Vol. 61, No. 8

INFECTION AND IMMUNrrY, Aug. 1993, p. 3503-3510 0019-9567/93/083503-08$02.00/0 Copyright © 1993, American Society for Microbiology

Influence of Intra-Amoebic and Other Growth Conditions on the Surface Properties of Legionella pneumophila J. BARKER,1t P. A. LAMBERT,2 AND M. R. W. BROWN2* Birmingham Regional Public Health Laboratory, Heartlands Hospital, Binningham B9 5ST, 1 and Pharmaceutical Sciences Institute, Department of Phannaceutical Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, 2 United Kingdom Received 23 February 1993/Accepted 18 May 1993

The surface properties of LegioneUla pneumophila were examined by analyzing outer membrane (OM) proteins, lipopolysaccharides (LPS), and cellular fatty acids after growth within Acanthamoeba polyphaga and in vitro under various nutrient-depleted conditions. Intra-amoeba-grown legionellae were found to differ in several respects from cells grown in vitro; most notably, they contained a 15-kDa OM protein and a monounsaturated straight-chain fatty acid (18:19'). These compounds were also found in abundant quantities in the host amoeba. Immunoblot analysis of intra-amoeba-grown legioneUlae with antiacanthamoebic serum revealed that both the bacterial whole ceUs and Sarkosyl-extracted OMs contained amoebic antigens. The findings suggest that the 15-kDa OM protein is likely to be of amoebic origin and associates with the OM of the bacterium. It is proposed that disruption of amoebic membranes, as a result of intra-amoebic infection, may liberate macromolecules, including a 15-kDa poypeptide, a major constituent of the amoebic membrane, which adhere to the surface of the legionellae. Growth under specific nutrient depletions also had a significant effect on the surface composition of L. pneumophila. CeUs grown under phosphate depletion were markedly sensitive to protease K digestion and contained lower levels of LPS, as observed by silver staining of the digests on polyacrylamide gels. Intra-amoeba-grown ceUs contained more bands than the in vitro-grown organisms, reflecting further differences in the nature of the LPS. The whole-cel fatty acids of the phosphate-depleted ceUs were appreciably different from those of cells grown under other nutritional conditions. We found no evidence for expression of iron-regulated OM proteins under iron depletion.

the imposition of nutrient deprivations causes modifications in bacterial physiology. For example, alterations in the cell surface which increase the affinity of cell surface components for growth-limiting substrates so that they can be taken into the cytosol more effectively can occur (9). Thus, in nutrient-limited environments, cells grow more slowly and have altered cell envelopes (8, 10, 17, 22, 25). The purpose of this investigation was to examine the effects of environmental conditions on the physiology of L. pneumophila grown intra-amoebically and in vitro under nutrient-depleted conditions. Although inorganic phosphate is a nutrient commonly limiting in aquatic oligotrophic habitats, any nutrient limitation could occur in such environments (37). In addition to phosphate depletion, we studied the effects of nitrogen deprivation, because nitrogen is a major nutrient requirement of legionellae (36) and because legionellae do not utilize carbohydrates (a characteristic of oligotrophic organisms [37]). The molecular composition of the cells was determined by examining alterations in outer membrane proteins (OMPs), lipopolysaccharides (LPS), and cellular fatty acids, as it is well established that microbial surfaces play a vital role in bacterial survival and virulence (11). The origin of a 15-kDa L. pneumophila surface protein, found only in cells grown in amoebae, was investigated.

Legionella pneumophila has been shown to infect and multiply in a variety of amoebic hosts, especially the genera Acanthamoeba, Naegleria, and Hartmannella (1, 33, 39, 45). It is now widely acknowledged that this host-parasite relationship contributes to the propagation and distribution of legionella in natural and man-made water systems (18). Although the precise mechanisms for the survival and multiplication of legionellae in amoebae have not been fully elucidated, there is no doubt that considerable phenotypic changes occur upon comparison to cells grown in vitro. Intra-amoeba-grown legionellae are small and highly motile (3, 39), whereas in vitro-grown cells appear to be nonmotile and are often filamentous (4, 35). Previous work has shown that intra-amoeba-grown L. pneumophila is significantly more resistant towards biocide inactivation (3), including chlorine treatment (32), than cells grown in vitro. This is further evidence that in vivo growth has a profound effect on the physiology of legionella and produces an altered phenotype. Although little is known about the nature of the growth environment within the amoebic host, intracellular multiplication may be subject to iron restriction, as in human macrophages, through the activity of iron-binding proteins such as lactoferrin (12). Legionellae are also capable of surviving in low-nutrient, oligotrophic, aquatic environments (20) in which the rate of growth (if it occurs without association with amoebae) is likely to be governed by the availability of critical nutrients. It is widely recognized that *

