OUP user menu

Free-living amoebae and their intracellular pathogenic microorganisms: risks for water quality

Vincent Thomas, Gerald McDonnell, Stephen P. Denyer, Jean-Yves Maillard
DOI: http://dx.doi.org/10.1111/j.1574-6976.2009.00190.x 231-259 First published online: 1 May 2010


An increasing number of microorganisms, including bacteria but also viruses and eukaryotes, have been described as benefiting from interaction with free-living amoebae (FLA). Beneficial interaction can be due to resistance to predation conferring ecological advantage, intracellular survival and/or intracellular proliferation. This review highlights the potential risk associated with amoebae by listing all known pathogenic microbial species for which growth and/or survival promotion by FLA (mainly Acanthamoeba spp.) has been demonstrated. It focuses on the susceptibility of amoebal and intra-amoebal bacteria to various categories of biocides, the known mechanisms of action of these biocides against trophozoites and cysts and the various methods used to demonstrate efficacy of treatments against FLA. Brief descriptions of FLA ecology and prevalence in domestic/institutional water systems and their intrinsic pathogenicity are also presented. The intention is to provide an informed opinion on the environmental risks associated with the presence of FLA and on the survival of cysts following biocidal treatments, while also highlighting the need to conduct research on the roles of amoebae in aquatic ecosystems.

  • amoebae
  • Acanthamoeba
  • intracellular
  • survival
  • biocides
  • cysts


Free-living amoebae (FLA) are widespread in nature and are normal inhabitants of freshwater microbial ecosystems (Rodriguez-Zaragoza et al., 1994; Khan et al., 2006). They are thought to have a major impact on the dynamics of multimicrobial biofilms by feeding on various microorganisms and contributing to nutrient recycling (Pedersen et al., 1982). Several FLA species are potentially pathogenic to humans and animals but infections are not commonly reported. FLA per se are not considered to constitute a major threat to public health, although some have been involved in specific diseases. However, it has been recently recognized that FLA can interact with a variety of microorganisms (Greub & Raoult, 2004; Khan et al., 2006), in a way that benefits those microorganisms (particularly studied with bacteria but also observed with fungi and viruses). Bacteria can benefit from interactions with FLA due to (1) their ability to escape predation and grow in the presence of a protozoan that would normally phagocytose and digest nonresistant bacterial species; (2) their ability to resist intracellular digestion (intracellular survival, with the possible subsequent survival within a protozoan cyst); and (3) their ability to resist digestion but also to grow within the protozoan vegetative form (trophozoite; intracellular multiplication). Furthermore, several studies have demonstrated that the virulence of known pathogenic bacteria toward their protozoan hosts can reflect virulence toward humans and/or animals (Fields et al., 1986; Fenner et al., 2006; Goy et al., 2007; Steinberg & Levin, 2007). It has been postulated that the newly discovered amoebae-resisting bacteria (ARB) are likely to be pathogenic for humans and/or animals (Greub & Raoult, 2004; Molmeret et al., 2005). Consequently, FLA have been used as a tool to isolate new potentially pathogenic ARB species from various sources (Collingro et al., 2005; Thomas et al., 2006b).

From a public health perspective, amoebae, and notably amoebal cysts, can be highly resistant to various physical and chemical stresses and can thus protect any intracellular microorganism from deleterious environmental conditions that would normally kill them (King et al., 1988; Kilvington & Price, 1990; Barker et al., 1992). This protective effect is of increasing concern because it is speculated that partially efficient biocide treatments might select for FLA together with their intracellular microorganisms. They may even provide favorable conditions directly [the biocide treatment itself stimulating FLA growth (Srikanth & Berk, 1993)] or indirectly [biocidal treatment killing extracellular bacteria that are then used as a food source by Legionella pneumophila as demonstrated by Temmerman et al. (2006)].

A good understanding of ways to control FLA in water or other liquids is therefore most important. There is general paucity of information on the efficacy and mechanisms of action of biocides against amoebae. Most of the information available concerns other water-transmitted protozoal species such as Cryptosporidium and Giardia spp. that have been involved in gastrointestinal disease outbreaks. For FLA, mainly Acanthamoeba spp. have been evaluated against biocides used in contact lens solutions because of their association with keratitis and contamination of these solutions. Several studies were also published concerning the resistance of Acanthamoeba spp., Hartmannella spp. and Naegleria spp. to drinking water treatments but there is a general lack of information and clear discrepancies between studies on the efficacy of biocides against amoebal trophozoites and particularly their resistant forms (cysts). The reported variability of results between studies can be attributed to the lack of an international consensus in standard efficacy protocols to grow amoebal cysts and measure the cysticidal activity of biocides (Mercer, 2008).

The prevalence of amoebae in water networks and their association with biofilms

FLA and other protozoa are normal inhabitants of freshwater sources and soils (Rodriguez-Zaragoza, 1994). In these environments, the exact FLA population composition is dependent on the actual physicochemical parameters present, such as annual temperature fluctuation and pH changes (Kyle & Noblet, 1986). FLA can also colonize domestic and institutional water systems. It has been demonstrated that although clarification steps used in drinking water production plants dramatically reduce their numbers, some cells can spread from the water source to the distribution network despite disinfection of water with ozone and chlorine (Hoffmann & Michel, 2001; Corsaro et al., 2008; Thomas et al., 2008). Once in the distribution network, low disinfectant levels have only limited activity on FLA (Thomas et al., 2004). They can thus colonize virtually any kind of water system and have been isolated from a number of diverse environments, some containing harsh physical and/or chemical conditions such as elevated temperature (hot tubes and cooling towers) or a high concentration of a biocide or combination of biocides such as chlorine or chlorine-releasing agents (CRAs), biguanides and other agents. By way of example, they have been recovered from domestic tap water (Jeong & Yu, 2005), hospital water networks (Rohr et al., 1998; Thomas et al., 2006b), swimming pools (Vesaluoma et al., 1995), hydrotherapy baths (Scaglia et al., 1983), dental unit waterlines (Singh & Coogan, 2005), eyewash stations (Paszko-Kolva et al., 1991) and cooling towers (Barbaree et al., 1986; Berk et al., 2006). In large studies including many sampling points, amoebae were found in 20–30% of domestic tap water samples (Shoff et al., 2008). An even higher prevalence has been reported in hospitals, with as much as 68.9% of samples collected from hot water faucets, showers, hot water tanks and cooling towers being colonized (Lasheras et al., 2006). In a study of FLA colonization of domestic water networks in homes of patients suffering keratitis, Kilvington et al. (2004) found 89% of homes colonized, with FLA recovered from 76% of bathroom sink cold taps sampled. They concluded that water storage tanks promote colonization of domestic water with FLA (Kilvington et al., 2004). Amoebae cultivated from domestic water systems mostly belong to the genera Acanthamoeba, Hartmannella and Naegleria, but species belonging to the Echinamoeba, Vannella, Vahlkampfia and Saccamoeba genera have also been isolated (Rohr et al., 1998; Barbeau & Buhler, 2001; Hoffmann & Michel, 2001; Thomas et al., 2006b, 2008) (Fig. 1). Most species of these genera can form cysts that present various degrees of resistance to harsh environmental conditions.

Figure 1

Example of various FLA isolates recovered from a drinking water treatment plant (reproduced from Thomas et al., 2008). (a) Isolate closely related to Playtamoeba genus; (b) isolate closely related to Acanthamoeba genus; (c) isolate closely related to Vanella genus; (d) isolate closely related to Acanthamoeba genus; (e) isolate closely related to Glaeseria genus; (f) isolate closely related to Hartmannella genus; (g) isolate closely related to Naegleria genus; (h) isolate closely related to Naegleria genus; (i) isolate closely related to Hartmannella genus; (j) isolate closely related to Echinamoeba genus; (k) isolate closely related to Acanthamoeba genus. Scale bar for all pictures=10 μm.

It has been demonstrated that suspended bacteria do not provide particularly favorable conditions for FLA growth (Pickup et al., 2007a) and that FLA proliferation in water systems is mainly due to grazing on bacterial biofilms (Barbeau & Buhler, 2001) with which FLA are integrally associated (Huws et al., 2005). Protozoal grazing on bacterial biofilms happens to be a key factor regulating biofilm composition and dynamics (Pedersen et al., 1982; Kalmbach et al., 1998). Not all bacteria seem to be equally suitable food sources for amoebae; this will depend on the specific amoebae and bacterial strains, but antipredatory mechanisms may arise, including microcolony formation, toxin production and the presence of an intact capsule that can prevent feeding by FLA on bacteria (Weekers et al., 1993; Pickup et al., 2007b). Some of these mechanisms are thought to be predation-driven and regulated by N-acylhomoserine lactone (AHL)-dependent quorum-sensing systems, as demonstrated with the violacein toxin production by Chromobacterium violaceum (Matz et al., 2004). Interestingly, AHLs might also directly interact with eukaryotic cells (Joint et al., 2002). Other mechanisms that are not under direct dependency of quorum sensing have been described, such as those demonstrated for Pseudomonas aeruginosa type III secretion system that is used to kill biofilm-associated amoebae (Matz et al., 2008). Of note, whereas it is well known that microorganisms embedded within a biofilm are less susceptible to the effects of antimicrobial treatments including disinfection and antibiotic chemotherapy, the fate of trophozoites and cysts within microbial biofilms following biocidal treatment is under-reported and requires further investigation.

