Are species clinically relevant to humans?

Robyn Anne Nagel MB, BS, FRACP

A thesis submitted for the degree of Doctor of Philosophy at the University of Queensland in 2015 School of Veterinary Science

Title page Culture of human faecal specimen Blastocystis organisms, vacuolated and granular forms, Photographed RAN: x40 magnification, polarised light

ii Abstract

Blastocystis spp. are the most common enteric parasites found in human stool and yet, the life cycle of the organism is unknown and the clinical relevance uncertain. Robust cysts transmit infection, and many animals carry the parasite. Infection in humans has been linked to (IBS). Although Blastocystis carriage is much higher in IBS patients, studies have not been able to confirm Blastocystis spp. are the direct cause of symptoms. Moreover, eradication is often unsuccessful. A number of approaches were utilised in order to investigate the clinical relevance of Blastocystis spp. in human IBS patients. Deconvolutional microscopy with time-lapse imaging and fluorescent spectroscopy of xenic cultures of Blastocystis spp. from IBS patients and healthy individuals was performed. Green autofluorescence (GAF), most prominently in the 557/576 emission spectra, was observed in the vacuolated, granular, amoebic and cystic Blastocystis forms. This first report of GAF in Blastocystis showed that a Blastocystis-specific fluorescein-conjugated antibody could be partially distinguished from GAF. Surface pores of 1m in diameter were observed cyclically opening and closing over 24 hours and may have nutritional or motility functions. Vacuolated forms, extruded a viscous material slowly over 12 hours, a process likely involving osmoregulation. Tear- shaped granules were observed exiting from the surface of an amoebic form but their identity and function could not be elucidated. Seventeen different subtypes of Blastocystis based on the 18S subunit of ribosomal DNA have been described. Using subtyping techniques, the human and animal household contacts of 11 symptomatic Blastocystis-positive patients were studied. All household contacts were found to be positive for Blastocystis spp. The specific subtypes within households had high similarity with the symptomatic patients (94% in human and 88% in animal contacts). Although 18S rDNA subtyping was not sufficient to discriminate genetically identical organisms, the ST findings suggest that intra-household transmission of Blastocystis is almost universal. IBS is a common, heterogeneous condition with no objective biomarker. We studied eradication rates of Blastocystis following administration of different antimicrobial drug regimes. Currently recommended first line therapy for eradication, 400 mg three times daily or sulfamethozazole/trimethoprim 160/80mg were administered to 11 symptomatic patients for 14 days. No patient was negative for Blastocystis carriage after therapy with either drug. In a later study 10 diarrhoea-predominant IBS Blastocystis- positive patients were given 14 days of triple antimicrobial therapy, namely diloxanide furoate 500mg and 400 mg three times daily, and trimethoprim- iii sulfamethoxazole 160/80 mg twice daily, with 60% of the patients eradicating the organism. It is not certain if particular Blastocystis organisms may have more pathogenic potential than others. Many different subtypes of Blastocystis were found in all symptomatic patients studied in this thesis and thus far there is no support for particular subtypes either being pathogenic or commonly transmitted from non-domestic animal contacts. Our study of household pets, combined with current knowledge in the literature, suggests that Blastocystis may well be a reverse zoonosis when it involves companion domestic pets such as dogs and cats. If there is no correlation between the Blastocystis subtype and pathogenic potential it is possible that the host immunological response determines symptomatology. The serological responses to infection in 40 symptomatic diarrhoea-predominant IBS patients (26 positive, 14 negative for Blastocystis) and 40 healthy controls (24 positive, 16 negative for Blastocystis) were examined. Low serum immunoglobulin A (IgA) levels were found to be associated with carriage of Blastocystis but not symptoms (p<0.001). All IBS patients, but more commonly the Blastocystis-positive group, had significantly greater reactivity with Blastocystis proteins of 27kDa (likely a cysteine protease), 50 and 75-95 kDa molecular weight. Specific anti-protease antibodies were shown to react with Blastocystis proteins suggesting that metalloproteases, as well as cysteine proteases, may be important Blastocystis antigens. The faecal microbiota (FM) is recognised to influence our health and is altered in IBS. We examined the FM in 40 patients with IBS (26 positive, 14 negative for Blastocystis) and 57 healthy controls (42 positive and 15 negative for Blastocystis). Characteristic changes in the FM profile of IBS patients at a phylum and genus level were observed. However, when the subgroups were analysed there were significant changes seen at different taxonomic levels between the Blastocystis positive and negative IBS groups. The lower serum IgA level present in Blastocystis carriers was hypothesised to result in changes in the FM that enable Blastocystis to flourish with increased epithelial cell contact. Other previously described host factors, such as cytokine polymorphisms in IBS patients that increase IL-8 production and decrease IL-10, may select subjects for immune activation and symptom development. In summary, Blastocystis spp. remain enigmatic organisms that are difficult to clear from our gastrointestinal tracts. Subtyping has increased our epidemiological understanding and infection within households is almost universal. Host immune

iv responses and FM are significantly different in Blastocystis-positive IBS patients suggesting that this organism may contribute to IBS in some subjects.

v Declaration by author This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my research higher degree candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis.

vi Publications during candidature Peer-reviewed papers

CR Stensvold, HV Smith, R Nagel, KEP Olsen, RJ Traub, 2010 Eradication of Blastocystis carriage with antimicrobials: reality or delusion? Journal of Clinical Gastroenterology 44 (2), 85-90

R Nagel, L Cuttell, CR Stensvold, PC Mills, H Bielefeldt-Ohmann, RJ Traub, 2012. Blastocystis subtypes in symptomatic and asymptomatic family members and pets and response to therapy. Internal Medicine Journal 42 (11), 1187-1195

R Nagel, H Bielefeldt-•Ohmann, RJ Traub, 2014 Clinical pilot study: efficacy of triple antibiotic therapy in Blastocystis positive irritable bowel syndrome patients. Gut pathogens 6 (1), 34

R Nagel, C Gray, H Bielefeldt-Ohmann, RJ Traub, 2015 Features of Blastocystis spp. in xenic culture revealed by deconvolutional microscopy. Parasitology Research 114, 3237-3245

R Nagel, RJ Traub, MMS Kwan, H Bielefeldt-Ohmann 2015 Blastocystis specific serum immunoglobulin in patients with irritable bowel (IBS) syndrome. Parasites and Vectors, 8, 453: 1-13

R Nagel, RJ Traub, MMS Kwan, RJN Allcock, H Bielefeldt-Ohmann 2015 Differences in faecal microbiota profile between Blastocystis-positive and negative Irritable bowel syndrome patients Submitted for publication 2015

vii Publications included in this thesis

1. R Nagel, L Cuttell, CR Stensvold, PC Mills, H Bielefeldt-Ohmann, RJ Traub, 2012. Blastocystis subtypes in symptomatic and asymptomatic family members and pets and response to therapy. Internal Medicine Journal 42 (11), 1187-1195 -incorporated as Chapter 2

Contributor Statement of contribution Nagel, Robyn (candidate) Study design (40%) Recruitment of study participants (100%) Collection of samples (100%) Analysis of results (50%) Wrote paper (80%) Cuttell, Leigh Technical analysis of samples (70%) Stensvold, Christian Rune Study design (5%) Technical contribution to analysis of samples (5%) Mills, Paul C Study design (5%) Bielefeldt-Ohmann, Helle Study design (5%) Edited paper (10%) Traub, Rebecca J Study design (45%) Analysis of results (50%) Technical analysis of samples (25%) Edited paper (10%)

viii 2. R Nagel, H Bielefeldt-•Ohmann, RJ Traub, 2014 Clinical pilot study: efficacy of triple antibiotic therapy in Blastocystis positive irritable bowel syndrome patients. Gut pathogens 6 (1), 34 -incorporated as Chapter 5

Contributor Statement of contribution

Nagel, Robyn (candidate) Study design (100%)

Recruitment of study participants

(100%)

Collection of samples (100%)

Technical analysis of samples (95%)

Analysis of results (70%)

Wrote paper (80%)

Bielefeldt-Ohmann, Helle Analysis of results (15%)

Edited paper (10%)

Traub, Rebecca J Analysis of results (15%)

Technical analysis of samples (5%)

Edited paper (10%)

ix 3. R Nagel, C Gray, H Bielefeldt-Ohmann, RJ Traub, 2015 Features of Blastocystis spp. in xenic culture revealed by deconvolutional microscopy. Parasitology Research 114; 3237-3245

-incorporated as Chapter 4

Contributor Statement of contribution Nagel, Robyn (candidate) Study design (80%) Recruitment of study participants (100%) Collection of samples (100%) Technical analysis of samples (80%) Analysis of results (70%) Wrote paper (85%) Gray, Christian Technical analysis of samples (20%) Edited paper (5%) Bielefeldt-Ohmann, Helle Analysis of results (20%) Edited paper (5%) Traub, Rebecca J Study design (20%) Analysis of results (10%) Edited paper (5%)

x 4. R Nagel, RJ Traub, MMS Kwan, H Bielefeldt-Ohmann 2015 Blastocystis specific serum immunoglobulin in patients with irritable bowel syndrome (IBS) versus healthy controls Parasites and Vectors, 8, 453: 1-13

-incorporated as Chapter 6

Contributor Statement of contribution Nagel, Robyn (candidate) Study design (40%) Recruitment of study participants (100%) Collection of samples (100%) Technical analysis of samples (90%) Analysis of results (75%) Statistical analysis of study (40%) Wrote paper (70%) Kwan, Marcella Statistical analysis of study (55%) Edited paper (5%) Traub, Rebecca J Study design (20%) Analysis of results (5%) Statistical analysis of results (5%) Edited paper (5%) Bielefeldt-Ohmann, Helle Study design (40%) Technical analysis of samples (10%) Analysis of results (20%) Edited paper (20%)

xi 5. R Nagel, R J Traub, MMS Kwan, RJN Allcock, H Bielelfeldt-Ohmann 2015 Differences in faecal microbiota profiles between Blastocystis-positive and negative Irritable Bowel Syndrome patients Submitted for publication September 2015 -incorporated as Chapter 7

Contributor Statement of contribution Nagel, Robyn (candidate) Study design (80%) Recruitment of study participants (100%) Collection of samples (100%) Metagenomic analysis of samples (50%) Analysis of results (80%) Wrote paper (70%) Traub, Rebecca J Study design (10%) Edited paper (5%) Kwan, Marcella Statistical analysis of results (100%) Edited paper (5%) Allcock, Richard Metagenomic analysis (50%) Edited paper (5%) Bielefeldt-Ohmann, Helle Study design (10%) Analysis of results (20%) Edited paper (15%)

xii Contributions by others to the thesis

No contributions by others

Statement of parts of the thesis submitted to qualify for the award of another degree

None

xiii Acknowledgements

This journey would not have been possible without the support of my family, supervisors, colleagues and patients. My children Caitlin, David and Jonathon have all in their own unique ways given me support for this ambitious project. My mother Elaine has always encouraged academic endeavours and provided positive directional guidance during my uncertain times. My loving husband Peter has cooked a thousand family dinners, taken up the family duty slack when I have been physically and mentally absent and coped bravely with his lonely evenings. This thesis would not have been possible without his great generosity of spirit. My supervisors, Associate Professor Rebecca Traub, Dr Helle Bielefeldt-Ohmann, Professor Paul Mills, Dr Sze Fui Hii and Dr Marcella Kwan have been encouraging from the start and continued that guidance and support over the five years. My especial thanks to Rebecca Traub and Helle Bielefeldt-Ohmann who both gave me patient, optimistic and knowledgeable help when the going got tough. Dr Christian Gray contributed his considerable expertise to the deconvolutional microscope work, Dr Linda Dunn from the Queensland University of Technology valuable technical advice and axenic cultures of Blastocystis spp. and Dr Marcella Kwan vital statistical input. I would like to acknowledge the many Day Endoscopy Unit colleagues, Ellen Jenson and Ros Young from Toowoomba Gastroenterology Clinic, UQ laboratory staff members and post-graduate students, especially Dr Lyn Knott, Associate Professor Malcolm Jones and Dr Wenqi Wang, of the School of Veterinary Science at Gatton, University of Queensland, Dr Kevin Tan from the National University of Singapore, Professor Richard Allcock from the University of Western Australia and Dr Jenny Robson, microbiologist with Sullivan and Nicolaides, who generously donated samples, scientific and academic advice. I am also appreciative of the financial support given to me by the Royal Australasian College of Physicians in the form of the Murray-Wills Rural Fellowship for rural physicians in 2012. None of this would have been possible without the continued interest of my patients who were all keen and willing to collect faecal samples and support this research into Blastocystis sp. and irritable bowel syndrome. Finally a last tribute must be given to the Blastocystis parasite itself. Secrets endowed by the ancients were defended with guile and vigour by these fascinating organisms.

xiv Keywords

Blastocystis, irritable bowel syndrome, faecal microbiome, antibiotic therapy, zoonosis,

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code 110803: Human medical parasitology, 40%

ANZSRC code 920105: Human digestive system disorders, 30%

ANZSRC code 110705: Human medical Immunology and Immunochemistry, 20%

ANZSRC code 111706: Human Epidemiology, 10%

Fields of research (Giacometti et al.) Classification

FoR code: Division 11: Medical and Health Sciences 1107: Immunology, 30% 1108: Medical microbiology, 50% 1117: Public Health and health services, 20%

xv Table of Contents Abstract...... iii Declaration by author………………………………………………………………………….…..vi Publications during candidature……………………………………………………………...….vii Publications included in this thesis…………………………………………………………...... viii Contributions by others to the thesis…………………………………………...... xiii Statement of parts of the thesis submitted to qualify for the award of another degree.…..xiii Acknowledgements……………………………………………………..………………………..xiv Keywords…………………………………………………………………………………………..xv Australian and New Zealand Standard Research Classifications (ANZSRC)……………...xv Fields of Research Classification………………………………………………….………….…xv

1 Review of literature……………………………………………………………………….2 1.1 Why investigate Blastocystis species?...... 3 1.2 Introduction to the faecal microbiota...... 3 1.2.1 Defining the faecal microbiota…………………………………………………….3 1.2.2 Function of the faecal microbiota…………………………………………………4 1.2.3 Interaction with the host……………………………………………………………5 1.2.4 Identification of the faecal microbiome………………………………………..…5 1.2.5 Factors influencing the faecal microbiota………………………………………..5 1.3 Irritable Bowel Syndrome...... 9 1.3.1 Introduction………………………………………………………………………….9 1.3.2 Epidemiology………………………………………………………………………..9 1.3.3 Costs of irritable bowel syndrome………………………………………………10 1.3.4 Pathogenesis of irritable bowel syndrome…………………………………....10 1.3.5 Diagnosis…………………………………………………………………………..11 1.3.6 Therapy…………………………………………………………………………….12 1.4 Introduction to Blastocystis species………………………………………………….…12 1.4.1 What are they?...... 12 1.4.2 Where are they found?...... 14 1.4.3 Prevalence…………………………………………………………………………14 1.4.4 Routes of transmission…………………………………………………………...15 1.4.5 Zoonotic potential…………………………………………………………………16 1.5 Morphology of Blastocystis species…………………………………………………….18 1.5.1 Introduction………………………………………………………………………..18

xvi 1.5.2 Vacuolated, Granular, Multi-Vacuolar and Avacuolar forms…………………19 1.5.3 Amoebic form……………………………………………………………………..21 1.5.4 Cysts……………………………………………………………………………….21 1.5.5 Functional components…………………………………………………………..22 1.5.6 Life cycle…………………………………………………………………………...25 1.5.7 Modes of ……………………………………………………………25 1.6 Laboratory Diagnosis……………………………………………………………………..27 1.6.1 Stool samples and microscopy………………………………………………….27 1.6.2 Stool samples and culture...... 27 1.6.3 Immunological techniques……………………………………………………….28 1.7 Molecular Techniques and Subtyping…………………………………………………..29 1.7.1 Subtyping the organism using the ribosomal RNA gene……………………..29 1.7.2 Epidemiology of Blastocystis subtypes…………………………………………30 1.7.3 Multilocus sequence tagging and clades……………………………………….31 1.7.4 Whole genome sequencing……………………………………………………...32 1.8 Clinical Association of Blastocystis in Humans………………………………………..32 1.8.1 Introduction………………………………………………………………………..32 1.8.2 Blastocystis and Irritable Bowel Syndrome……………………………………33 1.8.3 Immunocompromised patients ………………………………………………….33 1.8.4 Cancer patients……………………………………………………………………34 1.8.5 Inflammatory Bowel Disease…………………………………………………….34 1.9 Animal studies…………………………………………………………………………….34 1.10 Possible mechanisms of pathogenicity…………………………………………………35 1.10.1 Introduction………………………………………………………………………..35 1.10.2 Intrinsic factors…………………………………………………………………….35 1.10.3 Host factors………………………………………………………………………..38 1.10.4 Pathogenic passengers…………………………………………………………..38 1.11 Antimicrobial Therapy and Blastocystis.………………………………………………..40 1.11.1 Current therapeutic guidelines…………………………………………………..40 1.11.2 In vitro antibiotic sensitivity………………………………………………………40 1.11.3 In vivo antibiotic sensitivity……………………………………………………….43 1.11.4 Therapeutic challenges…………………………………………………………..45 1.12 Aims and objectives of this thesis...... 46 1.12.1 The Hypothesis……………………………………………………………………46 1.12.2 Investigative approaches and objectives………………………………………46

xvii 2 Blastocystis subtypes in symptomatic and asymptomatic family members and pets and response to therapy……………………………………………………50 2.1 Abstract…………………………………………………………………………………….51 2.2 Introduction………………………………………………………………………………..52 2.3 Methods……………………………………………………………………………………52 2.3.1 Study outline………………………………………………………………………52 2.3.2 Inclusion protocol…………………………………………………………………53 2.3.3 Study protocol……………………………………………………………………..53 2.3.4 Diagnostic methods………………………………………………………………54 2.3.5 Ethics approval……………………………………………………………………55 2.4 Results……………………………………………………………………………………..55 2.4.1 Clinical characteristics of patients and household members………………...55 2.4.2 Endoscopy…………………………………………………………………………57 2.4.3 Therapy…………………………………………………………………………….57 2.4.4 Faecal smears and culture in patients………………………………………….57 2.4.5 PCR and genetic subtyping of Blastocystis in patients……………………….58 2.4.6 Faecal smears, cultures and PCR subtyping in HC…………………………..59 2.4.7 Faecal smears, cultures and PCR subtyping in animal contacts……………60 2.4.8 Clinical follow-up at 6 weeks…………………………………………………….60 2.5 Discussion…………………………………………………………………………………61 2.6 Conclusion…………………………………………………………………………………63 2.7 Acknowledgements……………………………………………………………………….63

3 Axenic culture of Blastocystis...... ………………………………………………….65 3.1 Abstract…………………………………………………………………………………….66 3.2 Introduction to axenic culture……………………………………………………………66 3.2.1 Why attempt axenic culture?...... 66 3.2.2 Early experience with axenic culture using antibiotics………………………..67 3.2.3 Physical separation of contaminating microorganisms……………………….67 3.2.4 Agar colonies……………………………………………………………………...68 3.2.5 Requirements for successful axenisation………………………………………69 3.3 Methods……………………………………………………………………………………69 3.3.1 Sample selection and management…………………………………………….69 3.3.2 Preliminary subculture one………………………………………………………70 3.3.3 Preliminary subculture two……………………………………………………….70

xviii 3.3.4 Antibiotic supplementation……………………………………………………….70 3.3.5 Monitoring the subcultures……………………………………………………….70 3.3.6 Solid agar culture…………………………………………………………………70 3.3.7 Semi-solid agar culture…………………………………………………………..71 3.4 Results……………………………………………………………………………………..71 3.5 Discussion…………………………………………………………………………………76 3.6 Conclusions………………………………………………………………………………..77

4 Features of Blastocystis spp. in xenic cultures revealed by deconvolutional microscopy……………………………………………………………………………….79 4.1 Abstract………………………………………………………………………………….…80 4.2 Introduction………………………………………………………………………………..80 4.3 Materials and methods…………………………………………………………………...83 4.3.1 Sample preparation………………………………………………………………83 4.3.2 Slide preparation………………………………………………………………….83 4.3.3 Antibody staining………………………………………………………………….84 4.4 Results and discussion…………………………………………………………………..84 4.4.1 Auto- and immunofluorescence…………………………………………………84 4.4.2 Cell granules and time-lapse microscopy………………………………………88 4.4.3 Other surface changes observed over time……………………………...... 93 4.5 Summary…………………………………………………………………………………..98 4.6 Acknowledgements…………………………………………………………………….…98

5 Clinical pilot study: efficacy of triple antibiotic therapy in Blastocystis positive irritable bowel syndrome patients……………………………………….101 5.1 Abstract……………………………………………………………………………….….102 5.2 Background………………………………………………………………………………103 5.3 Materials and methods………………………………………………………………….105 5.3.1 Study outline…………………………………………………………………..…105 5.3.2 Inclusion protocol……………………………………………………………..…105 5.3.3 Exclusion protocol……………………………………………………………….106 5.3.4 Study protocol………………………………………………………………..….106 5.3.5 Diagnostic methods……………………………………………………………..106 5.3.6 Parasitological methods……………………………………………………...…107 5.3.7 Molecular methods………………………………………………………………107

xix 5.3.8 Ethics approval, statistics………………………………………………………108 5.4 Results……………………………………………………………………………………108 5.4.1 Clinical characteristics of the study group…………………………………….108 5.4.2 Antibiotic therapy and clinical follow up……………………………………….109 5.4.3 Faecal smears, cultures and PCR results……………………………………109 5.4.4 Clearance and clinical follow up……………………………………………….111 5.5 Discussion………………………………………………………………………………..113 5.6 Conclusions………………………………………………………………………………119 5.7 Acknowledgements………………………………………………………………..……120

6 Blastocystis specific serum immunoglobulin in patients with irritable bowel syndrome (IBS) versus healthy controls………………………………………….122 6.1 Abstract…………………………………………………………………………………..123 6.2 Introduction………………………………………………………………………………124 6.3 Methods………………………………………………………………………………..…127 6.3.1 Study outline……………………………………………………………………..127 6.3.2 Diagnosis of Blastocystis infection…………………………………………….128 6.3.3 Serum Immunoglobulin A………………………………………………………128 6.3.4 Western Blotting…………………………………………………………………128 6.3.5 Zymograms………………………………………………………………………130 6.3.6 Statistical Analysis………………………………………………………………131 6.4 Results……………………………………………………………………………………131 6.4.1 Study subjects…………………………………………………………………...131 6.4.2 Serum IgA levels………………………………………………………………...133 6.4.3 Blastocystis-specific serum IgG antibody bands…………………………….133 6.4.4 Further analysis of the Blastocystis antigens………………………………...140 6.5 Discussion……………………………………………………………………………….146 6.6 Conclusions……………………………………………………………………………..150 6.7 Acknowledgements……………………………………………………………………..151

7 Differences in faecal microbiota profiles between Blastocystis-positive and negative Irritable Bowel syndrome patients………………………………………153 7.1 Abstract…………………………………………………………………………………..154 7.2 Background………………………………………………………………………………155 7.3 Methods…………………………………………………………………………………..157

xx 7.3.1 Study outline……………………………………………………………………..157 7.3.2 Inclusion protocol………………………………………………………………..158 7.3.3 Exclusion protocol……………………………………………………………….158 7.3.4 Diagnostic methods……………………………………………………………..159 7.3.4.1 Metagenomic anlaysis…………………………………………………159 7.3.4.2 Analysis of sequences…………………………………………………159 7.3.5 Statistical analysis……………………………………………………………….160 7.4 Results……………………………………………………………………………………161 7.4.1 Subjects………………………………………………………………………….161 7.4.2 Bacterial phyla seen in the study participants………………………………..162 7.4.3 Comparison of bacterial profiles in subjects with and without IBS…………162 7.4.4 Comparison of bacterial phyla across the four clinical subgroups…………166 7.4.5 Comparison of bacterial genera across the four clinical subgroups……….166 7.5 Discussion………………………………………………………………………………..172 7.6 Summary…………………………………………………………………………………176

8 Discussions and Conclusions………………………………………………………178 8.1 Revisiting the aims of this thesis………………………………………………………178 8.1.1 Aim of the thesis and interest in the organism.………………………………178 8.1.2 Irritable bowel syndrome today………………………………………………...179 8.1.3 Investigating the clinical relevance of Blastocystis spp……………………..180 8.2 Life cycle of Blastocystis spp…………………………………………………………..181 8.2.1 Axenic culture (Chapter 3)……………………………………………………..181 8.2.2 Features of Blastocystis spp. in xenic culture revealed by deconvolutional microscopy (Chapter 4)…………………………………………………………182 8.3 Subtyping and epidemiology…………………………………………………………...184 8.3.1 Expectations of subtyping………….…………………………………………..184 8.3.2 Epidemiological advances……………………………………………………...185 8.3.3 Zoonotic potential………………………………………………………………..185 8.3.4 Subtypes and pathogenicity……………………………………………………185 8.4 Subtyping and transmission within households……………………………………...186 8.4.1 Subtyping for diagnosis of infection (Chapters 2,5,6,7)…………………….186 8.4.2 Transmission of infection within households (Chapter 2)…………………..186 8.4.3 Subtypes and pathogenicity (Chapters 2,5,6,7)……………………………..186 8.5 Antimicrobial therapy for Blastocystis infection………………………………………187

xxi 8.5.1 Eradication is problematic………………………………………………………187 8.5.2 Why bother?...... 187 8.5.3 Clinical pilot study: efficacy of monotherapy with metronidazole or trimethoprim/sulfamethoxazole in Blastocystis-positive IBS patients (Chapter 2)……………………………………………………………………….188 8.5.4 Clinical pilot study: efficacy of triple antibiotic therapy in Blastocystis- positive IBS patients (Chapter 5)………………………………………………188 8.6 Host responses to Blastocystis infection……………………………………………..189 8.6.1 Blastocystis and the host immune system……………………………………189 8.6.2 Serum immunoglobulin A in patients with IBS versus healthy controls (HC) (Chapter 6)……………………………………………………………………….190 8.6.3 Blastocystis specific serum immunoglobulin responses in patients with IBS compared to HC (Chapter 6)…………………………………………………...190 8.6.4 Identifying serum immunoglobulin responses to Blastocystis antigens (Chapter 6)……………………………………………………………………….191 8.7 The faecal microbiome, IBS and Blastocystis carriage……………………………..192 8.7.1 Introduction to the faecal microbiome, IBS and Blastocystis……………….192 8.7.2 Differences in faecal microbiota profiles between Blastocystis-positive and negative IBS patients (Chapter 7)……………………………………………..192 8.8 Limitations of studies……………………………………………………………………193 8.8.1 Acquisition of patients for clinical studies……………………………………..193 8.8.2 Axenic culture failure (Chapter 3)……………………………………………...193 8.9 A unifying hypothesis and future directions…………………………………………..194 8.9.1 A unifying hypothesis……………………………………………………………194 8.9.2 Future directions…………………………………………………………………194

Bibliography…………………………………………………………………………………….198

9 Appendices……………………………………………………………………………..225

xxii List of tables

Table Table Legend Page

1.1 Human bacterial faecal microbiome in health and disease 7

1.2 Rome III diagnostic criteria for irritable bowel syndrome 9

1.3 Blastocystis protozoan characteristics 13

1.4 Properties of Blastocystis cysts 22

1.5 Distribution of Blastocystis subtypes infecting humans in 31 different geographic regions

1.6 Blastocystis subtype variation in irritable bowel syndrome 36

1.7 Activity of drugs against Blastocystis axenic stock cultures 42

1.8 In vivo efficacy of antimicrobials against Blastocystis 44

2.1 Clinical profile of symptomatic Blastocystis patients 56

2.2 Blastocystis faecal smear and culture results in patients 58 Blastocystis subtype results in symptomatic patients pre- and post- 2.3 therapy 59

2.4 Blastocystis subtype results in patients, household contacts and animals 60 Blastocystis cell counts in fresh faecal cultures derived from six patients 3.1 with gastrointestinal symptoms, subcultured every 2-3 days, in 74 Jones/Horse serum 10% seven days, then IMDM/Horse serum 10% for following 11 weeks

xxiii List of tables

Blastocystis faecal microscopy, XIVC and PCR results before and after 5.1 triple antibiotic therapy 110

5.2 Clinical characteristics and outcome after triple antibiotic therapy 112 Demographic and epidemiological characteristics of subgroups 6.1 132 Serum antibody reactions to Blastocystis antigens detected by Western 6.2 immunoblotting in all subjects infected with different subtypes of 132 Blastocystis Serum antibody reactions to Blastocystis antigens detected by Western 6.3 immunoblotting in different clinical groups 134 Comparison of presence of Blastocystis and/or gastrointestinal symptoms 6.4 to specific sized antibody bands directed against Blastocystis proteins 135 using western immunoblotting

7.1 Characteristics of clinical subgroups 161 7.2a Abundance of bacterial phyla seen in clinical subgroups (%) 163 7.2b Firmicutes:Bacteroidetes Ratio in clinical subgroups 164

7.3 Comparison of bacterial profiles in subjects with and without IBS 167 Comparison of abundance of selected bacterial species across the four 7.4 clinical subgroups 169

xxiv List of figures

Figure Figure legend Page

1.1 Taxonomic classification of 13

1.2 Blastocystis subtypes in animals 17

1.3 Light microscopy of different morphological forms of Blastocystis 20 1.3.1 Vacuolar and granular form 1.3.2 Multivacuolar form 1.3.3 Amoeboid forms 1.3.4 Cysts

1.4 Electron microscopy of functional components of Blastocystis 23 1.4.1 Surface coat 1.4.2 Outer membrane 1.4.3 Mitochondria-like organelles 1.4.4 Outer plasma membrane 1.4.5 Phagocytosis of bacterium 1.4.6 Different morphological forms 1.4.7 Nucleus

1.5 Proposed life cycle 26

1.6 Barcode region of Blastocystis 18S ribosomal RNA gene identified by 30 different primers

1.7 Cysteine proteases and possible virulence mechanisms 38

3.1 Scanning electromicrograph of Blastocystis colony grown on solid agar 68 Blastocystis colonies grown on solid agar 17/8/12 School of Veterinary 3.2 Science, UQ 72

xxv List of Figures

3.3 Light microscopy of Blastocystis colony grown on solid agar 17/08/12 72 School of Veterinary Science, University Queensland

3.4 Light microscopy of Blastocystis colony grown in soft agar 17/07/12 73 School of Veterinary Science, University Queensland

3.5 Blastocystis cell counts in fresh faecal cultures derived from six patients 75 with gastrointestinal symptoms, subcultured every 2-3 days, in Jones/Horse serum 10% for 7 days, then IMDM/Horse serum 10%

4.1 Morphological forms of Blastocystis spp. in xenic culture stained with 82 acridine orange

4.2 Autofluoresence of different morphological forms of Blastocystis spp. 85

4.3 Different morphological forms of Blastocystis spp. stained with 86 Blastofluor

4.4 Vacuolated form of Blastocystis spp., 15-min time-lapse sequence 89

4.5 Xenic culture of Blastocystis granular forms, time-lapse images over 24 h 90

4.6 Blastocystis with moving central granule, sequential images over 15-min 91 time intervals

4.7 Granules exiting Blastocystis amoebic form in xenic culture, time-lapse 92 images taken over 24 h

xxvi List of Figures

4.8 Blastocystis form in xenic culture showing pores on surface 94

4.9 Xenic culture of Blastocystis vacuolated form with extrusion, time-lapse 95 images

4.10 Fluorescent microscopy of Blastocystis forms stained with DAPI 97

6.1 Serum antibodies from Blastocystis positive clinical subgroups IBS-P and 137 HC-P reacting with Blastocystis proteins in Western blot Serum antibodies from Blastocystis positive clinical subgroups IBS-P and 138 6.2 HC-P and Blastofluor Ab reacting with Blastocystis proteins in Western blot Serum antibodies from Blastocystis negative clinical subgroups IBS-N 139 6.3 and HC-N reacting with Blastocystis proteins in Western blot Serum antibodies from IBS-P clinical subgroup and monoclonal 142 6.4 antibodies MAb1D5 and MAb5 reacting with Blastocystis proteins in Western blot Serum antibodies in IBS-P clinical subgroup and MAb anti-proMMP-9 143 6.5 reacting with Blastocystis proteins in Western blot Anti-proMMp-9, MAb1D5, MAb5 reacting with Blastocystis proteins in a 144 6.S1 Western blot 145 6.S2 Zymogram of Blastocystis WR1 proteins without addition of EDTA

7.1 Comparison of bacterial phyla profiles in subjects with and without IBS 165

7.2a Comparison of major bacterial phyla across the four clinical subgroups 171

7.2b Comparison of minor bacterial phyla across the four clinical subgroups 171

xxvii List of appendices

Appendix Legend Page number

9.1 Jones medium 226

9.2 Lockes medium 226

9.3 Biphasic solid and liquid phase medium 226 Video 1: Blastocystis with moving central granule, sequential images over 9.4 15 minute time intervals Available online with Parasitol Res (doi:10.1007/s00436-015-4540-x) Video 2: Granules exiting Blastocystis amoebic form in xenic culture, time 9.5 lapse images taken over 24 hours Available online with Parasitol Res (doi:10.1007/s00436-015-4540-x) Video 3: Blastocystis form in xenic culture showing pores on the surface 9.6 Available online with Parasitol Res (doi:10.1007/s00436-015-4540-x) Video 4: Xenic culture of Blastocystis vacuolated form with extrusion 9.7 Available online with Parasitol Res (doi:10.1007/s00436-015-4540-x) 9.8 Supplementary File 5.1: Clinical Diary of patients 227

9.9 Copy of copyright permission (Nagel et al 2014) 228

xxviii List of abbreviations

Abbreviations used commonly throughout the thesis are listed below. All abbreviations used are defined in parentheses after their use in each chapter. ______C degrees Celsius g/ml microgram/millilitre mol/L micromolar AF amoebic form AGL Animal Genetics Laboratory AVF avacuolated form AXC axenic culture B-D Boeck-Drbohlav culture medium CI confidence intervals CNS central nervous system CV central Cy5 cyanine 5 DAPI 4’,6’-diamindino-2-phenylindole D-IBS diarrhoea-predominant irritable bowel syndrome DNA deoxyribonucleic acid dNTP deoxyribonucleotide triphosphate EDTA ethylenediaminetetraacetic acid ELISA enzyme-linked immune sorbent assay EM electron microscopy FASTA/q acronym for text based format for representing nucleotide or peptide sequences/ and its corresponding quality score F:B ratio Firmicutes:Bacteroidetes ratio FECT formalin ether concentration technique FE-EM freeze-etch electron microscopy FITC fluorescein isothiocyanate FM faecal microbiota FODMAP fermentable oligodisaccharides monosaccharides and polyols g Gravitational (g) force GAF green autofluorescence GALT Gut associated lymphoid tissue GF granular form xxix List of abbreviations

HBD human beta defensin HC household contact or healthy control HeC healthy control HC-P healthy control positive for Blastocystis HC-N healthy control negative for Blastocystis HIV human immunodeficiency virus HM haematological malignancy hr(s) hour(s) HS horse serum IBD Inflammatory Bowel disease IBS Irritable bowel syndrome IBS-C Constipation-predominant IBS IBS-D Diarrhoea-predominant IBS IBS-M Mixed type IBS IBS-N Irritable bowel syndrome negative for Blastocystis IBS-P Irritable bowel syndrome positive for Blastocystis 3 ID50 concentration of antibiotic required to inhibit 50% uptake of H- hypoxzanthine IgA immunoglobulin class A IgG immunoglobulin class G IgM immunoglobulin class M IL interleukin IMDM Iscove’s Modified Dulbecco’s medium ISP ion sphere particles kDa kiloDalton M cells microfold cells MAb monoclonal antibody MAM mucosa associated microbiota mg/kg milligram/kilogram

MgCl2 magnesium chloride MIFT merthiolate iodine formaldehyde concentration min minutes MLO mitochondria-like organelle MLST multi locus restriction fragment

xxx List of abbreviations

MMP promatrix metalloprotease 9 mmol/L millimolar MVF multivacuolated form of Blastocystis MW molecular weight MZ metronidazole NA not available NF-B nuclear factor kappa-light-chain-enhancer of activated B cells O/HPF organisms per high power field PBS phosphate buffered saline PBS/T phosphate buffered saline/Tween 20 PCR polymerase chain reaction QIIME Quantitative Insights into Microbial Ecology RFLP restriction-fragment-length-polymorphism rRNA ribosomal ribonucleic acid rDNA DNA coding for ribosomal ribonucleic acid S subunit SBP symptomatic patient positive for Blastocystis SCFA short chain fatty acid SDS-PAGE sodium dodecyl sulphate polyacrylamide gel electrophoresis sec seconds SEM scanning electron microscopy sp./spp. species (singular/plural) SPPS acronym for a particular statistical software package SSU short subunit ST subtype STS subtype specific SVS-UQ School of Veterinary Science, University Queensland TAB triple antibiotic therapy with diloxanide furoate, trimethoprim/sulfamethoxazole, secnidazole TEM transmission electron microscopy TMT/SMX trimethoprim/sulfamethoxazole TNF- tumour necrosis factor- TRITC tetramethylrhodamine TS trichrome stain xxxi List of abbreviations

VF vacuolated form v/v volume concentration XIVC xenic in vitro culture

xxxii

ii

1 Review of literature

2 Review of Literature 1.1 Why investigate Blastocystis species? Blastocystis sp. are often the most common faecal parasite identified worldwide and infection has been linked to human disease, most specifically chronic, watery diarrhoea, bloating and abdominal pain (Tan et al. 2010), typically irritable bowel syndrome (IBS) symptoms. The carriage rate in some developing countries has been reported to exceed 60% of the community (Tan 2008). Faecal-oral transmission of the parasite likely occurs via ingestion of cysts indirectly in contaminated water sources or directly by contact with humans and domestic and non-domestic animals (Tan 2008). The parasite was under-diagnosed in commercial and research laboratory testing until the last 6 years, because identification of the parasite in wet faecal smears is unreliable. Molecular techniques have more than doubled the diagnostic accuracy (Stensvold et al. 2007a). The organism displays diverse morphological forms and the relationship and contribution of these forms in the life cycle of the parasite is not understood. Attempts at eradication with antimicrobials have a high failure rate (Stensvold et al. 2010). There are large gaps in our knowledge with regard to this parasite that are important to address, most notably whether the organism causes disease in humans or is simply an innocent bystander. Understanding how this organism contributes to the human faecal microbiome and any mechanism of pathogenicity of this apparently non-invasive organism may offer insights into the mechanism of mucosal damage in other gastrointestinal diseases such as Inflammatory Bowel Disease (IBD). If Blastocystis organisms are pathogenic to humans there are implications for the monitoring of safe water supplies and hygiene guidelines for humans and animal handling. It will be difficult to suggest preventative strategies or specific therapy until the life cycle of this parasite is understood.

1. 2 Introduction to the faecal microbiota 1.2.1 Defining the faecal microbiota The human gut is sterile at birth and during early infancy is colonised by organisms originating from the mother, diet and environment. The human

3 adult is host to 100 trillion species of commensal organisms that live in facilitative harmony with the 10 trillion cells in our body, the “faecal microbiota” (Bonfrate et al. 2013). Over 90% of the faecal mass is comprised of microbes namely bacteria (93%), viruses (5.8%), archaea (0.8%) and (0.5%)(Arumugam et al. 2011). Using molecular techniques thirty-seven eukaryotic species were identified in the faeces of a healthy adult including Blastocystis sp., 18 plant species and 18 fungal species (Gouba et al. 2013). However the vast majority of the FM (Garduno et al. 2002) comprises bacterial species that have not been cultured. The microbial density per gram of luminal content changes along the length of the gut from 101 in the relatively sterile stomach to 104 organisms in the jejunum, 108 in the ileum, 1010 in the colon and 1012 in the faeces with anaerobes predominating in the colon (Bonfrate et al. 2013). The bacterial component of the FM has received most attention. The contribution to the FM is small but most often includes Blastocystis spp. A recent study of 105 healthy adults in Ireland showed the prevalence of Blastocystis carriage to be 56%, with diverse subtypes and with stable carriage seen over 6-10 years (Scanlan et al. 2014). Blastocystis spp. in many people are one of the components of a healthy FM.

1.2.2 Function of the faecal microbiota This intestinal microbiota is estimated to contain 3.3 million genes (the microbiome) vastly outnumbering our 20,000 human genes. The faecal microbiota/microbiome is increasingly recognised as an additional significant human organ with important neuroendocrine, metabolic, nutritional and protective functions (Albenberg & Wu 2014). The faecal microbiota produces micronutrients vitamin B and K and short chain fatty acids (SCFA) that are a primary energy source for colonic mucosal cells. The bacteria facilitate fermentation of food for host metabolism, produce neurotransmitters that are absorbed systemically, conduct immune priming, including differentiation of T helper cells, in the host and protect against other luminal pathogens (Furusawa et al. 2014).

4 1.2.3 Interaction with the host Normally a mucus layer ensures that there is little contact between the luminal biota and the host gut epithelial cells. The mucous glycoprotein layer in the small bowel is thin and discontinuous in contrast to the sterile, bi- layered continuous thick mucus layer in the colon (Merga et al. 2014). Interaction between luminal bacteria and the host may occur directly with the epithelial cell and its glycocalyx, with the tight intercellular junctions between these cells or with Peyer’s patches comprising lymphoid cells topped with a dome epithelium of microfold (M) cells that are adapted for sampling/immune presentation of luminal antigens. The luminal microbiota may be very different to the mucosa-associated bacteria.

1.2.4 Identification of the faecal microbiota Metagenomic analysis using culture-independent high throughput 16 short subunit (SSU) ribosomal ribonucleic acid (rRNA) quantitative gene sequencing and microarrays, has recently allowed genetic insights into faecal bacterial populations. This technology allows the bacteria to be classified according to Phylum, Class, Order, Genus and Species. Of the fifty bacterial phyla humans host ten are found in the gut, and 90% of the colonic microbiota consists of two dominant phyla, namely Bacteroidetes and Firmicutes with great individual variability seen at species and strain level. Three “enterotypes” based on genus have been described predominant in Bacteroides (of Bacteroidetes) (Type 1), Prevotella (of Bacteroidetes)(Type 2) and Ruminococcus (of Firmicutes)(Type 3)(Arumugam et al. 2011) although in reality graduated differences between these three groups are likely to be seen across the human population.

1.2.5 Factors influencing the faecal microbiota In the newborn infant the gut microbiota is aerobic, but within days becomes anaerobic, and within 2 years is relatively stable till advanced age (Arrieta et al. 2014). Some of the factors that may change the human faecal microbiota include mode of delivery (vaginal versus Caesarean), feeding patterns in early infancy (breast feeding duration, time of introduction of foods), immunisation, antibiotic usage, sanitation (Arrieta et al. 2014), gender

5 (Aguirre de Carcer et al. 2011) and long term dietary choices. Prevotella species dominate in rural Africa where the diet is high in vegetables, fruit and fibre compared to Europe (De Filippo et al. 2010). The faecal microbiome is altered in various gastrointestinal diseases including IBS, IBD and colonic cancer (Table 1.1), raising the question of whether the disease is due to the abnormal microbiota or whether the disease is the end result of an abnormal immune response to normal faecal flora.

6 Table 1.1 Human bacterial faecal microbiome in heath and disease

General Level 1 Level 2 Level 3 Level 4/5/6 Diet, Age,1,2 IBS1 IBD5 description Phylum Class Order Class/Family/Gen Gender,3 (Irritable bowel (Inflammatory us/Species Enterotype4 syndrome) bowel syndrome) (CD:Crohn’s disease UC:ulcerative colitis)

General Neonatal gut sterile  temporal instability Diversity in CD, UC Comment Early bacteria  no. coliforms mucosa assoc aerobes  aerobe:anaerobe ratio Bacteria in CD Gram Positive Actinobacter  Actinobacter IBS Actinobacter Streptomyces Present in soil Corynebacterineae Mycobacterium  M avium in CD and freshwater (produce many bioactive Propionibacterineae molecules such as Actinomycetaceae antibiotics,, Bifidobacteriaceae Bf.catenulation Bifidobacterium as IBS  CD, UC macrolides, teracyclines) (Lactobacillus bifidus) teenagers Coriobacteridae Collinsella IBS Gram Negative Bacteroidetes After age 5 years  Bacteroidetes IBS Bacteriodia Bacteroides Enterotype 1  Baceroides IBS Bacteroides CD, UC Anaerobic Porphyromonas Rod shaped Widely found in Prevotellaceae Prevotella Enterotype 2 Prevotella IBS Prevotella CD, UC No spores environment, soil, sea, Xylanibacter (high fibre diet) guts, skin  M vs Females African children with high fibre diet with Firmicutes, Proteobacteria, lactobacillus vs European children Rikenellaceae Gram Positive Firmicutes  Firmicutes  Firmicutes CD, UC Bacilli (aerobes) Bacillus Strong cell wall (Bacillales Staphylococcus Round or rod Lactobacillales Listeria Often produce Staphylococccacea) Lactobacillus after 5 years age   IBS  CD, UC endospores Clostridia Clostridium C. coccoides  IBS (anaerobes) C. cocleatum  IBS

7 C.difficile  CD, UC Coprococcus  IBS Ruminococcus Enterotype 3  IBS  Ruminocoocus gnavus CD Clostridium iX Faecalobacterium CD, UC prausnitzii (butyrate producer++) Lachnospiraceae or Roseburia in low carbohydrate  CD Clos cluster XIV, b diets Mollicutes Negativicutes Veilonella IBS-C CD deep ulcers Gram Negative Proteobacteria  Proteobacteria IBS  Proteobacteria IBD Acidobacteria Deltaproteobacteria Desulfovibrionales Desulfovibrio  CD, UC Metabolism (purple bacteria capable of Epsiloproteobacteria Campylobacter  CD, epsiloproteoB variable but forming many shapes, outer Helicobacter most membrane of Alphaproteobacteria facultatively lipopolysaccharides) Zetaproteobacteria anaerobic, use Gammaproteobacteri Enterobacteriales Ercherichia photosynthesis, enterobacteria enterobacteria IBS  enterobacteriaceae a many have infants with eg adherent invasive flagellae) necrotising E.Coli in CD enterocolitis Proteus Salmonella Vibrio Yersinia Klebsiella Pasteurellaceaea Haemophilus CD deep ulcers Betaproteabacteria Neisseriales  CD Gram Negative Fusobacteria CD Obligate Leptotrichiaceae Streptobacillus anaerobe Fusobacteriacaea Fusobacterium F. necrogenes Non-spores Asaccharolytic nature Clos.rectum F.ulcerans F.necrophorum Deinococcus-Thermus Mucin loving Verricomicrobia Verrucomicobia Akkermansia  in obesity  CD, UC 5% of normal mucinophila (oral dose increases Bacteria in gut depth mucus layer) 1 (Simren et al. 2013) 2 (Arrieta et al. 2014) 3(Aguirre de Carcer et al. 2011) 4(Berry & Reinisch 2013) 5 (Shanahan 2012)

8

1.3 Irritable Bowel Syndrome 1.3.1 Introduction Irritable bowel syndrome (IBS) is a chronic heterogeneous disorder characterised by a symptom complex of abdominal pain, bloating and variable bowel habit in the absence of morphological, histological or inflammatory diagnostic markers (Drossman 2007). The disorder is divided into three types characterised by the change in bowel habit, namely diarrhoea-predominant (IBS-D), constipation-predominant (IBS-C) or alternating and constipation (IBS-M). The IBS diagnostic criteria have been set out by a committee and the latest guidelines are the Rome III criteria (Drossman 2007) (Table 1.2). Table 1.2 Rome III diagnostic criteria for irritable bowel syndrome

Recurrent abdominal pain or discomfort** at least 3 days per month in the last 3 months, associated with 2 or more of the following: 1. Improvement with defecation 2. Onset associated with a change in frequency of stool 3. Onset associated with a change in form (appearance) of stool

*Criteria fulfilled for the last 3 months with symptom onset at least 6 months prior to diagnosis.

1.3.2 Epidemiology IBS is reported worldwide and affects 3-20% of the population, usually occurring before the age of 45 years and has a female: male predominance of 2:1 (Grundmann & Yoon 2010). IBS prevalence tends to be lower in Asian countries, highest in South America and the exact prevalence in developing countries is difficult to assess (Canavan et al. 2014a). The predominance of different types of IBS varies between countries. Socioeconomic status has no relationship to the prevalence of IBS but IBS is twice as common if there is a prior family history of IBS (Canavan et al. 2014a). Depression, anxiety and insomnia are reported in 30% of IBS patients, more than twice the prevalence seen in the community (Grundmann & Yoon 2010). The onset is often reported after overseas travel (Tuteja et al. 2008) or infectious enteritis (Schwille-Kiuntke et al. 2011). Post-infectious IBS was first described in 1962 (Chaudhary & Truelove 1962) and has been reported to

9 occur following infection with many organisms including Shigella, Campylobacter, Salmonella, viruses and parasites at rates varying from 4- 36% (Schwille-Kiuntke et al. 2011).

1.3.3 Costs of irritable bowel syndrome Although IBS does not increase mortality, patients complain of diminished quality of life and 30% of people will seek medical attention, often at least twice a year for ten to fifteen years (Canavan et al. 2014b). IBS costs include direct medical costs (doctor and hospital visits and over two million prescriptions are reported to be written annually for IBS) and indirect costs related to loss of productivity (absenteeism and presenteeism) and have been estimated to amount to $1,350 million annually in the United States of America (Canavan et al. 2014b).

