Previous Article | Next Article ![]()
Clinical Microbiology Reviews, April 2004, p. 413-433, Vol. 17, No. 2
0893-8512/04/$08.00+0 DOI: 10.1128/CMR.17.2.413-433.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
and Didier Raoult*
Unité des Rickettsies, Faculté de Médecine, Université de la Méditerranée, Marseille, France
SUMMARY INTRODUCTION FREE-LIVING AMOEBAE AS A TOOL FOR ISOLATION OF AMOEBA-RESISTANT INTRACELLULAR MICROORGANISMS Culture of Amoebae for Detecting ARB Practical Use of Amoebae for ARB Culture AMOEBA-RESISTANT MICROORGANISMS Holosporaceae Bradyrhizobiaceae Legionellaceae Legionella pneumophila. Legionella anisa and other Legionella spp. Legionella-like amoebal pathogens. Pseudomonaceae Parachlamydiaceae Mycobacteriaceae Mimivirus Enterovirus Other Endosymbionts of Free-Living Amoebae Rickettsia-like endosymbionts. Members of the Cytophaga-Flavobacterium-Bacteroides phylum. ß proteobacteria naturally infecting free-living amoebae. Other Microorganisms Shown In Vitro To Resist Destruction by Free-Living Amoebae Burkholderiaceae. Coxiella burnetii Francisella tularensis. Enterobacteriaceae. Vibrionaceae. Listeria monocytogenes. Helicobacter pylori. Mobiluncus curtisii. Cryptococcus neoformans. Microorganisms Recovered Using Amoebal Coculture FREE-LIVING AMOEBAE AS A RESERVOIR OF AMOEBA-RESISTANT MICROORGANISMS TRANSMISSION OF AMOEBA-RESISTANT MICROORGANISMS FREE-LIVING AMOEBAE AS AN EVOLUTIONARY CRIB Induction of Virulence Traits Adaptation to Macrophages CONCLUSION REFERENCES
| SUMMARY |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
Free-living amoebae have at least two developmental stages: the trophozoite, a vegetative feeding form, and the cyst, a resting form (Fig. 1). Some amoebae, such as Naegleria spp., have an additional flagellate stage. Others, such as Mayorella and Amoeba, are non-cyst-forming species (191). The trophozoite, the metabolically active stage, feeds on bacteria and multiplies by binary fission. Cysts generally have two layers, the ectocyst and the endocyst. A third layer, the mesocyst, is present in some species. The structure may explain why cysts are resistant to biocides used for disinfecting bronchoscopes (101) and contact lenses (31, 124, 250) as well as to chlorination and sterilization of hospital water systems (206, 214). Adverse pH, osmotic pressure, and temperature conditions cause amoebae to encyst. Encystment also occurs when food requirements are not fulfilled. Protozoa excyst again when environmental conditions become favorable.
|
(108). Acanthamoeba is the only pathogenic species isolated from marine water (11). Free-living amoebae feed mainly on bacteria, fungi, and algae by phagocytosis, digestion occurs within phagolysosomes. Some microorganisms have evolved to become resistant to protists, since they are not internalized or are able to survive, grow, and exit free-living amoebae after internalization. These "amoeba-resistant microorganisms" (definition given in Table 1) include bacteria, viruses, and fungi. Among the amoeba-resistant bacteria (ARB) (Table 2), some were recovered by amoebal coculture (Bosea spp.) while others were identified within free-living amoebae isolated by amoebal enrichment (Procabacter acanthamoeba). Some are obligate intracellular bacteria (Coxiella burnetii), while others are facultative intracellular bacteria (Listeria monocytogenes) or might even be bacteria without a known eucaryotic-cell association (Burkholderia cepacia and Pseudomonas aeruginosa). Some are established human pathogens (Chlamydophila pneumoniae), while others are emerging pathogens (Simkania negevensis) or nonpathogenic species (Bradyrhizobium japonicum). Among the ARB, some (the Legionella-like amoebal pathogens [LLAP]) were named according to their cytopathogenicity (28) and represent the paradigm of bacteria able to lyse amoebae, while others (such as Parachlamydia acanthamoeba) were considered endosymbionts, since a stable host-parasite ratio was maintained (7). The term "endosymbionts," defined by Büchner as "a regulated, harmonious cohabitation of two nonrelated partners, in which one of them lives in the body of the other" (38), refers to the intra-amoebal location of the bacteria [endo] and to its close relationship with the amoebae [symbiosis] (Table 1). In the present review, we generally use the term ARB instead of endosymbiont, since many bacteria able to resist destruction by free-living amoebae do not represent true endosymbionts and since even endosymbionts may be endosymbiotic or lytic while in a given amoeba, depending on environmental conditions (96).
