Development of Detection Methods for
Legionella longbeachae in Clinical and
Environmental Samples
Ali Mohammadi
A thesis submitted for the degree of
Doctor of Philosophy
at the University of Otago, Dunedin,
New Zealand
December 2019
Abstract
Legionella longbeachae is the predominant causative agent of Legionnaires’ disease (LD) in New Zealand with a peak infection period during the spring and summer months. This is a plant-associated microbe that can be found throughout the environment and is associated with potting mix and compost. LD is difficult to diagnose because of difficulty with obtaining respiratory samples.
Acid pre-treatment followed by culture is commonly used for the isolation of Legionella species in both clinical and environmental samples but has limited sensitivity because of high contamination loads. Detecting volatile organic compounds (VOCs) in breath is a minimally invasive technique that could be used to investigate patients with possible LD.
The aims of the current study were to: a) develop isolation methods to improve identification of L. longbeachae in clinical and environmental settings, b) develop an immunomagnetic separation (IMS) method to improve culture and PCR, c) identify possible VOCs using gas chromatography-mass spectrometry (GC-MS) that can be used as diagnostic biomarkers for LD. d) investigate potting mix productions for the presence of L. longbeachae, e) study the potential influence of physico-chemical characteristics of potting mix products that permit L. longbeachae to survive,
To develop the IMS method, a polyclonal antibody was raised in rabbits by immunising them with heat-killed L. longbeachae antigens with several booster injections. The antibody produced was separated from serum and purified by ion-exchange chromatography. This polyclonal antibody had high specificity and sensitivity in capturing L. longbeachae, with minimum cross-reactivity with other species. This antibody was coupled to immunomagnetic beads and used to separate organisms in stored respiratory samples for both culture and PCR. Feedstock samples used in the manufacture of the potting mix were taken from three sites around New Zealand for PCR and culture.
Culture results of sputum samples were significantly improved by using IMS. qPCR was used for this preliminary screening of environmental samples. The results showed that bark and bark-containing products had the highest number of PCR positive samples for L. longbeachae.
i
Culture results for these PCR positive samples were negative. Several isolation methods were developed to improve the culturability of L. longbeachae from environmental samples. The results showed that the combination of the Legionella selective GVPC antibiotic suspension with IMS reduces the contamination of cultures with other organisms and also improves the recovery of L. longbeachae from the samples.
Physico-chemical methods were used to determine the elemental make up of a subgroup of feedstock samples. The results showed that there were higher concentrations of boron and sulfur in PCR negative bark samples than PCR positive samples.
Finally, the headspace of the culture of L. longbeachae was pre-concentrated and analysed by GC-MS to generate a volatile profile of these bacteria. A peak of interest was found in this profile. Further work needs to be undertaken to identify the precise chemical structure of this compound.
In conclusion, this work has demonstrated that the rate of positive cultures of stored sputum samples is higher with IMS separation. This result needs validating in prospective studies in a diagnostic laboratory to determine if it can become a useful diagnostic tool. The sensitivity of culture of heavily contaminated environmental samples can be improved using IMS and GVPC decontamination and these techniques can be validated in future studies. Addition of elements such as sulphur and boron to potting mix should be evaluated as a method of manufacturing a safer potting mix product. Further studies are needed to precisely identify the chemical structure of a VOC identified and clinical trials conducted to determine whether this is of diagnostic value.
ii
Acknowledgements
Firstly, I would like to express my appreciation to my supervisory team, Prof Steve Chambers, Prof David Murdoch and Dr Amy Scott-Thomas. Thank you for your support and encouragement and providing me with this opportunity to conduct this research. I also thank the researchers at the infection group; Dr Sandy Slow, Dr Jonathan Williman and Dr Pippa Scott for their kind help in experimental design, data analysis and editing the thesis.
My special thanks to Mr Trevor Anderson and Mrs Ros Podmore for all their expertise and kind help. Going through this journey would not be possible without your great contribution. I also appreciate the amazing support and assistance I received from the staff of the microbiology, and virology departments at the Canterbury Health Laboratories.
I offer my sincere appreciation to Dr John Lewis and Dr Peter Holder for the great work of antibody production. A very special gratitude goes out to Dr Margaret Currie, Dr Jenny Jordan and Dr Ruth Helms for their care and huge encouragement to keep me going.
I would like to thank the University of Otago for providing me with the scholarship to undertake this PhD project.
I would like to acknowledge the great scientists I came to know during my PhD journey who kindly helped me to step into the Legionella world by inviting me to their laboratories, namely Mr David Harte of Legionella reference laboratory, ESR institute Kenepuru, Dr Harriet Whiley of the Flinders University, Adelaide, and Dr Carmen Buchrieser, the head of the biology of intracellular bacteria unit, Institut Pasteur, Paris.
I would also like to extend my thanks to all the staff and students at the Department of Pathology and Biomedical Science, and my friends on Christchurch campus who have always offered their help and support, and for making this whole experience more fun.
Special thanks to my aunt and uncle; Dr Sedigheh Delmaghani and Dr Asad Aghaei who inspired me to choose biomedical research as a labour of love. I also appreciate their amazing help making it possible for me to visit Institut Pasteur in Paris. Visiting its museum was a great motivational experience as well!
Finally, I wish to express my sincere gratitude to my family for their unconditional love and support during the hard days, and for believing in me. My lovely parents, Thank you!
iii
Publications and conference presentations pertaining to this thesis
1. Mohammadi, A., Chambers, S. T., Scott-Thomas, A., Lewis, J. G., Anderson, T., Podmore, R., Williman, J., Murdoch, DR., Slow, S. (2020). Enhancement of culture of Legionella longbeachae from respiratory samples using immunomagnetic separation and antimicrobial decontamination. Journal of Clinical Microbiology.
2. Presence of Legionella longbeachae in bark-containing potting soils, and bark from Pinus radiata and other tree species in New Zealand (in progress)
3. Enhancing the culturability and qPCR efficiency for Legionella longbeachae in sputum specimens using Immunomagnetic Separation. The 5th meeting of the ESCMID Study Group for Legionella Infections (ESGLI). Lyon, France. (2018)
4. Improved isolation of Legionella longbeachae bacteria from potting mix products. The 9th International Conference on Legionella. Rome, Italy. (2017)
5. The effect of media on the production of volatiles from Legionella species. International Association of Breath Research Conference. Vienna, Austria. (2015)
iv
Table of contents
Abstract i
Acknowledgements iii
Publications and conference presentations pertaining to this thesis iv
Table of contents v
List of tables xii
List of figures xiii
List of abbreviations xiv
1. Introduction and Literature Review 2
1.1. General introduction 2
1.3. Ecology of Legionella 3
1.4. Legionellosis 5
1.4.1. Clinical manifestations 5
1.4.2. Epidemiology 5
1.4.3. Presence of Legionella in growing media products 7
1.5. Pathogenesis 9
1.6. Prevention 12
1.7. Treatment 12
1.8. Laboratory diagnostics and limitations 13
1.8.1. Microscopy 14
1.8.2. Culture-based techniques 14
1.8.3. Immunology and Serology techniques 15
1.8.4. Nucleic acid amplification techniques 16
1.9. Breath volatiles and analysis 17
1.9.1. Microbial volatile organic compounds 18
1.9.2. Techniques to analyse VOCs in breath 18
1.10. Aims and Objectives 20
v
2. General Material and Methods 22
2.1. Materials 22
2.1.1. General materials 22
2.1.2. Buffers and solutions 22
2.1.2.1. HCl:KCl pH 2.2 buffer 23
2.1.2.2. Glycine vancomycin polymyxin cycloheximide (GVPC) antibiotic solution 23
2.1.2.3. Coupling and washing buffers for Dynabead M280 23
2.1.3. Synthetic sputum 24
2.1.4. Culture and identification of microorganisms 25
2.1.4.1 Bacterial strains 25
2.1.4.2. Primers and probes 26
2.1.4.2.1. ssrA gene screening assay PCR for diagnostic testing 27
2.1.4.5. Media 28
2.1.4.5.1. Buffered yeast extract broth (BYE) 28
2.1.4.6. Assay buffer for ELISA 29
2.2.4.7. Rabbits 29
2.2. Methods 29
2.2.1 Microbiological culture 29
2.2.2. Identification and confirmation of microorganisms 30
2.2.3. DNA extraction 30
2.2.4 Real-time PCR Assay 30
SECTION I: CLINICAL EXPERIMENTS 24
3. Development of a L. longbeachae-Specific Immunomagnetic Separation Method 34
3.1. Introduction 34
3.2. Materials and Methods 36
3.2.1. Bacterial strains 36
3.2.2. Rabbits for antibody production 37
vi
3.2.3. Production of polyclonal antiserum against L. longbeachae 37
3.2.3.1. Antigen preparation 37
3.2.3.2. Rabbit immunisation 37
3.2.3.3. Serum collection 38
3.2.3.4. Titration of the rabbit antisera 38
3.2.3.5. Purification of rabbit polyclonal antibody 39
3.2.3.6. ELISA assay for the purified IgG fractions 40
3.2.3.7. Determination of antibody concentration 40
3.2.3.8. Evaluation of serogroup-specific response of antiserum 40
3.2.3.9. Cross-reactivity test 41
3.2.4. Coupling antibody with magnetic beads 42
3.3. Results 43
3.3.1. Antigen preparation 43
3.3.2. ELISA screening of the serum 43
3.3.3. Purification of rabbit polyclonal antibody 44
3.3.4. Screening of the purified protein fractions 45
3.3.5. Cross-reactivity testing 47
3.3.6. Coupling antibody with magnetic beads 49
3.4. Discussion 50
4. Enhancement of culture and qPCR for respiratory samples using immunomagnetic separation 56
4.1. Introduction 56
4.2. Materials and Methods 59
4.2.1. Respiratory samples 59
4.2.2. Selective media and supplements 59
4.2.3. Acidic buffer 60
4.2.4. Acid pre-treatment of the respiratory specimens 60
vii
4.2.5. IMS-culture for the isolation of L. longbeachae from respiratory samples 60
4.2.6. DNA extraction kit 61
4.2.7. Primers and probes for ssrA qPCR 61
4.2.8. PCR mixture and thermal cycling programme 62
4.2.9. Generation of a standard curve for ssrA PCR using the synthetic sputum 62
4.2.10. DNA extraction and qPCR for Dynabead-treated samples 63
4.2.11. Stratification of Ct value of respiratory samples by pneumonia severity scores 63
4.2.12. Statistical analysis 64
4.3. Results 64
4.3.1. Bacterial culture of L. longbeachae after acid pre-treatment 64
4.3.2. IMS-Culture for isolation of L. longbeachae in respiratory samples 65
4.3.3. IMS-qPCR of respiratory samples from community acquired pneumonia patients 67
4.3.4. Assessment of disease severity in patients corresponding to the positive PCR for L. longbeachae and qPCR results 73
4.4. Discussion 74
5. Identifying Volatile Organic Compounds (VOCs) Released from L. longbeachae 81
5.1. Introduction 81
5.2. Materials and methods 83
5.2.1. Bacterial strains and culture 83
5.2.2. Breath samples 83
5.2.2.1. Headspace analysis of Legionella broth culture 84
5.2.2.2. SPME pre-concentration of culture headspace 84
5.2.2.3. Gas chromatography – mass spectrometry (GC-MS) 84
5.3. Results 85
5.3.1. Identification of a candidate volatile biomarker for Legionella 85
5.4. Discussion 89
SECTION II: ENVIRONMENTAL EXPERIMENTS 87
viii
6. Environmental Investigation of Potting Mix Products for L. longbeachae 92
6.1. Introduction 92
6.2. Materials and Methods 94
6.2.1. Study design 94
6.2.2. Production site sampling 95
6.2.3. DNA extraction 95
6.2.4. qPCR of potting mix materials for L. longbeachae 96
6.2.5. IMS-PCR 96
6.2.6. Data analysis 96
6.3. Results 97
6.3.1. qPCR of the potting mix products samples 97
6.3.2. IMS-PCR 100
6.4. Discussion 101
7. Culture of qPCR Positive Potting Mix and Development of The Isolation Methods for 107
7.1. Introduction 107
7.2. Material and Methods 108
7.2.1. Acid pre-treatment – culture for L. longbeachae 108
7.2.1.1. Stored feedstock and bagged product samples 108
7.2.1.2. Preparation of HCl:KCl buffer (pH 2.2) 108
7.2.1.3. Acid pre-treatment 108
7.2.2. Development of isolation methods for L. longbeachae from potting mix 109
7.2.2.1. Bacterial inoculum 109
7.2.2.2. Potting mix for spiking experiments 109
7.2.2.3. Dynabeads coupled with antibody 110
7.2.2.4. Anti-rabbit IgG antibody 110
7.2.2.5. GVPC solution 110
ix
7.2.2.6. Immunomagnetic separation prior to culture 110
7.2.2.7. Indirect immunomagnetic separation (i-IMS) 110
7.2.2.8. Sucrose gradient centrifugation 111
7.2.2.9. GVPC decontamination 111
7.2.2.10. GVPC/IMS 111
7.2.3. Isolation and identification of L. longbeachae 112
7.2.4. Determination of lower limit of detection 112
7.2.5. Statistical analysis 112
7.3. Results 113
7.3.1. Culture for qPCR positive feedstock and bagged potting mix products 113
7.3.2. Development of isolation methods for L. longbeachae from spiked potting mix 113
7.4. Discussion 115
8. The Physico-chemical Parameters and The Presence of L. longbeachae in Potting Mix Products 121
8.1. Introduction 121
8.2. Materials and Methods 123
8.2.1. Recording temperature of growing media samples 123
8.2.2. Measurement of pH of samples 124
8.2.3. Generating the elemental profile of raw materials used for potting mix products 124
8.2.4. Data analysis 126
8.3. Results 126
8.3.1. Evaluation of temperature of the growing media samples 126
8.3.2. Comparison of pH value of various sample types 129
8.3.3. ICP-OES analysis 130
8.3.4. Meteorological parameters 132
8.4. Discussion 133
9. General Conclusions and Future Perspectives 139
x
9.1. General discussion 139
9.2. Future perspectives 145
References 146
Appendices 177
Appendix 1: Details of the tested respiratory specimens 177
Appendix 2: Details of samples collected from three potting mix production sites 179
xi
List of tables
Table 2.1: List of materials and chemicals used in this work ...... 22 Table 2.2: Preparation of Daynabead coupling buffers ...... 24 Table 2.3: Ingredients for preparation of buffered based of synthetic sputum ...... 24 Table 2.4: Quantity of amino acids required for making synthetic sputum ...... 25 Table 2.5: List of microorganisms used in this study ...... 26 Table 2.6: Primers and probes utilised in ITS qPCR to detect genomic DNA ...... 26 Table 2.7: qPCR inhibitor control (ART) primers and probes ...... 27 Table 2.8 Component of the ITS qPCR assay for L. longbeachae ...... 27 Table 2.9: ssrA gene primers and probe ...... 27 Table 2.10: PCR reaction mix for the ssrA gene ...... 28 Table 2.11: Microbiological media purchased in this study ...... 28 Table 2.12: ITS PCR programme for L. longbeachae...... 31 Table 2.13: Thermal cycling condition for ssrA gene PCR ...... 31 Table 4.1: PCR cycling conditions ...... 62 Table 4.2: Clinical criteria for severity stratification for CAP ...... 64 Table 4.3: Cross-tabulation of culture results of acid treated samples ...... 65 Table 4.4: Comparison of stratified samples based on mean Ct values ...... 70 Table 6.1: Sample size of the three investigated potting mix products facilities ...... 94 Table 6.2: Details of sampling from three growing media production facilities ...... 95 Table 6.3: qPCR results of the samples collected from three different sites ...... 98 Table 6.4: Statistical description of Ct value of PCR positive samples ...... 99 Table 8.1: Details of samples randomly selected for ICP elemental analysis ...... 125 Table 8.2: Statistical description of samples temperatures ...... 127 Table 8.3: Mean temperature of collected samples in three potting mix production sites..... 127 Table 8.4: Comparison of temperature between samples PCR positive and negative for ..... 128 Table 8.5: Summary of the descriptive features for pH of the overall sample types...... 129 Table 8.6: Mean pH values of different sample types PCR positive and negative ...... 129 Table 8.7: Mean content of trace elements in samples determined by ICP/OES analysis..... 132
xii
List of figures
Figure 1.1: Main risk factors of contracting Legionnaires' diseases...... 6 Figure 3.1: Antibody titre in four rabbits based on light absorbance at 450nm...... 44
Figure 3.2: Average of light absorbance (OD450) of 1:10 dilution of ...... 45 Figure 3.3: Total protein fractions reactivity to anti-rabbit IgG HRP antibody...... 46 Figure 3.4: Specific reactivity of varying concentrations of Ig fractions...... 46 Figure 3.5: Specific reactivity of varying concentrations of protein fractions ...... 47 Figure 3.6: Cross-reactivity test between immobilised L. longbeachae serogroup 1 ...... 48 Figure 3.7: Cross-reactivity test between immobilised L. longbeachae sg 2 with...... 48 Figure 3.8: Amplification plot for DNA extracted from L. longbeachae sg1...... 49 Figure 4.1: Comparison of the recovery rate of L. longbeachae from ...... 66 Figure 4.2: A representative culture recovery comparison of...... 67 Figure 4.3: Standard curve for the spiked artificial sputum qPCR...... 68 Figure 4.4: Distribution of Ct values of qPCR positive samples for L. longbeachae...... 69 Figure 4.5: relationship between Ct value of initial qPCR and culture positivity...... 71 Figure 4.6: Comparison of Ct values of qPCR and culture results of acid treated samples .... 72 Figure 4.7: Comparison of mean Ct values and culture results for the stored specimens...... 73 Figure 4.8: Relationship between the Ct values of ITS qPCR and the severity score...... 74 Figure 5.1: Different respiratory sampling approaches categorised based on...... 81 Figure 5.2: Full scan chromatogram of BYE broth headspace...... 86 Figure 5.3: Experimental spectra for the discovered peak, obtained...... 87 Figure 5.4: Experimental spectra for the discovered peak, obtained under GC/MS EI...... 88 Figure 6.1: standard curve generated from the qPCR results of the spiked potting mix...... 97 Figure 6.2: Comparison of Ct value for qPCR with and without suing IMS method for ...... 100 Figure 7.1: Comparison of recovery rate of different methods for isolation of Legionella ... 114 Figure 7.2: Culture results of three different treatments on GVPC...... 114 Figure 8.1: Canonical discriminant analysis plot...... 130
xiii
List of abbreviations
ANOVA Analysis of variance ATCC American type culture collection SBA Sheep blood agar BCYE Buffered charcoal yeast extract BYE Buffered yeast extract CAP Community acquired pneumonia CFU Colony forming unit CNS Central nervous system CURB-65 Confusion, urea, respiratory rate, blood pressure, age over 65 DEAE Diethylaminoethyl DFA Direct fluorescent antibody EI Electron impact ELISA Enzyme-linked immunosorbent assay fmol/l Femto-mol per litre GC-MS Gas chromatography - mass spectrometry GU/L Genomic unit per litre GVPC Glycine vancomycin polymyxin cycloheximide HCl Hydrochloric acid HGT Horizontal gene transfer HRP Horseradish peroxidase HS Head-space HS-SPME Head-space - solid phase microextraction IMS Immunomagnetic separation KOH Potassium hydroxide LAI-1 Legionella autoinducer-1 LCV Legionella-containing vacuole LD Legionnaires’ disease m/z Mass to charge ratio MALDI-ToF Matrix-assisted laser desorption/ionization-time of flight MGB 3′-Minor groove binder mip Macrophage infectivity potentiator MWY Modified Wadowsky-Yee nmol/l Nanomol per litre PBS Phosphate buffered saline PFGE Pulsed-field gel electrophoresis PHB Poly-3-hydroxybuyrate PMN Polymorphonuclear leukocyte ppbv Parts per billion by volume PPR Pentatricopeptide repeat pptv Parts per trillion by volume qPCR Quantitative polymerase chain reaction RFLP Restriction fragment length polymorphism ROS Reactive oxygen species SPME Solid phase microextraction T4SS Type IV secretion system TMB Tetramethyl benzidine VBNC Viable but non-culturable VOCs Volatile organic compounds
xiv
CHAPTER 1
Introduction and Literature Review
1. Introduction and Literature Review 1.1. General introduction
Legionnaires’ disease (LD) is a type of serious but relatively uncommon, community acquired pneumonia (CAP) caused by Legionella species. These bacteria are present in a wide range of environmental niches and cause respiratory infections resulting in hospitalisation with substantial mortality rate especially in the elderly and immunocompromised individuals.
Research highlights an association between urban interventions to the environment with an increasing incidence of LD since the 20th century. For example, modern water storage and distribution systems and artificial water containers such as cooling towers, spas and fountains have been found contaminated with Legionella [1; 2], gardening activity is also connected with this infection amongst the elderly [3].
LD has a global incidence and a nationwide study showed that the average incidence rate of
LD is 5.4 per 100,000 of population in New Zealand [4]. The results of this study also revealed that Legionella longbeachae is the predominant cause of LD, three times more common than the globally dominant species of this genus; Legionella pneumophila in New Zealand. The peak infection period of LD due to L. longbeachae was also found to occur during spring and summer and cases were associated with the use of commercial growing media such as potting mix and compost products [4]. However, there is a lack of detailed knowledge of reservoirs and means of transmission from the environment to humans at present.
Recent studies in Canterbury, New Zealand have shown the high incidence of Legionella infections but many patients remain undiagnosed even with PCR technology due to a lack of appropriate sampling and diagnostic methods [5]. Therefore, development of new detection methods and improving the current methods is of high importance.
2
1.2. Microbiology of Legionella About half of the 55 different species of Legionella are known to be human pathogens, but generally these microorganisms can be considered as environmental microbes and accidental pathogens to humans [6]. Legionella are fastidious poorly motile gram negative coccobacillus- shaped bacteria with an average width of 0.6 μm and average length of 11 μm changing based on the colony age. The older bacterium expresses 1-2 flagella depending on the ambient temperature [7].
Several Legionella species are capable of parasitising free-living amoeba. The proliferation rate of different species of Legionella within protozoa is also different as L. pneumophila multiplies faster than Legionella micdadei does within Acanthamoeba species [8]. Legionella needs iron salts and the essential amino acid L-cysteine to grow on artificial media in the laboratory. They can grow on several media such as buffered charcoal yeast extract agar
(BCYE), Glycine Vancomycin Polymyxin Cycloheximide (GVPC) agar and Modified
Wadowsky-Yee (MWY) agar [8]. Legionella do not grow on sheep blood agar (SBA) media and cannot be identified by gram stain of sputum sample or blood culture [9].
1.3. Ecology of Legionella
Following the 1976 outbreak of LD ecological studies demonstrated this organism naturally lives in aquatic environments, contributes in biofilm formation and that these bacteria can be transmitted to humans via Legionella-containing water droplets [10; 11]. It has also been observed that this species can enter nematodes such as Caenorhabditis elegans and protozoa such as free-living amoeba in aquatic environments [12]. Metabolic studies on Legionella and its amoebic host such as Acanthamoeba castellanii using 13C labelled glucose source have revealed that the bacteria take up amino acids from their hosts and use them directly as the carbon and energy source, or use the host-derived carbon structures and exploit them for protein
3
biosynthesis [13]. The most important host amoebae identified to date include Tetrahymena,
Hartmanella and Acanthamoebae which act as Legionella reservoirs [14].
As Legionella and their amoebic host have a universal distribution in aquatic and soil environments, it is a challenge to keep water distribution and also composting systems clear from these microorganisms. L. pneumophila is the most commonly isolated species from water, but Legionella micdadei, Legionella bozemanii, Legionella dumoffii, Legionella anisa, and
Legionella feeleii can also be detected in these sources. The recovery of non-pneumophila species is more difficult than isolation of L. pneumophila, so that the true frequency of other species may be underestimated. Sediment and microflora of water systems can impose an inhibitory effect on growth of several species of Legionella.
Legionella can be washed through soil into groundwater or surface water, both of which are widely used as sources for drinking water [6]. Unlike L. pneumophila, L. longbeachae is often found in soil and composted plant material [15; 16] and human infection can occur through exposure to Legionella-containing aerosols or contaminated potting mix products [17].
These bacteria can utilise different internal and ambient signals in order to adapt to different ecological niches. Therefore, when these bacteria are exposed to different environments, their physiological features can be altered to exploit the surrounding environment. Legionella autoinducer-1 (LAI-1) is an example which is an α-hydroxyketone chemical to modify some of these characteristics. Horizontal genes transfer (HGT) is another strategy exploited by
Legionella to adjust to different environments. Interestingly, these bacteria possess both prokaryotic genomic islands and some eukaryotic genes which have a pivotal role in their battle against other microbes and also human phagocytes. Therefore, comprehensive knowledge of
Legionella ecology will guide us to better understanding of its pathogenesis for human [12].
4
1.4. Legionellosis
Legionellosis emerged during the last century and different species of Legionella can cause pneumonia and Pontiac fever. The main clinical manifestation of Legionella infection is pneumonia that can be fatal if not treated appropriately. Pontiac fever is a mild non-pneumonic self-limiting illness with headaches, fever, chills, and myalgia. Legionellosis can also be accompanied with generalised sepsis [18]. Legionella is considered as an incidental human pathogen and the onset of the disease starts after inhalation of an aerosolised water or soil containing the microbes, or aspiration of contaminated water [12].
1.4.1. Clinical manifestations
It is rare for healthy people to become sick after exposure to Legionella organisms. In contrast, vulnerable patients go through one of two clinical manifestations, LD or Pontiac fever [6].
Typically, there is lobar consolidation of the lung and cellular infiltration of neutrophils and macrophages into the alveolar spaces [19]. Most patients present with a fever up to 39.4oC.
Some other characteristics of LD include scant production of sputum, dyspnoea, mental status changes, diarrhoea, and depression after recovery. In Pontiac fever, patients experience an abrupt start of fever, chills, headache, 12-48 h after exposure to contaminated aerosols, but do not develop pneumonia in this form of legionellosis. Recovery takes 2-7 days without treatment
[20].
1.4.2. Epidemiology
The first outbreak of Legionella occurred in 1976 during an American Legion gathering in
Philadelphia, USA. Among the delegates, 182 individuals were infected and 29 patients died as a result [21]. This unknown microorganism was defined as a new bacterial genus called
“Legionella” in 1979. The causative agent of this outbreak was later named L. pneumophila
5
Philadelphia1. To-date the incidence of legionellosis disease worldwide has an increasing trend
[22; 23]. It is possible that the number of worldwide cases might be underestimated due to the difficulty in sampling and poor sensitivity of the tests performed, [18; 24].
Legionnaires’ disease has two different epidemiological settings: community-acquired and nosocomial infections. Some risk factors increase the susceptibility of individuals to Legionella infections including aging, underlying health background, poor immune system function, smoking and dose of bacterial exposure (Figure 1.1) [8; 18; 25-28].
Figure 1.1: Main risk factors of contracting Legionnaires' diseases.
The three main factors include: A) Susceptibility of patients depending on age, or compromised immune system due to underlying disease or immune suppressive medications; B) Transmission route and exposure-related factors such as the size of aerosols and droplets, aerosolisation and distance from source; C) Reservoirs providing favourable conditions for bacterial growth and survival. For example, Legionella amplification through infection of protozoa, and biofilm formation (Modified from Cooper et al., 2004 [29].
Nosocomial infections rarely happen and most of them are linked to hospital water systems.
The most frequent species responsible of these infections include L. pneumophila, L. micdadae,
6
L. bozemanii and L. dumoffii [30]. Control of water supply in hospitals is very important to ensure immune compromised patients are not exposed to Legionella species.
HIV infection is considered a risk factor for infection with Legionella species [31]. Immune suppression following bone marrow or solid organ transplantation increases the risk of L. pneumophila, L. micdadae, and L. bozemanii [31]. Outbreaks of LD has been reported with unusual Legionella species. For example, an outbreak of Legionella sainthelensi in an aged care home has been reported in Canada [32].
Legionellosis became a notifiable disease in New Zealand under the Health Act 1956 which has provided an important record of reported cases [33; 34]. These records show that cases of
LD caused by L. longbeachae are more frequent than those caused by L. pneumophila, but systematic epidemiological studies need to be conducted to find the scale of diseases caused by these infections in New Zealand. Between April and August 2005 there was an outbreak of
LD in Christchurch caused by Legionella. In this one outbreak, 125 cases of LD were reported.
[34; 35].
1.4.3. Presence of Legionella in growing media products
The first serogroup of L. longbeachae (sg1) was recognised from a clinical sample in Long
Beach, California [36]. The second serogroup (sg2) was isolated from the lung tissue of a pneumonia patient [37]. The route of transmission of Legionella from the soil is uncertain.
Most of the patients admitted to hospital are gardeners or people with a history of gardening and using potting mix products. This is considered as the main source of these infections.
Restriction fragment length polymorphism (RFLP) analysis of the clinical samples and soil isolates revealed that there is a close similarity between these isolates detected in soil that supported the idea of considering potting mix as a source of Legionella infection [38]. A study
7
was conducted with focus on epidemiology and the temporo-spatial distribution of different species of Legionella genus in New Zealand based on the available data collected for 29 years before 2007. The risk assessment of the environmental samples (n=420) showed that 58% of the cases were due to contact with compost [34]. The high risk of L. longbeachae following working with potting mix has been addressed in Australia as well [39].
L. longbeachae infection is not common in most countries of the world and this organism is responsible for less than 5% of cases of LD worldwide. A small number of cases have been documented in countries such as Scotland, Japan, Thailand and the USA [40], but studies indicate that L. longbeachae is more common in New Zealand and Australia [4; 41; 42]. Further studies on prevalence of L. longbeachae in different countries are required as the real number of LD cases due to L. longbeachae may be higher than what is considered at present.
L. longbeachae is the predominant species responsible for LD cases in Canterbury, New
Zealand, with its peak infectivity from September to March [5]. The high number of cases found in Canterbury was attributed to a change in the testing strategy for Legionella species instituted in the diagnostic laboratory. qPCR for Legionella was applied to all specimens from pneumonia patients submitted to the Canterbury Health Laboratories, Christchurch, New
Zealand (CHL). The dramatic increase in cases detected was unexpected but raised questions about the validity of the incidence of LD in previous reported studies [5].
It is possible that the number of worldwide cases might be underestimated due to the difficulty in sampling and poor sensitivity of the tests performed [18; 24]. Moreover, the surveillance data from different countries that used a variety of diagnostic tests should be compared with caution due to their different health surveillance systems. It is possible the higher incidence of
L. longbeachae cases in Australia and New Zealand are due to differences in the rate of contamination of commercial potting mix products with L. longbeachae.
8
Environmental investigation has demonstrated that several species of Legionella including L. longbeachae sg1 were present in about 75% of Australian potting mix samples containing pine species components. Samples from several European countries such as Greece, Switzerland and the United Kingdom had no Legionella contamination [43-45]. Several environmental factors such as temperature, humidity and availability of nutrients such as minerals have an impact on the viability and presence of Legionella in the environment. L. longbeachae can persist for up to 10 months in potting mix held at a temperature range between -20 and 35°C
[46]. The association between the presence of pine sawdust in potting mix and presence of
Legionella has not been determined yet. A culture-proven isolation of L. longbeachae sg1 from soil that does not contain pine products, has also been reported in Japan [47]; suggesting there are many sources of this organism. Pulsed-field gel electrophoresis (PFGE) has been used previously to investigating the relevance of compost isolates of L. longbeachae to clinical cases. This may also show genetic variabilities among different strains of this microorganism
[48]. However, with the development of advanced sequence-based methods such as whole genome sequencing, the source of Legionella infections can now be investigated accurately
[49].
1.5. Pathogenesis
In healthy individuals with a strong immune system, bacterial colonisation is controlled by a vigorous pro-inflammatory response comprising cell-mediated clearance of L. pneumophila from the lung [50]. L. pneumophila interacts with its hosts such as protozoa, nematodes and human alveolar macrophages via Dot / Icm type IV secretion system (T4SS). This system releases approximately 300 effector proteins containing eukaryotic-like domains which modulate signalling pathways and can induce the formation of the “Legionella-containing vacuole” (LCV) in order to evade the phagocytic killing of Legionella [12; 51].
9
Studies have demonstrated that a post-translational modification of histone H3 results in alterations in chromatin structure and consequently gene expression [52]. This phenomenon was due to a T4SS substrate Lpp1683 containing a eukaryotic SET (Su(var)3-9, Enhancer-of- zeste and Trithorax) domain which is known to catalyse lysine methylation, thus this substrate has the potential to remodel chromatin and influence the bacterial pathogenesis through the host chromatin modification [52]. Seemingly, this particular mechanism has evolved in
Legionella to support their replication in the host cell.
The life cycle of Legionella includes two forms based on their exposure to external stress: The transmissive form in which bacteria have motility, resistance to environmental stimuli, and infectivity. In contrast, the replicative form does not have these features but can replicate intracellularly [53]. The shift between transmissive and replicative forms helps the bacteria invade human phagocytes and proliferate inside the host cell’s phagosomal compartment. This change of mode occurs through metabolic regulations which switches this biphasic cycle in favour of one of these two modes, suggesting a robust connection between metabolism and virulence of these bacteria [51].
Caspase-3 activation is also exploited by Legionella to proliferate in the host macrophage by inducing apoptosis in these infected cells. Therefore, the bacteria form pores which result in the osmotic lysis of the host cell and release of these intracellular bacteria [54]. In a study using a human monocyte cell line U937 and Acanthamoeba polyphaga, it was observed that L. pnuemophila serogroup 1 has the ability to grow intracellularly inside the human macrophage and to induce pathogenicity. The level of pathogenicity of different serogroups of L. pneumophila is related to the difference in expression level of virulence of these bacteria in different hosts [55].