1.3.3 Pathogenesis of irritable bowel syndrome The pathophysiology of this disease is unknown and a number of different physiological aspects have been explored to investigate this disease. IBS links to the faecal microbiota have been suggested not only because post-infectious IBS is a common occurrence but because antibiotic and FM transplantation has been reported to be useful therapy for IBS (Kennedy et al. 2014). IBS is reported to be associated with small bowel bacterial overgrowth, increased temporal instability, increased aerobe:anaerobe ratio and an increased Firmicutes:Bacteroidetes phylum ratio in the FM (Simren et al. 2013). Various other differences at a species level have also been reported in IBS (Table 1.1), including an increase in Enterobacteriaea spp. (Proteobacteria phylum) and a decrease in Faecalobacterium prausnitzii spp (Firmicutes phylum) (Carroll et al. 2012), Lactobacillus and Bifidobacterium spp. (Firmicutes and Actinobacter genus respectively) (Malinen et al. 2005). A recent meta-analysis has shown that IBS patients are 2.3 times more likely to be infected with Blastocystis spp. and healthy male carriers of Blastocystis were found to have decreased numbers of Faecalibacterium prausnitzii (Nourrisson et al. 2014). Abdominal hypersensitivity (Ritchie 1973) and altered gut motility are typical of IBS and these symptoms may be due to a dysregulated gut-brain

10 axis involving the enteric, autonomic and central nervous system (CNS) (Posserud et al. 2006). The communication from the central nervous system to the gut wall is bidirectional and influenced by neural, hormonal and immunological factors. A number of studies have reported findings that suggest the FM is capable of influencing the CNS (Grenham et al. 2011). The enteric nervous system displays increased numbers of mast cells secreting serotonin adjacent to the dendritic cells in IBS patients (Matricon et al. 2012). Low level inflammation and immune activation is also reported in the gastrointestinal mucosa of patients with IBS, particularly increased numbers of mucosal mast cells and increased intestinal permeability (Chadwick et al. 2002). Mast cell densities can be 50-100% higher, particularly in the ileum, caecum and colon of IBS patients compared to controls (Matricon et al. 2012) and higher intraluminal colonic levels of tryptase and histamine are seen in the IBS patients. Pro-inflammatory cytokines, including interleukin (IL)-6, IL-8 and tumour necrosis factor- (TNF-α), have been reported to have higher levels in the plasma and peripheral blood mononuclear cells of IBS patients (Beatty et al. 2014). Increased levels of proteases are reported in the faeces of IBS patients but it is not clear if these are of human or bacterial origin

(Gecse et al. 2008). Human beta-defensin HBD2 is an antimicrobial peptide produced by intestinal epithelial cells and higher levels have been found in the faeces of IBS patients (Matricon et al. 2012).

1.3.4 Diagnosis Many known clinical diseases mimic IBS and the diagnosis of IBS is contingent upon the exclusion of disorders including celiac disease, fermentable oligodisaccharide, monosaccharide and polyol intolerance (FODMAP intolerance) and inflammatory bowel disease (IBD). IBS (subcategorised clinically as either IBS-D, IBS-C or IBS-M) may be a heterogeneous disorder that has many different aetiological causes that clinically appear similar. It would be useful to identify biomarkers for IBS. Serum biomarkers for IBS have been investigated with a panel of ten (Lembo et al. 2009) and fourteen (Jones et al. 2014) markers and although diagnostic sensitivity was

11 above 80% specificity was poor at 50%. Interestingly, different biomarkers within the panel distinguished different clinical subtypes suggesting IBS-D and IBS-C may have a different biological cause.

1.3.5: Therapy No specific therapy is reliably efficacious in IBS. A number of therapies are used including dietary modification (low FODMAP diet and high fibre), specific management of constipation or diarrhoea, antispasmodic agents, bile acid binding agents, broad spectrum antibiotics such as rifaximin, probiotics and neuromodulators/antidepressants such as amitriptyline.

1.4 Introduction to Blastocystis spp 1.4.1 What are they? Blastocystis spp. are common, anaerobic, unicellular, eukaryotic, enteric found in man and most species of animals. They may have been described as early as 1849 during the infamous cholera outbreak in London (Parija & Jeremiah 2013) but Alexeieff first described this parasite as a distinct unicellular organism present in the gut of lizards and leeches in 1911 (Zierdt 1991). A year later Brumpt described four species of Blastocystis including “Blastocystis hominis” (Dunn 1992). Over the last 100 years Blastocystis has been incorrectly classified as , cyst of a flagellate, an amoebae, and a sporozoa. Zierdt identified protozoan attributes of the organism in 1976 (Zierdt & Tan 1976b) (Table1.3) and subsequent phylogenetic analysis of the 18 bit-piece subunit of the ribosomal gene (18S rDNA) has classified it as a eukaryote in the class (Silberman et al. 1996)(Figure 1.1). The accuracy of this classification has been confirmed with molecular analysis of other Blastocystis proteins (Arisue et al. 2002). Great genetic diversity (up to 7% variation in amino acid sequences) has been identified within the 18S DNA coding for ribosomal RNA (rDNA) subtypes (Noel et al. 2005).

12 Table 1.3 Blastocystis protozoan characteristics (Zierdt 1991)

Figure 1.1 Taxonomic classification of Stramenopiles (Yoon 2002)

13 Stramenopiles, together with green plants, red , animals and fungi, separated nearly simultaneously 1,260 million years ago (Knoll 1994). The Stramenopile group, comprising 100,000 members, includes diverse uni- and multi-cellular organisms such as slime nets, water moulds and . Most Stamenopiles are characterised by mitochondria with tubular cristae, flagella and flagellar hairs. Uniquely Blastocystis does not possess hairs or flagellae and is the only stramenopile to live in the human intestine. Blastocystis is currently classified as  Eukaryote  Phylum Chromista  Class Stramenopiles  Order Blastocystae  Genus Blastocystis  Species Blastocystis subtype 1-17

1.4.2 Where are they found? Blastocystis spp. are found world-wide and have been described in the gut of every species of animal except Osteichthyes (fish with bony rather than cartilaginous skeletons) to date (Belova & Krylov 1994). The organism is found in the gastrointestinal lumen and is not invasive. In humans it is found in the caecum and terminal ileum (Stenzel & Boreham 1996) and no significant macroscopic mucosal changes are seen at endoscopy of small or large bowel (Zuckerman et al. 1994; Nagel et al. 2012). A recent study in pigs has reported Blastocystis organisms present predominantly in the lumen and adjacent to epithelial cells of the large bowel, with small bowel colonisation more common in immunosuppressed pigs, and no pathological changes seen in the underlying epithelial mucosa using light microscopy (Wang et al. 2014a).

1.4.3 Prevalence The true prevalence of Blastocystis carriage is not known as methods of diagnosis and data collection are problematic. It has been estimated that over one billion people may carry the organism (Parija & Jeremiah 2013). An

14 Australian study reported a prevalence of 6% (Hellard et al. 2000) in the asymptomatic community and this accords with reported rates in other developed countries from 10-30% (Stenzel & Boreham 1996; Scanlan & Stensvold 2013). Prevalence rates of 19% have been reported in Sydney hospital inpatients (Roberts et al. 2012). The lowest community carriage rates have been reported in Japan (1.5%) and Singapore (3.3%) (Yoshikawa 2012). Carriage rates in developing countries from every continent have been reported to be much higher, ranging from 20-66% (Tan 2008) and even 100% in 93 children from Senegal (El Safadi et al. 2014). Marked variation in regional carriage rates within countries such as China, Thailand and Turkey have been documented. Higher rates of carriage have been reported in lower socioeconomic groups, those with lower standards of hygiene (Yamada et al. 1987; Torres et al. 1992), and young adults (who tend to have a higher rate of infection compared to young children or older adults (Nimri 1993).

1.4.4 Routes of transmission It is likely that the cystic form of Blastocystis transmits infection. The vacuolated form of Blastocystis was shown to have low infectivity (<20%) when delivered orally although intra-caecal inoculation resulted in consistent infection in germ free guinea pigs (Phillips & Zierdt 1976). A robust cystic form was described in 1991 (Yamada et al. 1987; Stenzel & Boreham 1991) and subsequently shown to be infective when delivered orally to rats (Suresh et al. 1993), BALB/c mice (Moe et al. 1997) and birds (Tanizaki et al. 2005). The infection likely arises from indirect contamination of water supplies or by direct or indirect contact with humans or animals. Many studies suggest an association between Blastocystis infection and unfiltered or unboiled water (Leelayoova et al. 2004; Karanis et al. 2007) or water vegetable food contamination (Al-Binali et al. 2006; Li et al. 2007b). Blastocystis has been included as a microbial agent of interest in the 2008 water sanitation and health guidelines of the World Health Organisation (Organisation 2008). Exposure to animals is probably a risk factor as zookeepers and abattoir workers have also been reported to be associated with a higher rate of Blastocystis infection (Parkar et al. 2007). Influent and treated effluent sewage treatment samples both carry viable Blastocystis cysts (23% and 7%,

15 respectively) (Banaticla & Rivera 2011) and as few as 10-100 cysts may establish infection (Yoshikawa et al. 2004b). Direct and indirect (including water-borne) faecal-oral mode of transmission would be consistent with reports of frequent Blastocystis infection in campers or travellers returning from overseas destinations (particularly developing countries), children in day care centres and nursing home inpatients (Cirioni et al. 1999).

1.4.5 Zoonotic potential Humans have been reported to carry 9 of the 17 subtypes identified to date (Stensvold 2013) and harbour the greatest diversity of subtypes (Figure 1.2). Blastocystis subtypes 2, 4 and 7 but not 3 (all derived from human sources), were able to infect chickens and/or rats in a recent study (Iguchi et al. 2007) although some subtypes required larger inoculums of the parasite to establish infection.

16

Figure 1.2 Blastocystis subtypes in animals (Parkar et al. 2010)

17 Moderate animal host specificity is demonstrated in the various subtypes and subtypes from avian and non-avian hosts rarely overlap. Interestingly laboratory rats have been reported to have a high carriage of Blastocystis (60%) and in contrast laboratory mice were negative for Blastocystis (Chen et al. 1997; Chen XQ 1997). The carriage rate of Blastocystis has been reported to be as high as 71% in dogs and 67% in cats in Brisbane (Duda et al. 1998), although a later study has suggested that dogs are not a natural host and likely carry the parasite only when they have extensive contact with other reservoir hosts (Wang et al. 2013). Blastocystis carriage in zoo, farm and native animals is widespread and the prevalence of carriage in particular animals varies widely from country to country (Yoshikawa 2012). Zookeepers have been shown to carry unusual Blastocystis subtypes that are present in the animals they tend (Parkar et al. 2010). Pigs mostly carry ST5 and two studies report infection in pig keepers with ST5 in China (Yan et al. 2007) and Australia (Wang et al. 2014c). However, it is unlikely that livestock are a major reservoir for Blastocystis infection in humans as the major ST in these animals seen around the world is ST5 and ST10 (Alfellanin et al. 2013) and these subtypes are rare overall in humans. Human to animal transmission of infection, in either direction, is likely (Noel et al. 2005; Wang et al. 2014c) but cannot be proven until more accurate / definitive molecular markers are used.

1.5 Morphology of Blastocystis species 1.5.1 Introduction Blastocystis exist in multiple morphological forms and these have been broadly categorised into six groups. In reality the protozoa may take many forms, even within the same culture, that are not easily categorised. This pleomorphism has led to confusion and misclassification in the literature. It is not clear how many of these different Blastocystis shapes are artefactual created by exposure to unsuitable environmental conditions. The size of the organism may vary 100 fold and the size, shape and type of form of the

18 parasite depends on the particular subtype, and the age and culture conditions (Tan 2008). The six most commonly categorised forms are the vacuolated form (VF), granular form (GF), multi-vacuolar form (MVF), amoeboid (AF), avacuolar (AVF) and cyst. The most easily recognised form is the central vacuolated form (VF) and generally without the presence of the VF in a wet faecal smear it is not possible to identify Blastocystis. Faecal specimens from symptomatic patients with diarrhoea have been reported to have predominantly AF (Tan & Suresh 2006b) and larger and slower growing parasites (Tan et al. 2008).

1.5.2 Vacuolated, Granular, Multi-Vacuolar and Avacuolar forms The vacuolated form (VF) is the most distinctive form of Blastocystis found in faeces and culture (Figure 1.3.1)(Nimri 1993). It is a round or oval cell consisting of a large vacuole comprising 90% of the organism surrounded by a thin rim of . Compressed within the rim of cytoplasm are organelles consisting of 1-4 nuclei, multiple mitochondria-like organelles, Golgi apparatus and endosome-like (Stenzel et al. 1991). The size of the VF averages 10 m but ranges from 2-200 m. Cells exposed to air quickly deform, collapse and burst looking like “empty balloons” (Zierdt 1991). The granular form (GF) (Figure1.3.1) is similar to the VF with either stippled to dense central opacity or multiple distinct small granules within the central vacuole or the cytoplasm. The multi-vacuolar form (MVF) (Figure 1.3.2) is similar to a smaller VF (up to 8 m diameter), with only 1 or 2 nuclei and multiple small vacuoles present within the CV and cytoplasm. The avacuolated form (AVF) has been described as smaller (5 m diameter), with 1-2 nuclei, mitochondria with prominent cristae, Golgi apparatus, endoplasmic reticulum but no surface coat (Stenzel & Boreham 1996). The central vacuole is not present.

19 Figure 1.3 Light microscopy of different morphological forms of Blastocystis

1.3.1 Multiple smaller vacuolar forms, some larger granular forms Light microscopy, x40, polarised light acridine orange (0.05%) stain (image (Nagel et al. 2015a))

1.3.2 Large multi-vacuolar form with two smaller granular forms Light microscopy, x40 polarised light, acridine orange (0.05%) stain (image: R Nagel)

1.3.3 Two amoeboid forms, one with a large central vacuole and the filled with granules. Light microscopy, x40 polarised light, acridine orange (0.05%) stain (image(Nagel et al. 2015a)

1.3.4 Cysts (image: (Zaman 1996))

Bar 10uM

20 There is no precise definition of each of these forms. It is likely that they represent a continuum of each other although the direction of development is not clear. MVF reportedly become VF after short term culture (Stenzel et al. 1991) and VF become GF under adverse culture conditions. VF cultures can be induced to become GF predominant specifically by increasing the serum concentration, changing the culture medium, axenising the culture, or adding antibiotics (Dunn 1992). In the absence of definitive cell markers it may be prudent, when using light microscopy, to describe only VF or GF.

1.5.3 Amoebic form There are conflicting reports in the literature regarding the existence and morphology of the amoebic form (AF). They are reported to be smaller in size and occur more frequently in the stools of patients with diarrhoea (Tan & Suresh 2006b) and range from 8-200 m diameter. The AF is irregular in shape with , but no apparent motility, with a surface coat and a plasma membrane reported to be devoid of coated pits (Stenzel et al. 1989). Two forms of AF have been described (Tan & Suresh 2006a)(Figure 1.3.3 ) a dense granular sort (CV filled with dense granules) and another lucent form (no CV, cytoplasm with vacuoles and minimal organelles). The organelles (nuclei 1-2), mitochondria, inclusions, ribosomal strands) appeared to concentrate in pseudopodia that often encircle bacteria (Figure 1.4) and phagosomes containing bacteria were present in the cytoplasm.

1.5.4 Cysts The small cystic stage (2-5 m diameter) was not described until more recently (Mehlhorn 1988; Stenzel et al. 1991). It is round or oval, with a characteristic highly refractive, thick cell wall with or without a loose, fibrillar coat. The cell wall contains some chitin (Stenzel 1995). Multiple nuclei (1-4), mitochondria, endoplasmic reticulum, vesicles are seen within the cell wall that has a distinctive notch (Figure 1.3.4). Although rarely present in in vitro cultures, Blastocystis can be induced to encyst with the addition of conditioned bacterial products or trypicase (Suresh et al. 2009). This form is robust and can remain viable for 2 months at 4C (Yoshikawa et al. 2004b)

21 but is susceptible to drying, high temperatures and freezing (Table 1.4). Faecal cysts have been reported in only 21% of infected patients (Sun 1989) and it is likely they are shed intermittently and/or difficult to identify.

Table 1.4 Properties of Blastocystis cysts Condition Survival Time Survival Time Survival Time (Moe et al. 1996) (Yoshikawa et al. (Zaki et al. 2004b) 1996) 40-50C 12 hours 25 water 1-2 months 4C water 14-28 days 1-2 months -20C 24 hours Room temperature 3 days air 4C air 3 days 37C air <12 hours 2.5% potassium 2 weeks dichromate solution Chlorine ( sodium survived hypochlorite) 2.2ppm

1.5.5 Functional components This slimy coated organism contains multiple nuclei, mitochondria like organelles, Golgi apparatus, lysosomes, and endoplasmic reticulum compressed into a thin rim of cytoplasm surrounding the large central vacuole that often contains many granules (Figure 1.4). The surface coat of the Blastocystis parasite is a dramatic structure; this carbohydrate-rich, slimy coat has fibrils that may extend 10 m from the surface, although these shorten after prolonged culture (Stenzel et al. 1991) and may be lost off cysts and AVF’s altogether (figure 1.4.1). Bacteria and other parasites are often seen adhering to the coat, particularly in fresh faecal specimens, suggesting that entrapment may aid nutrition (Zaman et al. 1999). The coat appears to be continuously synthesised and may offer other mechanical or chemical protective functions (Cassidy et al. 1994) and binds Blastocystis organisms in culture together (Zhang et al. 2011).

22

Figure 1.4 Electron microscopy of Blastocystis functional components

1.4.1 Blastocystis hominis showing the outer surface coat with adherent bacteria (Zaman et al. 1999) 1.4.2 Vacuolar form of fresh monkey vacuolated from of Blastocystis showing puckered outer membrane of parasite (Cassidy et al. 1994) 1.4.3 Multiple large, elongated mitochondria-like organelles displaying internal cristae (Dunn 1992) 1.4.4.Plasma membrane of Blastocystis showing surface pits and fuzzy outer coat (Dunn 1992) 1.4.5 Amoebic form apparently phagocytosing a bacterium with pseudopod (Dunn 1992) 1.4.6 Different morphological forms of Blastocystis in the same culture (vacuolated and granular) (Dunn 1992) 1.4.7 Granular form of Blastocystis showing typical cap of chromatin on large nucleus, multiple granules and vacuoles and internalised bacteria (Dunn 1992)

2

1

3 4

5 6

7

23 The role of the central vacuole is not known. It has been proposed to have a storage or metabolic role, and stains positively with periodic acid Schiff or Alcian blue for glycogen/ carbohydrates (Yoshikawa et al. 1995b) and Sudan black B or Nile blue for respectively(Yoshikawa et al. 1995e). Neither 4’,6’-diamindino-2-phenylindole (DAPI) nor acridine orange staining detected DNA (Suresh et al. 1994b). Nutrients may accumulate in the central vacuole transferred in small vesicles in the cytoplasm that are transferred from the surface of the organism by clathrin-based endocytosis (Stenzel et al. 1989). The central vacuole may also be a repository for the storage of apoptotic bodies (Nasirudenn et al. 2004). The heterogenous granules seen in the vacuole have been described as myelin-like inclusions, crystalline granules and lipid droplets (Dunn 1992). A suggested reproductive role for these granules (Zierdt 1991) has never been proven. Cytoplasmic contents, often containing organelles, may invaginate and deposit filament or vesicle-like membrane bound structures into the central vacuole (Tan 2008). The plasma membrane of VF and GF (but not the AF) contains large coated pits (50 nm diameter) with clathrin present on the internal cytoplasmic surface (Stenzel et al. 1989)(Figure 1.4.2). These pits have a clear role in endocytosis. Ferritin particles (known to bind electrostatically to coated pits) are internalised almost immediately via coated vesicles. In VF ferritin was distributed to the CV, cytoplasmic vacuoles and tubules(Dunn 1992). The ferritin accumulation in GF was markedly different with loose accumulation only in the CV and no accumulation in the granules. The texture of the surface structure of Blastocystis (a function of the size of the coated pits), varies between different animal hosts and also becomes smoother in in vitro culture (Cassidy et al. 1994). Multiple nuclei are present in cells, usually at opposite ends of the cell, giving the VF a typical signet ring appearance. The nuclei are approximately 1m diameter, surrounded by a nuclear envelope with nuclear pores, and topped by a characteristic eccentric electron-opaque spot of dense chromatin material (Stenzel & Boreham 1996) (Figure 1.4.7).

24 Although Blastocystis is a strict anaerobe, the parasite contains multiple (sometimes hundreds) mitochondrial-like organelles (MLO’s) that are spherical or oval in shape, 1-3 m diameter with cristae (tubuar or sacculate) (Dunn 1992), contain DNA and have a measurable trans-membrane potential (Figure 1.4.3). This organelle is presumed to be an ancient remnant of an oxidative proteobacteria that adapted its enzyme profile to the anaerobic environment. As well as a role in energy production it has been proposed as a site of lipid synthesis (Zierdt 1991). Analysis of the mitochondrial genome shows that it lacks cytochrome C complex genes predicting that this organelle has an incomplete oxidative phosphorylation chain, a partial Kreb’s cycle, amino acid and fatty acid metabolisms and an iron-sulphur cluster (Denoeud et al. 2011). These MLO’s, unlike classic mitochondria, have enzymes capable of converting pyruvate into CO2 and H2. Golgi apparatus (usually adherent to the nucleus), endoplasmic reticulum and lysosomes are also seen in cytoplasm.

1.5.6 Life cycle It is generally agreed in the literature that the cyst is the robust infective form but there is no other consensus of the relationship of the particular morphological forms to each other. A number of different life cycles for this organism have been proposed but there is no general agreement (Figure 1.5)(Zierdt 1991; Tan 2008). Infection may persist for longer than 10 years.

1.5.7 Modes of reproduction Simple binary is the only widely accepted form of replication. The VF, GF or AF cell appears to divide in two and the organelles are divided between the two parts. Other modes of replication have been proposed, such as plasmotomy, endodyogeny, schizogeny, sporulation and invagination (Zierdt 1991; Singh 1995; Zhang et al. 2007). Binary fission of the VF and GF is relatively commonly seen. Division of nuclei within an intact outer membrane is also seen (schizogeny) suggesting this may be the first step into formation of two separate organisms.

25 Figure 1.5 Proposed lifecycle (Tan 2008)

26

1.6 Laboratory Diagnosis 1.6.1 Stool samples and microscopy The formol-ether concentration technique (FECT), commonly used for detecting parasite ova and cysts, is ineffective in detecting Blastocystis (Suresh & Smith 2004), even if this technique is optimised using faecal specimens preserved in formalin. Concentration and use of distilled water often lyse fragile Blastocystis cells. Microscopy of fresh faecal smears (unstained or stained with Lugol’s iodine or trichrome) is the tool commonly used by commercial pathology laboratories (polymerase chain reaction (PCR) testing of DNA has been combined with microscopy since 2013 in Australia). Detection rates of simple microscopy of wet faecal smears are modest (30- 50%). Some reports suggest that greater numbers of Blastocystis organisms (>5 organisms per high power field) seen in the faeces correlate with presentation with acute diarrhoea (Kain et al. 1987) whilst others report no correlation (Doyle et al. 1990). Multiple faecal specimens may be required as the parasite sheds intermittently (Suresh et al. 2009). Microscopy of faecal smears has been reported to be improved using a fluorescent Blastocystis specific antibody (Blastofluor)(Dogruman-Al et al. 2010).

1.6.2 Stool samples and culture Xenic in vitro culture (XIVC) of faeces is relatively cheap and increases the diagnostic yield by 50% (Dogruman-Al et al. 2010). Culture of Blastocystis in association with bacteria is possible in a number of mediums including Jones, Iscove’s modified Dulbecco, minimal essential medium, Robinson’s medium, or biphasic inspissated egg slant overlaid with Locke’s solution (Tan 2008). The organism appears tolerant of different osmotic conditions. Anaerobic culture is required at 37C (except for ST7 that prefers 40C). Although the presence of bacteria aids Blastocystis growth, bacterial or fungal overgrowth will kill the organisms (Barret 1921). Some XIVC may need longer incubation as particular subtypes are slow growing (Ho et al. 1993).

27 Axenic (without contaminating microorganisms) cultures (AXC) are useful for research into aspects of Blastocsytis including sensitivity to antimicrobial agents but are not used routinely for diagnosis. PCR analysis of the Blastocystis 18S ribosomal RNA gene (18S rDNA) (Stensvold et al. 2007a) in faecal specimens is commonly used in research and may be considered the “gold standard”. XIVC sensitivity and specificity closely approaches that of PCR techniques. If subtype analysis is required DNA is preferably extracted from the faecal sample rather than XIVC to avoid fast growing subtypes dominating in mixed subtype infections (Tan et al. 2010). PCR testing has been reported to be more sensitive with particular diagnostic kits (Yoshikawa et al. 2011) and false negative results may be due to the presence of faecal PCR inhibitors such as bile and complex polysaccharides. Differences between the subtypes have been described. Subtype 7 parasites prefer to grow at a higher temperature and have much longer doubling times in cultures (50-80 hours) compared to ST4 parasites that double every 20-30 hours (Tan 2008). XIVC will favour organisms with faster growth. Axenic cultures have shorter doubling times of 7-12 hours (Lanuza et al. 1997) compared to XIVC. Subtype 1 parasites are approximately 30% smaller in size in culture than subtype 4. Differences in sensitivity to metronidazole and have been demonstrated between axenically cultured ST4 and ST7 organisms (Mirza et al. 2010). Detection rates of Blastocystis vary widely between commercial laboratories (Utzinger et al. 2010) and Blastocystis is often undiagnosed. Some pathology laboratories do not report the presence of Blastocystis, even if seen, considering it not to be pathogenic or relevant.

1.6.3 Immunological techniques Immunological techniques are not used currently for diagnosis of Blastocystis infection. Different studies have reported variously that different Blastocystis subtypes have characteristic electrophoretic protein banding patterns (Chen et al. 1999) or similar banding patterns (Init et al. 2003). Antibodies specific for Blastocystis have been reported in human sera (IgM and IgG) (Zierdt & Nagy 1993; Garavelli et al. 1995) and faeces (IgA)

28 (Mahmoud & Saleh 2003) that are significantly higher in symptomatic patients. Antibody titres have been reported to be higher with length and severity of infection (Zierdt & Nagy 1993). The characteristics of antibody demonstrating a significantly raised titre in symptomatic patients has been variously reported as of the IgG2 class (Hussain et al. 1997), and detecting 12kD (Kaneda et al. 2000), 29kD (Gamra et al. 2011) 30 kD, 50kD and 118kD molecular weight proteins ((Hegazy et al. 2008). A cytopathic antibody, specific for a 30kD MW protein in Blastocystis, has been identified as reacting to the legumain (a cysteine proteinase) (Petersen AM 2012) present on the Blastocystis surface (Tan et al. 2001).

1.7 Molecular Techniques and Subtyping 1.7.1 Subtyping the organism using the ribosomal RNA gene Analyses of the Blastocystis ribosomal RNA gene began in 1993 (Clark 1997a) and proved useful to establish its taxonomic classification as a Stramenopile (Silberman et al. 1996). Analysis of the 18S rDNA using mainly PCR and restriction-fragment-length-polymorphism (RFLP) has revealed wide genetic variation. A collaborative effort in 2007 produced a consensus for the organisation and definition of Blastocystis 18S rDNA subtypes (ST’s) (Stensvold et al. 2007b). Scicluna primers (Scicluna et al. 2006) identify a 500bp length of the SSU rRNA gene and the PCR products are not as reliable for sequencing as the nested Wong primers (Wong et al. 2008) that produce a 1100bp nucleic acid sequence more distally towards the 3” end of this gene. Examination of the SSU-rRNA gene has identified a specific region that may allow separation of Blastocystis organisms into particular allelles/clades that allow more accurate genetic identification, within particular subtypes (Figure 1.6).

29

Figure 1.6 Barcode region of Blastocystis 18S ribosomal RNA gene identifed by different primers (Clark et al. 2013)

1.7.2 Epidemiology of Blastocystis subtypes Subtype-specific sequence-tagged-site primers have currently defined up to 17 subtypes (Figure 1.2) each displaying more than 5% divergence from each other. Subtypes 1-9 are found in man, the most commonly found subtypes being ST3 (42-92%), 1 (8-25%), 4 (10-23%) (Tan 2008). Within countries there is marked regional variation in ST prevalence patterns (Li et al. 2007a) and mixed ST infections are common. ST3 is almost universally the most common in humans in different countries (Tan 2008) (Table 1.5 ) and is thought to be the subtype of human origin. ST 1,2,3,4 and 6 are common in humans. ST8 is commonly seen in non-human primates and in contrast ST4 is rarely seen in this group of animals. Subtype 9 appears exclusive to humans. There is moderate species specificity and in general avian subtypes 5, 6, 7 and subtypes 11-13 (elephants, giraffes) do not infect humans. Subtypes found commonly in rodents (ST4), pigs (ST1, 5) and non-human primates (ST1,8) infect man. Interestingly ST4 is commonly detected in European and Australasian human

30 and rodent populations but almost absent in Egypt, Libya, Iran and Nepal (Clark et al. 2013).

Table 1.5 Distribution of Blastocystis subtypes infecting humans in different geographic regions (Tan 2008)

1.7.3 Multilocus sequence tagging (MLST) and clades Molecular techniques to define Blastocystis organisms more uniquely are required for epidemiological purposes. MLST sequencing of the mitochondrial DNA is used mainly in haploid genomes, usually in bacteria, but is possible in Blastocystis mitochondria and has now been developed for ST1, 3 and 4 organisms (Stensvold et al. 2012; Clark et al. 2013). Large intra- subtype genetic diversity has been confirmed, particularly in ST3, and in contrast ST4 has very little genetic diversity. Human ST3 organisms mostly fall within one clade but with significant sequence variation, with multiple

31 clades seen in non-human ST3 hosts. Human ST4 organisms fall within one clade as well but with little sequence variation suggesting that ST4 entered the human population relatively recently (Clark et al. 2013).

1.7.4 Whole genome sequencing Whole genome sequencing has been achieved for ST7 and ST4 (Denoeud et al 2011; Wawrzyniak et al, 2015). The ST7 genome is small (19Mb in size) containing 6000 genes, and this has allowed prediction of the Blastocystis metabolic and secretory proteins. These predictions may need to be interpreted with some caution as a recent study has reported that the parasite has another unique feature for a eukaryote, namely it uses polyadenylation-mediated termination codons in the nuclear DNA (Klimes et al. 2014) thereby requiring re-annotation of genome sequences deposited on GenBank1.

1.8 Clinical Association in Humans 1.8.1 Introduction The pathogenicity of Blastocystis in humans is hotly debated. Currently there is no reliable animal model to test Koch’s postulates developed in 1884, and even with the many modifications to these “rules” since the recognition of viruses, healthy carrier states and the advent of molecular biological techniques (Fredricks & Relman 1996), proof of pathogenicity remains challenging. Indisputably many healthy people carry Blastocystis but over the last 100 years increasingly frequent reports of diarrhoeal illness associated with the parasite have emerged. The number of Blastocystis organisms observed in some patients has been reported as high as several billion per stool and up to 200 fluid bowel actions per day (Mehlhorn 1988). Co-infection with other enteric parasites is common (Markell & Udkow 1986). Case controlled studies of symptomatic and asymptomatic individuals have reported conflicting results (Kain et al. 1987; Nigro 2003; El-Shazly et al. 2005b; Graczyk et al. 2005; Leder et al. 2005; Kuo et al. 2008). Symptoms that have

1 www.ncbi.nlm.nih.gov/genbank/

32 been attributed to Blastocystis include diarrhoea, abdominal discomfort, anorexia and bloating (Stenzel & Boreham 1996). In children it has been reported to be associated with diarrhoea (Graczyk et al. 2005), poor growth and presentation with “acute abdomen” (Andiran et al. 2006). Blastocystis has been detected frequently in traveller’s diarrhoea (Jelinek et al. 1997; Paschke et al. 2010). Isolated case reports have described bloody diarrhoea (al-Tawil et al. 1994), but generally the organism is only invasive opportunistically (Horiki et al. 1999). Chronic urticaria (Valsecchi et al. 2004; Gupta & Parsi 2006; Vogelberg et al. 2010; Zuel-Fakkar et al. 2011), and reactive arthritis (Lee et al. 1990) have also been reported to be associated with Blastocystis infection.

1.8.2 Blastocystis and Irritable Bowel Syndrome Case controlled studies of patients with IBS have shown positive (Giacometti et al. 1999; Yakoob et al. 2004; Stensvold et al. 2009c) and negative association with Blastocystis infection (Sun et al. 1989; Tungtrongchitr et al. 2004; Surangsrirat et al. 2010; Krogsgaard et al. 2014). Carriage of Blastocystis is reported to be as high as 49% in patients with IBS compared to 24% in control patients (Yakoob et al. 2004). Carriage in patients with IBS-D is even higher (73%) compared to healthy controls in a later study (Yakoob et al. 2010). A recent meta-analysis of eleven trials, consisting of 1,728 patients with IBS and 1,292 control subjects showed that IBS patients were 2.3 times more likely to be infected with Blastocystis spp. than healthy controls (Nourrisson et al. 2014).

Levels of immunoglobulin G2 (IgG2) class antibodies directed against Blastocystis were reportedly higher in patients with IBS carrying Blastocystis compared to asymptomatic control subjects positive or negative for Blastocystis carriage (Hussain et al. 1997).

1.8.3 Immune-compromised patients Parasite infections and symptomatic diarrhoea are more common in immune-compromised patients, including those with human immunodeficiency virus (HIV), malignancy, renal failure or immunosuppressant therapy (Stark et al. 2007; Poirier et al. 2011). Blastocystis carriage is reported in over 50% of

33 immune-compromised patients (Idris et al. 2010) and rates are also higher in men who have sex with men (Stark et al. 2007). Case controlled studies in immunocompromised patients have shown conflicting results (Tasova et al. 2000; Poirier et al. 2011), and although the prevalence of Blastocystis is accepted to be higher in immunocompromised patients gastrointestinal symptoms are common in Blastocystis-positive and negative patients.

1.8.4 Cancer patients In vivo studies have suggested that Blastocystis hominis may facilitate the growth of human colorectal cancer cells (Chandramathi et al. 2010). Blastocystis antigen derived from infected patients with gastrointestinal symptoms also caused an up-regulation of nuclear factor kappa-light-chain- enhancer of activated B cells (NF-) and protein coding gene CTSB in colorectal cancer cell HCT116 culture (Chan et al. 2012). No up-regulation was seen with antigen from non-symptomatic carriers. Subtype 3 was shown to cause more proliferation of the HCT116 cells than subtypes 1,2,4 and 5 (Kumarasamy et al. 2013). Blastocystis is found commonly in cancer patients (Tasova et al. 2000) and survives despite concurrent chemotherapy (Chandramathi et al. 2012) and it may be an opportunistic infection that contributes to chemotherapy diarrhoea.

1.8.5 Inflammatory Bowel Disease (IBD) Blastocystis carriage has been reported to be uncommon in patients with active ulcerative colitis (Petersen 2012) and rare in patients with Crohn’s disease in Denmark but equivalent rates of carriage (around 6%) were seen in Turkey in IBD patients and healthy controls (Cekin et al. 2012). Blastocystis has also been reported in association with active colitis with resolution of colitis with metronidazole therapy (Tai 2011).

1.9 Animal studies Despite the wide prevalence of Blastocystis carriage in animals Blastocystis infection with diarrhoea has only been reported to occur in pig

34 herds (Burden et al. 1978), a pig-tailed macaque (McClure et al. 1980) and a dog (Chapman et al. 2009). Animal models have been used to investigate Blastocystis infection and experimental infections in rats, mice, guinea pigs and chickens have been described (Tan 2008). Diarrhoea with caecal and colonic mucosal ulceration has been shown to occur in germ free guinea pigs (Phillips & Zierdt 1976), albino (Elwakil & Hewedi 2010) and BALB/c mice (Moe et al. 1997) infected with Blastocystis. Mucosal damage was more severe in rats infected from symptomatic human patients, particularly ST1 compared to asymptomatic patients (Hussein et al. 2008). These animal models have been useful to confirm the infectivity of the cystic form (Yoshikawa et al. 2004b; Tanizaki et al. 2005). Rats and chickens have been favoured as animal models as they have high rates of carriage (95%)(Lee & Stenzel 1999). Blastocystis ST 2,4,7 derived from humans could infect both rats and chickens but ST3 was not able to infect these animals (Iguchi et al. 2007), again suggesting moderate species subtype specificity. A major issue with these animal models is that the range of Blastocystis subtypes described in chickens and rats are not the most commonly seen in humans (Scanlan 2012a) and mice rarely carry Blastocystis naturally (Chen et al. 1997). Pigs may provide a useful animal model to study Blastocystis infection as they commonly carry the parasite, (although they predominantly carry ST5 organisms that rarely infect humans) (Wang et al. 2014a).

1.10 Possible Mechanisms of Pathogenicity 1.10.1 Introduction Although many Blastocystis carriers are asymptomatic this does not exclude pathogenicity, as we know that most infections with the known pathogens Giardia and Entamoebae histolytica are asymptomatic (Tan 2008). It is not clear if Blastocystis pathogenicity, if it exists at all, is related to an intrinsic property of the organism or to host factors or both.

1.10.2 Intrinsic factors

35 It is difficult to quantitate the number of Blastocystis organisms infecting a patient for a number of reasons. It is difficult to identify any form but the VF of Blastocystis in the stool, the number of parasites in the stool may not necessarily represent the number in the caecum, the quality of the faecal specimen may be important as the parasites lyse easily when exposed to fresh air and we know that the parasite is shed intermittently. However, using simple microscopy of faecal smears large numbers of Blastocystis organisms can be seen in symptomatic and asymptomatic individuals suggesting that parasite load does not determine symptoms (Sun 1989). It has been suggested that longitudinal quantitative PCR studies may allow better assessment of the relative importance of parasite load (Stensvold 2013a). A number of studies have explored whether the subtype of the Blastocystis has any association with symptoms, IBS or urticaria. ST1 (Kaneda et al. 2001; Tan et al. 2006; Yan et al. 2006; Hussein et al. 2008; Eroglu et al. 2009; Stensvold et al. 2009b), ST3 (Tan et al. 2008) and ST4 (Poirier et al. 2011) have variously been reported to occur more frequently in symptomatic patients. Other studies have not shown any significant difference in the ST pattern between patients and the control population (Yoshikawa et al. 2004a; Dogruman-Al et al. 2009; Elwakil & Talaat 2009; Rene et al. 2009; Souppart et al. 2010) or other asymptomatic household members (Nagel et al. 2012) (Table 1.6). The large genetic diversity within each 18S rDNA subtype makes it possible that this form of molecular characterisation is not sufficient to discriminate different organisms.

Table 1.6 Blastocystis subtype variation in irritable bowel syndrome (Scanlan 2012b)

36

Virulence factors intrinsic to the parasite have been explored in a number of ways. Axenic cultures of Blastocystis (ST4) have been shown to cause cytopathic effects on Chinese hamster ovary cells (Walderich et al. 1998). Cultured cells and lysates of the culture induced , disrupted epithelial barrier function in a contact independent manner, rearranged F-actin distribution and increased the permeability across rat colonic epithelial monolayers (Puthia et al. 2006). If Blastocystis causes pathogenic effects without adhering to the gut mucosa it suggests that the parasite may be secreting a virulence factor. Blastocystis proteases, particularly cysteine proteases, have been shown to cleave human secretory immunoglobulin A (Puthia et al. 2005) and induce pro-inflammatory cytokine IL-8 production (Puthia et al. 2008) from host cells via NF-B activation. Cysteine protease activity has been demonstrated to vary between different Blastocystis subtypes (Mirza & Tan 2009). Analysis of the genome of Blastocystis ST7 organism has predicted that the parasite would be able to produce all major classes of proteases, and particularly cysteine proteases (20/22 predicted to be cysteine proteases) (Denoeud et al. 2011). Cysteine proteases are known to degrade luminal immunoglobulin A (IgA) and disrupt the epithelial mucosal barrier (Figure 1.7). Cysteine protease legumain is secreted by Blastocystis and this legumain has been shown to be (unusually) located on the cell surface and important for the survival of the parasite (Wu 2010).

37

Figure 1.7 Cysteine proteases and possible virulence mechanisms (Denoeud et al. 2011)

1.10.3 Host factors Host factors that might influence the effect of Blastocystis carriage, apart from studies in immune-compromised patients, are relatively unexplored. A study that assessed differences in cytokine expression between symptomatic and asymptomatic carriers of Blastocystis found that symptomatic patients were more likely to have polymorphisms in IL-8 and 10 (Olivo-Diaz A 2012) but not IL-6 and TNF-α (Jimenez-Gonzalez et al. 2012). Host antibody responses to infection may vary between individuals. Faecal (Mahmoud & Saleh 2003) and serum (Zierdt & Nagy 1993) antibody titres to Blastocystis antigens have been reported to be higher in symptomatic patients. A serum antibody reacting to a 30kDa MW protein in Blastocystis has been reported to be more common in symptomatic patients (Gamra et al. 2011). This antibody may correlate with the antigen that has been identified as a cysteine protease present on the cell surface of Blastocystis organism (Wu et al. 2010).

1.10.4 Pathogenic passengers

38 The term endosymbiosis has no implication regarding pathogenicity for the host but simply means that one organism is able to live and replicate within another host organism. Some endosymbiont bacteria have evolved mechanisms to avoid phagosomal lysis by their host and this increases their survival time in chlorinated water and decreases their susceptibility to antibiotics (Winiecka-Krusnell & Linder 2001). Endosymbiosis is not uncommon and almost 30% of free living amoebae Acanthamoebae polyphaga had some type of endosymbiont (Lau & Ashbolt 2009). Bacteria pathogenic to humans that have been reported to replicate in other free living amoebae include Chlamydia pneumonia, Helicobacter pylori, Mycobacterium leprae & avium, Rickettsia-like organisms, Legionella, Franscisella tularensis, coliforms, Yersinia enterocolitica, Burkholderia spp., Listeria, Vibrio and Wolbachia spp.(Winiecka-Krusnell & Linder 2001). Endosymbionts in Blastocystis have been reported previously (Zierdt & Tan 1976a; Dunn 1992; Stenzel & Boreham 1994). Intracellular replicating gram negative rods/cocci were first noted in 1976 (Zierdt & Tan 1976a; Stenzel & Boreham 1997) and the endosymbiotic condition was found to be more prevalent in axenised cells, unaffected by many antibiotic therapies and more likely to occur in association with giant Blastocystis cells. Zierdt (Zierdt & Tan 1976a) was unsuccessful in culturing or identifying the endosymbiont. He reported that the endosymbiont had no cell wall, was not flagellated and there was no evidence of a bacteriophage. Endosymbiosis of axenically cultivated cells was stable and permanent. Blastocystis cells died when high dose chloramphenicol or tetracycline appeared to eradicate the bacteria. Stenzel (Stenzel & Boreham 1994) reported two instances only of endosymbiosis in Blastocystis derived from a duck (Figure 1.4) and a monkey faecal samples. No endosymbionts were seen in the central vacuole; the duck samples had organisms within the nucleus of the Blastocystis and the monkey sample showed the bacteria in the cytoplasm. The number of organisms per cell varied from 2 to 40. The size of the bacteria (a Gram-negative rod/cocci), differed in duck and monkey samples suggesting they were different organisms. Almost all Blastocystis organisms in those samples contained the endosymbionts, although overall this was a rare finding.

39 Viruses may also be endosymbionts. Virus-like particles have also been reported in the cytoplasm of a small number of Blastocystis forms, in parasites derived from Macaca monkeys (Stenzel & Boreham 1997) and seasnakes (Teow et al. 1992). Endosymbiosis can alter the virulence of the host or the endosymbiont in either direction. The endosymbiotic bacteria in the free-living amoebae described above often developed increased virulence after passage through the amoebae. In contrast, the host Filaria sp. parasites become more pathogenic if the endosymbiont Wolbachia remains viable. Although endosymbionts were only seen in a small number of Blastocystis cultures (<10%) they may be difficult to detect. It is possible that Blastocystis is an innocent bystander and pathogenicity derives from carriage of unidentified endosymbionts.

1.11 Antimicrobial Therapy 1.11.1 Current therapeutic guidelines for Blastocystis The Center for Disease Control and Prevention in the United States stated in the latest update in 2012 2 that although the clinical significance of Blastocystis infection is controversial the following three antibiotics are suggested as first line therapies, namely metronidazole, trimethoprim/sulfamethoxazole and .

1.11.2 In vitro antibiotic sensitivity In vitro studies have systematically assessed Blastocystis antibiotic sensitivity. Only a small number of organisms were tested, subtype data is only available on the most recent ones and practical application may be limited by restricted availability of amoebicides. Axenic strains from three symptomatic patients and one monkey tested for antibiotic sensitivity were classified in three groups namely inhibitory “reducing cell counts 50-100 fold” , (emetine, metronidazole, firazolidone, trimethoprim-sulfamethoxazole, Entero-Vioform and pantamidine),

2 http://www.cdc.gov/parasites/Blastocystis.

40 moderately inhibitory “reducing cell counts 5-50 fold” (floraquin and chloroquine) and not inhibitory, “reduced cell counts less than five fold” (diloxanide furoate and paromycin, other common antibacterial or antifungal drugs)(Zierdt et al. 1983). The sensitivity of various drugs against Blastocystis (axenic strain Netsky) was measured using incorporation of 3H-hypoxanthine to assess cell viability (Dunn & Boreham 1991). The most effective drugs in descending order were furazolidone, emetine, EnteroVioform, satranidazole, secnidazole, chloroquine and trimethoprim (Table 1.7).

41

Table 1.7 Activity of drugs against Blastocystis axenic stock cultures (Dunn & Boreham 1991)

The most recent study (Mirza et al. 2010) used a resazurin assay to measure intrinsic cellular activity in different subtypes of Blastocystis exposed to antimicrobial microassays. Variation in drug sensitivities between subtypes

42 were found, namely resistance to metronidazole and tindazole was observed in ST7 (but not another similar imidazole, ). Blastocystis was found to be sensitive to the antimalarial drug mefloquine and increased sensitivity was noted with a trimethoprim-sulfamethoxazole combination ratio of 1:2 (rather than the conventional ratio of 1:5). Herbs including ginger, black pepper, white cumin and garlic have also been tested recently in vivo and garlic has been found to be as effective as metronidazole in inhibiting growth in cell culture (Yakoob et al. 2011). The effects of garlic were ascribed to the presence of thiosulfinates that inhibit RNA and DNA synthesis and the other herbs contain alkaloids with antibacterial activity.

1.11.3 In vivo antibiotic sensitivity Clearance of Blastocystis infection is often unsuccessful (Stensvold et al. 2010). Many different drugs (metronidazole, , iodoquinol, trimethoprim/sulfamethoxazole, nitazonanide, , ketoconazole, ornidazole) have been reported to be useful in eradicating Blastocystis clinically (Table 1.8).

43

Table 1.8 In vivo efficacy of antimicrobials against Blastocystis (Stensvold et al. 2010)

44 The most commonly used drug has been metronidazole and eradication rates reported in the literature vary widely from 0-100% (Stensvold et al. 2010). Eradication rates in children treated with 10 days of probiotic Saccharomyces boulardii, 10 days of metronidazole therapy and 10 days with no therapy (evaluated using only faecal smears for diagnosis) were 77%, 66% and 40% respectively (Dinleyici 2011). A recent retrospective study has reported Blastocystis clearance rates of 77% using paromomycin sulphate (500 mg tds for 5-10 days) and much higher than comparable clearance with metronidazole (38%) or no therapy (22%) (van Hellemond et al. 2013). This high clearance rate is at odds with the finding that paromomycin has no in vivo inhibition against ST 4 and 7 isolates of Blastocystis (Mirza et al. 2010) and perhaps the broad spectrum antibacterial actions of paromomycin are more relevant for clearance (Mirza et al. 2010). A single transcolonic infusion of furazolidone, nitazoxanide and secnidazole followed by ten days of oral nitazoxanide was also reported to clear 100% of eighteen Blastocystis positive patients (Borody 2012). Comparison of eradication in different studies is difficult. False negative results may be due to the low sensitivity of different diagnostic tests and irregular shedding of the parasite. Perceived failure to clear the parasite may be due to drug resistance (Mirza et al. 2010), suboptimal dosing schedule, poor compliance or rapid re-infection from the environment. Eradication may be due to the effect of the drug on Blastocystis, an indirect effect on bacterial enteric flora or even occur spontaneously. The age of the patient and clinical presentation are other factors that may change response to therapy.

1.11.4 Therapeutic challenges We need to be able to identify the patients, if any, who warrant therapy for Blastocystis. Knowledge of the life cycle of Blastocystis is needed to understand how long therapy will be required for and perhaps what morphological forms should be targeted. Effective drugs for the particular organism and subtype need to be identified and a means developed to deliver them to the site of infection in the right colon. Reservoirs of infection in the community need to be identified and strategies designed to limit spread of infection.

45 1.12. Aims and objectives of this thesis 1.12.1 The hypothesis

Blastocystis carriage may cause irritable bowel syndrome symptoms in susceptible people.

Blastocystis carriage is common in the community and most carriers are healthy. Irritable bowel syndrome, diagnosed according to subjective clinical criteria, is also commonly diagnosed in the community. A multi- pronged approach is required to identify individuals where these two conditions may be causatively linked. This thesis will investigate i) features of the parasite’s life cycle through observation of live parasites in xenic cultures, ii) the association between prevalence of particular Blastocystis ST’s and symptomatology iii) antibiotic therapy efficacy and if possible, clinical outcomes following parasite eradication iv) host immune responses to infection and v) whether significant differences exist between the gut microbiome of clinical groups of Blastocystis- positive and negative IBS patients and healthy controls.

1.12.2 Investigative approaches and objectives Microscopy of the Blastocystis organism in cultures may offer insights into its life cycle. (i) I will use deconvolutional microscopy to examine xenic cultures of Blastocystis organisms. Using fluorescent spectroscopy and specific fluorescent-tagged antibodies I will examine the different morphological forms of Blastocystis. Time-lapse sequences will be used to monitor the cultures and examine possible morphological or physiological changes in the organisms over time.

Clinical studies of symptomatic patients and their household contacts will provide information about the ease and mode of transmission within human and animal household contacts. It will also provide information about the

46 subtypes encountered in these groups and whether subtypes are clinically relevant. (ii) I will study symptomatic Blastocystis-positive IBS patients and their household contacts, both human and animal, to assess rates of transmission within households and identify the subtypes found within these groups.

Assessment of different antimicrobial regimens will guide us to the most effective therapeutic regimen for eradication. Giardia duodenalis was only recognised to be pathogenic when it was possible to effectively eradicate it in symptomatic patients (Gardner & Hill 2001). We may be able to better assess whether Blastocystis is the cause of symptoms if we have a therapeutic regimen with a high success rate (>90%) for eradication. (iii) I will conduct two clinical studies using different antimicrobial regimens, namely monotherapy with either metronidazole or trimethoprim/sulfamethoxazole, and triple antibiotic therapy with secnidazole, diloxanide furoate and trimethoprim/sulfamethoxazole, to assess eradication rates.