The intrinsic pathogenicity of FLA and the speculative role of intracellular bacteria

FLA have been associated with a number of diseases in humans (for reviews, see Visvesvara et al., 2007 and Marciano-Cabral & Cabral, 2003). Encephalitis and meningoencephalitis are life-threatening infections affecting immunocompetent children and immunocompromised adults. They mainly involve Acanthamoeba spp. and Balamuthia mandrillaris for encephalitis and Naegleria fowleri for meningoencephalitis. Sappinia diploidea has also been isolated from one human case of encephalitis (Visvesvara et al., 2007). Treatment is problematic due to difficulty in diagnosis, speed of the infection in the case of meningoencephalitis, resistance to some therapeutic compounds and the dormancy of cysts that can lead to subsequent patient relapse (Schuster & Visvesvara, 2004). Other types of infections have been reported: local or disseminated infections in immunocompromised adults but also in immunocompetent children; these infections mainly involve Acanthamoeba spp. (Marciano-Cabral & Cabral, 2003). Acanthamoeba spp. are also the main cause of corneal infection leading to keratitis associated with contact lenses (Ahearn & Gabriel, 1997; Illingworth & Cook, 1998), although several cases involving Vahlkampfia sp., Vannella sp. and Hartmannella sp. have also been described (Aitken et al., 1996; Michel et al., 2000b). Interestingly, some of these keratitis-associated FLA were found to be infected by various bacterial species (Fritsche et al., 1993; Michel et al., 2000b; Xuan et al., 2007; Yu et al., 2007; Scheid et al., 2008), thus raising concerns about the possible role of these intracellular bacteria in amoebal resistance to biocides and/or pathogenicity (Murdoch et al., 1998; Cengiz et al., 2000). Marciano-Cabral et al. (2003) also recovered an infected acanthamoebal isolate from an immunosuppressed patient who was diagnosed postmortem with disseminated acanthamoebiasis. They observed gram-negative rods in the trophozoites but did not indicate to which bacterial species these rods belonged. They speculated that amoebal pathogenicity may be a result of both tissue damage from amoeba-secreted products and the induction of high levels of proinflammatory cytokines (tumor necrosis factor-α) caused by intracellular bacteria. It should also be mentioned that Mycobacterium avium and L. pneumophila-infected FLA have been proved to be infectious particles in murine models of infection, and in these cases infected FLA were more pathogenic than an equivalent number of bacteria or a coinoculum of bacteria and uninfected amoebae (Brieland et al., 1997; Cirillo et al., 1997).

FLA interactions with pathogenic microbial species

Symbiotic and pathogenic interactions with bacteria

The association of various bacterial species with amoebae is thought to be common. Microscopic observation of cytoplasmic endosymbionts in Acanthamoeba castellanii were reported in 1967 (Jeon & Lorch, 1967) and 1975 (Proca-Ciobanu et al., 1975). Over a 40-year period, Jeon and colleagues published several studies describing an endosymbiotic bacteria (X-bacteria) that infected the amoebae initially as intracellular parasites, transferred from one amoebal isolate to another and thus established a new endosymbiotic relationship with the new host, which ultimately became necessary for amoebal survival (Jeon & Jeon, 1976). They later demonstrated that X-bacteria could switch expression of host genes coding for S-adenosylmethionine synthetase (sams) from one sams-encoding gene to another (Jeon & Jeon, 2004). Interestingly, X-bacteria were further demonstrated to belong to the Legionella order, being proposed as ‘Legionella jeonii’ (Park et al., 2004). A major interest in the interactions between FLA and pathogenic bacteria was triggered by the finding that human pathogenic species such as L. pneumophila proliferate in various amoebal hosts (Rowbotham et al., 1980). Fritsche et al. (1993) later reported that 25% of clinical and environmental Acanthamoeba spp. isolates harbored a variety of endosymbiotic bacteria. Another study reported that amoebae were more frequently infected in cooling towers (55%) than in natural environments (7.5%), suggesting that artificial conditions might favor association of bacteria with amoebae (Berk et al., 2006). In several cases, a single amoebal host was even reported to be infected with two different bacterial species (Michel et al., 2006; Heinz et al., 2007).

Bacterial species pathogenic to humans are also likely to resist digestion by FLA

It has been proposed that through their capacity to resist digestion by FLA, potential intracellular bacterial species (many yet to be discovered) are also likely to resist digestion by macrophages and thus may represent new human pathogenic species (Greub & Raoult, 2004). In our review, we studied demonstrated pathogenic species and particularly screened the Environmental Protection Agency (EPA) list of microorganisms that are known to cause disease in humans (‘CCL 3 Universe’ list, available at http://www.epa.gov/safewater/ccl/pdfs/report_ccl3_microbes_universe.pdf). This was compared with the current list of bacterial species that have been demonstrated elsewhere in the literature to ‘benefit’ interaction with protozoa. The CCL 3 Universe list is based on the comprehensive review by Taylor et al. (2001) that identified 538 bacterial species pathogenic to humans and/or animals. The EPA list has since been extended by two species but both from the genus ‘Legionella,’ recognizing that all species that belong to this genus are potentially pathogenic. Overall, there are now 539 bacterial species in the February 2008 EPA list. Of these 539 species, a comprehensive review of the literature allowed us to identify 102 (18.9%) that have been described as surviving and/or flourishing when in contact with various amoebal species (Table 1). Among these 102 species, 40 (39.2%) were isolated by amoebal coculture without complete demonstration of intracellular survival, 30 (29.4%) have been demonstrated to survive in one or several amoebal species and 32 (31.4%) have been shown to survive and grow in one or several amoebal species. These are likely to be underestimates of the true level of FLA–bacterial associations, when the following limitations are considered. First, most of the published studies used Acanthamoeba polyphaga strain Linc-Ap1 (available at the Culture Collection of Algae and Protozoa as CCAP 1501/18) or A. castellanii ATCC 30234 as the cell background, thus considerably limiting the possible range of bacteria–amoeba interactions that are often host specific. For example, it has been demonstrated that a L. pneumophila strain virulent to humans can grow within Acanthamoeba lenticulata strain PD2 whereas another virulent strain cannot (Molmeret et al., 2001). This might be the reason why, to our knowledge, there is no published report demonstrating intraprotozoal survival for several Legionella species: Legionella birmighamensis, Legionella cherrii, Legionella cincinnatiensis, Legionella jordanis, Legionella lansigensis, Legionella sainthelensi, Legionella wadsworthii and Tatlockia maceachernii (now Legionella maceachernii but still listed under its previous name in the CCL3 list). It is probable that these listed pathogenic species can also grow in specific amoebal hosts that have not been identified so far. Second, various bacterial strains reported in the CCL3 list have obviously never been tested for their interactions with protozoa. For example, the 12 Rickettsia strains listed should be considered as being strongly suspect of being resistant to protozoa because one species belonging to this genus, Rickettsia bellii (not in the CCL3 list), has been demonstrated to be able to survive for at least 6 weeks in A. polyphaga Linc-Ap1 (Ogata et al., 2006). Furthermore, Ehrlichia-like organisms have also been reported in Saccamoeba species (Michel et al., 1995b). In this case, the lack of published evidence for other rickettsial species survival within protozoa might be due to the difficulty in cultivating these obligate intracellular microorganisms (the majority classified as biosafety level 3). Other BSL3 species such as Mycobacterium leprae and Yersinia pestis were initially cited in the literature as being resistant to amoebae (Jadin, 1975; Nikul'shin, 1992) but these studies are very limited and lack strong scientific evidence. A more recent study tends to demonstrate that the leprae bacillus can survive for at least 72 h in A. castellanii ATCC 30010 (Lahiri & Krahenbuhl, 2008). Furthermore, symbiotic interactions between Acanthamoeba lugdunensis and a mycobacterial species related to M. avium and Mycobacterium intracellulare has been described recently; mycobacteria were propagated in protozoa for >6 years without any obvious detrimental effect for both microorganisms (Yu et al., 2007), demonstrating that amoebae can readily act as a stable environmental reservoir for mycobacteria.