MATERIALS AND METHODS Organisms and cultivation. The virulent strain of L. pneu-

mophila serogroup 1 (subgroup Knoxville), the causative organism in an outbreak of Legionnaires' disease in Stafford, England (2), was used throughout (available from Legionella Reference Laboratory, Central Public Health Laboratory, Colindale, London, United Kingdom). A different clinical

Corresponding author.

t Present address: Division of Biomedical Sciences, School of Science, Sheffield Hallam University, Sheffield S1 1WB, United Kingdom. 3503

3504

BARKER ET AL.

isolate, L. pneumophila serogroup 1 (subgroup Benidorm), was used for confirming some experiments. The strains were stored at -70°C as described previously (5), and recovery was achieved by growth in ABCD medium (36) at 35°C for 4 days. Cells from the initial culture were harvested by centrifugation (2,300 x g, 20 min), and the pellet was washed twice and resuspended in sterile distilled water to give an A660 of 1.0. This suspension was used to initiate batch cultures grown under nutrient-sufficient and -depleted conditions by setting the inoculum densities at an A660 Of ca. 0.03. ABCD synthetic medium was prepared lacking either phosphate or serine (the major nitrogen source of the medium), and following membrane filtration (0.2-p,m-pore-size filter), KH2PO4 or serine was added to the basal medium to give a final concentration of 0.2 or 2.0 mmol/liter, respectively. Iron restriction of growth was imposed on the cells by omitting both ferrous sulfate and hematin from ABCD broth. Cultures were shaken at 35°C, and growth was monitored spectrophotometrically at A660 until stationary phase (ca. 4 days), when the cells were harvested as described before. For some experiments, cells were also grown on plates of BCYE agar for 4 days (15). Acanthamoeba polyphaga was obtained from T. Rowbotham, Leeds Public Health Laboratory, Leeds, United Kingdom. This strain was associated with an outbreak of Legionnaires' disease in the United Kingdom caused by L. pneumophila serogroup 1. A. polyphaga was maintained axenically at 35°C in PYG broth (38) as monolayers in 75-cm2 tissue culture flasks (3). Intra-amoebic cultivation and harvesting of L. pneumophila were achieved as described previously (3). Essentially, L. pneumophila was initially grown in ABCD broth, and water-washed cells (105 CFU/ml) were inoculated into amoebic saline (38) suspensions of A. polyphaga (105 trophozoites per ml). Cultures were monitored by phase-contrast microscopy, which indicated the presence of infective, highly motile, intra-amoebic legionellae after 3 days of incubation at 35°C. The coculture was vortexed for 1 min to release legionellae from the amoebae. Legionellae were harvested from the suspension by centrifugation initially at 400 x g for 6 min at room temperature to deposit amoebic cells and debris. Legionellae were then deposited from the supernatant by further centrifugation (2,080 x g, 15 min, at room temperature). Examination of the cell pellet by phase-contrast microscopy showed it to contain no obvious amoebic cells or fragments. Legionella cells were routinely passaged in amoebic saline cocultures to produce cell pellets for analysis. As reported previously (3), no growth occurred in the saline suspension outside of the amoebic cells. LPS preparation. LPS was prepared from legionellae by protease K digestion of whole-cell lysates as described by Nolte et al. (34), using a modified sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) system. Variously grown legionellae (intra-amoebic and in vitro stationary-phase cultures) were harvested, washed, and suspended in water to give an A660 of 1.0. By reference to calibration curves of A660 versus total cell number, we had established previously that 1.0-A660 suspensions of cells grown under all nutritional conditions contained essentially the same number of cells per milliliter. A 1.5-ml volume of this suspension was centrifuged in a microcentrifuge (11,600 x g, 10 min). The supernatant was discarded, and the pellet was lysed by boiling for 10 min in 50 ,ul of 1 M Tris-HCl buffer (pH 6.8) containing 2% (wt/vol) SDS, 4% (vol/vol) mercaptoethanol, 10% (vol/vol) glycerol, and 0.05% (wt/vol) bromophenol blue. To the boiled lysate was added 10 Il of a

