Legionella

To test whether the phenotype of the pilD mutant is due to loss of the type II protein secretion system and/or the type IV pilus assembly apparatus, the authors employed a genetic approach. First, the authors mutated two loci involved in type II protein secretion. Second, the authors analyzed the phenotypes of two mutants defective in the biogenesis of Legionella pneumophila type IV pilus. The protease, acid phosphatase, and pnitrophenyl phosphorylcholine (PNPPC) hydrolase activities were similarly reduced in the supernatant of the pilD, lspDE, and lspG mutants compared with the wild type and the pilE L and pilQ mutant strains. In addition, the lipase and phospholipase A activities were also reduced in the supernatants of both Isp mutants relative to the wild type and the pilE L mutant. These results demonstrate that five pilD-dependent activities are secreted by the L. pneumophila type II protein secretion pathway. Indeed, the small growth defect exhibited by some pilQ mutants in U937 cells could hint at a minor role of the type IV pilus assembly apparatus itself for macrophage infection. This defect, added to the small defect due to the loss of type II secretion, could account for the growth impairment of the pilD mutant in macrophages. Alternatively, PilD could control a third unidentified pathway promoting macrophage infection. Thus, continued analysis of L. pneumophila pilD, lsp, and pilus mutants should not only expand our understanding of Legionnaires’ disease but may provide new paradigms for protein secretion systems.

The identification of a pilD gene in Legionella pneumophila offered new insight into the molecular pathogenesis of legionellosis. Three mutants, NU254, NU255, and NU256, were selected from the screening and further studied. Supernatants from these mutants contained minimal acid phosphatase activity, while possessing normal levels of other pilD-dependent exoproteins. To determine if the map gene was necessary for intracellular replication of Legionella, macrophage-like U937 cells and Hartmannella vermiformis amoebae were infected at multiplicity of infection of 0.1 and 1, respectively. The authors have demonstrated that L. pneumophila has two acid phosphatases that can be differentiated by their different sensitivity to tartrate, and while the tartrate-sensitive Map seems not to be essential for intracellular replication, the role of the tartrate-resistant phosphatase has yet to be determined. The gene that was interrupted by the transposon was determined, and a basic local alignment search tool (BLAST) analysis of the sequence showed that the putative protein shared some homology to lipase enzymes. Supernatants from the mutant presented a defect in its ability to hydrolyze p-nitrophenyl (PNP) palmitate (indication of esterase/lipase) and in the liberation of free fatty acid from phosphatidylcholine (indication of phospholipase A).

Legionella pneumophila secretes several enzymes, including phospholipase A (PLA), via a pilD-dependent type II secretion pathway. Phospholipases possess distinct substrate specificities with respect to both the phospho-head groups and the length and saturation of acyl chains esterified to the glycerol backbone. Although more frequently described for phospholipases C, PLAs are suspected to be virulence factors of bacteria. PLAs has been found for bacterial infections, phospholipases may be responsible for the bacterial escape from phagosomes and host cells after intracellular multiplication, the destruction of macrophages and epithelial cells, the generation of signal transducers like lyso-phosphatidylcholine and derivatives of both arachidonic and linoleic acid, the destruction of lung surfactant, and the induction of inflammation. The absence of dipalmitoylphosphatidylethanolamine (DPPE)-hydrolysis in the present study could be due to the high purity of DPPE (99%) or due to the high gel-to-liquid crystalline phase transition temperature of the palmitoyl derivate used in these experiments, whereas surfactant comprises a mixture of several phospholipids with distinct esterified fatty acids. The hydrolysis of both monopalmitoyl lysophosphatidylcholine (MPLPC) and 1-monopalmitoylglycerol (1-MPG) by lysophospholipase A (LPLA) corresponds to the substrate specificities of the enzyme family to which LPLA may belong, comprising Upases, phospholipases A, esterases, and acyltransferases. Since PLA and LPLA of L. pneumophila are potent in destruction of a variety of phospholipids and the composition of membranes of different cell membranes varies considerably, they might affect many cell types.

This chapter characterizes the secretion kinetics of putative virulence factors of Legionella pneumophila in comparison with those of non-L. pneumophila species during the phase of growth, when nutritional factors may become limited. To investigate the secretion pattern of different Legionella species, culture supernatants were evaluated for acid phosphatase, protease, phospholipase A (PLA), and lysophospholipase A (LPLA) activities. The experiments revealed that the examined enzyme activities seem to be secreted throughout the whole Legionella genus, except L. longbeachae and L. micdadei lack the majority of the activities tested. Differences in PLA secretion of different isolates within one Legionella species (L. steigerwaltii) have been described. L. gormanii, L. steigerwaltii, and L. pneumophila exhibited the highest PLA activity, and both L. gormanii and one of the tested L. pneumophila strains exhibited the most prominent LPLA activity. To estimate the amount of lysophosphatidylcholine (LPC) generated from culture supernatants of different Legionella species, the authors incubated culture supernatants with dipalmitoylphosphatidylchohne (DPPC) and analyzed the lipids by thin-layer chromatography. Culture supernatants from the exponential growth phase of both L. pneumophila and L. steigerwaltii generated considerable amounts of LPC. LPC was also produced by L. gormanii but not before bacteria entered the stationary phase.

This chapter discusses the current understanding of the relationship between Legionella pneumophila and the metal iron. It summarizes the work of a number of laboratories that demonstrated the importance of iron for Legionella, and highlights recent advances toward uncovering mechanisms of L. pneumophila iron acquisition. The mutant displayed a 42% reduction in hemin binding, confirming that hbp potentiates hemin acquisition by L. pneumophila. Within U937 cells, NU216 and its allelic equivalent NU216R were approximately 100-fold more sensitive than the wild type to treatment with desferoxamine, confirming that they are defective for intracellular iron acquisition. The EDDA-hypersensitive mutant NU208 was also dramatically impaired for replication in U937 cells, and its infectivity defect was exacerbated by treatment of the macrophages with desferoxamine. A reconstruction of the NU208 mutation confirmed that the iron acquisition and infectivity defects were due to the transposon insertion and not a spontaneous second-site mutation. Sequence analysis demonstrated that the transposon disruption lies within a gene that is highly similar to the cytochrome c maturation gene, ccmC. ccmC is generally recognized for its role in the heme export step of cytochrome biogenesis. Quantitative infection assays demonstrated that NU229 was impaired ca. 80-fold in intracellular growth. Reconstruction of the mutant by allelic exchange proved that the defect was due to the inactivation of frgA and not a spontaneous second-site mutation. Subsequently, trans-complementation of the mutation demonstrated that the infectivity defect was directly due to the loss of FrgA.

Cationic antimicrobial peptides (CAMP) form an important part of the innate immune response in many organisms ranging from humans to amoebae. The fact that CAMP resistance mechanisms exist in Legionella pneumophila is suggested by several lines of evidence. Investigations were undertaken to investigate the role of the pagP-like gene in L. pneumophila in CAMP resistance and intracellular infections. To investigate the role of the pagP-like gene in CAMP resistance, the MICs of PmB, a bacterially derived CAMP, and C18G, a synthetic CAMP based on human platelet activating factor IV, were assessed for the mutants generated. Consequently, bacterial growth in low Mg2+ minimal media increases the CAMP resistance generated by pagP. CAMP resistance to PmB or C18G was increased by growth in low-Mg2+ CDM in both 130b and NU260. Thus, there was no obvious defect in the initial stages of infection, i.e., attachment and invasion. Intracellular growth of both wild type and mutant was also tested in a coculture assay with Hartmannella vermiformis. This amoeba is known to support the growth of L. pneumophila and was isolated from a clinical case of legionellosis.

This chapter focuses on the major infectivity potentiator protein (Mip) protein and the function and expression of the flagellum of Legionella pneumophila. The Mip protein forms a homodimer on the bacterial surface; each monomer consists of a proximal and a peripheral domain. The contact regions between the two monomers seem to be located at the N-terminal part of the protein. Motility is an important factor to find a new host for another cycle of intracellular replication and for colonization of new habitats. Flagellated bacteria have been found in lung alveolar spaces of patients with legionellosis. It was shown that expression of the virulent phenotype and motility is regulated coordinately. A gentamicin infection assay revealed a clear difference in the number of intracellular bacteria at the onset of multiplication. L. pneumophila is found in very different habitats and it is able to replicate intracellularly in many host cells. Researchers recently demonstrated that regulation of fiaA expression is modulated by different environmental factors, such as temperature, growth phase, osmolarity, viscosity, and the nutrient stage.

More than 10 years ago, genetic analysis successfully characterized for the first time a virulence-associated gene in Legionella pneumophila. The Mip protein has been shown to contribute to the infection of both protozoa and human macrophages. To get morphological evidence for the virulence association of the Mip protein, the immunogold technique was applied to localize this protein on legionellae before and after invasion of Acanthamoeba castellanii, one of the protozoan species that serve as a natural host in the environment. Mip protein of in vitro cultured legionellae is regularly distributed in the bacterial cell wall/cytomembrane complex. Four hours after uptake of agar-grown legionellae by A. castellanii cells, Mip protein was localized inside phagosomes of the host cells. In the last years, many studies used transmission electron microscopy for demonstrating the sequence of events of the intracellular infection of amoeba with legionellae, as well as the immunogold technique for labeling of host cell proteins such as BiP, which is involved in a chaperoning function of proteins across the membrane of the endoplasmatic reticulum.

It is well established that Legionella pneumophila is an important cause of nosocomial and community-acquired pneumonia. L. pneumophila serogroup 1, the most prevalent serogroup, can be divided into several subtypes by using monoclonal antibodies (MAb). Researchers showed recently that changes in the lipopolysaccharide (LPS) MAb binding patterns are associated with changes in the virulence properties of L. pneumophila. Several studies showed that a majority of clinical isolates, especially strains associated with outbreaks, carried an LPS epitope that reacts with MAb 2 of the international standard panel and the MAb 3/1. Researchers used genetic fingerprinting to investigate several L. pneumophila serogroup 1 strains originated from the same source that differed in their reactivity with MAb 3/1 but were indistinguishable or very similar. Furthermore, researchers investigated whether the loss of the reactivity with MAb 3/1 resulted in differences in the multiplication in Acanthamoeba castellanii and macrophage cells. The lag-1 gene was amplified from chromosomal DNA and sequenced using an ABI 377 sequencer.

Legionella pneumophila is an intracellular parasite of free-living amoeba species. In natural or man-made freshwater systems, L. pneumophila is frequently found in tight association with biofilms. In the human host, L. pneumophila is an intracellular pathogen of alveolar macrophages and blood monocytes. The authors have described the phase-variable expression of a lipopolysaccharide (LPS) epitope in L. pneumophila serogroup 1 strains. Associated with LPS phase variation, they observed the loss of virulence and serum resistance in the mutant strain. The profile of the primarily long-chain 3-hydroxylated fatty acids was shifted to shorter chains by about two carbons on average in phase-variant 811 compared with wild-type RC1. In the virulent wild-type strain RC1, the 30-kb element is located in the chromosome, whereas excision from the chromosome and replication as a high-copy plasmid resulted in the mutant phenotype. The colony material of the mutant strains is very sticky and appears slimy when suspended in aqueous solutions, which may facilitate adherence to and survival in biofilms outside a host cell and may protect the cells from dehydration as well.