Subtyping may not be able to discriminate Blastocystis organisms with regard to pathogenicity. Instead, assessment of host antibody responses to the parasite might establish whether significant differences exist between symptomatic and asymptomatic carriers and this may support the hypothesis that Blastocystis is pathogenic in some people and not in others. It might enable Identification of specific Blastocystis antigens associated with symptoms and may also provide insights into the mechanism of pathogenicity of Blastocystis spp. (iv) I will evaluate serum immunoglobulin reponses in IBS patients and healthy controls using Western blotting. Further analysis of subgroups based on Blastocystis carriage status will be carried out to identify any Blastocystis proteins of interest. I will attempt to identify any reacting proteins by probing with protease antibodies, including antibody specific for legumain (MAb1D5), which has already been identified as a protein of interest in symptomatic patients (Wawrzyniak et al. 2012).

47

We know that the faecal microbiome has a distinctive pattern in patients with IBS. IBS is likely a heterogenous clinical condition. If Blastocystis patients are not a distinct clinical subset of IBS patients the FM pattern should be similar in IBS patients positive and negative for Blastocystis carriage. If the FM pattern is different in these two groups it suggests Blastocystis carriage status defines a different clinical group within IBS patients, and would support the hypothesis that Blastocystis spp. cause IBS symptoms. (v) The FM in IBS and healthy controls will be assessed using metagenomic analysis. The bacterial profiles in these groups will be evauluted from phyla to genus and further subgroup analysis will assess any relationship with the Blastocystis carrier status

48 49

2 Blastocystis subtypes in symptomatic and asymptomatic family members and pets and response to therapy

Nagel, R., Cuttell, L., Stensvold C.R., Mills, P., Bielefeldt-Ohmann, H., Traub, R., 2012. Blastocystis subtypes in symptomatic and asymptomatic family members and pets and response to therapy. Internal Medicine Journal 42: 1187-1195

-presented as published

50 Blastocystis subtypes in symptomatic and asymptomatic family members and pets and response to therapy

Robyn Nagel*1, Leigh Cuttell1, C.R. Stensvold2, Paul C. Mills1, Helle Bielefeldt- Ohmann1 and Rebecca J. Traub1

1 School of Veterinary Science, The University of Queensland, Gatton 4343, Queensland, Australia.2 Department of Microbiological Diagnostics, Statens Serum Institut, 2300 Copenhagen S, Denmark and Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK

2.1 Abstract Background: Blastocystis is a common, enteric parasite. The pathogenicity of the organism is uncertain but subtypes (ST) 1,3 have been reported more likely to cause Irritable Bowel-like symptoms. Aims: We treated symptomatic patients positive for Blastocystis (BSP) with conventional therapy and analysed 16SSU3 rDNA to assess clearance and carriage rates and subtype prevalence of the parasite in the asymptomatic household members. Methods: In a longitudinal, prospective case study 11 symptomatic patients positive for Blastocystis (SBPs) underwent outpatient clinical assessment to exclude other diagnoses before 14 days of either metronidazole 400 mg three times daily or trimethoprim/sulfamethoxazole 160/800 mg twice-daily therapy. Faecal specimens were collected from patients at baseline, Day 15, 28 and 56 after therapy and from 17 family members and 8 pets at Day 15. Specimens were analysed by faecal smear, culture and PCR analysis of 16SSU rDNA. Results: No patient cleared the organism following therapy. Subtypes 1 (45%), 3 (36%), 4 (36%), and 6 (9%) were found in the SBPs and subtypes identified before and after therapy were identical in each individual. All household contacts were positive for Blastocystis and 16/17 (94%) contacts showed identical Blastocystis subtypes to the symptomatic family member. All pets were positive for Blastocystis with PCR testing, 7/8 (88%) demonstrating ST concordance with the SBP. Conclusions: Conventional therapy is ineffective for symptomatic Blastocystis infection. The high prevalence of Blastocystis infection within households suggested transmission

3 This chapter is reproduced as published but should read 18SSU (16SSU used for bacteria, 18SSU used for detecting eukaryotes) 51 between humans and their pets. Sub-typing analysis of SSU rDNA alone in Blastocystis does not appear to predict pathogenicity. Key Words: Blastocystis; Irritable Bowel Syndrome; Endoscopy; Metronidazole; Transmission

2.2 Introduction Blastocystis are ubiquitous, anaerobic parasites commonly identified in human and animal stool (Tan 2004). Reports suggest Blastocystis causes disease in humans (Tan 2008). Symptoms attributed to the organism are abdominal pain, bloating and diarrhoea (Zierdt 1991; Yakoob et al. 2004). The genus of Blastocystis comprises at least 13 subtypes (ST) that demonstrate moderate animal species-specificity (Stensvold et al. 2009a). There is regional variation in the prevalence of different Blastocystis subtypes found in humans, although ST3 is universally the most common. Recent studies have suggested ST1 (Tan et al. 2006; Yan et al. 2007; Hussein et al. 2008; Eroglu et al. 2009; Stensvold et al. 2009c) and/or ST3 (Tan et al. 2008) may be more likely to be pathogenic in humans. The parasite cyst is transmitted by the faecal-oral route (Leelayoova et al. 2004) and close contact with animals may be a risk factor for acquisiton of infection (Li et al. 2007b; Parkar et al. 2010). Metronidazole, the most commonly prescribed drug for Blastocystis, is reported to have widely varying rates of efficacy (Stensvold et al. 2010). These inconsistencies may be attributed to insensitive diagnostic methods, variability of Blastocystis subtype-response to drugs, rapid re-infection or drug resistance. In this pilot study we examined the Blastocystis STs of symptomatic patients and their respective human and animal household contacts and recorded the efficacy of conventional therapy in the patients. We aimed to evaluate antimicrobial efficacy, the rate of environmental re-infection, and explore ST variations in clinical presentation or response to therapy.

2.3 Methods 2.3.1 Study Outline This prospective longitudinal study was conducted in a rural specialist outpatient clinic (Toowoomba Gastroenterology Clinic). Patients who presented complaining of chronic gastrointestinal symptoms and were positive for Blastocystis carriage were screened to exclude other causes for symptoms. Faecal samples were collected from eligible patients before and after a course of antimicrobial therapy for Blastocystis, and at

52 completion of the antimicrobials faecal samples were collected from all the human and animal household contacts. 2.3.2 Inclusion protocol Thirteen adult patients presenting consecutively to the clinic from 09/04/2008 to 31/08/2009 positive for Blastocystis carriage and complaining of gastrointestinal symptoms of more than 6 weeks duration were assessed clinically. Blood tests, full blood count, serum calcium, thyroid function tests and tissue transglutaminase antibody, were checked. Faecal microscopy, culture and toxin assay for Clostridium difficile was performed to check for ova, cysts, oocysts and bacterial pathogens. In addition a faecal ELISA kit (“ProsSpecT, Remel Xpect” by Oxoid, UK) tested for Giardia and Cryptosporidium. A validated, in-house real-time faecal PCR analyses carried out by Statens Serum Institute laboratory excluded occult Dientamoeba fragilis, Giardia duodenalis, Cryptosporidium spp., and E. dispar infection. A 3-day culture for hookworm was performed. Helicobacter pylori infection was evaluated by gastric biopsy at endoscopy. One patient with ulcerative colitis and another patient unable to undergo endoscopy were excluded from the study. The remaining 11 symptomatic Blastocystis patients (SBP’s) proceeded to upper, lower endoscopy and capsule enteroscopy, performed by the same endoscopist. Mucosal biopsies were taken from the gastric antrum and duodenum for histology and estimation of disaccharidase levels. Biopsy specimens were taken from the ileum, right and left sides of the colon.

2.3.3 Study Protocol S y Three baseline faecal specimens were taken from 11 eligible SBP prior to starting m 14 days of metronidazole 400 mg three times daily or trimethoprim-sulfamethoxazole p t 160mg/800 mg twice daily. Further faecal samples were taken within 48 hours of ceasing o antibiotics and 2 and 6 weeks after cessation of therapy. m a Trimethoprim-sulfamethoxazole 160mg/800 mg twice daily for 14 days was given to t patients who had previously failed one course of the identical dose of metronidazole or i c who were intolerant of the medication. Patients were reviewed clinically initially 2 weeks prior to therapy, at the completion of the antibiotic therapy and 6 weeks later at the p a completion of the study. All patients kept a diary recording the number of daily bowel t movements and general well being (scored 1-10; poor-excellent). They also recorded i e whether any household contact or animal was experiencing any change in gastrointestinal n health. t 53 p o s All the household contacts (HC) of these patients were asked to provide fresh stool specimens within 48 hours of the symptomatic patient finishing the course of therapy. Stool specimens were also collected on this date from all the household animal contacts (AHC). All SBP stool specimens were subjected to routine pathology faecal specimen microscopy and culture throughout the study. All stool specimens (SBP, HC and AHC) were subject to smear, culture and PCR for detection of Blastocystis and sub-typing.

2.3.4 Diagnostic methods All samples were run in parallel for the presence of Blastocystis using a simple fresh unstained wet faecal smear, in vitro culture and PCR (confirmed as Blastocystis using DNA sequencing). An individual human or animal was considered positive for Blastocystis if any one of the tests were positive. Parasitological Methods Fresh unpreserved faecal samples from patients were divided into two samples upon receipt. One sample was screened using a faecal smear by a commercial diagnostic laboratory in Brisbane while the other was transported to the School of Veterinary Science, University of Queensland (SVS-UQ). Unstained fresh faecal smears were examined under 40x magnification by light microscopy. The load of Blastocystis organisms was classified as light, medium or heavy (<5, 5-10, >10 organisms per high power field). Additionally, 100 mg of stool was inoculated in Jones’ culture medium (Jones 1946) and incubated at 37 ºC for 48-72 hours. Cultures were examined for the presence of Blastocystis by light microscopy. Molecular methods DNA was extracted from unpreserved faecal samples and pelleted cultures of Blastocystis using the QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions, except that following the addition of lysis buffer, the faecal suspension was subjected to three freeze-thaw cycles in liquid nitrogen and a 95 ºC water bath, respectively. The genomic DNA from stool and cultured Blastocystis specimens from all patients, household members and animals were subjected to PCR analyses. Three different published PCRs capable of amplifying the small subunit (SSU) rRNA gene of STs 1-13 of Blastocystis (Bohm-Gloning et al. 1997; Clark 1997b; Stensvold et al. 2006; Wong et al. 2008) were utilised in order to maximize the likelihood of detecting mixed ST infections. All PCR products positive for Blastocystis were subjected to DNA sequencing and phylogenetic analysis to identify the subtype. PCR products were run on gels of 1.5%

54 agarose in 1×TAE buffer at 150V in a Biorad electrophoresis system and were purified using Purelink PCR Purification Kit (Invitrogen) prior to sequencing. Sequencing was done using an ABI 3130xl Genetic Analyzer (Applied Biosystems) using a Ready Reaction Kit Version 3.1 (PE Applied Biosystems, California, USA) and sequences edited and assembled using Finch TV v 1.4.0. DNA. Sequences were deemed as ‘mixed’ when sequence profiles were characterized by a clean sequence with multiple peaks at sites of nucleotide polymorphism. DNA sequences were aligned with 15 previously sequenced Blastocystis SSU rRNA genes sourced from GeneBank (NCBI4) representing the 13 established subtypes described from mammals and birds. Distance-based analyses were conducted on readable sequences using Kimura 2-parameter distance estimates and trees were constructed using the Neighbor Joining algorithm using Mega 3.1 (Kumar et al. 2001) software. lacerate (U37108) was used as an out-group. Bootstrap analyses were conducted using 1000 replicates.

2.3.5 Ethics Approval The study was approved by The University of Queensland Medical Research Ethics Committee and Animal Ethics Committee and registered with the Australian New Zealand Clinical Trials registry (www.ANZCTR.org.au) ACTRN12610001066077

2.4 Results 2.4.1 Clinical characteristics of patients and household members Male (n=3) and female (n=8) patients with an age range of 23-71 years participated in the study. The shortest symptom duration was 3 months in one patient, six patients had exhibited symptoms for 2 years and the other five patients had complained of symptoms for more than 10 years. Presenting clinical symptoms were bloating (7/11; 64%), diarrhoea (5/11; 45%), nausea (5/11; 45%), flatulence (4/11; 36%), variable bowel habit (4/11; 36%) abdominal pain (3/11; 27%) and fatigue (2/11; 18%) (Table 2.1)

4 www.ncbi.nlm.nih.gov/

55 Table 2.1 Clinical profile of symptomatic Blastocystis patients

Patient Main Duration Previous Reversible Small Baseline Week 6 ST at symptoms of therapy Causes bowel and week Daily Baseline symptoms Blastocystis? Treated before colonic median median Inclusion in Endoscopy daily BO/WB Study BO/WB

1 Pain 10 years Yes , x2 Giardia Prominent 1/8 1/8 3 Bloating 2 years ago lymphoid follicles TI 2 Bloating 11 years Yes, x2 Enterobius Mild focal 3/8 4/8 3 Alternating 5 years ago vermicularis active colitis BH colonic biopsy

3 Alternating 2 years No N 1/10 1/9 3

BH

Nausea

Aeromonas 4 Diarrhea 2 years Yes, x3 N 3/8 3/8 3 Bloating 1 year ago Low Lactase

Pain

5 Diarrhea 2 years No Colonic 2/9 2/8 4 polyp 6 Nausea 3 months No 2mm 1/7 1/7 1 Fatigue nodule mid- small bowel Colonic polyp

7 Nausea 2 years No Prominent 1/7 1/8 1 Pain lymphoid Alternating follicles TI BH 8 Bloating 10 years Yes, x3 N 1/9 1/9 1 Alternating 10 years ago BH 9 Nausea 2 years No N 2/6 1/8 1 Bloating Alternating BH 10 Bloating 30 years No Ursodeoxycholic Colonic 1/7 1/8 1 Fatigue Acid for small polyp bowel resection

11 Bloating 2 years No N 2/7 1/7 4+6 Alternating BH

BO= Bowel actions per day WB=general wellbeing, 10/10 excellent TI=terminal ileum BH= bowel habit

Five patients had previously received therapy for Blastocystis, which did not clear either the infection or their symptoms. The remaining patients were naïve to therapy. During the period of the study no SBP had any organism other than the Blastocystis noted in the faecal testing. No patient demonstrated eosinophilia on the peripheral full blood film. In total 17 HC submitted specimens; only one teenage HC was unable to donate a sample (three other HC specimens collected from that family). Eight AHC samples were collected, five dogs and three cats from seven households (one of two dogs in one household escaped submitting a specimen). Household water sources were varied. No HC or AHC was undergoing medical investigation or therapy for gastrointestinal symptoms.

56 2.4.2 Endoscopy All patients displayed normal stomach and duodenum and all were negative for Helicobacter pylori at gastric biopsy. Duodenal biopsy showed mild intra-epithelial lymphocytosis in 2/11 (SBP2, 4) patients (Table 2.1). Colonoscopy with ileoscopy was macroscopically normal in 8/11 patients. One patient (SBP1) displayed prominent lymphoid follicles in the terminal ileum, confirmed on ileal biopsy. All the other ileal mucosal biopsies were normal. Random colonic mucosal biopsies were normal in 10/11 patients; SBP2 had mild focal active colitis noted at histology only. Capsule endoscopy looked normal in 9 patients; two patients (SBP1 and 7) displayed prominent distal ileal lymphoid follicles.

2.4.3 Therapy Eight patients were prescribed metronidazole and one patient (SBP7) complained of severe nausea but completed the course. Three patients (SBP2, 4,9) were prescribed trimethoprim/sulfamethoxazole and one patient (SBP7) developed a rash and ceased the tablets on day 11 of the 14-day course.

2.4.4 Faecal smears and culture in patients Faecal smear results prior to the commencement of treatment demonstrated a parasite load of <5 organisms per high-powered field (O/HPF) in five patients and > 5 O/HPF in the remaining patients. All faecal smears or cultures performed at SVS-UQ were positive for Blastocystis at days 14, 28 and 56 (Table 2.2).

57 Table 2.2 Blastocystis faecal smear and culture results in patients

Patient Day 0 Day 14 Day 28 Day 56

UQs Therapy CLs UQs UQc CL s UQs UQc CLs UQs UQc

1 >5 MTZ + >5 + + >5 + + <5 +

2 >5 TS - >5 + + >5 + + >5 +

3 >5 MTZ + <5 + + <5 + n/a >10 +

4 >10 TS + <5 + + >10 + - - -

5 >10 MTZ + <5 + + <5 + + <5 +

6 <5 MTZ + <5 + - - + - <5 +

7 <5 MTZ + <5 + n/a <5 + + <5 +

8 <5 MTZ - <5 + - <5 - - <5 -

9 <5 MTZ - <5 + + <5 + - <5 +

10 <5 MTZ + <5 + + <5 + + <5 +

11 <5 MTZ - <5 + + >5 + + >5 +

MTZ= metronidazole TS=Trimethoprim/sulfamethoxazole c=culture s=smear= number of Blastocystis organisms/high power field in wet faecal smear n/a=not available: Time: 0 days =before antibiotic therapy 14 days = within 48 hours of finishing antibiotic therapy CL=commercial laboratory UQ=University of Queensland laboratory: Faecal smear results: organisms per high power field light microscopy x40magnification: <5, >5, >10

Faecal smears performed at the commercial laboratory (CL) showed negative results in 4/11 (36%), 2/11 (18%) and 3/9 (33%) at 14, 28 and 56 days, respectively. In 3/33 specimens examined at SVS-UQ, the faecal smear result was negative and the culture positive (SBP6 at 28 days; SBP8 at 28 and 56 days). Change in parasite excretion load pre- and post- therapy was variable within the study group (Table 2.2). Four SBP’s (SBP6, 7,9,10) had “occasional leukocytes” noted in less than 50% of each individual’s faecal microscopy specimens, all other specimens were negative for faecal leukocytes.

2.4.5 PCR and genetic sub-typing of Blastocystis in patients All patients were PCR positive for Blastocystis after antibiotic therapy. Sub-typing of Blastocystis showed that all patients retained the same subtype following therapy with antibiotics. Three patients (SBP 4, 9 and 11) were found to have mixed subtype infections and two of these patients appeared to acquire the second subtype after therapy. The STs

58 identified in the patients were ST1 (5/11, 45%), ST3 (4/11; 36%), ST4 (4/11; 36%) and ST6 (1/11; 9%)(Table 2.3).

Table 2.3 Blastocystis subtype results in symptomatic patients pre- and post- therapy

Patient Subtype Subtype before therapy after therapy

1 3 3

2 3 3 3 3 3 4 3 3+4 5 4 4 6 1 1 7 1 1 8 1 1

9 1 1+4

10 1 1 11 4+6 4+6

2.4.6 Faecal smears, cultures and PCR sub-typing in household contacts Seventeen household contacts were tested and 10/17 (59%) contacts were positive for Blastocystis on faecal smear. Blastocystis organisms were low in number and only 1/10 positive specimens contained >5 organisms per high power field. Cultures were positive in 5/17 household contacts (29%). However PCR analysis of these specimens showed that 17/17 (100%) of household contacts had evidence of Blastocystis carriage. Readable DNA sequences allowed subtype determination in all except one household member (SBP11/HC1). The patients and their household contacts had concordant Blastocystis subtypes in 10/11 cases. SBP8 was the only patient to have a different subtype to the household contact (Table 2.4).

59 Table 2.4 Blastocystis subtype results in patients, household contacts and animals

Patient Patient Household Household Household Household Animal Animal Animal Animal Subtypes Contact Fecal Fecal PCR Contact Fecal Fecal PCR Smear Culture Subtypes Smear Culture Subtypes 1 3 HC1 + + 3 2 3 HC1 + - 3 Dog - - 3 HC2 + + 3

3 3 HC1 + - 3 4 3+4 HC1 + - 3 Dog - + 3 5 4 HC1 + + 1+6 Cat + - 1+4 HC2 + - 4 6 1 HC1 - - 1 Dog - - 1

7 1 HC1 + + 7 Dog + - 1 HC2 + - 7 HC3 + - 1 8 1 HC1 - - 6 9 1+4 HC1 - - 1 10 1 HC1 - - 1 Cat1 - - 1 Cat2 - - 1

11 4+6 HC1 - - + Dog - + 1+4 HC2 - + 6 HC3 - - 6

2.4.7 Faecal Smears, cultures and PCR sub typing in animal contacts Seven households had contact with eight household animals. Only 2/8 (25%) animals were positive for Blastocystis on faecal smear, and none were positive on culture. However, all eight animals had positive PCR tests. All of the animals harboured at least one subtype in common with each of the corresponding seven patients. Most animals (7/8) also had subtypes in common with the corresponding household contacts (HC11 was the exception; Table 2.4)

2.4.8 Clinical follow up at 6 weeks Only one patient reported any change in diary entry (SBP9 reported 2 point-rise in general well-being at 6 weeks post therapy; Table 2.1). No household contact or animal was reported to have any change in gastrointestinal health.

60 2.5 Discussion In this study, none of the eleven patients with symptomatic succeeded in clearing infection following a 14-day course of metronidazole or trimethoprim/sulfamethoxazole. Faecal Blastocystis organisms have been detected 2 days after oral ingestion of cysts in mice (Moe et al. 1997). Persistence of the parasite within 48 hours of ceasing antibiotics in this study suggests failure to clear the parasite and not re- infection. Reported rates of eradication of Blastocystis using these antibiotics vary widely in the literature from 22-100% (Ok et al. 1999; Moghaddam et al. 2005). High false- negative rates of Blastocystis detection using routine faecal smears have previously been reported (Stensvold et al. 2007a) and false negative rates of 50% were confirmed in this study. Light microscopy is an unreliable tool for detecting this polymorphic parasite that exhibits highly variable daily faecal excretion rates (Suresh et al. 2009). The vacuolated form, which is the most readily identified form in the faecal smear, may have a 10-fold variation in size (Zierdt 1988) and small vacuolated forms, amoeboid or cystic forms (2-5 m) with low excretion rates may not be readily identified (Suresh & Smith 2004). Large clinical trials using more sensitive diagnostic tests such as PCR in addition to conventional methods will be required to accurately evaluate the in vivo efficacy of the numerous agents (Stensvold et al. 2010) that have been reported to be useful in treating Blastocystis and to evaluate new antimicrobials. The possible causes of antimicrobial failure in clinical Blastocystis cases are numerous. Metronidazole and trimethoprim have been studied in vitro (Zierdt et al. 1983; Dunn & Boreham 1991; Vdovenko & Williams 2000) and, although both drugs inhibit growth of Blastocystis, higher doses are needed to be cytocidal. Delivery of a cytocidal dose of the antimicrobial to the organism in vivo may be problematic. Both antibiotics are rapidly absorbed systemically, primarily in the small bowel (70% of a single oral dose of both agents is excreted in urine), and little active drug reaches the colon. In animal studies Blastocystis has been found primarily in the caecum and proximal colon (Phillips & Zierdt 1976; Moe et al. 1997). Endoscopic reports in humans demonstrate viable Blastocystis organisms in caecal fluid and in biopsy specimens of the large bowel mucosa but in humans generally the organism is not invasive (Zuckerman et al. 1994). Studies in humans do not report Blastocystis in the stomach or duodenum and no abnormality of the small bowel was noted in this study. It is likely that drugs effective against Blastocystis will need to be delivered in high concentration to the lumen of the caecum. Blastocystis subtypes may also vary in their susceptibility to various antimicrobial agents. Blastocystis STs have been shown to differ in their morphology and growth rates

61 (Cassidy et al. 1994; Tan et al. 2008) and a recent in vitro study has reported variation in metronidazole sensitivity between different subtypes in in vitro culture (Mirza et al. 2010). Studies assessing drug sensitivities may need to take account of the ST and regional origin of the organism. The pathogenicity of Blastocystis is uncertain. Blastocystis is commonly found in the stools of healthy people, which is in agreement with observations of healthy, yet positive household members in this study. The prevalence in an asymptomatic population in Australia has been reported to be 6% (Hellard et al. 2000) assessed with stained faecal smears, but is reported to be as high as 20-30% in other nations (Yaicharoen et al. 2005; Li et al. 2007b) when using faecal culture and DNA detection tools. However, the existence of healthy carriers does not necessarily exclude pathogenicity. Common parasitic pathogens, such as Giardia duodenalis and Entamoeba histolytica, are known to cause symptomatic disease in less than 50% of infected people (Pickering et al. 1984). Studies have found Blastocystis more commonly in the stools of patients with Irritable Bowel Syndrome (Yakoob et al. 2004; Yakoob et al. 2010), and Blastocystis has been linked to outbreaks of diarrhea in travellers and soldiers in camp (Jelinek et al. 1997; Leelayoova et al. 2004). Most large studies of symptomatic Blastocystis patients describe a non-inflammatory watery diarrhea with no mucosal ulceration or invasion (Zuckerman et al. 1994; Jelinek et al. 1997). This study is the first to examine the small bowel in Blastocystis infection, as previous studies have not reported inspecting the terminal ileum at lower endoscopy (Kain et al. 1987; Zuckerman et al. 1994; Chen et al. 2003). Prominent lymphoid tissue in the terminal ileum was the only mucosal abnormality seen in a minority of patients. One explanation for the apparent clinical paradox of symptomatic and asymptomatic infected household members may be that subtypes of Blastocystis vary in their pathogenicity, notably ST3 (Tan et al. 2008) and ST1 (Kaneda et al. 2001; Tan et al. 2006; Yan et al. 2006; Hussein et al. 2008; Eroglu et al. 2009; Stensvold et al. 2009c) has been reported more frequently in symptomatic patients. Other studies have not confirmed any significant difference in the distribution of particular subtypes between symptomatic and asymptomatic Blastocystis carriers or specific diseases (Yoshikawa et al. 2004a; Dogruman-Al et al. 2009; Elwakil & Talaat 2009; Rene et al. 2009; Souppart et al. 2010). Although ST1, rather than ST3 was the most common in our small study we found no evidence that a particular 16SSU rDNA subtype of Blastocystis was more likely to be associated with symptoms. If particular Blastocystis strains are pathogenic, then the differences in the organism may not be apparent with our current sub-typing tools directed

62 at rDNA of the dominant faecal strain or it may be that individual host/parasite interaction determines pathogenicity. The parasite load may also be a significant factor in causing disease and no reliable guide to quantification exists. Although previous studies have suggested that Blastocystis infection is considered more severe if five or more organisms per HPF are recovered on faecal smear, this measure could be unreliable considering the wide variation in daily excretion of the parasite (Suresh et al. 2009). In this study, human and animal household contacts had fewer parasites positively documented on smear and a lower rate of successful culture of Blastocystis. Many aspects of the life cycle of Blastocystis are unknown. At present it seems likely that only the thick walled cyst, which survives water chlorination (El-Shazly et al. 2005a), transmits infection (Moe et al. 1997). In this study the universal carriage of Blastocystis in household contacts and the concordance of subtypes suggest that Blastocystis is easily transmitted. A recent study has also reported concordance of STs between pets and their owners (Eroglu & Koltas 2010). Even if the primary infection is not acquired from the household animal these animals are likely to remain a reservoir for re- infection.

2.6 Conclusion The universal failure of single antimicrobials in this pilot study suggests that future studies may need to investigate efficacy of combinations of antimicrobial therapy as well as clearance rates in different patient subgroups (such as length of time of acquisition of infection or diarrhoea predominant illness). The high concordance of Blastocystis subtypes within households warrants further investigation in a larger study using more specific molecular tools to allow definitive comments on routes of transmission.

2.7 Acknowledgements Acknowledgements: Dr Jenny Robson, Microbiologist, Sullivan and Nicolaides Competing Interests: Nil Funding: Early Career Research Grant awarded to Dr R Traub, University of Queensland Trial Registry: The University of Queensland Medical Research Ethics Committee and Animal Ethics Committee and Australian New Zealand Clinical Trials registry (www.ANZCTR.org.au) ACTRN:12610001066

63

64

3 Axenic culture of Blastocystis spp.

65 Axenic cultures 3.1 Abstract Although Blastocystis spp. are the most common enteric parasite found worldwide there is a dearth of basic information about the biology of the parasite. Axenic cultures provide a simple system to investigate molecular, biochemical processes and antimicrobial sensitivities of Blastocystis. Historically axenisation has been difficult to achieve, and there is limited quantity and diversity of axenic stock available worldwide. Three previously reported successful techniques to develop axenic cultures of Blastocystis were trialled, namely culture into solid agar, soft agar and preliminary partial elimination of contaminating bacteria using a Ficoll-metrozoic acid gradient. Although growth in agar was achieved, further propagation of the parasite was not successful. One major factor in this failure was likely the quantity of inoculum of the parasite. Future success would likely be improved if the cultures were attended daily, only stool with large numbers of Blastocystis was selected for culture and antimicrobial sensitivity on the enteric flora was determined and used to select antibiotics. However, the most important step forward will be to understand the life cycle of this parasite and the parasite forms or priming factors necessary for reproduction. Despite documenting sufficient numbers for successful inoculation at six weeks of culture, it is well recognized that aging cultures will not succeed. Schizogony is a reported but not accepted mode of reproduction of Blastocystis. The apparent schizonts seen in this study cannot be confirmed without demonstrating genetic material within the daughter cells.

3.2 Introduction to axenic culture 3.2.1 Why attempt axenic culture? Blastocystis organisms living in a sterile, artificial environment are likely to differ in basic metabolic processes and possibly even life cycle from those living in and interacting with the luminal and microbial environment of the gut. Nevertheless, axenic cultures (homogenous culture devoid of any other contaminating microorganisms) provide a simpler model to investigate molecular and biochemical processes and the antimicrobial sensitivities of Blastocystis spp. However, reliable development of diverse axenic cultures of Blastocystis has been difficult to achieve and improving axenic culture protocols has been identified as one of the most important undertakings needed to improve our knowledge of this parasite (Poirier et al. 2012).

66 3.2.2. Early experience with axenic culture using antibiotics Axenic culture of Blastocystis sp. was first described in 1921 with the aim of proving that this parasite was a distinct species and not a degenerate faecal form of other parasites, or vegetable matter (Barret 1921). Twenty-five subcultures were successfully achieved in anaerobic conditions at the base of a test tube using a 10% human serum and 0.5% sodium chloride solution incubated at 37C. Bacterial overgrowth and subsequent culture death was commonly observed in the unsuccessful attempts of axenisation in this study. Subsequent studies cultured the parasite in small amounts with difficulty but could not isolate the parasite from bacteria (Ciferri & Redaelli 1938) or transfer to a solid media (Lynch 1922; Knowles & Das Gupta 1924). Finally in 1974 axenic culture of significant quantities was achieved in a strict anaerobic environment by the slow addition of antibiotics over months, with gradual reduction of the contaminating bacteria to three or four species and then final bacterial eradication (Zierdt & Williams 1974) using a bi-phasic medium consisting of egg medium slants overlayed by Locke’s solution (Boeck and Drbohlav (B-D) inspissated egg medium) (Boeck & Drbohlav 1925; Clark & Diamond 2002). Despite slow withdrawal of the antibiotic many cultures died when bacteria were finally eliminated. Later studies attained culture of significant quantity of parasite in a monophasic liquid medium (Molet et al. 1981).

3.2.3 Physical separation of contaminating microorganisms Initial physical methods of differential centrifugation to separate out the contaminating bacteria damaged the fragile Blastocystis form that “bursts like a balloon” (Zierdt 1991). Attempts to improve culture yields and reduce contaminating organisms by layering culture sediments on a Ficoll-metrozoic acid gradient were reported to be partially successful (Upcroft et al. 1989). A later study combined this separation method with repeated exposure to antibiotics selected according to the bacterial contaminants antimicrobial susceptibility cultured in a B-D biphasic medium (Lanuza et al. 1997). This study developed 81 different cultures of Blastocystis and reported successful axenisation in 25 (31%). These axenic cultures had a mean generation time of 10.5 hours, took 3-5 weeks to axenise and were observed for seven days to confirm no bacterial contamination. The final harvest was six tubes of axenic Blastocystis organisms with a concentration of Blastocystis organisms varying from 0.25-2.5 x 106 per tube. The broadspectrum antibiotics used included cephtazidine, ciproflacin, fosfomycin, imipenen and vancomycin. The authors reported that consecutive addition of antibiotics was more successful than adding a combination of antibiotics together.

67 The authors had no explanation for the failure of 70% of the cultures to axenise, but suggested that this may support previous suggestions that some strains of Blastocystis are dependent on companion bacteria for growth (Zierdt 1991; Boreham & Stenzel 1993).

3.2.4 Agar colonies To date the most successful axenic culture technique producing significant quantities of Blastocystis in a timely manner (three weeks) has used physical separation of colony growth on the surface of agar to complete the separation of the parasite from contaminating bacteria before expansion in liquid medium (Ng & Yan 1999). This method was further enhanced to increase clonal yield on the agar plates (Tan et al. 2000). These colonies grew to 3mm diameter in size by day eleven and were distinguishable from bacterial colonies by colour, domed shape and mucoid texture with maximum growth noted at day four or five. A later technique described successful growth after eight days in soft agar, but separation from the agar and transfer to liquid medium was more problematic (Valido & Rivera 2007). These agar colonies have been examined microscopically and appear to have an organised structure. Light microscopy showed large irregularly shaped amoeboid forms (AF) on the periphery of the colony with smaller (around 10m) mostly vacuolated forms (VF) at the centre (Valido & Rivera 2007). Scanning electron micrographs show two regions in the colonies, a central domed region and a flattened peripheral region (Tan et al. 2000) (Figure 3.1). A fine surface coat was present on most cells and this coat thickened as the colony aged. Figure 3.1 Scanning electromicrograph of Blastocystis colony grown on solid agar (Tan et al. 2000)

D: dome P: periphery of colony

68 3.2.5 Requirements for successful axenisation Successfully culturing vigorous axenic cultures, which have been subcultured many times without requiring antibiotic supplementation to prevent subsequent bacterial contamination, is difficult. All studies have shown that strict anaerobic conditions are required, even to the extent of pre-reducing the medium for 48 hours. The size of the inoculum is important, requiring an initial inoculum of at least 1-1.5 x106 Blastocystis cells (Zierdt & Williams 1974; Ho et al. 1993), although surprisingly, further propagation was possible by transferring only a small number of cells (<100) from one established agar colony to another agar plate. The type of cell that is transferred is important, a mixture of VF and granular forms is preferred, VF alone usually fail (Zierdt & Williams 1974). Media with different osmolality have been used and indicate that Blastocystis is tolerant in this regard. The age of the seeding culture is also important, cultures 7 and 14 days propagate well but those after 20 days usually fail. There are also local factors that may be important including the type of culture containers (some glass and plastic tubes are not suitable) (personal communication, Dr K.W.S Tan, NUS, Singapore). Currently many laboratories around the world would like to have readily available axenic cultures of Blastocystis spp.. Subtype 4 (WR1), ST1 (Netski), and ST7 are currently subcultured successfully in Department of Microbiology, National University Singapore, Singapore (Dr K.W.S Tan).

3.3 Methods 3.3.1 Sample selection and management Fresh faecal specimens from patients positive for Blastocystis, as determined by light microscopy and later confirmed with PCR (sixty five patients with a range of subtypes 1-8, 45 symptomatic for gastrointestinal symptoms), and from asymptomatic pigs were collected and subcultured within 2-6 hours of passage of stool. These specimens were subjected to preliminary subculture in high quality polypropylene test tubes (Sarstedt 10mL Cat no: 83.9923.929) and then transferred to agar for further culture. Two methods of axenic culture were used in our experiments, namely the solid agar (Ng & Yan 1999) and the semi-solid agar (Valido & Rivera 2007) method. Enrichment of Blastocystis passage was attempted using Ficoll-metrozoic acid (Histopaque, Sigma-Aldrich, cat no:10771) gradient to concentrate the organisms.

69 3.3.2 Preliminary subculture one A small amount of fresh faeces (approximately 10 gm) was cultured aerobically for 48 hours in culture solution of Jones medium (Appendix 3.1) /10% heat inactivated horse serum (HS) (Gibco, Life technologies, Cat No: 26050-088). Sediment (0.5-1.0 ml) from the base of each test tube was subcultured anaerobically using pre-reduced Iscove’s Modified Dulbecco’s medium (IMDM)/10%HS medium supplemented with antibiotics every 2-3 days. Anaerobic conditions were obtained using an anaerobic pouch GasPakTM (B-D, cat. No: 26001) system in a sealed jar. The sediment was examined microscopically and when a strong growth of Blastocystis was observed at the second, third or fourth subculture with relatively little bacterial contamination this was transferred to agar or subjected to Histopaque concentration.

3.3.3 Preliminary subculture two The same method as above was employed except that following culture in Jones medium the sediment was transferred to B-D biphasic medium (Appendix 3.2), subcultured every three–four days and then subsequently transferred to agar.

3.3.4 Antibiotic supplementation Antibiotics were added to the medium to give a final concentration of claforan 2,000 g/ml, ampicillin 4,000 g/ml, streptomycin 1000 g/ml and benzylpenicillin 100 IU/ml. When fungal overgrowth was seen amphotericin 6 g/ml was added.

3.3.5 Monitoring the subcultures Six symptomatic patients had their preliminary subcultures monitored for growth and morphological form over time. The pellet at the bottom of the tube was gently dispersed evenly in the whole test tube and cells were examined by light microscopy and counted using a haemocytometer.

3.3.6 Solid agar culture BactoTM agar (Difco, Cat no. 214520) prepared as a 3.6% solution with 0.1% sodium thioglycollate (Sigma-Aldrich, cat no. T0632) was sterilised in an autoclave and then kept at 56C in a waterbath. Two mL of this agar was added to 18ml of warmed IMDM/10% inactivated HS combined with ampicillin (50,000 g) and streptomycin (1,000 g) in a 50 mL centrifuge tube and gently mixed. Sediment from the preliminary subculture (0.5-1.0 mL) was added, the tube gently swirled and the mixture poured into a 100x15 mm

70 petri dish. A small amount of sediment (0.10 mL) was also sprayed onto the surface of the agar. The petri dish was placed in the freezer at -20C until the agar was set (approximately 3-5 minutes), then transferred to an anaerobic jar with GasPakTM and incubated undisturbed at 37C for ten days. At day ten the plate was observed for colony growth, if no growth was seen the plate was re-incubated for a further 4-7 days and then discarded if no growth was seen. A randomly selected sample of agar was examined with a light microscope to exclude unrecognised growth of Blastocystis colonies. If colony growth was observed the colonies were picked up separately with a sterile pipette and transferred gently to 2 mL microcentrifuge tubes containing pre-reduced IMDM/HS (with varying concentrations of HS from 10-30%), centrifuged at 11,000g for 10sec in a microfuge to remove the agar and incubated at 37C in IMDM/HS 10% for further subculture and propagation at 3-7 days.

3.3.7 Semi-solid agar culture BactoTM agar (Difco, Cat no. 214520) prepared as a 1.5% solution with 0.1% sodium thioglycollate (Sigma-Aldrich, cat no. T0632) was sterilised, and kept at 56C in a waterbath. The agar preparation and IMDM were added in equal proportion to a 50 ml centrifuge tube to a volume of 18 ml and two mL of inactivated HS and combined with ampicillin (50,000 g) and streptomycin (1,000 g) were added. Sediment from the preliminary subculture (0.5 mL) was gently pipetted onto a petri dish and 20 ml of agar mixture gently poured on top. The Petri dish was then treated as per the solid agar regimen.

3.4 Results Preliminary cultures were attempted 210 times between 15/2/11 and 23/07/13. Bacterial and fungal overgrowth of cultures was common despite antibiotics. Growth of subcultures was more vigorous using the biphasic medium, but it appeared more difficult to reduce the numbers of contaminating bacteria in this medium. Subcultures with good growth of Blastocystis were placed into agar 32 times with two solid agar plates (Plate 1 and Plate 2) derived from the same pig specimen growing small numbers of colonies (1- 3mm in diameter) (Figures 3.2, 3.3) and with two instances of poor growth in semi-solid agar from human and pig specimens (Figure 3.4).

71 Figure 3.2 Blastocystis colonies grown on solid agar 17/8/12 School of Veterinary Science, University of Queensland

Figure 3.3 Light microscopy of Blastocystis colony grown on solid agar 17/8/12

Arrows indicate cells possibly undergoing schizogeny Black circle indicates the direction of the centre of colony Bar: 10m

72 Figure 3.4 Light microscopy of Blastocystis colony grown in soft agar 17/7/12 School of Veterinary Science, UQ Large numbers of amoeboid cells are seen in the centre and round clusters free at the edge.

None of these partially successful cultures had had cells separated with HistopaqueTM. None of the Blastocystis organisms from Plate 1 could be further propagated in liquid media and the cultures died after 2-4 weeks. Plate 2 was placed back into the incubator for further colony growth, but examination a week later showed that all colonies had died (presumably the brief exposure to oxygen was detrimental to the organisms). Light microscopy showed mainly AF towards the centre of the colony or round VF free or very close to the periphery of the colony. Some of the cells had prominent granules at the edge of the vacuoles and one appeared to show round inclusions shaped like daughter cells inside the main cell suggesting reproduction via schizogony (Figure 3.3, arrow). Cell counts in the preliminary cultures fluctuated over time, showing peaks at 2-3 weeks and again at 5-6 weeks before decreasing (Table 3.1, Figure 3.5). Our VF cell counts were never high but reduced with time over the life and demise of the culture. As the VF and AF counts fell the GF counts increased with the aging culture.

73 Table 3.1 Blastocystis cell counts in fresh faecal cultures derived from 6 patients with gastrointestinal symptoms, subcultured every 2-3 days, in Jones/Horse serum 10% seven days, then IMDM/Horse serum 10% for following 11 weeks

Patient 1 VF AF GF Patient 2 VF AF GF Patient 3 VF AF GF Baseline 1 5 277 8 16 72 1 7 83 Week 1 0 8 271 2 17 188 0 4 275 Week 2 0 0 344 0 1 178 0 10 109 Week 3 0 2 656 0 5 524 1 4 202 Week 4 7 3 273 0 7 276 1 0 453 Week 5 1 10 858 0 8 742 2 3 412 Week 6 0 6 111 0 1 294 0 0 336 Week 7 0 0 50 0 0 199 0 0 157 Week 8 0 0 200 0 0 377 0 0 276 Week 9 8 1 277 2 0 226 1 0 231 Week 10 0 0 121 0 0 199 0 0 129 Week 11 0 0 15 0 0 27 0 0 49

Patient 4 VF AF GF Patient 5 VF AF GF Patient 6 VF AF GF Baseline 0 7 193 1 14 162 0 13 333 Week 1 0 18 399 0 5 335 0 3 337 Week 2 3 5 305 1 3 249 2 4 487 Week 3 0 14 580 0 1 468 1 7 379 Week 4 0 8 690 0 4 436 1 5 403 Week 5 1 4 640 0 3 777 1 3 784 Week 6 0 6 463 2 7 459 0 1 118 Week 7 0 0 218 0 0 400 0 0 212 Week 8 0 0 286 0 0 446 0 0 211 Week 9 1 0 265 0 0 216 1 1 127 Week 10 0 0 88 0 0 100 0 0 47 Week 11 0 0 51 0 0 16 0 0 25

cell count x104/mL VF: vacuolated form AF: amoebic form GF: granular

74 Figure 3.5 Blastocystis cell counts in fresh faecal cultures derived from six patient with gastrointestinal symptoms, subcultured every 2-3 days, in Jones/Horse serum 10% seven days, then IMDM/Horse serum 10% 3.5.Patient 1 3.5.Patient 2 3.5.Patient 3

1000 1000 500

VF 500 0 VF VF VF AF AF AF 0 0

GF VF GF VF GF

Week 2 Week

Baseline

Week 4 Week

Week 6 Week

Week 8 Week Week 10 Week

3.5.Patient 4 3.5.Patient 5 3.5.Patient 6

1000 1000 1000

VF VF VF 0 AF 0 AF 0 AF VF VF VF

GF GF GF

Week 2 Week

Week 4 Week

Baseline

Week 6 Week

Week 8 Week Week 10 Week

y axis: cell count x104/mL x axis: time in weeks VF: vacuolated form AF: amoebic

75 3.5 Discussion I failed to develop an axenic culture of Blastocystis and it would be useful to understand why the cultures failed. The limited success that I had occurred using pig samples that generally had greater numbers of Blastocystis in the stool compared to humans. Only four of the total of 65 patients had large numbers (3+ load of VF) of Blastocystis noted in their stool and only one of these patients was included in the six patients monitored with cell counts. Four of the six patients that had cell count monitoring all had baseline inoculum above the 1-1.5 x106 (Zierdt & Williams 1974; Ho et al. 1993) suggested to be necessary for successful transfer. The remaining two patients were only slightly below this level. Nevertheless it is likely that the higher the number of organisms in the inoculum the greater the chance of success. The timing of subculture is also known to be important and working on the project part time prevented daily assessment of the subcultures. It is possible that some subcultures would have been more vigorous, if they had been transferred after 24 hours and this may have reduced bacterial overgrowth. Although we used antibiotics in a combination used in previous studies it is likely that during the last 10-20 years there has been significant changes in the enteric bacterial antimicrobial sensitivity spectrum. Future attempts could attempt to identify the contaminating organisms and their antimicrobial sensitivities. Sequential addition of antibiotics rather than a cocktail may be more successful. The HistopaqueTM gradient did separate out Blastocystis from the contaminating bacteria for further subculture, but it was apparent that much smaller numbers of organisms were obtained with this method. Failure may again relate to number of organisms available for inoculation into the culture. Although we did achieve one culture on the solid agar the number of colonies was small although the diameter of the colonies was similar to that reported previously. The failure to progress from there to further subculture may have been due to residual agar around the cells that prevents growth (Ng & Yan 1999). In the future it would be useful to use the agar colonies, as these require a much smaller inoculum, to seed another agar plate in order to obtain another agar plate with stronger growth. We are all constrained in this work by the lack of knowledge of the life cycle of the parasite. The fact that inoculating large numbers of VF into culture is not as useful as a mixture of forms (Zierdt & Williams 1974) suggests that the VF themselves either need priming to reproduce or are not the primary mechanism of cell reproduction. The VF is

76 characteristic of subcultures (particularly attenuated axenic cultures) but is not the predominant form seen in fresh faecal cultures, and was reported to be rarely seen in a patient with severe diarrhoea and profuse excretion of Blastocystis organisms (Zierdt & Tan 1976b). Transmission of infection is reported to occur via the cystic form of Blastocystis (Zierdt 1991). Transformation of VF to the cystic form can be achieved by culturing in excystation mediums (Suresh et al. 1994a). Future attempts at axenic culture could include early subculture in encystation mediums that may help to stimulate the VF to progress along the path towards encystment. There appeared to be at least a bi-modal distribution of cell growth over time in our monitored patients. Most patients had highest cell numbers at 2-3 weeks and then again, usually a smaller peak at 5-6 weeks before the culture waned. Peak growth at 10-15 days was a consistent finding in other uncounted cultures. Yet, even though the cell numbers are still high at 5-6 weeks some vital factor is lost as it is reported that it is not useful to subculture after 20 days (Zierdt & Williams 1974). The only mode of reproduction of Blastocystis that is currently generally accepted is binary fission although other modes of asexual transmission namely plasmotomy, endodyogony and schizogony have been described (Zierdt 1991; Zhang et al. 2007). Our axenic culture showed two cells with multiple round/crescentic inclusions that looked like daughter cells. A number of authors (Tan & Stenzel 2003; Windsor et al. 2003) have urged caution in interpreting these forms. It has been suggested that schizonts reported previously may be stressed cells undergoing degeneration, exuberant granular inclusions within the parasite or other unknown faecal forms. No definite confirmation of schizogony has been possible with electron microscopy to date. The cells we have imaged were in a pure uncontaminated culture and appeared to be thriving. Only two cells in the whole slide had this appearance, suggesting that if it does represent schizogony it may be uncommon. However Blastocystis is polymorphic and many forms have multiple granular inclusions. It will be necessary to document nuclear material in these daughter inclusions in order to prove schizogony.

3.6 Conclusions Axenic culture of Blastocystis remains an important goal but is problematic. Understanding the life cycle of the parasite will aid progress with axenic culture methods.

77

78

4 Features of Blastocystis spp. in xenic culture revealed by deconvolutional microscopy

Robyn Nagel, Christian Gray, Helle Bielefeldt-Ohmann, Rebecca J Traub, 2015. Features of Blastocystis spp. in xenic culture revealed by deconvolutional microscopy. Parasitol Res 114: 3237-3245.

-presented as published

79 Features of Blastocystis sp. in Xenic Culture Revealed by Deconvolutional Microscopy

Robyn Nagel1, Christian Gray2, Helle Bielefeldt-Ohmann1,3, Rebecca Traub 4

1 School of Veterinary Science, University of Queensland, Gatton, Queensland, Australia 2 Plymouth University Peninsula schools of Medicine and Dentistry, Plymouth University, Plymouth, United KIngdom 3 Australian Infectious Diseases Research Centre, University of Queensland, Queensland, Australia 4 Faculty of Veterinary and Agricultural Sciences, University of Melbourne, Parkville, Victoria, Australia

4.1 Abstract Blastocystis organisms are a common human enteric parasite and have been reported to cause irritable bowel-like symptoms. Previous microscopy studies have described Blastocystis to possess complex morphology and ultrastructures. Cultured stool samples from patients with irritable bowel syndrome and asymptomatic healthy pigs were examined using time-lapse imaging and fluorescent spectroscopy on a deconvolutional microscope. Blastocystis organisms of the vacuolated, granular, amoebic and cystic forms were observed to autofluoresce in the 557/576 emission spectra. Autofluoresence could be distinguished from fluorescein-conjugated Blastocystis-specific antibody staining in vacuolated and granular forms. This antibody stained Blastocystis subtypes 1, 3 and 4 but not 5. Surface pores of 1m diameter were observed cyclically opening and closing over 24 hours. Vacuolated forms extruded a viscous material from a single surface point with coincident deflation that may demonstrate osmoregulation. Tear-shaped granules were observed exiting from the surface of an amoebic form but their origin and identity remain unknown. Key words: Blastocystis; deconvolutional microscopy; autofluorescence

4.2 Introduction Blastocystis spp. are common enteric, unicellular parasites found in almost every species of animal (Tan 2008). Seventeen different subtypes (ST’s) defined by the 18SSU of the ribosomal RNA gene are recognised and ST3 is almost universally the most common of the nine ST’s found in humans (Stensvold 2012). Blastocystis infection has been linked to irritable bowel

80 syndrome-like symptoms in humans but epidemiological studies are inconclusive (Scanlan 2012b). Many healthy people carry Blastocystis and it remains unclear if the parasite is pathogenic. Pathogenicity has been speculated to relate to parasite subtypes (Tan et al. 2006; Eroglu & Koltas 2010) as well as to the host’s immune response (Olivo-Diaz et al. 2012). Analysis of the genome of two different ST’s of Blastocystis has allowed prediction of genes and proteins (Denoeud et al. 2011; Wawrzyniak et al. 2015) that may be associated with the organism’s pathogenic potential. Nevertheless, a lack of knowledge of the basic life cycle and metabolic function of the parasite remains a major limitation to exploiting the use of animal models and in vitro systems to further explore the clinical significance and therapeutic options for the control of this organism. Studies utilising light microscopy, transmission (TEM), scanning (SEM) and freeze-etch (FE-EM) electron microscopy have described multiple forms of Blastocystis, of variable sizes, including vacuolated (VF), granular (GF), amoebic (AF) and cystic (CF) forms (Figure 4.1).

