Survey of Legionella Species Found in Thai Soil
Members of the Gram-negative genus Legionella are typically found in freshwater environments, with the exception of L. longbeachae, which is present in composts and potting mixes. When contaminated aerosols are inhaled, legionellosis may result, typically as either the more serious pneumonia Legionnaires’ disease or the less severe flu-like illness Pontiac fever. It is presumed that all species of the genus Legionella are capable of causing disease in humans. As a followup to a prior clinical study of legionellosis in rural Thailand, indigenous soil samples were collected proximal to cases’ homes and workplaces and tested for the presence of legionellae by culture. We obtained 115 isolates from 22/39 soil samples and used sequence-based methods to identify 12 known species of Legionella represented by 87 isolates.
Legionellosis is most often attributed to inhalation of contaminated aerosols from manmade water systems or, in the case of Legionella longbeachae, to inhalation of contaminated potting mixes or composts [1–3]. From 2003 to 2004, active population-based surveillance for atypical bacterial respiratory pathogens was performed in the province of Sa Kaeo, Thailand, to establish incidence . Immunologic testing performed on sera of suspect legionellosis cases found (5%) adult cases exhibited a fourfold rise in titer to L. longbeachae by the indirect immunofluorescence assay. L. longbeachae infection followed the typical legionellosis demographic, with incidences highest among adults over 34 years of age, increasing steadily with age, and infection peaking through September and October. As rice farming is quite prevalent in this area of Thailand, the extensive exposure of farm workers to moist soil would be a plausible means for transmitting L. longbeachae . These findings, coupled with limited access to composted or processed soil, led us to investigate the presence of legionellae, specifically L. longbeachae, in indigenous soils and exposure due to agricultural practices in Sa Kaeo.
2. Materials and Methods
In 2009, thirty-nine wet soil samples were collected from eight rural sites within Sa Kaeo province where prior laboratory-confirmed cases of legionellosis were identified and were shipped to the Centers for Disease Control and Prevention in Atlanta for culture . Cultures were performed using a modification of the procedure used to culture legionellae from water . Five grams of soil were weighed, and mixed with 50 mL sterile dH2O. The filtrate was then strained through sterile gauze into a 50 mL conical tube. A 1 : 5 dilution was made of the filtrate, and 500 μL of the diluted filtrate was acid treated for fifteen minutes with an equal part of KCl/HCl acid (pH 2.3). After acid treatment, 50 μL was cultured on one plate of buffered charcoal yeast extract (BCYE) supplemented with 160 mg/L cycloheximide, two plates additionally supplemented with 100,000 U/L polymyxin B and 5 mg/L vancomycin (PCV), and two plates further supplemented with 2 g/L glycine (GPCV). Plates were examined at four, seven, and fourteen days to check for the presence of Legionella. Any samples overgrown with non-Legionella organisms were retreated with acid in fifteen-minute increments and recultured, until nonrelevant organisms were reduced enough to allow for identification of legionellae. Colonies displaying Legionella morphology were checked for cysteine auxotrophy on BCYE biplates with and without L-cysteine. Colonies requiring L-cysteine for growth were considered presumptive Legionella species.
Monoclonal antibody (MAb) testing was used to identify L. pneumophila serogroup 1 (Lp1) as previously described [7, 8]. Other L. pneumophila serogroups were determined using direct fluorescent antibody staining performed on dried, formalin-fixed suspensions using L. pneumophila serogroup-specific fluorescein isothiocyanate labeled antibody . Select non-L. pneumophila isolates were tested by slide agglutination as described .
All non-Lp1 isolates were identified by sequencing the macrophage infectivity potentiator (mip) gene . The resulting sequence was then queried against the National Center for Biotechnology Information (NCBI) GenBank nucleotide database using NCBI’s alignment tool Basic Local Alignment Search Tool (BLAST) [12, 13]. Sequences with a minimum of 95% identity to a known Legionella species were assigned. Those sequences with less than 95% identity to sequences of submitted Legionella species were considered a potential novel Legionella organism. All L. pneumophila isolates were genotyped using the sequence-based typing (SBT) epidemiological typing scheme established by the European Working Group for Legionella Infections (EWGLI) as previously described [14, 15]. eBURST analysis was performed to observe the relatedness of the soil isolates to isolates from other countries previously submitted to the EWGLI SBT database (http://eburst.mlst.net/).
