Home Bacteria Legionella
Legionella PDF Print E-mail
Tuesday, 17 August 2010 00:00
Andreas Nocker
Mª Adela Yáñez
Elena Soria
Vicente Catalán

  • pathogenesis due to invasion and replication in human macrophages
  • omnipresent in water
  • transmitted by aspiration of aerosols
  • can tolerate high temperatures (up to 60°C) although growth occurs only up to app. 43-45°C
  • temperatures between 20 and 45°C, pH 5-7, high iron and carbonate content, plastic pipes, high organic material, presence of water microflora, and water stagnancy present favorable conditions for Legionella growth
  • temperatures below 20°C, low iron and carbonate content,  pH > 8, absence of water microflora, and high water exchange in pipes  present unfavorable conditions
  • the majority of Legionella in water is believed to be associated with biofilms or free-living protozoa
  • greatly enhanced survival and increased resistance to disinfection when ingested by amoebae
The genus Legionella comprises gram-negative bacteria belonging to the gamma proteobacteria. The representative species of the genus is L. pneumophila, which was the first species to be described. L. pneumophila and related Legionella bacteria cause Legionellosis, a disease that emerged in the second half of the 20th century. The severity of Legionellosis varies from mild febrile illness (Pontiac fever) to a potentially fatal form of pneumonia (Legionnaires' disease) that can principally affect anyone, but in most cases affects people with increased susceptibility (due to age, illness, immunosuppression, or other risk factors such as smoking). The number of species, subspecies and serogroups of Legionellae continuously increases with accumulating knowledge. This genus currently comprises at least 50 species with 70 distinct serogroups ( Wullings and van der Kooij 2006; Benson and Fields 1998; La Scola et al. 2004). Although around half of them have been associated with disease, L. pneumophila (mainly serogroup 1) is the most common and prominent pathogen and is the causative agent of more than 70% of Legionellosis infections, 20-30% are caused by other serogroups, and 5-10% are caused by non-pneumophila species ( Joseph C 2002). The majority of human infections with species other than L. pneumophila are pneumonic (Fang and Vickers 1989). Of the reported non-pneumophila infection, the causes of infection are: L. micdadei (60%), L. bozemanii (15%), L. dumoffii (10%), L. longbeachae (5%) and other species (10%).

The name Legionnaire's disease originates from the fact that a pneumonia outbreak caused by L. pneumophila affected members of the Pennsylvania American Legion during a meeting in 1976 ( Brooks et al. 2004). The milder pneumonia referred to as Pontiac fever, on the other hand, is named after the town Pontiac in Michigan where employees and visitors of the health department suffered an acute respiratory disease. Both diseases are caused by aspiration of contaminated aerosols as produced by shower heads, air-conditioning systems, whirlpools, fountains, or dental devices ( Brooks et al. 2004; Bollin et al. 1985; Steinert et al. 2002). It has been shown by autopsy that L. pneumophila can spread from the respiratory system to the body. Legionellae have been detected in the spleen, liver, kidney, myocardium, bone and bone marrow, joints, inguinal and intrathoracic lymph nodes and digestive tract (Lowry and Tompkins 1993). Pathogenesis is mainly due to the ability of L. pneumophila to invade and multiply within human macrophages ( Steinert et al. 2002). The ability to invade other cells also has important implications for their lifecycle in the environment: in water, the majority of the Legionella cells are believed to live intracellularly as protozoan parasites. Infected protozoa might be an important vehicle of transmission to humans when inhaled ( Harb et al. 1998). At the same time the sheltered environment provided by the protozoan envelope protects Legionella and lowers its susceptibility to disinfection.

Outbreaks and legislation. The introduction of Legionella into several man-made aquatic environments has resulted in increased presence of the pathogen in bioaerosols, which presents a significant hazard for human health ( Pascual et al. 2001). Outbreaks of L. pneumophila occur throughout the world. Due to its high frequency Legionellosis has been classified as a statutory notifiable disease in most industrialized countries. Nevertheless the priority given to Legionellosis surveillance may need to be even greater than suggested by local morbidity and mortality due to its heavy impact on industrial, touristic, and social activities.

Due to differences in public health surveillance systems the provision of Legionella data is determined by each country's public health system ( WHO 1999). A country's national surveillance of Legionnaires' disease depends on factors such as: infrastructure and public health laws, adopted surveillance principles and standard operating procedures, notification law, data protection, patient confidentiality and freedom of information legislation. The cases of Legionellosis disease can be grouped into the following categories: community acquired, domestically acquired, nosocomial (i.e. health-care acquired), travel associated.

Some examples of Legionella outbreaks occurred are listed below:

  • An outbreak caused by a whirlpool at a flower show in the Netherlands in 1999 resulted in 31 deaths and more than 200 cases of disease ( Wullings and van der Kooij 2006; Den Boer et al. 2002).
  • An explosive outbreak of Legionnaires' disease occurred in 2001 in Murcia, Spain. The high number of infections (800 suspected cases, of which 449 cases were confirmed) and a fatality rate of 1% made this incident the world's largest outbreak of the disease that has been reported to date. An epidemiologic investigation suggested the cooling towers at a city hospital to be the point-source of the outbreak ( García-Fulgueiras et al. 2003).
  • A community outbreak in England in 2003 involved 27 cases of Legionellosis and two deaths ( Gaia et al. 2003; Kirrage et al. 2007). The source of the outbreak was reported to be the cooling tower of an apple cider factory. The cooling tower was switched on once a year, when apples were delivered to the plant in order to initiate the process of cider production. The outbreak was stopped when the point source was identified and the cooling towers were shut down.

