The Role of Mast Cell Proteases in Respiratory Disease

Andrew Deane

BSc(Hons), MSc, PGCE.

Discipline of Immunology and Microbiology

School of Biomedical Science and Pharmacy

Faculty of Health

The University of Newcastle

Newcastle, NSW, Australia

Submitted in the fulfilment of the requirements for the award of a PhD degree

This dissertation contains no material which has been accepted for the award of any other degree or diploma in any university or other tertiary institution and, to the best of my knowledge and belief, contains no material previously published or written by another person, except where due reference has been made in the text. I give consent to this copy of my dissertation, when deposited in the University Library, being made available for loan and photocopying subject to the provisions of the Copyright Act 1968.

Andrew Deane

August 2016

Acknowledgements

Firstly, I would like to thank my primary supervisor, Prof. Philip Hansbro, whose support and supervision made this work possible. I’d also like to thank my co-supervisor; Dr

Andrew Jarnicki for his moral support, technical guidance and encouragement throughout this entire process. I would also like to thank Dr Shaan Gellatly for her surrogate supervision during this work, especially in bacteriology.

I’d like to thank all the staff and students at HMRI who have provided assistance and friendship especially Gang Liu, James Pinkerton, Alexandra Brown, Celeste Harrison and

Prema Monogar.

I would also like to thank my partner Chris Beeson, for his unending moral support and patience in putting up with me during this long process.

Table of Contents

List of Figures 9

List of Tables 17

Abbreviations 17 Synopsis 20

Chapter 1: Introduction 23 1.1 The innate immune system 23 1.1.1 Pathogen detection by innate cells 23 1.1.2 Complement 24 1.1.3 Innate responses to pathogen detection 26 1.1.4 Antigen presentation 26 1.2 Respiratory infections 29 1.2.1 Immune protection against infection 29 1.3 Common pathogens associated with respiratory disease 31 1.3.1 Streptococcus pneumoniae 31 1.3.2 Pseudomonas aeruginosa 39 1.3.3 Influenza A virus 45 1.4 The role of mast cells in respiratory infections 53 1.4.1 Mast cell activation and degranulation 54 1.4.2 Role of mast cells in bacterial infection 55 1.4.3 Role of mast cells in viral infection 57 1.4.4 Role of mast cells in S. pneumoniae infection 60 1.4.5 Role of mast cells in P. aeruginosa infection 60 1.4.6 Role of mast cells in Influenza infection 61 1.4.7 Mast cell proteases 62 1.4.8 Mast cell tryptases 63 1.4.9 Mast cell chymases 68 1.4.10 Mast cell related tryptases 69 1.4.11 Mast cell derived Heparin 69 1.4.12 Mast cell factor NDST2 in mast cell granule composition and tryptase activity 70 1.4.13 Mast cell related factor RasGRP4 71 1.5 Mouse genetics 73 1.5.1 C57 mice 73 1.5.2 mMCP5 null mouse 73

1.5.3 mMCP6 null mouse 73 1.5.4 mMCP6 null, mMCP7 knock in mouse 74 1.5.5 Prss31 null mouse 74 1.5.6 NDST2 null mouse 74 1.5.7 Prss22 null mouse 75 1.5.8 RasGRP4 null mouse 75 1.5.9 Caspase 11 deletion in mice with 129Sv background 75 1.6 Study rationale 76

Chapter 2: The role of mast cell proteases and associated factors in the pathogenesis of S. pneumoniae in a pneumococcal pneumonia model in mice.

77 2.1 Abstract 78 2.2 Introduction 79 2.3 Methods 80 2.3.1 Ethics statement 80 2.3.2 Streptococcus pneumoniae infection model 80 2.3.3 Cellular inflammation 81 2.3.4 Bacterial recovery 81 2.3.5 Cytokine expression in BALf 81 2.3.6 Histopathological scoring 82 2.3.7 Statistics 82 2.4 Model 83 2.4.1 Characterisation of S. pneumoniae infection 83 2.4.2 S. pneumoniae infection in mMCP6-/- mice 85 2.4.3 S. pneumoniae infection in mMCP6-/- mMCP7+/+ mice 90 2.4.4 S. pneumoniae infection in Prss31-/- mice 96 2.4.5 S. pneumoniae infection in NDST2-/- mice 101 2.4.6 S. pneumoniae infection in mMCP5-/- mice 107 2.4.7 S. pneumoniae infection in Prss22-/- mice 112 2.4.8 S. pneumoniae infection in RasGRP4-/- mice 117 2.4.9 Summary of results 122 2.5 Discussion 123

Chapter 3: The role of mast cell proteases, their related proteases and mast cell associated factors in the pathogenesis of P. aeruginosa in a pneumonia model in mice.

130 3.1 Abstract 131 3.2 Introduction 132 3.3 Methods 133 3.3.1 Ethics statement 133 3.3.2 Clinical score 133 3.3.3 Pseudomonas aeruginosa infection model 133 3.3.4 Heat killed bacteria inoculation model 134 3.3.5 Cellular inflammation 134 3.3.6 Bacterial recovery 135 3.3.7 Cytokine expression in BALf 135 3.3.8 Statistics 135 3.3.9 Histopathological scoring 136 3.4 Results 137 3.4.1 Characterisation of P. aeruginosa infection 137 3.4.2 P. aeruginosa infection in mMCP6-/- mice 140 3.4.3 P. aeruginosa infection in mMCP6-/- mMCP7+/+ mice 148 3.4.4 P. aeruginosa infection in Prss31-/- mice 158 3.4.5 P. aeruginosa infection in NDST2-/- mice 166 3.4.6 P. aeruginosa infection in mMCP5-/- mice 175 3.4.7 P. aeruginosa infection in Prss22-/- mice 183 3.4.8 P. aeruginosa infection in RasGRP4-/- mice 191 3.4.9 Summary of results 199 3.5 Discussion 200

Chapter 4: The role of mast cell proteases, their related proteases and mast cell associated factors in the pathogenesis of Influenza A virus in a A/WSN/33 model in mice.

210 4.1 Abstract 211 4.2 Introduction 212 4.3 Methods 214 4.3.1 Ethics statement 214 4.3.2 Clinical score 214 4.3.3 H1N1 A/WSN/33 infection model 215 4.3.4 Cellular inflammation 215 4.3.5 Viral BALf collection 216 4.3.6 Plaque assay 216 4.3.7 Cytokine expression in BALf 216 4.3.8 Statistics 216 4.3.9 Histopathological scoring 217 4.4 Results 218 4.4.1 Characterisation of H1N1 A/WSN/33 infection 218 4.4.2 Influenza A infection in mMCP6-/- mice 220 4.4.3 Influenza A infection in mMCP6-/- mMCP7+/+ mice 228 4.4.4 Influenza A infection in Prss31-/- mice 236 4.4.5 Influenza A infection in NDST2-/- mice 244 4.4.6 Influenza A infection in mMCP5-/- mice 252 4.4.7 Influenza A infection in Prss22-/- mice 260 4.4.8 Influenza A infection in RasGRP4-/- mice 268 4.4.9 Summary of results 276 4.5 Discussion 277

Chapter 5: Discussions and conclusions 284 5.1 Significance of research 284 5.1.1 mMCP6 can modulate infection associated inflammation 285 5.1.2 mMCP7 is deleterious during S. pneumoniae, P. aeruginosa and Influenza virus infections 286 5.1.3 Prss31 is detrimental during IAV infection 286 5.1.4 Prss31 role in the pathogenesis of bacterial infection is pathogen specific 287 5.1.5 The role of Prss22 in the pathogenesis of respiratory infection is pathogen specific 288 5.1.6 RasGRP4 is protective in the later time points of P. aeruginosa and IAV infections 289 5.1.7 Further knock out mouse generation using CRISPER/Cas9 technology 289 5.2 Publications 290 5.2.1 Accepted publications 290 References 291

Appendix: Submitted paper: Gene expression signature of cigarette smoke-induced lung damage in a mouse model correlates with human chronic obstructive pulmonary disease. 316

List of Figures

Figure 1.1.2.1: Schematic overview of the complement cascade 24 Figure 2.4.1.1: Characterisation of Spn infection 84 Figure 2.4.1.2: Overview of Spn model 85 Figure 2.4.2.1: Spn infected mMCP6-/- mice show enhanced bacterial clearance 86 Figure 2.4.2.2: Spn infected mMCP6-/- mice show unaltered inflammatory cell infiltration 87 Figure 2.4.2.3: Histopathological scoring of Spn infected mMCP6-/- mice show no differences from infected control C57 mice 88 Figure 2.4.2.4: Cytokine profiling of Spn infected mMCP6-/- mice show no differences from infected control C57 mice 89 Figure 2.4.3.1: Spn infected mMCP6-/- mMCP7+/+ mice show unaltered bacterial clearance 91 Figure 2.4.3.2: Spn infected mMCP6-/- mMCP7+/+ mice show unaltered inflammatory cell infiltration 92 Figure 2.4.3.3: Histopathological scoring of Spn infected mMCP6-/- mMCP7+/+ mice show no differences from infected control C57 mice 93 Figure 2.4.3.4: Cytokine profiling of Spn infected mMCP6-/- mMCP7+/+ mice show impaired IL-1β responses following Spn infection 94 Figure 2.4.4.1: Spn infected Prss31-/- mice show enhanced bacterial clearance 96 Figure 2.4.4.2: Spn infected Prss31-/- mice show an unchanged inflammatory response 97 Figure 2.4.4.3: Histopathological scoring of Spn infected Prss31-/- mice show no differences from infected control C57 mice 98 Figure 2.4.4.4: Cytokine profiling of Spn infected Prss31-/- mice identified reduced IL-1β and enhanced IL-6 responses following Spn infection 99 Figure 2.4.5.1: Spn infected NDST2-/- mice show impaired bacterial clearance 101 Figure 2.4.5.2: Spn infected NDST2-/- mice show greater inflammation 102 Figure 2.4.5.3: Histopathological scoring of Spn infected NDST2-/- mice are elevated compared to the scores seen in infected control C57 mice 103 Figure 2.4.5.4: Cytokine profiling of Spn infected NDST2-/- mice identified altered cytokine profiles with reduced TNFα, CXCL2 and IL-1β, and enhanced IL-6 responses following Spn infection 105 Figure 2.4.6.1: Spn infected mMCP5-/- mice show impaired bacterial clearance 107 Figure 2.4.6.2: Spn infected mMCP5-/- mice show reduced inflammatory cell infiltration 108 Figure 2.4.6.3: Histopathological scoring of Spn infected mMCP5-/- mice show no differences from infected control C57 mice 109

Figure 2.4.6.4: Cytokine profiling of Spn infected mMCP5-/- mice show reduced CXCL2 and IL-1β responses 110 Figure 2.4.7.1: Spn infected Prss22-/- mice show no changes in bacterial clearance 112 Figure 2.4.7.2: Spn infected Prss22-/- mice show no changes in inflammatory cell infiltration 113 Figure 2.4.7.3: Histopathological scoring of Spn infected Prss22-/- mice show no differences from infected control C57 mice 114 Figure 2.4.7.4: Cytokine profiling of Spn infected Prss22-/- mice identified no differences in cytokine responses 115 Figure 2.4.8.1: Spn infected RasGRP4-/- mice show enhanced bacterial clearance 117 Figure 2.4.8.2: Spn infected RasGRP4-/- mice show no changes in inflammatory cell infiltration 118 Figure 2.4.8.3: Histopathological scoring of Spn infected RasGRP4-/- mice show no differences from infected control C57 mice 119 Figure 2.4.8.4: Cytokine profiling of Spn infected RasGRP4-/- mice shows a significant impairment in all cytokine responses 121 Figure 3.4.1.1: Characterisation of PA14 infection. P. aeruginosa infection was assessed in mice that received PA14 138 Figure 3.4.1.2: Overview of the PA14 live and heat killed models 139 Figure 3.4.2.1: PA14 infected mMCP6-/- mice show unaltered bacterial clearance and bacteraemia at 12 hours post infection 140 Figure 3.4.2.2: PA14 infected mMCP6-/- mice show decreased inflammatory cell infiltration due to reduced neutrophils at 12 hours post infection 141 Figure 3.4.2.3: Histopathological scoring of PA14 infected mMCP6-/- mice at 12 hours show no differences from infected control C57 mice 142 Figure 3.4.2.4: Cytokine profiling of PA14 infected mMCP6-/- mice 12 hours post infection show impaired TNFα and IL-1β responses 143 Figure 3.4.2.5: PA14 infected mMCP6-/- mice show unaltered bacterial clearance at 24 hours post infection 144 Figure 3.4.2.6: mMCP6-/- infected mice show an increased incidence of bacteraemia as detected by blood cultures at 24 hours post infection 144 Figure 3.4.2.7: PA14 infected mMCP6-/- mice show comparable cell infiltration at 24 hours post infection 145 Figure 3.4.2.8: Histopathological scoring of PA14 infected mMCP6-/- mice at 24 hours show no differences from infected control C57 mice 146 Figure 3.4.2.9: Cytokine profiling of PA14 infected mMCP6-/- mice 24 hours post infection show impaired TNFα responses 147

Figure 3.4.3.1: PA14 infected mMCP6-/- mMCP7+/+ mice show a worse overall clinical score at 12 hours post infection 148 Figure 3.4.3.2: PA14 infected mMCP6-/- mMCP7+/+ mice show enhanced bacterial clearance in the lungs but show a greater incidence of bacteraemia at 12 hours post infection 149 Figure 3.4.3.3: PA14 infected mMCP6-/- mMCP7+/+ mice show decreased inflammatory cell infiltration attributed to reduced neutrophilic infiltration at 12 hours post infection 150 Figure 3.4.3.4: Histopathological scoring of PA14 infected mMCP6-/- mMCP7+/+ mice at 12 hours show more inflammation in lungs than infected control C57 mice 151 Figure 3.4.3.5: Cytokine profiling of PA14 infected mMCP6-/- mMCP7+/+ mice 12 hours post infection show impaired TNFα and IL-6 responses 152 Figure 3.4.3.6: mMCP6-/- mMCP7+/+ mice show greater susceptibility to PA14 infection 153 Figure 3.4.3.7: Heat killed PA14 inoculated mMCP6-/- mMCP7+/+ mice show inflammatory cell responses comparable to their C57 controls at 12 hours 154 Figure 3.4.3.8: Histopathological scoring of heat killed PA14 inoculated mMCP6-/- mMCP7+/+ mice at 12 hours shows no differences from infected control C57 mice 155 Figure 3.4.3.9: mMCP6-/- mMCP7+/+ mice inoculated with heat killed PA14 shows worse clinical scores that their C57 controls at 12 hours 156 Figure 3.4.3.10: Cytokine profiling of heat killed PA14 inoculated mMCP6-/- mMCP7+/+ mice at 12 hours demonstrate impaired TNFα and IL-6 responses 157 Figure 3.4.4.1: PA14 infected Prss31-/- mice show comparable bacterial clearance to their C57 controls in the lungs 158 Figure 3.4.4.2: PA14 infected Prss31-/- mice show comparable inflammatory cell infiltration attributed to neutrophilic infiltration at 12 hours post infection when compared to their C57 controls 159 Figure 3.4.4.3: Histopathological scoring of PA14 infected Prss31-/- mice at 12 hours shows no differences from infected control C57 mice 160 Figure 3.4.4.4: Cytokine profiling of PA14 infected Prss31-/- mice at 12 hours demonstrate an impaired CXCL2 response when compared to C57 controls 161 Figure 3.4.4.5: PA14 infected Prss31-/- mice show comparable bacterial clearance to their C57 controls at 24 hours post infection 163 Figure 3.4.4.7: Histopathological scoring of PA14 infected Prss31-/- mice at 24 hours show no differences from infected control C57 mice 164 Figure 3.4.4.8: Cytokine profiling of PA14 infected Prss31-/- mice at 24 hours are comparable to their C57 controls 165 Figure 3.4.5.1: PA14 infected NDST2-/- mice show comparable bacterial clearance to their C57 controls in the lungs with a reduced incidence of bacteraemia as measured from the liver at 12 hours post infection 166 11

Figure 3.4.5.2: PA14 infected NDST2-/- mice show enhanced inflammatory cell infiltration attributed to neutrophilic infiltration at 12 hours post infection 167 Figure 3.4.5.3: Histopathological scoring of PA14 infected NDST2-/- mice at 12 hours show no differences from infected control C57 mice 168 Figure 3.4.5.4: NDST2-/- infected mice show worse clinical scores that their C57 controls at 12 hours 169 Figure 3.4.5.5: Cytokine profiling of PA14 infected NDST2-/- mice at 12 hours demonstrates impaired TNFα and IL-6 responses together with enhanced CXCL1 and CXCL2 responses 170 Figure 3.4.5.6: NDST2-/- mice show greater susceptibility to PA14 infection 171 Figure 3.4.5.7: Heat killed PA14 inoculated NDST2-/- mice show comparable inflammatory cell infiltration when compared to their C57 controls at 12 hours post inoculation 172 Figure 3.4.5.8: Histopathological scoring of heat killed PA14 inoculated NDST2-/- mice at 12 hours show no differences from infected control C57 mice 173 Figure 3.4.5.9: Cytokine profiling of heat killed PA14 inoculated NDST2-/- mice at 12 hours demonstrate an enhanced CXCL1 response when compared to C57 controls 174 Figure 3.4.6.1: PA14 infected mMCP5-/- mice show enhanced bacterial clearance in the lungs compared to their C57 controls at 12 hours post infection 175 Figure 3.4.6.2: PA14 infected mMCP5-/- mice show comparably more inflammatory cell infiltration predominantly attributed to neutrophilic infiltration at 12 hours post infection when compared to their C57 controls 176 Figure 3.4.6.3: Histopathological scoring of PA14 infected mMCP5-/- mice at 12 hours show no differences from infected control C57 mice 177 Figure 3.4.6.4: Cytokine profiling of BALf from of PA14 infected mMCP5-/- mice at 12 hours demonstrate impaired TNFα and CXCL1 responses together with an enhanced IL-1β response 178 Figure 3.4.6.5: PA14 infected mMCP5-/- mice show enhanced bacterial clearance in the lungs compared to their C57 controls at 24 hours post infection 179 Figure 3.4.6.6: PA14 infected mMCP5-/- mice show an impaired inflammatory cell infiltration at 24 hours post infection when compared to their C57 controls 180 Figure 3.4.6.7: Histopathological scoring of PA14 infected mMCP5-/- mice at 24 hours show no differences from infected control C57 mice 181 Figure 3.4.6.8: Cytokine profiling of PA14 infected mMCP5-/- mice at 24 hours demonstrate impaired CXCL1 and IL-6 responses 182 Figure 3.4.7.1: PA14 infected Prss22-/- mice show unaltered bacterial clearance at 12 hours post infection 183 Figure 3.4.7.2: PA14 infected Prss22-/- mice show comparable inflammatory cell infiltration when impaired to their C57 controls at 12 hours post infection 184

Figure 3.4.7.3: Histopathological scoring of PA14 infected Prss22-/- mice at 12 hours show no differences from infected control C57 mice 185 Figure 3.4.7.4: Cytokine profiling of PA14 infected Prss22-/- mice 12 hours post infection show impaired TNFα response together with an enhanced CXCL1 and CXCL2 responses 186 Figure 3.4.7.5: PA14 infected Prss22-/- mice show a degree of protection from bacteraemia at 24 hours post infection 187 Figure 3.4.7.6: PA14 infected Prss22-/- mice show comparable inflammatory cell infiltration when impaired to their C57 controls at 24 hours post infection 188 Figure 3.4.7.7: Histopathological scoring of PA14 infected Prss22-/- mice at 24 hours show no differences from infected control C57 mice 189 Figure 3.4.7.8: Cytokine profiling of PA14 infected Prss22-/- mice 24 hours post infection show comparable cytokine responses compared to C57 mice controls 190 Figure 3.4.8.1: PA14 infected RasGRP4-/- mice show unaltered bacterial clearance and bacteraemia at 12 hours post infection 191 Figure 3.4.8.2: PA14 infected RasGRP4-/- mice show decreased inflammatory cell infiltration attributed to reduced neutrophilic infiltration at 12 hours post infection 192 Figure 3.4.8.3: Histopathological scoring of PA14 infected RasGRP4-/- mice at 12 hours show no differences from infected control C57 mice 193 Figure 3.4.8.4: Cytokine profiling of PA14 infected RasGRP4-/- mice 12 hours post infection show an enhanced TNFα response 194 Figure 3.4.8.5: PA14 infected RasGRP4-/- mice show unaltered bacterial clearance at 24 hours post infection 195 Figure 3.4.8.6: PA14 infected RasGRP4-/- mice show an increased inflammatory cell infiltration at 24 hours post infection 196 Figure 3.4.8.7: Histopathological scoring of PA14 infected RasGRP4-/- mice at 24 hours show no differences from infected control C57 mice 197 Figure 3.4.8.8: Cytokine profiling of PA14 infected RasGRP4-/- mice 24 hours post infection show an impaired TNFα response 198 Figure 4.4.1.1: Characterisation of A/WSN/33 infection 219 Figure 4.4.2.1: mMCP6-/- A/WSN/33 infected mice fail to recover weight by day 10 post infection 217 Figure 4.4.2.2: mMCP6-/- infected mice have a lower viral peak than infected C57 controls 221 Figure 4.4.2.3: A/WSN/33 infected mMCP6-/- mice show comparable inflammatory cell infiltration attributed to macrophage, neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls 222

Figure 4.4.2.4: Histopathological scoring of the lungs from A/WSN/33 infected mMCP6-/- mice at 7 day p.i show comparable scores to those seen in infected control C57 mice 223 Figure 4.4.2.5: Cytokine profiling of A/WSN/33 infected mMCP6-/- mice at 7 days p.i demonstrate comparable cytokines responses to their C57 infected controls 224 Figure 4.4.2.6: A/WSN/33 infected mMCP6-/- mice show elevated inflammatory cell infiltration attributed to macrophages and lymphocyte infiltration at day 10 p.i when compared to their C57 controls 225 Figure 4.4.2.7: Histopathological scoring of the lungs from A/WSN/33 infected mMCP6-/- mice at 10 day p.i are elevated compared to the scores seen in infected control C57 mice 226 Figure 4.4.2.8: Cytokine profiling of A/WSN/33 infected mMCP6-/- mice at 10 days p.i demonstrate impaired IL-10 and slightly elevated IL-6 responses 227 Figure 4.4.3.1: mMCP6-/- mMCP7+/+ A/WSN/33 infected mice show similar weight loss to infected C57 controls 228 Figure 4.4.3.2: mMCP6-/- mMCP7+/+ infected mice have a higher viral peak than infected C57 controls 229 Figure 4.4.3.3: A/WSN/33 infected mMCP6-/- mMCP7+/+ mice show reduced inflammatory cell infiltration at day 7 p.i when compared to their C57 controls 230 Figure 4.4.3.4: Histopathological scoring of the lungs from A/WSN/33 infected mMCP6-/- mMCP7+/+mice at 7 day p.i show comparable scores to those seen in infected control C57 mice 231 Figure 4.4.3.5: Cytokine profiling of A/WSN/33 infected mMCP6-/- mMCP7+/+ mice at 7 days p.i demonstrate impaired IL-10 IFNg and IL-6 responses 232 Figure 4.4.3.6: A/WSN/33 infected mMCP6-/- mMCP7+/+ mice show reduced inflammatory cell infiltration attributed to macrophages and lymphocytes infiltration at day 10 p.i when compared to their C57 controls 233 Figure 4.4.3.7: Histopathological scoring of the lungs from A/WSN/33 infected mMCP6-/- mMCP7+/+mice at 10 day p.i show elevated scores than those seen in infected control C57 mice 234 Figure 4.4.3.8: Cytokine profiling of A/WSN/33 infected mMCP6-/- mMCP7+/+ mice at 10 days p.i demonstrate impaired IL-10 and IL-6 responses 235 Figure 4.4.4.1: Prss31-/- A/WSN/33 infected mice lose less weight infected C57 controls 236 Figure 4.4.4.2: Prss31-/- infected mice have a higher viral peak than infected C57 controls 237 Figure 4.4.4.3: A/WSN/33 infected Prss31-/- mice show comparable inflammatory cell infiltration attributed to macrophage, neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls 238 Figure 4.4.4.4: Histopathological scoring of the lungs from A/WSN/33 infected Prss31-/- mice at 7 days p.i show comparable scores to those seen in infected control C57 mice 239 Figure 4.4.4.5: Cytokine profiling of A/WSN/33 infected Prss31-/- mice at 7 days p.i demonstrate impaired IL-10 and enhanced IFNg responses 240 14

Figure 4.4.4.6: A/WSN/33 infected Prss31-/- mice show reduced inflammatory cell infiltration attributed to macrophage, neutrophil and lymphocyte at day 10 p.i when compared to their C57 controls 241 Figure 4.4.4.7: Histopathological scoring of the lungs from A/WSN/33 infected Prss31-/- mice at 10 days p.i show comparable scores to those seen in infected control C57 mice 242 Figure 4.4.4.8: Cytokine profiling of A/WSN/33 infected Prss31-/- mice at 10 days p.i demonstrate impaired IL-10 responses 243 Figure 4.4.5.1: NDST2-/- A/WSN/33 infected mice show similar weight loss to infected C57 controls 244 Figure 4.4.5.2: NDST2-/- have a viral peak equal to that of infected C57 controls 245 Figure 4.4.5.3: A/WSN/33 infected NDST-/- mice show comparable inflammatory cell infiltration attributed to neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls 246 Figure 4.4.5.4: Histopathological scoring of the lungs from A/WSN/33 infected NDST2-/- mice at 7 day p.i show comparable scores to those seen in infected control C57 mice 247 Figure 4.4.5.5: Cytokine profiling of A/WSN/33 infected NDST2-/- mice at 7 days p.i demonstrate comparable responses to those seen in C57 infected controls 248 Figure 4.4.5.6: A/WSN/33 infected NDST2-/- mice show greatly elevated inflammatory cell response attributed to macrophage, neutrophil and lymphocyte infiltration at day 10 p.i when compared to their C57 controls 249 Figure 4.4.5.7: Histopathological scoring of the lungs from A/WSN/33 infected NDST2-/- mice at 10 day p.i show higher scores than those seen in infected control C57 mice 250 Figure 4.4.5.8: Cytokine profiling of A/WSN/33 infected NDST2-/- mice at 10 days p.i demonstrate impaired IL-6 and elevated IFNg responses 251 Figure 4.4.6.1: mMCP5-/- A/WSN/33 infected mice lose more weight during IAV infection than infected C57 controls 252 Figure 4.4.6.2: mMCP5-/- have a higher viral peak than infected C57 controls 253 Figure 4.4.6.3: A/WSN/33 infected mMCP5-/- mice show comparable inflammatory cell infiltration attributed to macrophage, neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls 254 Figure 4.4.6.4: Histopathological scoring of the lungs from A/WSN/33 infected mMCP5-/- mice at 7 days p.i show comparable scores to those seen in infected control C57 mice 255 Figure 4.4.6.5: Cytokine profiling of A/WSN/33 infected mMCP5-/- mice at 7 days p.i demonstrate comparable cytokines responses to their C57 infected controls 256 Figure 4.4.6.6: A/WSN/33 infected mMCP5-/- mice show comparable inflammatory cell infiltration at day 10 p.i to their C57 controls 257

Figure 4.4.6.7: Histopathological scoring of the lungs from A/WSN/33 infected mMCP5-/- mice at 10 days p.i show higher scores than those seen in infected control C57 mice 258 Figure 4.4.6.8: Cytokine profiling of A/WSN/33 infected mMCP5-/- mice at 10 days p.i demonstrate impaired IL-10 and slightly elevated IL-6 responses 259 Figure 4.4.7.1: Prss22-/- A/WSN/33 infected mice show similar weight loss to infected C57 controls 260 Figure 4.4.7.2: Prss22 have a higher viral peak than infected C57 controls 261 Figure 4.4.7.3: A/WSN/33 infected Prss22-/- mice show comparable inflammatory cell infiltration attributed to macrophage and lymphocyte infiltration at day 7 p.i when compared to their C57 controls 262 Figure 4.4.7.4: Histopathological scoring of the lungs from A/WSN/33 infected Prss22-/- mice at 7 days p.i show comparable scores to those seen in infected control C57 mice 263 Figure 4.4.7.5: Cytokine profiling of A/WSN/33 infected Prss22-/- mice at 7 days p.i elevated IFNg and IL-6 responses 264 Figure 4.4.7.6: A/WSN/33 infected Prss22-/- mice show reduced inflammatory cell infiltration in BALf at day 10 p.i when compared to their C57 controls 265 Figure 4.4.7.7: Histopathological scoring of the lungs from A/WSN/33 infected Prss22-/- mice at 10 days p.i show higher scores than those seen in infected control C57 mice 266 Figure 4.4.7.8: Cytokine profiling of A/WSN/33 infected Prss22-/- mice at 10 days p.i demonstrate impaired IL-6 responses 267 Figure 4.4.8.1: RasGRP4-/- A/WSN/33 infected mice show similar weight loss to infected C57 controls 266 Figure 4.4.8.2: RasGRP4-/- have a lower viral peak than infected C57 controls 269 Figure 4.4.8.3: A/WSN/33 infected RasGRP4-/- mice show comparable inflammatory cell infiltration attributed to neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls 270 Figure 4.4.8.4: Histopathological scoring of the lungs from A/WSN/33 infected RasGRP4-/- mice at 7 days p.i show comparable scores to those seen in infected control C57 mice 271 Figure 4.4.8.5: Cytokine profiling of A/WSN/33 infected RasGRP4-/- mice at 7 days p.i demonstrate an elevated IL-6 response 272 Figure 4.4.8.6: A/WSN/33 infected RasGRP4-/- mice show elevated inflammatory cell infiltration attributed to macrophages and lymphocytes at day 10 p.i when compared to their C57 controls 273 Figure 4.4.8.7: Histopathological scoring of the lungs from A/WSN/33 infected RasGRP4-/- mice at 10 days p.i show higher scores than those seen in infected control C57 mice 274 Figure 4.4.8.8: Cytokine profiling of A/WSN/33 infected RasGRP4-/- mice at 10 days p.i demonstrate impaired IL-6 response 275 16

List of Tables

Table 2.4.9.1: Summary of S. pneumoniae infected mice deficient in mast cell proteases, associated proteases or mast cell associated factors 122 Table 3.4.9.1: Summary of P. aeruginosa infected mice deficient in mast cell proteases, associated proteases or mast cell associated factors and different time points during infection 199 Table 4.4.9.1: Summary of Influenza infected mice deficient in mast cell proteases, associated proteases or mast cell associated factors at different points during the viral infection 276

Abbreviations

C57 C57BL/6 SPN S. pneumoniae D39 PA14 P. aeruginosa PA14 WSN Influenza A/WSN/33 IAV Influenza A virus PAMP Pathogen associated molecular patterns PRR Pattern recognition receptors DAMP Damage associated molecular patterns HMGB1 High mobility group box 1 MASP Mannose binding lectin associated protease APC Antigen presenting cells ChoP Adhesin phosphorylcholine PAF Platelet activating factor CbpA Choline binding protein A ECM Extra cellular matrix PavA Pneumococcal adhesion and virulence A MHC Major histocompatibility complex MAC Membrane attack complex CPS Capsular polysaccharide PCV CPS-protein conjugate vaccines IPD Invasive pneumococcal disease PnPS Pneumococcal capsular polysaccharide PGK Phosphoglycerate kinase

pIgR Poly immunoglobulin receptors CBP Choline binding CBD Choline binding domain NLR Nod-like receptor LTA Lipoteichoic acid CLIP Class II invariant chain peptide ROI Reactive oxygen intermediates NET Neutrophil extracellular traps TCR T cell receptor PLY Pneumolysin PGN Peptidoglycan BALf Bronchoalveolar lavage fluid GAG Glucosaminoglycan NDST2 Glucosaminyl N-deacetylase/N-sulphotransferase-2 LPS Lipopolysaccharide DC Dendritic cells GEF Guanine nucleotide exchange factors PGN Peptidoglycan mMCP Mast cell protease CLP Cecal ligation and puncture SCF Secreted stem cell factor COX2 Cyclooxygenase 2 hsp40 Heat shock protein 40 NS1 Non-structural protein 1 CPS Capsular polysaccharide PCV CPS-protein conjugate vaccines PBS Sterile phosphate buffered saline SEM Standard experimental mean ANOVA Analysis of variance NP Nuclear protein PKR RNA- dependent protein kinase COPD Chronic obstructive pulmonary disease CF Cystic fibrosis nCFB Non-CF bronchiectasis 18

LB Lauria-Bertani agar i.n Intranasally p.i Post infection TNFα Tumour necrosis factor alpha IFNg Interferon gamma CXCL1 Chemokine (C-X-C motif) ligand 1 CXCL2 Chemokine (C-X-C motif) ligand 2 IL-6 Interleukin 6 IL-1β Interleukin 1 beta CFU Colony forming units WHO World Health Organisation MDCK Madin-Darby Canine Kidney PFU Plaque forming unit

Synopsis

S. pneumoniae, P. aeruginosa and Influenza A virus are 3 common respiratory pathogens responsible for extensive disease and mortality. S. pneumoniae is responsible for

25% of preventable child deaths, P. aeruginosa is the leading cause of Gram negative nosocomial infections whilst seasonal epidemics of Influenza infections result in half a million deaths per year and pandemic outbreaks of the virus such as the 1918 Spanish flu pandemic infected up to 1/3 of the global population resulting in an estimated 50 million deaths.

Mast cells, synonymous with allergy, are long lived tissue resident sentinels of the immune system. They are strategically located at host interface environments, including the lungs where they comprise 2% of the cross sectional area of alveolar walls. Their expression of a wide range of PRRs means they can detect pathogens and release numerous proinflammatory mediators to recruit additional cells or aid in direct killing of invading pathogens1.

Mature mast cells develop large electron dense granules2 filled with mast cell restricted proteases, such as chymases, tryptases and carboxypeptidase A3, together with growth factors, lysosomal enzymes, proteoglycans of serglycin, preformed cytokines and biogenic amines such as histamine and serotonin1. Mast cells comprise of almost 50% by weight of protease-serglycin complexes3. Following mast cell activation and subsequent degranulation these mediators are released from the cell.

The role of these mast cell proteases, mast cell related proteases and mast cell associated factors in infection and disease is unclear.

Using transgenic mice modified to insert or remove a range of differing mast cell proteases, mast cell related proteases or mast cell associated factors this study investigated

the role of mMCP5, mMCP6, mMCP7, NDST2, Prss31, Prss22 and RasGRP4 during

S. pneumoniae, P. aeruginosa and Influenza A virus infections.

The first study looked at the role of these proteases and associated factors in a

S. pneumoniae model. By interpreting a combination of results from different transgenic mice

I show that mMCP7 expression in this disease results in impaired bacterial clearance.

Additionally, I demonstrate that mMCP5-/- mice have impaired macrophage recruitment in this infection showing that one role for mMCP5 in this infection is macrophage recruitment.

The second study looked at the role of these proteases and associated factors in a

P. aeruginosa model. I demonstrate that mMCP6 promotes protective inflammatory responses during this infection with mMCP6-/- presenting with elevated instances of bacteremia. Additionally, I show mMCP7 expression is detrimental during this infection with elevated mortality rates observed together with an impairment in trans epithelial migration. I also repeated the observation that mMCP5-/- mice have a defective macrophage recruitment, and postulate that mMCP5s role in bacterial infections is to promote macrophage recruitment.

I also demonstrate Prss22 as playing a negative role during this infection contributing towards bacteraemia, this observation suggests that the use of Prss22 specific inhibitors could be of therapeutic use in combatting bacteraemia during this infection.

The final study investigated the role of these proteases and associated factors in an

Influenza A model. I show the mMCP6 limits excessive inflammation during the resolution of this infection by moderating IP-10 concentrations. Additionally, I show Prss31-/- mice are protected from excessive inflammatory responses, I propose that Prss31 promotes IL-10 production resulting in elevated inflammation. I also link Prss22 expression with protection from excessive inflammation during the infection, pointing to a possible therapeutic use of this tryptase. Finally, I highlight an anti-inflammatory role for RasGRP4 during the resolution of Influenza infection protecting mice from excessive inflammation.

These findings have helped me identify potential roles for these proteases and associated factors in a range of disease and point to their roles being infection specific. I have identified key proteases whose functions point to promising avenues for disease treatment.

Chapter 1:

Introduction

1.1 The innate immune system

1.1.1 Pathogen detection by innate cells

The innate immune system comprises of multiple components working synchronously. Its represents the first line of defence against pathogens and therefore requires an immediate response to variety of external stimuli. The cellular mechanisms by which the innate immune system detects external stimuli has been studied extensively and has re- emerged as an important focus. Briefly, pathogenic microbes are recognised as foreign, or dangerous, in part by their expression of an indefinite variety of pathogen associated molecular patterns (PAMP). These PAMPs are sensed by any one of several classes of host pattern recognition receptors (PRR)4 expressed on the internal and external cell surfaces and within the cytosol of both the innate and adaptive immune cells. PRR activation results in, either signal transduction to the nucleus and subsequent gene expression, typically orchestrated via the transcription factor NFκB (e.g. in toll like receptor activation) or the direct activation of inflammasomes that mediate caspase 1 dependent cleavage of preformed proinflammatory cytokines IL-1β, IL-18 and IL-33 (as in NOD-like receptor activation)5,6.

In some circumstances such as pathogen immune evasion strategies7 or tissue trauma, traditional PAMPs are not detected, instead host cells express endogenous molecules, damage associated molecular patterns (DAMP), which include alarmins such as defensins, annexins and high mobility group box 1 (HMGB1) proteins together with additional PAMPs8.

Alarmins such as HMGB1, are typically released following non-programmed cell death.

However some immune cells, such as those of myeloid lineages9,10, NK cells11 and tissue

resident cells including endothelial cells and smooth muscle cells12, can actively secrete

HMGB1 via an atypical secretory mechanism following exposure to inflammatory stimuli.

HMGB1 binds to a class of PRRs called Toll-like receptions, namely TLR-2 and 4 as well as to the receptor for advanced glycation end products (RAGE). The binding of HMGB1 to these receptors results in the expression of inflammatory cytokines and chemokines that actively recruit innate immune cells such as macrophages, neutrophils and dendritic cells to sites of damage8,13.

1.1.2 Complement

The complement system is an integral part of the innate immune system. It is comprised of an array of soluble and membrane bound proteins, and is activated by the detection of PAMPs by PRRs. Some complement components such as mannan binding lectin and C1q are classified as PRRs in their own right14 and these PRRs can bind PAMPs directly or when coated in antibodies creating an immune complex15. Activation of the complement cascade can occur by any one of three pathways, the classical, lectin or alternative pathways which all converge at a C3 convertase16.

Figure 1.1.2.1: Schematic overview of the complement cascade. 24

In the classical pathway the binding of C1q to the immunoglobulins IgG or IgM induces a conformational change resulting in the activation of serine esterase’s via C1r activation. C1r autocatalysis results in proteolytically cleaved C1s, which in turn catalyses the cleavage of C4 into C4a and C4b. C4b then binds C2 which is subsequently cleaved by C1s forming C4b2a, the classical pathway involving C3 convertase. The lectin pathway is initiated by the binding of mannose binding lectin or filolins to carbohydrate PAMPs

(mannose repeats and N-acetyl glucosamine, respectively). Binding results in a conformational change similar to that of C1q and immunoglobulin binding, activating mannose binding lectin associated proteases 1 & 2 (MASPS 1 & 2), which function analogous to C1r and C1s, resulting in the formation of the C4b2a C3 convertase. The alternative pathway differs from the previous 2 pathways in that the C3 convertase generated has different core proteins. C3 contains a thioester bond which is spontaneously hydrolysed by water, liberating C3a and allowing the binding of factor B, which is cleaved by factor D liberating Ba and leaving C3bBb, the alternative pathway C3 convertase17,18.

Both C3 convertases once formed are able to amplify the cleavage of C3 to C3a and

C3b. C3b is able to bind to any hydroxyl groups found in carbohydrates and proteins. Host cells, unlike pathogens contain sialic acid residues, which bind factor H, the key regulatory protein of the complement system. Factor H binds C3b, and together with complement receptor 1 (CD35), membrane cofactor protein (CD46) and serum resident Factor I C3b is inactivated leaving iC3b. In the absence of these regulatory “self” components, C3b is not inactivated and continues to build up on the cell surface, acting as an opsonin. It can also bind to either C3 convertase, forming the C5 convertases, C4b2a3b & C3bBbC3b, respectively.

These convertases then cleave C5 to C5a and C5b. C5b, C6 and C7 form the

C5bC6C7complex, which together with C8 forms the membrane attack complex (MAC) that inserts itself into lipid membranes. The catalytic addition of 15 C9 molecules in a ring

produces an transmembrane channel in the target phospholipid bilayer resulting in cell death and lysis17.

C3a, C5a and C4a to some extent, are potent anaphylatoxins capable of recruiting inflammatory cells such as neutrophils and macrophages and triggering mast cell degranulation, increasing vascular permeability, smooth muscle contraction and inducing chemotaxis16. They have also been shown to act directly on macrophages and dendritic cells altering their interactions with T cells17.

1.1.3 Innate responses to pathogen detection

Activation of the innate immune system results in the production and release of inflammatory mediators such as TNFα, IL-1α, IL-1β and IL-6 together with chemokines that attract circulating and tissue resident immune cells. This system can be self-perpetuating, depending on the extent and quality of the stimulus. In addition, activation can result in the production of factors that augment microbicidal activity such as the respiratory burst in neutrophils and macrophages19. Innate activation can also lead to the production of mediators such as IFNγ, TGFβ, histamines or interleukins that facilitate subsequent B and T cell activation, including complement, which is required for the formation of germinal centres within lymph nodes20. Also, the type and duration of stimuli through different PRR results in differential dendritic cell signalling and subsequent Th1/Th2 polarization of T cells21,22.

1.1.4 Antigen presentation

The innate system processes and presents antigen (primarily pathogen derived) bound to either MHC I or MHC II molecules to receptors on T cells (TCRs). Recognition of

MHC/antigen complex in conjunction with co-stimulatory molecules results in the activation

of T cell responses, clonal expansion and subsequent instigation of a specific adaptive immune response23.

In particular, the processing of viral antigens involves their ubiquitination within the cystol, then proteolytically degraded by a cystol resident multi-subunit proteasome. In uninfected cells, protein degradation is mediated via the constitutively expressed cystolic 26s proteasome, however, following exposure to IFNγ, transcription and translation of 3 additional proteasome subunits, b1i, b2i and b5i occurs. These subunits substitute the b1, b2 and b5 subunits altering the substrate specificity of the proteasome24. The additional binding of the IFNg inducible PA28α and PA28β proteasome activator molecules results in conformational changes exposing the α annulus and allowing access to previously restricted polypeptides25. These changes result in faster degradation, and generation of peptides 8 to 16 amino acids in length, which are preferentially transported into the lumen of the endoplasmic reticulum by TAP1 and TAP2. Once in the ER, these short peptides associate with MHC I molecules, resulting in Golgi-directed transport to the cell surface and subsequent presentation23.

Bacteria and their associated bacterial antigens are typically acquired by phagocytosis and subsequently digested within cytoplasmic bodies called phagolysosomes, by proteases such as cathepsins B, D, S and L. These antigens do not typically enter the cystol of the cell, thus cannot be digested by the proteasome. MHC II molecules are transported to the phagolysosome from the ER bound to an invariant chain which blocks peptide access to the binding groves of MHC molecules. Fusion between ER transport vesicles and the phagolysosome results in the exposure of the invariant chain to proteases and it subsequent degradation leaving a class II invariant chain peptide (CLIP) bound within the peptide binding groove of the MHC II molecule. MHC like molecules subsequently catalyse the

release of the CLIP resulting in the binding of pathogen-derived peptide and trafficking of the

MCH II peptide complex to the cell surface for presentation23.

An alternative process referred to as cross presentation can occur in specialised circumstances, in which MHC I and II bound peptides are presented on the alternative MHC, usually as a result of pathogenic disruptions to atypical presentation26.

1.2 Respiratory infections

1.2.1 Immune protection against infection

Mucosal tissue is exposed directly to the environment and pathogens, it subsequently requires specific immune responses to protect against infection. At around 90m2 the human airway has 9 times the surface area of the gut and 45 times more than the skin27. The distal airways filter up to 9000 litres of air every day and are exposed to a plethora of inhaled environmental agents with varying immunogenicity.

Evolution has driven the development of robust defence mechanisms. This involves clearance of bacteria through mechanical mucociliary action in the upper respiratory tract.

Also both innate and adaptive responses protect the host from pathogens whilst maintaining immune homeostasis and curtailing inappropriate immune responses, involving antimicrobial peptides, complement and surfactant proteins28. The potential danger of the inhaled environmental agents is initially monitored and assessed by lung resident dendritic cells and macrophages. Typically non pathogenic material results in a tolerogenic response mediated by a subset of regulatory T cells and resident DC subsets27.

The epithelial lining of the lower respiratory tract comprises of a magnitude of adapted cells types including non-ciliated clara cells, neuro-epithelial, ciliated, basal, serous and goblet cells as well as alveolar type I & II epithelial cells. This array of cell types contribute towards optimised mucociliary clearance of pathogens and particles from the lung.

Together they produce a range of lysozymes, defensins and surfactant proteins that act as defence proteins29.

Airspace leukocytes consist of approximately 95% alveolar macrophages, 4% lymphocytes and 1% neutrophils. All these cells avidly phagocytose any inhaled particles or pathogens30 allowing them to rapidly detect and respond to any pathogen encountered. More

recently innate lymphoid cells have been described and play an important role in mucosal barrier integrity yet their role in the lung remains poorly characterised31.

Mast cells play a role in immune protection in the lung. They comprise around 2% of total cell numbers in the lung32 and are strategically located at air interfaces within the lungs, where they can act as the front line defence to invading pathogens. Detection of pathogenic organisms trigger a robust inflammatory response typically comprising of macrophages and neutrophils initially with leukocyte recruitment occurring at later time points. If this response is protracted, excessive or ineffective, damage can occur to the thin walls of the lung which can result in impaired gaseous exchange and in extreme cases respiratory failure and death.

Recent studies have shown that mast cells can modulate this inflammatory response.

1.3 Common pathogens associated with respiratory disease

1.3.1 Streptococcus pneumoniae

1.3.1.1 Introduction

S. pneumoniae, often referred to as pneumococcus is the causative agent for diseases such as pneumococcal sepsis, pneumococcal meningitis and pneumococcal pneumonia. It is a

Gram positive bacterium, classified into pneumococcal capsular serotypes determined by their expression of structurally and antigenically diverse capsular polysaccharides. To date 91 serotypes have been described with only a few serotypes being pathogenic, and their geographic and historical distribution are not fully known.

In 2007, around 14.8 million incidences of pneumococcal disease were estimated to have occurred33. Of these, 826,000 deaths were attributed to have of occurred in children under the age of 5, and it is believed S. pneumoniae infection is responsible for 25% of all preventable deaths in this age group34. There is a higher disease burden in developing countries especially in Africa and Asia35. The most severe complication in infections of S. pneumoniae, bacteraemia, is the leading infectious cause of life threatening illness in

Australia, especially in at risk groups36.

1.3.1.2 Colonisation

S. pneumoniae pathogenesis initiates at nasopharyngeal colonisation, with its typical niche being the mucosal surfaces of the upper respiratory tract of humans. It can also be found in larger mammals that live in close association with humans34. Colonisation is typically asymptomatic however if the pneumococcus gains access to airways a neutrophilic inflammatory response is initiated causing disease symptoms.

Initial colonisation is associated with horizontal dissemination of the bacterium within the community via person to person contact. The exact mechanisms are yet to be determined, however it is hypothesised to be via direct contact with the secretions of an infected individual. Primarily, colonisation occurs within the first year of life, and declines over time34. Rates of carriage in the first year of life typically exceed 50% with a consistent decline into adulthood where typical carriage rates are between 5-10%37. However, incidents of S. pneumoniae mediated disease are higher in at risk populations including the under 5s, over 65s and immunocompromised individuals. Following colonisation the period of carriage in an individual host responses varies from a week to a few months38.

Upon entry into the nasal cavity S. pneumoniae contacts the sialic acid rich mucopolysaccharides in the mucus secretions. As the pneumococcal capsular polysaccharide

(CPS) is negatively charged, this results in a repulsing effect, facilitating access to the epithelial surfaces34,39. During initial colonisation transparent phase variant pneumococcus show enhanced binding to host tissues, verses opaque variants40. The adhesin phosphorylcholine (ChoP), a structural component of both cell wall teichoic acids and the membrane bound lipoteichoic acid, mediates adherence to the widely expressed platelet activating factor receptor (PAFr) by mimicking host platelet activating factor (PAF), that also contains ChoP, activating host cell signalling41 Choline binding protein A (CbpA) also known as PspC is non-covalently bound to ChoP via choline binding domain repeats, it binds to an epithelial glycoprotein, human secretory protein, a constituent part of the polymeric immunoglobulin receptors and secretory immunoglobulins, facilitating the translocation of the pneumococcus across the epithelium42,43.

Human epithelial cells have an array of glycoconjugates containing multiple disaccharide configurations. The configuration of N-acetylglucoseaminosyl groups play an import role in host susceptibility to pneumococcus. Andersson et al. showed pneumococcus

is able to bind to N-acetylglucosamine-β-(1,3)-galactose but not to N-acetylglucosamine β-

(1,4)-galactose44. Additionally pneumococcus contains three surface associated exoglycosidases, NanA a neuraminidase, BgaA a β-galactosidase and StrH, a β-N- acetylglucosaminidase, that sequentially cleave terminal saccharides from glycoconjugates rendering them functionally impaired, uncovering sterically inaccessible receptors and providing a nutrient source34. Upon reaching the basement membrane pneumococcal hyaluronidase binds to tetra and hexasaccharide hyaluronan substrates present throughout connective tissues45. Adhesion to extra cellular matrix (ECM) components fibronectin and plasminogen are facilitated by pneumococcal adhesion and virulence A (PavA) and enolase respectively46,47.

Pneumococcal toxin pneumolysin (PLY), one of the main virulence factors, is a cholesterol dependent cytolysin present in almost all clinical isolates of S. pneumoniae48. It induces cell lysis by creating transmembrane pores in cholesterol containing membranes.

PLY is also believed to facilitate bacterial transmission to the blood from the lungs34. Several studies demonstrate that the absence of PLY results in reduced virulence in murine models48.

Osmotic stress induced by pneumolysin pore formation in epithelial cells triggers the release of neutrophilic chemokines49.

1.3.1.3 Host response

Clearance of S. pneumoniae is mediated via the innate immune system and postulated to be via phagocytosis and subsequent intracellular killing mediated by resident alveolar macrophages and recruited neutrophils50, indeed one of the defining features of this infection is acute neutrophilic inflammation51. Following recruitment to the lung, neutrophils are activated resulting in secretion of reactive oxygen intermediates (ROIs), antimicrobial proteins such as hydrolases and the release of extracellular traps (NETs) to bind, neutralise

and kill bacteria. The transient containment of the bacteria as a result of NET production is believed to be beneficial to the host by reducing the risk of bacterial escape into the bloodstream and subsequent bacteraemia. Some of these secretions, whilst being beneficial in their bactericidal activity can cause extensive damage to the host tissues51.

PAMPs expressed by the bacteria are recognised by PRRs which induce the innate immune response. Bacterial lipoteichoic acid (LTA) and lipoproteins from the cell wall are detected by TLR252,53, bacterial PLY is recognised by TLR454 whilst unmethylated CpG motifs from bacterial DNA are detected via TLR9 within endosomes55. Following detection, these PRRs mediate inflammatory responses via the production of inflammatory mediators such as CXCL1, IL-6, IFNs and CCL2 via the interferon regulatory factors 3 and 7 and NF- kB. The Nod-like receptor (NLR) NOD2 detects bacterial peptidoglycans subsequently binding and activating caspase 1 and triggering pro-IL1b and -IL-18 processing and secretion56. Collectively, detection of PAMPS and resulting cytokine production results in recruitment of additional macrophages and neutrophils to the lung57.

Mast cells have been show to play mixed roles during infection. S. pneumoniae can induce mast cells degranulation without triggering the release of preformed TNFa or the de novo synthesis of TNFa and IL-660. In the early stages of S. pneumoniae infection (< 6 hours) these cells, and specifically the extracellular traps they generate, enhance the killing of bacteria58, however in later stages (>6 hours), mast cell activity is detrimental59, with elevated bacterial counts and wide spread bacterial dissemination to the blood. This detrimental activity is in independent of mast cell degranulation.