|
|
|
|
| FREE-LIVING AMOEBAE AS A TOOL FOR ISOLATION OF AMOEBA-RESISTANT INTRACELLULAR MICROORGANISMS |
|---|
|
|
|---|
Culture remains the ultimate goal of pathogen identification, since it makes a microorganism available for further study (122). Culturing the microorganism is useful to reliably classify it, to test its ability to infect human macrophages, to determine its pathogenicity in animal models, to test its antibiotic susceptibility, to use it as antigen for serological testing, and to produce polyclonal or monoclonal antibodies.
The two-step strategy allowed the identification and culture of several ARB, such as Odyssella thessalonicensis (27), Parachlamydia acanthamoebae (7, 29), and several Legionella-like amoebal pathogens (LLAP) (82, 212). The procedures generally used for growing free-living amoebae were recently extensively reviewed in this journal (217) and are not presented here.
Amoebal coculture appeared useful for the recovery of L. pneumophila (210), L. anisa (156), and numerous
proteobacteria (150, 155). In addition, as amoebae graze on bacteria, amoebal coculture may help in selectively growing new ARB of unknown pathogenicity (150). Amoebal coculture could also be used to clean the samples from other more rapidly growing species that generally overwhelm the agar plates. Thus, using that technique, Rowbotham was able to grow L. pneumophila from human feces (213). Moreover, the amoebal coculture is a cell culture system that may be performed in the absence of antibiotics and is suitable for the recovery of new bacterial species of unknown antibiotic susceptibility. The main limitation of this technique is the decreased viability of amoebae such as Acanthamoeba and their encystment at high incubation temperature, which does not allow the recovery of bacteria requiring a temperature of
37°C. Hence, only 45% of A. polyphaga organisms are viable trophozoites after 4 days at 37°C, compared to 65 to 85% at 25 to 32°C (96).
Depending on the nature of the sample and of the bacteria sought, the inoculation may be processed differently. Thus, to apply both the amoebal coculture and the two-step approach to a sample, it is advisable to first centrifuge it at low speed (about 180 x g for 10 min) and to use the supernatant for amoebal coculture and the pellet for amoebal enrichment. However, when using only one approach, it is better to inoculate all of the specimen to increase the sensitivity of the amoebal coculture. To disrupt the amoebal cells present in order to release the intracellular microbes, the samples may be treated with a lytic solution, such as 2,3-hydroxy-1,4,-dithiolbutane or dithiothreitol (156, 210), or with repeated sequential exposure to liquid nitrogen and boiling water, the "hot-ice" procedure. However, the advantages of these lytic procedures in terms of sensitivity have not been evaluated yet. Acid decontamination (210) and addition of both colistin (500 U/ml) and vancomycin (10 µg/ml) (150) may help to selectively grow Legionella spp. After inoculation, centrifugation of the microplates (2,900 x g) may accelerate the contact between the inoculum and the amoebal background and thus increase the sensitivity of the procedure.
The inoculated microplates are incubated at 30 to 35°C in a humidified atmosphere. The humidified atmosphere should prevent amoebal encystment. Too high a temperature will cause early amoebal encystment, whereas too low a temperature will be associated with less bacterial growth. Depending on the encystment rate, subcultures on fresh amoebae should be performed after about 4 to 7 days. Moreover, amoebal cocultures should be examined every day or at least every 2 days for the appearance of amoebal lysis that will suggest the presence of an amoeba-resistant microorganism and will necessitate subculture on fresh amoebae. At the time of subculture, each well should be screened for the presence of intra-amoebal bacteria. This screening is best achieved by gently shaking the microplates to suspend amoebae. Then, about 200 µl of the suspension is cytocentrifuged (147) and slides are stained with Gram, Gimenez (87), or Ziehl-Neelsen strain, depending on the suspected bacteria. We prefer Gimenez staining because most fastidious ARB are Gimenez positive and because fuchsin-stained bacteria are easily seen within the malachite green-stained amoebae (Fig. 3). The slide may also be prepared for immunofluorescence testing with the patients' serum as the primary antibody. The latter approach is valuable for detecting any microorganism against which a given patient has developed antibodies.
