APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Mar. 1984, p. 467-471 0099-2240/84/030467-05$02.00/0 Copyright C) 1984, Amerian Soiety for Mirobiology Vol. 47, No. 3 Proliferation of Legionella pneumophila as an Intraellular Parasite of the Ciliated Protozoan Tetrahymena pyriformis BARRY S. FIELDS,1'2 EMMETT B. SHOTTS, JR.?,* JAMES C. FEELEY,2 GEORGE W. GORMAN,2 AND WILLIAM T. MARTIN2 Department of Medial Mirobiology, College of Veterinary Mediine, The University of Georgia, Athens, Georgia 30602,1 and Respiratory and Speial Pathogens Laboratory Branh, Division of Baterial Diseases, Center for Infetious Diseases, Centers for Disease Control, Atlanta, Georgia 303332 Reeived 11 August 1983/Aepted 8 Deember 1983 In a series of experiments, we have determined that Legionella pneumophila will proliferate as an intraellular parasite of the iliated holotrih Tetrahymena pyriformis in sterile tap water at 35 C. After 7 days of inubation, serpentine hains of _103 L. pneumophila ells were observed throughout the ytoplasm of the protozoan infeted initially with 1 to 30 L. pneumophila ells. The overall L. pneumophila population inreased from a. 1.0 x 102 to a. 5.0 x 104 ells per ml in the oulture within this time frame. The interations between the protozoan and the baterium appear to depend upon their onentratiohs as well as temperature of inubation. L. pneumophila did not multiply in sterile tap water alone, in suspensions of lysed T. pyriformis, or in ell-free filtrates of a T. pyriformis ulture. In addition to establishing an eologial model, we found that addition of T. pyriformis to environmental speimens served as an enrihment method that improved isolation of legionella from the speimens. Legionella pneumophila has been isolated from ooling towers, evaporative ondensers, streams, lakes, hospital shower heads, and soures of potable water (4, 7, 9). The ubiquitous distribution of L. pneumophila would appear to ontradit the stringent onditions required to grow the baterium in the laboratory (6). These observations have led some researhers to suggest that L. pneumophila is not a free-living aquati baterium (13). Tison et al. reported an assoiation of blue-green algae (yanobateria) and L. pneumophila and proposed that the baterium was utilizing algal extraellular produts as its arbon and energy soures (13). Rowbotham allowed amoebae of the genera Aanthamoeba and Naegleria to feed on L. pneumophila on an agar surfae (12) and observed that they beame heavily infeted or were lysed. From this he hypothesized that these amoebae ould be natural hosts for L. pneumophila and ould be a potential enrihment mehanism for the baterium in nature. Tyndall and Domingue subsequently reported the ability of L. pneumophila to infet the free-living amoebae Naegleria lovaniensis and Aanthamoeba royreba on nonnutrient agar plates (14). This study was initiated to determine whether L. pneumophila ould reat similarly with iliated protozoa in an aquati environment. We seleted the protozoan Tetrahymena pyriformis for study beause it (i) is found ubiquitously in the same types of water as Legionella spp. (1), (ii) feeds primarily on bateria, (iii) is easily maintained in axeni ulture, (iv) is extremely motile, and (v) is readily available beause of its wide use in biologial researh (10). MATERIALS AND METHODS Organisms and preparation of inoula. The L. pneumophila serogroup 1 (Philadelphia 1 strain) used in this study had been maintained by guinea pig passage and had never been grown on bateriologial agar. When needed, portions of infeted guinea pig spleen tissue whih had been quik frozen and stored at -70 C were inoulated onto a buffered * Corresponding author. haroal-yeast extrat agar plate and inubated for 3 days at 35 C under air and 2.5% CO2 (11). The T. pyriformis ultures were from a University of Georgia stok strain (no. 500, Midwest Culture Servie) grown in Elliot medium no. 2 at 25 C (5). When needed, T. pyriformis organisms were olleted on a 0.8-,um filter membrane (Millipore Corp., Bedford, Mass.) by gravity filtration, washed, and suspended in sterile tap water. In the experiment designed to determine optimum oulture onditions, some T. pyriformis organisms were grown in a suspension of heat-killed Esherihia oli (CDC no. A8188) in sterile tap water, and ells were olleted as desribed above. Mirosopy. We examined all oultures daily with a Zeiss mirosope. Slide preparations onsisted of a drop of the oulture smeared on a glass slide and stained by the Gimenez tehnique (8) or with immunofluoresent antiserum to L. pneumophila serogroup 1 (3). Determination of interation of legionella and protozoa. Two idential tests were performed, eah onsisting of one oulture, a suspension of L. pneumophila (ontrol 1), and a suspension of T. pyriformis (ontrol 2). Equal onentrations of the appropriate organisms were suspended in 25 ml of sterile tap water and inubated at 35 C. The numbers of extraellular L. pneumophila present at 0, 3, and 7 days