Printer Friendly

Isolation of Legionella pneumophila from cooling towers, public baths, hospitals, and fountains in Seoul, Korea, from 2010 to 2012.


Since Legionella pneumophila was identified as the causative agent in the mass outbreak of pneumonia at the 1976 American Legion Convention in Philadelphia, several mass outbreaks and sporadic individual incidences have been reported worldwide (Fraser et al., 1977). An increased incidence of legionellosis since 2000 has been reported in the U.S. (Hicks, Garrison, Nelson, & Hampton, 2012) and Europe (European Centre for Disease Prevention and Control, 2012). Legionella detection tests were performed on specimens collected from patients with community-acquired pneumonia in 15 different hospitals in Korea from 2005 to 2011; the results of these tests revealed two (0.9%) Legionella-positive cases. Therefore, Legionella spp. is also considered a cause of pneumonia in Korea (Cho et al., 2012).

In 2008, 4,938 sets of environmental water samples were collected from 16 local autonomous entities in Korea. Analysis of these water samples revealed 560 (11.3%) positive cases. L pneumophila was detected in 85% of these positive samples, in which L. pneumophila serogroup 1 (sg1) was predominant (Lee, Shim, Kim, Yu, & Kang, 2010). Pathogenic bacteria are often present as hidden contaminants in drinking water treatment. The Legionella sequences were found even after membrane purification, indicating the continued risk of legionellosis (Kwon, Moon, Kim, Hong, & Park, 2011).

Two cases of legionellosis were reported in Korea in 2001, and less than 10 cases were reported between 2002 and 2005. After 2006, however, the number of reported legionellosis cases increased to approximately 20-30 annually; in Seoul, an average of seven or eight cases of legionellosis have been reported each year since 2006 (except in 2011). From 2011 on, death due to legionellosis has been reported in Korea: one and three cases of deaths were reported in 2011 and 2012, respectively, of which one death in each year occurred in Seoul (Korea Centers for Disease Control & Prevention, 2013). Seoul has a very high population density of 16,936 people/[km.sup.2], with 10.25 million inhabitants in an area of 605.21 [km.sup.2]. An increasing number of people are being exposed to pathogens, often because people frequently visit public facilities such as hotels, shopping malls, hospitals, and public baths.

Pulsed-field gel electrophoresis (PFGE) patterns of samples from patients with legionellosis were identical to those observed in the water distribution system in five of the nine hospitals investigated from 1989 to 2006 (Garcia-Nunez et al., 2008). Another study suggested that more studies need to be conducted on environmental contamination by Legionella spp. (Jonas et al., 2000). In Korea, PFGE testing was used to identify L. pneumophila sg1 in environmental water and clinical specimens collected from 1985 to 2007. The results indicated that certain PFGE types persisted over several years and also showed that some types of PFGE patterns can be found in both environmental water and clinical specimens (Lee, Kang, & Yu, 2010).

The objectives of our study were to analyze water samples collected from the major habitats of L pneumophila (including cooling towers, public baths, hospitals, and fountains) as a step toward establishing preventive measures for legionellosis in Seoul, and to investigate the serological and molecular biological characteristics of L. pneumophila to collect epidemiological data and identify characteristic PFGE patterns. The epidemiological data obtained in our study will help quickly identify the sources of legionellosis outbreaks to prevent disease spread and recurrence.

Materials and Methods

From 2010 to 2012, 3,495 water samples were collected from several public facilities. The samples were categorized by origin (cooling towers, public baths, hospitals, or decorative fountains) and then analyzed. More than 40 water tests were performed on the samples collected from each of the 25 districts in Seoul. Water sterilization during the sample collection period was not considered. Water samples were collected annually from May to September. Isolates of Legionella spp. (n = 527) were identified by microbial culture and polymerase chain reaction (PCR) with mip--and rpoB-specific primers. Serological diagnosis and PFGE were also performed.

Culture of Water Samples

At each collection point, 1 L of water was obtained and filtered through 0.2-[micro]m nitrocellulose membranes under vacuum. Each membrane was plated on buffered charcoal yeast extract (BCYE) agar with L-cysteine and antibiotic supplements and incubated for 10 days at 37[degrees]C. Colonies growing on BCYE agar, but not on blood agar, were definitively identified as Legionella spp.


