Interaction between legionella pneumophila and biofilm forming organism pseudomonas aeruginosa

aeruginosa PAO1-CFP biofilms 50 3.7 Introduction of NALCO 7320 into developing and mature P.. aeruginosa PAO1 biofilm removal screening 68 3.10 Antimicrobial susceptibility testing of

AND BIOFILM FORMING ORGANISM PSEUDOMONAS AERUGINOSA

WON CHOONG YUN (B.Sc (Hons.), NUS)

A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF SCIENCE

DEPARTMENT OF MICROBIOLOGY NATIONAL UNIVERSITY OF SINGAPORE

Acknowledgements

I would like to express my heartfelt gratitude to the following people who have

made a difference in my life during the course of this study:

A/Prof Lee Yuan Kun for his invaluable guidance, constant encouragement and

patience throughout the course of this study

Dr Gamini Kumarasinghe from the Department of Laboratory Medicine, National

University Hospital, A/Prof Zhang Lian Hui from Institute of Molecular and Cell

Biology, and A/Prof Tim Tolker-Nielsen from BioCentrum-DTU, The Technical

University of Denmark, for kindly providing bacterial strains for this study

Mr Ma Xi from Nalco Company for his invaluable advice, generous assistance

and constant concern Dr Chen Hui and Mr Tim Lim, also from Nalco Company,

for their generous sharing of experiences and gracious assistance

Mr Low Chin Seng for his precious technical assistance and for being a

fatherly-figure in a laboratory setting Mdm Chew Lai Meng for her encouragement and

warm friendship

Ho Phui San, Lee Hui Cheng, Wang Shugui and especially Chow Wai Ling and

Janice Yong Jing Ying for their generous help, precious friendship and incredible

understanding when absentmindedness get the better of me Post-graduate life has

never been better without them!

Toh Yi Er and Lee Kong Heng from Confocal Microscopy Unit, and Toh Kok Tee

from Flow Cytometry Unit for their invaluable technical assistance

My family and husband, Clement Choo, for their generous love, unwavering

support and relentless encouragement through difficult time of my life Especially

