Long-term survival of Legionella pneumophila serogroup 1 under low-nutrient conditions and associated morphological changes

Long-term survival of Legionella pneumophila serogroup 1 under low-nutrient conditions and associated morphological changes

FEMS Microbiology Ecology 102 (1992) 45-55 © 1992 Federation of European Microbiological Societies 0168-6496/92/$05.00 Published by Elsevier 45 FEMS...

1MB Sizes 0 Downloads 43 Views

FEMS Microbiology Ecology 102 (1992) 45-55 © 1992 Federation of European Microbiological Societies 0168-6496/92/$05.00 Published by Elsevier

45

FEMSEC 00419

Long-term survival of Legionellapneumophila serogroup 1 under low-nutrient conditions and associated morphological changes C h r i s t i n e P a s z k o - K o l v a 1,a, M a n o u c h e h r S h a h a m a t b a n d R i t a R. Colwell b a Aduanced Technology Laboratories, Alta Loma, CA, USA and b Department of Microbiology, Uni~'ersity of Maryland, College Park, MD USA and Center of Marine Biotechnology, Maryland Biotechnology Institute, Uniuersity of Maryland, Baltimore, MD, USA Received 14 April 1992 Revision received and accepted 15 September 1992

Key words: Legionella pneumophila; Oligotrophic; Survival; Tap water

1. S U M M A R Y Extended survival of Legionella pneumophila, using both a clinical and an environmental isolate, was studied in drinking water, creek water, and estuarine water microcosms. Legionella populations were monitored by acridine orange direct counts ( A O D C ) and viable count on buffered charcoal yeast extract agar amended with alphaketoglutarate ( B C Y E a ) . Initial colony counts of the clinical isolate in drinking and creek water microcosms were 2 × 10 8 c f u / m l and, after incubation for 1.5 years, the plate counts decreased to 3 x 10 6 c f u / m l . The A O D C counts, however, did not change significantly. The clinical isolate in

Correspondence to: R.R. Colwell, Department of Microbiology, University of Maryland, College Park, MD 20742, USA. Present address: Advanced Technology Laboratories, Alta Loma, CA 91701, USA and Metropolitan Water District of Southern California, Water Quality Division, LaVerne, CA 91750, USA.

estuarine water decreased in plate counts to 10 2 ( c f u / m l ) over the same period. After incubation for 1.5 years at 15°C in the microcosms, Legionella plate counts of creek and drinking water decreased by two logs. Direct microscopic examination of aliquots removed from all microcosms revealed the presence of small bacilli, large bacilli and rare filamentous cells. The environmental isolate demonstrated only one colony morphology upon culture on B C Y E a . Interestingly, after four months incubation in the microcosm, upon plating the clinical isolate on B C Y E a , two distinct colony types were evident. Examination by immunofluorescent staining employing a monoclonal antibody against L. pneumophila revealed both bacillus and filamentous forms. The total cellular proteins of both morphotypes were examined by sodium dodecyl sulfate polyacrylyamide gel electrophoresis (SDS-PAGE), demonstrating identical protein patterns. Those "Legionella cells remaining culturable during 1.5 years of incubation grew rapidly when transferred to B C Y E a . Incubation was continued and it was found that some strains of L, pneumophila serogroup 1 can

46 remain viable for longer than 2.4 years under low-nutrient conditions.

prolonged incubation conditions.

2. I N T R O D U C T I O N

3. M A T E R I A L S AND M E T H O D S

A large number of microbial ecology studies have been done which have established the ubiquity of legionellae in the environment [1-5]. However, very few studies have involved investigating the long-term survival and morphological changes of Legionella spp. in aquatic habitats [6,7]. Morphology, physiology and response of Legionella to physical and environmental parameters can change significantly with fluctuations in temperature, osmolarity, a n d / o r nutrient levels [8-10]. For example, Legionella spp. isolated from a variety of environmental sources exhibit different sensitivities to disinfectants [3,6]. Thus, longterm residence in a given environmental habitat may influence many aspects of cell metabolism and structure. Changes in Legionella morphology associated with nutrient concentration, nutrient composition, and redox potential have been observed [6,11]. Legionella cells grown in a chemostat revealed pleomorphism [6], whereas in batch cultures, Pine et al. [11] observed drastic changes in cell surface-to-volume ratio, shape, and size which could be related to temperature, nutrient, and aeration conditions for both bacillus and filamentous forms of Legionella. It is important to note that morphological changes are often the direct result of physiological or genetic selection, often in response to environmental pressure. Transformation from the bacillus to the filamentous form has been associated with a partial loss of virulence of Legionella pneumophila serogroup 1 [12,13]. In addition, only the bacillus form has been observed in patient specimens when viewed by immunofluorescent staining, while the filamentous form is only observed following subculture onto laboratory media [13] or detected by direct fluorescent antibody staining in environmental samples. The objective of this study was to examine the morphology and culturability of clinical as well as environmental isolates of L. pneumophila during