were determined by removing 0.1-ml samples from eah suspension and ulturing them in tripliate on buffered haroalyeast extrat agar plates. The plates were inubated at 35 C under air and 2.5% C02, and the number of CFU was ounted after 7 days. The number of T. pyriformis present at these times was determined by plaing 0.1 ml of the suspensions on a glass slide and ounting the protozoa by visual inspetion with a Baush & Lomb disseting mirosope, (x8 to x40 magnifiation). Determination of nutritional soures. A 2-week-old ulture of T. pyriformis grown in 25 ml of Elliot medium no. 2 at 25 C was lysed by three onseutive freezings (-20 C) and thawings (25 C). The fragmented ell suspension was entrifuged at 1,500 rpm for 10 min. The supernatant was removed 467
468 FIELDS ET AL. APPL. ENVIRON. MICROBIOL. E 0~~~ 400- E 300- - CEIn s._ O 200-0 Oh. 0.0 2-00- E~~~TM.doii FIG. 1. Effets of oulture on the growth of L. pneumophila and T. pyriformis at 35 C. Symlbols: 0, L. pneumophila in oulture; *, L. pneumophila ontrol; - - -, T. pyriformis in oulture;, T. pyriformis ontrol. Values for this figure are ompiled data of two trials. and filtered through a 0.45-p>m Nalgene filter assembly. The resulting pellet of fragmented ells was suspended in 22.5 ml of sterile tap water by vortex blending. An inoulum of L. pneumophila (103/ml) was added to 22.5 ml of eah of the following: the suspension of lysed ells, the filtered supernatant, sterile tap water, and uninoulated Elliot medium no. 2. All were inubated and sampled as previously desribed to determine the number of CFU of L. pneumophila. Effets of temperature. A suspension of 2.5 ml ontaining a. i03 T. pyriformis organisms per ml was added to 22.5 ml of sterile tap water in eah of four test vessels and inubated for 7 days at 25, 30, 35, or 40 C, respetively. The numbers of protozoa present were determined as desribed above. Coultures were prepared by adding 2.5 ml of a suspension ontaining 10 T'. pyriformis organisms per ml to 22.5 ml of a suspension whih ontained 2.0 x 102 L. pneumophila ells per ml in sterile tap water. One vessel was inubated for 7 days at eah of the following temperatures: 25 C, 30 C, 35 C, 40 C. The suspensions were sampled to determine the number of organisms as desribed above. Optimization of test onditions. From the results of the first three experiments, we hypothesized that inreased multipliation of legionella in T. pyriformis oulture ould be ahieved by (i) prefeeding the T'. pyriformis on killed bateria, (ii) inreasing the ratio of T'. pyriformis to legionella, and (iii) preinubating the oulture at 25 C for 2 h before shifting it to 35 C. This was aomplished by adding 24 ml of i03 L. pneumophila ells per ml in sterile tap water to eah of four flasks. One flask (a) was seeded with 1.0 ml of T. pyriformis at 102 organisms per ml that were grown in a suspension of 1 ml of killed E. oli (107/ml) in sterile water; two flasks (b, ) eah were seeded with 7'. pyriformis at 102 organisms per ml grown in Elliot medium no. 2 at 25 C; and the fourth flask (d) was seeded with 7'. pyriformis grown in Elliot no. 2 medium to 103 organisms per ml. All oultures were immediately inubated at 35 C exept the oulture in flask that was inubated at 25 C for 2 h before being shifted to 35 C. Suspensions were sampled to determine the number of organisms as desribed above. Enrihment for naturally ourring L. pneumophila. Sam- TIME (days) FIG. 2. Effets of temperature on the growth of T. pyriformis in axeni ulture. Symbols: 0, 25 C; A, 30 C;, 35 C; - - -, 40 C. ples (25 ml) from eah of 10 water samples assoiated with an outbreak of Legionnaires disease and stored for 2 to 3 weeks at room temperature (25 C) were inoulated with 0.1 ml of a suspension ontaining 2.0 x 106 T. pyriformis. The samples were inubated at 25 C for 2 h and then held at 35 C for the remainder of the experiment. The samples were ultured for 0, 3, and 7 days by (i) ulturing 0.1 ml on eah of three plates of BCYE agar and (ii) ulturing 0.1 ml on eah of three plates of BCYE agar after aid treatment (2). Data were analyzed by the binomial test for determination of statistial signifiane. RESULTS The number of L. pneumophila CFU inreased by 3 to 4 logs within 1 week in eah of the oultures (Fig. 1). However, L. pneumophila did not multiply in axeni ulture in sterile tap water. The T. pyriformis organisms in both oultures and axeni ultures were destroyed over a 7-day period. Neither the suspension of lysed T. pyriformis ells nor the ell extrat of the spent T. pyriformis supported growth of L. pneumophila. T. pyriformis grew best at 250C in axeni uilture and in oulture with L. pneumophila. Higher temperatures redued the ability of T. pyriformis to multiply and survive (Fig. 2). Maximum multipliation of L. pneumophila ourred in oulture with T. pyriformis at 35 C (Fig. 3). Cl) -0 U, 0 -C, FIG. 3. Effets of temperature on the growth of L. pneumophila in oulture with T. pyriformis. Symbols: *, 25 C; A, 30 C; 35 C; -- -, 40 C.