my father, for his thought-provoking discussions and tremendous help in software

improvements for this study My son for sharing his precious life with me

2.1.5.1 Natural and man-made habitats 16

2.1.5.2 Distribution of Legionella in Singapore 18

2.1.5.3 Association of Legionella with protozoa 19

2.1.5.4 Association of Legionella with biofilm 21

2.1.5.5 Interaction of Legionella with Pseudomonas spp 24

2.2.2 General characteristics of biofilm 25

2.2.4 Stages of biofilm development 27

2.2.4.1 Stage 1: Reversible attachment 27

2.2.4.2 Stage 2: Irreversible attachment 28

2.2.5 Determinants of biofilm structure 31

2.2.6 Microbial diversity of biofilms 33

2.2.7 Microbial positioning in biofilm 34

2.3.4 Water treatment in cooling towers 38

3.1.3 Maintenance of stock cultures 42

3.2.1 Growth kinetics of L pneumophila 42

3.2.2 Growth kinetics of P aeruginosa PAO1 43

3.2.3 Growth kinetics of P aeruginosa PAO1-CFP 43

3.3 Determination of the influent flow rate (Q) for continuous culture in

3.4 Optimization of labelling processes 44

3.4.1 Optimization of L pneumophila labelling with CFDA-SE 44

3.4.2 Optimization of planktonic P aeruginosa PAO1-CFP labelling

3.4.4 Optimization of P aeruginosa PAO1-CFP biofilm labelling with PI 45

3.5 P aeruginosa PAO1-CFP biofilm formation in CDC Biofilm Reactor

3.5.2 Setup of CDC Biofilm Reactor assembly 47

3.5.3 P aeruginosa PAO1-CFP biofilm formation 48

3.6 Introduction of L pneumophila into P aeruginosa PAO1-CFP biofilms 50

3.7 Introduction of NALCO 7320 into developing and mature

P aeruginosa PAO1-CFP biofilms containing L pneumophila 51

3.8 Monitoring of each organism in CBR continuous flow system 52

3.8.2.1 Sampling bulk fluid 52

3.8.3.1 Preparation of coupons intended for enumeration 53

3.8.3.2 Preparation of coupons intended for visualization by CLSM 53

3.8.4 Disaggregation by homogenization 54

3.8.5 Enumeration of each organism 55

3.8.5.1 Enumeration of P aeruginosa PAO1-CFP by culture 55

3.8.5.2 Enumeration of L pneumophila by immunofluorescence 56

3.8.6 Detection of exogenous contaminants 58

3.8.7 Visualization and image acquisition by CLSM 59

3.8.8 Application of COMSTAT image analysis software package 60

3.8.8.1 Preparation of image stacks 60

3.9 Screening for effective P aeruginosa PAO1 biofilm-removing agent 65

3.9.1 Kinetics of P aeruginosa PAO1 biofilm formation in microtiter

3.9.3 Biofilm-removing agents used 67

3.9.4 P aeruginosa PAO1 biofilm removal screening 68

3.10 Antimicrobial susceptibility testing of NALCO 7320 69

4.2 Determination of the influent flow rate (Q) for continuous culture in CDC

4.3 Optimization of labelling processes 74

4.3.1 Optimization of L pneumophila labelling with CFDA-SE 74

4.3.2 Optimization of planktonic P aeruginosa PAO1-CFP labelling

4.3.3 Optimization of P aeruginosa PAO1-CFP biofilm labelling with PI 77

4.4 Kinetics of P aeruginosa PAO1-CFP biofilm formation in CDC

4.4.1 Kinetics of biofilm formation 80

4.4.2 Structure of biofilm by image analysis 81

4.5.4 Surface-to-biovolume ratio distributions of developing and

4.5.5 Porosity distributions of developing and mature biofilms 100

4.5.6 Correlation between SBR and porosity 103

4.5.7 Correlation between legionellae adhesion and parameters of

4.5.8 Localization of L pneumophila in P aeruginosa PAO1-CFP

4.6 Screening for effective P aeruginosa PAO1 biofilm removing agent 108

4.6.1 Kinetics of P aeruginosa PAO1 biofilm formation in microtiter

4.6.2 P aeruginosa PAO1 biofilm removal screening 109

4.7.1 Kinetics of P aeruginosa PAO1 biofilm removal 111

4.7.2 Antimicrobial susceptibility testing 112

4.8 Introduction of NALCO 7320 into developing and mature P aeruginosa

PAO1-CFP biofilms containing L pneumophila 114

4.8.1 Persistence of P aeruginosa PAO1-CFP in CBR 114

4.8.2 Structure of P aeruginosa PAO1-CFP biofilms treated by NALCO

4.8.3 Persistence of L pneumophila in P aeruginosa PAO1-CFP biofilms

4.8.4 Distribution of L pneumophila in P aeruginosa PAO1-CFP biofilms

4.8.5 Bio-volume distributions of developing and mature biofilms treated

List of Tables Table 3.1

Sampling points of 6 independent experiments for the study

of P aeruginosa PAO1-CFP biofilm formation

List of biofilm-removing agents used

Effect of treatment duration on staining and viability of L

pneumophila cells

Effect of treatment duration on staining of P aeruginosa

PAO1-CFP cells

Table showing Pearson’s correlation between Log (Number

of L pneumophila cells) and Log (Number of CFDA pixels

per µm3)

The ratio of SBR at the bottom 20% versus the top 20% of developing and mature biofilm

Comparing means of porosity over time

Table showing Pearson’s correlation between porosity and SBR

Table showing Pearson’s correlation between legionellae

adhesion to P aeruginosa PAO1-CFP biofilm (representing

the number of legionellae per coupon per 106 legionellae inoculated into CBR) and parameters of the biofilm

Efficacy of biofilm removing agents

Table showing Pearson’s correlation between bio-volume and legionellae loss

List of Figures Figure 3.1

Schematic diagram of the CDC Biofilm Reactor assembly

Growth curve of L pneumophila cultured in BCYE broth

at 37°C with shaking at 120rpm

Growth curve of P aeruginosa PAO1 cultured in MM

liquid media at 30°C with shaking at 120rpm

Growth curve of P aeruginosa PAO1-CFP cultured in

MM liquid media at 30°C with shaking at 120rpm

Graph of Ln(OD600nm) against time (hr) plotted for the

exponential growth phase of P aeruginosa PAO1-CFP

Histograms illustrating the number of events (cells) plotted against FL1-H (representing green fluorescence of CFDA-

stained cells) for L pneumophila cells that were (A) mock

treated, or treated with CFDA-SE for (B) 20mins, (C) 30mins, or (D) 40mins

Histograms illustrating the number of events (cells) plotted against PMT4 Log (representing red fluorescence of PI-

stained cells) for P aeruginosa PAO1-CFP cells that were

(A) mock treated, or treated with 1.0mg/ml PI for (B) 5mins, (C) 10mins, or (D) 15mins

Histograms illustrating the number of events (cells) plotted against PMT4 Log (representing red fluorescence of PI-

stained cells) for P aeruginosa PAO1-CFP cells that were

(A) mock treated, or treated with 0.1mg/ml PI for (B) 5mins, (C) 10mins, (D) 15mins, or (E) 30mins

CLSM images of a 7 days old P aeruginosa PAO1-CFP biofilm and adhered L pneumophila, stained with 0.1mg/ml PI for 5mins: (A) P aeruginosa PAO1-CFP biofilm (blue fluorescence), (B) CFDA-stained L

pneumophila (green fluorescence), (C) PI-stained P

aeruginosa PAO1-CFP biofilm, and (D) overlapping

display of the above 3 images

CLSM images of a 7 days old P aeruginosa PAO1-CFP biofilm and adhered L pneumophila, stained with 0.1mg/ml PI for 15mins: (A) P aeruginosa PAO1-CFP biofilm (blue fluorescence), (B) CFDA-stained L

pneumophila (green fluorescence), (C) PI-stained P

display of the above 3 images

CLSM images of a 7 days old P aeruginosa PAO1-CFP biofilm and adhered L pneumophila, stained with 0.1mg/ml PI for 30mins: (A) P aeruginosa PAO1-CFP biofilm (blue fluorescence), (B) CFDA-stained L

pneumophila (green fluorescence), (C) PI-stained P

aeruginosa PAO1-CFP biofilm, and (D) overlapping

display of the above 3 images

Viable cell counts of P aeruginosa PAO1-CFP biofilm

formed in CBR at 30°C with stirring at 120rpm

Bio-volume of P aeruginosa PAO1-CFP biofilm formed

in CBR at 30°C with stirring at 120rpm

Average thickness of P aeruginosa PAO1-CFP biofilm

formed in CBR at 30°C with stirring at 120rpm

Maximum thickness of P aeruginosa PAO1-CFP biofilm

formed in CBR at 30°C with stirring at 120rpm

Substratum coverage of P aeruginosa PAO1-CFP biofilm

formed in CBR at 30°C with stirring at 120rpm

Surface-to-biovolume ratio (SBR) of P aeruginosa

PAO1-CFP biofilm formed in CBR at 30°C with stirring at 120rpm

Roughness coefficient of P aeruginosa PAO1-CFP

biofilm formed in CBR at 30°C with stirring at 120rpm

Viable cell counts of planktonic P aeruginosa PAO1-CFP

in the bulk fluid of CBR at 30°C with stirring at 120rpm

CLSM image of a P aeruginosa PAO1-CFP biofilm (blue)

structure indicative of dispersion stage of biofilm

development, with adhered L pneumophila (green)

Adhesion of L pneumophila to different developmental stages of P aeruginosa PAO1-CFP biofilm

