SEARCH

SEARCH BY CITATION

Keywords:

  • amoebal coculture;
  • Legionella ;
  • Legionella pneumophila ;
  • natural soil;
  • rainwater;
  • source investigation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References

Aims

For the majority of sporadic Legionnaires’ disease cases the source of infection remains unknown. Infection may possible result from exposure to Legionella bacteria in sources that are not yet considered in outbreak investigations. Therefore, potential sources of pathogenic Legionella bacteria—natural soil and rainwater puddles on roads—were studied in 2012.

Methods and Results

Legionella bacteria were detected in 30% (6/20) of soils and 3·9% (3/77) of rainwater puddles by amoebal coculture. Legionella pneumophila was isolated from two out of six Legionella positive soil samples and two out of three Legionella positive rainwater samples. Several other species were found including the pathogenic Leg. gormanii and Leg. longbeachae. Sequence types (ST) could be assigned to two Leg. pneumophila strains isolated from soil, ST710 and ST477, and one strain isolated from rainwater, ST1064. These sequence types were previously associated with Legionnaires’ disease patients.

Conclusions

Rainwater and soil may be alternative sources for Legionella.

Significance and Impact of the Study

The detection of clinically relevant strains indicates that rainwater and soil are potential sources of Legionella bacteria and future research should assess the public health implication of the presence of Leg. pneumophila in rainwater puddles and natural soil.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References

Legionnaires’ disease (LD) can be contracted after inhalation of aerosols containing Legionella (Fields et al. 2002). These bacteria are ubiquitous in the environment, in freshwater and in soil (Fields et al. 2002; Amemura-Maekawa et al. 2012; Travis et al. 2012). Legionella bacteria have been shown to multiply intracellularly as parasites of free-living protozoa (Fields 1996; Lau and Ashbolt 2009). There are currently more than 50 known species of Legionella (Diederen 2008). Legionella pneumophila serogroup 1 is the cause of the majority of LD cases in Europe (Doleans et al. 2004; Harrison et al. 2007) and the USA (Benin et al. 2002). Legionella pneumophila has been shown to replicate in water at temperatures between 25 and 42°C, with an optimal growth temperature of around 35–37°C (Yee and Wadowsky 1982; Wadowsky et al. 1985). Known sources of Leg. pneumophila, often exhibiting these favourable growth temperatures, are, for example, cooling towers (van den Hoek et al. 2006; Nguyen et al. 2006), fountains (O'Loughlin et al. 2007; Palmore et al. 2009) and whirlpools (Den Boer et al. 2002; Coetzee et al. 2012). In Australia and New Zealand, Leg. longbeachae is an important cause of LD (Yu et al. 2002). Potting soil has been described as a source of Leg. longbeachae (Steele et al. 1990; Den Boer et al. 2007; Cramp et al. 2010; Pravinkumar et al. 2010).

Infection with Legionella occurs both sporadically and in outbreaks, with most cases of LD being sporadic (Fields et al. 2002). For the majority of sporadic cases of LD, the source of infection remains unknown (Den Boer et al. 2008). One reason for this could be that in source investigations, pathogenic Legionella bacteria go undetected in potential sources because of the failure of the methods used for the detection of Legionella in environmental samples. The standard method of detection of Legionella in environmental matrices is culture on Buffered Charcoal Yeast Extract (BCYE) medium plates (Edelstein 1981). These plates are easily overgrown by other bacteria in the sample, hence resulting in false negative results for Legionella bacteria. Furthermore, it is known that Legionella bacteria can enter a viable-but-not-culturable (VBNC) state and might therefore not be detected by culture methods using BCYE plates (Steinert et al. 2002). A second explanation could be that infection has occurred through exposure to Legionella bacteria from sources that are not considered in source investigations. Viable Leg. pneumophila bacteria were recently detected in rainwater on asphalt roads (Sakamoto et al. 2009; Kanatani et al. 2013) and rainwater from pluvial floods (Schalk et al. 2012). Rainwater may possible be an alternative source of infection, not yet targeted in routine source investigations. Several studies have reported an association between rainfall and LD. Hicks et al. (2007) reported an association between increase in rainfall and increase in LD incidence. Fisman et al. (2005) identified rainfall and increased humidity as predictors for the occurrence of LD. Karagiannis et al. (2009) established that in the Netherlands, for the warmest period of the year (April–September), precipitation intensity, relative humidity and temperature were positively associated with LD incidence. Although the underlying mechanism that can explain this association has not yet been clarified, the results from these studies indicate that the natural environment may play an important role in exposure of humans to Legionella. Furthermore, natural soil has been described as a reservoir of Legpneumophila in Thailand and Japan (Kuroki et al. 2007; Amemura-Maekawa et al. 2012; Kanatani et al. 2013). Although potting soil is a known source of Leg. longbeachae (Steele et al. 1990; Den Boer et al. 2007; Cramp et al. 2010; Pravinkumar et al. 2010), it is yet unclear whether exposure to Legionella species in natural soil can lead to disease.

