Differential Profiles of Mammalian and Mosquito Derived Ross River Virus

Home , Ross River virus

Differential profiles of mammalian and mosquito derived Ross River virus

Jade Redfern BForSc, BAppSc (Hons).

Faculty of Applied Science

University of Canberra,

Bruce, Canberra, ACT

Australia

A dissertation submitted as fulfilment of the requirements for the award of a Doctor of Philosophy in Applied Science.

April 2015

Abstract

Ross River virus (RRV) is a mosquito borne virus that results in polyarthritic disease with symptoms including rash, lethargy and myalgia.

Currently, RRV is responsible for around 5000 infections each year in

Australia and disease symptoms can persist for several months or years in

RRV infected patients. Early interaction of the virus with the host immune system has been identified as critical to the resulting disease. In addition, it has been shown that there are differences between mammalian and mosquito derived virus in the N-linked glycans on envelope proteins, the ability of the virus to infect cells in vitro and the resulting cytokine expression. However, the mechanism behind these differences and their effect in vivo is yet to be elucidated. The work reported in this thesis aimed to test the hypothesis that mammalian and mosquito derived RRV differ in replication fitness in vitro and in vivo, and are associated with differences in host response and the subsequent clinical disease.

RRV-T48, the common laboratory strain of RRV, showed differences in replication in vitro. Mammalian derived RRV T48 (Mam-RRV-T48) showed a replication advantage in mammalian (Vero) cells and a disadvantage in mosquito (C6/36) cells. In contrast, mosquito derived RRV T48 (Mos-RRV-

T48) demonstrated a replication advantage in C6/36 cells and a disadvantage in Vero cells. This is indicative of RRVs ability to adapt to different host cells and increase the infectivity in the host within one passage. Furthermore, Mos-

RRV-T48 showed a replication advantage in Raw 264.7 cells. Previous studies have also showed that mosquito derived virus had increased binding affinity

iii with cell surface receptors such as DC-SIGN, resulting in an increased infectivity. This is likely due to high mannose N-linked glycans present on mosquito derived virus only. Glycosylation differences have been found to result in strong IFN induction for mammalian but not mosquito derived virus in DCs. We confirmed this occurs in RRV-T48 infection of Raw 264.7 and

Jaws II cells.

The RRV-T48 studies were expanded further utilising RRV field isolates extracted from mosquitos in Western Australia. The aim of this study was to examine if the differential effect observed with RRV-T48 could also be observed using wild type circulating RRV isolates in vitro. The study confirmed that RRV isolates mimicked the same pattern of infection as seen in RRV-T48 study. Mam-RRV-isolates produced the highest titres in Vero cells while Mos-

RRV isolates produced the highest titres in C6/36 cells. Furthermore, approximately half of all Mam-RRV isolates replicated to higher titres than their Mos-RRV counterparts in Raw 264.7 cells; indicating that binding affinity is not the only factor involved in RRV replication.

Subsequent in vivo studies showed a significant difference between

Mam-RRV-T48 and Mos-RRV-T48. The in vivo pilot study also revealed that disease progression was influenced by the gender of mice. While female mice showed almost no difference in disease when inoculated with Mam-RRV-T48 or Mos-RRV-T48, disease measures in male mice were significantly more severe. This has led to the development of a male only mouse model of RRV disease which provides more robust and reproducible results.

In the in vivo model, Mos-RRV-T48 infection induced less weight gain and higher clinical scores compared to Mam-RRV-T48 inoculated mice.

Additionally, Mos-RRV-T48 infection resulted in more inflammatory infiltrates, greater tissue and bone destruction present in quadriceps and ankle samples compared to Mam-RRV-T48 inoculated mice. IFNβ induction was upregulated in Mam-RRV-T48 infected mice at peak disease. Isolates

RRV-M and RRV-R showed similar disease patterns as seen with RRV-T48.

This demonstrates that the differential induction of disease by Mam-RRV and

Mos-RRV also occurs for wild type circulating isolates.

In addition, Barmah Forest and Dengue viruses also showed differential profiles between mammalian and mosquito derived viruses in vitro, similar to

RRV-T48 findings. This suggests that the observations in RRV infection may be more widely applicable to other arboviruses and warrants investigation.

In conclusion, the cell culture findings in this thesis suggest that arboviruses should be cultured in mosquito cell lines, ensuring that the data generated is related to natural infection and host response. Furthermore, one of the most pivotal outcomes of this thesis is in the refinement of the current

RRV mouse model which now takes into account gender dimorphism. The improvement of the RRV mouse model is an important step forward in more reproducible arbovirus research.

Acknowledgements

I would like to thank Dr Suresh Mahalingham who started me on my doctoral journey and for making it possible for me to pursue my love of science. However, I would not have been able to complete this journey if it had not been for my primary supervisor Dr Reena Ghildyal. Her patience, encouragement, and support has made all the difference, and I thank her for her wisdom of knowing when I needed pushing and when I needed space. I would like to thank my other supervisor Dr Michelle Nelson who has stuck with me throughout the candidature and supplied hours of experimental and editorial advice.

I would also like to thank all of the general staff and students at the

University of Canberra, you have all helped me at one point or another whether technically or emotionally and I am thankful for being surrounded by such wonderful people. Particular credit goes to Jacqui Richards for her hours of work in the animal house.

To Dijana Townsend, my weathered war veteran who undertook this journey with me. There are no words to express my gratitude to you for all the times you have picked me up and helped me move forward. We have cried together and laughed together and our friendship means more to me than you will ever know. You lent me your strength, Thank you. I would also like to thank Thomas Townsend1, for his amazing patience and assistance with my formatting faux pas. Also for his willingness to volunteer his time, technical

1 All images, unless otherwise referenced, were created by Thomas Townsend under instruction from the candidate – Jade Redfern. vii expertise, and graphic design genius which allowed my sanity to remain intact.

I would especially like to thank my mother, father, family and friends.

I thank them for their understanding and patience through this process. For letting me off on all the visits that I missed and things I forgot over the past few years. I love you all. To my husband, Hamish Podger, for his encouragement and support. It is rare to find someone who will give you the freedom to follow your crazy dreams. Thank you my love. And lastly to my daughter, Saige Podger. You are my wings, my inspiration to be a better person every day and I hope I make you proud. I love you with all my heart and soul.

viii

Table of Contents

Arboviruses ...... 1 Mosquito-borne viruses of medical importance ...... 2 Taxonomy ...... 3 Togaviridae ...... 4 Alphavirus Virion and genome ...... 7 Alphavirus replication...... 7 RRV epidemiology and disease ...... 14 Flaviviridae ...... 17 Bunyaviridae ...... 22 Arbovirus immunobiology and pathogenesis ...... 25 Cells targeted by arboviruses in infection and disease ...... 25 Attachment and entry of arboviruses ...... 27 Early events in host response to arboviral infection ...... 28 Cytokines ...... 35 Type I interferons (IFN) and arbovirus infection ...... 39 Arboviruses derived from mammalian versus mosquito cells .. 43 Glycosylation differences between mammalian and mosquito derived virus ...... 44 Entry/attachment differences between mammalian and mosquito derived virus ...... 47 Differential immune responses to mammalian and mosquito derived virus ...... 48

Mouse model of RRV...... 49 Thesis objectives ...... 51

General reagents ...... 53 Cell lines and virus strains ...... 54 Cell banking and culture ...... 55 Virus propagation and purification ...... 57 RRV propagation and purification ...... 61 Barmah forest virus (BFV) ...... 61 Dengue virus (DenV) ...... 61 Virus titration by plaque assay (RRV and BFV) ...... 62 DenV titration by immuno-focus assay ...... 62 Purification of RRV and BFV ...... 64 Purification of DenV ...... 64 Virus kinetics studies ...... 65 Infection of adherent cells ...... 65 Infection of non-adherent cells...... 66 Mouse Studies ...... 67 Mouse strains and breeding ...... 67 Mouse model of disease...... 67 Expression studies ...... 69 RNA extraction from cultured cells ...... 69 RNA extraction from mouse tissue ...... 69 Preparation of complementary DNA (cDNA)...... 70 Real time PCR ...... 71 Statistical analysis ...... 73

Introduction ...... 75 Results ...... 77 Plaque morphology of Mam-RRV-T48 and Mos-RRV-T48 ...... 77 Multistep replication kinetics of Mam-RRV-T48 and Mos-RRV- T48 in Vero, C6/36, Raw 264.7 and Jaws II cells ...... 78

xii

Induction of IFNβ by Mam-RRV-T48 and Mos-RRV-T48 in Raw 264.7 and Jaws II cells ...... 81 Discussion ...... 83 Differential plaque morphology of mammalian and mosquito derived RRV ...... 83 Differential replication kinetics of mammalian and mosquito derived RRV in Vero and C6/36 cells ...... 84 Differential replication kinetics of mammalian and mosquito derived RRV in Raw 264.7 and Jaws II cells ...... 85 Differential induction of IFNβ by mammalian and mosquito derived RRV in Raw 264.7 and Jaws II cells ...... 87 Conclusion ...... 88

Introduction ...... 91 Results ...... 93 Multistep replication kinetics of Mam-RRV and Mos-RRV field isolates in Vero cells ...... 93 Mam-RRV and Mos-RRV field isolate trends in Vero cells ...... 94 Multistep replication kinetics of Mam-RRV and Mos-RRV field isolates in a mosquito cell line ...... 99 Mam-RRV and Mos-RRV field isolate trends in a mosquito cell line...... 100 Multistep replication kinetics of Mam-RRV and Mos-RRV field isolates in Raw 264.7 cells ...... 105 Mam-RRV and Mos-RRV field isolate trends in Raw 264.7 cells ...... 106 Discussion ...... 111 Differential replication kinetics in Vero and C6/36 cells ...... 111 Differential replication kinetics in Raw 264.7 cells...... 112 Conclusion ...... 114

Introduction ...... 115 Results ...... 119 Disease profile of IFNAR -/- mice in RRV infection ...... 119

xiii

Inactivated RRV in a mouse model of disease ...... 124 Mam-RRV-T48 and Mos-RRV-T48 in a mouse model of disease ...... 125 Cytokine expression of active and inactive Mam-RRV-T48 or Mos- RRV-T48 in a mouse model of disease...... 127 Comparison of gender in mouse model of RRV infection ...... 136 Viral titres in mice inoculated with Mam-RRV-T48 or Mos-RRV- T48 in mice...... 140 Histology of mice infected with Mam-RRV-T48 or Mos-RRV- T48 ...... 144 Cytokine profiles of mice inoculated with Mam-RRV-T48 or Mos- RRV-T48 ...... 154 Discussion ...... 164 IFNβ is critical for development of disease symptoms...... 164 UV inactivated virus does not cause arthritic disease symptoms in the RRV mouse model of disease...... 166 Mosquito derived RRV causes a decrease in percentage weight gain and an increase in clinical score in the RRV mouse model of disease...... 167 Cytokine profiles of RRV disease depend on viral cell origin and differ from inactivated virus...... 167 Results gained from the RRV mouse model are sexually dimorphic...... 169 Male mice show greater reproducible variation when inoculated with Mam-RRV or Mos-RRV...... 170 Viral titres are higher in quadriceps muscle and ankle joint of Mos-RRV-T48 inoculated mice at peak disease ...... 170 Mice have different histology when inoculated with Mam-RRV or Mos-RRV ...... 171 Mice have different cytokine profiles when inoculated with Mam- RRV or Mos-RRV ...... 172 Conclusion ...... 173

Introduction ...... 175 Results ...... 177

xiv

Weight gain in mice inoculated with Mam-RRV-T48 or Mos-RRV- T48 field isolates ...... 177 Clinical scores of mice inoculated with Mam-RRV or Mos-RRV field isolates...... 182 Disease progression in mice inoculated with Mam-RRV or Mos- RRV field isolates ...... 186 Viral titres in mice inoculated with Mam-RRV or Mos-RRV field isolates ...... 190 Quadriceps muscle histology of mice inoculated with Mam-RRV or Mos-RRV field isolates at 24 hours post infection ...... 194 Ankle joint histology of mice inoculated with Mam-RRV or Mos- RRV field isolates at 24 hours post infection ...... 198 Brain histology of mice inoculated with Mam-RRV and Mos-RRV field isolates at 24 hours post infection ...... 202 Quadriceps muscle histology of mice inoculated with Mam-RRV and Mos-RRV field isolates at peak disease...... 204 Ankle joint histology of mice inoculated with Mam-RRV or Mos- RRV field isolates at peak disease ...... 208 Brain histology of mice inoculated with Mam-RRV or Mos- RRV at peak disease ...... 213 Cytokine profiles of mice inoculated with Mam-RRV and Mos- RRV field isolates ...... 215 Discussion ...... 228 Weight gain and clinical score trends in mice inoculated with mammalian and mosquito derived RRV isolates...... 229 Investigation of differential viral titres in tissue of mice inoculated with RRV field isolates...... 231 Differential tissue pathology in mice inoculated with RRV field isolates at 24 hours post infection ...... 232 Differential tissue pathology in tissue of mice inoculated with RRV field isolates at peak disease post infection ...... 234 Cytokine profiles in quadriceps muscle of mice inoculated with RRV field isolates ...... 236 Cytokine profiles in ankle joints of mice inoculated with RRV field isolates ...... 238 Conclusion ...... 239