Legionellae are ubiquitous in natural and anthropogenic water environments and have been found in lakes and rivers ( Fliermans et al. 1981), drinking water distribution systems ( Henke and Seidel 1986; Lawrence et al. 1999), potable hot water systems ( Yáñez et al. 2005), cooling towers ( Castilla et al. 2007; Ragull et al. 2007), WWTPs ( Palmer et al. 1993), whirlpools (Den Boer et al. 2002), and to a somewhat lesser extent in groundwater ( Brooks et al. 2004; Riffard et al. 2001, 2005). Legionella spp. are well known to colonize surface-attached biofilms ( Thomas et al. 2004) and have recently also been identified in floating biofilms. Under certain conditions the occurrence of biofilms might be associated with the abundance of Legionellae. Costa et al. hypothesized that Legionella concentrations in groundwater were lower than those found in man-made systems or surface waters because of the smaller amount of biofilm in groundwater compared to sometimes massive biofilm development in the other environments ( Costa et al. 2005). Nevertheless, groundwater sources have been reported as a cause of disease ( Fields et al. 2002; Riffard et al. 2005). Densities above 104 to 105 CFU/liter are considered a potential threat to human health and were associated to legionellosis ( Meenhorst et al. 1985; Patterson et al. 1994). A concentration of 103 CFU/liter was referred to as a first alert value ( Joly et al. 2006). In distribution systems, the presence of dead-end loops and water stagnation in the plumbing system are considered technical risk factors ( Steinert et al. 2002). Also the age of the plumbing system and the materials used were found to influence colonization (Leoni K 2003). Although Legionella can be abundant in wastewater ( Catalán et al. 1997; Palmer et al. 1993), they do not enter water through fecal pollution ( Steinert and Heuner 2005). Fecal indicators are thus no option for Legionella detection. However, elevated Legionella concentrations in sewage-contaminated coastal waters and in reclaimed water have been reported ( Ortiz-Roque and Hazen 1987; Palmer et al. 1993; Palmer et al. 1995).

Examples of studies on occurrence in water are:

  • When studying the Legionella colonization frequency in hot and cold water distribution systems of 385 hotels in Greece, the Legionella count was >500 CFU per liter in hot water distribution systems of 80 hotels and ≥104 CFU liter in 25 hotels in at least one sample. Among the 277 isolates, L. pneumophila was the most frequently isolated species (87%). The authors reported that legionellae were not isolated at water temperatures above 60.3°C and below 23.7°C. Legionella levels did not correlate with coliform, E. coli, intestintal enterococci, or Clostridium perfringens counts, but with total plate counts. The isolation rates in Legionella-positive hotels were higher in summer (24.1%) than in the other seasons (16.2 - 17.5%)  ( Mouchtouri et al. 2007a; Mouchtouri et al. 2007b).
  • A study examining different hot water systems of randomly selected family residences in two German cities found no legionellae in houses with point-of-use instantaneous water heaters (50 samples) using plate counting. On the other hand, 12% (n=400) of houses with storage tanks and recirculating hot water tested positive with a mean count of 3,870 CFU/liter (maximum: 106 CFU/liter). Systems with district heating supply showed the highest rate of isolation with 50% prevalence. L. pneumophila accounted for 94% of all Legionella positive specimens, of which 72% belonged to serogroup 1. Newer systems (not older than 2 years) did not show signs of colonization. Temperature was seen as the most important determinant for multiplication of Legionella. Water with a constant temperature below 46°C was most frequently found colonized and provided the highest Legionella counts, whereas systems with a water temperature of 60°C and higher were Legionella-free ( Mathys et al. 2007).
  • Legionella spp. concentrations in 16 surface water (SW) prior postdisinfection and 81 groundwater (GW) drinking water supplies in the Netherlands with temperature below 15°C were tested using qPCR and semiquantitative dilution PCR. Numbers were significantly higher in SW treated with multiple barriers at 4°C than in GW treated at 9 to 12°C with aeration and filtration. The following median concentrations were reported (cells per liter; number of samples is given in brackets):