3. Results and Discussion
Twenty-two (56%) of the 39 soil samples received were positive for Legionella. In total, we obtained 115 isolates, 87 of which were known species, 25 potential novel species, and 3 isolates that were not typable by mip sequencing (Table 1). Nine of the species identified have been reported in association with human disease and represented 70% (80/115) of isolates obtained . Isolates of L. birminghamensis, L. lansingensis, L. pneumophila, L. rubrilucens, and L. sainthelensi were found exclusively in soil samples from the personal residence of suspect cases. Isolates of L. bozemanae, L. erythra, L. gormanii, L. quateirensis, L. quinlivanii, and the potentially novel species identified as Legionella sp. ST24644 (NCBI accession number GU083740; isolated from a cooling tower in Thailand) were found only in soil samples from the workplace of suspect cases. No isolates of L. longbeachae were identified. The highest diversity of species () was found in soils taken from the outdoor area for washing at the personal residence and within rice fields. L. pneumophila serogroup 1 was identified in the environment; however, none were positive for monoclonal antibody MAb2, the phenotypic subtype responsible for 65–100% of Lp1-caused legionellosis .
Five of 112 isolates tested reacted strongly with L. longbeachae serogroup 1 antisera by direct fluorescent antibody testing. Sequence analysis, however, indicated these isolates were L. bozemanae (), Legionella sp. ST24644 (), and a novel nonfluorescent Legionella species (). Although the species of the three mip untypable isolates remain unknown, they were identified as Legionella spp. using a pan-Legionella real-time PCR assay . Slide agglutination testing found these isolates were not L. geestiana, a species in which the mip gene is known to not amplify with the primers used .
Eight allelic profiles were identified by SBT analysis, seven of which were novel profiles (as of November 2, 2010). Two of the seven novel sequence types identified were found to be related to isolates from community-acquired and nosocomial cases through eBURST analysis (as of December 14, 2010; data not shown). The eighth allelic profile matched the existing ST260 which has been associated with community-acquired cases.
Although a primary goal in this study was to identify the environmental source(s) of the suspect causative agent L. longbeachae in the suspect pneumonia cases, we were unable to recover this species from these 39 soil samples. Interestingly, the water-saturated soils collected did support the growth of many other Legionella species not previously associated with this indigenous soil type.
These findings suggest that the L. longbeachae seropositive-pneumonia cases documented in 2003/4 were either (i) due to L. longbeachae that were no longer or never present in these environments or were not found in our limited number of samples as mentioned previously; (ii) due to a serologically cross-reactive strain of legionellae; (iii) serological cross-reaction with another pathogen; or (iv) false-positives. Cross-reactivity is an inherent problem in the use of serology for identification, and the findings of this study further highlight the need for a molecular-based method for identification [19–22]. When possible, a clinical isolate is preferred for diagnosis of suspect non-L. pneumophila cases because of insufficient specificity and sensitivity of non-pneumophila legionellae antisera .
For future studies, we wish to conduct prospective surveillance in Sa Kaeo and obtain legionellae isolates from patient samples. A clinical isolate would allow for sequence identification of the etiologic agent and detection of any novel species responsible for respiratory disease in tropical countries such as Thailand. Concurrent environmental sampling would allow identification of settings capable of transmitting these pathogens to susceptible hosts. Although L. longbeachae is widely accepted as the predominant pathogenic, soil-dwelling species of Legionella, the presence of legionellae in soils has been limited to composts and manufactured potting mixes. The findings in this study indicate native soils are a likely reservoir of multiple Legionella species in regions with a geography and climate similar to Sa Kaeo and may play a role in human disease.