View this table:
Table 1

Described interactions of pathogenic bacterial species with FLA

Bacterial speciesDescribed interaction with protozoaThreat listReferences
Emerging infectious diseasesCDC notifiableNIAID bioterror agentsFood and water pathogensHHS select agents
Achromobacter xylosoxidansCoculture and cell lysis (Ap Linc AP-1)Greub et al. (2004); Pagnier et al. (2008)
Acinetobacter baumanniiCoculture without cell lysis (Ap Linc AP-1, Ac ATCC 30010)xPagnier et al. (2008); Thomas et al. (2008)
Acinetobacter calcoaceticusCoculture without cell lysis (Ap Linc AP-1)xPagnier et al. (2008)
Acinetobacter haemolyticusCoculture (Ac ATCC 30010)Thomas et al. (2008)
Acinetobacter johnsoniiCo-culture (Ac ATCC 30010)Thomas et al. (2008)
Acinetobacter juniiCoculture without cell lysis (Ap Linc AP-1, Ac ATCC 30010)Pagnier et al. (2008); Thomas et al. (2008)
Acinetobacter lwoffiiCoculture (Ac ATCC 30010)Thomas et al. (2008)
Acinetobacter radioresistensCoculture (Ac ATCC 30010)xThomas et al. (2008)
Aeromonas caviaeIC survival (Ac ATCC 30234)Rahman et al. (2008)
Aeromonas hydrophilaIC survival (Ac ATCC 30234)Rahman et al. (2008)
Aeromonas veroniiIC survival (Ac ATCC 30234)Rahman et al. (2008)
Bacillus cereusIC multiplication (Ap Linc Ap-1)xxPagnier et al. (2008); Evstigneeva et al. (2009)
Bacillus licheniformisCoculture (Naegleria fowleri HB-1, Ac ATCC 30010)Lebbadi et al. (1995); Thomas et al. (2008)
Bacillus pumilusCoculture (Ac ATCC 30010)Thomas et al. (2008)
Bacillus subtilisCoculture without cell lysis (Ap Linc Ap-1)Pagnier et al. (2008)
Brevundimonas diminutaIC multiplication (Ap Linc Ap-1)Evstigneeva et al. (2009)
Brevundimonas vesicularisCoculture without cell lysis (Ap Linc Ap-1)Pagnier et al. (2008)
Burkholderia cepaciaIC multiplication (Ap Linc Ap-1), IC survival (Ap ATCC 50372)xLanders et al. (2000); Lamothe et al. (2004)
Burkholderia pseudomalleiIC survival (Acanthamoeba astronyxis CCAP 1534/1)xxxInglis et al. (2000)
Campylobacter coliIC multiplication (Ap Linc Ap-1)Axelsson-Olsson et al. (2007)
Campylobacter hyointestinalisIC multiplication (Ap Linc Ap-1)Axelsson-Olsson et al. (2007)
Campylobacter jejuniIC multiplication (Ap Linc Ap-1)xxxAxelsson-Olsson et al. (2007)
Campylobacter lariIC multiplication (Ap Linc Ap-1)Axelsson-Olsson et al. (2007)
Chlamydophila pneumoniaeIC survival (Ac ATCC 30234)xEssig et al. (1997)
Chromobacterium violaceumCoculture with complete cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Chryseobacterium meningosepticumCoculture with partial cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Citrobacter freundiiIC survival (Ac ATCC 30234), coculture (Ac ATCC 30010)King et al. (1988); Thomas et al. (2008)
Coxiella burnetiiIC survival (Ac ATCC 30234)xxxLa Scola & Raoult (2001)
Delftia acidovoransCoculture with partial cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Edwardsiella tardaIC survival (Tp)King et al. (1988)
Enterobacter aerogenesCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Enterobacter amnigenusCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Enterobacter cancerogenusCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Enterobacter cloacaeIC multiplication (Ap Linc Ap-1), IC survival (Ac ATCC 30234), coculture (Ac ATCC 30010)King et al. (1988); Thomas et al. (2008); Evstigneeva et al. (2009)
Escherichia coli (includ. 0157H7)IC multiplication (Ac, Ap), IC survival (Tp)xxxxBarker et al. (1999); Alsam et al. (2006); Steinberg & Levin (2007)
Francisella tularensisIC multiplication and IK survival (Ac ATCC 30234)xxxxAbd et al. (2003)
Hafnia alveiCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Helicobacter pyloriIC survival (Ac)xWiniecka-Krusnell et al. (2002)
Klebsiella oxytocaIC survival (Ac ATCC 30234), coculture without cell lysis (Ap Linc-Ap1)King et al. (1988); Pagnier et al. (2008)
Klebsiella pneumoniaeIC survival (Ac ATCC 30234), coculture without cell lysis (Ap Linc-Ap1)xKing et al. (1988); Pagnier et al. (2008)
Kluyvera cryocrescensCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Legionella anisaIC multiplication (Tp, Ac ATCC 30010, Ap Linc-Ap1)Fields et al. (1990); La Scola et al. (2001); Thomas et al. (2006a, b)
Legionella feeleiiIC multiplication (Tp), coculture (Acanthamoeba culbertsoni)Fields et al. (1986); Kuroki et al. (2007a)
Legionella hackeliaeIC multiplication (Tp)Fields et al. (1986)
Legionella longbeachaeIC multiplication (Acanthamoeba sp., Tp)Steele & McLennan (1996); Doyle et al. (1998)
Legionella oakridgensisIC multiplication (Tp)Fields et al. (1986)
Legionella pneumophilaIC multiplication and IK survival (>20 FLA species)xxFor a review, see Kuiper et al. (2006)
Legionella rubrilucensCoculture (Ap)La Scola et al. (2003b)
Listeria ivanoviiIC survival (Ac ATCC 30234)Zhou et al. (2007)
Listeria monocytogenesIC multiplication (Tp), IC survival (Ac ATCC 30234)xxxLy & Muller (1989); Ly & Muller (1990); Zhou et al. (2007)
Listeria seeligeriIC survival (Ac ATCC 30234)Zhou et al. (2007)
Listeria welshimeriIC survival (Ac ATCC 30234)Zhou et al. (2007)
Methylobacterium mesophilicum Coculture and cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Morganella morganiiCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Mycobacterium abscessusIC and IK survival (Ap Linc-Ap1)xAdekambi et al. (2006)
Mycobacterium aviumIC multiplication (Ac ATCC 30234 and CCAP 1501/1B, Ap Linc-Ap1 and CCAP 1501/3B,xCirillo et al. (1997); Steinert et al. (1998); Adekambi et al. (2006)
Dd AX2, Tp ATCC 30202), IK survival (Ap ATCC 30872 and Linc-Ap1)Strahl et al. (2001); Skriwan et al. (2002); Mura et al. (2006); Whan et al. (2006)
Mycobacterium bovisIC and IK survival (Ac CCAP 1501/1A)xTaylor et al. (2003)
Mycobacterium chelonaeIC and IK survival (Ap Linc-Ap1)Adekambi et al. (2006)
Mycobacterium fortuitumIC multiplication (Ac ATCC 30234), IC and IK survival (Ap Linc-Ap1)xCirillo et al. (1997); Adekambi et al. (2006)
Mycobacterium gordonaeIC and IK survival (Ap Linc-Ap1), coculture (Ac ATCC 30010)Adekambi et al. (2006); Thomas et al. (2006a, b); Thomas et al. (2008)
Mycobacterium kansasiiIC multiplication (Ac ATCC 30010), IC and IK survival (Ap Linc-Ap1)xAdekambi et al. (2006); Goy et al. (2007)
Mycobacterium lepraeIC survival (Acanthamoeba culbertsoni ATCC 30171)xJadin et al. (1975); Lahiri & Krahenbuhl (2008)
Mycobacterium malmoenseIC and IK survival (Ap Linc-Ap1)Adekambi et al. (2006)
Mycobacterium marinumIC multiplication (Ac ATCC 30234, Dd AX2), IC and IK survival (Ap Linc-Ap1)xCirillo et al. (1997); Solomon et al. (2003); Adekambi et al. (2006)
Mycobacterium mucogenicumIC and IK survival (Ap Linc-Ap1)Adekambi et al. (2006)
Mycobacterium peregrinumIC and IK survival (Ap Linc-Ap1)Adekambi et al. (2006)
Mycobacterium porcinumIC and IK survival (Ap Linc-Ap1)Adekambi et al. (2006)
Mycobacterium scrofulaceumIC multiplication and IK survival (Tp ATCC 30202)Strahl et al. (2001)
Mycobacterium simiaeIC and IK survival (Ap Linc-Ap1), IC survival (Ac)Krishna Prasad & Gupta (1978); Adekambi et al. (2006)
Mycobacterium smegmatisIC and IK survival (Ap Linc-Ap1), IC survival (Ac)Krishna Prasad & Gupta (1978); Adekambi et al. (2006)
Mycobacterium szulgaiIC and IK survival (Ap Linc-Ap1)Adekambi et al. (2006)
Mycobacterium ulceransIC survival (Ac, Ap)xKrishna Prasad & Gupta (1978); Eddyani et al. (2008)
Mycobacterium xenopiIC multiplication and IK survival (Ap Linc-Ap1), coculture (Ac ATCC 30010)xDrancourt et al. (2006); Thomas et al. (2006a, b)
Ochrobactrum anthropiCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Pantoea agglomeransCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Pasteurella multocidaIC multiplication (Ap ATCC 50372)Hundt & Ruffolo (2005)
Porphyromonas gingivalisIC multiplication (Ac ATCC 30234)Wagner et al. (2006)
Prevotella intermediaIC multiplication (Ac ATCC 30234)Wagner et al. (2006)
Providencia alcalifaciensCoculture with partial cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Providencia rettgeriCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Pseudomonas aeruginosaIC multiplication (Ap ATCC 30461, Echinamoeba sp.)xMichel et al. (1995a); Hwang et al. (2006)
Pseudomonas alcaligenesIC multiplication (Ap Linc-Ap1)Evstigneeva et al. (2009)
Pseudomonas fluorescensCoculture with cell lysis (Ap Linc-Ap1, Ac ATCC 30010)Pagnier et al. (2008); Thomas et al. (2008)
Pseudomonas putidaCoculture with partial cell lysis (Ap Linc-Ap1, Ac ATCC 30010)Pagnier et al. (2008); Thomas et al. (2008)
Rahnella aquatilisCoculture and cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Ralstonia pickettiiIC multiplication (Ac, Naegleria lovaniensis ATCC 30808)xMichel & Hauroder (1997)
Rhodococcus equiCoculture (Ac ATCC 30010)xThomas et al. (2008)
Rhodococcus erythropolisCoculture (Ac ATCC 30010)Thomas et al. (2008)
Rothia dentocariosaIC multiplication (Ap Linc Ap-1)Evstigneeva et al. (2009)
Salmonella typhimuriumIC multiplication (Ap Linc-Ap1)xxxxGaze et al. (2003)
Serratia ficariaCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Serratia marcescensCoculture with complete cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Serratia plymuthicaIC multiplication (Ap Linc Ap-1)Evstigneeva et al. (2009)
Serratia proteamaculansCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Shigella boydiiCoculture without cell lysis (Ap Linc-Ap1)xxxPagnier et al. (2008)
Shigella dysenteriaeIC multiplication (Ac ATCC 30234)xxxSaeed et al. (2009)
Shigella sonneiIC multiplication (Ac ATCC 30234 and 30010)xxxJeong et al. (2007a); Saeed et al. (2009)
Staphylococcus aureus (MRSA)IC multiplication (Ap), IK survival (?)xxMarciano-Cabral (2004); Huws et al. (2006)
Stenotrophomonas maltophiliaCoculture with complete cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
Streptococcus pneumoniaeIC multiplication (A. polyphaga Linc-Ap1)xEvstigneeva et al. (2009)
Tatlockia micdadei IC multiplication and IK survival (Ap Linc Ap-l, Ac, Hv, Tp, A. culbertsoni, H. cantabrigiensis)xFields et al. (1986); Fallon & Rowbotham (1990)
Vibrio choleraeIC multiplication and IK survival (Ac ATCC 30234, Naegleria gruberi 1518/1e)xxThom et al. (1992); Abd et al. (2005)
Vibrio parahaemolyticusCoculture (Ac ATCC 30234)xxxLaskowski-Arce & Orth (2008)
Yersinia enterocoliticaIC survival (Ac ATCC 30234)xxKing et al. (1988)
Yersinia pestisIC survival (Hartmannella rhysodes)xxNikul'shin et al. (1992)
Yersinia pseudotuberculosisCoculture without cell lysis (Ap Linc-Ap1)Pagnier et al. (2008)
  • Bacterial species classified as high priority infectious microorganisms in several official lists are indicated (Ecke et al., 2005). Interactions with the soil amoeba Dictyostelium discoideum and the ciliate Tetrahymena pyriformis are also reported but were not taken into account for bacteria–FLA interactions calculation.

  • Ac, Acanthamoeba castellanii; Ap, Acanthamoeba polyphaga; Dd, Dictyostelium discoideum; Hv, Hartmannella vermiformis; Tp, Tetrahymena pyriformis; IC, intracellular; IK, intracyst.