PCR analysis was performed to determine whether the L. pneumophila isolates possessed the rpoB (RNA polymerase B) (Ko et al., 2003; Nielsen, Hindersson, Hoiby, & Bangsborg, 2000) and mip (Lindsay, Abraham, & Fallon, 1994) genes, which encode toxin production. The reactions were conducted in a thermal cycler.

Agglutination Test

L. pneumophila isolates were cultured on BCYE agar, and their growth was confirmed using a serotyping kit. After the isolates were placed in physiological saline at a high concentration, they were heated at 100[degrees]C for one hour followed by serogroup determination based on the results of the slide agglutination test.

Chromosomal PFGE Analysis

Band patterns were obtained by Sfil digestion and by following an established protocol with minor modifications. The gels were illuminated with a UV light source and photographed. Macrorestriction patterns were analyzed with BioNumerics software (version 3.5) in a similar manner to that used for subgrouping.


L. pneumophila Detection

From 2010 to 2012, more than 1,000 water samples were collected annually. Among the 3,495 water samples collected throughout this period, L. pneumophila was identified by culture and PCR analysis. L. pneumophila was detected in 210 (17.5%), 195 (15.5%), and 122 (11.8%) samples in 2010, 2011, and 2012, respectively, indicating an average annual detection rate of 15.1%. Furthermore, L. pneumophila was detected in 239 (21.0%) of the 1,140 samples, 207 (17.0%) of the 1,215 samples, 75 (8.0%) of the 940 samples, and 5 (4.7%) of the 93 samples collected from cooling towers, public baths, hospitals, and fountains, respectively (Table 1).

Serological Diagnosis, PFGE Patterns, and Correlation Analysis According to the Sample Type and Serogroup

Serological diagnosis of sg1 through sg6 was performed for all 527 isolates. Moreover, 30% of these isolates were selected using proportional stratified sampling for each sample type based on the detection rate, and 170 (32%) isolates were analyzed by PFGE. These isolates were classified into groups A-D based on at least 65% homology; we confirmed that the serogroups and distribution of each sample type varied among the groups.

The Chi-square test revealed the relationship between PFGE patterns and serogroups. In Group A, sg1 isolates constituted a relatively high proportion (n = 36, 76.6%) of the whole sample population (60.6%). Compared to Group A, Group B possessed fewer sg1 isolates (n = 19, 47.5%) and more sg5 isolates (n = 9, 22.5%). Group C showed various serological distribution. Specifically, the number of sg1 isolates in Group C was slightly lower than the average distribution (n = 29, 50%), whereas 8 (13.8%) and 15 isolates (25.9%) were detected as sg3 and sg6, respectively, which was greater than the number indicated by the average distribution. Similar to Group A, Group D had 19 sg1 isolates (76.0%) and three sg2 isolates (12.0%), which was greater than the number indicated by the average distribution; Group D had no sg4, sg5, or sg6 isolates. In general, sg4 was not detected in any of the groups (Table 2).

As shown in Table 3, the Chi-square test was used to verify the relationship between the PFGE pattern and sample type. In Group A, 36 isolates (76.6%) were detected in cooling tower samples, which constituted a much higher proportion of the total isolates detected from cooling towers (46.4%). In Group B, a remarkably high number of isolates (n = 37, 92.5%) was detected in public bath samples. Interestingly, none of the isolates was detected in samples from the cooling towers, whereas a small number of isolates were identified in hospital samples (n = 3, 7.5%). In Group C, 20 isolates (35.7%) were detected in samples from cooling towers, which represents a lower distribution compared to the total number of isolates detected, whereas a relatively higher number of isolates (n = 15, 26.8%) were detected in samples from hospitals. Only two isolates of Group C were detected in fountain samples (data not shown because of the low overall frequency). In Group D, a high number of isolates (n = 22, 88.0%) were detected in cooling tower samples, indicating a pattern similar to Group A, and three isolates (12.0%) were detected in public bath samples. No strains were detected at the other facilities (Table 3).

Based on the fact the serogroups and sample types are correlated to yield specific PFGE patterns, we investigated the correlation between serogroups and sample types. In cooling tower samples, sg1 isolates constituted the highest proportion (n = 61, 78.2%). In samples from public baths and hospitals, sg1 isolates were detected in high numbers, whereas sg3, sg5, and sg6 isolates were detected at similar rates (constituting ~50%). Two isolates were detected in fountain samples (data not shown because of low frequency; Table 4).


