Legionella: State of the Art 30 Years after Its Recognition

Legionnaires’ disease was thought to be an atypical and easily recognized form of pneumonia for several years after the 1976 Philadelphia epidemic. Astute clinicians began to question the characterization of the illness as clinically unique, and in a seminal prospective study of community-acquired and nosocomial pneumonia, researchers showed that Legionnaire's disease could not be distinguished on presentation from non-Legionnaires’ disease cases of pneumonia. This study was followed by several others showing that neither community-acquired nor nosocomial Legionnaire's disease could be accurately distinguished on presentation from other common pneumonias on clinical, laboratory, and roentgenographic grounds. Each study found one or two differences between Legionnaire's disease and other pneumonias, often differing from one study to another. Another recent approach that has been taken to detect Legionnaires’ disease is the analysis of blood acute phase reactant levels. Researchers measured blood procalcitonin and neopterin levels in pneumonia patients with and without Legionnaires’ disease, which showed that patients with pneumococcal pneumonia had higher levels of procalcitonin and lower levels of neopterin than did Legionnaires’ disease patients and that procalcitonin levels increased with greater disease severity regardless of diagnosis. Several types of probably noninfectious extrapulmonary disease have also been recognized since the 1976 Legionnaires’ disease epidemic.

Legionnaires’ disease (LD) treatment recommendations are supported by data obtained from in vitro and cellular studies, experimental studies with animal models, and observational studies, some of which come from prospective clinical studies of community-acquired pneumonia (CAP). There is some controversy about the use of new fluoroquinolones versus macrolides as the treatment of choice for LD. The least inflammation is found in the animal model with azithromycin, while the most is observed with erythromycin, with the quinolones being intermediate. Extrapulmonary manifestations of legionellosis, which are rare but sometimes observed in the immunocompromized host, may indicate significant therapeutic connotations. Mixed infections in legionellosis should be kept in mind concerning the inmunocompromized population since there are many reports of death when clinicians fail to identify and treat the dual component of infection. Some international guidelines recommend combined therapy for severe episodes but no consistent evidence supports this suggestion. For most patients monotherapy with a macrolide or a selected fluoroquinolone usually leads to a more cost-effective outcome. Possible toxicities of adding more than one antibiotic may be a concern, particularly in the intensive care unit setting. In vitro and in vivo evidence shows a highest potency of rifampin against Legionella, although the theoretical possibility of rapid induction of rifampin-resistant strains when it is administered alone precludes its use as sole therapy. The only variable that remained statistically significant on multivariate logistic regression analysis was the APACHE score (OR 1.86) at intensive care unit admission.

Microbiological diagnostic methods play the key role in establishing the etiological diagnosis since the clinical presentation is not specific for Legionella infections. Currently available are culture, urinary antigen detection, direct fluorescent antigen testing, detection of nucleic acid, and serology. Culture is still the gold standard among the diagnostic methods for Legionella infections. The sensitivity of culture for the diagnosis of Legionnaire's disease has been estimated to be in the range of 11 to 65% in retrospective studies usually performed in reference laboratories. The most important technique for the identification of legionellae in the clinical laboratory is the serological characterization of isolated strain. The antigen excreted with urine has been characterized as heat-stable and resistant to enzymatic cleavage with a molecular weight of about 10 kDa. The advantages of urinary antigen detection are obvious: specimens are easy to obtain and can be investigated repeatedly; antigenuria is detectable very early and, therefore, often gives the first evidence for Legionella infection. Direct fluorescent antibody (DFA) testing of respiratory specimens is a rapid method for the detection of Legionella antigen. The excretion of DNA fragments in the urine is described for several bacterial pathogens, suggesting the suitability of urine PCR for the detection of Legionella DNA. The indirect fluorescent antibody (IFA) test is the only method for antibody detection that has been evaluated and standardized. None of the diagnostic tests presently available offers the desired quality with respect to sensitivity and specificity. Therefore, the standard technique is to use several diagnostic tests in parallel.

This chapter compares the individual risk factors, clinical features, and mortality of Legionella pneumonia in hospital and community scenarios. A comparative study was performed from 1983 to 2005, with 425 cases of pneumonia due to Legionella pneumophila being diagnosed during this time. The patients were divided into two groups based on the site of acquisition. Group 1 included 197 patients with community-acquired Legionella pneumonia (CALP), and group 2 included 123 patients with hospital-acquired Legionella pneumonia (HALP). Respiratory symptoms, gastrointestinal, neurological manifestations, and laboratory abnormalities prevailed in the community setting probably for two reasons: first, the CALP patients entered the hospital late because of the lack of specificity in initial symptomatology, allowing extrapulmonary symptoms to develop. Second, in hospitals with endemic legionellosis, the diagnosis and the onset of treatment for nosocomial Legionella pneumonia is immediate because of the high level of suspicion. This makes the extrapulmonary manifestations of Legionella infection less likely to be observed. The best way to prevent HALP is by culturing the hospital water supply, and, if colonized, Legionella tests should be used in any case of hospital-acquired pneumonia.

This chapter determines variables related to mortality of Legionnaire's disease (LD). An observational, comparative study was performed from 1983 to 2005, with 408 cases of Legionella pneumonia being diagnosed during this time. The variables studied included demographic characteristics, individual and aspiration risk factors, clinical manifestations, radiological manifestations, diagnostic data, treatment, and outcome. The variables were analyzed using univariate and multivariate logistic regression analysis by SPSS version 12.0. Regarding demographic data and risk factors, patients who died had a higher frequency of underlying diseases than the control group (89.8 versus 66.9%): mainly chronic respiratory disease (42.9 versus 29.6%), chronic heart failure (22.4 versus 10.2%), neoplasm (38.8 versus 15.8%), neurological sequelae (20 versus 7.6%), and inmunosuppressive therapy-mainly corticoids (45.8 versus 14.6%). Despite the improvement in the diagnostic procedures of LD and the use of more efficacious antibiotics against Legionella, chronic heart failure, hematological cancer, and corticotherapy are still bad prognostic factors of LD. Recommendations for prevention of LD should focus on settings in which there are persons at greatest risk for illness or serious outcome.

Legionella pneumophila has been recognized as an important cause of community-and hospital-acquired pneumonia in both healthy and immunosuppressed patients since its first description in 1976. The authors have systematically used urinary antigen testing as a specific diagnostic test for Legionella infection since 1994. It should also be pointed out that the reporting of Legionnaires’ disease (LD) has been mandatory in Catalonia since 1987. This chapter describes demographic data; clinical, analytical, and radiological characteristics; diagnostic methods; and therapeutic data of a series of community- and hospital-acquired cases of LD collected from 1983 to 2005. It determines trends in demographic and clinical data, individual risk factors, diagnostic methods, and outcome over time. Mortality by nosocomial and community-acquired LD combined has almost significantly decreased over time, which may be due to the earlier diagnosis and changes in treatment.

Bacterial pneumonia continues to be the most frequent pulmonary complication in human immunodeficiency virus (HIV)-infected patients, although its prevalence has decreased with the use of highly active antiretroviral therapy. The most common etiologic agent of community-acquired pneumonia (CAP) in infected and uninfected patients is Streptococcus pneumoniae. Legionella is at present one of the leading causes of CAP in the general population. This chapter compares the epidemiological data, clinical features, outcome, and mortality of pneumonia by S. pneumoniae and by Legionella spp. in HIV-infected patients. An observational, comparative study was performed in 15 HIV patients with CAP by Legionella (group 1) and 46 by S. pneumoniae (group 2). No statistically significant differences were observed between the two groups in either the use of an appropriate antibiotic treatment in the emergency department or in the delay from the presentation of the symptoms until the initiation of an appropriate antibiotic therapy. Legionella pneumonia (LP) should be considered as an opportunistic infection in HIV-infected patients, and the Legionella urinary antigen test should be mandatory whenever we are faced with CAP in an HIV-positive patient.

In Denmark (5,400,000 inhabitants), Legionella infection is a notifiable disease. Two departments at Statens Serum Institut (SSI) in Copenhagen are involved in collecting data about these cases, since (i) notifications based on clinical and local laboratory data are sent to the Department of Epidemiology; (ii) all cultured Legionella patient isolates are referred to the Department of Bacteriology, Mycology, and Parasitology (ABMP) from local departments of clinical microbiology for confirmatory identification and serotyping; and (iii) water samples from environmental investigations are sent to AMBP for culture and DNA typing. Every local clinical microbiology department in Denmark has its own infectious disease control unit, which is responsible for the surveillance and investigation of nosocomial infections. The diagnosis of Legionella infection was made by the detection of Legionella species in respiratory specimens by PCR, a positive urinary antigen test, or a significant rise in antibody titre. Diagnostic results and clinical data, including onset of disease and location of the patients, were retrieved from the laboratory information system at DCM Herlev. During the period of 2000 to 2004, several efforts were made to combat nosocomial Legionella infection, such as increasing the hot water temperature and replacing hot water tanks with heat exchangers.

The literature on Legionnaires’ disease (LD) widely cites 2 to 10 days (less frequently to 14 days) as the incubation period for the disease. This chapter summarizes the findings of reports with data adequate to examine the potential for incubation periods longer than 10 to 14 days. These findings suggest that while the majority of cases will develop by 14 days, 19 days may be the upper end of the incubation period. The scientific literature was searched, primarily via PubMed and Scirus, for reports on outbreaks of LD. The recovered reports were reviewed for information on the incubation period considered for case inclusion and for data on the actual periods between exposure and symptom onset for the cases. Investigators of outbreaks with well-defined sources and exposure windows should consider a longer case inclusion period from 14 to perhaps as long as 21 days to ensure good data capture and to help better define the latency period between exposure and symptoms for LD.

Fluoroquinolones are the most active drugs against Legionella in intracellular and animal models, achieving high intracellular levels and low MICs. In patients with human immunodeficiency virus infection, Legionnaires' disease (LD) has a more severe clinical presentation and worse evolution, with a mortality rate of up to 20%. This chapter presents a case of a severely inmunosuppressed woman with LD in whom treatment failed with levofloxacin and who recovered after the addition of azithromycin. The intracellular location of the pathogen is relevant to the efficacy of the antibiotic. At present, new macrolides such as clarithromycin and azithromycin and fluoroquinolones are the most active drugs in the treatment of Legionella pneumonia. Even though the advances in rapid specific diagnosis and the greater efficacy of the new macrolides and quinolones have led to a better prognosis of Legionella pneumonia, mortality continues to be very high in immunosuppressed patients. Thus, randomized trials comparing the efficacy of combination therapies with new macrolides and quinolones versus monotherapy should be performed to evaluate their real impact in reducing morbidity and mortality of Legionella pneumonia with bad prognostic factors.

Legionella pneumophila infections may progress among patients rapidly. Optimal results in the treatment of Legionnaire's disease may only be achieved with prompt effective antimicrobial therapy. This chapter presents a study in which in vitro activities of some antibiotics were assessed against L. pneumophila isolated from water samples. It has been determined that L. pneumophila strains obtained from the cooling tower are sensitive to all antibiotics tested. On the other hand, it has been found that 8% of tested L. pneumophila strains isolated from the cold water tank are resistant to azithromycin. Other findings have shown that 8% of strains isolated from the hot water tank are resistant to erythromycin, azithromycin, and doxycycline; 6% of strains isolated from hot water faucets are resistant to azithromycin; and 8% of strains isolated from hot water shower heads are resistant to erythromycin, azithromycin, ciprofloxacin, and doxycycline. According to these findings, it is clear that tested strains which have been found resistant to antibiotics have been isolated mostly from hot water shower heads. Consequently, this study has shown that ofloxacin and levofloxacin can be preferable antibiotics in the treatment of L. pneumophila infections where bactericidal activity is needed.

Diagnosis of Legionnaires’ disease (LD) in patients with pneumonia is based on phenotypic (culture, serologic testing, antigen detection in urine) and genotypic PCR methods. This chapter assesses the sensitivity and specificity of Legionella DNA detection using PCR on serum samples from patients with proven LD and patients with infections other than Legionella. Serum samples from two healthy volunteers were included as negative controls after every four samples. One patient was admitted to the emergency department of our hospital with proven LD caused by Legionella pneumophila serogroup 1 (urinary antigen test positive, culture positive, and sputum PCR positive). From this patient, consecutive serum samples were collected for Legionella-specific, real-time PCR. The results of PCR showed an increase in Ct values (corresponding with a logarithmic decrease of bacterial DNA) in the course of time and were found to mirror the clinical condition of the patient and c-reactive protein values during the acute stage of infection. Serologic methods to diagnose L. pneumophila infections are highly sensitive, but their utility is generally limited to epidemiologic studies due to the time lag needed to detect seroconversion. The advantages of PCR for the detection of L. pneumophila DNA in serum are evident; serum samples are readily obtainable and can be processed within a working day.

Despite extensive efforts, the etiology of about 40 to 60% of community-acquired pneumonia remains unclear; a number of these cases are attributed to Legionella. The incidence in this group may be as high as 38%. Cultivation is still supposed to be the gold standard, but there is doubt about the sensitivity in clinical practice. Researchers established a new method for the detection of Legionella by using PCR to amplify genomic ribosomal DNA and a reverse dot blot for the detection of L. pneumophila and non-pneumophila species. The authors used a multiplex-PCR and the coamplification of human DNA (pyruvate dehydrogenase gene) as an internal control to avoid false-positive results. To avoid false-positive results, incubation with uracil-N-glycosylase was used, and for avoidance of false-negative results the coamplification of the pyruvate dehydrogenase gene was used, creating an amplificate of 185 bp in length in each PCR. The authors used the established method to investigate clinical specimens of 93 patients taken out of a prospective randomized study about the epidemiology of community-acquired pneumonia in Berlin, Germany, from 1991 to 1992. They were able to demonstrate the ability to detect Legionella using PCR in conjunction with reverse dot blotting. This method allows the enhanced detection of non-pneumophila species in addition to established methods. Interpretation of the results for the incidence of non-pneumophila infections remains controversial because of a lack of facts about pathogenity of this group and comparable results in other studies.

Diagnosis of acute Legionella pneumophila in patients with community-acquired pneumonia can be time-consuming and difficult to confirm using current diagnostic tests. A research prototype assay using Gen-Probe's proprietary transcription-mediated amplification (TMA) technology was evaluated for detection of L. pneumophila in throat swabs from patients with documented community-acquired pneumonia. The TMA L. pneumophila assay procedure includes target capture, TMA, and the hybridization protection assay. Amplified methods are generally more sensitive than other laboratory methods, particularly for fastidious organisms such as Legionella. Therefore, in addition to the initial TMA L. pneumophila assay that targets a specific region of L. pneumophila rRNA, another target was utilized in an alternate region of L. pneumophila rRNA as a confirmatory TMA-based assay.

This chapter talks about a study in which a PCR-enzyme-linked immunosorbent assay (ELISA) was devised for the detection of Legionella and Legionella pneumophila–specific DNA in human respiratory secretions. The PCR-ELISA incorporates a solid-phase probe hybridization step to verify the PCR reaction. Briefly, DNA was extracted from a total of 343 routine respiratory secretions using Nucleospin tissue columns. The PCR-ELISA was performed using reagents from the Roche Diagnostic PCR-ELISA (Dig Detection) kit Cat No. 1636111 following the manufacturers’ instructions. A total of 343 respiratory samples were tested, of which 21 patients were identified as cases of legionellosis by fulfilling one or more of the criteria for a definitive case. With PCR-based assays becoming more sensitive and specific, perhaps the distinction of a definitive case should incorporate more molecular techniques. The 16S RNA PCR-ELISA has the same sensitivity as the 5S RNA PCR/SB technique but can also identify L. pneumophila and is more sensitive than either direct fluorescent antibody (DFA) or culture. The lack of suitable samples, i.e., respiratory secretions, should not inhibit investigation, as Legionella specific DNA has been found in serum and urine.

The genus Legionella comprises 48 species and 70 different serogroups, and approximately half of them have been associated with human disease. Serological methods have been widely used for species and serogroup identification, but the progressive characterization of new species has established that antigen cross-reactivity limits specificity and restricts their use to a few frequently isolated species. Serological and sequence-based methods were used by the authors in a study for comparing Legionella identification with 93 Legionella strains including clinical and environmental ones, L. pneumophila, and other species. The aims of the study were to determine the correlation between serological and sequence-based methods and to establish a better strategy to increase the number of species of Legionella that can be identified. Comparing serological methods, 90 of 93 (96.7%) strains presented agreeing results with the three methods used. Comparing gene sequencing methods, 85 of 93 (96.7%) strains presented agreeing results with both genes. The percentage of similarity for species allocation was 99 to 100% with the mip gene and 97 to 100% with the 16S rRNA gene. Serological methods have been the most frequently used for Legionella identification, but the number of species that could be identified is limited. Sequencing methods allow the identification of any strain, detecting all species described and new ones. In this study all strains were identified with both genes, but disagreements were detected when both gene results were compared.

An outbreak of Legionnaires' disease (LD) occurred in August 2004 in Sweden in the town of Lidköping, by Lake Vänern. Two genotypes of Legionella pneumophila serogroup (sg) 1 were found by culture in patient samples, and one belonging to subgroup Benidorm (genotype A) and the other to subgroup Bellingham (genotype B). The indirect immunofluorescent antibody technique (IFAT) is still a standard method for antibody assay in patients with legionellosis, allowing for the screening of antibodies against several species and subtypes of Legionella. A serological study was therefore undertaken to assess the impact of the legionella outbreak on all pneumonia cases that occurred in the town. Patient samples were also tested for antibodies against other atypical pneumonia agents, i.e. Mycoplasma pneumoniae and Chlamydophila pneumoniae. Local experience in the laboratory using more than one subgroup of antigen had been shown earlier to increase sensitivity. Optimal sensitivity of immunofluorescent antibody tests in serological diagnosis of L. pneumphila sg 1 infections requires the use of an antigen subgroup that is in agreement with the antigenic setup of the epidemic strain.