10
Most of the literature on pathogenesis of Legionella is focused on L. pneumophila. But recently, some studies have been accomplished on L. longbeachae to shed light on its pathogenicity. As these two Legionella species have different ecological niches, a genetic comparison of L. longbeachae D-4968 strain with L. pneumophila strains Philadelphia-1, Corby, Lens, and Paris was conducted [17]. The results suggested that several characteristic differences of L. longbeachae may explain its unique virulence and ecological adaptation. For example, due to lack of flagella encoding genes, L. longbeachae is incapable of affecting caspase-1 activation which triggers the different immune response in different strains of mice. For example, L. longbeachae is shown to colonise A/J, C57BL/6, and BALB/c mice while L. pneumophila cannot colonise BALB/c and C57BL/6 mice [17].
L. longbeachae has shown a characteristic virulence in mice. In an experimental study of
Legionella infection in mice, the majority of mortality among mice was observed following an intra-tracheal infection with L. longbeachae serogroup 1 [56]. Another unique feature of L. longbeachae is the presence of proteins with eukaryotic domains and degrading enzymes of the cell wall of plant cells in its proteome that suggests the possibility of L. longbeachae being a plant pathogen [57]. Pentatricopeptide repeat (PPR) proteins constitute one of the largest protein families in land plants. They are ubiquitous in eukaryotes with a mitochondrial genome but absent from most prokaryotes. The few bacteria such as L. longbeachae, that contain PPR proteins, are all pathogens or symbionts of eukaryotes [58]. There are still many unknown facts about pathogenesis of Legionella so further clarity of the pathogenesis of these bacteria will provide an appropriate diagnosis and treatment of LD.
11
1.6. Prevention
There is no vaccine available for the prevention of legionellosis and this disease is not considered as a communicable infectious disease as it is not transmitted from human to human.
However, there are some key measures that can be taken to reduce the risk of transmission of
Legionella from the potential reservoirs in the environment such as water distribution systems and composting facilities. Prudent measures can protect susceptible patients undergoing immunosuppressive therapies such as corticosteroids, cytotoxic chemotherapy and anti- rejection drugs for cancer treatment or transplants.
Several measures such as chlorination, superheating and copper-silver ionisation treatment have been suggested to prevent the growth of Legionella species in potable water [8]. It is also mandatory for the manufacturers to put warning labels on bagged potting mix about the potential contamination of L. longbeachae and instructions to open the bag using a protective face mask [8].
1.7. Treatment
Erythromycin used to be the drug of choice for LD, but because of its side-effects, especially for immunosuppressed patients, alternative antibiotics such as azithromycin, and quinolones such as moxifloxacin and levofloxacin are now prescribed most frequently [18; 59; 60].
Timely detection of Legionella and appropriate treatment including antibiotic therapy and management of respiratory, renal and central nervous system (CNS) complications are the main requirements for treating Legionnaires’ disease. Early administration of antibiotic therapy improves the outcome for those suffering from LD. Unlike the other causes of community acquired pneumonia (CAP) in which beta-lactam antibiotics are the first-line of treatment, these antibiotics are not effective against Legionella species. The reason for this lack of efficacy is
12
that Legionella are intracellular microorganisms, only the antibiotics which can penetrate into the host cells are effective. Therefore, macrolides and fluoroquinolones that have this feature are prescribed for these bacteria [61].
All Legionella species have a similar pattern of susceptibility to different antibiotic classes.
The antibiotic that is active on in vitro susceptibility testing works if it reaches an effective concentration inside the human macrophage cells. Fluoroquinolones such as levofloxacin, and ciprofloxacin are very effective. Erythromycin and newer macrolides including azithromycin, clarithromycin and roxithromycin have shown good activity against Legionella species but tetracyclines such as doxycycline are inactive. Newer macrolides also have a convenient once/twice daily administration and excellent oral bioavailability [8].
1.8. Laboratory diagnostics and limitations
The main factor in management of LD is employing a reliable discriminatory method with sufficient sensitivity and specificity to distinguish Legionella species from other causes of pneumonia. Application of different diagnostic methods for Legionella depends on the type of clinical specimen available as well. Urine antigen test is an accessible and cheap test that is mainly used only for one serogroup of L. pneumophila. Respiratory samples are sometimes difficult to collect, but if present qPCR is a reliable diagnostic test with high specificity and sensitivity [5]. Culture is insensitive but several precise techniques such as 16S rRNA analysis, cell wall fatty acid profiling and ubiquinone composition of the bacteria are available to identify them to species level [62]. These tests are available only in well-equipped specialised laboratories. Serological diagnosis relies on a rise in antibody titre and is only useful for epidemiological studies as results are not available in time to be clinically relevant.
13
Identification of L. longbeachae is carried out by assessment of antibody elevation, PCR and culture which requires reliable equipment and experience. In countries such as New Zealand,
Australia and Scotland, unlike the rest of the world, serology and PCR are commonly applied primarily to detect Legionella species including L. longbeachae [18]. The common specimen to be tested by PCR as the routine diagnostic method is sputum, however collection of this specimen from afflicted patients is difficult. With regard to the number of LD cases, the severity and potential fatality of this infection, and challenges for the sample collection, development of prompt, accurate and non-invasive detection methods for Legionella to prevent further complications is required.
1.8.1. Microscopy
Basic microbiological gram staining is not very helpful for detection of Legionella as it stains poorly [63]. Observing polymorphonuclear leukocytes (PMNs) without the presence of bacteria on a gram stained sample has some diagnostic value. Among Legionella species, L. micdadae is a weak acid-fast microbe following Kinyoun or modified Ziehl-Neelsen’s staining, and can therefore be mistaken as a Mycobacterial species [64].
1.8.2. Culture-based techniques
Culture of Legionella species from a respiratory sample is the most specific test for LD as
Legionella species do not colonise the upper or lower airways and is considered the gold standard for diagnosis [63]. Acid treatment of the sample is usually employed to inhibit the growth of contaminating microorganisms but can be inhibitory to growth of Legionella species reducing the sensitivity of this test [65]. Buffered Charcoal Yeast Extract agar (BCYE) is the
14
most commonly used medium. Addition of antibiotics however, has an inhibitory effect on the growth of some Legionella species [63].
Glycine Vancomycin Polymyxin Cycloheximide (GVPC) medium is a less inhibitory medium for Legionella species as all known species are resistant to vancomycin and polymyxin. Some
Legionella media contain dyes that can improve the detection of colonies of Legionella species.
For example, modified Wadowsky-Yee (MWY) medium contains bromocresol purple and bromthymol blue that facilitate the identification of bacterial colonies.
Colonies of different species of Legionella appear on the culture plates after 2-3 days and have the characteristic grainy “cut-glass” appearance [66]. On BCYE plates, L. pneumophila colonies are white/green, L. longbeachae appears grey, and both L. dumoffii and L. gormanii have a green appearance [67]. Although culture is a reliable diagnostic test, it is time- consuming to perform as Legionella has a fastidious nature and is less sensitive than PCR.
Potential cultures of Legionella species need to be confirmed with DNA or protein specific methods such as qPCR or matrix-assisted laser desorption ionisation time-of-flight mass spectrometry (MALDI-ToF) mass spectrometry [68].
1.8.3. Immunology and Serology techniques
Detecting the microbial antigens is aimed to find the active current infection, while the antibody testing shows the recent exposure of the patient to Legionella. Although antigen detection and sero-diagnosis (antibody) are used for Legionella diagnosis, they have some disadvantages.
Direct fluorescent antibody (DFA) test is a rapid test in which the presence of target antigens such as bacterial antigens is tested by direct binding to specific tagged antibodies and the result can be obtained in 2-4 hrs. This test has a high specificity of 96-99%, but its sensitivity is dependent on the number of bacteria present in the sample [67]. Therefore, the clinical samples
15
from the lower airway have a higher chance of detection of Legionella. The urine antigen test is a sensitive and very specific test which is commonly used in the laboratory for detecting L. pneumophila serogroup 1 antigens [69]. Therefore, the negative result cannot exclude other
Legionella infections. Immunofluorescence assay (IFA) and enzyme-linked immunosorbent assay (ELISA) are the most frequent serological test that are based on antibody presence methodology. 1:256 or higher titre has a diagnostic value for Legionella. 4-fold antibody increase between acute and convalescent serum is of diagnostic value as well [70].
Several antigen detection methods such as ELISA, radioimmunoassay (RIA) and latex agglutination tests are also used for the detection of Legionella lipopolysaccharide antigens that can be observed in most of the patients, but these are limited to detect L. pneumophila serogroup 1 which is considered to be a disadvantage [71].
1.8.4. Nucleic acid amplification techniques
Culture positive samples are often confirmed by PCR amplification of genes such as mip
(macrophage infectivity potentiator), 16S-23S rRNA and 23S–5S rRNA intergenic gene spacer regions [5; 72; 73]. Genotyping methods such as DNA sequencing, PFGE, RFLP and amplified fragment length polymorphism (AFLP) also help to distinguish different species of Legionella
[40; 74]. Real-Time PCR is another useful technique to detect Legionella from clinical and environmental samples. In terms of the Legionella quantity, approximating the bacterial load in the sample from PCR results should be undertaken carefully. For example, in an international study to monitor Legionella species in water systems, it was shown that the log mean difference between genomic unit per litre (GU/L) and colony forming unit (CFU) of bacteria of the same samples were significantly different, there are more GU/mL than CFU/mL [75]. Recently, a
PCR-based method called HIRA-TAN PCR (Human cell-controlled identification of the
16
respiratory agent and “TAN”, meaning sputum in Japanese) was developed to differentiate
Legionella species from commensal microorganisms in sputum samples [76].
1.9. Breath volatiles and analysis
Human breath is a complex mixture of many different small molecules including gases, vapours and aerosols representing the physiological status of the body. The four main groups of breath compounds can be classified as:
1. Endogenous human metabolome; metabolites originating from metabolic activity of the
cells.
2. Microbiome; metabolites produced by metabolic activities of microbiota or invading
microorganisms.
3. Food-derived metabolome; chemical compounds coming from digestion and
subsequent biotransformation of foods released by body tissues and the microbiota.
4. Exposome; compounds derived from inhaled ambient air or absorption via the skin.
Volatile organic compounds (VOCs) can be detected from many samples including breath.
These compounds are organic chemicals produced from metabolic activities of all living organisms. Some of these compounds are common to many species but others are relatively specific and can be applied as diagnostic targets for pathogenic microorganisms [77-80]. This is mostly appropriate for lower respiratory infections as obtaining clinical samples from patients is intrusive and can also be very laborious. Detection of VOCs in breath offers a possible diagnostic approach provided the causative agent does not colonise in the target organ and is metabolically distinct from other organisms.
17
1.9.1. Microbial volatile organic compounds
As part of microbial metabolism, different microorganisms such as fungi and bacteria produce a wide range of VOCs as the primary or secondary metabolites [81; 82]. Volatile organic compound profiles of a microbe depend on several factors such as species, nutritional sources and growth phase [83].
Most VOCs are common to different types of microbes, but some of them are unique to specific species. These microbial metabolites have been of increasing interest to be exploited as diagnostic targets. For example, four VOCs including methyl phenylacetate, methyl p-anisate, methyl nicotinate and o-phenylanisole have been identified from cultures of Mycobacterium tuberculosis and Mycobacterium bovis [79; 84].
The main advantages of breath analysis over many traditional “gold standards” is its non- invasive nature. It is also easy to repeat, and in most cases can be analysed rapidly. VOCs are promising targets for breath analysis as the kinetics of exhalation can be approximated according to substance solubility as well as most VOCs being stable [82]. Recent progress in analytical chemistry has assisted with the detection of many different substances in breath regardless of them being released as a direct result of host metabolism, host immune response or microorganism lung colonisation. Identification of these volatiles can provide helpful information for the diagnosis of many disorders such as pulmonary infection [77; 85], cancer
[86; 87] , organ dysfunction [88; 89] and metabolic disorders [90].
1.9.2. Techniques to analyse VOCs in breath
There has been an increasing interest in using breath analysis in clinical practice. However, the difficulty of detecting trace concentrations of some compounds has stimulated the development of different analytical techniques to identify and detect chemical compounds in exhaled breath.
18
By applying these methods, it is possible to study the VOC profile produced from different biological systems from various environments ranging from headspace of culture media to environmental air.
While there are many analytic techniques the most commonly applied techniques in breath analysis are gas chromatography mass spectrometry (GC-MS), selected ion flow tube mass spectrometry (SYFT) [91-93], proton transfer reaction mass spectrometry (PTR-MS) [94-96] and secondary electrospray ionisation mass spectrometry (SESI-MS) [97-100]. There are also other analytical techniques such as electronic nose [101-103], and also ion mobility spectrometry [104] techniques which are used to a lesser extent. Different analytical procedures have different detection limits where some are able to detect metabolites with concentrations as low as part per quadrillion (ppq) compared to others that can only detect part per billion
(ppb) concentration of molecules such as acetone, ethanol and ammonia in the laboratory air
[105].
Many studies reporting the identities of different VOCs in breath have used the GC-MS technique. This technique is a highly efficient method to study the chemical composition of samples such as headspace of culture and breath samples collected from individuals [82; 106].
19
1.10. Aims and Objectives
The aim of these studies is to elucidate the link between the environmental reservoir of this microorganism and human disease which is important for designing further epidemiological studies and preventive measurements to control the possible outbreaks and contamination sources. This requires an improvement of the current detection methods for both for clinical and environmental samples, specifically culture and qPCR methods. To achieve this goal, the following set of objectives were investigated to:
1. develop the isolation methods to improve identification of L. longbeachae in clinical and environmental samples,
2. develop an immunomagnetic separation (IMS) method to improve culture and qPCR for L. longbeachae.
3. identify possible VOCs in the headspace of cultures of Legionella species by GC-MS method that can be used as diagnostic biomarkers of LD.
4. investigate potting mix products for the presence of L. longbeachae,
5. study the potential role of physico-chemical features of potting mix products that permit
L. longbeachae to survive,
20
CHAPTER 2
General Material and Methods
2. General Material and Methods 2.1. Materials
2.1.1. General materials
All materials and chemicals used in this work are listed in Table 2.1.
Table 2.1: List of materials and chemicals used in this work
Material Provider Agar media Fort Richard laboratories, Auckland, NZ Gram staining kit Merk, Germany Ultrapure™ DNase/RNase-Free Distilled Invitrogen, Carlsbad, USA Water Bacto yeast extract Becton Dickinson L-Cysteine HCl Sigma Aldrich Iron pyrophosphate Sigma Aldrich ACES Buffer Sigma Aldrich Potassium hydroxide (KOH) Sigma Aldrich Hydrochloric acid (HCl) Sigma Aldrich L. longbeachae and internal control primers and ITS-P probe Integrated DNA Technologies (IDT)
TaqManTM MGB probe Life Technologies Complete Freund’s adjuant (CFA) Sigma Aldrich goat anti-rabbit horseradish peroxidase Jackson ImmunoResearch Laboratories, (HRP) conjugated antibody USA DEAE-Sephadex A-50 Sigma Aldrich Dynabeads M-280 Tosylactivated Invitrogen GenElute Bacterial Genomic DNA Kit Sigma Aldrich TaqManTM Gene Expression Master Mix Applied Biosystems Platinum® taq mix (10x) Life Technologies Sucrose – Molecular biology grade Sigma-Aldrich
2.1.2. Buffers and solutions
The following buffers and solutions were used in this work:
22
2.1.2.1. HCl:KCl pH 2.2 buffer
A mixture of hydrochloric acid and potassium chloride (HCl:KCl buffer) was used to pre-treat the stored respiratory specimens from Legionella PCR positive patients according to CHL microbiology laboratory protocol. The buffer was made fresh by mixing 39 mL of 0.2 M HCl with 250 mL of 0.2M KCl with a final molarity of 0.2 M. The pH was adjusted to 2.2 using
1M KOH, and was sterilised by autoclaving at 121°C for 15 min.
2.1.2.2. Glycine vancomycin polymyxin cycloheximide (GVPC) antibiotic solution
Glycine vancomycin polymyxin cycloheximide (GVPC) selective supplement (Oxoid,
Hampshire, England) was used in this work to decontaminate samples in order to isolated
Legionella. One vial of GVPC supplement was reconstituted into 500 mL sterile ultrapure water to be used as a washing solution. A vial of GVPC selective supplement contains ammonia free glycine (1.5 g), vancomycin hydrochloride (0.5 mg), polymyxin B sulphate (40000 IU) and cycloheximide (40.0 mg) per 500 mL.
2.1.2.3. Coupling and washing buffers for Dynabead M280
The following buffers described in Table 2.2 were prepared according to the manufactures’ instruction (Life technologies, Catalog No.1404) for coupling the antibody to Dynabeads:
23
Table 2.2: Preparation of Daynabead coupling buffers
Buffer Preparation 0.1 M sodium phosphate buffer, pH 2.62 g of NaH2PO4.H2O was mixed with 14.4 g Na2HPO4.2 7.4 H2O and its volume was adjusted with dH2O up to 1 Litre. 3 M ammonium sulphate buffer 39.64 g of (NH4)2SO4 was dissolved in 0.1 M sodium phosphate buffer and adjusted up to 100 mL. PBS pH 7.4 with 0.5% (w/v) tween 20 0.88 g NaCl was mixed with 0.5% (w/v) BSA and added to 80 mL of 0.01 M sodium phosphate buffer pH 7.4. The final volume was adjusted to 100 mL with 0.01 M sodium- phosphate pH 7.4. PBS pH 7.4 with 0.1% (w/v) tween 20 0.88 g NaCl was mixed with 0.1% (w/v) BSA and added to 80 mL of 0.01 M sodium phosphate buffer pH 7.4. The final volume was adjusted to 100 mL with 0.01 M sodium- phosphate pH 7.4.
2.1.3. Synthetic sputum
To prepare one litre of synthetic sputum, the buffered base was made with the following ingredients described in Table 2.3.
Table 2.3: Ingredients for preparation of buffered based of synthetic sputum
Ingredient quantity 0.2 M NaH2PO4 6.5 mL 0.2 M Na2HPO4 6.25 mL 1 M KNO3 0.348 mL NH2Cl 0.122 g KCl 1.114 g NaCl 3.03 g 10 mM MOPS 0.41 g ddH2O 779.6 mL
Next, 100 mM stock solutions of the amino acids described in Table 2.4 were prepared and required volume for each amino acid was added to the buffered base. Tryosine, aspartate & tryptophan were resuspended in 1 M, 0.5 M and 0.5M NaOH respectively.
24
Table 2.4: Quantity of amino acids required for making synthetic sputum
Amino acid* Volume Amino acid Volume Amino acid Volume (mL) (mL) (mL) Aspartate 8.27 Threonine 10.72 Serine 14.46 Glutamate.HCl 15.49 Proline 16.61 Glycine 12.03 Alanine 17.8 Cysteine.HCl 1.6 Valine 11.17 Methionine 6.33 Isoleucine 11.2 Leucine 16.09 Tyrosine 11.17 Phenylalanine 5.3 Ornithine.HCl 6.76 Lycine.HCl 21.28 Histidine.HCl 5.19 Tryptophan 0.13 Arginine.HCl 3.06 * Initial concentration of all the amino acids was 100 mM
After the addition of these amino acids to the buffered base, the pH was adjusted to 6.8 with 1
M NaOH and the mixture was filter sterilized through a 0.22 µm pore filter. After sterilization,
1.754 mL of 1 M CaCl2, 0.606 mL of 1 M MgCl2 and 1 mL of 3.6 mM FeSO4.7H2O were added to the mixture. The synthetic sputum was stored at 4°C to be used for the experiments on the following day.
2.1.4. Culture and identification of microorganisms
2.1.4.1 Bacterial strains
ATCC and clinical strains of L. longbeachae serogroups 1 and 2 were obtained from
Canterbury Health Laboratories, Christchurch, New Zealand that had been stored at -80°C.
These bacteria were cultured on BCYE agar (Fort Richard Laboratories, Auckland, New
Zealand) for three days. To check the purity of the culture, five colonies were selected.
Identification of the selected colonies was completed by MALDI-ToF mass spectrometry on a
Microflex LT system (Bruker Daltonics, Bremen, Germany), which used the MALDI Biotyper version 3.1 database for species identification.
The microorganism that were used in Chapter 3 are listed in Table 2.5.
25
Table 2.5: List of microorganisms used in this study
Microorganism Reference number L. longbeachae serogroups 1 CHL laboratories, Canterbury, NZ L. longbeachae serogroups 2 CHL laboratories, Canterbury, NZ L. pneumophila serogroup 1 ATCC 33152 L. pneumophila serogroup 4 NZESR 3001 L. micdadei NZESR 2609 Aspergillus fumigatus CHL laboratories, Canterbury, NZ Pseudomonas aeruginosa ATCC 27853
2.1.4.2. Primers and probes
To conduct qPCR for detection of Legionella, primers with a PCR product of 259 bp length and a L. longbeachae specific probe were used as described in Table 2.6 [73; 107].
Table 2.6: Primers and probes utilised in ITS qPCR to detect genomic DNA of L. longbeachae
Primers and probes Sequence Forward primer (LegITS) 5’-GTACTAATTGGC TGA TTGTCTTGACC-3’ Reverse primer (LegITS) 5’-CCTGGCGATGACCTACTTTCG-3’ MGB* TaqManTM probe (Llo-ITSp) 5’- VIC-TATCATGCCAATAATGCGCGA-3’BHQ
* MGB – Minor groove binder
An artificial construct (abbreviated as ART) was added to PCR templates as inhibition control.
Primers and probes for the internal control are described in Table 2.7. ART is constructed from a reference sequence with Accession number U17140 in GenBank, inserted into a plasmid carried by competent E.coli bacterial cells. This construct was designed and made by the CHL laboratories, Canterbury, New Zealand. ART PCR product is 158 bp in length.
26
Table 2.7: qPCR inhibitor control (ART) primers and probes
Primers and probes Sequence Forward primer (ARTF) 5’-AGCGGTGACGCATGCCTTCCA-3’ Reverse primer (ARTR) 5’-CAAAGGAGACATTCTCACGCTACAGTT-3’ TaqManTM Probe 5’CY5-AACACCAAGTGGCCTTTCAGGCTGCGCGACT-3’BHQ
The components of the reaction mix for ITS PCR are described in Table 2.8:
Table 2.8 Component of the ITS qPCR assay for L. longbeachae including the internal control (ART)
Component Volume/per well (µL) Final concentration TaqManTM Gene Expression Master Mix (2x) 12.5 1× 50 µM Forward primer 0.5 1.0 µM 50 µM Reverse primer 0.5 1.0 µM 20 µM MGB TaqManTM probe 0.5 0.2 µM 50 µM ART Forward primer 0.1 0.2 µM 50 µM ART Reverse primer 0.1 0.2 µM 10 µM ARTCy5 probe 0.25 0.1 µM H2O 4.55 - Template DNA 5 X ART internal control template 1 X Total 25 µL
2.1.4.2.1. ssrA gene screening assay PCR for diagnostic testing
Primers (500 nmol/L) and probe (100 nmol/L) specific for ssrA gene (Genbank accession no.AE017354) as Legionella genus specific PCR according to Benitez et al. 2013 are listed in
Table 2.9.
Table 2.9: ssrA gene primers and probe
Oligonucleotide Sequence ssrA Forward primer 5’-GGCGACCTGGCTTC-3’ ssrA Reverse primer 5’-GGTCATCGTTTGC ATTTATATTTA-3’ ssrA MGB probe 5’-FAM-ACGTGGGTTGCAA-NFQ-3’
The following mixture at Table 2.10 was prepared to conduct real-time PCR targeting ssrA gene.
27
Table 2.10: PCR reaction mix for the ssrA gene
Component Volume/per well (µL) Final concentration Platinum® Taq mix (10x) 2.5 1× MgCl2 (50 mM) 2 4 mM dNTP’s (10 µM) 0.5 0.2 mM 50 µM ssrA Forward primer 0.25 0.5 µM 50 µM ssrA Reverse primer 0.25 0.5 µM 10 µM ssrA probe 0.25 0.1 µM 50 µM ART Forward primer 0.1 0.2 µM 50 µM ART Reverse primer 0.1 0.2 µM 10 µM ARTCy5 probe 0.125 0.05 µM H2O 8.675 - Platinum® Taq DNA polymerase 0.25 2U / rxn Template DNA 10 X Total 25 µL
2.1.4.5. Media
Most of the microbiological media including and selective and non-selective media for
Legionella were purchased as listed in Table 2.11.
Table 2.11: Microbiological media purchased in this study
Medium provider Sheep blood agar buffered charcoal yeast extract (BCYE) agar Fort Richard Modified Wadowsky and Yee (MWY) agar Laboratories, Glycine Vancomycin Polymyxin Cycloheximide (GVPC) agar Auckland, New Zealand
Some media were prepared in the laboratory as below:
2.1.4.5.1. Buffered yeast extract broth (BYE)
BYE broth was applied during this research in order to prepare bacterial suspensions for spiking tests and also for the analysing the volatile compounds released from the Legionella.
To prepare a litre of BYE broth, base medium was first prepared including 10 g yeast extract,
ACES buffer 10 g and 980 mL distilled water. pH was adjusted to 6.9 using 1M KOH. The solution was autoclave at 121ºC. After cooling down to 50°C, Solution A (0.4 g L-Cysteine
28
HCl dissolved in 10 mL distilled water) and solution B (0.25 g iron-pyrophosphate dissolved in 10 g distilled water) which were filter sterilized by a 0.22 µm filter unit were added. pH was checked again and adjusted if necessary.
2.1.4.6. Assay buffer for ELISA
To prepare assay for buffer for ELISA, 100 mL PBS buffer was made containing 0.1% gelatin
(w/v) and 0.1% Tween 20 (v/v).
2.2.4.7. Rabbits
All animal work was performed at the Christchurch Animal Research Area (CARA) facility,
University of Otago Christchurch. Four female New Zealand white rabbits (Oryctolagus cuniculus) were housed in CARA under specific pathogen-free (SPF) conditions and were fed a standard commercial diet. This work was approved by the University of Otago Animal Ethics
Committee (AEC) approval C8/15C.
2.2. Methods
2.2.1 Microbiological culture
Legionella was cultured in two different ways depending on the nature of the experiments: broth culture and culture on solid agar plates. For sub-culturing Legionella, BCYE agar was used as a non-selective medium for Legionella. Sheep blood agar was also applied in this research as a negative control for Legionella growth. Broth culture was performed by growing the bacteria using BYE broth and incubating the medium at 36°C in an orbital shaking incubator
(Ratek, Newcastle, Australia) at 150 rpm for at least 36 h.
29
2.2.2. Identification and confirmation of microorganisms
Olympus SZX10 binocular zoom stereomicroscope (Olympus, Japan) was applied for examining the characteristics of Legionella-like colonies grown on the plates. To confirm the identity of the organisms, Microflex LT (Bruker Daltonics, Bremen, Germany) was used for
MALDI-ToF mass spectrometry method was applied.
2.2.3. DNA extraction
The bacterial genomic DNA was purified from different type of samples including environmental and clinical specimens using the commercial GenElute Bacterial Genomic DNA
Kit (Sigma Aldrich) following manufacturer's instructions. This kit was previously optimised and selected to use by Microbiology unit at CHL laboratories, Canterbury, New Zealand.
2.2.4 Real-time PCR Assay
PCR programme for Applied Biosystems® 7500 (ABI, Thermofisher Scientific, USA) to detect ITS target is described in Table 2.12.
30
Table 2.12: ITS PCR programme for L. longbeachae
Step Temperature (ºC) Time (h:min:s) Cycle number Denaturation 94 00:10:00 1 Amplification 94 00:00:15 50 60 00:01:00* Cool down 40 00:01:00 1
* Acquisition of fluorescence
The ssrA PCR was performed using a Roche Lightcycler 480 machine as previously optimised for ssrA gene as a clinical screening procedure at CHL laboratories. The thermal cycling conditions are described in Table 2.13.
Table 2.13: Thermal cycling condition for ssrA gene PCR
Step Temperature Time (h:min:s) Cycle number (ºC) Denaturation 95 00:10:00 1 Amplification 95 00:00:15 50 Final extension 60 00:01:00 50
31
SECTION I: CLINICAL EXPERIMENTS
I: CLINICAL EXPERIMENTS
CHAPTER 3
Development of a L. longbeachae-Specific Immunomagnetic Separation Method
3. Development of a L. longbeachae-Specific Immunomagnetic Separation Method 3.1. Introduction
The detection and isolation of Legionella is challenging for specialist laboratories. While culture and Gram staining have been used for the identification of Legionella, it is difficult to grow these bacteria in the laboratory. Legionella do not stain well and prove difficult to culture.
Legionellae are fastidious organisms and must be grown on specialised bacteriological media.
The bacterial colonies appear in culture after three to five days incubation on BCYE agar (pH
6.9), which is still the mostly common non-selective medium for culture since its development
[24]. The sensitivity of culture of Legionella from patient sputum is not satisfactory. In many cases, Legionella is absent by culture of respiratory specimens while positive for PCR and urinary antigen test [108]. There are five common methods widely used to detect Legionella in the diagnostic laboratory, as discussed in Chapter 1. These are culture and PCR from respiratory secretions, direct and indirect fluorescent antibody staining, and the urine antigen test [71].
Identification of Legionella at the genus level is sufficient for treatment as all species have similar antibiotic susceptibility. Improving diagnostic methods for the detection of Legionella is an ongoing challenge for diagnostic microbiology. Improved detection of these bacteria will be advantageous for epidemiological studies such as outbreak investigations and tracing the reservoirs of infection. Current pre-treatment methods for processing respiratory samples such as acid and heat treatment were developed using L. pneumophila, not L. longbeachae and these pre-treatment methods have not been optimised for L. longbeachae [109; 110].
These two organisms have very different ecological niches. L. pneumophila is naturally a water-borne organism and L. longbeachae is found in soil and composted plant materials [111].
As such, they both have their own specific characteristics that should be considered when developing isolation techniques for detection in clinical and environmental samples. For
34
example, L. pneumophila has flagella, while L. longbeachae does not and L. longbeachae but not L. pneumophila has a capsule [57]. Developing a novel method to improve the isolation of viable L. longbeachae cells has been a continuous effort in clinical microbiology laboratories.
Development of rapid laboratory methods has been progressing in biomedical research.
Immunomagnetic separation (IMS) is a technique of increasing interest as it can be combined to improve the current diagnostic methods such as culture, PCR, and flow cytometry [112-
114]. IMS has therefore been used to enhance identification of a wide variety of organisms including several species of bacteria, fungi, viruses and parasites [115-118]. The principle of
IMS is to employ paramagnetic beads coated with specific antibodies that can bind to antigens on the surface of cells. IMS can then be used to concentrate the target bacteria and separate them from undesirable organisms present in the sample [119]. The IMS method was originally established to isolate blood cells for clinical purposes [120], but the application of the IMS technology has been expanded for isolating microorganisms from a variety of matrices [121-
125].
Currently, PCR and culture are the most common methods used to identify L. longbeachae.
Depending on the sample matrix, there are accompanying organisms which can overgrow resulting in a failure to isolate Legionella [126]. Chemicals in the samples, such as humic acid, can cause failure of the PCR method [127]. Several preparation methods such as heat treatment, acid pre-treatment and serial dilution of samples have been performed to improve the outcome of these detection techniques. However, these preparation methods are time-consuming and decrease the chance of detecting low numbers of bacteria by diluting the samples.
IMS has been used previously to isolate some Legionella species. For example, in a study on water samples, magnetic beads coupled with a genus-specific monoclonal antibody against an epitope of macrophage infectivity potentiator (MIP) protein were used. However, the recovery
35
rate varied from 30% to 60% depending on the species of Legionella. Furthermore, there was a cross-reaction of non-Legionella organisms and non-specific growth of these microbes in culture. However, this study did not include L. longbeachae [128].
This chapter describes the development of a polyclonal antibody specific for L. longbeachae and an IMS method to capture this microorganism which will be combined with culture and
PCR methods in next chapters. The hypothesis was that IMS capture and separation used prior to PCR and culture will improve the sensitivity of these detection techniques particularly if low numbers of bacteria are present in the samples. To do so, the production of a polyclonal antibody in rabbits was designed. Polyclonal antibodies have their own advantages. The process is relatively fast and economic. Furthermore, as they are secreted by different clones of B lymphocytes, they are heterogeneous and can bind to different epitopes of the antigen of interest. The principle for the selection of appropriate animals to produce polyclonal antibodies is based on a) the required quantity of antibody, b) the feasibility of blood collection, c) the phylogenetic distance between the target antigen and the selected animal. Several types of mammals have been used to produce antibodies, of which rabbits are the most used. Rabbits are readily handled and bled, have reasonable life-span, and can generate high-titre of the antibodies in their sera with a good level of affinity [129]. Because of these advantages, rabbits were selected for this study. The developed IMS method in this chapter was applied in Chapter
4 for the respiratory specimens.
3.2. Materials and Methods
3.2.1. Bacterial strains
Clinical strains of L. longbeachae were used in this study as described in Section 2.2.4.1.
36
3.2.2. Rabbits for antibody production
Rabbits were used in this chapter as described earlier in Section 2.2.4.7. These rabbits were used for production of polyclonal antibody.
