Investigation of the Role of Treponema Denticola Motility and Uncharacterized Protein TDE0659 in Synergistic Biofilm Formation with Porphyromonas Gingivalis

Investigation of the role of Treponema denticola motility and uncharacterized protein TDE0659 in synergistic biofilm formation with Porphyromonas gingivalis

Hong Min Ng ORCID 0000-0002-4957-1424

Doctor of Philosophy

May 2018

Oral Health CRC Melbourne Dental School Faculty of Medicine, Dentistry and Health Sciences The University of Melbourne

A thesis submitted in total fulfilment of the degree

ABSTRACT

Chronic periodontitis has a polymicrobial biofilm etiology and interactions between key oral bacterial species such as Porphyromonas gingivalis, Treponema denticola and Tannerella forsythia contribute to disease progression. P. gingivalis and T. denticola both have a set of virulence factors that are believed to contribute to the initiation and development of disease. It has been previously shown that P. gingivalis and T. denticola exhibit strong synergy in growth, biofilm formation and virulence in an animal model of disease. The motility of T. denticola, although not considered as a classic virulence factor, is likely to be involved in synergistic biofilm development between P. gingivalis and T. denticola. In order to investigate this, an optimized protocol for the transformation of T. denticola was developed and used to produce a number of T. denticola mutants targeting the motility machinery. The resulting mutants lacked periplasmic flagella (∆flgE) or possessed periplasmic flagella that were either non-functional (∆motA and ∆motB) or non-regulatable (∆cheY). ∆cheY contained a large genomic excision and was thus omitted from the study, however analyses of ∆flgE, ∆motA and ∆motB showed that they were impaired in motility and growth. Quantitative proteomic analyses of mutant strains showed that the inactivation of these motility-associated genes, especially motA and motB, have far reaching effects beyond motility. The inactivation of motA and motB activated a cellular stress response in the mutants and directly or indirectly impacted the growth of the mutants through the change in abundance of a number of proteins. T. denticola motility mutant and WT strains were grown as mono- and dual-species biofilms with P. gingivalis. Results showed that T. denticola motility and/or spiral morphology are required for monospecies biofilm formation and T. denticola periplasmic flagella are essential for synergistic biofilm formation with P. gingivalis.

Zones of clearing were observed between T. denticola sibling colonies grown close to one another on agar, similar to the sibling killing phenomenon observed in Paenibacillus dendritiformis. Given that a sibling killing phenomenon could have a considerable effect on biofilm formation, this phenomenon was investigated further in T. denticola. Although T. denticola was found to possess a homologue (TDE0659) of the dendritiformis sibling bacteriocin (DfsB) produced by P. dendritiformis, the protein did not undertake the same function in T. denticola. Construction and analysis of a T. denticola mutant strain lacking TDE0659 showed that the loss of this gene product from T. denticola did not prevent

adjacent sibling colonies from forming zones of clearing on agar, further the mutant was able to form monospecies biofilms similar to WT. However, the lack of TDE0659 prevented the incorporation of P. gingivalis in a dual-species static biofilm with T. denticola, indicating that TDE0659 is essential for promoting synergistic biofilm formation with P. gingivalis. Together, the findings from this study will further our understanding about how P. gingivalis and T. denticola interact with one another and proliferate during disease.

DECLARATION

This is to certify that:

i. the thesis comprises only my original work towards the PhD except where indicated in the Preface, ii. due acknowledgement has been made in the text to all other material used, iii. the thesis is fewer than 100,000 words in length, exclusive of figures, tables, maps, bibliographies and appendices.

Hong Min Ng

iii

PREFACE

Several members of the Oral Health CRC, Melbourne Dental School and Bio21 Institute, The University of Melbourne contributed to the work contained within this thesis. Cloning and propagation of plasmid pCF382, a kind gift from Dr Christopher Fenno (School of Dentistry, University of Michigan), in E. coli hosts were carried out by Dr

Kheng Hui Tan. Genome sequencing of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520 using the Ion Torrent Personal Genome Machine (PGM) was performed by Ms. Brigitte Hoffmann and the resulting sequencing reads were analyzed by Dr Catherine Butler (Section 2.12.1). Samples of T. denticola ATCC 35405, ATCC 33520,

HL51, ∆flgE33520, ∆motA33520 and ∆motB33520 were prepared for cryo-EM and imaged by Dr Yu-Yen Chen (Section 2.12.4). Label-free proteomics by MaxQuant was carried out by Dr Paul Veith (Section 2.12.8). Statistical analysis of the data from the static biofilm assays of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520 using Kruskal-Wallis with Conover-Imam test was performed by Mr. Yong Kai Wong from RMIT University (Section 2.12.9). Real-time PCR for the enumeration of bacterial cells was conducted with the help of Ms. Lin Xin Kin (Section 2.12.10). Flow cell biofilm assays, imaging of biofilms and analysis of the confocal data sets with COMSTAT software were performed by Ms. Sze Wei Liu (Section 2.12.11 and Section 2.12.12). T. denticola HL51 was obtained from Dr Howard Kuramitsu. Part of the Introduction (Chapter 1) to this thesis and Figure 1.3 have been published as the review article, “Ng, H. M., Kin, L. X., Dashper, S. G., Slakeski, N., Butler, C. A. and Reynolds, E. C. (2016) 'Bacterial interactions in pathogenic subgingival plaque', Microbial Pathogenesis, 94, 60- 69”. The work contained within this thesis was supported by the NHMRC, the Oral Health CRC (established and supported under the Australian Government’s Cooperative Research Centers Program) and Melbourne Dental School, The University of Melbourne.

ACKNOWLEDGEMENTS

First and foremost, I would like to thank God for giving me the opportunity to undertake this PhD. I thank God for His abundant grace and mercy to me throughout this program. I acknowledged that without God, it is impossible for me to complete this PhD, for it is indeed a marathon that needs much perseverance and patience. God has sustained, encouraged and comforted me when all things seem to be not working. He has granted me peace that surpasses all understanding when I was worried and anxious. He has always provided me with a way out at His perfect timing and helped me to truly experience His grace and power. God is indeed the One who is in control of everything and “with God all things are possible” (Matthew 19:26, King James Version). I am grateful for the provisions of God over the past few years and I have lacked nothing under the care of my God Almighty. He has provided me with the financial support from the Research Training Program Scholarship (The University of Melbourne) and Oral Health CRC stipend top- up scholarship to help me with my tuition fees and allowance. God has also blessed me abundantly by enabling me to attend several local and international conferences with the financial support of both the Oral Health CRC and Melbourne Dental School.

I thank God for the many people that He has sent to help me throughout this PhD program. My appreciation goes to my supervisors Dr Nada Slakeski, Dr Catherine Butler and Professor Stuart Dashper. They have supported and guided me through my PhD journey with much patience and kindness. Their commitment, knowledge and enthusiasm for research are much valued. Their willingness to spend time to discuss with me about my project and their efforts in guiding me to prepare the thesis are much appreciated. I would also like to thank Melbourne Laureate Professor Eric Reynolds AO FICD FTSE FRACDS for his support in this program.

I am grateful for the kindness, support and advice from the fellow staff and students of the Oral Health CRC and Melbourne Dental School over the past four years. I would like to thank Ms. Caroline Moore, Mr. Steven Cleal, Ms. Brigitte Hoffmann, Ms. Deanne Catmull, Ms. Yan Tan and Dr. Kheng Hui Tan who had shared their expertise and knowledge in the laboratory work. Particular thanks go to Ms. Sze Wei Liu who has provided me with much help in the experiments especially near the end of the PhD. I am appreciative of the commitments of Dr Yu-Yen Chen, who has spent much time and effort, in obtaining the cryo-EM images. I am thankful also for the kindness and help from Dr.

Paul Veith and Ms. Dina Chen in mass spectrometry data acquisition and analysis and Dr Lianyi Zhang in helping me to get access to the spectrofluorometer in Bio21 Institute. Not forgetting Ms. Lin Xin Kin who is always ready and willing to lend a hand and whose friendship has made working in the lab a pleasure.

Finally, I would like to thank God for the support and care from my family and relatives to undertake this PhD. A special appreciation goes to my housemate, friend and sister-in- Christ, Ms. May Ju Gui, who has undertaken her PhD together with me. I am grateful for her unceasing encouragement and support over the past four years. I thank God for sending me this faithful companion who has walked side by side with me, picked me up when I fell and shared the highs and lows of this PhD journey. Her company and sharing have made this PhD journey an enjoyable one.

Most importantly, I thank God for drawing me to the saving knowledge of my Lord and Saviour Jesus Christ during the course of this PhD. His saving grace is the greatest gift that I have obtained throughout this course.

Table of Contents Abstract…………………………………………………………………………………..i

Declaration……………………………………………………………………………...iii

Preface…………………………………………………………………………………..iv

Acknowledgements……………………………………………………………………...v

Table of contents……………………………………………………………………….vii

Table of tables………………………………………………………………………….xv

Table of figures……………………………………………………………………….xviii

List of abbreviations………………………………………………………………….xxiii

List of units…………………………………………………………………………..xxvii

Chapter 1 Introduction ...... 1

1.1 Periodontal diseases ...... 2

1.1.1 Chronic periodontitis ...... 2

1.2 Porphyromonas gingivalis ...... 3

1.2.1 Characteristics ...... 3

1.2.2 Virulence factors of P. gingivalis ...... 4

1.2.2.1 Cysteine proteases ...... 4

1.2.2.2 Fimbriae ...... 7

1.2.2.3 Lipopolysaccharide (LPS) ...... 10

1.2.2.4 Outer membrane vesicles (OMVs) ...... 12

1.3 Treponema denticola ...... 15

1.3.1 Characteristics ...... 15

1.3.2 Virulence factors of T. denticola ...... 15

1.3.2.1 Dentilisin ...... 15

1.3.2.2 Major sheath protein (Msp) ...... 17

1.3.2.3 Motility and chemotaxis ...... 18

vii

1.3.3 Genetic manipulation of T. denticola ...... 22

1.4 Symbioses and synergism between P. gingivalis and T. denticola ...... 23

1.5 Polymicrobial biofilms ...... 24

1.5.1 Polymicrobial biofilms of P. gingivalis and T. denticola ...... 26

1.5.2 T. denticola motility and chemotaxis in synergistic biofilm formation with P. gingivalis ...... 27

1.5.3 Bacterial competitions in polymicrobial biofilms ...... 29

1.6 Thesis hypotheses and aims ...... 32

Chapter 2 Materials and methods ...... 33

2.1 Chemicals, media, supplements and antibiotics ...... 34

2.2 Bacterial strains, plasmids and growth conditions ...... 34

2.2.1 Bacterial strains ...... 34

2.2.2 Bacterial growth conditions and media ...... 34

2.2.3 Bacterial enumeration ...... 35

2.3 Centrifugation ...... 41

2.4 Extraction of genomic and plasmid dna from cells ...... 41

2.4.1 Genomic DNA isolation ...... 41

2.4.2 Plasmid isolation ...... 41

2.5 Manipulation of dna ...... 41

2.5.1 Restriction enzyme digestions ...... 41

2.5.2 DNA precipitation ...... 42

2.5.3 Agarose gel electrophoresis ...... 42

2.5.4 Ligation ...... 43

2.5.5 Genomic PCR amplification ...... 43

2.5.6 Colony PCR amplification ...... 43

2.5.7 Splicing by overlap extension (SOE) PCR ...... 44

2.6 Nucleotide sequencing ...... 45

viii

2.7 Transformation of bacteria ...... 56

2.8 Generation of recombinant strains ...... 57

2.8.1 Construction of P. gingivalis W50::FbFP strain (ECR771) ...... 57

2.8.2 Construction of T. denticola ATCC 35405::FbFPevoglow-Bs2-stop (ECR774)

and ATCC 35405::FbFPevoglow-C-Bs2-stop (ECR775) strains ...... 57

2.8.3 Construction of allele exchange T. denticola mutants ...... 58

2.9 Digestion of recombination cassettes with T. denticola lysates ...... 73

2.10 Fluorescence assays for the analyses of P. gingivalis W50 and T. denticola ATCC 35405 expressing FbFP ...... 73

2.10.1 Fluorescence microscopy ...... 73

2.10.2 Fluorescence spectroscopy ...... 74

2.11 Protein isolation and analysis ...... 74

2.11.1 Trichloroacetic acid (TCA) precipitation ...... 74

2.11.2 Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) …………………………………………………………………………...75

2.11.3 Immunoblotting ...... 75

2.11.4 In-gel trypsin digestion and MALDI-TOF MS or LC-MS/MS ...... 75

2.12 Characterization of T. denticola ATCC 33520 mutants ...... 76

2.12.1 Genomic sequencing...... 76

2.12.2 Colony morphology ...... 77

2.12.3 Swimming assay ...... 77

2.12.4 Cryo-electron microscopy (Cryo-EM) ...... 78

2.12.5 Growth curves ...... 78

2.12.6 Autoaggregation and coaggregation assays ...... 78

2.12.7 Reverse Transcription-Polymerase Chain Reaction (RT-PCR) ...... 79

2.12.8 Quantitative proteomic analyses ...... 80

2.12.9 Static biofilm assay ...... 80

2.12.10 Real-time PCR ...... 81

2.12.11 Growth of biofilms in flow cells ...... 82

2.12.12 Fluorescent staining of biofilms, confocal laser scanning microscopy and image analysis ...... 82

2.12.13 Inhibition assay ...... 83

2.13 Extraction and analysis of proteins secreted from T. denticola colonies ..... 84

2.14 Bioinformatic analyses ...... 84

Chapter 3 Establishment of an efficient and reproducible system for T. denticola transformation and generation of T. denticola mutants…………..…………………….86

3.1 Introduction ...... 87

3.2 Results ...... 89

3.2.1 Transformation of T. denticola ATCC 35405 with plasmid pCF382...... 89

3.2.2 Optimization of transformation efficiency of T. denticola strain ATCC 33520……………………………………………………………………………...94

3.2.3 Transformation of T. denticola ATCC 35405 with plasmid pHN-911 .... 96

3.2.3.1 Construction of pHN-911 suicide plasmid ...... 97

3.2.3.2 Generation of T. denticola ATCC 35405 allele exchange mutant ∆TDE0911 ...... 98

3.2.4 Generation of T. denticola ATCC 33520 and ATCC 35405 mutants using the optimized protocol ...... 99

3.2.5 Generation of T. denticola ATCC 35405 ∆TDE0911::kan mutant lacking TDE0911 ...... 102

3.2.5.1 Construction of the pHN-911kan suicide plasmid ...... 102

3.2.5.2 Generation of T. denticola ATCC 35405 allele exchange mutant ∆TDE0911::kan ...... 104

3.2.6 Digestion of recombination cassettes with T. denticola lysates ...... 104

3.2.7 Generation of mutants in T. denticola ATCC 35405 ∆TDE0911::kan .. 109

3.3 Discussion ...... 111

3.4 Conclusion ...... 117

Chapter 4 Investigation of biofilm development in real time using fluorescently-labelled T. denticola and P. gingivalis………………………………………………………….118

4.1 Introduction ...... 119

4.2 Results ...... 122

4.2.1 Generation of a T. denticola ATCC 35405 strain expressing FbFP (ATCC 35405::FbFP) ...... 122

4.2.1.1 Construction of the pHN-TdEvo suicide plasmid ...... 122

4.2.1.2 Generation of a T. denticola ATCC 35405 mutant expressing FbFP

from evoglow-Bs2-stop (ATCC 35405::FbFPevoglow-Bs2-stop) ...... 125

4.2.1.3 Visualization of T. denticola ATCC 35405 and ATCC

35405::FbFPevoglow-Bs2-stop fluorescence using confocal microscopy ...... 126

4.2.1.4 Immunoblot analyses of FbFP expression from T. denticola ATCC

35405 and ATCC 35405::FbFPevoglow-Bs2-stop transformants 1 and 2 ...... 128

4.2.1.5 Construction of the pHN-TdclosEvo suicide plasmid ...... 129

4.2.1.6 Generation of a T. denticola ATCC 35405 mutant expressing FbFP

from evoglow-C-Bs2-stop (ATCC 35405::FbFPevoglow-C-Bs2-stop) ...... 131

4.2.1.7 Visualization of T. denticola ATCC 35405 and ATCC

35405::FbFPevoglow-C-Bs2-stop transformant 10 fluorescence using confocal microscopy ...... 132

4.2.1.8 Immunoblot analyses of FbFP expression from T. denticola ATCC

35405 and ATCC 35405::FbFPevoglow-C-Bs2-stop transformants 1-11 ...... 134

4.2.2 Generation of a P. gingivalis W50 strain expressing FbFP (W50::FbFP) ………………………………………………………………………….137

4.2.2.1 Construction of the pHN-PgEvo suicide plasmid ...... 137

4.2.2.2 Generation of W50::FbFP ...... 138

4.2.2.3 Visualization of P. gingivalis W50 and W50::FbFP transformants 1-4 fluorescence using microscopy ...... 139

4.2.2.4 Fluorescence intensities of P. gingivalis W50 and W50::FbFP transformants 1-4 as determined by fluorescence spectroscopy ...... 141

4.2.2.5 Immunoblot analyses of FbFP expression from P. gingivalis W50 and W50::FbFP transformants 1-4 ...... 142

4.3 Discussion ...... 145

4.4 Conclusion ...... 149

Chapter 5 Defining the role of T. denticola motility in synergistic biofilm formation with P. gingivalis…………………………………………………………………………...150

5.1 Introduction ...... 151

5.2 Results ...... 155

5.2.1 Construction of the pHN-motA, pHN-motB, pHN-flgE and pHN-cheY suicide plasmids ...... 155

5.2.2 Transformation of T. denticola with pHN-motA, pHN-motB, pHN-cheY and pHN-flgE suicide plasmids ...... 161

5.2.3 Generation of T. denticola ATCC 35405 motility mutants lacking motA, motB, flgE and cheY ...... 162

5.2.4 Generation of motility mutants lacking flgE, motA, motB and cheY in T. denticola ATCC 35405 ∆TDE0911::kan ...... 163

5.2.5 Generation of T. denticola ATCC 33520 motility mutants lacking flgE, motA, motB and cheY ...... 164

5.2.6 Characterization of T. denticola ATCC 33520 motility mutants ∆motA33520,

∆motB33520, ∆flgE33520 and ∆cheY33520 ...... 168

5.2.6.1 Genomic sequencing of ∆motA33520, ∆motB33520, ∆flgE33520 and

∆cheY33520 ……………………………………………………………………..168

5.2.6.2 Morphological differences between T. denticola ATCC 33520 and the

motility mutants ∆motA33520, ∆motB33520 and ∆flgE33520 ...... 169

5.2.6.3 Swimming assay of T. denticola ATCC 33520, ∆motA33520, ∆motB33520

and ∆flgE33520 ...... 169

5.2.6.4 Cryo-electron microscopy of T. denticola ATCC 33520, ∆flgE33520,

∆motA33520, ∆motB33520 ...... 170

5.2.6.5 Growth of T. denticola ATCC 33520, ∆motA33520, ∆motB33520 and

∆flgE33520……………………………………………………………………...174 xii

5.2.6.6 Autoaggregation and coaggregation assays of T. denticola ATCC

33520, ∆flgE33520, ∆motA33520, ∆motB33520 and P. gingivalis W50 ...... 175

5.2.6.7 Reverse Transcription-Polymerase Chain Reaction (RT-PCR) analysis

of fla operon transcription in ∆flgE33520, ∆motA33520 and ∆motB33520 ...... 177

5.2.6.8 Identification of proteins that had changed in abundance in ∆flgE33520,

∆motA33520 and ∆motB33520 ...... 180

5.2.6.9 Static biofilm assay of T. denticola ATCC 33520, ∆flgE33520,

∆motA33520 and ∆motB33520 with P. gingivalis W50 ...... 202

5.2.7 T. denticola ATCC 35405 and HL51 ...... 204

5.2.7.1 Dual-species biofilms of T. denticola ATCC 35405 and HL51 ...... 206

5.3 Discussion ...... 208

5.4 Conclusion ...... 219

Chapter 6 Defining the role of T. denticola Dfsb homologue in T. denticola and in the interactions between T. denticola and P. gingivalis……………………………………220

6.1 Introduction ...... 221

6.2 Results ...... 223

6.2.1 Competing T. denticola ATCC 35405 sibling colonies ...... 223

6.2.2 Bioinformatic analyses ...... 223

6.2.3 Generation of T. denticola allele exchange mutants lacking putative sibling

lethal factor DfsBTd35405 ...... 231

6.2.3.1 Construction of the pHN-0659 suicide plasmid ...... 231

6.2.3.2 Generation of allele exchange mutants lacking DfsBTd35405 in T. denticola ATCC 35405 and ∆TDE0911::kan ...... 233

6.2.3.3 Generation of T. denticola ATCC 33520 allele exchange mutant

∆DfsBTd33520 ...... 233

6.2.4 Characterization of ∆DfsBTd33520 ...... 235

6.2.4.1 Growth curves ...... 235

6.2.4.2 Inhibition assay ...... 236

6.2.5 Isolation and identification of inhibiting material ...... 238

xiii

6.2.6 The role of DfsBTd33520 in the interaction between T. denticola and P. gingivalis ...... 245

6.2.6.1 Inhibition assay with P. gingivalis W50 ...... 245

6.2.6.2 Autoaggregation and coaggregation assays ...... 249

6.2.6.3 Static biofilm assay ...... 250

6.2.6.4 Real-time PCR to determine the number of T. denticola and P. gingivalis cells in the dual-species biofilms ...... 251

6.3 Discussion ...... 254

6.4 Conclusion ...... 261

Chapter 7 General discussion…………………………………………………………262

References…………………………………………………………………………….267

Appendices……………………………………………………………………………299

Appendix I…………………………………………………………………………..299

Appendix II………………………………………………………………………….300

Appendix III…………………………………………………………………………308

xiv

TABLES OF TABLES

Table 2.1 Strains and plasmids ...... 36 Table 2.2 Restriction enzymes used in the course of this study ...... 42 Table 2.3 Standard PCR amplification conditions ...... 44 Table 2.4 Oligonucleotide primers used for PCR amplification and nucleotide sequencing ...... 45 Table 2.5 Primer pairs for generating SOE fragments ...... 51 Table 2.6 Primers used for RT-PCR ...... 80 Table 3.1 T. denticola ATCC 35405 transformation using plasmid pCF382 under aerobic conditions...... 92 Table 3.2 Genes upstream and downstream of TDE2110 in ATCC 35405 and HMPREF9722_RS06010 in ATCC 33520...... 94 Table 3.3 Optimization of T. denticola ATCC 33520 transformation using plasmid pCF382 ...... 96 Table 3.4 The effect of DNA amount on the transformation efficiency of T. denticola ATCC 35405 ...... 99

Table 3.5 Transformation trials to generate ∆motA, ∆motB, ∆flgE and ∆DfsBTd35405 in T. denticola ATCC 35405 ...... 101 Table 3.6 Transformation of ∆TDE0911::kan ...... 109 Table 3.7 List of mutants generated in T. denticola ATCC 35405, ∆TDE0911::kan and ATCC 33520 ...... 110 Table 3.8 Top ten hits of ORF B homologues identified by PSI-BLAST...... 116 Table 5.1 Genes in the fla operon of T. denticola ATCC 35405 and ATCC 33520. ... 165 Table 5.2 Genes in the che operon of T. denticola ATCC 35405 and ATCC 33520. .. 165 Table 5.3 Total and unique proteins that were significantly changed in abundance in T. denticola ∆flgE33520, ∆motA33520 and ∆motB33520, grouped by COG category...... 182 Table 5.4 Proteins significantly changed in abundance in all three T. denticola mutants or

∆motA33520 and ∆motB33520 only, grouped by COG category...... 185

Table 5.5 Proteins significantly changed in abundance in all three T. denticola ∆motA33520,

∆motB33520 and ∆flgE33520 mutants relative to wild-type (ratio≥1.5 and ≤0.67, p<0.05) ...... 186

Table 5.6 Proteins that were not detected in T. denticola ATCC 33520 but were detected in ∆motA33520 and ∆motB33520 (p<0.05) ...... 191

Table 5.7 Proteins significantly increased in abundance in T. denticola ∆motA33520 and

∆motB33520 relative to wild-type (ratio≥1.5, p<0.05) ...... 192

Table 5.8 Proteins significantly decreased in abundance in T. denticola ∆motA33520 and

∆motB33520 relative to wild-type (ratio≤0.67, p<0.05) ...... 195 Table 5.9 Overall biometric parameters of the dual-species biofilms of T. denticola ATCC 35405 and HL51 with P. gingivalis W50 harvested at 90 h ...... 207

Table 6.1 The amino acid sequences of DfsB and DfsBTd35405 ...... 224

Table 6.2 Top seven hits of DfsBTd35405 structural homologues predicted by Phyre2 using the default settings ...... 226

Table 6.3 Top seven hits of DfsBTd35405 structural homologues predicted by Swiss-Model using the default settings ...... 227

Table 6.4 Top seven hits of DfsBTd35405 structural homologues predicted by HHpred using the default settings ...... 228

Table 6.5 DfsBTd35405 homologues in different T. denticola strains identified by blastp ...... 229

Table 6.6 DfsBTd35405 homologues in different bacterial species in the Spirochaete phylum identified by blastp ...... 230

Table 6.7 Genes upstream and downstream of DfsBTd35405 in ATCC 35405 and DfsBTd33520 in ATCC 33520 ...... 234 Table 6.8 MALDI-TOF MS identification of proteins in gel bands 1-14 ...... 240 Table 6.9 MALDI-TOF MS identification of proteins in gel bands 1-8 ...... 243 Table 6.10 Proteins absent in lane B that had high A/D and C/E ratios ...... 245 Table 6.11 A/D and C/E ratios of the proteins identified in Table 6.9 ...... 245 Table 6.12 Bacterial cell numbers in the dual-species biofilms formed by P. gingivalis

W50 grown with T. denticola ATCC 33520 or ∆DfsBTd33520 harvested after 5 days of anaerobic incubation...... 252 Table I.1 Chemicals, media, supplements and antibiotics suppliers…………………..299

Table II.1 Nucleotide polymorphisms of T. denticola ∆flgE33520 mutant determined by genomic sequencing…………………………………………………………………...300

Table II.2 Nucleotide polymorphisms of T. denticola ∆motA33520 mutant determined by genomic sequencing…………………………………………………………………...304

xvi

Table II.3 Nucleotide polymorphisms of T. denticola ∆motB33520 mutant determined by genomic sequencing…………………………………………………………………...306

Table III.1 Proteins significantly changed in abundance in T. denticola ∆flgE33520 mutant relative to wild-type (ratio≥1.5 and ≤0.67, p<0.05)……………………………………308

Table III.2 Proteins significantly changed in abundance in T. denticola ∆motA33520 mutant relative to wild-type (ratio≥1.5 and ≤0.67, p<0.05)……………………………………312

Table III.3 Proteins significantly changed in abundance in T. denticola ∆motB33520 mutant relative to wild-type (ratio≥1.5 and ≤0.67, p<0.05)……………………………………330

xvii

TABLES OF FIGURES

Figure 1.1 Structure of T. denticola ...... 19 Figure 1.2 Model for the T. denticola chemotaxis signalling pathway ...... 21 Figure 1.3 Proposed roles of T. denticola motility in development of P. gingivalis and T. denticola polymicrobial biofilms...... 29 Figure 2.1 Schematic overview of splicing by overlap extension (SOE) PCR method.. 55 Figure 2.2 Generation of P. gingivalis W50::FbFP mutant strain ECR771 ...... 60

Figure 2.3 Generation of T. denticola ATCC 35405::FbFPevoglow-Bs2-stop and ATCC

35405::FbFPevoglow-C-Bs2-stop mutant strains ECR774 and ECR775 respectively ...... 61 Figure 2.4 Generation of T. denticola ATCC 33520 ∆pyrF mutant strain ECR772 ...... 62 Figure 2.5 Generation of T. denticola ATCC 35405 ∆TDE0911 mutant strain ECR773 ...... 63

Figure 2.6 Generation of T. denticola ATCC 35405 ∆cheY35405 mutant strain ECR823 64 Figure 2.7 Generation of T. denticola ATCC 35405 ∆TDE0911::kan mutant strain ECR824 ...... 65 Figure 2.8 Generation of T. denticola ATCC 35405 ∆TDE0911::kan ∆cheY mutant strain ECR825...... 66 Figure 2.9 Generation of T. denticola ATCC 35405 ∆TDE0911::kan ∆flgE mutant strain ECR826 ...... 67

Figure 2.10 Generation of T. denticola ATCC 33520 ∆motA33520 mutant strain ECR827 ...... 68

Figure 2.11 Generation of T. denticola ATCC 33520 ∆motB33520 mutant strain ECR828 ...... 69

Figure 2.12 Generation of T. denticola ATCC 33520 ∆flgE33520 mutant strain ECR829 ...... 70

Figure 2.13 Generation of T. denticola ATCC 33520 ∆cheY33520 mutant strain ECR830 ...... 71

Figure 2.14 Generation of T. denticola ATCC 33520 ∆DfsBTd33520 mutant strain ECR831 ...... 72 Figure 2.15 T. denticola and P. gingivalis inhibition assay ...... 84 Figure 3.1 Plasmid pCF382 ...... 91 Figure 3.2 Generation of plasmid pHN-911 ...... 97

xviii

Figure 3.3 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-911 ...... 98 Figure 3.4 PCR to confirm homologous recombination of TDE0911::ermAM with T. denticola ATCC 35405 genome ...... 99 Figure 3.5 Generation of plasmid pHN-911kan ...... 103 Figure 3.6 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-911kan ...... 103 Figure 3.7 PCR to confirm homologous recombination of TDE0911::aphA2 with T. denticola ATCC 35405 genome ...... 104 Figure 3.8 Recombination cassettes amplified by PCR ...... 105 Figure 3.9 Digestion of the restriction enzyme-digested pCF382 with T. denticola ATCC 35405, ATCC 33520 and ∆TDE0911::kan cell lysates ...... 106 Figure 3.10 Digestion of the unmethylated recombination cassettes with T. denticola ATCC 35405, ATCC 33520 and ∆TDE0911::kan cell lysates ...... 107 Figure 3.11 Digestion of the methylated, NotI-digested pHN-motA, pHN-motB, pHN- flgE, pHN-cheY and pHN-0659 with T. denticola ATCC 35405, ATCC 33520 and ∆TDE0911::kan cell lysates ...... 108 Figure 4.1 Generation of plasmid pHN-TdEvo ...... 123 Figure 4.2 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicon from one E. coli colony harboring pHN-TdEvo using primers M13 Forward and M13 Reverse ...... 123 Figure 4.3 Expression of FbFP in E. coli ...... 124 Figure 4.4 PCR to confirm homologous recombination of TDE0911::evoglow::aphA2 with T. denticola ATCC 35405 genome ...... 125

Figure 4.5 Visualization of T. denticola ATCC 35405, ATCC 35405::FbFPevoglow-Bs2-stop1 and 2 using confocal microscopy ...... 127 Figure 4.6 Analysis of FbFP expression from T. denticola ATCC 35405 and ATCC

35405::FbFPevoglow-Bs2-stop 1-2 using immunoblots ...... 129 Figure 4.7 Generation of plasmid pHN-TdclosEvo...... 130 Figure 4.8 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-TdclosEvo ...... 131 Figure 4.9 PCR to confirm homologous recombination of TDE0911::clos- evoglow::aphA2 with T. denticola ATCC 35405 genome ...... 132

xix

Figure 4.10 Visualization of T. denticola ATCC 35405 with and without BacLight™ stain and ATCC 35405::FbFPevoglow-C-Bs2-stop 10...... 133 Figure 4.11 Analyses of FbFP expression in T. denticola ATCC 35405 and ATCC

35405::FbFPevoglow-C-Bs2-stop 1-11 using immunoblots ...... 135

Figure 4.12 Analyses of FbFP expression in T. denticola ATCC 35405::FbFPevoglow-Bs2-stop and ATCC 35405::FbFPevoglow-C-Bs2-stop using immunoblots ...... 136 Figure 4.13 Generation of plasmid pHN-PgEvo ...... 138 Figure 4.14 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-PgEvo ...... 138 Figure 4.15 PCR to confirm homologous recombination of mfaI::evoglow::ermFAM with P. gingivalis W50 genome...... 139 Figure 4.16 Visualization of P. gingivalis W50 and W50::FbFP 1-4 fluorescence using microscopy...... 140 Figure 4.17 Fluorescence emission spectra of P. gingivalis W50 and W50::FbFP 1-4 A. from unwashed cells in BHI. B. from washed cells in 0.85% (w/v) NaCl ...... 142 Figure 4.18 Analysis of FbFP expression in P. gingivalis W50 and W50::FbFP 1-4 using immunoblot ...... 143 Figure 4.19 Analyses of FbFP secretion from P. gingivalis W50 and W50::FbFP 1-4 via immunoblot ...... 144 Figure 5.1 Diagrammatic representation of the T. denticola chemotaxis signaling pathway and its periplasmic flagellum ...... 153 Figure 5.2 Generation of plasmid pHN-motA ...... 156 Figure 5.3 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-motA ...... 156 Figure 5.4 Generation of plasmid pHN-motB ...... 157 Figure 5.5 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-motB ...... 158 Figure 5.6 Generation of plasmid pHN-flgE ...... 159 Figure 5.7 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-flgE ...... 159 Figure 5.8 Generation of plasmid pHN-cheY ...... 160 Figure 5.9 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-cheY...... 160

Figure 5.10 NotI-digestion of the plasmids pHN-motA, pHN-motB, pHN-cheY and pHN- flgE ...... 161 Figure 5.11 PCR to confirm homologous recombination of cheY::ermAM with T. denticola ATCC 35405 genome ...... 162 Figure 5.12 PCR to confirm homologous recombination of cheY::ermAM with the T. denticola ∆TDE0911::kan genome ...... 163 Figure 5.13 PCR to confirm homologous recombination of flgE::ermAM with T. denticola ∆TDE0911::kan genome ...... 164 Figure 5.14 PCR to confirm homologous recombination of motA::ermAM with T. denticola ATCC 33520 genome ...... 166 Figure 5.15 PCR to confirm homologous recombination of motB::ermAM with T. denticola ATCC 33520 genome ...... 166 Figure 5.16 PCR to confirm homologous recombination of flgE::ermAM with T. denticola ATCC 33520 genome ...... 167 Figure 5.17 PCR to confirm homologous recombination of cheY::ermAM with T. denticola ATCC 33520 genome ...... 168

Figure 5.18 Colony morphologies of T. denticola ATCC 33520 and ∆motB33520 ...... 169

Figure 5.19 Swimming assay of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and

∆motB33520 ...... 170

Figure 5.20 Representative cryo-EM images of T. denticola ATCC 33520, ∆flgE33520,

∆motA33520 and ∆motB33520 ...... 172

Figure 5.21 Growth curves of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and

∆motB33520...... 175

Figure 5.22 Autoaggregation of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520,

∆motB33520 and P. gingivalis W50 ...... 176

Figure 5.23 Coaggregation of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and

∆motB33520 with P. gingivalis W50...... 177

Figure 5.24 RT-PCR analysis of fla operon transcription in ∆motA33520, ∆motB33520 and

∆flgE33520 ...... 178

Figure 5.25 Number of proteins that had changed in abundance in ∆flgE33520, ∆motA33520 and ∆motB33520 ...... 184 Figure 5.26 Mono- and dual-species static biofilms of T. denticola ATCC 33520,

∆flgE33520, ∆motA33520 and ∆motB33520 with P. gingivalis W50 ...... 203 Figure 5.27 Comparison of T. denticola ATCC 35405 and HL51 ...... 205 xxi

Figure 5.28 Representative CLSM images of P. gingivalis W50 and T. denticola ATCC 35405 or HL51 dual-species biofilms ...... 206 Figure 6.1 Competing T. denticola ATCC 35405 colonies ...... 223

Figure 6.2 DUF1706 domain found in DfsBTd35405 ...... 224

Figure 6.3 Alignment of DfsBTd35405 with DfsB using Phyre2 ...... 229

Figure 6.4 Multiple sequence alignment of DfsBTd35405 homologues from different T. denticola strains ...... 231 Figure 6.5 Generation of plasmid pHN-0659 ...... 232 Figure 6.6 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-0659 ...... 232

Figure 6.7 Pairwise sequence alignment of DfsBTd33520 and DfsBTd35405 ...... 234 Figure 6.8 PCR to confirm homologous recombination of TDE0659::ermAM with T. denticola ATCC 33520 genome ...... 235

Figure 6.9 Growth curves of T. denticola ATCC 33520 and ∆DfsBTd33520 ...... 236

Figure 6.10 Inhibition assay of (A.) T. denticola ATCC 33520 and (B.) ∆DfsBTd33520.237

Figure 6.11 The growth of T. denticola ATCC 33520 and ∆DfsBTd33520 within and outside of the center of four evenly-spaced T. denticola ATCC 33520 and ∆DfsBTd33520 colonies...... 237 Figure 6.12 Analysis of proteins extracted from different sections of OBGM agar with competing T. denticola colonies ...... 239 Figure 6.13 Analysis of proteins extracted from different sections of OBGM agar (containing filtered serum) with competing T. denticola colonies ...... 242 Figure 6.14 Analysis of whole proteins extracted from different sections of OBGM agar (no rabbit serum) with competing T. denticola colonies ...... 244 Figure 6.15 The growth of P. gingivalis W50 and ∆ABK within and outside of the zone of clearing between T. denticola ATCC 33520 and ∆DfsBTd33520 colonies ...... 247

Figure 6.16 The growth of T. denticola ATCC 33520 and ∆DfsBTd33520 within and without the regions between P. gingivalis W50 and ∆ABK colonies ...... 248 Figure 6.17 Autoaggregation (A.) and coaggregation (B.) of P. gingivalis W50, T. denticola ATCC 33520 and ∆DfsBTd33520 ...... 249

Figure 6.18 Static biofilm assay of T. denticola ATCC 33520 and ∆DfsBTd33520...... 251 Figure 6.19 Bacterial cell numbers in the dual-species biofilms of ATCC 33520 and

∆DfsBTd33520 grown with P. gingivalis W50 ...... 253

xxii

List of Abbreviations

Α alpha

Ax Absorbance, where x is the wavelength of light (nm) used to measure the absorbance aa amino acid(s)

Ampr ampicillin resistance

Arg- arginine-

ATCC American Type Culture Collection

BApNA N-α-benzoyl-L-arginine-p-nitroaniline

BHI brain heart infusion

CaCl2 calcium chloride

Camr chloramphenicol resistance

CFU colony forming unit

CO2 carbon dioxide

DLS Dynamic Light Scattering

DMSO di-methyl sulfoxide

DNA deoxyribonucleic acid dNTPs deoxynucleotide triphosphates

DTT dithiothreitol

E. coli Escherichia coli

EDSL electron dense surface layer

EDTA ethylenediaminetetraacetic acid

Ermr erythromycin resistance

FbFP flavin mononucleotide-based fluorescent proteins

xxiii

FMN flavin mononucleotide

H2 hydrogen

HagA hemagglutinin A

HBA horse blood agar

Kanr kanamycin resistance

KCl potassium chloride

IL interleukin

IM inner membrane

Kgp lysine-gingipain protease of P. gingivalis

LB lysogeny broth

LOS lipo-oligosaccharide

LpNA N-Tosylglycyl-L-prolyl-L-lysine 4-nitroanilide acetate salt

LPS lipopolysaccharide

Lys- lysine- mAb monoclonal antibody

MCPs methyl-accepting chemotaxis proteins

MgCl2 magnesium chloride

MgSO4 magnesium sulfate

MSP major sheath protein

N2 nitrogen

NaCl sodium chloride

NaOH sodium hydroxide nt nucleotide

NTC no template control

xxiv

O2 oxygen

OD optical density

OBGM oral bacterial growth medium

OM outer membrane

OMPs outer membrane proteins

OMVs outer membrane vesicles

ORF open reading frame

PBS phosphate buffered saline

PCR polymerase chain reaction

PF periplasmic flagella

PFA paraformaldehyde

P. gingivalis Porphyromonas gingivalis

PMF peptide mass fingerprinting

PrtP dentilisin

RT-PCR reverse transcription PCR r recombinant

RE restriction enzyme

RgpA arginine-gingipain A protease of P. gingivalis

RgpB arginine-gingipain B protease of P. gingivalis

RNA ribonucleic acid

RT room temperature

SDS sodium dodecyl sulfate

SDS-PAGE sodium-dodecyl-sulfate polyacrylamide gel electrophoresis

SEM Scanning Electron Microscope

xxv

SN supernatant

TAE tris-acetate buffer

TCA trichloroacetic acid

Tcr tetracycline resistance

Temp. temperature

T. denticola Treponema denticola

TEM Transmission Electron Microscopy

T. forsythia Tannerella forsythia

TLCK Nα-Tosyl-L-lysyl chloromethyl ketone hydrochloride

WC whole cells

xxvi

List of Units

Ω ohms

°C degrees Celsius

Au absorbance units bp base pairs cm centimeter fmol femtomole g relative centrifugal force h hour(s) kbp kilobase pairs kDa kilo Dalton kV kilo Volt

µg microgram

µg/mL microgram per milliliter

µL microliter

µg/µL microgram per microliter mg/mL milligram per milliliter mL milliliter mM millimolar

µM micromolar min minute(s) ms millisecond(s)

M Molar ng nanogram

xxvii

nm nanometer

% percentage pmol picomole rpm revolutions per minute s second(s)

U units

V Volt v/v volume:volume ratio w/v weight:volume ratio

xxviii

“Being confident of this very thing, that he which hath begun a good work in you will perform it until the day of Jesus Christ”

Philippians 1:6 King James Version

xxix

CHAPTER 1 INTRODUCTION

1.1 PERIODONTAL DISEASES

Periodontal diseases are among the most prevalent diseases in the world. They affect the supporting tissues and structures of teeth and are the most common factor for tooth loss, affecting 46% of adults aged 30 years and older in the United States, where 1 in 3 of them are affected by moderate to severe forms of the disease (Eke et al. 2015). The most common forms of periodontal disease are gingivitis and periodontitis (Armitage 2004), which arise from plaque-associated inflammation of the periodontium. The occurrence of damage to the supporting tissues of the teeth is the way to distinguish between gingivitis and periodontitis clinically (Armitage 1995). The milder gingivitis is characterized by the presence of gingival inflammation, which is manifested as red and swollen gums that bleed easily, without loss of connective tissue attachment (Armitage 1995). Gingivitis is reversible with daily cleaning of the teeth and regular removal of plaque by a dentist or dental hygienist. However, if left untreated, gingivitis can advance to the more aggressive periodontitis which is characterized by gingival inflammation with loss of connective tissue and alveolar bone. The symptoms of periodontitis include inflamed gum tissue, bleeding gums, gingival recession, deep-pocket formation between the gum and tooth surface and eventual loose teeth and possible tooth loss (Armitage 1995).

1.1.1 Chronic periodontitis

Chronic periodontitis is an inflammatory, plaque-associated disease of the supporting tissues of the teeth, which results in destruction of the tooth’s supporting tissues including the alveolar bone. It is one of the two categories of periodontitis, the other being localized aggressive periodontitis (Armitage and Cullinan 2010). Chronic periodontitis has been associated with increased risk of cardiovascular disease, preterm and underweight birth, rheumatoid arthritis and certain cancers, such as that of the orogastrointestinal tract and pancreas (Linden et al. 2013, Tonetti et al. 2013). Chronic periodontitis is a polymicrobial disease believed to be initiated by changes in the species composition of subgingival plaque, and subsequent alteration of the host immune response (Byrne et al. 2009). It is linked with the overgrowth of a small number of oral microbiota that form a polymicrobial biofilm (subgingival plaque) accreted to the tooth (Wiebe and Putnins 2000). Several bacterial species are thought to be associated with periodontal diseases however Treponema denticola, Porphyromonas gingivalis and Tannerella forsythia have together been categorized as a significant consortium named the ‘Red Complex’

(Socransky et al. 1998) due to their strong association with the clinical measurements of severe periodontal disease such as periodontal pocket depth and bleeding on probing (Lamont and Jenkinson 1998, Socransky et al. 1998, Holt and Ebersole 2005, Dashper et al. 2011). The synergistic virulence of the ‘Red Complex’ bacteria in periodontal disease progression was demonstrated by a study using a rat model of disease which showed that polymicrobial infections of intra-oral inoculation of P. gingivalis, T. denticola and T. forsythia at a 1:1:1 ratio (3.3 × 109 cells each) resulted in higher levels of bone resorption compared with mono-inoculations of each of the bacterial species (1010 cells) (Kesavalu et al. 2007). According to a report on the oral biofilm architecture on human teeth, Tannerella sp. were found in the intermediate layer of subgingival plaque, P. gingivalis was found as micro-colonies in the top layer while Treponemes were found outside of the top layer (Zijnge et al. 2010). In addition, it was shown that P. gingivalis was only found in the presence of T. forsythia and T. denticola, whereas both T. denticola and T. forsythia were often found in the absence of P. gingivalis, suggesting that T. forsythia and T. denticola may provide P. gingivalis with a suitable environment for its colonization and proliferation (Byrne et al. 2009). Using a comparative genome analysis, the symbiosis among the ‘Red Complex’ bacteria was more recently suggested to comprise of both cooperative and competitive interactions (Endo et al. 2015). The common features of the Red Complex bacteria are their extracellular proteolytic activity that is mediated by cell- surface-located proteases, their complex anaerobic fermentations of amino acids, production of toxic metabolites, and outer membrane (or sheath) vesicles.

1.2 PORPHYROMONAS GINGIVALIS

1.2.1 Characteristics

P. gingivalis is a non-motile, black-pigmented anaerobic Gram-negative coccobacillus with an average diameter of 1 μm (Tan et al. 2014). P. gingivalis belongs to the phylum Bacteroidetes and genus Porphyromonas, which is characterized by the production of large amounts of cell-associated protoheme (Holt et al. 1999). P. gingivalis has been implicated as one of the major pathogens in chronic periodontitis and has often been linked to several other diseases including rheumatoid arthritis and atherosclerosis (Gibson and Genco 2007, Scher and Abramson 2013, Mahalakshmi et al. 2017). Studies using a mouse periodontitis model suggested P. gingivalis as a keystone pathogen that has the ability to dysregulate the host immune response and to disrupt host homeostasis, causing

dysbiosis and favoring polymicrobial biofilm development (Hajishengallis et al. 2012). Byrne et al. (2009) demonstrated that the level of P. gingivalis in subgingival plaque above threshold levels (~10% of total bacterial cell load) is predictive of imminent clinical attachment loss (disease progression). For P. gingivalis to reach the level (10% or greater of total bacterial cell load of the plaque) necessary to produce dysbiosis and significant alveolar bone resorption, tissue inflammation is needed (Lam et al. 2014) and it is aided by the localization of P. gingivalis in the superficial layers of subgingival plaque adjacent to the periodontal pocket epithelium (Zijnge et al. 2010). P. gingivalis produces a variety of virulence factors in both cell-associated and secretory forms such as cysteine proteases, fimbriae, lipopolysaccharide and outer membrane vesicles.

1.2.2 Virulence factors of P. gingivalis

1.2.2.1 Cysteine proteases

Among the many proteases produced by P. gingivalis, cysteine proteases, or more commonly referred to as gingipains, have been implicated as the major virulence factors of P. gingivalis in periodontal diseases. The gingipains, predominantly cell surface- associated and also released on outer membrane vesicles (OMVs), are Arg- and Lys- specific proteases that are encoded by three genes designated rgpA, rgpB and kgp. Arg- specific gingipains (RgpA, RgpB), which are specific for arginine peptide bonds, are derived from rgpA and rgpB while the Lys-specific gingipain (Kgp), which is specific for lysine peptide bonds, is derived from kgp (Curtis et al. 2001). RgpA and Kgp are polyproteins made up of proteases associated with proteolytically processed C-terminal haemagglutinin-adhesin or cleaved adhesin domains (CADs) while RgpB does not associate with any CAD (Slakeski et al. 1998). Furthermore, RgpB lacks a C-terminal adhesin binding motif (ABM) that is found in the catalytic domains of RgpA and Kgp (Slakeski et al. 1998). ABM is proposed to be responsible for the non-covalent association of the RgpA and Kgp catalytic domains into the cell surface complexes (Slakeski et al. 1998, O’Brien-Simpson et al. 2003). Gingipains are one of the most widely studied classic virulence factors of P. gingivalis as they play an essential role at every step of infection: from attachment and colonization, to nutrient acquisition, to evasion of host defenses and to the formation of polymicrobial biofilm.

Gingipains are essential for the adherence and colonization of P. gingivalis. The CADs of RgpA and Kgp, which are also present in hemagglutinin A (HagA), play an important

role in the coaggregation of P. gingivalis with other bacteria. A P. gingivalis ATCC 33277 strain deficient in the CADs (∆hagA ∆rgpA ∆kgp triple mutant) exhibited no coaggregation activity with Actinomyces viscosus (Abe et al. 2004). The proteolytic activities of the gingipains are responsible for the processing of proteins containing CADs and thus production of active forms of the CADs, as the ∆rgpA ∆rgpB ∆kgp triple mutants lacking all three gingipains did not coaggregate with A. viscosus (Abe et al. 2004). The same results were also demonstrated where both of the P. gingivalis ATCC 33277 triple mutants showed significantly reduced coaggregation with T. denticola (Ito et al. 2010).

Rgp proteolytic activity is also required for the cleavage and maturation of P. gingivalis long fimbriae (FimA), another virulence factor of P. gingivalis that mediates the initial attachment of the bacterium to oral tissue (Nakayama et al. 1996, Shoji et al. 2004). This is evident from a study which showed decreased fimbriae expression on the surface of P. gingivalis 381 ∆rgpA mutant (Tokuda et al. 1996). Similarly, gingipains are required for the processing and maturation of a novel third fimbrillin recently identified in P. gingivalis (Nagano et al., 2017; M. J. Gui, unpublished data). Cleavage of the N-terminal region of precursors of MfaI, FimA and the third novel fimbrillin PGN_1808 is required for the donor strand complementation mechanism involved in polymerization of these native fimbrillin proteins (Kadowaki et al. 1998, Nagano et al. 2017, Nakayama et al. 1996, Shoji et al. 2004). The gingipains have also been shown to bind strongly to several extracellular matrix proteins, such as fibrinogen, fibronectin, laminin and collagen type I and V (Pathirana et al. 2006), suggesting their importance in P. gingivalis colonization of the gingival margin. Interestingly, the gingipains preferentially bind immobilized forms of matrix proteins over the soluble forms, suggesting an adaptive mechanism which facilitates P. gingivalis colonization of the host in the fluid environment of the oral cavity (McAlister et al. 2009). Furthermore, gingipains are involved in the modulation of P. gingivalis adherence to eukaryotic cells, including epithelial cells. Chen and Duncan (2004) suggested that the adherence of P. gingivalis to epithelial cells is mediated by the CADs of RgpA, Kgp and HagA while the detachment is mediated by the RgpA and RgpB catalytic domains. This was derived from the observation that antibody raised to the recombinant CAD of RgpA blocked bacterial attachment and a ∆rgpA ∆rgpB double mutant of P. gingivalis showed a higher level of adhesion to epithelial cell monolayers than the parent strain (Chen et al. 2001, Chen and Duncan 2004). The detachment of P. gingivalis from the epithelial cells may be essential for the escape and systemic spread of

P. gingivalis from the oral cavity which will then lead to P. gingivalis-associated systemic diseases.

Apart from functioning as adhesins, gingipains are important for nutrient acquisition. P. gingivalis is asaccharolytic and relies upon low-molecular-weight peptides for its carbon, nitrogen and energy sources (Milner et al. 1996). Gingipains, being responsible for at least 85% of the total proteolytic activity of P. gingivalis (Potempa et al. 1997), are therefore essential for the degradation of serum- and tissue-derived proteins to support the growth of the bacterium. This is evident through a study which showed that P. gingivalis mutants lacking gingipain activities were unable to grow on chemically defined media with the sole carbon and nitrogen source provided by serum proteins (Grenier et al. 2003). Moreover, P. gingivalis has an obligate requirement for iron and protoporphyrin IX (PPIX), both of which can be simultaneously derived from heme. The largest reserves of heme and iron in the human host are within the red blood cells and the ability to agglutinate erythrocytes (hemagglutination) and then lyse them to release hemoglobin (hemolysis) are essential for heme acquisition. P. gingivalis RgpA and Kgp, together with HagA, possess haemagglutinin activity through their CADs (Shi et al. 1999) and are therefore important for the survival of the bacterium in vivo.

Gingipains are also involved in the evasion of the host immune system. The ability of gingipains to degrade different components of the human complement system allows the bacterium to resist killing by serum complement. When compared to the wild type (WT) strain, P. gingivalis mutants lacking either RgpA, RgpB or Kgp were more susceptible, as shown by reduced cell viability, to human serum containing all the active components of the complement system (Grenier et al. 2003). In addition, the gingipains also contribute to serum resistance of P. gingivalis in a proteolysis-independent manner. Potempa et al. (2008) showed that gingipains are able to bind to the human complement inhibitor, C4b- binding protein (C4BP), thereby decreasing the deposition of the host’s membrane attack complex onto the P. gingivalis surface. The binding to C4BP is likely to be mediated by RgpA as a ∆rgpA mutant showed significantly lower binding capacity compared with other gingipain mutants (Potempa et al. 2008). Furthermore, despite having an identical catalytic domain, RgpB showed only a weak binding of C4BP, suggesting that the C4BP binding site may be located within the CAD of RgpA (Potempa et al. 2008). Gingipains also play a key role in dysregulating the cytokine signaling network for the bacterium to

evade host defenses and to change the local environment to suit its needs (Guo et al. 2010). They are able to degrade various cytokines of the interleukin (IL) family, such as IL-1β (Sharp et al. 1998), IL-4 (Yun et al. 2003), IL-8 (Mikolajczyk-Pawlinska et al. 1998), IL-12 (Yun et al. 2001) and IL-6 (Banbula et al. 1999), as well as interferon- gamma (IFN-γ) (Yun et al. 1999) and tumor necrosis factor-alpha (TNF-α) (Calkins et al. 1998). The fact that P. gingivalis is able to survive and even proliferate, as demonstrated by the substantial rise in numbers and proportions of this bacterium in the subgingival plaque under inflammatory conditions suggests that the bacterium is well-adapted to, and can even benefit from, the hostile environment of inflamed tissue (Socransky et al. 1999).

Gingipains are important for the synergistic biofilm formation and synergistic virulence of P. gingivalis with T. denticola. It has been shown that P. gingivalis ∆rgpB mutants were attenuated in synergistic biofilm formation with T. denticola in a static biofilm assay (Kuramitsu et al. 2005a), likely due to the loss of their ability to coaggregate with T. denticola (Ito et al. 2010). As opposed to P. gingivalis WT, a P. gingivalis W50ABK mutant lacking all three gingipains was not able to form polymicrobial biofilms synergistically with T. denticola in flow cells (Zhu et al. 2013). Furthermore, the majority of T. denticola cells lost their typical spiral morphology when grown with P. gingivalis W50ABK in polymicrobial biofilms. This phenomenon was also observed in T. denticola monospecies biofilms but not in polymicrobial biofilms with P. gingivalis WT (Zhu et al. 2013). The requirement for P. gingivalis gingipains in its synergistic virulence with T. denticola was demonstrated in a murine lesion model where a mixed infection of T. denticola with a P. gingivalis mutant deficient in gingipains induced a localized abscess characteristic of T. denticola infection, as opposed to a distant spreading lesion induced by the mixed infection of T. denticola with WT P. gingivalis (Kesavalu et al. 1998). However, precautions should be taken when interpreting these results as the mutations in gingipain genes may have pleotropic effects (Grenier et al. 2003), such as alteration of long fimbriae expression on the mutant cell surface as discussed above.

1.2.2.2 Fimbriae

Fimbriae, the non-flagellar proteinaceous appendages on the P. gingivalis cell surface, are important for the colonization and establishment of the bacterium in subgingival regions. For decades it was thought that only two distinct fimbriae, long or major (FimA) and short or minor (MfaI) fimbriae, are produced by P. gingivalis. However, more

recently a third novel fimbrillin has been described for P. gingivalis (Nagano et al., 2017; M. J. Gui, unpublished data) but there are no studies published characterizing this third fimbrillin yet. The subunit of the long fimbriae is FimA, a 41 kDa protein encoded by the fimA gene, while that of the short fimbriae is MfaI, a 67 kDa protein encoded by mfaI. P. gingivalis long fimbriae have been classified into six types (I, Ib, II, III, IV, V) based on the different nucleotide sequences of the fimA genes and type II FimA was found to possess greater adhesive and invasive capabilities for epithelial cells than other FimA types (Amano et al. 2004). Although both long and short fimbriae appear to contribute to the virulence of P. gingivalis, the role of long fimbriae is more well-established and is considered to be directly responsible for many of the adhesive properties of P. gingivalis. Long fimbriae were shown to be involved in the autoaggregation of P. gingivalis, which is important for the initial attachment and organization of biofilms by P. gingivalis (Kuboniwa et al. 2009). Besides that, the long fimbriae help with the attachment of P. gingivalis to host cells and other bacteria in the subgingival plaque in order to prevent the removal of unattached bacteria by constant salivary flow and shear forces. A fimA mutant that failed to express long fimbriae on the bacterial surface was shown to have only one- third of the adhesive capacity of WT to tissue-cultured human gingival fibroblasts and epithelial cells (Hamada et al. 1994), suggesting the importance of the FimA protein in the interaction of P. gingivalis with gingival tissue cells (Malek et al. 1994). Although it was reported previously that long fimbriae are required for the coaggregation of P. gingivalis with T. denticola, as the coaggregation of these bacteria was inhibited by preincubation of T. denticola with P. gingivalis fimbriae in a dose-dependent manner (Hashimoto et al. 2003), a more recent report demonstrated that a P. gingivalis FimA- deficient mutant did not show reduced coaggregation with T. denticola (Ito et al. 2010). This indicates that there are other surface components besides the long fimbriae that mediate the coaggregation of P. gingivalis with T. denticola, such as gingipains and HagA as mentioned above.

Long fimbriae are also required for the modulation of host immune responses by P. gingivalis. Long fimbriae are involved in epithelial cell invasion, a mechanism to impair host cells and resist clearance by the host immune system (Amano 2007). P. gingivalis long fimbriae are able to bind strongly to cellular α5β1 integrin and it was postulated that following the binding, P. gingivalis is captured by cellular pseudopodia which enable invagination through an actin-mediated pathway (Amano 2007). P. gingivalis fimA

mutants were shown to have approximately eightfold reduction in their capability to invade normal human gingival epithelial cells as compared to the WT (Weinberg et al. 1997). Following invasion, P. gingivalis gingipains, particularly Arg-gingipains, mediate the degradation of cellular focal adhesion components paxillin and focal adhesion kinase (FAK) (Amano 2007), leading to cellular functional impairment (see below). Besides that, long fimbriae, being the most exterior components of the cells, are likely to be the first point of interaction with host cell receptors and important in the induction of inflammatory responses. Several studies have shown that the long fimbriae are capable of inducing human peripheral macrophages and neutrophils to overproduce several proinflammatory cytokines such as IL-l, IL-6, and TNF-α (Ogawa et al. 1994a, Ogawa et al. 1994b). They were also able to stimulate IL-8 production by human gingival epithelial cells through activation of toll-like receptor 2 (TLR2) (Asai et al. 2001). The significant reduction in cytokine responses stimulated by the P. gingivalis fimA mutant may partly explain the inability of the mutant to induce periodontal bone loss in a gnotobiotic-rat model (Malek et al. 1994, Kato et al. 2007). In addition, long fimbriae are involved in immune subversion, thereby allowing P. gingivalis to avoid clearance by the host immune system. P. gingivalis long fimbriae interact with complement receptor 3 (CR3), which inhibits TLR2-induced IL-12 production via extracellular signal-regulated kinase (ERK1/2) signalling (Hajishengallis et al. 2007, Hajishengallis and Lambris 2011). IL- 12 is an important cytokine that regulates the production of IFN-γ, a potent activator of macrophage bactericidal activity (Trinchieri 2003). The reduction in the levels of IFN-γ as a result of CR3 activation by fimbriae thus allows P. gingivalis to suppress cell- mediated immunity (Enersen et al. 2013).

The short fimbriae of P. gingivalis also play a role in cell-cell interactions and induction of inflammatory responses. For example, short fimbriae are essential for the specific recognition of SspB proteins on the Streptococcus gordonii cell surface and development of P. gingivalis biofilms on streptococcal substrates (Lamont et al. 2002). An insertional mutation of the mfaI gene in P. gingivalis resulted in up to 80% reduction in binding to S. gordonii as compared to the parent strain and the mutant was defective in biofilm formation with S. gordonii (Lamont et al. 2002), implicating the importance of short- fimbriae-mediated adherence in P. gingivalis biofilm formation. Another study also showed that short fimbriae are required for P. gingivalis autoaggregation and microcolony formation in biofilm. Although the mfaI mutation did not affect the ability

of P. gingivalis to adhere to saliva-coated surfaces, it did affect the ability of the cells to form microcolonies (Lin et al. 2006). The possible role of short fimbriae in the induction of an inflammatory response and periodontal tissue breakdown was demonstrated by Hiramine et al. (2003), who showed that short fimbriae induce IL-1α, IL-1β, and TNF-α expression in mouse peritoneal macrophages (Hiramine et al. 2003). Furthermore, it was shown that mfaI deletion in P. gingivalis resulted in significant reduction in the ability of the mutant to induce periodontal bone loss in a rat model (Umemoto and Hamada 2003).

1.2.2.3 Lipopolysaccharide (LPS)

Lipopolysaccharide (LPS) is a structure found on the outer membrane of Gram-negative bacteria and it is a key inflammatory mediator due to its ability to potently activate host inflammatory and innate defence responses. LPS is made up of three covalently linked chemical structures: lipid A, core oligosaccharide and O-antigen polysaccharide. The bacterial endotoxic activity is confined to the lipid A, while immunological activity is contained within the O-antigen (Takada and Kotani 1992). Unlike the well-studied enterobacterial LPS, P. gingivalis LPS is known to exhibit very low endotoxic activities (Nair et al. 1983, Fujiwara et al. 1990, Takada et al. 1990), probably due to different phosphorylation and acylation patterns of P. gingivalis lipid A (Ogawa 1993). Despite its low endotoxic activity, P. gingivalis LPS is a potent stimulator of proinflammatory cytokines that are responsible for recruitment of inflammatory cells to the site of infection. For example, gingival fibroblasts have been reported to produce IL-1, IL-6, IL- 8 and TNF-α upon exposure to P. gingivalis LPS (Wang and Ohura 2002, Tardif et al. 2004). Furthermore, P. gingivalis LPS was demonstrated to enhance the expression of IL- 1β and IL-18 from human monocyte cells (Hamedi et al. 2009). The elevated secretion of pro-inflammatory cytokines may then result in inflammation that ultimately leads to tissue destruction.

P. gingivalis LPS contains an unusual amount of lipid A heterogeneity when compared with E. coli LPS (Kumada et al. 1995, Bainbridge et al. 2002, Darveau et al. 2004). For example, P. gingivalis strain W50 has two distinct LPS macromolecules containing different glycan repeating units, A-LPS and O-LPS. A-LPS contains an anionic polysaccharide (APS) with phosphorylated branched mannan repeating units linked to lipid A while O-LPS is a polysaccharide with O-antigen tetrasaccharide repeating units linked to lipid A (Rangarajan et al. 2008). The finding that a monoclonal antibody

(mAb1B5) originally raised against the catalytic domain of RgpA cross-reacted with A- LPS (Curtis et al. 1999), by recognizing the phosphorylated branched mannan in the APS repeat unit (Paramonov et al. 2005), led to the suggestion that A-LPS is the glycolipid outer-membrane anchor for a group of surface proteins with conserved C-terminal domain (CTD), including the gingipains (Veith et al. 2002). The CTD, approximately 70 to 80 amino acyl residues in length, is present in 34 P. gingivalis proteins and is involved in the secretion, maturation and cell surface attachment of CTD-containing proteins (Veith et al. 2002, Seers et al. 2006). The heterogeneous nature of lipid A in P. gingivalis LPS may allow for activation of different signal transduction pathways to elicit various, or even opposing, inflammatory responses. A-LPS has reduced ability to stimulate proinflammatory cytokines, including IL-1α, IL-1β, IL-6 and IL-8, from human monocytes as compared to total LPS (Rangarajan et al. 2008). In addition, P. gingivalis LPS contains both tetra- and penta-acylated lipid A structures that differ in the placement of fatty acids and the number of phosphate groups. The LPS isotype with a penta-acylated lipid A structure was shown to induce expression of E-selectin, a cell adhesion molecule from endothelial cells involved in inflammation, while the tetra-acylated form antagonized E-selectin expression (Reife et al. 2006). Further, penta-acylated lipid A significantly up-regulated the expression of IL-6 and IL-8 in human gingival fibroblasts (HGFs) but tetra-acylated lipid A did not induce a significant host response (Herath et al. 2011). The ability of P. gingivalis LPS heterogeneity to differentially modulate proinflammatory cytokine expression could potentially lead to immunological dysregulation, which is a feature commonly reported in chronic periodontitis.

The activation of toll-like receptors (TLRs), a group of pathogen recognition receptors that play a key role in the host innate immune response, by LPS is an important trigger for the release of chemokines and cytokines by host cells for recruitment of immune cells (Takeda and Akira 2007). Although accumulation of host immune cells such as neutrophils at the site of inflammation is essential in the host defense against infection, resolution of inflammation, a complex process involving induction of immune tolerance and the clearance of neutrophils, is equally important for the prevention of chronic inflammation that might lead to tissue destruction. Induction of immune tolerance leads to a reduced immune response to challenge with microbial antigens to which the host has been previously exposed (Medvedev et al. 2006). Apoptosis, or programmed cell death, of neutrophils and their subsequent removal by macrophages are essential for the

elimination of neutrophils (Jaattela et al. 2004). Two cytokines, IL-8 and TNF-α, are known to play a role in these processes. IL-8 is a potent chemoattractant of neutrophils (Shelburne et al. 2007) and its downregulation is important in the prevention of chronic inflammation caused by neutrophil infiltration, while TNF-α is responsible for the control of neutrophil life span and induction of apoptosis (Maggini et al. 2009). Using a human monocytic cell line, Zaric et al. (2010) showed that P. gingivalis LPS induces only partial immune tolerance, with continued high production of IL-8 but significantly reduced secretion of TNF-α after repeated challenge, suggesting that there is continued migration of neutrophils to the site of infection with prolonged life span (Zaric et al. 2010). In the oral cavity, this may then contribute to the chronic inflammation and tissue destruction seen in periodontitis.

1.2.2.4 Outer membrane vesicles (OMVs)

Outer membrane vesicles (OMVs) are secreted by most of the Gram-negative bacteria via blebbing of their outer membrane and are generally spherical, bilayered-structures with sizes typically in the range of 50 to 250 nm (Kulp and Kuehn 2010). Outer membrane proteins, LPS and other lipids are found on the membrane of OMVs while the vesicle lumen mainly contains periplasmic proteins (Ellis and Kuehn 2010). OMVs are an important virulence factor of bacteria as they allow transportation of insoluble as well as soluble material to distant targets in a concentrated, protected and targeted form (Kulp and Kuehn 2010). P. gingivalis has long been known to produce OMVs that are important for its nutrient acquisition, colonization and biofilm development, host interaction and impairment, dysregulation of host immune systems as well as destruction of periodontal tissues (Gui et al. 2016). Recently, the bacterium has been demonstrated to selectively incorporate certain proteins and virulence factors, such as gingipains, HagA and FimA, in its OMVs (Veith et al. 2014). Further, we have shown in our laboratory that the third novel fimbrillin (PG1881) is also enriched in OMVs (unpublished results). Due to the presence of gingipains and fimbriae, the OMVs of P. gingivalis have the potential to mediate autoaggregation and coaggregation of the bacterium with other bacteria, thereby enabling bacterial colonization and biofilm development. For example, it has been shown in vitro that P. gingivalis OMVs coaggregated various oral microorganisms including: Eubacterium saburreum with Capnocytophaga ochracea; Staphylococcus aureus with various Streptococcus spp.; Actinomyces spp. and mycelium-type Candida albicans; and

P. gingivalis with Prevotella intermedia in a gingipain-dependent manner (Grenier and Mayrand 1987, Kamaguchi et al. 2003a, Kamaguchi et al. 2003b).

The OMVs of P. gingivalis also have an important role in host cell detachment and impairment through their gingipain activity. Nakao et al. (2014) showed that the supernatant of P. gingivalis culture induced detachment of oral squamous epithelial cells in vitro but the cell detachment activity was significantly reduced after the removal of OMVs by ultracentrifugation. Further, purified OMVs were able to cause cell detachment in a dose-dependent manner. Since the cell detachment activity was inhibited in a dose- dependent manner by preincubating OMVs with the antibody raised against the catalytic domain of RgpA, it was postulated that gingipains, particularly Rgp, mediate the cell detachment activity of OMVs (Nakao et al. 2014). Due to their smaller sizes, OMVs are expected to be more efficient in gaining entry into host cells than whole bacterial cells. It was shown by Furuta et al. (2009b) that P. gingivalis OMVs swiftly (within 15 minutes) adhered to human epithelial carcinoma cells (HeLa) and immortalized human gingival epithelial cells (IHGE) and then entered via a lipid raft-dependent endocytic pathway. The internalization process is likely to be dependent on the presence of long fimbriae as OMVs prepared from a ∆fimA mutant were found to negligibly interact with both epithelial cell lines (Furuta et al. 2009b). After being taken up by endocytic compartments, the OMV-associated gingipains degrade the functional molecules such as TfR (transferrin receptor) and paxillin/FAK (focal adhesion kinase) in the host cells, causing cellular functional impairment that can persist for an extended period of time (Furuta et al. 2009a). Both TfR and paxillin are likely to be degraded while the OMVs are attached to the plasma membrane immediately before being isolated into the endocytic compartments. In addition, the localization of TfR on the interior surfaces of the endocytic compartments could also enable its degradation by the OMVs inside the endocytic compartments (Furuta et al. 2009a). TfR is the receptor required for the uptake of extracellular transferrin and iron is released from transferrin after the TfR-transferrin complex is internalized by the cell. The degradation of TfR by P. gingivalis OMV- associated gingipains, particularly Rgp, impairs the uptake of transferrin into the cell and depletes the cell of iron. This negatively affects DNA synthesis and adenosine triphosphate (ATP) generation processes in the cell that require iron, resulting in impaired cellular migration and proliferation (Furuta et al. 2009a). The internalized OMVs are eventually degraded by the cellular digestive machinery after they are sorted to lysosomal

organelles. The observation that the OMVs were able to survive within the lysosomal compartments for more than 24 h (Furuta et al. 2009b), significantly longer than the normal range (2 to approximately 3 h) for protein degradation (Sevlever et al. 2008), supports the idea that OMVs enable efficient delivery of virulence factors in a protected form.

In addition to host cell impairment, P. gingivalis OMVs also dysregulate the host immune system, allowing the bacteria to escape from the host defense response and/or induce an inflammatory response that leads to tissue destruction. For example, P. gingivalis OMV gingipains can degrade CD14, a receptor on human macrophages that recognizes LPS, thus reducing the ability of macrophages to trigger LPS-stimulated cytokine production (Duncan et al. 2004). The OMV gingipains can also degrade the immunoglobulins IgG and IgM as well as complement factor C3 (Grenier and Mayrand 1987, Grenier 1992a, Duncan et al. 2004), allowing the bacteria to escape from the host immune response. OMVs may also act as decoys for the immune system, enabling bacteria to evade immune detection during infection (Nakao et al. 2014). P. gingivalis OMVs have been shown to promote inflammation in several mouse studies. In one of the studies, P. gingivalis OMVs were shown to induce dose-dependent infiltration of neutrophils into connective tissue of mice injected with the vesicles (Srisatjaluk et al. 1999). In another study, P. gingivalis OMVs stimulated the production of nitric oxide (NO), a gaseous inorganic radical which mediates infection defense and inflammation, by murine macrophage cells (Imayoshi et al. 2011). P. gingivalis OMVs also mediate the biosynthesis and surface membrane expression of leukocyte adhesion molecules, E-selectin and intracellular adhesion molecule-1 (ICAM-1), by vascular endothelial cells and therefore may be involved in the initiation of adaptive immunity (Srisatjaluk et al. 1999). A strong mucosal immunogenicity of P. gingivalis OMVs was also shown in a mouse model (Nakao et al. 2011). Further, the antigens released from lysed OMVs in the lysosome may be recognized and processed by antigen presenting cells such as dendritic cells and macrophages, leading to induction of adaptive immunity, including pathogen-specific antibody production (Nakao et al. 2014). This is supported by the observation that the sera from periodontitis patients contain abundant immunoglobulins that specifically react with P. gingivalis OMVs (Nakao et al. 2014) and subcutaneous immunization of mice with P. gingivalis OMVs confers immune protection against challenge by homologous and even heterologous strains of P. gingivalis (Kesavalu et al. 1992).

1.3 TREPONEMA DENTICOLA

1.3.1 Characteristics

T. denticola is an obligate anaerobe that is highly proteolytic. It is a member of the Spirochaete phylum, which is distinct from both Gram-positive and Gram-negative bacteria, possibly due to extensive horizontal gene transfer with Archaea and eukaryotic organisms (Ibba et al. 1997, Brown et al. 1998, Bond and Francklyn 2000, Brown et al. 2001, Wolf et al. 2001, Henz et al. 2005, Paster and Dewhirst 2006). A total of 48 treponemes including T. amylovorum, T. lecithinolyticum, T. denticola, T. maltophilum, T. medium, T. parvum, T. socranskii and T. vincentii were listed in the Human Oral Microbiome Database (HOMD), suggesting the diversity of treponemes in the oral cavity. Among the oral treponemes, T. denticola is most readily cultivable and serves as a good model organism for the study of treponeme-host interactions in periodontal disease. Under the electron microscope, T. denticola is a long, thin, spiral shaped bacterium with an average length of 5 to 20 µm (Tan et al. 2014). T. denticola is a motile bacterium that moves by means of its periplasmic flagella located within the periplasmic space. A number of both classic and non-classic virulence factors of T. denticola are discussed in a review by Dashper et al. (2011). Amongst these virulence factors, the T. denticola surface protein dentilisin and major sheath protein (Msp) are undoubtedly the most- extensively studied classic virulence factors and T. denticola motility and chemotaxis systems are becoming a focal point of study as non-classic virulence factors. Besides contributing significantly to bacterial survival and proliferation, these virulence factors enable T. denticola to interact with other pathogenic bacteria either directly or indirectly, to damage the host and contribute to the initiation and progression of chronic periodontitis (Ishihara 2010).

1.3.2 Virulence factors of T. denticola

1.3.2.1 Dentilisin

Dentilisin (PrtP) is the major extracellular protease and virulence factor of T. denticola. It is also called prolyl phenylalanine-specific serine protease as it cleaves at phenylalanyl/alanyl and prolyl/alanyl bonds (Ishihara et al. 1996). Dentilisin (72 kDa) forms a protease complex with two auxiliary proteins, PrcA1 (~40 kDa) and PrcA2 (~30 kDa). The dentilisin protease complex is encoded by a three-gene operon containing prcB,

prcA, and prtP. The PrcA polypeptide is cleaved by PrtP to PrcA1 and PrcA2 which likely contribute to the anchoring of the protease complex in the outer membrane (Lee et al. 2002a, Godovikova et al. 2010). Although the role of PrcB has not yet been fully elucidated and it has not been considered as part of the active protease complex, the protein is required for the expression and activity of the dentilisin protease complex, probably by stabilizing the PrtP polypeptide during localization to the outer membrane (Godovikova et al. 2010). Dentilisin is proposed to contribute to T. denticola nutrient processing or acquisition as T. denticola PrtP-deficient mutants were unable to degrade high-molecular-weight media constituents (Bian et al. 2005). It is also believed to disrupt or modulate host intercellular signaling and degrade host cell matrix proteins, thereby contributing to disease progression. The protein may be involved in T. denticola penetration of epithelial cell layers and invasion of underlying tissue in advanced periodontitis through degradation of intercellular adhesion proteins (Chi et al. 2003). In addition, dentilisin modulates host cell immune responses by its ability to degrade interleukin-1β (IL-1β), IL-6, tumor necrosis factor alpha (TNF-α), and monocyte chemoattractant protein 1 (Miyamoto et al. 2006, Okuda et al. 2007). Dentilisin also plays an important role in the structural organization and processing of T. denticola cell-surface components, including the other widely studied virulence factor Msp (Ishihara et al. 1998, Dashper et al. 2011). This is supported by studies that showed that Msp production levels and oligomerization were defective in mutants that lack PrtP protease activity and the high molecular mass oligomeric protein characteristic of the outer sheath of T. denticola was reduced with the loss of PrtP activity (Ishihara et al. 1998) (Lee et al. 2002a).