The end of the 1980s was the “golden age” of Legionella pneumophila serotyping. In recent years molecular genotyping methods have been increasingly applied to the classification of legionellae. This chapter describes a short typing system for L. pneumophila that might be useful in epidemiological investigations. Serotyping of L. pneumophila is mostly done by using indirect immunofluorescence tests. However, in a European multicenter study for typing of L. pneumophila, it was demonstrated that this technique produces variable results, especially for the recognition of serogroup-cross-reactive epitopes by reagents LpH and LpK. These problems can be avoided by using an enzyme-linked immunosorbent assay (ELISA) that is a more reproducible technique because of reading of positive results with respect to the blank value, whereas microscopically judging is sometimes equivocal and influenced by the laboratory staff. ELISA allows the screening of many colonies in less than six hours. In addition, the selection of only 11 reagents combined with a blank without monoclonal antibodies (MAbs) as the 12th value was done to adapt the typing scheme to the microplate array.

Several molecular biological methods are increasingly being used for the identification of legionellae in man-made aquatic environments. Currently, the genus Legionella is known to include 51 species. Some of the species have been isolated only from environmental sources up to now, but it is generally accepted that all species may cause pneumonia, especially in immunocompromised persons. These huge numbers of Legionella spp. represent a very wide serological heterogeneity that can lead to unsatisfying sensitivity and specificity of serological identification tools. The specificity of Duopath Legionella was calculated by testing 50 bacterial strains isolated from water samples and grown on GVPC agar plates. The latex agglutination assay Legionella spp. recognizes seven of the most frequent Legionella non-pneumophila species causing Legionnaires’ disease but not the wide range of legionellae found in water systems, which are also suspected to be pneumonia pathogens. In many countries, water or environmental samples have to be analyzed for the presence of Legionella spp. instead of only Legionella pneumophila. Here, Duopath Legionella revealed a significant advantage over the latex assay and makes the phenotypic diagnostic gap significantly smaller. Therefore, Duopath Legionella can be considered a user-friendly, simple and reliable test for the simultaneous identification of L. pneumophila and non-pneumophila.

Legionella is an important cause of both community-acquired and nosocomial pneumonias. In this chapter, the authors produced 17 monoclonal antibodies (mAbs) against the recombinant PAL (rPAL) cloned from Legionella pneumophila serogroup (sg) 1 and investigated antigenic diversity of the peptidoglycan-associated lipoprotein (PAL) proteins from soluble fractions of 16 Legionella species including 22 serogroups by mAb reactivity patterns in ELISA. Soluble antigens from Legionella strains (L. pneumophila sg 1, 3, 4, 5, and 6; L. annisa; L. dumoffii; L. gormanii; L. jordanis; L. micdadei; L. oakridgensis; L. sainthelensi; L. bozemanii sg 1 and 2; L. longbeachae sg 1 and 2; L. Hackeliae sg 1) were prepared by the method of Berdal et al. The authors concluded that antigenic diversity of a Legionella species–common PAL protein is present among Legionella species and that developing diagnostic agents using a Legionella antigen, even if it is conserved, might be better for separately detecting L. pneumophila species and non-pneumophila Legionella species.

In a study on Legionella species, the authors developed a new immunochromatographic test (ICT), the SD Bioline Legionella antigen test (Standard Diagnostics, Inc., Korea), which uses antibodies specific to the peptidoglycan-associated lipoprotein antigen, and evaluated the performance of the test kit by comparing it with the Binax NOW Legionella urinary antigen test (Binax, Portland, Maine) and Biotest Legionella urine antigen EIA (Biotest AG,Dreieich, Germany). The authors tested urine samples from 99 adult patients with pneumonia. Results of 3 Legionella urinary antigen tests in 11 patients with Legionnaires' disease are provided in this chapter. Comparison of test performance of 3 Legionella urinary antigen assays was performed. Detection of Legionella soluble antigens among those artificially inoculated in normal urine samples are also provided.

The research latex product was done to identify Legionella pneumophila, Legionella anisa, and Legionella taurinensis. The comparison with two commercialized latex reagents was made using 190 Legionella wild-type strains (96 L. pneumophila, 63 L. anisa, 31 L. taurinensis); they corresponded to the French clinical and environmental distribution of the isolates. The three latex kits tested efficiently identified all the L. pneumophila isolates within 15 min. Only the research latex product is able to identify L. anisa and L. taurinensis. The research latex product is a rapid and easy tool to identify the majority of the clinical and environmental Legionella isolates. Latex agglutination is a rapid technique to identify Legionella isolates. Therefore, the authors developed latex reagents which allow the identification of the majority of the clinical and environmental Legionella in 15 min. The research latex product is a reliable and rapid tool to identify the most frequent Legionella in clinical and water samples.

The main diagnostic methods for Legionnaires’ disease (LD) are culture, urinary antigen detection, serology, and PCR, all of which are costly to perform and variably subject to false-positive results in low-prevalence populations. There was little support for no testing, even when empiric therapy for LD is given, as testing was regarded as important for epidemiologic and clinical reasons. There is now a considerable amount of data on nucleic acid amplification tests (NAATs), particularly PCR, for diagnosing LD. The need for rigorous assessment of nonrespiratory sample types was emphasized. The choice of empiric antibiotic therapy for CAP depends on several important considerations. Various ancillary treatments have been proposed for LD, including corticosteroids, oxygen, cytokine inhibitors, activated protein C, and glucose control. For some of these treatments there is evidence indicating benefits for treating severe infection/sepsis, although data are lacking for the specific treatment of LD. Corticosteroid therapy may be indicated for adult respiratory distress syndrome caused by LD or bronchiolitis obliterans organizing pneumonia, but this has not been systematically studied. Many who participated in a panel discussion on clinical aspects of LD, agreed that ancillary therapies shown to be beneficial for other types of pneumonia would likely be applicable to the treatment of LD.

The European surveillance scheme for travel-associated Legionnaires’ disease (EWGLINET) supports ongoing microbiological research and development for improved laboratory detection of clinical and environmental sources of infection using standardized methods within European countries. Adherence to the EWGLI guidelines, introduced in 2002, for control and prevention of travel-associated cases of Legionnaires’ disease is leading to greater protection of travelers in all European countries. This review covers EWGLINET’s role in the collection of national legionella data that enables trends over time at the aggregate level to be studied as well as historical trends within and between European countries. It covers the 10-year period 1995 to 2004 and mainly focuses on trends in aggregate data. The European data set from contributing countries between 1995 and 2004 is composed of 27,244 cases of Legionnaires’ disease. Community-acquired cases comprise the largest proportion of cases in Europe, at 38% overall, but vary from year to year depending on the number and size of reported outbreaks. This short review of surveillance data on cases of Legionnaires’ disease in Europe has mainly focused on descriptive aggregated data for the period 1995 to 2004. Major differences between countries in number of cases, category of exposure, number and range of outbreaks, and methods of diagnosis can only be highlighted rather than examined in depth. All countries with data in this review acknowledge the importance of international collaborations since these also benefit microbiological and epidemiological developments at the national level.

The main reason for typing Legionella pneumophila is to help identify environmental sources giving rise to cases of legionellosis, thus allowing control measures to be implemented and further cases to be prevented in a timely manner. Monoclonal antibody (mAB) subgrouping is probably the most rapid and reliable phenotypic method available and is excellent for initial screening of isolates. Only a few of the many genotypic methods described have been shown to meet these exacting requirements. Of these, pulsed-field gel electrophoresis is the most widely applied, but others have also been used. For example, a restriction-fragment length polymorphism (RFLP) typing method was used as the standard method in the United Kingdom for almost 20 years. Investigation of travel-associated legionellosis requires even more rigorously standardized systems, as not only must the types be clearly defined but they also must be recognizable and reproducible in laboratories in different countries (different laboratories, different times). The standardized sequence-based typing (SBT) method targets part of six genes (flaA, pilE, asd, mip, mompS, proA), each of which shows considerable allelic variation, thus providing an extremely discriminatory typing system. Furthermore, of the seven major-outbreak-associated strains for which SBT data are available, five are SBT 3,4,1,1,14,9. However, if substantiated, these observations have important and widespread implications for our understanding of the epidemiology of legionellosis and for the monitoring and control of L. pneumophila in the built environment.

It is widely known that Legionella longbeachae appears to be more common in Australia than anywhere else. Western Australia has reported up to 93% of positive cases being L. longbeachae. In a review of national data, 51% of Australian notifications were for Legionella pneumophila infections, and 42% were L. longbeachae. Victorians are about 10 times more likely to have a Legionella urinary antigen test than other Australians. The differences in reported proportions of Legionella species across Australia may also be influenced by different testing patterns among clinicians. Many would argue that it is not plausible that outbreaks of legionellosis are being routinely missed in other parts of Australia and that the number of outbreaks in Victoria simply indicates a greater risk environment, most clearly exemplified by the aquarium outbreak. In summary, it would appear that an iterative process has taken place in Victoria, with more testing leading to more case detection, which has led to more outbreaks being detected, raising clinician awareness, leading to more testing, and so on.

The European Surveillance Scheme for Travel Associated Legionnaires’ Disease (EWGLINET) has helped strengthen European national surveillance systems and improve case detection and reporting in travel-associated cases. The guidance and legislation introduced in England and Wales over the past 25 years aims to ensure that all relevant systems are properly installed and maintained and that investigation of these systems (e.g., wet cooling towers) in an outbreak situation is as efficient as possible. These measures are aimed at reducing the absolute number of cases of Legionnaires’ disease. England and Wales’ annual case reports severely underrepresent the true annual incidence of Legionnaires’ disease. The urinary antigen test was introduced in England and Wales in the 1990s and is now the primary diagnostic method for 80% of English and Welsh cases of Legionnaires’ disease. The test is quick and easy to perform and might therefore be enabling countries to detect milder cases of the disease that would otherwise have gone undiagnosed. However, in England and Wales the number of cases has remained relatively constant despite the increasing use of the test. The simple trend of an increasing number of cases of Legionnaires’ disease in England and Wales over the past 25 years masks a more complicated picture. Reporting systems that suffer from underdiagnosis and underreporting are likely to register an increase in case reports (as ascertainment improves) before any actual decrease from improved control and prevention in case numbers is detected.

Community-acquired and hospital-acquired legionellosis have been occasionally reported in Korea since the recognition of an outbreak of Pontiac fever in 1984. However, there was a lack of epidemiological information on the diagnosed patients that could be used as serological evidence. Researchers analyzed the epidemiological and clinical data and antibiotic treatment records for the patients diagnosed serologically as having legionellosis in Korea during 1999 to 2002. The common clinical symptoms included fever, cough, sputum, and dyspnea. Twenty-eight patients (25.9%) showed reactivity for Legionella gormanii, and 14 patients (13.0%) were associated with L. pneumophila serogroup 1 according to serological tests. Most of the legionellosis cases diagnosed with serological methods in Korea show epidemiological aspects similar to those reported in other countries. In sero-diagnosis of this study, it is remarkable that L. gormanii was the most prevalent species in Korea. Therefore, L. gormanii infection as well as other Legionella species other than L. pneumophila should be taken into consideration when serological diagnosis is performed in Korea.

This chapter presents the preliminary results of a large multicentric study carried out in Italy to detect Legionnaires’ disease among patients with pneumonia. The general objective was to better evaluate the prevalence of disease within various Italian regions in order to confirm the existing data on the disease frequency and to eventually verify differences related to the geographical area. A case-control study was also designed to investigate behavioral, personal, and environmental risk factors related to the disease. Among studied parameters, the presence of chemokine receptor (CCR) 5D32 and CCR264I mutations was investigated to establish their possible role in susceptibility to Legionnaires’ disease. From 2001 to 2004, an active surveillance has been carried out in six hospitals (between 800 and 1,000 beds) located in the north, center and south of Italy. Both community-acquired and suspected nosocomial pneumonia were considered. A brief questionnaire for collecting personal data and disease characteristics was filled out for each examined patient. In conclusion, to always perform laboratory tests for Legionnaires’ disease appears an effective method to increase the detection of cases. The active clinical surveillance scheme was able to identify 189 cases, contributing to increased reports in Italy, where in total about 600 cases for year are reported. The preliminary results of the case-control study show that cases have peculiar, although not sufficiently studied, alterations in both immune and biochemical parameters, which deserve further investigations. Furthermore, a more extensive study may support a possible involvement of CCR2 in resistance and susceptibility to Legionnaires’ disease.

Several surveys show that Legionella pneumophila is also present in the animal species living in the natural environment. This observational epidemiological survey aims, primarily, at evaluating the antibody coverage with particular reference to previous infection (IgG) by looking at predefined populations based on predictable risk versus Legionella exposure. Second, some demographic characteristics are evaluated, in particular those that reflect behaviors or situations that seem to facilitate Legionella infection. Double-sample tests were used to enable initial separation of positives from negatives; then the positive ones were analyzed to carry out IgG-IgM labeling. Preliminary conclusions show that seroprevalence rate seems to be very low. Legionellosis can be suspected clinically, but diagnosis can be confirmed only by laboratory testing. Higher serum positivity among frequent users of sport facilities seems to be significant, as does the lower positivity of older in-patients with several hospitalizations. Positive sample confirmation and the evaluation of doubtful samples by means of indirect fluorescent-antibody assay method will help provide a more clear and comprehensive picture.

In Spain, where Legionnaires’ disease has been a notifiable disease since 1997, the incidence of Legionnaires’ disease has increased considerably in the past few years, but a large interregional variability still persists, similar to that existing in other countries. Seroepidemiological studies of Legionella pneumophila have usually been cross-sectional and have, almost always, been utilized to ascertain rates of exposure to the bacteria in apparently healthy populations or in different risk groups. This chapter outlines the underlying hypotheses of a retrospective study of patients with pneumonia. It also discusses the scope of the seroepidemiological study. The average prevalence of seropositivity to L. pneumophila antibodies in all the samples analyzed was 7.03%.The highest prevalence corresponded to the Balearic Islands (8.66%) and the lowest to Andalusia (5.95%), with Catalonia occupying an intermediate position (6.86%). A prospective study would permit the elimination of a possible bias in the selection of patients remitted to the laboratory for analysis. A study that also analyzes the evolution of the seroprevalence in an apparently healthy population would be useful in extrapolating the results to the general population.

Travel-associated Legionnaires’ disease (LD) has caused concern since the early 1980s. As a result, the European Group on Legionella Infections (EWGLINET) was set up in 1986 in order to identify travel associated cases and clusters of the disease. This chapter measures and compares the risks associated with travel to different destinations in Spain for travelers from several European countries as well as for Spanish nationals. A total of 6,411 cases of LD of any origin (community, nosocomial, and travel associated) were reported trough the Spanish reporting system from 1999 to 2004. Major variations exist from year to year in the figures for visitors from individual countries. The authors have estimated the incidence rate of travel-associated Legionnaires’ disease by identifying the denominators as the number of days stayed at accommodation sites. They have used them as an approximation of risk exposure. This is a new approach, as normally calculations correspond to cases per number of tourists. The decrease of global incidence rates for foreigners and Spaniards from 2003 on, coinciding with the entrance of the regulating law for the control of legionellosis, permits us to be optimistic in terms of our capacity to prevent this disease in the future.

This chapter determines the prevalence of Legionella spp. and identifies the chromosomal DNA subtypes and the genomic stability of these isolates during a 20-month follow-up period in the potable-water distribution system of a hotel. From June 2003 to March 2005, 110 water samples were collected from a hotel water supply comprising 11 sample collections of water. Legionella pneumophila serogroup (sg.) 1 was isolated in 49 (44.55%) of the 110 samples tested. The cold water distal point was L. pneumophila–positive throughout the whole study period. It should be pointed out that this distal point of cold water was an infrequently used point (roof-top tap) with stagnant water, which may explain the low chlorine levels (range, 0.05 to 0.95 mg of Cl2 per liter), the heating of water during warmer seasons and thus, colonization by Legionella throughout the study period. Environmental isolates of Legionella were genotypically characterized by pulsed-field gel electrophoresis. The number of colonized points was uniform throughout the 20-month study. All L. pneumophila isolates analyzed showed the same chomosomal DNA genotype, which indicated the clone stability over the 20-month follow-up period.

This chapter evaluates the changing patterns of strains causing nosocomial Legionella infections (nLD) from a university hospital water system during a decade of active surveillance. All bronchoalveolar lavage specimens were routinely screened for Legionella. Bronchial secretions and urine probes from patients with suspicious atypical pneumonia were processed by routine culture methods. Isolates of Legionella spp. from patients and available corresponding environmental ones were genotyped using an AFLP technique previously described by EWGLI and further subtyped by monoclonal antibody analysis. Legionella-positive clinical specimens were reported to the infection control team, and every case was carefully followed up. Detection of Legionella was defined as nLD if the patient was admitted to the university hospital at least 10 days before and no other source of infection could be observed. Currently, the majority of Legionella infections in Germany are detected by a urine antigen test, and consequently the utilization rate of culture methods to detect Legionella has decreased over the past years to 11% in 2004, but only the latter method allows comparison of patient isolates with the ones isolated from the water system.

This chapter provides a representative overview of the scope of Legionnaire's disease (LD) and assess the type and frequency of implemented diagnostic methods and transmission control practices. A questionnaire was sent to the clinical directors (with two reminders) of each selected hospital. The anonymous survey included questions about hospital demographics (type of hospital, bed count, presence of risk patients [neonates, oncological, transplanted, HIV positive, and immunsuppressed ICU patients]), episodes of nosocomial and community-acquired LD in the past 5 years, diagnostic testing for LD, environmental sampling practices and the results obtained, as well as questions about prophylactic measurements and maintenance of hospital water systems in general and especially for high-risk patients. 56 out of 60 (93%) of the hospitals sample their water (peripheral and/or central) for Legionella, 31 (55%) with positive results. Of the two hospitals reporting nosocomial LD, only one had positive sampling results, while the other one did not detect Legionella. Considering the high costs of testing for Legionella and for LD prevention, further research is indicated to establish standard procedures concerning low- and high-risk patients.