The findings and the conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
G. E. Bollin, J. F. Plouffe, M. F. Para, and B. Hackman, “Aerosols containing Legionella pneumophila generated by shower heads and hot-water faucets,” Applied and Environmental Microbiology, vol. 50, no. 5, pp. 1128–1131, 1985.View at: Google Scholar
T. J. Dondero Jr., R. C. Rendtorff, and G. F. Mallison, “An outbreak of Legionnaires' disease associated with a contaminated air-conditioning cooling tower,” The New England Journal of Medicine, vol. 302, no. 7, pp. 365–370, 1980.View at: Google Scholar
T. W. Steele, “Legionnaires' disease in South Australia, 1979–1988,” Medical Journal of Australia, vol. 151, no. 6, pp. 322–328, 1989.View at: Google Scholar
C. R. Phares, P. Wangroongsarb, S. Chantra et al., “Epidemiology of severe pneumonia caused by Legionella longbeachae, Mycoplasma pneumoniae, and Chlamydia pneumoniae: 1-year, population-based surveillance for severe pneumonia in Thailand,” Clinical Infectious Diseases, vol. 45, no. 12, pp. e147–e155, 2007.View at: Publisher Site | Google Scholar
National Statistical Office Thailand, “Population and housing census,” 2000, http://web.nso.go.th/census/poph/finalrep/sakaeofn.pdf.View at: Google Scholar
J. M. Barbaree, G. W. Gorman, W. T. Martin, B. S. Fields, and W. E. Morrill, “Protocol for sampling environmental sites for legionellae,” Applied and Environmental Microbiology, vol. 53, no. 7, pp. 1454–1458, 1987.View at: Google Scholar
J. R. Joly, R. M. McKinney, and J. O. Tobin, “Development of a standardized subgrouping scheme for Legionella pneumophila serogroup 1 using monoclonal antibodies,” Journal of Clinical Microbiology, vol. 23, no. 4, pp. 768–771, 1986.View at: Google Scholar
G. N. Sanden, P. K. Cassiday, and J. M. Barbaree, “Rapid immunodot technique for identifying Bordetella pertussis,” Journal of Clinical Microbiology, vol. 31, no. 1, pp. 170–172, 1993.View at: Google Scholar
W. B. Cherry, B. Pittman, and P. P. Harris, “Detection of legionnaires disease bacteria by direct immunofluorescent staining,” Journal of Clinical Microbiology, vol. 8, no. 3, pp. 329–338, 1978.View at: Google Scholar
W. L. Thacker, H. W. Wilkinson, and R. F. Benson, “Comparison of slide agglutination test and direct immunofluoresence assay for identification of Legionella isolates,” Journal of Clinical Microbiology, vol. 18, no. 5, pp. 1113–1118, 1983.View at: Google Scholar
R. M. Ratcliff, J. A. Lanser, P. A. Manning, and M. W. Heuzenroeder, “Sequence-based classification scheme for the genus Legionella targeting the mip gene,” Journal of Clinical Microbiology, vol. 36, no. 6, pp. 1560–1567, 1998.View at: Google Scholar
S. Ratzow, V. Gaia, J. H. Helbig, N. K. Fry, and P. C. Lück, “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,” Journal of Clinical Microbiology, vol. 45, no. 6, pp. 1965–1968, 2007.View at: Publisher Site | Google Scholar
N. A. Kozak, R. F. Benson, E. Brown et al., “Distribution of lag-1 alleles and sequence-based types among Legionella pneumophila serogroup 1 clinical and environmental isolates in the United States,” Journal of Clinical Microbiology, vol. 47, no. 8, pp. 2525–2535, 2009.View at: Publisher Site | Google Scholar
K. A. Thurman, A. K. Warner, K. C. Cowart, A. J. Benitez, and J. M. Winchell, “Detection of Mycoplasma pneumoniae, Chlamydia pneumoniae, and Legionella spp. in clinical specimens using a single-tube multiplex real-time PCR assay,” Diagnostic Microbiology and Infectious Disease, vol. 70, no. 1, pp. 1–9, 2011.View at: Publisher Site | Google Scholar
R. F. Benson, W. L. Thacker, B. B. Plikaytis, and H. W. Wilkinson, “Cross-reactions in Legionella antisera with Bordetella pertussis strains,” Journal of Clinical Microbiology, vol. 25, no. 4, pp. 594–596, 1987.View at: Google Scholar
P. H. Edelstein, R. M. McKinney, and R. D. Meyer, “Immunologic diagnosis of legionnaires' disease: cross-reactions with anaerobic and microaerophilic organisms and infections caused by them,” Journal of Infectious Diseases, vol. 141, no. 5, pp. 652–655, 1980.View at: Google Scholar
V. Jimenez-Lucho, M. Shulman, and J. Johnson, “Bordetella bronchiseptica in an AIDS patient cross-reacts with Legionella antisera,” Journal of Clinical Microbiology, vol. 32, no. 12, pp. 3095–3096, 1994.View at: Google Scholar
C. Pelaz, L. G. Albert, and C. M. Bourgon, “Cross-reactivity among Legionella species and serogroups,” Epidemiology and Infection, vol. 99, no. 3, pp. 641–646, 1987.View at: Google Scholar