We detected Legionella species in 111 bath water samples and 95 cooling tower water samples by using a combination of conventional plate culture, quantitative polymerase chain reaction qPCR and qPCR combined with ethidium monoazide treatment EMA-qPCR methods. In the case of bath water samples, Legionella spp. were detected in 30 samples by plate culture, in 85 samples by qPCR, and in 49 samples by EMA-qPCR. Of 81 samples determined to be Legionella-negative by plate culture, 56 and 23 samples were positive by qPCR and EMA-qPCR, respectively. Therefore, EMA treatment decreased the number of Legionellapositive bath water samples detected by qPCR. In contrast, EMA treatment had no effect on cooling tower water samples. We therefore expect that EMA-qPCR is a useful method for the rapid detection of viable Legionella spp. from bath water samples.

Viable Legionella spp. in environmental water samples were characterized phylogenetically by a clone library analysis combining the use of ethidium monoazide and quantitative PCR. To examine the diversity of Legionella spp., six cooling tower water samples and three bath water samples were collected and analyzed. A total of 617 clones were analyzed for their 16S rRNA gene sequences and classified into 99 operational taxonomic units (OTUs). The majority of OTUs were not clustered with currently described Legionella spp., suggesting the wide diversity of not-yet-cultured Legionella groups harbored in cooling tower water environments.

No abstract

Acid treatment is performed in the detection of Legionella species from environmental water samples using the plate culture method. The acid treatment functions to eliminate the nontarget heterotrophic bacteria. Usually, an HCL-KCL buffer is used for the acid treatment of samples. However, the buffering capacity of the HCL-KCL buffer is not sufficient to maintain a constant low-pH in highly alkaline samples or highly buffered samples such as those from cooling towers or spas. Our results show that the pH of various water samples ranged from 2.4 to 7.2 after treatment with an HCL-KCL buffer. We demonstrate here that an acid-phosphate buffer produces a low pH in highly alkaline samples or highly buffered samples. The pH of the water samples ranged from 2.2 to 2.5 after treatment with an acid-phosphate buffer. In the results of the examination of 161 samples, the use of an acid-phosphate buffer instead of an HCL-KCL buffer decreased the effect of non-target heterotrophic bacteria from 3.1 % to 1.2 %. Therefore, efficient detection of Legionella from samples of varying water quality can be achieved following pretreatment with an acid-phosphate buffer.

No abstract