    Ozonation as a SW treatment step was reported to cause a 1-log-unit reduction. Sequencing of 16S rRNA genes revealed the presence of species with >97% similarity to the human pathogens L. bozemanii, L. lytica, L. pneumophila, L. dumoffii, L. anisa, L. drozanski, L. fallonii, and L. micdadei. Although non-L. pneumophila species were observed in all samples, L. pneumophila was reported not to be a common member of the microbial community in raw or treated GW and not in treated water at temperatures below 12°C ( Wullings and van der Kooij 2006). L. pneumophila was reported to represent only a small fraction (<1%) of the total number of Legionellae detected. The authors correlate this finding with the rare detection of L. pneumophila in GW samples in a German study ( Seidel et al. 1986).
  • Groundwater from various sites in the U.S. and Canada, which was not under the direct influence of surface water, showed Legionella spp. concentrations from 102 to 105 CFU/liter in water and up to 1.2 x 102 CFU/cm2 in biofilms ( Brooks et al. 2004). Legionella could also be readily recovered from groundwater samples (water and biofilms) in a study performed by Riffard et al. ( Riffard et al. 2001).
  • Costa et al. examined a total of 111 groundwater samples from six different boreholes in two geographical areas in Portugal over 7 years. Colonization of aquifers was not found uniform and the persistence of Legionella was independent of the plumbing system and pumps present in the borehole. In some locations, Legionella spp. were never detected, in others they persistently recovered. The majority of isolates in area 1 were identified as L. oakridgensis, in area 2 as L. pneumophila. Concentrations ranged between 50 and 2.4 x 104 CFU/liter ( Costa et al. 2005).
  • Naturally occurring floating biofilms were shown to be an ecological niche for Legionella. Legionella spp. were present in 100% and 81% of these biofilms from anthropogenic and natural aquatic systems, respectively. Concentrations were in the range of 10 to 100 cells/cm2. L. pneumophila was detected in 100 and 70% of these samples. In addition, two typical Legionella hosts, Naegleria spp. and Acanthamoeba spp., were present in high proportions in these biofilms. Acanthamoeba spp. were shown to be naturally infected with L. pneumophila using FISH probes. ( Declerck et al. 2007a). An in-vitro experiment could confirm the finding that addition of amoebae A. castellanii can result in increased levels of biofilm-associated L. pneumophila. About 90% of A. castellanii were filled with highly metabolically active L. pneumophila ( Declerck et al. 2007b).
  • A study in Belgium revealed that 56% of 46 aquatic samples (shower, industrial, natural and tap water) tested positive for Legionella spp. by cultivation and 98% by PCR. Legionella concentrations in tap water were reported to be between 50 to 4000 CFU/liter and in shower samples between 50 to 3,750 CFU/liter. One shower sample showed a concentration of >80,000 CFU/liter ( Devos et al. 2005).
  • In the U.S., L. pneumophila has been found in rivers and lakes (temperature range between 1- to 29°C) with concentrations between 104 to 107 cells per liter using FA methods ( Fliermans et al. 1981).
  • Despite is non-fecal origin, Legionella spp. were found in concentrations >103 cells/ml in wastewater as determined by PCR and direct fluorescent antibody staining. L. pneumophila concentrations were <103 cells/ml. No or only moderate (10-fold) reduction in cell numbers throughout the treatment process was reported ( Palmer et al. 1993). Also chlorinated effluent contained Legionella spp. although they were not culturable ( Palmer et al. 1995).
  • A group in Michigan isolated Legionella from five of 856 water samples from chlorinated public water supplies. Isolates were identified as L. pneumophila serogroup 1, L. dumoffii and L. jordanis. No coliform bacteria were detected in these samples ( Hsu et al. 1984).
Although disinfection seems to be successful in reduction of Legionella contamination or even short-term eradication, it has proven difficult to achieve sustainable long-term elimination from aquatic environments. Different cases of Legionella re-emergence after (hyper-)chlorination (e.g. on a cargo ship, in domestic hot water systems, or a hospital hot water distribution system) have been reported ( Caylà et al. 2001; García et al. 2007; Borella et al. 2004; Heimberger et al. 1991). Despite different treatments, the same Legionella strains persistently caused disease. The interaction of Legionella with free-living protozoa and other microorganisms might be a key factor in this respect ( García et al. 2007). Apart from obtaining nutrients and enabling them to survive under adverse environmental conditions, the association with protozoa and biofilms makes Legionella more resistant to disinfection ( Lemarchand et al. 2004). Resistance to 50 mg/L of chlorine has been reported when Legionella spp. were enclosed in cysts of amoeba ( Lemarchand et al. 2004). Significantly higher resistance against chlorine and monochloramine of L. pneumophila when associated with the protozoan Hartmannella vermiformis in a model multi-species biofilms was also reported by Donlan et al. ( Donlan et al. 2005). In addition to providing a sheltered environment to internalized Legionella, amoebae are suspected to be involved in the resuscitation of Legionella that have entered a VBNC state after disinfection ( García et al. 2007). Removal or reduction of biofilms and protozoa can therefore be considered essential when trying to reduce Legionella levels. Legionella log10 reductions were reported to be 2.07 and 2.11 (exposure to 0.5 mg/liter for 180 min) for chlorine and monochloramine in the absence of H. vermiformis, respectively, while reductions were only 0.67 and 0.81 in its presence. Monochloramine was considered a more efficient disinfectant than chlorine for Legionella eradication ( Fields et al. 2002). A reason might be their greater stability, which results in higher residual disinfectant concentrations over more extensive distribution systems ( Fields et al. 2002). In another study, chlorine dioxide was favored over chlorine for the same reason although both disinfectants were effective ( Loret et al. 2005). Disinfectants that combat biofilms more efficiently can be considered more sustainable for killing Legionella. Continuous treatment with chlorine dioxide was also identified as more efficient than other treatments for controlling L. pneumophila in a domestic water system with re-circulating loops and dead ends in the presence of biofilms in a study by Thomas et al. ( Thomas et al. 2004). A commonly used strategy in hot water systems for Legionella prevention is thermal disinfection. Maintaining the water temperatures above at least 50°C in hot water systems with occasionally increases to 65°C might be considered an effective control measure ( Blanc et al. 2005). The efficiency of heat treatment, however, also might depend on the presence of protozoa. The ingestion of Legionella by thermotolerant free-living amoebae was shown to increase heat tolerance of the intracellular pathogens ( Storey et al. 2004a). Protection from thermal disinfection might also derive to some extent from association with biofilms although the degree of protection would be expected to correlate with biofilms thickness ( Storey et al. 2004a; Storey et al. 2004b). A simplistic quantitative microbial risk assessment model presented by Storey et al. considered biofilms detachment and interaction with amoebae in water distribution systems as important ecological factors that significantly increase the risk of waterborne legionellosis ( Storey et al. 2004b).

The efficacy of various disinfection procedures has been reviewed in detail by Kim et al. ( Kim et al. 2002). The document includes thermal, UV, and chemical (metals, oxidizing and non-oxidizing agents) disinfection and concludes that thermal disinfection is effective at >60°C and among chemical disinfectants, oxidizing agents are in general more effective than non-oxidizing agents. UV irradiation by itself was considered insufficient for complete eradication of Legionella, although it can result in significant decrease of culturability within a short time.

The main drawbacks of the chemical disinfectants are that sometimes microorganism become resistant when subjected to a single disinfectant, and periodic alternation of disinfectants is often required to mitigate this effect. Moreover, depending on the disinfectant a high degree of corrosion in the facilities can occur. Non-chemical treatment methods are a desirable alternative, but any new treatment method must be carefully tested for its long-term anti-microbial efficacy potential adversely impacting scale formation, and corrosion or heat transfer efficiency of the overall system.  In the case of treatment of Legionella new alternative methods are proposed such as cavitation and photocatalysis. 