Introduction ...... 241 Results ...... 243 Plaque morphology of Mam-BFV and Mos-BFV ...... 243 Multistep replication kinetics of Mam-BFV and Mos-BFV in Raw 264.7 cells ...... 243 Differential IFNβ induction by mammalian and mosquito derived Barmah Forest virus in Raw 264.7 cells ...... 244 Plaque morphology of Mam-DenV and Mos-DenV ...... 246 Multistep replication kinetics of Mam-DenV and Mos-DenV in Jaws II cells ...... 246 Differential IFNβ induction by DenV in Jaws II cells ...... 247 Discussion ...... 248 Differential plaque phenotypes of arboviruses ...... 249 Differential growth kinetics of arboviruses in immune cell lines ...... 250 IFNβ induction by mammalian and mosquito derived arboviruses ...... 251 Conclusion ...... 252

Arobovirus growth kinetics in tissue culture ...... 255 Gender dimorphism in the RRV mouse model of disease ...... 259 Mammalian and mosquito derived RRV results in differential disease profiles in a mouse model of disease ...... 261 Trends in infection and disease caused by RRV field isolates in a male mouse model ...... 262 Inflammatory responses to RRV in vivo ...... 268 Conclusions and future directions ...... 273

xvi

List of Figures

Figure 1.1 Schematic representation of alphavirus virion and genome. .... 6

Figure 1.2 Alphavirus replication cycle ...... 9

Figure 1.3 Relationship between rainfall, mosquito density and RRV incidence in Brisbane ...... 15

Figure 1.4 Schematic representation of flavivirus virion and genome. .... 19

Figure 1.5 Flavivirus replication cycle ...... 20

Figure 1.6 Schematic representation of Bunyaviridae virion and genome...... 24

Figure 1.7 Comparative overview of mammalian and mosquito N-glycan subtypes ...... 46

Figure 3.1 Differential plaque morphology of Mam-RRV-T48 and Mos-RRV- T48 ...... 78

Figure 3.2 Growth kinetics of mammalian and mosquito derived RRV in various cell lines ...... 80

Figure 3.3 IFNβ induction by Mam-RRV-T48 and Mos-RRV-T48 ...... 82

Figure 4.1 RRV field isolate multistep kinetics in mammalian cells ...... 97

Figure 4.2 RRV field isolate growth kinetic trends in mammalian cells... 98

Figure 4.3 RRV field isolate multistep kinetics in mosquito cells ...... 103

Figure 4.4 RRV field isolate growth kinetic trends in mosquito cells ..... 104

Figure 4.5 RRV field isolate multistep kinetics in mammalian macrophage cells ...... 109

Figure 4.6 RRV field isolate growth kinetic trends in mammalian macrophages ...... 110

Figure 5.1 Type I interferon is important in development of clinical disease in mouse model of Ross river virus infection ...... 123

Figure 5.2 Disease progression in mice inoculated with RRV ...... 126

Figure 5.3 Inflammatory cytokine profiles differ between mammalian and mosquito derived RRV and also between active and inactive RRV in the quadriceps muscle...... 131

xvii

Figure 5.4 Inflammatory cytokine profiles differ between mammalian and mosquito derived RRV and also between active and inactive RRV in the ankle joint...... 135

Figure 5.5 Male C57Bl/6 mice infected with mosquito derived T48 RRV lose more weight and have higher clinical scores than female C57Bl/6 mice...... 138

Figure 5.6 Difference in disease and weights of male C57Bl/6 mice inoculated with T48 RRV is highly reproducible...... 139

Figure 5.7 Blood and tissue virus titres...... 143

Figure 5.8 Histology of quadriceps muscle in mice infected with RRV at 24 hours post infection...... 145

Figure 5.9 Histology of quadriceps muscle in mice infected with RRV at peak disease...... 146

Figure 5.10 Histology of ankle joint in mice infected with RRV at 24 hours post infection...... 149

Figure 5.11 Histology of ankle joint in mice infected with RRV at peak disease...... 151

Figure 5.12 Histology of brain tissue in mice infected with RRV at 24 hours and peak disease...... 153

Figure 5.13 Cytokine profiles in the quadriceps muscle of mice inoculated with mammalian and mosquito derived RRV...... 158

Figure 5.14 Cytokine profiles in the ankle joint of mice inoculated with mammalian and mosquito derived RRV...... 163

Figure 6.1 Weight gain of C57Bl/6J mice inoculated with Mam-RRV and Mos-RRV field isolates...... 181

Figure 6.2 Clinical score of C57Bl/6J mice inoculated with Mam-RRV and Mos-RRV isolates...... 185

Figure 6.3 Disease profile trends for mice inoculated with Mam-RRV and Mos-RRV...... 189

Figure 6.4 Viral titres recovered from blood, quadriceps muscle, ankle joint and brain of mice inoculated with Mam-RRV and Mos-RRV field isolates...... 193

Figure 6.5 Histology of quadriceps muscle in mice infected with RRV at 24 hours post infection ...... 197

xviii

Figure 6.6 Histology of ankle joint of mice infected with RRV at 24 hours post infection ...... 201

Figure 6.7 Histology of brain tissue in mice infected with RRV at 24 hours post infection ...... 203

Figure 6.8 Histology of quadriceps muscle in mice infected with RRV at peak disease ...... 207

Figure 6.9 Histology of ankle joint in mice infected with RRV at peak disease ...... 212

Figure 6.10 Histology of brain tissue in mice infected with RRV at peak disease ...... 214

Figure 6.11 Cytokine profiles in the quadriceps muscle of mice inoculated with mammalian and mosquito derived RRV field isolates. .. 220

Figure 6.12 Cytokine profiles in the ankle joint of mice inoculated with mammalian and mosquito derived RRV field isolates...... 226

Figure 7.1 Differential plaque morphology of mammalian and mosquito derived Barmah Forest virus ...... 243

Figure 7.2 Growth kinetics of mammalian and mosquito derived Barmah forest virus in Raw 264.7 cells...... 244

Figure 7.3 IFNβ induction by mammalian and mosquito derived Barmah forest virus ...... 245

Figure 7.4 Differential plaque morphology of mammalian and mosquito derived dengue virus ...... 246

Figure 7.5 Growth kinetics of mammalian and mosquito derived dengue virus in various Jaws II cells ...... 247

Figure 7.6 IFNβ induction of mammalian and mosquito derived dengue virus ...... 248

xix

List of Tables

Table 1.1 General information on arboviruses ...... 4

Table 1.2 Notifications of Ross River virus infection ...... 7

Table 1.3 Mosquito-borne viruses in the Flaviviridae family that cause human disease ...... 17

Table 1.4 Dengue disease grade classifications ...... 21

Table 2.1 General reagents and their preparation...... 53

Table 2.2 Growth media...... 54

Table 2.3 Cell lines and their normal growth conditions...... 55

Table 2.4 In-vitro transcription reaction mix ...... 59

Table 2.5 RRV field isolates ...... 60

Table 2.6 Clinical score of RRV disease in mice ...... 68

Table 2.7 Reverse transcriptase master mix reagent ...... 70

Table 2.8 Real-time PCR master mix reagent ...... 71

Table 2.9 Real time PCR parameters ...... 72

Table 8.1 Correlation between in vivo and in vitro profiles of RRV isolates ...... 263

xxi

List of Abbreviations

ADE Antibody-dependant enhancement ATF Activation Transcription Factor BFV Barmah Forest Virus BHK Baby Hamster Kidney Cells CP Capsid Protein C3 Complement Component 3 CDC Center for Disease Control cDNA Complementary DNA ChikV Chikungunya Virus CR3 Complement Receptor 3 DC Dendritic Cells DC-SIGN Dendritic Cell Specific Intercellular Adhesion Molecule-3- Grabbing Non-Integrin

DenV Dengue Virus DF Dengue Fever DHF Dengue Haemorrhagic Fever DSS Dengue Shock Syndrome E Envelope Protein EEEV Eastern Equine Encephalitis Virus GAS INF ɣ Activated Site GM-CSF Granulocyte Macrophage Colony Stimulating Factor HMG High Mobility Group Protein HS Heparin Sulfate IFN Interferon IFNAR Interferon Alpha Receptor IKK Inhibitor of Kappa Beta Kinase IL Interleukin IP Interferon ɣ Induced Protein

xxiii

IRF Interferon Regulator Factor ISGF IFN-Stimulated Gene Factor ISRE IFN-Stimulated Response Element Jak-STAT Janus Kinase – Signal Transducer and Activator of Transcription

JEV Japanese Encephalitis kb Kilobases KV Kunjin Virus LPS Lipopolysaccharide L-SIGN Liver/Lymph Node Intercellular Adhesion Molecule-3- Grabbing Non-Integrin

Mam Mammalian Derived MBL Mannose Binding Lectin MCP/CCL2 Monocyte Chemotactic Protein MHC Major Histocompatibility Complex MIF Macrophage Inhibitory Factor MIG Macrophage Induced Gene MIP Macrophage Inhibitor Complex Mos Mosquito Cell Derived mRNA Messenger RNA MVE Murray Valley Encephalitis NF-KB Nuclear Factor Kappa B NK Natural Killer cells NO Nitric Oxide NsP Non-Structural Protein ORF Open Reading Frame PCR Polymerase Chain Reaction PDC Plasmacytoid Dendritic Cells PKR Serine Threonine Protein Kinase R Ra Receptor Antagonist RdRp RNA-dependant RNA Polymerase

xxiv

RRV-T48 Ross River Virus T48 RVFV Rift Valley Fever SFV Semliki Forest Virus SinV Sinbis Virus SIRS Systemic Inflammatory Response Syndrome TBK Threonine Protein Kinase TC T Lymphocyte Cytotoxic Cells CD8+ TGF Transforming Growth Factor TH T Lymphocyte Helper Cells CD4+ TLR Toll Like Receptor TM T Lymphocyte Memory Cell TNF Tumour Necrosis Factor VEEV Venezuelan equine encephalitis virus WHO World Health Organisation WNV West Nile Virus Wt Wild Type YF Yellow Fever

xxv

Introduction

Arboviruses

Arboviruses (arthropod-borne viruses) are transmitted through an arthropod vector and are maintained in nature through a continuous cycle of transmission between a blood feeding arthropod vector and susceptible vertebrate host (Mellor, 2000). Arthropods that are able to function as vectors for viral diseases include ticks and biting flies, such as mosquitos (Lupi,

2011). Arboviruses cause significant clinical disease manifestations in both humans and animals. They are a growing medical concern, especially as global temperatures rise and the vectors are able to inhabit more areas; therefore increasing the spread of disease. Over 550 species of arboviruses were listed in the international catalogue of arboviruses held by the Centers for Disease Control and Prevention (CDC) in 1985 and more have emerged since (Liu et al., 2011a). Arboviruses are predominantly within the virus families Togaviridae, Flaviviridae, Bunyaviridae, Rhabdoviridae and

Reoviridae; however they also include one genus in the Orthomyxoviridae family and one virus in the Asfarviridae family (Ciota and Kramer, 2010).

Arboviruses are ribonucleic acid (RNA) viruses with the singular exception of

African swine fever virus which is a deoxyribonucleic acid (DNA) virus (Shope,

1994). An important feature of RNA viruses is their high mutation rate which can make it difficult to create effective vaccines. This is a concern, as arboviral infections are responsible for a significant burden to the public health system worldwide. There are over 100 arbovirus species that can cause

1 human disease with the majority being transmitted by a mosquito vector

(Ciota and Kramer, 2010).

Mosquito-borne viruses of medical importance

As far back as 100AD there are reports of fevers being spread by biting flies, however, despite suspicions that mosquitos were the carriers for yellow fever (YFV) in 1854, the first arbovirus was not identified until 1901 (Calisher,

2005). In 1930, only six arboviruses had been identified. A surge in advancement of viral isolation techniques at that time lead to a number of viruses being identified and isolated in quick succession; jumping to 16 arboviruses by 1940, 35 by 1950, 144 by 1960, 354 by 1970 and 450 arboviruses by 1980 (Chamberlain, 1982). More and more arboviruses continue to emerge, many of which cause significant human disease and use the mosquito as the primary vector.

Mosquito-borne arboviruses are prevalent around the world in locations where conditions allow for the breeding and existence of the vector species.

In addition to this, disease outbreaks are usually seasonal and depend on factors such as temperature, level of breeding and when the mosquitoes are taking a blood meal (Kilpatrick and Pape, 2013; Shope, 1994; Waldock et al.,

2013). It has been suggested that the spread of arthropod-borne infections is increasing due to an increase in human travel and changes in environmental conditions that favour vector survival, such as global warming and urban living (Russell, 1998).

Geographical distribution of arboviruses is highly reliant on the viability of susceptible vector populations in the area. Worldwide, Japanese encephalitis virus (JEV), Dengue virus (DenV), West Nile virus (WNV) and JF have been identified as having a major impact on health. In Australia the arboviruses of concern are DenV, Ross River virus (RRV), Barmah Forest

Virus (BFV), JEV, Murray Valley Encephalitis virus (MVE), Chikungunya virus

(ChikV), Kunjin virus (KV) and Rift Valley fever virus (RVFV) (Department of

Health, 2010).