For pathogenic bacteria, evasion from lysis by serum complement factors in the human host is essential for survival and virulence. Therefore, many pathogens have developed effective strategies to overcome eradication by complement. Such strategies include removal or destruction of complement factors, inhibition of complement activation, or imitation of complement protein by molecular mimicry. Legionella pneumophila activates complement via both pathways, the classical and the alternative. Lipopolysaccharide (LPS) phase variation is associated with serum sensitivity and loss of virulence. Wild-type RC1 and the phase-variant mutant 811 exhibited significant differences with regard to resistance to serum complement factors. Strain RC1 was relatively serum resistant, whereas when mutant 811 was incubated with 40% normal human serum at 37№C, no viable bacteria were recovered after 15 min. Therefore, the two strains provide a valuable tool for comparative investigation of serum resistance in L. pneumophila. In both strains, RC1 and 811, the most abundant portion of C3b is inactivated and iC3b is the predominant molecule deposited on the L. pneumophila surface.

The bacterial DnaJ-like protein DjlA is composed of two functional domains, an N-terminal transmembrane domain of type III topology and a J domain at the C terminus. The authors investigated the Legionella pneumophila DjlA protein. The plasmid constructs were used to transform L. pneumophila and Escherichia coli strains. Overexpression of the L. pneumophila djlA gene in E. coli K-12 resulted in a mucoid phenotype due to colanic acid capsule production. The djlA mutants of all four strains did not exhibit alterations in growth rate in buffered yeast extract (BYE) broth when compared with the appropriate wild type, suggesting that djlA is not an essential gene in L. pneumophila. Incubation with a pool of 40% normal human serum revealed that the djlA mutants were as serum-resistant as the parent wild-type strains. The obtained CFU counts for strains R458 and Corby and the corresponding mutants are depicted. Thus, deletion of the djlA gene resulted in reduced virulence in all investigated L. pneumophila strains. To our knowledge, this is the first report of an influence on virulence by the djlA gene product in pathogenic bacteria. However, the authors do not know by which mechanism virulence attenuation is achieved in the L. pneumophila djlA mutants.

A change in the lipopolysaccharide (LPS) upon phase variation in Legionella pneumophila serogroup 1, subgroup OLDA, wild-type strain RC1 was detected with the aid of LPS-specific monoclonal antibody MAb 2625. Chromosomal insertion and excision of a 30-kb instable genetic element of possibly phage origin was identified as the molecular mechanism for phase variation. The core of the LPS is a nonasaccharide that lacks heptose and phosphate, contains abundant 6-deoxy sugars, and is highly O- and N-acetylated. The MAb 2625 epitope is present in wild-type RC1 cells and is lost in the spontaneous mutant 811 upon phase variation. Lipid A was prepared by mild acid hydrolysis (0.1 M NaOAc-HOAc buffer, pH 4.4,100№C, 4 h) of LPS each of wild-type RC1 and phase-variant 811 followed by centrifugation and lyophilization of the pellet. LPS biosynthesis pathways involved in assembly of lipid A were also affected by phase variation, which resulted in a specifically altered lipid A structure with a modified profile of fatty acids of a particular type. Phase variation may affect a regulatory factor, which influences LPS biosynthesis, virulence, and serum resistance.

To persist in the environment, Legionella pneumophila that is confronted with feeding amoebae, must avoid digestion and instead establish a protected site for intracellular replication. After exhausting the resources of its host, the bacterial progeny must vacate the premises to search out a more fertile locale. When L. pneumophila encounters grazing amoebae, its prowess as a parasite becomes paramount. A model for the L. pneumophila life cycle in which growth phase determines virulence expression and replication vacuole biogenesis has been provided in this chapter. A striking observation made by Brenda Byrne motivated the authors' investigation of L. pneumophila virulence regulation. When broth cultures of L. pneumophila exit the exponential growth (E) phase, the bacteria become osmotically resistant, sodium sensitive, cytotoxic, motile, infectious, and competent to evade macrophage lysosomes. Results obtained by four independent experimental approaches strengthen the hypothesis that L. pneumophila survives and replicates within lysosomal compartments. First, virtually every vacuole that harbors more than five bacteria also contains LAMP-1. Second, bacteria within endosomal vacuoles are viable, as judged by their capacity to respond to a metabolic inducer by expressing a gfp reporter gene. Third, replicating bacteria obtained from macrophages, but not broth, are acid resistant. Finally, inhibition of vacuole acidification and maturation by bafilomycin A1 or other agents actually inhibits bacterial replication.

This chapter focuses on the cyclical nature of some morphological and physiological changes that take place in infected HeLa cells and during growth in vitro. Coverslip cultures of HeLa cells were infected with purified mature intracellular forms (MIFs), and changes in Gimenez staining were followed for up to 3 days. The bacterial strains used throughout these studies were Lp1-SVir and 2064, two virulent strains previously reported and characterized. Cultures with 100% green forms (after Gimenez staining) were identified upon growth of Legionella pneumophila on buffered charcoal-yeast extract (BCYE) plates. Rowbotham reported earlier that infection of amoebae resulted in the formation of small, highly motile, and infectious legionellae that are believed to be equivalent to the HeLa-derived MIFs reported. Purified MIFs were placed in buffered yeast extract (BYE) and their growth was followed through viable cell counts. Researchers concluded that the green forms arising in the BYE cultures were replicating bacteria that developed from the MIFs originally used as inoculum. In summary, L. pneumophila undergoes cyclical morphological and physiological changes that suggest the existence of a developmental program.

The complexities of the relationship between Legionella pneumophila and other microorganisms for the growth and survival of legionellae in the environment is still not fully understood, though it is generally believed that amoebae play an important role in the natural environment. Although strains of varying virulence are isolated from environmental sources, it is still generally accepted that intracellular replication is important for the ability of L. pneumophila to proliferate in the natural environment. The model water system allowed the development of reproducible biofilms. A diverse but fairly constant consortium of aquatic microorganisms including fungi, bacteria, and protozoa could be maintained in this system. To investigate if legionellae could have the potential to replicate without multiplication within a protozoal host, an avirulent strain of L. pneumophila serogroup 1 Pontiac (Corby Strain) (CAC) was added to the system. If intracellular multiplication is essential for the proliferation of L. pneumophila in aquatic systems, then elimination of trophozoites in the model system would prevent any further growth of legionellae, and consequently numbers would decrease due to dilution by the continuous culture medium.

Several approaches have been taken to understand the genetic basis for intracellular multiplication by Legionella pneumophila. The first was based on the properties of avirulent variants that had lost the ability to replicate inside macrophages and were also defective in preventing phagosome-lysosome fusion. Researchers reasoned that introduction of a wild-type region of the L. pneumophila genome that restored the ability to replicate within and kill human macrophages would provide information about the genes that were defective in the variants. A genomic library of wild-type L. pneumophila DNA was introduced to one of the avirulent variants, and complemented bacteria that regained the abilities to replicate intracellularly and kill host cells were identified with the aid of a plaque assay. The striking homology between the icm/dot genes and the IncI tra and trb genes involved in conjugal DNA transfer prompted us to examine if L. pneumophila could conjugally transfer DNA. The current model for how the Icm/Dot complex influences the intracellular fate of L. pneumophila postulates that there are two classes of Icm/Dot gene products. Several possible mechanisms could be used to specifically interrupt the fusion of phagosomes with lysosomes.

This chapter aims at obtaining the complete sequence of the Philadelphia 1 strain of Legionella pneumophila, derived from the original 1976 outbreak that established the clinical entity known as Legionnaires' disease. The GC content of L. pneumophila is ~38% over most of its length. From the genomic sequence so far (more than 85% of the sequence is currently contained in contigs greater than 1 kb), 3,000 open reading frames (ORF) were covered, about 2,000 of which are complete, and on the basis of homology to genes from other prokaryotes, more than 1,100 putative genes in L. pneumophila were identified. Genes in L. pneumophila Philadelphia 1 can be compared with genes in other pathogenic organisms. The amino acid sequence can guide structural and functional analyses of enzymes and other proteins of the organism. The authors used knockout strategies in the past to assess the function of genes discovered in the course of the sequencing project; and this continues to be their preferred method for functional analysis.

This chapter examines whether a chromosomal conjugation system exists in Legionella pneumophila. Mating experiments between K6 (approximately 2.6 X 109 CFU) and Chicago-2S (approximately 8.5 X 109 CFU) typically yielded 103 Kmr Smr LacZ+ recombinants. All recombinants tested (100/100) belonged to serogroup 6, which is the same serogroup as strain Chicago-2S. This result suggested that the donor was strain K6 and the recipient was strain Chicago-2S. The mechanism of the DNA transfer was most consistent with conjugal transfer. To genetically examine the DNA transferred from K6 to Chicago-2S, genotyping by arbitrarily primed PCR (AP-PCR) and pulsed-field gel electrophoresis (PFGE) were performed among strains K6, Chicago-2S, and their transconjugants. The restriction fragment patterns of the transconjugants were distinct from the pattern exhibited by K6 and were similar to that exhibited by Chicago-2S. The transfer of DNA from the chromosome of strain K6 to Chicago-2S could occur by a variety of mechanisms. Transposition would presumably transfer only the transposition sequences to random sites that would differ in different recombinants. Researchers showed that a thymine auxotroph of L. pneumophila was repaired by plasmid R68.45-mediated chromosomal mobilization of a prototrophic donor strain. Therefore, if L. pneumophila had acquired an onT by integration of these conjugative elements, it would become selfmobilizable like the Hfr of Escherichia coli.

Legionella pneumophila is a ubiquitous opportunistic pathogen that causes serious infection in humans, especially in immunocompromised individuals. Adaptive immunity consists of activated T cells, cytokines such as gamma interferon (IFN-γ), and activated macrophages that restrict the growth and spread of bacteria within its intracellular environment. Innate immunity to the bacteria, on the other hand, is perhaps even more complex and includes a variety of different host cells and cytokines. The innate immunity to L. pneumophila infection limits the growth of the bacteria, which generally occurs in macrophage. Therefore, the regulation of L. pneumophila growth in macrophages by proinflammatory cytokines, such as TNF-α, may be a main event in the early phase of infection. TNF-α, which is mainly produced by macrophages, is a potent cytokine and has extensive properties affecting the regulation of immunological and inflammatory cells. In general, the adaptive immunity to bacterial infections is characterized by the memory of the previous antigen stimulation, which is not necessarily the infection, and the transferable nature of the memory with lymphocytes.

Interleukin-12 (IL-12), one of the key cytokines in regulating the development of Th1 response, is a heterodimeric cytokine composed of two disulfide-linked p35 and p40 subunits. Both subunits have to be expressed within the same cell to produce biologically active p70 heterodimer. The production of IL-12 by monocytes/macrophages is induced by exposing responsive cells to a variety of microbial products such as lipopolysaccharide (LPS). The suppression of IL-12 by Legionella pneumophila infection may be occurring at the level of gene transcription. The analysis of the steady-state levels of IL-12 p35 and IL-12 p40 mRNA determined by reverse transcriptase-PCR (RT-PCR) showed that L. pneumophila infection induced the suppression of mRNA accumulation for the IL-12 p40 gene in response to LPS stimulation, but not the IL-12 p35 gene, which was constitutive in all macrophages treated. L. pneumophila infection induced production of MCP-1 protein and up-regulated the LPS-induced production of MCP-1 protein regardless of the virulence of L. pneumophila and its viability in the macrophages. The experimental mouse infection with L. pneumophila resulted in certain levels of IL-12. Since IL-12 production is regulated by multiple mechanisms, including cytokines produced by other cells, it can be speculated that the suppressed IL-12 by infection may be overcome by other systems in vivo, such as the host defense system.