|
proteobacteria) and should be incubated for prolonged periods. Subcultures on amoebal microplates are performed by inoculation of about 100 to 150 µl of the coculture on 1 ml of a fresh amoebal suspension. Preservation of the remaining sample at 80°C may be useful because some material is still available for additional subculture attempts. Subcultures on other cell lines may also be possible after filtration to eliminate the amoebae. | AMOEBA-RESISTANT MICROORGANISMS |
|---|
|
|
|---|
Another member of the Holosporaceae, Candidatus Odyssella thessalonicensis, has been identified from an amoeba collected from the drip tray of the air conditioning system of a hospital in Greece (27). A recent phylogenetic analysis of the 16S rRNA genes of all these symbionts of Acanthamoeba, of the ciliate Paramecium caudatum (C. caryophilus, H. obtusa, and H. elegans) and of the shrimp Penaeus vannamei (NHP bacterium) showed that Candidatus O. thessalonicensis was phylogenetically close to Candidatus P. acanthamoebae (22). Interestingly, the branching of host and endosymbiont phylogenetic trees was congruent (22, 114, 117). This suggested that the ancestor of the C. caryophilus-related endosymbionts lived within an amoebal progenitor and coevolved with their hosts during the diversification of the different Acanthamoeba sublineages (22). If this is true, C. caryophilus will have only recently transferred to the ciliate Paramecium caudatum. The R-body represents a traditional taxonomic key criterion of the genus Caedibacter, confering killer traits (195, 216). The restricted presence of genetic mobile elements encoding the R-body in C. caryophilus and their absence in the endosymbionts of Acanthamoeba suggest that R-body-encoding genetic elements were acquired by C. caryophilus since it began to live within the ciliate P. caudatum (22).
Candidatus O. thessalonicensis was successfully grown on A. polyphaga, suggesting that amoebae may be an useful tool for culture of amoebal symbionts and for the recovery of new species (see below). The host range of O. thessalonicensis is narrow, being restricted to Acanthamoeba spp. Interestingly, incubation at 37 and 30°C resulted in amoebal lysis after 4 and 7 days, respectively, whereas at 22°C O. thessalonicensis appeared to form a stable host-endosymbiont equilibrium during a 3-week coincubation period (27). This might correspond to a modulation of the virulence at higher temperatures (see below). It is another example of endosymbionts that behave as lytic or symbiotic bacteria depending on environmental conditions (96).
The human pathogenicity of symbionts related to C. caryophilus remains to be determined.
A. broomae, A. massiliensis, and A. birgiae were also shown to resist destruction by free-living amoebae (150, 153, 155). A. massiliensis and A. birgiae were recovered from water by amoebal coculture (153). By analogy to what has been learned from experiments with Legionella, it has been proposed that Afipia spp. may be agents of nosocomial pneumonia (153), especially since they are common in hospital water networks (150). A pathogenic role for Afipia is further suggested by its uptake within murine macrophages in a nonendocytic compartment (162) and by its isolation from a patient with osteomyelitis (36). Although exposure of intensive care unit patients to Afipia spp. has been documented serologically (155), the role of Afipia spp. as etiological agents of nosocomial pneumonia remains to be demonstrated. Indeed, A. clevelandensis may cross-react with Brucella spp. and Yersinia enterolitica (64), demonstrating the low intergenus specificity of immunoflurescence testing for that clade.
The genus Bosea was described by Das et al (57) based on a single isolate, Bosea thiooxydans, recovered from agricultural field soil during a study of chemolithotrophic bacteria (56). Three additional species, B. eneae, B. vestrisii, and B. massiliensis (154), were recovered by amoebal coculture from hospital water supplies (150), demonstrating that some representatives of that genus are ARB commonly present in water. More importantly, they were associated with severe nosocomial pneumonia in ventilated intensive care unit patients (152).
As long ago as 1980, Rowbotham suggested the role of amoebae, not only as a reservoir but also in the propagation and distribution of Legionella spp. in water systems and in the transmission of these bacteria to humans (209). He proposed that humans are infected not by inhaling free legionellae but by inhaling a vesicle or an amoeba filled with Legionella organisms (Fig. 4) (209). These vesicles filled with Legionella (8, 26) might contain as many as 104 bacteria (211). Since then, the relationship between L. pneumophila and free-living amoebae has been extensively studied, and the studies have confirmed that free-living amoebae are necessary for Legionella multiplication in water biofilms, although the bacteria may survive in a latent state in biofilms without amoebae (186).
