High Prevalence, Genetic Diversity and Intracellular Growth Ability of Legionella in Hot Spring Environments

Background Legionella is the causative agent of Legionnaires' disease, and hot springs are a major source of outbreaks of this disease. It is important from a public health perspective to survey hot spring environments for the presence of Legionella. Methods Prospective surveillance of the extent of Legionella pollution was conducted at three hot spring recreational areas in Beijing, China in 2011. Pulsed-field gel electrophoresis (PFGE) and sequence-based typing (SBT) were used to describe the genetic polymorphism of isolates. The intracellular growth ability of the isolates was determined by interacting with J774 cells and plating the dilutions onto BCYE agar plates. Results Overall, 51.9% of spring water samples showed Legionella-positive, and their concentrations ranged from 1 CFU/liter to 2,218 CFU/liter. The positive rates of Legionella were significantly associated with a free chlorine concentration of ≥0.2 mg/L, urea concentration of ≥0.05 mg/L, total microbial counts of ≥400 CFU/ml and total coliform of ≥3 MPN/L (p<0.01). The Legionella concentrations were significantly associated with sample temperature, pH, total microbial counts and total coliform (p<0.01). Legionella pneumophila was the most frequently isolated species (98.9%), and the isolated serogroups included serogroups 3 (25.3%), 6 (23.4%), 5 (19.2%), 1 (18.5%), 2 (10.2%), 8 (0.4%), 10 (0.8%), 9 (1.9%) and 12 (0.4%). Two hundred and twenty-eight isolates were analyzed by PFGE and 62 different patterns were obtained. Fifty-seven L. pneumophila isolates were selected for SBT analysis and divided into 35 different sequence types with 5 main clonal groups. All the 57 isolates had high intracellular growth ability. Conclusions Our results demonstrated high prevalence and genetic polymorphism of Legionella in springs in Beijing, China, and the SBT and intracellular growth assay results suggested that the Legionella isolates of hot spring environments were pathogenic. Improved control and prevention strategies are urgently needed.


Introduction
Legionellosis is an environment-related, acute respiratory infection that is caused by gram-negative bacteria of the genus Legionella. Legionella widely exist in environmental waters, and they are particularly prevalent in warm water. To date, more than 50 Legionella species have been characterized, and 25 species are known to cause human disease [1,2]. Most human infections are caused by Legionella pneumophila, which is responsible for approximately 90% of the identified clinical cases, and the predominant serogroup is serogroup 1 [1,3,4,5].
Transmission of bacteria from the environment to humans occurs via inhalation or aspiration of Legionella-containing aerosols [6,7]. The water systems of artificial facilities, including potable water systems [8,9], spa water [10] and cooling towers [11,12] in large buildings, hotels, hospitals and public baths that are contaminated by Legionella are the implicated sources of outbreaks and sporadic cases of Legionnaires' disease. Thus, it is important from a public health perspective to continually survey environmental water systems for the presence of Legionella [13,14,15,16]. In particular, the contamination of hot-water systems has gradually become recognized as an important risk factor all over the world [17,18,19,20,21,22], and many cases of Legionnaires' disease have been detected in Legionella-contaminated hot-water systems.
In recent years, hot springs have been recognized as a major source of outbreaks of legionellosis. Several outbreaks and a number of cases of Legionella infection that were associated with hot springs have been reported in Japan, Spain and France [23,24,25].
In 2002, a large outbreak of legionellosis that affected two hundred and ninety-five patients occurred in a hot spring in Japan [24,26]. In China, hot springs have become an increasingly popular form of recreation, and many people enjoy bathing in hot spring tubs. It is important to survey the contamination of Legionella in hot springs where people bath to avoid serious outbreaks of Legionnaires' disease. In our previous study that compared the ability of four methods (i.e., culture, PCR, quantitative real-time PCR and EMA-quantitative real-time PCR) to detect Legionella in different types of water systems, we isolated Legionella from spring water samples at a rate of up to 93.3% by quantitative real-time PCR. When using the culture method, the percent-positive rate for Legionella was 54.4%, and the concentrations of Legionella were as high as 1,109 CFU/liter [27]. These results showed high degree of pollution of Legionella in spring water samples in China. However, that study did not systematically focus on the Legionella in springs, and the spring water samples were only used to compare the detection ability of different methods.
The aim of our study is to detect Legionella in hot spring water systems by collecting samples for one year from three hot spring recreational areas in Beijing, China so that we may ascertain whether factors such as temperature, pH and total microbial counts are related to the extent of Legionella pollution. The serogroup of Legionella isolates were determined. Genetic characteristics were analyzed using pulsed-field gel electrophoresis (PFGE) and sequence-based typing (SBT) methods. Furthermore, the intracellular growth ability of the isolates was determined.