Taxonomy

Mosquito borne viruses of medical importance predominantly originate from three families, Togaviridae, Flaviviridae and Bunyaviridae. Viruses from the families Togaviridae and Flaviviridae involve a mosquito vector and the ability to utilise humans as the host (Table 1.1). Viruses from the

Bunyaviridae family infect vertebrate hosts other than humans as their primary host, with occasional zoonosis resulting in human infections. La

Crosse virus is perhaps one of the most significant and has been identified as the major cause of viral encephalitis in America, with 80-100 cases per year

(CDC, 2011).

Table 1.1 - General information on arboviruses

Family Togaviridae Flaviviridae Bunyaviridae

Single stranded Single stranded Single stranded Genome RNA RNA RNA +/- strand Positive Positive Negative

Envelope Yes Yes Yes Spherical (65-70 Spherical (40-64 Spherical (90-100 Virion morphology nm) nm) nm) Nucleocapsid Icosahedral Spherical Helical filaments morphology Genome Linear 10-12 kb Linear 10-12 kb Linear 11-19 kb Baltimore IV IV V classification Genus Alphavirus Flavivirus Hantavirus

Rubivirus Hepacivirus Nairovirus

Hepatitis G Virus Othobunyavirus

Pestivirus Phlebovirus

Tospovirus

Units: nanometers (nm), Kilobases (kb)

Togaviridae

The Togaviridae family consists of two genera, Alphavirus and

Rubivirus, the latter of which contains only rubella virus, which is not transmitted by a vector. In contrast, the Alphavirus genus consists of 40 viruses, most of which are transmitted by an arthropod vector. Alphaviruses have single stranded, non-segmented, positive sense RNA genomes that are

11-12 kilobases (kb) long (Figure 1.1). The genome contains two open reading frames (ORFs) which separate the translation of the structural and nonstructural proteins (NsP). The Alphavirus genome codes for four structural

4 proteins; the capsid protein (C) and three envelope glycoprotein (E1, E2 and

E3) as well as a small protein known as 6K whose role is not yet fully elucidated (Firth et al., 2008). Four NsPs are encoded in the 5’ section of the genome (NsP1-4) and constitute two thirds of the coding region (Strauss and

Strauss, 1994; Vaney et al., 2013).

Alphaviruses have been traditionally further broken down into New

World viruses and Old World viruses. New World viruses result in acute encephalitic type diseases, whilst Old World viruses result in chronic arthritic disease. There is some overlap of these definitions, as three Old World viruses, Sindbis (SinV), RRV and ChikV, primarily cause an arthritic disease but have been shown to be neuro-invasive and have the potential to cause encephalitic disease (Zacks and Paessler, 2010).

Alphaviruses are generally referred to as enzootic infections, as humans are not the preferred host for these viruses despite being susceptible to infection. They are maintained and amplified through a mosquito vector and a mammalian or marsupial vertebrate host (Atkins, 2013). Of all the

Alphaviruses, RRV is the most reported arboviral infection in Australia and is of significant concern to human health (Table 1.1).

Figure 1.1 - Schematic representation of alphavirus virion and genome.

Representation of a cross section of the Alphavirus virion showing the envelope glycoproteins E1 and E2, lipid bilayer, the capsid protein (CP). The Alphavirus RNA genome is a single molecule of plus strand RNA and 9-12 kilobases in length. The non-structural proteins are translated from the genome RNA whilst the structural proteins are translated from sub- genomic mRNA into two polyproteins which are then processed by host and viral proteases to produce proteins (Firth et al., 2008; Strauss et al., 1988).

Table 1.2 - Notifications of Ross River virus infection Year ACT NSW NT QLD SA TAS VIC WA Aust. 2008 21 1153 263 2843 197 77 231 863 5648 2009 4 909 431 2146 328 29 92 837 4776 2010 22 1083 336 2379 451 39 422 409 5141 2011 8 577 184 1220 979 7 1312 871 5158 2012 11 607 227 1944 221 18 281 1378 4687 2013 4 504 300 1787 167 8 170 1368 4308

Notifications of RRV infection data received from State and Territory health authorities for 2008 to 2013 (Department of Health, 2015)., Australian Capital Territory (ACT), New South Wales (NSW), Northern Territory (NT), Queensland (QLD), South Australia (SA), Tasmania (TAS), Victoria (VIC), Western Australia (WA).

Alphavirus Virion and genome

The genome of the T48 strain of RRV, commonly used in laboratory research, is 11854 nucleotides long (Genebank). Structural and Ns proteins are encoded in two distinct regions with separate ORFs. Two glycosylated proteins (E1 and E2) found in the envelope and the icosahedral nucleocapsid contained by it, make up the four commonly recognised structural proteins along with E3 which is although required for spike assembly, is not incorporated in the virion (Cheng et al., 1995). The 6K protein has been associated with formation of ion channels and is also encoded in the structural section of the genome.

Alphavirus replication

The spherical, enveloped virion is approximately 71 nanometres (nm) in diameter with a positive sense, single stranded, non-segmented RNA genome (Strauss and Strauss, 1994). Many studies examining replication

7 and protein formation of Alphaviruses have used SinV as a representative virus and thus much of the knowledge around the RRV genome, protein functions and viral replication has been inferred from studies using SinV, the

Alphavirus replication cycle is illustrated in Figure 1.2.

Protruding portions of the viral envelope glycoproteins, E2, as well as

E1 interact with host cell surface receptors and mediate membrane fusion

(Snyder and Mukhopadhyay, 2012). E2 is highly conserved among

Alphaviruses, but interestingly, a single point mutation in E2 in SinV virus resulted in a different effect on spike formation and viral assembly than in

RRV indicating that E2 folding is species specific (Snyder et al., 2012). Recent studies in RRV have indicated the same importance of E2 in virulence. Jupille et al. (2013) found that a single amino acid substitution at position 18 on the

RRV E2 glycoprotein resulted in attenuation of disease in mice, with less virus spread and a significant reduction in viremia. In vitro it was shown that this mutant also replicated more efficiently than RRV-T48 in mosquito cells but displayed lower titres in mammalian cells. It was proposed that amino acid switching at this position functions to regulate fitness in either vertebrate or invertebrate hosts. It has also been found that amino acid 218 of RRV can be modified to create heparan sulfate (HS) binding site and that this leads to an expansion of the host range of RRV to include avian cells (Heil et al., 2001).

Other studies involving the E2 glycoprotein of SinV found that deletions at positions 391 disrupted virus assembly in mammalian cells but did not have the same effect in insect cells (Hernandez et al., 2005).

Figure 1.2 - Alphavirus replication cycle

The Alphavirus virion enters the cell via receptor mediated endocytosis. The low pH of the endosome results in membrane fusion and allows the delivery of the viral genome into the host cell cytoplasm. Non-structural protein precursors are translated from the viral mRNA and processing by viral proteases results in the production of the non-structural proteins which leads to the synthesis of full length negative strand RNA intermediate. From this genomic and sub-genomic RNA is synthesised. Sub-genomic RNA results in the production of a polyprotein of which the capsid protein is the first to be cleaved. The remaining glycoproteins are further processed in the Golgi apparatus before undergoing further cleavage and being transported to the plasma membrane. The nucleocapsid, genomic RNA and glycoproteins bind together and bud from the host cell membrane as a mature virion (Schwartz and Albert, 2010).

E1 facilitates fusion of the viral envelope with a target membrane, additionally it generates an icosahedral lattice on the viral surface and forms ion permeable pores in the target membrane during virus entry (Lescar et al.,

2001; Wengler et al., 2003). The formation of ion pores occurs at low pH, in- vivo this happens in the low pH endosomes but the effect can also be induced in-vitro under low pH (White and Helenius, 1980). The formation of ion pores in the endosomal membrane is thought to facilitate core disassembly and translation of the viral genome. Rearrangement of proteins occurs in the acidic environment exposing the E1 fusion peptide. The E1 fusion peptide is then inserted into the host endosome membranes (Gibbons et al., 2003). It is noted that unlike infection in vertebrate cells, low pH may not be required in the infection of mosquito cells (Hernandez et al., 2001). Following fusion, the nucleocapsid core is release into the cytoplasm and RNA translation and transcription occur forming a genome plus strand RNA, complementary minus strand RNA and sub-genomic messenger RNA (mRNA).

Minus strand synthesis requires NsP1 which has been indicated in the initiation and maintenance of minus strand synthesis (Wang et al., 1991).

NsP1 has been identified as a critical determinant in mouse virulence where mutation can effect virulence and increase host type I interferon (IFN) response without effecting viral replication in both SinV and RRV (Cruz et al.,

2010). A role for NsP1 in mouse disease was further confirmed in a study which swapped the NsP1 regions of RRV-T48 and a mouse avirulent strain resulting in attenuation of RRV-T48 and restored virulence to the avirulent strain (Jupille et al., 2013). In ChikV infection, NsP1 is an antagonist to the

IFN stimulated gene, BST-2, that prevents infection of bystander cells by prevention of virion budding (Jones et al., 2013).

NsP2 plays a role in minus strand synthesis by blocking host cell translation and facilitation of unwinding of RNA during RNA replication and transcription (Peranen et al., 1990; Rikkonen, 1996; Rikkonen et al., 1994;

Sawicki et al., 2006; Tamm et al., 2008). It also cleaves the non-structural polyprotein (Hardy and Strauss, 1989).

NsP3 is critical for minus strand and sub-genomic RNA synthesis and has been shown to block the formation of stress granules as well as to interfere with host cell membrane trafficking, cell signalling and cytoskeletal organisation (Fros et al., 2012; Lemm et al., 1994; Neuvonen et al., 2011; Park and Griffin, 2009; Tuittila and Hinkkanen, 2003). It is important to note that

NsP3 has a different morphology and protein interaction depending on whether they are from New or Old World Alphavirus. It is believed, that despite these differences, their functions are likely similar (Foy et al., 2013a).

In Venezuelan equine encephalitis virus (VEEV) infection, it has been shown that NsP3 plays a critical role in virus replication of mosquito cells but not mammalian cells. It is unknown if this is also the case for Old World

Alphaviruses (Foy et al., 2013b). NsP4 is a RNA-dependent RNA polymerase

(RdRP) and is thought to interact with other NSP’s and may have a role in the regulation of host cell gene expression (Rupp et al., 2011).

The sub-genomic RNA is translated to produce structural polyprotein of which CP is the first protein to be cleaved, through autocatalytic cleavage

(Melancon and Garoff, 1987). The process occurs in the cytoplasm after which

11 it assembles into the nucleocapsid core and later associates with the mature glycoproteins at the plasma membrane (Forsell et al., 2000).

Meanwhile, the remaining polyprotein, containing E3+E2(PE2)-6K-E1 is translocated across the ER membrane where the polyprotein is cleaved into

E3-E2, 6K and E1 and the proteins undergo posttranslational modifications including the addition of high mannose chains on N-linked glycosylation sites.

Proteins then move through the Golgi network and the envelope protein spike complex matures before E3 is cleaved by furin and the E2-E1 spike complex is transported to the plasma membrane for assembly with the nucleocapsid.

Unlike the other envelope proteins, E3 is not incorporated into the virion.

Major roles of E3 include particle assembly, and mediation of spike folding and activation. E3 transports the structural polyprotein, once CP has been cleaved, to the ER. The polyprotein is then further cleaved to form E3-E2, often called pE2 (Zhang et al., 2003). Spike formation occurs in the ER and if E3 is not present correct folding of E2 does not occur and E2-E1 heterodimer spikes are not able to form (Snyder and Mukhopadhyay, 2012).

E3 must be cleaved from E2 by furin before the spikes become fusion competent and are transported through the host secretory system to the plasma membrane for budding (de Curtis and Simons, 1988; Sjoberg et al.,

2011). To highlight the importance that E3 plays in Alphavirus infectivity, it was found that mice immunised with E3 glycoprotein antibody were protected against lethal infection with VEEV (Parker et al., 2010). Additionally, studies indicate that alteration to the highly conserved cysteine residues in the glycoprotein highly alters infectivity (Parrott et al., 2009).

Virion budding occurs on the cell membrane surface as the nucleocapsid core and glycoproteins associate. This is facilitated by 6K which functions as a cation-selective ion channel in the lipid bilayer (Melton et al.,

2002). Complete deletion of the gene results in a virion that is indistinguishable to the wild type (Wt) virus but reduces the titre by 50 fold in baby hamster kidney (BHK) cells (Liljestrom et al., 1991). Loewy et al.

(1995) confirmed that 6 K plays a role in assembly and budding process. In conjunction with this, they also identified that lack of this protein makes

Semliki Forest virus (SFV) more thermolabile. Additionally, their studies showed that mosquito cells were more competent at growing the 6K deletion mutant than mammalian cells and that a large amount of infectious virus remained cell associated in both cell lines. When visualised on a protein gel,

6 K migrates as a doublet. Previously this has been explained as a result of varying degrees of acylation. Current research is investigating this doublet and has indicated that it is in fact due to frame shifting which results in the synthesis of an addition novel protein of 8 kDa, called the TransFrame (TF) protein. Whilst it used to be accepted that small amounts of 6 K were incorporated into the virion, these new studies suggest that it is actually TF that is included and not 6 K (Firth et al., 2008; Snyder et al., 2013).

Characterisation of this novel protein has only recently commenced, however it is indicated as having similar ion-channel activity to 6K. Additionally, TF was found to be not required for culture in cells but does alter the release of virus in both mammalian and insect cells. Furthermore, when TF was altered in SinV mice mortality fell to under 15 % compared to 95 % of mice infected with Wt virus (Snyder et al., 2013).