With time, utilization of public facilities has increased, leading to increased sharing of space by a large number of randomly gathered people. Legionella spp., the causative organism of legionellosis, is frequently isolated from public facilities, and more studies and information are needed to prevent legionellosis and to effectively block its transmission. Therefore, under the assumption that distinctive types of L. pneumophila are present in unique niches at different public places owing to the different environmental conditions, we investigated their distinctive characteristics using PFGE. In the analysis of PFGE patterns based on 65% homology, we verified the serogroups and characteristics of samples from public facilities according to the groups of L. pneumophila. Among all PFGE groups (A-D), sg1 was detected most frequently. In particular, isolates belonging to Groups A and D, which had the highest percentages of sg1 isolates, were detected in cooling tower samples in large quantities. In contrast, sg3, sg5, and sg6 were evenly distributed in Groups B and C. Many Group B isolates were detected in public bath samples. Group C isolates were distributed among the samples from cooling towers, public baths, and hospitals. In the 2008 Korean study, sg1 was the most prevalent serogroup based on sequence-based typing. That study also showed that the second most dominant strains differed with facility type (Lee, Shim, et al., 2010). In our study, we verified using the Chi-square test of independence that the sample type, serogroup, and PFGE pattern are closely correlated. On the basis of these results, we conclude that facility-dependent different ecological environments affect the niche-formation behaviors of L. pneumophila.

Because cooling towers are exposed to the external environment and thus are more vulnerable to contamination through the infiltration of fugitive dust and particulate organic matter, we anticipated we would observe various kinds of serotype distribution; however, the results were surprising. Only the distribution for Groups A and D showed varying genotype distribution. In contrast, the samples collected from public baths and hospitals contained various serogroups other than sg1. In Korea, bathing in public baths has become more popular in recent years. This raises hygienic concerns about various contaminants such as dead skin cells, cosmetics, and body lotions, as well as people entering the tub without taking a shower first, which creates a suitable living environment for L. pneumophila. Thus, we predicted that such public baths would have niches for different kinds of serogroups. Indeed, different kinds of Legionella species and serogroups (sg1, sg5, and sg6) were found in the hot water distribution system in spas in the Czech Republic (World Health Organization, 2007). In hospitals, inpatients, outpatients, medical staff, guardians, and others continuously enter and exit the building, and a number of medical devices, namely, humidifiers, nebulizers, and respiratory machines, are constantly used, offering favorable environments for the niche construction of various types of L. pneumophila. According to a recent report, sg1, sg5, and sg6 were also detected in medical facilities in Russia (Gruzdeva, Tartakovskii, & Mar'in, 2012).

Therefore, the PFGE characteristics of Legionella from each public facility may provide useful information for the rapid identification of pathogens and assist in the implementation of preventive measures for future outbreaks. The results of PFGE pattern analysis of isolates originating from facilities can be used in epidemiological investigations.


Our study identified L. pneumophila isolates in samples collected from public facilities and classified them according to their biological and molecular genetic characteristics. The PFGE analysis revealed different characteristics of L. pneumophila in each public facility; it was thus confirmed that different serogroups of L. pneumophila constructed varying niches according to the types of public facilities. Ml

Acknowledgements: The authors would like to thank Director Seokju Cho and Miok Song, PhD, from the Institute of Health and Environment, Seoul, Korea, for providing invaluable advice and input.


Cho, M.-C., Kim, H., An, D., Lee, M., Noh, S.-A., Kim, M.-N., Chong, Y.P., & Woo, J.H. (2012). Comparison of sputum and nasopharyngeal swab specimens for molecular diagnosis of Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella pneumophila. Annals of Laboratory Medicine, 32(2), 133-138.

European Centre for Disease Prevention and Control. (2012). Legionnaires' disease in Europe 2010. Stockholm: Author.

Fraser, D.W, Tsai, T.R., Orenstein, W, Parkin, W.E., Beecham, H.J., Sharrar, R.G., Harris, J., Mallison, G.F., Martin, S.M., McDade, J.E., Shepard, C.C., & Brachman, P.S. (1977). Legionnaires' disease: Description of an epidemic of pneumonia. The New England Journal of Medicine, 297(22), 1189-1197.

Garcia-Nunez, M., Sopena, N., Ragull, S., Pedro-Botet, M.L., Morera, J., & Sabria, M. (2008). Persistence of Legionella in hospital water supplies and nosocomial Legionnaires' disease. FEMS Immunology and Medical Microbiology, 52(2), 202-206.

