Comparative assessment of virulence traits in Legionella spp.

O. A. Terry Alli, Steven Zink, N. Katherine von Lackum and Yousef Abu-Kwaik

Department of Microbiology and Immunology, College of Medicine, University of Kentucky, Chandler Medical Center, Lexington, KY 40503, USA

Correspondence
Yousef Abu Kwaik
yabukw{at}uky.edu


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Legionella pneumophila is a facultative intracellular pathogen that accounts for the majority of cases of Legionnaires' disease in the USA and Europe, but other Legionella spp. have been shown to cause disease. In contrast, Legionella longbeachae is the leading cause of Legionnaires' disease in Australia. The hallmark of Legionnaires' disease caused by L. pneumophila is the intracellular replication within phagocytes in the alveolar spaces, and the Dot/Icm type IV secretion system is essential for intracellular replication. Although it has been presumed that intracellular replication within phagocytes is the hallmark of other virulent legionellae, the virulence traits of Legionella spp. apart from L. pneumophila are not well defined. In this study, 27 strains of Legionella spp. belonging to 16 species that have been isolated from humans or from the environment were examined for five virulence traits exhibited by L. pneumophila: cytopathogenicity, intracellular replication within macrophages, induction of apoptosis/DNA fragmentation, pore-formation-mediated cytolysis of the host cell, and the presence of the dot/icm loci. The strains were divided into two broad groups (low and high cytopathogenic groups) based on cytopathogenicity assays using U937 human-derived macrophages. The other four virulence traits were evaluated in the low and high cytopathogenic groups of Legionella species. Most L. pneumophila serogroup 1 strains were highly cytopathogenic after 72 h, manifested high levels of intracellular growth, induced apoptosis/DNA fragmentation, and exhibited pore-forming activity. The majority of the other species were the low cytopathogenic group that did not induce apoptosis, neither did they exhibit pore-forming activity. All the species of legionellae tested have all the dot/icm loci, when examined by DNA hybridization. No correlation was found between cytopathogenicity and the other four pathogenic traits.


Abbreviations: sRBC, sheep red blood cell; TUNEL, terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labelling


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Legionellae are facultative intracellular Gram-negative bacteria that are ubiquitous inhabitants of natural and man-made water systems, where they replicate intracellularly within protozoa (Harb et al., 2000). From this environment, legionellae can be transmitted to humans by inhalation of infectious droplets. At least 48 species of legionellae have been identified, of which five species have been designated Legionella-like amoebic pathogens (LLAPs) comprising 70 distinct serogroups with more than half of the species of legionellae implicated in disease (Abu Kwaik, 1998; Adeleke et al., 1996; Benson & Fields, 1998; Lo Presti et al., 1999, 2001). Legionella pneumophila accounts for the vast majority of cases in most of the world, with L. micdadei ranking distantly second (Benin et al., 2002; Joshi & Swanson, 1999). L. longbeachae and L. dumoffii ranked third and fourth, respectively, in the latest survey of Legionnaires' disease in the USA (Benin et al., 2002). In contrast to the incidence of L. pneumophila in the USA, L. longbeachae is the predominant species of Legionella responsible for Legionnaires' disease in many regions of Australia (Doyle et al., 1998). The majority of the species of legionellae that are not well known have been implicated in disease; for example, L. tucsonensis has been isolated from a renal transplant patient (Thacker et al., 1989), and L. birminghamensis has also been isolated from a cardiac transplant recipient (Wilkinson et al., 1987). As a result of this, it is plausible to infer that a common theme of infection caused by Legionella spp. other than L. pneumophila is the ability to cause disease in immunocompromised hosts. An example of a Legionella species that has not been implicated in any disease in a human population is L. spiritensis. However, there is a possibility that the organism could cause disease if provided with the necessary conditions to thrive in host cells.

There are at least 25 000 cases of pneumonia due to L. pneumophila reported annually that require hospitalization (Abu Kwaik, 1998). This number is thought to be underestimated due to difficulties in bacterial isolation and diagnosis from secretions (Abu Kwaik et al., 1998; Jaulhac et al., 1992; Koide & Saito, 1995). The Centers for Disease Control and Prevention (CDC) figures between 1980 and 1998 show a general decline in mortality after reaching a peak in 1988 (Benin et al., 2002). In England and Wales, the legionellae surveillance carried out showed a similar trend to what has been reported in the USA, reaching its peak in 1988 (Joseph et al., 1995, 1997). One major recent outbreak of Legionnaires' disease occurred in the Netherlands during which 188 individuals attending a flower show were hospitalized. This outbreak was linked to a whirlpool spa and a sprinkler system used to water the flowers (Den Boer et al., 2002). People at risk of contracting Legionnaires' disease are elderly individuals, smokers, and people with underlying respiratory or immunocompromising conditions (Marston et al., 1994).