Legionella pneumophila causes a severe atypical pneumonia in humans, the so-called Legionnaires' disease. Although the disease mostly occurs in immunocompromised subjects, several virulence factors of Legionella have been described that confer to the microorganism an intrinsic aggressive potential. The ability of this pathogen to replicate in host macrophages is considered a major virulence trait. For this reason, the susceptibility to the experimental infection with L. pneumophila observed in guinea pigs and in selected mouse strains has been generally explained by the permissiveness of macrophages of these animals for Legionella invasion and intracellular replication in vitro, as reported, for elicited peritoneal macrophages from susceptible A/J mice. Researchers found that the i.p. inoculation of virulent Legionella cells from the VIR+ strain was rapidly lethal for A/J mice, since 100 and 60% of the animals acutely died within 2 days following the challenge with 2 X 108 or 2 X 107 bacterial cells. Unchecked TNF-α production upon i.v. Legionella inoculation has recently been reported as the causative factor in the death of Legionella replication-unpermissive BALB/c mice. Accordingly, the protection afforded by IL-10 treatment or TNF-α neutralization in the mice model also indicates that the release of TNF-α, a well-known toxic shock-related cytokine, plays a major role in mediating the lethal effects of the virulent legionellae. Therefore induction of proinflammatory cytokines, in particular TNF-α, besides being an important factor facilitating the resolution of L. pneumophila infections, might also be regarded as a possible contributory factor to the pathogenic action of virulent L. pneumophila strains.

This chapter talks about the role of the flagellum in the observed protection from legionella infection that was investigated by the authors. First, the minimal lethal dose (MLD) of two different strains of Legionella pneumophila was determined by infecting intraperitoneally (i.p.) female A/J mice, aged 4 to 6 weeks, with different doses of an L. pneumophila serogroup 6 strain of human origin (VIR+), or from the L. pneumophila serogroup 1 Corby strain. A flagellar preparation was performed from a 6-day culture of the flagellated VIR+ strain on buffered charcoal-yeast extract (BCYE; Oxoid, Italy) agar plates at 36№C, in humidified atmosphere and 2.5% CO2. Mice were observed for signs of illness and for survival for 15 days. In flagella-immunized mice, no mortality was observed, compared with 100% mortality of similarly challenged mice immunized with the sham-flagellar preparation or unimmunized control mice. Involvement of humoral immune responses in the protective effect due to flagellar immunization was investigated, also considering that flagella are present in most Legionella species and serogroups, and their composition is common in all of them.

Legionellae were initially characterized by criteria typically used to describe an autonomous unicellular organism. Bacteria of the genus Legionella were described as gram-negative, aerobic, and rod shaped with one or more polar or lateral flagella. A more complete ecological profile of legionellae will provide the scientific basis for selecting the most appropriate host for pathogenesis studies as well as studies to identify procedures to control amplification of the bacteria in certain environments. Analysis of bacterial group behavior represents the most recent facet in the study of the ecology of legionellae. In building water systems, legionellae are most frequently detected in biofilms of plumbing fixtures and heating, ventilating, and air-conditioning (HVAC) equipment. The complex nutrients available with biofilms have led some researchers to propose that the biofilm might allow the survival and multiplication of legionellae outside a host cell. If legionellae can multiply extracellularly within biofilms, the study and characterization of this phenomenon could have tremendous impact on control strategies for the prevention of legionellosis. Researchers have only begun to characterize the interaction of legionellae with the microbial community. Controlling legionellae within biofilms may lead to the most effective control measures to prevent legionellosis.

Infections of mammalian and protozoan cells by Legionella pneumophila share a number of similarities but differ in several attributes. The mechanisms of killing of the host cell and release of intracellular bacteria after termination of intracellular replication are not known for L. pneumophila or any other vacuolar intracellular pathogen. During screening of the mini Tn10::kan mutant library the intracellular bacteria, L. pneumophila, belonging to the parental strain AA100 were released into the tissue culture medium with in 24 to 48 h postinfection. In contrast, despite the prolific intracellular replication of the five mutants, they were "trapped" within and failed to egress from macrophages and epithelial cells during the 48-h infection, and the majority of the infected cells remained viable and intact. Researchers have recently proposed a model of biphasic death of mammalian cells by L. pneumophila initiated by caspase-3-dependent apoptosis followed by necrosis, which is probably mediated by the pore-forming toxin. To confirm the roles played by the pore-forming activity in necrotic killing of Acanthamoeba polyphaga and subsequent release of the intracellular bacteria, researchers examined infections at a multiplicity of infection (MOI) of 500, for the following two reasons. First, infection at this MOI may ensure that most of the cells were infected by L. pneumophila and may allow better examination of whether the rib mutants were defective in exiting the protozoan host. Second, it has been shown that extracellular L. pneumophila induces rapid necrotic killing of mammalian cells within 20 to 180 min at an MOI of 500.

Legionella pneumophila is a facultative intracellular bacillus that causes nosocomial and community-acquired pneumonia and, rarely, extrapulmonary infections in humans. Signature-tagged mutagenesis employs uniquely tagged transposons that are used to randomly mutagenize a bacterial chromosome and create a library. A library of 700 mutant clones created by signature-tagged mutagenesis was screened using this negative-selection strategy in Acanthamoeba castellanii, a free-living amoeba that may serve as an environmental reservoir of legionellae. The efficiency of invasion was studied by incubating L. pneumophila strains grown to postexponential phase with A. castellanii, using gentamicin to kill extracellular organisms and then determining remaining intracellular CFU. The behavior of the mutants was also examined in human macrophages derived from peripheral blood mononuclear cells (PBMCs) and in phorbol myristate acetate (PMA)-differentiated U-937 cells. In contrast, all of the mutants replicated within PMA-differentiated U-937 cells, indicating that U-937 cells are deficient in killing L. pneumophila compared with freshly isolated macrophages. The homologous gene in Salmonella enterica serovar Typhimurium is linked to virulence since mutants are attenuated in their ability to kill mice. Two mutants, STM-1 and STM-4, appeared to be deficient in A. castellanii alone, and two other mutants, STM-2 and STM-5, led to intracellular replication defects that were observed in both A. castellanii and H. vermiformis.

Several pathogens exhibit a considerable host range. Legionella pneumophila, for example, can infect various protozoa species, experimentally inoculated guinea pigs, and human macrophages, as well as epithelial cells. Vegetative cells of Dictyostelium feed on bacteria and upon starvation aggregate and differentiate into pluricellular fruiting bodies. To evaluate whether D. discoideum is a suitable model system for studying Legionella pathogenicity, the authors compared the intracellular growth of different Legionella species in Dictyostelium with the established host model system Acanthamoeba castellanii. The FlaA-and the Mip-negative mutant of L. pneumophila Corby revealed moderate growth defects and the ligA-negative mutant was severely impaired to grow intracellularly. To examine host functions required for growth, the authors also investigated defined Dictyostelium mutants. Identification of Legionella spp., Dictyostelium spp., and Hartmannella spp. by fluorescence-labeled rRNA probes have been provided in this chapter. Available molecular tools to manipulate the host are transformation with integrating and nonintegrating eukaryotic vectors, homologous recombination, antisense techniques, and restriction enzyme mediated integration (REMI). The application of these methods should allow the elucidation of the interaction of bacterial virulence factors with specific host targets.

Legionella pneumophila is the major agent responsible for Legionnaires' disease. To identify proteins that participate in the interaction of Legionella and its host, the authors screened a genomic library of L. pneumophila strain Corby with anti-Corby antiserum. P16-related proteins and p16 sequences were present exclusively in strains belonging to L. pneumophila. Characterization on the protein level was done by screening whole bacteria with MAb 61/1 as dot blot. The mutant showed the expected reaction type-loss of reactivity with the P16-specific monoclonal antibody. By comparing growth on the solid medium it was determined that p16 is not essential for growth of L. pneumophila and that genetic manipulations did not generate additional effects that influenced growth or colony morphology. To investigate the importance of L. pneumophila P16 for attachment, uptake, and/or intracellular growth in its different host cells, infection assays were performed in differentiated U937 cells and in Acanthamoeba castellanii. The wild-type phenotype was restored in the complemented mutant. From these results the authors speculate that the surface-exposed 16-kDa protein promotes the attachment/initial uptake and might act as an adhesin in the receptor-mediated uptake mechanism in amoebae. Further infection assays with other known Legionclla natural hosts will show whether the described locus p16 might be responsible for the broad host spectrum of L. pneumophila in contrast to the other species of Legionella.

This chapter aims at defining a minimum multiplicity of infection (MOI) and determining if the presence of non-Legionella species of bacteria would influence the recovery of Legionella pneumophila CFU during replication and lysis of amoebae host cells. It also aims at documenting the impact of adding the free-swimming ciliate Tetrahymena to the amplified number of Legionella after they had lysed the population of amoebae host cells. If amoebae infection by L. pneumophila was the result of inadvertent contact and subsequent phagocytosis, an inhibitory effect might have been expected since addition of exogenous bacteria would likewise result in increased inadvertent contact with phagocytosis, and thus diminish potential contact and uptake of virulent L. pneumophila. As a direct result of L. pneumophila amplification, amoebae lysed, liberating the bacteria, which were then concentrated into vesicles by the Tetrahymena sp. A conclusion reached in this study is that L. pneumophila can initiate infection in amoebae at an MOI considerably lower than often used in laboratory studies.

The growth-promoting properties of 10 selected synthetic materials, as well as copper and stainless steel, were tested in static (batch) experiments at 25°C and also in a dynamic (flowthrough) test at ambient temperature simulating flow conditions in a household plumbing system. The study objectives were (i) to compare the microbial growth-promoting properties of various synthetic materials with those of copper and stainless steel under static and dynamic test conditions, (ii) to determine the effect of the materials on the multiplication of Legionella, and (iii) to evaluate the materials on the basis of the obtained results. The Legionella growth potential (LegGP) (CFU/cm2) of the materials in the biomass production potential (BPP) test was calculated from the concentrations of Legionella in water and in biofilm after 56, 84, and 112 days of incubation of the flasks with the material samples. The results of the static test reveal that materials may affect biofilm formation at long periods of stagnation and a temperature of 25°C. However, in practice also, water contributes to biofilm formation. Data about biofilm concentrations and Legionella in plumbing systems are needed to enable a quantitative evaluation of the BPP and biofilm formation potential (BFP) values as observed for the various materials in the static test.

Effective antimicrobial chemotherapy of Legionnaires’ disease is driven primarily by the location of Legionella pneumophila inside cells, primarily alveolar macrophages. Determining optimal chemotherapy for Legionnaires’ disease by conducting truly informative clinical trials is generally of little use since the disease is uncommon and because the fatality rates in most community-acquired cases are relatively low. For effective chemotherapy to occur there must be colocalization of the bacterium and antimicrobial agents, and the antimicrobial agent must retain bioactivity against the bacterium. The β-lactam drugs are instead found in the cytosol, probably the main reason why they are ineffective in the treatment of Legionnaires’ disease. In cell culture model, erythromycin therapy results in survival rates that are the same for azithromycin or active quinolone antimicrobial agents. A number of obstacles have prevented such studies, including the great cost of diagnostic testing for Legionnaires’ disease, the relative rarity of the disease, and the usually low mortality in most community-acquired cases of the disease. Erythromycin is relatively poorly active against Legionella pneumophila in cell culture and animal model studies. The superior drugs are azithromycin, levofloxacin, and some of the newer quinolone antimicrobial agents. There is little good evidence that addition of rifampin is beneficial for those receiving azithromycin, levofloxacin, or ciprofloxacin, and there is potential for undesirable toxicities or side effects from addition of rifampin.