Sample collection
We selected three hot spring recreational areas in Beijing as study points. At each area, five spa pools were selected for sampling every month in 2011 except in May, and one to four samples were collected from each spa pool based on the size of the pools. All 308 spa water samples were collected during the oneyear study period, and the procedure for collection and pretreatments of the environmental water samples was based on the protocol according to ISO 11731 (1998). Five hundred milliliters of each water sample was collected from a reservoir or condensation pan and placed in a sterile, screw-capped container. Chlorine in the water samples was inactivated by the addition of sodium thiosulfate.

Ethics statements
Water sample collection in this study was carried out on private lands, and we confirm that the owners of the lands gave permission to conduct the study on these sites. No specific permissions were required for these locations/activities, because there are no endangered or protected species in these hot spring recreational areas and this study did not involve any endangered or protected species.

Legionella detection
To detect Legionella, 200 microliters of the water samples were centrifuged at 6,000 g for 10 min in a 300 ml-capacity centrifuge bottle. The pellets were resuspended in 3 ml of distilled water, and the resuspended pellets were serially diluted 10-fold with sterile water. Diluted and undiluted samples (100 ml each) were plated onto GVPC agar (Oxoid, Hampshire, United Kingdom), and the plates were incubated at 37 uC for 10 days. Colonies were identified as Legionella using the L-cysteine requirement test, Gram staining and slide agglutination using polyclonal antisera (Denka Seiken, Tokyo, Japan).

Physical-chemical and microbiological analyses of water samples
Water temperature and its pH were measured in situ using a portable pH meter (D-24E, Horiba Co., Japan). The samples were transported to the lab immediately. The residual free chlorine concentration was assayed using DPD liquid (HACH, Loveland, USA). The urea concentration was measured using the Standard Method of the Chinese National Institute of Standards (GB/ T18204-2000).
The total microbial counts at 36 uC were obtained twice by the pour-plate method on plate count agar (Oxoid, Hampshire, United Kingdom). The plates were incubated at 36 uC for 48 h, and total coliform was measured by membrane filtration procedures.

PFGE
We used the 1-day, standardized PFGE protocol for L. pneumophila [28]. Cell suspensions were placed in polystyrene tubes (Falcon; 12675 mm), and their optical density was adjusted to 3.8-4.0 using a Densimat photometer (BioMérieux, Marcy l'Etoile, France). Legionella slices were digested using 30 U per slice of AscI (New England Biolabs, Ipswich, MA, USA) for 4 h at 37 uC, and electrophoresis was performed using a CHEF-DRIII system (Bio-Rad Laboratories, Hercules, CA, USA). Electrophoresis was run with a switch time of 6.8 s to 54.2 s for 19 h, and images were captured using a Gel Doc 2000 system (Bio-Rad) and converted to TIFF files. The TIFF files were analyzed using the BioNumerics version 5.1 software (Applied Maths, Kortrijk, Belgium). Similarity analysis of the PFGE patterns was performed by calculating the Dice coefficients (SD) [29], with customized tolerance for each EP, and clustering was created using the unweighted-pair group method with average linkages (UPGMA).

SBT and allelic diversity analyses
Genotyping was conducted via the standard, sequence-based typing (SBT) method of the European Working Group for Legionella Infections (EWGLI) using 7 genes (flaA, pilE, asd, mip, mompS, proA and neuA) [30,31]. The SBT database that was available on the EWGLI website (http://www.ewgli.org/) was used for nucleotide analysis, and the sequences were compared with those in the SBT database, which were also available on the website (http://www.hpa-bioinformatics.org.uk/legionella/ legionella_sbt/php/sbt_homepage.php). Additionally, sequence types (STs) that were not available in the EWGLI SBT database were represented alphanumerically in this study (e.g., ST-CN1).
The clonal complexes were analyzed using the eBURST V3 software (http://eburst.mlst.net), and clusters of related STs that descended from a common ancestor were defined as clonal groups (CGs). Single genotypes that corresponded to no CG were defined as singletons, and the BioNumerics software was used to create a minimum spanning tree. In the minimum spanning tree, the founder ST was defined as the ST with the greatest number of single-locus variants. Types are represented by circles, and the size of a circle indicates the number of strains of this particular type. Heavy solid lines connect two types that differ within a single locus; light solid lines connect double-locus variants; heavy dotted lines connect triple-locus variants; light dotted lines connect quadruple-locus variants; and gray circles represent STs that are not part of any clonal complex.