The aim of our study was to establish whether rainwater puddles on the road and natural soil near roads in the Netherlands could serve as reservoirs of Legionella and to investigate whether these potential sources harbour viable, pathogenic Legionella bacteria. We investigated the prevalence of Legionella spp. and Leg. pneumophila in rainwater puddles on roads in the spring and late summer of 2012. Based on more favourable growth conditions due to higher temperatures in summer, a higher prevalence was expected. Characteristics of the puddles and weather characteristics were also assessed in order to assess their relationship with Legionella prevalence. In late summer, we also investigated the prevalence of Legionella spp. and Leg. pneumophila in soil near roads. An amoebal coculture procedure was used for the detection of Legionella in the samples (Schalk et al. 2012). Amoebal coculture has been successfully applied in previous studies to isolate Legionella bacteria that could not be recovered by plating on BCYE plates (Rowbotham 1983; Fallon and Rowbotham 1990; La Scola et al. 2001; Schalk et al. 2012).

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References

Sampling

During two periods in 2012, water samples were taken from rainwater puddles on roads. In April and May, samples were collected at 43 different locations in the area around Utrecht, the Netherlands. From July to October, samples were taken at 40 locations that had also been sampled in the first period. GPS coordinates were used to document the locations of the rainwater puddles. An amount of approx. 125 ml water was scooped from puddles with sterile tubes or aspirated from the road surface with sterile syringes and then stored in sterile bottles. During the second period, soil samples were taken near roads at 20 locations where rainwater puddles had also been sampled. Within an area of 1 m2, four subsamples from the upper 2 cm of soil were taken, mixed and stored in a sterile bottle. Water and soil samples were cooled during transportation and stored at 4°C until analysis. During sampling the following parameters were measured: ambient temperature, soil temperature, temperature of the rainwater in the puddle, size and depth of the puddle. In addition, pH, turbidity and electrical conductance of the sampled rainwater were measured.

Pre-treatment of the samples

Rainwater samples were concentrated by filtration of 100 ml of the sample through a 0·2 μm polycarbonate membrane filter (Merck Millipore, Darmstadt, Germany). After filtration, the membrane was placed in a sterile container with a layer of glass beads and 5 ml of the original sample. The container was placed in an ultrasound tank for 5 min in order to recover the organisms from the membrane. Not all samples could be filtrated due to the particulate content of the water. These samples were directly subjected to analysis. Of the soil samples, 5 g were resuspended in 5 ml of sterile distilled water. These suspensions were vortexed for approx. 10 s and allowed to stand at room temperature for 1 h. Just before amoebal coculture (as described below), these soil suspensions were vortexed again.

Amoebal coculture method

The amoebal coculture method for detection of Legionella was performed as previously described by Schalk et al. (2012): Acanthamoeba castellanii ATCC #30234 (American Type Culture Collection, Rockville, MD) were grown in 75 cm2 culture flasks (Corning Inc., New York, NY) with 15 ml of peptone-yeast extract-glucose (PYG) broth (2% proteose peptone, 0·1% yeast extract, 0·1 mol l−1 D-glucose, 4 mmol l−1 MgSO4, 0·4 mmol l−1 CaCl2, 0·1% sodium citrate dehydrate, 0·05 mmol l−1 Fe(NH4)2(SO4)2.6H2O, 2·5 mmol l−1 NaH2PO3, 2·5 mmol l−1 K2HPO3) at 25°C. On the day of infection, the PYG broth was removed and the amoebae were resuspended in Page's amoebal saline (PAS) (Rowbotham 1983). The amoebae were washed three times with PAS followed by centrifugation at 850 g for 10 min and subsequent resuspension of the pellet in 15 ml PAS. After the last resuspension in PAS, the number of cells was counted in a Bürker-Türk counting chamber. Cells were seeded in a 12-well microplate (Corning) at a density of 5 × 105 cells ml−1 of PAS. Per well, 1 ml of PAS with amoeba was added and this was inoculated with 100 μl of sample. Each sample was tested in triplicate (three wells). The amoebal plates were subsequently incubated at 32°C. As a negative control, one well with amoebae was not inoculated. After 3 days of infection, the amoebae were resuspended by pipetting the contents of each well up and down and 100 μl of each suspension was subcultured on a new plate with freshly seeded amoebae, prepared as described above. The plates were incubated for another 3 days at 32°C. Subsequently, the content of each well was resuspended by pipetting up and down and was 10-fold serially diluted in PAS. Of the 103, 104, 105, 106 and 107-fold dilutions, 100 μl was plated on BCYE plates (Oxoid Ltd., Hampshire, UK). Plates were incubated at 32°C. After 4 and 7 days of incubation, the BCYE plates were inspected for Legionella-like colonies, with a stereo microscope (Olympus, magnification 40×). Since 300 μl was analysed for each sample, the theoretical detection limit for unfiltrated water samples was 3·3*103 CFU l−1. The theoretical detection limit is the calculated detection limit, assuming that one bacterium in 300 μl can be detected. The filtrated water samples were 20 times concentrated, resulting in a theoretical detection limit of 1·7*102 CFU l−1. For the soil samples, 5 g was resuspended in 5 ml of sterile distilled water. Of this 5 ml, 300 μl was inoculated with amoeba and the theoretical detection limit was 3·3 CFU g−1.