  • * Amoebal species that were demonstrated to support growth of Legionella pneumophila: Acanthamoeba castellanii, Acanthamoeba culbertsoni, Acanthamoeba griffini, Acanthamoeba lenticulata, Acanthamoeba palestinensis, Acanthamoeba royreba, Acanthamoeba polyphaga, Balamuthia mandrillaris, Dictyostelium discoideum, Echinamoeba exundans, Hartmannella cantabrigiensis, Hartmannella vermiformis, Naegleria australiensis, Naegleria fowleri, Naegleria jadini, Naegleria lovaniensis, Naegleria gruberi, Platyamoeba placida, Saccamoeba spp., Vahlkampfia jugosa, Vexillifera spp. and Willaertia spp.

  • Now ‘Legionella micdadei’. x, Bacterial species listed in threat list (Ecke et al., 2005).?, not clearly demonstrated.

Other reasons why various bacterial species listed in the CCL3 list have not been tested for their interaction with amoebae includes a limited number of laboratories investigating these interactions and the theoretical assumption that many of these interactions would not make sense. It is almost counterintuitive that obligate anaerobic bacteria such as Clostridium spp. (16 species in the CCL3 list), Prevotella spp. (20 species) and Porphyromonas spp. (seven species) might survive or even grow within amoebae. However, intra-amoebal persistence and/or multiplication have been demonstrated for several obligate anaerobic species that belong to these genera: Clostridium frigidicarnis (Pagnier et al., 2008), Porphyromonas gingivalis and Prevotella intermedia (Wagner et al., 2006). Persistence and multiplication in Acanthamoeba spp. has also been demonstrated for Mobiluncus curtisii, an obligate anaerobe that causes vaginosis (Tomov et al., 1999). Interestingly, other species that require microaerobic conditions for cultivation in artificial media, such as Campylobacter spp. and Helicobacter pylori, are also able to persist and grow within amoebae (Winiecka-Krusnell et al., 2002; Axelsson-Olsson et al., 2007). Thus, it seems that protozoa provide an ecological niche that can respond to the needs of various, and very different, bacterial species.

It should also be emphasized that various ARB bacterial species that are not in the CCL3 list have been demonstrated or are strongly suspected of being significant pathogens for humans and/or animals (Table 2). They consist mainly in several mycobacterial species (listed in Herdman & Steele, 2004), Alpha- and Gammaproteobacteria and new Chlamydiales species (Horn et al., 2008; Greub et al., 2009).

View this table:
Table 2

Described interactions of highly suspected pathogenic bacterial species with FLA

Bacterial speciesDescribed interaction with protozoaReferences
Parachlamydia acanthamoebaeIC multiplication (various Acanthamoeba spp.)Amann et al. (1997); Marrie et al. (2001); Borel et al. (2007); Baud et al. (2008)
Protochlamydia amoebophilaIC multiplication (Acanthamoeba UWC1, Dd AX2)Fritsche et al. (2000); Skriwan et al. (2002); Haider et al. (2008)
Protochlamydia naegleriophilaIC multiplication (various FLA species)Michel et al. (2000a); Casson et al. (2008)
Simkania negevensisIC multiplication (N. clarki, B. mandrillaris, Ac, Hv), IC multiplication and IK survival (Ac Linc-Ap1)Kahane et al. (1998); Kahane et al. (2001); Lieberman et al. (2002); Michel et al. (2005)
Waddlia chondrophilaIC multiplication (various FLA species)Dilbeck et al. (1990); Henning et al. (2002); Dilbeck-Robertson et al. (2003); Michel et al. (2004); Baud et al. (2007); Haider et al. (2008)
Agrobacterium tumefaciensIC multiplication (Ap Linc-Ap1)Evstigneeva et al. (2009)
Bosea sp.Coculture (Ap Linc-Ap1, Ac ATCC30010)La Scola et al. (2003b); Berger et al. (2006); Thomas et al. (2007)
Bradyrhizobium sp.Coculture (Ap Linc-Ap1, Ac ATCC30010)La Scola et al. (2003b); Berger et al. (2006), Thomas et al. (2006a, b)
Rasbo bacteriumCoculture (Ap Linc-Ap1, Ac ATCC30010)La Scola et al. (2003b); Berger et al. (2006); Thomas et al. (2006a, b)
Roseomonas gilardiiCoculture (Ac ATCC30010)Thomas et al. (2006a, b)
Aeromonas eucrenophilaPartial cell lysis (Ap Linc-Ap1)Evstigneeva et al. (2009)
Aeromonas salmonicidaIC multiplication (Ap Linc-Ap1)Evstigneeva et al. (2009)
Legionella-like amoebal pathogensIC multiplication (various FLA species depending on LLAP isolate considered)Birtles et al. (1996); Birtles et al. (1997); McNally et al. (2000)
Pseudomonas mendocinaIC multiplication (Ap Linc-Ap1)Evstigneeva et al. (2009)
Klebsiella variicolaComplete cell lysis (Ap Linc-Ap1)Evstigneeva et al. (2009)
CFB group bacteria
Chryseobacterium indologenesCoculture and cell lysis (Ap Linc AP-1)Pagnier et al. (2008)
Sphingobacterium multivorumIC multiplication (Ap Linc-Ap1)Evstigneeva et al. (2009)
High GC% Gram+
Kocuria kristinaeCoculture (Ac ATCC 30010)Thomas et al. (2008)
Mobiluncus curtisiiIC multiplication (A. culbertsoni A1 and others Acanthamoeba spp.)Tomov et al. (1999)
Mycobacterium bohemicumIC and IK survival (Ap Linc-AP1)Adekambi et al. (2006)
Mycobacterium goodiiIC and IK survival (Ap Linc-AP1)Adekambi et al. (2006)
Mycobacterium immunogenumIC and IK survival (Ap Linc-AP1)Adekambi et al. (2006)
Mycobacterium lentiflavumIC and IK survival (Ap Linc-AP1)Adekambi et al. (2006)
Mycobacterium mageritenseIC and IK survival (Ap Linc-AP1)Adekambi et al. (2006)
Mycobacterium septicumIC and IK survival (Ap Linc-AP1)Adekambi et al. (2006)
Mycobacterium tusciaeIC and IK survival (Ap Linc-AP1)Adekambi et al. (2006)
Staphylococcus pasteuriIC multiplication (Ap Linc-Ap1)Evstigneeva et al. (2009)
  • Ac, Acanthamoeba castellanii; Ap, Acanthamoeba polyphaga; Dd, Dictyostelium discoideum; Hv, Hartmannella vermiformis; IC, intracellular; IK, intracyst.

Interactions with fungi and parasitic protozoa

In addition to bacteria, other microorganisms pathogenic to humans have also been reported to interact with amoebae in a way that can promote their survival and their transmission to susceptible hosts (Table 3). Cryptococcus neoformans is able to grow inside (in the case of C. neoformans var. neoformans) or in the presence of (with C. neoformans var. gattii) A. castellanii ATCC 30324 (Steenbergen et al., 2001; Malliaris et al., 2004), whereas nonvirulent acapsular strains and phospholipase mutants cannot. Importantly, it has also been demonstrated that the murine virulence of C. neoformans was enhanced by passage through live Dictyostelium discoideum culture (Steenbergen et al., 2003). Similar studies with other fungi also demonstrated their capacity to use amoebae after ingestion (Sporothrix schenckii and Histoplasma capsulatum) or to exert extracellular cytotoxic effect (Blastomyces dermatitidis), leading in both cases to the release of nutrients from killed amoebae that might then be used for fungal growth (Steenbergen et al., 2004). It has also been demonstrated that H. capsulatum virulence to mice was enhanced after growth in the presence of A. castellanii ATCC 30324. Other studies suggested that Acanthamoeba spp. could assist in disseminating Cryptosporidium parvum oocysts; however, it was not clearly demonstrated that ingested forms were still alive (Stott et al., 2003; Gomez-Couso et al., 2006). The parasite Toxoplasma gondii was recently demonstrated to survive for up to 2 weeks in A. castellanii without reducing the infectivity and the pathogenicity of oocysts (Winiecka-Krusnell et al., 2009).

View this table:
Table 3

Described interactions of microorganisms other than bacteria with FLA

MicroorganismDescribed interaction with protozoaReference
Acanthamoeba polyphaga MimivirusIC multiplication (various Acanthamoeba spp. including Ap Linc-Ap1)La Scola et al. (2003a); Suzan-Monti et al. (2006)
Coxsackie virus b3IC and IK survival (Ac), IC survival (Tp)Teras et al. (1988); Mattana et al. (2006)
AdenovirusIC survival (various Acanthamoeba isolates)Lorenzo-Morales et al. (2007)
EchovirusesEnhanced survival (Ac strain Neff)Danes & Cerva (1981)
Cryptococcus neoformansIC multiplication (Ac ATCC 30324, Dd AX-4)Steenbergen et al. (2001); Steenbergen et al. (2003)
Blastomyces dermatitidisCoculture with cell lysis (Ac ATCC 30324)Steenbergen et al. (2004)
Sporothrix schenckiiCoculture with cell lysis (Ac ATCC 30324)Steenbergen et al. (2004)
Histoplasma capsulatumCoculture with cell lysis (Ac ATCC 30324)Steenbergen et al. (2004)
Toxoplasma gondiiIC survival (Ac)Winiecka-Krusnell et al. (2009)
  • * Indirect proof because only viral DNA was detected from axenic Acanthamoeba isolates.

  • Ac, Acanthamoeba castellanii; Ap, Acanthamoeba polyphaga; Dd, Dictyostelium discoideum; Tp, Tetrahymena pyriformis; IC, intracellular; IK, intracyst.

Interactions with viruses

Significantly, it has been demonstrated that several human pathogenic enteroviruses interact with protozoa. Enhanced survival of echoviruses bound to amoebae has been reported (Danes & Cerva, 1981). The study by Mattana et al. (2006) reported that coxsackie virus b3 was able to survive within A. castellanii trophozoites, and detected significant infectious virus after a 6 months cycle of encystment and excystment; to our knowledge, there is no other report of a well-known pathogenic virus being able to survive inside FLA. It has only been recently suggested that Acanthamoeba spp. could act as reservoirs for adenoviruses, with viral DNA (mostly adenovirus type 2) being detected from 34 of 236 (14.4%) environmental amoebal isolates (Lorenzo-Morales et al., 2007). A newly described giant virus ‘Acanthamoeba polyphaga Mimivirus’ has also been isolated from an environmental Acanthamoeba strain (La Scola et al., 2003a) and might be involved in pneumonia (La Scola, 2005; Raoult et al., 2006).