Legionella pneumophila is the etiologic agent of Legionnaires' disease, which is a very serious illness with pneumonia and a mortality of 15 to 20%. In addition to L. pneumophila, more than 40 other Legionella spp. are known (designated in this chapter as non-L. pneumophila species), of which several species have been shown to be pathogenic for humans. To improve diagnosis of Legionella infections, the authors designed a novel PCR to specifically amplify all Legionella DNA. Specific probes LPN and LSPP were used for discrimination of, respectively, L. pneumophila and non-L. pneumophila species. The sensitivity of PCR was determined by spiking bacteria in negative clinical material. The lower detection limit was found to be at least 0.1 CFU. A total of 208 samples from 208 patients clinically suspected of legionellosis were subjected to the Legionella PCR/probe procedure. Comparison of results of PCR, serology, culture, and urine antigen detection revealed that PCR-based detection of L. pneumophila, and serology in reconvalescent serum, yielded 2.8-fold more positives than Legionella culture. PCR may provide an important contribution to an early diagnosis of legionellosis. PCR is the only method suitable to diagnose non-L. pneumophila infections. Future studies could include sequencing of non-L. pneumophila PCR products to get an impression of the distribution of Legionella species in patient samples.

Legionella pneumophila is known as an important cause of nosocomial pneumonia. This chapter discusses three cases in an oncological hospital. A 47-year-old woman with malignant non-Hodgkin lymphoma, was hospitalized for more than 10 days, treated with high-dose chemotherapy. During treatment she developed a pneumonia, with a positive urinary Legionella antigen test (Biotest enzyme immunoassay urinary Legionella antigen). Despite therapy with macrolides and fluoroquinolones, the patient died from respiratory failure 1 week after the diagnosis. Immediately after the detection of the first case of Legionella pneumonia, the water distribution system was sampled in several sites, including the shower in the patient’s room, the intake from the municipal water supply, the water storage tank, and some distant outlets. A therapy with macrolides and fluoroquinolones was effective in curing another patient with high-risk breast carcinoma being treated with high-dose chemotherapy. It is remarkable that the diagnosis has always been performed by the detection of the urinary Legionella antigen, although the colonizing subtype was of the serogroup 2-14. A 54-year old man had received appropriate chemotherapy and general condition was improving, but at the end of the month, the patient developed Legionella pneumonia, as documented by a positive urinary Legionella antigen. Despite macrolide and fluoroquinolone treatment, the patient died 5 days later. The flush and heat decontamination method is not effective in eradicating L. pneumophila from the hot water distribution system, and it is therefore mandatory to keep surveillance measures for early diagnosis and prompt treatment of possible new cases.

This chapter compares human leukocyte antigen (HLA) frequency and HLA haplotype frequency in Legionella-positive patients with frequencies in healthy individuals from a local panel (control group). Sputum, bronchoalveolar lavage (BAL), or urine obtained from 114 patients after solid organ (heart, kidney, liver, pancreas) transplantation were examined for the presence of Legionella by using the culture method (sputum, BAL), and the direct fluorescent antibody assay method (sputum, BAL), and by the detection of urinary antigen (urine) during the 4 months after transplantation. A buffered charcoal-yeast extract medium (OXOID) was the culture method used in parallel with BMPA and GVPC selective supplements. Legionella serotypes were determined by using microagglutination test. The monoclonal antibody of MONOFLUO L. pneumophila IFA test kit was used for the direct fluorescent-antibody assay method. The urinary antigen was detected by using Legionella Urin antigen enzyme immunoassay. Differences were found in HLA-A antigen frequencies and HLA-B antigen frequencies between Legionella-positive patients and a control group. Differences in HLA A-B haplotypes and HLA A-B-DR haplotypes were not statistically significant between healthy individuals and Legionella-positive patients. Several published reports suggest that natural resistance or susceptibility to infection with Legionella might be genetically determined. However, no significant difference in HLA class II antigens and the frequencies of haplotypes HLA A-B and HLA A-B-DR was found between both defined groups.

From January 1994 to June 2000, legionellae were isolated by culture from 195 patients with Legionnaires’ disease in Denmark. In this survey the authors present the Legionella serogroup and Legionella subgroup distribution of the isolates stratified according to where the patients acquired the infection: in the community (in Denmark) or during hospitalization or during traveling abroad. Isolates from 190 patients were L. pneumophila of which 185 could be assigned to a definitive serogroup or subgroup. Five isolates were non-L. pneumophila: one was L. micdadei, two were L. hozemanii, and two were L. longbeachae. For cases where the infection was acquired in Denmark, only 39.7% (69 of 174) of cases were caused by L. pneumophila serogroup 1 (Pontiac subgroups only 27.6%). L. pneumophila serogroup 3 was a relatively common cause of community-acquired and nosocomial Legionnaires’ disease in Denmark, accounting for 20% of all culture-confirmed cases. The sensitivities of the commercially available Legionella urinary antigen tests for serogroups other than serogroup 1 is low and is also lower for serogroup 1 non-Pontiac subgroups than for Pontiac subgroups. It is therefore important to supplement the use of Legionella urinary antigen assays in the diagnosis of Legionnaires' disease with other methods such as PCR and culture, particularly for the groups of patients with underlying disease and for those who are infected during hospitalization.

A study by Benson et al. compared the usefulness of urinary antigen detection for diagnosing Legionnaires’ disease caused by L. pneumophila serogroup 1 of different monoclonal subgroups in addition to cases caused by other serogroups. A total of 152 urine samples from 152 patients with culture-proven legionellosis were tested. Urine samples were tested simultaneously using Binax Legionella urinary enzyme immunoassay (EIA) and Biotest Legionella Urin Antigen enzyme immunoassay (EIA). Taking all cases of legionellosis into account, a sensitivity of 77% for both Binax EIA and Biotest EIA was found. This is in good agreement with other studies. In addition, the authors demonstrated that the sensitivities for infections caused by monoclonal antibody (MAb) 3/1- and MAb 2-positive strains are much higher, exceeding 95%; thus, urinary antigen detection is the best method for diagnosing community-acquired as well as travel-associated cases. Diagnosis of such infections has especially far-reaching implications for nosocomially acquired legionellosis, which is significantly more often caused by strains belonging to monoclonal subgroups OLDA/Oxford and Bellingham or to serogroups 2 to 15 than are community-acquired infections.

The Binax enzyme immunoassay (EIA) and radioimmunoassay (RIA) both showed a certain cross-reactivity with non-serogroup 1 Legionella pneumophila, as is to be expected when using polyclonal rabbit antibodies for the antigen detection. The Biotest EIA is the former urinary antigen EIA from the Robert Koch Institute in Berlin developed by Fehrenbach for extended cross-reactivity with non-serogroup 1 L. pneumophila. Benson demonstrated that only 10 of 45 non-serogroup 1 urine samples previously positive in the broad-spectrum enzyme-linked immunosorbent assay described by Tang were positive with the Binax EIA (22%), whereas 13 of 42 of these samples were positive with the Biotest EIA (31%); when using a lower cut-off, 20 (48%) were positive with Biotest EIA. 28 non-serogroup 1 samples were evaluated from different countries provided by users of the test to demonstrate reactivity of the two urinary antigen tests with these urines. Binax EIA and Biotest EIA kits were used according to the test kit instructions. The Biotest EIA therefore demonstrates a considerably broader cross-reactivity with urines from non-serogroup 1 L. pneumophila infections than the Binax EIA. Certainly more potential exists by fully exploiting the advantages of polyclonal antibodies from rabbits, broadly cross-reactive with Legionella species in the Biotest EIA, in continuing efforts to further improve reactivity with Legionella species other than L. pneumophila.

Current methodologies for the diagnosis of legionellosis are time-consuming and labor-intensive, which dissuades many doctors from ordering the tests. To make the urinary rapid diagnostic, antibody was raised to the carbohydrate portion of Legionella serogroup 1. The antibody was immobilized on nitrocellulose and conjugated to colloidal gold particles to produce an immunochromatographic test (ICT). To evaluate the test, a retrospective study was performed in which 300 frozen archive urine samples were thawed and evaluated using the ICT device. Results were compared with culture diagnosis. A prospective study was then performed in which 93 urine samples were collected from symptomatic patients reporting to the hospital with lower respiratory symptoms or sepsis. The purpose of this study was to evaluate the test with fresh urine samples. These patients were evaluated by culture and the ICT. In these samples, the specificity of the test was 100%. No samples were positive either by culture or ICT for Legionella pneumophila. Since the test can be performed rapidly with accurate results, the test is leading to more patients being properly diagnosed and treated accordingly.

This chapter evaluates the performance of PCR with respect to sensitivity, specificity, and predictive value of a positive test result with the help of the results obtained by other Legionella diagnostic methods used in a routine laboratory. Laboratory results for other diagnostic methods were included if samples were collected within a period of 90 days relative to the sample for PCR. The amplification control thus contained the binding sites of the 16S rDNA primers, was purified by gel electrophoresis, and was added to the mastermix at a concentration producing a distinct amplicon in negative controls without increasing the level of detection for the positive controls. The amplicons of the Legionella PCR were analyzed by gel electrophoresis. The authors used two different assays: from 1995 to 1999 an inhouse assay was used and from 1999 the Biotest Legionella Urin antigen EIA was used. The performance of Legionella PCR as a routine diagnostic method for respiratory infections (Legionnaires' disease) caused by L. pneumophila was acceptable, although not all PCR- positive cases could be verified by other diagnostic methods. The sensitivity of Legionella PCR is probably higher than any other single method.

This chapter investigates the presence of Legionella-specific DNA in patient serum using a 5S RNA PCR. The DNA extraction was monitored by a control PCR , which amplified part of the β-globin gene, and the specificity of the PCR was verified by a combination of Southern blotting with a 50-oligomer Legionella-specific probe and TaqI restriction digestion of the PCR products. The Biotest urinary antigen enzyme-linked immunosorbent assay was performed as per the manufacturer's instruction. The serology and culture tests were performed. In patients with multiple samples there were variations in PCR reactivity depending on the time of sample. In patient 10, the 5S RNA PCR was positive in the first two serum samples (the acute- and early convalescent-phase samples) but negative in the later two convalescent-phase serum samples. The Southern blotting has the advantage that it can be performed on all PCR reactions, but the Taql digest requires a PCR product to confirm the specificity of the PCR reaction. The 5S RNA PCR was sensitive and specific and should increase laboratory diagnosis of Legionella infection when there is a nonvalidated serological response.

In this chapter the authors report the application of fluorescence in situ hybridization (FISH) for the direct detection and identification of Legionella spp. and L. pneumophila from respiratory tract specimens of patients with pneumonia using published 16S rRNA-based probes specific for the genus Legionella (Leg705) and for L. pneumophila (LegPne1). To evaluate the usefulness of FISH as a clinical diagnostic method, the authors analyzed 160 respiratory samples (bronchoalvealar lavage (BAL), pleural fluid, sputum, and tissue) from patients suspected of having pneumonia caused by Legionella spp. Probes Leg705 and LegPne1, one labeled with fluorescein and the other labeled with Cy3, were added simultaneously to each well, except for the negative-control well, where hybridization buffer alone was added. Almost two-thirds of Legionella pneumonia cases identified were caused by either L. longbeachae or L. micdadei. Given this background, development of a series of probes specific for the predominant species may be of value for identification and epidemiologic purposes.