3.1. Strain selection

under

nutrient-limiting

Two strains of L. pneumophila serogroup 1 were employed, a clinical and an environmental isolate. The clinical strain, designated 320, was isolated from a Buffalo, NY, patient with Legionnaires' Disease. The environmental strain, designated 082, was isolated by swabbing a dripping sub-basement wall in a building on the University of Maryland at College Park campus. Both strains were maintained in aquatic microcosms and regularly monitored for 1.5 years, with periodical sampiing for up to 2.4 years. Both strains were confirmed as L. pneumophila by cysteine requirement, serotyping, G + C mol percent over-all D N A base composition, and DNA-DNA hybridization [14].

3.2 Microcosm inoculation and sampling Drinking water, creek water, and estuarine water microcosms were employed in the study. Drinking water was collected from a faucet in the laboratory of the Microbiology Building at the University of Maryland at College Park, MD. Creek water samples were collected from a creek which runs through the University of Maryland campus, and the estuarine water samples were collected from the Chesapeake Bay, using widem o u t h polypropylene containers (Nalgene, Rochester, NY) pre-cleaned with dilute hydrochloric acid. The estuarine samples were collected aboard the Research Vessel R / V Ridgeley Warfield at Tolly Point (30 56.'0 N, 76 26.'0 W) in the Fall of 1988. Temperature of the estuarine water sample was 11.3°C and salinity was 14.8 ppt. Both L. pneumophila strains were grown in buffered yeast extract broth (BYE), amended with alpha-ketoglutarate, for 48 h at 37°C in a humid environment with 2.5% CO 2. Cells were harvested by centrifugation at 8000 × g for 15 min and washed three times in sterile drinking water, to reduce carry-over of nutrients, before being

47 added to filtered (0.2 /.~m polycarbonate filters; Nucleopore, Pleasanton, CA), autoclaved (121°C for 15 rain) drinking, creek, or estuarine water microcosms [7]. Both total- and free-chlorine concentrations were always less than 1.0 ppm (Hanna Instruments, Woonsocket, RI) prior to autoclaving and never detected following autoclaving. The clinical strain, L. pneumophila 320, was inoculated at an initial concentration of l0 s cells/ml, into separate 1-1 Pyrex bottles (Corning Glass Works, Corning, NY) containing 900 ml autoclaved drinking water, creek water, or estuarine water. The environmental isolate was inoculated at an initial concentration of 107 cells/ml into 900 ml of drinking or creek water. Different initial concentrations of L. pneumophila were added to the microcosms to assess the effect of initial concentration on long-term survival and to evaluate metabolic activity once cells became non-culturable with [3H]thymidine-uptake experiments. Incubation of all five flasks was in the dark at 15°C and on a shaker (100 rpm) for the entire study period. At established intervals, samples were aseptically removed from the microcosms, cultured on B C Y E a for viable count (colony forming units per ml; c f u / m l ) [15] and simultaneously enumerated by A O D C (cells/ml) as described by Hobbie et al. [16] and by D F A (cells/ml) as described by Cherry et al. [17] using antisera obtained from Scimedix (Denville, N J) and Genetic Systems (Seattle, WA). Duplicate 0.1-ml aliquots of samples were spread-plated on B C Y E a , using sterile bent glass rods and the plates were incubated for one week, after which colony forming units (cfu) were counted. Sampling over a period of 323 days reduced the water volume in the microcosm to 100 ml. At this time, creek water was not available, so to extend the incubation period and to provide stringent incubation conditions (in an attempt to induce a non-culturable state) 800 ml of filtered, autoclaved drinking water was added to each microcosm. Again all microcosms were sampled and incubated as before, 15°C, in the dark and on a shaker (100 rpm). Following dilution of all microcosms, future enumerations of c f u / m l and immunofluorescent counts (cells/ml) considered

