Molecular Methods for Campylobacter and Arcobacter Detection

Marwan Abu-Halaweh, MSc. (Hons)

A Thesis submitted in fulfillment of the requirements of the degree of Doctor of Philosophy in the School of Biomolecular and Biomedical Science, Faculty of, Griffith University, Nathan Campus, Queensland, Australia. January 2005

I

Statement of Originality This work has not previously been submitted for a degree or diploma in any university. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made in the thesis itself.

Marwan Abu-Halaweh

I

Acknowledgments First of all, thanks is to be directed to GOD. I thank Associate Professor Bharat Patel for the opportunity to work with him. His advice, support and assurance gave me enthusiasm and confidence during my research. I appreciate the advice, assistance and friendship of Dr Ben Mijts, Dr Mark Spanevello, Dr David Innes, Dr Sungwan Kanso, Dr Lyle McMillen, Mr Peter Bain and all members of our lab throughout the course of my research in Professor Patel’s laboratory. I acknowledge Dr John Bates, Miss Trudy and the Bacteriology Division of the Queensland Health Scientific Services and the staff for their help and assistance on site. I gratefully acknowledge the financial support from Queensland Health Scientific Services and Griffith University. Also I gratefully acknowledge the editors at WordsRU and Blayse Research for the proofreading of this thesis. All my friends who continued asking “So, how far to go now?” are mentioned for their ongoing companionship. Finally, I thank my family (especially my father and mother) for their continued interest, praying for me and encouraging me in all that I have done and will do, and my wife for her support during this time.

II

Publications and Proceedings Arising from this Thesis Research Paper Abu-Halaweh M, Bates J, Patel BKC: Rapid detection and differentiation of pathogenic Campylobacter jejuni and Campylobacter coli by real-time PCR. Res Microbiol 2005, 156(1):107-114. This paper generated from chapter 3 and chapter 4.

Conference Presentations (supported by a grant for conference support by Griffith University) ¾ Abu-Halaweh M., Bates J. and Patel B.K.C. (2004). Rapid detection and identification of Campylobacter jejuni and Campylobacter coli directly from chicken samples

by

real-time

PCR

Bahrain: The first GCC Genetic conference - Bahrain, 5-7 October 2003. This presentation generated from chapter 3. ¾ Abu-Halaweh M., Bates J. and Patel B.K.C. (2004). Rapid detection and identification of Campylobacter and Arcobacter species by real-time PCR. European Meeting on Molecular Diagnostics - Kurhaus Hotel The Hague / Scheveningen – Netherlands - 16th & 17th October 2003. This presentation generated from chapter 6. ¾ Abu-Halaweh M., Bates J. and Patel B.K.C. (2004). Rapid detection of Campylobacter species using Ligase Detection Reaction (LDR). European Meeting on Molecular Diagnostics - Kurhaus Hotel The Hague / Scheveningen – Netherlands - 16th & 17th October 2003. This presentation generated from chapter 9.

III

Publications in Preparation ¾ Rapid detection and identification of Arcobacter species by real-time PCR. ¾ Rapid detection and identification of C. coli and C. jejuni and C. lari from other Campylobacter species by Ligase Chain Reaction. ¾ A real-time PCR multiplexed for rapid detection and identification of C. coli and C. jejuni from other Campylobacter species.

IV

Abstract Twenty species and six subspecies of the genera Arcobacter and Campylobacter have been described to date. All are Gram-negative, microaerophilic, curved, spiral or Sshaped

cells,

and

are

members

of

the

order

Campylobacterales,

class

Epsilonproteobacteria phylum Proteobacteria. Though most members are pathogenic, C. jejuni, C. coli and A. butzleri are the most frequently isolated species from patients suffering from gastrointestinal illness. The current methods for their detection, identification, and differentiation are cumbersome, time consuming and lack specificity. DNA based molecular techniques including real-time Polymerase Chain Reaction (PCR)

and

Fingerprinting

methods

Terminal

Restriction

Fragments

Length

Polymorphism (T-RFLP) and Ligase Detection Reaction (LDR) have been used in this project to develop rapid detection and identification methods for Campylobacter and Arcobacter species. Five real-time PCR methods were developed which include: (a) rapid detection and identification of Campylobacter species using real-time PCR adjacent hybridisation probes, (b) rapid identification of C. jejuni using SYBR Green I, (c) rapid detection and differentiation of Arcobacter species using adjacent hybridisation probes, (d) rapid detection and differentiation of Arcobacter species and the Campylobacter group (C. coli, C. jejuni, C. lari, C. hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae, C lanienae) using melting temperature (Tm) of adjacent hybridisation probes, and (e) a one tube real-time PCR multiplex for the rapid detection and identification of Campylobacter species, C. coli and C. jejuni using a TaqMan Probe, in an iCycler iQTM (BioRad, USA) and Light CyclerTM (Idaho Technology, USA). The real-time PCR methods for a and b, was a two-tube assay which detected, identified and differentiated C. coli, C. lari and C. jejuni from other members of the family Campylobacteraceae. The first tube assay was based on the principle of Fluorescence V

Resonance Energy Transfer (FRET) in which the FRET signal from the hybridisation of two adjacent fluoroprobes, probe Cy5+1046 and a specific downstream 6-FAM probe to the target site within the 16S rRNA gene of the 681 base pair amplicons produced during PCR with primers F2 and Cam Reverse, was continuously measured in a LightCyclerTM (Idaho Technology, USA) enabling the differentiation of C. coli, C. lari and C. jejuni from all other Campylobacter species. Following detection of these species, a second tube assay was used to differentiate C. jejuni from C. coli and C. lari based on the continuous monitoring of an increase in fluorescence in a LightCyclerTM due to the binding of the intercalating dye SYBR Green I to DNA amplicons, produced by primers Hip-2214F and Hip-2474 targeting the hippuricase (hipO) gene known to be present in the C. jejuni genome but not in C. coli, C. lari and other Campylobacter species. The subsequent temperature dependent dissociation of the strands to produce a specific Tm of 85±0.5οC confirmed the presence of C. jejuni, whereas a Tm of 56οC indicated the presence of non-specific primer dimers. This two tube assay was successfully used to identify and differentiate 176 cultures isolated from animals, humans, plants and birds as C. coli and C. lari group (77 isolates), C. jejuni (88 isolates), and other Campylobacter species (11 isolates). Furthermore, this assay was used to identify and differentiate 30-enrichment cultures initiated from chicken samples as C. jejuni and C. coli and C. lari group (26) and C. jejuni (18). In addition, more than one Campylobacter species could also be detected in the same enrichment culture. Method c was a single tube real-time PCR, which employed adjacent hybridisation probes to differentiate pathogenic Arcobacter species using the FRET principle. In this assay, two specific 6-FAM labelled probes, probe Butz specific for A. butzleri and probe Skir-Cry specific for A. skirrowii and A. cryaerophilius, and an adjacent universal Cy5 labelled probe were used for the simultaneous detection, identification and differentiation of A. butzleri from A. skirrowii and A. cryaerophilius. The binding of

VI

either probe Butz or probe Skir-Cry in conjunction with the probe Cy5+1046 to their respective 16S rRNA target regions within the 317 base pair amplicons produced by primers F2 and R5, resulted in an increase in fluorescence during real-time PCR. The subsequent temperature dependent dissociation of the strands to produce specific Tms based on nucleotide heterogeneity of the probe to the target binding sites enabled the differentiation of the three Arcobacter species: a Tm of 67oC due to dissociation of probe Butz confirmed the presence of A. butzleri, and a Tm of 63oC and 65oC due to the dissociation of probe Skir-Cry differentiated A. skirrowii and A. nitrofigilis respectively. The method successfully identified all the 22 Arcobacter isolates obtained from humans and birds as A. butzleri (20 isolates), A. skirrowii (1 isolate), and A. nitrofigilis (1 isolate). The assay also successfully detected A. butzleri in 18 out of 30-enrichment cultures initiated from chicken samples and 9 as A. skirrowii and/or A. cryaerophilus. In addition, more than one Arcobacter species was also be detected in the same enrichment culture. Method d was a new, complex, single tube assay, termed multi FAM adjacent hybridisation real-time PCR assay. It has been developed for the simultaneous detection and differentiation of the Campylobacter group (C. coli, C. jejuni, C. lari, C. hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae, and C. lanienae), A. butzleri, and A. skirrowii. The binding of 6-FAM labelled probes (probe Jejuni-coli, probe Butz, and probe Skir-Cry) in conjunction with the universal Cy5 labelled probe to their specific target sites within the amplicon produced an increase in fluorescence signal, which was measured in a LightCyclerTM. The subsequent dissociation of the probes to produce specific Tms identified and differentiated the targeted species: a Tm of 67oC identified A. butzleri, a Tm of 63oC identified A. skirrowii and a Tm of 65oC identified the Campylobacter group (C. coli, C. jejuni, C. lari, C. hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae, C lanienae), and a Tm of 56oC detected other species.

VII

One hundred and sixty-seven isolates out of 198 Campylobacter and Arcobacter cultures of human, animal, plant and bird origin, were identified as members of the Campylobacter group (C. coli, C. jejuni, C. lari, C. hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae, C lanienae), A. butzleri (20 isolates), A. skirrowii (1 isolate) and other Campylobacter species (7 isolates). Furthermore, 27 out of 60enrichment cultures contained the Campylobacter group (C. coli, C. jejuni, C. lari, C. hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae and C lanienae), whereas 18 detected A. butzleri, and 9 detected A. skirrowii and/or A. cryaerophilus. More than one Arcobacter species was detected in some enrichment cultures. Purified DNA prepared by the phenol-chloroform-isoamyl alcohol method and CTAB method as well as crude DNA prepared by a simple cell lysis rapid boil and clarify method could be successfully used in all the assays. The methods were quantitative and as little as 192 fg/μl of C. jejuni DNA corresponding to 125 C. jejuni cells, or 113fg/μl of A. butzleri DNA corresponding to 110 A. butzleri cells could be successfully detected. In addition, after inoculating 5 colonies in 50 ml enrichment broth, real-time PCR assays could successfully detect C. coli and C. jejuni after 8 hours and A. butzleri after 24 hours incubation. The methods developed using the low-throughput LightCyclerTM were easily transferable to the high-throughput iCycler iQTM. The data obtained from both instruments were almost identical. Method e was a single tube reaction consisting of a 3-target multiplex real-time PCR assay for detecting all Campylobacter species, which could also simultaneously identify and discriminate between Campylobacter coli and Campylobacter jejuni, the most commonly reported food and water-borne gastrointestinal pathogens. The single tube, multiplex real-time assay, consists of three TaqMan probes, each of which is labelled at the 5’ with a different fluorophore reporter molecule (FAM, MAX, and Cy5), and coupled to a Black Hole Quencher (BHQ) at the 3’ end, for analysis in an iCycler iQTM. VIII

The specificity of each target was tested against a panel of pathogens including the closely related Arcobacter species. The assay was subsequently used to identify and differentiate the 176 Campylobacteraceae isolates of animal, human, plant and bird origin held in our culture collection, into C. coli (76 isolates), C. jejuni (88 isolates), and other Campylobacter species (12 isolates). Furthermore, 27 out of 30 enrichment cultures initiated from chicken samples were tested and detected the presence of Campylobacter species, C. jejuni was detected in 18 and C. coli in 11. More than one Campylobacter species was also detected in the same enrichment cultures. In addition, the method successfully detected 125 cells of C. jejuni and 110 cells of C. coli. The recently developed T-RFLP is a DNA fingerprinting method. It was used in this project for rapid detection and screening of the Campylobacter and Arcobacter species based on the 16S rRNA sequence information available online. In this method, two different primers were used: a forward primer, FAM-49F fluorescently labelled with 6FAM at the 5’ end, and a reverse primer, R2, used to amplify a 460- 465 bp amplicon. These amplicons were digested using five different restriction enzymes (DdeI, Sau3AI, KpnI, BtsI, and EcoRI), producing fragments of different lengths based of on the recognition sites of the restriction enzyme. The FAM-labelled fragments were separated using an ABI377 DNA sequencer and the size of the fragment identified as Campylobacter or Arcobacter species. This assay was successfully tested and identified 198 isolates from animal, human, plant and bird origin from our culture collection as C. jejuni, C. coli and C. lari group (165 isolates), other Campylobacter species (11 isolates) and Arcobacter species (22 isolates). Furthermore, of the 60 enrichment cultures Campylobacter species in 27, C. jejuni, C. coli and C. lari group was detected in 26, C. helviticus and/or C. rectus and/or C. concisus in 4, other Campylobacter species in 4 and Arcobacter species in 30 of the cultures. Moreover, two samples

IX

showed the presence of Mycoplasma like and /or Clostridium like, and/or Eubacterium like and/or Lactobacillus like species. LDR is an extremely powerful Single Nucleotide Polymorphism (SNP) DNA fingerprinting technique that has been developed for one base pair (bp) differentiation. LDR relies on the ability of the DNA ligase activity, to ligate two juxtaposed probes only if they have a 100% match to the target sequence at the nick junction. In this project, LDR has been used for detection and identification of Arcobacter species and Campylobacter species of C. coli, C. lari and C. jejuni in a single tube reaction. 16S rRNA universal forward (F2) and reverse (Rd1) primers were used to amplify 759 bp amplicon hybrids the target region for the adjacent LDR common and discriminative probes. Two discriminative probes designated probe LDRCam and probe LDRArco with the ability to specifically hybridise to Campylobacter and Arcobacter species respectively, and an adjacent common probe, designated probe LDRProteo, specific for Proteobacteria 16S rRNA, were used to detect and identify Campylobacter and Arcobacter species. Furthermore, primers LDRFla-7, and LDRFla-630 were used to amplify 643 bp amplicons of the flaA gene region to which the discriminative probes LDRJejuni, LDRColi, and LDRLari, and the LDRFlaA common probe hybridised, to distinguish C. jejuni, C. coli and C. lari respectively. The assay was subsequently used to identify and differentiate the 198 Campylobacteraceae isolates of animal, human, plant and bird origin held in our culture collection, into C. coli (76 isolates), C. jejuni (88 isolates), C. lari (1 isolate), other Campylobacter species (12 isolates), and others (22 isolates). In addition, of the 60 enrichment cultures initiated from chicken samples that were tested Campylobacter species were identifies in 27, C. jejuni in 18, C. coli in 11, C. lari in 2, and Arcobacter species in 22 of the cultures. More than one Campylobacter species was detected in some enrichment cultures.

X

Molecular methods have been used in this study due to their reliability and rapidity of results. Five different real-time PCR methods and two fingerprinting methods used for rapid detection and identification of Campylobacter and Arcobacter species are methods have proven to have a high specificity and sensitivity (100%) compared to the gold standard conventional methods used by QHSS. The results demonstrate that real-time PCR, T-RFLP and LDR, may prove to be a rapid alternative to the traditional cultivation method. The real-time assay allowed detection of Campylobacter and Arcobacter in a closed system reducing the risk of contamination by amplicon carry-over. LDR has the ability to accurately detect different Campylobacter and Arcobacter species in one step based on the size of the LDR products. Furthermore, T-RFLP proved to be a powerful screening technique, exemplified in this project by the detection of more than one Campylobacter species not detected by conventional methods or real-time or LDR assays.

XI

Table of Contents Statement of Originality ..................................................................................................................I Acknowledgments .......................................................................................................................... II Publications and Proceedings Arising from this Thesis ............................................................III

Research Paper .........................................................................................................III Conference Presentations (supported by a grant for conference support by Griffith University)...................................................................................................III Publications in Preparation ..................................................................................... IV Abstract ............................................................................................................................................V List of Figures ......................................................................................................................... XVIII List of Tables ............................................................................................................................. XXV Abbreviations List ....................................................................................................................XXVI Abbreviations List ....................................................................................................................XXVI 1.

Chapter 1 Introduction .......................................................................................................... 1

1.1. The Genus Campylobacter ............................................................................ 1 1.1.1. Natural Reservoirs .................................................................................... 2 1.1.2. Epidemiology and Transmission .............................................................. 3 1.1.3. Mechanisms of Pathogenesis.................................................................... 5 1.1.4. Molecular Biology.................................................................................... 6 1.1.5. Proteomes ................................................................................................. 6 1.2. The Genus Arcobacter .................................................................................... 8 1.2.1. Natural Reservoirs .................................................................................... 8 1.2.2. Pathogenesis and Transmission ................................................................ 9 1.2.3. Culturing of Arcobacter ......................................................................... 10 1.3. Methods of Identification and Differentiation of Campylobacter and Arcobacter .................................................................................................................. 11 1.3.1. Phenotypic Characterisation................................................................... 11 1.3.2. Molecular Methods for the Pathogen Detection..................................... 12 1.4. Recent Development in Molecular Methods for Pathogen Detection...... 13 1.4.1. Choice of Targets ................................................................................... 13 1.4.2. Choice of Techniques ............................................................................. 15 1.5. Real Time PCR ............................................................................................. 15 1.5.1. Equipment and Equipment Formats ....................................................... 16 1.5.1.1 LightCycleTM (Idaho Technology, USA) Real Time PCR Detection System ................................................................................................................ 16 1.5.1.2 iCycler iQTM (BioRad, USA) Real Time PCR Detection System.......... 18 1.5.2. Fluorophore Probe Technology and Chemistries ................................... 18 1.5.2.1 Adjacent Hybridisation Probes Assays .................................................. 20 1.6. Real Time PCR Multiplex............................................................................ 21 1.6.1. TaqMan Probe ........................................................................................ 25 1.6.2. MGB TaqMan Probe .............................................................................. 26 1.7. Fingerprinting Molecular Techniques........................................................ 29 1.7.1. Terminal Restriction Fragment Length Polymorphism (T-RFLP)......... 30 XII

1.7.2. DNA Ligases, and Ligase Based Molecular Techniques ....................... 33 1.7.2.1 Ligase detection reaction ........................................................................ 35 1.8. 2.

Project Aims and Rationale ......................................................................... 37

Chapter 2 Material and Methods ........................................................................................ 39

2.1. Bacterial Cultures, Sources and Growth.................................................... 39 2.1.1. Campylobacter and Arcobacter Enrichment Culture for Chicken sample 40 2.2. DNA Extraction ............................................................................................ 40 2.2.1. CTAB Method ........................................................................................ 41 2.2.2. Rapid DNA Extraction by Boiling ......................................................... 41 2.2.3. Phenol-chloroform-isoamyl Alcohol Method ........................................ 41 2.3. Agarose Gel Electrophoresis of DNA ......................................................... 42 2.3.1. Ultra Agarose Gel Electrophoresis ......................................................... 43 2.4. 16S rRNA Amplification and Sequencing .................................................. 44 2.4.1. 16S rRNA Primer Design ....................................................................... 44 2.4.2. 16S rRNA Amplification........................................................................ 44 2.4.3. Purification of PCR products.................................................................. 45 2.4.4. 16S rRNA Sequencing and Sequence Analysis ..................................... 45 2.5. Real Time PCR ............................................................................................. 48 2.5.1. Design of Adjacent Hybridisation Probes .............................................. 48 2.5.2. Design of TaqMan probe........................................................................ 49 2.5.3. LightCyclerTM Running Parameter for adjacent hybridisation probe assay 49 2.6. Real Time Detection of the Campylobacter Species SYBR Green I ......... 50 2.6.1. LightCyclerTM Detection of the Hippuricase (hipO) Gene ..................... 50 2.6.2. Real Time PCR Detection of hipO Gene using iCycler iQTM ................ 51 2.6.3. iCycler iQTM Triplex Detection of Campylobacter Species................... 53 2.7. Determining the Sensitivity of the Real Time Assay ................................. 55 2.7.1. Using Incubation Time as Parameter...................................................... 55 2.7.2. Using Different DNA Concentration...................................................... 55 2.7.3. Nucleic Acid Quantitation ...................................................................... 55 2.7.4. Using Different Cell Concentration........................................................ 56 2.8. Terminal Restriction Fragment Length Polymorphism (T-RFLP) for Detection and Identification of Campylobacter and Arcobacter Species .............. 56 2.8.1. PCR Amplification and Purification of the Products ............................. 56 2.8.2. Multiplex PCR Amplification for T-RFLP ............................................ 56 2.8.3. Restriction Endonucleases Digestion ..................................................... 57 2.8.4. Recovery of the DNA by Ethanol Precipitation ..................................... 57 2.8.5. T-RFLP Product Detection and Analysis ............................................... 57 2.9. LDR for Detection, Identification of Campylobacter and Arcobacter Species........................................................................................................................ 58 2.9.1. PCR Amplifications and Product Purification ....................................... 58 2.9.2. LDR Product Detection and Analysis .................................................... 58 2.10. Comparison between Culture methods and Molecular methods (Gold Standard)................................................................................................................... 59 3. Chapter 3 Rapid Identification of C. jejuni and C. coli by Targeting the 16S rRNA using Real-Time PCR ............................................................................................................................. 62 XIII

3.1.

Chapter Overview ........................................................................................ 62

3.2. Results............................................................................................................ 62 3.2.1. Detection and Identification of Campylobacter Species by Conventional Culture Methods ..................................................................................................... 62 3.2.2. Identification of Isolates (QHSS B287), (QHSS 00M2260), (QHSS 99M126) and (QHSS 99M2318) as C. coli, C. jejuni, C. upsaliensis and C. hyointestinalis by Partial Sequencing of 16S rRNA .............................................. 63 3.2.3. Development and Design of PCR Primers and Adjacent Hybridisation Probes 64 3.2.4. Real Time Detection of C. jejuni and C. coli Using Adjacent Hybridisation .......................................................................................................... 65 3.2.5. Identification of C. jejuni Isolates in the Culture Collection Using the Real Time Adjacent Hybridisation Probes Assay .................................................. 69 3.2.6. Comparison between Conventional Culture Assay and Adjacent Hybridisation Probes Real-time PCR Assay .......................................................... 70 3.2.7. Real-Time PCR Detection of C. jejuni from Different DNA Concentrations ........................................................................................................ 70 3.2.8. Real-Time PCR Detection of C. jejuni and C. coli from Single Colony Serial Dilutions ....................................................................................................... 73 3.2.9. Real Time Detection of C. jejuni and C. coli during Different Growth Phases 75 3.2.10. Improving Rapidity of the Real Time PCR Assay for the Detection of C. coli and C. jejuni..................................................................................................... 75

4.

3.3.

General Discussion ....................................................................................... 77

3.4.

Conclusion ..................................................................................................... 80

Chapter 4 Rapid Distinction Between C. jejuni and C. coli using SYBR Green I .......... 82

4.1.

Chapter Overview ........................................................................................ 82

4.2.

Materials and Methods ................................................................................ 82

4.3. Results............................................................................................................ 83 4.3.1. Detection of C. jejuni by Conventional Culture Methods ...................... 83 4.3.2. Design of PCR Hippuricase Gene Primer .............................................. 83 4.3.3. Development of a Real Time SYBR Green I PCR Assay using a LightCyclerTM ......................................................................................................... 85 4.3.4. Development of a Real Time PCR SYBR Green I Assay Using an iCycler IQTM ........................................................................................................... 87 4.3.5. Identification of C. jejuni Isolates in the QHSS Culture Collection and Enrichment Cultures using the Real Time SYBR Green I Assay .......................... 89 4.3.6. Comparison between Conventional Culture assay and SYBR green I assay 90

5.

4.4.

General Discussion ....................................................................................... 90

4.5.

Conclusion ..................................................................................................... 93

Chapter 5 Rapid Detection and Identification of Arcobacter Species.............................. 94

5.1.

Chapter Overview ........................................................................................ 94

5.2. Results............................................................................................................ 95 5.2.1. Identification of Strains (NCTC 12481), (NCTC 12713) and (NCTC 12251) as A. butzleri, A. skirrowii and A. nitrofigilis by Partial Sequencing of 16S rRNA 95 XIV

5.2.2. Development and Design of PCR Primers and Adjacent Hybridisation Probes 97 5.2.3. Detection and Identification of Arcobacter Species by Conventional Culture Methods ..................................................................................................... 99 5.2.4. Development of the Real-Time Assay for the Detection of A. butzleri, A. skirrowii and A. nitrofigilis using Adjacent Hybridisation Probes......................... 99 5.2.5. Identification of Arcobacter Species in the Culture Collection and Enrichment Cultures using the Real Time Adjacent Hybridisation Probes Assay100 5.2.6. Comparison between Conventional Culture Assay and Adjacent Hybridisation Probe Assay for the Detection and Identification of Arcobacter Species 102 5.2.7. Real Time PCR Detection of A. Butzleri from Different DNA Concentrations ...................................................................................................... 103 5.2.8. Real Time PCR Detection of A. butzleri from Single Colony Serial Dilutions 105 5.2.9. Real-Time PCR Detection of A. butzleri during Different Growth Phases 107 5.3.

General discussion ...................................................................................... 107

5.4.

Conclusion ................................................................................................... 111

6. Chapter 6 Rapid Identification of Arcobacter species and C. jejuni and C. coli using Tm of Adjacent Hybridisation Probes .............................................................................................. 112

6.1. Chapter Overview ...................................................................................... 112 TM 6.1.1. Real-Time PCR Assay in a LightCycler ........................................... 113 6.2. Results.......................................................................................................... 113 6.2.1. Multi-FAM One Tube Assay for Rapid Detection and Identification of C. jejuni, C. coli, A. butzleri and A. skirrowii ........................................................... 113 6.2.2. Identification of C. jejuni Isolates in the Culture Collection using the Real Time Multi-FAM Adjacent Hybridisation Probes Assay ............................ 116 6.2.3. Comparison between Conventional Culture assay and Multi-FAM Adjacent Hybridisation Probes Assay .................................................................. 117 6.3.

General discussion ...................................................................................... 117

6.4.

Conclusion ................................................................................................... 119

7. Chapter 7 DNA Multiplexing for Rapid Detection and Identification of C. coli and C. jejuni from the Other Campylobacter Species .......................................................................... 120

7.1.

Chapter preview ......................................................................................... 120

7.2. Material and Methods ................................................................................ 120 7.2.1. Methods ................................................................................................ 120 7.2.2. Design of TaqMan Probes .................................................................... 120 7.3. Results and Discussions.............................................................................. 121 7.3.1. Detection and Identification of Campylobacter Species by Conventional Culture Methods ................................................................................................... 121 7.3.2. Development and Design of PCR Primers and TaqMan (5’ Nuclease Assay) Probes ....................................................................................................... 121 7.3.3. Optimising conditions for the Multiplex Real-Time PCR.................... 125 7.3.4. Developing of a Multiplex Real Time PCR Assay .............................. 125 7.3.5. Real Time Detection of C. jejuni and C. coli Using 5’ Nuclease Assay Probe (TaqMan Probe) ......................................................................................... 126 XV

7.3.6. Identification of Campylobacter and C. jejuni and C. coli Isolates in the Culture Collection and Enrichment Cultures using the Triplex Real Time Assay126 7.3.7. Comparison between Conventional Culture assay and Triplex Real Time PCR Assay............................................................................................................ 127 7.3.8. Sensitivity of the Multiplex Real Time PCR Assay............................. 129 7.3.8.1 Multiplex Real-Time PCR detection of C. coli and C. jejuni from Different DNA Concentrations......................................................................... 129 7.3.8.2 Multiplex Real Time PCR detection of C. coli and C. jejuni from Single Colony Serial Dilution...................................................................................... 133 7.4.

General Discussion ..................................................................................... 134

7.5.

Conclusion ................................................................................................... 138

8. Chapter 8 Screening and Identification of Campylobacter and Arcobacter Species up to Species Level using T-RFLP...................................................................................................... 140

8.1.

Introduction ................................................................................................ 140

8.2. Methods ....................................................................................................... 142 8.2.1. Primer Design and Synthesis................................................................ 142 8.2.2. Computer Simulation of 16S rRNA TRF from RDP Database............ 142 8.3. Results.......................................................................................................... 143 8.3.1. Primer Design and Synthesis................................................................ 143 8.3.1.1 T-RFLP Computer Simulation ............................................................. 143 8.3.1.2 T-RFLP Analysis Using ABI 377 DNA Sequencer ............................. 145 48.................................................................................................................................................. 158

8.3.2. Identification of Campylobacter and Arcobacter Isolates in the Culture Collection and enrichment cultures using the T-RFLP Assay ............................. 159 8.3.3. Comparison between Conventional Culture assay and T-RFLP assay 159 8.4.

General discussion ...................................................................................... 159

8.5.

Conclusion ................................................................................................... 161

9. Chapter 9 Detection and Identification of Campylobacter Species, Arcobacter, C. jejuni, C. coli and C. lari using LDR..................................................................................................... 163

9.1.

Introduction ................................................................................................ 163

9.2. Material and Methods ................................................................................ 163 9.2.1. Probe and Primer Design and Synthesis............................................... 163 9.3. Results.......................................................................................................... 165 9.3.1. Design of 16S rRNA and Universal Common and Discriminative Probes 165 9.3.2. PCR Amplification and Ligase Detection Reaction............................. 170 9.3.3. Detection of LDR products using ABI377 (Applied BioSystems) ...... 171 9.3.4. Identification of Campylobacter and Arcobacter Isolates in the Culture Collection and Enrichment Cultures using the LDR Assay ................................. 177 9.3.5. Comparison between Conventional Culture Methods and LDR Assay 177 9.4.

General Discussion ..................................................................................... 177

9.5.

Conclusion ................................................................................................... 179

10.

Chapter 10 Conclusions and Future Directions.......................................................... 180

10.1. The Use of Molecular Methods for the Detection of Food-borne Pathogens................................................................................................................. 187 XVI

10.2. Advantages and Disadvantages of Molecular Methods Compared to Culture Methods ..................................................................................................... 189 11.

Appendices ..................................................................................................................... 192

11.1.

Appendix 1 .............................................................................................. 192

11.2.

Appendix 2 .............................................................................................. 198

11.3.

Appendix 3 .............................................................................................. 199

11.4.

Apendix 4................................................................................................. 201

11.5.

Appendix 5 .............................................................................................. 202

11.6.

Appendix 6 .............................................................................................. 203

11.7.

Appendix 7 .............................................................................................. 204

11.8.

Appendix 8 .............................................................................................. 205

11.9.

Appendix 9 .............................................................................................. 206

11.10.

Appendix 10 ............................................................................................ 208

11.11.

Appendix 11 ............................................................................................ 209

11.12.

Appendix 12 ............................................................................................ 210

11.13.

Appendix 13 ............................................................................................ 211

11.14.

Appendix 14 ............................................................................................ 212

11.15.

Appendix 15 ............................................................................................ 214

11.16.

Appendix 16 ............................................................................................ 215

12.

References ...................................................................................................................... 220

XVII

List of Figures Figure 1.1 Dendrogram showing relationship of Campylobacter (16 species and 6 sub species) and Arcobacter (4 species) within the class Epsilon Proteobacteria, phylum Proteobacteria. Sequences were extracted from Ribosomal Database Project (RDPII, version 8.1) [31] and GenBank (http://www3.ncbi.nlm.nih.gov/BankIt/), aligned using BioEdit (Version 5.0.6) and the dendrogram constructed from the comparison of 828 unambiguous nucleotides using TreeCon [32]. (ATCC = American Type Culture Collection), CCUG (Culture Collection, University of Gutenborg, Sweden), NCTC (National Collection of Type Cultures) CIP (Collection de institute Pasteur). Type cultures T are shown in the brackets. ....................................................................................................................3 Figure 1.2. Cases of Campylobacter and other food-borne infections by month of specimen collection; Centre for Disease Control and Prevention/U.S. Department of Agriculture/Food and Drug Administration Collaborating Sites Food borne Disease Active Surveillance Network, 1996 [5].......4 Figure 1.3 Schematic representation of real-time PCR adjacent hybridisation probes assay. http://www.gensetoligos.com/Products/Probes/prod_LightCycler.html ............................................22 Figure 1.4 Schematic representations of real-time PCR SYBR Green I assay. http://biochem.roche.com/lightcycler/lc_principles/lc_prin_dna_det01.htm .....................................22 Figure 1.5 Absorption spectra of the black hole quencher (BHQ-1, BHQ-2 and BHQ-3), and the suitable fluorescent dye for each BHQ members. ..............................................................................24 Figure 1.6 Schematic representation of real-time PCR TaqMan technique http://www.gensetoligos.com/Products/Probes/prod_Taqman.html...................................................26 Figure 1.7 DPI3 (dihydrocyclopyrroloindole tripeptide) type MGB probe http://www.epochbio.com/technology/MGBs.htm.............................................................................28 Figure 1.8 Water effect on MGB probe/DNA duplex http://www.epochbio.com/products/mgbe_how_it_works.htm..........................................................28 Figure 2.1 A schematic overview of the project is shown in which two approaches namely PCR and fingerprinting, culminating in three major techniques (real-time PCR, LDR and T-RFLP), were used for identifying and differentiating Campylobacter and Arcobacter species. Chapter 2 describes the material and methods, while chapters 3 to 9 contains the results and their discussion. The conclusions from these studies and future directions are outlined in chapter 10. Chapters 3 to 7 describe five different real-time PCR, while chapters 8 and 9 describe two fingerprinting assays (TRFLP and LDR), which differentiate Arcobacter and Campylobacter as well as discriminate species of these genera....................................................................................................................................60 Figure 3.1 The partial sequences of 16S rRNA (positions 817 to 1466 E. coli numbering [230])of the four Campylobacter species from QHSS (C. coli from chicken (QHSS P287) and C. jejuni (QHSS 00M2128, human source), C. upsaliensis (QHSS 99M126, human source) and C. hyointestinalis (QHSS 99M2318, human source) compared with database sequences of C. hyointestinalis (ATCC 35217, swine source) C. coli (CCUG 11283, porcine source), C. jejuni (CCUG 24567, human source ) and C. upsaliensis (CCUG 14913, from Canine faeces). .................................................................66 Figure 3.2 An alignment of the partial 16S rDNA sequences corresponding to position 783 to 1464 (E. coli numbering [230] of representative of Campylobacter species is shown, this region is amplified using primers F2 and Cam-Rev (indicated by left and right hand arrows) to produce a 681 bp amplicon. The universal fluoroprobe Cy5+1046 (specific for members of domain Bacteria) and the probe Jejuni-coli (specific to C. coli, C. jejuni and C. lari) bind adjacent to each other within the target regions of the amplicon. ...........................................................................................................67 Figure 3.3 An increase in fluorescence during specific adjacent hybridisation of the fluoroprobe 6FAM Jejuni-coli and fluoroprobe Cy5 1046+ to the target sites of the 16S rDNA amplicons of C. jejuni (ATCC 940565) (-■-), C. lari (ATCC 35223) (-●-) and C. coli (NCTC 11366) (-▲-) during PCR as measured in the LightCyclerTM. No increase was observed for C. hyointestinalis (QHSS 99M2318) (-×-), C. upsaliensis (QHSS 99M126) (-X-), E. coli (- -) or a negative control lacking a DNA template (-+-). The DNA templates were prepared using the CTAB method...........................68 Figure 3. 4. The melting peaks generated from the dissociation of the fluoroprobes from C. coli (NCTC 11366) C. lari (ATCC 35223) and C. jejuni (ATCC 940565) amplicon target sites at the end of run of real-time PCR produced a Tm of 65oC. But as expected a Tm for C. hyointestinalis, C. upsaliensis, E. coli and a negative control lacking template was not produced. ....................................................68 Figure 3.5 Agarose gel electrophoresis of PCR reaction products from C. upsaliensis (Lane 1), C. hyointestinalis (Lane 2), E. coli (Lane 3), C. coli (Lane 5), C. jejuni (Lane 6) C. lari (Lane 7) and a negative control lacking DNA template (Lane 8). Only C. coli C. lari and C. jejuni but none of the others produced the amplicons of 681 bp as expected. Molecular weight marker (100 bp) (Promega, USA) was used for base pair size comparison (Lane 4). ....................................................................69 XVIII

Figure 3.6 Real-time detection of C. jejuni CTAB-purified DNA at different concentrations. 192 ng/μl (-■-), 19.2 ng/μl (-▲-), 1.92 ng/μl (-+-), 192 pg/μl (-●-), 19.2 pg/μl (-×-), 1.92 pg/μl (-∆-) and 192 fg/μl (-X-)....................................................................................................................................71 Figure 3.7 Real-time PCR analysis of 10-fold serial dilutions of C. jejuni genomic DNA isolated using CTAB methods. Quantitation was performed by determining the cycle threshold (Ct). Ct values were then plotted against the log10 of DNA concentration. The dilution series ranges from 192 fg/μl to 192 ng/μl of C. jejuni DNA. Two microliters was used for each PCR reaction. The diagram shows one representative experiment. The straight line, which was calculated by linear regression [Ct =-0.9927 log10 (DNA concentration)+ 20.648], shows an R2 of 0.966. (The experiments were repeated three times with identical results.)...............................................................................................................72 Figure 3.8 Agarose gel electrophoresis of different DNA concentrations 192 ng/μl (Lane 1), 19.2 ng/μl (Lane 2), 1.92 ng/μl (Lane 3), 192 pg/μl (Lane 4), 19.2 pg/μl (Lane 5), 7 1.92 pg/μl (Lane 6), and 192 fg/μl (Lane 7) shows an expected 681 bp amplicons. The DNA 100 bp ladder was used as molecular markers (Promega, Australia) (Lane 8). ............................................................................72 Figure 3.9 A single colony of C. jejuni (ATCC 940565) was serially diluted, DNA extracted by the rapid boiling method and used in real-time adjacent hybridisation assays. The numbers of cells in each dilution was determined by plating onto BAPs. 500000 cells (-▲-), 50000 cells (- -), 5000 cells (-×), 500 cells (-∆-), 125 cells (-+-) and Negative control lacking template (-♦-). .................................74 Figure 3.10 A Standard curve for C. jejuni genomic DNA for 16S rRNA quantitative real-time PCR using adjacent hybridisation probes. Ct values were plotted against log10 cell numbers. The straight line, which was calculated by linear regression [Ct =-1.7022 log (cells number)+ 36.41], shows an R2 of the culture of C. jejuni serially diluted 10-fold is 0.9994. One microliter of DNA was used in a 10 μl PCR reaction. (The experiments were repeated three times with identical results.)...........................74 Figure 3.11 Gel analysis of PCR amplification products with one colony serial dilution of C. jejuni derived from 16S rRNA primer set F2 and R5. 500000 cells (Lane 1), 50000 cells (Lane 2), 5000 cells (Lane 3), 500 cells (Lane 4), 110 cells (Lane 5), and negative control lacking template (Lane7). A 100 DNA ladder was used for size evaluation (Promega, Australia) (Lane 6). ..............................75 Figure 3.12 C. jejuni (ATCC 940565) was grown for periods of up to 24 hours, DNA was extracted by the rapid boiling method and used in real-time adjacent hybridisation assay. 24 hours incubation (-♦), 8 hours (-■-), 6 hours (-▲-), 4 hours (-●-), 2 hours (-□-) and 0 hours (-●-). ..................................76 Figure 3.13 Real-time PCR detection of C. coli during different growth phases. 24 hours incubation (-♦-), 8 hours (-■-), 6 hours (-▲-), 4 hours (-*-), 2 hours (-□-) and 0 hours (-●-).......................................76 Figure 3.14 Real-time PCR of different DNA concentrations of C. jejuni prepared by the CTAB method and after the reduction of the amplification time to 22 minutes. DNA concentrations were 19.2 ng/μl (-♦-), 1.92 ng/μl (-■-), 192 pg/μl (-▲-) and 19.2 pg/μl (-●-). ............................................................77 Figure 4.1 Alignment of partial hipO gene sequences of C. jejuni (GenBank Accession numbers Z36940and AL139076) showing the forward primer, Hip-2214F (5’GTTGTTGCACCAGTGACTATGA-3’) and reverse primer, Hip-2474R (5’AGCTAGCTTCGCATAATAACTTG-3’) used in the development of real-time PCR. ...................84 Figure 4.2 Real-time PCR SYBR Green I assay with CTAB-purified DNA for the detection of the hipO gene. C. jejuni (ATCC 940565) (-Q-), C. coli (NCTC 11366) (-V-), C. upsaliensis (QHSS 99M126) (-□-), C. hyointestinalis (QHSS 99M2318) (-O-) and negative control lacking template (♦-) for the detection of hipO gene......................................................................................................86 Figure 4.3 The melting peaks generated from the dissociation of the fluoroprobes from C. jejuni (ATCC 940565) amplicon target sites at the end of a run of real-time PCR produces a Tm of 85oC and a Tm of 56 °C for C. coli, C. hyointestinalis, C. upsaliensis and negative control lacking template. 86 Figure 4.4 Agarose gel electrophoresis of the PCR products from real-time SYBR Green I assay showing the expected 260 bp specific product for C. jejuni (Lane 1). No products were generated from C. coli (Lane 2), C. hyointestinalis (Lane 3), C. upsaliensis (Lane 4) and a negative control lacking DNA template (Lane 5). A DNA 100 bp ladder was used for size comparison. ...................87 Figure 4.5 Real-time PCR SYBR Green I assay for the detection of C. jejuni hipO using an iCycler iQTM. CTAB-purified DNA from C. jejuni (ATCC 940565), C. coli (NCTC 11366), C upsaliensis (QHSS 99M126) and C. hyointestinalis (QHSS 99M2318) were used as templates. ........................88 Figure 4.6. The Tm generated from the dissociation of SYBR Green I from the PCR products during the gradual increase in temperature from 50°C to 94°C. A specific 260 bp product with a Tm of 85 C and a Tm of 56°C for C. jejuni (ATCC 940565), C. coli (NCTC 11366), C. upsaliensis (QHSS 99M126) and C. hyointestinalis (QHSS 99M2318). ..........................................................................89 Figure 4.7 Agarose gel electrophoresis of the PCR products from real-time SYBR Green I assay using iCycler iQTM showing the expected 260 bp specific product for C. jejuni (Lane 1) and no products generated from C. coli (Lane 2), C. hyointestinalis (Lane 3) C. upsaliensis (Lane 4) and No Template (Lane 5). A 100 bp DNA ladder is shown in Lane 6 (Promega, Australia)........................89 XIX