3.2.3. Production of polyclonal antiserum against L. longbeachae
3.2.3.1. Antigen preparation
L. longbeachae serogroup 1 and serogroup 2 were cultured on BCYE agar. Sheep blood agar
(5% SBA) was used to ensure there was no contamination. Three-day-old cultures were first examined by Gram stain according to the manufacturer’s instructions (77730-1KT-F bacterial staining kit, Merk®, Germany) and their identification was confirmed using MALDI-ToF mass spectrometry. Proven L. longbeachae colonies were harvested and suspended in 500 µL sterile phosphate buffered saline (PBS), pH 7.2. The suspension was centrifuged for 3 min at 7000 × g, re-suspended in 1 mL PBS and incubated at 60°C for 30 min to kill the bacteria. This temperature was selected instead of boiling at 100°C in order to keep the structural integrity of the surface antigens of the bacterial cells. The final concentration of the bacterial suspension was adjusted to 200 μg/mL in sterile PBS. To evaluate the viability, an aliquot of the bacterial suspension (100 µL) was plated onto BCYE agar and incubated for three days (Fallon, 1982).
3.2.3.2. Rabbit immunisation
The rabbits were immunised according to an in-house protocol developed in the Endocrinology and Steroid laboratory, Canterbury Health Laboratories. The heat-killed cells of L. longbeachae serogroup 1 and serogroup 2 were washed twice in PBS. The prime inoculum was
500 μL of the bacterial suspension (200 μg/mL) emulsified with an equal volume of Freund’s complete adjuvant (Sigma-Aldrich, St. Louis, MO, USA). Four 6-month–old female New
37
Zealand White rabbits were selected for the immunisation with two rabbits per serogroup
(rabbit A and B were injected with the serogroup 1, and rabbit C and D with the serogroup 2 bacterial suspension). The rabbits were injected subcutaneously with the antigenic emulsion.
The first and second booster injections were done according to CHL routine protocols with injections administered on days 20 and 35 respectively, using incomplete Freund’s adjuvant.
A final injection on day 40 used sterile PBS buffer without any adjuvants in order to increase the blood volume in rabbits before taking blood from them [130].
3.2.3.3. Serum collection
Blood was drawn from the rabbits through cardiac puncture 10 days after the second booster injection and was stored at 4°C for 2 h to allow the sample to clot. The coagulated blood was separated from the serum by centrifugation 1500 g for 5 min.
3.2.3.4. Titration of the rabbit antisera
Collected sera were tested for L. longbeachae antibodies by an ELISA assay to measure the titre of rabbit antibody production. To coat the ELISA plates (Falcon 3912 Microtest III,
Becton Dickinson Co., Oxnard, California, USA), 2 µL of heat-killed L. longbeachae suspension in 10 mL of 6M aqueous guanidine-hydrochloride (Sigma-Aldrich, St. Louis, MO) with 100 µL/well concentration was added and the plate was incubated overnight at room temperature. Following the coating, the plates were washed 4 times with wash solution (normal saline containing 0.1% v/v Tween 20) and then blocked by the addition of 200 µL/well of assay buffer (section 2.2.4.6).
Serial dilutions of the rabbit sera in assay buffer were added to the L. longbeachae coated plates for 1 h at room temperature. The plates were then washed and goat anti-rabbit horseradish
38
peroxidase (HRP) conjugated antibody (Jackson ImmunoResearch Laboratories, Inc., West
Baltimore, PA, USA) at a concentration of 0.8 mg/mL was added (100 µL/well diluted 1:1000 in assay buffer) and incubated for 1 h at room temperature. The plates were finally washed with a PBS buffer. To prepare the substrate, a 600 mL aqueous solution containing 8.2 g anhydrous sodium acetate and 3.6 g citric acid was added to 400 mL of methanol containing 270 mg of tetramethyl benzidine (TMB). The substrate was added (100 µL/well) to each well for colour development. This reaction was stopped after incubation at room temperature for 5 min, through the addition of 1 M HCl (100 µL/well) to the plate and the absorbance was read at 450 nm on a Fluostar Galaxy reader (BMG Technologies, Germany). The serum with the highest antibody titre was selected for purification of the target polyclonal antibody.
3.2.3.5. Purification of rabbit polyclonal antibody
Ion-exchange chromatography was used to purify the antibody against L. longbeachae. A chromatography column (height 10 cm, 3 cm diameter) with a bed volume of 50 mL was prepared by packing it with DEAE-Sephadex A-50 (Sigma Aldrich). The column was equilibrated with 10mM phosphate buffer, pH 6.5. Ten mL of serum from rabbit A was added to the column and then washed with 10mM phosphate buffer, pH 6.5. The unbound material was washed off to obtain a quantitative immunoglobulin yield from serum according to the original method described by Baumstark [131]. Thirty fractions were collected (5-10 mL). The light absorbance monitored at 280 nm (OD280) with the UV-Visual spectrophotometer Cary-50 v3.0 (Agilent, Santa Clara, CA, USA). Proteins absorb UV light at 280 nm strongly due to three main constituent amino acids; tryptophan, tyrosine and cysteine.
39
3.2.3.6. ELISA assay for the purified IgG fractions
A semi-quantitative ELISA assay was performed to determine rabbit immunoglobulin concentration in each column fraction. Based on the generated chromatogram from OD280 absorbance reading, fractions 1 to 6 contained no protein and were hence discarded. Fractions
7 to 30 were diluted (1:10, 1:100, 1:1000 and 1:10000) in PBS and were added (100 µL) to the microtiter ELISA plate and incubated overnight at room temperature.
Following coating, the wells were washed and then emptied by inversion. 1:1000 dilution of
Conjugated goat anti-rabbit IgG HRP in assay buffer was added to the wells (100 µL/well).
After 15 min incubation at room temperature, the plate was washed and TMB substrate added as described previously. Following colour development, the plate was stopped, and the absorbance was read at 450 nm as before.
3.2.3.7. Determination of antibody concentration
Direct Detect® Infrared-based spectrometry (Merck Millipore, Danvers, USA) was used to determine the concentration of proteins in the serum. The purified antibody (2 µL) was added onto three spots on a sample card surrounded by a hydrophilic polytetrafluoroethylene (PTFE) membrane. The fourth position was used as the blank by adding 2 µL of 10mM phosphate buffer, pH 6.5. The IR-based quantification of the amide bonds in protein chains was carried out by reading the sample card and the average concentration of proteins was calculated by its software.
3.2.3.8. Evaluation of serogroup-specific response of antiserum
The specific responsiveness of the fractionated antiserum from rabbit A immunised with L. longbeachae sg1 antigens against both serogroups of this microorganism was evaluated by an
40
ELISA assay. Two separate microplates were set up by coating the wells overnight with heat- killed L. longbeachae serogroup 1 or serogroup 2 antigen preparation (each 5 µg/well) in 6M aqueous guanidine HCl. The plates were washed and blocked in assay buffer as described previously, and polyclonal antiserum fractions at four dilutions (1:10; 1:100; 1:1000 and
1:10,000) in assay buffer were added and incubated at room temperature for 1h. Plates were then washed and incubated with anti-rabbit IgG-HRP (1:1000 in assay buffer) followed by washing, TMB substrate was added and further processing as described previously. The antibody fractions with the best titres against L. longbeachae serogroup 1 were pooled and subsequently used for the conjugation to magnetic beads.
3.2.3.9. Cross-reactivity test
An ELISA assay was designed to determine the specificity of the raised antibody and any possible nonspecific cross-reactions with other species of Legionella and non-Legionella microorganisms [132]. This experiment was conducted according to the established procedure of the department of biochemistry, Canterbury health laboratories, New Zealand (Dr. John
Lewis, Personal communication). Several non-L. longbeachae species of Legionella and other organisms that can be found either in environmental or respiratory tract samples were selected, including L. pneumophila serogroup 1 (ATCC 33152) and serogroup 4 (NZESR 3001), L. micdadei (NZESR 2609), Aspergillus fumigatus (clinical isolate, CHL laboratories) and
Pseudomonas aeruginosa (ATCC 27853).
To prepare heat killed antigenic suspensions of these organisms, the Legionella strains were grown on BCYE agar and the other organisms were grown on 5% sheep blood agar plates. The grown organisms were collected from the plates and transferred into screw-capped tubes containing 2 mL of sterile PBS buffer pH 7.2. The suspensions were washed twice and centrifuged for 3 min at 7000 × g, reconstituted in 1 mL PBS and placed into a water bath at
41
100°C for 15 min. The culture result was negative confirming that Legionella were inactivated by heat treatment. Legionella.
The protein concentration of the suspensions was measured by Infrared spectrometry (Direct
Detect®, Merck Millipore, Danvers, USA) as outlined previously. All antigenic suspensions were adjusted to the concentration of 10 µg/100 µL. Aliquots of 100 µL of the suspensions were stored at -80°C. Microtiter plates were coated with 10-fold dilutions of rabbit antibodies raised against serogroup 1 and serogroup 2 of L. longbeachae (1:1000 dilution to further 10- fold dilutions. The wells of microtiter plates were then coated with both serogroup 1 and 2 antigenic suspension in 6 M guanidine HCl as described. Light absorbance of the samples was measured on a Fluostar reader to evaluate the extent of shared antigenic determinants due to possible protein sequence homology between these organisms and L. longbeachae sergroup1 and serogroup 2.
3.2.4. Coupling antibody with magnetic beads
M-280 Tosylactivated Dynabeads (Invitrogen, Norway) were used to couple the high titre polyclonal antibody harvested from rabbit A. Coupling and washing buffers were prepared as described in 2.2.2.2, and the antibody was coupled to the beads with the final concentration of
100 µg antibody / 5 mg beads. The washed and unbound antibodies were collected separately and quantified with Direct Detect® Infrared spectrometer to evaluate the efficiency of the coupling.
To assess the capturing of Legionella cells by the antibody-coupled beads, a bacterial suspension of 106 CFU/mL of L. longbeachae serogroup 1 using PBS buffer was prepared.
Four dilutions of L. longbeachae serogroup 1 were prepared and 50 µL of the coupled beads were added to 1 mL of these bacterial suspensions in 2 mL microtubes, followed by incubation
42
at room temperature for 30 min. During the incubation period, the microtubes were inverted several times to ensure the antibody-coupled beads were well mixed in the bacterial suspension.
After the incubation period, the tube was placed on the magnet to allow the magnetic bead- bound bacterial cells within the sample to stick to the side of the tube wall. Any unbound protein in the sample was washed off by BPS buffer, and the remaining Dynabead-bound bacteria were washed once more with PBS buffer. To evaluate the efficiency of IMS for DNA extraction, washed beads (200 µL) were used as samples for DNA extraction using GenElute bacterial genomic DNA kit protocol (Sigma) and L. longbeachae specific qPCR was conducted.
3.3. Results
3.3.1. Antigen preparation
The identity of 3-day-old colonies grown on BCYE agar was confirmed as L. longbeachae by
MALDI-ToF mass spectrometry. After heat killing the bacteria, there was no growth on sheep blood agar or BCYE agar confirming lack of viability of the bacterial suspension to be used for inoculation the rabbits.
3.3.2. ELISA screening of the serum
The results of the ELISA test on the collected sera indicated that all four rabbits produced antibodies against L. longbeachae antigens and had a considerable quantity of antibody in their serum. As shown in Figure 3.1, the serum collected from rabbit A immunised with L. longbeachae serogroup 1 antigens, the antibody titre was the highest. Serum from rabbit A was selected for purification of IgG fraction.
43
Figure 3.1: Antibody titre in four rabbits based on light absorbance at 450nm.
These graphs from ELISA assay reads illustrate that all four rabbits (A to D) were responsive and produced antibodies against L. longbeachae, but at different concentrations. Antiserum from rabbit A had the highest antibody titre against both serogroup 1 (sg1) and serogroup 2 (sg2) of L. longbeachae. OD450 numbers are the average of triple measurements of light absorbance.
3.3.3. Purification of rabbit polyclonal antibody
Light absorbance (280nm) of fractions of rabbit A serum eluted from the ion-exchange chromatography column, showed that fractions 7 to 16 had the highest level of protein fractions. These fractions were pooled for downstream procedures. Infra-Red based spectrometry showed that the mean concentration of these pooled fractions was 1.4 mg/mL
(SD=0.12).
44
3.3.4. Screening of the purified protein fractions
ELISA assay of the purified fractions showed that the antibody raised in rabbit A reacted with anti-rabbit IgG HRP antibody. The same protein fraction containing immunoglobulins was responsive to antigens of both L. longbeachae serogroup 1 and serogroup 2. The level of this response was higher against the L. longbeachae serogroup 1 antigens as shown in Figure 3.2.
Figure 3.2: Average of light absorbance (OD450) of 1:10 dilution of fractions of crude protein following chromatography.
The reaction of total Ig with anti-rabbit IgG HRP antibody was as shown with the blue line. The reaction of protein fractions against the antigens of L. longbeachae sg1 and sg2 as measured by ELISA. (Orange and grey lines). The OD450 measurements were repeated three times and the average was plotted.
The ELISA assay demonstrated that the purified protein fractions contained antibodies reactive to the anti-rabbit antibody. This responsiveness was evaluated by measuring the reactivity of goat anti-rabbit HRP-conjugated IgG against four dilutions of the chromatographic fractions
(Figure 3.3).
45
Figure 3.3: Total protein fractions reactivity to anti-rabbit IgG HRP antibody.
Protein fractions at dilutions of 1:10, 1:100, 1:1000 and 1:10000 were added to the plates coated with HRP-conjugated goat anti-rabbit IgG antibody in an ELISA microplate. Optical density was measured at 450 nm wavelength.
Evaluation of reactivity of the purified protein fractions with varying concentrations against L. longbeachae serogroup 1 antigens showed that reactivity reached the plateau in fraction 16 and shows a declining trend afterwards (Figure 3.4).
Figure 3.4: Specific reactivity of varying concentrations of Ig fractions with L. longbeachae sg 1 bacterial suspension.
Purified fractions with varying dilutions of 1:10, 1:100, 1:1000 and 1:10000 were added to wells in a microplate coated with the bacterial suspension.
46
The same evaluation of reactivity against L. longbeachae serogroup 2 antigens showed the same pattern of reactivity of the rabbit protein fractions (Figure 3.5).
Figure 3.5: Specific reactivity of varying concentrations of protein fractions with L. longbeachae sg 2 bacterial suspension.
Purified fractions with varying dilutions of 1:10, 1:100, 1:1000 and 1:10000 were added to wells in a microplate coated with the bacterial suspension.
3.3.5. Cross-reactivity testing
The results of the cross-reactivity test showed no cross-reaction of the anti-L. longbeachae antibody with non-longbeachae species of Legionella and A. fumigatus and P. aeruginosa, but high specificity for L. longbeachae serogroup 1. There was a partial cross-reactivity with L. longbeachae serogroup 2. Figures 3.6 and 3.7 show the cross-reactivity scores based on the light absorbance at 450nm. Varying concentrations of antibody were tested from 1.4 µg/mL
(1400 ng/mL) to 0.14 pg/mL. As seen in Figure 3.7, the results showed that there were interactions between anti-L. longbeachae antibodies produced in rabbits with L. longbeachae serogroup 1 organisms. There was no interaction with other species of Legionella and also non-
Legionella microbial suspensions. Reactivity to L. longbeachae serogroup 1 was observed in concentrations up to 1.4 ng/mL.
47
Figure 3.6: Cross-reactivity test between immobilised L. longbeachae serogroup 1 suspension and with several other organisms using the rabbit anti-L. longbeachae polyclonal antibody.
Scores closer to zero represent more reaction between antigen and antibody and scores close to 1 and above show no interaction. Protein fractions used for this test were from 1400 ng/mL to 0.14 pg/mL concentrations respectively.
Similar results were obtained from ELISA assay for cross-reactions for L. longbeachae serogroup 2 as shown in Figure 3.7.
Figure 3.7: Cross-reactivity test between immobilised L. longbeachae sg 2 with several other organisms using the rabbit anti-L. longbeachae polyclonal antibody.
Protein fractions used for this test were from 1400 ng/mL to 0.14 pg/mL concentrations respectively. Scores closer to zero represent more reaction between antigen and antibody and scores close to 1 and above show no interaction.
48
3.3.6. Coupling antibody with magnetic beads
After the coupling steps were completed, the mean concentration of unbound antibodies in the discarded supernatant was 0.46 µg/mL implying that 66% of the anti- L. longbeachae antibodies were coupled to the beads. The culture result of a suspension of L. longbeachae with the antibody coupled beads followed by colony count showed that the beads were able to capture approximately 65% of the bacteria from the suspension. The identity of the captured colonies was confirmed as L. longbeachae using MALDI-ToF mass spectrometry. Quantitative
PCR of DNA samples from the bacteria captured by the IMS method (Figure 3.8) showed the bacteria were captured by antibody-Dynabead complexes.
Figure 3.8: Amplification plot for DNA extracted from L. longbeachae sg1 cells captured by IMS method.
The fluorescent signal (∆Rn) is plotted versus the cycle numbers. The three purple curves (A-C) show fluorescence signal from DNA extracted from bacterial suspensions. Curve A with Ct value of 20 is related to bacterial suspension with 106 CFU/mL, curve B shows Ct = 25 for 104 CFU/mL, and curve C shows Ct = 30 for 102 CFU/mL respectively.
49
3.4. Discussion
The aim of the work described in this chapter was to develop a polyclonal antibody against L. longbeachae bacterial antigens. This antibody was used for developing an immunomagnetic technique with the potential to capture bacteria from both clinical and environmental samples that contain multiple microbial populations.
The experiments were set up according to the polyclonal antibody production method used at the Endocrinology and Steroid laboratories, Canterbury Health Laboratories. In order to stimulate antibody production in rabbits, Legionella antigens were prepared for inoculation into the animals. The bacteria were heat killed at 60°C rather than boiling in 100°C or using formalin, to maintain the antigenic integrity of the organism and retain superficial antigenic epitopes [133].
Direct Detect® Spectrometer was used to measure the concentration of the raised antibody.
This method had several advantages over the traditional colorimetric assays such as the
Bradford assay. It is very simple and can be performed in only two minutes using only 2 μL of the sample per analysis. The Infrared spectrometry method has an improved accuracy in comparison with the conventional protein quantitation assays and avoids the problem of interfering detergents [134].
The results of cross-reactivity test were important to show that the antibody reacted with L. longbeachae successfully and there was no cross-reactivity with L. pneumophila and L. micdadei. Lack of cross-reactivity is important for clinical application such as isolation of L. longbeachae from respiratory specimens. In particular, there was no cross-reactivity with L. pneumophila and L. micdadei which are the most commonly identified non-L. longbeachae species in the clinical samples [5]. However, cross reactivity between Legionella species would
50
not have a major impact clinically, as possible cross-reacting species can be identified by
MALDI-ToF mass spectrometry or qPCR in the diagnostic laboratory.
Cross-reactivity with any of a large number of other environmental Legionella species would limit the usefulness of the antibody in isolating L. longbeachae from the environmental samples. Although Legionella are adapted to persist in varying conditions in the environment, their isolation from environmental samples is challenging [10]. Previous studies have shown the presence of L. pneumophila and other species of Legionella are mainly found in soil containing water and waste materials. For example, isolation of several Legionella species including L. pneumophila, L. feeleii, L. dumoffii, L. longbeachae, and L. jamestownensis from soil contaminated with industrial waste in Japan has been reported [135]. In addition, L. pneumophila, L. longbeachae, L. bozemanii, L. dumoffii, L. gormanii, L. jordanis, L. micdadei and L. anisa were isolated from potting soil in Greece [44]. One limitation of our cross- reactivity experiment is that there may be unknown environmental species of Legionella genus that are also difficult to culture. However, cross-reacting species that might be immunocaptured and further grown on Legionella microbiological media can be identified by the qPCR method and MALDI-ToF mass spectrometry.
There was no cross-reaction with P. aeruginosa and A. fumigatus. These organisms were chosen because they are respiratory pathogens and found in the environment as well. For example, patients with respiratory diseases such as with cystic fibrosis (CF) and chronic obstructive pulmonary disease (COPD) are at higher risk of getting infected with these microorganisms [136-139].
Pathogenic microorganisms such as P. aeruginosa and A. fumigatus are also environmental organisms that are commonly found in soil, water bodies and potting mix products [140]. These
51
organisms were therefore considered to be useful indicator organisms of non-Legionella organisms that may be present when testing environmental samples.
Cross-reactions of antibodies against Legionella species with other organisms have been reported suggesting that these organisms share similar antigens. For example, cross-reactions between L. pneumophila and Burkholderia pseudomallei, Coxiella burnetii and L. micdadei has been shown in previous studies [141-143]. It was not feasible to test a wide range of environmental organisms that could potentially cross-react with the antibody. Confirmatory techniques such as MALD-TOF mass spectrometry may be used to identify organisms isolated from the environmental samples but the accuracy of these methods will depend on whether the spectra are entered into the library. Other testing such as sequencing may be needed to establish the identity of organisms identified if there is not a good match in the database.
Combining the application of magnetic beads with the specific antibodies can help the isolation of these microorganisms from samples with complex microbial communities. Magnetic beads with 2.8 µm diameter covered with toluenesulfonyl (tosyl) groups are recommended by the manufacturer when the antibody targets bacterial or fragile eukaryotic cells. These beads are the smallest available beads and have been shown to capture small cells such as bacteria [144].
These magnetic beads have hydrophobic residues that are able to covalently bind to the primary amino or sulfhydryl groups. The tosyl group is a highly reactive group due to the stability of the resonance structure. It has a distributed negative charge rather than a localized charge. The tosyl group serves as a protecting group which is readily cleavable and thus acts as a leaving group. There is a slightly positively charged carbon atom which reacts with available negative charged groups on a target molecule. The negative group on an antibody will attack the positive part of the tosyl group and the antibody will replace the tosyl group becoming covalently bound
(K. Pedersen, Life Sciences Solutions, Norway. Personal Communication, 31/1/2017).
52
The IMS method was successfully applied to both culture and PCR of L. longbeachae due to the specific binding of the polyclonal antibody attached to the magnetic beads to the intact L. longbeachae cells. The capture was linear over a range of 102 to 106 CFU/mL. To our knowledge, this is the first study to use IMS method specifically for L. longbeachae. The current literature available is mainly about L. pneumophila. For example, an IMS/culture has been evaluated using a monoclonal antibody against the lipopolysaccharides (LPS). This study has used the identical magnetic bead as our work. The culture limit was found to be 103
CFU/mL [145].
Antibody-coupled magnetic beads have several important advantages for studies that may be conducted over long time periods. They do not need to be prepared fresh for each experiment.
Coupling of the antibody to Dynabeads can be performed in high volume and the coupled beads can be stored as aliquots in the refrigerator. This will help save time for carrying out IMS experiments during the study. The shelf life can vary from 24-36 months depending on the antibody coated on the magnetic Dynabeads. Storage of coupled Dynabeads is straight forward as they just required to be covered with buffer and refrigerated at 2-8oC in the upright position.
Freezing of coupled-beads is not generally recommended by the manufacturers. Freeze- thawing can cause damage to the surface of the beads. This is especially important for the
Dynabeads coated with antibodies on the surface as freezing of the beads can influence the functionality of the antibodies coated on the bead surface. If they have been frozen, it is important to re-suspend the beads completely and wash them properly before use. If freezing is required, the beads need to be frozen in the buffer they are supplied in by the manufacturer.
Repeated freezing and thawing should be avoided as ice crystals formed can damage the surface of the Dynabeads (ThermoFisher troubleshooting support centre webpage).
53
In conclusion, the polyclonal anti-L. longbeachae antibody produced in rabbits was very specific for L. longbeachae sg1 and could also bind to serogroup 2 with a lower affinity. There was no cross-reactivity between this antibody and the other Legionella species and also two non- Legionella microorganisms available at the time of conducting the experiments. This antibody was successfully coupled with Dynabeads and captured L. longbeachae bacteria for culture and DNA extraction. As the antibody produced for this study was successfully tested by an ELISA assay to detect L. longbeachae serogroup 1, development of a specific immunoassay for detection of L. longbeachae in different types of clinical and environmental samples can be considered for future studies.
54
CHAPTER 4
Enhancement of Culture and qPCR for Respiratory Samples Using Immunomagnetic Separation
4. Enhancement of culture and qPCR for respiratory samples using immunomagnetic separation 4.1. Introduction
A range of methods has been developed to isolate and detect Legionella in a variety of clinical and environmental sample types as described in Chapter 1. Detecting, isolating and identifying these bacteria is clinically important for patient treatment because it enables the optimal antibiotic regimen for treatment to be determined. The identification of Legionella species from clinical samples has implications for public health because it allows a timely and appropriate epidemiological investigation of potential LD outbreaks.
The isolation of Legionella species is a laborious task regardless of the availability of current diagnostic techniques such as bacterial culture and use of selective media [71]. The low number of bacterial cells in the sample and overgrowth of co-existing non-Legionella and fungi requires experienced laboratory personnel to spend a considerable amount of time examining the cultures. Samples often contain mixed organisms from the lung microbiome and other microorganisms such as Burkholderia spp., Proteus spp., and fungi that can grow on the
Legionella agar and overgrow the Legionella colonies causing false negative results [146; 147].
Acid pre-treatment of samples was developed to reduce the number of non-Legionella organisms and improve the isolation of Legionella. This decontamination method is imperfect and may inhibit the growth and recovery of Legionella, especially in specimens with a low bacterial load leading to a false negative result [65]. Acid pre-treatment has been widely used to reduce contamination because of the relative resistance of Legionella species to a low pH, but this technique may reduce the recovery of these organisms as well. For example, the acid pre-treatment of water samples spiked with L. pneumophila reduced the recovery rate by 50% on both BCYE non-selective and MWY selective media [148].
56
It has also been reported that different strains of Legionella have varying degrees of sensitivity to acid pre-treatment. For example, some strains of L. pneumophila isolated from environmental water samples were more resistant to HCl:KCl buffer pH 2.2 treatment in comparison to the strains stored in the laboratory [149].
The current protocol for the isolation of Legionella species from respiratory samples at
Canterbury Health Laboratories (CHL) is based on the Clinical & Laboratory Standards
Institute (CLSI) recommendations that provides standards and guidelines for medical laboratories (T. Anderson, Canterbury Health Laboratories, Personal communication,
5/2/2018). This protocol uses Dithiothreitol (Sputolysin ® , Calbiochem, USA) to homogenise the mucoid specimens and then acid pre-treatment to inhibit commensal flora before culture on
BCYE and MWY medium (R. Podmore, Microbiology Department, CHL laboratories,
Personal communication, 8/2/ 2018). Sputolysin can free bacteria from the surrounding proteinaceous matrix of the specimens such as sputa but also has a negative effect on Legionella viability and therefore the recovery of Legionella spp. A study conducted at CHL showed that prolonged exposure to sputolysin decreases Legionella growth with a 50% (CFU/mL) reduction in viability over 24 hours (Unpublished data). This effect can lead to a false negative culture result and thus should be considered as a potential growth inhibitor. Nonetheless, sputolysin is an essential step for sample homogenisation for culture and the IMS procedure
CHL staff have also standardised the acid pre-treatment procedure as this has been found to be time critical. Overexposure to acid decreases organism viability and therefore reduces the
Legionella recovery rate (T. Anderson, Microbiology Department, CHL laboratories, Personal communication, 21/5/2019). An exposure time of three minutes is currently used.
The introduction of real-time PCR technology as a fast and accurate tool of molecular diagnostics has changed the gold standard of Legionella diagnosis from culture to real-time
PCR [150]. This method has been dramatically improved by selecting different target genes
57
such as Legionella macrophage infectivity potentiator (mip), 5S ribosomal DNA and 16S ribosomal DNA and the 23S-5S spacer regions [150-153] for the assay. Behets and colleagues developed TaqMan™ qPCR for the detection of L. pneumophila and reported it as a reliable, sensitive and specific detection method [154].
In a large study on 1843 respiratory samples using real-time PCR with a L. longbeachae- specific hybridisation probe, Murdoch and colleagues demonstrated that all 2.1% (39/1843) culture positive samples were detected and an additional 2.4% (45/1843) samples with negative culture results were positive [5]. This doubled the number of cases in which Legionella species were implicated and suggested that the main reason for the discrepancy between culture and
PCR results was the detection of DNA from non-viable organisms that may have been killed or inactivated by prior antibiotic therapy or sample pre-treatment such as acid pre-treatment.
Real-time PCR is not a perfect test as the presence of inhibitors in biological and environmental samples can hamper DNA extraction or amplification. This can also give rise to false negative results. Inhibitors can also shift the Ct values and cause false negative results. For example, the haemoglobin from blood in respiratory specimens such as sputum or bronchoalveolar lavage
(BAL) can inhibit Taq polymerase activity [155]. Therefore, the addition of internal controls to the samples is needed to identify inhibition of the reaction.
The successful application of the IMS method for microbial identification in different types of samples such as foods [156], clinical and animal specimens [157; 158] and environmental samples [159] illustrates that application IMS to clinical samples has potential to reduce the contamination load while not influencing the viability of Legionella cells [145]. IMS could capture organisms and concentrate the Legionella load prior to both culture and DNA extraction for PCR testing. The IMS procedure as a pre-treatment step for DNA extraction from complex sample matrix such as sputum may also remove potential PCR inhibitors such as high
58
concentrations of DNA (human and other microorganisms) and further improve the detection rate of L. longbeachae. This chapter compares the recovery rate of L. longbeachae by culture and PCR in respiratory samples with and without IMS enhanced recovery. The aim was to improve the efficiency of diagnosis of L. longbeachae in clinical settings for both culture and qPCR.
4.2. Materials and Methods
4.2.1. Respiratory samples
Residual portions of 62 respiratory specimens that were processed at CHL and Legionella positive by ITS qPCR were available. These stored specimens were not acid treated prior the storage. The ITS qPCR is genus specific. These samples had been collected throughout 2017 and stored as 2 mL aliquots at -80oC freezer by CHL. No cryo-preservative was added to the samples prior to storage.
4.2.2. Selective media and supplements
GVPC solution was used to decontaminate the specimens. This solution was prepared as described in Section 2.2.2.2. Modified Wadowsky and Yee (MWY) agar was used to culture
Legionella in this work.
59
4.2.3. Acidic buffer
HCl:KCl acidic buffer pH 2.2 was prepared as described in Section 2.2.2.1.
4.2.4. Acid pre-treatment of the respiratory specimens
The stored sputa were thawed at room temperature after removal from the -80oC freezer. Equal volumes of HCl:KCl acidic buffer and specimen were mixed in 2 mL microtubes. The mixture was left at room temperature for 4 min. The sample was streak plated (100 µL) onto MWY agar (Fort Richard Laboratories, Auckland, New Zealand). The plates were incubated at 36oC for up to six days and examined visually and by stereomicroscopy daily from day three for
Legionella-like colonies. The identity of presumptive Legionella colonies was confirmed by
MALDI-ToF mass spectrometry (Bruker Daltonik GmbH, Bremen, Germany).
4.2.5. IMS-culture for the isolation of L. longbeachae from respiratory samples
To reduce the contamination rate of respiratory samples, an equal volume of GVPC supplement dissolved in sterile ultrapure water was added to the samples and incubated at room temperature for 4 min.
Following this, 50 µL of Dynabeads™ coated with anti- L. longbeachae antibody (prepared earlier as described in Chapter 3) was added to a 2 mL microcentrifuge tube. One mL of each
GVPC treated sample was added to a 2 mL microcentrifuge tube. The sample was incubated at room temperature with gentle agitation for 2 h. The microcentrifuge tube was placed on a magnetic stand for 2 min to allow the magnetic beads to be pulled from the solution. The bead- bacteria complexes were washed twice with 1 mL of sterile ultrapure water (pH 7.0). The obtained pellet as the final sample was resuspended in 1 mL GVPC suspension and 100 µL was inoculated onto a selective MWY plate. To make sure all the bacteria had been captured,
60
100 µL of supernatants was inoculated onto a MWY plates for culture as well. All the MWY plates were incubated at 36oC in an aerobic atmosphere for 3-7 days. Legionella-like colonies were examined under a stereomicroscope and MALDI-ToF mass spectrometry was applied to confirm the identity of these colonies as Legionella.
4.2.6. DNA extraction kit
The GenElute bacterial genomic DNA Kit (Sigma-Aldrich, St. Louis, MO, USA) was used to extract DNA from the respiratory samples.
4.2.7. Primers and probes for ssrA qPCR
To perform qPCR on the samples, the established screening test for clinical specimens at the
CHL laboratory was used, which uses Legionella genus specific primers (ssrA) described in
Section 2.2.3.3. All primers and probes were obtained from Integrated DNA Technologies Inc.
(Coralville, IA, USA), except the minor groove binding (MGB) PanLeg-P1 probe, which was custom made by Applied Biosystems. This probe was labelled at its 5′-end with FAM and at its 3′-end the MGB-NFQ quencher.
61
4.2.8. PCR mixture and thermal cycling programme
The PCR reaction mix was prepared as outlined in section 2.2.4.3. PCR reaction mix (15 µL) and DNA templates (10µL) were loaded into the LightCycler® 480 multiwell Plates 96
(Roche Diagnostics GmbH, Mannheim, Germany) and the plates were sealed. All qPCR assays were carried out using the LightCycler 480 real-time platform (Roche Diagnostics,
New Zealand) using the conditions summarised in Table 4.1.
Table 4.1: PCR cycling conditions
Step Temperature (oC) Time (HH:MM:SS) Cycle Number Hold 50oC 00:02:00 1 Denaturation 95oC 00: 05:00 1 Amplification 95oC 00:00:15 50 60oC * 00:01:00 * Acquisition of fluorescence
4.2.9. Generation of a standard curve for ssrA PCR using the synthetic sputum
To generate a standard curve for qPCR, synthetic sputum with a similar content as human sputum was prepared and applied as described in Section 2.2.3. The prepared synthetic sputum was spiked with L. longbeachae sg1 ATCC33462 strain from a 3-day-old culture on BYE agar.