81 Figure 4.1 Morphological forms of Blastocystis sp. In xenic culture stained with acridine orange

Bar 10m solid arrow:granular forms dashed arrow:vacuolated form Star: ameboid forms

82 The relationship of these different forms to each other is unclear (Tan 2008), although it is certain that the robust cystic form transmits infection (Moe et al. 1997). Microscopic images have often been obtained from attenuated or axenic cultures. These elegant studies have been useful in describing the intricate ultrastructure and surface morphology of the various forms of Blastocystis but their limitation is that they capture still images of a dead organism separated from the usual microbial environment. In this study, we employed deconvolutional microscopy of xenic cultures of live Blastocystis to obtain time lapse and 3-dimensional images of Blastocystis organisms. This microscope also has the capability to record fluorescence in various light spectra facilitating utilisation of Blastocystis- specific fluorescent antibodies and fluorescent double stranded deoxyribonucleic acid (DNA) stain, DAPI (4’,6-diamidino-2-phenylindole) (Sigma-Aldrich, Australia).

4.3 Materials and methods 4.3.1 Sample preparation Fresh faecal specimens were obtained from irritable bowel syndrome patients positive for Blastocystis carriage and from pigs at the University of Queensland (UQ) Gatton Campus piggery, obtained in accordance with University of Queensland Medical and Animal Ethics Committee approvals numbers 2011000454 and 2012000069, respectively. Ten grams of faeces was subcultured anaerobically for 24-48 hours in Jones medium (Jones 1946) supplemented with 10% heat-inactivated horse serum. The sediment at the interface between the basal residue in the test-tube and the liquid medium was removed for microscopic analysis.

4.3.2 Slide preparation 15µL of fresh sediment was mixed gently with 10µL phosphate buffered saline and placed on a glass slide. The edges of the slide cover were sealed to prevent air entry to the specimen. An inverted slide was placed into a Deltavision Elite deconvolution microscope (Applied Precision, GE

83 Healthcare, Berthold Australia, Bundoora, Victoria, Australia) and examined with polarised light and with fluorescent filters. Four different filters that allowed excitation and emission ranges from 350 nm to 700 nm, including DAPI, FITC (fluorescein isothiocyanate), TRITC (tetramethylrhodamine) and Cy5 (cyanine 5) spectroscopy ranges were used for examination. The images were magnified 20 times or 40-60 times with oil immersion and repeated images were taken at cross sections through the specimen or at different time intervals. Time-lapse microscopy was performed over 24 hours with images taken every 15 minutes. The environmental chamber was maintained at 37C throughout the microscopy session.

4.3.3 Antibody staining Fluorescein (FITC)-conjugated Blastocystis specific polyclonal antibody raised against axenic ST3 (Blastofluor; Antibodies Inc, David, CA, USA) was also used to assess the specimens. Two hundred microliters of sediment was incubated with 5µL of Blastofluor for thirty minutes at 37C. Staining of double stranded deoxyribonucleic acid (DNA) with DAPI was performed by adding 2µL of stock solution of DAPI (0.5ug DAPI in 1mL PBS) to the prepared microscopy slide and gently mixed before incubating in the dark for 10 minutes.

4.4 Results and Discussion 4.4.1 Auto and Immuno fluorescence Autofluorescence was observed in VF, GF and AF and the CF (Figure 4. 2) of Blastocystis. The autofluorescence was greatest in the TRITC excitation/emission light spectra (557/576-nM) and appeared to be brightest in the central vacuole, with spots of light emission observed at the surface of the organism. The autofluorescence faded with time and appeared to be optimal in fresh VF forms, CF and least in AF. The autofluorescence was distracting but could be distinguished from the Blastofluor stain in the VF and GF. The Blastofluor stained the external cell membrane of VF, GF but not the AF (Figure 4.3) and was brightest using the FITC (495/515-nm) light spectrum.

84 Figure 4.2 Autofluoresence of different morphological forms of Blastocystis spp.

VF (ST4) GF (ST4) AF (ST 6) Cyst (ST 6) Polarised light

TRITC fluorescence spectra

TRITC: tetramethylrhodamine VF: vacuolated form GF: granular form AF: amoebic form CF cystic form Horizontal bar equals 20 m ST: subtype

85 Figure 4.3 Different morphological forms of Blastocystis spp. stained with Blastofluor

VF VF GF cluster GF cluster

Polarised light Polarised light Polarised light TRITC: No Blastofluor

FITC/Blastofluor FITC/Blastofluor FITC/Blastofluor FITC/Blastofluor VF: Vacuolated form GF: Granular form FITC: fluorescein isothiocyanate spectra TRITC: tetramethylrhodamine spectra Bar: 20m

86 The Blastofluor stain could not be reliably differentiated from the strong autofluorescence seen in the CF. The Blastofluor antibody stained the VF and GF of ST’s 1, 3 and 4 specimens but did not stain these forms in ST5 pig samples. DAPI staining of VF stained the nuclei on opposite poles of the central vacuole a dense bright blue. Non-specific green autofluorescence (GAF) was originally thought to be a rare occurrence in microalgae that might prove a useful taxonomic feature. However, over time GAF has been identified in many different organisms including algae, dinoflagellates, , cyanobacteria and (Tang & Dobbs 2007). The GAF occurs in the cytoplasm of these microalgae as well as in cysts that have little cytoplasm and is affected by various types of stress on the cell such as light, heat, drying and is caused by molecules other than chlorophyll. The chemical nature of GAF is unknown but is likely caused by multiple compounds or multiple derivatives of a single compound, such as flavin-like molecules, or luciferin compounds (Tang & Dobbs 2007). A variety of stains that fluoresce in the green light spectrum are used to assess morphology and physiological states of organisms and GAF has been reported to interfere with the interpretation of these stains (Tang & Dobbs 2007). This is the first report of autofluoresence by Blastocystis organisms. Blastocystis diverged in evolution at the time of red algae and the genome suggests incorporation of cyanobacteria genes (Denoeud et al. 2011) so the presence of GAF is not unexpected. The GAF was seen strongly in the central vacuole of the VF. Little is known of the chemical constituents of this vacuole except that it stains positively for carbohydrates (Yoshikawa et al. 1995a) and lipids (Yoshikawa et al. 1995d). The GAF did not interfere with the ability to detect Blastofluor staining at the surface of the organism in the VF and GF’s but could not be differentiated from GAF in cysts. Blastocystis cysts are difficult to identify in stools but as GAF is a relatively non-specific finding in microalgae screening with fluorescent filters may not aid detection of the CF. Blastofluor did not stain the AF of Blastocystis suggesting that the surface antigens of this form differ from the VF and GF. Blastofluor is a polyclonal antibody raised in

87 rabbits to axenic cultures of ST3 and has been shown to react to Blastocystis organisms with no cross reactivity with Giardia duodenalis, Entamoebae histolytica or Cryptosporidium spp. (Dogruman-Al et al. 2010). Blastocystis immunolabeling with Blastofluor has previously been shown to stain faecal samples of steers, goats and humans, but not pigs after prolonged (up to a year) storage in formalin (Gould & Boorom 2013). This is consistent with our own observations in which no Blastofluor staining was detected in our ST5 pig samples. ST5 organisms are commonly detected in pigs worldwide and locally 100% pigs have previously been shown to carry Blastocystis sp. ST5, and 7% harbour mixed infection with ST1 (Wang et al. 2014c). This suggests that the surface antigens display homology between some ST’s (ST 1,3,4) but that ST5 organisms differ in some respect to some surface antigens.

4.4.2 Cell granules and time-lapse microscopy Detailed views of Blastocystis organisms were obtained but apart from observing binary fission developing in vacuolated and granular forms, we did not observe changes from one morphological form to another. It was difficult to keep the Blastocystis organisms in the field of view despite using the cell tracking facility of the microscope. The Blastocystis cells were fragile and would often either rupture (Figure 4.4) or drift away out of the field of view over the 24 hours of recording, preventing observation of progressive cell changes past a certain point (Figure 4.5). One GF, observed over 6 hours in total, that appeared stationary on the slide displayed a large central granule moving rapidly within the GF (Figure 4.6, Supplementary video file 4.1).

88

Figure 4.4 Vacuolated form of Blastocystis sp., 15 minute time-lapse sequence

1 2 3 4

Bar 20M

89

Figure 4.5 Xenic culture of four Blastocystis granular forms, two showing development / rupture of cysts at cell surface

Bar 20m Arrow shows cystic structures; three time lapse images taken over three hours

90 Figure 4.6 Blastocystis with moving central granule sequential images over 15 minute time intervals

0’ 15’ 30’

45’ 60’ 75’

90’ 105’ 120’

135’ 150’ 165’

10uM ______

’ ’ 91

Another time-lapse study of a granular AF showed the form enlarging and losing granularity on the surface with subsequent expulsion of small tear- shaped granules. These granules appeared to arise from the surface of the organism or track along fissures to the surface (Figure 4.7). These granules varied in size from 1-3 µm and appeared to have a central dark spot surrounded by clear cytoplasm with a thick dark surface membrane. The release of the granules from the surface was best appreciated in the video images (Supplementary video file 4.2). Granules that appear to have tracked along fissures are identified as an area of interest in Figure 4.7.

Figure 4.7 Granules exiting Blastocystis amoebic form in xenic culture Timelapse images taken over 24 hours

Bar 15m Circles indicate areas of interest

92 The life cycle of Blastocystis remains elusive and the function of the multiple inclusion granules present in the GF is still unknown. These granules have been proposed to have functions related to metabolism, storage or reproduction (Zierdt 1991). DAPI has been shown to stain VF double stranded DNA strongly in the nucleus and weakly in the mitochondria-like organelles (Stenzel & Boreham 1996) but did not stain the central vacuole or the granules of GF’s (Matsumoto et al. 1987). These granules have been classified into three different types based on TEM studies (Tan & Zierdt 1973), namely myelin-like inclusions, crystalline granules and lipid droplets (Dunn et al. 1989). None of these previous microscopy studies have supported the hypothesis that these central granules constitute a reproductive phase. The cause of the movement of the large central granule seen within the GF in this study (Figure 4.6) is not understood and simple Brownian motion cannot be excluded. The small granules that were expelled from the granular AF are intriguing (Figure 4.7) but DAPI stains were not performed on the AF to confirm a reproductive role. AF are reported to occur commonly in faeces (Stenzel & Boreham 1996), particularly in patients with symptomatic diarrhoea (Tan & Suresh 2006b), but only rarely in culture. Asexual binary fission of VF and GF occur commonly in culture but other routes of reproduction have not been proven (Tan & Stenzel 2003). It is possible that the gut microbiota is necessary for completion of the normal life cycle of Blastocystis and that sexual replication may occur only in primed AF in the gut.

4.4.3 Other surface changes observed over time Surface pores, with an approximate diameter of 1µm, were seen on the external surface of the cell membrane of a partially ruptured ST3 Blastocystis organism (either a VF or a GF). The time-lapse recording of this event shows a total of 15 pores opening and closing (Supplementary video file 4.3). They appear to open and close fully over one hour during four 15 minutes interval observations. The regular movement of the pores continued for the length of the 24 hours study. A time interval sequence is shown with two pores in the foreground demonstrating fully open and closed pore positions in Figure 4.8.

93 Figure 4.8 Blastocystis form in xenic culture showing pores on surface

Open Pores Closed Pores

Bar: 20M Arrows mark two pores of interest

The presence of surface pores concurs with previous EM findings. Using TEM four morphological types of electron dense pits have been described on the surface membrane of the cell wall of Blastocystis, namely bars, cup-shaped pits, pits and tubular pits (Dunn et al. 1989). SEM studies have also confirmed the presence of small indentations on the external cell wall membrane, with fewer indentations noted on cells in prolonged culture (Cassidy et al. 1994). FE-EM has shown pores, approximately 0.5 µm in diameter, on the external surface of the outer cytoplasmic membrane that correspond to indentations on the inner surface of the outer cytoplasmic membrane (Tan et al. 1974), suggesting that pores may connect the outer and inner cytoplasmic membrane into the central vacuole. Unfortunately FE-EM damages the external layer of the outer cytoplasmic membrane so that FE-EM cannot confirm if these pores connect all the way through to the cell surface. Another time-lapse series of images showed extrusion of a viscous material apparently from one single point on the external cell membrane of two VFs of Blastocystis (Figure 4.9) (Supplementary video file 4.4).

94

Figure 4.9 Xenic culture of Blastocystis vacuolated form with extrusion; timelapse images . 1 2 3 4

5 6 7 8

9 10 11 12

Bar: 0.9M

1-2 images: time interval: 13 hours 2-12 images: 11.5 hour

95

The Blastocystis organisms were first seen to increase in size over 13 hours then deflate, but without disintegrating, as this substance extruded. This extrusion process took 11.5 hours in total and the extruded material had an amoebic shape. A separate DAPI stained slide (with multiple images taken at cross sections through the slide, but not time-lapse) stained nuclei in adjacent VFs bright blue but there was no discrete nuclear material staining seen in the adjacent amoebic shape. This amoebic-shaped structure appeared very similar in shape to the extruded material observed in the previous slide (Figure 4.10). The centre of this amoebic structure stained a light diffuse blue and the surface is covered in granules that do not fluoresce with the DAPI stain. This curious extrusion of viscous material from the VF occurring over many hours was observed to occur with deflation but conservation of the VF (Figure 4.9). Typically VF’s have only a fine rim of cytoplasm surrounding a large central vacuole. The large amount of this viscous material extruded from a single point occurring simultaneously with deflation of the VF is consistent with a direct connection from the surface to the central vacuole. This fluid extrusion may be occurring through the surface pores although it was interesting to note that the fluid only appeared to exit from one place on the surface wall and not from multiple pores. The regular, repetitive movement of pores opening and closing in these time-lapse images suggests that the movement is active and requiring energy, even in this disrupted cell. Surface pores have been reported to be important for nutrition and osmoregulation in other protozoa (Bokhari et al. 2008). The possibility that this viscoid extrusion represented the formation of a different morphological form, namely the development of an AF from a VF was considered. The similar amoeboid shape recorded on a different day showed no evidence of nuclear material within it (Figure 4.10) and provided no support for this hypothesis.

96 Figure 4.10

Fluorescent microscopy of Blastocytis forms stained with DAPI

Bar: 20m Solid arrows: Vacuolated forms of Blastocystis with nuclei at tip of arrow Dashed arrow: Amoebic shape DAPI: 4’,6-diamidino-2-phenylindole

Over the 24 hours of the study the material on the slide would have depleted nutrients and become progressively dehydrated, and it is possible this fluid extrusion is an osmoregulatory process occurring in a stressed VF. Blastocystis has been reported to survive in varying osmotic conditions (Ho et al. 1993) although only cysts survive immersion in fresh water. Osmosensing and osmoregulation have been reported in yeasts, microalgae and protozoa (Suescun-Bolivar & Thome 2015). Protozoans have

97 been reported to osmoregulate variously by changing the concentration of alanine, effluxing ions and water from either contractile vacuoles or intracellular compartments as well as inhibiting mitochondrial function during hyperosmolar conditions (Suescun-Bolivar & Thome 2015).

4.5 Summary Deconvolutional microscopy of xenic cultures of Blastocystis spp. produced images of living organisms in their natural microbial environment but proved challenging due to the fragility of the organisms. Fluorescent spectrometry revealed that Blastocystis exhibits green autofluoresence in common with many other microalgae. GAF was able to be distinguished from fluorescein-specific Blastocystis antibody labeling in VF and GF but not cysts. Care will need to be taken in future studies to distinguish GAF from other stains fluorescing in the green light spectrum. Surface pores were observed and appeared to be opening and closing actively. Extrusion of a viscoid material from VF may be occurring via these pores. Coincident with the extrusion the VF appeared to deflate suggesting a direct connection from the surface to the central vacuole. This extrusion may be an osmoregulatory mechanism used by Blastocystis sp. to maintain cell homeostasis and integrity. Further studies could use time-lapse images to examine changes in organisms immersed in hypo-osmolar and hyperosmolar medium to further explore this hypothesis. The nature of the tear-shaped granules observed exiting from an amoebic form is unclear. Amoebic forms are common in diarrheal faeces and rarely observed in culture. It is possible these are reproductive granules occurring only in amoebic forms that have been primed by the gut faecal microbiota. Fluorescent nuclear stains to prove transmission of genetic material could be combined in future time-lapse studies of amoebic forms in xenic cultures.

4.6 Acknowledgements I am grateful to Dr Rick Gould (Antibodies Inc, USA) for generously supplying the Blastofluor antibody, to Dr Lyn Knott (School of Veterinary Science, Gatton, University of Queensland) for invaluable help in setting up the

98 microscope and to Associate Professor Kevin Tan (Department of Microbiology, National University Singapore) for helpful guidance in interpretation of some videos. These experiments comply with the current laws of the country in which they were performed.

99 100

5 Clinical pilot study: efficacy of triple antibiotic therapy in Blastocystis positive irritable bowel syndrome patients

Nagel, R., Bielefeldt-Ohmann, H., Traub, R., 2014 Clinical pilot study: efficacy of triple antibiotic therapy in Blastocystis positive irritable bowel syndrome patients. Gut Pathogens 6:34

-presented as published

References

101 Clinical pilot study: Efficacy of triple antibiotic therapy in Blastocystis positive Irritable Bowel Syndrome patients

R.Nagel1, H.Bielefeldt-Ohmann1, 2, R Traub3

1School of Veterinary Science, The University of Queensland, Gatton, Queensland, Australia; 2Australian Infectious Diseases Research Centre, The University of Queensland, St. Lucia, Australia; 3Faculty of Veterinary Science, The University of Melbourne, Parkville, Victoria 3052, Australia

5.1 Abstract Background: Blastocystis species are common human enteric parasites. Carriage has been linked to Irritable Bowel Syndrome (IBS). Treatment of Blastocystis spp. with antimicrobials is problematic and insensitive diagnostic methods and re-infection complicate assessment of eradication. We investigated whether triple antibiotic therapy comprising diloxanide furoate, trimethoprim/sulfamethoxazole and secnidazole (TAB) given to diarrhoea- predominant IBS (D-IBS) patients positive for Blastocystis would achieve eradication. Methods: In a longitudinal, prospective case study 10 D-IBS Blastocystis- positive patients took 14 days of diloxanide furoate 500mg thrice daily, trimethoprim/sulfamethoxazole 160/80mg twice daily and secnidazole 400 mg thrice daily. Faecal specimens were collected at baseline, day 15 and 4 weeks after completion of TAB. Specimens were analysed using faecal smear, culture and polymerase chain reaction (PCR) of the 16 SSU rRNA. Patients kept a concurrent clinical diary. Results: Six (60%) patients cleared Blastocystis spp. after TAB, including three who had failed previous therapy. Subtypes detected were ST3 (60%), ST4 (40%), ST1 (20%) and ST7, 8 (10%); four patients had mixed ST infections. Serum immunoglobulin A (IgA) levels were low in 40% of patients. Higher rates of Blastocystis clearance were observed in patients symptomatic for less than a year (Mann-Whitney, p=0.032, 95% confidence) with no associations found with age, previous antibiotic therapy, faecal parasite load, ST, IgA level or clinical improvement.

102 Conclusions: Clearance of Blastocystis spp. was achieved with TAB in 60% of D-IBS patients, an improvement over conventional monotherapy. Higher clearance rates are needed to facilitate investigation of the relevance of this parasite in clinically heterogenous IBS.

Keywords Blastocystis, Therapy, Irritable Bowel Syndrome, PCR

5.2 Background Blastocystis spp.are unicellular, anaerobic enteric parasites found world wide in almost all species of animals (Tan 2008). The cystic form survives for days in the environment and likely transmits infection by the faecal-oral route either by direct contact or via contaminated water supplies. Blastocystis spp. are the most common parasite found in human faeces and carriage has been reported to be associated with Irritable Bowel Syndrome (IBS)–like symptoms (Tan 2008; Dogruman-Al et al. 2010; Jimenez-Gonzalez et al. 2012). IBS is a chronic heterogenous disorder present in up to 10% of the global population characterised by a symptom complex of abdominal pain and variable bowel habit in the absence of gross morphological, histological or inflammatory diagnostic markers (Drossman 2007). Although IBS does not increase mortality the disease impairs quality of life and costs billions of dollars annually in direct and indirect health care costs (Inadomi et al. 2003). Genetic, environmental, gender and psychological factors are known to be important in the development of the disease but the pathophysiology is unknown. As many as 10% of cases of IBS are recognised to develop after a bout of acute infectious gastroenteritis and this has focused research on microbial interactions with the luminal gut mucosa (Beatty et al. 2014). Faecal microbiota profiles are reported to be significantly different in IBS patients, most notably in the diarrhoea predominant IBS (D-IBS) subgroup (Collins 2014). IBS patients have diminished faecal bacterial diversity with temporal instability, fewer aerobic bacteria and decreased Bifidobacterium spp./ Enterobacter spp. ratio (Collins 2014; DuPont 2014). Altered intestinal motility and abdominal hypersensitivity are typically seen in

103 IBS, and many researchers have reported subtle alterations of the enteric immune and nervous system of IBS patients (de Salonen et al. 2010). Increased numbers of mast cells secreting serotonin that are closely aligned to mucosal dendritic cells are consistently observed in the gastrointestinal mucosa of patients with IBS (Matricon et al. 2012). Changes in cytokine secretion are reported not only with many luminal enteric microbial infections but increased levels of pro-inflammatory cytokines (interleukin (IL)-6, IL-8, and tumour necrosis factor- (TNF-)) have been reported in the plasma and peripheral blood mononuclear cells of IBS patients (Beatty et al. 2014). Many of these cytokines increase epithelial permeability by disrupting the mucosal tight junctions (Bruewer et al. 2003). Proteases are also reported to be elevated in the faeces of IBS patients (Gecse et al. 2008; Steck et al. 2014) although it is not clear if these proteases derive from enteric microorganisms, human intestinal cells or both. Blastocystis organisms have been shown to activate IL-8 gene expression in human colonic epithelial cells in vitro (Puthia et al. 2008), increase epithelial permeability (Puthia et al. 2006) and degrade intestinal Immunoglobulin A (IgA) (Puthia et al. 2005), mediated possibly by action of cysteine proteases (Mirza & Tan 2009) reported to be present on the cell surface (Wu et al. 2010). The prevalence of Blastocystis carriage is increased in all IBS patients but highest in D-IBS patients (73% compared to 27% in healthy control patients) (Yakoob et al. 2010), suggesting that this subgroup may provide a more homogenous IBS patient group to study. Seventeen different subtypes (ST) of Blastocystis spp. have been described to date (Stensvold 2013) revealing moderate host ST specificity and regional ST variation in humans. Early reports that particular subtypes of Blastocystis may be associated with clinical pathogenicity in humans have not been confirmed (Tan 2008). Treatment of Blastocystis spp. with antimicrobials is problematic and insensitive diagnostic methods and possible re-infection complicates assessment of successful eradication (Stensvold et al. 2010; Nagel et al. 2012). Monotherapy with metronidazole is the most commonly recommended drug and published eradication rates vary from 0-100% (Stensvold et al. 2010). If Blastocystis spp. could be predictably successfully eradicated in

104 symptomatic patients then it may be possible to more easily evaluate whether infection is associated with clinical symptoms. In this pilot study we treated ten patients with diarrhoea-predominant IBS (IBS-D) who were positive for Blastocystis carriage with triple antibiotic therapy (diloxanide furoate, trimethoprim-sulfamethoxazole, secnidazole) and assessed clinical status, ST and clearance rates.

5.3 Materials and Methods 5.3.1 Study Outline This prospective, longitudinal study was conducted in a rural specialist outpatient clinic (Toowoomba Gastroenterology Clinic). Patients presenting with chronic diarrhea were clinically assessed and patients who were subsequently diagnosed with diarrhoea-predominant IBS (IBS-D) (Drossman 2007) and were positive for Blastocystis carriage were invited to participate in the study. Faecal specimens were collected from study participants at baseline, after two weeks of triple antibiotic therapy and six weeks after completion of antibiotics.

5.3.2 Inclusion protocol Adult patients presenting to the clinic with chronic diarrhoea from 1/8/11 to 20/02/14 were assessed clinically. Blood tests, including full blood count, serum calcium, thyroid function tests, serum immunoglobulin A (IgA) and coeliac antibody tests were performed. Serum IgA testing was performed using a nephelometry assay using a BNII device (Siemens, Germany). The reference range for a healthy population provided by the manufacturer was 1.24-4.16 g/L. Selective IgA deficiency is defined as <50% of the lower range of normal (0.61 g/L). Faecal microscopy for ova, cysts and parasites (including Dientamoeba fragilis trophozoites), culture (for bacterial pathogens including Salmonella sp, Shigella sp, Vibrio cholerae, Campylobacter sp, Aeromonas sp) and PCR analysis for Giardia duodenalis, Clostridium difficile and Entamoebae histolytica (and a 3-day culture for hookworm) were performed by a commercial pathology laboratory to exclude known pathogens and initially detect the presence of Blastocystis spp. All patients proceeded to

105 an upper and lower endoscopy that included gastric antral, duodenal, ileal and colonic biopsies. Ten consecutive, eligible symptomatic patients positive for Blastocystis carriage who had no other cause for symptoms identified and who fulfilled the Rome criteria (Drossman 2007) for D-IBS, namely symptoms of chronic abdominal pain, occurring frequently over the last 3 months, associated with defecation and the passage of predominantly loose stool, were enrolled in the study. All patients who were invited to participate consented to enrolment and completed the study.

5.3.3 Exclusion protocol Only non-pregnant adults between 18 and 75 years of age were recruited for the study. Patients with a known allergy to sulphur drugs or patients with significant systemic diseases or co-morbidities were excluded.

5.3.4 Study protocol One baseline faecal specimen was collected from patients prior to 14 days therapy with diloxanide furoate 500mg three times daily, trimethoprim/sulfamethoxazole 160/800mg twice daily and secnidazole 400 mg three times daily. Two further faecal samples were taken, the first within 48 hours of ceasing antibiotics and another 4 weeks after ceasing antibiotics. Patients were reviewed clinically at the start of the study, after completion of antibiotics and 6 weeks later. All patients kept a diary which recorded the number of daily bowel movements (number per day), consistency of the stool (scored: 1=very hard, 2=hard, 3=formed, 4=loose, 5=watery) and general well being (scored 1-10; from poor to excellent). Routine electrolytes, liver function tests and a full blood count were repeated 4 weeks after ceasing the antibiotics.

5.3.5 Diagnostic methods All samples were run in parallel for the presence of Blastocystis spp. using a simple unstained wet faecal smear, xenic in vitro culture (XIVC) and PCR (confirmed as Blastocystis spp. using DNA sequencing). A patient was considered to be positive if any one of the tests was positive.

106 5.3.6 Parasitological methods Fresh unpreserved faecal specimens from patients were examined as a fresh faecal smear under 40x magnification by light microscopy. The load of Blastocystis organisms was classified as light, medium or heavy (<5, 5-10 and >10 organisms per high powered field (O/HPF). Additionally 100mg of fresh stool was inoculated in Jones’ culture medium (Jones 1946) and incubated at 37C for 48 hours. Cultures were examined for the presence of Blastocystis spp. by light microscopy.

5.3.7 Molecular methods DNA was extracted from the unpreserved faecal samples and pelleted cultures of Blastocystis using the QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturers instructions, except that following the addition of lysis buffer, the faecal suspension was subjected to three freeze-thaw cycles in liquid nitrogen and a 95C water bath respectively. The genomic DNA from stool and cultured Blastocystis specimens from all patients was subjected to PCR analyses. All samples were analysed using the Wong protocol (Wong et al. 2008) that amplified a 1000bp of the (SSU) rRNA gene of ST 1-17 of Blastocystis spp. All positive PCR products were subjected to DNA sequencing and phylogenetic analysis to identify the particular ST. All Blastocystis spp. positive samples were also analysed separately with three specific ST primers (ST1, 3, 4) in order to detect mixed infections. The ST1 specific protocol was developed by Dr CR Stensvold and the ST3 and ST4 protocols were developed in-house. These primers amplify highly ST- specific smaller fragments of the (SSU) rRNA gene and obviate the need for PCR product sequencing. The ST specific primers 1 and 4 were specific for ST when tested against Netski (ST1) and WR1 (ST4) axenic cultures. The STS primer 3 demonstrated mild cross reactivity with ST1 and ST4 and if disparity was found between the Wong primer results and the STS primer result the PCR product was sequenced. The ST specific primers sequences are as follows: ST1: Forward:TAATACATGAGAAAGTCCTCTGG Reverse: CTCAATCCAATTGCAAGAC Annealing temp: 57 C; amplicon size: 336 bp

107 ST3: Forward: GTAGTTGTGTAATGAATACATTCG Reverse: GTCCCTTTTAATCTTATTCTCATG Annealing temp: 62 C; amplicon size: 250 bp ST4: Forward: GTGATCTTCGGATGACGTGAATC Reverse: CAGAATTTCACCTCTGACTATTG Annealing temp: 60C amplicon size 290bp PCR products were run on gels of 1% agarose in 1 x sodium borate

(SB)(0.4gmNaOH/2.25g Boric acid/1L H2O) buffer at 200V in a Biorad electrophoresis system and were purified using a Purelink PCR Purification Kit (Invitrogen, Grand Island, NY, USA) prior to sequencing. Sequencing was done using an ABI 3130xl Genetic Analyzer (Applied Biosystems) using a Ready Reaction Kit Version 3.1 (PE Applied Biosystems, California, USA). Sequences were then edited and assembled using Finch TV v 1.4.0. DNA sequences were aligned with 15 reference sequences of Blastocystis SSU rRNA genes sourced from GeneBank (NCBI) representing the established subtypes described from mammals and birds. Distance-based analyses were conducted on readable sequences using Kimura 2-parameter distance estimates and trees were constructed using the Neighbour Joining algorithm using Mega 3.1(Kumar et al. 2001) software. Preoteromonas lacertae (U37108) was used as an out-group. Bootstrap analyses were conducted using 1000 replicates.

5.3.8 Ethics approval, statistics The study was approved by the University of Queensland Medical Research Ethics Committee and registered with the Australian and New Zealand Clinical Trials registry (http://www.ANZCTR.org.au) ACTRN: 12611000918921. The results were analysed using the Mann-Whitney U test (Wizard statistical program).

5.4 Results 5.4.1 Clinical characteristics of the study group Male (1) and female (9) patients with an age range of 28-73 years (median age 49 years) participated in the study. The shortest symptom duration was 3

108 months and the longest 10 years and the presenting symptoms were chronic diarrhoea and abdominal pain (10/10). Six patients had received prior antibiotic therapy (including metronidazole, norfloxacillin, nitazoxanide) for Blastocystis infection that did not clear either the parasite in the faeces or their symptoms. The remaining patients were naïve to therapy. During the period of the study no patient had any organisms other than Blastocystis spp. noted on faecal testing. No patient demonstrated eosinophilia on the peripheral blood film. There were no abnormalities noted in the baseline electrolyte, liver function tests or full blood films blood tests. Five patients recorded a low baseline serum IgA level (<1.24 g/L).

5.4.2 Antibiotic therapy and clinical follow up All patients completed 14 days of the triple antibiotic therapy (TAB). No serious adverse effects were noted; four patients complained of nausea, two patients of headache and one developed oral thrush and bloating during the antibiotic therapy course. At four weeks post antibiotic therapy repeat electrolytes, liver function tests and full blood count testing remained normal in all patients.

5.4.3 Faecal smears, cultures and PCR results Faecal smear results prior to the commencement of treatment demonstrated a parasite load of <5 O/HPF in five patients, 5-10 O/HPF in four patients and >10 O/HPF in one patient. Five patients had cleared the Blastocystis parasite from faeces on testing with faecal smear, XIVC and PCR at Day 15 and six patients had similarly cleared the organism at 6 weeks. All baseline samples positive to either microscopy or culture were positive on PCR testing using the Wong primers. Four patients had mixed ST infections that were confirmed with the ST-specific primers and subsequent sequencing if indicated. The Blastocystis ST’s detected were ST3 (60%), ST 4 (40%), ST 1 (20%) and ST7 and ST8 (10%). Day 15 and week six samples demonstrated good correlation between faecal smear/XIVC results and PCR testing (Table 5.1).

109 Table 5.1 Blastocystis faecal microscopy, XIVC and PCR results before and after triple antibiotic therapy

Patient Time Faecal microscopy PCR-WONG* PCR-ST1 specific** PCR-ST3 specific*** PCR-ST4*** Clearance Identity

Faeces XIVC Faeces XIVC Faeces XIVC Faeces XIVC Faeces XIVC 1 Baseline + + ST4 ST6 Negative Negative ST3* Negative ST4 Day 15 Negative Negative ST4 Negative Negative Negative ST3 Negative Negative Negative Yes Week 6 Negative Negative Negative Negative Yes

2 Baseline + + ST7 Negative Negative Negative Negative

Day 15 + + + Negative No

Week 6 + + ST7 Negative Negative Negative Negative No

3 Baseline + ST1 ST1 Negative Negative Day 15 + + + Negative Negative Negative No Week 6 + + Negative Negative No 4 Baseline + + Negative ST3 Negative Negative Negative ST3 Negative Negative Day 15 Negative Negative Negative Negative Yes Week 6 Negative Negative Negative Negative Yes 5 Baseline + + ST4 Negative Negative ST3* ST4 Day 15 Negative + Negative ST6 Negative ST3* Negative No Week 6 Negative + ST4 Negative Negative ST3* ST4 No 6 Baseline + + ST4 Negative Negative Negative ST4 Day 15 Negative Negative ST1 Negative Negative Negative Negative Yes Week 6 Negative Negative Negative Negative Yes 7 Baseline + Negative Negative ST4 Negative ST3* Negative Day 15 Negative Negative Negative Negative Yes Week 6 Negative Negative Negative Negative Yes 8 Baseline + + ST8 ST8 Negative Negative Negative Negative Negative Negative

Day 15 + ST8 Negative Negative Negative No

Week 6 + + ST8 ST8 Negative Negative Negative Negative Negative Negative No

9 Baseline + + Negative ST3 Negative ST3 Negative Day 15 Negative Negative Yes Week 6 Negative Negative Negative Negative Yes 10 Baseline + + Negative ST3 Negative ST3 Negative Day 15 + + ST3 Negative Negative ST3 Negative No Week 6 Negative Negative Negative Negative Yes

XIVC: xenic invitro culture Polymerase chain reaction PCR *Sequenced

110 5.4.4 Clearance and Clinical Follow up Overall 60% of patients cleared the organism from their faeces at 6 weeks. Three of these patients had had previous courses of antibiotics. The patient’s were reviewed clinically and their patient diary scored. The average score for the baseline week and week 4 following TAB therapy was calculated taking the weekly mean of the number of daily bowel actions, consistency of stool and general wellbeing. The final clinical score was calculated as the sum of improvement in frequency of bowel habit, consistency and wellbeing from baseline to week four after antibiotic therapy (Apppendix: Supplementary File 5.1). Six patients reported that they felt clinically improved and this correlated with a positive change of greater than 1.0 diary score from baseline to six weeks. However, of these six clinically improved patients only three demonstrated faecal clearance of the organism. In contrast, four patients did not report clinical improvement and two of these patients demonstrated faecal clearance of Blastocystis spp. This data was analysed to assess whether clinical improvement or Blastocystis clearance had any relationship to age of patient, duration of symptoms, naïve therapy status, original Blastocystis faecal load, the particular Blastocystis subtype, serum IgA level or clinical outcome. Significantly higher rates of Blastocystis spp. clearance were observed if the patient had been symptomatic for less than a year (Mann- Whitney, p=0.032, 95% confidence), but no other associations were found (Table 5.2).

111 Table 5.2 Clinical characteristics and outcome after triple antibiotic therapy

Patient Age Duration Previous Blastocystis Blastocystis Serum Diary Clinical Clearance yrs of AB load in ST’s IgA score Score at symptoms Therapy baseline g/L >=1 point 6 weeks (years) faeces 1 35 4 Yes (F,O) 2+ 3,4,6 1.51 +4.1 Yes Yes 2 59 5 No 2+ 7 1.12 (L) +2.4 Yes No 3 28 1.5 Yes (F) 1+ 1 2.1 -0.3 No No 4 35 0.25 Yes (F,O) 1+ 3 2.2 +0.2 No Yes 5 51 1.5 Yes (F) 1+ 3,4,6 2.82 +2.0 Yes No 6 58 0.5 No 2+ 1,4 1.03 (L) +1.1 Yes Yes 7 73 0.75 No 1+ 3,4 1.22 +4.5 Yes Yes 8 49 10 Yes (F,O) 3+ 8 1.94 +1.3 Yes No 9 63 1 No 1+ 3 0.75 (L) -0.8 No Yes 10 36 1 Yes (F) 2+ 3 0.98 (L) -0.6 No Yes

F=Flagyl O=other antibiotics (metronidazole, nitazoxanide, norfloxacillin) L=low, <1.24g/L, level defined by laboratory

112 5.5 Discussion Successful faecal eradication of Blastocystis spp.occurred in 60% of patients at the end of this study following fourteen days of triple antibiotic therapy using diloxanide furoate, trimethoprim/sulfamethoxazole and secnidazole (TAB). Our previous study (Nagel et al. 2012) used metronidazole as a monotherapy and achieved 0% successful faecal eradication, suggesting that this TAB regimen may be useful to treat Blastocystis infection. Although six patients in our study had already received at least one course of metronidazole therapy previously, three of these patients proceeded to clear the organism after TAB therapy indicating that triple therapy is still a worthwhile alternative in this group. One recent study has reported 100% failure to clear in five patients with Blastocystis ST3 infection treated with TAB combination (Roberts et al. 2014). However, these five patients had faecal specimens retested 12 months following the TAB therapy and it was not possible to exclude reinfection in this group. Triple antibiotic therapy has been shown to be useful in treating infecting organisms such as Helicobacter pylori and Mycobacterium spp. and it may be necessary to explore combinations of antibiotics and methods of delivery to further improve the clearance rate of Blastocystis infection. Widely varying rates of eradication (0-100%) of Blastocystis using a variety of antibiotics including ketoconazole, ornidazole, tinidazole, rifaximin and paromomycin have been reported (Stensvold et al. 2010) and some of this apparent inconsistency may be due to using insensitive faecal smears as a sole means of diagnosis. Diagnostic methods such as faecal XIVC or PCR have been reported to increase accuracy by 30-50% for detection of Blastocystis spp. (Matricon et al. 2012) . Metronidazole resistance has also been reported (Dunn et al. 2012) and differences in antimicrobial sensitivities to metronidazole and emetine demonstrated between axenic ST4 and ST7 cultures (Mirza et al. 2010). Resistance may therefore contribute to regional variation in eradication rates. A number of antimicrobials have been tested for activity against Blastocystis spp. in vitro, in both axenic and xenic conditions (Zierdt et al. 1983; Dunn & Boreham 1991; Vdovenko & Williams 2000) and this data is a useful starting point for choice of antimicrobial. The Blastocystis

113 organism appears to establish in the anaerobic environment of the human ileum and caecum, thrives in the presence of bacteria, and only in extremely rare circumstances is invasive to the gastrointestinal mucosa(Tan 2008). An intrinsic difficulty in treating Blastocystis spp. may be delivering the appropriate drugs to the intraluminal site of infection or understanding how to alter the faecal microbiota to produce an environment that is unsupportive or hostile to the parasite. Secnidazole, similar to metronidazole, is a 5-imidazole drug substituted at position 1 and has been shown to have significant (although 50% less than metronidazole) inhibitory activity against Blastocystis spp. in vitro tested on axenic cultures (Dunn & Boreham 1991). All drugs are rapidly absorbed from the small bowel, metabolised in the liver and excreted primarily in the urine. They exert their antimicrobial action in an anaerobic intracellular environment after reduction to a toxic metabolite that disrupts DNA linkage (Lamp et al. 1999). Metronidazole is the most common empiric therapy used against Blastocystis (Coyle et al. 2012) and as many patients that present for therapy have already failed metronidazole we felt it was useful to use a different imidazole particularly as resistance to metronidazole has been shown not to extend across the entire class of 1-sustituted 5-imidazoles (Mirza et al. 2010). The pharmacodynamics of secnidazole are not as well defined as those of metronidazole but it is known that secnidazole has reduced excretion of parent drug and metabolites in the urine (50% compared to 90% with metronidazole) and therefore may have increased faecal excretion. It also has more than double the excretion half-life of metronidazole at 29 hours (Mirza et al. 2010). These pharmacologic properties may be an advantage in treating Blastocystis infection. Trimethoprim/sulfamethoxazole duotherapy is often proposed as second line therapy for Blastocystis spp. infection and eradication rates of 22- 100% have been reported for this combination therapy (Stensvold et al. 2010). These two drugs act synergistically to interfere with the production of dihydrofolate reductase that is required for purine and including DNA synthesis. Blastocystis ST7 does not take up Hthymidine (Dunn & Boreham 1991; Wu et al. 2010), is likely to synthesize purines de novo, and consequently likely to have a dihydrofolate reductase that can be targeted.

114 Trimethoprim is rapidly absorbed from the small bowel (Menardi & J.P. 1984) is lipophilic with wide tissue distribution and 80% is excreted in the urine unchanged. Trimethoprim therapy alone has been shown to have inhibitory activity similar to secnidazole in vitro (Vdovenko & Williams 2000). Sulfamethoxazole similarly is rapidly absorbed in the small bowel (Menardi & J.P. 1984) metabolised in the liver, and over 90% excreted in the urine. They both have a half-life of around 10 hours. Although sulfamethoxazole alone was found to be non-inhibitory in vitro (Mirza et al. 2010) a combination of trimethoprim:sulfamethoxazole (in the ratio of 1:5, Bactrim) was found to be much less effective at inhibiting Blastocystis growth in axenic cultures than a trimethoprim/sulfamethoxazole 1:2 ratio which may indicate that the synergistic effect of the sulfamethoxazole plays an important role in directly inhibiting Blastocystis spp. growth. Enteric coating this duotherapy compounded in a 1:2 ratio may be a useful therapeutic avenue to explore. An alternative therapeutic approach would be to use an antimicrobial drug that has already been described to have intraluminal amoebocidal properties. A number of these drugs tested in vitro such as emetine, furazolidone, nitazoxanide, mefloquine, quinicrine, quinine, iodoquinol, (Entero-Vioform) and chloroquine showed moderate to high inhibition on Blastocystis cell growth (Zierdt et al. 1983; Dunn & Boreham 1991; Vdovenko & Williams 2000; Mirza et al. 2010). Emetine and clioquinol therapy have serious adverse reactions reported and are not used currently in clinical practice. Other intraluminal drugs such as diloxanide furoate, paromomycin, doxycycline and pyrimethamine showed no inhibitory effect on Blastocystis (Zierdt et al. 1983; Vdovenko & Williams 2000). Paromomycin is a relatively safe, poorly absorbed with activity against not only gram-negative but many gram-positive bacteria, including Mycobacterium tuberculosis, and protozoa such as Leishmania spp. (Davidson et al. 2009). Interestingly two recent studies have reported high Blastocystis clearance rates (77-100%) (van Hellemond et al. 2013; Roberts et al. 2014) using paromomycin and if confirmed this would suggest the broad spectrum aminoglycoside is achieving success by an indirect effect on the faecal microbiota. The largest study (van Hellemond et al. 2013) retrospectively

115 analysed faecal microscopy results on 52 symptomatic Blastocystis carriers treated with paromomycin alone and clearance rates were estimated to be 77%. Rifaximin is an oral broad-spectrum antibiotic derived from rifamycin that is not absorbed systemically (97% excreted in the faeces) (Adachi & DuPont 2006). It has been used successfully to treat, hepatic encephalopathy, travellers diarrhoea (Adachi & DuPont 2006) and non-constipated IBS patients (Pimental et al. 2011). Only one case report of its successful use in Blastocystis infection has been published (Amenta et al. 1999). Diloxanide furoate was chosen for our study as it had a good safety profile, the drug remains intraluminal until it is hydrolysed to diloxanide, absorbed and subsequently excreted in the urine. It is known to effectively clear intestinal although the mechanism of action is not known (Gadkariem et al. 2004). It was not possible to assess whether the addition of diloxanide furoate increased the efficacy of the regimen in this study. The addition of an intraluminal antimicrobial to a therapeutic regimen seems logical and paromomycin may be the most appropriate choice to add to a regimen at the moment. Unfortunately monotherapy is usually a cheaper option (the cost of 14 days of appropriate doses of metronidazole, trimethoprim/sulfamethoxazole, diloxanide furoate, paromomycin, secnidazole and rifaximin being AUS $24, $24, $55, $70, $80, $80 respectively) and the cost of TAB rising to AUS$159. If an effective drug combination could be identified it is likely the costs of a combination therapy would reduce. This small pilot study was not designed to interrogate specific variables that might be relevant to Blastocystis spp. symptoms, clearance or clinical outcome. The literature is inconsistent but some studies have suggested that particular subtypes, namely ST 1, 3 and 4 of Blastocystis may be pathogenic (Tan 2008). In contrast, and in accord with recent studies (Nagel et al. 2012; Roberts et al. 2014) our patients demonstrated a diverse range of subtypes and no association was seen between subtype and faecal clearance or resolution of symptoms. Although previous studies have reported mixed infection rates of 5-30% (Li et al. 2007a; Nagel et al. 2012) almost half of the patients (40%) in this study demonstrated mixed infections. The use of ST specific primers and XIVC techniques that allow ST’s present in small

116 numbers to grow to detectable numbers are factors likely to increase identification of mixed infections. It is well recognised that many Blastocystis carriers are healthy. Much research has been directed towards identifying genetic characteristics specific to the parasite, such as ST, that may indicate pathogenicity. Some studies have investigated host factors that may allow Blastocystis sp. to cause clinical symptoms. Epidemiological studies conducted in patients with compromised immune function (HIV carriage, chronic renal failure, haematological malignancies with immunosuppressant therapy (Tasova et al. 2000; Poirier et al. 2011) have shown increased rates of Blastocystis carriage but no clear consensus has emerged linking carriage to symptoms. IBS (Beatty et al. 2014) and Blastocystis infection (Puthia et al. 2008) have both been linked to changes in inflammatory cytokine expression. The prevalence of single nucleotide gene polymorphisms present in IL-8 (pro-inflammatory) and IL-10 (anti-inflammatory) cytokines is reported to be significantly increased not only in IBS patients but also in IBS patients positive for Blastocystis infection (Olivo-Diaz et al. 2012) suggesting that factors regulating the immune response in the host are important in expression of clinical disease. IgA is also an important part of host mucosal immunity. Blastocystis proteases have been reported to cleave human secretory IgA (Puthia et al. 2005). Interestingly 40% of our group of IBS-D Blastocystis had low serum IgA and this was much higher than the 1.4% (1 in 19) recorded in our healthy adult volunteers negative for Blastocystis spp. (unpublished observations). The levels of serum IgA recorded in our patients are compatible with partial selective IgA deficiency but did not reach <50% of the lower limit of normal defined as selective IgA deficiency. Selective IgA deficiency (prevalence of 0.17% in Australia) is usually asymptomatic, but may be associated with an increased risk of respiratory, urinary and gastrointestinal (including giardiasis) sepsis, and allergic and autoimmune diseases (Singh et al. 2014). There was no apparent association between IgA levels and subsequent clearance. Serum IgA levels recorded in children are generally lower than adults (Stoop et al. 1969) and if IgA levels influence the risk of carriage of Blastocystis then children may be more susceptible.

117 In this paper we describe successful clearance of the Blastocystis organism from faeces after testing at day 15 and week 4 after antibiotics. Blastocystis organisms may be excreted irregularly(Suresh et al. 2009) so it was useful to have had two clear faecal specimens 4 weeks apart to confirm faecal clearance. We used three methods of detecting Blastocystis in the stool, taking care to include XIVC and PCR amplification of Blastocystis DNA as the latter two methods have approximately a 90% sensitivity detection rate compared with only around 50% for microscopy alone (Stensvold et al. 2007a). Nevertheless false negative results may occur in a small percent of cases, where complete clearance of Blastocystis spp. in the ileum may not have occurred. It would be useful to be able to distinguish symptomatic and asymptomatic patients positive for Blastocystis faecal carriage. Although Blastocystis spp. are non-invasive organisms serum antibodies against Blastocystis spp. have been reported in humans infected with the parasite (Hussain et al. 1997). The presence of antibodies often only signifies exposure to that Blastocystis antigen some time in the past. However one study has reported that an antibody reacting to a 29kDa protein, thought to be a cysteine protease present on the plasma membrane of Blastocystis spp., is more commonly seen in symptomatic patients (Abou Gamra et al. 2011). If confirmed this finding and similar approaches may allow us to utilise the expression of host response to infection to determine need for therapy. The only variable that was found to be significant was duration of GIT symptoms and eventual clearance of the parasite. If we assume that the GIT symptoms were related to the Blastocystis carriage this finding may suggest that Blastocystis infection that persists in the host for years may be more difficult to eradicate, particularly as spontaneous clearance has been reported to occur in only 19% of symptomatic patients over 12 months (van Hellemond et al. 2013). The patients in this study were a careful selection of D-IBS patients with no other identifiable enteric pathogen or other cause for symptoms. Nevertheless only 50% of patients who reported clinical improvement had cleared the organism. Perhaps reducing parasite load was enough to achieve clinical improvement in these patients. The two patients who reported dramatic clinical improvement (>4 points) both cleared the organism. In contrast 50% of

118 the patients who did not report clinical improvement cleared the organism, suggesting that Blastocystis carriage was not the cause or not the only cause of their D-IBS. IBS is a heterogenous clinical condition and many diseases masquerade as IBS. Over time definitive specific diagnostic tests or therapies that allow us to discriminate other causes of IBS may become available. Although Giardia duodenalis was identified microscopically in the 1700’s acceptance of its pathogenicity only occurred in the last forty years when effective therapy with imidazoles (>90% successful eradication) demonstrated clinical improvement (Gardner & Hill 2001)[51]. Although 60% patients cleared in this study clinical cause and effect will be able to be more easily evaluated when we can more reliably and effectively eradicate Blastocystis spp.