Potential impact on public health

From a public health perspective, it is important to note that a large number of bacterial species interacting successfully with protozoa are included in various lists of pathogenic microorganisms created by several medical and governmental organizations (see Ecke et al., 2005 for a complete review of these lists) (Table 1). Furthermore, 20 bacterial species in the CCL3 list have been shown to survive in amoebal cysts; these are mainly mycobacteria (16 species) interacting with Acanthamoeba Linc-Ap1 (Adekambi et al., 2006; Thomas & McDonnell, 2007). Other species include Francisella tularensis (Abd et al., 2003), L. pneumophila (Kilvington & Price, 1990), Legionella micdadei (Fallon & Rowbotham, 1990) and Vibrio cholerae (Thom et al., 1992). This may be of importance because when bacteria internalized within trophozoites have already increased resistance to biocides (King et al., 1988; Howard & Inglis, 2005), this level of resistance increases significantly for bacteria internalized in amoebal cysts (Kilvington & Price, 1990). Some species impair encystation once inside their amoebal host (Garcia et al., 2007). An impressive picture of amoebal cysts heavily infected with Staphylococcus aureus has been published by Marciano-Cabral (2004) (Fig. 2), although there is no study on the role of cysts in survival and protection of S. aureus in this paper. Additional studies testing intracyst survival of waterborne pathogens such as Pseudomonas spp. and Burkholderia spp. are also warranted. Resuscitation of bacteria existing in the ‘viable-but-nonculturable’ (VBNC) state due to starvation or biocide treatment has been demonstrated for L. pneumophila (Steinert et al., 1997) and Campylobacter jejuni (Axelsson-Olsson et al., 2005) after coculture with FLA, and might thus be investigated for other bacterial species.

Figure 2

An Acanthamoeba cyst containing Staphylococcus aureus (arrow). Scale bar=1 μm. Reproduced from Marciano-Cabral (2004) with kind authorization of author and editors.

Amoebal coculture as a tool to isolate new microorganisms

Because of the similarities between mechanisms allowing microorganisms to escape phagocytosis and/or digestion by FLA and the mechanisms allowing these same microorganisms to escape phagocytosis and/or digestion by macrophages, FLA have been proposed as a tool to recover potentially new pathogenic species from various environments (Greub & Raoult, 2004). The method is relatively straightforward and has been extensively described in previous publications (La Scola et al., 2000; Greub & Raoult, 2004; Thomas & McDonnell, 2007). Briefly, it mainly consists in seeding samples onto axenic amoebal cells and observing whether microorganisms develop in amoebae using appropriate staining methods such as Gimenez staining. If microorganisms are observed, and depending on the results obtained with specific colorations such as Ziehl–Neelsen staining for mycobacteria, samples can be subcultured on specific media or on new axenic amoebae in case of obligate intracellular organisms (Fig. 3). PCR and sequencing with universal primers targeting 16S rRNA- or RpoB-encoding genes can be used to identify isolated bacteria. When using classical staining and microscopic methods to observe potentially infected amoebae, one should remember that FLA have also been found to be infected by small eukaryotic cells (microsporidies) (Hoffmann et al., 1998; Michel et al., 2000b) that can resemble bacteria, and that giant viruses infecting amoebae can resemble gram-positive cocci (La Scola et al., 2003a).

Figure 3

Proposed method used to recover FLA and intracellular microorganisms from various samples. Reproduced from Thomas et al. (2008) and Thomas & McDonnell (2007). *Gimenez-staining of Candidatus‘Criblamydia sequanensis’ in Acanthamoeba castellanii ATCC 30010 (unpublished data); **Zielh–Neelsen staining of Mycobacterium fluoranthenivorans in A. castellanii ATCC 30010 (Loret et al., 2008a).

Using this technique with complex environmental samples generally leads to retrieval of a wide variety of microorganisms, including Alpha-, Beta- and Gammaproteobacteria species (La Scola et al., 2000; Thomas et al., 2006b; Pagnier et al., 2008), members of the CytophagaFlavobacteriumBacteroides group (Horn et al., 2001; Pagnier et al., 2008; Evstigneeva et al., 2009), Actinomycetales species (Wang et al., 2006; Pagnier et al., 2008, 2009), bacilli and clostridii (Pagnier et al., 2008; Evstigneeva et al., 2009) and new Chlamydiales (Amann et al., 1997; Horn et al., 2000; Thomas et al., 2006a; Corsaro et al., 2009). Attempts to use amoebae to isolate microorganisms directly from clinical samples are less common. Legionella pneumophila was isolated from feces and sputa of patients with community-acquired legionnaires' disease (Rowbotham et al., 1998), Legionella anisa from sputum of an immunocompromised man with pneumonia for whom conventional diagnostic tests were negative (La Scola et al., 2001). A new mycobacterial species, Mycobacterium massiliense, was isolated from the sputum and bronchoalveolar fluid of a patient with hemoptoic pneumonia by plating on axenic media and amoebal coculture with A. polyphaga (Adekambi et al., 2004). Mura et al. (2006) used cocultivation with A. polyphaga to examine gut samples from patients with various intestinal disorders; after cocultivation, they were able to detect M. avium ssp. paratuberculosis from 13 of 39 coculture samples using PCR and in situ hybridization (Mura et al., 2006). Initial surgical samples were all Ziehl–Neelsen negative whereas auramine–rhodamine staining detected mycobacteria in six of the 13 coculture samples demonstrated positive by molecular methods, suggesting that in these cases acid-fast phenotype changed from negative to positive during incubation with amoebae. Of note, albeit Chlamydia-like bacteria are thought to be responsible for pneumonia and miscarriages, to our knowledge, they have never been isolated from clinical samples using the amoebal coculture method; indirect evidence of their role in these infections was provided using only molecular and serological methods. However, the first described Parachlamydia isolate was recovered from an Acanthamoeba sp. cultivated from the nasal mucosa of a female volunteer (Michel et al., 1994; Amann et al., 1997), suggesting that amoebae and their intracellular host may colonize humans in close proximity of potential infection sites. A new Alphaproteobacteria species proposed as ‘Rhodobacter massiliensis’ was isolated from the nose of a patient with aspiration pneumonia using amoebal coculture with A. polyphaga (Greub & Raoult, 2003b). Authors used the same method in another study and isolated seven gram-negative bacteria (Alpha- and Betaproteobacteria, bacteria belonging to the Bacteroides–Cytophaga–Flexibacter group) from the nose of patients presenting severe concomitant viral and bacterial infections for most of them (Greub et al., 2004).

Thus, amoebal coculture can be considered as a valuable method to isolate fastidious species from complex environmental samples as well as from clinical samples. However, as described above, the main limitation of this technique might be the restricted host range that was described for several intracellular species (Michel et al., 2004, 2005), including clearly demonstrated pathogenic species (Dey et al., 2009). Further optimization of the method is thus needed and could be achieved by proposing the concomitant or sequential use of a range of selected amoebal species/strains, in parallel with a shell-vial culture assay using mammalian cell lines for clinical samples (Gouriet et al., 2005).

Amoebal survival strategies to inimical factors

Upon exposure to detrimental conditions, both physical (temperature changes, UV light and radiation) or chemical (e.g. pH changes, exposure to biocide and starvation) trophozoites can undergo encystation (Greub & Raoult, 2004). Cysts can survive for many years in the environment; Mazur et al. (1995) showed that 14 of 17 encysted acanthamoebal isolates stored at +4 °C in water survived for 24 years without apparent loss of virulence in the mouse model. Acanthamoebal cysts have also been reactivated after storage for >20 years in a completely dry environment (Sriram et al., 2008). Encystation is a relatively rapid process that can be divided into a number of development stages (Table 4) (Chavez-Munguia et al., 2005, 2009). During encystation, the forming cysts become resistant to a number of agents and these events can be used as markers of encystations progress (Turner et al., 2000a, b; Lloyd et al., 2001). Pre-emergent A. castelanii cysts (committed to excystation) have been found to be less resistant than mature cysts when challenged with chlorhexidine (CHA) or polyhexamethylene biguanide (PHMB), although trophozoites were still more susceptible (Khunkitti et al., 1998b). Several proteins have been demonstrated to be secreted mainly during encystation: this includes but is not limited to the proteins necessary for cellulose synthesis, several cyst cell wall proteins and polyphenol oxidase (Chen et al., 2004). Preliminary studies also demonstrated that gene expression is modulated during encystment, with upregulation of genes coding for cyst-specific protein 21, protein kinase C, proteasome, heat shock protein, various proteinases, cullin 4, autophage protein 8 and ubiquitin-conjugating enzymes (Moon et al., 2008). Cysts are a dehydrated structure with a double wall composed of cellulose and relatively small numbers of proteins (Turner et al., 2000b; Lloyd et al., 2001). The outer ectocyst wall is composed mainly of protein and lipid-containing materials (Bauer et al., 1967; Neff & Neff, 1969; Rubin et al., 1976) and its appearance is fibrillar in nature, associated with ill-defined amorphous substances (Bowers & Korn, 1969). In contrast, the inner endocyst wall contains cellulose and its structure appears to be composed of fine fibrils embedded in a granular matrix (Bowers & Korn, 1969). Both walls meet at ostioles, covered by opercula (Fig. 4). The exact composition and morphological aspect of the cyst wall, and notably the spatial separation between endocyst and ectocyst (Smirnov & Michel, 1999), may vary between species and strains but also depends upon the composition of the media used during encystation (Griffiths & Hughes, 1969; Stratford & Griffiths, 1971; Chagla & Griffiths, 1974; De Jonckheere & Brown, 2005). Such a media effect accounts in part for discrepancies in inactivation data between studies (Kilvington & Anger, 2001; Hughes et al., 2003). Interestingly, it has also been demonstrated recently that Acanthamoeba strains lose their ability to encyst synchronously after prolonged axenic culture, suggesting that in the ‘perfect’ environments FLA downregulate genes that are no longer required for survival (Kohsler et al., 2008).