10-fold dilutions from 103 to 106 CFU/mL were made by using sterile saline. One mL of each dilution was added to 9 mL of synthetic sputum. DNA was extracted using the same procedure as described in 4.2.6. Ten microlitres of each DNA sample was added to the master mix and ssrA PCR was performed. A standard curve was generated by plotting the Ct values and known
Legionella concentrations as described above.
62
4.2.10. DNA extraction and qPCR for Dynabead-treated samples
To extract DNA, 200 µL of the processed sample was used. DNA was extracted from the samples using GenElute Bacterial Genomic DNA Kit (Sigma-Aldrich) according to the manufacturer’s instructions. However, DNA templates used for the initial qPCR were extracted using the NucliSens® easyMag automated platform at the CHL laboratory. The initial ITS real- time PCR program was specific for Legionella genus that was used in Chapter 3. Real-time
PCR was conducted as described in Section 4.2.1.5. Inhibited samples were diluted 1:10 and qPCR was repeated on the diluted samples.
4.2.11. Stratification of Ct value of respiratory samples by pneumonia severity scores
To examine the association between the Ct values of quantitative PCR and disease severity in patients, anonymised CURB-65 scores of patients whom the specimens belonged to were used.
The CURB-65 score is a validated severity scoring system that predicts mortality and is derived from patient information collected at admission to hospital [160]. These criteria are listed in
Table 4.2. One point is given for each characteristic present and the numbers added together.
A CURB-65 score of 0 is mild and 5 the most severe category with highest mortality. Professor
Stephen Chambers retrieved the clinical data from the clinical records and provided an anonymised CURB-65 score that was then matched with the Ct values.
63
Table 4.2: Clinical criteria for severity stratification for CAP
CURB-65 criteria* for CAP cases Confusion Urea concentration > 7 mmol/L Respiratory rate ≥ 30 Systolic blood pressure (BP) < 90 mmHg or diastolic BP ≤ 60 mmHg Age ≥ 65 * For each of these criteria applied to the patients, one point is considered
4.2.12. Statistical analysis
Statistical analyses were carried out using R 3.5.0 (R Core Team, 2018). To assess whether there was a significant difference between the isolation rates after applying the acid pre- treatment and IMS methods, paired proportions were generated using the McNemar's Chi- squared test with continuity correction and ANOVA Tukey Post-Hoc test. The resulted data were plotted by R programme as a generalised additive regression model to illustrate the relationship of the culture results and Ct values of qPCR. To examine the relationship between the Ct values and CURB-65 scores of patients, linear regression was used. The graphs were generated using R. P < 0.05 was considered as an indicator of the significant difference.
4.3. Results
4.3.1. Bacterial culture of L. longbeachae after acid pre-treatment
Of the 62 stored samples available from CHL 53 were positive for L. longbeachae, eight for L. pneumophila and one L. micdadei according to the recorded results of culture and microscopy, qPCR and MALDI-ToF tests performed by CHL microbiology laboratory (R. Podmore, CHL laboratories, Personal communication). Of the 53 samples containing L. longbeachae DNA, 15
(30%) culture positive for L. longbeachae after acid pre-treatment prior to cryopreservation at
-80oC stored in 2 mL Cryovial tubes (Greiner Bio-One GmbH - Frickenhausen, Germany) at
64
CHL laboratory. Following storage, the remainder of these specimens were acid treated in our laboratory and 10 (18.9%) samples were culture positive (Appendix 1). The proportion of samples that were culture positive after acid pre-treatment (18.9%) was lower after storage in comparison with the fresh samples (30%), but not significantly different when examined by
McNemar’s Chi-squared test (P = 0.096). Cross-tabulation (Table 4.3) suggests that most samples that were culture positive after storage, had also been culture positive initially before freezing the samples (7/10). However, less than half of the samples that had been culture positive before storage were still positive after storage (7/17).
Table 4.3: Cross-tabulation of culture results of acid treated samples before and after storage
Second acid treatment Initial acid treatment Culture negative Culture positive Total
Culture negative 33 3 36 Culture positive 10 7 17 Total 43 10 53
4.3.2. IMS-Culture for isolation of L. longbeachae in respiratory samples
Results of culture on selective MWY agar showed that from 53 samples originally qPCR positive for L. longbeachae DNA prior to cryopreservation (CHL), 26 samples (49%) were culture positive after cryopreservation following the concentration of specimens with the IMS treatment (Figure 4.1). The proportion that were culture positive using IMS (49%), was significantly higher compared with the acid-treated stored samples (10/53, 18.9%; P < 0.001).
Figure 4.1 shows the recovery rate of the culture on the acid-treated fresh, acid treated stored and IMS-treated stored specimens. Culture following IMS separation of the 9 CHL culture positive non-L. longbeachae samples were all negative.
65
Figure 4.1: Comparison of the recovery rate of L. longbeachae from respiratory specimens.
There was no significant difference between the recovery rate from fresh (purple) and cryopreserved (green) samples treated with acidic buffer (n = 53, P = 0.15). There was a significant increase in recovery rate using IMS (orange) than both initial culture (P < 0.05) and the stored samples (P < 0.01).
No formal comparison was made of the colony counts on plates after preparation of the stored samples by direct plating, acid pre-treatment and IMS/GVPC methods but a representative example of plates is shown in Figure 4.2. There were many colonies of contaminating respiratory flora using the direct plating method. Acid pre-treatment and IMS eliminated this contamination. In this example the number of colonies seen on the plate prepared with acid pre-treatment is lower than with and IMS which is consistent with the qPCR before cryopreservation (CHL) results. Culture results of all supernatants from the IMS-washed samples were negative for L. longbeacahe.
66
A B C
Figure 4.2: A representative culture recovery comparison of a direct plated (A), acid washed (B) and IMS (C) treated sample.
This figure shows the culture result of L. longbeachae sg1 from a given specimen using direct plating, acid pre-treatment and IMS/GVPC treatment from left to right respectively on selective MWY agar. Direct plating (A) resulted in the growth of contaminating microorganisms present in the sample and failure of the growth of Legionella. Acid pre-treatment (B) reduced the contamination on culture, however it reduced the recovery level of L. longbeachae. Combined IMS/GVPC method followed by culture resulted in an increased recovery level of L. longbeachae and no contamination of the culture.
4.3.3. IMS-qPCR of respiratory samples from community acquired pneumonia patients
The standard curve for the ssrA PCR (after cryopreservation) results for the Legionella genus in synthetic sputum was generated as shown below (Figure 4.3). The PCR reaction for four
IMS-treated samples was inhibited initially (data available in appendix 1). However, re-assay of a 1:10 dilution of these samples, showed that they were not detected on repeat qPCR testing.
67
Figure 4.3: Standard curve for the spiked artificial sputum qPCR.
This curve demonstrated the relationship between the Ct values and log of varying DNA concentration of L. longbeachae. Synthetic sputum was spiked with 10-fold dilutions of L. longbeachae sg1 from 103 to 106 CFU/mL (R² = 0.9977).
To illustrate the relationship of the frequency of culture-positive samples for L. longbeachae and IMS-qPCR, the results were plotted against Ct values from the qPCR before cryopreservation (CHL) and ssrA qPCR after cryopreservation (UOC) assays. As shown in
Figure 4.4, after the initial acid pre-treatment of the fresh respiratory samples (CHL), positive culture samples were clustered below the Ct value of 36, while the culture negative samples had Ct values up to 42 (panel A). After using IMS method on cryopreserved specimens, the range of culture positive samples shifted with a higher number found with Ct values greater than 34 using the initial PCR conducted by CHL laboratory (panel B). One sample that was
PCR positive with a high Ct value of 42 before storage, was PCR and culture negative using
IMS method after cryopreservation. IMS-treated cryopreserved samples were tested using the qPCR targeting the ssrA gene. The results showed a similar distribution with culture positive samples with higher Ct values (panel C).
68
Figure 4.4: Distribution of Ct values of qPCR positive samples for L. longbeachae.
Panel A shows the CHL laboratories results of original qPCR (ITS) and culture after acid pre-treatment of the respiratory specimens. Panel B shows the association between the ITS qPCR Ct values with the culture results of the IMS processed stored specimens. Panel C shows results of IMS-culture and ssrA qPCR (after cryopreservation) for Legionella spp. of the same samples after storage for 12 months following IMS method prior to processing.
69
The culture results following the three treatments and their related Ct values are shown in Table
4.4. The mean Ct values were significantly different between culture positive and negative samples for the fresh and stored respiratory specimens treated with acid pre-treatment. The difference was not significant for IMS-treated samples. The number of culture positive samples for the initial acid pre-treatment, acid treatment of the stored samples, and IMS-treated samples were 17, 10 and 27 respectively.
Table 4.4: Comparison of stratified samples based on mean Ct values of different treatments
Treatment Mean Ct value Number of Difference Confidence P-value samples of Ct interval Culture Culture Culture Culture negative positive negative positive AW1 35.52 30.36 36 17 5.15 [2.64 - 7.67] < 0.001 AW2 34.78 29.92 43 10 4.86 [0.85 - 8.88] 0.022 IMS 34.72 32.98 26 27 1.74 [-0.73 - 4.21] 0.164 AW1: acid pre-treatment of fresh respiratory samples AW2: acid pre-treatment of stored respiratory samples IMS: Immunomagnetic separation for stored respiratory samples
The relationship between culture positivity after acid pre-treatment of the fresh and stored specimens and IMS pre-treated stored specimens is shown as a simple logistic regression model in Figure 4.5 based on 95% confidence interval. This graph demonstrates the decline of culture positivity for the stored respiratory specimens. IMS-treated samples are shown to have greater possibility of culture positivity at higher Ct values containing a lower DNA load of L. longbeachae.
70
Figure 4.5: relationship between Ct value of initial qPCR and culture positivity.
This is a generalised additive regression model generated by R programme in which Ct value of the original qPCR (ITS) applied to fresh respiratory samples for three different treatments was fitted as a continuous linear predictor of culture positivity. The red line shows the mean predicted culture positivity based on Ct values of DNA extracted from fresh respiratory samples treated with acid pre- treatment. The green line, represents this for the identical respiratory samples frozen and stored prior the acid pre-treatment. The blue line shows the culture positivity for the specimens treated with IMS method prior culture. The coloured areas show 95% confidence intervals. (AW1: acid wash of the fresh respiratory specimens; AW2: acid pre-treatment of the stored respiratory specimens; IMS: immunomagnetic separation using the stored respiratory specimens). Figure produced in R studio® version 1.2.1335.
To explore further relationship between Ct values and culture results, the proportion of culture positive fresh and stored specimens after acid pre-treatment was plotted using grouped of Ct values (CHL) as shown in Figure 4.6.
71
Figure 4.6: Comparison of Ct values of qPCR and culture results of acid treated samples before and after storage.
This plot shows the proportion of culture positive samples pre-treated by acid pre-treatment that are grouped by four Ct value categories of 22-30, 30 - < 35, 35 - < 38 and above 38. Fresh samples are shown in red and stored samples are shown in blue. This classification was carried out based on the median Ct values and interquartile ranges obtained from the ITS qPCR conducted by CHL laboratory. The dots represent the mean Ct values of tested respiratory specimens. Figure produced in R studio® version 1.2.1335.
The same comparison was conducted for culture results of the stored respiratory specimens pre-treated with acid pre-treatment or IMS that is shown as Figure 4.7. For acid treated stored samples, mean Ct value of culture positive samples (29.92) was significantly lower than the mean Ct value of culture negative samples (34.78) with P < 0.05.
72
Figure 4.7: Comparison of mean Ct values and culture results for the stored specimens treated with acid wash and IMS method.
Median Ct values and interquartile range of the ITS region targeted qPCR (CHL) was stratified by culture results for the acid treated stored samples in red (AW2), and IMS-treated stored samples are shown in blue. Red lines represent negative cultures and blue lines represent positive cultures. The dots represent the mean values. Figure produced in R studio® version 1.2.1335.
Comparison of Ct values of two PCR runs before and after cryopreservation of the respiratory specimens by R programme showed a relative similarity of the results with Pearson correlation
(r) = 0.83, (95%CI = 071 to 0.90).
4.3.4. Assessment of disease severity in patients corresponding to the positive PCR for L. longbeachae and qPCR results
CURB-65 scores of 44 patients with culture positive samples for L. longbeachae was available.
The distribution of CURB-65 scores for the patients from whom respiratory samples obtained were 0 (9/44, 20%), score 1 (12/44, 27.27%), score 2 (13/44, 29.54%), score 3 (8/44, 18.18%), and score 4 (3/44, 6.81%). None of the recruited patients matched the most severe score of 5.
73
Categorising the Ct values based on CURB-65 criteria showed that there is an apparent correlation between cycle threshold cut-off of qPCR and the severity of pneumonia in patients from whom the respiratory samples were obtained (Figure 4.8). The lower Ct values represent the higher genomic unit (DNA load) of the given sample. An analysis of variance (ANOVA) showed that the median Ct values varied across the groups (P < 0.001). The mean Ct value of samples taken from patients with CURB score of 4 (23.66±4.72) were significantly different
(P < 0.001) from Ct values belonging to the samples related to score 0 (35.72±3.46), score 1
(34.90±4.28) and score 2 (33.58±3.90).
Figure 4.8: Relationship between the Ct values of ITS qPCR and the severity score (CURB65) of the disease.
Ct values were plotted according to their corresponding CURB65 scores. Score 0 indicates the mildest level of pneumonia while a score of 4 indicates the most severe in the cohort of patients that provided the samples for this assessment. Dots represent the mean Ct values and horizontal lines show the medians and interquartile range for Ct value of each group. Ct value belonging to the score 4 samples was significantly different from samples in category 0, 1 and 2 (P < 0.001).
4.4. Discussion
The results of this study demonstrate that the IMS-culture method has potential to improve the sensitivity of L. longbeachae culture in the clinical setting. There was a significant increase in number of positive L. longbeachae culture samples following the IMS procedure (49%), in comparison with the results after the acid pre-treatment from both initial (31%) and second
74
(18%) acid treatments. The most likely reason for this is that the organisms are not exposed to the acid pre-treatment with IMS and the acid pre-treatment is inhibitory to stored organisms.
Applying IMS method can also help the target organism to be removed from the sputum matrix which may contain inhibitory compounds released by contaminating organism, whereas these may remain in the sample after acid pre-treatment and reduce the recovery rate. For example, the inhibitory effect of a strain of Staphylococcus warneri on L. longbeachae ATCC33484 has been demonstrated which was through the release of the bacteriocin peptide compounds [161].
There is no information on the inhibitory effects of common upper respiratory organisms on the growth of Legionella species but this could be investigated.
A further possible explanation for the difference in recovery rates of the acid and IMS-treated samples is that the antibody-bead complex collects the bacterial cells with an intact structure without disrupting the cell wall as a possible damage using acid pre-treatment method.
Legionella species are also known to persist in a viable but non-culturable (VBNC) state that may be detected by PCR [162]. Little is known about the conditions that trigger the change to a culturable form but it is possible the process of IMS may facilitate this. Previous work has shown that the attachment of captured bacteria to the paramagnetic beads has no unfavourable effect on their growth [163] but it is possible that the bacterial load may be underestimated by colony counts. Steric hindrance and spatial limitations during the attachment of the antibodies on the surface of the beads to the surface of bacterial cells, are supposedly random but one bead may get attached to several organisms [164]. This will produce only one colony when cultured on a plate and may lead to an underestimate of the number or viable organisms present in a given sample.
The inhibitory effect of the acid pre-treatment, as it is used to reduce bacterial contamination of sputum samples has not been well studied. The acid pre-treatment method was developed to eliminate the contaminating microorganisms from the samples to help Legionella grow on the
75
microbiological media following the observations that they survived the low pH [109; 165].
Although the HCl:KCl buffer has a pH value of 2.2 in which Legionella can resist for up to 30 min, prolonging this treatment or changing the pH of the sample to lower values can influence the growth of Legionella cells. For example, L. pneumophila Philadelphia 1 strain cells could not survive in pH 1.9 for more than 15 min [165]. Furthermore, a study on the recovery rate of
Legionella with a focus on L. pneumophila serogroups 1-14 and L. micdadei from seeded sterile water samples, revealed that the recovery of Legionella after acid pre-treatment was not satisfactory [166]. Further studies are needed to determine precisely how sensitive L. longbeachae is to acid conditions and the critical timing of acid exposure, but it seems probable that undue exposure will reduce the rate of recovery on culture.
There was an apparent decline in the proportion of culture positive results when the respiratory samples were stored at -80oC, and then were acid treated prior to the culture stage. These results should not be compared directly as they were done in different laboratories, with different preparation of reagents and by different laboratory staff. Nevertheless, it is possible that there was a loss of viability of some samples as no cryopreservative was added to the samples prior to the storage.
The literature has very limited information on the recovery rate of Legionella species following storage at -80oC. Edelstein stated no loss of viability of Legionella after long term freezing of lung and sputum samples at -70oC. However, their main target species was several L. pneumophila strains and they did not provide quantitative data or any information about the other species [167]. Our observations suggest that storage may decrease the recovery rate of L. longbeachae. It is possible that the dual insults of storage and acid pre-treatment both contribute to low recovery rate and further investigation is needed to determine the main reason of this reduction.
76
Loss of viability after freezing has been demonstrated for other bacterial species. Recovery of
P. aeruginosa, from sputum of patients with bronchiectasis after freezing at -20°C for 24 and
48 h resulted in a significant reduction of the bacterial density [168]. In a study on the recovery rate of Haemophilus influenza from the sputum of bronchiectasis patients stored up to 7 days, the optimal recovery of these bacteria happened only from the specimens preserved in glycerol and stored at -70°C [169]. Similar investigations need to be conducted to determine the optimal techniques and storage conditions needed for respiratory specimens in respect to the recovery of L. longbeachae.
Bacterial cells stored by freezing can be damaged by dehydration, formation of ice microcrystals and generation of reactive oxygen species (ROS) while frozen. However, many bacteria have evolved survival mechanisms against freezing conditions. This includes the formation of storage granules such as polyhydroxyl alkanoates [170]. These granules help bacterial cells resist the stress of freezing [171]. Poly-3-hydroxybuyrate (PHB) is a well-studied example of these compounds and it has been demonstrated that Legionella can form PHB as well [172]. This can apply to the stored samples in this study, so that the bacteria were viable during the storage, but needed the optimal condition to grow. Applying IMS treatment to fresh respiratory samples may remove any interfering effect of storage. Further studies need to be carried out with fresh samples to directly compare the rates of recovery of L. longbeachae by culture after acid pre-treatment and the application of IMS.
There was a close correlation between the Ct values of the original ITS qPCR (CHL) and those carried out after storage by ssrA PCR (UOC) (r-value = 0.83, r2 = 0.68; P < 0.01). The minor differences of Ct values of the original ITS qPCR (CHL) and those carried out after storage by ssrA PCR (UOC) may be due to the target difference and the repeated freeze-thawing of respiratory samples which can have an impact on the microbial content of the samples and therefore the DNA extraction outcome (Cuthbertson, 2015). There were also differences in the
77
DNA extraction techniques used. The initial results are from DNA samples extracted from specimens using the automated easyMag system at CHL laboratory, while a manual DNA extraction kit was used in the repeat experiment. It would be preferable to compare the efficiency of combining IMS with DNA extraction using the same DNA extraction and PCR method to remove possible variations in the methods. Despite this the results are remarkably similar and demonstrated that the DNA had not degraded during the storage period.
The results of qPCR and culture, clearly demonstrated that lower Ct values were associated with a higher probability of recovery of L. longbeachae after both acid pre-treatment and IMS.
While with samples with high Ct values, there was a lower probability of recovering an organism after acid pre-treatment. This has important implications for laboratories when isolating the organisms is of value for outbreak investigations, determining virulence factors in individual strains, or monitoring the development of antibiotic resistance [173; 174].
As an exploratory investigation the Ct values and culture results were plotted against the pneumonia severity scores (CURB-65). These demonstrated that the Ct values of ITS qPCR were lower in more severe cases of LD. As the lower Ct value of qPCR results represent higher bacterial load, and the bacterial load has a direct relation with the severity of infections. These findings were consistent with the results of a study conducted by Murdoch and colleagues testing the clinical specimens for Legionella by qPCR [5]. The recovery of organisms from the extremes of disease severity provides a possible tool to investigate the pathogenic factors of organisms in these disease phenotypes, or genetic adaptive changes in the organisms following attack from the host immune system.
In conclusion, the results of our study showed that IMS-Culture resulted in an improvement in the recovery of L. longbeachae bacteria to 49% in comparison with the 31% in fresh acid treated samples and 18% in stored acid treated specimens. It is important that the technique
78
will be tested prospectively in a clinical laboratory, but may provide a reliable technique that could eliminate the use of acid pre-treatment. The IMS technique can have a particular application in research laboratories when culture of L. longbeachae is needed to study on epidemiology or pathogenesis of LD. Our in-house antibody showed a high specificity for L. longbeachae bacteria as the supernatants of the IMS-treated samples were all PCR negative for this organism. A limitation of using this technique in long term is the variation of antibody quality from different batches e.g. from different rabbits Therefore, developing this method using a highly-specific monoclonal antibody has been considered for further work.
The value of the IMS techniques is apparent in L. longbeachae and this technique may be applicable to L. pneumophila. This would require the development of multiple antibodies but there does not seem any reason why the magnetic beads could not be pooled. Both techniques have potential drawbacks as preparing the acid buffer from hydrochloric acid can be potentially hazardous to the staff and use of magnetic beads requires patience and skilful laboratory staff.
The IMS method is amenable to automation which could be applied in diagnostic laboratories where there are high volume workloads.
79
CHAPTER 5
Identifying Volatile Organic Compounds (VOCs) Released from L. longbeachae
5. Identifying Volatile Organic Compounds (VOCs) Released from L. longbeachae 5.1. Introduction
There are many microbial pathogens that can be the causative agents of pneumonia in humans.
Legionella species are among these pathogens and the pneumonia they provoke cannot be distinguished clinically or radiographically from other causes of pneumonia. Some cases of
Legionnaires’ disease can be diagnosed by a urine test but this only detects serogroup 1 of L. pneumophila [6; 147]. All other types are detected by culture or PCR of sputum samples, which can be very difficult to obtain in confused, elderly or very young patients. In a study on 107 cases of L. longbeachae infection, only about half of the patients were reported to have productive cough and 38% of patients had dry cough. Therefore, collection of respiratory specimens is difficult and Legionnaires’ disease often escapes diagnosis [175]. Obtaining respiratory samples is a difficult task as well and some procedures are invasive to the patients
(Figure 5.1). As such, a breath test which could detect the presence of Legionella species within the lung is an attractive diagnostic option.
Figure 5.1: Different respiratory sampling approaches categorised based on the invasiveness of their nature.
This gradient shows a range of different respiratory specimen collection method starting from lung biopsy as the most invasive method (red) to broncho-alveolar lavage collection, induced sputum, expectorated sputum, breath condensate collection and finally gaseous exhaled breath (green) as the non-invasive collection method.
81
Exhaled breath contains an abundance of compounds produced from metabolism. Monitoring the volatile organic compounds in breath samples taken from patients is an expanding area in development of diagnostics for many diseases. There are several commercially available techniques to detect and quantify VOCs with different sensitivity and specificity levels as described in Chapter 1. GC-MS method can provide a very sensitive measurement of these compounds in a given breath sample. Obtaining information about the metabolic composition in breath samples and VOCs concentrations can supply the clinicians with valuable information about the underlying health problems in a patient. Detecting and quantifying the characteristic
VOCs released from microbes can also be a useful diagnostic approach to identify the colonising microbes inside the body. Collection of VOCs in the headspace (HS) of broth culture or the breath sample can be performed by pre-concentration using solid‐phase micro-extraction
(SPME) fibres to be analysed by GC-MS [78; 176].
This chapter describes the preliminary experiments conducted to establish a VOC profile of L. longbeachae by analysing the VOCs released from these bacteria grown in broth culture using
HS-SPME/GC-MS method. Furthermore, focusing on unique mass spectra and searching their identity through NIST mass spectral library (National institute of standards and technology) as a reference database was planned to identify specific biomarkers for Legionella.
82
5.2. Materials and methods
5.2.1. Bacterial strains and culture
Three clinical strains belonging to the species L. longbeachae including the standard
ATCC33462 strain and two clinical isolates were obtained from Canterbury Health
Laboratories for this study. Bacteria were cultured on BCYE agar at 36oC for 3 days. A sheep blood agar control plate was used to ensure pure cultures of L. longbeachae. BYE broth was prepared according to the protocols to use for liquid culture of bacteria in order to collect the headspace gases released by the cultured Legionella cells in screw-capped vials (Section
2.2.4.5.1). A 1% inoculum from an initial culture was added to 10 mL fresh BYE broth in a 50 mL bottle. The cultures were incubated at 36oC shaking at 100 rpm. Growth curves were performed by taking a 100 µL sample of culture over a period of 72 h; this was diluted in 900
µL phosphate buffered saline (PBS) and an optical density reading performed at an OD600
(8453 UV-Vis, Agilent, USA).
5.2.2. Breath samples
One hundred community acquired pneumonia patients were assessed and 64 patients who fitted into the study criteria established for respiratory research at the University of Otago
Christchurch were recruited. The study was approved by the Southern Health and disability ethics committee (HDEC) (ref: URB/07/07/026/AM03). All patients provided written informed consent. Breath samples were collected in sterile glass bulbs and SPME method was applied for pre-concentration of breath volatiles with overnight incubation at room temperature. The collected exhaled breath samples were analysed by SPMS/GC-MS as a full- scan mode in order to compare the profiles with the candidate biomarkers found in pure culture of bacteria.
83
5.2.2.1. Headspace analysis of Legionella broth culture
The growth and headspace analysis was performed in three replicates for each strain. Time point sampling started from time zero for up to 72 h as the stationary phase of the bacterial growth. Full-scan analysis was performed using solid phase micro-extraction coupled with gas chromatography-mass spectrometry.
5.2.2.2. SPME pre-concentration of culture headspace
To achieve this, volatile compounds were pre-concentrated from the culture headspace using a
DVB/CAR/PDMS StableFlex (2 cm) fibre (Supelco, Bellefonte, PA, USA). Each SPME fibre was pre-conditioned in an injector port at 250°C for 10 min, one chromatogram was recorded as a test. Then the activated clean fibre was inserted into a bacterial culture vial through the septum without agitation and was exposed to the headspace of in vitro culture for a period of 1 min.
5.2.2.3. Gas chromatography – mass spectrometry (GC-MS)
A Saturn 2200 GC/MS system (Varian, Palo Alto, USA) was used to perform GC-MS analysis.
Thermal desorption was used to release the absorbed molecules on the fibre to the gas chromatography device by directly injection of the fibre into the GC/MS port. A ZB-wax 30m
× 0.25 m × 0.25 µm column (Phenomex, Auckland, New Zealand) was coupled to a programmable temperature vapouriser (PTV-1079) injector. Temperature of the injector, ion trap, manifold and transfer line were 250, 200, 60 and 250°C, respectively. The oven program commenced at 60°C for 2 min and was raised to 260°C at a rate of 10°C/min, at which temperature was maintained for a further 2 min. Zero grade helium flow was set to a constant rate of 1.2 mL/min and the split vent was opened to a ratio of 1:50 after 1 min.
84
Initial identification of markers was performed in the electron impact (EI)-mode as full scan
(m/z 10-450). VOCs released from BYE broth that was not inoculated with bacteria was also analysed to use the generated profile as the control. To achieve this aim, 10 mL of fresh non- inoculated BYE broth inside a glass vial that was incubated at 37oC was taken out of the incubator and a clean desorbed SPME fibre was dipped into the headspace area of the vial through a septum for 1 min to collect the VOCs. This analysis was carried out at time zero, 12 h, 24 h, 48 h and 72 h post-incubation of BYE liquid medium.
5.3. Results
5.3.1. Identification of a candidate volatile biomarker for Legionella
Headspace of BYE broth was analysed, and a full scan volatiles profile of this medium was obtained as shown in Figure 5.2. This is a chromatogram obtained from headspace of the medium without inoculation with Legionella over time, starting from time zero, and continuing at time intervals of 12, 24, 48 and 72 h. There are several peaks that increase over time. These peaks were not identified as this served as the control for these experiments.
85
Figure 5.2: Full scan chromatogram of BYE broth headspace.
This profile was obtained using GC-MS under electron impact (EI) condition as the blank control resulting from the across results at different time points at time zero, 12, 24, 48 and 72 hours after incubation at 36oC.
Figure 5.3 demonstrates the mass spectral profile of headspace released from in vitro culture of L. longbeachae serogroup 1 cultured in BYE broth after 12, 24, 36, 48, and 52 h post- incubation. A peak of interest started to appear with retention time of 3.940 min with increasing trend over time. Full scan analysis showed this peak of interest was not identified in media control with m/z ratio of 55 was measured for this peak of interested using MS/MS. This peak of interest was identified in all replicates.
86
A
B
C
Figure 5.3: Experimental spectra for the discovered peak, obtained under GC/MS EI mode, from in vitro L. longbeachae sg1 culture.
Panel A shows the full scan chromatogram of 1 min SPME headspace exposure for L. longbeachae sg1 after 12, 24, 36, 48 and 52 h of incubation at 36oC. The peak of interest can be seen at 3.940 min to be increasing over time. Panel B shows mass spectral changes of the increasing peak of interest (1A) over time up to 52 h. Panel C shows mass spectra of ions of the peak of interest under EI conditions, based on the mass to charge ration (m/z).
A number of volatile compounds were identified using NIST library identification. The search results showed that nearest chemical compounds for the peak of interest include 3-
87
methylhexane (C7H16) [589-34-4], 2-cyclopropylbutane (C7H14) [5750-02-7] and aminomethyl-cyclopropane (C4H9N) [2516-47-4]. Analysing the headspace of L. longbeachae serogroup 2, showed that the identical peak is present in its metabolic profile as well with retention time of 3.956 min as shown in Figure 5.4.
A
B
Figure 5.4: Experimental spectra for the discovered peak, obtained under GC/MS EI mode, from in vitro L. longbeachae sg2 culture.
Panel A shows a chromatogram of the analysed headspace of L. longbeachae sg2 culture with increasing peaks at different time intervals up to 72 h. Panel B shows the mass spectrum of the peak of interest (1A) with m/z of 55 at retention time of 3.956.
Due to some technical limitations, the clinical study was halted. Therefore, it was not possible to examine the breath samples for the presence of the peak found among the VOCs released from the in vitro culture.
88
5.4. Discussion
The main finding of this chapter was to discover a peak of interest in both serogroup 1 and serogroup 2 of L. longbeachae, while it was absent in BYE liquid medium. The molecules with similar m/z ratio on NIST spectral library were 3-methylhexane (C7H16), 2-cyclopropylbutane
(C7H14) and aminomethyl-cyclopropane (C4H9N). Discovery of novel volatile biomarkers is of an increasing interest in diagnostics including infectious diseases such as respiratory infections.
This research area has been well established in our group and several human pathogens such as P. aeruginosa, A. fumigatus and M. tuberculosis have been analysed for their volatile compounds [77; 80; 82].
Similar to other areas of Legionella microbiology, the literature is very limited about biomarkers of Legionella. In a study conducted by El Qader and colleagues on identifying
VOCs in bacterial and viral respiratory infections, L. pneumophila was among the tested microorganism [177]. The results of GS-MS analysis showed that the bacterial strains produced
VOCs identified as heptane (C7H16) and methylcyclohexane (C7H14) which are connected to the consequences of oxidative stress as part of the lung inflammation events. Interestingly, methylated cycloalkanes have been found in VOC profile of breath of patients infected with
M. tuberculosis [178].
To detect the peak of interest found in pure bacterial culture in patients, breath samples from pneumonia patients to search for this compound in their breath volatile profile as a novel breath test for diagnosis of L. longbeachae bacteria. Full scan chromatograms of the breath samples were obtained from GC-MS these data could be compared with the results of the in vitro culture
GC-MS data by performing principal component analysis to compare retention times to identify the compounds and also the concentration of the compounds as area under peak.
However, the GC-MS instrument failed, so it was impossible to continue conducting these experiments. A new GC-MS device has been purchased subsequently and it is expected that in
89
near future, this study will be repeated using the new instrument. The obtained results can be compared with the preliminary results obtained in this work.
Alongside the identification of Legionella VOCs and investigation of breath samples of pneumonia patients for them, several other experiments can be conducted in future work to expand knowledge about Legionella VOCs and their potential application as a diagnostic laboratory method. These include a) detection and identification of VOCs released by
Legionella incubated with human macrophage and neutrophils to compare this VOCs profile with VOCs released from in vitro culture of Legionella , b) determination of limit of detection and the best sampling parameters for the sensitive detection of identified Legionella VOCs, c) investigation into the prevalence of the identified volatiles in food products, beverages, medications and the environment, which may cause confounding results.
90
SECTION II: ENVIRONMENTAL EXPERIMENTS
SECTION II: ENVIRONMENTAL EXPERIMENTS
CHAPTER 6
Environmental Investigation of Potting Mix Products for L. longbeachae
6. Environmental Investigation of Potting Mix Products for L. longbeachae 6.1. Introduction
Legionella bacteria are widely distributed in the environment. L. pneumophila is commonly isolated from water and aquatic environments such as spa pools and cooling towers, whereas
L. longbeachae is found in the soil, composted plant materials and potting mix products [42].
L. longbeachae was first identified from a patient in Long Beach, California in 1981, after which the second serogroup of this species was reported the following year [36]. Sporadic cases of LD with L. longbeachae have been reported globally. For example, L. longbeachae serogroup 1 was first reported in Europe as the potential causative agent of LD in a patient in
Sweden in 1987 [179]. In Australia, the first cases of LD due to L. longbeachae were identified in 1988. By 1989, 30 patients were diagnosed with LD related to this Legionella species [46].