the dilution effect on the final calculations. Sampies of 10 ml were removed from the clinical and environmental drinking water microcosms at approx. 2.4 years, concentrated by centrifugation, and prepared for electron microscopy. Colonies recovered from microcosms on B C Y E a at day 735 were transferred to Legionella transparent medium (LTM) [18], incubated at 37°C in a humid environment with 2.5% CO2, and examined by scanning electron microscopy.

3.3. Gel electrophoresis A 72-h culture of each colony type obtained by plating from the 2.4-year-old microcosms (containing the clinical isolate) was harvested from BCYEc~ medium and the ceils suspended in 10 ml of SDS buffer (0.06 M Tris • HCL, pH 6.8, 2% SDS, 10% glycerol, 5% 2-mercaptoethanol, and 0.001% bromphenol-blue; Bio-Rad, Richmond, CA). The sample mixture and protein standards (Bio-Rad) were heated for 5 rain in a boiling water bath and centrifuged at 1000 × g for 10 rain at room temperature. Following centrifugation, aliquots containing 30 /zg of the protein sample and 10 ~g of the standard were loaded onto the gel. The samples and standards were subjected to a constant current of 60 mA for 3.5 h (Bio-Rad 3000/300 power source) using a 1.5-cm thick slab gel of 11.5% acrylamide as a running gel and a 5% acrylamide stacking gel in a discontinuous buffer system, a modification of the Laemmli method [19]. Following electrophoresis, gels were stained with Coomassie brilliant blue, destained, and photographed.

3.4. Electron microscopy Samples taken directly from the drinking water microcosms containing L. pneumophila strains 082 and 320, were concentrated by centrifugation (7000 X g for 20 rain) and plated on B C Y E a as previously described. Samples were then seeded on carbon-coated grids, stained with 1% phosphotungstic add, dried, and viewed under a Zeiss transmission electron microscope (TEM) (model EM-10CR, 80 KV). In the case of the clinical isolate (320) the two morphotypes were fixed in glutaraldehyde directly on plates (72-h growth on B C Y E a or LTM) prior to processing for scan-

48

(a)

10 9

10 9

"'"

,lO 8 cn

r.,.

o

G) O ilO

i0 v,

9 []

8

10 6

b)

i0

320

DW

320

DW

!

AODC Viable

10 6

!

10 9

.108

10 ~

'~

i0

7

7.

,107~ r.j

o

,10 6

10 ~. ,n i05

8

320 , 320

CW

AODC

CW,Viable

,

,

,

10 5

,c,

o IT...L

,10 9

- [

.10 8

.10 7 . 1 0 6 ~:= r~

i0 s

105~ o

104

104

o

103

103 320

EW

AODC

320

EW Viable

102

102 4 i0 I

,

0

i00

,

200

,

,

,

300

400

500

101 60q

Day

Fig. 1. Legionellapneumophila serogroup 1 (clinical strain 320) incubated in sterilized (a) drinking water, (b) creek water and, (c) estuarine water microcosms at 15°C for 535 days. Total counts (AODC) and colony counts ( c f u / m | ) are compared. The line at day 323 represents the addition of drinking water to each microcosm.

49

ning electron microscopy (SEM). Samples were fixed on poly-L-lysine coated glass coverslips in 4% glutaraldehyde, post-fixed in 1% osmium tetroxide, dehydrated in acetone in graded steps, and critical point dried. Samples were mounted on aluminum stubs, sputter-coated with gold-palladium, and examined by SEM using a Hitachi electron microscope (model $500, 20KV).