Dentilisin is also known to act as an adhesin on the outer membrane of T. denticola and is important for the coaggregation between T. denticola and other bacteria, such as T. forsythia and P. gingivalis. Coaggregation is an important process in the development of dental plaque, which contributes to the pathogenesis of periodontal diseases, as it allows colonization and adherence of late colonizers to an established biofilm and subsequently increases the numbers of these organisms in the dental plaque (Kolenbrander et al. 1993). Dentilisin has been proposed to interact with the fimbriae of P. gingivalis, thus allowing the bacteria to coaggregate and localize closely together in a mature biofilm (Hashimoto et al. 2003). Close association of bacterial species are beneficial as it enables metabolic communication while minimizing dilution of metabolites and signaling molecules (Grenier 1992b). Dentilisin has also been shown to contribute to the coaggregation

between T. denticola and T. forsythia. The coaggregation between T. denticola and T. forsythia was completely lost with the inactivation of dentilisin and the degree of coaggregation with T. forsythia for different T. denticola strains was proportional to their level of dentilisin activity (Sano et al. 2014). The protease activity of dentilisin is not directly involved in the coaggregation reaction as treatment of T. denticola cells with the serine protease inhibitor, phenylmethylsulfonyl fluoride (PMSF), did not affect coaggregation (Sano et al. 2014). Thus, further analysis is required to identify the mechanism by which dentilisin is involved in the coaggregation of T. denticola and T. forsythia but it is likely that dentilisn is required for the processing and maturation of specific cell surface coaggregation factors.

1.3.2.2 Major sheath protein (Msp)

Major sheath protein (Msp) is the most abundant protein in the T. denticola outer membrane (or sheath). It is a 53 kDa integral β-barrel membrane protein with dual functions: as an adhesin that mediates adhesion to other bacteria and colonization of host tissues, as well as a porin that acts as a permeable pore in the membrane. MSP contributes to the coaggregation of T. denticola with P. gingivalis and F. nucleatum through its protein and carbohydrate moieties, respectively (Rosen et al. 2008). The surface-exposed N-terminal loops of Msp are able to bind to a variety of host proteins, including fibronectin, fibrinogen and laminin (Fenno and McBride 1998, Ellen 2006). In a study done by Bamford et al. (2007), the binding of T. denticola to fibrinogen was reduced by 15-20% when Msp was inactivated. The observation that the binding was diminished even more (~80%) when dentilisin was inactivated (Bamford et al. 2007) further reinforces the importance of dentilisin in the processing of Msp or the formation of an Msp complex within the outer membrane of the bacterium. The binding of T. denticola to fibronectin potentially serves several purposes. Firstly, soluble and degraded fibronectin may serve as a nutrient source for the bacteria. Secondly, the decoration of the bacterial cell surface with fibronectin may divert the host immune defense mechanisms and reduce bacterial clearance. Thirdly, the colonization and subsequent host tissue invasion by T. denticola may be enhanced through its binding to immobilized fibronectin, which is commonly found associated with the extracellular matrix or deposited onto oral surfaces (Bamford et al. 2007). The binding of recombinant Msp to heterologous Treponema species with low fibronectin-binding capability enhanced the bacterial fibronectin-binding capability by 10-fold (Edwards et al. 2005), suggesting an

advantageous phenotypic complementation and synergy in adherence amongst mixed populations of oral spirochetes containing T. denticola (Edwards et al. 2005). However, a T. denticola ∆msp mutant and WT strains bound to human gingival epithelial cells similarly (Abiko et al. 2014a), indicating the involvement of other factors in the interactions of T. denticola with host cells.

Msp has also been shown to form a pore with large channel diameter in planar lipid bilayers and has cytopathic pore-forming activity against cultured epithelial cells (Fenno et al. 1998). It is involved in the release of extracellular matrix-degrading neutral proteinases from human polymorphonuclear leukocytes (PMN) which contributes to tissue destruction during periodontal inflammation. Although the mechanism by which it works in vivo is unclear, an in vitro study using purified Msp suggested the fusion of the protein to the PMN membranes and the subsequent lysosomal discharge of the proteases (Ding et al. 1996). In addition to the adherence and cytotoxic activity, Msp also has an immunomodulatory role. Being the most abundant protein in T. denticola which is exposed on the cell surface, Msp is a major antigen that is easily accessible to immune cells (Abiko et al. 2014a). A purified Msp complex was shown to induce innate immune responses of murine macrophages through Toll-like receptor 2-myeloid differentiation factor 88 (TLR2-MyD88) (Nussbaum et al. 2009). It was also shown to activate a proinflammatory response by inducing the production of TNF-α, IL-1β, IL-6 and matrix metalloproteinase 9 (MMP-9) by primary human peripheral blood monocytes (Gaibani et al. 2010). Besides that, interaction with Msp leads to host cell actin remodeling and reorganization, which likely impairs neutrophil chemotaxis and phagocytic activity (Dashper et al. 2011). The observation that macrophages became tolerant to stimulation by enterobacterial LPS after activation by purified Msp suggests that T. denticola may play a broader role in the persistence of the bacterial consortia found associated with chronic periodontitis by allowing them to evade clearance through suppression of the host response to LPS (Nussbaum et al. 2009).

1.3.2.3 Motility and chemotaxis

Although not generally referred to as classic virulence factors in most human pathogens, motility and chemotaxis are undoubtedly important in bacterial-host interactions and disease progression. Chemotaxis allows motile bacteria to move towards or away from environmental stimuli. The combination of a chemotaxis system and motility enables

efficient nutrient acquisition, avoidance of toxic substances and translocation to optimal colonization sites by the bacteria. It is thus important in the survival and proliferation of the bacteria, especially for the nutritionally fastidious organisms such as the oral treponemes. Approximately 5-6% of the genome of sequenced treponemes is responsible for motility and chemotaxis (Seshadri et al. 2004). The motility of T. denticola is dependent on its periplasmic flagella (PF) which, unlike the exposed flagella of most motile bacteria, are located in the periplasmic space between the outer and cytoplasmic membranes of the cells (Chi et al. 1999). T. denticola has four PF, two anchored at each end of the cell which extend towards the center of the cell and overlap at mid-cell (Izard et al. 2008). The periplasmic flagella of T. denticola are made up of a hook basal body complex and a flagellar filament. The filament consists of three filament core proteins (FlaB1-3) and three filament outer layer proteins (FlaA1-3). The majority of T. denticola cells at exponential phase adopt a highly irregular twisted morphology attributed to the PF; this irregular morphology was not observed in the absence of PF (Ruby et al. 1997). It is beneficial for T. denticola penetration and invasion of periodontal tissues to have the PF protected in the periplasmic space from the immobilization effects of highly viscous environments and the flagella-specific antibodies produced by the host response to infection (Fenno and McBride 1998, Charon and Goldstein 2002).

Figure 1.1 Structure of T. denticola. Periplasmic flagella are located within the periplasmic space of T. denticola. Major outer sheath protein (Msp) and dentilisin are embedded on the outer sheath of T. denticola (adapted from Ishihara 2010).

Chemotaxis is the mechanism by which motile bacteria respond to their environment. It allows the bacteria to migrate towards favorable stimuli, such as nutrients, and away from

unfavorable stimuli, such as toxins. Genome sequencing and bioinformatic analyses of T. denticola revealed that it not only possesses a complete flagellum-based motility system but also a complete set of chemotactic proteins required for signal perception, transduction and adaptation (Dashper et al. 2011), indicating the importance of chemotaxis for this bacterium. Like the most widely studied chemotaxis system of Escherichia coli, a chemotaxis gene cluster or operon containing three chemotaxis genes, cheA, cheW and cheY, was observed in the genome of T. denticola (Greene and Stamm 1999). However, the T. denticola che operon contains an additional open reading frame, cheX, between cheW and cheY. CheR and CheB, which are responsible for adaptation in T. denticola, are encoded by another operon in the genome of T. denticola (Seshadri et al. 2004). An unusually high number (~20) of methyl-accepting chemotaxis proteins (MCPs), inner membrane-spanning chemoreceptors used to detect environmental stimuli, were identified in the T. denticola genome which may reflect the complexity of the ecological niche of this bacterium (Seshadri et al. 2004). Two of these proteins, DmcA and DmcB, were shown to be required for T. denticola chemotaxis towards nutrients (Kataoka et al. 1997, Li et al. 1999). The chemotaxis system of T. denticola works in a similar manner to that of E. coli. Chemotactic stimuli are detected by the MCPs and the signals are transmitted via CheA, CheW and CheY to the flagellar motor. The histidine kinase CheA forms a membrane-associated complex with the MCPs as well as CheW, which is required for the formation of the MCP-CheA-CheW ternary complex, and undergoes ATP-dependent autophosphorylation in compliance with the stimulus bound to the respective MCP. The nature of the stimulus is then communicated to the motor via phosphotransfer reactions from CheA to the response regulator CheY, which controls the direction of flagellar motor rotation according to its phosphorylation level (Figure 1.2). Repellents increase autophosphorylation of CheA, whereas attractants decrease this activity (Lux et al. 2002). The adaptation in T. denticola is achieved via methylation or demethylation of the MCPs by CheR and CheB (Sim et al. 2005).

Figure 1.2 Model for the T. denticola chemotaxis signalling pathway. Chemotactic stimuli are detected by the methyl-accepting chemotaxis proteins (MCPs) which traverse the inner membrane. CheA, which forms a membrane-associated complex with the MCPs as well as CheW, undergoes ATP-dependent autophosphorylation and the signals are transmitted to the periplasmic flagellar motor via phosphotransfer reactions from CheA to the response regulator CheY, which controls the direction of flagellar motor rotation. FliM and FliY are part of the motor proteins (adapted from Sim et al. 2005).

The ability to penetrate tissue allows pathogenic spirochetes to migrate to sites favorable for bacterial proliferation and is important for bacterial virulence. It also plays a role in tissue destruction in chronic periodontitis. By using an oral epithelial cell line-based tissue penetration assay, both motility and chemotaxis were shown by Lux et al. (2001) to be involved in tissue penetration by T. denticola. In this study, a non-motile mutant of T. denticola (∆flgE) was unable to penetrate the tissue layer, indicating that active bacterial movement rather than passive translocation via endocytosis by the eukaryotic cells or other cellular processes is important for tissue invasion. The tissue penetration of the chemotaxis mutants, ∆cheA, ∆dmcA and ∆dmcB, was also significantly impaired as compared to the WT cells although they were all motile. The ∆cheA mutant showed only about 2 to 3% of the WT penetration rate while the ∆dmcA and ∆dmcB mutants showed about 30 and 10% of the WT penetration rate respectively (Lux et al. 2001). Although the relationship between chemotaxis and tissue penetration was unclear, a possible explanation could be derived from the observation that T. denticola exhibited an

exponential rather than linear or hyperbolic increase of penetration efficiency over time (Lux et al. 2001). The known chemoattractants for T. denticola include glucose, serum, and albumin, which are all possible indicators of damaged host tissues (Umemoto et al. 2001, Ruby et al. 2008). During the penetration process, the release of these compounds from damaged tissues could attract T. denticola, leading to chemotaxis-guided targeting of the bacteria to damaged tissue where it can take advantage of the weak spots (Lux et al. 2001).

1.3.3 Genetic manipulation of T. denticola

Genetic manipulation, i.e. manipulation of the genes of an organism, is important for the functional analysis of genes and their gene products and is a vital tool in the investigation of the molecular basis of bacterial virulence. The first efforts to genetically modify spirochetes were performed in 1992 using electroporation to transfer plasmid DNA into Brachyspira hyodysenteriae, the causative agent of swine dysentery (ter Huurne et al. 1992). The second spirochete that was successfully transformed was the Lyme disease agent Borrelia burgdorferi in 1994 by electroporation (Samuels et al. 1994). T. denticola was the third spirochete with which a genetic manipulation system was successfully developed by electroporation of a wide host range plasmid into the bacterium (Li and Kuramitsu 1996). In the same year, 1996, homologous recombination into the T. denticola genome was achieved by electroporation, resulting in inactivation of the flgE gene by the Bacteroides fragilis erythromycin resistance markers ermF and ermAM (Li et al. 1996). Over the twenty plus years since, the progress in the transformation of T. denticola with allelic replacement cassettes and shuttle plasmids has been very slow, mainly because of the low transformation efficiency of T. denticola, i.e. low number of transformants per microgram of input DNA. The availability of a small number of selectable markers and strain-limited shuttle plasmids has further hindered complementation analyses and the expression of heterologous genes in T. denticola (Kuramitsu et al. 2005b, Bian and Li 2011, Godovikova et al. 2015, Li et al. 2015). The transformation efficiency differs among different strains of T. denticola and the transformation of the type strain, ATCC 35405, is particularly challenging. The presence of unique restriction-modification systems and the absence of a natural plasmid in T. denticola ATCC 35405 were proposed to be the main contributors to the low transformation efficiency of this strain of T. denticola (Chi et al. 2002, Bian and Li 2011). Another T. denticola strain, ATCC 33520, which lacks the restriction-modification systems and harbors a natural plasmid is reported

to have a higher transformation efficiency than ATCC 35405 (Ivic et al. 1991, Bian and Li 2011). Nevertheless, the overall transformation efficiency of T. denticola ATCC 33520, usually <200 colony forming unit (CFU) per microgram of DNA, is still much lower than that of other bacteria. Although more efficient methods to generate T. denticola allelic exchange mutants and to facilitate their complementation by plasmids have been developed, the protocols are not routinely reproducible. Furthermore, the high rate of “false-positives” that often accompanies the transformation of T. denticola impedes efforts to identify true recombinants (Ishihara et al. 2010, Bian and Li 2011). Therefore, it is important to optimize the transformation protocol of T. denticola to allow routine reproducible transformation. During the course of this PhD study, a significant increase of at least 100-fold in the plasmid transformation efficiency of the type strain ATCC 35405 using a modified protocol was reported by Godovikova et al. (2015). The modified protocol included minimizing T. denticola cell exposure to oxygen in all processing steps to prepare electrocompetent cells and in the steps during and after electroporation (Godovikova et al. 2015). Using the modified protocol, the authors were able to routinely produce 103 to 104 antibiotic-resistant colonies using 1 µg of input plasmid DNA, much higher than that obtained in the previous studies, which is normally <200 (and in most cases <10) using the typically high amount (5-10 µg) of input plasmid DNA in any T. denticola host strain (Godovikova et al. 2015). Based on the study of Godovikova et al. (2015), it appears that keeping the cells in a strict anaerobic environment during the whole transformation process, from the preparation of electrocompetent cells to the recovery of cells after electroporation, is a critical determinant in increasing the transformation efficiency of T. denticola.

1.4 SYMBIOSES AND SYNERGISM BETWEEN P. GINGIVALIS AND T. DENTICOLA

Culture-dependent isolation techniques have previously characterized most diseases as monomicrobial. However, due to the advancement of culture-independent community analysis methodologies, it has been acknowledged that some diseases, particularly chronic diseases such as dental caries and chronic periodontitis, are polymicrobial in nature (Cripps and Otczyk 2006, Peters et al. 2012). It is also now well accepted that interactions among selected members of the microbiota are essential for disease progression. P. gingivalis and T. denticola are members of the Red Complex bacteria that are often found to co-localize in subgingival plaque and co-exist in deep periodontal

pockets. This suggests a strong ecological relationship and potential interactions between the two bacteria that may contribute to the persistence of the bacteria in the periodontal pocket and the progression of chronic periodontitis (Socransky et al. 1998, Tan et al. 2014). For example, several reports have shown that T. denticola and P. gingivalis display symbioses in protein degradation, nutrient utilization and growth promotion (Grenier 1992b, Nilius et al. 1993, Kigure et al. 1995, Hollmann and Van der Hoeven 1999, Grenier and Mayrand 2001, Yoneda et al. 2005). Using continuous co-culture of T. denticola and P. gingivalis, Tan et al. (2014) showed that the cell densities of both P. gingivalis and T. denticola increased significantly, by 54% and 30% respectively, in co- culture compared with mono-culture. In addition, the presence of P. gingivalis caused an upregulation of T. denticola glycine catabolism pathways and T. denticola conditioned medium stimulated free glycine production by P. gingivalis. Since free glycine is an important carbon source for T. denticola, a metabolic symbiosis between the two bacteria in co-culture is proposed (Tan et al. 2014). Besides symbioses, the two bacteria also display synergistic virulence. Kesavalu et al. (1998), using an in vivo murine abscess model, demonstrated that at low doses of P. gingivalis alone where minimal or no lesions were induced, the addition of T. denticola significantly enhanced the formation of spreading lesions (Kesavalu et al. 1998). Furthermore, Orth et al. (2011), using a murine experimental model of periodontitis, showed a 40-fold reduction in total bacterial cell number needed to induce significant periodontal bone loss in co-inoculation of a 1:1 ratio of P. gingivalis and T. denticola as compared with inoculation of P. gingivalis alone (Orth et al. 2011). The observed synergistic virulence of T. denticola and P. gingivalis co- infections in animal models may be partly explained by the upregulation of T. denticola genes encoding virulence factors, such as dentilisin and Msp, during the co-culture of the two bacterial species (Tan et al. 2014). Taken together, these results support the idea that T. denticola facilitates the invasion and tissue damage by P. gingivalis and the two bacteria act synergistically in progression of chronic periodontitis (Dashper et al. 2011) (Orth et al. 2011).

1.5 POLYMICROBIAL BIOFILMS

The majority of microbes in nature form biofilms that attach to biotic and abiotic surfaces and rarely exist entirely in planktonic forms. The formation of bacterial biofilms has been considered an important virulence factor in many chronic and persistent infections. Biofilms present a significant threat as they are often resistant to the host immune

response and the highest deliverable levels of antibiotics. Chronic biofilm-associated infections frequently result in host tissue damage that may have a long-term impact on the patients. Because of their resistance to antibiotics, biofilm-associated infections can be life threatening to immune-compromised patients (Li and Tian 2012). Biofilm communities in most environments, including the human body, tend to be polymicrobial in nature (del Pozo and Patel 2007). Polymicrobial biofilms are defined as “a varied collection of organisms (fungi, bacteria, and viruses) that exist at a phase or density interface and are coated in a self- and/or host-derived hydrated matrix, often consisting of polysaccharide” (Brogden et al. 2005).

Several advantages are gained through the formation of a polymicrobial biofilm, such as passive resistance (Weimer et al. 2011), metabolic cooperation (Fischbach and Sonnenburg 2011, Elias and Banin 2012), byproduct influence (Carlsson 1997), quorum sensing systems (Elias and Banin 2012), more efficient DNA sharing (Madsen et al. 2012) and other synergies. Passive resistance occurs when the resistance factor possessed or produced by one member in the biofilm confers resistance to the other members of the biofilm that do not have the factor. One such example is the passive resistance of Streptococcus pneumoniae to a β-lactam antibiotic in the presence of the β-lactamase- producing Haemophilus influenzae strain (Weimer et al. 2011). Metabolic cooperation occurs when the metabolic byproduct of one member is utilized by another member in the biofilm. It has been reported in several bacteria including those within the human oral cavity including Aggregatibacter actinomycetemcomitans and Streptococcus gordonii, Actinomyces naeslundii and S. gordonii, as well as T. denticola and P. gingivalis (Jakubovics et al. 2008, Ramsey et al. 2011, Tan et al. 2014). Byproducts or end-products of metabolism, such as ammonia, lactic acid, and carbon dioxide, from one member can have a significant influence on the surrounding members of a biofilm. For instance, ammonia production from the fermentation of nitrogenous compounds by Fusobacterium nucleatum, Prevotella intermedia and P. gingivalis may contribute to the acid neutralizing activity in dental plaque and promote the survival of acid-sensitive bacteria (Takahashi 2003). Quorum sensing is a well-studied mechanism in which individual organisms in a biofilm communicate with one another through the release of small diffusible signaling molecules. Several quorum sensing systems for intra- and/or interspecies communications that are important for biofilm development and coordinated physiological activities among the members in a biofilm have been reviewed in Elias and

Banin (2012), Li and Tian (2012) and Ng et al. (2016). Horizontal gene transfer is a mechanism whereby one bacterium obtains specific traits from another bacterium through the transfer of genetic materials such as plasmids. With multiple species located in close proximity in a single community, biofilms provide an enlarged gene pool with more efficient DNA sharing which is important for the adaptation of bacteria in the biofilm (Madsen et al. 2012).

During cohabitation and cocolonization, the microbes that formed the biofilm community have developed mutualistic or synergistic relationships or competitive antagonistic approaches that gives them a competitive advantage over other microbes (Peters et al. 2012). Given the unique biology of bacterial interactions in biofilms, it is essential to understand these interactions in order to develop effective therapeutic strategies to prevent or treat biofilm infections. Biofilms are generally studied using static or flow cell systems. Static biofilms are easier to set up and require less specialized skill, but its static nature may result in nutrient limitation and bacterial waste accumulation (Merritt et al. 2005). On the other hand, the biofilms developed in a flow system, such as a flow cell, are continuously irrigated with fresh medium under constant flow, thus allowing nutrient replenishment and bacterial waste removal (Merritt et al. 2005). In addition, the adherent surface in the static biofilm model is generally polystyrene whilst that in the flow cell model is generally glass. Due to the differences in their settings and adherent surfaces, static biofilm and flow cell models can produce different results (Vanhommerig et al. 2014).

1.5.1 Polymicrobial biofilms of P. gingivalis and T. denticola

The oral cavity is a habitat where microbial biofilms are the preferred and predominant bacterial lifestyle due to the constant removal of unattached bacteria by shear forces and bulk phase movement. In vivo, P. gingivalis and T. denticola are frequently detected together in the superficial layers of human subgingival plaque associated with a chronic periodontitis lesion (Kigure et al. 1995). In vitro, P. gingivalis and T. denticola have been shown to form polymicrobial biofilms synergistically using both static biofilm models and flow cell assays. Using a static biofilm model, it was shown that the polymicrobial biofilms of P. gingivalis and T. denticola have higher quantities of colonization compared with each organism alone and they adhere more tightly to the substratum than the monospecies biofilms (Yamada et al. 2005). The biofilm formation was not enhanced

when the two bacteria were separated by filters (Yamada et al. 2005), indicating the importance of bacterial coaggregation in initiation and development of biofilm formation. Using a flow cell biofilm model, it was shown that there was an approximately three-fold and six-fold increase in the total biovolumes of P. gingivalis and T. denticola respectively in polymicrobial biofilms as compared to monospecies biofilms. The average and maximum thickness of the polymicrobial biofilms was also significantly increased as compared to the homotypic biofilms (Zhu et al. 2013), suggesting a strong synergy in biofilm formation between P. gingivalis and T. denticola.

1.5.2 T. denticola motility and chemotaxis in synergistic biofilm formation with P. gingivalis

Bacterial motility and chemotaxis, although not considered as classic virulence factors, are likely to be essential for the formation and development of biofilms. There are three proposed mechanisms in which motility may be involved in biofilm formation and development. First, the flagella may act as an adhesin to enable bacterial attachement to abiotic surfaces and facilitate the initiation of biofilm formation. Second, motility may be required to breach surface tension at the medium-surface interface and enable the bacteria to reach a certain substratum. Finally, motility may facilitate the spread of a developing biofilm by allowing bacteria within the biofilm to move along the substratum (Pratt and Kolter 1998, Merritt et al. 2007). In the study of non-motile E. coli mutants, it was observed that both the aflagellated (∆fliC and ∆flhD) and flagellated (∆motA, ∆motB and ∆motAB) mutants were deficient in biofilm formation, suggesting that flagella do not function as inert adhesins and active motility is required for biofilm formation (Pratt and

Kolter 1998). In addition, motility was essential for the detachment of P. aeruginosa cells from the biofilm and spread across surfaces for the expansion of the biofilm (An et al. 2006).

Unlike motility, the role of chemotaxis in biofilm formation is unclear and the requirement for chemotaxis during biofilm formation appears to be species-specific. For example, chemotaxis was shown to be essential for Aeromonas spp. and A. tumefaciens biofilm formation. The inactivation of the chemotaxis sensor kinase cheA in Aeromonas spp. resulted in more than 80% reduction in biofilm formation as compared to the WT (Kirov et al. 2004). Similarly, an A. tumefaciens ΔcheA mutant showed reduction in attachment efficiency and aberrant biofilm phenotype in flow cells. The biomass and

height of the A. tumefaciens ΔcheA mutant biofilms were approximately threefold lower than those of WT biofilms, with reduced overall substratum coverage (Merritt et al. 2007). In E. coli however, chemotaxis was shown to be fully dispensable for biofilm formation. Using a static biofilm assay, it was shown that a motile but non-chemotactic mutant of E. coli, ΔcheA-Z which had the chemotaxis operon inactivated, formed biofilms that were indistinguishable from the WT (Pratt and Kolter 1998). A T. denticola mutant lacking one of the MCPs, DmcA, had a similar biofilm forming ability as WT with P. gingivalis under static culture (Vesey and Kuramitsu 2004). However, since the inactivation of one or two chemoreceptors is unlikely to represent the whole chemotaxis system, the importance of T. denticola chemotaxis in synergistic biofilm formation with P. gingivalis may be better tested using a mutant deficient in cheA, the central kinase of the chemotaxis pathway.

Among the three Red Complex bacteria that are strongly associated with chronic periodontitis, T. denticola is the only motile species. It was observed previously in a flow cell assay that T. denticola lost its typical spiral morphology in monospecies biofilms but retained it in the polymicrobial biofilms with P. gingivalis and T. forsythia (Zhu et al. 2013). Since the morphology of T. denticola is closely related to its internal periplasmic flagella (Ruby et al. 1997) and periplasmic flagella are required for the motility of T. denticola, T. denticola motility is likely to be involved in its synergistic biofilm formation with P. gingivalis. Several studies have supported this hypothesis. In a static biofilm model, a non-motile and aflagellated T. denticola ∆flgE mutant was significantly attenuated in colonizing preformed P. gingivalis biofilms (Vesey and Kuramitsu 2004). Another study showed a reduction in coaggregation and polymicrobial biofilm formation between a T. denticola ∆flgE mutant and P. gingivalis (Yamada et al. 2005). The ∆flgE mutant was also unable to form a polymicrobial biofilm synergistically with P. gingivalis and T. forsythia in flow cells (Zhu et al. 2013). The ability of T. denticola to move in highly viscous environments (Klitorinos et al. 1993), which is expected in a biofilm, is beneficial for the bacterial movement through the biofilms. It has been shown that motile bacteria can create pores in the biofilm matrix (Houry et al. 2012). In a biofilm where there is often a nutrient gradient, the chemotactic ability of T. denticola may be useful in directing the bacterium to potential sources of nutrition. As T. denticola cells move through the biofilm, they may leave behind pores which then allow for nutrient and waste exchange between the biofilm and the environment. This may then enable a higher

biofilm biomass to be achieved (Ng et al. 2016) (Figure 1.3). The known chemoattractants for T. denticola which include glucose, serum and albumin are all possible indicators of damaged host tissues. During chronic periodontitis, the release of these components from inflamed tissues might attract T. denticola to the site of infection. As they move, T. denticola could potentially transport attached P. gingivalis to the site of inflammation, where they could cause further damage to the host tissues, leading to disease progression (Ng et al. 2016). This hypothesis is supported by a study which showed, using a capillary assay, that a non-motile bacterium, Lachnoanaerobaculum saburreum, can be translocated by piggybacking on T. denticola, with the piggyback mediated by P. gingivalis OMVs (Grenier 2013).

Figure 1.3 Proposed roles of T. denticola motility in development of P. gingivalis and T. denticola polymicrobial biofilms. T. denticola cells may utilize their chemotactic ability to move towards nutrient-rich regions and as they move, pores could be created in the biofilm matrix which then allows for the influx of nutrients and the outflow of metabolic waste. P. gingivalis could potentially “piggyback” on the motile T. denticola for its dispersal. Purple coccobacilli represent P. gingivalis while green spirals represent T. denticola.

1.5.3 Bacterial competitions in polymicrobial biofilms

Life in polymicrobial biofilms is characterized by high cell density and close proximity between different species of bacteria. For instance, dental plaque is well-known for its vast biodiversity (>700 species of bacteria) and high cell density (1011 cells per gram wet weight of plaque) (Li and Tian 2012). The bacterial cells in biofilms do not randomly stick to one another but form a well-organized multispecies community with specific spatiotemporal configurations. They are normally involved in complex social interactions

mediated by direct physical contact or secretion of small diffusible signal molecules. The interactions can occur between and within species and can be either competitive or cooperative. Cooperative interactions between different species of bacteria in biofilms, such as bacterial coaggregation, nutritional symbioses and metabolic cooperation, are well-studied and they often play important roles in the development of polymicrobial biofilms. For example, coaggregation or coadhesion allows late colonizers to colonize and adhere to an established biofilm with antecedent bacteria and subsequently increase the composition and number of participating microorganisms in the biofilm (Kolenbrander et al. 1993). Coaggregation was frequently observed among members of the oral microbiota, including P. gingivalis and T. denticola (Hashimoto et al. 2003, Ito et al. 2010, Yamada et al. 2005), and this process often involves bacterial surface proteins. Nutritional and metabolic symbioses benefit the bacteria through division of labor and reducing the bacterial energy investment which can then be used to increase cellular biomass. Nutrient sharing and metabolic symbioses were also observed between P. gingivalis and T. denticola in co-culture and polymicrobial biofilms (Zainal-Abidin et al. 2012, Tan et al. 2014).

While cooperative interactions evidently contribute to the development of polymicrobial biofilms, the role of competitive interactions is less well understood but they are believed to be equally important for the maintenance of a balanced relationship between bacteria in biofilms, thereby promoting biodiversity and homeostasis in the biofilms (Li and Tian 2012). Competitive behaviors can be broadly categorized into two forms: passive and active. In passive competition, different species of bacteria compete with one another through resource exploitation while in active competition, bacteria secrete antimicrobial compounds, such as antibiotics and bacteriocins, to eliminate their competitors. Many bacteria in dental biofilms were found to produce bacteriocins, proteinaceous toxins that, unlike traditional antibiotics, often have a narrower killing spectrum and inhibit the growth of related organisms (Kuramitsu et al. 2007). For example, bacteriocin or bacteriocin-like activities have been reported in many oral streptococcal species and oral bacteria, such as Streptococcus mutans, P. gingivalis, Prevotella intermedia, Capnocytophaga ochracea, Aggregatibacter actinomycetemcomitans, Haemophilus influenzae, Fusobacterium nucleatum, and Eikenella corrodens (Kuramitsu et al. 2007). The genome of T. denticola also encodes three potential bacteriocin secretion systems and three potential ATP binding cassette (ABC)-type bacteriocin exporter proteins

(Seshadri et al. 2004, Tanaka-Kumazawa et al. 2017), suggesting bacteriocin production by this bacterium. The production of these bacteriocins is normally tightly regulated to ensure that the bacteriocins are produced at the right time and place for optimal usage within the biofilm community. The regulation of bacteriocin production is affected by both genetic factors and environmental conditions such as cell density, nutritional availability and pH (Li and Tian 2012).

Although there is a clear advantage in targeting and eliminating bacterial competitors of different strains or species in the biofilms, some bacteria produce bacteriocins that target genetically-identical sibling cells under certain circumstances. These include the Gram- positive bacteria Bacillus subtilis which triggers the process of cannibalism during starvation (Gonzalez-Pastor et al. 2003, Ellermeier et al. 2006) and Streptococcus pneumoniae which activates fratricide to lyse non-competent cells and benefit from the released DNA (Guiral et al. 2005). Paenibacillus dendritiformis, another Gram-positive bacterium, also kills sibling cells through the production of a bacteriocin called Slf (sibling lethal factor) that is cleaved from a larger protein called DfsB (dendritiformis sibling bacteriocin). The production of Slf for sibling killing is believed to be cell-density dependent and important for the regulation of intercolony competition (Be’er et al. 2010). The advantage conferred by these sibling killing mechanisms in biofilms, especially in polymicrobial biofilms, is unclear but the programmed cell lysis of a subpopulation of the sibling cells in biofilms under specific conditions may promote the survival and fitness of the remaining sibling cells in the biofilms or even contribute to the overall fitness of the polymicrobial biofilm community by promoting biodiversity and homeostasis in the biofilm.

1.6 THESIS HYPOTHESES AND AIMS

Although T. denticola periplasmic flagella were shown to be important for the synergistic biofilm development between P. gingivalis and T. denticola, the role of T. denticola motility has not been elucidated. A routine reproducible transformation protocol for T. denticola is required in order to produce mutant strains deficient in genes associated with motility and strains containing fluorescent reporter genes.

Zones of clearing have been observed between adjacent T. denticola colonies grown on agar plates. A similar phenomenon in P. dendritiformis has also been observed where it has been shown that DfsB, the sibling killing protein is responsible for this phenomenon. The T. denticola genome encodes a homologue of DfsB, but to date, no published studies have investigated this protein.

The aims of this study are:

1) To establish an efficient and reproducible system for the transformation of T. denticola 2) To generate fluorescently-labelled P. gingivalis and T. denticola strains for investigation of biofilm development in real time 3) To generate T. denticola mutants lacking functional flagella to determine the role of T. denticola motility in synergistic biofilm formation with P. gingivalis 4) To generate a T. denticola mutant lacking a DfsB homologue to determine the role of the protein in T. denticola and in T. denticola synergistic biofilm formation with P. gingivalis

CHAPTER 2 MATERIALS AND METHODS

2.1 CHEMICALS, MEDIA, SUPPLEMENTS AND ANTIBIOTICS Analytical grade chemicals, media, supplements and antibiotics were obtained from the suppliers listed in Appendix I unless otherwise stated in text. All buffers and solutions were prepared with deionized water and autoclaved for sterility.

2.2 BACTERIAL STRAINS, PLASMIDS AND GROWTH CONDITIONS 2.2.1 Bacterial strains E. coli, P. gingivalis and T. denticola strains used in this study are listed in Table 2.1.

2.2.2 Bacterial growth conditions and media E. coli strains used in this study were grown aerobically at 37ºC in lysogeny broth (LB) [1% (w/v) tryptone, 1% (w/v) NaCl, 0.5% (w/v) yeast extract], with shaking at 200-250 revolutions per minute (rpm) or on LB agar plates [LB broth supplemented with 1.5% (w/v) Bacto-agar]. Plate and broth media were supplemented with appropriate antibiotics (Table 2.1). The plasmids generated and used in this study are listed in Table 2.1. Antibiotic was used for selection and maintenance of transformants of interest. All plasmids were amplified in the E. coli α-Select strain (Bioline Aust Pty. Ltd., Australia). E. coli cultures were stored on LB agar at 4ºC for up to a month and in LB broth mixed with 50% (v/v) glycerol for long-term storage at -80ºC.

P. gingivalis and T. denticola strains were maintained anaerobically (80% N2, 10% O2, and 10% CO2) in an anaerobic chamber (MK3 anaerobic workstation; Don Whitley Scientific Ltd., England) at 37°C. P. gingivalis strains were grown on horse blood agar [HBA; 3.7% (w/v) Blood agar base No. 2, 10% (v/v) lysed horse blood] or in brain heart infusion (BHI) broth supplemented with 5 μg/mL haemin, 0.5 mg/mL cysteine hydrochloride and appropriate antibiotics (Table 2.1). P. gingivalis colonies on plates were passaged onto fresh HBA every seven days. Only cells less than five passages were used. P. gingivalis cultures (20 mL) were prepared by inoculating supplemented BHI broth with several colonies grown on HBA and incubating overnight. P. gingivalis sub- cultures (20 or 40 mL) were prepared with inoculum from a 20 mL primary culture before overnight incubation. Optical density of batch cultures was monitored at 650 nm using a spectrophotometer (Cary 50 UV-Vis, Varian Inc. Scientific Instruments, USA). Culture purity was assessed by Gram stain and colony morphology.

T. denticola strains were grown in Oral Bacterial Growth Medium (OBGM) or on OBGM agar [OBGM supplemented with 0.8% (w/v) UltraPure™ Low Melting Point Agarose].

To make 1 L of OBGM, brain heart infusion (12.5 g), tryptone soya broth (10 g), yeast extract (7.5 g), sodium chloride (2 g), D-glucose (2 g), ascorbic acid (2 g), sodium pyruvate (1 g), asparagine (0.25 g) and sodium thioglycolate (0.5 g) were added to 800 mL of water, mixed and autoclaved. After autoclaving, the solution was supplemented with 200 mL of filter-sterilized solutions containing cysteine hydrochloride (1 g), hemin (5 mg), menadione (1 mg), ammonium sulfate (2 g), thiamine pyrophosphate (6 mg), sodium hydrogen carbonate (2 g), filtered heat-inactivated rabbit serum [added to 2% (v/v)] and volatile fatty acids mix added to 0.5% (v/v) (0.1 M potassium hydroxide containing 0.5% v/v of isobutyric acid, DL-2-methybutyric acid, isovaleric acid and valeric acid). The pH of the medium was adjusted to 7.4. T. denticola cultures were passaged into fresh OBGM every seven days. Only cells less than four passages were used.

2.2.3 Bacterial enumeration

P. gingivalis cell numbers in BHI were determined spectrophotometrically where A650 of 0.6, which is within the exponential phase of growth, is equivalent to approximately 2.5 9 × 10 cfu/mL. Monoculture bacterial cell density was determined by correlating the A650 to the A650 versus cell number standard curve derived by Orth et al. (2010). The equation 9 8 used to determine the number of P. gingivalis cells in OBGM is: 2 × 10 × A650 + 2 × 10

(effective A650: 0.1 to 1.0). The equation used to determine the number of T. denticola 9 7 cells in OBGM is: 3 × 10 × A650 + 9 × 10 (effective A650: 0.1 to 0.25).

Table 2.1 Strains and plasmids.

Bacterial Plasmid DescriptionB Antibiotic resistance Reference strainA selectionC, D, E E. coli α-Select NAF F- deoR endA1 recA1 relA1 gyrA96 hsdR17(rk-, mk+) NAF Bioline Aust Pty. supE44 thi-1 phoA ∆(lacZYA argF)U169 Φ80lacZ∆M15λ- Ltd., Australia ECR535 pCF382 [Kanr, Ermr]; pSTBlue-1::SOE insert for pyrF knockout. Kan, Erm (Capone et al. 2007) NovaBlue host ECR539 pCF382 [Kanr, Ermr]; pSTBlue-1::SOE insert for pyrF knockout. Kan, Erm (Capone et al. 2007); SCS110 host This laboratory ECR761 pHN-911 [Ampr, Ermr]; pGEM®-T Easy::SOE insert for TDE0911 Amp, Erm This study knockout. α-Select host ECR762 pHN-PgEvo [Ampr, Ermr]; pGEM®-T Easy::SOE insert with evoglow- Amp, Erm This study Bs2-stop for insertion into mfaI locus. α-Select host ECR763 pHN-TdEvo [Ampr, Kanr]; pGEM®-T Easy::SOE insert with evoglow- Amp, Kan This study Bs2-stop for replacement of TDE0911. α-Select host ECR764 pHN- [Ampr, Kanr]; pGEM®-T Easy::SOE insert with evoglow- Amp, Kan This study TdclosEvo C-Bs2-stop for replacement of TDE0911. α-Select host ECR765 pHN-motA [Ampr, Ermr]; pGEM®-T Easy::SOE insert for motA Amp, Erm This study knockout. α-Select host

ECR766 pHN-motB [Ampr, Ermr]; pGEM®-T Easy::SOE insert for motB Amp, Erm This study knockout. α-Select host ECR767 pHN-cheY [Ampr, Ermr]; pGEM®-T Easy::SOE insert for cheY Amp, Erm This study knockout. α-Select host ECR768 pHN-flgE [Ampr, Ermr]; pGEM®-T Easy::SOE insert for flgE Amp, Erm This study knockout. α-Select host ECR769 pHN- [Ampr, Kanr]; pGEM®-T Easy::SOE insert for Amp, Kan This study 911kan TDE0911kan knockout. α-Select host ECR770 pHN-0659 [Ampr, Ermr]; pGEM®-T Easy::SOE insert for TDE0659 Amp, Erm This study knockout. α-Select host NAF pGEM®-T Ampr; T-Overhang cloning plasmid NAF Promega, USA Easy P. gingivalis W50 NAF Wild-type NAF Kind gift from Professor Philip Marsh (University of Leeds, UK) ECR771 NAF [Ermr]; P. gingivalis W50 pHN-PgEvo allele exchange Erm This study mutant with evoglow-Bs2-stop in mfaI, W50::FbFP

∆ABK NAF [Tcr, Camr, Ermr]; P. gingivalis W50 allele exchange Tc, Cam, Erm This laboratory mutant with kgp ORF deleted and replaced by ermF in W50 (Dashper et al. 2004) ∆AB, ∆rgpArgpBkgp T. denticola HL51 NAF [Ermr]; T. denticola ATCC 35405 ∆flgE mutant with flgE Erm (Li et al. 1996) gene fragment insertionally inactivated with ermFAM gene ATCC NAF Wild-type NAF Purchased from 35405 American Type Culture Collection (ATCC), USA ATCC NAF Wild-type NAF Purchased from 33520 ATCC ECR772 NAF [Ermr]; T. denticola ATCC 33520 pCF382 allele exchange Erm This study mutant, ∆pyrF ECR773 NAF [Ermr]; T. denticola ATCC 35405 pHN-911 allele exchange Erm This study mutant, ∆TDE0911 ECR774 NAF [Kanr]; T. denticola ATCC 35405 pHN-TdEvo allele Kan This study exchange mutant with evoglow-Bs2-stop replacing

TDE0911, ATCC 35405::FbFPevoglow-Bs2-stop

ECR775 NAF [Kanr]; T. denticola ATCC 35405 pHN-TdclosEvo allele Kan This study exchange mutant with evoglow-C-Bs2-stop replacing

TDE0911, ATCC 35405::FbFPevoglow-C-Bs2-stop ECR823 NAF [Ermr]; T. denticola ATCC 35405 pHN-cheY allele Erm This study

exchange mutant, ∆cheY35405 ECR824 NAF [Kanr]; T. denticola ATCC 35405 pHN-911kan allele Kan This study exchange mutant, ∆TDE0911::kan ECR825 NAF [Kanr, Ermr]; T. denticola ATCC 35405 ∆TDE0911::kan Kan, Erm This study pHN-cheY allele exchange mutant, ∆TDE0911::kan ∆cheY ECR826 NAF [Kanr, Ermr]; T. denticola ATCC 35405 ∆TDE0911::kan Kan, Erm This study pHN-flgE allele exchange mutant, ∆TDE0911::kan ∆flgE ECR827 NAF [Ermr]; T. denticola ATCC 33520 pHN-motA allele Erm This study

exchange mutant, ∆motA33520 ECR828 NAF [Ermr]; T. denticola ATCC 33520 pHN-motB allele Erm This study

exchange mutant, ∆motB33520 ECR829 NAF [Ermr]; T. denticola ATCC 33520 pHN-flgE allele Erm This study

exchange mutant, ∆flgE33520 ECR830 NAF [Ermr]; T. denticola ATCC 33520 pHN-cheY allele Erm This study

exchange mutant, ∆cheY33520

ECR831 NAF [Ermr]; T. denticola ATCC 33520 pHN-0659 allele Erm This study

exchange mutant, ∆DfsBTd33520 A Freeze-dried and glycerol stock cultures of E. coli, P. gingivalis W50, T. denticola ATCC 33520 and T. denticola ATCC 35405 strains were obtained from the culture collection of the Cooperative Research Centre for Oral Health Science, The University of Melbourne, Australia. Additional strains were generated in the course of this study. B Abbrevation for plasmids: Ampr, ampicillin resistance; Ermr, erythromycin resistance; Kanr, kanamycin resistance, Tcr, tetracycline resistance, Camr, chloramphenicol resistance; :: represents insertion of gene; ∆ represents deletion of gene; [ ] brackets indicate phenotype. C Antibiotic resistance selection in E. coli strains with resistance genes: 100 μg/mL ampicillin for Ampr, 150 μg/mL erythromycin for Ermr or 50 μg/mL kanamycin for Kanr. D Antibiotic resistance selection in P. gingivalis strains with resistance genes: 10 μg/mL erythromycin for Ermr, 10 μg/mL tetracycline for Tcr or 10 μg/mL chloramphenicol for Camr. E Antibiotic resistance selection in T. denticola strains with resistance genes: 40 μg/mL erythromycin for Ermr or 75 μg/mL kanamycin for Kanr. F NA: Not applicable.

2.3 CENTRIFUGATION Centrifugation of cells and subcellular fractions were done using a variety of rotors and centrifuges including: a JA-12 rotor (Beckman Coulter™ Inc.) with or without 15 mL tube adapters or a FiberLite® F10 rotor (Thermo Fisher Scientific Inc., USA) in an Avanti® J-25I high performance centrifuge (Beckman Coulter™ Inc.); a C0650 rotor (Beckman Coulter™ Inc.) in an Allegra® X-30R refrigerated benchtop centrifuge (Beckman Coulter™ Inc.).

2.4 EXTRACTION OF GENOMIC AND PLASMID DNA FROM CELLS 2.4.1 Genomic DNA isolation Genomic DNA was purified using the DNeasy® Blood and Tissue kit (QIAGEN, Clifton Hill, VIC, Australia) according to the manufacturer’s instructions. Genomic DNA was resuspended in deionized water or 5 × diluted AE buffer (QIAGEN). DNA concentration was determined spectrophotometrically at a wavelength of 260 nm using a NanoDrop™ Spectrophotometer ND-1000 (Thermo Fisher Scientific Inc., USA) or fluorometrically using a Qubit 1.0 Fluorometer (Thermo Fisher Scientific Inc., USA).

2.4.2 Plasmid isolation Plasmid DNA was purified from E. coli strains using commercial kits, the QIAprep® Spin Miniprep kit (QIAGEN) and the PureLink® HiPure Plasmid Maxiprep Kit (Invitrogen, USA) according to the manufacturer’s instructions and stored in deionized water at -20°C. DNA concentration was determined spectrophotometrically at a wavelength of 260 nm using a NanoDrop™ Spectrophotometer ND-1000.

2.5 MANIPULATION OF DNA 2.5.1 Restriction enzyme digestions DNA was digested with various restriction enzymes (RE) following the manufacturer’s instructions (New England Biolabs Inc., USA) (Table 2.2). Progress of digestion was monitored via agarose gel electrophoresis. Digestion was terminated by heat-inactivation at 65°C or 80°C for 20 min where appropriate. The enzyme was removed using DNA precipitation or the QIAquick Gel Extraction Kit (QIAGEN) according to the manufacturer’s instructions.

Table 2.2 Restriction enzymes used in the course of this study.

Restriction Restriction site Reaction conditions enzyme Buffer Temperature Incubation (°C) time (h)

EcoRI G↓AATTC NEB buffer 2.1 37 2

CTTAA↑G

BsaBI GATNN↓NNATC NEB buffer 2.1 60 2

CTANN↑NNTAG

PstI CTGCA↓G NEB buffer 3.1 37 2

G↑ACGTC

NotI GC↓GGCCGC CutSmart® Buffer 37 3

CGCCGG↑CG

2.5.2 DNA precipitation DNA precipitation was carried out by adding 0.1 × sample volume of 3 M sodium acetate (pH 5.2) and 3 × sample volume of ice-cold ethanol to the DNA solution, mixed well and incubated at -80°C for 30 min. DNA was pelleted by centrifugation at 13 000 rpm for 30 min at 4°C. The supernatant was removed. Ethanol [1 mL of 70% (v/v)] was added to wash the DNA pellet. The ethanol was discarded after centrifugation at 13 000 rpm for 15 min at 4°C. The ethanol wash step was repeated before the DNA pellet was air-dried and suspended in an appropriate volume of deionized water.

2.5.3 Agarose gel electrophoresis Electrophoresis was carried out to separate DNA fragments using 0.8% (w/v) agarose gels prepared with Tris-acetate buffer (TAE; 40 mM Tris-acetate, 1 mM EDTA, pH 8) containing SYBR SafeTM gel stain [0.004% (v/v)]. One-fifth volume of 5 × DNA Loading Buffer Blue (Bioline) was mixed with DNA samples prior to electrophoresis. Molecular markers [3.5-10 µL of HyperLadder™ 1kb (Bioline)] were used for DNA sizing. Samples underwent electrophoresis at 80-100 V for an appropriate amount of time. DNA bands

were visualized by fluorescence using the Fujifilm LAS-3000 Imager (Fujifilm, Berthold Australia Pty. Ltd., Australia).

2.5.4 Ligation Ligation reaction (10 μL) containing 200 U of T4 DNA ligase (New England Biolabs

Inc., USA) and 1 × T4 DNA Ligase Reaction Buffer [50 mM Tris-HCl, 10 mM MgCl2, 1 mM ATP, 10 mM DTT, pH 7.5] was incubated at 14°C overnight. The enzyme was inactivated by incubation at 65°C for 20 min.

A vector:insert molar ratio of 1:6 was used and the amount of insert needed was calculated according to the formula below (provided by Promega Corporation):

ng of vector × kb size of insert ng of insert = × vector:insert molar ratio kb size of vector

2.5.5 Genomic PCR amplification PCR reactions were performed using standard conditions with annealing and extension temperatures as appropriate for each reaction (Table 2.3). Salt adjusted annealing temperatures were determined using OligoCalc (Kibbe 2007). Lyophilized oligonucleotide primers (GeneWorks, Australia) were suspended in deionized water to a concentration of 10 µM and used at 400 nM final concentration. Purified T. denticola ATCC 33520, ATCC 35405 or P. gingivalis W50 genomic DNA (50-100 ng) was amplified using the appropriate primer pairs (Table 2.4) and Herculase II Fusion DNA Polymerases (Agilent Technologies, United States) according to the manufacturer’s instructions. Thermal cycling was carried out using a G-Storm GS1 thermal cycler (GeneWorks Pty. Ltd., Australia) according to the conditions in Table 2.3. The size and purity of PCR amplicons were examined using agarose gel electrophoresis and purified using the QIAquick Gel Extraction Kit as per manufacturer’s instructions.

2.5.6 Colony PCR amplification Colony PCR was used to screen for plasmids containing inserts in E. coli transformants. Single E. coli colonies were picked using sterile pipette tips, patched onto an LB plate with the appropriate antibiotic and mixed into individual tubes containing PCR reaction mix [20 µL; 200 nM of M13 Forward and M13 Reverse primers (Table 2.4, 1 × MyTaq Red Reaction Buffer, 0.3 units of MyTaq DNA polymerase (Bioline) and sterile deionized water]. For P. gingivalis, since the cells are inhibitory to PCR, a sample of purified genomic DNA (10-30 ng) was used as a template. For T. denticola, a small volume of

culture was used as the template. Thermal cycling was carried out using a G-Storm GS1 thermal cycler according to the conditions in Table 2.3 with the exception of a longer initial denaturation step (up to 5 min) in order to lyse the cells. PCR amplicons were examined using agarose gel electrophoresis.

Table 2.3 Standard PCR amplification conditions.

Step Temperature (°C) Time Cycles

Initial denaturation 95 1-2 min 1

Denaturation 95 15 s

Annealing 55-60 15-20 s 30-35

Extension 72 30-45 s

Final extension 72 3 min 1

2.5.7 Splicing by overlap extension (SOE) PCR The SOE PCR method (Horton et al. 1989) was used to construct allele exchange suicide plasmids for homologous recombination with the P. gingivalis and T. denticola chromosome and produced various mutant strains of P. gingivalis and T. denticola. A schematic overview of the procedure is shown in Figure 2.1. Briefly, fragments that were to be spliced were generated in separate polymerase chain reactions. The PCR products have short complementary sequences that were annealed with an initial five cycles of amplification without primers prior to a further round of PCR with primers that produced a spliced amplicon. Specific details are described in the following sections. PCR amplification was carried out using Herculase II Fusion DNA Polymerases or Platinum™ PCR SuperMix High Fidelity (Thermo Fisher Scientific Inc., USA) according to the manufacturer’s instructions. Thermal cycling was carried out using a G-Storm GS1 thermal cycler. The size and purity of PCR amplicons were examined using agarose gel electrophoresis. PCR amplicons and DNA in agarose gels were purified and extracted using the QIAquick Gel Extraction Kit according to the manufacturer’s instructions.

2.6 NUCLEOTIDE SEQUENCING Appropriate amounts of plasmid DNA along with 5 ρmol of primers (Table 2.4) were sent for nucleotide sequencing at Applied Genetic Diagnostics, Department of Pathology, University of Melbourne (Melbourne, Australia). Nucleotide sequence data and chromatogram files were analyzed using Chromas (Technelysium Pty. Ltd.).

Table 2.4 Oligonucleotide primers used for PCR amplification and nucleotide sequencing.

Primer Sequence (5’ → 3’) Location of namek nucleotide mfaI5’-F ATGCCTCAAGAAGAAACTTTTAGG 200458- 200481a, i mfaI5’-R TTCAATATAAGCGACAAGGGATAATCATAATACTC 201117- GGTAGCATTTG 201092a, i, j Kgpprom-F CTACCGAGTATTATGATTATCCCTTGTCGCTTATA 1939571- TTGAAAACATG 1939596a, i, j Kgpprom-R AACGACTGGAACGACGCCATACTTTAAAACAATTT 1939820- ATGGTCGTGATTC 1939793a, i, j Evoglow-F ACCATAAATTGTTTTAAAGTATGGCGTCGTTCCAG 830-810c, i, j TCGTTC Evoglow-R AGCAATAGCGGAAGCTATCGTCACTCGAGCAGCTT 440-417c, i, j TTCATATTC ermFAM-F ATGAAAAGCTGCTCGAGTGACGATAGCTTCCGCTA 9-30e, i, j TTGCTTT ermFAM-R TGATCTTCCCACAGACCGGCAGCTGTCAGTAGTAT 2151-2126e, i, j ACCTAATAATT mfaI3’-F TAGGTATACTACTGACAGCTGCCGGTCTGTGGGAA 201118- GATC 201136a, i, j mfaI3’-R GAGATCAACCTCATAGGAATGAAC 201840- 201817a, i tde911-F CGGCTGTCAGACTACAAATG 929885- 929904b, i

tde911-R TTGTTCATATTAGATAGTGACATTTTGTATC 930627- 930605b, i, j tde911-3’F GAGGAAATAAAAAAAATCTGGATTTGTGTATGT 931390- 931412b, i, j tde911-3’R CACACGTAAAACAGATGAC 932074- 932056b, i ermAM-F CACTATCTAATATGAACAAAAATATAAAATATTCT 3189-3213f, i, j C ermAM-R CCAGATTTTTTTTATTTCCTCCCGTTAAATAATAG 3926-3903f, i, j tde911-5’R- TCTTTACTGAGCCCTGTCATATTAGATAGTGACAT 930627- Msp TTTGTATCTCCTC 930600b, i, j Msp prom-F ACAAAATGTCACTATCTAATATGACAGGGCTCAGT 455389- AAAGAAG 455410b, i, j Msp prom-R AACGACTGGAACGACGCCATAAAAAATTCCTCCTT 455609- GTTATTTGTTTG 455583b, i, j Evo-F-Msp ATAACAAGGAGGAATTTTTTATGGCGTCGTTCCAG 830-810c, i, j TCGTTC Evo-R-Kan GCTGGCAATTCCGGTTCGCTTCACTCGAGCAGCTT 440-417c, i, j TTCATATTC Kan-F ATGAAAAGCTGCTCGAGTGAAGCGAACCGGAATTG 1015-1034g, i, CCAGC j Kan-R ATACATACACAAATCCAGATTCAGAAGAACTCGTC 1953-1933g, i, AAGAAG j tde911-3’F- TTCTTGACGAGTTCTTCTGAATCTGGATTTGTGTA 931395- Kan TGTATGGTAG 931419b, i, j tde911-3’R- TCTATACACACGTAAAACAGATGAC 932080- long 932056b, i Msp prom- TGAAAACTTGCGGATCCCATAAAAAATTCCTCCTT 455609- closEvo-R GTTATTTGTTTG 455583b, i, j closEvo-Msp- ATAACAAGGAGGAATTTTTTATGGGATCCGCAAGT 455-488d, i, j F TTTCAAAGTTTTGGAATAC closEvo-Kan- GCTGGCAATTCCGGTTCGCTTTATTCAAGAAGCTT 851-874d, i, j R TTCATATTC

Kan-closEvo- ATGAAAAGCTTCTTGAATAAAGCGAACCGGAATTG 1015-1034g, i, F CCAGC j tde9115’- CAATTTGAGCATTAAGGTTTATTAGATAGTGACAT 930627- ermAM-R TTTGTATCTCCTC 930600b, i, j ermAM ACAAAATGTCACTATCTAATAAACCTTAATGCTCA 994-1021e, i, j prom-F AATTGTTTGTTTG ermAM AATCCATCTTGTTCAATCATGTAATCACTCCTTCT 1301-1328e, i, j prom-R TAATTACAAATTT Kan-ermAM- AATTAAGAAGGAGTGATTACATGATTGAACAAGAT 1159-1182g, i, F GGATTGCAC j TDE2767-F GTCTCCCGCAGGGCAAG 2821755- 2821739b, i TDE2767-R CCTCTAGAGTCGACCTGCAGAAACGACGCTATATC 2820947- CATATCTAC 2820970b, i, j ermAM- ATATGGATATAGCGTCGTTTCTGCAGGTCGACTCT 2179-2161e, i, j 2767-F AGAG ermAM-fla GAGCAGCAAAAATAATTTGCAAACCTTAATGCTCA 994-1021e, i, j prom-R AATTGTTTGTTTG fla prom- CAATTTGAGCATTAAGGTTTGCAAATTATTTTTGC 2824837- ermAM-F TGCTCTCAC 2824814b, i, j fla prom- TGTTTCTTTTTCCTTGCCATATGAACCTCCATAAA 2824508- 2765-R AACTTTTTGCAG 2824534b, i, j TDE2765-F AAGTTTTTATGGAGGTTCATATGGCAAGGAAAAAG 2820183- AAACACGGTC 2820159b, i, j TDE2765-R GGGTATAAGACCTGCACCTC 2819329- 2819348b, i TDE2766-F AGGAATATTCGGAGGTATAGCAG 2820944- 2820922b, i TDE2766-R CCTCTAGAGTCGACCTGCAGCCTTGCCATTTTAAT 2820175- CCTTTAATAC 2820199b, i, j

ermAM- AAAGGATTAAAATGGCAAGGCTGCAGGTCGACTCT 2179-2161e, i, j 2766-F AGAG fla prom- ATTAAGTCATTATCAGCCATATGAACCTCCATAAA 2824508- 2764-R AACTTTTTGCAG 2824534b, i, j TDE2764-F AAGTTTTTATGGAGGTTCATATGGCTGATAATGAC 2819421- TTAATGGATG 2819397b, i, j TDE2764-R TCTGTATCCGTATCTCCTGAG 2818788- 2818808b, i TDE1493-F TTGAGTACCATGATTGTACTCATC 1538473- 1538496b, i TDE1493-R CCTCTAGAGTCGACCTGCAGGTCCTGTTTCGATAT 1539233- CATAATAAAAC 1539208b, i, j ermAM- TTATGATATCGAAACAGGACCTGCAGGTCGACTCT 2179-2161e, i, j 1493-F AGAG ermAM- TAGGAAATACCTAGGACACCAAACCTTAATGCTCA 994-1021e, i, j 1495-R AATTGTTTGTTTG TDE1495-F CAATTTGAGCATTAAGGTTTGGTGTCCTAGGTATT 1539636- TCCTAATC 1539658b, i, j TDE1495-R CATAGCCCGAACATCCTTTTAAC 1540427- 1540405b, i TDE2768-F TGTTGCCTCTAAAGAAGCCTAC 2823380- 2823401b, i TDE2768-R CCTCTAGAGTCGACCTGCAGTAATTATTGCCTCCT 2822600- AATTGTTATCTG 2822626b, i, j ermAM- CAATTAGGAGGCAATAATTACTGCAGGTCGACTCT 2179-2161e, i, j 2768-F AGAG fla prom- AGCCGCGTTACCTGTATCATATGAACCTCCATAAA 2824508- 2767-R AACTTTTTGCAG 2824534b, i, j TDE2767-F AAGTTTTTATGGAGGTTCATATGATACAGGTAACG 2821160- (flgE KO) CGGCTAAAC 2821183b, i, j TDE2767-R AGCATTTACTAAGGATATCCATGTG 2820524- (flgE KO) 2820548b, i

TDE0660-F CCTTTCTCCAATAGATGAGATAC 696995- 697017b, i TDE0660-R CCTCTAGAGTCGACCTGCAGAGGTTTCTCCCATAA 696293- AAATATGTAAAC 696319b, i, j ermAM- TATTTTTATGGGAGAAACCTCTGCAGGTCGACTCT 2179-2161e, i, j TDE0660-F AGAG ermAM- CAGCCCTTGTTTGATACTATAAACCTTAATGCTCA 994-1021e, i, j TDE0658-R AATTGTTTGTTTG TDE0658-F CAATTTGAGCATTAAGGTTTATAGTATCAAACAAG 695761- GGCTGTTC 695783b, i, j TDE0658-R ACTCGCCCTATATTCCGTATTC 695060- 695081b, i FORWARD TCCGTTTTTCCGTTTTTGTC 2133719- TestpyrF 2133738b REVERSE AGCCTTCCCGTATCTTGGTT 2137570- TestpyrF 2137551b Pg0177-F GGATCTGATGCAATGGCTTAATG 200318- 200340a Pg0179-R ATAAGTCTGCGAATATCCATATGC 201960- 201937a tde910-F TGGCTCAACTCCTGTGAATATG 929720- 929741b tde913-R AATGAACATTTATCGCTGCCTTG 932263- 932241b ermAM-F ATGAACAAAAATATAAAATATTCTCAAAAC 3189-3218f short ermAM-R TTATTTCCTCCCGTTAAATAATAGATAAC 3926-3898f short 5’motA-F CTTCGGTGAAGCCCACGAG 2821971- 2821953b 3’motA-R CTAAGGCGACATCTACTATCAAG 2819097- 2819119b

5’motB-F TATTGAGTGTAATCCGGATGTTAC 2821121- 2821098b 3’motB-R TACGTAAGCTGTTCTACGGTTG 2818593- 2818614b 5’cheY-F TTCGAATGCTCCCATGCTTTC 1538351- 1538371b 3’cheY-R TCCGCAGCCGTTGGGTC 1540532- 1540516b 5’flgE-F GATGCTTCAGCAGACTTTGTAC 2823532- 2823553b 3’flgE-R CATGTTAGGGCCGAGAGCA 2820422- 2820440b 5’TDE0659-F TCCTTCTCAGTCGGAAGCAG 697221- 697240b 3’TDE0659- AGATTATGTTGTGCAGCTTTGAAG 694853- R 694876b M13 Forward GTAAAACGACGGCCAGTG 2979-2996h M13 Reverse GGAAACAGCTATGACCATGA 190-171h a Location of nucleotides in P. gingivalis W83 genome; Genbank accession number NC_002950.2. b Location of nucleotides in T. denticola ATCC 35405 genome; Genbank accession number NC_002967.9. c Location of nucleotides in pGLOW-Bs2-stop (Jena Bioscience, Jena, Germany). d Location of nucleotides in pGLOW-CKXN-Bs2 (Jena Bioscience, Jena, Germany). e Location of nucleotides in shuttle vector pHS17 ermF and ermAM; Genbank accession number AF219231. f Location of nucleotides in pCF382 (Capone et al. 2007). g Location of nucleotides in pCR4-TOPO; vector sequence (Thermo Fisher Scientific Inc., USA). h Location of nucleotides in pGEM®-Teasy; vector sequence (Promega, USA). i Oligonucleotide primers used for SOE PCR. j Overlapping base pairs were underlined. k All oligonucleotide primers were obtained from GeneWorks Pty. Ltd., Australia.

Table 2.5 Primer pairs for generating SOE fragments.

Plasmid Reaction Fragment Primer pairsa pHN-PgEvo PCR-1.1 5’ mfaI mfaI5’-F, mfaI5’-R PCR-1.2 kgp promoter Kgpprom-F, Kgpprom-R PCR-1.3 evoglow-Bs2-stop Evoglow-F, Evoglow-R PCR-1.4 ermFAM ermFAM-F, ermFAM-R PCR-1.5 3’ mfaI mfaI3’-F, mfaI3’-R SOE-1.1 5’ mfaI-kgp promoter mfaI5’-F, Kgpprom-R SOE-1.2 ermFAM-3’ mfaI ermFAM-F, mfaI3’-R SOE-1.3 5’ mfaI-kgp promoter- evoglow-Bs2-stop mfaI5’-F, Evoglow-R SOE-1.4 mfaI::evoglow::ermFAM mfaI5’-F, mfaI3’-R pHN-911 PCR-2.1 5’ TDE0911 tde911-F, tde911-R PCR-2.2 ermAM ermAM-F, ermAM-R PCR-2.3 3’ TDE0911 tde911-3’F, tde911-3’R SOE-2.1 ermAM-3’ TDE0911 ermAM-F, tde911-3’R SOE-2.2 TDE0911::ermAM tde911-F, tde911-3’R pHN-TdEvo PCR-3.1 5’ TDE0911 tde911-F, tde911-5’R-Msp PCR-3.2 msp promoter Msp prom-F, Msp prom-R PCR-3.3 evoglow-Bs2-stop Evo-F-Msp, Evo-R-Kan

PCR-3.4 aphA2 Kan-F, Kan-R PCR-3.5 3’ TDE0911 tde911-3’F-Kan, tde911-3’R-long SOE-3.1 5’ TDE0911-msp promoter tde911-F, Msp prom-R SOE-3.2 aphA2-3’ TDE0911 Kan-F, tde911-3’R-long SOE-3.3 evoglow-Bs2-stop-aphA2-3’ TDE0911 Evo-F-Msp, tde911-3’R-long SOE-3.4 TDE0911::evoglow::aphA2 tde911-F, tde911-3’R-long pHN- PCR-4.1 5’ TDE0911 tde911-F, tde911-5’R-Msp TdclosEvo PCR-4.2 msp promoter Msp prom-F, Msp prom-closEvo-R PCR-4.3 evoglow-C-Bs2-stop closEvo-Msp-F, closEvo-Kan-R PCR-4.4 aphA2 Kan-closEvo-F, Kan-R PCR-4.5 3’ TDE0911 tde911-3’F-Kan, tde911-3’R-long SOE-4.1 5’ TDE0911-msp promoter tde911-F, Msp prom-closEvo-R SOE-4.2 aphA2-3’ TDE0911 Kan-closEvo-F, tde911-3’R-long SOE-4.3 5’ TDE0911-msp promoter- evoglow-C-Bs2-stop tde911-F, closEvo-Kan-R SOE-4.4 TDE0911::clos-evoglow::aphA2 tde911-F, tde911-3’R-long pHN-motA PCR-5.1 5’ motA TDE2767-F, TDE2767-R PCR-5.2 ermAM ermAM-2767-F, ermAM-fla prom-R PCR-5.3 fla promoter fla prom-ermAM-F, fla prom-2765-R PCR-5.4 3’ motA TDE2765-F, TDE2765-R SOE-5.1 5’ motA-ermAM TDE2767-F, ermAM-fla prom-R

SOE-5.2 fla promoter-3’ motA fla prom-ermAM-F, TDE2765-R SOE-5.3 motA::ermAM TDE2767-F, TDE2765-R pHN-motB PCR-6.1 5’ motB TDE2766-F, TDE2766-R PCR-6.2 ermAM ermAM-2766-F, ermAM-fla prom-R PCR-6.3 fla promoter fla prom-ermAM-F, fla prom-2764-R PCR-6.4 3’ motB TDE2764-F, TDE2764-R SOE-6.1 5’ motB-ermAM TDE2766-F, ermAM-fla prom-R SOE-6.2 fla promoter-3’ motB fla prom-ermAM-F, TDE2764-R SOE-6.3 motB::ermAM TDE2766-F, TDE2764-R pHN-cheY PCR-7.1 5’ cheY TDE1493-F, TDE1493-R PCR-7.2 ermAM ermAM-1493-F, ermAM-1495-R PCR-7.3 3’ cheY TDE1495-F, TDE1495-R SOE-7.1 5’ cheY-ermAM TDE1493-F, ermAM-1495-R SOE-7.2 cheY::ermAM TDE1493-F, TDE1495-R pHN-flgE PCR-8.1 5’ flgE TDE2768-F, TDE2768-R PCR-8.2 ermAM ermAM-2768-F, ermAM-fla prom-R PCR-8.3 fla promoter fla prom-ermAM-F, fla prom-2767-R PCR-8.4 3’ flgE TDE2767-F (flgE KO), TDE2767-R (flgE KO) SOE-8.1 5’ flgE-ermAM TDE2768-F, ermAM-fla prom-R SOE-8.2 fla promoter-3’ flgE fla prom-ermAM-F, TDE2767-R (flgE KO)

SOE-8.3 flgE::ermAM TDE2768-F, TDE2767-R (flgE KO) pHN-0659 PCR-9.1 5’ TDE0659 TDE0660-F, TDE0660-R PCR-9.2 ermAM ermAM-TDE0660-F, ermAM-TDE0658-R PCR-9.3 3’ TDE0659 TDE0658-F, TDE0658-R SOE-9.1 5’ TDE0659-ermAM TDE0660-F, ermAM-TDE0658-R SOE-9.2 TDE0659::ermAM TDE0660-F, TDE0658-R pHN-911kan PCR-10.1 5’ TDE0911 tde911-F, tde9115’-ermAM-R PCR-10.2 ermAM promoter ermAM prom-F, ermAM prom-R PCR-10.3 aphA2 Kan-ermAM-F, Kan-R PCR-10.4 3’ TDE0911 tde911-3’F-Kan, tde911-3’R-long SOE-10.1 5’ TDE0911-ermAM promoter tde911-F, ermAM prom-R SOE-10.2 aphA2-3’ TDE0911 Kan-ermAM-F, tde911-3’R-long SOE-10.3 TDE0911::aphA2 tde911-F, tde911-3’R-long a Primer sequences listed in Table 2.4.

Figure 2.1 Schematic overview of splicing by overlap extension (SOE) PCR method. The double stranded DNA and primers are represented by solid colour lines. Primers are denoted by lower-case letters and PCR products are denoted by pairs of capital letters corresponding to the primer pairs used to generate that product. The primers consist of two parts, one is required for initial annealing to the DNA template (same colour as the DNA template) while the other is overlapping sequence required for the initial annealing of two PCR products in step 2. Step 2 represents the initial five rounds of amplification without primers where the denatured fragments anneal at the overlapping sequence and are extended 3’ by Herculase II Fusion DNA polymerase (dashed line) to form a mutant SOE product. By adding additional primers ‘a’ and ‘d’ at Step 3 the mutant SOE product is further amplified by PCR (Horton et al. 1989).

2.7 TRANSFORMATION OF BACTERIA E. coli α-Select gold efficiency (50 µL; Bioline) commercial chemically competent cells were transformed using heat-shock with ligation reaction products and plasmid DNA as per manufacturer’s instructions. Transformants were selected on LBA containing appropriate antibiotics (Table 2.1) and incubated at 37°C overnight or at RT for >48 h. Colony PCR was used to screen for the presence of the appropriate insert in the resulting putative recombinant plasmids. Plasmid DNA was then purified and sequenced with the appropriate oligonucleotide primers (Table 2.4).

Electroporation of P. gingivalis cells with a linearized suicide plasmid was performed as previously described (Slakeski et al. 2011). Transformants were selected on HBA plates containing the appropriate antibiotics (Table 2.1) and incubated anaerobically at 37°C. To screen for the presence of the appropriate homologous recombination event, gDNA was purified from the resulting transformants and used in standard PCR with the appropriate oligonucleotide primers (Table 2.4).

Electroporation of T. denticola cells with a linearized suicide plasmid was performed as previously published (Li et al. 1996) with some modifications. Briefly, two-day

(exponential phase; A650 of 0.10-0.16 for ATCC 35405 and ∆TDE0911::kan; A650 of 0.16- 0.28 for ATCC 33520) T. denticola cultures were decanted into centrifuge tubes in the anaerobic chamber. The cells were harvested by centrifugation (4,000 g, 4°C) and the subsequent washing and resuspension steps were conducted at 4°C. All resuspension steps were performed using 1 mL pipette tips which had approximately 1 cm cut off to reduce shear forces on cells. The cells were washed two times instead of three times to reduce handling and loss of cells. Sterile 10% (v/v) glycerol used for all wash and resuspension steps was pre-reduced in the anaerobic chamber for at least 16 h before use. Electroporation was carried out as previously published (Li et al. 1996), typically producing a time constant of 4.0 to 4.7 ms. T. denticola cells were then immediately suspended in 1.2 mL of pre-reduced OBGM in the anaerobic chamber and incubated overnight at 37°C. Transformants were selected on OBGM agar plates containing the appropriate antibiotics (Table 2.1) and incubated anaerobically at 37°C. To screen for the presence of the appropriate homologous recombination event, the resulting transformants were grown in OBGM (2 mL) containing the appropriate antibiotics whereupon a small volume of culture (~ 25 µL) was used in colony PCR with the appropriate oligonucleotide primers (Table 2.4).

2.8 GENERATION OF RECOMBINANT STRAINS 2.8.1 Construction of P. gingivalis W50::FbFP strain (ECR771) Plasmid pHN-PgEvo was generated by SOE PCR. An overview of the SOE PCR method is shown in Figure 2.1 and the oligonucleotides used are as shown in Table 2.4. The primer pairs required to generate the corresponding amplicons are listed in Table 2.5. To construct pHN-PgEvo, the 5’ and 3’ flanking regions of mfaI (5’ mfaI and 3’ mfaI), kgp promoter, evoglow-Bs2-stop and ermFAM were amplified by PCR with five pairs of primers as listed in Table 2.5. The primers mfaI5’-R, Kgpprom-F, Kgpprom-R, Evoglow- F, Evoglow-R, ermFAM-F, ermFAM-R and mfaI3’-F contain overlapping base pairs (underlined in Table 2.4). In the second step, 5’ mfaI and kgp promoter were fused by PCR using mfaI5’-F and Kgpprom-R primers while ermFAM and 3’ mfaI were fused by PCR using ermFAM-F and mfaI3’-R primers. Herculase II Fusion DNA polymerase was used to extend the template sequences for five rounds of amplification before the respective primers were added to the PCR reaction tube and amplification continued for a further 30 cycles (Step 2, Figure 2.1). In the third step, the constructed 5’ mfaI-kgp promoter and evoglow-Bs2-stop were fused by PCR using mfaI5’-F and Evoglow-R primers. In the final step, 5’ mfaI-kgp promoter-evoglow-Bs2-stop and ermFAM-3’ mfaI were further merged by PCR using primers mfaI5’-F and mfaI3’-R. The final PCR product (mfaI::evoglow::ermFAM) was cloned into the pGEM®-T Easy vector (Promega, Madison, WI), generating the construct pHN-PgEvo. The plasmid was linearized using NotI and used to transform P. gingivalis W50 by electroporation (Figure 2.2). Recombinant colonies were selected on HBA plates containing erythromycin (5 μg/mL) and confirmation of DNA integration was performed by PCR analysis.

2.8.2 Construction of T. denticola ATCC 35405::FbFPevoglow-Bs2-stop (ECR774) and

ATCC 35405::FbFPevoglow-C-Bs2-stop (ECR775) strains Plasmids pHN-TdEvo and pHN-TdclosEvo were generated by SOE PCR. The oligonucleotides and primer pairs required to generate the corresponding amplicons are shown in Table 2.4 and Table 2.5 respectively. To construct pHN-TdEvo, the 5’ and 3’ flanking regions of TDE0911 (5’ TDE0911 and 3’ TDE0911), msp promoter, evoglow- Bs2-stop and aphA2 were first amplified by PCR with five pairs of primers as listed in Table 2.5. The tde911-5’R-Msp, Msp prom-F, Msp prom-R, Evo-F-Msp, Evo-R-Kan, Kan-F, Kan-R and tde911-3’F-Kan contain overlapping base pairs (underlined in Table 2.4). In the second step, 5’ TDE0911 and msp promoter were fused by PCR using tde911-

F and Msp prom-R primers while aphA2 and 3’ TDE0911 were fused by PCR using Kan- F and tde911-3’R-long primers. In the third step, the constructed aphA2-3’ TDE0911 and evoglow-Bs2-stop were fused by PCR using Evo-F-Msp and tde911-3’R-long primers. In the final step, 5’ TDE0911-msp promoter and evoglow-Bs2-stop-aphA2-3’ TDE0911 were further merged by PCR using primers tde911-F and tde911-3’R-long. The final PCR product (TDE0911::evoglow::aphA2) was cloned into the pGEM®-T Easy vector, generating the construct pHN-TdEvo. The plasmid was linearized using NotI and used to transform T. denticola ATCC 35405 by electroporation (Figure 2.3). Recombinant colonies were selected on OBGM agar plates containing kanamycin (75 μg/mL) and confirmation of DNA integration was performed by colony PCR analysis. Plasmid pHN- TdclosEvo was constructed in a similar manner except that the evoglow-C-Bs2-stop gene was used instead of evoglow-Bs2-stop and the primers Msp prom-R, Evo-F-Msp, Evo-R- Kan and Kan-F were replaced with Msp prom-closEvo-R, closEvo-Msp-F, closEvo-Kan- R and Kan-closEvo-F respectively. The final PCR product (TDE0911::clos- evoglow::aphA2) was cloned into the pGEM®-T Easy vector, generating the construct pHN-TdclosEvo. The plasmid was linearized using NotI and used to transform T. denticola ATCC 35405 by electroporation (Figure 2.3). Recombinant colonies were selected on OBGM agar plates containing kanamycin (75 μg/mL) and confirmation of DNA integration was performed by colony PCR analysis.