Infection by legionellae occurs after inhalation of contaminated aerosol, after which the organism enters and multiplies within alveolar macrophages. Within the intracellular environment of mammalian phagocytic (Horwitz, 1983) and alveolar epithelial cells (Gao et al., 1998), legionellae replicate within a rough endoplasmic reticulum (RER)-surrounded phagosome. The dot/icm loci are composed of 24 genes that are involved in the assembly of a type IV secretion apparatus that is central to pathogenesis. The dot/icm loci are essential for enhancement of phagocytosis by human-derived cells (Hilbi et al., 2001), macropinocytic uptake by A/J mice-derived macrophages (Watarai et al., 2001), evasion of lysosomal fusion and intracellular replication (Segal et al., 1998; Vogel et al., 1998), induction of apoptosis (Zink et al., 2002), and pore-formation-mediated lysis of the host cell and bacterial egress upon termination of intracellular replication (Alli et al., 2000; Molmeret et al., 2002b). Biphasic killing of mammalian cells by L. pneumophila (Alli et al., 2000; Gao & Abu Kwaik, 1999b) has been proposed in which apoptosis is first initiated, followed by a temporal induction of necrosis and lysis of the host upon growth transition into the post-exponential phase (Alli et al., 2000; Byrne & Swanson, 1998; Kirby et al., 1998). The pore-forming activity mediates lysis of the host cell, and mutants defective in pore-forming activity are defective in lysis of the host cell and are delayed in subsequent egress from mammalian (Alli et al., 2000) and protozoan cells (Gao & Abu Kwaik, 2000b). The C-terminus of IcmT has been shown to be essential for pore-formation-mediated bacterial egress from the host cell (Molmeret et al., 2002a, b). Importantly, pore-forming activity plays a major role in pulmonary cytotoxicity and inflammation in experimental animals (Alli et al., 2000; Molmeret et al., 2002b). These pathogenic mechanisms have been studied for L. pneumophila but little is known about their role in the pathogenesis of other species of legionellae.

Based on previous studies on L. dumoffii and L. micdadei, the general assumption is that replication within phagocytes is the hallmark of virulent legionellae (Levi et al., 1987; Moffat & Tompkins, 1992; O'Connell et al., 1995). The pathogenic traits of other Legionella spp. apart from L. pneumophila are not well defined. Previous studies have addressed the replication of legionellae within human macrophage cell lines (O'Connell et al., 1996) and guinea pigs (Doyle et al., 2001; Izu et al., 1999). The growth of Legionella spp. in protozoan and human macrophage cell lines has also been compared (Neumeister et al., 1997). However, these studies have examined a single virulence trait in the limited number of strains of legionellae studied. In this study, we aimed to define traits responsible for the pathogenesis of Legionella spp. in 27 strains belonging to 16 species, by examining five pathogenic traits: cytopathogenicity, intracellular multiplication in U937 macrophages, pore-formation-mediated cytolysis, the induction of apoptosis, and the presence of the dot/icm loci. We selected strains of legionellae that have been shown to cause disease along with strains that have not been associated with Legionnaires' disease with the hope of defining the virulence trait(s) that differentiate L. pneumophila strains from the rest. We found no virulence trait that can differentiate L. pneumophila from the remaining species of legionellae using the five aforementioned pathogenic traits.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, cell culture and media.
The Legionella spp. used in this study with the exception of L. pneumophila serogroup I AA100 and GR159 (a mutant derived from AA100) were a kind gift of R. Benson, B. Fields and J. Pruckler from the Centers for Disease Control and Prevention (CDC); the passage history of the strains has been described before (O'Connell et al., 1996). All the strains used have been implicated in human disease with the exception of L. adelaidensis, L. gratiana, L. moravica, L. parisiensis, L. santicrucis and L. spiritensis. For the details of the strains used, see Table 1. The strains have been passaged not more than three times in all cases. All strains of legionellae were grown on buffered charcoal-yeast extract (BCYE) plates or in buffered yeast extract (BYE) broth. U937 cells were cultured at 37 °C in RPMI-1640 containing 10 % fetal bovine serum (Gibco) in a humidified atmosphere containing 5 % CO2. U937 cells were differentiated with phorbol 12-myristate 13-acetate (Sigma) for 48 h before use.


View this table:
[in this window]
[in a new window]
 
Table 1. Strains of Legionella species used

 
Cytopathogenicity assay.
Legionella spp. were grown on BCYE plates for 3 days prior to infection of U937 macrophages. Infection was performed, in triplicate, in 96-well plates containing 105 cells per well at an m.o.i. of 1 for 1 h for U937 followed by gentamicin (50 µg ml-1) treatment for 1 h, followed by three washes to remove extracellular bacteria and further incubation at 37 °C in 5 % CO2. At several time points, the monolayers were treated for 4 h with 10 % Alamar Blue dye (Alamar Bioscience) as previously described (Abu Kwaik & Pederson, 1996). Viability of the monolayers was determined by measuring the absorbance of Alamar-Blue-treated monolayers by using a VMAX Kinetic Microplate reader (Molecular Devices) and expressed as percentage cell death compared with uninfected cells by using the formula [1-(mean absorbance of treated cells/mean absorbance of untreated cells)]x100.

Growth kinetics of Legionella spp. in U937 macrophages.
Infections of U937 macrophages by Legionella spp. were performed, in triplicate, in 96-well plates containing 105 cells per well at an m.o.i. of 1 for U937 macrophages as previously described (Alli et al., 2000). At the end of the infection period, the monolayers were treated with gentamicin (50 µg ml-1) for 1 h as described above. The number of bacteria in the monolayers at several time intervals after washing of the gentamicin was determined.

Contact-dependent pore formation assay.
Contact-dependent pore formation in the plasma membrane was determined by examining haemolysis of sheep red blood cells (sRBCs) by Legionella spp. at an m.o.i. of 25 following 2 h of bacterial–sRBC contact at 37 °C, as previously described (Kirby et al., 1998). Briefly, sRBCs (Remel) were diluted in RPMI 1640 medium, and washed three times by centrifugation for 10 min at 2000 g until the supernatant did not show any sign of haemolysis; the cells were then counted with a haemocytometer. Reactions were set up in a final volume of 1 ml with final concentrations of 1x108 sRBC ml-1 and 2·5x109 bacteria ml-1 and incubated at 37 °C for 2 h. At the end of incubation period, the pellets were resuspended by vortexing and repelleted by centrifugation for 2 min at 17 000 g. Supernatants were transferred to cuvettes and the absorbance was read at 415 nm.

DNA fragmentation analysis and TUNEL assays.
DNA fragmentation analysis was carried out as previously described (Gao & Abu Kwaik, 1999b). Differentiated U937 macrophages were plated in six-well plates (1·5x106 cells per well) and infected with Legionella spp. at an m.o.i. of 50 for 1 h. At the end of the infection period, the monolayers were washed three times to remove unattached extracellular bacteria and maintained at 37 °C in culture medium. At 3 h post-infection, the cells in each well were lysed in 500 µl lysis buffer [10 mM Tris (pH 7·5), 20 mM EDTA (pH 8·0), 0·5 % Triton X-100] for 30 min on ice. The lysates were treated with 0·5 % SDS and 300 µg proteinase K ml-1 for 2 h and DNA extracted with phenol and chloroform before precipitation with ethanol. The precipitates were dissolved in 10 mM Tris (pH 8·0)/1 mM EDTA containing 0·5 µg RNase ml-1, electrophoresed in 1·8 % agarose gel, and stained with ethidium bromide; individual lanes were examined for the presence of DNA fragmentation. DNA fragmentation was scored as positive in each strain when compared with the negative control that was not infected with any bacteria.

Terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labelling (TUNEL) assays were performed exactly as we described previously (Gao & Abu Kwaik, 1999b). Briefly, cells attached to 96-well plates were infected for 1 h with the strains of bacteria at an m.o.i. of 50, and then extracellular bacteria were washed off. For labelling of apoptotic nuclei, the cells were subjected to FITC-conjugated TUNEL using an apoptosis detection kit, according to the manufacturer's instructions (Boehringer Mannheim). Cells were examined using an Axiovert S100 Zeiss fluorescence microscope. A minimum of 100 cells per sample was counted, and apoptosis was quantified as the percentage of apoptotic cells (TUNEL-positive nuclei). The ability of the bacteria to induce apoptosis was scored as positive if the percentage of apoptotic cells was greater or equal to 50 %. Multiple independent samples were examined.

DNA hybridization analysis.
The dot/icm genes have been shown to play an important role in the virulence of L. pneumophila (Segal et al., 1998; Vogel et al., 1998). We screened all the Legionella spp. using Southern hybridization probed with the PCR products derived from all regions of the dot/icm loci from L. pneumophila AA100. A cocktail of probes that contained 2·1 kb PCR product of icmTSRQ, 1·6 kb PCR product of dotDCB, 1·8 kb PCR product of icmJB, 1·0 kb PCR product of icmWX and 1·0 kb PCR product of icmLK was used. The 2·1 kb PCR product of icmTSRQ was generated using primers P1 (5'-CACAGTTAAAACTTCAAGCTGAACC-3') and P2 (5'-CTGCTCAGAGCTATTTTT-3'). The 1·6 kb PCR product of dotDCB was generated using primers P3 (5'-CGATTGGTCTGGTCCGATTGA-3') and P4 (5'-TCTCGAATAATGGAAGCTAACAATGTC-3'). The 1·8 kb PCR product of icmJB was generated using primers P5 (5'-TGCCATGTTCTTTTTTGTGCTATTAC-3') and P6 (5'-GAGCGTAAACCAGATCAATCCAAGTAG-3'). The 1·0 kb PCR product of icmWX was generated using primers P7 (5'-TGGGTTGGTTCCTGAGGTATGA-3') and P8 (5'-TGGGGCGCTGAAATTTTGATAT-3'). The 1·0 kb PCR product of icmLK was generated using primers P9 (5'-CGGAAGGCTGGGACCAATT-3') and P10 (5'-CCACTCGATAATCCACGGCTTTC-3'). Labelling of DNA probes and Southern hybridizations were performed as described previously (Abu Kwaik et al., 1997). High-stringency hybridization and washes were performed at 60 °C; low-stringency hybridization and washes that allowed for 20 % mismatch were performed at 42 °C.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Cytopathogenicity of Legionella spp. to U937 macrophages
On the basis of the results in Fig. 1 and Table 2, the Legionella spp. were divided into two main broad groups. The six group I strains were able to kill 50 % or more of the U937 macrophages (Fig. 1a), while the remaining group II strains were able to kill 49 % or less of the U937 macrophages after 72 h of infection (Fig. 1b–d). Members of the group I were L. pneumophila serogroup 1 (AA100, SG1-62, SG1-66 and SG1-67), L. micdadei Rivera strain and L. dumoffii. Interestingly, not all L. pneumophila serogroup I strains belonged to group I (Fig. 1b), suggesting variation in the degree of virulence. All the four strains of L. longbeachae tested in this study belonged to group II.



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 1. Cytopathogenicity of Legionella spp. to U937 macrophages. Values for cytopathogenicity are expressed as the percentage of host cells killed as a result of the infection, compared with uninfected cells as determined by Alamar Blue assays following 72 h of infection. These data are representative of at least three independent experiments performed in triplicate. The absence of error bars indicates very small standard deviations that could not be displayed. Abbreviations used in Figs 1–3 are: L. adelaide., L. adelaidensis; L. b'mensis, L. birminghamensis; L. cincinn., L. cincinnatiensis; L. longb., L. longbeachae; L. m., L. micdadei; L. maceach., L. maceachernii.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Pathogenic traits of Legionella spp.

 
Intracellular growth kinetics of Legionella spp.
Four patterns of growth within the U937 macrophage cells emerged when compared with the virulent L. pneumophila AA100 strain (Fig. 2a–f). The first pattern (Table 2) comprised strains that were able to grow in an unrestricted manner like AA100. These strains had an intracellular growth index (c.f.u. at 72 h/c.f.u. at 0 h) greater than 100 after 72 h of macrophage infection. This group comprised the majority of L. pneumophila serogroup 1 including the mutant GR159, L. micdadei strains, L. parisiensis, L. hackeliae, L. longbeachae strains, L. maceachernii, L. cherrii and L. dumoffii (Table 2, Fig. 2). The bacterial strains with the highest growth index (106) were L. pneumophila AA100 and L. micdadei (Rivera and Tatlock) strains. The results in Fig. 2(a), for group I strains according to the above classification, show that L. pneumophila SG1-62 was able to reach the highest level of intracellular infection in 24 h, followed by a decline, which was in sharp contrast to all the other strains in this group. The second pattern was those of strains that were unable to replicate in the first 24 h of infection but increased their number by 72 h post-infection. L. birminghamensis exhibited this growth pattern within macrophages. The third pattern was for those bacterial strains that maintained their numbers without a detectable change for 72 h. The numbers increased slightly after 24 h of infection but their numbers declined afterwards. L. pneumophila SG1 Knoxville and L. cincinnatiensis exhibited this pattern of growth within macrophages (Table 2, Fig. 2c, d). The growth index at 72 h in this case could be misleading because it did not show a decline in growth after 24 h for L. cincinnatiensis. L. spiritensis did not show appreciable growth after 72 h infection. The fourth most notable pattern was for those species that were unable to grow within macrophages and were gradually killed as shown by a decline in their viability. These strains were L. wadsworthii, L. gratiana and L. santicrucis, with intracellular growth indices ranging from 10-1 at 24 h to 10-5 at 72 h (Table 2, Fig. 2c, e).



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 2. Intracellular growth kinetics of Legionella spp. in U937 macrophage-like cells. These data are representative of at least three independent experiments performed in triplicate. The absence of error bars indicates very small standard deviations that could not be displayed. See Fig. 1 for abbreviations.

 
Induction of apoptosis in U937 cells
Apoptosis during early stages of the infection and pore-formation-mediated cytolysis upon termination of intracellular bacterial replication have been implicated in death of the host cell infected by L. pneumophila (Alli et al., 2000; Gao & Abu Kwaik, 1999b). DNA fragmentation by gel electrophoresis and TUNEL assay were used as two independent indicators of apoptosis in infected U937 macrophages (Gao & Abu Kwaik, 1999b). Results of both assays were consistent for all the strains (Table 2). The results showed that only L. pneumophila serogroup 1 strains (AA100, SG1-62, SG1-66 and GR159), along with L. spiritensis and L. moravica, induced apoptosis while L. pneumophila SG1-64, SG1-65 and SG1-67 did not (Table 2). All the other strains used in this study did not induce apoptosis (Table 2), suggesting that there is no correlation between cytopathogenicity and induction of DNA fragmentation/apoptosis.

The pore-forming activity of Legionella spp.
The pore-forming activity of L. pneumophila has been shown to contribute to cytotoxicity (Kirby et al., 1998) and the ability of the organism to egress from the host cell after cessation of intracellular replication (Alli et al., 2000; Gao & Abu Kwaik, 2000a). Contact-dependent haemolysis of sRBC assay was performed to examine pore-forming activity, as previously described (Kirby et al., 1998). The L. pneumophila rib mutant GR159, which is defective in pore-forming activity (Alli et al., 2000) was used as a negative control in this assay. The group I and II strains demonstrated variable pore-forming activity (Fig. 3a, b, d). Taken together, the data showed that there was no correlation between pore-forming activity and cytopathogenicity. The data also showed that only the L. pneumophila serogroup 1 (AA100, SG1-62, 65, 66 and 67) strains along with L. spiritensis had pore-forming activity, while L. pneumophila SG1-64 and Knoxville strains did not (Fig. 3). All the other strains tested in this study did not show any significant pore-forming activities.



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 3. Pore-forming activity of Legionella spp. as determined by contact-dependent haemolysis of sRBC measured as A415. These data are representative of at least three independent experiments performed in triplicate. The absence of error bars indicates very small standard deviations that could not be displayed. See Fig. 1 for abbreviations.

 
Prevalence of the dot/icm genes in Legionella
Southern analyses revealed the presence of two to six EcoRI genomic fragments hybridizing to the dot/icm loci under low and high stringencies in all Legionella spp. (Fig. 4a, b and data not shown). More than two bands were observed for all the L. pneumophila SG1 strains including AA100, while two to six bands were observed in other species of Legionella at both low and high stringencies. All the L. pneumophila SG1 strains, including the Knoxville and AA100 strains, used in this study had the same pattern of restriction fragment length polymorphism bands. When the dot/icm probes were used individually in Southern hybridization, they all hybridized to all chromosomal DNA of all the Legionella spp. (data not shown). The data show that the dot/icm loci are present within the Legionella spp. and not only limited to L. pneumophila.



View larger version (84K):
[in this window]
[in a new window]
 
Fig. 4. Southern blot analysis for the presence of dot/icm genes in Legionella spp. The mixture of PCR products from different regions of the dot/icm loci was hybridized to EcoRI-digested chromosomal DNA at low stringency (a) and high stringency (b). Lanes: 1, L. dumoffii; 2, L. longbeachae Long Beach 4; 3, L. micdadei (Tatlock strain); 4, L. spiritensis; 5, L. maceachernii; 6, L. santicrucis; 7, L. parisiensis; 8, L. pneumophila SG1 Knoxville; 9, L. hackeliae; 10, L. gratiana; 11, L. moravica; 12, L. cincinnatiensis; 13, L. cherrii; 14, L. micdadei (Rivera strain); 15, L. pneumophila SG1-62; 16, L. pneumophila SG1-64; 17, L. pneumophila SG1-65; 18, L. pneumophila SG1-67; 19, L. pneumophila AA100.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Two phases of macrophage killing by L. pneumophila have been proposed: the first phase is through caspase-3-mediated induction of apoptosis during early stages of infection (Gao & Abu Kwaik, 1999a, b); the second phase is characterized by pore-formation-mediated lysis and bacterial egress from the host cell upon termination of intracellular replication (Alli et al., 2000; Molmeret & Abu Kwaik, 2002). In view of this, we examined these phenotypic traits along with intracellular replication in macrophages and presence of the dot/icm genes to elucidate the virulence traits for different species and strains of legionellae. The classification of the bacterial strains used in this study into two broad groups provides important clues to the pathogenic features of some of the Legionella spp. Group I comprised the most pathogenic strains used in this study and was dominated by L. pneumophila serogroup 1. This classification is consistent with the high prevalence of L. pneumophila SG1, which accounts for more than 80 % of cases of Legionnaires' disease (Marston, 1995). Within the L. pneumophila strains, 50 % of them induced apoptosis/DNA fragmentation and 63 % showed pore-forming activity. One pathogenic trait that is common in the majority of L. pneumophila strains was the high level of intracellular growth. It should also be noted that two L. pneumophila SG1 strains (SG1-64 and SG1-65) and GR159 are placed in group II in spite of a high intracellular growth index after 72 h of macrophage infection due to their being non-cytopathogenic to the macrophage cell line after 72 h of infection, suggesting that high cytopathogenicity is not a general feature of L. pneumophila SG1. It is interesting that L. pneumophila SG1 Knoxville strain lacked all of the pathogenic traits associated with this particular serogroup. Taken together, our results show that there is heterogeneity in the virulence/pathogenic traits in L. pneumophila. Heterogeneity in virulence traits was also observed for L. micdadei.

L. pneumophila GR159, a rib mutant defective in pore-forming activity, has already been characterized extensively by our laboratory (Alli et al., 2000; Molmeret et al., 2002a, b). Pore-forming activity plays an important role in the cytopathogenicity of L. pneumophila since all the strains that possessed this phenotypic trait belong to group I. However, we found that this is not the only factor responsible for this high cytopathogenicity since L. dumoffii and L. micdadei Rivera strains did not possess this phenotypic trait but were still highly cytopathogenic. Pore-forming activity is a phenotypic trait that could aid in the identification and diagnosis of L. pneumophila in the laboratory, since all the strains of L. pneumophila possessed this particular phenotype with the exception of SG1-64 and SG1 Knoxville. Genetic characterization of these strains need to be carried out to ascertain the basis for lack of pore-forming activity. Mutation within the dot/icm loci cannot be ruled out, as we have shown in our previous studies that a base-pair deletion within icmT is sufficient to change the pore-forming activity phenotype from positive to negative (Molmeret et al., 2002a, b).

Classification of isolates of Legionella spp. has been carried out previously using mouse and guinea pig macrophages. Izu et al. (1999) grouped 20 reference strains into four groups. In contrast, our classification based on cytopathogenicity divided 27 strains of Legionella spp. into two broad main groups. Group I classification by Izu et al. (1999) includes most of the group II strains in our study along with L. dumoffii. There is no direct correlation with the grouping in this study, which could be due to the differences in the strains and/or methods of classification. This is not surprising as Izu et al. (1999) found no significant characteristics common to another grouping done by Neumeister et al. (1997), which was based on the bacterial doubling time in Mono Mac 6 cells and Acanthamoeba castellanii. Our data on L. cincinnatiensis are consistent with the finding of Izu et al. (1999) that this species of Legionella does not grow within U937 cells despite the fact there might be some strain differences and different host cell lines used. However, the main objective of this study was not to provide a classification method for Legionella spp., but rather to provide a better understanding of the pathogenic trait(s) that could be responsible for variation in the virulence of these bacteria.

Interestingly, L. dumoffii and L. micdadei Rivera, which did not exhibit pore-forming activity, and did not induce apoptosis/DNA fragmentation in macrophages, belonged to high cytopathogenic group I. The lack of two pathogenic traits in these strains despite their ability to cause cell death indicates that other factors in addition to apoptosis and pore-forming activity are involved in cytopathogenicity of these species. Cytopathogenicity could be the result of unrestricted growth of L. micdadei (Rivera) and L. dumoffii. However, high levels of intracellular growth have also been demonstrated among the group II strains, suggesting that high growth rate within macrophages is not sufficient to cause cell death for these strains. It is possible that variation in the combination of the three phenotypic traits (induction of apoptosis in host cells, pore-forming activity, and high intracellular growth index) could be what determines the variability in the incidence of Legionella spp. in causing Legionnaires' disease. L. pneumophila and L. micdadei are the two most common species of legionellae that cause Legionnaires' disease, with L. pneumophila and L. micdadei responsible for 85 % and 5 % respectively, of cases (Dowling et al., 1983; Halberstam et al., 1992; Hebert et al., 1980; Pasculle et al., 1980). Importantly, these two species belonged to group I in our classification scheme, suggesting that the grouping can discriminate between the Legionella spp. that are responsible for high and low incidence of infection. Cytopathogenicity is the only pathogenic trait that gives direct correlation with high incidence of a particular species to cause disease in a human population. The L. longbeachae strains used in this study failed to induce apoptosis/DNA fragmentation, which is in sharp contrast to a previous study that showed this particular species of legionellae did induce apoptosis (Arakaki et al., 2002). The differences in results could be due to strain differences, as we have already shown heterogeneity in virulence trait(s) in L. micdadei and L. pneumophila.

Southern analyses revealed the presence of the dot/icm loci in all the Legionella spp., indicating the highly conservative nature of these loci in all the species of legionellae. Previous studies have shown the presence of dotA and icmX genes in L. micdadei Pittsburg, L. micdadei Tatlock, L. bozemanii, L. gratiana (Matthews & Roy, 2000), and of dotA, dotB, dotE and dotFG in L. micdadei 31B, Rivera, Camileri and D-2676 (Joshi & Swanson, 1999). The IcmX protein has been demonstrated in L. micdadei and L. gratiana (Matthews & Roy, 2000). In our studies, all strains of legionellae investigated harboured all the dot/icm loci. In spite of the presence of the dot/icm loci, two different groups of Legionella spp. based on the degree of cytopathogenicity can be demonstrated. It remains to be seen whether all the proteins of the dot/icm genes are expressed and functional in all the Legionella spp. The roles of the dot/icm loci in pathogenesis have been well established for L. pneumophila (Segal et al., 1998; Vogel et al., 1998), but the contributions of these loci to the pathogenesis of other species are yet to be determined. Our data suggest that the dot/icm regions of L. pneumophila may represent a sensitive and powerful molecular tool for definitive diagnosis and identification of legionellae in a clinical microbiology laboratory since all the strains of legionellae tested in this study harbour these loci. However, a large-scale screen is required to confirm the specificity of hybridization to all other strains of legionellae and not to other genera of bacteria.

In conclusion, most of the group II strains of legionellae, with low cytopathogenicity, are unable to replicate within macrophages and lack pore-forming activity, and the majority of them are unable to induce apoptosis; this could be a general feature of non-pathogenic strains of legionellae. In contrast, strains belonging to group I are able to replicate within macrophages, with the majority of them exhibiting pore-forming activity, and some of them can induce apoptosis/DNA fragmentation. No single virulence traits were found to correlate with the cytopathogenicity of this genus.


   ACKNOWLEDGEMENTS
 
We thank H. Shuman and R. Isberg for their kind gifts of the dot/icm clones. We thank R. Benson, B. Fields and J. Pruckler for the kind gift of the Legionella spp. used in this study. Y. A. is supported by Public Health Service Award RO1AI43965.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Abu Kwaik, Y. (1998). Fatal attraction of mammalian cells to Legionella pneumophila. Mol Microbiol 30, 689–696.[CrossRef][Medline]

Abu Kwaik, Y. & Pederson, L. L. (1996). The use of differential display-PCR to isolate and characterize a Legionella pneumophila locus induced during the intracellular infection of macrophages. Mol Microbiol 21, 543–556.[Medline]

Abu Kwaik, Y., Gao, L.-Y., Harb, O. S. & Stone, B. J. (1997). Transcriptional regulation of the macrophage-induced gene (gspA) of Legionella pneumophila and phenotypic characterization of a null mutant. Mol Microbiol 24, 629–642.[Medline]

Abu Kwaik, Y., Gao, L.-Y., Stone, B. J. & Harb, O. S. (1998). Invasion of mammalian and protozoan cells by Legionella pneumophila.: Bull Inst Pasteur 96, 237–247.[CrossRef]

Adeleke, A., Pruckler, J., Benson, R., Rowbotham, T., Halablab, M. & Fields, B. S. (1996). Legionella-like amoebal pathogens – phylogenetic status and possible role in respiratory disease. Emerg Infect Dis 2, 225–229.[Medline]

Alli, O. A. T., Gao, L.-Y., Pedersen, L. L., Zink, S., Radulic, M., Doric, M. & Abu Kwaik, Y. (2000). Temporal pore formation-mediated egress from macrophages and alveolar epithelial cells by Legionella pneumophila. Infect Immun 68, 6431–6440.[Abstract/Free Full Text]

Arakaki, N., Higa, F., Koide, M., Tateyama, M. & Saito, A. (2002). Induction of apoptosis of human macrophages in vitro by Legionella longbeachae through activation of the caspase pathway. J Med Microbiol 51, 159–168.[Abstract/Free Full Text]

Benin, A. L., Benson, R. F. & Besser, R. E. (2002). Trends in Legionnaires' disease, 1980–1998: declining mortality and new patterns of diagnosis. Clin Infect Dis 35, 1039–1046.[CrossRef][Medline]

Benson, R. F. & Fields, B. S. (1998). Classification of the genus Legionella. Semin Respir Infect 13, 90–99.[Medline]

Byrne, B. & Swanson, M. S. (1998). Expression of Legionella pneumophila virulence traits in response to growth conditions. Infect Immun 66, 3029–3034.[Abstract/Free Full Text]

Den Boer, J. W., Yzerman, E. P., Schellekens, J. & 11 other authors (2002). A large outbreak of Legionnaires' disease at a flower show, the Netherlands, 1999. Emerg Infect Dis 8, 37–43.[Medline]

Dowling, J. N., Kroboth, F. J., Karpf, M., Yee, R. B. & Pasculle, A. W. (1983). Pneumonia and multiple lung abscesses caused by dual infection with Legionella micdadei and Legionella pneumophila. Am Rev Respir Dis 127, 121–125.[Medline]

Doyle, R. M., Steele, T. W., McLennan, A. M., Parkinson, I. H., Manning, P. A. & Heuzenroeder, M. W. (1998). Sequence analysis of the mip gene of the soilborne pathogen Legionella longbeachae. Infect Immun 66, 1492–1499.[Abstract/Free Full Text]

Doyle, R. M., Cianciotto, N. P., Banvi, S., Manning, P. A. & Heuzenroeder, M. W. (2001). Comparison of virulence of Legionella longbeachae strains in guinea pigs and U937 macrophage-like cells. Infect Immun 69, 5335–5344.[Abstract/Free Full Text]

Gao, L.-Y. & Abu Kwaik, Y. (1999a). Activation of caspase-3 in Legionella pneumophila-induced apoptosis in macrophages. Infect Immun 67, 4886–4894.[Abstract/Free Full Text]

Gao, L.-Y. & Abu Kwaik, Y. (1999b). Apoptosis in macrophages and alveolar epithelial cells during early stages of infection by Legionella pneumophila and its role in cytopathogenicity. Infect Immun 67, 862–870.[Abstract/Free Full Text]

Gao, L.-Y. & Abu Kwaik, Y. (2000a). Hijacking the apoptotic pathways of the host cell by bacterial pathogens. Microb Infect 2, 1705–1719.[CrossRef][Medline]

Gao, L.-Y. & Abu Kwaik, Y. (2000b). The mechanism of killing and exiting the protozoan host Acanthamoeba polyphaga by Legionella pneumophila. Environ Microbiol 2, 79–90.[CrossRef][Medline]

Gao, L.-Y., Stone, B. J., Brieland, J. K. & Abu Kwaik, Y. (1998). Different fates of Legionella pneumophila pmi and mil mutants within human-derived macrophages and alveolar epithelial cells. Microb Pathog 25, 291–306.[CrossRef][Medline]

Halberstam, M., Isenberg, H. D. & Hilton, E. (1992). Abscess and empyema caused by Legionella micdadei. J Clin Microbiol 30, 512–513.[Abstract]

Harb, O. S., Gao, L.-Y. & Abu Kwaik, Y. (2000). From protozoa to mammalian cells: a new paradigm in the life cycle of intracellular bacterial pathogens. Environ Microbiol 2, 251–265.[CrossRef][Medline]

Hebert, G. A., Thomason, B. M., Harris, P. P., Hicklin, M. D. & McKinney, R. M. (1980). "Pittsburgh pneumonia agent": a bacterium phenotypically similar to Legionella pneumophila and identical to the TATLOCK bacterium. Ann Intern Med 92, 53–54.[Medline]

Hilbi, H., Segal, G. & Shuman, H. A. (2001). Icm/Dot-dependent upregulation of phagocytosis by Legionella pneumophila. Mol Microbiol 42, 603–617.[CrossRef][Medline]

Horwitz, M. A. (1983). Formation of a novel phagosome by the Legionnaires' disease bacterium (Legionella pneumophila) in human monocytes. J Exp Med 158, 1319–1331.[Abstract]

Izu, K., Yoshida, S., Miyamoto, H., Chang, B., Ogawa, M., Yamamoto, H., Goto, Y. & Taniguchi, H. (1999). Grouping of 20 reference strains of Legionella species by the growth ability within mouse and guinea pig macrophages. FEMS Immunol Med Microbiol 26, 61–68.[CrossRef][Medline]

Jaulhac, B., Nowicki, M., Bornstein, N., Meunier, O., Prevost, G., Piemont, Y., Fleurette, J. & Moteil, H. (1992). Detection of Legionella spp. in bronchoalveolar lavage fluids by DNA amplification. J Clin Microbiol 30, 920–924.[Abstract]

Joseph, C. A., Hutchinson, E. J., Dedman, D., Birtles, R. J., Watson, J. M. & Bartlett, C. L. (1995). Legionnaires' disease surveillance: England and Wales 1994. Commun Dis Rep CDR Rev 5, R180–R183.[Medline]

Joseph, C. A., Harrison, T. G., Ilijic-Car, D. & Bartlett, C. L. (1997). Legionnaires' disease in residents of England and Wales: 1996. Commun Dis Rep CDR Rev 7, R153–R159.[Medline]

Joshi, A. D. & Swanson, M. S. (1999). Comparative analysis of Legionella pneumophila and Legionella micdadei virulence traits. Infect Immun 67, 4134–4142.[Abstract/Free Full Text]

Kirby, J. E., Vogel, J. P., Andrews, H. L. & Isberg, R. R. (1998). Evidence for pore-forming ability by Legionella pneumophila. Mol Microbiol 27, 323–336.[CrossRef][Medline]

Koide, M. & Saito, A. (1995). Diagnosis of Legionella pneumophila infection by polymerase chain reaction. Clin Infect Dis 21, 199–201.[Medline]

Levi, M. H., Pasculle, A. W. & Dowling, J. N. (1987). Role of the alveolar macrophage in host defense and immunity to Legionella micdadei pneumonia in guinea pig. Microb Pathog 2, 269–282.[Medline]

Lo Presti, F., Riffard, S., Meugnier, H. & 8 other authors (1999). Legionella taurinensis sp. nov., a new species antigenically similar to Legionella spiritensis. Int J Syst Bacteriol 49, 397–403.[Abstract]

Lo Presti, F., Riffard, S., Meugnier, H. & 9 other authors (2001). Legionella gresilensis sp. nov. and Legionella beliardensis sp. nov., isolated from water in France. Int J Syst Evol Microbiol 51, 1949–1957.[Abstract/Free Full Text]

Marston, B. J. (1995). Epidemiology of community-acquired pneumonia. Infect Dis Clin Pract 4, S232–S239.

Marston, B. J., Lipman, H. B. & Breiman, R. F. (1994). Surveillance for Legionnaires' disease: risk factors for mortality and morbidity. Arch Intern Med 154, 2417–2422.[Abstract]

Matthews, M. & Roy, C. R. (2000). Identification and subcellular localization of the Legionella pneumophila IcmX protein: a factor essential for establishment of a replicative organelle in eukaryotic host cells. Infect Immun 68, 3971–3982.[Abstract/Free Full Text]

Moffat, J. F. & Tompkins, L. S. (1992). A quantitative model of intracellular growth of Legionella pneumophila in Acanthamoeba castellanii. Infect Immun 60, 296–301.[Abstract]

Molmeret, M. & Abu Kwaik, Y. (2002). How does Legionella pneumophila exit the host cell? Trends Microbiol 10, 258–60.[CrossRef][Medline]

Molmeret, M., Alli, O. A. T., Marina, R., Milorad, S., Mijenko, D. & Abu Kwaik, Y. (2002a). The C-terminus of icmT is essential for the pore formation and for intracellular trafficking of L. pneumophila within A. polyphaga. Mol Microbiol 43, 1139–1150.[CrossRef][Medline]

Molmeret, M., Alli, O. A., Zink, S., Flieger, A., Cianciotto, N. P. & Kwaik, Y. A. (2002b). icmT is essential for pore formation-mediated egress of Legionella pneumophila from mammalian and protozoan cells. Infect Immun 70, 69–78.[Abstract/Free Full Text]

Neumeister, B., Schoniger, S., Faigle, M., Eichner, M. & Dietz, K. (1997). Multiplication of different Legionella species in Mono Mac 6 cells and in Acanthamoeba castellanii. Appl Environ Microbiol 63, 1219–1224.[Abstract]

O'Connell, W. A., Bangsborg, J. M. & Cianciotto, N. P. (1995). Characterization of a Legionella micdadei mip mutant. Infect Immun 63, 2840–2845.[Abstract]

O'Connell, W. A., Dhand, L. & Cianciotto, N. P. (1996). Infection of macrophage-like cells by Legionella species that have not been associated with disease. Infect Immun 64, 4381–4384.[Abstract]

Pasculle, A. W., Feeley, J. C., Gibson, R. J. & 7 other authors (1980). Pittsburgh pneumonia agent: direct isolation from human lung tissue. J Infect Dis 141, 727–732.[Medline]

Segal, G., Purcell, M. & Shuman, H. A. (1998). Host cell killing and bacterial conjugation require overlapping sets of genes within a 22-kb region of the Legionella pneumophila chromosome. Proc Natl Acad Sci U S A 95, 1669–1674.[Abstract/Free Full Text]

Thacker, W. L., Benson, R. F., Schifman, R. B., Pugh, E., Steigerwalt, A. G., Mayberry, W. R., Brenner, D. J. & Wilkinson, H. W. (1989). Legionella tucsonensis sp. nov. isolated from a renal transplant recipient. J Clin Microbiol 27, 1831–1834.[Medline]

Vogel, J. P., Andrews, H. L., Wong, S. K. & Isberg, R. R. (1998). Conjugative transfer by the virulence system of Legionella pneumophila. Science 279, 873–876.[Abstract/Free Full Text]

Watarai, M., Derre, I., Kirby, J., Growney, J. D., Dietrich, W. F. & Isberg, R. R. (2001). Legionella pneumophila is internalized by a macropinocytotic uptake pathway controlled by the Dot/Icm system and the mouse Lgn1 locus. J Exp Med 194, 1081–1096.[Abstract/Free Full Text]

Wilkinson, H. W., Thacker, W. L., Benson, R. F., Polt, S. S., Brookings, E., Mayberry, W. R., Brenner, D. J., Gilley, R. G. & Kirklin, J. K. (1987). Legionella birminghamensis sp. nov. isolated from a cardiac transplant recipient. J Clin Microbiol 25, 2120–2122.[Medline]

Zink, S. D., Pedersen, L., Cianciotto, N. P. & Abu Kwaik, Y. (2002). The Dot/Icm type iv secretion system of Legionella pneumophila is essential for the induction of apoptosis in human macrophages. Infect Immun 70, 1657–1663.[Abstract/Free Full Text]

Received 6 September 2002; revised 6 December 2002; accepted 19 December 2002.