4. RESULTS In Fig. 1 (a-c) survival of the clinical strain of

L. pneumophila in sterilized drinking, creek, and estuarine water microcosms, respectively, is illustrated. Colony counts of the clinical isolate, shown in Fig. la declined slightly after incubation for approx. 300 days in drinking water. The most significant decline followed the addition of fresh drinking water, which severely diluted the nutrient pool as evidenced by the increased decline in culturable counts. The primary source of carbon and energy for the Legionella are amino acids. Legionella have been reported to exhibit brief periods of growth in both autoclaved drinking water [7], and in unsterilized drinking water [20]. As shown in Fig. lb, the clinical isolate incubated in creek water demonstrated a slow decline in culturable counts, while the same isolate incubated, in estuary water (Fig. lc) showed a sharper decline, i.e. a one log decline after incubation for 30 days. Total counts, measured by AODC, showed no significant change during the entire test period. After incubation for 1.5 years, the clinical strain showed a two-log decline in culturable counts of both drinking and creek water. Those cells incubated in estuarine water exhibited the greatest decrease in culturable counts (6 logs), approx, a 3-log decline after 1.5 years and an additional 3-log decline after the addition of the drinking water. Results of cultural and total counts of the environmental isolate (082; not shown) in drinking water and creek water showed a similar two-log decline in cultural counts, with no significant change in total count (AODC). After incubation for four months in the microcosms, only the clinical isolate (L. pneumophila strain 320) exhibited aberrant changes in colony

morphology. Two colony types were observed on B C Y E a : a dry, sticky colony, purple colored at the edges, revealing extremely filamentous cells upon Gram staining and a creamy, smooth, white-colored colony, which demonstrated rodshaped cells upon Gram staining. Both morphotypes were confirmed as L. pneumophila through DNA-DNA hybridization and immunofluorescent antibody staining with both polyclonal and monoclonal antisera against L. pneurnophila serogroup 1. In addition, contamination was not found following routine plating (throughout the study period) on both blood and nutrient agars (media on which Legionella does not grow). Interestingly, upon initial culture at 4 months the bacillus colony was predominant, however, over time the fila-

1

2

3

92. 5 k Da 66. 2 kDa

45 kDa

31 kDa

21 kDa

14 kDa

Fig. 2. Gel patterns of total m e m b r a n e proteins of Legionella pneumophila serogroup 1 (clinical strain 320), filamentous form (lane 1) and bacillus form (lane 2) stained with Comassie blue. Lane 3 contains protein standards.

50

mentous form became more apparent, and after a year and a half the ratio of bacillus to filamentous colonies on B C Y E a was close to 50:50; similar results were obtained by direct examination ( A O D C ) of the microcosms containing the clinical strains. Acridine orange staining of the

filamentous form (320), after culturing on B C Y E a for 72 h, showed bright, orange-red fluorescent filaments. The filaments were non-motile, and if agitated, the resultant smaller filaments were also non-motile and stained orange-red. Similar examination of the rod-shaped cells of the creamy

i/!! #

A

c

B

io

Fig. 3. Legionellapneumophila serogroup 1 (strains 320 and 082) incubated in drinking water for > 2.4 years, negatively stained with P T A and examined by TEM. (A) Environmental isolate (strain 082) exhibiting varying electron densities ( x 33000). (B) Cell (082) appears to be undergoing division ( × 20000). (C) Cells (320) with varying length and cell width and injured cell with variable electron transparency (X16000). (D) Clinical isolate ( 3 2 0 ) w i t h uniform electron density and convoluted surface structure ( x 50000). Bars = 0.5/xm.

51

colony, showed both green and orange fluorescing, rod-shaped cells when stained by acridine orange. The environmental isolate of L. pneumophila strain 082 did not exhibit alteration in colony morphology upon culture on B C Y E a at any time during the study, although bacillus and rare filamentous forms could be observed upon microscopic examination (AODC and DFA) of the microcosms. After the microcosms had been incubated for 1.5 years, fluorescent antibody staining was done, using samples from each microcosm and staining with anti-L, pneumophila SG1 polyclonal antisera. In aliquots removed from the microcosm, cells of both the clinical and environmental isolates stored in drinking water, stained well with polyclonal antisera by DFA, with a brilliant yellow-green color, while cells from creek and estuary microcosms (clinical strain only) stained slightly less brilliantly. DFA staining of samples from microcosms containing the clinical isolate revealed yellow-green, brightly fluorescing, rod-shaped and filamentous ceils. DFA staining of L. pneumophila serogroup 1 cells in microcosms was again done after incubation of the cells for 2.4 years. Again, the strongest staining reaction was observed for cells in drinking water, followed by creek and estuarine water.