2.8.3 Construction of allele exchange T. denticola mutants Plasmid pCF382 was digested with EcoRI and BsaBI to release insert before it was used to transform T. denticola ATCC 35405 or ATCC 33520 by electroporation (Figure 2.4). Recombinant colonies were selected on OBGM agar plates containing erythromycin (40 μg/mL) and confirmation of DNA integration was performed by colony PCR analysis. Plasmids pHN-911, pHN-911kan, pHN-motA, pHN-motB, pHN-flgE, pHN-cheY and pHN-0659 were generated as allele exchange suicide plasmids by SOE PCR. The oligonucleotides and primer pairs required to generate the corresponding amplicons needed for each SOE PCR are shown in Table 2.4 and Table 2.5 respectively. Plasmid pHN-911 was generated as previously described (Bian and Li 2011). Briefly, the 5’ and 3’ flanking regions of TDE0911 (5’ TDE0911 and 3’ TDE0911) and ermAM were amplified by PCR with three pairs of primers as listed in Table 2.5. The tde911-R, tde911- 3’F, ermAM-F and ermAM-R primers contain overlapping base pairs (underlined in Table 2.4). In the second step, ermAM and 3’ TDE0911 were fused by PCR using primers

ermAM-F and tde911-3’R. In the final step, the constructed ermAM-3’ TDE0911 fragment and 5’ TDE0911 were further merged by PCR using primers tde911-F and tde911-3’R. The final PCR product (TDE0911::ermAM) was cloned into the pGEM®-T Easy vector, generating the construct pHN-911. pHN-cheY and pHN-0659 were generated in a similar manner using their corresponding primers. Similarly, pHN-motA, pHN-motB and pHN-flgE were generated using their corresponding primer pairs, except that an additional fragment (fla promoter) was fused to the constructs. pHN-911kan was generated in a similar manner using its corresponding primers, except that aphA2 was used instead of ermAM and an additional fragment (ermAM promoter) was fused to the construct. All plasmids were linearized with NotI and used to transform T. denticola ATCC 35405, ∆TDE0911::kan or ATCC 33520 by electroporation (Figure 2.5-Figure 2.14). Recombinant colonies were selected on OBGM agar plates containing erythromycin (40 μg/mL) and confirmation of DNA integration was performed by colony PCR analysis.

Figure 2.2 Generation of P. gingivalis W50::FbFP mutant strain ECR771. Schematic overview showing homologous recombination between pHN-PgEvo and the chromosome of P. gingivalis W50. Large arrows indicate orientation of the open reading frames (ORFs). The name of each gene is indicated within the arrow. PG0177, transposase; PG0178, MfaI fimbrillin; PG0179, fimbrillin-A associated anchor protein Mfa2; Pkgp, kgp promoter; evo, FbFP from evoglow-Bs2-stop; ermFAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.3 Generation of T. denticola ATCC 35405::FbFPevoglow-Bs2-stop and ATCC 35405::FbFPevoglow-C-Bs2-stop mutant strains ECR774 and ECR775 respectively. Schematic overview showing homologous recombination between pHN-TdEvo or pHN-TdclosEvo and the chromosome of T. denticola ATCC 35405. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. TDE0910, hypothetical protein; TDE0911, type II restriction endonuclease TdeIII; TDE0912, hypothetical protein; TDE0913, hypothetical protein; Pmsp, msp promoter; evo, FbFP from evoglow-Bs2-stop; cevo, FbFP from evoglow-C-Bs2-stop; aphA2, kanamycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.4 Generation of T. denticola ATCC 33520 ∆pyrF mutant strain ECR772. Schematic overview showing homologous recombination between pCF382 and the chromosome of T. denticola ATCC 33520. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. HMPREF9722_ RS06000, hypothetical protein; HMPREF9722_ RS06005, hypothetical protein; pyrF, orotidine 5'- phosphate decarboxylase; HMPREF9722_ RS06015, DEAD/DEAH box helicase; ermFAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.5 Generation of T. denticola ATCC 35405 ∆TDE0911 mutant strain ECR773. Schematic overview showing homologous recombination between pHN-911 and the chromosome of T. denticola ATCC 35405. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. TDE0910, hypothetical protein; TDE0911, type II restriction endonuclease TdeIII; TDE0912, hypothetical protein; TDE0913, hypothetical protein; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.6 Generation of T. denticola ATCC 35405 ∆cheY35405 mutant strain ECR823. Schematic overview showing homologous recombination between pHN-cheY and the chromosome of T. denticola ATCC 35405. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. cheW, chemotaxis protein CheW; cheX, chemotaxis protein CheX; cheY, chemotaxis protein CheY; TDE1495, hypothetical protein; TDE1496, chromosome partition protein SmC; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.7 Generation of T. denticola ATCC 35405 ∆TDE0911::kan mutant strain ECR824. Schematic overview showing homologous recombination between pHN-911kan and the chromosome of T. denticola ATCC 35405. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. TDE0910, hypothetical protein; TDE0911, type II restriction endonuclease TdeIII; TDE0912, hypothetical protein; TDE0913, hypothetical protein; Perm, ermAM promoter; aphA2, kanamycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.8 Generation of T. denticola ATCC 35405 ∆TDE0911::kan ∆cheY mutant strain ECR825. Schematic overview showing homologous recombination between pHN-cheY and the chromosome of T. denticola ATCC 35405 ∆TDE0911::kan. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. cheW, chemotaxis protein CheW; cheX, chemotaxis protein CheX; cheY, chemotaxis protein CheY; TDE1495, hypothetical protein; TDE1496, chromosome partition protein SmC; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.9 Generation of T. denticola ATCC 35405 ∆TDE0911::kan ∆flgE mutant strain ECR826. Schematic overview showing homologous recombination between pHN-flgE and the chromosome of T. denticola ATCC 35405 ∆TDE0911::kan. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. motA, motility protein A; TDE2767, flagellar protein-like protein; flgE, flagellar hook protein FlgE; flgD, flagellar basal body rod modification protein FlgD; fliK, flagellar hook-length control protein FliK; Pfla, fla operon promoter; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.10 Generation of T. denticola ATCC 33520 ∆motA33520 mutant strain ECR827. Schematic overview showing homologous recombination between pHN-motA and the chromosome of T. denticola ATCC 33520. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. flgE, flagellar hook protein FlgE; HMPREF9722_RS03030, flagellar protein-like protein; motA, motility protein A; motB, flagellar motor protein MotB; fliL, flagellar basal body-associated protein FliL; Pfla, fla operon promoter; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.11 Generation of T. denticola ATCC 33520 ∆motB33520 mutant strain ECR828. Schematic overview showing homologous recombination between pHN-motB and the chromosome of T. denticola ATCC 33520. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. HMPREF9722_RS03030, flagellar protein-like protein; motA, motility protein A; motB, flagellar motor protein MotB; fliL, flagellar basal body-associated protein FliL; fliM, flagellar motor switch protein FliM; Pfla, fla operon promoter; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.12 Generation of T. denticola ATCC 33520 ∆flgE33520 mutant strain ECR829. Schematic overview showing homologous recombination between pHN-flgE and the chromosome of T. denticola ATCC 33520. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. fliK, flagellar hook-length control protein FliK; flgD, flagellar basal body rod modification protein FlgD; flgE, flagellar hook protein FlgE; HMPREF9722_RS03030, flagellar protein-like protein; motA, motility protein A; Pfla, fla operon promoter; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.13 Generation of T. denticola ATCC 33520 ∆cheY33520 mutant strain ECR830. Schematic overview showing homologous recombination between pHN-cheY and the chromosome of T. denticola ATCC 33520. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. HMPREF9722_RS08875, chromosome partition protein SmC; HMPREF9722_RS08880, hypothetical protein; cheY, chemotaxis protein CheY; cheX, chemotaxis protein CheX; cheW, chemotaxis protein CheW; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

Figure 2.14 Generation of T. denticola ATCC 33520 ∆DfsBTd33520 mutant strain ECR831. Schematic overview showing homologous recombination between pHN-0659 and the chromosome of T. denticola ATCC 33520. Large arrows indicate orientation of the ORFs. The name of each gene is indicated within the arrow. HMPREF9722_RS12905, putative transcriptional regulator; HMPREF9722_RS12910, DfsB homologous protein; HMPREF9722_RS12915, ABC transporter, permease protein; ermAM, erythromycin resistance gene. Restriction enzyme recognition sites are indicated by bold vertical lines and enzyme name. Regions of homology are indicated by dashed lines. Position and orientation of primers used to confirm homologous recombination are indicated with small arrows. Not to scale.

2.9 DIGESTION OF RECOMBINATION CASSETTES WITH T. DENTICOLA LYSATES The recombination cassettes were amplified from their respective plasmids by PCR using the primer pair M13 Forward/M13 Reverse. pCF382 was digested with BsaBI and EcoRI to release the recombination cassette from the vector. The PCR and digested products were purified using the QIAquick Gel Extraction Kit as per manufacturer’s instructions. To prepare methylated plasmids for digestion, plasmids isolated from E. coli α-select cells were digested with NotI and the digested products were purified using DNA precipitation. T. denticola cell lysates were prepared as previously described (Bian and Li 2011). Briefly, the cells were grown to exponential phase before 1.5 × 109 cells were harvested by centrifugation and resuspended in 300 µL of PBS. The cells were lysed by sonication on ice using a CPX750 Ultrasonic processor (Cole Parmer Instrument Company, USA) for one 10 second and one 20 second pulse with a 10 second interval. After centrifugation at 6,000 g for 5 min at 4°C, the supernatants were collected and stored at 4°C as crude cell extracts. For the DNA digestion analysis of PCR-amplified recombination cassettes and pCF382, 1 µg of each recombination cassette and digested pCF382 was mixed with 4 µL of crude cell extract, 2 µL of 10× NEB buffer 4 (New England BioLabs) and deionized water to a total volume of 20 µL. For the DNA digestion analysis of methylated, NotI-digested plasmids, 5 µg of each plasmid was mixed with 20 µL of crude cell extract, 4 µL of 10× NEB buffer 4 and deionized water to a total volume of 40 µL. The reaction mixtures were incubated at 37°C. A sample of the reaction mixture collected after 30 min and 2 h of incubation was analyzed by agarose gel electrophoresis.

2.10 FLUORESCENCE ASSAYS FOR THE ANALYSES OF P. GINGIVALIS W50 AND T. DENTICOLA ATCC 35405 EXPRESSING FBFP 2.10.1 Fluorescence microscopy Fluorescence images of E. coli colonies on LB agar plates were obtained after a 1 or 2 s exposure using the Fujifilm LAS-3000 Imager (470 nm wavelength excitation filter; 515 nm wavelength emission filter). Overnight cultures of E. coli were concentrated 5 fold and a small volume (~40 µL) of the cultures was aliquoted onto a microscope slide. Fluorescence images were obtained using the Fujifilm LAS-3000 Imager with the same settings.

For imaging of planktonic T. denticola cells, the cells were grown to exponential phase and pelleted at 4,000 g for 10 min at RT. The cells were washed twice with PBS, resuspended in PBS, placed on glass slides and allowed to dry. For preparation of fixed cells, 4% (v/v) paraformaldehyde (PFA) in PBS was added to the dried cells and incubated for 30 min. The slides were then washed once each with PBS and water before curing with ProLong® Gold Antifade Mountant according to the manufacturer’s protocols. The cells were visualized on a confocal laser scanning microscope as described previously (Zainal-Abidin et al. 2012). Overnight grown, planktonic E. coli cells were prepared and visualized in a similar manner.

For imaging of planktonic P. gingivalis cells, the cells were grown to exponential phase and pelleted at 8,000 g for 10 min at RT. The cells were washed twice in PBS before a final resuspension in PBS. The cells were visualized on a combined fluorescence and phase contrast microscope with the help of Mr. Cameron Nowell at the imaging, flow cytometry and analysis core facility at Monash Institute of Pharmaceutical Science.

2.10.2 Fluorescence spectroscopy The fluorescence intensities of P. gingivalis transformants were compared with that of WT using a Cary Varian spectrofluorometer. P. gingivalis cells were grown to exponential phase and a sample of cells were collected as unwashed cells. The remaining cells were pelleted at 8,000 g for 10 min at RT. The cell pellet was washed twice in 0.85% (w/v) NaCl before resuspension in 0.85% (w/v) NaCl. The cells were collected as washed cells. The unwashed and washed cells were aliquoted into a quartz cuvette and the fluorescence intensity was measured across a range of wavelength from 450 nm to 600 nm with the excitation wavelength fixed at 450 nm.

2.11 PROTEIN ISOLATION AND ANALYSIS 2.11.1 Trichloroacetic acid (TCA) precipitation Protein samples were precipitated by adding TCA (6.1 M stock solution, Sigma-Aldrich) to a final concentration of 4% (v/v) and incubated at -30°C for 30 min. Protein was pelleted by centrifugation at 16 100 g for 1 h at 4ºC. The supernatant was removed. Ice- cold acetone [100% (v/v)] was added to wash the protein pellet. The acetone was discarded after centrifugation at 16 100 g for 15 min at 4ºC. The protein pellet was air- dried and suspended in an appropriate volume of deionized water or 1 × NuPAGE LDS Sample Buffer (Thermo Fisher Scientific Inc., USA).

2.11.2 Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) Whole cell samples or TCA precipitated culture supernatants and protein samples were analyzed by SDS-PAGE. TCA precipitated samples suspended in 1 × NuPAGE LDS Sample Buffer were supplemented with 50 mM DTT while samples in solution were mixed with 4 × NuPAGE LDS Sample Buffer to a final 1 × concentration and 50 mM DTT. Samples were heat-denatured at 100°C for 5 min before they were resolved on a 10% (w/v) or 12% (w/v) NuPAGE® Novex® Bis-Tris or a 4-12% (w/v) gradient NuPAGE® Novex® Bis-Tris precast mini-gel using standard NuPAGE® protocols (Invitrogen Corporation, USA). SeeBlue® Plus2 Pre-Stained Standard (Invitrogen Corporation) was used as the molecular mass standard. Gels were run until the dye front of the SeeBlue® Plus2 Pre-Stained Standard reached the foot of the gel and stained with SimplyBlue™ SafeStain (Invitrogen Corporation) according to the manufacturer’s instructions. Stained gels were imaged with a Fujifilm LAS-3000 Imager.

2.11.3 Immunoblotting Proteins were resolved by SDS-PAGE and transferred onto a nitrocellulose membrane (Protran Nitrocellulose Hybridization Transfer Membrane, 0.45 µm; PerkinElmer, USA) using NuPAGE® Transfer Buffer according to the standard NuPAGE® protocols (Invitrogen Corporation). To visualize the protein transferred, Ponceau S (SIGMA) staining of the membrane was carried out. The membrane was washed in water to remove the stain before it was blocked with blocking buffer [5% (w/v) non-fat skim milk in PBST (0.1% (v/v) Tween-20 in PBS)] overnight at 4°C or at RT for 1 h. The membrane was then probed with primary antibody, evoglow antibodies Bs (evocatal GmbH, Germany), in blocking buffer at RT for 1 h, washed several times with PBST and then probed with IgG horseradish peroxidase (HRP)-conjugated anti-rabbit antibody in blocking buffer at RT for 1 h. The membrane was washed several times with PBST and developed with Immobilon® Western chemiluminescent HRP substrate (Merck Millipore, a subsidiary of Merck, Germany) according to the manufacturer’s instructions. Chemiluminescence was visualized with a Fujifilm LAS-3000 Imager.

2.11.4 In-gel trypsin digestion and MALDI-TOF MS or LC-MS/MS Gel bands were excised from the SimplyBlue™ SafeStain stained SDS-PAGE gel and cut into small pieces (approximately 1 mm3 cubes) using scalpels. The gel cubes were dehydrated with 100 µL of ethanol and then reduced with 100 µL of 10 mM DTT in 50 mM ammonium bicarbonate (ABC; pH 8.0) at 56°C for 1 h. The supernatant was removed

before the sample was alkylated with 50 µL of 50 mM iodoacetamide in 50 mM ABC at RT for 30 min in the dark. The supernatant was removed and the gel cubes were washed 4 times at RT with rotation for 15 min using a series of solutions (500 µL each): 50 mM ABC; 50% (v/v) ethanol in 20 mM ABC; 20 mM ABC and finally 50% (v/v) ethanol in 20 mM ABC. The gel cubes were then dehydrated with 50 µL of ethanol and supernatant was removed. In-gel trypsin digestion was carried out by addition of sequencing-grade trypsin (10 ng/µL in 20 mM ABC; Promega) to the sample followed by incubation at 37°C overnight. After incubation, acidification was performed by the addition of TFA to a final concentration of 0.1% (v/v). The reaction was centrifuged and the supernatant was transferred to a new tube for analysis by MALDI-TOF MS or LC-MS/MS.

For analysis of whole proteins extracted from different sections of OBGM agar, three additional peptide extractions were carried out with 0.1% (v/v) TFA, 20% (v/v) acetonitrile (ACN) in 0.1% (v/v) TFA and 50% (v/v) ACN in 0.1% (v/v) TFA solutions. For each extraction, 80 µL of extraction solution was added to the gel pieces and sonicated for 10 min in a sonicator bath. The extracts were pooled and frozen in liquid nitrogen before they were concentrated in a freeze dryer until totally dehydrated. The samples were resuspended in 10 µL of 8% (v/v) ACN in 0.1% (v/v) TFA. The solution was then sonicated for 10 min in a sonicator bath to aid dissolution and then stored at -80°C. Ten microliters of 0.1% (v/v) TFA was added to the samples which were then centrifuged at 16,000 g for 10 min. The supernatant was removed to a new tube and sent for LC-MS/MS analysis.

MALDI-TOF MS analysis was performed as previously described (O'Brien-Simpson et al. 2005) except that the proteins were identified by peptide mass fingerprinting against the T. denticola database instead of P. gingivalis database using an in-house Mascot search engine. LC-MS/MS analysis was carried out as previously described (Glew et al. 2017).

2.12 CHARACTERIZATION OF T. DENTICOLA ATCC 33520 MUTANTS 2.12.1 Genomic sequencing All aspects of genome sequencing from template preparation through to the use of the Ion Torrent Personal Genome Machine (PGM) were performed by Ms. Brigitte Hoffmann using the kits and detailed protocols of the manufacturer Thermo Fisher Scientific, unless otherwise stated. Briefly, 1 µg genomic DNA was physically fragmented to

approximately 400 bp in size using the Covaris M220 Focused-ultrasonicator™ (TrendBio, Australia). A 1 µL aliquot of sheared DNA was visualised using the LabChip GX Touch 24 Nucleic Acid Analyzer (PerkinElmer, USA) to ensure a peak fragment size of 400 bp. Upon confirmation of appropriate shearing, the DNA was end-repaired (Ion Xpress™ Plus Fragment Library Kit), purified (Agencourt™ AMPure™ XP Kit, Beckman Coulter), barcoded adaptors ligated to the DNA then nick-repaired (Ion Xpress™ Plus Fragment Library Kit; Ion Xpress™ Barcode Adapters Kit) and purified again (Agencourt™ AMPure™ XP Kit, Beckman Coulter). The labelled library was then size-selected again using the Pippin Prep™ DNA Size Selection System (Sage Science), aiming for a target-peak size of ~480 bp collected over a specified range. Following sample purification (Agencourt™ AMPure™ XP Kit, Beckman Coulter) the concentration of the unamplified library was determined using qPCR (Ion Library TaqMan® Quantitation Kit). All of the prepared libraries were at an adequate concentration which did not require further amplification. As up to 8 genomes can be sequenced at a time, barcoded libraries were pooled in equimolar amounts of 26 pM to ensure an equal representation of each barcoded library in the sequencing run. The library was then used to prepare enriched, template-positive Ion PGM™ Hi-Q™ Ion Sphere™ Particles (ISPs) using the Ion OneTouch™ 2 System (Ion PGM™ Hi-Q™ OT2 Kit). The recovered template-positive ISPs were enriched using the Ion OneTouch™ ES Instrument and Ion OneTouch™ ES Supplies Kit, then loaded onto an Ion 318™ Chip v2 BC and sequenced using the Ion PGM™ Hi-Q™ Sequencing Kit and Ion PGM™ Instrument. The resulting sequencing reads were downloaded from the Torrent Server and analyzed by Dr Catherine Butler using Geneious R8.1.9 (Biomatters Ltd, New Zealand) with comparison made to the available genome reference sequences.

2.12.2 Colony morphology Images of T. denticola colonies on OBGM agar plates and swimming assays were obtained using the Fujifilm LAS-3000 Imager.

2.12.3 Swimming assay To compare the motility of the T. denticola mutants to that of the WT ATCC 33520 or ATCC 35405, a swimming assay in semisolid OBGM agar medium [OBGM supplemented with 0.4% (w/v) Molecular Grade Agarose (Bioline) and 1% (w/v) BD Difco™ gelatin (Bacto Laboratories Pty. Ltd.)] was performed. T. denticola cells at exponential phase of growth were harvested by centrifugation (4 000 g, 6 min, RT) and

resuspended in an appropriate volume of deionized water. The bacterial suspension (2 µL; 107 or 108 cells) was spotted in the semisolid OBGM agar medium. The plates were dried for 20 min at RT before they were incubated anaerobically at 37°C for 7 or 10 days. The area of turbid plaque was measured.

2.12.4 Cryo-electron microscopy (Cryo-EM) T. denticola cultures were grown to exponential phase as previously described. Whole culture or whole cells were prepared for cryo-EM and imaged by Dr Yu-Yen Chen as previously published (Chen et al. 2011) with the following modifications. The cells were not washed. The cryo-EM instrument, FEI Tecnai G2 F30, was operated at 200 or 300 kV and equipped with either a Gatan US1000 2k × 2k CCD camera or an FEI Ceta 4k × 4k CMOS camera.

2.12.5 Growth curves T. denticola cells at stationary phase of growth were inoculated into fresh OBGM to an

A650 of 0.02-0.03. Cell density was monitored over a period of 264 h by measuring the

A650 of the culture using a UV-visible spectrophotometer (Varian). The generation time of each culture was calculated as follows. The growth curves were plotted as log10 (A650 × 100) against time. A linear regression was drawn using the points in exponential phase of growth and an equation was obtained. The log10 (A650 × 100) at the start of the time interval and log10 (A650 × 100) at the end of the time interval were calculated from the equation. The generation time was calculated with the following equation:

log 2 G = 푡 × b−B where G = generation time in hours t = time interval in hours b = log10 (A650 × 100) at the end of the time interval

B = log10 (A650 × 100) at the start of the time interval

2.12.6 Autoaggregation and coaggregation assays Autoaggregation of T. denticola strains and P. gingivalis W50 as well as coaggregation between T. denticola strains and P. gingivalis W50 were evaluated as previously described (Abiko et al., 2014b) with some modifications. T. denticola and P. gingivalis cells were grown to exponential phase before they were harvested by centrifugation at 4,000 g for 10 min at RT. Cells were washed twice with coaggregation buffer (20 mM 78

phosphate buffer, pH 8.0 supplemented with 1 mM CaCl2, 1 mM MgCl2, and 150 mM

NaCl) and suspended in the coaggregation buffer to an A650 of 0.5. For the autoaggregation assay, 1 mL of the cell suspension was placed in a cuvette and the height of the bacterial aggregates was monitored using a ruler for 7 h. For the coaggregation assay, equal volumes of T. denticola and P. gingivalis cell suspensions were combined in a cuvette to give an A650 of 0.5. The height of the bacterial aggregates was monitored for 7 h.

2.12.7 Reverse Transcription-Polymerase Chain Reaction (RT-PCR)

Whole cells of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520 from cultures at exponential phase were harvested by centrifugation at 4,000 g at 4°C for 10 min. The cell pellets were snap frozen in liquid N2 and stored at -80°C until use. Total RNA was extracted from the frozen cell pellets using TRIzol® reagent (Invitrogen, Australia) according to the manufacturer's instructions. Cells were homogenized by chemical lysis using the TRIzol® reagent and mechanical lysis using 100 µM 0.1 mm silica beads (Lysing Matrix B, MP Biomedicals, USA). Contaminating genomic DNA was removed with TURBO™ DNase (Ambion, Australia) according to the manufacturer's instructions. PCR amplification using the primer pair TDE1495- F/TDE1495-R (Table 2.4) was performed to assess DNA contamination in the RNA samples. The concentration of extracted RNA was determined spectrophotometrically using the NanoDrop™ Spectrophotometer ND-1000. The integrity of RNA in the samples was determined by agarose gel electrophoresis and by 260/280 ratio from the NanoDrop™ Spectrophotometer ND-1000.

The RNA extracted from different T. denticola strains was reverse transcribed into cDNA using the Invitrogen™ SuperScript™ VILO™ cDNA Synthesis Kit (Thermo Fisher Scientific Inc., USA) according to the manufacturer's instructions. The cDNA generated was amplified by PCR for 20 cycles using oligonucleotide primers specific to each gene as listed in Table 2.6. PCR amplicons were analyzed using agarose gel electrophoresis.

Table 2.6 Primers used for RT-PCR.

Gene Forward primerA Reverse primerA Expected product size (bp) fliK/flgD TDE2768-F TDE2768-R 802 motA/motB TDE2766-F TDE2766-R 770 motB/fliL TDE2765-F TDE2765-R 855 TDE2767/motA TDE2767-F (flgE TDE2767-R (flgE 660 KO) KO) fliL/fliM TDE2764-F TDE2764-R 634 A Primer sequences listed in Table 2.4.

2.12.8 Quantitative proteomic analyses Raw MS files were analyzed by MaxQuant (Ver 1.5.3.30) using label-free quantification (LFQ) searching against the whole T. denticola ATCC 35405 protein database containing a total of 2786 protein sequences obtained from the Comprehensive Microbial Resource Website (cmr.jcvi.org). The default parameters were used for LFQ. Data analysis was performed in Excel, using the MaxQuant output file ‘proteinGroups.txt.’. Any contaminating eukaryotic proteins were removed from the total proteins. LFQ intensity values were used to produce ratios of paired samples and the geometric mean of these ratios was calculated. Statistical analysis was performed using paired, two-tailed Student’s T-tests in Excel and a p value of <0.05 was considered to be statistically significant. A protein’s abundance was considered to have significantly changed between two samples if there was ≥1.5- or ≤0.67-fold change in the geometric mean and a p value of <0.05.

2.12.9 Static biofilm assay Static biofilm assay was carried out as previously described (Dashper et al., 2014) with some modifications. Briefly, T. denticola and P. gingivalis cells were grown to exponential phase (A650 of 0.10-0.30 for T. denticola and A650 of 0.6-0.8 for P. gingivalis) and diluted with fresh, pre-reduced OBGM to A650 of 0.15 where necessary. For monospecies biofilms, 2 mL of each diluted culture was aliquoted into each well of a Greiner CELLSTAR® 12-well flat-well plate (Sigma-Aldrich). For dual-species biofilms, equal volumes of T. denticola and P. gingivalis diluted cultures were aliquoted into each well of the 12-well flat-well plate and mixed. After inoculation, the plates were

incubated at 37°C anaerobically for 1 h, sealed with a Greiner multiwell plate sealer (Sigma-Aldrich) and further incubated at 37°C anaerobically for 5 days. After incubation, the medium in each well was decanted and each well was gently rinsed with 2.1 mL of deionized water. The plates were air-dried and the biofilms were stained with 0.1% (v/v) crystal violet for 30 min at RT. The crystal violet was removed and each well was rinsed twice with 2.5 mL of deionized water. The plates were air-dried and destained with 2.1 mL of 99% (v/v) ethanol. The A540 of each well was measured using a Wallac VICTOR3™ 1420 Multilabel Counter (PerkinElmer, USA). The non-parametric counterpart of the (one-way) ANOVA, the Kruskal-Wallis rank sum test (Kruskal and Wallis 1952), was conducted with the Conover-Imam test (Conover and Iman 1979) for the statistical analysis of the biofilm data and a p value of <0.05 was considered to be statistically significant. For real-time PCR analysis, each well was rinsed with 1 mL of OBGM to remove loosely attached and planktonic cells after the medium in the well was decanted. The biofilm was then removed from the substratum into 1 mL of OBGM and the bacterial cells were enumerated by real-time PCR.

2.12.10 Real-time PCR DNA from each biofilm sample was extracted using the DNeasy PowerBiofilm Kit (Qiagen) following the manufacturer’s instructions. Real-time PCR was carried out as previously described (Byrne et al. 2009, Dashper et al. 2014). Briefly, real-time PCRs were performed in triplicate using oligonucleotide primers targeting the 16S rRNA genes of T. denticola and P. gingivalis. A 25 µL reaction was set up with 12.5 µL Platinum SYBR green quantitative PCR (qPCR) Supermix UDG (Invitrogen), 9.5 µL DNase-free deionized water, a 200 nM final concentration of forward and reverse primers and 20 ng of template. All real-time PCR conditions consisted of an initial heating step at 50°C for 2 min and denaturation at 95°C for 2 min, followed by 35 cycles of denaturation at 95°C for 15 s, annealing at 58°C for 30 s, and extension at 72°C for 30 s. Fluorescence data were collected immediately following the extension step of each cycle. Tenfold serial dilutions of DNA of known concentration as determined using a NanoDrop™ Spectrophotometer ND-1000 were used to construct standard curves for quantification of each species. The total cell number of each bacterial species was then calculated for each biofilm sample. Student’s T-test was used for the statistical analysis of the data and a p value of <0.05 was considered to be statistically significant.

2.12.11 Growth of biofilms in flow cells The growth of biofilms in flow cells was performed by Ms. Sze Wei Liu as previously published (Zhu et al. 2013). Briefly, biofilms were cultured in a 3D-printed block with VeroClear, RGD810, which is a rigid and nearly colourless printing material. The block has a single track that is 50 mm long, 15 mm wide and 2 mm deep and covered by a standard-sized, uncoated glass microscope coverslip (ProSciTech, QLD, Australia). The coverslip which served as the attachment substratum for the biofilm was secured to the flow cell with a silicone adhesive (GE Silicones, General Electric Company, Waterford, NY). To ensure sterility, 0.5% (v/v) sodium hypochlorite was pumped through the flow cell system for 3 h at 3 mL/h, followed by 1% (v/v) hydrogen peroxide for 10 min at 3 mL/h. The flow cell system was then rinsed with sterile deionized water overnight to flush out the hydrogen peroxide. The flow cell system was transferred into a MG500 anaerobic workstation at 37°C, filled with pre-reduced 20% (v/v) OBGM, and reduced overnight to condition the glass surface with medium. Snap frozen pellets of bacteria prepared from exponential-phase cultures (A650 of 0.6 for P. gingivalis and A650 of 0.1 for T. denticola) were resuspended in 1 mL of pre-reduced OBGM. Equal volume of P. gingivalis and T. denticola was mixed with an additional 1 mL of pre-reduced OBGM and used as the inocula for polymicrobial biofilms. After inoculation, the system was incubated for 1 h with the coverslip side facing down prior to a constant flow (3 mL/h) of 20% (v/v) OBGM with the coverslip side facing up. Biofilms adhered to the glass coverslips were harvested 90 h after the commencement of constant medium flow. PBS was pumped through the flow cell for 30 min at 6 mL/h to remove culture medium and unattached bacterial cells. The biofilms were fixed with 4% (v/v) PFA for 30 min at 6 mL/h and then rinsed with PBS for 30 min at 6 mL/h to remove the PFA. The biofilm was embedded in 25% (w/v) acrylamide with 0.025% (w/v) ammonium persulfate and 0.8% (v/v) N,N,N’,N’- tetramethylethylenediamine (TEMED). The gels were left to set at RT overnight. The coverslips were then removed and the acrylamide slab with embedded biofilms was stored in PBS at 4°C prior to analysis.

2.12.12 Fluorescent staining of biofilms, confocal laser scanning microscopy and image analysis Biofilms were subjected to fluorescent in situ hybridization (FISH) using species-specific probes by Ms. Sze Wei Liu as previously described (Zainal-Abidin et al. 2012). Fluorescently-labelled biofilms were visualized on an Axiovert 200 M inverted

microscope (Carl Zeiss, Germany) fitted with a Zeiss LSM 510 META Confocal scan head, using the 458/477/488 nm argon and 543 nm HeNe laser lines. A 63×, 1.2- numerical aperture oil immersion lens was used to record 5 image stacks in random positions from each of three biological replicates. The confocal data sets were analyzed with COMSTAT software to determine the biometric parameters of the biofilm (Heydorn et al. 2000, Vorregaard 2008). Student’s T-test was used for the statistical analysis of the biometric data and a p value of <0.05 was considered to be statistically significant. Three- dimensional reconstructed images were produced using Zeiss LSM image browser (Carl Zeiss, Germany). All visualization and analyses were done by Ms. Sze Wei Liu.

2.12.13 Inhibition assay T. denticola cells at exponential phase of growth were harvested by centrifugation (1 000 g, 10 min, RT) and resuspended in an appropriate volume of deionized water. The bacterial suspension (2 µL; 108 cells) was spotted in the semisolid OBGM agar medium according to Figure 2.15A. The plates were dried for 20 min at RT before they were incubated anaerobically at 37°C until zones of clearing were observed between the colonies.

To test for the inhibition of T. denticola and P. gingivalis growth in the center of four evenly-spaced T. denticola colonies, T. denticola cells (2 µL; 108 cells) were spotted in the semisolid OBGM agar medium according to Figure 2.15A. When zones of clearing were observed between the colonies after 9 days of anaerobic incubation, T. denticola or P. gingivalis cells (2 µL; 108 cells) were spotted in the middle of the T. denticola colonies and near the edge of the plates as shown in spots a and b, respectively, in Figure 2.15A. The plates were further incubated anaerobically at 37°C for up to 24 days.

To test for the inhibition of T. denticola growth in the center of four evenly-spaced P. gingivalis colonies, P. gingivalis cells (2 µL; 108 cells) were spotted in the semisolid OBGM agar medium according to Figure 2.15A. After 4 days of anaerobic incubation, T. denticola cells (2 µL; 108 cells) were spotted in the middle of the P. gingivalis colonies and near the edge of the plates. The plates were further incubated anaerobically at 37°C for 8 days.

Figure 2.15 T. denticola and P. gingivalis inhibition assay. A. T. denticola and P. gingivalis cells (2 µL; 108 cells) were spotted 1 cm-apart in the semisolid OBGM agar medium. “a” and “b” indicate the positions where the cells were spotted in the middle of colonies and near the edge of the plate respectively. B. Regions (A-E) of the agar where the protein was isolated from.

2.13 EXTRACTION AND ANALYSIS OF PROTEINS SECRETED FROM T. DENTICOLA COLONIES T. denticola colonies were grown in semisolid OBGM agar, with or without rabbit serum, until zones of clearing were observed. Secreted proteins were extracted from the agar as previously published (Be'er et al. 2009). Briefly, sections of agar from zones A-E as labelled in Figure 2.15B were excised from 20 plates and suspended in 10 mL of sterilized deionized water for 3 days. The supernatant was filtered through a 0.22 µm filter, TCA precipitated and run on SDS-PAGE. In-gel trypsin digestion was carried out as previously described (Section 2.11.4). After the final acidification with 0.1% (v/v) TFA, the reaction was centrifuged and the supernatant was transferred to a new tube for analysis using MALDI-TOF MS or LC-MS/MS.

2.14 BIOINFORMATIC ANALYSES The percentage of nucleotide identity between the upstream and downstream genes of pyrF in T. denticola ATCC 35405 and ATCC 33520 was determined using EMBOSS needle nucleotide pairwise sequence alignment tool on EMBL-EBI Job Dispatcher framework (Rice et al. 2000, McWilliam et al. 2013, Li et al. 2015). The homologues of ORF B in pTD1, a naturally occurring plasmid in T. denticola ATCC 33520 (Ivic et al. 1991, Caudry et al. 1995), were identified by using the sequence of ORF B in PSI-BLAST

(Position-Specific Iterated BLAST) on NCBI. The DfsB homologue in T. denticola

ATCC 35405, TDE0659 (designated as DfsBTd35405 in this study), was identified by Dr Catherine Butler through searching for DUF1706 family proteins in T. denticola ATCC 35405. Subsequent bioinformatic analyses were done using Compute pI/Mw tool on the ExPASy server (Bjellqvist et al. 1993, Bjellqvist et al. 1994, Gasteiger et al. 2005) for prediction of the size and pI of the proteins, EMBOSS needle protein pairwise sequence alignment tool on EMBL-EBI Job Dispatcher framework (Rice et al. 2000, McWilliam et al. 2013, Li et al. 2015) for protein pairwise sequence alignment, SignalP 4.1 Server (Petersen et al. 2011) for prediction of the presence of signal peptide and PSORTb v3.0 program for prediction of the subcellular localization of the proteins (Yu et al. 2010). The protein homology detection and prediction servers, Phyre2 (Kelley et al. 2015), Swiss- Model (Biasini et al. 2014) and HHpred (Soding et al. 2005), were used to predict and analyze protein structure. DfsBTd35405 homologues were identified by using the sequence of DfsBTd35405 in BLASTP (protein-protein BLAST) on NCBI. A multiple sequence alignment of DfsBTd35405 homologues from different T. denticola strains was conducted using the ClustalW2 multiple sequence alignment tool on EMBL-EBI.

CHAPTER 3 ESTABLISHMENT OF AN EFFICIENT AND REPRODUCIBLE SYSTEM FOR T. DENTICOLA TRANSFORMATION AND GENERATION OF T. DENTICOLA MUTANTS

3.1 INTRODUCTION T. denticola is an obligate anaerobe that is frequently associated with P. gingivalis in severe periodontal disease. It is a slow-growing, highly fragile bacterium (Salvador et al. 1987, Wardle 1997) with fastidious nutrient requirements, making its cultivation and handling in laboratory very difficult. The genetic analysis of T. denticola has been progressing slowly since its first successful transformation about 20 years ago (Li and Kuramitsu 1996, Li et al. 1996), mainly because of its extremely low transformation efficiency, strain-limited shuttle plasmid and the availability of a small number of selectable markers (Kuramitsu et al. 2005b, Bian and Li 2011, Godovikova et al. 2015, Li et al. 2015). The transformation efficiency of T. denticola ranges from 0.5 cfu/µg to 1500 cfu/µg of DNA (Li and Kuramitsu 1996, Li et al. 1996, Chi et al. 1999). The slow growth rate of this bacterium has further made its transformation a time-consuming process, taking at least two weeks from the growth of the cells to the detection of transformants (Kuramitsu et al. 2005b). Despite two decades passing, a relatively small number of T. denticola transformants have been reported in the literature and many laboratories have reported modifications of the original published protocol for successful transformation (Ishihara et al. 1998, Limberger et al. 1999, Chi et al. 2002, Lux et al. 2002, Abiko et al. 2014b, Godovikova et al. 2015), indicating the extreme difficulty in T. denticola transformation.

Futhermore, the transformation efficiency varies among different strains of T. denticola. T. denticola ATCC 33520 was reported to have a higher transformation efficiency than the type strain ATCC 35405 (Bian and Li 2011). For example, the E. coli-T. denticola shuttle plasmid pKMCou which was successfully transformed into ATCC 33520 was unable to be transformed into ATCC 35405. This is probably due to the incapability of T. denticola ATCC 35405, which naturally lacks a plasmid, to maintain the shuttle plasmid (Chi et al. 2002). Since T. denticola ATCC 33520 naturally harbors a plasmid pTD1(Ivic et al. 1991), it might express a system for maintaining plasmids; alternatively the plasmid pTD1 may be important in maintaining the shuttle vector (Chi et al. 2002). In addition, it has been shown that the restriction modification systems, particularly TDE0911, present uniquely in T. denticola ATCC 35405 and not ATCC 33520, are a major reason for the low transformation efficiency of ATCC 35405. Inactivation of the gene encoding TDE0911 allowed T. denticola ATCC 35405 to be transformed with a shuttle vector at a similar efficiency as ATCC 33520 (Bian and Li 2011). Together, these studies showed

that the absence of a naturally occurring plasmid and the presence of unique restriction modification systems in T. denticola ATCC 35405 contribute to its low transformation efficiency.

The transformation protocol for T. denticola has been optimized continually over the years in different studies due to the lack of reproducibility of the original protocol. For example, one study introduced a heat-treatment step (50°C for 30 min) before chilling and washing the cells and it was reported that this heat-treatment step greatly increased the transformation efficiency of T. denticola, probably because it minimized the effects of T. denticola restriction modification systems on transforming DNA (Chi et al. 2002). Besides that, the amount of DNA used for different transformations can vary from 0.5 µg to as high as 50 µg (Chi et al. 2002, Abiko et al. 2014b). The growth media, wash buffer and agar used for the selection of transformants are also different in different studies. In addition to low transformation efficiency, a high “false-positive” transformation frequency, i.e. a high number of antibiotic-resistant mutants that do not contain the desired mutation, are found in T. denticola transformation. For example, in the study by Ishihara et al. (2010) and Bian and Li (2011), only 8% and 13% respectively of the erythromycin- resistant T. denticola ATCC 35405 colonies were true transformants. All these complexities suggest the need to develop a reliable and reproducible genetic manipulation system for T. denticola to increase its transformation efficiency and at the same time decrease the “false-positive” transformation frequency. This study attempted to achieve this aim by systematically investigating a number of variables based on the published standard protocol and determining the most important factors in improving the transformation efficiency of T. denticola. The optimized transformation protocol was then used to generate T. denticola mutants expressing heterologous genes or lacking specific genes for subsequent experiments described in the following chapters. The problems faced while generating T. denticola mutants and the possible reasons for them are also discussed in this study.

3.2 RESULTS 3.2.1 Transformation of T. denticola ATCC 35405 with plasmid pCF382 Plasmid pCF382 [Figure 3.1A; (Capone et al. 2007)], a kind gift from Dr Christopher Fenno (School of Dentistry, University of Michigan), was used to establish an efficient and reproducible system for T. denticola transformation. This plasmid was chosen as it had been successfully used to generate a mutant lacking the pyrF gene (TDE2110) in the T. denticola type strain ATCC 35405 (Capone et al. 2007). Previous attempts in this laboratory to generate T. denticola ATCC 35405 ∆pyrF mutant with pCF382 using the original protocol were unsuccessful (K. H. Tan, personal communication). Furthermore, there was a high number of spontaneous erythromycin-resistant colonies on the transformation plates (K. H. Tan, personal communication). At the beginning of this study, numerous variables were trialed based on the original protocol in cell preparation and electroporation conditions, but no viable cells were observed on the positive control plates, i.e. cells electroporated without plasmid DNA and plated on non-antibiotic- containing OBGM agar (results not shown). This indicates that T. denticola is extremely fragile and prone to being killed during the transformation process. The cells were then handled in a gentler manner such as using cut tips (1 mL pipette tips which had approximately 1 cm cut off the tip) for all resuspension steps to reduce shear forces on the cells. The gentle handling of cells was proven successful as viable cells were obtained on the positive control plates. After establishing the conditions required for viable cells, a number of variables were investigated based on the published standard protocol (Table 3.1) but they were not exhaustive due to the large number of samples and plates, long growth times and time constraints within an experiment.

Variables in the preparation of competent cells including the growth phase, temperature of preparation and freezing of cells were investigated. T. denticola ATCC 35405 was grown to exponential (A650 of 0.14-0.16) or stationary (A650 of 0.30) phase, harvested by centrifugation, washed twice or thrice in 10% (v/v) glycerol and resuspended in 10% (v/v) glycerol. All of these steps were done under aerobic conditions following the existing published protocols at that point in time. The purpose of the wash steps was to remove the conductive solutes in the medium which would compromise the transformation by causing arcing during electroporation. Since no arcing was observed from the cells that were washed twice and to reduce cell loss from the extra wash step, the cells were washed only twice from the 4th trial onwards.

Preparation of plasmid pCF382 trialed no digestion, linearization or double digestion, and the optimal amount of purified plasmid used for transformation was also investigated. Linearization of pCF382 was performed with EcoRI or PstI (Figure 3.1B) on plasmid prepared from E. coli NovaBlue (ECR535) which is used for general plasmid cloning. The double digestion of pCF382 was carried out with EcoRI and BsaBI to release the recombination cassette (4.7 kbp) from the background vector (2.7 kbp; Figure 3.1C). The plasmid was prepared from E. coli SCS110 (ECR539) which lacks the Dam and Dcm methylases as the activity of BsaBI is blocked when the DNA is methylated by Dam methylase. After digestion, DNA precipitation was carried out to purify the DNA. The digested products were resuspended in deionized water and used to transform T. denticola ATCC 35405 by electroporation.

For the optimal electroporation conditions; the temperature and voltages of electroporation were investigated. In order to reduce the number of samples handled, the cells were only electroporated at 1.8 kV from the 4th trial onwards. After electroporation, the cells were allowed to recover overnight in an anaerobic chamber in pre-reduced OBGM. Cells electroporated with plasmid DNA were plated in OBGM agar supplemented with fresh erythromycin (40 µg/mL) for selection of transformants while cells electroporated without plasmid DNA were plated in non-antibiotic-containing OBGM agar as positive controls to show cell viability. T. denticola ATCC 35405 was not transformable under any of the conditions tested although the cells were viable (Table 3.1), proving the difficulty in the transformation of this strain of T. denticola.

Figure 3.1 Plasmid pCF382. A. Diagram of pCF382 showing the restriction sites. Red arrows indicate the restriction sites used to linearize the plasmid or release the insert from the vector. B. Linearization of pCF382 with PstI. Lane 1, undigested pCF382; Lane 2, digested pCF382 showing the expected product size of 7.4 kbp. Linearization of pCF382 with EcoRI showed a similar profile. C. Digestion of pCF382 with the restriction enzymes EcoRI and BsaBI. Lane 1, undigested pCF382; Lane 2, digested pCF382. Expected product sizes of 2.7 kbp for vector and 4.7 kbp for insert were observed. Agarose gel electrophoresis [0.8% (w/v) in TAE] was performed to monitor the progress of the digestions. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Table 3.1 T. denticola ATCC 35405 transformation using plasmid pCF382 under aerobic conditions.

No. Cell preparation Plasmid Electroporation Electroporated Ermr conditions cell viability colonies Growth phase of Temp. No. of Fresh/frozen RE digestionA Amount Temp. Voltage (CFU/mL)B cells at harvest (°C) washes cells (µg) (°C) (kV) 1 Exponential 4 3 Frozen single 10 RT 1.6 [>3 × 105] 0 2 4 3 Frozen single 10 RT 1.8 [>3 × 105] 0 3 4 3 Frozen single 10 RT 2.0 [>3 × 105] 0 4 4 2 Frozen single 50 4 1.8 [>3 × 105] 0 5 4 2 Frozen double 50 4 1.8 [>3 × 105] 0 6 4 2 Frozen none 50 4 1.8 [>3 × 105] 0 7 4 2 Fresh single 10 4 1.8 [>3 × 105] 0 8 4 2 Fresh double 10 4 1.8 [>3 × 105] 0 9 RT 2 Fresh single 10 RT 1.8 [>3 × 105] 0 10 RT 2 Fresh none 10 RT 1.8 [>3 × 105] 0 11 Stationary RT 2 Fresh single 10 RT 1.8 [>3 × 105] 0 12 RT 2 Fresh none 10 RT 1.8 [>3 × 105] 0 A RE: restriction enzyme; single digestion indicates linearization of the plasmid with EcoRI or PstI; double digestion indicates the release of the recombination cassette from the background vector using EcoRI and BsaBI.

B Cells electroporated without plasmid DNA and plated in non-antibiotic-containing OBGM agar. CFU/mL values in brackets [>3 × 105] represent conditions for which the plates yielded CFU too numerous to count when one-fifth volume of the recovered cells were plated.

3.2.2 Optimization of transformation efficiency of T. denticola strain ATCC 33520 The failure in transforming T. denticola ATCC 35405 with pCF382 under all the conditions tested led to the use of another T. denticola strain for the optimization of the transformation protocol. T. denticola ATCC 33520 was used as it had been reported to have a higher transformation efficiency than ATCC 35405, probably because of the presence of a plasmid in ATCC 33520 which is absent in ATCC 35405 (Caudry et al. 1995) and the absence of the restriction-modification systems found in ATCC 35405 (Bian and Li 2011). The flanking regions in the recombination cassette of pCF382 comprise the genes upstream (TDE2109) and downstream (TDE2111) of the pyrF gene in T. denticola ATCC 35405. Since the pyrF gene and its upstream and downstream genes are found in the genome of T. denticola ATCC 33520 (Table 3.2), the plasmid pCF382 could be used in ATCC 33520 to create a mutant strain lacking the pyrF gene. Using nucleotide alignment, the percentage of nucleotide identity between the upstream genes, TDE2109 and HMPREF9722_RS06015, is 97.6% whilst that between the downstream genes, TDE2111 and HMPREF9722_RS06005, is 95.8%.

Table 3.2 Genes upstream and downstream of TDE2110 in ATCC 35405 and HMPREF9722_RS06010 in ATCC 33520.

T. denticola T. denticola Description ATCC 35405 ATCC 33520 gene designation gene designation pyrF TDE2110 HMPREF9722_ Orotidine-5’-monophosphate RS06010 decarboxlyase 5’ adjacent TDE2109 HMPREF9722_ ATP-dependent RNA gene to pyrF RS06015 helicase, DbpA 3’ adjacent TDE2111 HMPREF9722_ tRNA [guanosine(46)-N7]- gene to pyrF RS06005 methyltransferase, TrmB

Plasmid pCF382 was digested with EcoRI and BsaBI as previously described to release the recombination cassette (4.7 kbp) from the background vector (2.7 kbp; Figure 3.1C). The digested products were purified, resuspended in deionized water and used to transform T. denticola ATCC 33520 by electroporation under a variety of conditions as described in Table 3.3. T. denticola ATCC 33520 cells were grown to exponential phase

(A650 of 0.16-0.28), harvested by centrifugation, washed twice in 10% (v/v) glycerol and

resuspended in 10% (v/v) glycerol. DNA (10 µg) was added to the cells and electroporation was carried out without delay. After electroporation, the cells were allowed to recover overnight in an anaerobic chamber in pre-reduced OBGM. Transformed cells were plated in OBGM agar medium pre-equilibrated to 37°C and supplemented with fresh erythromycin (40 µg/mL) for selection of transformants.

T. denticola ATCC 33520 was not transformable under aerobic conditions when the cells were prepared and electroporated at room temperature (RT), despite the different voltages used for electroporation (Table 3.3). T. denticola ATCC 33520 was transformable under aerobic conditions when the cells were prepared and electroporated at 4°C. Fresh cells were prepared and electroporated on the same day. Frozen cells were prepared, snap- frozen in liquid nitrogen and stored at -80°C until use. Frozen cells were thawed on ice before use. Electroporation at 1.6 kV yielded a similar number of transformants for both fresh and frozen cells. Electroporation at 1.8 kV yielded more transformants with fresh cells than frozen cells. The transformation efficiency of T. denticola ATCC 33520 greatly increased when fresh cells were prepared and electroporated at 4°C under anaerobic conditions. Electroporation at 1.8 kV produced a higher transformation efficiency than 1.6 kV. The transformation efficiency increased by 12-fold with electroporation at 1.6 kV and 15-fold with electroporation at 1.8 kV under anaerobic conditions compared with aerobic conditions (Table 3.3).

Table 3.3 Optimization of T. denticola ATCC 33520 transformation using plasmid pCF382A.

B Electroporation Cell preparation Electroporated conditions Ermr cell viability colonies Temp. Aerobic/anaerobic Fresh/frozen Temp. Voltage (CFU/mL)C (°C) conditions cells (°C) (kV) 1.6 [>3 × 105] 0 RTD Aerobic Fresh RTD 1.8 [>3 × 105] 0

2.0 0 0 1.6 [>3 × 105] 27 4 Aerobic Frozen 4 1.8 [>3 × 105] 5

2.0 [>3 × 105] 4 1.6 1.1 x 107 30 4 Aerobic Fresh 4 7 1.8 1.2 x 10 49 1.6 7.7 x 108 354 4 Anaerobic Fresh 4 8 1.8 6.8 x 10 730 A Ten micrograms of double-digested pCF382 were used for all transformations. B T. denticola cells were harvested at exponential phase and washed twice in 10% (v/v) glycerol. C Cells electroporated without plasmid DNA and plated in non-antibiotic-containing OBGM agar to show viability. Absolute CFU/mL values were obtained from dilutions of the recovered cells. CFU/mL values in brackets [>3 × 105] represent conditions for which the plates yielded CFU too numerous to count when 1/1200 volume of the recovered cells were plated.

3.2.3 Transformation of T. denticola ATCC 35405 with plasmid pHN-911 The optimized protocol developed in T. denticola ATCC 33520 which includes preparation of competent cells under anaerobic conditions was used to transform T. denticola ATCC 35405 with plasmid pCF382. However, the transformation was unsuccessful. TDE0911 has been reported to be a main restriction enzyme involved in the low transformation efficiency of T. denticola ATCC 35405 and the absence of this restriction enzyme in ATCC 33520 is proposed to be the major factor resulting in the higher transformation efficiency of ATCC 33520 (Bian and Li 2011). Therefore, plasmid pHN-911 was constructed in an attempt to create a mutant lacking the TDE0911 gene in T. denticola ATCC 35405.

3.2.3.1 Construction of pHN-911 suicide plasmid The plasmid was generated as described by Bian and Li (2011) as a suicide plasmid consisting of ermAM flanked by DNA corresponding to the genome upstream (5’ flanking region) and downstream (3’ flanking region) of TDE0911 in T. denticola ATCC 35405. The plasmid was generated by SOE PCR using the corresponding primers as listed in Table 2.5. The 5’ and 3’ flanking regions of TDE0911 (5’ TDE0911 and 3’ TDE0911) were amplified from T. denticola ATCC 35405 gDNA, whilst ermAM was amplified from plasmid pCF382. The PCR amplicons with the expected sizes of 751 bp, 695 bp and 760 bp for 5’ TDE0911, 3’ TDE0911 and ermAM respectively (Figure 3.2A) were purified. The purified 3’ TDE0911 was fused with ermAM by SOE PCR (Figure 3.2B). The PCR amplicon with the desired size of 1446 bp was purified and fused by SOE PCR with 5’ TDE0911 to yield a final product (TDE0911::ermAM) of 2166 bp (Figure 3.2C). The purified TDE0911::ermAM was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Five transformed colonies selected on LB Amp plates were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. All transformants examined possessed the desired 2.4 kbp amplicon (Figure 3.3). Plasmid from clone one was sequenced to confirm the fidelity of the recombination cassette in pHN-911. E. coli α-Select harboring pHN-911 was designated ECR761.

Figure 3.2 Generation of plasmid pHN-911. A. PCR amplicons of 5’ TDE0911 (Lane 1; 751 bp), 3’ TDE0911 (Lane 2; 695 bp) and ermAM (Lane 3; 760 bp). B. PCR amplicon of 3’ TDE0911-ermAM (1446 bp). C. PCR amplicon of the final SOE product (TDE0911::ermAM; 2166 bp). D. NotI-digestion of pHN-911 resulting in expected products of 2.2 kbp for insert and 3 kbp for vector. Lane 1, undigested pHN-911; Lane 2, NotI-digested pHN-911. The PCR amplicons and restriction digestion of pHN-911 were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Figure 3.3 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-911. Lanes 1-5, PCR amplicons from E. coli colonies 1-5 generated using primers M13 Forward and M13 Reverse show the expected amplicon size of 2.4 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel. 3.2.3.2 Generation of T. denticola ATCC 35405 allele exchange mutant ∆TDE0911 Plasmid pHN-911 was linearized by digestion with NotI to release the recombination cassette (2.2 kbp) from the background vector (3 kbp; Figure 3.2D). After digestion, DNA precipitation was carried out to remove the enzyme. The digested products were resuspended in deionized water and used to transform T. denticola ATCC 35405 by electroporation. T. denticola ATCC 35405 was grown to exponential phase (A650 of 0.14), harvested by centrifugation, washed twice with 10% (v/v) glycerol and resuspended in 10% (v/v) glycerol. The cells were prepared at 4°C under anaerobic conditions, following the optimized protocol established in T. denticola ATCC 33520 (Section 3.2.2). The cells were electroporated at 4°C at 1.8 kV. T. denticola ATCC 35405 was transformable under anaerobic conditions using plasmid pHN-911 (Table 3.4). T. denticola ATCC 35405 was transformable with as little as 50 ng of plasmid with a high efficiency of 320 CFU/µg of DNA (Table 3.4). Increasing the amount of plasmid for transformation from 50 ng to 10 µg dramatically increased the number of colonies (CFU) but decreased the efficiency to 83 CFU/µg of DNA. Increasing the amount of plasmid for transformation from 10 µg to 50 µg did not significantly increase the number of CFUs and consequently lowered the transformation efficiency markedly.

Thirty T. denticola ATCC 35405 transformants selected on OBGM agar plates containing 40 µg/mL erythromycin were subcultured in OBGM containing 40 µg/mL erythromycin whereupon a small volume of each culture was used for PCR analysis with primers tde910-F and ermAM-R short (Table 2.4) to confirm DNA integration at the correct locus. In the study done by Bian and Li (2011), only eight out of sixty-three (13%) erythromycin-resistant colonies were true positives. However, all transformants examined in this study possessed the desired amplicons of 1.6 kbp (Figure 3.4) and were true positives. Transformant one was designated ECR773.

Table 3.4 The effect of DNA amount on the transformation efficiency of T. denticola ATCC 35405A.

Amount of digested pHN- Ermr colonies/µg of Ermr colonies 911 DNAB 10 µg 834 83.4

20 µg 929 46.5

50 µg 906 18.1

50 ng 16 320 A T. denticola ATCC 35405 cells were prepared and transformed with digested pHN-911 as described in the text. B The transformation efficiency was expressed as Ermr colonies/µg of DNA.

Figure 3.4 PCR to confirm homologous recombination of TDE0911::ermAM with T. denticola ATCC 35405 genome. C shows no PCR amplification product using WT gDNA as template. Lanes 1-30, PCR amplicons from ∆TDE0911 transformants 1-30. The primer pairs used were tde910-F and ermAM-R short. The expected amplicon size of 1.6 kbp was observed in ∆TDE0911 transformants and there was no product in the WT control as expected. The PCR amplicons were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

3.2.4 Generation of T. denticola ATCC 33520 and ATCC 35405 mutants using the optimized protocol Since both of the T. denticola strains were transformable with high efficiency using the protocol established above, the new protocol was used to generate a number of T. denticola mutants. These included T. denticola transformants harboring fluorescent proteins, T. denticola mutants impaired in motility and a T. denticola mutant lacking a putative DfsB (dendritiformis sibling bacteriocin) homologue. The details of the generation of each mutant are described in the following Chapters. In order to generate a fluorescently-labelled T. denticola, the TDE0911 gene was replaced with a gene (evoglow-Bs2-stop or evoglow-C-Bs2-stop) that encodes a flavin mononucleotide-based

fluorescent protein (FbFP). The gene was successfully inserted into the genome of T. denticola ATCC 35405 and replaced the TDE0911 gene. T. denticola motility mutants were generated by deleting the genes involved in the motility of T. denticola including cheY, motA, motB and flgE. All of the mutants were successfully generated in T. denticola ATCC 33520, but only ∆cheY was successfully generated in T. denticola ATCC 35405. The failure to generate ∆motA, ∆motB and ∆flgE in T. denticola ATCC 35405 was not due to the competency of the cells as these cells were successfully transformed with TDE0911::ermAM, the positive control, and cheY::ermAM (Table 3.5). A T. denticola mutant lacking a DfsB homologue was generated by deleting the TDE0659 (herein designated DfsBTd35405) gene, which encodes the putative DfsB homologue. The DfsBTd gene was successfully deleted from the genome of T. denticola ATCC 33520 but not

ATCC 35405. Similar to the motility mutants, the failure to generate ∆DfsBTd in T. denticola ATCC 35405 was not due to the competency of the cells, as the cells were successfully transformed with TDE0911::ermAM (Table 3.5).

100

Table 3.5 Transformation trials to generate ∆motA, ∆motB, ∆flgE and ∆DfsBTd35405 in T. denticola ATCC 35405A.

Recombination Amount of digested Electroporated cell Ermr No. cassette plasmid DNA viability (CFU/mL)B colonies cheY::ermAM 10 µg 4 1 motA::ermAM 10 µg 1.2 x 106 0

motB::ermAM 10 µg 0 10 µg 0 2 motA::ermAM 50 ng 0

10 µg 0 motB::ermAM 4.1 x 107 50 ng 0 10 µg 93 TDE0911::ermAM 50 ng 16 motA::ermAM 10 µg 0 3 motB::ermAM 10 µg 2.8 x 108 0

TDE0911::ermAM 10 µg 136 motA::ermAM 10 µg 0 4 motB::ermAM 10 µg 2.8 x 108 0

TDE0911::ermAM 10 µg 475 motA::ermAM 10 µg 0 5 TDE0911::ermAM 10 µg 4.7 x 106 412

TDE0659::ermAM 10 µg 0 TDE0911::ermAM 10 µg 140 6 3.0 x 106 TDE0659::ermAM 10 µg 0 TDE0911::ermAM 10 µg 242 7 3.9 x 107 flgE::ermAM 10 µg 0 A T. denticola ATCC 35405 cells were prepared using the optimized protocol and transformed with NotI-digested plasmids which had the recombination cassette released from the vector. B Cells electroporated without plasmid DNA and plated in non-antibiotic-containing OBGM agar to show viability.

101

3.2.5 Generation of T. denticola ATCC 35405 ∆TDE0911::kan mutant lacking TDE0911 To investigate whether TDE0911 was responsible for the inability to generate ∆pyrF,

∆motA, ∆motB, ∆flgE and ∆DfsBTd mutants in T. denticola ATCC 35405, TDE0911 was removed and the resulting mutant was transformed with the recombination cassettes. Since all of the recombination cassettes were constructed with the ermAM gene which confers erythromycin resistance to the transformants, aphA2 which confers kanamycin resistance to the transformants was used to select for a TDE0911 knockout in T. denticola ATCC 35405.

3.2.5.1 Construction of the pHN-911kan suicide plasmid Plasmid pHN-911kan was generated as a suicide plasmid consisting of the aphA2 gene driven by the ermAM promoter, flanked by DNA corresponding to the genome upstream and downstream of the TDE0911 gene in T. denticola ATCC 35405. The 5’ and 3’ flanking regions of TDE0911 (5’ TDE0911 and 3’ TDE0911) were amplified from T. denticola ATCC 35405 WT gDNA. The aphA2 gene was amplified from the plasmid pCR4-TOPO, whilst the ermAM promoter was amplified from the shuttle vector pHS17. The PCR amplicons of the expected sizes of 763 bp, 375 bp and 835 bp for 5’ TDE0911, ermAM promoter and aphA2 respectively (Figure 3.5A) were purified. 5’ TDE0911 was fused with the ermAM promoter by SOE PCR to yield a 1098 bp PCR amplicon while aphA2 was fused with 3’ TDE0911 by SOE PCR to yield a 1501 bp PCR amplicon (Figure 3.5B). The PCR amplicons were purified and fused together by SOE PCR for a final product (TDE0911::aphA2) of 2559 bp (Figure 3.5C). The purified TDE0911::aphA2 was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Eight transformant colonies selected on LB Amp plates were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. All transformants examined possessed the desired 2.8 kbp amplicon (Figure 3.6). Plasmid from clone 7 was sequenced to confirm the fidelity of the recombination cassette in pHN-911kan. E. coli α-Select harboring pHN- 911kan was designated ECR769.

102

Figure 3.5 Generation of plasmid pHN-911kan. A. PCR amplicons of 5’ TDE0911 (Lane 1, 763 bp), ermAM promoter (Lane 2, 375 bp) and aphA2 (Lane 3, 835 bp). B. PCR amplicons of 5’ TDE0911-ermAM promoter (Lane 1, 1098 bp) and aphA2-3’ TDE0911 (Lane 2, 1501 bp). C. PCR amplicon of TDE0911::aphA2 (2559 bp). D. NotI-digestion of pHN-911kan showing the expected products of 2.6 kbp for insert and 3 kbp for vector. Lane 1, undigested pHN-911kan; Lane 2, NotI-digested pHN-911kan. The PCR amplicons and restriction digestion of pHN-911kan were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Figure 3.6 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-911kan. Lanes 1-8, PCR amplicons from E. coli colonies 1-8 generated using primers M13 Forward and M13 Reverse show the expected amplicon size of 2.8 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

103

3.2.5.2 Generation of T. denticola ATCC 35405 allele exchange mutant ∆TDE0911::kan Plasmid pHN-911kan was linearized by digestion with NotI to release the recombination cassette (2.6 kbp) from the background vector (3 kbp; Figure 3.5D). After digestion and DNA precipitation, the digested products were resuspended in deionized water and used to transform T. denticola ATCC 35405 by electroporation. Four transformants randomly selected from OBGM agar plates containing 75 μg/mL kanamycin were subcultured in OBGM containing 75 μg/mL kanamycin whereupon a small volume of each culture was used for PCR analysis to confirm DNA integration at the correct locus. All transformants produced the desired amplicons of 2.9 kbp and 2 kbp for the primer pairs tde910- F/tde913-R and tde910-F/Kan-R respectively (Table 2.4; Figure 3.7). Transformant one was designated ECR824.

Figure 3.7 PCR to confirm homologous recombination of TDE0911::aphA2 with T. denticola ATCC 35405 genome. Lanes 1 and 6, PCR amplicons from T. denticola ATCC 35405 WT gDNA as template; Lanes 2-5 and 7-10, PCR amplicons from ∆TDE0911::kan transformants 1-4 as templates; Lanes 1-5, PCR products amplified using primers tde910- F and tde913-R showing the expected amplicon size for WT (2.5 kbp; lane 1) and for ∆TDE0911::kan transformants 1-4 (2.9 kbp; lanes 2-5, respectively); Lanes 6-10, PCR products amplified using primers tde910-F and Kan-R showing no PCR amplicon generated from WT (lane 6) and 2 kbp PCR amplicons from the ∆TDE0911::kan clones 1-4 (lanes 7-10, respectively) as expected. The PCR amplicons were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

3.2.6 Digestion of recombination cassettes with T. denticola lysates In order to determine whether TDE0911 restriction enzyme digestion of the recombination cassettes was responsible for the inability to transform T. denticola ATCC 35405 with these constructs, digestion of the recombination cassettes with T. denticola ATCC 35405, ATCC 33520 and ∆TDE0911::kan cell lysates was carried out. All of the

104

recombination cassettes used for the transformation of T. denticola were included. Digestions of TDE0911::ermAM, TDE0911::aphA2, TDE0911::evoglow::aphA2 and TDE0911::clos-evoglow::aphA2 recombination cassettes with cell lysates from ATCC 33520 and ∆TDE0911::kan were not carried out as these recombination cassettes were not used to transform these two T. denticola strains. The recombination cassettes were amplified from their respective plasmids by PCR using the primer pair M13 Forward/M13 Reverse. The PCR products with the desired size were purified (Figure 3.8). Since pCF382 was constructed in a different background vector and there was no readily available primer to amplify the recombination cassette, the recombination cassette was released from the vector using restriction digestion. Both the recombination cassette and vector were subjected to digestion with T. denticola lysates (Figure 3.9).

Figure 3.8 Recombination cassettes amplified by PCR. The recombination cassettes were amplified by PCR from their respective plasmids using the primers M13 Forward and M13 Reverse and examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. Lane 1, TDE0911::ermAM (expected 2.4 kbp); Lane 2, cheY::ermAM (expected 3.0 kbp); Lane 3, TDE0911::evoglow::aphA2 (expected 3.3 kbp); Lane 4, TDE0911::clos- evoglow::aphA2 (expected 3.3 kbp); Lane 5, motA::ermAM (expected 3.3 kbp); Lane 6, motB::ermAM (expected 3.1 kbp); Lane 7, flgE::ermAM (expected 3.2 kbp); Lane 8, TDE0911::aphA2 (expected 2.8 kbp); Lane 9, TDE0659::ermAM (expected 2.9 kbp). The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

Preparation of the T. denticola cell lysates required growth of the cells to exponential phase, harvesting by centrifugation, resuspension in PBS and sonication of 1.5 x 109 cells. After lysis and centrifugation, the supernatants were collected as crude cell lysates. Cell lysates were mixed with 1 µg of each recombination cassette and incubated at 37°C. A sample of the reaction mixture collected after 30 min and 2 h of incubation was analyzed by agarose gel electrophoresis. It was observed that all of the recombination cassettes, except TDE0911::ermAM and TDE0659::ermAM, were digested by T. denticola ATCC 35405 lysate within 30 min of incubation. In comparison, digestion of the motA::ermAM,

105

motB::ermAM, flgE::ermAM, cheY::ermAM, TDE0659::ermAM and pCF382 cassettes by T. denticola ATCC 33520 lysate occurred at a much slower rate, as the bands became partially-digested only after 2 h of incubation. An absence of TDE0911 completely abolished the digestion of the recombination cassettes by ATCC 35405 cell lysate even after 2 h of incubation (Figure 3.9 and Figure 3.10), showing that TDE0911 is the main contributor to the digestion of the unmethylated recombination cassettes by ATCC 35405. Since the plasmids used for the transformation of T. denticola were produced in E. coli α-select cells which are capable of methylation, the experiments were repeated with methylated, NotI-digested plasmids pHN-motA, pHN-motB, pHN-flgE, pHN-cheY and pHN-0659. Similar results as the unmethylated, PCR-amplified recombination cassettes were obtained from digestions by T. denticola ATCC 35405 and ∆TDE0911::kan lysates, showing that TDE0911 is also involved in the digestion of the methylated recombination cassettes by ATCC 35405. However, the methylated plasmids were not cleaved by ATCC 33520 lysate even after 2 h of incubation (Figure 3.11).

Figure 3.9 Digestion of the restriction enzyme-digested pCF382 with T. denticola ATCC 35405, ATCC 33520 and ∆TDE0911::kan cell lysates. Exponential-phase T. denticola cells (1.5 x 109 cfu) were harvested and lysed by sonication. The crude cell lysates collected after centrifugation were mixed with 1 µg of restriction enzyme-digested pCF382. The reaction mixture was incubated at 37°C for 30 min and 2 h before analysis on agarose gels. C indicates control with no cell lysate, 35405 indicates T. denticola ATCC 35405 cell lysate, 33520 indicates T. denticola ATCC 33520 cell lysate and ∆TDE0911::kan indicates ∆TDE0911::kan cell lysate. The digestion was analyzed by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

106

Figure 3.10 Digestion of the unmethylated recombination cassettes with T. denticola ATCC 35405, ATCC 33520 and ∆TDE0911::kan cell lysates. Exponential-phase T. denticola cells (1.5 x 109 cfu) were harvested and lysed by sonication. The crude cell lysates collected after centrifugation were mixed with 1 µg of each PCR-amplified recombination cassette. The reaction mixture was incubated at 37°C for 30 min and 2 h before analysis on agarose gels. C indicates control with no cell lysate, 35405 indicates T. denticola ATCC 35405 cell lysate, 33520 indicates T. denticola ATCC 33520 cell lysate and ∆TDE0911::kan indicates ∆TDE0911::kan cell lysate. The digestion was analyzed by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

107

Figure 3.11 Digestion of the methylated, NotI-digested pHN-motA, pHN-motB, pHN-flgE, pHN-cheY and pHN-0659 with T. denticola ATCC 35405, ATCC 33520 and ∆TDE0911::kan cell lysates. Exponential-phase T. denticola cells (1.5 x 109 cfu) were harvested and lysed by sonication. The crude cell lysates collected after centrifugation were mixed with 5 µg of each NotI-digested plasmid. The reaction mixture was incubated at 37°C for 30 min and 2 h before analysis on agarose gels. C indicates control with no cell lysate, 35405 indicates T. denticola ATCC 35405 cell lysate, 33520 indicates T. denticola ATCC 33520 cell lysate and ∆TDE0911::kan indicates ∆TDE0911::kan cell lysate. The digestion was analyzed by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

108

3.2.7 Generation of mutants in T. denticola ATCC 35405 ∆TDE0911::kan Since the deletion of TDE0911 abolished the digestion of the recombination cassettes flgE::ermAM, motA::ermAM, motB::ermAM, cheY::ermAM and pCF382, these recombination cassettes were used to transform ∆TDE0911::kan (ECR824) in order to generate the deletion mutants. Although TDE0911 was not involved in the digestion of TDE0659::ermAM as shown above, the recombination cassette was nevertheless tested in ∆TDE0911::kan. The cells were prepared and electroporated using the protocol described previously (Section 3.2.3.2). Transformation with cheY::ermAM and flgE::ermAM were successful but transformation with all other recombination cassettes was unsuccessful (Table 3.6). The transformation with cheY::ermAM had increased from four transformants in ATCC 35405 to 223 transformants in ∆TDE0911::kan. The failure to transform ∆TDE0911::kan with motA::ermAM, motB::ermAM, TDE0659::ermAM and pCF382 was not due to the competency of the cells, as the cells were successfully transformed with flgE::ermAM and cheY::ermAM (Table 3.6). These results suggest that TDE0911 is not the only factor contributing to the low transformation efficiency of T. denticola ATCC 35405. The mutants that had been successfully generated in each strain of T. denticola used in this study are listed in Table 3.7.

Table 3.6 Transformation of ∆TDE0911::kanA.

Recombination cassette Ermr colonies motA::ermAM 0 motB::ermAM 0 cheY::ermAM 223 flgE::ermAM 2 TDE0659::ermAM 0 pCF382 0 A T. denticola ∆TDE0911::kan cells were prepared using the optimized protocol and transformed with digested plasmids which had the recombination cassette released from the vector.

109

Table 3.7 List of mutants generated in T. denticola ATCC 35405, ∆TDE0911::kan and ATCC 33520.

T. denticola strain Mutant ∆TDE0911 ∆TDE0911::kan

ATCC 35405 ∆cheY35405

ATCC 35405::FbFPevoglow-Bs2-stop

ATCC 35405::FbFPevoglow-C-Bs2-stop ∆TDE0911::kan ∆flgE ATCC 35405 ∆TDE0911::kan ∆TDE0911::kan ∆cheY

∆pyrF33520

∆flgE33520

ATCC 33520 ∆cheY33520

∆motA33520

∆motB33520

∆DfsBTd33520

110

3.3 DISCUSSION An efficient genetic manipulation system is important for the functional analysis of genes and their gene products and is a vital tool in the investigation of the molecular basis of the pathogenicity of a bacterium. The ability to modify the genome of a bacterium enables the construction of mutant strains lacking specific genes which may then be used to determine the potential virulence factors in the bacterium. T. denticola is well-known for its extremely low transformation efficiency which impedes the characterization of the molecular details of its virulence. The slow growth, fastidious nutrient requirements, strain-limited shuttle plasmid and the presence of limited selectable markers have further slowed down progression in the genetic analysis of T. denticola (Kuramitsu et al. 2005b, Bian and Li 2011, Godovikova et al. 2015, Li et al. 2015). Although a transformation protocol for T. denticola was developed more than 20 years ago (Li and Kuramitsu 1996, Li et al. 1996), there are only a small number of existing T. denticola mutants, suggesting the lack of reproducibility of the original protocol. The aim of this study was to develop a reliable and reproducible genetic manipulation system for T. denticola. Using T. denticola strain ATCC 33520 which was reported to be more accessible to genetic manipulations than the type strain ATCC 35405 (Bian and Li 2011), an optimized transformation protocol which yielded a high number of transformants was developed.

Among the conditions tested, preparation of electrocompetent cells under anaerobic conditions was the most important factor in improving the transformation efficiency of T. denticola ATCC 33520. This result is consistent with a recent study which reported a 100-fold higher transformation efficiency in T. denticola ATCC 35405 with a shuttle plasmid using a modified protocol which included preparation of electrocompetent cells under anaerobic conditions (Godovikova et al. 2015). Since T. denticola is a strict anaerobe which requires an anaerobic environment for survival (Lai and Chu 2008), a minimal exposure to oxygen during the preparation of electrocompetent cells might have increased cell viability and thus resulted in a higher number of transformants. This is reflected in the electroporated cell viability which showed a higher number of surviving cells when they were prepared under anaerobic conditions as compared to aerobic conditions (Table 3.3). For this reason, the wash buffer and growth medium used for the recovery of electroporated cells were pre-reduced in the anaerobic chamber for at least 16 h. Further, the cells were only washed twice instead of thrice as used in the original protocol due to the fragility of T. denticola cells. Cellular damage to fragile spirochetes

111

during sample processing is a well-noted phenomenon (Salvador et al. 1987, Wardle 1997) so minimal handling of the cells is preferable. Similar to the modified protocol used by Godovikova et al. (2015), electroporation was carried out immediately after the addition of DNA to avoid DNA degradation by endonucleases. The OBGM agar medium was pre-equilibrated to 37°C before plating of the electroporated cells as T. denticola is sensitive to heat shock.