Intracellular growth assay
The isolated Legionella strains were grown in BYE broth until they reached early stationary phase. Approximately 2610 9 bacteria were pelleted, resuspended and diluted (1:1,000) in RPMI 1640 tissue culture medium. The bacteria were then added to J774 cells (2610 5 per well) in 24-well dishes to give a multiplicity of infection (MOI) of approximately 10. The infected cells were incubated at 37 uC under 5% CO 2 -air for 1.5 h and washed three times with PBS to remove extracellular bacteria. To measure bacterial internalization, 1 ml of sterile, distilled water was added to the wells to release intracellular bacteria from the host cells, and the CFUs were determined by plating dilutions on BCYE agar plates. To each of the wells, 0.5 ml of fresh tissue culture medium was added, and the intracellular and extracellular bacteria in each well were combined at 24-h intervals. The total number of CFUs was determined by plating the dilutions onto BCYE agar plates.
The J774 cell monolayers were prepared on cover slips by the same procedures as described above. The cells were infected with L. pneumophila philadelphia-1 and the other isolated strains. After 48 h of infection, the infected J774 cells were stained by Gimenez staining and observed under a light microscope.

Statistical analysis
Data were analyzed using SPSS for the Windows Release 11.0.1 software (SPSS Inc, Chicago, IL, USA). The chi square test or Fisher exact test were used for analyzing qualitative data, while the t test or Mann-Whitney test were used to analyze quantitative data. The odds ratios (OR) and 95% confidence interval (CI) were calculated to assess categorical risk variables associated with Legionella-positive test results. The bacterial concentrations of L. pneumophila in intracellular growth assay were analyzed by the t test. The results were considered statistically significant when the P value was ,0.05.

Results
The degree of pollution of Legionella in spring water samples All 308 spring water samples were tested, and Legionella was cultured from 160 of those samples. Viable counts for the positive samples ranged from 1 CFU/liter to 2,218 CFU/liter with a median of 97 CFU/liter. Overall, the data presented a positive rate of Legionella of 51.9% (160/308) in the spring water samples of this study. Of the samples, 37.5% were isolated over 100 CFU/ liter, and 7.5% exceeded 1,000 CFU/liter.
All of the 15 studied spa pools were positive for Legionella in at least one sample during the one-year study period. In January, all 15 spa pools were positive for Legionella, and in June, only six spa pools were positive for Legionella. The monthly average Legionella viable contents for each pool ranged from 0.5 CFU/liter to 569 CFU/liter, and two pools exhibited a monthly average concentration of Legionella that was higher than 100 CFU/liter.

Site and seasonal variations in the level of Legionella
The isolation rates and concentrations of Legionella were different between the three hot spring recreational areas. The isolation rates were 85.5%, 45.5% and 43.6% in areas A, B and C, respectively. The average bacterial counts of Legionella in each month at area A ranged from 0.3 to 829 CFU/liter, and at areas B and C, the average counts were 0-52 CFU/liter and 0-104 CFU/ liter, respectively. The isolation rate and concentration of Legionella from area A were significantly higher than those of areas B and C (P,0.01).
The isolation rates and concentrations of Legionella were higher during the spring (73.3%, 132 CFU/liter) and winter (62.2%, 134 CFU/liter) than in summer (50.0%, 49 CFU/liter) or autumn (48.9%, 58 CFU/liter) (Figure 1). Figure 1 also shows a positive association between Legionella concentrations and passenger flow. In spring and winter, the numbers of passengers visiting hot spring recreational areas were higher than in summer and autumn.

Correlation of the occurrence of Legionella with water quality parameters
We have measured the temperature, pH, free chlorine concentration and urea concentration of every spring water sample that was obtained. The average temperature and pH ranged from 31 uC to 49 uC and 7.80 to 8.04, respectively. No significant association was found between the Legionella positive rate and sample temperature $40 uC (OR = 1.820, p = 0.06) or pH$7.9 (OR = 0.685, p = 0.286) ( Table 1). However, the Legionella concentration was significantly associated with the sample temperature (correction coefficient = 0.226, p,0.01) and pH (correction coefficient = 20.349,p,0.01). Legionella colonization was positively associated with a free chlorine concentration of $0.2 mg/L or urea concentration of $0.05 mg/L, but no significant association was found between the Legionella concentration and free chlorine or urea concentration (p = 0.135 and p = 0.078, respectively).
The microbiological parameters of the water samples that were examined included the total microbial counts and total coliform. Sixty percent of the tested water samples showed total microbial counts $400 CFU/ml and total coliform $3 MPN/L. A significant association was found between the positive rate of Legionella and total microbial counts $4001 CFU/ml or total coliform $3 MPN/L. The Legionella concentration was also found to be associated with the total microbial counts (correction coefficient = 0.185, p,0.01) and total coliform (correction coefficient = 0.298, p,0.01).