Figure 5.1 The partial sequences of 16S rDNA (positions 783 to 1120, E. coli numbering according to Brosius et al. 1978 [230]) of the three Arcobacter species from QHSS and NCTC (A. skirrowii (NCTC 12713, lamb source) and A. nitrofigilis (NCTC 12251, Alterniflora roots). A. butzleri (QHSS 99M3958 human source) compared with database sequences of A. skirrowii (CCGU 10374, lamb source), A. nitrofigilis (CCUG 15893, Alterniflora roots) and A. butzleri (CCUG 10373 human source). 96 Figure 5.2 The partial sequences of 16S rRNA (positions 783 to 1120, E. coli numbering [259]) from representative isolates of Arcobacter species. Boxes represent the universal bacterial forward (F2) and reverse (R5) primers, the universal fluoroprobe Cy5+1046 specific for members of domain Bacteria and fluoroprobe 6-FAM specific for detecting A. butzleri, A. skirrowii, A. nitrofigilis and A. cryaerophilius.....................................................................................................................................98 Figure 5.3 The increase in fluorescence during specific hybridisation of probes; probe Butz, probe Skir-Cry and Universal Cy5 to the target site in the 16S rRNA of A. butzleri (-■-), A. skirrowii (-▲), A. nitrofigilis (-+-), C. coli (-●-), C. jejuni (-♦-) and no template (-∆-). Suring PCR as measured in the LightCyclerTM. Purified DNA was prepared using CTAB method. Amplicons produced were run on an agarose gel (Figure 5.5) and the Tm determined (Figure 5.4). ...............................................101 Figure 5.4 The melting peaks generated from the dissociation of the fluoroprobes from A. butzleri, A. skirrowii and A. nitrofigilis amplicon target sites at the end of real-time PCR produces Tm's of 63oC for A. skirrowii, 65oC for A. nitrofigilis and of 67°C for A. butzleri but as expected the probe did not hybridise to the amplicons of C. coli, C. jejuni, E. coli and no template (Negative control), therefore no Tm was produced..........................................................................................................................101 Figure 5.5 Agarose gel electrophoresis of PCR products from no DNA template (Lane 2), A. butzleri (Lane 3), A. skirrowii (Lane 4), A. nitrofigilis (Lane 5), C. coli (Lane 6) and C. jejuni (Lane 7). A 100 bp DNA Ladder is in Lane 1 (Promega, Australia). ..................................................................102 Figure 5.6 Agarose gel electrophoresis of different DNA concentrations 113 ng/μl ((Lane 2), 11.3 ng/μl (Lane 3), 1.13 ng/μl (Lane 4), 113 pg/μl (Lane 5), 11.3 pg/μl (Lane 6), 1.13 pg/μl (Lane 7), 113fg/μl (Lane 8) and 100 bp DNA ladder (Promega, Australia) (Lane 1)......................................104 Figure 5.7 Real-time detection of A. butzleri CTAB-purified DNA at different concentrations. 113 ng/μl (-χ-), 11.3 ng/μl (-■-), 1.13 ng/μl (-▲-), 113 pg/μl (-●-), 11.3 pg/μl (-+-), 1.13 pg/μl (-X-) and 113 fg/μl (-∆-)...................................................................................................................................104 Figure 5.8 A Standard curve for A. butzleri genomic DNA for 16S rRNA quantitative real-time PCR using adjacent hybridisation probes. The DNA was serially diluted 10-fold. Ct values were plotted against the log10 DNA concentration and the equation and linear correlation coefficient (R2) determined in Microsoft Excel. The straight line, which was calculated by linear regression [Ct =1.1323 log (DNA concentration) + 20.648], shows an R2 of 0.9825. One microliter of standard DNA was used in a 10 μl PCR reaction. (The experiments were repeated three times with identical results.) 105 Figure 5.9 A representative real-time PCR analysis of a dilution series of A. butzleri culture ranging from 440000 to 110 cells. One microliter was used for each PCR reaction. The increase in fluorescence signal related to cycle number is shown. 440000 cells (-■-), 44000 cells (-●-), 4400 cells (-▲-), 440 cells (-ж-) and 110 cells (-♦-). ...............................................................................106 Figure 5.10 A Standard curve for A. butzleri genomic DNA for 16S rRNA quantitative real-time PCR using adjacent hybridisation probes. One colony was serially diluted 10-fold. Ct values were plotted against log10 cell numbers. The straight line, which was calculated by linear regression [Ct =-1.0893 log10 (concentration)+ 36.339], shows an R2 of 0.9944. One microliter of DNA was used in a 10 μl PCR reaction. (The experiments were repeated three times with identical results.).........................106 Figure 5.11 A. butzleri was grown for periods of up to 24 h, DNA was extracted by the rapid boiling method and used in real-time PCR adjacent hybridisation probes assays. (- -) 0, (-▲-) 2, (-x-) 4, (♦-) 6 hours, (-Δ-) 8 hours and (-■-) 24 hours...................................................................................107 Figure 6.1 Real-time detection of Campylobacter species and Arcobacter species using probe SkirCry, probe Butz and probe Jejuni-coli, C. jejuni (QHSS 00M2260) (- -), C. coli (P287/96) (-▲-), A. butzleri (ATCC 12481) (-X-), A. skirrowii (ATCC 12713) (-∆-), C. upsaliensis (QHSS 99M126), (■-) and a negative control lacking template (-♦-). ...........................................................................115 Figure 6.2 The melting peaks generated from the dissociation of the fluoroprobes from A. butzleri (NCTC 12481) (Tm 67 oC), A. skirrowii (NCTC 12713) (Tm 63 oC), C. jejuni (ATCC 940565) (Tm 65 o C) and C. coli (NCTC 11366) (Tm 65 oC), while E. coli, C. upsaliensis, and C. fetus and negative control lacking template DNA does not produce a Tm. ....................................................................115 Figure 6.3 Agarose gel electrophoresis of PCR products from a negative control lacking DNA template (Lane 1), A. butzleri (ATCC 12481) (Lane 2), A. skirrowii (ATCC 12713) (Lane 3), C. jejuni (QHSS 00M2260), (Lane 4), C. coli (QHSS P287/96) (Lane 5), C. upsaliensis (QHSS 00M2260) (Lane 6) and 100 DNA ladder (Lane 7)..........................................................................116

XX

Figure 7.1 The partial sequences of 16S rRNA (corresponding to position 936 to 1087, E. coli numbering [230]) from representative isolates of Campylobacter species. Boxes represent Campylobacter species forward primer (TaqCam-936F) and Campylobacter reverse primer (TaqCam-1087R), Campylobacter species-specific TaqMan probe (TaqCam-1034). .....................123 Figure 7.2 The partial sequences of C. jejuni and C. coli orfA gene. (A) The partial sequences of C. jejuni hipO gene (corresponding to position 1754 to 1924, C. jejuni numbering). Boxes represent C. jejuni forward primer (HipTaq-1754) and C. jejuni reverse primer (HipTaq-1924), C. jejuni specific TaqMan probe (TaqJej-1864). (B) The partial sequences of C. coli hipO gene (corresponding to position 4024 to 4129, C. coli numbering). Boxes represent C. coli forward primer (OrfATaq-4024) and C. coli reverse primer (OrfATaq-4109), C. coli specific TaqMan probe (TaqColi-4075).........124 Figure 7.3 Single tube TaqMan amplification of DNA extracted from Campylobacter isolates, using iCycler iQTM, with primers and probe for identifying and detecting Campylobacter 16S rRNA genes, C. jejuni hipO genes and C. coli orfA genes. The fluorescence of each sample during TaqMan hydrolysis was plotted against the PCR cycle number. C. jejuni (ATCC 940565) 16S rRNA gene (♦-), C. coli (NCTC 11366) 16S rRNA gene (-▲-), C. hyointestinalis (QHSS 99M2318) 16S rRNA gene (-X-) and C. upsaliensis (QHSS 99M126) 16S rRNA gene (-□-), C. jejuni (ATCC 940565) hipO gene (-●-), and C. coli (NCTC 11366) orfA gene (-■-) and negative control lacking template (∆-). Similar results were obtained with a further 206 samples (see appendices 1and 10). ...............128 Figure 7.4 Multiplex TaqMan amplification of DNA extracted from representative Campylobacter isolates, using iCycler iQTM, with primers and probes for identifying and detecting Campylobacter 16S rRNA genes, C. jejuni hipO genes and C. coli orfA genes. The fluorescence of each sample during TaqMan hydrolysis was plotted against the PCR cycle number. Tube 1 contains one set of primers and a probe to simultaneously detect C. hyointestinalis (QHSS 99M2318) 16S rRNA genes, (-□-). Tube 2 contains one set of primers and a probe to simultaneously detect C. upsaliensis (QHSS 99M126) 16S rRNA genes (-◊-). Tube 3 contains two sets of primers to simultaneously detect C. coli (NCTC 11366) orfA genes (-♦-), and C. coli (NCTC 11366) 16S rRNA genes (-●-). Tube 4 contains three sets of primers and probes to simultaneously detect C. jejuni (ATCC 940565) hipO genes (-▲-), C. coli (NCTC 11366) orfA genes (-∆-) and C. hyointestinalis (QHSS 99M2318) 16S rRNA genes (-■-). Similar results were obtained with a further 206 samples (see appendices 1 and 10). 128 Figure 7.5 A representative agarose gel electrophoresis of multiplex real-time PCR (TaqMan probes) of DNA extracted from Campylobacter isolates, using iCycler iQTM, with primers and probes for identifying and detecting Campylobacter 16S rRNA genes, C. jejuni hipO genes and C. coli orfA genes. (Figure 7.4) The assay generated three amplicons 105, 151 and 170 bp, representing C. coli, Campylobacter species and C. jejuni respectively. Lanes 3 to 8 shows three PCR products generated from a single tube containing three sets of primers and probes. Lane 2 shows PCR product generated from one tube containing two sets of primers and probes. Lanes 9 and 10 shows PCR products generated from a single tube containing one set of primers and a probe. Similar results were obtained with a further 206 samples (see appendices 1 and 10)......................................................................129 Figure 7.6 Sensitivity of the Campylobacter 16S rRNA gene probe and primers set (Section 7.3.1) in detecting C. jejuni. Purified C. jejuni DNA was used as the template in different concentrations of 156 ng/μl (-▲-), 15.6 ng/μl (-+-), 1.56 ng/μl (-×-), 156 pg/μl (-Χ-), 15.6 pg/μl (-■-), 1.56 pg/μl (-∆-), and 156 fg/μl (-♦-), representing Ct values in the range 13-26. .......................................................130 Figure 7.7 A Standard curve for 16S rRNA C. jejuni genomic DNA quantitative real-time PCR using TaqMan probes. The DNA was serially diluted 10-fold. The Ct values were plotted against log10 DNA concentration and the equation and R2 determined in Microsoft Excel. The straight line, which was calculated by linear regression [Ct =-1.0392 log (DNA concentration)+ 18.498], shows an R2 of 0.9675. (The experiments were repeated three times with identical results.) ...................................130 Figure 7.8 Simultaneous detection and amplification of the 10-fold serial dilution of purified C. coli DNA. Detection was done using a sequence-specific TaqMan probe, targeting C. coli orfA gene. Purified C. coli DNA was used as the template in concentrations of quantities of 142 ng/μl (-■-), 14.2 ng/μl (-▲-), 1.42 ng/μl (-+-), 142 pg/μl (-X-), 14.2 pg/μl (-●-), 1.42 pg/μl (∆--), 142 fg/μl (-♦-), representing Ct values in the range 15-29.........................................................................................131 Figure 7.9 A Standard curve for 10-fold serial dilutions of purified C. coli genomic DNA for orfA gene quantitative real-time PCR using TaqMan probes. Quantification was performed by determining Ct values. Ct values were then plotted against log10 DNA concentration and the equation and R2 determined in Microsoft Excel. The straight line, which was calculated by linear regression [Ct =0.9461 log (DNA concentration)+ 20.439], shows an R2 of 0.996. One microliter of DNA was used in a 20 μl PCR reaction. (The experiments were repeated three times with identical results.) ........131 Figure 7.10 Sensitivity of the C. jejuni hipO gene probe and primers set in detecting C. jejuni CTAB DNA. 10 –fold serially diluted purified C. jejuni DNA was used as the template in concentration of

XXI

156 ng/μl (-▲-), 15.6 ng/μl (-+-), 1.56 ng/μl (-×-), 156 pg/μl (-●-), 15.6 pg/μl (-∆-), 1.56 pg/μl (-■-), and 156 fg/μl (-♦-), representing Ct values in the range 15-30. .......................................................132 Figure 7.11 A Standard curve for C. jejuni genomic DNA for hipO gene quantitative real-time PCR using TaqMan probes. The DNA was serially diluted 10-fold. The Ct values were then plotted against log10 DNA concentration and the equation and R2 determined in Microsoft Excel. The straight line, which was calculated by linear regression [Ct =-1.1323 log (DNA concentration)+ 21.325], shows an R2 of 0.9825. One microliter of DNA was used in a 20 μl PCR reaction. (The experiments were repeated three times with identical results.) ....................................................................................................132 Figure 7.12 A single colony of C. jejuni (ATCC 940565) was serially diluted and DNA was extracted by the rapid boiling method and used in multiplex real-time PCR using primers and probes targeting hipO gene. The number of cells in each dilution was determined by plating onto BAPs 500000 cells (▲), 50000 cells (■), 5000 cells (♦), 500 cells (Δ), 125 cells (×) and 50 cells (●)...........................134 Figure 7.13 A standard curve showing the relation ship between Ct and log10 cells number for serialy dilution of one colony of C. jejuni. The straight line, which was calculated by linear regression [Ct =1.0934 log10 (Cell number)+ 33.713], shows an R2 of 0.9872. Two microliter of DNA was used in a 20 μl PCR reaction. ..........................................................................................................................134 Figure 8.1 Representative TAP sorted display windows, showing RDP organism identifiers, TRF sizes, organism names and restriction enzymes (DdeI, EcoRI and KpnI) used in this assay. Organisms are sorted alphabetically in ascending order by their RDP sequence identifiers. ...................................141 Figure 8.2 An alignment of the partial 16S rRNA sequences corresponding to position 49 to 536 (E. coli numbering according to [230]) of representatives of the Campylobacter and Arcobacter species is shown. This region is amplified using Fam-49F primer and R2 primer (indicated by left and right hand arrows) to produce a 460 bp amplicons. ..................................................................................144 Figure 8.3 Agarose gel electrophoresis of PCR products generated with Fam-49 forward and R2 reverse primers. C. coli (NCTC 11366) (Lane 2), C. jejuni (ATCC 940565) (Lane 3) C. upsaliensis (QHSS 99M126) (Lane 4), A. butzleri (NCTC 12481) (Lane 5), A. skirrowii (NCTC 12713) (Lane 6), shows an expected amplicon size of 462 bp, and the100 bp DNA ladder (Promega, Australia) (Lane 1). 145 Figure 8.4 A typical electropherogram of T-RFLP profiles obtained after DdeI restriction enzyme digestion of the 16S rRNA amplicons produced with primers FAM-49F and R2 and DNA of pure and selective enrichment cultures of Campylobacter and Arcobacter. A 53 bp TRF identifying a pure culture of A. butzleri (NCTC 12481) (Panel 8.4A) and a 246 bp TRF identifying C. jejuni (ATCC 940565) (Panel 8.4B) are shown. A profile of four TRFs representing 48, 72, 104 and 246 bp from DNA of a selective enrichment culture grown from the chicken sample (C-P1000) is shown in Panel 8.4C. The 48 bp TRF identifies C. hyointestinalis and/or C. fetus, the 72 bp TRF identifies Clostridium, and/or Lactobacillus species, and/or Mycoplasma species, the 104 bp TRF specifically detects C. sputrum, and the 246 bp TRF identifies C. jejuni. A similar complicated but different profile was observed when DNA from enrichment cultures initiated from chicken samples (C-P1006, and P10-A) was amplified and the amplicons digested with DdeI (Panel 8.4D). The 53 bp TRF identified Arcobacter species, the 136 bp TRF identified C. hyointestinalis subspecies hyointestinalis and the 246 bp TRF identified C. jejuni and/or C. lari and or C. coli. GeneScan®-500 (TAMRA) internal (Applied BioSystems, Australia) with a size range between 50 bp and 500 bp (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500 bp) was used as an internal standard (Panel 8.4E). .....................................................................................................................................153 Figure 8.5 A typical electropherogram of T-RFLP patterns obtained after Sau3AI restriction enzyme digestion of the 16S rRNA PCR products amplified with FAM-49F and R2 primers and the DNA of pure and selective enrichment cultures of Campylobacter and Arcobacter species. A profile of two TRFs representing 255 and 465 bp from DNA of pure culture is shown (Panel 8.5A). A 255 bp TRF identifying C. jejuni (ATCC 940565), and the 465 bp TRF represent the undigested PCR product respectively. A single 260 bp TRF identifying A. butzleri (NCTC. 12481) is shown (Panel 8.5B). Two distinctive TRFs representing a 255 and 260 bp TRF generated from a PCR products of mixed DNA templates from Campylobacter and Arcobacter enrichments (C-P1002 and P10-A) amplified using 16S rRNA gene FAM-49 and R2 primers pair is shown (Panel 8.5C). The 255 bp TRF identifies Campylobacter species, whereas the 260 identify Arcobacter species. GeneScan®-500 (TAMRA) internal standard (Applied BioSystems, Australia) with a size range between 50 bp and 500 bp (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500 bp) was used for TRF size comparison (Panel 8.5D)...........................................................................................................154 Figure 8.6 A representative electropherogram of T-RFLP patterns obtained after restriction endonuclease digestion with KpnI and EcoRI of the 16S rRNA PCR amplicons amplified with FAM-49F and R2 primers and the DNA of pure and enrichment cultures of Campylobacter and Arcobacter species. A 405 and 465 bp TRFs, generated from the digestion of the C. jejuni PCR amplicons with EcoRI is shown (Panel 8.6A). A 405 bp TRF identifies C. jejuni (ATCC 940565), while the 465 bp TRF represents the undigested PCR product. Further two TRFs representing 415 XXII

and 465 bp generated from the PCR products of C. jejuni (ATCC 940565) using KpnI is shown (Panel 8.6B). A complex T-RFLP profiles shows two distinctive TRF representing 405 and 415 bp, and two large size TRFs representing 460 and 465 bp TRF generated from the digestion of PCR amplicons of Campylobacter enrichment cultures (C-P1015) with both restriction endonucleases KpnI and EcoRI (Panel 8.6C). The 405 bp TRF identifies C. lari or C. jejuni or C. coli group, whereas the 415 bp identifies Campylobacter species. The remaining 460 and 465 bp TRFs represent the undigested PCR products. Panel 8.6D shows the GeneScan®-500 (TAMRA) internal standard (Applied BioSystems, Australia) (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500). 155 Figure 8.7 A typical electropherogram of T-RFLP patterns obtained after both BtsI and DdeI restriction endonuclease digestion of the 16S rRNA PCR products amplified with FAM-49F and R2 primers and the DNA of pure and enrichment cultures of Campylobacter and Arcobacter species. The presence of the 313 and 465 bp TRF generated from the digestion of the 16S rRNA PCR products of A. butzleri (NCTC 12481) using BtsI restriction endonuclease is shown (Panel 8.7A). The 313 bp TRF specifically detects and identifies A. butzleri (NCTC 12481), whereas the 465 bp TRF represents the undigested PCR products. A single 465 bp TRF representing the undigested PCR products, when BtsI is used to digest the PCR products of DNA from C. jejuni (ATCC 940565) (Panel 8.7B). A complex T-RFLP of 4 TRFs representing 53, 72, 242, 313 bp generated from digestion of the multiplex PCR amplicons of mixed DNA templates from Campylobacter and Arcobacter enrichments (P11-A and C-P1009) using BtsI and DdeI restriction enzymes is shown (Panel 8.7C). The 53 and 313 bp TRF identifies Arcobacter species, while the 242 bp TRF identifies either C. concisus or C. rectus. The 72 bp TRF represents the Clostridium species. Panel 8.7D shows the GeneScan®-500 (TAMRA) internal standard (Applied BioSystems, Australia) (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500). ...................................................................156 Figure 8.8 A representative electropherogram of T-RFLP patterns obtained after Sau3AI, EcoRI, KpnI, BtsI and DdeI restriction enzymes digestion of the multiplex PCR products amplified FAM49F and R2 targeting the 16S rRNA genes of the DNA extracted from the enrichment cultures (CP1017, and P12-A). A very complicated T-RFLP profile of seven TRFs representing 72, 260, 242, 255, 405, 415, and 465 bp is shown (Panel 8.8A). The 72 bp TRF could identify Clostridium species or any other bacterial species generated after digestion of the PCR amplicons with one of the restriction enzymes used in this study. The 313 and 260 bp TRF identifies Arcobacter species generated from the digestion of the PCR products with DdeI, BtsI and Sau3AI respectively. The 242 bp TRF represents C. rectus or C. concisus generated from the digestion of the PCR products using DdeI. The 405 bp TRF could identify C. coli or C. lari or C. jejuni generated from the digestion of the PCR products with EcoRI. The 255 and 415 bp TRFs identifying Campylobacter species, is generated from the digestion of PCR products with Sau3AI and KpnI respectively. Panel 8.8B shows the GeneScan®-500 (TAMRA) internal standard (Applied BioSystems, Australia) (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500). ...................................................................157 Figure 9.1: Schematic representation of LDR assay used for rapid detection and identification of Campylobacter and Arcobacter species. The discriminative probe corresponding to the Campylobacter or Arcobacter species is shown in green bold and has a G at its 3´ end. The common probe is shown in black bold. In this example, the template has been amplified from Campylobacter or Arcobacter species. Thus only the Campylobacter and Arcobacter species discriminative probe is ligated to the common probe, generating a single type of ligation product. The thermal cycling is repeated twenty times in order to generate sufficient ligation product for detection........................164 Figure 9.2 An alignment of the partial 16S rRNA sequences corresponding to position 783 to 1542 (E. coli numbering according to [291]) of representatives of Campylobacter and Arcobacter, this region is amplified using F2 16S rRNA forward primer and Rd1 16S rRNA revers primer (indicated by left- and right-hand arrows) to produce a 759 bp amplicon. The common probe LDRProteo (specific for members of Proteobacteria) and the LDR discriminative probes (LDRCam and LDRArco specific to Campylobacter and Arcobacter species respectively) bind adjacent to each other within the target regions of the amplicon. ...............................................................................168 Figure 9.3 An alignment of the partial FlaA sequences corresponding to position 7 to 649 representative of Campylobacter species is shown. This region is amplified using LDRFla-7 primer and LDRFla-630 primer (indicated by left- and right-hand arrows) to produce a 643 bp amplicon. The common probe LDRFla (specific for C. coli, C. jejuni and C. lari) and the LDR discriminative probes (LDRJejuni, LDRColi and LDRLari specific for C. jejuni, C. coli and C. lari respectively) hybridise adjacent to each other within the target regions of the amplicon......................................169 Figure 9.4 Amplification of the 16S rRNA gene of Campylobacter and Arcobacter species by PCR with the primers F2 and Rd1. The PCR products were electrophoresed in 1% agarose gel. C. coli (NCTC 11366) (Lane 2), C. jejuni (ATCC 940565) (Lane 3), C. hyointestinalis (QHSS 99M2318) (Lane 4), A. butzleri (NCTC 12481) (Lane 5), shows an expected amplicon size of 759 bp. A ladder of 100 bp molecular markers is shown (Promega, Australia) (Lane 11). .........................................170 XXIII

Figure 9.5 Agarose gel electrophoresis of flaA gene PCR amplicons produced with LDRFla forward and LDRFla reverse primer. C. coli (NCTC 11366) (Lane 1), C. jejuni (ATCC 940565) (Lane 2), chicken isolate enriched in Campylobacter enrichment media (Lanes 3 and 4,) shows an expected amplicon size of 644 bp. No products are generated from C. upsaliensis (QHSS 99M126) consistent with expectations (Lane 5). A 100 bp DNA ladder is shown (Promega, Australia) (Lane 11). .......171 Figure 9.6 A typical electropherogram of LDR patterns generated from the hybridisation and ligation of LDR probes to target regions of the amplicons of 16S rRNA genes (LDRCam, LDRArco, LDRProteo probe) for identifying Campylobacter and Arcobacter and target regions of the amplicons of FlaA genes (LDRJejuni, LDRColi, LDRLari and LDR-Fla) for identifying C. lari, C. coli and C. jejuni from the DNA of pure and enrichment cultures. When the 16S rRNA discrimination hybridisation and ligation system is used, a 43 bp identifies a pure culture of Arcobacter species (Panel 9.6A), while a 46 bp identifies a pure culture of Campylobacter species (Panel 9.6B). When the flaA gene discrimination hybridisation and ligation system is used, a 52 bp LDR product generated with the DNA of C. jejuni pure culture (panel 9.6C). The combined use of both 16S rRNA and flaA genes from pure and enrichments cultures generated 5 LDR products able to identify Campylobacter and Arcobacter species (Panels 9.6D and 9.6E). The 46 and 48 bp LDR products identify Campylobacter species using 16S rRNA and C. coli using flaA gene respectively (Panel 9.6D). A profile of three LDR products representing the sizes of 43, 46, and 48 bp generated from the hybridisation and ligation of the LDR probe to the multiplex PCR amplicons of 16S rRNA and flaA gene of DNA from Campylobacter and Arcobacter enrichments (C-P1019 and P22-A) is shown (Panel 9.6E). The 43 and 46 bp identifies Arcobacter and Campylobacter 16S rRNA gene respectively, whereas the 48 specifically identifies the C. jejuni flaA gene. Multiple LDR products generated from unpurified 16S rRNA PCR products of C. jejuni is shown (Panel 9.6F). A negative control shows no LDR products generated from E. coli (Panel 9.6G). Panel H shows the internal GeneScan standard. This figure has been redrawn to scale from the row data in Appendix 13. ......175 Figure 9.7 A schematic representation of the LDR assay, showing the identification of Campylobacter species and Arcobacter species using 16 S rRNA and detection of C. lari, C. coli and C. jejuni using FlaA gene based on the LDR product size of each genera or species...............................................176

XXIV

List of Tables Table 1.1 Pathogenicity genes of C. jejuni (a). ......................................................................................7 Table 1.2 Representative of real-time PCR Instruments available in the market. ..............................17 Table 1.3 Type of fluorophore probe technology suitable for molecular techniques. ................................19 Table 1.4 The maximum absorbance and quenching range for four different non-fluorescent quenchers currently available. .............................................................................................................................24 Table 2.1 Primers used for Amplification and Sequencing of 16Sr DNA..........................................47 Table 2.2 Show the iCycler iQTM filters excitation and emission wavelength for FAM, MAX and Cy5...54 Table 3.1 Biochemical tests used to distinguish between closely related C. jejuni, C. coli and C. lari according to the Australian standard methods....................................................................................63 Table 8.1 This table summarises the results shown in figure 8.4-8.8. It shows the TRF’s size of the five restriction enzymes used in this study to digest the PCR products of 16S rRNA gene amplified using FAM-49 and R2 primers as described in sections 2.8.1, 2.8.2 and 2.8.3. The TRFs size were determined by the ABI prism 377 as described in section 2.8.5. This table shows the TRFs resulting from the digestion of the PCR amplicons using a single restriction enzyme or a combination of these enzymes. ...........................................................................................................................................158 Table 8.2 Theoretical TRF size of the five restriction endonuclease enzymes used in this study, to digest 16S rRNA amplified using FAM-49Fand R2 primers, as described in Section 8.1.3, by TAP computer simulation and BioEdit. ....................................................................................................162

XXV

Abbreviations List ATCC

American Type Culture Collection Centre

BLAST

Basic Local Alignment Search Tool For Nucleic Acids

bp

Base Pair

BSA

Bovine Serum Albumin

CCD

Charge-Coupled Device

CFU

Colony Forming Unit

CTAB

Hexadecyltrimethyl Ammonium Bromide

dH2O

Deionised Water

ddH2O

Double Distilled Water

DNA

Deoxyribonucleic Acid

DATP

Deoxyadenosine Triphosphate

DCTP

Deoxycytidine Triphosphate

DGTP

Deoxyguanosine Triphosphate

DNTP

Deoxynucleoside Triphosphate

DSMZ

Deutsche Sammlung von Mikroorganismen und Zellkulturen

DTTP

Deoxythymidine Triphosphate

EDTA

Ethylenediamine-Tetraacetic Acid

EtBr

Ethydium Bromide

FAM

Fluorescein Isothiocyanate

FRET

Fluorescence Resonance Energy Transfer

XXVI

g

Gram

H

GT

h

Hours

K

ACT

kDa

Kilo Dalton

l

Litre

M

AC

M

Molar

MGB

Minor Groove Binder

Mol

Mole

MRNA

Messenger Ribonucleic Acid

MW

Molecular Weight

Mb

Mega bases

NCTC

National Collection of Type Cultures

Nt

Nucleotide

PCR

Polymerase Chain Reaction

QHSS

Queensland Health Scientific Services

RDNA

Ribosomal Deoxyribonucleic Acid

RFLP

Restriction Fragment Length Polymorphism

RNA

Ribonucleic Acid

RRNA

Ribosomal Ribonucleic Acid

XXVII

RNAse

Ribonuclease

ROX

Carboxy-X-rhodamine

S

CG

SSU

Small Sub Unit

Subsp.

Sub species

T

Type Strain

TAE

Tris/Acetate/Ethylenediamine Tetra-Acetic Acid (Buffer)

TE

Tris-EDTA

TRIS

Tris (hydroxymethyl) aminomethane

Vol

Volume

W

AT

Wt

Weight

Y

CT

XXVIII

1. Chapter 1 Introduction The genera Campylobacter and Arcobacter are microaerophiles and members of the order Campylobacterales; class Epsilon Proteobacteria of phylum Proteobacteria. They are Gram-negative and are natural inhabitants of the intestinal tracts of poultry and warm-blooded domestic animals where microaerophilic conditions and warm body temperatures are ideal conditions for their continuous growth [1-5]. Consequently, raw milk and under-cooked poultry products are the leading vehicles for outbreaks of related gastrointestinal diseases [6]. To date, a total of twenty species and six subspecies (Figure 1.1) of the genera Campylobacter and Arcobacter have been identified, all of which are presumed to be human gastrointestinal pathogens [7, 8]. Recent reports suggest that Campylobacter and Arcobacter species have the ability to co-infect their hosts at the same time and hence identifying and distinguishing them although very important is hampered by their low biochemical reactiveness [9]. 1.1. The Genus Campylobacter The characteristics of all members of the genus Campylobacter (Figure 1.1) are curved, spiral or S-shaped Gram-negative cells, with tapered ends [10, 11]. The cells measure 1.5-6.0 μm x 0.2-0.9 μm and are actively motile by an unsheathed polar flagellum located at one or both ends of the cell [12, 13]. Drying and freezing reduces their numbers in contaminated raw meat [14, 15]. In 1886, Escherich first observed organisms similar to Campylobacter species in stool samples of children with diarrhoea [16]. Later, Campylobacter was misidentified as “Vibrio fetus” and recognised as a pathogen associated with infertility and abortion in cattle and sheep by McFadyean and Stockman in 1913 [16]. Subsequently, a thermophilic vibrio-like bacterium (growth temperature of 42°C) was reported by King (1957) to be associated with acute human enteritis [16]. In an outbreak of acute enteritis amongst inmates of two adjacent prisons 1

following the consumption of contaminated raw milk, 355 inmates were hospitalised with symptoms of vomiting, abdominal cramps, diarrhoea, fever, headache and backache [17]. Microbiological investigation of the mucoid diarrhoeal specimen revealed vibrios [18]. With the development of identification techniques, these Vibriorelated bacterial isolates were transferred from the genus Vibrio to a newly created genus Campylobacter as C. jejuni [Campylobacter (Greek, curved rod) jejuni (Greek, maeger)] in 1963 [19, 20]. 1.1.1. Natural Reservoirs C. jejuni is a normal flora of birds and animals, which act as natural reservoirs. Campylobacter can be easily spread in animal and bird populations through a common water source or through contact with infected faeces. For example, when an infected bird is slaughtered, Campylobacter can be transferred from its intestines to the meat [21, 22]. It has been estimated that more than half of the raw chickens in the United States market are tainted with Campylobacter [23]. Campylobacter is also present in giblets, especially the liver [5]. Many chicken flocks are silent carriers of Campylobacter species [24]. Milk from Campylobacter infected cows can be contaminated via the udder or via splashing of manure [25, 26]. Although rodents and pets, including dogs, cats and birds, harbour and transmit Campylobacter to humans, they are normally harmless in such hosts [27, 28]. Lindblom et al. (1995) [29] reported that many domestic pets are colonised with C. upsaliensis, a familiar Campylobacter species isolated from children. Full appreciation of Campylobacter species as human pathogens was only recognised after routine laboratory methods (selective cultures and biochemical tests) for their isolation and detection were developed and incidence reporting had begun [30].

2

Figure 1.1 Dendrogram showing relationship of Campylobacter (16 species and 6 sub species) and Arcobacter (4 species) within the class Epsilon Proteobacteria, phylum Proteobacteria. Sequences were extracted from Ribosomal Database Project (RDPII, version 8.1) [31] and GenBank (http://www3.ncbi.nlm.nih.gov/BankIt/), aligned using BioEdit (Version 5.0.6) and the dendrogram constructed from the comparison of 828 unambiguous nucleotides using TreeCon [32]. (ATCC = American Type Culture Collection), CCUG (Culture Collection, University of Gutenborg, Sweden), NCTC (National Collection of Type Cultures) CIP (Collection de institute Pasteur). Type cultures T are shown in the brackets.

1.1.2. Epidemiology and Transmission Campylobacter was the most common bacterial disease causing 46% of all diarrhoeal illness of laboratory confirmed cases of bacterial gastroenteritis, reported by the Centre for Disease Control and Prevention (CDC) U.S.A. in 1996, as compared to 28% for Salmonella, 17% for Shigella and 5% for Escherichia coli O157 (Figure 1.2) [5]. These figures are undoubtedly underestimated, as many cases are not reported because the infection may not require treatment, or it may go undiagnosed. The true estimate of 3

human campylobacteriosis (illness ranging from loose stools to dysentery) in the U.S.A. is thought to be as high as 2.1-2.4 million [33]. Infection with Campylobacter occurs much more frequently in summer than in winter [2, 34].

Figure 1.2. Cases of Campylobacter and other food-borne infections by month of specimen collection; Centre for Disease Control and Prevention/U.S. Department of Agriculture/Food and Drug Administration Collaborating Sites Food borne Disease Active Surveillance Network, 1996 [5].

Most Campylobacter related human illnesses are caused by C. jejuni and C. coli (95%), with the remaining 5% caused by other Campylobacter species such as C. lari, C. hyointestinalis, C. upsaliensis and C. fetus [35-38]. In addition, C. rectus, C. gracilis, and C. showae have been implicated in human periodontal disease, whereas three cases of extraoral abscess have been reported to be caused by C. rectus or C. curvus [39]. Low doses of Campylobacter can cause gastrointestinal illness. For example, waterborne outbreak data shows that doses as low as 500 cells can induce illness [37, 38, 40]. The potential of C. jejuni infection is higher in AIDS and immunocompromised patients than in the general population. For example, in Los Angeles County between 1983 and 1987, campylobacteriosis cases was reported in 519 patients per 100,000 of population in patients with AIDS (Acquired Immunodeficiency Syndrome), 39 times higher than

4

the rate in the crude population [37, 41]. In developed countries, C. jejuni infection is extremely common in early childhood and five to ten episodes may occur in the first two years of a child’s life [42]. In most cases, the duration of the illness is very short, usually three to ten days even without antibiotic treatment, as most of the cases selfrecover [12]. Campylobacter are isolated from infants and young adults (15 to 25 years) more frequently than from other age groups, and from males more frequently than females [37]. Although Campylobacter illness is not considered lethal, it has been reported that around 100 to 730 persons die due to Campylobacter infections each year in the U.S.A [43-46]. The symptoms of Campylobacter infections are fever, abdominal cramping and diarrhoea with or without blood, which normally remain from several days to more than one week [46]. It is also known that Campylobacter infections mimic appendicitis symptoms and misdiagnosis can lead to unnecessary surgical removal of the appendix [5]. On occasions, the infection may be asymptomatic. 1.1.3. Mechanisms of Pathogenesis Animal model studies have indicated that Campylobacter species invade the mucosal surface of the lower gastrointestinal tract by their inherent property of motility, which allows them to move easily in this viscous environment to cause pathogenesis [47]. The flagellum is thought to facilitate the binding of C. jejuni and other Campylobacter species to epithelial cells [48]. Some people may develop autoimmune diseases such as, Reiter’s

syndrome,

Miller

Fisher

syndrome

or

reactive

arthritis

following

campylobacteriosis [49], whereas others may develop a rare disease called GuillainBarré Syndrome (GBS) that affects the nerves of the body, beginning several weeks after the diarrhoeal illness, and that can lead to paralysis that can last for up to several weeks [50]. GBS results from deamyelination of peripheral nerves, and has long been considered to

be

immunologically

mediated 5

[51]. Approximately

0.1% of

campylobacteriosis cases develop into GBS after two to three weeks of infection [50]. These results indicate that the patient immune response rather than C. jejuni acute toxicity may be responsible for GBS [51, 52]. 1.1.4. Molecular Biology Nuijten et al. (1990) [53] constructed a preliminary physical map of the C. jejuni UA580 genome and determined C. coli and C. jejuni genome sizes to be 1.7 Mb. All the remaining Campylobacter species have genome sizes ranging between 1.6 Mb and 1.7 Mb with the exception of C. upsaliensis, which has an estimated size of 2 Mb [13, 54]}. All Campylobacter genomes are Adenine and Thymine (AT) rich with Guanine and Cytosine (GC) contents of around 30%. Multiple copies of 16S rRNA genes (3-7 copies) have been described for the genus Campylobacter [53]. The C. jejuni (NCTC 11168) genome has been completely sequenced [55]. The circular chromosome has a G+C content of 30.6%, and consists of 1,641,481 base pairs (bp) predicted to encode 1,654 proteins and 54 stable RNA species [55]. Several genes for C. jejuni pathogenicity have been identified (Table 1.1). 1.1.5. Proteomes The study of proteomes of eukaryotes and prokaryotes has the advantage that it provides data on expressed proteins and therefore knowledge of their functions. The data from proteome studies can supplement and complement genome sequence information and the combined data can significantly advance knowledge on cellular processes. However, as proteome tools are still being developed, limited proteomic studies have been undertaken (http://www.proteome.org.au/). A partial proteome of C. jejuni is currently being

developed

and

is

available

http://www.eabiotech.com/2dgel2_images.htm.

6

at

the

URL

Table 1.1

Pathogenicity genes of C. jejuni (a).

Gene Name

GenBank Accession

Function in Pathogenesis

CdtC

Cj0077c

Cytolethal distending toxin

CdtB

Cj0078c

Cytolethal distending toxin

CdtA

Cj0079c

Cytolethal distending toxin

-----

Cj0183

Putative integral membrane protein with

-----

Cj0223

haemolysin domain Pseudogene (IgA protease family)

TlyA

Cj0588

Putative haemolysin

MrsA

Cj0637c

Putative peptide methionine sulfoxide

-----

Cj0914c

reductase CiaB protein

-----

Cj1279c

Putative fibronectin domain-containing

Cj1349c

lipoprotein Possible fibronectin / fibrinogen-binding

PldA

Cj1351

protein Phospholipase A

-----

Cj1371

Putative periplasmic protein (vacJ

CadF

Cj1478c

homolog) Outer membrane fibronectin-binding

-----

protein (a)

Source: http://www.sanger.ac.uk/Projects/C_jejuni/functional_classes/4.I.shtml

7

1.2. The Genus Arcobacter Arcobacter species were identified as Campylobacter until 1977 when Ellis et al. described them as a taxon that contained gram-negative, spirillum-like bacteria isolated from bovine and porcine fetuses [56]. Currently Arcobacter is differentiated into four species: A. butzleri, A. cryaerophilus, A. skirrowii, and A. nitrofigilis [13, 54, 57, 60, 61]. All members of the genus Arcobacter (Figure 1.1) are curved to slightly curved, Sshaped or helical, non-spore forming Gram-negative rods which may appear as spiral (0.2 to 0.9 μm wide and 1 to 3 μm long). They are actively motile by an unsheathed polar flagellum located at one or both ends of the cell [13, 54, 57, 60, 61]. They are microaerophiles and can grow aerobically and anaerobically at temperatures between 15οC and 30οC [62]. They also show variable growth when cultured at temperatures between 37οC and 42οC [63]. They cannot tolerate drying or freezing, which reduces their numbers in raw meat [64]. Most Arcobacter species are non-haemolytic except for A. skirrowii, and some A. butzleri strains which can be alpha haemolytic [65]. 1.2.1. Natural Reservoirs Arcobacter species were first isolated from bovine fetuses by Ellis et al. in 1977 [56]. Since then Arcobacter have been isolated from asymptomatic carrier pigs, from aborted and normal foetuses and those with reproductive problems [66-68], but have been taxonomically misidentified as Campylobacter. A. butzleri, A. cryaerophilus and A. skirrowii have been implicated in human diarrhoeal illnesses and bacteremia [58, 66, 69-72]. They have also been frequently associated with animal diseases including abortion, mastitis, and enteritis and from food of animal origin and water [63, 69, 73]. They have also been isolated from faecal samples of clinically healthy cows [74]. A. nitrofigilis is a nitrogen fixing bacterium associated with the roots of the salt marsh plant, Spartina alterniflora [63, 73, 75].

8

Recently Arcobacter species have been isolated from symptomatic and asymptomatic chickens and other poultry, but eggs do not seem to be carry the organisms [76]. In France, A. butzleri was isolated from 81% of chicken carcasses from five different processing plants and another study in Canada produced a 97% isolation rate [73]. A. skirrowii and A. cryaerophilus were also isolated from chicken abattoirs as well as from retail sale but with less frequency [63]. A survey of mechanically separated turkey samples has suggested that this meat was significantly contaminated with Arcobacter species [77]. Despite its common isolation from chicken and other poultry carcasses, A. butzleri is also occasionally isolated from caecal samples suggesting post-slaughter contamination [76]. Natural infection of live chicken and other poultry has also been reported recently [78-80]. Water probably also plays a major role in transmission of Arcobacter species to animals and humans. For example, A. butzleri has been isolated from canal and well water [81]. From these limited studies, it appears that the primary habitat of Arcobacter is largely unknown. 1.2.2. Pathogenesis and Transmission Arcobacter species can cause animal diseases such as abortion, diarrhoea and mastitis [82]. Arcobacter species have been isolated from patients with diarrhoea [83, 84] and from patients with bacteremia [79]. Tee et al. (1988) [66] reported the first case of A. cryaerophilus 1B infection in a homosexual male with diarrhoea. Arcobacter clinical isolates have now been reported in children with diarrhoea [85]. In addition, utero infections due to Arcobacter species have also been suggested [79]. The isolation of Arcobacter species from human and food animals has lead to speculation that these are likely to be zoonotic disease causing bacteria [74]. To date, the definitive relationship between human diseases and Arcobacter remains uncertain. In humans, contact with contaminated water may be the main cause of the infection, but person-to-person 9

transmission was confirmed in an outbreak in an Italian nursery school using a PCRmediated DNA fingerprinting technique [74]. 1.2.3. Culturing of Arcobacter Arcobacter was first isolated using a Leptospira medium but subsequently modified Campylobacter protocols have been used. Currently, there are no standard methods for isolating Arcobacter, but several methods/media are popular. A two stepped culture approach, which consists of a pre-enrichment step followed by isolation using a filter membrane method, is popular for the isolation of Arcobacter species. A pre-enrichment stage in Arcobacter broth CM965 (Oxoid) supplemented with either CAT selective supplement SR174E (cefoperazone 8 mg l teicoplanin 4 mg l 32 mg l

1

1

1

, amphotericin B 10 mg l

1

,

) (Oxoid) or CCDA-selective supplement SR155 (cefoperazone

, amphotericin B 10 mg l

1

) (Oxoid) , followed by isolation using a filter

membrane method has been routinely used [85]. For the latter, a serially diluted enrichment is filtered through a 0.45 μm filter membrane [86] and the membrane placed on mCCDA (modified cefoperazone, charcoal, deoxycholate agar) plates or CAT agar plates. The plates are incubated at 37οC for 24 hours [87, 88]. In other cases, the addition of piperacillin to Arcobacter selective broth has been suggested to prevent the growth of Pseudomonas species [65, 87, 89]. Recently, new commercial enrichment broths namely LabM Campylobacter enrichment broth (LEM) and Arcobacter Basel Medium (ABM) have also been formulated as enrichment media for Arcobacter isolation. So far, there are no universally acceptable methods for the recovery of Arcobacter from clinical samples.