RRV epidemiology and disease

RRV was first isolated by Doherty et al. (1963) from Aedes vigilax mosquitoes, taken from an area near the Ross River in Townsville,

Queensland, Australia. Doherty further identified the virus as the causative agent of the epidemic polyarthritis that had been reported around the warmer states of Australia in 1972. RRV requires a mosquito vector as well as a mammalian/marsupial host reservoir and is maintained through a transmission cycle between the two. Marsupials have been identified as the most competent hosts and in the Northern Territory, the agile wallaby

(Macropus agilis) and the dusky rat (Rattus colletti) have been identified as potential major reservoirs of RRV. In addition, kangaroos, other wallabies and possums have also been identified as potential hosts (Harley et al., 2001;

Kay et al., 1982). Human infection by RRV is thought to be due to spill over from an epidemic in a mammalian reservoir population, however once this has happened human-mosquito-human transmission is thought to be the main cycle during a human epidemic (Carver et al., 2009). As RRV requires a mosquito vector, the disease is limited to areas in which susceptible mosquito species can survive. The most prominent mosquito vectors identified are

Ochlerotatus vigilax (formerly Aedes vigilax), a salt water mosquito and Culex annulirostris, a fresh water mosquito; however it is important to note that RRV has been associated with over 35 species of mosquitos (Jacups et al., 2008).

Hu et al. (2010) found that the incidence of human RRV infection was directly related to mosquito abundance, which was scored between 0-100 according to mosquito population density. There was an 89 % chance of an RRV outbreak if Oc. vigilax abundance was scored between 64 and 90 in coastland

Queensland and an 80 % chance of an RRV outbreak if Cx. annulirostris abundance was scored between 53 and 73 in inland Queensland.

Furthermore, pre-emptive mosquito control programs that include both species have been shown to result in decreased incidence of RRV disease

(Tomerini et al., 2011). Additionally, a clear relationship exists between an increase in mosquito population due to increased rainfall and increase in RRV infection Figure 1.3 (Tong et al., 2008). RRV is endemic in Australia, Papua

New Guinea, Fiji, Indonesia, Cook Islands, and Samoa.

Figure 1.3 - Relationship between rainfall, mosquito density and RRV incidence in Brisbane

Increased rainfalls in Brisbane correlate with an increase in moquito population peaks which in turn corresponds to a peak in RRV incidence within two months of peak mosqito populations. Graph used with permission from Tong et al., (2008).

Human infection by RRV occurs when an infected mosquito takes a blood meal from a person. Once infected, the incubation period averages eight

15 days with a range of 3-21 days (Harley et al., 2001). Early symptoms include non-itchy macropapular rash, headache, fever and lethargy, myalgia or arthritic symptoms can occur acutely or chronically in the disease. While joint pain is present in more than 95 % of patients and tiredness in more than

90 %, other symptoms such as rash, headache and fever only present in approximately half of all patients (Barber et al., 2009). Arthritic symptoms primarily affect the joints of the limbs, such as the carpals and metacarpals and can last between 3-12 months with associated pain ranging from moderate to severe. Acute RRV infection has been associated with a risk of developing post-infective syndrome and chronic fatigue syndrome, however this is thought to be due to the host response to disease and not virus specific

(Hickie et al., 2006). Meningitis and encephalitis have also only been reported a few times and are considered extremely rare (Harley et al., 2001).

Presently there is no vaccine for RRV disease and treatment of disease is aimed at alleviating symptoms. There is a promising vaccine candidate which has been shown to provide protection to mice from viremia and also eliminate mortality and disease in IFN α/β receptor (IFNAR) knock out (-/-) mice. Furthermore, no antibody dependent enhancement (ADE) was seen and antibodies were cross protective against infection with ChikV (Holzer et al.,

2011). Human trials have been promising with 92.9 % of participants seroconverted however up to 30 % of candidates were symptomatic (Aichinger et al., 2011). Until a vaccine is realised, the main treatment is use of non- steroidal anti-inflammatory drugs and paracetamol. It has been estimated

16 that the annual disease burden cost to Australia is at least US$10 million

(Jones et al., 2010)

Flaviviridae

The family Flaviviridae consists of a group of positive strand, enveloped

RNA viruses. This large family is further broken down into three genera:

Flavivirus, Pestivirus and Hepacivirus. Of these, only the genus Flavivirus contains mosquito borne viruses of human disease significance (Table 1.3).

Table 1.3 - Mosquito-borne viruses in the Flaviviridae family that cause human disease Virus Vector Primary Distribution Disease host Dengue 1 Mosquito Human World Wide DF, DHF, DSS Dengue 2 Mosquito Human World Wide DF, DHF, DSS Dengue 3 Mosquito Human World Wide DF, DHF, DSS Dengue 4 Mosquito Human World Wide DF, DHF, DSS Japanese Mosquito Bird/Pig Asia Encephalitis encephalitis Kunjin Mosquito Bird Australia Encephalitis Murray valley Mosquito Bird Australia Encephalitis encephalitis North Powassan Tick Rodent America/Europe Encephalitis

North St Louis America/Central Mosquito Bird Encephalitis encephalitis America/South America

West Nile Mosquito Bird Africa/Europe/Asia Encephalitis

Meningitis, hepatitis and Yellow fever Mosquito Monkey Africa/South America haemorrhagic fever

Modified from Monath and Heinz, (1996) with additional information from Munoz-Jordan et al. (2005). DF – Dengue Fever, DHF – Dengue Haemorrhagic Fever, DSS – Dengue Shock Syndrome

It is interesting to note that Flaviviruses were originally grouped with

Alphaviruses in the Togaviridae family based on similarities in their viral structure, however, distinct differences in genomes, proteins and their functions, as well as translational mechanisms led to the reclassification of

Flaviviruses into the Flaviviridae family (Strauss and Strauss, 2001).

Flaviviruses have a non-segmented single stranded positive sense RNA genome which is approximately 11 kb long and contains a single ORF. The genome encodes for three structural proteins: C, glycoprotein precursor of matrix (M) protein (prM) and E, and seven NsPs: NsP1, NsP2A, NsP2B, NsP3,

NsP4, NsP4B and NsP5 (Pierson and Diamond, 2012).

Although Flaviviruses can target many host cells, infection of dendritic cells (DC) has been found to be of particular importance. It has been shown that interaction with dendritic cell-specific intercellular adhesion molecule-3- grabbing non-integrin (DC-SIGN) and liver/lymph node-specific intercellular adhesion molecule-3-grabbing integrin (L-SIGN) is involved in the attachment of Flavivirus to DC’s (Davis et al., 2006; Martina et al., 2008; Martins Sde et al., 2012; Tassaneetrithep et al., 2003).

Once attached, the Flavivirus virion enters the host cell by clathrin- mediated endocytosis and the replication cycle is outlined in Figure 1.5 (Chu et al., 2006; Peng et al., 2009). The virus NsPs assist in the evasion of the immune response by inhibiting IFN induced signalling by interfering with both the janus kinase signal transducer and activator of transcription (Jak-STAT) pathway and through prevention of activation of interferon regulatory factor

(IRF)-3 resulting in a delay in IFNβ production and thus a down regulation of

18 host immune response (Best et al., 2005; Fredericksen and Gale, 2006; Lin et al., 2006).

Figure 1.4 - Schematic representation of flavivirus virion and genome.

Representation of a cross section of the Flavivirus virion showing the envelope glycoprotein homodimer (E), matrix protein (M), lipid bilayer, capsid protein (CP) and genomic RNA. The positive sense Flavivirus RNA genome is translated into a polyprotein which is then cleaved by host and viral proteases to produce indicated structural and non-structural proteins (NsP) (Chambers et al., 1990; Hernandez et al., 2014).

Figure 1.5 - Flavivirus replication cycle

The Flavivirus virion enters the cell via receptor mediated endocytosis. The low pH of the endosome results in membrane fusion and allows the delivery of the viral genome into the host cell cytoplasm. A single polyprotein is translated from the viral mRNA, processing by host and viral proteases results in the production of proteins and replication of the viral RNA occurs. Assembly of the virus occurs on the endoplasmic reticulum membranes. Virions bud from the ER and undergo further processing in the Golgi apparatus with virion maturation occurring when prM is cleaved by furin and the mature virion buds from the plasma membrane (Pierson and Diamond, 2012).

Dengue Virus (DenV)

Without a doubt, the virus in the Flavivirus genera that has had the most impact on human health is DenV. DenV is emerging as one of the world’s most important arboviruses with an estimated 2.5 billion people at risk of infection (WHO, 2001). DenV can cause three clinical conditions in humans: dengue fever (DF), dengue haemorrhagic fever (DHF) and dengue shock syndrome (DSS), with each increasing in severity (Alejandria, 2004). A disease classification system for these conditions has been devised by the

World Health Organisation (WHO), based on clinical presentations (Table 1.4)

Table 1.4 - Dengue disease grade classifications Disease Grade Symptoms Dengue Fever - Fever with two or more of the following signs: headache, retro-orbital pain, myalgia, arthralgia Dengue Haemorrhagic I Fever with above signs. The only Fever haemorrhagic manifestation is a positive tourniquet test and/or easy bruising Dengue Haemorrhagic II Above signs plus spontaneous bleeding Fever usually in the forms of skin or other haemorrhages Dengue Haemorrhagic III Above signs followed by circulatory failure Fever with Dengue manifested by a rapid, weak pulse and Shock Syndrome narrowing of pulse pressure or hypotension, with the presence of cold, clammy skin Dengue Haemorrhagic IV Profound shock with undetectable blood Fever with Dengue pressure and pulse Shock Syndrome

Table modified from Wichmann and Jelinek (2004) and WHO (1997).

Despite the global importance of DenV disease, the mechanisms by which this virus causes disease are still not completely understood

(Stephenson, 2005). It has been found, however, that DenV is able to subvert the innate immune system through several strategies that target important

21 antiviral cytokines such as type I IFN. Several NsPs have specific actions that inhibit IFN production. For example, NsP2B3 cleaves host factors resulting in the interruption of IRF-3 phosphorylation which is involved in the transcription of type I IFNs. In addition, NsP2A, NsP4A, NsP4B and NsP5 have all been shown to cause disruption to the IFN I activation pathway

(Morrison et al., 2012). Of particular note, NsP5 is able to directly bind to

STAT2 and inhibits phosphorylation. This in turn disables the production of the antiviral IFN response (Mazzon et al., 2009).

Further understanding of DenV and its subsequent disease has been slowed by a lack of suitable mouse model. While several papers have now published proposed models they are all imperfect due to resistance of Wt mice to DenV. This means that to date mouse models involve severely immune compromised mice, human-mice chimeras or massive viral loading all of which have individual problems and limitations (Bente et al., 2005; Chen et al., 2007; Chen et al., 2009; Huang et al., 2000; Kuruvilla et al., 2007; Shresta et al., 2006). As such, a reliable and robust model is still being researched.

Bunyaviridae

Bunyaviridae is the largest virus family, with five genera covering over

350 viruses. However, it contains the least number of medically significant arboviruses. Bunyaviradae have a negative sense, single stranded, segmented RNA genome Figure 1.6. The three segments exist in a circular configuration inside the nucleocapsid protein and are denoted by size, large, medium and small (Walter and Barr, 2011). The genome encodes for four

22 structural proteins, two glycoproteins, the nucleoprotein and the viral polymerase L.

Some, but not all of the Bunyaviridae viruses, also encode NsPs (Guu et al., 2012). Entry of Bunyaviridae into host cells is through phagocytosis, clathrin-mediated or clathrin independent endocytosis (Guu et al., 2012;

Hollidge et al., 2012; Walter and Barr, 2011). Of most concern to human health are La Crosse virus, RVFV and California encephalitis virus. La Crosse virus and Californian encephalitis virus both cause encephalitic disease symptoms that range from fever to vomiting, stiff neck, seizures, mental impairment, coma and death. Although infection is quite common, progression to severe disease is rare (Hollidge et al., 2010). Of the two, La

Crosse virus has been identified as the major cause of paediatric encephalitis in the USA, with approximately 100 cases reported per year (CDC, 2011).

Outbreaks of human infections of RVFV occur mainly in Africa and are associated with contact with infected livestock. Disease outcomes of RVFV include asymptomatic infection, retinitis, hepatitis, renal failure, necrotic encephalitis, severe haemorrhagic fever and death (Hollidge et al., 2010).

Figure 1.6 - Schematic representation of Bunyaviridae virion and genome.

Representation of a cross section of the Bunyaviridae virion showing the embedded glycoproteins Gn and Gc, the lipid bilayer, nucleoprotein (N) and the segmented genomic RNA. The Bunyaviridae consists of three genomic RNA segments, S, M and L which are each encapsulated by the nucleoprotein and are in association with the L protein molecules (Ikegami and Makino, 2011; van Knippenberg and Elliott, 2015)

Arbovirus immunobiology and pathogenesis

Cells targeted by arboviruses in infection and disease

Although the primary cells targeted by RRV have not been fully elucidated, several immune cells have been shown to interact with the virus early in the infection process and several have been determined to contribute to disease development. Shabman et al (2008) suggest that Langerhans cells and dermal myeloid DCs are early targets for viral replication after introduction of virus from a mosquito bite, as they are located in the skin tissue. Immune cells such as T lymphocytes and natural killer cells have also been implicated in RRV pathogenesis; however to date it is the macrophage that has been shown to contribute most significantly to RRV disease symptoms. Macrophages have been found to be the main cell infiltrate in inflamed tissues during disease in a mouse model (Morrison et al., 2006).