Legionellae are part of the microbial community of aquatic ecosystems, natural as well as man-made, which explains why legionellosis occurs worldwide. In many countries Legionnaires’ disease is a notifiable disease. This chapter presents the preliminary results of the distribution of Legionella genotypes cultured from patients and environmental sources. Between August 2002 and September 2005, sero- and genotyping of 130 patient isolates and 220 environmental isolates showed that 98% of the patient strains were from the Legionella pneumophila genus. The genotypes 004 Lyon and 010 London, responsible for almost one-third of all Legionnaires’ disease patients in the authors' study period in The Netherlands, were not found in the environmental samples collected from the potential sources that resulted from the structured interviews according to the questionnaires. The preliminary results of the study indicate that systematic collection and sampling gives insight to the distribution of the L. pneumophila genus and to serotypes and genotypes in humans and in the environment. Based on the findings, actions in The Netherlands should be more aggressive for the European Working Group for Legionella Infections (EWGLI) genotypes 004 Lyon and 010 London. Legionellae are capable of infecting humans by aerosol inhalation or by drinking and subsequent aspiration of water.

During outbreak investigations molecular analytical methods are necessary to discriminate among clinical isolates and to establish a link with environmental isolates. In addition to monoclonal antibodies used as an initial subtyping method, other methods employed have been pulse-field gel electrophoresis, amplified fragment length polymorphosim (AFLP), arbitrarily primed PCR, multilocus variable-number tandem repeat, and sequence-based typing (SBT). In the present study, the authors used SBT to study the relationships among clinical and environmental isolates of Legionella pneumophila isolated from the same hotel during a 22-year period. Water samples were collected from the hotel and cultured using standard procedures for the recovery of Legionella isolates from environmental sources. To determine the discriminatory power of the SBT procedure, eight isolates that share the same monoclonal antibody (MAb) subtype were included. These isolates represent four enzyme types based on previous studies using multilocus enzyme electrophoresis (MLEE). All isolates were tested with a panel of MAbs to determine their subtype. Isolates were compared by AFLP. Patient isolates and representative environmental isolates from both outbreaks were analyzed by a sequence-based typing scheme.

The sequence-based typing (SBT) method was used to determine the allelic profiles of three sporadic clinical isolates as well as seven environmental isolates of Legionella pneumophila serogroup 6, isolated at the Bacteriology Laboratory Unit of the Karolinska University Hospital in 2004. The SBT may prove to be an effective aid for the epidemiological investigation of nosocomial Legionnaires’ disease (LD). Legionella species, especially Legionella pneumophila is a well-known cause of nosocomially acquired pneumonia, occurring frequently in immunocompromised subjects. The SBT method was used in a study to determine the allelic profiles of three L. pneumophila serogroup 6 clinical isolates from suspected cases of nosocomial LD as well as seven environmental L. pneumophila serogroup 6 isolates collected from drinking water sources in the vicinity of the patients being investigated. The aim of this study was to correlate the nosocomial infections to the environmental strain in the water systems of the hospital. Interestingly, the four control isolates collected from other hospitals in Sweden shared the same SBT profile, a finding that can be explained by the possible limited genomic heterogeneity within L. pneumophila serogroup 6 compared to serogroup 1. Further typing of a larger number of L. pneumophila serogroup 6 isolates is needed to validate this theory.

Direct aspiration of water during submersion has caused Legionnaires’ disease (LD) cases. This chapter presents one such case caused by Legionella pneumophila with a novel source of infection. Appropriate antimicrobial therapy was given once LD was identified, in addition to the antibiotics targeted against other pathogens. The patient suffered from bilateral lung infiltrates, pleural effusion, and cavitation of lung parenchyma, all symptoms of LD. The patient had LD and Pseudomonas aeruginosa and fungal infections. The pulmonary damage led to multiorgan failure, and the secondary or co-infections contributed to his death. LD cases have also occurred in neonates after water births. Immersion in river water has resulted in at least two reported LD cases caused by L. pneumophila sg 10 and sg 13.

A case control study was performed in South Australia to determine if Legionella longbeachae infection was associated with recent handling of commercial potting mix and to examine possible modes of transmission. Recent use of potting mix was associated with illness (odds ratio [OR] 4.74; 95% confidence interval [95% CI], 1.65 to 13.55; P=0.004) in bivariate analysis. This chapter suggests there are other factors within the gardening environment, as well as intrinsic and behavioral host factors, that are better predictors of L. longbeachae infection than recent use of potting mix. An association between illness and proximity to dripping, hanging pots supports inhalation of contaminated aerosols produced during watering as a possible mode of transmission. Another possible mode of transmission is ingestion of organisms via contaminated hands, which is supported by the association between illness and eating or drinking after gardening before washing hands. Further information on risk factors for L. longbeachae infection that explores the interface between potting mix, other gardening exposures, and behaviors while gardening need to be addressed in a larger study. Long-term smokers and possibly people with preexisting medical conditions such as respiratory and cardiac illness should be warned about their increased risk of L. longbeachae infection. Raising people’s awareness of a possible health risk when using potting mix should continue in order to protect against L. longbeachae infection.

The recently described sequence-based typing (SBT) scheme for epidemiological typing of Legionella pneumophila offers many advantages over other methods. However, clinical and environmental isolates are not always available to allow epidemiological typing to be performed. For this reason the authors have sought to apply the SBT technique directly to clinical and environmental samples where isolates are not available or before they become available. This has the potential to allow more rapid determination of typing data or the possibility of obtaining such data when there is no alternative, i.e., in the absence of isolates. One aim was to determine if this direct approach was feasible. This was established using environmental concentrates, together with environmental and clinical isolates, obtained during the investigation of the largest United Kingdom outbreak to date, in Barrow-in-Furness in 2002. The case had antibody levels consistent with infection by L. pneumophila, but no isolate was obtained. In 2005 a cluster of three cases was reported from the same complex. Respiratory samples from two suspected cases (cases 1 and 2) were obtained, but L. pneumophila was only isolated from one (case 1). PCR analysis of the extract from case 2 was also undertaken using an in-house L. pneumophila–specific real-time PCR assay targeting the macrophage infectivity potentiator (mip) gene. SBT was performed on DNA extracted from both the isolate (case 1) and the clinical sample (case 2). Direct SBT was also successfully demonstrated using clinical samples including respiratory and postmortem specimens.

This chapter investigates the feasibility of providing a dedicated database allowing the online identification of putative Legionella spp. following standard PCR amplification and sequencing using a previously published identification scheme for Legionella. A standard PCR and sequencing protocol based on a previously published sequence-based identification scheme was used to generate a DNA sequence from the macrophage infectivity potentiator gene (mip). Identification of user-generated sequences was performed using tools showing: (i) an alignment of all the sequences from the reference alignment, top five database matches, and the user sequence; (ii) a neighbor-joining tree of the alignment, including the reference species, five closest matches from the database, and the user sequence; and (iii) an alignment of the eight sequences from the combined alignment that are most similar to the user sequence together with percentage similarity scores. It is anticipated that additional genes, including those coding for 16S rRNA, RpoB, and RnpB, will be added to the current identification system to aid in the characterization and identification of known and potential novel members of this genus.

Pulsed-field gel electrophoresis (PFGE) is considered to be one of the most discriminative epidemiological methods for subtyping Legionella pneumophila strains and for elucidating the sources of infection. On the other hand, sequence-based typing (SBT), which was recently developed, is also a powerful epidemiological method. All L. pneumophila serogroup 1 isolates used in a study were independently obtained from a wide variety of Japanese locations. The primers used for SBT and the reaction mixture and conditions were the same as those used by Gaia et al. Putative novel variants found in this study were submitted to the curators of the European Working Group for Legionella Infections SBT database for verification and assignment of new allelic numbers according to the curators’ instructions. Both PFGE and SBT indicated that the spa-bath isolates from Japan were highly diverse. The water used in public spa baths in Japanese resorts is mostly obtained from hot springs. The first clinical isolate in Japan was not assigned to the cooling tower SB-type. Only four SB types were common between Europe and Japan. Although PFGE is the most widely used technique and is generally accepted to be highly effective in discriminating genomic differences, it may have certain drawbacks with regard to interlaboratory reproducibility. SBT appears to be less effective at discriminating between strains than PFGE. The authors have to consider the advantages and limitations of both methods and apply the most suitable method according to the requirements.

Assessment of the sequence quality is a skilled task and can often be very subjective. Therefore, the primary aim of the study described in this paper was to develop novel web-based tools to facilitate: (i) database curation, designation of novel allele types, and allelic profiles; (ii) automated quality assessment of sequence chromatograms. The outcome of the study is an online tool, which is now accessible via the European Working Group for Legionella infections website to allow users to assemble sequence traces (from the sequence-based typing (SBT) scheme) and to identify matching preexisting alleles. Most SBT types are derived from at least two sequences, typically one forward and one reverse sequence. Extensive sequence analysis and manipulation of the sequences and assembled contig is carried out using the BioPerl toolkit of objects. The processes involved include quality-trimming, assessment of the proportion of the contig that is double-stranded, and consensus generation. Data analysis from multicenter proficiency SBT analysis has confirmed that misidentification readily occurs when poor-quality trace files are used. This online tool will help systemically check sequence quality and will assist in the determination of minimum standards for a range of sequence quality parameters, ensuring that only high-quality data is accepted for allele and profile identification.

This chapter describes alternate approaches that take advantage of special properties of Legionella pneumophila strains that allowed the formulation of simple screens using toothpicks and petri dishes containing bacteriological medium. Developing readouts in the microorganism that allow screens on solid medium has allowed a number of translocated substrates to be identified. Interestingly, it was found that dotL mutants producing low levels of DotA were able to grow on agar plates. This important finding allowed considerable manipulation of phenotypes, because the behavior of mutants that would otherwise be inviable could now be evaluated during intracellular growth. The great majority of mutations, in fact, appeared to directly affect the assembly of the Dot/Icm system or disrupt function of DotL, and many of the impaired proteins are involved in either membrane biogenesis or membrane protein folding. In several T4SSs,Mob protein can be transferred into recipient cells in the absence of DNA mobilization, indicating that protein translocation into bacteria can occur in the same fashion that is observed when the pathogen contacts a mammalian cell. Furthermore, elimination of multiple paralogs of a single gene family does not solve the redundancy issue, suggesting that a similar amino acid sequence does not necessarily mean the proteins have similar functions. Development of functional and biochemical assays for these proteins will be necessary to sort out their roles in intracellular growth.

The availability of a variety of genetic tools, a complete genome sequence, and several convenient host systems offers the opportunity to study the molecular mechanisms of the interactions between Legionella and its hosts. Examination of the Legionella genome sequence revealed the existence of a relatively large number of genes that resemble eukaryotic rather than prokaryotic orthologs. Alternative approaches based on genome inspection or the use of clever genetic screens for other phenotypes eventually led to the identification of several effectors. In one case, Legionella effectors that vaguely resemble components of vesicle fusion proteins (e.g. EEA1, interaptin) were shown to play a role in a previously recognized nonlytic release pathway in protozoa that results in packets of bacteria being exocytosed prior to host cell death. Frustrated by the lack of information based on classical genetic approaches, researchers sought to examine the genome for additional clues using more targeted approaches than the original BLAST-based annotation. The majority of leg gene products appear to contain motifs involved in protein:protein interactions such as ankyrin repeats, leucine-rich repeats, or coiled coils. A combination of bioinformatics and novel genetic approaches has resulted in the identification of several interesting translocated effectors. The challenge is now to determine how these effectors alter host cell organelle trafficking.

This chapter focuses on one major class of virulence factors, the dot/icm genes. The initial set of dot/icm genes was independently identified via a plate selection and an enrichment strategy in the Shuman and Isberg laboratories, respectively. Type IV secretion systems consist of both plasmid transfer systems and adapted conjugation systems used by pathogens to export substrates. Coxiella burnetii, the causative agent of Q fever, also contains a type IVB secretion system that strongly resembles the Legionella pneumophila Dot/Icm T4SS. It was initially believed, based on the homology to a plasmid transfer system, that the L. pneumophila T4SS might transfer a DNA substrate into the host cell similar to the plant pathogen Agrobacterium tumefaciens. The L. pneumophila Dot/Icm T4SS has the ability to transfer plasmids between bacterial cells and can export a variety of protein substrates into eukaryotic host cells.

Legionella pneumophila is the causative agent of Legionnaires’ disease, a fatal form of pneumonia in elderly and immunocompromised people. L. pneumophila replicates inside host cells by using a type IV secretion system (T4SS) encoded by the dot/icm genes. The L. pneumophila dot/icm genes are essential for intracellular multiplication of L. pneumophila. To determine the localization of the L. pneumophila T4SS apparatus, the authors initially investigated the distribution of one Dot protein, DotF. DotF, predicted to localize to the inner membrane by hydrophilicity analysis, has been shown to interact with secreted substrates via a bacterial two-hybrid screen. Quantitative analysis showed that more than 95% of cells had at least one focus of fluorescence at a pole. The majority of the cells showed bipolar DotF staining. The polar staining was specific to DotF because it was not apparent in a strain lacking the dotF gene. Agrobacterium tumefaciensVirB T4SS has been proposed to provide a selective advantage since this plant pathogen attaches to its host cell in a polar manner, thereby allowing intimate contact between the secretion system and the plant cell. Based on the observations, and the work on the A. tumefaciensVirB system, it is possible that polar localization is a conserved feature of all adapted type IV secretion systems. Alternatively, as in the case of adapted conjugation systems and IcsA, there may be a selective advantage or molecular requirement for the proteins to be localized to the bacterial pole.

The ability of Legionella pneumophila to cause disease is primarily due to its ability to survive inside alveolar macrophages, immune cells capable of destroying most bacteria. The Dot/Icm proteins make up a type IVB secretion system that is required for delivery of multiple protein substrates into the host cell cytoplasm, where they are believed to alter the endocytic pathway and allow the bacterial phagosome to avoid fusion with lysosomal markers. Type IV secretion systems (T4SSs) are used by many pathogens to deliver substrates to host cells, including Helicobacter pylori, Bordetella pertussis, and Brucella abortus. The secretion systems used by these pathogens are ancestrally related to plasmid transfer systems and are thus referred to as adapted conjugation systems. Our goal is to characterize the molecular details of how the Dot/Icm secretion complex assembles and functions to export substrates. The approaches to understanding these processes are based on techniques that have proven successful in characterizing the canonical T4SS, the VirB/D4 system from the plant pathogen Agrobacterium tumefaciens. To identify functional subcomplexes of the Dot/Icm T4SS, researchers are currently examining protein interactions between Dot/Icm components. One approach researchers are taking to accomplish this is to determine effects on protein stability caused by deletions of other dot/icm genes. This approach is based on the fact that proteins that interact in a complex often require their interaction partners for stability.

Legionella pneumophila possesses a large variety of lipolytic enzyme activities that affect phospholipids. In addition to the phospholipid-degrading phospholipase A (PLA) and lysophospholipase A (LPLA) activities, glycerophospholipid:cholesterol acyltransferase (GCAT) activity has been described in this chapter. First, the authors were interested in whether all of the 11 PLP genes are expressed in L. pneumophila Philadelphia-1 during growth in laboratory media. In order to assess expression of the PLP genes, mRNA was isolated at four growth phases (early logarithmic, mid-logarithmic, late logarithmic, and early stationary) in standard laboratory media and subsequently used for reverse transcriptase PCR with L. pneumophila Philadelphia-1 PLP gene-specific primers. L. pneumophila wild type and the patA/vipD mutants were grown to late exponential phase, and cell lysates and culture supernatants of the bacteria were tested for PLA, LPLA, and lipase activities. Importantly, in coinfection assays with amoebae and macrophages, the L. pneumophila patA/vipD mutant strains were severely impaired for intracellular replication. Thus, PatA/VipD is a new type IVB secreted phospholipase of L. pneumophila with essential importance during host cell infection.

The IcmG and IcmS proteins are not constituents of the membrane-spanning part of the Icm/Dot T4SS; rather, these proteins bind to secreted effector proteins, thus possibly facilitating their translocation. Perhaps this specific function accounts for the only partial phenotype in the amoebae plate test (APT) of Legionella pneumophila strains lacking IcmG or IcmS. We used icmS and icmG mutants to screen by the APT an L. pneumophila genomic library for multicopy suppressors. Overexpression of lcsC but not its paralogue lpxB in an L. pneumophila icmG mutant strain was cytotoxic for Acanthamoeba castellanii. L. pneumophila LcsC and its paralogue LpxB share 31% identity on an amino acid level and are 30% or 42% identical, respectively, to the LpxB orthologue from E. coli. A closer inspection revealed that this oxidoreductase is 39% identical to the GnnA protein of Acidithiobacillus ferrooxidans, which together with the GnnB protein catalyzes the conversion of UDP-N-acetylglucosamine (UDP-GlcNAc) to UDP-GlcNAcN3. Finally, not only the glycosyl transferase gene lpxB but also the acyl trans-ferase genes lpxA and lpxD are present in at least two copies in the L. pneumophila genome. The Icm/Dot T4SS is not only required for the upregulation of phagocytosis and intracellular replication of L. pneumophila, but also for growth of L. pneumophila in the presence of A. castellanii on agar plates. The APT established here may prove useful to discover other bacterial factors relevant for interactions with amoeba.