10

1.3. Methods of Identification and Differentiation of Campylobacter and Arcobacter Arcobacter species are closely related to Campylobacter species, in that they do not ferment carbohydrates but hydrolyse indoxyl acetate, and show Campylobacter-like motility characteristics that are also found in C. coli, C. jejuni and C. upsaliensis [86, 90]. In addition, some Arcobacter species have fastidious growth requirements and because they are relatively biochemically inert and morphologically similar to Campylobacter species, correct identification is often difficult or impossible by routine methods [88]. Many different methods have been used to identify and differentiate Campylobacter and Arcobacter species. These techniques include phenotypic and genotypic methods such as the use of selective media, biochemical tests, DNA-DNA hybridisation, gene sequencing,

PCR

and,

more

recently,

MALDI

(Matrix

Assisted

Laser

Desorption/Ionization) / mass spectrometry of proteins . 1.3.1. Phenotypic Characterisation The routine clinical diagnosis of Campylobacter and Arcobacter from gastrointestinal infections is based on a presumptive identification, isolation from stool samples and identification from cultures. Paradise (1996) [48] noted that some laboratories firstly examine wet mounts of all diarrhoeal stools for polymorphonuclear leucocytes as well as examine for the presence of bacterial forms suggestive of Campylobacter species. Paradise contends that it is unlikely that Campylobacter species will be recovered in clinically significant numbers in stool specimens in the absence of polymorphonuclear leucocytes, and hence these specimens do not proceed further [48]. The appearance of colonies on Campylobacter selective media at 42οC indicates the presence of a thermophilic Campylobacter species. The identification is usually further confirmed by performing rapid catalase and cytochrome oxidase tests, which are 11

positive for C. jejuni, C. coli and C. lari. Further biochemical tests (e.g., hippurate hydrolysis) are performed to identify species such as C. jejuni [91]. Whilst the above methods are simple to perform, Campylobacter are asaccharolytic, fastidious bacteria, thus limiting the phenotypic tests by which isolates may be detected and differentiated [92]. Isolation and identification of Campylobacter is laborious, expensive, and time consuming, requiring up to four days. Furthermore, identification of the Campylobacter to species level is hindered by some technical limitations including variations in methodology and subjective interpretation of biochemical test results [93]. There are also non-culturable isolates of Campylobacter, called viable non-culturable (VNC) Campylobacter [9, 13]. Furthermore, atypical phenotypes of Campylobacter isolates may cause the mis-identification of the species. Linton et al. (1997) [91] demonstrated that the detection and differentiation of C. jejuni from C. coli depends on the hippurate hydrolysis by C. jejuni, but certain atypical isolates fail to hydrolyse. Eyers et al. (1993) [94] have suggested that only 90% of the hippurate hydrolysis tests are accurate. 1.3.2. Molecular Methods for the Pathogen Detection Biochemical tests depend on biochemical pathways but their disruption, for example due to gene mutations, can lead to products, not being synthesised, leading to negative (false) results. In other cases, the biochemical reaction may be weak and therefore may not be detectable and hence give a negative (false) result. It is been suggested that it would be more accurate and consistently reproducible to detect the gene that encodes for a specific biochemical phenotype rather than detecting the phenotype [95]. For example, it is more accurate to detect Campylobacter hippurate hydrolysis genes than to test the ability of Campylobacter species to hydrolyse hippurate. Hippurate hydrolysis is the most popular phenotypic characterisation test that is used to distinguish between C. jejuni and the other Campylobacter species including C. coli [96-98].

12

Steinbruecker et al. (1999) [9] compared three different methods for Campylobacter identification

which

included

normal

biochemical

techniques,

Gas

Liquid

Chromatography (GLC) and PCR, and noted PCR techniques to be the most sensitive method. Denis et al. (1999) [99] observed that PCR tests provided 100% efficiency compared to only 34% with the biochemical tests. In addition to PCR, many other genotypic methods such as 16S rRNA sequencing, DNA-DNA hybridisation, calorimetric reverse DNA hybridisation assays [9, 98, 100-102], Pulsed Field Gel Electrophoresis (PFGE) [103], Amplified Fragment Length Polymorphism (AFLP) [104-106] and automated ribotyping assays [103, 107] have been used to identify Campylobacter. PCR tests have also been developed for the detection and identification of Campylobacter species from stool specimens as well as from food samples [86]. Harmon et al. (1997) [108] developed primers specifically for differentiating C. jejuni and C. coli directly from poultry specimens. Linton et al. (1997) [91] used PCR to identify Campylobacter species directly from diarrhoeic specimens. Winters and Slafik (2000) [109] used multiplex PCR to differentiate C. jejuni and A. butzleri and Wesley et al. (1995) [86] used PCR for detection and identification of A. butzleri. 1.4. Recent Development in Molecular Methods for Pathogen Detection Subsequent to enrichment and isolation, phenotypic analysis and serotyping has been the cornerstone of pathogen detection and identification. However, these conventional methods, which have a number of disadvantages, have been overcome by the use of methods that target a range of molecules (e.g., DNA, RNA, Protein, etc.). 1.4.1. Choice of Targets The choice of genes depends on a number of factors such as the presence or absence of the gene and the copy numbers of the gene(s). This choice of gene(s) together with the choice of the technique (PCR, SNP, RFLP (Restriction Fragment Length

13

Polymorphism), DGGE (Denaturing or Temperature Gradient Gel Electrophoresis), PFGE, LDR, etc.) determines the sensitivity of detection / differentiation to the required level of discriminations; i.e., family, genus, species or isolate. The recent advent of genomics and proteomic studies have facilitated molecular identification methods by determining specific genes or proteins present in different species and also the number of molecular targets that can be used for the purposes of identification and differentiation. These, together with the advent of high in silico storage and processing capacity and common availability of bioinformatics tools, have led to instantaneous electronic access to generalised molecular biology databases (e.g., GenBank at URL http://www3.ncbi.nlm.nih.gov/) and specialised databases (e.g., Protein database at URL http://www.expasy.org/). An extensive bacterial 16S rRNA sequence database, which holds 96,000 sequences, including those for Campylobacter and Arcobacter species, is currently available electronically [31]. Comparative sequence analysis of the bacteria 16S rRNA genes generally permits phylogenetic relationships to be established down to species level [110]. 23S rRNA, which is almost twice the size of 16S rRNA, contains more information and may provide greater discrimination and identification to the subspecies level [110]. However, the lack of adequate sequence representation in databases of all genera of the domain Bacteria has precluded its use. 5S rRNA (approximately 120 nucleotides), 16S rRNA (approximately 1600 nucleotides) and 23S rRNA (approximately 3000 nucleotides), the three ribosomal RNA genes, are present in all bacteria [110]. The rRNA, are highly conserved and usually multi-copied and species having 70% or more genomic DNA relatedness usually have greater than 97% sequence identity within their 16S rRNA sequences [111]. Although multiplexing PCR molecular methods that simultaneously target rRNA and other non-

14

rRNA genes are highly discriminatory and sensitive, they are cumbersome to perform routinely due to the high technical demands. 1.4.2. Choice of Techniques A number of powerful yet simple and rapid molecular detection techniques are currently under investigation for use in identification and differentiation of infectious agents. Among these, real-time PCR has received the most attention, and has the advantages of simplicity and ease of use and is a candidate for high-throughput application in a similar manner to the cantilever and microarray techniques. Real-time PCR has the potential to replace the existing conventional PCR methods first devised by Saiki et al. (1985) [112]. On the other hand, the technique of T-RFLP is very simple, cost-effective, and rapid, but relies on the availability of sequence information. T-RFLP has been used extensively with 16S rRNA. The technique of LDR is extremely specific and has the potential to discriminate sequences with only one-nucleotide difference. 1.5. Real Time PCR There are several different methods for detecting molecular targets and the choice of target is based on the selection of the technique. In the case of PCR, the double stranded PCR products can be detected by gel electrophoresis or hybridisation probes. PCR products are separated by agarose gel electrophoresis and visualised under UV light after staining with EtBr (Ethidium Bromide), whereas DNA probe hybridisation detects products confirming the products’ identity [110]. DNA hybridisation assays (heterogeneous assays) are more time-consuming and complicated to perform and require that the products or probes be immobilised on membrane or micro-titre plate surfaces. Morrison et al. (1989) [113] described a new detection method based on the hybridisation of fluorogenic probes to the target region of PCR products (homogeneous assays).

15

The principle of real-time PCR depends on continuous monitoring of fluorescent signals derived from FRET [110]. FRET is dependent on the use of specialised, fluorescently labelled PCR oligonucleotides used as probes. Since the first report on the technique of real-time PCR using FRET fluoroprobes in the mid-1990s with the ABI PRISM® [114] and the LightCyclerTM [115], there have been major innovations in terms of both instrumentation design (Table 1.2) and probe technology (Table 1.3). 1.5.1. Equipment and Equipment Formats The improvements to real-time PCR instrumentation include the efficiency of the light source for fluorophore excitation, the detection system (fluorometry and imaging technology), the efficiency of heat generation and heat transfer capabilities to the sample and also the availability of a wide range of filters and fluorophores for multiplexing (Table 1.2). Our laboratory has been using both the hydrolysis and hybridisation probes and two different types of real-time PCR instruments. The descriptions of the systems used in our laboratory are given below. 1.5.1.1 LightCycleTM (Idaho Technology, USA) Real Time PCR Detection System LightCyclerTM is a micro-volume multi-sample PCR machine with a built-in fluorometer produced by Idaho Technology. In this system, the fluorescent probe signals generated by the hybridisation or hydrolysis probes are detected and measured by the in-built fluorometer, simultaneously confirming product identity [116]. Both hybridisation (Dual hybridisation probe) or hydrolysis (TaqMan probe) probes, internal to the amplicon, are added to the PCR reaction and their hybridisation with the newlyformed products’ target region is detected by continuous monitoring during PCR, thus eliminating the risk of carryover contamination associated with handling of PCR products. In addition, the LightCyclerTM consists of a high intensity light source, which

16

Table 1.2

Representative of real-time PCR Instruments available in the market.

Instrument

LightCyclerTM

LightCyclerTM

Rotor GeneTM

iCycler iQTM

ABI 770TM

MX 4000TM

Company

Idaho Technology

Roche

Corbett Research

BioRad

Perkin Elmer

Stratagene

Capacity

24 Glass capillary

32 Glass capillary

36 x 0.2ml tubes

96-well plate

96-well plate

96-well plate

OR 72 x 0.1ml strip Heating

Air

Air

Air

Block heater

Block heater

Block heater

Detector

Fluorometer

Fluorometer

Fluorometer

CCD camera

CCD camera

CCD camera

Multiplex

No

Yes

Yes

Yes

Yes

Yes

Xenon Arc lamp

Xenon Arc lamp

Dual light (470

Tungsten

Laser light

Quartz tungsten

(250 –1000 nm)

(250 –1000nm)

&530nm)

halogen lamp

halogen lamp

continuos

continuos

(350-1000nm)

(350-750nm)

methods

capability Light source

17

heats the air of the sample chamber. The circulating hot air is vortexed and the heat transfer to and in the samples is made efficient due to the small reaction volumes contained within the glass capillaries [117-119]. 1.5.1.2 iCycler iQTM (BioRad, USA) Real Time PCR Detection System The iCycler iQTM system is an improvement on the iCyclerTM (BioRad, USA) PCR system, which is a heat block system. The iCycler iQTM features a broad-spectrum light source that offers maximum flexibility in selecting fluorescent chemistries. The iCycler iQTM detection methods are based in a Charged-coupled Device (CCD) placed on the top of the 96-well plate. This detection system has the multiplexing ability that gave the iCycler iQTM one of its most important features, which is the ability to detect four different fluoroprobes at the same time. In addition to the multiplexing ability, the iCycler iQTM has the facility to run temperature gradients, which is an important parameter for optimising the conditions for real-time PCR. The detection result is displayed by user-friendly software. 1.5.2. Fluorophore Probe Technology and Chemistries In general, probes can be separated into two general functional classes, namely, hydrolysis probes and hybridisation probes. However, a wide range of probe designs are available, especially for the functional category of hybridisation probes, and a number of different fluorophores are available for tagging the probes. However, only limited reports on their effectiveness are currently available. There are numerous probe designs available on the market today (Table 1.3) but the TaqMan and the dual hybridisation probe technologies were the first to be developed and reported. Our laboratory has had extensive experience with both the hydrolysis and dual hybridisation probe technologies [116, 120-124].

18

Table 1.3 Type of fluorophore probe technology suitable for molecular techniques.

Fluoroprobe

URL

¾

http://biochem.roche.com/lightcycler/lc_pri

DNA binding dye e.g. SYBR green, ethidium bromide…etc

¾

nciples/lc_prin_dna_det01.htm

Oligonucleotides probe •

Hydrolysis probes (TaqMan,

bes/prod_Taqman.html

PE) •

http://www.gensetoligos.com/Products/Pro

Minor Groove Binder (MGB)

http://www.syntheticgenetics.com/eclipse/ http://www.gensetoligos.com/Products/Pro

•

bes/prod_LightCycler.html Adjacent hybridisation probes (Roche Dual probe)

¾

http://www.gensetoligos.com/Products/Pro bes/prod_Molecular.html

•

Hair Pin Probes

•

Molecular Beacon

•

Sunrise UniPrimer

711.html

•

Scorpion

http://www.bostonprobes.com/pages/scienc

Peptide nucleic Acid probes

e/tech-support/sci-tech.html

http://www.talron.co.il/home.html http://www.probes.com/handbook/figures/0

(PNA)

As shown in Table 1.3, there are three different types of real-time PCR fluoroprobes: Oligonucleotides probes, DNA binding dye probes and Peptide nucleic acid probes. Oligonucleotide probes have different techniques such as an adjacent hybridisation probes assay and TaqMan assays. Adjacent hybridisation probe assays (Figure 1.3) are based on the fact that two probes, donor and acceptor, are separated by one base gap and labelled with two different fluorophores, such as fluorescein and Cy5; in this case fluorescein labelled the 3’ end of the upstream probe, Cy5 labelled the 5’ end of the downstream probe [125]. During PCR amplification, fluorescein is excited when it is hybridised to accumulative PCR products and then emits the excited energy to the

19

adjacent Cy5 probes, which will increase the Cy5-to-fluorescein ratio [125]. During PCR Cy5-fluorescein will keep increasing until the DNA template is consumed. The second technique is the TaqMan probe, in which both fluorophores are found in one probe. The TaqMan probe is described in detail in Section 1.6.1. The dyes SYBR Green I (Figure 1.4) and YO-PRO-1 bind to dsDNA in a similar manner to EtBr and the resultant fluorescence emission can be measured in a real-time PCR apparatus [116, 120-123]. These dyes bind to specific or non-specific products (primer dimer) and therefore lack the ability of discrimination. Chapter 4 illustrates how both LightCyclerTM and iCycler iQTM SYBR Green I assays have been used to distinguish C. jejuni from all other Campylobacter species based on the amplification of the hippuricase gene using specific primers. The adjacent hybridisation probes assay has an advantage over TaqMan and SYBR Green I assays. Adjacent hybridisation probes are cheaper than TaqMan probes and both are more accurate than SYBR Green I assays. The other advantage is related to the 16S rRNA universal Cy5 probes that have been developed close to the high variable region of bacterial domain. This means that only one Cy5 probe is required, saving costs for synthesising multiple Cy5 labelled probes. 1.5.2.1 Adjacent Hybridisation Probes Assays Wittwer et. al., [115] developed the first real-time PCR assay which used adjacent hybridisation probes labelled with Cy5 and fluorescein, as the resonance energy transfer pair. Choice of the fluorophore pair depends on the spectral wavelength overlap in order to minimise the direct excitation of the acceptor; for this reason, Cy5 and fluorescein have been selected [115]. Real-time PCR can be used to detect particular species through the use of a fluorescent probe that is complementary to the investigated DNA in the LightCyclerTM and iCycler iQTM. The ability to measure the accumulated PCR product during PCR amplification is

20

called real-time PCR. The LightCyclerTM has been successfully employed in real-time PCR to identify and differentiate pathogenic Leptospira from non-pathogenic Leptospira by using different specifically labelled probes [122, 123]. Probes were designed that bound to target 16S rRNA gene sequences in the genome of Leptospira, and the use of fluorescence monitoring showed whether the DNA of interest was being amplified [122]. Real-time PCR is capable of detecting small amounts of DNA and is therefore valuable in bacterial detection and identification as well as in studying microbial diversity. Real-time PCR has advantages over the conventional and competitive PCR, as there is no loss in the yield of PCR product. The second advantage is that real-time PCR does not require post-PCR processing, which results in increased throughput with reduced carryover contamination and post-PCR process errors [102, 126]. Finally, real-time PCR is extremely rapid compared to all other PCR methods. Idaho (1993) claims that a PCR reaction can be run in less than 15 minutes because the reaction is monitored continuously [127]. For the purposes of this study, the FRET dyes are to be placed on two hybridisation probes. Increase in the FRET should be observed due to the two florescent dyes being brought together by hybridisation [127]. The fluorescence from the acceptor fluoroprobe is then proportional to the concentration of accumulated PCR product. 1.6. Real Time PCR Multiplex Different coloured dyes are frequently used to increase the density of information obtained from a sample. These dyes can either be attached to a primer or probe and also can be detected either during PCR amplification or by post amplification (e.g., agarose gel electrophoresis). Multiplexing by colour is used widely in flow cytometry and DNA sequencing. Colour multiplexing is partly limited by the broad emission spectrum

21

characteristic of common dyes. Fluorescent dyes are easily measured, have few biological safety hazards and fluorescence detection can be multiplexed, allowing process controls or higher throughput assays [128].

Figure 1.3 Schematic representation of real-time PCR adjacent hybridisation probes assay. http://www.gensetoligos.com/Products/Probes/prod_LightCycler.html

Figure 1.4 Schematic representations of real-time PCR SYBR Green I assay. http://biochem.roche.com/lightcycler/lc_principles/lc_prin_dna_det01.htm

PCR multiplexing can be defined as using multiple primers to produce multiple amplicons within a single tube PCR reaction. Multiplexing is an important application of conventional PCR. Real-time PCR multiplexing is defined as the use of multiple fluorogenic probes for the detection and discrimination of multiple amplicons. Compared to conventional PCR multiplexing, real-time PCR multiplexing is not easy to 22

perform due to the lack of the fluorophore combination that can be used in a single tube in addition to the use of a monochromatic energising light source [129]. Even though excitation by a single wavelength generates bright emissions from an amenable fluorophore, this restricts the number of fluorophores that can be included [130]. One of the most important developments in real-time PCR is the discovery of the nonfluorescent quenchers which librate some wavelengths that were occupied previously by the emissions of the fluorescent quencher [130-132]. This allows the use of a greater number of spectrally detectable fluoroprobes in one tube. Different non-fluorescent quencher families, such as the black hole quencher (BHQ-1, BHQ-2- and BHQ-3) (Figure 1.5), QSY (QSY-35, QSY-21, QSY-7 and QSY-9) and the Iowa Black quencher (Iowa Black FQTM and Iowa Black RQTM) are now used to cover the 400-700 nm wavelength, which highlights the need for only one non-fluorescent quencher that can cover this wavelength range (Figure 1.5 and Table 1.4). The newly discovered dark quencher is a chromophore that is characterised by capturing energy from the excited reporter molecule without any subsequent emission of a detectable fluorescent signal of its own, thereby dispersing energy as heat rather than fluorescence [130]. The advantages of the dark quencher over the fluorescent quencher are: Dark quenchers capture light energy from an excited reporter molecule. Dark quenchers have advantages over fluorescent quenchers in many applications: (i) Probes incorporating dark quenchers are more sensitive due to lower background fluorescence compared to the fluorescent quenchers; (ii) they enable the simultaneous use of a wide range of reporter dyes, expanding the options available for multiplex assays; and. (iii) they are compatible with a broad range of image analysis instruments [133].

23

Table 1.4 The maximum absorbance and quenching range for four different nonfluorescent quenchers currently available.

Quencher

Maximum absorbance (nm) Quenching range (nm)

BHQ-1

534

480-580

BHQ-2

579

550-650

BHQ-3

672

620-730

QSY-35

472

410-500

QSY-7

560

500-600

QSY-9

562

500-600

QSY-21

660

590-720

DABCYL

453

380-530

Iowa Black FQTM

532

420-620

Iowa Black RQTM 645

500-700

Figure 1.5 Absorption spectra of the black hole quencher (BHQ-1, BHQ-2 and BHQ3), and the suitable fluorescent dye for each BHQ members.

24

1.6.1. TaqMan Probe Holland et al. (1991) [132] described a technique where they detected the amplicons by monitoring the affect of the Taq polymerase /DNA nuclease effect on specific radio labelled probe/target DNA duplexes. This technique was the basis for the development of the TaqMan assay by researchers at Perkin Elmer [134]. The following criteria are important for an oligoprobe labelled dye. It: (i) should be easy to attach the dye to the probe, (ii) should be able to be detected at low concentrations, (iii) should be able to produce a detectable signal upon specific hybridisation, (iv) should be stable at high temperatures, (v) should not interfere with Taq DNA polymerase activity, (vi) should be compatible with a broad range of image analysis instruments, and (vii) should be biologically safe [129-131]. In 1993, the TaqMan probe was combined with PCR to develop the first real-time PCR fluorescence excitation and detection assay [135]. TaqMan real-time PCR assays were initially run on the Applied BioSystems nucleic acid detection system (ABI 7700). A TaqMan probe (Figure 1.6) is conjugated to a fluorescent reporter dye (FAM, TET, Cy5, etc.) at the 5’end and a quencher, either a fluorescent quencher such as TAMRA or a non-fluorescent quencher such as BHQ-1, at the 3’ end. It requires 5’ exonuclease activity to separate the two probes [129]. During the extension stage of the PCR cycle, the fluorogenic probes hybridise with the DNA template in the region that is located between the reverse and forward primers. After that, degradation of the TaqMan probe, by the 5’->3’ exonuclease activity of Taq DNA polymerase, frees the reporter dye from the quenching activity of the quencher, and thus the fluorescence intensity will increase with the increase in cleavage of the probe, which is related to the amount of PCR product formed [136]. This fluorescence change is irreversible with temperature and Tm analysis impossible. 25

Figure 1.6 Schematic representation of real-time PCR TaqMan http://www.gensetoligos.com/Products/Probes/prod_Taqman.html

technique

1.6.2. MGB TaqMan Probe TaqMan chemistry has recently undergone significant improvements, which has led to the development of the TaqMan MGB probe, which includes: addition of minor groove binders (MGB) to the probe to increase sensitivity and stability of the probe hybridisation, as well as using a non-fluorescent quencher (NFQ) such as a black hole quencher (BHQ) to replace the standard fluorescent quencher (TAMRA) [129]. Minor groove binders (MGB) were first developed for therapeutic application, but toxicity limited their use [137]. Epoch Bioscience recognised their ability to stabilise DNA hybridisation and developed the TaqMan MGB probe technology [138]. Due to their performance in stabilising DNA probe target hybridisation, and their ease of synthesis, Epoch Bioscience developed the DPI3 (dihydrocyclopyrroloindole tripeptide) type MGB probe (Figure 1.7) [138]. MGB is a derivative of naturally occurring antibiotics that specifically bind to the duplex DNA in the minor groove [137]. The development of MGB probe chemistry and application of hyper-stabilised duplexes with the complementary DNA [135, 139]. NMR studies by Kumar et al. (1998) [140] have indicated that the MGB, dihydrocyclopyrroloindole tripeptide (DPI3), folds into the 26

minor groove formed by the terminal 5-6 bp. An MGB is a crescent–shaped, long and flat molecule that fits snugly (binds isohelically) into the minor groove of duplex DNA [135]. MGB stabilised in the minor groove by either Van der Waals contacts or hydrophobic and electrostatic interaction [141] (Figure 1.8). During PCR amplification, after the TaqMan MGB probe hybridises, the MGB stabilises the annealing by folding into the minor groove of the DNA duplex created between the probe and the complementary sequence of the target DNA [142]. The binding of MGBs in the minor groove can be very strong, particularly with AT-rich B-form DNA [135], resulting in an increased probe Tm [143]. Kutyavin et al. (2000) [142] found that the length of a MGB probe that varies between 12-20 bp will fulfil the Tm requirement of PCR reaction, while the other TaqMan probe lengths are between 14 and 40 nucleotides depending on the GC content. However, most MGB’s are sensitive to the structure of the DNA duplexes and bind poorly to GC rich DNA [135]. MGB probe G-C content is not as important compared to the other probe, as the addition of the MGB to probes with A-T rich sequences give a larger increase in Tm than the probes with G-C rich sequence [142]. For example, in 2004, Kulesh et al. [143] redesigned their TaqMan probe as TaqManMGB, thereby reducing the probe sequence from 26 bp (Tm=60.4) to 18 bp (Tm=69.9). In contrast, MGBs from the lexitropsin family were shown to limit the strict AT preference [144], exhibiting in some cases exclusively GC site binding [145].

27

Figure

1.7 DPI3 (dihydrocyclopyrroloindole tripeptide) http://www.epochbio.com/technology/MGBs.htm

Figure

1.8 Water effect on MGB probe/DNA http://www.epochbio.com/products/mgbe_how_it_works.htm

28

type

MGB

probe

duplex

1.7. Fingerprinting Molecular Techniques Species richness (the numbers of different species) and species evenness (population density of each species) are two important factors in defining microbial community structure and diversity [146]. Quantitation of these two factors is limited by culturedependent methods because large fractions (>85 to 99.999%) of the organisms existing in nature are yet to be cultured [146-149]. Due to the complexity of the microbial communities and their metabolic diversity, monitoring microbial population dynamics in an environmental system can be a sensitive method to track changes [150]. The purpose of fingerprinting is often to rapidly profile the kinds of organisms found in a sample [151]. 16S rRNA genes are used as phylogenetic markers and have been used with many DNA fingerprinting techniques [152-155]. This technique has proven to be accurate, rapid, robust, and a cost-effective method to assess microbial diversity. To characterise bacterial communities, different fingerprinting techniques that avoid culture dependent methods have been developed recently, including RFLP [152], 16S rRNA cloning and sequencing [147, 156], Denaturing or Temperature Gradient Gel Electrophoresis (DGGE or TGGE) of 16S rRNA [157], and rDNA-targeted in-situ hybridisation [158-162]. Other techniques, based on the ligase enzyme activity such as Oligonucleotide Ligation Assay (OLA) [163], Ligase Chain Reaction (LCR) [164, 165], LDR [166], have all been developed for rapid detection and identification as well as characterisation of bacterial communities. These techniques can provide important information, which is useful for the assessment of the phylogenetic groups of the prokaryotes. However, the reliability of these techniques is hindered by different variables, such as DNA isolation, PCR amplification, and the selection of the clones as well as result analysis.

29

Clone libraries are the most sensitive fingerprinting technique and offer the highest degree of phylogenetic resolution, but they can be cumbersome for microbial community analysis [167]. In addition, using clone libraries for community composition comparison can be problematic if the libraries offer incomplete coverage of a community, although progress has been made in this area [168]. The construction and screening of these clone libraries is time consuming, laborious, and expensive, therefore, the clone library size in most cases is reduced to less than 50 individual clones per sample, and still only a few samples can be processed in a reasonable time [169]. RFLP is able to characterise and profile bacterial isolates [170] and complex bacterial communities [171]. It is usually run on horizontal or vertical high-resolution polyacrylamide gel electrophoresis. At the end of the electrophoresis, all the fragments bands are labelled with a fluorescent dye such as EtBr, SYBR Green I, or any other similar dye [171]. The third fingerprinting technique that can be used is DGGE. In this technique, DNA fragments obtained from PCR amplification of microbial communities, using specially designed primers (GC clamp primer) are separated using a highresolution polyacrylamide vertical gel electrophoresis according to their melting behaviour, which is determined by their nucleotide sequence [172, 173]. Recently, TRFLP has been shown to be an automated, rapid, sensitive and cost-effective technique for the detection and characterisation of bacterial communities by using capillary electrophoresis (CE) together with laser-induced fluorescence (LIF) detection [174]. TRFLP and DGGE are the most common fingerprinting techniques that are used to study marine bacterial communities [175-187]. 1.7.1. Terminal Restriction Fragment Length Polymorphism (T-RFLP) Cancilla et al. (1992) [188] were one of the first groups to use automated sequencing gel electrophoresis for strain typing, using a fluorescently labelled specific primer targeting 30

the Repetitive Extragenic Plaindromic (REP). Avaniss-Aghajani et al. (1994) [174] combined sequencing technology and 16S rRNA to develop a new screening technique to identify Mycobacteria. It was described as a highly sensitive, cost-effective and rapid assay for detecting and identifying a broad spectrum of bacterial species [174]. T-RFLP is a modification of RFLP and PCR in which a forward or reverse primer is labelled with a fluorescent dye (e.g., FAM), and the product is digested with one or more restriction enzymes, producing one or more Terminal Restriction Fragments of different lengths, depending on the enzyme specificity and DNA sequence. The digested PCR products are analysed on an automated sequencing instrument (capillary electrophoresis or gel electrophoresis), which determines the exact sizes of the fluorescent fragments by comparing the fragment size to known internal standards. The resulting data is often displayed to look similar to a chromatogram, with each peak representing a different species, genus, or bacterial group that share the same restriction site [151]. However, because most of the microbial communities have not been identified so far, one cannot be absolutely sure if a particular peak comes from the expected organism or not, as there can be unknown organisms that have the same terminal fragment lengths [168]. Current technology does not include the ability to isolate the DNA from the chosen peak to sequence and compare it to the microbial sequences available online. Although the T-RFLP fingerprints can show which groups are the most profuse and which are uncommon in the screened sample, they can only provide semi-quantitative data, as it is not possible to expect the relative sizes of peaks to precisely reflect the abundance of the representative groups in the community [151]. This is because the peaks represent the digested PCR products amplified from the target DNA regions, and the amount of PCR product is affected by differences in DNA per cell, the number of

31

copies of the 16S rRNA genes per genome, reaction incubation time and other factors [189]. The differences in the terminal fragment lengths are due to the presence of restriction sites in a particular region of the PCR product for each microbial group, genus, species, and strain; sometimes even the absence of the restriction site will lead to large differences in terminal fragment lengths. Additionally, for closely related groups that share the same restriction sites, any insertions or deletions usually lead to relatively small differences in terminal fragment lengths [190]. T-RFLP techniques are easier to perform due to the development of bioinformatic tools. T-RFLP has been successfully used to study microbial community dynamics, diversity and richness in soils [156], marine waters [182], contaminated aquifer samples [146, 191], complex bacterial communities [146] and Archaea in the gut of marine fish [192]. A new software found on the Ribosomal Database Project (RDPII) web page has been developed (http://www.rdp.cme.msu.edu/html/TAP-T-RFLP.html#program), to analyse T-RFLP, based on the selection of amplification primer(s) and the restriction enzyme(s) used for digestion. Therefore, the theoretical results determined by RDPII can be compared with the experimental T-RFLP results. The advantages of the RDPII software is that it can retrieve and compare a huge number of the DNA sequences available online. This data could then be used to make a prediction of the TRFs from the thousands of available organisms. Five different analysis methods (2nd order least square, 3rd order least square, cubic spline, global southern and local southern) available in the GeneScan 3.1.2 software (Applied BioSystems, USA) can be used to analyse TRF data. Different analysis methods generate different standard curves for the internal ladder, thus creating differences on the observed TRF results [193]. The local southern method produces a standard curve with the least TRF drift, where TRF drift is the observed TRF length 32

minus the true TRF length [194]. TRF drift have been reported previously with estimates ranging from as little as 1 bp to as much as 7 bp [146, 193-195]. Electropherogram data analysis has suggested that the major source of the TRF drift is the differential migration of ladder and DNA sample, which is due to the difference between the internal ROX-labelled ladder and the 6-FAM labelled sample DNA, most likely due to the ROX label having 12 more carbon atoms than the 6-FAM label (Applied BioSystems, USA). 1.7.2. DNA Ligases, and Ligase Based Molecular Techniques DNA-based technologies have been used to study molecular genetics, microbiology, immunology and oncology in clinical settings [196]. PCR-based methods have been used to design numerous molecular diagnostic techniques for the detection and identification of pathogens, mutations and for genotyping [196-200]. Most of the recently developed techniques are finding applications in nucleic acid-based diagnostic areas. DNA ligase detection and identification techniques have recently been developed and reported. DNA ligases are ubiquitous cell proteins that are essential for a number of important cellular processes, including DNA replication, recombination and repair systems present in all living cells, from viruses to humans [201]. These enzymes catalyse the formation of a phosphodiester bond at single-strand breaks by esterfication of adjacent 5’-phosphoryl and a 3’-hydroxyl in double stranded DNA [201]. The ligation reaction mechanism can be split into three sequential steps: enzyme adenylation, substrate adenylation (deadenylation) and nick-closure [202]. The enzyme adenylation step involves the transfer of an adenosine monophosphate (AMP) group from the cofactor nicotinamide adenine dinucleotide (NAD) or adenosine 5'triphosphate (ATP) to a lysine residue in the adenylation motif KXDG through a phosphoamide linkage [203]. The substrate adenylation step involves the transfer of the 33

AMP (Adenosin monophosphate) group to the 5' phosphate at the nick through a pyrophosphate linkage to form a DNA-adenylate intermediary [202]. The nick-closure step is involved in the formation of a phosphodiester bond, which seals the nick and releases AMP [203]. DNA ligases can be divided into two families, those that require ATP (Adenosine triphosphate) as a cofactor and those that require NAD (Nicotinamide adenine dinucleotide). ATP-dependent ligases have a molecular weight between 30 to >100 kDa and are found in most living cells (eukaryotes which include yeast and mammals [204, 205] and prokaryotes (Bacteria, Archaea and Eubacteria) [206, 207], and in viruses [208]. NAD-dependent ligases are found mostly in bacteria [209, 210] are highly homologous and are monomeric proteins of 70–80 kDa [201]. Many eukaryotic organisms encode multiple ATP-dependent ligases, while some eubacteria encode both NAD- and ATP-dependent ligases [202]. In contrast, there are four different ligase genes, ligases I, III, and IV and V, found only in humans [204, 211, 212]. The LDR technique used in this experiments utilises Taq DNA Ligase (New England Biolabs, Australia) which is an NAD+-dependent DNA ligase, purified from an E. coli isolate containing the recombinant Thermus aquaticus HB8 ligase gene [213]. Based on the ligase activity, different molecular techniques have been developed recently, such as OLA, LCR, Gap-LCR, and LDR. In OLA, two probes are hybridised to the sequence of interest to form a nick. One probe contains a reporter, which is 32Plabelled or fluorescently labelled (e.g. FAM, HEX, TET, etc.) [202]. The second probe contains a biotin group. The ligase enzyme will covalently join the two probes only if they have a 100% match with target sequences at the nick junction. The fluorescence or radio labelled signal resulting from the ligation will be detected by a fluorometer or a radiometer. OLA is used to detect and identify the βA and βs alleles of human β-globin in cells, plasmid and genomic DNA [214]. Two pairs of complementary probes are 34

annealed at both strands of the target sequence in the Ligase Chain Reaction LCR [165, 215]. The ability of the ligation products of one strand to serve as a template for the complementary pairs of probes in subsequent cycles leads to exponential amplification. LCR is used for the detection and identification of human papilloma virus, HIV-1 (Human Immunodeficiency Virus) and Mycobacterium tuberculosis, as well as for genetic disorder diagnosis [164, 215, 216]. Gap-LCR is similar to the conventional LCR except that Gap-LCR has a gap between the juxtaposed primer pair. Abbott Laboratories (Abbott Park, IL) have developed a few tests for rapid detection of Mycobacterium tuberculosis, Chlamydia trachomatis, HIV, other pathogens and for genotyping beta-2 adrenergic receptor genes [217-220]. 1.7.2.1 Ligase detection reaction LDR has been reported for rapid detection of the mutations associated with colorectal cancer [221], viral detection [222, 223], parasite species causing malaria in humans [224] and bacterial discrimination [166]. As mentioned previously (Section 1.7.2), the high sensitivity of the ligase-based techniques gives them the ability to detect and identify minority-mutated targets from large normal cells; for example, cancer cells, which result from a somatic mutation, found in a large population of normal cells [202, 221]. The discriminative probe, as can be inferred from its name, is used to discriminate strain, species, genus or family from other microbial communities, or to distinguish mutated genes from normal ones, as in case of a cancer cell. One bp difference at the 3’ end of the discriminative probe will be able to differentiate a target sequence from crude populations; for example, in a microbial community, a difference in a single nucleotide along the 16S rRNA can be employed for rapid detection and identification of different micro-organisms [166]. On the other hand, the common probe is universal; for example,

35

in microbial identification, a common probe will detect / bind to large members of the microbial community. A common probe in the case of microarray detection is normally attached to a Zip Code complement, which will direct the PCR / LDR product to the designated Zip Code address on the DNA universal microarray [225]. Microarrays are fabricated on a glass slide, where a unique 24-nucleotide artificial sequence composed of six tetramers having similar Tm values, termed Zip Code is attached to a glass slide. The Zip Code concept is similar to molecular tags developed for bacterial and yeast detection and identification, as well as human genetic disorder diagnosis [226]. The LDR technique involves the amplification of the target DNA using the conventional PCR assay followed by the ligation of two probes. The discriminative probe is labelled with fluorescent dye (6-FAM) at the 5' end, and the common probe labelled with a phosphate group (PO4) at the 3' end [166]. If the probes are perfectly matched to the target sequence of the PCR product, DNA ligase will ligate those two probes, forming the LDR product, which binds to the location on the microarray where the Zip Code sequence has been spotted [225]. Laser scanning of the array will then detect the fluorescence released from the LDR product and the LDR product/s will be identified by its location within the array [166]. In addition to that, LDR products can be detected by using DNA sequencers such as an ABI 377 DNA gel sequencer or by using capillary electrophoresis sequencers such as ABI 310 and 337 [227]. Using a common probe allows the development of multiplex screening and identification of certain genes or species by varying the size of the discrimination probe or the fluorescent material used [227]. The formation of the LDR product that could be detected by either microarray or ABI DNA gel sequencer systems indicates the presence of the target, while LDR product length will designate the target exactly (e.g., bacterial genus or species). Furthermore, such assays are ideal for multiplexing as shown in chapter 9 of this thesis, since several primer sets can hybridise along a gene without the interference

36

encountered in polymerase-based assays [221]. Developing PCR multiplexing that produces equivalent amounts of each PCR product can be difficult, time consuming, expensive and laborious [227]. This is due to the difference in the annealing temperature of each set of the primers in the reaction. The primer concentration, template concentration, MgCl2 and other salt concentration can affect Tm. Therefore, as the number of the PCR amplicons increase, it becomes more difficult to optimise the reaction condition to obtain an equal amount of each amplicon. Using the PCR/LDR approach, we targeted the 16S rRNA sequence using Campylobacter, Arcobacter and representatives of other bacterial groups, selected from GenBank and Ribosomal Database Project II. Butsi et al. (2003) [166] determined different 16S rRNA common probe locations, which were tested with Campylobacter and Arcobacter to determine if we could use these regions to design discriminative and common probes to differentiate Campylobacter and Arcobacter from each other and from other bacterial groups. Furthermore, the FlaA gene was used to differentiate C. coli and C. jejuni from the other Campylobacter species after determining discriminative and common probe regions. 1.8. Project Aims and Rationale From the above discussions, it is apparent that the unique shape of the cell is extremely useful in Gram-stain identification of members of the family Campylobacteraceae. It is also clear that certain species are more frequently associated with gastrointestinal infections than others and therefore need to be distinguished by specific and sensitive tests. For example, C. coli is the second most important pathogen after C. jejuni and easy discriminatory methods are required. However, despite various developments in chemotaxonomic methods, species and subspecies detection and identification beyond the family level is problematic due to their relative biochemical inactivity [228]. It is therefore not surprising that to-date, 67 different biochemical tests and molecular 37

techniques such as Polymerase Chain Reaction Restriction Fragment Length Polymorphism (PCR-RFLP), 16S rRNA gene sequencing and DNA – DNA hybridisation have been devised for the rapid identification of Campylobacter species. However, not all of these techniques are suitable for routine testing in a microbiology laboratory, as they may be cumbersome, time-consuming and / or expensive to use. As a consequence of an increase in food- and water-borne related diseases and our inability to apply sensitive methods to discriminate pathogens such as members of the family Campylobactericea, new molecular targets and methods are currently being devised. The aim of this project is to develop new, rapid, easy-to-use, sensitive, accurate and cost-effective

molecular

methods

for

the

identification,

differentiation

and

quantification of members of the Campylobacter and Arcobacter species from enriched and isolated cultures. The approach that has been developed includes the use of bioinformatics tools to identify gene targets (such as those that may be potentially linked to pathogenesis) and the use of new emerging techniques (such as real-time PCR). Once such methods are developed, they could be used for monitoring disease outbreaks and the rapid identification of pathogens in patients for efficient medical attention and the monitoring of the health of the environment (such as food quality and water quality). Rapid responses due to these new methods will be crucial in limiting the severity of disease outbreaks. In addition to real-time PCR, two fingerprinting techniques, T-RFLP and LDR have been developed in our laboratory for rapid detection and identification of Campylobacter and Arcobacter species (Figure 2.1)

38

2. Chapter 2 Material and Methods 2.1. Bacterial Cultures, Sources and Growth All cultures were grown on blood agar plates (BAP) or in Preston Campylobacter broth medium (PCMB) for up to 48 hours at 42ºC under CO2-enriched microaerophilic conditions. C. jejuni (ATCC 940565) and C. coli (NCTC 11366) were purchased from the American Type Culture Collection Centre (ATCC) and the National Collection of Type Cultures (NCTC), respectively. C. upsaliensis (QHSS 99M126) and C. hyointestinalis (QHSS 99M2318) were obtained from Public Health Microbiology Laboratory, Queensland Health Scientific Services (QHSS) collection. C. lari (ATCC 35223), C. sputorum biovar sputorum (ATCC 33562), C. fetus subsp. fetus (NCTC 10842) and C. fetus subsp. venerealis (NCTC 10354) were kindly provided by Dr Victoria Korolik. Another 168 isolates obtained from animals, humans, plants and birds were resuscitated from −80ºC stock cultures (Appendix 1). These had previously been identified as Campylobacter isolates by the Public Health Microbiology Laboratory, QHSS, which acts as a referral centre for human isolates of Campylobacter, as well as being involved in a significant amount of food and water related testing for this group of organisms. Isolates were characterised phenotypically by their ability to hydrolyse hippurate, their production of catalase, their sensitivity to nalidixic acid and cephalothin and their growth on Preston and blood agar plates at 25ºC, 37ºC and 42ºC and in the presence of a special atmosphere (80% N2, 10% H2, 10% CO2) [13]. Similarly, three Arcobacter type cultures, A. butzleri (NCTC 12481), A. skirrowii (NCTC 12713), and A. nitrofigilis (NCTC 12251) were purchased from NCTC and another 22 isolates from animals, humans, plants and birds were resuscitated from 39

−80ºC stock cultures (Appendix 2) and were grown under the same conditions (except that Arcobacter grows at 37ºC). Furthermore, E. coli, Staphylococcus aureus, Clostridium perfringens, Micrococcus luteus, and Pseudomonas aeruginosa were used as negative control. 2.1.1. Campylobacter and Arcobacter Enrichment Culture for Chicken sample Campylobacter enrichments were initiated from 30 chicken pieces purchased from four poultry outlets (Appendix 5). They were rinsed in Preston Campylobacter broth medium (PCMB) (Oxoid, Australia) fortified with polymyxin B (5 IU/ml), rifampicin (5 µg/ml), trimethoprim (5 µg/ml) and cycloheximide (50 µg/ml) and incubating the broth at 42ºC under CO2-enriched microaerophilic conditions for 24 hours (Appendix 5). After incubation, a loopful of culture was streaked on to Preston selective agar plates (Oxoid, Australia) and incubated for 48 h at 42ºC under CO2-enriched microaerophilic conditions. Cell appearance and biochemical tests were then used to identify Campylobacter. The same procedures were used to identify Arcobacter but Arcobacter enrichment broth (AEB) (Oxoid, Australia) and Arcobacter agar (Oxoid, Australia) were used for culturing and all incubations for Arcobacter were at 37°C (Appendix 6). 2.2. DNA Extraction Three different methods of DNA extraction were used in this study, as described below. The sources for DNA extraction were pure cultures and enrichment cultures initiated from chicken samples as described section 2.1.1. 1·5 ml of Campylobacter and Arcobacter enrichment cultures (Section 2.1.1) used to precipitate of the residue and then 1 ml of the supernatant was centrifuged for 10 min at 13 000 g to pellet the cells. After removing the supernatant, the pellet was suspended in 400 μl TE buffer.