Additionally, disease symptoms in mice have been shown to be abrogated by removal of macrophages (Lidbury et al., 2000). Macrophage migration inhibitory factor (MIF), a cytokine released by macrophages that increases host inflammatory response, has also been shown to be crucial to development of arthritic disease symptoms in a mouse model of RRV (Herrero et al., 2011). In addition, RRV and BFV have been shown to replicate effectively in DCs, however in contrast, ChikV cannot replicate in DCs showing that Alphavirus target cells are virus specific. Human epithelial and endothelial cells as well as fibroblasts and macrophages have all been shown to be permissive to ChikV infection and have been suggested as possible target

25 cells (Shabman et al., 2007; Sourisseau et al., 2007). Recent studies have shown that primary human osteoclasts can be productively infected by RRV and ChikV and that the infection of osteoblasts and the resulting pro- inflammatory mediators released are a critical factor in Alphavirus disease

(Chen et al., 2014a; Chen et al., 2014b; Noret et al., 2012).

The primary target cells of DenV infection in vivo have long been thought to be blood monocytes and tissue macrophages (Chen and Wang,

2002; Halstead, 2002; Moreno-Altamirano et al., 2004; Mosquera et al., 2005;

Wang et al., 2002). However, there has been a growing body of evidence that suggests that DCs may in fact be the primary target cells (Alen et al., 2012;

Wu et al., 2000). Libraty et al. (2001) have shown that DenV can infect and replicate in immature human myeloid DCs. Wu et al. (2000) demonstrated that immature monocyte-derived DCs and human skin Langerhans cells were highly permissive to DenV infection, with Langerhans and DCs being ten times more permissive to infection than either monocytes or macrophages.

Several other cell lines have been implicated as targets of DenV infection including those of myeloid, lymphoid, epithelial, endothelial and fibroblastic lineages (Wang et al., 2002).

It is important to acknowledge that although the method and site of viral introduction in the host is the same for all arboviruses, the cells that they target and productively infect in the host can differ between viruses and impacts the resulting infection and disease progression in the host.

Attachment and entry of arboviruses

Although the category of arbovirus covers several virus families, there are similarities in the viruses’ attachment to target cells and replication cycles. Both Alphaviruses and Flaviviruses have been shown to interact with

DC-SIGN and L-SIGN as attachment ligands which mediate viral entry to host cells (Geijtenbeek et al., 2000; Klimstra et al., 2003; Pohlmann et al., 2001;

Ryman and Klimstra, 2008; Tassaneetrithep et al., 2003). It has been proposed that interaction with these C-type lectins by Alphavirus and

Flavivirus can alter the response to infection and enhance viral replication whilst reducing antiviral activity in the host (Shabman et al., 2007; Silva et al., 2007). It has further been inferred that DC-SIGN and L-SIGN may act as common receptors to arboviruses and influence the rate of DC infection early in the host response (Ryman and Klimstra, 2008).

HS glycosaminoglycans are polysaccharides found at the cell surface of animal cells. Interaction between HS and arboviruses has been implicated in increased virulence of arboviruses in mammalian cells (Artpradit et al., 2013;

Dalrymple and Mackow, 2011; Gardner et al., 2011; Smit et al., 2002; Zhu et al., 2010). Enhanced binding and infectivity was conferred on SinV by mutations in E2 that accumulated during tissue culture passage and allowed the use of cell surface HS as attachment ligands. In addition, after only three passages in mammalian cells, SinVs ability to utilise HS binding increased by over 12 fold, and this resulted in a 100 fold increase in viral titres.

Interestingly, the authors state that Wt isolates do not utilise HS binding sites but can adapt in as little as one passage to utilise these sites. The data

27 described in the paper cites only one field isolate that was recreated through a complementary DNA (cDNA) clone so it remains possible that this is an isolate dependent effect (Klimstra et al., 1998). In fact, a study comparing a known high affinity DenV strain with DenV obtained from the serum of human patients found that the HS binding residues were conserved indicating a role for high affinity HS strains in vivo (Conway et al., 2014). In contrast, studies have shown that an increase in the binding of SinV to HS in tissue culture can lead to increased clearance by the liver in a mouse system compared to virus with low HS binding affinity (Byrnes and Griffin, 2000; Smit et al., 2001). This finding is supported by more recent studies that show HS binding attenuates alphaviral disease that is dependent on high titre viremia but can lead to an increase in virulence due to encephalitic disease (Ryman et al., 2007a; Ryman et al., 2007b). Although RRV-T48 has previously been found to bind independent of HS, it was found that a single substitution in the E2 glycoprotein conferred HS binding and expanded host range in vitro

(Byrnes and Griffin, 1998; Heil et al., 2001; Zhang et al., 2005). In addition to mounting evidence that HS plays a role in arbovirus infection, a recent study identified the ability of mosquito saliva to enhance in vitro infection by increasing viral attachment to HS (Conway et al., 2014).

Early events in host response to arboviral infection

In a study using SinV it was found that very early differences in virus replication within the first few infected host cells had the ability to change the outcome from 100 % mortality to 100 % survival (Ryman et al., 2007b). This clearly demonstrates the importance of early events in host immunity to viral

28 infection. The innate immune system is the host’s front line of defence against invading pathogens and is not reliant on previous exposure. It involves white blood cells such as monocytes, macrophages, neutrophils, basophils, eosinophils, T-cells and natural killer (NK) cells as well the complement system and the release of cytokines (Ma and Suthar, 2015). In the case of arboviruses, the virus is introduced into the skin and circulating blood, both of which are areas rich with immune cells that recognise pathogens and mount an immune response, some of these cells are thought to play a critical role in the development of arboviral disease and include DCs, macrophages,

NK and T cells (Reviewed in Briant et al., 2014).

DCs are able to capture antigen and present it to T-cells and B-cells to help modulate host immune response to infection. They are present in the mucosal linings and blood and also in the skin as Langerhans cells, a specialised form of DC. These cells are thought to be one of the first encountered by arboviral infection and have been shown to contribute to the release of antiviral cytokines such as type I IFN and tumour necrosis factor

(TNF)α resulting in initiation of the innate immune response after exposure to arbovirus (Klimstra et al., 2003; Libraty et al., 2001; Shabman et al., 2008).

Several studies have shown that DCs are not only permissive to infection by a range of arboviruses but that the expression of DC-SIGN is capable of binding to the high mannose N-glycans on arboviruses facilitating viral attachment to the cell (Feinberg et al., 2001; Johnston et al., 2000; Klimstra et al., 2003; Morizono et al., 2010; Richter et al., 2014). DC-SIGN is also

29 expressed on macrophages, another cell involved in early stages of arboviral infection.

Macrophages are phagocytic cells found in tissue and in the blood stream which play an important role in innate immunity and the inflammatory response. Monocytes are recruited to the site of infection through various cytokines and chemokines which also result in their activation and differentiation into macrophages. Two forms of activated macrophages have been proposed, M1 macrophages are cytotoxic, associated with inflammation and differentiate in response to IFNγ and lipopolysaccharide (LPS); they release high amounts of the pro-inflammatory cytokine interleukin (IL)12. M2 macrophages result in the presence of IL4 and are associated with tissue repair and wound healing. They act to reduce inflammation by releasing anti-inflammatory cytokines such as IL10 and transforming growth factor (TGF) β (Gordon and Martinez, 2010; Laskin,

2009; Verreck et al., 2004). Macrophages have been shown to play a critical role in the development of viral arthritis caused by Alphavirus due to the release of cytotoxic, pro and anti-inflammatory cytokines as well as fibrogenic mediators and subsequent tissue destruction in a dysmodulated immune response (Assuncao-Miranda et al., 2010; Lidbury et al., 2008; Morrison et al., 2006). In a mouse model of RRV infection it was shown that removal of macrophages resulted in significantly reduced RRV disease and tissue damage (Herrero et al., 2011; Lidbury et al., 2011; Lidbury et al., 2000;

Stoermer et al., 2012). Additionally, it has been shown that macrophages present in RRV and ChikV infection are skewed towards a wound healing and

30 immunosuppressive/anti-inflammatory profile (M2) as opposed to a pro- inflammatory/host defence profile (M1) at late time points (7 days post infection). An M2 profile is associated with expression of arginase 1 by macrophages/monocytes and it was shown that deletion of arginase 1 dramatically reduced viral loads and improved tissue pathology indicating that virus clearance and disease resolution is inhibited in an arginase 1- dependent manner (Stoermer et al., 2012). In addition, macrophages have been shown to contribute to disease pathology of Flaviviruses through the release of cytokines such as TNFα, IL6 and nitric oxide (NO) which has been shown to correlate with severe disease (Kreil and Eibl, 1996; Levy et al., 2010)

Another cell capable of killing infected cells during viral infection is the

NK cell, a type of cytotoxic lymphocyte that is able to recognise stressed cells and induce apoptosis without requiring antigen presentation on the affected cell and is up-regulated in the presence of type I IFN (Petitdemange et al.,

2014). The role of NK cells in arboviral infection is still poorly elucidated and is a continuing area of research. It has been suggested that NK cells may play a role in arthritic symptoms of RRV due to increased destruction of stressed tissue in the joints and have been shown, along with macrophages, to make up a large percentage of the inflammatory infiltrates in affected tissue (Aaskov et al., 1987; Herrero et al., 2011). Additionally, NK cells are recruited to the sites of DenV infection by mast cells and present with an up regulation of cell adhesion molecules (Azeredo et al., 2006; St. John et al., 2011). This results in an increase in the production of pro-inflammatory cytokines by activated

NK cells and maintenance of an inflammatory state in infected tissue (Ahmad and Alvarez, 2004; Dalbeth and Callan, 2002; Green et al., 2005).

T-cells, like NK cells, are a type of lymphocyte that originate in the bone marrow. They mature in the thymus under different positive and negative pressures into two separate T-cell subsets, CD4+ and CD8+ cells. They leave the thymus as naïve cells and are found throughout the body in circulating blood, lymphatic tissues and other organs such as bone marrow (Di Rosa and

Pabst, 2005). Interaction of the naïve T-cell with an antigen and the presence of cytokines, such as IFN and TNFα, results in activation of the naïve cells.

Activated T-cells have been shown to regulate bone homeostasis through the release of cytokines that act directly on bone tissue cells such as bone marrow and osteoblasts but have also been shown to trigger osteoclastogenesis, resulting in bone loss and joint destruction (Kong et al., 1999). This has important implications in the development of arthritic alphaviral infections.

Activation of T-cells results in the development of several different T-cell subsets, including Helper CD4+ (TH), Cytotoxic CD8+ (TC) and Memory (TM) T- cells, which can be derived from either CD4+ or CD8+ cells.

TM cells are part of the adaptive immune response and are antigen- specific; they remain after infection to improve the response of the immune system to subsequent infection by the same antigen, usually resulting in successful amelioration of the resulting disease or milder disease symptoms.

TC cells recognise virally infected cells through major histocompatibility complex (MHC) class I molecules bound to viral peptide antigen and presented on the surface of infected cells. The cytotoxic cells release granules which

32 induce apoptosis of the infected cell. Removal of TC cells in encephalitic

Alphaviruses infection in mice resulted in prevention of lesions of demyelination which occur during the inflammatory response post alphaviral infection (Subak-Sharpe et al., 1993). Additionally, TC cells have been shown to be significantly up-regulated in the synovial fluid of psoriatic arthritis patients and that this correlates with disease and bone erosion (Menon et al.,

2014). TC cells have also been shown to play an assisting role in the clearance of RRV from persistently infected macrophages indicating that they may play a dual role in disease (Linn et al., 1998).

TH cells can differentiate into several different subsets of cells, each of which secretes different characteristic cytokines. These include TH1, TH2 and

TH17 cells (Raghupathy et al., 1998). TH1 cells release cytokines such as IL2,

IFNγ and TNFα which are involved in cell mediated immunity, inflammatory reactions and delayed type hypersensitivity disorders, on the other hand TH2 cells secrete cytokines such as IL4, IL5, IL10 and IL13 which are important in regulating humoral immunity and antibody production (Sun and Kochel,

2013). In general, TH1 and TH2 cells are thought of as opposing and balancing forces and a skew to one profile can influence the outcome of infection. For instance, recovery from viral infection is linked with a TH1 immune response, while a TH2 immune response is associated with the development of more severe disease (Chaturvedi et al., 1999b). Additionally, TH1 response is associated with M1 macrophage differentiation and TH2 with an M2 macrophage profile; to date the contribution of TH17 cytokines to macrophage activation is unclear (Chinetti-Gbaguidi and Staels, 2011).

TH17 cells have only recently been described and one theory states that these are the primary cells responsible for inflammation and that TH1 cells are more related to antigen presentation and cellular immunity.

Differentiation into TH17 cells is driven by IL6, IL1β and IL23 and they release

IL17, granulocyte-macrophage colony-stimulating factor (GM-CSF), macrophage inflammatory protein (MIP)-3α and TNFα. IL17 is an important inflammatory cytokine linked to rheumatoid arthritis (RA). Furthermore, studies have shown that TH17 cells result in the production of pro- inflammatory cytokines that can also drive osteoclastogenesis resulting in bone resorption (Jovanovic et al., 1998). Additionally, TH17 cells have been shown to have a pathological effect in encephalitic alphaviral infection in mice.