Cavitation. The cavitation is the formation, growth, and implosion of vapor bubbles in a liquid. It can be created by sound waves, lasers, or by fluctuations in fluid pressure (hydrodynamic cavitation). Cavitation has been reported to kill bacteria through chemical reactions, pressure pulses and micro-jets, and by localized high temperatures. In aqueous liquids, cavitation leads to the formation of hydroxyl radicals and hydrogen peroxide. These short-lived reactive radicals are capable of secondary oxidation and reduction reactions in the immediate vicinity of the bubble. This method has been applied to several cooling water systems ( Gaines et al. 2007) and results were very promising in terms of reducing heterotrophic plate counts and the presence of L. pneumophila. A corrosion study revealed no notable corrosion except for galvanized steel. The high variance in the heat transfer data indicates that further studies are required to fully verify a positive impact. Nevertheless, the implementation of physical disinfection systems offers a number of potential benefits, including lower operating costs, elimination of hazardous chemicals, the potential to "de-skill" the labor workforce, simplified equipment complexity and operation, reduced regulatory reporting requirement, and an enhanced corporate image by using a non-chemical "green" technology.

Photocatalysis. Heterogeneous photocatalysis using UV radiation and semiconductor oxide, which in most cases is the anatase form of TiO2, has emerged as an innovative method for water treatment. By using an aqueous oxygenated suspension of polycrystalline TiO2 and UV light it is possible to carry out the oxidation of organic species adsorbed at a solid suface. This oxidation process, which is started by photogeneration of active oxygenated radicals such as OH or O2-, can also be used to attack the cell membrane of microorganisms and to cause their inactivation ( Coronado et al. 2005). The method has been applied for the Legionella disinfection in different pilot facilities ( Cheng et al. 2007) and results indicated that the photocatalytic oxidation (PCO) has been efficacious in the inactivation of different strains of Legionella. A higher lethality of the photocatalytic method was observed towards Legionella in comparison to the other aerobic bacteria. This fact is explained by the hypothesis that the radicals attack the Legionella secretion system, which is highly active for virulence, and thus, promotes adhesion to the TiO2 surface. This attack would then be able to inactivate L. pneumophila without dispersing the outer membrane. Apart from this, the water flow through the catalyst fibres can facilitate the bacteria transport towards the TiO2 surface, and additionally the generated shear stress may help adhesion, at least for some bacteria as E. coli, contributing further to improve the photokilling efficiency; both factors may contribute to the efficiency of this photoreactor configuration ( Coronado et al. 2005).

Selected studies on disinfection are summarized in the following:

  • In a reactor fed with tap water containing 0.15 ppm chlorine, the number of culturable L. pneumophila cells in preconditioned drinking water-associated biofilms (grown under high-shear turbulent flow) and in outlet water was studied ( Lehtola et al. 2007). Culturability in outlet water decreased during the experiment, whereas culturability in biofilms did not change significantly over the 2 week study period. L. pneumophila was shown to survive substantially longer in drinking water-associated biofilms than E. coli (weeks compared to days) making the latter a bad indicator for this pathogen in biofilms and its release ( Lehtola et al. 2007).
  • A L. pneumophila strain was reported to have persisted for 15 years in the water system of a hospital in Northern Italy that had undergone disinfection procedures multiple times since 1990 and had been equipped with a continuous disinfection system. The strain caused an outbreak of Legionnaire's disease in 2004 ( Scaturro et al. 2007).
  • Hypochlorite concentrations of 256 ppm or less were found sufficient to reduce culturability of seven pure culture L. pneumophila strains (initial concentration: 5 x 106 CFU ml-1) to undetectable levels. When ingested by Acanthamoeba polyphaga, the same strains could resist hypochlorite concentrations up to 1024 ppm. It was further shown that Legionella that were non-culturable after chlorination, were resuscitated when co-cultured with A. polyphaga. Intracellular pathogens proliferated robustly within the amoebae. Interestingly, intracellular L. pneumophila seemed to block encystation of A. polyphaga when exposed to hypochlorite. Increased resistance to this disinfectant was observed for both the intracellular pathogen and the host. The authors recommended that co-culturing of Legionella with protozoa should be employed when performing viability tests of L. pneumophila ( García et al. 2007).
  • A study looking at the efficiency of thermal and chlorine treatments showed that the association of Legionella with amoebae resulted in increased resistance to heat treatment. All six different Legionella species examined could be located in either of two thermotolerant Acanthamoebae (one clinical and one environmental isolate). Exposure up to 50°C resulted in a maximally 10-fold higher loss of culturability of planktonic Legionella compared to amoebae-ingested Legionella. Temperatures of 60°C, 70°C, and 80°C on the other hand led to a 100-fold greater loss of culturability in the absence of amoebae. In other word, the interaction with amoebae increased the heat resistance of Legionella 10- to 100 fold. In contrast to other studies, association with amoebae did not result in overall higher chlorine resistance. However, Acanthamoebae cysts were found to maintain viability after exposure to 100 mg/l chlorine for 10 min and to 80°C heat implying that they might provide protection of internalized pathogens even under harsh disinfection conditions ( Storey et al. 2004a).
  • Long-term starvation was shown to increase the resistance of Legionella against thermal and chlorine disinfection. Legionella is likely to encounter periods of starvation in stagnant water systems and respond to this environmental condition by entering a VBNC state. In this study both culturable and starved L. pneumophila were subjected to heat treatment (50-70°C) and chlorination (0.5 - 20 ppm). Loss of culturability occurred after starvation of 33-40 days. The authors suggested that the non-culturable state induced long-term starvation might be associated with increased resistance to thermal and chlorine disinfection ( Chang et al. 2007).
  • Efficacies of different disinfection procedures for removal of L. pneumophila  were compared in a model domestic water systems with galvanized steel re-circulation loops and copper dead legs. Biofilm populations were allowed to form by circulating sand-filtered river water through the system for 2 weeks followed by progressive dilution with de-chlorinated drinking water for 1 month. This resulted in a complex ecosystem naturally contaminated with L. pneumophila and different amoebae. The natural Legionella contamination was increased by seeding with lab-grown L. pneumophila, before applying disinfectants to separate loops. Continuous application of chlorine (2.5 mg per liter) and chlorine dioxide (0.5 mg per liter) were found efficient in reducing microbial flora including L. pneumophila. Ozone (0.5 mg per liter) was somewhat less efficient in reducing heterotrophic bacteria, whereas contamination was reported significantly higher with monochloramine (0.5 mg per liter) and copper-silver ionization (0.8/0.02 mg per liter). The latter two disinfectants were not successful in removing biofilms and during copper-silver treatment Legionella re-growth was observed.  No treatment could eliminate amoebae leading to recovery of Legionella to initial levels in all cases after ending the treatment. The authors pointed out that constant maintenance of disinfection residuals and control of amoebae is important for Legionella control ( Thomas et al. 2004).
  • Comparing different disinfection methods (chlorine, ozone, heat, and UV light) in a model plumbing system with a suspension of 107 CFU/ml planktonic L. pneumophila, UV light at a dose of 30 mW-s/cm-2 at 254 nm resulted in a 4 to 5 log decrease in culturability within 20 min. However, no further inactivation was observed after 20 min for up to 6 hours. Both continuous chlorination (4 to 6 ppm) and continuous ozonation (1 to 2 mg per liter) resulted in a gradual 5-6 log decrease within 6 hours. Heat (50 to 60°C) seemed to be the most efficient disinfection procedure resulting in complete loss of culturability in less than 3 hours ( Muraca et al. 1987). These data from 1987 have to be seen in context, however, with the more recent finding that disinfection efficacy is strongly influenced by Legionella's interaction with biofilms and amoebae.
Legionellae are able to survive under a wide range of water conditions, including temperatures from 0 to 63°C and a pH range of at least 5.0-8.5 ( Health Canada 2006).