Status of L pneumophila in our continuous flow CBR

Percentage loss of L pneumophila in developing P

aeruginosa PAO1-CFP biofilm

Percentage loss of L pneumophila in mature P aeruginosa

PAO1-CFP biofilm

Bio-volume distribution of (A) developing, and (B) mature

P aeruginosa PAO1-CFP biofilms

Surface-to-biovolume ratio (SBR) distribution of (A)

developing, and (B) mature P aeruginosa PAO1-CFP

biofilms

Porosity of P aeruginosa PAO1-CFP biofilm

Porosity distribution of (A) developing, and (B) mature P

aeruginosa PAO1-CFP biofilms

Scatterplot of porosity and SBR both obtained from all data of 6 independent experiments

CLSM images of P aeruginosa PAO1-CFP biofilm (blue) with adhered L pneumophila (green) taken on different

occasions: (A) 3hrs after legionellae introduction to developing biofilm (3-days-old), (B) 4 days after legionellae introduction to developing biofilm, (C) 3hrs after legionellae introduction to mature biofilm (7-days-old), and (D) 4 days after legionellae introduction to mature biofilm

Kinetics of P aeruginosa PAO1 biofilm formation in

microtitre plate at 30°C

Highest percentage biofilm removal of various removing agents

biofilm-Kinetics of biofilm removal by NALCO 7320

Visual determination of minimum inhibitory concentration (MIC)

Determination of minimum bactericidal concentration (MBC) of NALCO 7320

Viable cell counts of P aeruginosa PAO1-CFP biofilms

treated with NALCO 7320

Viable cell counts of planktonic P aeruginosa PAO1-CFP

in CBR treated with NALCO 7320

Bio-volume of P aeruginosa PAO1-CFP biofilm in CBR

treated with NALCO 7320

Average thickness of P aeruginosa PAO1-CFP biofilm in

CBR treated with NALCO 7320

Maximum thickness of P aeruginosa PAO1-CFP biofilm

in CBR treated with NALCO 7320

Substratum coverage of P aeruginosa PAO1-CFP biofilm

in CBR treated with NALCO 7320

Surface-to-biovolume ratio of P aeruginosa PAO1-CFP

biofilm in CBR treated with NALCO 7320

Roughness coefficient of P aeruginosa PAO1-CFP

biofilm in CBR treated with NALCO 7320

Persistence of L pneumophila in P aeruginosa

PAO1-CFP biofilms treated with NALCO 7320

Cell counts of planktonic L pneumophila in CBR treated

with NALCO 7320

Scatterplot of bio-volume and legionellae loss, obtained from 4 independent experiments

Effect of NALCO 7320 on the distribution of L

pneumophila in (A) developing, and (B) mature P

aeruginosa PAO1-CFP biofilms

Effect of NALCO 7320 on the distribution of bio-volume

in (A) developing, and (B) mature P aeruginosa

PAO1-CFP biofilms

Porosity of P aeruginosa PAO1-CFP biofilm in CBR

treated with NALCO 7320

Effect of NALCO 7320 on porosity distribution of (A) developing, and (B) mature P aeruginosa PAO1-CFP

List of Abbreviations 3OC 12 -HSL

Hydraulic residence time Buffered charcoal yeast extract CDC biofilm reactor

5-(and-6)-carboxyfluorescein diacetate, succinimidyl ester

Colony forming unit Confocal laser scanning microscope 2,2-Dibromo-3-nitrilopropionamide Direct florescent antibody

Deionized water Extracellular polysaccharides

Minimal media + 60µg/ml gentamicin

Paraformaldehyde Propidium iodide Parts per million Nutrient influent flow rate Correlation coefficient Revolutions per minute Surface to bio-volume ratio Sodium dodecyl sulphate Doubling time

Maximum volume of bulk fluid in CBR during continuous flow

Viable but non-culturable

Summary

In present study, a reproducible model Pseudomonas aeruginosa PAO1-CFP

biofilm, with distinct stages of biofilm development, was established in CDC

Biofilm Reactor continuous flow system using defined minimal media at 30°C

Splitting certain data, such as bio-volume and surface-to-biovolume ratio (SBR),

into 5 sections along biofilm thickness and applying a novel method of biofilm

porosity quantification in a 3-dimensional context provided greater insights of

biofilm structures and properties Consequently, biofilm structures and

development were better described, and the first physical evidence of porous

channels within biofilm cell cluster was observed

Legionella pneumophila adhesion study revealed that legionellae adhesion to

biofilms was independent of developmental stage of the latter Instead, biofilm

structure and porosity were found to determine the amount and even localization

of legionellae adhesion to biofilm Additionally, L pneumophila persistence study

revealed that legionellae was least likely to get desorbed at bottom 60% of the

biofilms, especially at bottom 20%, and unbalanced advective transport of

legionellae towards biofilm surface commenced upon biofilm maturation, most

probably due to unbalanced cell growth

Eight commercially available biofilm-removing agents were screened using

microtitre plate assay for one with the highest efficacy Subsequently, application

of the selected biocide, NALCO 7320, (at bactericidal concentration to planktonic

P aeruginosa PAO1) to P aeruginosa PAO1-CFP biofilms yielded complete

disinfection of developing biofilm while a resistant subpopulation was found in

the remains of mature biofilm

Porosity distribution and biofilm structural analysis suggested that NALCO 7320

caused biofilm detachment by affecting the nature of extracellular polysaccharides

(EPS) matrix that bound the microbial cells together as a microcolony, while

applying biocidal effect on P aeruginosa PAO1-CFP cells within the biofilm

Legionellae persistence in biocide-treated biofilms was found to be independent

on the stage of biofilm development and loss of biomass, but regions of the

biofilms in which legionellae best persist were detected Since EPS is a major

component in biofilm matrix, it was hypothesized to play an important role in

legionellae persistence in biocide-treated biofilms

Chapter 1: Introduction

L pneumophila, the main species of the genus Legionella, was first recognized as

a pathogen after an outbreak of acute pneumonia, called Legionnaires’ disease that

occurred at the convention of the American Legion in Philadelphia, USA, during

1976 (Fraser et al., 1977) To date, forty eight species of Legionella have been

described, including 70 distinct serogroups (Borella et al., 2005) Approximately

half of the 48 species of legionellae have been associated with legionellosis, but it

is likely that most legionellae can cause human disease under appropriate

conditions (Fields, 1996) L pneumophila is responsible of approximately 91% of

all reported community cases of legionellosis and among the 15 serogroups of this

species, L pneumophila serogroup 1 accounts for the 84% of confirmed cases (Yu

et al., 2002)