CD4 T cell mediated immunity is an important component in clearing pneumococcus from the murine nasopharynx, with MHC II deficient mice showing impaired pneumococcal clearance51. Additionally, opsonisation of pneumococci by complement or serotype specific

antibodies contributes towards the clearance via phagocytosis and cell mediated killing, however antibodies alone are not essential in this process37.

Complement also plays a role vital role in aiding the elimination of S. pneumoniae via opsonisation prior to neutrophil phagocytosis61. The complement cascade can be initiated via

C reactive protein (CRP) binding to ChoP, a constituent part of cell wall associated acids and lipoteichoic acids, subsequently interacting with complement component C1q initiating the classical pathway of complement activation62. Mannan binding lectin also contributes towards the opsonisation of S. pneumoniae, however in this infection, it fails to trigger the complement lectin pathway directly63. The absence of sialic acid residues on the bacteria and subsequent lack of Factor H binding means the spontaneous activation of the alternative pathway also contributes towards host responses.

Figure 1.3.1.3.1: Schematic overlay of a S. pneumoniae infection. 1. S. pneumoniae pathogenesis initiates at nasopharyngeal colonisation, a process that is typically asymptomatic. The bacteria gain access to the basement membrane and expresses adhesins such a ChoP that mediates adhesion to the widely expressed platelet activating factor receptor. 2. S. pneumoniae gains access to the airways allowing colonisation of the lower respiratory tract. 3. The expression of virulence factors, such as exoglycosidases to liberate a nutrient source from glycosolated proteins and pneumolysin which forms pours in cholesterol membranes triggering cell lysis, liberating more nutrients whilst facilitating bacterial access to the blood and subsequent systemic dissemination resulting in bacteraemia, aid the pathogenesis of the bacteria. Host responses are typically innate mediated with phagocytosis and intracellular killing orchestrated initially by alveolar macrophages and then neutrophils. Neutrophils release reactive oxygen intermediates and extracellular traps (NETs) to neutralise and kill bacteria. Complement plays a key role in opsonising bacteria prior to neutrophilic phagocytosis and killing.

1.3.1.4 Virulence factors

S. pneumoniae has evolved several strategies to circumvent host immune responses.

It has a novel capacity to resist NET mediated killing via a surface endonuclease (endA), that effectively enables the bacteria to degrade the DNA scaffold of NETs and escape64. The prevalence of endA in pathogenic S. pneumoniae serotypes is common with endA activity reported in serotypes 1, 2, 4, 7f, 14 and 19f 64.

S. pneumoniae inhibit humoral responses through the secretion of antibody degrading proteases. At the mucosal surface, pneumococcus is opsonised by host IgA1, the major mucous antibody component comprising >90% of mucosal immunoglobulins. IgA1 is subsequently cleaved by secreted S. pneumoniae zinc metalloproteases resulting in pneumococcal antigen bound to the IgA1 Fab fragments effectively neutralising Fc mediated inflammatory responses34.

Another bacterial factor ChoP, binds a number of proteins (10-15) known as choline binding proteins (CBP) via a 20 amino acid repeat sequence known as the choline binding domain (CBD). One such CBP, PspA is a lactoferrin binding protein, it provides protection from the bactericidal activity of apolactoferrin and disrupts complement C3 binding to the bacterial cell surface41. Another CBP, Choline binding protein A (CbpA), binds to the ectodomain of human poly-immunoglobulin receptors (pIgR) facilitating adherence41 and possibly transepithelial transport34. It also binds to complement factor H preventing the formation of C3b and the activation of the alternative pathway and opsonisation65. The bacterium is further protected from complement via phosphoglycerate kinase (PGK) which can modulate complement by integrating with C5, C7 and C9, disrupting the formation of the

MAC and preventing subsequent killing66.

The decline in pneumococcus carriage correlates with increasing levels of serological and mucosal antibodies to pneumococcal capsular polysaccharide (PnPS). This together with

the decreased rates of carriage in vaccinated populations whose anti-PnPS antibody titre has been boosted led, to the assumption that an anti-PnPS antibody titre is key to preventing the carrier state37 and that anti PnPS antibody cross reactivity to other homotypic pneumococcus

PnPS, while not protecting from colonisation, contributes towards a shortened carriage duration37,38,67.

1.3.1.5 Treatment and prevention

With growing levels of antibiotic resistance to penicillin, erythromycin and fluoroquinolones complicating treatment options and contributing to increased health care associated costs, prophylactic vaccination covering the most prevalent serotypes offers strong potential benefits in patient outcomes68. Immunisation has helped reduce the incidence of invasive pneumococcal disease (IPD) in developed nations, in some cases by >90%69, but developing nations, often battling other epidemics such as HIV, carry a greater burden with countries such as Mozambique suffering from high incidences70,71.

There is growing concern with the wide spread resistance to penicillin and reports of resistance to other antibiotics72, thus vaccination is in reality, the only feasible way of controlling IPD long term71. Older vaccines only targeted the capsular polysaccharide (CPS) of S. pneumoniae, with the 23 valent vaccine formulation still in use today being effective against approximately 90% of pathogenic serotypes. Unfortunately CPSs are poorly immunogenic as they are T cell independent antigens and show limited protection in young children (under 2) due to a poor antibody response to polysaccharide vaccines in this age group73. More recently CPS-protein conjugate vaccines (PCV) have been licensed and show considerable efficacy in the previously poorly protected high at risk groups such as children74 and the elderly75.

PCVs remain serotype specific and cost implications have resulted in the scope of serotype coverage to be cut to just the 13 most prevalent serotypes. Ultimately ubiquitous

PCV usage could simply apply selective pressure on serotype distribution resulting in promoting non PCV covered serotypes to replace those covered, indeed there is growing evidence to support this notion76. Serotype prevalence varies from country to country making serotype surveillance and region specific PCV development more challenging for developing countries ultimately contributing to additional costs rendering the PCV prohibitively expensive for those most at need76.

1.3.2 Pseudomonas aeruginosa

1.3.2.1 Introduction

Pseudomonas aeruginosa is one of the most common Gram negative bacteria associated with nosocomial respiratory infections77,78. The rod shaped facultative anaerobe, ubiquitous in the environment, is found in a range of habitats, from soil and aquatic niches to animal and human hosts79,80, and is resistant to many commonly used antibiotics. It is a plant and protozoa pathogen but can also infect mammals81. In humans, host immune responses are typically sufficient to prevent infection, however in instances of underlining pathology, such as burns, immunodeficiencies, immunocompromised individuals and chronic pulmonary disorders such as chronic obstructive pulmonary disease (COPD), Cystic fibrosis (CF) or non-CF bronchiectasis (nCFB), P. aeruginosa acts as an opportunistic pathogen. It is also the leading cause of chronic cystic fibrosis lung disease and is found in up to 80% of all CF patients by the age of 1882.

Data shows that P. aeruginosa chronically infects between 10-30% of nCFB and

COPD patients with this infection being associated with a poor prognosis83,84. Infection is also a major morbidity and mortality risk factor for the critically ill and immunocompromised

with P. aeruginosa associated pneumonia and sepsis having high mortality rates, with some institutions report ventilator acquired pneumonia mortality rates up to 30%80,81. Evidence also shows that the levels of drug resistance in this species is rising85, limiting therapeutic options to combat this opportunistic pathogen and contributing to a growing burden on health care systems globally.

The genome of P. aeruginosa is larger than most other sequenced bacteria, with

PA14s genome being 6.538 megabases long86 and encoding 5892 protein open reading frames79. P. aeruginosa contains a large repertoire of genes involved in the transport, regulation catabolism and efflux of organic compounds which together with its putative chemotaxis systems gives P. aeruginosa the ability to survive and thrive in a plethora of ecological niches86. It also contains an array of virulence factors enabling pathogenesis in a variety of milieu.

With its capacity to survive in a range of environmental niches, and its potential to cause disease in a range of compromised patients, P. aeruginosa is a major risk to patient health and a growing burden on health care systems80,85.

Whilst there has been extensive research in the area, to date there are no approved vaccines to P. aeruginosa. For patients with COPD, CF or nCFB immunisation against this pathogen isn't as yet possible.

1.3.2.2 Colonisation

P. aeruginosa’s genome contains a variety of genes encoding a multitude of nutrient acquisition pathways, together with virulence factors and defence mechanisms, which allow it to thrive in a range of environments81. One of the key factors aiding pseudomonads ability to colonise such a broad range of niches is attributed to their ability to form a biofilm, which is particularly useful in colonising the lung.

Upon reaching the pulmonary epithelium, P. aeruginosa can establish either a transient (acute) infection, like those seen in ventilator acquired pneumonia, or a persistent

(chronic) infection, typical seen in CF and some nCFB and COPD patients80.

In transient infections, planktonic P. aeruginosa is propelled and attaches to the pulmonary epithelial cell surface via fimbria, a type IV pili, binding directly to mucin or the glycolipid asialoGM1s b-GalNAc(1-4)bGal residues that are enriched on repairing epithelium87,88. Next, using its type III secretion system P. aeruginosa injects its 4 effector proteins. The first, ExoU, a phospholipase disrupts the extracellular singling kinases 1 & 2 p38 pathways whilst simultaneously initiating the proapoptotic JNK1/2 pathways. The next two ExoS and ExoT inhibit the functions of several GTPases altering cytoskeletal function and disrupting epithelial tight junctions, cellular polarity and repair facilitating basolateral invasion89. The final effector protein ExoY acts to increase cAMP 800 fold in the target cell, impairing bacterial uptake by the cell while disrupting the endothelial barrier function28.

Collectedly these effector proteins damage host cells facilitating access to the circulation.

Other virulence factors such as quorum sensing molecules, rhamnolipids, exotoxin A and elastase contribute towards disrupting the permeability of the respiratory epithelium promoting further invasion. Simultaneously P. aeruginosa forms biofilms by secreting exopolysaccharide matrix, providing protection from neutrophil mediated phagocytosis and reducing the efficacy of antibiotics81. The biofilm acts as a reservoir for the bacteria with active or mechanical processes resulting in subsequent biofilm dispersion.

In persistent infections the colonisation process is subtler with many patients being asymptomatic. Unlike transient infections, persistent infections are typically restricted to the mucus, without contact with the epithelial surface. There is a phenotypic switch to a mucoid colony, with marked down regulation of many of the bacteria virulence factors81. One virulence factor that is found only in persistently infected lungs is the copolymer

exopolysaccharide Alginate90, which has been shown to inhibit local immune responses promoting persistence over invasiveness81.

Figure 1.3.2.2.1: Schematic overlay of a P. aeru gin osa infection. 1. P. aeruginosa gains access to the lower respiratory tract and rapidly forms a biofilm. 2. Upon reaching the pulmonary epithelium the bacteria can establish transient or a persistent infection. During the acute infection the bacteria attached to the basement membrane via fimbria, it then injects its type 3 secretion system proteins into the host cells disrupting epithelial tight junctions. 3. Once the epithelial barrier is disrupted the bacteria is disseminated via the circulation causing widespread bacteraemia, whilst at the same time forming biofilms on the respiratory epithelium. Infection is rapidly followed by an extensive neutrophilic inflammatory response. The Neutrophils release reactive oxygen intermediates and extracellular traps (NETs) to neutralise and kill bacteria. Complement plays a key role in opsonising bacteria prior to neutrophilic phagocytosis and killing.

1.3.2.3 Host response

The hallmark of pseudomonal respiratory infections is a robust neutrophilic response, this response results in destructive pulmonary inflammation and IL-8 production, a potent neutrophil chemoattractant. P. aeruginosa quorum pheromone N-3-oxododecanoyl homoserine lactone amplifies IL-8 production91, resulting in an excessive neutrophilic influx, leading to the overproduction of proteolytic enzymes causing extensive tissue destruction and pulmonary failure91.

Airway epithelial cell and alveolar macrophages PRRs TLR4 and TLR5 recognise the

PAMPs LPS and flagellin from P. aeruginosa respectively92, triggering the production of inflammatory CXCL1, CXCL2, IL-6 and TNFa. Whilst the T3SS induces NLRC4 dependent caspase 1 activation and subsequent processing of pro-IL-1b to its active form, which is subsequently secreted92 inducing airway epithelial cells to further promote the inflammatory responses by releasing neutrophil chemokines such as KC and MIP-293.

Neutrophils recruited to the lungs facilitate pathogen elimination, mediated by their serine proteases, neutrophil elastase, cathepsin G and proteinase 3.

Maintaining the epithelial integrity of the respiratory tract is essential for an effective host response, and mast cells have been shown to be essential in maintaining epithelial integrity during infection. They secrete an as yet unknown factor that is essential for the maintenance of this integrity in a P. aeruginosa model94.

In addition to cellular responses host complement is also involved in aiding elimination of P. aeruginosa either via opsonisation or direct killing via the assembly of the membrane attack complex. Studies showing that mice deficient in C5aR and C5, whilst capable of mounting a robust neutrophilic inflammatory response, show profound susceptibility to P. aeruginosa supporting the essential role compliment plays in this infection95.

1.3.2.4 Treatment and prevention

Treatment of P. aeruginosa can be complicated by its ability to form biofilms and aggregates on mucosal membranes96 resisting antibiotic action. This together with its intrinsic chromosomally encoded antibiotic resistance to b-lactams, macrolides, tetracyclines and most fluoroquinolones and capacity to acquire new resistance via horizontal gene transfer or mutations80 to aminoglycosides, b-lactam/b-lactamase inhibitor combinations, carbapenems, carboxypenicillins, 3rd and 4th generation cephalosporins, monobactams some fluoroquinolones, and the polymyxins97, makes treatment particularly challenging.

Treatment challenges demonstrate the need for a preventative vaccine for those patients at risk and despite 50 years of research, a licensed vaccine for clinical use is yet to be brought to market. Dilemmas include developing a prophylactic vaccine that must be administered prior to colonisation by P. aeruginosa thus limiting its efficacy to those high risk patients80 who are uncolonized. The development of a therapeutic vaccine for chronically infected patients also presents major challenges given underlining immunopathology and variable immunological status in different high risk cohorts.

Given the challenges of treating P. aeruginosa and its capacity to develop a broad range of resistance to many of our current antibiotic interventions, the need for a preventative or therapeutic vaccine is paramount.

1.3.3 Influenza A virus

1.3.3.1 Introduction

The Orthomyxoviridae family of viruses contain an 8 segmented negative strand ssRNA genome encoding up to 16 proteins, although not all viruses express all 1698. There are three distinct types, A, B and C99, type A circulates in human, avian, canine, equine and swine populations and is classified by the expression of two glycoproteins, hemagglutinin

(HA, 18 subtypes) and neuraminidase (NA, 9 sybtypes).

In humans, the virus is restricted to the upper respiratory system and can cause symptoms for 7-10 days100. Annually, Influenza is responsible for between 300,000 and

500,000 deaths globally101, with infants, elderly and immunocompromised being at greatest risk of presenting with enhanced disease severity and higher mortality rates102. In the recent swine flu pandemic pregnant women and obese patients with other underlining medical conditions demonstrated elevated mortality rates103.

The absence of a proofreading capacity in the Influenza polymerase results in a high frequency of point mutations, with approximately one error per copied viral genome, effectively meaning that one infected cell is capable of generating approximately 10,000 new viral mutants104, this process drives antigenic drift and subsequent seasonal epidemics.

Occasionally two different subtypes of Influenza virus will infect a host, often from animal reservoirs resulting in gene reassortment between the different Influenza viruses (antigenic shift) generating novel subtypes that are capable of causing pandemics. Widespread distribution of animal reservoirs often living in close proximity to human populations means the potential for a gene reassortment with a subtype capable of causing disease in humans is an ever-present risk requiring constant surveillance of circulating virus subtypes in animal reservoirs is required. The recent Swine flu pandemic had relatively low virulence, yet this outbreak demonstrates that pandemics can occur suddenly and unexpectedly despite our

constant global surveillance. Currently the Influenza A virus (IAV) H5N1 avian strain is of significant concern given the high virulence rates observed and Influenza’s ability to adapt.

The frequency of pandemics in humans since the mid eighteenth century varies between 10-

40 years105, so it is not a matter of if, but when the next pandemic will arise. In addition to the risk to human health, outbreaks in domestic poultry or swine result in mass culling and significant economic impact99.

1.3.3.2 Viral entry and disease

IAV gains entry to the host via the nasal or oral cavities and comes into contact with the mucus coating the respiratory epithelium, which consists of highly glycosolated mucins rich in terminal sialic acids. IAV HA binds to the sialic acid residues becoming trapped in the mucus. Some virions are subsequently released by NA cleaving the decoy sialic acid residues allowing virion progression to the epithelial cell membrane binding to HA sialic acid side- chains of epithelial cell glycoproteins and glycolipids106. Different HA demonstrate differing tissue tropism. Typically human viruses have a tropism for α2,6-linked sialic acid to galactose whilst avian and equine viruses show a tropism towards α2,3-linked sialic acid to galactose106. This tropism towards alternative linkages of sialic acid residues to galactose restricts the virus to tissues rich in such linkages, namely the upper respiratory epithelia of humans and swine or the enteric tracts of avians107,108. Human lower respiratory tracts also contain α2,3-linked sialic acid to galactose residues on the epithelial cells which explains the ability of avian viruses to infect the lower respiratory tract of humans albeit rarely. Changes to the amino acid sequences of HA can result in altered tropism and even host specificity.

Once the virus reaches the respiratory epithelium it attaches to terminal sialic acid residues and is internalised via receptor mediated endocytosis. The virus passes through the membrane contained within an endosome where low pH triggers confirmation changes within

the virion, leading to delivery of the viral genome to the nucleus of the cell109. Host transcriptional and translational machinery is hijacked and new virus capsids are assembled and egress to the cell membrane and bud. NA removes sialic acid from cell surface releasing the new viruses to infect other cells. From the initial infected cell the virus spreads to both immune cells including macrophages, dendritic cells and mast cells, as well as non-immune cells.

1.3.3.3 Innate and adaptive immune responses to Influenza virus

Respiratory symptoms of Influenza are caused by the virus’s cytopathology of the respiratory epithelium and subsequent release of proinflammatory cytokines such at TNFa and IL-6. The release these mediators induces the recruitment of additional neutrophils, monocytes and natural killer cells to the lung. A range of cells produce these proinflammatory mediators including pneumocytes, epithelial cells, alveolar macrophages and tissue resident mast cells. Initially natural killer cells kill infected epithelial cells while recruited monocytes and neutrophils, along with alveolar macrophages phagocytose infected/dead cells110,111. If initial innate immune responses are insufficient to prevent the virus establishing a foothold, then the adaptive immune will be required to clear the virus.

Infected cells detect viral components via cytosolic and endosomal innate PRRs and present viral antigens within the MHC1 complex on their surfaces, triggering the production of inflammatory cytokines such as IL-1b, IL-18, and CCL2. These cytokine trigger neighbouring cells to release TGFb amplifying additional cytokine production by resident cells112. This response triggers the maturation of tissue resident dendritic cell and the recruitment and maturation of monocyte-derived dendritic cells. These cells take up antigen and migrate to the draining lymphatics entering lymph nodes via the afferent lymphatic, where they present antigens to T and B cells in the paracortex or germinal centres of the

lymph node respectively. If the T and B cells encounter antigen with specificity to their unique T cell receptor (TCR) or B cell receptor (BCR) they undergo a process of clonal expansion. Some T cells mature into effector cells and migrate back to the lung. Here they can induce T cell mediated killing of infected cells via perforin and granzyme containing granules113, or induce apoptosis via Fas, which augments innate immune responses clearing the infection. B cells generate specific antibody responses that neutralise viruses, halting their spread. They also opsonise virus directly and trigger complement mediated lysis of the virion.

Another role of antibodies is in antibody dependent cellular cytotoxicity (ADCC) in which the antibodies bind to surface expressed viral antigens flagging the infected cells to be killed by effector cells112.

1.3.3.4 Host cellular response

Following infection, both innate and adaptive immune responses are stimulated. Viral

RNA, the main PAMP associated with IAV, is detected by host cell PRRs. The endosomal toll like receptor TLR7, and in humans TLR8, and the NLR family pyrin domain containing 3

(NLRP3) protein all detect ssRNA, while Retonoic acid inducible gene I (RIG-I) detects viral

5′-triphosphate RNA114. IAV M2 protein has also been shown to activate the NLPR3 inflammasome pathway115.

Detection of viral PAMPs via these PRRs triggers proinflammatory cytokine and type

1 interferon production116. IFNb production is regulated by interferon regulatory factor 7

(IRF7) and amplified via a positive feedback loop, with both IFNa and IFNb being stimulated117. These type 1 interferons trigger antiviral states in neighbouring and infected cells by autocrine or paracrine signalling via heterodimeric IFNAR1 and IFNAR2 receptors.

Triggering of this signalling cascade results in the enhanced transcription of hundreds of different interferon responsive genes which mediates antiviral responses such as cell cycle

arrest, enhanced presentation of antigens via the upregulation of antigen processing machinery and MHC1, and dendritic cell maturation, enhancing the functions of CD8+ T cell and NK cells118. Collectively the interferon and cytokine produced following viral infection shape the adaptive immune responses driving antibody production and T cell activation. The magnitude of this response has direct consequences for the host with an excessive response, often referred to as a cytokine storm, being associated with increased risk of hospitalisation and higher rates of morbidity and mortality119.

1.3.3.5 Viral virulence factors

IAV non-structural protein 1 (NS1) has been shown to aid virus replication by inhibiting the activation of PRRs. NS1 contains a RNA binding domain that binds viral RNA masking it from intracellular PRRs such as TLRs and RIG-I, effectively delaying the onset of the interferon response and subsequent induction of an antiviral state116,119. NS1 disrupts

RIG-I viral detection and activation via binding to tripartite motif-containing protein 25 that ubiquninates RIG-1 caspase recruitment domain. It also forms a complex with RNA- dependent protein kinase (PKR), preventing dsRNA detection and halting PKR mediated translational arrest and subsequent disruption to viral protein synthesis116. Another IAV protein, nuclear protein (NP) also disrupts PKR, activating its inhibitor P58ipk through disassociating it from heat shock protein 40 (hsp40). Once disassociated, P58ipk binds to

PKR, resulting in its inactivation120. Another IAV protein, M2, also activates P58ipk but by binding to hsp40, preventing its association with P58ipk and results in inhibited protein synthesis, apoptosis and release of virus116,121.

1.3.3.6 Treatment and prevention

Control measures such as annual vaccinations using either live attenuated or inactivated Influenza vaccines comprised of several of the currently circulating virus subtypes offers some protection. There has been many studies focusing on developing a universal Influenza vaccine however as yet and despite some progress, a universal vaccine is not available122. Current vaccines do generate virus specific antibody response and result in a protective B cell memory population123, unfortunately protection in this population typically declines rapidly, with reduced antibody titres falling over time to below a concentration capable of protecting against a subsequent infection124. Additionally whilst vaccination provides full protection in 75-80% of healthy vaccinated individuals, the vaccine efficacy drops to around 40% in some care home patient populations125.

The use of live attenuated influenza vaccines is preferentially recommended in non- asthmatic children from 2 -18 years of age in many countries, and elicits a longer lasting and broader humoral immunity than seen in the inactivated virus vaccine126.

Vaccine generation takes approximately 6 months to reach production from emergence of a new subtype, leaving the population unprotected during the initial waves of a new pandemic127. Two classes of anti-viral medication are currently available to help fill this prevention gap. The first, amantadines, inhibit the function of the M2 ion channel preventing the influx of H+ into the virion, preventing the uncoating of the virion. Currently these drugs are only effective on IAV as IBV do not contain M2 ion channels125. The second class, neuraminidase inhibitors target the active site of neuraminidase, preventing it digesting the hemagglutinin receptors binding the new virions to the cell, halting subsequent budding and infection of other cells125.

A recent review by the Cochrane institute found no evidence of important public health benefits that would warrant the widespread use of the neuraminidase inhibitors

oseltamivir and zanamivir in infected patients. They also report that whilst these drugs are effective in a prophylaxis setting, the balance between benefits and harm need to be carefully considered128.

Furthermore, wide spread resistance to antivirals is a growing concern. In 2003 a

S31N M2 gene mutation conferring amantadine resistance began circulating in south east

Asia. The mutation continued to spread in countries that prescribed amantadine sparingly, such as Australia, suggesting the resistance mutation was spreading in the absence of drug pressure129. These seasonal IAV are now extinct, having been replaced by A(H1N1)pdm09, which due to its Eurasian Swine linage, also had the S31N mutation thus the 2009 pandemic flu strain was resistant to amantadine. At present all currently circulating influenza viruses show amantadine resistance.

Mutations that convey resistance to neuraminidase inhibitors, such as H275Y NA mutation that results in resistance to oseltamivir have also been observed in seasonal IAV infections, albeit at a relatively low frequency (~1%)129. This mutation, together with a few others that convey resistance129, demonstrate that our current antiviral treatment options are limited and could rapidly become exhausted.

More recently, therapeutic interventions with the aim of reducing the negative impact of the host defence caused by responses to the pathogen130 have been explored. Excessive leukocyte infiltration into the lung during IAV infection as a result of a cytokine storm has been shown to cause widespread tissue damage and alterations to the underlining lung architecture. Modulating this inflammation using COX2 (cyclooxygenase 2) inhibitors reduces immunopathology driven by excessive prostaglandin production during infection without impacting pathogen burden130. Studies in mice show the cholesterol reducing drug gemfibrozil possesses additional therapeutic properties, including inhibiting the release of inflammatory cytokines131 and increased survival rates following IAV infection132. Another

study using an analogue of the cell membrane component sphingosine during IAV infection found that the analogue, AAL-R, reduced cytokine production while maintaining a protective neutralising cytotoxic T cell and antibody response. Additionally, there was less leukocyte infiltration into the lungs and also less pulmonary tissue damage. Inhibiting mast cell degranulation also provided protection from tissue damage133, and mast cell deficient mice showed reduced inflammatory cytokine production and lower levels of subsequent immunopathology134. Collectively these data outline the need to develop timely and effective interventions in the treatment of Influenza infections due to the lack of effective acute current therapies.

1.4 The role of mast cells in respiratory infections

Mast cells differentiate from myeloid progenitor cells in the bone marrow. Mast cells and Basophils differentiate from a common bipotent basophil/mast cell progenitor, innate mast cells migrate from the bone marrow to prior to connective tissues135, where upon exposure to either surface expression or secreted stem cell factor (SCF)136 and the localised cytokines IL-3, IL-4, IL-9, and IL-10, mature into mast cells137. Once mature, mast cells develop characteristic features such as large electron dense granules2, filled with mast cell restricted proteases, such as chymases, tryptases and carboxypeptidase A3, together with growth factors, lysosomal enzymes, proteoglycans of serglycin, preformed cytokines (such as

TNFa and IL-4), and biogenic amines including histamine and serotonin1. Protease serglycin complexes, stored in the secretory granules account for approximately 50% of the weight of a mast cell138.

While they are synonymous with deleterious effects in allergic responses, evidence suggests that they may also have a beneficial role in some situations. Mast cells act as the sentinel of the immune system, being strategically located at host interface environments.

Their expression of a wide range of PRRs means they can detect the presence of a wide range of pathogens and release numerous proinflammatory mediators to recruit addition cells or aid in direct killing of invading pathogens1.

Mast cells function as professional phagocytes, they can phagocytose opsonised

(compliment or antibody) bacteria and destroy them via oxidative and non-oxidative bactericidal mechanisms139,48,140, as well as contribute towards microbial defence by releasing meditators such as cathelicidins that act as antimicrobial proteins and recently mast cells have been found to release extracellular traps141.

They can be broadly divided into two groups based on whether they contain heparin proteoglycans or not. Their granules can contain a combination of chymases and tryptases,

with the exact composition of granules being determined by the milieu in which the mast cell matures. In the mouse, mast cell proteases (mMCP) 1, mMCP2, mMCP4 and mMCP5 are chymases, whereas mMCP6 and mMCP7 are tryptases142. The expression of mMCP1 and mMCP2 is restricted to mucosal mast cells where their production is promoted by the presence of IL-10. The remaining chymases, mMCP4 and mMCP5 are restricted to the granules of connective tissue mast cells, as are the two closely related tryptases mMCP6 and mMCP7143. Human mast cell granules contain 4 tryptases α1, βI, βII & γ1 together with chymases and cathepsin G144. Unlike mast cells from mice, pulmonary mast cells from humans lack chymase expression, instead chymase positive mast cells are typically found close to glands144.

While mast cell numbers are relatively constant within these connective tissues, mast cells from mucosal tissues show considerable variation in numbers under steady state conditions145.

1.4.1 Mast cell activation and degranulation

Mast cells are activated by various mechanisms, the best characterised via FcεRI cross linking. FcεRI has a very high affinity for IgE consequently almost all IgE is bound to these receptors, with very low concentrations remaining in the serum. Mast cells with IgE bound to their FcεRI are considered sensitised and the binding of antigen to these antibodies triggers cross linking within the receptor cluster, initiating a signalling cascade resulting in mast cell degranulation146.

Mast cells can also be activated by other stimuli including complement derived anaphylatoxins c3a and c5a, IgG, neuropeptides, venom components, bacterial cell wall components and pathogens derived peptides1. TLR ligands can also trigger mast cell

activation without degranulation resulting in the recreation of lipid mediators together with cytokines and chemokines.

Full activation induces immediate degranulation: the release of pre-stored mediators including histamine, serglycin, serotonin and proteases, is followed within a few hours by the de-novo synthesis of mediators such as prostaglandins, leukotrienes and proinflammatory cytokines and chemokines145. This delayed response produces an exceptionally potent inflammatory response.

1.4.2 Role of mast cells in bacterial infection

Experimentally, the roles of mast cells in infections have been studied both through genetically mutated mice and chemical inhibition of their activation. Mice with a mutation in the c-Kit gene (kit), referred to as WBB6F1-W/Wv mice, have defective signalling though c-

Kit, as a consequence they are unable to respond to the vital mast cell growth factor SCF. The subsequent lack of a growth signal prevents mast cells from maturing and results in an almost total ablation of mast cells in the mouse147. Unfortunately given these mice are semi- syngeneic with C57-bgJ/bgJ "beige" mice, they share the predisposition of Chediak-Higashi syndrome as a result in a mutation in the lysosomal trafficking regulator protein148,149.

Additionally these mice also present with several other phenotypic abnormalities including neutropenia, sterility, macrocytic anaemia impaired melanogenisis, and almost absolute absence interstitial cells of Cajal150,1.

Studies using this mouse showed mast cells are essential for protection from enterbacteria induced sepsis following cecal ligation and puncture (CLP) with mast cell deficient mice being 20 fold less effective at clearing enterobacteria151. Following i.p injection of Klebsiella pneumoniae or Escherichia coli mast cell deficient mice demonstrate impaired neutrophilic inflammatory responses. As mast cells are a major producer of TNFα

their absence in these model results in impaired recruitment of neutrophils. The group also reported that other neutrophil chemoattractants such as leukotriene B4 and IL-8 play a significant role in mast cell induced neutrophilic inflammation following bacterial challenge, demonstrating that mast cells can modulate neutrophilic inflammation following detection of bacteria140. Other models have also shown that mast cells are protective against the Gram negative bacteria Citrobacter rodentium152, P. aeruginosa153, Helicobacter pylori154,

Francisella tularensis155 and the Gram positive bacteria’s Listeria monocytogenes156 and

Streptococcus pyogenes157.

More recently an alternative mast cell deficient mouse model has been developed on a

C57 mouse background (C57 KitW-sh/KitW-sh)158 mice. The specific mutation, W-sh, is an inversion in the transcriptional regulatory elements of kit159 and the resultant mice have fewer phenotypic abnormalities than the WBB6F1W/Wv mice. They are fertile, yet show impaired skin pigmentation together with profound mast cell deficiency in multiple tissues160, they also show elevated numbers of neutrophils and basophils161. These mice are not protected in a moderately severe CLP model161 and also show that mast cells are protective in Mycoplasma pneumoniae and E. Coli infections162,163.

Whilst demonstrating the protective nature of mast cells in moderately severe CLP,

Piliponsky et al. also showed that mast cells can have a negative role in a more extensive

CLP model161, demonstrating the contextual nature of the mast cell response to infection.

Mast cell generated TNFa appeared to be the major factor associated with a detrimental outcome in this study.

The detrimental role of mast cells in CLP was also confirmed in another study utilising an inducible mast cell deficient mouse, these mice have been genetically altered to add a cassette composed of an internal ribosomal entry site together with the bright red td-tomato (tdT) fluorescent protein, a cleavage sequence (2A) and human diphtheria toxin

receptor (hDTR) to the 3′-UTR of Ms4a2, encoding FcεRI β chain164. Their mast cells (and basophils), both expressing FcεRI β, are fluorescent red and can be depleted by the addition of diphtheria toxin to the water of the mice. This depletes all FcεRI β chain expressing cells, with basophil numbers returning to normal with 1 week, yet mast cells numbers remaining depleted for months, indeed at 2 months only 6% of normal mast cells numbers seen rising to

50% by 6 months164. This group found that mast cells were detrimental in a CLP model and this was attributed to mast cell secreted IL-4 that suppressed the phagocytic activity of macrophages.

Several studies have shown the negative role mast cells can play during infections. In an E. coli bladder infection model tissue resident mast cells and the IL-10 they secrete was shown to suppress humoral and cell mediated responses within the bladder resulting in bacterial persistence165. In a Streptococcus pyogenes dermatitis model mast cells exacerbate the proinflammatory effects of streptolysin O166, whilst in a Staphylococcus aureus infection

TLR2 and NOD1 detection of bacterial peptidoglycan (PGN) was shown to induce degranulation releasing mediators, including histamine and serotonin, that had synergistic effects in PGN induced diarrhoea167. In a S. aureus dermatitis model mast cells were shown to contribute to IL-4 production, aggravating the inflammation seen in this model168.

1.4.3 Role of mast cells in viral infection

The role of mast cells during viral infections is less clear than that in bacterial infections. Mast cells can directly recognise several viruses including Influenza, Dengue,

Sendai and HIV resulting in varying degrees of activation. Their effects however may be due to induction of other cell mediated effects, such as the promotion of CD8+ cells recruitment together with inducing type 1 interferon responses169.

Mast cell activation and degranulation have been shown to contribute towards immunopathology during H5N1 Influenza infections in mice. Using a mast cell stabiliser, which inhibits degranulation, Hu et al. demonstrated reduced immunopathology and enhanced survival133. A study by Graham et al. showed that KitW-sh/W-sh mice were protected from significant disease and immunopathology when infected with H1N1

A/WSN/33. The group showed that mast cells can be infected by this viral strain, however the virus cannot replicate in these cells. Recognition of viral dsRNA in a RIG-I dependent manner results in mast cell activation134. Transcriptome analysis by Josset et al. comparing seasonal IAV infections with highly pathogenic infections identified genetic signatures for macrophages and neutrophils associated with severe IAV infection. They also identified a third signature, that of mast cells170, but the authors didn’t explore this result, allowing other groups to postulate that together with macrophages and neutrophils, mast cells contribute towards the excessive inflammatory responses and increased vascular permeability seen in the highly pathogenic IVA infections171.

Another infection mast cells have been associated with is the mosquito bourne flavivirus, Dengue virus. The virus can induce mast cell degranulation and subsequent release of cytokines and chemokines, altering vascular permeability. In animal models of dengue the treatment with mast cells stabilisers restore vascular integrity and infection in C57 KitW- sh/W-sh mice further demonstrated the detrimental role mast cells play during the infection, with mice absent of mast cells showing reduced symptoms and immunopathology. These findings correlate with observations in humans where there is a correlation with mast cell activation and disease severity, indeed mast cell chymases is a predictive marker for dengue haemorrhagic fever172. These observations correspond with those seen during Newcastle disease virus infection in poultry where mast cell degranulation is attributed to the intestinal tissue damage and decreased survival173.

The parainfluenza virus Sendai virus is highly transmissible in both rodents and swine, resulting in epizootic infections. Infection results in proliferation of bronchiolar mast cells and recruitment of circulating progenitors independent of mast cell degranulation. The increase in numbers persisted for several months and contributed towards increased airway responsiveness following infection174.

Mast cells are essential for effective control of Herpes Simplex Virus 2 (HSV-2).

Infection in KitW/Wv mice demonstrated increased viral titers, elevated clinical severity and higher mortality rates. These hallmarks of infection were reversed following reconstitution of these mice with wild type bone marrow derived mast cells but not bone marrow derived mast cells from TNFa-/- or IL-6-/- mice, indicating that mast cell derived TNFa and IL-6 are required for the effective control of the virus175.

TLR2 and 4 agonists, peptidoglycan (PGN) and LPS, have been shown to induce the production of Th2 skewing cytokines (IL-5, IL-10 & IL-13) and inflammatory mediators

(TNFα & histamine), in human cord blood-derived mast cells176. Yet, LPS stimulation alone failed to induce degranulation directly, and required IL-4 and serum to induce TNFα.

Conversely, PGN was able to induce degranulation, measured by histamine release, without

IL-4 priming or the presence of serum176. This capacity of PGN to induce mast cell degranulation via TLR2176 indicates an important role for bacterial components in inducing mast cell degranulation. Finally, LPS is known as a strong activator for mast cells inducing

TLR4 activation and subsequent expression and secretion of TNFα and IL-6 into the milieu177.

1.4.4 Role of mast cells in S. pneumoniae infection

The role mast cells play in the pathogenesis of S. pneumoniae has been investigated by several groups. Exposure to the bacteria induce, mast cell degranulation without triggering the release of preformed TNFa or de novo synthesis of TNFa and IL-6. This induction of degranulation minus preformed cytokine release was postulated to be an evasion strategy by the bacterium and has only observed in live infections 60. In vitro, human mast cells have been shown to kill S. pneumoniae directly via secretion of the cathelicidin LL-37 in response to pneumolysin, this limits the dissemination of the bacteria in the early stages of an invasive infection48.

The prevailing notion in the literature is that mast cells are generally beneficial in experimental infections by promoting bacteria clearance and reducing lethality. However, more recent data show mast cells, in a S. pneumoniae model, play a negative role, and that this is detrimental activity is independent of mast cell degranulation suggesting little to no role for mast cell protease in this infection59.

Taken together the literature clearly demonstrates that mast cells play an important role during S. pneumoniae infection, however the precise role is unclear with contradicting reports on the roles these cells play and inconsistent reports on how degranulation contributes towards the pathogenesis of this infection. This lack of clarity warrants further study.

1.4.5 Role of mast cells in P. aeruginosa infection

The role mast cells in a P. aeruginosa infection has also been investigated by several groups. In vitro studies of both murine bone marrow derived and human cord blood derived mast cells show exposure of mast cells to P. aeruginosa results in the activation and production of IL-6 in mast cells and that this activation is mediated by protein phosphatase

2A and protein kinase C which are physically associated in resting mast cells. It has been

suggested that an inability to quench the excessive IL-6 production in mast cells contributes to the excessive inflammation seen in P. aeruginosa mediated pneumonia178.

Whilst typically considered an extracellular bacterium, during invasive infections

P. aeruginosa acquires the ability to enter and grow within host cells. In vitro studies show

P. aeruginosa induces autophagy in mast cells which leads to bacterial death. In addition, the study also demonstrated that in vivo P. aeruginosa respiratory infection induced mast cell degranulation179. Later studies by the same group demonstrated that maintaining the epithelial integrity of the respiratory tract is essential for an effective host response, and that mast cells were essential in maintaining this epithelial integrity during infection. They secrete an as yet unknown factor that is essential for the maintenance of this integrity in a P. aeruginosa model94.

Again, the literature supports the notion that mast cells play an important role in the pathogenesis of this infection. Additional studies are required to elucidate the precise role played by these cells and their secreted factors in a P. aeruginosa respiratory infection.

1.4.6 Role of mast cells in Influenza infection

Mast cells have traditionally been overlooked in relation to viral infections. However recent studies by Hu et al. demonstrate that they have a detrimental role. Inhibiting mast cell degranulation during H5N1 IAV infection protected mice from extensive immunopathology caused by viral induced apoptosis and high mortality rates133.

Other studies by Graham et al. utilising a H1N1 A/WSN/33 demonstrated that in vivo mast cells are directly infected by the virus but the virus fails to replicate in these cells134. The infection activated the mast cells triggering degranulation directly, an observation that was repeated using other viral isolates suggesting a common feature of IAV and IBV. The group also confirmed the role mast cells play in immunopathology. Mast cell deficient mice had

less proinflammatory cytokines and chemokine production, resulting in less inflammatory cell recruitment and subsequent lung pathology and vascular leakage134. Taken together these two studies point to a mast cell mediators being released following mast cell degranulation that is involved in excessive inflammation and immunopathology.

1.4.7 Mast cell proteases

Approximately 50% of the protein content of mature connective tissue mast cells comprises of 16 neutral proteases density packed within the acidic granules ionically bound to chondroitin sulfate or heparin180. Therefore, these molecules are likely to have substantial impact when released following mast cell degranulation. There are 3 broad classes of proteases associated with mast cells. These include the chymases, tryptases, and carboxypeptidase A which are based partly on their particular substrate specificity. The chymases, which are chymotrypsin like proteases, cleave after aromatic amino acids and trypsin like proteases typically cleave after Lys/Arg residues. In addition, a number of tryptase related molecules have be shown to be important in disease. These proteases are tightly regulated with their expression, activation and storage defined in molar amounts180.

An abundance of inhibitors control their enzymatic functions outside the cell and macrophages, fibroblasts, and endothelial cells have been shown to endocytose these proteases and subsequently degrade them.

1.4.8 Mast cell tryptases

Mast cells tryptases have been associated with several inflammatory conditions including asthma181, rheumatoid arthritis182, ulcerative colitis and Crohn’s disease183. Mast cell-derived human β-tryptase levels correlate with COPD severity184, and in mice, cigarette smoke exposure results in mMCP6 release185. Previous studies within the Hansbro group found that mMCP6 was involved in the development of COPD in cigarette smoke induced

COPD model in mice186.

In humans approximately 23% of the population have a-tryptase deficiency, dividing data into ethnicity demonstrates wide variations with Caucasians showing a tryptase deficiency rates of 45%, whilst Chinese-American rates are just 10%187. It is believed that aabb heterotetramers exist, yet this is to be confirmed in-vivo, however their existence would infer variable functional tetramers with ab heterotetramers having reduced catalytic activity versus a homotetrameric bbbb tryptase that would have the highest catalytic activity188. The impact this variation in tryptase catalytic activity has in inflammatory diseases is unknown and warrants further study.

a and d tryptases have catalytic domain defects resulting in reduced activity188. a-tryptases have key mutation in the active site whilst d have a premature stop codon losing the C-terminal region containing resides believed to be essential for specificity and catalytic competence188. They also seem to contain a mutation within their pro-peptide that prevents effective autolytic processing and as a consequence they are often secreted inactive. These mutations and deficiencies make a and d tryptases non priority candidate to investigate further.

Mice express 2 mast cell b-tryptases mMCP6 and mMCP7, and one g-tryptase,

Prss31. The b-tryptases share 78.4% and 76.7% sequence homology to their human orthologs

β-tryptases, respectively143,189 and 76% homology between one another190. The g-tryptase shares 45% and 46% sequence homology with mMCP6 and mMCP7 respectively. b-tryptase expression is not exclusive to mast cells, they are also present within basophils188.

In humans, the mMCP6 homologue, TPSB2, together with the homologue for mMCP7, TPSAB1, are the predominant proteases present within mast cell granules180. mMCP6 and mMCP7 are translated as zymogens which possess hydrophobic signal peptides together with a 10 amino acid pro-peptide that is proteolytically cleaved resulting in the preferential formation of homotypic tetramers, with the active sites of each monomer facing towards a central core191. Heterotypic tetramers have been reported in vitro192. These tetramers are subsequently stabilised by ionically binding to serglycin proteoglycans, which collectively make up 50% of the mass of tissue mast cells138, they are stored within the low pH environment of mast cell granules193. Additionally, active monomers of each tryptase have also been reported194.

Although the two tryptases share significant sequence homology, in vitro studies suggest that they have different substrate specificities195. mMCP6, unlike other MCPs, has a binding site for heparin very close to its active site, suggesting that heparin glycosaminoglycans act sterically, altering its substrate specificity195. To date no specific substrate for mMCP6 has been described2. mMCP7, also referred to as mature mast cell tryptase-2142 is functionally distinct to that of mMCP6195, despite sharing 71% primary amino acid sequence homology196. It has differences in substrate specificities with fibrinogen being identified as one of mMCP7s substrates, indicating that it may play a role in anti- coagulation197. Expression of tryptases in mice is also species specific. C57 mice have an exon2/intron2 splice site mutation in their mMCP7 gene rendering the gene inactive193.

Therefore, the generation of a mMCP6-/- mouse on a C57J background138 yields a double

tryptase knockout that allows investigations into the roles of mast cell tryptases in a number of different conditions.

I have shown that mMCP6 contributes to COPD in a novel smoking model developed in our laboratory186. mMCP6-/- mice have significantly less macrophages and neutrophils in

BALf versus controls and reduced emphysema. mMCP6-/- also have decreased neutrophilic responses in Klebsiella pneumoniae infections and as a result fail to clear the bacteria138. The addition of human recombinant β1-tryptase prior to K. pneumoniae infections led to a reduction in viable bacteria recovery in a mouse mast cell knock out model193. mMCP6 alongside mMCP7 was shown to also be involved in the development of experimental rheumatoid arthritis182.

mMCP6 does not readily dissociate from its serglycin proteoglycan, consequently it remains within the extracellular matrix for extended periods. Murine serum contains inhibitors to mMCP6195 but to date no specific mMCP6 or human β tryptase inhibitor has been identified in blood138. mMCP7 however shows pH dependent binding to serglycin proteoglycans, demonstrating strong binding in the low pH (5.5) seen within mast cell granules yet readily dissociates with serglycin proteoglycan at neutral pH. As a consequence, degranulation results in mMCP7 dissemination away from the mast cell and can be detected in the circulation190 where it is free to cleave the a chain of fibrinogen, helping to reduce fibrin/platelet mediated clotting during mast cell induced inflammation197.

Studies have shown that intraperitoneal injection of heparin bound recombinant- mMCP6 and human recombinant β tryptase into the lung or peritoneal cavity of mice results in neutrophilia, but does not increase eosinophils, basophils, platelets or mast cell numbers195,193. Intraperitoneal injection with mMCP7 does not elicit the same response and does not induce neutrophilia195. This prolonged mMCP6 retention and rapid inactivation in

serum would infer the tryptase has a preferential effect locally195, acting to recruit neutrophils.

Other experiments show that tryptases are required to be bound to serglycin proteoglycan to be effective. Neutrophils failed to migrate to human endothelial cells following exposure to recombinant mMCP6 or mMCP7. However, when heparin bound mMCP6 and human endothelial cells were co-incubated, neutrophil migration occurred.

Subsequent investigations demonstrated that the human endothelial cells, when co-cultured with mMCP6 bound with heparin, secrete CXCL8, but no RANTES or TNFα production192.

Taken together, these data infer that mMCP6 possesses proinflammatory functions that are orchestrated via the activation of bystander cells and results in the production of chemokines for neutrophils such as CXCL1, CXCL8, MMP-3, MMP-13143.

As mMCP6 and 7 are typically found within mast cells (and basophil) granules, together with other chymases and tryptases, purification of sufficient quantities for functional studies is challenging. One alternative method that is used to investigate mMCP6 and mMCP7 is to use a bacclovirus system to make recombinant pseudozymogen tryptase that can be activated ex vivo using an enterokinase195. Another method commonly used is to culture bone marrow in the presence of IL-3. However, caution must be used when considering this method, as connective tissue mast cells contain predominantly heparin and small amounts of chondroitin sulphate (CS) whereas bone marrow derived mast cells contain predominantly CS with low levels of heparin198. This shift in the glucosaminoglycan (GAG) could result in changes to mMCP6 substrate specificity, making functional studies difficult to interpret. A more prudent methodology for studying mMCP6 could be to culture bone marrow derived mast cells with 3T3 fibroblasts, which has been shown to yield mast cells with higher levels of heparin than IL-3199.

The g-tryptase Prss31 is a mast cell associated serine protease that is membrane bound200. Whilst being present and active in humans and mice g-tryptases are not ubiquitous in mammalian populations. Additionally Prss31 transcription in mice is strain dependent with

C57 mice expressing the highest whilst other stains have little to no expression, additionally tissue specific expression patterns are also seen with higher amounts detected in the skin and lower amounts in the lung196. Prss31 g-tryptase may not play as an important role in all mammalian biology given its absence in the genomes of the dog and chimpanzee188 but could play in role in human biology given that we have retained this gene.

Following mast cell activation and subsequent degranulation this mast cell protease remains anchored to the outer membrane of the mast cell by a unique c terminal membrane spanning peptide200, restricting Prss31 activity to cells that are in physical contact with mast cells.

Whilst Prss31 has some overlapping substrate specificity with b tryptases, studies have shown that g-tryptases have substrate preferences for P1 Arg over Lys, P2 aromatic residues, and P3 basic residues200. Profiling the effects of inhibitors on g-tryptases showed that this tryptase was inhibited by serpin-class inhibitors that b2 tryptases are resistant to suggesting an active site distant from that of b tryptases that has more similarities to trypsin than other tryptases201,188.

In a study by Fricker et al., Prss31 null mice were shown to have increased baseline airway reactivity to methacholine. Additionally, Prss31 was found to be detrimental during a cigarette smoke induced COPD model with Prss31-/- mice having reduced associated inflammatory cell influx in the bronchoalveolar lavage fluid (BALf), lower histopathological scores and were protected from collagen deposition associated with the model. The group also showed that Prss31-/- mice are protected to a degree from experimental colitis.

Collectively this data points to a proinflammatory role for Prss31202.

1.4.9 Mast cell chymases

mMCP5 is one of 4 chymases expressed in the mouse, while humans only express one. The most phylogenically and structurally similar chymase in both species is being mMCP5203. mMCP5 has elastase like cleavage properties and is sometime referred to as an elastase204.

Interestingly mMCP5 knockout mice lack protein expression of carboxypepsidase A3

(CPA3), this observation was reciprocated in a CPA3 knock out mouse that lacked mMCP5 protein expression. This interdependence at the protein level makes determining observations seen in mMCP5-/- difficult to interpret alone without considering CPA3205.

Purified human chymase from skin induced neutrophilic and esonophilic inflammation in guinea pigs with increases of 60 fold and 12 fold respectively. This proinflammatory role was much more pronounced when repeated in BALB/c mice, with just

5 ng of injected chymase eliciting a 700 fold increase in neutrophils, 21 fold increase in eosinophils, 19 fold increase in lymphocytes and a 7 fold increase in macrophages at 16 hours206. This data clearly points to a proinflammatory role for this chymase.

Further studies using an mMCP5 null mice show they are protected from tissue injury following 2nd degree burns207, and mMCP5 was show to have cytotoxic activity in a

Ischemia with subsequent reperfusion (IR) injury model208.

To date the precise role of human chymase is unclear, it is associated with CPA3 protein expression and induces a robust inflammatory response in different species. This ability to modulate inflammation could be important during infectious disease.

1.4.10 Mast cell related tryptases

Human e-tryptase, PRSS22, is abundantly expressed in epithelial cells, located in the adult trachea and oesophagus. It is not abundantly expressed in adult lungs but is extensively found in foetal lung209. It shares 40% amino acid sequence homology with the b tryptases I and II yet structural studies indicate this tryptase has substrate specificities different to those of other tryptases210.

Studies point to PRSS22 having a restricted substrate specificity and the pro-tryptase can autoactivate211. The tryptase is not inhibited by any know protease inhibitors found in human plasma or a1-antitrypsins or secretory leukocyte protease inhibitors present in the lung. It has been shown to activate the serine protease urokinase type plasminogen activator

(uPA). uPA activated the serine protease plasmin by cleaving its inactive form plasminogen, this triggers a cascade that promotes the degradation of the extracellular matrix and fibrinolysis211. The ability of PRSS22 to autoactivate and to initiate this cascade suggests that together with its ability to promote the migration of smooth muscle cell, this tryptase may play a role in airway remodelling211.

1.4.11 Mast cell derived Heparin

Heparin is the most potent anionic polyelectrolyte present in mammals and is a major component of secretory granules in mast cells212,213. A glucosaminoglycan (GAG) exclusively produced by mast cells of connective tissue origin, its hexuronic acid residues, unlike other GAGs, are restricted to D-glucosamine and L-iduronic acid212. Glucosaminyl

N-deacetylase/N-sulphotransferase-2 (NDST2) has been shown to be essential in the N- deacetylation and N-sulphonation of heparin214.