Distribution of serogroups of Legionella isolates
We selected one to five isolates from each water sample to determine their species and serogroups. In total, 268 isolates of Legionella from 160 samples were examined. To our surprise, of the 160 Legionella-positive samples, two or more Legionella species or serogroups were identified from 70 samples (43.8%). In nine samples, strains belonging to four or five different L. pneumophila serogroups were detected, and these cases were observed in all three hot spring recreational areas.
L. pneumophila serogroup 3 was found most frequently and was isolated from 67 samples of 14 pools throughout the one-year study period. This group was followed in predominance by L. pneumophila serogroups 6, 5, 1 and 2 ( Table 2). These serogroups were isolated for more than eight months; however, no L. pneumophila serogroup 2 or 6 strains were isolated from area C.  Other L. pneumophila serogroups and non-L. pneumophila strains were isolated less frequently. For example, L. pneumophila serogroup 8 was only isolated from area C in January; L. pneumophila serogroup 12 was only isolated from area A in April; L. pneumophila serogroup 9 was isolated from area B in March and October; two L. pneumophila serogroup 10 strains were isolated from area B in December and from area B in October; and L. micdadei and L.bozemanii were only isolated from area B in June and August, respectively.

PFGE and SBT analysis of Legionella isolates
Total 228 isolates were selected to study by PFGE, and 62 different patterns were obtained (Figure 3). The pattern LPA16.CN0042 was the most frequently occurring pattern and contained 31 strains that were isolated from three hot spring recreational areas over six months. Thirty-five patterns contained more than one isolate, and among these, 15, 21 and 32 patterns contained isolates from different hot spring recreational areas, serogroups and months, respectively. These results indicated a high genetic polymorphism among these tested isolates.
To further show the high genetic polymorphism of strains that were isolated from the spa water samples, we selected another 55 isolates from one sample from area A that was collected in April. These 55 isolates were divided into 16 different patterns with one to sixteen in each pattern ( Figure 4). The pattern LPA16.CN0144 was dominant and contained sixteen (29.1%) isolates. Among these sixteen patterns, nine occurred during other months in area A, and five occurred in other areas. However, the second-most dominant pattern LPA16.CN0141 only occurred in area A in April.
For SBT, 57 L. pneumophila isolates were selected based on their PFGE patterns and serogroups and isolated, and these isolates were differentiated by SBT into 35 different sequence types (STs). Ten profiles (ST59, ST81, ST87, ST246, ST354, ST377, ST595, ST961, ST1101 and ST1279) could be found in the EWGLI SBT database, but the profile of ST-CN1 to ST-CN25 could not be found in the database.
According to the results of our eBURST analysis, 35 STs belonged to 5 CGs and 4 singleton STs (Table 3, Figure 5). Of the 5 CGs, CG1, which was the prevalent clonal group, included 21 isolates that belonged to ST59, ST1101,

Intracellular growth ability
The 57 isolates that were analyzed by SBT were also tested for their intracellular growth ability. For all of the isolates, intracellular bacteria could be observed under a light microscope by Gimenez staining on day 2 of J774 cell infection, which suggested that all isolates have intracellular growth ability in J774 cells ( Figure 6). To further evaluate the intracellular growth ability of the isolates, plate counting was performed. All isolates, except isolate FS_4_1103abu, had bacterial concentrations of 10 4 -10 5 CFU/ml, 10 5 -10 6 CFU/ml, 10 6 -10 7 CFU/ml and 10 7 -10 8 CFU/ml on day 0, 1, 2 and 3 of infection, respectively, and showed no difference in bacterial concentrations of L. pneumophila Table 2. The number of spa water samples, spa pools and months contaminated by different Legionella species and serogroups.   Legionella in Hot Spring PLOS ONE | www.plosone.org philadelphia-1 by the t test (p.0.05). FS_4_1103abu exhibited bacterial concentrations of 6.8610 5 CFU/ml and 5.6610 6 CFU/ ml on day 2 and 3 of infection, respectively, and was significantly different from that of L. pneumophila philadelphia-1 ((p,0.05). However, FS_4_1103abu still showed high proliferation ability in J774 macrophage cells (Figure 7).