Translocation of fully or partially folded proteins across bacterial membranes is a remarkable property of the type II secretion (T2S) system and the twin-arginine translocation (Tat) pathway. This chapter summarizes our current knowledge of the significance of T2S and Tat for Legionella pneumophila. Mutational analysis showed that the Legionella secretion pathway (Lsp) promotes the secretion of at least 11 degradative enzymes, including acid phosphatases, chitinase, zinc-metalloprotease, ribonuclease, mono-, di-, triacylglycerol lipases, phospholipase A (PLA), lysophospholipases A (LPLA) and phospholipases C (PLC). Although all these strategies have been successful, they all have their limitation in defining the complete set of proteins secreted by the Lsp pathway. First, the genes encoding the secreted chitinase, ribonuclease, and tartrate-resistant acid phosphatase have not yet been identified. Second, genetic analysis indicates that some identified exoenzymes do not account for all the corresponding Lsp-dependent activity; e.g., supernatants of the plcA-negative strain retain 50% of the PLC activity of the wild type, indicating that there is more than one Lsp-secreted PLC. Finally, there are probably some additional Lsp-secreted proteins whose function has not been tested for. In conclusion, the Lsp system promotes the secretion of degradative enzymes, growth at low temperature, intracellular replication within amoebae and macrophages, as well as virulence in A/J mice. Moreover, the Tat pathway facilitates secretion of phospholipase C, cytochrome c–dependent respiration, growth in iron-limiting conditions, as well as intracellular replication. Further identification and characterization of Lsp and Tat substrates should therefore enhance our understanding of the pathogenesis of L. pneumophila.

The type II protein secretion system of Legionella pneumophila is important for bacterial survival in protozoa, macrophages, and the lungs of A/J mice. Presently, L. pneumophila is known to express both type II and type IV secretion systems. Type II secretion (lsp) mutants of L. pneumophila grow normally in bacteriologic media at 30 to 37°C. The low-temperature defect of lsp mutants of L. pneumophila serogroup 1 strain 130b was partially due to the presence of 0.25 g/liter ferric pyrophosphate in the growth media, since bacterial growth on agar plates at 25°C improved significantly when that supplement was omitted. The inhibitory effect of iron was specific to the lsp mutant and to low-temperature growth, since the growth of wild type at 25°C did not change with variations in the amount of supplemental iron, and the growth of the lsp mutant was not negatively influenced by iron supplementation when incubated at 37°C. In the environment, L. pneumophila is found in freshwater and man-made water systems. In sum, the results presented indicate that the L. pneumophila type II secretion system is not only important for growth at 35 to 37°C in host cells but is also critical for extracellular growth and survival at lower temperatures, including iron-rich bacteriological media and environmental water samples. Thus, type II secretion appears to play a central and multifaceted role in the natural history of Legionnaires’ disease.

Different secretion pathways have been shown to play a role in the virulence of Legionella pneumophila. Recently the authors showed the presence of the twin-arginine translocation (Tat) pathway in L. pneumophila Philadelphia-1 and its importance in intracellular replication and biofilm formation. The Tat pathway translocates folded proteins across the cytoplasmic membrane. In order to study the importance of the Tat pathway in the virulence of L. pneumophila, the identification of Tat substrates and their possible involvement in virulence was initiated. Since some Tat substrates might be transported across the outer membrane following Tat-dependent transport across the cytoplasmic membrane, the authors looked for differential spots in culture media of the wild-type strain and two Tat secretion mutants (tatB and tatC mutant) by two-dimensional protein gel electrophoresis analysis. Three proteins were found to be absent from the culture medium of the Tat secretion mutants: LvrE, Lpg1962 (a peptidyl-prolyl cis-trans isomerase), and Lpg2320 (a hypothetical protein). The lvrE gene is situated in between the lvh (Legionella vir homologues) genes on the L. pneumophila Philadelphia-1 genome that encode the Lvh type IV secretion system. Based on two-dimensional analysis, three proteins were found to be absent in the culture media of the L. pneumophila tatB and tatC mutant. For one of these proteins, LvrE, Tat dependence was confirmed using specific antibodies.

This chapter describes a first putative type I secretion system (Lss) of Legionella pneumophila which is encoded by the lssXYZABD locus. In order to identify substrates of this Lss secretion system, the authors performed comparative two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) analysis of extracellular proteins of L. pneumophila Corby wild type, Corby lssB mutant, and complemented lssB mutant strain. In this study it was shown that the identified VirK protein of L. pneumophila is secreted in an LssB- but not a TatC-dependent manner. Referring to the results, it is implied that the Lss secretion system is also linked to the Lsp type II secretion system in L. pneumophila. It is speculated that VirK and p14 are secreted via a concerted action of the Lss and the Lsp system which includes cleavage of the signal sequences during secretion. In Agrobacterium and Rhizobium VirK is associated with the vir-regulon, and it was postulated that VirK should have a function in the interaction of the bacteria with the host cells. Therefore, the authors started to generate a virK mutant strain of L. pneumophila Corby and an antibody against VirK to further characterize the expression, secretion, and function of VirK of L. pneumophila.

The macrophage infectivity potentiator (Mip) protein of Legionella pneumophila is one of the most studied Legionella proteins and has long been known to promote virulence. Mip protein has been purified and shown to possess peptidyl-prolyl cis/trans isomerase (PPI-ase) activity. Comparison of primary structures and crystallographic studies shows that Mip proteins have an N-terminal and a C-terminal domain. Type II secretion is one of five protein secretion systems that can mediate the export of proteins across the gram-negative outer membrane into the extracellular milieu and/or into target cells, and the authors have shown that the type II protein secretion system of L. pneumophila is required for optimal replication in macrophages, amoebae, and mice. Interestingly, one of transposon mutants (previously designated NU247) proved to contain a single transposon insertion in the mip gene, suggesting, for the first time, that Mip might influence protein secretion. This chapter talks about the NU247 reproducibly that showed a 40 to 70% reduction in p-NPPC hydrolase activity in its culture supernatants in comparison to the wild-type strain 130b. Since Mip is highly conserved in the Legionella genus, and surface and secreted Mip-like proteins are present in other pathogenic microorganisms, Mip and Mip-like proteins might promote the secretion of other important effectors.

Legionella pneumophila possesses several virulence mediating factors; for example, it has two protein export systems: the type II (Lsp) and the type IVB (Dot/Icm) secretion systems. Both systems transport effector molecules, including lipolytic proteins such as phospholipases. L. pneumophila possesses two major phospholipase activities, phospholipase A (PLA) and lysophospholipase A (LPLA), both capable of hydrolyzing phospholipids and generating the reaction products: free fatty acids (PLA and LPLA), cytotoxic lysophospholipids (via the action of PLA), and water-soluble phosphodiesters, like glycerophosphorylcholine (via the subsequent action of LPLA on lysophospholipids). Patatins are a group of plant storage glycoproteins that show lipid acyl hydrolase activity. One of the corresponding proteins, PatA, was found to be an LPLA when expressed in Escherichia coli. The patA to patK genes of L. pneumophila have been determined to be expressed during bacterial growth in laboratory media. Furthermore, by the construction of knockout mutants, it was shown that patA contributes to secreted bacterial PLA and LPLA activities and is essential for intracellular replication of L. pneumophila, during both amoeba and macrophage infection, which implies an important role for lipolytic enzymes with respect to the pathogenic behavior of the bacterium. The high number of lipolytic enzymes present in L. pneumophila implies that lipids may be an important source of nutrients for the bacterium, and/or lipid hydrolysis is an essential step in bacterial establishment, especially within the host cell.

The major proteins of the anion-exchange fractions with phospholipases A (PLA) activity were N-terminally sequenced. This chapter analyzes the bacterial membrane lipid composition of the Legionella pneumophila aas mutants and the 130b wild type by thin-layer chro-matography (TLC). The role of Aas during L. pneumophila intracellular infection was assessed in the three host models, U937 macrophages, A549 epithelial cells, and Acanthamoeba castellanii amoebae, and Aas was found to be dispensable, because L. pneumophila aas mutants showed the same increase in CFU in all three hosts during 72 h of infection as the wild type. The chapter presents data that show that L. pneumophila Aas contributes to the incorporation of phospholipids into the bacterial membrane, and in addition to L. pneumophila PlaA, is another enzyme involved in the detoxification of lysophospholipids. The authors examined the corresponding L. pneumophila 130b unk1 mutants for hydrolysis of diacylphospho-lipids, monoacylphospholipids, and 1-MPG and found that Unk1 did not contribute to the secreted PLA and L. pneumophila phospholipases A (LPLA) activities of L. pneumophila. The data indicate an essential role for Unk1 in infections of amoebae and macrophages by L. pneumophila. The authors identified three proteins, Aas, LvrE, and Unk1, present in the L. pneumophila culture supernatant. Aas contributes to the cell membrane integrity of L. pneumophila, but is dispensable for the infection of host cells. Since bacterial lipolytic enzymes are important bacterial tools for surviving both inside and outside of hosts, their characterization promotes one's understanding of the life cycle of L. pneumophila.

Legionella pneumophila possesses secreted phospholipase A (PLA) and lysophospholipase A (LPLA) activities which might be involved in bacterial virulence, because they represent tools for modification and lysis of host cell membranes and for the interference with the host signal transduction pathways. The L. pneumophila genomes Philadelphia-1, Paris, and Lens code for three members of the GDSL family of lipolytic proteins, namely PlaA, PlaC, and PlaD. The authors aimed to identify the factor which directly activates PlaC. Furthermore, they sought to characterize PlaD, the third L. pneumophila GDSL enzyme, and to compare the lipolytic properties as well as the capacity of L. pneumophila plaD mutants to multiply intracellularly with L. pneumophila wild type and plaA and plaC mutants. The PlaC-activating factor, characterized by its ability to enhance glycerophospholipid:cholesterol acyltransferase (GCAT) activity of PlaC expressed in Escherichia coli, was biochemically purified from the culture supernatant of an L. pneumophila 130b plaC mutant by anion exchange chromatography (AEX). The main protein in the activating fractions consistently was the zinc metalloprotease ProA. To investigate whether L. pneumophila Corby plaD could confer PLA and LPLA activities to E. coli, cell lysates as well as the culture supernatants from E. coli clones carrying plaD in trans were examined for hydrolysis of PLA, LPLA, and lipase substrates.

Bacteria have evolved numerous ways to acquire iron from the environment, but a common mechanism is the production of siderophores. Siderophores are electronegative, iron-regulated, low-molecular-weight compounds produced by bacteria and fungi that bind ferric iron and facilitate its internalization via specific receptors. Siderophores are secondary metabolites mainly synthesized by nonribosomal peptide synthetases that are similar to those used for antibiotic synthesis. Iron-bound siderophores are usually recognized at the cell surface by specific receptors in both gram-positive and gram-negative bacteria that internalize the ferrisiderophore. Siderophores are extremely effective in binding the Fe3+ ion because they contain the most effective iron binding ligands in nature, consisting of hydroxamate, catecholate, and α-hydroxycarboxylate ligands that form hexadentate Fe3+ complexes, thus satisfying the six coordination sites on ferric ions. The authors showed that L. pneumophila could produce a high-affinity iron-chelator. When grown at 37°C in a low-iron chemically defined medium (CDM), L. pneumophila secretes a low-molecular-weight substance that is reactive in the Chrome Azurol S (CAS) assay. Indeed, legiobactin does not extract into common solvents that are used to extract catecholate and hydroxamate siderophores. The legiobactin peak is the portion of L. pneumophila supernatants that promotes growth of iron-starved legionellae and is absent in the lbtA mutant supernatants. As was the case using 13 C nuclear magnetic resonance (NMR) analysis, proton NMR analysis of purified legiobactin demonstrates that the siderophore contains only aliphatic residues. Currently, the authors are working toward determining the structure of legiobactin based on two-dimensional-NMR, elemental analysis, and Maldi experiments.

Bioinformatic analysis of the Legionella pneumophila Philadelphia 1 genome revealed four novel open reading frames (ORFs) that have no known prokaryotic homologues but are predicted to encode products with significant domain homology to eukaryotic proteins. The authors postulated that the gene products of these eukaryotic-like ORFs may interact with eukaryotic host cell proteins and thus play a role in the subversion of host cellular trafficking pathways by L. pneumophila and the establishment of the replicative niche inside host cells. RNA was extracted from stationary phase broths of L. pneumophila 130b using the Epicentre Masterpure RNA Purification Kit (MCR85102). cDNA was synthesized, and specific primers for each gene were used to amplify 200 to 1,100 base pair portions of each ORF. This revealed that all four ORFs were transcribed in stationary phase. To investigate the potential role that each eukaryotic-like gene under investigation may play in interactions with host cells, insertional mutants were constructed for each ORF. The chapter focuses on localization of the protein products of eukaryotic-like ORFs. To characterize the products of lpg1905 and lpg2644, a combination of epitope-tagging and specific polyclonal antibodies were used to localize the proteins within the bacteria.

Several studies have compared the virulence traits of different Legionella spp., yet little is known about the genetic basis of these phenotypic differences. To investigate genetic differences between L. pneumophila and L. micdadei, the authors performed a low stringency genomic subtractive hybridization between a serogroup 1 isolate of L. pneumophila and a clinical isolate of L. micdadei. Subtractive hybridization revealed 151 open reading frames (ORFs) present in L. pneumophila 02/41 and absent in L. micdadei 02/42. LadC is predicted to be an adenylate cyclase, as it possesses an intact HAMP signal transduction region and the catalytic domain for adenylate cyclases. There are several mechanisms by which LadC, as a putative adenylate cyclase, could influence the invasion of host cells. LadC, through control of cAMP levels within the bacterial cell, may influence the regulation of L. pneumophila virulence determinants or it may play a direct role in host-pathogen interactions. This investigation has demonstrated that L. pneumophila possesses virulence determinants that are absent in other Legionella species.

This chapter reports that through genetic and functional studies in yeast and mammalian cells, the authors have demonstrates that Legionella pneumophila Hsp60 alters signaling cascades in eukaryotic cells, modifies the actin cytoskeleton of mammalian cells, and alters the trafficking of mitochondria in mammalian cells, making Hsp60 a potential effector in L. pneumophila’s intracellular establishment. The authors determined that dotA and dotB mutant derivatives of Lp02 had virtually no protease-accessible HtpB; i.e., these mutants displayed no surface-exposed HtpB. In addition, immunogold electron microscopy, performed with an HtpB-specific rabbit antiserum and a secondary gold-conjugated antibody against rabbit immunoglobulin, showed that the Lp02 dot mutants, particularly the dotB mutant, had an increased amount of periplasmic gold particles. This interesting result suggested that a nonfunctional Dot/Icm system results in the accumulation of HtpB in the periplasm. Recombinant HtpB caused Chinese hamster ovary (CHO) cells to lose their stress fibers. The abilities of HtpB to specifically alter eukaryotic signaling pathways, cytoskeletal organization, and organellar traffic are indeed functional characteristics that fit well into HtpB’s potential role as an L. pneumophila virulence effector.

This chapter demonstrates for the first time to what extent the lipopolysaccharide (LPS) architecture is involved in extracellular structures of Legionella pneumophila bacteria. Furthermore, the intraphogosomal/ intracellular shedding of LPS components is investigated. The authors' investigation of the surface architecture of strain Lp02 and its dotA and letA mutants substantiate that the LPS equipment depends at least on both genes. Therefore, the valuation of the influence of dotA and letA on pathogenicity also has to consider the phenotypic changes of the LPS surface structures, because they initiate the first step of interaction between bacteria and host cells. The amounts of LPS were quantified by enzyme-linked immunosorbent assay. LPS was immobilized in microtiter wells and detected by monoclonal antibody (MAb) 3/1 followed by antimouse-IgG horseradish peroxidase. Moreover, until the postexponential growth phase in broth cultures only approximately 5% of bacteria of strains Corby and Lp02 can be labeled by MAb 59/1. After lysis of host cells it was observed that legionellae were released either as bacterial clusters positive for both MAbs or as single MAb 3/1-positive bacteria being MAb 59/1-positive or negative. The ability to modify and shed LPS allows L. pneumophila to build lamellar structures and vesicles and to release soluble LPS that may modulate host cells, enhance survival in aerosols, or stabilize biofilms. In this chapter, the authors have demonstrated that the whole LPS architecture is influenced by at least two systems, namely the type IV secretion apparatus and letA/letS regulatory elements.

This chapter determines whether developmental changes in the lipopolysaccharide (LPS) affect the ability of Legionella pneumophila to alter phagosome maturation. Modifications of the LPS of L. pneumophila serogroup 1 strains have been correlated to virulence and serum sensitivity. Spontaneous mutants lacking the lag-1 gene lose reactivity with MAb 3/1, and expression of the gene in trans restores MAb 3/1 reactivity as well as acetyltransferase activity. To test whether acetylation of the O-antigen affects phagosome maturation, the authors constructed a lag-1 mutant and compared its fate in murine macrophages to a wild-type serogroup 1 L. pneumophila strain (Lp02). Thus, the authors examined any dot-specific effects by also analyzing a lag-1 dotA double mutant in addition to a dotA single mutant. The authors first investigated whether the LPS of the lag-1 mutant differed from wild type by using immunofluorescence, Western blot, and bacterial adherence to hydrocarbon (BATH) assays. Cytotoxicity was determined by analyzing macrophage viability after a 1-h incubation with bacteria at a high multiplicity of infection. Accordingly, their working model is that, by acetylating its LPS during the replicative phase, intracellular bacteria release the inhibition to fusion with endosomal vacuoles while increasing their resistance to lysosomal enzymes. By modifying the pathogen’s surface according to growth phase, the LPS armor can alternately contribute both to transmission and to replication within key defenders of the human immune system.

This chapter presents the description of the components of outer membrane (OM) vesicles produced by Legionella pneumophila. The composition of purified vesicles was compared by immumoblot analysis and enzyme-linked immunosorbent assay using the panel of monoclonal antibodies (MAbs) listed. Additional components of the vesicles included the chaperone Hsp60, the amoebal uptake protein (Aup), and the Mip PPIase protein which is necessary for full virulence. The authors assumed that lipid A remains anchored to intravesical structural proteins, whereas carbohydrate moieties of lipopolysaccharide (LPS) are altered, either before or after vesicle formation. The LPS-enriched vesicles are composed of several structural as well as virulence-associated proteins of L. pneumophila, e.g., Mip, Aup, and Hsp60. The detection of particular dot/icm gene products inside the OM vesicles would be one approach to investigate how OM vesicles contribute to L. pneumophila pathogenesis during phagosome formation and maturation, as well as during biofilm establishment and persistence. In addition to growth phase-dependent regulation, the authors demonstrated that the LPS content of OM vesicles varies between L. pneumophila serotypes.