Heparin when nebulized and given to asthmatics was shown to improve FEV1 scores and reduce eosinophils and lymphocytes in the BAL 14 days later215. Another study found

heparin to be effective in preventing bronchoconstriction in exercise induced asthma216. Both studies postulated that the mechanism of action for heparin was the modulation of mediator release than direct action on smooth muscle.

Whilst the therapeutic benefits of heparin in treating asthma are clear, its use in other inflammatory conditions is less so. Two separate trails of heparin treatment in ulcerative colitis patients showed contradictory findings in the efficacy of heparin as a treatment option in this disease217,218.

The role of heparin in infectious disease is less studied, although it is known that heparin sulphate proteoglycans play a role cell surface adhesion for many microbes219-221.

1.4.12 Mast cell factor NDST2 in mast cell granule composition and tryptase activity

In a novel model, Forsberg et al.214, disrupted the NDST2 gene by generating a knock out mouse on a C57 background. These mice lack sulphonated heparin, demonstrating the requirement for NDST2 in the production of biologically active heparin. Additionally, the prevalence of mast cells in NDST2-/- mice is reduced and the protease repertoire of those that remain is altered. Whilst comparable mRNA expression of mMCP2, 4, 5 and 6 was detected in NDST2-/- vs NDST2+/+ mice, tryptase and chymase activity was almost absent in the former. Indeed, transmission electron microscopy of NDST2-/- mast cells reveal few granules and empty vacuoles214.

Proteases, both tryptases and chymases bind to serglycin bound heparin on serglycin proteoglycans213 within the granules of connective tissue mast cells169. However, other mast cells (e.g. mucosal) who’s granules lack heparin, contain chymases and tryptases bound to

CS222. NDST2 plays no role in the biosynthetic pathway for CS, which is verified by the observation that mucosal mast cells from NDST2-/- mice contain electron dense granular structures, consistent with chymase containing granules214. 70

In NDST2-/- mice from a non C57 background, bone marrow derived mast cells express normal levels of enzymatically active mMCP6213. Conversely, NDST2-/- on a C57 background have connective tissue mast cells that produce mMCP6 transcripts but no protein by western blot analysis214. As heparin is required for stabilisation of mMCP4, 5 & 6 together with CPA3 these are all absent from secretory granules of mast cells that would reside in connective tissues, including the lung. Thus, mMCP6 protein detection differs in the different mouse strains with NDST2-/-. A possible explanation may be that non C57 mouse strains contain mMCP7, which can form heterotetramers with mMCP6, possibly stabilising mMCP6 in the absence of heparin205,208.

Given the multiple mast cell protease deficiencies caused by knocking out NDST2, most research utilising these mice have been in the context of these associated proteases, for example, the roles of mMCP6 and mMCP7 in experimental arthritis182.

1.4.13 Mast cell related factor RasGRP4

Ras guanine nucleotide releasing protein 4 (RasGRP4) is guanine nucleotide exchange factor and diacylgycerol/phorbol ester receptor typically located in the cytoplasm and systolic side of the plasma membrane. This association with the plasma membrane points towards its involvement in early signalling events223. It is a member of the Ras-GRP family of guanine nucleotide exchange factors (GEFs), but shares less than 50% of its amino acid sequence with its closest family member224. This protein activates members of the Ras superfamily of GTP-binding proteins and has been shown to be a signal transduction protein that contributes to the final stages of mast cell development but it is not essential for the development of mature mast cell granules225. Whilst the specific singling pathways RasGRP4 participates in is unknown, studies suggest it is involved in in CD117 (c-Kit) signalling224.

RasGRP4 was until recently believed to be restricted to mast cells and their non- granulated progenitors224, however, recent data has highlighted the expression of RasGRP4 in

CD117+ splenic dendritic cells (DC). In RasGRP4-/- mice DCs mediated stimulation of natural killer cells resulted in the production of reduced amounts of the proinflammatory cytokine IFNg in response to LPS225.

Gene linkage studies places human RasGRP4 within an asthma susceptibility locus226 and animal models of experimental arthritis and colitis show that RasGRP4 deficient mice were protected from these inflammatory conditions223, collectively indicating that RasGRP4 promotes inflammatory disease conditions.

1.5 Mouse genetics

In this study I have used 7 different transgenic mice and wild type controls.

1.5.1 C57 mice

These wild type mice have an exon2/intron2 splice site mutation in their mMCP7 gene rendering the gene inactive193.

1.5.2 mMCP5 null mouse

mMCP5 knock out mice were generated by utilising homologous recombination to disrupt Cma1 in embryonic stem cells from 129Sv mice, these were then placed in C57 blastocysts and backcrossed with wild type C57 mice. The resulting heterozygous mMCP5-/+ mice were back crossed 3 times and mMCP5-/- mice selected and backcrossed another 4 times. These mice were utilised in the studies of schema reperfusion injury by Abonia et al.208. Prior to these mice being transferred to Australia, these mice were back crossed another 6 times and the mMCP5-/- mice used in this study have been backcrossed a total of 10 times.

mMCP5-/- mice lack expression of mCPA3, thus any results seen in these mice need to be considered in the context of the combined absence of the serine protease mCPA3.

1.5.3 mMCP6 null mouse

The generation of the mMCP6-/- mouse was first described by Thakurdas et al.138.

Briefly homologous recombination was used to disrupt mMCP6 replacing the first 3 exons with Cre recombinase resulting in a mMCP6 lacking critical amino acids for catalytic function. The resultant protein is non-functional.

This mouse differs from mMCP6 mice used by Shin et al.227 in that our mouse have a disrupted mMCP6 from a C57 background whereas Shins et al’s mice have used homologous recombination within an embryonic stem cell from 129Sv mice. These mice unlike the C57 mice used in this thesis, express mMCP7 and very little Prss31, meaning that the mMCP6-/- mice in their studies will be mMCP7+/+ and Prss31nil unlike our mice who are mMCP6-/- mMCP7-/- Prss31+/+.

1.5.4 mMCP6 null, mMCP7 knock in mouse

As mentioned above, Shin et al.227 disrupted mMCP6 using homologous recombination within an embryonic stem cell from 129Sv mice and back crossed these mice with C57 mice. These mice were subsequently mMCP6-/- but due to their 129Sv back ground, these mice were also mMCP7+/+ and Prss31 nil. These mMCP7+/+ mice were used during this study.

1.5.5 Prss31 null mouse

A Prss31-/- mouse was generated as described by Fricker et al.202. Briefly the translation initiation site within exon 1 was removed by homologous recombination and a

Cre/loxP cassette inserted to embryonic stem cells from C57 mice. These were then placed in

C57 blastocysts and backcrossed with wild type C57 mice. The resulting phenotypes of these mice are mMCP6+/+, mMCP7-/- and Prss31-/-.

1.5.6 NDST2 null mouse

NDST2-/- mice were first described in 1996213. They were generated using homologous recombination to disrupt NDST2 in embryonic stem cells from 129Sv mice.

These were then placed in C57 blastocysts and backcrossed with wild type C57 mice. The

absence of sulphated heparin resulted in mice with mast cells that lacked mMCP 4, 5, and 6 together with mCPA3 but contained lots of mMCP7213.

1.5.7 Prss22 null mouse

A Prss22-/- mouse was generated using a Cre/locP homologous recombination to delete the first 4 exons of Prss22 in embryonic stem cells from C57 mice. These were placed in C57 blastocysts and backcrossed with wild type C57 mice. The phenotype is Prss22-/- and given its C57 background it is also mMCP7-/-.

1.5.8 RasGRP4 null mouse

A RasGRP4-/- mouse was generated using homologous recombination in embryonic stem cells from C57 mice to disrupt the RasGRP4 gene. These were then placed in C57 blastocysts and backcrossed 5 times with wild type C57 mice.

1.5.9 Caspase 11 deletion in mice with 129Sv background

The mMCP5-/-, mMCP7+/+ and NDST2-/- mice all originate from a 129Sv background, and as such contain a 5 base pair deletion to the Casp4 locus introducing a premature stop codon228. This stop codon attenuates all caspase 11 expression, which I turn could impact on caspase 1 mediated IL-1β activation229.

1.6 Study rationale

The literature demonstrates mast cells contribute towards host defences against pathogens, which can have both beneficial and detrimental effects. A common element of the mast cell response to pathogens following PAMP detection by PRRs is activation and degranulation, releasing a plethora of potent pre-stored mediators, including the mast cell restricted serine proteases into the milieu. It is clear that studies are required to determine which individual components involved in mast cell activation contribute to these different effects.

Previous studies have highlighted the role of a few of these proteases in some infections settings but as yet the roles of mMCP5, mMCP6, mMCP7, NDST2, Prss31, Prss22 and RasGRP4 in S. pneumoniae, P. aeruginosa and IAV infections are unknown.

The experiments in this thesis are designed to determine the role these mast cell proteases, mast cell associated proteases and factors play in the pathogenesis of

S. pneumoniae, P. aeruginosa and IAV.

Chapter 2:

The role of mast cell proteases and associated factors in the pathogenesis of S. pneumoniae in a pneumococcal pneumonia model in mice.

2.1 Abstract

Mast cells are sentinels of the innate immune system, they are enumerated at sites of pathogen colonisation and are known to play a key role in protection against bacterial pathogens. S. pneumoniae, the causative agent of pneumococcal pneumonia is responsible for

25% of global preventable deaths in children under 5 years. The literature shows that mast cells are detrimental during pneumococcus infection in animal models, however the reason for this are unclear.

Using transgenic mice, deficient or competent in a range of mast cell proteases, mast cell related proteases and mast cell associated factors I investigated the role these components play in the pathogenesis of S. pneumoniae in a pneumococcal pneumonia infection model.

I show that the proteases mMCP5, mMCP6 and Prss31 are detrimental during pneumococcal infection, impairing bacterial clearance, and that the tryptases mMCP7 and

Prss22 play no role in the pathogenesis of this infection. I demonstrate the protective role

NDST2 plays in moderating the bacterial induced inflammation and cytokine expression.

Finally, I show that RasGRP4 contributes to the inflammatory response in this infection.

In conclusion, the dichotomy in the role of mast cell proteases, their associated proteases and mast cell related factors is demonstrated in the detrimental effects of some mast cell proteases in pneumococcal infection, I exclude the involvement of some proteases in this infection and have identified individual protease targets that could be targeted for therapeutic intervention. Finally, I have characterised the important roles played by NDST2 and

RasGRP4 in pneumococcal pneumonia.

2.2 Introduction

S. pneumoniae, previously referred to a pneumococcus, is a Gram positive bacteria that is the causative agent for diseases such as pneumococcal sepsis, pneumococcal meningitis, and pneumococcal pneumonia. Epidemiological data shows it to be responsible for 25% of preventable deaths of children under 5 years globally33 with developing nations carrying the greatest disease burden.

The development of capsular polysaccharide (CPS) vaccines has been superseded by

CPS-protein conjugate vaccines (PCV) in developed nations such as here in Australia where dramatic declines in infections rates have been seen. However current vaccine serotypic coverage is limited with current PCVs providing immunity to only the 13 most prevalent pathogenic serotypes. Widespread antibiotic resistance and suggestions the current PCV is skewing S. pneumoniae serotype distribution within vaccinated populations by applying selective pressures and promoting the replacement of PCV serotypes with non PCV covered ones means that S. pneumoniae will remain a major risk to human health in the coming decades. Understanding the pathogenesis of S. pneumoniae is essential to combatting its diseases long term.

Mast cells are synonymous with allergy and much of the research into mast cells has been focused on their deleterious pathological effects as they pertain to allergic disease.

However, mast cells are a key component of the innate immune system and act as sentinel cells being one of the first cells types to come into contact with invading S. pneumoniae.

They are indispensable in host defence against bacterial infections151. Infectious models utilising mast cell deficient mice have demonstrated the important role of mast cells in innate immunity, however to date the role played by mast cell derived proteases, their related proteases and other mast cell factors in S. pneumoniae remains unclear.

In this study I use 7 transgenic mice lines in a murine S. pneumoniae D39 (Spn) respiratory infection model that demonstrates the hallmarks of pneumococcal pneumonia with characteristic neutrophilic inflammatory responses to elucidate the roles of mast cell proteases, their associated proteases and mast cell factors in the pathogenesis of S. pneumoniae infection.

2.3 Methods

2.3.1 Ethics statement

These studies were conducted in accordance with the New South Wales Animal

Research Act 1985, abiding by the animal research regulations (2010) of the said act and the

National Health and Medicine Research Councils Australian code for the care and use of animals for scientific purposes 8th Edition (2013). All protocols used were approved by the

University of Newcastle’s Animal Care and Ethics Committee (permit number 987/0111) under ethics application number A-2012-217.

2.3.2 Streptococcus pneumoniae infection model

Streptococcus pneumoniae D39 serotype 2 (Spn) was plated onto heart infused agar

(ThermoFisher Scientific, Waltham, MA) and incubated over night at 37°C and 5% CO2. The plates were washed and the bacteria resuspended in sterile phosphate buffered saline (PBS).

7-9 week old male mice were anesthetised via an intraperitoneal (i.p) injection of 200mg/kg ketamine and 20mg/kg xylazane then inoculated intratracheally (i.t) with 40µl containing

2x106 Spn (5x107/ml) in PBS. After 8 hours the mice were sacrificed by an i.p. injection of pentobarbital and tissues collected for analysis.

2.3.3 Cellular inflammation

Bronchoalveolar lavage fluid (BALf) was collected from the multilobed (right) lung, processed then analysed. Briefly, the left lung was tied off with cotton thread and 2 x 700µl of PBS was used to wash the right lung via a blunt cannula. The BALf cells were pelleted and resuspended in 750µl of red cell lysis buffer (10mM KHCO3, 150Mm NH4Cl, 0.1Mm EDTA

Na2) on ice for 5 minutes; the supernatant was aspirated and snap frozen in liquid nitrogen for later cytokine analysis. The cells were washed and resuspended in PBS, total cell numbers were calculated using a haemocytometer, then placed in microscope slide containing cassettes and cytocentrifuged at 300g for 5 minutes (ThermoFisher Scientific, Waltham, MA). After drying overnight, the slides were stained with May-Grünwald-Giemsa and observed using a light microscope. Leukocyte enumeration was conducted raised of morphological criteria with a total of 200 cells being counted.

2.3.4 Bacterial recovery

The multi lobed right lung was collected and homogenised in 1 ml of sterile PBS. The lung homogenate and an aliquot of BALf were serially diluted by 1:10, 1:100, 1:1000,

1:10,000 and 1:100,000 in PBS and plated onto heart infused agar (ThermoFisher Scientific,

Waltham, MA) and incubated over night at 37°C and 5% CO2. Colony counts were conducted after 24 hours yielding colony forming units per lung and colony forming units per ml BALf. Both values were combined to give total colony forming units.

2.3.5 Cytokine expression in BALf

Cytokine concentrations in BALf were determined using DuoSet® ELISA kits for

TNFα, CXCL1, CXCL2, IL-1β and IL-6 (R&D Systems, Minneapolis, MN) as per manufactures instructions.

2.3.6 Histopathological scoring

Left lungs were perfused, inflated and fixed in formalin. Fixed tissues were embedded, sectioned (4-6µm) and stained with hematoxylin and eosin. Histopathological scoring was assessed blind in replicates of 3 by myself, using the table below and as previously described230.

Score 1: Airways Inflammation Score /4

0 Lack of inflammatory cells around airways - Absent 1 Some airways have small numbers of cells - Mild 2 Some airways have significant inflammation - Moderate 3 Majority of airways have some inflammation - Marked 4 Majority of airways are significantly inflamed – Severe

Score 2: Vascular Inflammation Score /4

0 Lack of inflammatory cells around vessels - Absent 1 Some vessels have small numbers of cells - Mild 2 Some vessels have significant inflammation - Moderate 3 Majority of vessels have some inflammation - Marked 4 Majority of vessels are significantly inflamed – Severe

Score 3: Parenchymal Inflammation (at 10X magnification) Score /5

0 <1% affected 1 1-9% affected 2 10-29% affected 3 30-49% affected 4 50-69% affected 5 >70% affected

Score 1 + Score 2+ Score 3 = Histopathological score

2.3.7 Statistics

Statistical analyses were performed with GraphPad Prism version 6f. All results were expressed as a mean ± SEM from 6-8 mice, in duplicate. Analysis performed are as indicated in the figures. Values of a p less than 0.05 was deemed a significant difference.

2.4 Model

2.4.1 Characterisation of S. pneumoniae infection

In order to investigate the roles of mast cell proteases, their associated proteases and other mast cell factors in the pathogenesis of S. pneumoniae, a murine model was established in 8 week old male C57 mice and characterised. Intratracheal inoculation with 2x106 S. pneumonia D39 (Spn) while in log growth phase produced a productive infection. Bacterial numbers declined during the first 4 hours following inoculation, and recovered at 8 hours yielding a significant increase in total numbers (Figure 2.4.1.1 A). Infection induced a strong inflammatory response (Figure 2.4.1.1B), comprising of both an increase in BALf macrophage numbers as well as a large increase in the number of neutrophils (Figure 2.4.1.1

C) and neutrophils (Figure 2.4.1.1 D). Peak macrophage response occurred at 8 hours which corresponded with the increase in infection (compared to the 4 hour time point) peak of infection responses observed in BALf.

A B 1×106 2.0×105 *** * 1×105 1.5×105

1×104 1.0×105

1×103 5.0×104 CFU/ml BALf + Lung + BALf CFU/ml

1×102 Total leukocytes/ml BALf 0.0

0 Hrs 4 Hrs 8 Hrs 0 Hrs 4 Hrs 8 Hrs 12 Hrs 24 Hrs 12 Hrs 24 Hrs Time post infection Time post infection

C D

5 5 1×10 ** 1.2×10 ** 5 8×104 1.0×10 8.0×104 6×104 6.0×104 4×104 4.0×104 4 2×10 2.0×104 Neutrophils/ml BALf Neutrophils/ml Macrophages/ml BALfMacrophages/ml 0 0.0

0 Hrs 4 Hrs 8 Hrs 0 Hrs 4 Hrs 8 Hrs 12 Hrs 24 Hrs 12 Hrs 24 Hrs Time post infection Time post infection

Figure 2.4.1.1: Characterisation of Spn infection. S. pneumoniae infection was assessed in mice that received Spn. Bacterial recovery was determined from a homogenate of BALf and lung tissue (A) Airway inflammation was represented by total leukocyte counts in BALf (B) and May-Grünwald-Giemsa staining enabled macrophage and neutrophil differential counts (C and D respectively). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001.

Following the outcome of the characterisation of Spn in C57 mice it was decided to use 8 hours post infection as a suitable endpoint for analysis (Figure 2.4.1.2).

Figure 2.4.1.2: Overview of Spn model. Groups of mice are infected with during hour 0 with 2x106 S. pneumoniae D39 and culled at 8 hours post infection.

2.4.2 S. pneumoniae infection in mMCP6-/- mice

To determine the role of the mast cell specific tryptase mMCP6 in the pathogenesis of

S. pneumoniae, mMCP6-/- mice were infected intratracheally with 2x106 Spn while in log phase. mMCP6-/- mice demonstrated enhanced bacterial clearance (Figure 2.4.2.1) with a typical neutrophilic dominated inflammatory response comparable to C57 infected controls.

A B 1×106 * * 1×107

1×106 1×105

CFU/Lung 1×105 Cfu/ml BALf Cfu/ml

1×104 1×104

C57 Spn C57 Spn mMCP6-/- Spn mMCP6-/- Spn

C * 1×107

1×106

5 Total CFU 1×10

1×104

C57 Spn

mMCP6-/- Spn

Figure 2.4.2.1: Spn infected mMCP6-/- mice show enhanced bacterial clearance. Bacterial returns from homogenised lung and BALf were determined (A&B) and counts combined (C) from 8-hour post Spn infected mice. Data are means ± SEM, differences in bacterial returns were determined by Students t-test, *P<0.05.

Despite the enhance bacterial clearance mMCP6-/- infected mice show typical neutrophilic dominated inflammatory response comparable to C57 infected controls (Figure

2.4.2.2).

Figure 2.4.2.2: Spn infected mMCP6-/- mice show unaltered inflammatory cell infiltration. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B) and neutrophils (C). Data are means ± SEM and counts are by one-way ANOVA with Tukey’s multiple comparisons test *P<0.05, **P<0.01.

Histopathological scoring of lung tissue from Spn infected mice show significant inflammation, however no differences were observed in the magnitude of these responses between the mMCP6-/- infected mice C57 mice (Figure 2.4.2.3 A-E).

Figure 2.4.2.3: Histopathological scoring of Spn infected mMCP6-/- mice show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4- 6µm) lung tissues was stained with hematoxylin and eosin was conducted (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 Spn infected (C), mMCP6-/- uninfected (D), and mMCP6-/- Spn infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001.

Cytokine profiles from BALf demonstrated significant increases in TNFα, CXCL1,

CXCL2, IL-1β and IL-6 following Spn infection however no differences were observed between mMCP6-/- infected mice and their C57 infected controls (Figure 2.4.2.4).

Figure 2.4.2.4: Cytokine profiling of Spn infected mMCP6-/- mice show no differences from infected control C57 mice. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D) and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test.

In summary following S. pneumoniae respiratory infection mMCP6-/- mice show enhanced bacterial clearance with unaltered inflammatory cell recruitment in the BALf, comparable histopathological score’s and cytokine profiles that mirror those of wild type controls. This enhanced bacterial clearance points to a negative role for mMCP6 in this infection.

2.4.3 S. pneumoniae infection in mMCP6-/- mMCP7+/+ mice

C57 mice are deficient in mast cell specific tryptase mMCP7, due to an exon2/intron2 splice site mutation at mMCP7 locus193. C57 mice do express the tryptase mMCP6 which shares considerable sequence homology with mMCP7, therefore to determine the role of mMCP7 alone in the pathogenesis of S. pneumoniae mMCP6-/- mMCP7+/+ mice were infected intratracheally with 2x106 log phase Spn. mMCP6-/- mMCP7+/+ infected mice demonstrated comparable bacterial recovery from the lungs and BALf (Figure 2.4.3.1).

A B

1×107 1×106

1×106

5 1×10 CFU/Lung CFU/ml BALf BALf CFU/ml

1×104 1×105

C57 Spn C57 Spn

mMCP-6 -/- mMCP7+/+ Spn mMCP-6 -/- mMCP7+/+ Spn

1×107

1×106 Total CFU

1×105

C57 Spn

mMCP-6 -/- mMCP7+/+ Spn

Figure 2.4.3.1: Spn infected mMCP6-/- mMCP7+/+ mice show unaltered bacterial clearance. Bacterial returns from homogenised lung and BALf were determined (A&B) and counts combined (C) from 8-hour post Spn infected mice. Data are means ± SEM, differences in bacterial returns were determined by Students t-test.

Together with the unaltered bacterial clearance mMCP6-/- mMCP7+/+ infected mice show a typical neutrophilic dominated inflammatory response comparable to C57 control groups (Figure 2.4.3.2).

A B 1.5×105 **** *** 1×105 8×104 1.0×105 6×104

4×104 5.0×104 2×104 Leukocytes/ml BALf

0.0 BALf Cells/mL of Number 0

C57 Spn C57 Spn C57 Sham C57 Sham

mMCP-6 -/- mMCP7+/+ Spn mMCP-6 -/- mMCP7+/+ Spn mMCP-6 -/- mMCP7+/+ Sham mMCP-6 -/- mMCP7+/+ Sham

C 6×104 *** ****

4×104

2×104

Number of Cells/mL BALf Cells/mL of Number 0

C57 Spn C57 Sham

mMCP-6 -/- mMCP7+/+ Spn mMCP-6 -/- mMCP7+/+ Sham

Figure 2.4.3.2: Spn infected mMCP6-/- mMCP7+/+ mice show unaltered inflammatory cell infiltration. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B) and neutrophils (C). Data are means ± SEM and counts are by one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001 and ****P<0.0001.

Histopathological scoring again demonstrated significant changes in inflammation following Spn challenge, however there was no observable differences in inflammation seen between the infected mMCP6-/- mMCP7+/+ mice versus the C57 infected control (Figure

2.4.3.3).

Figure 2.4.3.3: Histopathological scoring of Spn infected mMCP6-/- mMCP7+/+ mice show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues was stained with hematoxylin and eosin was conducted (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 Spn infected (C) mMCP6-/- mMCP7+/+ uninfected (D) and mMCP6-/- mMCP7+/+ Spn infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiling from BALf demonstrated differences between the significant increases in TNFα, CXCL1, CXCL2, IL-1β and IL-6 (Figure 2.4.3.4) in response to infection with Spn, however the mMCP6-/- mMCP7+/+ mice infected groups show a reduced IL-1β response following Spn infection (Figure 2.4.3.4 D).

A B TNFα CXCL1 800 **** **** 800 **** ****

600 600

400 400 pg/ml BALf pg/ml BALf 200 200

0 0

C57 Spn C57 Spn C57 Sham C57 Sham

mMCP6-/-mMCP7+/+ Spn mMCP6-/-mMCP7+/+ Spn C mMCP6-/-mMCP7+/+Sham D mMCP6-/-mMCP7+/+Sham CXCL2 IL-1β 1500 **** **** 150 **** **** ** 1000 100

500 50 pg/ml BALf pg/ml BALf

0 0

C57 Spn C57 Spn C57 Sham C57 Sham

mMCP6-/-mMCP7+/+ Spn mMCP6-/-mMCP7+/+ Spn E mMCP6-/-mMCP7+/+Sham mMCP6-/-mMCP7+/+Sham IL-6 400 **** ****

300

200 pg/ml BALf 100

C57 Spn C57 Sham

mMCP6-/-mMCP7+/+ Spn mMCP6-/-mMCP7+/+Sham

Figure 2.4.3.4: Cytokine profiling of Spn infected mMCP6-/- mMCP7+/+ mice show impaired IL-1β responses following Spn infection. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, **P<0.01, ***P<0.001, ****P<0.0001.

In summary following S. pneumoniae respiratory infection mMCP6-/- mMCP7+/+ mice show comparable bacterial clearance with unaltered inflammatory cell recruitment in the

BALf, comparable histopathological score’s and cytokine profiles that mirror those of wild type controls with the exception of a reduced IL-1β response. The implications of this reduced IL-1β are unclear, however collectively these data point to mMCP7 playing little to no role in the pathogenesis of this infection.

2.4.4 S. pneumoniae infection in Prss31-/- mice

To determine the role of mast cell restricted g-tryptase, Prss31, in the pathogenesis of

S. pneumoniae Prss31-/- mice were infected intratracheally with 2x106 log phase Spn. Prss31-/- infected mice demonstrated enhanced bacterial recovery from the lungs and BALf of Prss31-/- infected mice (Figure 2.4.4.1).

A B 1×106 * *** 1×107

1×106 1×105 CFU/Lung 1×105 Cells/ml BALf Cells/ml

1×104 1×104

C57 Spn C57 Spn Prss31-/- Spn Prss31-/- Spn

C * 1×107

1×106

5 Total CFU 1×10

1×104

C57 Spn

Prss31-/- Spn

Figure 2.4.4.1: Spn infected Prss31-/- mice show enhanced bacterial clearance. Bacterial returns from homogenised lung and BALf were determined (A&B) and counts combined (C) from 8-hour post Spn infected mice. Data are means ± SEM, differences in bacterial returns were determined by Students t-test, *P<0.05, ***P<0.001.

Despite the enhanced bacterial clearance seen, there were no differences observed in

BALf with the characteristic neutrophilic dominated inflammatory response comparable to

C57 infected controls (Figure 2.4.4.2).

A B

5 4 1.5×10 * * 8×10

6×104 1.0×105

4×104

5.0×104 2×104 Leukocytes /ml BALf Macrophages/ml BALfMacrophages/ml 0.0 0

C57 Spn C57 Spn C57 Sham C57 Sham Prss31-/- Spn Prss31-/- Spn Prss31-/- Sham Prss31-/- Sham

4 8×10 * *

6×104

4×104

2×104 Neutrophils/ml BALf Neutrophils/ml

C57 Spn C57 Sham Prss31-/- Spn Prss31-/- Sham

Figure 2.4.4.2: Spn infected Prss31-/- mice show an unchanged inflammatory response. Airway inflammation was determined by BALf leucocyte counts (A), differential counts of macrophages (B), and neutrophils (C). Data are means ± SEM and counts are by one-way ANOVA with Tukey’s multiple comparisons test *P<0.05.

Histopathological scores concurred with the observations in BALf data that following

Spn challenge there was a significant influx of inflammatory cells into the lungs, with overall levels of inflammation within the two infected groups being comparable.

Figure 2.4.4.3: Histopathological scoring of Spn infected Prss31-/- mice show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues was stained with hematoxylin and eosin was conducted (A). Representative micrographs representing lungs from C57 uninfected (B), C57 Spn infected (C), Prss31-/- uninfected (D), and Prss31-/- Spn infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

BALf cytokine profiles show significant increases in TNFα, CXCL1, CXCL2, IL-1β and IL-6 following infection in both genotypes. There were significant increases in the levels of IL-1β following infection however when compared with the C57 infected control group, these concentrations were significantly suppressed in Prss31-/- infected mice. Conversely the concentration of IL-6 was significantly higher in Prss31-/- infected mice versus their C57 control mice, this elevated IL-6 supports the observation of enhanced bacterial clearance seen in Prss31-/- mice given that IL-6 enhances respiratory burst capacity of neutrophils and their rate of phagocytosis together with promoting superoxide anion generation231 (Figure 2.4.4.4).

A B TNFα CXCL1 1000 *** *** 1000 **** *** 800 800

600 600

400 400 pg/ml BALf pg/ml BALf 200 200

0 0

C57 Spn C57 Spn C57 Sham C57 Sham Prss31-/- Spn Prss31-/- Spn Prss31-/- Sham Prss31-/- Sham

C D IL-1β CXCL2 150 1500 **** *** **** **** ** 100 1000

50 500 pg/ml BALf pg/ml BALf

0 0

C57 Spn C57 Spn C57 Sham C57 Sham Prss31-/- Spn Prss31-/- Spn Prss31-/- Sham Prss31-/- Sham

E IL-6 600 **** **** ****

400

200 pg/ml BALf

C57 Spn C57 Sham Prss31-/- Spn Prss31-/- Sham

Figure 2.4.4.4: Cytokine profiling of Spn infected Prss31-/- mice identified reduced IL-1β and enhanced IL-6 responses following Spn infection. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E) Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, **P<0.01, ***P<0.001, ****P<0.0001.

In summary following S. pneumoniae respiratory infection Prss31-/- mice show enhanced bacterial clearance with unaltered inflammatory cell recruitment in the BALf, comparable histopathological score’s and cytokine profiles that mirror those of wild type controls with the exception of a reduced IL-1β and elevated IL-6 response. The enhanced bacterial clearance yet comparable inflammatory responses in BALf and histopathological scoring suggest more effective bacterial clearance in these mice. Something that has not previously been reported.

100

2.4.5 S. pneumoniae infection in NDST2-/- mice

To determine the role of Glucosaminyl N-deacetylase/N-sulphotransferase-2

(NDST2), an enzyme essential in the N-deacetylation and N-sulphonation of heparin, in the pathogenesis of S. pneumoniae NDST2-/- mice were infected intratracheally with 2x106 log phase Spn. NDST2-/- infected mice show significantly impaired bacterial clearance (Figure

2.4.5.1).

A B * 1×108 1×107 *

1×107 1×106

6 5 1×10 CFU/Lung 1×10 Cells/ml BALF Cells/ml

1×105 1×104

C57 Spn C57 Spn

NDST2-/- Spn NDST2-/- Spn

1×108 **

1×107

6 Total CFU 1×10

1×105

C57 Spn

NDST2-/- Spn

Figure 2.4.5.1: Spn infected NDST2-/- mice show impaired bacterial clearance. Bacterial returns from homogenised lung and BALf were determined (A&B) and counts combined (C) from 8-hour post Spn infected mice. Data are means ± SEM, differences in bacterial returns were determined by Students t-test, *P<0.05, **P<0.01.

101

In addition to the inpaired bacterial clearance NDST2-/- infected mice also had enhanced neutrophilic dominated inflammatory response however lymphocytes also contributed to the enhanced response seen, albeit in small overall numbers. Interestingly macrophages whilst showing increases in total numbers, these increases were not statistically significant (Figure 2.4.5.2).

A B

5 * 4 1.5×10 **** 8×10

**** 6×104 1.0×105

4×104

5.0×104 2×104 Leukocytes/ml BALf Macrophages/ml BALfMacrophages/ml 0.0 0

C57 Spn C57 Spn C57 Sham C57 Sham NDST2-/- Spn NDST2-/- Spn NDST2-/- Sham NDST2-/- Sham

C 1.5×105 **** ****

1.0×105

5.0×104 * Neutrophils/ml BALf Neutrophils/ml

0.0

C57 Spn C57 Sham NDST2-/- Spn NDST2-/- Sham

Figure 2.4.5.2: Spn infected NDST2-/- mice show greater inflammation. Airway inflammation was determined by BALf leucocyte counts (A), and differential counts of macrophages (B), and neutrophils (C). Data are means ± SEM and counts are by one-way ANOVA with Tukey’s multiple comparisons test *P<0.05, ****P<0.0001.

102

Histopathological scores concurred with the observations in BALf data, in that, following challenge there is a significant influx of inflammatory cells to the lungs and that

NDST2-/- infected mice show significantly worse inflammation when compared to their C57 wildtype infected control group (Figure 2.4.5.3).

Figure 2.4.5.3: Histopathological scoring of Spn infected NDST2-/- mice are elevated compared to the scores seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues was stained with hematoxylin and eosin was conducted (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 Spn infected (C), Prss31-/- uninfected (D), and Prss31-/- Spn infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, ****P<0.0001.

103

BALf cytokine profiles show significant deviations between NDST2-/- and C57 Spn infected control groups in CXCL1, CXCL2, IL-1β and IL-6 (Figure 2.4.5.4). One-way

ANOVA analysis shows a significant increase in TNFα responses to infection in C57 infected mice, however no significant difference in TNFα responses in NDST2-/- mice following infection was seen (Figure 2.4.5.3 A). Significant differences in the concentrations of

CXCL2, IL-1β and IL-6 were also observed between the C57 infected control group and the

NDST2-/- infected mice. The NDST2-/- infected mice demonstrated reduced CXCL2 and

IL-1β concentrations whilst IL-6 concentrations were higher. (Figure 2.4.5.4 C,D,E). These altered cytokine profiles concur with the enhanced cellular migration seen together with impaired bacterial killing in the NDST2-/- mice.

104

Figure 2.4.5.4: Cytokine profiling of Spn infected NDST2-/- mice identified altered cytokine profiles with reduced TNFα, CXCL2 and IL-1β, and enhanced IL-6 responses following Spn infection. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, **P<0.01, ****P<0.0001.

105

-/- In summary following S. pneumoniae respiratory infection NDST2 mice show impaired bacterial clearance with elevated inflammatory cell recruitment in the BALf, higher histopathological score’s and cytokine profiles show reduced TNFα, CXCL2, IL-1β and increased IL-6 response. Taken together this data demonstrates the importance of NDST2 in mounting an effective immunological response to this infection.

106

2.4.6 S. pneumoniae infection in mMCP5-/- mice

To determine the role of mast cell restricted chymase mMCP5 in the pathogenesis of

S. pneumoniae mMCP5-/- mice were infected intratracheally with 2x106 log phase Spn. mMCP5-/- mice demonstrated enhanced bacterial clearance (Figure 2.4.6.1).

A B

1×107 1×107 1×106 * 6 1×105 1×10 1×104 1×105 1×103

CFU/Lung 2 1×10 BALf CFU/ml 1×104 1×101 1×100 1×103

Spn Spn C57 Spn C57 Spn

mMCP5-/- mMCP5-/-

1×107 *

1×106

5 Total CFU 1×10

1×104

Spn C57 Spn

mMCP5-/-

Figure 2.4.6.1: Spn infected mMCP5-/- mice show impaired bacterial clearance. Bacterial returns from homogenised lung and BALf were determined (A&B) and counts combined (C) from 8-hour post Spn infected mice. Data are means ± SEM, differences in bacterial returns were determined by Students t-test, *P<0.05.

107

The enhanced bacterial clearance was accompanied with reduced neutrophilic dominated inflammatory response comparable to C57 control groups (Figure 2.4.6.2). The muted inflammatory responses seen in the mMCP5-/- mice were are associated with an apparent lack of macrophage infiltration following the infection as demonstrated by a reduction in macrophages numbers seen in the BALf (Figure 2.4.6.2 B).

A B 1.5×105 *** * 8×104 * * 6×104 1.0×105

4×104

5.0×104 2×104 Leukocytes/ml BALf Macrophages/ml BALfMacrophages/ml 0.0 0

Spn Spn Sham Sham C57 Spn C57 Spn C57 Sham C57 Sham mMCP5-/- mMCP5-/- mMCP5-/- mMCP5-/-

4 4×10 * ***

3×104

2×104

1×104 Neutrophils/ml BALf Neutrophils/ml

Spn Sham C57 Spn C57 Sham mMCP5-/- mMCP5-/-

Figure 2.4.6.2: Spn infected mMCP5-/- mice show reduced inflammatory cell infiltration. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B) and neutrophils (C). Data are means ± SEM and counts are by one-way ANOVA with Tukey’s multiple comparisons test *P<0.05, ***P<0.001.

108

Histopathological scoring demonstrated significantly increase inflammation following

Spn challenge however there was no observable difference in inflammation seen between mMCP5-/- mice and their infected C57 control mice.

Figure 2.4.6.3: Histopathological scoring of Spn infected mMCP5-/- mice show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4- 6µm) lung tissues was stained with hematoxylin and eosin was conducted (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 Spn infected (C) mMCP5-/- uninfected (D) and mMCP5-/- Spn infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiles from BALf demonstrated significant increases in TNFα, CXCL1,

CXCL2, IL-1β and IL-6 in response to infection with Spn (Figure 2.4.6.4). However mMCP5-/- infected mice show reduced concentrations of CXCL2 and IL-1β in the BALf

(Figure 2.4.6.4 C, D) which point to altered migration and cell function in the mMCP5-/- infected mice, consistent with observations.

109

A B TNFα CXCL1 1000 ** *** 1500 *** **** 800 1000 600

400 500 pg/ml BALf pg/ml BALf 200

0 0

C57 Spn C57 Spn C57 Sham C57 Sham

mMCP5-/- Spn mMCP5-/- Spn mMCP5-/- Sham mMCP5-/- Sham

C D CXCL2 IL-1β 2000 **** ** 150 **** **** * ** 1500 100

1000

50 pg/ml BALf pg/ml BALf 500

0 0

C57 Spn C57 Spn C57 Sham C57 Sham mMCP5-/- Spn mMCP5-/- Spn mMCP5-/- Sham mMCP5-/- Sham

E IL-6 800 *** ****

600

400 pg/ml BALf 200

C57 Spn C57 Sham mMCP5-/- Spn mMCP5-/- Sham

Figure 2.4.6.4: Cytokine profiling of Spn infected mMCP5-/- mice show reduced CXCL2 and IL-1β responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E) Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

110

In summary following S. pneumoniae respiratory infection mMCP5-/- mice show enhanced bacterial clearance with reduced inflammatory cell recruitment in the BALf, comparable histopathological score’s and cytokine profiles that mirror those of wild type controls with the exception of a reduced IL-1β and CXCL2 response. Together this data suggests that mMCP5 plays a detrimental role during this infection.

111

2.4.7 S. pneumoniae infection in Prss22-/- mice

To determine the role of Prss22 in the pathogenesis of S. pneumoniae. Prss22-/- mice were infected intratracheally with 2x106 log phase Spn. Prss22-/- mice demonstrated no observable changes in bacterial clearance (Figure 2.4.7.1).

A B BALf CFU Lung CFU 1×107 1×106

1×106 1×105 CFU/ml CFU/ml

1×105 1×104

C57 Spn C57 Spn

Prss22-/- Spn Prss22-/- Spn

C Total CFU 1×107

1×106 CFU/ml

1×105

C57 Spn

Prss22-/- Spn

Figure 2.4.7.1: Spn infected Prss22-/- mice show no changes in bacterial clearance. Bacterial returns from homogenised lung and BALf were determined (A&B) and counts combined (C) from 8-hour post Spn infected mice. Data are means ± SEM, differences in bacterial returns were determined by Students t-test.

112

Together with no changes in bacterial clearances Prss22-/- infected mice also show a typical neutrophilic dominated inflammatory response comparable to C57 infected controls

(Figure 2.4.7.2).

A B BALf Macrophages 4 4 8×10 * * 4×10

6×104 3×104

4×104 2×104

4 BALf Cells/ml 4 Cells/ml BALf Cells/ml 2×10 1×10

0 0

C57 Spn C57 Spn C57 Sham C57 Sham Prss22-/- Spn Prss22-/- Spn Prss22-/- Sham Prss22-/- Sham

C Neutrophils 4×104 ** *

3×104

2×104

Cells/ml BALf Cells/ml 1×104

C57 Spn C57 Sham Prss22-/- Spn Prss22-/- Sham

Figure 2.4.7.2: Spn infected Prss22-/- mice show no changes in inflammatory cell infiltration. Airway inflammation was determined by BALf leucocyte counts (A), and differential counts of macrophages (B), and neutrophils (C). Data are means ± SEM and counts are by one-way ANOVA with Tukey’s multiple comparisons test *P<0.05, **P<0.01.

113

Histopathological scoring demonstrated significant increases in inflammation following Spn challenge however again there was no observable differences between the

Prss22-/- infected group and the C57 infected control group.

Figure 2.4.7.3: Histopathological scoring of Spn infected Prss22-/- mice show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues was stained with hematoxylin and eosin was conducted (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 Spn infected (C), Prss22-/- uninfected (D), and Prss22-/- Spn infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiles from BALf demonstrated significant increases in TNFα, CXCL1,

CXCL2, IL-1β and IL-6 in response to infection with Spn (Figure 2.4.7.4). There were no observable differences seen between Prss22-/- infected mice and their C57 wild time controls.

114

A TNFα B CXCL1 600 800 **** **** **** **** 600 400

400

200 pg/ml BALf pg/ml BALf 200

0 0

C57 Spn C57 Spn C57 Sham C57 Sham Prss22-/- Spn Prss22-/- Spn Prss22-/- Sham Prss22-/- Sham

C D CXCL2 IL-1β 1000 **** **** 150 **** **** 800 100 600

400 50 pg/ml BALf pg/ml BALf 200

0 0

C57 Spn C57 Spn C57 Sham C57 Sham Prss22-/- Spn Prss22-/- Spn Prss22-/- Sham Prss22-/- Sham

E IL-6 250 **** ****

200

150

100 pg/ml BALf 50

C57 Spn C57 Sham Prss22-/- Spn Prss22-/- Sham

Figure 2.4.7.4: Cytokine profiling of Spn infected Prss22-/- mice identified no differences in cytokine responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

115

In summary following S. pneumoniae respiratory infection Prss22-/- mice show comparable bacterial clearance, inflammatory cell recruitment, histopathological score’s and cytokine profiles. These data demonstrate that Prss22 is not involved in the immunological responses to this infection.

116

2.4.8 S. pneumoniae infection in RasGRP4-/- mice

To determine the role of RasGRP4 in the pathogenesis of S. pneumoniae RasGRP4-/- mice were infected intratracheally with 2x106 Log phase Spn. RasGRP4-/- mice demonstrated enhanced bacterial clearance (Figure 2.4.8.1).

A B * 1×106 1×107 *

1×105 1×106

4 5 CFU/Lung 1×10 1×10 Cells/ml BALf Cells/ml

1×103 1×104

C57 Spn C57 Spn

RasGRP4-/- Spn RasGRP4-/- Spn

C 1×107 *

1×106

5 Total CFU 1×10

1×104

C57 Spn

RasGRP4-/- Spn

Figure 2.4.8.1: Spn infected RasGRP4-/- mice show enhanced bacterial clearance. Bacterial returns from homogenised lung and BALf were determined (A&B) and counts combined (C) from 8-hour post Spn infected mice. Data are means ± SEM, differences in bacterial returns were determined by Students t-test, *P<0.05.

117

Together with enhanced bacterial clearance RasGRP4-/- infected mice show altered inflammatory profile with overall leukocyte counts masking stark changes seen in the neutrophil and macrophage responses, with an increased neutrophil response and the absence of a macrophage response (Figure 2.4.8.2).

A B 1×105 * * 8×104 ** **** 8×104 6×104 6×104 4×104 4×104 2×104 2×104 Leukocytes/ml BALf Macrophages/ml BALfMacrophages/ml 0 0

C57 Spn C57 Spn C57 Sham C57 Sham

RasGRP4-/- Spn RasGRP4-/- Spn RasGRP4-/- Sham RasGRP4-/- Sham

C 8×104 * ** 6×104

4×104

2×104 Neutrophils/ml BALf Neutrophils/ml

C57 Spn C57 Sham

RasGRP4-/- Spn RasGRP4-/- Sham

Figure 2.4.8.2: Spn infected RasGRP4-/- mice show no changes in inflammatory cell infiltration. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B) and neutrophils (C). Data are means ± SEM and counts are by one-way ANOVA with Tukey’s multiple comparisons test *P<0.05, **P<0.01, ***P<0.001.

118

Histopathological scoring demonstrated significant changes in inflammation following Spn challenge however despite the differences observed in the inflammatory cell composition in BALf there were no observable histopathological differences between the

RasGRP4-/- infected mice lungs and the C57 infected control groups.

Figure 2.4.8.3: Histopathological scoring of Spn infected RasGRP4-/- mice show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4- 6µm) lung tissues was stained with hematoxylin and eosin was conducted (A). Representative micrographs representing lungs from C57 uninfected (B), C57 Spn infected (C), RasGRP4-/- uninfected (D), and RasGRP4-/- Spn infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

119

Cytokine profiles from BALf demonstrated significant increases in TNFα, CXCL1,

CXCL2, IL-1β and IL-6 in response to infection with S. pneumoniae (Figure 2.4.8.4).

Significant differences in the magnitude of these responses was observed. Infected

RasGRP4-/- mice show significantly impaired production of cytokines with decreases seen in

TNFα, CXCL1, CXCL2, IL-1β and IL-6 respectively when compared to C57 infected controls. The reduction in CXCL1 and CXCL2 would be expected to reduce cell migration and could in part explain the observations of reduced macrophages seen in the BALf of

RasGRP4-/- infected mice. Reductions in TNFα, IL-1β and IL-6 are a little more puzzling given the enhanced bacterial clearance see in RasGRP4-/- infected mice, however these cytokines do work synergistically together and would facilitate the greater bacterial killing seen in the enhance neutrophilic response of the RasGRP4-/- infected mice.

120

A TNFα B CXCL1 600 **** * 800 **** **** ** **** 600 400

400

200 pg/ml BALf pg/ml BALf 200

0 0

C57 Spn C57 Spn C57 Sham C57 Sham

RasGRP4-/- Spn RasGRP4-/- Spn RasGRP4-/- Sham RasGRP4-/- Sham

C D CXCL2 IL-1β 1000 **** **** 150 **** **** *** 800 * 100 600

400 50 pg/ml BALf pg/ml BALf 200

0 0

C57 Spn C57 Spn C57 Sham C57 Sham

RasGRP4-/- Spn RasGRP4-/- Spn RasGRP4-/- Sham RasGRP4-/- Sham

E IL-6 400 **** **** **** 300

200 pg/ml BALf 100

C57 Spn C57 Sham

RasGRP4-/- Spn RasGRP4-/- Sham

Figure 2.4.8.4: Cytokine profiling of Spn infected RasGRP4-/- mice shows a significant impairment in all cytokine responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D) and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

121

In summary following S. pneumoniae respiratory infection RasGRP4-/- mice show

enhanced bacterial clearance with an altered inflammatory cell recruitment profile in the

BALf, comparable histopathological score’s and a cytokine profile demonstrating impaired

production of all cytokines measured. These data indicate that RasGRP4 contributes to the

normal inflammatory response to this infection, and that in its absence an alternative

inflammatory profile is seen. This is a previously unknown role for RasGRP4.

2.4.9 Summary of results

Transgenic Bacterial Airway Histopathology Cytokine Profile Mouse Line Clearance Inflammation Scores Enhanced Comparable to controls Comparable to controls Comparable to controls mMCP6-/-

Comparable to Comparable to controls Comparable to controls Reduced IL-1β mMCP6-/- mMCP7+/+ controls response.

Enhanced Comparable to controls Comparable to controls Reduced IL-1β & Prss31-/- enhanced IL-6 responses Impaired Increased neutrophilic Elevated scores Reduced TNFα, CXCL2 NDST2-/- inflammation & IL-1β and enhanced IL-6 responses Enhanced Macrophage Comparable to controls Reduced CXCL2 & IL-1β inflammatory response responses mMCP5-/- absent

Comparable to Comparable to controls Comparable to controls Comparable to controls Prss22-/- controls Enhanced Overall Leukocyte count Comparable to controls Reduced TNFα, CXCL1, comparable, however CXCL2, IL-1β & IL-6 RasGRP4 -/- elevated neutrophils responses and reduced macrophages.

Table 2.4.9.1: Summary of S. pneumoniae infected mice deficient in mast cell proteases, associated

proteases or mast cell associated factors.

122

2.5 Discussion

This research builds on the observations of Clement et al.232 and van den Boogaard59 showing that mast cells play a role in host defences to pneumococcal pneumonia by examining the role of individual mast cell proteases using transgenic mice models. Using a

S. pneumoniae infection murine model, which elicits immune responses with the hallmarks of human pneumococcal pneumonia infection, I have shown that individual mast cell proteases, their related proteases and associated factors play important roles in driving innate immune responses to pneumococcal infection.

Mast cells are the sentinels of the immune system, strategically located and enriched at the entry points in the body where they are likely to encounter bacteria, including

S. pneumoniae. They play a key role in triggering and sculpting the inflammatory responses and have been shown to influence innate immunity as well as delaying adaptive immunity during infections59. There is growing evidence that mast cell derived proteases play import roles during infections204.

This study shows that mice devoid of the β-tryptase mMCP6 have enhanced bacterial clearance with no discernible differences in the magnitude of inflammatory responses mounted or alterations in cytokine response, suggesting an alternate mechanism to inflammation. Generally following degranulation, mMCP6 is released, however its activity is restricted as it does not readily dissociate from its serglycin proteoglycans, effectively restricting its activities to within the immediate extracellular matrix for extended periods.

Moreover, inhibitors to mMCP6 are abundant in murine serum is abundant with inhibitors to mMCP6195 further limiting its activity. Therefore, it is unlikely mMCP6 is directly involved in bacterial killing, it is more probable it acts on other circulating cells.

Protease activity may cleave exogenous proteins, which can restrict bacterial clearance i.e. through PAR2 activation on the cell surface of neighbouring cells233. As well as cleave other

123

peptides to inactivate immune responses through cleavage of proinflammatory chemokines and cytokines234. Others have shown that mMCP6 can affect T helper cell function, including upregulating Foxp3235, important in regulating host response to infection.

mMCP6 has no effect of the magnitude of the inflammatory responses in this model.

These findings are in apparent contradiction to the previous finding of Thakurdas et al. who showed that mMCP6 was a potent instigator of neutrophil inflammation and was required for the clearance of Klebsiella pneumoniae138; however Thakurdas used an acute intraperitoneal infection model, so direct comparisons to the lung infection model used here are difficult given my model is specific to the lung, which has a considerably larger surface area than that of an intraperitoneal injection site, and where mast cell prevalence is lower than that seen in the skin. The Spn model used in this thesis represents a lower relative dose of bacteria in a given host pathogen interface, thus maybe insufficient to induce the inflammatory responses observed by Thakurdas.

The genetic addition of β-tryptase mMCP7 to mMCP6-/- mice had no effect on bacterial clearance or the magnitude of the inflammatory responses when compared to C57 control mice. However, when considering the observations of mMCP6-/- mMCP7+/+ and mMCP6-/- infected mice together the data infers that the addition of mMCP7 impairs bacterial clearance given that mMCP6-/- infected mice have enhance bacterial clearance whilst mMCP6-/- mMCP7+/+ infected mice do not.