Discussion
This study illustrates a high degree of pollution of Legionella in hot spring water samples in Beijing, China. Legionella were detected in 51.9% of collected hot spring water samples from three hot spring recreation areas during the one-year study period. The isolation rate in this study was higher than what was reported by previous studies in Japan (28.7%) [32] and Taiwan (27.8%) [33]. The positive rate showed the severity of contamination of hot spring water in Beijing. If culture methods were utilized that might underestimate the presence of Legionella species, the actual level of contamination could be greater. The concentration of Legionella was low in this study, with only 7.5% of the samples giving a Legionella concentration that exceded 1000 CFU/liter, which was lower than what was found in Japan [32]. In the study carried out in Japan, nearly half of the samples exhibited Legionella concentrations that exceeded 1000 CFU/liter. The above findings clarified that although the number of the bacteria was low, Legionella widely and long-standingly inhabits hot spring water in Beijing.
In this study, the Legionella pollution and concentrations at three different hot spring recreational areas differed. These results were consistent with those obtained in other studies [32,33]. The Legionella pollution and concentration were also different between seasons in this study: the levels were higher in spring and winter but lower in summer and autumn. In another study that looked at cooling towers in China, the highest isolation rate and concentration was observed in autumn and showed a positive association between Legionella presence in cooling towers and increasing air temperatures [15]. However, in this study, the association between the presence of Legionella in spring water samples and increasing air temperatures was negative. Because the air temperature could significantly impact the water temperature of cooling towers but not that of spring water, we speculate that no association between the presence of Legionella in spring water samples and air temperatures will occur. Another factor, passenger flow, may explain the differences in the presence of Legionella between seasons. Spring and winter were the peak periods for Chinese to visit hot spring recreational areas, and a positive association between the presence of Legionella and the increasing passenger flow was shown in this study. Based on our results, we suggest that more disinfection measures are needed in spring and winter than in summer and autumn.
The growth and survival of all organisms is constrained by specific ranges of physicochemical parameters. The temperature and pH of water samples have been considered major impact factors of contamination by Legionella [34]. However, in this study, no significant association was found between the Legionella positive rate and sample temperature $40 uC or pH$7.9. We believe that this is due to the small interval of temperature (31 uC to 49 uC) and pH (7.80 to 8.04) of the spring water samples that were collected in this study. Legionella are able to reproduce at temperatures between 25 and 50 uC and survive in temperatures of up to 55-60 uC; therefore, they thrive easily in naturally and artificially hot water systems [35]. It is known that Legionella easily reproduce in high  temperature water systems, and an inverse association between water temperature and the concentration of Legionella was found [36,37,38]. In a previous study, Legionella species were found in water with temperatures ranging from 23.7 uC to 60.3 uC, and a significant association between the Legionella positive rate and sample temperature of$55 uC was found [34]. However, the highest sample temperature (49 uC) in this study was lower than 55 uC, and no significant difference between the Legionella positive rate that was found between 31-40 uC and 41-49 uC were found in the present study. In Furuhata's study, the isolation rates were significantly different when the pH was lower or higher than 3. The pH of all spring water samples in this study was near 8.0, which was an optimal pH for Legionella in an aquatic environment [20,33,37]. The free chlorine concentration is another important impact factor for contamination of spring water by Legionella. Mouchtouri et al. found that Legionella colonization was positively associated with a free chlorine concentration of $0.2 mg/L, which is in agreement with the results of the present study [34]. The microbiological parameters of spring water that were examined here included the total microbial counts and total coliform, and contamination by Legionella was associated with total microbial counts and total coliform levels in samples that were collected from the same water distribution system. This association may be explained by the ecological water environment being suitable for colonization and reproduction of Legionella and it being beneficial for other microorganisms. Therefore, the total microbial published) in Beijing; however, no direct epidemiological relationship was found between this case and that of the hot springs. Invasion and intracellular replication of Legionella within protozoa in the environment play major roles in the transmission of Legionnaires' disease, and important relationships between the intracellular growth ability of Legionella within protozoa and macrophages and human legionellosis have been seen [44,45]. Theoretically, Legionella isolates with invasive and intracellular growth ability can infect the human body. In this study, all isolates that were studied using the intracellular growth assay could enter a macrophage and had high intracellular growth ability, which suggests that these hot spring isolates have the potential to cause human legionellosis.
In conclusion, our results demonstrated high prevalence and genetic polymorphism of Legionella in springs in Beijing, China. Furthermore, the SBT and intracellular growth assay results suggested that the Legionella isolates of hot spring environments were pathogenic. Because of the potential for spring water systems to harbor, amplify and disseminate Legionella, improved control and prevention strategies are urgently needed.