Temperature tolerance. Whereas growth is restricted to temperatures between 20 and app. 43°C, L. pneumophila has been isolated from water at temperatures below 10°C (surface water) up to 60°C (artificial water systems) ( Wullings and van der Kooij 2006). In a study examining Legionella spp. in hot water systems (with recirculation), colony counts were reported up to temperatures of about 56°C. The range between 55°C and 60°C seemed to be a critical temperature region above which survival in hot water systems is affected ( Darelid et al. 2002; Mathys et al. 2007). Exact temperature tolerance values vary between different studies as might be assumed due to strain variability, different exposure times, different bacterial 'history' and different experimental conditions. Although outbreaks caused by natural water were mainly associated with temperatures higher than 30°C ( Lemarchand et al. 2004), Legionella can survive for long periods of time at low temperatures. Paszko-Kolva et al. reported the potential of clinical and environmental L. pneumophila isolates to persist and survive up to 2.5 years under low nutrient conditions when seeded into drinking and creek water monitored by acridine orange direct counts and viable counts ( Paszko-Kolva et al. 1992). It was speculated that L. pneumophila detected in cold water systems are survivors from summer periods when water temperatures were above 20°C ( Wullings and van der Kooij 2006).

Nutrient and mineral requirements. Legionella proliferates in water when nutrients are available. The nutrients may be given in form of dissolved organic constituents or indirectly by the presence of other microorganisms which undergo decay resulting in nutrient release ( Grimes DJ 1991). The presence of iron is important as it promotes and assists the growth of Legionella. Studies performed in several facilities such as storage cisterns and service pipe works have shown a clear correlation between presence of iron oxide and growth of Legionella ( Borella et al. 2004). The iron is not only a determining factor for planktonic, but also for intracellular growth of L. pneumophila. Although virtually all pathogens require iron, L. pneumophila has relatively high metabolic requirements of this metal. The multiplication of L. pneumophila in human mononuclear phagocytes depends on the availability of labile iron that is obtained primarily from three proteins: transferrin, lactoferrin, and ferritin. It has been shown that agents that inhibit the availability of iron inhibit the intracellular multiplication of bacteria (Byrd and Horwitz 1989; Byrd and Horwitz 1990; Byrd and Horwitz 1991). Also the presence of other salts and minerals within the water have an effect on survival ( Heller et al. 1998; States et al. 1985). States et al. found that low levels of certain metals like zinc, or potassium (which are found in plumbing systems) enhance growth of L. pneumophila, although elevated concentrations are toxic ( States et al. 1985). Also low concentrations of sodium chloride (0.1 - 0.5%) were found beneficial for survival ( Heller et al. 1998). Under low-nutrient or adverse environmental conditions, survival might also be ensured by adapting a dormant state ( Steinert et al. 2002). Viable-but not culturable cells might represent a large portion of the environmental Legionellae population ( Steinert et al. 1997). This would correlate with the fact that higher cell numbers are typically obtained with molecular methods compared to culture isolation.

pH preference. Legionellae are acid-tolerant (they can withstand exposure to pH 2.0 for short periods) and have been isolated from environmental sources ranging from pH 2.7 to 8.3 ( Anand et al. 1983). Ohno et al. reported an optimal range between pH 6.0 and 8.0 in a microcosm study ( Ohno et al. 2003).

Association with other microorganisms. Although planktonic Legionella spp. have been reported to survive for extended time periods in the environment, their extraordinary ability to interact with other microorganisms is highly beneficial for survival. They attach to surfaces in aquatic environments and readily form biofilms together with other microorganisms. Apart from providing a protective environment and increasing survival of community members, biofilms allow interaction with other microorganisms. The beneficial impact of the presence of non-Legionella species is known to influence Legionella growth. Moreover, Legionella are well known to maintain parasitic relationships with protozoa like Hartmanella, Acanthamoeba, Valkampfia, Naegleria spp., and other protozoa like Tetrahymena ciliates ( Lemarchand et al. 2004; Steinert et al. 2002). As Legionella have the capacity to survive and replicate intracellularly in their hosts, the interaction with protozoa is considered a key factor in the Legionella amplification process ( Steinert et al. 1997). Indeed, although most Legionella species can be cultivated under laboratory conditions, it is not known to what extent they occur in the free planktonic state in water ( Brooks et al. 2004). In the water environment, Legionellae might live mainly or even exclusively within or associated with other organisms. It was suggested that Legionella only replicate within protozoan and infected mammalian cells or on laboratory media ( Kwaik YA 1998; Fields B 1996), but growth in the planktonic state or in biofilms cannot be excluded. Legionellae living intracellularly are believed to have increased resistance to harsh environmental conditions and better survival characteristics.