The real number of cases of Legionnaires’ disease is unknown, although in the

USA, it is estimated that the incidence is 20 cases per million population (Borella

et al., 2005) In Europe, during the period 2003-2004, a total of 10,322 cases of

Legionnaires’ disease was reported, with national infection rates ranging from 0 to

28.7 cases per million population (Ricketts and Joseph, 2005) The mean annual

incidence rates were 0.9 (Heng et al., 1997) and 1.7 (Goh et al., 2005) per 100,000

population in Singapore, during the period 1986-1996 and 1998-2002

respectively Because of the difficulty in distinguishing Legionella associated

diseases from other forms of pneumonia and influenza, many cases are

unreported Nevertheless, the overall case-fatality rate is high especially among

seriously immunosuppressed individuals, at 24% for the adequately treated and

80% for those without treatment (Fliermans, 1995)

Legionella associated diseases have emerged in the last half of the 20th century

because of human alteration of the environment Legionella spp is part of the

natural aquatic environment and the bacterium is capable of surviving extreme

ranges of environmental conditions (Fliermans et al., 1981) However, when

allowed to remain in their natural habitat, legionellae are rarely the causative

agents of human disease since natural freshwater environments have not been

implicated as reservoirs of legionellosis Main sources of L pneumophila are

waters from hot distribution systems and cooling towers Numerous cases of

legionellosis have been found to occur after exposure to contaminated waters in

offices, hotels, hospitals and cruise ships, among other locations (Borella et al.,

2005)

Factors leading to outbreaks or sporadic cases are not completely understood, but

certain events are considered prerequisites for infection These include the

presence of the bacterium in aquatic environment, amplification to an unknown

infectious dose and transmission via aerosol to a human host that is susceptible to

infection (Fliermans, 1995) Although amoebae are key factors in Legionella

amplification process (Fields, 1996), this pathogen is able to survive as free

organism for long periods within biofilms which are widespread in man-made

water systems Its persistence has been attributed to survival within biofilms

(Rogers and Keevil, 1992; Rogers et al., 1994) Additionally, association of

Legionella to biofilm may explain, at least in part, why legionellae are relatively

hard to eradicate in water systems, as biofilms exhibit a marked resistance to

biocidal compounds and chlorination (LeChevallier et al., 1988) Therefore a

more extensive knowledge on biofilm-associated legionellae may lead to the most

effective control measures to prevent legionellosis

Majority of Legionella-biofilm studies employed naturally occurring microbial

biofilm communities, and failed to identify all the organisms present and their

contribution to the survival and multiplication of legionellae Additionally,

Pseudomonas aeruginosa PAO1, a wound isolate (Holloway, 1955), is generally

found in the same aquatic environments as L pneumophila (Murga et al., 2001), is

the most widely used P aeruginosa laboratory strain (Stover et al., 2000) and its

biofilm development has been well documented (Sauer et al., 2002) Therefore in

the present study, a reproducible model P aeruginosa PAO1-CFP biofilm was

established in a CDC Biofilm Reactor continuous flow system using defined

minimal media at 30°C Since P aeruginosa PAO1 biofilms are structurally and

dynamically complex biological systems with regulated developmental stages

(Sauer et al., 2002), it was hypothesized that legionellae interacts differently with

biofilms at different developmental stages and responds differently to biocidal

treatments while residing in biofilms at different developmental stages

To allow further insights into biofilm development, current method of quantifying

biofilm structures was improved by splitting up certain descriptive data into 5

sections along the thickness of the biofilm and a novel method of quantifying

biofilm porosity in a 3-dimensional context was developed Using the model and

better descriptive methods of biofilm structure and porosity, it was determined if

there is any difference (in numbers and distribution pattern) in accumulation and

persistence of L pneumophila in developing and mature biofilm, and if the

structure or porosity of biofilm plays a role in the accumulation and persistence of

L pneumophila

In a bid to deepen the knowledge on the effect of biocide on

legionellae-associated to biofilms, a biocide was first selected by screening through eight

commercially available biofilm-removing agents for one with the highest efficacy

using microtitre plate assay Subsequently, the effects of the selected biocide,

NALCO 7320 (at bactericidal concentration to planktonic P aeruginosa PAO1)

on the persistence and structure of P aeruginosa PAO1-CFP biofilm, and the

persistence of biofilm-associated legionellae were characterized

Chapter 2: Literature Review

2.1 Legionella

2.1.1 Introduction to Legionella

The terror of the unknown is seldom better displayed than by the response of a

population to the appearance of an epidemic, particularly when the epidemic

strikes without apparent cause Between July 22 and August 3, 1976, there was a

remarkable incidence of febrile respiratory disease among persons who had

attended the American Legion Convention in Philadelphia from July 21 to 24

“Legionnaires’ disease” (LD) is the term used to describe the illness that occurred

among persons attending the convention (Fraser et al., 1977)

The etiologic agent of LD was first isolated in guinea pigs from lung specimens

collected on autopsy and subsequently, serologic evidence for the etiological role

of the bacterium, designated L pneumophila subsp pneumophila, was obtained by

indirect fluorescent antibody staining (McDade et al., 1977) In fact, the first

strains of Legionella were already isolated in guinea pigs by using procedures for

the isolation of Rickettsia by Tatlock in 1943 (McDade et al., 1979)