Besides low transformation efficiency, the generation of T. denticola mutants is also complicated by high “false-positive” transformation frequency. In this study, however, there was no “false-positive” transformation result observed from the various transformations done using cells prepared under anaerobic conditions (Figure 3.4 and additional data not shown). The prolonged incubation period (~7-14 days) required for the growth of T. denticola transformants may cause the antibiotic to become unstable, resulting in a sublethal concentration that promotes the development of antibiotic resistance in WT T. denticola. This is especially true for erythromycin, which is commonly used in the selection of T. denticola transformants, as it is a bacteriostatic antibiotic that does not kill the bacteria but stops their growth. When it becomes unstable, the growth of T. denticola WT cells may resume, thus resulting in “false-positive” transformation results. Therefore, it is important to prepare fresh antibiotics and add them to the OBGM agar medium just before use.

A variety of bacterial stress responses as a result of exposure to environmental stresses such as nutrient starvation, oxidative stress, heat stress and ribosomal stress have been linked to antimicrobial resistance (Poole 2012). It was proposed that the induction of a stress response in Lactococcus lactis is associated with its resistance to ribosomally active antibiotics (Dorrian et al. 2011). Erythromycin targets the bacterial ribosome and a 23S rRNA point mutation is reported to contribute to erythromycin resistance in T. denticola (Lee et al. 2002b). 23S rRNA mutations are widely reported to confer macrolide resistance in a wide variety of bacterial species (Vester and Douthwaite 2001). It is possible that the exposure of T. denticola cells to oxygen during the competent cell preparation induces stress responses that contribute to mutational changes in the ribosomal RNA and prevent the binding of erythromycin, thus altering the natural susceptibility of the bacterium to erythromycin. This is further supported by the observation in this study that T. denticola ATCC 33520 cells prepared under aerobic conditions produced only 12 true transformants out of the 35 Ermr colonies examined

112

whereas all transformants using cells prepared under anaerobic conditions were routinely true positives (data not shown).

During the optimization process, it was observed that T. denticola ATCC 33520 cells prepared and electroporated at room temperature were not transformable although they were viable, except those electroporated at 2.0 kV (Table 3.3). Room temperature preparation and electroporation of electrocompetent cells were shown to increase the transformation efficiency of several bacteria (Lofblom et al. 2007, Tu et al. 2016), probably by facilitating DNA entry into the cell through membranes that are more permeable at higher temperatures. Room temperature allows for a higher membrane fluidity and may be useful in the transient opening of the cell membranes upon electroporation. However, the higher membrane fluidity also results in faster sealing of the transient pores in the membrane. At low temperature, the pores in the membrane may remain open for a longer time due to the reduced movement of the lipids. In the case of T. denticola where a high amount of DNA is usually used for the transformation, a longer opening of the pores in the cell membranes may be beneficial to DNA entry into the cells. Besides that, the reduced activity of restriction and degradative enzymes in the cells at 4°C might also increase the efficiency of transformation. Furthermore, the periplasmic flagella which wrap around the cell body may be less active at low temperature and thus causes less interference for DNA entry into the cells. At a lower voltage, fresh and frozen cells gave similar numbers of transformants but at a higher voltage, fresh cells gave more transformants than frozen cells, suggesting that frozen cells might be more susceptible to the higher voltage.

Although extra care was taken in the treatment of cells and maintenance of anaerobic conditions during preparation of competent cells, a high variability in cell viability and transformation efficiency were observed in the transformation of T. denticola ATCC 35405 by TDE0911::ermAM (Table 3.5). The variability in transformation efficiency was not due to cell viability as the same number of viable cells produced different number of transformed cells (see no. 3 and 4 of Table 3.5). The batch to batch variation in the transformation efficiency of ATCC 35405 cells is also observed in previous studies (Lux et al. 2002, Godovikova et al. 2015) but the reason for this is unclear. It could be caused by factors outside of practical handling, such as slight variations in the anaerobic environment due to a high usage of the anaerobic chamber or slight variations in the concentration and purity of the transforming DNA.

113

This study has further confirmed that T. denticola ATCC 33520 is more amenable to transformation than ATCC 35405. The digested plasmid pCF382 which was readily transformed into ATCC 33520 could not be transformed into ATCC 35405 under the many conditions tested in this study and in other studies done in this laboratory (K. H. Tan, personal communication). All mutants were successfully generated in ATCC 33520 but not in ATCC 35405. Based on the results obtained with the controls, there was no problem with cell competency. TDE0911, one of the type II restriction endonucleases found in ATCC 35405, was shown to recognize and digest unmethylated DNA (Bian and Li 2011). In this study, TDE0911 was shown to be involved in the digestion of both methylated and unmethylated recombination cassettes since the deletion of TDE0911 abolished the digestion of both methylated and unmethylated recombination cassettes. Although deleting TDE0911 improved the transformation with cheY::ermAM and flgE::ermAM, the ∆TDE0911::kan mutant was still unable to be transformed by the recombination cassettes motA::ermAM, motB::ermAM and pCF382, suggesting that there are other factors which contribute to the inability to generate these mutants in ATCC 35405. This is further supported by the fact that TDE0659::ermAM was not digested by ATCC 35405 lysate yet was unable to be transformed into ATCC 35405.

Recombination cassette and locus of transformation may play a role in the transformation efficiency of ATCC 35405. A study on the transformation of Xylella fastidiosa showed that increasing the size of non-homologous region in the recombination cassette decreases the recombination efficiency (Kung et al. 2013). However, the size of non-homologous regions in the recombination cassettes is unlikely to be the main factor in the unsuccessful transformation of T. denticola ATCC 35405 by motA::ermAM, motB::ermAM, TDE0659::ermAM and pCF382. This is because cheY::ermAM, which has a similar size of non-homologous region as TDE0659::ermAM, and TDE0911::evoglow::aphA2, which has a similar size of non-homologous region as motA::ermAM and motB::ermAM, were successfully used to transform T. denticola ATCC 35405. The efficiency of gene inactivation may be locus specific. This is supported by a study which reported that the inactivation of TDE1430 (which encodes a putative β-1,4-galactosyltransferase) has a higher efficiency compared to the inactivation of other genes including flgE (Y. Li et al. 2015). The secondary structures of DNA such as hairpins or loops may affect recombination efficiency. On the one hand, the presence of DNA secondary structures could interfere with recombination at certain loci. For example, DNA secondary

114

structures were related to the disruptions in PCR, sequencing and BAC (bacterial artificial chromosome) recombineering within the murine Foxd3 locus (Nelms and Labosky 2011). On the other hand, DNA secondary structures were known to induce recombination, for instance in recombination of the var gene in Plasmodium falciparum during DNA replication (Sander et al. 2014). Another example is pilE recombination in Neisseria gonorrhoeae which requires the presence of a specific DNA structure called a G- quadruplex (G4) upstream of the pilE locus (Cahoon and Seifert 2009). However, further analyses are needed to determine the presence of DNA secondary structures in the T. denticola loci that give a lower transformation efficiency and to investigate the specific effect of these DNA secondary structures on homologous recombination at these loci.

The other factor that may potentially be involved in the successful transformation of ATCC 33520 but not ATCC 35405 is the presence of an endogenous plasmid in ATCC 33520. The plasmid, pTD1, is 2.6 kbp in size and its sequence revealed two open reading frames, ORF A and ORF B (Ivic et al. 1991, Caudry et al. 1995). While ORF A was determined to be a replicative protein required for the replication of the plasmid, the function of ORF B remained unclear (MacDougall et al. 1992). However, when submitted to PSI-BLAST, the protein sequence of ORF B matched with a group of plasmid recombination enzymes or recombinases from different bacteria including E. coli, Salmonella enterica and Klebsiella pneumoniae (Table 3.8). Since recombinases are enzymes involved in homologous recombination (Kowalczykowski et al. 1994), it is thus speculated that the presence of pTD1 in ATCC 33520 helps with the transformation of this T. denticola strain and the absence of this plasmid in ATCC 35405 reduces the transformability of this T. denticola strain. Further study is needed to confirm that the presence of pTD1 increases the transformation efficiency of T. denticola.

115

Table 3.8 Top ten hits of ORF B homologues identified by PSI-BLAST.

Query Identity Protein accession cover E value Description Organism (%) (%) Plasmid WP_052318681.1 98 3e-104 30 recombination E. coli enzyme WP_069190824.1 98 8e-104 30 Recombinase E. coli

Plasmid WP_095530623.1 98 2e-103 30 recombination E. coli enzyme Plasmid WP_079961789.1 98 4e-103 30 recombination S. enterica enzyme Plasmid K. WP_094327214.1 98 7e-103 30 recombination pneumoniae enzyme Plasmid Shigella WP_094323980.1 98 1e-102 30 recombination sonnei enzyme Plasmid WP_097312066.1 98 1e-102 29 recombination E. coli enzyme Plasmid Shigella WP_050597603.1 98 4e-102 30 recombination flexneri enzyme Plasmid Ancylostoma KIH50631.1 97 7e-102 30 recombination duodenale enzyme Plasmid WP_072749431.1 98 7e-102 30 recombination E. coli enzyme

116

3.4 CONCLUSION In conclusion, this study has developed an optimized transformation protocol for T. denticola which yielded a high transformation efficiency with low “false-positive” transformation results. Preparation of electrocompetent cells under anaerobic conditions was the most important factor in the transformation of T. denticola. This finding has affirmed the results from a recently published study which showed that minimum exposure to oxygen in pre- and post-electroporation steps is important in the transformation of T. denticola. This study also showed that the activity of the restriction enzyme TDE0911 was not the main factor for the low transformation efficiency of T. denticola ATCC 35405. Other factors such as the presence of DNA secondary structures may complicate the homologous recombination in certain loci of T. denticola genome. The higher transformation efficiency of T. denticola ATCC 33520 than ATCC 35405 may be attributed to its naturally occurring plasmid that might encode for a recombinase that enables homologous recombination. These findings expand the current knowledge on the genetic manipulation of T. denticola and will be useful for the future molecular analysis of T. denticola virulence. Using the optimized protocol, several T. denticola mutants harboring heterologous genes or lacking specific genes were generated and used in the subsequent studies.

117

CHAPTER 4 INVESTIGATION OF BIOFILM DEVELOPMENT IN REAL TIME USING FLUORESCENTLY- LABELLED T. DENTICOLA AND P. GINGIVALIS

118

4.1 INTRODUCTION Real-time visualization of biofilm development is useful to provide an insight of how biofilms develop over time. Although it is possible to capture different developmental stages of biofilms by harvesting the biofilms grown to different time points and visualizing them with fluorescent in situ hybridization (FISH) probes, a real-time examination that allows for the continual examination of biofilms will provide more details on the complexity and dynamics of the biofilms. Besides that, the use of FISH may pose a problem in disrupting the biofilm structure in the process of fixing the samples with chemicals prior to the probe hybridization, thus making time-course studies difficult (Franklin et al. 2015). Real-time biofilm visualization is also beneficial in saving time and resources as it removes the need to fix and stain the biofilms. A real-time examination of polymicrobial biofilms made up of two or more bacteria is also useful in visualizing the changes in the spatial and temporal distributions of the different bacteria in the biofilms over time. The most common method for real-time biofilm study is the use of fluorescent proteins combined with confocal laser scanning microscopy (CLSM). CLSM is ideal for imaging biofilms, which have an extensive three-dimensional (3D) structure, as there is no out-of-focus haze that is typical for 3D specimens imaged using conventional light or epifluorescence microscopy (Franklin et al. 2015). The use of fluorescent proteins is beneficial as they are produced by the bacteria themselves, normally non-toxic and their ability to fluoresce in the absence of any chemical cofactors, proteins or substrates (Chalfie et al. 1994) helps to eliminate the introduction of extra variables that may affect the results.

FISH staining has been the most common method used to visualize and quantitate the biofilms formed by anaerobic bacteria in flow cell assays. This is partly because of the lack of equipment to maintain an anaerobic environment for the strict anaerobic bacteria while viewing their biofilm development in real-time. All of the microscopes are set up in aerobic environment that is unfavorable for the growth of anaerobic bacteria including T. denticola and P. gingivalis. The biofilms of T. denticola and P. gingivalis take days to mature (Kuramitsu et al. 2005a, Zainal-Abidin et al. 2012, Zhu et al. 2013) and it is necessary that they are maintained under anaerobic conditions at all times. Recent advancements in microscopy technology has allowed for the use of microscopes equipped with incubation chambers that can be maintained at 37°C under CO2 gas. This is a major improvement that is beneficial for the real-time biofilm study of anaerobic bacteria. The

119

other hurdle in the real-time study of anaerobic bacteria is the lack of a fluorescent protein that can be used under anaerobic conditions. For many years, genetically encoded green fluorescent protein (GFP) from the jellyfish Aequorea victoria and its derivatives have been used for real-time imaging of living specimens. They have been used in a broad range of different organisms ranging from bacteria to mammals. However, their requirement for oxygen as a co-factor for the synthesis of their chromophores limited their utility in the study of anaerobic bacteria which are maintained in an oxygen-free environment (Drepper et al. 2007).

The development of the ‘‘oxygen-independent’’ flavin mononucleotide (FMN)-based fluorescent proteins (FbFPs) has been a major breakthrough in the real-time study of anaerobic bacteria. They were engineered based on the blue-light photoreceptors YtvA from Bacillus subtilis or SB2 from Pseudomonas putida (Drepper et al. 2007). These fluorescent proteins are small (∼11–16 kDa), form either monomers or dimers and harbor the conserved LOV (light oxygen voltage) domain, which belongs to the structurally conserved PAS (PerArntSim) superfamily (Drepper et al. 2007, Drepper et al. 2013). The LOV domain predominantly binds FMN, which is ubiquitously distributed and can be provided by any host organism, as a chromophore. FbFPs show a cyan-green fluorescence

(λmax = 495 nm) upon excitation with blue light (λmax = 450 nm) (Drepper et al. 2013). Unlike GFP and its derivatives, FbFPs are able to fluoresce under both aerobic and anaerobic conditions (Drepper et al. 2007). FbFPs were successfully used as an in vivo fluorescence reporter protein in a number of organisms including Rhodobacter capsulatus, Roseobacter denitrificans, E. coli, the facultative anaerobic yeasts Saccharomyces cerevisiae and Candida albicans as well as the obligate anaerobes P. gingivalis, Bacteroides fragilis and Clostridium cellulolyticum (Drepper et al. 2007, Piekarski et al. 2009, Choi et al. 2011, Lobo et al. 2011, Drepper et al. 2013). Recently, a study had also successfully generated a T. denticola strain expressing FbFP from a shuttle plasmid (Godovikova et al. 2015). The ability to visualize these FbFP-labelled anaerobic bacteria including P. gingivalis, B. fragilis and T. denticola inside human host cells (Choi et al. 2011, Lobo et al. 2011, Godovikova et al. 2015) has opened up possibilities for the applications of FbFPs in the study of complex anaerobic processes including host–pathogen interactions during infection.

T. denticola and P. gingivalis have been shown to form polymicrobial biofilms synergistically in flow cell assays. It was observed that the mature polymicrobial biofilms

120

formed by T. denticola, P. gingivalis and Tannerella forsythia were dominated by P. gingivalis and T. denticola, with the former having a higher biomass, while the immature biofilms were dominated by T. forsythia and T. denticola (Zainal-Abidin et al. 2012, Zhu et al. 2013). This shows a dynamic shift in species proportions over time as the biofilms develop. T. forsythia was present in the mature polymicrobial biofilms in extremely low abundance and the dual-species biofilms formed by P. gingivalis and T. denticola showed a similar synergistic effect when compared with three-species biofilms, suggesting that T. forsythia does not interact closely with P. gingivalis and T. denticola (Zainal-Abidin et al. 2012, Zhu et al. 2013). Therefore, in this study, the dual-species biofilms made up of only P. gingivalis W50 and T. denticola ATCC 35405 were investigated. In order to examine the dual-species biofilm development in real time, the aim of this study was to generate fluorescently labelled P. gingivalis W50 and T. denticola ATCC 35405 by expressing FbFP that is able to fluoresce under anaerobic conditions. The protocol optimized in the previous chapter was used for the generation of fluorescently labelled T. denticola ATCC 35405. The successful generation of fluorescently labelled P. gingivalis W50 and T. denticola ATCC 35405 will also be useful in the subsequent experiments for the comparison of the biofilm formation abilities of WT T. denticola and its isogenic mutants.

121

4.2 RESULTS 4.2.1 Generation of a T. denticola ATCC 35405 strain expressing FbFP (ATCC 35405::FbFP) 4.2.1.1 Construction of the pHN-TdEvo suicide plasmid Plasmid pHN-TdEvo was generated as a T. denticola suicide plasmid where a recombination cassette consisting of the msp promoter, evoglow-Bs2-stop and aphA2 flanked by DNA corresponding to the genome upstream and downstream of the TDE0911 gene in T. denticola ATCC 35405 was cloned into pGEM®-T Easy. The 5’ and 3’ flanking regions of TDE0911 (5’ TDE0911 and 3’ TDE0911) as well as the msp promoter were amplified from T. denticola ATCC 35405 WT gDNA whilst aphA2 was amplified from the plasmid pCR4-TOPO. evoglow-Bs2-stop was amplified from the plasmid pGLOW-Bs2-stop. The PCR amplicons with the expected sizes of 763 bp, 706 bp, 261 bp, 454 bp and 979 bp for 5’ TDE0911, 3’ TDE0911, msp promoter, evoglow-Bs2-stop and aphA2 respectively were purified (Figure 4.1A). The PCR amplicon with the desired size of 1645 bp from the fusion of aphA2 with 3’ TDE0911 (Figure 4.1B) was purified and fused with evoglow-Bs2-stop by SOE PCR (Figure 4.1C). Non-specific amplifications were observed with the fusion of 5’ TDE0911 with msp promoter (984 bp) and fusion of aphA2-3’ TDE0911 with evoglow-Bs2-stop (2059 bp) (Figure 4.1C). Thus the PCR amplicons with the desired size (as indicated by the red boxes in Figure 4.1C) were purified from the agarose gel and fused by SOE PCR to make the final product (TDE0911::evoglow::aphA2) of 3003 bp (Figure 4.1D). The purified TDE0911::evoglow::aphA2 was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Of the transformed colonies selected on LB Amp plates, one was analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. The transformant possessed the desired 3.3 kbp amplicon and was designated ECR763 (Figure 4.2). Plasmid from ECR763 was subjected to nucleotide sequencing to confirm the fidelity of the recombination cassette and was designated pHN-TdEvo. When illuminated with blue LED light and viewed using confocal laser scanning microscope, the colonies and cells of ECR763 fluoresced brightly (Figure 4.3).

122

Figure 4.1 Generation of plasmid pHN-TdEvo. A. PCR amplicons of 5’ TDE0911 (Lane 1; 763 bp), 3’ TDE0911 (Lane 2; 706 bp), msp promoter (Lane 3; 261 bp), evoglow- Bs2-stop (Lane 4; 454 bp) and aphA2 (Lane 5; 979 bp). B. PCR amplicon of aphA2-3’ TDE0911 (1645 bp). C. PCR amplicons of 5’ TDE0911-msp promoter (Lane 1; 984 bp) and evoglow-Bs2-stop-aphA2-3’ TDE0911 (Lane 2; 2059 bp). D. PCR amplicon of the final SOE product, TDE0911::evoglow::aphA2 (3003 bp). E. NotI-digestion of pHN- TdEvo where 3 kbp was expected for both insert and vector. Lane 1, undigested pHN- TdEvo; Lane 2, NotI-digested pHN-TdEvo. The red boxes indicate DNA purified from the agarose gel. The PCR amplicons and NotI-digestion of pHN-TdEvo were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Figure 4.2 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicon from one E. coli colony harboring pHN-TdEvo using primers M13 Forward and M13 Reverse. The expected PCR amplicon is 3.3 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

123

Figure 4.3 Expression of FbFP in E. coli. E. coli carrying plasmids pHN-TdEvo (ECR763) and pHN-911 (ECR761) were grown on LB agar with appropriate antibiotics. The agar plates were then photographed under UV light (470 nm wavelength excitation filter; 515 nm wavelength emission filter) to visualize FbFP expression. A drop of each ECR763 and ECR761 culture was placed on a microscope slide and photographed under UV light to visualize FbFP expression. The photos of E. coli colonies and cultures were taken with a Fujifilm LAS-3000 Imager. ECR763 and ECR761 cells washed twice in 0.85% (w/v) NaCl were visualized using the confocal microscope.

124

4.2.1.2 Generation of a T. denticola ATCC 35405 mutant expressing FbFP from

evoglow-Bs2-stop (ATCC 35405::FbFPevoglow-Bs2-stop) Plasmid pHN-TdEvo was linearized by digestion with NotI to release the recombination cassette (3 kbp) from the background vector (3 kbp; Figure 4.1E). Since the size of the recombination cassette was similar to that of the background vector, their bands were unable to be resolved on an agarose gel (Figure 4.1E). The digested products were purified by DNA precipitation and resuspended in deionized water and used to transform T. denticola ATCC 35405 by electroporation. Twelve transformants selected on OBGM agar plates containing 75 µg/mL kanamycin were grown in OBGM containing 75 µg/mL kanamycin and a small volume of each culture was used for PCR analysis to confirm DNA integration at the correct locus. All transformants examined possessed the desired amplicons of size 3.4 kbp for the primer pair tde910-F/tde913-R (Table 2.4; Figure 4.4).

Transformant one, T. denticola ATCC 35405::FbFPevoglow-Bs2-stop 1, was designated ECR774.

Figure 4.4 PCR to confirm homologous recombination of TDE0911::evoglow::aphA2 with T. denticola ATCC 35405 genome. A. PCR amplicon from T. denticola ATCC 35405 gDNA control using primers tde910-F and tde913-R showing the expected size of 2.5 kbp. B. PCR amplicons from ATCC 35405::FbFPevoglow-Bs2-stop transformants 1-12 using primer pair tde910-F/tde913-R showing the expected size of 3.4 kbp. The PCR amplicons were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

125

4.2.1.3 Visualization of T. denticola ATCC 35405 and ATCC 35405::FbFPevoglow-Bs2-

stop fluorescence using confocal microscopy

To check for fluorescence from T. denticola ATCC 35405 and ATCC 35405::FbFPevoglow-

Bs2-stop transformants 1 and 2, the cells were grown to exponential phase, washed twice in 0.85% (w/v) NaCl and resuspended in 0.85% (w/v) NaCl. The cells were viewed under a confocal microscope (Figure 4.5). The fluorescence of ATCC 35405::FbFPevoglow-Bs2-stop 1 and 2 was similar to that of the background fluorescence emitted by WT. Furthermore, the fluorescence observed from T. denticola ATCC 35405 and ATCC 35405::FbFPevoglow-

Bs2-stop 1 and 2 was blurry with no distinct cell shape visible, unlike the fluorescence observed from T. denticola ATCC 35405 upon staining with the LIVE/DEAD™ BacLight™ Bacterial viability kit where distinct cells were observed. Similar results were observed from cells fixed using 4% (v/v) paraformaldehyde (PFA) in PBS (data not shown).

126

Figure 4.5 Visualization of T. denticola ATCC 35405, ATCC 35405::FbFPevoglow-Bs2- stop1 and 2 using confocal microscopy. T. denticola cells washed twice in 0.85% (w/v) NaCl were visualized using the confocal microscope. For T. denticola WT stained with BacLight™, the cells were washed twice in 0.85% (w/v) NaCl then stained with BacLight™ before visualization. Images shown were obtained using a 63x oil immersion lens under laser excitation (488 nm wavelength) to visualize FbFP (ATCC 35405::FbFPevoglow-Bs2-stop 1 & 2) and SYTO™ 9 Green Fluorescent Nucleic Acid stain (WT stained with BacLight™).

127

4.2.1.4 Immunoblot analyses of FbFP expression from T. denticola ATCC 35405

and ATCC 35405::FbFPevoglow-Bs2-stop transformants 1 and 2 An immunoblot was carried out to check for the expression of FbFP (evoglow) from the whole cell lysates of ATCC 35405::FbFPevoglow-Bs2-stop transformants 1 and 2. E. coli harboring plasmid pHN-TdEvo (ECR763) was used as a positive control as it was found to express the FbFP while E. coli harboring plasmid pHN-911 (ECR761) was used as a negative control. The expected size of FbFP was 16 kDa. A band between 14 and 19 kDa was observed only in the lane loaded with ECR763 but not in all the other lanes on the immunoblot, indicating that FbFP was not expressed in ATCC 35405::FbFPevoglow-Bs2-stop 1 and 2 (Figure 4.6B). The FbFP was expressed in ECR763 at a level detectable only by immunoblot as the band was not observed on an SDS-PAGE gel (Figure 4.6A). A band of approximately 14 kDa was observed on the SDS-PAGE gel in the lane loaded with the negative control strain ECR761, however this was not FbFP as pHN-911 does not contain the gene for FbFP expression. The anti-evoglow antibody had poor specificity as it cross- reacted with mid to high molecular weight proteins from wild-type T. denticola ATCC 35405.

128

Figure 4.6 Analysis of FbFP expression from T. denticola ATCC 35405 and ATCC 35405::FbFPevoglow-Bs2-stop 1-2 using immunoblots. Whole cell lysates of E. coli harboring plasmid pHN-TdEvo (ECR763) and plasmid pHN-911 (ECR761), T. denticola ATCC 35405 and transformants ATCC 35405::FbFPevoglow-Bs2-stop 1-2 were analyzed via A. SDS-PAGE using NuPAGE® 12% Bis-Tris gel and B. immunoblotting on a nitrocellulose membrane. Each T. denticola sample (2.5 x 108 cells) was lysed by sonication before being loaded on the gel. M indicates the SeeBlue® Plus2 pre-stained molecular mass standard. The immunoblots were probed with evoglow antibodies Bs (1/5000 dilution) followed by IgG HRP-conjugated anti-rabbit antibody (1/3000 dilution). The immunoblots were developed with Immobilon® Western HRP Substrate and exposed for 60 min. Arrows indicate bands of interest.

4.2.1.5 Construction of the pHN-TdclosEvo suicide plasmid As the genome of T. denticola has a low GC content of 37.9%, the construction of a T. denticola evoglow expression cassette was repeated using an alternative evoglow gene that had been codon optimized for use in low GC Clostridia. Plasmid pHN-TdclosEvo was generated in the same manner as pHN-TdEvo except that the evoglow-C-Bs2-stop gene was used instead of evoglow-Bs2-stop and it was amplified from the plasmid pGlow- CKXN-Bs2. The desired size of evoglow-C-Bs2-stop is 460 bp (Figure 4.7A). 5’ TDE0911

129

was fused with the msp promoter to yield a 984 bp PCR amplicon, whilst 3’ TDE0911 was fused with aphA2 to yield a 1645 bp PCR amplicon (Figure 4.7B). The 5’ TDE0911- msp promoter was purified and fused with evoglow-C-Bs2-stop for a 1404 bp product which was then fused with 3’ TDE0911-aphA2 to yield a final product (TDE0911::clos- evoglow::aphA2) of 3009 bp (Figure 4.7D). The purified TDE0911::clos- evoglow::aphA2 was then ligated to pGEM®-T Easy and used to transform E. coli α- Select cells. Six transformed colonies selected on LB Amp plates were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. All transformants examined possessed the desired 3.3 kbp amplicon and clone one was selected and designated ECR764 (Figure 4.8). Plasmid from ECR764 was subjected to nucleotide sequencing to confirm the fidelity of the recombination cassette and was designated pHN- TdclosEvo. When illuminated with UV light and viewed using confocal microscope, the colonies and cells of ECR764 fluoresced brightly, similar to that of ECR763 (Figure 4.3; results not shown).

Figure 4.7 Generation of plasmid pHN-TdclosEvo. A. PCR amplicons of evoglow-C- Bs2-stop (Lane 1; 460 bp), msp promoter (Lane 2; 261 bp) and aphA2 (Lane 3; 979 bp). B. PCR amplicons of 5’ TDE0911-msp promoter (Lane 1; 984 bp) and aphA2-3’ TDE0911 (1645 bp). C. PCR amplicon of 5’ TDE0911-msp promoter-evoglow-C-Bs2- stop (1404 bp). D. PCR amplicon of the final SOE product, TDE0911::clos- evoglow::aphA2 (3009 bp). E. NotI-digestion of pHN-TdclosEvo. Expected fragment size of 3 kbp for both insert and vector. The PCR amplicons and NotI-digestion of pHN- TdclosEvo were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

130

Figure 4.8 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-TdclosEvo. Lane 1-6, PCR amplicons from E. coli colonies 1-6 using primers M13 Forward and M13 Reverse. Expected amplicon size was 3.3 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

4.2.1.6 Generation of a T. denticola ATCC 35405 mutant expressing FbFP from

evoglow-C-Bs2-stop (ATCC 35405::FbFPevoglow-C-Bs2-stop) Plasmid pHN-TdclosEvo was linearized by digestion with NotI to release the recombination cassette (3 kbp) from the background vector (3 kbp; Figure 4.7E). The purified digested products were used to transform T. denticola ATCC 35405 by electroporation. Twelve transformants selected on OBGM agar plates containing 75 µg/mL kanamycin were grown in OBGM containing 75 µg/mL kanamycin and a small volume of each culture was used for PCR analysis to confirm DNA integration at the correct locus. All transformants examined possessed the desired amplicons of size 2.5 kbp and 1.8 kbp for the primer pairs tde910-F/Kan-R and Kan-closEvo-F/tde913-R respectively (Table 2.4; Figure 4.9). Transformant ten, T. denticola ATCC 35405 ATCC

35405::FbFPevoglow-C-Bs2-stop 10, was designated ECR775.

131

Figure 4.9 PCR to confirm homologous recombination of TDE0911::clos- evoglow::aphA2 with T. denticola ATCC 35405 genome. A. PCR amplicons from ATCC 35405::FbFPevoglow-C-Bs2-stop transformants 1-12 (lanes 1-12) and ATCC 35405 WT control (lane 13) using primer pair tde910-F/Kan-R. No product was expected from WT and 2.5 kbp amplicons were expected from ATCC 35405::FbFPevoglow-C-Bs2-stop transformants. B. PCR amplicons from ATCC 35405::FbFPevoglow-C-Bs2-stop transformants 1-12 (lanes 1-12) and ATCC 35405 WT control (lane 13) using primer pair Kan-closEvo- F/tde913-R. No product was expected from WT and 1.8 kbp amplicons were expected from ATCC 35405::FbFPevoglow-C-Bs2-stop transformants. The PCR amplicons were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

4.2.1.7 Visualization of T. denticola ATCC 35405 and ATCC 35405::FbFPevoglow-C-

Bs2-stop transformant 10 fluorescence using confocal microscopy

T. denticola ATCC 35405 and ATCC 35405::FbFPevoglow-C-Bs2-stop 10 (ECR775) cells were grown to exponential phase, washed twice in 0.85% (w/v) NaCl and resuspended in 0.85% (w/v) NaCl. The cells were viewed using a confocal microscope. Although fluorescence was observed from ATCC 35405::FbFPevoglow-C-Bs2-stop 10, its intensity was similar to that of the WT. Furthermore, the fluorescence observed from T. denticola

ATCC 35405 and ATCC 35405::FbFPevoglow-C-Bs2-stop 10 was blurry with no distinct cell shape, unlike the fluorescence observed from T. denticola ATCC 35405 stained with the LIVE/DEAD™ BacLight™ Bacterial viability kit where distinct cells were observed. Similar results were observed from cells fixed using 4% (v/v) PFA in PBS (data not shown).

132

Figure 4.10 Visualization of T. denticola ATCC 35405 with and without BacLight™ stain and ATCC 35405::FbFPevoglow-C-Bs2-stop 10. T. denticola cells washed twice in 0.85% (w/v) NaCl were visualized using confocal microscopy. For T. denticola WT stained with BacLight™, the cells were washed twice in 0.85% (w/v) NaCl, stained with BacLight™ and visualized using confocal microscopy. Images shown were obtained using a 63x oil immersion lens under laser excitation (488 nm wavelength) to visualize FbFP (ATCC 35405::FbFPevoglow-C-Bs2-stop 10) and SYTO™ 9 Green Fluorescent Nucleic Acid stain (WT stained with BacLight™).

133

4.2.1.8 Immunoblot analyses of FbFP expression from T. denticola ATCC 35405

and ATCC 35405::FbFPevoglow-C-Bs2-stop transformants 1-11 Immunoblots were carried out to check for the expression of FbFP from ATCC

35405::FbFPevoglow-C-Bs2-stop clones 1-11. E. coli harboring the plasmid pHN-TdclosEvo (ECR764) was used as a positive control. The expected size of FbFP was 16 kDa. A band between 14 and 19 kDa was observed in the lanes loaded with the whole cell lysates from

ECR764 and ATCC 35405::FbFPevoglow-C-Bs2-stop 1-11, but not in the WT ATCC 35405 lane (Figure 4.11B). This indicated that FbFP was expressed from ATCC

35405::FbFPevoglow-C-Bs2-stop 1-11 but its expression was very low. Direct comparison between clones containing either the codon optimized or non-optimized evoglow gene showed that the band between 14 and 19 kDa was only observed in the cell lysates of

ATCC 35405::FbFPevoglow-C-Bs2-stop 7 and 10, the two clones with the seemingly highest cell expression of FbFP, but not ATCC 35405::FbFPevoglow-Bs2-stop 1 and 2 (Figure 4.12).

This further confirmed that FbFP was expressed in ATCC 35405::FbFPevoglow-C-Bs2-stop which contained the GC codon optimized evoglow gene but not ATCC

35405::FbFPevoglow-Bs2-stop which contained the non-optimized evoglow gene.

134

Figure 4.11 Analyses of FbFP expression in T. denticola ATCC 35405 and ATCC 35405::FbFPevoglow-C-Bs2-stop 1-11 using immunoblots. Whole cell lysates of E. coli harboring plasmid pHN-TdclosEvo (ECR764) and plasmid pHN-911 (ECR761), T. denticola ATCC 35405 and transformants ATCC 35405::FbFPevoglow-C-Bs2-stop 1-11 were analyzed via A. SDS-PAGE using a NuPAGE® 12% Bis-Tris gel and B. immunoblotting on a nitrocellulose membrane. Each T. denticola sample (1.25 x 108 cells) was pelleted and resuspended in NuPAGE® LDS Sample Buffer before being loaded on the gel. M indicates the SeeBlue® Plus2 pre-stained molecular mass standard. Immunoblots were probed with evoglow antibodies Bs (1/5000 dilution) followed by IgG HRP-conjugated anti-rabbit antibody (1/3000 dilution). Immunoblots were developed with Immobilon® Western HRP Substrate with an exposure time of 5 min (ECR 764) and 75 min (T. denticola WT and transformants). Arrows indicate bands of interest.

135

Figure 4.12 Analyses of FbFP expression in T. denticola ATCC 35405::FbFPevoglow- Bs2-stop and ATCC 35405::FbFPevoglow-C-Bs2-stop using immunoblots. Whole cell lysates of T. denticola ATCC 35405::FbFPevoglow-Bs2-stop transformants 1 and 2 and ATCC 35405::FbFPevoglow-C-Bs2-stop transformants 7 and 10 were analyzed via A. SDS-PAGE using a NuPAGE® 10% Bis-Tris gel and B. immunoblotting on a nitrocellulose membrane. Each T. denticola sample (1.25 x 108 cells) was pelleted and resuspended in NuPAGE® LDS Sample Buffer before being loaded on the gel. M indicates the SeeBlue® Plus2 pre-stained molecular mass standard. Immunoblots were probed with evoglow antibodies Bs (1/10000 dilution) followed by IgG HRP-conjugated anti-rabbit antibody (1/3000 dilution). Immunoblots were developed with Immobilon® Western HRP Substrate with an exposure time of 25 min. Arrows indicate bands of interest.

136

4.2.2 Generation of a P. gingivalis W50 strain expressing FbFP (W50::FbFP) 4.2.2.1 Construction of the pHN-PgEvo suicide plasmid Plasmid pHN-PgEvo was generated as a P. gingivalis suicide plasmid where pGEM®-T Easy contained a recombination cassette consisting of kgp promoter, evoglow-Bs2-stop and ermFAM flanked by DNA corresponding to the genome upstream and downstream of the mfaI gene in P. gingivalis W50. The 5’ and 3’ flanking regions of mfaI (5’ mfaI and 3’ mfaI), as well as the kgp promoter, were amplified from P. gingivalis W50 WT gDNA, whilst ermFAM was amplified from the shuttle vector pHS17. The evoglow-Bs2- stop gene was amplified from the plasmid pGLOW-Bs2-stop (evocatal GmbH, Germany). The PCR amplicons with the expected size of 680 bp, 743 bp, 290 bp, 454 bp and 2183 bp for 5’ mfaI, 3’ mfaI, kgp promoter, evoglow-Bs2-stop and ermFAM respectively were purified (Figure 4.13A-B). A PCR amplicon with the desired size of 930 bp was obtained from the fusion of 5’ mfaI with the kgp promoter by SOE PCR (Figure 4.13C). The purified 5’ mfaI-kgp promoter was then fused with evoglow-Bs2-stop to yield a 1344 bp PCR amplicon, whilst ermFAM was fused with 3’ mfaI to yield a 2886 bp PCR amplicon (Figure 4.13D). The purified 5’ mfaI-kgp promoter-evoglow-Bs2-stop was fused by SOE PCR with ermFAM-3’ mfaI for a final product (mfaI::evoglow::ermFAM) of 4190 bp (Figure 4.13E). The purified mfaI::evoglow::ermFAM was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Of the transformant colonies selected on LB Amp plates, four were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. All transformants examined possessed the desired 4.5 kbp amplicon (Figure 4.14). Plasmid from clone two was sequenced to confirm the fidelity of the recombination cassette in pHN-PgEvo. E. coli α-Select harboring pHN-PgEvo was designated ECR762.

137

Figure 4.13 Generation of plasmid pHN-PgEvo. A. PCR amplicons of 5’ mfaI (Lane 1; 680 bp), 3’ mfaI (Lane 2; 743 bp), kgp promoter (Lane 3; 290 bp) and evoglow-Bs2- stop (Lane 4; 454 bp). B. PCR amplicon of ermFAM (2183 bp). C. PCR amplicon of 5’ mfaI-kgp promoter (930 bp). D. PCR amplicon of 5’ mfaI-kgp promoter-evoglow-Bs2- stop (Lane 1; 1344 bp) and ermFAM-3’ mfaI (Lane 2; 2886 bp). E. PCR amplicon of the final SOE product, mfaI::evoglow::ermFAM (4190 bp). F. NotI-digestion of pHN-PgEvo showing the expected products of 4.2 kbp for insert and 3 kbp for vector. Lane 1, undigested pHN-PgEvo; Lane 2, NotI-digested pHN-PgEvo. The PCR amplicons and NotI-digestion of pHN-PgEvo were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Figure 4.14 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-PgEvo. Lanes 1-4, PCR amplicons of the expected size of 4.5 kbp generated from E. coli colonies 1-4 using primers M13 Forward and M13 Reverse. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

4.2.2.2 Generation of W50::FbFP Plasmid pHN-PgEvo was linearized by digestion with NotI to release the recombination cassette (4.2 kbp) from the background vector (3 kbp) (Figure 4.13F). The digested products were purified and used to transform P. gingivalis W50 by electroporation. Genomic DNA was isolated from four transformants selected on HBA plates containing erythromycin. Confirmation of DNA integration at the correct locus was performed by PCR with Pg0177-F and Pg0179-R primers which anneal 5’ and 3’ respectively to the desired homologous recombination site of the recombination cassette. All four transformants possessed the appropriate 4.5 kbp amplicon (Figure 4.15). Transformant one was designated P. gingivalis W50 W50::FbFP or ECR771.

138

Figure 4.15 PCR to confirm homologous recombination of mfaI::evoglow::ermFAM with P. gingivalis W50 genome. Lane 1, PCR amplicon from P. gingivalis wild-type gDNA control; Lane 2-5, PCR amplicons from W50::FbFP transformants 1-4. All amplicons are of the expected sizes of 1.6 kbp for WT and 4.5 kbp for W50::FbFP. The PCR amplicons were examined by agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

4.2.2.3 Visualization of P. gingivalis W50 and W50::FbFP transformants 1-4 fluorescence using microscopy To assay fluorescence of P. gingivalis W50 and W50::FbFP 1-4, cells were grown to exponential phase, washed twice in PBS and resuspended in PBS. The cells were viewed under a combined fluorescence and phase contrast microscope which has the ability to view cells with both fluorescence and phase contrast modes (Figure 4.16). The phase contrast mode was used to confirm the bacterial cells in the samples. Fluorescence was observed from wild-type P. gingivalis, suggesting that P. gingivalis is auto-fluorescent. The fluorescence observed from W50::FbFP clones 1 – 4 was similar to that of the wild- type, suggesting that FbFP was either not being expressed or not expressed at a detectable level above WT background level.

139

Figure 4.16 Visualization of P. gingivalis W50 and W50::FbFP 1-4 fluorescence using microscopy. Anaerobically grown, exponential-phase cells were washed twice in PBS, resuspended in PBS, aliquoted into an ibidi 1 μ-Slide and visualized under a combined fluorescence and phase contrast microscope. Images shown in the left column were obtained during fluorescence mode while the images shown on the right were obtained during phase contrast mode (Magnification of 63x).

140

4.2.2.4 Fluorescence intensities of P. gingivalis W50 and W50::FbFP transformants 1-4 as determined by fluorescence spectroscopy The fluorescence intensities of P. gingivalis W50 and W50::FbFP 1-4 were measured using a spectrofluorometer. The fluorescence intensities of the cells were measured across a wavelength of 450 nm to 600 nm with the excitation wavelength fixed at 450 nm (the excitation wavelength of FbFP). Exponential-phase P. gingivalis cells were collected as unwashed cells. The remaining cells were washed twice in 0.85% (w/v) NaCl and resuspended in 0.85% (w/v) NaCl as washed cells. 0.85% (w/v) NaCl was used instead of PBS due to the concern that PBS may contribute to background fluorescence. The unwashed WT cells exhibited a low level of auto-fluorescence, further confirming the results obtained from the fluorescence microscopy. The fluorescence emission spectra of W50::FbFP 1-4 are similar to that of WT, washed or unwashed. This suggests that W50::FbFP 1-4 do not fluoresce differently to WT, thus any fluorescence from these clones is due to auto-fluorescence and not the FbFP protein.

141

Figure 4.17 Fluorescence emission spectra of P. gingivalis W50 and W50::FbFP 1- 4 A. from unwashed cells in BHI. B. from washed cells in 0.85% (w/v) NaCl. Anaerobically grown, exponential-phase cells were collected as the unwashed cells. Cells washed twice in 0.85% (w/v) NaCl and resuspended in 0.85% (w/v) NaCl were collected as washed cells. An excitation wavelength of 450 nm was used and fluorescence intensity was measured across a range of wavelengths from 450 nm to 600 nm.

4.2.2.5 Immunoblot analyses of FbFP expression from P. gingivalis W50 and W50::FbFP transformants 1-4 An immunoblot was carried out to check for the expression of FbFP in W50::FbFP transformants 1-4. E. coli strains ECR763 and ECR761 were used as a positive control and a negative control respectively. The expected size of FbFP is 16 kDa. A band between 14 and 19 kDa was observed in the lane loaded with ECR763 but not in the lane loaded with ECR761, confirming the expression of FbFP in ECR763 but not ECR761. The FbFP band was not observed in any P. gingivalis cells including the transformants W50::FbFP

142

1-4, suggesting that FbFP was not expressed at a detectable level in the transformants during either exponential or stationary phase growth (Figure 4.18). In order to determine whether FbFP was secreted from the cells, cell-free supernatants from P. gingivalis WT and transformants W50::FbFP 1-4 were TCA-precipitated and analyzed by immunoblot. No band with a size between 14 and 19 kDa was observed in any P. gingivalis supernatant samples (Figure 4.19), indicating that FbFP was not secreted from the cells. The antibody had poor specificity as it cross-reacted with mid to high molecular weight proteins from wild-type P. gingivalis W50 and E. coli.

Figure 4.18 Analysis of FbFP expression in P. gingivalis W50 and W50::FbFP 1-4 using immunoblot. Whole cell lysates of E. coli harboring plasmid pHN-TdEvo (ECR763) and plasmid pHN-911 (ECR761), P. gingivalis WT and transformants W50::FbFP 1-4 were analyzed via A. SDS-PAGE using NuPAGE® 12% Bis-Tris gel and B. immunoblotting on a nitrocellulose membrane. P. gingivalis cells were resuspended to a final concentration of 2.5 x 107 cells/µL, lysed by sonication and an equal volume of each lysed sample was loaded on the gel. M indicates the SeeBlue® Plus2 pre-stained molecular mass standard. The immunoblot was probed with evoglow antibodies Bs (1/5000 dilution; Jena Bioscience, Jena, Germany) followed by IgG HRP-conjugated anti-rabbit antibodies (1/3000 dilution). The immunoblot was developed with Immobilon® Western HRP Substrate and exposed for 5 min. Arrows indicate bands of interest.

143

Figure 4.19 Analyses of FbFP secretion from P. gingivalis W50 and W50::FbFP 1- 4 via immunoblot. A whole cell lysate of E. coli harboring plasmid pHN-TdEvo (ECR763) and supernatants from P. gingivalis WT and transformants W50::FbFP 1-4 were analyzed via A. SDS-PAGE using a NuPAGE® 12% Bis-Tris gel and B. immunoblotting on a nitrocellulose membrane. The supernatant (1 mL) of different P. gingivalis samples was TCA-precipitated, resuspended in NuPAGE® LDS Sample Buffer and loaded on the gel. M indicates the SeeBlue® Plus2 pre-stained molecular mass standard. The immunoblot was probed with evoglow antibodies Bs (1/5000 dilution) followed by IgG HRP-conjugated anti-rabbit antibodies (1/3000 dilution). The immunoblot was developed with Immobilon® Western HRP Substrate and exposed for 7.5 min. Arrows indicate bands of interest.

144

4.3 DISCUSSION T. denticola and P. gingivalis have been shown to form polymicrobial biofilms synergistically in static and flow cell assays (Yamada et al. 2005, Zhu et al. 2013). Their biofilms are commonly visualized and quantified by fluorescent in situ hybridization (FISH) using species-specific probes and confocal scanning laser microscopy imaging (Zainal-Abidin et al. 2012, Zhu et al. 2013). Although this is a powerful technique to visualize and quantify biofilms, it requires the biofilms to be fixed before staining, thus rendering a real-time examination of the biofilm development impossible. In this study, an attempt to construct fluorescently labelled T. denticola ATCC 35405 and P. gingivalis W50 were undertaken in an effort to develop strategies to examine dual-species biofilm development in real time. The evoglow gene which encodes a fluorescent protein reported to fluoresce under anaerobic conditions (FbFP) (Drepper et al. 2007) was used as T. denticola and P. gingivalis were both anaerobic bacteria and their biofilm development will need to be carried out under anaerobic conditions.

Using the optimized protocol developed in the previous chapter, T. denticola ATCC 35405 was successfully transformed with the recombination cassette carrying the evoglow-Bs2-stop gene from the plasmid pGLOW-Bs2-stop. The evoglow-Bs2-stop gene expression was under the influence of T. denticola major outer sheath protein promoter region (msp promoter) as Msp is a constitutively expressed, highly abundant protein in T. denticola and msp transcript was found at a high level in T. denticola (Fenno et al. 1996), indicating that the msp promoter is a strong promoter. A strong promoter is necessary to ensure the optimal expression of FbFP for visualization of the cells. Moreover, the sequence of msp promoter had been previously characterized (Fenno et al. 1996). TDE0911, which encodes a type II restriction enzyme, was chosen as the locus of insertion of the evoglow gene as previous experiments (Chapter 3) had shown that this locus was amenable to inactivation by allelic exchange.

Despite successful incorporation of the evoglow-Bs2-stop gene into the genome of ATCC

35405::FbFPevoglow-Bs2-stop, no fluorescence was observed in the transformants under confocal microscopy. Immunoblot analyses showed that FbFP was not expressed in the transformants under the conditions examined in this study. The cross-reactivity of evoglow antibodies Bs with mid to high molecular weight proteins from T. denticola whole cell lysates suggested its poor specificity. FbFP contains a LOV domain which is a subset of the PAS superfamily (Buckley et al. 2015). PAS domain is widespread in

145

proteins from all kingdoms of life and is highly conserved in three-dimensional structure (Henry and Crosson 2011). It is possible that the evoglow antibodies Bs raised against the FbFP protein recognized its PAS domain and cross-reacted with other PAS-containing proteins in T. denticola. This is supported by the finding that a PAS-containing protein in T. denticola, TDE1970 (Sarkar et al. 2010), has a molecular weight of 46 kDa which correlates well with the size of the high molecular weight band observed in the immunoblot of T. denticola whole cell lysates (Figure 4.6).

Since T. denticola ATCC 35405 expressing FbFP was successfully constructed in a study by using an alternative evoglow-C-Bs2-stop gene from the plasmid pGlow-CKXN-Bs2 (Godovikova et al. 2015), the construction of the expression cassette was repeated using this alternative gene. FbFP was expressed in ATCC 35405::FbFPevoglow-C-Bs2-stop transformants, proving that this evoglow gene that had been codon optimized for use in low GC Clostridia is more useful in T. denticola. However, FbFP was expressed at a very low level and it was not secreted from the cells. The amount of FbFP expressed in the cells was not enough for its fluorescence to be detected by confocal microscopy. In the previous study which successfully produced a fluorescently tagged T. denticola ATCC 35405, the evoglow-C-Bs2-stop gene was fused with the fla promoter in the shuttle vector pCF728 (Godovikova et al. 2015). In this study, FbFP was expressed from the genome instead of a shuttle vector due to the concern that the shuttle vector may not be maintained in the absence of selective antibiotic pressure during intended biofilm development studies. The inclusion of an antibiotic is not desirable as it may affect the survival of P. gingivalis and it introduces an extra variable that might affect the biofilm formation. Although the copy number of pCF728 in T. denticola ATCC 35405 is unclear, it is likely that there will be more copies of evoglow-C-Bs2-stop transcript expressed in the reported strain with the shuttle vector than in the strain constructed in this study that only contains one copy of the evoglow-C-Bs2-stop gene in its genome. It was observed in T. denticola ATCC 33520 that Msp is 100 times more abundant than FlgE (the most abundant protein expressed from the fla operon), suggesting that the msp promoter is a stronger promoter than the fla promoter. Therefore, the difference in promoter used for the expression of FbFP in T. denticola is unlikely to be the reason for a low expression of FbFP in this study. However, it should be noted that the expression of the proteins might be different in T. denticola ATCC 35405 than ATCC 33520. Nevertheless, the authors who constructed the fluorescently labelled T. denticola (Godovikova et al. 2015) have

146

communicated that they are still improving their evoglow cassette to increase the expression of FbFP (C. Fenno, personal communication) and the fact that FbFP expression was only detectable in T. denticola cells after fixation and barely detectable in live cultures (C. Fenno, personal communication) makes it unsuitable for real-time studies.

Upon transformation with pHN-TdEvo and pHN-TdclosEvo, a high level of FbFP expression was obtained in the E. coli α-Select cells harboring these plasmids (ECR763 and ECR764 respectively). This observation supports a previous study which suggested that E. coli is able to use the msp promoter which has the composition and spacing of the proposed -35 and -10 regions in almost perfect identity with the E. coli promoter consensus sequence (Fenno et al. 1996). The high degree of identity to the E. coli consensus promoter sequence allows recognition by E. coli RNA polymerase and thus leads to the expression of FbFP. The expression of FbFP in the E. coli cell pellets was high enough to be detected by the naked eye and the fluorescence of the colonies was readily visualized under UV light (Figure 4.3), possibly because of the high copy number of pHN-TdEvo and pHN-TdclosEvo in the cells since the background pGEM®-T Easy vector is a high-copy-number vector. Besides showing a desirable codon efficiency in the evoglow genes for successful expression of FbFP in E. coli cells, the high expression of FbFP in the ECR763 and ECR764 also shows the stability and non-toxicity of the protein in the cells and it has further confirmed that there was no problem in the construction of the recombination cassettes as functional protein product was produced.

In addition to T. denticola ATCC 35405, the construction of a fluorescently labelled P. gingivalis W50 was attempted. The evoglow-Bs2-stop gene was fused with the kgp promoter region as Kgp is one of the most abundant proteins expressed in P. gingivalis and it was shown to have stable expression under different conditions (N. Slakeski, personal communication). Furthermore, the kgp promoter region has been identified previously (Jackson et al. 2000). mfaI which encodes the minor fimbriae fimbrillin was chosen as the locus of insertion of the evoglow gene, as the mfaI gene of W50 is disrupted by an insertion element (Dashper et al. 2017) and thus is non-functional. This locus has also been used successfully as a point of insertion in other studies in our laboratory without any observable changes to WT phenotype. The evoglow-Bs2-stop gene was successfully incorporated into the genome of W50::FbFP transformants, but the fluorescence of the transformants was similar to that of the WT cells (Figure 4.16).

147

Fluorescence was observed under fluorescence microscopy from wild-type P. gingivalis W50 cells washed with PBS but no fluorescence was detected by fluorescence spectroscopy from wild-type P. gingivalis W50 cells washed with 0.85% (w/v) NaCl. This suggests that PBS might contribute to the fluorescence of P. gingivalis W50 cells. P. gingivalis WT and W50::FbFP transformants emitted fluorescence in BHI but not in 0.85% (w/v) NaCl (Figure 4.17), suggesting that BHI might contain components that emit fluorescence when excited at 450 nm, such as lipofuscin, a high-molecular weight ﬂuorescent material that has been shown to be enriched in neurons and cardiac muscles (Schonenbrucher et al. 2008). Immunoblot analyses confirmed that FbFP was neither expressed in nor secreted from the P. gingivalis cells. A previous study which had successfully produced fluorescently tagged P. gingivalis used a shuttle vector in strain ATCC 33277 (Choi et al. 2011). The evoglow-Bs2-stop gene was fused with the fimA promoter region and cloned into the P. gingivalis-E. coli shuttle vector pT-COW, which has a copy number of 5-10 in Bacteriodetes (Cho et al. 2001, Choi et al. 2011). The use of a shuttle vector may be more advantageous as it gave a higher copy number of evoglow- Bs2-stop gene. However, the use of a shuttle vector is not desirable in this study due to the reasons outlined above for T. denticola. The difference in P. gingivalis strains used in this study (W50) and the previous study (ATCC 33277) may also be a reason for the difference in FbFP expression. Nevertheless, we have received communication that the authors have lost the fluorescently labelled P. gingivalis clone due to a mishap related to their freezer system and they are now experiencing difficulty in remaking the clone (O. Yilmaz, personal communication), suggesting the lack of reproducibility in the construction of fluorescently labelled P. gingivalis.

148

4.4 CONCLUSION In conclusion, the attempts to generate fluorescently labelled P. gingivalis and T. denticola in this study by expressing FbFP were unsuccessful. The evoglow-C-Bs2-stop gene from the plasmid pGlow-CKXN-Bs2 proved to be more useful than the evoglow-Bs2- stop gene from the plasmid pGLOW-Bs2-stop, as T. denticola transformants carrying evoglow-C-Bs2-stop but not evoglow-Bs2-stop were able to express FbFP, albeit at a very low level. Although the use of shuttle vectors may be helpful to increase the copy number and thus the expression of FbFP, the possibility of losing the shuttle vectors in the absence of selective antibiotic pressure over the long time of biofilm development remained a concern. It is possible to introduce shuttle vectors that confer the same antibiotic resistance to both P. gingivalis and T. denticola, but the addition of antibiotic would add an extra variable into the biofilm study. To date, there is only one study each that has reported on the use of FbFP in P. gingivalis and T. denticola and subsequent correspondence has indicated problems with both these strains. A lack of publications using such a valuable tool in the widely researched P. gingivalis and T. denticola (and in anaerobic bacteria in general) is indicative that the use of FbFP for the construction of fluorescently labelled anaerobic bacteria is not a readily useable strategy and a lot of work is required to optimize the usage of FbFP.

149

CHAPTER 5 DEFINING THE ROLE OF T. DENTICOLA MOTILITY IN SYNERGISTIC BIOFILM FORMATION WITH P. GINGIVALIS

150

5.1 INTRODUCTION T. denticola and P. gingivalis are pathobionts frequently associated with severe chronic periodontitis and the study of their interactions in polymicrobial biofilms is important to understand disease progression and initiation. T. denticola is one of only a few motile bacteria in the oral cavity while P. gingivalis is a sessile bacterium. T. denticola motility, although not considered as a classical virulence factor, is important for its pathogenesis. The role of motility in biofilm formation is unclear but it is believed that it contributes to pore formation in the biofilm matrix which then results in an increase of nutrient flow into the biofilm (Houry et al. 2012). Unlike extracellularly flagellated bacteria, T. denticola remains motile in a highly viscous environment, likely due to the protection of the flagella in the periplasmic space, away from direct contact with the external environment (Klitorinos et al. 1993). A biofilm is a dynamic structure made up of microbial cell aggregates enclosed within a matrix of extracellular polymeric substances. Channels within the biofilm contain a liquid phase which forms the viscous part of the biofilm (Lembre et al. 2012). The ability of T. denticola to move in highly viscous environments may be beneficial for its movement through the biofilms. Together with the presence of a chemotaxis system that allows it to move in response to environmental stimuli, it is speculated that T. denticola movement in the biofilm remodels the biofilm, creating pores which allows for better nutrient penetration and waste removal, thereby contributing to a larger biofilm biomass.

A spirochete periplasmic flagellum is similar to a typical bacterial flagellum in that it can be divided into three parts: a basal body, hook and filament (Figure 5.1). The basal body is composed of a rod and a motor-switch complex embedded in the inner membrane. The motor is made up of two parts: the rotor and the stator. The rotor is made up of FliF, FliG, FliM and FliY. FliG, FliM and FliY also act as a molecular switch involved in the switching of the direction of flagellar rotation (Limberger 2004, Morimoto and Minamino 2014). Each stator is composed of two integral membrane proteins, MotA and MotB, and a peptidoglycan binding motif within the C-terminal periplasmic domain of MotB is postulated to anchor the stator to the peptidoglycan layer. The stator is a proton channel complex coupling transmembrane ion movement with flagellar rotation (Kojima and Blair 2001, Chevance and Hughes 2008, Kojima et al. 2009, Morimoto and Minamino 2014). The flagellar hook structure consists of one major polypeptide, FlgE, and the disruption of flgE gene in T. denticola has been shown to produce a non-motile mutant

151

that is deficient in both flagellar hook and filament structures (Li et al. 1996, Limberger 2004). The flagellar hook serves as a universal joint that transmits torque produced by the motor in the basal body to the flagellar filament, whose rotation results in specific movements of the cell body that enables the cell to move (Berg 1976, Limberger 2004). The flagellar filament of T. denticola is made up of three filament core proteins (FlaB1- 3) and three filament outer layer proteins (FlaA1-3). The flagellar filaments originate at opposite ends of the same cell and overlap at mid-cell (Chan et al. 1993). Each T. denticola cell possesses four periplasmic flagella, two from each end of the cell (Izard et al. 2008). T. denticola also possesses a complete set of chemotaxis proteins required for signal perception, transduction, and adaptation. The methyl-accepting chemotaxis proteins (MCPs) which traverse the inner membrane are responsible for detecting environmental stimuli and the signals are transmitted via the cytosolic chemotaxis proteins CheA, CheW and CheY to the flagellar motor (Figure 5.1)(Dashper et al. 2011). CheY is the chemotaxis response regulator that is phosphorylated by CheA, the central kinase in the chemotaxis pathway and in E. coli, the direction of flagellar rotation changes when phosphorylated CheY binds to the motor switch (Hess et al. 1988).

152

Figure 5.1 Diagrammatic representation of the T. denticola chemotaxis signaling pathway and its periplasmic flagellum. The methyl-accepting chemotaxis protein (MCP) detects environmental stimuli and the signals are transmitted via CheA, CheW and CheY to the flagellar motor switch (FliG, FliM, FliY). The periplasmic flagellum is made up of three parts: a basal body, hook and filament. The flagellar hook (FlgE) connects and transmits torque produced by the motor (rotor and stator) in the basal body to the flagellar filament which then rotates. Diagram was adapted and modified from Limberger (2004) and Sim et al. (2005).

Previous studies have shown that T. denticola and P. gingivalis form biofilms synergistically (Yamada et al. 2005, Zhu et al. 2013). The dual-species biofilms of T. denticola and P. gingivalis were larger than the sum of their mono-species biofilms. It was also observed that most of the T. denticola cells in mono-species biofilms lost their characteristic spiral morphology but retained it when grown with P. gingivalis (Zhu et al. 2013). The morphology of T. denticola is closely related to its periplasmic flagella (Ruby

153

et al. 1997) and periplasmic flagella are required for the motility of T. denticola (Li et al. 1996). Therefore, it is suggested that T. denticola motility may play a role in dual-species biofilm formation and development with P. gingivalis and the aim of this study was to investigate the role of T. denticola motility in its synergistic biofilm formation with P. gingivalis. Using the optimized protocol developed in Chapter 3, a series of T. denticola mutants lacking flgE, motA, motB and cheY were created. These genes were targeted, as their inactivation in T. denticola or other bacteria had resulted in mutants with impaired motility (Li et al. 1996, Lane et al. 2005, Houry et al. 2010, Sultan et al. 2015). The mutants generated were then characterized for their morphology, motility, growth, binding with P. gingivalis and protein expression profiles. Finally, the ability of the mutants to form dual-species biofilms with P. gingivalis were studied using both static and flow cell biofilm assays.

154

5.2 RESULTS In order to study the role of T. denticola motility in synergistic biofilm formation with P. gingivalis, four T. denticola mutants lacking flgE, motA, motB and cheY were generated. FlgE is the flagellar hook protein, MotA and MotB are the flagellar stator proteins and CheY is a chemotaxis protein involved in changing the direction of flagellar rotation (Figure 5.1). Inactivation of the flgE gene is expected to produce a non-motile mutant due to a lack of flagellar hook and filament, whereas the MotA, MotB and CheY mutants are expected to be impaired in motility but have intact flagella. By comparing the ability of these mutants to form biofilms with P. gingivalis, the role of T. denticola motility or periplasmic flagella in synergistic biofilm formation with P. gingivalis can be examined. The mutants were produced by homologous recombination with suicide plasmids pHN- flgE, pHN-motA, pHN-motB and pHN-cheY constructed as described below.

5.2.1 Construction of the pHN-motA, pHN-motB, pHN-flgE and pHN-cheY suicide plasmids Plasmids pHN-motA, pHN-motB and pHN-flgE were generated as T. denticola suicide plasmids where pGEM®-T Easy contained recombination cassettes consisting of ermAM and the fla promoter flanked by DNA corresponding to the genome upstream and downstream of the target genes in T. denticola ATCC 35405. The motA, motB and flgE genes are all in the same operon expressed from the fla promoter. The introduction of the fla promoter in the recombination cassette is to ensure transcription of all operon genes downstream of the deleted target gene. Plasmid pHN-cheY was generated in the same manner as the others, but without the introduction of a promoter as it is the last gene in the operon. The plasmids were generated by SOE PCR using the corresponding primers as listed in Table 2.5.

The 5’ and 3’ flanking regions of motA (5’ motA and 3’ motA), as well as the fla promoter, were amplified from T. denticola ATCC 35405 WT gDNA whilst ermAM was amplified from the shuttle vector pHS17. The PCR amplicons with the expected sizes of 829 bp, 875 bp, 370 bp and 1226 bp for 5’ motA, 3’ motA, fla promoter and ermAM respectively, were purified (Figure 5.2A-B). The purified 5’ motA and 3’ motA were fused with ermAM and fla promoter respectively, by SOE PCR. Non-specific amplifications were observed with the fusion of 5’ motA with ermAM (1205 bp) and the fusion of fla promoter with 3’ motA (2015 bp) (Figure 5.2C-D). The PCR amplicons of the desired sizes (as indicated by the red boxes in Figure 5.2C-D) were purified from the agarose gel and fused for a

155

final product (motA::ermAM) of 3180 bp (Figure 5.2E). The purified motA::ermAM was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Of the transformed colonies selected on LB Amp plates, twelve were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. Transformants 2-4, 9, 10 and 12 possessed the desired 3.3 kbp amplicon (Figure 5.3). Plasmid from clone 9 was subjected to nucleotide sequencing to confirm the fidelity of the recombination cassette in pHN- motA. E. coli α-Select harboring pHN-motA was designated ECR765.

Figure 5.2 Generation of plasmid pHN-motA. A. PCR amplicons of 5’ motA, 3’ motA and ermAM. Lane 1, 5’ motA (829 bp); Lane 2, 3’ motA (875 bp); Lane 3, ermAM (1226 bp). B. PCR amplicon of fla promoter (370 bp). C. PCR amplicon of 5’ motA-ermAM (1205 bp). D. PCR amplicon of fla promoter-3’ motA (2015 bp). E. PCR amplicon motA::ermAM (3180 bp). All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The red boxes indicate DNA purified from the agarose gel. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Figure 5.3 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-motA. Lanes 1-12, PCR amplicons from E. coli colonies 1-12 generated using primers M13 Forward and M13 Reverse. The expected amplicon size is 3.3 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

156

Plasmid pHN-motB was generated in the same manner as pHN-motA with some modifications. The 5’ and 3’ flanking regions of motB (5’ motB and 3’ motB) were amplified from T. denticola ATCC 35405 WT gDNA. The PCR amplicons with the desired sizes of 790 bp and 654 bp for 5’ motB and 3’ motB respectively, were purified (Figure 5.4A). Non-specific amplifications were observed with the fusion of 5’ motB with ermAM (1976 bp) and the fusion of fla promoter with 3’ motB (984 bp) (Figure 5.4B-C). The PCR amplicons of the desired sizes (as indicated by the red boxes in Figure 5.4B-C) were purified from the agarose gel and fused for a final product (motB::ermAM) of 2920 bp (Figure 5.4D). The purified motB::ermAM was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Of the transformed colonies selected on LB Amp plates, six were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. Transformant 4 possessed the desired 3.1 kbp amplicon (Figure 5.5). Plasmid from clone 4 was subjected to nucleotide sequencing to confirm the fidelity of the recombination cassette in pHN-motB. E. coli α-Select harboring pHN-motB was designated ECR766.

Figure 5.4 Generation of plasmid pHN-motB. A. PCR amplicons of 5’ motB, 3’ motB, fla promoter and ermAM. Lane 1, 5’ motB (790 bp); Lane 2, 3’ motB (654 bp); Lane 3, fla promoter (370 bp); Lane 4, ermAM (1226 bp). B. PCR amplicon of 5’ motB-ermAM (1976 bp). C. PCR amplicon of fla promoter-3’ motB (984 bp). D. PCR amplicon of motB::ermAM (2920 bp). All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The red boxes indicate DNA purified from the agarose gel. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

157

Figure 5.5 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-motB. Lanes 1-6, PCR amplicons from E. coli colonies 1-6 generated using primers M13 Forward and M13 Reverse. The expected amplicon size is 3.1 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

Plasmid pHN-flgE was generated in the same manner as pHN-motA with some modifications. The 5’ and 3’ flanking regions of flgE (5’ flgE and 3’ flgE) were amplified from T. denticola ATCC 35405 WT gDNA. PCR amplicons with the desired sizes of 822 bp, 680 bp, 2008 bp and 1010 bp for 5’ flgE, 3’ flgE, 5’ flgE-ermAM and fla promoter-3’ flgE respectively (Figure 5.6A-B), were purified and fused by SOE PCR. A final product (flgE::ermAM) of 2978 bp was obtained (Figure 5.6C). The purified flgE::ermAM was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Of the transformed colonies selected on LB Amp plates, eight were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. All transformants examined except clone 3 possessed the desired 3.2 kbp amplicon (Figure 5.7). Plasmid from clone 8 was subjected to nucleotide sequencing to confirm the fidelity of the recombination cassette in pHN-flgE. E. coli α-Select harboring pHN-flgE was designated ECR768.

158

Figure 5.6 Generation of plasmid pHN-flgE. A. PCR amplicons of 5’ flgE, 3’ flgE, fla promoter and ermAM. Lane 1, 5’ flgE (822 bp); Lane 2, 3’ flgE (680 bp); Lane 3, fla promoter (370 bp); Lane 4, ermAM (1226 bp). B. PCR amplicon of 5’ flgE-ermAM and fla promoter-3’ flgE. Lane 1, 5’ flgE-ermAM (2008 bp); Lane 2, fla promoter-3’ flgE (1010 bp) C. PCR amplicon of flgE::ermAM (2978 bp). All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Figure 5.7 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-flgE. Lanes 1-8, PCR amplicons from E. coli colonies 1-8 generated using primers M13 Forward and M13 Reverse. The expected amplicon size is 3.2 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

Plasmid pHN-cheY was generated in the same manner as pHN-motA with some modifications. The 5’ and 3’ flanking regions of cheY (5’ cheY and 3’ cheY) were amplified from T. denticola ATCC 35405 WT gDNA. The expected sizes of PCR amplicons for 5’ cheY and 3’ cheY were 781 bp and 792 bp, respectively (Figure 5.8A). The expected size of PCR amplicon for 5’ cheY-ermAM is 1967 bp. The PCR amplicon with the desired size (as indicated by the red box in Figure 5.8B) was purified from the agarose gel and fused by SOE PCR with 3’ cheY for a final product (cheY::ermAM) of 2739 bp (Figure 5.8). The purified cheY::ermAM was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Of the transformed colonies selected on LB Amp

159

plates, ten were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. All transformants examined except clones 1, 4 and 8 possessed the desired 3 kbp amplicon (Figure 5.9). Plasmid from clone five was subjected to nucleotide sequencing to confirm the fidelity of the recombination cassette in pHN-cheY. E. coli α-Select harboring pHN-cheY was designated ECR767.

Figure 5.8 Generation of plasmid pHN-cheY. A. PCR amplicons of 5’ cheY, 3’ cheY and ermAM. Lane 1, 5’ cheY (781 bp); Lane 2, 3’ cheY (792 bp); Lane 3, ermAM (1226 bp). B. PCR amplicon of 5’ cheY-ermAM (1967 bp). C. PCR amplicon of cheY::ermAM (2739 bp). All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The red box indicates DNA purified from the agarose gel. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Figure 5.9 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-cheY. Lanes 1-10, PCR amplicons from E. coli colonies 1-10 generated using primers M13 Forward and M13 Reverse. The expected amplicon size is 3 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

160

5.2.2 Transformation of T. denticola with pHN-motA, pHN-motB, pHN-cheY and pHN-flgE suicide plasmids Plasmids pHN-motA, pHN-motB, pHN-cheY and pHN-flgE were linearized by digestion with NotI to release the recombination cassettes with the desired sizes of 3.2 kbp, 2.9 kbp, 2.7 kbp and 3 kbp respectively, from the background vector (3 kbp). Only a single band was observed in all the digestions as the size of the background vector is similar to that of the recombination cassettes (Figure 5.10). After digestion, DNA precipitation was carried out to remove the enzyme. The digested products were resuspended in deionized water and used to transform T. denticola cells by electroporation. Transformants selected on OBGM Erm (40 µg/mL) agar plates were grown in OBGM Erm broth whereupon a small volume of each culture was used for PCR analysis to confirm DNA integration at the correct locus.

Figure 5.10 NotI-digestion of the plasmids pHN-motA, pHN-motB, pHN-cheY and pHN-flgE. A. NotI-digestion of pHN-motA where 3.2 kbp was expected for insert and 3 kbp was expected for vector. Lane 1, undigested pHN-motA; Lane 2, digested pHN- motA. B. NotI-digestion of pHN-motB where 2.9 kbp was expected for insert and 3 kbp was expected for vector. Lane 1, undigested pHN-motB; Lane 2, digested pHN-motB. C. NotI-digestion of pHN-cheY where 2.7 kbp was expected for insert and 3 kbp was expected for vector. D. NotI-digestion of pHN-flgE where 3 kbp was expected for both insert and vector. Lane 1, undigested pHN-flgE; Lane 2, digested pHN-flgE. All digestions were monitored via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

161

5.2.3 Generation of T. denticola ATCC 35405 motility mutants lacking motA, motB, flgE and cheY Transformation of T. denticola ATCC 35405 with cheY::ermAM yielded four colonies on OBGM Erm plates which were analyzed by PCR to confirm DNA integration at the correct locus. All transformants possessed the desired amplicons of sizes 3.0 kbp, 1.7 kbp and 2.0 kbp for the primer pairs 5’cheY-F/3’cheY-R, 5’cheY-F/ermAM-F short and ermAM-R short/3’cheY-R respectively (Table 2.4, Figure 5.11). Transformant one was designated T. denticola ATCC 35405 ∆cheY35405 (ECR823).

Figure 5.11 PCR to confirm homologous recombination of cheY::ermAM with T. denticola ATCC 35405 genome. Lane 1, 4 and 7, PCR amplicons generated using primers 5’cheY-F and 3’cheY-R showing the expected size of 2.2 kbp for WT and 3.0 kbp for ∆cheY35405 transformants; Lane 2, 5 and 8, PCR amplicons generated using primers 5’cheY-F and ermAM-F short showing the expected size of 1.7 kbp for ∆cheY35405 transformants and no product for WT; Lane 3, 6 and 9, PCR amplicons generated using primers ermAM-R short and 3’cheY-R showing the expected size of 2.0 kbp for ∆cheY35405 transformants and no product for WT. All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Transformations of T. denticola ATCC 35405 with motA::ermAM, motB::ermAM and flgE::ermAM were unsuccessful despite many trials (Table 3.5). As discussed previously in Chapter 3, the failure was not due to the competency of the cells as the cells were successfully transformed with the positive control TDE0911::ermAM.

162

5.2.4 Generation of motility mutants lacking flgE, motA, motB and cheY in T. denticola ATCC 35405 ∆TDE0911::kan The recombination cassettes flgE::ermAM, motA::ermAM, motB::ermAM and cheY::ermAM were used to transform T. denticola ATCC 35405 ∆TDE0911::kan (ECR824). Two hundred and twenty three colonies were obtained from the transformation with cheY::ermAM and four transformants were analyzed by PCR using the primers 5’cheY-F, 3’cheY-R, ermAM-fla prom-R and ermAM-2767-F to confirm DNA integration at the correct locus. All transformants examined possessed the desired amplicons of sizes 3.0 kbp, 2.1 kbp and 2.1 kbp for the primer pairs 5’cheY-F/3’cheY-R, 5’cheY-F/ ermAM-fla prom-R and ermAM-2767-F/3’cheY-R respectively (Table 2.4; Figure 5.12). Transformant one was designated T. denticola ∆TDE0911::kan ∆cheY (ECR825).

Figure 5.12 PCR to confirm homologous recombination of cheY::ermAM with the T. denticola ∆TDE0911::kan genome. Lanes 1, 7 and 13, PCR amplicons from ΔTDE0911::kan control; Lanes 2, 8 and 14, NTC; Lanes 3-6, 9-12 and 15-18, PCR amplicons from ∆TDE0911::kan ∆cheY transformants 1-4; Lanes 1-6, PCR amplicons generated using primers 5’cheY-F and 3’cheY-R showing the expected size of 2.2 kbp for ΔTDE0911::kan and 3.0 kbp for ∆TDE0911::kan ∆cheY; Lanes 7-12, PCR amplicons generated using primers 5’cheY-F and ermAM-fla prom-R showing the expected size of 2.1 kbp for ∆TDE0911::kan ∆cheY and no product for ΔTDE0911::kan; Lanes 13-18, PCR amplicons generated using primers ermAM-2767-F and 3’cheY-R showing the expected size of 2.1 kbp for ∆TDE0911::kan ∆cheY and no product for ΔTDE0911::kan. All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

163

Two colonies were obtained from the transformation of T. denticola ∆TDE0911::kan with flgE::ermAM and analyzed by PCR to confirm DNA integration at the correct locus. Both transformants possessed the desired amplicons of sizes 3.2 kbp, 2.1 kbp and 2.3 kbp for the primer pairs 5’flgE-F/3’flgE-R, 5’flgE-F/ermAM-fla prom-R and ermAM-2767- F/3’flgE-R respectively (Figure 5.13). Transformant one was designated T. denticola ∆TDE0911::kan ∆flgE (ECR826).

Figure 5.13 PCR to confirm homologous recombination of flgE::ermAM with T. denticola ∆TDE0911::kan genome. Lane 1, PCR amplicons from T. denticola ATCC 35405 WT control; Lanes 2-3, PCR amplicons from ∆TDE0911::kan ∆flgE transformants 1-2; Lane 4, NTC. A. PCR amplicons generated using primers 5’flgE-F and 3’flgE-R showing the expected size of 3.1 kbp for WT and 3.2 kbp for ∆TDE0911::kan ∆flgE. B. PCR amplicons generated using primers 5’flgE-F and ermAM-fla prom-R showing the expected size of 2.1 kbp for ∆TDE0911::kan ∆flgE and no product for WT. C. PCR amplicons generated using primers ermAM-2767-F and 3’flgE-R showing the expected size of 2.3 kbp for ∆TDE0911::kan ∆flgE and no product for WT. All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Transformations of T. denticola ∆TDE0911::kan with motA::ermAM or motB::ermAM were unsuccessful but the failure was not due to the competency of the cells as discussed in Chapter 3.