Given their 129Sv background and inherent caspase 11 deficiency, mMCP6-/- mMCP7+/+ infected mice did produce IL-1b following infection albeit at a lower levels than

C57 infected mice. This pleiotropic cytokine has been shown to be essential for facilitating infiltration of leukocytes to inflamed sites and activating them to secrete a cascade of additional proinflammatory cytokines propagating ongoing inflammatory responses. IL-1b-/- mice have previously demonstrated increased susceptibility to S. pneumoniae infection236,

124

thus one would expect mMCP6-/- mMCP7+/+ infected mice to show greater susceptibility to

S. pneumoniae infection at later time points.

Interestingly Prss31-/- mice, like the mMCP6-/- mice, show enhanced bacterial clearance with unaltered inflammatory cell influx. They do however both show impaired

IL-1b responses coupled with an enhanced IL-6 responses. This data is interesting given that mMCP6-/- mMCP7+/+ mice are expected to contain negligible amounts of Prss31 on account of their 129Sv cross background and the close proximity of Prss31 to mMCP7 within the murine genome, just some 2.3kb apart196. Indeed, studies by Wong et al. demonstrated that

Prss31 expression was also strain dependent with expressions levels in 129Sv derived bone marrow derived mast cells (BMDMC) being below the levels of detection.

Additionally, the unaltered inflammatory responses seen in the C57 and Prss31-/- Spn infected mice is somewhat contradictory to the increased IL-6 concentrations observed in the

Spn infected Prss31-/-. Typically elevated IL-6 levels would correspond with increased inflammation237. Given that this model was 8 hours post infection, elevated IL-6 concentrations could result in greater inflammation in Prss31-/- infected mice at later time points.

The cytokine data showing reduced IL-1b responses in Prss31-/- infected mice allow one to postulate that Prss31 may play a modest role in promoting IL-1b responses following

Spn challenge however this role remains unclear. This could together with the inherent caspase 11 deficiency in these mice, help to explain the lower IL-1b response seen in the Spn infection of mMCP6-/- mMCP7+/+ mice. Exactly how Prss31 could promote IL-1b is unclear, however, noting that Prss31 is a membrane bound tryptase this effect would be localised to the mast cell and its immediate proximity.

When taking the mMCP6-/- and Prss31-/- infection data together, my findings that mMCP6 and Prss31 impair bacterial clearance are in direct contradiction with Boogaard et al.

125

who concluded that mast cells unfavourable role in Spn infection is independent of mast cell degranulation59. My data demonstrates that the tryptases mMCP6 and Prss31 contribute towards elevated bacterial loads in the lung. As these tryptases are only released following degranulation, my data demonstrates that degranulation of mast cells does play a role in mediating the unfavourable functions of mast cells observed in Spn infections. The impact the absence of mMCP6 and Prss31 would have in a LD50 infection remains unclear and such experiments are prohibited in Australia.

Any heparin bound tryptase data needs to be considered in the context of the

NDST2-/- infection data, given that NDST2-/- mast cells granules are depleted of heparin proteoglycans that bind and stabilise proteases including the tryptases mMCP6 and mMCP7.

The results show that NDST2-/- mice have impaired bacterial clearance, significantly higher levels of inflammation with altered cytokine profiles following infection, with impairments in

TNFa, CXCL2 and IL-1b responses and a somewhat enhanced IL-6 response compared to

C57 infected controls. This CXCL2 data is somewhat inconsistent with the observation of a significantly enhanced neutrophil dominated inflammatory response in the NDST2-/- infected group. It is clear these mice mount a robust inflammatory response but it is somehow impaired in its capacity to clear bacteria. The precise mechanism at play in this model is unclear and warrants further study, however a word of caution must also be given owing to

NDST2 being required for the production of heparin, a glycosaminoglycan that is known to stabilise many mast cell derived proteases. As a consequence, mast cells derived from

NDST2-/- mice contain granular compartments that have a reduced ability to store a range of different proteases205. Data gathered from experiment’s utilising these mice could must be interpreted with this altered protease composition in mind. Whilst it is clear NDST2 is essential for effective bacterial clearance and moderation of the inflammatory response during disease the precise mechanism of how NDST mediates this is far from clear. The

126

extensive disruption to the composition of mast cell granules together with an inherent caspase 11 deficiency seen in NDST2-/- mice make pinpointing specific functions difficult.

When interpreting these results one must also consider the observations in the context of other factor(s) intricately associated with N-deacylated and N-sulphonated heparin may have played a role. In the absence of heparin mast cell proteases bind to chondroitin sulphate

E the synthesis of which isn’t impacted by the absence of NDST2. Chondroitin sulphate E, like heparin stabilises mast cell proteases and may alter their substrate specificity222. This shift in the ratios of heparin / chondroitin sulphate E protease binding and subsequent substrate specificity changes could also be contributing to my observations. Whilst NDST2-/- are useful in elucidating the roles of mast cells in pneumococcal disease, interpreting the data is difficult given the polyfactorial consequences of this specific knock out.

In the absence of the chymase mMCP5 there was enhanced Spn clearance with reduced inflammation and lower CXCL2 and IL-1b production following infection. During this infection the typical increase in macrophage numbers seen in this model were absent.

This reduction in macrophages numbers corresponds with the reduction seen in CXCL2, a cytokine known to promote macrophage migration. Whilst mMCP5 appears to be required for recruiting macrophages in this model, its absence corresponds with enhanced bacterial killing.

Again, like tryptases, chymase expression is altered in NDST2-/- mice which have very low levels of mMCP5208. Previous studies have suggested even these small amount of mMCP5 in NDST2-/- mice are sufficient to induce additional inflammation208. It is true that the NDST2-/- infected mice appear to have worse inflammation than the mMCP5-/- mice, however these two experiments were run at different times making direct comparisons impractical. C57 mice have normal levels of mMCP5 and during Spn infection were

127

somewhat protected compared to NDST2-/- indicating, that in this model at least, the role of mMCP5 in NDST2-/- mice infected with S. pneumoniae is negligible if anything.

Data from Prss22-/- infected mice shows no differences in bacterial clearance, inflammatory cell recruitment or cytokine responses leading to the conclusion that the epithelial cell restricted epsilon tryptase Prss22 plays no role in the pathogenesis of

S. pneumoniae.

At the start of this study RasGRP4 was believed to be exclusively expressed in mast cells, however more recent work has shown it to be present on splenic CD117+ dendritic cells225. This new finding introduces new questions that the initial study had not planned to address. In my model, RasGRP4-/- infected mice showed enhanced bacterial clearance compared to their infected C57 counterparts, and an equivalent overall inflammatory response that masked an ablated macrophage yet enhanced neutrophilic response. Cytokine profiles demonstrated drastically altered profiles with significant reductions in all cytokines measured. Interestingly, whilst cytokines were reduced, only CXCL1 was absent in the

RasGRP4-/- infected mice. The observation of a reduced CXCL2 and an absent CXCL1 response to Spn on the surface appears to be contradictory given the observation of enhance neutrophilic inflammatory responses seen in RasGRP4-/- infected mice, however when considering the inherent functional redundancy of cytokines, with different molecules having overlapping functions, it is probable that an untested cytokine, could be elevated explaining the increased neutrophilic inflammation.

These results clearly demonstrate that RasGRP4 is detrimental in Spn infection, which is consistent with finding in other studies of this protein in an inflammatory context223.

TNFα recruits inflammatory cells via the up regulation of adhesion molecules238 and mast cell derived TNFα has been shown to enhance neutrophilic inflammation239. Moreover, reports of RasGRP4-/- BMDMCs secreting less TNFα and IL-1b following PMA

128

stimulation223 tallies with the cytokine profiles seen in this study. Furthermore, the observation that Spn infected RasGRP4-/- mice have enhanced bacterial clearance, together with an altered inflammatory cell profile dominated by a neutrophilic inflammatory responses coupled with lower levels of proinflammatory cytokine production permit the intriguing notion of the potential therapeutic benefits of a RasGRP4 inhibitor such as Galectin-3240.

Such an inhibitor could potential alter the kinetics of the infection providing an additional treatment option for this serious life threatening disease.

In conclusion, I have shown that the proteases mMCP5, mMCP6 and Prss31 play a deleterious role in the pathogenesis of S. pneumoniae with all impairing bacterial clearance and I report a novel roles for Prss31 in enhancing IL-1b responses and mMCP5 in recruiting macrophages. I demonstrate that the mast cell related epsilon tryptase Prss22 expressed on epithelial cells together with mast cell tryptase mMCP7 plays no role in the pathogenesis of

S. pneumoniae and show that previously thought to be mast cell restricted RasGRP4 is also detrimental to disease pathogenesis by contributing to impaired bacterial clearance and promoting proinflammatory cytokine production. Finally, I show NDST2 is protective in

S. pneumoniae infection.

129

Chapter 3:

The role of mast cell proteases, their related proteases and mast cell associated factors in the pathogenesis of P. aeruginosa in a pneumonia model in mice.

130

3.1 Abstract

Studies have shown that mast cells protect mice against pseudomonal lung injury via secreted factor(s). Using transgenic mice deficient or competent in a range of mast cell proteases, mast cell related proteases and mast cell associated factors I investigated the role these components play in the pathogenesis of Pseudomonas aeruginosa in a pneumonia infection model.

I show that the protease mMCP6 and mast cell related factor NDST2 play protective roles during P. aeruginosa infections. Whilst demonstrating the deleterious roles played by the proteases mMCP7 and Prss22. I highlight the immunomodulatory functions for mMCP5 and RasGRP4 and propose a novel new role for mMCP5 in the recruitment of macrophages during a P. aeruginosa pneumonia infection.

I demonstrate that the protease Prss31 plays no role in the pathogenesis of this infection whilst identifying Prss22 as a novel therapeutic target that could be inhibited during therapeutic intervention to protect from bacteraemia.

In conclusion, the dichotomy in the role of mast cell proteases, their associated proteases and mast cell related factors is demonstrated in the sometimes polar effects observed during P. aeruginosa infection.

131

3.2 Introduction

Pseudomonas aeruginosa is one of the most common Gram negative bacteria associated with nosocomial respiratory infections77,78. This opportunistic pathogen is a particular risk to those with underlining pathology such as burns, immunodeficiency and chronic pulmonary disorders such as chronic obstructive pulmonary disease (COPD), cystic fibrosis (CF) or non-CF bronchiectasis (nCFB). With its capacity to survive in a range of environmental niches, and its ability to cause disease in a range of patients P. aeruginosa is a major risk to health and a growing burden to health care systems80,85.

Mast cells are synonymous with allergy and much of the research into mast cells has been focused on their deleterious pathological effects. However, mast cells are a key component of the innate immune system and act as sentinel cells being one of the first cells types to come into contact with invading P. aeruginosa151. Studies have shown that mast cells protect mice against pseudomonal lung injury via a unknown secreted factor94. To date the role played by specific mast cell derived proteases, their related proteases and other mast cell factors in P. aeruginosa infection remains unclear despite studies highlighting a potential role.

In this study, I use 7 transgenic mice lines that are deficient or competent in a range of mast cell proteases, related proteases and related factors, in a murine P. aeruginosa PA14 respiratory infection model that demonstrates the hallmarks of pneumonia with characteristic neutrophilic inflammatory responses to elucidate the roles of mast cell proteases, their associated proteases and mast cell factors in the pathogenesis of P. aeruginosa infection.

132

3.3 Methods

3.3.1 Ethics statement

These studies were conducted in accordance with the New South Wales Animal

Research Act 1985, abiding by the animal research regulations (2010) of the said act and the

National Health and Medicine Research Councils Australian code for the care and use of animals for scientific purposes 8th Edition (2013). All protocols used were approved by the

University of Newcastle’s Animal Care and Ethics Committee (permit number 987/0111) under ethics application number A-2012-217.

3.3.2 Clinical score

Animals were monitored throughout the experiment and assigned a clinical score based of the following:

Clinical Clinical signs Score

1 Healthy with no signs of illness

2 Consistently ruffled fur, especially on neck

3 Piloerection, breathing may be deeper and mice less alert

4 Labored breathing. Frequently showing tremors and lethargy

5 Frequently emaciated. May show cyanosis of tail & ears.

6 Death *Animals scoring 4 or 5 are euthanized immediately

3.3.3 Pseudomonas aeruginosa infection model

Pseudomonas aeruginosa PA14 from glycerol stocks was plated onto 2% Lauria-

Bertani (LB) agar (ThermoFisher Scientific, Waltham, MA) overnight and incubated in 37°C and 5% CO2. The next day a single colony was transferred to LB broth and incubated in a

133

shaking incubator overnight. The broth was then diluted in LB and incubated in a shaking incubator until the bacteria were in log phase. The bacteria were pelleted and washed in PBS before being resuspended in PBS at an appropriate concentration. 7-9 weeks old male mice were anesthetised using isoflurane and inoculate intranasally (i.n) with 30µl containing 1x106 cfu PA14. After 12 or 24 hours the mice were sacrificed using pentobarbital overdose and tissues collected for endpoint analysis.

3.3.4 Heat killed bacteria inoculation model

P. aeruginosa PA14 from glycerol stocks was prepared as above, then heat killed by placing the bacteria into 65°C water bath for 2 hours. The culture was stored at -80°C prior to use. Bacterial death was determined by plating out the bacteria to asses any growth.

3.3.5 Cellular inflammation

Bronchoalveolar lavage fluid (BALf) was collected from the multilobed (right) lung, processed then analysed. Briefly, the left lung was tied off with cotton thread and 2 x 700µl of PBS was used to wash the right lung via a blunt cannula. The BALf cells were pelleted and re-suspended in 750µl of red cell lysis buffer (10mM KHCO3, 150Mm NH4Cl, 0.1Mm EDTA

Na2) on ice for 5 minutes; the supernatant was aspirated and snap frozen in liquid nitrogen for later cytokine analysis, The cells were washed and re-suspended in PBS, total cell numbers were calculated using a haemocytometer, then placed in microscope slide containing cassettes and cytocentrifuged at 300g for 5 minutes (ThermoFisher Scientific, Waltham, MA). After drying overnight, the slides were stained with May-Grünwald-Giemsa and observed using a light microscope. Leukocyte enumeration was conducted raised of morphological criteria with a total of 200 cells being counted.

134

3.3.6 Bacterial recovery

The multilobed right lung and the left liver lobe were collected and homogenised in 1 ml of sterile PBS. Serial dilutions of 1:10, 1:100, 1:1000, 1:10,000 and 1:100,000 were prepared and plated onto 2% Lauria-Bertani agar (ThermoFisher Scientific, Waltham, MA) and incubated overnight at 37°C 5%CO2. Colony counts were conducted after 24 hours yielding colony forming units per lung or per left liver lobe. At the same time 20µl of blood from cardiac puncture was plated onto 2% Lauria-Bertani agar and also incubate as above.

These yielded a colony forming unit per ml of blood.

3.3.7 Cytokine expression in BALf

Cytokine concentrations in BALf were determined using DuoSet® ELISA kits for

TNFα, CXCL1, CXCL2, IL-1β and IL-6 (R&D Systems, Minneapolis, MN) as per manufactures instructions.

3.3.8 Statistics

Statistical analysis was performed with GraphPad Prism version 6f. All results were expressed as a mean ± SEM from 5-8 mice. Analysis performed is as indicated in the figures.

Values of a p < 0.05 was deemed a significant difference.

135

3.3.9 Histopathological scoring

Score 1: Airways Inflammation Score /4

Score 2: Vascular Inflammation Score /4

Score 3: Parenchymal Inflammation (at 10X magnification) Score /5

0 <1% affected 1 1-9% affected 2 10-29% affected 3 30-49% affected 4 50-69% affected 5 >70% affected

Score 1 + Score 2+ Score 3 = Histopathological score

136

3.4 Results

3.4.1 Characterisation of P. aeruginosa infection

In order to investigate the roles of mast cell proteases, their associated proteases and other mast cell factors in the pathogenesis of P. aeruginosa, a murine model was established in 8 weeks old male C57 mice and characterised. Intranasal inoculation with 1x106 log phase

P. aeruginosa PA14 induced a strong inflammatory response comprising of both macrophages, lymphocytes and neutrophils with the neutrophilic influx being more profound in scope.

Bacterial numbers declined during the first 24 hours following inoculation rebounding at 48 hours (Figure 3.4.1.1 A) At 12 hours post infection significant inflammation was seen in the BALf, increasing further at 24 hours. (Figure 3.4.1.1 B) Differential analysis revealed that this increase in inflammation is comprised predominantly of macrophages and neutrophils (Figure 3.4.1.1 C,D) lymphocytes also increase significantly following infection

(Figure 3.4.1.1 E). By 48 hours mice had lost 15% of their body mass and the experiments was halted and the mice culled in accordance with our ethics boards rules.

137

A B 1×106 8×106 * 5 1×10 6 **** 6×10 * 1×104 ** ** 4×106 1×103 ** ** CFU/Lung 2×106 1×102 Leukocytes/ml BALf 1×101 0 0 1 4 0 1 4 12 24 48 12 24 Timepoint (h) Timepoint (h)

C D

1.5×106 *** * ** 6 **** 6×10 *

1.0×106 4×106

5.0×105 2×106 Macrophages/ml BALfMacrophages/ml 0.0 BALf Neutrophils/ml 0 1 4 12 24 0 0 1 4 Timepoint (h) 12 24 Timepoint (h)

E 8×104 *

6×104

4×104

2×104 Lymphocytes/ml BALf 0 0 1 4 12 24 Timepoint (h)

Figure 3.4.1.1: Characterisation of PA14 infection. P. aeru gin osa infection was assessed in mice that received PA14. Analysis of bacterial recovery from the multilobed lung homogenate was determined (A) Airway inflammation was represented by total leukocyte counts in BALf (B) and May-Grünwald-Giemsa staining enabled macrophage, neutrophil, and lymphocyte differential counts (C, D & E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ***P<0.0001.

138

Following the characterisation of this infection in C57 mice the time points of 12 and

24 hours were selected as they yielded significant increases in inflammatory cells (Figure

3.4.1.2). In those experiments where heat killed bacteria were used, a 12 hour time point was selected due to adverse events seen in mMCP6-/- mMCP7+/+ and NDST2-/- mice at 24 hours post live infection (Figure 3.4.1.2).

Figure 3.4.1.2: Overview of the PA14 live and heat killed models. Groups of mice are infected at 0 hours with 1x 106 live or heat killed bacteria and culled at 12 & 24 hours in live infections and 12 hours in heat killed ones.

139

3.4.2 P. aeruginosa infection in mMCP6-/- mice

To determine the role of mast cell specific tryptase mMCP6 in the pathogenesis of

P. aeruginosa mMCP6-/- mice were infected intranasally with 1x106 log phase P. aeruginosa

PA14. At 12 hours post infection mMCP6-/- infected mice show no changes in bacterial clearance in the lung or any subsequent differences in the systemic dissemination of the bacteria to the blood or liver when compared to C57 control mice (Figure 3.4.2.1 A-C).

A B C

1×105 150 60

100 40

CFU/Lung 50 20 CFU/ml Blood CFU/ml CFU/Left Liver lobe Liver CFU/Left

1×104 0 0

C57 PA14 C57 PA14 C57 PA14 mMCP6-/- PA14 mMCP6-/- PA14 mMCP6-/- PA14

Figure 3.4.2.1: PA14 infected mMCP6-/- mice show unaltered bacterial clearance and bacteraemia at 12 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A-C). Bacterial returns are by Students t-test with none reaching significance (Figure 3.4.2.1).

140

There was however a considerable reduction in the magnitude of the inflammatory cell influx in the BALf from the mMCP6-/- infected mice. Differential analysis shows this reduction was due to lower neutrophil numbers, an observation that was statistically significant when compared with C57 control mice responses (Figure 3.4.2.2 A-D).

A B 5 6×106 **** **** 2.0×10

1.5×105 6 4×10 **** 1.0×105

2×106 5.0×104 Macrophages/ml BALfMacrophages/ml Leukocytes/ml BALf 0 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14

mMCP6-/- ShammMCP6-/- PA14 mMCP6-/- ShammMCP6-/- PA14

C D 5 6×106 **** **** 1.5×10 **

5 4×106 1.0×10 ****

4 2×106 5.0×10 Neutrophils/ml BALf Neutrophils/ml Lymphocytes/ml BALf 0 0.0

C57 PA14 C57 Sham C57 PA14 C57 Sham

mMCP6-/- PA14 mMCP6-/- ShammMCP6-/- PA14 mMCP6-/- Sham

Figure 3.4.2.2: PA14 infected mMCP6-/- mice show decreased inflammatory cell infiltration due to reduced neutrophils at 12 hours post infection. Airway inflammation was determined by BALf leucocyte counts (A), and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

141

Histopathological scoring of lung tissue at 12 hours shows the characteristic inflammatory responses seen following infection in both the mMCP6-/- infected mice and their C57 infected controls. Analysis shows that the magnitude of this change was comparable between the two infected groups (Figure 3.4.2.3).

Figure 3.4.2.3: Histopathological scoring of PA14 infected mMCP6-/- mice at 12 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C), mMCP6-/- uninfected (D), and mMCP6-/- PA14 infected (E) mice Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001 and ****P<0.0001.

Cytokine profiles from the BALf from mMCP6-/- PA14 infected mice demonstrated drastically altered production of TNFα and IL-1β with both cytokines being significantly reduced compared to C57 infected control groups. Indeed, the TNFα response was completely absent in this experiment (Figure 3.4.2.4 A-E).

142

A TNFα B CXCL1 2500 **** **** 1500 2000 **** **** 1000 1500

1000 500 pg/ml BALf pg/ml BALf 500

0 0

C57 Sham C57 PA14 C57 Sham C57 PA14

mMCP6-/- ShammMCP6-/- PA14 mMCP6-/- ShammMCP6-/- PA14

C D IL-1β CXCL2 300 **** * 1500 * **** **** 200 1000

100 500 pg/ml BALf pg/ml BALf

0 0

C57 Sham C57 PA14 C57 Sham C57 PA14

mMCP6-/- ShammMCP6-/- PA14 mMCP6-/- ShammMCP6-/- PA14

E IL-6 2000 **** **** 1500

1000 pg/ml Balf 500

C57 Sham C57 PA14

mMCP6-/- ShammMCP6-/- PA14

Figure 3.4.2.4: Cytokine profiling of PA14 infected mMCP6-/- mice 12 hours post infection show impaired TNFα and IL-1β responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

143

With initial results at 12 hours showing reduced inflammation in BALf together with reduced TNFα and IL-1β production the experiment was repeated at the 24-hour time point.

Just like 12 hours, the mMCP6-/- PA14 infected mice show no desirable differences in bacterial recovery from the lungs (Figure 3.4.2.5). There were however, distinct differences in the prevalence of bacteraemia observed between the mMCP6-/- infected group and their

C57 infected controls, with all the mMCP6-/- infected mice testing positive to bacteraemia by blood culturing (n=5) compared to 1 (n=7) from the C57 control groups doing the same

(Figure 3.4.2.6). This data indicates that mMCP6 has a protective roel against bacteraemia in this model.

Figure 3.4.2.5: PA14 infected mMCP6-/- mice show unaltered bacterial clearance at 24 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A- C). Bacterial returns were analysed by Students t-test.

*** 100

% Group with Bacteremia Bacteremia with Group % 0

C57 PA14 mMCP6-/- PA14

Figure 3.4.2.6: mMCP6-/- infected mice show an increased incidence of bacteraemia as detected by blood cultures at 24 hours post infection. The presence of PA14 in blood was determined. Differences in bacterial numbers were analysed by Students t-test ***P<0.001.

144

At 12 hours post PA14 infection the leukocyte counts from BALf of mMCP6-/- infected mice demonstrated a reduction in inflammation, attributed to a reduction in neutrophils. These changes were no longer observable in mMCP6-/- infected mice at 24 hours, indeed the overall leukocyte counts and differential counts are comparable to their C57 control group (Figure 3.4.2.7 A-D).

Figure 3.4.2.7: PA14 infected mMCP6-/- mice show comparable cell infiltration at 24 hours post infection. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

145

Histopathological scoring of lung tissue at 24 hours again show the characteristic inflammation following infection in both mMCP6-/- infected mice and C57 infected controls.

Analysis identified that the difference was similar to data at 12 hours, and comparable between the two infected groups (Figure 3.4.2.8).

Figure 3.4.2.8: Histopathological scoring of PA14 infected mMCP6-/- mice at 24 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C), mMCP6-/- uninfected (D), and mMCP6-/- PA14 infected (E) mice Data are means ± SEM, One-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

At 12 hours post infection mMCP6-/- infected mice demonstrated drastically altered production of TNFα and IL-1β, this observation was partly replicated at the 24-hour time point with the continued absence of a TNFα response and a somewhat ablated IL-1β response. Indeed, half of the mMCP6-/- infected mice show no IL-1β response at all. All other cytokine responses are comparable to C57 infected control mice (Figure 3.4.2.9).

146

A TNFα B CXCL1 400 300 **** **** **** **** 300 200

200

100 pg/ml BALf pg/ml BALf 100

0 0

C57 Sham C57 PA14 C57 Sham C57 PA14

mMCP6-/- ShammMCP6-/- PA14 mMCP6-/- ShammMCP6-/- PA14

C D CXCL2 IL-1β 50000 50 **** **** * 40000 40

30000 30

20000 20 pg/ml BALf pg/ml BALf 10000 10

0 0

C57 Sham C57 PA14 C57 Sham C57 PA14

mMCP6-/- ShammMCP6-/- PA14 mMCP6-/- ShammMCP6-/- PA14

E IL-6 50 *** ** 40

20 pg/ml BALf 10

C57 Sham C57 PA14

mMCP6-/- ShammMCP6-/- PA14

Figure 3.4.2.9: Cytokine profiling of PA14 infected mMCP6-/- mice 24 hours post infection show impaired TNFα responses. Common Inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

147

This data, taken together with the 12-hour data demonstrates that, in this model, mMCP6 is required to mount a robust TNFα and IL-1β proinflammatory cytokine response to

P. aeruginosa. The absence of this cytokine response contributes towards the systemic dissemination of the bacteria to the blood and subsequent bacteraemia. Overall the data demonstrates that mMCP6 plays a protective role in the pathogenesis of P. aeruginosa in this model.

3.4.3 P. aeruginosa infection in mMCP6-/- mMCP7+/+ mice

C57 mice are deficient in mast cell specific tryptase mMCP7, due to an exon2/intron2 splice site mutation at mMCP7 locus193. C57 mice do express the tryptase mMCP6 which shares considerable sequence homology with mMCP7, therefore to determine the role of mMCP7 alone in the pathogenesis of P. aeruginosa mMCP6-/- mMCP7+/+ mice were infected intranasally with 1x106 log phase P. aeruginosa PA14.

Clinical scores from mice at 12 hours clearly demonstrated that the mMCP6-/- mMCP7+/+ infected mice fared worse during this infection with their clinical scores being around 50% greater than C57 infected controls (Figure 3.4.3.1).

4 ***

1 Clinical score (/6) score Clinical

C57 sham C57 PA14

mMCP6-/- 7+/+ mMCP6-/-sham 7+/+ PA14

Figure 3.4.3.1: PA14 infected mMCP6-/- mMCP7+/+ mice show a worse overall clinical score at 12 hours post infection. Data are means ± SEM, Clinical score figures are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001.

148

At 12 hours post infection mMCP6-/- mMCP7+/+ infected mice demonstrated a statistically significant impairment in bacterial clearance coupled with a more extensive systemic dissemination of the bacteria to the blood or liver when compared to C57 infected control mice (Figure 3.4.3.2 A-C).

A B C ** 1×106 1000 1×106 * ** 800 1×105 1×105 600 1×104

400 1×103 4 CFU/Lung 1×10 cfu/ml (blood) cfu/ml 200 1×102 CFU/Left Liver lobe Liver CFU/Left

1×103 0 1×101

C57 PA14 C57 PA14 C57 PA14

mMCP6-/- 7+/+ PA14 mMCP6-/- 7+/+ PA14 mMCP6-/- 7+/+ PA14

Figure 3.4.3.2: PA14 infected mMCP6-/- mMCP7+/+ mice show enhanced bacterial clearance in the lungs but show a greater incidence of bacteraemia at 12 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A-C). Bacterial returns are by Students t-test *P<0.05, **P<0.01, ***P<0.001.

This reduction in bacterial clearance and subsequent systemic bacterial dissemination corresponded with observations from BALf that shows reductions in total leukocyte counts

(Figure 3.4.3.3 A), primarily attributed to a reduction in the magnitude of neutrophilic inflammation (Figure 3.4.3.3 C) seen in mMCP6-/- mMCP7+/+ infected mice. Further reductions were observed in lymphocyte numbers albeit from a somewhat lower baseline. All the reductions observed were of statistical significance when compared to C57 infected control mice.

149

A B

1.5×106 *** ** 4×104

3×104 1.0×106

2×104

5.0×105 1×104 Leukocytes/ml BALf Macrophages/ml BALfMacrophages/ml 0.0 0

C57 sham C57 PA14 C57 sham C57 PA14

mMCP6-/- 7+/+ mMCP6-/-sham 7+/+ PA14 mMCP6-/- 7+/+ mMCP6-/-sham 7+/+ PA14

C D 2.0×106 *** ** 4×104 *** ** 1.5×106 3×104

1.0×106 2×104

5.0×105 1×104 Neutrophils/ml BALf Neutrophils/ml Lymphocytes/ml BALf 0.0 0

C57 sham C57 PA14 C57 sham C57 PA14

mMCP6-/- 7+/+ mMCP6-/-sham 7+/+ PA14 mMCP6-/- 7+/+ mMCP6-/-sham 7+/+ PA14

Figure 3.4.3.3: PA14 infected mMCP6-/- mMCP7+/+ mice show decreased inflammatory cell infiltration attributed to reduced neutrophilic infiltration at 12 hours post infection. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both **P<0.01, ***P<0.001.

150

Histopathological scoring of lung tissue at 12 hours shows the characteristic inflammation typical following this infection in both the mMCP6-/- mMCP7+/+ infected mice and their C57 infected controls. Analysis shows that this change was significantly different between the two infected groups, with the mMCP6-/- mMCP7+/+ mice showing elevated histopathological scores (Figure 3.4.3.4).

Figure 3.4.3.4: Histopathological scoring of PA14 infected mMCP6-/- mMCP7+/+ mice at 12 hours show more inflammation in lungs than infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C), mMCP6-/- mMCP7+/+ uninfected (D), and mMCP6-/- mMCP7+/+ PA14 infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001 and ****P<0.0001.

Cytokine profiling demonstrated an almost total absence of TNFα and IL-6 production following infection in mMCP6-/- mMCP7+/+ mice. Other measured cytokines are comparable to C57 control mice (Figure 3.4.3.5).

151

A TNFα B CXCL1 2500 **** **** 1000 **** **** 2000 800

1500 600

1000 400 pg/ml BALf pg/ml BALf 500 200

0 0

C57 Sham C57 PA14 C57 Sham C57 PA14

mMCP6-/-7+/+ ShammMCP6-/-7+/+ PA14 mMCP6-/-7+/+ ShammMCP6-/-7+/+ PA14

C D CXCL2 IL-1β 800 **** **** 400 *** ** 600 300

400 200 pg/ml BALf pg/ml BALf 200 100

0 0

C57 Sham C57 PA14 C57 Sham C57 PA14

mMCP6-/-7+/+ ShammMCP6-/-7+/+ PA14 mMCP6-/-7+/+ ShammMCP6-/-7+/+ PA14

E IL-6 2500 **** **** 2000

1500

1000 pg/ml BALf 500

C57 Sham C57 PA14

mMCP6-/-7+/+ ShammMCP6-/-7+/+ PA14

Figure 3.4.3.5: Cytokine profiling of PA14 infected mMCP6-/- mMCP7+/+ mice 12 hours post infection show impaired TNFα and IL-6 responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D) and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

152

This observations of impaired bacterial clearance, tightened systemic dissemination of the bacteria, impaired inflammatory responses, worse clinical scoring and altered cytokine profiles demonstrate the detrimental role mMCP7+/+ contributes in this model.

Given the observations seen at 12 hours post infection, ethically I would not have investigated this infection in mMCP6-/- mMCP7+/+ mice at 24 hours. However, I conducted the 24-hour time point first, I did not anticipate the severity of this infection in this transgenic mouse. Subsequently the outcome of the experiment resulted in the submission of an adverse event report to the University of Newcastle’s Animal Care and Ethics Committee (permit number 987/0111) under ethics number A-2012-217.

mMCP6-/- mMCP7+/+ infected mice at 12 hours demonstrated a plethora of negative indicators for mMCP7s role in PA14 infection. Taking this model out to 24 hours these mice show a marked increase in mortality rates, indeed more that 50% of the mice succumbed to the infection, a figure that is alone significant.

Figure 3.4.3.6: mMCP6-/- mMCP7+/+ mice show greater susceptibility to PA14 infection. Survival curve comparison between PA14 infected mMCP6-/- mMCP7+/+ mice and C57 controls. Statistical analysis is via Gehan-Breslow-Wilcoxon test. *P<0.05.

Mindful of the severity of the PA14 infection in mMCP6-/- mMCP7+/+ mice, I sought to investigate if the observations I had previously seen were as a result of an aberrant immune

153

response to components of the P. aeruginosa bacterium or if the response was due to an invasive infection.

12 hours following inoculation with heat killed PA14 mMCP6-/- mMCP7+/+ mice like their C57 control counterparts mount a robust inflammatory response comprised of neutrophils (Figure 3.4.3.7 C) and to a lesser degree, lymphocytes (Figure 3.4.3.7 D).

Interestingly there are significant reductions in the number of macrophages in mMCP6-/- mMCP7+/+ infected mice, something that wasn’t replicated in their C57 controls (Figure

3.4.3.7 B). There were also no differences in clinical scores observed between the groups

(data not shown).

Figure 3.4.3.7: Heat killed PA14 inoculated mMCP6-/- mMCP7+/+ mice show inflammatory cell responses comparable to their C57 controls at 12 hours. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B) neutrophils (C) and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01. 154

Histopathological scoring of lung tissue at 12 hours post inoculation with heat killed

PA14 shows the characteristic, albeit muted, inflammatory response following exposure in both the mMCP6-/- mMCP7+/+ inoculated mice and their C57 inoculated controls. Analysis shows that the magnitude of these changes, in mMCP6-/- mMCP7+/+ inoculated mice, were significantly different (Figure 3.4.3.8).

Figure 3.4.3.8: Histopathological scoring of heat killed PA14 inoculated mMCP6-/- mMCP7+/+ mice at 12 hours shows no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C), mMCP6-/- mMCP7+/+ uninfected (D), and mMCP6-/- mMCP7+/+ PA14 infected (E) mice. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001 and ****P<0.0001.

155

Together with the elevated inflammatory response in the lung, the clinical scores from mMCP6-/- mMCP7+/+ infected mice at 12 hours clearly demonstrated that these mice fared worse during this infection in relation to their clinical scores with them being around 50% greater than C57 infected controls (Figure 3.4.3.9).

Figure 3.4.3.9: mMCP6-/- mMCP7+/+ mice inoculated with heat killed PA14 shows worse clinical scores that their C57 controls at 12 hours. Clinical scores at 12 hours post inoculation. Data are means ± SEM, one- way ANOVA with Tukey’s multiple comparisons test, *P<0.05.

Cytokine profiling of BALf from the 12 hour heat killed experiment replicated observations seen in the 12-hour live infection time point in that mMCP6-/- mMCP7+/+ mice failed to mount any TNFα or IL-6 responses to the bacteria. Other cytokine responses are comparable to C57 infected controls.

156

Figure 3.4.3.10: Cytokine profiling of heat killed PA14 inoculated mMCP6-/- mMCP7+/+ mice at 12 hours demonstrate impaired TNFα and IL-6 responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

157

This data concurs with the 12-hour live infection data in that mMCP6-/- mMCP7+/+ mice appear to have defects in TNFα and IL-6 responses following inoculation and present with higher clinical scores. The data supports the notion that mMCP7 is detrimental in both our live and heat killed PA14 models.

This data taken together demonstrates that mMCP7 plays a detrimental role in the pathogenesis of P. aergonisia in this model.

3.4.4 P. aeruginosa infection in Prss31-/- mice

To determine the role of mast cell restricted g-tryptase Prss31 in the pathogenesis of

P. aeruginosa Prss31-/- mice (10-13 weeks) were infected intranasally with 1x106 log phase

P. aeruginosa PA14. At 12 hours post infection Prss31-/- infected mice show no changes in bacterial clearance in the lung when compared to C57 control mice (Figure 3.4.4.1).

Figure 3.4.4.1: PA14 infected Prss31-/- mice show comparable bacterial clearance to their C57 controls in the lungs. Bacterial returns from homogenised lung were determined. Bacterial returns are by Students t-test.

158

Analysis of BALf also shows no differences in the magnitude of inflammatory responses seen following infection in total leukocytes or differential counts between samples from Prss31-/- infected mice and their C57 controls (Figure 3.4.4.2). Additionally, the clinical scores shows no discernible differences between the infected groups.

A B

1.5×106 ** * 1.5×105

1.0×106 1.0×105

5.0×105 5.0×104 Leukocytes/ml BALf Macrophages/ml BALfMacrophages/ml 0.0 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14 Prss31 Sham Prss31 PA14 Prss31 Sham Prss31 PA14

C D

6 6×104 2.5×10 * * * ** 2.0×106 4×104 1.5×106

1.0×106 2×104 5.0×105 Leukocytes/ml BALf Neutrophils/ml BALf Neutrophils/ml

0.0 0

C57 Sham C57 PA14 C57 Sham C57 PA14 Prss31 Sham Prss31 PA14 Prss31 Sham Prss31 PA14

Figure 3.4.4.2: PA14 infected Prss31-/- mice show comparable inflammatory cell infiltration attributed to neutrophilic infiltration at 12 hours post infection when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

159

Histopathological scoring of lung tissue at 12 hours post infection with PA14 shows the characteristic inflammatory response following exposure in both the Prss31-/- infected mice and their C57 infected controls. Analysis shows that the magnitude of these changes are comparable compared two infected groups (Figure 3.4.4.3).

Figure 3.4.4.3: Histopathological scoring of PA14 infected Prss31-/- mice at 12 hours shows no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C) Prss31-/- uninfected (D), and Prss31-/- PA14 infected mice (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiling at 12 hours shows a reduction in CXCL2 responses in the Prss31-/- infected mice compared to C57 mice, an observation in keeping with the observation of reduced neutrophils in BALf. All other measured cytokines demonstrated a similar response following infection.

160

Figure 3.4.4.4: Cytokine profiling of PA14 infected Prss31-/- mice at 12 hours demonstrate an impaired CXCL2 response when compared to C57 controls. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

161

With initial results at 12 hours showing unchanged bacterial clearance and inflammation in BALf, comparable clinical scores and just a moderate change in cytokine production, namely a reduction in CXCL2 production the experiment was repeated at the 24- hour time point.

Like the 12-hour time point, the 24-hour data shows no differences in bacteria clearance from the lung, also there were no differences in the systemic dissemination of

PA14 in Prss31-/- mice compared to the C57 infected controls.

Figure 3.4.4.5: PA14 infected Prss31-/- mice show comparable bacterial clearance to their C57 controls at 24 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A-C). Bacterial returns are by Students t-test.

162

Analysis of BALf also demonstrated no significant differences between the infected groups in the inflammatory response to PA14 (Figure 3.4.4.6). Clinical scoring also demonstrated no differences between infected groups.

A B 5 6×106 **** **** 2.0×10

1.5×105 4×106 1.0×105

2×106 5.0×104 Leukocytes/ml BALf Macrophages/ml BALfMacrophages/ml 0 0.0

C57 PA14 C57 Sham C57 Pa14 C57 Sham Prss31-/- PA14 Prss31-/- ShamPrss31-/- PA14 Prss31-/- Sham

C D

6 1.5×105 6×10 **** **** ** ***

5 4×106 1.0×10

4 2×106 5.0×10 Neutrophils/ml BALf Neutrophils/ml Lymphocytes/ml BALf 0 0.0

C57 PA14 C57 Sham C57 PA14 C57 Sham Prss31-/- PA14 Prss31-/- ShamPrss31-/- PA14 Prss31-/- Sham

Figure 3.4.4.6: PA14 infected Prss31-/- mice show comparable inflammatory cell infiltration primarily attributed to neutrophilic infiltration at 24 hours post infection when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B) neutrophils (C) and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

163

Histopathological scoring of lung tissue at 24 hours again shows the characteristic inflammation following infection in both the Prss31-/- infected mice and their C57 infected controls. Analysis shows that the magnitude of this change was again similar compared to the

12 hours, being comparable between the two infected groups (Figure 3.4.4.7).

Figure 3.4.4.7: Histopathological scoring of PA14 infected Prss31-/- mice at 24 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C) Prss31-/- uninfected (D) and Prss31-/- PA14 infected mice (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Looking at the cytokine profiles of mice from 24 hours the impaired CXCL2 responses observed at 12 hours are no longer present in Prss31-/- infected mice, indeed all cytokine responses from the infected Prss31-/- mice are comparable to their C57 controls

(Figure 3.4.4.8).

164

A TNFα B CXCL1 200 **** 300 **** **** **** 150 200

100

100 pg/ml BALf pg/ml BALf 50

0 0

C57 PA14 C57 PA14 C57 Sham C57 Sham

Prss31-/- ShamPrss31-/- PA14 Prss31-/- ShamPrss31-/- PA14

C D CXCL2 IL-1β 50000 30 **** **** *** ** 40000 20 30000

20000 10 pg/ml BALf pg/ml BALf 10000

0 0

C57 PA14 C57 PA14 C57 Sham C57 Sham

Prss31-/- ShamPrss31-/- PA14 Prss31-/- ShamPrss31-/- PA14

E IL-6 100 ****

80 ****

40 pg/ml BALf 20

C57 PA14 C57 Sham

Prss31-/- ShamPrss31-/- PA14

Figure 3.4.4.8: Cytokine profiling of PA14 infected Prss31-/- mice at 24 hours are comparable to their C57 controls. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), and IL-6 (D). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, **P<0.01, ***P<0.001, ****P<0.0001.

This data taken together demonstrates that Prss31 plays little to no role in the pathogenesis of P. aergonisa in this model.

165

3.4.5 P. aeruginosa infection in NDST2-/- mice

To determine the role of Glucosaminyl N-deacetylase/N-sulphotransferase-2

(NDST2), an enzyme essential in the N-deacetylation and N-sulphonation of heparin, in the pathogenesis of P. aeruginosa. NDST2-/- mice were infected intranasally with 1x106 log phase P. aeruginosa PA14. At 12 hours post infection NDST2-/- infected mice show no changes in bacterial clearance in the lung but demonstrated reduced incidences of bacteraemia as detected in the liver when compared to C57 control mice (Figure 3.4.5.1A-C).

Figure 3.4.5.1: PA14 infected NDST2-/- mice show comparable bacterial clearance to their C57 controls in the lungs with a reduced incidence of bacteraemia as measured from the liver at 12 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A-C). Bacterial returns are by Students t-test *P<0.05, **P<0.01, ***P<0.001.

There was however a considerable increase in leukocyte counts from BALf from

NDST2-/- infected mice, and differential analysis shows this increase was attributed to neutrophilic inflammation. Unusually lymphocytes were not observed in this model and there was a reduction in macrophages seen in NDST2-/- infected mice when compared with C57 controls (Figure 3.4.5.2 A-C).

166

A B 1.5×106 * ** 1.5×106 *

1.0×106 1.0×106 ** **

5.0×105 5.0×105 Neutrophils/ml BALf Neutrophils/ml Leukocytes/ml BALf 0.0 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14

NDST2-/- ShamNDST2-/- PA14 NDST2-/- ShamNDST2-/- PA14

C 3×104 ***

2×104

1×104 Macrophages/ml BALfMacrophages/ml 0

C57 Sham C57 PA14

NDST2-/- ShamNDST2-/- PA14

Figure 3.4.5.2: PA14 infected NDST2-/- mice show enhanced inflammatory cell infiltration attributed to neutrophilic infiltration at 12 hours post infection. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B) neutrophils (C) and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

167

Histopathological scoring of lung tissue at 12 hours post infection with PA14 shows the characteristic inflammatory response following exposure in both the NDST2-/- infected mice and their C57 infected controls. Analysis shows that the magnitude of these changes are comparable between the two infected groups (Figure 3.4.5.3).

Figure 3.4.5.3: Histopathological scoring of PA14 infected NDST2-/- mice at 12 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C), NDST2-/- uninfected (D), and NDST2-/- PA14 infected mice (E) Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

168

Together with the enhanced inflammatory responses seen in BALf, the clinical scores from NDST2-/- infected mice at 12 hours clearly demonstrated that the NDST2-/- infected mice fared worse during this infection with their clinical scores being again around 50% greater than C57 infected controls (Figure 3.4.5.4).

6 *

2 Clinical score (/6) score Clinical

C57 Sham C57 PA14

NDST2-/- PA14 NDST2-/- Sham

Figure 3.4.5.4: NDST2-/- infected mice show worse clinical scores that their C57 controls at 12 hours. Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05.

At 12 hours post infection BALF from NDST2-/- infected mice demonstrated altered cytokine production with an increase in CXCL1 and CXCL2 respectively, and decreases in

TNFα and IL-6 when compared to C57 controls (Figure 3.4.5.5).

169

A B CXCL1 TNFα ** 2500 **** **** 1500 **** **** 2000 1000 1500

1000 500 pg/ml BALf pg/ml BALf 500

0 0

C57 Sham C57 PA14 C57 Sham C57 PA14

NDST2-/- ShamNDST2-/- PA14 NDST2-/- ShamNDST2-/- PA14

C D CXCL2 IL-1β 1200 ** 50 **** **** 1000 40 800 30 600 20 400 pg/ml BALf pg/ml BALf

200 10

0 0

C57 Sham C57 PA14 C57 Sham C57 PA14

NDST2-/- ShamNDST2-/- PA14 NDST2-/- PA14NDST2-/- Sham

E IL-6 1500 **** **** **** 1000

500 pg/ml BALf

C57 Sham C57 PA14

NDST2-/- ShamNDST2-/- PA14

Figure 3.4.5.5: Cytokine profiling of PA14 infected NDST2-/- mice at 12 hours demonstrates impaired TNFα and IL-6 responses together with enhanced CXCL1 and CXCL2 responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

170

Again, given the observations seen at 12-hour post infection, ethically I would not have investigated this infection in NDST2-/- mice at 24-hours. However, I conducted the 24- hour time point first (and in conjunction with the mMCP6-/- mMCP7+/+ PA14 24-hour experiment), I did not anticipate the severity of this infection in this transgenic mouse.

Subsequently the outcome of the experiment resulted in the submission of an adverse event report to the University of Newcastle’s Animal Care and Ethics Committee (permit number

987/0111) under ethics number A-2012-217.

The NDST2-/- infected mice at 12 hours demonstrated a protective role for NDST2 in our PA14 infection. Taking this model out to 24 hours these mice show a marked increase in mortality rates, indeed more that 80% (6/7) of the mice succumbed to the infection, a figure that is alone significant.

Figure 3.4.5.6: NDST2-/- mice show greater susceptibility to PA14 infection. Survival curve comparison between PA14 infected NDST2-/- mice and their C57 controls. Statistical analysis is via Gehan-Breslow- Wilcoxon test. *P<0.05, **P<0.005.

Mindful of the severity of the live PA14 infection in NDST2-/- mice, I sought to investigate if the observations I had previously seen were as a result of an aberrant immune response to components of the P. aeruginosa bacterium or if the response was due to an invasive live infection.

171

Following inoculation with heat killed PA14, NDST2-/- mice like their C57 control counterparts mounted a robust inflammatory response comprised of neutrophils (Figure

3.4.5.7 C) and to a lesser degree, lymphocytes (Figure 3.4.5.7 D); the two mouse genotypes are comparable. Additionally, no differences in clinical scores were observed between the groups (data not shown).

A B 2.0×105 ** 8×104

1.5×105 * 6×104

1.0×105 4×104

5.0×104 2×104 Leukocytes/ml BALf Macrophages/ml BALF Macrophages/ml 0.0 0

C57 sham C57 sham

C57 heat killedNDST2-/- sham C57 heat killedNDST2-/- sham

NDST2-/- heat killed NDST2-/- heat killed

C D

1.5×105 *** 3000

1.0×105 2000 *

5.0×104 1000 Neutrophils/ml BALF Neutrophils/ml Lymphocytes/ml BALF 0.0 0

C57 sham C57 sham

C57 heat killedNDST2-/- sham C57 heat killedNDST2-/- sham

NDST2-/- heat killed NDST2-/- heat killed

Figure 3.4.5.7: Heat killed PA14 inoculated NDST2-/- mice show comparable inflammatory cell infiltration when compared to their C57 controls at 12 hours post inoculation. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

172

Histopathological scoring of lung tissue at 12 hours post inoculation with heat killed

PA14 shows a decreased, albeit muted, inflammatory response following exposure in both the

NDST2-/- inoculated mice and their C57 inoculated controls. Analysis shows that the magnitude of these changes are comparable between the two inoculated groups (Figure

3.4.5.8).

Figure 3.4.5.8: Histopathological scoring of heat killed PA14 inoculated NDST2-/- mice at 12 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C), NDST2-/- uninfected (D), and NDST2-/- PA14 infected (E) Data are means ± SEM, One-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiling of BALf shows an increase in the CXCL1 response to the heat killed bacteria in NDST2-/-, an observation also seen in the live infection at 12 hours. All other cytokines responses are comparable to those of the C57 controls. Interestingly IL-1β was not detected in this experiment.

173

A TNFα B CXCL1 250 *** 500 **** 200 400 *** ** 150 300

100 200 pg/ml BALf pg/ml BALf 50 100

0 0

C57 Sham C57 Sham C57 HK PA14 C57 HK PA14 NDST2-/- Sham NDST2-/- Sham NDST2-/- HK PA14 NDST2-/- HK PA14

C D CXCL2 IL-6 80 100 **** **** **** *** 80 60 60 40 40 pg/ml BALf pg/ml BALf 20 20

0 0

C57 Sham C57 Sham C57 HK PA14 C57 HK PA14 NDST2-/- Sham NDST2-/- Sham NDST2-/- HK PA14 NDST2-/- HK PA14

Figure 3.4.5.9: Cytokine profiling of heat killed PA14 inoculated NDST2-/- mice at 12 hours demonstrate an enhanced CXCL1 response when compared to C57 controls. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), and IL-6 (D). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, **P<0.01, ***P<0.001, ****P<0.0001.

This data taken together demonstrate the protective role NDST2 plays in the pathogenesis of P. aeruginosa in this model.

174

3.4.6 P. aeruginosa infection in mMCP5-/- mice

To determine the role of mast cell restricted chymase mMCP5 in the pathogenesis of

P. aeruginosa mMCP5-/- mice were infected intranasally with 1x106 log phase P. aeruginosa

PA14. At 12 hours post infection mMCP5-/- infected mice show enhanced bacterial clearance in the lung with no changes in the systemic dissemination of the bacteria to the blood or liver when compared to C57 control mice (Figure 3.4.6.1 A-C).

A B C 105 80 600 * 60 400

104 40

CFU/Lung 200

20 Blood CFU/ml CFU/Left Liver lobe Liver CFU/Left

103 0 0

C57 PA14 C57 PA14 C57 PA14 mMCP5-/- PA14 mMCP5-/- PA14 mMCP5-/- PA14

Figure 3.4.6.1: PA14 infected mMCP5-/- mice show enhanced bacterial clearance in the lungs compared to their C57 controls at 12 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A-C). Bacterial returns are by Students t-test *P<0.05.

There was also an enhanced inflammatory response seen in the mMCP5-/- infected mice with a significantly greater leukocyte count in BALf seen in the mMCP5-/- infected group, indeed differential analysis show this increase was attributed to increased neutrophil numbers and to a lesser degree lymphocyte influx, these observations were statistically significant when compare with C57 control mice responses (Figure 3.4.6.2 A-D).

Additionally, no differences were seen between the infected groups when comparing the clinical score (data not shown).

175

A B 1×106 * * 5 **** 1.5×10 ** ** 8×105

6×105 1.0×105

4×105 5.0×104 2×105 Leukocytes/ml BALf

0 BALfMacrophages/ml 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14 mMCP5-/- ShammMCP5-/- PA14 mMCP5-/- ShammMCP5-/- PA14

C D

6 * * 4 1×10 **** 2.0×10 * 8×105 1.5×104 6×105 1.0×104 4×105 5.0×103 2×105 Neutrophils/ml BALf Neutrophils/ml Lymphocytes/ml BALf 0 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14

mMCP5-/- ShammMCP5-/- PA14 mMCP5-/- ShammMCP5-/- PA14

Figure 3.4.6.2: PA14 infected mMCP5-/- mice show comparably more inflammatory cell infiltration predominantly attributed to neutrophilic infiltration at 12 hours post infection when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

176

Histopathological scoring of lung tissue at 12 hours show the characteristic inflammation following infection in both the mMCP5-/- infected mice and their C57 infected controls. Analysis show that the magnitude of this change was comparable between the two infected groups (Figure 3.4.6.3).