Legionella pneumophila has a complex life cycle composed of at least three developmental states, each characterized by a different cell type: the replicative, transmissive, and mature intracellular forms. This chapter talks about how during its life cycle in mouse macrophages L. pneumophila first delays and then exploits phagosomelysosome fusion, a virulence strategy that the authors have named “the pregnant pause’’. The authors tested whether these vesicles retain the same surface properties and activities as whole microbes. Specifically, the hypothesis that L. pneumophila releases LPS-rich vesicles that arrest phagosome maturation while also remodeling the bacterial surface during differentiation to the replicative form, was also tested. Vesicles shed by L. pneumophila or Escherichia coli into broth were purified by gradient ultracentrifugation. As it differentiates to the transmissive form, L. pneumophila gains the capacity to inhibit phagosome maturation and also alters the composition of the glycoconjugates on its surface. Like Salmonella, L. pneumophila fills its phagosome with membrane vesicles. The O-chain of L. pneumophila LPS is a homopolymer of legionaminic acid, a unique sugar structurally similar to sialic acid whose striking features are the lack of free hydroxyl groups and a high degree of O-acetylation. During this pregnant pause, L. pneumophila alters the composition of its LPS not only to release the block to fusion but also to tolerate the harsh conditions of the lysosomal compartment, which the progeny exploit as a source of nutrients and membrane to support proliferation.

Legionella pneumophila is a gram-negative intracellular pathogen that often causes a serious and life-threatening pneumonia. In lung tissue, bacteria multiply in several types of host cells, including macrophages, monocytes, and alveolar epithelial cells. In this paper, the authors present topics concerning the pathogenesis of Legionella infections, such as exaggeration of acute lung injury by hyperoxia and a role of Pseudomonas quorum-sensing molecules in the ecological niche of Legionella organisms. The authors first examined the effects of hyperoxia on survival in permissive A/J and nonpermissive C57BL/6 mice with Legionella pneumonia. Importantly, hyperoxia treatment alone (90 to 94% oxygen for 60 h), without infection, induced no death of mice in either strain. Histone-associated DNA fragments are a marker for DNA fragmentation, one of the main characteristics of apoptosis, whereas caspase-3 is an essential protease mediating apoptosis. Terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling staining of infected lung sections demonstrated increased apoptosis in hyperoxic mice, predominantly in macrophages (alveolar and interstitial) and alveolar epithelial cells. The growth of L. pneumophila was completely inhibited by 50 μM of N-3-oxododecanoyl-L-homoserine lactone (3-oxo-C12-HSL), and interestingly significant suppressions of virulence factor genes were demonstrated in L. pneumophila exposed to Pseudomonas 3-oxo-C12-HSL.

Legionella pneumophila has been shown to induce apoptosis within macrophages, monocytic cell lines, and alveolar epithelial cells. This chapter addresses the Dot/Icm-mediated induction of apoptosis in genetically susceptible A/J mice during L. pneumophila infection. In an experiment A/J and BALB/c mice were intratracheally inoculated with 106 cfu of L. pneumophila AA100- green fluorescent protein (GFP) or the dotA-GFP mutant. At different time points after infection (2, 24, 48, and 72 h) mice were sacrificed and the lungs were removed. It has been recently shown that robust activation of caspase-3 is exhibited throughout the early and exponential intracellular replication of L. pneumophila, but apoptosis is delayed and is not triggered in the infected cells until the late stages of infection, concomitant with termination of intracellular replication. L. pneumophila AA100 induces apoptosis in vivo during the late stages of Legionnaires’ disease in experimental animals. Genetically susceptible A/J mice are susceptible to the L. pneumophila Dot/Icm-mediated induction of apoptosis. Apoptosis is triggered at the late stages of infection, concomitant with termination of intracellular replication. BALB/c mice are resistant to infection of L. pneumophila AA100 as well as induction of apoptosis.

A central aspect of the pathogenesis of Legionella pneumophila is its ability to differentiate in response to nutrient availability. The phagosomal transporter (Pht) proteins are members of the major facilitator superfamily of proteins which are ATP-independent transporters that perform diverse functions in both prokaryotic and eukaryotic cells. One of these transporters, PhtA, is required by intracellular L. pneumophila for threonine acquisition, differentiation, and growth. Accordingly, the authors have developed a model in which Pht proteins are responsible for acquisition of nutrients in the nascent phagosome; have postulated that this process is absolutely required both for L. pneumophila differentiation and growth within host macrophages. Knowing that nutrient starvation in broth triggers differentiation of replicative cells to the transmissive form, the authors tested if threonine limitation, caused by mutation of the phtA locus, impaired differentiation of intracellular bacteria. Analysis of mutants that lack one of each of the other members of the Pht family indicated that some of the putative transporters are required for growth within mouse macrophages, but others are not. In particular, the phtB, phtG, phtH, phtIK genes are not essential for growth within macrophages, while phtC, phtD, phtE, phtF, and phtJ are required for replication.

Legionella pneumophila is a facultative intracellular bacterium which replicates within amoebae and macrophages by a similar mechanism. Intracellular replication within both protozoa and mammalian cells requires the bacterial icm/dot genes, encoding a conjugation apparatus related to type IV secretion systems (T4SSs). The inositol carbohydrate moiety of phosphoinositides (PIs) is phosphorylated at positions 3, 4, and/or 5 by specific kinases or dephosphorylated by phosphatases, respectively. Phagocytosis of L. pneumophila by Dictyostelium was determined by a gentamicin protection assay and by flow cytometry. Wild-type L. pneumophila (but not ΔicmT) replicated more efficiently in ΔPI3K1/2 or in wild-type Dictyostelium treated with PI3K inhibitors. To investigate whether Icm/Dot secreted proteins bind in vitro to PIs immobilized on nitrocellulose membranes, the authors performed a lipid protein overlay assay using glutathione S-transferase fusion proteins and an anti-glutathione S-transferase antibody. Thus, SidC and its paralogue SdcA were found to specifically bind to PI phosphate but not to other PIs or lipids. The result was confirmed in a pull-down assay, where the authors employed phospholipid vesicles containing either PI(4)P or other PIs.

This chapter aims at determining whether the interaction of Legionella pneumophila with freshwater ciliates provides additional insights into the factors that may predispose L. pneumophila to infect human cells. When ciliates feed on L. pneumophila Philadelphia-1 strains Lp1-SVir and Lp02, the rate of pellet formation varies proportionally to the bacteria-to-ciliate ratio. The plaquing efficiency of isolated pellets was higher than that of free legionellae grown in vitro, an observation that could be accounted for by the fact that each pellet was actually formed by ~100 bacteria. The Tetrahymena-mediated packaging of free legionellae into infectious units with a payload of ~100 bacteria may be an important predisposing factor in the infection of human cells. It should be remembered here that the intracellular replication of L. pneumophila in cultured human macrophages is not conducive to the effective formation of mature intracellular form (MIFs). The authors specifically determined whether pelleted legionellae were more resistant to desiccation than free legionellae. The wells containing the dried legionellae were seeded with L929 cells in minimal essential medium. In summary, the interaction with ciliates may predispose L. pneumophila to infect humans by packaging legionellae into respirable-size infectious units containing hundreds of MIFs, a process also associated with some environmental fitness advantages.

Genetics can be a powerful approach with which to dissect the biological relationships between pathogens and their hosts. Genetic studies have established that polymorphisms in neuronal apoptosis inhibitory protein 5 (Naip5) appear to explain the entire difference in permissiveness of B6 and A/J macrophages to Legionella replication. The authors examined whether permissiveness of mouse macrophages to Legionella growth was affected by inhibitors of the mitogen activated protein (MAP) kinases, another important class of signaling molecules in the immune system. Importantly, therefore, it was found that macrophages from B6-backcrossed caspase-1 knockout mice were also more permissive for Legionella growth than were wild-type B6 macrophages. Currently, Legionella flagellin is detected by B6 macrophages in a manner dependent on Naip (and possibly Ipaf ), leading to caspase-1 activation, rapid cell death, and nonpermissiveness for Legionella growth. Whether Naip or Ipaf is a direct intracellular receptor for flagellin will likely be difficult to establish convincingly, as even the much more thoroughly characterized Toll and Nod proteins have not been unequivocally demonstrated to bind directly to their putative ligands. It has nevertheless been extremely satisfying to see how the concerted application of mouse and bacterial genetics has led to several insights into the nature of innate macrophage resistance to Legionella.

Genetic analysis in the mouse has been used to identify host genes and proteins that play important roles in natural defences against a broad range of infectious diseases. In mouse, susceptibility to infection with Legionella pneumophila is genetically controlled. Intracellular pathogens have evolved different strategies to inhibit or escape the bacteriostatic or bactericidal defense mechanisms of host macrophages. Electron microscopy studies indicated that Legionella-containing phagosome (LCP) morphology diverges from classic phagosomes: within the first 5 min of infection, LCPs become surrounded by host vesicles; 15 min postinfection, the thickness of the phagosomal membrane resembles that of the endoplasmic reticulum (ER), and 6 h after infection, ribosomes are found attached to the cytoplasmic face of LCPs. Nontransgenic permissive A/J peritoneal and bone marrow-derived macrophages allowed massive L. pneumophila replication over a 72-h infection period (2.8 ± 0.3 and 1.6 ± 0.4 log CFU respectively) compared to macrophages from Birc1e/Naip5-resistant transgenic animals (0.74 ± 0.18 and -0.52 ± 0.26 log CFU respectively). It has been proposed that, like other members of the NLR family, Birc1e/Naip5 may function as an intracellular sensor of L. pneumophila products and may trigger immune responses through inflammatory caspases.

This chapter focuses on identifying the loci responsible for regulating Legionella pneumophila replication in the lungs, by comparing susceptible B10.A/Niid with other resistant inbred strains. First, the susceptibility of B10.A/Niid mice might be a general response against the L. pneumophila species. Second, in order to identify the locus that is responsible for the susceptibility phenotype on the major histocompatibility complex (MHC) region of the H-2a haplotype, the authors analyzed the influence of the MHC haplotypes on L. pneumophila infection by using three types of B10.A intra-MHC-congenic mice, i.e., B10.A (2R), B10.A (3R), and B10.A (5R), as well as the parental mice, B10.A/SgSnJ, obtained from the Jackson laboratory. Third, to identify the responsible locus, the authors performed a whole-genome single-point linkage analysis on 280 (AKR/NxB10.A/Niid)F2 mice using 81 simple sequence-length markers that are polymorphic for B10.A/Niid and AKR/N. On the basis of this analysis, a L. pneumophila susceptibility locus with significant linkage was mapped between D13Die26 and D13Mit287 (mapped position 54 to 57 cM) on chromosome 13 (P<0.00001); no additional significant linkages were observed. Finally, the authors compared the replication in thioglycolate-elicited macrophages obtained from B10.A/Niid mice and the B10.A/Niid- Lgn1 s congenic mouse strain that was constructed by introgressively backcrossing (B10.A/ NiidxA/J) F1 mice to B10.A/Niid mice.

Legionella pneumophila has earned a reputation as a public scourge, and rightly so. With the goal of identifying the molecular basis of L. pneumophila pathogenesis, investigators quickly turned to animal models. By exploiting the observation that susceptibility to L. pneumophila behaves as a single Mendelian trait, the Dietrich and Gross laboratories identified naip5 as a critical determinant of mouse resistance to this pathogen. Based on these and other pioneering pathology and immunology studies in the 1980s and 1990s, workers in the field today continue to depend on small animal models to investigate the host-pathogen interactions that govern the outcome of L. pneumophila infection. Flagellar-based motility is mediated by a sophisticated organelle whose components are strictly regulated and whose assembly is carefully orchestrated. Extracellular flagellin is recognized by a receptor on the surface of eukaryotic cells, Toll-like receptor 5 ( TLR-5; 17). Knowing that bacterial flagellins trigger a rapid, proinflammatory innate immune response, the authors postulated that lysis of macrophages by motile L. pneumophila was not a pathogen tactic, but rather a host defense. L. pneumophila can induce apoptosis in numerous cell types, including human peripheral blood monocytes, human monocyte, epithelial and T cell lines, and mouse alveolar macrophages and epithelial cells. If the macrophage response to flagellin is indeed a host mechanism to combat infection, the authors postulated that L. pneumophila that lack flagellin would escape the Naip5.

The innate immune system has an important function in activation and shaping of the adaptive immune response through the induction of costimulatory molecules and cytokines. The toll-like receptors (TLRs), which have recently been characterized as receptors of innate immunity, are one of the most important pattern recognition receptor families. TLRs recognize various classes of pathogen-associated molecular pattern. A 19-kDa Legionella peptidoglycan-associated lipopro-tein (PAL) is a species-common antigen that successfully induces the PAL-specific immune responses in mice delivered with plasmid DNA3-PAL. The purpose of this study was to determine whether peptidoglycan-associated lipoprotein (PAL) can activate TLRs involved in innate immunity that might be linked to development of specific immune responses. Recombinant-PAL (rPAL) of Legionella pneumophila serogroup 1 was prepared. It also showed that r-PAL-induced cytokine production was inhibited significantly in the presence of TLR2-blocking antibody. These results indicate that the Legionella PAL is likely to activate TLR2-mediated signaling in murine macrophages, and they have important implications for the development of immune responses by Legionella surface antigens.

In this chapter, primary murine (BALB/c) bone marrow-derived dendritic cells (DCs), a phagocytic monocytic cell essential for innate immunity to intracellular microorganisms like Legionella, were infected in vitro with the bacteria. In related studies, it was recently reported that epigallocatechin gallate (EGCG) inhibited interleukin-12 (IL-12) production by lipopolysaccaride-treated DCs. EGCG has also been reported to decrease LPS-induced TNF- α production in a dose-dependent manner in the murine macrophage cell line RAW 264.7 and to similarly inhibit LPS-induced TNF- α production in elicited BALB/c mouse peritoneal macrophages, effects attributed in part through blocking NF- κ activation. In cultured human peripheral blood mononuclear cells, EGCG was also reported to stimulate production of TNF- α. All of these reports suggest that cytokine modulatory effects of catechins have varied effects depending upon the immune cell subpopulation studied. However, it is likely that concentrations of catechins at tissue sites are higher than in the blood. The short half-life of catechins in vivo might be overcome, however, by repeated administration, which is feasible given the reported low toxicity of this catechin and given that even high doses, such as 1,600 mg, are well tolerated by human subjects. In conclusion, authors present evidence that EGCG can inhibit IL-12 production while enhancing TNF- α production in bone marrow-derived DCs infected with Legionella.

In this chapter, the authors aimed to characterize the cascade of flaA regulation to obtain more information about mechanisms of gene regulation and about virulence of Legionella pneumophila. To analyze the cascade of flagellar gene regulation, the authors identified, cloned and characterized the regulators which may be involved in flaA regulation. Specific mutants of the identified regulators were generated and analyzed for flagellar gene expression. The authors also showed, that the sigma-54 activator protein FleQ is the master regulator of the flagellar regulon. Recently, the ability to analyze flagellar gene expression by transcriptome analysis using wild-type L. pneumophila strain Paris and the isogenic fliA mutant strain was developed. The regulation of the flagellum is linked to the expression of the virulent phenotype of Legionella. Phenotypic characteristics which are associated with the virulence of L. pneumophila have the ability to lyse human erythrocytes, to infect host cells, and to replicate inside host cells. It was shown recently that mutations in Toll-like receptor 5, which mediates the flagellin-dependant recognition of bacteria by the innate immune system in humans, is associated with the incidence of L. pneumophila infection. These results together indicate that various phenotypes associated with the flagellar regulon are of importance for virulence of L. pneumophila, demonstrating the importance of research into the flagellar regulon.

In a search for Legionella pneumophila Philadelphia proteins with sequence similarity to members of the LuxR family of transcriptional regulators, three novel proteins were identified. These were designated LpnR1 (lpg 2557), LpnR2 (lpg 1946), and LpnR3 (lpg 1448), and although these proteins were not quorum-sensing related, they act as transcriptional regulators. For the malate dehydrogenase gene (maeA), a semiquantitative RT-PCR analysis was performed to check transcriptional control by LpnR3. Total RNA was isolated from L. pneumophila, and L. pneumophila ΔlpnR3 was grown to exponential and stationary growth phase in culture broth. RT-PCR with maeA-specific primers (5' -atggatccggcaatcaaagtaac-3' and 5' -atggatcctagaacatatggttc-3') was performed on 100 ng of total RNA. Activation of transcription of two other genes, i.e., the triacyl glycerol lipase and the acetoacetate decarboxylase genes, on the other hand, is dependent on LpnR3, but rather indirectly, or requires an additional factor.

Hsp60 mediates invasion of HeLa cells, blocks organelle trafficking, and is the most abundant protein secreted into the phagosomes of host cells throughout the course of intracellular multiplication. In Escherichia coli the heat shock response is controlled by sigma factor 32 (RpoH) encoded by rpoH. Moreover, we do not know to what extent RpoH is either expressed or required for gene expression. In a study the authors focused on controlling levels of Hsp60 production by modulating expression of rpoH in order to evaluate the role of heat-shock protein (Hsp60) and the heat shock response in pathogenesis. The authors developed three strategies to alter rpoH expression: (i) construct an rpoH knockout mutant by allelic replacement; (ii) replace the endogenous rpoH promoter with an isopropyl- β-D-thio-galactopyranoside (IPTG)-inducible promoter; and (iii) to knock down levels of RpoH and HtpAB with antisense RNA. Either the rpoH gene is located in a dead zone for recombination or more likely the introduction of strong promoters into this locus causes constitutive expression of downstream genes associated with cell division (ftsYEX). Preliminary data presented in a report in this chapter suggest that antisense can be used to knock down protein expression levels in L. pneumophila.