A summary of selected studies on survival is given in the following:

  • Survival of L. pneumophila Flint 1 serogroup 1 was studied in different water types. The population showed almost no decline in sterilized or non-sterilized tap water up to 21 days at 5, 24, or 35°C. For longer incubations lower temperatures were highly favorable for survival. In non-sterilized tap water, 20% and 0.3% of cells were still culturable after 299 days at 5°C and 24°C, respectively. At 35°C no culturability was observed after 70 days of exposure. In sterilized tap water survival was better with 80% (5°C) and 10% (24°C) of cells still being culturable after 299 days. A concentration of at least 0.2 mg of chlorine per ml was required to eliminate 90% of Legionella in 2 hours ( Hsu et al. 1984).
  • Steinert et al. reported successful resuscitation of dormant L. pneumophila after passage through Acanthamoeba castellanii. Bacteria lost culturability in sterilized tap water (20°C) after incubation for 125 days ( Steinert et al. 1997). After an additional 55 days of non-culturability, 1 ml aliquots were co-incubated at 37°C with 105 amoeba. Culturable Legionella reoccurred and colony counts increased over time. After 3 days of co-incubation, cell numbers exceeded the initial concentration (8 x 106 cells/ml compared to 104 cells/ml) indicating growth. Infectivity of reactivated cells in human monocytes and infected guinea pigs was comparable with a culture of the same strain grown to log-phase. Dormant cells, on the other hand, which were not passaged through amoeba could not be resuscitated in guinea pigs ( Steinert et al. 1997).
  • In a 42°C tap water microcosm seeded with 106 CFU/ml of exponential of post-exponential L. pneumophila, cells lost culturability (but not their metabolic activity) within 61 and 41 days, respectively. Addition of microcosm samples to well-attached Acanthamoeba spp. resulted in reappearance of colonies. However, only a small portion of the L. pneumophila population could be revived (in contrast to the study by Steinert et al. ( Steinert et al. 1997) and resuscitation with other strains was not successful. When A. castellanii was added into a hot spring microcosm (42°C, pH 5) which normally resulted in complete loss of L. pneumophila culturability in 2 months, metabolic activity was maintained much longer in the presence of amoebae and bacterial cell numbers increased up to 10-fold after incubation for 1 additional month. Interestingly, no colonies formed ( Ohno et al. 2003).
  • L. pneumophila was reported to be able to replicate in biofilms in water distribution pipes in the presence of amoeba. A reactor was preconditioned with Aeromonas hydrophila, E. coli, Flavobacterium breve and Pseudomonas aeruginosa, forming a mixed biofilm. When L. pneumophila was added, cells were detectable in the biofilm already after 2 h indicating fast biofilm colonization. Biofilm-associated Legionella initially increased after addition of Acanthamoeba castellanii and intracellular replication within the amoebae. More than 50% of the amoebae population was filled with intracellular L. pneumophila as estimated by FISH. Lysis of the amoebae after 3 days resulted in an increase in planktonic L. pneumophila, whereas biofilm-associated Legionella decreased ( Declerck et al. 2007b).
Although some studies have investigated the infective dose and the obtained results have indicated that Legionella spp. represent a health risk to humans when a concentration of 104 to 105 CFU per liter potable water is exceeded ( Delgado-Viscogliosi et al. 2005), the median infective dose (LD50) of Legionella is unknown for healthy adults. Outbreaks affecting mainly immunocompromised people can occur at low concentrations ( Patterson et al. 1994; Meenhorst et al. 1985). It is known that hospital patients have become infected after exposure of only a few minutes to the sources of some outbreaks, whereas the infection has also been reported at up to 3.2 km from the source of other outbreaks (WHO). Nevertheless, infection depends on many factors such as the bacterial load of the water/microbial quality, the degree of aerosol formation and the resulting dissemination of bacteria through air, host factors, the virulence of the particular strain of Legionella and their viability. Inhaled protozoa might serve as cofactors in pathogenicity: L. pneumophila-infected protozoa were more pathogenic in mice than the equivalent number of bacteria or a co-inoculum of bacteria and uninfected protozoa ( Brieland et al. 1997). Another study showed that L. pneumophila was at least 100-fold more invasive for human epithelial cells and 10-fold more invasive for mouse macrophages when grown in the presence of Acanthamoeba castellani compared to pure cultures ( Cirillo et al. 1994).

Many countries have developed guidelines or regulations for the control of Legionella in water systems and for the prevention of legionellosis. Guidelines are advisory, whereas regulations and codes of practice have a more formal standing and are supported by legislative enforcement (including, in the case of regulations, specific information on managerial responsibility and operator competency). The guidelines include NHMRC (1988), CDC (2003), Allegheny County Health Department, Pennsylvania (1997), ASHRAE (2000), Ehrlich, Steele & Sabatini (2000), Standards Association of Australia/Standard Association of New Zealand (2002), EWGLINET (2003), and WHO (2004, 2006). Regulations include HSC (2000), IEE (2001), Spanish RD 865 (2003), and Victorian Department of Human Services (2001). In general, each country has their own regulations and guidelines, but all serve to achieve efficient risk management. These regulations and guidelines include the necessity of testing programs and all of them include the culture isolation method based on norms such as ISO11731, the joint Australian/New Zealand Standard AS/NZS 3896, and AFNOR T90-431.

The "gold standard" isolation of Legionella is based on ISO11731 as a reference. This norm specifies how to proceed to isolate and enumerate Legionella from water samples: sample concentration by membrane filtration, different treatments aimed at decreasing the competing flora, plating onto Buffered Charcoal Yeast Extract (BCYE) supplemented with L-Cysteine and glycine, polymixin, vancomycin and cycloheximide (GVPC) agar plates, and incubation under proper conditions during up to 10 days.