5.2.5 Generation of T. denticola ATCC 33520 motility mutants lacking flgE, motA, motB and cheY The organization of flgE, motA, motB and cheY in the genome of T. denticola ATCC 33520 is similar to that of ATCC 35405 (Table 5.1 and Table 5.2). The fla operon and chemotaxis operon of T. denticola ATCC 33520 has 98.4% and 98.9% identity respectively with that of T. denticola ATCC 35405 when aligned using the EMBOSS needle protein pairwise sequence alignment tool.

164

Table 5.1 Genes in the fla operon of T. denticola ATCC 35405 and ATCC 33520.

Gene name T. denticola ATCC 35405 T. denticola ATCC 33520 fliP TDE2760 HMPREF9722_RS03065 orf10 TDE2761 HMPREF9722_RS03060 fliY TDE2762 HMPREF9722_RS03055 fliM TDE2763 HMPREF9722_RS03050 fliL TDE2764 HMPREF9722_RS03045 motB TDE2765 HMPREF9722_RS03040 motA TDE2766 HMPREF9722_RS03035 orf4 TDE2767 HMPREF9722_RS03030 flgE TDE2768 HMPREF9722_RS03025 flgD TDE2769 HMPREF9722_RS03020 fliK TDE2770 HMPREF9722_RS03015

Table 5.2 Genes in the che operon of T. denticola ATCC 35405 and ATCC 33520.

Gene name T. denticola ATCC 35405 T. denticola ATCC 33520 cheA TDE1491 HMPREF9722_RS08900 cheW TDE1492 HMPREF9722_RS08895 cheX TDE1493 HMPREF9722_RS08890 cheY TDE1494 HMPREF9722_RS08885

Transformations of T. denticola ATCC 33520 with flgE::ermAM, motA::ermAM, motB::ermAM and cheY::ermAM were successful. Of the colonies transformed with motA::ermAM selected on OBGM Erm agar plates, six were analyzed by PCR using primers 5’motA-F and 3’motA-R. All transformants examined possessed the desired amplicon of size 3.6 kbp (Figure 5.14). Transformant two was designated T. denticola

ATCC 33520 ∆motA33520 (ECR827).

165

Figure 5.14 PCR to confirm homologous recombination of motA::ermAM with T. denticola ATCC 33520 genome. Lane 1, PCR amplicon from T. denticola ATCC 33520 WT control; Lanes 2-7, PCR amplicons from ∆motA33520 transformants 1-6. Primers 5’motA-F and 3’motA-R were used. The expected sizes of PCR amplicons were 2.9 kbp for WT and 3.6 kbp for ∆motA33520. All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

Of the colonies transformed with motB::ermAM selected on OBGM Erm agar plates, six were analyzed by PCR using primers 5’motB-F and 3’motB-R. All transformants examined possessed the desired amplicon of size 3.3 kbp (Figure 5.15). Transformant one was designated T. denticola ATCC 33520 ∆motB33520 (ECR828).

Figure 5.15 PCR to confirm homologous recombination of motB::ermAM with T. denticola ATCC 33520 genome. Lane 1, PCR amplicon from T. denticola ATCC 33520 WT control; Lanes 2-7, PCR amplicons from ∆motB33520 transformants 1-6. Primers 5’motB-F and 3’motB-R were used. The expected sizes of PCR amplicons were 2.5 kbp for WT and 3.3 kbp for ∆motB33520. All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

166

Of the colonies transformed with flgE::ermAM selected on OBGM Erm agar plates, four were analyzed by PCR. All transformants examined possessed the desired amplicons of sizes 3.2 kbp, 2.1 kbp and 2.3 kbp for the primer pairs 5’flgE-F/3’flgE-R, 5’flgE- F/ermAM-fla prom-R and ermAM-2767-F/3’flgE-R respectively (Figure 5.16).

Transformant two was designated T. denticola ATCC 33520 ∆flgE33520 (ECR829).

Figure 5.16 PCR to confirm homologous recombination of flgE::ermAM with T. denticola ATCC 33520 genome. A. Lane 1, PCR amplicon from T. denticola ATCC 33520 WT control; Lane 2, NTC; Lanes 3-6, PCR amplicons from ∆flgE33520 transformants 1-4 using primers 5’flgE-F and 3’flgE-R. The expected sizes of PCR amplicons were 3.1 kbp for WT and 3.2 kbp for ∆flgE33520. B. Lanes 1 and 7, PCR amplicons from T. denticola ATCC 33520 WT control; Lanes 2 and 8, NTC; Lanes 3-6 and 9-12, PCR amplicons from ∆flgE33520 transformants 1-4; Lanes 1-6, PCR amplicons using primers 5’flgE-F and ermAM-fla prom-R showing the expected size of 2.1 kbp for ∆flgE33520 and no product for WT; Lanes 7-12, PCR amplicons using primers ermAM- 2767-F and 3’flgE-R showing the expected size of 2.3 kbp for ∆flgE33520 and no product for WT. All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Of the colonies transformed with cheY::ermAM selected on OBGM Erm agar plates, four were analyzed by PCR. All transformants examined possessed the desired amplicons of sizes 2.9 kbp, 2.1 kbp and 2.1 kbp for the primer pairs 5’cheY-F/3’cheY-R, 5’cheY-F/ ermAM-fla prom-R and ermAM-2767-F/3’cheY-R respectively (Figure 5.17).

Transformant one was designated T. denticola ATCC 33520 ∆cheY33520 (ECR830).

167

Figure 5.17 PCR to confirm homologous recombination of cheY::ermAM with T. denticola ATCC 33520 genome. Lanes 1, 7 and 13, PCR amplicons from T. denticola ATCC 33520 WT control; Lanes 2, 8 and 14, NTC; Lanes 3-6, 9-12 and 15-18, PCR amplicons from ∆cheY33520 transformants 1-4; Lanes 1-6, PCR amplicons using primers 5’cheY-F and 3’cheY-R showing the expected size of 2.2 kbp for WT and 3.0 kbp for ∆cheY33520; Lanes 7-12, PCR amplicons using primers 5’cheY-F and ermAM-fla prom-R showing the expected size of 2.1 kbp for ∆cheY33520 and no product for WT; Lanes 13- 18, PCR amplicons using primers ermAM-2767-F and 3’cheY-R showing the expected size of 2.1 kbp for ∆cheY33520 and no product for WT. All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

5.2.6 Characterization of T. denticola ATCC 33520 motility mutants ∆motA33520,

∆motB33520, ∆flgE33520 and ∆cheY33520

5.2.6.1 Genomic sequencing of ∆motA33520, ∆motB33520, ∆flgE33520 and ∆cheY33520 Genomic sequencing was performed to confirm the authenticity of the various T. denticola ATCC 33520 mutants, ∆motA33520 (ECR827), ∆motB33520 (ECR828), ∆flgE33520

(ECR829) and ∆cheY33520 (ECR830). All the mutants had the appropriate target genes deleted from their genomes, but ∆cheY33520 had an additional large genomic excision of 53 kbp and thus was not further characterized. No other recombination events or excisions were detected in the remaining mutants and although some single nucleotide polymorphisms (SNPs) were found (Appendix II), they were not predicted to affect the phenotype of the mutants. These mutants were subjected to further examination.

168

5.2.6.2 Morphological differences between T. denticola ATCC 33520 and the

motility mutants ∆motA33520, ∆motB33520 and ∆flgE33520

∆motA33520, ∆motB33520 and ∆flgE33520 grown on OBGM agar plates appeared as small, dense and pinpoint-shaped colonies. The edges of the colonies were defined instead of diffusing outward like that of ATCC 33520 (Figure 5.18).

Figure 5.18 Colony morphologies of T. denticola ATCC 33520 and ∆motB33520. T. denticola cells were inoculated into OBGM agar [OBGM supplemented with 0.8% (w/v) UltraPure™ Low Melting Point Agarose] and poured into petri dishes which were then incubated anaerobically for two weeks. T. denticola ATCC 33520 formed colonies with diffuse edges while ∆motB33520 formed colonies with defined edges. The colonies of ∆motB33520 were smaller and denser than those of ATCC 33520. The colonies formed by ∆motA33520 and ∆flgE33520 had a similar morphology to the colonies of ∆motB33520.

5.2.6.3 Swimming assay of T. denticola ATCC 33520, ∆motA33520, ∆motB33520 and

∆flgE33520 A swimming assay on semisolid agar was used to check for bacterial motility. Motile bacteria are able to move outward from the point of inoculation to form a turbid plaque. T. denticola cells (107 cfu) were grown to exponential phase before they were harvested, resuspended in deionized water and spotted onto semisolid OBGM agar medium. The area of turbidity was measured after 10 days of anaerobic incubation. The area of turbidity 2 2 2 was 330.1 ± 62.7 mm for ATCC 33520, 9.7 ± 3.3 mm for ∆flgE33520, 14.4 ± 10.1 mm 2 for ∆motA33520 and 8.6 ± 1.8 mm for ∆motB33520 (Figure 5.19). The small area of turbidity and the lack of diffused edge indicated that the mutants did not swim outward from the point of inoculation.

169

Figure 5.19 Swimming assay of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520. T. denticola cells were grown to exponential phase, harvested and spotted on semisolid OBGM agar [OBGM supplemented with 0.4% (w/v) Molecular Grade Agarose and 1% (w/v) BD Difco™ gelatin]. The plates were incubated anaerobically for 10 days before the area of turbid plaque was measured. A. The data are presented as means ± standard deviations (N=29) and were analyzed by Students’ T-test. Values that were significantly different (p<0.05) from the value for T. denticola ATCC 33520 are indicated by an asterisk (*). B. Representative image of the swimming assay of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520 after 10 days of anaerobic incubation.

5.2.6.4 Cryo-electron microscopy of T. denticola ATCC 33520, ∆flgE33520,

∆motA33520, ∆motB33520

T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520 cells at exponential phase were imaged with cryo-electron microscopy (cryo-EM) to check for the presence of periplasmic flagella and to determine cellular morphology. Numerous cryo-EM micrographs were obtained for each strain and compared. Periplasmic flagella were observed in T. denticola ATCC 33520, ∆motA33520 and ∆motB33520 but not ∆flgE33520 as expected. Up to four or five periplasmic flagella were distinguishable in T. denticola ATCC 33520 but only two or three periplasmic flagella were distinguishable in

∆motA33520 and ∆motB33520. Not all cells or cell segments of ∆motA33520 and ∆motB33520 showed visible flagella. Similar with the finding of Ruby et al. (1997), T. denticola ATCC

170

33520 cells showed irregular twisted morphology, with both planar and helical regions.

∆motA33520 and ∆motB33520 cells appeared straighter than ATCC 33520 while many cells or cell segments of ∆flgE33520 had limited or no spirality and appeared rod shaped. The cells of ∆flgE33520, ∆motA33520 and ∆motB33520 were more elongated than the cells of ATCC 33520 (Figure 5.20).

171

33520 ∆flgE33520

172

∆motA33520 ∆motB33520

Figure 5.20 Representative cryo-EM images of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520. T. denticola cells were grown to exponential phase and viewed directly under cryo-EM. Red arrows point to the periplasmic flagella.

173

5.2.6.5 Growth of T. denticola ATCC 33520, ∆motA33520, ∆motB33520 and ∆flgE33520

The growth of T. denticola ATCC 33520, ∆motA33520 (ECR827), ∆motB33520 (ECR828) and ∆flgE33520 (ECR829) was examined by monitoring the absorbance (A650) of the batch cultures (N=3) at 6-24 h intervals for a total duration of 264 h (Figure 5.21). The maximum A650 at stationary phase, i.e. A650 after 168 h, achieved by the mutants was less than that of the WT (0.44 ± 0.04) (p<0.05). The maximum A650 of ∆motA33520 was 0.32 ±

0.04, the maximum A650 of ∆flgE33520 was 0.31 ± 0.04 and the maximum A650 of

∆motB33520 was 0.26 ± 0.06. These results suggest that the mutants have a lower maximum cell density at stationary phase than the WT. ∆flgE33520 entered exponential phase and reached stationary phase faster than the WT. It entered exponential phase after 24 h whereas the WT entered exponential phase after 48 h. ∆flgE33520 reached stationary phase at 96 h whereas the WT reached stationary phase at 142 h. Although ∆motA33520 entered exponential phase at approximately the same time as the WT (48 h), it reached stationary phase slower than the WT (168 h). ∆motB33520 entered exponential phase at a much later time (72 h) but it reached stationary phase at a similar time as ∆motA33520 (168 h).

The mean generation time was calculated to determine the growth rate of each strain of

T. denticola. It was calculated based on the rate of change of A650 during the exponential growth phase of each culture. The mean generation time was determined between 48 and

72 h post-inoculation for ATCC 33520, between 24 and 72 h for ∆flgE33520, between 24 and 72 h for ∆motA33520 and between 48 and 96 h for ∆motB33520. The mean generation times for ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520 are 14 ± 3, 22 ± 3, 22 ± 3 and 24 ± 5 h respectively. The increase in generation time of the mutants (p<0.05) indicates that all the mutants have lower growth rate than the WT.

174

Figure 5.21 Growth curves of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520. T. denticola cultures at stationary phase were inoculated into fresh OBGM medium at t=0. A650 of the cultures was measured at 6-24 h intervals for 264 h. The data are presented as means ± standard deviations (N=3). Student’s T-test showed that the maximum A650 of the WT was significantly higher than that of all the mutants (p<0.05). The mean generation time was calculated based on the rate of change of A650 during the exponential growth phase of cultures (N=3).

5.2.6.6 Autoaggregation and coaggregation assays of T. denticola ATCC 33520,

∆flgE33520, ∆motA33520, ∆motB33520 and P. gingivalis W50 Autoaggregation and coaggregation assays are used to determine the ability of a bacterium to bind to itself or with another bacterium. Large aggregates form as a result of bacterial binding and sink to the bottom of the tube. T. denticola and P. gingivalis cells at exponential growth phase were harvested, washed twice with coaggregation buffer and adjusted to A650 of 0.5 with the coaggregation buffer. For the coaggregation assay, T. denticola and P. gingivalis cell suspensions were combined to give an A650 of 0.5. The height of bacterial aggregates in a cuvette was monitored for 7 h. The decrease in height indicates that the bacteria have autoaggregated or coaggregated and the aggregates have sunk to the bottom of the cuvette.

175

T. denticola ATCC 33520 autoaggregated at the highest rate, with all bacterial aggregates sinking to the bottom of the cuvette within 2 h of incubation. ∆motA33520 autoaggregated slower than the WT as the height decreased slowly and reached the bottom of the cuvette at the end of the 7 h incubation. ∆motB33520 autoaggregated slower than ∆motA33520 and the bacterial aggregates did not sink completely to the bottom of the cuvette at the end of the incubation time. ∆flgE33520 autoaggregated at the slowest rate, with the height of its bacterial aggregates starting to slowly decrease after 1 h of incubation and eventually reaching approximately halfway by the end of the 7 h incubation. P. gingivalis W50 did not autoaggregate as the height of its suspension remained constant throughout the 7 h of incubation (Figure 5.22).

When T. denticola ATCC 33520, ∆motA33520 and ∆motB33520 were mixed with P. gingivalis W50, the height of their bacterial aggregates decreased rapidly and all bacterial aggregates sank to the bottom of the cuvette within 2 h of incubation. ∆flgE33520 coaggregated slower with P. gingivalis W50 than the others but all of the bacterial aggregates had sunk to the bottom of the cuvette by 5 h of incubation (Figure 5.23).

Figure 5.22 Autoaggregation of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520, ∆motB33520 and P. gingivalis W50. T. denticola and P. gingivalis cells at exponential growth phase were harvested, washed twice with coaggregation buffer and adjusted to A650 of 0.5. The height of bacterial aggregates was monitored for 7 h. The data are presented as means ± standard deviations (N=3). Standard deviation bars of P. gingivalis W50 are invisible because they are very small.

176

Figure 5.23 Coaggregation of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520 with P. gingivalis W50. T. denticola and P. gingivalis cells at exponential growth phase were harvested, washed twice with coaggregation buffer and adjusted to A650 of 0.5 with the coaggregation buffer. Equal volumes of T. denticola and P. gingivalis cell suspensions were combined. The height of bacterial aggregates was monitored for 7 h. The data are presented as means ± standard deviations (N=3).

5.2.6.7 Reverse Transcription-Polymerase Chain Reaction (RT-PCR) analysis of

fla operon transcription in ∆flgE33520, ∆motA33520 and ∆motB33520 The genes encoding FlgE, MotA and MotB (flgE, motA and motB) are part of an 11 gene operon starting with fliK and ending with fliP. RT-PCR was performed using primers specific to each gene to examine transcription of the fla operon. PCR of the negative controls, which are ‘no template’ and ‘RNA’, exhibited no bands, whilst PCRs of the WT gDNA positive control generated bands of the appropriate size (802 bp for fliK/flgD, 770 bp for motA/motB, 855 bp for motB/fliL, 660 bp for TDE2767/motA and 634 bp for fliL/fliM). The upstream and downstream genes of flgE, motA and motB were transcribed, indicating that the genomic deletion did not have serious deleterious effects on transcription of neighboring genes in the operon. These results showed that the native fla promoter fragment reintroduced downstream of deletion points was functional. However, due to a lack of an appropriate constitutive control (i.e. a housekeeping gene) for normalization, the quantitative differences in transcriptional level between these genes in WT and mutants could not be accurately determined.

177

Figure 5.24 RT-PCR analysis of fla operon transcription in ∆motA33520, ∆motB33520 and ∆flgE33520. A. Diagrammatic representation of the fla operon in T. denticola ATCC 33520 genome. The fla operon is made up of 11 genes from HMPREF9722_RS03015 to HMPREF9722_RS03065. Predicted functions of gene products are annotated. RT-PCR transcripts amplified by specific forward and reverse primer pairs are indicated by

178

forward and reverse arrows, respectively. B. Investigation of transcription of fliK/flgD (upstream genes) and motA/motB (downstream genes) in ∆flgE33520. RT-PCR was performed using MyTaq™ HS DNA polymerase (Bioline) and primer pairs specific to fliK/flgD (TDE2768-F and TDE2768-R) and motA/motB (TDE2766-F and TDE2766-R). C. Investigation of transcription of fliK/flgD (upstream genes) and motB/fliL (downstream genes) in ∆motA33520. RT-PCR was performed using MyTaq™ HS DNA polymerase (Bioline) and primer pairs specific to fliK/flgD (TDE2768-F and TDE2768-R) and motB/fliL (TDE2765-F and TDE2765-R). D. Investigation of transcription of TDE2767/motA (upstream genes) and fliL/fliM (downstream genes) in ∆motB33520. RT-PCR was performed using MyTaq™ HS DNA polymerase (Bioline) and primer pairs specific to TDE2767/motA [TDE2767-F (flgE KO) and TDE2767-R (flgE KO)] and fliL/fliM (TDE2764-F and TDE2764-R). The templates used for PCR were no template (lane 1), mutant RNA (lane 2), T. denticola ATCC 33520 gDNA (lane 3), reverse transcribed mutant RNA (lane 4) and reverse transcribed T. denticola ATCC 33520 RNA (lane 5). All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

179

5.2.6.8 Identification of proteins that had changed in abundance in ∆flgE33520,

∆motA33520 and ∆motB33520 In order to examine proteins that had changed in abundance in the motility mutants relative to the WT ATCC 33520, their total protein expression was investigated using quantitative proteomics on the whole cell lysates. A total of 1330 proteins (48% of the total proteins) were identified in the whole cell lysates of the WT and mutants. There were 64, 337 and 326 proteins that had changed in abundance in ∆flgE33520, ∆motA33520 and ∆motB33520 respectively as compared to ATCC 33520 (Appendix III). The proteins FlgE (TDE2768), MotA (TDE2766) and MotB (TDE2765) were found in ATCC 33520 but not in ∆flgE33520, ∆motA33520 and ∆motB33520 respectively, confirming that the deletion of the genes had abolished the expression of the proteins in the respective mutants. Proteins significantly changed in abundance were sorted into functional categories on the basis of clusters of orthologous groups (COG) (Tatusov et al. 1997, Tatusov et al. 2003).

Of the 64 proteins in ∆flgE33520, 42 were significantly decreased in abundance and 22 were significantly increased in abundance. Most of the proteins (34%) were assigned to COGs related to cellular processes and signaling. Of these, COG category N (cell motility) had the highest number of differentially expressed proteins (19%) (Table 5.3), indicating that the proteins that had changed in abundance in ∆flgE33520 were more focused on flagella and motility.

Of the 337 proteins in ∆motA33520, 213 were significantly decreased in abundance and 124 were significantly increased in abundance. Ninety-one (27%) of the differentially expressed proteins were assigned to COGs related to metabolism, 87 proteins (26%) were assigned to COGs related to cellular processes and signaling and 36 proteins (11%) were assigned to COGs related to information storage and processing. COG categories T (signal transduction mechanisms) and E (amino acid transport and metabolism) had the highest number of differentially expressed proteins (Table 5.3).

Of the 326 proteins in ∆motB33520, 229 were significantly decreased in abundance and 97 were significantly increased in abundance. Ninety-two proteins (28%) were assigned to COGs related to metabolism, 80 proteins (25%) were assigned to COGs related to cellular processes and signaling and 32 proteins (10%) were assigned to COGs related to information storage and processing. Similar to ∆motA33520, COG categories T (signal transduction mechanisms) and E (amino acid transport and metabolism) had the highest

180

number of differentially expressed proteins (Table 5.3). The number of proteins in each

COG category is similar between ∆motA33520 and ∆motB33520.

181

Table 5.3 Total and unique proteins that were significantly changed in abundance in T. denticola ∆flgE33520, ∆motA33520 and ∆motB33520, grouped by COG category.

Total proteinsA Unique proteinsB C COG ∆flgE33520 ∆motA33520 ∆motB33520 ∆flgE33520 ∆motA33520 ∆motB33520 Information storage J 3 17 18 1 7 6 and processing K 1 7 5 - 4 1 L 2 12 9 - 6 3 Cellular processes and D 1 6 4 1 2 - signaling M 3 18 14 3 8 4 N 12 17 14 - 2 1 O 3 12 12 1 3 3 T 1 21 23 - 8 9 U 2 7 9 - 1 2 V - 6 4 - 3 1 Metabolism C 5 13 12 4 6 5 E 2 20 22 1 5 7 F 3 13 10 1 4 1 G 1 15 16 1 5 6 H 1 5 7 1 1 3 I 2 7 6 2 1 - P 3 16 17 - 4 5 Q - 2 2 - 1 1 Poorly or not characterized R - 8 8 - - - S 17 91 94 8 31 30 N/A 2 24 20 1 7 3 Total 64 337 326 25 109 91 A Total number of proteins that had changed in abundance in each mutant. B Proteins that had changed in abundance in the specific mutant only and not in the other two mutants.

182

C One-letter abbreviations for the functional COG categories: J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosome partitioning; V, defense mechanisms; T, signal transduction mechanisms; M, cell wall/membrane/envelope biogenesis; N, cell motility; U, intracellular trafficking, secretion, and vesicular transport; O, posttranslational modification, protein turnover, chaperones; C, energy production and conversion; G, carbohydrate transport and metabolism; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; R, general function prediction only; S, function unknown.

183

Twenty-four proteins had changed in abundance in all three mutants (Figure 5.25, Table 5.4 and Table 5.5). The direction of change in protein abundance was consistent between

∆motA33520 and ∆motB33520 but not necessarily with ∆flgE33520. Most of the proteins (42%) clustered in category N. All of these proteins, including the detectable proteins expressed from the fla operon (TDE2764-TDE2768), FlaB (TDE1004), FlaA (TDE1408, TDE1409, TDE1712) and a flagellar hook-associated protein FlgL (TDE2353), were significantly decreased in abundance in all three mutants, except TDE0119 (FliS) which was decreased in abundance in ∆motA33520 and ∆motB33520 but was increased in abundance in ∆flgE33520. Notably, the adenosine triphosphate (ATP) binding cassette (ABC) transporter proteins

TDE0758, TDE0984-TDE0986 were increased in abundance in ∆motA33520 and

∆motB33520 but decreased in abundance in ∆flgE33520. Two other proteins, TDE1754 (desulfoferrodoxin/neelaredoxin) and TDE2721 (helicase domain protein), were also increased in abundance in ∆motA33520 and ∆motB33520 but decreased in abundance in

∆flgE33520. TDE1208 (DNA topoisomerase I) was increased in abundance in all three mutants.

Figure 5.25 Number of proteins that had changed in abundance in ∆flgE33520, ∆motA33520 and ∆motB33520. The Venn diagram summarizes the number of proteins that had changed in abundance in the motility mutants as compared to the WT ATCC 33520. There were 64, 337 and 326 proteins that had changed in abundance in ∆flgE33520, ∆motA33520 and ∆motB33520 respectively. Twenty-four proteins were identified from all three mutants and 224 proteins were identified from ∆motA33520 and ∆motB33520. One hundred and nine, 91 and 25 proteins were changed in abundance in ∆motA33520, ∆motB33520 and ∆flgE33520 only, respectively.

184

Table 5.4 Proteins significantly changed in abundance in all three T. denticola mutants or ∆motA33520 and ∆motB33520 only, grouped by COG category.

Number of proteins changed in increased in decreased in COGA abundance in all abundance in abundance in three mutants ∆motA33520 and ∆motA33520 and ∆motB33520 ∆motB33520 Information J - 7 3 storage K - 1 2 and processing L 2 4 2 Cellular processes D - 3 1 and signaling M - - 10 N 10 - 13 O - 2 6 T - 5 8 U - 2 4 V - - 3 Metabolism C 1 4 3 E 1 6 9 F 2 3 6 G - 3 7 H - 2 2 I - - 6 P 3 2 10 Q - - 1 Poorly or not R - 2 6 characterized S - 14 45 N/A 5 5 12 Total 24 65 159 A One-letter abbreviations for the functional COG categories: J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosome partitioning; V, defense mechanisms; T, signal transduction mechanisms; M, cell wall/membrane/envelope biogenesis; N, cell motility; U, intracellular trafficking, secretion, and vesicular transport; O, posttranslational modification, protein turnover, chaperones; C, energy production and conversion; G, carbohydrate transport and metabolism; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; R, general function prediction only; S, function unknown.

185

Table 5.5 Proteins significantly changed in abundance in all three T. denticola ∆motA33520, ∆motB33520 and ∆flgE33520 mutants relative to wild-type (ratio≥1.5 and ≤0.67, p<0.05). Shading indicates proteins predicted to be organized in an operon.

Wild-type ∆motA ∆motB ∆flgE Locus Tag Protein description 33520 33520 33520 COGC AbundanceA ratioB ratioB ratioB iron compound ABC transporter, periplasmic iron compound- TDE0758 6661050 1.69 4.14 0.23 P binding protein, putative oligopeptide/dipeptide ABC transporter, permease protein, TDE0984 5710433 5.94 8.79 0.00 P putative oligopeptide/dipeptide ABC transporter, periplasmic peptide- TDE0985 321778333 3.19 5.20 0.33 E binding protein, putative TDE0986 oligopeptide/dipeptide ABC transporter, ATP-binding protein 2915917 13.84 12.92 0.00 P TDE1208 DNA topoisomerase I (TopA) 60318667 1.94 1.93 1.79 L TDE1234 hypothetical protein 363805 2.95 3.81 1.51 - TDE1754 desulfoferrodoxin/neelaredoxin 68045667 2.68 2.71 0.54 C TDE2721 helicase domain protein 3375000 2.28 2.70 0.67 L TDE0119 flagellar protein FliS 1486100 0.00 0.00 2.71 N TDE0423 hypothetical protein 18472000 0.25 0.27 0.53 - TDE0463 purine nucleoside phosphorylase (DeoD) 54131167 0.43 0.55 0.62 F TDE0501 hypothetical protein 9236483 0.00 0.00 0.50 - TDE1004 flagellar filament core protein (FlaB) 528188333 0.33 0.24 0.01 N TDE1318 hypothetical protein 759610 0.00 0.00 0.00 - TDE1408 flagellar filament outer layer protein (FlaA) 591515000 0.30 0.29 0.04 N TDE1409 flagellar filament outer layer protein (FlaA) 558983333 0.28 0.28 0.04 N TDE1712 flagellar filament outer layer protein (FlaA) 1851283333 0.20 0.18 0.02 N TDE2085 amino acid kinase family protein 72132167 0.43 0.46 1.55 F TDE2353 flagellar hook-associated protein (FlgL) 1834867 0.00 0.00 0.59 N

186

TDE2764 flagellar protein FliL 64643667 0.14 0.12 0.50 N TDE2765 flagellar motor rotation protein B (MotB) 34357167 0.11 0.00 0.35 N TDE2766 motility protein A (MotA) 34589333 0.00 0.42 0.44 N TDE2768 flagellar hook protein FlgE 43882167 0.47 0.33 0.00 N TDE2779 hypothetical protein 59470000 0.11 0.10 0.31 - A The abundance of each protein in the wild-type T. denticola ATCC 33520 was calculated from the average Intensity Based Absolute Quantification (IBAQ) intensity from three replicates. B Geometric mean of ratios, from three replicates, produced from the LFQ intensity of protein in mutant relative to that of protein in wild-type. Ratio of ≥1.5 indicates that the protein had increased in abundance in mutant relative to wild-type and ratio of ≤0.67 indicates that the protein had decreased in abundance in mutant relative to wild-type. C One-letter abbreviations for the functional COG categories: J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosome partitioning; V, defense mechanisms; T, signal transduction mechanisms; M, cell wall/membrane/envelope biogenesis; N, cell motility; U, intracellular trafficking, secretion, and vesicular transport; O, posttranslational modification, protein turnover, chaperones; C, energy production and conversion; G, carbohydrate transport and metabolism; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; R, general function prediction only; S, function unknown.

187

Two hundred and twenty-four proteins were differentially regulated in both ∆motA33520 and ∆motB33520, with 65 proteins increased in abundance and 159 decreased in abundance in both mutants (Figure 5.25, Table 5.6-Table 5.8). The direction of change in protein abundance was similar between ∆motA33520 and ∆motB33520 except for one protein, 54 TDE2079 (σ -dependent transcriptional regulator), which was upregulated in ∆motA33520 but downregulated in ∆motB33520. The COG category with the highest number of proteins was N (cell motility) and all of the 13 proteins were decreased in abundance. All of the 10 proteins in category M (cell wall/membrane/envelope biogenesis), 3 proteins in category V (defense mechanisms), 6 proteins in category I (lipid transport and metabolism) and 1 protein in category Q (secondary metabolites biosynthesis, transport and catabolism) were also decreased in abundance. Of the proteins that were increased in abundance, 20 proteins (31%) were assigned to COGs related to metabolism, 12 proteins (19%) were assigned to COGs related to information storage and processing while 11 proteins (17%) were assigned to COGs related to cellular processes and signaling. COG category J (translation, ribosomal structure and biogenesis) contained the highest number of proteins that had increased in abundance. These proteins include the ribosomal proteins L6 (TDE0782), L18 (TDE0783) and S9 (TDE0853) as well as several aminoacyl-tRNA synthetases such as valyl-tRNA synthetase (TDE1364), tryptophanyl-tRNA synthetase (TDE1588) and phenylalanyl-tRNA synthetase (TDE1927). As aminoacyl-tRNA synthetases are enzymes that catalyze the attachment of a given amino acid to its cognate tRNA (Ling et al. 2009), the increased abundance of these aminoacyl-tRNA synthetases may suggest the increased usage of the corresponding amino acids valine, tryptophan and phenylalanine in ∆motA33520 and ∆motB33520. There was also an increase in the abundance of proteins involving in replication, recombination and repair (category L) and transcription (category K), such as recombination protein A (RecA; TDE0872), DNA topoisomerase I (TopA; TDE1208) and transcription termination factor Rho (Rho; TDE1503). Several proteins involved in cell cycle control, cell division, chromosome partitioning (category D) were also increased in abundance, including a rod-shaped determining protein MreB (TDE1349) and a chromosome partition protein SmC (TDE1496). The metabolic enzymes FAD-dependent oxidoreductase (TDE2643), Na+- translocating NADH-quinone reductase E subunit (NqrE; TDE0834), thiamine biosynthesis protein ThiI (TDE2568), glycosyl hydrolase (TDE1730) and glycine reductase complex selenoprotein (GrdA; TDE0745) were among the proteins that had highly increased in abundance in both ∆motA33520 and ∆motB33520. Two methyl-accepting 188

chemotaxis proteins (MCPs; TDE1009 and TDE2549) and several ABC transporter proteins (TDE0758, TDE0983-TDE0987) also increased in abundance in both mutants.

Of the proteins that were decreased in abundance in both ∆motA33520 and ∆motB33520, 45 proteins (28%) were assigned to COGs related to cellular processes and signaling, 44 proteins (28%) were assigned to COGs related to metabolism, seven proteins (4%) were assigned to COGs related to information storage and processing. COG category N contained the highest number of proteins that were decreased in abundance. These included the proteins expressed from the fla operon (TDE2761-TDE2768), FlaB (TDE1004, TDE1477), FlaA (TDE1408, TDE1409, TDE1712) and a flagellar hook- associated protein FlgL (TDE2353). A MCP (TDE2783), 2 chemotaxis proteins (CheX and CheY) and 9 ABC transporter proteins (TDE0386, TDE0425, TDE0813, TDE1226, TDE1271, TDE1416, TDE2092, TDE2356-TDE2357) were also among the proteins that had decreased in abundance. The virulence factors of T. denticola including Msp (TDE0405) and hemolysin (TDE1669) as well as the glycine reductase complex proteins GrdD (TDE0239) and GrdE2 (TDE2120) were also decreased in abundance.

Twenty-five proteins were changed in abundance in ∆flgE33520 only, with 15 proteins increased in abundance and 10 proteins decreased in abundance (Figure 5.25, Table 5.3). Most of these proteins were assigned to COGs related to metabolism (40%), followed by cellular processes and signaling (20%) then information storage and processing (4%). COG category C contained the highest number of proteins, followed by M and I. The OmpA family protein TDE0664 was increased in abundance. A ribosomal protein (TDE0777) was decreased in abundance.

One hundred and nine proteins were changed in abundance in ∆motA33520 only, with 58 proteins increased in abundance and 51 proteins decreased in abundance (Figure 5.25, Table 5.3). Most of these proteins were assigned to COGs related to cellular processes and signaling (25%) as well as metabolism (25%), followed by information storage and processing (16%). COG category T and M contained the highest number of proteins, followed by J. The MCP TDE0347 was decreased in abundance while TDE0169 and TDE0181 were increased in abundance. The chemotaxis protein methyltransferase (CheR; TDE0647) was also decreased in abundance. The ribosomal proteins (TDE0772, TDE0881, TDE1751) had increased in abundance. The bacteriocin ABC transporter protein (TDE0426) was decreased in abundance while the other ABC transporter proteins (TDE1067, TDE1068, TDE1311, TDE1417, TDE1516, TDE1689, TDE1850, TDE1947)

189

were increased in abundance. Two OmpA family proteins (TDE1663, TDE1992), the cell division protein (FtsA; TDE1203) and the antigenic membrane lipoprotein TmpC

(TDE1950) were also increased in abundance. Dentilisin (TDE0762) and associated PrcA protein (TDE0761) were decreased in abundance.

Ninety-one proteins were changed in abundance in ∆motB33520 only, with 30 proteins increased in abundance and 61 proteins decreased in abundance (Figure 5.25, Table 5.3). Most of these proteins were assigned to COGs related to metabolism (31%), followed by cellular processes and signaling (22%) then information storage and processing (11%). COG category T contained the highest number of proteins, followed by E, G and J. The MCP TDE0072 was decreased in abundance. The ribosomal proteins (TDE0778, TDE0885, TDE1678) were increased in abundance. The ABC transporter proteins TDE0143, TDE0398, TDE1075, TDE1076, TDE1273 and TDE2226 were decreased in abundance while TDE0183 and TDE0385 were increased in abundance.

190

Table 5.6 Proteins that were not detected in T. denticola ATCC 33520 but were detected in ∆motA33520 and ∆motB33520 (p<0.05). Proteins predicted to be organized in an operon were shaded.

B Locus Tag Protein description ∆motA33520 ∆motB33520 COG abundanceA abundanceA TDE0607 ParA family ATPase 1437833 1263620 D TDE0650 membrane protein, putative 1930400 2015500 R TDE0834 Na+-translocating NADH/quinone reductase E subunit (nqrE) 13615000 15786333 C TDE0983 oligopeptide/dipeptide ABC permease, frameshift mutation 10107800 13752333 - TDE0987 oligopeptide/dipeptide ABC transporter, ATP-binding protein 18876667 10296000 E TDE1388 conserved hypothetical protein 5229600 5336233 S TDE1588 tryptophanyl-tRNA synthetase, putative 2403000 1787400 J TDE1713 hypothetical protein 4326167 4973100 S TDE1730 glycosyl hydrolase, family 2 1515967 1384733 G TDE2285 conserved hypothetical protein 3341067 3744233 S TDE2325 conserved hypothetical protein 6432867 8986167 S TDE2568 thiamine biosynthesis protein ThiI 1973467 1764700 H TDE2643 oxidoreductase, FAD-dependent 14274333 17407000 C TDE2673 hypothetical protein 34092333 38270667 - TDE2771 conserved hypothetical protein 392363 750760 - A The abundance of each protein was calculated from the average IBAQ intensity from three replicates. B One-letter abbreviations for the functional COG categories: J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosome partitioning; V, defense mechanisms; T, signal transduction mechanisms; M, cell wall/membrane/envelope biogenesis; N, cell motility; U, intracellular trafficking, secretion, and vesicular transport; O, posttranslational modification, protein turnover, chaperones; C, energy production and conversion; G, carbohydrate transport and metabolism; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; R, general function prediction only; S, function unknown.

191

Table 5.7 Proteins significantly increased in abundance in both T. denticola ∆motA33520 and ∆motB33520 relative to wild-type (ratio≥1.5, p<0.05). Proteins predicted to be organized in an operon were shaded.

C Locus Tag Protein description Wild-type ∆motA33520 ∆motB33520 COG AbundanceA ratioB ratioB TDE0055 flagellar biosynthesis protein FlhA 2858667 1.86 1.64 U TDE0129 PyrBI protein 164263333 1.67 2.23 F TDE0626 hypothetical protein 52321333 1.99 1.65 S TDE0651 inorganic pyrophosphatase, frameshift mutation 96338667 1.74 1.82 - TDE0652 membrane protein, putative 40358667 1.60 1.99 S TDE0745 glycine reductase complex selenoprotein GrdA 27500000 4.60 4.59 C TDE0758 iron compound ABC transporter, periplasmic iron compound-binding 6827467 1.69 4.14 P protein, putative TDE0782 ribosomal protein L6 (RplF) 16268333 2.30 1.54 J TDE0783 ribosomal protein L18 (RplR) 39484667 2.62 2.55 J TDE0840 hypothetical protein 7915900 2.07 1.70 S TDE0853 ribosomal protein S9 (RpsI) 68605000 1.99 2.36 J TDE0872 RecA protein 6319867 4.93 2.12 L TDE0984 oligopeptide/dipeptide ABC transporter, permease protein, putative 6575533 5.94 8.79 R TDE0985 oligopeptide/dipeptide ABC transporter, periplasmic peptide-binding 321453333 3.19 5.20 E protein, putative TDE0986 oligopeptide/dipeptide ABC transporter, ATP-binding protein 3993533 13.84 12.92 P TDE1009 methyl-accepting chemotaxis protein 28383333 2.74 1.68 T TDE1208 DNA topoisomerase I (TopA) 69289000 1.94 1.93 L TDE1234 hypothetical protein 283517 2.95 3.81 S TDE1349 rod shape-determining protein MreB 3359300 6.64 2.78 D TDE1364 valyl-tRNA synthetase (ValS) 12635000 1.59 1.51 J

192

TDE1373 HD domain protein 1702133 1.93 1.93 S TDE1415 nucleotidyl transferase/aminotransferase, class V 32576667 1.82 1.57 E TDE1444 metallo-beta-lactamase family protein 16029000 2.79 2.59 J TDE1496 chromosome partition protein SmC, putative 5179200 2.08 2.01 D TDE1499 adenylosuccinate lyase, putative 52923667 1.66 1.66 F TDE1503 transcription termination factor Rho 32425333 1.66 1.63 K TDE1541 metallo-beta-lactamase family protein 12947000 2.15 2.25 S TDE1584 lipoprotein, putative 384266667 2.49 2.30 S TDE1628 hypothetical protein 58951667 1.95 2.91 - TDE1643 PASTA domain protein 9700633 1.82 1.50 T TDE1671 trigger factor 90896000 1.80 1.58 O TDE1687 conserved hypothetical protein 8830867 1.70 1.84 S TDE1754 desulfoferrodoxin/neelaredoxin 73828000 2.68 2.71 C TDE1882 glycosyl hydrolase, family 1 15417000 1.80 1.81 G TDE1896 phosphoribosylaminoimidazolecarboxamide formyltransferase/IMP 8421800 1.67 1.56 F cyclohydrolase (PurH) TDE1898 preprotein translocase, SecA subunit (SecA) 28523667 1.59 1.55 U TDE1927 phenylalanyl-tRNA synthetase, beta subunit (PheT) 16258000 1.84 1.71 J TDE2079 sigma-54 dependent transcriptional regulator, putative 1443567 1.68 0.00 T TDE2167 pyridine nucleotide-disulphide oxidoreductase family protein 10058800 2.24 2.27 E TDE2235 methylaspartate ammonia-lyase 123063333 1.87 2.40 E TDE2244 conserved hypothetical protein 1813500 6.52 3.38 S TDE2328 efflux pump component MtrF 23938000 1.57 1.62 H TDE2480 chaperone protein HtpG 80220667 2.40 2.13 O TDE2549 methyl-accepting chemotaxis protein 1364663 6.62 7.13 T TDE2566 conserved domain protein 6948100 1.57 1.73 S

193

TDE2573 glucose-6-phosphate isomerase 9746033 2.15 2.08 G TDE2616 cyclic nucleotide-binding protein 11704000 1.97 2.23 T TDE2682 conserved hypothetical protein 6990233 1.81 1.64 L TDE2721 helicase domain protein 3375000 2.28 2.70 L TDE2776 proline iminopeptidase (Pip) 24991000 1.67 1.78 E A The abundance of each protein in the wild-type T. denticola ATCC 33520 was calculated from the average IBAQ intensity from three replicates. B Geometric mean of ratios, from three replicates, produced from the LFQ intensity of protein in mutant relative to that of protein in wild-type. C One-letter abbreviations for the functional COG categories: J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosome partitioning; V, defense mechanisms; T, signal transduction mechanisms; M, cell wall/membrane/envelope biogenesis; N, cell motility; U, intracellular trafficking, secretion, and vesicular transport; O, posttranslational modification, protein turnover, chaperones; C, energy production and conversion; G, carbohydrate transport and metabolism; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; R, general function prediction only; S, function unknown.

194

Table 5.8 Proteins significantly decreased in abundance in both T. denticola ∆motA33520 and ∆motB33520 relative to wild-type (ratio≤0.67, p<0.05). Proteins predicted to be organized in an operon were shaded. In the case where two or more operons were arranged consecutively, different darkness of shadings was used to differentiate the operons.

D Locus Tag Protein description Wild-type ∆motA33520 ∆motB33520 COG AbundanceA ratioB,C ratioB,C TDE0008 copper-translocating P-type ATPase 1737167 0.00 0.00 P TDE0011 alkyl hydroperoxide reductase/peroxiredoxin 5732033333 0.37 0.00 O TDE0012 carbon starvation protein CstA, putative 461243333 0.27 0.01 T TDE0042 phosphate acetyltransferase (Pta) 270816667 0.42 0.55 C TDE0065 conserved hypothetical protein 8750800 0.00 0.00 S TDE0066 conserved hypothetical protein 6329767 0.28 0.30 S TDE0081 hypothetical protein 1744067 0.00 0.00 S TDE0103 aminotransferase, class-V 1789900 0.00 0.00 E TDE0119 flagellar protein FliS 1433833 0.00 0.00 U TDE0149 DNA-binding response regulator 1913133 0.00 0.00 T TDE0151 integral membrane protein, YeeE/YedE family 34126667 0.49 0.47 S TDE0175 pyrrolidone-carboxylate peptidase (Pcp) 98789667 0.38 0.40 O TDE0186 hypothetical protein 596683333 0.60 0.63 P TDE0200 tetrapyrrole methylase family protein 2478933 0.00 0.00 R TDE0207 permease, GntP family 38581667 0.33 0.38 E TDE0212 conserved hypothetical protein 122642000 0.46 0.44 P TDE0213 TPR domain protein 3607700 0.65 0.59 U TDE0239 glycine reductase complex protein GrdD 231270000 0.62 0.59 I TDE0300 cytosol aminopeptidase family protein 388300000 0.63 0.61 E TDE0301 tetracenomycin polyketide synthesis O-methyltransferase TcmP, putative 5128000 0.00 0.00 Q TDE0337 glucosamine-6-phosphate isomerase (NagB) 49480667 0.35 0.41 G

195

TDE0354 general stress protein 14 18201333 0.34 0.40 R TDE0358 cinnamoyl ester hydrolase (CinI) 2591833 0.00 0.00 I TDE0386 ABC transporter, periplasmic substrate-binding protein 585636667 0.50 0.45 P TDE0389 (R)-2-hydroxyglutaryl-CoA dehydratase, beta subunit, putative 89464667 0.54 0.48 E TDE0405 major outer sheath protein 3398466667 0.44 0.54 M TDE0418 lipoprotein, putative 30177333 0.29 0.30 - TDE0419 hypothetical protein 27165667 0.11 0.14 - TDE0423 hypothetical protein 18472000 0.25 0.27 - TDE0425 bacteriocin ABC transporter, ATP-binding/permease, putative 730670 0.00 0.00 V TDE0448 deoxyribose-phosphate aldolase (DeoC) 40219000 0.45 0.55 F TDE0451 arginine deiminase (ArcA) 92357667 0.66 0.62 E TDE0463 purine nucleoside phosphorylase (DeoD) 53346667 0.43 0.55 F TDE0501 hypothetical protein 9163200 0.00 0.00 S TDE0523 conserved hypothetical protein 4614200 0.00 0.00 S TDE0536 hypothetical protein 3203433 0.00 0.00 S TDE0557 hypothetical protein 3581133 0.00 0.00 S TDE0593 internalin-related protein 536157 0.00 0.00 S TDE0594 cysteine protease domain, YopT-type 2557167 0.00 0.00 - TDE0601 malonyl CoA-acyl carrier protein transacylase, putative 23808000 0.66 0.63 I TDE0605 hypothetical protein 72829000 0.40 0.48 - TDE0641 UDP-N-acetylglucosamine 1-carboxyvinyltransferase (MurA) 27933000 0.51 0.46 M TDE0654 peptidase, M20/M25/M40 family 167696667 0.66 0.63 R TDE0677 conserved hypothetical protein 125676667 0.38 0.41 S TDE0784 ribosomal protein S5 (RpsE) 109160000 0.61 0.65 J TDE0793 conserved hypothetical protein 22475667 0.46 0.42 S TDE0799 glycerophosphoryl diester phosphodiesterase, putative 6033100 0.54 0.49 C

196

TDE0813 ABC transporter, ATP-binding protein/permease 1060683 0.00 0.00 V TDE0870 phosphatase/nucleotidase 114953333 0.61 0.64 F TDE0877 conserved hypothetical protein 15203667 0.43 0.46 I TDE0880 nicotinate-nucleotide-dimethylbenzimidazole phosphoribosyltransferase 1499430 0.00 0.00 H (CobT) TDE0951 lipoprotein, putative 91908333 0.64 0.63 R TDE0971 arginyl-tRNA synthetase (ArgS) 16741667 0.66 0.67 J TDE1001 orotate phosphoribosyltransferase (PyrE) 27871667 0.52 0.61 F TDE1004 flagellar filament core protein (FlaB) 466846667 0.33 0.24 N TDE1016 probable ATP-dependent protease LA, frameshift mutation 3193867 0.59 0.52 - TDE1020 dicarboxylate transporter, periplasmic dicarboxylate-binding protein, 59879000 0.60 0.59 G putative TDE1021 lipoprotein, putative 14540667 0.00 0.00 S TDE1072 lipoprotein, putative 689326667 0.54 0.60 R TDE1086 conserved hypothetical protein TIGR00255 7883233 0.57 0.56 S TDE1088 conserved hypothetical protein 6955700 0.00 0.00 S TDE1089 conserved hypothetical protein 39941667 0.44 0.48 S TDE1102 hypothetical protein 89104667 0.54 0.59 S TDE1111 transporter, putative 5874533 0.00 0.00 C TDE1122 anti-anti-sigma factor 17954000 0.00 0.00 T TDE1174 hypothetical protein 20867333 0.42 0.49 S TDE1226 zinc ABC transporter, periplasmic zinc-binding protein (TroA) 12195867 0.65 0.58 P TDE1231 hypothetical protein 32748333 0.67 0.61 S TDE1236 triosephosphate isomerase (TpiA) 66671000 0.51 0.60 G TDE1271 oligopeptide/dipeptide ABC transporter, ATP-binding protein 753637 0.00 0.00 E TDE1300 conserved hypothetical protein 20554000 0.41 0.41 S

197

TDE1314 penicillin-binding protein 847027 0.00 0.00 M TDE1318 hypothetical protein 725143 0.00 0.00 U TDE1328 hypothetical protein 9560867 0.00 0.00 S TDE1333 hflC protein, putative 31090000 0.62 0.67 O TDE1352 penicillin-binding protein 663403 0.00 0.00 M TDE1370 YjeF-related protein 1344200 0.00 0.00 G TDE1399 prolipoprotein diacylglyceryl transferase (Lgt) 2294100 0.00 0.00 M TDE1408 flagellar filament outer layer protein (flaA) 576866667 0.30 0.29 N TDE1409 flagellar filament outer layer protein (flaA) 559370000 0.28 0.28 N TDE1416 ABC transporter, permease protein 2626667 0.00 0.00 M TDE1435 hypothetical protein 1232367 0.00 0.00 S TDE1464 SsrA-binding protein (SmpB) 7125867 0.00 0.00 O TDE1477 flagellar filament core protein (FlaB) 892906667 0.34 0.27 N TDE1478 conserved hypothetical protein 8751233 0.00 0.00 S TDE1480 conserved hypothetical protein 189256667 0.10 0.07 S TDE1484 hypothetical protein 6296500 0.00 0.00 - TDE1489 hypothetical protein 32149333 0.53 0.64 - TDE1493 chemotaxis protein CheX 177560000 0.34 0.41 N TDE1494 chemotaxis protein CheY 171210000 0.53 0.53 T TDE1509 lipoprotein releasing system, ATP-binding protein, putative 1456133 0.00 0.00 V TDE1511 pathogen-specific surface antigen, putative 1545400000 0.51 0.54 P TDE1519 conserved hypothetical protein 59522333 0.47 0.51 S TDE1533 conserved hypothetical protein 2624633 0.00 0.00 S TDE1548 conserved hypothetical protein TIGR00103 5305767 0.00 0.00 S TDE1639 sporulation related repeat protein 58553333 0.60 0.56 S TDE1657 N utilization substance protein B (NusB) 8365633 0.00 0.00 K

198

TDE1669 hemolysin 1258966667 0.20 0.00 E TDE1699 hypothetical protein 3118300 0.00 0.00 - TDE1712 flagellar filament outer layer protein (FlaA) 1851533333 0.20 0.18 N TDE1725 conserved domain protein 1368067 0.00 0.00 S TDE1745 conserved hypothetical protein 6718533 0.00 0.00 S TDE1748 GTP-binding protein, GTP1/Obg family 1590000 0.00 0.00 S TDE1848 hypothetical protein 18270667 0.45 0.57 S TDE1849 hypothetical protein 45571667 0.30 0.31 S TDE1858 transcriptional regulator, PadR family 12206667 0.00 0.00 K TDE1859 hypothetical protein 45146000 0.47 0.37 S TDE1862 conserved domain protein 245266667 0.58 0.52 S TDE1914 Holliday junction DNA helicase RuvB 1728917 0.00 0.00 L TDE2009 conserved hypothetical protein 4332433 0.00 0.00 S TDE2029 hydrolase, TatD family 2696533 0.00 0.00 L TDE2039 lipoprotein releasing system, permease protein, putative 2799733 0.00 0.00 M TDE2061 conserved hypothetical protein 11339367 0.00 0.00 P TDE2074 radical SAM enzyme, Cfr family 2664000 0.00 0.00 R TDE2085 amino acid kinase family protein 76823333 0.43 0.46 F TDE2089 signal recognition particle protein (Ffh) 852297 0.00 0.00 U TDE2092 amino acid ABC transporter, amino acid-binding protein, putative 1351333 0.00 0.00 T TDE2120 glycine reductase complex proprotein GrdE2 459900000 0.43 0.59 S TDE2148 OmpA family protein 7732233 0.48 0.34 M TDE2200 methionine gamma-lyase (MegL) 596470000 0.57 0.65 E TDE2210 hypothetical protein 35590000 0.53 0.51 S TDE2217 galactose/glucose-binding lipoprotein (MglB) 254480000 0.52 0.59 G TDE2257 5-nucleotidase family protein 845120000 0.55 0.60 F

199

TDE2280 transglycosylase, SLT family 5223867 0.38 0.32 M TDE2342 1-deoxy-D-xylulose 5-phosphate reductoisomerase (Dxr) 1289963 0.00 0.00 I TDE2344 undecaprenyl diphosphate synthase (UPPS) 2083200 0.00 0.00 I TDE2353 flagellar hook-associated protein 3 1834867 0.00 0.00 N TDE2356 iron compound ABC transporter, periplasmic iron compound-binding 5151000 0.00 0.00 P protein, putative TDE2357 iron compound ABC transporter, permease protein 4943200 0.00 0.00 P TDE2376 conserved hypothetical protein 3882400 0.00 0.00 S TDE2390 hypothetical protein 336300000 0.53 0.39 - TDE2410 hemolysin 163846667 0.38 0.12 E TDE2414 hypothetical protein 7559067 0.00 0.00 - TDE2451 S-adenosylmethionine:tRNA ribosyltransferase-isomerase (QueA) 948770 0.00 0.00 J TDE2466 conserved hypothetical protein 27652000 0.00 0.00 O TDE2469 hexokinase family protein 2270100 0.00 0.00 G TDE2485 conserved hypothetical protein 27640667 0.66 0.65 S TDE2501 response regulator 16410667 0.46 0.61 T TDE2515; conserved hypothetical proteinconserved hypothetical protein 2488367 0.00 0.00 S TDE0945 TDE2517 DNA repair protein RadA 3838300 0.63 0.43 O TDE2542 antigen, putative 2992200 0.00 0.00 - TDE2551 hypothetical protein 1475633 0.00 0.00 S TDE2580 GGDEF domain protein 3335033 0.42 0.29 T TDE2591 rhodanese-like domain protein 26251000 0.47 0.62 P TDE2592 DbpA RNA binding domain protein 31204000 0.55 0.58 S TDE2594 conserved hypothetical protein 3912567 0.00 0.00 S TDE2597 ribulose-phosphate 3-epimerase (RPE) 10352933 0.00 0.00 G

200

TDE2602 outer membrane protein, putative 279783333 0.56 0.58 M TDE2614 ApbE family protein 1673733 0.00 0.00 H TDE2685 flagellar synthesis regulator FleN 1086167 0.00 0.00 D TDE2761 conserved hypothetical protein 1946067 0.00 0.00 N TDE2762 flagellar motor switch protein FliY 40587000 0.42 0.23 N TDE2764 flagellar protein FliL 65052333 0.14 0.12 N TDE2765 flagellar motor rotation protein B (MotB) 34386667 0.11 0.00 N TDE2766 motility protein A (MotA) 34589333 0.00 0.42 N TDE2768 flagellar hook protein FlgE 43923333 0.47 0.33 N TDE2775 lipoprotein, putative 8312167 0.39 0.51 S TDE2779 hypothetical protein 59837667 0.11 0.10 S TDE2783 methyl-accepting chemotaxis protein 2736267 0.00 0.00 T A The abundance of each protein in the wild-type T. denticola ATCC 33520 was calculated from the average IBAQ intensity from three replicates. B Geometric mean of ratios, from three replicates, produced from the LFQ intensity of protein in mutant relative to that of protein in wild-type.

C Zero ratio indicates that the protein was identified in ATCC 33520 but not in ∆motA33520 or ∆motB33520. D One-letter abbreviations for the functional COG categories: J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosome partitioning; V, defense mechanisms; T, signal transduction mechanisms; M, cell wall/membrane/envelope biogenesis; N, cell motility; U, intracellular trafficking, secretion, and vesicular transport; O, posttranslational modification, protein turnover, chaperones; C, energy production and conversion; G, carbohydrate transport and metabolism; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; R, general function prediction only; S, function unknown.

201

5.2.6.9 Static biofilm assay of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and

∆motB33520 with P. gingivalis W50 A static biofilm assay was used as a preliminary screening method to determine the ability of T. denticola ATCC 33520 and its motility mutants to form mono- or dual-species biofilms with P. gingivalis W50. T. denticola and P. gingivalis cells were grown to exponential phase, diluted with fresh, pre-reduced OBGM to A650 of 0.15 and aliquoted into 12-well plates. For dual-species biofilms, equal volumes of T. denticola and P. gingivalis diluted cultures were mixed in the 12-well plate. The plate was incubated anaerobically for 1 h, sealed and incubated anaerobically for 5 days. Following removal of the growth media and washing, crystal violet was used to stain the biofilms in order to determine the total biofilm biomass. P. gingivalis W50 formed mono-species biofilms of significantly greater biomass than T. denticola ATCC 33520 (p<0.05) (Figure 5.26). The mono-species biofilms formed by the motility mutants ∆flgE33520, ∆motA33520, ∆motB33520 had significantly less biomass than that formed by ATCC 33520 (p<0.05) (Figure 5.26). The biomass of the dual-species biofilms formed by T. denticola ATCC 33520 with P. gingivalis W50 was approximately twice that of the sum of their mono-species biofilms (p<0.05), suggesting synergistic biofilm formation (Figure 5.26). The biomass of the dual-species biofilms of ∆motA33520 with P. gingivalis W50 was similar to that of ATCC

33520 with P. gingivalis W50. ∆motB33520 and ∆flgE33520 formed biofilms of significantly less biomass with P. gingivalis W50 than ATCC 33520 (p<0.05). Interestingly, ∆flgE33520 dual-species biofilms were smaller than P. gingivalis W50 mono-species biofilms (p<0.05).

202

Figure 5.26 Mono- and dual-species static biofilms of T. denticola ATCC 33520, ∆flgE33520, ∆motA33520 and ∆motB33520 with P. gingivalis W50. T. denticola and P. gingivalis cells were grown to exponential phase, diluted to A650 of 0.15 and grown anaerobically for 5 days in a 12-well plate, either in a monoculture or dual-species culture. The resultant biofilm was stained with crystal violet and the total biomass determined spectrophotometrically. The data are presented as means ± standard deviations (N=23- 27) and were analyzed using Kruskal-Wallis with Conover-Imam test. All values were significantly different (p<0.05) except between the following pairs: ∆motA33520 (mono) and ∆motB33520 (mono); ATCC 33520 (mono) and ∆flgE33520 (dual); as well as ATCC 33520 (dual) and ∆motA33520 (dual).

203

5.2.7 T. denticola ATCC 35405 and HL51 HL51 is a mutant with flgE gene insertionally inactivated in T. denticola ATCC 35405 (Li et al. 1996) and it was obtained from Dr Howard Kuramitsu. The HL51 mutant was phenotypically similar to ∆flgE33520. It was unable to swim outward from the point of inoculation to form a turbid plaque like its parent strain ATCC 35405 (Figure 5.27A-B). Periplasmic flagella observed in ATCC 35405 were absent in HL51 and HL51 appeared to be more rod-shaped than ATCC 35405 (Figure 5.27C). These results were consistent with the observations made of T. denticola ∆flgE33520 (Figure 5.20).

204

Figure 5.27 Comparison of T. denticola ATCC 35405 and HL51. A. Swimming assay of T. denticola ATCC 35405 and HL51. T. denticola cells (108 cfu) grown to exponential phase were harvested and spotted on semisolid OBGM agar medium. The plates were incubated for 7 days in an anaerobic chamber before the area of turbid plaque was measured. The data are presented as means ± standard deviations (N=10) and were analyzed by Students’ T-test. The area of turbidity of HL51 was significantly reduced (p<0.05) compared to T. denticola ATCC 35405 and is indicated by an asterisk (*). B. Representative image of the swimming assay of T. denticola ATCC 35405 and HL51 after 7 days of anaerobic incubation. C. Representative cryo-EM images of T. denticola ATCC 35405 and HL51. T. denticola cells were grown to exponential phase and viewed directly under cryo-EM. Red arrow points to the periplasmic flagella.

205

5.2.7.1 Dual-species biofilms of T. denticola ATCC 35405 and HL51 The ability of HL51 to form dual-species biofilms with P. gingivalis W50 was compared to its parent strain ATCC 35405 in flow cells. The flow cell assays, which better resemble the environment in the periodontal pocket where there is a constant flow of gingival crevicular fluid, were carried out in parallel with the static biofilm assays to show the role of T. denticola periplasmic flagella in the synergistic biofilm formation of T. denticola and P. gingivalis. The biomass of the dual-species biofilms formed by HL51 grown together with P. gingivalis W50 (5.87 ± 4.51 µm3/µm2) was reduced by 41% when compared to the biomass of the dual-species biofilms formed by ATCC 35405 grown together with P. gingivalis W50 (9.96 ± 3.16 µm3/µm2) (Table 5.9). There was also a significant decrease in the ratio of HL51 mutant relative to P. gingivalis W50 (3:8) in their dual-species biofilms as compared with that of ATCC 35405 relative to P. gingivalis W50 (3:2) in their dual-species biofilms (p<0.05). The dual-species biofilms formed by HL51 had a higher surface to biovolume ratio than the dual-species biofilms formed by ATCC 35405, suggesting a different biofilm structure. The dual-species biofilms formed by HL51 had a similar maximum thickness as the dual-species biofilms of ATCC 35405 (Table 5.9).

Figure 5.28 Representative CLSM images of P. gingivalis W50 and T. denticola ATCC 35405 or HL51 dual-species biofilms. The biofilms were stained with species- specific FISH probes (red, T. denticola; green, P. gingivalis) after 114 h incubation in a flow cell model. A. P. gingivalis W50 and ATCC 35405 biofilm. B. P. gingivalis W50 and HL51 biofilm.

206

Table 5.9 Overall biometric parameters of the dual-species biofilms of T. denticola ATCC 35405 and HL51 with P. gingivalis W50 harvested at 90 h.

Dual-species biofilm with P. gingivalis W50 % change in Biometric mutant relative parameters T. denticola ATCC to wild-typeA T. denticola HL51 35405

Biomass 9.96 ± 3.16 5.87 ± 4.51* -41 (µm3/µm2)

% of T. 90 27 -70 denticola

% of P. 61 79 +30 gingivalis

Ratio of T. denticola to P. 3:2 3:8 NAB gingivalis

Average thickness 12.66 ± 3.60 11.25 ± 6.31 -11 (Biomass) (µm)

Maximum 21.07 ± 3.92 17.6 ± 7.49 -16 thickness (µm)

Surface to biovolume ratio 1.59 ± 0.42 2.81 ± 1.77* +77 (µm2/µm3)

Data are expressed as means ± standard deviations of two biological replicates, from five CLSM images taken at random positions from each biological replicate. All images were analysed using COMSTAT software. *Significantly different to ATCC 35405 dual-species biofilm values, as determined by Student’s T-test (p<0.05). A Percentage change in mutant parameter relative to wild-type parameter. Negative values indicate a reduction in the mutant parameter relative to wild-type parameter whereas positive values indicate an increase in the mutant parameter relative to wild-type parameter. B NA: Not applicable.

207

5.3 DISCUSSION P. gingivalis and T. denticola are periodontopathogens that have been shown to form polymicrobial biofilms synergistically in vitro. P. gingivalis is a sessile bacterium while T. denticola is a motile bacterium with periplasmic flagella. Although there is increasing evidence suggesting motility as one of the virulence factors of T. denticola (Dashper et al. 2011), its role in biofilm formation is unclear. In this study, the importance of periplasmic flagella and/or motility of T. denticola in synergistic biofilm formation with P. gingivalis was investigated. The genes motA, motB, flgE and cheY were selectively deleted from the genome of T. denticola in order to create mutants impaired in motility. The motA, motB and flgE genes are part of the fla motility operon of T. denticola (Stamm and Bergen 1999) while cheY is part of the che chemotaxis operon of T. denticola (Greene and Stamm 1999). MotA and MotB form the stator which is part of the flagellar motor consisting of rotor and stator (Morimoto et al. 2010b, Morimoto and Minamino 2014). FlgE is the flagellar hook protein (Li et al. 1996) while CheY is a response regulator involved in the transmission of sensory signals from the chemotaxis pathway to the flagellar motors (Park et al. 2004). Since all of the mutants were successfully generated in T. denticola ATCC 33520 but not in the type strain ATCC 35405 (see Chapter 3 discussion), ATCC 33520 and its isogenic mutants were used in this study. Furthermore, as there was an unexpected genomic excision of 53 kbp in the genome of ∆cheY33520 in addition to the expected deletion of the cheY gene, the results for this mutant were omitted.

The motility mutants ∆motA33520, ∆motB33520 and ∆flgE33520 were non-motile, as shown by their inability to spread on agar plates (Figure 5.18 and Figure 5.19). This is consistent with previous studies which showed that the inactivation of these genes produced non- motile mutants in T. denticola and other bacteria (Li et al. 1996, Chi et al. 2002, Ito et al. 2005, Houry et al. 2010, Sultan et al. 2015). Although T. denticola ATCC 33520 is commonly observed to have four periplasmic flagella, two from each end of the cell which overlap at mid-cell (Izard et al. 2008), in this study up to five periplasmic flagella were also occasionally observed (Figure 5.20). As expected, the ∆flgE33520 mutant was deficient in periplasmic flagella as it lacks the flagellar hook protein, FlgE, necessary for flagellar assembly (Li et al. 1996). Periplasmic flagella were observed in ∆motA33520 and

∆motB33520 mutants, but with a lower number than the parent strain ATCC 33520 and not all cell segments of ∆motA33520 and ∆motB33520 showed visible flagella (Figure 5.20). A

208

similar result was shown in Bacillus subtilis where the deletion of both motA and motB resulted in a reduction in the number and length of flagella (Ito et al. 2005). Quantitative proteomic analysis of ∆motA33520 and ∆motB33520 whole cell lysates revealed that all three predicted flagellar filament core proteins (TDE1004, TDE1475 and TDE1477), with the exception of TDE1475 in ∆motB33520, and all three flagellar filament outer layer proteins

(TDE1408, TDE1409 and TDE1712) were downregulated in ∆motA33520 and ∆motB33520 as compared to the WT. The reduction in periplasmic flagella number in ∆motA33520 and

∆motB33520 might have resulted from the deficiency in these proteins. Furthermore, it was observed that FleN, a protein reported to be involved in the regulation of flagella number in Pseudomonas aeruginosa, was also downregulated in ∆motA33520 and ∆motB33520 mutants. A study of a T. denticola non-motile mutant lacking FliG, another protein of the flagellar motor, revealed that the mutant had a markedly decreased number of flagellar filaments and the flagellar filaments were usually shorter in length than those of the wild- type. Furthermore, there was a reduction in the FlaA and FlaB proteins in the ∆fliG mutant (Slivienski-Gebhardt et al. 2004). The similar phenotype of the ∆fliG mutant with

∆motA33520 and ∆motB33520 suggested that the flagellar motor might be involved in the normal assembly of periplasmic flagella in T. denticola.

The flagellar filaments of T. denticola consist of three outer layer sheath proteins (FlaA1- 3; TDE1408, TDE1409, TDE1712) and three core proteins (FlaB1-3; TDE1004, TDE1475, TDE1477). In this study, it was observed that the relative abundance of these flagellar filament subunits in T. denticola ATCC 33520 cells was different. For FlaA, the abundance of TDE1408 and TDE1409, which were predicted to be organized in an operon, was similar but they were 3 times less abundant than TDE1712, the major FlaA protein. For FlaB, there were 13 and 17 times more TDE1004 and TDE1477 respectively than TDE1475. These results show that under normal growth conditions, T. denticola ATCC 33520 flagella are relatively enriched with TDE1477 and TDE1712. The level of reduction in the abundance of each flagellar filament subunit in the mutants compared with wild-type was different. In ∆flgE33520, there was a greater reduction in the abundance of TDE1475 and TDE1477 as compared to TDE1004 and there was a greater reduction in the abundance of TDE1712 as compared to TDE1408 and TDE1409. As a result of this, TDE1004 and TDE1712 became the most abundant flagellar filament proteins in

∆flgE33520, albeit at a much lower level than ATCC 33520. In ∆motA33520, there was a greater reduction in the abundance of TDE1004 and TDE1477 as compared to TDE1475

209

and there was a greater reduction in the abundance of TDE1712 as compared to TDE1408 and TDE1409. A similar pattern was observed in ∆motB33520 except that TDE1475 in

∆motB33520 was not significantly reduced. Similar to the WT ATCC 33520, TDE1477 and

TDE1712 were the most abundant flagellar filament proteins in ∆motA33520 and

∆motB33520, suggesting that the stoichiometry of the flagella in ∆motA33520 and ∆motB33520 was similar to that of ATCC 33520.

The cellular morphology of ∆motA33520, ∆motB33520 and ∆flgE33520 was different to that of T. denticola ATCC 33520. A previous study showed that the periplasmic flagella of ATCC 33520 contributed to its irregular twisted morphology with both planar and helical regions on the cells. This irregular morphology was the predominant form adopted by the cells in exponential phase, the other being the regular right-handed helical form (Ruby et al. 1997). Wild-type T. denticola cells with the outer membrane removed and T. denticola mutants lacking periplasmic flagella lost their irregular morphology and adopted a helical form (Ruby et al. 1997). In this study, an irregular morphology was observed in ATCC

33520 WT cells. However, unlike the previous study, the ∆flgE33520 mutant cells lacking periplasmic flagella adopted a rod shape (Figure 5.20) instead of a helical shape (Ruby et al. 1997). This result is more similar to that observed in Borrelia burgdorferi where a ∆flaB mutant deficient in periplasmic flagella lost its flat-wave morphology and adopted a rod shape (Motaleb et al. 2000). The reason for the discrepancy in the observed morphology of ∆flgE mutant in this study with the previous study is unclear since the same strain of T. denticola harvested at the same growth phase (late exponential phase) was used and both mutants were deficient in periplasmic flagella. The only differences are the growth media and the temperature of incubation (Ruby et al. 1997) but it is uncertain how these would affect the morphology of the cells.

As compared to the parent ATCC 33520 strain, ∆motA33520 and ∆motB33520 appeared less spiral and more rod-like (Figure 5.20). This is possibly because of the reduction in the number and length of the periplasmic flagella as periplasmic flagella contribute to the morphology of T. denticola cells. Besides that, proteins such as penicillin binding proteins (PBPs) and rod shape-determining protein MreB that were reported to be involved in the morphology of bacteria including Leptospira (another spirochete) (Slamti et al. 2011), likely through their interactions with the peptidoglycan layer (Divakaruni et al. 2007), had changed in abundance in ∆motA33520 and ∆motB33520. The PBPs (TDE1314 and TDE1352) reduced in abundance while MreB (TDE1349) increased in abundance,

210

suggesting that the mutants might have switched to a different mechanism of maintaining the cell morphology. It was shown in a previous study that the MreB cytoskeleton, prokaryotic actin homologues that form spiral structures immediately underneath the cell membrane, of Bacillus subtilis is sensitive to changes in the membrane potential. The proton motive force (PMF) was shown to be required for the cellular localization of MreB and disruption of the PMF resulted in a rapid delocalization and loss of helicity of MreB (Strahl and Hamoen 2010). As it was predicted that the PMF is likely to be disrupted

(discussed later) in ∆motA33520 and ∆motB33520, it is possible that the cells increased the production of MreB in order to compensate for the delocalization of MreB. Furthermore, the altered cellular morphology could also be caused by the lack of flagellar rotation. A study on B. burgdorferi showed that flagellar rotation is required for its flat-wave morphology as it promotes the formation of an ordered flagellar ribbon structure by enhancing the interaction between the flagellar ribbon and the cell cylinder (Sultan et al. 2015). The deletion of motB in B. burgdorferi distorted the flagellar ribbon in the mutant cells and caused part of the cells to be rod shaped (Sultan et al. 2015). Together, these results suggest that both the presence of properly assembled periplasmic flagella and flagellar rotation could be important determinants of the characteristic morphology of T. denticola.

The loss of motility impaired the growth of ∆motA33520, ∆motB33520 and ∆flgE33520 in liquid cultures. The mutants had a lower maximum cell density and a lower growth rate than the parent strain ATCC 33520 (Figure 5.21). One of the reasons for this could be the lack of nutrient accessibility as a result of the loss of motility. Alternatively, motility might be important for the separation of replicating cells. Compared to the WT cells,

∆motA33520, ∆motB33520 and ∆flgE33520 cells were more elongated in shape, suggesting that the daughter cells may not have been completely separated from the mother cells (Figure 5.19). In B. burgdorferi, flagellar proteins are required for proper cell division. The cells of the non-motile mutants of B. burgdorferi ∆flaB, ∆motB, ∆flgE and ∆fliF often grow in chains and the mutants are defective in growth as well as cell division (Sal et al. 2008, Sultan et al. 2013, Sultan et al. 2015). The increase in abundance of a ParA family

ATPase (TDE0607) in both ∆motA33520 and ∆motB33520 mutants may be an indicator for the defect in chromosomal segregation and cell division in the mutants. The parA gene is encoded adjacent to a parB gene (TDE0606) in the T. denticola genome, suggesting that it belongs to the chromosome-encoded ParABS system which has been implicated in

211

chromosome segregation between bacterial daughter cells prior to cell division (Ringgaard et al. 2011). This system consists of three components: ParA ATPase, ParB DNA-binding protein, and parS sequence, a partition site on the DNA target (Hwang et al. 2013). It was reported that ATP-bound ParA interacts with the chromosome via nonspecific DNA (nsDNA) binding and ParB loads onto the DNA via parS to form the partition complex. ParA ATPase activity is stimulated by ParB and is essential for the release of ParA from the nsDNA during the partition process (Le Gall et al. 2016). Besides partition activities, ParA also acts as a transcriptional repressor by binding specifically to the par operator and regulating expression of the par genes. Both ATP- and ADP-bound ParA are able to bind to the par operator but ADP-bound ParA has ~5 to 10-fold higher binding affinity than ATP-bound ParA (Fung et al. 2001). The predicted disruption in the transmembrane ion gradient in ∆motA33520 and ∆motB33520 cells could have caused a reduction in the production and availability of intracellular ATP. The increased abundance of ParA in the mutants would mean that there were more ADP- bound forms than ATP-bound forms. As ParA binds nsDNA in an ATP-dependent manner and ATP hydrolysis is important for the partition activity of ParA (Fung et al. 2001), the increase in ParA not only resulted in repression of the par genes, but also reduction in chromosomal segregation and cell division. Another possible reason is that the generation of a PMF through the proton channel complex formed by MotA and MotB is an important source of energy for the survival of T. denticola or is required for the efflux of certain growth inhibitors from the cells. This is supported by a study which suggested that the PMF of T. denticola is involved in the efflux of human β-defensin 3 from its cytoplasm (Brissette and Lukehart 2007). Furthermore, it was observed that

∆motB33520 was more impaired in growth than ∆motA33520. This is possibly because MotA alone is capable of incorporating into the flagellar motor (Morimoto et al. 2010b) even without MotB, but the absence of MotB, especially the plug segment required for the suppression of undesirable proton flow through the proton channel of the MotA/B complex (Hosking et al. 2006, Morimoto et al. 2010a), results in proton leakage that negatively affects cell growth.

Bacterial autoaggregation and coaggregation with another bacterium are thought to play an important role in the development of a biofilm. As compared to ATCC 33520, all motility mutants had reduced ability to autoaggregate, with ∆flgE33520 having the lowest ability (Figure 5.22). This result suggests that motility and more importantly the presence

212

of periplasmic flagella are required for the autoaggregation of T. denticola. However, it should be noted that the mutants were less spiral than the wild-type and the reduced autoaggregation could be a result of the changed morphology of the mutants. As for coaggregation with P. gingivalis W50, only ∆flgE33520 displayed a reduced rate of coaggregation (Figure 5.23), suggesting that periplasmic flagella rather than motility play a major role in the coaggregation of T. denticola and P. gingivalis. This result is in agreement with a previous study which showed that the ∆flgE mutant created in ATCC 35405 had a reduced rate of coaggregation with P. gingivalis 381 as compared to its parent strain (Yamada et al. 2005). The reduced rate of coaggregation might be contributed to by changes in the surface proteins involved in the binding of T. denticola and P. gingivalis. Dentilisin and the major sheath protein are two major proteins known to be involved in the binding of T. denticola and P. gingivalis (Hashimoto et al. 2003, Rosen et al. 2008), however quantitative proteomic analyses showed that the levels of these proteins in ∆flgE33520 were comparable to those of ATCC 33520. It is possible that there are changes in other unidentified surface components that are involved in the binding of T. denticola and P. gingivalis. Alternatively, the cell morphology of T. denticola might be important for its coaggregation with P. gingivalis W50 as ∆flgE33520 was more rod-shaped than the WT.