VOL. 47, 1984 PROLIFERATION OF LEGIONELLA PNEUMOPHILA IN PROTOZOA 469 Ef 7- a 0. o E 6 L a)i e- - wl.-, 4- O 3. 0 -J 3 TIME (days) K-.- 1- - - FIG. 4. Effets of various fators on the growth of L. pneumophila in oulture with T. pyriformis. Symbols: 0, inreased onentration (103/ml) of T. pyriformis (broth grown) at 35 C; A, shift in temperature of inubations (25 to 35 C);, T. pyriformis (broth grown) at 35 C; - - -, T. pyriformis (bateria grown) at 35 C. Broth-grown protozoa supported greater multipliation of L. pneumophila (1 log) than the suspension of bateria-fed protozoa (Fig. 4). L. pneumophila CFU inreased by 5 logs when the temperature of inubation of the oulture was hanged, whereas inreasing the onentration of T. pyriformis by 1 log resulted in a final onentration of 6 logs of L. pneumophila per ml (Fig. 4). Using aid treatment and diret plating proedures before enrihment, we found that 5 of the 10 environmental water samples ontaining naturally ourring Legionella speies 7 L I were positive for the baterium. After 7 days of inubation with T. pyriformis, an additional 4 samples were found positive for Legionella speies, resulting in 9 of 10 positive samples (P = 0.0625). Immunofluoresent staining repeatedly revealed lusters of 10 to over 100 brightly staining baterial ells both within the ytoplasm of the protozoa and free in the menstruum of the oulture. The appearane of these lusters of ells was simultaneous with the onset of multipliation of L. pneumophila. Gimenez staining (8) of test samples indiated that these lusters of L. pneumophia ells were ontained in food vauoles of T. pyriformis. These same preparations also revealed many T. pyriformis organisms heavily infeted with L. pneumophila throughout their ytoplasm (Fig. 5 and 6). Some heavily infeted protozoa appeared as a sphere of baterial ells and were often partially ruptured. DISCUSSION Our study showed that L. pneumophila an infet and multiply within protozoa in an aquati environment. Detetion of large numbers of L. pneumophila in both food vauoles and ytoplasm of T. pyriformis and the inability of legionella to multiply on extraellular produts or lysed T. pyriformis ells support this onlusion. We also determined that temperature had an influene on the overall population of legionella in oulture with T. pyriformis. The fat that inreased multipliation of L. pneumophila ourred when oultures were inubated at 35 C may have been due to the ompromised ondition of the protozoa. The initial drop in T. pyriformis onentration after 24 h was attributed to the osmoti shok whih ourred when the organisms were removed from the ulture medium and plaed in the sterile tap water. Mirosopi examination of these oultures revealed that many protozoa were greatly enlarged or had ruptured a short time after being transferred to the tap water..~~ FIG. 5. T. pyriformis ontaining small numbers of L. pneuimoplhila within vauoles. Gimenez stain. N. Nuleus; L. L. pneumnophila; V, vauole. Bar, a. 10,um.