Figure 3.4.6.3: Histopathological scoring of PA14 infected mMCP5-/- mice at 12 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C), mMCP5-/- uninfected (D), and mMCP5-/- PA14 infected mice (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiling of the BALf show that mMCP5-/- infected mice have a reduction in

TNFα and CXCL1 responses to infection yet demonstrate an enhanced IL-1β response. All other cytokine responses are comparable to C57 control mice (Figure 3.4.6.4).

177

A TNFα B CXCL1 2500 **** 2500 **** **** **** 2000 2000

1500 1500

1000 1000 pg/ml Balf pg/ml Balf 500 500

0 0

C57 PA14 C57 PA14 C57 Sham C57 Sham

mMCP5-/- ShammMCP5-/- PA14 mMCP5-/- ShammMCP5-/- PA14

C D CXCL2 IL-1β 1500 **** 300 * ** *** **** 1000 200

pg/ml Balf 500

pg/ml Balf 100

0 0

C57 PA14 C57 Sham C57 PA14 C57 Sham mMCP5-/- ShammMCP5-/- PA14 mMCP5-/- ShammMCP5-/- PA14

E IL-6 2000 **** **** 1500

1000 pg/ml Balf 500

C57 PA14 C57 Sham

mMCP5-/- ShammMCP5-/- PA14

Figure 3.4.6.4: Cytokine profiling of BALf from of PA14 infected mMCP5-/- mice at 12 hours demonstrate impaired TNFα and CXCL1 responses together with an enhanced IL-1β response. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

178

With initial results showing mMCP5-/- infected mice at 12 hours display enhanced bacterial clearance, greater inflammation in BALf, altered cytokine responses with reduced

TNFα and CXCL1 countered by an enhanced IL-1β response but an overall unchanged clinical score the experiment was repeated at the 24-hour time point.

Just like 12 hours, the mMCP5-/- PA14 infected mice show enhanced bacterial clearance and unchanged systemic dissemination of the bacteria. (Figure 3.4.6.5 A-C) when compared to their C57 infected controls.

A B C

1×104 ** 1×104 1×104

1×103

1×103 1×103 1×102 CFU/Lung

CFU/ml Blood CFU/ml 1×101 CFU/Left Liver lobe Liver CFU/Left

1×102 1×102 1×100

C57 PA14 C57 PA14 C57 PA14 mMCP5-/- PA14 mMCP5-/- PA14 mMCP5-/- PA14

Figure 3.4.6.5: PA14 infected mMCP5-/- mice show enhanced bacterial clearance in the lungs compared to their C57 controls at 24 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A-C). Bacterial returns are by Students t-test **P<0.005.

179

At 24 hours post PA14 infection the leukocyte counts from BALf of mMCP5-/- infected mice show a dramatic reversal of the 12 hour observations with a significant reduction that was attributed to lower numbers of neutrophils and to a lesser extent, macrophages. Intriguingly lymphocytes were absent (Figure 3.4.6.6 A-C). No differences were seen in clinical scores between the two infected groups (data not shown).

A B C

2.5×106 4×106 **** **** **** **** 2.0×105 *** * 2.0×106 3×106 1.5×105 1.5×106 ** 2×106 ** 1.0×105 1.0×106 1×106 5.0×105 5.0×104 Neutrophils/ml BALf Neutrophils/ml Leukocytes/ml BALf

0.0 0 BALfMacrophages/ml 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14 C57 Sham C57 PA14 mMCP5-/- ShammMCP5-/- PA14 mMCP5-/- ShammMCP5-/- PA14 mMCP5-/- ShammMCP5-/- PA14

Figure 3.4.6.6: PA14 infected mMCP5-/- mice show an impaired inflammatory cell infiltration at 24 hours post infection when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

180

Histopathological scoring of lung tissue at 24 hours again shows the characteristic inflammation following infection in both the mMCP5-/- infected mice and their C57 infected controls. Analysis also shows that the magnitude of this change was again like the observation at 12 hours, comparable between the two infected groups (Figure 3.4.6.7 A-E).

Figure 3.4.6.7: Histopathological scoring of PA14 infected mMCP5-/- mice at 24 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 PA14 infected (C), mMCP5-/- uninfected (D), and mMCP5-/- PA14 infected mice (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiling shows an ongoing defect in CXCL1 responses observed at 12 hours in the mMCP5-/- infected mice, indeed at 24 hours CXCL1 responses would appear to be absent. TNFα and IL-1β responses are restored to comparable levels seen in C57 infected control mice whilst IL-6 responses in the mMCP5-/- infected mice appear muted at 24 hours

(Figure 3.4.6.8). These muted CXCL2 responses could explain the reduction in the scope of inflammatory responses seen in the BALf of the mMCP5-/- infected mice.

181

A TNFα B CXCL1 250 **** **** 150 **** **** 200 100 150

100 50 pg/ml BALf pg/ml BALf 50

0 0

C57 PA14 C57 PA14 C57 Sham C57 Sham

mMCP5-/- ShammMCP5-/- PA14 mMCP5-/- ShammMCP5-/- PA14

C D CXCL2 IL-1β 50000 80 **** **** *** 40000 60 30000 40 20000 pg/ml BALf pg/ml BALf 20 10000

0 0

C57 PA14 C57 PA14 C57 Sham C57 Sham

mMCP5-/- ShammMCP5-/- PA14 mMCP5-/- ShammMCP5-/- PA14

E IL-6 50 **** * 40 **** 30

20 pg/ml BALf 10

C57 Pa14 C57 Sham

mMCP5-/- ShammMCP5-/- Pa14

Figure 3.4.6.8: Cytokine profiling of PA14 infected mMCP5-/- mice at 24 hours demonstrate impaired CXCL1 and IL-6 responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

This data taken together demonstrates that mMCP5 plays a detrimental role in the pathogenesis of P. aeruginosa in this model.

182

3.4.7 P. aeruginosa infection in Prss22-/- mice

To determine the role of Prss22, an epithelial restricted e-tryptase that shares 40% primarily amino acid sequence homology with the b-tryptases, in the pathogenesis of P. aeruginosa Prss22-/- mice were infected intranasally with 1x106 log phase P. aeruginosa

PA14. At 12 hours post infection Prss22-/- infected mice show no changes in bacterial clearance in the lung when compared to C57 control mice (Figure 3.4.7.1) Cultures of homogenised liver and blood yielded now bacterial cultures (data not shown).

105

104

103

102 CFU/Lung

101

100

C57 PA14 Prss22 PA14

Figure 3.4.7.1: PA14 infected Prss22-/- mice show unaltered bacterial clearance at 12 hours post infection. Bacterial returns from homogenised lung. Bacterial returns are by Students t-test.

183

There were no observable differences in leukocyte or differential counts from BALf demonstrating that the magnitude of the inflammatory responses seen in the Prss22-/- infected mice comparable with C57 control mice responses (Figure 3.4.7.2 A-D). Clinical scoring of the mice also shows no differences between the infected groups (data not shown).

A B 8×104 1.5×106 ** **

6×104 1.0×106 4×104

5 5.0×10 2×104 Macrophages/ml BALfMacrophages/ml Leukocytes/ml BALf 0 0.0

C57 Sham C57 PA14 C57 PA14 C57 Sham Prss22 Sham Prss22 PA14 Prss22 Sham Prss22 PA14

C D 8×106 * * 6×104 ** 6×106 4×104

4×106

2×104 2×106 Neutrophils/ml BALf Neutrophils/ml Lymphocytes/ml BALf 0 0

C57 Sham C57 PA14 C57 Sham C57 PA14 Prss22 Sham Prss22 PA14 Prss22 Sham Prss22 PA14

Figure 3.4.7.2: PA14 infected Prss22-/- mice show comparable inflammatory cell infiltration when impaired to their C57 controls at 12 hours post infection. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

184

Histopathological scoring of lung tissue at 12 hours shows the characteristic inflammation following infection in both the Prss22-/- infected mice and their C57 infected controls. Analysis shows that the magnitude of this changes are comparable between the two infected groups (Figure 3.4.7.3 A-E).

Figure 3.4.7.3: Histopathological scoring of PA14 infected Prss22-/- mice at 12 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 PA14 infected (C), Prss22-/- uninfected (D), and Prss22-/- PA14 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Interestingly, despite there being no apparent differences seen in bacterial clearance, inflammatory cells in BALf or clinical score, cytokine profiling demonstrated some major differences. The TNFα response in the infected Prss22-/- was lower than levels observed in controls whilst the CXCL1 and CXCL2 responses were higher than in controls. All other measured cytokines are comparable to C57 controls.

185

A B

TNFα CXCL1 3000 8000 **** **** ** ** 6000 2000

4000

1000 pg/ml BALf pg/ml BALf 2000 *

0 0

Sham PA14 Sham PA14 C57 PA14 C57 PA14 C57 Sham C57 Sham

Prss22-/- Prss22-/- Prss22-/- Prss22-/-

C D

CXCL2 IL-1β 1000 250 **** **** **** 800 **** **** 200

600 150

400 100 pg/ml BALf pg/ml BALf 200 50

0 0

Sham PA14 Sham PA14 C57 PA14 C57 PA14 C57 Sham C57 Sham

Prss22-/- Prss22-/- Prss22-/- Prss22-/-

IL-6 1500 **** ****

1000

500 pg/ml BALf

Sham PA14 C57 PA14 C57 Sham

Prss22-/- Prss22-/-

Figure 3.4.7.4: Cytokine profiling of PA14 infected Prss22-/- mice 12 hours post infection show impaired TNFα response together with an enhanced CXCL1 and CXCL2 responses. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, **P<0.01, ****P<0.0001.

186

With initial results at 12 hours showing unchanged bacterial clearance and inflammation in BALf, comparable clinical scores and just a moderate alteration in cytokine production, namely a reduction in TNFα and increase in CXCL1 and CXCL2 production respectively in the Prss22-/- infected mice, the experiment was repeated at the 24-hour time point.

24 hours post infection Prss22-/- infected mice show no changes in bacterial clearance in the lung, however just 2/7 and 0/7 had detectable bacteria counts in the liver and blood respectively, demonstrating a degree of protection from the systemic dissemination of bacteria to the liver and blood when compared to C57 control mice (Figure 3.4.7.5).

A B C ns 104 1500 * 20000 * 15000 1000

103 10000

500

CFU/ml Blood CFU/ml 5000 CFU/Single lobe CFU/Single CFU/Left Liver lobe Liver CFU/Left

102 0 0

C57 PA14 C57 PA14 C57 PA14 Prss22-/- PA14 Prss22-/- PA14 Prss22-/- PA14

Figure 3.4.7.5: PA14 infected Prss22-/- mice show a degree of protection from bacteraemia at 24 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A- C). Bacterial returns are by Students t-test *P<0.05.

187

There were no observable differences in leukocyte counts from BALf demonstrating that the magnitude of the inflammatory responses seen in the Prss22-/- infected mice are comparable with C57 control mice responses (Figure 3.4.7.6 A-D). However, differential counts indicate a significant increase in lymphocytes and a reduction in macrophages in the

Prss22-/- infected groups. Clinical scoring of the mice also shows no differences between the infected groups (Data not shown).

A B

2.0×106 6 * 5×10 * *

4×106 1.5×106

3×106 1.0×106 2×106 5.0×105 1×106 Leukocytes/ml BALf Macrophages/ml BALfMacrophages/ml 0 0.0

C57 Pa14 C57 Sham C57 Sham C57 PA14 Prss22-/- Sham Prss22-/- Pa14 Prss22-/- ShamPrss22-/- PA14 C D 4×106 ** ** 2.5×104 * 2.0×104 3×106 1.5×104 2×106 1.0×104

1×106 5.0×103 Neutrophils/ml BALf Neutrophils/ml Lymphocytes/ml BALf 0 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14

Prss22-/- ShamPrss22-/- PA14 Prss22-/- ShamPrss22-/- PA14

Figure 3.4.7.6: PA14 infected Prss22-/- mice show comparable inflammatory cell infiltration when impaired to their C57 controls at 24 hours post infection. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

188

Histopathological scoring of lung tissue at 24 hours again shows the characteristic inflammation following infection in both the Prss22-/- infected mice and their C57 infected controls. Analysis also shows that the magnitude of this change identical to the 12 hours time point, comparable between the two infected groups (Figure 3.4.7.7 A-E).

Figure 3.4.7.7: Histopathological scoring of PA14 infected Prss22-/- mice at 24 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 PA14 infected (C), Prss22-/- uninfected (D), and Prss22-/- PA14 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiling at 24 hours didn’t replicated the reduced TNFα responses and elevated CXCL1 and CXCL2 responses seen at 12 hours, with all cytokines measured being comparable to C57 controls (Figure 3.7.6). In this experiment, IL-6 was not detected in any group.

189

Figure 3.4.7.8: Cytokine profiling of PA14 infected Prss22-/- mice 24 hours post infection show comparable cytokine responses compared to C57 mice controls. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), and IL-1β (D). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

The absence of bacteremia in the Prss22-/- infected mice points to Prss22 as detrimental in the pathogenesis of P. aeruginosa in this model.

190

3.4.8 P. aeruginosa infection in RasGRP4-/- mice

To determine the role of RasGRP4 in the pathogenesis of P. aeruginosa RasGRP4-/- mice were infected intranasally with 1x106 log phase P. aeruginosa PA14. At 12 hours post infection RasGRP4-/- infected mice show no changes in bacterial clearance in the lung or degree of systemic dissemination to the blood when compared to C57 control mice (Figure

3.4.8.1).

Figure 3.4.8.1: PA14 infected RasGRP4-/- mice show unaltered bacterial clearance and bacteraemia at 12 hours post infection. Bacterial returns from homogenised multi lobe lung and blood were determined (A-B). Bacterial returns are by Students t-test.

Despite no changes in bacterial clearance or dissemination, leukocyte counts from

BALf shows infected RasGRP4-/- have reduced inflammatory cells (Figure 3.4.8.2 B), with differential counts attributing this to a significant reduction in the magnitude of the neutrophil inflammation (Figure 3.4.8.2 C). All other differential counts show comparable inflammatory responses to infection to those of C57 controls and clinical scoring shows no differences between groups (data not shown).

191

Figure 3.4.8.2: PA14 infected RasGRP4-/- mice show decreased inflammatory cell infiltration attributed to reduced neutrophilic infiltration at 12 hours post infection. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

192

Histopathological scoring of lung tissue at 12 hours shows the characteristic inflammation following infection in both RasGRP4-/- infected mice and C57 infected controls. Analysis shows that the magnitude of this changes are comparable between the two infected groups. (Figure 3.4.8.3 A-E).

Figure 3.4.8.3: Histopathological scoring of PA14 infected RasGRP4-/- mice at 12 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 PA14 infected (C), RasGRP4-/- uninfected (D), and RasGRP4-/- PA14 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001.

Interestingly, despite the reduction in inflammation seen in infected RasGRP4-/- mice cytokine profiling demonstrated comparable responses in CXCL1, CXCL2, IL-1β and IL-6 production to those seen in C57 controls. There was however one difference, with RasGRP4-/- infected mice showing an increase in TNFα compared to C57 controls (Figure 3.8.3 A).

193

Figure 3.4.8.4: Cytokine profiling of PA14 infected RasGRP4-/- mice 12 hours post infection show an enhanced TNFα response. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), IL-1β (D), and IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, ****P<0.0001.

194

With initial results at 12 hours showing unchanged bacterial clearance and a reduction of inflammation in BALf, comparable clinical scores and just a moderate alteration in cytokine production, namely a slightly enhanced TNFα response the experiment was repeated at the 24-hour time point.

24 hours post infection Prss22-/- infected mice show no changes in bacterial clearance in the lung, or differences in the degree of systemic dissemination of bacteria to the liver and blood when compared to C57 control mice (Figure 3.4.8.5).

A B C 105 2000 300

1500 104 200

1000

3 CFU/Lung 10 100

500 (blood) cfu/ml CFU/Left Liver lobe Liver CFU/Left

102 0 0

C57 PA14 C57 PA14 C57 PA14 RasGRP4-/- PA14 RasGRP4-/- PA14 RasGRP4-/- PA14

Figure 3.4.8.5: PA14 infected RasGRP4-/- mice show unaltered bacterial clearance at 24 hours post infection. Bacterial returns from homogenised lung, homogenised left liver lobe and blood were determined (A- C). Bacterial returns are by Students t-test.

Differences were seen in the magnitude of the inflammatory responses to infection in the RasGRP4-/- mice reversing observations seen at 12 hours. Leukocyte counts at 24 hours shows significantly more inflammation in the BALf (Figure 3.4.8.6 A) with subsequent differential counts revealing that these differences were attributed to an enhanced neutrophil inflammation (Figure 3.4.8.6 C) and to a lesser degree higher macrophage numbers (Figure

3.4.8.6 B). Like 12 hours, clinical scores at 24 hours shows no differences between infected groups (data not shown).

195

A B

3×106 * 2.5×105 ** **** 2.0×105 ** 2×106 *** 1.5×105

1.0×105 1×106 5.0×104 Lymphocytes/ml BALf BALfMacrophages/ml 0 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14

RasGRP4-/- ShamRasGRP4-/- PA14 RasGRP4-/- ShamRasGRP4-/- PA14

C D 6 **** 4 5×10 * 2.0×10 4×106 1.5×104 3×106 ** 1.0×104 2×106 5.0×103 1×106 Neutrophils/ml BALf Neutrophils/ml Lymphocytes/ml BALf 0 0.0

C57 Sham C57 PA14 C57 Sham C57 PA14

RasGRP4-/- ShamRasGRP4-/- PA14 RasGRP4-/- ShamRasGRP4-/- PA14

Figure 3.4.8.6: PA14 infected RasGRP4-/- mice show an increased inflammatory cell infiltration at 24 hours post infection. Airway inflammation was determined by BALf leucocyte counts (A) and differential cell counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, differences in cell numbers were analysed by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

196

Histopathological scoring of lung tissue at 24 hours again shows the characteristic inflammation following infection in both RasGRP4-/- infected mice and C57 infected controls. Analysis also shows that the magnitude of this change was again like the observation at 12 hours, comparable between the two infected groups (Figure 3.4.8.7 A-E).

Figure 3.4.8.7: Histopathological scoring of PA14 infected RasGRP4-/- mice at 24 hours show no differences from infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 PA14 infected (C), RasGRP4-/- uninfected (D), and RasGRP4- /- PA14 infected (E) Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

Cytokine profiling, like the leukocyte and differential counts, shows an apparent reversal of results obtained at 12 hours, with the RasGRP4-/- infected mice TNFα responses showing a reduction compared with C57 controls. All other cytokines measured are comparable to controls. Interestingly IL-6 was undetectable in this experiment.

197

A B TNFα CXCL1 300 * 400 **** **** **** **** 300 200

200

100 pg/ml BALf pg/ml BALf 100

0 0

C57 PA14 C57 PA14 C57 Sham C57 Sham

RasGRP4 ShamRasGRP4 PA14 RasGRP4 ShamRasGRP4 PA14

C D

CXCL2 IL-6 **** **** 80 25000 *** *** 20000 60 15000 40 10000 pg/ml BALf pg/ml BALf 20 5000

0 0

C57 PA14 C57 PA14 C57 Sham C57 Sham

RasGRP4 ShamRasGRP4 PA14 RasGRP4 ShamRasGRP4 PA14

Figure 3.4.8.8: Cytokine profiling of PA14 infected RasGRP4-/- mice 24 hours post infection show an impaired TNFα response. Common inflammatory cytokines were determined by ELISA. TNFα (A), CXCL1 (B), CXCL2 (C), and IL-6 (D) Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, ***P<0.001, ****P<0.0001.

Taken together this data shows that RasGRP4 plays a mixed role in the pathogenesis of P. aeruginosa in this model. At 12 hours the absence of RasGRP4 results in reduced inflammation yet at 24 hours this observation is reversed with its absence resulting in elevated inflammation.

198

3.4.9 Summary of results

Bacterial Transgenic Time Airway Histopathology Cytokine Clearance & Mouse Line point Inflammation Scores Profile Bacteremia Comparable to Reduced neutrophilic Comparable to Reduced TNFα & 12 Hrs controls response controls IL-1β responses -/- CFUs in lung Overall inflammation mMCP6 comparable, comparable, yet Comparable to Reduced TNFα & 24 Hrs however extensive macrophage counts controls IL-1β responses bacteremia elevated Impaired bacterial clearance and Impaired neutrophilic Impaired TNFα & 12 Hrs Elevated scores extensive inflammatory response IL-6 responses -/- bacteremia mMCP6 Extensive mortality Extensive mortality Extensive mortality Extensive mortality +/+ 24 Hrs mMCP7 observed observed observed observed 12 Hrs Impaired TNFα & Heat N/A Comparable to controls Elevated scores IL-6 responses Killed Comparable to Comparable to Reduced CXCL2 12 Hrs Comparable to controls controls controls response Prss31-/- Comparable to Comparable to Comparable to 24 Hrs Comparable to controls controls controls controls Reduced TNFα & Elevated inflammation Comparable to Comparable to enhanced CXCL1, 12 Hrs coupled with impaired controls controls CXCL2 & IL-6 macrophage response NDST2-/- responses Extensive mortality Extensive mortality Extensive mortality Extensive mortality 24 Hrs observed observed observed observed

Reduced TNFα & Enhanced bacterial Enhanced neutrophilic Comparable to 12 Hrs CXCL1 & enhanced clearance inflammation controls & IL-1β responses mMCP5-/- Impaired CXCL1 & Enhanced bacterial Reduced neutrophilic Comparable to 24 Hrs reduced IL-6 clearance inflammation controls responses Reduced TNFα & Comparable to Comparable to IL-1β & enhanced 12 Hrs Comparable to controls controls controls CXCL1 & CXCL2 responses Prss22-/- Comparable Over all inflammation bacterial Comparable to controls, Comparable to Comparable to 24 Hrs clearance. yet macrophage counts controls controls Protection from reduced bacteremia Comparable to Comparable to Elevated TNFα 12 Hrs Reduced inflammation controls controls response RasGRP4 -/- Comparable to Comparable to Reduced TNFα 24 Hrs Enhanced inflammation controls controls response

Table 3.4.9.1: Summary of P. aeru gin osa infected mice deficient in mast cell proteases, associated

proteases or mast cell associated factors and different time points during infection.

199

3.5 Discussion

This research builds on the research of work of Junkins et al., who demonstrated the role of an unknown secreted mast cell factor in the maintenance of respiratory epithelium integrity during a P aeruginosa infection94. Mast cells have traditionally been associated with a proinflammatory allergic response but more recently they have been also shown to have immunomodulatory roles.

Using a P. aeruginosa infection model that elicits immune responses with the hallmarks of an acute human P. aeruginosa pneumonia in transgenic mice, I show that mast cell proteases, their related proteases and associated factors play important roles in driving innate immune responses during a transient P. aeruginosa pneumonia.

Mast cells are sentinels of the immune system, strategically located and enriched at the entry points for P. aeruginosa. They play a key role in triggering and sculpting the inflammatory responses and have been shown to influence innate immunity as well as delaying adaptive immunity during infections59. There is growing evidence that mast cell derived proteases play import roles during infection204.

This study shows that mice devoid of the β-tryptase mMCP6 have an impaired inflammatory response to P. aeruginosa at 12 hours that altered at 24 hours compared to controls. This reduction in inflammation coupled with an apparent absence of a TNFα response at 12 hours post infection could explain the subsequent widespread bacteraemia seen in the infected mMCP6-/- mice at 24 hours, given that TNFα recruits inflammatory cells via the up regulation of adhesion molecules238. Any delay in the typical robust neutrophil response gives P. aeruginosa an opportunity to colonise the respiratory epithelium cell surface, utilising virulence factors to disrupt the integrity of the epithelium and facilitating dissemination into the blood. These observations are consistent with the findings of

Thakurdat et al. who shows mMCP6 to be a potent inducer of neutrophilic inflammation3.

200

Whilst the precise mechanism are not fully described experiments have shown that mMCP6 induces IL-8 in human endothelial cells and it is postulated that mMCP6 works by inducing bystander cells to produce large amounts of neutrophil specific chemotactic factor138.

Collectively this data is supportive to the unpublished findings mentioned in a paper194 where

Adachi and Stevens report that this mMCP6-/- mice strain demonstrated a higher susceptibility to Pseudomonas194. Running this model out to 48 hours one would expect fatalities given the extensive bacteraemia observed at 24 hours post infection.

Taken together this data demonstrate a positive role for mMCP6 during P. aeruginosa infection, the observation that mMCP6-/- mice fail to generate a TNFα response at both 12 and 24 hours post infection indicates a potential new role for mMCP6, specifically promoting

TNFα production and protecting from bacteraemia.

The genetic introduction of β-tryptase mMCP7 to mice proved to be lethal in this model with 50% of infected mMCP6-/- mMCP7+/+ mice succumbing to infection within 24 hours.

The impaired bacterial clearance, widespread systemic dissemination of P. aeruginosa and higher clinical scores match the observation of reduced inflammation in BALf. This points to an impairment in inflammatory cell migration across the epithelial membrane following infection. The absence of TNFα and IL-6 responses in the BALf, two cytokines that work synergistically together, would restrain the respiratory burst capacity of neutrophils and impair the rate of phagocytosis and superoxide anion generation231. This is consistent with the observations of enhanced systemic dissemination of the bacteria and impaired bacterial clearance observed in the lungs at 12 hours.

These results are interesting given the inherent caspase 11 deficiency of these mice, they would be expected to show greater resistance to sepsis229. Why the opposite is true remains unclear.

201

Studies of mMCP6-/- mMCP7+/+ mice inoculated with heat killed PA14 revealed similar neutrophilic inflammatory responses and histopathological scores to that of C57 inoculated controls, whilst as expected the magnitude of the response was drastically lower than seen in a live infection, with neutrophilic inflammation around 100 x lower than the typical counts seen in a live infection and histopathological scoring being approximately 1/2 the typical score of a live infection at 12 hours. Whilst these responses were lower, clinical symptoms in the mMCP6-/- mMCP7+/+ inoculated mice were much worse resulting in a higher clinical score than that of the C57 inoculated controls. Indeed, the elevated clinical scoring in the mMCP6-/- mMCP7+/+ inoculated mice resulted in the halting of the experiment and the sacrificing of the mice at 12 hours.

Another interesting observation in the mMCP6-/- mMCP7+/+ mice exposed to heat killed

PA14 was the absence of the cytokine TNFa following challenge. The wild type infection also resulted in no TNFa detection in the BALf. In chapter 2, I demonstrate that TNFa is produced following an S. pneumoniae infection in these mice. Collectively these data point towards mMCP6-/- mMCP7+/+ mice having a defective TNFa response to P. aeruginosa.

Why the mMCP6-/- mMCP7+/+ inoculated mice faired so poorly in the heat killed model is unclear. A possible explanation toxic shock, or perhaps they are less able to handle an assault of antigen provided by the heat killed bacteria, however given the inherent caspase 11 deficiency in these mice and that previous studies have shown caspase 11 deficient mice to be protected from sepsis229. The underlining cause points to the detrimental addition of mMCP7 in this model acting in a caspase 11 independent manner.

Collectively these results show that a live P. aeruginosa infection is required to induce the robust neutrophil response associated with this model and that mMCP6-/- mMCP7+/+ mice have an impaired TNFα and IL-6 responses following exposure to live or heat killed

P. aeruginosa. How this response is impaired is unclear but caution must be used when

202

interpreting these results alone. mMCP7 forms hetrotetramers with itself and mMCP6, the absence of mMCP6 in this mouse leads to the ubiquitous production of mMCP7 homotetramers that would be prevalent in much higher concentrations than in normal physiological conditions192. Indeed these homotetramers would be absent in C57 mice given their exon2/intron2 splice site mutation at mMCP7 locus193. These mMCP7 homotetramers would not be confined to the extracellular matrix as they can readily disassociate from their serglycin proteoglycans, unlike their heterotetramer counterparts. This ability to readily disseminate together with mMCP7s potent anticoagulant activity197 could explain the observations of widespread systemic dissemination of bacteria seen at 12 hours in mMCP6-/- mMCP7+/+ mice.

BALB/c mice, unlike C57 mice have retained a functional mMCP7 locus. In a study comparing the lung defences of eleven different strains of mice to P. aeruginosa challenge

C57 mice demonstrated the greatest resistance to infection241. Whether this resistance is attributed to their mMCP7-/- status specifically is unclear but BALB/c mice do not demonstrate the vulnerability to P. aeruginosa that the mMCP6-/- mMCP7+/+infected mice do.

Experiments deleting mMCP7 in these mice would help address this issue. Given the ability of mMCP6 and mMCP7 to form heterotetramers with one another an infection profile from a mMCP6+/+ mMCP7+/+ infected mouse could provide further clarification on the role of mMCP7 homotetramers versus mMCP6/mMCP7 heterotrtramers during this infection.

The overall observations that the mMCP6-/- mMCP7+/+ mice are more vulnerable to

P. aeruginosa infection cannot be attributed to the presence of mMCP7 alone. The mMCP6-/- mMCP7-/- also demonstrate phenotypic attributes that also contribute to the pathogenesis of P. aeruginosa infection in these mice. Overall the data shows that mMCP7 is extremely deleterious in the pathogenesis of P. aeruginosa infection.

203

As mentioned in previous chapters, any heparin bound tryptase data needs to be considered in the context of the NDST2-/- infection data. The results here indicated that mMCP6 is proinflammatory during P. aeruginosa infections and given that NDST2-/- mast cells granules are depleted of heparin proteoglycans affecting mMCP6 and mMCP7, the increases in the inflammatory responses seen in the NDST2-/- infected mice at 12 hours must be independent of mMCP6.

NDST2-/- infected mice had comparable bacterial clearance in the lungs at 12 hours compared to C57 infected controls and indicated a degree of protection from the systemic dissemination of bacteria as no bacteria was recovered from the livers of the NDST2-/- infected mice. The NDST2-/- infected mice demonstrated a more intense inflammatory response dominated by neutrophils yet shows a reduction in macrophage numbers when compared to C57 infected mice. Although it must be noted the NDST2-/- uninfected mice shows higher macrophage levels than found in controls.

The enhanced intensity of the inflammatory response seen in this model can be explained in part by the enhanced CXCL1 and CXCL2 responses in BALf of NDST2-/- infected mice.

The other changes in cytokine production seen in NDST2-/- infected mice could explain the fatality rates of 80% seen in these mice at 24 hours. The drastically impaired TNFα and IL-6 responses would impair respiratory burst capacity of neutrophils together with the impairing the rate of phagocytosis and superoxide anion generation in neutrophils231. These factors could contribute to bacterial pathogenesis resulting in the lethal outcomes seen.

Looking at the data from the heat killed 12 hour experiment it was unfortunate that there were no discernible differences seen that could indicate why NDST2-/- infected mice were so susceptible to this infection.

It is clear these NDST2-/- infected mice mount a stronger leukocyte response than their

C57 infected controls, this together with the absence of a robust TNFα and IL-6 points to

204

these recruited cells being defective, or suppressed in their response. The precise mechanism at play in this model is unclear and warrants further study. However as NDST2 is required for the production of heparin, a protein that stabilises many mast cell derived proteases, NDST-/- mice mast cells have a significantly altered protease composition. Any alterations to proteases content could contribute during this infection and result in the observations seen. It is evident NDST2 is essential for moderation of the inflammatory response during

P. aeruginosa infection, and indeed the previous chapter shows the same is true during

S. pneumoniae infection. The extensive disruption to the composition of mast cell granules seen in NDST-/- mice make pinpointing specific functions difficult.

Again the inherent deficiency in caspase 11 present in these mice, owing to their 129Sv background also needs to be considered, given this defect confers resistance to LPS induced sepsis229. The high death rates seen at 24 hours is again intriguing and warrants further study.

The absence of IL-1β production in response to heat killed P. aeruginosa points to the necessity of a live infection to induce IL-1β production, this taken together with the inherent caspase 11 deficiency in these mice demonstrates that IL-1β production in a live

P. aeruginosa is independent of caspase 11.

Prss31-/- infected mice show comparable bacterial clearance, inflammatory cell responses and histopathological scores together with cytokine profiles at 12 and 24 hours post infection similar to those of C57 infected controls. The only notable difference was the CXCL2 production in BALf at 12 hours, which in the Prss31-/- infected mice was around 50% lower than C57 infected controls. This reduced CXCL2 response didn’t seem to have any observable impact on the overall host response to the infection. Collectively this data indicates that Prss31 plays no discernible role in the pathogenesis of P. aeruginosa infection.

Interestingly given that mMCP6-/- mMCP7+/+ mice are expected to contain negligible amounts of Prss31 on account of their 129Sv cross background and the close proximity of

205

Prss31 to mMCP7 within the murine genome, just some 2.3kb apart196. This data infers that the observations seen in the mMCP6-/- mMCP7+/+ mice are wholly attributed to the introduction of mMCP7 to the model and not a reduction in Prss31.

In the absence of the chymase mMCP5 there was enhanced P. aeruginosa clearance at both 12 and 24 hours post infection when compared to C57 infected controls. At 12 hours there was an enhanced leukocyte recruitment in the BALf that by 24 hours had reversed becoming a reduced response. At 24 hours in the mMCP5-/- infected mice the reduced inflammation observed together with the enhanced bacterial clearance from the lung is consistent with observations in Chapter 2 that indicate mMCP5 is deleterious in an inflammatory environment of S. pneumoniae associated lung inflammation and now at the

24-hour stage in P. aeruginosa infection. The reason why the inflammation resolved at 24 hours could be attributed to the impaired CXCL1 responses seen in the BALf of mMCP5-/- infected mice at 24 hours.

Interestingly there was an ablated macrophage response at 24 hours that whilst not typically significant in this model, in this instance the absence of any increase in numbers above that of baseline was. There was also a difference seen at 12 hours, however, the mMCP5-/- uninfected group demonstrated unusually high macrophage levels so interpreting this observation is difficult. The 24-hour observation of an ablated macrophage response is particularly poignant given similar observations were seen in during S. pneumoniae infection

(Chapter 2). As one of the biggest producers of CXCL1 are macrophages, the reduction in macrophages in mMCP5-/- mice may explain the reduced amounts of this cytokine detected.

Reduced CXCL1 levels typically point to reduced neutrophil recruitment, however, this was not the observation in this model. As mentioned previously, cytokines often have overlapping functionality, it is likely that an untested cytokine is compensating for the reductions of

CXCL1 and CXCL2 seen in the mMCP5-/- infected mice.

206

The absence of sepsis in these mice allows for the exclusion of caspase 11 deficiency as a primary cause for the sepsis observed in the mMCP6-/- mMCP7+/+ and NDST2-/- infected mice.

Again, like tryptases, chymase expression is altered in NDST2-/- mice which have very low levels of mMCP5208. Previous studies have suggested even these small amounts of mMCP5 in NDST2-/- mice can be sufficient to induce additional inflammation208. C57 mice have normal levels of mMCP5 and during P. aeruginosa infections they were somewhat protected compared to NDST2-/- mice indicating, that in this P. aeruginosa model, just like the S. pneumoniae model, the role of mMCP5 in the NDST-/- mice infected with P. aeruginosa is negligible if anything.

Overall this data shows that in the absence of mMCP5-/- there is elevated inflammation in the BALf at 12 hours with enhanced bacterial clearance. This inflammation is greatly reduced by 24 hours with continued enhanced bacterial clearance when compared with C57 infected controls. This data suggests mMCP5 has immunomodulatory functions dependent upon the stage of infection, acting as an anti-inflammatory mediatory early in the infection and a proinflammatory mediator in later stages of the infection.

Data from Prss22-/- infected mice at 12 hours post infection show no differences in bacterial clearance or inflammatory cell recruitment but did demonstrate significant alterations in cytokine responses with impaired TNFα and enhanced CXCL1 and CXCL2 responses. Higher levels of the chemotactic chemokines CXCL1 and CXCL2 recruit more inflammatory cells to the lung, contributing to in part, to the heightened histopathological scores seen in Prss22-/- infected mice when compared to their C57 infected controls. At 24 hours Prss22-/- infected mice are protected from dissemination of bacteria to the blood and liver. There was a trend to impaired bacterial clearance in the lungs of Prss22-/- mice, but this was not statistically significant. There was also reduced macrophage inflammation seen in the

207

Prss22-/- infected mice at 24 hours, something unexpected given the enhanced CXCL1 and

CXCL2 responses seen at 12 hours. A limitation of the 12 and 24-hour time points is that they provide only snapshots; it would be interesting to repeat the experiment in order to cover more time points so that the kinetics of the infection may be better described. Overall this data demonstrates the epithelial cell restricted epsilon tryptase Prss22 is detrimental during

P. aeruginosa pathogenesis by promoting the dissemination of bacteria to the blood and liver.

How Prss22-/- mice are protected from bacteraemia at 24 hours is unclear, but given that bacteraemia is a serious complication of P. aeruginosa associated pneumonia with high mortality rates, understanding the role Prss22 plays during this infection could yield potential therapeutic targets to protect patients from bacteraemia.

At the start of this study RasGRP4 was believed to be exclusively expressed in mast cells, however more recent work has shown it to be present on splenic CD117+ dendritic cells225.

This new finding introduces new questions that the initial study was not planned to address.

In my model RasGRP4-/- infected mice at 12 hours show reduced inflammation that was predominantly due to reduced neutrophilic influx in BALf and subtle increase in TNFα production. There were no other discernible differences seen. However, at 24 hours the trends were reversed showing enhanced inflammation in BALf attributed to neutrophils and to a lesser degree macrophages. There was also a subtle reduction in TNFα responses detected in

BALf. Together this data suggests RasGPR4 is proinflammatory in the inital stages of

P. aeruginosa infection, an observation that is consistent with the S. pneumoniae findings in

Chapter 2 and with published studies223. However, data from 24 hours post infection shows the reverse with RasGRP4 appearing to play an anti-inflammatory role. This finding of a potential anti-inflammatory role for RasGRP4 contradicts the findings of other groups who’s data, albeit in a colitis model, points to RasGRP4 as being proinflammatory223. Collectively this data reveals a mixed role for RasGRP4-/- during early and late infection.

208

In conclusion, I show that mMCP6 and NDST2 play protective roles during P. aeruginosa infections, whereas mMCP7 and Prss22 play deleterious roles contributing to mortality and systemic bacterial dissemination respectively. Additionally, I demonstrated that Prss31 plays a limited role in the pathogenesis of P. aeruginosa and described a novel role for mMCP5 in immunomodulation and the recruitment of macrophages during infection. Finally, I show that

RasGRP4 plays a mixed immunomodulatory role in this model which is dependent upon the stage of the infection.

209

Chapter 4:

The role of mast cell proteases, their related proteases and mast cell associated factors in the pathogenesis of Influenza A virus in a A/WSN/33 model in mice.

210

4.1 Abstract

Recent studies suggest mast cells play a key role in host responses to Influenza virus infection. Studies have shown that mast cell degranulation contributes towards excessive inflammatory cell influx and immune pathology during IAV infections. Using transgenic mice deficient or competent in a range of mast cell proteases, mast cell related proteases and mast cell associated factors I investigated the role these components play in the pathogenesis of IAV infection model.

I demonstrate that the tryptases mMCP6 and Prss22 are, together with mast cell associated factors NDST2 and RasGRP4, immunomodulatory in the pathogenesis of IAV infection. Additionally, I show Prss31 to be proinflammatory during the viral clearance phase of this infection and reveal a mixed role for mMCP7 acting as anti-inflammatory at the peak of the infection but becoming proinflammatory during the resolution phase. Finally, I show that mMCP5 appears to play a neutral role in the pathogenesis of IAV infection.

In conclusion, I again demonstrate the diverse role of mast cell proteases, their associated proteases and mast cell related factors play during an infection. I have identified

Prss31 as a potential therapeutic target to promote the tolerance of IAV infection whilst reducing infection associated inflammation.

211

4.2 Introduction

Influenza A virus, a member of the Orthomyxoviridae family of viruses contain an 8 segmented negative strand ssRNA genome encoding up to 16 proteins, although not all viruses express all 1698. In humans, the virus is typically restricted to the upper respiratory system and can cause symptoms for 7-10 days100. Annually Influenza is responsible for between 300,000 and 500,000 deaths globally101 with infants, elderly and immunocompromised individuals being at greatest risk of presenting with enhanced disease severity and carry a higher mortality rates102.

The virus has a high mutation rate, attributed to the absence of a proofreading capacity in the Influenza polymerase. This high mutation rate contributes towards antigenic drift and is responsible for seasonal epidemics of the virus. Occasionally two different subtypes of Influenza virus will infect a host, often in animal reservoirs, resulting in genetic re-assortment between the Influenza strains, generating novel subtypes that could trigger pandemics. This process is called antigenic shift and leads to new pandemics every 10-40 years.

With treatment options limited to a few antivirals, with diminishing efficacy, such as

Tamiflu whose usefulness has been brought into doubt in recent Cochrane reviews, and seasonal vaccinations used as a prophylactic intervention in at risk groups. The need for new treatments has never been higher. Research in to the pathogenesis of IAV infection can highlight potential new therapeutic options for patient with this disease.

Mast cells have traditionally been overlooked during IAV infections. Recently studies have shown mast cells play a detrimental role during IAV infection, contributing towards excessive inflammation and immune pathology133,134. These studies demonstrate that IAV can directly infect mast cells and trigger degranulation. Collectively this data demonstrates mast cells play a pivotal role in the pathogenesis of IAV infection.

212

In this study, I use 7 transgenic mice lines that are deficient or competent in a range of mast cell proteases, related proteases and related factors, in a murine A/WSN/33 respiratory infection model to elucidate the roles of mast cell proteases, their associated proteases and mast cell factors in the pathogenesis of IAV infection. A/WSN/33 has been shown to directly infect mast cells triggering their degranulation134 and was selected due to this ability and the mildness of the resultant infection.

213

4.3 Methods

4.3.1 Ethics statement

These studies were conducted in accordance with the New South Wales Animal

Research Act 1985, abiding by the animal research regulations (2010) of the said act and the

National Health and Medicine Research Councils Australian code for the care and use of animals for scientific purposes 8th Edition (2013). All protocols used were approved by the

University of Newcastle’s Animal Care and Ethics Committee (permit number 987/0111) under ethics application number A-2012-217.

4.3.2 Clinical score

Animals were monitored throughout the experiment and assigned a clinical score based of the following:

Clinical Clinical signs Score

1 Healthy with no signs of illness

2 Consistently ruffled fur, especially on neck

3 Piloerection, breathing may be deeper and mice less alert

4 Labored breathing. Frequently showing tremors and lethargy

5 Frequently emaciated. May show cyanosis of tail & ears

6 Death *Animals scoring 4 or 5 are euthanized immediately

214

4.3.3 H1N1 A/WSN/33 infection model

H1N1 A/WSN/33 was kindly provided by Patrick Reading, from WHO Collaborating

Centre for Reference and Research of Influenza, Victoria, Australia. Diluted working stocks of 15000 plaque forming units per ml (pfu/ml) of A/WSN/33 were prepared and stored at -80°C until required. Working concentrations and viability were verified by plaque assay.

7-9 weeks old male C57 or relevant transgenic mice were anesthetised using isoflurane and inoculate intranasally (i.n) with 50µl containing 750pfu (15000pfu/ml). After 7 or 10 days the mice were sacrificed using pentobarbital overdose and tissues collected for endpoint analysis.

4.3.4 Cellular inflammation

Na2) on ice for 5 minutes; the supernatant was aspirated and snap frozen in liquid nitrogen for later cytokine analysis, the cells were washed and re-suspended in PBS, total cell numbers were calculated using a haemocytometer, then placed in microscope slide containing cassettes and cytocentrifuged at 300g for 5 minutes (ThermoFisher Scientific, Waltham, MA). After drying overnight, the slides were stained with May-Grünwald-Giemsa and observed using a light microscope. Leukocyte enumeration was conducted raised of morphological criteria with a total of 200 cells being counted.

215

4.3.5 Viral BALf collection

Bronchoalveolar lavage fluid (BALf) was collected from the mutlilobed (right) lung, briefly, the left lung was tied off with cotton thread and 2 x 700µl of PBS was used to wash the right lung via a blunt cannular. The returned volume was noted and half was snap frozen in liquid nitrogen then transferred to -80°C storage.

4.3.6 Plaque assay

Madin-Darby Canine Kidney (MDCK) cells were grown to approximately 85% confluence then washed in PBS and submerged in Leibovitz’s L-15 (L-15) medium

(Invitrogen, MA) supplemented with HEPES (Invitrogen, MA) and trypsin-TPCK

(Invitrogen, MA). Snap frozen BALf samples were serially diluted and incubated with cells for 1 hour at 37°C 5% CO2. Following incubation, the inoculations as removed and cells were covered in a thin overlay of 1.8% low melt agarose (Sigma, MO) in L-15 media. Cells were then incubated at 37°C 5% CO2 for 3 days then plaques enumerated.

4.3.7 Cytokine expression in BALf

Cytokine concentrations in BALf were determined using DuoSet® ELISA kits for

TNFα, IL-10, IP-10, IL-1β, IFNg and IL-6 (R&D Systems, Minneapolis, MN) as per manufactures instructions.

4.3.8 Statistics

Statistical analysis was performed with GraphPad Prism version 6f. All results were expressed as a mean ± SEM from 5-8 mice. Analysis performed is as indicated in the figures.

Values of a p < 0.05 was deemed a significant difference.

216

4.3.9 Histopathological scoring

Score 1: Airways Inflammation Score /4

Score 2: Vascular Inflammation Score /4

Score 3: Parenchymal Inflammation (at 10X magnification) Score /5

0 <1% affected 1 1-9% affected 2 10-29% affected 3 30-49% affected 4 50-69% affected 5 >70% affected

Score 1 + Score 2+ Score 3 = Histopathological score

217

4.4 Results

4.4.1 Characterisation of H1N1 A/WSN/33 infection

In order to investigate the roles of mast cell proteases, associated proteases and other mast cell factors play in the pathogenesis of A/WSN/33 a murine model was established in 8 week old male C57 mice and characterised. Intranasal inoculation with 750 pfu A/WSN/33 produced a productive infection that induced a strong inflammatory response comprising of macrophages and neutrophils and lymphocytes.

Infected mice begin to lose significant amounts of weight by day 8 with these losses levelling off by day 9 and some indications of weight recover at day 10 (Figure 4.4.1.1 A).

Virus titres climb during the first week following infection peaking at day 7 post infection and being cleared by day 10 (Figure 4.4.1.1 B). Inflammatory cells counts reveal little increase in inflammatory cells at day 3 p.i however there are significant increase in macrophages, neutrophils and lymphocytes by day 7 p.i. Whilst neutrophil responses plateau at day 7 p.i macrophage and lymphocyte numbers continue to rise at day 10 p.i (Figure

4.4.1.1 C, D, E and F).

218

Figure 4.4.1.1: Characterisation of A/WSN/33 infection. IAV infection was assessed in 8-9 week C57 male mice. Analysis of weight loss (A) and viral return from BALf (B). Airway inflammation was represented by total leukocyte counts in BALf (C) and May-Grünwald-Giemsa staining enabled macrophage, neutrophil and lymphocyte differential counts (D, E & F). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001 ****P<0.0001.

219

4.4.2 Influenza A infection in mMCP6-/- mice

To determine the role of mast cell specific tryptase mMCP6 in the pathogenesis of

Influenza A mMCP6-/- mice were infected intranasally with 750pfu of A/WSN/33. mMCP6-/- infected mice followed a similar weight loss profile to that of their C57 infected controls up to 8 days post infection (p.i) at which point the mMCP6-/- mice failed to gain weight like their

C57 controls. At day 10 p.i the difference between the two infected groups had reached statistical significance (Figure 4.4.2.1).

Figure 4.4.2.1: mMCP6-/- A/WSN/33 infected mice fail to recover weight by day 10 post infection. Data are means ± SEM, two-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001.

220

At day 7 p.i mMCP6-/- infected mice show a lower viral peak when compared to C57 control mice (Figure 4.4.2.2).

Figure 4.4.2.2: mMCP6-/- infected mice have a lower viral peak than infected C57 controls. BALf viral titers were determined at day 7 post infection by plaque assay on MDCK cells. Data are means ± SEM, plaque forming units are by Students t-test *P<0.05.

221

Despite the reduced viral peak seen in mMCP6-/- infected mice, the magnitude of the inflammatory responses seen in BALf is comparable to C57 infected controls. Differential analysis show the inflammatory response to be comprised of neutrophils, macrophages and lymphocytes. With these increases being statistically significant for each cell type compared to uninfected controls (Figure 4.4.2.3).

Figure 4.4.2.3: A/WSN/33 infected mMCP6-/- mice show comparable inflammatory cell infiltration attributed to macrophage, neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

222

Histopathological scoring of lung tissue at 7 days p.i show the characteristic inflammation typical following A/WSN/33 infection in both the mMCP6-/- infected mice and their C57 infected controls.

Figure 4.4.2.4: Histopathological scoring of the lungs from A/WSN/33 infected mMCP6-/- mice at 7 day p.i show comparable scores to those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 A/WSN/33 infected (C), mMCP6-/- uninfected (D), and mMCP6-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

At 7 days p.i cytokine profiles from the BALf of mMCP6-/- Influenza infected mice revealed reduction in IL-6, IFNg and increase in IL-10, each being of statistical significance compared to C57 infected control mice. All other cytokine responses are comparable to C57 infected control mice.

223

Figure 4.4.2.5: Cytokine profiling of A/WSN/33 infected mMCP6-/- mice at 7 days p.i demonstrate comparable cytokines responses to their C57 infected controls. Common inflammatory cytokines were determined by ELISA. TNFα (A) IL-10 (B), IP-10 (C), IFNg (D), IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

224

Both mMCP6-/- and C57 mice cleared the A/WSN/33 infection by day 10 p.i, with no plaques detected by plaque assay (data not shown). Enumeration of cells from BALf show that the overall inflammatory cell influx was greater at 10 days p.i than at 7 days p.i. leukocyte counts were elevated in mMCP6-/- infected mice at day 10 day p.i when compared to their C57 control group. Differential analysis revealed that this increase was attributed to elevated macrophage and lymphocyte infiltration (Figure 4.4.2.6 A-D).

Figure 4.4.2.6: A/WSN/33 infected mMCP6-/- mice show elevated inflammatory cell infiltration attributed to macrophages and lymphocyte infiltration at day 10 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

225

Histopathological scoring of A/WSN/33 infected mMCP6-/- mice lung tissue at 10 days p.i corresponds with the elevated inflammatory cell counts observed in BALf. These mice have extensive disease associated pathology compared to their C57 infected controls.

Figure 4.4.2.7: Histopathological scoring of the lungs from A/WSN/33 infected mMCP6-/- mice at 10 day p.i are elevated compared to the scores seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 A/WSN/33 infected(C), mMCP6- /- uninfected(D), and mMCP6-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

At 7 days p.i cytokine profiles from the BALf of mMCP6-/- Influenza infected mice revealed an increase in IL-10, IP-10 and IFNg. At day 10 p.i IL-10 responses remained elevated, but IL-6 responses had returned to comparable levels seen on C57 infected controls.

In this model IFNg and TNFα concentrations in the BALf at day 10 p.i are typically below the threshold for detection. However, the infected mMCP6-/- mice at day 10 p.i all had IFNg present, albeit at relatively low concentrations. The expression of this proinflammatory cytokine corresponds with the worsening inflammation seen in the BALf and

226

histopathological scores of these mice. Additionally, elevated levels of the lymphocyte chemokine IP-10 were also corresponding with the observed elevated lymphocyte inflammation in BALf. TNFα and IL-1β were not detected (data not shown).

Figure 4.4.2.8: Cytokine profiling of A/WSN/33 infected mMCP6-/- mice at 10 days p.i demonstrate impaired IL-10 and slightly elevated IL-6 responses. Common inflammatory cytokines were determined by ELISA. IL-10 (A), IP-10 (B), IFNg (C) IL-6 (D). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

In summary following A/WSN/33 infection mMCP6-/- mice failed to begin to regain weight by day 10, showed a reduced peak viral titre, had greater numbers of inflammatory cell in the BALf, had elevated histopathology scores and showed elevated IL-10, IP-10 and

227

IFNg. These data taken together show mMCP6, at day 10 post infection, to be anti- inflammatory, protecting against excessive inflammation in the lung.