INFECT. IMMUN.

fresh solution of protease K (2.5 mg/ml in lysate buffer; Sigma). Digestion was carried out at 60°C for 1 h, and then the samples were centrifuged for 3 min to remove cell debris. LPS from Salmonella enteritidis grown in nutrient broth was similarly extracted. LPS was analyzed by SDS-PAGE with the buffer system of Lugtenberg et al. (27), using 5% stacking and 15% separating gels each containing 4 M urea. Two volumes of the LPS preparation were mixed with 1 volume of sample buffer containing 3% (wt/vol) SDS, 9% (vol/vol) mercaptoethanol, 30% (wtlvol) sucrose, and 0.003% (wt/vol) bromophenol blue in 120 mM Tris-HCl buffer (pH 6.8), and then the preparation was boiled for 10 min. Purified LPS from Eschenchia coli 0111 (Sigma) was run as a control. Electrophoresis was carried out in a Mini-Protean system (Bio-Rad Ltd.) at 10 mA per gel until the tracking dye was about 1 cm from the bottom of the gel. After electrophoresis, the gels were fixed in 40% ethanol-10% acetic acid and stained for LPS by the method of Tsai and Frasch (43). Fatty acid analysis. Whole cells of L. pneumophila, grown either in ABCD medium or intra-amoebically, and trophozoites ofA. polyphaga were washed in water and diluted to give suspensions with anA660 of ca. 1.0. These standardized suspensions were then freeze-dried, and accurate dry weights were obtained. For cells grown under various nutrient-depleted conditions and on BCYE agar, the standardized suspensions were used without freeze-drying. Fatty acids were extracted and converted to methyl esters, using the alkaline saponification method described by Moss et al. (30). Fatty acid methyl esters were analyzed by gas-liquid chromatography, using a 2.5-m (4-mm internal diameter) packed glass column of 3% SP-2100 DOH on 100/120 Supelcort (Supelco). A Pye Unicam chromatograph operating at 20-ml/ min nitrogen flow with a temperature gradient of 150 to 225°C at 2°C/min was used with a flame ionization detector. Peak areas were calculated with a CDP1 integrator (Pye Unicam). Individual fatty acids were identified by comparison of retention times with those of authentic bacterial standards (Supelco). For quantification of fatty acids, tridecanoic acid was used as an internal standard and added to the test suspensions before derivatization. Membrane proteins. Sarkosyl-insoluble membranes of legionellae and acanthamoebae were prepared with standardized suspensions, as described above for fatty acid analysis. Cells (1.5 ml) were disrupted ultrasonically in an MSE Soniprep disintegrator (MSE Scientific Instruments, Crawley, United Kingdom) operating with a 3-mm-diameter probe at maximum power and 20-pm amplitude. Five cycles of 30-s sonication were used, with intervening periods of 30-s cooling on ice. After sonic disruption, 0.15 ml of 20% (wt/vol) N-lauroyl sarcosinate, sodium salt, was added to dissolve the cytoplasmic membrane, as described by Lambert (26). Residual cell debris was pelleted by centrifuging the suspensions in a microcentrifuge (11,600 x g) for 3 min. The supernatant was then centrifuged at 11,600 x g for 2 h at 4°C in the microcentrifuge to deposit the outer membrane (OM). The deposited OMs were suspended in 25 ,ul of water and then mixed with an equal volume of sample denaturing buffer containing 4% SDS, 20% glycerol, 2% mercaptoethanol, 0.002% bromophenol blue, and 0.125 M Tris-HCl (pH 6.8). Samples were analyzed with a 5% stacking gel and a 12% separating gel in the Lugtenberg et al. buffer (27), and electrophoresis was continued at 30 mA until the tracking dye was 1 cm from the bottom of the gel. The gels were protein stained with 0.1% Coomassie blue in 50% ethanol10% acetic acid. OMs were also prepared by sucrose density gradient centrifugation of membranes prepared from soni-