2.1.2 General characteristics of Legionella

Members of the genus Legionella are faintly staining Gram-negative, aerobic rods

or filaments (usually found after growth in enriched laboratory media), 0.3-0.9µm

in width and 2-20µm or more in length (Brenner et al., 1985) They are neither

encapsulated nor acid-fast; they do not form endospores or microcysts (Brenner et

al., 1985) They are chemoorganotrophic, where amino acids are utilized as

carbon and energy sources, while carbohydrates are neither fermented nor

oxidized (Tesh and Miller, 1981) Furthermore, they are nutritiously fastidious

where L-cysteine-HCL is absolutely necessary for their growth and iron salts in

the medium enhance their growth (Feeley et al., 1978)

The cell wall is made up of two three-layered unit membranes (Brenner et al.,

1985) and is predominated by branched chain fatty acids (Fisher-Hoch et al.,

1979) The fatty acid composition of the cell wall varies among the different

species belonging to the genus Legionella, thus fatty acid analysis is useful for the

differentiation of Legionella species (Diogo et al., 1999) In addition, the cellular

fatty acid composition of the bacteria is found to be similar to that of known

thermophilic bacteria (Fliermans, 1995) Therefore, it is not surprising to see

Legionella associated with thermally elevated habitats (Verissimo et al., 1991)

Various L pneumophila strains and isolates of species other than L pneumophila

are able to produce flagella (Heuner et al., 1995), which are later shown to be a

positive predictor for virulence in Legionella (Bosshardt et al., 1997) Ott et al

(1991) demonstrated that the expression of the gene flaA, encoding the flagella

subunit, is temperature-dependent Further studies in the same laboratory revealed

that the expression of flaA is also influenced by the growth phase, the viscosity

and the osmolarity of the medium, and by amino acids (Heuner et al., 1999)

Similar to a number of other Gram-negative bacteria, Legionella is able to enter a

viable but non-culturable (VBNC) state under low-nutrient conditions (Hussong et

al., 1987) Also, the loss of culturability appeared to be accelerated at higher

temperature of 37°C, as compared to 4°C (Hussong et al., 1987) Many

procedures to reactivate VBNC legionellae have failed, with the exception of the

passage through Acanthamoeba castellani (Steinert et al., 1997) Both

amoeba-reactivated cells and plate-grown L pneumophila cells had the same capacity for

intracellular survival in human monocytes and intraperitoneally infected guinea

pigs, which is considered a parameter for virulence However, reactivation of

VBNC cells was not observed in the animal model Although there is a correlation

of Legionella infection of amoeba, human cell lines and animal models, it cannot

be excluded that VBNC forms are virulent for human

2.1.3 Taxonomy of Legionella

The family Legionellaceae consists of the single genus Legionella (Fields et al.,

2002) At least 48 species comprising 70 serogroups have been distinguished

(Fields et al., 2002; Borella et al., 2005) Legionella pneumophila consists of 15

serogroups, of which serogroup 1 is the most common, followed by serogroups 4

and 6 (Den Boer and Yzerman, 2004) The number of species and serogroups of

legionellae continues to increase Phylogenetically, the nearest relative to the

Legionellaceae is Coxiella burnetti, the etiologic agent of Q fever (Adeleke et al.,

1996 and Swanson and Hammer, 2000) These organisms have similar

intracellular lifestyles and may utilize common genes to infect their host

Some legionellae cannot be grown on routine Legionella media and has been

termed Legionella-like amoebal pathogens (LLAPs; Adeleke et al., 1996) LLAP

was first recovered and isolated from the sputum of a patient with pneumonia by

cocultivating with its host amoebae (Fry et al., 1991) Additional LLAP strains

may be human pathogens as well, but proving this is difficult because they cannot

be detected by conventional techniques used for legionellae (Fields et al., 2002)

2.1.4 Legionella and Diseases

2.1.4.1 Clinical presentation

Diseases caused by Legionella are collectively termed legionellosis Legionellosis

classically presents as two distinct clinical entities, Legionnaires’ disease (LD;

Fraser et al., 1977), a severe multisystem disease involving pneumonia, and

Pontiac fever (PF; Glick et al., 1978), a self-limited flu-like illness

Features of LD include fever, non-productive cough, headache, myalgias, rigors,

dyspnea, diarrohea and delirium (Tsai et al., 1979) Histological reports describe

intra- and extracellular bacteria in phagocytes, fibroblasts and epithelial cells

(Fields, 1996) Chest X-rays often show evidence of pneumonia, but it is

impossible to distinguish LD from other types of pneumonia on the basis of

symptoms alone (Edelstein, 1993) As a result, many cases go probably

unreported This assumption is supported by serologic surveys which show that

many persons in an apparently healthy population have antibodies against

legionellae (Paszko-Kolva et al., 1993)

The clinically distinct self-limited and non-pneumonic PF is a milder,

influenza-like form of disease (Fields et al., 1990) It usually appears on an epidemic mode

(Tossa et al., 2006) but many persons who seroconvert to Legionella will be

entirely asymptomatic (Boshuizen et al., 2001) Because of its benignity and lack

of specificity, the occurrence of PF is often undiagnosed and is therefore even less

reported than LD

To date, there is no consensus on the duration of the incubation period, on its

clinical symptoms, nor on the causal species of Legionella Since the microbe has

never been isolated from PF patients, it has been speculated that PF is caused by

VBNC forms of Legionella (Steinert et al., 1997) Other hypotheses to explain PF

include changes in virulence factors, toxic or hypersensitivity reactions to bacteria

(Kaufmann et al., 1981) or their products; high levels of endotoxin in aerosolized

water may be responsible for clinical symptoms (Fields et al., 2001)

2.1.4.2 Diagnosis

Although Legionella species are gram-negative bacilli, they are rarely visualized

on Gram stains of clinical material (Stout et al., 2003) A Gram stain of a sputum

specimen showing polymorphonuclear leukocytes without bacteria can be a

valuable clue to Legionella infection (Muder and Yu, 2002)