View this table:
Table 4

Encystment and excystment stages

InductionAmeboidDegradation of cellular componentsNeff et al. (1964); Neff & Neff (1969)
Immature cystSphericalCell wall synthesis – first cell wall layer observedWeisman et al. (1976)
Mature cystSphericalSynthesis of second layer wallWeisman et al. (1976)
Initiation stageCytoplasm of the cyst loses its granular appearance and larger globules are observedMattar & Byers (1971)
Pre-emergent stageDetachment of the amoeba from the endocyst wall and reappearance of a contractile vacuoleWeisman et al. (1976)
Emergence stageDigestion of an operculumAmoeba passes out of the cyst wallsWeisman et al. (1976)
Figure 4

Structure of an Acanthamoeba castellanii ATCC 30010 cyst obtained after 5–7 days incubation of trophozoites at 33°C in Page's modified Neff's amoeba saline (magnification 4500 ×). Endocyst (Endo), ectocyst (Ecto), ostioles (Os), operculum (Op) and mitochondria (M) are indicated. In this experiment, amoebal trophozoites were used to grow the new Chlamydiales species: Criblamydia sequanensis (Thomas et al., 2006a) that can be observed as star-shaped extracellular bacteria in this picture.

Because cysts are the major contributor to protozoal resistance to biocides, encystation is a crucial step in producing resistance and the factors inducing encystation need to be understood. A number of chemical biocides have been shown to induce the formation of cysts in Acanthamoeba spp., including diamidines such as diminazene aceturate and pentamidine isethionate and CHA at a low concentration (Byers et al., 1991; Chomicz et al., 2005). A study investigating the development of resistance during encystation highlighted that exponentially growing trophozoites of A. castellanii were more susceptible to CHA and PHMB than pre-encystation trophozoites (i.e. trophozoites committed to encystation) or mature (7 days old) cysts (Khunkitti et al., 1998b). Indeed, other investigations showed a correlation between encystation, notably the synthesis of the cellulose endocyst wall, and the development of resistance to different cationic biocides and oxidizing agents (Turner et al., 2000a, b; Lloyd et al., 2001). The development of the cyst wall has been shown to correspond to a decrease in CHA and PHMB adsorption during encystment in A. castellanii (Turner et al., 2004).

Overall, little is known about the mechanisms of resistance of Acanthamoeba spp. to biocides. The double cyst wall is presumed to represent a permeability barrier and an intrinsic resistance mechanism for CHA and PHMB (McDonnell & Russell, 1999) as well as various diamidines (Perrine et al., 1995). Uptake isotherms of cysts challenged with CHA and PHMB (Khunkitti et al., 1997b), as well as the kinetics of biocide action (Perrine et al., 1995; Khunkitti et al., 1996), provide possible evidence for penetration barrier type resistance mechanisms. Additionally, the metabolically dormant nature of the cyst may negate some of the effects of biocides such as the aromatic diamidines, which inhibit polyamine synthesis and mitochondrial function (Byers et al., 1991). Superior homologues of these biocides such as hexamidine and octamidine were found to be more efficient against trophozoites and cysts. These compounds were found to diffuse better through the cytoplasmic membrane and cell wall although their lethal activity was linked to the protonated amide groups rather than their longer side chain (Perrine et al., 1995).

The resistance of Acanthamoeba spp. to chlorine has been described for 30 years (De Jonckheere & van de Voorde, 1976). Comparatively, there have been few reports of protozoal resistance to PHMB (Murdoch et al., 1998; Wysenbeek et al., 2000). Chlorine and CRAs are heavily used in the water industry, while PHMB is used in both the water industry and the medical contact lens industries.

Although the relative biocide sensitivity for most commonly isolated FLA genera is more or less predictable (Acanthamoeba spp. cysts>Hartmannella spp. cysts>Naegleria spp. cysts), it is more difficult to classify species according to this criterion. Concerning Acanthamoeba genus, species that were historically defined based on morphological criteria were later divided into 15 genotypes based on rRNA gene sequences (Visvesvara et al., 2007). Genotype T4, which includes A. castellanii and A. polyphaga, is the most frequently encountered in human infection (Booton et al., 2002, 2005), but the authors have demonstrated that isolates belonging to suspected nonpathogenic genotypes can be more resistant to disinfectants (Shoff et al., 2007). Furthermore, and similar to bacteria, different strains of FLA but of the same taxonomic species may present rather different susceptibilities to biocides (Jeong et al., 2007b). This was observed by Srikanth & Berk (1993) with the Acanthamoeba hatchetti ATCC isolate being more susceptible to cooling tower biocides than an environmental isolate that belonged to the same species, the latter even being speculated to grow using biocides as a carbon source. The same authors observed in a further study that isolates can adapt to biocides and reported emerging cross-resistance in amoebae following treatment with subinhibitory levels of biocides. Such a phenomenon is of particular interest considering the common but controversial practice of rotating biocides for water treatment (Srikanth & Berk, 1994).

Efficacy of chemical biocides and physical agents against amoebae

Although the lethal effects of biocides against trophozoites and cysts have been described in a number of amoebae (Illingworth & Cook, 1998; Lindquist et al., 1998; Turner et al., 1999, 2004), our understanding of the interaction between cysts or trophozoites and biocides is limited. The mechanisms of biocidal action against bacteria and other organisms are often presumed to be similar to those upon comparable structures present in amoebae. Some studies aiming to elucidate the effect of cationic agents against Acanthamoeba spp. have been published. A transmission electron microscope study has demonstrated both structural and cytoplasmic membrane damage to A. castellanii trophozoites treated with biguanides (CHA and PHMB) and this is comparable to the effects observed during similar studies using bacteria (Khunkitti et al., 1998a, 1999). Both CHA and PHMB have been shown to induce pentose leakage in trophozoites at low concentrations, which is also thought to be an indication of cytoplasmic membrane damage (Khunkitti et al., 1997b; Turner et al., 2004). Biocides generally lack selective toxicity and many target sites are likely involved in their lethal effects (Khunkitti et al., 1998a). More specific interactions have been sometimes described. For instance, diamidines have been shown to inhibit S-adenosylmethionine decarboxylase and consequently polyamine synthesis in trophozoites of Acanthamoeba (Byers et al., 1991); it has also been demonstrated that they can affect mitochondria (Akins & Byers, 1980; Sands et al., 1985) and nucleic acids (Lindquist et al., 1998).

The efficacy of a limited number of biocides, principally biguanides, quaternary ammonium compounds (QACs), chlorine and CRAs, and oxidizing agents, have been investigated in studies focusing on medical application (contact lenses) and water disinfection. Some discrepancies in results can be observed in trophocidal and cysticidal activity, and these will be discussed later.

Activity of cationic biocides

Trophozoites of Acanthamoeba spp., A. castellanii, A. polyphaga and Acanthamoeba culbertsoni have been found to be susceptible to low concentrations (0.005–0.006%) of CHA (Penley et al., 1989; Silvany et al., 1990; Connor et al., 1991; Hugo et al., 1991). Brandt et al. (1989) reported CHA (0.005%) in formulation with thimerosal 0.001% (an organomercury compound) and/or EDTA 0.1% to be effective against acanthamoebal cysts within 24 h.

PHMB was described as trophocidal and cysticidal against A. castellanii (Silvany et al., 1991; Burger et al., 1994; Khunkitti et al., 1996), A. polyphaga (Silvany et al., 1991; Burger et al., 1994) and clinical isolates of Acanthamoeba spp. associated with keratitis (Elder et al., 1994; Hay et al., 1994; Seal et al., 1996; Kim & Hahn, 1999). However, very low concentrations of PHMB (0.00005%), polyaminopropyl biguanide (0.00005%) and the QAC polyquaternium-1 (0.055–0.001%) were reported to be inactive against A. castellanii (Penley et al., 1989; Davies et al., 1990; Silvany et al., 1990; Zanetti et al., 1995; Cengiz et al., 2000), A. polyphaga (Penley et al., 1989; Davies et al., 1990; Silvany et al., 1990) and A. culbertsoni (Connor et al., 1991). Several investigations showed that the concentration of polymeric biguanides in multipurpose solutions commercially available was not cysticidal within the manufacturer's recommended contact time (Buck et al., 1998; Kilvington et al., 1998; Niszl & Markus, 1998). Indeed, a prolonged contact time of 24 h was necessary for significant cysticidal activity of PHMB (0.02%) against A. castellanii (Aksozek et al., 2002). The QAC benzalkonium chloride, at a concentration of 0.04%, was shown to be efficacious against trophozoites and cysts of A. castellanii (Silvany et al., 1991; Zanetti et al., 1995; Khunkitti et al., 1996) and A. polyphaga (Silvany et al., 1991). Much lower concentrations (0.003–0.004%) were found to be noncysticidal (Penley et al., 1989; Connor et al., 1991; Hugo et al., 1991), although Turner et al. (2000b) showed that the minimum cysticidal concentration for A. castellanii was 0.004%.

Oxidizing agents

Chlorine at 10 mg L−1 (free and combined) was reported to be effective against Hartmannella vermiformis cysts after 30 min exposure (Kuchta et al., 1993), but it is clearly ineffective against acanthamoebal cysts because they can resist exposure to 100 mg L−1 chlorine for 10 min (Storey et al., 2004) or 50 mg L−1 for 18 h (Kilvington & Price, 1990). Limited activity of chlorine was also reported against B. mandrillaris cysts and trophozoites (Siddiqui et al., 2008), whereas Chang (1978) reported that exposure to 1–7 mg L−1 free chlorine (for 5–30 min) was cysticidal against Naegleria spp. and that cyst inactivation conformed to a first-order kinetic. Differences in cysticidal activity to chlorine, and also to chlorine dioxide and ozone, have been reported with Acanthamoeba spp. being less susceptible than Naeglaria spp. (Cursons et al., 1980). Monochloramine at 3.9 mg L−1 killed Naegleria lovanensis cysts within 1 h (Ercken et al., 2003). However, this biocide has also been demonstrated to induce VBNC state in L. pneumophila that could then be resuscitated by coculture with A. castellanii (Alleron et al., 2008).