|
No case of person-to-person transmission of L. pneumophila has been reported, suggesting that genetic variants of Legionella that survived the strong selective pressure exerted by the alveolar macrophages are not maintained (230). Conversely, the virulence traits selected by intra-amoebal life will persist in the "species" genome. These virulence traits include motility (211), resistance to cold (16), invasive phenotype (44), resistance to antibiotics such as erythromycin (19), and resistance to biocides (16). Increased motility is associated with increased spread of the microorganism while extracellular. However, this virulence trait was not expected to be the result of intracellular growth within amoebae since neither motility nor flagella are required for intracellular growth (177, 196). This apparent paradox may be explained by a model of differential phenotypic expression of L. pneumophila depending on growth conditions (39). Thus, when growth conditions are optimal in host cells, L. pneumophila expresses functions to replicate maximally, and when amino acids become limiting, L. pneumophila produces factors to lyse the exhausted host cell, to survive osmotic stress, to spread in the environment, and to evade lysosomal degradation in the new intracellular niche (39).
These virulence traits expressed during the postexponential phase and selected over millions of years of replication within its protozoan host might explain the adaptation of L. pneumophila to life in human macrophages. Indeed, the life cycle of L. pneumophila within macrophages is very similar to that within amoebae (Table 4), and at the molecular level, some identical strategies are used to adhere to, enter, escape from, replicate in, and exit from both (see below).
|
The interactions between water, free-living amoebae, and Legionella spp. were better understood after the investigation of an outbreak of Pontiac fever. A total of 34 persons attending conferences at a hotel in California presented with Pontiac fever, a disease characterized by acute fever and upper respiratory tract illness (74). They were presumably infected by L. anisa since five of eight subjects exhibited a fourfold rise in antibody titers and since L. anisa and Hartmanella vermiformis were isolated from the decorative fountain in the hotel lobby (74, 75). In contrast to the wide amoebal host range of L. pneumophila (211), that strain of L. anisa grew only within H. vermiformis. It could not be grown within another protist, Tetrahymena pyriformis, or within human mononuclear cells (75). The narrow host range of L. anisa may explain its role as an agent of Pontiac fever (75), a milder disease than legionellosis, and may also explain why it was recognized mainly in immunocompromised patients (156, 235). We also showed the role of L. anisa as an agent of ventilator-acquired pneumonia, causing a cryptic epidemic due to contamination of intensive care unit tap water (152).
Legionella-like amoebal pathogens. In 1956, Drozanski described an obligate intracellular parasite of free-living amoebae that causes lysis of the amoeba cells (66). Though initially named Sarcobium lyticum (67), this species was reclassified within the genus Legionella as L. lytica (112). In the meantime, Rowbotham reported the isolation of a Legionella-like bacterium, which he named Legionella-like amoebal pathogen 1 (LLAP-1), since, like L. pneumophila, it was able to induce amoebal lysis (211). Although LLAP-1 did not fluoresce with L. pneumophila, L. micdadei, and L. feeleii SG1 antisera and could not be grown on BCYE agar, it exhibited a fatty acid profile suggesting its placement within the genus Legionella (211). Phenotypic and genotypic characterization later allowed its classification as Legionella drozanskii (5). In 1991, another LLAP (LLAP-3) was identified within an amoeba enriched from the sputum of a patient with pneumonia (82). LLAP-3 was later shown to be a member of the species L. lytica (28).
Then, additional LLAP were recovered from environmental sources (28, 181). LLAP-9, LLAP-7FL (fluorescent), and LLAP-7NF (nonfluorescent) were also shown to be members of the species L. lytica (5), while LLAP-6, LLAP-10, and LLAP-12 represented three new species of Legionella: L. rowbothamii, L. fallonii, and L. drancourtii, respectively (5, 151).
Since the recovery of L. lytica (LLAP-3) from a patient with pneumonia who seroconverted against this strain and whose infection improved with macrolide therapy (82), there has been growing evidence that these emerging species of Legionella account for some of the pneumonias of unknown etiology (4, 174, 212; C. E. Benson, W. Drozanski, T. J. Rowbotham, I. Bialkowska, D. Losos, J. C. Butler, H. B. Lipman, J. F. Plouffe, and B. S. Fields, Abstr. 95th Gen. Meet. Am. Soc. Microbiol. 1995, abstr. C-200, p. 35, 1995). Serological evidence of 10 L. lytica infections was obtained from screening more than 5,000 patients (212). More importantly, 4.3% of 255 patients hospitalized for a community-acquired pneumonia were seropositive for L. drancourtii (LLAP-4), compared to 0.4% of 511 healthy controls (p = 0.045).
