Free Living Amoebae as Opportunistic Hosts for Intracellular Bacterial Parasites
Since the original identification of Legionella pneumophila as the causative agent of Legionnaires' disease, the number of described species in the genus has grown to over 30. Nearly one half of these species have been associated with human illness, however over 80% of the reported cases of Legionnaires' disease are caused by L. pneumophila [4]. A disease characteristic is the replication of bacteria within alveolar macrophages. The widespread occurrence of Legionella species in conjunction with the fastidious nature of the genus suggested they might require interactions with other microorganisms to replicate and sustain themselves in the environment. In vitro studies have clearly demonstrated the ability of L. pneumophila (and some other species in the genus) to use amoebae as a host cell for intracellular replication. More recently the occurrence of undescribed Legionella species that are pathogens of common free living amoebae has been reported. Collectively, the bacteria have been described as the Legionella-like, amoebal pathogens (LLAPs) because they are capable of multiplying in the cytoplasm of amoebae and are fastidious in their growth requirements. LLAPs were originally described in Europe and were taken from a variety of environmental sources [2].
Some intracellular events following infection, such as the appearance of ribosomes and mitochondria in proximity to the membrane-enclosed bacteria, are common to both amoebae and human mononuclear phagocytes infected with L. pneumophila. The amoebae, Acanthamoeba and Hartmannella species, commonly found in natural and human made aquatic environments [1], have been used most frequently in laboratory studies. Generally Legionella cells were added to amoebae at a multiplicity of infection (MOI) of 10:1 to 200:1 (Legionella to amoebae) in various types of media.
The purpose of present work was to examine natural and man-made environments for the presence of amoebae containing intracellular parasitic bacteria, and to characterize the bacteria. In addition, a study of infectivity using low MOIs was conducted to determine a minimum MOI necessary for L. pneumophila replication in Acanthamoeba. Such MOIs may be more representative of ratios found in natural and human-made environments.
MATERIALS AND METHODS
Screening of samples of water, soil, sediment or biofilm for presence of infected amoebae was carried out by placing approximately 30 μl of water or slurry on nonnutrient agar streaked with heat killed Escherichia coli. The presence of intracellular amoebal pathogens was determined by examination of the agar plates at 100X and 400X magnification as previously described [5]. Agar plugs containing infected amoebae were removed with a scalpel and transferred to monolayers of axenic Acanthamoeba polyphaga to maintain the infection. Alternatively, plates were rinsed with a sterile buffer, and the rinse was placed into wells of 96-well microplates for easy observation under an inverted microscope. Suspensions of A polyphaga were added to wells in which infected amoebae from original plates were detected. Attempts to isolate the infecting bacteria were made using various media, including buffered charcoal yeast extract (BCYE) agar for Legionella species.
To determine a minimum MOI for infectivity, Legionella pneumophila (AA100, obtained from Y. Abu Kwaik) was cultured on BCYE agar at 37°C. Acanthamoeba polyphaga (ATCC 30461), was maintained by passage in Trypticase soy broth (TSB) at 25°C in 25 cm2 tissue culture flasks. Amoebae were harvested from TSB and suspended in sterile spring water at a concentration of 3 × 105 cells per ml and placed in multiwell tissue culture dishes. L. pneumophila was added at MOI's of 100:1, 10:1, 1:1, 0.1:1, 0.01:1, and 0.001:1. Cultures were incubated for 96 h at 30°C. The number of colony forming units (CFU) was determined by 10-fold serial dilutions of 100-μl aliquots of the Legionella/amoebae co-cultures in sterile spring water and spread onto BCYE agar.