5.6 Conclusions Successful faecal eradication of Blastocystis spp. occurred in 60% of D-IBS patients at the end of this study following fourteen days of triple antibiotic therapy using diloxanide furoate, trimethoprim/sulfamethoxazole and secnidazole (TAB). Metronidazole monotherapy achieved 0% successful eradication in our previous study and as some isolates of Blastocystis spp. have also been shown to be resistant to metronidazole, this drug should no longer be recommended as first line therapy. Surprisingly paromomycin, a drug that exhibits little inhibitory activity in vitro, has been reported to have eradication rates above 75%. If these results are confirmed in future studies using sensitive diagnostic markers, paromomycin may prove to be a useful basis for a double or triple antibiotic regimen. Predictable successful eradication (>90%) of Blastocystis spp. will expedite investigation of the clinical significance of Blastocystis infection. No meaningful analysis of clearance and symptoms was possible in this trial but 40% Blastocystis patients demonstrated low serum IgA levels. It may be worth focusing future studies on investigating immune profiles in the host and including host IgA levels.

119 5.7 Acknowledgements We thank Dr CR Stensvold, Department of Microbiological Diagnostics, Statens Serum Institut, Copenhagen, Denmark for providing ST1 specific primers, and Dr Linda Dunn, Queensland Insititute of Medical Research for donation of axenic cultures of Blastocystis subtypes Netski and WR1. This study was funded by the Royal Australian College of Physicians “Murray-Will Fellowship for Rural Physicians” (awarded to R.Nagel 2012)

120 121

6 Blastocystis specific serum immunoglobulin in patients with irritable bowel syndrome (IBS)

R Nagel, RJ Traub, MMS Kwan, H Bielefeldt-Ohmann, 2015 Blastocystis specific serum immunoglobulin in patients with irritable bowel syndrome. Parasites and Vectors 8: 453 pp1-13

-presented as published

122 Blastocystis specific serum immunoglobulin in patients with irritable bowel syndrome (IBS) versus healthy controls Robyn Nagel1, Rebecca J. Traub1,2, Marcella M.S. Kwan3, Helle Bielefeldt- Ohmann1,4

1 School of Veterinary Science, The University of Queensland, Gatton Campus, Queensland 4343, Australia. email: [email protected] 2 Faculty of Veterinary and Agricultural Sciences, University of Melbourne, Parkville, Victoria 3052, Australia. email: [email protected] 3 Rural Clinical School, School of Medicine, The University of Queensland, Toowoomba, 4350, Australia, email: [email protected] 4 Australian Infectious Diseases Research Centre, The University of Queensland, St. Lucia, Queensland 4072, Australia. email: [email protected]

6.1 Abstract Background: Blastocystis species are common enteric human parasites and carriage has been linked to Irritable Bowel Syndrome (IBS), particularly diarrhoea-predominant IBS. The spectrum of immune reactivity to Blastocystis proteins has been reported previously in symptomatic patients. We investigated differences in serum immunoglobulin profiles between patients with IBS, both positive and negative for Blastocystis carriage, and healthy controls (HC). Methods: Forty diarrhoea-predominant IBS patients (26 patients positive for Blastocystis sp., 14 negative patients) and forty HC (24 positive, 16 Blastocystis-negative) were enrolled. Age, gender, ethnicity and serum immunoglobulin A (IgA) levels were recorded and faecal specimens were analysed using smear, culture and polymerase chain reaction amplification of ribosomal DNA. Sera were tested in Western blots and the reactivities compared to known targets using monoclonal antibodies Blastofluor (Blastocystis specific antibody), MAb1D5 (cytopathic to Blastocystis cells), anti-promatrix metalloprotease-9 (anti-MMP-9) and SDS-PAGE zymograms.

123 Results: Levels of serum IgA were significantly lower in Blastocystis carriers (p<0.001) but had no relationship to symptoms. Western blots demonstrated serum IgG antibodies specific for Blastocystis proteins of 17, 27, 37, 50, 60-65, 75-90, 95-105 and 150 kDa MW. Reactivity to the 27, 50 and 75-95 kDa proteins were found more frequently in the IBS group compared to the HC’s (p<0.001) and correlation was greater for Blastocystis-positive IBS patients (p<0.001) than for negative IBS patients (p<0.05). MAb1D5 reacted with proteins of 27 and 100 kDa, and anti-MMP-9 with 27, 50 and 75-100 kDa proteins. Bands were seen in zymograms around 100 kDa. Conclusions: Low serum IgA levels are associated with Blastocystis carriage. All IBS patients were more likely to demonstrate reactivity with Blastocystis proteins of 27 kDa (likely a cysteine protease), 50 and 75-95 kDa MW compared to HC. The presence of antibodies to these Blastocystis proteins in some Blastocystis-negative subjects suggests either prior exposure to Blastocystis organisms or antibody cross reactivities. The anti-proMMP-9 reaction at 50 and 75-100 kDa and the zymogram result suggest that metalloproteases may be important Blastocystis antigens. Word Count: 322

Trial registration Australian and New Zealand Clinical Trials registry ACTRN: 12611000918921 Key Words Blastocystis; Irritable Bowel Syndrome; Immunoglobulin A; Western blotting, proteases

6.2 Introduction Irritable bowel syndrome (IBS) is a common clinical condition affecting up to 10% of the global population and characterised by abdominal pain, bloating and disturbance of bowel habit (Canavan et al. 2014a).The underlying cause of IBS is not known although differences in gut motility, the enteric immune and nervous system, intestinal permeability, systemic

124 cytokine production, faecal protease excretion, faecal microflora and psychological profiles have been reported in this condition (Whorwell 2014). The disease is defined by a clinical symptom complex description (Rome III) (Longstreth et al. 2006) that has a sensitivity of 70% and a specificity of 80% in differentiating IBS patients from patients with other gastrointestinal diseases (Corsetti et al. 2014). This clinical definition is the gold standard, but assessment can be subjective. Reliable, clinically applicable IBS biomarkers, particularly if they enabled therapeutic choices, would be useful. Serum biomarkers for IBS have been investigated in two previous studies (Lembo et al. 2009; Jones et al. 2014) and both were less accurate than using the Rome III criteria. Post-infectious IBS is known to comprise 10% of the total cases of IBS and this has stimulated interest in the faecal microbiota as a possible contributing cause of IBS (Beatty et al. 2014). The carriage of the enteric organism Blastocystis has been reported to be three times higher in patients with diarrhoea predominant IBS (D-IBS) compared to healthy controls (Yakoob et al. 2010), making it an organism of interest in IBS. Blastocystis sp. are the most common parasites found in human stool (Tan 2008). Sub-typing of the18S ribosomal DNA (rDNA) has identified 17 different subtypes (ST’s) and nine have been identified in humans (Stensvold 2013). Carriage of Blastocystis sp. is increased in patients with various types of immunosuppression (Tan 2008) and in patients with irritable bowel syndrome (Yakoob et al. 2010). Nevertheless, many carriers are healthy and a definite association between carriage and illness has not been proven in epidemiological studies (Poirier et al. 2012). Blastocystis sp. reside in the intestinal lumen, establishing in the ileum and caecum adherent to the outer layer of mucus (Phillips & Zierdt 1976; Wang et al. 2014a), with only rare reports of mucosal invasion. Antibodies specific for Blastocystis antigens have been demonstrated in the faeces and the serum of carriers (Zierdt & Nagy 1993; Kaneda et al. 2000; Hegazy et al. 2008; Gamra et al. 2011; Wang et al. 2014b) and antibody titres have been reported to be higher with length and severity of infection (Zierdt & Nagy 1993). These antibodies have been described in all immunoglobulin classes. IgA antibodies specific for Blastocystis sp. have been shown to be present in pig faeces (Wang et al. 2014b) and in both faeces and serum of infected

125 humans (Mahmoud & Saleh 2003). Our previous pilot study demonstrated lower serum IgA levels in IBS patients positive for Blastocystis carriage than in other patient groups or healthy controls (Nagel et al. 2014). Blastocystis specific immunoglobulin G (IgG) antibodies have also been detected in serum (Zierdt & Nagy 1993; Garavelli et al. 1995; Mahmoud & Saleh 2003) and faeces of Blastocystis infected humans (Mahmoud & Saleh 2003), with lower titres in the latter compartment. Using SDS-PAGE and Western blot analysis, serum IgG has been reported to react with Blastocystis-specific proteins of 12 kDa (Kaneda et al. 2000), 29 kDa, 50 kDa and 118 kDa molecular weights (MWs) (Hegazy et al. 2008). Antibody reactivity to a 29 kDa MW Blastocystis protein has been reported to be more common in the serum of symptomatic patients compared to asymptomatic individuals (Gamra et al. 2011), and subsequent protein sequencing of a 30 kDa protein showed it to possess 50% homology with known cysteine protease (legumain type) peptide sequences (Wu et al. 2010). A monoclonal IgM antibody 1D5 (MAb1D5), known to be cytopathic to Blastocystis organisms (Tan et al. 2001) as well as a human legumain antibody both bind to this 30 kDa Blastocystis antigen in Western blots. Protease secretion is a recognised virulence mechanism for parasites (McKerrow 1989) facilitating tissue/cell invasion, protein activation and immunoevasion. These molecules are also highly immunogenic. Analysis of the genome of Blastocystis ST7 has led to the predictions that the parasite would be able to produce all major classes of proteases, including serine proteases, metalloproteases and perhaps as many as 20 different cysteine proteases (Denoeud et al. 2011; Wawrzyniak et al. 2012). Notably Blastocystis proteases have been shown to cleave human secretory IgA (Puthia et al. 2005) and induce production of the pro-inflammatory cytokine interleukin-8 (IL-8) by enterocytes in an in vitro model system (Puthia et al. 2006). Our aim was to explore the clinical relevance of Blastocystis sp. in patients with diarrhoea predominant IBS by assessing serum antibody reactivities specific for Blastocystis. In this study we compared the serum IgA levels in IBS patients, either positive (IBS-P) or negative (IBS-N) for Blastocystis carriage, and healthy controls (positive (HC-P) and negative (HC-

126 N) for Blastocystis). IgG serological responses to specific Blastocystis antigens were examined in all these subgroups using Western blotting techniques. Identification of specific Blastocystis antigens was attempted by probing the Western blots with Blastocystis specific antibody (Blastofluor, Antibodies Inc), MAb1D5 and pro-matrix metalloprotease 9 (anti-MMP-9) antibody.

6.3 Methods 6.3.1 Study outline Forty patients presenting with diarrhoea-predominant IBS to the Toowoomba Gastroenterology Clinic and forty healthy volunteers were enrolled in the study and signed written consent forms. The study was approved by the University of Queensland Medical Research Ethics Committee and was part of a clinical trial that is registered with the Australian and New Zealand Clinical Trials registry (http://www.ANZCTR.org.au) ACTRN: 12611000918921. Single baseline faecal and serum samples were collected from all participants. Total serum IgA levels were measured in all participants and faecal and serum specimens underwent further microbiological, molecular and Western blot analysis. Inclusion protocol Adult patients presenting with chronic diarrhoea and abdominal pain from 1/8/11 to 20/02/14 were assessed clinically. Baseline blood tests, including full blood count, serum calcium, thyroid function tests, serum IgA and coeliac antibody tests were performed. Other pathogens known to cause chronic diarrhoea such as Salmonella sp, Shigella sp, Vibrio sp., Campylobacter sp, Aeromonas sp),Giardia duodenalis, Clostridium difficule and Dientamoeba histolytica were excluded with faecal microscopy , culture and PCR testing performed by a commercial pathology laboratory. All patients proceeded to an upper and lower endoscopy that included gastric antral, duodenal, ileal and colonic biopsies. Forty eligible symptomatic patients who had no other cause for symptoms and who fulfilled the Rome criteria (Drossman 2007) for diarrhoea predominant IBS were enrolled in the study. Of these, 26 patients were positive for Blastocystis (IBS-P) and 14 were

127 negative (IBS-N). Forty healthy volunteers working at the University of Queensland and with no gastrointestinal symptoms in the preceding 12 months were tested for Blastocystis sp. infection, 24 were positive (HC-P) and 16 were negative (HC-N). Details of age, gender and ethnicity of participants were recorded. Exclusion protocol Only non-pregnant adults between 18 and 75 years of age were recruited for the study. Patients with significant systemic diseases or co- morbidities were excluded.

6.3.2 Diagnosis of Blastocytis infection All samples were run in parallel for the presence of Blastocystis sp. using a simple unstained wet faecal smear, xenic in vitro culture (XIVC) and PCR (confirmed as Blastocystis sp. using DNA sequencing). A study participant was considered to be positive if any one of the tests was positive as described previously (Nagel et al. 2014).

6.3.3 Serum Immunoglobulin A Serum samples were stored at -20C prior to analysis at Sullivan and Nicolaides Pathology Service, Brisbane, Queensland using a Siemens BNII Nephelometer (Siemens, Munich, Germany) and Siemens reagents.

6.3.4 Western Blotting Antigen preparation An axenic strain of Blastocystis sp. ST4 (WR1) was cultured anaerobically in pre-reduced Iscove’s Modified Dulbecco’s medium (IMDM) (Sigma-Aldrich, St Louis, USA: 13390) enriched with 10% heat-inactivated horse serum (Gibco: 26050088). WR1 was originally obtained from the stool of Wistar rats in Singapore but has been in continuous axenic culture for more than seventeen years. Blastocystis antigen was prepared using the proteinase inhibitor Complete Lysis-B (2x), EDTA-free kit (Roche, Mannheim, Germany: 04719948001) and stored at -20C until required.

128 SDS-PAGE protocol Approximately 300 ng of Blastocystis antigen was mixed with equal volumes of loading buffer (Laemmli sample buffer mixed with 2M 2- mercaptoethanol in a ratio of 19:1 vol:vol (Bio- Rad, California, USA)) and loaded into the large well of a preparative 4-15% Tris-HCl Ready Gel (Bio- Rad: 161-1140). A separate single protein standard ladder well in the gel was loaded with 15 uL of Precision Plus Protein Western C Standard (Bio-Rad: 161-0385). The gel was electrophoresed in M Tris/Glycine/SDS buffer (Bio- Rad: 161-0732) at 100V for 40-60 minutes at room temperature and then rinsed three times with deionised water. The proteins were transferred to a nitrocellulose membrane using 1M Tris/Glycine/20% methanol buffer electrophoresed at 100V for 60 minutes at 4C. The membrane was rinsed three times with 0.1% 1M phosphate buffered saline (PBS)/Tween 20 (PBST) (Sigma Aldrich, St Louis, USA: P5927) solution and then blocked in 100 ml of blocking solution (0.1% PBST, 4gm skim milk and 1gm bovine serum albumin per 100 ml (Sigma Aldrich: A2153) for 60 minutes at room temperature. The membrane was cut into strips and stored at -20C till required. Immunolabeling protocol Membrane strips were incubated with primary antibody (patient sera or MAbs) diluted in 1 mL of Signal BoostTM Immunoreaction Enhancer Kit (Calbiochem, Darstadt, Germany: 407207) Solution One for 24 hours on a rocker platform at 4C before washing consecutively three times for 15 minutes in PBST. The strips were then incubated with secondary antibody diluted in Signal Boost Solution Two at room temperature for 1 hour on a rocker platform and the washing steps repeated. The patient sera were diluted 1:10 and the secondary antibody was a 1: 2000 dilution of horseradish peroxidase (HRP)-conjugated goat anti-human IgG Ab (Sigma Aldrich: A8786). An additional four antibodies were tested for reactivity with the Blastocystis antigens: (i) a mouse derived Blastocystis-cytopathic monoclonal IgM antibody, MAb1D5 (purified Mab against surface legumain of Blastocystis ST7, IgM, 16 g/L) (Tan et al. 1996; Tan et al. 1997; Wawrzyniak et al. 2012), (ii) a control unrelated mouse derived monoclonal IgM antibody, MAb5,

129 (iii) a rabbit polyclonal anti-Blastocystis ST3 antibody (Blastofluor)(Gould & Boorom 2013), and (iv) a rabbit polyclonal anti-human matrix metalloproteinase 9 (Anti-MMP-9) (Calbiochem: 444236). These primary antibodies were diluted 1:100, and the secondary antibodies 1:1000, using HRP-conjugated-goat anti-mouse IgM (Sigma, St Louis, USA: Cat no: A8786) and HRP-conjugated anti-rabbit polyclonal Ab, respectively. All blots were run with a protein standard ladder and a negative control. Negative controls were prepared according to the protocol above except exclusion of serum/primary antibody in step one followed by incubation with the secondary antibody. The antibody binding was visualized using Clarity Western ECL Substrate (Bio Rad, California USA) and a Bio Rad Chemi-DocTM XRS Gel Documentation system. The brightness and contrast was adjusted over entire images to optimise the image but no differential image manipulation was used on separate lanes.

6.3.5 Zymograms The Blastocystis antigen was prepared as per the Western blotting protocol and combined with Zymogen buffer (Bio-Rad: 161-0764) in a vol:vol ratio of 1:1. Approximately 50 ng of buffered Blastocystis antigen was loaded into each well of a 10 % gelatin Ready Gel Zymogram Gel (Bio-Rad: 161- 1167) and the gel was electrophoresed in 1M Tris/Glycine/SDS buffer (Bio- Rad: 161-0732) at 90V for 40-60 minutes at room temperature. The gel was cut in half and both gels subjected to renaturation and development steps. The gels were incubated in 100mL of Zymogen Renaturation buffer (Bio-Rad: 161-0765) for 1 hour at room temperature on a rocker platform and then incubated in 100mL of Zymogen Development buffer (Bio-Rad: 161-0766) for 24 hours at 37C without or with protease inhibitor, 1mM EDTA (ethylenediaminetetraacetic acid), added to the two solutions for one gel. The buffer was decanted and the gels were stained with 100 mL Aqua stain (Bulldog Bio, Portsmouth, UK) for 15 minutes on a rocker platform and rinsed with deionised water.

130 6.3.6 Statistical Analysis Statistical analysis was carried out using SPSS v.22 (IBM SPPS 2013). The level of IgA was transformed using natural logarithm to get a normal distribution. A one-way Analysis of Covariance (ANCOVA) was used to compare the level of IgA between the four groups, adjusting for potential confounders, such as seasonality (Weber-Mzell et al. 2004). Pearson’s chi- square/ Fisher exact test were used to assess the significance of association between subjects with IBS and healthy controls with the presence of serum IgG antibodies detected against Blastocystis proteins of various sizes.

6.4 Results 6.4.1 Study subjects The eighty study participants comprised forty symptomatic IBS patients (26 positive and 14 negative for Blastocystis carriage) and forty asymptomatic healthy controls (24 positive and 16 negative for Blastocystis). The average participants’ age was 45.2 ± 14.4 years (range 16 - 75 years). There were no significant between-group differences in terms of age (F (3,76)=0.81, p=0.49) and gender distribution (λ2=4.37, p=0.22). However, there were significantly more Caucasians compared to non-Caucasians in the IBS-P, IBS-N and HC- P groups (λ2= 16.83, p=0.001). The Blastocystis sp. subtypes detected were ST3 (34%), ST4 (26%), ST1 (16%), ST2 and ST7 (8%), ST8 (6%) and ST5 (2%). Characteristics of the 80 subjects and their Blastocystis subtypes are given in Table 6.1 and 6.2.

131 Table 6.1 Demographic and epidemiological characteristic of subgroups IBS-P IBS-N (n=14) HC-P HC-N (n= 26) (n=24) (n=16) Age (years; 44.5±15.2 44.5±14.2 48.7 ±13.2 41.7±15.0 mean±SD) Female (n, %) 19 11 12 11

Caucasian (n, %) 26 (100) 13 (93) 21 (88) 9 (56)

Season recruited (n) Spring 2 3 19 0 Summer 8 5 4 1 Autumn 11 4 0 7 Winter 5 2 1 8

Blastocystis ST distribution ST1 2 6 ST2 1 3 ST3 9 8 ST4 9 4 ST5 0 1 ST6 0 0 ST7 3 1 ST8 2 1

Serum IgA* (g/L; 1.63±0.95 2.19±1.01 1.54±0.76 3.06±1.61 mean±SD)

SD: standard deviation ST: subtype IgA: Immunoglobulin A IBS-P/N: Irritable bowel syndrome subjects positive/negative for Blastocystis HC-P/N: Healthy control subjects positive/negative for Blastocystis *unadjusted mean; results IgA of different subtypes in IBS-P & HC-P groups were pooled for analysis purposes

Table 6.2 Serum antibody reactions to Blastocystis antigens detected by Western immunoblotting in all subjects infected with different subtypes of Blastocystis Protein band Number of subjects infected with particular subtype of at molecular Blastocystis weight ST1 ST2 ST3 ST4 ST5 ST7 ST8 (n=8) (n=4) (n=17) (n=13) (n=1) (n=4) (n=3) 17kDa 2 3 1 27 kDa 7 3 16 10 1 2 3 37 kDa 1 2 2 1 50 kDa 7 2 14 11 1 4 3 60-65 kDa 7 4 16 11 1 3 3 75-90 kDa 8 4 12 9 1 3 2 95-105 kDa 5 1 10 7 1 2 2 150 kDa 5 3 8 2 1 2

ST: subtype kDa: kiloDalton Absence of bands

132 6.4.2 Serum IgA levels Serum IgA levels were higher in male (2.36±1.36 g/L) compared to female participants (1.79±1.09 g/L) (t=4.15, p<0.05). Levels of IgA were not found to be different between Caucasian and Asian participants, nor across the different Blastocystis subtypes (both p>0.05). The average level of serum IgA was lowest in the HC-P group, followed by IBS-P, IBS-N and HC-N (Table 6.1). Participants who tested Blastocystis- positive (combined IBS-P and HC-P subgroups) had a significantly lower serum IgA level (1.59±0.54 g/L) than their Blastocystis- negative counterpart (combined IBS-N and HC-N) subgroups (2.65±1.41 g/L) (t(78)=4.06, p<0.001). Planned contrasts revealed that participants in the HC-N group had significantly higher serum IgA levels compared to those in the IBS-P group, t(75)= -3.31, p=0.001, and HC-P group, t(75)= -4.30, p<0.001. Pairwise comparisons showed that IgA levels in IBS-N participants were higher than those in the HC-P group (p<0.01). There was a significant difference in IgA levels between the four groups after controlling for the potential confounding effect of seasonality and gender, F(3,73)=9.05, p<0.001. The gender covariate was significantly related to the level of IgA, (F(1,73)=7.93, p<0.01), whereas seasonality, was not (F(1,75) = 3.42, p= 0.07).

6.4.3 Blastocystis-specific serum IgG antibody bands Western blots were employed to assess the serum IgG response to Blastocystis proteins in the four clinical groups and the analysis of these results is shown in Tables 6. 2, 6.3 and 6.4. Reactivity was seen to Blastocystis proteins of approximately 27, 50, 60-65, 75-90, 95-105 and 150 kDa MW (Yakoob et al.) in all groups of patients (Figures 6.1-6.3). These results were examined to assess whether serum reactivity to Blastocystis proteins of different sizes was associated with carriage of a particular subtype of Blastocystis sp., or either the presence of Blastocystis sp. or IBS symptoms. If Blastocystis spp. were to prove to be a cause of IBS then it is possible that IBS-N patients may have had Blastocystis infection in the relatively recent past and the subgroups with IBS were analysed separately in order to reduce any contribution from pre-existing antibodies (Table 6.4).

133 Table 6.3 Serum antibody reactions to Blastocystis antigens detected by Western immunoblotting in different groups

Protein band Number of subjects with presence of protein bands at molecular (n, %) weight IBS-P (n=26) IBS-N (n=14) HC-P (n=24) HC-N (n=16) 17kDa 1 (4) 0 (0) 5 (21) 0 (0)

27kDa 25 (96) 12 (86) 17 (71) 7 (44)

37kDa 3 (12) 0 (0) 3 (13) 0 (0)

50kDa 26 (100) 13 (93) 16 (67) 8 (50)

60-65kDa 26 (100) 12 (86) 19 (80) 13 (81) 75-90kDa 25 (96) 13 (93) 14 (58) 9 (56) 95-105kDa 14 (54) 3 (21) 14 (58) 2 (13) 150kDa 5 (19) 1 (7) 16 (67) 1 (6) IBS-P/N: Irritable bowel syndrome subjects positive/negative for Blastocystis HC-P/N: Healthy control subjects positive/negative for Blastocystis n: number kDa: kiloDa

134 Table 6.4 Comparison of presence of Blastocystis and/or gastrointestinal symptoms to specific sized antibody bands directed against Blastocystis proteins using Western immunoblotting

Presence of antibody to Blastocystis protein at band MW (n, %) Gp1 Gp2 17kDa 27kDa 37kDa 50kDa 60-65kDa 75-90kDa 95-105kDa 150kDa Gp1 Gp2 Gp1 Gp2 Gp1 Gp2 Gp1 Gp2 Gp1 Gp2 Gp1 Gp2 Gp1 Gp2 Gp1 Gp2 Blastocystis presence IBS-P IBS-N 6 (12) 0 (0)* 42 19 6 (12) 0 (0)* 42 21 45 25 39 22 28 5 21 2 & & (84) (63)* (84) (70) (90) (83) (78) (73) (56) (17)*** (42) (7)*** HC-P HC-N (n=50) (n=30) IBS-P HC-N 6 (12) 0 (0) 42 7 6 (12) 0 (0) 42 8 45 13 39 9 (56) 28 2 21 1 (6)** & (n=16) (84) (44)*** (84) (50)** (90) (81) (78) (56) (13)** (42) HC-P (n=50) IBS-P IBS-N 1 (4) 0 (0) 25 12 3 (12) 0 (0) 26 13 26 12 25 13 14 3 5 (19) 1 (7) (n=26) (n=14) (96) (86) (100) (93) (100) (86)* (96) (93) (54) (21)*

HC-P HC-N 5 (21) 0 (0)^ 17 7 (44) 3 (13) 0 (0) 16 8 (50) 19 13 14 9 (56) 14 2 16 1 (n=24) (n=16) (71) (67) (79) (81) (58) (58) (13)** (67) (6)***

GIT symptoms IBS-P HC-P 1 5 (13) 37 24 3 (8) 3 (8) 39 24 38 32 38 23 17 16 6 (15) 17 & & (Bailar (93) (60)*** (98) (60)*** (95) (80)* (95) (58)*** (43) (40) (43)** IBS-N HC-N & (n=40) (n=40) Mostel ler) IBS-P HC-P 1 (4) 5 (13) 25 24 3 (12) 3 (8) 26 24 26 32 25 23 14 16 5 (19) 17 (n=26) & (96) (60)*** (100) (60)*** (100) (80)* (96) (58)*** (54) (40) (43)* HC-N

135 (n=40) IBS-P HC-P 1 (4) 5 (21) 25 17 3 (12) 3 (13) 26 16 26 19 25 14 14 14 5 (19) 16 (n=26) (n=24) (96) (71)* (100) (67)*** (100) (80)* (96) (58)*** (54) (58) (67)***

IBS-P HC-N 1 (4) 0 (0) 25 7 3 (12) 0 (0) 26 8 26 13 25 9 14 2 5 (19) 1 (6) (n=26) (n=16) (96) (44)*** (100) (50)*** (100) (81)* (96) (56)*** (54) (13)**

IBS-N HC-P 0 (0) 5 (13) 12 24 0 (0) 3 (8) 13 24 12 32 13 23 3 (21) 16 1 (7) 17 (n=14) & (86) (60) (93) (60)* (86) (80) (93) (58)* (40) (43)* HC-N (n=40) IBS-N HC-N 0 (0) 0 (0) 12 7 0 (0) 0 (0) 13 8 12 13 13 9 3 (21) 2 (13) 1 (7) 1 (6) (n=14) (n=16) (86) (44)* (93) (50)* (86) (81) (93) (56)*

Gp: group of study subjects n: number of subjects kDa: kiloDalton MW: molecular weight IBS-P/N: Irritable bowel syndrome subjects positive/negative for Blastocystis HC-P/N: healthy control subjects positive/negative for Blastocystis *p≤0.05 **p≤0.01 ***p≤0.001

136 Figure 6.1 Serum antibodies from Blastocystis positive clinical subgroups IBS-P and HC-P reacting with Blastocystis

proteins in Western blot

L |______IBS-P______||______HC-P______| N kDa 250

150

100 75

50

37

25 20

15

IBS-P= Irritable bowel Syndrome patients positive for Blastocystis HC-P=Healthy controls positive for Blastocystis N=negative control L=protein standard ladder kDa=kiloDalton molecular weight

137 Figure 6.2 Serum antibodies from Blastocystis positive clinical subgroups IBS-P and HC-P and Blastofluor Ab reacting with Blastocystis proteins in Western blot

|______IBS-P______| |______HC-P______| N Bf

250 kDa

150

100

75 50

37

25

IBS-P=Irritable Bowel Syndrome patient positive for Blastocystis HC-P=Healthy controls positive for Blastocystis N=negative control Bf=Blastofluor antibody kDa=kiloDalton molecular weight

138 Figure 6.3 Serum antibodies from Blastocystis negative clinical subgroups IBS-N and HC-N reacting with Blastocystis proteins in Western blot

L HC-N kDa N L |______|IBS-N |______|

50

100

75

50

25

IBS-N=Irritable Bowel Syndrome patient negative for Blastocystis HCN=Healthy controls negative for Blastocystis N=negative control L=standard ladder kDa=kiloDalton molecular weight

139

The numbers of Blastocystis-positive subjects in each subtype group were small but serum reactivity to all the protein bands was present in every subtype with the exception of 17 kDa, 37 kDa and 150 kDa bands (Table 6.2). No significant associations were found between Blastocystis sp. subtypes and reactivity to particular protein bands based on Fisher’s exact test results (all p>0.05). Reactivity with proteins of approximately 17 and 37 kDa MW were observed in 7.5% of total participants and these were only present in those subjects who tested positive for the presence of Blastocystis sp. Proportions of participants with reactivity for proteins of 95-105 kDa and 150 kDa MW, were significantly lower in the combined Blastocystis-negative group (λ2= 11.97, p≤0.001 and λ2= 11.43, p≤0.001 respectively) (Table 6.3). The association of antibody-reactivity to the 95-105 kDa protein(s) with Blastocystis carriage was statistically significant in the intra-subgroup comparisons in both HC and IBS groups (p≤0.05) (Table 6.4). Reactivity to the 27, 50 and 75-90 kDa MW bands were detected in descending order of frequency in the IBS-P group (96%, 100%, 96%), the IBS-N group (86%, 93%, 93%), the HC-P (71%, 67%, 58%) and the HC-N (44%, 50%, 56%) groups (Table 6.3). Reactivity to each of these three proteins was found to be significantly associated with IBS symptoms, regardless of Blastocystis status (combined IBS-P + IBS-N (p≤0.001), IBS-P (p≤0.001) or IBS-N (p<0.05), respectively, versus HC (Table 6.4).

6.4.4 Further Analysis of the Blastocystis Antigens The Blastofluor antibody was tested in parallel with IBS-P and HC-P sera on a Western blot and showed reactivity with all the protein bands noted in the study subgroups, namely 17, 27, 37, 50, 60-65, 75-90, 95-105, 150 and 250 kDa MW (Table 6.2). The MAb1D5 and MAb5 monoclonal antibodies were tested in parallel with four sera derived from BSP patients on a Western blot. The MAb1D5 reacted strongly with two protein bands of approximately 100 kDa MW and

140 less strongly with a 27 kDa protein. The MAb5 gave no reactivity on the blot (Figure 6.4). The anti-MMP-9 antibodies were tested in parallel with two BSP sera on a Western blot and the anti-MMP-9 antibodies reacted with proteins of 27, 50 and 75-100 kDa MW (Figure 6.5). The MAb1D5, MAb5, anti-MMP-9 were tested in parallel on a Western blot, using a 7.5% gel in an attempt to separate out the bands around 75-100 kDa MW, and this demonstrated that the MAb1D5 antibody reacted with a protein closer to 100 kDa MW compared to the anti-MMP-9 that reacted with a protein of approximately 80 kDa (Supplementary data: Figure 6.1S). Both antibodies reacted with a protein of approximately 25 kDa MW. A zymogram was run with and without EDTA (metalloprotease inhibitor). In the absence of EDTA two distinct bands above and below the 100kDa protein marker (approximately 95 and 105 kDa, respectively) were apparent. Addition of EDTA to the zymogram abolished these two bands (Supplementary data: Figure 6.2S).

141 Figure 6.4 Serum antibodies from IBS-P clinical subgroup and monoclonal antibodies MAb1D5 and MAb5 reacting with Blastocystis proteins in Western blot

L MAb5 MAb1D5 |______IBS-P 1,2,3,4______|

150kDa

100kDa

75kDa

50kDa

37kDa

25kDa

15kDa

IBS-P= Irritable bowel syndrome patient positive for Blastocystis ; Patient 1,2,3,4 Mab=monoclonal antibody L-protein standard ladder kDa=kiloDalton molecular weight

142 Figure 6.5 Serum antibodies in IBS-P clinical subgroup and MAb anti-proMMP-9 reacting with Blastocystis proteins in Western blot

IBS-P anti-proMMP-9 1 2 kDa

100

75 50

25

IBS-P= Irritable Bowel Syndrome patient positive for Blastocystis: Patient 1,2 MAb: monoclonal antibody anti-proMMP-9=antibody to pro-matrix metalloprotease-9 kDa=kiloDalton molecular weight

143 Figure: 6.S1 Anti-proMMP-9, MAb1D5, MAb5 reacting with Blastocystis proteins in a Western blot*

anti-proMMP-9 MAb1D5 MAb5 kDa 250

150

100

75

50 35 25

anti-proMMP-9=antibody to pro-matrix metalloprotease-9 Lane 1 MAb=monoclonal antibody MAb1D5 (Lane 2) and MAb5 (Lane 3) kDa=kiloDalton molecular weight *7.5% gel

144 Figure: 6.S2 Zymogram of Blastocystis WR1 proteins without addition of EDTA

kDa Ladder WR1 WR1 250

150  100 

Arrows indicate level of positive zymogram bands

145 6.5 Discussion Serum IgA levels were found to be significantly lower in subjects that were positive for Blastocystis sp. even after adjusting for the confounding variables of gender, age, season and Asian ethnicity that have been reported to influence serum IgA levels (Stoop et al. 1969; Weber-Mzell et al. 2004). The level of IgA correlated with carriage but not symptoms suggesting that a low serum IgA may permit Blastocystis sp. to establish in the gut lumen but may not influence pathogenicity. The association of Blastocystis carriage and low serum IgA would be consistent with the fact that IgA is the major immunoglobulin class involved in mucosal defence (Aghamohammadi et al. 2009) and is known to prevent adherence of pathogens to the gut luminal surface, bind toxins and inhibit antigen absorption. IgA production increases after three months of age in parallel with the change from a sterile gut to one with a complex commensal intestinal microbiota (Macpherson et al. 2008). Secretory IgA can be effective against pathogens as an innate non-specific local response or develop into an adaptive specific systemic response after presentation of luminal antigen to gut associated lymphoid tissue (GALT) and subsequent generation of IgA-producing plasma cells in the intestinal mucosa. The complex relationship of IgA to the regulation or monitoring of the intestinal microbiota is not understood. Proteases derived from Blastocystis cell lysates have been shown to be capable of degrading secretory IgA (Puthia et al. 2005). Low serum levels of IgA may be due to a number of causes including partial or complete selective IgA deficiency (Latiff & Kerr 2007) and likely influence host and microbial interactions. IgA deficiency is associated with an increased risk of giardiasis (caused by another non-invasive luminal parasite) as well as respiratory infections and allergic and autoimmune diseases (Aghamohammadi et al. 2009). Patients with autoimmune diseases and selective IgA deficiency more commonly exhibit the major histocompatibility complex (MHC) haplotype 8.1 and other non-MHC gene defects have been reported (Singh et al. 2014). IgG serum antibodies from many study participants in all subgroups reacted with proteins of approximately 27, 50, 60-65, 75-90, 95-105, and

146 150kDa MW in the Western blots using axenic WR1 (ST4) as antigen source. Although differences in the quantity or affinity of the antibodies were suggested by the intensity of the signals, no attempt was made to apply quantitation by densitometry. The rabbit-derived Blastofluor antibody, raised in rabbits primed with axenic ST3, showed a very similar reactivity pattern to proteins of 10, 17, 27, 37, 50, 75, 100 KDa MW when tested in parallel with human sera (Figure 6.2), suggesting that these particular Blastocystis antigens are common to both subtypes 3 and 4. Whilst some researchers have described different SDS-PAGE protein patterns in different Blastocystis organisms (Kukoschke & Muller 1991; Chen et al. 1999) others have found little difference in the protein sizes between different Blastocystis subtypes(Tan et al. 1996; Init et al. 2003). Over 40% of healthy asymptomatic patients negative for Blastocystis sp. reacted to Blastocystis antigens at 27, 50, 60-65 and 75-90 kDa, respectively (Table 6.3), suggesting that positive reactions with these proteins occurred either due to past exposure to Blastocystis sp. or as a result of a cross reactivity with the Blastocystis antigens. Only the antibody band sizes 95-105 and 150 kDA were significantly more common in those with current carriage of Blastocystis sp., occurring in approximately 60% of these individuals. However IBS groups, either positive or negative for Blastocystis carriage, when compared to HC displayed a different pattern with significantly higher number of reactions to 27, 50 and 75-95 kDa MW proteins. Overall these proteins were present in approximately 86-100% of IBS patients compared to 44-71% of HC. Although presence of these proteins was greatest in the Blastocystis-positive IBS group, it was still present in the IBS group negative for current carriage of Blastocystis sp.. This finding again suggests the positive reactions in the IBS-N group were either due to non- specific antibody cross- reactions or due to residual antibody from past exposure to Blastocystis organisms. The gut epithelial permeability is increased in IBS (Simren et al. 2013; Beatty et al. 2014) and it is probable that IBS patients have greater exposure to intraluminal antigens. Although the number of subjects in our clinical subgroups was relatively small our findings

147 suggest that exploration of the nature of these 27, 50, 75-95 kDa MW proteins may be useful in identifying biomarkers for IBS (current biomarkers are not able to reliably exclude organic disease (Corsetti et al. 2014)), identifying patients who are symptomatic due to Blastocystis infection and informing development of effective therapies. Several studies have reported that an antigenic protein with a MW of approximately 30 kDa is present in Blastocystis organisms (Lanuza et al. 1999; Hegazy et al. 2008) and differences in reactivity to this protein(s) between symptomatic patients and healthy controls have been reported (Hegazy et al. 2008; Abou Gamra et al. 2011). No reactivity with proteins of this size was reported in sera from patients with other parasitic infections suggesting the 30 kDa protein might be specific for Blastocystis organisms. A cytopathic monoclonal antibody, MAb1D5, previously shown to bind to the external surface of Blastocystis sp.(Tan et al. 1997), reacts with a 30kDa protein from Blastocystis isolates derived from humans (ST’s unknown), but reportedly not with antigen derived from WR1 (Tan et al. 2001). Peptide sequences derived from material obtained by 2-dimensional electrophoresis and Western blotting suggest that this 30kDa protein is closely related to a cysteine protease of the legumain type. Although no reactivity of the MAb1D5 was reported previously with subtype WR1 (Tan et al. 2001),and this was the subtype we used as an antigen source in our study, the 27 kDa protein- reactivity that we observed more frequently in the symptomatic group may nevertheless correspond to this previously described 30 kDa protein. The difference in reactivity may be accounted for by increased purity of MAb1D5 used in our study and differences in antigen preparation technique, and the MW size difference accounted for by differences in Blastocystis antigen source and preparation (e.g., subtype, level of glycosylation). The MAb1D5 tested in parallel with our human serum samples reacted strongly with a double band around 100 kDa MW and also with a 27 kDa protein in the Western blots (Figure 6.4). This pattern of reactivity (100 and ~30 kDa) against an antigen preparation using ST7 organisms was shown previously (Wawrzyniak et al. 2012) and suggests that the 30 kDa antigen may be a processed or degraded form of the 100kDa protein.

148 We also noted differences in IBS patients (IBS-P and IBS-N) with regard to reactivity to 50 and 75-95 kDa proteins. These proteins may correspond to the 50 and 118 kDa antigens (Hegazy et al. 2008), or the 60- 100 kDa sized proteases (Rajamanikam & Govind 2013) reported previously in symptomatic Blastocystis infected patients. Parasitic proteases are known to be highly immunogenic and may facilitate tissue invasion in the host. We searched for a protease that might be consistent with this role and sizes. Matrix metalloproteinase-9 (MMP-9) is one member of the zinc binding matrix metalloproteinases (MMP) that degrade components of the extracellular matrix including gelatin (gelatinase B), collagens, and elastin, thereby potentially aiding tissue invasion (Geurts et al. 2012), and has been shown to interact with the immune system including potentiation of IL-8 (Geurts et al. 2012). MMP-9’s are found in the cytosol, vesicles and cell membranes of cells (Vandooren et al. 2013). They exist as a pro-enzyme at around 92 kDa, an activated form around 84 kDa and truncated forms at lower sizes around 50 kDa in humans (Shimokawa et al. 2002). The pro-MMP-9’s are activated by other proteases, detergents and heat. We tested an antibody specific for human pro-MMP-9 in parallel with human sera in our Western blots and found that it reacted with Blastocystis proteins of approximately 75-100, 50 and 27 kDa MW, respectively, and that the pattern differed from the MAb1D5 reactivity. MAb1D5 reacted with a different distinct protein of approximately 95 kDa as well as a 27 kDa protein. A subsequent zymogram performed using antigen that had been pre- treated with protease inhibitors minus EDTA showed a positive double band around 100 kDa that was abolished with EDTA. These findings suggest that a metalloproteinase around 100 kDa MW is present in Blastocystis sp.. Inherent differences in gel runs and species variation of enzymes make it difficult to assess if this 100 kDa zymogram band is related to the Western blot 75-95 kDa bands we noted to be of significance in symptomatic patients. The findings nevertheless suggest that a MMP-9 isoform is present in Blastocystis organisms, and the pro-MMP-9 band may correspond to the 75-95 kDa protein, while the 50 kDa protein may possibly be the enzymatically active form and the 27 kDa form a truncated fragment.

149 The detection of a 27 kDa protein with both the anti-MMP-9 and the MAb1D5 suggest that this band may contain more than one protein, including a truncated fragment of a MMP-9 isoform. Previous studies investigating proteases present in Blastocystis sp. have suggested that cysteine proteases constitute the major class of proteases in this organism (80% protease activity shown to be inhibited by cysteine protease inhibitors) (Sio et al. 2006). However it may be difficult to account for proteases that are present as pro- enzymes that require activation or that may be inhibited from converting to active forms by other proteases. Protease activity at 32 kDa was reported to be significantly increased in symptomatic patients (Abdel-Hameed & Hassanin 2011; Abou Gamra et al. 2011) but may be due to a mixture of enzymes of that size. The genome of ST7 has been sequenced and this has allowed the prediction that all classes of proteases (cysteine, serine and metalloproteases) could be secreted by the parasite (Denoeud et al. 2011) . The majority (91%) are predicted to be of the cysteine protease and none were specifically predicted, at this level of genome testing sensitivity, to be a metalloprotease of the MMP-9 class. Preliminary analysis of the ST4 genome has shown the presence of up to 30% genes with no ortholog present in the ST7 genome (Wawryniak et al. 2015) emphasising the genetic diversity present in this organism. Additionally, a recent study has discovered that 15% of nuclear genes in Blastocystis sp. Subtype 7 uniquely use a polyadenylation-mediated creation of termination codons (Klimes et al. 2014) and predicted genes may need to be re-evaluated in the light of this new information.

6.6 Conclusions Low serum IgA levels may facilitate Blastocystis carriage but are not associated with the development of IBS gastrointestinal symptoms. Antibodies to Blastocystis organisms are commonly found in the serum of many healthy individuals, but antibodies to 27, 50 and 75-95kDa MW proteins are significantly increased in IBS patients regardless of Blastocystis status. Some of these serum antibodies are likely to be reacting to a previously reported

150 cysteine protease, but others may be reacting with another class of proteases, the metalloproteases.

6.7 Acknowledgments We are grateful for the kind donations of important research material by Dr Linda Dunn, QIMR, Australia (axenic cultures of Blastocystis for antigen preparation), Dr Rick Gould, Antibodies Inc, USA (Blastofluor antibody) and Dr Kevin Tan, NUS, Singapore (MAb 1D5, MAb5), who also read the manuscript and contributed useful comments. This study was funded by the Royal Australian College of Physicians “Murray-Will Fellowship for Rural Physicians” (awarded to R.Nagel, 2012).

151 152

7 Differences in faecal microbiota profiles between Blastocystis-positive and negative irritable bowel syndrome patients

R Nagel, RJ Traub, RJN Allcock, MMS Kwan, H Bielefeldt-Ohmann

Submitted for publication September 2015

153

Differences in faecal microbiota profiles between Blastocystis-positive and negative irritable bowel syndrome (IBS) patients

Robyn Nagel, Rebecca J Traub, Richard JN Allcock, Marcella MS Kwan, Helle Bielefeldt-Ohmann

7.1 Abstract Background and aim: Characteristic differences in the faecal microbiota (FM) have been described in irritable bowel syndrome (IBS). Blastocystis carriage has also been associated with IBS although overall the organism constitutes less than 1% of the total FM. We investigated whether the carriage of Blastocystis in IBS patients affects the FM profile. Methods and results: Forty patients with diarrhoea-predominant IBS (26 Blastocystis-positive and 14 Blastocystis–negative) and fifty-seven healthy controls (HC) (42 Blastocystis-positive and 15 Blastocystis-negative) submitted faecal samples for high throughput culture-independent metagenomic analysis using 16S ribosomal RNA gene sequencing. Differences in abundance of bacteria in these IBS and HC groups were evaluated from phylum to genus level. Significant changes were seen in the two dominant phyla, namely a rise in Firmicutes and a statistical significant reduction in relative abundance of Bacteroidetes (with a three fold increase in the Firmicutes: Bacteoridetes ratio) were observed in IBS patients, regardless of Blastocystis infection status. Further subgroup analysis showed Blastocystis-positive IBS patients had significant reductions in abundance at the phylum level of Actinobacteria (p= 0.015) and Firmicutes (p=0.013), and significant increases in Bacteroidetes (p= 0.05), “Other” (unidentified bacteria) (p=0.004) and Proteobacteria (p=0.018) compared to Blastocystis-negative IBS patients. These changes were reflected in a lower average Firmicutes: Bacteroidetes ratio of 6.19 in Blastocystis-positive compared to 9.00 in Blastocystis-negative

154 IBS patients. Minor differences at genus level were also seen between the IBS groups in Anaerostipes, Blautia and Alistepes spp. Conclusion: Significant differences in the FM between Blastocystis-positive and Blastocystis-negative diarrhoea-predominant IBS patients were found suggesting that Blastocystis carriage either alters the FM directly, or changes in the FM allow Blastocystis spp. to flourish. These differences in FM profile may define a separate clinical subtype of IBS. Key Words: faecal microbiome, Blastocystis , Irritable Bowel Syndrome

7.2 Background The human gut is sterile at birth and during early infancy is colonised by organisms originating from the mother, diet and environment. The adult human gut is host to 100 trillion commensal organisms that live in facilitative harmony with the 10 trillion cells in the body (Bonfrate et al. 2013). Over 90% of the faecal mass is comprised of microbes, the “faecal microbiome” (FM), comprising bacteria (93%), viruses (5.8%), archaea (0.8%) and eukaryotes (0.5%)(Arumugam et al. 2011). This intestinal microbiota is estimated to contain 3.3 million gene copies (the microbiome) vastly outnumbering the estimated 20,000 human genes and is increasingly recognised as an additional “human organ” with important neuroendocrine, metabolic, nutritional and protective functions (Albenberg & Wu 2014). The FM produces micronutrients Vitamin B and K and short chain fatty acids (SCFA) that are a primary energy source for colonic mucosal cells. The bacteria facilitate fermentation of food for host metabolism, produce neurotransmitters that are absorbed systemically, conduct immune priming, including differentiation of T helper cells, in the host and protect against other luminal pathogens (Furusawa et al. 2014). Metagenomic analysis, using culture-independent high throughput 16S ribosomal ribonucleic acid (rRNA) quantitative gene sequencing and microarrays, allows genetic insights into the FM with most attention given to the bacterial populations. This technique allows taxonomic classification of organisms from phylum to species. Of the fifty bacterial phyla humans host, fifteen are found in the gut and 90% of the colonic microbiota consists of two

155 dominant phyla, namely Firmicutes and Bacteroidetes, with great individual variability seen at species and strain level. Over 80% of bacterial species identified with these techniques remain uncultured (Shanahan 2012). Although eukaryotes comprise less than 1% of the total FM, metagenomic analysis has identified thirty-seven eukaryotic species in the faeces of a healthy adult including Blastocystis spp., eighteen plant species and eighteen fungal species (Gouba et al. 2013). A recent study of 105 healthy adults in Ireland showed the prevalence of Blastocystis carriage to be as high as 56%, with diverse subtypes and with stable carriage seen over a duration of 6-10 years (Scanlan et al. 2014) and concluded that Blastocystis carriage is likely one of the components of a healthy FM. In the newborn infant the gut microbiota is aerobic, but within days becomes anaerobic, and within two years is relatively stable till advanced age (Arrieta et al. 2014). Some of the factors that influence the human faecal microbiota include mode of delivery (vaginal versus Caesarean), feeding patterns in early infancy (breast feeding duration, time of introduction of foods), immunisation, antibiotic usage, sanitation (Arrieta et al. 2014), gender (Aguirre de Carcer et al. 2011) and long-term dietary choices. For example, Prevotella species (Bacteroidetes phylum) dominate in rural Africa, where the diet is high in vegetables, fruit and fibre compared to Europe (De Filippo et al. 2010). Irritable bowel syndrome (IBS) is a chronic heterogeneous condition affecting approximately 10% of the population worldwide (Canavan et al. 2014a). The disease is defined by a clinical symptom complex description (Rome III) and classified according to the predominant bowel habit, namely diarrhoea, constipation or “mixed” diarrhoea/constipation (Longstreth et al. 2006). Although the disease is benign, there are high direct and indirect health costs associated with diagnosis and management of this condition (Canavan et al. 2014b). The cause of IBS is not known, although 10% of cases of IBS arise after an enteric infection (Beatty et al. 2014), and improvement has been reported after broad spectrum antibiotic therapy (Laterza et al. 2015). Blastocystis spp. are three times more likely to be found in the stools of patients with diarrhoea-predominant IBS compared to healthy controls making this an organism of particular interest in the FM of patients

156 with IBS (Yakoob et al. 2010). No definite association between pathogenicity and specific Blastocystis subtype has been found. The FM is altered in various gastrointestinal diseases including IBS, leading researchers to ask whether this disease is caused by particular organisms within this altered FM or whether the disease is the end result of an abnormal host immune response to certain FM (Shanahan 2012). In adults with IBS the FM has characteristically displayed decreased diversity of organisms, temporal instability, changes in the dominant phyla, namely an increased Firmicutes:Bacteroidetes ratio, and changes in the minor phyla (decreased Actinobacter and increased Proteobacteria) (Simren et al. 2013). Changes in the abundance of certain bacterial families/species are also reported including Bifidobacterium, Lactobacillus (phylum:Actinobacter), Bacteroides, Prevotella (phylum:Bacteroidetes), Blautia, Clostridium, Collinsella, Coprococcus, Staphylococcus, Ruminococcus, Veilonella (phylum:Firmicutes) and Enterobacteria (phylum:Proteobacteria) (Malinen et al. 2005; Lyra et al. 2009; Rajilić–Stojanović et al. 2011; Jeffery et al. 2012b; Bye et al. 2014). A previous study has suggested that irritable bowel subtypes may be able to be characterised by their FM profile and that these subtypes do not necessarily correspond to their clinical categorisation (Jeffery et al. 2012b). IBS symptoms may have multiple causes and we hypothesized that Blastocystis spp. are one specific cause of IBS symptoms, and therefore it may be possible to identify differences in the bacterial FM between Blastocystis-positive and Blastocystis-negative IBS patients. In this study we compared the FM in diarrhoea-predominant IBS patients, positive and negative for Blastocystis and healthy controls, positive and negative for Blastocystis carriage.