To determine whether the variation of colony types observed for the clinical isolate were accompanied by changes in protein composition, the total cellular proteins were examined using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The protein patterns, shown in Fig. 2, were identical for both cell types. Cells of the environmental isolate (082) incubated in drinking water for 2.4 years were examined by electron microscopy and appeared electron transparent, with central electron dense areas (Fig. 3A, B). The clinical strain (320) of L. pneumophila, incubated in drinking water for 2.4 years, demonstrated varying cell lengths, a convoluted cell wall, and uniform electron density (Fig. 3C, D). After incubation for 2.4 years, an aliquot was removed from the drinking water microcosm containing the clinical isolate grown on B C Y E a for 72 h and the bacillus and filamentous colonies were identified. The bacillus form (strain 320) was isolated and negatively stained, revealing smooth, finely grained surfaces (Fig. 4A). The filamentous form of L. pneurnophila strain 320 was processed as described above and examined by electron microscopy, which revealed clumps of filamentous cells, with many filaments greater

Fig. 4. T E M examination of Legionella pneumophila strain 320 (clinical isolate) held for 2.4 years in the microcosm and grown on B C Y E a for 72 h. (A) Bacillus form with smooth surface and non-parallel sides ( x 10000). (B) Filamentous form ( x 2000). The bars in figure (A) and (B) represent 0.5/xm and 5.0 ~ m , respectively.

52

than 100 ~ m in length (Fig. 4B). L. pneumophila bacillus and filamentous forms were grown on B C Y E a for 72 h and examined by SEM as shown in Fig. 5A and 5B, respectively. The bacillus form was transferred from BCYEc~ to LTM media and incubated for 72 h. Examination by SEM showed

#

slightly elongated cells (Fig. 5C). The filamentous form of L. pneurnophila also grew well on LTM media, examination by SEM, revealed cells which reverted to the bacillus form without filaments (Fig. 5D).

f

Fig. 5. Legionella pneumophila strain 320 (clinical isolate) held for 2.4 years and grown on B C Y E a for 72 h, (A) Cells examined from creamy colonies by SEM are short and rod shaped; note cell with single polar flagellum, (B) Examination of 'sticky' colonies by SEM revealed filamentous cell morphology. (C) Bacillus colony from B C Y E a transferred to LTM for 72 h. (D) Filamentous colony taken from BCYEeL and transferred to LTM for 72 h; note reversion to the bacillus form on this media. Bar = 0.5 ~m.

53 5. D I S C U S S I O N The association of L. pneumophila with water and water systems is now well established [3,14,20-22]. Because legionellae are ubiquitous in the natural environment, it is often difficult to 'track', the source of this organism. The question has been raised as to whether L. pneumophila is continually re-seeded into water systems or, in fact, derives from or occurs naturally in aquatic reservoirs, i.e. is autochthonous. It was useful, therefore, to establish longevity of L. pneumophila in drinking water, surface water, and estuaries. Clearly, in natural systems, biotic factors, e.g. predation, play a significant role in the dynamics of Legionella populations [23]. In the studies reported here, the incubation temperature of 15°C was selected because it represented the average temperature of the environmental waters tested [23]. Large increases in viable counts were not observed in any of the microcosms. Viable counts of the clinical isolate incubated in estuarine water for 1.5 years showed the greatest decrease in culturable counts (over both drinking and river water), when plated on BCYEc~. Salinity was, therefore, concluded to be implicated in the decline in culturable count. In fact, it has long been known that the inherent toxicity of the sodium ion has an inhibitory effect on growth of L. pneumophila [15,24]. The addition of drinking water to the estuarine water at day 323 appeared to have little effect in diluting the toxic effects of the sodium ion, and the decrease in viable counts on B C Y E a continued. It is important to note that the survival of the Legionella cells in the microcosms after 323 days, when fresh drinking water was added to all microcosms, can only be interpreted with caution since the composition of the river and estuarine waters were altered. L. pneumophila cells, in general, are pleomorphic, with three morphological types (small rods, large bacilli, and filamentous rods) reported [6,11,22,25]. Anecdotal reports in the literature describe the emergence of filamentous forms of Legionella upon subculture of a clinical isolate on artificial media [11,12,26]; this is possibly due to recovery on antibiotic media. Initially, it was