This chapter tests the feasibility of replacing the sacB gene with another suicide gene, rdxA, from Helicobacter pylori, which encodes an NADPH oxygen insensitive nitroreductase. The authors show that the improved suicide vector pKBOXR (rdxA) and use of metronidazole in counterselection is superior to the sacB in Legionella pneumophila. Helicobacter pylori rdxA encodes an oxygen insensitive NADPH nitroreductase which exhibits high substrate specificity for the prodrug metronidazole, which is rapidly reduced to DNA damaging adducts of hydroxylamine, a bactericidal agent, thereby acting as a potential and novel counterselectable marker To create a gene knockout construct, flanking regions of ~500 bp of the targeted gene were individually PCR amplified and ligated into the multiple cloning site of an appropriate high-copy cloning vector (i.e., pBluescript, pUC19), after which an antibiotic cassette (kanamycin or gentamicin) was inserted between the two flanking regions. To date, utilization of the novel allelic exchange vector pKBOXR in our laboratory has generated several successful chromosomal gene knockouts of ahpC1, ahpC2D, ihfA, and ihfB in L. pneumophila Lp02. The pKBOXR vector appears to be more efficient than pBOC20 in generating knockout gene mutant strains, with the additional benefit that L. pneumophila is not subjected to osmotic stress during the counterselectable process.

To understand the genetic basis of the ability of Legionella to adapt to a wide variety of different environmental settings, the authors have initiated a long-term genome-wide gene expression profiling study of Legionella pneumophila wild-type strains Philadelphia 1 and JR32 and several Legionella mutants during their growth in rich axenic media as well as at different stages of Acanthamoeba infection using whole-genome microarrays. The gene expression measurements for key operons were independently confirmed by real-time PCR. Gene expression profiles of wild type and rpoS mutants were compared during growth in rich axenic media. About 30% of the bacterial genome was found to be expressed under any single condition tested in the axenic growth experiments. Infection was initiated by centrifugation of the bacteria onto the adherent amoebae. About 500 genes were found to be expressed during the internalization and early stages of infection; 78 of them were also expressed during growth in liquid culture. This chapter provides a comparison of expressed gene complements between axenic growth and infection and some gene ontology (GO) functional gene categories expressed during the Acanthamoeba infection. Further studies should provide a more detailed picture of coordinated gene expression during Legionella infection as well as the molecular events that accompany it.

A whole-genome microarray was constructed which contained 3,823 gene-specific oligonucleotides representing every gene predicted in the Legionella pneumophila strain Paris genome plus each loci specific to strains Lens (302 genes) and Philadelphia. In this short article, some results of the in vivo transcriptional program of the three L. pneumophila strains during their life cycles in the natural eukaryotic host Acanthamoeba castellanii will be presented. A biphasic life cycle is clearly evident from the transcriptional data, with sets of replicative phase (RP) genes upregulated at the earlier time point and transmissive phase (TP) genes upregulated at the later time points. The global transcriptional programs of all three L. pneumophila strains were very similar, indicating that common regulatory mechanisms govern their life cycles. When switching from a fast-growing bacterium to a motile, highly infectious microbe, the pathogen fundamentally alters its transcriptional program. Among the most significantly upregulated genes in the TP, the authors identified four two-component systems, six LysR-family members, four additional transcriptional regulators [CpxR, PaiA-, MoxR and TetR-like], and two integration host factors. The findings provide a valuable resource and conceptual framework to increase knowledge of the Legionella life cycle and the host functions exploited during infection of amoebae and macrophages.

Legionella pneumophila is found in diverse ecological niches such as axenic cultures, biofilms with other microbes, and intracellular vacuoles of protozoa and human cells. This may have provided ample opportunities for the organisms to rearrange their genomes and to accumulate genes that help them survive in these different environments. Legionella's ability to survive in harsh conditions such as in plumbing systems treated with potent biocides may require additional genes necessary for detoxification. The needed gene pool can be assembled via gene family expansions and horizontal gene transfer from other organisms. Genome comparisons of the Philadelphia 1 strain and those of Paris and Lens, indicating positions of the large-scale genome rearrangements. Phylogenetic and gene composition analysis indicate that different L. pneumophila strains display distinct acquisition histories for these gene subsets, suggesting gene rearrangements as well as repeated horizontal gene transfer events. Several dozen eukaryotic gene homologs with distinct phenotypes which appear to be acquired by gene transfer were identified. Overall, the genome of L. pneumophila reveals features supporting genome rearrangement events of different scales. Given its intracellular existence, it is not surprising that examination of the genome sequence offers evidence that L. pneumophila can be an active participant in horizontal gene transfer. As more strains of L. pneumophila are sequenced, along with other species of Legionella, much more will be learned about their evolutionary origins, lifestyle, and potential for pathogenicity.

In order to explore the genetic diversity in a large number of strains, a Legionella biodiversity array was designed containing probes (internal PCR products) specific for all those genes that are variable among the three strains as well as of conserved ones that have known or putative implication in virulence. The gene content of 180 Legionella isolates, 165 Legionella pneumophila strains, and 25 representatives of Legionella species other than L. pneumophila isolated from the environment during epidemics or from sporadic cases were analyzed by DNA-DNA hybridization using the Legionella biodiversity array. L. pneumophila serogroup 1 is largely overrepresented among human legionellosis cases, suggesting that virulence differences exist among different L. pneumophila serogroups and other Legionella species. In order to address this question and to try to find a possible link between virulence differences and a specific pattern of virulence genes, the authors analyzed the distribution of these factors among 180 strains. This study also reports for the first time the distribution of the eukaryotic-like genes identified during sequence analysis among L. pneumophila. In order to better understand plasticity and evolution in L. pneumophila the authors used hierarchical clustering to analyze the hybridization patterns obtained. Despite the high genome plasticity observed, hierarchical clustering of the hybridization results of the genes selected for the Legionella biodiversity array correlates with results of rpoB sequence typing and therefore with phylogeny. The Legionella biodiversity array allowed subdivision of the otherwise indistinguishable group of Paris strains.

The current trend in publishing new Legionella species is to utilize culture and phenotypic characteristics, serology, cellular fatty acids and isoprenoid quinone profiles, DNA relatedness, and a molecular phylogenetic analysis targeting usually more than one gene. To investigate the greater use of sequence-based phylogenetic analyses in lieu of serology, fatty acid and isoprenoid quinone profiles, and even perhaps DNA relatedness, 56 strains representing 33 potentially novel species were phylogenetically assessed with 6 gene targets. First, following a phylo-genetic analysis based on sequence from the mip gene, 49 strains representing 30 potentially novel species were selected because their genetic distance from recognized species was greater than that between recognized species. Second, DNA studies on two strains indicated that they were novel, even though their mip sequence-derived genetic distance was closer to a recognized species than that between recognized species. Last, the remaining five strains were selected because they were genetically related to either of these latter two strains, based on their macrophage infectivity potentiator (mip) sequence-derived genetic distance. The six gene targets used in this analysis, 16S rRNA gene, the mip gene , the RNA polymerase β-subunit (rpoB), the RNase P RNA gene (rnpB), and the DNA gyrase A subunit (gyrA), together with the zinc metallo-protease (proA, also known as mspA) gene. In the interim, the use of a combined approach using the sequenced- based phylogeny to identify “nearest neighbor” species will greatly simplify the use of total DNA relatedness studies.

Understanding the ecological diversity within Legionella may help in identifying all environmental reservoirs that can sustain the different potential Legionella pathogens. Here the authors apply the community phylogeny method to demarcate the ecologically distinct populations within the genus Legionella, based on the sequence of the macrophage infectivity potentiator (mip) gene from 496 Legionella isolates, obtained from Rodney M. Ratcliff, representing all characterized species within Legionella, as well as many uncharacterized groups. The community phylogeny analysis evaluates different sets of parameter values for their likelihood of yielding a phylogeny consistent with the observed clade sequence diversity of the group. This chapter provides a broad overview of the maximum likelihood approach for finding the trio of parameter values best fitting the observed curve of clade sequence diversity for Legionella, given the number of sequences sampled. The community phylogeny analysis provides an objective and theory-based method for identifying sequence clusters that are likely to correspond to ecotypes with a long history of coexistence. When a putative ecotype identified by community phylogeny is shown to be ecologically distinct in nature, it will have demonstrated the attributes of species. The authors hope that a more precise taxonomy, giving names to all the long-coexisting and ecologically distinct groups within a named species, will allow ecologists and epidemiologists to better predict the properties of newly isolated strains.

This chapter explores issues related to Legionella genomics. A critical breakthrough in this area recently occurred with the determination of the genome sequences of three clinical, serogroup 1 isolates of Legionella pneumophila subsp. pneumophila. The three strains are L. pneumophila Paris, Lens and Philadelphia-1. The genome sequences obtained from these three strains have already proven to be incredibly useful as a genetic tool and they have provided numerous new and valuable insights into the biology of L. pneumophila. Analysis of the completed genome sequences of L. pneumophila strains Paris, Lens and Philadelphia-1 revealed that each consists of 3.3 to 3.5 million base pairs, encodes approximately 3,000 genes, and has an average G+C content of 38%. Since the three analyzed strains belong to the same species and same serogroup, this level of diversity is remarkable. Several additional insights that have been derived from genomic analyses include the marked plasticity of the genomes, the evidence of large-scale rearrangements, and the presence of a large number of eukaryotic-like proteins. The availability of a large number of different genome sequences of Legionella will clearly benefit the community and pave the way for in-depth comparative genomics. In summary, the field of Legionella genomics has only begun and promises a bright future.

The majority of Legionella/biofilm studies that have been conducted employ naturally occurring microbial communities. These have evaluated the effect of temperature and surface materials on the growth of Legionella pneumophila as well as the effect of biocides on planktonic and sessile bacteria. This study indicated that L. pneumophila may persist in biofilms in the absence of amoebae, but in the model, the amoebae were required for multiplication of the bacteria. A number of bacterial cell structures and factors have been shown to be critical for biofilm formation. These include flagella, quorum-sensing factors, polysaccharides, and pili. The biofilm model described by Murga et al. was used to compare these strains in the presence of heterotrophic bacteria (Pseudomonas aeruginosa, Klebsiella pneumoniae, and Flavobacterium spp.) with and without the amoeba Hartmannella vermiformis. Both strain BS100 and NU243 show a significant decrease in retention in the biofilm in the absence of amoebae. The bacteria were harvested from buffered charcoal yeast extract agar plates and inoculated in various broth media. Tap water was pumped into the reactor, creating the flow and eventually diluting out the broth. The total concentration of L. pneumophila (viable count) shows an indistinguishable response to challenge with live or killed P. aeruginosa or Escherichia coli. The addition of host amoeba cells allows L. pneumophila to persist in the sessile and planktonic fractions of the biofilm reactor for extended periods of time.

The haploid social amoeba Dictyostelium discoideum is a well-defined model organism which is amenable to diverse biochemical, cell biological, and molecular genetic approaches. The similarities between the Dictyostelium and mammalian cells extend to membrane trafficking, endocytic transit and sorting events. Therefore, Dictyostelium opens the exciting possibility of investigating and manipulating both sides of the Legionella-host interaction. The intrinsic features of Dictyostelium and a set of well-established molecular tools enabled the authors to prove that Dictyostelium is a representative cellular model for Legionella infection. The research areas presented in this chapter include the analysis of the transcriptional host cell response to Legionella infection and the use of custom tailored Dictyostelium mutant cells to identify determinants of susceptibility and resistance. In order to describe the transcriptional host cell response to Legionella infection, the authors employed DNA-microarrays which carried approximately half of the Dictyostelium genes. Mammalian Nramp1 decorates endosomes, phagosomes, and postlysosomal vesicles of macrophages, and in D. discoideum it has been shown that Nramp1 depletes iron from the phagolysosome in an ATP-dependent process. Three approaches were used to analyze the complex cross-talk between Legionella and Dictyostelium. First, Dictyostelium microarrays were used to study the systemic host response to Legionella infection. Second, fluorescently tagged host cell factors were used to get new insights into the temporal and dynamic interactions between the pathogen and the host. Third, custom tailored mutants and specific inhibitors were used to functionally analyze infection relevant determinants of host susceptibility and resistance.

Recently the authors demonstrated through several uptake experiments that some bacterial species do not act as competitors on Legionella pneumophila uptake by Acanthamoeba castellanii, but are able to influence intracellular Legionella replication. Biofilm experiments were performed at 35°C in a rotating annular reactor. Experiments were repeated at least three times and always consisted of two parts: during the first part, without present A. castellanii (days 0 to 36), the authors investigated if L. pneumophila were able to attach and eventually grow in the biofilm. During the second part, A. castellanii trophozoites were added to the system and the evolution of biofilm-associated L. pneumophila was investigated. Addition of A. castellanii on day 36 resulted within 48 h in a significant decrease of biofilm-associated non-Legionella bacteria due to predation, while biofilm-associated L. pneumophila increased with 1.5 log units due to intracellular replication in the present A. castellanii trophozoites. The association between Legionnaires’ disease and the presence of high numbers of human pathogenic L. pneumophila bacteria in man-made water supplies has already been confirmed frequently. Therefore, the results stress the importance of further obtaining data concerning interactions between L. pneumophila and potential amoeba host populations associated with biofilms.

This chapter analyzes biofilm formation of Legionella pneumophila in rich medium which supports extracellular proliferation of the bacteria. To investigate biofilm formation of Legionella under static conditions, a 1:1 mixture of L. pneumophila labeled with the fluorescent proteins EGFP or DsRed-Express was inoculated in a glass-bottom 35-mm petri dish. Biofilm formation was analyzed with an inverted confocal microscope (Axiovert 200M, 100 X oil objective Plan Neofluar; Zeiss). Biofilm formation of L. pneumophila in rich medium was also quantified by crystal violet staining in upright polystyrene microtiter plates or on polystyrene pins of "inverse" lids under static (no medium exchange) or quasi-static conditions (medium replaced twice a day). The L. pneumophila fliA mutant strain reproducibly accumulated 30% less biomass within 5 days, demonstrating that bacterial factors contribute to biofilm formation. It is noteworthy that mutants lacking rpoS or letA, both of which are required for the expression of transmissive (virulence) traits of L. pneumophila, were not affected in biofilm formation. A continuous-flow chamber system was set up to further test the hypothesis that plank-tonic cells are crucial for biofilm formation on surfaces and to noninvasively monitor in real-time the adherence and biofilm formation of L. pneumophila under dynamic flow conditions.

The association of legionellae with microorganisms in the aquatic environment has been documented earlier. This relationship includes its cohabitation with protozoa, algae, and prokaryotes within biofilms. Potential candidates are the intercellular signals produced by bacterial species commonly isolated from biofilm communities. Acyl-homoserine lactones (AHSL) and furanosyl borate diesters (AI-2) are two prokaryotic signaling compounds that have been isolated from biofilms. The objective of this chapter is to evaluate the capacity of Legionella pneumophila to respond to interspecific signals. From evaluation of protein expression after exposure to various bacterial supernatants, multiple proteins were identified as products of L. pneumophila's response to heterospecific signals. Spots of interest were manually excised from two-dimensional electrophoresis gels, washed, and subjected to in situ digestion with trypsin, and the resultant peptides were desalted and concentrated with C18ZipTips (Millipore). AI-2 has been detected in numerous prokaryotes, both gram-positive and gram-negative, and is commonly referred to as a universal communication signal among microorganisms.

Humans have frequently used plants to treat common infectious diseases, and some of these traditional medicines are still part of the habitual treatment of various maladies. This chapter presents a study that investigated antibacterial activity of some lichens against Legionella pneumophila serogroup (sg) 1 and L. pneumophila sg 2 to 14 strains. The inhibition effects of lichen aqueous extract (AE) and ethanolic extract (EE) against L. pneumophila sg 1 ATCC 33152 (A), sg 1 (C, E), and sg 2 to 14 (B, D) were performed according to the agar well diffusion method by using buffered charcoal yeast extract agar. It was found that the antimicrobial activity of the AE was greater than 70% of the activity of EE. Lichens now constitute important sources for prospecting for new bioactive molecules, either by the direct use of their secondary metabolites or by employing their biyosynthetic or semisynthetically derived compounds, which are produced with the aim of attaining higher effectiveness, improved absorption, or even decreased toxicity. The establishment disinfection procedure may not prevent the occurrence of Legionella in the water system. It is known that Legionnaires’ disease frequently appears because of contaminated water systems and it can be treated by antibiotics.

The chapter describes the partial characterization of a new bacteriocin, produced by Staphylococcus warneri RK, active against Legionella pneumophila. A bacterial strain, isolated from the environment, displayed an anti-Legionella activity by a spot on lawn assay. Various bacterial strains, gram-positive and gram-negative, were used as target organisms to test their sensitivity to S. warneri RK by a spot on lawn assay. The anti-Legionella molecule was then purified from the raw extract during a three-step purification procedure. First, the raw extract was diluted (50:50) with sodium acetate buffer (20 mM, pH 5) and applied to a weak cation exchange chromatography column (Hiprep 16/10 Carboxy-Methyl FF, Amersham Biosciences). Second, the latter fraction was applied onto a solid phase extraction C18 cartridge (Sep-pak plus, Waters), washed successively with 5 ml of 0, 10, 20, 30, 40, and 80% acetonitrile containing 0.1% trifluoroacetic acid. Third, the active fraction (80% acetonitrile) was lyophilized, solubilized with 1 ml of 40% acetonitrile, and injected on a Kromasil C8 reverse-phase HPLC analytical column (5μm, 100 Å, 4.6X250 mm, A.I.T.). The anti-Legionella activity has been eluted, during the last step, at 100% acetonitrile, indicating that the corresponding peptide is highly hydrophobic. This anti-Legionella-active fraction was subjected to mass spectrometry analysis, giving two molecular masses of 2613.8 Da and 2449.4 Da. This result shows that two peptides were present in the fraction, one of which likely corresponded to the anti-Legionella bacteriocin. The anti-Legionella molecule is therefore an original peptide.