The hypothesis about the link between clinical cases and environmental sources of LD due to
L. longbeachae was strengthened by the isolation of these bacteria from the compost products.
In a study in Southern Australia, 58% of the potting mix samples were culture positive for L. longbeachae [111]. A similar study was conducted in Japan and 37.5% (9/24) samples were culture positive for L. longbeachae using an enrichment method. These potting mix samples contained composted wood products [47]. Several studies in Scotland revealed the relationship between LD and exposure to plant growing media. The first study found that L. longbeachae isolated from patients and potting mix associated with the cases, had identical genotypes for two cases. Both of these patients used the same potting mix product [40]. In another study in
2012, four cases of LD were found to be caused by L. longbeachae serogroup 1 between 2008 and 2010. All these patients were exposed to compost [180]. Similar work conducted by Potts and colleagues showed that six cases were due to L. longbeachae and all the affected individuals were amateur gardeners with frequent exposure to horticultural growing media.
92
Interestingly, L. longbeachae was isolated from the plant growing media related to five cases
[181].
Further investigations need to be conducted on the environmental sources of L. longbeachae including the components of the potting mix products, and also transmission of L. longbeachae from these sources to humans. The aim of this study was to determine the risk of L. longbeachae contamination in New Zealand made potting mix products. Testing the commercially-prepared potting mix products requires good collaboration of the industrial producers with researchers. This opportunity became available for us as Professor David
Murdoch developed a collaboration with a major gardening products supplier, which had three manufacturing sites around New Zealand.
In view of the success of the testing strategy for L. longbeachae used in CHL laboratories for clinical samples using qPCR method in the past decade, this opportunity was given to our study to investigate the potting mix products by this method. Isolation of the bacteria by culture was followed by confirmatory identification using MALD-ToF. qPCR, which has a higher sensitivity than culture, was chosen as the preliminary screening tool. However, culture was also carried out (Chapter 7) in order to use the isolated environmental strains in future studies for genetic analysis and comparison with the clinical strains.
93
6.2. Materials and Methods
6.2.1. Study design
The study design was based on the probability sampling strategy for random selection of samples to enable us to determine the contamination in the areas of interest and also in different types of feed stock accurately. Methods for sampling were determined when on site as we did not know what to expect in terms of weather conditions and the type of potting mix products available for sample on the collection days. Samples were taken based on the simple random sampling method. qPCR was used as the preliminary screening tool for this evaluation, followed by microbiological culture as the secondary endpoint. All production sites used pine bark but other feedstock varied between sites which were obtained from local sources (Table
6.1).
Table 6.1: Sample size of the three investigated potting mix products facilities
Production facility Sample size Site 1 348 Site 2 250 Site 3 570 Overall sample size 1168
Samples were taken from around the base of feed stock piles and between 15 and 50 cm depth approximately 2 meters apart covering the available surface. Rare event sample size calculator from https://www.statstodo.com/SSizRareEvent_Pgm.php suggesting that a sample size about
100 (95) would give 85% power to get at least one positive sample in feed stock at each of three sites, assuming the true proportion of samples is 2%. Input for this was from previous studies in which L. longbeachae was found in 33/55 samples in Australian potting mixes, 0/19 in Europe and 2/24 in potting mix samples in Japan [46; 47; 111; 180; 181]. However, there have been no systematic studies and no data from New Zealand potting mix products.
94
6.2.2. Production site sampling
Three different potting mix production facilities around New Zealand were selected for this investigation. Samples were collected from these locations at various times throughout 2017
(Table 6.2).
Table 6.2: Details of sampling from three growing media production facilities
Facility Sampling date Temperature Humidity Wind Pressure Lower South island 12th April 12oC 96% 28km/h SE 1017mbar (site 1) Upper South island 17th May 17oC 94% 23km/h 1008mbar (site 2) NE Upper North island 31th May 18oC 71% 5km/h NE 1013mbar (site 3)
Approximately 50 g of samples were collected, from each site into individual 70 mL collection containers and were transferred to the laboratory for testing. Details of samples are listed in
Appendix 2.
6.2.3. DNA extraction
To extract DNA for qPCR, five grams of the samples was weighed and transferred to sterile zipper plastic bags. The bags were crushed using a mortar and pestle to grind their contents to fine powder. The processed samples were transferred aseptically to a Falcon tube containing
50 mL ultrapure™ distilled water (ThermoFisher Scientific, Waltham, USA). The sample was vortexed and shaken vigorously and left to stand for 30 min at room temperature. DNA was extracted from samples using the commercial GenElute Bacterial Genomic DNA kit (Sigma-
Aldrich, St. Louis, MO, USA) following the manufacturer’s instructions. DNA was extracted
95
from 3-day old colony of L. longbeachae ATCC33462 to act as a positive control alongside a negative control of ultrapure water.
6.2.4. qPCR of potting mix materials for L. longbeachae
TaqManTM quantitative real-time PCR targeting ITS region of L. longbeachae was carried out for preliminary screening of the collected samples. An artificial DNA sequence (ART) was included as an internal control of PCR reaction inhibition. First, a standard curve for potting mix products was generated by using DNA extracted from autoclaved potting mix samples spiked with six different dilutions of L. longbeachae sg1 serially diluted from 102 to 107
CFU/g. DNA was purified as described before. This was repeated three times in form of biological replicates and the mean was applied as the standard curve. DNA templates were applied for a TaqmanTM qPCR reaction using primers and a probe specific for L. longbeachae.
The resulted Ct values were plotted versus the log of bacterial concentrations for generating the curve and evaluating the linearity of the results.
6.2.5. IMS-PCR
To evaluate the possible impact of applying IMS method on DNA extraction and subsequent qPCR for the environmental samples. Eight environmental samples with positive qPCR results were selected. One mL of these processed samples was added to 50 µL of prepared Dynabeads coupled with anti-Legionella antibody prior DNA extraction and qPCR.
6.2.6. Data analysis
Data obtained were analysed using SPSS Statistics software version 25.0 (SPSS Inc, Chicago,
IL). One-way ANOVA analysis was conducted and Tukey Post Hoc test were carried out to compare the means and variances of different groups. T-test statistical analysis was used to
96
compare two groups of PCR results with and without application of IMS for DNA extraction of the DNA templates.
6.3. Results
6.3.1. qPCR of the potting mix products samples
A standard curve was generated (Figure 6.1) for the sterile potting mix sample spiked with L. longbeachae sg1. The plot below shows the curve of the qPCR result.
Figure 6.1: standard curve generated from the qPCR results of the spiked potting mix samples.
DNA extracted from six commercial potting mix samples spiked with 2×102 to 107 CFU/mL of L. longbeachae sg1 ATCC33462 was tested by qPCR. The resulted Ct values of these samples were plotted versus the log of the bacterial concentrations (cell/mL) to generate the standard curve with efficiency of 78%.
97
The overall qPCR results from sampling of all three production sites are detailed in Table 6.3, categorised by sample type. The results of qPCR for L. longbeachae showed that in site 1, 14% of the collected samples were PCR positive for L. longbeachae. Bark samples had the highest percentage of PCR positivity (30%), followed by sawdust (17.8%) and samples collected from the mixing area in the production site (5.5%). There was no PCR positive sample among the peat samples. In site 2, only 2% of the tested samples were PCR positive for L. longbeachae.
The only positive samples belonged to the bark-based samples (2%) and compost (10%).
There was no PCR positive sample among other sample types including mixing area, sawdust, peat moss and potting mix. Site 3, had the lowest positivity PCR for L. longbeachae with overall percentage of 1.75%. However, it was found that all the positive samples were among the bagged final products stored in an open area outside (5.5%), while there was no positive sample among the feed stock from the indoor area of the production site. The positive bagged products included compost 5/18 (27.7%), wood chips 3/18 (16.6%), potting mix 1/9 (11.1%) and a lawn preparation mixture 1/9 (11.1%).
Table 6.3: qPCR results of the samples collected from three different sites
site 1 site 2 site 3 Total Bark 30/100 (30%) 4/150 (2%) 0/90 34/340 (9.7%) Sawdust 17/95 (17.8%) 0/40 0/50 17/185 (9.1%) Peat 0/100 0/5 0/100 0/205 Composted bark 0 2/20 (10%) 0/130 2/150 (1.3%) Green waste 0 0 0/130 0/130 Pumice 0 0 0/20 0/20 Mixing area 3/53 (5.5%) 0/20 0 3/73 (4%) Bagged products 0 0 10/180 (5.5 %) 10/180 (5.5%) Total 50 / 348 (14.3%) 6/235 (2.55%) 10/700(1.4%) 66/1283 (5.1%)
98
The overall results of all three sites altogether showed that 5.5% of the samples were qPCR positive for L. longbeachae. Based on sample type, 8.9% of bark and bark containing products,
9.1% of sawdust and 3.7% of bagged potting mix products were PCR positive. All peat and green waste samples were negative. Table 6.4 describes details of mean and the lowest and highest Ct values for each sample type. Categorising the PCR positive samples of all three sites based on the sample type revealed that the bark-based samples had the highest frequency
(9.7%), followed by sawdust (9.1%), bagged potting mix (5.5%), and composted bark (1.3%).
The lowest mean Ct belonged to the compost (43.81) and the lowest minimum Ct value was from a bark sample (38.38).
Table 6.4: Statistical description of Ct value of PCR positive samples
Number of positive qPCR Mean SD Min Max Bark 34 44.81 2.53 38.38 49.12 Compost 7 43.81 2.3 40.3 47.85
Potting mix 6 46.03 3.73 38.74 48.56
Sawdust 17 46.06 1.76 43.18 48.75
Total 64 45.15 2.51 38.38 49.12
Compost samples had the lowest mean Ct value. Comparison of mean Ct values of two groups by student T-test showed that the only significant difference was between the Ct values belonging to the compost and sawdust samples with the two-tailed P-value of 0.01 (95% CI:
From -4.0404 to -0.4596, df = 22).
99
6.3.2. IMS-PCR
Results of IMS-qPCR using the eight examined potting mix products showed that IMS-qPCR lowered the Ct value (M = 44.34, SD = 1.87) which represents higher DNA load identified in the samples. The mean Ct value of the initial qPCR was higher (M = 47.2, SD = 1.67). T-test statistical analysis showed that the difference was statistically significant (P < 0.01) (Figure
6.2).
Figure 6.2: Comparison of Ct value for qPCR with and without suing IMS method for DNA extraction.
Results of PCR and IMS-PCR of eight positive bark samples are shown in this graph. Application of IMS for DNA extraction to be used as DNA template for qPCR, reduced the Ct values (in green) in comparison of performing qPCR for the identical bark samples without using IMS for DNA extraction (in blue).
100
6.4. Discussion
The current study confirmed the presence of L. longbeachae bacteria at all three different potting mix production sites at different locations in New Zealand. Overall, 5.5% from the total of 1168 tested samples were qPCR positive for L. longbeachae. Potting mix from these sites are sold widely across both the North Island and South Island of New Zealand. This is consistent with the widespread distribution of Legionnaires disease in New Zealand.
The qPCR positivity rates varied between sites with 14.3%, 2% and 1.7% for site 1, site 2 and site 3 respectively. Several factors might contribute to the difference of prevalence, such as the source and proportion of the raw materials, composting methods applied at different sites, weather conditions of each location such as temperature and humidity and storage conditions of the materials in each site. For example, the lowest temperature was at site 1 (12oC), while site 2 and site 3 had higher outdoor temperatures of 17oC and 18oC respectively. The highest humidity on the day of sampling at site 1 with 96% humidity, followed by site 2 (94%) and site
3 (71%). Interestingly, site 1 had the highest contamination rate with 14.3% qPCR positive samples for L. longbeachae. Further studies could indicate how conditions on the day of sampling or over the preceding weeks influence the load of Legionella in feedstock or the final potting mix product. These factors might be associated with variation in rates of reporting of identification of L. longbeachae in the environment in different countries worldwide. It is also possible that the storage is a contributing factor for Legionella presence in these products. The three visited sites were storing the materials short term (up to three months). Interestingly,
Legionella were not found in the short-term storage facilities, while positive samples were among the ones from long-term storage sites and the composting facilities.
All qPCR positive samples in this study were pine bark (34/1283) or pine bark containing products (30/1168). The results of qPCR were in accordance with the previous studies demonstrating the presence of Legionella in pine bark containing potting mix products.
101
Composted bark-based materials such as hammer-milled pine bark and sawdust are the primary component of potting mix products for soil amendment in Australia and New Zealand [111].
This suggests that pine bark is an important source of L. longbeachae in New Zealand as many tons are used each year for manufacturing potting mix.
The source of the bark may contribute to the different prevalence of Legionella in different products. Unlike Australia and New Zealand, the Japanese potting mix products contain composted materials from oak tree (different species of genus Quercus) and Japanese oak tree
(Quercus mongolica var. crispula), with only 8.3% (2/24) isolation rate of L. longbeachae, compared to the higher rate (58%) among the samples containing pine bark in Australia (Koide,
2001). A study in Switzerland reported 89% (41/46) of the potting mix samples were found to be culture positive for Legionella spp. including 4.3% positive for L. longbeachae , but the source of the bark in the potting mix is unclear [43].
An important characteristic of Legionella is the ability to survive under acidic conditions.
Wooden materials and pine products such as pine needles have acidic pH values ranging from
3.5 to 7 depending on the plant species which may favour growth of Legionella over less acid tolerant species [182]. Genetic studies have suggested that L. longbeachae contains coding regions responsible to produce cellulase and beta-glucosidase [57]. These enzymes can degrade carbon structures of plant materials and provide an energy source and help provide a niche of these organism [183].
All 130 tested green waste samples were PCR negative for L. longbeachae which was unexpected as green waste contains some wood and bark. There are several possible reasons for this result. Composting itself can influence the presence of the microorganisms. However, these samples were from one batch of feedstock and more samples need to be tested at different time points in future studies. The manufacturer of the green waste that was tested stressed that
102
this was a very well controlled industrial process which ensured high temperatures were reached and maintained. These conditions may be different for other manufacturers and home based composting so the results may not be generalisable to other green waste products.
The literature is limited regarding the presence of Legionella in green waste. For example, one study conducted in Switzerland investigated the presence of Legionella by culture method in eight green waste collection sites including composting and storage facilities [184]. Although all fresh green waste samples were culture negative in 7 out of 8 facilities, some samples were positive for several species of Legionella in one site including L. micdadei, L. oakridgensis, L. jamestowniensis and L. cincinnatiensis. However, L. longbeachae was not reported among these culture positive samples. The source of these green waste materials were plant materials such as branches, leaves or grass collected from five community centres which were composted in three commercial facilities. The type and species of the plants was not addressed in this study but different plants may provide niches preferred by the various Legionella species [184].
Peat samples in this study were also PCR negative for L. longbeachae. This is consistent with results of other studies conducted in Europe and Japan, in which peat samples were also free of L. longbeachae contamination [47; 111]. Bog peat is still the predominant material used in growing media in European Union due to its availability and lower price compared to other materials such as bark or coir [185]. Composted materials such as bark, sawdust, wood fibre, green waste and timber by-products are of increasing use in growing media industry as there is a concern about the environmental consequences of peat mining. Monitoring for the presence of Legionella in potting mix products is necessary after using the peat alternatives in these countries.
There are several limitations for this study. Firstly, the qPCR method applied for this study was optimised for potting mixed products. Based on the generated standard curve, Ct value of 48
103
was used as the cut-off value. This value was identified by extrapolating the curve backwards to an estimated 10 CFU/mL of L. longbeachae sg1 ATCC33462. To validate the high Ct values as true positives, eight samples were tested using IMS for DNA extraction followed by qPCR method. These results added certainty to the previous results of qPCR for the tested potting mix products. It was also clear that the Ct value of the environmental samples was different from the clinical specimens (Chapter 4), this might be due to the complex nature of the potting mix products and the chemicals they contain. The difference of Ct value of positive PCR samples of different materials may be related to the difference of bacterial load in different types of samples, effect of chemical make-up of these materials for the presence and growth of
Legionella, or variability of Ct values for environmental samples due to their complex matrix.
It is possible that some samples contained very low levels of L. longbeachae below the level of detection of the assay and using IMS would improve the sensitivity of the assay. In a pilot study during this work, IMS was used prior to DNA extraction for qPCR and the results of qPCR showed an improved Ct values of the selected sample which were tested qPCR positive for L. longbeachae. The Ct values of these samples were up to five cycles lower using IMS-
PCR which is an improvement, given the fact that the cycles represent exponential values of
DNA load of the samples.
Another limitation of qPCR method is not differentiating live and dead bacteria as it identifies any target DNA in the samples. The use of viability dyes such as propidium monoazide (PMA) with ability to bind to double-stranded DNA of viable cells through the photo-activation mechanism could be used to differentiate viable cells from dead cells in qPCR. Legionella are exposed to high temperature and other environmental factors during composting process which may kill the bacteria [186].
104
In conclusion, the only products that were qPCR positive were bark of bark-containing products such as sawdust and potting mix. Pine trees might be a natural reservoir for L. longbeachae. This work established the foundation for further studies to examine the presence of L. longbeachae in living trees as their natural reservoir. Further studies are required to elucidate the clinical relevance of the presence of L. longbeachae in these manufactured potting mix products.
105
CHAPTER 7
Culture of qPCR Positive Potting Mix and Development of The Isolation Methods for
L. longbeachae
7. Culture of qPCR Positive Potting Mix and Development of The Isolation Methods for 7.1. Introduction
Isolation of Legionella from environmental samples is a challenging task. Compost and potting mix products have a wide range of environmental microorganisms that complicate the isolation of the target organism. The standard method to reduce the number of competing organisms on culture is the acid pre-treatment. This inhibits acid sensitive organisms but spares Legionella species that are relatively tolerant to low pH [109] and reduces heterotrophic growth of the contaminating organisms.
The current culture method was developed based on the initial studies conducted on L. pneumophila and water samples which is the basis on international standards such as the international organization for standardization (ISO 11731:2017). The only standard culture protocol developed for potting mix products was created jointly by standards Australia and standards New Zealand (AS/NZS 5024(Int):2005) has been withdrawn without replacement indicating dissatisfaction with quality of the protocol.
Previous studies have also shown that different strains of Legionella such as environmental strains and laboratory stock strains show different levels of sensitivity to acid pre-treatment
(Roberts, 1987). Further studies need to be conducted for various compost and potting mix products and different strains of L. longbeachae to determine the sensitivity and the optimal pH value for pre-treatment of environmental samples such as potting mix.
The aim of this study was firstly; to confirm the presence of L. longbeachae in qPCR positive samples using a standard protocol for acid pre-treatment, and secondly; to compare a variety of alternative techniques for the isolation of L. longbeachae in an attempt to improve upon the current acid pre-treatment method for the isolation of L. longbeachae from environmental samples. These include the IMS method developed in this research (Chapter 3), indirect-IMS
107
using a second antibody, sucrose gradient centrifugation, GVPC pretreatment, and a combination of GVPC pre-treatment and IMS.
7.2. Material and Methods
7.2.1. Acid pre-treatment – culture for L. longbeachae
7.2.1.1. Stored feedstock and bagged product samples
Samples of feedstock bark, sawdust and final bagged product taken from the manufacturing sites were labelled and stored in plastic bags at room temperature after DNA extraction prospectively during the experiments (Chapter 6). The qPCR positive samples were retrieved for testing. A standard protocol for culture of L. longbeachae was used. This has been described in detail for sputum in Chapter 4 but required adaption for environmental samples.
7.2.1.2. Preparation of HCl:KCl buffer (pH 2.2)
A mixture of hydrochloric acid and potassium chloride (HCl:KCl buffer) was used to pre-treat the stored samples according to CDC Laboratory Guidance for Processing Environmental
Samples [187]. Buffer was made before use by mixing 39 mL of 0.2M HCl with 250 mL of
0.2M KCl. The pH was adjusted to 2.2 using 1M KOH, and was sterilised by autoclaving at
121°C for 15 min. pH value was checked after autoclaving.
7.2.1.3. Acid pre-treatment
In brief, 5 g of a commercial potting mix sample that had been previously softened using a sterile pestle and mortar was weighed, and up to 50 mL of sterile distilled water added. This was mixed by shaking and vortexing and allowed to stand for 30 min. One mL of the supernatant was treated with equal volume of HCl:KCl buffer pH 2.2. 100 µL of Tween 80 was
108
added to the suspension and mixed well. The mixture was incubated at room temperature for
10 min, then sonicated for 4 min in a water bath at 36°C. The sample was serially diluted in
HCl:KCl buffer from 1:10 to 1:104 dilutions. 100 µL of each dilution was spread plated onto
GVPC agar. Plates were incubated in a 36°C incubator for up to seven days and were checked from day three onward for the presence of L. longbeachae growth.
7.2.2. Development of isolation methods for L. longbeachae from potting mix
7.2.2.1. Bacterial inoculum
L. longbeachae sg1 ATCC33462 was cultured on CYE agar and after three days, the colonies were used to make a standard inoculum by diluting a bacterial suspension equivalent to 0.5
McFarland standard (~1.5×108 CFU/mL).
7.2.2.2. Potting mix for spiking experiments
To develop a method for the isolation of Legionella, 5 g of a commercial potting mix product was autoclaved at 121oC for 15 min. The sample was then spiked with 1 mL of 1.5×106
CFU/mL of L. longbeachae sg1 ATCC33462 bacterial suspension. The same bacterial suspension was used to spike a non-sterile potting mix sample as well. Non-sterile potting mix was confirmed to be negative for Legionella by acid treatment-culture and qPCR before spiking the samples.
109
7.2.2.3. Dynabeads coupled with antibody
M280 Tosylactivated Dynabeads coupled with the rabbit polyclonal anti-L. longbeachae antibodies were used as described previously (Chapter 3). The coupled magnetic beads were stored at 4oC during the experiments.
7.2.2.4. Anti-rabbit IgG antibody
Goat anti-rabbit IgG antibody (Jackson ImmunoResearch Laboratories, Inc., West Baltimore,
PA, USA) was purchased for the secondary antibody. The antibody was stored at 4oC.
7.2.2.5. GVPC solution
A vial of selective GVPC supplement (Oxoid SR0152) was added to 500 mL of ultrapure distilled water to be used in the experiments. The final concentration of these antibiotics were as follows per litre: Glycine (3 g/L), Vancomycin (1 mg/L), Polymyxin B (80,000 IU/L) and
Cycloheximide (80 mg/L).
7.2.2.6. Immunomagnetic separation prior to culture
One mililitre of the washed samples were mixed with 50 µL of Dynabead M280 Tosylactivated coupled with anti- L. longbeachae rabbit polyclonal antibody and incubated for 30 min at room temperature. IMS method was conducted for the potting mix sample as described in Chapter 3.
7.2.2.7. Indirect immunomagnetic separation (i-IMS)
IMS was performed as described in Section 7.2.2.6 but with an extra step of using a secondary antibody against the rabbit anti-Legionella antibody. To carry out indirect immunomagnetic separation method, the sample was first treated with 50 µL of rabbit polyclonal antibody. After
110
incubation at room temperature in a shaking incubator, Dynabeads coupled with AffiniPure
Goat Anti-rabbit IgG (Jackson ImmunoResearch Immunoresearch Labs. USA) were added to the sample to capture the polyclonal antibody. After washing the final mixture, three replicates as 50 µL of washed Dynabeads captured by the magnet were inoculated onto GVPC agar to examine the growth of any L. longbeachae.
7.2.2.8. Sucrose gradient centrifugation
One mililitre of five solutions of sucrose ranging from 10% to 50% (w/v) dissolved in distilled water, was added to a 15 mL tube. The gradient was established by layering the various sucrose concentrations in a stepwise manner beginning with the 50% solution. One mL of the spiked sample was added to the tube, and centrifuged at 240 rpm for 30 min. 100µL of the top layer of supernatant was plated onto a GVPC plate and incubated at 36°C for up to seven days. The three plates were checked from day three on a daily basis for the presence of Legionella-like colonies.
7.2.2.9. GVPC decontamination
One mL of suspended potting mix sample with equal volume of GVPC antibiotic supplement was left at room temperature for 30 min. 100µL was inoculated onto a GVPC plate. The plates were incubated at 36oC for up to 7 days and checked from day three.
7.2.2.10. GVPC/IMS
After treating 1 mL of suspended potting mix sample with an equal volume of GVPC solution,
1 mL of the mixture was transferred to a 1.5 mL microtube and 50µL of coupled Dynabeads was added and mixed well. This mixture of antibody-coupled Dynabeads and the suspended
111
potting mix was rotated at room temperature for 30 min and the beads were isolated using the magnet and resuspended in 100 µL distilled water. Next, this suspension was inoculated onto a GVPC plate. The plates were incubated at 36oC for up to 7 days and checked from day three.
7.2.3. Isolation and identification of L. longbeachae
Aliquots of the final mixture of all experimental methods were diluted serially (10-fold) by using relevant diluents and were inoculated onto GVPC plates, and incubated at 36°C for seven days. Colony count was carried out for the bacteria on the surface of the plates. Identification of these colonies were confirmed as Legionella by using MALDI-ToF mass spectrometry.
7.2.4. Determination of lower limit of detection
After completion of the experimental methods, the best method was chosen to be evaluated for its limit of detection. The use of viability dyes such as propidium monoazide (PMA) with ability to bind to double-stranded DNA of viable cells through the photo-activation mechanism could be used to differentiate viable cells from dead cells in qPCR.
7.2.5. Statistical analysis
To compare the efficiency of different isolation methods, results were analyses statistically by using One-way ANOVA analysis for independent samples, followed by post-hoc Tukey HSD test to conduct the pair-wise comparisons of sample means.
112
7.3. Results
7.3.1. Culture for qPCR positive feedstock and bagged potting mix products
All 66 qPCR positive samples were culture negative for L. longbeachae after acid wash pre- treatment. There were some Legionella-like colonies growing which were identified as
Burkholderia species such as B. xenovorans, B. tropica, B. fungorum and B. phenazinium and one isolate was identified as Enterobacter cloacae using MALDI-ToF mass spectrometry method.
7.3.2. Development of isolation methods for L. longbeachae from spiked potting mix
L. longbeachae was recovered from all treated samples. Figure 7.1 shows the recovery of bacteria using different treatments. ANOVA analysis showed that among the three methods used for sterile potting mix, there was a significant difference in the recovery of bacteria (df
=2, P < 0.0001). Comparison of the mean CFU/g (mean±SD) by post-hoc Tukey HSD Test showed that recovery of L. longbeachae by IMS (1.5×105±11547) was significantly higher than acid pre-treatment (9.3×102±707; P < 0.01) and indirect IMS (1.5×103±556; P < 0.01).
Analysis of the methods used for non-sterile potting mix samples showed a significant difference among the methods (df =2, P < 0.0001). Sucrose gradient centrifugation (1.46×106±
57735) and GVPC/IMS (1.43×106±115470) had higher recovery than GVPC decontamination
(1.93×105±51316; P < 0.01). Using GVPC solution resulted in removal of contaminating microorganisms in culture. However, using sucrose gradient centrifugation method for the stored potting mix resulted in overgrowth of yeasts and fungi.
As GVPC/IMS method provided the best results, it was tested for recovery at lower concentrations of L. longbeachae. The limit of detection for the combined GVPC/IMS method was determined as 102 CFU/mL (Figure 7.1).
113
Figure 7.1: Comparison of recovery rate of different methods for isolation of Legionella from spiked autoclaved and non-autoclaved potting mix.
Potting mix samples spiked with 106 CFU/mL of L. longbeachae. A) For autoclaved samples (shown in blue): Acid wash, IMS, indirect-IMS (i-IMS) methods were used. B) For non-autoclaved samples (shown in green), Sucrose gradient centrifugation (SGC), GVPC decontamination, and combined GVPC/IMS were used. Recovery data is shown on a log scale with base of 10 for CFU/g. Error bars represent the standard error.
Untreated potting mix GVPC decontaminated GVPC + IMS Figure 7.2: Culture results of three different treatments on GVPC.
Examples of 5-day culture of untreated, GVPC decontaminated and GVPC decontaminated combined with Immunomagnetic separation are shown. Samples of non-autoclaved potting mix were spiked with L. longbeachae sg1 ATCC33462 strain.
114
7.4. Discussion
Using acid pre-treatment and culture resulted in negative results for all qPCR positive samples identified in Chapter 5. This could be due to several reasons. Firstly, as the results of qPCR suggested, most of these samples contained an estimated bacterial load of less than 102
CFU/mL of L. longbeachae. This low number of bacterial cells in the samples were likely to be below the limit of detection of the available culture methods for isolation of these bacteria from potting mix products.
Secondly, acid wash, which is the current conventional method for isolation of Legionella from potting mix and compost, is adapted for water samples followed by filtration. The parameters developed have not been optimised for potting mix and compost. There is little evidence on the sensitivity of L. longbeachae to acid conditions, the optimal pH, exposure time to the acid, or the buffering effect of potting mix on the acid. Most studies have been conducted on environmental water samples aiming to identify L. pneumophila. There is evidence that the type and amount of contaminating material in the samples may alter the performance of the buffer (0.2 M HCl buffer (pH 2.2)) on environmental water samples. For example, Kasuga and colleagues found that the final pH of the combined HCl buffer and water samples was ranged from pH 3.0 to 7.4, while when the investigators used 0.1 M potassium citrate (pH 2.2) with the same water samples, pH reached to the final level of 2.5 to 2.7, with 3 times recovery rate of Legionella from water samples and more efficient inhibitory effects on contaminating microbes [188].
Previous studies have also shown that different strains of Legionella bacteria such as environmental strains and laboratory stock strains show different levels of sensitivity to acid pre-treatment [149]. Further studies need to be conducted for various compost and potting mix products and different strains of L. longbeachae to determine the sensitivity and the optimal pH value for pre-treatment of environmental samples such as potting mix.
115
One possible reason is that the samples tested were stored for different lengths of time after sampling days at room temperature of about 20oC in the laboratory. These conditions may favour the growth of other organisms rather than L. longbeachae which is favoured by warmer conditions. It is possible competing environmental organisms such as fungi release Legionella inhibitory chemicals reducing their viability. Legionella are also known to persist in the VBNC mode under undesirable conditions such as chemical stressors so they may be detectable by
PCR but not be able to be cultured with standard techniques [189].
There was a long time interval between obtaining the samples and screening them with qPCR, due to limited access to a PCR instrument. Our workflow was that only those samples that were qPCR positive would be cultured. It is possible this had a direct effect on the viability of
Legionella organisms. In addition, it is possible that some samples contained the anti-
Legionella chemicals which may have been inhibitory to L. longbeachae. Finally, Legionella are intracellular bacteria and may be present within environmental amoeba making the sample
PCR positive but may not be released to grow under laboratory conditions. All of these factors may need to be investigated in future studies.
To identify the best method for culture of Legionella from potting mix products, five other methods were investigated. The study results showed that sucrose gradient centrifugation and combination of immunomagnetic separation with GVPC antibiotic decontamination
(IMS/GVPC) methods were the best isolation methods. However, using sucrose gradient centrifugation of a spiked sample from an old stored potting mix bag resulted in overgrowth of other organisms such as fungi and yeasts. IMS had a greater recovery (106 CFU/g) compared with the current acid wash-culture technique (103 CFU/g). Antibiotic decontamination also removed many undesirable organisms within the samples. Overall, the IMS/GVPC method was the most efficient method. It reduced the contamination rate while enhancing the maximum
116
recovery rate and specific isolation of L. longbeachae more than the other comparative methods.
IMS has been studied for L. pneumophila and results suggested an improved isolation rate with method with less contamination in comparison with samples treated with conventional acid wash method [145]. However, a further study applied the IMS method to isolation of L. pneumophila that was spiked into different types of water samples such as distilled water. This study found that the recovery was significantly affected by the type of water sample, with the recovery rate of Legionella from about 60% from spiked distilled water to 36% from cooling tower water [186]. The authors of this study suggested that the application of IMS is limited in highly contaminated samples with variable recovery rate due to the nature of samples. Small aggregates of the co-existing bacterial cells and filamentous fungi may be accumulated in the antibody-bead complex, confounding the isolation technique. For this reason, they suggested further improvement was needed to develop more efficient methods for purification of
Legionella from water samples. The strategy of adding GVPC as a decontamination step to the
IMS method substantially reduced the non-Legionella species on culture.
GVPC did not eliminate all contaminating organisms. For example, several species of
Burkholderia bacteria with natural resistance to all GVPC supplement antibiotics were isolated during the culture improvement tests. Modifying the type and quantity of antibiotics in selective media may further improve the culture results as well. For example, Casati and colleagues found better culture results for L. pneumophila and several other species of
Legionella present in compost by adding propiconazole (5×10−4mg/mL) as a fungicide to control the growth of moulds commonly found in compost [184]. Further studies with this or other similar agents could be considered.
117
The limit of detection for the combined GVPC/IMS method was determined as 102 CFU/g of
L. longbeachae. It was not possible to isolate bacteria from potting mix spiked with 10
CFU/mL. This limit might be due to the ability of the polyclonal antibody and magnetic beads used in this study to isolate the target bacteria from potting mix. The Dynabead M280 tosylactivated used for IMS were the best available commercial magnetic beads but development of monoclonal antibodies against specific antigenic targets of L. longbeachae, or an alternative type of magnetic bead may help make a more specific and sensitive method possible. In this study only 5 grams of spiked samples was used. Scaling up this to 50 g and concentrating the washed sample may also increase the chance of isolating the bacteria in potting mix. It is important to note that the IMS method may not identify bacteria sheltered inside protozoan host organisms.