4.4.3 Influenza A infection in mMCP6-/- mMCP7+/+ mice

To determine the role of mast cell specific tryptase mMCP7 in the pathogenesis of

Influenza A mMCP6-/- mMCP7+/+ mice were infected intranasally with 750pfu of A/WSN/33. mMCP6-/- infected mice followed a similar weight loss profile to that of their C57 infected controls. The uninfected controls also show a similar weight gain profile up to day 7 p.i when the two groups diverged with the mMCP6-/- mMCP7+/+ mice gaining weight at a faster pace that the C57 uninfected controls. By day 10 p.i this difference had reached statistical significance (Figure 4.4.3.1).

Figure 4.4.3.1: mMCP6-/- mMCP7+/+ A/WSN/33 infected mice show similar weight loss to infected C57 controls. Data are means ± SEM, two-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

228

At day 7 p.i mMCP6-/- mMCP7+/+ infected mice an elevated viral peak in the lung when compared to C57 control mice (Figure 4.4.3.2).

Figure 4.4.3.2: mMCP6-/- mMCP7+/+ infected mice have a higher viral peak than infected C57 controls. BALf viral titers were determined at day 7 post infection by plaque assay on MDCK cells. Data are means ± SEM, plaque forming units are by Students t-test ***P<0.0005.

Together with an elevated viral peak mMCP6-/- mMCP7+/+ infected mice show reduced levels of leukocytes in BALf. Differential analysis shows that this reduction was attributed to a somewhat muted macrophage response together with a slightly lower lymphocyte count, however neither of these were significant in their own right (Figure

4.4.3.3 A-D).

229

Figure 4.4.3.3: A/WSN/33 infected mMCP6-/- mMCP7+/+ mice show reduced inflammatory cell infiltration at day 7 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

230

Histopathological scoring of lung tissue at 7 days p.i shows the characteristic inflammation typical following A/WSN/33 infection in both the mMCP6-/- mMCP7+/+ infected mice and their C57 infected controls. While the mMCP6-/- mMCP7+/+ histopathological scores are slightly lower than the C57 infected controls these observations do not reflect the observations of reduced inflammatory cell influx observed in the BALf.

Figure 4.4.3.4: Histopathological scoring of the lungs from A/WSN/33 infected mMCP6-/- mMCP7+/+ mice at 7 day p.i show comparable scores to those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 A/WSN/33 infected (C), mMCP6-/- mMCP7+/+ uninfected (D), and mMCP6-/- mMCP7+/+ A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001 and ****P<0.0001.

At 7 days p.i cytokine profiles from the BALf of mMCP6-/- mMCP7+/+ Influenza infected mice revealed reduced IL-10 production and a defective IFNg response. All other cytokine responses are comparable to C57 infected control mice.

231

Figure 4.4.3.5: Cytokine profiling of A/WSN/33 infected mMCP6-/- mMCP7+/+ mice at 7 days p.i demonstrate impaired IL-10, IFNg and IL-6 responses. Common inflammatory cytokines were determined by ELISA. TNAα (A) IL-10 (B), IP-10 (C), IFNg (D), IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

232

Both mMCP6-/- mMCP7+/+ and C57 mice cleared the A/WSN/33 infection by day 10 p.i, with no plaques detected by plaque assay (data not shown). Enumeration of cells from

BALf shows that the overall inflammatory cell influx was greater at 10 days p.i than at 7 days p.i. There were reduced leukocyte counts in mMCP6-/- mMCP7+/+ infected mice at day 10 p.i when compared to their C57 control group. Differential analysis revealed that this reduction was attributed to an absence of the macrophage inflammatory response, an observation that tallies with the same observation at day 7 p.i, suggesting that in this model, mMCP7 may play a role in curtailing macrophage inflammation (Figure 3.4.3.6 A -D).

Figure 4.4.3.6: A/WSN/33 infected mMCP6-/- mMCP7+/+ mice show reduced inflammatory cell infiltration attributed to macrophages and lymphocytes infiltration at day 10 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

233

The observation of comparable histopathological scores seen at day 7 p.i are reversed at day 10 p.i with mMCP6-/- mMCP7+/+ mice having an increase in their scores. Again, these results do not tally the observations seen in BALf of a reduced inflammatory cell count, but could point to impaired cell migration into the airway lumen in these mMCP7 knock in mice.

Figure 4.4.3.7: Histopathological scoring of the lungs from A/WSN/33 infected mMCP6-/- mMCP7+/+ mice at 10 day p.i show elevated scores than those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 A/WSN/33 infected (C), mMCP6-/- mMCP7+/+ uninfected (D), and mMCP6-/- mMCP7+/+ A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, **P<0.01 and ****P<0.0001.

The observed reduction in concentrations of IL-10 seen at day 7 p.i in mMCP6-/- mMCP7+/+ infected mice was repeated in these mice at day 10 p.i. Additionally there were reduced levels of IL-6 detected, an observation in keeping with the reduced inflammatory cell influx observed in the BALf, TNFα, IFNg, and IL-1b were not detected (data not shown).

234

Figure 4.4.3.8: Cytokine profiling of A/WSN/33 infected mMCP6-/- mMCP7+/+ mice at 10 days p.i demonstrate impaired IL-10 and IL-6 responses. Common inflammatory cytokines were determined by ELISA. IL-10 (A), IP-10 (B), IL-6 (C) Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

In summary following A/WSN/33 infection mMCP6-/- mMCP7+/+ mice showed similar weight loss and recovery profiles, showed an elevated peak viral titre, had reduced numbers of inflammatory cell in the BALf, yet had elevated histopathology scores and showed reduced levels of IL-10, IL-6 and IFNg. These data overall show mMCP7 playing a detrimental role in IAV infection by inducing excessive pathology.

235

4.4.4 Influenza A infection in Prss31-/- mice

To determine the role of mast cell specific tryptase Prss31 in the pathogenesis of

Influenza A Prss31-/- mice were infected intranasally with 750pfu of A/WSN/33. Prss31-/- infected mice followed a similar weight loss profile to that of their C57 infected controls up to day 6 p.i. The Prss31-/- infected mice lost less weight their C57 infected controls, a difference that reached significance on days 7, 9 & 10 p.i. (Figure 4.4.4.1).

Figure 4.4.4.1: Prss31-/- A/WSN/33 infected mice lose less weight infected C57 controls. Data are means ± SEM, two-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

236

At day 7 p.i Prss31-/- infected mice show an elevated viral peak in the lung when compared to C57 control mice (Figure 4.4.4.2).

Figure 4.4.4.2: Prss31-/- infected mice have a higher viral peak than infected C57 controls. BALf viral titers were determined at day 7 post infection by plaque assay on MDCK cells. Data are means ± SEM, plaque forming units are by Students t-test *P<0.05, **P<0.005, ***P<0.0005.

The elevated viral peak seen in Prss31-/- infected mice didn’t correspond to any additional increases in leukocyte counts from BALf. Differential analysis shows the inflammatory response to be comparable and comprised of neutrophils, macrophages and lymphocytes. With these increases being statistically significant for each cell type compared to uninfected controls (Figure 4.4.4.3 A-D).

237

Figure 4.4.4.3: A/WSN/33 infected Prss31-/- mice show comparable inflammatory cell infiltration attributed to macrophage, neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

238

Histopathological scoring of lung tissue at 7 days p.i shows the characteristic inflammation typical following A/WSN/33 infection with no discernible differences seen between the Prss31-/- infected mice and their C57 infected controls.

Figure 4.4.4.4: Histopathological scoring of the lungs from A/WSN/33 infected Prss31-/- mice at 7 days p.i show comparable scores to those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 A/WSN/33 infected (C), Prss31-/- uninfected (D), and Prss31-/-A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ****P<0.0001.

At 7 days p.i cytokine profiles from the BALf of Prss31-/- Influenza infected mice revealed reduced IL-10 production together with elevated IFNg production. All other cytokine responses are comparable to C57 infected control mice.

239

Figure 4.4.4.5: Cytokine profiling of A/WSN/33 infected Prss31-/- mice at 7 days p.i demonstrate impaired IL-10 and enhanced IFNg responses. Common inflammatory cytokines were determined by ELISA. TNAα (A), IL-10 (B), IP-10 (C), IFNg (D), IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

240

Both Prss31-/- and C57 mice cleared the A/WSN/33 infection by day 10 p.i, with no plaques detected by plaque assay (data not shown). Enumeration of cells from BALf shows that the overall inflammatory cell influx was greater at 10 days p.i than at 7 days p.i . There were reduced leukocyte counts in Prss31-/- infected mice at day 10 p.i when compared to their

C57 control group. Differential analysis revealed that this reduction was attributed to a reduction of macrophages, neutrophils and lymphocytes (Figure 4.4.4.6 A-D).

Figure 4.4.4.6: A/WSN/33 infected Prss31-/- mice show reduced inflammatory cell infiltration attributed to macrophage, neutrophil and lymphocyte at day 10 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one way ANOVA with Tukey’s multiple comparisons test, for both **P<0.05, ***P<0.005, ****P<0.0005.

241

Histopathological scoring of lung tissue at 10 days p.i shows the characteristic inflammation typical following A/WSN/33 infection with no discernable differences seen between the Prss31-/- infected mice and their C57 infected controls, an observation that is counterintuitive to the observations of reduced cells seen in the BALf. This again points towards defective cell migration to the airway lumen.

Figure 4.4.4.7: Histopathological scoring of the lungs from A/WSN/33 infected Prss31-/- mice at 10 days p.i show comparable scores to those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 A/WSN/33 infected (C), Prss31-/- uninfected (D), and Prss31-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

242

The observed reduction in concentrations of IL-10 seen at day 7 p.i in Prss31-/- infected mice was repeated in these mice at day 10 p.i whilst the elevated IFNg responses seen at day 7 p.i had resolved with no IFNg being detected in the BALf. Furthermore, no

TNFα or IL-1b were detected (data not shown). All other cytokine responses are comparable to C57 infected control mice.

Figure 4.4.4.8: Cytokine profiling of A/WSN/33 infected Prss31-/- mice at 10 days p.i demonstrate impaired IL-10 responses. Common inflammatory cytokines were determined by ELISA. IL-10 (A), IP-10 (B), IL-6 (C). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

243

In summary following A/WSN/33 infection Prss31-/- mice showed less weight loss, an elevated peak viral titre, had reduced numbers of inflammatory cell in the BALf, with comparable histopathology scores and showed reduced levels of IL-10 and increased IFNg.

These data overall show Prss31 is detrimental in the pathogenesis of IAV infection.

4.4.5 Influenza A infection in NDST2-/- mice

To determine the role of Glucosaminyl N-deacetylase/N-sulphotransferase-2

(NDST2), an enzyme essential in the N-deacetylation and N-sulphonation of heparin, in the pathogenesis of Influenza A NDST2-/- mice were infected intranasally with 750pfu of

A/WSN/33. NDST2-/- infected mice followed a similar weight loss profile to that of their C57 infected controls, albeit with slightly less weight loss, this difference however was not statistically significant (Figure 4.4.5.1).

Figure 4.4.5.1: NDST2-/- A/WSN/33 infected mice show similar weight loss to infected C57 controls. Data are means ± SEM, two-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

244

At day 7 p.i NDST2-/- infected mice show no changes in viral peak in the lung when compared to C57 control mice (Figure 4.4.5.2).

Figure 4.4.5.2: NDST2-/- have a viral peak equal to that of infected C57 controls. BALf viral titers were determined at day 7 post infection by plaque assay on MDCK cells. Data are means ± SEM, plaque forming units are by Students t-test.

Together with a comparable viral peak NDST2-/- infected mice show comparable leukocyte counts in BALf with differential analysis revealing the inflammatory response to be comprised of neutrophils and lymphocytes. With these increases being statistically significant for both cell types compared to uninfected controls. The typical increased macrophage inflammatory response at day 7 p.i was absent in this experiment (Figure 4.4.5.3

A-D).

245

Figure 4.4.5.3: A/WSN/33 infected NDST-/- mice show comparable inflammatory cell infiltration attributed to neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005

246

Histopathological scoring of lung tissue at 7 days p.i shows the characteristic inflammation typical following A/WSN/33 infection with no discernible differences seen between the NDST2-/- infected mice and their C57 infected controls.

Figure 4.4.5.4: Histopathological scoring of the lungs from A/WSN/33 infected NDST2-/- mice at 7 day p.i show comparable scores to those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 A/WSN/33 infected (C), NDST2-/- uninfected (D), and NDST2-/-A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

At 7 days p.i cytokine profiles from the BALf of Prss31-/- Influenza infected mice are comparable to C57 infected control mice.

247

Figure 4.4.5.5: Cytokine profiling of A/WSN/33 infected NDST2-/- mice at 7 days p.i demonstrate comparable responses to those seen in C57 infected controls. Common inflammatory cytokines were determined by ELISA. TNAa (A), IL-10 (B), IP-10 (C), IFNg (D), IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

248

Both NDST2-/- and C57 mice cleared the A/WSN/33 infection by day 10 p.i, with no plaques detected by plaque assay (data not shown). Enumeration of cells from BALf shows that the overall inflammatory cell influx was greater at 10 days p.i than at 7 days p.i . There were very elevated leukocyte counts in NDST2-/- infected mice at day 10 p.i when compared to their C57 control group, an observation that wasn’t seen at day 7 p.i. Differential analysis revealed that increase in inflammatory cells was attributed to macrophages, neutrophils and lymphocytes, each of these increase being significantly elevated in contrast to the inflammatory cell counts from C57 infected mice (Figure 4.4.5.6 A -D).

Figure 4.4.5.6: A/WSN/33 infected NDST2-/- mice show greatly elevated inflammatory cell response attributed to macrophage, neutrophil and lymphocyte infiltration at day 10 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

249

Histopathological scoring of lung tissue at 10 days p.i shows elevated pathology in the NDST2-/- A/WSN/33 infected mice compared to their C57 infected controls. This observation matches the increases seen in inflammatory cell counts from BALf in the group.

Figure 4.4.5.7: Histopathological scoring of the lungs from A/WSN/33 infected NDST2-/- mice at 10 day p.i show higher scores than those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 A/WSN/33 infected (C), NDST2-/- uninfected (D), and NDST2-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

Cytokine profiling at day 10 p.i unlike those at day 7 p.i revealed stark differences from those seen in the C57 infected control mice. In this model IFNg and TNFα concentrations in the BALf at day 10 p.i are typically below the threshold for detection.

However the infected NDST2-/- mice at day 10 p.i all had IFNg present, albeit at relatively low concentrations. The expression of this proinflammatory cytokine corresponds with the worsening inflammation seen in these mice. Additionally, IL-6 production in these mice is markedly reduced. Given the capacity of IL-6 to enhance neutrophil function together with

250

the elevated inflammatory cell influx seen in BALf, these observations suggest a somewhat impaired inflammatory response. Further more, no TNFα or IL-1β were detected (data not shown) and all other cytokine responses are comparable to C57 infected control mice.

Figure 4.4.5.8: Cytokine profiling of A/WSN/33 infected NDST2-/- mice at 10 days p.i demonstrate impaired IL-6 and elevated IFNg responses. Common inflammatory cytokines were determined by ELISA. IL-10 (A), IP-10 (B), IL-6 (C). Data are means ± SEM, one way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

In summary following A/WSN/33 infection NDST2-/- mice comparable weight loss profiles and peak viral titres, however there was significantly elevated numbers of inflammatory cells in the BALf, and higher histopathology scores at day 10 post infection.

251

Cytokine profiles showed reduced IL-6 and increased IFNg. These data overall demonstrate

NDST2 is protective during this infection.

4.4.6 Influenza A infection in mMCP5-/- mice

To determine the role of the mast cell chymases mMCP5 in the pathogenesis of

Influenza A mMCP5-/- mice were infected intranasally with 750pfu of A/WSN/33. mMCP5-/- infected mice followed a similar weight loss profile to that of their C57 infected controls, however the mMCP5 mice lost significantly more weight at days 8 and 9 p.i than their C57 infected controls (Figure 4.4.6.1).

Figure 4.4.6.1: mMCP5-/- A/WSN/33 infected mice lose more weight during IAV infection than infected C57 controls. Data are means ± SEM, two way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

252

At day 7 p.i mMCP5-/- infected mice show an elevated viral peak in the lung when compared to C57 control mice (Figure 4.4.6.2).

Figure 4.4.6.2: mMCP5-/- have a higher viral peak than infected C57 controls. BALf viral titers were determined at day 7 post infection by plaque assay on MDCK cells. Data are means ± SEM, plaque forming units are by Students t-test *P<0.05, **P<0.005.

Despite the elevated viral peak seen in mMCP5-/- infected mice no differences were seen in the scope of the leukocyte influx in counts from BALf. Differential analysis shows inflammatory responses to be comparable to C57 infected controls and comprised of neutrophils, macrophages and lymphocytes. With the increases in neutrophils and lymphocytes being statistically significant compared to uninfected controls (Figure 4.4.6.3 A-

D).

253

Figure 4.4.6.3: A/WSN/33 infected mMCP5-/- mice show comparable inflammatory cell infiltration attributed to macrophage, neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leukocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

254

Histopathological scoring of lung tissue at 7 days p.i shows the characteristic inflammation typical following A/WSN/33 infection with no discernible differences seen between the mMCP5-/- infected mice and their C57 infected controls.

Figure 4.4.6.4: Histopathological scoring of the lungs from A/WSN/33 infected mMCP5-/- mice at 7 days p.i show comparable scores to those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 A/WSN/33 infected (C), mMCP5-/- uninfected (D), and mMCP5-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01.

At 7 days p.i cytokine profiles from the BALf of mMCP5-/- Influenza infected mice are comparable to C57 infected control mice.

255

Figure 4.4.6.5: Cytokine profiling of A/WSN/33 infected mMCP5-/- mice at 7 days p.i demonstrate comparable cytokines responses to their C57 infected controls. Common inflammatory cytokines were determined by ELISA. TNFα (A), IL-10 (B), IP-10 (C), IFNg (D), IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

256

Both mMCP5-/- and C57 mice cleared the A/WSN/33 infection by day 10 p.i, with no plaques detected by plaque assay (data not shown). Enumeration of cells from BALf shows that the overall inflammatory cell influx was higher at 10 days p.i than at 7 days p.i .

Leukocyte counts in mMCP5-/- infected mice at day 10 p.i were similar to their C57 control group. Differential analysis revealed that the increased numbers of macrophages, neutrophils and lymphocytes, was again comparable to controls with these increases following infection being statistically significant to uninfected controls (Figure 4.4.6.6 A-D).

Figure 4.4.6.6: A/WSN/33 infected mMCP5-/- mice show comparable inflammatory cell infiltration at day 10 p.i to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

257

Histopathological scoring of lung tissue at 10 days p.i shows the characteristic inflammation typical following A/WSN/33 infection with no discernible differences seen between the mMCP5-/- infected mice and their C57 infected controls.

Figure 4.4.6.7: Histopathological scoring of the lungs from A/WSN/33 infected mMCP5-/- mice at 10 days p.i show higher scores than those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 A/WSN/33 infected (C), mMCP5-/- uninfected (D), and mMCP5-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, **P<0.01 and ***P<0.001.

Cytokine profiling at day 10 p.i unlike those at day 7 p.i revealed differences from those seen in the C57 infected control mice. mMCP5-/- infected mice had reduced IL-10 and elevated IL-6 concentrations in their BALf compared to C57 infected controls. Furthermore, no TNFα, IFNg or IL-1b were detected (data not shown) and all other cytokine responses are comparable to C57 infected control mice.

258

Figure 4.4.6.8: Cytokine profiling of A/WSN/33 infected mMCP5-/- mice at 10 days p.i demonstrate impaired IL-10 and slightly elevated IL-6 responses. Common inflammatory cytokines were determined by ELISA. IL-10 (A), IP-10 (B), IL-6 (C). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001.

In summary following A/WSN/33 infection mMCP5-/- mice showed more weight loss, and an elevated peak viral titre, the numbers of inflammatory cell in the BALf are comparable to controls, as is histopathology scores. Cytokine profiles show reduced IL-10 and elevated IL-6 at day 10, the other cytokines are comparable to controls. These data overall show mMCP5 plays no role in the pathogenesis of IAV infection.

259

4.4.7 Influenza A infection in Prss22-/- mice

To determine the role of mast cell related tryptase Prss22 in the pathogenesis of

Influenza A Prss22-/- mice were infected intranasally with 750pfu of A/WSN/33. Prss22-/- infected mice followed a very similar weight loss profile to that of their C57 infected controls. The uninfected controls also shows a similar weight gain profiles (Figure 4.4.7.1).

Figure 4.4.7.1: Prss22-/- A/WSN/33 infected mice show similar weight loss to infected C57 controls. Data are means ± SEM, two-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

260

At day 7 p.i Prss22-/- infected mice show an elevated viral peak in the lung when compared to C57 control mice (Figure 4.4.7.2).

Figure 4.4.7.2: Prss22 have a higher viral peak than infected C57 controls. BALf viral titers were determined at day 7 post infection by plaque assay on MDCK cells. Data are means ± SEM, plaque forming units are by Students t-test *P<0.05.

261

Despite the elevated viral peak seen in Prss22-/- infected mice no differences were seen in the scope of the leukocyte influx in counts from BALf. Differential analysis shows inflammatory responses to be comparable to C57 infected controls and comprised of macrophages and lymphocytes. With the increases in macrophages and lymphocytes being statistically significant when compared to uninfected controls (Figure 4.4.7.3 A-D).

Figure 4.4.7.3: A/WSN/33 infected Prss22-/- mice show comparable inflammatory cell infiltration attributed to macrophage and lymphocyte infiltration at day 7 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

262

Histopathological scoring of lung tissue at 7 days p.i shows the characteristic inflammation typical following A/WSN/33 infection with no differences seen between the

Prss22-/- infected mice and their C57 infected controls.

Figure 4.4.7.4: Histopathological scoring of the lungs from A/WSN/33 infected Prss22-/- mice at 7 days p.i show comparable scores to those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 A/WSN/33 infected (C), Prss22-/- uninfected (D), and Prss22-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

Cytokine profiles from the BALf of Prss22-/- Influenza infected mice at day 7 p.i revealed elevated IFNg and IL-6 concentrations. All other cytokines concentrations are comparable to C57 infected control mice.

263

Figure 4.4.7.5: Cytokine profiling of A/WSN/33 infected Prss22-/- mice at 7 days p.i elevated IFNg and IL- 6 responses. Common inflammatory cytokines were determined by ELISA. TNAa (A), IL-10 (B), IP-10 (C), IFNg (D), IL-6 (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

264

Both Prss22-/- and C57 mice cleared the A/WSN/33 infection by day 10 p.i, with no plaques detected by plaque assay (data not shown). Enumeration of cells from BALf shows that the overall inflammatory cell influx was higher at 10 days p.i than at 7 days p.i.

Leukocyte counts in Prss22-/- infected mice at day 10 p.i were generally lower than seen in their C57 control group. Differential analysis revealed that there were slight reductions in macrophages, neutrophils and lymphocytes, however these differences were not significant

(Figure 4.4.7.6 A-D).

Figure 4.4.7.6: A/WSN/33 infected Prss22-/- mice show reduced inflammatory cell infiltration in BALf at day 10 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

265

Figure 4.4.7.7: Histopathological scoring of the lungs from A/WSN/33 infected Prss22-/- mice at 10 days p.i show higher scores than those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 A/WSN/33 infected (C), Prss22-/- uninfected (D), and Prss22-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001 and ****P<0.0001.

Cytokine profiles from the BALf of Prss22-/- Influenza infected mice at day 10 p.i shows a reversal of the elevated IL-6 concentrations detected in samples from day 7 p.i. All other cytokines are comparable to C57 infected control mice. Furthermore, no TNFα, IFNg or

IL-1b was detected (data not shown).

266

Figure 4.4.7.8: Cytokine profiling of A/WSN/33 infected Prss22-/- mice at 10 days p.i demonstrate impaired IL-6 responses. Common inflammatory cytokines were determined by ELISA. IL-10 (A), IP-10 (B), IL-6 (C). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

In summary following A/WSN/33 infection Prss22-/- mice showed comparable weight loss, and an elevated peak viral titre, the numbers of inflammatory cell in the BALf are comparable to controls at day 7, as are histopathology scores. Day 10 BALf and histopathological scores show elevated inflammation and pathology. Cytokine profiles show elevated IFNg and IL-6 at day 7 with reduced IL-6 at day 10, the other cytokines are

267

comparable to controls. These data overall show Prss22 plays a beneficial role in the pathogenesis of IAV infection by promoting an anti-inflammatory state.

4.4.8 Influenza A infection in RasGRP4-/- mice

To determine the role of mast cell associated factor RasGRP4 in the pathogenesis of

Influenza A RasGRP4-/- mice were infected intranasally with 750pfu of A/WSN/33.

RasGRP4-/- infected mice followed a very similar weight loss profile to that of their C57 infected controls. The uninfected controls also shows a similar weight gain profiles (Figure

4.4.8.1).

Figure 4.4.8.1: RasGRP4-/- A/WSN/33 infected mice show similar weight loss to infected C57 controls. Data are means ± SEM, two-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

268

At day 7 p.i RasGRP4-/- infected mice show a reduced viral peak in the lung when compared to C57 control mice (Figure 4.4.8.2).

Figure 4.4.8.2: RasGRP4-/- have a lower viral peak than infected C57 controls. BALf viral titers were determined at day 7 post infection by plaque assay on MDCK cells. Data are means ± SEM, plaque forming units are by Students t-test **P<0.005.

Despite the elevated lower peak seen in RasGRP4-/- infected mice no differences were seen in the scope of the leukocyte influx in counts from BALf. Differential analysis shows inflammatory responses to be comparable to C57 infected controls and comprised of neutrophils and lymphocytes. With the increases in these cells being statistically significant when compared to uninfected controls. The typical macrophage inflammatory response at day

7 p.i was muted in this experiment (Figure 4.4.8.3 A-D).

269

Figure 4.4.8.3: A/WSN/33 infected RasGRP4-/- mice show comparable inflammatory cell infiltration attributed to neutrophil and lymphocyte infiltration at day 7 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B), neutrophils (C), and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

270

Figure 4.4.8.4: Histopathological scoring of the lungs from A/WSN/33 infected RasGRP4-/- mice at 7 days p.i show comparable scores to those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from C57 uninfected (B), C57 A/WSN/33 infected (C), RasGRP4-/- uninfected (D), and RasGRP4-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001 and ****P<0.0001.

At 7 days p.i cytokine profiles from the BALf of RasGRP4-/- Influenza infected mice show an elevated IL-6 response compared with C57 infected control mice. IFNg whilst typically seen at this time point was not detected, all other tested cytokines are comparable to

C57 infected control mice.

271

Figure 4.4.8.5: Cytokine profiling of A/WSN/33 infected RasGRP4-/- mice at 7 days p.i demonstrate an elevated IL-6 response. Common inflammatory cytokines were determined by ELISA. TNAa (A) IL-10 (B), IP-10 (C), IL-6 (D). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Both RasGRP4-/- and C57 mice cleared the A/WSN/33 infection by day 10 p.i, with no plaques detected by plaque assay (data not shown). Enumeration of cells from BALf shows that the overall inflammatory cell influx was higher at 10 days p.i than at 7 days p.i.

Leukocyte counts in RasGRP4-/- infected mice at day 10 p.i were higher than those seen in their C57 control group. Differential analysis revealed elevated numbers of macrophages and lymphocytes, with these being statistically significant compared to infected controls (Figure

4.4.8.6 A-D).

272

Figure 4.4.8.6: A/WSN/33 infected RasGRP4-/- mice show elevated inflammatory cell infiltration attributed to macrophages and lymphocytes at day 10 p.i when compared to their C57 controls. Airway inflammation was determined by BALf leucocyte counts (A) and differential counts of macrophages (B) neutrophils (C) and lymphocytes (D). Data are means ± SEM, cell counts are by one-way ANOVA with Tukey’s multiple comparisons test, for both *P<0.05, **P<0.05, ***P<0.005, ****P<0.0005.

Histopathological scoring of lung tissue at 10 days p.i shows elevated pathology scores in the RasGRP4 A/WSN/33 infected group when compared to their C57 infected control. This observation tallies the elevated inflammatory cell counts observed in BALf from this group.

273

Figure 4.4.8.7: Histopathological scoring of the lungs from A/WSN/33 infected RasGRP4-/- mice at 10 days p.i show higher scores than those seen in infected control C57 mice. Histopathological scoring was assessed in fixed, paraffin embedded, sectioned (4-6µm) lung tissues stained with hematoxylin and eosin (A). Representative micrographs representing lungs from, C57 uninfected (B), C57 A/WSN/33 infected (C), RasGRP4-/- uninfected (D), and RasGRP4-/- A/WSN/33 infected (E). Data are means ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, ***P<0.001 and ****P<0.0001.

Cytokine profiling at day 10 p.i unlike those at day 7 p.i revealed a reversal in the elevated IL-6 responses seen with the Day 10 p.i observations showing a total absence of

IL-6. Given IL-6s capacity to enhance neutrophil function together with the elevated inflammatory cell influx seen in BALf, this observation points to an impaired inflammatory response. Furthermore, no TNFα, IFNg or IL-1b were detected (data not shown) and all other cytokine responses are comparable to C57 infected control mice.

274

Figure 4.4.8.8: Cytokine profiling of A/WSN/33 infected RasGRP4-/- mice at 10 days p.i demonstrate impaired IL-6 response. Common inflammatory cytokines were determined by ELISA. IL-10 (A), IP-10 (B), IL-6 (C) Data are means ± SEM, one way ANOVA with Tukey’s multiple comparisons test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

In summary following A/WSN/33 infection RasGRP4-/- mice showed comparable weight loss, and lower peak viral titre, the numbers of inflammatory cell in the BALf are comparable to controls at day 7, as are histopathology scores. Day 10 BALf and histopathological scores show elevated inflammation and pathology. Cytokine profiles show

275

elevated IL-6 at day 7 with reduced IL-6 at day 10, the other cytokines are comparable to

controls. These findings demonstrate RasGRP4 mediating anti-inflammatory responses in

this infection.

4.4.9 Summary of results

Transgenic Time Viral Weight Airway Histopathology Cytokine Mouse point Titer Loss Inflammation Scores Profile Line Elevated IL-10 Comparable Comparable Comparable to & IL-6 & Day 7 Lower to controls to controls controls reduced IFNg responses mMCP6-/- Increased inflammation Elevate IL-10, Comparable dominated by Day 10 Greater Elevated scores IP-10 & IFNg to controls Macrophages and responses Lymphocytes. Reduced IL-10, Comparable Reduced overall Comparable to Day 7 Higher IFNg & IL-6 -/- to controls inflammation controls mMCP6 responses mMCP7+/+ Reduced macrophage Reduced IL-10 Comparable Comparable Day 10 numbers and lower overall Elevated scores & IL-6 to controls to controls inflammation responses Reduced IL-10 Comparable Comparable to Day 7 Higher Lower & increased to controls controls Prss31-/- IFNg responses Comparable Reduced overall Comparable to Reduced IL-10 Day 10 Lower to controls inflammation controls response Comparable Comparable Comparable Comparable to Comparable to Day 7 to controls to controls to controls controls controls Increased inflammation NDST2-/- Reduced IL-6 & Comparable Comparable comprising of Day 10 Elevated scores increased IFNg to controls to controls Macrophages, Neutrophils responses & Lymphocytes Comparable Comparable to Comparable to Day 7 Higher Greater to controls controls controls mMCP5-/- Comparable Comparable Comparable Comparable to Increased Il-6 Day 10 to controls to controls to controls controls response Increased IL-6 Comparable Comparable Comparable to Day 7 Higher & IFNg to controls to controls controls Prss22-/- responses Comparable Comparable Reduced overall Reduced IL-6 Day 10 Elevated scores to controls to controls inflammation response Comparable Comparable Comparable to Increased IL-6 Day 7 Lower to controls to controls controls response Increased inflammation RasGRP4 -/- Comparable Comparable comprising of Impaired IL-6 Day 10 Elevated scores to controls to controls Macrophages and response Lymphocytes

Table 4.4.9.1: Summary of Influenza infected mice deficient in mast cell proteases, associated proteases or mast cell associated factors at different points during the viral infection.

276

4.5 Discussion

This research builds on the work by Hu et al. and Graham et al. who demonstrated that mast cell degranulation leads to elevated immunopathology during IAV infection. Hu et al. showed that inhibiting mast cell degranulation protected mice from extensive immunopathology by reducing viral induced apoptosis and reduced the high mortality rates in a H5N1 infection133. A later study by Graham et al. demonstrated that mast cells are infected and directly activated by H1N1 A/WSN/33 triggering degranulation but without the production of virions. Additionally they also reinforced the findings of Hu et al., in that using kitW-sh mice they showed the absence of mast cells protected mice from excessive cytokine and chemokine production resulting in reduced lung pathology and vascular leakage134. These studies point to a mast cell mediator(s) released following degranulation that plays a negative role in pathogenesis of IAV infection. Utilising our unique knock out and knock in mouse lines I sought to determine the roles of the mast cell proteases mMCP5, mMCP6, mMCP7,

Prss31, their associated protease Prss22 and related factors NDST2 and RasGRP4 play during

IAV pathogenesis.

Using a H1N1 A/WSN/33 infection model that elicits immune responses with the hallmarks of a human IAV infection in transgenic mice, I show that mast cell proteases, their related proteases and associated factors play important roles in driving innate immune responses. Mast cells are sentinels of the immune system, strategically located at air interface within the lungs where they can act as the front line in the defence from invading pathogens and entry points for IAV. They can be directly infected resulting in activation and subsequent degranulation134 and play a key role in triggering and sculpting the inflammatory responses.

They influence innate immunity as well as delaying adaptive immunity during infections, and growing evidence demonstrates that mast cell derived proteases play import roles during infection.

277

This study shows that mice devoid of the β-tryptase mMCP6 have a similar clinical score to their C57 infected control mice and leukocyte enumeration in BALf at 7 day p.i. mMCP6-/- mice demonstrate equivalent inflammatory cell counts with differential counts showing no discernible differences between the two infected groups. Plaque assays revealed that the mMCP6-/- infected mice have a viral titre peak at day 7 p.i that was lower that their

C57 infected controls, yet both strains had cleared the virus by day 10 p.i. Cytokine profiling revealed a reduction in IL-6 and IFNg concentration and an increase in IL-10. Analysis from day 10 p.i shows elevated leukocyte enumeration in BALf, with differential analysis revealing increased numbers of macrophages and lymphocytes, this elevated cell influx coincided with higher histopathological scores seen in the mMCP6-/- infected groups.

Cytokine profiles concurred with the inflammatory cell influx and histopathological scores with elevated concentrations of the lymphocyte chemokine IP-10, together with elevated

IL-10 and IFNg observed in the mMCP6-/- infected mice.

Day 10 BALf counts revealed lower inflammation, approximately 50% lower, than typically expected for this model. Cytokine profiles are consistent with other studies using this model. As BALf for cytokine profiling was removed prior to differential count processing it is possible that human error during the differential count preparation has resulted in a this observed reduction. As such this specific differential count result should be considered with this in mind. Confidence in the other data points for this experiment remain high.

Collectively this data points to mMCP6 being beneficial during IAV infection by promoting tolerance. Whilst the presence of mMCP6 results in a higher viral titre, it seems to promote reduced inflammation and less associated pathology. mMCP6 is best known for its ability to induce neutrophilic inflammation195 and possess proinflammatory functions that are orchestrated via the activation of bystander cells , resulting in the production of chemokines

278

for neutrophils such as CXCL1, CXCL8, MMP-3, MMP-13143. Given that this anti- inflammatory role is not prevalent at 7 days p.i points to mMCP6 playing a new role in modulating excessive lymphocyte and macrophage inflammation into the lung during IAV infection by modulating IP-10 production. This would point to a novel role for mMCP6 in

Influenza infection.

The role of mMCP7 in the pathogenesis of IAV infection is mixed. The genetic addition of β-tryptase mMCP7 to mice resulted in an elevated viral peak and reduced inflammatory cell influx in the BALf at day 7 p.i, an observation that persisted to day 10 p.i, suggesting that mMCP7 promotes tolerance in IAV infection. This notion is supported by cytokine profiles at day 7 p.i, showing reduced IFNg and IL-6 with IL-6 remaining reduced compared to C57 infected controls at day 10 p.i. Conversely the elevated histopathological scores seen at day 10 p.i points to enhanced pathology caused by mMCP7. This increased histopathology scoring was contrary to the reduced inflammatory cell influx into the BALf.

This could point towards a defect in trans epithelial migration in mMCP7+/+ mice, an observation that was seen in the PA14 infection model described in chapter 3.

Whilst the mMCP6-/- mMCP7+/+ mice effectively clear the virus by day 10 p.i, like their C57 infected controls, collectively this data points to mMCP7 playing a detrimental role in IAV infection by inducing excessive pathology, possibly as a consequence of impaired inflammatory cell migration to the airway lumen. Additionally the absence of mMCP6, which has been shown to promote pathogen tolerance and mediate excessive leukocyte inflammation in this model could also be contributing towards the results observed.

The membrane anchored mast cell g-tryptase Prss31 is detrimental in the pathogenesis of IAV infection. Prss31-/- infected mice lose less weight during the course of the infection, demonstrate reduced inflammation in BALf at day 10 p.i, whilst having a higher viral peak at day 7 p.i, all virus is cleared by day 10 p.i. This suggests that the Prss31-/- infected mice can

279

tolerate IAV infection whilst mitigating excessive inflammatory responses. How Prss31 contributes to inflammation during this infection is unclear. However considering the data from the mMCP6-/- mMCP7+/+ infected mice, that are, due to their knock-in background being from 129Sv mice, also Prss31-/-, together with the Prss31-/- infected mouse data, both mice have reduced IL-10 concentrations in BALf at day 7 p.i. IL-10 can play a detrimental role in IAV infection with studies showing IL-10-/- mice are protected from lethal IAV doses242 thus it’s possible that Prss31, via an unknown mechanism, promotes IL-10 production that subsequently has a negative consequence for mice in this model.

The proinflammatory role attributed of Prss31 seen in this model agree with studies by Fricker et al. who showed Prss31-/- mice have reduced inflammation in a cigarette smoke induced COPD model and an experimental colitis model. Collectively these data suggest that a Prss31 specific inhibitor could be a beneficial therapeutic intervention in IAV infections and IAV infectious exacerbation of COPD. Inhibiting Prss31 could reduce inflammation during IAV infection, reducing the clinical manifestations of the disease in an otherwise health individual. COPD is a chronic inflammatory disease of the lungs, its progression is closely associated by exacerbations (environmental or infectious in nature), thus any treatment to mitigate inflammatory responses in already diseased lungs could be of great therapeutic benefit.

As mentioned in previous chapters, any heparin bound tryptase data needs to be considered in the context of the NDST2-/- infection data. NDST2-/- show no differences between C57 infected control mice at day 7 p.i however by day 10 p.i they have elevated inflammatory cell influx into the BALf, comprising of macrophages, neutrophils and lymphocytes. They also demonstrate higher histopathological scores. Given that NDST2-/- mast cells granules are depleted of heparin proteoglycans including mMCP6 the observation of elevated inflammatory cell influx into the BALf can be in part explained by an absence of

280

mMCP6. NDST2-/- mast cells do contain mMCP7, however unlike the mMCP6-/- mMCP7+/+ mice NDST2-/- mice do not have elevated peak viral counts suggesting that mMCP7 doesn’t contribute towards this observation in these mice. Perhaps it is the absence of Prss31 in mMCP6-/- mMCP7+/+ mice, given their 129Sv background that contributed to the observation of higher viral peak numbers. mMCP7s presence could help explain the elevated histopathological scores seen however the elevated inflammatory cells seen in the BALf casts doubt on notion of mMCP7 causing defective inflammatory cell migration. Yet given the extensive disruption to the composition of mast cell granules seen in NDST-/- mice, pinpointing specific functions of missing proteases is difficult and all data needs to be considered collectively.

In this IAV infection NDST2 is protective, an observation that is likely attributed to the ability of NDST2 competent mice to produce heparin and store mast cell associated proteases.

Mast cell chymase, mMCP5 plays no role in the pathogenesis of IAV. mMCP5-/- infected mice lose slightly more weight during the infection however by day 10 p.i their weight losses are once again comparable to C57 infected controls. They also have elevated peak viral counts but show no other differences from that of their C57 infected controls at day

7 p.i. At day 10 p.i they have cleared the virus, with histopathological scores and inflammatory cell influx into BALf remain comparable to C57 infected controls. There are some differences in the cytokine profile of mMCP5-/- infected mice at day 10 p.i with elevated IL-6 decreased IL-10 production. The lower IL-10 in this model appears to be benign. Collectively this data point to mMCP5 playing little to no role in the pathogenesis of

Influenza infection. Studies using mMCP5-/- mice also need to be mindful than mMCP5-/- mice do not express CPA3205, thus in this study mMCP5 and CPA3 appear to play no role in the pathogenesis of IAV.

281

Mice deficient in mast cell related and epithelial cell restricted e-tryptase Prss22 have higher viral peaks at day 7 p.i but unaltered weightloss, inflammatory cell counts in BALf, histopathological scoring and with the exception of elevated IL-6 and IFNg, have comparable cytokine production to C57 infected control mice at day 7 p.i. Interestingly at day 10 p.i these mice demonstrate very high histopathology scores yet reduced inflammatory cell recruitment detected in BALf pointing to Prss22 playing a beneficial role in the pathogenesis of IAV infection. How this protease protects mice from pathology is unclear, but given its expression by airway epithelial cells and this anti-inflammatory role, the prospect of this tryptase being artificially delivered to the airways to combat excessive inflammation is intriguing.

Whilst the data show the establishment of a productive Influenza infection in these mice, at day 7 BALf counts revealed elevated inflammation in the Prss22-/- and control mice, being 2-3 times higher than typically observed in this model. The day 10 data were consistent with other influenza models run in these mice. Why the day 7 data show elevated inflammation is unclear, however processing differences in the differential count preparation could have contributed to this change. As the BALf for cytokine profiling was removed prior to any processing, this could in part explain why cytokine profiles are comparable to other influenza models run and shown in this thesis.

At the start of this study RasGRP4 was believed to be exclusively expressed in mast cells, however more recent work has shown it to be present on splenic CD117+ dendritic cells. This new finding introduces new questions that the initial study was not planned to address. In my model RasGRP4-/- infected mice shows a reduced peak viral counts that like

C57 infected controls are cleared by day 10 p.i. They are comparable to C57 infected controls in weight loss, inflammatory cell counts from BALf, histopathological scoring and cytokine production, with the exception of an elevated IL-6 response. At day 10 p.i the RasGRP4-/- mice demonstrated elevated inflammatory cell influx in the BALf together with higher

282

histopathological scores and an absence of IL-6 response, demonstrating that RasGRP4 protects mice from IAV induced inflammation and pathology.

These findings of RasGRP4 mediating anti-inflammatory responses in an infection setting goes against other studies that show RasGRP4 as being proinflammatory in inflammatory conditions such as arthritis and colitis223.

In conclusion, I show that the tryptases mMCP6 and Prss22 are together with mast cell associated factors NDST2 and RasGRP4 are immunomodulatory in the pathogenesis of

IAV infection. All are beneficial by contributing towards reducing excessive inflammation towards the end of the acute phase of the infection during viral clearance (day7 p.i to day 10 p.i). Additionally, I show Prss31 to be proinflammatory during the viral clearance phase of this infection and demonstrated a mixed role for mMCP7 with its presence reducing inflammatory cell influx into the BALf at day 7 p.i yet contributing towards elevated histopathological scores at day 10 p.i. Finally, I show that mMCP5 appears to play a neutral role in the pathogenesis of IAV infection.

283

Chapter 5:

Discussions and conclusions 5.1 Significance of research

Respiratory infections are a leading cause of infection induced morbidity and mortality. This research points to new roles for mast cell proteases, their associated proteases and related factors in the pathogenesis of three respiratory infections S. pneumoniae,

P. aeruginosa and Influenza virus.

Mast cells are typically associated with allergy, consequently much research has focused on this area. Recently, the role these sentinel cells play in orchestrating the immune responses to infectious disease has been investigated revealing a role for these cells in various infections. Studies shown that mast cells are potent modulators of immune function, mediating the release of proinflammatory mediators, recruiting addition cells or aiding in direct killing of invading pathogens1.

Recently studies have been limited to using mast cell deficient mice and subsequently reconstituting these mice with bone marrow derived mast cells to demonstrate reported observations are indeed due to the absence of mast cells and not related to the other phenotypic characteristics mast cell deficient mice typically possess. These studies simply demonstrate the involvement of mast cells during an infection. Inhibiting mast cell degranulation is useful to determine if the observations are a result of mast cell degranulation, but again, the findings do not give any indication as to which of the many components of mast cells granules contribute to the observed results.

Utilising novel transgenic mice with mast cell proteases deleted or inserted, or deletions in mast cell related proteases, and mast cell associated factors, I have been able to build on the current literature and decipher the specific roles of these proteases and associated factors play in 3 common respiratory infections.

284

5.1.1 mMCP6 can modulate infection associated inflammation

These studies clearly establish mMCP6 is protective during P. aeruginosa infection in mice. My findings concur with findings reported by other groups3, demonstrating mMCP6s proinflammatory nature, and build on them by, for the first time, demonstrating a physiological role for mMCP6 in promoting inflammation and protection from subsequent bacteraemia during P. aeruginosa infection in mice.

Contrasting with the proinflammatory role of mMCP6 in P. aeruginosa infection, for the first time, I show that mMCP6, in the context of IAV infection, promotes an anti- inflammatory response during viral resolution, tipping the balance towards immune control and away from pathology. This discovery has significant implications for potential therapeutic interventions, especially in preventing IAV induced exacerbations in patients with pre-existing lung pathology, such as patients with cystic fibrosis, asthma or COPD.

The precise mechanism of how mMCP6 promotes immune control over pathology is unclear. My data shows that IP-10 responses are reduced. A this cytokine is made by many cell types present in the lung including, NK cells, monocytes, endothelial cells, macrophages,

T cells and dendritic cells identifying which cells contribute towards the observed reduction

IP-10 production would held elucidate the mechanism of how mMCP6 moderates inflammation in IAV infections.

Rerunning this model and investigating the resolution phase of this infection in more detail and by conducting a more extensive cytokine expression profile will provide a clearer picture of the extent of other changes in cytokines and chemokine expression. Additionally, identifying those cells responsible for the altered cytokine profile using FACS analysis will help determine a mechanism of mMCP6s moderation of the inflammatory response during

IAV resolution.

285

5.1.2 mMCP7 is deleterious during S. pneumoniae, P. aeruginosa and Influenza virus infections

These studies demonstrate for the first time that mMCP7 promotes pathology in both

P. aeruginosa and Influenza virus infections. I also show mMCP7 impairs bacterial clearance during S. pneumoniae infection. This data, collectively, contributes towards our current understanding of mast cell biology by demonstrating a physiological role for this tryptase in the context of three common respiratory pathogens and identifies mMCP7 as a potential therapeutic target to mediate inflammation during P. aeruginosa and Influenza virus infections.

The data in these studies show that mMCP7 expression results in reduced transepithelial migration in the lung during IAV and P. aeruginosa infection. With elevated inflammation seen in histopathological scoring without corresponding increases in BALf the quality of responses needs to be determined, which would include further investigation into the expression of adhesion molecules and chemokine expression.

5.1.3 Prss31 is detrimental during IAV infection

These studies clearly demonstrate that’s Prss31 is associated with IL-10 production during IAV infection. IL-10 has been shown to enhance mortality during IAV infection, with

IL-10-/- mice being protected from lethal IAV doses242. I show in Prss31-/- and mMCP6-/- mMCP7+/+ mice (also Prss31 nil) that IL-10 is supressed during IAV infection.

Other studies demonstrated Prss31-/- mice have reduced inflammation in noninfectious diseases such as in a cigarette smoke induced COPD model and an experimental colitis model202. Collectively, my data concurs and adds to the literature, supporting the proinflammatory properties of Prss31.

286

For the first time I clearly demonstrate the detrimental role of Prss31 in IAV infection and my data underlines the potential therapeutic benefits of inhibition of Prss31 during

Influenza infection. My data supports the hypothesis that Prss31 inhibition promotes viral tolerance over excessive inflammation, effectively reducing immunopathology. This finding is especially relevant to patients with pre-existing inflammatory lung conditions such as

COPD and Cystic fibrosis, where additional infection associated inflammation is particularly problematic.

5.1.4 Prss31 role in the pathogenesis of bacterial infection is pathogen specific

My Data demonstrates that Prss31 impairs bacterial clearance in a S. pneumoniae infection, a previously unknown role for Prss31. Conversely Prss31 appears to play no role in the pathogenesis of P. aeruginosa.

Collectively these data demonstrate for the first time that a membrane bound tryptase plays a role during respiratory infections, and that this role is pathogen specific. The data implies that therapeutic inhibition of this tryptase during specific respiratory infections, such as S. pneumoniae could yield clinically useful patient outcomes and offer a viable treatment option in COPD patients given that exacerbations are closely associated with poor progression of the disease243.

I have also demonstrated that Prss31 is associated with IL-10 production in my IAV models and this is broadly protective. The precise nature of this protection is yet to be determined however cytokine analysis shows Prss31-/- infected mice to have reduced levels of

IL-10. This immunoregulatory cytokine has been associated with enhanced mortality with

IL-10-/- mice showing protection for lethal doses of IAV. Investigating how Prss31 promotes

IL-10 expression will point towards a mechanism that could prove to be therapeutically useful.

287

5.1.5 The role of Prss22 in the pathogenesis of respiratory infection is pathogen specific

For the first time, I demonstrate that epithelial cell derived e-tryptase, Prss22, plays a key role in the pathogenesis of respiratory infections and that its role, like that of g-tryptase

Prss31, is pathogen specific.

I show that during P. aeruginosa respiratory infection Prss22-/- mice are protected from bacteraemia, indicating that inhibitors to Prss22 could provide a prophylactic treatment for sepsis.

Given that Prss22 activates uPA which subsequently activates plasminogen, triggering a cascade that promotes the degradation of the extracellular matrix and fibrinolysis211, delivering aerosolised inhibitors of Prss22 into the lung during a P. aeruginosa respiratory infection would elucidate the usefulness of inhibiting the Prss22 mediated break down of clots during P. aeruginosa infection. This would determine the efficacy of Prss22 inhibitors in preventing bacterial escape in to the blood and sequent bacteraemia.

I also demonstrate a previously unknown role for Prss22 during IAV infection. I show that Prss22 modulates inflammation during the post viral resolution phase of IAV infection, resulting in reduced pathology. The novel role in this infection context also points to potential therapeutic uses.

Demonstrating reduced histopathology and inflammation by providing aerosolised

Prss22 during a IAV infection would provide a proof of principle treatment option for IAV infected patents. Reducing IAV related inflammation and pathology in patients with pre- existing respiratory disease such as COPD and cystic fibrosis would be of significant therapeutic benefit and would slow underlining disease progression.

288

5.1.6 RasGRP4 is protective in the later time points of P. aeruginosa and IAV infections

Finally, I demonstrate that RasGRP4 plays an immunomodulatory role during IAV infection and in the later time points of P. aeruginosa infection acting to moderate inflammation and pathology. These findings are the first to ever demonstrate a role for

RasGRP4 mediating anti-inflammatory responses in an infection setting. The findings seem to challenge the current understanding that RasGRP4 is proinflammatory in inflammatory conditions such as arthritis and colitis223.

To further elucidate the role of RasGRP4 in these infections the use of a RasGRP4 inhibitor such as Galectin-3, as used by Shalom-Feuerstein et al.240 could prove informative and provide a proof of principle experiment that points towards potential new therapeutic intervention for IAV and P. aeruginosa infection.

5.1.7 Further knock out mouse generation using CRISPER/Cas9 technology

Many of the mice used in this experiment took many years to develop and cost

$100,000s. With the relatively recent development of CRISPER/Cas9 technology, the cost of generating a knock out mouse has dropped considerably. This new and cheap technique opens up the possibility of generating new knock out mice to further explore the role of mast cell proteases.

The generation of a mMCP6+/+ mMCP7+/+ mouse would complement the studies in this thesis by permitting the investigation of the hetertetrametic mMCP6/7 protease in the pathogenesis of S. pneumoniae, P. aeruginosa and IAV infection. These observations together with the those found from mMCP6-/- mMCP7+/+, mMCP6-/- mMCP7+/+ and NDST2-/- mice studies would improve our knowledge of the roles of these tryptases.