RESULTS AND DISCUSSION
When soil, water, sediment or biofilm samples were placed on agar with E. coli, amoeba populations replicated and were easily observed after 72 h at 30°C. When viewed at 100X and 400X magnification, organelles such as the nucleus and vacuoles were clearly visible. In some samples amoebae appeared swollen and contained numerous bacteria that were highly motile. Such infected amoebae were also observed in microwells from washed plates. Often the internal morphological features of infected amoebae were not apparent, and the bacteria occupied a significant portion of the amoebic cytoplasmic space. When the amoebae were transferred to a monolayer of A. polyphaga the bacteria were capable of infecting the laboratory cultures of ameobae. During a typical replication cycle in amoebae, the intracellular bacteria rapidly increase in number between 24 h and 96 h of incubation at 30°C. Giemsa staining revealed that the bacteria initially accumulate within vacuoles. Differential interference contrast microscopy also revealed that infected amoebae of some samples contained motile bacteria throughout the cytoplasm, and were not confined strictly to food vacuoles. Over time the amoebae lost their ability to adhere to the surface. By 96 h after infection the infected amoebae lacked well defined internal morphological features such as a cell nucleus and vacuoles. The bacteria continued to be highly motile inside the amoebae. None of the amoebal parasites examined to date have been able to multiply on conventional laboratory agar media such as blood agar or Trypticase soy agar. Some isolates were capable of growth on BCYE, but the incubation time required for the appearance of colonies was longer (up to 7 days) than for appearance of L. pneumophila colonies. In cases where colonies were eventually isolated, confirmation that the agar-grown bacteria were indeed the amoebal parasites was obtained by re-infecting A. polyphaga with the bacteria and observing that the bacteria would not grow on blood agar. In addition, in situ hybridization techniques corroborated these observations. Two of the isolated amoeba pathogens, one from soil and one from a cooling tower, did appear to be novel strains of Legionella (Farone, M.B., Farone, A.L., Ventrice, J.A., Berk, S.B., Newsome, A.L., & Gunderson, J.H. 2001. A novel Legionella-like bacterium from infected soil amoebae. Abstract Q-420, p.669, American Society for Microbiology Annual Meeting, Orlando, F1. and Reed, A.D., Bowman, W.C., Gunderson, J.H., Wimberly, D.N., Berk, S.B., Newsome, A.L., Scott, T.L., & Johnson, R.A., 1999. Characterization of a Legionella-like amoebal pathogen form a hospital cooling tower. Abstract N-31, p.453, American Society For Microbiology Annual Meeting, Chicago, II.). Observations based on differential interference contrast microscopy also revealed that several of the samples contained amoebae infected with bacteria unlikely to be related to Legionella, as some bacteria were very large wide rods, others were coccoid, and molecular techniques showed that primers for Legionella sequences did not produce a PCR product. Various manifestations of the infections were observed and suggested that several non-legoinellae species were also agents of amoebal infections.
As previously reported (McNealy, T., Newsome, A.L., Johnson, R., & Berk, S.B., 2000. Impact of amoebae, bacteria and Tetrahymena on Legionella pneumophila multiplication in an aquatic environment. Abstract p-23, p. 40, 5th International Conference on Legionella, Ulm, Germany), a MOI of 100:1 resulted in a 1.8-fold increase in L. pneumophila CFU (from 1.4 × 108/ml to 2.5 × 108/ml). A reduction in MOI did not diminish the ability of L. pneumophila to infect and multiply within amoebae. A MOI to 0.1: 1 resulted in a 459-fold increase from 1.57 × 105 to 7.23 × 107 CPU ml. Decreasing the MOI to as little as 0.001:1 resulted in a dramatic increase (7,400- fold) in the number of recoverable L. pneumophila from an initial CPU count of 1.46 × 103 CPU/ml. The incubation period was increased to 144 h for the lowest MOI to allow numbers of CPUs to stabilize. At this lowest MOI tested 98% of all amoebae host cells were infected by 96 h and the number of intracellular bacteria were too numerous to count. At 144 h all amoebae trophozoites were lysed as a result of intracellular Legionella replication. In spring water at a low initial MOI some infected amoebae became encysted, and trapped motile L. pneumophila within the cysts.
The recovery of infected amoebae from a variety of habitats suggests that the bacteria responsible may be widespread in the environment. They might be expected to occur whenever free living amoebae are present, and cooling tower environments appear to favor such associations. The clinical relevance of these bacterial amoebic pathogens remains to be established. Recently, other investigators have shown that intracellular bacterial pathogens such as Chlamydia, mycobacteria, and Listeria may reside and benefit from association with free-living amoebae such as Acanthamoeba [4]. Collectively, these studies suggest amoebae may serve as hosts to a variety of intracellular bacterial parasites that are of potential significance to human health.
It is likely that L. pneumophila must have an appropriate host cell for multiplication in the environment. Laboratory studies have often used A. polyphaga as a host cell model. The ability of a low MOI to eventually infect a large population of amoebae may show how a low MOI could overwhelm an immunocompromised patient because of intracellular replication in macrophages. The finding that low MOI ratios could result in bacteria becoming trapped within cysts may affect measures taken to control such bacteria in human-made environments. In high dose MOI experiments, bacteria invade and may overwhelm amoebae before encystment can occur. Low MOIs, more likely to be encountered in the environment, may give the amoebae an opportunity to ingest bacteria and encyst prior to their destruction. Such entrapment within amoebae may provide protection from water treatment chemicals or biocides applied to reduce numbers of potential bacterial pathogens.
Overall, results of the work described above suggest that Legionella and LLAPs may represent only a small portion of the potential intracellular bacterial pathogens that can be associated with amoebae, and that such associations may be widespread. Knowledge of the details of the interactions, between these bacteria and amoebae, such as the finding that low MOIs result in bacteria being trapped in cysts will be important for designing useful procedures for the control of these potential pathogens in human-made environments.
ACKNOWLEDGMENTS
Although the research described in this article has been funded wholly or in part by the United States Environmental Protection Agency through grant number R 825352–01–0 and grant number R 827111–01–0 to Sharon G. Berk, it has not been subjected to the Agency's required peer and policy review. It therefore does not necessarily reflect the views of the Agency, and no official endorsement should be inferred.