VOL. 61, 1993

cated cell suspensions. Cells were broken as described above for the Sarkosyl method, and remaining unbroken cells were deposited by centrifugation at 5,000 x g for 10 min at 5°C. The supernatant containing cytoplasm and crude membranes was centrifuged (50,000 x g, 30 min) to concentrate the total inner membranes and OMs. The total membrane pellet was resuspended in 20% (wt/wt) sucrose and applied to the top of a 35 to 65% (wt/wt) sucrose density gradient prepared in seven steps (1.6 ml each) in a disposable polyallomer thin-walled ultracentrifuge tube (95 by 14 mm; volume, 14 ml). This was centrifuged in a swing-out rotor (Beckman SW40 Ti) for 18 h at 35,000 rpm and 5°C (Beckman L8 ultracentrifuge). The OMs formed a sharp band at a sucrose concentration of 50% (wt/wt). After the membranes were washed in water, they were examined by SDS-PAGE as described above. Investigation of the association of amoebic proteins with OMs. Cells of L. pneumophila were grown in ABCD broth, washed, and incubated in a suspension of sonically disrupted A. polyphaga in amoebic saline for 24 h at 22°C. During this time, no increase in bacterial cell numbers occurred, as no other nutrients were added. The bacteria were harvested by centrifugation, and OMs were prepared by the Sarkosyl extraction method. Immunoblotting. Cellular proteins of L. pneumophila and A. polyphaga separated by SDS-PAGE as described above were transferred onto nitrocellulose membranes (0.45-jim pore size; Bio-Rad Laboratories Ltd., Watford, Herts, United Kingdom), using the Western blotting (immunoblotting) procedure described by Towbin et al. (42). Transfer was carried out at pH 8.3 in 25 mM Tris buffer containing 192 mM glycine and 20% (vol/vol) methanol, using a Bio-Rad Mini Trans Blot apparatus at 100 V for 1 h at 4°C. To ensure that efficient transfer of proteins had taken place, the gels were stained with Coomassie blue after transfer. The nitrocellulose blots were blocked by washing in TTBS (Tween-Tris-buffered saline; 0.3% [vol/vol] Tween 20-0.9% [wt/vol] NaCl in 10 mM Tris-HCl, pH 7.4) for 1 h at room temperature. The blots were then given three 5-min washes in TBS (Tris-buffered saline; 0.9% NaCl in 10 mM Tris-HCl, pH 7.4). Reaction with antisera was carried out overnight at 4°C by immersing the blots in serum diluted 1:200 in TTBS. Sera used were raised in rabbits against whole cells of L. pneumophila serogroup 1, subgroup Knoxville, and against A. polyphaga. The antilegionella serum was obtained from T. G. Harrison, Central Public Health Laboratory, London, United Kingdom. It was raised against formalin-killed whole cells grown on BCYE agar. The antiacanthamoeba serum was obtained from D. Warhurst, London School of Hygiene & Tropical Medicine, London, United Kingdom. This was raised against sonicated whole cells ofA. polyphaga in saline with Freund's complete adjuvant. In addition, nonimmunized rabbit serum diluted 1:200 in TTBS was used as a negative control. After probing, the blots were given three 5-min washes in TBS and immersed in -TlBS containing 0.25 ,ug of staphylococcal protein A-horseradish peroxidase conjugate (Sigma) per ml for 3 h at room temperature. The blots were washed again in TBS as described above, and the color was developed by adding a solution containing H202 (0.01%, vol/vol) and 4-chloro-1-naphthol (25 ,ug/ml; Sigma) in 10 mM Tris-HCl buffer (pH 7.4). Fluorescence microscopy. Whole cells of intra-amoebagrown L. pneumophila were examined by indirect immunofluorescence microscopy by reacting with the rabbit antilegionella (subgroup Knoxville) serum (as used in the Western

SURFACE PROPERTIES OF LEGIONELLAE

3505

blot analysis) and goat anti-rabbit fluorescein isothiocyanate conjugate (Sigma).