7.3 Methods 7.3.1 Study outline Forty patients presenting with diarrhoea-predominant IBS to the Toowoomba Gastroenterology Clinic and fifty-seven healthy volunteers (HC) were enrolled in the study. Single baseline faecal samples were collected

157 from all participants and tested for the carriage of Blastocystis and subjected to metagenomic analysis. The faeces were frozen at -20C within 4 hours of collection. The faecal microbiome profile of all patients was examined and comparative analysis was made between subjects with IBS and HC, and between Blastocystis-positive and Blastocystis-negative IBS (IBS-P, IBS-N) and HC (HC-P, HC-N) subjects. The study was approved by the University of Queensland Medical Research Ethics Committee and was part of two clinical trials registered with the Australian and New Zealand Clinical Trials registry (http://www.ANZCTR.org.au) ACTRN: 12610001066077 and 12611000918921.

7.3.2 Inclusion protocol Patients presenting to the clinic with chronic diarrhoea from 09/04/2008 to 20/2/14 were assessed clinically (Nagel et al. 2014). Forty consecutive, eligible symptomatic patients who had no other cause for symptoms identified and who fulfilled the Rome criteria for diarrhoea-predominant IBS (Longstreth et al. 2006), namely symptoms of chronic abdominal pain, occurring frequently over the last 3 months, associated with defecation and the passage of predominantly loose stool, were enrolled in the study. Healthy volunteers were recruited from the University of Queensland and from asymptomatic members of households containing a symptomatic Blastocystis-positive patient. All patients who were invited to participate consented to enrolment and completed the study. A full medical history was taken only from symptomatic patients. No record was taken in any subject of diet, pre- or pro- biotic intake.

7.3.3 Exclusion protocol Non-pregnant subjects between 15 and 75 years of age were recruited for the study. Patients with significant systemic diseases or co-morbidities were excluded. Subjects were excluded if they had had a course of any antibiotic in the preceding six weeks prior to stool collection.

158 7.3.4 Diagnostic methods All samples were run in parallel for the presence of Blastocystis spp. using a simple unstained wet faecal smear, xenic in vitro culture (XIVC) and PCR (confirmed as Blastocystis spp. using DNA sequencing to enable subtyping) as previously described (Nagel et al. 2014) . A patient was considered to be positive if any one of the tests was positive.

7.3.4.1 Metagenomic analysis 16S rRNA PCR Extracted DNA was quantified using a Qubit fluorometer and 1ng samples were amplified using the 16S V4/5 primers (515F/806R). The primer sequences and protocol was based on Caporaso et al 2012, with local modifications. Specifically, we used a mixture of gene-specific primers and gene-specific primers tagged with ion torrent-specific sequencing adaptors and barcodes. The tagged and untagged primers were mixed at a ratio of 90:10. Using this method, the approximately 10 cycle inhibition observed by using long tagged primers could be reversed and hence we achieved amplification of all samples using 18-20 cycles, thus minimising primer-dimer formation and allowing streamlined downstream purification. Amplification was confirmed by agarose gel electrophoresis and product formation was quantified by fluorometry. Up to 100 amplicons were diluted to equal concentrations and adjusted to a final concentration of 15pM. Templated Ion Shere Particles (ISP) were generated on an Ion One Touch 2 (Life Technologies) using 400bp templating kit and sequenced on a PGM (LifeTechnologies) for 800 cycles using 400bp sequencing kit yielding a modal read length of 309bp. Reads were trimmed for quality purposes using TorrentSuite 4.0.2(Caporaso et al. 2010).

7.3.4.2 Analysis of sequences Metagenomic analysis using culture-independent high throughput 16S rRNA quantitative gene sequencing and microarrays was performed on the PCR derived sequences. The data was analysed using software analysis program Quantitative Insights into Microbial Ecology (QIIME, version 1.7)(Caporaso et al. 2010). The following commands were applied to the derived 16S rRNA sequences (Cock et al. 2010) (i) the rRNA sequence

159 FASTq reads were separated into two separate libraries, one containing “sequences (FASTA files)” and the other “quality of DNA information (QUAL)” scores) (ii) Each file in the sequence library was assigned a unique subject identity barcode, creating a “mapping” library (iii) PCR “mixed sequence” chimeras were removed using a reference file and identification of “de novo” chimeric sequences (iv) Operational taxonomic units based on 97% specific16S rRNA gene sequence identities were used to distinguish different species of microbe and these were grouped into their most likely phylum/class/order/family/genus using GreenGenes database, Version 12_10) (McDonald et al. 2012). The reads were repeated to refine the dataset to a consistent number of reads and taxonomic identification. Genomic analysis was obtained from taxonomic Level 1-6, but not including Level 7 species subtype identification (Caporaso et al. 2010).

7.3.5 Statistical analysis Statistical analysis was carried out using IBM SPSS Statistics (IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp). Percentages of gut microbial flora at phylum (L2) and species level (L6) across the four clinical groups were analysed using Kruskal-Wallis test. Those species with a significant overall difference were further analysed for between group differences using the following equation(Siegel & Castellan 1988): RBari - RBarj | >= Z*Sqrt[ (N*(N+1)/12)*(1/ni + 1/nj) ] where RBari, RBarj, ni, nj are the mean of the ranks and the sample sizes associated with the i-th and j-th groups. N is the total sample size and Z is the critical value from the standard normal curve (Z = 2.638 for k=4 groups and where alpha=0.05/(k*(k -1)) = 0.0083333.). Firmicutes: Bacteiodetes ratio is calculated by dividing individual abundance of Firmicutes by individual abundance of Bacteroidetes. Mann-Whitney test was used to test any significance between total IBS and total HC groups.

160 7.4 Results 7.4.1 Subjects Ninety-seven participants comprised of 40 symptomatic diarrhoea- predominant irritable bowel syndrome (IBS) patients and 57 asymptomatic healthy controls (HC) were included in the study. Mean ages of IBS and control subjects were 45.2 (range, 23 – 77) and 42.0 (range, 15 - 74) years old, respectively. Subjects were further classified as to whether they tested positive or negative for the parasite Blastocystis. A significant difference amongst the four participant groups was found in terms of gender (λ2=15.25, p<0.05), but not age (F=1.11, p= 0.30). Blastocystis subtype 3 (ST-3) was most common in both IBS and HC groups, ST-4 was seen more commonly in the IBS group but this did not reach statistical significance. Characteristics of participants in clinical subgroups are shown in Table 7.1.

Table 7.1 Characteristics of clinical subgroups IBS-P IBS-N HC-P HC-N (n=26) (n=14) (n=42) (n=15) Age 45.6 ± 13.6 44.5 ± 14.3 41.8 ± 15.6 42.4 ± 15.2 (mean±sd) Female 20 (76.9) 11 (78.6) 18 (35.7) 10 (66.7) (n, %) Blastocystis subtypes (n, %) ST1 5 (19.2) 12 (28.6) ST3 8 (30.8) 12 (28.6) ST4 7 (26.9) 6 (14.3) Other subtypes 6 (23.1) 7(28.6) (including ST2,5-8) Medications (n, %)

Subjects on PPI/H2Bl 7 (27%) 4 (29%) Nil or OCP only 14 (54%) 4 (29%)

IBS-P: Patients with irritable bowel syndrome positive for Blastocystis IBS-N: Patients with irritable bowel syndrome negative for Blastocystis HC-P: Healthy controls positive for Blastocystis HC-N: Healthy controls negative for Blastocystis

PPI=proton pump inhibitor therapy H2Bl=Histamine 2 blocker therapy OCP=oral contraceptive pi

161 Twenty-eight symptomatic patients were taking prescribed medications (16 of the IBS-P, 12 of the IBS-N), including amitriptyline, oral contraceptive pill, proton pump inhibitors, histamine 2 blockers, thyroxine, cholesytramine, antihypertensives and non-steroidal agents (Table 7.1).

7.4.2 Bacterial phyla seen in the study participants The major phyla seen in all subjects were Firmicutes (46.27% of all phyla) and Bacteroidetes (40.99%). The other most common phyla in descending order of abundance were Proteobacteria, Actinobacteria, Verrucomicrobia and Fusobacteria and this pattern was observed in the first two phyla and varied in the remaining phyla in all the subject subgroups (IBS, IBS-P, IBS-N, HC, HC-P, HC-N) (Table 7.2a). Eight phyla (including “Unclassified”) were present in only a few patients (<15% of total cohort) in one or more of the clinical groups (see Table 7.2a) and comprised less than 0.05% abundance (of the clinical subgroup mean value). The OD1 phylum was not found in any participant. The Spirochaetes, TM7 and Tenericutes phyla were only found in three individuals in the HC-P group. The Unclassified “phylum” was found in five individuals in the IBS-P group. The phylum Elusimicrobia was not found in the IBS-N group. 7.4.3 Comparison of bacterial profiles in subjects with and without IBS The FM was compared between IBS and HC from phylum to species level. The two major phyla in subjects in both IBS and HC groups were the Firmicutes (IBS vs HC: 51% vs 43% of total FM, p=0.12) and Bacteroidetes (IBS vs HC: 34% vs 46% of total FM, p<0.05) (7.2a; Figure 7.1). Bacteroidetes abundance was significantly reduced in the IBS group and the Firmicutes:Bacteroidetes ratio was three times higher in the IBS group compared to the HC (Table 7.2b) (p=0.02). Increased abundance of Proteobacteria and Actinobacter phyla in the IBS group was observed but did not reach statistical significance.

162 Table 7.2a Abundance of bacterial phyla seen in clinical subgroups (%)

Phyla Total IBS IBS-P IBS-N Total HC HC-P HC-N (n=40) (n=26) (n=14) (n=57) (n=42) (n=15) Actinobacteria 3.565 2.906# 4.789 2.014 0.668 5.784 Bacteroidetes 33.780* 39.171+ 23.769 46.045 48.467 39.263 Cyanobacteria/ Chloroplast 0.031 0.045 0.006 0.022 0.025 0.013 Elusimicrobia 0.02 0.025 0 0.002 0.0005 0.003 Firmicutes 51.037 44.350# 63.456++ 42.924 41.970 45.593 Fusobacteria 0.272 0.031 0.720 0.082 0.110 0.003 Lentisphaerae 0.020 0.025 0.010 0.018 0.020 0.011 Other 3.786 4.893 1.729 3.185 3.481 2.356 Proteobacteria 6.857 8.032+ 4.676 5.286 4.758 6.764 Spirochaetes 0 0 0 0.001 0.002 0 Synergistetes 0.007 0.004 0.011 0.003 0.003 0.004 TM7 0 0 0 0.0004 0.0005 0 Tenericutes 0 0 0 0.004 0.005 0 Verrucomicrobia 0.309 0.217 0.479 0.297 0.347 0.159 Unclassified 0.004 0.006 0 0 0 0

163 Table 7.2b Firmicutes: Bacteroidetes Ratio in clinical subgroups (abundance of Firmicutes/abundance of Bacteroidetes)

Total IBS IBS-P IBS-N Total HC HC-P HC-N Firmicutes:Bacteroidetes Ratio 7.1313.40 6.1914.85 9.0010.16++ 2.287.19 1.421.179 5.0814.51 (mean  standard deviation) * significant difference in Total IBS cf Total HC # significant reduction in IBS-P cf IBS-N +significant increase in IBS-P cf IBS-N ++ significant increase in IBS-N cf HC-N IBS-P: Patients with irritable bowel syndrome positive for Blastocystis IBS-N: Patients with irritable bowel syndrome negative for Blastocystis HC-P: Healthy controls positive for Blastocystis HC-N: Healthy controls negative for Blastocystis cf=compare

164 Figure 7.1 Comparison of bacerial phyla profiles in subjects with and without IBS

165 A number of genera of microbes showed differences in abundance between IBS and HC subjects, and many of these differences reached statistical significance (Table 7.3). At genus level the organisms that were statistically increased in IBS patients were Methanobrevibacter, Actinomyces, Gordonibacter, Enterococcus, Streptococcus, Blautia, Lachnospiacea_incertae_sedis, Papillobacter, Ruminococcus and Gemminger. The genera significantly decreased in IBS patients were Olsenella, Parabacteroides, Peptococcus, Cantenibacterium, Alisonella, Dialister, Megamonas and Megaspaera (Table 7.3).

7.4.4 Comparison of bacterial phyla across the four clinical subgroups Differences were found between bacterial phyla profiles in IBS-P and IBS-N patients (Table 7.2a; 7. 2a,b). Comparison of IBS-P to IBS-N subjects showed significant reductions in abundance in Actinobacteria (p= 0.015) and Firmicutes (p=0.013), and significant increases seen in Bacteroidetes (p= 0.05), Other (unidentified bacteria) (p=0.004) and Proteobacteria (p=0.018). The only other difference seen at phylum level between the clinical subgroups was an increased abundance of Firmicutes in IBS-N compared to HC-N (p=0.05) (Table 7.4). The mean FB ratio of the IBS-P patients (6.19) was lower than the ratio found in the IBS-N patients (9.00), and similarly the FB ratio in the HC-P (1.42) was lower than that seen in the HC-N (5.06) although these trends did not reach statistical significance (7. 2b). The minor phyla only have small numbers of subjects in each group making meaningful statistical interpretation difficult (7.2b).

7.4.5 Comparison of bacterial genera across the four clinical subgroups Significant differences in bacterial profiles at genus level were found between the clinical subgroups (Table 7.4). When comparison was made between the IBS-P and the IBS-N group a reduction in bacterial abundance was found in Anaerostipes, Blautia (Firmicutes phyla) and Other spp and increased abundance was noted in Alistepes spp. (Bacteroidetes phyla) (p<0.05).

166 Table 7.3 Comparison of bacterial profiles in subjects with and without IBS Phylum (L2) Class (L3) Order (L4) Family (L5) Genus (L6) Euryarchaeota Methanobacteria  Methanobacteriales  Methanobacteriaceae  Methanobrevibacter 

Actinobacteria Actinobacteria Actinomycetales  Actinomycetaceae  Actinomyces  Bifidobacteriales Bifidobacteriaceae Other  Coriobacteriales Coriobacteriaceae Gordonibacter  Olsenella  Bacteroidetes  Bacteroidia  Bacteroidales  Porphyromonadaceae Parabacteroides 

Firmicutes Bacilli Lactobacillales Enterococcaceae Enterococcus  Streptococcaceae  Streptococcus  Clostridia  Clostridiales  Lachnospiraceae  Blautia  Lachnospiracea_incertae_sedis  Peptococcaceae 1  Peptococcus  Rumincoccaceae Papillibacter  Ruminococcus  Erysipelotricha Erysipelotrichales Erysipelotrichaceae Cantenibacterium  Other  Negativicutes  Selenomonadales  Veillonellaceae  Allisonella  Dialister 

167 Megamonas  Megaspaera 

Proteobacteria Alphaproteobacteria  Rhizobiales Hyphomicrobiaceae  Gemmiger  (Unclassified)  Other  Other  Other  Other  L=level Bold entries indicate significant difference between groups (p<0.05)  indicate significant (p<0.05) increase or decrease in IBS relative to healthy subjects, respectively

168

Table 7.4 Comparison of abundance of selected** bacterial species across the four clinical subgroups Species Mean ± SD (%) IBS-P (n=26) IBS-N (n=14) HC-P (n=42) HC-N (n=15) Actinomyces spp. ^ 0.019± 0.052 0.033± 0.042 0.014± 0.005 0.016± 0.022 Bifidobacterium spp. ^ 2.440± 8.130 3.636± 5.267 0.408± 1.039 5.265± 9.671 2.1.1.2.1.4 Other 0.039± 0.156 0.023± 0.031 0.003± 0.013 0.085± 0.237 Eggerthella spp. ^ 0.018± 0.058 0.116± 0.182 0.003± 0.010 0.037± 0.112 Gordonibacter spp. 0.005± 0.012 0.011± 0.028 0 0.003± 0.007 Olsenella spp. 0 0 0.006± 0.023 0 Butyricimonas spp. 0.188±0.267 0.127±0.413 0.351±0.521 0.481±0.133 Prevotella spp. ^ 2.336± 5.513 0.034± 0.112 4.344±8.583 3.917± 11.803 Alistipes spp. * 6.142± 5.326 1.843± 2.460 5.473±5.802 4.336± 5.389 Lauconostoc spp. ^# 0.035± 0.058 0.001±0.005 0.063±0.101 0.004± 0.011 Weissella spp. # 0.038±0.066 0.003±0.011 0.090±0.172 0.001± 0.005 Streptococcus spp. ^ 1.276±2.612 0.646±0.631 0.192±0.273 0.516± 0.587 Eubacterium spp. 0.004± 0.013 0.043±0.101 0.004±0.013 0.013± 0.036 Anaerostipes spp.*^ 0.248±0.498 2.067±2.149 0.123±0.157 0.586± 1.461 Blautia spp. *^ 1.130±1.923 8.011±8.001 0.450± 0.366 4.943± 9.926 Lachnospiracea_incertae_sedis spp. ^ 2.092±3.592 3.163±2.645 0.813±0.769 2.400± 4.333 Clostridium XI spp. ^ 0.224±0.378 0.850±0.766 0.255±0.505 0.453± 0.602 Clostridium IV spp. # 0.829±1.501 0.487±0.422 1.037±1.373 0.268± 0.376

169 Oscillibacter spp. 0.675± 1.052 0.354± 0.284 0.655±0.534 0.360± 0.334 Papillibacter spp. 0 0.023± 0.065 0 0 Canternibacter spp. 0.004± 0.016 0 0.154±0.546 0.040± 0.118 Coprobacillus spp. 0.008± 0.032 0.070± 0.138 0.006±0.019 0.017± 0.051 Allisonella spp. 0.008± 0.028 0 0.016±0.049 0.129± 0.414 Dialister spp. % 0.227± 0.727 2.009± 4.290 3.799± 11.633 3.761± 5.844 2.9.1.1.1.1 Other spp. *% 4.893± 5.681 1.726± 2.834 3.481±3.747 2.356± 5.774 2.10.2.1.4.3 Other spp. 0.049± 0.080 0.006± 0.017 0.050± 0.095 0.019± 0.067 Acinetobacter spp. 0.017±0.036 0.001±0.005 0.031±0.106 0 Bacteria, Other spp. 0.006±0.015 0 0 0 ** Selection based on overall statistical significance across all clinical groups, determined by Kruskal-Wallis test with p<0.05. * significant post-hoc difference IBS-P vs IBS-N ^ significant post-hoc difference IBS-N vs HC-P # significant post-hoc difference HC-P vs HC-N % significant post-hoc difference IBS-P vs HC-N

170 Figure 7.2a Comparison of major bacterial phyla across the four clinical subgroups

Figure 7.2b Comparison of minor bacterial phyla across the four clinical subgroups

171

No significant differences were noted between the IBS-P and HC-P but significantly decreased abundance was found between IBS-P and HC-N in Dialister spp. (p<0.05). No significant differences were found between the IBS-N and HC-N groups but between IBS-N and HC-P groups significantly increased abundance of Actinomyces, Bifidobacterium, Eggerthella, Streptococcus, Anaerostipes, Blautia, Lachnospiracea_incertae_sedis, Clostridium spp. and decreased abundance of Prevotella and Lauconostoc spp. was seen (p<0.05).

7.5 Discussion Previously reported changes in the FM of the two dominant phyla with a raised Firmicutes:Bacteroidetes ratio in IBS patients compared to HC were confirmed in this study comprising diarrhoea–predominant IBS patients. Similarly a trend of increased Proteobacteria abundance was consistent with previous reports in IBS patients but did not reach statistical significance. Accord with previous studies was also observed with significantly increased abundance in a number of species such as Ruminococcus, Papillobacter, Actinomyces, Lachnospiracaea_incertae_sedis, Blautia and Peptococcus spp. (Rajilić–Stojanović et al. 2011; Jeffery et al. 2012b; Simren et al. 2013; Bye et al. 2014). The L.incertae_sedis spp in particular have previously been reported to be enriched in diarrhoea-predominant IBS patients (De Filippo et al. 2010). Reductions in abundance were in accord for Parabacteroides spp., but results found in our study for Dialister, Veillonellaceae and Methanobrevibacter spp. contradicted previous results reported for (constipation-predominant) IBS (Bye et al. 2014). Veillonellaceae (and also Coprococcus spp.) have been reported to be more abundant in the constipation-predominant subgroup of IBS patients (Simren et al. 2013). Methanobrevibacter spp. were found to be increased in our study, whereas previously levels have been reported to be low, particularly in constipation- predominant IBS patients (Rajilić–Stojanović et al. 2011), and this had been thought to explain the increased flatulence reported by IBS patients.

172 Although our findings were consistent with other FM/IBS studies there were some notable differences. Actinomyces, Bifidobacterium and Faecalibacterium spp. are often reported to be reduced in abundance in IBS but were not significantly different from HC in this study. Bifidobacterium spp. may be underrepresented with qPCR techniques due to amplification biases (Simren et al. 2013). Although we found Bacteroidetes were significantly reduced in IBS, the genus Prevotella, which is often reported to be decreased in IBS patients (Simren et al. 2013) and increased in subjects with a high fibre intake (De Filippo et al. 2010), displayed no significant difference in abundance in our study. Currently three human gut enterotypes are described with predominance of Bacteroides, Prevotella and Ruminococcus (Arumugam et al. 2011) genera. The Ruminoccocus enterotype is less clearly defined than the first two and has been linked to IBS (Simren et al. 2013) and would be consistent with findings in this study. However, although this enterotype categorisation has been a useful starting point for understanding changes in the FM it now seems more likely that enterotypes follow a continuous gradient (Jeffery et al. 2012a) of profiles in the community rather than these three discrete enterotypes. Many studies of the FM in IBS patients have not separated out clinical subtypes of IBS (diarrhoea, constipation, or mixed-predominant) and as these have been shown to have different FM patterns (Lyra et al. 2009) this may account for differences in results between studies. Other confounding factors in most FM studies are the variable influences of non-standardised diets, pre- and probiotic therapy and other potential FM altering medications. In our study almost half the IBS patients were taking either no medication or only the oral contraceptive pill (OCP) and numbers of patients on medication were high in both IBS-P and IBS-N groups. Universally IBS has a female predominance (Canavan et al. 2014a). Sex hormone modulation of the gut microbiota has been reported (Mulak et al. 2014) and it is likely OCP therapy has some impact on the FM. Approximately one third of subjects with IBS were taking acid suppression therapy that has been reported to change the gastric microbiome significantly but have much less effect on the FM (Tsuda et al. 2015).

173 A multilayered mucosal barrier separates the luminal microbes from intestinal epithelial cells. The mucosa associated microbiota (MAM) has been shown to be different to the FM (Merga et al. 2014) and more abundant in IBS subjects compared to controls (Swidsinski et al. 2005). Discordant results in the microbiota profile have been reported between MAM and the FM (Codling et al. 2010; Ng et al. 2013; Bye et al. 2014) The FM, utilised in this study, constitutes 80% of the luminal microbiota and is much easier to sample by obtaining a simple faecal specimen. Differences between studies can still arise depending on whether fresh or frozen (and if so how quickly after passage of stool) specimens are used. The majority of the MAM are found on the luminal surface layer of mucus and only a small number of micro- organisms that possess suitable adhesion molecules can penetrate the mucus layer and directly interface with the epithelial cell (Hong & Rhee 2014). MAM, although low in abundance, may have more capacity to influence the intestinal epithelial barrier and immune system. Analysis of MAM requires an endoscopic procedure with a mucosal biopsy that by necessity, with the exception of distal colonic specimens, requires a cleansing bowel preparation that has been shown to disrupt the microbial/mucus niche environment (Hong & Rhee 2014). The main aim of this study was to determine whether particular subtypes of diarrhoea-predominant IBS, positive or negative for Blastocystis carriage, could be distinguished by their FM profile. Previous studies have investigated subgroups of IBS. Significant divergence in the FM between diarrhoea-predominant IBS and other IBS groups (constipation-predominant and mixed-symptom) has been reported (Lyra et al. 2009), highlighting the importance of categorising IBS patients clinically. In contrast another study has described three different FM profiles in IBS patients, one profile similar to HC, and with no correlation found between the FM profiles and the clinical categorisation (Jeffery et al. 2012b). In our study significant differences between the IBS-P and IBS-N groups were seen at the phyla and genus level although only small numbers of subjects were available for comparison at genus level. At the phylum level it was notable that the Firmicutes:Bacteroidetes ratio in IBS-P patients was lower than in the IBS-N group and this ratio mirrored the changes seen in abundance in these two

174 phyla. The minor phyla also showed differences in our study; the rise in Proteobacteria was significantly greater in the IBS-P compared to N and conversely the rise in Actinobacteria greater in the IBS-N group. Differences between IBS-P and IBS-N have been found in this study suggesting Blastocystis spp. are either directly or indirectly influencing the local gut microbiome or that the FM in some subgroups of IBS patients is more conducive to the carriage of Blastocystis. If IBS-P patients constitute a separate clinical subgroup to IBS-N patients, therapy may need to be different. There are many possible pathophysiological mechanisms that could account for differences in the FM in these two IBS groups. Although no definite biomarkers have been identified for IBS, an impaired intestinal mucosal barrier (IMB) and evidence of mild chronic immune activation are frequent findings. The IMB is a physical, biochemical and immune barrier that allows absorption of water and nutrients across the semipermeable barrier while providing defence against invasion of luminal microorganisms (Sanchez de Medina et al. 2014). The physical elements are the type and density of secreted mucin, secretory immunoglobulin A (sIgA), and the integrity of epithelial cells and the intervening tight junctions. Biochemical antimicrobial peptides are secreted from crypt cells and innate and adaptive immune responses defend the basal surface of the epithelial cell lining. We have previously reported that Blastocystis carriage, but not symptoms, is associated with lower levels of serum IgA (Nagel et al. 2015b). IgA is produced by plasma cells and 80% of these are located in the gut lamina propria. More IgA is produced than any other immunoglobulin type

(Suzuki et al. 2010) and it is the predominant mucosal immunoglobulin. sIgA is transported into the gut lumen via epithelial cells and physically prevents bacterial translocation across the epithelium. However, more importantly sIgA has modulatory effects on the FM. Lack of sIgA causes an expansion of anaerobic bacteria in the bowel and conversely increased IgA presence is reported to induce “tolerance” in host-microbial relationships (Fagarasan 2008). Blastocystis is strictly anaerobic and may prefer an expanded anaerobic microbiota environment. It is likely that the Blastocystis-positive subjects may have changes in the FM related to the differences in IgA levels.

175 However, this would not explain why only some Blastocystis positive subjects are symptomatic. Blastocystis spp. constitute only <1% of the FM. However, their presence in the lumen and close contact with the epithelial cells (Wang et al.

2014a), possibly facilitated by low sIgA levels, may allow greater interaction with the intestinal epithelial barrier and host-microbial immune pathways. Although some Blastocystis organisms may have unique, as yet undefined, pathological attributes, the host immune response may be an important factor in determining clinical response to Blastocystis infection. IBS patients are more likely to have polymorphisms that are associated with increased production of pro-inflammatory cytokines interleukin (IL)-2, 6, 8 and tumour necrosis factors and reduced production of anti-inflammatory IL-10 (Hughes et al. 2013). Polymorphisms of cytokines have also been reported in Blastocystis carriers, but interestingly although changes causing high IL-8 and low IL-10 levels were associated with IBS only the IL-8 polymorphisms were associated with Blastocystis carriage (Olivo-Diaz et al. 2012). We can speculate that Blastocystis organisms flourish in an anaerobic luminal microbial environment induced by low sIgA that also facilitates intimate contact of Blastocystis or other microbes with the IMB and reduces “immune tolerance”. The host- microbial response is characterised by high IL-8 levels in the Blastocystis carrier, and combined with low levels of IL-10 in some susceptible subjects causes chronic immune system activation and IBS symptoms.

7.6 Summary Differences in the FM in the dominant phyla and the Firmicutes: Bacteroidetes ratio are confirmed in diarrhoea-predominant IBS patients compared to HC subjects. Significant differences in the Actinobacteria, Bacteroidetes, Firmicutes and Proteobacteria phyla abundance are seen between IBS patients positive and negative for Blastocystis carriage. These differences might indicate that IBS patients with Blastocystis may constitute a separate clinical IBS group. These changes in the FM in symptomatic patients may be facilitated by low levels of IgA and IL-10 combined with high IL-8 levels in the host.

176

177

8 Discussion and conclusions

178 Discussions and Conclusions 8.1 Re-visiting the aims of this thesis 8.1.1 Aim of the thesis and interest in the organism Blastocystis spp. are the most common enteric parasite found in faecal samples submitted for examination in symptomatic patients. My aim was to investigate if these parasites are clinically relevant to humans. Blastocystis spp. were first described in 1911 (Zierdt 1991) and finally correctly classified taxonomically as a Stramenopile in 1996 (Silberman et al. 1996). Scientific interest in the organism over the first eight decades was limited. However, over the last twenty years citations in Pubmed45 have risen almost fourfold (from 24 to 81 articles published per year identified using “Blastocystis” as a search term). Research studies in Blastocystis spp. have been facilitated by technical developments such as PCR-based sub-typing and culture independent metagenomic techniques. Increasing interest in the pathogenesis of functional bowel disorders and the persistent reports of association of Blastocystis with clinical disease, particularly irritable bowel syndrome (IBS), in humans and the further possibility that infection may be zoonotic in nature has further boosted research interest in this organism.

8.1.2 Irritable bowel syndrome today Although benign IBS is a chronic disorder affecting approximately 7-21% of the population worldwide, usually with female predominance (Chey et al. 2015), direct IBS healthcare costs are estimated to be approximately US$5,000 per patient per year and a further US$2000 per patient cost to industry for absenteeism and presenteeism (Canavan et al. 2014b). These high costs do not include the financial cost of self-medication (up to 45% of affected individuals), the loss of productivity of carers or the non-financial costs of reduced quality of life. IBS is characterised by a symptom complex of abdominal pain and altered bowel habits (Longstreth et al. 2006) and as yet no reliable biomarkers have been identified (Corsetti et al. 2014). IBS is not a single disease but has been described as a symptom cluster resulting from diverse pathologies (Chey et al. 2015). Therapy for IBS is currently problematic and symptoms often persist for years. If different

5 http://www.ncbi.nlm.nih.gov/pubmed

179 causes of IBS could be identified, then it is possible that specific therapy could be targeted to that particular cause of IBS. Previous studies of IBS have identified quantifiable changes in intestinal permeability, gut immune activation status, gut motility, visceral sensitivity and enteric-brain communication (Chey et al. 2015). It is recognised that more than 10% of cases of IBS are reported to follow an infectious gastroenteritis (Schwille-Kiuntke et al. 2011) and broad-spectrum antibiotics may be useful therapy (Laterza et al. 2015). Recent research has further investigated microbial relationships in IBS and characteristic changes in the faecal microbiome have been described (Hong & Rhee 2014). Additionally, the carriage of Blastocystis is reported to be 2-3 times higher in IBS patients (Yakoob et al. 2004) suggesting that this organism may be a contributing cause of IBS.

8.1.3 Investigating the clinical relevance of Blastocystis spp. Investigation into any common clinically heterogeneous condition lacking reliable biomarkers, such as IBS, is challenging. Investigating the clinical effects of an organism that is difficult to eradicate and likely causing no adverse health outcomes in the majority of carriers is also problematic. Combining both investigations is fraught with difficulties, and thus, it was clear that a multi-pronged approach to the question of the clinical relevance of Blastocystis was required. In the first instance it was important to define as clearly as possible the symptomatic group of patients under investigation. It was necessary to exclude other pathology and ensure that no other aetiology for their symptoms could be found. Three clinical subgroups of IBS are described, and in order to reduce variables, wherever possible, only the diarrhea-predominant IBS group was studied in this project. All patients were recruited after presenting with symptoms to the Toowoomba Gastroenterology Clinic and having had appropriate clinical investigation. This recruitment process allowed tight selection of the IBS group under study and was an advantage that many previous studies lacked. In addition, many symptomatic patients actively seeking therapy and within an appropriate ethical framework, were offered specific antimicrobial therapy. Follow up of these patients enabled evaluation of eradication rates and to a lesser extent, clinical outcome.

180 Specific investigation into Blastocystis was conducted using deconvolutional microscopy to investigate the life cycle and PCR sub-typing to investigate possible subtype –associated pathogenicity and transmission routes. In order to evaluate whether immune responses in the host, rather than specific organisms determined symptoms, immunoglobulin levels, specific antibody/antigen reactions and finally changes in the faecal microbiome of the host were investigated.

8.2 Life cycle of Blastocystis spp. 8.2.1 Axenic cultures (Chapter 3) Although the interaction of Blastocystis spp. with the human gut was our primary interest, axenic cultures can provide a simpler model to investigate molecular and biochemical properties of the organism. Axenic culture was first described in1921 (Barret 1921) and although basic prerequisites such as neutral pH and strict anaerobic conditions have been defined, the culture technique remains difficult and unreliable. The main difficulties in separating the parasite from contaminating bacteria arise because Blastocystis is physically fragile, limiting methods of separation. More importantly, Blastocystis appears to require the presence of bacteria for sustained growth (Lanuza et al. 1997). Difficulty in developing axenic cultures of Blastocystis has been nominated, in 2012, as one of the significant barriers to Blastocystis research (Poirier et al. 2012). We attempted to develop axenic cultures of Blastocystis with the intended ultimate aim of investigating the intrinsic properties of specific organisms associated with symptomatic clinical presentation and also in order to derive pure Blastocystis protein for use in our Western blot antibody/antigen studies. Despite trialling three different culture methods, and culturing 210 times over 29 months, axenic Blastocystis grew in agar only four times and had no success in further continuous subculture in liquid medium. Albeit limited, the success we achieved was associated with the transfer of a large inoculum of Blastocystis organisms (>106) and transfer to agar in the first few days of the initial subculture from faeces. Initial subcultures had a bi-modal rate of growth pattern, peaking at 2-3 weeks and then again at 5-6 weeks before culture death. Our findings suggest that there may be some priming factor present in the gut

181 environment that stimulates reproduction and this factor is lost over time in relatively sterile culture conditions.

8.2.2 Features of Blastocystis spp. in xenic culture revealed by deconvolutional microscopy (Chapter 4) The Blastocystis parasite displays a bewildering array of morphological forms with great variation in size (Tan 2008). These attributes not only make it difficult to reliably identify the organism in stool but have confounded attempts to identify the separate stages of its life cycle. The forms that have been described include the vacuolated (VF), granular (GF), amoebic (AF) and cystic form (CF). Although it is certain that the robust cystic form transmits infection (Moe et al. 1997), the relationship of the other forms to each other is unclear. The function of the large vacuole compressing the peripheral cytoplasm of the VF and the multiple different granules present in the GF are unknown. The organism contains organelles such as nuclei, nucleoli, Golgi apparatus, lysosomes, endoplasmic reticulum but also contains often hundreds of unique mitochondria-like organelles (MLO’s) despite being a strict anaerobe (Stechmann et al. 2008). The VF and GF display a slimy carbohydrate-rich coat that extends for up to 10 m from the pitted surface membrane and often contains entangled bacteria (Zaman et al. 1999). These and other features have been described using light microscopy (LM) and transmission (TEM), scanning (SEM), freeze-etch electron (FEM) microscopy studies. These studies have often been conducted on dead organisms growing in axenic culture. We used deconvolutional microscopy (DM) with time-lapse imaging and fluorescent spectroscopy of xenic cultures of Blastocystis from stool samples of IBS patients and healthy pigs in order to observe these living organisms in their natural microbiological environment (Nagel et al. 2015a). Autofluorescence was observed in the VF, GF, cysts and less well in the AF in the TRITC excitation/emission light spectra (557/576-nM). The fluorescence was greatest in the central area of the cell with spots of light emission observed at the surface. This is the first report of green autofluoresence (GAF) in Blastocystis organisms. GAF is not a rare finding in microalgae and is known to occur in cyanobacteria. The Blastocystis genome suggests incorporation of cyanobacteria genes (Denoeud et al. 2011) and consequently the finding of GAF is not unexpected.

182 The chemical nature of GAF is not known although contribution from flavin-like molecules or luciferin compounds are suspected (Tang & Dobbs 2007). GAF was seen strongly in the central vacuole of the VF. Little is currently known about the contents of this vacuole except that it stains positively for carbohydrates (Yoshikawa et al. 1995c) and lipids (Yoshikawa et al. 1995d). FITC-conjugated Blastocystis- specific polyclonal antibody (Blastofluor) stained the surface of VF and GF and could be distinguished from GAF. Interestingly, Blastofuor did not stain the surface of AF suggesting different surface antigens are present in this form, and could not be distinguished from strong GAF in cysts. If “green spectrum” fluorescent-conjugated dyes are used in future studies care will need to be taken to differentiate the fluorescent dye from GAF. Detailed views of Blastocystis organisms were obtained with DM, but apart from observing binary fission developing in VF and GF, we did not observe changes form one morphological form to another. The fragile VF and GF‘s often abruptly ruptured over the period of observation. One time-lapse study of a large granular AF showed the form enlarging, with loss of surface granularity, and then subsequent expulsion of tear shaped granules that tracked along fissures to the surface. Although AF’s have been reported to be more prevalent in diarrheal stool the nature of the granules present in AF has not been studied. We were therefore unable to make any conclusions as to the identity of these intriguing tear-drop granules. Indentations on the surface of Blastocystis were previously well described using LM and EM, and channels apparently connecting the central vacuole to the subsurface have been described using FEM but not confirmed to reach the surface membrane. The use of DM allowed description of the existence of multiple surface pores of 1m diameter that actively opened and closed every 15 minutes over the 24hour period of observation. This finding would be consistent with active control of channels connecting to the central vacuole although the purpose is not clear. Surface pores have important nutritive and osmoregulatory functions in other protozoa (Bokhari et al. 2008). The presence of centrally connecting pores would be consistent with previous findings that ferritin particles, known to bind electrostatically to the lining of deep tubular pits present on the surface of the parasite, are rapidly internalised to the central vacuole (Dunn 1992).

183 Another curious phenomenon was recorded using DM, namely the extrusion of a viscous material from one or two points only on the surface of a VF. The extrusion occurred slowly over 11 hours and the organisms appeared to deflate but not rupture during this process leaving an adjacent amoebic-looking form. Typically VF has only a fine rim of cytoplasm surrounding the large central vacuole and this material most likely exuded from the vacuole. Some of the odd shaped forms seen in Blastocystis culture may be these extruded shapes, but we were not able to identify nuclear material within these using DAPI staining. Blastocystis is known to survive in varying osmotic conditions. The purpose of the vacuole is not known but this extrusion may represent an osmoregulatory process. In summary, using DM did not shed further knowledge on the life cycle of the parasite, but did allow observations of physiological properties and activities in the organisms that may inform our knowledge of function.

8.3 Subtyping and epidemiology 8.3.1 Expectations of subtyping Two years before I started this thesis standardisation of Blastocystis sub- types (ST’s) (based on the 18S rRNA gene) was published (Stensvold et al. 2007b) and in addition to improving diagnosis of Blastocystis carriage, this allowed a much more comprehensive comparison of studies than had been possible before using ribosome sequences. Previously diagnosis of Blastocystis carriage relied on light microscopy or culture and PCR techniques improved the diagnostic yield by 50% (Tan et al. 2010). Using these techniques, carriage of Blastocystis in healthy individuals has recently been reported to be 20-53% (Scanlan et al. 2014; O'Brien Andersen et al. 2015) and over 60% in some developing countries (Tan 2004). Most workers in the field expected that knowledge of the ST would allow markers for pathogenicity to be identified, in addition to giving insights into epidemiology and understanding of the zoonotic potential of this parasite. The latter two expectations have been realised but not the first.

184 8.3.2 Epidemiology advances Advances in our understanding of the molecular epidemiology of Blastocystis spp. have revealed subtype variation to occur in different countries and even different counties. Currently 17 different ST’s are described and ST1-9 shown to occur in humans (Stensvold et al. 2012). Overall ST3 remains the dominant human ST of Blastocystis with only Thailand reporting ST1 as the dominant ST (Tan 2008). The other common ST’s 1, 2, 4 and 6 show large regional variation in prevalence. ST4 has generated a lot of interest as it is almost absent from Asia, Middle Eastern countries and South America (Alfellani et al. 2013). Multi locus sequence analysis (MLST) of the mitochondrial genome of ST4 organisms has shown organisms to be almost clonal with much less genetic variation compared the other subtypes suggesting ST4 is a relatively recent mutation of the parasite (Clark et al. 2013).

8.3.3 Zoonotic potential The zoonotic potential of the parasite has also been explored and with the exception of close personal contact with exotic zoo animals (Parkar et al. 2010), non- human primates (Stensvold et al. 2009a) and pigs (Wang et al. 2014c) there is little overlap with the ST’s found in most animals and birds and those reported in human infections (Stensvold et al. 2009a). In particular, a recent study has shown a very low prevalence of Blastocystis in dogs from Australian pounds and Cambodian rural villages (2.3-1.3%), although a higher prevalence was reported in stray city dogs from India (24%) (Wang et al. 2013) ST’s observed in dogs varied widely.These findings suggest that dogs, at least in Australia, are not an important reservoir for infection for humans. Rodents do commonly carry ST4 and these animals are a potential reservoir of Blastocystis infection, although ST4 is a minor contributor to Blastocystis carriage in humans and is not present at all in many countries (Clark et al. 2013).

8.3.4 Subtypes and pathogenicity There is currently no consensus in the literature regarding specific subtypes and their pathogenic potential. ST1, 3 and 4 have variously been reported to have a higher prevalence in symptomatic patients (Tan et al. 2008; Eroglu et al. 2009; Poirier et al. 2011) while other studies have reported no difference (Dogruman-Al et al. 2009; Elwakil & Talaat 2009).

185

8.4 Subtyping and transmission within households 8.4.1 Subtyping for diagnosis of infection (Chapters 2, 5, 6, 7) We performed four clinical studies in Blastocystis-positive patients with IBS. In each study three methods of diagnosis of infection were used, namely light microscopy identification in wet faecal smears, culture in Jones medium for 48 hours and sub-typing using PCR techniques that identified the 18S rRNA gene. This allowed the most accurate identification of Blastocystis infection.

8.4.2 Transmission of infection within households (Chapter 2) In our first study eleven Blastocystis-positive patients with irritable bowel syndrome, seventeen human and eight animal household contacts were monitored for Blastocystis carriage before antimicrobial therapy was administered to the symptomatic patients. All human household contacts were positive for Blastocystis carriage with 94% demonstrating similarity with the ST found in the symptomatic patient. All animal household contacts were positive for Blastocystis carriage with 88% demonstrating similarity with the ST found in the symptomatic patient at the 18S rRNA gene. Blastocystis organisms were seen in fewer numbers in the stool of animals and the organism was difficult to grow from these specimens. Despite this similarity, due to the large genetic variability found within STs on more variable loci, it is not possible to be sure that these organisms were identical on the 18S rRNA gene, but it seems likely. These findings suggest that transmission within households is common and that infection is easily spread. In conjunction with the later finding of low prevalence in dogs in the community pounds (Wang et al. 2013), but a high prevalence in our household pets, I conclude that dogs are not natural hosts and have likely acquired the infection from their human owners or less likely from a common environmental source, not present in Queensland dog pounds.

8.4.3 Subtypes and pathogenicity (Chapters 2,5,6,7) Two of our clinical studies looked at symptomatic Blastocystis-positive patients only and the later two studies included positive healthy controls as well. Blastocystis sub-typing was performed in all these subjects. The largest cohort was

186 documented in Chapter 7 and this included forty symptomatic IBS patients and fifty- seven healthy positive controls. The ST’s seen in decreasing prevalence in the IBS patients were ST3 (31%), ST4 (27%), ST1 (19%) and others (ST’s 2,5,6,7,8) (23%). The ST’s seen in decreasing prevalence in the healthy controls were ST3 (29%), ST1 (29%), ST4 (14%) and others (ST’s 2, 5, 6, 7, 8) (29%). Although there was a trend for ST4 to be more prevalent in our IBS patients compared to healthy controls this difference never reached statistical significance. My findings do not support an association of specific subtypes with pathogenicity.

8.5 Antimicrobial therapy for Blastocystis infection 8.5.1 Eradication is problematic Multiple different antimicrobials have been reported to be used as eradication therapy for Blastocystis infection with widely varying reports of success (Stensvold et al. 2010). The most notable of these is metronidazole, used most commonly in clinical practice and currently recommended as first line therapy by the Centre for Disease Control of USA. It is reported to have clearance rates ranging from 3-100% (Stensvold et al. 2010). Although in vitro sensitivity to metronidazole is high (Zierdt et al. 1983; Dunn & Boreham 1991), resistance has been reported (Dunn et al. 2012), and there are also problems delivering the active drug to the site of interest (likely the ileum) and understanding the length of the course of antimicrobial therapy needed (predicated upon knowledge of lifecycle parameters that are unknown at present).

8.5.2 Why bother? If Blastocystis spp. are proven to cause symptoms then it will be useful to have an effective therapy. However, more to the point at this uncertain stage in Blastocystis research is, if Blastocystis infection could be reliably eradicated this would allow large clinical trials to be performed in symptomatic patients, who could be closely monitored for clinical outcomes. In that context it is important to remember that Giardia duodenalis was identified in the 1700’s but not confirmed as a pathogen until reliable therapy with imidazoles cured both infection and symptoms (Gardner & Hill 2001).

187 8.5.3 Clinical pilot study: efficacy of monotherapy with metronidazole or trimethoprim/sulfamethoxazole in Blastocystis-positive IBS patients (Chapter 2) Eleven patients were treated with the most commonly recommended mono- therapies, namely metronidazole (nine patients with 400 mg three times daily) and trimethoprim/sulfamethoxazole (two patients intolerant of metronidazole, 160/800mg twice daily). These patients were treated for 14 days and faecal samples taken at baseline, within 24 hours of completion of therapy and later at 28 and 56 days post- therapy. No patient cleared the Blastocystis; there was a 100% failure to eradicate at the time of completion of the antimicrobial course. The later samples had been taken to check on the possibility of re-infection following primary clearance from environmental / person or animal sources, but proved unnecessary. We conclude that monotherapy with metronidazole for Blastocystis infection should not be recommended as first line therapy and a later study has also concurred with this statement (Roberts et al. 2014).

8.5.4 Clinical pilot study: efficacy of triple antibiotic therapy in Blastocystis positive IBS patients (Chapter 5) Some organisms are difficult to clear with monotherapy, and Helicobacter pylori and Mycobacterium sp. are well known examples of infections that are managed with triple antimicrobial therapy. We decided to explore this approach and gave 10 Blastocystis-positive patients with diarrhoea-predominant IBS diloxanide furoate 500 mg three times daily, trimethoprim/sulfamethoxazole 160/80mg twice daily and secnidazole 400 mg three times daily for 14 days. These particular drugs were chosen based on the following considerations. Secnidazole, an imidazole similar to metronidazole, demonstrated moderately high antimicrobial inhibitory action in in vitro testing (Dunn & Boreham 1991), no documented intrinsic antimicrobial resistance (Mirza et al. 2010) and double the serum half-life of metronidazole with likely more excretion through the gut (Gillis & Wiseman 1996). Trimethoprim/sulfamethoxazole has previously been reported to have eradication rates ranging from 22-100% (Stensvold et al. 2010). This antimicrobial combination works synergistically to disrupt production of the enzyme dihydrofolate reductase required for nuclear purine synthesis and is primarly excreted in the urine. Blastocystis organisms are likely to synthesize DNA, and require this enzyme (Wu et al. 2010). Trimethoprim, but not sulfamethoxazole, has

188 been shown to have moderate antimicrobial inhibitory action in vitro (Dunn & Boreham 1991). However, the combination therapy has been shown to have additive inhibitory effects in vitro (Mirza et al. 2010). Diloxanide furoate showed no antimicrobial inhibitory action in vitro (Zierdt et al. 1983), but it was chosen because it is an effective amoebocidal drug that remains intraluminal within the gut and has a high safety profile (Gadkariem et al. 2004). Faecal specimens were obtained at baseline, within 24 hours and 4 weeks of ceasing antibiotics. Six of ten patients (60%) cleared Blastocystis at the end of the therapy (with no recurrence at 4 weeks). This was a small pilot study and the main aim was to assess antibiotic clearance rates. However, analysis of this small cohort showed higher clearance rates in patients with symptoms for less than one year. No association with clearance was found with age, previous antibiotic therapy, faecal parasite load, Blastocystis subtype, serum IgA levels or clinical improvement. The lack of association of clearance with clinical improvement is not necessarily a deterrent to the argument that Blastocystis sp. may cause symptoms. It may be that reducing the load of Blastocystis organisms, but without complete clearance, resolves symptoms or that the IBS patients who did not improve with clearance had IBS unrelated to Blastocystis infection.