In conclusion, the presence of a vast variety of organisms in samples such as potting mix and compost is a big challenge. The qPCR results from the investigations in Chapter 5 yielded Ct values of 43> longbeachae are related to samples containing less than 102 CFU/mL of L. longbeachae, which is below to the limit of detection of the GVPC/IMS culture method. It is clear that isolating bacteria in low numbers in potting mix products is difficult and needs further improvement. The acid wash protocol used was not the most sensitive method for the isolation of Legionella from potting mix products. The best alternative from these experiments was
IMS/GVPC method which enhanced consistency of the results by minimising the background contamination and increased recovery of L. longbeachae from potting mix samples.
IMS/GVPC may also remove any inhibitory chemicals potentially present in potting mix during the washing steps.
Disadvantages of this technique include the complexity and time-consuming nature of the method that may make it impractical for large studies. In addition, IMS separating and culturing bacteria located within host protozoa may not be possible with this technique. Development of
118
co-culture methods as an amplification step for intracellular Legionella is another possible tool for improving the resuscitation and isolation of these bacteria from potting mix products.
Finding simple and rapid methods is needed for further work. The IMS/GVPC protocol is promising but needs to be tested prospectively alongside collecting samples in further studies.
Although qPCR is the current gold standard for Legionella detection, isolating these bacteria from the environment contributes to understanding of its epidemiology and biology. Improving culture is an important tool for this purpose. Improvement of Legionella isolation from potting mix samples spiked with Legionella showed that combining the IMS method and GVPC decontamination can help to isolate these bacteria in future studies. IMS is useful to capture viable cells as the mechanism is based on attachment to Legionella cells with intact cell envelope.
119
CHAPTER 8
The Physico-chemical Parameters and The Presence of L. longbeachae in Potting Mix Products
8. The Physico-chemical Parameters and The Presence of L. longbeachae in Potting Mix Products 8.1. Introduction
Microbial growth is a multifactorial phenomenon and all bacteria require some general and also specific conditions for survival and growth. Understanding the role of environmental factors such as temperature, pH, humidity and nutritional requirements is key to unravelling the ecology of Legionella. L. pneumophila has been the main species of Legionella studied mainly due to its presence in aquatic systems found near to populated areas and plumbing systems. L. longbeachae has been found in potting mix and environmental samples. L. pneumophila has also been isolated from soil as well [190]. Therefore, these findings suggest the necessity to consider the role of environmental parameters including weather conditions on the presence, growth and spread of Legionella in the environment.
L. pneumophila has been successfully recovered from environmental samples within a wide range of temperature from 5.7 to 63°C [10]. These bacteria have also grown in the laboratory at different temperatures from 25 to 45°C [191]. However, a comparative study on the thermal sensitivity of four strains of L. pneumophila and L. longbeachae ATCC33462 strain, revealed that L. longbeachae was the most sensitive species with 5-log inactivation by exposure to 50°C for 20 min and 55°C for only 5 min [192]. The adaptive mechanisms evolved in L. longbeachae to survive in soil and potting mix products might explain the lack of this resistance in this species, but further studies need to be conducted to elucidate this feature.
Previous studies have demonstrated that L. pneumophila can grow and propagate at a wide pH range is from 2.7 to 9.2 in water samples [191; 193]. However, the optimal pH range for the laboratory-grown strains of L. pneumophila was found to be limited to a range of 6.5 to 7.5 by others [194].
121
With the exception of some bacteria such as Lactobacillus plantarum, almost all living organisms have an absolute requirement for iron [195]. Legionellae are known as fastidious microorganisms. During the preliminary examination of the role of trace elements on
Legionella growth, it was found that iron is a crucial growth promoter of Legionella in the form of ferric pyrophosphate in varying concentrations from 0.001 to 1% of the media [194; 196].
Therefore, iron pyrophosphate is used in microbiological media for the growth of Legionella in the laboratory. Early investigations on optimum medium to grow Legionella in vitro revealed that there is an absolute requirement for seven L-form enantiomer of amino acids including cysteine, arginine, serine, threonine, valine, methionine, leucine, and isoleucine [197-199].
Therefore, L-cysteine is added to the media to improve the growth of Legionella [194].
Rowbotham and colleagues showed that environmental protozoa such as amoeba can provide the nutritional requirements for Legionella [200]. Legionella can also form biofilms and exploit other microorganisms as another tool to meet their required materials for growth and survival in the environment [201]. Previous studies have demonstrated that environmental factors such as temperature and nutrients also influence the virulence of Legionella. For example, the sensitivity of L. pneumophila to sodium differs between virulent and avirulent strains and also among cells at different growth phases [202]. However, the knowledge about L. longbeachae virulence is in its infancy and further studies are required to be carried out to fill this gap [17;
203].
Similar to many other aspects of Legionella biology, L. pneumophila has been the main target of studies regarding the nutritional requirements. For example, the role of iron in L. pneumophila growth and the acquisitions mechanism has been studied extensively [204-206].
Iron has been identified as a required micronutrient for L. pneumophila growth both in vitro and in vivo [207]. However, it was shown that if the iron concentration exceeds 1 mg/L, higher concentrations are toxic to these bacteria [208]. The presence of elements such as cobalt,
122
manganese and zinc has also been found to be related to the persistence of L. pneumophila in water systems [196; 209]. On the contrary, copper has been shown to have inhibitory effects on L. pneumophila growth and its biofilm formation [210; 211].
A better understanding of the chemical and physical conditions for the survival and growth of
Legionella would allow risk factors for infection to be identified and be useful for the epidemiological investigations of outbreaks of Legionnaires’ disease. As L. longbeachae has been identified in compost and potting mix products, the basic environmental parameters of temperature, pH and elemental profile of different types of samples including pine bark, sawdust, peat and green waste were studied. The findings can contribute towards on understanding of the relationship between the above-mentioned parameters and the presence and survival of Legionella in the potting mix products.
The aim of this chapter was to monitor and measure several important physical and chemical parameters of the collected samples. The relationship of values of these parameters in L. longbeachae qPCR positive and negative samples was then determined.
8.2. Materials and Methods
8.2.1. Recording temperature of growing media samples
Temperature recordings were conducted on-site for all three locations. The temperature of samples was measured using a non-contact infrared thermometer with dual laser targeting
(Digitech QM7221, Australia). Samples were taken from the surface and up to 30 cm deep from the piles or from the bagged products. For site 1, due to inclement weather with heavy rain during the sampling, temperature measurements were recorded for two grades of pine bark and sawdust only.
123
8.2.2. Measurement of pH of samples pH value of samples was measured by applying a pen style pH meter (SX610; Sanxin,
Shanghai, China). To determine the pH of solid material, 0.5 g of ground samples were added to 10 mL vial with 5 mL of ultrapure water. Samples were left at room temperature (21°C) for
4 h and shaken gently occasionally. After the solid material had settled the pH meter was placed into the liquid and pH was measured.
8.2.3. Generating the elemental profile of raw materials used for potting mix products
To determine the elemental profile of the samples, a set of 95 samples were selected using randomised computer generated sample numbers based on Bernoulli distribution. These samples included pine bark (both PCR positive and negative for L. longbeachae), and other sample types including sawdust, peat and green waste which were all qPCR negative for L. longbeachae (Table 8.1).
124
Table 8.1: Details of samples randomly selected for ICP elemental analysis
Sample type Number Bark (from Pinus radiata) negative PCR for L. longbeachae 20 Bark (from Pinus radiata) positive PCR for L. longbeachae 15 Green waste 20 Peat 20 Sawdust 20 Total 95
The elemental measurement was conducted by the scientists at the department of soil and physical sciences, Lincoln University, Canterbury, New Zealand. In order to generate the elemental profile of solid samples, Inductively Coupled Plasma Optical Emission Spectrometry
(ICP-OES) was employed. Using the collected data, the measurements were statistically analysed and modelled at Otago University Christchurch. The concentration of the following
17 elements were measured for this experiment: aluminium (Al), arsenic (As), boron (B), calcium (Ca), cadmium (Cd), chromium (Cr), copper (Cu), iron (Fe), potassium (K), magnesium (Mg), manganese (Mn), sodium (Na), nickel (Ni), phosphorus (P), lead (Pb), sulphur (S) and zink (Zn).
In order to prepare samples for this analysis, samples were digested by a microwave-assisted dissolution method using a CEM Mars Express microwave (CEM Corporation, Matthews, NC,
USA). Portions of the dried, ground and mixed samples were weighed (0.2 g) into a Teflon
PFA® and Kevlar shielded vessel. Samples were then dissolved in 2 mL of trace element grade concentrated nitric acid (69%, w/w, Merck, Germany) and 2 mL of hydrogen peroxide (30%).
Samples were next heated through a heating programme as follows: ramp to 90oC over 15 mins; hold for 5 min; ramp to 185oC over 10 min; hold for 15 min. The solution obtained was analysed by Varian 720 ICP-OES according to the available guideline (Nölte, 2003). Each analysis was carried out three times. For the quality control, a NIST library standard reference
125
material and also a tomato leaf sample were examined together with the tested samples as a measure of reliability.
8.2.4. Data analysis
Statistical analysis was performed by using ANOVA and Tukey Post Hoc test were used to examine the difference between groups, and multivariate analysis of variance (MANOVA) test for ICP analysis results. Multiple comparisons were performed to conduct the pairwise comparisons for each of the sample types for each of the analysed elements. This provided us with the mean difference (with 95% CI) in the analysed element concentrations among the samples. Tukey HSD was also used as the Post Hoc test at P = 0.001 to compare different groups. SPSS software version 25.0 (IBM SPSS, Chicago, USA) was used for this work.
Canonical discriminant analysis was also used to generate a plot showing the difference of elemental profile of the tested groups of samples. The alpha level of 0.05 was used as a significance criterion of P-values for these experiments.
Comparison of ICP elemental profiles between bark samples with positive and negative PCR for L. longbeachae was conducted using non-parametric Mann Whitney U-test, with null hypothesis (H=0) stipulating that these two independent groups of bark samples have the same distribution of the measured elements. pH values were first transformed to H+ ion concentrations according to the pH = -log [H+] formula prior to the analysis.
8.3. Results
8.3.1. Evaluation of temperature of the growing media samples
Statistical descriptions of recorded temperatures of total collected samples from three sites is shown is Table 8.2.
126
Table 8.2: Statistical description of samples temperatures
Sample type Mean (°C) N SD Min (°C) Max (°C)
Bark 22.67 283 10.05 4.5 72 Compost 17.91 83 6.9 10 36.8 Green waste 22.51 100 3.58 9.2 29.9 Peat 22.73 95 2.65 14.3 28.5 Potting mix 17 156 8.6 8 48.3 Pumice 15.12 10 0.81 14.2 16.5 Sawdust 21.33 167 9.67 4.1 49
ANOVA analysis showed that there is a significant difference of temperature between different sample types, [F (6,887) = 11.88, P < 0.001].
Among all sample types, peat (M=22.73, SD=2.65) and bark (M=22.67, SD=10.05) had the highest mean temperatures, which was significantly different from potting mix (P < 0.001).
Pumice had the lowest mean temperature (M=15.12, SD=0.81). However, the number of collected pumice samples was low. ANOVA analysis of temperature for samples collected from different sites showed that there was also a significant difference between sites [F (2, 891)
= 11.36, p = 0.00001]. Statistical descriptions are shown in Table 8.3.
Table 8.3: Mean temperature of collected samples in three potting mix production sites
N Mean (°C) Std. Deviation Site 1 136 24.10 12.48 Site 2 245 20.20 10.00 Site 3 513 20.39 6.11
The mean temperature of samples from site 1 (M=24.10, SD=12.48) was significantly higher than the other two (P < 0.001). ANOVA analysis for temperature among all samples collected
127
from three sites categorised for PCR positivity showed that temperature was significantly different between the groups [F(1,892) = 6.27, P = 0.012] and PCR positive samples had higher temperature (M=23.86, SD=8.61) than PCR negative samples (M=20.73, SD=8.11). The comparison of temperature for samples PCR positive for L. longbeachae and PCR negative samples was performed based on each sample type as shown in Table 8.4.
Table 8.4: Comparison of temperature between samples PCR positive and negative for L. longbeachae
qPCR positive (°C) qPCR negative (°C) P-value Bark 24.79 ± 15.73† 22.48 ± 9.41 0.292 Compost (bagged) 11.74 ± 1.14 18.49 ± 6.94 0.013 Potting mix 24.46 ± 20.64 16.86 ± 8.29 0.130 Sawdust 27.49 ± 13.00 20.64 ± 9.01 0.005 †mean ± standard deviation
The result showed that temperature was significantly higher in PCR positive sawdust samples in comparison to PCR negative samples (P = 0.005), while PCR positive bagged compost samples had significantly lower temperatures in comparison with PCR negative samples (P =
0.013). There was no significant difference in the temperature between PCR positive and PCR negative samples among bark and potting mix samples.
128
8.3.2. Comparison of pH value of various sample types
Statistical descriptive for pH measurement of total sample types is shown in Table 8.5.
Table 8.5: Summary of the descriptive features for pH of the overall sample types.
Sample type N Mean pH SD Min pH Max pH Bark 294 5.23 0.69 2.8 8 Compost 83 7.47 0.74 5.4 9.5 Green waste 100 5.85 0.44 4.7 7 Potting mix 159 6.8 1.02 4.1 9 Pumice 10 7.13 0.41 6.5 7.7 Sawdust 167 5.71 0.62 3 8.1
The results showed that pH was significantly higher in PCR positive sawdust in comparison with the PCR negative sawdust samples (P = 0.001). There was a significant difference of pH between PCR positive and PCR negative sawdust samples, while no statistically significant difference was found among other sample types (Table 8.6).
Table 8.6: Mean pH values of different sample types PCR positive and negative for L. longbeachae
Sample type pH value pH value P-value (qPCR positive) (qPCR negative) Bark 5.25 ± 0.42 5.23 ± 0.72 0.871 Sawdust 6.18 ± 0.60 5.66 ± 0.61 0.001* Potting mix 6.55 ± 1.34 6.81 ± 1.01 0.542 Bagged compost 7.34 ± 0.47 7.48 ± 0.76 0.631 * Significant difference between the two groups at P = 0.001 level
129
8.3.3. ICP-OES analysis
The results of ICP-OES elemental analysis are summarised as Table 8.7. Canonical discrimination analysis showed the differences and similarities of elemental profiles among various sample types (Figure 8.1). This canonical plot demonstrates that sample types fall into three distinct non-overlapping clusters. Green waste had the highest concentrations of most elements and was distinct from all the other sample types. Peat had the lowest elemental content except for its sulphur content.
Figure 8.1: Canonical discriminant analysis plot.
This graph shows the degree of similarity and differences of various sample types samples including bark PCR positive and negative, sawdust, peat and green waste analysed by ICP-OES method. The function coefficients (the proportion of within-groups’ sums of squares to the total sums of squares) were generated in order to evaluate the significance level of difference between the groups by using the Wilks’ Lambda distribution factor.
130
Table 8.7 shows the concentration of measured elements as milligram per kilogram (mg/kg) of dried weight for each sample. MANOVA analysis showed that there were significant differences of elemental content among sample types. This comparison showed that concentrations of 15 elements were substantially higher in green waste than the other sample types and only two elements, magnesium (3021.16 mg/kg) and nickel (8.28 mg/kg) were lower than found in other sample types.
Peat had the lowest concentration of most elements. This difference was statistically significant from other sample types for only Iron (Fe) and Manganese (Mn). However, peat had the highest concentrations of sulphur concentration with a significant difference from the other sample types (P < 0.001).
Bark and sawdust had similar elemental profiles. However, sawdust had a significantly lower concentration of sodium (P = 0.01) in comparison with PCR positive bark samples and zinc (P
= 0.05) in comparison with PCR positive and negative bark samples. Sawdust had the highest concentration of magnesium among all sample types as well. Bark samples had the lowest concentration of sulphur, which was significantly different from the other tested samples.
Comparison of bark samples that were positive and negative PCR for L. longbeachae was conducted using non-parametric Mann Whitney U-test, showed that concentration of boron (P
= 0.036) and sulphur (P = 0.017) were significantly higher in PCR negative samples than PCR positive samples.
131
Table 8.7: Mean content of trace elements in samples determined by ICP/OES analysis.
Bark negative Bark positive Sawdust Green waste Peat Al 4060.98 3816.07 1325.4 23901.79 1906.82 (2817.57) (2889.44) (772.50) (3105.01) (381.51) As 0.93 1.01 0.88 11.35 0.31 (0.74) (0.84) (0.29) (2.26) (0.21) B 10.26 7.27* 7.20 22.63 5.02 (4.26) (3.11) (3.16) (2.26) (0.69) Ca 5718.42 3811.87 6417 54857 2779.3 (3897.81) (1000.00) (8112.32) (9787.80) (882.00) Cd 0.05 0.03 0.03 0.27 0.04 (0.04) (0.02) (0.01) (0.07) (0.03) Cr 10.92 10.76 6.20 23.65 1.76 (7.18) (12.13) (2.65) (22.78) (0.96) Cu 5.83 5.78 3.78 47.13 2.02 (2.83) (3.20) (1.14) (5.30) (0.60) Fe 5605.78 4574.65 5398.71 11338.66 1384.97 (4493.26) (3866.27) (3397.91) (1609.34) (487.74) K 1760.7 2125.53 940.75 5446.8 388.52 (594.62) (626.30) (397.91) (1550.38) (88.08) Mg 9943.6 6139.11 12848.41 3021.16 2881.05 (13044.19) (7657.82) (12225.10) (335.28) (978.85) Mn 180.36 153.76 211.53 690.39 23.14 (106.03) (71.66) (86.22) (105.98) (11.67) Na 439.3 534.91 226.37 1467.2 600.45 (140.78) (141.98) (153.8) (531.65) (107.09) Ni 46.95 29.63 64.32 8.28 3.13 (60.69) (41.27) (73.00) (14.81) (4.36) P 252.92 256.42 165.45 8565.91 295.09 (105.61) (97.83) (79.33) (1624.73) (46.40) Pb 2.05 1.69 0.89 24.91 0.95 (1.89) (1.83) (0.82) (9.85) (0.74) S 613.22 347.29 * 988.15 3108.53 4726.22* (342.78) (59.31) (521.65) (261.65) (260.54) Zn 32.87 33.89 16.79 198.19 9.74 (13.74) (11.27) (7.01) (25.48) (3.55) * Significant difference (P < 0.05) of an elemental concentration in a given sample type from the other samples types
Measurements are based on mg/kg. Standard deviations are shown in brackets. N = 20 with the exception for PCR positive bark sample (N = 15).
8.3.4. Meteorological parameters
The information about the weather condition in three sampling sites was collected as described in Table 6.2 in Chapter 6.
132
8.4. Discussion
This chapter focused on physicochemical factors of temperature, pH and elemental profile that play crucial role for Legionella growth in the environment. The mean temperature of various sample types was different between sites; significantly higher in site 1 (lower South Island) than the other two potting mix production sites. Interestingly it was raining heavily in site 1 on sampling day and samples were exposed to the weather, but at site 3, the feedstock was stored indoor. Peat, sawdust and green waste had the highest temperatures among the sample types.
Sawdust and potting mix samples PCR positive for L. longbeachae had a higher mean temperature than PCR negative samples. High temperatures in composting materials may contribute to this finding.
In the waste management industry, the temperatures may increase to over 70oC as a result of metabolic activities of biodegradation as part of the composting process of plant materials. This results in dramatic changes in diversity and make-up of the microbial populations in favour of those with better activity and tolerance of higher temperatures in the decomposing materials
[212]. Reduced-oxygen scavenging enzyme such as superoxide dismutase and catalase are found among Legionella species and L. longbeachae has a high catalase activity [213].
Possessing this enzymatic suite provides protection from harmful effects of reactive oxygen species (ROS) during metabolic activities. Survival of Legionella in water has been well studied and it is found that they can grow in a wide range from 0oC to 63oC, with 30-42oC range as the permissive range of their multiplication [214; 215]. In contrast, most environmental organisms do not grow at temperatures over 37oC and tolerating extreme conditions requires metabolic adjustments.
Comparison of pH among sample types showed that it was different between sample types.
Compost had the highest mean pH (7.47) and bark had the lowest mean pH (5.2) among all sample types. However, there was no difference of mean pH between sites. The only significant
133
difference in pH was between PCR positive (pH = 6.18) and PCR negative (pH = 5.66) sawdust samples. pH is a vital factor for the optimal growth of microorganisms. Previous studies have shown that Legionella can live in an acidic pH when encountering harsh environmental conditions. These cells can also survive the acidic conditions inside human alveolar macrophages [216]. This acidic condition can also reduce the growth of other organisms in favour of Legionella. The relation between the pH of plant materials used in potting mix production and presence of Legionella should be further studied in future work. In Chapter 5, it was demonstrated that a large proportion of PCR positive results belonged to bark and bark- containing samples. Various tree species have specific bark chemical makeup including electrical conductivity, ash content and pH. The pH has been shown to be lower in bark from coniferous tree species, when compared to trees of other species. For example, Scot pine bark has a measured pH range of 2.92 – 3.73 [217]. This may favour the persistence and growth of
Legionella in the environment.
One important factor that helps Legionella survive the harsh conditions found in composting materials and production of potting mix products is the presence of environmental amoeba and ciliates such as Acanthamoeba, Tetrahymena, Naegleria, Echinamoeba, and Cyclidium [15;
218; 219]. Legionella may be engulfed by the amoeba and live and even multiply inside the amoebic hosts. This can be beneficial for these bacteria to protect them against drying and heating which are part of composting process. Intra-protozoan life of Legionella can provide abundant nutrients from their host, while the environmental nutritional condition is impoverished for them [220]. When amoeba encyst due to adverse environmental conditions,
Legionella inside them may persist, probably by changing to a dormant state. It has been shown that the interaction of Legionella bacteria with their host amoeba increases at higher temperatures. For example, interactions of L. pneumophila with its Acanthamoeba host was
134
greater at 35oC than at 20oC [221]. Further research needs to be conducted to expand the knowledge about this intracellular life of L. longbeachae inside its environmental hosts.
The elemental profile of these four types of raw materials was found to vary extensively. Green waste was dramatically different from the other sample types. Peat had differences particularly with the highest sulphur content. This was in accord with previous studies. For example, in a study conducted in Switzerland, measuring sulphur concentration in peat bog showed that it had a range of 1 to 5 grams per kilogram of the samples [222].
Bark and sawdust had similar elemental profiles, mainly because of their common source, namely Pinus radiata trees. It was also found that PCR negative bark samples had greater concentration of boron and sulphur than PCR positive bark samples.
ICP-OES elemental measurement revealed interesting results that may be important for survival of Legionella. First, green waste had a highest concentration of most of the measured elements in comparison with the other tested sample types. One or more of these high concentrations might be toxic for Legionella, inhibiting their growth and be the reason for negative PCR results of all green waste samples. High concentrations of heavy metals with positive charge, such as copper and silver can inhibit the growth of bacterial cells and even kill them by attaching to cell walls with that have a negative charge [223]. However, it is important to consider that although a concentration of 0.2-0.4 mg/L of copper is recommended to remove
L. pneumophila from water systems, this is the ionic form of copper, and concentrations of free elements in the tested samples are unknown. Alternatively, the reason for a PCR negative results could be because no Legionella was present in the tested samples. Arsenic is another element that can inhibit enzymatic reactions in biological systems through degradation of proteins or active sites of enzymes [224]. However, there are some studies reporting the
135
presence of bacteria with ability to metabolise arsenic with the potential application of bioremediation of arsenic in contaminated soils [225].
Interestingly, the concentration of two elements magnesium and nickel were higher in bark and sawdust samples than green waste, and these are both from P. radiata source. Both of these elements have important roles for plant cells such as enzymatic activities and cell growth. For example, nickel is an important part of the active site of urease enzyme which is crucial to lyse urea as this is toxic to plant cells [226; 227].
All these trace elements were present in all tested samples types. Therefore, deficiency of the elements in unlikely to explain the difference of PCR results. However, the low concentration of iron in peat might contribute to negative PCR results for this sample type. Previous studies have focused on metal requirements of L. pneumophila [196]. It was shown that Legionella growth was promoted by addition of low concentrations of calcium, iron, magnesium, and zinc elements to the medium. In another study, it was found that low concentrations of elements such as calcium, potassium, iron, zinc, magnesium, manganese and copper can increase the growth of Legionella [194]. Calcium and magnesium are two important elements that contribute to the enzymatic activities and bacterial structures such as cell wall. Iron plays a vital role in biological activities such as oxidation-reduction complexes and as enzyme cofactors. The nutritional requirement of Legionella also contributes to their intracellular growth inside environmental protozoa or human alveolar macrophages. For example, when more iron was added to the culture of the ciliate Dictyostelium discoideum, uptake of L. pneumophila was increased [228]. Interestingly, smoking is considered a risk factor for
Legionnaires’ disease. It has been shown that smoking results in iron overload in the lungs which can provide the required iron for Legionella to grow inside alveolar macrophage as well
[205].
136
Boron and sulphur were the only elements with a significantly higher concentrations in PCR negative bark samples than PCR positive bark samples. Boron has been shown to be involved in the structure of a furanosyl borate ester autoinducer (AI-2) that contributes to reporting the environmental changes and quorum sensing [229].
Although some bacteria are resistant to high concentrations of boron and can accumulate it, other bacteria are sensitive to its toxicity due to permeability of boron into lipid bilayers [230].
There is no published information of inhibitory effects of boron on Legionella. Future studies need to be conducted to determine the possible effects. The inhibitory effect of boron on
Legionella growth is a possible explanation for this difference. Sulphur is an important element for biosynthesis of nucleic acids, proteins and amino acids such as cysteine and methionine
[231; 232].
The bioavailability of these elements needs to be considered besides the biological content of these elements to determine the optimal and toxic levels of these elements for Legionella. For instance, copper availability depends on the pH of the sample, and has an inverse relationship with pH value, as a change of pH from 6.5 to 7.4 decreases the availability of this element from
4 mg/L to 1.3 mg/L [233]. Therefore, the combination of environmental factors found in various sample types may contribute to the presence of Legionella. Therefore, further studies are required to have a better understanding of the role of these elements on the growth and survival of L. longbeachae in different conditions.
In conclusion, the physico-chemical factors such as temperature, pH, and trace elements may contribute to the persistence and growth of Legionella in potting mix products. However, this needs to be further investigated by examining the effect of individual factors alone or in combination on bacterial growth. This approach may help us find better ways to control these bacteria in potting mix products.
137
CHAPTER 9
General Conclusions and Future Perspectives
9. General Conclusions and Future Perspectives
9.1. General discussion
Legionnaires’ disease is a bacterial infection with global distribution is caused by several human pathogen species of the Legionella genus. L. longbeachae is mainly found in composted plant materials and is the predominant species with an increasing incidence in New Zealand
[4]. Previous studies have suggested a link between gardening activities and exposure to potting mix products with contracting the disease [45; 234]. One hypothesis of this study was that L. longbeachae are present among feedstock and final products in potting mix production sites.
The aim of this study was to improve the detection methods for L. longbeachae. Therefore, several methods were examined to find an improved detection and isolation method for L. longbeachae in clinical and environmental samples.
To improve the isolation methods for L. longbeachae for more efficient culture and PCR, an immunomagnetic separation method was developed (Chapter 3). This included the production of a highly specific and sensitive antibody. A polyclonal antibody against L. longbeachae was raised in rabbits by provoking their immune system using heat-killed L. longbeachae antigens.
After several booster injections, serum containing polyclonal antibody against L. longbeachae was separated and purified using ion-exchange chromatography.
ELISA assays showed that this antibody has a high affinity for serogroup 1 of L. longbeachae, with a lower affinity for serogroup 2. Therefore, it can be used for isolation of both serogroups of L. longbeachae. Cross-reactivity test with several other microorganisms including other species of Legionella genus and several ubiquitous environmental microorganisms showed that this polyclonal antibody was highly specific for L. longbeachae and had no cross-reaction with them. This is of great importance as samples often contain many co-existing microorganisms.
Although the assessed non-Legionella organisms were not closely related to L. longbeachae, a wider range of microorganisms including those more closely related with possible cross-
139
reactivity were not available for this project. Testing a wider range of organisms including those more closely related to L. longbeachae is something that should be considered for future work.
This polyclonal antibody was coupled with 2.8 µm magnetic beads covered with toluensulfonyl groups which results in strong binding of the antibody to the beads through covalent bonds.
This magnetic bead is recommended by the manufacturer for isolation of small cells from complex matrices. The coupled beads with anti-L. longbeachae polyclonal antibody were applied prior to culture and also to DNA extraction, which was followed by real-time PCR.
The results confirmed that the IMS was successful in capturing L. longbeachae from samples for culture or PCR.
The IMS method was designed to isolate L. longbeachae from both clinical and environmental samples with complex matrices. To validate the IMS method, it was tested in clinical setting combined with both culture and qPCR methods. Sputum samples obtained from pneumonia patients over a year that were stored at -80°C was tested by IMS method (Chapter 4). The results showed that IMS-culture was a promising method with significantly improved culture sensitivity (49%) in comparison with culture results after acid treatment of fresh (31%) and stored (18%) respiratory specimens. In a study by Murdoch and colleagues, only 15% of the
PCR positive samples, were culture positive [5].
The sensitivity of frozen specimens to acidic buffer and the presence of contaminating microorganisms in sputum may contribute to inhibitory effects on recovery of Legionella bacteria. Applying the IMS method is a good strategy to improve Legionella culture by protecting these bacteria from exposure to acidic buffer and competing microorganisms. Using the coupled bead and antibody may help isolation of bacterial cells without damaging their sensitive structural integrity after storage. Specific immunocapture of L. longbeachae can
140
provide a better growth condition for Legionella on culture by removing the contaminating microorganisms and reducing the presence of inhibitory molecules present in the samples during washing steps. It is possible that the IMS method may also facilitate shift of Legionella cells from VBNC into vegetative form [235].
Acid pre-treatment of samples to isolate Legionella bacteria has been developed based on the fact that they can survive in acidic environments, such as the human macrophage. However, prolonging their presence in low pH such as 2.2 might damage their cell wall [65]. Further studies need to be conducted to explain the pH and time range these bacteria can tolerate exposure to such harsh conditions.
The findings of this study suggest that there may be a link between storage and loss of viability of Legionella bacteria. This has been observed with other bacteria [168]. However, the possible effects of storage on viability of L. longbeachae and the reasons behind it needs to be investigated. Conducting prospective studies on IMS using fresh specimens in the future, will also elucidate the difference with retrospective studies using cryopreserved samples.
IMS-PCR results in this study confirmed that lower Ct values were associated with a higher probability of culture positivity for Legionella in specimens. A similar pattern was observed when assessing the severity of pneumonia in patients based on CURB-65 scores [5]. Those with more severe scores had higher load of Legionella DNA (lower Ct values) in real-time
PCR.
Volatile organic compounds (VOCs) are metabolites released from all organisms and some can be used as biomarkers to identify microorganisms. This can be used as a non-invasive method such as a breath test on pneumonia patients as collecting the respiratory specimen is a problem.
This can assist diagnosis of Legionnaires’ diseases without doing invasive testing on the patient in a short time period compared with traditional culture and PCR methods that are time-
141
consuming. To achieve this aim, 100 pneumonia patients were recruited prospectively, and their breath samples were collected and analysed by SPME/GC-MS method to study the volatile profile of their breath samples (Chapter 5). The metabolic profile of L. longbeachae was also studied alongside by culturing these bacteria in broth culture and collecting the released gases in the culture headspace analysed by GC/MS method. A peak of interest was found in VOC profile of the pure culture. Due to the technical breakdown of the GC/MS machine, it was not possible to search for this peak in the collected breath samples.
To examine potting mix products as the source of L. longbeachae, an environmental investigation was conducted by sampling raw materials and bagged final products at three different production sites nationwide (Chapter 6). PCR results showed that L. longbeachae were present in 5.1% of the total samples collected. Site 1 had the highest ratio of PCR positivity (14.3%) in comparison with the site 2 (2.55%), and site 3 (1.4%). Several factors may contribute to these differences, including source of the raw materials, composting approaches at different production sites, weather conditions and also storage conditions of the tested material. The study findings showed that pine bark and bark-containing products had the highest proportion of PCR positive samples for L. longbeachae which was comparable to the findings of previous studies, suggesting bark-containing potting mix products as an environmental source of L. longbeachae [46]. There was no PCR positive sample among all tested green waste and peat samples. These findings suggest that pine trees may be a significant reservoir of L. longbeachae and pine bark is an important ongoing source of these organism into the manufacturing process of potting mix. This does not mean that other sources of L. longbeachae do not contribute to the presence of L. longbeachae in potting mix products.
PCR results showed a limitation of environmental testing for L. longbeachae as well. The acquired Ct values were higher than of those for clinical samples. This may be due to the low number of Legionella persisting in the environment. The low dose of these bacteria might be
142
enough to infect human through aerosols and multiply in favourable conditions. It is possible that low number of bacterial cells can be under the limit of detection for PCR. The IMS method can be used in future work to capture very low number of Legionella cells.
Real-time PCR does not differentiate live bacterial cells from dead cells either. Using a commercially available fluorescence dyes such as PMA and establishing appropriate methods to perform PCR with ability to evaluate the live/dead ratio of the Legionella cells will provide valuable information [236].