The deletion of motA, motB and flgE in ∆motA33520, ∆motB33520 and ∆flgE33520 respectively influenced the level of other motility-related proteins. Quantitative proteomic analyses of the whole cell lysates of all the mutants showed a reduction in the level of proteins encoded from the genes derived from within the fla operon where motA, motB and flgE are located. It was reported that completion of the hook-basal body structure serves as a checkpoint for transcriptional regulation of flagellum synthesis (Hughes et al. 1993). This system involves FliA (a motility-specific transcription factor σ28), an anti-sigma factor FlgM, and σ28 promoter recognition sequences associated with the motility genes (Hughes et al. 1993). Homologs of FliA and FlgM are found in T. denticola, encoded by TDE2683 and TDE0201 respectively (Frederick et al. 2011), and a putative σ28-like promoter was found upstream of fliK, the first gene in the fla operon of T. denticola (Limberger et al. 1999, Stamm and Bergen 1999). This suggests that T. denticola may regulate its fla operon through this system and the deletions of motA, motB and flgE negatively regulated the transcription of genes upstream and downstream of the mutations in the fla operon. It was also observed that the levels of proteins whose genes are not located in the same fla

213

operon as flgE, motA and motB were altered. These proteins included a flagellar chaperone (FliS), a flagellar hook-associated protein (FlgL), all three filament core proteins (FlaB; TDE1004, TDE1475, TDE1477) and filament outer layer proteins (FlaA; TDE1408, TDE1409, TDE1712). The decreased levels of FlaA and FlaB proteins in

∆flgE33520 is in agreement with the study done on B. burgdorferi where the reduction was shown to be mediated at the post-transcriptional level (Sal et al. 2008). Moreover, the presence of MotA and MotB might also act as a checkpoint for the expression of chemotaxis genes as several chemotaxis-related proteins were differentially regulated in

∆motA33520 and ∆motB33520. Together, these results suggest that flagellum synthesis and the expression of chemotaxis and motility genes utilizes a finely regulated process where there is a series of transcriptional and/or post-transcriptional controls.

Besides motility-related proteins, the abundance of an oligopeptide/dipeptide ABC transporter made up of proteins TDE0983-TDE0987 changed in abundance in all three

∆motA33520, ∆motB33520 and ∆flgE33520 mutants. ABC transporters are multi-subunit protein pumps that utilize the free energy of ATP hydrolysis to drive the transport of substrates across the inner cell membrane. It was observed in this study that the abundance of TDE0983-TDE0987 was highly increased in ∆motA33520 and ∆motB33520 while highly decreased in ∆flgE33520. This could show that oligopeptides or dipeptides usually transported into the cell by PMF driven systems were affected in the ∆motA33520 and

∆motB33520 mutants. In response to this, the ATP-dependent ABC transporter TDE0983- TDE0987 was increased several folds as shown in Table 5.6 and Table 5.7, to compensate for the reduction in oligopeptide or dipeptide transport. This ATP-dependent system will be far more inefficient than the PMF driven system and will result in ∆motA33520 and

∆motB33520 growing slower than the WT, which is consistent with the growth studies. In the case of ∆flgE33520, the PMF saved from powering flagella rotation, due to the absence of flagella, may be directed to the transport of oligopeptides or dipeptides by PMF driven systems into the cell. Therefore, there was a lower demand for the ATP-dependent system and it resulted in the downregulation of TDE0983-TDE0987 in ∆flgE33520. This hypothesis is also supported by the growth studies which showed that ∆flgE33520 entered exponential phase faster than the WT and the other two mutants, likely due to the higher efficiency of PMF driven system than ATP-dependent system. Another protein, TDE0758, which is the periplasmic binding component of an ABC transporter involved in iron uptake into the cytoplasm also changed in abundance in all three mutants.

214

TDE0758 increased in abundance in ∆motA33520 and ∆motB33520 but decreased in abundance in ∆flgE33520. All the other putative iron compound ABC transporters

(TDE2356-TDE2357 and TDE0385-TDE0386) that changed in abundance in ∆motA33520 and ∆motB33520, except TDE0385, had reduced abundance. This may show that the mutants switched to a different iron acquisition pathway or it may reflect a reduced demand for iron.

The disruption of MotA and MotB may have caused stress to the cells of ∆motA33520 and

∆motB33520. This was observed in the increased abundance of several proteins involved in the bacterial stress response. These include desulfoferrodoxin/neelaredoxin (TDE1754), which was shown in Treponema pallidum to be an iron-binding superoxide reductase used to cope with oxidative stress (Jovanovic et al. 2000), and the RecA protein (TDE0872), which plays a role in homologous recombination, DNA repair and induction of the SOS response. DNA topoisomerase I (TopA; TDE1208), which was shown to be important for efficient transcriptional activation of the recA promoter during the E. coli SOS response to antibiotics and stress challenge, had also increased in abundance (Liu et al. 2011). The transcription termination factor Rho (TDE1503) which uses its RNA–DNA helicase function to release nascent mRNA from the RNA polymerase transcription complex in an ATP-dependent manner (Brennan et al. 1987) and was shown to be required for oxidative stress survival in Caulobacter crescentus (Italiani et al. 2002), also increased in abundance in ∆motA33520 and ∆motB33520. The observed reduction in the abundance of the penicillin-binding proteins (PBPs) TDE1314 and TDE1352, which are essential for the synthesis of the peptidoglycan layer of the cell (Sauvage et al. 2008), suggests that

∆motA33520 and ∆motB33520 mutants might experience peptidoglycan stress. The increased abundance of the metallo-beta-lactamase family proteins TDE1444 and TDE1541 is likely to be a response to the peptidoglycan stress as the cells may perceive the reduction in peptidoglycan synthesis as a result of PBPs inhibition and thus produce more metallo- beta-lactamases, the enzymes that catalyze the hydrolysis of beta-lactam antibiotics that inhibit PBPs (Palzkill 2013). Furthermore, two aminoacyl-tRNA synthetases, valyl-tRNA synthetase (TDE1364) and phenylalanyl-tRNA synthetase (TDE1927), involved in the editing of misacylated tRNAs (Ling et al. 2009) and proposed to be critical for bacterial stress responses and survival (Bullwinkle and Ibba 2016) were increased in abundance in

∆motA33520 and ∆motB33520. Overall, these results suggest that MotA and MotB are

215

important components of the cells as their absence activates a cellular stress response in the mutants.

The Na+-translocating NADH/quinone reductase E subunit (NqrE; TDE0834) increased + in abundance in both ∆motA33520 and ∆motB33520 mutants. NqrE is part of the Na - translocating NADH/quinone oxidoreductase (Na+-NQR) respiratory complex found in prokaryotes (Barquera 2014). Na+-NQR pumps Na+ across the membrane, generating an electrochemical Na+ gradient which energizes important functions in the cell, including rotation of the Na+-dependent flagella for motility. NqrE has six transmembrane helices which is able to form the multi-helix bundles that allow ion transport through the cell membrane (Barquera 2014). An FAD-dependent oxidoreductase (TDE2643) which contains a 2Fe-2S (one of the cofactors in Na+-NQR) binding domain had also increased in abundance in ∆motA33520 and ∆motB33520. It was reported that an FAD-dependent oxidoreductase containing-domain 1 protein (FOXRED1), is required for the assembly of the mitochondrial respiratory Complex I (NADH/ubiquinone oxidoreductase) (Fassone et al. 2010), a functional homologue of Na+-NQR. Therefore, TDE2643 could play a similar role in the assembly of Na+-NQR. The increase in abundance of TDE0834 and

TDE2643 in ∆motA33520 and ∆motB33520 may suggest that the mutants have switched to the usage of Na+ as a coupling ion for flagellar motor rotation instead of H+ as a result of PMF disruption. The membrane-embedded stators, Mot complexes, harness energy of either transmembrane H+ or Na+ ion gradients to power flagellar rotation. There are two distinct types of Mot stators with different ion specificities. In B. subtilis, MotA and MotB constitute the H+-coupled Mot while MotP and MotS constitute a Na+-coupled Mot (Ito et al. 2005, Terahara et al. 2008). In Vibrio, the force-generating unit of the Na+-driven flagellar motor is composed of four components: PomA, PomB, MotX, and MotY (Li et al. 2011). The use of Na+ ion gradients is often associated with elevated pH and sodium concentrations (Mulkidjanian et al. 2008). The switching to the use of Na+ instead of or in addition to H+ for flagellar motor rotation may be an adaptation of T. denticola to the increase in pH and salinity in a periodontal pocket during chronic periodontitis. It was reported that the progression of periodontitis is associated with a rise in pH of the gingival sulcus (Zilm et al. 2010). Maintaining high levels of PMF would become increasingly difficult for T. denticola as the external concentration of H+ ions become lower. In addition, the concentration of Na+ ions in the periodontal pocket may increase as a result of bleeding and tissue inflammation. Switching to the use of Na+ ions may be thus more

216

advantageous for T. denticola in this relatively alkaline and sodium-rich environment. This hypothesis is supported by a study which shows that the MotAB of Bacillus clausii, an alkaliphilic bacterium, is bifunctional with respect to ion-coupling capacity in that it couples motility to sodium at high pH (pH>8.5) but uses protons at lower pH (pH<8.5) (Terahara et al. 2008). B. clausii MotAB increases its use of sodium as the pH becomes increasingly alkaline (Terahara et al. 2008) and presumably the PMF is getting smaller. + The observed increase in the component of Na -NQR in ∆motA33520 and ∆motB33520 with the predicted disruption in PMF may thus suggest that the ion-coupling pattern of T. denticola MotAB changes as the relative magnitudes of the transmembrane sodium and proton motive forces change. However, further experiments are needed to demonstrate this.

T. denticola ATCC 33520 was able to form a detectable biofilm under static conditions although it was reported previously that T. denticola ATCC 35405 was unable to form a biofilm on inert surfaces such as glass, polystyrene and PVC unless the plate was pre- coated with P. gingivalis 381 or the host protein fibronectin (Vesey and Kuramitsu 2004). The discrepancy was not due to strain difference as the mono-species biofilms formed by ATCC 35405 were comparable to that of ATCC 33520 (data not shown). P. gingivalis W50 which is generally regarded as a poor biofilm former (Davey 2006, Biyikoglu et al. 2012), was also able to form a biofilm under static conditions and its biofilm forming ability was even higher than that of T. denticola (Figure 5.26). This may be due to the plates used in this study being specially treated to improve cell adhesion and protein binding (Greiner CELLSTAR® 12-well, flat-well plates). Compared to the WT,

∆motA33520, ∆motB33520 and ∆flgE33520 had reduced abilities to form a mono-species biofilm under static conditions (Figure 5.26), suggesting the importance of T. denticola motility in mono-species biofilm formation under static conditions. Together with the results which showed that these mutants were impaired in autoaggregation, this may indicate that motility is required for the bacteria to bind together to form a biofilm. Alternatively, as mentioned above, the spiral shape of T. denticola might be important for its autoaggregation and thus its ability to form a biofilm.

This study showed that T. denticola ATCC 33520 and P. gingivalis W50 formed dual- species biofilms synergistically under static conditions. Unexpectedly, the abilities of

∆motA33520 and ∆motB33520 to form dual-species biofilms with P. gingivalis were different although both of them are non-motile and displayed similar phenotypes. MotB was shown

217

to be more important than MotA in the dual-species biofilm formation of T. denticola with P. gingivalis as ∆motB33520 had reduced ability in forming a dual-species biofilm with P. gingivalis W50 while ∆motA33520 did not. The abilities of the mutants to form dual-species biofilms with P. gingivalis were not dependent on their growth as ∆motA33520 had a lower growth rate and grew to a lower maximum cell density than ATCC 33520 yet had a similar ability to ATCC 33520 in forming a dual-species biofilm with P. gingivalis.

Because of the disparity in the results of ∆motA33520 and ∆motB33520, it is uncertain whether motility per se is required for the synergistic biofilm formation between T. denticola and P. gingivalis. However, T. denticola periplasmic flagella were shown to be important for the synergistic biofilm formation under both static and flow conditions. In agreement with the previous studies (Vesey and Kuramitsu 2004, Yamada et al. 2005), the absence of periplasmic flagella in ∆flgE33520 attenuated its biofilm forming ability with P. gingivalis W50 in static biofilm assays. Interestingly, the dual-species biofilms of

∆flgE33520 grown together with P. gingivalis W50 were smaller than the mono-species biofilms of P. gingivalis W50, suggesting an inhibitory effect from the absence of periplasmic flagella to dual-species biofilm formation under static conditions. The flow cell biofilm assays performed on T. denticola ATCC 35405 and HL51, which displayed similar phenotypes as ∆flgE33520, confirmed the importance of T. denticola periplasmic flagella in its synergistic biofilm formation with P. gingivalis. HL51 formed poorer biofilms with P. gingivalis W50 than ATCC 35405 and it was attenuated in its ability to incorporate into the dual-species biofilms with P. gingivalis W50 in flow cells (Table 5.9). There was a smaller reduction in the biomass of the dual-species biofilms of the

∆flgE mutant grown together with P. gingivalis W50 under flow conditions (41%; Table 5.9) compared to static conditions (79%; Figure 5.26), probably because the flow conditions compensated for the lack of motility of the mutant cells by both helping with the spreading of cells from the biofilms and also providing constant access to nutrients.

218

5.4 CONCLUSION In conclusion, this study had generated three non-motile mutants of T. denticola lacking flgE, motA and motB. The characterization of these mutants showed interesting and unexpected phenotypes. The quantitative proteomics analyses on the whole cell lysates of the mutants generated a comprehensive dataset and provided valuable insights into the effects of inactivating motility genes in T. denticola proteomes. Whilst the inactivation of flgE had its effect mainly on flagella and motility, inactivation of motA and motB appeared to have far reaching effects which are beyond motility on the cell, likely due to a collapse in the PMF as a result of disruption of the MotAB proton channel complex. This study also showed that T. denticola periplasmic flagella and/or spiral morphology are essential for its synergistic biofilm formation with P. gingivalis, particularly in its incorporation into the dual-species biofilms.

219

CHAPTER 6 DEFINING THE ROLE OF T. DENTICOLA DFSB HOMOLOGUE IN T. DENTICOLA AND IN THE INTERACTIONS BETWEEN T. DENTICOLA AND P. GINGIVALIS

220

6.1 INTRODUCTION Bacteria in nature do not live in isolation and they are often found to co-exist with other bacteria. There are multiple types of interactions between bacterial species that can benefit or disadvantage the interacting species or strains. These interactions are dynamic and interchangeable depending on the environmental conditions. When the resources become limited, the bacteria will compete with one another to ensure their own survival. Bacteria use a variety of antimicrobial compounds directed against other bacteria. These include the classic antibiotics which normally have a broad killing spectrum and also other bacteriocins that have a narrower spectrum of activity that kill only closely related species or strains (Ghoul and Mitri 2016). Although it has long been recognized that competition occurs between distantly related or unrelated bacteria, recent studies have shown that sibling/sister cells, i.e. cells that are genetically identical, also kill and eliminate one another under specific conditions. Two well-characterized sibling killing phenomena, cannibalism and fratricide, have been observed in the Gram-positive bacteria, B. subtilis and S. pneumoniae respectively (Gonzalez-Pastor et al. 2003, Guiral et al. 2005, Ellermeier et al. 2006). In B. subtilis, during starvation, a fraction of the cells in the colony which has entered the pathway to sporulate secrete extracellular killing factors to lyse their non-sporulating, non-immune sibling cells and feed on the released nutrients to delay sporulation (Gonzalez-Pastor et al. 2003, Ellermeier et al. 2006). Fratricide occurs when competent cells of S. pneumoniae trigger the lysis of noncompetent cells and benefit from the release of DNA to increase genetic diversity or from the release of virulence factors to increase virulence during infection (Guiral et al. 2005). For both cannibalism and fratricide, competition occurs between members of single colonies.

A third sibling killing phenomenon which occurs between two neighboring sibling colonies was identified in Paenibacillus dendritiformis, a Gram-positive bacterium that forms complex branching patterns on agar. Neighboring colonies of P. dendritiformis grew away from each other on an agar plate, leaving a zone of clearing between them containing dead cells (Be'er et al. 2009). In P. dendritiformis, a 12 kDa protein extracted from the agar between two competing colonies was reported to be responsible for the killing effect and the protein was named Slf (sibling lethal factor). Slf is derived from a larger precursor protein named DfsB (dendritiformis sibling bacteriocin) that was not active by itself and required cleavage by subtilisin to produce Slf, the active peptide

221

(Be’er et al. 2010). The determination of the crystal structure of Slf revealed that it is a helix-rich polypeptide with a C-terminal helix rich in basic and aromatic residues that may function as a cationic antimicrobial peptide (Taylor et al. 2016). Slf is not produced under normal growth conditions but is produced when the local concentration of subtilisin exceeds the threshold for regulated colony growth, as in the case of approaching colonies (Be’er et al. 2010). The concentration of Slf is highest in the region between two competing colonies and decreases with increasing distance from the zone of inhibition. At high concentration, Slf kills the bacteria at the leading edges of the colonies while at lower concentration, Slf induces the transition of motile P. dendritiformis rods to small vegetative cocci that lack flagella. Unlike rods, cocci are resistant to killing by Slf and unlike the dormant spores of P. dendritiformis, they are able to replicate (Be'er et al. 2011). These cocci are useful in the monitoring of the environment and quick transition to rods when the environment favors rod formation (Be'er et al. 2011).

T. denticola is a motile bacterium that forms diffuse colonies on agar. In this study, it was observed that zones of clearing formed between T. denticola colonies grown on agar. A search of the literature revealed the similar phenomenon in P. dendritiformis (Be'er et al. 2009). Bioinformatic analyses were carried out and identified the presence of a DfsB homologue in T. denticola. Due to the high similarity of the T. denticola homologue with DfsB, a mutant lacking the putative DfsB homologue was generated, using the protocol optimized in Chapter 3, and characterized. Meanwhile, an attempt to isolate the DfsB homologue and/or identify other potential inhibiting materials from the agar region of inhibition between adjacent colonies was carried out. Although the attempts were complicated by the presence of rabbit serum albumin in the agar medium and a large number of T. denticola outer sheath vesicle-associated proteins, several potential targets were identified. Since T. denticola is often co-isolated with P. gingivalis from subgingival plaque and is often found to interact with P. gingivalis synergistically and symbiotically, the role of T. denticola DfsB homologue in the interactions of T. denticola with P. gingivalis was investigated.

222

6.2 RESULTS 6.2.1 Competing T. denticola ATCC 35405 sibling colonies An interesting phenomenon was observed on OBGM agar plates when T. denticola ATCC 35405 was grown anaerobically for more than two weeks, where a zone of clearing formed between colonies (Figure 6.1). This phenomenon is unusual as the two colonies originated from the same culture and are expected to have the same offence and defense mechanisms.

Figure 6.1 Competing T. denticola ATCC 35405 colonies. A. T. denticola cells were inoculated into OBGM agar [0.8% (w/v) UltraPure™ Low Melting Point Agarose] and poured into petri dishes. The agar plates were then incubated anaerobically for two weeks. Red circles indicate colonies with zones of clearing. B. T. denticola cells at exponential growth phase were harvested, concentrated and spotted in semisolid OBGM agar medium [0.4% (w/v) agarose and 1% (w/v) gelatin]. The plates were incubated anaerobically until zones of clearing were observed between the colonies.

6.2.2 Bioinformatic analyses P. dendritiformis Slf is the cleaved, active polypeptide form of a larger protein, DfsB. DfsB belongs to the DUF1706 family (Taylor et al. 2016). A search for DUF1706 and T. denticola resulted in the identification of TDE0659, herein designated DfsBTd35405. The

DUF1706 covers 96% of the entire length of DfsBTd35405, from residues 3 to 165 (Figure

6.2). DfsBTd35405 (170 aa) has a similar length to DfsB (173 aa). DfsB is 20 kDa in size with a predicted pI of 6.92 while DfsBTd35405 is 20 kDa with a predicted pI of 8.90 (Bjellqvist et al. 1993, Bjellqvist et al. 1994, Wilkins et al. 1999). DfsB is cleaved by subtilisin to produce Slf, which is involved in the sibling killing mechanism of P. dendritiformis (Be’er et al. 2010). DfsB is cleaved at a proline residue at position 72 that is preceded by a positively-charged lysine residue. A similar cleavage site was found in

DfsBTd35405 where the proline residue at position 72 is preceded by a lysine residue (Table

223

6.1). The putative cleaved peptide of DfsBTd35405 (SlfTd35405) is 99 aa in length, similar to

Slf which is 102 aa in length. SlfTd35405 has a predicted molecular weight of 11.5 kDa, similar to the apparent molecular weight of Slf (12 kDa) and has a basic pI of 9.29, similar to Slf, which has a predicted pI of 10 (Bjellqvist et al. 1993, Bjellqvist et al. 1994, Wilkins et al. 1999). No signal peptide was predicted in the sequences of both DfsB and

DfsBTd35405 using SignalP 4.1 Server (Petersen et al. 2011) and the subcellular localization of these proteins predicted using PSORTb v3.0 program (Yu et al. 2010) was unknown, suggesting that these two proteins were either not secreted or secreted in a Sec- independent pathway. Subtilisin, on the other hand, was predicted to contain a signal peptide and localize extracellularly (Yu et al. 2010, Petersen et al. 2011).

Figure 6.2 DUF1706 domain found in DfsBTd35405. A graphics view of the hypothetical protein DfsBTd35405 in T. denticola ATCC 35405. The DUF1706 domain spans 96% of the entire length of DfsBTd35405, from residues 3 to 165.

Table 6.1 The amino acid sequences of DfsB and DfsBTd35405.

a Protein Length (aa) Sequence DfsB 173 MASYEYTSKEELKKTIHAAYLLLDGEFEGIDDSQKDN RVPEVDRTPAEIIAYQLGWLHLVMGWDRDELAGKPVI MPAPGYKWNQLGGLYQSFYAAYADLSLTELRRLFRDT ERQWLDWIDTLSEEDLFTQSVRKWTGDKPNWPMARWI HINSAAPFKTFRAKIRKWKKHQRQA DfsBTd35405 170 MPRPTTKEDLIFASNEQFEKLWNLIDGMTEKERTADI NPNERDKNVRDVLVHLYEWHCLLLNWIKSNTKGKPAA FLPEAYNWKTYPEMNVEFWKKHQGTPYADAEKMLKKT HKEVMKLIETFSNEELFSKAVFNWTGTTTLGSYCVSA TSSHYDWAIKDIKKALKKYRGK a Bold letters indicate the sequences of Slf and the predicted SlfTd35405.

224

To determine the structural similarity of DfsBTd35405 to DfsB, the DfsBTd35405 protein sequence was submitted to protein homology detection and prediction servers including Phyre2, Swiss-Model and HHpred. The DfsB crystal structure was the fourth highest match in Phyre2 with 100% confidence and 19% identity (Table 6.2); the top three matches had up to 25% identity with DfsB and were all uncharacterized proteins. The DfsB crystal structure was the closest match in Swiss-Model (Table 6.3) and again, the fourth highest match in HHpred with a 99.9% probability to be a true positive and a low

E-value of 1.7E-23 (Table 6.4). The predicted C-terminal helix of DfsBTd35405 is rich in aromatic (Y, W and H) and basic residues (K and R), similar to that of DfsB. All of these results suggested that DfsBTd35405 is a likely candidate for a DfsB homologue in T. denticola.

DfsBTd35405 homologues were identified in many strains of T. denticola and other bacterial species in the Spirochaete phylum (Table 6.5 and Table 6.6). Multiple sequence alignment of DfsBTd35405 homologues from different T. denticola strains showed that the sequence of DfsBTd35405 is highly conserved (Figure 6.4). The cleavage site in DfsB and

DfsBTd35405 is also found in most of the homologues. In some homologues, the proline residue at position 72 is replaced with a leucine residue (Figure 6.4).

225

Table 6.2 Top seven hits of DfsBTd35405 structural homologues predicted by Phyre2 using the default settings.

Template Alignment coverage Confidence % Template information identity

c4n6cB 96% (residues 1-165) 100.0 21 crystal structure of the b1rzq2 protein from Streptococcus pneumoniae

c5cofA 97% (residues 1-166) 100.0 25 crystal structure of uncharacterized protein q1r1x2 from E. coli UTI89

c5cqvB 95% (residues 3-165) 100.0 19 crystal structure of uncharacterized protein q8dwv2 from Streptococcus agalactiae

c5civA 95% (residues 3-166) 100.0 19 sibling lethal factor precursor DfsB

c5cogB 96% (residues 1-167) 100.0 21 crystal structure of yeast IRC4

d1rxqa 93% (residues 7-166) 99.8 19 YfiT-like putative metal-dependent hydrolases

c2rd9C 98% (residues 3-170) 99.8 14 crystal structure of a putative Yfit-like metal-dependent hydrolase (BH0186) from Bacillus halodurans

226

Table 6.3 Top seven hits of DfsBTd35405 structural homologues predicted by Swiss-Model using the default settings.

Template Sequence identity Alignment coverage Oligo state Template information (%)

5civ.1.A 20.00 97% (residues 1-168) monomer Sibling bacteriocin (DfsB)

5cof.1.A 25.15 96% (residues 2-169) monomer Uncharacterized protein

5cog.1.A 21.82 97% (residues 1-168) monomer IRC4

5com.1.A 21.47 96% (residues 2-167) monomer Putative conjugative transposon protein Tn1549-like, CTn5-Orf2

4n6c.1.A 20.86 96% (residues 2-167) monomer Uncharacterized protein

5cqv.1.A 18.40 96% (residues 2-167) homo-dimer Uncharacterized protein

5cqv.1.B 18.40 96% (residues 2-167) homo-dimer Uncharacterized protein

227

Table 6.4 Top seven hits of DfsBTd35405 structural homologues predicted by HHpred using the default settings.

Template Probability of template E-value Sequence Template information to be a true positive identities (%)

5COF_A 99.9 2.4E-25 24 Uncharacterized protein

4N6C_B 99.9 6.6E-24 21 Uncharacterized protein

4N6C_A 99.9 6.6E-24 21 Uncharacterized protein

5CIV_A 99.9 1.7E-23 19 Sibling bacteriocin (DfsB)

5COG_B 99.9 2E-23 20 IRC4

3CEX_A 99.8 1E-21 14 Uncharacterized protein

1RXQ_B 99.8 5E-21 18 YfiT

228

Figure 6.3 Alignment of DfsBTd35405 with DfsB using Phyre2. The sequence of DfsBTd35405 was submitted to Phyre2 and aligned with the sequence of DfsB. Query sequence refers to that of DfsBTd35405 and template sequence refers to that of DfsB. The predicted C-terminal helix of DfsBTd35405 is boxed.

Table 6.5 DfsBTd35405 homologues in different T. denticola strains identified by blastp.

T. denticola strain Protein accession number

ATCC 35405 TDE0659

ATCC 35404 HMPREF9721_02064 ATCC 33521 HMPREF9735_02301 OTK HMPREF9723_01793

MYR-T HMPREF9727_02499

H1-T HMPREF9725_00144

ATCC 33520 HMPREF9722_RS12910

ASLM HMPREF9729_01532

US-Trep HMPREF9728_00715

H-22 HMPREF9726_02071

SP33 HMPREF9733_01037

SP32 HMPREF9732_00861

F0402 HMPREF9353_02127

229

SP37 HMPREF9724_01363

SP44 HMPREF9734_00430

SP23 HMPREF9731_02225

Table 6.6 DfsBTd35405 homologues in different bacterial species in the Spirochaete phylum identified by blastp.

Genus Species TDE0659 homologuesA Treponema denticola 17 putidum 2 maltophilum 1 vincentii 4 medium 2 pedis 4 OMZ 838 1 phagedenis 3 bryantii 3 C6A8 1 JC4 1 azotonutricium 1 socranskii 1 Sediminispirochaeta smaragdinae 1 Leptospira kirschneri 1 Brachyspira CAG:700 1 alvinipulli 1 A One DfsBTd35405 homologue was identified per strain of bacterial species.

230

Figure 6.4 Multiple sequence alignment of DfsBTd35405 homologues from different T. denticola strains. The protein sequences of DfsBTd35405 homologues were aligned using the ClustalW2 multiple sequence alignment tool. The arrow points to the putative cleavage site in the protein.

6.2.3 Generation of T. denticola allele exchange mutants lacking putative sibling

lethal factor DfsBTd35405 6.2.3.1 Construction of the pHN-0659 suicide plasmid Plasmid pHN-0659 was generated as a T. denticola suicide plasmid where pGEM®-T Easy contained a recombination cassette consisting of ermAM flanked by DNA corresponding to the genome upstream (5’ flanking region) and downstream (3’ flanking region) of TDE0659 (DfsBTd35405) in T. denticola ATCC 35405. The plasmid was generated by SOE PCR using the corresponding primers as listed in Table 2.5. The 5’ and 3’ flanking regions of TDE0659 (5’ TDE0659 and 3’ TDE0659) were amplified from T. denticola ATCC 35405 WT gDNA whilst ermAM was amplified from the shuttle vector pHS17. The PCR amplicons with the expected size of 745 bp, 744 bp and 1226 bp for 5’ TDE0659, 3’ TDE0659 and ermAM respectively were purified (Figure 6.5A-B). The purified 5’ TDE0659 was fused with ermAM by SOE PCR. The PCR amplicon with the desired size of 1921 bp was purified (Figure 6.5C) and fused by SOE PCR with 3’ TDE0659 to yield a final product (TDE0659::ermAM) of size 2635 bp (Figure 6.5D). The

231

purified TDE0659::ermAM was then ligated to pGEM®-T Easy and used to transform E. coli α-Select cells. Fifteen transformant colonies selected on LB Amp plates were analyzed by colony PCR using the M13 Forward/M13 Reverse primer pair. All transformant clones except 9 and 10 possessed the desired 2.9 kbp amplicon (Figure 6.6). Plasmid from clone four was sequenced to confirm the fidelity of the recombination cassette in pHN-0659. E. coli α-Select harboring pHN-0659 was designated ECR770.

Figure 6.5 Generation of plasmid pHN-0659. A. PCR amplicons of 5’ TDE0659 (Lane 1; 745 bp) and ermAM (Lane 2; 1226 bp). B. PCR amplicon of 3’ TDE0659 (744 bp). C. PCR amplicon of 5’ TDE0659-ermAM (1921 bp). D. PCR amplicon of the final SOE product (TDE0659::ermAM; 2635 bp). E. NotI-digestion of pHN-0659 showing the expected products of 2.6 kbp for insert and 3 kbp for vector. Lane 1, undigested pHN- 0659; Lane 2, digested pHN-0659. All PCR amplicons and the restriction enzyme digestion were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gels.

Figure 6.6 Agarose gel electrophoresis [0.8% (w/v) in TAE] of colony PCR amplicons from E. coli colonies harboring pHN-0659. Lanes 1-15, PCR amplicons generated from E. coli colonies 1-15 using primers M13 Forward and M13 Reverse. The expected size of product is 2.9 kbp. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

232

6.2.3.2 Generation of allele exchange mutants lacking DfsBTd35405 in T. denticola ATCC 35405 and ∆TDE0911::kan Plasmid pHN-0659 was linearized by digestion with NotI to release the recombination cassette of 2.7 kbp from the background vector of 3 kbp (Figure 6.5E). After digestion, DNA precipitation was carried out to remove the enzyme. The digested products were resuspended in deionized water and used to transform T. denticola ATCC 35405 or ∆TDE0911::kan by electroporation. The generation of ∆DfsB in T. denticola ATCC 35405 and ∆TDE0911::kan were unsuccessful.

6.2.3.3 Generation of T. denticola ATCC 33520 allele exchange mutant ∆DfsBTd33520

HMPREF9722_RS12910, herein designated DfsBTd33520, is the homologue of DfsBTd35405 in T. denticola ATCC 33520. It has 98% sequence similarity with DfsBTd35405 when aligned using the EMBOSS needle protein pairwise sequence alignment tool (Figure 6.7) and when used in Phyre2, the fourth match is the DfsB crystal structure with 100% confidence and 19% identity, the same as the result for DfsBTd35405 (Table 6.2).

DfsBTd33520 is 170 aa in length with a predicted molecular weight of 20 kDa and a predicted pI of 8.87. The putative cleavage site in DfsBTd35405 was also found in

DfsBTd33520 (Figure 6.7). The putative cleaved product is 99 aa in length with a predicted molecular weight of 11.7 kDa and a predicted pI of 9.25. The organization of the genes upstream and downstream of DfsBTd33520 is similar to that of DfsBTd35405 (Table 6.7). Of the T. denticola ATCC 33520 transformed colonies selected on OBGM agar plates containing 40 µg/mL erythromycin, two were grown in OBGM containing 40 µg/mL erythromycin and a small volume of culture was used for PCR analysis to confirm DNA integration at the correct locus. Primers 5’TDE0659-F, 3’TDE0659-R, ermAM-fla prom- R and ermAM-2767-F were used. Both transformants possessed the desired amplicons of size 3.1 kbp, 2.1 kbp and 2.1 kbp for the primer pairs 5’TDE0659-F/3’TDE0659-R, 5’TDE0659-F/ermAM-fla prom-R and ermAM-2767-F/3’TDE0659-R respectively (Table 2.4; Figure 6.8). Transformant 1 was designated ECR831.

233

Figure 6.7 Pairwise sequence alignment of DfsBTd33520 and DfsBTd35405. The protein sequences of DfsBTd33520 and DfsBTd35405 were aligned using EMBOSS needle protein pairwise sequence alignment tool. There is a 96% identity and 98% similarity between the two protein sequences. Arrow indicates the putative cleavage site.

Table 6.7 Genes upstream and downstream of DfsBTd35405 in ATCC 35405 and DfsBTd33520 in ATCC 33520.

T. denticola T. denticola ATCC Description ATCC 35405 33520

DfsB DfsBTd35405 DfsBTd33520 ClbS/DfsB family four- homologue helix bundle protein

Upstream TDE0660 HMPREF9722_RS12905 putative transcriptional gene regulator

Downstream TDE0658 HMPREF9722_RS12915 ABC transporter, gene permease protein

234

Figure 6.8 PCR to confirm homologous recombination of TDE0659::ermAM with T. denticola ATCC 33520 genome. Lane 1, 5 and 9, PCR amplicons from T. denticola ATCC 33520 control; Lane 2, 6 and 10, NTC; Lane 3, 7 and 11, PCR amplicons from ∆DfsBTd33520 transformant 1; Lane 4, 8 and 12, PCR amplicons from ∆DfsBTd33520 transformant 2. The primer pairs used were indicated at the top of the gel photo. The expected amplicon size from 5’TDE0659-F/3’TDE0659-R was 2.4 kbp from WT and 3.1 kbp from ∆DfsBTd33520. The expected amplicon size from 5’TDE0659-F/ermAM-fla prom-R was 2.1 kbp from ∆DfsBTd33520 and no product from WT. The expected amplicon size from ermAM-2767-F/3’TDE0659-R was 2.1 kbp from ∆DfsBTd33520 and no product from WT. All PCR amplicons were examined via agarose gel electrophoresis [0.8% (w/v) in TAE]. The position of HyperLadder™ 1kb molecular weight markers quantified in kbp are shown beside the gel.

6.2.4 Characterization of ∆DfsBTd33520 6.2.4.1 Growth curves

The growth curves of T. denticola ATCC 33520 and its isogenic mutant ∆DfsBTd33520 (ECR831) were plotted by monitoring batch culture growth (N=3) at 6-24 h intervals for a total period of 264 h. The maximum A650 achieved by the mutant was similar to that of ATCC 33520 (p>0.05), suggesting that they have a similar maximum cell density. The lag phase of ATCC 33520 and ∆DfsBTd33520 was similar in length. The mean generation times determined in the exponential growth phase were 14 ± 3 h for ATCC 33520 and 12

± 4 h for ∆DfsBTd33520. The similar generation times (p>0.05) suggest that the ATCC

33520 and ∆DfsBTd33520 have similar growth rates.

235

Figure 6.9 Growth curves of T. denticola ATCC 33520 and ∆DfsBTd33520. A650 of cultures was measured at 6-24 h intervals. The data are presented as means ± standard deviations (N=3). Student’s T-test showed that the maximum A650 achieved by ∆DfsBTd33520 was similar to that of ATCC 33520 (p>0.05). The mean generation time was calculated based on the rate of change of A650 during the exponential growth phase of cultures (N=3).

6.2.4.2 Inhibition assay

An inhibition assay was carried out to determine if DfsBTd33520 is involved in the sibling killing phenomenon in T. denticola. T. denticola cells were grown to exponential phase, concentrated and spotted in the semisolid OBGM agar medium. The plates were incubated anaerobically until zones of clearing were observed between the colonies. Zones of clearing were observed between the colonies of both wild-type ATCC 33520 and ∆DfsBTd33520 mutant (Figure 6.10). Furthermore, the growth of both ATCC 33520 and

∆DfsBTd33520 within the regions between the colonies of ATCC 33520 and ∆DfsBTd33520 was inhibited but their growth outside of the regions was not inhibited (Figure 6.11).

These results suggest that DfsBTd33520 is not involved in the sibling killing phenomenon in T. denticola ATCC 33520.

236

Figure 6.10 Inhibition assay of (A.) T. denticola ATCC 33520 and (B.) ∆DfsBTd33520. The cells at exponential growth phase were harvested, concentrated and spotted in semisolid OBGM agar medium [0.4% (w/v) agarose and 1% (w/v) gelatin]. The plates were incubated anaerobically until zones of clearing were observed between the colonies.

Figure 6.11 The growth of T. denticola ATCC 33520 and ∆DfsBTd33520 within and outside of the center of four evenly-spaced T. denticola ATCC 33520 and ∆DfsBTd33520 colonies. T. denticola cells at exponential phase were harvested and spotted in semisolid OBGM agar medium [0.4% (w/v) agarose and 1% (w/v) gelatin]. The plate was incubated anaerobically for 9 days until zones of clearing were observed between the colonies. T. denticola cells at exponential phase were harvested and spotted within and without the center of the four evenly-spaced T. denticola colonies. The plate was further incubated for 7 days.

237

6.2.5 Isolation and identification of inhibiting material Another approach was taken to identify other potential inhibiting materials from the agar region of inhibition between adjacent T. denticola colonies. Inhibition assay was carried out using T. denticola ATCC 35405 cells (108 cfu) harvested at exponential phase of growth. Sections of agar were excised from the plates after zones of clearing were observed between the T. denticola colonies and the extracted proteins were analyzed by SDS-PAGE. Lane A was derived from an agar sample taken from the center of four evenly-spaced colonies, lane B was derived from an agar sample taken outside the colonies, lane C was taken from an agar sample at the diffused edge of two colonies and lane D was taken from an agar sample at the diffuse edge of a colony as shown in Figure 6.12A. The agar samples of lanes C and D contained cells while that of lanes A and B did not. The sample in lane B was taken as a control to show the basal level and composition of proteins in the medium, so that protein bands that are found in other lanes but not lane B would indicate proteins from T. denticola. The protein profiles of lanes A and C were compared to that of lanes B and D so that proteins that were found only in lanes A and C had a high probability of involvement in the sibling killing phenomenon. The gel banding profiles of Lanes B-D were very similar except that the band at ~9 kDa in Lane C was not found in Lane B and it had a higher intensity than that in Lane D. The band above the ~9 kDa band, which is ~12 kDa, in Lane C had a lower intensity as compared to that in Lanes B and D. The sample in Lane C contained a relatively small amount of protein than that of Lanes B and D. The gel banding profile of Lane A is different from that of Lanes B-D and the overall protein intensity in Lane A is higher than that in Lanes B-D. Several bands in Lane A, including a large band between 49 and 62 kDa, were of interest as they were either absent or have a higher intensity than their respective bands in Lanes B-D. These gel bands of interest were excised from stained SDS-PAGE gels (Figure 6.12), subjected to in-gel trypsin digestion and analyzed by MALDI-TOF MS. Although there were a few T. denticola proteins identified, such as TDE2022 in band 2, TDE1399 in band 3, TDE1369 in band 6, TDE1463 in band 9, TDE1150 in band 10 and TDE2279 in band 12, the bands matched up predominantly to rabbit serum albumin, a component in the OBGM agar medium (Table 6.8).

238

Figure 6.12 Analysis of proteins extracted from different sections of OBGM agar with competing T. denticola colonies. A. Exponential phase T. denticola cells (108 cfu) were spotted 1 cm apart in semisolid OBGM agar medium [0.4% (w/v) agarose and 1% (w/v) gelatin] and incubated until zones of clearing were observed between the colonies. B. Proteins extracted from different sections of agar (A-D) were analyzed via SDS-PAGE using NuPAGE® 4-12% Bis-Tris gel. M indicates the SeeBlue® Plus2 pre-stained molecular mass standard. Red boxes indicate bands of interest that were excised for MALDI-TOF MS analysis.

239

Table 6.8 MALDI-TOF MS identification of proteins in gel bands 1-14.

Band T. denticola protein Protein identity Protein number accession number scoreA 1 NAB Rabbit serum albumin 56 2 TDE2022 YD repeat protein 49 NAB Rabbit serum albumin 69 3 TDE1399 Prolipoprotein diacylglyceryl 50 transferase (lgt) NAB Rabbit serum albumin 48 4 NAB Rabbit serum albumin 56 5 NAB Rabbit serum albumin 81 6 TDE1369 excinuclease ABC, B subunit 49 NAB Rabbit serum albumin 76 7 NAB Rabbit serum albumin 67 8 NAB Rabbit serum albumin 74 9 TDE1463 hypothetical protein 48 10 TDE1150 hypothetical protein 48 11 NAB Rabbit serum albumin 55 12 TDE2279 HD domain protein 47 NAB Rabbit serum albumin 54 13 NAB Rabbit serum albumin 44 14 unidentified NAB NAB A Protein scores greater than 47 are significant (p<0.05). B NA: Not applicable.

The presence of serum albumin complicated the analysis as the gel banding patterns observed might be due to serum proteins that have been partially digested by T. denticola proteinases. Therefore, the experiment was repeated with filtered serum where the rabbit serum was filtered through a 10 kDa membrane to remove the serum albumin before it was added into the OBGM agar medium. The colonies in the OBGM agar medium with filtered rabbit serum formed smaller and denser turbid plaques than the colonies in the normal medium (compare Figure 6.12 and Figure 6.13), possibly due to a lower viscosity of the medium. The proteins extracted from different sections of agar were analyzed by

240

SDS-PAGE (Figure 6.13). Lane A was derived from an agar sample taken from the center of four evenly-spaced colonies, lane A2 was derived from the remaining agar from the center of the colonies after sample A was taken, lane B was derived from an agar sample taken near the edge of the plate and far from the colonies, lane C was derived from an agar sample taken at the diffused edge of two colonies, lane D was derived from an agar sample taken near the diffused edge of a colony and lane E was derived from an agar sample taken at the diffuse edge of a colony (Figure 6.13A). The agar samples of lanes C and E contained cells while that of the other lanes did not. The sample in lane B serves as a control to show the basal level and composition of proteins in the medium. The protein profiles of lanes A, A2 and C were compared to that of lanes B, D and E so that proteins that were found only in lanes A, A2 and C would have a high probability of involvement in the sibling killing phenomenon. The overall band profile and intensity observed in this gel was different from that obtained previously for unfiltered rabbit serum and the prominent band slightly below 62 kDa observed in the previous gel was absent in this gel. These could be due to the removal of the rabbit serum albumin from the agar medium. The gel banding profile of lane B was very different from that obtained previously, possibly due to protein overloading, as the profile was more smeared and the only individual band observed was at 62 kDa. The gel banding profiles of lanes C-E were very similar, although with different intensities. There were only a few bands observed in lane A and all of them were found in lanes C-E. Lane A2 contained a relatively higher amount of protein than other lanes. All of the bands found in lane A2 were also found in lanes C-E, but with lower intensities. The gel bands of interest were excised (Figure 6.13) and subjected to MALDI-TOF MS analysis. Band 1 was identified as dentilisin (TDE0762), band 2 was identified as protease complex-associated polypeptide (TDE0761), band 3 was identified as TDE1693, band 4 was identified as TDE2673, band 5 was identified as TDE2308 and band 7 was identified as Snf2 family protein (TDE0006). TDE1693, TDE2673 and TDE2308 are hypothetical proteins. Bands 6 and 8 could not be identified (Table 6.9).

241

Figure 6.13 Analysis of proteins extracted from different sections of OBGM agar (containing filtered serum) with competing T. denticola colonies. A. Exponential phase T. denticola cells (108 cfu) were spotted 1 cm apart in semisolid OBGM agar medium [0.4% (w/v) agarose and 1% (w/v) gelatin] and incubated until zones of clearing were observed between the colonies. B. Proteins extracted from different sections of agar (A-E) were analyzed via SDS-PAGE using NuPAGE® 4-12% Bis-Tris gel. A2 is the remaining section of agar within the zone of clearing after A was excised. M indicates the SeeBlue® Plus2 pre-stained molecular mass standard. Red boxes indicate bands of interest that were excised for MALDI-TOF MS analysis.

242

Table 6.9 MALDI-TOF MS identification of proteins in gel bands 1-8.

Band T. denticola protein Protein identity Protein number accession number scoreA

1 TDE0762 dentilisin 99

2 TDE0761 protease complex-associated 90 polypeptide, prcA

3 TDE1693 hypothetical protein 103

4 TDE2673 hypothetical protein 48

5 TDE2308 hypothetical protein 51

6 unidentified NAB NAB

7 TDE0006 Snf2 family protein 48

8 unidentified NAB NAB

A Protein scores greater than 47 are significant (p<0.05). B NA: Not applicable.

In order to compare the relative level of proteins present in different sections of agar, it was necessary to analyze the proteins from the whole sample instead of separating the extracted proteins on an SDS-PAGE gel and excising bands of interest. This was done by running the proteins for a short time on SDS-PAGE gel and excising the whole section of stained gel, which contained all extracted protein, for in-gel trypsin digestion and Orbitrap MS analysis. Lane A was derived from an agar sample taken from the center of four evenly-spaced colonies, lane B was derived from an agar sample taken near the edge of the plate and far from the colonies, lane C was derived from an agar sample taken at the diffused edge of two colonies, lane D was derived from an agar sample taken near the diffused edge of a colony and lane E was derived from an agar sample taken at the diffuse edge of a colony (Figure 6.14A). The agar samples of lanes C and E contained cells while that of the other lanes did not. The sample in lane B serves as a control to show the basal level and composition of proteins in the medium. In order to identify potential candidates involved in the sibling killing phenomenon, the intensity of the proteins in lane A was compared to that in lane D and the intensity of the proteins in lane C was compared to that in lane E. Proteins that have both A/D and C/E ratios higher than 1.5 and that were not found in lane B were potential candidates. Several proteins were found to meet these

243

criteria. These included cheB (TDE0648), an ABC transporter ATP-binding protein (TDE1592), a lipoprotein (TDE1072) and a hypothetical protein (TDE0889) (Table 6.10). T. denticola proteins identified in Table 6.8 were not found in this analysis. All T. denticola proteins identified in Table 6.9, except TDE0006, were found in this analysis to have an A/D ratio of more than 1, but not all had a C/E ratio of more than 1 (Table 6.11). All of these proteins were identified in lane B.

Figure 6.14 Analysis of whole proteins extracted from different sections of OBGM agar (no rabbit serum) with competing T. denticola colonies. A. Exponential phase T. denticola cells (108 cfu) were spotted 1 cm apart in semisolid OBGM agar medium [0.4% (w/v) agarose and 1% (w/v) gelatin] and incubated until zones of clearing were observed between the colonies. B. Proteins extracted from different sections of agar (A-E) were subjected to SDS-PAGE on a NuPAGE® 10% Bis-Tris gel for 13 min. Red boxes indicate sections of gel that were excised for Orbitrap MS analysis.

244

Table 6.10 Proteins absent in lane B that had high A/D and C/E ratios.

Protein accession A/D C/E Protein identity number ratioA ratioA protein-glutamate methylesterase, TDE0648 4.6 1.8 cheB TDE0889 hypothetical protein 2.7 2.2 ABC transporter ATP-binding TDE1592 2.0 2.7 protein TDE1072 putative lipoprotein 1.7 1.5 A The ratios were calculated based on the LFQ intensities of the proteins in each lane. Only proteins with both ratios higher than 1.5 were listed.

Table 6.11 A/D and C/E ratios of the proteins identified in Table 6.9.

Protein accession numberA A/D ratioB C/E ratioB TDE0762 1.9 0.8 TDE0761 1.7 0.7 TDE1693 1.3 0.7 TDE2673 1.2 1.2 TDE2308 2.0 1.1 A All of the proteins were found in lane B. B The ratios were calculated based on the LFQ intensities of the proteins in each lane.

6.2.6 The role of DfsBTd33520 in the interaction between T. denticola and P. gingivalis 6.2.6.1 Inhibition assay with P. gingivalis W50 As T. denticola and P. gingivalis are commonly detected together in the subgingival plaque associated with chronic periodontitis and they often display a synergistic and symbiotic relationship (Zhu et al. 2013, Tan et al. 2014), it is of interest to examine the role of T. denticola DfsB homologue in their interactions. An inhibition assay on agar plates was carried out using T. denticola ATCC 33520, ∆DfsBTd33520 and P. gingivalis W50. Since DfsB needs to be cleaved in order to be active, P. gingivalis W50 triple mutant ∆ABK lacking all three RgpA, RgpB and Kgp gingipains (Dashper et al. 2004) was included in order to examine the role of gingipains in the processing and activation of DfsBTd33520. T. denticola cells were spotted in semisolid OBGM agar medium and

245

incubated anaerobically for 9 days until zones of clearing were observed. P. gingivalis cells at exponential phase were harvested, concentrated and spotted within and outside of the zone of clearing between T. denticola colonies. The plate was further incubated anaerobically for seven days until P. gingivalis colonies were observed. The growth of P. gingivalis W50 and ∆ABK were inhibited within the zone of inhibition but not outside of the zone of inhibition. When the plate was further incubated for seven days, T. denticola colonies facing the P. gingivalis W50 colony outside of the zone of inhibition appeared to grow better than those not facing the P. gingivalis W50 colony. This observation was less pronounced for T. denticola colonies facing P. gingivalis ∆ABK colony. After prolonged incubation (24 days after the P. gingivalis cells were spotted in the plate), the growth of T. denticola colonies towards the P. gingivalis W50 colony was inhibited but the growth of T. denticola colonies towards the P. gingivalis ∆ABK colony was not inhibited. The same observations were obtained for both T. denticola ATCC 33520 and

∆DfsBTd33520 (Figure 6.15).

246

Figure 6.15 The growth of P. gingivalis W50 and ∆ABK within and outside of the zone of clearing between T. denticola ATCC 33520 and ∆DfsBTd33520 colonies. T. denticola cells at exponential phase were harvested and spotted 1 cm apart in semisolid OBGM agar medium [0.4% (w/v) agarose and 1% (w/v) gelatin]. The plate was incubated anaerobically for 9 days. P. gingivalis cells at exponential phase were harvested and spotted within and outside of the zone of clearing between T. denticola colonies. The plate was further incubated for 7, 14 and 24 days as indicated.

247

To check whether T. denticola cells were inhibited by P. gingivalis, P. gingivalis cells were spotted in semisolid OBGM agar medium and incubated anaerobically for 4 days until the colonies were observed. T. denticola cells were then spotted inside and outside of the region between the P. gingivalis colonies. The plate was further incubated anaerobically for 8 days. The growth of both T. denticola ATCC 33520 and ∆DfsBTd33520 were inhibited between the colonies of P. gingivalis W50 but not between the colonies of

P. gingivalis ∆ABK. The growth of both T. denticola ATCC 33520 and ∆DfsBTd33520 spotted out of the region between P. gingivalis colonies were not inhibited.

Figure 6.16 The growth of T. denticola ATCC 33520 and ∆DfsBTd33520 within and without the regions between P. gingivalis W50 and ∆ABK colonies. P. gingivalis cells at exponential phase were harvested and spotted in semisolid OBGM agar medium. The plate was incubated anaerobically for 4 days. T. denticola cells at exponential phase were harvested and spotted within and out of the regions between P. gingivalis colonies. The plate was further incubated for 8 days.

248

6.2.6.2 Autoaggregation and coaggregation assays Autoaggregation and coaggregation assays were done as described in Section 5.2.6.6.

∆DfsBTd33520 autoaggregated at the same rate as T. denticola ATCC 33520 and all bacterial aggregates sank to the bottom of the cuvette within 2 h of incubation while P. gingivalis W50 did not autoaggregate as the height of its suspension remained constant throughout the 7 h of incubation (Figure 6.17). ∆DfsBTd33520 coaggregated with P. gingivalis W50 at the same rate as ATCC 33520 with P. gingivalis W50. All bacterial aggregates sank to the bottom of the cuvette within 2 h of incubation.

Figure 6.17 Autoaggregation (A.) and coaggregation (B.) of P. gingivalis W50, T. denticola ATCC 33520 and ∆DfsBTd33520. T. denticola and P. gingivalis cells at exponential growth phase were harvested, washed twice with coaggregation buffer and adjusted to A650 of 0.5. For coaggregation assay, equal volumes of T. denticola and P. gingivalis cell suspensions were combined. The height of bacterial aggregates was monitored for 7 h. The data are presented as means ± standard deviations (N=3). Standard deviation bars of P. gingivalis W50 autoaggregation are not visible because they are very small.

249

6.2.6.3 Static biofilm assay Static biofilm assay was carried out as described in Section 5.2.6.9. The monospecies biofilms formed by ATCC 33520 and ∆DfsBTd33520 were similar in biomass, showing that

∆DfsBTd33520 has a similar ability as the wild-type ATCC 33520 to form a monospecies biofilm. The biomass of the dual-species biofilms of ATCC 33520 and ∆DfsBTd33520 grown together with P. gingivalis W50 were significantly higher than the biomass of the respective T. denticola monospecies biofilms (p<0.05). The biomass of the dual-species biofilms of ATCC 33520 grown together with P. gingivalis W50 were significantly greater than the sum of the biomasses of their monospecies biofilms (p<0.05), suggesting synergistic biofilm formation. The biomass of the dual-species biofilms of ∆DfsBTd33520 was reduced by 53% as compared to that of ATCC 33520, demonstrating that the deletion of DfsBTd33520 impaired synergistic biofilm formation between T. denticola ATCC 33520 and P. gingivalis W50.

250

Figure 6.18 Static biofilm assay of T. denticola ATCC 33520 and ∆DfsBTd33520. T. denticola and P. gingivalis cells were grown individually to exponential phase, diluted to A650 of 0.15 and grown anaerobically for 5 days in a 12-well plate, either in a monoculture or a polymicrobial culture. The resultant biofilm was stained with crystal violet and the total biomass determined spectrophotometrically. The data are presented as means ± standard deviations (N=9). Student’s T-test showed that the dual-species biofilms formed by each T. denticola strain grown together with P. gingivalis W50 were larger than their individual monospecies biofilms (p<0.05). The dual-species biofilms of T. denticola ATCC 33520 grown together with P. gingivalis W50 were significantly larger than those of ∆DfsBTd33520 grown together with P. gingivalis W50 (p<0.05). There was no significant difference in the biomass of the monospecies biofilms of T. denticola ATCC 33520 and ∆DfsBTd33520. 6.2.6.4 Real-time PCR to determine the number of T. denticola and P. gingivalis cells in the dual-species biofilms The numbers of T. denticola and P. gingivalis W50 cells in the static biofilms were determined in order to investigate whether the decrease in the dual-species biofilms formed by ∆DfsBTd33520 and P. gingivalis W50 was due to a decrease in one species or both. Dual-species biofilms composed of T. denticola ATCC 33520 or ∆DfsBTd33520 and P. gingivalis W50 were cultured in OBGM in 12-well flat-well plates for 5 days. The dual-species biofilms were removed from the substratum after being washed and resuspended in OBGM before the DNA was extracted using the DNeasy PowerBiofilm

251

Kit (Qiagen). The bacterial cell numbers from each biofilm were determined by real-time PCR (Table 6.12 and Figure 6.19). There was a 34% reduction in the total bacterial cell numbers in the dual-species biofilms of ∆DfsBTd33520 as compared to that of ATCC 33520 (p<0.05). There was a 99% reduction in the number of P. gingivalis W50 cells in the dual- species biofilms formed with ∆DfsBTd33520 than that formed with ATCC 33520 (p<0.05). There was a 136% increase in the number of T. denticola cells in the dual-species biofilms of ∆DfsBTd33520 as compared to that of ATCC 33520 (p<0.05). There were more P. gingivalis W50 cells than T. denticola ATCC 33520 cells (ratio of T. denticola ATCC 33520 to P. gingivalis W50 cells is 0.38 ± 0.17) in their dual-species biofilms while there were more ∆DfsBTd33520 than P. gingivalis W50 (ratio of T. denticola ∆DfsBTd33520 to P. gingivalis W50 cells is 194.22 ± 84.46) in their dual-species biofilms (p<0.05).

Table 6.12 Bacterial cell numbers in the dual-species biofilms formed by P. gingivalis W50 grown with T. denticola ATCC 33520 or ∆DfsBTd33520 harvested after 5 days of anaerobic incubation.

Dual-species biofilm with P. gingivalis W50 % change in mutant relative ATCC 33520 ∆DfsBTd33520 to wild-typeA Total bacterial cell 1.4 × 108 ± 5.0 × 107 9.3 × 107 ± 2.0 × 107* -34 numbers P. gingivalis W50 1.0 × 108 ± 3.3 × 107 5.4 × 105 ± 2.5 × 105* -99 cell numbers T. denticola cell 3.9 × 107 ± 2.3 × 107 9.2 × 107 ± 2.0 × 107* +136 numbers Ratio of T. denticola to P. 0.38 ± 0.17 194.22 ± 84.46* NAB gingivalis W50 cells The data are presented as means ± standard deviations (N=8). *Significantly different to ATCC 33520 dual-species biofilm values, as determined by Student’s T-test (p<0.05). A Percentage change in mutant parameter relative to wild-type parameter. Negative values indicate a reduction in the mutant parameter relative to wild-type parameter whereas positive values indicate an increase in the mutant parameter relative to wild-type parameter. B NA: Not applicable.

252

Figure 6.19 Bacterial cell numbers in the dual-species biofilms of ATCC 33520 and ∆DfsBTd33520 grown with P. gingivalis W50. Dual-species biofilms composed of T. denticola ATCC 33520 or ∆DfsBTd33520 and P. gingivalis W50 were cultured in OBGM in 12-well flat-well plates for 5 days. The dual-species biofilms were removed from the substratum after being washed and resuspended in OBGM. DNA extracted from the biofilms was used to determine the bacterial cell numbers by real-time PCR. The data are presented as means ± standard deviations (N=8). Students’ T-test showed that the numbers of bacterial cells for each species were significantly different (p<0.05) between the dual-species biofilms of ATCC 35405 grown with P. gingivalis W50 and that of ∆DfsBTd33520 grown with P. gingivalis W50.

253

6.3 DISCUSSION Bacteria in nature are often surrounded by different strains and species with which they compete for resources necessary for their survival. These bacteria have adapted different strategies to outcompete and displace their competitors. There are two types of competitive phenotypes, one is passive competition through resource exploitation and the other one is active competition through elimination of the competitors. The classical example of active competition is the secretion of antimicrobial compounds which are normally targeted at distantly related bacteria that are not immune to the antimicrobial compounds. However, it was reported that genetically identical sibling bacteria also kill and eliminate one another under certain circumstances (Gonzalez-Pastor et al. 2003, Guiral et al. 2005, Ellermeier et al. 2006). The current study was initiated by the observation of zones of clearing between T. denticola sibling colonies on agar and the similarity of this observation with the sibling killing phenomenon reported in P. dendritiformis (Be'er et al. 2009).

Bioinformatic analyses showed that T. denticola TDE0659 (DfsBTd35405) is very similar to P. dendritiformis DfsB in sequence, length, molecular weight and structure. DfsBTd35405 contains sequence similar to the DfsB cleavage site and the potentially cleaved product

(SlfTd35405) is highly similar to Slf in length, molecular weight and pI. It was predicted that the C-terminal helix of Slf, which is rich in aromatic and basic residues, has a functional mechanism akin to that for cationic antimicrobial peptides (Taylor et al. 2016).

Similarly, the predicted C-terminal helix of SlfTd35405 is rich in aromatic and basic residues (Figure 6.3), suggesting that it may function in a similar way as Slf. In P. dendritiformis, subtilisin is responsible for the cleavage of DfsB to Slf and interestingly, dentilisin of T. denticola is a serine protease of the subtilisin family (Correia et al. 2003). This suggested that T. denticola could use dentilisin for the cleavage of DfsBTd35405 to SlfTd35405. If this is true, the inactivation of dentilisin would abolish the production of SlfTd35405 and the zones of clearing between T. denticola colonies. However, zones of clearing were observed between the colonies of a dentilisin-deficient mutant (data not shown), indicating that either dentilisin is not responsible for the cleavage of DfsBTd35405 or there are other proteins involved in the cleavage of DfsBTd35405 to SlfTd35405. The latter explanation is supported by the identification of a putative oligopeptidase B, OpdB (TDE1195) in T. denticola which may function as a lysine-specific protease (Veith et al. 2009, Dashper et al. 2011). Since DfsBTd35405 is predicted to be cleaved after a lysine residue (K71) (Table

254

6.1), it is possible that TDE1195 is involved in the processing and activation of

DfsBTd35405.

The high similarity between DfsBTd35405 and DfsB has led to the attempt to construct a T. denticola mutant lacking the putative DfsB homologue. Although the deletion of

DfsBTd35405 was unsuccessful in T. denticola ATCC 35405, a T. denticola ATCC 33520 mutant lacking the HMPREF9722_RS12910 (DfsBTd33520) gene was successfully generated. The similar growth rate and maximum cell density of ∆DfsBTd33520 with the

WT ATCC 33520 suggested that DfsBTd33520 is not essential for normal cell growth in liquid cultures. An inhibition assay showed that the sibling colonies of ∆DfsBTd33520 formed zones of clearing similar to that of ATCC 33520 and the growth of both T. denticola wild-type and ∆DfsBTd33520 was inhibited when grown within the center of four evenly-spaced ∆DfsBTd33520 colonies (Figure 6.10 and Figure 6.11), indicating that

DfsBTd33520 is not a sibling killing factor. This is further supported by the fact that

DfsBTd35405 was not found in any of the agar sections extracted from between or around the colonies of T. denticola (data not shown).

Unlike the study of P. dendritiformis, the isolation and identification of the potential inhibiting material produced by competing T. denticola colonies was much more complicated. In P. dendritiformis, only three proteins (flagellin, subtilisin and Slf) were isolated from the agar between the competing colonies and by testing the effect of each protein on the growth of a colony, Slf was identified as the lethal factor (Be’er et al. 2010). The first trial to isolate the inhibiting material produced by competing T. denticola colonies was complicated by the presence of rabbit serum albumin which is one of the components in the OBGM growth medium used to culture T. denticola. The presence of rabbit serum albumin impeded the detection of T. denticola proteins because the protein profiles observed may be due to the cleavage of serum albumin and not the presence of specific T. denticola proteins (Figure 6.12). Nevertheless, a protein of interest, TDE2022, was identified from the agar sample taken from the center of four evenly-spaced T. denticola colonies (Table 6.8). TDE2022 belongs to the YD repeat proteins which are widely distributed in bacteria and eukaryotes. The YD repeat proteins have been implicated in contact-dependent growth inhibition, bacterial interactions with eukaryotic host cells, coordination of multicellular behavior and intercellular communications (Koskiniemi et al. 2013). Rhs (rearrangement hotspot) proteins are among the most widely studied YD repeat proteins. Enterobacterial Rhs proteins are composed of four

255

distinct domains: an N-terminal domain that is conserved within, but not between, subfamilies of rhs, a core domain (776-888 aa) flanked by conserved motifs, a hyperconserved domain (61 aa) ending in PxxxxDPxGL and a variable C-terminal domain (21-168 aa) (Jackson et al. 2009). Another YD repeat protein, the insecticidal toxin-complex C protein (TccC) shares a similar architecture with Rhs (Waterfield et al. 2001). In both of these proteins, the variable C-terminal domains were shown to carry toxin domains that are involved in inhibition of neighboring bacterial cell growth (Rhs) and killing of insect host cells (TccC) (Waterfield et al. 2001, Koskiniemi et al. 2013). TDE2022 is one of the four YD repeat proteins produced by T. denticola ATCC 35405, the other three being TDE1558, TDE1560 and TDE2020. All of these proteins share a similar architecture with Rhs and TccC proteins but the hydrophobic “L” in the PxxxxDPxGL peptide motif, which demarcates the C-terminal domain, is replaced by polar residues “R” (TDE1560 and TDE2022), “E” (TDE2020) or “K” (TDE1558). When submitted to Phyre2, the sequence of TDE2022 matched with a high confidence (100%) and sequence coverage (75%, residues 170-1299) to the crystal structure of TcdB2-TccC3 proteins of Photorhabdus luminescens (pdb 4O9X; data not shown). Together, these results suggest that TDE2022 may be a potential inhibiting factor that plays a role in the T. denticola sibling killing phenomenon. However, the Rhs systems also encode sequence-diverse immunity proteins (RhsI) that are encoded from the same operon as the rhs genes and specifically neutralize cognate Rhs toxins to protect the cells from autoinhibition (Koskiniemi et al. 2013). Similarly, TDE2022 is associated with a small ORF which could possibly encode for an immunity protein. If all of the cells are able to produce the immunity protein, it will be then unclear how TDE2022 specifically mediates the killing of T. denticola cells between colonies and not within a single colony. It might be possible that the protein plays a broader role in the communication between cells, thereby providing information about population density and coordinating multicellular behavior. This hypothesis is supported by a study which shows that the Rhs protein in Myxococcus xanthus is involved in its social motility, the gliding over solid surfaces as groups of cells (Youderian and Hartzell 2007).

The problem of sample contamination with rabbit serum proteins was resolved by filtering the serum albumin through a 10 kDa filter before use or removing the rabbit serum from the OBGM agar medium. Although the growth of T. denticola in OBGM with filtered serum and without serum was not affected (data not shown), the bacterium

256

appeared to form a smaller and denser turbid plaque in agar. This is probably due to the reduced motility of T. denticola in the OBGM agar medium that has a lower viscosity as a result of filtering or removing the rabbit serum (Ruby and Charon 1998). The observation of a high level of proteins present in the agar between T. denticola colonies (Lane A in Figure 6.12) indicated that the growth constraint was unlikely to be a result of nutrient depletion, although it could be a depletion in inorganic or nonproteinaceous nutrients. A relatively high amount of proteins were observed in the SDS-PAGE lane derived from agar samples taken from the center of colonies (lane A) than the control lanes (lanes B-D; Figure 6.12), probably because the amount of agar excised was unequal. In the latter experiment (Figure 6.14), an equal amount of agar was excised for each lane in order to better compare the level of proteins in different agar regions. A high number of proteins were observed in the region between the colonies and in the controls even after the removal of serum albumin by filtration. This could be explained by the release of outer sheath vesicles, which are packed with proteins, from T. denticola into the agar medium as all of the identified proteins except TDE0006 were enriched in T. denticola outer sheath vesicles (P. Veith, personal communication). As all of these protein bands were found in the controls (lanes C-E), they are unlikely to be the inhibiting material produced specifically by competing T. denticola colonies.

Due to the large number of proteins present in all agar sections, it was necessary to quantitatively compare the proteins extracted from the different agar sections (Figure 6.14). Several proteins, TDE0648, TDE0889, TDE1592 and TDE1072, were found to have a higher level at the regions between competing colonies than at the regions outside of the colonies (Table 6.10). All of these proteins except TDE1592 were found to be enriched in T. denticola outer sheath vesicles (P. Veith, personal communication). TDE1072 is one of the major membrane proteins and it is an ABC transporter substrate- binding protein that is predicted to be lipidated (Abiko et al. 2014a). TDE1592 is an ABC transporter ATP-binding protein and TDE0648 is the chemotaxis response regulator cheB. It was reported previously that CheB overproduction inhibits the chemotactic swarming ability of E. coli (Stewart 1993). A high level of CheB in the region between the competing colonies could suggest that the chemotactic motility of T. denticola cells in this region was impaired. TDE0889 is a hypothetical protein that is 214 aa in length. It is small and positively charged with predicted molecular weight and pI of 24 kDa and 9 respectively. These characteristics make TDE0889 a likely potential candidate for sibling

257

killing in T. denticola as Slf of P. dendritiformis is also a small and positively charged protein. However, there was no structural similarity between TDE0889 and Slf and a Phyre2 search using TDE0889 sequence did not identify a similarity with any known protein crystal structures.

An interesting phenomenon was observed after prolonged incubation of T. denticola colonies with a P. gingivalis colony. T. denticola colonies facing a P. gingivalis W50 colony spotted outside of the zone of clearing appeared to grow better than the colonies without a neighboring P. gingivalis colony, but their growth was later inhibited when they approached the P. gingivalis W50 colony (Figure 6.15). A similar observation was reported in P. dendritiformis where there is faster expansion at the facing fronts of two competing colonies at the early stages of growth, likely because of the added low level of subtilisin from the neighboring colony (Be’er et al. 2015). It is possible that P. gingivalis produces a chemoattractant, such as an amino acid, peptide, lipid or signaling molecule, for T. denticola that draws T. denticola cells towards its colony. As they grow, both T. denticola and P. gingivalis colonies secrete proteases into the agar to break down proteins for nutrient sources. The synergy between T. denticola and P. gingivalis proteases promotes the growth of T. denticola colonies facing the P. gingivalis colony. However, when T. denticola cells approach the P. gingivalis W50 colony, P. gingivalis gingipains, released into the surrounding agar in soluble or OMV-bound forms, limit the growth of T. denticola either directly or indirectly through the release of certain factor(s). The growth of T. denticola colonies facing P. gingivalis ∆ABK colony spotted outside of the zone of clearing was promoted but was not inhibited when they approached the P. gingivalis ∆ABK colony (Figure 6.15), probably because P. gingivalis ∆ABK still produces the chemoattractant and proteases that promote T. denticola growth but is unable to limit T. denticola growth due to the lack of gingipains. Together with the results which showed that the growth of T. denticola was inhibited between P. gingivalis W50 colonies but not between P. gingivalis ∆ABK colonies (Figure 6.16), these results show that P. gingivalis gingipains play a major role in the inhibition of T. denticola growth by P. gingivalis colonies on plate.

The results obtained above are interesting, as gingipains were shown to be important for the synergistic biofilm formation between T. denticola and P. gingivalis. Unlike P. gingivalis W50, P. gingivalis ∆ABK mutant was unable to form polymicrobial biofilms with T. denticola and T. forsythia synergistically in flow cells (Zhu et al. 2013). The

258

close association between T. denticola and P. gingivalis W50 observed in the polymicrobial biofilms suggests a non-competitive interaction between the two bacteria. However, the complex social interactions that occur between organisms in biofilms can be either competitive or cooperative and competitive interactions may be as important as cooperative interactions in maintaining a balanced relationship between microbes in biofilms (Li and Tian 2012). The inhibition of T. denticola growth by P. gingivalis colonies on an agar plate could reflect a gingipain-mediated competitive interaction between P. gingivalis and T. denticola in dual-species biofilms. P. gingivalis W50 which was shown to poorly attach to the glass substratum of a flow cell initially became the dominant species in the mature polymicrobial biofilms (Zainal-Abidin et al. 2012). However, T. denticola which dominated the immature polymicrobial biofilms comprised only less than 10% of the total cells in the mature polymicrobial biofilms (Zainal-Abidin et al. 2012). It is possible that T. denticola, which attaches better to the substratum of the flow cells initially, helps with the incorporation and proliferation of P. gingivalis in the dual-species biofilms at earlier time points. T. denticola DfsB homologue may be involved in this process as discussed later. With the increase of P. gingivalis cells and subsequently the level of gingipains, the growth of T. denticola may be inhibited. This inhibition of T. denticola growth may be beneficial to maximize the synergy and symbioses between the two bacteria by maintaining an optimal ratio of P. gingivalis and T. denticola cells. Therefore, gingipains may play a role in the competitive interactions between P. gingivalis and T. denticola in dual-species biofilms that will ultimately benefit the survival of both bacteria in the biofilms. However, more experiments need to be carried out to prove this hypothesis and it should be noted that a specific phenotype observed on an agar plate does not necessarily reflect biofilm development. The plate inhibition assays look at the interactions between two bacterial monocultures that are grown separately while the static biofilm assays look at the interactions between bacteria that are grown together. Gingipains that were involved in the competitive interactions between P. gingivalis and T. denticola on plate may play a role in the cooperative interactions between the two bacteria in biofilms.

DfsBTd33520 was not involved in the monospecies biofilm formation of T. denticola ATCC 33520 but was involved in the synergistic biofilm formation between T. denticola ATCC 33520 and P. gingivalis W50. There was a significant reduction in the biomass of the dual-species biofilms of ∆DfsBTd33520 grown with P. gingivalis W50 as compared to that

259

of ATCC 33520 grown with P. gingivalis W50 (Figure 6.18). P. gingivalis W50, being the dominant species in the dual-species biofilms with ATCC 33520, was almost excluded from the dual-species biofilms with ∆DfsBTd33520 (Figure 6.19 and Table 6.12), suggesting that DfsBTd33520 is required for the incorporation of P. gingivalis into the dual- species biofilms with T. denticola. There are several possible explanations for this.

Firstly, DfsBTd33520 may be an adhesin required for the binding of P. gingivalis to T. denticola and the lack of this protein reduces the incorporation of P. gingivalis into T. denticola biofilms. However, the lack of DfsBTd33520 did not affect the coaggregation between P. gingivalis W50 and ∆DfsBTd33520, indicating that it is unlikely to be an adhesin for their binding. Secondly, DfsBTd33520 may be a growth stimulant for P. gingivalis W50 or a protective factor that protects P. gingivalis W50 against the inhibiting materials produced by T. denticola. To test for this hypothesis, P. gingivalis W50 cells were grown in the spent media of T. denticola ATCC 33520 and ∆DfsBTd33520. There was no difference in the growth of P. gingivalis W50 in the spent media of both T. denticola strains (data not shown), indicating that DfsBTd33520 is not involved in the stimulation or protection of

P. gingivalis W50 cells. Although it may be argued that DfsBTd33520 and/or the inhibiting factor are expressed only when the two bacteria are grown together and not when T. denticola is grown alone, DfsBTd35405 was not differentially expressed during the co- culture of T. denticola ATCC 35405 with P. gingivalis W50 and in the polymicrobial biofilms of T. denticola, P. gingivalis and T. forsythia (Zainal-Abidin et al. 2012, Tan et al. 2014). Other than the DfsB crystal structure, the sequence of DfsBTd35405 also matched with a high confidence with the crystal structures of several metal-dependent hydrolases (Table 6.2). The common characteristic of these metal-dependent hydrolases is the presence of a histidine triad in their active sites that functions to coordinate a metal such as nickel or zinc (Taylor et al. 2016). When aligned with these metal-dependent hydrolases, it was observed that DfsBTd35405 contained the first two conserved histidines of the histidine triad. Although the third histidine was replaced with an alanine residue, it is possible that DfsBTd35405 possess a function similar to that of a metal-dependent hydrolase and may be an enzyme involved in a crucial pathway of biofilm formation and development between T. denticola and P. gingivalis.

260

6.4 CONCLUSION In conclusion, sibling T. denticola colonies inhibit the growth of one another in a similar pattern as P. dendritiformis colonies and homologues of proteins necessary for the sibling killing mechanisms in P. dendritiformis are found in T. denticola. Despite having a high similarity to DfsB of P. dendritiformis, DfsBTd33520 is not involved in the sibling killing of T. denticola colonies. Several potential targets were identified from the agar extracted between the competing T. denticola colonies but they need to be further tested. Although synergy and symbioses are often observed between T. denticola and P. gingivalis, their colonies competed and inhibited the growth of one another on agar plates, likely through mechanisms that involve gingipains. DfsBTd33520 is important for the development of dual-species biofilms between T. denticola and P. gingivalis, particularly in the incorporation of P. gingivalis in the biofilms, but more experiments are needed to show the role that it plays in this process. The results from this study support the notion that the interactions between T. denticola and P. gingivalis are dynamic and it is the sum of the competitive and cooperative interactions that helps to sustain the interspecies growth symbiosis over an extended period of time.

261

CHAPTER 7 GENERAL DISCUSSION

262

Chronic periodontitis has a polymicrobial biofilm etiology. Changes in the subgingival plaque bacterial species composition and subsequent alteration of the host immune response are believed to initiate chronic periodontitis (Byrne et al. 2009) and interactions between key oral bacterial species are essential for the disease progression. In vivo, P. gingivalis and T. denticola are frequently found to co-exist in deep periodontal pockets and have been co-localized to the superficial layers of subgingival plaque as microcolony blooms adjacent to the pocket epithelium (Kigure et al. 1995, Zijnge et al. 2010). This suggests a possible interbacterial interaction between P. gingivalis and T. denticola that contributes to the initiation and development of chronic periodontitis. When grown as monospecies biofilms, T. denticola lost its typical spiral morphology but when grown in polymicrobial biofilms containing P. gingivalis, the cells retained the spiral morphology (Zhu et al. 2013). This observation suggested that T. denticola motility plays an important role in synergistic biofilm formation with P. gingivalis. In order to investigate this, three non-motile T. denticola mutants, ∆motA33520, ∆motB33520 and ∆flgE33520, were generated using an optimized transformation protocol. The optimized protocol which included the preparation of electrocompetent cells under anaerobic conditions significantly improved the transformation efficiency of T. denticola and reduced “false-positive” transformation frequency. This optimized protocol will be an important addition to the current genetic tools of T. denticola which will further contribute to the molecular analysis of T. denticola virulence. Although complementation of chromosomal mutations is important and is routinely performed to confirm the involvement of specific genes in particular phenotypes, it was not performed in this study due to time constraints and the need to prioritize the characterization of the mutants. Instead whole genome sequencing was performed to show that the only deletion and/or rearrangement was that of the targeted gene.