The dichotomy between LLAP and other Legionella spp. may appear artificial, being based only on the fact that LLAP grow poorly or not at all on BCYE agar (28). However, it has helped to demonstrate that LLAP are emerging agents of pneumonia (4, 170, 212; Benson et al., Abstr. 95th Gen. Meet. Am. Soc. Microbiol. 1995) and that BCYE is not suitable for in vitro cultivation and consequently for the diagnosis of LLAP-related pneumonia. Amoebal coculture is an important tool for diagnosing these emerging infections.
In nature, free-living amoebae probably also feed on Pseudomonas spp. that are widely distributed in water and that are present at low concentration. Their encounter may be facilitated by the better adherence of Pseudomonas (than of E. coli.) to Acanthamoeba (33). However, some Pseudomonas spp. evolved to become resistant to amoebae, as demonstrated by the isolation of Acanthamoeba naturally infected with P. aeruginosa (178). Hence, free-living amoebae might also play a role as a reservoir for some amoeba-resistant strains of Pseudomonas, similar to what has been shown for Legionella spp. This is important, given the role of P. aeruginosa as an agent of pneumonia (85). Whether there is a correlation between resistance to Acanthamoeba and pathogenicity remains to be determined. Models of virulence are discussed below.
|
More importantly, there is a growing body of evidence that Parachlamydia is an emerging pathogen of clinical relevance (98). The first hint was the identification of Parachlamydia sp. strain Hall coccus within an amoeba isolated from the source of an outbreak of humidifier fever in the United States (29) and a related serological study (29). In another serological study, fourfold-increased titers of antibodies against Parachlamydia in 2 of 500 patients with pneumonia was observed (Benson et al., Abstr. 95th Gen. Meet. Am. Soc. Microbiol. 1995). More recently, 8 (2.2%) of 371 patients with community-acquired pneumonia were seropositive (titer, >1/50) for Parachlamydia, compared to 0 of 511 healthy subjects (170). The amplification of DNA of Parachlamydiaceae from bronchoalveolar lavage fluid and sputum (50, 52) and from mononuclear cells of a patient with bronchitis (190) provided additional hints of a potential pathogenicity. Our own work provided evidence that Parachlamydia may be an agent of community-acquired pneumonia, at least in human immuno deficiency virus-infected patients with low CD4 counts (94), and may also be an agent of aspiration pneumonia, at least in severely head-injured trauma patients (95). Moreover, Parachlamydia enters and multiplies within human macrophages (97), another point in favor of its pathogenic role. In conclusion, human exposure to Parachlamydia could be a cause of upper respiratory tract infection, bronchitis, aspiration pneumonia, and community-acquired pneumonia.
The possible role of Parachlamydia in the pathogenesis of Kawasaki disease (170), a vasculitis associated with respiratory infections (23, 58), and in the pathogenesis of atherosclerosis (95, 190) merits further study.
Simkania negevensis is an agent of pneumonia in adults and of bronchiolitis in children (134, 159); it is phylogenetically related to the Chlamydiaceae and to the Parachlamydiaceae (72). Its ability to multiply within A. polyphaga and to survive within cysts has been documented (133), suggesting that free-living amoebae might also act as a reservoir and a selective environment ground for this clade. An extensive review has been recently published (78).
Chlamydophila pneumoniae may survive within Acanthamoeba castelanii but, in contrast to Parachlamydiaceae and Simkaniaceae, does not grow within this species of amoeba (71). The role of free-living amoebae as a reservoir for this established agent of lung infections remains to be tested. Whether additional species of Chlamydiaceae may resist destruction by free-living amoebae also remains to be established.
|
Rickettsia-like endosymbionts. The presence of two endosymbionts of Acanthamoeba spp. phylogenetically related to members of the order Rickettsiales was demonstrated in 1999 by Fritsche et al. (80). These two Rickettsia-like strains were closely related, with 99.6% 16S rRNA sequence homology. They were more closely related to Rickettsia sibirica and R. typhi, with sequence similarities of 85.4% (80). Phylogenetic analysis confirmed that they share a common ancestor with Rickettsia spp. However, important sequence divergence was demonstrated by deep branching in the phylograms (80). Given the fact that Rickettsiales symbionts apparently have a narrow host range and that Rickettsia spp. may be considered commensal endosymbionts of ticks, this deep branching may correspond to the time of divergence of the protozoan and arthropods or to the time of their acquisition by ancestral ticks. The human pathogenicity of this Rickettsia-like lineage remains to be defined, as do its host range, prevalence, distribution, and interactions with free-living amoebae.