Gruzdeva, O.A., Tartakovskii, I.S., & Mar'in, G.G. (2012). Examination of the contamination aetiological agent of legionellosis in the water supply systems of medical treatment facilities [Article in Russian]. Voenno-medisinskiizhurnal, 333(5), 34-38.

Hicks, L.A., Garrison, L.E., Nelson, G.E., & Hampton, L.M. (2012). Legionellosis--United States, 2000-2009. American Journal of Transplantation, 12(1), 250-253.

Jonas, D., Meyer, H.-G.W, Matthes, P., Hartung, D., Jahn, B., Daschner, F.D., & Jansen, B. (2000). Comparative evaluation of three different genotyping methods for investigation of nosocomial outbreaks of Legionnaires' disease in hospitals. Journal of Clinical Microbiology, 38(6), 2284-2291.

Ko, K.S., Hong, S.-K., Lee, K.-H., Lee, H.K., Park, M.-Y., Miyamoto, H., & Kook, Y.-H. (2003). Detection and identification of Legionella pneumophila by PCR-restriction fragment length polymorphism analysis of the RNA polymerase gene (rpoB). Journal of Microbiological Methods, 54(3), 325-337.

Korea Centers for Disease Control & Prevention. (2013). Infectious diseases surveillance yearbook, 2012. Seoul: Author.

Kwon, S., Moon, E., Kim, T.-S., Hong, S., & Park, H.-D. (2011). Pyrosequencing demonstrated complex microbial communities in a membrane filtration system for a drinking water treatment plant. Microbes and Environments, 26(2), 149-155.

Lee, H.K., Kang, Y.H., & Yu, J.Y. (2010). Genomic diversity of Legionella pneumophila serogroup 1 from environmental water sources and clinical specimens using pulsed-field gel electrophoresis (PFGE) from 1985 to 2007, Korea. Journal of Microbiology (Seoul, Korea), 48(5), 547-553.

Lee, H.K., Shim, J.I., Kim, H.E., Yu, J.Y., & Kang, Y.H. (2010). Distribution of Legionella species from environmental water sources of public facilities and genetic diversity of L. pneumophila serogroup 1 in South Korea. Applied and Environmental Microbiology, 76(19), 6547-6554.

Lindsay, D.S.J., Abraham, W.H., & Fallon, R.J. (1994). Detection of mip gene by PCR for diagnosis of Legionnaires' disease. Journal of Clinical Microbiology, 32(12), 3068-3069.

Nielsen, K., Hindersson, P., Hoiby, N., & Bangsborg, J.M. (2000). Sequencing of the rpoB gene in Legionella pneumophila and characterization of mutations associated with rifampin resistance in the Legionellaceae. Antimicrobial Agents and Chemotherapy, 44(10), 2679-2683.

World Health Organization. (2007). Legionella and the prevention of legionellosis. Geneva: Author.

Changkyu Kim, MS

Sujin Jeon, MS

Jihun Jung, MS

Younghee Oh, PhD

Seoul Metropolitan Government

Research Institute of Public Health

and Environment

Yeonsun Kim, PhD

Department of Public Health

Dankook University

Jaein Lee, PhD

Sungmin Choi, PhD

Youngzoo Chae, PhD

Seoul Metropolitan Government

Research Institute of Public Health

and Environment

Young-ki Lee, PhD

Department of Public Health

Dankook University

Corresponding Author: Young-ki Lee, Professor, Department of Health, Graduate School of Health & Welfare, Dankook University, 119 Dandaero, Dongnam-gu Cheonan-si Chungnam, Korea 330-714. E-mail:
Number of Positive Samples for L. pneumophila Isolated From Public
Facilities in Seoul Over Three Years

Year                            Isolates (%) (a)

            Cooling       Public     Hospital    Fountain   Others
             Tower         Bath

2010       95 (22.2)    85 (19.4)    28 (10.7)   2 (3.9)    0 (0.0)
2011       91 (22.3)    71 (16.2)    32 (9.0)    0 (0.0)    1 (3.7)
2012       53 (17.4)    51 (15.1)    15 (4.7)    3 (11.5)   0 (0.0)
Subtotal   239 (21.0)   207 (17.0)   75 (8.0)    5 (4.7)    1 (1.1)

Year           Total

2010         210 (17.5)
2011         195 (15.5)
2012         122 (11.8)
Subtotal   527 (15.1) (b)

(a) Water sampling for cooling towers was done at the cooling
tower basins of large-scale shopping malls, department houses,
hotels, large buildings, and hospitals; at public bath houses in
taps, showers, and hot tubs; and at hospitals in toilets and
showers. The fountains were decorative fountains in parks and

(b) The average detection rate was 15.1% in decreasing order of
cooling towers, public baths, hospitals, and fountains.