Interestingly, recent studies examining the TH17 response have linked this with an ‘opening’ of the blood-brain barrier to viral infection increasing neuro immune pathology in some viral infections (Barkhordarian et al., 2015).

The complement system consists of over 30 proteins which contribute to an enzymatic cascade that is part of innate immunity but which also aids the adaptive immune response. When triggered, the complement cascade in turn triggers a series of potent anti-antigen and inflammatory events such as opsonisation, chemotaxis and cell lysis (Orlowsky and Kraus, 2015). In regards to the involvement of the complement system in arboviral infection, the mannose binding lectin pathway of complement activation has been found to be essential for the development of RRV disease symptoms and complement component 3 (C3) has been shown to contribute to disease development through mediating inflammatory tissue destruction (Gunn et al., 2012;

Morrison et al., 2008). Complement receptor 3 (CR3) is highly expressed on monocytes/macrophages and NK cells which have been shown to be up- regulated in alphaviral infection. Additionally, C3 is able to activate pro- inflammatory cells, mediating phagocytosis and initiating cytotoxic degranulation (Ehlers, 2000; Ross, 2000). Further to this, CR3 has been shown to contribute to induction of smooth muscle cell apoptosis by activated macrophages (Vasudevan et al., 2003).

A direct correlation to C3 and inflammatory damage caused by RRV infection has been demonstrated in mice where C3 deficient mice developed attenuated disease symptoms and had significantly less tissue damage than

Wt mice. Of note is the fact that this decrease in disease symptoms and tissue damage occurred with no reduction in cell infiltrates to the site and the attenuated profile in mice was attributed to the diminished expression of pro- inflammatory cytokines that require CR3 signalling (Morrison et al., 2008).

Cytokines

Cytokines are a group of proteins whose actions affect a variety of functions, including cellular metabolism, growth, differentiation and immune functions. Cytokines are released by many cell types in response to viral infection or upon recognition of viral infection in neighbouring cells

(Mahalingam et al., 2003). Host response to infection can result in systemic inflammatory response syndrome (SIRS) which occurs when the host’s pro- inflammatory response is dysregulated due to pathogens or trauma. This dysregulation results in a destructive cytokine storm and can be the cause of

35 disease manifestations (Ryman and Klimstra, 2008). Additionally, viruses have evolved ways to circumvent this system and to skew it in the favour of virus replication and survival. Cytokines can be sorted by many different grouping systems including pro-inflammatory and TH1 or anti-inflammatory and TH2, as previously discussed. During the first few hours after alphaviral infection, critical events occur that can determine the level of virus production, host immune response and the resulting disease progression

(Frolov et al., 2012).

Severe RRV disease in mice correlates with an increase in the inflammatory cytokine MIF in both serum and tissue. MIF deficient mice had attenuated disease symptoms and reduced tissue damage whilst having comparable virus titres to Wt mice (Herrero et al., 2011). Monocyte chemotactic protein (MCP) has been found to be crucial in the development of disease symptoms in a mouse model of RRV disease. If synthesis of MCP was obliterated through the use of bindarit treatment, mice developed mild disease, had significantly reduced tissue destruction and reduced inflammatory cell recruitment despite equivalent viral load in tissues when compared to non-treated mice (Rulli et al., 2009). MCP-1, IFNγ and TNFα have also been shown to be significantly up-regulated in mice infected with

RRV as opposed to mock infected controls when assessed by real time polymerase chain reaction (PCR) at 10 days post infection, which would indicate a TH1 type response at peak disease; both MCP-1 and IFNγ have also been detected in synovial effusions of infected mice. It is thought that these cytokines may contribute to inflammatory disease pathology (Herrero et al.,

2011; Rulli et al., 2007). In support of this, injection of mice with neutralising antibodies for TNFα, IFNγ or MCP-1 reduced clinical scores (Lidbury et al.,

2008). In contrast, another study found that when mice were treated with the TNFα inhibitor, etanercept, disease was greatly exacerbated and mortality increased from 0 % to 100 % (Zaid et al., 2011). This discrepancy is most likely due to timing, although the first study does not specifically state when the neutralising antibodies were administered, it appears that this was only done at peak disease; whilst in the second study the TNFα inhibitor was administered from day one every 2 days until peak disease. Recently it has been proposed that a switch from a TH1 a TH2 cytokine profile occurs and that this may contribute to persistence of RRV arthritis and also in antibody dependent enhancement of subsequent infections (Suhrbier and La Linn,

2004). A small plaque variant of RRV induced lower levels of type I IFN and was able to inhibit type I IFN signalling. As a result, mice infected with this

RRV variant had increased disease symptoms and mortality (Lidbury et al.,

2011).

Interestingly, ChikV patients have significantly increased serum levels of IL1 receptor antagonist (Ra), IL2 receptor (R), IL6, IL8, IL10, interferon gamma-induced protein (IP)10, macrophage induced gene (MIG), MCP-1 when compared to healthy controls, suggesting a TH2 type response

(Chaaitanya et al., 2011). Additionally, type I IFN up regulation occurs at very early time points in the course of ChikV infection and before type I IFN levels return to normal (Labadie et al., 2010). In VEEV infection of mice, pro- inflammatory cytokines are detectable in draining lymph nodes 6 hours after

37 inoculation. The disease then progresses to fatality within 2 weeks (Grieder et al., 1997) which is likely due to increased permeability of blood brain barrier caused by a cytokine storm (Wang et al., 2004).

In less severe forms of DenV infections, such as DF, a TH1 response is observed but if the disease progresses to DHF then a switch to TH2 cytokine response is evident (Chaturvedi et al., 1999b). For example, IL12 levels are highest in patients with DF and levels drop significantly as disease severity increases, to the point where no IL12 could be detected in patients with grade

III or IV DHF (Pacsa et al., 2000). Additionally, patients with DF have higher

IFNγ levels than patients with DHF and patients with DHF had significantly higher IL10 levels than those with DF (Chen et al., 2005). Further supporting the shift theory are studies with secondary DenV infections which indicate that a TH1 response occurs in the first three days post infection and that from the fourth day onwards there is a shift to a TH2 response (Chaturvedi et al.,

1999a). In WNV infection of mice, IL10 is up-regulated and is associated with more severe disease. Increase in IL10 was seen as early as 6 hours post infection in serum but increased again at 72 hours and peaked at 5 days, indicating a possible shift to a TH2 response after an initial TH1 response. IL10 deficient mice were found to have higher levels of other antiviral cytokines and this was associated with better clearance of the virus and less severe disease

(Bai et al., 2009).

Type I interferons (IFN) and arbovirus infection

IFNs are a family of cytokines that help regulate the immune system in response to infection, cancer or antigenic challenge (Chadha et al., 2004).

They also help to regulate the immune response and favour the production of

TH1 cytokines whilst suppressing TH2 cytokine production (Tilg and Kaser,

2004). IFNs can be divided into three groups, type I, II and III. In humans, type I IFNs include IFNα and IFNβ, which are the predominant cytokines involved in the innate immune response against viral infections, and are secreted by many cell types following infection, the primary producers being plasmacytoid dendritic cells (pDC), (Baccala et al., 2005; Taniguchi and

Takaoka, 2001). The type II class of IFNs includes IFNγ, a cytokine secreted by activated T lymphocytes and NK cells in response to antigen or cytokine stimulation (Taniguchi and Takaoka, 2001). The role and mechanism of type

III IFNs is poorly characterised. Debate still exists in the literature as to whether these are a true distinct group or if they are IFN stimulated genes with IFN-like activities (Ank et al., 2008; Ank et al., 2006; Levy et al., 2011;

Zhou et al., 2007). In either case type III IFN’s are believed to have similar actions to type I IFN’s and are thought to be activated in virus infection through IRF-3 and IRF-7 dependant pathways (Osterlund et al., 2007).

Transcription of type I IFNs is induced in cells in response to viral stimuli, such as viral entry and release of viral components, for example RNA, which is detected by mannose receptors and toll like receptors (TLR) (Munoz-

Jordan et al., 2003). (Malmgaard, 2004). Once bound to cell surface receptors, several downstream signalling pathways are activated resulting in

39 the induction of transcription factors necessary for driving gene transcription.

This event occurs in two phases, the initial induction of IFNβ followed by the transcriptional activation of IFN-inducible genes, which are required for the amplification and production of both IFNβ and IFNα (Smith et al., 2005).

The binding of TLR3 and TLR4 to viral RNA results in the phophorylation of IRF-3 through the recruitment of threonine protein kinase

(TBK1) (Baccala et al., 2005). IRF-3 resides in the cytoplasm of normal cells, and upon phosphorylation, translocates to the nucleus where it initiates the transcriptional activation of IFNβ (Munoz-Jordan et al., 2005; Taniguchi and

Takaoka, 2001). Serine threonine protein kinase R (PKR) is also activated upon binding to RNA (Basler and Garcia-Sastre, 2002). Once activated, PKR phosphorylates inhibitor of nuclear factor kappa beta (NF-κB) kinase (IκK)

(Smith et al., 2005). This activates IκK, which then phosphorylates the inhibitor molecule of nuclear factor kappa B (IκB), which is an important transcription factor for the induction of type I IFN. Once phosphorylated, IκB is ubiquinated and subsequently degraded, freeing NF-κB and allowing it to translocate to the nucleus (Goodbourn et al., 2000). Once in the nucleus, NF-

κB forms a complex called the enhanceasome with IRF-3, as well as either high mobility group protein-1/y (HMG-1/y) or activation transcription factor-

2 (ATF-2) homodimers, or ATF-2/c-Jun heterodimers (Goodbourn et al.,

2000). The enhanceasome assembles at the IFNβ gene enhancer and regulates transcription of the gene (Seth et al., 2006) Once transcribed, IFNβ then induces IRF-7 by binding to the IFNAR. This results in the translocation

40 of IRF-7 to the nucleus and the induction of IFNα expression (Baccala et al.,

2005).

Binding of IFNα and IFNβ to their shared receptor IFNAR results in the transcription of approximately 100 genes that induce an antiviral state

(Munoz-Jordan et al., 2003). After induction, IFNα/β binds to the IFNAR located on the cell surface, which is made up of two subunits, IFNAR-1 and

IFNAR-2 both of which are required for downstream signal transduction

(Bekisz et al., 2004; Taniguchi and Takaoka, 2001). Associated with IFNAR-

1 and IFNAR-2 are two Janus protein tyrosine kinases, Jak1 and Tyk2

(Baccala et al., 2005). These two receptor-associated proteins phosphorylate tyrosine residues on the IFNAR, and provide docking sites for signal transducer and activator of transcription (STAT) proteins (Bekisz et al., 2004).

STAT-2 attaches to IFNAR-1 where it is phosphorylated and subsequently binds STAT-1 (Taniguchi and Takaoka, 2001). STAT-1 and STAT-2 form a heterodimer in association with IRF-9, (also known as p48) and form IFN- stimulated gene factor 3 (ISGF3) complex (Jones et al., 2005). ISGF3 translocates to the nucleus of the cell and binds to the IFN-stimulated response element (ISRE) in promoter regions of IFN response genes, initiating their transcription (Bekisz et al., 2004). The signalling pathway initiated by

IFNα/β binding can also result in the formation of STAT-1 homodimers that stimulate IFNγ activated sites (GAS), thus up-regulating transcription of genes in the IFNγ signalling pathway (Baccala et al., 2005).

Alphavirus infection leads to the production of type I IFN in vivo (Frolov et al., 2012). The role of type I IFN response is primarily to protect uninfected

41 cells from subsequent rounds of infection and give the host time to mount an adaptive immune response (Frolov et al., 2012). In a mouse model, RRV virulence has been directly related to type I IFN sensitivity (Stoermer Burrack et al., 2014). Mice lacking the IFN α/β receptor subunit (IFNAR -/-) are unable to mediate type I IFN activity. When infected with Alphavirus they are unable to mount an effective immune response despite the competence of the rest of the immune system (Muller et al., 1994). In the case of SinV and SFV,

IFNAR -/- mice experience mortality within days whilst Wt mice show no disease symptoms (Fragkoudis et al., 2007; Muller et al., 1994; Zhang et al.,

2007). It is interesting to note that within the Alphavirus genera it has been shown that whilst VEEV induces type I IFN there is no IFN induction by

Eastern Equine encephalitis virus (EEEV) in infected mice. It is proposed that this is linked to inability of EEEV to infect cells of myeloid lineage and its absence in lymphatic tissue. It is proposed that this avoidance of host immunity is responsible for the increased neurovirulence of EEEV (Gardner et al., 2008). Later studies by Gardner et al. (2009) showed that this was strain specific and that whilst the virulent North American EEEV did initially replicate slower than its avirulent South American counterpart, the North

American strain also failed to induce significant levels of IFN whilst IFN induction was prominent in the South American strain.

Type I IFNs prevent the accumulation of DenV RNA in a variety of human cells, rather than inhibiting binding or cellular entry of the virus

(Diamond et al., 2000). Cells that were pre-treated with IFNα/β were resistant to DenV infection; however, addition of IFNα/β post infection did not prevent

42 viral replication in either antibody-dependent or antibody-independent infection. Diamond et al. (2000) suggest that this may be due to IFNα/β inhibiting the virus at an early stage and/or that DenV actively interrupts the

IFNα/β pathway.