Members of the Cytophaga-Flavobacterium-Bacteroides phylum. The first member of the Cytophaga-Flavobacterium-Bacteroides phylum identified within a free-living amoeba had a fatty acid profile indicating its affiliation with Cytophaga (187). Additional work showed its closer relationship to Flavobacterium succinicans (99% 16S rRNA gene sequence homology) (116). This Flavobacterium sp. and another strain related to Flavobacterium johnsoniae (98% 16S rRNA sequence homology), also identified within an Acanthamoeba organism, were able to grow on sheep blood agar, demonstrating a facultative intracellular lifestyle (116). Whether Flavobacterium spp. use the free-living amoebae as a reservoir and whether these bacteria play a role as human pathogens remains to be defined.
Another member of the Cytophaga-Flavobacterium-Bacteroides phylum was identified within an Acanthamoeba organism isolated from lake sediment in Malaysia (116). In contrast to the Flavobacteriaceae, this strain, named Amoebophilus asiaticus, did not grow on agar-based media and was thus considered an obligate intracellular endosymbiont of Acanthamoeba (116). Of note, the host range of that symbiont was restricted to the host Acanthamoeba and attempts to infect other Acanthamoeba spp., Naegleria spp., Hartmanella vermiformis, Vahlkampfia ovis, Balamuthia mandrillaris, Willaertia magna, and Dictyostelium discoideum failed (116). This narrow host range and the fact that A. asiaticus is closely related to another endosymbiont of Acanthamoeba apparently isolated in Hungary (116) and, to a lesser degree, to Ixodes and Encarsia symbionts suggest that their common ancestor acquired or developed adaptative features which allow its descendants to successfully infect different eucaryotic lineages (116). The human pathogenicity of A. asiaticus also remains to be defined.
ß proteobacteria naturally infecting free-living amoebae. Ralstonia pickettii (180) and Procabacter acanthamoeba (115) are the only two species of ß proteobacteria shown to naturally infect free-living amoebae. R. pickettii may act as an opportunistic pathogen (249). However, it is mainly incriminated as a contaminant of solutions used either for patient care (39, 146, 169) or for laboratory diagnosis (34), sometimes associated with pseudobacteremia (34).
Procabacter acanthamoeba, named in honor of Proca-Ciobanu, was identified within Acanthamoeba spp. recovered by amoebal enrichment of environmental samples and corneal scrapings (115). Its pathogenic role is largely unknown, but given its obligate intracellular lifestyle, infection may remain undiagnosed if axenic cultures alone are used.
Burkholderiaceae. B. cepacia is associated with severe lung infections, especially in cystic fibrosis patients (76, 234) and, to a lesser extent, in intensive care unit patients. The role of free-living amoebae as a reservoir of B. cepacia has not been demonstrated. However, their role in the transmission of B. cepacia is possible, since expelled vesicles filled with B. cepacia bacteria have been reported (168).
B. pseudomallei causes melioidosis, an infection that may present as a fatal acute septicemia (239) or as subacute or chronic relapsing form (41). The fact that B. pseudomallei is resistant to amoebae (127) may explain its association with water (126) and its ability to survive within macrophages (132).
Coxiella burnetii C. burnetii, an obligate intracellular bacterium, is the agent of Q fever. Phylogenetically related to Legionella spp. (245), it was also reported to resist destruction by free-living amoebae (158). Human infection occurs mainly by inhalation of infected aerosols in areas of livestock breeding, as demonstrated by the increased prevalence of Q fever in populations living downwind from breeding areas (240). The knowledge that C. burnetii is able to survive within free-living amoebae may shed some light on the epidemiology of that fastidious organism, suggesting that the bacteria expelled into the environment might persist for months within amoebae and might use them for transmission. It may also explain the resistance of C. burnetii to biocides (219) and its adaptation to life within phagolysosomes at acidic pH (172, 175).
Francisella tularensis. Tularemia is a zoonotic disease caused by F. tularensis. It may be acquired by exposure to various mammals such as ground squirrels, rabbits, hares, voles, muskrats, and water rats or to ticks, flies, and mosquitos (70). There is evidence that F. tularensis can persist in water courses. Beavers and other water mammals, including lemming carcasses, might play the role of reservoir for the bacteria in water. However, the fact that free-living amoebae filled with F. tularensis can be observed, including subsequent lysis of the amoebae (2, 25), suggests that the protists might be an important water reservoir.