8.6 Host responses to Blastocystis infection 8.6.1 Blastocystis and the host immune system Blastocystis carriage is reported to be higher in immuno- compromised people, including immuno-compromised children (Noureldin et al. 1999; Idris et al. 2010), Human immunodeficiency virus (HIV) infections (Stark et al. 2007), patients with haematological malignancies (Tasova et al. 2000; Poirier et al. 2011) or those undergoing cancer chemotherapy (Chandramathi et al. 2012; Mirza & Tan 2012). However, although diarrhoea was common in these groups of patients Blastocystis carriage was not shown to be significantly higher in symptomatic patients (Tasova et al. 2000; Poirier et al. 2011). Immunoglobulin A (IgA) secreted into the gastrointestinal lumen is an important host defence. Blastocystis organisms and lysates from cell cultures have been shown to lyse IgA (Puthia et al. 2005). Additionally, although Blastocystis sp. are luminal organisms with only rare reports of opportunistic mucosal invasion,

189 serum antibodies to Blastocystis antigens in every immunoglobulin class have been found by many researchers (Zierdt & Nagy 1993; Garavelli et al. 1995; Mahmoud & Saleh 2003). Antigen interaction occurs between the gut lumen and the submucosal immune system and the nature of this interaction with secretory IgA, specific Blastocystis faecal and serum antibodies and cytokine T-cell secretion will modulate the host’s response to infection. Blastocystis organisms are reported to cause increased production of the pro-inflammatory cytokine interleukin (IL)-8 in an enterocyte cell culture system (Puthia et al. 2006). IBS patients are also reported to have increased prevalence of polymorphisms causing increased production of IL-8 and reduced production of the anti-inflammatory cytokine IL-10, but the IL-8 polymorphisms were only reported to be present in the Blastocystis-positive IBS patients (Olivo-Diaz et al. 2012).

8.6.2 Serum immunoglobulin A in patients with IBS versus healthy controls (Chapter 6) Forty-diarrhoea predominant IBS patients (26 patients positive for Blastocystis sp., 14 negative patients) and 40 HC (24 positive, 16 Blastocystis-negative) were enrolled in our recent study. Age, gender, ethnicity, season of blood test, ST of Blastocystis and serum IgA were recorded. Mean levels of serum IgA were 25-50% lower in Blastocystis carriers (p<0.001) but had no relationship to symptoms or the formerly mentioned variables. A lower IgA may allow the Blastocystis to establish in the gut lumen but does not appear to influence pathogenicity.

8.6.3 Blastocystis specific serum immunoglobulin responses in patients with IBS compared to HC (Chapter 6) In this same study (8.6.2) serum samples from the study groups were run against Western blots of Blastocystis (WR1, ST4) antigen. This Western blot analysis revealed serum IgG antibodies specific for Blastocystis proteins of 17, 27, 37, 50, 60-65, 75-90, 95-105, and 150 kDa molecular weight (MW). Reactivity to the 27, 50 and 75-95 kDa proteins were found significantly more frequently in the IBS group compared to the HC and correlation was greater with the Blastocystis-positive IBS patients compared to the negative IBS patients. Although the difference in reactivity between the groups was significant overall, these proteins were recognized

190 by the host immune response in 86-100% of IBS patients and 44-71% of HC. The reaction to these Blastocystis proteins in negative patients suggests either a non- specific cross reactivity with other antigens or prior exposure to Blastocystis organisms.

8.6.4 Identifying serum immunoglobulin responses to Blastocystis antigens (Chapter 6) Specific immunoglobulin responses to Blastocystis antigens in IBS patients have previously been reported at 30kDa (Abou Gamra et al. 2011), 50 and 118 kDa (Hegazy et al. 2008), and 60-100 kDa (Rajamanikam & Govind 2013) and these reports correlate with our study findings of proteins of interest at 27, 50 and 75-95 kDa in symptomatic patients. We used specific antibodies, namely a polyclonal Blastocystis ST3 antibody (Blastofluor), a monoclonal cytopathic antibody to surface legumain (a cysteine protease) of Blastocystis ST7 (MAb1D5), a polyclonal anti-human matrix metalloproteinase 9 (Anti-MMP-9) and a control negative monoclonal antibody MAb5, to probe the Blastocystis antigens present on the Western blot in an attempt to identify the proteins of interest. The Blastofluor antibody result showed a consistent Blastocystis antigen pattern in ST3 and ST4 organisms. The MAb1D5 antibody reacted with a double protein band around 100kDa and also at 27kDa. MAb1D5 has been shown to react mainly at 30, but also at 100 kDa in a previous study and protein sequencing at this level has identified sequences consistent with a cysteine protease (Wawrzyniak et al. 2012). We suspect the 27kDa reaction in our study is identical to the 30kDa protein found in other studies and that size differences can be accounted for by small differences in antigen preparation and Western blotting technique. Cysteine proteases are predicted in the Blastocystis genome, identified in Blastocystis lysates and in general are well recognised to be associated with mechanisms of pathogenicity in parasites (Denoeud et al. 2011). Our results suggest that the 100kDa protein is an intact enzyme and the 27kDa protein is a degraded fragment of a cysteine protease. The anti-MMP-9 antibody was selected for trial as metalloproteases are found on the surface of parasites, are a well recognised pathogenic mechanism for host attack, their presence is predicted by Blastocystis genome studies, and anti-MMP-9

191 reportedly reacts with proteins at 50kDa. In our study the anti-MMP-9 reacted with Blastocystis antigen at 27, 50 and approximately 80kDa. These results suggest that metalloproteases are produced by Blastocystis spp. A subsequent zymogram performed with and without a metalloprotease inhibitor (EDTA) showed abolition of reaction with a protein around 100kDa with the addition of the inhibitor. We propose that the 80 and 50kDa proteins are either intact pro- or active-enzymes and that the 27kDa protein is a degraded fragment of enzyme. The reactions at 27kDa may be due to more than one type of degraded enzyme fragment.

8.7 The faecal microbiome, IBS and Blastocystis carriage 8.7.1 Introduction to the faecal microbiome, IBS and Blastocystis The faecal microbiome (FM) comprises bacteria, viruses, archaea and eukaryotes, the cell count exceeding by 10 times that of its human host. The 3.3 million FM genes produce micronutrients and neurotransmitters, metabolise food, and facilitate immune priming and defence. Metagenomic analysis, using culture- independent gene sequencing techniques, has allowed insights into the FM that were impossible previously (Bonfrate et al. 2013). The FM is altered in various disease states and characteristic changes described in IBS are decreased diversity of organisms, temporal instability and changes in the dominant bacterial phyla with increases in the Firmicutes: Bacteroidetes ratio (Simren et al. 2013). Three different FM patterns have previously been described in IBS patients, with no correlation found with clinical categories of IBS (Jeffery et al. 2012a). IBS symptoms may have multiple causes and we hypothesized that if Blastocystis sp. are one specific cause of IBS it may be possible to identify differences in the FM between Blastocystis- positive and negative-IBS patients.

8.7.2 Differences in faecal microbiota profiles between Blastocystis-positive and negative IBS patients (Chapter 7) Forty patients with diarrhoea-predominant IBS (26 Blastocystis-positive and 14 Blastocystis–negative) and 57 healthy controls (HC) (42 Blastocystis-positive and 15 Blastocystis-negative) submitted faecal samples for high throughput culture- independent metagenomic analysis using 16S ribosomal RNA gene sequencing. Differences in abundance of bacteria in these IBS and HC groups were evaluated

192 from phylum to genus level. Significant changes in the two dominant phyla, namely a rise in Firmicutes and a statistical significant reduction in relative abundance of Bacteroidetes (with a threefold increase in the Firmicutes: Bacteoridetes ratio) were observed in IBS patients, regardless of Blastocystis infection status. Further subgroup analysis showed Blastocystis-positive IBS patients had significant reductions in abundance at the phylum level of Actinobacteria (p= 0.015) and Firmicutes (p=0.013), and significant increases in Bacteroidetes (p= 0.05), “Other” (unidentified bacteria) (p=0.004) and Proteobacteria (p=0.018) compared to Blastocystis-negative IBS patients. These changes were reflected in a lower average Firmicutes: Bacteroidetes ratio of 6.19 in Blastocystis-positive compared to 9.00 in Blastocystis-negative IBS patients. Minor differences at genus level were also seen between the IBS groups in Anaerostipes, Blautia and Alistepes spp. These significant differences in the FM between Blastocystis-positive and Blastocystis-negative diarrhoea-predominant IBS patients suggest that Blastocystis carriage either alters the FM directly, or changes in the FM allow Blastocystis to flourish. These differences in FM profile may define a separate clinical subtype of IBS.

8.8 Limitations of studies 8.8.1 Acquisition of patients for the clinical studies There are no reliable biomarkers for IBS. The strength of our clinical studies was that the IBS patients were selected rigorously and had full clinical screening by one person to exclude other causes of symptoms. This allowed the clinical group to be as well-defined as possible. This advantage was also a disadvantage as it meant the numbers of patients in the clinical groups was small and it took many months to recruit patients.

8.8.2 Axenic culture failure (Chapter 3) Axenic cultures could have provided an abundance of Blastocystis antigen from different subtypes and increased quantities of antigen for our Western blot studies. Axenic cultures are problematic worldwide and we were not able to resolve this obstacle and I have provided suggestions for different approaches to axenic cultivation in Chapter 3. It may have been to our advantage that consequently only

193 one ST of Blastocystis was used (WR1) as that reduced the variables in the experiment. Large quantities of frozen purified Blastocystis antigen were lost in the electricity blackout in Gatton that occurred in the devastating 2011 floods.

8.9 A unifying hypothesis and future directions 8.9.1 A unifying hypothesis I propose that in most individuals with normal to high serum levels of IgA the

Blastocystis parasite is not able to establish a chronic infection. The sIgA, once translocated to the gut lumen, directly lyses the parasite. Additionally these IgA levels are associated with a faecal microbiome characterised by a greater prevalence of aerobic bacteria that is not a supportive environment for the strict anaerobe Blastocystis. In individuals with a low serum IgA the parasite is able to establish a chronic infection. Lower levels of sIgA mean that the parasite is not lysed immediately and

Blastocystis is able to survive in the gut lumen. Lower sIgA levels allow the parasite to position itself in the mucosa associated microbiota layer, adjacent to the gastrointestinal mucosal surface. Although eukaryotes, such as Blastocystis, comprise less than 1% of the faecal microbiome any organism that has closer apposition to the mucosal surface will have a proportionately greater opportunity to interact physically and immunologically with the host. However, even though some individuals are more susceptible to Blastocystis infection only a minority of these people will develop symptomatic disease. We know that Blastocystis spp. produce many proteases that are highly immunogenic. I have shown in this thesis that antibodies to cysteine proteases, and I would propose metalloproteases, produced by this enteric organism can be found in the serum of healthy and symptomatic patients, suggesting that these are potent antigens. I believe the host response to these antigens determines clinical outcome. Host responses are likely to vary between individuals. One possible immune response that has already been described relates to changes in gut mucosal cytokine production. Individuals with genetically determined cytokine polymorphisms causing low levels of protective anti- inflammatory IL-10 and high levels of pro- inflammatory IL-6 are more likely to develop IBS. Only individuals with these cytokine changes and in addition high levels of pro-inflammatory IL-8 have IBS and carry

194 Blastocystis. These cytokine polymorphisms will augment the host inflammatory response allowing the parasite to damage the host epithelial cells via activated cysteine and metallo-proteinases. This group of Blastocystis-positive IBS patients can be distinguished from other IBS patients by changes in the faecal microbiome and, I propose, by the presence of higher titres of specific serum antibodies to the parasite proteases.

8.9.2 Future directions Our studies raise questions that could be explored further in many different ways and one obstacle that would be useful to overcome is the difficulty with axenic cultures. Chapter four describes granules exiting from an amoebic form (AF). It may be that Blastocystis organisms need to be primed in the gut environment to convert to this form and it would be of interest to examine AF’s and these granules with stains for genetic material. We need an antimicrobial regimen that achieves high, probably higher than 90%, eradication rates to allow accurate evaluation of clinical response in treated IBS patients. Two recent studies have reported high rates of clearance using oral paromomycin, a poorly absorbed aminoglycoside compound that intriguingly has not been shown to have any antimicrobial inhibitory action on Blastocystis sp. in vitro (Zierdt et al. 1983). The smaller study (Roberts et al. 2014) describes three of three patients given paromomycin as second line therapy who all cleared, and the larger study (van Hellemond et al. 2013) involves retrospective review of the hospital charts of 52 patients treated with paromomycin 500mg three times daily for 5-7 days and reports an eradication rate of 77%. The latter study used only microscopic examination of faecal smears for diagnosis of Blastocystis infection. Further trials using different combinations of antimicrobials are indicated, including consideration of topical antibiotic therapy administered via colonic irrigation. We have identified three Blastocystis antigens at 27, 50 and 75-100 kDa, recognized by the host immune response and which may be associated with pathogenicity. If large amounts of Blastocystis antigen could be produced, the proteins at these bands could be sequenced, identified, amplified using recombinant protein technology and ELISA kits could be developed to evaluate these reacting

195 proteins in larger numbers of patients with various gastrointestinal diseases, including IBS. These proteins may be biomarkers for a specific IBS subgroup. It would also be interesting to check cytokine polymorphisms in all patients in future clinical studies that would also combine faecal microbiome profiling and serum antibody testing. Future studies would benefit from a larger and more diverse sample size of subjects and would most likely need to be multicentred. It will be important to include patients with other gastrointestinal diseases into the studies to establish if findings are specific to Blastocystis carriage, IBS or common non-specific findings in patients with damaged intestinal mucosa.

196

197 Bibliography

Abdel-Hameed, DM & Hassanin, OM 2011, 'Proteaese activity of Blastocystis hominis subtype3 in symptomatic and asymptomatic patients', Parasitol Res, vol. 109, no. 2, pp. 321-7.

Abou Gamra, MM, Elwakil, HS, El Deeb, HK, Khalifa, KE & Abd Elhafiz, HE 2011, 'The potential use of 29 kDa protein as a marker of pathogenicity and diagnosis of symptomatic infections with Blastocystis hominis', Parasitology Research, vol. 108, no. 5, pp. 1139-46.

Adachi, JA & DuPont, HL 2006, 'Rifaximin: A novel Nonabsorbed Rifamycin for Gastrointestinal disorders', Clinical Infectious Diseases, vol. 2006, no. 42, pp. 541-7.

Aghamohammadi, A, Cheraghi, T, Gharagozlou, M, Movahedi, M, Rezaei, N, Yeganeh, M, Parvanch, N, Abolhassani, H, Pourpak, Z & Moin, M 2009, 'IgA deficiency: Correlation between Clinical and Immunologic Phenotypes', J Clin Immunol, vol. 29, pp. 130-6.

Aguirre de Carcer, D, Cuiv, PO, Wang, T, Kang, S, Worthley, D, Whitehall, V, Gordon, I, McSweeney, C, Leggett, B & Morrison, M 2011, 'Numerical ecology validates a biogeographical distribution and gender-based effect on mucosa- associated bacteria along the human colon', ISME J, vol. 5, no. 5, pp. 801-9.

Al-Binali, AM, Bello, CS, El-Shewy, K & Abdulla, SE 2006, 'The prevalence of parasites in commonly used leafy vegetables in South Western, Saudi Arabia', Saudi Med J, vol. 27, no. 5, pp. 613-6. al-Tawil, YS, Gilger, MA, Gopalakrishna, GS, Langston, C & Bommer, KE 1994, 'Invasive Blastocystis hominis infection in a child', Arch Pediatr Adolesc Med, vol. 148, no. 8, pp. 882-5.

Albenberg, LG & Wu, GD 2014, 'Diet and the Intestinal Microbiome: Associations, Functions and Implications for Health and Disease', Gastroenterology, vol. 146, no. 6, pp. 1564-72.

Alfellani, MA, Stensvold, CR, Vidal-Lapiedrac, A, Onuohad, ESU, Fagbenro- Beyiokud, AF & Clark, CG 2013, 'Variable geographic distribtion of Blastocystis subtypes and its potential implications', Acta Tropica, vol. 126, pp. 11-8.

Alfellanin, MA, Taner-Mulla, D, Jacob, AS, Imeede, CI, Yoshikawa, H, Stensvold, CR & Clark, CG 2013, 'Genetic diversity of Blastocystis in Livestock and Zoo animals', , vol. 164, no. July, pp. 497-509.

Amenta, M, Dalle Nogare, ER, Colomba, C, Prestileo, TS, Di Lorenzo, F, Fundaro, S, Colomba, A & Ferrieri, A 1999, 'Intestinal protozoa in HIV- infected patients: effect of rifaximin in and Blastocystis hominis infections', J Chemother, vol. 11, no. 5, pp. 391-5.

198

Andiran, N, Acikgoz, ZC, Turkay, S & Andiran, F 2006, 'Blastocystis hominis-- an emerging and imitating cause of acute abdomen in children', J Pediatr Surg, vol. 41, no. 8, pp. 1489-91.

Arisue, N, Hashimoto, T, Yoshikawa, H, Nakamura, Y, Nakamura, G, Nakamura, F, Yano, TA & Hasegawa, M 2002, 'Phylogenetic position of Blastocystis hominis and of stramenopiles inferred from multiple molecular sequence data', J Eukaryot Microbiol, vol. 49, no. 1, pp. 42-53.

Arrieta, MC, Stiemsma, LT, Amenyogbe, N, Brown, EM & Finlay, B 2014, 'The intestinal microbiome in early life: health and disease', Front Immunol, vol. 5, no. 427, pp. 1-18.

Arumugam, M, Raes, J, Pelletier, E, Le Paslier, D, Yamada, T, Mende, DR, Fernandes, GR, Tap, J, Bruls, T, Batto, JM, Bertalan, M, Borruel, N, Casellas, F, Fernandez, L, Gautier, L, Hansen, T, Hattori, M, Hayashi, T, Kleerebezem, M, Kurokawa, K, Leclerc, M, Levenez, F, Manichanh, C, Nielsen, HB, Nielsen, T, Pons, N, Poulain, J, Qin, J, Sicheritz-Ponten, T, Tims, S, Torrents, D, Ugarte, E, Zoetendal, EG, Wang, J, Guarner, F, Pedersen, O, de Vos, WM, Brunak, S, Dore, J, Meta, HITC, Antolin, M, Artiguenave, F, Blottiere, HM, Almeida, M, Brechot, C, Cara, C, Chervaux, C, Cultrone, A, Delorme, C, Denariaz, G, Dervyn, R, Foerstner, KU, Friss, C, van de Guchte, M, Guedon, E, Haimet, F, Huber, W, van Hylckama-Vlieg, J, Jamet, A, Juste, C, Kaci, G, Knol, J, Lakhdari, O, Layec, S, Le Roux, K, Maguin, E, Merieux, A, Melo Minardi, R, M'Rini, C, Muller, J, Oozeer, R, Parkhill, J, Renault, P, Rescigno, M, Sanchez, N, Sunagawa, S, Torrejon, A, Turner, K, Vandemeulebrouck, G, Varela, E, Winogradsky, Y, Zeller, G, Weissenbach, J, Ehrlich, SD & Bork, P 2011, 'Enterotypes of the human gut microbiome', Nature, vol. 473, no. 7346, pp. 174-80.

Bailar, JC, 3rd & Mosteller, F 1988, 'Guidelines for statistical reporting in articles for medical journals. Amplifications and explanations', Ann Intern Med, vol. 108, no. 2, pp. 266-73.

Banaticla, JE & Rivera, WL 2011, 'Detection and subtype identification of Blastocystis isolates from wastewater samples in the Philippines', J Water Health, vol. 9, no. 1, pp. 128-37.

Barret, H 1921, 'A method for the cultivation of Blastocystis', Ann Trop Med Parasitol, vol. 15, pp. 113-6.

Beatty, JK, Bhargava, A & Buret, AG 2014, 'Post-infectious irritable bowel syndrome: Mechanistic insights into chronic disturbances following enteric infection', World J Gastroenterol, vol. 20, no. 14, pp. 3976-85.

Belova, LM & Krylov, MV 1994, '[The specificity of blastocysts (Rhizopoda: Lobosea)]', Parazitologiia, vol. 28, no. 2, pp. 105-9.

199 Berry, D & Reinisch, W 2013, 'Intestinal microbiota: a source of novel biomarkers in inflammatory bowel diseases?', Best Pract Res Clin Gastroenterol, vol. 27, no. 1, pp. 47-58.

Boeck, WC & Drbohlav, J 1925, 'The cultivation of Entamoebae histolytica', American Journal of Hygiene, vol. 5, pp. 271-407.

Bohm-Gloning, B, Knobloch, J & Walderich, B 1997, 'Five subgroups of Blastocystis hominis from symptomatic and asymptomatic patients revealed by restriction site analysis of PCR-amplified 16S-like rDNA', Trop Med Int Health, vol. 2, no. 8, pp. 771-8.

Bokhari, AA, Solomon, T & Desai, SA 2008, 'Two distinct mechanisms of transport through the plasmodial surface anion channel', J Membr Biol, vol. 226, no. 1-3, pp. 27-34.

Bonfrate, L, Tack, J, Grattagliano, I, Cuomo, R & Portincasa, P 2013, 'Microbiota in health and irritable bowel syndrome syndrome: current knowledge, perspectives and therapeutic options', Scand J Gastroenterol, vol. 48, pp. 995-1009.

Boreham, PF & Stenzel, DJ 1993, 'Blastocystis in humans and animals: morphology, biology, and epizootiology', Adv Parasitol, vol. 32, pp. 1-70.

Borody, TJ 2012, 'Single Transcolonoscopic infusion of three anti-protozoal agents for difficult Blastocystis Hominis infections', UEGW, p. Poster.

Bruewer, M, Luegering, A, Kucharzik, T, Parkos, CA, Madara, JL, Hopkins, AM & Nusrat, A 2003, 'Proinflammatory cytokines disrupt epithelial barrier function by apoptosis-independent mechanisms', J Immunol, vol. 171, pp. 6164-72.

Burden, DJ, Anger, HS & Hammet, NC 1978, 'Blastocystis sp infections in pigs', Vet Micro, vol. 3, pp. 227-34.

Bye, W, Ishaq, N, Bolin, TD, Duncombe, VM & Riordan, SM 2014, 'Overgrowth of the indigenous gut microbiome and irritable bowel syndrome', World J Gastroenterol, vol. 20, no. 10, pp. 2449-55.

Canavan, C, West, J & Card, T 2014a, 'The epidemiology of irritable bowel syndrome', Clin Epidemiol, vol. 6, pp. 71-80.

Canavan, C, West, J & Card, T 2014b, 'Review article: the economic impact of the irritable bowel syndrome', Aliment Pharmacol Ther, vol. 40, no. 9, pp. 1023-34.

Caporaso, JG, Kuczynski, J, Stombaugh, J, Bittinger, K, Bushman, FD, Costello, EK, Fierer, N, Pena, AG, Goodrich, JK, Gordon, JI, Huttley, GA, Kelley, ST, Knights, D, Koenig, JE, Ley, RE, Lozupone, CA, McDonald, D, Muegge, BD, Pirrung, M, Reeder, J, Sevinsky, JR, Turnbaugh, PJ, Walters,

200 WA, Widmann, J, Yatsunenko, T, Zaneveld, J & Knight, R 2010, 'QIIME allows analysis of high-throughput community sequencing data', Nat Methods, vol. 7, no. 5, pp. 335-6.

Carroll, IM, Ringel-Kulka, T, Siddle, JP & Ringel, Y 2012, 'Alterations in composition and diversity of the intestinal microbiota in patients with diarrhea- predominant irritable bowel syndrome', Neurogastroenterol Motil, vol. 24, no. 6, pp. 521-30, e248.

Cassidy, MF, Stenzel, DJ & Boreham, PF 1994, 'Electron microscopy of surface structures of Blastocystis sp. from different hosts', Parasitol Res, vol. 80, no. 6, pp. 505-11.

Cekin, AH, Cekin, Y, Adakan, Y, Tasdemir, E, Koclar, FG & Yolcular, BA 2012, 'Blastocystosis in patients with gastrointestinal symptoms: a case– control study', BMC Gastro, vol. 12, p. 122.

Chadwick, VS, Chen, W, Shu, D, Paulus, B, Bethwaite, P, Tie, A & Wilson, I 2002, 'Activation of the mucosal immune system in irritable bowel syndrome', Gastroenterology, vol. 122, no. 7, pp. 1778-83.

Chan, KH, Chandramathi, S, Suresh, K, Chua, KH & Kuppusamy, UR 2012, 'Effects of symptomatic and asymptomatic isolates of Blastocystis hominis on colorectal cancer cell line, HCT116', Parasitology Research, vol. 110, no. 6, pp. 2475-80.

Chandramathi, S, Suresh, K, Anita, ZB & Kuppusamy, UR 2012, 'Infections of Blastocystis hominis and microsporidia in cancer patients: are they opportunistic?', Trans R Soc Trop Med Hyg, vol. 106, no. 4, pp. 267-9.

Chandramathi, S, Suresh, K & Kuppusamy, UR 2010, 'Solubilized antigen of Blastocystis hominis facilitates the growth of human colorectal cancer cells, HCT116', Parasitol Res, vol. 106, no. 4, pp. 941-5.

Chapman, S, Thompson, C, Wilcox, A & Russell, K 2009, 'What is your diagnosis? Rectal scraping from a dog with diarrhea', Veterinary Clinical Pathology, vol. 38, no. 1, pp. 59-62.

Chaudhary, NA & Truelove, SC 1962, 'The irritable bowel syndome: A study of the clinical features, predisposing causes, and prognosis in 130 cases', QJ Med, vol. 31, pp. 307-22.

Chen, TL, Chan, CC, Chen, HP, Fung, CP, Lin, CP, Chan, WL & Liu, CY 2003, 'Clinical characteristics and endoscopic findings associated with Blastocystis hominis in healthy adults', Am J Trop Med Hyg, vol. 69, no. 2, pp. 213-6.

Chen, XQ, Singh, M, Ho, LC, Moe, KT, Tan, SW & Yap, EH 1997, 'A survey of Blastocystis sp. in rodents', Lab Anim Sci, vol. 47, no. 1, pp. 91-4.

201 Chen, XQ, Singh, M, Ho, LC, Tan, SW & Yap, EH 1999, 'Characterization of protein profiles and cross-reactivity of Blastocystis antigens by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblot analysis', Parasitol Res, vol. 85, no. 4, pp. 343-6.

Chen XQ, SM, Ho LC, Moe KT, Tan SW, Yap EH 1997, 'A survey of Blastocystis sp. in rodents', Lab Animal Sci, vol. 47, pp. 91-4.

Chey, WD, Kurlander, J & Eswaran, S 2015, 'Irritable bowel syndrome: a clinical review', JAMA, vol. 313, no. 9, pp. 949-58.

Ciferri, R & Redaelli, P 1938, 'A new hypothesis on the nature of Blastocystis', Mycopathologica, vol. 1, pp. 3-6.

Cirioni, O, Giacometti, A, Drenaggi, D, Ancarani, F & Scalise, G 1999, 'Prevalence and clinical relevance of Blastocystis hominis in diverse patient cohorts', European Journal of Epidemiology, vol. 15, pp. 389-93.

Clark, CG 1997a, 'Extensive genetic diversity in Blastocystis hominis', Mol Biochem Parasitol, vol. 87, no. 1, pp. 79-83.

Clark, CG 1997b, 'Riboprinting: a tool for the study of genetic diversity in microorganisms', J Eukaryot Microbiol, vol. 44, no. 4, pp. 277-83.

Clark, CG & Diamond, LS 2002, 'Methods for Cultivation of Luminal Parasitic of Clinical Importance', Clinical Microbiology Reviews, vol. 15, no. 3, pp. 329-41.

Clark, CG, van der Giezen, M, Alfellani, MA & Stensvold, CR 2013, 'Recent Developments in Blastocystis research', Adv Parasitol, vol. 82, pp. 1-32.

Cock, PJ, Fields, CJ, Goto, N, Heuer, ML & Rice, PM 2010, 'The Sanger FASTQ file format for sequences with quality scores, and the Solexa/Illumina FASTQ variants', Nucleic Acids Res, vol. 38, no. 6, pp. 1767-71.

Codling, C, O'Mahony, L, Shanahan, F, Quigley, EM & Marchesi, JR 2010, 'A molecular analysis of fecal and mucosal bacterial communities in irritable bowel syndrome', Dig Dis Sci, vol. 55, no. 2, pp. 392-7.

Collins, S 2014, 'A role for the gut microbiota in IBS', Nat Rev Gastroenterol Hepatol, vol. 11, no. 8, pp. 497-505.

Corsetti, M, Van Oudenhove, L & Tack, J 2014, 'The quest for biomarkers in IBS-where should it lead us?', Neurogastroenterology & Motility, vol. 26, no. 12, pp. 1669-76.

Coyle, CM, Varughese, J, Weiss, LM & Tanowitz, HB 2012, 'Blastocystis: to treat or not to treat', Clin Prac, vol. 54, pp. 105-7.

202 Davidson, RN, den Boer, M & Ritmeijer, K 2009, 'Paromomycin', Trans R Soc Trop Med Hyg, vol. 103, no. 7, pp. 653-60.

De Filippo, C, Cavalieri, D, Di Paola, M, Ramazzotti, M, Poullet, JB, Massart, S, Collini, S, Pieraccini, G & Lionetti, P 2010, 'Impact of diet in shaping gut microbiota revealed by a comparative study in children from Europe and rural Africa', Proc Natl Acad Sci U S A, vol. 107, no. 33, pp. 14691–6. de Salonen, A, Vos, WM & Palva, A 2010, 'Gastointestinal microbiota in irritable bowel syndrome: present state and perspectives', Microbiology, vol. 156, pp. 3205-15.

Denoeud, F, Roussel, M, Noel, B, Wawrzyniak, I, Da Silva, C, Diogon, M, Viscogliosi, E, Brochier-Armanet, C, Couloux, A, Poulain, J, Segurens, B, Anthouard, V, Texier, C, Blot, N, Poirier, P, Ng, GC, Tan, KSW, Artiguenave, F, Jaillon, O, Aury, J, Delbac, F, Wincker, P, Vivarès, CP & El Alaoui, H 2011, 'Genome sequence of the stramenopile Blastocystis, a human anaerobic parasite', Genome Biology, vol. 12, no. 3, p. R29.

Dinleyici, EC, Eren M, Dogan N, Reyhanioglu S, yargic ZA, Vandenpla Y 2011, 'Clinical efficacy of Saccharomyces boulardii or metronidazole in symptoamtic children with Blastocystis hominis infection', Parasitol Res, vol. 108, pp. 541-5.

Dogruman-Al, F, Simsek, Z, Boorom, K, Ekici, E, Sahin, M, Tuncer, C, Kustimur, S & Altinbas, A 2010, 'Comparison of methods for detection of Blastocystis infection in routinely submitted stool samples, and also in IBS/IBD Patients in Ankara, Turkey', PLoS One, vol. 5, no. 11, p. e15484.

Dogruman-Al, F, Yoshikawa, H, Kustimur, S & Balaban, N 2009, 'PCR-based subtyping of Blastocystis isolates from symptomatic and asymptomatic individuals in a major hospital in Ankara, Turkey', Parasitol Res, vol. 106, pp. 263-8.

Doyle, P, Helgason, M, Mathias, R & Proctor, E 1990, 'Epidemiology and Pathogenicity of Blastocystis hominis', J Clin Microbiol, vol. 28, no. 1, pp. 116- 21.

Drossman, DA 2007, 'Introduction. The Rome Foundation and Rome III', Neurogastroenterol Motil, vol. 19, no. 10, pp. 783-6.

Duda, A, Stenzel, DJ & Boreham, PF 1998, 'Detection of Blastocystis sp in domestic dogs and cats', Veterinary Parasitology, vol. 76, pp. 9-17.

Dunn, LA 1992, 'Variation among cultured stocks of Blastocystis hominis (Brumpt 1912)', PhD thesis, The University of Queensland, Brisbane, Australia thesis, University of Queensland.

Dunn, LA & Boreham, PF 1991, 'The in-vitro activity of drugs against Blastocystis hominis', J Antimicrob Chemother, vol. 27, no. 4, pp. 507-16.

203

Dunn, LA, Boreham, PFL & Stenzel, DJ 1989, 'Ultrastructural variation of Blastocystis hominis stocks in culture', Int J Parasitol, no. 19, pp. 43-56.

Dunn, LA, Tan, KS, Vanelle, P, Juspin, T, Crozet, MD, Terme, T, Upcroft, P & Upcroft, JA 2012, 'Development of metronidazole-resistant lines of Blastocystis sp.', Parasitol Res, vol. 111, pp. 441-50.

DuPont, HL 2014, 'Review article: evidence for the role of gut microbiota in irritable bowel syndorme and its potential influence on therapeutic targets', Aliment Pharmacol Ther, vol. 39, pp. 1033-42.

El Safadi, D, Gaayeb, L, Meloni, D, Cian, A, Poirier, P, Wawryzyniak, I, Delbac, F, Dabboussi, F, Delhaes, L, Seck, M, Hamze, M, Riveau, G & Viscogliosi, E 2014, 'Children of the Senegal River basin show the highest prevalence of Blastocystis sp. ever observed worldwide', BMC Infectious Disease, vol. 14, p. 164.

El-Shazly, AM, Abdel-Magied, AA, El-Beshbishi, SE, El-Nahas, HA, Fouad, MAH & Monib, SM 2005a, 'Blastocystis Hominis among symptomatic and asymptomatic individuals in Talkha Centre Dakahlia Governate Egypt', J Egyptian Society of Parasitologists, vol. 35, no. 2, pp. 653-66.

El-Shazly, AM, Abdel-Magied, AA, El-Beshbishi, SN, El-Nahas, HA, Fouad, MA & Monib, MS 2005b, 'Blastocystis hominis among symptomatic and asymptomatic individuals in Talkha Center, Dakahlia Governorate, Egypt', J Egypt Soc Parasitol, vol. 35, no. 2, pp. 653-66.

Elwakil, HS & Hewedi, IH 2010, 'Pathogenic potential of Blastocystis hominis in laboratory mice', Parasitol Res, vol. 107, no. 3, pp. 685-9.

Elwakil, HS & Talaat, RM 2009, 'Genetic analysis of Blastocystis hominis isolated from symptomatic and asymptomatic human hosts in Egypt', J Egypt Soc Parasitol, vol. 39, no. 1, pp. 99-109.

Eroglu, F, Genc, A, Elgun, G & Koltas, IS 2009, 'Identification of Blastocystis hominis isolates from asymptomatic and symptomatic patients by PCR', Parasitol Res, vol. 105, no. 6, pp. 1589-92.

Eroglu, F & Koltas, IS 2010, 'Evaluation of the transmission mode of B. hominis by using PCR method', Parasitol Res, vol. 107, no. 4, pp. 841-5.

Fagarasan, S 2008, 'Evolution, development, mechanism and function of IgA in the gut', Curr Opin Immunol, vol. 20, no. 2, pp. 170-7.

Fredricks, D & Relman, DA 1996, 'Sequence-based Identification of Microbial pathogens: a reconsideration of Koch's postulates', Clin Microbiol Rev, pp. 18- 33.

204 Furusawa, Y, Obata, Y & Hase, K 2014, 'Commensal microbiota regulates T cell fate decision in the gut', Semin Immunopathol, vol. 37, no. 1, pp. 17-25.

Gadkariem, EA, Belal, F, Abounassif, MA, El-Obeid, HA & Ibrahim, KEE 2004, 'Stability studies on diloxanide furoate: effect of pH, temperature, gastric and intestinal fluids', Il Farmaco, vol. 59, pp. 323-9.

Gamra, MM, Elwakil, HS, El Deeb, HK, Khalifa, KE & Elhafiz, HE 2011, 'The potential use of 29 kDa protein as a marker of pathogenicity and diagnosis of symptomatic infections with Blastocystis hominis', Parasitol Res, vol. 108, pp. 1139-46.

Garavelli, PL, Zierdt, CH, Fleisher, TA, Liss, H & Nagy, B 1995, 'Serum antibody detected by fluorescent antibody test in patients with symptomatic Blastocystis hominis infection', Recenti Prog Med, vol. 86, no. 10, pp. 398- 400.

Gardner, TB & Hill, DR 2001, 'Treatment of giardiasis', Clin Microbiol Rev, vol. 14, no. 1, pp. 114-28.

Garduno, RA, Garduno, E, Hiltz, M & Hoffman, PS 2002, 'Intracellular Growth of Legionella pneumophila Gives Rise to a Differentiated Form Dissimilar to Stationary-Phase Forms', Infection and Immunity, vol. 70, no. 11, pp. 6273- 83.

Gecse, K, Roka, R, Ferrier, L, Leveque, M, Eutamene, H, Cartier, C, Ait- Belgnaoui, A, Rosztoczy, A, Izbeki, F, Fioramonti, J, Wittmann, T & Bueno, L 2008, 'Increased faecal serine protease activity in diarrhoeic IBS patients: a colonic luminal factor impairing colonic permeability and sensitivity', Gut, vol. 57, no. 5, pp. 591-9.

Geurts, N, Opdenakker, G & Van den Steen, PE 2012, 'Matrix metalloproteinases as therapeutic targets in protozoan parasitic infections', Pharmacol Ther, vol. 133, no. 3, pp. 257-79.

Giacometti, A, Cirioni, O, Fiorentini, A, Fortuna, M & Scalise, G 1999, 'Irritable bowel syndrome in patients with Blastocystis hominis infection', Eur J Clin Microbiol Infect Dis, vol. 18, no. 6, pp. 436-9.

Gillis, JC & Wiseman, LR 1996, 'Secnidazole. A review of its antimicrobial activity, pharmacokinetic properties and therapeutic use in the management of protozoal infections and bacterial vaginosis', Drugs, vol. 51, no. 4, pp. 621- 38.

Gouba, N, Raoult, D & Drancourt, M 2013, 'Plant and fungal diversity in gut microbiota as revealed by molecular and culture investigations', PLoS One, vol. 8, no. 3, p. e59474.

205 Gould, R & Boorom, K 2013, 'Blastocystis surface antigen is stable in chemically preserved stool samples for at least 1 year', Parasitol Res, vol. 112, pp. 2469-71.

Graczyk, TK, Shiff, CK, Tamang, L, Munsaka, F, Beitin, AM & Moss, WJ 2005, 'The association of Blastocystis hominis and Endolimax nana with diarrheal stools in Zambian school-age children', Parasitol Res, vol. 98, no. 1, pp. 38-43.

Grenham, S, Clarke, G, Cryan, JF & Dinan, TG 2011, 'Brain-gut-microbe communication in health and disease', Front Physiol, vol. 2, no. 94, pp. 1-15.

Grundmann, O & Yoon, SL 2010, 'Irritable bowel syndrome: epidemiology, diagnosis and treatment: an update for health-care practitioners', J Gastroenterol Hepatol, vol. 25, no. 4, pp. 691-9.

Gupta, R & Parsi, K 2006, 'Chronic urticaria due to Blastocystis hominis', Australas J Dermatol, vol. 47, no. 2, pp. 117-9.

Hegazy, MM, Maklouf, LM, El Hamshary, EM, Dawoud, HA & Eida, AM 2008, 'Protein profile and morphometry of cultured human Blastocystis hominis from children with gastroenteritis and healthy ones', J Egypt Soc Parasitol, vol. 38, no. 2, pp. 453-64.

Hellard, ME, Sinclair, MI, Hogg, GG & Fairley, CK 2000, 'Prevalence of enteric pathogens among community based asymptomatic individuals', J Gastroenterol Hepatol, vol. 15, no. 3, pp. 290-3.

Ho, LC, Singh, M, Suresh, G, Ng, GC & Yap, EH 1993, 'Axenic culture of Blastocystis hominis in Iscove's modified Dulbecco's medium', Parasitol Res, vol. 79, no. 7, pp. 614-6.

Hong, SN & Rhee, PL 2014, 'Unraveling the ties between irritable bowel syndrome and intestinal microbiota', World J Gastroenterol, vol. 20, no. 10, pp. 2470-81.

Horiki, N, Kaneda, Y, Maruyama, M, Fujita, Y & Tachibana, H 1999, 'Intestinal blockage by carcinoma and Blastocystis hominis infection', Am J Trop Med Hyg, vol. 60, no. 3, pp. 400-2.

Hughes, PA, Zola, H, Penttila, IA, Blackshaw, LA, Andrews, JM & Krumbiegel, D 2013, 'Immune activation in irritable bowel syndrome: can neuroimmune interactions explain symptoms?', Am J Gastroenterol, vol. 108, no. 7, pp. 1066-74.

Hussain, R, Jaferi, W, Zuberi, S, Baqai, R, Abrar, N, Ahmed, A & Zaman, V 1997, 'Significantly increased IgG2 subclass antibody levels to Blastocystis hominis in patients with irritable bowel syndrome', Am J Trop Med Hyg, vol. 56, no. 3, pp. 301-6.

206 Hussein, EM, Hussein, AM, Eida, MM & Atwa, MM 2008, 'Pathophysiological variability of different genotypes of human Blastocystis hominis Egyptian isolates in experimentally infected rats', Parasitol Res, vol. 102, no. 5, pp. 853-60.

IBM SPPS, S 2013, 'IBM SPPS Statistics for Windows, Version 22.0'.

Idris, NS, Dwipoerwantoro, PG, Kurniawan, A & Said, M 2010, 'Intestinal parasitic infection of immunocompromised children with diarrhoea: clinical profile and therapeutic response', J Infect Dev Ctries, vol. 4, no. 5, pp. 309-17.

Iguchi, A, Ebisu, A, Nagata, S, Saitou, Y, Yoshikawa, H, Iwatani, S & Kimata, I 2007, 'Infectivity of different genotypes of human Blastocystis hominis isolates in chickens and rats', Parasitology International, vol. 56, no. 2, pp. 107-12.

Inadomi, JM, Fennerty, MB & Bjorkman, D 2003, 'The Economic Impact of Irritable Bowel Syndrome', Alimentary Pharmacology and Therapeutics, vol. 18, no. 7, pp. 671-82.

Init, I, Mak, JW, Top, S, Zulhainan, Z, Prummongkol, S, Nissapatorn, V, Wan- Yusoff, WS & Anuar, AK 2003, 'Polypeptides associated with in vitro cyst formation of Blastocystis hominis', Southeast Asian J Trop Med Public Health, vol. 34, no. 4, pp. 727-32.

Jeffery, IB, Claesson, MJ, O'Toole, PW & Shanahan, F 2012a, 'Categorization of the gut microbiota: enterotypes or gradients?', Nature Reviews Microbiology, vol. 10, no. 9, pp. 591-2.

Jeffery, IB, O'Toole, PW, Ohman, L, Claesson, MJ, Deane, J, Quigley, EM & Simren, M 2012b, 'An irritable bowel syndrome subtype defined by species- specific alterations in faecal microbiota', Gut, vol. 61, no. 7, pp. 997-1006.

Jelinek, T, Peyerl, G, Loscher, T, von Sonnenburg, F & Nothdurft, HD 1997, 'The role of Blastocystis hominis as a possible intestinal pathogen in travellers', J Infect, vol. 35, no. 1, pp. 63-6.

Jimenez-Gonzalez, DE, Martinez-Flores, WA, Reues-Gordillo, J, Ramirez- Miranda, ME, Arroyo-Escalante, S, Romero-Valdovinos, M, Stark, D, Souza- Saldivar, V, Martinez-Flores, F, Flisser, A, Olivo-Diaz, A & Maravilla, P 2012, 'Blastocystis infection is associated with irritable bowel syndrome in a Mexican patient population', Parasitol Res, vol. 110, no. 3, pp. 1269-75.

Jones, DR 1946, 'A medium for investigating the breakdown of pectin by bacteria', Nature, vol. 158, no. 4018, p. 625.

Jones, MP, Chey, WD, Singh, S, Gong, H, Shringarpure, R, Hoe, N, Chuang, E & Talley, NJ 2014, 'A biomarker panel and psychological morbidity differentiates the irritable bowel syndrome from health and provides novel pathophysiological leads', Aliment Pharmacol Ther, vol. 39, no. 4, pp. 426-37.

207 Kain, KC, Noble, MA, Freeman, HJ & Barteluk, RL 1987, 'Epidemiology and clinical features associated with Blastocystis hominis infection', Diagn Microbiol Infect Dis, vol. 8, no. 4, pp. 235-44.

Kaneda, Y, Horiki, N, Cheng, X, Tachibana, H & Tsutsumi, Y 2000, 'Serologic response to Blastocystis hominis infection in asymptomatic individuals', Tokai J Exp Clin Med, vol. 25, no. 2, pp. 51-6.

Kaneda, Y, Horiki, N, Cheng, XJ, Fujita, Y, Maruyama, M & Tachibana, H 2001, 'Ribodemes of Blastocystis hominis isolated in Japan', Am J Trop Med Hyg, vol. 65, no. 4, pp. 393-6.

Karanis, P, Kourenti, C & Smith, H 2007, 'Waterborne transmission of protozoan parasites: a worldwide review of outbreaks and lessons learnt', J Water Health, vol. 5, no. 1, pp. 1-38.

Kennedy, PJ, Cryan, JF, Dinan, TG & Clarke, G 2014, 'Irritable bowel syndrome: A microbiome-gut-brain axis disorder?', World J Gastroenterol, vol. 20, no. 39, pp. 14105-25.

Klimes, V, Gentekaki, E, Roger, AJ & Elias, M 2014, 'A large number of nuclear genes in the human parasite blastocystis require mRNA polyadenylation to create functional termination codons', Genome Biol Evol, vol. 6, no. 8, pp. 1956-61.

Knoll, AH 1994, 'Proterozoic and early Cambrian protists: evidence for accelerating evolutionary tempo', Proc Natl Acad Sci U S A, vol. 91, no. 15, pp. 6743-50.

Knowles, R & Das Gupta, BM 1924, 'On the nature of Blastocystis hominis', Indian Journal of Medical Research, vol. 12, pp. 31-8.

Krogsgaard, LR, Engsbro, AL, Stensvold, CR, Nielsen, HV & Bytzer, P 2014, 'The Prevalence of Intestinal Parasites Is Not Greater Among Individuals With Irritable Bowel Syndrome: a Population-Based Case-Control Study', Clin Gastroenterol Hepatol.

Kukoschke, K & Muller, H 1991, 'SDS-PAGE and immunological analysis of different axenic Blastocystis hominis strains', J Med Microbiol, vol. 35, pp. 35- 9.

Kumar, S, Tamura, K, Jakobsen, IB & Nei, M 2001, 'MEGA2:Molecular Evolutionary Genetics Analysis Software.', Tempe (AZ).

Kumarasamy V, Kuppusamy, UR, Samudi, C & Kumar, S 2013, 'Blastocystis species subtype 3 triggers higher proliferation of human colorectal cancer cells, HCT116', Parasitol Res, vol. 2013, no. 112, pp. 3551-5.

Kuo, HY, Chiang, DH, Wang, CC, Chen, TL, Fung, CP, Lin, CP, Cho, WL & Liu, CY 2008, 'Clinical significance of Blastocystis hominis: experience from a

208 medical center in northern Taiwan', J Microbiol Immunol Infect, vol. 41, no. 3, pp. 222-6.

Lamp, KC, Freeman, CD, Klutman, NE & Lacy, MK 1999, ' and Pharmacodynamics of the Nitroimidazole Antimicrobials', Clin Pharmacokinet, vol. 36, no. 5, pp. 353-73.

Lanuza, MD, Carbajai, JA, Villar, J & Borras, R 1997, 'Description of an improved method for Blastocystis hominis culture and axenisation', Parasitol Res, vol. 83, pp. 60-3.

Lanuza, MD, Carbajal, JA, Villar, J, Mir, A & Borras, R 1999, 'Solube-protein and antigenic heterogeneity in axenic Blastocystis hominins isolates: pathogenic implications', Parasitol Res, vol. 85, pp. 93-7.

Laterza, L, Ianiro, G, Scoleri, I, Landi, R, Bruno, G, Scalaferri, F, Gaetani, E, Campanale, M & Gasbarrini, A 2015, 'Rifaximin for the treatment of diarrhoea- predominant irritable bowel syndrome', Exp Opin Pharmacother, vol. 16, no. 4, pp. 607- 15.

Latiff, AHA & Kerr, MA 2007, 'The clinical significance of immunoglobulin A deficiency', Ann Clin Biochem, vol. 44, pp. 131-9.

Lau, HY & Ashbolt, NJ 2009, 'The role of biofilms and protozoa in Legionella pathogenesis: implications for drinking water', Journal of Applied Microbiology, vol. 107, no. 2, pp. 368-78.

Leder, K, Hellard, ME, Sinclair, MI, Fairley, CK & Wolfe, R 2005, 'No correlation between clinical symptoms and Blastocystis hominis in immunocompetent individuals', J Gastroenterol Hepatol, vol. 20, no. 9, pp. 1390-4.

Lee, MG, Rawlins, SC, Didier, M & DeCeulaer, K 1990, 'Infective arthritis due to Blastocystis hominis', Ann Rheum Dis, vol. 49, no. 3, pp. 192-3.

Lee, MG & Stenzel, DJ 1999, 'A survey of Blastocystis in domestic chickens', Parasitol Res, vol. 85, no. 2, pp. 109-17.

Leelayoova, S, Rangsin, R, Taamasri, P, Naaglor, T, Thathaisong, U & Mungthin, M 2004, 'Evidence of waterborne transmission of Blastocystis hominis', Am J Trop Med Hyg, vol. 70, no. 6, pp. 658-62.

Lembo, AJ, Neri, B, Tolley, J, Barken, D, Carroll, S & Pan, H 2009, 'Use of serum biomarkers in a diagnostic test for irritable bowel syndrome', Aliment Pharmacol Ther, vol. 29, no. 8, pp. 834-42.