Confirmation and typing of Legionella colonies

Suspected Legionella colonies were tested for their inability to grow on BCYE medium without cysteine (Oxoid Ltd., Hampshire, UK). Strains unable to grow on media without cysteine were further identified by sequence analysis of part of the mip gene (Schalk et al. 2012). To this end, colonies were frozen and subsequently diluted 1 : 10 in distilled water and heated in a heating block for 5 min at 95°C. A PCR was performed on 5 μl of this lysate with the degenerated primers Mip-FP1 (5′-GAASARCAAATGAAAGAYGTTC-3′) and Mip-RP8 (5′-CCAGGRATAACTTGYGAWAC-3′). PCR was performed in a total volume of 50 μl PCR mixture consisting of 1 × PCR buffer II (Roche, Almere, the Netherlands), 2·5 mmol l−1 MgCl2 (Roche), 0·2 mmol l−1 dNTPs (Roche), 0·8 μmol l−1 of each primer and 1·5 U Taq polymerase. The PCR programme was as follows: 5′ 95°C, 35 cycli of 30″ 95°C, 30″ 50°C, 30″ 72°C, followed by 10′ 72°C. The size of the PCR product (320 base pairs) was checked by DNA gel electrophoresis on a 2% (w/v) agarose gel. PCR products were subsequently treated with ExoSAP (GE Healthcare, Diegem, Belgium). These PCR products were sequenced with a BigDye Terminator Cycle Sequencing Ready Reaction kit (Perkin-Elmer, Applied Biosystems, Foster City, CA) with the Mip-FP1 and Mip-RP8 primers as forward and reverse primers. Sequences were analysed with BioNumerics software, ver. 6.6 (Applied Maths, Kortrijk, Belgium) and compared to Legionella sequences present in the GenBank database to subsequently type the isolated Legionella strains. Phylogenetic trees were constructed using the 320 nt sequence of the mip gene derived from the isolated strains and available sequences in the GenBank database, using the neighbour-joining (NJ) method, as implemented in the BioNumerics software. The accession numbers of the reference strains used in the phylogenetic tree are shown in Table 2. Legionella pneumophila strains were further genotyped by the standard sequence-based typing (SBT) method of the European Working Group for Legionella Infections (EWGLI) using seven genes (flaA, pilE, asd, mip, mompS, proA and neuA) (Gaia et al. 2005; Ratzow et al. 2007; Farhat et al. 2011). The Leg. pneumophila strains were genotyped by our laboratory, as well as by the National Legionella Reference Laboratory. Legionella isolates were identified by polyclonal antisera, serogroup 1 and serogroups 2-14, coupled to latex beads (Legionella latex test, Oxoid Limited, Hampshire, England). The identification of the separate serogroups of Leg. pneumophila was performed with monovalent antisera prepared by hyper-immunising rabbits with reference strains (Denka Seiken Co. Ltd., Tokyo, Japan).

Statistical analyses

The unpaired t-test was performed to evaluate statistically significant differences between the mean values of the following parameters in period 1 and period 2: temperature, pH, turbidity and electrical conductance of the puddles (Excel 2010, Microsoft, Schiphol, the Netherlands). Statistical significance was set at P ≤ 0·05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References

Detection of Legionella in rainwater from puddles on the road

In this study, rainwater puddles on the road were sampled for Legionella during two periods. In the first sampling period, from 19 April 2012 to 23 May 2012, 46 puddles were sampled at 43 locations on five sampling days. Of the 46 samples, nine were excluded from the results as cross-contamination occurred during analysis. Of 37 rainwater samples, 18 were filtrated. Filtration of the remaining samples was not possible due to particulate content. Legionella was detected in 1 out of 37 (2·7%) rainwater samples by the amoebal coculture method (Table 1). The isolated strains were typed as Leg. gormanii (Table 2, sample A).

Table 1. Characteristics of the rainwater puddles and soil, sampled in periods 1 and 2
Period122
MatrixRainwaterRainwaterSoil
Total number of samples374020
Number of Legionella spp-positive samples (Leg. pneumophila-positive samples)1 (0)2 (2)6 (2)
Puddle temperature (°C) Mean (Range)14·9 (8·3–24·4)17·7 (12·3–24·4)
Soil temperature (°C) Mean (Range)17·7 (12·5–24·6)
pH Mean (Range)7·32 (6·29–8·02)7·25 (6·69–8·75)
Turbidity (FTU) Mean (Range)228·1 (25·85–1468)201·6 (9·98–1000)
Electrical conductance (μS) Mean (Range)143·5 (22·4–589)105·6 (18·8–370)
Table 2. Serotyping and genotyping results of Legionella strains isolated from rainwater and soil samples
Rainwater sampleSoil sampleDate of samplingLegionella spp.a(# colonies analysed)Serotype (# colonies analysed)ST or allelic profileb
  1. ST, sequence type; 7–14, one of the serogroups 7–14, exact serogroup not further determined; –, not applicable; nt, not further typed.