This chapter describes the theoretical framework within which typing networks can be established. The definitions used for identifying bacterial strains and clones will be provided, and the clinical value of typing will be illustrated as well. The chapter discusses the current tools available for molecular typing of Legionella pneumophila and their reproducibility. Some of the current developments in the field of bacterial genetic identification are pinpointed, and ongoing and future developments in the field of establishing typing networks are summarized. The most common overall methodological distinction is the one between phenotyping and genotyping procedures. Genes can be specifically amplified and analyzed for sequence variability on the basis of restriction site polymorphism or variable length of the PCR products. Random amplification of polymorphic DNA (RAPD) analysis and arbitrary primed PCR are methods that are essentially based on random priming of PCR primers, often using quite relaxed reaction conditions. Amplification fragment length polymorphism (AFLP) analysis is another recent PCR-mediated procedure with which subsets of DNA restriction fragments can be successfully amplified. DNA sequencing technology has revolutionized microbiology in general because it enabled whole genome sequencing for a variety of microorganisms. Multilocus sequence typing (MLST) is one example: the sequence polymorphism detected in a number of slowly evolving genes allows for the categorization of strains on the basis of allelic diversity. Future technology improvement may increase the self-sustained development of exchangeable data subdirectories in individual laboratories, but we still have to go along way before we get there.

Sequence-based genotyping schemes have become widespread as sequencing costs reduce and suitable sequence databases are created. The genus Legionella is an ideal choice: the number of species is large and the strains are relatively inert when utilizing traditional biochemical tests. The effects of redundancy within the genetic code result in much less constrained, and hence more informative, genetic variation, especially at the third codon position. However, the homology at primer target sites is correspondingly reduced, making primer design more difficult, and often primers targeting protein-encoding genes must incorporate multiple bases at one or several sites to achieve complete consensus. In conclusion, the genotyping scheme targeting the mip gene is currently the most tested and discriminatory of all published schemes for identifying Legionella strains, short of performing whole chromosome DNA hybridization studies. As additional gene targets are reported for Legionella, multilocus sequence typing will become possible, which in turn will ensure that not only will homologous recombination events be detected, but the resolution of significant strain relationships at a subspecies level will be possible.

One patient isolate and one environmental isolate of Legionella pneumophila serogroup 6 were obtained from an outbreak investigation associated with a whirlpool spa, in which 2 cases of Legionnaires' disease and 10 cases of Pontiac fever were diagnosed. The authors used amplified fragment length polymorphism (AFLP) analysis to type the patient and environmental isolates. They compared the patterns obtained from the outbreak-related strains with the electrophoretic types obtained from 48 strains analyzed by multilocus enzyme electrophoreis (MLEE). The previous analysis by MLEE had grouped the 48 unrelated strains of serogroup 6 in to 11 electrophoretic types. L. pneumophila serogroup 6 strains Chicago 2, Johannesburg 5, Albany 1, Oxford 1, SRP 39, Denver 3, Sydney 1, LD 82-683, ED 38, and Vasteras 57/1, the whirlpool and patient isolates, were selected for evaluation of the AFLP technique on the basis of previous subtyping performed by MLEE. Analysis of fragment sizes by agarose gel electrophoresis indicated that the patient and environmental strains were identical; however the 10 unique strains were divided into 7 patterns. By incorporating a fluorescent primer and separating the fragments with the ABI 310, all 10 unrelated strains could be shown to have unique patterns that were separate from the patient and environmental strains. An advantage of AFLP compared with pulsed-field gel electrophoresis or MLEE is that it is easier to perform and less time-consuming.

This chapter investigates the epidemiological relatedness between Legionella strains, isolated from a patient with Legionnaires' disease and from the hospital water supply. Strains of L. pneumophila serogroup 1 isolated from the patient and from the hospital environment, and clinical and environmental unrelated strains were typed by pulsed-field gel electrophoresis (PFGE) and by random amplified polymorphic DNA PCR (RAPD-PCR). L. pneumophila serogroup 1 strains isolated from the patient and 11 strains isolated from hot water samples collected from different sites of the hospital water supply showed the same profile by PFGE with Notl and Sfil. These strains shared identical profiles also by RAPD-PCR with the two primers used. The profiles of L. pneumophila serogroup 1 Philadelphia 1 type strain and of the clinical and environmental unrelated strains were different from the above strains by PFGE and by RAPD-PCR. The molecular typing results demonstrated that the strains of L. pneumophila serogroup 1, isolated from the patient and from the hospital hot water, were indistinguishable and thus genetically related, showing that the hospital water was the source of infection and thus confirming the nosocomial origin.

Molecular typing of pathogenic relevant strains of Legionella pneumophila serogroup 1 is indispensable in determining the epidemiology of the disease. In addition to classical serotyping and subtyping using monoclonal antibodies, a variety of molecular methods have been successfully employed for epidemiological purposes including amplified fragment length polymorphism (AFLP) analysis, pulsed-field gel electrophoresis, random amplified polymorphic DNA, and many other PCR-based methods. These techniques are currently used because of their feasibility, rapidity, and great potential for discrimination. Multilocus sequence typing (MLST) is a sequence-based variation of multilocus enzyme electrophoresis (MLEE), where several genes are compared simultaneously at their sequence level. In the future, the authors plan to validate MLST method for the whole panel of strains, and also to test other target genes, to evaluate the usefulness of MLST for L. pneumophila serogroup 1. MLST appears to be a promising typing technology, and by the selection of novel target genes this method has the potential of becoming the gold standard for L. pneumophila population studies.

Legionella is a naturally occurring aquatic bacterium found in lakes, rivers, hot springs, ponds, and moist soils. It causes Legionnaires' disease, a respiratory illness characterized by pneumonia. The disease is predominantly acquired by inhaling mist from contaminated water sources, and it is associated with considerable morbidity and mortality in immunocompromised hosts. Serological subtyping is insufficiently discriminatory when a given serogroup comprises only a few antigenically distinct subtypes, as for Legionella pneumophila serogroup 6. Furthermore, genotypic differences have been reported in phenotypically similar organisms. Discriminatory molecular subtyping methods may be more informative to establish the environmental source of the illness. This chapter establishes and evaluates PCR and random amplified polymorphic DNA (RAPD) for detection and typing of Legionella from environmental sources. The authors' data show that RAPD analysis can be a better tool to discriminate between isolates compared with serotyping. Thus, it could be more informative in epidemiological studies to determine the source of infection.

To estimate the risk of legionellosis from the plumbing systems, it is a common practice to count CFU in water samples. Interestingly, Legionella can survive outside of protozoa for a longer period of time and can fall-still alive-into a viable but nonculturable (VBNC) physiological state, from which it can be recovered by addition of amoebae. The technique of quantitative competitive-template (QC)-PCR is mainly used in virology. The principle of QC-PCR is based on the simultaneous amplification of the genomic template and a recombined internal PCR standard (IS), which usually can be distinguished as the smaller template. There are a large number of useful PCR targets and primers for specific detection of Legionella. To demonstrate legionellae-specific QC-PCR in principle, the authors chose a 16S rRNA gene-based assay, which has proven to be reliable in the routine diagnostics of clinical microbiology and has recently been proposed as one method for identification of the most common Legionella species. QC-PCR could be evaluated as a fast and simple method for enumeration in infection models or in environmental water samples. This approach might even detect Legionella genomes, which cannot be recovered on agar plates because of low plating efficiency or organisms present in a VBNC state.

When detailed epidemiologic characterization is required, standardized molecular typing methods are indispensable for the subtyping of Legionella pneumophila serogroups. Therefore, a comparison was made between pulsed-field gel electrophoresis (PFGE) and amplified fragment length polymorphism (AFLP) analysis for genotyping of L. pneumophila, using standardized protocols from the European Working Group on Legionella Infections (EWGLI). DNA fingerprint patterns derived from AFLP seem to be superior to the patterns derived from PFGE, in both speed and interpretability. AFLP and PFGE have similar discriminatory power. AFLP and PFGE are powerful tools for the epidemiological subtyping of L. pneumophila. The authors emphasize the importance of the standardization of data analysis and data exchange using the latest software (BioNumerics). Studies are under way to determine both the intralaboratory and interlaboratory reproducibility more extensively for further standardization of these powerful subtyping methods used by an increasing number of European countries.

This chapter evaluates the fluorescent in situ hybridization (FISH) test with a specific 16S rRNA-targeted probe for detecting the presence of viable Legionella pneumophila in water. The results of this FISH test are available after 6 h (without inactivation) or 24 h (with activation). A large number of water samples, mostly from hot water systems, were tested with both test varieties and results were compared with those obtained with the culture method. FISH is a detection technique using a specific fluorescence-labeled DNA probe targeting the rRNA of the cells of the selected microorganism(s). Effect of activation on the result of the FISH test for detecting L. pneumophila in water samples is discussed in the chapter. Interpretation of the FISH test results, in terms of hygienic significance, is even more complicated than such interpretation of colony counts. On the basis of the FISH results following conclusions can be drawn: (i) FISH allows detection of L. pneumophila in samples of water and biofilms within 24 h after sampling, but activation is needed, (ii) in many cases the numbers of cells detected with FISH are much higher than numbers detected with the culture method, and (iii) further improvement of the FISH test, e.g., a level of detection similar to the culture method, is needed.

Legionella spp. are gram-negative bacteria responsible for epidemic and sporadic cases of pneumonia after inhalation of contaminated water droplets from a variety of water sources. This chapter presents the distribution of Legionella species, serogroups, and monoclonal subtypes of Legionella strains isolated in Germany and typed at the National Reference Laboratory for Legionella. Environmental strains were isolated on selective buffered charcoal-yeast extract agar from water samples collected at various locations in Germany. The monoclonal subtypes were named according to the reactivity with reference strains. For macrorestriction analysis, chromosomal DNAs were digested overnight with SfiI, AscI, and Notl (New England Biolabs, Schwahlbach, Germany) and separated with the CHEF III System (BioRad Laboratories, Munich, Germany). In this study, all patients had a radiographically confirmed pneumonia confirmed as legionellosis by isolating a Legionella strain from clinical specimens. Cases were classified as nosocomial if the patient had been hospitalized continuously for a minimum of 3 days at the time of disease onset or had been discharged from the hospital less than 10 days before the onset of symptoms. Of the 133 strains grown from clinical samples, 88 (66%) belonged to serogroup 1, 38 (29%) belonged to serogroups 2 to 15, and 7 (5%) were typed as other Legionella species. These data are in good agreement with results from Europe and the United States.

The isolation of Legionella species from environmental specimens has multiple steps subject to many potential errors, which have a cumulative effect. The growth medium is relatively complex and contains heat-sensitive ingredients, and the preparation involves multiple aseptic additions after autoclaving. The selective media and pretreatments do not suppress all background flora, and the recognition of colonies of Legionella, particularly non-L. pneumophila species, requires experience. In addition background bacterial flora can sometimes suppress the growth of colonies of Legionella completely. It is important that all laboratories in different countries produce equivalent reliable results. To achieve this, each laboratory needs a comprehensive quality assurance program including good internal quality control and participation in external quality assurance (EQA) schemes. The beneficial effects of belonging to an EQA scheme have been demonstrated by the gradual improvement in the consistency of the laboratories. This in turn is leading to more reliable quantification, which is important for monitoring the effectiveness of control measures throughout the world.