Clincal specimens used for culture of Legionella species include sputum or

bronchoalveolar lavage specimens, bronchial aspirates, lung biopsy specimens and

blood (Den Boer and Yzerman, 2004) Isolation of Legionella species from a

clinical specimen on selective media provides a definitive diagnosis Buffered

charcoal yeast extract agar that contains antibiotics to suppress commensal flora is

commercially available However, certain media formulations that are selective

for L pneumophila may inhibit growth of other Legionella species (Muder and

Yu, 2002) Preheating steps and acid washing procedures were developed to

reduce overgrowth by other microorganisms, thus serve as additional means of

increasing the sensitivity of sputum culture (Den Boer and Yzerman, 2004)

Species-specific DFA testing is more often applied directly to clinical specimens

(Stout et al., 2003) Since the sensitivity of the DFA stain is much lower than for

culture (range, 25 to 75%), it was suggested that this test should not be performed

routinely (Edelstein, 1993) It should be noted that the sensitivity and specificity

of DFA testing of clinical specimens is not precisely known for species other than

L pneumophila (Muder and Yu, 2002)

The commercially available Legionella urinary antigen test reliably detects only

infection due to L pneumophila serogroup 1 Urinary antigen test results are

occasionally positive in cases of disease due to other Legionella species, but the

sensitivity is low; consequently, a negative test result is of little value in excluding

Legionella infection (Muder and Yu, 2002) Potentially, PCR could detect all

known Legionella species However, so far the sensitivity of the test varies from

11 to 100% and many publications report specificities of lower than 99% (Den

Boer and Yzerman, 2004)

At present, optimal sensitivity for diagnosis of LD will be achieved by using a

combination of culture, serological investigation and urinary antigen detection

(Den Boer and Yzerman, 2004) After reviewing the diagnostic methods of LD,

Den Boer and Yzerman (2004) postulated that easy-to-perform PCR test with high

sensitivity and a specificity of above 99% may become accepted as new gold

standard for diagnosis of LD in the future On the contrary, the sensitivity and

specificity of detecting seroconversion to Legionella species other than L

pneumophila is still uncertain While seroconversion alone can be used for the

diagnosis of infection due to other species, such diagnoses should be regarded as

presumptive unless there are supporting microbiologic or epidemiologic data

(Muder and Yu, 2002)

2.1.4.3 Epidemiology

Studies have estimated that between 8,000 and 18,000 persons are hospitalized

with legionellosis annually in the United States (Marston et al., 1997) Failure to

utilize available diagnostic tools may result in the mistaken impression that

Legionella infections are not occurring in a hospital or a community For

Legionella infections in particular, national extrapolations are potentially

misleading because of the critical importance of local microenvironment

As summarized by Fliermans (1995), the overall case-fatality rate is high Among

previously healthy individuals, 7-9% die when treated with erythromycin, while

25% die when hospitalized but not treated with appropriate antibiotics Among

seriously immunosuppressed individuals, the mortality rate is 24% for the

adequately treated and 80% for those without treatment

In the same study carried out by Marston et al (1997) involving patients with

community-acquired pneumonia requiring hospitalization, Legionella spp was

found to be responsible for 2–5% of the cases studied In the USA, 91% of

isolates from Legionnaires’ disease patients are typed as Legionella pneumophila

serogroup 1 (Marston et al., 1997) This is in contrast to the situation in Australia

and New Zealand, where 30% of the cases of Legionnaires’ disease are caused by

Legionella longbeachae (Yu et al., 2002)

In an international collaborative study conducted by Yu et al (2002), community

acquired LD is dominated by L pneumophila serogroup 1 (84.2% of all isolates)

Species other than L pneumophila were rare: L longbeachae (3.9%) and L

bozemanii (2.4%) accounted for most of the nonpneumophila cases L micdadei,

L feeleii, L dumoffii, L wadsworthii and L anisa combined accounted for 2.2%

of the remaining cases Hospital-acquired pneumonia have also involved

serogroups other than L pneumophila serogroup 1 (especially serogroups 4 and 6)

and Legionella species other than L pneumophila, especially L micdadei, L

dumoffii and L bozemanii (Fang et al., 1989) Nevertheless, L pneumophila

serogroup 1 is still the dominant cause of legionellosis

Epidemiological studies indicate that Legionella is an opportunistic pathogen,

with elderly and immuno-compromised patients being most susceptible

(Fliermans, 1996) Other risk factors for the disease include smoking, male sex,

chronic lung disease, hematologic malignancies, end-stage renal disease, lung

cancer, immunosuppresion and diabetes (Marston et al., 1994) Differences in host

susceptibility and bacterial virulence make it difficult to clearly define an

infectious dose (Steinert et al., 2002)

The best-documented route for transmission of infection is the generation of an

infective aerosol from a Legionella contaminated water source (Winn, 1999)

Aspiration of contaminated potable water is another probable mechanism for

infection of the lower respiratory tract (Blatt et al., 1993) Entry through the

gastrointestinal tract has been suggested to explain abdominal infections, although

this portal of entry has not been proved Direct entry of bacteria into flesh wound

may also cause nosocomial Legionella infection (Winn, 1999) There has been no

evidence of human-to-human transmission or documented laboratory infections

2.1.4.4 Epidemiology in Singapore

To find out if the disease occurs in Singapore, legionellosis was made

administratively notifiable in 1985 and legally notifiable in 2000, to the

Quarantine and Epidemiology Department, Ministry of the Environment The first

local case of LD, a 27 year old Chinese male plumber, was admitted to Toa Payoh

Hospital on 4th February, 1986 (Lim et al., 1986) Clinical suspicion of LD was

confirmed by the presence of serum antibody to L pneumophila (titre 1:512) by

an indirect fluorescent antibody test and it is not known where the patient acquired