RESULTS Fatty acid analysis. The effects of the growth environment on the ratio of cellular fatty acids in variously grown L. pneumophila whole cells are shown in Table 1. Although there was considerable variation in the relative amounts of fatty acids, reflecting differences in the growth conditions, 16-carbon-chain-length and certain branched-chain acids (a15:0, i-16:0, a-17:0, 18:0, and 20:0) were present under all growth conditions. Indeed, the branched-chain acids constituted 35.4% of the total acids in the intra-amoeba-grown cells and over 65% in the cells grown under iron depletion. For cells grown intra-amoebically, on BCYE agar, and under iron-depleted conditions, the most abundant fatty acid (25.2 to 35.3% of the total) was a saturated, branched-chain, 16-carbon acid (i-16:0). After growth in ABCD medium under nutrient-sufficient conditions, the 16-carbon straightchain saturated acid (16:0) was the most abundant (18.8%) and the i-16:0 acid was the next most abundant (16.2%). By contrast, under phosphate-depleted conditions, the monounsaturated, 16-carbon, straight-chain acid (16:19) was the predominant fatty acid (21.4%); 16:0 (16.8%) and i-16:0 (15.2%) acids were next. A major fatty acid component of intra-amoeba-grown legionellae was a monounsaturated, 18-carbon, straightchain acid (18:19) which made up 23.6% of the total and was the next most abundant acid after i-16:0 (25.2%). In addition, the amount of the branched-chain a-15:0 acid was considerably reduced (4.5%) in the intra-amoeba-grown cells compared with cells grown under iron-depleted conditions, when it constituted up 19.4% of the total. The fatty acid contents (percentage of dry weight) of legionellae and uninfected acanthamoebic trophozoites are shown in Table 2. Intra-amoebically grown legionellae contained about one-third less fatty acid in total than cells grown in ABCD medium, although they contained similar amounts of the i-16:0 branched-chain acid, a major fatty acid component of legionella species (19, 29, 31). Intra-amoebic growth reduced the amounts of a-15:0, a-17:0, and the monounsaturated 16:19 acid by up to two-thirds. Only 10 fatty acids were detected in A. polyphaga, and these were responsible for over 7% of the dry weight of the organism. However, the 18:19 fatty acid was a major component, accounting for 3.04% of the dry weight. Essentially similar results were obtained when the same experiments were carried out with L. pneumophila serogroup 1, subgroup Benidorm, grown in vitro and in A. polyphaga. LPS profiles. SDS-PAGE analysis of protease K-digested, whole-cell lysates of L. pneumophila revealed a compact banding pattern on silver staining which was much tighter than those of the S. enteritidis or E. coli controls (Fig. 1). The banding of the legionella LPS was seen only when samples were analyzed in 15% gels by the protease K digestion method of Nolte et al. (34). Lower concentrations of acrylamide in the separating gel produced an unresolved, densely stained band at the gel front in the position associated with rough LPS of enterobacteria. Although the LPS profiles of legionellae grown under various conditions were basically similar, there were some minor differences. The cultures grown under either nutrient-sufficient or iron-depleted conditions in ABCD medium, or on BYCE agar, produced LPS bands which were intensely stained with silver. Intra-amoeba-grown cells had about ten bands in the

BARKER ET AL.

3506

INFECr. IMMUN.

TABLE 1. Cellular fatty acid composition of L. pneumophila grown under various conditionsa Fatty acidb

3-OH-12:0 14:0 a-15:0 15:0

i-16:19 i-16:0

16:19 16:0

a-17:0 A17:0 17:0

18:19 18:0 19:0 20:0 Unknown Total branched-chain acids

Intra-amoebic

growth 2.3 1.9 4.5 0.5

Relative % of total fatty acids after growth under given condition ABCD medium BCYE agar Iron P04l3ACmeim BYagr depleted