The purpose of the current study was to determine the prevalence of Legionella species in public whirlpool spas (in the absence of disease) by examining a large population of these spas over the past 18 years. Additionally, the presence of Legionella was correlated with other microbiological parameters, including the presence of another disease-causing bacterium (Pseudomonas aeruginosa) that may be transmitted via whirlpool spas, and the total bacterial numbers as a measure of the overall microbiology and disinfection status of the spas. A similar relationship was found between P. aeruginosa and high total bacterial counts. On the basis of these results, it would appear that Legionella and P. aeruginosa proliferate primarily in spas where the disinfection (i.e., free halogen concentrations) and other maintenance procedures are inadequate to control the total bacterial numbers. In conclusion, the current data would support a conclusion that Legionnaires' disease is a preventable illness, at least as associated with whirlpool spas. Prevention of disease would be related to tight control of free available halogen levels and other regular maintenance procedures, coupled with bacterial monitoring to assess the effectiveness of these procedures.

This chapter evaluates the efficacy of international detection methods compared with an in house-developed method for the optimum recovery of Legionella from cooling water, determines the effect of heat and acid pretreatment procedures as well as the use of selective supplement on Legionella recovery, investigates possible antibody cross-reactions with nonlegionellae using a polyvalent direct fluorescent antibody conjugate specific for Legionella species known to cause most outbreaks of legionellosis worldwide, and formulates guidelines to assist industry in the management and control of Legionella levels in cooling water. The results of a preliminary study showed that in 50 spiked tap, source, and cooling water samples, acid treatment had the most damaging effect on the recovery of Legionella from tap water, followed by source water and then cooling water. Heat-treated spiked cooling water gave poor Legionella recovery rates, followed by source water and tap water. The most probable number (MPN) method is proposed as the national standard procedure for the detection of Legionella from cooling water in South Africa. In the case of the ISO method, all representative Legionella colonies need to be confirmed using blood agar, buffered charcoal-yeast extract (BCYE) (without cysteine), latex agglutination, and/or direct fluorescent antibodies.

Large numbers of legionellae in water distribution systems present a potentially serious health risk to workers and the general public. Sporadic infections are diagnosed with increasing frequency, clearly illustrating the importance of appropriate methods for Legionella detection. A Legionella Action Group formed in 1995, took up a study as part of an initiative to provide South African laboratories and industries with guidelines for the most appropriate culturing method for the environment, and with information on the prevalence of Legionella in industrial waters. Two of the internationally accepted culture methods and a locally developed adaptation of the most probable number (MPN) method, used by some laboratories in South Africa, were evaluated. The International Standard (ISO) method is widely accepted as a standard and used by some laboratories in South Africa. The quantitative MPN method has been adapted for enumeration of Legionella in water samples by South African workers. Single colonies with the typical groundglass appearance of legionellae were tested for cysteine dependence by inoculating buffered charcoal-yeast extract (BCYE) agar and nutrient agar and incubating these until growth was observed on the BCYE agar, as indicated in the ISO and Australian Standard (AS) methods.

In France, the risk of legionellosis associated with cooling systems had been underestimated for a long time because until 1998, no community outbreak had been related to these devices. The health risk was not taken into account and no risk assessment was developed to manage the cooling system. However, recommendations for prevention of legionellosis were provided in governmental guidelines published in 1997. The maintenance of cooling towers and control of Legionella were not mandated by law. Only compressors associated with cooling systems were submitted for regulation as detrimental devices for the environment. This chapter describes a Parisian experience, including two outbreaks, which should lead companies to improve the health aspects of managing cooling towers. Legionella and L. pneumophila were determined in accordance with the French standard procedure XP T 90-431. The two patient isolates (L. pneumophila, L. pneumophila serogroup 1) had indistinguishable fingerprints using pulsed-field gel electrophoresis and arbitrarily primed PCR to strains isolated from water of one particular cooling tower. Legionella recovered from cooling towers examined during two outbreaks in Paris have been provided in the chapter.

This chapter aims at estimating Legionella contamination of dental-unit waters and antimicrobial susceptibility of isolated strains. The membrane filtration method and growth on a buffered charcoal-yeast extract medium with GVPC selective supplement (glycine, vancomycin, polymyxin B, and cycloheximide) was used. For the estimation of antibiotic resistance, four groups of antibiotics were used: (i) inhibitors of synthesis of cell walls β-lactam antibiotics), (ii) inhibitors of synthesis of proteins (tetracyclines, macrolides, aminoglycosides), (iii) inhibitors of synthesis of nucleic acids (quinolones, rifampin), and (iv) inhibitors of synthesis of cell membranes (polymyxin). L. pneumophila is sensitive to a number of antibiotics; however, in vitro susceptibility studies do not always correlate with clinical efficacy because Legionella is an intracellular pathogen. Water for use with drills in oral surgery should be held in sterile containers and pass down freshly sterilized tubes to sterilized handpieces. Access to the water lines to enable regular flushing with powerful disinfectant would also be desirable.

In Denmark more than 50% of known cases of Legionnaires' disease are community acquired with no known association with traveling. To assess the role of hot water systems in private homes and institutions as possible sources for infection, a pilot study was performed from October 1999 to May 2000. A total of 46 hot water systems in large buildings situated in or around Copenhagen were included in the study. Twenty-four systems were in public buildings in one municipality. The systems included 13 blocks of flats, 14 schools, seven nursing homes, eight sport centers, two industries, and two other institutions. The water samples were analyzed at The National Centre for Hospital Hygiene, Statens Serum Institute, within 2 days of sampling. Two to five colonies from each water sample were further tested by monoclonal antibodies (Dresden MAb panel) that distinguish between 15 serogroups of Legionella pneumophila. All isolates determined to be L. pneumophila serogroup 1 were further characterized by an Enzyme Immuno Assay, distinguishing between MAb 3/1-reactive isolates (Pontiac group) and MAb 3/1-negative isolates (non-Pontiac group). Five of the six serogroups of L. pneumophila identified as causing community acquired Legionnaires' disease in Denmark were isolated from the water samples. The results of this pilot study are in accordance with the possibility that domestic hot water systems in private homes and institutions may be the sources of some of the community-acquired cases in Denmark.

Legionella pneumophila serogroup 1 is the serogroup of Legionella species most frequently isolated from infected patients in Japan and all over the world. The recent findings suggest that one of the most discriminative epidemiological methods is pulsed-field gel electrophoresis (PFGE). To develop the method of epidemiological surveillance, PFGE and monoclonal antibodies were used for the analysis of 27 clinical and 20 environmental L. pneumophila serogroup 1 isolates in Japan. Researchers reported five anti-L. pneumophila serogroup 1 monoclonal antibodies that were specific for serogroup 1 but showed different reactivities, previously. Monoclonal antibody-sensitized latex was prepared and the latex agglutination test on slide plate was performed. There was a close correlation between monoclonal antibody subtyping and PFGE analysis. The result by the two analyses suggested that the clinical isolates, the isolates from cooling towers, and the isolates from hot springs in Japan formed the distinctive genetically or antigenic clusters.

Legionella species are widespread in aquatic environments. Legionella pneumophila is a pathogenic bacteria and is responsible for Legionnaires' disease. Man-made water systems are contaminated by legionellae through the municipal water supply. L. pneumophila is generally found in hot water systems and biofilms that are present in water supply systems and multiplies there within protozoan cells. Bacterial transmission to humans occurs through droplets generated from cooling towers, shower heads, and other man-made devices that generate aerosols. The relationship between L. pneumophila and total viable bacteria counts (22°C) was performed by linear regression. A cultural analysis of water systems of 139 hotels showed that 57 of the hotels (41%) were contaminated by L. pneumophila. L. pneumophila serogroup 1 was isolated from 29 (23.3%) water samples, whereas L. pneumophila serogroups 2 to 14 were isolated from 95 (76.6%) water samples. The mean water temperature varied from 15 to 62°C in cold and hot water samples. It was found that the effective growth temperature of L. pneumophila ranged from 28°C to 41°C. An epidemiologic case has been notified in one of the hotels surveyed in which L. pneumophila serogroup 1 was isolated.

Identification of the organism responsible for Legionnaires' disease was first associated with an outbreak in 1976 in Philadelphia, Pa., where 189 cases and 29 deaths occurred among American Legion veterans who had attended a hotel-based convention. European Working Group for Legionella Infections (EWGLI) provides expert epidemiological, microbiological, and environmental health advice to clinical and public health officials, tour operators, national and international government departments, and other relevant groups such as hoteliers and members of the public. This chapter mainly presents data from 1999, which was provided by 28 of the 31 countries that currently participate in the European scheme. It also compares trends and incidence of cases in Europe from 1993. Since 1993, the overall rate of infection per million residents has ranged from 3.35 to 4.46. In 1999, 32% of the overall cases were reported as community acquired, 9% were hospital acquired, and 21% were associated with travel either in their own country or abroad. The objectives of the December 1999 meeting of experts convened by the European Commission to review the Belgium and The Netherlands whirlpool spa outbreaks, were to advise the Commission on the need for European recommendations and guidelines for the control and prevention of future whirlpool spa outbreaks and to produce specific recommendations that could be addressed by the Commission. EWGLI is assisting data contribution for effective surveillance at the international level through its sharing of information on cases, outbreaks, sources of infection, and continued developments in epidemiological, microbiological, and environmental aspects of legionella infection.

The recent outbreak of Legionnaires' disease at the Melbourne Aquarium has prompted widespread comment by public health officials regarding testing for Legionella. Cooling tower maintenance and biocide dosage have been shown to be major contributors to risk management, and review papers have stated that with proper management, Legionella can be readily controlled. Regular cleaning regimens and inspection and review of the cooling towers and water treatment equipment are also important aspects of a control strategy. Operation of cooling towers and cooling water temperature has been shown to positively correlate with concentrations of Legionella in cooling towers. Prevailing weather conditions such as cloud cover, ultraviolet light intensity, air temperature, and relative humidity have been shown to influence the survival and dispersion of Legionellae in aerosol. Routine maintenance procedures by owner/operators and water treatment personnel could be electronically reported to the system and risk assessments for each system could be made.

This chapter presents biweekly sample data from 28 small cooling towers colonized by Legionella over a 4-month summer period. These data were used to appraise the validity of risk assessments on the basis of routine Legionella sampling and to investigate the temporal variation in concentrations of Legionella bacteria in cooling towers. All the cooling towers were maintained to comply with the Australian Standard AS 3666:1989, including weekly water treatment chemical dosage. Median Legionella concentrations and ranges were calculated for each cooling tower. Assessment of the risk of Legionnaires' disease outbreaks based solely on counts of Legionella may be misleading. The factors contributing to successful control of the bacteria are well established, and to some extent generic, although individual systems are almost always unique in some aspect. Risk assessment of cooling towers for Legionella dissemination should be focused upon the established and quantifiable criteria. The role of Legionella culture in determining risk should be subordinate, due to the absence of a quantifiable relationship between culture results, system status, and dose response.