While the deletion of flgE produced an expected aflagellated mutant, the deletions of motA and motB resulted in an unexpected reduction in flagellar number, indicating the importance of a functional stator or motor in the proper assembly of T. denticola periplasmic flagella. The reduction in flagellar number in ∆motA33520 and ∆motB33520 could be partly attributed to the reduced abundance of FlaA and FlaB proteins in these two mutants. Nevertheless, the stoichiometry of the flagella in ∆motA33520 and ∆motB33520 was similar to that of the WT ATCC 33520. Consistent with the studies which showed that the morphology of spirochetes is the result of complex interactions between the cell

263

cylinder and the periplasmic flagella (Ruby et al. 1997, Motaleb et al. 2000), the absence

(∆flgE33520) or reduction of flagella (∆motA33520 and ∆motB33520) resulted in the mutants being less spiral and more rod shaped than the WT. This result has also reinforced the importance of a properly assembled and functional periplasmic flagella in maintaining the typical spiral morphology of T. denticola.

The comprehensive quantitative analysis of the proteomes of T. denticola mutants revealed the effects of motility gene inactivation on T. denticola proteomes and could be practiced routinely in the future on other T. denticola mutants in order to identify changes in the protein expression that might affect the mutant phenotypes. The deletion of motA, motB and flgE had influenced the level of motility-related proteins encoded from the genes derived from within or outside of the fla operon, as well as several chemotaxis- related proteins. This result demonstrated that the expression of chemotaxis and motility genes is a finely regulated and coordinated process where the expression of one gene serves as a checkpoint in the transcriptional and/or post-transcriptional controls of other genes. The importance of the transmembrane ion motive force (IMF) generated by the MotA/B ion channel complex in T. denticola was revealed in this study as the potential disruption of IMF in ∆motA33520 and ∆motB33520 resulted in the increased abundance of several proteins that are implicated in the cell stress response. The potential disruption of

IMF had also directly or indirectly impacted the growth of ∆motA33520 and ∆motB33520 through the change in abundance of a number of proteins including a ParA family ATPase (TDE0607) and the ATP-dependent ABC transporter TDE0983-TDE0987.

The loss of motility impacted the growth of ∆motA33520, ∆motB33520 and ∆flgE33520. This showed that T. denticola motility plays an important role in its normal growth, probably in nutrient acquisition and proper cell division. The motility and/or spiral morphology of T. denticola were shown to be required for its autoaggregation which in turn was essential for monospecies biofilm formation as the mutants that were less spiral than the WT had reduced abilities to autoaggregate and form monospecies biofilms. Furthermore, this study has confirmed the importance of T. denticola periplasmic flagella in its synergistic biofilm formation with P. gingivalis, particularly for its incorporation into the dual- species biofilms.

In this study, T. denticola sibling colonies were shown to inhibit the growth of one another on an agar plate in a similar manner to that seen in P. dendritiformis which was proposed

264

to be mediated by the dendritiformis sibling bacteriocin (DfsB) (Be'er et al. 2009). Given that the sibling killing phenomenon in P. dendritiformis is proposed to be cell density- dependent (Be'er et al. 2009, Be’er et al. 2010, Be’er et al. 2015) and biofilms are characterized by high cell density, the sibling killing phenomenon could have a significant effect on T. denticola biofilm formation, or possibly its synergistic biofilm formation with P. gingivalis. Bioinformatic analyses identified a DfsB homologue in T. denticola (TDE0659 or DfsBTd) that was highly similar to P. dendritiformis DfsB. Production and testing of a T. denticola mutant lacking this protein demonstrated that it was not involved in the sibling killing mechanism. Nevertheless, several proteins of interest were identified from the agar between T. denticola competing colonies, including the YD repeat protein TDE2022 which may play a role in either contact-dependent growth inhibition or coordination of multicellular behavior, and TDE0889 which has significant similarity to Slf, the cleaved product of DfsB. These proteins will need to be further tested for their roles in T. denticola sibling killing. For example, T. denticola mutants lacking each of the proteins could be generated and tested in the inhibition assay as mentioned in this study to determine if the proteins are involved in the sibling killing phenomenon in T. denticola. Alternatively, these proteins could be expressed as recombinant proteins from E. coli and used to investigate their effect on the growth of T. denticola.

T. denticola and P. gingivalis colonies inhibited the growth of one another on agar plates. Interestingly, gingipains were shown to be essential for the synergistic biofilm formation of T. denticola and P. gingivalis (Zhu et al. 2013) yet were involved in the inhibition of T. denticola growth by P. gingivalis colonies on plate. This indicated a gingipain- mediated competitive interaction between P. gingivalis and T. denticola that may balance their symbiotic interactions, leading to a state of homeostasis which would ultimately benefit the survival of both bacteria in biofilms. If this is the case, it will add significantly to the belief that the overall fitness of a multispecies biofilm community is reliant on the sum of competitive and cooperative interactions between the members in the biofilm.

Unexpectedly, DfsBTd which appeared to play no role in the interactions of P. gingivalis and T. denticola on agar plates was shown to be involved in their synergistic biofilm. The similarities in structure of the T. denticola DfsBTd homologue with several metal- dependent hydrolases and the possession of two conserved histidines suggested that it

265

may be a metal-dependent hydrolase involved in a crucial pathway of T. denticola and P. gingivalis synergistic biofilm formation.

Overall, the improved transformation protocol developed in this study has been used successfully to generate a number of T. denticola mutants including the motility mutants and ∆DfsBTd33520. The mutants have been used to show the importance of T. denticola periplasmic flagella and DfsB homologue in synergistic biofilm formation with P. gingivalis.

266

REFERENCES

Abe, N., Baba, A., Takii, R., Nakayama, K., Kamaguchi, A., Shibata, Y., Abiko, Y., Okamoto, K., Kadowaki, T. and Yamamoto, K. (2004) 'Roles of Arg- and Lys- gingipains in coaggregation of Porphyromonas gingivalis: identification of its responsible molecules in translation products of rgpA, kgp, and hagA genes', Biological Chemistry, 385(11), 1041-1047.

Abiko, Y., Nagano, K., Yoshida, Y. and Yoshimura, F. (2014a) 'Characterization of Treponema denticola mutants defective in the major antigenic proteins, Msp and TmpC', PLOS One, 9(11), e113565.

Abiko, Y., Nagano, K., Yoshida, Y. and Yoshimura, F. (2014b) 'Major membrane protein TDE2508 regulates adhesive potency in Treponema denticola', PLOS One, 9(2), e89051.

Amano, A. (2007) 'Disruption of epithelial barrier and impairment of cellular function by Porphyromonas gingivalis', Frontiers in Bioscience (Landmark Edition), 12, 3965-3974.

Amano, A., Nakagawa, I., Okahashi, N. and Hamada, N. (2004) 'Variations of Porphyromonas gingivalis fimbriae in relation to microbial pathogenesis', Journal of Periodontal Research, 39(2), 136-142.

An, D. D., Danhorn, T., Fuqua, C. and Parsek, M. R. (2006) 'Quorum sensing and motility mediate interactions between Pseudomonas aeruginosa and Agrobacterium tumefaciens in biofilm cocultures', Proceedings of the National Academy of Sciences of the United States of America, 103(10), 3828-3833.

Armitage, G. C. (1995) 'Clinical evaluation of periodontal diseases', Periodontology 2000, 7, 39-53.

Armitage, G. C. (2004) 'Periodontal diagnoses and classification of periodontal diseases', Periodontology 2000, 34, 9-21.

Armitage, G. C. and Cullinan, M. P. (2010) 'Comparison of the clinical features of chronic and aggressive periodontitis', Periodontology 2000, 53, 12-27.

Asai, Y., Ohyama, Y., Gen, K. and Ogawa, T. (2001) 'Bacterial fimbriae and their peptides activate human gingival epithelial cells through Toll-like receptor 2', Infection and Immunity, 69(12), 7387-7395.

267

Bainbridge, B. W., Coats, S. R. and Darveau, R. P. (2002) 'Porphyromonas gingivalis lipopolysaccharide displays functionally diverse interactions with the innate host defense system', Annals of Periodontology, 7(1), 29-37.

Bamford, C. V., Fenno, J. C., Jenkinson, H. F. and Dymock, D. (2007) 'The chymotrypsin-like protease complex of Treponema denticola ATCC 35405 mediates fibrinogen adherence and degradation', Infection and Immunity, 75(9), 4364-4372.

Banbula, A., Bugno, M., Kuster, A., Heinrich, P. C., Travis, J. and Potempa, J. (1999) 'Rapid and efficient inactivation of IL-6 gingipains, lysine- and arginine-specific proteinases from Porphyromonas gingivalis', Biochemical and Biophysical Research Communications, 261(3), 598-602.

Barquera, B. (2014) 'The sodium pumping NADH:quinone oxidoreductase (Na+-NQR), a unique redox-driven ion pump', Journal of Bioenergetics and Biomembranes, 46(4), 289-98.

Be'er, A., Florin, E. L., Fisher, C. R., Swinney, H. L. and Payne, S. M. (2011) 'Surviving bacterial sibling rivalry: inducible and reversible phenotypic switching in Paenibacillus dendritiformis', mBio, 2(3), e00069-11.

Be'er, A., Zhang, H. P., Florin, E. L., Payne, S. M., Ben-Jacob, E. and Swinney, H. L. (2009) 'Deadly competition between sibling bacterial colonies', Proceedings of the National Academy of Sciences of the United States of America, 106(2), 428- 33.

Be’er, A., Ariel, G., Kalisman, O., Helman, Y., Sirota-Madi, A., Zhang, H. P., Florin, E. L., Payne, S. M., Ben-Jacob, E. and Swinney, H. L. (2010) 'Lethal protein produced in response to competition between sibling bacterial colonies', Proceedings of the National Academy of Sciences, 107(14), 6258-6263.

Be’er, A., Benisty, S., Ariel, G. and Ben-Jacob, E. (2015) 'Interplay between sibling bacterial colonies' in Hagen, S. J., ed., The Physical Basis of Bacterial Quorum Communication, New York, NY: Springer New York, 145-162.

Berg, H. C. (1976) 'How spirochetes may swim', Journal of Theoretical Biology, 56(2), 269-273.

Bian, J. and Li, C. (2011) 'Disruption of a type II endonuclease (TDE0911) enables Treponema denticola ATCC 35405 to accept an unmethylated shuttle vector', Applied and Environmental Microbiology, 77(13), 4573-4578.

268

Bian, X. L., Wang, H. T., Ning, Y., Lee, S. Y. and Fenno, J. C. (2005) 'Mutagenesis of a novel gene in the prcA-prtP protease locus affects expression of Treponema denticola membrane complexes', Infection and Immunity, 73(2), 1252-1255.

Biasini, M., Bienert, S., Waterhouse, A., Arnold, K., Studer, G., Schmidt, T., Kiefer, F., Gallo Cassarino, T., Bertoni, M., Bordoli, L. and Schwede, T. (2014) 'SWISS- MODEL: modelling protein tertiary and quaternary structure using evolutionary information', Nucleic Acids Research, 42(Web Server issue), W252-258.

Biyikoglu, B., Ricker, A. and Diaz, P. I. (2012) 'Strain-specific colonization patterns and serum modulation of multi-species oral biofilm development', Anaerobe, 18(4), 459-470.

Bjellqvist, B., Basse, B., Olsen, E. and Celis, J. E. (1994) 'Reference points for comparisons of two-dimensional maps of proteins from different human cell types defined in a pH scale where isoelectric points correlate with polypeptide compositions', ELECTROPHORESIS, 15(3-4), 529-539.

Bjellqvist, B., Hughes, G. J., Pasquali, C., Paquet, N., Ravier, F., Sanchez, J. C., Frutiger, S. and Hochstrasser, D. (1993) 'The focusing positions of polypeptides in immobilized pH gradients can be predicted from their amino acid sequences', ELECTROPHORESIS, 14(1), 1023-1031.

Bond, J. P. and Francklyn, C. (2000) 'Proteobacterial histidine-biosynthetic pathways are paraphyletic', Journal of Molecular Evolution, 50(4), 339-347.

Brennan, C. A., Dombroski, A. J. and Platt, T. (1987) 'Transcription termination factor rho is an RNA-DNA helicase', Cell, 48(6), 945-52.

Brissette, C. A. and Lukehart, S. A. (2007) 'Mechanisms of decreased susceptibility to β- defensins by Treponema denticola', Infection and Immunity, 75(5), 2307-2315.

Brogden, K. A., Guthmiller, J. M. and Taylor, C. E. (2005) 'Human polymicrobial infections', The Lancet, 365(9455), 253-255.

Brown, J. R., Douady, C. J., Italia, M. J., Marshall, W. E. and Stanhope, M. J. (2001) 'Universal trees based on large combined protein sequence data sets', Nature Genetics, 28(3), 281-285.

Brown, J. R., Zhang, J. and Hodgson, J. E. (1998) 'A bacterial antibiotic resistance gene with eukaryotic origins', Current Biology, 8(11), R365-367.

269

Buckley, A. M., Petersen, J., Roe, A. J., Douce, G. R. and Christie, J. M. (2015) 'LOV- based reporters for fluorescence imaging', Current Opinion in Chemical Biology, 27, 39-45.

Bullwinkle, T. J. and Ibba, M. (2016) 'Translation quality control is critical for bacterial responses to amino acid stress', Proceedings of the National Academy of Sciences of the United States of America, 113(8), 2252-2257.

Byrne, S. J., Dashper, S. G., Darby, I. B., Adams, G. G., Hoffmann, B. and Reynolds, E. C. (2009) 'Progression of chronic periodontitis can be predicted by the levels of Porphyromonas gingivalis and Treponema denticola in subgingival plaque', Oral Microbiology and Immunology, (6), 469.

Cahoon, L. A. and Seifert, H. S. (2009) 'An alternative DNA structure is necessary for pilin antigenic variation in Neisseria gonorrhoeae', Science, 325(5941), 764-767.

Calkins, C. C., Platt, K., Potempa, J. and Travis, J. (1998) 'Inactivation of tumor necrosis factor-alpha by proteinases (gingipains) from the periodontal pathogen, Porphyromonas gingivalis. Implications of immune evasion', Journal of Biological Chemistry, 273(12), 6611-6614.

Capone, R. F., Ning, Y., Pakulis, N., Alhazzazi, T. and Fenno, J. C. (2007) 'Characterization of Treponema denticola pyrF encoding orotidine-5'- monophosphate decarboxylase', FEMS Microbiology Letters, 268(2), 261-267.

Carlsson, J. (1997) 'Bacterial metabolism in dental biofilms', Advances in Dental Research, 11(1), 75-80.

Caudry, S., Klitorinos, A., Gharbia, S. E., Pssara, N., Siboo, R., Keng, T. and Chan, E. C. (1995) 'Distribution and characterization of plasmids in oral anaerobic spirochetes', Oral Microbiology and Immunology, 10(1), 8-12.

Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W. and Prasher, D. C. (1994) 'Green fluorescent protein as a marker for gene expression', Science, 263(5148), 802-805.

Chan, E. C., Siboo, R., Keng, T., Psarra, N., Hurley, R., Cheng, S. L. and Iugovaz, I. (1993) 'Treponema denticola (ex Brumpt 1925) sp. nov., nom. rev., and identification of new spirochete isolates from periodontal pockets', International Journal of Systematic Bacteriology, 43(2), 196-203.

Charon, N. W. and Goldstein, S. F. (2002) 'Genetics of motility and chemotaxis of a fascinating group of bacteria: the spirochetes', Annual Review of Genetics, 36, 47- 73.

270

Chen, T. and Duncan, M. J. (2004) 'Gingipain adhesin domains mediate Porphyromonas gingivalis adherence to epithelial cells', Microbial Pathogenesis, 36(4), 205-9.

Chen, T., Nakayama, K., Belliveau, L. and Duncan, M. J. (2001) 'Porphyromonas gingivalis gingipains and adhesion to epithelial cells', Infection and Immunity, 69(5), 3048-3056.

Chen, Y. Y., Peng, B., Yang, Q., Glew, M. D., Veith, P. D., Cross, K. J., Goldie, K. N., Chen, D., O'Brien-Simpson, N., Dashper, S. G. and Reynolds, E. C. (2011) 'The outer membrane protein LptO is essential for the O-deacylation of LPS and the co-ordinated secretion and attachment of A-LPS and CTD proteins in Porphyromonas gingivalis', Molecular Microbiology, 79(5), 1380-1401.

Chevance, F. F. and Hughes, K. T. (2008) 'Coordinating assembly of a bacterial macromolecular machine', Nature Reviews Microbiology, 6(6), 455-465.

Chi, B., Chauhan, S. and Kuramitsu, H. (1999) 'Development of a system for expressing heterologous genes in the oral spirochete Treponema denticola and its use in expression of the Treponema pallidum flaA gene', Infection and Immunity, 67(7), 3653-3656.

Chi, B., Limberger, R. J. and Kuramitsu, H. K. (2002) 'Complementation of a Treponema denticola flgE mutant with a novel coumermycin A1-resistant T. denticola shuttle vector system', Infection and Immunity, 70(4), 2233-2237.

Chi, B., Qi, M. and Kuramitsu, H. K. (2003) 'Role of dentilisin in Treponema denticola epithelial cell layer penetration', Research in Microbiology, 154(9), 637-643.

Cho, K. H., Cho, D., Wang, G. R. and Salyers, A. A. (2001) 'New regulatory gene that contributes to control of Bacteroides thetaiotaomicron starch utilization genes', Journal of Bacteriology, 183(24), 7198-7205.

Choi, C. H., DeGuzman, J. V., Lamont, R. J. and Yilmaz, Ö. (2011) 'Genetic transformation of an obligate anaerobe, P. gingivalis for FMN-green fluorescent protein expression in studying host-microbe interaction', PLOS One, 6(4), e18499.

Conover, W. J. and Iman, R. L. (1979) On multiple-comparisons procedures, Technical Report LA-7677-MS, Los Alamos Scientific Laboratory.

Correia, F. F., Plummer, A. R., Ellen, R. P., Wyss, C., Boches, S. K., Galvin, J. L., Paster, B. J. and Dewhirst, F. E. (2003) 'Two Paralogous Families of a Two-Gene Subtilisin Operon Are Widely Distributed in Oral Treponemes', Journal of Bacteriology, 185(23), 6860-6869.

271

Cripps, A. W. and Otczyk, D. C. (2006) 'Prospects for a vaccine against otitis media', Expert Review of Vaccines, 5(4), 517-534.

Curtis, M. A., Aduse-Opoku, J. and Rangarajan, M. (2001) 'Cysteine proteases of Porphyromonas gingivalis', Critical Reviews in Oral Biology & Medicine, 12(3), 192-216.

Curtis, M. A., Thickett, A., Slaney, J. M., Rangarajan, M., Aduse-Opoku, J., Shepherd, P., Paramonov, N. and Hounsell, E. F. (1999) 'Variable carbohydrate modifications to the catalytic chains of the RgpA and RgpB proteases of Porphyromonas gingivalis W50', Infection and Immunity, 67(8), 3816-3823.

Darveau, R. P., Pham, T. T., Lemley, K., Reife, R. A., Bainbridge, B. W., Coats, S. R., Howald, W. N., Way, S. S. and Hajjar, A. M. (2004) 'Porphyromonas gingivalis lipopolysaccharide contains multiple lipid A species that functionally interact with both toll-like receptors 2 and 4', Infection and Immunity, 72(9), 5041-5051.

Dashper, S., O'Brien-Simpson, N., Liu, S. W., Paolini, R., Mitchell, H., Walsh, K., D'Cruze, T., Hoffmann, B., Catmull, D., Zhu, Y. and Reynolds, E. (2014) 'Oxantel disrupts polymicrobial biofilm development of periodontal pathogens', Antimicrobial Agents and Chemotherapy, 58(1), 378-385.

Dashper, S. G., Cross, K. J., Slakeski, N., Lissel, P., Aulakh, P., Moore, C. and Reynolds, E. C. (2004) 'Hemoglobin hydrolysis and heme acquisition by Porphyromonas gingivalis', Oral Microbiology and Immunology, 19(1), 50-56.

Dashper, S. G., Mitchell, H. L., Seers, C. A., Gladman, S. L., Seemann, T., Bulach, D. M., Chandry, P. S., Cross, K. J., Cleal, S. M. and Reynolds, E. C. (2017) 'Porphyromonas gingivalis uses specific domain rearrangements and allelic exchange to generate diversity in surface virulence factors', Frontiers in Microbiology, 8, 48.

Dashper, S. G., Seers, C. A., Tan, K. H. and Reynolds, E. C. (2011) 'Virulence factors of the oral spirochete Treponema denticola', Journal of Dental Research, 90(6), 691- 703.

Davey, M. E. (2006) 'Techniques for the growth of Porphyromonas gingivalis biofilms', Periodontology 2000, 42, 27-35. del Pozo, J. L. and Patel, R. (2007) 'The challenge of treating biofilm-associated bacterial infections', Clinical Pharmacology & Therapeutics, 82(2), 204-209.

272

Ding, Y., Uitto, V. J., Haapasalo, M., Lounatmaa, K., Konttinen, Y. T., Salo, T., Grenier, D. and Sorsa, T. (1996) 'Membrane components of Treponema denticola trigger proteinase release from human polymorphonuclear leukocytes', Journal of Dental Research, 75(12), 1986-1993.

Divakaruni, A. V., Baida, C., White, C. L. and Gober, J. W. (2007) 'The cell shape proteins MreB and MreC control cell morphogenesis by positioning cell wall synthetic complexes', Molecular Microbiology, 66(1), 174-188.

Dorrian, J. M., Briggs, D. A., Ridley, M. L., Layfield, R. and Kerr, I. D. (2011) 'Induction of a stress response in Lactococcus lactis is associated with a resistance to ribosomally active antibiotics', The FEBS Journal, 278(21), 4015-4024.

Drepper, T., Eggert, T., Circolone, F., Heck, A., Krauss, U., Guterl, J. K., Wendorff, M., Losi, A., Gartner, W. and Jaeger, K. E. (2007) 'Reporter proteins for in vivo fluorescence without oxygen', Nature Biotechnology, 25(4), 443-445.

Drepper, T., Gensch, T. and Pohl, M. (2013) 'Advanced in vivo applications of blue light photoreceptors as alternative fluorescent proteins', Photochemical & Photobiological Sciences, 12(7), 1125-1134.

Duncan, L., Yoshioka, M., Chandad, F. and Grenier, D. (2004) 'Loss of lipopolysaccharide receptor CD14 from the surface of human macrophage-like cells mediated by Porphyromonas gingivalis outer membrane vesicles', Microbial Pathogenesis, 36(6), 319-325.

Edwards, A. M., Jenkinson, H. F., Woodward, M. J. and Dymock, D. (2005) 'Binding properties and adhesion-mediating regions of the major sheath protein of Treponema denticola ATCC 35405', Infection and Immunity, 73(5), 2891-2898.

Eke, P. I., Dye, B. A., Wei, L., Slade, G. D., Thornton-Evans, G. O., Borgnakke, W. S., Taylor, G. W., Page, R. C., Beck, J. D. and Genco, R. J. (2015) 'Update on Prevalence of Periodontitis in Adults in the United States: NHANES 2009 - 2012', Journal of Periodontology, 1-18.

Elias, S. and Banin, E. (2012) 'Multi-species biofilms: living with friendly neighbors', FEMS Microbiology Reviews, 36(5), 990-1004.

Ellen, R. P. (2006) 'Virulence determinants of oral Treponemes' in Radolf, J. D. and Lukehart, S. A., eds., Pathogenic Treponema: molecular and cellular biology, Wymondham, Norfolk, UK: Caister Academic Press, 357-386.

273

Ellermeier, C. D., Hobbs, E. C., Gonzalez-Pastor, J. E. and Losick, R. (2006) 'A three- protein signaling pathway governing immunity to a bacterial cannibalism toxin', Cell, 124(3), 549-559.

Ellis, T. N. and Kuehn, M. J. (2010) 'Virulence and immunomodulatory roles of bacterial outer membrane vesicles', Microbiology and Molecular Biology Reviews, 74(1), 81-94.

Endo, A., Watanabe, T., Ogata, N., Nozawa, T., Aikawa, C., Arakawa, S., Maruyama, F., Izumi, Y. and Nakagawa, I. (2015) 'Comparative genome analysis and identification of competitive and cooperative interactions in a polymicrobial disease', The ISME Journal, 9(3), 629-642.

Enersen, M., Nakano, K. and Amano, A. (2013) 'Porphyromonas gingivalis fimbriae', Journal of Oral Microbiology, 5, 10.3402/jom.v5i0.20265.

Fassone, E., Duncan, A. J., Taanman, J. W., Pagnamenta, A. T., Sadowski, M. I., Holand, T., Qasim, W., Rutland, P., Calvo, S. E., Mootha, V. K., Bitner-Glindzicz, M. and Rahman, S. (2010) 'FOXRED1, encoding an FAD-dependent oxidoreductase complex-I-specific molecular chaperone, is mutated in infantile-onset mitochondrial encephalopathy', Human Molecular Genetics, 19(24), 4837-4847.

Fenno, J. C., Hannam, P. M., Leung, W. K., Tamura, M., Uitto, V. J. and McBride, B. C. (1998) 'Cytopathic effects of the major surface protein and the chymotrypsin-like protease of Treponema denticola', Infection and Immunity, 66(5), 1869-1877.

Fenno, J. C. and McBride, B. C. (1998) 'Virulence factors of oral treponemes', Anaerobe, 4(1), 1-17.

Fenno, J. C., Muller, K. H. and McBride, B. C. (1996) 'Sequence analysis, expression, and binding activity of recombinant major outer sheath protein (Msp) of Treponema denticola', Journal of Bacteriology, 178(9), 2489-2497.

Fischbach, M. A. and Sonnenburg, J. L. (2011) 'Eating For Two: How Metabolism Establishes Interspecies Interactions in the Gut', Cell Host & Microbe, (4), 336- 347.

Franklin, M. J., Chang, C., Akiyama, T. and Bothner, B. (2015) 'New technologies for studying biofilms', Microbiology Spectrum, 3(4), 10.1128/microbiolspec.MB- 0016-2014.

Frederick, J. R., Sarkar, J., McDowell, J. V. and Marconi, R. T. (2011) 'Molecular signaling mechanisms of the periopathogen, Treponema denticola', Journal of Dental Research, 90(10), 1155-1163.

274

Fujiwara, T., Ogawa, T., Sobue, S. and Hamada, S. (1990) 'Chemical, immunobiological and antigenic characterizations of lipopolysaccharides from Bacteroides gingivalis strains', Journal of General Microbiology, 136(2), 319-326.

Fung, E., Bouet, J. Y. and Funnell, B. E. (2001) 'Probing the ATP‐binding site of P1 ParA: partition and repression have different requirements for ATP binding and hydrolysis', The EMBO Journal, 20(17), 4901-4911.

Furuta, N., Takeuchi, H. and Amano, A. (2009a) 'Entry of Porphyromonas gingivalis outer membrane vesicles into epithelial cells causes cellular functional impairment', Infection and Immunity, 77(11), 4761-4770.

Furuta, N., Tsuda, K., Omori, H., Yoshimori, T., Yoshimura, F. and Amano, A. (2009b) 'Porphyromonas gingivalis outer membrane vesicles enter human epithelial cells via an endocytic pathway and are sorted to lysosomal compartments', Infection and Immunity, 77(10), 4187-4196.

Gaibani, P., Caroli, F., Nucci, C. and Sambri, V. (2010) 'Major surface protein complex of Treponema denticola induces the production of tumor necrosis factor alpha, interleukin-1 beta, interleukin-6 and matrix metalloproteinase 9 by primary human peripheral blood monocytes', Journal of Periodontal Research, 45(3), 361- 366.

Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, S., Wilkins, M. R., Appel, R. D. and Bairoch, A. (2005) 'Protein Identification and Analysis Tools on the ExPASy Server' in Walker, J. M., ed., The Proteomics Protocols Handbook, Totowa, NJ: Humana Press, 571-607.

Ghoul, M. and Mitri, S. (2016) 'The Ecology and Evolution of Microbial Competition', Trends Microbiol, 24(10), 833-845.

Gibson, F. C., 3rd and Genco, C. A. (2007) 'Porphyromonas gingivalis mediated periodontal disease and atherosclerosis: disparate diseases with commonalities in pathogenesis through TLRs', Current Pharmaceutical Design, 13(36), 3665- 3675.

Glew, M. D., Veith, P. D., Chen, D., Gorasia, D. G., Peng, B. and Reynolds, E. C. (2017) 'PorV is an outer membrane shuttle protein for the type IX secretion system', Scientific Reports, 7, 8790.

Godovikova, V., Goetting-Minesky, M. P., Shin, J. M., Kapila, Y. L., Rickard, A. H. and Fenno, J. C. (2015) 'A modified shuttle plasmid facilitates expression of a flavin mononucleotide-based fluorescent protein in Treponema denticola ATCC 35405', Applied and Environmental Microbiology, 81(18), 6496-6504.

275

Godovikova, V., Wang, H. T., Goetting-Minesky, M. P., Ning, Y., Capone, R. F., Slater, C. K. and Fenno, J. C. (2010) 'Treponema denticola PrcB is required for expression and activity of the PrcA-PrtP (dentilisin) complex', Journal of Bacteriology, 192(13), 3337-3344.

Gonzalez-Pastor, J. E., Hobbs, E. C. and Losick, R. (2003) 'Cannibalism by sporulating bacteria', Science, 301(5632), 510-513.

Greene, S. R. and Stamm, L. V. (1999) 'Molecular characterization of a chemotaxis operon in the oral spirochete, Treponema denticola', Gene, 232(1), 59-68.

Grenier, D. (1992a) 'Inactivation of human serum bactericidal activity by a trypsinlike protease isolated from Porphyromonas gingivalis', Infection and Immunity, 60(5), 1854-1857.

Grenier, D. (1992b) 'Nutritional interactions between two suspected periodontopathogens, Treponema denticola and Porphyromonas gingivalis', Infection and Immunity, 60(12), 5298-5301.

Grenier, D. (2013) 'Porphyromonas gingivalis outer membrane vesicles mediate coaggregation and piggybacking of Treponema denticola and Lachnoanaerobaculum saburreum', International Journal of Dentistry, 2013, 305476.

Grenier, D. and Mayrand, D. (1987) 'Functional characterization of extracellular vesicles produced by Bacteroides gingivalis', Infection and Immunity, 55(1), 111-117.

Grenier, D. and Mayrand, D. (2001) 'Cleavage of human immunoglobulin G by Treponema denticola', Anaerobe, 7(1), 1-4.

Grenier, D., Roy, S., Chandad, F., Plamondon, P., Yoshioka, M., Nakayama, K. and Mayrand, D. (2003) 'Effect of inactivation of the Arg- and/or Lys-gingipain gene on selected virulence and physiological properties of Porphyromonas gingivalis', Infection and Immunity, 71(8), 4742-4748.

Gui, M. J., Dashper, S. G., Slakeski, N., Chen, Y. Y. and Reynolds, E. C. (2016) 'Spheres of influence: Porphyromonas gingivalis outer membrane vesicles', Molecular Oral Microbiology, 31(5), 365-378.

Guiral, S., Mitchell, T. J., Martin, B. and Claverys, J. P. (2005) 'Competence-programmed predation of noncompetent cells in the human pathogen Streptococcus pneumoniae: genetic requirements', Proceedings of the National Academy of Sciences of the United States of America, 102(24), 8710-8715.

276

Guo, Y., Nguyen, K. A. and Potempa, J. (2010) 'Dichotomy of gingipains action as virulence factors: from cleaving substrates with the precision of a surgeon's knife to a meat chopper-like brutal degradation of proteins', Periodontology 2000, 54(1), 15-44.

Hajishengallis, G., Darveau, R. P. and Curtis, M. A. (2012) 'The keystone-pathogen hypothesis', Nature Reviews Microbiology, 10(10), 717-725.

Hajishengallis, G. and Lambris, J. D. (2011) 'Microbial manipulation of receptor crosstalk in innate immunity', Nature Reviews Immunology, 11(3), 187-200.

Hajishengallis, G., Shakhatreh, M. A., Wang, M. and Liang, S. (2007) 'Complement receptor 3 blockade promotes IL-12-mediated clearance of Porphyromonas gingivalis and negates its virulence in vivo', Journal of Immunology, 179(4), 2359-2367.

Hamada, N., Watanabe, K., Sasakawa, C., Yoshikawa, M., Yoshimura, F. and Umemoto, T. (1994) 'Construction and characterization of a fimA mutant of Porphyromonas gingivalis', Infection and Immunity, 62(5), 1696-1704.

Hamedi, M., Belibasakis, G. N., Cruchley, A. T., Rangarajan, M., Curtis, M. A. and Bostanci, N. (2009) 'Porphyromonas gingivalis culture supernatants differentially regulate interleukin-1beta and interleukin-18 in human monocytic cells', Cytokine, 45(2), 99-104.

Hashimoto, M., Ogawa, S., Asai, Y., Takai, Y. and Ogawa, T. (2003) 'Binding of Porphyromonas gingivalis fimbriae to Treponema denticola dentilisin', FEMS Microbiology Letters, 226(2), 267-271.

Henry, J. T. and Crosson, S. (2011) 'Ligand binding PAS domains in a genomic, cellular, and structural context', Annual Review of Microbiology, 65, 261-286.

Henz, S. R., Huson, D. H., Auch, A. F., Nieselt-Struwe, K. and Schuster, S. C. (2005) 'Whole-genome prokaryotic phylogeny', Bioinformatics, 21(10), 2329-2335.

Herath, T. D., Wang, Y., Seneviratne, C. J., Lu, Q., Darveau, R. P., Wang, C. Y. and Jin, L. (2011) 'Porphyromonas gingivalis lipopolysaccharide lipid A heterogeneity differentially modulates the expression of IL-6 and IL-8 in human gingival fibroblasts', Journal of Clinical Periodontology, 38(8), 694-701.

Hess, J. F., Oosawa, K., Kaplan, N. and Simon, M. I. (1988) 'Phosphorylation of three proteins in the signaling pathway of bacterial chemotaxis', Cell, 53(1), 79-87.

277

Heydorn, A., Nielsen, A. T., Hentzer, M., Sternberg, C., Givskov, M., Ersboll, B. K. and Molin, S. (2000) 'Quantification of biofilm structures by the novel computer program COMSTAT', Microbiology, 146 ( Pt 10), 2395-2407.

Hiramine, H., Watanabe, K., Hamada, N. and Umemoto, T. (2003) 'Porphyromonas gingivalis 67-kDa fimbriae induced cytokine production and osteoclast differentiation utilizing TLR2', FEMS Microbiology Letters, 229(1), 49-55.

Hollmann, R. and Van der Hoeven, H. J. (1999) 'Inability of intact cells of Treponema denticola to degrade human serum proteins IgA, IgG and albumin', Journal of Clinical Periodontology, 26(7), 477-479.

Holt, S. C. and Ebersole, J. L. (2005) 'Porphyromonas gingivalis, Treponema denticola, and Tannerella forsythia: the 'red complex', a prototype polybacterial pathogenic consortium in periodontitis', Periodontology 2000, 38, 72-122.

Holt, S. C., Kesavalu, L., Walker, S. and Genco, C. A. (1999) 'Virulence factors of Porphyromonas gingivalis', Periodontology 2000, 20, 168-238.

Horton, R. M., Hunt, H. D., Ho, S. N., Pullen, J. K. and Pease, L. R. (1989) 'Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension', Gene, 77(1), 61-68.

Hosking, E. R., Vogt, C., Bakker, E. P. and Manson, M. D. (2006) 'The Escherichia coli MotAB proton channel unplugged', Journal of Molecular Biology, 364(5), 921- 937.

Houry, A., Briandet, R., Aymerich, S. and Gohar, M. (2010) 'Involvement of motility and flagella in Bacillus cereus biofilm formation', Microbiology, 156(Pt 4), 1009- 1018.

Houry, A., Gohar, M., Deschamps, J., Tischenko, E., Aymerich, S., Gruss, A. and Briandet, R. (2012) 'Bacterial swimmers that infiltrate and take over the biofilm matrix', Proceedings of the National Academy of Sciences, 109(32), 13088-13093.

Hughes, K. T., Gillen, K. L., Semon, M. J. and Karlinsey, J. E. (1993) 'Sensing structural intermediates in bacterial flagellar assembly by export of a negative regulator', Science, 262(5137), 1277-1280.

Hwang, L. C., Vecchiarelli, A. G., Han, Y.-W., Mizuuchi, M., Harada, Y., Funnell, B. E. and Mizuuchi, K. (2013) 'ParA-mediated plasmid partition driven by protein pattern self-organization', The EMBO Journal, 32(9), 1238-1249.

278

Ibba, M., Morgan, S., Curnow, A. W., Pridmore, D. R., Vothknecht, U. C., Gardner, W., Lin, W., Woese, C. R. and Soll, D. (1997) 'A euryarchaeal lysyl-tRNA synthetase: resemblance to class I synthetases', Science, 278(5340), 1119-1122.

Imayoshi, R., Cho, T. and Kaminishi, H. (2011) 'NO production in RAW264 cells stimulated with Porphyromonas gingivalis extracellular vesicles', Oral Diseases, 17(1), 83-89.

Ishihara, K. (2010) 'Virulence factors of Treponema denticola', Periodontology 2000, 54(1), 117-135.

Ishihara, K., Kuramitsu, H. K., Miura, T. and Okuda, K. (1998) 'Dentilisin activity affects the organization of the outer sheath of Treponema denticola', Journal of Bacteriology, 180(15), 3837-3844.

Ishihara, K., Miura, T., Kuramitsu, H. K. and Okuda, K. (1996) 'Characterization of the Treponema denticola prtP gene encoding a prolyl-phenylalanine-specific protease (dentilisin)', Infection and Immunity, 64(12), 5178-5186.

Ishihara, K., Wawrzonek, K., Shaw, L. N., Inagaki, S., Miyamoto, M. and Potempa, J. (2010) 'Dentipain, a Streptococcus pyogenes IdeS protease homolog, is a novel virulence factor of Treponema denticola', Biological Chemistry, 391(9), 1047- 1055.

Italiani, V. C., Zuleta, L. F. and Marques, M. V. (2002) 'The transcription termination factor Rho is required for oxidative stress survival in Caulobacter crescentus', Molecular Microbiology, 44(1), 181-194.

Ito, M., Terahara, N., Fujinami, S. and Krulwich, T. A. (2005) 'Properties of motility in Bacillus subtilis powered by the H+- coupled MotAB flagellar stator, Na+-coupled MotPS or hybrid stators MotAS or MotPB', Journal of Molecular Biology, 352(2), 396-408.

Ito, R., Ishihara, K., Shoji, M., Nakayama, K. and Okuda, K. (2010) 'Hemagglutinin/Adhesin domains of Porphyromonas gingivalis play key roles in coaggregation with Treponema denticola', FEMS Immunology and Medical Microbiology, 60(3), 251-260.

Ivic, A., MacDougall, J., Russell, R. R. and Penn, C. W. (1991) 'Isolation and characterization of a plasmid from Treponema denticola', FEMS Microbiology Letters, 62(2-3), 189-193.

279

Izard, J., Hsieh, C. E., Limberger, R. J., Mannella, C. A. and Marko, M. (2008) 'Native cellular architecture of Treponema denticola revealed by cryo-electron tomography', Journal of Structural Biology, 163(1), 10-17.

Jaattela, M., Cande, C. and Kroemer, G. (2004) 'Lysosomes and mitochondria in the commitment to apoptosis: a potential role for cathepsin D and AIF', Cell Death Differ, 11(2), 135-136.

Jackson, A. P., Thomas, G. H., Parkhill, J. and Thomson, N. R. (2009) 'Evolutionary diversification of an ancient gene family (rhs) through C-terminal displacement', BMC Genomics, 10, 584.

Jackson, C. A., Hoffmann, B., Slakeski, N., Cleal, S., Hendtlass, A. J. and Reynolds, E. C. (2000) 'A consensus Porphyromonas gingivalis promoter sequence', FEMS Microbiology Letters, 186(1), 133-138.

Jakubovics, N. S., Gill, S. R., Iobst, S. E., Vickerman, M. M. and Kolenbrander, P. E. (2008) 'Regulation of gene expression in a mixed-genus community: stabilized arginine biosynthesis in Streptococcus gordonii by coaggregation with Actinomyces naeslundii', Journal of Bacteriology, 190(10), 3646-3657.

Jovanovic, T., Ascenso, C., Hazlett, K. R., Sikkink, R., Krebs, C., Litwiller, R., Benson, L. M., Moura, I., Moura, J. J., Radolf, J. D., Huynh, B. H., Naylor, S. and Rusnak, F. (2000) 'Neelaredoxin, an iron-binding protein from the syphilis spirochete, Treponema pallidum, is a superoxide reductase', Journal of Biological Chemistry, 275(37), 28439-28448.

Kadowaki, T., Nakayama, K., Yoshimura, F., Okamoto, K., Abe, N. and Yamamoto, K. (1998) 'Arg-gingipain acts as a major processing enzyme for various cell surface proteins in Porphyromonas gingivalis', The Journal of Biological Chemistry, 273(44), 29072-29076.

Kamaguchi, A., Nakayama, K., Ichiyama, S., Nakamura, R., Watanabe, T., Ohta, M., Baba, H. and Ohyama, T. (2003a) 'Effect of Porphyromonas gingivalis vesicles on coaggregation of Staphylococcus aureus to oral microorganisms', Current Microbiology, 47(6), 485-491.

Kamaguchi, A., Ohyama, T., Sakai, E., Nakamura, R., Watanabe, T., Baba, H. and Nakayama, K. (2003b) 'Adhesins encoded by the gingipain genes of Porphyromonas gingivalis are responsible for co-aggregation with Prevotella intermedia', Microbiology, 149(Pt 5), 1257-1264.

280

Kataoka, M., Li, H., Arakawa, S. and Kuramitsu, H. (1997) 'Characterization of a methyl- accepting chemotaxis protein gene, dmcA, from the oral spirochete Treponema denticola', Infection and Immunity, 65(10), 4011-4016.

Kato, T., Kawai, S., Nakano, K., Inaba, H., Kuboniwa, M., Nakagawa, I., Tsuda, K., Omori, H., Ooshima, T., Yoshimori, T. and Amano, A. (2007) 'Virulence of Porphyromonas gingivalis is altered by substitution of fimbria gene with different genotype', Cellular Microbiology, 9(3), 753-765.

Kelley, L. A., Mezulis, S., Yates, C. M., Wass, M. N. and Sternberg, M. J. E. (2015) 'The Phyre2 web portal for protein modeling, prediction and analysis', Nature Protocols, 10(6), 845-858.

Kesavalu, L., Ebersole, J. L., Machen, R. L. and Holt, S. C. (1992) 'Porphyromonas gingivalis virulence in mice: induction of immunity to bacterial components', Infection and Immunity, 60(4), 1455-1464.

Kesavalu, L., Holt, S. C. and Ebersole, J. L. (1998) 'Virulence of a polymicrobic complex, Treponema denticola and Porphyromonas gingivalis, in a murine model', Oral Microbiology and Immunology, 13(6), 373-377.

Kesavalu, L., Sathishkumar, S., Bakthavatchalu, V., Matthews, C., Dawson, D., Steffen, M. and Ebersole, J. L. (2007) 'Rat model of polymicrobial infection, immunity, and alveolar bone resorption in periodontal disease', Infection and Immunity, 75(4), 1704-1712.

Kibbe, W. A. (2007) 'OligoCalc: an online oligonucleotide properties calculator', Nucleic Acids Research, 35(Web Server issue), W43-46.

Kigure, T., Saito, A., Seida, K., Yamada, S., Ishihara, K. and Okuda, K. (1995) 'Distribution of Porphyromonas gingivalis and Treponema denticola in human subgingival plaque at different periodontal pocket depths examined by immunohistochemical methods', Journal of Periodontal Research, 30(5), 332- 341.

Kirov, S. M., Castrisios, M. and Shaw, J. G. (2004) 'Aeromonas flagella (polar and lateral) are enterocyte adhesins that contribute to biofilm formation on surfaces', Infection and Immunity, 72(4), 1939-1945.

Klitorinos, A., Noble, P., Siboo, R. and Chan, E. C. (1993) 'Viscosity-dependent locomotion of oral spirochetes', Oral Microbiology and Immunology, 8(4), 242- 244.

281

Kojima, S. and Blair, D. F. (2001) 'Conformational change in the stator of the bacterial flagellar motor', Biochemistry, 40(43), 13041-13050.

Kojima, S., Imada, K., Sakuma, M., Sudo, Y., Kojima, C., Minamino, T., Homma, M. and Namba, K. (2009) 'Stator assembly and activation mechanism of the flagellar motor by the periplasmic region of MotB', Molecular Microbiology, 73(4), 710- 718.

Kolenbrander, P. E., Ganeshkumar, N., Cassels, F. J. and Hughes, C. V. (1993) 'Coaggregation: specific adherence among human oral plaque bacteria', The FASEB Journal, 7(5), 406-413.

Koskiniemi, S., Lamoureux, J. G., Nikolakakis, K. C., t'Kint de Roodenbeke, C., Kaplan, M. D., Low, D. A. and Hayes, C. S. (2013) 'Rhs proteins from diverse bacteria mediate intercellular competition', Proceedings of the National Academy of Sciences of the United States of America, 110(17), 7032-7037.

Kowalczykowski, S. C., Dixon, D. A., Eggleston, A. K., Lauder, S. D. and Rehrauer, W. M. (1994) 'Biochemistry of homologous recombination in Escherichia coli', Microbiological Reviews, 58(3), 401-465.

Kruskal, W. H. and Wallis, W. A. (1952) 'Use of Ranks in One-Criterion Variance Analysis', Journal of the American Statistical Association, 47(260), 583-621.

Kuboniwa, M., Amano, A., Hashino, E., Yamamoto, Y., Inaba, H., Hamada, N., Nakayama, K., Tribble, G. D., Lamont, R. J. and Shizukuishi, S. (2009) 'Distinct roles of long/short fimbriae and gingipains in homotypic biofilm development by Porphyromonas gingivalis', BMC Microbiology, 9, 105-105.

Kulp, A. and Kuehn, M. J. (2010) 'Biological functions and biogenesis of secreted bacterial outer membrane vesicles', Annual Review of Microbiology, 64, 163-184.

Kumada, H., Haishima, Y., Umemoto, T. and Tanamoto, K. (1995) 'Structural study on the free lipid A isolated from lipopolysaccharide of Porphyromonas gingivalis', Journal of Bacteriology, 177(8), 2098-2106.

Kung, S. H., Retchless, A. C., Kwan, J. Y. and Almeida, R. P. P. (2013) 'Effects of DNA size on transformation and recombination efficiencies in Xylella fastidiosa', Applied and Environmental Microbiology, 79(5), 1712-1717.

Kuramitsu, H. K., Chen, W. and Ikegami, A. (2005a) 'Biofilm formation by the periodontopathic bacteria Treponema denticola and Porphyromonas gingivalis', Journal of Periodontology, 76(11 Suppl), 2047-2051.

282

Kuramitsu, H. K., Chi, B. and Ikegami, A. (2005b) 'Genetic manipulation of Treponema denticola', Current Protocols in Microbiology, Chapter 12, Unit 12B.2.

Kuramitsu, H. K., He, X., Lux, R., Anderson, M. H. and Shi, W. (2007) 'Interspecies interactions within oral microbial communities', Microbiology and Molecular Biology Reviews, 71(4), 653-670.

Lai, Y. and Chu, L. (2008) 'Novel mechanism for conditional aerobic growth of the anaerobic bacterium Treponema denticola', Applied and Environmental Microbiology, 74(1), 73-79.

Lam, R. S., O'Brien-Simpson, N. M., Lenzo, J. C., Holden, J. A., Brammar, G. C., Walsh, K. A., McNaughtan, J. E., Rowler, D. K., Van Rooijen, N. and Reynolds, E. C. (2014) 'Macrophage depletion abates Porphyromonas gingivalis-induced alveolar bone resorption in mice', Journal of Immunology, 193(5), 2349-2362.

Lamont, R. J., El-Sabaeny, A., Park, Y., Cook, G. S., Costerton, J. W. and Demuth, D. R. (2002) 'Role of the Streptococcus gordonii SspB protein in the development of Porphyromonas gingivalis biofilms on streptococcal substrates', Microbiology, 148(Pt 6), 1627-1636.

Lamont, R. J. and Jenkinson, H. F. (1998) 'Life below the gum line: pathogenic mechanisms of Porphyromonas gingivalis', Microbiology and Molecular Biology Reviews, 62(4), 1244-1263.

Lane, M. C., Lockatell, V., Monterosso, G., Lamphier, D., Weinert, J., Hebel, J. R., Johnson, D. E. and Mobley, H. L. T. (2005) 'Role of motility in the colonization of uropathogenic Escherichia coli in the urinary tract', Infection and Immunity, 73(11), 7644-7656.

Le Gall, A., Cattoni, D. I., Guilhas, B., Mathieu-Demaziere, C., Oudjedi, L., Fiche, J. B., Rech, J., Abrahamsson, S., Murray, H., Bouet, J. Y. and Nollmann, M. (2016) 'Bacterial partition complexes segregate within the volume of the nucleoid', Nature Communications, 7, 12107.

Lee, S. Y., Bian, X. L., Wong, G. W. K., Hannam, P. M., McBride, B. C. and Fenno, J. C. (2002a) 'Cleavage of Treponema denticola PrcA Polypeptide To Yield Protease Complex-Associated Proteins Prca1 and Prca2 Is Dependent on PrtP', Journal of Bacteriology, 184(14), 3864-3870.

Lee, S. Y., Ning, Y. and Fenno, J. C. (2002b) '23S rRNA point mutation associated with erythromycin resistance in Treponema denticola', FEMS Microbiology Letters, 207(1), 39-42.

283

Lembre, P., Lorentz, C. and Martino, P. D. (2012) 'Exopolysaccharides of the biofilm matrix: a complex biophysical world' in Karunaratne, D. N., ed., The Complex World of Polysaccharides, Rijeka: IntechOpen, Ch. 13.

Li, H., Arakawa, S., Deng, Q. D. and Kuramitsu, H. (1999) 'Characterization of a novel methyl-accepting chemotaxis gene, dmcB, from the oral spirochete Treponema denticola', Infection and Immunity, 67(2), 694-699.

Li, H. and Kuramitsu, H. K. (1996) 'Development of a gene transfer system in Treponema denticola by electroporation', Oral Microbiology and Immunology, 11(3), 161- 165.

Li, H., Ruby, J., Charon, N. and Kuramitsu, H. (1996) 'Gene inactivation in the oral spirochete Treponema denticola: construction of an flgE mutant', Journal of Bacteriology, 178(12), 3664-3667.

Li, N., Kojima, S. and Homma, M. (2011) 'Sodium-driven motor of the polar flagellum in marine bacteria Vibrio', Genes Cells, 16(10), 985-999.

Li, W., Cowley, A., Uludag, M., Gur, T., McWilliam, H., Squizzato, S., Park, Y. M., Buso, N. and Lopez, R. (2015) 'The EMBL-EBI bioinformatics web and programmatic tools framework', Nucleic Acids Research, 43(W1), W580-584.

Li, Y., Ruby, J. and Wu, H. (2015) 'Kanamycin resistance cassette for genetic manipulation of Treponema denticola', Applied and Environmental Microbiology, 81(13), 4329-4338.

Li, Y. H. and Tian, X. (2012) 'Quorum sensing and bacterial social interactions in biofilms', Sensors (Basel), 12(3), 2519-2538.

Limberger, R. J. (2004) 'The periplasmic flagellum of spirochetes', Journal of Molecular Microbiology and Biotechnology, 7(1-2), 30-40.

Limberger, R. J., Slivienski, L. L., Izard, J. and Samsonoff, W. A. (1999) 'Insertional inactivation of Treponema denticola tap1 results in a nonmotile mutant with elongated flagellar hooks', Journal of Bacteriology, 181(12), 3743-3750.

Lin, X., Wu, J. and Xie, H. (2006) 'Porphyromonas gingivalis minor fimbriae are required for cell-cell interactions', Infection and Immunity, 74(10), 6011-6015.

Linden, G. J., Lyons, A. and Scannapieco, F. A. (2013) 'Periodontal systemic associations: review of the evidence', Journal of Periodontology, 84(4 Suppl), S8- S19.

284

Ling, J., Reynolds, N. and Ibba, M. (2009) 'Aminoacyl-tRNA synthesis and translational quality control', Annual Review of Microbiology, 63, 61-78.

Liu, I. F., Sutherland, J. H., Cheng, B. and Tse-Dinh, Y.-C. (2011) 'Topoisomerase I function during Escherichia coli response to antibiotics and stress enhances cell killing from stabilization of its cleavage complex', Journal of Antimicrobial Chemotherapy, 66(7), 1518-1524.

Lobo, L. A., Smith, C. J. and Rocha, E. R. (2011) 'Flavin mononucleotide (FMN)-based fluorescent protein (FbFP) as reporter for gene expression in the anaerobe Bacteroides fragilis', FEMS Microbiology Letters, 317(1), 67-74.

Lofblom, J., Kronqvist, N., Uhlen, M., Stahl, S. and Wernerus, H. (2007) 'Optimization of electroporation-mediated transformation: Staphylococcus carnosus as model organism', Journal of Applied Microbiology, 102(3), 736-747.

Lux, R., Miller, J. N., Park, N. H. and Shi, W. (2001) 'Motility and chemotaxis in tissue penetration of oral epithelial cell layers by Treponema denticola', Infection and Immunity, 69(10), 6276-6283.

Lux, R., Sim, J. H., Tsai, J. P. and Shi, W. (2002) 'Construction and characterization of a cheA mutant of Treponema denticola', Journal of Bacteriology, 184(11), 3130- 3134.

MacDougall, J., Margarita, D. and Saint Girons, I. (1992) 'Homology of a plasmid from the spirochete Treponema denticola with the single-stranded DNA plasmids', Journal of Bacteriology, 174(8), 2724-2728.

Madsen, J. S., Burmolle, M., Hansen, L. H. and Sorensen, S. J. (2012) 'The interconnection between biofilm formation and horizontal gene transfer', FEMS Immunology and Medical Microbiology, 65(2), 183-195.

Maggini, J., Raiden, S., Salamone, G., Trevani, A. and Geffner, J. (2009) 'Regulation of neutrophil apoptosis by cytokines, pathogens and environmental stressors', Frontiers in Bioscience (Landmark Edition), 14, 2372-2385.

Mahalakshmi, K., Krishnan, P. and Arumugam, S. B. (2017) '"Association of periodontopathic anaerobic bacterial co-occurrence to atherosclerosis" - A cross- sectional study', Anaerobe, 44, 66-72.

Malek, R., Fisher, J. G., Caleca, A., Stinson, M., van Oss, C. J., Lee, J. Y., Cho, M. I., Genco, R. J., Evans, R. T. and Dyer, D. W. (1994) 'Inactivation of the

285

Porphyromonas gingivalis fimA gene blocks periodontal damage in gnotobiotic rats', Journal of Bacteriology, 176(4), 1052-1059.

McAlister, A. D., Sroka, A., Fitzpatrick, R. E., Quinsey, N. S., Travis, J., Potempa, J. and Pike, R. N. (2009) 'Gingipain enzymes from Porphyromonas gingivalis preferentially bind immobilized extracellular proteins: a mechanism favouring colonization?', Journal of Periodontal Research, 44(3), 348-353.

McWilliam, H., Li, W., Uludag, M., Squizzato, S., Park, Y. M., Buso, N., Cowley, A. P. and Lopez, R. (2013) 'Analysis tool web services from the EMBL-EBI', Nucleic Acids Research, 41(Web Server issue), W597-600.

Medvedev, A. E., Sabroe, I., Hasday, J. D. and Vogel, S. N. (2006) 'Tolerance to microbial TLR ligands: molecular mechanisms and relevance to disease', Journal of Endotoxin Research, 12(3), 133-150.

Merritt, J. H., Kadouri, D. E. and O'Toole, G. A. (2005) 'Growing and analyzing static biofilms', Current Protocols in Microbiology, Chapter 1, Unit 1B.1.

Merritt, P. M., Danhorn, T. and Fuqua, C. (2007) 'Motility and chemotaxis in Agrobacterium tumefaciens surface attachment and biofilm formation', Journal of Bacteriology, 189(22), 8005-8014.

Mikolajczyk-Pawlinska, J., Travis, J. and Potempa, J. (1998) 'Modulation of interleukin- 8 activity by gingipains from Porphyromonas gingivalis: implications for pathogenicity of periodontal disease', FEBS Letters, 440(3), 282-286.

Milner, P., Batten, J. E. and Curtis, M. A. (1996) 'Development of a simple chemically defined medium for Porphyromonas gingivalis: requirement for alpha- ketoglutarate', FEMS Microbiology Letters, 140(2-3), 125-130.

Miyamoto, M., Ishihara, K. and Okuda, K. (2006) 'The Treponema denticola surface protease dentilisin degrades interleukin-1 beta (IL-1 beta), IL-6, and tumor necrosis factor alpha', Infection and Immunity, 74(4), 2462-2467.

Morimoto, Y. V., Che, Y. S., Minamino, T. and Namba, K. (2010a) 'Proton-conductivity assay of plugged and unplugged MotA/B proton channel by cytoplasmic pHluorin expressed in Salmonella', FEBS Letters, 584(6), 1268-1272.

Morimoto, Y. V. and Minamino, T. (2014) 'Structure and function of the bi-directional bacterial flagellar motor', Biomolecules, 4(1), 217-234.

Morimoto, Y. V., Nakamura, S., Kami-ike, N., Namba, K. and Minamino, T. (2010b) 'Charged residues in the cytoplasmic loop of MotA are required for stator

286

assembly into the bacterial flagellar motor', Molecular Microbiology, 78(5), 1117- 1129.

Motaleb, M. A., Corum, L., Bono, J. L., Elias, A. F., Rosa, P., Samuels, D. S. and Charon, N. W. (2000) 'Borrelia burgdorferi periplasmic flagella have both skeletal and motility functions', Proceedings of the National Academy of Sciences of the United States of America, 97(20), 10899-10904.

Mulkidjanian, A. Y., Dibrov, P. and Galperin, M. Y. (2008) 'The past and present of the sodium energetics: May the sodium-motive force be with you', Biochimica et Biophysica Acta, 1777(7-8), 985-992.

Nagano, K., Hasegawa, Y., Yoshida, Y. and Yoshimura, F. (2017) 'Novel fimbrilin PGN_1808 in Porphyromonas gingivalis', PLOS One, 12(3), e0173541.

Nair, B. C., Mayberry, W. R., Dziak, R., Chen, P. B., Levine, M. J. and Hausmann, E. (1983) 'Biological effects of a purified lipopolysaccharide from Bacteroides gingivalis', Journal of Periodontal Research, 18(1), 40-49.

Nakao, R., Hasegawa, H., Ochiai, K., Takashiba, S., Ainai, A., Ohnishi, M., Watanabe, H. and Senpuku, H. (2011) 'Outer membrane vesicles of Porphyromonas gingivalis elicit a mucosal immune response', PLOS One, 6(10), e26163.

Nakao, R., Takashiba, S., Kosono, S., Yoshida, M., Watanabe, H., Ohnishi, M. and Senpuku, H. (2014) 'Effect of Porphyromonas gingivalis outer membrane vesicles on gingipain-mediated detachment of cultured oral epithelial cells and immune responses', Microbes and Infection, 16(1), 6-16.

Nakayama, K., Yoshimura, F., Kadowaki, T. and Yamamoto, K. (1996) 'Involvement of arginine-specific cysteine proteinase (Arg-gingipain) in fimbriation of Porphyromonas gingivalis', Journal of Bacteriology, 178(10), 2818-2824.

Nelms, B. L. and Labosky, P. A. (2011) 'A predicted hairpin cluster correlates with barriers to PCR, sequencing and possibly BAC recombineering', Scientific Reports, 1, 106.

Ng, H. M., Kin, L. X., Dashper, S. G., Slakeski, N., Butler, C. A. and Reynolds, E. C. (2016) 'Bacterial interactions in pathogenic subgingival plaque', Microbial Pathogenesis, 94, 60-69.

Nilius, A. M., Spencer, S. C. and Simonson, L. G. (1993) 'Stimulation of in vitro growth of Treponema denticola by extracellular growth factors produced by Porphyromonas gingivalis', Journal of Dental Research, 72(6), 1027-1031.

287

Nussbaum, G., Ben-Adi, S., Genzler, T., Sela, M. and Rosen, G. (2009) 'Involvement of Toll-like receptors 2 and 4 in the innate immune response to Treponema denticola and its outer sheath components', Infection and Immunity, 77(9), 3939-3947.

O'Brien-Simpson, N. M., Pathirana, R. D., Paolini, R. A., Chen, Y. Y., Veith, P. D., Tam, V., Ally, N., Pike, R. N. and Reynolds, E. C. (2005) 'An immune response directed to proteinase and adhesin functional epitopes protects against Porphyromonas gingivalis-induced periodontal bone loss', Journal of Immunology, 175(6), 3980- 3989.

O’Brien-Simpson, N. M., Veith, P. D., Dashper, S. G. and Reynolds, E. C. (2003) 'Porphyromonas gingivalis gingipains: the molecular teeth of a microbial vampire', Current Protein & Peptide Science, 4(6), 409-426.

Ogawa, T. (1993) 'Chemical structure of lipid A from Porphyromonas (Bacteroides) gingivalis lipopolysaccharide', FEBS Letters, 332(1-2), 197-201.

Ogawa, T., Ogo, H., Uchida, H. and Hamada, S. (1994a) 'Humoral and cellular immune responses to the fimbriae of Porphyromonas gingivalis and their synthetic peptides', Journal of Medical Microbiology, 40(6), 397-402.

Ogawa, T., Uchida, H. and Hamada, S. (1994b) 'Porphyromonas gingivalis fimbriae and their synthetic peptides induce proinflammatory cytokines in human peripheral blood monocyte cultures', FEMS Microbiology Letters, 116(2), 237-242.

Okuda, T., Kimizuka, R., Miyamoto, M., Kato, T., Yamada, S., Okuda, K. and Ishihara, K. (2007) 'Treponema denticola induces interleukin-8 and macrophage chemoattractant protein 1 production in human umbilical vein epithelial cells', Microbes and Infection, 9(7), 907-913.

Orth, R. K. H., O'Brien-Simpson, N. M., Dashper, S. G. and Reynolds, E. C. (2011) 'Synergistic virulence of Porphyromonas gingivalis and Treponema denticola in a murine periodontitis model', Molecular Oral Microbiology, 26(4), 229-240.

Palzkill, T. (2013) 'Metallo-β-lactamase structure and function', Annals of the New York Academy of Sciences, 1277, 91-104.

Paramonov, N., Rangarajan, M., Hashim, A., Gallagher, A., Aduse-Opoku, J., Slaney, J. M., Hounsell, E. and Curtis, M. A. (2005) 'Structural analysis of a novel anionic polysaccharide from Porphyromonas gingivalis strain W50 related to Arg- gingipain glycans', Molecular Microbiology, 58(3), 847-863.

288

Park, S. Y., Chao, X., Gonzalez-Bonet, G., Beel, B. D., Bilwes, A. M. and Crane, B. R. (2004) 'Structure and function of an unusual family of protein phosphatases: the bacterial chemotaxis proteins CheC and CheX', Molecular Cell, 16(4), 563-574.

Paster, B. J. and Dewhirst, F. E. (2006) 'The phylogenetic diversity of the genus Treponema' in Radolf, J. D. and Lukehart, S. A., eds., Pathogenic Treponema: molecular and cellular biology, Wymondham, Norfolk, UK: Caister Academic Press, 9-18.

Pathirana, R. D., O'Brien-Simpson, N. M., Veith, P. D., Riley, P. F. and Reynolds, E. C. (2006) 'Characterization of proteinase-adhesin complexes of Porphyromonas gingivalis', Microbiology, 152(Pt 8), 2381-2394.

Peters, B. M., Jabra-Rizk, M. A., O'May, G. A., Costerton, J. W. and Shirtliff, M. E. (2012) 'Polymicrobial interactions: impact on pathogenesis and human disease', Clinical Microbiology Reviews, 25(1), 193-213.

Petersen, T. N., Brunak, S., von Heijne, G. and Nielsen, H. (2011) 'SignalP 4.0: discriminating signal peptides from transmembrane regions', Nature Methods, 8(10), 785-786.

Piekarski, T., Buchholz, I., Drepper, T., Schobert, M., Wagner-Doebler, I., Tielen, P. and Jahn, D. (2009) 'Genetic tools for the investigation of Roseobacter clade bacteria', BMC Microbiology, 9, 265.

Poole, K. (2012) 'Bacterial stress responses as determinants of antimicrobial resistance', Journal of Antimicrobial Chemotherapy, 67(9), 2069-2089.

Potempa, J., Pike, R. and Travis, J. (1997) 'Titration and mapping of the active site of cysteine proteinases from Porphyromonas gingivalis (gingipains) using peptidyl chloromethanes', Biological Chemistry, 378(3-4), 223-230.

Potempa, M., Potempa, J., Okroj, M., Popadiak, K., Eick, S., Nguyen, K. A., Riesbeck, K. and Blom, A. M. (2008) 'Binding of complement inhibitor C4b-binding protein contributes to serum resistance of Porphyromonas gingivalis', Journal of Immunology, 181(8), 5537-5544.

Pratt, L. A. and Kolter, R. (1998) 'Genetic analysis of Escherichia coli biofilm formation: roles of flagella, motility, chemotaxis and type I pili', Molecular Microbiology, 30(2), 285-293.

Ramsey, M. M., Rumbaugh, K. P. and Whiteley, M. (2011) 'Metabolite cross-feeding enhances virulence in a model polymicrobial infection', PLOS Pathogens, 7(3), e1002012.

289

Rangarajan, M., Aduse-Opoku, J., Paramonov, N., Hashim, A., Bostanci, N., Fraser, O. P., Tarelli, E. and Curtis, M. A. (2008) 'Identification of a second lipopolysaccharide in Porphyromonas gingivalis W50', Journal of Bacteriology, 190(8), 2920-2932.

Reife, R. A., Coats, S. R., Al-Qutub, M., Dixon, D. M., Braham, P. A., Billharz, R. J., Howald, W. N. and Darveau, R. P. (2006) 'Porphyromonas gingivalis lipopolysaccharide lipid A heterogeneity: differential activities of tetra- and penta-acylated lipid A structures on E-selectin expression and TLR4 recognition', Cellular Microbiology, 8(5), 857-868.

Rice, P., Longden, I. and Bleasby, A. (2000) 'EMBOSS: the European Molecular Biology Open Software Suite', Trends in Genetics, 16(6), 276-277.

Ringgaard, S., Schirner, K., Davis, B. M. and Waldor, M. K. (2011) 'A family of ParA- like ATPases promotes cell pole maturation by facilitating polar localization of chemotaxis proteins', Genes & Development, 25(14), 1544-1555.

Rosen, G., Genzler, T. and Sela, M. N. (2008) 'Coaggregation of Treponema denticola with Porphyromonas gingivalis and Fusobacterium nucleatum is mediated by the major outer sheath protein of Treponema denticola', FEMS Microbiology Letters, 289(1), 59-66.

Ruby, J. D. and Charon, N. W. (1998) 'Effect of temperature and viscosity on the motility of the spirochete Treponema denticola', FEMS Microbiology Letters, 169(2), 251- 254.

Ruby, J. D., Li, H., Kuramitsu, H., Norris, S. J., Goldstein, S. F., Buttle, K. F. and Charon, N. W. (1997) 'Relationship of Treponema denticola periplasmic flagella to irregular cell morphology', Journal of Bacteriology, 179(5), 1628-1635.

Ruby, J. D., Lux, R., Shi, W., Charon, N. W. and Dasanayake, A. (2008) 'Effect of glucose on Treponema denticola cell behavior', Molecular Oral Microbiology, 23(3), 234- 238.

Sal, M. S., Li, C., Motalab, M. A., Shibata, S., Aizawa, S. and Charon, N. W. (2008) 'Borrelia burgdorferi uniquely regulates its motility genes and has an intricate flagellar hook-basal body structure', Journal of Bacteriology, 190(6), 1912-1921.

Salvador, S. L., Syed, S. A. and Loesche, W. J. (1987) 'Comparison of three dispersion procedures for quantitative recovery of cultivable species of subgingival spirochetes', Journal of Clinical Microbiology, 25(11), 2230-2232.

290

Samuels, D. S., Mach, K. E. and Garon, C. F. (1994) 'Genetic transformation of the Lyme disease agent Borrelia burgdorferi with coumarin-resistant gyrB', Journal of Bacteriology, 176(19), 6045-6049.

Sander, A. F., Lavstsen, T., Rask, T. S., Lisby, M., Salanti, A., Fordyce, S. L., Jespersen, J. S., Carter, R., Deitsch, K. W., Theander, T. G., Pedersen, A. G. and Arnot, D. E. (2014) 'DNA secondary structures are associated with recombination in major Plasmodium falciparum variable surface antigen gene families', Nucleic Acids Research, 42(4), 2270-2281.

Sano, Y., Okamoto-Shibayama, K., Tanaka, K., Ito, R., Shintani, S., Yakushiji, M. and Ishihara, K. (2014) 'Dentilisin involvement in coaggregation between Treponema denticola and Tannerella forsythia', Anaerobe, 30, 45-50.

Sarkar, J., Frederick, J. and Marconi, R. T. (2010) 'The Hpk2-Rrp2 two-component regulatory system of Treponema denticola: a potential regulator of environmental and adaptive responses', Molecular Oral Microbiology, 25(4), 241-251.

Sauvage, E., Kerff, F., Terrak, M., Ayala, J. A. and Charlier, P. (2008) 'The penicillin- binding proteins: structure and role in peptidoglycan biosynthesis', FEMS Microbiology Reviews, 32(2), 234-258.

Scher, J. U. and Abramson, S. B. (2013) 'Periodontal disease, Porphyromonas gingivalis, and rheumatoid arthritis: what triggers autoimmunity and clinical disease?', Arthritis Research & Therapy, 15(5), 122-122.

Schonenbrucher, H., Adhikary, R., Mukherjee, P., Casey, T. A., Rasmussen, M. A., Maistrovich, F. D., Hamir, A. N., Kehrli, M. E., Jr., Richt, J. A. and Petrich, J. W. (2008) 'Fluorescence-based method, exploiting lipofuscin, for real-time detection of central nervous system tissues on bovine carcasses', Journal of Agricultural and Food Chemistry, 56(15), 6220-6226.

Seers, C. A., Slakeski, N., Veith, P. D., Nikolof, T., Chen, Y. Y., Dashper, S. G. and Reynolds, E. C. (2006) 'The RgpB C-terminal domain has a role in attachment of RgpB to the outer membrane and belongs to a novel C-terminal-domain family found in Porphyromonas gingivalis', Journal of Bacteriology, 188(17), 6376- 6386.

Seshadri, R., Myers, G. S., Tettelin, H., Eisen, J. A., Heidelberg, J. F., Dodson, R. J., Davidsen, T. M., DeBoy, R. T., Fouts, D. E., Haft, D. H., Selengut, J., Ren, Q., Brinkac, L. M., Madupu, R., Kolonay, J., Durkin, S. A., Daugherty, S. C., Shetty, J., Shvartsbeyn, A., Gebregeorgis, E., Geer, K., Tsegaye, G., Malek, J., Ayodeji, B., Shatsman, S., McLeod, M. P., Smajs, D., Howell, J. K., Pal, S., Amin, A., Vashisth, P., McNeill, T. Z., Xiang, Q., Sodergren, E., Baca, E., Weinstock, G. M., Norris, S. J., Fraser, C. M. and Paulsen, I. T. (2004) 'Comparison of the

291

genome of the oral pathogen Treponema denticola with other spirochete genomes', Proceedings of the National Academy of Sciences, 101(15), 5646-5651.

Sevlever, D., Jiang, P. and Yen, S. H. (2008) 'Cathepsin D is the main lysosomal enzyme involved in the degradation of alpha-synuclein and generation of its carboxy- terminally truncated species', Biochemistry, 47(36), 9678-9687.

Sharp, L., Poole, S., Reddi, K., Fletcher, J., Nair, S., Wilson, M., Curtis, M., Henderson, B. and Tabona, P. (1998) 'A lipid A-associated protein of Porphyromonas gingivalis, derived from the haemagglutinating domain of the RI protease gene family, is a potent stimulator of interleukin 6 synthesis', Microbiology, 144 ( Pt 11), 3019-3026.

Shelburne, C. E., Coopamah, M. D., Sweier, D. G., An, F. Y. and Lopatin, D. E. (2007) 'HtpG, the Porphyromonas gingivalis HSP-90 homologue, induces the chemokine CXCL8 in human monocytic and microvascular vein endothelial cells', Cellular Microbiology, 9(6), 1611-1619.

Shi, Y., Ratnayake, D. B., Okamoto, K., Abe, N., Yamamoto, K. and Nakayama, K. (1999) 'Genetic analyses of proteolysis, hemoglobin binding, and hemagglutination of Porphyromonas gingivalis. Construction of mutants with a combination of rgpA, rgpB, kgp, and hagA', Journal of Biological Chemistry, 274(25), 17955-17960.

Shoji, M., Naito, M., Yukitake, H., Sato, K., Sakai, E., Ohara, N. and Nakayama, K. (2004) 'The major structural components of two cell surface filaments of Porphyromonas gingivalis are matured through lipoprotein precursors', Molecular Microbiology, 52(5), 1513-1525.

Sim, J. H., Shi, W. and Lux, R. (2005) 'Protein-protein interactions in the chemotaxis signalling pathway of Treponema denticola', Microbiology, 151(Pt 6), 1801-1807.

Slakeski, N., Bhogal, P. S., O'Brien-Simpson, N. M. and Reynolds, E. C. (1998) 'Characterization of a second cell-associated Arg-specific cysteine proteinase of Porphyromonas gingivalis and identification of an adhesin-binding motif involved in association of the prtR and prtK proteinases and adhesins into large complexes', Microbiology, 144 ( Pt 6), 1583-1592.

Slakeski, N., Seers, C. A., Ng, K., Moore, C., Cleal, S. M., Veith, P. D., Lo, A. W. and Reynolds, E. C. (2011) 'C-terminal domain residues important for secretion and attachment of RgpB in Porphyromonas gingivalis', Journal of Bacteriology, 193(1), 132-142.

292

Slamti, L., de Pedro, M. A., Guichet, E. and Picardeau, M. (2011) 'Deciphering morphological determinants of the helix-shaped Leptospira', Journal of Bacteriology, 193(22), 6266-6275.

Slivienski-Gebhardt, L. L., Izard, J., Samsonoff, W. A. and Limberger, R. J. (2004) 'Development of a novel chloramphenicol resistance expression plasmid used for genetic complementation of a fliG deletion mutant in Treponema denticola', Infection and Immunity, 72(9), 5493-5497.

Socransky, S. S., Haffajee, A. D., Cugini, M. A., Smith, C. and Kent, R. L., Jr. (1998) 'Microbial complexes in subgingival plaque', Journal of Clinical Periodontology, 25(2), 134-144.

Socransky, S. S., Haffajee, A. D., Ximenez-Fyvie, L. A., Feres, M. and Mager, D. (1999) 'Ecological considerations in the treatment of Actinobacillus actinomycetemcomitans and Porphyromonas gingivalis periodontal infections', Periodontology 2000, 20, 341-362.

Soding, J., Biegert, A. and Lupas, A. N. (2005) 'The HHpred interactive server for protein homology detection and structure prediction', Nucleic Acids Research, 33(Web Server issue), W244-248.

Srisatjaluk, R., Doyle, R. J. and Justus, D. E. (1999) 'Outer membrane vesicles of Porphyromonas gingivalis inhibit IFN-gamma-mediated MHC class II expression by human vascular endothelial cells', Microbial Pathogenesis, 27(2), 81-91.

Stamm, L. V. and Bergen, H. L. (1999) 'Molecular characterization of a flagellar (fla) operon in the oral spirochete Treponema denticola ATCC 35405', FEMS Microbiology Letters, 179(1), 31-36.

Stewart, R. C. (1993) 'Activating and inhibitory mutations in the regulatory domain of CheB, the methylesterase in bacterial chemotaxis', Journal of Biological Chemistry, 268(3), 1921-1930.