speculated in these reports that nutrient concentration and composition were responsible for the observed morphological changes, based on the presence of distinct morphologies only under specific growth conditions in batch culture [11]. Other investigators studying the growth of Legionella in a chemostat, however, observed all three forms to occur simultaneously [6]. In our studies, all tl~ree morphological types were observed in microcosms inoculated with the clinical strain; however the large rods are believed to originate from the filamentous forms and were therefore grouped together. The pleomorphism observed in our studies appears to arise from fragmentation of filamentous cell. The filamentous bacilli were non-motile and extremely fragile, with rod-shaped cells appearing when cultures containing the filaments were shaken vigorously (Paszko-Kolva and Colwell, unpublished data). A O D C staining of the filamentous form revealed that all such cells stained fluorescent redorange, interpreted as indicating high R N A concentration and metabolically active cells, according to Hobbie et al. [16]. Examination of the bacillus form by A O D C showed yellow-green fluorescence, with some cells staining green (possibly metabolically inactive) and others orange (active). Because of the resistance of Legionella to nalidixic acid, it was not possible to determine the proportion of the viable but non-culturable cells enumerated by A O D C that remained substrate responsive [27]. Strong D F A reactions were observed for all samples throughout the study period. Positive D F A reactions suggests little or no change in cell-wall antigens throughout the 2.4 years that the samples were examined. However, surface antigen integrity may be affected by composition of the water, since intensity of the fluorescence was stronger in drinking water than creek or estuarine water microcosms. When total protein composition of the morphotypes of L. pneumophila were examined by SDS-PAGE, the total cellular protein composition was found to be identical for both morphological types. S D S - P A G E results showed no unusual proteins in the filamentous form. Legionella are facultatively oligotrophic bacte-

54 ria, often adapting quickly to fluctuations in nutrient concentration. Obligately oligotrophic bacteria are believed by some investigators to lack this ability [28]. Some facultatively oligotrophic bacteria are unable to grow in nutrient-rich media on initial isolation and will not be detected [29], which explains in part the difficulty in culturing Legionella even though it is observed to be present in large numbers when samples are examined by DFA. Many microorganisms in aquatic environments have been found to demonstrate survival strategies, including reductive division, increasing the number of high-affinity transporters, formation of coccoid cells, and cell elongation (increasing the surface to volume ratio, with the net effect of increasing nutrient uptake sites), and dormant cells or 'somnicells' [30-32]. Thus, nutrient concentration appears to be significant in regulating cell division, metabolism and morphology. To survive in a constantly changing aquatic environment, with attendant variations in nutrient concentration, as well as chemical and physical environmental factors, microbial cells must adapt rapidly to such changes. Other investigators [12,13] have shown that there is often a subpopulation of L. pneurnophila serogroup 1 cells which are able to revert to the filamentous form upon culture on laboratory media. Interestingly, this subpopulation can convert from a bacillus form (considered virulent) to a filamentous form (believed to be avirulent); this conversion is believed to be a two-way phenomenon. Interestingly, the filamentous conversion observed for clinical Legionella isolates upon transfer from oligotrophic microcosms to a high nutrient media may represent a nutrient response, with the filamentous form an adaptive response by the cell. The filamentous cells have been shown to be non-motile and avirulent. Again, the filamentous form may be a survival strategy employed by these bacteria. The non-motile, filamentous form is fragile, breaking into irregular-sized bacilli upon agitation, which may aid in the dissemination of this pathogen in the environment. However, the processing procedures (centrifugation or filtration) employed in Legionella detection may prevent the filamentous form from being routinely ob-

served in environmental samples when the samples have been processed for immunofluorescent staining. Nutrient 'shock' has been offered as an explanation for the observation that Legionella spp. detected by immunofluorescent staining are cultured only with great difficulty on nutrient media [21,33]. Hussong et al. [7] suggested that legionellae in the aquatic environment may enter a viable, but non-culturable state, akin to a form of sporulation. In addition to the survival strategies cited above, other tactics most certainly are employed by Legionella spp. for protection from environmental exigencies, such as changes in temperature a n d / o r pH, presence of disinfectant residuals, etc. Association with cyanobacteria, protozoa, and location within, as well as formation of, biofilms most certainly must be included in strategies employed by Legionella spp. for long-term survival in aquatic environments [1,5,14,22]. Whatever the strategy and regardless of culturability, environmental strains of Legionella can survive for periods greater than 2.4 years under low nutrient conditions and still be detected by direct fluorescent-antibody staining.