In this chapter, the authors report on an epidemiological study on the immune responses of residents in homes with heavy and with minor Legionella contaminations in their hot water systems. In one of the two residential areas with a central hot water system, Legionella contamination of up to 1 million CFU/liter had been known for several years. In the other residential area, decentralized hot water systems were installed in all houses so that no Legionella problem would be assumed. Using the criterion of antibody titer of 1:128 (immunofluorescence), 1.9% of the exposed and 1.1% of the control persons exhibited positive antibody titers against L. pneumophila serogroups 1 to 14, and 7.5% of the exposed and 6.5% of the control persons had positive antibody titers against Legionella species in total. In Eberswalde, a small city near Berlin, the prevalences of positive antibody titers in 246 healthy inhabitants were 10.2% in the exposed group and 13.4% in the control group. In the authors' investigation the participants exposed to heavy Legionella contamination in their hot water systems at home had higher antibody titers against Legionella than participants living in the control area.

A voluntary, laboratory-based reporting system of legionellosis was initiated in Sweden in 1981 and became compulsory in January 1996. According to the Swedish Act for Communicable Diseases, clinicians have to report all diagnosed cases of legionellosis since 1 July 1989. The incidence rate for reported cases of legionellosis during 1995 to 1999 ranged from 0.57 to 1.15 per 100,000 inhabitants. An estimation of sources of infection indicates that about one-third of the cases had a domestic, community-acquired infection and another third contracted legionellosis during travel abroad. The first recognized indigenous outbreak of legionellosis in Sweden occurred as early as 1979, when 58 persons contracted legionellosis at an indoor shopping center in a medium-sized Swedish town. The information is reported to the European Working Group for Legionella Infections (EWGLI) database. Five persons received treatment against legionellosis at a hospital after a weekend trip to Paris in 1988. In 1990, several Nordic tourists including 13 Swedes were reported to have contracted legionellosis during a visit to Majorca. At least 20 Swedish tourists had shown symptoms of legionellosis. Several persons in a group of bridge players, who went to Kusadasi in Turkey in 1996, also contracted legionellosis.

The incidence of Legionnaires' disease in Slovakia is low (0.2/million people), though underestimated; only a few sporadic cases have been documented. This chapter presents a study focused on an investigation of a potable water system for Legionella contamination and an evaluation of the risk of possible exposure, reflecting seroreactivity or even infection of nosocomial origin in patients of the hospital, where laboratory diagnostic tests for Legionella infection are not performed. The quantitative contamination of the water distribution system was 104 to 107 CFU/liter and was supported by the presence of legionellae in water from central and P3 tanks. Contamination was supported by legionellae surviving in the central tank, promoted by the nonadequate maintenance of the tank, especially in a temperature regimen (50 to 52№C) and hypochlorination. To assess the risk of exposure to contaminated tap water for inpatients, the authors did a serosurvey for antibodies in patients submitted to a normal treatment regimen from block A, colonized with legionellae in 20 to 45% from 1995 to 1999. A small group of patients suffering from pneumonia (more than 100) in the same hospital in which Legionella laboratory tests are not in routine use, was investigated in order to find another explanation for the absence of nosocomial legionellosis. There was a lack of convalescent-phase sera and samples for cultivation due to insufficient cooperation with clinicians and notably missing detection by Legionella urinary test. This test is widely recommended for active surveillance of legionellosis in colonized hospitals.

Legionellosis is a new infectious disease emerging in Poland although not yet registered. A Polish working group aimed at selecting laboratory methods for investigating human Legionella infections and detecting bacterium in the environment. The specimens from patients were examined for Legionella antigen in urine and for the level of antibodies to Legionella pneumophila serogroup 1 in patient sera. The standard commercial BIOTEST was used for the determination of Legionella antigen in urine. The procedure was evaluated according to the European program of the European Working Group for Legionella Infections for quality of testing of Legionella antigen determination in urine. Human serum samples of 180 patients with pneumonia and other pulmonary infectious diseases were collected and examined by microagglutination test. The antibodies to L. pneumophila serogroup 1 titer as high as 2,048 were found in only 2 of 69 sera from patients with atypical pneumonia: in one convalescent patient's serum and in serum of one nosocomial case with symptoms of pneumonia. The urine antigen of Legionella tests determined by the enzyme immunoassay method was also positive in both cases. The risk of L. pneumophila infection still seems to be low for the general population in Poland. It might be higher for some groups, including patients in rehabilitation and transplant units, and for staff in certain institutions such as hospitals.

A Legionnaires' disease outbreak was detected among visitors (50,000) and workers (830) of a commercial fair being held in Kapellen from 29 October to 7 November, 1999. An environmental investigation and a cohort study among stand workers were initially carried out to trace the source of the outbreak. A case-control study was launched in February 2000 to identify the source of the outbreak. All cases (n = 93) were persons attending the fair who developed symptoms of Legionnaires' disease within 2 weeks after their fair visit. Due to time (3 months after the outbreak) and logistical constraints, controls were selected from an available register from a nearby hospital. This register collected information about medical complaints of fair visitors who consulted the hospital, mostly to get reassured about their health status. Data were analyzed with the SPSS 9.0 statistical package. The analysis was repeated on confirmed cases to address the issue of potential misclassification of presumptive and possible cases. Additionally, the analysis was repeated on cases and controls who visited the fair once during the first long weekend (4 days) of the fair to account for a different risk of getting the disease related to the day of the fair visit. A total of 45 (48%) cases were confirmed, whereas 48 cases (52%) were presumptive or possible. The response rate for cases was 74.2% (69 of 93), and for controls was 46.3% (162 of 350). In Legionella outbreaks, timely epidemiological investigations are crucial, particularly to guide environmental investigations.

Ten cases of indigenously acquired Legionnaires' disease were identified by routine laboratory surveillance in Wales, United Kingdom, over a period of 7 months. Epidemiological investigations started with a case-control study. Case definitions used were those of the European Working Group for Legionella Infections; controls were taken from people who accompanied the three patients who had visited the hotel as part of a group. A cohort study was designed with staff who worked during the weekend that two patients had visited the hotel. The hypothesis was that more seroconversions would have occurred among people working close to the source, so that the average titer in this group would be higher than in people not exposed to the source. Six patients had a lung infiltrate on chest X ray. The first patient visited the hotel in July 1999, patients 2 and 3 visited in December 1999, patients 4 and 5 visited in January 2000, and patient 6 visited in February 2000. Only one stayed overnight and dined in the hotel, one used the swimming pool and spa for a week, three attended a lunch, and one attended an evening meal. Three of six patients were active smokers; none were immunosuppressed. Neither the case-control study nor the cohort study indicated a possible source of infection. Public Health officials and local authorities were notified of the possible role of ultrasonic humidifiers in the transmission of legionellae.

Water systems contaminated with legionellae cause outbreaks and sporadic cases of Legionella pneumonia. Legionellae are ubiquitous in water systems. In most clinical cases the source of infection can be confirmed only by using typing methods that discriminate between the patient and the environmental isolates. Water samples were analyzed using Legionella isolation standard method ISO 11731, completed by additional concentration with centrifugation (6,000 X g, 10 min). The legionella concentrations varied from 5 to 2,000 CFU/liter; the highest concentration was found in the tap water sample. The strains were L. pneumophila serogroup 1. The thermophilic heterotrophic bacteria concentration was highest (3.0 X 106 CFU/liter) in the tap water sample that contained the highest legionella counts. The mesophilic bacteria content was the highest in the legionella positive shower sample (5.9 X 107 CFU/liter). The L. pneumonia case of the 61-year-old man was detected only by the urinary antigen method. Therefore, conclusive evidence of the source of this particular infection could not be confirmed. Since the incubation time for legionella infection can vary from 2 to 10 days, it is possible that the infection was contracted elsewhere, such as the patient's home.

The population of Western Australia (WA) is concentrated mainly in the Perth metropolitan area and the coastal hinterland of the southwestern tip of the continent. Its climate is dry and warm and is described as Mediterranean. Yet Legionella infection in WA has been associated with exposure to garden products such as potting mix contaminated with L. longbeachae. The Legionella urinary antigen test is not available in WA. The majority of cases of Legionella infection are therefore diagnosed by serological means. Nucleic acid amplification methods for diagnosis of Legionella infection (a mip-based PCR protocol) were only introduced to WA in 1998 and have led to more diagnoses of L. pneumophila serogroup 1 infection than L. longbeachae infection since then, though the L. longbeachae type culture strain is negative by the mip-PCR. There was no evidence of temporal clustering of seropositive Legionella cases in WA from mid-1994 to mid-2000. During 1997, L. longbeachae was isolated from cooling tower water on three consecutive monthly samplings. Nevertheless, they suggest that the epidemiology of Legionella infection in WA is significantly different from what has been described elsewhere. The authors have been unable to find any laboratory-based evidence for the occurrence of time or space clusters of legionellosis cases in WA. Further studies into the epidemiology of Legionella infection in WA may generate useful insight into the environmental biology of this disease.

Legionella longbeachae was first isolated in 1980 from a patient with pneumonia in Long Beach, Calif. Legionellae were isolated from 11 of the 13 composted wood products samples and all 11 potting mix samples after amoebic enrichment. L. longbeachae was isolated from 4 and 5 samples, respectively. The most predominant species of Legionella obtained from potting soils was Legionella bozemanii serogroup 1, which was isolated from 13 samples. In contrast to L. longbeachae, L. bozemanii was isolated without enrichment from six samples. In the present survey, most of the Legionella species isolated from Japanese potting soils have been associated with disease in humans, an important finding with respect to public health. In Japan, various types of potting soils are used for gardening, such as composted wood products, potting mixes, peat moss, and peat moss-sand mixes. Japanese potting soils contain composted wood products produced from broadleaves such as oak and Japanese oak (sometimes cryptomeria and cypress), in contrast to pines or eucalyptus in Australia.

Domestic water supply, and particularly electric hot water systems, have been associated with Legionella colonization, and potential clinical implications have been suggested. In an effort to establish a link between sporadic legionellosis and hot water system colonization, patients hospitalized with documented sporadic community-acquired legionellosis (CAL) at Maisonneuve-Rosemont hospital in Montéral, Canada, were prospectively investigated. Only three culture-confirmed cases of CAL were established during the 30-month prospective epidemiological surveillance study. An epidemiological link between those patients' pneumonia and their contaminated home hot water tank was established in only one patient. The first water samples from this patient's home were obtained 45 days after the patient's admission to the hospital. Additional water samples from the hot water tank taken 14 weeks after the patient's admission showed a persistence in the system of the same isolate, underlying the long-term sustained colonization of contaminated hot water systems. Debilitation by chronic alcoholism and smoking, as well as recent plumbing repair and frequent exposures to aerosols generated from the shower head, likely played predominant physiopathological roles in the patient's acquisition of Legionnaires' disease. A study by the authors focused on severe CAL that required hospitalization and found a very low incidence linked to contaminated hot water tanks. Individual risk factors (i.e., smoking, age, chronic lung diseases) and immunoincompetence rather than environmental factors likely represent the major contributing factors in the acquisition of severe sporadic Legionnaires' disease.

Following the discovery of Legionella pneumophila as the etiological agent of Legionnaires' disease, many different serogroups and related species of this bacterium have been detected. A newly developed PCR method was used to generate strain-specific DNA fingerprinting profiles. This method is compared with the results of typing the same isolates with a set of monoclonal antibodies (MAbs). To identify different strains of Legionella, the authors used different primers to amplify DNA fragments in crude bacterial lysates to generate banding profiles. The complete Dresden MAb panel contains 98 MAbs related to lipopolysaccharide characteristics. Testing of the isolates was made with these MAbs utilized in the enzyme-linked immunosorbent assay method. The authors conclude that a combination of Göttingen direct fluorescent antibody technique and PCR methods can be a useful tool for subtyping L. pneumophila serogroup 5 strains.