Arboviruses derived from mammalian versus mosquito cells

Arboviruses are able to effectively evade host immune responses; however, the effect of propagation in a mosquito cell as opposed to a mammalian cell on virus morphology and consequent interaction with host immunity is still poorly defined. Some studies have shown that vector factors, such as the mosquito saliva, can increase infection in a mammalian host, for example type I IFN is down regulated in SinV infection in the presence of mosquito salivary gland extract and a TH2 cytokine profile is up-regulated

(Schneider et al., 2004). This is supported by analogous findings in ChikV where suppression of TH1 and up-regulation of TH2 cytokines was observed in mosquito bite infections compared to needle infections in mice (Thangamani et al., 2010). Other virion factors, such as differences in glycoprotein composition, have been shown to be important in differentiating host reactions to mammalian and mosquito derived virus.

Glycosylation differences between mammalian and mosquito derived virus

Virus grown in mammalian cells has been shown to have different glycosylation structures to virus grown in mosquito cells. This has been confirmed in RRV, where mammalian cell derived (Mam-) RRV was found to have complex oligosaccharides at E1 and E2 glycosylation sites as well as high mannose/hybrid at the other E2 glycosylation site. In contrast, mosquito cell derived (Mos-) RRV was found to have high mannose oligosaccharides at both

E2 glycosylation sites, whilst the E1 glycosylation site was found to comprise predominately of paucimannose oligosaccharides, as outlined in Figure 1.7

(Shabman et al., 2008). Additionally, SinV grown in mosquito cells has N- linked mannose glycans (Man3GlcNAc2) located at E glycosylation sites, while virus derived from mammalian cells has complex glycans at E glycosylation sites (Hsieh and Robbins, 1984). Similarly, N-linked glycans on Mam-DenV E were found to be a mix of high-mannose and complex glycans, whilst Mos-

DenV E have a mix of high-mannose and paucimannose glycans (Hacker et al., 2009). The DenV E protein contains two N-linked glycosylation sites,

Asparagine-67 and Asparagine-153. Deletion of one or both glycosylation sites did not affect propagation of DenV in mosquito cells, however removal of

Asparagine-67 resulted in virus that could infect mammalian cells but was unable to produce infective viral progeny and removal of Asparagine-153 resulted in reduced infectivity (Mondotte et al., 2007). Recent studies have found that basic residues at the N terminus of CP are critical for propagation of DenV in human cells, however a deletion of these residues only caused a

44 reduction in titre in mosquito cells (Samsa et al., 2012). Deletion of all three

N-linked glycosylation sites on E1 and E2, significantly reduced infectivity in

RRV infection of DC’s for both Mam-RRV and Mos-RRV. However even with the reduction in titres, Mam-RRV still had higher infectivity than Mos-RRV in

DCs. The presence of complex N-linked glycans on E2 is required for the increase in type I IFN induction which occurs in Mam-RRV infection of DC’s.

While Mos-WNV actively supresses antiviral and pro-inflammatory cytokines in macrophages, Mos-RRV was not found to actively suppress IFNβ when used to infect DC’s, further studies would need to be done to see if active suppression by Mos-RRV is occurring (Arjona et al., 2007; Shabman et al.,

2008). BFV mutants where the E2 glycosylations were deleted was able to induce high levels of type 1 interferon when derived from either mammalian or mosquito cells, however the Wt form only induces interferon when grown in mammalian cells; suggesting that the glycosylations on Mos- BFV are able to actively suppress type I IFN induction (Shabman et al., 2007). The different glycosylation exhibited may be responsible for the different viral attachment mechanisms between mammalian and mosquito derived virus (Jones et al.,

2005). However, it is interesting to note that a point mutation in NsP4B of

DenV4 was found to decrease replication capacity in mosquito cells whilst increasing replication in Vero and human cells. This mutation was also found to enhance viral replication in mice, and it is thus proposed that NsP4B may be involved in mediating the balance of effective replication between hosts, implying that there may be other factors involved in differences between mammalian and mosquito derived virus that are not solely dependent on N- linked glycosylations (Hanley et al., 2003). In support of this it has been

45 demonstrated that an RNA sequence in the viral 3’ untranslated region of the

DenV genome is essential for replication in mosquito cells but is not required for replication in mammalian cells (Groat-Carmona et al., 2012; Villordo and

Gamarnik, 2013).

Figure 1.7 - Comparative overview of mammalian and mosquito N-glycan subtypes

Mammalian processing in the Golgi apparatus generates high-mannose, hybrid and complex N-glycan subtypes. Most cell surface and secreted N-glycans are of the complex subtype. Mosquito cell processed N-glycans include high-mannose, paucimannose and hybrid N- glycan structures (de Vries et al., 2012; Mathieu-Rivet et al., 2014; Shi and Jarvis, 2007; Vasta, 2009).

Entry/attachment differences between mammalian and mosquito derived virus

Alphaviruses are very effective replicators in vivo with mature virus being released from cells within 6 hours post infection and within 16 hours each infected cell can produce over 100 virions and the second replication round is initiated (Frolov et al., 2012). It has recently been reported that arboviruses grown in mosquito cells differed in level of infectivity when compared to the same virus grown in mammalian cells (Davis et al., 2006;

Klimstra et al., 2003). It has been suggested that Mam-RRV and Mos-RRV may differentially interact with lectin receptors on the surface of cells or that they may enter cells differently but this is yet to be elucidated (Shabman et al., 2008). The mannose binding lectin (MBL) pathway has been identified as essential for the development of RRV disease in a mouse model. MBL deficient mice showed a marked reduction in disease symptoms and tissue damage when compared to Wt mice infected with RRV (Gunn et al., 2012). Thus, it is possible that lectin receptors may be triggered differentially by mammalian and mosquito derived virus depending on glycosylation, leading to different initial immune reactions that can alter the course of disease. In addition, other Alphaviruses have been shown to have differential infectivity, with both

Mos-SinV and Mos-WNV infecting DCs more efficiently than Mam-SinV and

Mam-WNV respectively (Davis et al., 2006; Klimstra et al., 2003).

Furthermore, Mos-WNV infected monocyte-derived DCs using DC-SIGN, but

Mam-WNV attached more efficiently to the DC-SIGN homologue L-SIGN and only weakly to DC-SIGN.

Differential immune responses to mammalian and mosquito derived virus

Alphaviruses are adept at interfering with the innate immune response through inhibition of cellular transcription. Recent studies have shown that the first two to four hours post infection are critical to the induction of cell defences. If this critical period is interrupted or dysregulated, the virus is able to establish in the host and cause disease (Frolov et al., 2012).

Mos-RRV has been found to be a poor inducer of type I IFN in DC cultures, although the virus efficiently infected DCs. In contrast, Mam-RRV has been found to be a potent inducer of type I IFN in DCs. Furthermore, this trend was observed in both BFV and VEEV where virus grown in mosquito cells resulted in low expression levels of type I IFNs in DCs (Shabman et al.,

2007). A recent study (Lidbury et al., 2011) identified a RRV variant that was type I IFN resistant. This variant resulted in increased mortality and disease symptoms in mice. It is thus reasonable to infer that perhaps the decreased induction of type I IFN by Mos-RRV may also lead to increased disease in an in-vivo model. To date, the differential induction of cytokines by Mam-RRV and Mos-RRV has not been investigated. Studies have shown that MBL- dependent activation of the complement cascade neutralized Mos-WNV but not Mam-WNV (Fuchs et al., 2010). Similar studies with DenV showed that while Mos-DenV was more susceptible (100 %), MBL-dependent complement activation also had a neutralising effect on Mam-DenV (60 %), indicating that differences in mammalian and mosquito derived Flaviviruses result in different immune reactions early in infection (Avirutnan et al., 2011).

Mouse model of RRV

Mouse models that are able to replicate human disease are vital to the development and expansion of knowledge in arboviral infections. Whilst cell culture provides many answers, it is not possible to understand the immune response and the effects on disease pathology without all of the factors present in the host immune response. As a result, the complete picture cannot be analysed without all of the factors at work in a live system.

The mouse model used in this thesis was developed under the direction of Dr Suresh Mahlingam and in collaboration with Dr Mark Heise, at the

University of North Carolina. In this model, 18-24 day old C57Bl/6J mice develop disease with comparable characteristics to human disease. Following inoculation, viral titres peak at around 24 hours and mice fail to gain weight compared to controls. Additionally, mice show disease symptoms characterised by ruffled fur, restricted movement and hind limb impairment, by day 5 and peak disease occurs between day 7 and day 10. At this point, tissue destruction and bone erosion are visible in histological samples. After peak disease is reached, mice recover slowly and symptoms resolve (Herrero et al., 2011; Morrison et al., 2006; Rulli et al., 2009).

RRV infection in mice causes severe muscle damage and inflammation in the hind limbs as a result of infiltration of macrophages and activation of the complement system (Morrison et al., 2007; Morrison et al., 2006).

Specifically, complement activation by the MBL pathway, but not the alternative or classical complement pathway has been identified as essential for the development of RRV disease (Gunn et al., 2012). Further it was shown

49 that depletion of macrophages prior to RRV infection reduced in-situ tissue damage but inflammatory infiltrates were still present (Rulli et al., 2007).

Investigations into the role of cytokines in the mouse model of RRV disease have found that MIF is significantly up-regulated at 24 hours post infection and at day five post infection which supports that massive influx of inflammatory cells (Herrero et al., 2011). Additionally, TNFα, IFNγ, IL6 and arginase-1 have been shown to be up-regulated in the quadriceps muscles and the ankle of infected mice at ten days post infection (Herrero et al., 2014).

It is interesting to note that although encephalitis is seen in the RRV mouse disease model it has been deemed non-relevant to the natural pathogenesis of arthritogenic Alphaviruses (Ryman and Klimstra, 2008).

Thesis objectives

This thesis aims to expand the knowledge on the differences between mammalian and mosquito derived virus and the effect of these differences in an in vitro and an in vivo model. The first aim of this thesis was to show dimorphic infection and replication kinetics for Mam-RRV and Mos-RRV in a tissue culture system. Expanding on this, the study included an examination of 18 field isolates collected from mosquitos in Western Australia to see if this was an isolate specific mechanism. Secondly, we aimed to investigate if the dimorphic profiles were also present in vitro and in vivo systems; this led not only to the study of RRV in a mouse model but also a refinement of the current animal model used. Thirdly, we examined the effect of Mam-RRV and Mos-

RRV field isolates on the cytokine profile elicited by the host in response to infection as well as an examination of histological pathology. Finally, we explored if these effects were unique to RRV or could have implication more broadly, in other Alphaviruses and further, in arboviruses in general.

Materials and Methods

General reagents

Reagents, chemicals and equipment used are cited in using the supplier name from which they have been sourced. However, the composition of general solutions and buffers not cited in text, are listed in Table 2.1 below.

Table 2.1 - General reagents and their preparation. Reagent Abbreviation Ingredients 1:49 w/v of Agar was added to Luria- Agar Plates N/A Bertani broth. Diethylpyro 1 mL/L diethylpyrocarbonate (Sigma- carbonate-treated DEPC-water Aldrich, USA) was added to NANOpure water water. Shaken overnight and autoclaved. Ethylene 1.861 g/L di-sodium-ethylene-diaminie- diaminetetraacetic EDTA tetraacetate.2H2O was added to acid NANOpure water. Adjusted to pH 8.0. 10 g NaCl, 10 g Tryptone and 5 g yeast Luria-Bertani LB broth per litre of NANOpure water. Adjusted to broth pH 7.0 and autoclaved for sterilization. 4 g sodium phosphate monobasic and 6.5 Neutral buffered g sodium phosphate dibasic was added to NA formalin 900 mL of NANOpure water. 100 mL of 37 % formaldehyde was added 8g NaCl, 0.2 g KCl, 1.44g Na2HPO4, 0.24 Phosphate g KH2PO4, per litre NANOpure water. PBS buffered saline (1x) Solution adjusted to pH 7.4 and autoclaved. 10 mL 1M Tris and 2 mL 0.5 M EDTA Tris-EDTA buffer TE buffer were added per litre of solution. Solution was autoclaved for sterilisation. 8 g NaCl, 0.2 g KCl, 3 g, Tris base, 0.015 Tris-buffered g Phenol red, were added per litre TBS saline NANOpure water. Solution was adjusted to pH 7.4 and sterilised by autoclaving. 2:2:96 2.5 % trypsin (Sigma-Aldrich, Trypsin N/A Australia): EDTA:PBS.

Cell lines and virus strains

All cells lines were obtained from the American Type Culture

Collection’s (ATCC) official Australian distributor, In Vitro Technologies Pty.

Ltd. For virus propagation Vero (ATCC, CCL-81) and C6/36 cells (ATCC, CRL-

1660) were used. These cells along Raw 264.7 (ATCC, TIB-71) and Jaws II cells (ATCC CRL-11904) were used for experimental work. These cell lines utilise different medium (Table 2.2) and temperature requirements, as listed below (Table 2.3). All culture media used was made as per manufactures recommendations.

