may not always be reproducible.
Plot of the rpoB G+C content vs. complete genome for 13 Leptospira species (Leptospira genomospecies 1, Leptospira santarosai, Leptospira interrogans, Leptospira noguchii, Leptospira borgpetersenii, Leptospira wolbachii, Leptospira biflexa, Leptospira meyeri, Leptospira genomospecies 3 each, Leptospira genomospecies 5, Leptospira weilii, Leptospira inadai, Leptospira alexanderi). Apart from Leptospira genomospecies 3 (dot), all species are represented by squares. A regression line is fitted to the data. The coefficient of determination and the tendency curve equation are indicated.
As is the case with rrs, a single pair of primers enables us to amplify partial rpoB of any representative strain of the 15 species/genomospecies belonging to this diverse genus. However, in Leptospira species, the analysis of partial rpoB offers two additional advantages. Firstly, a 600 bp fragment of rpoB may be amplified and sequenced in two runs of sequence using the same pair of primers used for PCR amplification. This is not the case for large fragments such as rrs, which requires approximately six primers and thus six runs of sequence to be accurately sequenced. Secondly, compared with rrs, the degree of polymorphism is higher in rpoB amplicons. This means that rpoB may be more useful than rrs for the identification of Leptospira species. For examples, in L. kirschneri Cynopteri and L. interrogans Canicola, the positions of only three nucleotides are different in 1500 bp sequences of rrs. In contrast, for the same species, the positions of 51 nucleotides are different in 600 bp fragment of rpoB. Owing to the convenience of this method, we amplified genomospecies 1, 3 and 5. To our knowledge, there is no published report that has examined the phenotypic characters of these genomospecies. Previously, it has been suggested that genomospecies 3 and 5 might be free–2 living species. Our data (Fig. 1) confirm this notion, for both genomospecies 3 and 5 cluster with Leptospira wolbachii, L. biflexa and Leptospira meyeri. In contrast, genomospecies 1 clusters with pathogenic Leptospira. Genomospecies 1 comprises two serovars, both of which have been isolated from frogs in China. The other serovars isolated from frogs include Ranarum and Bim in the USA and Barbados, respectively. The former serovar belongs to L. meyeri, a free–2 living species and the latter to L. kirschneri, a pathogenic species.
Several molecular techniques have been evaluated for the identification of Leptospira species or serovars. These include random amplified polymorphic DNA, arbitrarily primed PCR, use of insertion sequences in PCR-based assays, restriction length polymorphism, specific probes, variable number tandem repeat analysis and pulsed-field gel electrophoresis (Barocchi et al., 2001; Levett, 2001; Majed et al., 2005). However, none of these techniques is based on sequence analysis. This may be problematic as far as the reproducibility among different laboratories is concerned. The Leptospira genes coding for outer membrane proteins have been sequenced for six pathogenic species. These genes transfer horizontally from one species to the other. This makes these genes unsuitable for the identification of various Leptospira species (Haake et al., 2004).
Using rpoB sequence analysis, we were able to separate 11 L. interrogans serovars tested in this study and place them under seven genotypes. As had been the case with rrs (Postic et al., 2000), this technique separated several genogroups. However, it failed to distinguish clearly the member of one serovar from that of the other. This confirms the previously held notion (Herrmann et al., 1992; Brenner et al., 1999; Levett, 2001) that the same serovar can be found in different genetic groups/species. Analyses of several strains of the most commonly encountered serovars may prove useful for defining the contribution of rpoB sequencing to the molecular screening of Leptospira isolates. Such information may, in turn, prove invaluable for the determination of various serovars of Leptospira. Alternatively, analysis of a segment of rpoB may be useful as an initial screening test for the identification of a new isolate of Leptospira using a system of similarity cut-off to define species (La Scola et al., 2003). If the partial rpoB similarity of a test isolate is lower than 92%, it should be regarded as a new species. On the other hand, if such a value goes above 97%, the isolate under scrutiny should be regarded as being representative of a known species. However, to be used for the routine identification of Leptospira, these cut-offs will have to be validated on large collections of isolates, as we did for Corynebacterium sp. (Khamis et al., 2005). Finally, this technique may be useful for the detection as well as identification of Leptospira in clinical or environmental samples, provided the specificity of the primers has been verified previously.
(1999) Futher determination of DNA relatedness between serogroups and serovars in the family Leptospiraceae with a proposal for Leptospira alexanderi sp. nov and four new Leptospira genomospecies. Int J Syst Bacteriol 49: 839–858.