The elevated water temperatures and large wetted surface area in cooling water systems present ideal conditions for microbial growth. To reduce contamination in cooling water systems, attention must be paid to equipment design, installation, operation, and maintenance. The aim is to minimize microbial multiplication in these environments, to ensure water treatment is adequate, and to minimize the production and release of aerosols. The approach taken in Australia has been to produce a prescriptive standard, which has been incorporated into uniform building regulations and most of the State Health acts and regulations within Australia. Both chemical and physical monitoring of the water-treatment program are essential for system life and performance, as well as for controlling microorganisms such as Legionella spp. In case study 1, an outbreak of Legionnaires' disease was associated with a hotel. Legionella pneumophila serogroup 1 count was 2.8 X 107 CFU/ml. In case study 2, several cases of Legionnaires' disease occurred at an industrial site at which there were both large process cooling towers (L. pneumophila serogroup 1 count, 1,000 CFU/ml) and small comfort air-conditioning towers (not sampled for legionellae). At both locations, monitoring programs, including water sampling for legionellae, were implemented. No further clinical cases nor significant Legionella spp. detections have been experienced. These case studies are examples only of the increasing awareness of Legionella as an environmental pathogen that has led to control and monitoring strategies on a wide scale in Australia; sampling for legionellae may now be important as a defense under common law.

American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE's) Environmental Health Committee recognized the need for a comprehensive, nationally accepted guideline to address the public health issues surrounding Legionnaires' disease. The primary focus is on control and prevention of contaminated water sources, but it is limited to discussion of cooling towers and domestic water systems. Guideline Project Committee (GPC)-12 was composed of members with diverse backgrounds ranging from physicians to consulting engineers and microbiologists to manufacturers. Investigation of Legionnaires' disease outbreaks suggested that, in most instances, transmission to humans occurred when water containing the organism was aerosolized into droplets less than 5 μm in diameter, allowing inhalation into the lungs of a susceptible host. Conventional open cooling towers are evaporative heat-transfer devices in which atmospheric air mixes with and cools warm water by evaporating a portion of the water. Heated spas generally operate in the temperature range of 32 to 40№ C, which is in the range that favors amplification. ASHRAE Guideline 12 presents information on several budding water systems that can amplify and disseminate Legionella. This chapter thus focuses on three of these amplifiers: cooling towers, evaporative condensers, and heated spas. Specific guidance is presented to minimize the risk of legionellosis associated with these devices.

In Germany, several standards and regulations have been published with the objective of preventing cases and outbreaks of legionellosis; these standards and regulations need to be considered for the planning, operation, and maintenance of technical water systems. The following are environmental factors that generally favor the multiplication of Legionella: Inorganic deposits in heating elements, distribution beams, pipes, and fittings. These deposits have large surfaces, which can become colonized by microorganisms; the use of materials such as rubber or silicon (e.g., in washers, membrane expansion vessels, shower hoses), which quickly can become colonized by microorganisms generating a biofilm; and stagnant water in parts of the installation with inadequate or absent circulation (e.g., dead ends), in which Legionella can reach high concentrations. The filters in water treatment units for swimming and hot whirlpools are the most critical areas for colonization with Legionella. Usually, first the filters are colonized, and a detectable contamination of the pool water occurs only after Legionella grows through the filter material. Water samples can be obtained directly from the cooling water reservoirs of evaporation condensers and cooling towers. A section provides an overview of the objectives and reference values for testing for Legionella in hot water systems, swimming pools and hot whirlpools, and air-conditioning systems, according to German regulations and recommendations. In Germany, most water samples tested for Legionella are relatively unpolluted and usually have only little sediment (e.g., drinking or recreational water, humidifier water).

Much of what we know about the epidemiology and prevention of Legionnaires' disease (LD) comes from outbreak investigations and special studies. In this chapter the author focuses on three areas where one's knowledge is expanding and where new prevention opportunities have emerged: surveillance for travel-related legionellosis, the use of LD diagnostic tests, and new directions in water disinfection. Detection of travel-related outbreaks offers the potential for providing new prevention opportunities. The European Working Group for Legionella Infections conducts surveillance for travel-related LD in Europe. This system has been quite successful at identifying clusters of travel-associated LD. Several studies indicate that the ability to identify outbreaks of LD is inadequate and that the adoption of new diagnostic tests may be further eroding our capacity to identify certain types of Legionella infections. To prevent LD in hospitals, Centers for Disease Control and Prevention (CDC) recommends a strategy based on proper maintenance of water systems, universal testing of appropriate patients, and thorough investigations in situations where disease transmission occurs. The shift in diagnostic approach also has imphcations for the ability of the public health community to respond to outbreaks of LD. Monochloramine is currently used by nearly 25% of municipalities as a residual disinfectant in their public water distribution system. In addition to epidemiologic studies supporting the use of monochloramine, laboratory data demonstrate increased efficacy of monochloramine compared with chlorine in eliminating biofilm-associated Legionella.

The Queen's Medical Centre (QMC) is a large teaching hospital housing approximately 1,400 beds. Each block has its own independent circulating hot water system (HWS). The West Block and Medical School have storage calorifiers (water heaters) equipped with antistratification pumps, and the water is circulated at 60 to 62№C. The circulation of hot water at 60№C so that it is delivered to every outlet at 50№C within 1 min of opening is the recommended strategy for the control of Legionella in the United Kingdom. A number of problems are associated with the installation of thermostatic mixer valves (TMVs). They are expensive to purchase; require much maintenance, particularly if the water is hard; and the failure of the nonreturn valves fitted within them can result in cold water passing into the hot or vice versa, depending on the relative pressure differences. At the end of March 2000, the cold water supply to the ward was equipped with a chlorine dioxide dosing system, which ensures that a chlorine dioxide residual of at least 0.5 mg/liter reaches beyond the TMV to the shower hose and head. To date, this does not appear to have significantly reduced the legionellae. It may be that legionellae can only be maintained at undetectable levels by routinely removing and disinfecting the TMVs and shower components. To enable this, the units would have to be designed for quick release so that they can be rapidly recycled.

Aquatic bacteria of the genus Legionella are causative agents of severe nosocomial pneumonia with high case-fatality rates. There are different methods for the control and eradication of legionellae in water systems. Focal modalities include UV light, instantaneous heating, and ozone. An investigation took place in an internal hospital department with 375 beds (11,331 patients in 1999) with a separate copper pipe water system. This investigation determined whether maintaining a hot water temperature in water heaters at >50№C (122№F) with an additional central flowthrough UV sterilizer (power 85 W, 5 m3/h) was as efficient as maintaining a hot water temperature in water heaters at >60№C to control legionellae in a hospital water system. In all the rooms A, B, and C, the differences of legionellae counts between a hot water temperature in the water tank at >60ºC and a temperature of >50№C with UV light were significant. The differences in temperature were significant in room C and outlet D. There was no significant difference in legionella count in the probes of location D. The water sampled closest to the sterilizer (outlet D) has the same results with both types of water preparation in spite of different water temperatures at the outlet. That means a hot water temperature of >50№C with UV sterilization kills legionellae in the incoming water as efficiently as a water temperature of >60№C.

This chapter aims at determining the susceptibility of biofilm-associated Legionella pneumophila to free chlorine and monochloramine. A biofilm reactor was developed to grow biofilms containing L. pneumophila on 24 replicate stainless-steel surfaces in potable water. The ability of this reactor to produce reproducible biofilms was validated by the fact that the standard deviations of the base biofilm densities on stainless-steel coupons (n = 3) ranged from 0.06 to 0.18 log CFU per coupon. When Legionella containing biofilms were exposed to the same dosages of monochloramine for identical contact periods, the treatments were significantly more effective. In summary, the authors have shown that biofilm-associated L. pneumophila are significantly less susceptible to chlorine than are planktonic L. pneumophila, while susceptibility of planktonic and biofilm-associated L. pneumophila to monochloramine are similar. When monochloramine and free chlorine were compared under identical conditions, monochloramine was significantly more effective, indicating that monochloramine may be an effective disinfectant for the inactivation of L. pneumophila within potable water distribution systems. Further research using open system biofilm reactors and model distribution systems is needed to determine the utility of monochloramine as a disinfectant against biofilm-associated L. pneumophila.

Legionella and Legionella-like organisms live as facultative intracellular parasites of amoebae in the biofilm that covers the inside of tanks and pipes in water systems. Recent investigations, discussed in this chapter, indicate that it is possible to prevent 90% of drinking-water-associated Legionnaires' disease (community acquired as well as nosocomial; sporadic disease as well as outbreaks) through the use of monochloramine for residual municipal water disinfection. Monochloramine (combined chlorine) is formed when ammonia and free chlorine are mixed in water in the correct ratio. Recent research has shown that monochloramine may be considerably more effective against Legionella and Legionella-like organisms than free chlorine. An interesting hypothesis is that monochloramine may also be better at killing the amoebae that serve as hosts to Legionella and Legionella-like organisms. To this author's knowledge, monochloramine is the only disinfectant that has been shown to reduce the incidence of Legionnaires' disease when used at the municipal level. This finding is not just important for control of nosocomial infections. More than 75% of cases of Legionnaires' disease acquire their infection in the community. Residual disinfection of municipal drinking water with monochloramine seems an inexpensive and efficient way to prevent many of these cases. In addition, supplemental injection of monochloramine may become one of the preferred methods for control of Legionella and Legionella-like organisms in hospitals, hotels, and similar institutions.

This chapter evaluates the efficacy of copper-silver ionization that is reported to be a useful tool to control growth of Legionella. Copper-silver ionization was tested over a period of 1 year as the first step in a multiple-barrier system controlling Legionella. In Germany, additions of silver and copper to drinking water need special approval of the Ministry of Health and are permitted only with substantial restrictions. Measurable effects in Legionella reduction can be achieved only at silver concentrations significantly exceeding the German standard of 10 μg/liter. Copper values exceeded national standards, preventing a short-term high dosing of silver. Concentrations of both metals vary depending on water flow. Silver cannot be analyzed directly with in the system, causing a considerable feedback delay. Legionella counts stayed above targeted values (working standard, Legionella counts <1 CFU/ml) . Reduction is not as efficacious as in hot water systems maintaining about 60№C. Long-term effects (e.g., development of heavy-metal resistance) of adding silver and copper are yet to be evaluated.

Travelling and staying in hotels are a known risk factor of infections by Legionella. In Europe, it is known that more than 20% of all declared cases of legionellosis are associated with travelling. In Spain, a country that receives 50 million tourists every year and which has one of the most important tourism industries in the world, the travel-associated Legionnaires' disease (TALD) cases have followed a parallel evolution. The Hotel Inspection Programme, that was in operation for 1 year (June 1999 to July 2000), adequately accomplished the task of providing hygiene-health training of the management and maintenance staff of the hotels involved. Collaterally, it acted as a means of detecting certain generalized hygienic deficiencies in the water and airconditioning systems of the hotels, and is likely to have had a positive influence on their being remedied. The data derived from the evaluations that were made show that the programme was very well received by those taking part and that, consequently, this experience may be used to advantage in other tourism areas in the Mediterranean. Although it is early to venture conclusions on the epidemiological impact of the programme, the data available to us from the first year of its application are very encouraging and suggest that the educational factor may play a key role in the prevention of legionellosis in the tourist sector.