  2. a

    Legionella reference strains: Leg. gormanii [GenBank: AF047748], Leg. pneumophila [GenBank: EU047264, EU047246], Leg. wadsworthii [GenBank: GQ265831], Leg. gratiana [GenBank: U92206], Leg. feeleii [GenBank: AF022340], Leg. longbeachae [GenBank: X83036]. Leg. spp strains showed 87% or less similarity compared to mip sequences of known Legionella species present in the GenBank database. For the remaining strains, more than 92% similarity was observed with the reference strains, based on the 320 nt sequence of the mip gene.

  3. b

    If no ST could be assigned the allelic profile is provided, ‘X’ meaning that the gene target failed to amplify.

A 11 MayLeg. gormanii (2)
B 6 AugustLeg. pneumophila (21)7–14 (6)1064
C 26 AugustLeg. pneumophila (2)7–14 (2)5, 1, 22, 30, 6, 10, X
 D23 AugustLeg. wadsworhtii (3)
 D23 AugustLeg. gratiana (17)
 D23 AugustLeg. spp. (2)
 E18 SeptemberLeg. gormanii (43)
 F3 OctoberLeg. gormanii (1)
 G3 OctoberLeg. spp. (1)
 H4 OctoberLeg. pneumophila (12)1 (4)6, 10, 23, 21, 33, 14, 15
 I4 OctoberLeg. feeleii (13)
 I4 OctoberLeg. longbeachae (4)
 I4 OctoberLeg. pneumophila (16)1 (2)12, 10, 2, 5, 3, 17, 15
 I4 October 1 (2)12, 9, 2, 21, 3, 17, 39
 I4 October 1 (1)477
 I4 October 1 (1)nt
 I4 October 3 (1)710
 I4 October 3 (4)nt
 I4 October 13 (1)nt
 I4 October 7–14 (1)2, 10, 5, 10, 18, 14, 207
 I4 October 7–14 (1)2, 10, 14, 47, 18, 14, 207
 I4 October 7–14 (1)nt

In the second sampling period, which lasted from 30 July 2012 until 4 October 2012, 40 puddles were sampled on eight sampling days. These samples were taken at 40 of the 43 locations that were sampled in the first period. Road construction work at three of the 43 initially sampled locations prevented resampling at these sites. Of the 40 rainwater samples, 18 were filtrated. The remaining samples could not be filtrated due to particulate content. Legionella was detected in 2 of 40 (5·0%) rainwater samples by amoebal coculture method (samples B and C). The strains from both samples were typed as Leg. pneumophila non-serogroup 1 (non-SG1). Three isolates originating from sample B could be further typed by SBT as sequence type (ST) 1064 (Table 2). The non-SG1 isolates from rainwater sample C did not yield a seven-allele profile, as the neuA gene could not be amplified (Table 2).

Characteristics of sampled puddles and weather characteristics

Table 1 shows the characteristics of the puddles in the two periods, such as temperature, pH and turbidity, as well as the temperature of the soil near the puddles. The water temperature was significantly higher in period 2 than in period 1 (P < 0·0024). There was no significant difference in pH, turbidity or electrical conductance between the rainwaters sampled in period 1 and period 2. Table 3 shows the ambient temperature on the sampling days as well as the mean temperature in the 14 days preceding the sampling days and the total precipitation in the 14 days before sampling. Of the 37 puddles in period 1 and 40 puddles in period 2, 64·9% (24/37) and 62·5% (25/40), respectively, were completely located on the road, and 35·1% (13/37) and 37·5% (15/40), respectively, were partly on the road and partly on the side of the road. The roads were constructed of asphalt (period 1: 24/37, period 2: 24/40), pavement (period 1: 11/37, period 2: 12/40), or a combination of asphalt and pavement (period 1: 2/37, period 2: 4/40).

Table 3. Ambient temperature and precipitation at and before sampling days of rainwater puddles
Sampling dateNo. of rainwater samples (positive/total)No. of soil samples (positive/total)Mean T (°C) sampling dayMean T (°C) preceding daysTotal precipitation (mm) preceding days
  1. The number of Legionella-positive samples per total rainwater or soil samples that were collected is indicated for each sampling day in columns 1 and 2 respectively. The mean ambient temperature (°C) on the sampling day, the mean T (°C) in the 14 days preceding the sampling day and the total precipitation (mm) in the 14 days preceding the sampling day are indicated. The bold values indicate the samplings that contain at least one positive sample.

19 April0/7 8·96·924·9
22 April0/10 8·87·529·1
9 May0/8 15·711·728·7
11 May 1/7  13·812·642·8
23 May0/5 20·313·534·9
30 July0/5 15·417·745·6
6 August 1/6  17·218·748·6
8 August0/2 16·918·258·6
23 August  1/6 18·119·91·6
26 August 1/7 0/216·020·39·4
30 August0/30/415·819·633·5
18 September0/7 1/2 12·815·88·6
3 October0/9 2/3 12·812·341·4
4 October0/1 2/3 11·612·636·3

The prevalence of viable Legionella spp. in rainwater puddles on roads appeared not to markedly differ between period 1 and period 2 (overall prevalence 3/77, 3·9%). The three positive samples had not been filtrated before analysis. The Legionella prevalence was too low to draw any conclusions about the association with puddle or weather characteristics. The three positive rainwater samples were all completely located on the road, but the roads were constructed of different materials. The puddles had different sizes and depths, and they were situated at different locations.

