Sequence-Based Classification Scheme for the GenusLegionella Targeting the 5S rRNA Gene

We read with interest the article by Ratcliff et al. (3) concerning a sequence-based classification ofLegionella targeting the mip gene. We have used a similar classification system based on the smaller 5S rRNA gene (104 bp) which, while less discriminatory than the mip gene, enables species-specific identification of most Legionellaspecies implicated in human disease.

Our classification scheme was constructed from the 5S rRNA sequences of type strains deposited in the GenBank database by Heidrich et al. (1, 4) and others. All species in the database, including eight species thus far isolated only from environmental sources, can be distinguished from each other, apart from L. longbeachaeserogroup 2, L. dumoffii, and L. hackeliae, which have identical 5S rRNA sequences. Several other species exhibit differences of only one or two bases. However, L. pneumophila is easily distinguished from other species by a signature sequence at bp 61 to 63 (TTT). While the 5S rRNA genes of many L. pneumophila strains have been sequenced, the corresponding sequences for other species are currently based on only a few strains.

We have used this scheme to identify both Legionellaisolates and Legionella DNA detected in urine and serum samples by PCR using the following methods. DNA from bacterial isolates was extracted by an alkaline lysis method, where 50 μl of bacterial suspension (10⁹ CFU/ml) was added to 50 μl of 0.05 M NaOH, heated at 95°C for 15 min, and then neutralized by the addition of 10 μl of 1 M Tris (pH 7.7). Urine (200 μl) and serum (100 μl) samples were incubated in a 550-μl lysis solution containing 8.6 mM Tris (pH 8.3), 1.1 mM EDTA, 0.01% (wt/vol) glycogen, 0.9% (wt/vol) sodium dodecyl sulfate, and 0.45 mg of proteinase K/ml for 30 min at 56°C, and DNA was extracted with phenol and chloroform. The 5S rRNA gene was amplified by PCR (2) with M13-tailed primers. Products to be sequenced were purified by using a High Pure PCR Product Purification kit (Boehringer, Mannheim, Germany). Cycle sequencing reactions were performed by using a Thermo Sequenase fluorescence-labelled primer cycle sequencing kit with 7-deaza-dGTP (Amersham Life Science, Little Chalfont, United Kingdom) and IRD-41-labelled infrared M13 forward and reverse primers. Sequencing reaction products were analyzed on a LI-COR 4000L automated DNA sequencer (LI-COR Inc., Lincoln, Nebr.).

We have tested cultures of the following species using these methods:L. pneumophila serogroups 1 to 14, L. anisa,L. israelensis, L. birminghamensis, L. micdadei, L. longbeachae serogroup 1, L. wadsworthii, L. oakridgensis, L. dumoffii,L. jordanis, L. feelei serogroup 2, L. hackeliae, L. macearchernii, L. sainthelensi, and L. bozemanii serogroup 2. All results confirmed identifications by conventional methods. Of the nine wild strains tested (L. pneumophila [4],L. longbeachae [2], L. bozemanii[2], and L. feelei), the L. pneumophila strains showed 100% homology with the database, while the other species exhibited one or two base differences. This approach has also enabled us to identify to species level LegionellaDNA detected in urine and serum samples from patients with pneumonia (unpublished data).

The 5S rRNA target has insufficient discriminatory power if genotypingLegionella species is the primary goal. However, this scheme may be useful for laboratories committed to a 5S rRNA PCR system, especially for the differentiation of L. pneumophila from other species.

• Copyright © 1999 American Society for Microbiology