40

2.2.1. CTAB Method Campylobacter and Arcobacter colonies were harvested from BAP and the enrichment cultures (Section 2.2) suspended in a 400 μl TE buffer. 8μl lysozyme (50 mg/ml) and 40 μl achrompeptidase (20 mg/ml) were added and the mixture incubated at 37°C for 1 hour. Then 30 μl 10% v/v SDS and 10 μl Proteinase K (20 mg/ml) were added and incubated at 37°C for 10 minutes, followed by the addition of 100 μl NaCl and 80 μl CTAB/NaCl and incubated at 65°C for 10 minutes. For DNA purification, 750 μl chloroform/isoamyl alcohol (25:24) was added followed by centrifugation at 17.970 g for 5 minutes. The upper aqueous layer was removed and transferred to a new tube, where 750 μl of phenol/chloroform/isoamyl alcohol (25:24:1) added and the mixture centrifuged for another 5 minutes. The upper DNA containing layer was transferred to a new tube and the DNA precipitated using 450μl 90% isopropanol or 100% ethanol, centrifuged at 17.970 g for 10 minutes and the pellet washed with 200 μl 70% ethanol and air-dried. The pellet was resuspended in 50-100 μl TE buffer (pH 7.4) before use. 2.2.2. Rapid DNA Extraction by Boiling Five to ten colonies were harvested from BAP and resuspended in 100 μl TE buffer, while for enrichment cultures cell pellet suspended in 100 μl of TE buffer as described in section 2.2. The suspension was boiled for 10 minutes, centrifuged briefly and 1 μl of supernatant used as template real-time DNA for PCR in the LightCyclerTM (Sections 3.2.3, 4.4.2, 4.4.3, 5.3.2 and 6.3.1). 2.2.3. Phenol-chloroform-isoamyl Alcohol Method Campylobacter and Arcobacter colonies were harvested from BAP and suspended in 500 μl Tris (5mM, pH 8.0) containing 100 μl lysozyme (50 mg/ml). The mixture was incubated on ice for 5 minutes, then 600 μl EDTA (0.5M, pH 7.5) and 40 μl

41

achrompeptidase (20 mg/ml) was added, and incubated on ice for a further 5 minutes. 0.25 ml of 20% (w/v) SDS was added and the sample was gently mixed and incubated on ice for a further 5 minutes. 100 μL of 20 mg/ml proteinase K was added and the sample incubated at 55ºC overnight in order to allow both cell lysis and protein digestion to proceed to completion. During this time, a decrease in turbidity and increase in viscosity indicative of cell lysis was observed and the upper DNA containing aqueous

layer

removed

and

transferred

to

a

new

tube.

750

μl

phenol/chloroform/isoamyl alcohol (25:24:1) was then added and the mixture recentrifuged. DNA was precipitated by the addition of three volumes of ice cold 100% ethanol, spooled using a heat-sealed, sterile Pasteur pipette and transferred to a fresh tube containing a 500 μl TE buffer. 5 μL of 100 mg/ml RNase (QIAGEN) was added and incubated at 37°C for 30 minutes. Genomic DNA was stored at –20ºC. 2.3. Agarose Gel Electrophoresis of DNA Agarose gel electrophoresis was used as a visual check for DNA purity and for the estimation of molecular weights and concentrations. For this, a 1% agarose gel was prepared by suspending 0.8 g of agarose in 80 ml of 1x TAE buffer (50 X TAE buffer (per litre): 242g Tris, 57.1 ml glacial acetic acid, 100 ml 0.5M EDTA (pH 8.0). The agarose dissolved by heating in a microwave (~1 min). Any liquid lost due to evaporation as a result of microwaving was replaced with distilled water (this was achieved by weighing the solution before and after microwaving). The solution was cooled to approximately 55°C and 10 µl of 5mg/ml EtBr were added with care, as EtBr is a carcinogen. The solution was poured into a gel tray and an appropriate comb inserted to form the wells. Once the gel had set, the comb was removed and the gel submerged in an electrophoresis chamber containing a 1x TAE running buffer. DNA samples containing

42

0.2 volume of gel loading solution (6 X DNA loading dye (per 10 ml): 0.025g bromophenol blue, 1.5 ml Ficoll) were loaded into the wells and electrophoresed at room temperature at 80 volts for 30-40 minutes. The DNA was visualised under ultraviolet light and a digital image taken using a UVP GDAS 1200 Gel Documentation Analysis System (Pathtech Pty Ltd, Australia) and edited using a simple image editor (e.g. PaintShop Pro v4.12). 2.3.1. Ultra Agarose Gel Electrophoresis A 1% agarose gel electrophoresis used as an aid to determine the size of products could not differentiate between small variations in the amplicons size produced in the multiplex PCR assays. Therefore, a high-resolution agarose method was chosen. For this 4.5 g of ultra agarose 1000 (Gibco, USA) was suspended in 100 ml of 1x TAE buffer and dissolved by heating in a microwave for 20 seconds intervals followed by swirling (~2 minutes). Any liquid lost due to evaporation as a result of microwaving was replaced with distilled water (this was achieved by weighing the solution before and after microwaving). The solution was cooled to approximately 55°C and 10 µl of 5mg/ml EtBr added. The solution was poured into a gel tray and an appropriate comb used to form the wells. The agarose gel was allowed to solidify at room temperature for 30 minutes. Once the gel had set, the comb was removed and the gel submerged in an electrophoresis chamber containing a 1x TAE running buffer. DNA samples containing 0.2 volume of gel loading solution were loaded into the wells and electrophoresed at room temperature at 100 volts for 60 minutes. The DNA was visualised under ultraviolet light and a digital image taken using a UVP imaging system.

43

2.4. 16S rRNA Amplification and Sequencing 2.4.1. 16S rRNA Primer Design A number of software packages are available electronically (e.g., Primer Design, Xprimer and Primer Selection), which aid in the design of primers. These software packages follow a number of criteria, listed below [110]: •

Primer Tm is always greater than 45οC.

•

Primers are at least 18 bp long in order to ensure good hybridisation to the target site.

•

Repeating bases are avoided wherever possible (i.e., avoid three or more similar bases in succession, especially Gs or Cs).

•

The primer GC content is at least 50% in order to have a high Tm. If the GC content is less than 50%, the numbers of base are increased in order to increase the Tm.

The above criteria were used in designing the primers used in this project. Amplification and sequencing primers designed previously in our lab (Table 2.1) were also used in this project.

Primer

specificity

was

always

checked

by

using

BLAST

(http://www.ncbi.nlm.nih.gov/BLAST/). 40 nmol oligonucleotides synthesised by Proligo, Australia, were diluted to 200 pmol/μl and stored at –20°C as stock solutions. Dilutions from these stocks were used at a final concentration of 50 and 3.2 pmol/μl for amplification and sequencing respectively. 2.4.2. 16S rRNA Amplification 16S rRNA was amplified from genomic DNA using bacterial specific primers Fd1 (forward primer) and Rd1 (reverse primer). Genomic DNA was prepared by either the CTAB method (Section 2.2.1) or the Phenol-chloroform-isoamylalcohol method

44

(Section 2.2.3). The primer and template were prepared in a conventional master mix using amplification protocols described below. The PCR was used to amplify the 16S rRNA genes from genomic DNA. Reactions were prepared on ice in sterile 0.2 ml thin-wall tubes (Quantum Scientific Products Pty Ltd, Australia). Reactions consisted of 5 μl of 10x Taq buffer (without MgCl2, Promega), 3.5 μl of MgCl2 (25 mM, Promega), 0.5 μl of 20 mM dNTPs (5 mM dATP, 5 mM dGTP, 5 mM dCTP, and 5 mM dTTP), 1 μl of 50 μM Fd1 primer (see Table 2.1), 1 μl of 50 μM Rd1 primer (see Table 2.1), 0.08 μl of 5 U/μl of Taq DNA Polymerase (Promega Corp., USA), 2 μl of 20 ng chromosomal DNA and 37.0 μl of sterile ddH2O. An overlay of 40 μl of sterile mineral oil was added. The PCR was carried out in a Corbett Research Thermal Sequencer (model FTS-1) with the following parameters: 1 cycle of 94ºC for 3 minutes; 30 cycles of 94ºC for 1 minute, 55ºC for 1 minute, 72ºC for 1 minute and 1 second, 1 cycle of 55ºC for 1 minute and 72ºC for 20 minutes. No mineral oil was added to the PCR. Each PCR run contained a negative control (2 μl sterile dH2O instead of template DNA) and a positive control (2 μl of known amplifiable DNA instead of template DNA). A 5 μl aliquot of each PCR was checked by agarose gel electrophoresis. 2.4.3. Purification of PCR products PCR products were purified using a QIAquick® PCR Purification Kit as per manufacturer’s instructions (QIAGEN, Inc., Valencia, CA, USA). 2.4.4. 16S rRNA Sequencing and Sequence Analysis Fluorescent sequence reactions were prepared on ice in sterile 0.6ml tubes. Reactions consisted of 20 ng purified PCR product, 2 μl of 3.2 μM primer (Table 2.1), 8 μl of ABI PRISM® BigDye™ Terminator Cycle Sequencing Ready Reaction Mix (Applied Biosystems, Australia), and sterile ddH2O to a final volume of 20 μl. An overlay of 40 45

μl of sterile mineral oil was added. Thermal cycling was carried out in a Corbett Research Thermal Sequencer (model FTS-1) following the ABI recommended cycling program of 25 cycles of 96ºC for 30 seconds, 50ºC for 15 seconds and 60ºC for 4 minutes. Alternatively, a Rapid Cycler (Idaho Technology, USA) was also used for cycle sequence. For this, the sequence reaction containing 20 ng purified PCR product, 1.5 μl of 3.2 μM primer, 4 μl of ABI PRISM® BigDye™ Terminator Cycle Sequencing Ready Reaction Mix (Applied Biosystems, Australia) was pipetted into a 10 μl capillary tube and the cycling carried out in a Rapid Cycler using the same program as described above. The sequencing products were purified using ABI’s recommended ethanol precipitation for BigDye™ Terminators by adding the 20 μl sequencing reactants to 80μL of 80% ethanol at room temperature. After 15 minutes to < 24 hours, the samples were centrifuged at 17.970 g for 20 minutes. The supernatant was carefully removed by aspiration, 250 μl of 70% ethanol added, recentrifuged for 10 minutes and the pellet dried by heating to 95ºC for 1 minute in a Thermoline heating block (DB-1). Sequence reaction products were stored in the dark at 4ºC prior to electrophoresis on a 4.8% denaturing polyacrylamide gel on an ABI 377 automated DNA sequencer at the Molecular Biology Facility, Griffith University. The DNA sequence data that were obtained were assembled using BioEdit and the assembled sequences analysed using BLAST and FASTA at the NCBI server (http://www.ncbi.nlm.nih.gov/).

46

Table 2.1

Primers used for Amplification and Sequencing of 16Sr DNA.

Name

E. coli position

Sequence (5'-->3')

Fd1

8-27

CAGAGTTTGATCCTGGCTCAG

F1

339-357

CTC CTA CGG GAG GCA GCA G

F1.1

519-536

CAG CAG CCG CGG TAA TAC

F2

783-803

CAG GAT TAG ATA CCC TGG TAG

F3

907-926

AAA CTC AAA GGA ATT GAC GG

F3.1

1100-1115

CAA CGA GCG CAA CCC C

F4

1391-1406

TGT ACA CAC CGC CCG T

F5

1493-1511

AGT CGT AAC AAG GTA ICC G

R1

357-342

CTG CTG CCT CCC GTA G

R2

536-519

GTA TTA CCG CGG CTG CTG

R3

802-785

CCA GGG TAT CTA ATC CTG

R4

926-907

CCG TCA ATT CCT TTG AGT TT

R5

1115-1100

GGG GTT GCG CTC GTT G

Cam-Rev

1443-1464

GGA CGG TAA CTA GTT TAG TAT T

R6

1513-1494

TAC GGT TAC CTT GTTA CGA C

Rd1

1526-1542

AAGGAGGTGATCCAGCC

Note: The 16S rRNA primer sequences [229] and the E. coli numbering is based on Brosius et al. (1981) [230].

47

2.5. Real Time PCR 2.5.1. Design of Adjacent Hybridisation Probes Raw sequence data generated from the 16S rRNA sequences of the 20 Campylobacter and Arcobacter species and subspecies were edited for accuracy and aligned with other 16S rRNA sequences using BioEdit (Hall, 1999). Closely related sequences and representatives from various phyla of domain Bacteria were extracted from the GenBank and Ribosomal Database Project (RDPII) to reveal genus / species-specific sequences as described (Section 2.4.1). Bacteria and/or Campylobacter and/or Arcobacter specific forward and reverse primers were selected. A universal, domain Bacteria specific Cy5 labelled probe, with the potential of binding to the target site within the amplicon produced by the primers, and a 6-FAM labelled genus / species probe specific for binding to the downstream region of the universal Cy5 probe, were designed according to the following general guidelines [110]: •

Probe Tm should be around 10οC greater than the primer Tm's.

•

Labelled at the 3’ end of the upstream probe by fluorescein, which serves as donor in the FRET and blocks the extension from the probes.

•

Labelled at the 5’ in the downstream probe by Cy5, which will serve as acceptor in the FRET then the 3’end will be phosphorated to block the extension of the PCR product.

•

The probes should be separated by one base.

•

The probes should be placed on one strand near the primer on the opposite strand.

The specificity of probes and primers was checked against the GenBank nucleotide sequence database by the FASTA program. The fluoroprobes were made by Proligo, Australia. 48

2.5.2. Design of TaqMan probe TaqMan probes were designed according to the following general guidelines [110]: •

Probes should be 20-40 bp long.

•

GC content should be 40-60%.

•

Probe Tm should be at least 5οC greater than the annealing / extension temperature.

•

Avoid secondary structure in primers, probe and target strands.

•

Probes should not hybridise to the forward and reverse primers.

•

Avoid probes with long runs of a single base.

•

Probes should be labelled with 6-FAM or any other fluorescent dye as a reporter dye at the 5’ end, and with fluorescence quencher or black hole quencher (BHQ1, BHQ-2, and BHQ-3) at the 3’ end. 2.5.3. LightCyclerTM Running Parameter for adjacent hybridisation probe assay

Template DNA was prepared according to Section 2.2 and a 10 μl reaction mix prepared. In order to bring the initial fluorescence reading (about 8) within the 10-point scale of the LightCyclerTM, it was necessary to lower the concentration of the fluoresceinlabelled probe to 0.2 μM, as recommended by Idaho Technology. 10 µl glass capillary tubes were used. Reactions consisted of 1 μl of 10x PCR buffer 20 mM MgCl2 (part # 1771, Idaho Technology, USA), 1 μl of 2 mM dNTPs (0.5 mM dATP, 0.5 mM dGTP, 0.5 mM dCTP, and 0.5 mM dTTP), 1 μl of 5 μM F2 (Forward 16S rRNA primer (Table 2.1)), 1 μl of 5 μM Cam reverse primer (Table 2.1), 1 μl of 2 μM 6-FAMprobe, 1 μl of 2 μM Cy5 probe, 0.08 μl of 5 U/μl of Taq DNA Polymerase (Promega Corp., USA), 1 μl (around 20 ng) of chromosomal DNA, and 2.92 μl of sterile dd H2O. 49

The reaction mix was pipetted into a glass capillary tube and snap-sealed by a plastic cap (part # 1720, Idaho Technology, USA). The PCR was carried out in a LightCycler™. The PCR conditions were as following: 1 cycle of 94°C for 30 seconds and 45 cycles of 94°C for 1 second, 55°C for 30 seconds and 72°C for 30 seconds with a ramp speed of 20°C/second. Fluorescence emissions were monitored and recorded for 100 milliseconds during the annealing step (55°C for 30 seconds) by the LightCycler™. A melting curve to determine the optimal annealing temperature of hybridisation probes was obtained immediately after PCR, by increasing the temperature from 45°C to 94°C at a rate of 0.5°C per second over a 2 minute period with the signal acquisition mode set to be continuous. A personal computer and the accompanying software provided an interface for the operation of the LightCycler™ and analysis of the data. Cycle-to-cycle fluorescence readings were plotted on the computer screen for continuous monitoring of PCR products. At the end of each run, raw fluorescence data were saved to text files, which were analysed with the software or imported into a spreadsheet program for normalisation in order to minimise the effect of sample-to-sample variations. Normalisations were performed by subtracting the minimum value from all data points, the result divided by the maximum value, and multiplied by 100. For verifying PCR results, the reactions were removed from the capillary tubes and examined by electrophoresis on 1% agarose gel in 1x Tris-acetate-EDTA (TAE) buffer. The λ Hind III was used as a size marker. 2.6. Real Time Detection of the Campylobacter Species SYBR Green I 2.6.1. LightCyclerTM Detection of the Hippuricase (hipO) Gene A 10 µl PCR reaction contained 0.5 µM of the Hip-2214F and Hip-2474R primers, 2 ng of genomic DNA (less than 2 ng for templates prepared by the boiling method), 0.25 mM dNTP, 1 μl of (2 mM MgCl2, 50 mM Tris-HCl, 0.25 mg/ml BSA (Sigma A2153))

50

or 1 μl of 10X buffer 20 mM MgCl2 (Part # 1771, Idaho Technology, USA), 0.4 units of Taq DNA polymerase (Promega, Madison, WI), and 1 µl of 1:10,000 dilution of SYBR Green I concentrate (Molecular Probes, Eugene, OR, USA). The dye concentrate is supplied as a viscous DMSO solution, which makes accurate pipetting difficult. It was necessary to adjust the concentration of the SYBR Green I dye empirically so as to obtain a baseline fluorescence reading of about 2 on the 10-point scale of the LightCycler ™. After the reaction mix was loaded into a glass capillary tube and snapsealed by a plastic cap (part # 1720, Idaho Technology), thermal cycling was carried out as follows: initial denaturation at 94°C for 30 seconds, followed by 22 cycles of denaturation at 94°C for 2 seconds, annealing at 55°C for 30 seconds and extension at 72°C for 30 seconds. The temperature transition rate was set at 20°C per second. Acquisition of fluorescence signals from up to 24 samples was carried out at the end of every annealing step for 50 milliseconds. For distinguishing specific products from nonspecific products and primer dimers, a melting curve was obtained immediately after amplification by holding the temperature at 45°C for 20 seconds followed by a gradual increase in temperature to 95°C at a rate of 0.2°C per second, with the signal acquisition mode set at step. For verification of the real-time detection, the PCR products were also examined on a 1.0% agarose gel in Tris-acetate-EDTA buffer. The 100 bp DNA ladder (Promega, Australia) was used as a size marker. 2.6.2. Real Time PCR Detection of hipO Gene using iCycler iQTM A 20-µl PCR reaction contained 2 μl of 5 µM of the Hip-2218F, 2 μl of Hip-2470R primers, 2 μl of genomic DNA (around 4 ng if prepared by CTAB, or less than 4 ng for templates prepared by the boiling method), 2 μl 0.25 mM dNTP, 2 μl of (2 mM MgCl2, 50 mM Tris-HCl (pH 8.3), 0.25 mg/ml BSA (Sigma A2153)) or 2 μl of 10 X buffer 20

51

mM MgCl2 (Part # 1771, Idaho Technology, USA), 0.08 μl of 5U/μl of Taq DNA polymerase (Promega, Madison, WI), and 0.2 µl of 1:10,000 dilution of SYBR Green I concentrate (Molecular Probes, Eugene, OR, USA). Sterile distilled water was added to make the final volume 20 μl. The master mix was thoroughly vortexed and 20 μl aliquots were pipetted into a 96-well PCR plate. After the reaction mix was loaded and sealed with optical-quality sealing film (Part # 223-9444, BioRad, USA), it was briefly spun by hand to bring all the reagents to the bottom of the wells. The plate was placed in the iCycler iQTM detection system and thermal cycling was carried out as follows: initial denaturation at 94°C for 30 seconds, followed by 22 cycles of denaturation at 94°C for 10 seconds, annealing at 55°C for 30 seconds and extension at 72°C for 30 seconds. Acquisition of fluorescence signals from up to 96 samples was carried out at the end of every second annealing step. For distinguishing specific products from nonspecific products and primer dimers, a melting curve was obtained immediately after amplification by holding the temperature at 65°C for 10 seconds. Fluorescence readings were taken after each increase in temperature to 95°C at a rate of 0.5°C per 10 seconds. The camera was set after each increase in the temperature. For verification of the melting curve results, the PCR reactions were also examined on a 1.0% agarose gel in Tris-acetate-EDTA buffer. The 100 bp DNA ladder (Promega, Australia) was used as a size marker. A personal computer and the accompanying software provided an interface for the operation of the iCycler iQTM and analysis of the data. Cycle-to-cycle fluorescence readings were plotted on the computer screen for continuous monitoring of PCR products.

52

2.6.3. iCycler iQTM Triplex Detection of Campylobacter Species All real-time PCR assays were performed in 96-well plates (Part number 223-9441, BioRad, USA) and thermal cycling and fluorescence measured in real-time in an iCycler iQTM. Real-time PCR in which a single TaqMan probe was used in each assay contained 2 μl of 5 µM of the forward primer (TaqCam- 936F or TaqHip-1754F or TaqOrfA-4024F), 2 μl of 5 µM of the corresponding respective reverse primer (TaqCam-1087R or TaqHip1924R or TaqOrf-4129R), 2 μl of genomic DNA (approximately 4 ng if prepared by CTAB, or less than 4 ng for templates prepared by the boiling method), 2 μl 0.25 mM dNTP, 2 μl of 10x buffer 30 mM MgCl2 (1770, Idaho Technology, USA), 0.08 μl of 5U/μl of Taq DNA polymerase (Promega, Madison, WI), and 2μl of TaqMan probe (TaqCam-1034 probe, TaqHip-1864 probe or TaqOrf-4075 probe, respectively). Sterile distilled water was added to a final volume to 20 μl and the reactants mixed, the plates sealed with an optical-quality sealing film (223-9444, BioRad, USA) and briefly centrifuged to collate the mixture to the bottom of the wells. Thermal cycling was carried out using the following conditions: initial denaturation at 94 °C for 30 sec, followed by 40 cycles of denaturation at 94 °C for 30 sec. The optimisation for annealing and extension temperatures, which ranged from 57 °C to 64 °C for 60 sec, was implemented by activating the gradient options of the instrument. Acquisition of fluorescence signals from up to 96 samples was carried out at the end of every second of the annealing and extension step using the excitation and emission filters (Table 2.2). The CCD was set after each increase in temperature. A PC and the accompanying software provided an interface for the operation of the iCycler iQTM and analysis of the data. Cycle-to-cycle fluorescence readings were plotted on the computer screen for continuous monitoring of PCR products.

53

Multiplex real-time PCR reactions were essentially the same as described for the realtime PCR in which a single TaqMan probe was used (described above) except that all the three pairs of primers and probes were used in a single reaction: The reaction mixture contained 1 μl each of 5 µM of the forward primers (TaqCam- 936F, TaqHip1754F and TaqOrfA-4024F), 1 μl each of 5 µM of the reverse primers (TaqCam-1087R, TaqHip-1924R and TaqOrfA-4129R), 1 μl of genomic DNA (approximately 2 ng if prepared by the CTAB method, or less than 2 ng for templates prepared by the boiling method), 3 μl 0.25 mM dNTP, 3 μl of 10x buffer 30 mM MgCl2 (1771, Idaho Technology, USA), 0.12 μl of 5U/μl of Taq DNA polymerase (Promega, Madison, WI), and 1 µl of 2 µM TaqMan probes (TaqCam-1034 probe, TaqHip-1864 probe and TaqOrfAm-4075 probe). Sterile distilled water was added to a final volume 20 μl. The subsequent steps and the thermal cycling parameters used were carried out as described above with the exception that the annealing and extension was performed at 60oC and not by using the gradient option, as this temperature was determined at the optimal temperature.

Table 2.2 Show the iCycler iQTM filters excitation and emission wavelength for FAM, MAX and Cy5

Excitation

Emission

Fluorophores

490/20 nm (Nano meter)

530/30 nm

FAM

530/30 nm

575/20 nm

MAX

635/30 nm

680/30 nm

Cy5

54

2.7. Determining the Sensitivity of the Real Time Assay 2.7.1. Using Incubation Time as Parameter The minimum time that was required for detecting the growth of C. coli and C. jejuni in liquid medium was determined in PCMB. For this, C. jejuni [(ATCC 940565) and (QHSS OOM2177)] and C. coli [(NCTC 11366) and (QHSS 98M729)] were streaked onto BAPs. Five colonies from each plate was picked after 24 h incubation, resuspended in 50 ml PCMB and incubated at 42°C under anaerobic conditions. Aliquots were removed at intervals of 0, 4, 6, 8, and 24 h during growth, the samples placed on ice, and DNA extracted by the rapid boiling method. 1 ml of each culture was taken and the DNA extracted by rapid boiling methods (Section 2.2.2) and PCR run as described previously (Section 2.5.3 and 2.6.2) 2.7.2. Using Different DNA Concentration To determine the sensitivity of the assay used in this project, a DNA sample was harvested by CTAB assay and phenol-chloroform-isoamylalcohol (25:24:1) from the bacterial cultures of C. coli, C. jejuni and A. butzleri. After that, serial dilution of the purified DNA was used to check the sensitivity of the assay. 2.7.3. Nucleic Acid Quantitation The concentration of DNA in solution was determined by spectrophotometry using a Cintra 20 UV-Visible spectrometer (GBC Scientific Equipment Pty Ltd, Australia). The absorbance was measured at 260 nm, while the purity of a nucleic acid sample was determined by measuring the ratio of the absorbance at 260 nm (A260) and 280 nm (A280), as previously described [91]. Each OD unit at A260 nm corresponds to 50 μg/ml of double stranded DNA or 33 μg/ml of single stranded DNA [231]. Alternatively, the DNA concentration of solutions was estimated by visualisation on an agarose gel

55

relative to known DNA standards. A pure DNA solution has an A260:A280 ratio of 1.8 ± 1. 2.7.4. Using Different Cell Concentration The detection limits of the real-time assays to cell numbers of C. coli and C. jejuni was determined using the following approach. C. jejuni [(ATCC 940565) and (QHSS OOM2177)] and C. coli [(NCTC 11366) and (QHSS 98M729)] were streaked onto BAPs, and after 24 h incubation a colony from each plate was picked and resuspended in 100 μl TE buffer. 75 μl aliquots were used to extract DNA using the rapid boiling method and the remaining 25 μl aliquots were serially diluted up to 10-12 and 10 μl spread onto BAP to determine colony forming units (CFU). This allowed the estimation of the numbers of cells used in the real-time PCR assay. 2.8. Terminal Restriction Fragment Length Polymorphism (T-RFLP) for Detection and Identification of Campylobacter and Arcobacter Species 2.8.1. PCR Amplification and Purification of the Products 16S rRNA fragments of Campylobacter and Arcobacter type cultures and chicken isolate were amplified using a Fam-49F forward primer (6-FAM on the 5' end, Proligo, Australia) and a universal 16S rRNA gene universal reverse primer (R2), producing a 482 bp fragment as described in section 2.4.2. PCR products were purified as described in section 2.4.3. 2.8.2. Multiplex PCR Amplification for T-RFLP A 2 μl of 1:1 boiled DNA sample (section 2.2) from different Campylobacter and Arcobacter enrichment cultures initiated from chicken sample, as described in section 2.1.1, was used as a template for PCR reaction as described in section from 2.8.1.

56

2.8.3. Restriction Endonucleases Digestion Five restriction enzymes EcorI (5’…G^AATC…3’), KpnI (5’…GGTAC^C…3’) (Fermentas, USA), BtsI (5’…GCAGTGNN^…3’), Sau3A (5’…^GATC…3’), and DdeI (5’…C^TNAG…3’)(New England BioLabs, USA) were used for digestion of the16S rRNA PCR fragments that had been amplified as described above. These restriction enzymes were the most suitable for detecting and differentiating Campylobacter and Arcobacter species as determined by TAP simulation analysis. Restriction enzyme digestion (total volume 20 μl) was performed according to the manufacturer’s instructions. 2.8.4. Recovery of the DNA by Ethanol Precipitation The T-RFLP products that ensued from the digestion were purified by adding 16 µl of sterile distilled water and 64 µl 100% ethanol to the entire 20μl reactant. The mixture (100 μl) was vortexed briefly and incubated at room temperature for 15 minutes before centrifuging (Sigma, USA) at 17.970 g for 10 minutes. The supernatant was removed carefully by aspiration, and the precipitant containing the T-RFLP products washed by adding 250 μl of 70% ethanol and again centrifuged at 17.970 g for 10 minutes. The supernatants were removed by aspiration, and the pellet dried at 60ºC for 15 minutes in a thermal cycler (Corbet Research, Australia). If necessary, T-RFLP products were stored in the dark at 4ºC prior to electrophoresis on a 4.8% denaturing polyacrylamide gel. 2.8.5. T-RFLP Product Detection and Analysis Detection of T-RFLP products was performed on an Applied BioSystems 377 DNA sequencer with 96-lane upgrade (Applied BioSystems, Australia). Samples were prepared by combining 2.0 µl of purified T-RFLP product, 2.5 µl of deionised formamide, 0.5 µl of 25mM EDTA (pH 8.0) containing 5% (wt/vol) blue dextran, and

57

0.50 µl GeneScan®-500 TAMRA. The standard contains single strands of DNA with a single TAMRA fluorophore GeneScan®-500 (TAMRA) internal Standard (PE applied BioSystems, Australia) with a size range of 35, 50, 75, 100, 139, 140, 150, 200, 250, 300, 350, 400, 450, 480, and 500 bp. Samples were mixed by pipetting, denatured at 94○C, and immediately cooled to 4○C. 1.0 µl aliquots were loaded onto a 0.2mm thick comb on 96 well-to-read plates, 4.8 denaturing polyacrylamide gel and electrophoresed at 51○C for 4 hours with limits of 3 kV and 60mA. Data were collected in GeneScan mode, and the lengths of control and sample TRFs were calculated by comparison with the internal standard by using the local southern method [232]. 2.9. LDR for Detection, Identification of Campylobacter and Arcobacter Species 2.9.1. PCR Amplifications and Product Purification The DNA region coding for the 16S rRNA was amplified with universal primers F2 and Rd1, producing a 759 bp fragment. Within the 759 bp is a hyper variable region, which discriminates Campylobacter and Arcobacter species. The DNA region coding for the FlaA gene was amplified using LDRFla-7 primer and LDRFla-630 primer, to produce a 643 bp amplicon. PCR reaction was performed using an iCycler iQTM thermal cycler, as described in section 2.4.2. Following PCR, 5 µl of proteinase K (1 mg/ml) was added to the PCR mixture, the reaction heated at 70°C for 10 minutes and then at 94°C for 15 minutes to denature Taq DNA polymerase enzyme activity. PCR products were purified as described in section 2.4.3. Purified PCR products were eluted in 50 μl of TE pH 7.4 and checked by agarose gel electrophoresis. 2.9.2. LDR Product Detection and Analysis Ligation reaction was carried out in a final volume of 20 μl containing 1.5 μl of 10x Taq DNA Ligase Buffer (20 mM Tris-HCl (pH 7.5), 25 mM potassium acetate, 10 mM

58

magnesium acetate, 0.1% Triton X-100, 1mM NAD, 10 mM DTT (New England BioLabs, England), 2 pmol of each discriminating probe (LDRArco, LDRCam, LDRJejuni, LDRColi, and LDRLari), and common probe (LDRProteo common probe and LDRFlaA common probe) 100–500 fmol of purified PCR products and 1 µl of 40 U/µl T4 DNA ligase (New England Bio lab, England). The LDR was cycled for 20 rounds of 94°C for 30 seconds and 66°C for 4 minutes in an iCycler iQTM thermal cycler. After cycle completion, 3μl of 100mM EDTA was added to the LDR product to prevent any further DNA ligase activity. LDR products were purified as described in Section 2.8.3 and detected using an applied BioSystems 377 DNA sequencer with 96-lane upgrade (Applied BioSystems, Australia) as described in Section 2.8.5. 2.10.

Comparison between Culture methods and Molecular

methods (Gold Standard) In this project different molecular methods have been developed for rapid detection and identification of Campylobacter and Arcobacter from cultured isolates or directly from enrichment cultures initiated from chicken samples, to determine the efficiency of the molecular methods compared with conventional culture methods based on sensitivity and specificity. Accuracy is defined as the fraction of a measured value to a reference value taken as a "gold standard,". In this study, molecular methods are the reference method (Ref). Accuracy is calculated as culture positives / Ref positives. Sensitivity is calculated as 1 – false negative rate, while specificity is calculated as 1 – false positive rate. A false negative rate is calculated as the number of culture methods (Negative), Ref (Positive) samples / Total Ref (Positive), whereas a false positive rate is calculated as the culture method (Positive), Ref (Negative) samples / Total Ref (Negative) samples.

59

A. nitrofigilis Arcobacter (Chapter 5)

Campylobacter (Chapter 3)

A. skirrowii

Adjacent hybridization Probe

C. coli, and C. lari

A. butzleri C. coli, C. lari and C. jejuni

Adjacent hybridization probe

SYBR Green I (Chapter 4)

C. jejuni

Other Campylobacter

Real Time PCR (Chapters 3-7)

A. butzleri A. butzleri A. skirrowii Campylobacter species (Chapter 6)

Adjacent hybridization probe (Multi FAM probe)

A. skirrowii * Campylobacter group Other bacterial species

Campylobacter species. (Chapter 7)

TaqMan probes Multiplex

C. coli C. jejuni Other Campylobacter species

T-RFLP (Chapter 8)

Campylobacter Species

Campylobacter and Arcobacter species

EcoRI

Sau3A1, DdeI, BtsI and KpnI

C. coli, C. jejuni and C. lari

Arcobacter species

DNA Finger- printing (Chapter 8 and 9)

Formatted

C. Coli

LDR (Chapter 9)

Campylobacter and Arcobacter species

16S rRNA gene

Campylobacter species

FlaA gene

C. jejuni C. lari

Arcobacter species

Figure 2.1 A schematic overview of the project is shown in which two approaches namely PCR and fingerprinting, culminating in three major techniques (real-time PCR, LDR and T-RFLP), were used for identifying and differentiating Campylobacter and Arcobacter species. Chapter 2 describes the material and methods, while chapters 3 to 9 contains the results and their discussion. The conclusions from these studies and future directions are outlined in chapter 10. Chapters 3 to 7 describe five different real-time PCR, while chapters 8 and 9 describe two fingerprinting assays (T-RFLP and LDR), which differentiate Arcobacter and Campylobacter as well as discriminate species of these genera.

60

61

3. Chapter 3 Rapid Identification of C. jejuni and C. coli by Targeting the 16S rRNA using Real-Time PCR 3.1. Chapter Overview This chapter describes the development of a rapid and sensitive real-time PCR method, targeting the 16S rRNA gene, for the detection and identification of C. jejuni, C. lari and C. coli, the most common and routinely isolated pathogenic members of the genus Campylobacter (Figure 2.1). A summary of this work is described below and the methods used are described fully in Chapter 2. Five species of Campylobacter (C. coli (QHSS B287), C. jejuni (QHSS 00M2260), C. upsaliensis (QHSS 99M126), C. lari (ATCC 35223T) and C. hyointestinalis (QHSS 99M2318), which had been identified serologically, were cultured, the DNA extracted using the CTAB method (Section 2.2.1), and their identity confirmed as members of the genus Campylobacter by partial sequencing of the 16S rRNA genes. The real-time PCR was developed using the DNA of the four species. The real-time PCR method was validated by using the DNA template from 176 Campylobacter isolates and 30 enrichment cultures derived from chicken samples. Furthermore, E. coli, Staphylococcus aureus, Clostridium perfringens, Micrococcus luteus, and Pseudomonas aeruginosa were used as negative control. 3.2. Results 3.2.1. Detection

and

Identification

of

Campylobacter

Species

by

Conventional Culture Methods Australian standard methods of colony appearance and biochemical testing (Appendix 3) were used to detect and identify the presence of Campylobacter species in 62

Campylobacter enrichment cultures initiated from chicken samples. The appearance of colonies on Campylobacter selective media at 42οC indicates the presence of a thermophilic Campylobacter species. The identification is usually further confirmed by performing rapid catalase, which is positive for C. jejuni, C. coli and C. lari. Further biochemical tests (Table 3.1 and appendix 4) are performed to identify species such as C. jejuni, C. coli and C. lari [91, 125, 233]. Of the 30 samples tested using the conventional culture methods described in section 2.1.1, 25 (83%) samples were positive and 5 (17%) samples were negative for Campylobacter. Identification to the species level revealed that 17 (57%) isolates were identified as C. jejuni, 7 (23%) were identified as C. coli and 1 (3%) was identified as C. lari (Appendix 5).

Table 3.1 Biochemical tests used to distinguish between closely related C. jejuni, C. coli and C. lari according to the Australian standard methods.

C. jejuni

C. coli

C. lari

Nalidixic Acid

Sensitive

Sensitive

Resistance

Cephalothin

Resistance

Resistance

Resistance

Hippurate hydrolysis

Positive

Negative

Negative

3.2.2. Identification of Isolates (QHSS B287), (QHSS 00M2260), (QHSS 99M126) and (QHSS 99M2318) as C. coli, C. jejuni, C. upsaliensis and C. hyointestinalis by Partial Sequencing of 16S rRNA As expected, DNA from all four isolates [C. coli (QHSS B287), C. jejuni (QHSS 00M2260), C. upsaliensis (QHSS 99M126) and C. hyointestinalis (QHSS 99M2318)] produced an amplicon of approximately 1.5 kb with primers Fd1 and Rd1. The following primers, F2, F3, F4, F5, R3, R4 and Rd1 (Table 2.1) were then used to 63

sequence approximately 759 nucleotides of the 16S rRNA corresponding to the position 783 to 1542 (E. coli numbering) from the amplicons. This region was chosen as it contains homologous, variable and hyper variable tracts of nucleotides useful in differentiating Campylobacter species (Figure 3.1). BLAST analysis identified isolates B287, 00M2260, 99M126 and 99M2318 as C. coli, C. jejuni, C. upsaliensis, and C. hyointestinalis respectively. The tentative identification of the isolates concurs with the 16S rRNA based molecular analysis. Therefore, the use of these sequences and other database sequences strengthened confidence in sequence accuracy and therefore the subsequent design of fluoroprobes for real-time PCR assays. 3.2.3. Development

and

Design

of

PCR

Primers

and

Adjacent

Hybridisation Probes 16S rRNA sequence alignments of representative members of domain Bacteria and Campylobacter species identified several regions that could be useful for amplification and as targets for the hybridisation of fluorescent probes. A 16S rRNA universal forward primer designated F2 (5’-ATCTAATGGCTTAACCATTAAAC-3’) and a C. jejuni, C. coli and C. lari specific reverse primer designated primer Cam-Rev (5’AATACTAAACTAGTTACCGTC-3’) [91], targeting positions 783 and 1464 respectively (according to the E. coli numbering [230]) were selected to produce a 681 bp amplicon. Two fluoroprobes with the potential of targeting adjacent regions within the 681 bp amplicon were then identified. A domain Bacteria specific designated probe Cy5+1046 (5’-Cy5-AGGTGITGCATGGITGTCGTTGTCGTGT-PO4-3’), [232, 234] targeting position 1075-1046 and a 6-FAM labelled Jejuni-coli probe (5’GTGCTAGCTTGCTAGAACTTAGAGA-FAM-3’) targeting position 1019-1044 were designed (Figure 3.2). GenBank database searches using BLAST and FASTA indicated that the combined use of the forward and reverse primers would specifically amplify the 16S rDNA of C. 64

jejuni and C. coli-C. lari producing an amplicon of 681 bp. FASTA analysis confirmed our previous results that probe Cy5+1046 was a universal probe that would bind to all members of domain Bacteria [232, 234] due to the inosine introduced degeneracy at positions 1051 and 1059 and that probe Jejuni-coli would only hybridise to the Campylobacter group (C. jejuni, C. coli, C .lari, C. hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae, and C lanienae) (Figure 3.2). 3.2.4. Real Time Detection of C. jejuni and C. coli Using Adjacent Hybridisation An increase in fluorescence was observed during real-time PCR with DNA templates prepared using the CTAB method and the rapid boiling methods for C. jejuni, C. coli and C. lari but not for C. hyointestinalis, C. upsaliensis and E. coli (Figure 3.3). Additionally, Staphylococcus aureus, Clostridium perfringens, Micrococcus luteus, and Pseudomonas aeruginosa, were also tested and showed negative results (result not shown). This result was expected and in accordance with the theoretical primer and probe data. The Tm determined from the dissociation of the hybridised fluoroprobes from the target at the end of PCR was 65°C for C. jejuni, C. coli and C. lari (Figure 3.4) and was further proof of fluoroprobe specificity. Subsequent agarose gel electrophoresis showed that 681 bp amplicons were present from the DNA of C. coli, C. jejuni and C. lari but not from C. hyointestinalis, C. upsaliensis, C. fetus and E. coli (Figure 3.5), and

65

Figure 3.1 The partial sequences of 16S rRNA (positions 817 to 1466 E. coli numbering [230])of the four Campylobacter species from QHSS (C. coli from chicken (QHSS P287) and C. jejuni (QHSS 00M2128, human source), C. upsaliensis (QHSS 99M126, human source) and C. hyointestinalis (QHSS 99M2318, human source) compared with database sequences of C. hyointestinalis (ATCC 35217, swine source) C. coli (CCUG 11283, porcine source), C. jejuni (CCUG 24567, human source ) and C. upsaliensis (CCUG 14913, from Canine faeces).

66

Escherichia coli----------------------------------GTG—--CCTTCGGG—AACCGTGAGACAGGTGCTGCATGGCTGTCGTCAGCTCGTGT---------------------C. sputorum

ATCC 33491---------------------------GTGTCTGCTTGCAGAAATGTTAAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT ---------------------

C. gracilis ATCC33236------------------------ ----GTGTCTGCTTGCAGAAATGTTAAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. concisus ATCC33237-----------------------------GTGTCTGCTTGCAGAAATGTTAAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. curvus ATCC35224-------------------------------GTGTCTGCTTGCAGAAATGTTAAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. mucosalis CCUG6822-----------------------------GTGCTAGCTTGCTAGAATGTTGAGACAGGTGCTNCACGGCTGTCGTCAGCTCGTGT --------------------C. rectus ATCC33238-------------------------------GTGTCTGCTTGCAGAAATGTTAAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. hyointestinalis ATCC35217----------------------GTGCTTGTTTACAAGAAATTAGTGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. fetus ATCC19438--------------------------------GTGCTAGCTTGCTAGAAAGTTGAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. upsaliensis CCUG14913C ------------------------GTGCTAGCTTGCTAGAATGTTGAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. helveticus NCTC12470---------------------------GTGTCTGCTTGCAGAAATGTTAAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. showae CCUG11641-------------------------------GTGTCTGCTTGCAGAAATGTTAAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. lari NCTC11352---------------------------------GTGCTAGCTTGCTAGAACTTAGAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. coli CCUG11283---------------------------------GTGCTAGCTTGCTAGAACTTAGAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------C. jejuni CCUG24567-------------------------------GTGCTAGCTTGCTAGAACTTAGAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT --------------------Jejuni-coli Probe---------------------------------GTGCTAGCTTGCTAGAACTTAGAGA Cy5 +1046 Probe-------------------------------------------------------------AGGTGITGCACGGITGTCGTCAGCTCGTGT

CAGGATTAGATACCCTGGTAG Primer F2 (783 – 805)

Jejuni-coli probe 5’ 1019-1044

Probe Cy5+1046 FAM 3’

5’Cy5

1046-1075

GGGGTTGCGCTCGTTG 3’P

Primer Cam-Rev (1443-1464)

Figure 3.2 An alignment of the partial 16S rDNA sequences corresponding to position 783 to 1464 (E. coli numbering [230] of representative of Campylobacter species is shown, this region is amplified using primers F2 and Cam-Rev (indicated by left and right hand arrows) to produce a 681 bp amplicon. The universal fluoroprobe Cy5+1046 (specific for members of domain Bacteria) and the probe Jejuni-coli (specific to C. coli, C. jejuni and C. lari) bind adjacent to each other within the target regions of the amplicon.

67

Figure 3.3 An increase in fluorescence during specific adjacent hybridisation of the fluoroprobe 6-FAM Jejuni-coli and fluoroprobe Cy5 1046+ to the target sites of the 16S rDNA amplicons of C. jejuni (ATCC 940565) (-■-), C. lari (ATCC 35223) (-●-) and C. coli (NCTC 11366) (-▲-) during PCR as measured in the LightCyclerTM. No increase was observed for C. hyointestinalis (QHSS 99M2318) (-×-), C. upsaliensis (QHSS 99M126) (-X-), E. coli (- -) or a negative control lacking a DNA template (+-). The DNA templates were prepared using the CTAB method.

Figure 3. 4. The melting peaks generated from the dissociation of the fluoroprobes from C. coli (NCTC 11366) C. lari (ATCC 35223) and C. jejuni (ATCC 940565) amplicon target sites at the end of run of real-time PCR produced a Tm of 65oC. But as expected a Tm for C. hyointestinalis, C. upsaliensis, E. coli and a negative control lacking template was not produced.