Contingency Table Examining the Relationship Between the PulsedField
Gel Electrophoresis (PFGE) Group and Serogroup

PFGE                Serogroup (%) (a)

               1           2          3

A          36 (76.6)    2 (4.3)    0 (0.0)
B          19 (47.5)    3 (7.5)    4 (10.0)
C          29 (50.0)    1 (1.7)    8 (13.8)
D          19 (76.0)    3 (12.0)   3 (12.0)
Subtotal   103 (60.6)   9 (5.3)    15 (8.8)

PFGE          Serogroup (%) (a)       Total

               5           6

A          5 (10.6)     4 (8.5)    47 (100.0)
B          9 (22.5)    5 (12.5)    40 (100.0)
C           5 (8.6)    15 (25.9)   58 (100.0)
D           0 (0.0)     0 (0.0)    25 (100.0)
Subtotal   19 (11.2)   24 (14.1)   170 (100.0)

Note. df = 12, p = .95, [chi square] = 33.3 (Chi-square observed)
> [chi square] = 21.0 (Chi-square expected), p = .001. PFGE
groups and serogroups were confirmed to be correlated.

(a) Overall, sg1 accounted for the largest proportion in all four
groups. Groups A and D made up the highest percentages of sg1
with highly concentrated distribution patterns, whereas Group B
and C showed even distributions of sg3, sg5, and sg6 apart from

Contigency Table Examining the Relationship Between the PulsedField
Gel Electrophoresis (PFGE) Group and Sample Type

PFGE Group               Sample Type (%) (a)        Total

              Cooling     Public     Hospital
               Tower       Bath

A            36 (76.6)   6 (12.8)    5 (10.6)    47 (100.0)
B             0 (0.0)    37 (92.5)    3 (7.5)    40 (100.0)
C            20 (35.7)   21 (37.5)   15 (26.8)   56 (100.0)
D            22 (88.0)   3 (12.0)     0 (0.0)    25 (100.0)
Subtotal     78 (46.4)   67 (39.9)   23 (13.7)   168 (100.0)

Note. df = 6, p = .95, [chi square] = 91.7 (Chi-square observed)
> [chi square] = 12.6 (Chi-square expected), p = .000. PFGE
groups and sample types were confirmed to be correlated.

(a) Groups A and D isolates showed the highest distributions at
cooling towers, and Group B at public baths, whereas Group C
isolates were evenly distributed at cooling towers, public baths,
and hospitals.

Contingency Table Examining the Relationship Between Serogroup
and Sample Type

Sample Type (a)                Serogroup (%)

                      1           2         3

Cooling Tower     61 (78.2)    4 (5.1)   1 (1.3)
Public Bath       32 (47.8)    4 (6.0)   8 (11.9)
Hospital          10 (43.5)    1 (4.3)   5 (21.7)
Subtotal          103 (61.3)   9 (5.4)   14 (8.3)

Sample Type (a)        Serogroup (%)      Total

                      5           6

Cooling Tower      4 (5.1)    8 (10.3)    78 (100.0)
Public Bath       12 (17.9)   11 (16.4)   67 (100.0)
Hospital          3 (13.0)    4 (17.4)    23 (100.0)
Subtotal          19 (11.3)   23 (13.7)   168 (100.0)

Note. df = 8, p = .95, [chi square] = 24.1 (Chi-square observed)
> [chi square] = 15.5 (Chi-square expected), p = .002. Serogroups
and sample types were confirmed to be correlated.

(a) Cooling towers showed the highest number of sg1 with a highly
concentrated distribution pattern, whereas public baths and
hospitals showed even distributions of sg3, sg5, and sg6 apart
from sg1.
COPYRIGHT 2015 National Environmental Health Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kim, Changkyu; Jeon, Sujin; Jung, Jihun; Oh, Younghee; Kim, Yeonsun; Lee, Jaein; Choi, Sungmin; Chae
Publication:Journal of Environmental Health
Article Type:Report
Geographic Code:9SOUT
Date:Jan 1, 2015
Previous Article:Hand washing among Palestinians in the West Bank and Gaza Strip: attitudes and practices.
Next Article:Association between [PM.sub.10] and respiratory hospital admissions in different seasons in heavily polluted Lanzhou city.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters