Potential Virulence Role of the Legionella pneumophila ptsP Ortholog

We previously identified the Legionella pneumophila ptsP (phosphoenolpyruvate phosphotransferase) ortholog gene as a putative virulence factor in a study of signature-tagged mutagenesis using a guinea pig pneumonia model. In this study, we further defined the phenotypic properties of L. pneumophila ptsP and its complete sequence. The L. pneumophila ptsP was 2,295 bases in length. Its deduced amino acid sequence had high similarity with ptsP orthologs of Pseudomonas aeruginosa, Azotobacter vinelandii, and Escherichia coli, with nearly identical lengths. Here we show that while the mutant grew well in laboratory media, it was defective in both lung and spleen multiplication in guinea pigs. It grew slowly in guinea pig alveolar macrophages despite good uptake into the cells. Furthermore, there was minimal growth in a human alveolar epithelial cell line (A549). Transcomplementation of the L. pneumophila ptsP mutant almost completely rescued its growth in alveolar macrophages, in A549 cells, and in guinea pig lung and spleen. The L. pneumophila ptsP mutant was capable of evasion of phagosome-lysosome fusion and resided in ribosome-studded phagosomes. Pore formation activity of the mutant was normal. The L. pneumophila ptsP mutant expressed DotA and IcmX in apparently normal amounts, suggesting that the ptsP mutation did not affect dotA and icmX regulation. In addition, the mutant was resistant to serum and neutrophil killing. Taken together, these findings show that L. pneumophila ptsP is required for full in vivo virulence of L. pneumophila, most probably by affecting intracellular growth.     

Intracellular growth of Legionella pneumophila in Dictyostelium discoideum, a system for genetic analysis of host-pathogen interactions

Conditions were established in which Legionella pneumophila, an intracellular bacterial pathogen, could replicate within the unicellular organism Dictyostelium discoideum. By several criteria, L. pneumophila grew by the same mechanism within D. discoideum as it does in amoebae and macrophages. Bacteria grew within membrane-bound vesicles associated with rough endoplasmic reticulum, and L. pneumophila dot/icm mutants, blocked for growth in macrophages and amoebae, also did not grow in D. discoideum. Internalized L. pneumophila avoided degradation by D. discoideum and showed evidence of reduced fusion with endocytic compartments. The ability of L. pneumophila to grow within D. discoideum depended on the growth state of the cells. D. discoideum grown as adherent monolayers was susceptible to L. pneumophila infection and to contact-dependent cytotoxicity during high-multiplicity infections, whereas D. discoideum grown in suspension was relatively resistant to cytotoxicity and did not support intracellular growth. Some known D. discoideum mutants were examined for their effect on growth of L. pneumophila. The coronin mutant and the myoA/B double myosin I mutant were more permissive than wild-type strains for intracellular growth. Growth of L. pneumophila in a G(beta) mutant was slightly reduced compared to the parent strain. This work demonstrates the usefulness of the L. pneumophila-D. discoideum system for genetic analysis of host-pathogen interactions.     

lvgA,a novel Legionella pneumophila virulence factor

Several novel Legionella pneumophila virulence genes were previously discovered by use of signature-tagged mutagenesis (P. H. Edelstein, M. A. Edelstein, F. Higa, and S. Falkow, Proc. Natl. Acad. Sci. 96:8190-8195, 1999). One of these mutants appeared to be defective in multiplication in guinea pig lungs and spleens, yet it multiplies normally in guinea pig alveolar macrophages. Here we report further characterization of the mutated gene and its protein and the virulence role of the gene. The complete sequence of the gene, now called lvgA, is 627 bp long, and its protein product is approximately 27 kDa in size. lvgA was present in all 50 strains of L. pneumophila tested. No significant nucleic acid or protein homology was found in the GenBank database for the gene, nor were any distinctive motifs discovered in a search of other databases. The expression of both DotA and IcmX in the lvgA mutant was normal. Subcellular fractionation studies localized LvgA to the outer membrane fraction, and protease digestion studies suggested that at least some of the protein is surface expressed. No change in bacterial lipopolysaccharide composition or reactivity to serogroup-specific antisera was detected in the mutant. Growth competition studies with alveolar macrophages showed that the mutant was outcompeted by its parent 3-fold in 24 h and 24-fold in 48 h, in contrast to what was observed with the null phenotype in parallel testing with alveolar macrophages or with the A549 alveolar epithelial cell line. This macrophage defect of the mutant bacterium was due to slower growth, as the mutant invaded alveolar macrophages normally. Electron microscopy showed that the mutant bacterium resided in a ribosome-studded phagosome in alveolar macrophages, with no distinction from its parent. The lvgA mutant was outcompeted by its parent about sixfold in guinea pig lungs and spleens; prolonged observation of infected animals showed no late-onset virulence of the mutant. Transcomplementation of the mutant restored the parental phenotype in guinea pigs. The lvgA mutant was twofold more susceptible to killing by human beta-defensin 2 but not to killing by other cationic peptides, serum complement, or polymorphonuclear neutrophils. lvgA is a novel virulence gene that is responsible for pleiotropic functions involving both extracellular and intracellular bacterial resistance mechanisms.     

Characterization of the Legionella pneumophila gene ligA

Legionella pneumophila is a bacterial pathogen that resides and multiplies in macrophages as well as in its natural aquatic hosts, the protozoa. Different bacterial factors contribute to pathogenicity and accompanying eukaryotic intracellular events. Sequencing of mip flanking regions revealed a gene of 2610 bp, ligA, that has no significant similarity to any of the genes identified previously. Epidemiological studies indicate that this gene is present in Legionella pneumophila, the species most often associated with cases of the Legionnaires' disease, but not in Legionella species other than L. pneumophila. The isogenic ligA deletion mutant was resistant to NaCl, and showed decreased cytotoxicity to human monocytes and decreased hemolytic activity to red blood cells. However, the most prominent effect of the L. pneumophila ligA mutant strain LEPF1 was the nearly completely reduced replication within the natural host Acanthamoeba castellanii. Since this gene is L. pneumophila specific and regulates numerous bacterial properties we designated this gene ligA for Legionella pneumophila infectivity gene A.     

Characterization of the gene encoding the major secreted lysophospholipase A of Legionella pneumophila and its role in detoxification of lysophosphatidylcholine

We previously showed that Legionella pneumophila secretes, via its type II secretion system, phospholipase A activities that are distinguished by their specificity for certain phospholipids. In this study, we identified and characterized plaA, a gene encoding a phospholipase A that cleaves fatty acids from lysophospholipids. The plaA gene encoded a 309-amino-acid protein (PlaA) which had homology to a group of lipolytic enzymes containing the catalytic signature GDSL. In Escherichia coli, the cloned gene conferred trypsin-resistant hydrolysis of lysophosphatidylcholine and lysophosphatidylglycerol. An L. pneumophila plaA mutant was generated by allelic exchange. Although the mutant grew normally in standard buffered yeast extract broth, its culture supernatants lost greater than 80% of their ability to release fatty acids from lysophosphatidylcholine and lysophosphatidylglycerol, implying that PlaA is the major secreted lysophospholipase A of L. pneumophila. The mutant's reduced lipolytic activity was confirmed by growth on egg yolk agar and thin layer chromatography and was complemented by reintroduction of an intact copy of plaA. Overexpression of plaA completely protected L. pneumophila from the toxic effects of lysophosphatidylcholine, suggesting a role for PlaA in bacterial detoxification of lysophospholipids. The plaA mutant grew like the wild type in U937 cell macrophages and Hartmannella vermiformis amoebae, indicating that PlaA is not essential for intracellular infection of L. pneumophila. In the course of characterizing plaA, we discovered that wild-type legionellae secrete a phospholipid cholesterol acyltransferase activity, highlighting the spectrum of lipolytic enzymes produced by L. pneumophila.     

Cross-reactions between Legionella pneumophila (serogroup 1) and twenty-eight other bacterial species, including other members of the family Legionellaceae.

Cross-reactions between Legionella pneumophila serogroup 1 and 28 other bacterial species were studied by various quantitative immunoelectrophoretic techniques. A sonicated L. pneumophila antigen and purified homologous rabbit antibody were used as a reference system. Few antigens (0 to 6) cross-reacted with non-Legionellaceae, but two were found in nearly all gram-negative bacteria tested (antigens no. 1 and 66). Antigen no. 66 of the L. pneumophila reference system was shown to be antigenically similar to the "common antigen" of Pseudomonas aeruginosa reported in many gram-negative bacteria. Greater than 85% of the antigens from L. pneumophila serogroup 1 cross-reacted with the other six serogroups of L. pneumophila. By contrast, Fluoribacter (Legionella) bozemanae, F. (L.) dumoffii, F. (L.) gormanii, and Tatlockia (Legionella) micdadei cross-reacted with only 45, 53, 39, and 43% of the reference system antigens, respectively. The antigenic relatedness of members of the Legionellaceae, expressed as a matching coefficient, is discussed in terms of its taxonomic significance. Serogroup-, genus-, and family-specific antigens are identified in the L. pneumophila reference system.     

Growth of 28 Legionella species on selective culture media: a comparative study.

We compared the growth of 28 Legionella spp. on four manufacturers' buffered charcoal-yeast extract (BCYE) agar media and selective BCYE media that contained polymyxin B, anisomycin, and vancomycin or cefamandole. With BCYE as a "gold standard," growth for Legionella pneumophila was significantly better than for the nonpneumophila species on all media tested. L. pneumophila and 24 other Legionella spp. grew on vancomycin-containing media, while L. santicrucis, L. rubrilucens, and L. erythra grew poorly. In contrast, 11 of 28 species (notably L. micdadei and L. bozemanii) did not grow on cefamandole-containing media and 8 of 28 species only grew marginally. We demonstrated that selective BCYE media that contain vancomycin or cefamandole may not support the growth of all Legionella spp. One commercial manufacturer's media were consistently suboptimal. Laboratories should not rely on a manufacturer's quality control testing in lieu of their own.     

Lcl of Legionella pneumophila is an immunogenic GAG binding adhesin that promotes interactions with lung epithelial cells and plays a crucial role in biofilm formation

Legionellosis is mostly caused by Legionella pneumophila and is defined by a severe respiratory illness with a case fatality rate ranging from 5 to 80%. In vitro and in vivo, interactions of L. pneumophila with lung epithelial cells are mediated by the sulfated glycosaminoglycans (GAGs) of the host extracellular matrix. In this study, we have identified several Legionella heparin binding proteins. We have shown that one of these proteins, designated Lcl, is a polymorphic adhesin of L. pneumophila that is produced during legionellosis. Homologues of Lcl are ubiquitous in L. pneumophila serogroups but are undetected in other Legionella species. Recombinant Lcl binds to GAGs, and a Δlpg2644 mutant demonstrated reduced binding to GAGs and human lung epithelial cells. Importantly, we showed that the Δlpg2644 strain is dramatically impaired in biofilm formation. These data delineate the role of Lcl in the GAG binding properties of L. pneumophila and provide molecular evidence regarding its role in L. pneumophila adherence and biofilm formation.     

Protein kinase LegK2 is a type IV secretion system effector involved in endoplasmic reticulum recruitment and intracellular replication of Legionella pneumophila

Legionella pneumophila is the etiological agent of Legionnaires' disease. Crucial to the pathogenesis of this intracellular pathogen is its ability to subvert host cell defenses, permitting intracellular replication in specialized vacuoles within host cells. The Dot/Icm type IV secretion system (T4SS), which translocates a large number of bacterial effectors into host cell, is absolutely required for rerouting the Legionella phagosome. Many Legionella effectors display distinctive eukaryotic domains, among which are protein kinase domains. In silico analysis and in vitro phosphorylation assays identified five functional protein kinases, LegK1 to LegK5, encoded by the epidemic L. pneumophila Lens strain. Except for LegK5, the Legionella protein kinases are all T4SS effectors. LegK2 plays a key role in bacterial virulence, as demonstrated by gene inactivation. The legK2 mutant containing vacuoles displays less-efficient recruitment of endoplasmic reticulum markers, which results in delayed intracellular replication. Considering that a kinase-dead substitution mutant of legK2 exhibits the same virulence defects, we highlight here a new molecular mechanism, namely, protein phosphorylation, developed by L. pneumophila to establish a replicative niche and evade host cell defenses.     

A Legionella pneumophila gene encoding a species-specific surface protein potentiates initiation of intracellular infection.

To investigate the pathogenesis of Legionnaires disease at a molecular level, we mutated by directed allelic exchange a gene encoding a Legionella pneumophila-specific 24,000-dalton (Da) surface protein. Southern hybridization and immunoblot analyses demonstrated that the predicted DNA rearrangement occurred in L. pneumophila with a specific loss of 24-kDa antigen expression. Compared with its isogenic parent, the mutant was significantly impaired in its ability to infect transformed U937 cells, a human macrophagelike cell line; i.e., the bacterial inoculum of the mutant strain that was required to initiate infection of the macrophage monolayer was ca. 80-fold greater than that of the isogenic parent strain. The mutant strain regained full infectivity on reintroduction of a cloned 24-kDa protein gene, indicating that the reduced infectivity was due specifically to the mutation in that gene. Compared with the parent strain, the mutant strain was recovered at titers that were ca. 10-fold lower shortly after infection, but it exhibited a similar intracellular growth rate over the next 40 h, indicating that the mutant was defective in its ability to initiate macrophage infection rather than in its ability to replicate intracellularly. When opsonized, the mutant strain was still significantly less infectious than the parent strain, despite equivalent macrophage association, suggesting that the mutant was not merely missing a ligand for macrophage attachment. The mutant also exhibited reduced infectivity in explanted human alveolar macrophages, demonstrating the relevance of the U937 cell model for analyzing this mutant phenotype. These results represent the first identification of a cloned L. pneumophila gene that is necessary for optimal intracellular infection; we designate this gene mip, for macrophage infectivity potentiator.     

Legionella pneumophila feoAB promotes ferrous iron uptake and intracellular infection

In order to determine the role of ferrous iron transport in Legionella pathogenesis, we identified and mutated the feoB gene in virulent Legionella pneumophila strain 130b. As it is in Escherichia coli, the L. pneumophila feoB gene was contained within a putative feoAB operon. L. pneumophila feoB insertion mutants exhibited decreased ferrous but not ferric iron uptake compared to the wild type. Growth on standard buffered charcoal yeast extract agar or buffered yeast extract broth was unaffected by the loss of L. pneumophila FeoB. However, the L. pneumophila feoB mutant had a reduced ability to grow on buffered charcoal yeast extract agar with a reduced amount of its usual iron supplementation, a phenotype that could be complemented by the addition of feoB in trans. In unsupplemented buffered yeast extract broth, the feoB mutant also had a growth defect, which was further exacerbated by the addition of the ferrous iron chelator, 2,2'-dipyridyl. The feoB mutant was also 2.5 logs more resistant to streptonigrin than wild-type 130b, confirming its decreased ability to acquire iron during extracellular growth. Decreased replication of the feoB mutant was noted within iron-depleted Hartmannella vermiformis amoebae and human U937 cell macrophages. The reduced intracellular infectivity of the feoB mutant was complemented by the introduction of a plasmid containing feoAB. The L. pneumophila feoB gene conferred a modest growth advantage for the wild type over the mutant in a competition assay within the lungs of A/J mice. Taken together, these results indicate that L. pneumophila FeoB is a ferrous iron transporter that is important for extracellular and intracellular growth, especially in iron-limited environments. These data represent the first evidence for the importance of ferrous iron transport for intracellular replication by a human pathogen.     

Phagocytosis of Wild-Type Legionella pneumophila Occurs through a Wortmannin-Insensitive Pathway

Wild-type Legionella pneumophila grows in human macrophages within a replicative phagosome, avoiding lysosomal fusion, while nonreplicative mutants are killed in lysosomes. Wortmannin, a phosphatidylinositol 3-kinase (PI3K) inhibitor, blocks phagocytosis of an avirulent mutant, but not of wild-type L. pneumophila, without affecting membrane ruffling and actin polymerization. These results show that wild-type and mutant Legionella strains use different entry pathways. They suggest that PI3Ks are involved in phagocytosis of an avirulent L. pneumophila mutant and regulate the ability of microorganisms to generate a replicative phagosome.     

Identification of Legionella species by lipopolysaccharide antigen pattern.

Electrophoretic analysis of lipopolysaccharide (LPS) extracts from 430 previously serotyped Legionella isolates and 28 American Type Culture Collection (ATCC) non-Legionella pneumophila Legionella reference strains representing different Legionella species and serogroups has been performed. LPS was prepared from Legionella suspensions by sonication and proteinase K digestion. Following sodium dodecyl sulfate-polyacrylamide gel electrophoresis, LPS bands were either stained with silver nitrate or transferred onto a nitrocellulose membrane and detected with rabbit antibodies raised against L. pneumophila serogroup 5, which was known to cross-react with L. pneumophila serogroups 1 to 14. Silver staining revealed that each of the 28 ATCC non-L. pneumophila Legionella strains possessed an individual and characteristic LPS banding pattern. The LPS profile was defined by the molecular weight of the visualized bands and/or the individual ladder-like LPS pattern. It was demonstrated by immunoblotting that non-L. pneumophila Legionella strains did not react with the serogroup 5 antiserum, thus allowing for the differentiation between L. pneumophila and non-L. pneumophila species.     

Genetic, immunological, and cytotoxic comparisons of Legionella proteolytic activities.

Several strains of Legionella pneumophila and other species of Legionella with proteolytic activities were compared by assays, including Southern hybridizations and Western immunoblots, to determine their proteolytic, hemolytic, and cytotoxic activities. Only proteases from strains of L. pneumophila were both hemolytic and cytotoxic, and proteolytic activities extracted from other species of Legionella possessed only hemolytic activity. A 4.0-kilobase DNA sequence encoding the 38-kilodalton metalloprotease from L. pneumophila Philadelphia 1 that we showed previously was responsible for the observed hemolytic and cytotoxic phenotypes (F. D. Quinn and L. S. Tompkins, Mol. Microbiol., 3:797-805, 1989) was used in Southern hybridizations to probe chromosomal DNA from several strains of L. pneumophila and other Legionella species. The probe hybridized to the chromosomal DNA of all serogroups of L. pneumophila but not to any strains of L. dumoffii, L. micdadei, L. feeleii, or L. jordanis that we examined. Additionally, Western immunoblots done with rabbit antisera made to the cloned L. pneumophila protease demonstrated cross-reactions among 38-kilodalton proteins from strains of L. pneumophila, but no reactions were observed with proteins from other species of Legionella. Similarly, the cloned protease from L. pneumophila reacted with convalescent-phase sera from patients infected with L. pneumophila, but not with antisera isolated from patients infected with other Legionella species. Thus, despite some similarities among the proteolytic activities of members of the genus Legionella, including proteolytic and hemolytic phenotypes, metal requirements for zinc or iron, sensitivity to EDTA, and temperature and pH optima, we documented distinct genetic, immunological, and cytotoxicity differences among the proteolytic activities produced by Legionella species.     

Evaluation of the Gen-Probe DNA probe for the detection of legionellae in culture.

A commercial DNA probe kit designed to detect rRNA from legionellae was evaluated for its ability to correctly discriminate between legionellae and non-legionellae taken from culture plates. The probe kit, made by the Gen-Probe Corp. (San Diego, Calif.), was radiolabeled with 125I, and probe bacterial RNA hybridization, detected in a simple one-tube system hybridization assay, was quantitated with a gamma counter. A total of 156 Legionella sp. strains were tested, of which 125 were Legionella pneumophila and the remainder were strains from 21 other Legionella spp. A total of 106 gram-negative non-legionellae, isolated from human respiratory tract (81%) and other body site (19%) specimens, were also tested; 14 genera and 28 species were represented. The probe easily distinguished all of the legionellae from the non-legionellae. The average legionellae/non-legionellae hybridization ratio was 42:1, and the lowest ratio was 2:1; a minor modification in the procedure increased the lowest ratio to 5:1. In addition to correctly identifying all Legionella species, the probe was able to separate some of the various species of Legionella. L. pneumophila strains hybridized more completely to the probe than did the other Legionella spp.; L. wadsworthii and L. oakridgensis hybridized only about 25% of the probe relative to L. pneumophila. Some strains of phenotypically identified L. pneumophila had much lower hybridization to the probe than other members of the species and may represent a new Legionella species. The simplicity of the technique and specificity of the probe make it a good candidate for confirming the identity of legionellae in culture.     

Cloning and characterization of the gene encoding the major cell-associated phospholipase A of Legionella pneumophila, plaB, exhibiting hemolytic activity