68

Figure 3.5 Agarose gel electrophoresis of PCR reaction products from C. upsaliensis (Lane 1), C. hyointestinalis (Lane 2), E. coli (Lane 3), C. coli (Lane 5), C. jejuni (Lane 6) C. lari (Lane 7) and a negative control lacking DNA template (Lane 8). Only C. coli C. lari and C. jejuni but none of the others produced the amplicons of 681 bp as expected. Molecular weight marker (100 bp) (Promega, USA) was used for base pair size comparison (Lane 4).

is in line with expectation as the primer pair (F2 and Cam-Rev) was specific for the former but not the latter. These results confirmed the excellent discriminatory ability and specificity of the adjacent hybridisation probes assay system. Primer dimer and non-specific PCR products were present in the PCR products generated from the three different kinds of DNA extracted templates used in this assay. However, these non-specific products did not adversely affect hybridisation probe specificity. Different pairs of primers were tested but non-specific PCR products still appeared, which indicated that primer dimer was not the cause of the non-specific products. 3.2.5. Identification of C. jejuni Isolates in the Culture Collection Using the Real Time Adjacent Hybridisation Probes Assay Out of 176 pure culture isolates, 165 isolates were identified as C. jejuni, C. coli and C. lari group (Appendix 1). Furthermore, 26 out of 30 enrichment cultures initiated from chicken samples successfully identified the presence of C. jejuni, C. coli and C. lari (Appendix 7). 69

3.2.6. Comparison between Conventional Culture Assay and Adjacent Hybridisation Probes Real-time PCR Assay Sensitivity and specificity are two major parameters (besides rapidity, simplicity and the economic aspects) that could determine if the detection methods are able to replace the available methods or not. All molecular methods (real–time PCR, T-RFLP and LDR) used in this thesis gave similar sensitivity, specificity and accuracy, and therefore can be used as references for comparison with the conventional culture methods. Real-time PCR methods used for the detection and identification of member of the genera Campylobacter showed 96% accuracy. The false negative rate was 4% and the false positive rate was 0%. Therefore, the sensitivity and specificity of the assay is 96 % and 100 % respectively. Furthermore, the real-time PCR method developed in this chapter provides quantitative detection of DNA fragments that is highly specific to Campylobacter species in the QHSS culture collection and enrichment cultures initiated from chicken samples. 3.2.7. Real-Time PCR Detection of C. jejuni from Different DNA Concentrations CTAB prepared DNA (Section 2.2.1) with concentrations between 192 fg/μl to 192 ng/μl could be used in real-time PCR assays (Figure 3.6). DNA concentration and purity were estimated as described in section 2.7.3 Primer dimer and non-specific PCR products were present in PCR products generated from the three different kinds of DNA extracted templates used in this assay. However, these non-specific products did not adversely affect hybridisation probe specificity. Different pairs of primers were tested but non-specific PCR products still appeared, which indicated that primer dimer was not the cause of the non-specific products.

70

With a DNA concentration of 192 ng/μl μg the increase in fluorescence started at cycle 14 and in the case of a DNA concentration of 192 fg/μl, the initial increase in fluorescence started at cycle 28. A standard curve based on known concentrations of DNA purified from a culture of C. jejuni was used for real-time PCR quantification. A standard curve was constructed after real-time PCR amplification of seven different DNA concentrations, ranging from 192 fg/μl to 192 ng/μl of DNA. The linear correlation coefficient of the standard curve was R2 = 0.966, demonstrating that the assay is suited for quantitative measurements (Figure 3.7). As expected, PCR amplicons were observed on agarose gels with all DNA concentrations. Furthermore, the highest DNA concentration showed the most fluorescence and the lowest the least fluorescence (Figure 3.8). A range beyond these concentrations was not tested but published reports indicate that DNA concentrations as low as 5 fg/μl, 6 fg/μl, 20 fg/μl, 33 fg/μl can be detected [235], [234], [236] and [237] respectively.

70

Normalized fuoresence

60

Error!

50 40 30 20 10

43

40

37

34

31

28

25

22

19

16

13

10

7

4

1

0 Cycle number

Figure 3.6 Real-time detection of C. jejuni CTAB-purified DNA at different concentrations. 192 ng/μl (-■-), 19.2 ng/μl (-▲-), 1.92 ng/μl (-+-), 192 pg/μl (-●-), 19.2 pg/μl (-×-), 1.92 pg/μl (-∆-) and 192 fg/μl (-X-).

71

y = -0.9927Ln(x) + 20.648 R2 = 0.966

35

Cycle threshold (C t)

30 25

0.000192 0.00192 0.0192

20

0.192

1.92 19.2

15

192

10 5 0 1.00E-04

1.00E-02

1.00E+00

1.00E+02

1.00E+04

Log10 DNA concentration

Figure 3.7 Real-time PCR analysis of 10-fold serial dilutions of C. jejuni genomic DNA isolated using CTAB methods. Quantitation was performed by determining the cycle threshold (Ct). Ct values were then plotted against the log10 of DNA concentration. The dilution series ranges from 192 fg/μl to 192 ng/μl of C. jejuni DNA. Two microliters was used for each PCR reaction. The diagram shows one representative experiment. The straight line, which was calculated by linear regression [Ct =-0.9927 log10 (DNA concentration)+ 20.648], shows an R2 of 0.966. (The experiments were repeated three times with identical results.)

Figure 3.8 Agarose gel electrophoresis of different DNA concentrations 192 ng/μl (Lane 1), 19.2 ng/μl (Lane 2), 1.92 ng/μl (Lane 3), 192 pg/μl (Lane 4), 19.2 pg/μl (Lane 5), 7 1.92 pg/μl (Lane 6), and 192 fg/μl (Lane 7) shows an expected 681 bp amplicons. The DNA 100 bp ladder was used as molecular markers (Promega, Australia) (Lane 8).

72

3.2.8. Real-Time PCR Detection of C. jejuni and C. coli from Single Colony Serial Dilutions Cell numbers were determined as described in section 2.7.4. It was found that dilution beyond 10-6 did not show any growth. As few as 125 cells were detected by the assay when a DNA template was extracted by the rapid boiling or CTAB method from a serial dilution of a single colony of C. jejuni or C. coli, indicating that the method was very sensitive. As shown in figure 3.9, the DNA extracted from 500000 cells shows an increase in fluorescence starting at cycle 14 and in the case of 125 cells the initial increase started at cycle 28. The square regression coefficients (R2 = 0.9994) showed a good correlation between the cell number and Ct values, demonstrating the accuracy of the real-time PCR-based quantification (Figure 3.10). Agarose gel electrophoresis showed 681 bp amplicons with all cell numbers as expected (Figure 3.11). Furthermore, the highest cell number showed the most fluorescence and the lowest the least fluorescence. These results are similar to other studies where 100 Leptospira cells were detected [116]. Though a number of reports have suggested that single cells can be detected, this requires significant optimisation and may not be suitable for use in routine high-throughput diagnostic laboratories. In our hands, increasing the sensitivity of the assay also leads to unacceptably noisy signals, which could lead to result misinterpretation.

73

Figure 3.9 A single colony of C. jejuni (ATCC 940565) was serially diluted, DNA extracted by the rapid boiling method and used in real-time adjacent hybridisation assays. The numbers of cells in each dilution was determined by plating onto BAPs. 500000 cells (-▲-), 50000 cells (- -), 5000 cells (-×-), 500 cells (-∆-), 125 cells (-+-) and Negative control lacking template (-♦-).

Cycle threshold (Ct)

30 25

y = -1.7022Ln(x) + 36.41 1.25E+02 2 5.00E+02 R = 0.9994 5.00E+03

20

5.00E+04

15

5.00E+05

10 5 0 1.00E+ 1.00E+ 1.00E+ 1.00E+ 1.00E+ 1.00E+ 1.00E+ 00 01 02 04 05 06 03 Log10 cells number

Figure 3.10 A Standard curve for C. jejuni genomic DNA for 16S rRNA quantitative realtime PCR using adjacent hybridisation probes. Ct values were plotted against log10 cell numbers. The straight line, which was calculated by linear regression [Ct =-1.7022 log (cells number)+ 36.41], shows an R2 of the culture of C. jejuni serially diluted 10-fold is 0.9994. One microliter of DNA was used in a 10 μl PCR reaction. (The experiments were repeated three times with identical results.)

74

Figure 3.11 Gel analysis of PCR amplification products with one colony serial dilution of C. jejuni derived from 16S rRNA primer set F2 and R5. 500000 cells (Lane 1), 50000 cells (Lane 2), 5000 cells (Lane 3), 500 cells (Lane 4), 110 cells (Lane 5), and negative control lacking template (Lane7). A 100 DNA ladder was used for size evaluation (Promega, Australia) (Lane 6).

3.2.9. Real Time Detection of C. jejuni and C. coli during Different Growth Phases The earliest point at which C. jejuni, C. coli and C. lari were detected from growth in broth cultures was 8 hours after incubation with no fluorescence detected in samples taken prior to this time (Figures 3.12 and 3.13). 3.2.10. Improving Rapidity of the Real Time PCR Assay for the Detection of C. coli and C. jejuni Some reports have indicated that rapid PCR can be run in 15 minutes or less in a RapidCyclerTM [115, 118, 119, 127]. The LightCyclerTM is a sister machine of the RapidCyclerTM. Hence, modifications were introduced and the assay time was decreased from 65 minutes to 22 minutes (Figure 3.14). This was achieved by decreasing the annealing and extension time from 30 seconds each to 10 seconds each. However, the data produced using the modification were very noisy, possibly due to the reduced time, which interfered with the annealing and extension time during amplification and probe binding, and hence this line of research was not pursued further.

75

Figure 3.12 C. jejuni (ATCC 940565) was grown for periods of up to 24 hours, DNA was extracted by the rapid boiling method and used in real-time adjacent hybridisation assay. 24 hours incubation (-♦-), 8 hours (-■-), 6 hours (-▲-), 4 hours (-●-), 2 hours (-□-) and 0 hours (-●-).

Figure 3.13 Real-time PCR detection of C. coli during different growth phases. 24 hours incubation (-♦-), 8 hours (-■-), 6 hours (-▲-), 4 hours (-*-), 2 hours (-□-) and 0 hours (-●-).

76

Figure 3.14 Real-time PCR of different DNA concentrations of C. jejuni prepared by the CTAB method and after the reduction of the amplification time to 22 minutes. DNA concentrations were 19.2 ng/μl (-♦-), 1.92 ng/μl (-■-), 192 pg/μl (-▲-) and 19.2 pg/μl (-●-).

3.3. General Discussion C. jejuni, C. coli, and C. lari are among the most frequently isolated food-borne pathogens associated with human illness. Campylobacter species occur naturally in the intestines of birds, livestock, household pets and other environmental sources such as water. Contaminated food products of poultry and cattle and sheep, etc. (red meat) are known to be the major route of human infections. It has been reported that 11 out of 57 Campylobacter infection outbreaks that occurred between 1978 and 1986 were waterborne [238]. An early diagnosis of gastrointestinal disease is of critical importance for the management of disease control, as there is a wide group of bacterial species that could be the causative agents, and at the same time these species could be resistant to multitude of anti-microbial substances. Culture methods used for detection of Campylobacter species have several disadvantages including: time-consuming nature, lack of sensitivity and specificity, and being technically demanding. Due to these disadvantages molecular methods, which are highly sensitive and specific as their 77

detection is based on specific bacterial sequence signatures, has become attractive. Since the inception of real-time PCR 8 years ago, these methods for detection of foodborne pathogens have been recognised as valuable diagnostic procedures [115]. The availability of commercial real-time PCR kits (e.g. LightCycler foodproof Salmonella Detection Kit, and the LightCycler Listeria monocytogenes Detection Kit, etc.) has alleviated problems typically associated with new technology, such as specificity, sensitivity and accuracy compared to conventional culture methods. Development of an easy, rapid real-time PCR assay that is able to detect C. jejuni, C. coli and C. lari is vital for the identification of these species from clinical, environmental and food samples. This will help to reduce the cost of the treatment and aid in the prevention of the diseases. For this reason, adjacent hybridisation probe assays have been described in this chapter. The difference in the Tm between the primer pair and the adjacent hybridisation probes enables the probes to hybridise with the target regions of the amplicons during the annealing and extension segment of each PCR cycle. This will cause an increase in the Cy5 fluorescence emission (F2) with a concomitant decrease in fluorescence emission (F1) (measured as a ratio of F2/F1). Accordingly, an increase in the F2/F1 ratio was observed for all used DNA templates prepared by the three different DNA extraction methods. C. coli, C. jejuni and C. lari were detected using the Jejuni-coli probe but not with any other closely related probe of Campylobacter species or E. coli (Figure 3.3). The assay used in this project was based on continuous monitoring of PCR products using fluorogenic probes. Each run consisted of 45 cycles (Section 2.5.2) and took an average of 62 minutes to complete. It was found that the same method could be run in a shorter cycle time of about 22 minutes but a noise problem appeared which interfered with the results, as shown in Figure 3.13.

78

In most runs, an increase in the fluorescence could be detected after 12 cycles. As mentioned previously, the LightCyclerTM detection procedure took an average of 62 minutes. In contrast, if the same sample were run in a conventional PCR heat block machine, it would take around 150 minutes plus another 35 minutes for the agarose gel electrophoresis. Rapid Cycle PCR and continuous monitoring of fluorescence is achieved because of the fast temperature transition in air, the use of thin-walled glass capillaries for holding samples (which ensures efficient heat transfer) and the built-in fluorometer for precise optical measurement of hybridisation signals. For bacterial DNA prepared by boiling, an increase in the F2/F1 ratio was also observed, and agarose gel electrophoresis showed the expected PCR product. The results from DNA prepared by CTAB or phenol chloroform: isoamylalcohol can be used as a template for real-time PCR. The simple and quick boiling DNA extraction method is, however, very effective and time saving for use in the real-time PCR assay. Our ultimate goal of this project was to develop a rapid identification method for C. jejuni, C. lari and C. coli. Further work is needed to apply the methods described in this chapter to clinical specimens. The real-time PCR assay relies on the hybridisation of fluorogenic probes and/or fluorescent dyes such as SYBR Green I to PCR products. The method used in this chapter offers several advantages such as simplicity, specificity and ease of use and is also a candidate for high-throughput in a similar manner to that for cantilever, and microarray techniques, and therefore has the potential to replace and/or complement existing methods. In addition, the wide range of fluorescent labels [239] and probe technologies [240] available in the market today makes real-time PCR an extremely attractive diagnostic tool. A number of papers describing Campylobacter detection by real-time PCR have been published recently. Cirattoli et al. (2002) [241] and Dionisi et al. (2004) [242] used adjacent hybridisation probes to detect and screen for ciprofloxacin (gyrA gene) resistance in Italian strains of C. coli whereas Best et al.

79

(2003) [243] used TaqMan hydrolysis probes to target membrane-associated protein (mapA) and periplasmic substrate binding protein (ceuE) genes, and LaGier et al (2004) [126] used hipO and serine hydroxymethyltransferase (glyA) genes to distinguish between C. jejuni and C. coli respectively. Yang et al. (2003) [244] used TaqMan hydrolysis probes and Perelle et al. (2004) [240] used adjacent hybridisation probes against 16S rRNA genes. These reports differ in the probing techniques, the instrument, the gene or the target regions selected within the gene from the method described in our report. We have developed a real-time PCR adjacent hybridisation assay, which detects C. jejuni, C. coli and C. lari. 3.4. Conclusion In order to prevent human infection by Campylobacter species, it is necessary to prevent contaminated food (contaminated poultry and red meats) from reaching the consumer. The major drawback of conventional culture methods is that enrichment and detection procedures require up to 5 days, but fresh poultry, for example, is usually processed and becomes available to costumers within 24-48 hours [245]. Real-time PCR has been used in this study for the rapid detection of C. jejuni, C. coli, and C. lari, which are responsible for most of the Campylobacter infections associated with human illness. Real-time PCR detection of Campylobacter can be achieved in a highly standardised format in less than 10 hours. The format includes an enrichment culture step and realtime PCR. As the handling of PCR products is not required, this reduces the risk of amplicon carry-over leading to false positive results. In the future, the detection time could be reduced further by developing a rapid DNA extraction method directly from food samples. The real-time PCR method described in this chapter could be also employed to detect Campylobacter species present in human samples. The quantitative ability of this method has been examined successfully and as little as 125 cells or 192 fg/μl can be detected for C. jejuni. This method is very sensitive and specific, and 80

dramatically speeds up detection, and thereby improves the management of Campylobacter species outbreaks compared to conventional culture methods.

81

4. Chapter 4 Rapid Distinction Between C. jejuni and C. coli using SYBR Green I 4.1. Chapter Overview The hipO gene codes for the enzyme hippuricase which catalyses the hydrolysis of Nbenzoyleglycine (hippuric acid) to glycine and benzoic acid [97]. Glycine is detected by a ninhydrin-base reagent system [246], whereas benzoic acid is detected by precipitation with a ferric chloride reagent [247] or, more reliably, by gas-liquid chromatography [96, 248]. The hipO is present in C. jejuni but not in any other of the Campylobacter species and is therefore the most reliable biochemical assay that can discriminate C. jejuni from all other Campylobacter species including the closely related C. coli [249]. However, the occurrence of C. jejuni hippuricase-negative variants, the inability of the assay to detect low level producers of hippuricase and the inoculum size dependent expression of hipO are problematic in separation of these species [97, 250, 251]. Fortunately, the hipO gene is highly conserved and does not show any significant polymorphism [252], and is therefore an attractive target for the development of a real-time PCR assay for the confirmation of C. jejuni. This chapter describes a rapid method for the detection and identification C. jejuni by targeting the hipO gene using real-time PCR (Figure 2.1). The real-time PCR assay was compared using two different instruments, namely the LightCyclerTM and the iCycler iQTM. 4.2. Materials and Methods C. jejuni (ATCC 940565), C. coli (NCTC 11366), C. upsaliensis (QHSS 99 M 126), and C. hyointestinalis (QHSS 99M2318) were cultured on BAPs as described in Chapter 2 (Section 2.1). The DNA was purified from the isolates using the CTAB method (Section 2.2.1) and used as templates for the development of the real-time 82

SYBR Green I PCR assays. Replicate samples of the assay were continuously monitored in a LightCyclerTM and iCycler iQTM and the performance of the two instruments compared. Subsequently, DNA from 176 serotyped Campylobacter isolates, and 30 enrichment cultures initiated from chicken samples, were prepared by the rapid boil method and used in the assay. 4.3. Results 4.3.1. Detection of C. jejuni by Conventional Culture Methods Australian standard methods of colony appearance and biochemical testing (Section 3.2.1) were used to detect and identify C. jejuni from enrichment cultures initiated from chicken samples as described in section 2.1.1. Of the 30 samples tested using conventional culture, 17 (57%) samples were positive for C. jejuni.

4.3.2. Design of PCR Hippuricase Gene Primer The hipO genes sequences of C. jejuni were extracted from GenBank and aligned using BioEdit software (Figure 4.1). The C. jejuni reverse primer, Hip-2474R (5’AGCTAGCTTCGCATAATAACTTG-3’) described by Linton, (1997) and a forward primer, Hip-2214F (5’-GTTGTTGCACCAGTGACTATGA-3’) were identified and expected to produce a 260 bp amplicon. These two primers were specific to the hipO gene of C. jejuni as indicated by BLAST analysis against the GenBank database and had a theoretical calculated Tm value of 62 °C and 60 °C respectively.

83

C. jejuni NCTC 11168 ---------------GTTGTTGCACCAGTGACTATGA-------TTTTGCTAGTGAGGTTGCAAAAGAATTAT-----------CAAGTTATTATGCGAAGCTAGC C. jejuni strain KOR-H812 ---------GTTGTTGCACCAGTGACTATGA-------TTTTGCTAGTGAGGTTGCAAAAGAATTAT-----------CAAGTTATTATGCGAAGCTAGC

Schematic representation

primer Hip-2214F (2214-2236)

primer Hip-24744R (2452-2474)

5’GTTGTTGCACCAGTGACTATGA 3`

5’CAAGTTATTATGCGAAGCTAGC…3`

Figure 4.1 Alignment of partial hipO gene sequences of C. jejuni (GenBank Accession numbers Z36940and AL139076) showing the forward primer, Hip-2214F (5’-GTTGTTGCACCAGTGACTATGA-3’) and reverse primer, Hip-2474R (5’-AGCTAGCTTCGCATAATAACTTG3’) used in the development of real-time PCR.

84

4.3.3. Development of a Real Time SYBR Green I PCR Assay using a LightCyclerTM The SYBR Green I real-time PCR assay was initially developed using hipO gene specific primers, to check for the presence of the hipO gene in the DNA template (CTAB-purified and rapid boiling methods) of C. jejuni (ATCC 940565), C. coli (NCTC 11366), C. upsaliensis (QHSS 99M126) and C. hyointestinalis (QHSS 99M2318). An early increase in fluorescence (cycle 15) was consistently observed with C. jejuni (ATCC 940565) DNA templates, but a late increase in fluorescence (cycle 22) was observed from DNA templates from C. coli (NCTC 11366), C. upsaliensis (QHSS 99 M 126), and C. hyointestinalis (QHSS 99M2318) as depicted in Figure 4.2. This was consistent with expectations as SYBR Green I is known to bind to all double stranded DNA products whether these products are specific, non-specific (e.g. primer dimers). Calculation of the Tm from the melting curves generated from the dissociation of SYBR Green I from the PCR products suggested that the C. jejuni (ATCC 940565) amplicons, which had a Tm of 85оC were specific, but those from C. coli (NCTC 11366), C. upsaliensis (QHSS 99M126) and C. lari (ATCC 35223) which had a lower Tm of 56оC, were non-specific, perhaps primer dimers (Figure 4.3). Agarose gel electrophoresis confirmed that 260 bp amplicons were produced from C. jejuni (ATCCC 940565) DNA templates but not from the other 3 strains in line with the expected theoretical size (Figure 4.4).

85

Error!

Figure 4.2 Real-time PCR SYBR Green I assay with CTAB-purified DNA for the detection of the hipO gene. C. jejuni (ATCC 940565) (-Q-), C. coli (NCTC 11366) (V-), C. upsaliensis (QHSS 99M126) (-□-), C. hyointestinalis (QHSS 99M2318) (-O-) and negative control lacking template (-♦-) for the detection of hipO gene.

Figure 4.3 The melting peaks generated from the dissociation of the fluoroprobes from C. jejuni (ATCC 940565) amplicon target sites at the end of a run of real-time PCR produces a Tm of 85oC and a Tm of 56 °C for C. coli, C. hyointestinalis, C. upsaliensis and negative control lacking template.

86

Figure 4.4 Agarose gel electrophoresis of the PCR products from real-time SYBR Green I assay showing the expected 260 bp specific product for C. jejuni (Lane 1). No products were generated from C. coli (Lane 2), C. hyointestinalis (Lane 3), C. upsaliensis (Lane 4) and a negative control lacking DNA template (Lane 5). A DNA 100 bp ladder was used for size comparison.

4.3.4. Development of a Real Time PCR SYBR Green I Assay Using an iCycler IQTM The SYBR Green I real-time PCR reactions used in the LightCyclerTM above were also used in an iCycler iQTM. As with the LightCyclerTM results, all C. jejuni that had been tested in this experiment gave a positive result, while C. coli, C. lari and other Campylobacter species samples gave negative results. A gradual increase in fluoresence for C. jejuni started at cycle 13 due to the accumulation of the PCR specific products. C. coli and the other Campylobacter species also showed an increase in fluorescence but after cycle 25. This is due to the accumulation of the non-specific products (e.g. primer dimer) (Figure 4.5). The initial increase in fluoresence is dependent on the initial DNA concentration, so that the lower the DNA concentration the delayed the rise in fluoresence, and conversely, the higher the DNA concentration, the faster the rise. When the melting program was performed on all the C. jejuni positve samples, a melting curve with a Tm at about 85-86оC was generated, which is the Tm for the specific PCR product (Figure 4.6). Another melting temperature was generated at 58оC for negative samples, which is related to the non-specific products that were amplified. 87

The presence of large non-specific products is due to the high initial concentration of the template DNA. The agarose gel electrophoresis of the PCR samples verified the expected 260 bp PCR products produced by the two primers used in this experiment and this PCR product was found only on the positively identified samples (Figure 4.7). This protocol was shown to be rapid and robust, and useful for rapid detection and identification of C. jejuni in addition to distinguishing it from other Campylobacter species.

Figure 4.5 Real-time PCR SYBR Green I assay for the detection of C. jejuni hipO TM using an iCycler iQ . CTAB-purified DNA from C. jejuni (ATCC 940565), C. coli (NCTC 11366), C upsaliensis (QHSS 99M126) and C. hyointestinalis (QHSS 99M2318) were used as templates.

88

Figure 4.6. The Tm generated from the dissociation of SYBR Green I from the PCR products during the gradual increase in temperature from 50°C to 94°C. A specific 260 bp product with a Tm of 85 C and a Tm of 56°C for C. jejuni (ATCC 940565), C. coli (NCTC 11366), C. upsaliensis (QHSS 99M126) and C. hyointestinalis (QHSS 99M2318).

Figure 4.7 Agarose gel electrophoresis of the PCR products from real-time SYBR Green I assay using iCycler iQTM showing the expected 260 bp specific product for C. jejuni (Lane 1) and no products generated from C. coli (Lane 2), C. hyointestinalis (Lane 3) C. upsaliensis (Lane 4) and No Template (Lane 5). A 100 bp DNA ladder is shown in Lane 6 (Promega, Australia).

4.3.5. Identification of C. jejuni Isolates in the QHSS Culture Collection and Enrichment Cultures using the Real Time SYBR Green I Assay The above validated assay was then tested with the DNA templates from 176

Campylobacter isolates in our collection, prepared by the rapid boiling method. All the tested isolates serotyped as C. jejuni (88 isolates) gave positive results with the assay

89

but none of the other non-C. jejuni isolates (88 isolates) were found to be positive (Appendix 1). Furthermore, C. jejuni has been detected in 18 enrichment cultures initiated from chicken samples (Appendix 7).

4.3.6. Comparison between Conventional Culture assay and SYBR green I assay In this chapter a real-time PCR assay employing SYBR green I for the distinguishing of

C. jejuni from other Campylobacter species showed similar sensitivity, specificity and accuracy to the triplex real-time PCR assay (chapter 7), and the LDR assay (chapter 9). Conventional culture methods provide 94.4% accuracy compared to real-time PCR assay used in this chapter, as shown in section 2.10. Furthermore, the false negative rate was 5.5%, and the false positive was 0%. Therefore, sensitivity and specificity, were 94.4% and 100% respectively. The high number of false negative results obtained by conventional culture could be explained by the presence of an atypical C. jejuni that gave false negative results with the hippurate hydrolysis test, or the presence of VNC Campylobacter, or because C.

jejuni may be present in low numbers together with high numbers of competitor organisms.

4.4. General Discussion C. jejuni is one of 16 species in the genus Campylobacter. C. jejuni is responsible for 80- 90% of Campylobacter infections in humans, while C. coli accounts for 5-10% [11, 37, 253]. Conventional culture methods have been used to distinguish between the two species based on hippurate hydrolysis activity, which is a critical biochemical reaction that distinguishes C. jejuni from other Campylobacter species. However, the presence of atypical C. jejuni increases the chances of false negative results. It has been reported that the detection of a gene is more accurate than the detection of a phenotypic

90

characteristic of that gene. For this reason the hipO gene has been used in this study for the rapid detection and identification of C. jejuni from the closely related C. coli and C.

lari, as well as from all other Campylobacter species. Molecular detection methods (real-time PCR, T-RFLP, LDR, etc.) have several advantages over culture methods including: (i) shorter assay times (ii) the capability of identifying different forms of

Campylobacter, and (iii) the ability to target different genes, making them favourable as a replacement technique over culture methods. In this chapter, SYBR green I real-time PCR assay targeting the hipO gene was chosen for rapid detection and identification of

C. jejuni. Cycle numbers of the real-time PCR assay used in this chapter were reduced from 40 to 25 cycles to overcome the increase in the fluorescence signal after 25 cycles due to the accumulation of non-specific products (e.g. primer dimer), using both LightCyclerTM and iCycler iQTM. Results from the LightCyclerTM are comparable to those obtained using an iCycler iQTM. In both instruments, an increase in fluorescence began at cycle 15 for C. jejuni (ATCC 940565), whereas an increase in fluorescence was observed after cycle 23 for C. coli (NCTC 11366), C. upsaliensis (QHSS 99 M 126), and C. hyointestinalis (QHSS 99M2318). Both machines also produced similar results for the Tm in which specific products have melting peaks at 85-86οC while the non-specific products have melting temperature at 56οC. One advantage of the LightCyclerTM over the iCycler iQTM is the time taken to perform reactions. While the iCycler iQTM required two and a half hours to complete the run (the melt and the real-time monitoring), the LightCyclerTM only required 62 minutes to finish the run. The second advantage is that the LightCyclerTM required a smaller amount of master mix (PCR mix) of only 10 μl while the iCycler iQTM required at least double the quantity of the LightCyclerTM per reaction. An advantage of the iCycler iQTM over the LightCyclerTM is its ability to read more than 91

one fluorophore signal simultaneously, and therefore it can be used for real-time multiplex PCR. This is a significant advantage but has not been reported extensively. Another advantage of the iCycler iQTM is its ability to run 96 samples at the same time as opposed to the 24 or 32 in the LightCyclerTM. Hippuricase activity encoded by the hipO gene is often used as the only phenotypic test capable of differentiating C. jejuni from all other Campylobacter species including the closely related C. coli and C. lari species [249, 254]. Other researchers have used molecular techniques targeting the hipO gene for the purpose of differentiation. For example, Wainø et al. (2003) [255] used both phenotypic tests and species-specific PCR assays for the identification of Campylobacter species isolated from Danish broilers (chickens). Steinhauserova et al. (2001) [256] used phenotypic and molecular methods to identify Campylobacter species. Linton et al. (1997) [91] used a PCR assay based on

hipO to differentiate C. jejuni from other Campylobacter species, whereas Gehua et al. (2002) [257] developed a colony multiplex PCR assay for identification and differentiation of C. jejuni, C. coli, C. lari, C. upsaliensis, and C. fetus subsp. Fetus. However, several disadvantages that have been associated with conventional PCR assays include its technically demanding nature, lack of quantitation, being prone to carry-over contamination due to the post PCR procedures, and time-consuming aspects. The combination of the two real–time PCR assays described in this chapter and chapter 3 for the detection and differentiation of C. jejuni, from the closely related species C.

coli and C. lari, offers an attractive and alternative assay to conventional culture and conventional PCR methods for the clinical, food and environmental microbiology laboratories as described by the paper merged from these two chapters [125].

92

4.5. Conclusion SYBR Green I real-time PCR assay targeting hipO gene has been described in this chapter for the rapid detection and identification of C. jejuni from all other

Campylobacter species. The method described is easy to use, and rapid, requiring less than 10 h to detect C. jejuni, and this includes the setting up of enrichment cultures. Also, it is highly sensitive and specific compared to culture methods, and cost effective compared to the other real-time PCR assays employing adjacent hybridisation probes or TaqMan probes. The number of cycles has been reduced to 25 to reduce the effects of non-specific products that would interfere with real-time PCR results. Distinguishing specific C. jejuni products and non-specific products can also be achieved based on Tm, thereby removing any false results (e.g. from the primer dimer).

93

5. Chapter 5 Rapid Detection and Identification of Arcobacter Species 5.1. Chapter Overview Four Arcobacter species, A. butzleri, A. skirrowii, A. cryaerophilius and A. nitrofigilis, are closely related to each other. A. butzleri is the most commonly associated human gastrointestinal pathogen, with A. cryaerophilius and A. skirrowii have also been implicated as a human pathogen, but to lesser extent. A. butzleri, A. skirrowii, and A.

cryaerophilius are implicated in animal infections. The development of a real-time PCR method for differentiating the human pathogenic Arcobacter species from the nonhuman pathogens and from the related, morphologically and biochemically indistinguishable Campylobacter species is described in this chapter. Three strains of Arcobacter (A. butzleri (NCTC 12481), A. skirrowii (NCTC 12713) and

A. nitrofigilis (NCTC 12251) identified serologically were selected, cultured (Section 2.1) and their DNA purified using a CTAB method (Section 2.2.1). Their identities as members of the genus Arcobacter were confirmed by partial sequencing of the 16S rRNA genes (Figure 5.1). Amplification primers and potentially useful targets for 6FAM and Cy5 labelled adjacent hybridisation probes were identified (Figure 5.2) in sequences that had been extracted, edited and aligned as described in Chapter 2 (Sections 2.4.1 and 2.4.4). Additionally, the probes were designed according to the general guidelines described in Chapter 2 (Section 2.5.1). The CTAB-purified DNA (Section 2.2.1) from the three strains A. butzleri (NCTC 12481), A. skirrowii (NCTC 12713) and A. nitrofigilis (NCTC 12251) was used to develop a real-time PCR assay for the identification and differentiation of A. butzleri from A. skirrowii and A. nitrofigilis. A total of 22 Arcobacter isolates (Appendix 3) including the three types of culture strains used previously, and 30 enrichment cultures initiated from chicken samples (Appendix 6), were cultured, the DNA extracted using 94

the rapid boiling method and their identities determined using the newly devised realtime PCR assay. Furhtermore, E. coli, Staphylococcus aureus, Clostridium

perfringens, Micrococcus luteus, and Pseudomonas aeruginosa, were used as negative control. The sensitivity of the method was tested by using purified DNA and/or DNA extracted by the rapid boiling method for the A. butzleri (NCTC 12481).

5.2. Results 5.2.1. Identification of Strains (NCTC 12481), (NCTC 12713) and (NCTC 12251) as A. butzleri, A. skirrowii and A. nitrofigilis by Partial Sequencing of 16S rRNA The identities of three strains [A. butzleri (NCTC 12481), A. skirrowii (NCTC 12713) and A. nitrofigilis (NCTC 12251)] that had been isolated and typed serologically by QHSS were confirmed by partial sequencing of the 16S rRNA gene. As expected, DNA from all three strains produced an amplicon of approximately 1.5 kb with primers Fd1 and Rd1. Five primers (F2, F3, F4, F5 and Rd1) were then used to sequence approximately 759 nucleotides (corresponding to E. coli positions 783 to 1542, Figure 5.1). Sequence comparison revealed that the similarity between the species was less than 96%. This meant that all are different species, with the exception of A. skirrowii and A. cryaerophilius that showed 97% similarity. To distinguish the later two species, DNA-DNA hybridisation, serological and phenotypic characterisation will be necessary. A region corresponding to positions 783-1100 (E. coli numbering) was chosen as it contains heterogenicity for differentiating Arcobacter species. The conting assembly, alignment and BLAST analysis indicated that strains (NCTC 12481), (NCTC 12713) and (NCTC 12251) were strains of A. butzleri, A. skirrowii and A. nitrofigilis. The initial identification of the strains concurs with the 16S rRNA gene molecular analysis. The use of these sequences and other database sequences allowed an accurate design of species-specific fluoroprobes for real-time PCR assays. 95

Figure 5.1 The partial sequences of 16S rDNA (positions 783 to 1120, E. coli numbering according to Brosius et al. 1978 [230]) of the three Arcobacter species from QHSS and NCTC (A. skirrowii (NCTC 12713, lamb source) and A. nitrofigilis (NCTC 12251, Alterniflora roots). A. butzleri (QHSS 99M3958 human source) compared with database sequences of A. skirrowii (CCGU 10374, lamb source), A. nitrofigilis (CCUG 15893, Alterniflora roots) and A. butzleri (CCUG 10373 human source).

96

5.2.2. Development

and

Design

of

PCR

Primers

and

Adjacent

Hybridisation Probes Two probes, designated probe Butz (5’GTGTCTGCTTGCAGAAACTTGCATA FAM3’E. coli position 1019-1044) specific for binding to the 16S rRNA of A. butzleri, and probe Skir-Cry (5’GTGTCTGCTTGCAGAAACTTATATA-FAM-3’, E. coli position 1019-1044), with the ability to hybridise to A. skirrowii, A. cryaerophilius, were designed after analysis of the aligned sequences. The region immediately downstream from the target sites of probe Butz and probe Skir-Cry (E. coli position 1046 to 1075) was deemed to be suitable for binding of the universal Cy5 as labelled probe (5’Cy5AGGTGCTGCACGGCTGTCGTCAGCTCGTGT-P3’) and has been described in Chapter 3 (Section 3.2.3). 16S rRNA bacterial domain universal primers F2 and R5 (Table 2.1), expected to produce a 317 bp PCR product, were chosen for amplifying the region straddling the real-time PCR assay targets. The design of the amplification primer pair and the adjacent probes is schematically represented in Figure 5.2. GenBank database searches using BLAST and FASTA indicated that the combined use of the forward and reverse primers and adjacent hybridisation probes would specifically amplify and detect the 16S rRNA of A. butzleri A. skirrowii, A. cryaerophilius and A.

nitrofigilis producing an amplicon of 317 bp. FASTA analysis confirmed our previous results that probe Cy5+1046 was a universal probe that would bind to all members of domain Bacteria [258] due to the inosine introduced degeneracy at positions 1051 and 1059 (Figure 5.2). The primer was synthesised by the Centre for Molecular and Cellular Biology, Southern Cross University, Lismore, NSW, Australia, and the probes were synthesised by Proligo, Australia.

97

Escherichia coli------------CAGGATTAGATACCCTGGTAG--------GTG—--CCTTCGGG—AACCGTGAGACAGGTGCTGCATGGCTGTCGTCAGCTCGTGT-----GGGGTTGCGCTCGTTG----C. coli CCUG11283----------------------------------------GTGCTAGCTTGCTAGAACTTAGAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT ------------------------C. jejuni CCUG24567--------------------------------------GTGCTAGCTTGCTAGAACTTAGAGACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT ------------------------A. cryaerophilus CCUG17801----------- -------------------GTGTCTGCTTGCAGAAACTTATATACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT-------------------------A. butzleri CCUG10373------------------------------------GTGTCTGCTTGCAGAAACTTGCATACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT-------------------------A. nitrofigilis CCUG15893--------------------------------GTGTCTGCTTGCAGAAACTTACATACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT-------------------------A. skirrowii CCUG10374-----------------------------------GTGTCTGCTTGCAGAAACTTATATACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT-------------------------A. cryaerophilus CCUG17801-------------------------------GTGTCTGCTTGCAGAAACTTATATACAGGTGCTGCACGGCTGTCGTCAGCTCGTGT-------------------------Skir-Cry Probe-------------------------------------------GTGTCTGCTTGCAGAAACTTATATA Butz. Probe----------------------------------------------GTGTCTGCTTGCAGAAACTTGCATA Cy5 +1046 Probe----------------------------------------------------------------------AGGTGCTGCACGGCTGTCGTCAGCTCGTGT

Schematic representation

Primer F2 785-805 5’CAGGATTAGATACCCTGGTAG 3`

6-FAM probe 5’

1019-1044

FAM

c

3` 5’

probe Cy5+1046 Cy5

1046-1075

Primer R5 1100-1115 3`P

5’ GGGGTTGCGCTCGTTG 3`

Figure 5.2 The partial sequences of 16S rRNA (positions 783 to 1120, E. coli numbering [259]) from representative isolates of Arcobacter species. Boxes represent the universal bacterial forward (F2) and reverse (R5) primers, the universal fluoroprobe Cy5+1046 specific for members of domain Bacteria and fluoroprobe 6-FAM specific for detecting A. butzleri, A. skirrowii, A. nitrofigilis and A. cryaerophilius.

98

5.2.3.

Detection and Identification of Arcobacter Species by Conventional

Culture Methods Australian standard methods of colony appearance and biochemical tests (Appendix 4) were used to detect and identify the presence of Arcobacter species in Arcobacter enrichment cultures initiated from chicken samples. Arcobacter species are not able to be distinguished to species level by biochemical tests so for this reason they were identified as Arcobacter species only. Thirty samples were tested using conventional culture methods described in section 2.1.1. Sixteen (53%) samples were positive for Arcobacter, whereas 14 (47%) samples were negative for Arcobacter (Appendix 6).

5.2.4. Development of the Real-Time Assay for the Detection of A. butzleri, A. skirrowii and A. nitrofigilis using Adjacent Hybridisation Probes As shown in Figure 5.3, which represents the real-time monitoring of the PCR products during amplification, it appears that both probes are binding to target sites in template DNA: the probe Skir-Cry to A. skirrowii and A. cryaerophilius, while the probe Butz binds to A. butzleri. This result was expected and was in accordance with the theoretical primer and probe data. As a result, Arcobacter species showed an increase in fluorescence indicating that PCR amplification was occurring. The resulting 317 bp PCR product had incorporated the correct fluorescent probe and produced an emission of energy, which was detected by the LightCyclerTM. The probes did not generate an increase in fluorescence for the Campylobacter species and E. coli (Figure 5.3). Also,

Staphylococcus aureus , Clostridium perfringens, Micrococcus luteus, Pseudomonas aeruginosa, have been tested and did not generate an increase in fluorescence (result not shown). The melting curve and melting peaks (Figure 5.4) showed the difference in the Tm of the two specific probe target groups. This result was consistent with the

99

theoretical primer and probe data. These show that A. butzleri has a Tm of 67°C, A.

skirrowii of 63°C and A. nitrofigilis of 65°C. C. coli, C. jejuni and the negative control did not incorporate the specific probe, and therefore did not generate a Tm. This was further proof of the fluoroprobe specificity. The difference in the Tm of the probetemplate hybrid is an important tool in rapid identification of Arcobacter species. Additionally, agarose gel electrophoresis (Figure 5.5) showed that all samples generated PCR products, indicating that the PCR reactions were working. The result confirms the specificity of the adjacent hybridisation probes assay used in the experiment. Both probe Butz and probe Skir-Cry have the ability to detect A. nitrofigilis, as there is only one nucleotide difference between the sequence on the target region of the two probes (probe Butz and probe Skir-Cry) and the target region of A. nitrofigilis. Therefore, it is difficult to differentiate this species from the other Arcobacter species based on realtime monitoring. In this project, it was found that the Tm was an important tool in the differentiation of A. nitrofigilis from the other Arcobacter species (Figure 5.4).

5.2.5. Identification of Arcobacter Species in the Culture Collection and Enrichment Cultures using the Real Time Adjacent Hybridisation Probes Assay This method successfully identified the 22 Arcobacter pure cultures from the QHSS collection including type strains, as A. butzleri (20 isolates), A. skirrowii and/or A.

cryaerophilius (1 isolate) and A. nitrofigilis (1 isolate) (Appendix 3). Furthermore, 23 out of 30 enrichment cultures initiated from chicken samples were tested and identified as Arcobacter species, A. butzleri (18) and A. skirrowii and/or A. cryaerophilius (9) (Appendix 8). In addition, more than one Arcobacter species was detected in the same enrichment culture.

100

Figure 5.3 The increase in fluorescence during specific hybridisation of probes; probe Butz, probe Skir-Cry and Universal Cy5 to the target site in the 16S rRNA of A. butzleri (-■-), A. skirrowii (-▲-), A. nitrofigilis (-+-), C. coli (-●-), C. jejuni (-♦-) TM and no template (-∆-). Suring PCR as measured in the LightCycler . Purified DNA was prepared using CTAB method. Amplicons produced were run on an agarose gel (Figure 5.5) and the Tm determined (Figure 5.4).

Figure 5.4 The melting peaks generated from the dissociation of the fluoroprobes from A. butzleri, A. skirrowii and A. nitrofigilis amplicon target sites at the end of real-time PCR produces Tm's of 63oC for A. skirrowii, 65oC for A. nitrofigilis and of 67°C for A. butzleri but as expected the probe did not hybridise to the amplicons of C. coli, C. jejuni, E. coli and no template (Negative control), therefore no Tm was produced.

101

Figure 5.5 Agarose gel electrophoresis of PCR products from no DNA template (Lane 2), A. butzleri (Lane 3), A. skirrowii (Lane 4), A. nitrofigilis (Lane 5), C. coli (Lane 6) and C. jejuni (Lane 7). A 100 bp DNA Ladder is in Lane 1 (Promega, Australia).