Detection of Legionella in soil

During period 2, soils near roads were sampled at 20 locations. Legionella spp. were detected in six out of 20 (30%) soil samples by amoebal coculture (Table 1). Legionella pneumophila serogroup 1 (SG1) was detected in two of the samples. For one of these samples (sample H), the four analysed strains had an identical but unknown allelic profile (Table 2). The other sample (sample I) contained a mixture of different Leg. pneumophila SG1 strains and non-SG1 strains. Typing of the Leg. pneumophila SG1 strains yielded two unknown allelic profiles and one known profile, namely ST477. Typing of the Leg. pneumophila non-SG1 strains also yielded two unknown profiles and one known profile for a SG3 strain, namely ST710. In addition, Leg. feeleii and Leg. longbeachae were detected in this sample. In another soil sample (sample D), a mixture was found of Leg. wadsworthii, Leg. gratiana and an undefined Legionella species, for which the sequence of the mip gene did not match Legionella sequences present in the NCBI database. In two samples (samples E and F) Leg. gormanii was detected, and the sixth sample (sample G) contained an undefined Legionella species.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References

The use of amoebal coculture enables studies of complex environmental matrices that may represent yet unknown sources of Legionella. In this study, Legionella was isolated from rainwater puddles and soil using an amoebal coculture method with A. castelanii cells. For the rainwater samples, also a direct plating method on Glycine-Vancomycin-PolymyxinB-Cycloheximide (GVPC) medium plates was used. However, no Legionella bacteria could be isolated. Other bacteria from the samples probably inhibited the growth of Legionella. Heat and acid treatments of the samples hardly reduced the presence of other microorganisms on the culture plates (data not shown).

Two Leg. pneumophila non-SG1 strains were detected in rainwater. One of these strains, ST1064, has previously been isolated from two patients in the Netherlands in 2011 (EWGLI). The prevalence of Legionella spp. in rainwater puddles during spring and late summer was low in comparison to other studies (3·9%). Schalk et al. (2012) detected Legionella in four out of 13 (31%) pluvial floods in the Netherlands by amoebal coculture, sampled in June and August 2011. Two of 13 samples (15%) contained Leg. pneumophila. The pluvial flood samples were all taken after heavy rainfall. Legionella in rainwater puddles may possibly originate from surrounding soil, and the higher prevalence in pluvial floods could be explained by the fact that water enters pluvial floods by runoff from large surfaces of surrounding soil. In the current study, the number of positive samples was too small to be able to draw any conclusions about the effect of rainfall intensity on Legionella prevalence. Another explanation for the lower prevalence in rainwater samples could be that Legionella bacteria in shallow rainwater puddles are subject to more UV light than in pluvial floods, inactivating the bacteria (Dutka 1984). Furthermore, prevalence of Legionella could be affected by temperature. In a study by Sakamoto et al. (2009), Legionella prevalence was related to the mean ambient temperature on the sampling date. Three of 19 samples (16%) were positive at temperatures below 20°C, six of 14 samples (43%) were positive at temperatures between 20 and 25°C, and seven of 12 samples (58%) were positive at temperatures above 25°C. In our study, the mean measured ambient temperature on sampling days varied between 8·8 and 20·3°C. These relatively low temperatures could explain the low prevalence of Legionella in our study; however, similar ambient temperatures were measured in June and August 2011, when pluvial flood samples were studied and a higher Legionella prevalence was found by Schalk et al. (2012). In contrast to the study by Sakamoto et al. (2009), Kanatani et al. (2013) found the prevalence of Legionella in puddles not to be related to ambient temperature on the date of sampling. Kanatani et al. (2013) analysed puddles on roads for the presence of Legionella spp. by culture at six fixed locations in Japan. Throughout the year, samples were taken monthly and Legionella spp. were detected at temperatures ranging from −0·6°C to 32·2°C.

Sakamoto et al. (2009) analysed rainwater that was collected directly from the air but did not detect any Legionella bacteria by culture. However, Legionella DNA was detected in the samples and possibly viable-but-not-culturable Legionella bacteria are present in raindrops (Sakamoto et al. 2009). Legionella bacteria were previously isolated from rainwater collected from roofs (Simmons et al. 2008; Schets et al. 2010). However, it is not clear from these studies whether Legionella were already present in the raindrops or were present on the roof surfaces and were washed off by rain.