289

5.2 Publications

5.2.1 Accepted publications

Emma L. Beckett, B Biomed Sci, Richard L. Stevens, Andrew G. Jarnicki, Richard Y.

Kim, Irwan Hanish, Nicole G. Hansbro, Andrew Deane, Simon Keely, Jay C. Horvat, Ming

Yang, Brian G. Oliver, Nico van Rooijen, Mark D. Inman, Roberto Adachi, Roy J.

Soberman, Sahar Hamadi, Peter A. Wark, Paul S. Foster, Philip M. Hansbro, “A new short- term mouse model of chronic obstructive pulmonary disease identifies a role for mast cell tryptase in pathogenesis” Journal of Allergy and Clinical Immunology

MJ Gold, PR Hiebert, HY Park, D Stefanowicz, A Le, MR Starkey, A Deane, AC

Brown, G Liu, JC Horvat, ZA Ibrahim, MB Sukkar, PM Hansbro, C Carlsten, S VanEeden,

DD Sin, KM McNagny, DA Knight and JA Hirota “Mucosal production of uric acid by airway epithelial cells contributes to particulate matter-induced allergic sensitization”

Mucosal immunology

Michael Fricker, Andrew Deane & Philip M Hansbro “Animal models of chronic obstructive pulmonary disease” Expert Opinion in Drug Discovery

290

References

1. Johnzon, C.-F., Rönnberg, E. & Pejler, G. The Role of Mast Cells in Bacterial

Infection. AJPA 186, 4–14 (2016).

2. Stevens, R. L. & Adachi, R. Protease-proteoglycan complexes of mouse and human

mast cells and importance of their beta-tryptase-heparin complexes in inflammation

and innate immunity. Immunological Reviews 217, 155–167 (2007).

3. Thakurdas, S. M. et al. The mast cell-restricted tryptase mMCP-6 has a critical

immunoprotective role in bacterial infections. Journal of Biological Chemistry 282,

20809–20815 (2007).

4. Medzhitov, R. & Janeway, C. A. Decoding the patterns of self and nonself by the

innate immune system. Science 296, 298–300 (2002).

5. Opitz, B., van Laak, V., Eitel, J. & Suttorp, N. Innate immune recognition in

infectious and noninfectious diseases of the lung. Am J Respir Crit Care Med 181,

1294–1309 (2010).

6. Lamkanfi, M. & Dixit, V. M. Inflammasomes: guardians of cytosolic sanctity.

Immunological Reviews 227, 95–105 (2009).

7. Finlay, B. B. & McFadden, G. Anti-immunology: evasion of the host immune system

by bacterial and viral pathogens. Cell 124, 767–782 (2006).

8. Bianchi, M. E. DAMPs, PAMPs and alarmins: all we need to know about danger. J.

Leukoc. Biol. 81, 1–5 (2007).

9. Gardella, S. et al. The nuclear protein HMGB1 is secreted by monocytes via a non-

classical, vesicle-mediated secretory pathway. EMBO Rep. 3, 995–1001 (2002).

10. Bonaldi, T. et al. Monocytic cells hyperacetylate chromatin protein HMGB1 to

redirect it towards secretion. EMBO J. 22, 5551–5560 (2003).

11. Semino, C. et al. The maturation potential of NK cell clones toward autologous

291

dendritic cells correlates with HMGB1 secretion. J. Leukoc. Biol. 81, 92–99 (2007).

12. Porto, A. et al. Smooth muscle cells in human atherosclerotic plaques secrete and

proliferate in response to high mobility group box 1 protein. FASEB J. 20, 2565–

2566 (2006).

13. Lotze, M. T. & Tracey, K. J. High-mobility group box 1 protein (HMGB1): nuclear

weapon in the immune arsenal. Nat. Rev. Immunol. 5, 331–342 (2005).

14. Köhl, J. Self, non-self, and danger: a complementary view. Adv. Exp. Med. Biol. 586,

71–94 (2006).

15. Bohlson, S. S., O'Conner, S. D., Hulsebus, H. J., Ho, M.-M. & Fraser, D. A.

Complement, c1q, and c1q-related molecules regulate macrophage polarization.

Front. Immunol. 5, 402 (2014).

16. Zhou, W. The new face of anaphylatoxins in immune regulation. Immunobiology

217, 225–234 (2012).

17. Walport, M. J. Complement. First of two parts. N Engl J Med 344, 1058–1066

(2001).

18. Markiewski, M. M. & Lambris, J. D. The role of complement in inflammatory

diseases from behind the scenes into the spotlight. AJPA 171, 715–727 (2007).

19. Underhill, D. M. & Ozinsky, A. Phagocytosis of microbes: complexity in action.

Annu. Rev. Immunol. 20, 825–852 (2002).

20. Fischer, M. B. et al. Regulation of the B cell response to T-dependent antigens by

classical pathway complement. The Journal of Immunology 157, 549–556 (1996).

21. Maldonado-López, R. & Moser, M. Dendritic cell subsets and the regulation of

Th1/Th2 responses. Semin. Immunol. 13, 275–282 (2001).

22. Kapsenberg, M. L. Dendritic-cell control of pathogen-driven T-cell polarization. Nat.

Rev. Immunol. 3, 984–993 (2003).

292

23. Vyas, J. M., Van der Veen, A. G. & Ploegh, H. L. The known unknowns of antigen

processing and presentation. Nature Publishing Group 8, 607–618 (2008).

24. Cascio, P., Hilton, C., Kisselev, A. F., Rock, K. L. & Goldberg, A. L. 26S

proteasomes and immunoproteasomes produce mainly N-extended versions of an

antigenic peptide. EMBO J. 20, 2357–2366 (2001).

25. Whitby, F. G. et al. Structural basis for the activation of 20S proteasomes by 11S

regulators. Nature 408, 115–120 (2000).

26. Ackerman, A. L. & Cresswell, P. Cellular mechanisms governing cross-presentation

of exogenous antigens. Nat. Immunol. 5, 678–684 (2004).

27. Kopf, M., Schneider, C. & Nobs, S. P. The development and function of lung-

resident macrophages and dendritic cells. Nat. Immunol. 16, 36–44 (2014).

28. Lau, G. W., Hassett, D. J. & Britigan, B. E. Modulation of lung epithelial functions

by Pseudomonas aeruginosa. Trends in Microbiology 13, 389–397 (2005).

29. Whitsett, J. A. Intrinsic and innate defenses in the lung: intersection of pathways

regulating lung morphogenesis, host defense, and repair. J. Clin. Invest. 109, 565–

569 (2002).

30. Martin, T. R. & Frevert, C. W. Innate immunity in the lungs. Proc Am Thorac Soc 2,

403–411 (2005).

31. Lai, D.-M., Shu, Q. & Fan, J. The origin and role of innate lymphoid cells in the

lung. Mil Med Res 3, 25 (2016).

32. Lamb, D. & Lumsden, A. Intra-epithelial mast cells in human airway epithelium:

evidence for smoking-induced changes in their frequency. Thorax 37, 334–342

(1982).

33. O'Brien, K. L. et al. Burden of disease caused by Streptococcus pneumoniae in

children younger than 5 years: global estimates. The Lancet 374, 893–902 (2009).

293

34. Kadioglu, A., Weiser, J. N., Paton, J. C. & Andrew, P. W. The role of Streptococcus

pneumoniae virulence factors in host respiratory colonization and disease. Nat Rev

Micro 6, 288–301 (2008).

35. World Health Organization. WHO Weekly Epidemiological Record. 87, 129–144

(world health organization, 2012).

36. Health, A. G. D. O. Pneumococcal disease.

37. McCool, T. L. & Weiser, J. N. Limited role of antibody in clearance of Streptococcus

pneumoniae in a murine model of colonization. Infect. Immun. 72, 5807–5813

(2004).

38. Gwaltney, J. M., Sande, M. A., Austrian, R. & Hendley, J. O. Spread of

Streptococcus pneumoniae in Families. II. Relation of Transfer of S. pneumoniae to

Incidence of Colds and Serum Antibody. J. Infect. Dis. 132, 62–68 (1975).

39. Nelson, A. L. et al. Capsule Enhances Pneumococcal Colonization by Limiting

Mucus-Mediated Clearance. Infect. Immun. 75, 83–90 (2006).

40. Weiser, J. N., Austrian, R., Sreenivasan, P. K. & Masure, H. R. Phase variation in

pneumococcal opacity: relationship between colonial morphology and

nasopharyngeal colonization. Infect. Immun. 62, 2582–2589 (1994).

41. Bergmann, S. & Hammerschmidt, S. Versatility of pneumococcal surface proteins.

Microbiology (Reading, Engl.) 152, 295–303 (2006).

42. Hammerschmidt, S., Talay, S. R., Brandtzaeg, P. & Chhatwal, G. S. SpsA, a novel

pneumococcal surface protein with specific binding to secretory immunoglobulin A

and secretory component. Mol. Microbiol. 25, 1113–1124 (1997).

43. Rosenow, C. et al. Contribution of novel choline-binding proteins to adherence,

colonization and immunogenicity of Streptococcus pneumoniae. Mol. Microbiol. 25,

819–829 (1997).

294

44. Andersson, B. et al. Identification of an active disaccharide unit of a glycoconjugate

receptor for pneumococci attaching to human pharyngeal epithelial cells. J. Exp.

Med. 158, 559–570 (1983).

45. Jedrzejas, M. J., Mello, L. V., de Groot, B. L. & Li, S. Mechanism of hyaluronan

degradation by Streptococcus pneumoniae hyaluronate lyase. Structures of

complexes with the substrate. Journal of Biological Chemistry 277, 28287–28297

(2002).

46. Holmes, A. R. et al. The pavA gene of Streptococcus pneumoniae encodes a

fibronectin-binding protein that is essential for virulence. Mol. Microbiol. 41, 1395–

1408 (2001).

47. Bergmann, S., Rohde, M., Chhatwal, G. S. & Hammerschmidt, S. alpha-Enolase of

Streptococcus pneumoniae is a plasmin(ogen)-binding protein displayed on the

bacterial cell surface. Mol. Microbiol. 40, 1273–1287 (2001).

48. Cruse, G. et al. Human lung mast cells mediate pneumococcal cell death in response

to activation by pneumolysin. J. Immunol. 184, 7108–7115 (2010).

49. Ratner, A. J. et al. Epithelial cells are sensitive detectors of bacterial pore-forming

toxins. Journal of Biological Chemistry 281, 12994–12998 (2006).

50. Barbuti, G., Moschioni, M., Fumarulo, R., Censini, S. & Montemurro, P.

Streptococcus pneumoniae modulates the respiratory burst response in human

neutrophils. FEMS Immunology & Medical Microbiology 60, 57–62 (2010).

51. van Rossum, A. M. C., Lysenko, E. S. & Weiser, J. N. Host and bacterial factors

contributing to the clearance of colonization by Streptococcus pneumoniae in a

murine model. Infect. Immun. 73, 7718–7726 (2005).

52. Schröder, N. W. J. et al. Lipoteichoic acid (LTA) of Streptococcus pneumoniae and

Staphylococcus aureus activates immune cells via Toll-like receptor (TLR)-2,

295

lipopolysaccharide-binding protein (LBP), and CD14, whereas TLR-4 and MD-2 are

not involved. Journal of Biological Chemistry 278, 15587–15594 (2003).

53. Yoshimura, A. et al. Cutting edge: recognition of Gram-positive bacterial cell wall

components by the innate immune system occurs via Toll-like receptor 2. The

Journal of Immunology 163, 1–5 (1999).

54. Malley, R. et al. Recognition of pneumolysin by Toll-like receptor 4 confers

resistance to pneumococcal infection. PNAS 100, 1966–1971 (2003).

55. Albiger, B. et al. Toll-like receptor 9 acts at an early stage in host defence against

pneumococcal infection. Cellular Microbiology 9, 633–644 (2007).

56. Hsu, L.-C. et al. A NOD2-NALP1 complex mediates caspase-1-dependent IL-1beta

secretion in response to Bacillus anthracis infection and muramyl dipeptide. Proc.

Natl. Acad. Sci. U.S.A. 105, 7803–7808 (2008).

57. Koppe, U., Suttorp, N. & Opitz, B. Recognition of Streptococcus pneumoniae by the

innate immune system. Cellular Microbiology 14, 460–466 (2012).

58. Köckritz-Blickwede, von, M. et al. Phagocytosis-independent antimicrobial activity

of mast cells by means of extracellular trap formation. Blood 111, 3070–3080 (2008).

59. van den Boogaard, F. E. et al. Mast cells impair host defense during murine

Streptococcus pneumoniae pneumonia. J. Infect. Dis. 210, 1376–1384 (2014).

60. Barbuti, G. et al. Streptococcus pneumoniae induces mast cell degranulation.

International Journal of Medical Microbiology 296, 325–329 (2006).

61. Hyams, C., Camberlein, E., Cohen, J. M., Bax, K. & Brown, J. S. The Streptococcus

pneumoniae capsule inhibits complement activity and neutrophil phagocytosis by

multiple mechanisms. Infect. Immun. 78, 704–715 (2010).

62. Szalai, A. J., Briles, D. E. & Volanakis, J. E. Human C-reactive protein is protective

against fatal Streptococcus pneumoniae infection in transgenic mice. The Journal of

296

Immunology 155, 2557–2563 (1995).

63. Brown, J. S. et al. The classical pathway is the dominant complement pathway

required for innate immunity to Streptococcus pneumoniae infection in mice. PNAS

99, 16969–16974 (2002).

64. Beiter, K. et al. An endonuclease allows Streptococcus pneumoniae to escape from

neutrophil extracellular traps. Curr. Biol. 16, 401–407 (2006).

65. Cheng, Q., Finkel, D. & Hostetter, M. K. Novel purification scheme and functions

for a C3-binding protein from Streptococcus pneumoniae. Biochemistry 39, 5450–

5457 (2000).

66. Blom, A. M., Bergmann, S., Fulde, M., Riesbeck, K. & Agarwal, V. Streptococcus

pneumoniae phosphoglycerate kinase is a novel complement inhibitor affecting the

membrane attack complex formation. J. Biol. Chem. 289, 32499–32511 (2014).

67. McCool, T. L., Cate, T. R., Moy, G. & Weiser, J. N. The immune response to

pneumococcal proteins during experimental human carriage. J. Exp. Med. 195, 359–

365 (2002).

68. Reynolds, C. A., Finkelstein, J. A., Ray, G. T., Moore, M. R. & Huang, S. S.

Attributable healthcare utilization and cost of pneumonia due to drug-resistant

streptococcus pneumonia: a cost analysis. Antimicrob Resist Infect Control 3, 16

(2014).

69. Black, S. et al. Efficacy, safety and immunogenicity of heptavalent pneumococcal

conjugate vaccine in children. Northern California Kaiser Permanente Vaccine Study

Center Group. Pediatr. Infect. Dis. J. 19, 187–195 (2000).

70. Isaacman, D. J., McIntosh, E. D. & Reinert, R. R. Burden of invasive pneumococcal

disease and serotype distribution among Streptococcus pneumoniae isolates in young

children in Europe: impact of the 7-valent pneumococcal conjugate vaccine and

297

considerations for future conjugate vaccines. International Journal of Infectious

Diseases 14, e197–e209 (2010).

71. Roca, A. et al. Invasive pneumococcal disease in children<5 years of age in rural

Mozambique. Trop. Med. Int. Health 11, 1422–1431 (2006).

72. Appelbaum, P. C. Resistance among Streptococcus pneumoniae: Implications for

drug selection. Clin. Infect. Dis. 34, 1613–1620 (2002).

73. Leinonen, M. et al. Antibody response to 14-valent pneumococcal capsular

polysaccharide vaccine in pre-school age children. Pediatr Infect Dis 5, 39–44

(1986).

74. Kaplan, S. L. et al. Early trends for invasive pneumococcal infections in children

after the introduction of the 13-valent pneumococcal conjugate vaccine. Pediatr.

Infect. Dis. J. 32, 203–207 (2013).

75. Bonten, M. J. M. et al. Polysaccharide Conjugate Vaccine against Pneumococcal

Pneumonia in Adults. N Engl J Med 372, 1114–1125 (2015).

76. Musher, D. M., Sampath, R. & Rodriguez-Barradas, M. C. The potential role for

protein-conjugate pneumococcal vaccine in adults: what is the supporting evidence?

Clin. Infect. Dis. 52, 633–640 (2011).

77. Tam, V. H. et al. Prevalence, resistance mechanisms, and susceptibility of multidrug-

resistant bloodstream isolates of Pseudomonas aeruginosa. Antimicrobial Agents and

Chemotherapy 54, 1160–1164 (2010).

78. Richards, M. J., Edwards, J. R., Culver, D. H. & Gaynes, R. P. Nosocomial

infections in combined medical-surgical intensive care units in the United States.

Infect Control Hosp Epidemiol 21, 510–515 (2000).

79. Klockgether, J. Pseudomonas aeruginosa genomic structure and diversity. 1–18

(2011). doi:10.3389/fmicb.2011.00150/abstract.

298

80. Grimwood, K., Kyd, J. M., Owen, S. J., Massa, H. M. & Cripps, A. W. Vaccination

against respiratory Pseudomonas aeruginosainfection. Human Vaccines &

Immunotherapeutics 11, 14–20 (2014).

81. Williams, B. J., Dehnbostel, J. & Blackwell, T. S. Pseudomonas aeruginosa: Host

defence in lung diseases. Respirology 15, 1037–1056 (2010).

82. Gaspar, M. C., Couet, W., Olivier, J. C., Pais, A. A. C. C. & Sousa, J. J. S.

Pseudomonas aeruginosa infection in cystic fibrosis lung disease and new

perspectives of treatment: a review. Eur J Clin Microbiol Infect Dis 32, 1231–1252

(2013).

83. Almagro, P. et al. Pseudomonas aeruginosa and Mortality after Hospital Admission

for Chronic Obstructive Pulmonary Disease. Respiration 84, 36–43 (2012).

84. Loebinger, M. R. et al. Mortality in bronchiectasis: a long-term study assessing the

factors influencing survival. European Respiratory Journal 34, 843–849 (2009).

85. Obritsch, M. D., Fish, D. N., MacLaren, R. & Jung, R. National surveillance of

antimicrobial resistance in Pseudomonas aeruginosa isolates obtained from intensive

care unit patients from 1993 to 2002. Antimicrobial Agents and Chemotherapy 48,

4606–4610 (2004).

86. Lee, D. G. et al. Genomic analysis reveals that Pseudomonas aeruginosa virulence is

combinatorial. Genome Biol. 7, R90 (2006).

87. Feldman, M. et al. Role of flagella in pathogenesis of Pseudomonas aeruginosa

pulmonary infection. Infect. Immun. 66, 43–51 (1998).

88. Pizarro-Cerdá, J. & Cossart, P. Bacterial adhesion and entry into host cells. Cell 124,

715–727 (2006).

89. Kazmierczak, B. I., Mostov, K. & Engel, J. N. Epithelial cell polarity alters Rho-

GTPase responses to Pseudomonas aeruginosa. Mol. Biol. Cell 15, 411–419 (2004).

299

90. Wozniak, D. J. et al. Alginate is not a significant component of the extracellular

polysaccharide matrix of PA14 and PAO1 Pseudomonas aeruginosa biofilms. PNAS

100, 7907–7912 (2003).

91. Smith, R. S. et al. IL-8 Production in Human Lung Fibroblasts and Epithelial Cells

Activated by the Pseudomonas Autoinducer N-3-Oxododecanoyl Homoserine

Lactone Is Transcriptionally Regulated by NF- B and Activator Protein-2. The

Journal of Immunology 167, 366–374 (2001).

92. Lavoie, E. G., Wangdi, T. & Kazmierczak, B. I. Innate immune responses to

Pseudomonas aeruginosa infection. Microbes and Infection 13, 1133–1145 (2011).

93. Mijares, L. A. et al. Airway epithelial MyD88 restores control of Pseudomonas

aeruginosa murine infection via an IL-1-dependent pathway. J. Immunol. 186, 7080–

7088 (2011).

94. Junkins, R. D. et al. Mast Cells Protect against Pseudomonas aeruginosae Induced

Lung Injury. AJPA 184, 2310–2321 (2014).

95. Höpken, U. E., Lu, B., Gerard, N. P. & Gerard, C. The C5a chemoattractant receptor

mediates mucosal defence to infection. Nature 383, 86–89 (1996).

96. Sun, H.-Y., Fujitani, S., Quintiliani, R. & Yu, V. L. Pneumonia due to Pseudomonas

aeruginosa: part II: antimicrobial resistance, pharmacodynamic concepts, and

antibiotic therapy. CHEST 139, 1172–1185 (2011).

97. Driscoll, J. A., Brody, S. L. & Kollef, M. H. The epidemiology, pathogenesis and

treatment of Pseudomonas aeruginosa infections. Drugs (2007).

doi:10.1111/liv.13031

98. Shi, Y., Wu, Y., Zhang, W., Qi, J. & Gao, G. F. Enabling the ‘host jump’: structural

determinants of receptor-binding specificity in influenza A viruses. Nature

Publishing Group 12, 822–831 (2014).

300

99. Pulendran, B. & Maddur, M. S. in Influenza Pathogenesis and Control - Volume II

(eds. Oldstone, M. B. A. & Compans, R. W.) 386, 23–71 (Springer International

Publishing, 2014).

100. Valkenburg, S. A. et al. Immunity to seasonal and pandemic influenza A viruses.

Microbes Infect. 13, 489–501 (2011).

101. Molinari, N.-A. M. et al. The annual impact of seasonal influenza in the US:

measuring disease burden and costs. Vaccine 25, 5086–5096 (2007).

102. Taubenberger, J. K. & Morens, D. M. The pathology of influenza virus infections.

Annu Rev Pathol 3, 499–522 (2008).

103. Karlsson, E. A., Marcelin, G., Webby, R. J. & Schultz-Cherry, S. Review on the

impact of pregnancy and obesity on influenza virus infection. Influenza Other Respir

Viruses 6, 449–460 (2012).

104. Drake, J. W. Rates of spontaneous mutation among RNA viruses. PNAS 90, 4171–

4175 (1993).

105. World Health Organization. Weekly epidemiological record. 87, 461–476 (2012).

106. Gamblin, S. J. & Skehel, J. J. Influenza hemagglutinin and neuraminidase membrane

glycoproteins. J. Biol. Chem. 285, 28403–28409 (2010).

107. Baum, L. G. & Paulson, J. C. Sialyloligosaccharides of the respiratory epithelium in

the selection of human influenza virus receptor specificity. Acta Histochem. Suppl.

40, 35–38 (1990).

108. Couceiro, J. N., Paulson, J. C. & Baum, L. G. Influenza virus strains selectively

recognize sialyloligosaccharides on human respiratory epithelium; the role of the

host cell in selection of hemagglutinin receptor specificity. Virus Res. 29, 155–165

(1993).

109. Karlsson Hedestam, G. B. et al. The challenges of eliciting neutralizing antibodies to

301

HIV-1 and to influenza virus. Nat Rev Micro 6, 143–155 (2008).

110. Gazit, R. et al. Lethal influenza infection in the absence of the natural killer cell

receptor gene Ncr1. Nat. Immunol. 7, 517–523 (2006).

111. Hashimoto, Y., Moki, T., Takizawa, T., Shiratsuchi, A. & Nakanishi, Y. Evidence

for phagocytosis of influenza virus-infected, apoptotic cells by neutrophils and

macrophages in mice. The Journal of Immunology 178, 2448–2457 (2007).

112. Jegaskanda, S., Reading, P. C. & Kent, S. J. Influenza-specific antibody-dependent

cellular cytotoxicity: toward a universal influenza vaccine. J. Immunol. 193, 469–475

(2014).

113. Braciale, T. J., Sun, J. & Kim, T. S. Regulating the adaptive immune response to

respiratory virus infection. Nature Publishing Group 12, 295–305 (2012).

114. Iwasaki, A. & Pillai, P. S. Innate immunity to influenza virus infection. Nat. Rev.

Immunol. 14, 315–328 (2014).

115. Ichinohe, T., Pang, I. K. & Iwasaki, A. Influenza virus activates inflammasomes via

its intracellular M2 ion channel. Nat. Immunol. 11, 404–410 (2010).

116. Kreijtz, J. H. C. M., Fouchier, R. A. M. & Rimmelzwaan, G. F. Immune responses to

influenza virus infection. Virus Res. 162, 19–30 (2011).

117. Sato, M. et al. Positive feedback regulation of type I IFN genes by the IFN-inducible

transcription factor IRF-7. FEBS Lett. 441, 106–110 (1998).

118. Randall, R. E. & Goodbourn, S. Interferons and viruses: an interplay between

induction, signalling, antiviral responses and virus countermeasures. Journal of

General Virology 89, 1–47 (2008).

119. Hale, B. G., Albrecht, R. A. & García-Sastre, A. Innate immune evasion strategies of

influenza viruses. Future Microbiology 5, 23–41 (2010).

120. Sharma, K. et al. Influenza A virus nucleoprotein exploits Hsp40 to inhibit PKR

302

activation. PLoS ONE 6, e20215 (2011).

121. Guan, Z. et al. Interaction of Hsp40 with influenza virus M2 protein: implications for

PKR signaling pathway. Protein Cell 1, 944–955 (2010).

122. Pica, N. & Palese, P. Toward a universal influenza virus vaccine: prospects and

challenges. Annu. Rev. Med. 64, 189–202 (2013).

123. Sasaki, S. et al. Comparison of the influenza virus-specific effector and memory B-

cell responses to immunization of children and adults with live attenuated or

inactivated influenza virus vaccines. Journal of Virology 81, 215–228 (2007).

124. Castilla, J. et al. Decline in influenza vaccine effectiveness with time after

vaccination, Navarre, Spain, season 2011/12. Euro Surveill. 18, (2013).

125. Stiver, G. The treatment of influenza with antiviral drugs. CMAJ 168, 49–56 (2003).

126. Sridhar, S., Brokstad, K. A. & Cox, R. J. Influenza Vaccination Strategies:

Comparing Inactivated and Live Attenuated Influenza Vaccines. Vaccines (Basel) 3,

373–389 (2015).

127. Webster, R. G. & Govorkova, E. A. Continuing challenges in influenza. Ann. N. Y.

Acad. Sci. 1323, 115–139 (2014).

128. Jefferson, T. et al. Neuraminidase inhibitors for preventing and treating influenza in

healthy adults and children. Cochrane Database Syst Rev 19, CD008965 (2014).

129. Oh, D. Y. & Hurt, A. C. A Review of the Antiviral Susceptibility of Human and

Avian Influenza Viruses over the Last Decade. Scientifica 2014, 430629–10 (2014).

130. Medzhitov, R., Schneider, D. S. & Soares, M. P. Disease tolerance as a defense

strategy. Science 335, 936–941 (2012).

131. Cabrero, A., Laguna, J. C. & Vázquez, M. Peroxisome proliferator-activated

receptors and the control of inflammation. Curr Drug Targets Inflamm Allergy 1,

243–248 (2002).

303

132. Budd, A. et al. Increased survival after gemfibrozil treatment of severe mouse

influenza. Antimicrobial Agents and Chemotherapy 51, 2965–2968 (2007).

133. Hu, Y. et al. Mast Cell-Induced Lung Injury in Mice Infected with H5N1 Influenza

Virus. Journal of Virology 86, 3347–3356 (2012).

134. Graham, A. C., Hilmer, K. M., Zickovich, J. M. & Obar, J. J. Inflammatory response

of mast cells during influenza A virus infection is mediated by active infection and

RIG-I signaling. J. Immunol. 190, 4676–4684 (2013).

135. Mast cell progenitors: origin, development and migration to tissues. 63, 9–17 (2015).

136. Anderson, D. M. et al. Alternate splicing of mRNAs encoding human mast cell

growth factor and localization of the gene to chromosome 12q22-q24. Cell Growth

Differ. 2, 373–378 (1991).

137. Amin, K. The role of mast cells in allergic inflammation. Respir Med 106, 9–14

(2012).

138. Thakurdas, S. M. et al. The mast cell-restricted tryptase mMCP-6 has a critical

immunoprotective role in bacterial infections. Journal of Biological Chemistry 282,

20809–20815 (2007).

139. Féger, F., Varadaradjalou, S., Gao, Z., Abraham, S. N. & Arock, M. The role of mast

cells in host defense and their subversion by bacterial pathogens. Trends Immunol.

23, 151–158 (2002).

140. Malaviya, R., Ikeda, T., Ross, E. & Abraham, S. N. Mast cell modulation of

neutrophil influx and bacterial clearance at sites of infection through TNF-alpha.

Nature 381, 77–80 (1996).

141. Di Nardo, A., Vitiello, A. & Gallo, R. L. Cutting edge: mast cell antimicrobial

activity is mediated by expression of cathelicidin antimicrobial peptide. The Journal

of Immunology 170, 2274–2278 (2003).

304

142. McNeil, H. P. et al. Isolation, characterization, and transcription of the gene

encoding mouse mast cell protease 7. PNAS 89, 11174–11178 (1992).

143. Hamilton, M. J., Sinnamon, M. J. & Lyng, G. D. Essential role for mast cell tryptase

in acute experimental colitis. (Proceedings of the …, 2011).

doi:10.1073/pnas.1005758108/-/DCSupplemental

144. Miller, H. R. P. & Pemberton, A. D. Tissue-specific expression of mast cell granule

serine proteinases and their role in inflammation in the lung and gut. Immunology

105, 375–390 (2002).

145. Mortaz, E., Folkerts, G. & Redegeld, F. Mast cells and COPD. Pulm Pharmacol Ther

24, 367–372 (2011).

146. Rivera, J., Fierro, N. A., Olivera, A. & Suzuki, R. New insights on mast cell

activation via the high affinity receptor for IgE. Adv. Immunol. 98, 85–120 (2008).

147. Kitamura, Y., Go, S. & Hatanaka, K. Decrease of mast cells in W/Wv mice and their

increase by bone marrow transplantation. Blood 52, 447–452 (1978).

148. Galli, S. J. & Kitamura, Y. Genetically mast-cell-deficient W/Wv and Sl/Sld mice.

Their value for the analysis of the roles of mast cells in biologic responses in vivo.

Am. J. Pathol. (1987).

149. Kaplan, J., De Domenico, I. & Ward, D. M. Chediak-Higashi syndrome. Curr. Opin.

Hematol. 15, 22–29 (2008).

150. Grimbaldeston, M. A. et al. Mast cell-deficient W-sash c-kit mutant Kit W-sh/W-sh

mice as a model for investigating mast cell biology in vivo. AJPA 167, 835–848

(2005).

151. Echtenacher, B., Männel, D. N. & Hültner, L. Critical protective role of mast cells in

a model of acute septic peritonitis. Nature 381, 75–77 (1996).

152. Wei, O. L., Hilliard, A., Kalman, D. & Sherman, M. Mast cells limit systemic

305

bacterial dissemination but not colitis in response to Citrobacter rodentium. Infect.

Immun. 73, 1978–1985 (2005).

153. Siebenhaar, F. et al. Control of Pseudomonas aeruginosa skin infections in mice is

mast cell-dependent. AJPA 170, 1910–1916 (2007).

154. Velin, D., Bachmann, D., Bouzourene, H. & Michetti, P. Mast cells are critical

mediators of vaccine-induced Helicobacter clearance in the mouse model.

Gastroenterology 129, 142–155 (2005).

155. Ketavarapu, J. M. et al. Mast cells inhibit intramacrophage Francisella tularensis

replication via contact and secreted products including IL-4. Proc. Natl. Acad. Sci.

U.S.A. 105, 9313–9318 (2008).

156. Edelson, B. T., Li, Z., Pappan, L. K. & Zutter, M. M. Mast cell-mediated

inflammatory responses require the alpha 2 beta 1 integrin. Blood 103, 2214–2220

(2004).

157. Matsui, H. et al. Dermal mast cells reduce progressive tissue necrosis caused by

subcutaneous infection with Streptococcus pyogenes in mice. J. Med. Microbiol. 60,

128–134 (2011).

158. Duttlinger, R. et al. The Wsh and Ph mutations affect the c-kit expression profile: c-

kit misexpression in embryogenesis impairs melanogenesis in Wsh and Ph mutant

mice. PNAS 92, 3754–3758 (1995).

159. Nagle, D. L., Kozak, C. A., Mano, H., Chapman, V. M. & Bućan, M. Physical

mapping of the Tec and Gabrb1 loci reveals that the Wsh mutation on mouse

chromosome 5 is associated with an inversion. Hum. Mol. Genet. 4, 2073–2079

(1995).

160. Wolters, P. J. et al. Tissue-selective mast cell reconstitution and differential lung

gene expression in mast cell-deficient Kit(W-sh)/Kit(W-sh) sash mice. Clin. Exp.

306

Allergy 35, 82–88 (2005).

161. Piliponsky, A. M. et al. Mast cell-derived TNF can exacerbate mortality during

severe bacterial infections in C57-KitW-sh/W-sh mice. Am. J. Pathol. 176, 926–938

(2010).

162. Xu, X. et al. Mast cells protect mice from Mycoplasma pneumonia. Am J Respir Crit

Care Med 173, 219–225 (2006).

163. Shelburne, C. P. et al. Mast cells augment adaptive immunity by orchestrating

dendritic cell trafficking through infected tissues. Cell Host Microbe 6, 331–342

(2009).

164. Dahdah, A. et al. Mast cells aggravate sepsis by inhibiting peritoneal macrophage

phagocytosis. J. Clin. Invest. 124, 4577–4589 (2014).

165. Chan, C. Y., St John, A. L. & Abraham, S. N. Mast cell interleukin-10 drives

localized tolerance in chronic bladder infection. Immunity 38, 349–359 (2013).

166. Metz, M. et al. Mast cells determine the magnitude of bacterial toxin-induced skin

inflammation. Exp. Dermatol. 18, 160–166 (2009).

167. Feng, B.-S., He, S.-H., Zheng, P.-Y., Wu, L. & Yang, P.-C. Mast cells play a crucial

role in Staphylococcus aureus peptidoglycan-induced diarrhea. AJPA 171, 537–547

(2007).

168. Nakamura, Y. et al. Staphylococcus δ-toxin induces allergic skin disease by

activating mast cells. Nature 503, 397–401 (2013).

169. Abraham, S. N. & St John, A. L. Mast cell-orchestrated immunity to pathogens.

Nature Publishing Group 10, 440–452 (2010).

170. Josset, L. et al. Implication of inflammatory macrophages, nuclear receptors, and

interferon regulatory factors in increased virulence of pandemic 2009 H1N1

influenza A virus after host adaptation. Journal of Virology 86, 7192–7206 (2012).

307

171. Graham, A. C., Temple, R. M. & Obar, J. J. Mast Cells and Influenza A Virus:

Association with Allergic Responses and Beyond. Front. Immunol. 6, 499–12

(2015).

172. St John, A. L., Rathore, A. P. S., Raghavan, B., Ng, M.-L. & Abraham, S. N.

Contributions of mast cells and vasoactive products, leukotrienes and chymase, to

dengue virus-induced vascular leakage. Elife 2, e00481 (2013).

173. Sun, Q. et al. Evidence for a role of mast cells in the mucosal injury induced by

Newcastle disease virus. Poult. Sci. 88, 554–561 (2009).

174. Sorden, S. D. & Castleman, W. L. Virus-induced increases in bronchiolar mast cells

in Brown Norway rats are associated with both local mast cell proliferation and

increases in blood mast cell precursors. Lab. Invest. 73, 197–204 (1995).

175. Aoki, R. et al. Mast cells play a key role in host defense against herpes simplex virus

infection through TNF-α and IL-6 production. J. Invest. Dermatol. 133, 2170–2179

(2013).

176. Varadaradjalou, S. et al. Toll-like receptor 2 (TLR2) and TLR4 differentially activate

human mast cells. Eur. J. Immunol. 33, 899–906 (2003).

177. McCurdy, J. D., Lin, T. J. & Marshall, J. S. Toll-like receptor 4-mediated activation

of murine mast cells. J. Leukoc. Biol. 70, 977–984 (2001).

178. Boudreau, R. T. M., Garduno, R. & Lin, T.-J. Protein phosphatase 2A and protein

kinase Calpha are physically associated and are involved in Pseudomonas

aeruginosa-induced interleukin 6 production by mast cells. Journal of Biological

Chemistry 277, 5322–5329 (2002).

179. Junkins, R. D., Shen, A., Rosen, K., McCormick, C. & Lin, T.-J. Autophagy

Enhances Bacterial Clearance during P. aeruginosa Lung Infection. PLoS ONE 8,

e72263–13 (2013).

308

180. Douaiher, J. et al. Development of mast cells and importance of their tryptase and

chymase serine proteases in inflammation and wound healing. Adv. Immunol. 122,

211–252 (2014).

181. Fiorucci, L. & Ascoli, F. Mast cell tryptase, a still enigmatic enzyme. Cell. Mol. Life

Sci. 61, 1278–1295 (2004).

182. McNeil, H. P. et al. The mouse mast cell-restricted tetramer-forming tryptases mouse

mast cell protease 6 and mouse mast cell protease 7 are critical mediators in

inflammatory arthritis. Arthritis Rheum 58, 2338–2346 (2008).

183. Raithel, M. et al. Release of mast cell tryptase from human colorectal mucosa in

inflammatory bowel disease. Scand. J. Gastroenterol. 36, 174–179 (2001).

184. Zhang, X. et al. Tryptase enzyme activity is correlated with severity of chronic

obstructive pulmonary disease. Tohoku J. Exp. Med. 224, 179–187 (2011).

185. Mortaz, E., Givi, M. E., Da Silva, C. A., Folkerts, G. & Redegeld, F. A. A relation

between TGF-β and mast cell tryptase in experimental emphysema models. BBA -

Molecular Basis of Disease 1–7 (2012). doi:10.1016/j.bbadis.2012.03.006

186. Beckett, E. L. et al. A new short-term mouse model of chronic obstructive

pulmonary disease identifies a role for mast cell tryptase in pathogenesis. J. Allergy

Clin. Immunol. 131, 752–762 (2013).

187. Soto, D., Malmsten, C., Blount, J. L., Muilenburg, D. J. & Caughey, G. H. Genetic

deficiency of human mast cell a-tryptase. Clinical & Experimental Allergy 32,

1000–1006 (2002).

188. Caughey, G. H. Mast cell tryptases and chymases in inflammation and host defense.

Immunological Reviews 217, 141–154 (2007).

189. Huang, R. & Hellman, L. Genes for mast-cell serine protease and their molecular

evolution. Immunogenetics 40, 397–414 (1994).

309

190. Ghildyal, N. et al. Fate of two mast cell tryptases in V3 mastocytosis and normal

BALB/c mice undergoing passive systemic anaphylaxis: prolonged retention of

exocytosed mMCP-6 in connective tissues, and rapid accumulation of enzymatically

active mMCP-7 in the blood. J. Exp. Med. 184, 1061–1073 (1996).

191. Pereira, P. J. et al. Human beta-tryptase is a ring-like tetramer with active sites facing

a central pore. Nature 392, 306–311 (1998).

192. Huang, C. et al. Formation of enzymatically active, homotypic, and heterotypic

tetramers of mouse mast cell tryptases. Dependence on a conserved Trp-rich domain

on the surface. Journal of Biological Chemistry 275, 351–358 (2000).

193. Hunt, J. E. et al. Natural disruption of the mouse mast cell protease 7 gene in the C57

mouse. Journal of Biological Chemistry 271, 2851–2855 (1996).

194. Zheng, D. & Stevens, R. L. Chapter 593 Mouse Mast Cell Tetramer-forming

Tryptases. Handbook of Proteolytic Enzymes 2698–2705 (Elsevier Ltd, 2012).

doi:10.1016/B978-0-12-382219-2.00593-7

195. Huang, C. et al. Induction of a selective and persistent extravasation of neutrophils

into the peritoneal cavity by tryptase mouse mast cell protease 6. The Journal of

Immunology 160, 1910–1919 (1998).

196. Wong, G. W. et al. Identification of a new member of the tryptase family of mouse

and human mast cell proteases which possesses a novel COOH-terminal hydrophobic

extension. Journal of Biological Chemistry 274, 30784–30793 (1999).

197. Huang, C. et al. The tryptase, mouse mast cell protease 7, exhibits anticoagulant

activity in vivo and in vitro due to its ability to degrade fibrinogen in the presence of

the diverse array of protease inhibitors in plasma. Journal of Biological Chemistry

272, 31885–31893 (1997).

198. Razin, E., Stevens, R. L., Akiyama, F., Schmid, K. & Austen, K. F. Culture from

310

mouse bone marrow of a subclass of mast cells possessing a distinct chondroitin

sulfate proteoglycan with glycosaminoglycans rich in N-acetylgalactosamine-4,6-

disulfate. Journal of Biological Chemistry 257, 7229–7236 (1982).

199. Levi-Schaffer, F., Austen, K. F., Gravallese, P. M. & Stevens, R. L. Coculture of

interleukin 3-dependent mouse mast cells with fibroblasts results in a phenotypic

change of the mast cells. PNAS 83, 6485–6488 (1986).

200. Trivedi, N. N. & Caughey, G. H. Chapter 592 γ-Tryptase. Handbook of Proteolytic

Enzymes 2694–2697 (Elsevier Ltd, 2012). doi:10.1016/B978-0-12-382219-2.00592-5

201. Yuan, J. et al. Expression and characterization of recombinant gamma-tryptase.

Protein Expr. Purif. 49, 47–54 (2006).

202. Hansbro, P. M. et al. Importance of mast cell Prss31/transmembrane

tryptase/tryptase-γ in lung function and experimental chronic obstructive pulmonary

disease and colitis. J. Biol. Chem. 289, 18214–18227 (2014).

203. Caughey, G. H. Chapter 590 Chymases. Handbook of Proteolytic Enzymes 2675–

2683 (Elsevier Ltd, 2012). doi:10.1016/B978-0-12-382219-2.00590-1

204. Wernersson, S. & Pejler, G. Mast cell secretory granules: armed for battle. Nature

Publishing Group 1–17 (2014). doi:10.1038/nri3690

205. Galli, S. J. et al. in 126, 45–127 (Elsevier, 2015).

206. He, S. & Walls, A. F. Human mast cell chymase induces the accumulation of

neutrophils, eosinophils and other inflammatory cells in vivo. Br. J. Pharmacol. 125,

1491–1500 (1998).

207. Younan, G. et al. The inflammatory response after an epidermal burn depends on the

activities of mouse mast cell proteases 4 and 5. J. Immunol. 185, 7681–7690 (2010).

208. Abonia, J. P. et al. Mast cell protease 5 mediates ischemia-reperfusion injury of

mouse skeletal muscle. The Journal of Immunology 174, 7285–7291 (2005).

311

209. Heusel, J. W., Hanson, R. D., Silverman, G. A. & Ley, T. J. Structure and expression

of a cluster of human hematopoietic serine protease genes found on chromosome

14q11.2. Journal of Biological Chemistry 266, 6152–6158 (1991).

210. Wong, G. W. et al. Human Tryptase (PRSS22), a New Member of the Chromosome

16p13.3 Family of Human Serine Proteases Expressed in Airway Epithelial Cells.

Journal of Biological Chemistry 276, 49169–49182 (2001).

211. Yasuda, S. et al. Urokinase-type plasminogen activator is a preferred substrate of the

human epithelium serine protease tryptase epsilon/PRSS22. Blood 105, 3893–3901

(2005).

212. Kjellén, L. & Lindahl, U. Proteoglycans: structures and interactions. Annu. Rev.

Biochem. 60, 443–475 (1991).

213. Humphries, D. E. et al. Heparin is essential for the storage of specific granule

proteases in mast cells. Nature 400, 769–772 (1999).

214. Forsberg, E. et al. Abnormal mast cells in mice deficient in a heparin-synthesizing

enzyme. Nature 400, 773–776 (1999).

215. Fal, A. M., Kraus-Filarska, M., Miecielica, J. & Małolepszy, J. [Mechanisms of

action of nebulized low molecular weights heparin in patients with bronchial

asthma]. Pol. Merkur. Lekarski 15, 543–545 (2003).

216. Ahmed, T., Garrigo, J. & Danta, I. Preventing bronchoconstriction in exercise-

induced asthma with inhaled heparin. N Engl J Med 329, 90–95 (1993).

217. Panés, J. et al. Comparison of heparin and steroids in the treatment of moderate and

severe ulcerative colitis. Gastroenterology 119, 903–908 (2000).

218. Ang, Y. S. et al. Randomized comparison of unfractionated heparin with

corticosteroids in severe active inflammatory bowel disease. Aliment. Pharmacol.

Ther. 14, 1015–1022 (2000).

312

219. Oduah, E., Linhardt, R. & Sharfstein, S. Heparin: Past, Present, and Future.

Pharmaceuticals 9, 38–12 (2016).

220. García, B., Fernández-Vega, I. & García-Suárez, O. The role of heparan sulfate

proteoglycans in bacterial infections. J Med Microb … (2014).

221. Shafti-Keramat, S. et al. Different heparan sulfate proteoglycans serve as cellular

receptors for human papillomaviruses. Journal of Virology 77, 13125–13135 (2003).

222. Metcalfe, D. D., Baram, D. & Mekori, Y. A. Mast cells. Physiol. Rev. 77, 1033–1079

(1997).

223. Adachi, R. et al. Ras guanine nucleotide-releasing protein-4 (RasGRP4) involvement

in experimental arthritis and colitis. J. Biol. Chem. 287, 20047–20055 (2012).

224. Yang, Y. RasGRP4, a New Mast Cell-restricted Ras Guanine Nucleotide-releasing

Protein with Calcium- and Diacylglycerol-binding Motifs. Identification of defective

variants of this signaling protein in Asthma, Mastocytosis and mast cell leukemia

patients and demonstration of the importance of RasGRP4 in mast cell development

and function. Journal of Biological Chemistry 277, 25756–25774 (2002).

225. Zhou, S. et al. CD117+ Dendritic and Mast Cells Are Dependent on RasGRP4 to

Function as Accessory Cells for Optimal Natural Killer Cell-Mediated Responses to

Lipopolysaccharide. PLoS ONE 11, e0151638–18 (2016).

226. The Collaborative Study on the Genetics of Asthma (CSGA). A genome-wide search

for asthma susceptibility loci in ethnically diverse populations. The Collaborative

Study on the Genetics of Asthma (CSGA). Nat. Genet. 15, 389–392 (1997).

227. Shin, K. et al. Mouse mast cell tryptase mMCP-6 is a critical link between adaptive

and innate immunity in the chronic phase of Trichinella spiralis infection. The

Journal of Immunology 180, 4885–4891 (2008).

228. Kenneth, N. S. et al. An inactivating caspase 11 passenger mutation originating from

313

the 129 murine strain in mice targeted for c-IAP1. Biochem. J. 443, 355–359 (2012).

229. Kayagaki, N. et al. Non-canonical inflammasome activation targets caspase-11.

Nature 479, 117–121 (2011).

230. Horvat, J. C. et al. Neonatal chlamydial infection induces mixed T-cell responses that

drive allergic airway disease. Am J Respir Crit Care Med 176, 556–564 (2007).

231. Mullen, P. G., Windsor, A. C., Walsh, C. J., Fowler, A. A. & Sugerman, H. J. Tumor

necrosis factor-alpha and interleukin-6 selectively regulate neutrophil function in

vitro. J. Surg. Res. 58, 124–130 (1995).

232. Clement, C. G. et al. Stimulation of Lung Innate Immunity Protects against Lethal

Pneumococcal Pneumonia in Mice. Am J Respir Crit Care Med 177, 1322–1330

(2008).

233. Marshall, J. S. Mast-cell responses to pathogens. Nat. Rev. Immunol. 4, 787–799

(2004).

234. Pang, L., Nie, M., Corbett, L., Sutcliffe, A. & Knox, A. J. Mast cell beta-tryptase

selectively cleaves eotaxin and RANTES and abrogates their eosinophil chemotactic

activities. The Journal of Immunology 176, 3788–3795 (2006).

235. Liu, Z.-Q. et al. Mast cell-derived serine proteinase regulates T helper 2 polarization.

Sci. Rep. 4, (2014).

236. Kafka, D. et al. Contribution of IL-1 to resistance to Streptococcus pneumoniae

infection. International Immunology 20, 1139–1146 (2008).

237. Descoteaux, G. A. D. A. A. Macrophage cytokines: involvement in immunity and

infectious diseases. 1–12 (2014). doi:10.3389/fimmu.2014.00491/abstract

238. Ming, W. J., Bersani, L. & Mantovani, A. Tumor necrosis factor is chemotactic for

monocytes and polymorphonuclear leukocytes. The Journal of Immunology 138,

1469–1474 (1987).

314

239. Zhang, Y., Ramos, B. F. & Jakschik, B. A. Neutrophil recruitment by tumor necrosis

factor from mast cells in immune complex peritonitis. Science 258, 1957–1959

(1992).

240. Shalom-Feuerstein, R., Levy, R., Makovski, V., Raz, A. & Kloog, Y. Galectin-3

regulates RasGRP4-mediated activation of N-Ras and H-Ras. Biochimica et

Biophysica Acta (BBA) - Molecular Cell Research 1783, 985–993 (2008).

241. Wilson, K. R., Napper, J. M. & Denvir, J. Defect in early lung defence against

Pseudomonas aeruginosa in DBA/2 mice is associated with acute inflammatory lung

injury and reduced bactericidal activity in …. … 153, 968–979 (2007).

242. McKinstry, K. K. et al. IL-10 deficiency unleashes an influenza-specific Th17

response and enhances survival against high-dose challenge. J. Immunol. 182, 7353–

7363 (2009).

243. Bauer, C. M. T. et al. Treating viral exacerbations of chronic obstructive pulmonary

disease: insights from a mouse model of cigarette smoke and H1N1 influenza

infection. PLoS ONE 5, e13251 (2010).

315

Appendix

In addition to the research and publications referenced within the thesis body, the following paper is a co-first authored paper submitted for review to PLOS one on 11th June

2016.

Gene expression signature of cigarette smoke-induced lung damage in a mouse model correlates with human chronic obstructive pulmonary disease.

Authors: Andrew Deane Andrew Jarnicki Richard Y. Kim Gang Liu Alan Hsu Peter Wark Joshua D. Campbell Richard L. Stevens Philip M. Hansbro

Corresponding author: Name: Professor Philip M. Hansbro Department: Immunology/Microbiology HMRI Building University of Newcastle Callaghan, NSW 2308 e-mail: [email protected] phone no.: +61 (02) 4042 0187 Fax no.: +61 (02) 4042 0024

316

Affiliations:

AJ, AD, RYK, GL: Centre for Asthma and Respiratory Disease, School of

Biomedical Sciences and Pharmacy, Faculty of Health, University of Newcastle and Hunter

Medical Research Institute, Callaghan, Australia.

AC-YH: Centre for Asthma and Respiratory Disease, The University of Newcastle,

Newcastle, New South Wales, Australia

PW: Centre for Asthma and Respiratory Disease, The University of Newcastle,

Newcastle, New South Wales, Australia, Department of Respiratory and Sleep Medicine,

John Hunter Hospital, Newcastle, New South Wales, Australia

JDC: Divison of Computational Biomedicine, Department of Medicine, Boston

University School of Medicine, 72 East Concord Street, Boston, MA 02118, USA.

Bioinformatics program, Boston University, 44 Cummington Street, Boston, MA 02215,

USA.

RLS: Department of Medicine, Division of Rheumatology, Immunology, and

Allergy, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts,

USA.

MeSH terms: Cigarette smoke: Chronic obstructive pulmonary disease, emphysema,

Gene expression profiling.

317

Abstract

Background: Chronic obstructive pulmonary disease represents a major disease which currently is untreatable. Identification Preclinical models that reflect key aspects of cigarette smoke-induced COPD represent potential opportunities to identify molecular components involved in disease development and progression as well as key networks involved.

Methods: Gene array analysis was performed on the lungs of two strains cigarette smoke-exposed mice before and after key pathological features were evident. Differential gene expression profiles were determined by setting expression thresholds at a fold change

≥2.0 with a p-value ≤0.05. Unsupervised hierarchical clustering was used to determine expression profiles amongst samples for experimental time points. Also, gene expression profiles were determined on human lung samples and correlated to the severity of emphysema. The mouse and human gene expression profiles were compared to determine the level of molecular compatability. Also, gene expression patterns were used to determine key pathways that are involved in COPD and potentially warrant intervention.