the illness from

In an attempt to determine the level of antibodies to L pneumophila serogroups 1

to 4 in 150 young normal adults who are blood bank donors, Nadarajah et al

(1987) discovered a 20% overall prevalence of antibodies to L pneumophila in

the normal population studied In addition, a national serologic survey of the

general population conducted in 1993 showed a prevalence of 10.3% in those

below 20 years of age and 21.9% in those 20 years of age and above (Heng et al.,

1997) These results suggest that L pneumophila is wide spread in the

environment in Singapore, which is confirmed by a surveillance of Legionella

bacteria in various artificial water systems (Heng et al., 1997) Despite the

ubiquitous distribution of Legionella in artificial water systems in Singapore and

high prevalence of sero-converted individuals, there has been no clustering of

cases by person and place and no common source outbreak linked to any artificial

water system since the disease was made notifiable in 1985 Goh et al (2005)

suggested that the absence of outbreak was due to the low prevalence of the highly

pathogenic Pontiac subtype of L pneumophila locally or low Legionella counts in

the cooling towers and other water systems here (only one fifth with Legionella

colony count above 10 colony-forming units (CFU) /ml)

During the period 1986 to 1996, a total of 258 sporadic cases of

community-acquired legionellosis was reported, giving a mean annual morbidity rate of 0.9

per 100,000 population (Heng et al., 1997) However, a total of 273 cases,

including 37 imported cases, were reported during the period 1998-2002, giving a

mean annual incidence rate of 1.7 per 100,000 population (Goh et al., 2005)

These are lower than that of the USA (20 per 100,000 population per year; Borella

et al., 2005) and Scotland (5.1 per 100,000 population), and comparable to that of

Denmark (1.8 per 100,000 population), Germany (1.6 per 100,000 population) and

England and Wales (0.2 per 100,000 population; Heng et al., 1997) In both

studies conducted in Singapore, cases were reported predominantly among males,

ethnic Indians, the elderly and those with concurrent medical conditions The

overall case-fatality rate was 14.7% for the period 1986-1996 (Heng et al., 1997)

and 5.5% for 1998-2002 (Goh et al., 2005)

Legionella pneumonia accounted for 2% to 7% of the community-acquired

pneumonia among hospitalized patients in Singapore (Ong and Eng, 1995) The

incidence of community-acquired pneumonia due to legionellosis in USA is 2-5%

(Marston et al., 1997), in Italy is 5.9% (Montagna et al., 2006) and in Brazil is

5.1% (Chedid et al., 2005) Thus in most countries, it is less than 10%

2.1.4.5 Treatment

In order to administer accurate treatment to patients with Legionnaires’ disease,

correct diagnosis is critical Unfortunately, it is not possible to clinically

distinguish patients with Legionnaires’ disease from patients with other types of

pneumonia (Edelstein, 1993) Furthermore, delay in starting appropriate therapy

has been associated with increased mortality (Heath et al., 1996) Thus, Bartlett et

al (2000) proposed that empirical therapy for persons hospitalized with

community-acquired pneumonia should include coverage for Legionnaires’

disease

Historically, erythromycin has been the drug of choice for Legionnaires’ disease

(Fields et al., 2002) In vitro data suggest that azithromycin and many

fluoroquinolone agents have superior activity against Legionella spp

Additionally, these agents have fewer side effects than erythromycin (Edelstein,

1998) Since azithromycin and levofloxacin has been licensed by the Food and

Drug Admininstration for the treatment of Legionnaires’ disease, thus they are

preferred over erythromycin (Fields et al., 2002)

2.1.5 Ecology of Legionella

2.1.5.1 Natural and man-made habitats

Water is the major reservoir for legionellae and the bacteria are found in

freshwater environments worldwide (Fliermans et al., 1981) Legionellae have

been detected in as many as 40% of freshwater environments by culture and in up

to 80% of freshwater sites by PCR (Fields, 2002) Furthermore, Legionella has

been shown to survive in marine waters (Heller et al., 1998) and even ocean

waters receiving treated sewage have been found to contain Legionella species

(Palmer et al., 1993) In contrast with the aquatic environment, L longbeachae is

a frequent isolate from potting soil (Steele et al., 1990) This species is the leading

cause of legionellosis in Australia and occurs in gardeners and those exposed to

commercial potting soil (Ruehlemann and Crawford, 1996)

L pneumophila multiplies at temperatures between 25ºC and 42ºC, with an

optimal growth temperature of 35ºC (Katz and Hammel, 1987) However, in

nature or in association with algae, the optimum growth temperature of Legionella

spp may be 45°C or higher (Fliermans et al., 1981) Most cases of legionellosis

can be traced to human-made aquatic environments where the water temperature

is higher than ambient temperature (Fields et al., 2002) Thermally altered aquatic

environments can shift the balance between protozoa and bacteria, resulting in

rapid multiplication of legionellae, which can translate into human disease

Legionellosis is a disease that has emerged in the last half of the 20th century

because of human alteration of the environment Left in their natural state,

legionellae would be an extremely rare cause of human disease, as natural

freshwater environments have not been implicated as reservoirs of outbreaks of

legionellosis Furthermore, the population densities of Legionella spp in

freshwater are extremely low and at the highest densities measured Legionella

spp account for less than 1% of the total bacterial population (Fliermans et al.,

1981)

Human infection occurs exclusively by inhalation of contaminated aerosols which

can be produced by air conditioning systems, cooling towers, whirlpools, spas,

fountains, ice machines, vegetable misters, dental devices and even shower heads

(Atlas, 1999) In addition, the presence of dead-end loops, stagnation in plumbing

systems and periods of non-use or construction have been shown to be technical

risk factors (Ciesielski et al., 1984; Atlas, 1999) Also, the material of the piping

system has been shown to influence the occurrence of high bacterial

concentrations In this respect, the use of copper as plumbing material may help to

minimize the risk of legionellosis whereas plastic materials support high numbers

of L pneumophila (Rogers et al., 1994)