In soil samples, the prevalence of viable Legionella spp. and Leg. pneumophila (30% and 10% respectively) was higher than in rainwater puddles. There was also a greater diversity of Legionella strains in the soil samples. Two of the soil samples contained a mixture of Legionella spp. and from one of these samples at least six different Leg. pneumophila strains were isolated. Strain ST477 has been previously isolated from five patients in the Netherlands, Italy, Austria, the UK and Germany, according to the EWGLI SBT-database (EWGLI). Strain ST710 has been previously isolated from two patients in Canada and Germany. To our knowledge, our study is the first to detect Legionella bacteria in natural soil in Europe. Evstigneeva et al. (2009) tested 11 soil samples by amoebal coculture from the city of Marseilles, France, but did not detect Legionella. Two recent studies described the isolation of clinically relevant Leg. pneumophila strains from natural soils in Thailand and Japan. Travis et al. (2012) studied the prevalence of Legionella spp. in soil in Thailand and found 22 of 39 (56·4%) samples, taken at eight rural sites, positive for Legionella spp. Of a total of 115 isolates, 17 typed as Leg. pneumophila. For eight strains allelic profiles were identified by SBT, and one identified ST had been previously associated with community-acquired cases. In a study by Amemura-Maekawa et al. (2012), 35 Leg. pneumophila isolates from soils in Japan were typed by SBT. Eleven different STs were assigned, and nine of these STs had previously been detected in clinical isolates. Our study and the studies by Travis et al. (2012) and Amemura-Maekawa et al. (2012) provide evidence that natural soil may serve as a reservoir for Leg. pneumophila. However, it is yet unknown whether humans are exposed to Legionella from natural soil.

Potting soil has been described as a source of Leg. longbeachae causing Legionnaires’ disease (Steele et al. 1990; Den Boer et al. 2007; Pravinkumar et al. 2010). However, there is little evidence that exposure to Leg. pneumophila strains in natural soil may cause infection and possibly disease. Only Wallis and Robinson (2005) reported on a case in Australia where the same Leg. pneumophila SG1 pulsovar was found in soil and in the patient. The soil was sampled from a field where the patient had spent several days potting plants.

Whether and how people are exposed to Legionella in natural soil and rainwater is unclear. During handling of soil, i.e. digging the ground, soil particles with Legionella bacteria may possibly be aerosolised and aerosols inhaled. Wind could also be a mechanism for aerosolization of soil particles (Jones and Harrison 2004), and it may be that people acquire LD by inhalation of aerosolised soil particles from the environment. Legionella in rainwater puddles could be aerosolised by splashing or by wind.

In conclusion, soil and rainwater may be alternative sources for Legionella, especially in periods when conditions for growth and aerosolization of Legionella are favourable, such as elevated temperature and relative humidity. The public health implication of the presence of Leg. pneumophila in natural soil and rainwater puddles is yet unclear and should not be overlooked. The Legionella concentration in rainwater puddles and soil and the rate of aerosolization should be quantified and new studies should focus on the effects of environmental conditions on the viability, growth and virulence of Legionella in rainwater and diverse soil types. Risk of infection from exposure to Legionella in the natural environment can be estimated by means of quantitative microbial risk assessment (QMRA), highlighting the need for estimations of exposure levels and dose–response relations (Bouwknegt et al. 2013).

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References

This work was performed on behalf of and for the account of the Ministry of Health, Welfare and Sport (V/210321).