Isolating Legionella from these PCR positive potting mix products was not successful (Chapter
7). Culture included the conventional acid pre-treatment prior to inoculation of the treated samples onto GVPC selective media. It was obvious that the number and diversity of co- existing microorganisms in the potting mix product samples was greater than those grown from sputum samples. This shows the difficulty of culturing L. longbeachae from environmental samples. Another possible reason for the negative culture results from PCR positive samples might be the very low number of Legionella in the tested samples which are approximately less than 102 CFU/g. Furthermore, acid pre-treatment was developed for the water samples and L. pneumophila [109]. New methods for isolating and growing L. longbeachae from environmental samples are needed. To improve the isolation of these bacteria from potting mix products, several new methods were tested using potting mix samples spiked with L. longbeachae (Chapter 8). These methods included direct and indirect IMS, sucrose gradient centrifugation, GVPC decontamination and combined GVPC-IMS method. Culture results showed that GVPC decontamination followed by IMS separation, resulted in the highest recovery rate among the tested methods. GVPC decontaminated helped the reduction of contaminating microorganisms of the plates, while IMS captured L. longbeachae bacteria from the treated samples. Determining limit of detection for different sample types, developing
143
improved concentration methods, and using GVPC-IMS method for the freshly collected environmental samples may improve isolation of L. longbeachae from such samples.
As the PCR results were varying for different sample types, the role of physico-chemical parameters such as such as temperature, pH and elemental profiles on presence of Legionella was studied. The results of this analysis are discussed comprehensively in Chapter 8. These results suggest that physico-chemical parameters should be considered for their influence on persistence and survival of Legionella in the environment. The chemical analysis of pine bark may give some leads for reducing the load of L. longbeachae in potting mix. Both sulphur and boron were associated with lower PCR positivity rates in the bark samples and could be additives in the manufacturing process of potting mix. The results of this study are limited as we did not know whether the elements were in a free or bound form, and further studies are needed to determine whether adding the elements will be of benefit.
As pine bark had a considerable PCR positive samples for L. longbeachae, a pilot study was carried out alongside testing pine tree and several other types of trees for the presence of L. longbeachae. I contributed to this by advising the research assistants on the methods of processing the bark samples and performing the qPCR results. Preliminary results showed that there is a significant change in the number of qPCR positive over the year of samples. There was a much higher number of PCR positive pine trees in spring compared with winter and lower mean Ct values in the positive samples. The ratio of trees with at least one positive sample was: January (6%); April (2%); July (12%); October (54%); (P < 0.001). Further studies are required to produce a more comprehensive understanding of this finding in future work.
144
9.2. Future perspectives
To improve the efficiency of IMS method for isolation of L. longbeachae, monoclonal antibodies could be developed. Our in-house polyclonal antibody had a high level of specificity and sensitivity, but producing such a high-quality polyclonal antibody is difficult, has limited availability, and depends on many factors such as the host animal as well.
The results of IMS testing on sputum samples suggested that storage can influence the culturability of Legionella cells and it is suggested to use IMS method in future as a prospective study to understand the true value of using this method. Co-culture with type strain and wild- type amoeba and other protozoan cells is another avenue than should be explored in future studies to improve culture results [237; 238]. The intracellular life of Legionella can benefit this method and yields more efficient culture results. Co-culture with macrophage cell lines is another option for rapid isolation of Legionella bacteria that can be examined in future studies
[239]. As literature is poor regarding the co-culture of L. longbeachae, and most laboratories suffer from the lack of a well-established method to isolate L. longbeachae from environmental samples, it is valuable to establish a model in order to provide a comprehensive and reliable isolation method. It is also suggested that the combined GVPC-IMS method could be compared concurrently with current methods to determine its value for isolation of L. longbeachae and other Legionella species from on the freshly collected environmental samples.
As a new GC/MS machine is purchased, it is planned to repeat the volatile analysis for L. longbeachae using this device to study whether compounds found previously can be discovered with similar retention times can be found in the mass spectrometry results. The candidate compounds can be further studied by searching through NIST library to discover further about their structure and identity.
145
References
1. Sabria, M., Alvarez, J., Dominguez, A., Pedrol, A., Sauca, G., Salleras, L., Lopez, A.,
Garcia-Nunez, M., Parron, I., & Barrufet, M. (2006). A community outbreak of
Legionnaires' disease: evidence of a cooling tower as the source. Clinical Microbiology
and Infection, 12(7), 642-647.
2. Smith, A. D., Cowan, J. O., Brassett, K. P., Herbison, G. P., & Taylor, D. R. (2005).
Use of exhaled nitric oxide measurements to guide treatment in chronic asthma. N Engl
J Med, 352(21), 2163-2173. doi:10.1056/NEJMoa043596
3. den Boer, J. W., Yzerman, E. P., Jansen, R., Bruin, J. P., Verhoef, L. P., Neve, G., &
van der Zwaluw, K. (2007). Legionnaires' disease and gardening. Clin Microbiol
Infect, 13(1), 88-91. doi:10.1111/j.1469-0691.2006.01562.x
4. Priest, P. C., Slow, S., Chambers, S. T., Cameron, C. M., Balm, M. N., Beale, M. W.,
Blackmore, T. K., Burns, A. D., Drinkovic, D., Elvy, J. A., Everts, R. J., Hammer, D.
A., Huggan, P. J., Mansell, C. J., Raeder, V. M., Roberts, S. A., Robinson, M. C.,
Sathyendran, V., Taylor, S. L., Thompson, A. W., Ussher, J. E., van der Linden, A. J.,
Williams, M. J., Podmore, R. G., Anderson, T. P., Barratt, K., Mitchell, J. L., Harte, D.
J., Hope, V. T., & Murdoch, D. R. (2019). The burden of Legionnaires' disease in New
Zealand (LegiNZ): a national surveillance study. Lancet Infect Dis, 19(7), 770-777.
doi:10.1016/s1473-3099(19)30113-6
5. Murdoch, D. R., Podmore, R. G., Anderson, T. P., Barratt, K., Maze, M. J., French, K.
E., Young, S. A., Chambers, S. T., & Werno, A. M. (2013). Impact of routine systematic
polymerase chain reaction testing on case finding for Legionnaires' disease: a pre-post
comparison study. Clin Infect Dis, 57(9), 1275-1281. doi:10.1093/cid/cit504
146
6. Fields, B. S., Benson, R. F., & Besser, R. E. (2002). Legionella and Legionnaires'
disease: 25 years of investigation. Clin Microbiol Rev, 15(3), 506-526.
doi:10.1128/cmr.15.3.506-526.2002
7. Ott, M., Messner, P., Heesemann, J., Marre, R., & Hacker, J. (1991). Temperature-
dependent expression of flagella in Legionella. J Gen Microbiol, 137(8), 1955-1961.
doi:10.1099/00221287-137-8-1955
8. Muder, R. R., & Yu, V. L. (2002). Infection due to Legionella species other than L.
pneumophila. Clin Infect Dis, 35(8), 990-998. doi:10.1086/342884
9. Reynaldos Canton Migotto, T., Gyorik Lora, S., Gaia, V., & Pagnamenta, A. (2014).
ARDS with septic shock due to Legionella longbeachae pneumonia in a patient with
polymyalgia rheumatica. Heart Lung Vessel, 6(2), 114-118.
10. Fliermans, C. B., Cherry, W. B., Orrison, L. H., Smith, S. J., Tison, D. L., & Pope, D.
H. (1981). Ecological distribution of Legionella pneumophila. Appl Environ Microbiol,
41(1), 9-16.
11. Harf, C., & Monteil, H. (1988). Interactions between free-living amoebae and
Legionella in the environment. Water Science and Technology, 20(11-12), 235-239.
12. Hilbi, H., Hoffmann, C., & Harrison, C. F. (2011). Legionella spp. outdoors:
colonization, communication and persistence. Environmental microbiology reports,
3(3), 286-296.
13. Schunder, E., Gillmaier, N., Kutzner, E., Herrmann, V., Lautner, M., Heuner, K., &
Eisenreich, W. (2014). Amino acid uptake and metabolism of Legionella pneumophila
hosted by Acanthamoeba castellanii. Journal of Biological Chemistry, 289(30), 21040-
21054.
14. Brown, M. R., & Barker, J. (1999). Unexplored reservoirs of pathogenic bacteria:
protozoa and biofilms. Trends in microbiology, 7(1), 46-50.
147
15. Steele, T., & McLennan, A. (1996). Infection of Tetrahymena pyriformis by Legionella
longbeachae and other Legionella species found in potting mixes. Appl. Environ.
Microbiol., 62(3), 1081-1083.
16. Koide, M., Saito, A., Okazaki, M., Umeda, B., & Benson, R. F. (1999). Isolation of
Legionella longbeachae serogroup 1 from potting soils in Japan. Clinical infectious
diseases, 29(4), 943-944.
17. Kozak, N. A., Buss, M., Lucas, C. E., Frace, M., Govil, D., Travis, T., Olsen-
Rasmussen, M., Benson, R. F., & Fields, B. S. (2010). Virulence factors encoded by
Legionella longbeachae identified on the basis of the genome sequence analysis of
clinical isolate D-4968. Journal of bacteriology, 192(4), 1030-1044.
18. Phin, N., Parry-Ford, F., Harrison, T., Stagg, H. R., Zhang, N., Kumar, K., Lortholary,
O., Zumla, A., & Abubakar, I. (2014). Epidemiology and clinical management of
Legionnaires' disease. The Lancet infectious diseases, 14(10), 1011-1021.
19. Hilbi, H. (2014). Molecular Mechanisms in Legionella Pathogenesis (Vol. 376):
Springer.
20. WHO, Bartram, J., Chartier, Y., Lee, J., Pond, K., & Surman-Lee, S. Legionella and
the prevention of legionellosis. Geneva: World Health Organization (WHO); 2007. In.
21. Fraser, D. W., Tsai, T. R., Orenstein, W., Parkin, W. E., Beecham, H. J., Sharrar, R. G.,
Harris, J., Mallison, G. F., Martin, S. M., & McDade, J. E. (1977). Legionnaires'
disease: description of an epidemic of pneumonia. New England Journal of Medicine,
297(22), 1189-1197.
22. Hicks, L. A., Rose, C., Fields, B., Drees, M., Engel, J., Jenkins, P., Rouse, B., Blythe,
D., Khalifah, A., & Feikin, D. (2007). Increased rainfall is associated with increased
risk for legionellosis. Epidemiology & Infection, 135(5), 811-817.
148
23. De Vito, C., Marzuillo, C., D'Andrea, E., Romano, F., & Villari, P. (2011). Emerging
and re-emerging infectious diseases: tackling the challenge on a global level. Italian
Journal of Public Health, 8(1).
24. Reller, L. B., Weinstein, M. P., & Murdoch, D. R. (2003). Diagnosis of Legionella
infection. Clinical infectious diseases, 36(1), 64-69.
25. Storch, G., BAINE, W. B., FRASER, D. W., BROOME, C. V., CLEGG, H. W.,
COHEN, M. L., GOINGS, S. A., POLITI, B. D., TERRANOVA, W. A., & TSAI, T.
F. (1979). Sporadic community-acquired Legionnaires' disease in the United States: A
case-control study. Annals of Internal Medicine, 90(4), 596-600.
26. Nash, T., Libby, D., & Horwitz, M. (1984). Interaction between the legionnaires'
disease bacterium (Legionella pneumophila) and human alveolar macrophages.
Influence of antibody, lymphokines, and hydrocortisone. The Journal of clinical
investigation, 74(3), 771-782.
27. Tkatch, L. S., Kusne, S., Irish, W. D., Krystofiak, S., & Wing, E. (1998). Epidemiology
of legionella pneumonia and factors associated with legionella-related mortality at a
tertiary care center. Clinical infectious diseases, 27(6), 1479-1486.
28. Friedman, H., Yamamoto, Y., Newton, C., & Klein, T. (1998). Immunologic response
and pathophysiology of Legionella infection. Paper presented at the Seminars in
respiratory infections.
29. Cooper, A. J., Barnes, H. R., & Myers, E. R. (2004). Assessing risk of Legionella.
Ashrae Journal, 46(4), 22-27.
30. Lode, H., Kemmerich, B., Schäfer, H., Grothe, R., Hartmann, R., Ehret, W., &
Ruckdeschel, G. (1987). Significance of non-pneumophila Legionella species in adult
community-acquired and nosocomial pneumonias. Klinische Wochenschrift, 65(10),
463-468.
149
31. Blatt, S. P., Dolan, M. J., Hendrix, C. W., & Melcher, G. P. (1994). Legionnaires'
disease in human immunodeficiency virus-infected patients: eight cases and review.
Clinical infectious diseases, 18(2), 227-232.
32. Loeb, M., Simor, A. E., Mandell, L., Krueger, P., McArthur, M., James, M., Walter, S.,
Richardson, E., Lingley, M., & Stout, J. (1999). Two nursing home outbreaks of
respiratory infection with Legionella sainthelensi. Journal of the American Geriatrics
Society, 47(5), 547-552.
33. Doyle, R. M., Steele, T. W., McLennan, A. M., Parkinson, I. H., Manning, P. A., &
Heuzenroeder, M. W. (1998). Sequence analysis of the mip gene of the soilborne
pathogen Legionella longbeachae. Infection and immunity, 66(4), 1492-1499.
34. Graham, F., White, P., Harte, D., & Kingham, S. (2012). Changing epidemiological
trends of legionellosis in New Zealand, 1979–2009. Epidemiology & Infection, 140(8),
1481-1496.
35. White, P., Graham, F., Harte, D., Baker, M., Ambrose, C., & Humphrey, A. (2013).
Epidemiological investigation of a Legionnaires' disease outbreak in Christchurch,
New Zealand: the value of spatial methods for practical public health. Epidemiology
& Infection, 141(4), 789-799.
36. McKINNEY, R. M., PORSCHEN, R. K., EDELSTEIN, P. H., BISSETT, M. L.,
HARRIS, P. P., BONDELL, S. P., STEIGERWALT, A. G., WEAVER, R. E., EIN, M.
E., & LINDQUIST, D. S. (1981). Legionella longbeachae species nova, another
etiologic agent of human pneumonia. Annals of Internal Medicine, 94(6), 739-743.
37. Bibb, W., Sorg, R., Thomason, B., Hicklin, M., Steigerwalt, A., Brenner, D., & Wulf,
M. (1981). Recognition of a second serogroup of Legionella longbeachae. Journal of
clinical microbiology, 14(6), 674-677.
150
38. Lanser, J., Adams, M., Doyle, R., Sangster, N., & Steele, T. (1990). Genetic relatedness
of Legionella longbeachae isolates from human and environmental sources in
Australia. Appl Environ Microbiol, 56(9), 2784-2790.
39. O'connor, B., Carman, J., Eckert, K., Tucker, G., Givney, R., & Cameron, S. (2007).
Does using potting mix make you sick? Results from a Legionella longbeachae case-
control study in South Australia. Epidemiology & Infection, 135(1), 34-39.
40. Pravinkumar, S., Edwards, G., Lindsay, D., Redmond, S., Stirling, J., House, R., Kerr,
J., Anderson, E., Breen, D., & Blatchford, O. (2010). A cluster of Legionnaires' disease
caused by Legionella longbeachae linked to potting compost in Scotland, 2008-2009.
Eurosurveillance, 15(8), 19496.
41. Ruehlemann, S. A., & Crawford, G. R. (1996). Panic in the potting shed: the
association between Legionella longbeachae serogroup 1 and potting soils in
Australia. Medical journal of Australia, 164(1), 36-38.
42. Whiley, H., & Bentham, R. (2011). Legionella longbeachae and legionellosis.
Emerging infectious diseases, 17(4), 579.
43. Casati, S., Gioria‐ Martinoni, A., & Gaia, V. (2009). Commercial potting soils as an
alternative infection source of Legionella pneumophila and other Legionella species in
Switzerland. Clinical Microbiology and Infection, 15(6), 571-575.
44. Velonakis, E., Kiousi, I., Koutis, C., Papadogiannakis, E., Babatsikou, F., &
Vatopoulos, A. (2010). First isolation of Legionella species, including L. pneumophila
serogroup 1, in Greek potting soils: possible importance for public health. Clinical
Microbiology and Infection, 16(6), 763-765.
45. Currie, S. L., & Beattie, T. K. (2015). Compost and Legionella longbeachae: an
emerging infection? Perspectives in public health, 135(6), 309-315.
151
46. Steele, T. W., Lanser, J., & Sangster, N. (1990). Isolation of Legionella longbeachae
serogroup 1 from potting mixes. Appl. Environ. Microbiol., 56(1), 49-53.
47. Koide, M., Arakaki, N., & Saito, A. (2001). Distribution of Legionella longbeachae
and other legionellae in Japanese potting soils. Journal of Infection and Chemotherapy,
7(4), 224-227.
48. Montanaro-Punzengruber, J., Hicks, L., Meyer, W., & Gilbert, G. (1999). Australian
isolates of Legionella longbeachae are not a clonal population. Journal of clinical
microbiology, 37(10), 3249-3254.
49. Bacigalupe, R., Lindsay, D., Edwards, G., & Fitzgerald, J. R. (2017). Population
genomics of Legionella longbeachae and hidden complexities of infection source
attribution. Emerging infectious diseases, 23(5), 750.
50. Kozak-Muiznieks, N. A., Lucas, C. E., Brown, E., Pondo, T., Taylor, T. H., Jr., Frace,
M., Miskowski, D., & Winchell, J. M. (2014). Prevalence of sequence types among
clinical and environmental isolates of Legionella pneumophila serogroup 1 in the
United States from 1982 to 2012. J Clin Microbiol, 52(1), 201-211.
doi:10.1128/jcm.01973-13
51. Manske, C., & Hilbi, H. (2014). Metabolism of the vacuolar pathogen Legionella and
implications for virulence. Front Cell Infect Microbiol, 4, 125.
doi:10.3389/fcimb.2014.00125
52. Rolando, M., Sanulli, S., Rusniok, C., Gomez-Valero, L., Bertholet, C., Sahr, T.,
Margueron, R., & Buchrieser, C. (2013). Legionella pneumophila effector RomA
uniquely modifies host chromatin to repress gene expression and promote intracellular
bacterial replication. Cell host & microbe, 13(4), 395-405.
53. Molofsky, A. B., & Swanson, M. S. (2004). Differentiate to thrive: lessons from the
Legionella pneumophila life cycle. Molecular microbiology, 53(1), 29-40.
152
54. Bitar, D. M., Molmeret, M., & Kwaik, Y. A. (2004). Molecular and cell biology of
Legionella pneumophila. International Journal of Medical Microbiology, 293(7-8),
519-527.
55. Patrizia, M., Annalisa, B., Immacolata, A., Isabella, M., de Niederhäusern, S., &
Moreno, B. (2013). Protozoa and human macrophages infection by Legionella
pneumophila environmental strains belonging to different serogroups. Archives of
microbiology, 195(2), 89-96.
56. Gobin, I., Susa, M., Begic, G., Hartland, E. L., & Doric, M. (2009). Experimental
Legionella longbeachae infection in intratracheally inoculated mice. Journal of
medical microbiology, 58(6), 723-730.
57. Cazalet, C., Gomez-Valero, L., Rusniok, C., Lomma, M., Dervins-Ravault, D., Newton,
H. J., Sansom, F. M., Jarraud, S., Zidane, N., & Ma, L. (2010). Analysis of the
Legionella longbeachae genome and transcriptome uncovers unique strategies to cause
Legionnaires' disease. PLoS Genet, 6(2), e1000851.
58. Barkan, A., & Small, I. (2014). Pentatricopeptide repeat proteins in plants. Annual
review of plant biology, 65, 415-442.
59. Perry, C. M., Ormrod, D., Hurst, M., & Onrust, S. V. (2002). Gatifloxacin: a review of
its use in the management of bacterial infections. Drugs, 62(1), 169-207.
60. Balfour, J. A. B., & Wiseman, L. R. (1999). Moxifloxacin. Drugs, 57(3), 363-373.
61. Zarogoulidis, P., Alexandropoulou, I., Romanidou, G., Konstasntinidis, T. G., Terzi,
E., Saridou, S., Stefanis, A., Zarogoulidis, K., & Constantinidis, T. C. (2011).
Community-acquired pneumonia due to Legionella pneumophila, the utility of PCR,
and a review of the antibiotics used. Int J Gen Med, 4, 15-19. doi:10.2147/ijgm.s15654
153
62. Lambert, M., & Moss, C. W. (1989). Cellular fatty acid compositions and isoprenoid
quinone contents of 23 Legionella species. Journal of clinical microbiology, 27(3), 465-
473.
63. Edelstein, P. H., & Christian, L. (2015). Legionella. In Manual of Clinical
Microbiology, Eleventh Edition (pp. 887-904): American Society of Microbiology.
64. Bentz, J. S., Carroll, K., Ward, J. H., Elstad, M., & Marshall, C. J. (2000). Acid‐ fast‐
positive Legionella pneumophila: A possible pitfall in the cytologic diagnosis of
mycobacterial infection in pulmonary specimens. Diagnostic cytopathology, 22(1), 45-
48.
65. Inoue, H., Iwasawa, T., Saruwatari, Y., & Agata, K. (2004). Improved acid
pretreatment for the detection of Legionella species from environmental water samples
using the plate culture method. Biocontrol Science, 9(3), 43-50.
66. Nazarian, E. J., De Jesus, M., & Musser, K. A. (2015). Chapter 90 - Legionella. In Y.-
W. Tang, M. Sussman, D. Liu, I. Poxton, & J. Schwartzman (Eds.), Molecular Medical
Microbiology (Second Edition) (pp. 1611-1625). Boston: Academic Press.
67. Percival, S. L., & Williams, D. W. (2014). Chapter Eight - Legionella. In S. L. Percival,
M. V. Yates, D. W. Williams, R. M. Chalmers, & N. F. Gray (Eds.), Microbiology of
Waterborne Diseases (Second Edition) (pp. 155-175). London: Academic Press.
68. Gaia, V., Casati, S., & Tonolla, M. (2011). Rapid identification of Legionella spp. by
MALDI-TOF MS based protein mass fingerprinting. Systematic and Applied
Microbiology, 34(1), 40-44.
69. Kashuba, A. D., & Ballow, C. H. (1996). Legionella urinary antigen testing: potential
impact on diagnosis and antibiotic therapy. Diagnostic microbiology and infectious
disease, 24(3), 129-139.
154
70. Farshy, C. E., Klein, G. C., & Feeley, J. C. (1978). Detection of antibodies to
legionnaires disease organism by microagglutination and micro-enzyme-linked
immunosorbent assay tests. Journal of clinical microbiology, 7(4), 327-331.
71. Pierre, D. M., Baron, J., Victor, L. Y., & Stout, J. E. (2017). Diagnostic testing for
Legionnaires’ disease. Annals of Clinical Microbiology and Antimicrobials, 16(1), 1-
4.
72. Diederen, B. M., Kluytmans, J. A., Vandenbroucke-Grauls, C. M., & Peeters, M. F.
(2008). Utility of real-time PCR for diagnosis of Legionnaires' disease in routine
clinical practice. J Clin Microbiol, 46(2), 671-677. doi:10.1128/jcm.01196-07
73. Yang, G., Benson, R., Pelish, T., Brown, E., Winchell, J. M., & Fields, B. (2010). Dual
detection of Legionella pneumophila and Legionella species by real-time PCR
targeting the 23S-5S rRNA gene spacer region. Clin Microbiol Infect, 16(3), 255-261.
doi:10.1111/j.1469-0691.2009.02766.x
74. Montanaro-Punzengruber, J. C., Hicks, L., Meyer, W., & Gilbert, G. L. (1999).
Australian isolates of Legionella longbeachae are not a clonal population. J Clin
Microbiol, 37(10), 3249-3254. doi:10.1128/jcm.37.10.3249-3254.1999
75. Lee, J. V., Lai, S., Exner, M., Lenz, J., Gaia, V., Casati, S., Hartemann, P., Lück, C.,
Pangon, B., Ricci, M. L., Scaturro, M., Fontana, S., Sabria, M., Sánchez, I., Assaf, S.,
& Surman-Lee, S. (2011). An international trial of quantitative PCR for monitoring
Legionella in artificial water systems. J Appl Microbiol, 110(4), 1032-1044.
doi:10.1111/j.1365-2672.2011.04957.x
76. Hirama, T., Minezaki, S., Yamaguchi, T., Kishi, E., Kodama, K., Egashira, H.,
Kobayashi, K., Nagata, M., Ishii, T., & Nemoto, M. (2014). HIRA-TAN: a real-time
PCR-based system for the rapid identification of causative agents in pneumonia.
Respiratory Medicine, 108(2), 395-404.
155
77. Scott-Thomas, A. J., Syhre, M., Pattemore, P. K., Epton, M., Laing, R., Pearson, J., &
Chambers, S. T. (2010). 2-Aminoacetophenone as a potential breath biomarker for
Pseudomonas aeruginosa in the cystic fibrosis lung. BMC pulmonary medicine, 10(1),
56.
78. Scott-Thomas, A., & Chambers, S. T. (2017). Volatile Organic Compounds: Upcoming
Role in Diagnosis of Invasive Mould Infections. Current Fungal Infection Reports,
11(3), 117-123.
79. Syhre, M., Scotter, J. M., & Chambers, S. T. (2008). Investigation into the production
of 2-Pentylfuran by Aspergillus fumigatus and other respiratory pathogens in vitro and
human breath samples. Medical Mycology, 46(3), 209-215.
80. Chambers, S. T., Bhandari, S., Scott-Thomas, A., & Syhre, M. (2011). Novel
diagnostics: progress toward a breath test for invasive Aspergillus fumigatus. Medical
Mycology, 49(Supplement_1), S54-S61.
81. Schulz, S., & Dickschat, J. S. (2007). Bacterial volatiles: the smell of small organisms.
Natural product reports, 24(4), 814-842.
82. Chambers, S. T., Scott-Thomas, A., & Epton, M. (2012). Developments in novel breath
tests for bacterial and fungal pulmonary infection. Current opinion in pulmonary
medicine, 18(3), 228-232.
83. Korpi, A., Järnberg, J., & Pasanen, A.-L. (2009). Microbial volatile organic
compounds. Critical reviews in toxicology, 39(2), 139-193.
84. Syhre, M., Manning, L., Phuanukoonnon, S., Harino, P., & Chambers, S. T. (2009).
The scent of Mycobacterium tuberculosis–part II breath. Tuberculosis, 89(4), 263-266.
85. Dummer, J., Storer, M., Swanney, M., McEwan, M., Scott-Thomas, A., Bhandari, S.,
Chambers, S., Dweik, R., & Epton, M. (2011). Analysis of biogenic volatile organic
156
compounds in human health and disease. TrAC Trends in Analytical Chemistry, 30(7),
960-967.
86. Phillips, M., Herrera, J., Krishnan, S., Zain, M., Greenberg, J., & Cataneo, R. N. (1999).
Variation in volatile organic compounds in the breath of normal humans. Journal of
Chromatography B: Biomedical Sciences and Applications, 729(1-2), 75-88.
87. Saalberg, Y., & Wolff, M. (2016). VOC breath biomarkers in lung cancer. Clinica
Chimica Acta, 459, 5-9.
88. Davies, S., Spanel, P., & Smith, D. (1997). Quantitative analysis of ammonia on the
breath of patients in end-stage renal failure. Kidney international, 52(1), 223-228.
89. Van den Velde, S., Nevens, F., van Steenberghe, D., & Quirynen, M. (2008). GC–MS
analysis of breath odor compounds in liver patients. Journal of Chromatography B,
875(2), 344-348.
90. Di Francesco, F., Fuoco, R., Trivella, M. G., & Ceccarini, A. (2005). Breath analysis:
trends in techniques and clinical applications. Microchemical journal, 79(1-2), 405-
410.
91. Adams, N., & Smith, D. (1976). The selected ion flow tube (SIFT); a technique for
studying ion-neutral reactions. International Journal of Mass Spectrometry and Ion
Physics, 21(3-4), 349-359.
92. Smith, D. (1992). The ion chemistry of interstellar clouds. Chemical reviews, 92(7),
1473-1485.
93. Senthilmohan, S. T., McEwan, M. J., Wilson, P. F., Milligan, D. B., & Freeman, C. G.
(2001). Real time analysis of breath volatiles using SIFT-MS in cigarette smoking.
Redox Report, 6(3), 185-187.
94. Lindinger, W., Hansel, A., & Jordan, A. (1998). On-line monitoring of volatile organic
compounds at pptv levels by means of proton-transfer-reaction mass spectrometry
157
(PTR-MS) medical applications, food control and environmental research.
International Journal of Mass Spectrometry and Ion Processes, 173(3), 191-241.
95. Lourenço, C., & Turner, C. (2014). Breath analysis in disease diagnosis:
methodological considerations and applications. Metabolites, 4(2), 465-498.
96. Amann, A., Poupart, G., Telser, S., Ledochowski, M., Schmid, A., & Mechtcheriakov,
S. (2004). Applications of breath gas analysis in medicine. International Journal of
Mass Spectrometry, 239(2-3), 227-233.
97. Bean, H. D., Zhu, J., & Hill, J. E. (2011). Characterizing bacterial volatiles using
secondary electrospray ionization mass spectrometry (SESI-MS). JoVE (Journal of
Visualized Experiments)(52), e2664.
98. Zhu, J., Bean, H. D., Kuo, Y.-M., & Hill, J. E. (2010). Fast detection of volatile organic
compounds from bacterial cultures by secondary electrospray ionization-mass
spectrometry. Journal of clinical microbiology, 48(12), 4426-4431.
99. Zhu, J., Bean, H. D., Jiménez-Díaz, J., & Hill, J. E. (2013). Secondary electrospray
ionization-mass spectrometry (SESI-MS) breathprinting of multiple bacterial lung
pathogens, a mouse model study. Journal of applied physiology, 114(11), 1544-1549.
100. Zhu, J., & Hill, J. E. (2013). Detection of Escherichia coli via VOC profiling using
secondary electrospray ionization-mass spectrometry (SESI-MS). Food microbiology,
34(2), 412-417.
101. Pavlou, A. K., Magan, N., McNulty, C., Jones, J. M., Sharp, D., Brown, J., & Turner,
A. P. (2002). Use of an electronic nose system for diagnoses of urinary tract infections.
Biosensors and Bioelectronics, 17(10), 893-899.
102. Röck, F., Barsan, N., & Weimar, U. (2008). Electronic nose: current status and future
trends. Chemical reviews, 108(2), 705-725.
158
103. Staerz, A., Roeck, F., Weimar, U., & Barsan, N. (2020). Electronic Nose: Current
Status and Future Trends. Surface and Interface Science: Volume 9: Applications of
Surface Science I, 9, 335-379.
104. Westhoff, M., Litterst, P., Freitag, L., Urfer, W., Bader, S., & Baumbach, J. I. (2009).
Ion mobility spectrometry for the detection of volatile organic compounds in exhaled
breath of patients with lung cancer: results of a pilot study. Thorax, 64(9), 744-748.
doi:10.1136/thx.2008.099465
105. Španěl, P., & Smith, D. (1996). Selected ion flow tube: a technique for quantitative
trace gas analysis of air and breath. Medical and Biological Engineering and
Computing, 34(6), 409-419. doi:10.1007/BF02523843
106. Thomas, K. W., Pellizzari, E. D., & Cooper, S. D. (1991). A canister-based method for
collection and GC/MS analysis of volatile organic compounds in human breath. J Anal
Toxicol, 15(2), 54-59. doi:10.1093/jat/15.2.54
107. Herpers, B. L., de Jongh, B. M., van der Zwaluw, K., & van Hannen, E. J. (2003). Real-
time PCR assay targets the 23S-5S spacer for direct detection and differentiation of
Legionella spp. and Legionella pneumophila. J Clin Microbiol, 41(10), 4815-4816.
doi:10.1128/jcm.41.10.4815-4816.2003
108. Chen, D. J., Procop, G. W., Vogel, S., Yen-Lieberman, B., & Richter, S. S. (2015).
Utility of PCR, Culture, and Antigen Detection Methods for Diagnosis of Legionellosis.
J Clin Microbiol, 53(11), 3474-3477. doi:10.1128/jcm.01808-15
109. Bopp, C. A., Sumner, J. W., Morris, G. K., & Wells, J. G. (1981). Isolation of
Legionella spp. from environmental water samples by low-pH treatment and use of a
selective medium. J Clin Microbiol, 13(4), 714-719. doi:10.1128/jcm.13.4.714-
719.1981
159
110. Edelstein, P. H., Snitzer, J. B., & Bridge, J. A. (1982). Enhancement of recovery of
Legionella pneumophila from contaminated respiratory tract specimens by heat. J Clin
Microbiol, 16(6), 1061-1065. doi:10.1128/jcm.16.6.1061-1065.1982
111. Steele, T., Moore, C., & Sangster, N. (1990). Distribution of Legionella longbeachae
serogroup 1 and other legionellae in potting soils in Australia. Appl Environ Microbiol,
56(10), 2984-2988.
112. Chapman, P., Ellin, M., Ashton, R., & Shafique, W. (2001). Comparison of culture,
PCR and immunoassays for detecting Escherichia coli O157 following enrichment
culture and immunomagnetic separation performed on naturally contaminated raw
meat products. International journal of food microbiology, 68(1-2), 11-20.
113. Jeníková, G., Pazlarová, J., & Demnerová, K. (2000). Detection of Salmonella in food
samples by the combination of immunomagnetic separation and PCR assay.
International Microbiology, 3(4), 225-229.
114. Jung, Y. S., Frank, J. F., & Brackett, R. E. (2003). Evaluation of antibodies for
immunomagnetic separation combined with flow cytometry detection of Listeria
monocytogenes. Journal of food protection, 66(7), 1283-1287.