2.1.5.2 Distribution of Legionella in Singapore

The first L pneumophila was isolated from hospital cooling towers in Singapore

(Nadarajah and Goh, 1986) Subsequently, Meers et al (1989) took 87 water

samples from 48 cooling-towers on 15 sites and isolated 19 strains of Legionella

from 7 of the sites All of the isolates are known to cause legionellosis (Yu et al.,

2002) 53% of the isolates belong to L pneumophila serogroup 1, which is the

dominant causal agent for both community- and hospital-acquired legionellosis

(Fang et al., 1989; Yu et al., 2002)

For the period of 1991 to 1996, the overall isolation rate of Legionella bacteria

was 36% (1107 positive samples / 3095 samples taken) for cooling towers, 33%

(2/6) for public showers, 29% (30/103) for indoor decorative fountains, 15%

(10/68) for outdoor decorative fountains, 15% (4/26) for outdoor man-made

decorative waterfalls and 2% (1/48) for spa pools (Heng et al., 1997) The

isolation rate was not correlated with rainfall The majority of the isolates (85.6%)

belonged to L pneumophila while 46.9% belonged to serogroup 1

Based on the samples collected during epidemiologic investigations in the period

of 1998 to 2002, Legionella bacteria were isolated from 550 (59.6%) of 923

cooling towers, 41 (38.3%) of 107 water fountains, 6 (16.2%) of 37 mist fans and

23 (23.7%) of 97 water taps/shower heads (Goh et al., 2005) Of 188 Legionella

bacteria isolated from cooling towers, L pneumophila was found to be the

predominant species (65.4%) while 50.4% belonged to serogroup 1 The

non-pneumophila isolates are known to cause legionellosis (Yu et al., 2002) and they

were L bozemanii (14.9%), L anisa (6.4%) and L dumoffii (1.6%) Legionella

bacteria count was also found to be generally low, with 39 (20.7%) cooling towers

exceeding 10 CFU/ml

2.1.5.3 Association of Legionella with protozoa

A universal trait of legionellae and LLAP organisms is their intracellular

existence These bacteria are capable of infecting and multiplying within a variety

of mammalian and protozoan cell lines (Fields, 1996) Most of these studies were

conducted with L pneumophila, primarily because this species is responsible for

the majority of legionellosis (Marston, 1994) It appears that L pneumophila may

also have the most extensive host range of legionellae (Fields, 1996)

Protozoa do not only provide nutrients for the intracellular legionellae, but also

represent a shelter when environmental conditions become unfavorable (Thomas

et al., 2004) Compared to in vitro grown L pneumophila, amoeba-grown bacteria

have been shown to be highly resistance to chemical disinfectants and to treatment

with biocides (Barker et al., 1992) Particularly inside Acanthamoeba cysts, the

bacteria are able to survive high temperatures, disinfection procedures and drying

(Rowbotham, 1986; Kilvington and Price, 1990; Winiecka-Krusnell and Linder,

1999) Furthermore, cooling tower amoebae containing legionellae may adapt to

biocides and may even be stimulated by biocides (Srikanth and Berk, 1993)

Legionella may also use protozoa to colonize new habitats where inhaled protozoa

represent a vehicle for effective transmission to humans (Cirillo et al., 1994) In

addition, these vehicles of respirable size of 1-5µm containing L pneumophila are

highly resistant to biocides (Berk et al., 1998) Interestingly, the same study

showed that more intense vesicle formation has been noticed before encystations

and when the amoeba are exposed to a mixed bacterial population, corresponding

to the conditions occurring in their natural environment Interaction of Legionella

and protozoa also contributes and enhances the infection process itself (Cirillo et

al., 1994; Cirillo et al., 1999) However, the underlying mechanisms of this

phenomenon are not well elucidated yet (Steinert et al., 2002) In addition,

Brieland et al (1997) demonstrated that L pneumophila-infected amoeba were

more pathogenic than an equivalent number of bacteria or co-inoculum of the

bacteria and amoeba A passage through Acanthamoeba castellanii was found to

reactivate viable but non-culturable (VBNC) Legionella into culturable state

(Steinert et al., 1997)

While protozoa are the natural hosts of legionellae, the infection of human

phagocytic cells is opportunistic (Fields et al., 2002) Much of our understanding

of the pathogenesis of legionellae has come from an analysis of the infection

process in both protozoa and human host cells Studies contrasting the role that

virulent factors play in these two host populations allow speculation on the

bacteria’s transition from their obligatory relationship with protozoa to their

opportunistic relationship with humans

2.1.5.4 Association of Legionella with biofilm

Biofilms are the primary site of Legionella growth and persistence In natural

hydrothermal areas, Legionella spp were isolated in higher numbers from

biofilms than from water (Marrão et al., 1993) Similarly, in a man-made model

potable water system, the bacteria are more easily detected from swab samples of

biofilm than from flowing water (Rogers et al., 1994) Thus, suggesting that the

majority of the legionellae are biofilm associated Furthermore, the association of

Legionella to biofilm may explain, at least in part, why legionellae are relatively

hard to eradicate in water systems, as biofilms exhibit a marked resistance to

biocidal compounds and chlorination (LeChevallier et al., 1988)

Only a limited number of studies attempted to characterize the bacteria’s

association within these complex ecosystems (Rogers and Keevil, 1992; Walker et

al., 1993; Rogers et al., 1994; Rogers et al., 1995) Rogers and Keevil (1992)

demonstrated that legionellae occurred in microcolonies within aquatic biofilm in

the absence of amoebae, thus providing the first evidence that legionellae is able

to grow extracellularly within the biofilm Walker et al (1993) evaluated the

effect of surface materials on growth of L pneumophila using gas

chromatography-mass spectrometry analysis of genus-specific hydroxy fatty

acids, while Rogers et al (1994) evaluated the effect of temperature and surface

materials on the growth of L pneumophila Rogers et al (1995) used biofilm

models to evaluate silver efficacy against L pneumophila and this study

represents a vast improvement over previous studies, which primarily evaluated

the susceptibility of agar-grown bacteria in sterile water