Conflict of Interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References

The authors declare that they have no competing interests.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  • Amemura-Maekawa, J., Kikukawa, K., Helbig, J.H., Kaneko, S., Suzuki-Hashimoto, A., Furuhata, K., Chang, B., Murai, M. et al. (2012) Distribution of monoclonal antibody subgroups and sequence-based types among Legionella pneumophila serogroup 1 isolates derived from cooling tower water, bathwater, and soil in Japan. Appl Environ Microbiol 78, 42634270.
  • Benin, A.L., Benson, R.F. and Besser, R.E. (2002) Trends in legionnaires disease, 1980–1998: declining mortality and new patterns of diagnosis. Clin Infect Dis 35, 10391046.
  • Bouwknegt, M., Schijven, J.F., Schalk, J.A. and de Roda Husman, A.M. (2013) Quantitative risk estimation for a Legionella pneumophila infection due to whirlpool use. Risk Anal 33, 12281236.
  • Coetzee, N., Duggal, H., Hawker, J., Ibbotson, S., Harrison, T.G., Phin, N., Laza-Stanca, V., Johnston, R. et al. (2012) An outbreak of Legionnaires’ disease associated with a display spa pool in retail premises, Stoke-on-Trent, United Kingdom, July 2012. Euro Surveill 17, 20271.
  • Cramp, G.J., Harte, D., Douglas, N.M., Graham, F., Schousboe, M. and Sykes, K. (2010) An outbreak of Pontiac fever due to Legionella longbeachae serogroup 2 found in potting mix in a horticultural nursery in New Zealand. Epidemiol Infect 138, 1520.
  • Den Boer, J.W., Yzerman, E.P., Schellekens, J., Lettinga, K.D., Boshuizen, H.C., Van Steenbergen, J.E., Bosman, A., Van den Hof, S. et al. (2002) A large outbreak of Legionnaires’ disease at a flower show, the Netherlands, 1999. Emerg Infect Dis 8, 3743.
  • Den Boer, J.W., Yzerman, E.P., Jansen, R., Bruin, J.P., Verhoef, L.P., Neve, G. and van der Zwaluw, K. (2007) Legionnaires’ disease and gardening. Clin Microbiol Infect 13, 8891.
  • Den Boer, J.W., Bruin, J.P., Verhoef, L.P., Van der Zwaluw, K., Jansen, R. and Yzerman, E.P. (2008) Genotypic comparison of clinical Legionella isolates and patient-related environmental isolates in The Netherlands, 2002–2006. Clin Microbiol Infect 14, 459466.
  • Diederen, B.M. (2008) Legionella spp. and Legionnaires’ disease. J Infect 56, 112.
  • Doleans, A., Aurell, H., Reyrolle, M., Lina, G., Freney, J., Vandenesch, F., Etienne, J. and Jarraud, S. (2004) Clinical and environmental distributions of Legionella strains in France are different. J Clin Microbiol 42, 458460.
  • Dutka, B.J. (1984) Sensitivity of Legionella pneumophila to sunlight in fresh and marine waters. Appl Environ Microbiol 48, 970974.
  • Edelstein, P.H. (1981) Improved semiselective medium for isolation of Legionella pneumophila from contaminated clinical and environmental specimens. J Clin Microbiol 14, 298303.
  • Evstigneeva, A., Raoult, D., Karpachevskiy, L. and La Scola, B. (2009) Amoeba co-culture of soil specimens recovered 33 different bacteria, including four new species and Streptococcus pneumoniae. Microbiology 155, 657664.
  • EWGLI Sequence-Based Typing Database for Legionella pneumophila. [http://www.hpa-bioinformatics.org.uk/legionella/legionella_sbt/php/sbt_homepage.php] (accessed 17 December 2012).
  • Fallon, R.J. and Rowbotham, T.J. (1990) Microbiological investigations into an outbreak of Pontiac fever due to Legionella micdadei associated with use of a whirlpool. J Clin Pathol 43, 479483.
  • Farhat, C., Mentasti, M., Jacobs, E., Fry, N.K. and Lück, C. (2011) The N-acylneuraminate cytidyltransferase gene, neuA, is heterogenous in Legionella pneumophila strains but can be used as a marker for epidemiological typing in the consensus sequence-based typing scheme. J Clin Microbiol 49, 40524058.
  • Fields, B.S. (1996) The molecular ecology of Legionellae. Trends Microbiol 4, 286290.
  • Fields, B.S., Benson, R.F. and Besser, R.E. (2002) Legionella and Legionnaires’ disease: 25 years of investigation. Clin Microbiol Rev 15, 506526.
  • Fisman, D.N., Lim, S., Wellenius, G.A., Johnson, C., Britz, P., Gaskins, M., Maher, J., Mittleman, M.A. et al. (2005) It's not the heat, it's the humidity: wet weather increases legionellosis risk in the greater Philadelphia metropolitan area. J Infect Dis 192, 20662073.
  • Gaia, V., Fry, N.K., Afshar, B., Luck, P.C., Meugnier, H., Etienne, J., Peduzzi, R. and Harrison, T.G. (2005) Consensus sequence-based scheme for epidemiological typing of clinical and environmental isolates of Legionella pneumophila. J Clin Microbiol 43, 20472052.
  • Harrison, T.G., Doshi, N., Fry, N.K. and Joseph, C.A. (2007) Comparison of clinical and environmental isolates of Legionella pneumophila obtained in the UK over 19 years. Clin Microbiol Infect 13, 7885.
  • Hicks, L.A., Rose, C.E. Jr, Fields, B.S., Drees, M.L., Engel, J.P., Jenkins, P.R., Rouse, B.S., Blythe, D. et al. (2007) Increased rainfall is associated with increased risk for legionellosis. Epidemiol Infect 135, 811817.
  • van den Hoek, J.A., EP, I.J. and Coutinho, R.A. (2006) Legionella outbreak in Amsterdam: a cooling tower as the source. Ned Tijdschr Geneeskd 150, 18081811.