Results: In comparison to air controls, the lungs of cigarette smoke treated mice had significant alteration of 61 specific genes. Comparison to a human array dataset which identified a gene expression profile of genes strongly associated with emphysema severity identified 49 genes in the human cohort that overlap with the mouse cigarettes smoke induced signature, representing 79% of the significantly regulated genes identified in the mouse model. The majority of the significantly altered transcripts are associated with the dysregulation of the immune system, with sustained inflammation with this signature predominantly indicated sustained changes in the inflammatory response. Network and pathway associations were identified to determine the major biological functions involving these genes. These included the alteration of leukocyte activation, as well as their adhesion

318

and subsequent migration into and out of the lung. In addition, pathways involving ROS production, phagocytosis, control of bacterial infection, increased matrix metalloproteinase production and apoptosis were identified.

Conclusions: Array analysis of the lungs from mice with cigarette smoke-induced

COPD identified a gene expression signature that identifies genes involved in pathogenesis. that signature that Cigarette smoking This mouse model therefore is strongly representative of the molecular aspects of COPD pathogenesis and provides a good basis in the identification of genes and pathways with the potential to identify genes in the early diagnosis of COPD and well as providing targets for both management and treatment of COPD.

319

Introduction

Chronic obstructive pulmonary disease (COPD) is a heterogeneous disease with complex genomic and environmental interactions. It is an inflammatory lung syndrome, characterized by limitation in expiratory airflow that deteriorates over a period of years [1].

The underlying causes of this reduction in lung function can include a loss in lung elastic recoil, damage to alveoli, fibrosis, oedema and smooth muscle contraction [2]. Typical clinical symptoms include chronic bronchitis and emphysema which contribute to a decreased quality of life, brought about by a shortness of breath, that, over a period of years, progresses towards either chronic hypoxemic or hypercarbic respiratory failure [3]. COPD affects approximately 64 million people worldwide [4] and is projected to be the 3rd most common cause of disease-related mortality by 2020 [5]. Cigarette smoking is the primary cause of the development of COPD with an estimated 15% of smokers likely to develop the disease [6]. Cigarette smoke exposure induces destruction of alveoli, mucus hypersecretion and mucociliary dysfunction. These effects are associated with an increased and persistent inflammatory state within the lungs, which promotes airway remodelling eventually leading to airflow obstruction [7-10].

Clinically, COPD is diagnosed through several techniques including spirometry, X- rays or CT scans of the chest, pulmonary function testing, and oximetry or arterial blood gas testing. However, these tests have a number of limitations, including the inability to often distinguish between alternative lung diseases that affect airflow, and COPD. Also, the diverse nature of COPD represents a number of different combinations of symptoms that may require selective treatments based on the set of clinical presentations and molecular signatures [11].

These tests are often performed when the disease is well advanced. While cessation of smoking represents the best option for reduction of disease features, importantly, as COPD

320

continues to develop, early invention may also be helpful in slowing disease progression and improving quality of life [12]. Therefore early identification of COPD is likely to be important. In particular, molecular biomarkers that can identify key aspects of the disease may result in information about predisposition, diagnosis, and prognosis of COPD with the possibility of utilizing these potential biomarkers to reduce the disease morbidity [13,14].

High throughput microarrays are a valuable way of profiling gene expression to identify genes and molecular pathways associated with COPD and smoking-related diseases.

A number of human studies have previously performed arrays on lung tissue in patients exposed to cigarette smoke and who have developed aspects of COPD (refs). These have involved a number of different approaches, including examining whole lung tissue, comparing areas of specific tissue damage (such as emphysema) to neighbouring healthy tissue, examining individual cell types. These approaches can prove informative, however, as

COPD is a highly complex disease involving the likely interaction between multiple cells types and pathways.

A major factor that has hindered the understanding and development of specific therapies for COPD is the lack of a mouse model that recapitulates the hallmark pathological features and molecular changes associated with cigarette smoke exposure in a short time frame. We have recently published data on a mouse model of COPD induced by delivering tightly controlled doses of cigarette smoke directly to the airways [15]. The model is characterised by the progressive presentation of hallmark features of COPD, and is representative of a number of aspects of the human disease. Importantly, alveoli become enlarged and lung function reduced after just 8wks of exposure. Other models use uncontrolled, whole body exposure and require 6mths to induce these features. Significantly,

321

this model enables us to investigate both the induction and progression phases of COPD in a reasonable time frame. Through the development of weight loss, acute and chronic inflammatory responses characterised by neutrophil, macrophage and CD8+ T cell influx into the airways, destruction of alveolar tissue and alveolar enlargement, increased numbers of mucus secreting cells around the airways, and reduced lung function. As in humans, the pathology is not reversible after smoking cessation, thus our model recapitulates the crucial inflammatory, histologic and pathophysiologic features of human disease.

Here we set out to determine the gene expression signature associated with cigarette smoke induced COPD in mouse model to identify pathways involved in pathogenesis and potential targets for therapy. In addition, we compared this gene expression fingerprint with a distinctive human cohort to show the potential correlation between the 2 subsets. Two strains of mice were exposed to physiological levels of cigarette smoke for up to 12 weeks and gene array performed on total lungs over that time period. As development of clinical COPD can often take decades, identification of potential biomarkers through initial screening of whole lung tissue was performed as more likely to result in the identification of transcripts involved in initiation of COPD pathology. In comparison to air controls, cigarette smoke treatment resulted in the significant alteration over the course of 12 weeks of 61 specific genes. The genes associated with this signature predominantly indicated sustained changes in the inflammatory response. The major biological functions involving these genes included the activation, adhesion and migration of leukocytes, cellular apoptosis, control of bacterial infection, phagocytosis, ROS production, and production of increase in matrix metalloproteinases. Comparison to a human dataset representing a gene signature of patients with emphysema indicated good concordance, with 49 genes identified in the human cohort that overlap with the mouse cigarettes smoke induced signature, which represents 79% of the

322

significantly regulated genes identified in the mouse model. This mouse model therefore is strongly representative of the molecular aspects of COPD pathogenesis and provides a good basis in the identification of genes and pathways with the potential to identify genes in the early diagnosis of COPD and well as providing targets for both management and treatment of

COPD.

323

Methods and materials

Mice

Wild type BALB/c and C57 were bred and housed in specific pathogen free conditions. All animal usage was conducted in accordance with guidelines issued by

NHMRC, and approved by the AEC of the University of Newcastle, Australia.

Smoke exposure

In each experiment mice were simultaneously exposed to cigarette smoke (twelve

3R4F reference cigarettes (University of Kentucky, Lexington, Ky) twice per day and 5 times per week for up to 12 weeks) by using a custom-designed and purpose-built nose-only, directed-flow inhalation and smoke-exposure system (CH Technologies, Westwood, NJ) housed in a fume and laminar flow hood [15]. Each exposure lasted 75 minutes.

Tissue collection and RNA isolation

Whole lung tissue was collected and stored in RNAlater®(Ambion). Isolation of total

RNA from lung homogenates was achieved by the guanidinium thiocyanate-phenol- chloroform extraction method using TRIzol®(Invitrogen, Mount Waverley, Aust.). The integrity of total RNA samples was assessed with an Agilent 2100 Bioanalyzer (Agilent

Technologies, Santa Clara, CA) where an RNA Integrity Number (RIN) of >8 was accepted for further analysis.

Microarray profiling

Microarray profiling was performed using GeneChip® Mouse Genome 430 2.0 Arrays

(in GeneChip® HT MG-430 PM 24-Array Plate format; Affymetrix, Santa Clara, CA, USA) containing probe sets to examine >39,000 transcripts. All staining, hybridization and 324

scanning of arrays was performed by the Ramaciotti Centre for Genomics (University of New

South Wales, Sydney, Australia). Briefly, total RNA was amplified and biotin-labeled using the GeneChip® HT 3’ IVT Express Kit (Affymetrix, Santa Clara, CA, USA) and array hybridization and washes were performed according to manufacturer’s protocol. The arrays were scanned using the ??? scanner (presumably a GeneChip 3000).

Gene expression and pathway analysis of mouse array

GeneSpring GX 11.3 software (Agilent Technologies, Palo Alto, CA) was used to normalise murine microarray datasets and identify significant differential gene expression between smoke and non-smoke exposed mice at each given time point analysed. Differential gene expression thresholds were set at a fold change ≥2.0 with a p-value ≤0.05. Unsupervised hierarchical clustering was used to determine expression profiles amongst samples for experimental time points.

In order to determine network and pathway associations, each of the 5 experimental time point datasets consisting of differentially expressed genes (fold change ≥2.0 with a p- value ≤0.05) were analysed using Ingenuity Pathway Analysis (IPA; Ingenuity ® Systems) and mapped to Ingenuity's Knowledge Base. Comparison and subsequent core analyses were used to identify significant associations between datasets and biological function and canonical pathways by dividing the total number of genes in a dataset with the total number of molecules associated with a given pathway with p- values determined using Fisher’s exact test.

325

Quantification of mRNA expression by real-time quantitative PCR

Total RNA was treated with DNAseI (Sigma-Aldrich, Castle Hill, NSW, Aust.), reverse transcribed using Bioscript (Bioline, Alexandria, NSW, Aust.) and random hexamer primers (Invitrogen). Products were amplified with specific custom designed primers (IDT,

Coralville, IA) in a ViiA 7 Real Time PCR system (Life Technologies, Grand Island, NY) utilising SYBR Green real time PCR master mixes (Invitrogen, Grand Island, NY). The relative abundance of each specific cDNA was calculated using HPRT as a reference gene.

Human cohort gene expression analysis

The characteristics of the human cohort used in this study have been previously described [16]. For the study of gene expression in lung tissue of patients with severe COPD, raw CEL files were obtained from GEO (GSE27597) and normalized with RMA using the custom Entrez Gene CDF v16 [17]. Gene expression profiles associated with increasing emphysema severity were identified as previously described [16]. Briefly, linear mixed-effect models were used which included position within the lung and mean linear intercept as fixed effects and patient from which each sample was derived as a random effect.

Alternative analysis

In addition to the genespring normalisation, the microarray data was also normalised using control group GAPDH fluorescence. As low levels of transcripts may have less influence on pathology, the transcripts that were of expressed in relatively low amounts compared to GAPDH genes (<1%) were excluded. These transcripts were then further selected by analysing for 2-fold changes in the lungs of smoked mice compared to controls.

326

Statistical analysis

Unless otherwise stated, data were analysed using unpaired t tests assuming Gaussian distribution with PRISM V6.0d software (Graph Pad). Data are means +/- SEM of 6-8 mice/group, p-values of <0.05 were considered statically significant.

Results

Distinct gene expression profiling and validation of cigarette smoke induced lung damage To determine a murine gene expression signature associated with cigarette smoke exposure, gene expression array was performed on whole lungs from mice exposed to cigarette smoke or air. Genespring analysis identified a subset of 61 genes that were consistently differentially expressed by at least 2 fold (p<0.05) in smoke treated BALB/c mice compared to air treated controls from 4 weeks (when pathological aspects of COPD are not evident), through to 12 weeks, where key features of COPD are observed (Table 1).

Comparative analysis of an alternate mouse strain, C57J at 8 weeks sh0wed that the transcript profiles were similar between these two strains (Table 1). A number of key genes associated with different aspects of cigarette smoking and the development and progression of lung disease were dysregulated. These include increased inflammation, mucus production, development and maintenance of emphysema, and chronic bronchitis. For example, serum amyloid A3 (SAA3), increased expression of which is associated with general inflammation, was highly up-regulated. Matrix metalloproteinase 12 (MMP12), associated with lung destruction was also highly upregulated throughout the timecourse. The Slc26A4 gene, which encodes for the molecule Pendrin, is involved in mucus production and is increased throughout the course of 12 weeks in BALB/c mice and at 8 weeks in C57J.

327

Technical validation was performed by RT-PCR on a number of key genes associated with the mouse molecular fingerprint. MMP12, Cxcl1, Cxcl5, MARCO and MSR1 were all upregulated in BALB/c mice throughout the experiment. These transcripts were also validated in C57J mice at week 8, indicating concordant relative abundance and increase of transcripts with the values identified in the array (Figure 1).

Identification of a distinct gene signature from human with emphysema related lung destruction Array data obtained from humans that had emphysema associated lung damage has been been previously performed [16]. A gene signature was identified that correlated transcript expression to the degree of COPD/emphysema severity. The resultant analysis identified 127 distinctly regulated transcripts. Analysis comparing the 127 gene signature identified from this human dataset to the mouse gene expression signature was performed.

The two datasets showed good concordance between each other. Of the differentially regulated transcripts identified in the mouse dataset (n=61 transcripts, fold change>2, p<0.05), 49 of these transcripts were associated with the human dataset, representing 79% of the mouse gene expression signature (Table 2). Validation by RT-PCR was performed on a number of transcripts that were commonly expressed in both the mouse and human datasets.

Independent validation of gene expression signature

In order to determine further confirm the validity of the gene expression signature, an independent analysis was undertaken of the initial array data. A more stringent analysis resulted in the identification of a more constricted gene expression list comprising of a subset of 17 transcripts (Table 3). Analyses of the 2 mouse datasets and the human dataset identified

15 common transcripts (Figure 2).

328

Major biological functions associated with genetic signature

Differentially expressed transcripts which were concordant in both human and mouse lung datasets (49 genes in total, Table 2) were analysed using Ingenuity Pathway Analysis

(IPA) to determine the major biological functions associated with cigarette smoke induced lung damage. In particular, pathways associated with that immune regulaton were highly system were , further defining this process as central to the lung disease initiation and development (Table 4). These functions included increased diapedesis, leukocyte numbers, movement and activation. Other functions that were identified involved lung fibrosis, altered phagocytosis, cell death (in particular apoptosis), and ROS induction through H2O2 production.

Network analysis of common gene signatures

We performed network analysis to determine important pathways and molecular associations with the genes differentially expressed. The most significant networks were identified based on IPA ranking of significance which was determined by the known connections identified between the different molecules (Figure 4). The most significant networks show that ERK1/2, and CCL2 and CCL3, molecules involved in the chemoattraction of polymorphonuclear cells, are particularly involved (Figure 3A). The second most significant networks involves intricate interactions between AKT, P38 MAPK, TGFb and collagen production (Figure 3B). The third most significant pathway indicates a role for polyubiquitin, ubiquitin C. and potentially the disturbance of multiple processes involving cellular homeostasis by disrupting the ubiquitination-proteasome pathway (Figure 3C).

329

Discussion

Cigarette smoke associated COPD is a significant problem that is often difficult to detect until well advanced, and while smoking rates have reduced, incidence and death continue to rise due to the increase in population size. In addition, disease can progress after smoking cessation. While there is currently no cure, early intervention can likely increase quality of life and be beneficial in reducing the rate of disease progression [12]. Mouse models then can more accurately reflect key aspects of human COPD would be most useful in assessing disease causality and the development of efficacious therapeutic agents, with the aim of inhibiting disease progression and perhaps repair of lung damage. Here we identify a transcriptional profile associated with the development and progression of COPD in mice exposed to cigarette smoke. The aim was to identify a unique gene expression signature to aid in disease identification and development of therapeutic targets. In addition, this murine signature was compared to a human dataset comprising of patients with gene expression correlated with the emphysema severity i.e. greater transcript expression changes associated with worsening disease.

A distinct transcript signature was identified in mice comprising of 61 genes that were differentially regulated. When compared with the human dataset, which showed that of the

127 genes differentially expressed and associated with emphysema associated lung damage,

49 transcripts were identical to those identified in the mouse dataset, representing 79% concordance between the 2 subsets. This relationship represents a relatively good similarity, given that comparisons between different human subsets and differences between animal models can often be substantial. The strong correlation between the mouse and human datasets may indicate a mouse model that may have relevance in translating therapeutic value of drugs. Here we examined the whole lung in the mouse as the heterogenous nature of the

330

disease and the many cell types involved in its pathogenesis make it important to determine the gene expression in a global sense [18]. Also, within the lung, selective regional identification of a gene transcript signature before clinical features manifest is not possible.

The array identified that gene expression changes was particularly associated with inflammation, and that they were increased days after cigarette smoke exposure, and were maintained throughout the period of smoking exposure. This includes general markers of inflammation a such as SAA3, an acute phase inflammatory protein whose expression is increased and maintained during smoke exposure, can stimulate cell activation through p38, in a manner dependent on TLR4, MD-2, and MyD88 [19] [20]. Other inflammatory associated gene expression changes are associated with activation of cells associated with

COPD, in particular alveolar and interstitial macrophages, which represent one of the main contributing cellular sources to disease development and progression [15]. Their influx and activation state are crucial steps in disease development. Its is also well known that phagocytosis is impaired in COPD, and that this has detrimental effects on disease [21]. Here the combination of several pathways involved in macrophage movement and activation, as well as phagocytosis. This includes the upregulation of Trem2. Increased expression of

Trem2 can result in increased association with DAP12, a molecule involved activating immune responses. The interaction between these two can result in the increased production of CCL2 [22], and therefore can contribute to the influx of macrophages monocytes and neutrophils. MARCO, a scavenger receptor involved in bacterial clearance has been reported to be reduced when stimulated in vitro with CS [23]. However, in both the mouse model and in human samples, analysis revealed an increase in MARCO expression. This may either indicate an alternative role for MARCO. Apart from their anti-bacterial role, it has been suggested that the increase in scavenger receptors can result in increased macrophage adhesion, retention and activation through increased binding with collagens whose structure

331

is altered by cigarette smoke [24]. In addition, recruitment of macrophages to the lungs exposed to CS is likely to be further enhanced by the increase in the macrophage chemokine

CCL5 expression [25] suggesting markers of macrophage recruitment as well as survival and proliferation as particularly prominent.

Neutrophil infiltration is a significant feature in COPD [26]. A large body of evidence implicates neutrophil and their products, including elastases, cytokines and chemokines as effectors of lung damage and continual decline in lung function. are important in neutrophil function cytokines in modulating neutrophil function and survival. A number of neutrophil chemoattractants and activators are increased including CCL9, CXCL1, CXCL2 and CXCL5.

Therefore as numerous chemokines are involved in neutrophil attraction, and their persistence in expression, suggest that negating neutrophil influx and activation may require complex treatments. CH25H is an enzyme that converts cholesterol into 25-HC. CH25H has been shown to be increased in COPD and localized in alveolar macrophages and pneumocytes of

COPD patients [27]. It has been associated with various aspects of lung disease, including cytokine release from primary human bronchial epithelial cells [28], contributing to fibroblast-mediated lung tissue remodeling by promoting myofibroblast differentiation and the excessive release of extracellular matrix protein and MMPs via an NF-κB-TGF-β dependent pathway [29]. Particularly, neutrophil inflammation is positively correlated with

CH25H production, partly through increased CCL5 production [30].

Excessive production of airway mucus is a feature of bronchial COPD and contributes to morbidity and mortality. Pendrin, encoded by the SLC26A4 (PDS) gene as a molecule responsible for airway mucus production. In both asthma and COPD mouse models, pendrin is up-regulated at the apical side of airway epithelial cells in association with mucus

332

overproduction. Pendrin induced expression of MUC5AC, a major product of mucus in asthma and COPD, in airway epithelial cells. Finally, the enforced expression of pendrin in airway epithelial cells in vivo, using a Sendai virus vector, rapidly induced mucus overproduction in the lumens of the lungs together with neutrophilic infiltration in mice.

These findings collectively suggest that pendrin can induce mucus production in airway epithelial cells and may be a therapeutic target candidate for COPD [31].

A number of factors are identified here that have been shown to have a strong effect on alveolar septum destruction resulting in alveolar enlargement and reduction of gas exchange area. NOXO1, through regulation of NOX1, and MMP12 are both involved in alveolar destruction. NOX1 has been shown to upregulate ROS production through a Stat3 dependent mechanism which results in alveolar cell death in Acute Respiratory Distress

Syndrome [32]. MMPs represent powerful proteinases that can cleave a variety of targets in the lung resulting in tissue destruction. Snps in the MMP12 gene have been associated with a lower risk of developing COPD [33], and others have shown that MMP12 is increased in human lungs [34]. Experimentally, KO mice have demonstrated the importance of MMP12 in the development of emphysema [35].

Therefore the array data confirms the complexity of the COPD, and the data indicates molecular targets and their associated pathways that are likely to be good candidates for future studies and development of therapeutics. The mouse model also has strong concordance with a human dataset, suggesting its potential in translating potential findings to the clinic.

333

References

1. The Global Strategy for the Diagnosis, Management and Prevention of COPD, Global

Initiative for Chronic Obstructive Lung Disease (GOLD) 2014 update.

http://www.goldcopd.org/. (2014)

2. Spurzem JR, Rennard SI (2005) Pathogenesis of COPD. Semin Respir Crit Care Med

26: 142–153. doi:10.1055/s-2005-869535.

3. Tuder RM, Petrache I (2012) Pathogenesis of chronic obstructive pulmonary disease. J

Clin Invest 122: 2749–2755. doi:10.1172/JCI60324.

4. WHO, COPD Fact sheet N°315, February 2011.

5. Murray CJ, Lopez AD (1997) Global mortality, disability, and the contribution of risk

factors: Global Burden of Disease Study. Lancet 349: 1436–1442. doi:10.1016/S0140-

6736(96)07495-8.

6. Lopez AD (2006) Chronic obstructive pulmonary disease: current burden and future

projections. European Respiratory Journal 27: 397–412.

doi:10.1183/09031936.06.00025805.

7. Wright JL, Cosio M, Churg A (2008) Animal models of chronic obstructive

pulmonary disease. AJP: Lung Cellular and Molecular Physiology 295: L1–L15.

doi:10.1152/ajplung.90200.2008.

8. Sutherland ER, Martin RJ (2003) Current reviews of allergy and clinical immunology.

Allergy 112: 819–827. Available:

http://scholar.google.com/scholar?cluster=5498047819725132311&hl=en&num=30&

334

as_sdt=0,5.

9. Wen Y, Reid DW, Zhang D, Ward C, Wood-Baker R, et al. (2010) Assessment of

airway inflammation using sputum, BAL, and endobronchial biopsies in current and

ex-smokers with established COPD. COPD 5: 327–334. doi:10.2147/COPD.S11343.

10. Schleimer RP, Schleimer RP, Schleimer RP (2005) Innate Immune Responses and

Chronic Obstructive Pulmonary Disease: "Terminator" or "Terminator 2"?

Proceedings of the American Thoracic Society 2: 342–346. Available:

http://pats.atsjournals.org/cgi/doi/10.1513/pats.200504-030SR.

11. Lam DCL, Hui CKM, Ip MSM (2012) Issues in pulmonary function testing for the

screening and diagnosis of chronic obstructive pulmonary disease. Curr Opin Pulm

Med 18: 104–111. doi:10.1097/MCP.0b013e32834feae7.

12. Decramer M, Celli B, Kesten S, Lystig T, Mehra S, et al. (2009) Effect of tiotropium

on outcomes in patients with moderate chronic obstructive pulmonary disease

(UPLIFT): a prespecified subgroup analysis of a randomised controlled trial. Lancet

374: 1171–1178. doi:10.1016/S0140-6736(09)61298-8.

13. Patel A, Hurst J, Wedzicha J (2010) The Potential Value of Biomarkers in Diagnosis

and Staging of COPD and Exacerbations. Semin Respir Crit Care Med 31: 267–275.

doi:10.1055/s-0030-1254067.

14. Dahl M, Nordestgaard BG (2009) Markers of early disease and prognosis in COPD.

COPD 4: 157–167.

15. Beckett EL, Stevens RL, Jarnicki AG, Kim RY, Hanish I, et al. (2013) A new short-

term mouse model of chronic obstructive pulmonary disease identifies a role for mast

335

cell tryptase in pathogenesis. J Allergy Clin Immunol 131: 752–762.

doi:10.1016/j.jaci.2012.11.053.

16. Campbell JD, McDonough JE, Zeskind JE, Hackett TL, Pechkovsky DV, et al. (2012)

A gene expression signature of emphysema-related lung destruction and its reversal by

the tripeptide GHK. Genome Medicine 4: 67.

17. Dai M (2005) Evolving gene/transcript definitions significantly alter the interpretation

of GeneChip data. Nucleic Acids Research 33: e175–e175. doi:10.1093/nar/gni179.

18. Savarimuthu Francis SM, Larsen JE, Pavey SJ, Duhig EE, Clarke BE, et al. (2011)

Genes and Gene Ontologies Common to Airflow Obstruction and Emphysema in the

Lungs of Patients with COPD. PLoS ONE 6: e17442.

doi:10.1371/journal.pone.0017442.s011.

19. Deguchi A, Tomita T, Omori T, Komatsu A, Ohto U, et al. (2013) Serum Amyloid A3

Binds MD-2 To Activate p38 and NF- B Pathways in a MyD88-Dependent Manner.

The Journal of Immunology 191: 1856–1864. doi:10.4049/jimmunol.1201996.

20. Bozinovski S, Uddin M, Vlahos R, Thompson M, McQualter JL, et al. (2012) Serum

amyloid A opposes lipoxin A₄ to mediate glucocorticoid refractory lung inflammation

in chronic obstructive pulmonary disease. Proc Natl Acad Sci USA 109: 935–940.

doi:10.1073/pnas.1109382109.

21. Donnelly LE, Barnes PJ (2012) Defective phagocytosis in airways disease. CHEST

141: 1055–1062. doi:10.1378/chest.11-2348.

22. Koth LL, Cambier CJ, Ellwanger A, Solon M, Hou L, et al. (2010) DAP12 Is Required

for Macrophage Recruitment to the Lung in Response to Cigarette Smoke and

336

Chemotaxis toward CCL2. The Journal of Immunology 184: 6522–6528.

doi:10.4049/jimmunol.0901171.

23. Baqir M, Chen C-Z, Martin RJ, Thaikoottathil J, Case SR, et al. (2008) Cigarette

smoke decreases MARCO expression in macrophages: implication in Mycoplasma

pneumoniae infection. Respiratory Medicine 102: 1604–1610.

doi:10.1016/j.rmed.2008.05.002.

24. Kirkham PA, Spooner G, Ffoulkes-Jones C, Calvez R (2003) Cigarette smoke triggers

macrophage adhesion and activation: role of lipid peroxidation products and scavenger

receptor. Free Radical Biology and Medicine 35: 697–710. doi:10.1016/S0891-

5849(03)00390-3.

25. Balamayooran G, Batra S, Cai S, Mei J, Worthen GS, et al. (2012) Role of CXCL5 in

Leukocyte Recruitment to the Lungs during Secondhand Smoke Exposure. American

Journal of Respiratory Cell and Molecular Biology 47: 104–111.

doi:10.1165/rcmb.2011-0260OC.

26. Quint JK, Wedzicha JA (2007) The neutrophil in chronic obstructive pulmonary

disease. Journal of Allergy and Clinical Immunology 119: 1065–1071.

doi:10.1016/j.jaci.2006.12.640.

27. Sugiura H, Koarai A, Ichikawa T, Minakata Y, Matsunaga K, et al. (2012) Increased

25-hydroxycholesterol concentrations in the lungs of patients with chronic obstructive

pulmonary disease. Respirology 17: 533–540. doi:10.1111/j.1440-1843.2012.02136.x.

28. Koarai A, Yanagisawa S, Sugiura H, Ichikawa T, Kikuchi T, et al. (2012) 25-

Hydroxycholesterol enhances cytokine release and Toll-like receptor 3 response in

airway epithelial cells. Respiratory Research 13: 63. doi:10.1186/1465-9921-13-63.

337

29. Ichikawa T, Sugiura H, Koarai A, Kikuchi T, Hiramatsu M, et al. (2013) 25-

Hydroxycholesterol promotes fibroblast-mediated tissue remodeling through NF.

Experimental Cell Research 319: 1176–1186. doi:10.1016/j.yexcr.2013.02.014.

30. Diczfalusy U, Olofsson KE, Carlsson AM, Gong M, Golenbock DT, et al. (2009)

Marked upregulation of cholesterol 25-hydroxylase expression by lipopolysaccharide.

The Journal of Lipid Research 50: 2258–2264. doi:10.1194/jlr.M900107-JLR200.

31. Nakao I, Kanaji S, Ohta S, Matsushita H, Arima K, et al. (2008) Identification of

pendrin as a common mediator for mucus production in bronchial asthma and chronic

obstructive pulmonary disease. J Immunol 180: 6262–6269.

32. Carnesecchi S, Dunand-Sauthier I, Zanetti F, Singovski G, Deffert C, et al. (2014)

NOX1 is responsible for cell death through STAT3 activation in hyperoxia and is

associated with the pathogenesis of Acute Respiratory Distress Syndrome. Int J Clin

Exp Pathol : 537–551.

33. Hunninghake GM, Cho MH, Tesfaigzi Y, Soto-Quiros ME, Avila L, et al. (2009)

MMP12,Lung Function, and COPD in High-Risk Populations. N Engl J Med 361:

2599–2608. doi:10.1056/NEJMoa0904006.

34. Molet S, Belleguic C, Lena H, Germain N, Bertrand CP, et al. (2005) Increase in

macrophage elastase (MMP-12) in lungs from patients with chronic obstructive

pulmonary disease. Inflamm res 54: 31–36. doi:10.1007/s00011-004-1319-4.

35. Hautamaki RD (1997) Requirement for Macrophage Elastase for Cigarette Smoke-

Induced Emphysema in Mice. Science 277: 2002–2004.

doi:10.1126/science.277.5334.2002.

338

Figure Legends

Figure 1. Schematic of the features that develop during the course of cigarette smoke exposure and development of COPD.

Figure 2. Quantitative RT- PCR validation of representative genes shared within the human and mouse datasets. Gene expression relative to HPRT was validated in BALB/c mice over a period of 12 weeks smoking. Expression was confirmed in C57J mice also after 8 weeks of smoking. Data represents means +/- SEM of 6-8 mice/group. Students t-test p**<0.005p***<0.0005.

Figure 3. Quantitative RT-PCR validation of representative genes shared within the human and mouse datasets. Gene expression relative to HPRT was validated in human samples. Data represents the means +/- SEM of 4-5 subjects. Students t-test p**<0.005p***<0.0005.

Figure 4. A Venn diaGramof gene expression signature classified as differentially expressed in human and mouse lungs with COPD. Mouse lungs were analysed by two different methods (see Materials and Methods). The total number of non-redundant transcripts with significant p values and changes >2-fold (in brackets) were detected by microarray analysis. Overlaps in commonly expressed transcripts with changed values that were shared among humans and 2 methods of validation of the mouse array data analyses are identified.

339

Figure 5. The top three molecular networks identified by Ingenuity pathway analysis. The three most significant molecular networks as determined by IPA pathway enrichment analysis proceeding from most (A) to less (C) significance.

340

Table 1: Distinctive gene expression signature in the lungs of mice exposed to cigarette smoke.

61 unique genes were differentially expressed in the lungs of smoked mice by >2 fold (p<0.05) compared to air control BALB/c mice over 12wks and C57Bl at 8 weeks.

Gen 4 6 8 1 8

e Wk Wk Wk 2 Wk Wk

BALB/c 57Bl

2 1 2 3 3

Saa3 6.283 0.418 6.462 1.906 0.005

Mmp 1 1 1 3 9

12 7.990 4.619 7.966 1.317 .215

Cxcl 1 1 1 1 8

1 5.596 5.466 2.004 5.460 .835

Gpn 1 1 1 1 1 mb 0.300 0.310 1.395 4.693 1.456

Ch25 1 1 1 1 6 h 0.180 0.778 0.955 0.676 .518

Slc2 3 2 9 1 6

6a4 0.891 7.878 .635 2.883 .415

Orm 9 3 9 9 1

1 .387 .074 .306 .634 6.562

Cxcl 3 3 8 6 4

5 7.752 2.873 .712 .014 .357

Vnn 8 8 7 1 6

1 .883 .127 .559 0.358 .135

6 5 6 1 2

Ccl9 .547 .186 .922 0.576 .350

1100 6 4 6 7 3

001G20Rik .540 .868 .908 .800 .444

Trem 3 3 6 6 6

2 .985 .513 .279 .598 .605

6 5 6 9 7

Ctsk .460 .678 .246 .328 .683

341

AA4 4 3 6 1 5

67197 .266 .140 .209 2.008 .162

5 4 6 1 3

Pigr .362 .071 .101 1.935 .441

Zran 4 4 5 5 1 b3 .734 .842 .996 .006 1.969

Slc6 5 3 5 1 4 a20a .957 .924 .649 2.399 .673

Marc 5 4 5 1 1 o .702 .376 .033 1.276 5.660

Inhb 5 3 4 5 2 a .020 .122 .895 .382 .503

Nox 4 4 4 5 5 o1 .863 .667 .788 .471 .436

1300 3 3 4 5 5

002K09Rik .512 .186 .557 .128 .075

Slc3 5 4 4 5 1

9a2 .511 .977 .415 .331 4.310

Cxcl 6 5 4 4 6

2 .594 .246 .165 .305 .603

3 3 4 4 5

Msr1 .793 .141 .024 .745 .263

4 5 3 3 3

Ptgir .267 .063 .816 .600 .908

Clec 4 2 3 4 5

5a .015 .773 .794 .856 .208

5 4 3 4 3

Lcn2 .160 .624 .765 .465 .386

3 4 3 3 2

Itgax .595 .562 .647 .986 .745

Myo 2 2 3 3 3

5a .891 .347 .439 .270 .884

Cd20 3 3 3 3 3

0r4 .032 .006 .262 .984 .892

Atp6 3 3 3 3 3 v0d2 .607 .329 .251 .863 .695

3 3 4 5

342

Lhfp .484 4 .152 .332 .201 l2 .087

3 3 3 2 4

Il1rn .868 .398 .068 .947 .003

3 2 2 4 3

Cd68 .801 .949 .975 .156 .374

3 4 2 4 2

F7 .167 .045 .966 .179 .448

Mmp 3 3 2 2 2

19 .200 .738 .943 .350 .506

Kyn 2 2 2 2 2 u .489 .605 .936 .815 .320

3 4 2 3 3

Olr1 .576 .017 .921 .984 .370

Prun 3 2 2 3 2 e2 .085 .388 .920 .370 .171

2 2 2 2 1

Ccl3 .836 .325 .892 .587 0.710

Lem 2 2 2 3 2 d1 .551 .651 .835 .285 .804

Mcol 2 2 2 2 3 n3 .677 .862 .835 .773 .536

3 2 2 3 3

Ccl6 .100 .651 .777 .497 .620

Vnn 3 2 2 2 2

3 .923 .918 .718 .580 .925

Slc7 2 3 2 3 2 a2 .873 .530 .713 .319 .527

Cd20 3 3 2 4 2

0r1 .824 .747 .711 .302 .909

4 3 2 3 2

Cd14 .891 .719 .681 .722 .308

2 2 2 3 4

Spp1 .735 .216 .593 .722 .679

Lrp1 2 2 2 2 3

2 .415 .655 .505 .986 .488

Rgs1 2 2 2 2 6

343

6 .741 .081 .432 .297 .926

Niac 2 2 2 3 3 r1 .769 .529 .387 .604 .241

2 2 2 6 2

Cybb .883 .582 .365 .213 .222

2 2 2 2 2

Ly75 .098 .657 .303 .479 .102

Clec 3 2 2 3 3

4n .284 .168 .245 .470 .161

2 2 2 2 2

Pld3 .290 .316 .193 .236 .611

4 2 2 2 2

Has3 .067 .191 .177 .419 .853

1600 4 3 2 6 2

029D21Rik .034 .326 .167 .169 .706

Csf2 3 3 2 2 2 rb2 .887 .917 .073 .761 .122

Fabp - - - - 5

1 3.650 3.110 2.046 2.400 .174

- - - - 2

Bcan 2.588 2.587 2.403 3.855 .705

- - - - -

Ppbp 2.328 3.485 3.135 2.224 2.317

344

Table 2: Transcripts associated with human emphysema associated lung destruction correlate with mouse transcript signature. This table represents the 49 genes that are regulated in response to cigarette smoke associated lung damage between human subjects and a mouse model of COPD.

Gen Coeffici p-

e ent value

LY7 0.45373 0.0024

5 4992 19035

PPB 0.59104 0.0038

P 6072 91716

PIG 0.31045 0.0053

R 7077 16497

HAS 0.21611 0.0096

3 1977 35137

CD2 0.28219 0.0178

00R1 8416 68133

CTS 0.32519 0.0289

K 1361 67935

ATP 0.34974 0.0387

6V0D2 9019 606

TRE 0.28030 0.0635

M2 5134 33476

MS 0.26104 0.0953

R1 3461 95111

CCL 0.25218 0.1182

3 7343 65691

GPN 0.22916 0.1358

MB 7066 82472

CX 0.30329 0.1538

345

CL1 2956 30922

IL1 0.20182 0.1671

RN 2203 02704

CLE 0.16527 0.1928

C5A 0342 00899

CX 0.17172 0.2030

CL5 9207 47376

VN 0.16920 0.2068

N1 5433 70806

MM 0.23462 0.2102

P19 0844 76541

MM 0.21724 0.2144

P12 1588 84092

LRP 0.08613 0.2288

12 2526 06668

VN 0.15075 0.3335

N3 5167 7385

ZRA 0.02442 0.3755

NB3 5208 85882

LHF 0.06329 0.3857

PL2 1103 68159

PLD 0.07454 0.3920

3 1506 9649

PTG 0.07156 0.4301

IR 1596 53122

CH2 0.13548 0.4729

5H 4728 87735

CX 0.19070 0.5002

CL2 0368 56025

INH 0.07852 0.5197

BA 8001 5454

0.02667 0.5377

F7 2262 86405

346

OLR 0.10570 0.5713

1 3222 9801

SLC 0.02965 0.5755

39A2 2298 46585

CY 0.07632 0.5854

BB 057 38094

MY 0.04187 0.5874

O5A 5845 07549

RGS 0.07228 0.5923

16 4559 12421

LE 0.02296 0.6386

MD1 1684 69819

CD1 0.05917 0.6475

4 2097 58528

NO 0.02132 0.6586

XO1 1744 23967

MC 0.05445 0.6694

OLN3 1917 6072

SLC 0.05729 0.6919

26A4 4643 47925

PRU 0.06339 0.7138

NE2 2865 00831

LCN 0.10925 0.7231

2 6607 87433

SLC 0.03720 0.7657

7A2 3711 7338

MA 0.03457 0.7735

RCO 9053 40425

KY 0.03461 0.8051

NU 0099 4131

FAB 0.01080 0.8404

P1 6785 22637

BC 0.00784 0.8630

347

AN 1254 37003

ITG 0.02208 0.8635

AX 4359 5931

OR 0.04664 0.8726

M1 2002 60199

CD6 0.00584 0.9580

8 265 99107

SPP 0.00029 0.9988

1 5854 82285

Table 3: A more stringent comparative analysis confirms gene expression regulation by genespring in mouse lungs exposed to cigarette smoke.

Ge Wk Wk Wk Wk Wk

ne 4 6 8 12 8

BA C5

LB/c 7Bl

Saa 15. 10. 37. 22. 18.

3 7 6 3 8 9

M 12. 13. 14. 23 5.9 mp12 1 8 7

Gp 10. 10. 10. 7.1 8.1 nmb 4 9 1

Ma 10. 4.2 4.1 6.5 8.4 rco 9

Vn 6 7.6 6.3 7.1 4.3 n1

Ch 10. 7.1 6 7.8 6 25h 1

348

Cts 4.5 5.4 5.6 6.9 7 k

Slc 21. 25. 22. 4.7 5 26a4 2 9 8

Slc 12. 3.9 4.8 4.4 4 39a2 4

Slc 4.1 4.2 4.4 4 4.3 6a20a

Zra 10. 3.2 7.6 4 3.7 nb3 3

Msr 2.6 3.3 3.5 3.4 3.9 1

No 3.6 4.1 2.9 4 4.8 xo1

Cxc 5.1 4.9 2.8 x 6.4 l2

Cd 2.6 2.8 2.5 3 3 68

Tre 2.7 3.3 -2.6 4.7 5.8 m2

Fab -5.3 -3.2 -4.8 -3.3 -6.1 p1

349

Table 4: Major biological functions enriched which are associated with the gene signatures identified in mouse and human COPD. p-value is calculated within Ingenuity pathway analysis using the cumulative number of literature references supporting the prediction of increase in activity from the overlapping mouse and human COPD dataset i.e.

49 genes.

Predicted function p-value Molecules # Molecules

CCL3L1/CCL3L3, Ccl6, Ccl9, CD14, CLEC5A, CSF2RB, CXCL2, CXCL3, CXCL6, CYBB, F7, HCAR2, IL1RN, leukocyte migration 4.34E-14 24 INHBA, ITGAX, LCN2, MMP12, MSR1, OLR1, PIGR, PPBP, RGS16, SPP1, TREM2 CCL3L1/CCL3L3, Ccl6, Ccl9, CD14, CLEC5A, CXCL2, cell movement 2.22E-12 CXCL3, CXCL6, CYBB, F7, IL1RN, INHBA, ITGAX, LCN2, 21 MMP12, MSR1, PIGR, PPBP, RGS16, SPP1, TREM2 CCL3L1/CCL3L3, CD14, CLEC5A, CXCL2, CXCL3, CXCL6, inflammatory response 5.20E-12 CYBB, F7, IL1RN, INHBA, LCN2, LY75, MARCO, MSR1, 20 OLR1, PIGR, PPBP, SLC7A2, SPP1, VNN1 Ccl6, Ccl9, CD14, CLEC5A, CSF2RB, CXCL2, CXCL3, activation of cells 2.04E-10 CXCL6, CYBB, F7, IL1RN, INHBA, ITGAX, LCN2, MSR1, 21 PPBP, PTGIR, SLC7A2, SPP1, TREM2, VNN1 1300002K09Rik, CCL3L1/CCL3L3, CD14, CD68, CSF2RB, binding of cells 9.91E-10 CXCL2, CXCL3, HAS3, INHBA, ITGAX, MARCO, MSR1, 15 OLR1, PIGR, SPP1 CCL3L1/CCL3L3, CD14, CD200R1, CSF2RB, CXCL2, adhesion of immune cells 1.13E-09 CXCL3, ITGAX, MARCO, MSR1, OLR1, PPBP, RGS16, 13 SPP1 CCL3L1/CCL3L3, CXCL2, CXCL3, CXCL6, CYBB, IL1RN, cell movement of myeloid cells 2.63E-09 INHBA, ITGAX, LCN2, MMP12, MSR1, PIGR, PPBP, SPP1, 15 TREM2 CCL3L1/CCL3L3, Ccl6, Ccl9, CD14, CLEC5A, CSF2RB, CXCL2, CXCL3, CXCL6, CYBB, F7, HAS3, HCAR2, IL1RN, migration of cells 2.82E-09 27 INHBA, ITGAX, LCN2, MMP12, MMP19, MSR1, OLR1, PIGR, PPBP, PTGIR, RGS16, SPP1, TREM2 CD14, CLEC5A, CXCL3, CXCL6, IL1RN, INHBA, LCN2, activation of myeloid cells 3.15E-09 11 MSR1, SLC7A2, SPP1, TREM2 CCL3L1/CCL3L3, Ccl6, CD14, CSF2RB, CXCL2, CXCL3, recruitment of leukocytes 3.15E-08 11 HCAR2, IL1RN, LCN2, OLR1, SPP1 CCL3L1/CCL3L3, CLEC5A, CSF2RB, CXCL2, CXCL3, infiltration of cells 5.90E-08 12 CYBB, IL1RN, LCN2, MMP12, MSR1, PPBP, SPP1 CCL3L1/CCL3L3, CD14, CXCL6, CYBB, IL1RN, ITGAX, Bacterial Infection 7.21E-08 12 LCN2, MARCO, MMP12, MSR1, PIGR, SPP1 CD14, CH25H, CSF2RB, ITGAX, LY75, MARCO, MCOLN3, phagocytosis 1.46E-07 10 MSR1, MYO5A, TREM2 CD14, CH25H, CLEC5A, CSF2RB, CXCL3, ITGAX, immune response of phagocytes 4.57E-07 8 MARCO, TREM2 CD14, CH25H, CLEC5A, CSF2RB, CXCL3, ITGAX, response of myeloid cells 5.38E-07 8 MARCO, TREM2 activation and influx of neutrophils 6.26E-07 CD14, CXCL3, CXCL6, IL1RN, LCN2, SPP1 6 CXCL2, CXCL3, IL1RN, LCN2, MMP12, MSR1, PPBP, infiltration of granulocytes 1.11E-06 8 SPP1 migration of neutrophils 1.69E-06 CCL3L1/CCL3L3, CXCL2, CXCL3, CXCL6, CYBB, SPP1 6 chronic obstructive pulmonary disease 1.82E-06 CXCL3, CYBB, LCN2, MMP12, PTGIR, SPP1, TREM2 7 CD14, CLEC5A, CSF2RB, CXCL3, CXCL6, F7, IL1RN, activation of leukocytes 1.99E-06 13 INHBA, LCN2, MSR1, SLC7A2, SPP1, TREM2 CCL3L1/CCL3L3, CD14, CXCL2, F7, ITGAX, LCN2, atherosclerosis 3.40E-06 11 MMP12, MSR1, OLR1, PTGIR, SPP1 apoptosis of myeloid cells 4.75E-06 CD14, CXCL2, CYBB, HCAR2, IL1RN, LCN2, MSR1 7 CCL3L1/CCL3L3, Ccl6, Ccl9, CXCL2, CXCL3, CXCL6, F7, chemotaxis of cells 6.11E-06 11 INHBA, PIGR, PPBP, SPP1 CD14, CH25H, CSF2RB, ITGAX, LY75, MARCO, MSR1, phagocytosis of cells 6.63E-06 8 TREM2 CD14, CD200R1, CSF2RB, CYBB, MARCO, MMP12, function of antigen presenting cells 7.40E-06 8 MSR1, SPP1 adhesion of phagocytes 1.01E-05 CCL3L1/CCL3L3, CXCL2, CXCL3, ITGAX, MARCO, PPBP 6 damage of lung 1.01E-05 CCL3L1/CCL3L3, Ccl6, CD14, MARCO, MMP12, OLR1 6 CCL3L1/CCL3L3, Ccl9, CD200R1, CSF2RB, CXCL2, quantity of leukocytes 1.19E-05 CXCL6, CYBB, FABP1, IL1RN, INHBA, LCN2, MMP12, 15 PIGR, PTGIR, SPP1 infiltration by neutrophils 1.94E-05 CXCL2, CXCL3, IL1RN, MMP12, PPBP, SPP1 6 CCL3L1/CCL3L3, Ccl6, CD14, CSF2RB, CXCL3, CYBB, inflammation of lung 2.25E-05 9 MMP12, PTGIR, SPP1 cell movement of monocytes 2.52E-05 CCL3L1/CCL3L3, CXCL3, INHBA, ITGAX, MMP12, SPP1 6 atherosclerotic lesion 2.81E-05 CCL3L1/CCL3L3, CXCL2, F7, MMP12, MSR1, SPP1 6 distribution of fatty acid 4.19E-05 FABP1, LCN2 2 innate immune response 4.67E-05 CLEC5A, CLEC6A, CYBB, LCN2, VNN1 5 apoptosis of neutrophils 4.67E-05 CXCL2, CYBB, HCAR2, IL1RN 4 cell movement of natural killer cells 6.68E-05 CCL3L1/CCL3L3, Ccl6, CXCL3, SPP1 4 function of macrophages 7.42E-05 CD14, CYBB, MARCO, MMP12, MSR1, SPP1 6 clearance of bacteria 7.48E-05 CD14, MARCO, OLR1 3 phagocytosis of phagocytes 8.15E-05 CD14, CH25H, CSF2RB, MARCO, TREM2 5 phagocytosis of myeloid cells 8.48E-05 CD14, CH25H, CSF2RB, MARCO, TREM2 5 quantity of antigen presenting cells 8.65E-05 Ccl9, CD200R1, FABP1, IL1RN, LCN2, MMP12, SPP1 7 acute inflammatory response 9.78E-05 CXCL2, LCN2, VNN1 3 synthesis of hyaluronic acid 9.78E-05 HAS3, IL1RN, SPP1 3 adhesion of neutrophils 1.04E-04 CXCL2, CXCL3, ITGAX, PPBP 4 CD14, CSF2RB, CXCL2, CYBB, HCAR2, IL1RN, INHBA, cell death of immune cells 1.12E-04 11 LCN2, MSR1, PPBP, SPP1 accumulation of phagocytes 1.38E-04 CXCL2, CXCL3, IL1RN, ITGAX, SPP1 5 chemotaxis of bone marrow-derived neutrophils1.39E-04 CXCL2, CXCL3 2 respiratory burst 1.47E-04 CD14, CXCL2, CXCL3, CYBB 4 quantity of interleukin 1.59E-04 CD14, CXCL6, IL1RN, OLR1, SPP1 5 migration of macrophages 1.73E-04 IL1RN, MMP12, SPP1, TREM2 4 chemoattraction of phagocytes 2.57E-04 CCL3L1/CCL3L3, CXCL3, PPBP 3 CCL3L1/CCL3L3, CD200R1, CSF2RB, CXCL6, CYBB, quantity of granulocytes 2.61E-04 7 LCN2, MMP12 abnormal quantity of TNF 3.07E-04 CCL3L1/CCL3L3, CD14, TREM2 3 abnormality of immune system 3.49E-04 CCL3L1/CCL3L3, CD14, PIGR, SPP1, TREM2 5 phagocytosis by macrophages 3.68E-04 CD14, CH25H, MARCO, TREM2 4 immune response of antigen presenting 4.12E-04cells CD14, CD68, CH25H, MARCO, TREM2 5 release of hydrogen peroxide 4.96E-04 CXCL3, PPBP 2 proliferation of hematopoietic cells 5.16E-04 CCL3L1/CCL3L3, CSF2RB, CXCL2, CXCL3, INHBA, SPP1 6 production of reactive oxygen species 5.38E-04 CD14, CYBB, F7, ITGAX, MSR1, NOXO1, SPP1 7 CCL3L1/CCL3L3, Ccl6, CXCL3, INHBA, ITGAX, MMP12, cell movement of mononuclear leukocytes5.48E-04 8 RGS16, SPP1 binding of macrophages 6.06E-04 CD14, MARCO, MSR1 3 function of granulocytes 6.36E-04 CD14, CSF2RB, CXCL6, CYBB 4 synthesis of reactive oxygen species 6.57E-04 CD14, CYBB, F7, ITGAX, MSR1, NOXO1, OLR1, SPP1 8 fibrosis of lung 7.41E-04 CTSK, IL1RN, MMP12, PTGIR, RGS16 5 transmigration of leukocytes 7.85E-04 CCL3L1/CCL3L3, CXCL2, CXCL3, ITGAX 4 injury of lung 8.12E-04 CD14, MARCO, MMP12, OLR1 4 quantity of macrophages 9.33E-04 CD200R1, FABP1, LCN2, MMP12, SPP1 5 stimulation of cells 9.61E-04 CXCL3, IL1RN, INHBA, LCN2, PPBP, SPP1 6

350

Figure 1.

351

Figure 2.

BALB/c C57Bl

0.5 0.5 *** *** *** 0.4 0.4 ** 0.3 0.3

0.2 0.2

0.1 0.1 MSR1(relative expression) MSR1(relative expression) 0.0 0.0 4 6 8 12 weeks post smoking

2.0 4 *** *** 1.5 3 *** *** *** 2 1.0 ***

0.5 1

0.0 0 MARCO(relative expression) MARCO(relative expression) 4 6 8 12 weeks post smoking

0.3 0.3 **** ** **** ** ** 0.2 0.2

0.1 0.1

CXCL1 (relative expression) (relative CXCL1 0.0

CXCL1 (relative expression) (relative CXCL1 0.0 4 6 8 12

weeks post smoking

0.3 0.5 **** *** 0.4 ***

0.2 ** 0.3

0.2 0.1 ** 0.1 CXCL5 (relative expression) (relative CXCL5

CXCL5 (relative expression) (relative CXCL5 0.0 0.0 4 6 8 12 weeks post smoking 2.5 2.5 *** ** 2.0 2.0

1.5 **** **** ** 1.5

1.0 1.0

0.5 0.5

0.0 MMP12(relative expression) 0.0

MMP12(relevative expression) 4 6 8 12 weeks post smoking

352

Figure 3.

0.0006 **

0.0004

0.0002

CXCL1 (relative expression) (relative CXCL1 0.0000

0.010 * 0.008

0.006

0.004

0.002

CXCL5 (relative expression) (relative CXCL5 0.000

0.004 * 0.003

0.002

0.001

MMP12(relative expression) 0.000

353

Figure 4.

Human (127)

34 0 15

11 1 1

Mouse analysis 1 (61) Mouse analysis 2 (17)

Figure 5.

A" B" C"

354

-. --- .-- / .. / .- -- / ..-. .-. . .

355