In other fields distinct from the environmental testing, such as clinical diagnosis, culture isolation following ISO11731 is used in cases of severe respiratory failure, in compromised patients, nosocomial infections and in cases in which disease is caused by any Legionellae other than L. pneumophila serogroup 1. In all other situations, alternative methods are more frequently used due to the drawbacks of culture isolation of Legionella (slow growth-rate, inability to detect viable-and-non-culturable bacteria, and difficulty of isolating Legionella in samples with a high background of competing flora). In this sense, one of the most commonly employed methods is the urinary antigen detection by enzyme immunoassay (EIA). The outcome of the disease with Legionnaires' disease greatly depends on the speed with which a diagnosis can be performed and the initiation of appropriate therapy (Heath et al. 1996). Urinary antigen detection for the diagnosis of Legionnaires' disease has several advantages compared with other methods. It is a rapid, sensitive, and specific method for the detection of serum antibody levels. In addition to the EIA method described above, the market offers a variety of commercial kits with different characteristics and specifications ( Dominguez et al. 1998; Helbig et al. 2001).
As mentioned above, culture isolation of Legionella present a number of drawbacks such as that is time-consuming due to the slow growth rate of the Legionella, the inability to detect viable-and-non-culturable bacteria, and the difficulty of isolating Legionella in samples with a high background of competing bacterial flora. Nevertheless, the culture isolation procedure based on ISO11731 is the generally accepted standard procedure for Legionella testing in environmental samples to fulfill the regulations.
Alternative molecular methods have been described in order to overcome some disadvantages of Legionella culture isolation. Some of them are listed below:

  • Nested PCR-AGE: An assay by Catalán et al. targeting the mip gene could detect as few as 2 genome copies per reaction ( Catalán et al. 1994). The method was used successfully after removal of PCR inhibitors to detect L. pneumophila in wastewater ( Catalán et al. 1997). Devos et al. compared the Legionella-specific primer pairs JFP-JRP ( Jonas et al. 1995, modified protocol) and LEG225-LEG 858 ( Miyamoto et al. 1997) in combination with 16S primers (first step PCR) in a nested PCR reaction. The two approaches were applied on 46 aquatic samples: 41% tested positive for Legionella spp. using the plate count technique ISO 1131, 56% were positive with the LEG-primers and 98% were positive with primers JFP-JRP. Sequencing of amplicons showed that 70% of the JFP-JRP amplicons comprised L. pneumophila. It was concluded that the JFP-JRP primers in a nested assay were sensitive enough for testing of environmental samples ( Devos et al. 2005).
  • Seminested PCR-AGE with IAC targeting dotA gene of L. pneumophila. The assay reliably detected 15 genome copies in pure samples ( Yáñez et al. 2007). The inclusion of the IAC allows the detection of potential PCR inhibition, but may results in slightly higher detection limits compared to PCR assays without IAC.
  • Dilution PCR-AGE: Undiluted and diluted sample DNA was tested with PCR targeting 16S rRNA genes for Legionella spp. and the mip gene for specific detection of L. pneumophila. Amplification products were analyzed using AGE. Initial DNA concentrations are calculated from the highest DNA dilution that still yields PCR product ( Wullings and van der Kooij 2006).
  • qPCR: An increasing number of different real-time PCR approaches can be found.
    - Ballard et al. presented a real-time PCR LightCycler assay  for detection of L. pneumophila. The detection limit was reported to be 2.5 CFU/reaction, corresponding to app. 1,000 cells/liter in the original water sample. Due to DNA loss during extraction, the authors estimated the sensitivity of 2.5 CFU corresponds to app. 25 genome copies per reaction ( Ballard et al. 2000). The assay was applied by Levi et al. to 100 random environmental water samples. The detection limit was reported to be 800 cells per liter ( Levi et al. 2003).
    - Wellinghausen et al. described a qPCR assay for detection of Legionella spp. (based on 16S rRNA genes) and L. pneumophila (based on the mip gene) in potable water. Detection limits of 16S rRNA detection was less than one genome copy, the sensitivity of the mip assay was reported similar. The test was successfully validated with 100 ml water samples ( Wellinghausen et al. 2001).
    -  Joly et al.: The primers and probes published by Jonas et al. ( Jonas et al. 1995) were applied in a qPCR assay with samples from hot water systems and cooling towers. Detection and quantification limits depended on the samples and were estimated between 0.6-10 genome copies per assay for the 16S rRNA genes (corresponding to 30-500 genome copies per liter; for detection of Legionella spp.) and 6-100 genome copies per assay for the mip genes (corresponding to 300 and 5,000 genome copies per liter; for detection of L. pneumophila). Correlation between numbers produced by qPCR and plate counting was very poor with PCR numbers being sometimes dramatically higher ( Joly et al. 2006).
    -  Morio et al. presented a qPCR for detecting L. pneumophila targeting the mip gene. The limit for reproducible detection was reported to be 600 genome copies per liter. Quantification was reliable with concentrations exceeding 800 genome copies per liter. These thresholds depended on the water volume. The assay was validated wit hot water samples from hospitals. The correlation with cultivation-based detection was very weak as seen in other studies ( Morio et al. 2007).
    -  Yaradou et al: An integrated qPCR assay was developed by GeneSystems (Bruz, France) and was designed for routine detection of Legionella in water samples. The system includes a sample filtration unit and a DNA extraction unit. The detection limit was reported to be 5 genome copies per reaction. The method was evaluated by testing samples from hot water systems and cooling towers. Obtained numbers showed an acceptable correlation with culture results for the hot water samples (r2=0.732), but not for the cooling tower samples. The higher complexity of the cooling tower samples were suspected to be thre reason. Concentrations suggested by qPCR were on average 3.1 and 4.5-fold higher than the concentrations obtained by cultivation for the two samples types, which correlates with other studies ( Yaradou et al. 2007).
    -  Yáñez et al: A IMS + qPCR with IAC assay targets the defective organelle trafficking (dotA) gene. The detection limit is app. 7 genome copies per reaction. The assay was applied to potable water and cooling tower samples. In the positive samples, the numbers of cells calculated by PCR was 20% higher than the culture values ( Yáñez et al. 2005). The use of IMS was also evaluated in a study by Riffard et al. ( Riffard et al. 2005).
  • ChemScanRDI: The method is based on solid-phase cytometry and was reported to allow enumeration of the majority of L. pneumophila serogroups within 3 to 4 hours. The assay is based on flow-cytometric detection of cells which were labeled with FITC-conjugated antibodies. Sensitivity was reported to be in the range between 10-100 cells/liter. When applied to water samples from hospital hot water systems, the obtained counts were higher than the ones obtained by a standard culture method. The ratios between the two methods ranged from 1.4 to 38 (except two very high ratios of 148 and 325) (Aurell et al. 2004).
  • Immunological double-staining: The method is based on epifluorescence microscopy of filterable water samples stained with Legionella-specific antibodies and the viability marker ChemChrome V6. ChemChrome V6 gets cleaved by esterases in viable cells converting the red ChemChrome into a green fluorescent product which is trapped within the cells. Heat-killed cells do not show esterase activity. Two commercial polyclonal fluorescein isothiocyonate-conjugated antibodies (red) were used for detection of L. pneumophila and other Legionella species. Stained cells were counted under the microscope with viable cells showing green and red fluorescence and dead cells staining only red. For natural water samples containing > 3 log10 CFU/liter there was good correlation between culture counts and the double-staining method. For lower concentrations the correlation was lost with the microscopically determined numbers being too high. The detection limit was found dependent ton the water volume filtered (higher volumes contain more particles) and the number of miscroscope fields examined. Under clean experimental conditions the detection limit was 176 cells/liter (compared to 50 CFU/liter using the French standard culture method). Also with filterable water samples, the method was reported to be at least as sensitive as culturing. Microscopic counts were normally higher than plate counts. Filtration of volumes larger than 100 ml was not recommended. ( Delgado-Viscogliosi et al. 2005)