  • Jones, A.M. and Harrison, R.M. (2004) The effects of meteorological factors on atmospheric bioaerosol concentrations–a review. Sci Total Environ 326, 151180.
  • Kanatani, J., Isobe, J., Kimata, K., Shima, T., Shimizu, M., Kura, F., Sata, T. and Watahiki, M. (2013) Close genetic relationship between Legionella pneumophila serogroup 1 isolates from sputum specimens and puddles on roads, as determined by sequence-based typing. Appl Environ Microbiol 79, 39593966.
  • Karagiannis, I., Brandsema, P. and van der Sande, M. (2009) Warm, wet weather associated with increased Legionnaires’ disease incidence in The Netherlands. Epidemiol Infect 137, 181187.
  • Kuroki, H., Miyamoto, H., Fukuda, K., Iihara, H., Kawamura, Y., Ogawa, M., Wang, Y., Ezaki, T. et al. (2007) Legionella impletisoli sp. nov. and Legionella yabuuchiae sp. nov., isolated from soils contaminated with industrial wastes in Japan. Syst Appl Microbiol 30, 273279.
  • La Scola, B., Mezi, L., Weiller, P.J. and Raoult, D. (2001) Isolation of Legionella anisa using an amoebic coculture procedure. J Clin Microbiol 39, 365366.
  • Lau, H.Y. and Ashbolt, N.J. (2009) The role of biofilms and protozoa in Legionella pathogenesis: implications for drinking water. J Appl Microbiol 107, 368378.
  • Nguyen, T.M., Ilef, D., Jarraud, S., Rouil, L., Campese, C., Che, D., Haeghebaert, S., Ganiayre, F. et al. (2006) A community-wide outbreak of Legionnaires’ disease linked to industrial cooling towers–how far can contaminated aerosols spread? J Infect Dis 193, 102111.
  • O'Loughlin, R.E., Kightlinger, L., Werpy, M.C., Brown, E., Stevens, V., Hepper, C., Keane, T., Benson, R.F. et al. (2007) Restaurant outbreak of Legionnaires’ disease associated with a decorative fountain: an environmental and case-control study. BMC Infect Dis 7, 93.
  • Palmore, T.N., Stock, F., White, M., Bordner, M., Michelin, A., Bennett, J.E., Murray, P.R. and Henderson, D.K. (2009) A cluster of cases of nosocomial Legionnaires’ disease linked to a contaminated hospital decorative water fountain. Infect Control Hosp Epidemiol 30, 764768.
  • Pravinkumar, S.J., Edwards, G., Lindsay, D., Redmond, S., Stirling, J., House, R., Kerr, J., Anderson, E. et al. (2010) A cluster of Legionnaires’ disease caused by Legionella longbeachae linked to potting compost in Scotland, 2008–2009. Euro Surveill 15, 19496.
  • Ratzow, S., Gaia, V., Helbig, J.H., Fry, N.K. and Luck, P.C. (2007) Addition of neuA, the gene encoding N-acylneuraminate cytidylyl transferase, increases the discriminatory ability of the consensus sequence-based scheme for typing Legionella pneumophila serogroup 1 strains. J Clin Microbiol 45, 19651968.
  • Rowbotham, T.J. (1983) Isolation of Legionella pneumophila from clinical specimens via amoebae, and the interaction of those and other isolates with amoebae. J Clin Pathol 36, 978986.
  • Sakamoto, R., Ohno, A., Nakahara, T., Satomura, K., Iwanaga, S., Kouyama, Y., Kura, F., Kato, N. et al. (2009) Legionella pneumophila in rainwater on roads. Emerg Infect Dis 15, 12951297.
  • Schalk, J.A.C., Docters van Leeuwen, A.E., Lodder, W.J., de Man, H., Euser, S., den Boer, J.W. and de Roda Husman, A.M. (2012) Isolation of Legionella pneumophila from pluvial floods by amoebal coculture. Appl Environ Microbiol 78, 45194521.
  • Schets, F.M., Italiaander, R., van den Berg, H.H. and de Roda Husman, A.M. (2010) Rainwater harvesting: quality assessment and utilization in The Netherlands. J Water Health 8, 224235.
  • Simmons, G., Jury, S., Thornley, C., Harte, D., Mohiuddin, J. and Taylor, M. (2008) A Legionnaires’ disease outbreak: a water blaster and roof-collected rainwater systems. Water Res 42, 14491458.
  • Steele, T.W., Lanser, J. and Sangster, N. (1990) Isolation of Legionella longbeachae serogroup 1 from potting mixes. Appl Environ Microbiol 56, 4953.
  • Steinert, M., Hentschel, U. and Hacker, J. (2002) Legionella pneumophila: an aquatic microbe goes astray. FEMS Microbiol Rev 26, 149162.
  • Travis, T.C., Brown, E.W., Peruski, L.F., Siludjai, D., Jorakate, P., Salika, P., Yang, G., Kozak, N.A. et al. (2012) Survey of Legionella species found in thai soil. Int J Microbiol 2012, 218791.
  • Wadowsky, R.M., Wolford, R., McNamara, A.M. and Yee, R.B. (1985) Effect of temperature, pH, and oxygen level on the multiplication of naturally occurring Legionella pneumophila in potable water. Appl Environ Microbiol 49, 11971205.
  • Wallis, L. and Robinson, P. (2005) Soil as a source of Legionella pneumophila serogroup 1 (Lp1). Aust N Z J Public Health 29, 518520.
  • Yee, R.B. and Wadowsky, R.M. (1982) Multiplication of Legionella pneumophila in unsterilized tap water. Appl Environ Microbiol 43, 13301334.
  • Yu, V.L., Plouffe, J.F., Pastoris, M.C., Stout, J.E., Schousboe, M., Widmer, A., Summersgill, J., File, T. et al. (2002) Distribution of Legionella species and serogroups isolated by culture in patients with sporadic community-acquired legionellosis: an international collaborative survey. J Infect Dis 186, 127128.