5.2.6. Comparison between Conventional Culture Assay and Adjacent Hybridisation Probe Assay for the Detection and Identification of Arcobacter Species Real-time PCR assays described in this chapter, like all other molecular methods described in this thesis, show 100% accuracy, specificity and sensitivity. In comparison, the conventional culture method used in this study 16 out of 30 enrichment cultures detected the presence of Arcobacter species by conventional culture based methods, and 22 by real-time PCR, representing only 73% accuracy, 100% specificity and 73% sensitivity. In addition, the real-time PCR detects the presence of more than one

Arcobacter species in the same enrichment culture, which is usually not possible with conventional culture methods. Different studies have shown that the presence of A. butzleri in poultry carcasses can range from 0% to 97% [259]. This high percentage of Arcobacter detected in poultry samples concurs with the published studies [259]. The high contamination of chicken samples could be explained by the lack of hygiene in the slaughterhouses, as many 102

reports suggest that Arcobacter species do not colonize the intestinal tract of the chicken [8, 76, 260-262]. Alternatively this could be a post slaughterhouse contamination, which occurs in supermarkets or during packaging.

5.2.7. Real Time PCR Detection of A. Butzleri from Different DNA Concentrations Quantitative PCR assays have been shown to be useful clinically, and are becoming increasingly important tools in diagnostics laboratories for monitoring pathogens. Quantitative PCR was used for determining the sensitivity of the assay that is described in the previous section. DNA templates were prepared by CTAB (Section 2.1.1) in order to determine the sensitivity of the detection method. Both the agarose gel electrophoresis (Figure 5.6) and continuous monitoring (Figure 5.7) show the effect of different template DNA concentrations: the highest DNA concentration (113 ng/μl) shows an increase in fluorescence at cycle 12 while the lowest concentration (113 fg/μl) shows an increase after the cycle 25. A standard curve for A. butzleri was generated by using data derived from the serial dilution of A. butzleri 16S rRNA gene (Figure 5.7). The linear correlation coefficient of the standard curve was R2 = 0.9825, demonstrating the accuracy of the real-time PCR-based quantification (Figure 5.8). This standard curve could be used to approximate DNA concentrations in unknown samples.

103

Figure 5.6 Agarose gel electrophoresis of different DNA concentrations 113 ng/μl ((Lane 2), 11.3 ng/μl (Lane 3), 1.13 ng/μl (Lane 4), 113 pg/μl (Lane 5), 11.3 pg/μl (Lane 6), 1.13 pg/μl (Lane 7), 113fg/μl (Lane 8) and 100 bp DNA ladder (Promega, Australia) (Lane 1).

Normalized flouresence

120 100 80 60 40 20

43

40

37

34

31

28

25

22

19

16

13

10

7

4

1

0 Cycle number

Figure 5.7 Real-time detection of A. butzleri CTAB-purified DNA at different concentrations. 113 ng/μl (-χ-), 11.3 ng/μl (-■-), 1.13 ng/μl (-▲-), 113 pg/μl (-●-), 11.3 pg/μl (-+-), 1.13 pg/μl (-X-) and 113 fg/μl (-∆-).

104

35

Fluorescence

30 25

y = -1.1323Ln(x) + 19.218 R2 = 0.9825

0.000142 0.00142 0.0142 0.142

20

1.42 14.2

15

142

10 5 0 1.0E-04

1.0E-02

1.0E+00

1.0E+02

1.0E+04

Log DNA concentration

Figure 5.8 A Standard curve for A. butzleri genomic DNA for 16S rRNA quantitative real-time PCR using adjacent hybridisation probes. The DNA was serially diluted 10-fold. Ct values were plotted against the log10 DNA concentration and the equation and linear correlation coefficient (R2) determined in Microsoft Excel. The straight line, which was calculated by linear regression [Ct =-1.1323 log (DNA concentration) + 20.648], shows an R2 of 0.9825. One microliter of standard DNA was used in a 10 μl PCR reaction. (The experiments were repeated three times with identical results.)

5.2.8. Real Time PCR Detection of A. butzleri from Single Colony Serial Dilutions In order to determine the sensitivity of the detection method, a dilution series was prepared from a single colony of A. butzleri and the DNA extracted by the rapid boiling method (Section 2.2.2). Figure 5.9 shows that as few as 110 cells of A. butzleri could be detected, confirming the sensitivity of this method. Fluorescence increase started at cycle 27 for the 110 cells, whereas the initial increase started at cycle 20 for the 440.000 cells. These results are similar to those described in section 3.2.8. The linear correlation coefficient of the standard curve was R2 = 0.9994, demonstrating the accuracy of the real-time PCR-based quantification (Figure 5.10).

105

Figure 5.9 A representative real-time PCR analysis of a dilution series of A. butzleri culture ranging from 440000 to 110 cells. One microliter was used for each PCR reaction. The increase in fluorescence signal related to cycle number is shown. 440000 cells (-■-), 44000 cells (-●-), 4400 cells (-▲-), 440 cells (-ж-) and 110 cells (♦-).

Figure 5.10 A Standard curve for A. butzleri genomic DNA for 16S rRNA quantitative realtime PCR using adjacent hybridisation probes. One colony was serially diluted 10-fold. Ct values were plotted against log10 cell numbers. The straight line, which was calculated by linear regression [Ct =-1.0893 log10 (concentration)+ 36.339], shows an R2 of 0.9944. One microliter of DNA was used in a 10 μl PCR reaction. (The experiments were repeated three times with identical results.)

106

5.2.9. Real-Time PCR Detection of A. butzleri during Different Growth Phases Samples were taken at various time points during the growth of different Arcobacter species in order to determine the earliest time in the culture incubation at which these organisms could be effectively detected. Figure 5.11 shows that Arcobacter species were detected after 24 hours incubation. No increase in fluorescence was observed at 8 hours or less.

Figure 5.11 A. butzleri was grown for periods of up to 24 h, DNA was extracted by the rapid boiling method and used in real-time PCR adjacent hybridisation probes assays. (- -) 0, (-▲-) 2, (-x-) 4, (-♦-) 6 hours, (-Δ-) 8 hours and (-■-) 24 hours.

5.3. General discussion Arcobacter species have been recognised as emerging food-borne pathogens associated with human gastroenteritis [73]. Arcobacter have been isolated from poultry and other livestock and their presence has also been reported in water [72, 263, 264]. Arcobacter species have the ability to survive in well water stored at 5°C for 16 days and this is considered to be a major route of infection [71]. A different route of infection comes from contaminated foods such as poultry [66, 73]. Controlling the spread of Arcobacter 107

could be improved by early detection of Arcobacter from contaminated habitats. Though common food storage conditions are not conducive to the growth of Arcobacter species, their ability to survive at low temperatures (5°C) and their ability to infect individuals at a low infective dose, (100 - 500 cells) requires the development of rapid identification methods [8, 109]. The methods should be flexible in order to detect their presence from all food and other environmental sources including water. Given that Arcobacter species are biochemically inert, an array of phenotypic tests are carried out to identify and differentiate them at a species level [85, 89, 264]. Arcobacter isolates are differentiated into two subgroups based on

-haemolytic and catalase

activities [63, 265]. Currently, few molecular methods have been reported that discriminate between Arcobacter species though there is a clear need for such methods. In this chapter, two different 6-FAM arcobacters probes (probe Butz and probe SkirCry) have been developed and used for rapid detection and identification of Arcobacter species A. butzleri and (A. skirrwoii and A. cryaerophilius) respectively, in conjunction with a universal Cy5 labelled probe. Both FAM probes are differentiated by 2 bp. This difference will allow for the differentiation of A. butzleri from A. skirrowii and A.

cryaerophilius by a difference in Tm, due to the probes dissociation, which is indicated by the melting peaks on the graph. On the other hand, A. nitrofigilis, which is normally isolated from plant roots and not associated with human or animal infection, could be detected by both probes but is simultaneously differentiated from the other species by the difference in its Tm, as shown in Figure 5.4. Agarose gel electrophoresis in all positive and negative bacterial samples revealed the presence of a 317-bp PCR product except in the reaction with no template as expected. The presence of PCR products in all samples, but the absence of fluorescence signal from E. coli, C. coli, and C. jejuni as well as Staphylococcus aureus, Clostridium

perfringens, Micrococcus luteus, Pseudomonas aeruginosa indicates the high 108

specificity of the hybridisation probes. Primer dimer and non-specific products were present in all detection reactions, but these products did not affect the specificity of the probe. The levels of increase in the fluorescence due to probe hybridisation corresponded to the DNA concentration (Figure 5.6). Accordingly, fluorescence could be detected as early as cycle 12 for high DNA concentrations (113 ng/μl) and in cycle 29 for low DNA concentrations (113 fg/μl). When the number of cells required for detection was determined, it was found that as few as 110 cells (around 330 16S rRNA gene copies) could be detected for Arcobacter, which is similar to the Campylobacter species result (Section 3.2.5). It has been found that Arcobacter species could not be detected after 8 hours incubation, which may be due to the slower growth of Arcobacter species compared to Campylobacter species. The increase in the fluorescence ratio in the DNA sample of A. butzleri and A. skirrowii probes due to the binding of the specific probe (probe Butz and probe Skir-Cry) to the target site, but not for the other Campylobacter or E. coli indicates that the probes were highly specific. To determine the Tm of the probes, a melt program was run, as described in Section 2.5.2. Under the melt program conditions, the dissociation of the fluorogenic hybridisation probes from the PCR product started at around 60°C, with complete dissociation at 70°C. Tms for the Arcobacter species were as follows: A. butzleri: 67°C;

A. skirrowii: 63°C; and A. nitrofigilis: 65°C. The real-time PCR assay used in this study relies on continuous monitoring of PCR products using fluorogenic probes. This assay is rapid, highly specific, robust, highthroughput and automated. Each run consisted of 45 cycles, which took an average of 62 minutes to complete. The ultimate goal of this project to develop rapid identification methods for Arcobacter species was successful. Further work will be required to apply these methods to clinical specimens. The rapidity of the method described here provides a valuable tool for the food industry and could assist in identifying the Arcobacter found 109

in infected food products. This will help to develop procedures to control their transfer from infected birds or animals to non-infected ones in the slaughterhouses. It would also greatly facilitate the monitoring of these bacteria throughout the food processing chain and may therefore contribute to enhanced public health protection. A number of papers describing Arcobacter detection by culture methods, conventional PCR, SDS-PAGE and 16S rRNA gene sequencing have been published recently. For example, Houf et al. (2001) [266] used enriched Arcobacter broth with the selective supplement for isolation and quantification of Arcobacter species from poultry products. Conventional culture methods are the most commonly used methods for the detection of Arcobacter species and other pathogens, in pathology, food, and environmental microbiology laboratories, as they are easy to perform. However, these methods have several disadvantages including a lack of sensitivity and specificity; a slow process that requires 2-5 days for analysis; their inability to detect VNC and their ability to detect and discriminate dominant species only. Molecular methods are very sensitive as identification is based on a unique signature sequence for each species or subspecies. There are numerous examples of the power of molecular methods for clinical diagnosis of Arcobacter species. For examples, Woo et al. (2001) [258] used 16S ribosomal RNA gene sequencing for identification of Arcobacter cryaerophilius isolated from a traffic accident victim with bacteremia. The main disadvantage of the sequencing method includes the high cost aspects, the technically demanding nature, and the intensive labour content. Atabay et al. (2003) [259] used SDS-PAGE to detect

Arcobacter species in chicken carcasses sold in retail markets in Turkey. Harmon and Wesley (1997) [103] developed a multiplex PCR assay for the identification of

Arcobacter and differentiation of A. butzleri from other arcobacters. Furthermore, Houf et al. (2000) [88] developed a multiplex PCR assay for the simultaneous detection and identification of A. butzleri, A. cryaerophilius and A. skirrowii, while Suarez et al.

110

(1997) [267] used PCR for the detection of Arcobacter species in gastric samples from swine.

5.4. Conclusion Conventional PCR methods have been proven to be valuable for the detection of

Arcobacter species, and are considered very sensitive compared to the culture based methods. The real-time PCR assay developed in this study saves time and labour, and generates results within 1 hcompared to more than 3 h required for conventional PCR. In addition, detection of conventional PCR products by gel electrophoresis requires the use of EtBr, a known carcinogen which requires strict handling practices. This report has also demonstrated that a simple boiling method, which we have used previously to develop real-time PCR assays for Campylobacter species (chapter 3), and Leptospira,

Leptonema and various thermophilic bacteria [116, 120-123, 232], can also be used for Arcobacter. According to my knowledge this is the first time that a real-time PCR adjacent hybridisation assay has been used for rapid detection and identification of

Arcobacter species.

111

6. Chapter 6 Rapid Identification of Arcobacter species and C. jejuni and C. coli using Tm of Adjacent Hybridisation Probes 6.1. Chapter Overview Previously, Arcobacter and Campylobacter species were classified in a single genus due to their similar biochemical and cultural characteristics [109, 197]. Molecular techniques are regarded as important tools in their differentiation and a real-time PCR technique is reported in this chapter. The methods used for this study have been described in detail (Chapter 2), and are therefore described only briefly in this chapter, with reference to relative sections in brackets. Briefly, a total of 198 isolates [176 Campylobacter isolates (Appendix 1) and 22

Arcobacter isolates (Appendix 3) including the type cultures A. butzleri (NCTC 12481), A. skirrowii (NCTC 12713), A. nitrofigilis (NCTC 12251), C. jejuni (ATCC 940565) and C. coli (NCTC 11366)] were selected and cultured (Section 2.1). Also another 60 enrichment cultures initiated from chicken samples were used. DNA was extracted using three methods (Section 2.2). The identities of the isolates as members of the Campylobacter and Arcobacter species were confirmed by partial sequencing of the 16S rRNA genes (Figures 3.1 and 5.1). Useful target regions were identified and amplification primers, 6-FAM and Cy5 labelled adjacent hybridisation probes were designed against these targets (Figures 3.2 and 5.2, as described in Chapter 2 (Sections 2.4.1 and 2.5.1). For real-time PCR, a modification in which all three probes were used in one tube is described in this chapter (Figure 2.1).

112

6.1.1. Real-Time PCR Assay in a LightCycler

TM

10-µl glass capillary tubes were used in this assay. The reaction tubes contained 1 μl of 10x PCR buffer 20 mM MgCl2 (part 1771, Idaho Technology, USA), 1μl of 2 mM dNTPs (0.5 mM dATP, 0.5 mM dGTP, 0.5 mM dCTP, and 0.5 mM dTTP), 1 μl of 5

μM F2 (Forward 16S rRNA primer) (Table 2.1), 1 μl of 5 μM R5 (Reverse 16S rRNA primer) (Table 2.1), 1 μl of 2μM probe Jejuni-coli FAM, 1 μl of 2 μM probe Skir-Cry, 1 μl of 2 μM probe Butz, 1 μl of 2 μM Cy5 probe, 0.08 μl of 5 U/μL of Taq DNA Polymerase (Promega Corp., USA), 1 μl of chromosomal DNA, and 0.92 μl of sterile ddH2O. PCR thermal cycling for the amplification of the 16S rRNA target region was described in Section 2.5.3.

6.2. Results 6.2.1. Multi-FAM One Tube Assay for Rapid Detection and Identification of C. jejuni, C. coli, A. butzleri and A. skirrowii Three different probes (probe Jejuni-coli, probe Butz and probe Skir-Cry) (Sections 3.2.3 and 5.2.2) were developed and used to rapidly distinguish between C. coli and C.

jejuni from Arcobacter species. It was shown in Chapter 3 and Chapter 5 that these probes when used singly were highly specific, allowing highly accurate detection of the target species. All three probes were used simultaneously in one tube for genera/species identification by determining the specific Tm of the probe-template hybrid directly after the PCR amplification (Section 2.5.3). Figure 6.1 shows the real-time monitoring of the PCR products during amplification. It appears that each one of the three probes bound perfectly to its specific target sites resulting in an increase in the fluorescence signal during PCR and indicating that amplification was occurring. The resulting 317 bp amplicons hybridised to the correct fluorescent probe and produced a fluorescence 113

energy that was emitted and detected in the LightCyclerTM. Fluorescence increase started between cycle 18 and 22 due to the variation in the DNA concentration. The Tm and melting peaks (Figure 6.2) showed the differences in the Tm after the probes had been dissociated from the target region of the amplified amplicons as a result of increased temperature at the end of PCR amplification. The Tm generated from the dissociation of the perfect target was found to be 65°C for C. coli, C. jejuni, C. lari and

C. hyointestinalis, 67°C for A. butzleri, and 63°C for A. skirrowii, the difference in Tm being due to the difference in the sequence composition of the probes. As expected, C.

upsaliensis, C. fetus, E. coli and negative control lacking template did not generate a Tm. In this assay, the Campylobacter group (C. hyoilei, C. helviticus, C. hyointestinalis, C.

insulaenigrae, C lanienae, C. coli and C. jejuni) will be detected using the probe Jejunicoli and probe Cy5+1046 due to the use of the universal 16S rRNA forward (F2) and reverse (R5) primers. This is an important assay in rapid detection and identification of

Campylobacter and Arcobacter, which is similar to other results obtained in this study, as shown in figures 5.4 and 6.2. Agarose gel electrophoresis shows that all samples generated 317 PCR amplicons as expected, indicating that the PCR reaction was working (Figure 6.3). Results obtained from this study confirm the specificity of the adjacent hybridisation probes assay used in this study, as shown in Chapter 3 and Chapter 5.

114

Figure 6.1 Real-time detection of Campylobacter species and Arcobacter species using probe Skir-Cry, probe Butz and probe Jejuni-coli, C. jejuni (QHSS 00M2260) (- -), C. coli (P287/96) (-▲-), A. butzleri (ATCC 12481) (-X-), A. skirrowii (ATCC 12713) (-∆-), C. upsaliensis (QHSS 99M126), (-■-) and a negative control lacking template (-♦-).

Figure 6.2 The melting peaks generated from the dissociation of the fluoroprobes from A. butzleri (NCTC 12481) (Tm 67 oC), A. skirrowii (NCTC 12713) (Tm 63 oC), C. jejuni (ATCC 940565) (Tm 65 oC) and C. coli (NCTC 11366) (Tm 65 oC), while E. coli, C. upsaliensis, and C. fetus and negative control lacking template DNA does not produce a Tm.

115

Figure 6.3 Agarose gel electrophoresis of PCR products from a negative control lacking DNA template (Lane 1), A. butzleri (ATCC 12481) (Lane 2), A. skirrowii (ATCC 12713) (Lane 3), C. jejuni (QHSS 00M2260), (Lane 4), C. coli (QHSS P287/96) (Lane 5), C. upsaliensis (QHSS 00M2260) (Lane 6) and 100 DNA ladder (Lane 7).

6.2.2. Identification of C. jejuni Isolates in the Culture Collection using the Real Time Multi-FAM Adjacent Hybridisation Probes Assay 170 out of 198 isolates from human, animal, plant and bird origin were identified as members of the Campylobacter group (C. coli, C. jejuni, C. lari, C. hyoilei, C.

helviticus, C. hyointestinalis, C. insulaenigrae, and C lanienae), A. butzleri (20 isolates), A. skirrowii and A. cryaerophilius group (1 isolate) and other species (7 isolates) (Appendix 1 and 3). Furthermore, 27 out of 60 enrichment cultures successfully detected the presence of the Campylobacter group (C. coli, C. jejuni, C.

lari, C. hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae, C lanienae) (Appendix 6). A. butzleri was detected in 18 samples and A. skirrowii and/or A.

cryaerophilius detected in 9 samples (Appendix 9). More than one Arcobacter species were also detected from the same enrichment culture.

116

6.2.3. Comparison between Conventional Culture assay and Multi-FAM Adjacent Hybridisation Probes Assay Accuracy, Sensitivity and specificity of these methods is similar to those described in chapter 3, and chapter 5.

6.3. General discussion Campylobacter and Arcobacter species are associated with many human and animal diseases, and are difficult to differentiate morphologically. Arcobacter species grow in the air, whereas Campylobacter are microaerophiles. Arcobacter species were misidentified as Campylobacter species until 1977 when Ellis et al. described the later as gram-negative, spirillum-like bacteria [56]. Distinguishing between the two genera is very important for epidemiological studies. Jacob et al., 1996 [268] reported that due to a lack of sensitivity of conventional culture methods, Arcobacter were commonly misidentified as atypical Campylobacter. The lack of standardisation of conventional culturing methods and the occurrence of atypical strains has fueled an interest in molecular detection and identification methods [91, 228, 269]. Real-time PCR assays targeting 16S rRNA genes have been developed for rapid detection and identification of food-borne pathogens including Campylobacter,

Salmonella, and E. coli as these methods promise high throughput [125, 270]. Real-time PCR assays for the detection of Campylobacter species and Arcobacter species have been described fully in chapters 3, 4, and 5. The assays described are based on the continuous monitoring of PCR products using fluorogenic probes. The assay developed in this chapter is rapid, highly specific and has the potential of automation. Primer dimer and non-specific products, which appear in the agarose gel, did not affect the specificity of the probes either in the continuous monitoring or during melting. The results

117

described in this chapter indicate that Arcobacter and Campylobacter species can be distinguished as a result of different Tms after real-time PCR. These reactions were performed separately for Campylobacter and Arcobacter specific probes, as described in Chapter 3 and Chapter 5. In this chapter, we have extended this to include all three probes in one tube for differentiation of Campholobacter (C. coli, C. jejuni, C. lari, C.

hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae, C lanienae group), and Arcobacter species (A. butzleri, and A. skirrowii and A. cryaerophilius group). Development of a rapid, accurate, high-throughput, sensitive and cost-effective assay for the detection and identification of these closely related members of the two genera is a very important challenge. For this reason, many molecular techniques have been developed in the last decade. Most researchers have reported on the use of conventional PCR for the rapid detection and identification of Campylobacter and Arcobacter species [109, 197]. Though these methods are sensitive and rapid, they carry several limitations. These include cross contamination due to post-PCR steps; non-quantitation; timeconsuming, which includes setup of product separation on agarose gels and their subsequent detection by visualization under UV light, and the use of EtBr which is a carcinogen. An alternative approach of RFLP analysis of the 16S rRNA gene for the rapid identification of Campylobacter, Arcobacter and Helicobacter isolates was used by Marshall et al. (1999) [152]. This method has limited use as a routine diagnostic method because it is time-consuming, laborious, and requires highly skilled scientists, since the data that are generated are very complicated to interpret. Real-time PCR assays have been reported on extensively for the rapid detection and identification of pathogens, in general and should be the method of choice for developing rapid detection methods for Arcobacter and Campylobacter species. A number of studies describing the detection and identification of bacterial species, viruses and mutation based on the Tm of the PCR products, either by using adjacent hybridisation probes or SYBR Green I

118

assays, have been reported. Poddar (2003) [271] used a real-time PCR beacon probe and probe-target melting assay for the detection and discrimination of B. pertussisi and B.

holmesii while Zhang et al. (2002) [272] used adjacent hybridisation probes and melting curve analysis for rapid detection of hepatitis B virus mutations. Furthermore, Steffensen et al. (2003) [273] used adjacent hybridisation probe assays and melting analysis for rapid genotyping of MBL2 gene mutations. Hernandez et al. (2003) [274] developed a Tm-based SYBR Green I real time multiplex PCR assay to detect genetically modified organisms. This is the first time that adjacent hybridisation probe real-time assays and Tm analysis of the PCR products have been used to detect and identify different species of the closely related genera Arcobacter and Campylobacter.

6.4. Conclusion The melting temperature of fluoroprobes (multi-FAM adjacent hybridisation probes) was used to detect and differentiate Campylobacter and Arcobacter species associated with human infections. Campylobacter and Arcobacter enrichment cultures initiated from chicken samples and BAP cultures were used to detect and distinguish between Campylobacter, and Arcobacter species based on the Tm difference of the adjacent hybridisation probes. The rapid boiling DNA extraction method is a simple method for template preparation, thus it was used in this study as it has the advantage of saving time compared to other DNA extraction methods. The multiFAM assay described in this chapter is sensitive, rapid and relatively cost-effective compared to the conventional culture method.

119

7. Chapter 7 DNA Multiplexing for Rapid Detection and Identification of C. coli and C. jejuni from the Other Campylobacter Species 7.1. Chapter preview This chapter describes the development of a rapid and sensitive real-time PCR multiplex method, targeting 16S rRNA genes, orfA genes and hipO genes, for the detection and identification of Campylobacter species C. coli, and C. jejuni respectively (Figure 2.).

C. coli and C. jejuni are the most common and routinely isolated pathogenic members of the genus Campylobacter.

7.2. Material and Methods 7.2.1. Methods The methods used for this study have been described in detail (Chapter 2). Specific methods relating to this chapter are described below. Briefly, a total of 206 isolates, 176 Campylobacter isolates (Appendix 2), including the type strains already detailed and the enrichment cultures initiated from chicken samples, were used to validate the sensitivity and specificity of this assay. Furthermore,

Arcobacter species, E. coli, Staphylococcus aureus, Clostridium perfringens, Micrococcus luteus, and Pseudomonas aeruginosa, were used as negative control. 7.2.2. Design of TaqMan Probes 16S rRNA gene sequences from Campylobacter and various phyla of domain Bacteria were extracted from the GenBank and Ribosomal Database Project (RDPII) [275]. C.

jejuni hipO and C. coli orfA gene sequences were extracted from GenBank [276]. BioEdit [231] was used to align the sequences and to reveal genus and species-specific

120

signature sequences against which primers and probes were designed using the criteria outlined (Section 2.4.1 and 2.5.2).

7.3. Results and Discussions

7.3.1. Detection

and

Identification

of

Campylobacter

Species

by

Conventional Culture Methods Conventional culture methods described in section 2.1.1 successfully detected

Campylobacter in 25 (83%) samples, whereas 5 (17%) samples were negative for Campylobacter. Identification to the species level showed that 17 (56.6%) isolates were identified as C. jejuni, 7 (20%) were identified as C. coli and 1 (3.3%) was identified as

C. lari (Appendix 5). 7.3.2. Development and Design of PCR Primers and TaqMan (5’ Nuclease Assay) Probes Primers and probe target regions were identified for three 3 different genes as described in 7.2.2. A Campylobacter genus specific 16S rRNA target region useful for developing a TaqMan probe and primer sets was identified after examination and analysis of the multiple sequence alignment of representatives of domain Bacteria. The sequences in the dataset were extracted from GenBank Database (version 127) and Ribosomal Database Project II (RDPII) and also included all the validated type species of

Campylobacter and the closely related Arcobacter [31]. The forward primer TaqCam936F (5’-CAAGCGGTGGAGCATGTGGTTT-3’3’, E. coli position 936) and the reverse primer TaqCam-1087R primer (5’-CAACATCTCACGACACGAGCTG-3’, E.

coli position 1087) were selected. This primer combination was deemed to be Campylobacter species 16S rDNA specific and would produce a 148 bp amplicon during PCR. A region specific for the Campylobacter species within the 148 bp

121

sequence, corresponding to position 1034-1062, was found to be suitable target region for designing a fluorogenic TaqMan probe. The probe, which was designated TaqCam1034

(5’-Cy5-CAGCCGTGCAGCACCTGTCTCTAAGTTCT-BHQ-2-3’),

was

labelled at the 5’ with Cy5 and at the 3’ end was attached to BHQ-2 (Figure 7.1). The hipO gene is unique and is only present in the genome of C. jejuni and hence was used as a target gene for designing primers and probes for its specific detection. A forward

primer,

designated

TaqHip-1754F

(5’-

TGGTGCTAAGGCAATGATAGAAGA-3’), and a reverse primer designated TaqHip1924R (5’-TGACCACCTCTTCCAATAACTTCA-3’), corresponding to positions 1754-1777 and 1901-1924 respectively of the C. jejuni hipO gene, were designed, and were expected to produce an amplicon of 170 bp during PCR. A probe, designated TaqJej-1864

(5’-Max-AACTATCCGAAGAAGCCATCATCGCACCTT-BHQ-1-3’),

labelled with the fluorophore Max at the 5’ and a BHQ at the 3’ end was designed to bind to the amplicons corresponding to positions 1864-1894 produced by this primer pair (Figure7.2 A). A gene (OrfA) for a conserved hypothetical protein identified in the genome sequence of C. coli had significant divergence from other homologs and hence was selected as a target gene for the specific detection of C. coli. A forward primer, designated TaqOrf4024F (5’-GCACTCATCCAATACTTACAAGA-3’) and a reverse primer designated, TaqOrf-4129R (5’-CATTATGGTGTATTCCGCCCA-3’), corresponding to positions 4024-4047 and 4109-4129 of the C. coli orfA gene, were designed. The primer pair was expected to produce an amplicon of 105 bp during PCR. A probe designated TaqColi4075 (5’-Fam-AAGTTCCATCTGACGCTGAAGCTACTCAAG-BHQ-1-3’) labelled with 6-FAM at the 5’, and BHQ-1 at the 3’ end, was designed to bind to the amplicons (positions 4075-4104) produced by the primer pair (Figure 7.2 B).

122

Escherichia coli------------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. sputorum

ATCC 33491-----------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG------

C. gracilis ATCC33236------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG------C. concisus ATCC33237-------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. curvus ATCC35224---------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. mucosalis CCUG6822-------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. rectus ATCC33238---------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. hyointestinalis ATCC35217------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. fetus ATCC19438----------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. upsaliensis CCUG14913C --------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. helveticus NCTC12470-----------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. showae CCUG11641---------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. lari NCTC11352-----------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. coli CCUG11283-----------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----C. jejuni CCUG24567---------------CAAGCGGTGGAGCATGTGGTTT------ CAGCCGTGCAGCACCTGTCTCTAAGTTCT--------------- CAACATCTCACGACACGAGCTG-----Cam-Multi-TaqMan------------ ----------------------------------CAGCCGTGCAGCACCTGTCTCTAAGTTCT

Schematic representation

CAAGCGGTGGAGCATGTGGTTT

TaqCam-936F primer

CAGCCGTGCAGCACCTGTCTCTAAGTTCT

5’ Cy5

TaqCam-1034

BHQ2 3`

CAACATCTCACGACACGAGCTG

TaqCam-1087R primer

Figure 7.1 The partial sequences of 16S rRNA (corresponding to position 936 to 1087, E. coli numbering [230]) from representative isolates of Campylobacter species. Boxes represent Campylobacter species forward primer (TaqCam-936F) and Campylobacter reverse primer (TaqCam-1087R), Campylobacter species-specific TaqMan probe (TaqCam-1034).

123

C. jejuni hippuricase gene--TGTGCTAAGGCAATGATAGAAGA----------CGATCTTAAAAAAGGTGCGATGATGGCTTCTTCGGATAGTTATAGCATTGAAGTTATTGGAAGAGGTGGT-----

Schematic representation (A) C. coli OrfA gene---

Schematic representation (B)

TGGTGCTAAGGCAATGATAGAAGA HipTaq-1754 (1754-1777)

AACTATCCGAAGAAGCCATCATCGCACCTT TGACCACCTCTTCCAATAACTTCA 5’-MAX TaqJej-1864 probe BHQ-2 3` HipTaq-1924 (1901-1924)

----GCACTCATCCAATACTTACAAGAAAACAATTTAAAAGTTCCATCTGACGCTGAAGCTACTCAAGGCAAAGGCAAGGGGGCGAATATGGGCGGAATACACCATAATG----GCACTCATCCAATACTTACAAGA OrfATaq-4024 (4024-4047)

AAGTTCCATCTGACGCTGAAGCTACTCAAG 5’Fam

Coli-Taq probe

3`BHQ1

CATTATGGTGTATTCCGCCCA OrfATaq-4109 (4109-4129)

Figure 7.2 The partial sequences of C. jejuni and C. coli orfA gene. (A) The partial sequences of C. jejuni hipO gene (corresponding to position 1754 to 1924, C. jejuni numbering). Boxes represent C. jejuni forward primer (HipTaq-1754) and C. jejuni reverse primer (HipTaq-1924), C. jejuni specific TaqMan probe (TaqJej-1864). (B) The partial sequences of C. coli hipO gene (corresponding to position 4024 to 4129, C. coli numbering). Boxes represent C. coli forward primer (OrfATaq-4024) and C. coli reverse primer (OrfATaq4109), C. coli specific TaqMan probe (TaqColi-4075).

124

Theoretical analysis using BLAST and FASTA against the GenBank database (version 127) indicated that the combined use of the 3 sets of primers and probes allowing detection and identification of Campylobacter species, C. jejuni and C. coli in a multiplex real-time PCR. The fluoroprobes and primers were synthesised by Proligo (Australia), Qiagen (Australia) and Synthegen (USA).

7.3.3. Optimising conditions for the Multiplex Real-Time PCR Separate real-time PCR assays were used for the optimisation of annealing and extension parameters for each of the corresponding paired primer and TaqMan probe set. The reactions performed using the temperature gradient option (57 to 64oC) in an iCycler IQTM. The results indicated that the optimum annealing and extension for each of the paired primer and TaqMan probe set was 60оC. The result obtained was in accordance with the theoretical calculations.

7.3.4. Developing of a Multiplex Real Time PCR Assay After optimisation of the annealing and extension temperature (60оC), the real-time PCR multiplex was optimised by placing each set of primers and probes in separate wells of a 96-well PCR plate (Figure 7.3). Finally, a one tube real-time PCR multiplex reaction (Figure 7.4) was optimised by fixing the annealing/extension temperature and changing the concentration of the primers and probes from 0.2 to 0.75 μM, and dNTPs from 0.15 to 0.4 μM. Optimised conditions were used in this study as described in section 7.3.3. The same assays have also been performed on the LightCyclerTM under the same conditions except the reaction volume, which was reduced to 10μl. All samples gave positive results using either one probe in a separate run, three probes in one run but in separate tubes, or the three probes in a single reaction.

125

7.3.5. Real Time Detection of C. jejuni and C. coli Using 5’ Nuclease Assay Probe (TaqMan Probe) Figure 7.3 and 7.4 shows an increase in fluorescence was observed during real-time PCR multiplex with DNA templates prepared using CTAB methods and rapid boiling methods for C. coli, C. jejuni, C. hyointestinalis and C. upsaliensis, C. lari, C. sputorum and C. fetus, and enrichments cultures initiated from chicken samples. This indicated that the primer sets used had primed the target region generating three different amplicons 105, 151 and 170 bp representing C. coli, C. jejuni and Campylobacter species respectively, as expected (Figures 7.5). Both figures (Figure 7.3 and 7.4) show the increase in the fluorescence when one tube, one fluoroprobe or multi fluoroprobes were used. Furthermore, Arcobacter species, E. coli, Staphylococcus aureus,

Clostridium perfringens, Micrococcus luteus, and Pseudomonas aeruginosa, have been tested and showed negative results (result not shown).

7.3.6. Identification of Campylobacter and C. jejuni and C. coli Isolates in the Culture Collection and Enrichment Cultures using the Triplex Real Time Assay The assay was subsequently used to identify and differentiate 176 Campylobacteraceae isolates of animal, human, plant and bird origin held in our culture collection, into C.

coli (76 isolates), C. jejuni (88 isolates) and other Campylobacter species (12 isolates) (Appendix 1). Furthermore, 27 out of 30 enrichment cultures initiated from chicken samples identified as Campylobacter species, 11 as C. jejuni and 22 as C. coli (Appendix 10). In addition, more than one Campylobacter species could also be detected in the enrichment culture.

126

7.3.7. Comparison between Conventional Culture assay and Triplex Real Time PCR Assay In this chapter, the results of Triplex real-time PCR assay show a similar result to that obtained in Chapter 3 and 4, which compared the assay with conventional culture methods. Moreover, C. coli detected by real-time PCR targeting the orfA gene concurred 100% with the LDR results. Conventional methods used here provided 63.6% accuracy compared to molecular methods, where the sensitivity and specificity were 36.3% and 100% respectively. Detection of C. coli by conventional methods showed very low sensitivity compared to the molecular methods. This is perhaps due to the presence of C. coli in low numbers below the colony number that could be detected by conventional culture methods, or because of the presence of VNC C. coli. On the other hand, molecular methods are able to detect the DNA fragment target of any bacterial species under investigation, especially with the ability of the molecular methods to detect a very low cell number (in this study 125 cells). Also, C. jejuni shows 100% similarities to those results obtained in Chapter 4 using SYBR green I assay targeting

hipO gene and Chapter 9 LDR assay by targeting flaA gene. Conventional culture methods provide 94% accuracy. Furthermore, the false negative rate was 6%, and the false positive rate was 0%. Therefore, the sensitivity and specificity were 94% and 100% respectively. This lack of sensitivity and specificity of culture methods could be due to the presence of a typical C. jejuni and/or the ability of culture methods to detect the dominant species.

127

37

34

31

28

25

22

19

16

13

10

7

4

180 160 140 120 100 80 60 40 20 0 1

Fluorescence

Figure 7.3 Single tube TaqMan amplification of DNA extracted from Campylobacter isolates, using iCycler iQTM, with primers and probe for identifying and detecting Campylobacter 16S rRNA genes, C. jejuni hipO genes and C. coli orfA genes. The fluorescence of each sample during TaqMan hydrolysis was plotted against the PCR cycle number. C. jejuni (ATCC 940565) 16S rRNA gene (-♦-), C. coli (NCTC 11366) 16S rRNA gene (-▲-), C. hyointestinalis (QHSS 99M2318) 16S rRNA gene (X-) and C. upsaliensis (QHSS 99M126) 16S rRNA gene (-□-), C. jejuni (ATCC 940565) hipO gene (-●-), and C. coli (NCTC 11366) orfA gene (-■-) and negative control lacking template (-∆-). Similar results were obtained with a further 206 samples (see appendices 1and 10).

Cycle Num ber

Figure 7.4 Multiplex TaqMan amplification of DNA extracted from representative Campylobacter isolates, using iCycler iQTM, with primers and probes for identifying and detecting Campylobacter 16S rRNA genes, C. jejuni hipO genes and C. coli orfA genes. The fluorescence of each sample during TaqMan hydrolysis was plotted against the PCR cycle number. Tube 1 contains one set of primers and a probe to simultaneously detect C. hyointestinalis (QHSS 99M2318) 16S rRNA genes, (-□-). Tube 2 contains one set of primers and a probe to simultaneously detect C. upsaliensis (QHSS 99M126) 16S rRNA genes (-◊-). Tube 3 contains two sets of primers to simultaneously detect C. coli (NCTC 11366) orfA genes (-♦-), and C. coli (NCTC 11366) 16S rRNA genes (-●-). Tube 4 contains three sets of primers and probes to simultaneously detect C. jejuni (ATCC 940565) hipO genes (-▲-), C. coli (NCTC 11366) orfA genes (-∆-) and C. hyointestinalis (QHSS 99M2318) 16S rRNA genes (-■-). Similar results were obtained with a further 206 samples (see appendices 1 and 10).

128

Figure 7.5 A representative agarose gel electrophoresis of multiplex real-time PCR (TaqMan probes) of DNA extracted from Campylobacter isolates, using iCycler iQTM, with primers and probes for identifying and detecting Campylobacter 16S rRNA genes, C. jejuni hipO genes and C. coli orfA genes. (Figure 7.4) The assay generated three amplicons 105, 151 and 170 bp, representing C. coli, Campylobacter species and C. jejuni respectively. Lanes 3 to 8 shows three PCR products generated from a single tube containing three sets of primers and probes. Lane 2 shows PCR product generated from one tube containing two sets of primers and probes. Lanes 9 and 10 shows PCR products generated from a single tube containing one set of primers and a probe. Similar results were obtained with a further 206 samples (see appendices 1 and 10)

7.3.8. Sensitivity of the Multiplex Real Time PCR Assay 7.3.8.1 Multiplex Real-Time PCR detection of C. coli and C. jejuni from Different DNA Concentrations The increase in the fluorescence signal due to TaqMan probe hybridisation and hydrolysis during real-time PCR was directly related to the initial DNA concentration.

C. coli and C. jejuni CTAB prepared DNA (Section 2.2.1) with a concentration from 142 fg/μl to 142 ng/μl for C. coli and 156 fg/μl to 156 ng/μl for C. jejuni was used in real-time PCR (Figures 7.6, 7.8, and 7.10). Accordingly, a detectable fluorescence increase started as early as cycle 15, 15 and 13 for C. jejuni, C. coli and Campylobacter species respectively, and in the case of DNA concentration of 156 fg/μl and 142 fg/μl the initial increase in fluorescence started at cycles 30, 29 and 26, for C. jejuni, C. coli and Campylobacter species respectively. Plotting the Log10 DNA concentration against the cycle threshold values generated a straight line describing the relationship between 129

Figure 7.6 Sensitivity of the Campylobacter 16S rRNA gene probe and primers set (Section 7.3.1) in detecting C. jejuni. Purified C. jejuni DNA was used as the template in different concentrations of 156 ng/μl (-▲-), 15.6 ng/μl (-+-), 1.56 ng/μl (×-), 156 pg/μl (-Χ-), 15.6 pg/μl (-■-), 1.56 pg/μl (-∆-), and 156 fg/μl (-♦-), representing Ct values in the range 13-26.

Figure 7.7 A Standard curve for 16S rRNA C. jejuni genomic DNA quantitative real-time PCR using TaqMan probes. The DNA was serially diluted 10-fold. The Ct values were plotted against log10 DNA concentration and the equation and R2 determined in Microsoft Excel. The straight line, which was calculated by linear regression [Ct =-1.0392 log (DNA concentration)+ 18.498], shows an R2 of 0.9675. (The experiments were repeated three times with identical results.)

130

Figure 7.8 Simultaneous detection and amplification of the 10-fold serial dilution of purified C. coli DNA. Detection was done using a sequence-specific TaqMan probe, targeting C. coli orfA gene. Purified C. coli DNA was used as the template in concentrations of quantities of 142 ng/μl (-■-), 14.2 ng/μl (-▲-), 1.42 ng/μl (-+-), 142 pg/μl (-X-), 14.2 pg/μl (-●-), 1.42 pg/μl (∆--), 142 fg/μl (-♦-), representing Ct values in the range 15-29.

Figure 7.9 A Standard curve for 10-fold serial dilutions of purified C. coli genomic DNA for orfA gene quantitative real-time PCR using TaqMan probes. Quantification was performed by determining Ct values. Ct values were then plotted against log10 DNA concentration and the equation and R2 determined in Microsoft Excel. The straight line, which was calculated by linear regression [Ct =-0.9461 log (DNA concentration)+ 20.439], shows an R2 of 0.996. One microliter of DNA was used in a 20 μl PCR reaction. (The experiments were repeated three times with identical results.)

131

Figure 7.10 Sensitivity of the C. jejuni hipO gene probe and primers set in detecting C. jejuni CTAB DNA. 10 –fold serially diluted purified C. jejuni DNA was used as the template in concentration of 156 ng/μl (-▲-), 15.6 ng/μl (-+-), 1.56 ng/μl (-×-), 156 pg/μl (-●-), 15.6 pg/μl (-∆-), 1.56 pg/μl (-■-), and 156 fg/μl (-♦-), representing Ct values in the range 15-30.

Figure 7.11 A Standard curve for C. jejuni genomic DNA for hipO gene quantitative realtime PCR using TaqMan probes. The DNA was serially diluted 10-fold. The Ct values were then plotted against log10 DNA concentration and the equation and R2 determined in Microsoft Excel. The straight line, which was calculated by linear regression [Ct =-1.1323 log (DNA concentration)+ 21.325], shows an R2 of 0.9825. One microliter of DNA was used in a 20 μl PCR reaction. (The experiments were repeated three times with identical results.)

132

those two factors (Figures 7.7, 7.9, and 7.11). A standard curve for C. jejuni was generated by using data derived from the serial dilution of C. jejuni 16S rRNA gene. The linear correlation coefficient of the standard curve was R2 = 0.9675, 0.996, 0.9825 representing Campylobacter, C. coli, and C. jejuni standard curves respectively, demonstrating the accuracy of the real-time PCR-based quantification.