Legionella pneumophila, the causative agent of Legionnaires' disease, is an intracellular pathogen of amoebae, macrophages, and epithelial cells. The pathology of Legionella infections involves alveolar cell destruction, and several proteins of L. pneumophila are known to contribute to this ability. By screening a genomic library of L. pneumophila, we found an additional L. pneumophila gene, plaB, which coded for a hemolytic activity and contained a lipase consensus motif in its deduced protein sequence. Moreover, Escherichia coli harboring the L. pneumophila plaB gene showed increased activity in releasing fatty acids predominantly from diacylphospho- and lysophospholipids, demonstrating that it encodes a phospholipase A. It has been reported that culture supernatants and cell lysates of L. pneumophila possess phospholipase A activity; however, only the major secreted lysophospholipase A PlaA has been investigated on the molecular level. We therefore generated isogenic L. pneumophila plaB mutants and tested those for hemolysis, lipolytic activities, and intracellular survival in amoebae and macrophages. Compared to wild-type L. pneumophila, the plaB mutant showed reduced hemolysis of human red blood cells and almost completely lost its cell-associated lipolytic activity. We conclude that L. pneumophila plaB is the gene encoding the major cell-associated phospholipase A, possibly contributing to bacterial cytotoxicity due to its hemolytic activity. On the other hand, in view of the fact that the plaB mutant multiplied like the wild type both in U937 macrophages and in Acanthamoeba castellanii amoebae, plaB is not essential for intracellular survival of the pathogen.     

Legionella anisa, a possible indicator of water contamination by Legionella pneumophila

Legionella anisa is one of the most frequent species of Legionella other than Legionella pneumophila in the environment and may be hospital acquired in rare cases. We found that L. anisa may mask water contamination by L. pneumophila, suggesting that there is a risk of L. pneumophila infection in immunocompromised patients if water is found to be contaminated with Legionella species other than L. pneumophila.     

Intrapulmonary Hartmannella vermiformis: a potential niche for Legionella pneumophila replication in a murine model of legionellosis.

The potential role of inhaled protozoa as a niche for intrapulmonary replication of Legionella pneumophila was investigated in vivo with mutant strains of L. pneumophila which have reduced virulence for the amoeba Hartmannella vermiformis. L. pneumophila AA488 and AA502 were derived from wild-type strain AA100 after transposon mutagenesis. These mutants have reduced virulence for H. vermiformis but are fully virulent for mononuclear phagocytic cells. A/J mice, which are susceptible to replicative L. pneumophila lung infections, were inoculated intratracheally with L. pneumophila AA100, AA488, or AA502 (10[6] bacteria per mouse) or were coinoculated with one of the L. pneumophila strains (10[6] bacteria per mouse) and uninfected H. vermiformis (10[6] amoebae per mouse). The effect of coinoculation with H. vermiformis on intrapulmonary growth of each L. pneumophila strain was subsequently assessed. In agreement with our previous studies, coinoculation with H. vermiformis significantly enhanced intrapulmonary growth of the parent L. pneumophila strain (AA100). In contrast, intrapulmonary growth of L. pneumophila AA488 or AA502 was not significantly enhanced by coinoculation of mice with H. vermiformis. These studies demonstrate that L. pneumophila virulence for amoebae is required for maximal intrapulmonary growth of the bacteria in mice coinoculated with H. vermiformis and support the hypothesis that inhaled amoebae may potentiate intrapulmonary growth of L. pneumophila by providing a niche for bacterial replication.     

Virulence conversion of Legionella pneumophila: a one-way phenomenon.

Previous investigations have shown that Legionella pneumophila converts from virulence to avirulence after passage on supplemented Mueller-Hinton (SMH) agar and may convert back to virulence after passage in guinea pigs. However, there is no additional information concerning the apparent interconversion of virulent and avirulent derivatives of L. pneumophila cultures. We investigated the stability of a parental virulent culture and its avirulent derivatives and the growth and viability of these cultures on charcoal-yeast extract (CYE) and SMH agars. Avirulent derivatives of a highly virulent L. pneumophila culture were obtained by passage of the virulent parent culture on SMH agar. The only time a virulent L. pneumophila culture was recoverable from an avirulent culture was when the avirulent culture was derived from a saline suspension of a virulent culture which had been passaged only five times on SMH agar. When an avirulent culture was derived from a virulent culture passaged 25 times on SMH agar or from an isolated colony which grew on a SMH agar plate, we were unable to recover a virulent culture after successive passage through guinea pigs. These results suggest that the conversion process which occurs between virulent and avirulent forms of L. pneumophila is a one-way phenomenon from virulence to avirulence and that stable avirulent derivatives can be isolated. Furthermore, our findings suggest that SMH agar acts as a selective medium for the growth of avirulent L. pneumophila, and growth on SMH agar may be a phenotypic marker for avirulence. Virulent cells, although unable to grow on SMH agar, may remain viable for several passages on SMH agar and propagate when inoculated into guinea pigs.