depleted

N2 depleted

2.3 1.5 15.6 2.3

6.4 0.3 16.5 0.8

3.4 0.4 19.4 0.6

0 0 12.8 0

1.6 2.5 11.9 1.2

0.7 25.2 8.1 8.0

0.3 16.2 11.7 18.8

4.1 35.3 11.5 5.8

0.8 28.5 13.2 6.8

0 15.2 21.4 16.8

3.1 28.5 11.2 7.9

5.0

10.3

7.2

17.0

12.7

8.9

3.0

7.4

2.9

0

0

2.9

1.5

2.0

1.3

1.2

0

1.9

23.6 4.5 0.6 5.5

0 5.1 1.1 4.8

0.3 2.0 0.8 3.0

0.8 2.6 1.5 3.0

0 7.1 0 13.5

0 5.2 1.9 10.4

5.1

0.6

1.8

0.8

0.5

0.9

35.4

42.4

63.1

65.7

40.7

43.5

a All data presented were obtained reproducibly in at least three separate experiments. b For each fatty acid, numbers to the left of the colon refer to the number of carbon atoms, numbers to the right refer to the saturation number, and the superscript number gives the position of the double bond. i, iso branched; a, anti-iso branched; A, cyclopropane ring.

LPS ladder, whereas cells grown under other nutrient conditions in vitro had five to eight bands. The LPS from nitrogen-depleted cells and those grown intra-amoebically were stained with a similar intensity. Phosphate depletion had a profound effect on the cells, resulting in an LPS ladder which was only faintly stained. Presumably this reflects a greatly reduced LPS content. Quantitative comparison of the LPS profiles of cells grown under different conditions is made on the basis of the cell suspensions subjected to protease K digestion containing similar numbers of cells and the same dry weight of cellular material. No attempt was made to measure LPS content of the cell suspensions before protease K digestion because of the small numbers of cells available from intra-amoebic growth. We assumed that the silver stain reacts equally with LPS present in the cells grown under different conditions. It was noted that the variously grown legionella cells responded differently to protease K digestion. Phosphate-depleted legionellae and S. enteritidis were equally susceptible, producing no residual whole-cell deposit after digestion with the enzyme. By contrast, intra-amoeba-grown legionellae were much more resistant, with a small pellet of undigested cells remaining after treatment with the enzyme. Membrane proteins. SDS-PAGE protein profiles of Sarkosyl-prepared OMs from in vitro- and intra-amoeba-grown legionellae are shown in Fig. 2. As with the LPS profiles, we attempted to make a quantitative comparison of protein profiles of the variously grown cells by preparing suspensions containing the same number and dry weight of cells. We were unable to estimate the total protein content of the subsequent OM preparations because of the small amount of intra-amoeba-grown legionellae available. Therefore, SDSPAGE OMP analysis revealed by Coomassie blue staining shows the relative amounts of different proteins present in the OMs. A 29-kDa protein, presumed to be the major OMP

(16), was expressed under all growth conditions. The presence of a 15-kDa protein was a notable finding for intraamoeba-grown cells (lane 5), although the relative amount of this protein varied between batches of cells (cf. lane 6). This protein was not expressed in legionellae grown in vitro. Proteins in the range of 45 to 97 kDa were poorly expressed or absent in the intra-amoeba-grown bacteria. Minor protein bands at 21.5 and 14 kDa were exhibited in both the nitrogenand the iron-depleted cultures. Further SDS-PAGE analysis was done on whole-cell lysates and Sarkosyl-prepared fractions of in vitro- and intra-amoeba-grown legionellae and A. polyphaga trophozoites (Fig. 3). This confirmed the presence of a 15-kDa protein in the whole-cell and OM fractions of intra-amoeba-grown legionellae (lanes 5 and 6), but it also revealed that the amoebae possess a similar intensely stained protein band which is resistant to treatment with the Sarkosyl detergent (lane 2). To check whether the presence of an amoebic protein was an artifact introduced by the Sarkosyl OM preparation method, we also examined OMs prepared by sucrose density centrifugation. Examination of the OM band by SDS-PAGE again showed the presence of the 15-kDa protein, although the amount detected was considerably reduced compared with Sarkosyl-extracted OMs. When legionellae were incubated with a sonicated extract of amoebae, the 15-kDa protein was not detected in OMs prepared by either method. Western blots. The results of immunoblots with antiacanthamoeba serum against Sarkosyl-insoluble amoebic membranes and legionella OMs are shown in Fig. 4a. As expected, numerous antigens were strongly recognized in the amoebic whole-cell (lane 1) and Sarkosyl-insoluble (lane 2) membranes. However, the antiacanthamoeba serum did not appear to recognize the 15-kDa protein band in the Sarkosyl preparation even though it was an intensely staining band in