115. Yatsomboon, A., Sermswan, R. W., & Wongratanacheewin, S. (2018). Development of
an immunomagnetic separation-ELISA for the detection of Burkholderia pseudomallei
in blood samples. Asian Pac J Allergy Immunol.
116. Kassimi, L. B., Gonzague, M., Boutrouille, A., & Cruciere, C. (2002). Detection of
encephalomyocarditis virus in clinical samples by immunomagnetic separation and
one-step RT-PCR. Journal of virological methods, 101(1-2), 197-206.
117. Apaire-Marchais, V., Kempf, M., Lefrançois, C., Marot, A., Licznar, P., Cottin, J.,
Poulain, D., & Robert, R. (2008). Evaluation of an immunomagnetic separation method
to capture Candida yeasts cells in blood. BMC microbiology, 8(1), 157.
160
118. Gao, S., Zhang, M., Amer, S., Luo, J., Wang, C., Wu, S., Zhao, B., & He, H. (2014).
Development of an immunomagnetic bead separation-coupled quantitative PCR
method for rapid and sensitive detection of Cryptosporidium parvum oocysts in calf
feces. Parasitology research, 113(6), 2069-2077.
119. Stevens, K. A., & Jaykus, L.-A. (2004). Bacterial separation and concentration from
complex sample matrices: a review. Critical reviews in microbiology, 30(1), 7-24.
120. Lea, T., Vartdal, F., Davies, C., & Ugelstad, J. (1985). Magnetic monosized polymer
particles for fast and specific fractionation of human mononuclear cells. Scandinavian
journal of immunology, 22(2), 207-216.
121. Orlofsky, E., Gillor, O., Melli, A., Miller, W., Wuertz, S., Bernstein, N., & Shapiro, K.
(2013). Simultaneous detection of Giardia lamblia and Cryptosporidium parvum (oo)
cysts in soil using immunomagnetic separation and direct fluorescent antibody
staining. Journal of microbiological methods, 94(3), 375-377.
122. Rochelle, P. A., De Leon, R., Johnson, A., Stewart, M. H., & Wolfe, R. L. (1999).
Evaluation of immunomagnetic separation for recovery of infectious Cryptosporidium
parvum oocysts from environmental samples. Appl Environ Microbiol, 65(2), 841-845.
123. Enroth, H., & Engstrand, L. (1995). Immunomagnetic separation and PCR for detection
of Helicobacter pylori in water and stool specimens. Journal of clinical microbiology,
33(8), 2162-2165.
124. Pelagiadis, I., Dimitriou, H., & Kalmanti, M. (2007). CD133 Immunomagnetic
Separation: How Effective Is the Method for CD133+ Cell Isolation from Umbilical
Cord Blood? In: American Society of Hematology.
125. Yu—qun, Z., Wei—guo, X., Wei, L., Yong—jun, J., & Zhan—pei, X. (2010). The
diagnosis value of immunomagnetic bead separation of mycobacterium tuberculosis
161
from sputum for smear and polymerase chain reaction. Jiangsu Journal of Preventive
Medicine(2), 6.
126. Toze, S., Cahill, M., Sly, L., & Fuerst, J. (1994). The effect of Aeromonas strains on the
growth of Legionella. Journal of applied bacteriology, 77(2), 169-174.
127. Albers, C. N., Jensen, A., Bælum, J., & Jacobsen, C. S. (2013). Inhibition of DNA
polymerases used in Q-PCR by structurally different soil-derived humic substances.
Geomicrobiology Journal, 30(8), 675-681.
128. Riffard, S., Douglass, S., Brooks, T., Springthorpe, S., Filion, L., & Sattar, S. (2001).
Occurrence of Legionella in groundwater: an ecological study. Water Science and
Technology, 43(12), 99-102.
129. Leenaars, M., & Hendriksen, C. F. (2005). Critical steps in the production of polyclonal
and monoclonal antibodies: evaluation and recommendations. Ilar Journal, 46(3), 269-
279.
130. Vaitukaitis, J., Robbins, J., Nieschlag, E., & Ross, G. (1971). A method for producing
specific antisera with small doses of immunogen. The Journal of Clinical
Endocrinology & Metabolism, 33(6), 988-991.
131. Baumstark, J., Laffin, R. J., & Bardawil, W. (1964). A preparative method for the
separation of 7S gamma globulin from human serum. Archives of Biochemistry and
Biophysics, 108(3), 514-522.
132. Ditommaso, S., Giacomuzzi, M., Gentile, M., & Zotti, C. M. (2008). Antibody detection
and cross-reactivity among species and serogroups of Legionella by indirect
immunofluorescence test. Journal of microbiological methods, 75(2), 350-353.
133. Wilkinson, H. W., & Brake, B. J. (1982). Formalin-killed versus heat-killed Legionella
pneumophila serogroup 1 antigen in the indirect immunofluorescence assay for
legionellosis. Journal of clinical microbiology, 16(5), 979-981.
162
134. Strug, I., Utzat, C., & Nadler, T. (2012). IR-Based Protein & Peptide Quantitation:
New System Focuses on Intrinsic Amide Bonds. Genetic Engineering & Biotechnology
News, 32(19), 30-31.
135. Kuroki, H., Miyamoto, H., Fukuda, K., Iihara, H., Kawamura, Y., Ogawa, M., Wang,
Y., Ezaki, T., & Taniguchi, H. (2007). Legionella impletisoli sp. nov. and Legionella
yabuuchiae sp. nov., isolated from soils contaminated with industrial wastes in Japan.
Systematic and Applied Microbiology, 30(4), 273-279.
136. Rodrigo-Troyano, A., Melo, V., Marcos, P. J., Laserna, E., Peiro, M., Suarez-Cuartin,
G., Perea, L., Feliu, A., Plaza, V., & Faverio, P. (2018). Pseudomonas aeruginosa in
chronic obstructive pulmonary disease patients with frequent hospitalized
exacerbations: a prospective multicentre study. Respiration, 96(5), 417-424.
137. Malhotra, S., Hayes, D., & Wozniak, D. J. (2019). Cystic fibrosis and Pseudomonas
aeruginosa: the host-microbe interface. Clin Microbiol Rev, 32(3).
138. Tong, X., Cheng, A., Xu, H., Jin, J., Yang, Y., Zhu, S., & Li, Y. (2018). Aspergillus
fumigatus during COPD exacerbation: a pair-matched retrospective study. BMC
pulmonary medicine, 18(1), 55.
139. Engel, T. G., Erren, E., Driessche, K. S. V., Melchers, W. J., Reijers, M. H., Merkus,
P., & Verweij, P. E. (2019). Aerosol transmission of Aspergillus fumigatus in cystic
fibrosis patients in the Netherlands. Emerging infectious diseases, 25(4), 797.
140. Green, S. K., Schroth, M. N., Cho, J. J., Kominos, S. D., & Vitanza-Jack, V. B. (1974).
Agricultural plants and soil as a reservoir for Pseudomonas aeruginosa. Applied
microbiology, 28(6), 987-991.
141. Klein, G. C. (1980). Cross-reaction to Legionella pneumophila antigen in sera with
elevated titers to Pseudomonas pseudomallei. J Clin Microbiol, 11(1), 27-29.
doi:10.1128/jcm.11.1.27-29.1980
163
142. Finidori, J. P., Raoult, D., Bornstein, N., & Fleurette, J. (1992). Study of cross-reaction
between Coxiella burnetii and Legionella pneumophila using indirect
immunofluorescence assay and immunoblotting. Acta Virol, 36(5), 459-465.
143. Musso, D., & Raoult, D. (1997). Serological cross-reactions between Coxiella burnetii
and Legionella micdadei. Clin Diagn Lab Immunol, 4(2), 208-212.
144. Šafařıḱ , I., & Šafařıḱ ová, M. (1999). Use of magnetic techniques for the isolation of
cells. Journal of Chromatography B: Biomedical Sciences and Applications, 722(1-2),
33-53.
145. Allegra, S., Girardot, F., Grattard, F., Berthelot, P., Helbig, J., Pozzetto, B., & Riffard,
S. (2011). Evaluation of an immunomagnetic separation assay in combination with
cultivation to improve Legionella pneumophila serogroup 1 recovery from
environmental samples. J Appl Microbiol, 110(4), 952-961.
146. Edelstein, P. H. (1981). Improved semiselective medium for isolation of Legionella
pneumophila from contaminated clinical and environmental specimens. Journal of
clinical microbiology, 14(3), 298-303.
147. Lindsay, D. S., Abraham, W. H., Findlay, W., Christie, P., Johnston, F., & Edwards, G.
F. (2004). Laboratory diagnosis of legionnaires’ disease due to Legionella
pneumophila serogroup 1: comparison of phenotypic and genotypic methods. Journal
of medical microbiology, 53(3), 183-187.
148. Boulanger, C. A., & Edelstein, P. H. (1995). Precision and accuracy of recovery of
Legionella pneumophila from seeded tap water by filtration and centrifugation. Appl
Environ Microbiol, 61(5), 1805-1809.
149. Roberts, K. P., August, C. M., & Nelson, J. (1987). Relative sensitivities of
environmental legionellae to selective isolation procedures. Appl Environ Microbiol,
53(12), 2704-2707.
164
150. Ballard, A., Fry, N., Chan, L., Surman, S., Lee, J., Harrison, T., & Towner, K. (2000).
Detection of Legionella pneumophilausing a real-time PCR hybridization assay.
Journal of clinical microbiology, 38(11), 4215-4218.
151. Hayden, R., Uhl, J., Qian, X., Hopkins, M., Aubry, M., Limper, A. H., Lloyd, R., &
Cockerill, F. (2001). Direct detection of Legionella species from bronchoalveolar
lavage and open lung biopsy specimens: comparison of LightCycler PCR, in situ
hybridization, direct fluorescence antigen detection, and culture. Journal of clinical
microbiology, 39(7), 2618-2626.
152. Wellinghausen, N., Frost, C., & Marre, R. (2001). Detection of legionellae in hospital
water samples by quantitative real-time LightCycler PCR. Appl Environ Microbiol,
67(9), 3985-3993.
153. Robinson, P. N., Heidrich, B., Tiecke, F., Fehrenbach, F. J., & Rolfs, A. (1996).
Species-specific detection of Legionella using polymerase chain reaction and reverse
dot-blotting. FEMS microbiology letters, 140(2-3), 111-119.
154. Behets, J., Declerck, P., Delaedt, Y., Creemers, B., & Ollevier, F. (2007). Development
and evaluation of a Taqman duplex real-time PCR quantification method for reliable
enumeration of Legionella pneumophila in water samples. Journal of microbiological
methods, 68(1), 137-144.
155. Akane, A., Matsubara, K., Nakamura, H., Takahashi, S., & Kimura, K. (1994).
Identification of the heme compound copurified with deoxyribonucleic acid (DNA) from
bloodstains, a major inhibitor of polymerase chain reaction (PCR) amplification.
Journal of Forensic Science, 39(2), 362-372.
156. Gracias, K. S., & McKillip, J. L. (2004). A review of conventional detection and
enumeration methods for pathogenic bacteria in food. Canadian journal of
microbiology, 50(11), 883-890.
165
157. Lee, H., Lee, K., Kim, D., Choi, Y., Park, K., & Seong, J. (2017). Application of
immunomagnetic separation for detection of human coxsackievirus group b using
polyclonal antibodies.
158. Grant, I. R., & Stewart, L. D. (2015). Improved detection of Mycobacterium bovis in
bovine tissues using immunomagnetic separation approaches. In Veterinary infection
biology: molecular diagnostics and high-throughput strategies (pp. 153-161): Springer.
159. Tsai, H.-L., Hsu, B.-M., Hsu, T.-K., Huang, K.-H., Shih, F.-C., Chen, J.-S., Wang, H.-
J., Kao, P.-M., & Su, H.-C. (2015). Evaluation of immunomagnetic separation for the
improvement of Salmonella detection in aquatic environment. Environmental Earth
Sciences, 73(12), 7909-7914.
160. Lim, W., Van der Eerden, M., Laing, R., Boersma, W., Karalus, N., Town, G., Lewis,
S., & Macfarlane, J. (2003). Defining community acquired pneumonia severity on
presentation to hospital: an international derivation and validation study. Thorax,
58(5), 377-382.
161. Héchard, Y., Ferraz, S., Bruneteau, E., Steinert, M., & Berjeaud, J.-M. (2005). Isolation
and characterization of a Staphylococcus warneri strain producing an anti-Legionella
peptide. FEMS microbiology letters, 252(1), 19-23.
162. Dusserre, E., Ginevra, C., Hallier-Soulier, S., Vandenesch, F., Festoc, G., Etienne, J.,
Jarraud, S., & Molmeret, M. (2008). A PCR-based method for monitoring Legionella
pneumophila in water samples detects VBNC that can recover their cultivability. Appl
Environ Microbiol.
163. Olsvik, O., Popovic, T., Skjerve, E., Cudjoe, K. S., Hornes, E., Ugelstad, J., & Uhlen,
M. (1994). Magnetic separation techniques in diagnostic microbiology. Clin Microbiol
Rev, 7(1), 43-54.
166
164. McCloskey, K. E., Chalmers, J. J., & Zborowski, M. (2003). Magnetic cell separation:
characterization of magnetophoretic mobility. Analytical chemistry, 75(24), 6868-
6874.
165. WANG, W. L. L., BLASER, M. J., CRAVENS, J., & JOHNSON, M. A. (1979). The
Microorganism: Growth, Survival, and Resistance of the Legionnaires' Disease
Bacterium. Annals of Internal Medicine, 90(4), 614-618.
166. Bartie, C., Venter, S., & Nel, L. (2003). Identification methods for Legionella from
environmental samples. Water research, 37(6), 1362-1370.
167. Edelstein, P. H. (1983). Culture diagnosis of Legionella infections. Zentralblatt für
Bakteriologie, Mikrobiologie und Hygiene. 1. Abt. Originale. A, Medizinische
Mikrobiologie, Infektionskrankheiten und Parasitologie, 255(1), 96-101.
168. Murray, M. P., Doherty, C. J., Govan, J. R., & Hill, A. T. (2010). Do processing time
and storage of sputum influence quantitative bacteriology in bronchiectasis? Journal
of medical microbiology, 59(7), 829-833.
169. Hill, A., Zhang, Y., Kang, K., Clarke, A., & Donaldson, S. (2016). Optimum Condition
for Storage of Sputum from Patients with Bronchiectasis Infected with Haemophilus
influenzae. Arch Clin Microbiol, 7, 1-4.
170. Ayub, N. D., Tribelli, P. M., & López, N. I. (2009). Polyhydroxyalkanoates are
essential for maintenance of redox state in the Antarctic bacterium Pseudomonas sp.
14-3 during low temperature adaptation. Extremophiles, 13(1), 59-66.
171. Obruca, S., Sedlacek, P., Krzyzanek, V., Mravec, F., Hrubanova, K., Samek, O.,
Kucera, D., Benesova, P., & Marova, I. (2016). Accumulation of poly (3-
hydroxybutyrate) helps bacterial cells to survive freezing. PloS one, 11(6), e0157778.
172. Gillmaier, N., Schunder, E., Kutzner, E., Tlapák, H., Rydzewski, K., Herrmann, V.,
Stämmler, M., Lasch, P., Eisenreich, W., & Heuner, K. (2016). Growth-related
167
metabolism of the carbon storage poly-3-hydroxybutyrate in Legionella pneumophila.
Journal of Biological Chemistry, 291(12), 6471-6482.
173. Subbaram, K., Kannan, H., & Masadeh, M. M. A. (2017). Isolation, identification,
characterization and antibiotic sensitivity profile of pathogenic Legionella
pneumophila isolates from different water sources. Asian Pacific Journal of Tropical
Biomedicine, 7(5), 411-415.
174. Isenman, H., Anderson, T., Chambers, S. T., Podmore, R. G., & Murdoch, D. R. (2018).
Antimicrobial susceptibilities of clinical Legionella longbeachae isolates. Journal of
Antimicrobial Chemotherapy, 73(4), 1102-1104.
175. Isenman, H. L., Chambers, S. T., Pithie, A. D., MacDonald, S. L., Hegarty, J. M.,
Fenwick, J. L., Maze, M. J., Metcalf, S. C., & Murdoch, D. R. (2016). Legionnaires'
disease caused by Legionella longbeachae: Clinical features and outcomes of 107
cases from an endemic area. Respirology, 21(7), 1292-1299.
176. Zainol Hilmi, N. H., Idris, A. S., & Mohd Azmil, M. N. (2019). Headspace solid‐
phase microextraction gas chromatography–mass spectrometry for the detection of
volatile organic compounds released from Ganoderma boninense and oil palm wood.
Forest Pathology, 49(4), e12531.
177. Abd El Qader, A., Lieberman, D., Shemer Avni, Y., Svobodin, N., Lazarovitch, T.,
Sagi, O., & Zeiri, Y. (2015). Volatile organic compounds generated by cultures of
bacteria and viruses associated with respiratory infections. Biomedical
Chromatography, 29(12), 1783-1790.
178. Phillips, M., Cataneo, R. N., Condos, R., Erickson, G. A. R., Greenberg, J., La
Bombardi, V., Munawar, M. I., & Tietje, O. (2007). Volatile biomarkers of pulmonary
tuberculosis in the breath. Tuberculosis, 87(1), 44-52.
168
179. Eitrem, R., Forsgren, A., & Nilsson, C. (1987). Pneumonia and acute pancreatitis most
probably caused by a Legionella longbeachae infection. Scandinavian journal of
infectious diseases, 19(3), 381-382.
180. Lindsay, D., Brown, A., Brown, D., Pravinkumar, S., Anderson, E., & Edwards, G.
(2012). Legionella longbeachae serogroup 1 infections linked to potting compost.
Journal of medical microbiology, 61(2), 218-222.
181. Potts, A., Donaghy, M., Marley, M., Othieno, R., Stevenson, J., Hyland, J., Pollock, K.,
Lindsay, D., Edwards, G., & Hanson, M. (2013). Cluster of Legionnaires’ disease cases
caused by Legionella longbeachae serogroup 1, Scotland, August to September 2013.
Eurosurveillance, 18(50), 20656.
182. Barney, D. L., & Colt, W. M. (1991). Using bark and sawdust for mulches, soil
amendments and potting mixes. Current information series-Cooperative Extension
Service, University of Idaho (USA).
183. Cunha-Queda, A., Ribeiro, H., Ramos, A., & Cabral, F. (2007). Study of biochemical
and microbiological parameters during composting of pine and eucalyptus bark.
Bioresource technology, 98(17), 3213-3220.
184. Casati, S., Conza, L., Bruin, J., & Gaia, V. (2010). Compost facilities as a reservoir of
Legionella pneumophila and other Legionella species. Clinical Microbiology and
Infection, 16(7), 945-947.
185. Schmilewski, G. (2007). Growing medium constituents used in the EU. Paper presented
at the International Symposium on Growing Media 2007 819.
186. Yáñez, M. A., Carrasco-Serrano, C., Barberá, V. M., & Catalan, V. (2005). Quantitative
detection of Legionella pneumophila in water samples by immunomagnetic purification
and real-time PCR amplification of the dotA gene. Appl Environ Microbiol, 71(7),
3433-3441.
169
187. CDC. (2018). Legionnaires Disease Procedures Manual for Recovery. Retrieved from
https://www.cdc.gov/Legionella/labs/procedures-manual.html
188. Kasuga, O., Takagi, K., Misawa, H., Onozawa, M., Chiba, E., Tani, K., & Ikedo, M.
(2002). Effect of the pretreatment with organic acid buffer solutions on the recovery of
Legionella species from environmental water samples. Kansenshogaku zasshi. The
Journal of the Japanese Association for Infectious Diseases, 76(12), 1010-1015.
189. Oliver, J. D. (2005). The viable but nonculturable state in bacteria. Journal of
Microbiology (Seoul, Korea), 43, 93-100.
190. Wallis, L., & Robinson, P. (2005). Soil as a source of Legionella pneumophila
serogroup 1 (Lp1). Australian and New Zealand journal of public health, 29(6), 518-
520.
191. Wadowsky, R. M., Wolford, R., McNamara, A., & Yee, R. B. (1985). Effect of
temperature, pH, and oxygen level on the multiplication of naturally occurring
Legionella pneumophila in potable water. Appl Environ Microbiol, 49(5), 1197-1205.
192. Cervero-Aragó, S., Rodríguez-Martínez, S., Puertas-Bennasar, A., & Araujo, R. M.
(2015). Effect of common drinking water disinfectants, chlorine and heat, on free
Legionella and amoebae-associated Legionella. PloS one, 10(8), e0134726.
193. Sheehan, K. B., Henson, J. M., & Ferris, M. J. (2005). Legionella species diversity in
an acidic biofilm community in Yellowstone National Park. Appl Environ Microbiol,
71(1), 507-511.
194. Feeley, J., Gorman, G., Weaver, R., Mackel, D. C., & Smith, H. (1978). Primary
isolation media for Legionnaires disease bacterium. Journal of clinical microbiology,
8(3), 320-325.
195. Wooldridge, K. G., & Williams, P. H. (1993). Iron uptake mechanisms of pathogenic
bacteria. FEMS microbiology reviews, 12(4), 325-348.
170
196. Reeves, M., Pine, L., Hutner, S., George, J., & Harrell, W. K. (1981). Metal
requirements of Legionella pneumophila. Journal of clinical microbiology, 13(4), 688-
695.
197. Pine, L., George, J., Reeves, M., & Harrell, W. K. (1979). Development of a chemically
defined liquid medium for growth of Legionella pneumophila. Journal of clinical
microbiology, 9(5), 615-626.
198. George, J. R., Pine, L., Reeves, M. W., & Harrell, W. K. (1980). Amino acid
requirements of Legionella pneumophila. J Clin Microbiol, 11(3), 286-291.
doi:10.1128/jcm.11.3.286-291.1980
199. Tesh, M. J., & Miller, R. D. (1981). Amino acid requirements for Legionella
pneumophila growth. J Clin Microbiol, 13(5), 865-869. doi:10.1128/jcm.13.5.865-
869.1981
200. Rowbotham, T. J. (1980). Preliminary report on the pathogenicity of Legionella
pneumophila for freshwater and soil amoebae. J Clin Pathol, 33(12), 1179-1183.
doi:10.1136/jcp.33.12.1179
201. Keevil, C., Glenister, D., Salamon, K., Dennis, P., & West, A. (1988). A continuous
culture biofilm model for the study of medical and industrial corrosion. Paper presented
at the Proceedings of the Spring meeting Biodetioration.
202. Byrne, B., & Swanson, M. S. (1998). Expression of Legionella pneumophilavirulence
traits in response to growth conditions. Infection and immunity, 66(7), 3029-3034.
203. Doyle, R. M., Cianciotto, N. P., Banvi, S., Manning, P. A., & Heuzenroeder, M. W.
(2001). Comparison of Virulence ofLegionella longbeachae Strains in Guinea Pigs and
U937 Macrophage-Like Cells. Infection and immunity, 69(9), 5335-5344.
204. Johnson, W., Varner, L., & Poch, M. (1991). Acquisition of iron by Legionella
pneumophila: role of iron reductase. Infection and immunity, 59(7), 2376-2381.
171
205. Cianciotto, N. P. (2007). Iron acquisition by Legionella pneumophila. Biometals, 20(3-
4), 323-331.
206. Portier, E., Bertaux, J., Labanowski, J., & Hechard, Y. (2016). Iron Availability
modulates the persistence of Legionella pneumophila in complex biofilms. Microbes
and environments, ME16010.
207. Cianciotto, N. P. (2015). An update on iron acquisition by Legionella pneumophila:
new pathways for siderophore uptake and ferric iron reduction. Future microbiology,
10(5), 841-851.
208. Conley, L., Ceraso, M., Stephenson, T., Wolford, R., Wadowsky, R., McNamara, A.,
& Yee, R. (1985). Effects of metals on Legionella pneumophila growth in drinking
water plumbing systems. Appl Environ Microbiol, 50(5), 1149-1154.
209. Bargellini, A., Marchesi, I., Righi, E., Ferrari, A., Cencetti, S., Borella, P., & Rovesti,
S. (2011). Parameters predictive of Legionella contamination in hot water systems:
association with trace elements and heterotrophic plate counts. Water research, 45(6),
2315-2321.
210. Rogers, J., Dowsett, A., Dennis, P., Lee, J., & Keevil, C. (1994). Influence of
temperature and plumbing material selection on biofilm formation and growth of
Legionella pneumophila in a model potable water system containing complex microbial
flora. Appl Environ Microbiol, 60(5), 1585-1592.
211. Kusnetsov, J., Iivanainen, E., Elomaa, N., Zacheus, O., & Martikainen, P. J. (2001).
Copper and silver ions more effective against legionellae than against mycobacteria in
a hospital warm water system. Water research, 35(17), 4217-4225.
212. de Bertoldi, M. d., Vallini, G. e., & Pera, A. (1983). The biology of composting: a
review. Waste Management & Research, 1(2), 157-176.
172
213. Pine, L., Hoffman, P., Malcolm, G., Benson, R., & Keen, M. (1984). Determination of
catalase, peroxidase, and superoxide dismutase within the genus Legionella. Journal of
clinical microbiology, 20(3), 421-429.
214. Atlas, R. M. (1999). Legionella: from environmental habitats to disease pathology,
detection and control. Environmental Microbiology, 1(4), 283-293.
215. Rakic, A., Peric, J., & Foglar, L. (2012). Influence of temperature, chlorine residual
and heavy metals on the presence of Legionella pneumophila in hot water distribution
systems. Annals of Agricultural and Environmental Medicine, 19(3).
216. Katz, S. M., & Hammel, J. M. (1987). The effect of drying, heat, and pH on the survival
of Legionella pneumophila. Annals of Clinical & Laboratory Science, 17(3), 150-156.
217. Poikolainen, J. (1997). Sulphur and heavy metal concentrations in Scots pine bark in
northern Finland and the Kola Peninsula. Water, air, and soil pollution, 93(1-4), 395-
408.
218. Barbaree, J. M., Fields, B. S., Feeley, J. C., Gorman, G. W., & Martin, W. T. (1986).
Isolation of protozoa from water associated with a legionellosis outbreak and
demonstration of intracellular multiplication of Legionella pneumophila. Appl Environ
Microbiol, 51(2), 422-424.
219. Breiman, R. F., Fields, B. S., Sanden, G. N., Volmer, L., Meier, A., & Spika, J. S.
(1990). Association of shower use with Legionnaires' disease: possible role of
amoebae. Jama, 263(21), 2924-2926.
220. Wadowsky, R., Butler, L., Cook, M., Verma, S., Paul, M., Fields, B., Keleti, G., Sykora,
J., & Yee, R. (1988). Growth-supporting activity for Legionella pneumophila in tap
water cultures and implication of hartmannellid amoebae as growth factors. Appl
Environ Microbiol, 54(11), 2677-2682.
173
221. Skinner, A., Anand, C., Malic, A., & Kurtz, J. (1983). Acanthamoebae and
environmental spread of Legionella pneumophila. The Lancet, 322(8344), 289-290.
222. Jeker, P., & Krähenbühl, U. (2001). Sulfur profiles of the twentieth century in peat bogs
of the Swiss midlands measured by ICP-OES and by IC. CHIMIA International Journal
for Chemistry, 55(12), 1029-1032.
223. Yu-sen, E. L., Vidic, R. D., Stout, J. E., & Victor, L. Y. (2002). Negative effect of high
pH on biocidal efficacy of copper and silver ions in controlling Legionella
pneumophila. Appl Environ Microbiol, 68(6), 2711-2715.
224. Das, S., Jean, J.-S., Kar, S., & Chakraborty, S. (2013). Effect of arsenic contamination
on bacterial and fungal biomass and enzyme activities in tropical arsenic-contaminated
soils. Biology and fertility of soils, 49(6), 757-765.
225. Anderson, C. R., & Cook, G. M. (2004). Isolation and characterization of arsenate-
reducing bacteria from arsenic-contaminated sites in New Zealand. Current
microbiology, 48(5), 341-347.
226. Brown, P. H., Welch, R. M., & Cary, E. E. (1987). Nickel: A micronutrient essential
for higher plants. Plant physiology, 85(3), 801-803.
227. Sirko, A., & Brodzik, R. (2000). Plant ureases: roles and regulation. Acta Biochimica
Polonica, 47(4), 1189-1195.
228. Buracco, S., Peracino, B., Andreini, C., Bracco, E., & Bozzaro, S. (2018). Differential
effects of iron, zinc, and copper on dictyostelium discoideum cell growth and resistance
to Legionella pneumophila. Front Cell Infect Microbiol, 7, 536.
229. Coulthurst, S. J., Whitehead, N. A., Welch, M., & Salmond, G. P. (2002). Can boron
get bacteria talking? Trends in biochemical sciences, 27(5), 217-219.
174
230. Miwa, H., & Fujiwara, T. (2009). Isolation and identification of boron-accumulating
bacteria from contaminated soils and active sludge. Soil Science and Plant Nutrition,
55(5), 643-646.
231. Kredich, N. M. (2008). Biosynthesis of cysteine. EcoSal Plus, 3(1).
232. Ewann, F., & Hoffman, P. S. (2006). Cysteine metabolism in Legionella pneumophila:
characterization of an L-cystine-utilizing mutant. Appl Environ Microbiol, 72(6), 3993-
4000.
233. Cuppett, J. D., Duncan, S. E., & Dietrich, A. M. (2006). Evaluation of copper speciation
and water quality factors that affect aqueous copper tasting response. Chemical senses,
31(7), 689-697.
234. Kenagy, E., Priest, P. C., Cameron, C. M., Smith, D., Scott, P., Cho, V., Mitchell, P.,
& Murdoch, D. R. (2017). Risk factors for Legionella longbeachae Legionnaires’
disease, New Zealand. Emerging infectious diseases, 23(7), 1148.
235. Schrammel, B., Petzold, M., Cervero-Aragó, S., Sommer, R., Lück, C., & Kirschner,
A. (2018). Persistent presence of outer membrane epitopes during short-and long-term
starvation of five Legionella pneumophila strains. BMC microbiology, 18(1), 75.
236. Yáñez, M. A., Nocker, A., Soria-Soria, E., Múrtula, R., Martínez, L., & Catalán, V.
(2011). Quantification of viable Legionella pneumophila cells using propidium
monoazide combined with quantitative PCR. Journal of microbiological methods,
85(2), 124-130.
237. Hägele, S., Köhler, R., Merkert, H., Schleicher, M., Hacker, J., & Steinert, M. (2000).
Dictyostelium discoideum: a new host model system for intracellular pathogens of the
genus Legionella. Cell Microbiol, 2(2), 165-171. doi:10.1046/j.1462-
5822.2000.00044.x
175
238. Conza, L., Casati, S., & Gaia, V. (2013). Detection limits of Legionella pneumophila in
environmental samples after co-culture with Acanthamoeba polyphaga. BMC
Microbiol, 13, 49. doi:10.1186/1471-2180-13-49
239. Dietersdorfer, E., Kirschner, A., Schrammel, B., Ohradanova-Repic, A., Stockinger,
H., Sommer, R., Walochnik, J., & Cervero-Aragó, S. (2018). Starved viable but non-
culturable (VBNC) Legionella strains can infect and replicate in amoebae and human
macrophages. Water Res, 141, 428-438. doi:10.1016/j.watres.2018.01.058
176
Appendices
Appendix 1: Details of the tested respiratory specimens
Initial qPCR IMS-PCR Stored No. AW/culture IMS-Culture Ct value Ct value AW/culture
1 30.9 33.08 N N N 2 28 27.15 N N P 3 39.4 N N N P 4 40 N N N N 5 40 40 N N N 6 30.6 31.35 P P P 7 34 37.15 P P P 8 36 36.23 N N N 9 37 40 N N N 11 38 40 N N P 12 38 N N N N 16 34 37.19 N N P 17 34 N N N N 18 23 25.9 P P P 19 36 35.45 N N P 20 37 37.87 N N N 21 35 39.36 N N P 22 23.6 25.54 P P P 23 38 40 N N P 24 39 37.74 N N N 25 35 33.45 N N N 26 31 36.02 P N P 27 36 40 N N N 28 36 39.49 N N N 29 36 38.15 P N N 30 40.9 40 N N P 31 29 28.58 P N N 32 34 5 N N P 33 37 40 N N N 34 38.6 N N N N 35 22 27.62 P P P 36 32 35.54 P N N 37 27 30.9 N N N 38 37 37.79 N P P 39 37 34.81 N N P 40 33 32.12 P N N 41 37 38.88 N N N 42 35 34.68 P N N 43 31.7 36.94 N N N
177
46 29 29.56 P P P 47 36 36.54 N N P 48 32 32.82 P N P 49 26 28.71 P N P 50 35 34.88 P N N 52 35 37.46 N P P 53 35 35.82 N N N 55 27.3 33.2 N N N 56 34 33.15 N N P 58 30 35.02 P P P 59 30 30.47 N N N 60 35 35.04 P N N 61 39 30.24 N N P 62 35 36.06 N P P AW: Acid wash; N: Negative; P: Positive
178
Appendix 2: Details of samples collected from three potting mix production sites
Sample type Number Facility number Mixing area 53 Bark (5 grades) 100 Site 1 Sawdust 95 Peat 100 Mixing area 20 Sawdust (2 types) 40 Bark ( 8 types) 150 Site 2 Potting mix 15 Compost 20 Peat moss 5 Raw materials Bark (3 types) 90 Compost (2 types) 130 Sawdust 50 Pumice (two grades) 20 Peat 100 Site 3 Final products (bagged and stored) Potting mix (7 types) 81 Compost (4 types) 36 Bark and woodchips 63 Total 1168
179