470 FIELDS ET AL. APPL. ENVIRON. MICROBIOL. N FIG. 6. T. pyriformis heavily infeted with L. pneumophila. L. pneumophila are in serpentine hains throughout the ytoplasm of the protozoa. Gimenez stain. L, L. pneumophila; N, nuleus; V, vauole. Bar, a. 10 p.m. However, surviving protozoa did ontinue to multiply for a period of less than 7 days. Beause inubation at 35 C will eventually destroy the Tetrahymena ells, it is diffiult to determine if their death is the result of temperature, infetion by L. pneumophila, or both. In ontrast, the inrease in the overall population of legionella that resulted when the protozoa were held at 25 C for 2 h before being shifted to 35 C an be explained by the fat that the protozoa at this temperature were very healthy and therefore atively ingested bateria. Our preliminary test showing that a protozoan an be infeted by a single ell of L. pneumophila and that a burst size of 102 to 103 legionella ourred for eah infeted T. pyriformis ell suggests the use of higher onentrations of T. pyriformis to obtain inreased numbers of L. pneumophila. Our study also determined that the oulture system ould be used to inrease the number of legionella in environmental water samples. However, doumentation of L. pneumophila multiplying in unsterilized tap water (14) has made it apparent that these tests should have inorporated dupliate samples inubated at 35TC without the addition of T. pyriformis. These ontrols would allow the determination of the singular effet of addition of the protozoa as an enrihment soure. Mirosopy revealed that the T. pyriformis ells were at least partially responsible for the multipliation of legionella in these samples. We are presently onduting further tests to determine the effiay of this method as a standard enrihment proedure. Our study douments that L. pneumophila an infet iliated protozoa in an aquati system and that our oulture system an serve as a laboratory eologial model. Furthermore, we suggest the oulture method ould be used as a laboratory enrihment for legionella in environmental samples. ACKNOWLEDGMENTS The authors thank D. Howard, Laboratory Program Offie, Center for Infetious Disease, Centers for Disease Control, Atlanta, Ga., for his tehnial assistane in photography. LITERATURE CITED 1. Bik, H. 1972. Ciliated protozoa; 74. World Health Organization, Geneva. 2. Bopp, C. A., J. W. Sumner, G. K. Morris, and J. G. Wells. 1980. Isolation of Legionella spp. from environmental water samples by low-ph treatment and use of a seletive medium. J. Clin. Mirobiol. 13:714-719. 3. Cherry, W. B., and R. M. MKinney. 1979. Detetion of the Legionnaires' disease bateria in linial speimens by diret immunofluoresene, p. 92-103. In G. L. Jones and G. A. Hebert (ed.), "Legionnaires' ": the disease, the baterium and methodology. Center for Disease Control, Atlanta. 4. flondero, T. J., Jr., R. C. Rendtorff, G. F. Mallison, R. M. Weeks, J. S. Levy, E. W. Wong, and W. Shaffner. 1979. An outbreak of Legionnaires' disease assoiated with a ontaminated air-onditioning ooling tower. N. Engl. J. Med. 302:365-370. 5. Elliot, A. M. (ed.). 1973. biology of Tetrahymena. Dowden, Huthinson and Ross, Stroudsbury, Pa. 6. Feeley, J. C., G. W. Gorman, R. E. Weaver, D. C. Makel, and H. W. Smith. 1978. Primary isolation media for Legionnaires disease baterium. J. Clin. Mirobiol. 8:320-325. 7. Fliermans, C. B., W. B. Cherry, L. H. Orrison, and L. Thaker. 1979. Isolation of Legionella pneumophila from nonepidemirelated aquati habitats. Appl. Environ. Mirobiol. 37:1239-1242. 8. Gimenez, D. F. 1964. Staining of rikettsia in yolk-sa ultures. Stain. Tehnol. 39:135-140. 9. Morris, G. K., C. M. Patton, J. C. Feeley, S. E. Johnson, G. W. Gorman, W. T. Martin, P. Skaily, G. F. Mallison, B. D. Politi, and D. C. Makel. 1979. Isolation of Legionnaires' disease
VOL. 47, 1984 PROLIFERATION OF LEGIONELLA PNEUMOPHILA IN PROTOZOA 471 baterium from environmental samples. Ann. Intern. Med. 90:664-666. 10. Nilsson, J. R. 1979. Phagotrophy in Tetrahymena, p. 339-379. In M. Lerandowsky and S. H. Hunter (ed.), Biohemistry and physiology of protozoa, 2nd ed., vol. 2. Aademi Press, In., New York. 11. Pasulle, A. W., J. C. Feeley, R. J. Gibson, L. G. Cordes, R. L. Myerowitz, C. M. Patton, G. W. Gorman, L. L. Carmak, J. W. Ezzell, and J. N. Dowling. 1980. Pittsburg pneumonia agent: diret isolation from human lung tissue. J. Infet. Dis. 141:727-732. 12. Rowbotham, T. J. 1980. Preliminary report on the pathogeniity of Legionella pneumophila for freshwater and soil amoeba. J. Clin. Pathol. 33:1179-1183. 13. Tison, D. L., D. H. Hope, W. B. Cherry, and C. B. Fliermans. 1980. Growth of Legionella pneumophila in assoiation with blue-green algae (yanobateria). Appl. Environ. Mirobiol. 39:456-459. 14. Tyndall, R. L., and E. L. Domingue. 1982. Coultivation of Legionella penumophila and free-living amoebae. Appl. Environ. Mirobiol. 44:954-959. 15. Yee, R. B., and R. M. Wadowsky. 1982. Multipliation of Legionella pneumnophila in unsterilized tap water. Appl. Environ. Mirobiol. 43:1330-1334.