The diversity of sources and degree of bacterial contamination limit the use of a single method for the recovery of legionellae from environmental samples. In addition, direct fluorescence antibody (DFA) stain and coculture with axenic amoeba are important methods to detect legionellae at low concentrations. In this case, the procedures can be applied in environments with nutritional restriction and submitted to biocide treatment or extreme ranges of pH and temperature. This chapter talks about techniques used to screen the occurrence and diversity of Legionella strains from the aquatic environment in Brazil. Water samples were collected from the main reservoirs of Sao Paulo, dental units, and cooling towers and boilers from hospitals and industries. The nine isolates obtained in a study were identified as Legionella pneumophila through the sequencing and comparison of the mip gene sequences deposited in the GenBank database. The phylogenetic tree obtained from the sequences of nine isolates from this study showed the presence of three distinct clusters and the diversity of the L. pneumophila strains (BR strains). In addition, Sanden and collaborators reported that coculture of water samples with free-living amoebae for several days increased the recovery of Legionella spp., which was related to proliferation of amoebae in the samples. Detection of legionellae in environmental samples by using PCR and gene probes in conjunction with DFA staining was previously proposed.

The use of culture-independent molecular screening techniques has allowed the description of Legionella as a member of the microbial community structure of extreme environments such as polar regions. The authors used traditional culture and coculture methods to isolate Legionella spp. and independent-culture methods to verify and compare the diversity of the Legionellaceae population in the microbial communities of two Antarctic lakes. Water samples (5 liters) were collected from North and South Lakes, near the Brazilian Scientific Station Comandante Ferraz, located in the Keller Peninsula, King George Island, Antarctica. Colonies were obtained only from the North Lake samples. Selected colonies were used for total genomic DNA extraction. Denaturing gradient gel electrophoresis (DGGE) analysis from the water samples showed 28 to 30 bands representing abundant and similar phylotypes between the bacterial structures of both lakes. The DGGE technique demonstrated that Legionella is a significant member of the community in both lakes. Such findings could be related to the presence of ferric ions determined in the studied areas of Keller Peninsula. The isolation of Legionella from these Antarctic lakes will allow future studies in cold-resistance mechanisms of mesophilic bacteria in polar environments.

Legionellae are a predominantly aquatic family of bacteria commonly associated with water systems in the built environment. There are substantially fewer reports of the presence and survival of Legionella in the marine environment. Researchers have investigated the survival of Legionella pneumophila serogroup (sg) 3 in saline solutions and seawater in laboratory microcosms. The results of these investigations showed that both increasing salinity and temperature reduced survival of the organisms. The limitations of the investigations were that heat sterilization of the seawater would have two major effects: (i) chemical alteration of the nature of the seawater and (ii) removal of other microflora. Legionella grow in conjunction with other microorganisms, and this has been clearly demonstrated in nonsterile tap water cultures. Legionella were cultured by direct inoculation of 0.1 ml of sample onto GVPC agar plates followed by incubation at 35°C for 7 days. Representative colony morphologies were enumerated and presumptively confirmed as Legionella species by subculture onto blood agar and buffered charcoal yeast extract. The survival of L. pneumophila in nonsterile seawater decreased with increasing salinity. These results suggested that the organism was not capable of multiplication in the seawater. The protective role of amoebae and biofilm in Legionella survival and multiplication cannot be disregarded in this context. Terrestrial runoff and transient survival of Legionella is a plausible explanation for previous reports of relatively high concentrations of Legionella in coastal waters.

Many studies describe Legionella's ability to survive and to multiply in about 30 species of amoebae, especially Acanthamoeba, Naegleria, and Hartmanella, and in two species of ciliates. This chapter describes isolation of Legionella from hospital water supplies and from air conditioning systems during periodic controls sometimes performed after water disinfection. The culture of amoebae was also performed, and the interactions with legionellae were studied. Legionella was isolated from 251 of 471 (53.3%) hospital water samples (239 hot and 12 cold), corresponding to 11 of 12 (91.7%) hospitals. Amoebae alone were isolated from 27 of 471 (5.7%) samples, while Legionella alone were isolated from 196 of 471 (41.6%) samples. The distribution of Legionella and amoebae in all water samples in relation to temperature have been discussed in the chapter. The results of this study confirmed those observed by other authors reporting similar data for Legionella isolation (55%). In conclusion, the Legionella isolation rate was higher than that of amoebae from hospital water supplies, but amoebae were found in higher percents from air conditioning systems (64%, the despite limited number of samples examined) where Legionella was absent.

This chapter focuses on optimizing and applying cultivation-independent methods to obtain information about the presence and identity of free-living protozoa in river water (RW), treated water (TW), and tap water (Tap) in The Netherlands. This is the first description of protozoal diversity in freshwater by using molecular techniques. The PCR products were purified and cloned into Escherichia coli JM109 by using the Promega pGEM-T easy vectors system. Free-living protozoa represented a major part of the total eukaryotic community. The clone libraries show that the eukaryotic community in surface water, treated water, and tap water consisted, respectively, of 44, 62 and 20% of free-living protozoa. Blast analysis of the clones confirmed the presence of a highly diverse group of free-living protozoa in all three water types. Further optimization of terminal restriction fragment length polymorphism (T-RFLP) assays is needed to achieve better resolution, distributing the T-RFs of dominant free-living protozoa over the T-RFLP pattern from 50 to 500 bp. Clone libraries are a powerful tool to determine the diversity and identity of eukaryotic communities, but this technique is time-consuming. The application of real-time PCR assays for selected protozoa genera and cultivation methods will further extend one's knowledge of free-living protozoa serving as hosts for L. pneumophila in tap water installations.

In recent years, many public bathhouses introduced bathing water circulating systems for extended use, in which a sand filtration unit was installed. However, this resulted in several large scale outbreaks of legionellosis due to the microbiologically insufficient maintenance of the bathing facilities. Occurrence of Legionella in bathing water circulating systems, appears to be common and is a serious public health concern in Japan. The authors constructed a life-size model plant of a bathing water circulating system for the simulation experiment. These experiments are aimed at monitoring changes in the microbial constituents, especially a possible occurrence of Legionella in a bathing water circulating system, and developing preventive measures and intervention strategies. As a result of experiment 1, Legionella was detected in both the bathing water and the filter water at concentrations of 6.6 X 102 CFU/100 ml on the 3rd day after residual chlorine disappeared. The number of amoebae in the filter water fluctuated and amounted to 12 cells/ml at the end of the experiment. In the experiments, it was clearly demonstrated that Legionella occurred in the bathing water circulating system within a short period in a sequential manner of microbial growth. Namely, concentration of organic matter (dirt) in the bathing water can be monitored as the KMnO4 consumption value increased in correlation to the number of bathers. The deposited dirt allows bacteria to rapidly undergo multiplication in the bathing water, which consequently supports the occurrence of a large number of host amoebae.

Colonization of cooling towers (CTs) by Legionella is frequent and has been implicated in many community outbreaks of Legionnaires´ disease (LD). This chapter describes the fluctuation in Legionella counts in CTs over a 1-year period. Ten colonies of each positive culture were selected and characterized by nutritional requirements and by the immunoglutination latex assay test (OXOID). Legionella pneumophila was isolated in 13 of 15 CTs (86.6%). All positive CTs were colonized by L. pneumophila serogroup 1. No seasonal influence was observed with respect to variations in L. pneumophila counts. All of the systems would not have been considered a risk at some sample time during the same season.

Legionella pneumophila has frequently been found in cooling towers (CT) with different DNA subtypes being implicated in community-acquired Legionella infections. The aims of this chapter were to investigate the genotypic variability of L. pneumophila in cooling towers and determine the persistence of the DNA subtypes over time in two different areas. In six cooling towers the authors observed a single DNA subtype, in five, two DNA subtypes, in three cooling towers, three DNA subtypes and in two cooling towers the authors observed fewer than four DNA subtypes. In three of the seven cooling towers the authors observed a single DNA subtype, in two the authors found two DNA subtypes, and in another two, three DNA subtypes were reported. When authors studied the persistence of the DNA subtypes it was found that in 19 of 23 cooling towers (82%) the same DNA subtype was recovered after at least 6 months of follow-up. These results demonstrate the great genotypic variability of Legionella in the cooling towers.

Legionellae are ubiquitous in the environment. Molecular techniques have been applied, but they too have inherent limitations and rarely provide quantitative data. Consequently, there is a real need for a rapid and direct detection method. Each DNA probe has its own specific requirements, and the hybridization procedure is highly sensitive to changes in temperature, pH, and ionic conditions. This causes problems when working with environmental samples. An alternative is to use peptide nucleic acid (PNA) probes. PNAs are synthetic molecules in which the sugar phosphate backbone has been replaced by 2-aminoethyl-glycine. They exhibit sequence-specific recognition of both DNA and RNA, obeying Watson-Crick hydrogen bonding rules. A study has been undertaken to directly compare the binding efficiency of two published DNA probes against two newly designed PNA probes. The DNA probes tested were LEG226 and LegPNE1, which target all species of Legionella and only Legionella pneumophila, respectively. The thermal stability and chemical resistance of PNA probes make them a valuable tool for use with complex environmental samples and will permit the simultaneous labeling of several target species. It has been shown previously that Legionellae are widespread in potable water systems, but there is no information on population sizes and the proportion associated with biofilm compared to the bulk water. As outbreaks of Legionnaires’ disease continue, it is imperative that the authors increase their understanding of the distribution and prevalence of this serious human pathogen.

This chapter deals about the direct detection of Legionella from hospital water samples by commercial real-time (RT)-PCR system, that offers rapid results and reduced risk of cross-contamination. The preliminary results of 61 water samples from 6 hospitals and 15 water samples from air conditioning systems are presented and compared with those of the isolation method. Water samples were concentrated by filtration through cellulose acetate membrane filter and resuspended in 10 ml of the same water. The principle steps were cell lysis with proteinase K, purification on spin column and elution. Culture was positive in 28 (45.9%) hospital water samples, and the strains were identified as Legionella pneumophila serogroup 1, 3, 6, and 2-14 and Legionella spp. The results of qualitative RT-PCR were as follows: Legionella spp. was positive (detection limit: 133 GU/liter) in 60 (98.4%) samples and L. pneumophila was positive in 44 (72.1%). The results of culture and qualitative analysis for L. pneumophila by RT-PCR are shown. All hospitals examined showed some positive samples for culture and PCR, except for one that was culture negative and PCR positive. All water samples from air conditioning systems were culture negative, but RT-PCR was positive in 15 water samples for Legionella spp. and in 2 for L. pneumophila. In conclusion, Legionella detection by RT-PCR from environmental methods seems to be promising. However, at the moment, correlation between the method-result interpretations and public health significance need further evaluations.

Environmental monitoring of water systems for Legionella is performed for routine microbiological surveillance and outbreak investigations. The recovery of Legionella from water samples using conventional culture on selective media can be improved by sample concentration and/or sample decontamination techniques, but there is a reduction in recovery associated with these processes. A new immunoseparation method has been developed to improve the recovery of Legionella by concentrating and selectively separating Legionella spp. directly from water samples using immunomagnetic separation (IMS). Immunomagnetic separation has been used successfully for detection of other pathogens including Salmonella spp. The authors conducted the study in two phases and evaluated the IMS process alone and in combination with concentration and/or acid pretreatment in detecting Legionella spp. and Legionella pneumophila serogroup 1 (sg 1). The IMS procedure was performed using the BeadRetriever apparatus (Dynal Biotech Ltd., Preston, United Kingdom) for automated IMS using Dynabeads anti-Legionella (Dynal Biotech Ltd.). With the 11 known potable water samples, there was concordance (100% sensitivity) of culture results between the two methods, regardless of concentration or pretreatment of samples. The unconcentrated, untreated samples showed the greatest agreement in CFU/ml. Compared to conventional culture for L. pneumophila sg 1 only, the IMS method demonstrated a specificity of 98.5% (65 were negative and 2 were false-positive).

The usefulness of sampling waters for the detection of Legionella remains a point of discussion among researchers and industry for a variety of reasons. This chapter describes a novel detection system utilizing two liquid media (medium 1 and medium 2, formulations are currently under provisional patent applications) and real-time PCR that is capable of detecting viable Legionella in a variety of waters within 2 to 3 days. Over 30 strains of Legionella have been tested by this method, and the real-time PCR detection system has been validated by authors' laboratory. Real-time PCR utilizing Legionella 16S rRNA primers was performed on a Corbett Rotor-Gene 3000 utilizing SYTO9 as a DNA intercalating dye for PCR. The GMS method was challenged with 82 field samples (cooling tower, evaporative tower, and potable waters) and run in parallel to AS/NZS 3896:1998. Comparison of methods using field samples showed that the GMS method is 100% specific (no false positives) and 92% sensitive, with a single discrepant sample. The GMS method described is rapid, simple, cost-effective, less laborious than conventional culture, and easy to implement in any modern day microbiology laboratory.

Legionella bacterium is present naturally in ground water and has been described as a new pathogen growing in water distribution systems. Real-time PCR is a technology that permits us to quantify, in a rapid, sensitive, and reliable way, the genomic units of bacteria in water. The authors of this chapter have used real-time PCR to study the presence of Legionella bacteria in the environment and in water distribution networks. They have quantified Legionella spp. and L. pneumophila in natural water, water distributed in the urban network following disinfectant treatment, and waste water from domestic and industrial use. The water microbiology was studied by ATP measurements and enumeration of Legionella by culture and by real-time PCR. Thirty natural water samples were collected along the Vidourle river. Six waste water samples were collected in the three towns. PCR results were higher in natural water than in distributed water and lower than in waste water. In this study, real-time PCR shed light on the presence of Legionella species flora in natural water along Vidourle Valley watershed. Proliferating in biofilms and in amoebae, the Legionella species may be a useful indicator for the estimation of the total microbiological quality of water. ATP-metry can also be an interesting tool for the characterization of the microbiological quality of water.

This chapter investigates on the efficacy of detection of Legionella pneumophila serogroup 1 (sg 1) in cooling waters using the Binax Equate Legionella water test. The test kit offers a rapid detection method (45 min) for L. pneumophila sg 1 in environmental water samples. A high proportion of Binax positive results were also between 1.0 and 1.9 OD450. Legionella species were cultured from 22% (n = 31) of all cooling water samples. Presumptive Pseudomonas spp. were isolated from 42.9% of all cooling water samples. Two samples grew to confluence, indicating very high concentrations of Pseudomonas spp. in those samples. Presumptive P. fluorescens was isolated from 21.9% of all cooling water samples. The post-treatment results for the heat-treated sample showed that the average absorbance decreased following the treatment. A large number of false-positive Binax results were recorded whether or not Legionella spp. were cultured from the sample. Alternatively, the test also recognized heat-killed cells and false-positive results may have been due to the presence of dead L. pneumophila in the cooling systems. The results suggest that this system lacks the required specificity for routine use in cooling towers or other water systems. It seems likely that the test method would produce false-positive results in systems where disinfection was applied. False-positive results from cooling tower samples in this study may have been due to the presence of killed L. pneumophila sg 1. The results point to the necessity of indication of viability in future environmental tests for Legionella.

Metalworking fluids (MWFs) are highly susceptible to microbiological contamination from the environment due to high water content, available nutrients, and optimum growth temperature for most environmental microorganisms. Outbreaks of Pontiac fever and Legionnaires' disease were associated with MWFs at two unrelated automotive facilities. Legionella feeleii and L. pneumophila serogroup 1 (sg 1) were identified as the probable causative organisms. The objective of this chapter was to develop a selective and sensitive culture method for detection of Legionella in MWFs in order to prevent and control possible occupational health-related problems. Recovery efficiency was evaluated by spiking MWF field samples, highly contaminated (107 to 108/ml) with nonlegionella organisms with high levels (106 to 107/ml) of L. pneumophila sg 1 and L. feeleii. Several sample treatment conditions to reduce the interference of the high background population, including acidification, heat treatment, and use of antibiotics in the recovery medium, were evaluated.

This chapter relates to the microbial ecology of building water systems, aspects of Legionella virulence and dose response, and the notion of generic risk factors. Total heterotrophic plate count (TPC) evaluations can be a rapid and reliable means to augment Legionella culture for risk management in spa pools. Legionella virulence is extremely variable within and between species and serogroups. Cooling tower-associated outbreaks of disease are almost exclusively associated with Legionella pneumophila, and for the most part with L. pneumophila serogroup 1. It has been suggested that Legionella concentrations in water relate to risk, as they can be used as predictors of dose. Action directives based on the colony-forming units have been promoted as reliable risk management tools. The currently available evidence base demonstrates that health risk is variable between water systems. Factors such as total microbial load, Legionella species, disease transmission, and population exposure change dramatically with the nature of the disseminating system. Although there are common ecological determinants of Legionella growth and infection, the risks cannot rationally be generically applied. Available reports strongly suggest that good design, knowledge, and maintenance of water systems in the built environment are the most critical factors in preventing disease. An accurate and cost-effective risk assessment can only be made by considering these factors and the exposed demography.

Acute-care and long term-care facilities continue to experience cases of hospital-acquired Legionnaires’ disease. One of the major unresolved issues is whether the recommendations found in the guidelines will, if followed, result in the control and prevention of hospital-acquired Legionnaires’ disease. An evidence-based approach has been suggested as a way to resolve many of these issues. If applied to a guideline, the criteria should be that (i) the recommendations should be prospectively validated under controlled studies using a step-wise approach, (ii) the evaluation should be a prolonged observational period (>1 year) to evaluate the efficacy of the recommendations, and (iii) the recommended approach/actions should achieve the expected result-prevention of the disease through environmental control. Legionnaires’ disease is an environmentally acquired illness. One recommendation often found in guidance documents is “remove showerheads and aerators monthly for cleaning with chlorine bleach’’. The role of environmental monitoring in Legionella prevention has been a source of debate for many years. A number of disinfection methods have been used for control of Legionella in hospital water systems. These include thermal eradication (heat and flush), hyperchlorination, copper-silver ionization, point-of-use filters, and chlorine dioxide. Each of these methods has completed some of the evaluation criteria. All four steps of the evaluation criteria have been fulfilled for copper-silver ionization. The path to effective control of hospital acquired Legionnaires’ disease must be evidence based. Patients and healthcare facilities suffer when unconfirmed and untested recommendations become part of prevention guidelines.