7.3.8.2 Multiplex Real Time PCR detection of C. coli and C. jejuni from Single Colony Serial Dilution As few as 125 cells were detected by the assay when a DNA template extracted by the rapid boiling or CTAB methods, from a serial dilution of a single colony of C. jejuni or

C. coli, was used indicating that the method was very sensitive (Figure 7.12, Result for C. coli and Campylobacter species not shown). As shown in figure 7.12, the DNA extracted from the 125 cells shows an increase in fluorescence starting at cycle 28 and in the case of 500000 cells, the initial increase starts at cycle 20. A standard curve for the hipO gene was established using 10-fold dilution series of one colony of C. jejuni. The plots of the Ct values against log10 cell numbers were linear between 500000 cells and 110 cells. The R2 of the standard curve was R2 = 0.9872, demonstrating the accuracy of the real-time PCR-based quantification (Figure 7.13). These results are similar to our previous studies (Sections 3.2.8 and 5.2.8).

133

Figure 7.12 A single colony of C. jejuni (ATCC 940565) was serially diluted and DNA was extracted by the rapid boiling method and used in multiplex real-time PCR using primers and probes targeting hipO gene. The number of cells in each dilution was determined by plating onto BAPs 500000 cells (▲), 50000 cells (■), 5000 cells (♦), 500 cells (Δ), 125 cells (×) and 50 cells (●).

Figure 7.13 A standard curve showing the relation ship between Ct and log10 cells number for serialy dilution of one colony of C. jejuni. The straight line, which was calculated by linear regression [Ct =-1.0934 log10 (Cell number)+ 33.713], shows an R2 of 0.9872. Two microliter of DNA was used in a 20 μl PCR reaction.

7.4. General Discussion Rapid detection and identification of C. coli and C. jejuni from other Campylobacter species is vital from a clinical point of view. This is because it provides valuable data 134

for disease surveillance and epidemiological studies, which have implications for the design of intervention and prevention strategies [277]. The necessity also stems from the fact that up to 68% of C. coli are resistant to erythromycin, the standard antibiotic used for the treatment of Campylobacter infections, and thus rapid assay for detection will be useful in targeting/monitoring resistant outbreaks [278]. Differentiation of Campylobacter species by culture methods is problematic due to their slow and fastidious growth requirements, as well as their biochemical inertness [10, 11, 279]. For instance, discrimination between C. jejuni and C. coli by conventional culture methods requires up to 5 days. Moreover, the discrimination between C. jejuni and C.

coli is based solely on the ability of C. jejuni to hydrolyse hippurate, while C. coli lacks this ability [126]. However, 10% of C. jejuni isolates fail to hydrolyse hippurate in vitro, thereby giving erroneous results [91, 98, 280]. Growth on minimal media and αhemolytic activity studies do not provide absolute definition between C. coli and C.

jejuni and are rarely used in standard clinical analysis. For this reason, a variety of molecular methods have been developed including PCR, SDS-PAGE, and PCR-RFLP [197, 252, 281]. In this study, three different TaqMan (5’ nuclease assay) probes targeting three different genes were used to detect all Campylobacter species, as well as to distinguish between C. coli and C. jejuni from each other and from all other

Campylobacter species in one tube real-time PCR assay. The theoretical calculation of the Tm for the primers and probes was estimated to be between 56 oC and 60.6 oC, while the three TaqMan probes had a Tm range between 66 o

C and 71 oC. The Tm for the fluorogenic probe is 5oC to 10oC higher than that of the

amplification primers, enabling the probe to hybridise with the target region of the amplicons during the annealing / extension step of each PCR cycle. After that, the 5’ -> 3’ exonuclease activity of Taq DNA polymerase will hydrolyse TaqMan probe, thereby separating the quencher from the reporter during extension [136]. Due to the release of

135

the quenching effect on the reporter, the fluorescence intensity of the reporter dye will increase, which is then captured by the CCD detection camera system. The increase in fluorescence intensity of the reporter dye is a direct consequence of a successful PCR, this procedure can be used in the detection of specific DNA sequences. TaqMan assay offers a sensitive, reliable, cost-effective and automated method to determine the presence or absence of specific sequences. An increase in fluorescence was observed during real-time PCR with DNA templates from all Campylobacter species, but not from the closely related genus Arcobacter 16S rRNA gene. Furthermore, an increase in fluorescence was also detected from C. coli only using the orfA gene. C. jejuni PCR primer sets target the hipO gene, and generate a 170 bp amplicon to which the TaqJej-1864 TaqMan probes will hybridise to generate an increase in the fluorescence signal during the PCR amplification (Figures 7.3 and 7.4). As expected, three different amplicons of approximately 104, 151, and 170 bp were observed on agarose gels for the DNA templates from Campylobacter species, C. jejuni and C. coli respectively, but not from E. coli or Arcobacter using the 16S rRNA, and C.

hyointestinalis, C. upsaliensis using hipO and orfA genes (Result not shown). These results confirm the specificity of the primers and probes used as well as the TaqMan (5’ nuclease assay) probe assay detection abilities. To visualise the multiplex PCR product, a 4.5% ultra pure agarose 1000 (Invitrogen, Australia) gel was used, due to the poor resolution of the 1% DNA grade agarose, which did not distinguish between the 170 and 151 bp amplicons (Figure 7.5). As mentioned above, the Idaho LightCylerTM was used in this study, but this instrument is unable to differentiate different fluorophores signal simultaneously and hence cannot do multiplex real-time PCR, as done using the iCycler iQTM. However, agarose gel electrophoresis analysis of the real-time PCR products performed using LightCyclerTM showed bands similar to the PCR products bands generated by the iCycler iQTM.

136

Efficiency of probe hybridisation depends on reaction conditions including annealing and extension temperature, MgCl2 concentration, DNA concentration, probe Tm, and presence of secondary structure in probe and template. Multiplex real-time PCR can be performed on three- to four-channel detection systems, which may allow monitoring of three-to four probes in a single reaction using TaqMan probes or any other probe technology currently available. The multiplex TaqMan probe assay has been used widely to detect and discriminate between closely related pathogens, mutations and viruses from clinical specimens and environmental samples [116, 133, 235, 237, 270, 282-286]. Rapid detection and identification of C. jejuni and

C. coli from other Campylobacter species is very important as these two species cause more than 95% of the reported cases of campylobacteriosis. Previous studies for rapid detection and identification of Campylobacter species used a variety of different molecular techniques including PCR targeting the 16S rRNA gene, as well as many other genes [102, 125]. However, these methods show poor detection and are relatively time-consuming to perform. Real-time PCR described in this study has several advantages over conventional PCR. These include the decreased risk of carry over contamination because analysis of PCR products is carried out without opening the reaction vials. Results are available in real time PCR and the Tms are obtained after completion of the amplification reaction. PCR-RFLP is also used for differentiation between C. jejuni and C. coli [125], PCR-RFLP is a very powerful typing assay and has the ability to differentiate Campylobacter up to species or sub species level. But this technique has several disadvantages including the use of the carcinogen EtBr for detection of dsDNA. The technique is also time consuming, and the resulting analysis and interpretation are very difficult requiring highly qualified scientists for data interpretation. Recently, several different real-time PCR methods have been described for the detection and identification of Campylobacter species, including C. jejuni and C.

137

coli. A Tm analysis developed by Logan et. al. (2001) [102], and multiplex detection of C. jejuni and C. coli developed by LaGier et al (2004) [126] have been reported. The method described by Logan et al. (2001), is very sensitive but it lacks the ability to distinguish between C. jejuni and C. coli, whereas LaGier et al.’s (2001) assay successfully distinguishes between C. coli and C. jejuni, but lacks the ability to detect all other Campylobacter species. We have previously described a combined use of a two-tube adjacent hybridisation real-time PCR assay and SYBR green I assay for the rapid detection and identification of C. coli and C. jejuni from other Campylobacter species (Chapter 3, 4) [125] . This method, as described in chapter 3 and 4, is very sensitive and specific but the assay lacks the ability to identify C. coli from other

Campylobacter species as it does for C. jejuni. 7.5. Conclusion The triplex assay described in this study can be completed within 1.5 hours. It is automated and therefore the risk of carry over contamination is reduced, the assay is flexible and can be modified by including additional probes to identify further species. Our study employed an iCycler IQTM with the capacity for four-channel detection, and 96-samples. Multiplex real-time PCR assays can play a major role both in the diagnostic laboratory and in environmental studies of the incidence and significance of campylobacters in human disease, especially for those species that are not acquiescent to routine culture. The triplex real-time PCR assay has been successfully used in this study for the rapid detection of Campylobacter species, and to distinguish between the closely related species C. jejuni and C. coli. TaqMan probes have been employed in this study, targeting three different genes: 16S rRNA gene, hipO and orfA, to detect and identify

Campylobacter species, C. jejuni, and C. coli respectively. Methods described in this chapter prove to be sensitive, specific, rapid and relatively cost-effective compared to 138

the conventional culture methods. The methods described here are able to detect bacterial levels as low as 125 CFU and 156 fg/μl for C. jejuni, and as low as 110 CFU and 142 fg/μl for and C. coli. The rapid boiling DNA extraction method proved to be a very efficient method for the extraction of Campylobacter DNA from enrichment cultures as well as from BAP cultures in this assay. The iCycler IQTM has been used successfully in this study to perform real-time PCR multiplexing, while the LightCyclerTM successfully amplified the target region but its ability to read only one fluorophore limited its use.

139

8. Chapter 8 Screening and Identification of Campylobacter and Arcobacter Species up to Species Level using T-RFLP 8.1. Introduction T-RFLP has been previously used for community screening, and to study changes in the bacterial communities in soil and water (see chapter 1, section 1.7.1). In most previous cases, tetrameric (4 base cutters) restriction enzymes and bacterial universal primers for 16S rRNA genes have been used. From these studies, it was concluded that the T-RFLP is a sensitive technique that could be a valuable tool for the rapid identification of pure cultures or selectively enriched mixed microbial populations. This chapter describes the use of T-RFLP for the detection and identification of Campylobacter and Arcobacter (Figure 2.1). The main objective was to develop and validate a high-throughput procedure for the detection and identification of Campylobacter and Arcobacter in enrichment cultures from chicken samples by using a 16S rRNA PCR primer set, followed by restriction digest, and sizing of the resulting TRFs. TRFs were analysed using the T-RFLP analysis program (TAP) which forms part of the Ribosomal Database Project (RDP) suite of programs (http://www.cme.msu.edu/rdp/TRFLP/#program). TAP is an online multiplatform program located on a web server, which uses a common gateway interface to handle the priming digest calculations and accesses the RDP database for analysis. The service interface is implemented as an applet in the Java 2 programming language [287]. TAP results are displayed in three different configurations: the organism’s genus and species name, the RDP identifier, and one or more fragment sizes and enzyme pairs.

140

A schematic representation of this is shown in Figure 8.1. The default data output configuration is the prokaryotic phylogenetic hierarchy list but these results can be sorted by the user.

Figure 8.1 Representative TAP sorted display windows, showing RDP organism identifiers, TRF sizes, organism names and restriction enzymes (DdeI, EcoRI and KpnI) used in this assay. Organisms are sorted alphabetically in ascending order by their RDP sequence identifiers.

141

8.2. Methods The general methods used in this study have been described in detail in Chapter 2 and only specific methods relating to this chapter are described below.

8.2.1. Primer Design and Synthesis 16S rRNA sequence data of Campylobacter and Arcobacter species and representatives of various phyla of domain Bacteria were extracted from GenBank and the RDP database and aligned using BioEdit software [31, 275, 276]. Several primer sets for TRFLP were selected for synthesis using the guidelines described in section 2.4.1.

8.2.2. Computer Simulation of 16S rRNA TRF from RDP Database The RDP-II Release 8.1 [276] which includes more than 16,277 small subunit (SSU) rRNA sequences, of which 4,430 type cultures sequences were used in T-RFLP computer

simulation

analysis

via

the

T-RFLP

analysis

program

(TAP)

(http://www.cme.msu.edu/rdp/T-RFLP/#program). For TAP simulation analysis, the labelled forward or reverse primer sequence that contains non-Watson-Crick International Union of Biochemistry (IUB) characters, the restriction enzymes that target the restriction sites, the position and the maximum bp mismatch number allowed from the 5’end of the primer are inputted into the program. Using this information, TAP accesses the RDP database and attempts to prime each RDP sequence with the submitted user information under the selected methods match conditions. The output from this simulation allows the researcher to answer a number of questions such as; (i) What restriction endonuclease enzyme(s) will provide the highest discriminating activity for estimating the population diversity? (ii) What enzyme(s) will give the best resolution for the phylogenetic group(s) of interest? (iii) What primer-enzyme combination will be suitable for the community under investigation? [287].

142

8.3. Results 8.3.1. Primer Design and Synthesis A forward primer designated FAM-49F (5’-GTGYCTAAYACATGCAAGTCGAACG3’) (E. coli position 49-74) fluorescently labelled with 6-FAM at the 5’end and reverse primer R2 (5’-GTATTACCGCGGCTGCTG-3’) (E. coli position502-519) were designed after inspection of the aligned 16S rRNA gene sequences of Campylobacter,

Arcobacter species and representatives of various phyla of domain Bacteria that had been extracted from GenBank and the RDP database (Figure 8.2). Theoretical analysis using BLAST and FASTA against the GenBank database (version 127) indicated that primer FAM-49F was not 100% specific to Arcobacter and

Campylobacter species but it could also target a number of environmental clones, members of Sphingobacterium, Proteobacteria, Listeria monocytogenes, Abiotrophia,

Lactobacilli, Corynebacterium, Enterococcus, Mycoplasam, Acholeplasma laidlawii and a few unclassified taxa. In the case of primer R2, it is known that it is a universal primer and targets almost the entire domain Bacteria. The two primers FAM-49F and R2 were expected to produce 462 to 466 bp amplicons from the 16S rRNA genes of

Campylobacter and Arcobacter species and this was found to be the case (Figure 8.3). Other bacterial species were expected to produce 460-485 bp amplicons.

8.3.1.1 T-RFLP Computer Simulation T-RFLP discriminative ability to distinguish phylogenetic groups of bacterial species was theoretically evaluated using a computer simulation of the 16S rRNA for the entire bacterial species deposited in the RDP with primers FAM-49F forward and R2 reverse. This simulation showed that the DdeI restriction enzyme could digest 3782 sequences;

Sau3AI could digest 3760 sequences; KpnI, 1838 sequences, and EcoRI could digest 22

143

Escherichia coli-------------------------------------------------------GGGTGAGTAATGTCTGGGAAACTGCCTGATGGAGG------CAGCAGCCGCGGTAATAC----C. sputorum

ATCC35980-----------TAATACATGCAAGTCGAACG------------------GGGTGAGTAATGTATAGCTAATCTGCCCCATAGAG------CAGCAGCCGCGGTAATAC-----

C. gracilis ATCC33236--------------------------------------------------GGGTGAGTAATATATAGCTAATCTGCCCCTTGCTG----------------------------C. concisus ATCC33237--------------------------------------------------GGGTGAGTAATGTATAGCTAATCTGCCCCATAGTG----------------------------C. curvus ATCC35224----------------------------------------------------GGGTGAGTAATGTATAGCTAATCTGCCCCATAGTG----------------------------C. mucosalis CCUG6822--------------------------------------------------GGGTGAGTAATGTATAGCTAATCTGCCCTATGCAG----------------------------C. rectus ATCC33238----------------------------------------------------GGGTGAGTAATATATAGCTAATCTGCCCTACACTA----------------------------C. hyointestinalis ATCC35217-------------------------------------------GGGTGAGTAATGTATAGTTAATCTGCCCTACACTG----------------------------C. fetus ATCC19438-----------------------------------------------------GGGTGAGTAATGTATAGTTAATCTGCCCTACACTG----------------------------C. upsaliensis CCUG14913C ---------------------------------------------GGGTGAGTAAGGTATAGTTAATCTGCCCTACACTG----------------------------C. helveticus NCTC12470------------------------------------------------GGGTGAGTAAGGTATAGTTAATCTGCCCTACACTG----------------------------C. showae CCUG3054-----------------------------------------------------GGGTGAGTAATATATAGCTAACTTGCCCATTACTA----------------------------C. lari CCUG23947------------------------------------------------------GGGTGAGTAAGGTATAGTTAATCTGCCCTACACAA----------------------------C. coli CCUG33450------------------------------------------------------GGGTGAGTAAGGTATAGTTAATCTGCCCTACACAA----------------------------C. jejuni CCUG24567----------------------------------------------------GGGTGAGTAAGGTATAGTTAATCTGCCCTACACAA----------------------------A. butzleri CCUG10373--------------------------------------------------GGGTGAGTAATGTATAGGTAATATGCCTCTTACTA----------------------------A. skirrowii CCUG10374----------------------------- -------------------GGGTGAGTAATGTATAGGTAATATGCCTCTTACTA----------------------------A. nitrofigilis CCUG15893----------------------------------------------GGGTGAGTNATATATAGGTAACATGCCCTAGAGAG----------------------------A. cryaerophilus ATCC49615---------------------------------------------GGGTGAGTAATGTATAGGTAATATGCCTCTTACTA-----------------------------

Schematic representation

5’ Fam-TAATACATGCAAGTCGAACG 3`

5’CAGCAGCCGCGGTAATAC 3`

TRFLP-49F primer

R2 536R primer

Figure 8.2 An alignment of the partial 16S rRNA sequences corresponding to position 49 to 536 (E. coli numbering according to [230]) of representatives of the Campylobacter and Arcobacter species is shown. This region is amplified using Fam-49F primer and R2 primer (indicated by left and right hand arrows) to produce a 460 bp amplicons.

144

Figure 8.3 Agarose gel electrophoresis of PCR products generated with Fam-49 forward and R2 reverse primers. C. coli (NCTC 11366) (Lane 2), C. jejuni (ATCC 940565) (Lane 3) C. upsaliensis (QHSS 99M126) (Lane 4), A. butzleri (NCTC 12481) (Lane 5), A. skirrowii (NCTC 12713) (Lane 6), shows an expected amplicon size of 462 bp, and the100 bp DNA ladder (Promega, Australia) (Lane 1).

computer simulation, it was found that combination of FAM-49F forward primer and R2 reverse primer followed by digestion with Sau3AI or DdeI could distinguish between Campylobacter and Arcobacter species while KpnI and BtsI could only digest

Campylobacter and Arcobacter species respectively (Figure 2.1). On the other hand, EcoRI could detect C. coli, C. lari and C. jejuni from other Campylobacter and Arcobacter species whilst BtsI was specific for Arcobacter species, allowing the differentiation of Arcobacter species from Campylobacter species (Figure 2.1 and Table 8.2).

8.3.1.2 T-RFLP Analysis Using ABI 377 DNA Sequencer 258 bacterial isolates, including pure Campylobacter and Arcobacter (QHSS culture collection) and enrichment cultures initiated from chicken samples (Section 2.1) were used to determine the usefulness of T-RFLP. PCR products were generated using 16S rRNA primer pairs (FAM-49 and R2), digested using five different restriction endonucleases (DdeI, Sau3AI, KpnI, BtsI and EcoRI), and purified by ethanol precipitation. These fragments were then electrophoresed using

145

an ABI 377 DNA sequencer and the sizes of TRF determined using the Global Southern analysis method part of GeneScan 3.0 software (Applied BioSystems, USA). Using this approach TRF from 35-500 bp in length were accurately sized to + 2 bp, consistent with other published results [146]. A TRF profile from DdeI restriction endonuclease enzyme digestion of the PCR products of the DNA of Campylobacter and Arcobacter pure and enrichment cultures is shown in Figures 8.4 and Table 8.1. In case of the pure cultures, a simple TRF profile is observed; a 246 bp TRF is the only fragment seen for pure cultures of Campylobacter (Panel 8.4A), whereas a 53 bp TRF can only be seen from preparations of pure cultures of Arcobacter (Panel 8.4B). However, a more complex TRF profile is observed from the

Campylobacter enrichment culture C-P1000 (see section 2.1.1 for details) in which four TRF fragments of 48, 72, 104 and 246 bp are observed (Panel 8.4C). The 48 bp TRF could represent either C. hyointestinalis or C. fetus, the 104 bp TRF fragment could identify C. sputrum and/or C. helviticus and/or C. upsaliensis, and the 246 bp TRF fragment specifically identified the C. jejuni, C. coli and C. lari group. The 72 bp TRF fragment was not expected in the profile and could theoretically represent, as analysed by the RDP software, members of Clostridium, Mycoplasma, Lactobacillus and/or

Eubacterium. The T-RFLP technique described here is reliant on PCR, which makes it a powerful and sensitive detection method. Consequently, the T-RFLP technique could detect the additional 72 bp fragment. The enrichment culture C-P1000 was initiated in PCMB under microaerophilic conditions which is selective for Campylobacter species. It is likely that Clostridium, Eubacterium and Lactobacillus, which are also microaerophilic to anaerobes, could grow under these conditions though their growth in PCMB has not been reported in the literature. Mycoplasma on the other hand are aerobic intracellular parasites and their growth under these conditions is unlikely. In addition, it is possible that members of these four genera are contaminants (dead, alive

146

or in a VNC state) in the chicken sample, a source of inoculum for the enrichment cultures (see section 2.1.1). Future experiments should be undertaken to determine whether members of Clostridium, Eubacterium, and Lactobacillus could grow in PCMB. The VNC state and / or their low numbers in PCMB may make the task of detection in culture difficult. In case of preparations of mixed DNA from enrichment cultures of Campylobacter (enrichment culture number C-P1006) and Arcobacter (enrichment culture number P10A), three TRF fragments with sizes of 53, 136 and 246 bp were observed. In this case, the TRF profile matched that of Campylobacter and Arcobacter species (Panel 8.4D). The 53 bp TRF fragment identified the Arcobacter species; the 136 bp TRF fragment identified the C. hyointestinalis, subspecies hyointestinalis, and the 246 bp TRF fragment identified the C jejuni C. coli and C. lari group. It is assumed that the enrichments did not contain any contaminating species. This suggests that the source samples may vary in their microbiological composition. Figure 8.5 (Panels 8.5A –8.5C) shows the T-RFLP patterns obtained after the Sau3A restriction endonuclease digestion of 16S rRNA amplicons were amplified using FAM49F and R2 for the DNA of Campylobacter and Arcobacter pure and enrichment cultures (Figure 8.5 and table 8.1). A 255 bp TRF fragment specifically identifies the pure culture of C. jejuni (ATCC 940565) (Panel 8.5A), while a 260 bp TRF fragments also specifically identifies the pure culture of A. butzleri (NCTC 12481) (Panel 8.5B). When amplicons, generated from mixed DNA templates of Campylobacter (C-P1002) and Arcobacter (P10-A) enrichment cultures were digested with Sau3AI, two distinctive TRF fragments, 255 and 260 bp, were found (Panel 8.5B). The 255 and 260 bp TRF fragments specifically identify Campylobacter and Arcobacter species, respectively. This demonstrates the ability of the T-RFLP to specifically detect and identify those two genera, after the digestion of the PCR amplicons generated by the FAM-49 and R2 16S

147

rRNA primer pair using the Sau3AI restriction enzymes. The presence of the 465 bp TRF fragments on panel 8.5A may be due to insufficient restriction enzyme in the reaction mixture, or insufficient incubation time to complete DNA digestion. A T-RFLP profile was obtained of the KpnI and EcoRI restriction endonuclease digestion of PCR products, amplified using 16S rRNA primer pair (FAM-49F and R2) from the DNA of Campylobacter and Arcobacter pure and enrichment cultures (Figure8.6 and table 8.1). Two TRF fragments sized at 405 and 465 bp were generated using EcoRI (panel 8.6A). The 405 bp TRF fragment identifies C. jejuni (ATCC 940565), whereas the 465 bp represents the undigested PCR products. Another two TRF fragment profiles sized at 415 and 465 bp were generated using the KpnI restriction endonuclease (Panel 8.6B). The 415 bp TRF fragment detects and identifies C. jejuni (ATCC 940565), while the 465 bp TRF fragments represents the undigested PCR product. A profile of multiple digestions of the PCR products generated from the primers FAM49 and R2 with the DNA from Campylobacter enrichments cultures (C-P1015) using both restriction endonucleases KpnI and EcoRI, shows three different TRFs sized at 405, 415 and 465 bp (Panel 8.6C). The 405 bp TRF fragments detect and specifically identify C. coli and/or C. jejuni and/or C. lari using EcoRI, whereas the 415 bp specifically identifies the Campylobacter species using KpnI restriction endonucleases, while the 465 fraction represents the undigested PCR products. In summary, figure 8.6 shows the ability of KpnI and EcoRI to detect and differentiate the Campylobacter species, since KpnI digests all the PCR products of Campylobacter species generated with FAM-49 and R2 primer pairs, but not the Arcobacter species, whereas EcoRI will only digest the PCR products of C. jejuni, C. coli, and C. lari, thus proving the ability of T-RFLP to detect and identify bacteria to species level.

148

A T-RFLP profile was obtained after the DdeI and BtsI restriction enzymes digestion of 16S rRNA PCR products using FAM-49F and R2 forward and reverse primers for the DNA of Campylobacter and Arcobacter pure and enrichment cultures (Figure 8.7 and table 8.1). Two TRF fragments, sized at 313 and 465 bp, were observed in panel 8.7A. The 313 bp TRF fragments generated from the digestion of the PCR amplicons, using

BtsI, specifically identified A. butzleri (NCTC 12481), while the 465 bp represents the undigested PCR products. On the other hand, a single 465 bp TRF fragment represents the undigested PCR products of C. jejuni (ATCC 940565), observed when the BtsI restriction endonuclease was used for digestion, due to the fact the C. jejuni and all other Campylobacter species lack the BtsI restriction site (Panel 8.7B). Panel 8.7C shows a very good example of the combination of two different restriction enzymes on the digestion of a multiplex of amplicons generated from mixed DNA templates from

Arcobacter (P11-A) and Campylobacter (C-P1009) enrichment cultures. Four different TRF fragments sized at 53, 72, 242 and 313 bp were observed (Panel 8.7C). The 53 and 313 bp sized TRF fragments that generated from using DdeI and BtsI, respectively, specifically identified the Arcobacter species. The 72 bp TRF fragment could represent

Clostridium species, Mycoplasma species, Lactobacillus species or any other unidentified bacterial species that has the ability to grow in the Campylobacter and

Arcobacter selective enrichment media, or it may have resulted from a post–selective culture contamination. The 242 bp sized TRF fragments identifies C. concisus and/or C.

rectus. In summary, figure 8.7 shows the ability of BtsI and DdeI to detect and differentiate

Campylobacter and Arcobacter species, as DdeI digests all Campylobacter and Arcobacter species, whereas BtsI will only digest the PCR product generated from the Arcobacter species DNA, using FAM-49 and R2 primer pairs.

149

Figures 8.4 to 8.7 demonstrate that as a consequence of the variation in the restriction site positions in the 16S rRNA gene sequence, the resolution of T-RFLP analysis is often reduced from the species level to that of higher order groups. For example, in this study, the T-RFLP resolution analysis has been reduced from the species level for the

Campylobacter species to the genus level for the Arcobacter species, due to the variation in the DdeI restriction site for both genera (Table 8.2). Furthermore, EcoRI has a unique restriction site, which is only found in C. coli, C. jejuni, and C. lari, allowing the discrimination of those three species from all the other Proteobacteria species, when using the 465 bp PCR amplicons generated from 16S rRNA gene primer pairs Fam-49 and R2 for the DNA of Campylobacter and Arcobacter pure and enrichment cultures. These results are in accordance with the theoretical results obtained with TAP. Figures 8.4 to 8.7 illustrate the T-RFLP results obtained from the Campylobacter and

Arcobacter species PCR products that were generated using FAM-49 and R2 with one or two restriction enzymes. Figure 8.8 shows the ability of T-RFLP assays, using multiple restriction enzymes, to detect and identify the Arcobacter and Campylobacter species. A T-RFLP profile was obtained by using five different restriction enzymes (DdeI,

Sau3AI, KpnI, BtsI and EcoRI) to digest the 16S rRNA PCR products amplified using FAM-49F and R2 for the DNA of the Campylobacter and Arcobacter (C-P1002, and P12-A) enrichment cultures. A very complex profile of seven TRF fragments sized at 72, 242, 255, 260, 405, 415, and 465 bp were observed (8.8A). The 72 bp TRF fragment could identify the Clostridium species or any other bacterial species that could be amplified using FAM-49 and R2 primers, followed by the digestion of the generated PCR products, with any of the five restriction endonucleases used in this study. The 260 bp TRF fragments specifically identified the Arcobacter species generated with Sau3AI. The presence of the different Campylobacter species in this mixed sample is represented

150

by the 242 bp TRF fragment that could identify C. helviticus and/or C. rectus generated using DdeI, whereas the 405 bp TRF fragment generated using EcoRI could identify C.

coli or C. jejuni or C. lari. Furthermore, the presence of those five species or any other Campylobacter species is confirmed by the presence of 255 and 415 bp sized TRF fragments generated with Sau3AI and KpnI, respectively. On the other hand the 465 bp TRF fragment represents the undigested PCR products. Figure 8.8 shows that the combined use of restriction endonucleases DdeI, Sau3AI,

EcoRI and KpnI with preparation from the enrichment culture number C-P1017 and P12-A producing a very complex TRF pattern. This makes predictions of bacterial groups or species difficult but Panel 8.8A suggests otherwise as a very accurate prediction appears possible. The presence of Campylobacter species can be confirmed by the presence of 242, 255, 405 and 415 bp TRF fragments which are also generated from individual digestion by DdeI, Sau3AI, EcoRI and KpnI, respectively (Figures 8.5, 8.6, 8.7 and table 8.1) and in accordance with theoretical results described in table 8.2. The additional 72 bp TRF fragment, generated from the activity of any of these restriction endonucleases could be similar to that described previously for the enrichment culture number C-P1000, will makes the identification of the species or genus difficult. This figure shows the ability of T-RFLP to detect and differentiate bacteria to the species level. It also shows that the 53 TRF fragment peak, representing the Arcobacter species, which is expected to appear using DdeI based on the results obtained by TAP analysis, is missing. This may be due to the insufficient restriction enzyme being added, and/or the low concentration of template in the reaction mixture. The same may occur with C. jejuni, C. coli, and C. lari, which is digested by DdeI and produces a 246 bp TRF. Furthermore, the electropherograms generated from enrichment cultures of the chicken samples showed that some peaks have a higher fluorescence ratio than others. This is

151

related to the availability of the restriction enzyme, the competition between the restriction enzyme to bind with and digest the restriction site, and whether the restriction site in the highly variable region or the highly conserved region of the 16S rRNA gene sequence. When the restriction sites are in the highly conserved region, the T-RFLP results appear as a cluster of large peaks; which contain representatives of broadly different taxa (species, genus, group or family, etc.) in the electropherograms because many different taxa have similar TRFs. For example, the Campylobacter 16S rRNA TRF fragments generated using KpnI, had a higher fluorescence signal compared to the fluorescence signal generated from the Campylobacter 16S rRNA using EcoRI (Panel 8.6C). This is due to the presence of KpnI endonuclease recognition site in the highly conserved region of the Campylobacter 16S rRNA genes, while EcoRI is found only in

C. coli, C. jejuni and C. lari. On the other hand, if the restriction enzyme is in the highly variable region, the result may appear as a peak that represents an individual species or strain. This peak may be large if the species is very common in the sample. Therefore

KpnI will digest all the Campylobacter species including C. coli, C. jejuni, and C. lari (8.6A).

152

53

A 246

48 72 104

53

136

B

246

C

246 D

E

Figure 8.4 A typical electropherogram of T-RFLP profiles obtained after DdeI restriction enzyme digestion of the 16S rRNA amplicons produced with primers FAM-49F and R2 and DNA of pure and selective enrichment cultures of Campylobacter and Arcobacter. A 53 bp TRF identifying a pure culture of A. butzleri (NCTC 12481) (Panel 8.4A) and a 246 bp TRF identifying C. jejuni (ATCC 940565) (Panel 8.4B) are shown. A profile of four TRFs representing 48, 72, 104 and 246 bp from DNA of a selective enrichment culture grown from the chicken sample (C-P1000) is shown in Panel 8.4C. The 48 bp TRF identifies C. hyointestinalis and/or C. fetus, the 72 bp TRF identifies Clostridium, and/or Lactobacillus species, and/or Mycoplasma species, the 104 bp TRF specifically detects C. sputrum, and the 246 bp TRF identifies C. jejuni. A similar complicated but different profile was observed when DNA from enrichment cultures initiated from chicken samples (C-P1006, and P10-A) was amplified and the amplicons digested with DdeI (Panel 8.4D). The 53 bp TRF identified Arcobacter species, the 136 bp TRF identified C. hyointestinalis subspecies hyointestinalis and the 246 bp TRF identified C. jejuni and/or C. lari and or C. coli. GeneScan®-500 (TAMRA) internal (Applied BioSystems, Australia) with a size range between 50 bp and 500 bp (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500 bp) was used as an internal standard (Panel 8.4E).

153

255

A

260

B

255 260

C

465

D

Figure 8.5 A typical electropherogram of T-RFLP patterns obtained after Sau3AI restriction enzyme digestion of the 16S rRNA PCR products amplified with FAM49F and R2 primers and the DNA of pure and selective enrichment cultures of Campylobacter and Arcobacter species. A profile of two TRFs representing 255 and 465 bp from DNA of pure culture is shown (Panel 8.5A). A 255 bp TRF identifying C. jejuni (ATCC 940565), and the 465 bp TRF represent the undigested PCR product respectively. A single 260 bp TRF identifying A. butzleri (NCTC. 12481) is shown (Panel 8.5B). Two distinctive TRFs representing a 255 and 260 bp TRF generated from a PCR products of mixed DNA templates from Campylobacter and Arcobacter enrichments (C-P1002 and P10-A) amplified using 16S rRNA gene FAM-49 and R2 primers pair is shown (Panel 8.5C). The 255 bp TRF identifies Campylobacter species, whereas the 260 identify Arcobacter species. GeneScan®-500 (TAMRA) internal standard (Applied BioSystems, Australia) with a size range between 50 bp and 500 bp (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500 bp) was used for TRF size comparison (Panel 8.5D).

154

A

B

C

405

415

405 415

460 465

D

Figure 8.6 A representative electropherogram of T-RFLP patterns obtained after restriction endonuclease digestion with KpnI and EcoRI of the 16S rRNA PCR amplicons amplified with FAM-49F and R2 primers and the DNA of pure and enrichment cultures of Campylobacter and Arcobacter species. A 405 and 465 bp TRFs, generated from the digestion of the C. jejuni PCR amplicons with EcoRI is shown (Panel 8.6A). A 405 bp TRF identifies C. jejuni (ATCC 940565), while the 465 bp TRF represents the undigested PCR product. Further two TRFs representing 415 and 465 bp generated from the PCR products of C. jejuni (ATCC 940565) using KpnI is shown (Panel 8.6B). A complex T-RFLP profiles shows two distinctive TRF representing 405 and 415 bp, and two large size TRFs representing 460 and 465 bp TRF generated from the digestion of PCR amplicons of Campylobacter enrichment cultures (C-P1015) with both restriction endonucleases KpnI and EcoRI (Panel 8.6C). The 405 bp TRF identifies C. lari or C. jejuni or C. coli group, whereas the 415 bp identifies Campylobacter species. The remaining 460 and 465 bp TRFs represent the undigested PCR products. Panel 8.6D shows the GeneScan®-500 (TAMRA) internal standard (Applied BioSystems, Australia) (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500).

155

A

313

B

53, 72

465

C

242

313

D

Figure 8.7 A typical electropherogram of T-RFLP patterns obtained after both BtsI and DdeI restriction endonuclease digestion of the 16S rRNA PCR products amplified with FAM-49F and R2 primers and the DNA of pure and enrichment cultures of Campylobacter and Arcobacter species. The presence of the 313 and 465 bp TRF generated from the digestion of the 16S rRNA PCR products of A. butzleri (NCTC 12481) using BtsI restriction endonuclease is shown (Panel 8.7A). The 313 bp TRF specifically detects and identifies A. butzleri (NCTC 12481), whereas the 465 bp TRF represents the undigested PCR products. A single 465 bp TRF representing the undigested PCR products, when BtsI is used to digest the PCR products of DNA from C. jejuni (ATCC 940565) (Panel 8.7B). A complex T-RFLP of 4 TRFs representing 53, 72, 242, 313 bp generated from digestion of the multiplex PCR amplicons of mixed DNA templates from Campylobacter and Arcobacter enrichments (P11-A and C-P1009) using BtsI and DdeI restriction enzymes is shown (Panel 8.7C). The 53 and 313 bp TRF identifies Arcobacter species, while the 242 bp TRF identifies either C. concisus or C. rectus. The 72 bp TRF represents the Clostridium species. Panel 8.7D shows the GeneScan®-500 (TAMRA) internal standard (Applied BioSystems, Australia) (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500).

156

72

A

242 255 260

405 415

465

B

Figure 8.8 A representative electropherogram of T-RFLP patterns obtained after Sau3AI, EcoRI, KpnI, BtsI and DdeI restriction enzymes digestion of the multiplex PCR products amplified FAM-49F and R2 targeting the 16S rRNA genes of the DNA extracted from the enrichment cultures (C-P1017, and P12-A). A very complicated T-RFLP profile of seven TRFs representing 72, 260, 242, 255, 405, 415, and 465 bp is shown (Panel 8.8A). The 72 bp TRF could identify Clostridium species or any other bacterial species generated after digestion of the PCR amplicons with one of the restriction enzymes used in this study. The 313 and 260 bp TRF identifies Arcobacter species generated from the digestion of the PCR products with DdeI, BtsI and Sau3AI respectively. The 242 bp TRF represents C. rectus or C. concisus generated from the digestion of the PCR products using DdeI. The 405 bp TRF could identify C. coli or C. lari or C. jejuni generated from the digestion of the PCR products with EcoRI. The 255 and 415 bp TRFs identifying Campylobacter species, is generated from the digestion of PCR products with Sau3AI and KpnI respectively. Panel 8.8B shows the GeneScan®500 (TAMRA) internal standard (Applied BioSystems, Australia) (50, 75, 100, 139, 150, 200, 250, 300, 340, 350, 400, 450, 490, and 500).

157

Table 8.1 This table summarises the results shown in figure 8.4-8.8. It shows the TRF’s size of the five restriction enzymes used in this study to digest the PCR products of 16S rRNA gene amplified using FAM-49 and R2 primers as described in sections 2.8.1, 2.8.2 and 2.8.3. The TRFs size were determined by the ABI prism 377 as described in section 2.8.5. This table shows the TRFs resulting from the digestion of the PCR amplicons using a single restriction enzyme or a combination of these enzymes.

Restriction enzyme

DdeI

TRF length (bp) 48

Bacteria species or genus

53

A. butzleri

72

Clostridium, Lactobacillus and Mycoplasma C. Sputrum

104 136

C. hyointestinalis C. fetus

242

C. hyointestinalis subspecies hyointestinalis C. concisus, and C. rectus

246

C. coli, C. jejuni and C. lari

255

Campylobacter species

260

Arcobacter species

KpnI

415

Campylobacter species

EcoRI

405

C. jejuni, C. coli and C. lari

BtsI

313

Arcobacter species

All 5 restriction enzyme

72, 242, 255, 260, 405, 415 460-465

As shown in the other part of the table.

Sau3AI

Any of the 5 restriction enzyme

Undigested PCR product

158

8.3.2. Identification of Campylobacter and Arcobacter Isolates in the Culture Collection and enrichment cultures using the T-RFLP Assay T-RFLP assay was subsequently used to identify and differentiate the 198

Campylobacteraceae isolates of animal, human, plant and bird origin held in our culture collection, into C. jejuni, C. coli and C. lari group (165 isolates), other Campylobacter species (11) and Arcobacter species (22) (Appendices 1 and 2). Furthermore, 60 enrichment cultures were tested and successfully identified Campylobacter species (27),

C. jejuni, C. coli and C. lari group (26), C. helviticus and/or C. rectus and/or C. concisus (4) other Campylobacter species (4), and Arcobacter species (22) (Appendices 11, 12 and 13). More than one Campylobacter species could also be detected in the same enrichment culture.

8.3.3. Comparison between Conventional Culture assay and T-RFLP assay T-RFLP sensitivity is in accordance with the other real-time PCR (chapters 3 to 7), and LDR assays (Chapter 9). Furthermore, T-RFLP shows the same degree of accuracy, sensitivity and specificity to other molecular methods already described in chapters 3 to 7.

8.4. General discussion This chapter describes the development of a method, which is based on T-RFLP. To my knowledge, this is the first time this method has been used to identify and differentiate

Campylobacter and Arcobacter species. The method uses a primer pair, designated FAM-49F and R2, to generate 460-465 bp amplicons from 16S rRNA genes. The amplicons digested with DdeI, Sau3AI, BtsI, KpnI and EcoRI produced several different profile patterns, which can be used to identify Arcobacter and Campylobacter species. Examples of this include the use of Sau3AI, which produces a 255 bp TRF from

Campylobacter species and 260 bp from Arcobacter species, allowing the detection and 159

identification of those two genera (Figure 8.5). Also KpnI produces 415 bp TRF from

Campylobacter species whereas Arcobacter species, which lacks a KpnI restriction site, does not generate any TRF fragments. In addition, BtsI produces a 313 bp TRF from

Arcobacter species but no TRFs are generated from the Campylobacter species as they lack a BtsI restriction site. The T-RFLP method developed as part of my study is also useful for identifying and differentiating Arcobacter and Campylobacter species from selective enrichment cultures, although several other non-Campylobacter and non-Arcobacter species could be detected theoretically as part of the profile e.g., Clostridium species,

Sphingobacterium, Proteobacteria, Listeria monocytogenes, Abiotrophia, Lactobacilli, Corbacterium, Enterococcus, Mycoplasam, Acholeplasma laidlawii. It can be perceived that the presence of the TRF fragment is due to post-selective culture contamination, or due to the presence of species with VNC state resulting from the storage, culture or incubation conditions used for growing Campylobacter. Clostridium is an anaerobe, and is a ferment, which may mean that it grows in Prestons but very slowly. Alternatively they may be residual contaminates in the chicken samples. As small amounts of DNA can easily be amplified, it is possible that the TRF is represented to this species. There are no reports that describe the ability of these other species to grow under

Campylobacter or Arcobacter selective enrichment conditions. The T-RFLP assay described in this chapter is a continuation of the previous studies reported in the earlier chapter on the development of a rapid molecular assay for detecting and identifying Campylobacter and Arcobacter species. The T-RFLP technique is characterised by its ability to study single bacterial species as well as bacterial communities. Different molecular technique such as AFLP [104, 105], PCRRFLP [152], and PFGE [288] have been used to detect and identify Campylobacter species. These methods are very laborious, technically demanding and time-consuming.

160

Most of the studies conducted with T-RFLP so far involve the analysis of bacterial community diversity in different environments including activated sludge, soil and marine communities [182, 183], and bacteria from positive blood cultures [289], and bacterial communities from the termite hind-gut [146]. T-RFLP is a high-sensitive, costeffective, and high-throughput technique that could be used for rapid detection and screening of a bacterial community. The T-RFLP assay developed in this study could be used in the case of an epidemic or an outbreak of disease. It could also be the cornerstone for the development of other rapid detection methods for other bacterial species.

8.5. Conclusion In conclusion, the analysis of 16S rRNA genes by TRF profiling allows the identification of Campylobacter and Arcobacter process to be completed within 8 hours after enrichment cultures of the environmental or chicken sample or clinical specimens. Reduced diagnostic time has implications for outbreak containment and the duration of infection, affecting the cost of patient care, length of hospitalisation, development of broad-spectrum antibiotic resistance, and mortality due to the campylobacteriosis. TRF profiling represents a high-throughput and predictive source for identification of many organisms associated with bacteremia. In this chapter it was demonstrated that T-RFLP was used successfully to detect and identify Campylobacter and Arcobacter species. However, the assay was unable to distinguish between C. jejuni, C. coli, and C. lari. This assay was able to detect other species (C. sputrum, C. helviticus and C. fetus) that were present in the enrichment cultures initiated from chicken samples that were not detected by conventional culture method, and were not previously targeted by real-time PCR and LDR (chapters 3, 4, 5, 6, 7 and 9).

161

Table 8.2

Theoretical TRF size of the five restriction endonuclease enzymes used in this study, to digest 16S rRNA amplified using FAM-49Fand R2 primers, as described in Section 8.1.3, by TAP computer simulation and BioEdit.