3507

SURFACE PROPERTIES OF LEGIONELLAE

VOL. 61, 1993

TABLE 2. Cellular fatty acid composition of L. pneumophila and A. polyphaga % Dry wt of the organism (relative amt of fatty acids as % of total)

A. poEyphaga grown in PYG broth

L. pneumophila grown in:

Fatty acida polyphaga

ABCD medium

0.09 (2.6) 0.07 (2.0) 0.165 (4.8) .0.02 (0.58)

0.11 (2.1) 0.08 (1.6) 0.78 (15.4) 0.115 (2.2)

0 0.45 (6.2) 0 0

A.

3-OH-12:0 14:0 a-15:0 15:0

i-16:19 i-16:0

0.025 (0.73) 0.83 (24.4)

0.015 (0.3) 0.81 (16.0)

0 0

16:0

0.32 (9.4)

0.94 (18.5)

0.35 (4.8)

a-17:0 A17:0 17:0

0.18 (5.2) 0.1 (2.9) 0.055 (1.6)

0.51 (10.0) 0.37 (7.3) 0.10 (2.0)

0 0 0

18:29,n12 16:19 18:19

0 034(0)

18:0

0.84 (24.7) 0.16 (4.7)

00.8114 0.15 (3.0) 0.25 (4.9)

0.31 0.13 (4.3) (10.7) 3.04 (41.9) 0.30 (4.1)

19:0 20:0

0.02 (0.58) 0.2 (5.8)

0.05 (1.0) 0.24 (4.7)

0 0

20:1 20:2 20:3 20:4

0 0 0 0

0 0 0 0

0.03 0.78 0.82 0.69

Unidentified

0

0

0.36 (5.0)

% of total dry wt

3.435 (100)

5.07 (100)

7.26 (100)

2.115 (41.7)

0

1.2 (35.13) Total branched-chain acids a See footnote b, Table 1, for explanation of fatty acid designations.

the corresponding Coomassie blue-stained SDS-PAGE gel (Fig. 3, lane 2). The recognition of at least eight distinct protein bands in the whole-cell intra-amoeba-grown legionellae by the antiacanthamoeba serum was a surprising finding (Fig. 4a, lane 5). The corresponding OM extract of intraamoeba-grown cells revealed two distinct protein bands and three less intensely stained bands (lane 6), but as with the 1

2

I

5

6

f 7

(0.4) (10.7) (11.3) (9.5)

amoebic preparations, the 15-kDa protein band was not recognized by the antiacanthamoeba serum. Two bands of 31.5 and 28 kDa were recognized more intensely by the antiacanthamoeba serum in the immunoblot of OMs prepared from intra-amoeba-grown legionellae (Fig. 4a, lane 6) than in the corresponding whole-cell extract (lane 5). The parallel Coomassie blue-stained SDS-PAGE gel (Fig. 3) revealed that the 15-kDa protein was the predominant protein; the 29-kDa major OMP was not as strongly expressed.

8 I

2

3

4

5

6

kDal

97,

66 45

31

21 I I. >

FIG. 1. SDS-PAGE analysis of LPS extracted from L. pneumophila. Growth conditions: lane 3, nutrient-sufficient ABCD medium; lane 4, iron-depleted ABCD medium; lane 5, nitrogen-depleted ABCD medium; lane 6, phosphate-depleted ABCD medium; lane 7, intra-amoebic growth; lane 8, BYCE agar. Controls: lane 1, E. coli-purified LPS; lane 2, S. enteritidis grown in nutrient broth.

-

-

29

-

-

15

FIG. 2. SDS-PAGE profiles of L. pneumophila Sarkosyl-extracted OMs. Growth conditions: lane 1, nutrient-sufficient ABCD medium; lane 2, nitrogen-depleted ABCD medium; lane 3, irondepleted ABCD medium; lane 4, phosphate-depleted ABCD medium; lanes 5 and 6, intra-amoebic growth.

.,-f:i>