Hydrogen peroxide formulations are commonly used for the disinfection of contact lenses and the trophocidal and cysticidal efficacy of liquid hydrogen peroxide against Acanthamoeba spp. has been well studied (Brandt et al., 1989; Davies et al., 1990; Silvany et al., 1990, 1991; Niszl & Markus, 1998; Hughes & Kilvington, 2001; Aksozek et al., 2002). Efficacy has been reported to depend upon the type of formulation and contact time (Hughes & Kilvington, 2001), with two studies reporting commercial products containing 3% hydrogen peroxide to be ineffective against Acanthamoeba spp. within the manufacturer's recommended contact time of 30 min (Ludwig et al., 1986; Zanetti et al., 1995), but they were cysticidal after 4 h (Aksozek et al., 2002). Hydrogen peroxide as a gas has been demonstrated to be effective against A. polyphaga and A. castellanii cysts (Thomas & McDonnell, 2008).

When tested against Acanthamoeba and Naegleria spp. trophozoites, chlorine dioxide in water was efficacious after treatment for 30 min at concentrations of approximately 2 mg L−1 (Naegleria spp.) or 3 mg L−1 (Acanthamoeba spp.) (Cursons et al., 1980). However, the same active had only limited effect on A. polyphaga cysts exposed to 5 mg L−1 for 60 min (Loret et al., 2008b), and continuous injection in water pipes at 0.5 mg L−1 could not completely inactivate FLA (Thomas et al., 2004).

Ozone, the most potent oxidizing agent, has been shown to considerably reduce the FLA population in a drinking water plant (Thomas et al., 2008) and to be an efficient cysticidal against A. polyphaga (Loret et al., 2008b). However, its use is limited to the treatment of a limited volume of circulating water and it has no residual activity against FLA (Thomas et al., 2004).

Unformulated peracetic acid (PAA) was demonstrated to be effective against Acanthamoeba and Naegleria trophozoites after exposure to 15 mg L−1 for 2 h, but activity against Acanthamoeba cysts required longer incubation (18 h) and higher concentration (150 mg L−1). One commercial PAA-based product, when tested at room temperature, was shown to kill all trophozoites of A. polyphaga within 30 min, but failed to completely inactivate cysts within 24 h (Greub & Raoult, 2003a). Other FLA species seem more susceptible to inactivation by PAA, and cysticidal activity against N. lovanensis was observed after exposure to 5.33 mg L−1 for 1 h (Ercken et al., 2003).

Other biocides

The trophocidal and cysticidal activities of some other biocides have also been reported, but to a lesser extent. Iodine and bromine were found to have some cysticidal activity against N. fowleri (De Jonckheere & van de Voorde, 1976). These actives have generally poor efficacy against a number of protozoa (De Jonckheere & van de Voorde, 1976), and they have been proven to be inefficient against Acanthamoeba spp. cysts (Lim et al., 2000) under the conditions tested. Only a few studies reporting testing of aldehydes against FLA have been published. In the work by Aksozek et al. (2002), 10% formalin was reported to be cysticidal within 30 min against A. castellanii. Glutaraldehyde might be less active because a study focusing on A. polyphaga reported that a large proportion of trophozoites were still viable after exposure to a commercial formulation containing 2% glutaraldehyde for 30 min to 3 h (Greub & Raoult, 2003a). Isothiazolones consist of a range of biocides extensively used in the water industry, and as a consequence, their cysticidal activity, as well as their activity against intracellular bacterial pathogens, has been investigated. Cysts of environmental isolates of A. hatchetti and Cochliopodium bilimbosum were shown to be resistant to an isothiazolin derivative (5-chloro-2-methyl-4-isothiazolin-3-one) used according to the manufacturer's recommendations for the disinfection of cooling towers (Srikanth & Berk, 1993; Sutherland & Berk, 1996) and were shown to reproduce faster at a low concentration of the biocide (Srikanth & Berk, 1993). Diamidines have been shown to be trophocidal but they are not cysticidal (Osato et al., 1991; Perrine et al., 1995; Gray et al., 1996; Kim & Hahn, 1999; Wysenbeek et al., 2000). Although propamidine is poorly cysticidal, other homologous biocides such as hexamidine and octamidine appeared to have greater activity (Perrine et al., 1995). Interestingly, products containing 20% isopropyl alcohol were demonstrated to present good activity against A. polyphaga, A. castellanii and A. culbertsoni cysts (Penley et al., 1989; Connor et al., 1991; Aksozek et al., 2002). There are very few publications available on the efficacy of metallic salts against amoebae. Rohr et al. (2000) showed that the concentrations of copper (100 μg L−1) and silver (10 μg L−1) used within the limit of drinking water regulations were not active against H. vermiformis and the ciliate Tetrahymena pyriformis.

Combination of treatments

Owing to the low cysticidal activity of biocides when used over short contact times, combinations of biocides have been tested for their efficacy. A combination of hydrogen peroxide (3%) with catalase and potassium iodide (50 μM) significantly enhanced the cysticidal activity against A. polyphaga (Hughes et al., 2003); this may be due the increased oxidation potential of these mixtures. Sokmen et al. (2008) showed that the combination of silver-titanium dioxide activated by UV light, while inducing irreversible damage to the cell wall and intracellular structure of Giardia intestinalis cysts, was ineffective against cysts of A. castellanii. Heat pretreatment of H. vermiformis rendered them more susceptible to chlorine in the study by Kuchta et al. (1993). Likewise, although a number of publications have described the lack of efficacy of a range of QACs against environmental amoebal isolates, a combination of QACs with tributyltin neodecanoate (TBT/QAC) produced better activity (Srikanth & Berk, 1993; Sutherland & Berk, 1996). CHA (0.005%) was also found to be effective against cysts within 24 h (Brandt et al., 1989) or in combination with thiomersal (0.005%) within 6–9 h (Zanetti et al., 1995) or 6–24 h (Brandt et al., 1989).

Efficacy of physical agents, heat and irradiation against protozoa


Moist heat is generally considered to be trophocidal and cysticidal, although higher temperatures are needed to inactivate cysts. This difference in inactivation may arise as a consequence of dehydration of the cyst cytoplasm during cyst cell wall synthesis (Turner et al., 2000b; Lloyd et al., 2001).

Turner et al. (2000b) observed that trophozoites of A. castellanii were inactivated following a 30-min exposure at 46 °C, while a temperature of 56 °C was necessary to inactivate the same number of cysts. Aksozek et al. (2002) reported that a temperature of 65 °C for >5 min was cysticidal for A. castellanii, while Ludwig et al. (1986) reported that cysts of A. castellanii and A. polyphaga were inactivated only after exposure to moist heat at 80 °C for 10 min. However, the same exposure conditions did not completely inactivate thermotolerant Acanthamoeba spp. cysts in other studies (Storey et al., 2004). Balamuthia mandrillaris trophozoites may be more resistant to heat, because a temperature of 60 °C was not trophocidal within 60 min; however, a temperature of 80 °C maintained for at least 60 min did inactivate cysts (Siddiqui et al., 2008). Thermotolerant Hartmannella strains have been cultivated at 53 °C (Rohr et al., 1998), but cysts of H. vermiformis (environmental strain) were inactivated after exposure at 60 °C for 30 min (Kuchta et al., 1993). One Echinamoeba sp. strain has been shown to grow at 57 °C, but the cyst temperature resistance was not studied (Baumgartner et al., 2003). Naegleria spp. are generally considered more sensitive to heat (Chang, 1978; Rohr et al., 1998).

UV disinfection

While UV radiation is widely used for water disinfection, its efficacy against FLA has not been widely reported when compared with other protozoa such as C. parvum. Hijnen et al. (2006) provide a comprehensive overview of the ability of UV radiation to inactivate cysts and reported that Acanthamoeba spp. are highly UV resistant (Hijnen et al., 2006). UV exposure has good activity against trophozoites (Maya et al., 2003), but cysts are much more resistant. For example, Acanthamoeba spp. cysts were demonstrated to be resistant to exposure to UV-C at 253.7 nm, 1.1 mJ s−1 cm−2 (Hwang et al., 2004); A. castellanii cysts were shown to be resistant to UV-B irradiation (800 mJ cm−2) and B. mandrillaris cysts to exposure at 200 mJ cm−2 UV irradiation (Siddiqui et al., 2008). Others studies reported a 4 log10 reduction of A. polyphaga cysts after exposure to 40 mJ cm−2 (Loret et al., 2008a), suggesting that there might be some differences between strain and species sensitivities.

Exposure of A. polyphaga cysts to simulated global solar irradiance in water (85 mJ s−1 cm−2) did not achieve cyst inactivation when the temperature was kept <40 °C but gave better results at 45 °C (1.2 log10 reduction after 6 h), 50 °C (>3.6 log10 reduction after 6 h) and 55 °C (>3.3 log10 reduction after 4 h) (Heaselgrave et al., 2006). Sixty one percent of Naegleria gruberi cysts were still alive after UV irradiation with 21.6 mJ cm−2, and treating amoebae with inhibitors of DNA repair mechanisms improved UV irradiation efficacy (Hillebrandt & Muller, 1991). Photosensitized inactivation of Acanthamoeba palestinensis in the cystic stage has also been demonstrated by incubating cysts with tetracationic Zn(II)-phthalocyanine before exposure to 600–700-nm wavelength light sources (Ferro et al., 2006).


Physical removal by clarification and filtration processes has been proposed as an effective way to remove protozoa from drinking water (Loret et al., 2008a). Several studies demonstrated that a clarification step by sand filtration removed approximately 2–3 log10 FLA from water, with successive filtration on granular activated carbon (GAC) removing an additional 1–2 log10 (Hoffmann & Michel, 2001; Jeong & Yu, 2005; Thomas et al., 2008; Loret et al., 2008a). Of note, some of the FLA isolated in these studies harbored endosymbiotic bacterial species. Interestingly, most of the species that were recovered after the sand and GAC filters in these studies belonged to Hartmannella and Echinamoeba genera, suggesting that these small amoebae (cyst diameter generally <10 μm) are more likely to spread through drinking water treatment processes than Acanthamoeba and Naegleria spp. (cyst diameter generally >10 μm). Nanofiltration of water as a final stage of drinking water production is also used to remove microbial contaminants; however, membrane fouling is a major constraint in this process (Her et al., 2007), and it does not prevent recolonization downstream in a drinking water network. Terminal point-of-use filters are also effective (Exner et al., 2005). In our view, however, additional studies are needed to investigate the potential growth of biofilms and amoebae on these supports, as well as intracellular growth and release of bacteria that can pass through 0.2-μm filters (Hahn et al., 2004; Silbaq et al., 2009) that might then grow in the presence of FLA, such as Microbacterium spp. (Thomas et al., 2008).