Bacterial species

DdeI

BtsI

KpnI

EcoRI

A. nitrofigilis

53

312

260

A. skirrowi

53

312

260

A. cryaerophilius

53

312

260

A. butzleri

53

312

260

C. lari

245

415

C. Lanienae

245

413

C. jejuni

246

415

405

254

C. coli

246

415

405

254

C. hominis

92

412

254

412

254

405

Sau3AI

255 255

C. hyointestinalis subsp 50 hyointestinalis C. hyointestinalis subsp 137 lawsonii C. upsaliensis 106

412

254

415

254

C. fetus subsp fetus

50,

416

256

C. fetus subsp venerealis C. sputorum subsp sputorum C. sputorum subsp bubulus C. helviticus

247

416

256

104

411

256

162

336

176

106

415

254

413

253

412

252

414

254

309

E. coli C. showae C. concisus

243

C. mucosalis C. rectus

243

373

254

C. curvus

46

373

213

C. gracilis

245

414

254

162

9. Chapter 9 Detection and Identification of Campylobacter Species, Arcobacter, C. jejuni, C. coli and C. lari using LDR 9.1. Introduction LDR utilises the ability of thermostable DNA ligase to specifically link two adjacent probes when hybridised to a complementary target only when the nucleotides are perfectly base-paired at the nick junction [213, 290]. A single bp difference at the nick junction prevents ligation/amplification and is consequently differentiated from a perfect match (Figure 9.1). In this chapter, LDR technique is developed for the detection and identification of Campylobacter and Arcobacter from each other at one stage, as well as distinguishing C. coli, C. lari and C. jejuni from other Campylobacter and

Arcobacter species (Figure 2.1). 9.2. Material and Methods The general methods used in this study have been described in detail in Chapter 2. Specific methods relating to this chapter are described below.

9.2.1. Probe and Primer Design and Synthesis 16S rRNA sequences from Campylobacter species, four Arcobacter species and a few other species representing various phyla of the domain Bacteria available in the Ribosomal Database Project II Release 8.0 (http://rdp.cme.msu.edu/html/) were imported into BioEdit. The sequences were assembled and aligned using the Clustal W algorithm. Universal common probes were chosen in the 16S rRNA bacteria conserved region, which is downstream to the bacterial hyper variable region, from which the discriminative probes were designed. Specificity of each probe pair (universal common

163

Figure 9.1: Schematic representation of LDR assay used for rapid detection and identification of Campylobacter and Arcobacter species. The discriminative probe corresponding to the Campylobacter or Arcobacter species is shown in green bold and has a G at its 3´ end. The common probe is shown in black bold. In this example, the template has been amplified from Campylobacter or Arcobacter species. Thus only the Campylobacter and Arcobacter species discriminative probe is ligated to the common probe, generating a single type of ligation product. The thermal cycling is repeated twenty times in order to generate sufficient ligation product for detection.

164

probe and discriminating probe) was checked using BLAST. All probes selected had a Tm value between 55 and 72°C. Discriminating probes were labelled with a FAM molecule at their 5' end, while universal common probes were labelled with a phosphate in the same position (Geneworks, Australia). For discrimination of C. coli, C. lari and

C. jejuni, the FlaA gene was used to design a universal common and discriminative probe as described for 16S rRNA. FlaA forward and reverse primers were designed for synthesis following the guidelines described in section 2.4.1.

9.3. Results 9.3.1. Design of 16S rRNA and Universal Common and Discriminative Probes Forward and reverse 16S rRNA universal primers for the domain Bacteria, F2 (5’ CAGATTAGATACCCTGGTAG

3’,

E.

coli,

position

783)

and

Rd1

(5’

TACGGYTACCTTGTTACGAC 3’, E. coli, position 1542) were chosen and synthesised to amplify a 759 bp amplicon hybrid in the target regions of the LDR common and discriminative probes. Common and discriminative probes were designed to be complementary to one of the hyper variable regions identified by Busti et. al. (2002) [166]. This region was chosen to fulfil the LDR requirements, as shown in Figure 9.2. Discriminating probes were labelled with 6-FAM at the 5’ end with 3’ position unique to the Campylobacter and Arcobacter in the 16S rRNA. Two discriminative probes designated LDRCam and LDRArco were designed, LDRCam (5’ TGGAGCAAATCTATAAAATRYG3’,

Campylobacter

E.

species,

coli, and

position

1272)

LDRArco

specific

for (5’

TGACGTGGAGCRAATCTYAAAAATRYC3’, E. coli, position 1267) specific for

Arcobacter species. A common probe labelled with the phosphate group (PO4) at the 5’ end designated LDRProteo (5’TCCCAGTTCGGATTGKWSTCTGCAAC3’, E. coli, position 1295) was designed, downstream to the discriminative probe, specific for the 165

entire Proteobacteria groups including both genera Campylobacter and Arcobacter (Figure 9.2). FASTA analysis, which permits the verification of the probes against the GenBank bacterial sequence, which is not considered in the alignment, indicated that the common probe would bind to all Proteobacteria members due to the K, W and S bp that introduce degeneracy at positions 1311, 1312 and 1313. The LDRCam and LDRArco discriminative probes are only specific to Campylobacter and Arcobacter species respectively. The flaA gene was used in this study to detect, identify and distinguish C. coli, C. jejuni and C. lari. Thus, a forward primer named Fla-7 (5’ TTCGTATTAACACAAATGKT 3’) and reverse primer named Fla-650 (5’CWGATAARGCTATGGATGA3’) were designed and used to generate a 643 bp amplicon (Figure 9.3). Three different discriminative probes, C. jejuni specific discriminative probe designated LDRJejuni (5’STAAAAGTTTAGATGCTTCTTTAA3’, C jejuni position 64-87), C. coli specific discriminative probe designated LDRColi (5’AGAGCATTAGATAGCTCTTTAG3’, C

jejuni position 68-87) and C. lari specific discriminative probe named LDRLari (5’AGCATTAGATCAATCACTTT 3’, C jejuni position 66-87), were designed within the flaA gene hyper variable region that could be amplified with flaA gene primer pair. A

common

probe

designated

LDRFla

(5’

SMAGACTYAGTTCAGGTYTKAGAATHAAYT 3’, C jejuni position 88-117) designed downstream of the flaA gene discriminative probe to hybridise to C. coli, C.

lari, and C. jejuni, due to the S, M, Y, Y, K, H, and Y bp that introduces degeneracy at positions 66, 68, 73, 83, 85, 91 and 100 (Figure 9.2). Both flaA genes primers and LDR probes were designed based on the information available from the alignment of the flaA genes from C. jejuni, C. coli and C. lari.

166

167

E. coli str. K-12.---------------------CAGGATTAGATACCCTGGTAG-------GAAGCGGACCTCATAAAGTGCGTCGTAGTCCGGATTGGAGTCTGCAAC------AAGGAGGTGATCCAGCC-C. mucosalis CCUG6822----------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------C. hyointestinalis ATCC 35217--------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. lanienae--------------------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. gracilis ATCC33236----------------------------------------------TGGAGCAAATCTAAAAAATATGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------C. hominis---------------------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------C. fetus subsp. venerealis ATCC19438-------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------C. jejuni subsp. doylei CCUG24567----------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------C. sputorum subsp. sputorum ATCC35980------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. upsaliensis CCUG14913-------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------C. hyointestinalis subsp. Lawsonii---------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. coli ATCC33559-------------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. lari CCUG23947--------------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. concisus--------------------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. fetus subsp. fetus CIP5396--------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. helveticus NCTC 12470-------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------C. showae CCUG3054-------------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------C. jejuni subsp. jejuni ATCC33560----------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. rectus CCUG19168------------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------C. curvus ATCC35224------------------------------------------------TGGAGCAAATCTATAAAATACGTCCCAGTTCGGATTGGAGTCTGCAAC------------------------A. butzleri CCUG10373----------------------------------------------TGGAGCAAATCTAAAAAATACCTCCCAGTTCGGATTGTAGTCTGCAAC------------------------A. cryaerophilus ATCC49615-----------------------------------------TGGAGCGAATCTCAAAAATGCCTCCCAGTTCGGATTGTAGTCTGCAAC------------------------A. nitrofigilis CCUG15893------------------------------------------TGGAGC-AATCTCAAAAATGCCTCCCAGTTCGGATTGTAGTCTGCAAC------------------------A. skirrowii CCUG10374--------------------------------------------TGGAGCAAATCTTAAAAATATCTCCCAGTTCGGATTGTAGTCTGCAAC------------------------LDRCam Probe-------------------------------------------------------TGGAGCAAATCTATAAAATATGTCCCAGTTCGGATTGTTCTCTGCAAC------------------------LDRArco probe---------------------------------------------------------AGCRAATCTYAAAAATRYCTCCCAGTTCGGATTGTTCTCTGCAAC-------------------------

Schematic representation

Primer F2 783-804 5’ CAGGATTAGATACCCTGGTAG 3`

LDR CAM. Probe 5’FAM

1272-1294

C 3`

LDRProteo Common Probe

5’Po4

1295-1320

3`

Primer Rd1 1542-1526 5’ AAGGAGGTGATCCAGCC 3`

Figure 9.2 An alignment of the partial 16S rRNA sequences corresponding to position 783 to 1542 (E. coli numbering according to [291]) of representatives of Campylobacter and Arcobacter, this region is amplified using F2 16S rRNA forward primer and Rd1 16S rRNA revers primer (indicated by left- and right-hand arrows) to produce a 759 bp amplicon. The common probe LDRProteo (specific for members of Proteobacteria) and the LDR discriminative probes (LDRCam and LDRArco specific to Campylobacter and Arcobacter species respectively) bind adjacent to each other within the target regions of the amplicon.

168

C. jejuni FlaA gene----TTTCGTATTAACACAAATGT-------CTAAAAGTTTAGATGCTTCTTTAAGCAGACTTAGTTCAGGTCTTAGAATTAACT--------------GGAACAGGACTTGGAGCTTT-C. lari FlaA gene------TTCGTATTAACACAAATGGT---------AGAGCATTAGATAGCTCTTTAGCAAGACTTAGTTCAGGTTTGAGAATAAATT--------------GGAACAGGACTTGGAGCTTT-C. coli FlaA gene------TTCGTATTAACACAAATGTT-----------AGCATTAGATCAATCACTTTCAAGACTCAGTTCAGGTCTTAGAATCAACT--------------GGAACAGGTCTTGGAGCTTT-LDRJejuni------------------------------------------TAAAAGTTTAGATGCTTCTTTAA--LDRLari---------------------------------------------AGAGCATTAGATAGCTCTTTAG--LDRColi-----------------------------------------------AGCATTAGATCAATCACTTT--LDRFla common probe--------------- ---------------------------------------SMAGACTYAGTTCAGGTYTKAGAATHAAYT LDRFla-7 primer (7-26) 5’TTCGTATTAACACAAATGKT 3`

LDRFla Discriminative probe

LDR-FLA Common Probe

5’Po4-SMAGACTYAGTTCAGGTYTKAGAATHAAYT-3`

LDRFla-630 primer (630-649) 5’AAAGCTCCAAGTCCTGTTCC 3`

Figure 9.3 An alignment of the partial FlaA sequences corresponding to position 7 to 649 representative of Campylobacter species is shown. This region is amplified using LDRFla-7 primer and LDRFla-630 primer (indicated by left- and right-hand arrows) to produce a 643 bp amplicon. The common probe LDRFla (specific for C. coli, C. jejuni and C. lari) and the LDR discriminative probes (LDRJejuni, LDRColi and LDRLari specific for C. jejuni, C. coli and C. lari respectively) hybridise adjacent to each other within the target regions of the amplicon.

169

9.3.2. PCR Amplification and Ligase Detection Reaction In this project DNA was extracted (Section 2.2) from 258 samples of Campylobacter and Arcobacter pure culture and Campylobacter and Arcobacter enrichment cultures initiated from chicken samples (Section 2.1) (Appendices 1, 3, 5 and 6). As expected, an amplicon of approximately 759 bp was observed by agarose gel electrophoresis using the 16S rRNA F2 universal forward primer and 16S rRNA Rd1 universal reverse primer (Figure 9.4), while LDRFla-7 forward and LDRFla-650 reverse primers successfully generating a 643 bp amplicon, as shown in Figure .9.5.

Figure 9.4 Amplification of the 16S rRNA gene of Campylobacter and Arcobacter species by PCR with the primers F2 and Rd1. The PCR products were electrophoresed in 1% agarose gel. C. coli (NCTC 11366) (Lane 2), C. jejuni (ATCC 940565) (Lane 3), C. hyointestinalis (QHSS 99M2318) (Lane 4), A. butzleri (NCTC 12481) (Lane 5), shows an expected amplicon size of 759 bp. A ladder of 100 bp molecular markers is shown (Promega, Australia) (Lane 11).

170

Figure 9.5 Agarose gel electrophoresis of flaA gene PCR amplicons produced with LDRFla forward and LDRFla reverse primer. C. coli (NCTC 11366) (Lane 1), C. jejuni (ATCC 940565) (Lane 2), chicken isolate enriched in Campylobacter enrichment media (Lanes 3 and 4,) shows an expected amplicon size of 644 bp. No products are generated from C. upsaliensis (QHSS 99M126) consistent with expectations (Lane 5). A 100 bp DNA ladder is shown (Promega, Australia) (Lane 11).

9.3.3. Detection of LDR products using ABI377 (Applied BioSystems) LDR has been shown to be a powerful discriminative technique in the detection of cancer, disease gene mutation, as well as for the rapid detection and identification of bacterial species [222, 227]. In this study LDR assays have been used for rapid detection and identification of Campylobacter and Arcobacter from 258 pure and enrichment cultures initiated from chicken samples. LDR consists of two sequential steps, the first being the PCR, where the forward and reverse primer, which targets the 16S rRNA and flaA gene is used to amplify the 759 and 643 bp PCR amplicons, respectively (Section. 2.9.1). After that, the PCR amplicons are used in the ligase reaction, where the common and discriminative probes will hybridise to the target region within the 16S rRNA and flaA genes PCR amplicons. After that the ligation of the discriminative and common probe will occur only if there is a 100% match at the 3’ end of the discriminative probe. The LDR products generated from the hybridisation 171

and ligation of the LDR common and discriminative probe will be electrophoresed and the fluorescent product will be detected using the ABI370 DNA sequencer. As mentioned in section 8.3.4.2, the global southern analysis method was used to analyse the LDR data. From this analysis, LDR products between 35 to 500 bp long were accurately sized to + 2 bp, which is similar to the T-RFLP result (Section 8.3.4.2). Other studies have reported accuracy ranging from as little as 1 bp to as much as 7 bp [149, 193, 196]. Figures 2.1, 9.6 and 9.7 shows a 43, 46, 48 50 and 52 bp LDR product identifying

Arcobacter species, Campylobacter species, C. coli, C. lari, and C. jejuni respectively, that have been generated after the hybridisation and ligation of the of LDR probes, [(LDRCam, LDRArco, and LDRProteo) and (LDRJejuni, LDRColi, LDRLari and LDR-Fla)] to the target regions of the Campylobacter and Arcobacter 16S rRNA and

flaA genes PCR products. A single 43 bp LDR product identified the Arcobacter species that was generated from the hybridisation and ligation of the LDRProteo common probe and the LDRArco discriminative probe to the target region of the PCR products, this being amplified using the F2 and Rd1 16S rRNA universal primers of the DNA from

Arcobacter pure cultures (Panel 9.6A). Another single LDR product of 46 bp identifies Campylobacter species generated from the hybridisation and ligation of the LDRProteo and LDRCam to the Campylobacter species 16S rRNA gene (Panel 9.6B). A single 52 bp LDR product specifically identified C. jejuni (ATCC 940565) generated from the ligation of the LDRjejuni and LDR-FLA after their hybridisation to the target region of the flaA gene PCR amplicons (Panel 9.6C). On the other hand, no LDR products were generated from the E. coli 16S rRNA PCR products amplified with F2 and Rd1 (Panel 9.6G). The results obtained from the above data describe the sensitivity and discriminative ability of the LDR technique used to detect and identify Campylobacter species, Arcobacter species, C. jejuni, C. lari and C. coli. (The result for C. coli and C.

172

lari using flaA gene separately are not shown.) These results are in accordance with our expectation. Furthermore, the sensitivity and specificity of this assay was also confirmed for the DNA sample extracted from the Campylobacter and Arcobacter selective enrichment cultures incubated for 24 hours. After the sensitivity and specificity of the LDR assays had been proven, the next step was to confirm the ability of this assay to detect multiple samples (LDR multiplex). This was tested at the same time using the common and discriminative probe targeting the 16S rRNA and flaA gene. Two LDR products, one dominant over the other, represented the sizes of 46 and 48 bp (Panel 9.6D). The 46 bp LDR product identified the Campylobacter species, whereas the 48 bp LDR product specifically detected and identified C. coli. This panel verified the presence of C. coli as confirmed by both products, with the Campylobacter LDR peak being the dominant one, possibly due to the presence of more than one Campylobacter species in the sample. Three discriminative LDR products of sizes 43, 46, and 48 bp were generated from the hybridisation and ligation of the LDR discriminative and common probes to the target region within the flaA and 16S rRNA multiplex PCR products generated from

Campylobacter and Arcobacter enrichments (C-P10018 and P22-A) (Panel 9.6E). The 43 and 46 bp LDR products identify the 16 rRNA genes of Arcobacter and

Campylobacter species respectively, while the other LDR products 48, identify C. coli flaA gene.) (Panel 9.6F). Furthermore, such assays are ideal for multiplexing as shown in this chapter, since several primer sets can hybridise along a gene without the interference encountered in polymerase-based assays. Developing PCR multiplexing assays that produce equivalent amounts of each PCR product can be difficult, time consuming, expensive and laborious [229]. This is due to the difference in the annealing temperature of each set of the primers in the reaction. The primer concentration, template concentration, MgCl2 and 173

other salt concentration can affect Tm. Therefore, as the number of the PCR amplicons increase, it becomes more difficult to optimise the reaction conditions to obtain equal amounts of each amplicon. The success of the PCR / LDR typing assay make it very easy to perform multiplex detection or identification to species level in the case of bacteria identification, or detection of mutation in the case of single nucleotide polymorphisms. Detection and identification of the closely related species C. coli, C. jejuni and C. lari has been made. Discrimination between these three species is a very important task, and for this reason the LDR technique was tested and was shown to be effective and sensitive enough to discriminate the three species. LDR results show that it is very important to inhibit the Taq DNA polymerase activity and to purify the PCR product from primer dimers, non-specific products and Taq polymerase. If this was not the case then Taq polymerase activity will continue during ligation, as the ligation temperature is compatible with Taq polymerase activity. Consequently, larger than expected LDR products are produced. It has been found that incubation of PCR products in 1:10 volume of 1mg/ml of proteins K in 50 mM EDTA at 37 oC for 30 min and the purification of PCR products using spin columns are required to inhibit Taq DNA polymerase activity (Panel 9.6F).

174

43

A 46

B 52

43

35

C

46 48

D

46 48

E

Multi peaks

F

Negative control

G

Internal standard

H

50

75

Figure 9.6 A typical electropherogram of LDR patterns generated from the hybridisation and ligation of LDR probes to target regions of the amplicons of 16S rRNA genes (LDRCam, LDRArco, LDRProteo probe) for identifying Campylobacter and Arcobacter and target regions of the amplicons of FlaA genes (LDRJejuni, LDRColi, LDRLari and LDR-Fla) for identifying C. lari, C. coli and C. jejuni from the DNA of pure and enrichment cultures. When the 16S rRNA discrimination hybridisation and ligation system is used, a 43 bp identifies a pure culture of Arcobacter species (Panel 9.6A), while a 46 bp identifies a pure culture of Campylobacter species (Panel 9.6B). When the flaA gene discrimination hybridisation and ligation system is used, a 52 bp LDR product generated with the DNA of C. jejuni pure culture (panel 9.6C). The combined use of both 16S rRNA and flaA genes from pure and enrichments cultures generated 5 LDR products able to identify Campylobacter and Arcobacter species (Panels 9.6D and 9.6E). The 46 and 48 bp LDR products identify Campylobacter species using 16S rRNA and C. coli using flaA gene respectively (Panel 9.6D). A profile of three LDR products representing the sizes of 43, 46, and 48 bp generated from the hybridisation and ligation of the LDR probe to the multiplex PCR amplicons of 16S rRNA and flaA gene of DNA from Campylobacter and Arcobacter enrichments (C-P1019 and P22-A) is shown (Panel 9.6E). The 43 and 46 bp identifies Arcobacter and Campylobacter 16S rRNA gene respectively, whereas the 48 specifically identifies the C. jejuni flaA gene. Multiple LDR products generated from unpurified 16S rRNA PCR products of C. jejuni is shown (Panel 9.6F). A negative control shows no LDR products generated from E. coli (Panel 9.6G). Panel H shows the internal GeneScan standard. This figure has been redrawn to scale from the row data in Appendix 13.

175

LDR Campylobacter and Arcobacter species 16S rRNA gene

LDRArco probe

LDRCam Probe Campylobacter species (46 bp LDR products)

Arcobacter species (43 bp LDR products)

FlaA gene

LDRColi probe

C. coli (48 bp LDR products)

LDRLari probe

LDRJejuni probe

C. lari (50 bp LDR products)

C. jejuni (52 bp LDR products)

Figure 9.7 A schematic representation of the LDR assay, showing the identification of Campylobacter species and Arcobacter species using 16 S rRNA and detection of C. lari, C. coli and C. jejuni using FlaA gene based on the LDR product size of each genera or species.

176

9.3.4. Identification of Campylobacter and Arcobacter Isolates in the Culture Collection and Enrichment Cultures using the LDR Assay This assay was successfully tested and identified the 198 isolates of animal, human, plant and bird origin from our culture collection, into C. coli (76 isolates), C. jejuni (88 isolates), C. lari (1) other Campylobacter species (12 isolates), and Arcobacter species (22 isolates) (Appendix 1 and appendix 3). In addition, 27 out of 60 enrichment cultures initiated from chicken samples were tested and identified as Campylobacter species, 18 as C. jejuni, 11 as C. coli, 2 as C. lari and 22 as Arcobacter species (Appendices 14 and 15).

9.3.5. Comparison between Conventional Culture Methods and LDR Assay LDR sensitivity, specificity and accuracy are in accordance with the other real-time PCR and T-RFLP assays (chapters 3 to 8) and culture methods. However, LDR was able to detect C. lari from one enrichment culture. This enrichment culture showed negative results by culture methods and C. lari was not targeted by other molecular methods, as it is responsible for a very small proportions of human infection associated with Campylobacter species.

9.4. General Discussion Real-time PCR has been used successfully in all studies reported in this thesis for the rapid detection and identification of Campylobacter and Arcobacter species. The realtime PCR assay is able to detect up to four fluoroprobes when used in a real-time PCR systems such as iCycler IQ. This means that only four targeted genes are possible making this a disadvantage of real-time PCR techniques. LDR on the other hand, has been used in this study to detect five probes targeting 3 different species and 2 genera. It has been reported that LDR is able to detect and distinguish 19 possible mutations in the human K-ras gene [166]. 177

In this chapter, the LDR method was successfully developed to detect and identify both

Campylobacter and Arcobacter to the species level. The 16S rRNA bacterial domain universal primers F2 and Rd1 were used to amplify a 759 bp fragment, hybridise the target region of the LDR common and discriminative probes, and distinguish

Campylobacter and Arcobacter species based on the LDR products size of 45 and 43 bp respectively (Figures 9.6 and 9.7). The DNA region coding for flaA gene was amplified using LDRFla-7 forward primer, and LDRFla-630 reverse primer, generating a 643 bp amplicon, that hybridised to the target region of the LDR common and discriminative probes, allowing the discrimination of C. coli, C. jejuni and C. lari by producing 48, 50 and 52 bp LDR products respectively (Figures 9.6 and 9.7). These results concur with our expectation based on the alignment, and BLAST and FASTA analysis of the 16S rRNA and flaA gene. Generations of LDR products in this assay indicate the presence of Campylobacter and

Arcobacter species, while the length of an LDR product identifies which genus or species is present. When one set of probes and primers target one species or bacterial group using the LDR assay, an excess of LDR probes over template ensures efficient hybridisation and ligation (Panel 9.6A, B and C). Many papers describing the ability of LDR techniques for rapid detection and identification of bacterial species, mutations and viruses have been published. LDR have been used extensively to distinguish 19 possible mutations in K-ras [166], over 30 mutations in the cystic fibrosis transmembrane conductance regulator gene [292], the HER-2/neu gene amplification in multiple human breast cancer specimens [227], and bacterial discrimination [43]. This is the first time that LDR has been used to detect

Campylobacter and Arcobacter species, and to distinguish C. coli, C. jejuni and C. lari in a one-step, one-tube reaction. LDR reactions are compatible with a universal DNA array [182, 183, 191, 194, 293-297], and DNA sequencer [43] detection schemes.

178

The LDR assay described in this chapter represents a highly reproducible, sensitive and accurate assay for the rapid detection and identification of Campylobacter and

Arcobacter species from pure and enrichment cultures. Several features unique to LDR may account for its ability to detect bacterial species. In contrast to other quantitative real-time and conventional PCR assays, LDR uses an internal standard that precisely determines the size of the bacterial target sequence fragment with the exception of a two–bp substitution. The close identity of the detected LDR products ensures the efficiency of the PCR amplification and the hybridisation and ligation process between common and discriminative probes to their targets. The other advantage of LDR is based on single-strand DNA recognition, which avoids potential quantitative inaccuracies associated with heteroduplex formation. These factors combine to enable the rapid and accurate detection and specific identification of the Campylobacter and

Arcobacter species by targeting the 16S rRNA, as well as to distinguish C. coli, C. jejuni and C. lari simultaneously in multiple specimens. 9.5. Conclusion LDR has been used in this study for the rapid detection and identification of food-borne pathogens, Campylobacter and Arcobacter. The methods also differentiate C. coli, C.

jejuni, and C. lari, which are recognised as the most common Campylobacter species associated with human infections. This method has been used successfully to detect their presence in both enrichment cultures and BAP cultures, providing a rapid, specific, and sensitive detection method. LDR is able to identify a broad range of closely related species, though in this study only three species and two genera were targeted because they are the most common pathogens. The rapid boiling DNA extraction method described in this thesis proved to be a very efficient method for the extraction of

Campylobacter DNA from enrichment cultures, as well as from BAP cultures.

179

10. Chapter 10 Conclusions and Future Directions The genus Campylobacter, family Campylobacteraceae, class Epsilonproteobacteria of phylum Proteobacteria currently consists of sixteen species and eight subspecies. All are natural inhabitants of the intestinal tracts of poultry and warm-blooded domestic animals where microaerophilic conditions and the warm body temperature is an ideal environment for their continuous growth. The consumption of contaminated food and water of some of these species causes gastrointestinal illness in humans. Most reported Campylobacter related human illnesses are caused by C. jejuni and to a lesser extent by other Campylobacter such as C. coli, C. lari, C. hyointestinalis, C. upsaliensis and C. fetus. Arcobacter species are also found to be associated with human infection. It has been estimated that around 680 to 730 persons die due to Campylobacter infections each year in the United States. A number of methods reliant on real-time PCR, T-RFLP and LDR have been developed in this project for the rapid detection and identification of Campylobacter and Arcobacter species (Figure 2.1): (A) Two tube real-time PCR assays in which an adjacent hybridisation probes targeting the 16S rRNA gene, and SYBR Green I targeting the hipO gene were developed for the rapid detection and differentiation of C. coli and C. jejuni. (B) A two-probe method was designed for the specific detection of A. butzleri and for the detection of other Arcobacter species (A. nitrofigilis and A. skirrowii). Further species differentiation was achieved using Tm differences between C. coli, C. jejuni, A. butzleri and A. skirrowii. (C) Multi-FAM one tube assay for the rapid detection and identification of C. jejuni, C. coli, A. butzleri and A. skirrowii was developed.

180

(D) Real-time PCR was used to develop a high throughput, one tube rapid detection and differentiation multiplex method for C. coli and C. jejuni, allowing differentiation from each other as well as from the other Campylobacter species. (E) Concurrent with the multiplex system, a T-RFLP typing method was developed. This method was used to detect Campylobacter and Arcobacter species directly from food samples and would not be reliant on strain isolation, as is the case with real-time PCR methods A-D, described above. (F) LDR was used for rapid discrimination between Campylobacter and Arcobacter species. LDR is a high-throughput, specific and accurate technique that is based on the Taq DNA ligase activity, which allows a one bp discrimination. The two-tube real-time PCR for rapid detection and identification of C. coli and C. jejuni from other bacteria and Campylobacter species is a very simple, accurate and robust method, which employs adjacent hybridisation probes and SYBR Green I. The adjacent hybridisation probes target the 16S rRNA gene and is used to differentiate C. coli and C. jejuni from other Campylobacter species. SYBR Green I is used in a different tube to differentiate C. jejuni from C. coli and targets the hipO gene, which is only present in C. jejuni. The multiFam real-time assay employed adjacent hybridisation probes described in Chapter 5 and Chapter 6 for the rapid detection and identification of Campylobacter and Arcobacter species based on their Tm differences. In Chapter 5, two different adjacent probes (probe Butz, specific for A. butzleri and probe Skir-Cry specific for A. skirrowii and A. cryaerophilius), were used to differentiate three Arcobacter species based on their Tm differences where a Tm of 67oC due to dissociation of probe Butz confirmed the presence of A. butzleri and a Tm of 63oC and 65oC due to the dissociation of probe SkirCry differentiated A. skirrowii from A. nitrofigilis respectively. The method successfully identified all 22 Arcobacter cultures isolated from humans and birds. This assay was 181

mainly used to distinguish between A. butzleri and other Arcobacter species. A one bp difference was used to detect and distinguish A. nitrofigilis (not associated with human

infections) from A. skirrowii and A. cryaerophilius. In future, a 2 or more bp differences in probe or primer design could be developed for species or subspecies discrimination. Alternatively, a second probe targeting a different region could be investigated for differentiation between A. skirrowii and A. cryaerophilius.

In Chapter 6, a multi FAM-real-time assay were used to differentiate A. butzleri, and A. skirrowii from C. coli, C. jejuni, C. lari, C. hyoilei, C. helviticus, C. hyointestinalis, C. insulaenigrae, C lanienae based on their Tm (probe Butz, probe Skir-Cry), and a C. coli / C. jejuni specific probe (5’-GTGCTAGCTTGCTAGAACTTAGAGA-FAM-3’). Three different Tm’s had been generated at the end of the real-time PCR. These three Tm’s namely 67oC, 63oC and 65oC confirmed the presence of A. butzleri, A. skirrowii and C. coli / C. jejuni respectively. This technique is accurate, robust and cost-effective, but it is very difficult to design more than one probe in the chosen area of the 16S rRNA that has the ability to differentiate between closely related species, as in the case of Arcobacter and Campylobacter species. Multiplex real-time PCR is one of the most important developments in the real-time PCR approach, where more than one fluoroprobe is specific for certain species / genus. In Chapter 7, multiplex real-time PCR has been used for rapid detection and identification of C. coli and C. jejuni from other Campylobacter species, employing a TaqMan probe targeting three different genes, orfA, hipO, and 16S rRNA respectively. This method is accurate, automated and robust like all other real-time PCR. In addition, it is more cost-effective compared to the two-tube real-time assay, which is also a time consuming assay. Real-time PCR is a simple, accurate, robust, cost-effective and easy-to-perform diagnostic method that could be used for rapid detection and discrimination of

182

Campylobacter species, compared to the other conventional culturing techniques, biochemical tests and conventional PCR methods, which are less accurate, more laborious and time consuming. Due to the advances in molecular techniques, different methods have been developed recently or improved in the last five years. T-RFLP as mentioned previously is a modification of RFLP, where one of the PCR primers is labelled with a fluorescent dye such as FAM, Cy5, and HEX. T-RFLP is normally used for community screening of bacterial species from different environmental samples such as soil, water, and sludge [184, 186, 193, 294-297]. In this project, T-RFLP was developed to be used for screening, and rapid detection and discrimination of Campylobacter and Arcobacter species. T-RFLP is a powerful, high-throughput technique that could be used for screening and detection of Campylobacter and Arcobacter in the case of an epidemic or an outbreak of disease. T-RFLP methods described in chapter 8 were able to detect C. sputrum and/or C. helviticus and/or C. upsaliensis, C. hyointestinalis and/or C. fetus, which have not been detected by culture methods and have not been targeted in our study as they are responsible for a very low percentage of Campylobacter human infections.

However, other food-borne pathogens (Clostridium species and/or Mycoplasma species, and/or Lactobacillus species, and /or Eubacterium species) associated with human infections have been detected by T-RFLP which could be due to post amplification contamination or due to these species having a VNC status. T-RFLP assay was not able to distinguish between the closely related species C. lari C.

coli and C. jejuni. This problem could be solved by targeting other genes such as hipO and orfA genes, which could be the basis for future projects. Further distinction between the four Arcobacter species could also be achieved by targeting different genes in future. T-RFLP described in chapter 8 could be modified for the development of other assays for the detection of closely related genera, or targeting pathogens that infect

183

certain human systems such as the respiratory or gastrointestinal tract. Like other

techniques described, the disadvantages of this method are that it is time consuming, laborious and requires a highly qualified scientist. Although this method is time consuming it could be a method of choice for community screening because of its highthroughput and sensitivity. While T-RFLP is suitable for community screening, we investigated other methods that could offer greater sensitivity in identifying species. An example of such a method is LDR. LDR is a relatively recent technique. In this project, LDR has been used for rapid detection and identification of Campylobacter and Arcobacter species. The ability to perform multiplex PCR followed by multiplex LDR that could be used to detect and differentiate more than one target at the same time, is one of the most important features of this technique. In this project, five different target probes have been used to differentiate Campylobacter and Arcobacter species at the same time. LDR is the only technique that could detect and identify C. lari from the other closely related species, C. coli and C. jejuni. LDR has a higher discrimination power compared to T-RFLP, because the primer that is used for T-RFLP targets various phyla of the domain Bacteria while the LDR discriminating probe targets a specific sequence. In the future LDR techniques described in this thesis could be the corner-stone for developing microarray biochips that could be used in the diagnosis of Campylobacter and Arcobacter as well as other bacterial species. Real-time PCR is reported to be a simple, accurate, robust, cost-effective and easy-toperform diagnostic method that could be used for rapid detection and discrimination of Campylobacter species, when compared with culturing techniques, biochemical tests and conventional PCR methods, which are in general, less accurate, more laborious and time-consuming [126, 129, 298]. The use of the real-time PCR assays described in this thesis complement the reports of other authors. The real-time PCR assay described in

184

chapter 3 allows the detection of C. coli, C lari and C. jejuni, which are responsible for more than 95% of Campylobacter associated human infection. The combination of the assays described in chapters 3 and 4 not only detect the mentioned species but are also able to identify C. jejuni, a species which is responsible for 80-90% of Campylobacter infections. Although these methods are very sensitive and applicable for the detection of Campylobacter species, they lack the ability to differentiate between C. coli and C. jejuni, which is important for rapid epidemiological studies. For example, it is essential that C. coli is detected quickly as some strains are resistant to erythromycin and ciprofloxacin, antibiotics routinely used for treating Campylobacter infections. For this reason an easy to use, rapid triplex real-time PCR assay has been developed and described in chapter 7. Detection and differentiation of Arcobacter species could be also achieved using real-time PCR as described in chapter 5, but such a method has not yet been reported in the literature. The use of a multi-FAM assay to distinguish between Arcobacter species and Campylobacter species has been described in chapter 6. All the real-time PCR assays described in this thesis can be adapted for use in pathological laboratories having high sample loads for processing and having access to real-time PCR infrastructures. However, triplex assays require more recently developed real-time PCR instruments with the ability to read more than one fluorophore at the same time. Furthermore, T-RFLP is a powerful typing method able to distinguish between Arcobacter and Campylobacter at genera level, as well as distinguishing different Campylobacter species into different groups. LDR has the highest discrimination level compared to the other molecular methods used in this thesis, as it is able to distinguish between the Arcobacter and Campylobacter at genera level, and distinguish between C. coli, C lari, C. jejuni and other Campylobacter species. The last two methods could be used during disease outbreaks, as they are able to distinguish between the closely

185

related species and determine the species or group of species responsible for the outbreak, which could help in the prevention and treatment procedures. This project has shown that real-time PCR is a viable method to develop different diagnostic tests that could be used to detect and discriminate Campylobacter and Arcobacter species from BAP cultures (clinical, animal and environmental samples) and enrichment cultures. This method is easy to use and can be performed in any pathological laboratory that has a real-time PCR machine. T-RFLP was used in this project for detection and identification of Campylobacter and Arcobacter species, but this technique is still very expensive and requires a highly qualified scientist, which makes it more difficult to be used routinely in a pathology laboratory. LDR has the highest discrimination power of the molecular methods used in this study, as three species and two genera can be differentiated at the same time. On the other hand, if the LDR assay for Campylobacter and Arcobacter can be developed to be employed into biochips then it could be used in pathological laboratories routinely. Its use otherwise would be only by reference and research laboratories, which is important in the case of epidemics. Furthermore, these methods could be further developed for the detection of a wide range of pathogens targeting different groups of bacteria (pathogens causing respiratory tract infection, or gastrointestinal tract infection, sexual transmitted diseases, etc.). Comparative sequence analysis of the available 16SrRNA genes of Campylobacter and Arcobacter species is in keeping with general observations that the information content of these genes are suitable for establishing phylogenetic relationship to subspecies level. The results data of Chapter 7 and 9 demonstrated the importance of availability of sequence data in designing primers for the identification of closely related species. The presence of flaA, hipO and orfA genes were used to design probes and primers for the detection and identification of C. coli, C. jejuni and C. lari, which are usually

186

misidentified as C. jejuni. These species are difficult to discriminate using 16SrRNA genes as the target regions suitable for identification are highly homologous. In our study, the specificity of the methods developed was initially inferred theoretically from the sequence database and other bioinformatic tools available online. Subsequently a number of closely related species to Campylobacter and Arcobacter were tested experimentally, providing the proof of assay specificity. However, in order that these methods are used as standard methods, evaluating the sensitivity with other closely related species especially gastrointestinal bacteria would be of great benefit. In conclusion, the true significance of infections of Campylobacter and Arcobacter species is yet to be determined. However, from the available data it seems that in addition to the human suffering, the social and financial cost could be quite large. Therefore, development of methods to better and more accurately identify infection with C. coli, C. jejuni, and all other bacterial agents is important and this project is a step forward in this direction. 10.1.

The Use of Molecular Methods for the Detection of Food-

borne Pathogens Recently developed molecular methods such as real time PCR have been extensively used in the last 5 years for the rapid detection and identification of bacterial species. In 2003, after the approval of the research project, FOOD-PCR (http://www.PCR.dk) by the European Commission, an international expert group of the European Committee for Standardization was established to describe protocols that were aimed at validating and standardising the use of diagnostic PCR for the detection of pathogenic bacteria in foods [270, 299]. The principle driving force was that a standardised PCR-based method should accomplish various criteria such as analytical and diagnostic accuracy, high sensitivity and specificity, robotic ability (including an internal amplification control [IAC]), reduction of carry over contamination, and be an easy-to-perform by user187

friendly protocols for its application and interpretation [299]. The development of the real-time PCR has the potential to meet all of these criteria by combining amplification and detection in a one-step closed-tube reaction. Different reports have been published describing the use of real-time PCR for the detection of Clostridium, Salmonella, E. coli and various other food-borne pathogens [270, 300-305]. Many other molecular techniques such as the PCR-RFLP assay for the rapid identification of Listeria species, E. coli, and Clostridium species [306-308] and DGGE for the detection of E. coli, Clostridium, and Campylobacter [107, 173, 309] have also been developed and used to identify food-borne pathogens. Recently commercial diagnostic and biotechnology companies have been involved in a race for developing molecular diagnostic kits that could be used for the detection of pathogens as well as autoimmune diseases and genetic disorders. For example, Roche Applied Science developed different real-time PCR kits as shown below: LightCycler foodproof Salmonella Detection Kit. The LightCycler Listeria monocytogenes Detection Kit. The LightCycler foodproof E. coli O157 Detection Kit. The LightCycler GMO Maize Quantification Kit In a bid to develop more robust and sensitive real-time PCR kits, commercial companies have improved the fluorophores (FAM, HEX, TET, Cy5, MAX, etc.) for probe labelling, and have developed different probe assays (Adjacent hybridisation probe, TaqMan, TaqMan MGB probe, Molecular beacon, etc.) and developed better instrumentation. Examples of improved instrumentation include detection in less than one hour by the LightCyclerTM (Roche, USA). As described in chapter 3, an assay was developed for detection of C. jejuni within 22 minutes. Increased sample handling was introduced in the LightCycler (Roche, USA) from 24 samples to 384 samples. The ability for detecting multiple fluorophores in a single reaction tube was introduced in 188

2000 and the iCycler IQTM is able to read four fluorophores whereas five fluorophores can be read in the ABI 7500® System (ABI, USA). 10.2.

Advantages and Disadvantages of Molecular Methods

Compared to Culture Methods Several advantages have been given for molecular methods described in this study, including high sensitivity for identification and discrimination of Campylobacter and Arcobacter species as well as the ability to distinguish C. jejuni from the closely related C. coli and from all other Campylobacter species. Molecular methods used in our study show 100% sensitivity and specificity when compared to each other. The molecular methods described in this study are rapid compared to the conventional methods. Real-time PCR can be performed within a very short time of around 9 h for Campylobacter and less than 24 h for Arcobacter species which includes the enrichment step. LDR and T-RFLP need around 36 h which also includes an enrichment step for Arcobacter and less than 24 h for Campylobacter. This compares with up to 5 days for culture methods (Gold standard methods) [85, 87]. The methods described here reduce the time for detection leading to prediction of contaminated food products before being released to the consumer. Such rapid methods could be vital in setting up diagnosis and treatment for food–borne or water-borne disease outbreaks. Therefore, these methods are useful in assisting the food industry to satisfy government regulations regarding the presence of harmful bacteria in food products. Application of quantitative molecular methods, coupled with full or partial automation, to the screening of large number of samples could prove advantageous for a food processor concerned with timely release of safe products. One factor that would greatly facilitate these molecular methods gain widespread acceptance is developing a standard or a range of standard methods for the DNA extraction directly from human samples such as stools, blood and tissue, or

189

directly from the food sources (e.g. chicken, red meat, etc.) without the need for the enrichment cultures. Detection of more than one species by conventional culture methods is very limited, as generally only the dominant species will be detected. On the other hand, molecular methods are able to detect all the targeted pathogens in a sample, and have the ability to detect very low number of CFU. (In this study as low as 110-125 CFU were detected.) One of the most important achievements of molecular methods is the detection of the VNC Campylobacter and Arcobacter. VNC is prevalent in both Campylobacter and Arcobacter as they transfer to the coccoid VNC state under stress conditions such as conditions used for food storage. The presence of VNC Campylobacter and Arcobacter may be the cause for infection after the consumption of tainted food and their subsequent recovery in human or other host. Another advantage of the molecular methods described in this study is the ability to provide automated and specific detection of the Campylobacter and Arcobacter species. Combining automated molecular methods with DNA extraction protocols can prove to be amenable to complete automation and lead to high throughput screening of food and pathological samples, fulfilling the requirements for the standardised methods determined by the European Commission expert group described earlier. However, one of the disadvantages of the molecular methods is the cross-reaction with closely related 16S rRNA genes from other bacteria. This problem could be solved by targeting more than one gene and avoiding the design of primers and probes in the conserved regions of the genes, thereby increasing the specificity of the reaction. Molecular methods by their nature are relatively expensive, but these costs could be balanced by the fact that these methods are timesaving due to decreased turnaround times. Hence, laboratory personnel would be free to perform other tasks, leading to an increase in productivity and provide general economic benefit. Furthermore, the faster 190

the pathogens are detected the sooner the treatment regime can be initiated reducing hospitalisation times for human infection. A decrease in the spread of food-borne infection necessarily decreases the use of antibiotics and therefore the issue of antibiotic resistance would improve. It has been estimated than more than 81 million persons are infected with food-borne pathogens each year, with an associated cost of 8 to 10 billion dollars (US) [310].

191