ACKNOWLEDGEMENTS The authors gratefully acknowledge the excellent assistance of Sue Fisher, Biological Sciences Division, University of California at Irvine, with the electron microscopy. The authors thank Mary Emad and Ellen Canter for assistance in typing the manuscript. This research was supported by a grant-in-aid for research to R.R.C. from the Department of Environmental Safety, University of Maryland and Cooperative Agreement No. CR817791-01 between the U.S. Environmental Protection Agency and the University of Maryland.

REFERENCES [1] Anand, C.M., Skinner, A.R., Malic, A. and Kurtz, J.B. (1983) Interaction of Legionellapneumophila and a freeliving amoeba (Acanthamoeba palestinensis). J. Hyg. 91, 167-168.

55 [2] Benson, R.F., Thacker, W.L., Waters, R.P., Quinlivan, P.A., Mayberry, W.R., Brenner, D.J. and Wilkinson, H.W. (1989) Legionella quinliuanii sp. nov. isolated from water. Curr. Microbiol. 18, 195-197. [3] Fliermans, C.B., Cherry, W.B., Orrison, L.H., Smith, S.J., Tison, D.L. and Pope, D.H. (1981) Ecological Distribution of Legionella pneumophila. Appl. Environ. Microbiol. 41, 9-16. [4] Shahamat, M., Paszko-Kolva, C., Keiser, J. and Colwell, R.R. (1991) Sequential culturing method improves recovery of Legionella spp. from contaminated environmental samples. Zbl. Bakt. 275, 312-319. [5] Tison, D.L., Pope, D.H., Cherry, W.B. and Fliermans, C.B., (1980) Growth of Legionella pneumophiila in association with blue green algae (cyanobacteria). Appl. Environ. Microbiol. 39, 456-459. [6] Berg, J.D., Hoff, J.C., Roberts, P.V. and Matin, A. (1985) Growth of Legionella pneumophila in continous culture. Appl. Environ. Micro. 49, 1534-1537. [7] Hussong, D., Colwell, R.R., O'Brien, M., Weiss, A.D., Pearson, A.D., Weiner, R.M. and Burge, W.D. (1987) Viable Legionella pneumophila not detectable by culture on agar media. Biotechnology 5, 947-950. [8] Csonka, C.N. (1989) Physiological and genetic responses of bacteria to osmotic stress. Microbiol. Rev. 53, 121-147. [9] LeChevallier, M.W., Cawthon, C.D. and Lee, R.G. (1988) Factors promoting survival of bacteria in chlorinated water supplies. Appl. Environ. Microbiol. 54, 649-654. [10] Warren, W.J. and Miller, R.D. (1979) Growth of Legionnaires' disease bacterium in a chemically defined medium. J. Clin. Microbiol. 10, 50-55. [11] Pine, L., George, J.R., Reeves, M.W. and Harrell, W.H. (1979) Development of a chemically defined medium for growth of Legionella pneumophila. J. Clin. Microbiol. 9, 615-626. [12] Bornstein, N., Nowicki, M. and Fleurette, J. (1984) Loss of virulence of Legionella pneumophila serogroup 1 with conversion of cells to long filamentous rods. In: Legionella Proceedings of the 2nd International Symposium. Atlanta, Georgia. (Thornsberry, C., Ballows, A., Feeley, J.C. and Jakubowski, W., Eds.), pp. 70-71. American Society for Microbiology, Washington, DC. [13] McDade, J.E. and Shepard, C.C. (1979) Virulent to avirulent conversion of Legionnaires' disease bacterium (Legionella pneumophila). Its effect on isolation technique. J. Infect. Dis. 139, 707-711. [14] Paszko-Kolva, C., Yamamoto, H., Shahamat, M., Sawyer, T.K., Morris, G. and Colwell, R.R. (1991) Isolation of amoebae and Pseudomonas and Legionella spp. from eyewash stations. Appl. Environ. Microbiol. 57, 163-167. [15] Feely, J.C., Gibson, R.J., Gorman, G.W, Langford, N.C. Rasheed, K., Mackel, D. and Baine, W.B. (1979) Charcoal-yeast extract agar: primary isolation medium for Legionella pneumophila. J. Clin. Microbiol. 10, 437-441. [16] Hobbie, J.E., Daley, R. and Jasper, S. (1977) Use of Nucleopore filters for counting bacteria by fluorescent microscopy. Appl. Environ. Microbiol. 33, 1225-1228. [17] Cherry, W.B., Pittman, P.P., Harris, G.A., Herbert, B.M., Thomason, L., Thacker, W.L. and Weaver, R.E. (1978)