Strahl, H. and Hamoen, L. W. (2010) 'Membrane potential is important for bacterial cell division', Proceedings of the National Academy of Sciences of the United States of America, 107(27), 12281-12286.

Sultan, S. Z., Manne, A., Stewart, P. E., Bestor, A., Rosa, P. A., Charon, N. W. and Motaleb, M. A. (2013) 'Motility is crucial for the infectious life cycle of Borrelia burgdorferi', Infection and Immunity, 81(6), 2012-2021.

Sultan, S. Z., Sekar, P., Zhao, X., Manne, A., Liu, J., Wooten, R. M. and Motaleb, M. A. (2015) 'Motor rotation is essential for the formation of the periplasmic flagellar

293

ribbon, cellular morphology, and Borrelia burgdorferi persistence within Ixodes scapularis tick and murine hosts', Infection and Immunity, 83(5), 1765-1777.

Takada, H., Hirai, H., Fujiwara, T., Koga, T., Ogawa, T. and Hamada, S. (1990) 'Bacteroides lipopolysaccharides (LPS) induce anaphylactoid and lethal reactions in LPS-responsive and -nonresponsive mice primed with muramyl dipeptide', Journal of Infectious Diseases, 162(2), 428-434.

Takada, H. and Kotani, S. (1992) 'Structure-function relationships of lipid A' in Morrison, D. and Ryan, J., eds., Bacterial endotoxic lipopolysaccharides, New York: CRC Press, 107-134.

Takahashi, N. (2003) 'Acid-neutralizing activity during amino acid fermentation by Porphyromonas gingivalis, Prevotella intermedia and Fusobacterium nucleatum', Oral Microbiology and Immunology, 18(2), 109-113.

Takeda, K. and Akira, S. (2007) 'Toll-like receptors', Current Protocols in Immunology, Chapter 14, Unit 14 12.

Tan, K. H., Seers, C. A., Dashper, S. G., Mitchell, H. L., Pyke, J. S., Meuric, V., Slakeski, N., Cleal, S. M., Chambers, J. L., McConville, M. J. and Reynolds, E. C. (2014) 'Porphyromonas gingivalis and Treponema denticola exhibit metabolic symbioses', PLOS Pathogens, 10(3), e1003955.

Tanaka-Kumazawa, K., Kikuchi, Y., Sano-Kokubun, Y., Shintani, S., Yakushiji, M., Kuramitsu, H. K. and Ishihara, K. (2017) 'Characterization of a potential ABC- type bacteriocin exporter protein from Treponema denticola', BMC Oral Health, 17, 18.

Tardif, F., Ross, G. and Rouabhia, M. (2004) 'Gingival and dermal fibroblasts produce interleukin-1 beta converting enzyme and interleukin-1 beta but not interleukin- 18 even after stimulation with lipopolysaccharide', Journal of Cellular Physiology, 198(1), 125-132.

Tatusov, R. L., Fedorova, N. D., Jackson, J. D., Jacobs, A. R., Kiryutin, B., Koonin, E. V., Krylov, D. M., Mazumder, R., Mekhedov, S. L., Nikolskaya, A. N., Rao, B. S., Smirnov, S., Sverdlov, A. V., Vasudevan, S., Wolf, Y. I., Yin, J. J. and Natale, D. A. (2003) 'The COG database: an updated version includes eukaryotes', BMC Bioinformatics, 4, 41.

Tatusov, R. L., Koonin, E. V. and Lipman, D. J. (1997) 'A genomic perspective on protein families', Science, 278(5338), 631-637.

294

Taylor, J. D., Taylor, G., Hare, S. A. and Matthews, S. J. (2016) 'Structures of the DfsB protein family suggest a cationic, helical sibling lethal factor peptide', Journal of Molecular Biology, 428(3), 554-560. ter Huurne, A. A., van Houten, M., Muir, S., Kusters, J. G., van der Zeijst, B. A. and Gaastra, W. (1992) 'Inactivation of a Serpula (Treponema) hyodysenteriae hemolysin gene by homologous recombination: importance of this hemolysin in pathogenesis in mice', FEMS Microbiology Letters, 71(1), 109-113.

Terahara, N., Krulwich, T. A. and Ito, M. (2008) 'Mutations alter the sodium versus proton use of a Bacillus clausii flagellar motor and confer dual ion use on motors ', Proceedings of the National Academy of Sciences, 105(38), 14359-14364.

Tokuda, M., Duncan, M., Cho, M. I. and Kuramitsu, H. K. (1996) 'Role of Porphyromonas gingivalis protease activity in colonization of oral surfaces', Infection and Immunity, 64(10), 4067-4073.

Tonetti, M. S., Van Dyke, T. E. and working group 1 of the joint EFP/AAP workshop (2013) 'Periodontitis and atherosclerotic cardiovascular disease: consensus report of the Joint EFP/AAP Workshop on Periodontitis and Systemic Diseases', Journal of Periodontology, 84(4 Suppl), S24-29.

Trinchieri, G. (2003) 'Interleukin-12 and the regulation of innate resistance and adaptive immunity', Nature Reviews Immunology, 3(2), 133-146.

Tu, Q., Yin, J., Fu, J., Herrmann, J., Li, Y., Yin, Y., Stewart, A. F., Muller, R. and Zhang, Y. (2016) 'Room temperature electrocompetent bacterial cells improve DNA transformation and recombineering efficiency', Scientific Reports, 6, 24648.

Umemoto, T. and Hamada, N. (2003) 'Characterization of biologically active cell surface components of a periodontal pathogen. The roles of major and minor fimbriae of Porphyromonas gingivalis', Journal of Periodontology, 74(1), 119-122.

Umemoto, T., Jinno, T., Taiji, Y. and Ogawa, T. (2001) 'Chemotaxis of oral treponemes toward sera and albumin of rabbit', Microbiology and Immunology, 45(8), 571- 577.

Vanhommerig, E., Moons, P., Pirici, D., Lammens, C., Hernalsteens, J. P., De Greve, H., Kumar-Singh, S., Goossens, H. and Malhotra-Kumar, S. (2014) 'Comparison of biofilm formation between major clonal lineages of methicillin resistant Staphylococcus aureus', PLOS One, 9(8), e104561.

Veith, P. D., Chen, Y. Y., Gorasia, D. G., Chen, D., Glew, M. D., O'Brien-Simpson, N. M., Cecil, J. D., Holden, J. A. and Reynolds, E. C. (2014) 'Porphyromonas

295

gingivalis outer membrane vesicles exclusively contain outer membrane and periplasmic proteins and carry a cargo enriched with virulence factors', Journal of Proteome Research, 13(5), 2420-2432.

Veith, P. D., Dashper, S. G., O'Brien-Simpson, N. M., Paolini, R. A., Orth, R., Walsh, K. A. and Reynolds, E. C. (2009) 'Major proteins and antigens of Treponema denticola', Biochimica et Biophysica Acta, 1794(10), 1421-1432.

Veith, P. D., Talbo, G. H., Slakeski, N., Dashper, S. G., Moore, C., Paolini, R. A. and Reynolds, E. C. (2002) 'Major outer membrane proteins and proteolytic processing of RgpA and Kgp of Porphyromonas gingivalis W50', Biochemical Journal, 363(Pt 1), 105-115.

Vesey, P. M. and Kuramitsu, H. K. (2004) 'Genetic analysis of Treponema denticola ATCC 35405 biofilm formation', Microbiology, 150(Pt 7), 2401-2407.

Vester, B. and Douthwaite, S. (2001) 'Macrolide Resistance Conferred by Base Substitutions in 23S rRNA', Antimicrobial Agents and Chemotherapy, 45(1), 1- 12.

Vorregaard, M. (2008) Comstat2 - a modern 3D image analysis environment for biofilms, in Informatics and Mathematical Modelling, Technical University of Denmark: Kongens Lyngby, Denmark.

Wang, P. L. and Ohura, K. (2002) 'Porphyromonas gingivalis lipopolysaccharide signaling in gingival fibroblasts-CD14 and Toll-like receptors', Critical Reviews in Oral Biology & Medicine, 13(2), 132-142.

Wardle, H. M. (1997) 'The challenge of growing oral spirochaetes', Journal of Medical Microbiology, 46(2), 104-116.

Waterfield, N. R., Bowen, D. J., Fetherston, J. D., Perry, R. D. and ffrench-Constant, R. H. (2001) 'The tc genes of Photorhabdus: a growing family', Trends in Microbiology, 9(4), 185-191.

Weimer, K. E., Juneau, R. A., Murrah, K. A., Pang, B., Armbruster, C. E., Richardson, S. H. and Swords, W. E. (2011) 'Divergent mechanisms for passive pneumococcal resistance to beta-lactam antibiotics in the presence of Haemophilus influenzae', The Journal of Infectious Diseases, 203(4), 549-555.

Weinberg, A., Belton, C. M., Park, Y. and Lamont, R. J. (1997) 'Role of fimbriae in Porphyromonas gingivalis invasion of gingival epithelial cells', Infection and Immunity, 65(1), 313-316.

296

Wiebe, C. B. and Putnins, E. E. (2000) 'The periodontal disease classification system of the American Academy of Periodontology--an update', Journal of the Canadian Dental Association, 66(11), 594-597.

Wilkins, M. R., Gasteiger, E., Bairoch, A., Sanchez, J. C., Williams, K. L., Appel, R. D. and Hochstrasser, D. F. (1999) 'Protein identification and analysis tools in the ExPASy server', Methods in Molecular Biology, 112, 531-552.

Wolf, Y. I., Rogozin, I. B., Grishin, N. V., Tatusov, R. L. and Koonin, E. V. (2001) 'Genome trees constructed using five different approaches suggest new major bacterial clades', BMC Evolutionary Biology, 1, 8-8.

Yamada, M., Ikegami, A. and Kuramitsu, H. K. (2005) 'Synergistic biofilm formation by Treponema denticola and Porphyromonas gingivalis', FEMS Microbiology Letters, 250(2), 271-277.

Yoneda, M., Yoshikane, T., Motooka, N., Yamada, K., Hisama, K., Naito, T., Okada, I., Yoshinaga, M., Hidaka, K., Imaizumi, K., Maeda, K. and Hirofuji, T. (2005) 'Stimulation of growth of Porphyromonas gingivalis by cell extracts from Tannerella forsythia', Journal of Periodontal Research, 40(2), 105-109.

Youderian, P. and Hartzell, P. L. (2007) 'Triple Mutants Uncover Three New Genes Required for Social Motility in Myxococcus xanthus', Genetics, 177(1), 557-566.

Yu, N. Y., Wagner, J. R., Laird, M. R., Melli, G., Rey, S., Lo, R., Dao, P., Sahinalp, S. C., Ester, M., Foster, L. J. and Brinkman, F. S. (2010) 'PSORTb 3.0: improved protein subcellular localization prediction with refined localization subcategories and predictive capabilities for all prokaryotes', Bioinformatics, 26(13), 1608- 1615.

Yun, L. W., Decarlo, A. A., Collyer, C. and Hunter, N. (2003) 'Enhancement of Th2 pathways and direct activation of B cells by the gingipains of Porphyromonas gingivalis', Clinical and Experimental Immunology, 134(2), 295-302.

Yun, P. L., Decarlo, A. A., Collyer, C. and Hunter, N. (2001) 'Hydrolysis of interleukin- 12 by Porphyromonas gingivalis major cysteine proteinases may affect local gamma interferon accumulation and the Th1 or Th2 T-cell phenotype in periodontitis', Infection and Immunity, 69(9), 5650-5660.

Yun, P. L., DeCarlo, A. A. and Hunter, N. (1999) 'Modulation of major histocompatibility complex protein expression by human gamma interferon mediated by cysteine proteinase-adhesin polyproteins of Porphyromonas gingivalis', Infection and Immunity, 67(6), 2986-2995.

297

Zainal-Abidin, Z., Veith, P. D., Dashper, S. G., Zhu, Y., Catmull, D. V., Chen, Y. Y., Heryanto, D. C., Chen, D., Pyke, J. S., Tan, K., Mitchell, H. L. and Reynolds, E. C. (2012) 'Differential proteomic analysis of a polymicrobial biofilm', Journal of Proteome Research, 11(9), 4449-4464.

Zaric, S., Shelburne, C., Darveau, R., Quinn, D. J., Weldon, S., Taggart, C. C. and Coulter, W. A. (2010) 'Impaired immune tolerance to Porphyromonas gingivalis lipopolysaccharide promotes neutrophil migration and decreased apoptosis', Infection and Immunity, 78(10), 4151-4156.

Zhu, Y., Dashper, S. G., Chen, Y. Y., Crawford, S., Slakeski, N. and Reynolds, E. C. (2013) 'Porphyromonas gingivalis and Treponema denticola synergistic polymicrobial biofilm development', PLOS One, 8(8), e71727.

Zijnge, V., van Leeuwen, M. B., Degener, J. E., Abbas, F., Thurnheer, T., Gmur, R. and Harmsen, H. J. (2010) 'Oral biofilm architecture on natural teeth', PLOS One, 5(2), e9321.

Zilm, P. S., Mira, A., Bagley, C. J. and Rogers, A. H. (2010) 'Effect of alkaline growth pH on the expression of cell envelope proteins in Fusobacterium nucleatum', Microbiology, 156(Pt 6), 1783-1794.

298

APPENDICES

Appendix I

Table I.1 Chemicals, media, supplements and antibiotics suppliers. Reagent Supplier Acetone CHEM SUPPLY Agarose PROMEGA Ampicillin SIGMA ATP Bioline Bacto-Agar BD (Difco Laboratories) Calcium chloride SIGMA Ethanol 100% Undenatured CHEM SUPPLY Ethylenediaminetetraacetic acid disodium Univar salt (EDTA disodium salt) Ficoll SIGMA Glycerol CHEM SUPPLY Hydrochloric acid MERCK Pty. Ltd. Kanamycin SIGMA lysozyme SIGMA Potassium chloride CHEM SUPPLY Sodium chloride MERCK Pty. Ltd. Sodium hydroxide pellet CHEM SUPPLY SYBR® Safe gel stain Molecular Probes Inc. Tryptone OXOID Yeast Extract OXOID • BD (Difco Laboratories) Becton, Dickinson and Company, USA • Bioline (Aust) Pty. Ltd, Alexandria, NSW, Australia • CHEM SUPPLY, Australia • MERCK Pty. Ltd., Australia • Molecular Probes Inc., USA • OXOID Ltd., Basingstoke, Hampshire, England • PROMEGA, Promega Corporation, USA • SIGMA, Sigma-Aldrich Chemie, GmH, Germany • Univar, Ajax Finechem, Nuplex Industries (Aust) Pty. Ltd., Australia

299

Appendix II

Table II.1 Nucleotide polymorphisms of T. denticola ∆flgE33520 mutant determined by genomic sequencing. Locus Nucleotide Change Coverage Polymor Strand Variant Variant Amino Coding Codon Protein Protein Tag Position phism -Bias Frequency P- Acid region Change Effect Descripti Type (%) (%) Value Change (CDS) on (approx Position imate) HMP 350530 C -> G 38 SNP 52.60 100.00 2.50E- F -> L 75 TTC -> Substit- hypotheti REF9 (transver 103 TTG ution cal 722_ sion) protein RS01 685 HMP 660965 C -> G 24 SNP 50.00 100.00 2.50E- T -> R 128 ACG -> Substit hypotheti REF9 (transver 46 AGG ution cal 722_ sion) protein RS03 030 HMP 890944 +C 41 Insertion 51.20 100.00 1.30E- ANA 2,710 ANA Frame hypotheti REF9 119 Shift cal 722_ protein

300

RS03 985 HMP 926365 T -> G 48 SNP 50.00 100.00 4.00E- L -> V 157 TTA -> Substit hypotheti REF9 (transver 39 GTA ution cal 722_ sion) protein RS04 145 HMP 1139367 C -> T 41 SNP 58.50 100.00 2.50E- A -> V 332 GCT -> Substit GAF REF9 (transitio 66 GTT ution domain- 722_ n) containin RS05 g protein 155 HMP 1207840 T -> G 48 SNP 52.10 100.00 6.30E- L -> V 1,186 TTG -> Substit methylas REF9 (transver 140 GTG ution partate 722_ sion) ammonia- RS05 lyase 470 HMP 1222362 C -> A 33 SNP 54.50 100.00 2.00E- F -> L 1,014 TTC -> Substit glucose/g REF9 (transver 96 TTA ution alactose- 722_ sion) binding

301

RS05 lipoprotei 535 n HMP 1565026 C -> T 13 SNP 53.80 100.00 1.60E- S -> L 272 TCA -> Substit CPBP REF9 (transitio 21 TTA ution family 722_ n) intramem RS07 brane 065 metallopr otease, frameshif t mutation HMP 1838055 C -> T 23 SNP 56.50 100.00 2.00E- E -> K 214 GAG - Substit hypotheti REF9 (transitio 21 > AAG ution cal 722_ n) protein RS08 470 HMP 1950827 G -> A 37 SNP 56.80 100.00 4.00E- A -> T 952 GCA -> Substit HD-GYP REF9 (transitio 119 ACA ution domain- 722_ n) containin RS09 g protein 015

302

HMP 1950864 G -> T 33 SNP 57.60 100.00 2.00E- R -> L 989 CGT -> Substit HD-GYP REF9 (transver 63 CTT ution domain- 722_ sion) containin RS09 g protein 015 HMP 2330455 C -> T 46 SNP 56.50 100.00 1.60E- A -> T 1,969 GCT -> Substit hypotheti REF9 (transitio 129 ACT ution cal 722_ n) protein RS10 740 HMP 2446501 A -> T 24 SNP 50.00 100.00 4.00E- S -> T 127 TCC -> Substit hypotheti REF9 (transver 27 ACC ution cal 722_ sion) protein RS11 225 A NA: Not applicable.

303

Table II.2 Nucleotide polymorphisms of T. denticola ∆motA33520 mutant determined by genomic sequencing. Locus Nucleotide Change Coverage Polymor Strand Variant Variant Amino CDS Codon Protein Protein Tag Position phism -Bias Frequency P-Value Acid Position Change Effect Descript Type (%) (%) (approxi Change ion mate) HMP 660630 A -> G 45 SNP 60.00 100.00 1.00E- I -> V 1,210 ATT -> Substit flagellar REF9 (transitio 126 GTT ution hook 722_ n) protein RS03 FlgE 025 HMP 660841 G -> A 53 SNP 52.80 100.00 4.00E- V -> I 4 GTA -> Substit hypothet REF9 (transitio 149 ATA ution ical 722_ n) protein RS03 030 HMP 660965 C -> G 37 SNP 54.10 100.00 5.00E-71 T -> R 128 ACG -> Substit hypothet REF9 (transver AGG ution ical 722_ sion) protein RS03 030

304

HMP 661866 AA -> 29 Substitut 55.20 100.00 1.00E-87 K -> S 29 AAA - Substit flagellar REF9 GC ion > AGC ution motor 722_ protein RS03 MotB 040

305

Table II.3 Nucleotide polymorphisms of T. denticola ∆motB33520 mutant determined by genomic sequencing. Locus Nucleotide Change Coverage Polymor Strand Variant Variant Amino CDS Codon Protein Protein Tag Position phism -Bias Frequency P-Value Acid Position Change Effect Descript Type (%) (%) (approxi Change ion mate) HMP 350530 C -> G 32 SNP 53.10 100.00 6.30E-84 F -> L 75 TTC -> Substit hypothet REF9 (transver TTG ution ical 722_ sion) protein RS01 685 HMP 1565026 C -> T 11 SNP 54.50 100.00 3.20E-17 S -> L 272 TCA -> Substit CPBP REF9 (transitio TTA ution family 722_ n) intrame RS07 mbrane 065 metallop rotease, frameshi ft mutation

306

HMP 1862166 C -> T 25 SNP 56.00 100.00 1.00E-60 A -> T 304 GCT -> Substit RHS REF9 (transitio ACT ution repeat- 722_ n) associat RS08 ed core 580 domain- containi ng protein HMP 2446501 A -> T 27 SNP 55.60 100.00 2.00E-30 S -> T 127 TCC -> Substit hypothet REF9 (transver ACC ution ical 722_ sion) protein RS11 225

307

Appendix III

Table III.1 Proteins significantly changed in abundance in T. denticola ∆flgE33520 mutant relative to wild-type (ratio≥1.5 and ≤0.67, p<0.05). Proteins predicted to be organized in an operon were shaded.

C Locus Tag Protein description Wild-type ∆flgE33520 COG abundanceA ratioB TDE0051 alcohol dehydrogenase, iron-containing 24014333 1.88 C TDE0114 iron-dependent transcriptional regulator 34609000 0.55 K TDE0119 flagellar protein FliS (fliS) 1433833 2.71 N TDE0173 FtsK/SpoIIIE family protein 1914600 1.62 D TDE0174 nicotinate phosphoribosyltransferase, putative 28828000 1.53 H TDE0383 hypothetical protein 5615867 0.00 S TDE0423 hypothetical protein 18472000 0.53 - TDE0446 fibronectin type III domain protein 5530000 1.74 S TDE0455 kinase, GHMP family 2111867 2.10 G TDE0463 purine nucleoside phosphorylase (deoD) 53346667 0.62 F TDE0501 hypothetical protein 9163200 0.50 S TDE0533 hypothetical protein 7672000 0.44 S TDE0602 3-oxoacyl-(acyl-carrier-protein) synthase III (fabH) 5136167 0.45 I TDE0664 OmpA family protein 21103333 2.87 M TDE0689 5-methylthioribose kinase 1505633 0.00 S

308

TDE0758 iron compound ABC transporter, periplasmic iron compound-binding protein, 6827467 0.23 P putative TDE0777 ribosomal protein L14 (rplN) 93704333 0.66 J TDE0781 ribosomal protein S8 (rpsH) 46230333 0.65 J TDE0823 (3R)-hydroxymyristoyl-(acyl-carrier-protein) dehydratase, putative 37107000 0.56 I TDE0843 conserved hypothetical protein 187536667 0.65 S TDE0845 conserved hypothetical protein TIGR00266 61545333 0.62 S TDE0984 oligopeptide/dipeptide ABC transporter, permease protein, putative 6575533 0.00 P TDE0985 oligopeptide/dipeptide ABC transporter, periplasmic peptide-binding protein, 321453333 0.33 E putative TDE0986 oligopeptide/dipeptide ABC transporter, ATP-binding protein 3993533 0.00 P TDE1004 flagellar filament core protein 466846667 0.01 N TDE1208 DNA topoisomerase I (topA) 69289000 1.79 L TDE1211 heat shock protein HslVU, ATPase subunit HslU (hslU) 39877000 2.02 O TDE1234 hypothetical protein 283517 1.51 S TDE1318 hypothetical protein 725143 0.00 U TDE1408 flagellar filament outer layer protein FlaA, putative 576866667 0.04 N TDE1409 flagellar filament outer layer protein FlaA, putative 559370000 0.04 N TDE1475 flagellar filament core protein 44744667 0.00 N TDE1483 conserved hypothetical protein 127703333 0.30 S

309

TDE1521 hydro-lyase, tartrate/fumarate family, beta subunit 52919667 0.60 C TDE1640 3-dehydroquinate dehydratase, type 1, putative/shikimate 5-dehydrogenase, putative 5431933 1.59 E TDE1684 adenine phosphoribosyltransferase 7008600 0.67 F TDE1712 flagellar filament outer layer protein (flaA) 1851533333 0.02 N TDE1727 conserved hypothetical protein 102531333 0.46 O TDE1754 desulfoferrodoxin/neelaredoxin 73828000 0.54 C TDE1884 hypothetical protein 2661367 2.37 S TDE1916 glycerol kinase (glpK) 10331700 1.85 C TDE1917 cytidylyltransferase domain protein 2718300 1.88 M TDE1918 conserved hypothetical protein 2047400 1.58 S TDE1919 conserved domain protein 16423667 1.90 S TDE2030 lipoprotein, RlpA family 54538667 1.58 M TDE2043 signal recognition particle-docking protein FtsY (ftsY) 5551833 0.64 U TDE2048 conserved hypothetical protein 17770000 1.70 S TDE2085 amino acid kinase family protein 76823333 1.55 F TDE2087 translation initiation factor IF-1 (infA) 43004667 0.61 J TDE2130 hypothetical protein 21761000 0.00 - TDE2198 pyruvate-ferredoxin oxidoreductase 28485000 1.59 C TDE2302 HD domain protein 10887667 0.64 T TDE2353 flagellar hook-associated protein 3 1834867 0.59 N

310

TDE2565 hypothetical protein 2246400 0.00 S TDE2611 conserved hypothetical protein 2623300 1.52 S TDE2644 pyridine nucleotide-disulphide oxidoreductase family protein 416603 0.00 O TDE2693 ankyrin repeat protein 38558333 2.06 S TDE2721 helicase domain protein 3375000 0.67 L TDE2763 flagellar motor switch protein FliM (fliM) 14706667 0.46 N TDE2764 flagellar protein FliL (fliL) 65052333 0.50 N TDE2765 flagellar motor rotation protein B (motB) 34386667 0.35 N TDE2766 motility protein A (motA) 34589333 0.44 N TDE2768 flagellar hook protein FlgE (flgE) 43923333 0.00 N TDE2779 hypothetical protein 59837667 0.31 S A The abundance of each protein in the wild-type T. denticola ATCC 33520 was calculated from the average IBAQ intensity from three replicates.

B Geometric mean of ratios, from three replicates, produced from the LFQ intensity of protein in ∆flgE33520 relative to that of protein in wild-type. Ratio of ≥1.5 indicates that the protein had increased in abundance in ∆flgE33520 relative to wild-type and ratio of ≤0.67 indicates that the protein had decreased in abundance in ∆flgE33520 relative to wild-type. C One-letter abbreviations for the functional COG categories: J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosome partitioning; V, defense mechanisms; T, signal transduction mechanisms; M, cell wall/membrane/envelope biogenesis; N, cell motility; U, intracellular trafficking, secretion, and vesicular transport; O, posttranslational modification, protein turnover, chaperones; C, energy production and conversion; G, carbohydrate transport and metabolism; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; R, general function prediction only; S, function unknown.

311

Table III.2 Proteins significantly changed in abundance in T. denticola ∆motA33520 mutant relative to wild-type (ratio≥1.5 and ≤0.67, p<0.05). Proteins predicted to be organized in an operon were shaded. In the case where two or more operons were arranged consecutively, different darkness of shadings was used to differentiate the operons.

C Locus Tag Protein description Wild-type ∆motA33520 COG abundanceA ratioB TDE0008 copper-translocating P-type ATPase 1737167 0.00 P TDE0011 alkyl hydroperoxide reductase/peroxiredoxin 5732033333 0.37 O TDE0012 carbon starvation protein CstA, putative 461243333 0.27 T TDE0042 phosphate acetyltransferase (pta) 270816667 0.42 C TDE0046 formiminotransferase-cyclodeaminase family protein 187016667 0.65 E TDE0055 flagellar biosynthesis protein FlhA (flhA) 2858667 1.86 U TDE0065 conserved hypothetical protein 8750800 0.00 S TDE0066 conserved hypothetical protein 6329767 0.28 S TDE0081 hypothetical protein 1744067 0.00 S TDE0086 conserved hypothetical protein 1468300 0.63 L TDE0103 aminotransferase, class-V 1789900 0.00 E TDE0119 flagellar protein FliS (fliS) 1433833 0.00 U TDE0125 GGDEF domain protein 6712833 0.27 T TDE0129 PyrBI protein (pyrBI) 164263333 1.67 F TDE0149 DNA-binding response regulator 1913133 0.00 T

312

TDE0151 integral membrane protein, YeeE/YedE family 34126667 0.49 S TDE0168 transcriptional regulator, GntR family 8614867 0.00 K TDE0169 methyl-accepting chemotaxis domain protein 2943800 1.82 S TDE0175 pyrrolidone-carboxylate peptidase (pcp) 98789667 0.38 O TDE0181 methyl-accepting chemotaxis protein 8900967 1.57 T TDE0186 hypothetical protein 596683333 0.60 P TDE0198 conserved hypothetical protein 67631667 1.72 E TDE0200 tetrapyrrole methylase family protein 2478933 0.00 R TDE0205 adenylate/guanylate cyclase catalytic domain protein 56331333 1.60 T TDE0207 permease, GntP family 38581667 0.33 E TDE0212 conserved hypothetical protein 122642000 0.46 P TDE0213 TPR domain protein 3607700 0.65 U TDE0237 HDIG domain protein 26933333 1.89 S TDE0239 glycine reductase complex protein GrdD (grdD) 231270000 0.62 I TDE0249 flavoredoxin, putative 20076667 0.43 S TDE0296 formiminotransferase, putative 376556667 0.64 E TDE0300 cytosol aminopeptidase family protein 388300000 0.63 E TDE0301 tetracenomycin polyketide synthesis O-methyltransferase TcmP, putative 5128000 0.00 Q TDE0325 hypothetical protein 35163333 0.62 - TDE0337 glucosamine-6-phosphate isomerase (nagB) 49480667 0.35 G

313

TDE0341 ribonucleoside-diphosphate reductase, alpha subunit (nrdA) 2758667 1.89 F TDE0347 methyl-accepting chemotaxis protein DmcA (dmcA) 6901500 0.53 T TDE0354 general stress protein 14 18201333 0.34 R TDE0358 cinnamoyl ester hydrolase (cinI) 2591833 0.00 I TDE0386 ABC transporter, periplasmic substrate-binding protein 585636667 0.50 P TDE0389 (R)-2-hydroxyglutaryl-CoA dehydratase, beta subunit, putative 89464667 0.54 E TDE0400 transcriptional regulator, LysR family 6230933 1.63 K TDE0405 major outer sheath protein 3398466667 0.44 M TDE0418 lipoprotein, putative 30177333 0.29 - TDE0419 hypothetical protein 27165667 0.11 - TDE0423 hypothetical protein 18472000 0.25 - TDE0425 bacteriocin ABC transporter, ATP-binding/permease, putative 730670 0.00 V TDE0426 bacteriocin ABC transporter, bacteriocin-binding protein, putative 1997300 0.00 M TDE0437 PBS lyase HEAT-like repeat domain protein 15326333 1.57 C TDE0448 deoxyribose-phosphate aldolase (deoC) 40219000 0.45 F TDE0449 ferritin, putative 199496667 0.67 P TDE0451 arginine deiminase (arcA) 92357667 0.66 E TDE0460 transporter, putative 1801133 1.54 S TDE0463 purine nucleoside phosphorylase (deoD) 53346667 0.43 F TDE0501 hypothetical protein 9163200 0.00 S

314

TDE0523 conserved hypothetical protein 4614200 0.00 S TDE0536 hypothetical protein 3203433 0.00 S TDE0557 hypothetical protein 3581133 0.00 S TDE0585 hypothetical protein 18091000 1.51 S TDE0593 internalin-related protein 536157 0.00 S TDE0594 cysteine protease domain, YopT-type 2557167 0.00 - TDE0601 malonyl CoA-acyl carrier protein transacylase, putative 23808000 0.66 I TDE0605 hypothetical protein 72829000 0.40 - TDE0607 ParA family ATPase 0 NAD D TDE0614 precorrin-4 C11-methyltransferase (cobM) 10980000 0.00 H TDE0626 hypothetical protein 52321333 1.99 S TDE0641 UDP-N-acetylglucosamine 1-carboxyvinyltransferase (murA) 27933000 0.51 M TDE0647 chemotaxis protein methyltransferase (cheR) 23099333 0.47 T TDE0650 membrane protein, putative 0 NAD R TDE0651 inorganic pyrophosphatase, frameshift mutation 96338667 1.74 - TDE0652 membrane protein, putative 40358667 1.60 S TDE0654 peptidase, M20/M25/M40 family 167696667 0.66 R TDE0677 conserved hypothetical protein 125676667 0.38 S TDE0709 peptide methionine sulfoxide reductase (msrA) 20537000 2.10 O TDE0731 hypothetical protein 211176667 0.57 -

315

TDE0743 thioredoxin reductase (trxB) 824083333 0.62 O TDE0745 glycine reductase complex selenoprotein GrdA 27500000 4.60 C TDE0758 iron compound ABC transporter, periplasmic iron compound-binding protein, 6827467 1.69 P putative TDE0761 protease complex-associated polypeptide (prcA) 54532667 0.23 S TDE0762 Serine protease (dentilisin) from NCBI 115740000 0.17 - TDE0772 ribosomal protein L22 (rplV) 36534667 1.65 J TDE0782 ribosomal protein L6 (rplF) 16268333 2.30 J TDE0783 ribosomal protein L18 (rplR) 39484667 2.62 J TDE0784 ribosomal protein S5 (rpsE) 109160000 0.61 J TDE0793 conserved hypothetical protein 22475667 0.46 S TDE0799 glycerophosphoryl diester phosphodiesterase, putative 6033100 0.54 C TDE0813 ABC transporter, ATP-binding protein/permease 1060683 0.00 V TDE0834 Na+-translocating NADH/quinone reductase, E subunit (nqrE) 0 NAD C TDE0836 Na(+)-translocating NADH-quinone reductase, C subunit (nqrC) 71888667 0.63 C TDE0839 rhodanese-like domain protein 5911267 0.00 P TDE0840 hypothetical protein 7915900 2.07 S TDE0841 conserved hypothetical protein 29571667 0.34 S TDE0843 conserved hypothetical protein 187536667 0.63 S TDE0853 ribosomal protein S9 (rpsI) 68605000 1.99 J

316

TDE0870 phosphatase/nucleotidase 114953333 0.61 F TDE0872 recA protein (recA) 6319867 4.93 L TDE0874 transcriptional regulator, putative 22149667 0.59 K TDE0877 conserved hypothetical protein 15203667 0.43 I TDE0880 nicotinate-nucleotide-dimethylbenzimidazole phosphoribosyltransferase (cobT) 1499430 0.00 H TDE0881 ribosomal protein S16 (rpsP) 26367667 2.46 J TDE0938 conserved hypothetical protein 1113573 0.00 L TDE0942 long-chain-fatty-acid--CoA ligase, putative 17097000 1.69 I TDE0949 enolase (eno) 225013333 0.67 G TDE0951 lipoprotein, putative 91908333 0.64 R TDE0965 conserved hypothetical protein TIGR00282 0 NAD S TDE0971 arginyl-tRNA synthetase (argS) 16741667 0.66 J TDE0980 asparaginyl-tRNA synthetase (asnS) 55270000 1.60 J TDE0982 dihydroorotate dehydrogenase/oxidoreductase, FAD-binding 5524567 2.03 F TDE0983 oligopeptide/dipeptide ABC permease, frameshift mutation 0 NAD - TDE0984 oligopeptide/dipeptide ABC transporter, permease protein, putative 6575533 5.94 R TDE0985 oligopeptide/dipeptide ABC transporter, periplasmic peptide-binding protein, 321453333 3.19 E putative TDE0986 oligopeptide/dipeptide ABC transporter, ATP-binding protein 3993533 13.84 P TDE0987 oligopeptide/dipeptide ABC transporter, ATP-binding protein 0 NAD E

317

TDE1001 orotate phosphoribosyltransferase (pyrE) 27871667 0.52 F TDE1004 flagellar filament core protein 466846667 0.33 N TDE1009 methyl-accepting chemotaxis protein 28383333 2.74 T TDE1013 hemerythrin-related protein 0 NAD P TDE1016 probable ATP-dependent protease LA, frameshift mutation 3193867 0.59 - TDE1020 dicarboxylate transporter, periplasmic dicarboxylate-binding protein, putative 59879000 0.60 G TDE1021 lipoprotein, putative 14540667 0.00 S TDE1026 acyl carrier protein (acpP) 12719000 1.62 Q TDE1037 hypothetical protein 3106700 0.00 S TDE1050 hypothetical protein 91977667 0.62 T TDE1067 oligopeptide/dipeptide ABC transporter, ATP-binding protein 11394267 2.12 E TDE1068 oligopeptide/dipeptide ABC transporter, ATP-binding protein 1422667 2.44 P TDE1072 lipoprotein, putative 689326667 0.54 R TDE1086 conserved hypothetical protein TIGR00255 7883233 0.57 S TDE1088 conserved hypothetical protein 6955700 0.00 S TDE1089 conserved hypothetical protein 39941667 0.44 S TDE1102 hypothetical protein 89104667 0.54 S TDE1111 transporter, putative 5874533 0.00 C TDE1121 hypothetical protein 11731000 1.59 S TDE1122 anti-anti-sigma factor 17954000 0.00 T

318

TDE1174 hypothetical protein 20867333 0.42 S TDE1190 hypothetical protein 52163333 0.56 - TDE1203 cell division protein FtsA (ftsA) 3190000 3.41 D TDE1208 DNA topoisomerase I (topA) 69289000 1.94 L TDE1211 heat shock protein HslVU, ATPase subunit HslU (hslU) 39877000 2.46 O TDE1216 flagellar motor switch protein FliG (fliG) 7788267 1.68 N TDE1226 zinc ABC transporter, periplasmic zinc-binding protein (troA) 12195867 0.65 P TDE1231 hypothetical protein 32748333 0.67 S TDE1232 conserved domain protein 11368333 1.66 S TDE1234 hypothetical protein 283517 2.95 S TDE1236 triosephosphate isomerase (tpiA) 66671000 0.51 G TDE1244 TPR domain protein 26016333 0.63 U TDE1271 oligopeptide/dipeptide ABC transporter, ATP-binding protein 753637 0.00 E TDE1279 phosphoribosylformylglycinamidine synthase II (purL) 32458667 1.54 F TDE1296 ribosomal subunit interface protein, putative 49660333 0.39 S TDE1300 conserved hypothetical protein 20554000 0.41 S TDE1311 ABC transporter, ATP-binding/permease protein 3001367 1.77 V TDE1314 penicillin-binding protein 847027 0.00 M TDE1318 hypothetical protein 725143 0.00 U TDE1327 hypothetical protein 50248333 0.55 S

319

TDE1328 hypothetical protein 9560867 0.00 S TDE1333 hflC protein, putative 31090000 0.62 O TDE1349 rod shape-determining protein MreB (mreB) 3359300 6.64 D TDE1352 penicillin-binding protein 663403 0.00 M TDE1357 aldose 1-epimerase (galM) 19556667 0.67 G TDE1364 valyl-tRNA synthetase (valS) 12635000 1.59 J TDE1370 YjeF-related protein 1344200 0.00 G TDE1373 HD domain protein 1702133 1.93 S TDE1381 V-type ATPase, E subunit (atpE) 31526000 1.61 C TDE1388 conserved hypothetical protein 0 NAD S TDE1390 conserved hypothetical protein 2873567 1.50 C TDE1397 flagellar synthesis regulator FleN, putative 0 NAD D TDE1399 prolipoprotein diacylglyceryl transferase (lgt) 2294100 0.00 M TDE1406 hypothetical protein 38966000 0.56 S TDE1408 flagellar filament outer layer protein FlaA, putative 576866667 0.30 N TDE1409 flagellar filament outer layer protein FlaA, putative 559370000 0.28 N TDE1415 nucleotidyl transferase/aminotransferase, class V 32576667 1.82 E TDE1416 ABC transporter, permease protein 2626667 0.00 M TDE1417 ABC transporter, ATP-binding protein 1782700 2.00 M TDE1435 hypothetical protein 1232367 0.00 S

320

TDE1441 dTDP-glucose 4,6-dehydratase (rfbB) 5799633 1.83 M TDE1444 metallo-beta-lactamase family protein 16029000 2.79 J TDE1450 MutS2 family protein 2457867 1.90 L TDE1464 SsrA-binding protein (smpB) 7125867 0.00 O TDE1472 flagellar hook-associated protein 2 (fliD) 1970700 1.83 N TDE1475 flagellar filament core protein 44744667 0.63 N TDE1477 flagellar filament core protein 892906667 0.34 N TDE1478 conserved hypothetical protein 8751233 0.00 S TDE1480 conserved hypothetical protein 189256667 0.10 S TDE1484 hypothetical protein 6296500 0.00 - TDE1488 glyceraldehyde-3-phosphate dehydrogenase, type I (gap) 491943333 0.60 G TDE1489 hypothetical protein 32149333 0.53 - TDE1493 chemotaxis protein CheX (cheX) 177560000 0.34 N TDE1494 chemotaxis protein CheY (cheY) 171210000 0.53 T TDE1496 chromosome partition protein SmC, putative 5179200 2.08 D TDE1499 adenylosuccinate lyase, putative 52923667 1.66 F TDE1503 transcription termination factor Rho (rho) 32425333 1.66 K TDE1509 lipoprotein releasing system, ATP-binding protein, putative 1456133 0.00 V TDE1511 pathogen-specific surface antigen, putative 1545400000 0.51 P TDE1516 ABC transporter, ATP-binding protein, putative 2504867 2.86 V

321

TDE1519 conserved hypothetical protein 59522333 0.47 S TDE1533 conserved hypothetical protein 2624633 0.00 S TDE1538 DNA ligase, NAD-dependent (ligA) 1856767 1.84 L TDE1541 metallo-beta-lactamase family protein 12947000 2.15 S TDE1546 hypothetical protein 1954333 1.55 S TDE1548 conserved hypothetical protein TIGR00103 5305767 0.00 S TDE1552 hypothetical protein 7616833 0.00 - TDE1582 glycogen synthase (glgA) 3216367 0.00 G TDE1584 lipoprotein, putative 384266667 2.49 S TDE1588 tryptophanyl-tRNA synthetase, putative 0 NAD J TDE1594 pyridine nucleotide-disulphide oxidoreductase family protein 20400333 1.66 E TDE1628 hypothetical protein 58951667 1.95 - TDE1639 sporulation related repeat protein 58553333 0.60 S TDE1643 PASTA domain protein 9700633 1.82 T TDE1644 methionyl-tRNA formyltransferase (fmt) 1815733 0.00 J TDE1645 polypeptide deformylase (def) 0 NAD J TDE1657 N utilization substance protein B (nusB) 8365633 0.00 K TDE1660 leucine Rich Repeat domain protein 1216527 2.05 S TDE1663 OmpA family protein 54837333 1.66 M TDE1664 conserved domain protein 52141333 2.83 S

322

TDE1669 hemolysin 1258966667 0.20 E TDE1671 trigger factor (tig) 90896000 1.80 O TDE1687 conserved hypothetical protein 8830867 1.70 S TDE1689 ABC transporter, ATP-binding protein 0 NAD V TDE1699 hypothetical protein 3118300 0.00 - TDE1702 hypothetical protein 9087800 0.25 - TDE1712 flagellar filament outer layer protein (flaA) 1851533333 0.20 N TDE1713 hypothetical protein 0 NAD S TDE1725 conserved domain protein 1368067 0.00 S TDE1730 glycosyl hydrolase, family 2 0 NAD G TDE1745 conserved hypothetical protein 6718533 0.00 S TDE1748 GTP-binding protein, GTP1/Obg family 1590000 0.00 S TDE1751 ribosomal protein L21 (rplU) 36553333 3.95 J TDE1754 desulfoferrodoxin/neelaredoxin 73828000 2.68 C TDE1848 hypothetical protein 18270667 0.45 S TDE1849 hypothetical protein 45571667 0.30 S TDE1850 ABC transporter, permease protein, putative 24612333 1.61 M TDE1858 transcriptional regulator, PadR family 12206667 0.00 K TDE1859 hypothetical protein 45146000 0.47 S TDE1862 conserved domain protein 245266667 0.58 S

323

TDE1882 glycosyl hydrolase, family 1 15417000 1.80 G TDE1896 phosphoribosylaminoimidazolecarboxamide formyltransferase/IMP cyclohydrolase 8421800 1.67 F (purH) TDE1898 preprotein translocase, SecA subunit (secA) 28523667 1.59 U TDE1912 DNA-binding protein p25 20686000 0.52 K TDE1914 Holliday junction DNA helicase RuvB (ruvB) 1728917 0.00 L TDE1921 membrane protein, putative 15679000 0.00 S TDE1927 phenylalanyl-tRNA synthetase, beta subunit (pheT) 16258000 1.84 J TDE1930 conserved hypothetical protein 3623700 0.59 S TDE1947 ABC transporter, permease protein 0 NAD S TDE1950 membrane lipoprotein TmpC, putative 422300000 0.61 S TDE1951 metallo-beta-lactamase family protein 1763233 3.09 S TDE1992 OmpA family protein 0 NAD M TDE2009 conserved hypothetical protein 4332433 0.00 S TDE2029 hydrolase, TatD family 2696533 0.00 L TDE2039 lipoprotein releasing system, permease protein, putative 2799733 0.00 M TDE2044 lipoprotein, putative 481677 0.00 S TDE2061 conserved hypothetical protein 11339367 0.00 P TDE2074 radical SAM enzyme, Cfr family 2664000 0.00 R TDE2079 sigma-54 dependent transcriptional regulator, putative 1443567 1.68 T

324

TDE2083 anti-anti-sigma factor 34820333 0.00 T TDE2085 amino acid kinase family protein 76823333 0.43 F TDE2089 signal recognition particle protein (ffh) 852297 0.00 U TDE2092 amino acid ABC transporter, amino acid-binding protein, putative 1351333 0.00 T TDE2110 orotidine 5-phosphate decarboxylase (pyrF) 2703833 0.00 F TDE2118 topoisomerase IV, A subunit, putative 11541467 1.51 L TDE2120 glycine reductase complex proprotein GrdE2 (grdE-2) 459900000 0.43 S TDE2148 OmpA family protein 7732233 0.48 M TDE2163 conserved hypothetical protein 48750000 1.80 S TDE2167 pyridine nucleotide-disulphide oxidoreductase family protein 10058800 2.24 E TDE2185 lipoprotein, putative 2383133 0.00 G TDE2200 methionine gamma-lyase (megL) 596470000 0.57 E TDE2210 hypothetical protein 35590000 0.53 S TDE2217 galactose/glucose-binding lipoprotein (mglB) 254480000 0.52 G TDE2235 methylaspartate ammonia-lyase 123063333 1.87 E TDE2244 conserved hypothetical protein 1813500 6.52 S TDE2257 5-nucleotidase family protein 845120000 0.55 F TDE2275 tRNA pseudouridine synthase A (truA) 3460867 0.00 J TDE2280 transglycosylase, SLT family 5223867 0.38 M TDE2285 conserved hypothetical protein 0 NAD S

325

TDE2293 conserved hypothetical protein 476043 4.08 S TDE2325 conserved hypothetical protein 0 NAD S TDE2328 efflux pump component MtrF (mtrF) 23938000 1.57 H TDE2341 membrane-associated zinc metalloprotease, putative 8392433 1.74 M TDE2342 1-deoxy-D-xylulose 5-phosphate reductoisomerase (dxr) 1289963 0.00 I TDE2344 undecaprenyl diphosphate synthase (uppS) 2083200 0.00 I TDE2350 lipoprotein, putative 37905667 0.66 S TDE2353 flagellar hook-associated protein 3 1834867 0.00 N TDE2356 iron compound ABC transporter, periplasmic iron compound-binding protein, 5151000 0.00 P putative TDE2357 iron compound ABC transporter, permease protein 4943200 0.00 P TDE2369 conserved domain protein 425736667 0.53 O TDE2376 conserved hypothetical protein 3882400 0.00 S TDE2390 hypothetical protein 336300000 0.53 - TDE2395 Jag protein, putative 39802667 1.60 S TDE2403 metallo-beta-lactamase family protein 10069733 0.41 S TDE2410 hemolysin 163846667 0.38 E TDE2414 hypothetical protein 7559067 0.00 - TDE2451 S-adenosylmethionine:tRNA ribosyltransferase-isomerase (queA) 948770 0.00 J TDE2466 conserved hypothetical protein 27652000 0.00 O

326

TDE2469 hexokinase family protein 2270100 0.00 G TDE2480 chaperone protein HtpG (htpG) 80220667 2.40 O TDE2485 conserved hypothetical protein 27640667 0.66 S TDE2498 metallo-beta-lactamase family protein 11175933 0.65 C TDE2501 response regulator 16410667 0.46 T TDE2508 hypothetical protein 530330000 0.67 - TDE2512 aldehyde dehydrogenase (NADP) family protein 13819333 1.79 C TDE2515; conserved hypothetical proteinconserved hypothetical protein 2488367 0.00 S TDE0945 TDE2517 DNA repair protein RadA (radA) 3838300 0.63 O TDE2542 antigen, putative 2992200 0.00 - TDE2549 methyl-accepting chemotaxis protein 1364663 6.62 T TDE2551 hypothetical protein 1475633 0.00 S TDE2566 conserved domain protein 6948100 1.57 S TDE2568 thiamine biosynthesis protein ThiI (thiI) 0 NAD H TDE2573 glucose-6-phosphate isomerase (pgi) 9746033 2.15 G TDE2575 conserved hypothetical protein 4240900 0.40 S TDE2580 GGDEF domain protein 3335033 0.42 T TDE2586 DNA polymerase III, gamma and tau subunits, putative 5400767 2.96 L TDE2591 rhodanese-like domain protein 26251000 0.47 P

327

TDE2592 DbpA RNA binding domain protein 31204000 0.55 S TDE2594 conserved hypothetical protein 3912567 0.00 S TDE2597 ribulose-phosphate 3-epimerase (rpe) 10352933 0.00 G TDE2601 surface antigen, putative 80258333 1.51 M TDE2602 outer membrane protein, putative 279783333 0.56 M TDE2608 conserved hypothetical protein 18458000 1.84 S TDE2614 ApbE family protein 1673733 0.00 H TDE2616 cyclic nucleotide-binding protein 11704000 1.97 T TDE2617 cyclic nucleotide-binding protein 9485267 2.04 T TDE2643 oxidoreductase, FAD-dependent 0 NAD C TDE2673 hypothetical protein 0 NAD - TDE2682 conserved hypothetical protein 6990233 1.81 L TDE2685 flagellar synthesis regulator FleN (fleN) 1086167 0.00 D TDE2716 HAD-superfamily hydrolase, subfamily IA 14140667 0.56 S TDE2721 helicase domain protein 3375000 2.28 L TDE2761 conserved hypothetical protein 1946067 0.00 N TDE2762 flagellar motor switch protein FliY (fliY) 40587000 0.42 N TDE2763 flagellar motor switch protein FliM (fliM) 14706667 0.29 N TDE2764 flagellar protein FliL (fliL) 65052333 0.14 N TDE2765 flagellar motor rotation protein B (motB) 34386667 0.11 N

328

TDE2766 motility protein A (motA) 34589333 0.00 N TDE2768 flagellar hook protein FlgE (flgE) 43923333 0.47 N TDE2771 conserved hypothetical protein 0 NAD - TDE2775 lipoprotein, putative 8312167 0.39 S TDE2776 proline iminopeptidase (pip) 24991000 1.67 E TDE2779 hypothetical protein 59837667 0.11 S TDE2783 methyl-accepting chemotaxis protein 2736267 0.00 T A The abundance of each protein in the wild-type T. denticola ATCC 33520 was calculated from the average IBAQ intensity from three replicates. Zero value indicates the protein abundance was under the detectable limit.

B Geometric mean of ratios, from three replicates, produced from the LFQ intensity of protein in ∆motA33520 relative to that of protein in wild-type. Ratio of ≥1.5 indicates that the protein had increased in abundance in ∆motA33520 relative to wild-type and ratio of ≤0.67 indicates that the protein had decreased in abundance in ∆motA33520 relative to wild-type. C One-letter abbreviations for the functional COG categories: J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosome partitioning; V, defense mechanisms; T, signal transduction mechanisms; M, cell wall/membrane/envelope biogenesis; N, cell motility; U, intracellular trafficking, secretion, and vesicular transport; O, posttranslational modification, protein turnover, chaperones; C, energy production and conversion; G, carbohydrate transport and metabolism; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; R, general function prediction only; S, function unknown.

D NA: Not applicable as the protein was not detected in the wild-type. Refer to Table 5.6 for the protein abundance in ∆motA33520. The abundance of the proteins that are not present in Table 5.6 are 1761733 for TDE0965, 1620567 for TDE1013, 3477300 for TDE1397, 1333133 for TDE1645, 2114933 for TDE1689, 2733633 for TDE1947 and 636957 for TDE1992.

329

Table III.3 Proteins significantly changed in abundance in T. denticola ∆motB33520 mutant relative to wild-type (ratio≥1.5 and ≤0.67, p<0.05). Proteins predicted to be organized in an operon were shaded. In the case where two or more operons were arranged consecutively, different darkness of shadings was used to differentiate the operons.

C Locus Tag Protein description Wild-type ∆motB33520 COG abundanceA ratioB TDE0008 copper-translocating P-type ATPase 1737167 0.00 P TDE0011 alkyl hydroperoxide reductase/peroxiredoxin 5732033333 0.00 O TDE0012 carbon starvation protein CstA, putative 461243333 0.01 T TDE0014 conserved hypothetical protein 71764000 1.52 S TDE0034 hypothetical protein 15677667 0.57 - TDE0042 phosphate acetyltransferase (pta) 270816667 0.55 C TDE0043 TPR domain protein 3528667 0.60 U TDE0047 imidazolonepropionase (hutI) 65041667 0.61 Q TDE0048 hypothetical protein 18486000 0.00 - TDE0055 flagellar biosynthesis protein FlhA (flhA) 2858667 1.64 U TDE0064 phosphofructokinase (pfk) 44330667 0.61 G TDE0065 conserved hypothetical protein 8750800 0.00 S TDE0066 conserved hypothetical protein 6329767 0.30 S TDE0072 methyl-accepting chemotaxis protein 3616100 0.55 T TDE0081 hypothetical protein 1744067 0.00 S

330

TDE0089 transglutaminase-like domain protein 24740333 0.48 E TDE0103 aminotransferase, class-V 1789900 0.00 E TDE0114 iron-dependent transcriptional regulator 34609000 0.59 K TDE0119 flagellar protein FliS (fliS) 1433833 0.00 U TDE0129 PyrBI protein (pyrBI) 164263333 2.23 F TDE0143 thiamine ABC transporter, thiamine-binding protein 5358300 0.67 H TDE0149 DNA-binding response regulator 1913133 0.00 T TDE0151 integral membrane protein, YeeE/YedE family 34126667 0.47 S TDE0154 adenylate/guanylate cyclase catalytic domain protein 2167467 0.00 T TDE0175 pyrrolidone-carboxylate peptidase (pcp) 98789667 0.40 O TDE0183 ABC transporter, permease protein 6048633 1.80 P TDE0186 hypothetical protein 596683333 0.63 P TDE0200 tetrapyrrole methylase family protein 2478933 0.00 R TDE0207 permease, GntP family 38581667 0.38 E TDE0210 cob(I)alamin adenosyltransferase (cobO) 55988667 1.56 H TDE0212 conserved hypothetical protein 122642000 0.44 P TDE0213 TPR domain protein 3607700 0.59 U TDE0239 glycine reductase complex protein GrdD (grdD) 231270000 0.59 I TDE0294 TrkA domain protein 6565700 0.54 P TDE0300 cytosol aminopeptidase family protein 388300000 0.61 E

331

TDE0301 tetracenomycin polyketide synthesis O-methyltransferase TcmP, putative 5128000 0.00 Q TDE0336 membrane protein, putative 3610600 1.52 S TDE0337 glucosamine-6-phosphate isomerase (nagB) 49480667 0.41 G TDE0352 ribose 5-phosphate isomerase B (rpiB) 68881000 0.35 G TDE0354 general stress protein 14 18201333 0.40 R TDE0358 cinnamoyl ester hydrolase (cinI) 2591833 0.00 I TDE0374 glycosyl transferase, group 1 family protein 1152617 0.00 M TDE0383 hypothetical protein 5615867 0.00 S TDE0385 ABC transporter, ATP-binding/permease protein 0 NAD V TDE0386 ABC transporter, periplasmic substrate-binding protein 585636667 0.45 P TDE0389 (R)-2-hydroxyglutaryl-CoA dehydratase, beta subunit, putative 89464667 0.48 E TDE0392 (R)-2-hydroxyglutaryl-CoA dehydratase, beta subunit, putative 34354333 0.62 E TDE0398 oligopeptide/dipeptide ABC transporter, periplasmic peptide-binding protein 17442000 0.47 E TDE0405 major outer sheath protein 3398466667 0.54 M TDE0418 lipoprotein, putative 30177333 0.30 - TDE0419 hypothetical protein 27165667 0.14 - TDE0423 hypothetical protein 18472000 0.27 - TDE0425 bacteriocin ABC transporter, ATP-binding/permease, putative 730670 0.00 V TDE0448 deoxyribose-phosphate aldolase (deoC) 40219000 0.55 F TDE0451 arginine deiminase (arcA) 92357667 0.62 E

332

TDE0463 purine nucleoside phosphorylase (deoD) 53346667 0.55 F TDE0473 heavy metal translocating P-type ATPase 1095620 0.00 P TDE0501 hypothetical protein 9163200 0.00 S TDE0523 conserved hypothetical protein 4614200 0.00 S TDE0536 hypothetical protein 3203433 0.00 S TDE0540 hypothetical protein 12222000 0.65 S TDE0544 conserved hypothetical protein 12669667 0.41 S TDE0557 hypothetical protein 3581133 0.00 S TDE0558 hypothetical protein 10495600 0.00 S TDE0593 internalin-related protein 536157 0.00 S TDE0594 cysteine protease domain, YopT-type 2557167 0.00 - TDE0601 malonyl CoA-acyl carrier protein transacylase, putative 23808000 0.63 I TDE0605 hypothetical protein 72829000 0.48 - TDE0607 ParA family ATPase 0 NAD D TDE0615 cobalamin biosynthesis protein CbiG (cbiG) 1606633 0.00 H TDE0626 hypothetical protein 52321333 1.65 S TDE0630 sigma factor regulatory protein 6252300 0.67 T TDE0631 sigma factor regulatory protein, putative 7132733 0.41 T TDE0641 UDP-N-acetylglucosamine 1-carboxyvinyltransferase (murA) 27933000 0.46 M TDE0649 hypothetical protein 16920667 1.67 -

333

TDE0650 membrane protein, putative 0 NAD R TDE0651 inorganic pyrophosphatase, frameshift mutation 96338667 1.82 - TDE0652 membrane protein, putative 40358667 1.99 S TDE0654 peptidase, M20/M25/M40 family 167696667 0.63 R TDE0665 pyruvate ferredoxin/flavodoxin oxidoreductase family protein 348846667 1.92 C TDE0677 conserved hypothetical protein 125676667 0.41 S TDE0689 5-methylthioribose kinase 1505633 0.00 S TDE0728 FAD-binding protein, putative 2323400 0.00 C TDE0745 glycine reductase complex selenoprotein GrdA 27500000 4.59 C TDE0758 iron compound ABC transporter, periplasmic iron compound-binding protein, 6827467 4.14 P putative TDE0778 ribosomal protein L24 (rplX) 2249933 1.82 J TDE0781 ribosomal protein S8 (rpsH) 46230333 1.61 J TDE0782 ribosomal protein L6 (rplF) 16268333 1.54 J TDE0783 ribosomal protein L18 (rplR) 39484667 2.55 J TDE0784 ribosomal protein S5 (rpsE) 109160000 0.65 J TDE0793 conserved hypothetical protein 22475667 0.42 S TDE0799 glycerophosphoryl diester phosphodiesterase, putative 6033100 0.49 C TDE0813 ABC transporter, ATP-binding protein/permease 1060683 0.00 V TDE0834 Na+-translocating NADH/quinone reductase, E subunit (nqrE) 0 NAD C

334

TDE0840 hypothetical protein 7915900 1.70 S TDE0853 ribosomal protein S9 (rpsI) 68605000 2.36 J TDE0856 L-fuculose phosphate aldolase, putative 5552000 2.14 G TDE0865 conserved hypothetical protein 8134433 0.00 S TDE0870 phosphatase/nucleotidase 114953333 0.64 F TDE0872 recA protein (recA) 6319867 2.12 L TDE0877 conserved hypothetical protein 15203667 0.46 I TDE0880 nicotinate-nucleotide-dimethylbenzimidazole phosphoribosyltransferase (cobT) 1499430 0.00 H TDE0885 ribosomal protein L19 (rplS) 117815667 1.69 J TDE0951 lipoprotein, putative 91908333 0.63 R TDE0957 glycerophosphoryl diester phosphodiesterase family protein 6840667 0.00 C TDE0971 arginyl-tRNA synthetase (argS) 16741667 0.67 J TDE0983 oligopeptide/dipeptide ABC permease, frameshift mutation 0 NAD - TDE0984 oligopeptide/dipeptide ABC transporter, permease protein, putative 6575533 8.79 R TDE0985 oligopeptide/dipeptide ABC transporter, periplasmic peptide-binding protein, 321453333 5.20 E putative TDE0986 oligopeptide/dipeptide ABC transporter, ATP-binding protein 3993533 12.92 P TDE0987 oligopeptide/dipeptide ABC transporter, ATP-binding protein 0 NAD E TDE1001 orotate phosphoribosyltransferase (pyrE) 27871667 0.61 F TDE1004 flagellar filament core protein 466846667 0.24 N

335

TDE1009 methyl-accepting chemotaxis protein 28383333 1.68 T TDE1016 probable ATP-dependent protease LA, frameshift mutation 3193867 0.52 - TDE1020 dicarboxylate transporter, periplasmic dicarboxylate-binding protein, putative 59879000 0.59 G TDE1021 lipoprotein, putative 14540667 0.00 S TDE1027 ribonuclease III (rnc) 1776967 0.00 K TDE1049 translation elongation factor G (fusA-2) 126813333 1.50 J TDE1072 lipoprotein, putative 689326667 0.60 R TDE1075 oligopeptide/dipeptide ABC transporter, ATP-binding protein 26094333 0.43 P TDE1076 oligopeptide/dipeptide ABC transporter, ATP-binding protein 46452000 0.55 E TDE1086 conserved hypothetical protein TIGR00255 7883233 0.56 S TDE1088 conserved hypothetical protein 6955700 0.00 S TDE1089 conserved hypothetical protein 39941667 0.48 S TDE1094 conserved hypothetical protein 1494383 0.00 G TDE1102 hypothetical protein 89104667 0.59 S TDE1111 transporter, putative 5874533 0.00 C TDE1122 anti-anti-sigma factor 17954000 0.00 T TDE1124 hypothetical protein 10137967 1.85 S TDE1174 hypothetical protein 20867333 0.49 S TDE1197 conserved hypothetical protein TIGR00242 5088967 1.91 S TDE1198 S-adenosyl-methyltransferase MraW (mraW) 8303700 1.77 M

336

TDE1206 TPR domain protein 19359667 2.14 S TDE1208 DNA topoisomerase I (topA) 69289000 1.93 L TDE1226 zinc ABC transporter, periplasmic zinc-binding protein (troA) 12195867 0.58 P TDE1231 hypothetical protein 32748333 0.61 S TDE1234 hypothetical protein 283517 3.81 S TDE1236 triosephosphate isomerase (tpiA) 66671000 0.60 G TDE1259 amino acid carrier family protein 151753333 1.81 E TDE1271 oligopeptide/dipeptide ABC transporter, ATP-binding protein 753637 0.00 E TDE1273 oligopeptide/dipeptide ABC transporter, peptide-binding protein 320013333 0.62 E TDE1277 Fe-hydrogenase large subunit family protein 0 NAD C TDE1299 conserved hypothetical protein 8090767 0.64 S TDE1300 conserved hypothetical protein 20554000 0.41 S TDE1314 penicillin-binding protein 847027 0.00 M TDE1318 hypothetical protein 725143 0.00 U TDE1328 hypothetical protein 9560867 0.00 S TDE1333 hflC protein, putative 31090000 0.67 O TDE1342 conserved hypothetical protein 2159733 0.00 S TDE1349 rod shape-determining protein MreB (mreB) 3359300 2.78 D TDE1352 penicillin-binding protein 663403 0.00 M TDE1364 valyl-tRNA synthetase (valS) 12635000 1.51 J

337

TDE1370 YjeF-related protein 1344200 0.00 G TDE1373 HD domain protein 1702133 1.93 S TDE1388 conserved hypothetical protein 0 NAD S TDE1398 conserved hypothetical protein 9771067 0.53 L TDE1399 prolipoprotein diacylglyceryl transferase (lgt) 2294100 0.00 M TDE1407 conserved hypothetical protein 990537 0.00 S TDE1408 flagellar filament outer layer protein FlaA, putative 576866667 0.29 N TDE1409 flagellar filament outer layer protein FlaA, putative 559370000 0.28 N TDE1415 nucleotidyl transferase/aminotransferase, class V 32576667 1.57 E TDE1416 ABC transporter, permease protein 2626667 0.00 M TDE1432 conserved domain protein 0 NAD S TDE1435 hypothetical protein 1232367 0.00 S TDE1444 metallo-beta-lactamase family protein 16029000 2.59 J TDE1460 conserved domain protein 16383000 0.00 S TDE1464 SsrA-binding protein (smpB) 7125867 0.00 O TDE1467 HD domain protein 1107947 1.94 T TDE1468 glycoprotease family protein 3424067 0.00 O TDE1477 flagellar filament core protein 892906667 0.27 N TDE1478 conserved hypothetical protein 8751233 0.00 S TDE1480 conserved hypothetical protein 189256667 0.07 S

338

TDE1483 conserved hypothetical protein 127703333 1.52 S TDE1484 hypothetical protein 6296500 0.00 - TDE1489 hypothetical protein 32149333 0.64 - TDE1490 hypothetical protein 2719867 1.59 S TDE1493 chemotaxis protein CheX (cheX) 177560000 0.41 N TDE1494 chemotaxis protein CheY (cheY) 171210000 0.53 T TDE1496 chromosome partition protein SmC, putative 5179200 2.01 D TDE1499 adenylosuccinate lyase, putative 52923667 1.66 F TDE1503 transcription termination factor Rho (rho) 32425333 1.63 K TDE1509 lipoprotein releasing system, ATP-binding protein, putative 1456133 0.00 V TDE1511 pathogen-specific surface antigen, putative 1545400000 0.54 P TDE1519 conserved hypothetical protein 59522333 0.51 S TDE1533 conserved hypothetical protein 2624633 0.00 S TDE1541 metallo-beta-lactamase family protein 12947000 2.25 S TDE1548 conserved hypothetical protein TIGR00103 5305767 0.00 S TDE1584 lipoprotein, putative 384266667 2.30 S TDE1588 tryptophanyl-tRNA synthetase, putative 0 NAD J TDE1616 glycerate dehydrogenase (hprA) 11295100 1.83 C TDE1628 hypothetical protein 58951667 2.91 - TDE1639 sporulation related repeat protein 58553333 0.56 S

339

TDE1643 PASTA domain protein 9700633 1.50 T TDE1657 N utilization substance protein B (nusB) 8365633 0.00 K TDE1668 amino acid permease family protein 225143333 0.00 E TDE1669 hemolysin 1258966667 0.00 E TDE1671 trigger factor (tig) 90896000 1.58 O TDE1672 ATP-dependent Clp protease, proteolytic subunit ClpP (clpP-1) 11738800 2.24 U TDE1678 ribosomal protein S6 (rpsF) 66441667 1.85 J TDE1687 conserved hypothetical protein 8830867 1.84 S TDE1699 hypothetical protein 3118300 0.00 - TDE1712 flagellar filament outer layer protein (flaA) 1851533333 0.18 N TDE1713 hypothetical protein 0 NAD S TDE1725 conserved domain protein 1368067 0.00 S TDE1726 polyA polymerase family protein 4512833 2.23 J TDE1727 conserved hypothetical protein 102531333 0.54 O TDE1730 glycosyl hydrolase, family 2 0 NAD G TDE1745 conserved hypothetical protein 6718533 0.00 S TDE1748 GTP-binding protein, GTP1/Obg family 1590000 0.00 S TDE1754 desulfoferrodoxin/neelaredoxin 73828000 2.71 C TDE1848 hypothetical protein 18270667 0.57 S TDE1849 hypothetical protein 45571667 0.31 S

340

TDE1858 transcriptional regulator, PadR family 12206667 0.00 K TDE1859 hypothetical protein 45146000 0.37 S TDE1862 conserved domain protein 245266667 0.52 S TDE1881 methyltransferase, putative 2937400 0.00 L TDE1882 glycosyl hydrolase, family 1 15417000 1.81 G TDE1896 phosphoribosylaminoimidazolecarboxamide formyltransferase/IMP cyclohydrolase 8421800 1.56 F (purH) TDE1898 preprotein translocase, SecA subunit (secA) 28523667 1.55 U TDE1914 Holliday junction DNA helicase RuvB (ruvB) 1728917 0.00 L TDE1919 conserved domain protein 16423667 0.63 S TDE1927 phenylalanyl-tRNA synthetase, beta subunit (pheT) 16258000 1.71 J TDE1969 sigma-54 dependent transcriptional regulator/response regulator 1960267 0.50 T TDE1986 adenylate cyclase, putative 26364000 0.57 T TDE2009 conserved hypothetical protein 4332433 0.00 S TDE2029 hydrolase, TatD family 2696533 0.00 L TDE2039 lipoprotein releasing system, permease protein, putative 2799733 0.00 M TDE2043 signal recognition particle-docking protein FtsY (ftsY) 5551833 0.00 U TDE2061 conserved hypothetical protein 11339367 0.00 P TDE2073 TPR domain protein 53432667 0.52 G TDE2074 radical SAM enzyme, Cfr family 2664000 0.00 R

341

TDE2079 sigma-54 dependent transcriptional regulator, putative 1443567 0.00 T TDE2085 amino acid kinase family protein 76823333 0.46 F TDE2087 translation initiation factor IF-1 (infA) 43004667 1.68 J TDE2089 signal recognition particle protein (ffh) 852297 0.00 U TDE2092 amino acid ABC transporter, amino acid-binding protein, putative 1351333 0.00 T TDE2093 conserved hypothetical protein 21469000 0.41 S TDE2115 HAMP domain protein 6668467 0.56 T TDE2120 glycine reductase complex proprotein GrdE2 (grdE-2) 459900000 0.59 S TDE2144 conserved hypothetical protein 11541333 0.31 S TDE2148 OmpA family protein 7732233 0.34 M TDE2167 pyridine nucleotide-disulphide oxidoreductase family protein 10058800 2.27 E TDE2183 hypothetical protein 25323333 0.67 S TDE2189 M23/M37 peptidase domain protein 6483900 0.47 M TDE2200 methionine gamma-lyase (megL) 596470000 0.65 E TDE2210 hypothetical protein 35590000 0.51 S TDE2217 galactose/glucose-binding lipoprotein (mglB) 254480000 0.59 G TDE2226 ABC transporter, substrate-binding protein, putative 21784333 0.48 P TDE2235 methylaspartate ammonia-lyase 123063333 2.40 E TDE2244 conserved hypothetical protein 1813500 3.38 S TDE2257 5-nucleotidase family protein 845120000 0.60 F

342

TDE2280 transglycosylase, SLT family 5223867 0.32 M TDE2281 LysM domain protein 50345333 1.68 S TDE2285 conserved hypothetical protein 0 NAD S TDE2302 HD domain protein 10887667 0.00 T TDE2325 conserved hypothetical protein 0 NAD S TDE2328 efflux pump component MtrF (mtrF) 23938000 1.62 H TDE2342 1-deoxy-D-xylulose 5-phosphate reductoisomerase (dxr) 1289963 0.00 I TDE2344 undecaprenyl diphosphate synthase (uppS) 2083200 0.00 I TDE2346 translation elongation factor Ts (tsf) 108308333 0.60 J TDE2352 flagellar hook-associated protein FlgK (flgK) 20756000 0.58 N TDE2353 flagellar hook-associated protein 3 1834867 0.00 N TDE2356 iron compound ABC transporter, periplasmic iron compound-binding protein, 5151000 0.00 P putative TDE2357 iron compound ABC transporter, permease protein 4943200 0.00 P TDE2376 conserved hypothetical protein 3882400 0.00 S TDE2390 hypothetical protein 336300000 0.39 - TDE2391 peptidyl-prolyl cis-trans isomerase 192730000 0.63 O TDE2392 hypothetical protein 22553667 0.66 S TDE2410 hemolysin 163846667 0.12 E TDE2414 hypothetical protein 7559067 0.00 -

343

TDE2451 S-adenosylmethionine:tRNA ribosyltransferase-isomerase (queA) 948770 0.00 J TDE2466 conserved hypothetical protein 27652000 0.00 O TDE2469 hexokinase family protein 2270100 0.00 G TDE2480 chaperone protein HtpG (htpG) 80220667 2.13 O TDE2485 conserved hypothetical protein 27640667 0.65 S TDE2501 response regulator 16410667 0.61 T TDE2503 conserved domain protein 2092333 0.00 S TDE2504 O-sialoglycoprotein endopeptidase (gcp) 2471633 0.00 O TDE2515; conserved hypothetical protein;;conserved hypothetical protein 2488367 0.00 S TDE0945 TDE2517 DNA repair protein RadA (radA) 3838300 0.43 O TDE2536 conserved hypothetical protein 11292067 0.65 S TDE2542 antigen, putative 2992200 0.00 - TDE2549 methyl-accepting chemotaxis protein 1364663 7.13 T TDE2551 hypothetical protein 1475633 0.00 S TDE2566 conserved domain protein 6948100 1.73 S TDE2567 hypothetical protein 6569867 1.55 S TDE2568 thiamine biosynthesis protein ThiI (thiI) 0 NAD H TDE2571 conserved hypothetical protein 10394233 0.00 S TDE2573 glucose-6-phosphate isomerase (pgi) 9746033 2.08 G

344

TDE2580 GGDEF domain protein 3335033 0.29 T TDE2591 rhodanese-like domain protein 26251000 0.62 P TDE2592 DbpA RNA binding domain protein 31204000 0.58 S TDE2594 conserved hypothetical protein 3912567 0.00 S TDE2596 hypothetical protein 7303767 0.19 S TDE2597 ribulose-phosphate 3-epimerase (rpe) 10352933 0.00 G TDE2602 outer membrane protein, putative 279783333 0.58 M TDE2611 conserved hypothetical protein 2623300 0.32 S TDE2614 ApbE family protein 1673733 0.00 H TDE2616 cyclic nucleotide-binding protein 11704000 2.23 T TDE2643 oxidoreductase, FAD-dependent 0 NAD C TDE2670 glycosyl transferase, group 4 family protein 1350233 3.20 M TDE2673 hypothetical protein 0 NAD - TDE2682 conserved hypothetical protein 6990233 1.64 L TDE2685 flagellar synthesis regulator FleN (fleN) 1086167 0.00 D TDE2692 CTP synthase (pyrG) 16071333 1.54 F TDE2706 membrane protein, putative 4022967 1.91 G TDE2708 hypothetical protein 11600000 0.45 S TDE2709 BNR domain protein 9424600 0.55 S TDE2712 hypothetical protein 34280333 2.10 S

345

TDE2721 helicase domain protein 3375000 2.70 L TDE2725 cyclic nucleotide binding domain/GGDEF domain protein 3803833 0.00 T TDE2730 hydrolase, TatD family 43281333 0.65 L TDE2734 hypothetical protein 36143000 0.23 S TDE2748 acetyltransferase, GNAT family 33495000 0.63 S TDE2761 conserved hypothetical protein 1946067 0.00 N TDE2762 flagellar motor switch protein FliY (fliY) 40587000 0.23 N TDE2764 flagellar protein FliL (fliL) 65052333 0.12 N TDE2765 flagellar motor rotation protein B (motB) 34386667 0.00 N TDE2766 motility protein A (motA) 34589333 0.42 N TDE2768 flagellar hook protein FlgE (flgE) 43923333 0.33 N TDE2771 conserved hypothetical protein 0 NAD - TDE2775 lipoprotein, putative 8312167 0.51 S TDE2776 proline iminopeptidase (pip) 24991000 1.78 E TDE2779 hypothetical protein 59837667 0.10 S TDE2783 methyl-accepting chemotaxis protein 2736267 0.00 T A The abundance of each protein in the wild-type T. denticola ATCC 33520 was calculated from the average IBAQ intensity from three replicates. Zero value indicates the protein abundance was under the detectable limit.

B Geometric mean of ratios, from three replicates, produced from the LFQ intensity of protein in ∆motB33520 relative to that of protein in wild-type. Ratio of ≥1.5 indicates that the protein had increased in abundance in ∆motB33520 relative to wild-type and ratio of ≤0.67 indicates that the protein had decreased in abundance in ∆motB33520 relative to wild-type.

346

D NA: Not applicable as the protein was not detected in the wild-type. Refer to Table 5.6 for the protein abundance in ∆motB33520. The abundance of the proteins that are not present in Table 5.6 are 2374033 for TDE0385, 2939633 for TDE1277 and 679163 for TDE1432.

347

348

Minerva Access is the Institutional Repository of The University of Melbourne

Author/s: Ng, Hong Min

Title: Investigation of the role of Treponema denticola motility and uncharacterized protein TDE0659 in synergistic biofilm formation with Porphyromonas gingivalis

Date: 2018

Persistent Link: http://hdl.handle.net/11343/218171

File Description: Investigation of the role of Treponema denticola motility and uncharacterized protein TDE0659 in synergistic biofilm formation with Porphyromonas gingivalis

Terms and Conditions: Terms and Conditions: Copyright in works deposited in Minerva Access is retained by the copyright owner. The work may not be altered without permission from the copyright owner. Readers may only download, print and save electronic copies of whole works for their own personal non-commercial use. Any use that exceeds these limits requires permission from the copyright owner. Attribution is essential when quoting or paraphrasing from these works.