Other physical treatments

Pulsed electric fields have been used against Naegleria lovaniensis trophozoites (Vernhes et al., 2002); however, this is not a particularly challenging organism. Acanthamoeba castellanii cysts were resistant to 2500 Gy of γ irradiation (Aksozek et al., 2002), and N. gruberi cysts have been demonstrated to resist X-ray irradiation with a dose of 1.74 × 104 Gy (Hillebrandt & Muller, 1991), suggesting that some, if not all, FLA have very efficient DNA repair mechanisms. Cryotherapy (repeated quick freezing to around −100 °C) has been used to treat Acanthamoeba spp. cysts in vitro and in clinical settings but was not fully efficient (Meisler et al., 1986; Matoba et al., 1989). Repeated freeze–thawing cycles were also demonstrated to have limited cysticidal activity against B. mandrillaris (five cycles from −80 to 37 °C) (Siddiqui et al., 2008) and against A. castellanii (five cycles from −160 to 45 °C) (Aksozek et al., 2002).

Biocide efficacy against intracellular pathogens

Bacteria associated with FLA generally display higher resistance to biocides than their planktonic counterparts. Two kinds of mechanisms can potentially explain this observation: the physical barrier to the action of biocides constituted by the trophozoite and the cyst and the resistant bacterial phenotype triggered by intracellular growth. These two phenomena are intimately associated and might be difficult to differentiate. They should, however, be taken into account in future studies aiming at comparing resistance to biocides of bacteria grown under different conditions. For example, comparing the resistance of intracellular vs. planktonic Legionella spp. to biocides should consist in treating both infected FLA and intracellular bacteria just released from their host; comparing infected FLA with bacteria grown on agar plates might lead to different results. When growing in FLA, L. pneumophila differentiate into mature intracellular forms (MIFs) (Garduno et al., 2002). These MIFs are short, stubby rod-like structures with an electron-dense, laminar outer membrane layer. They contain inclusions of poly-β-hydroxybutyrate and laminations of internal membranes originating from the cytoplasmic membrane (Garduno et al., 2002). Stationary phase bacteria obtained after growth on artificial media are morphologically distinct. They appear as dull rods and present a typical gram-negative cell wall ultrastructure (Garduno et al., 2002). Interestingly, MIFs have a very low respiration rate, increased resistance to detergent-mediated lysis and tolerate higher pHs compared with stationary-phase bacteria (Garduno et al., 2002). In addition, it has also been observed that genes that are necessary for intracellular infection and virulence increase L. pneumophila's resistance to the bactericidal effects of cationic antimicrobial peptides (Robey et al., 2001). It should also be recognized that intracellular pathogens use a variety of different mechanisms to survive oxidative stress encountered in phagocytic cells. These same mechanisms can also enhance resistance to disinfection treatments. For example, this was demonstrated for the expression of the DNA-binding protein gene (dpsX) from starved cells in ‘Candidatus Legionella jeonii’; the expression of this gene was enhanced by phagocytic activities, conferring upon bacteria higher resistance to liquid hydrogen peroxide (Park et al., 2006). Barker et al. (1992) reported the lack of efficacy of isothiazolinones against L. pneumophila grown within A. polyphaga. The decreased susceptibility was linked to an iron-depletion phenotype. Differences in phenotypes of a bacterial strain grown in different amoebal hosts are also thought to influence biocide susceptibility of released bacteria; this was recently evidenced with L. pneumophila replicated from H. vermiformis showing greater chlorine resistance than the cells replicated from A. castellanii (Chang et al., 2009).

Other studies have focused on the fate of bacterial pathogens within amoebae following biocide exposure from a more ‘mechanical’ perspective, without investigation of phenotypic changes induced by intracellular growth. King et al. (1988) investigated the survival and ‘resistance’ of a range of intracellular bacterial pathogens to free chlorine and concluded that A. castellanii trophozoites and, to some extent, T. pyriformis had a predominant role in the survival of these pathogens. Similar studies were reported with Burkholderia pseudomallei being more resistant to monochloramine, chlorine and UV once protected in Acanthamoeba astronyxis trophozoites (Howard & Inglis, 2005). Hwang et al. (2006) reported the decreased efficacy of silver (0.1 mg L−1) and copper (1 mg L−1) against L. pneumophila and P. aeruginosa within A. polyphaga trophozoites. Campylobacter survival to chemical disinfection with an iodine-based product has been reported to increase when associated with A. castellanii and T. pyriformis trophozoites (Snelling et al., 2005). Encystment of FLA is preceded by the expulsion of food vacuoles and vesicles (Schuster et al., 1979). These vesicles can also contain intracellular bacteria that have been observed to stay viable for up to 6 months (Bouyer et al., 2007). They can also directly protect intracellular bacteria from the effect of biocides (Berk et al., 1998). Encystation itself might be either beneficial or detrimental to the intracellular bacteria. Mycobacteria (Adekambi et al., 2006), F. tularensis (Abd et al., 2003) and V. cholerae (Thom et al., 1992; Abd et al., 2005) survive in cysts and might therefore benefit from protection toward biocide treatments. However, in some instances, the bacterial pathogen loses its viability within the cyst. This was reported with encystation of L. monocytogenes in A. castellanii (Ly & Muller, 1990). An additional phenomenon has been recently described with L. pneumophila infection of A. polyphaga trophozoites. Although both microorganisms seemed to exhibit an increase in resistance to sodium hypochlorite (available chlorine concentration was not measured), the amoebal host lost its ability to encyst. It thus provided an intracellular niche immediately available for resuscitation of extracellular L. pneumophila that enter a VBNC state upon treatment with chlorine (Garcia et al., 2007).

Methods used for inactivation studies

Antimicrobial testing against amoebae is compounded by the absence of a standard test method (Mercer et al., 2008). A variety of test protocols have been used (e.g. flow cytometric analysis, plaque assay and colorimetric assay) and ultimately it is difficult to compare the results between these studies (Khunkitti et al., 1997a; McBride et al., 2005). Buck et al. (2000) reviewed several studies that evaluated the efficacy of contact lens preservative/disinfectant against Acanthamoeba spp. cysts; she reported great variability in test organisms, growth conditions, inoculum preparation, neutralization, recovery and quantitation methods for survivors. The critical parameters have been demonstrated to be the age of the test culture, the type of medium used for induction of cysts and the maturity of cysts (Kilvington & Anger, 2001; Hughes et al., 2003). Cysts produced from isolates that have not been subcultivated for many generations in axenic media are more resistant than isolates cultivated in artificial broth, as are cysts obtained from monoculture with bacteria (Hughes et al., 2003). Various neutralization methods have been used to halt the activity of test biocides after specified contact times, and the Dey–Engley neutralizing broth, which is recommended in European Standards for bactericidal efficacy testing, has been proven to be nontoxic to amoebal trophozoites (Buck & Rosenthal, 1996). Despite this, neutralization should always be confirmed as an internal control and can be product/biocide specific. It has been reported that centrifugation might be more readily adapted than filtration for the recovery of trophozoites; both methods giving the same recovery rates with cysts (Pernin et al., 1998). Another critical step is the method used to evaluate and enumerate survivors after treatment. Excystation is more efficient with cysts grown on a live bacteria lawn than with cysts grown on artificial media, and the most probable number technique for the enumeration of amoebae grown on Escherichia coli lawns has been proposed as the basis of a national standard for amoebicidal efficacy testing (Beattie et al., 2003). The choice of bacteria used may be a further variable in the efficiency of recovery. Alternative enumeration methods based on AlamarBlue staining (McBride et al., 2005) and viability staining with fluorescein diacetate and propidium iodide have been proposed (Khunkitti et al., 1997a; Borazjani et al., 2000), but they may prove difficult to use with cysts due to low metabolic activity and their low permeability to chemicals.

Concluding remarks

There can be little doubt that amoebae play a major role in the composition of microbial flora of water systems. Currently, however, this role is largely overlooked. The microbiological quality of water is still based on the presence of coliforms and the presence of amoebae, and more generally of protozoa and ciliates is not considered unless they are pathogenic (i.e. Cryptosporidium spp. and Giardia spp.) and involved in infection. In some environments such as in hospitals and dental practices, amoebae are seldom, if at all, considered. When they are investigated in these environments, amoebae are found together with a range of intracellular microorganisms, some being demonstrated pathogenic species and others for which a pathogenic role remains to be elucidated. FLA offer a protective intracellular environment in which to harbor microbial pathogens, thereby significantly reducing the efficacy of biocide treatment. In some cases, they might also favor persistence and extracellular proliferation of pathogenic bacterial species, thus leading to selection of these predation-resistant microorganisms in water systems. A number of investigations are needed to increase our understanding of these microorganisms together with the means for their control. The ecology of microorganisms in water systems and the impact of biocides have to be approached from a global perspective, taking into account interactions of microorganisms with biofilms and amoebae and the potential selection of pathogenic species during disinfection. These are suggested to include studies investigating (1) the interaction of physical and chemical biocides with amoebal trophozoites and cysts and the mechanisms of resistance to disinfection, (2) the survival of amoebae in complex microbial biofilms, (3) the fate of intracellular pathogens following biocidal treatment and (4) trophozoite and cyst detection. Finally, the standardization of cyst production and the efficacy testing for disinfection and sterilization investigations are required to produce meaningful and reproducible data. FLA have also been demonstrated to constitute a preferential place for intra- and interspecies genetic exchanges between intracellular microorganisms and their hosts (Ogata et al., 2006). They could thus act as true reservoirs and cross-kingdom vehicles of genetic information, including bacterial virulence or resistance factors. The availability of A. castellanii Neff genome (see http://www.hgsc.bcm.tmc.edu) should bring valuable information in this field of investigation and will trigger other FLA genome sequencing.

Increasing knowledge of the resistance of trophozoites and cysts to disinfection, their impact on selecting potential pathogenic species from complex microbial flora and their capacity to release intracellular pathogens will undoubtedly raise important questions. It will be our responsibility to use this information to develop new mechanisms for control.


  • Editor: Colin Berry


View Abstract