The degree of Legionella pneumophila contamination in hospital water supplies has been shown to correlate with the incidence of nosocomial Legionnaires’ disease (LD). Within the context of the Italian Multicentric Study of Legionellosis, the authors carried out a 4-year active prospective LD surveillance program in a large university hospital in Rome. They assessed the usefulness of the hospital water monitoring program to predict the risk of nosocomial LD. A monthly decontamination procedure of the intermediate tanks is carried out, which consists of mechanically cleaning out the tanks to remove the formed organics, followed by washing out the tanks with sodium hypochlorite. Once the nosocomial LD case had occurred, epidemiological and environmental investigations were performed in the hospital ward involved. The low incidence of nosocomial LD cases in the hospital during the study period seems to be correlated to the low contamination level of L. pneumophila (20% of positive samples in 4 years) in hospital wards and to a low percentage of positive water samples per semester of surveillance. An infection control system for nosocomial LD should therefore be based on both environmental and clinical surveillance, together with the appropriate maintenance of the hospital water distribution system.

This chapter analyzes variables in the rebuilding program resulting in Legionella in the domestic hot water system (DHWS), legionella monitoring of water and air-conditioning systems, and systems implemented to control contamination. Standard methods for Legionella culture of environmental samples and identification of isolates were followed for both water samples and swabs from showerheads. The Infection Control Committee advised the clinical staff to test patients with possible nosocomial pneumonia for Legionella pneumophila sg 1 infection by urine antigen tests. Environmental culture found many sites positive for L. pneumophila sg 1 including feed water to the calorifiers and a sample taken just after tempering of the water exiting the calorifier. The peripheral contamination with L. pneumophila sg 1 was discovered within months, but when a patient was recognized as possibly being infected, extensive changes to the DHWS were carried out. A UV irradiation plant was installed near the exit of the cold water tanks to reduce further colonization.

This chapter reviews the evidence of different disinfection methods to stop outbreaks of Legionnaires’ diseases (LD). Articles describing original outbreaks of Legionella were selected for the review. Full reports of potentially relevant publications were obtained and checked for eligibility. Decisions on which trials to include were based on full text articles. The following parameters were extracted: country, year of publication, year of outbreak, duration of outbreak, source of outbreak, facility and department, number of patients involved, number of health care workers involved, mortality, disinfection methods, and success of methods. Legionella outbreaks have a significant impact on patient morbidity and mortality in hospitals. In 68% of outbreaks, the source was the hospital water system or showers. In these cases, disinfection methods focusing on the water system should be chosen. The success of these disinfection methods depends on the status of the water system and the disinfection method used. This implies that technical measurements are very important factors for termination of outbreaks, despite other effective disinfection methods such as chlorination (70.6%) and elevating temperature once per month (71.4%). The safest interventions are sterile water for patient care or point-of-use filters. Interestingly, the number of published outbreaks has decreased since the 1980s, whereas the literature about Legionella did not. Whether or not this is due to better plumbing design of hospital water systems and more successful disinfection methods remains to be proven.

Quantitative microbial risk assessment (QMRA) for Legionella can provide a means of setting risk based limits on air concentration or risk based limits on the Legionella concentrations in water. Guinea pigs provide a reasonable animal model and dose-response data on which to base human risk projections due to similarities in the course of the disease in guinea pigs and humans and in vitro uptake and replication rates of Legionella in human and guinea pig alveolar macrophages. Most mouse and rat strains appear to be relatively resistant to Legionella infection due to less compliant alveolar macrophages. A two-zone box model was used to estimate the air concentrations in CFU/m3, with stochastic input distributions using Monte Carlo simulation to yield results as probability distributions. The respective exposure distributions were fed into the dose-response model using Monte Carlo simulation. The resulting estimated risk distributions were then compared to the reported rates for the outbreaks. The guinea pig infection data adequately predict the reported subclinical (seroprevalence) rates. For the outbreaks used for model evaluation, the reported rates of disease span an order of magnitude. Thus, the QMRA may not extrapolate to other Legionella species and strains.

This chapter investigates the overall Legionella colonization of thermal water and cold and hot water of particular hotels, including the detailed identification and typing of species and serogroups. The aim was also to evaluate the risk of particular hydrotherapy procedures, to identify the source of the Philadelphia strain in the spa complex, and to propose corrective actions to minimize the risk of future Legionella infections. Permanent chlorine dioxide disinfection was recommended for the hot water systems of the two hotels. A dilemma caused by the legislation has arisen as to how to treat the thermal water and its distribution system. Oxidizing biocides and ionizers are not permitted by law. Regular thermal disinfection is recommended, but its application is often unfeasible for the rapid settling of mineral deposits and scale, which usually clog pipes and valves. The bath equipment (mostly very sophisticated) contained many different kinds of tubing and hoses and air and water jets that were almost all colonized with biofilms harboring Legionella. Thermal and hot water distribution systems and hydrotherapy procedures present the sources, while aerosol inhalation and drinking water appear to be the transmission.

This chapter summarizes the results of an extensive study of biological treatment plants (BTPs) from 43 paper mills in Sweden. During September and October 2005, each paper mill with a BTP was sampled systematically. High concentrations of Legionella, up to 109 CFU /liter, were found in investigated treatment plants. In total, 66% of all investigated paper mills were positive for Legionella. The biomass that undergoes sedimentation in secondary ponds is reinoculated into the system by return-loops to stimulate the growth of the microflora. The remaining biomass is dewatered and used as an energy source at the mill and for soil-making outside the plant. As much as 30 tons of biosludge is produced daily and is often mixed with fibrous waste. Conditions for legionella growth in these systems are almost optimal with a temperature of approximately 37°C, available nutrients and iron sources, and massive aeration. In an investigation the authors found that the cooling towers seem to have little impact on the introduction of legionella downstream into BTPs. Other international studies of biological treatment of sewage have detected legionella with PCR but not by culture. A local risk assessment should be done together with the public health authorities to identify risks for spreading legionella by aerosols and to minimize the spread of disease. Maintenance of cooling towers including regular cleaning is also recommended.

In most of hot spring baths in Japan, bathing water is circulated for extended use to conserve hot spring water. In recent years, massive outbreaks of Legionnaires’ disease among hot spring bath users have been reported in many districts in Japan. In the present chapter, the authors used a bath model to investigate the effectiveness of backwashing the filtering medium using a high concentration of chlorine for disinfection and growth inhibition of Legionella. They then assessed the usefulness of this method from the perspective of hygiene control of circulating bathing water. Chlorine backwashing by the filter refreshment method (5 to 10 mg/liter) was performed once a day for 9 days, and the bathing water and the water from the filter unit were collected every day prior to backwashing to determine the presence of Legionella and amoebae. The number of Legionella in both bathing water and filter water was maintained at a level lower than 10 to 70 CFU/100 ml by repeated backwashing with chlorinated water alone once a day; the Legionella growth was greatly inhibited compared to that under non-disinfection conditions. On the basis of these results, daily backwashing by the filter refreshment method is considered to be effective for growth inhibition of both Legionella and host amoebae in circulating bathing water. Together with daily use of the filter refreshment method, addition of chlorine into the bathing water to a minimum concentration of 0.2 to 0.4 mg/liter may ensure the supply of circulating bathing water with increased microbial safety.

The contaminated hospital water system (HWS) is often identified as the source of Legionella spp.. In this case, disinfection of the HWS is required, aiming at the prevention of nosocomial Legionnaires’ disease (LD). This chapter analyzes the literature for evidence of different disinfection methods and their success in preventing nosocomial LD. The authors found 45 studies describing disinfection methods in HWS, 16 in which the outcome LD was included in this review. Nine authors reported three or more interventions to achieve disinfection of the HWS. Based on the study-quality rating system, rates for each disinfection method was calculated. Chlorine dioxide (rate 2.0) was used in one study with the outcome being LD only, where several disinfection methods for LD control were implemented. Once a system is technically well maintained, and a low Legionella spp. count is achieved, a single method, e.g., copper/ silver ions or high warm-water maintenance temperature (>55°C at distal outlets) for the prevention of LD might be sufficient.

Different methods have been applied to control Legionella in water systems such as copper-silver ionization, chlorine dioxide, thermal treatments, hyperchlorination, UV radiation. In this chapter, authors evaluated the efficacy of a continuous system based on a silver-hydrogen peroxide mixture in two nursing home water systems. Recent studies showed a decline of Legionnaire's disease (LD) in health care facility-acquired cases where environmental monitoring of the water system and installation of a disinfection method were performed. Therefore, when the plumbing system is heavily contaminated by Legionella, it is important to apply control measures to reduce Legionella colonization and infection risk. Studies on the efficacy of the silver-hydrogen peroxide disinfection system are currently lacking, and to our knowledge, the present study is the first reported for Legionella control in hot water systems. In vitro studies demonstrated that hydrogen peroxide is a weak disinfectant in the absence of a catalyst. The combined action of hydrogen peroxide with silver as the catalyst characterize the disinfection power of the mixture used in this study. Results of this study, although preliminary, show that such disinfection systems can effectively control Legionella water system colonization even if some increase of Legionella levels at distal sites is observed.

Thermal eradication of Legionella bacteria (superheat and flush) is only a temporary palliative when Legionella levels pose an immediate problem. Legionella bacteria are inhibited at 60°C (140°F), but it may be very difficult to maintain this water temperature throughout the system due to heat loss from the pipes, as water frequently takes a long and tortuous path from leaving the heating system to its return via the hot water circulation loop. It has been shown that prior to fitting thermostatic mixing valves (TMVs) devices Legionella bacteria had been absent or in very low numbers, but following the installation of these devices, high levels of Legionella were detected at outlets. Treatment of the hot water may consist of raising the water temperature and/or injecting a chemical such as chlorine, chlorine dioxide, or nonchemical ionization. Micro-pore filters fitted to outlets and showers are very effective at preventing the dissemination of bacteria when present at outlets, particularly in sensitive areas such as intensive care units, operating theaters, and areas where patients have reduced immunity. TMVs were introduced to prevent the scalding of the aged and those unable to react quickly when presented with very hot water. If water temperatures can be reduced and protected by an effective biocidal regimen then the problems and cost associated with TMVs are avoided.

The cooling towers gather together many factors for bacteria proliferation and possible transmission to human beings. To eliminate dissemination in the atmosphere, the objective of the presented new system is to eliminate the vectorization of bacteria by eliminating the aerosol production and dispersion from cooling towers. A review of the published techniques of aerosol size measurement has been carried out. The most suitable technique is an optical technique relevant to flow and able to accurately determine particle size distributions, number density, and total mass. The aerosol spectrometer principle is based on the diffraction of white light. The first measurements have been realized at the air exit and at 10 cm and 20 cm above the air exit of the cooling tower. Other measurements have been performed to identify the influence of the ventilation velocity in a cooling tower. Additional measurements have been realized with and without the drift eliminators on the same cooling tower and for the same climatic conditions. The measurements done on cooling towers and the modeling of countercurrent flow show that efforts must be concentrated on water distribution and on water recovery. A new concept has been developed by the CEP and Climespace, patented to eliminate aerosol production and any direct crossing of air flow and water flow.

The use of a specific chemical product to minimize the population density of bacteria is the required and usual approach to prevent cooling towers from becoming amplifiers of Legionella. Sampling the cooling water to control bacterial populations is the most common way to evaluate the water treatment program’s efficiency. This chapter compares Legionella population counts obtained from different trials on the system to determine the most representative sampling point of the cooling water. Then the authors checked the results obtained after 3, 5, and 10 days of incubation in order to test the relevancy of the results. Experiments were performed on Climespace cooling tower plants in Paris, France. On site, the basin towers were connected. The results obtained from the different points showed that the tower basin provides irregular Legionella concentrations. The results obtained from the air flush-out sample show scattered distribution. To date, the use of other techniques which make it possible to evaluate all the Legionella (like the technique by PCR) could lead to an abusive use of the chemical treatments generating a significant risk to the environment.

Legionnaires’ disease is a potentially fatal form of pneumonia caused by the bacterium Legionella. Legionella control in cooling water systems depends on many parameters (conception, hydraulic management, maintenance) favorable to biological deposits. Any variation in the parameters measured (microbiological, physicochemical and environmental parameters) made it possible to identify and destroy Legionella growth sources. The cooling tower system is an ecosystem in which Legionella is always present and can proliferate any time of year. The percentage of samples in which Legionella is detected increased after April. A meticulous analysis of the risks made it possible to identify and manage all the factors of risk of proliferation of Legionella. The Legionella risk management cannot be carried out without taking account of the protozoa which are resistant to biocides. To fight continuously against the formation of biofilm seems to be the method most adapted to fight effectively against the proliferation of Legionella and limit the use of biocides.

Monochloramine can be more effective than chlorine at reducing Legionella colonization of potable water systems in large buildings, which are key sources of community-and hospital-acquired Legionnaire's disease. Because monochloramine use is associated with lower concentrations of trihalomethanes, many utilities are converting their residual disinfectant from chlorine to monochloramine. In this chapter, the author observed that monochloramine conversion was associated with decreased Legionella colonization of buildings served by the municipal water system. To determine whether this decreased risk was sustainable over a longer period of time and at a larger number of distal sites, a larger colonization survey was performed in San Francisco before and after conversion to monochloramine. Legionella were isolated from water samples and biofilm swabs, speciated, and serogrouped using standard methods. In the chlorine phase, 29% (45 of 157) of samples collected from water heaters yielded Legionella; fewer than 1% (1 of 159) of water heaters sampled during the monochloramine phase were positive (P<0.001). After controlling for water heater temperature, building height, and disruptions in service, monochloramine use was associated with a 96% reduction in the prevalence of water heater colonization (P<0.001).

Efficacy of monochloramine against mature biofilms formed in a model recirculating water system was compared on different surfaces under identical conditions. A laboratory-scale cooling tower was designed to simulate a wet evaporative cooling system, which was experimentally seeded with Legionella pneumophila. Rather low inoculum was added to the model system to mimic the natural entry of L. pneumophila from the water supply. Biofilm had been allowed to grow on coupons, and the model system was disinfected after a 180-day period with monochloramine. Three coupons of each material were removed from the basin and dip-rinsed in sterile phosphate buffer to remove unattached cells. The bacterial densities in biofilms from different materials after disinfection (180 min, 1.5 ppm dose) suggested a material-dependent activity of monochloramine. Monochloramine was found ineffective against microorganisms on polypropylene (PP) and galvanized steel surfaces. Results indicated that monochloramine shows material-dependent activity in the cooling tower model system and it has long residual activity at high temperatures and pH levels, leading to improved performance in recirculating water. This observation supports the material-dependent activity of monochloramine and it could be explained by the formation of dissimilar biofilms on different surfaces, which affects the architecture of the biofilm matrix. Chemical disinfection is strongly recommended from the beginning of the tower operation to prevent or reduce the occurrence of Legionnaires’ disease. Monochloramine disinfection is inexpensive and could be used in automatic injection devices instead of chlorine.

The aim of the authors' work was to study the impact of monochloramine on sessile and planktonic Legionella pneumophila populations. The authors produced Legionella biofilms on glass beads (0.5 mm diameter) in buffered-yeast extract (BYE) broth. Bacteria were grown for one week at 37°C, under static conditions. Planktonic cells were obtained in same growth conditions without glass beads. Planktonic and sessile cells were washed twice with sterile water and then treated with 0.25 to 10 mg liter-1 monochloramine solutions. Sessile bacteria were collected from glass beads by sonication and enumerated on buffered charcoal yeast extract (BCYE) agar. In order to detect metabolic activity in viable-but-nonculturable (VBNC) cells, planktonic L. pneumophila Lens cells were treated with 0.75 mg liter-1 of monochloramine in order to obtain no culturable cells. After 8 days in the resting medium, the authors were able to detect esterase activity and membrane integrity by fluorescence microscopy using the ChemChrom V6 substratum (Chemunex) in samples without culturable cells. This demonstrates that VBNC cells produced by monochloramine retained metabolic activity. Recovery of culturability was observed for sessile bacteria treated with 1 mg liter-1 of monochloramine and coincubated with Acanthamoeba castellanii.

Most cases of legionellosis result from exposure to building water systems containing Legionella. There are two hazard analysis and control systems recommended for use today in legionellosis prevention: water safety plans (WSP) and hazard analysis critical control point (HACCP) plans. Prevention of waterborne disease is most effectively achieved by use of hazard analysis and control systems. In its publication in preparation, Legionella and the Prevention of Legionellosis, the use of the WSP hazard analysis and control system is highly promoted by the World Health Organization. For hazard analysis and control schemes to be useful, they must be practical, easy to develop and cost-effectively implemented. In France, the HACCP system was specified in 2005 as the preferred hazard analysis and control system for preventing legionellosis associated with cooling towers. Development of HACCP plans in the food processing/manufacturing industry is driven either by a legal mandate or by buyers of food who require suppliers to have such systems in place. In response to quantitatively documented success in the United States, the World Health Organization has adopted HACCP as the preferred recommended system for preventing food borne disease worldwide.

This chapter presents the panel responses to questions posed to the round table discussion “Environmental Sampling Data to Determine Risk". A risk assessment (RA) should be carried out in all premises that have public access and should include all healthcare premises, factories and office blocks. An RA should include a thorough site survey for all water systems in use, cooling towers, hot and cold water systems, spa pools, irrigation systems and should take into account the population using the premises. PCR can be used with the following provisos: there are still some issues around the sensitivity and specificity of PCR for legionellae, especially around the target sequences for the genus. The problems associated with thermostatic mixing valves (TMVs) are mainly associated with the fact that the temperature downstream of these is within the ideal range for microbial growth. A period of intensive monitoring should be undertaken including aerobic colony counts (not dipslides) combined with Legionella testing, ideally using both culture and PCR. The ecological determinants of Legionella colonization are common to all systems. Temperature, nutrients, presence of other microorganisms, and stagnation are all major predictors of Legionella growth. As a result, ecological determinants can be readily standardized across a wide range of systems, but risk assessment cannot.