Enterobacteriaceae. Enterobacteriaceae and several nonfermentative gram-negative bacteria (such as Stenotrophomonas maltophilia) are among the preferred nutrient sources of free-living amoebae. Thus, E. coli, E. aerogenes, and S. maltophilia appear to be better nutrient sources for amoebae than are Staphylococcus epidermidis, Serratia marcescens, and Pseudomonas spp. (243, 244). This explains why Enterobacter spp., Klebsiella spp., and E. coli are preferred for amoebal enrichment procedures (193, 217). Interestingly, viable E. coli organisms provide a higher yield of trophozoites than do nonviable ones (131). Conversely, E. coli is able to multiply in the presence of Acanthamoeba (224). However, unlike Legionella, which multiplies within the trophozoites, growth of E. coli is also possible when the bacteria are separated from Acanthamoeba by a semipermeable membrane (224). Although at the population or species level Enterobacteriaceae and Acanthamoeba spp. may be seen as mutualists, at the individual level the amoebae are better considered to be predators of the Enterobacteriaceae. The fact that the predator does not completely eliminate its prey is a generally accepted concept that may occur as a result of a variety of mechanisms (6). These mechanisms include interactions among predators (interference with their grazing activity), genetic feedback (mutants arise that are resistant to the predator or parasite), physical refuge (the presence of small pores of soil), switching to another prey, density dependence of attack by the predator, and increased replication of the prey that compensates for killing (6). Whether Enterobacteriaceae should be added to the growing list of ARB remains to be determined. Indeed, the amoeba-bacterium ratio and the viability and fitness of the amoebae should also be taken into account to validly determine the resistance of these species to free-living amoebae. Probably, some strains or some clones have a resistant phenotype, which may prove to be transient or persistent. Thus, the Vero cytotoxin-producing E. coli strains clearly represent ARB, since their population increased significantly in coculture (17).
Vibrionaceae. Vibrio cholerae is a member of the Vibrionaceae that is reported to survive and multiply within A. polyphaga and N. gruberi (237). Its survival within cysts of N. gruber i suggests that free-living amoebae may protect the bacteria while encysted.
Listeria monocytogenes. Since Listeria has been isolated from soil samples, sewage, and wastewater and since it resists destruction in human macrophages, Ly and Müller supposed that Listeria might be resistant to free-living amoebae (163). They confirmed their hypothesis by showing that L. monocytogenes multiplies within Acanthamoeba (163). However, after 1 month, most amoebal cells were encysted, preventing further Listeria survival (163).
Helicobacter pylori. H. pylori may be present in water (113, 125, 173, 192), suggesting that free-living amoebae could play the role of reservoir for these fastidious microorganisms. This hypothesis is sustained by the demonstration that H. pylori is able to grow when cocultured with A. castellanii (248). More important, the viability of H. pylori could be maintained for up to 8 weeks in coculture with A. castellanii in the absence of microaerobic conditions (248). Additional studies are needed to determine the role played in vivo by free-living amoebae in the transmission of H. pylori.
Mobiluncus curtisii. Only one species of anaerobe (Mobiluncus curtisii) has been reported to resist destruction by free-living amoebae (241). This obligate nonsporeforming anaerobic bacterium, causing vaginosis (218) and, more rarely, abcesseses (69) and bacteremia (91, 109), persisted for up to 4 to 6 weeks under aerobic conditions as a result of its internalization in Acanthamoeaba (241). Like H. pylori, M. curtisii may multiply aerobically when cocultured with free-living amoebae, while it requires otherwise strict atmospheric conditions for in vitro culture. This suggests that some strictly anaerobic bacteria are able to find refuge within amoebae. Further work is needed to determine the role of free-living amoebae as reservoirs of anaerobes. In particular, it may be interesting in the future to determine whether the vagina is colonized with free-living amoebae and whether such colonization may be associated with gynecological infections. The pathogenicity of M. curtisii, which is a commensal inhabitant of the vaginal flora and an established agent of vaginoses (218), might be due to a disequilibrium in the ratio of bacteria to amoebae.