Li, LH, Zhang, XP, Lv, S, Zhang, L, Yoshikawa, H, Wu, Z, Steinmann, P, Utzinger, J, Tong, XM, Chen, SH & Zhou, XN 2007a, 'Cross-sectional surveys and subtype classification of human Blastocystis isolates from four epidemiological settings in China', Parasitol Res, vol. 102, no. 1, pp. 83-90.

209

Li, LH, Zhou, XN, Du, ZW, Wang, XZ, Wang, LB, Jiang, JY, Yoshikawa, H, Steinmann, P, Utzinger, J, Wu, Z, Chen, JX, Chen, SH & Zhang, L 2007b, 'Molecular epidemiology of human Blastocystis in a village in Yunnan province, China', Parasitol Int, vol. 56, no. 4, pp. 281-6.

Longstreth, GF, Thompson, WG, Chey, WD, Houghton, LA, Mearin, F & Spiller, RC 2006, 'Functional Bowel Disorders', Gastroenterology, vol. 130, no. 5, pp. 1480-91.

Lynch, KM 1922, 'Blastocystis species in culture. A preliminary communication.', American Journal of Tropical Medicine, vol. 2, pp. 215-22.

Lyra, A, Rinttilä, T, Nikkilä, J, Krogius-Kurikka, L, Kajander, K, Malinen, E, Mättö, J, Mäkelä, L & Palva, A 2009, 'Diarrhoea-predominant irritable bowel syndrome distinguishable by 16S rRNA gene phylotype quantification', World Journal of Gastroenterology, vol. 15, no. 47, pp. 5936-45.

Macpherson, AJ, McCoy, KD, Johansen, FE & Brandtzaeg, P 2008, 'The immune geography of IgA induction and function', Mucosal Immunol, vol. 1, no. 1, pp. 11-22.

Mahmoud, MS & Saleh, WA 2003, 'Secretory and humoral antibody responses to Blastocystis hominis in symptomatic and asymptomatic human infections', J Egypt Soc Parasitol, vol. 33, no. 1, pp. 13-30.

Malinen, E, Rinttila, T, Kajander, K, Matto, J, Kassinen, A, Krogius, L, Saarela, M, Korpela, R & Palva, A 2005, 'Analysis of the fecal microbiota of irritable bowel syndrome patients and healthy controls with real-time PCR', Am J Gastroenterol, vol. 100, no. 2, pp. 373-82.

Markell, EK & Udkow, MP 1986, 'Blastocystis hominis: pathogen or fellow traveler?', Am J Trop Med Hyg, vol. 35, no. 5, pp. 1023-6.

Matricon, J, Meleine, M, Gelot, A, Piche, T, Dapoigny, M, Muller, E & Ardid, D 2012, 'Review article:associations between immune activation, intestinal permeability and the irritable bowel syndrome', Aliment Pharmacol Ther, vol. 36, pp. 1009-31.

Matsumoto, Y, Yamada, M & Yoshida, Y 1987, 'Light-microsocpical appearance and ultrastructure of Blastocstys hominis an intestinal parasite of man', Zentralbl Bakteriol, vol. 264, pp. 379-85.

McClure, H, Strobert, E & GR., H 1980, 'Blastocystis hominis in a pig-tailed macaque: a potential enteric pathogen for nonhuman primates', Laboratory Animal Science, vol. 30, no. 5, pp. 890-4.

McDonald, D, Price, MN, Goodrich, J, Nawrocki, EP, DeSantis, TZ, Probst, A, Andersen, GL, Knight, R & Hugenholtz, P 2012, 'An improved Greengenes

210 with explicit ranks for ecological and evolutionary analyses of bacteria and archaea', ISME J, vol. 6, no. 3, pp. 610-8.

McKerrow, JH 1989, 'Parasite proteases', Exp Parasitol, vol. 68, pp. 111-5.

Mehlhorn, H 1988, 'Blastocystis hominis, Brumpt 1912: are there different stages or species?', Parasitol Res, vol. 74, no. 4, pp. 393-5.

Menardi, G & J.P., G 1984, ' of oral antibiotics in children with short bowel syndrome', J Paediatr Surg, vol. 19, no. 1, pp. 84-6.

Merga, Y, Campbell, BJ & Rhodes, JM 2014, 'Mucosal barrier, bacteria and inflammatory bowel disease: possibilities for therapy', Dig Dis, vol. 32, no. 4, pp. 475-83.

Mirza, H & Tan, K 2012, Clinical Aspects of Blastocsytis infections: Advancements Amidst Controversies, Blastocystis: Pathogen or Passenger, Springer, Berlin.

Mirza, H & Tan, KS 2009, 'Blastocystis exhibits inter- and intra-subtype variation in cysteine protease activity', Parasitol Res, vol. 104, no. 2, pp. 355- 61.

Mirza, H, Teo, JDW, Upcroft, J & Tan, KSW 2010, 'A Rapid, High-Throughput Viability Assay for Blastocystis spp. Reveals Metronidazole Resistance and Extensive Subtype-Dependent Variations in Drug Susceptibilities', Antimicrobial Agents and Chemotherapy, vol. 55, no. 2, pp. 637-48.

Moe, KT, Singh, M, Howe, J, Ho, LC, Tan, SW, Chen, XQ, Ng, GC & Yap, EH 1997, 'Experimental Blastocystis hominis infection in laboratory mice', Parasitol Res, vol. 83, no. 4, pp. 319-25.

Moe, KT, Singh, M, Howe, J, Ho, LC, Tan, SW, Ng, GC, Chen, XQ & Yap, EH 1996, 'Observations on the ultrastructure and viability of the cystic stage of Blastocystis hominis from human feces', Parasitol Res, vol. 82, no. 5, pp. 439- 44.

Moghaddam, DD, Ghadirian, E & Azami, M 2005, 'Blastocystis hominis and the evaluation of efficacy of metronidazole and trimethoprim/sulfamethoxazole', Parasitol Res, vol. 96, no. 4, pp. 273-5.

Molet, B, Werler, C & Kremer, M 1981, 'Blastocystis hominis: improved axenic cultivation', Transactions of the Royal Society of Tropical Medicine and Hygiene, vol. 75, pp. 752-3.

Mulak, A, Tache, Y & Larauche, M 2014, 'Sex hormones in the modulation of irritable bowel syndrome', World J Gastroenterol, vol. 20, no. 10, pp. 2433-48.

211 Nagel, R, Bielefeldt-Ohmann, H & Traub, RJ 2014, 'Clinical pilot study:efficacy of triple antibiotic therapy in Blastocystis positive irritable bowel syndrome patients', Gut Pathogens, vol. 6, no. 34, pp. 1-9.

Nagel, R, Cuttell, L, Stensvold, CR, Mills, PC, Bielefeldt-Ohmann, H & Traub, RJ 2012, 'Blastocystis subtypes in symptomatic and asymptomatic family members and pets and response to therapy', Intern Med J, vol. 42, no. 11, pp. 1187-95.

Nagel, R, Gray, C, Bielefeldt-Ohmann, H & Traub, RJ 2015a, 'Features of Blastocystis spp. in xenic culture revealed by deconvolutional microsocpy', Parasitol Res, vol. 114, pp. 3237-45.

Nagel, R, Traub, RJ, Kwan, M & Bielefeldt-Ohmann, H 2015b, 'Blastocystis specific serum immunoglobulin in patients with irritable bowel sydrome versus healthy controls', Parasites and Vectors, vol. 8, no. 453, pp. 1-13.

Nasirudenn, AM, Hian, YE, Singh, M, . & Tan, KS 2004, 'Metronidazole induces programmed cell death in the protozoan parasite Blastocystis hominis', Microbiology, vol. 150, pp. 33-43.

Ng, GC & Yan, KS 1999, 'Colony growth as a step towards axenization of Blastocystis isolates', Parasitol Res, vol. 85, pp. 678-9.

Ng, SC, Lam, EF, Lam, TT, Chan, Y, Law, W, Tse, PC, Kamm, MA, Sung, JJ, Chan, FK & Wu, JC 2013, 'Effect of probiotic bacteria on the intestinal microbiota in irritable bowel syndrome', J Gastroenterol Hepatol, vol. 28, no. 10, pp. 1624-31.

Nigro L, LL, Massarelli L, Patamia I, Minniti S, Palermo F, Cacopardo B 2003, 'A Placebo-Controlled Treatment Trial of Blastocystis hominis Infection with Metronidazole', J Travel Med, vol. 10, pp. 128-30.

Nimri, LF 1993, 'Evidence of an epidemic of Blastocystis hominis infections in preschool children in northern Jordan', J Clin Microbiol, vol. 31, no. 10, pp. 2706-8.

Noel, C, Dufernez, F, Gerbod, D, Edgcomb, VP, Delgado-Viscogliosi, P, Ho, LC, Singh, M, Wintjens, R, Sogin, ML, Capron, M, Pierce, R, Zenner, L & Viscogliosi, E 2005, 'Molecular phylogenies of Blastocystis isolates from different hosts: implications for genetic diversity, identification of species, and zoonosis', J Clin Microbiol, vol. 43, no. 1, pp. 348-55.

Noureldin, MS, Shaltout, AA, El Hamshary, EM & Ali, ME 1999, 'Opportunistic intestinal protozoal infections in immunoompormised children', J Egypt Soc Parasitol, vol. 29, no. 3, pp. 951-61.

Nourrisson, C, .,, Scanzi, J, Pereira, B, NkoudMongo, C, Wawryzyniak, I, Cian, A, Viscogliosi, E, Livrelli, V, Delbac, F, Dapoigny, M & Poirier, P 2014,

212 'Blastocystis Is Associated with Decrease of Fecal Microbiota Protective bacteria: Comparative Analysis between Patients with Irritable Bowel Syndrome and Control Subjects', PLoS One, vol. 9, no. 11, p. e111868.

O'Brien Andersen, L, Bonde, I, Nielsen, HB & Stensvold, CR 2015, 'A Retrospective Metagenomics Approach to studying Blastocystis', FEMS Microbiol Ecol, vol. June, pp. 1-32.

Ok, UZ, Girginkardesler, N, Balcioglu, C, Ertan, P, Pirildar, T & Kilimcioglu, AA 1999, 'Effect of trimethoprim-sulfamethaxazole in Blastocystis hominis infection', Am J Gastroenterol, vol. 94, no. 11, pp. 3245-7.

Olivo-Diaz A, R-VM, Gudino-Ramirez A et al 2012, 'Findings related to IL-8 and IL-10 gene polymorphisms in a Mexican patient population with irritable bowel syndrome infected with Blastocystis', Parsitol Res, vol. 111, pp. 487-91.

Olivo-Diaz, A, Romero-Valdovinos, M, Gudino-Ramirez, A, Reyes-Gordillo, J, Jimenez-Gonzalez, DE, Ramirez-Miranda, M, Martinez-Flores, WA, Martinez- Hernandez, F & Flisser, A 2012, 'Findings related to IL-8 and IL-10 gene polymorphisms in a Mexican patient population with irritable bowel syndrome infected with Blastocystis', Parasitol Res, vol. 111, pp. 487-91.

Organisation, WH 2008, Guidelines for drinking water quality (incorp first and second addenda), WHO, Geneva.

Parija, SC & Jeremiah, SS 2013, 'Blastocystis:Taxonomy, biology and virulence', Trop Parasitol, vol. Jan-Jun, pp. 17-25.

Parkar, U, Traub, RJ, Kumar, S, Mungthin, M, Vitali, S, Leelayoova, S, Morris, K & Thompson, RC 2007, 'Direct characterization of Blastocystis from faeces by PCR and evidence of zoonotic potential', Parasitology, vol. 134, no. Pt 3, pp. 359-67.

Parkar, U, Traub, RJ, Vitali, S, Elliot, A, Levecke, B, Robertson, I, Geurden, T, Steele, J, Drake, B & Thompson, RC 2010, 'Molecular characterization of Blastocystis isolates from zoo animals and their animal-keepers', Vet Parasitol, vol. 169, no. April 19, pp. 8-17.

Paschke, C, Apelt, N, Fleischmann, E, Perona, P, Walentiny, C, Loscher, T & Herbinger, KH 2010, 'Controlled study on enteropathogens in travellers returning from the tropics with and without diarrhoea', Clin Microbiol Infect, vol. 17, no. 8, pp. 1194-200.

Petersen AM, SC, Mirsepasi H, Engberg J, Fris-Moller A, Porsbo LJ, Hammerum AM, Nordgaard-Lassen I, Nielsen HV, Krogfelt KA 2012, 'Active ulcerative colitis associated with low prevalence of Blastocystis and Dientamoboe fragilis infection', Scand J Gastroenterol, no. Mar 25.

Phillips, BP & Zierdt, CH 1976, 'Blastocystis hominis: pathogenic potential in human patients and in gnotobiotes', Exp Parasitol, vol. 39, no. 3, pp. 358-64.

213

Pickering, LK, Woodward, WE, DuPont, HL & Sullivan, P 1984, 'Occurrence of Giardia lamblia in children in day care centers', J Pediatr, vol. 104, no. 4, pp. 522-6.

Pimental, M, Lembo, A, Chey, W, Zakko, S, Ringel, Y, Yu, J, Mareya, SM, Shaw, AL, Bortey, E & Forbes, W 2011, 'Rifaximin Therapy for patients with Irritable Bowel Syndrome without consipation', NEJM, vol. 364, pp. 22-32.

Poirier, P, Wawrzyniak, I, Albert, A, El Alaoui, H, Delbac, F & Livrelli, V 2011, 'Development and evaluation of a real-time PCR assay for detection and quantification of blastocystis parasites in human stool samples: prospective study of patients with hematological malignancies', J Clin Microbiol, vol. 49, no. 3, pp. 975-83.

Poirier, P, Wawrzyniak, I, Vivares, CP, Delbac, F & El Alaoui, H 2012, 'New insights into Blastocystis species; a potential link with irritable bowel syndrome', PLoS Pathogens, vol. 8, no. 3, p. e1002545.

Posserud, I, Ersryd, A & Simren, M 2006, 'Functional findings in irritable bowel syndrome', WJG, vol. 12, no. 18, pp. 2830-8.

Puthia, MK, Lu, J & Tan, KS 2008, 'Blastocystis ratti contains cysteine proteases that mediate interleukin-8 response from human intestinal epithelial cells in an NF-kappaB-dependent manner', Eukaryot Cell, vol. 7, no. 3, pp. 435-43.

Puthia, MK, Sio, SWS, Lu, J & Tan, KSW 2006, 'Blastocystis ratti Induces Contact-Independent Apoptosis, F-Actin Rearrangement, and Barrier Function Disruption in IEC-6 Cells', Infection and Immunity, vol. 74, no. 7, pp. 4114-23.

Puthia, MK, Vaithilingam, A, Lu, J & Tan, KSW 2005, 'Degradation of human secretory immunoglobulin A by Blastocystis', Parasitol Res, vol. 97, no. 5, pp. 386-9.

Rajamanikam, A & Govind, SK 2013, 'Amoebic forms of Blastocystis spp.- evidence for a pathogenic role', Parasites and Vectors, vol. 6, p. 295.

Rajilić–Stojanović, M, Biagi, E, Heilig, HGHJ, Kajander, K, Kekkonen, RA, Tims, S & de Vos, WM 2011, 'Global and Deep Molecular Analysis of Microbiota Signatures in Fecal Samples From Patients With Irritable Bowel Syndrome', Gastroenterology, vol. 141, no. 5, pp. 1792-801.

Rene, BA, Stensvold, CR, Badsberg, JH & Nielsen, HV 2009, 'Subtype analysis of Blastocystis isolates from Blastocystis cyst excreting patients', Am J Trop Med Hyg, vol. 80, no. 4, pp. 588-92.

Ritchie, J 1973, 'Pain from distension of the pelvic colon by inflating a balloon in the irritable bowel syndrome', Gut, vol. 14, pp. 125-32.

214 Roberts, T, Ellis, J, Harkness, J, Marriott, D & Stark, D 2014, 'Treatment failure in patients with chronic Blastocystis infection', J Med Microbiol, vol. 63, pp. 252-7.

Roberts, T, Stark, J & Ellis, J 2012, 'Subtype distribution of Blastocystis isolates identified in a Sydney population and pathogenic potential of Blastocystis', Eur J Clin Microbiol Infect Dis, vol. 2013, no. 32, pp. 335-43.

Sanchez de Medina, F, Romero-Calvo, I, Mascaraque, C & Martinez- Augustin, O 2014, 'Intestinal Inflammation and Mucosal Barrier Function', Inflammatory Bowel Diseases, vol. 20, no. 12, pp. 2394-404.

Scanlan 2012a, 'Blastocystis: past pitfalls and future perspectives', Trends in Parasitology, vol. 28, pp. 327-34.

Scanlan, PD 2012b, 'Blastocystis: past pitfalls and future perspectives', Trends in Parasitology, vol. 28, no. 8, pp. 327-34.

Scanlan, PD & Stensvold, CR 2013, 'Blastocystis: getting to grips with our guileful guest', Trends in Parasitology, vol. 29, no. 11, pp. 523-5.

Scanlan, PD, Stensvold, CR, Rajilic-Stojanovic, M, Heilig, HG, De Vos, WM, O'Toole, PW & Cotter, PD 2014, 'The microbial eukaryote Blastocystis is a prevalent and diverse member of the healthy human gut microbiota', FEMS Microbiol Ecol, vol. 90, no. 1, pp. 326-30.

Schwille-Kiuntke, J, Frick, JS, Zanger, P & Enck, P 2011, 'Post-infectious irritable bowel syndrome--a review of the literature', Z Gastroenterol, vol. 49, no. 8, pp. 997-1003.

Scicluna, SM, Tawari, B & Clark, CG 2006, 'DNA Barcoding of Blastocystis', Protist, vol. 157, pp. 770-85.

Shanahan, F 2012, 'The colonic microbiota and colonic disease', Curr Gastroenterol Rep, vol. 14, no. 5, pp. 446-52.

Shimokawa, K, Katayama, M, Matsuda, Y, Takahashi, H, Hara, I, Sato, H & Kaneko, S 2002, 'Matric metalloproeteinase (MMP)-2 and MMP-9 activities in human seminal plasma', Molecular Human Reproduction, vol. 8, no. 1, pp. 32- 6.

Siegel, S & Castellan, NJ 1988, Nonparametric statistics for the behavioral sciences, McGraw-Hill, New York.

Silberman, JD, Sogin, ML, Leipe, DD & Clark, CG 1996, 'Human parasite finds taxonomic home', Nature, vol. 380, no. 6573, p. 398.

Simren, M, Barbara, G, Flint, HJ, Spiegel, BM, Spiller, RC, Vanner, S, Verdu, EF, Whorwell, PJ, Zoetendal, EG & Rome Foundation, C 2013, 'Intestinal

215 microbiota in functional bowel disorders: a Rome foundation report', Gut, vol. 62, no. 1, pp. 159-76.

Singh, K, Chang, C & Gershwin, ME 2014, 'IgA deficiency and autoimmunity', Autoimmun Rev, vol. 13, no. 2, pp. 163-77.

Singh M, SK, Ho L, Ng G, Yap E 1995, 'Elucidation of the life cycle of the intestinal protozoan Blastocystis hominis', Parsitol Res, vol. 81, pp. 446-50.

Sio, SWS, Puthia, MK, Lee, AS, Lu, J & Tan, KS 2006, 'Protease activity of Blastocystis hominis', Parasitol Res, vol. 99, no. 2, pp. 126-30.

Souppart, L, Moussa, H, Cian, A, Sanciu, G, Poirier, P, El Alaoui, H, Delbac, F, Boorom, K, Delhaes, L, Dei-Cas, E & Viscogliosi, E 2010, 'Subtype analysis of Blastocystis isolates from symptomatic patients in Egypt', Parasitol Res, vol. 106, pp. 505-11.

Stark, D, Fotedar, R, van Hal, S, Beebe, N, Marriott, D, Ellis, JT & Harkness, J 2007, 'Prevalence of enteric protozoa in human immunodeficiency virus (HIV)- positive and HIV-negative men who have sex with men from Sydney, Australia', Am J Trop Med Hyg, vol. 76, no. 3, pp. 549-52.

Stechmann, A, Hamblin, K, Pérez-Brocal, V, Gaston, D, Richmond, G, van der Giezen, MC, Clark, G & Roger, AJ 2008, 'Organelles in Blastocystis that Blur the Distinction between Mitochondria and ', Curr Biol, vol. 18, pp. 580-5.

Steck, N, Mueller, K, Schermann, M & Haller, D 2014, 'Bacterial proteases in IBD and IBS', Gut, vol. 61, pp. 1610-8.

Stensvold, CR 2012, 'Thinking Blastocystis out of the box', Trends in Parasitology, vol. 28, no. 8, p. 305.

Stensvold, CR, Alfellani, M & Clark, CG 2012, 'Levels of genetic diversity vary dramatically between Blastocystis subtypes', Infect Genet Evol, vol. 12, pp. 263-73.

Stensvold, CR, Alfellani, MA, Norskov-Lauritsen, S, Prip, K, Victory, EL, Maddox, C, Nielsen, HV & Clark, CG 2009a, 'Subtype distribution of Blastocystis isolates from synanthropic and zoo animals and identification of a new subtype', Int J Parasitol, vol. 39, no. 4, pp. 473-9.

Stensvold, CR, Arendrup, MC, Jespersgaard, C, Molbak, K & Nielsen, HV 2007a, 'Detecting Blastocystis using parasitologic and DNA-based methods: a comparative study', Diagn Microbiol Infect Dis, vol. 59, no. 3, pp. 303-7.

Stensvold, CR, Lewis, HC, Hammerum, AM, Porsbo, LJ, Nielsen, SS, Olsen, KE, Arendrup, MC, Nielsen, HV & Molbak, K 2009b, 'Blastocystis: unravelling potential risk factors and clinical significance of a common but neglected parasite', Epidemiol Infect, vol. 137, no. 11, pp. 1655-63.

216

Stensvold, CR, Lewis, HC, Hammerum, AM, Porsbo, LJ, Nielsen, SS, Olsen, KEP, Arendrup, MC, Nielsen, HV & MØLbak, K 2009c, 'Blastocystis: unravelling potential risk factors and clinical significance of a common but neglected parasite', Epidemiology and Infection, vol. 137, no. 11, p. 1655.

Stensvold, CR, Smith, HV, Nagel, R, Olsen, KE & Traub, RJ 2010, 'Eradication of Blastocystis carriage with antimicrobials: reality or delusion?', J Clin Gastroenterol, vol. 44, no. 2, pp. 85-90.

Stensvold, CR, Suresh, GK, Tan, KSW, Thompson, RCA, Traub, RJ, Viscogliosi, E, Yoshikawa, H & Clark, CG 2007b, 'Terminology for Blastocystis subtypes – a consensus', Trends in Parasitology, vol. 23, no. 3, pp. 93-6.

Stensvold, R, Brillowska-Dabrowska, A, Nielsen, HV & Arendrup, MC 2006, 'Detection of Blastocystis hominis in unpreserved stool specimens by using polymerase chain reaction', J Parasitol, vol. 92, no. 5, pp. 1081-7.

Stenzel, DJ 1995, 'Ultrastructural and cytochemical studies of Blastocystis spp'.

Stenzel, DJ & Boreham, PF 1991, 'A cyst-like stage of Blastocystis hominis', Int J Parasitol, vol. 21, no. 5, pp. 613-5.

Stenzel, DJ & Boreham, PF 1994, 'Bacteria-like endosymbionts in Blastocystis sp', Int J Parasitol, vol. 24, no. 1, pp. 147-9.

Stenzel, DJ & Boreham, PF 1996, 'Blastocystis hominis revisited', Clin Microbiol Rev, vol. 9, no. 4, pp. 563-84.

Stenzel, DJ, Boreham, PF & McDougall, R 1991, 'Ultrastructure of Blastocystis hominis in human stool samples', Int J Parasitol, vol. 21, no. 7, pp. 807-12.

Stenzel, DJ & Boreham, PJ 1997, 'Virus-like paricles in Blastocystis sp from simian faecal materials', Int J Parasitol, vol. 27, pp. 345-8.

Stenzel, DJ, Dunn, LA & Boreham, PF 1989, 'Endocytosis in cultures of Blastocystis hominis', Int J Parasitol, vol. 19, no. 7, pp. 787-91.

Stoop, JW, Zegers, BJ, Sanders, PC & Ballieux, RE 1969, 'Serum immunoglobulin levels in helathy children and adults', Clin Exp Immunol, vol. 4, pp. 101-12.

Suescun-Bolivar, LP & Thome, PE 2015, 'Osmosensing and osmoregulation in unicellular eukaryotes', World J Microbiol Biotechnol, vol. 31, no. 3, pp. 435-43.

217 Sun, T, Katz, S, Tanenbaum, B & Schenone, C 1989, 'Questionable clinical significance of Blastocystis hominis infection', Am J Gastroenterol, vol. 84, no. 12, pp. 1543-7.

Surangsrirat, S, Thamrongwittawatpong, L, Piyaniran, W, Naaglor, T, Khoprasert, C, Taamasri, P, Mungthin, M & Leelayoova, S 2010, 'Assessment of the association between Blastocystis infection and irritable bowel syndrome', J Med Assoc Thai, vol. 93 Suppl 6, pp. S119-24.

Suresh, K, Howe, J, Chong, SY, Ng, GC, Ho, LC, Loh, AK, Ramachandran, NP, Yap, EH & Singh, M 1994a, 'Ultrastructural changes during in vitro encystment of Blastocystis hominis', Parasitol Res, vol. 80, pp. 327-35.

Suresh, K, Ng, GC, Ho, LC, Yap, EH & Singh, M 1994b, 'Differentiation of the various stages of Blastocystis hominis by acridine orange staining', Int J Parasitol, vol. 24, no. 4, pp. 605-6.

Suresh, K, Ng, GC, Ramachandran, NP, Ho, LC, Yap, EH & Singh, M 1993, 'In vitro encystment and experimental infections of Blastocystis hominis', Parasitol Res, vol. 79, no. 6, pp. 456-60.

Suresh, K & Smith, H 2004, 'Comparison of methods for detecting Blastocystis hominis', Eur J Clin Microbiol Infect Dis, vol. 23, no. 6, pp. 509- 11.

Suresh, K, Venilla, GD, Tan, TC & Rohela, M 2009, 'In vivo encystation of Blastocystis hominis', Parasitol Res, vol. 104, no. 6, pp. 1373-80.

Suzuki, K, Kawamoto, S, Maruya, M & Fagarasan, S 2010, 'GALT: organization and dynamics leading to IgA synthesis', Adv Immunol, vol. 107, pp. 153-85.

Swidsinski, A, Weber, J, Loening-Baucke, V, Hale, LP & Lochs, H 2005, 'Spatial organization and composition of the mucosal flora in patients with inflammatory bowel disease', J Clin Microbiol, vol. 43, no. 7, pp. 3380-9.

Tai WP, HP, Wu J, Lin XC 2011, 'Six ulcerative colitis patients with refractory symptoms co-infective with Blastocystis hominis in China', Parasitol Res, vol. 108, pp. 1207-10.

Tan, HK, Harrison, M & Zierdt, CH 1974, 'Freeze-etch studies of the Granular and Vacuolated forms of Blastocystis hominis', Z Parasitenkd, vol. 44, pp. 267-78.

Tan, HK & Zierdt, CH 1973, 'Ultrastructure of Blastocystis hominis', Z Parasitenkd, vol. 42, pp. 315-24.

Tan, KS 2004, 'Blastocystis in humans and animals: new insights using modern methodologies', Vet Parasitol, vol. 126, no. 1-2, pp. 121-44.

218 Tan, KS, Ibrahim, M, Ng, GC, Nasirudeen, AMA, Ho, LC, Yap, EH & Singh, M 2001, 'Exposure of Blastocystis species to a cytotoxic monoclonal antibody', Parasitol Res, vol. 87, no. 7, pp. 534-8.

Tan, KS & Stenzel, DJ 2003, 'Multiple reproductive processes in Blastocystis: proceed with caution', Trends Parasitol, vol. 19, no. 7, pp. 290-1; author reply 1-2.

Tan, KSW 2008, 'New insights on classification, identification, and clinical relevance of Blastocystis spp', Clin Microbiol Rev, vol. 21, no. 4, pp. 639-65.

Tan, KSW, Mirza, H, Teo, JDW, Wu, B & MacAry, PA 2010, 'Current Views on the Clinical Relevance of Blastocystis spp', Current Infectious Disease Reports, vol. 12, no. 1, pp. 28-35.

Tan, KSW, Ng, GC, Quek, E, Howe, NP, Ramachandran, NP, Yap, EH & Singh, M, . 2000, 'Blastocystis hominis: A simplified, High-efficiency method for clonal growth on solid agar', Exp Parasitol, vol. 96, pp. 9-15.

Tan, SW, Ho, LC, Moe, KT, Chen, XQ, Ng, GC, Yap, EH & Singh, M 1996, 'Production and characterization of murine monoclonal antibodies to Blastocystis hominis', Int J Parasitol, vol. 26, no. 4, pp. 375-81.

Tan, SW, Singh, M, Ho, LC, Howe, J, Moe, KT, Chen, XQ, Ng, GC & Yap, EH 1997, 'Survival of Blastocystis hominis clones after exposure to a cytotoxic monoclonal antibody', Int J Parasitol, vol. 27, no. 8, pp. 947-54.

Tan, TC & Suresh, KG 2006a, 'Amoeboid form of Blastocystis hominis - a detailed ultrastructural insight', Parasitol Res, vol. 99, no. 6, pp. 737-42.

Tan, TC & Suresh, KG 2006b, 'Predominance of amoeboid forms of Blastocystis hominis in isolates from symptomatic patients', Parasitol Res, vol. 98, no. 3, pp. 189-93.

Tan, TC, Suresh, KG & Smith, HV 2008, 'Phenotypic and genotypic characterisation of Blastocystis hominis isolates implicates subtype 3 as a subtype with pathogenic potential', Parasitol Res, vol. 104, no. 1, pp. 85-93.

Tan, TC, Suresh, KG, Thong, KL & Smith, HV 2006, 'PCR fingerprinting of Blastocystis isolated from symptomatic and asymptomatic human hosts', Parasitol Res, vol. 99, no. 4, pp. 459-65.

Tang, YZ & Dobbs, FC 2007, 'Green autofluorescence in dinoflagellates, diatoms, and other microalgae and its implications for vital staining and morphological studies', Appl Environ Microbiol, vol. 73, no. 7, pp. 2306-13.

Tanizaki, A, Yoshikawa, H, Iwatani, S & Kimata, I 2005, 'Infectivity of Blastocystis isolates from chickens, quails and geese in chickens', Parasitol Res, vol. 96, no. 1, pp. 57-61.

219 Tasova, Y, Sahin, B, Koltas, S & Paydas, S 2000, 'Clinical significance and frequency of Blastocystis hominis in Turkish patients with hematological malignancy', Acta Med Okayama, vol. 54, no. 3, pp. 133-6.

Teow, WL, Ho, LC, Ng, GC, Chan, YC, Yap, EH & Chan, PP 1992, 'Virus-like particles in a Blastocystis species from the sea snake, Lapemis hardwickii', Int J Parasitol, vol. 22, pp. 1029-32.

Torres, P, Miranda, JC, Flores, L, Riquelme, J, Franjola, R, Perez, J, Auad, S, Hermosilla, C & Riquelme, S 1992, '[Blastocystosis and other intestinal protozoan infections in human riverside communities of the Valdivia River basin, Chile]', Rev Inst Med Trop Sao Paulo, vol. 34, no. 6, pp. 557-64.

Tsuda, A, Suda, W, Morita, H, Takanashi, K, Takagi, A, Koga, Y & Hattori, M 2015, 'Influence of Proton-Pump Inhibitors on the Luminal Microbiota in the ', Clin Transl Gastroenterol, vol. 6, p. e89.

Tungtrongchitr, A, Manatsathit, S, Kositchaiwat, C, Ongrotchanakun, J, Munkong, N, Chinabutr, P, Leelakusolvong, S & Chaicumpa, W 2004, 'Blastocystis hominis infection in irritable bowel syndrome patients', Southeast Asian J Trop Med Public Health, vol. 35, no. 3, pp. 705-10.

Tuteja, AK, Talley, NJ, Gelman, SS, Alder, SC, Thompson, C, Tolman, K & Hale, DC 2008, 'Development of functional diarrhea, constipation, irritable bowel syndrome, and dyspepsia during and after traveling outside the USA', Dig Dis Sci, vol. 53, no. 1, pp. 271-6.

Upcroft, JA, Dunn, LA, Dommett, LS, Healey, A, Upcroft, P & Boreham, PFL 1989, 'Chromosomes of Blastocystis hominis', Int J Parasitol, vol. 19, pp. 879- 83.

Utzinger, J, Botero-Kleiven, S, Castelli, F, Chiodini, PL, Edwards, H, Kohler, N, Gulletta, M, Lebbad, M, Manser, M, Matthys, B, N'Goran, EK, Tannich, E, Vounatsou, P & Marti, H 2010, 'Microscopic diagnosis of sodium acetate- acetic acid-formalin-fixed stool samples for helminths and intestinal protozoa: a comparison among European reference laboratories', Clin Microbiol Infect, vol. 16, no. 3, pp. 267-73.

Valido, EM & Rivera, WL 2007, 'Colony growth of Philippine isolates of Blastocystis hominis in simplified, soft agar medium', Parasitology Research, vol. 101, no. 1, pp. 213-7.

Valsecchi, R, Leghissa, P & Greco, V 2004, 'Cutaneous lesions in Blastocystis hominis infection', Acta Derm Venereol, vol. 84, no. 4, pp. 322-3. van Hellemond, J, Molhoek, N, Koelewij, R, Wismans, PJ & Genderen, PJ 2013, 'Is paromomycin the drug of choice for eradication of Blastocystis in adults?', J Infect Chemother, vol. 19, pp. 545-8.

220 Vandooren, J, Van den Steen, PE & Opdenakker, G 2013, 'Biochemistry and molecular biology of gelatinase B or matrix metalloproteinase-9 (MMP-9): the next decade', Crit Rev Biochem Mol Biol, vol. 48, no. 3, pp. 222-72.

Vdovenko, AA & Williams, JE 2000, 'Blastocystis hominis: neutral red supravital staining and its application to in vitro drug sensitivity testing', Parasitol Res, vol. 86, no. 7, pp. 573-81.

Vogelberg, C, Stensvold, CR, Monecke, S, Ditzen, A, Stopsack, K, Heinrich- Grafe, U & Pohlmann, C 2010, 'Blastocystis sp. subtype 2 detection during recurrence of gastrointestinal and urticarial symptoms', Parasitol Int, vol. 59, no. 3, pp. 469-71.

Walderich, B, Bernauer, S, Renner, M, Knobloch, J & Burchard, GD 1998, 'Cytopathic effects of Blastocystis hominis on Chinese hamster ovary (CHO) and adeno carcinoma HT29 cell cultures', Trop Med Int Health, vol. 3, no. 5, pp. 385-90.

Wang, W, Bielefeldt-Ohmann, H, Traub, RJ, Cuttell, L & Owen, H 2014a, 'Location and pathogenic potential of Blastocystis in the porcine intestine', PLoS One, vol. 9, no. 8, p. e103962.

Wang, W, Cuttell, L, Bielefeldt-Ohmann, H, Inpankaew, T, Owen, H & Traub, RJ 2013, 'Diversity of Blastocystis subtypes in dogs in different geographical settings', Parasites and Vectors, vol. 5, p. 215.

Wang, W, Cuttell, L, Traub, RJ, Owen, H & Bielefeldt-Ohmann, H 2014b, 'Characterisation of the Blastocystis specific faecal IgA immune response in pigs', Parasite Immunol, vol. 36, no. 10, pp. 503-8.

Wang, W, Owen, H, Traub, RJ, Cuttell, L, Ipankaew, T & Beilefeldt-Ohmann, H 2014c, 'Molecular epdiemiology of Blastocystis in pigs and their in-contact humans in Southeast Queensland, Australia, and Cambodia', Veterinary Parasitology, vol. 203, no. 3-4, pp. 264-9.

Wawryniak, I, Courtine, D, Osman, M, Hubans-Pierlot, C, Cian, A, Nourrisson, C, Chabe, M, Poirier, P, Bart, A, Polonais, V, Delgado-Viscogliosi, P, El Alaoui, H, Belkorchia, A, van Gool, T, Tan, KSW, Ferreira, S, Viscogliosi, E & Delbac, F 2015, 'Draft genome sequence of the intestinal parasite Blastocystis subtype 4-isolate WR1', Genomics Data, vol. 4, pp. 22-3.

Wawrzyniak, I, Courtine, D, Osman, M, Hubans-Pierlot, C, Cian, A, Nourrisson, C, Chabe, M, Poirier, P, Bart, A, Polonais, V, Delgado-Viscogliosi, P, El Alaoui, H, Belkorchia, A, van Gool, T, Tan, KSW, Ferreira, S, Viscogliosi, E & Delbac, F 2015, 'Draft genome sequence of the intestinal parasite Blastocystis subtype 4-isolate WR1', Genomics Data, vol. 4, pp. 22-3.

Wawrzyniak, I, Texier, C, Poirier, P, Viscogliosi, E, Tan, KSW, Delbac, F & El Alaoui, H 2012, 'Characterization of two cysteine proteases secreted by

221 Blastocystis ST7, a human intestinal parasite', Parasitology International, vol. 61, no. 3, pp. 437-42.

Weber-Mzell, D, Kotanko, P, Hauer, AC, Goriup, U, Haas, J, Lanner, N, Erwa, W, Ahmaida, IA, Haitchi-Petnehazy, S, Stenzel, M & Deutsch, J 2004, 'Gender, age and seasonal effects on IgA deficiency, a study of 7293 Caucasians', Eur J Clin Investigation, vol. 34, pp. 224-8.

Whorwell, PJ 2014, 'IBS in 2014: Developments in pathophysiology, diagnosis and management', Nat Rev Gastroenterol Hepatol, vol. 12, no. 2, pp. 72-4.

Windsor, JJ, Stenzel, DJ & Macfarlane, L 2003, 'Multiple reproductive processes in Blastocystis hominis', Trends Parasitol, vol. 19, no. 7, pp. 289- 90; author reply 91-2.

Winiecka-Krusnell, J & Linder, E 2001, 'Bacterial infections of free-living amoebae', Res Microbiol, vol. 152, pp. 613-9.

Wong, KH, Ng, GC, Lin, RT, Yoshikawa, H, Taylor, MB & Tan, KS 2008, 'Predominance of subtype 3 among Blastocystis isolates from a major hospital in Singapore', Parasitol Res, vol. 102, no. 4, pp. 663-70.

Wu, B, Yin, J, Texier, C, Roussel, M & Tan, KS 2010, 'Blastocystis legumain is localized on the cell surface, and specific inhibition of its activity implicates a pro-survival role for the enzyme', J Biol Chem, vol. 285, no. 3, pp. 1790-8.

Yaicharoen, R, Sripochang, S, Sermsart, B & Pidetcha, P 2005, 'Prevalence of Blastocystis hominis infection in asymptomatic individuals from Bangkok, Thailand', Southeast Asian J Trop Med Public Health, vol. 36 Suppl 4, pp. 17- 20.

Yakoob, J, Abbas, Z, Beg, MA, Naz, S, Awan, S, Hamid, S & Jafri, W 2011, 'In vitro sensitivity of Blastocystis hominis to garlic, ginger, white cumin, and black pepper used in diet', Parasitol Res, vol. 109, pp. 379-85.

Yakoob, J, Abbas, Z, Usman, MW, Sultana, A, Islam, M, Awan, S, Ahmad, Z, Hamid, S & Jafri, W 2014, 'Cytokine changes in colonic mucosa associated with Blastocystis spp. subtypes 1 and 3 in diarrhea-predominant irritable bowel syndrome', Parasitology, vol. 141, pp. 957-67.

Yakoob, J, Jafri, W, Beg, MA, Abbas, Z, Naz, S, Islam, M & Khan, R 2010, 'Irritable bowel syndrome: is it associated with genotypes of Blastocystis hominis', Parasitol Res, vol. 106, no. 5, pp. 1033-8.

Yakoob, J, Jafri, W, Jafri, N, Khan, R, Islam, M, Beg, MA & Zaman, V 2004, 'Irritable bowel syndrome: in search of an etiology: role of Blastocystis hominis', Am J Trop Med Hyg, vol. 70, no. 4, pp. 383-5.

222 Yamada, M, Yoshikawa, H, Tegoshi, T, Matsumoto, Y, Yoshikawa, T, Shiota, T & Yoshida, Y 1987, 'Light microscopical study of blastocystis spp in monkeys and fowls', Parasitol Res, vol. 73, pp. 527-31.

Yan, Y, Su, S, Lai, R, Liao, H, Ye, J, Li, X, Luo, X & Chen, G 2006, 'Genetic variability of Blastocystis hominis isolates in China', Parasitol Res, vol. 99, no. 5, pp. 597-601.

Yan, Y, Su, S, Ye, J, Lai, X, Lai, R, Liao, H, Chen, G, Zhang, R, Hou, Z & Luo, X 2007, 'Blastocystis sp. subtype 5: a possibly zoonotic genotype', Parasitology Research, vol. 101, no. 6, pp. 1527-32.

Yoshikawa, H 2012, Epidemiology, transmission and zoonotic potential of Blastocystis in human and animals, Blastocystis: Pathogen or passenger, Springer, Berlin.

Yoshikawa, H, Dogruman-Ai, F, Turk, S, Kustimur, S, Balaban, N & Sultan, N 2011, 'Evaluation of DNA extraction kits for molecular diagnosis of human Blastocystis subtypes from fecal samples', Parasitol Res, vol. 109, no. 4, pp. 1045-50.

Yoshikawa, H, Kuwayama, N & Enose, Y 1995a, 'Histochemical Detection of Carbohydrates of Blastocystis hominis', J Eukaryot Microbiol, vol. 42, pp. 70- 4.

Yoshikawa, H, Kuwayama, N & Enose, Y 1995b, 'Histochemical detection of carbohydrates of Blastocystis hominis', J Eukaryot Microbiol, vol. 42, no. 1, pp. 70-4.

Yoshikawa, H, Kuwayama, N & Enose, Y 1995c, 'Histochemical detection of carbohydrates of Blastocystis hominis.', J Eukaryot Microbiol., vol. 42, no. 1, pp. 70-4.

Yoshikawa, H, Satoh, J & Enose, Y 1995d, 'Light and electron localisation of lipids in Blastocystis hominins', J Electron Microsc, vol. 44, pp. 100-3.

Yoshikawa, H, Satoh, J & Enose, Y 1995e, 'Light and electron microscopic localization of lipids in Blastocystis hominis', J Electron Microsc (Tokyo), vol. 44, no. 2, pp. 100-3.

Yoshikawa, H, Wu, Z, Kimata, I, Iseki, M, Ali, IK, Hossain, MB, Zaman, V, Haque, R & Takahashi, Y 2004a, 'Polymerase chain reaction-based genotype classification among human Blastocystis hominis populations isolated from different countries', Parasitol Res, vol. 92, no. 1, pp. 22-9.

Yoshikawa, H, Yoshida, K, Nakajima, A, Yamanari, K, Iwatani, S & Kimata, I 2004b, 'Fecal-oral transmission of the cyst form of Blastocystis hominis in rats', Parasitol Res, vol. 94, no. 6, pp. 391-6.

223 Zaki, M, Zaman, V & Sheikh, NA 1996, 'Resistance of blastocystis hominis cysts to chlorine', J Pak Med Assoc, vol. 46, no. 8, pp. 178-9.

Zaman, V 1996, 'The diagnosis of Blastocystis hominis cysts in human faeces', J Infect, vol. 33, no. 1, pp. 15-6.

Zaman, V, Howe, J, Ng, M & Goh, TK 1999, 'Scanning electron microscopy of the surface coat of Blastocystis hominis', Parasitol Res, vol. 85, no. 12, pp. 974-6.

Zhang, X, Qiao, JY, Zhou, XJ, Yao, FR & Wei, ZC 2007, 'Morphology and reproductive mode of Blastocystis hominis in diarrhea and in vitro', Parasitol Res, vol. 101, no. 1, pp. 43-51.

Zhang, X, Zhang, S, Qiao, JY, Wu, X, Zhao, L, Liu, Y & Fan, X 2011, 'Ultrastructural insights into morphology and reproductive mode of Blastocystis hominis', Parasitology Research, vol. 110, no. 3, pp. 1165-72.

Zierdt, CH 1988, 'Blastocystis hominis, a long-misunderstood intestinal parasite', Parasitol Today, vol. 4, no. 1, pp. 15-7.

Zierdt, CH 1991, 'Blastocystis hominis--past and future', Clin Microbiol Rev, vol. 4, no. 1, pp. 61-79.

Zierdt, CH & Nagy, B 1993, 'Antibody response to Blastocystis hominis infections', Ann Intern Med, vol. 118, no. 12, pp. 985-6.

Zierdt, CH, Swan, JC & Hosseini, J 1983, 'In vitro response of Blastocystis hominis to drugs', J Protozool, vol. 30, no. 2, pp. 332-4.

Zierdt, CH & Tan, HK 1976a, 'Endosymbiosis in Blastocystis hominis', Exp Parasitol, vol. 39, no. 3, pp. 422-30.

Zierdt, CH & Tan, HK 1976b, 'Ultrastructure and light microscope appearance of Blastocystis hominis in a patient with enteric disease', Parasitol Res, vol. 50, no. 3, pp. 277-83.

Zierdt, CH & Williams, R 1974, 'Blastocystis hominis: axenic cultures', Exp Parasitol, vol. 36, pp. 233-43.

Zuckerman, MJ, Watts, MT, Ho, H & Meriano, FV 1994, 'Blastocystis hominis infection and intestinal injury', Am J Med Sci, vol. 308, no. 2, pp. 96-101.

Zuel-Fakkar, NM, Abdel Hameed, DM & Hassanin, OM 2011, 'Study of Blastocystis hominis isolates in urticaria: a case-control study', Clin Exp Dermatol, vol. 36, no. 8, pp. 908-10.

224

9 Appendices

225 9.1 Appendix 1 Jones Medium

Distilled water 962.5ml

Na2HPO4 disodium hydrogen orthophsophate 1.244 g

KH2PO4 monopotassium phospahte 0.397 g NaCl sodium chlordie 7.087 g

Discard 12.5 ml of solution -12.5ml

Prepare 1% yeast solution Add the 100 ml yeast 1g yeast extract in 100ml distilled water

Adjust pH to 7.0 using 1M NaOH sodium hydroxideas required

Autoclave at 121 C for 30 minutes

9.2 Appendix 2 Lockes medium

Distilled water 1 litre Glucose 5 g NaCl sodium chloride 16 g Calcium Cl calcium chloride 0.4 g KCl potassium chloride 0.4 g Magnesium chloride 0.02 g Sodium phosphate dibasic 4 g Sodium bicarbonate 0.8 g Potassium phosphate monobasic 0.6 g

Boil this solution, allow to settle overnight, filter through Whitmans paper no1

Add 0.02% phenol red (1:25, v/v), autoclave, Keep at 4 C till use

9.3 Appendix 3 Biphasic solid and liquid phase

SOLID PHASE Beat eggs with Locke solution and then run through a gauze filter 4 parts egg: 1 part Locke solution Dispense the ovosuspension into 5ml aliquots ?45ml per egg : 12.5 ml Locke solution Into small screw topped jars Put at 30 C slant for 30mins at 80 C Autoclave, then store at 4 C

LIQUID PHASE Locke’s solution plus heat inactivated horse serum 1: Add 4mL liquid phase to solid phase in Pre-reduce

226

Appendix 5.1 Supplementary File 5.1: Clinical Diary of patients 1 2 3 4 5 6 7

Time/ Baseline AB Week 1 AB Week 2 After AB After AB After AB After AB Clinical Patient Week 1 Week 2 Week 3 Week 4 Score F C WB F C WB F C WB F C WB F C WB F C WB F C WB 1 2.6 3.4 6 2.9 3.6 5.6 3.3 3.9 7.4 2.4 3.3 7.4 2.7 3.0 7.4 2.4 3.4 7.9 2.1 3 9 +4.1 2 1.5 3 5.1 1.3 2.7 5.6 1.1 2.6 6.9 1.3 2.9 6.1 1.4 3.0 7.0 1.3 2.4 7.0 0.9 2.8 6.6 +2.4 3 3.0 3.3 5.6 2.1 2.7 4.4 3.7 3.0 4.0 3.3 3.1 4.0 2.4 3.0 4.3 2.3 2.7 5.0 2.7 3.3 5.0 -0.3 4 2.6 3.5 7.2 2.6 3.1 6.9 3.1 3.4 6.8 3.2 3.5 6.9 2.8 3.6 6.9 2.8 3.5 7.0 2.5 3.6 7.4 +0.2 5 1.6 3.4 7.1 1.3 2.2 8.0 1.0 2.6 8.0 1.3 2.3 8.0 1.4 3.1 8.0 1.3 3.0 8.1 1.0 3.0 8.1 +2.0 6 1.3 3.4 7.1 1.8 3.3 7.1 1.4 3.7 6.9 1.3 3.5 7.5 1.3 3.5 7.4 1.3 3.2 7.5 1.2 3.1 7.8 +1.1 7 4.4 4.0 5.9 2.0 2.7 6.0 1.3 2.5 6.0 2.1 3.0 6.9 2.0 2.9 6.6 2.0 2.9 6.7 2.2 2.8 7.0 +4.5 8 2.6 3.0 7.0 3.0 3.1 6.7 2.7 3.3 7.3 1.3 3.0 7.6 1.4 3.0 7.0 1.6 3.4 7.1 1.6 3.1 7.4 +1.3 9 1.6 2.0 9.0 2.0 3.0 7.8 4.4 3.0 7.3 4.4 3.0 8.3 2.0 3.0 8.8 1.0 2.2 9.0 2.0 2.2 8.8 -0.8 10 2.0 2.4 6.7 2.4 3.2 3.6 2.0 2.6 6.4 2.1 2.1 4.1 1.9 2.2 5.2 1.2 2.5 6.5 1.6 2.7 6.0 -0.6

F=mean bowel frequency over 7 days C=mean consistency over 7 days WB=mean well being score over 7 days AB=antibiotic therapy F=number of bowel movements per 24 hours C=consistency of bowel action scored: 1= very hard 2=hard 3=formed 4=loose 5=watery WB= feeling of wellbeing assessed daily scored 1=feels terrible 10=best ever Final Clinical score = Improvement in stool frequency (1-7) + improvement in stool consistency (1-7) +improvement in wellbeing (7-1)

227

228

229