Detection of Legionnaires' disease bacteria by direct immunofluorescent staining. J. Clin. Microbiol. 8, 329338. [18] Armon, R. and Payment, P. (1990)A transparent medium for isolation of Legionella pneumophila from environmental water sources. J. Microbiol. Methods 11, 65-71. [19] Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of the bacteriophage T4. Nature 227, 680-685. [20] Yee, R.B. and Wadowsky, R.M. (1982) Multiplication of Legionella pneumophila in unsterilized tap water. Appl. Environ. Microbiol. 43, 1330-1334. [21] Fliermans, C.B., Cherry, C.B., Orrison, L.H. and Thacker, L. (1979) Isolation of Legionella pneumophila from nonepidemic related aquatic habitats. Appl. Environ. Microbiol. 37, 1239-1242. [22] Wright, J.B., Ruseska, I, Athar, M.A., Cobert, S. and Costerton, J.W. (1989) Legionella pneumophila grows adherent to surfaces in vitro and in situ. Infect. Contol. Hosp. Epidemiol. 10, 408-415. [23] Paszko-Kolva, C., Shahamat, M., Yamamoto, H., Sawyer, T.K. and Colwell, R.R. (1991) Survival of Legionella pneumophila in the Aquatic environment. Microb. Ecol. 22, 75-83. [24] Saito, A., Rolfe, R.D., Edelstein, P. and Finegold, S. (1981) Comparison of liquid growth media for Legionella pneumophila. J. Clin. Microbiol. 14, 623-627. [25] Flesher, A.R., Ito, S., Mansheim, B.J. and Kasper, D.L. (1979) The cell envelope of the Legionnaires' disease bacterium. Morphologic and biochemical characteristics. Ann. Intern. Med. 90, 628-630. [26] Surgot, M., Barioz, C.M., Nowicki, M., Bornstein, N. and Fleurette, J. (1988) An electron microscopy study of Legionella pneumophila after in vitro and in vivo culture. Zbl. Bakt. Hyg. A 269, 26-33. [27] Kogure, K., Simidu, U. and Taga N. (1979) A tentative direct microscopic method for counting living marine bacteria. Can. J. Microbiol. 25, 415-420. [28] Ishida, Y., Imai, I., Miyagaki, T., and Kadota, H. (1982) Growth and uptake kinetics of facultatively oligotrophic bacterium at low nutrient concentrations. Microbiol. Ecol. 8, 23-32. [29] Kuznetsov, S.I., Dubinina, G.A. and Lapteva, N.A. (1979) Biology of oligotrophic bacteria. Ann. Rev. Microbiol. 33, 377-387. [30] Herbert, D. (1958) Some principles of continous culture. In: Recent Progress in Microbiology, VIIth International Congress of Microbiology, pp. 381-396. [31] Reichardt, W. and Morita, R.Y. (1982) Survival stages of a psychrotrophic Cytophaga johnsonae strain. Can. J. Microbiol. 28, 841-850. [32] Rozak, D.B. and Colwell, R.R. (1987,) Survival strategies of bacteria in the natural environment. Microbiol. Rev. 51,365-379. [33] Barbaree, J.M., Gorman, G.W., Martin, W.T., Fields, B.S. and Morrill, W.E. (1987) Protocol for sampling environmental sites for Legionellae. Appl. Environ. Micro. 51,422-424.