Cryptococcus neoformans. Cryptococcus neoformans is a soil fungus that causes life-threatening meningitis in immunocompromised patients. It is one of the best examples of common adaptations to both amoebae and human macrophages (see below), suggesting that certain aspects of cryptococcal human pathogenesis are derived from mechanisms used by fungi to survive within environmental amoebae (223). Indeed, nonvirulent, nonencapsulated strains do not survive in amoebae whereas virulent strains do.
| FREE-LIVING AMOEBAE AS A RESERVOIR OF AMOEBA-RESISTANT MICROORGANISMS |
|---|
|
|
|---|
By definition, a microorganism able to resist free-living amoebae is able to enter, multiply within, and exit its amoebal host (Fig. 2); thus, free-living amoebae may be considered to be reservoirs of any amoeba-resistant microorganism (246). The implications of such a reservoir in terms of ecology, epidemiology, and public health remain to be better defined.
Since amoebae graze on bacteria, the entry of the bacteria by phagocytosis (Fig. 7) may be relatively easy, potentially being a passive process from the point of view of the ARB. This hypothesis is sustained by the fact that, in contrast to entry into macrophages, the entry of Legionella into amoebae is not inhibited by cytochalasin D (136). The intra-amoebal environment then favors the multiplication of a large variety of microorganisms (see above). Finally, exit from the amoebae may occur as expelled vesicles or by amoebal lysis (99, 211). Amoebal lysis is associated with the liberation of large numbers of bacteria. The mechanism of lysis has only been partially elucidated for L. pneumophila. Osmotic lysis of macrophages infected with L. pneumophila was shown to be mediated by the insertion of a pore into the plasma membrane (139). Gao and Abu Kwaik showed that this pore-forming activity of L. pneumophila was involved in the lysis of A. polyphaga (83). Thus, the wild-type bacterial strain was shown to cause the lysis of all the A. polyphaga cells within 48 h after infection, and all the intracellular bacteria are released into the culture medium (83). In contrast, all cells infected by the mutants remain intact, and the intracellular bacteria are "trapped" within A. polyphaga after the termination of intracellular replication (83). The icmT gene was shown to be essential for this pore formation-mediated lysis (185). For other ARB, the mechanism of lysis is unknown. Interestingly, the lysis was shown to be dependent on environmental conditions such as temperature. Indeed, we recently showed that Parachlamydia acanthamoebae is lytic for A. polyphaga at 32 to 37°C and endosymbiotic at 25 to 30°C (96). This suggests that A. polyphaga may serve as a reservoir for ARB at lower temperatures (for instance, when colonizing the nasal mucosa) and is liberated by lysis at higher temperatures (for instance when reaching the human lower respiratory tract) (Fig. 2). The role of free-living amoebae in resuscitating viable but nonculturable microorganisms should be further defined. Indeed, starvation conditions such as those present in low-nutrient-containing water or other stress such as exposure to antibiotics may induce some gram-negative bacteria to enter a viable but nonculturable state (40, 48, 49, 65). Up to now, a role for free-living amoeba in resuscitating viable but nonculturable bacteria has been demonstrated only for L. pneumophila (225).
|
Cysts are highly resistant to extreme conditions of temperature, pH, and osmolarity (204). The survival of ARB within cysts has been especially well documented for Mycobacterium avium (224) and Simkania negevensis (133). After 79 days at 4°C, the infectivity of S. negevensis was still greater than 50% of the initial infectivity for Acanthamoeba, while in the absence of amoebal cysts and trophozoites, the bacteria did not survive for 12 days at 4°C (133). Hence, for the internalized microorganism, the cyst may represent more than a protection: it may play a role in the persistence of the microorganism in the environment.
Moreover, Acanthamoeba spp. cysts are highly resistant to biocides used for contact lens disinfection (31, 207, 250). Consequently, when encysted, free-living amoeba could protect the internalized bacteria (137). This could explain the observed increased resistance to chlorine of A. felis (157) and L. pneumophila (135) within A. polyphaga.
| TRANSMISSION OF AMOEBA-RESISTANT MICROORGANISMS |
|---|
|
|
|---|
Free-living amoebae may also increase the transmission of ARB by producing vesicles filled with bacteria (8, 26, 99, 168, 211). These vesicles, first described for L. pneumophila (8, 26, 211), may increase the transmission potential of Legionella spp. and may lead to underestimation of the risk by colony plate count methods (26). Vesicles have also been reported to contain Burkholderia cepacia (168) and Parachlamydia acanthamoebae (99).
| FREE-LIVING AMOEBAE AS AN EVOLUTIONARY CRIB |
|---|
|
|
|---|