Typing methodologies for epidemiological studies

Epidemiological linkage of clinical and environmental Legionella isolates is invaluable in locating the source and extent of infection, allowing implementation of corrective measures and treatment to prevent further infection. Previous studies have demonstrated the usefulness of differentiating isolates belonging to serogroup 1 of L. pneumophila in order to confirm or refute epidemiological associations in outbreak investigations ( Edelstein et al. 1986), but only in a local setting. Isolates of L. pneumophila sg 1 can be rapidly subtyped by using monoclonal antibody (MAb) subgrouping with panels based on the international MAb subgrouping panel (Harrison et al. 1998). However, MAb subgrouping lacks discrimination, dividing L. pneumophila sg 1 into only 8 to 10 phenons, depending on which MAb panels are used. Consequently, in addition to this phenotypic typing method, a number of molecular methods have been successfully employed locally for epidemiological purposes. These include ribotyping, amplified fragment length polymorphism (AFLP) analysis, pulsed-field gel electrophoresis (PFGE), restriction fragment length polymorphism (RFLP) analysis, restriction endonuclease analysis (REA), and arbitrarily primed PCR ( Fry et al. 1999). One of these methods, AFLP analysis ( Valsangiacomo et al. 1995), has been adopted as an international standard and it is now widely used by members of the European Working Group for Legionella Infections (EWGLI). Interlaboratory studies ( Fry et al. 1999) have assessed the reliability and discriminatory power of AFLP analysis for epidemiological typing of L. pneumophila sg 1 isolates. While these studies have demonstrated that the method is robust and rapid and allows the exchange of data between laboratories without the need for exchange of isolates, a program of proficiency testing showed that a significant proportion of laboratories could not correctly identify isolates 100% of the time (Fray et al. 2000). Incorrect identification was typically a result of data analysis issues (e.g., difficulties in comparing the AFLP banding patterns) rather than of the method itself. For this reason, a universal typing method was developed that would be simple, rapid, discriminatory, and truly "portable" and that might in the future be applied by all microbiology laboratories to aid in the investigation of legionellosis. The method is based on the sequence of housekeeping genes and named Multilocus Sequence Typing (MLST) (Enright and Spratt 1998). A modification of this method, "sequence-based typing" (SBT), includes, apart from housekeeping genes, genes under selective pressure, e.g. virulence genes, which offer more information about variability of closely related strains ( Gaia et al. 2003; Gaia et al. 2005). The results demonstrated the advantages of SBT of L. pneumophila sg 1 over other current methods. By using 5 loci, this methodology has the potential to become a new "gold standard" for the epidemiological typing of L. pneumophila ( Gaia et al. 2005).

The following pictures are part of the Public Health Image Library (PHIL) from the Centers for Disease Control and Prevention (CDC). The photos are in the public domain and thus free of copyright restrictions. We would like to express our appreciation for providing these images.


Figure 1:

                                                  Legionella_photo 1                

Under a moderately-high magnification of 8000X, this colorized scanning electron micrograph (SEM) depicted a large grouping of Gram-negative Legionella pneumophila bacteria. Of particular importance, is the presence of polar flagella, and pili, or long streamers, which due to their fragile nature, in some of these views seem to be dissociated from any of the bacteria.

Source: http://phil.cdc.gov/phil/home.asp
Photo ID: 11152
Content provider(s): Centers for Disease Control and Prevention/ Margaret Williams, PhD; Claressa Lucas, PhD;Tatiana Travis, BS
Credit: Janice Haney Carr

Figure 2:

                                                  Legionella_photo 2             

Under a moderately-high magnification of 5011X, this 2002 scanning electron micrograph (SEM) revealed some of the ultrastructural morphology exhibited by small grouping of Hartmannella vermiformis amoebae trophozoites.

Source: http://phil.cdc.gov/phil/home.asp
Photo ID: 11183
Content provider(s): Centers for Disease Control and Prevention/ Dr. Rodney Donlan
Credit: Janice Haney Carr

Last Updated on Tuesday, 03 April 2012 02:40


Hepatitis A
Hepatitis E


Copyright © 2017 Waterborne Pathogens. All Rights Reserved. Powered by SuSanA