Table 2.2 - Growth media. Growth Growth Medium/Supplement full Medium/Supplement Manufacturer name Abbreviation

Dulbecco's Modified Eagle DMEM Gibco, Australia Medium Reduced Serum Medium/Minimal Essential Opti-MEM Gibco, Australia Medium Roswell Park Memorial RPMI Gibco, Australia Institute medium Sigma-Aldrich, Medium 199 M199 Australia Granulocyte macrophage Sigma-Aldrich, GM-CSF colony-stimulating factor Australia Thermotrace, Foetal calf serum FCS Australia

Table 2.3 - Cell lines and their normal growth conditions.

e (ºC) e

Growth

Species

FCS (%) (%) FCS

Medium

Adherent

Morphology

Designation

Temperatur

Raw Mouse Monocyte/ Yes DMEM 10 37 264.7 (BALB/c) macrophage

African green Vero Epithelial Yes Opti-MEM 3 37 monkey

Asian tiger C6/36 N/A Yes RPMI 10 28 mosquito Mouse RPMI + 5 ng/mL Jaws II Monocyte No 10 37 (C57BL/6) GM-CSF

Cell banking and culture

Thawing of cells from liquid nitrogen cryostorage

Cells removed from liquid nitrogen cryostorage were washed twice with

PBS prior to being grown in sterile 25 cm2 tissue culture flask (Corning

Incorporated, USA) with growth media. Cells were incubated under normal growth conditions (Table 2.3) for 24 hours, after which the media was discarded and replaced with fresh medium to remove dead cells. Cells were incubated until a confluent monolayer was formed for subsequent passaging or in the case of semi adherent cells (such as Jaws II), passaging occurred when cell density became high.

Cell culture propagation

Cells were propagated in tissue culture flasks (Corning Incorporated) and all cells were washed twice with PBS. Adherent cells were either

55 trypsinised or scraped with a flask cell scraper (Nunc, Thermo Scientific,

Australia) while non adherent cells were spun down at 1200 rpm (HD

Scientific, Australia) and resuspended in fresh PBS or media. The amount of cells added to the new flask was dependent on the desired growth rate of the cells; however, in most cases a 1:8 ratio cell split was performed. Cells were checked daily and split 2-3 times per week when cells were approximately 90

% confluent or reaching high density,. This was also dependent on the growth rate of the individual cell lines. Cultures were discarded when they had reached 20 passages.

Cell viability assay

Viable cell counts were performed by diluting 1:1 cell suspension with trypan blue (Sigma-Aldrich, Australia) and delivering an appropriate amount to a Neubauer haemocytometer (Hirschmann Em Techolor, Germany). The average of four sections was multiplied by 2 x 104 to calculate the number of cells per mL in the cell suspension.

Cryopreservation of cells

In order to store cell lines, a 75 cm2 confluent flask (Corning

Incorporated) or one containing a high cell density of semi adherent cells of desired cell type was prepared. For adherent cells, cells were washed once with PBS before cells were trypsinsed and added to a centrifuge tube. For non-adherent cell lines, cell suspension was centrifuged at 1200 rpm (HD

Scientific) to collect cells. Cells were then washed twice with PBS before re- suspension in 1 mL growth medium before being added to the freezing medium, (1 mL growth media, 2.5 mL FCS (Thermotrace) and 0.5 mL dimethyl

56 sulfoxide (DMSO) (Univar, Asia Pacific Specialty Chemicals)) and 1 mL aliquots transferred to plastic cryogenic vials (Corning Incorporated, USA).

The vials were placed in a Nalgene Cryo 1 °C freezing container (Nalge Nunc

International, USA) and stored at -80 ºC for 24 hours before being transferred to a liquid nitrogen Dewar.

Virus propagation and purification

Ross River virus (RRV)-T48 was derived from the infectious clone and then passaged through Vero and C6/36 cells. In addition to this, different

RRV isolates were supplied by Dr Cheryl Johansen (University of Western

Australia) and Barmah Forrest virus (BFV) was generously supplied by

Westmead Hospital. Dengue virus (DenV) used in this study was DenV2 New

Guinea C and was generously supplied by Dr Paul Young (University of

Queensland).

RRV T48 derivation

RRV (T48 strain) was produced from the pRR64 infectious clone (Kuhn et al., 1991) as described below.

E. coli containing the infectious clone pRR64, was grown overnight at

37 °C, 300 rpm, in a starter culture of 1mL LB with 100 µg/mL ampicillin

(Invitrogen, Australia). The culture was then streaked on an LB agar plate containing 50 mg/mL ampicillin and overnight growth at 37 °C. A single colony was used to inoculate 5 mL of LB broth with ampicillin (100 µg/mL)

(Invitrogen), followed by incubation at 37 °C with shaking for 8 hours. The

57 resultant bacterial culture was diluted to 100 mL in LB with 100 µg/mL ampicillin (Invitrogen) and grown overnight at 37 °C with shaking.

A Plasmid Maxi Kit (Qiagen, Netherlands) was used for plasmid DNA extraction following the instructions provided in the Qiagen Plasmid

Purification Handbook (Qiagen 2005). In brief, 100 mL culture was collected by centrifugation at 5000 g (HD Scientific) for 15 minutes (min) at 4 °C and resuspended in 10 mL of buffer P1. This was followed by an addition of 10 mL of buffer P2 which was mixed thoroughly and the solution incubated at room temperature for 5 min. Subsequently, 10 mL of P3 was added, mixed thoroughly and incubated on ice for 20 min. Precipitated proteins were removed by centrifugation at 20000 g for 30 min at 4 °C. Supernatant was applied to a Qiagen-tip 500 (pre-equilibrated with buffer QBT). The column was washed and bound DNA eluted in 15 mL Buffer QF. DNA was precipitated by adding 10.5 mL isopropanol, mixing well and centrifuging at 15000 g for

30 min at 4 °C. The DNA pellet was washed with 5 mL 70 % ethanol and air dried for 10 min before being resuspended in 100 µL TE buffer. Purified pRR64 (25 µL) was digested with 3 µL of SacI in a total volume of 50 µL with recommended buffer for 24 hours at 37 °C. Salts and enzymes were removed by extraction with phenol:chloroform:isoamyl alcohol (25:24:1) (PCIA) (Sigma-

Aldrich, USA). The aqueous phase was collected and re-extracted. The resultant aqueous phase was mixed with an equal volume of chloroform and centrifuged at 16300 g for 5 min at room temperature in order to collect the aqueous phase. Chloroform treatment was repeated. The DNA from the aqueous phase was precipitated overnight at -20 °C using 1 part DNA solution

58 to 9 parts 3 M sodium acetate (Univar, USA) and 2.5 volumes of 95 % ice cold absolute ethanol (Sigma-Aldrich, USA). The DNA pellet was recovered by centrifugation at 15000 g for 15 min at 4 °C, washed twice with 70 % ethanol, air dried and resuspend in 30 µL of DEPC water. The reaction was set up as listed in the Table 2.44 and incubated for 2 hours at 37 °C. RNA was precipitated for 30 min at -20 °C with 1 µL of 5mM ammonium acetate

(Univar, USA), 3 µL of ice cold 95 % ethanol. RNA was then collected by centrifugation at 15000 g for 15 min at 4 °C, washed once with 70 % ethanol and resuspended in DEPC Water.

Table 2.4 - In-vitro transcription reaction mix Reagent Manufacturer Volume (µL) 10 mM rNTPs Promega, Australia 1 40 MM Cap Analog Promega, Australia 0.5 10 x SP6 Buffer Promega, Australia 1 200 mM DTT Promega, Australia 0.5 40 U/µL RNAsin Promega, Australia 0.5 5 mg/mL BSA Sigma-Aldrich, Australia 1 20 U/µL SP6 polymerase Promega, Australia 1 1 µg DNA template Section 2.6.3 1 DEPC water Refer to Table 2.1 3.5

Vero cells were grown as described previously. Confluent cells were trypsinised and growth medium added before being centrifuged at room temperature and washed with DEPC-PBS. Cells were resuspended at 1.7 x

107 cells/mL in DEPC-PBS. Precipitated RNA was added to 0.7 mL cell suspension in a sterile Gene pulser electroporation cuvette (Bio-Rad, USA).

Electroporation was performed using a Gene Pulser II (Bio-Rad USA) with voltage set at 450V and Capacitance set at 50 F. Four pulses were used with

2 seconds between each pulse. Cells were transferred to a T150 flask (Corning

Incorporated) containing growth media and incubated for 24 hours at 37 °C.

Culture media was then removed and centrifuged. Supernatant was collected and titre determined using plaque assay as detailed in section 2.3.5.

RRV Field isolates

Field isolates were supplied by Dr Cheryl Johansen from the Arbovirus

Surveillance and Research Laboratory at the University of Western Australia.

All isolates received had undergone less than five passages in tissue culture, however, information on cell lines used and exact number of passages was not available. Information provided with the isolates is listed in Table 2.5.

Table 2.5 - RRV field isolates Abbreviation Isolate Year Phenotype Locality RRV-A DC42554 2007 Northern/eastern Waroona RRV-B DC43633 2007 Northern/eastern Mandurah RRV-C K66000 2007 Northern/eastern Broome RRV-D DC41773 2006 Northern/eastern Mandurah RRV-E SW79118 2006 Northern/eastern Busselton RRV-F K61297 2006 Northern/eastern Kununurra RRV-G DC39663 2005 Northern/eastern Mandurah RRV-H DC39901 2005 Northern/eastern Murray RRV-I K50605 2003 Northern/eastern Billiluna RRV-J K51922 2003 Northern/eastern Kununurra RRV-K DC36022 2003 Variant northern/eastern Mandurah RRV-L DC36669 2003 Variant northern/eastern Waroona RRV-M DC34889 2002 Variant northern/eastern Waroona RRV-N SW70003 2003 Variant northern/eastern Harvey RRV-O SW74249 2004 South western Harvey RRV-P DC36486 2003 South western Mandurah RRV-Q DC36571 2003 South western Mandurah RRV-R K50081 2003 unknown Broome

RRV propagation and purification

Cells were grown in vented tissue culture flasks (Corning Incorporated) until 80 % confluent. Vero cells were used to propagate Mam-RRV, while

C6/36 were used to propagate Mos-RRV. In both cases, cells were washed twice with PBS before adding virus inoculum and incubating for 1 hour at the appropriate temperature (Table 2.3). After the 1 hour adsorption period, normal growth media was added and flasks returned to the incubator for 48 hours. Virus was collected by remove supernatant after the flasks had been freeze thawed and solution clarified by centrifugation. All viral stocks were stored at -80 ºC and viral titres determined by plaque assay (Section 2.3.6. ).

Barmah forest virus (BFV)

The BFV used for experiments was the Bateman’s Bay isolate, generously supplied by Westmead Hospital, NSW. Growth of BFV stocks was performed as for RRV, section 2.3.2. except that virus was harvested on the third day of incubation. Vero cells were used to propagate Mam-BFV, while

C6/36 were used to propagate Mos-BFV.

Dengue virus (DenV)

The DenV used for all experiments was propagated from virus generously supplied by Dr Paul Young, University of Queensland and was serotype DenV-2; strain New Guinea C (NGC, PR1599 strain). DenV-2, NGC is the prototype DenV-2 strain and was first isolated in 1944 (CDC, 2005).

Growth of DenV stocks was performed as for RRV, section 2.3.2. with the following changes. Adsorption was performed for 2 hours and cells were

61 incubated for 7 days prior to harvest. Vero cells were used to propagate Mam-

DenV, while C6/36 were used to propagate Mos-DenV.

Virus titration by plaque assay (RRV and BFV)

A 24 well flat bottom tissue culture plate (Corning Incorporated) was seeded with Vero cells (2 x 104 cells/well) and incubated for 24 hours at 37

ºC with 5 % CO2. Cells were washed with PBS and then washed with serum free (SF) Opti-MEM medium (Gibco). Serial dilutions of the virus inoculum were performed and two wells per plate were left uninfected as controls.

The plate was incubated for 1 hour at 37 ºC with 5 % CO2. Wells were overlaid with 500 μL/well of Opti-MEM + 3 % FCS + 0.1 % agarose and incubated for an additional 48 hours at 37 ºC with 5 % CO2. At the end of the incubation period, medium was aspirated and 0.01 % crystal violet in 20

% ethanol (Sigma-Aldrich, USA) was added to each well for 2-5 min. Wells were rinsed with deionized water and left to air dry before plaques were counted and titres determined. The virus titre was determined by applying the following formula to calculate plaque forming units (pfu) per mL:

푛푢푚푏푒푟 표푓 푝푙푎푞푢푒푠 푥 푑푖푙푢푡푖표푛 푓푎푐푡표푟 푝푓푢/푚퐿 = 푖푛푓푒푐푡푖표푛 푣표푙푢푚푒 (푚퐿)

DenV titration by immuno-focus assay

A 96 well flat bottom tissue culture microplate (Corning Incorporated,

USA) was seeded with 2 x 104 Vero cells per well and incubated for 24 hours at 37 ºC with 5 % CO2 to achieve a confluent monolayer. Cells were washed with once with PBS and once with SF Opti-MEM (Gibco). Serial virus dilutions

62 were done in a separate 96 well round bottom plate and 50 µL of each dilution was transferred to Vero cell plates. The flat bottom microplate was incubated for 2 hours at 37 ºC with 5 % CO2 and overlaid with 150 μL/well of 1 x M199 medium (Sigma-Aldrich) containing 2.5 % FCS and 1.5 % carboxymethyl- cellulose (CMC) (Sigma-Aldrich, Australia) solution. The microplate was incubated for 4 days at 37 ºC with 5 % CO2.

Medium was aspirated and the plate washed twice with PBS and then incubated at -20 °C with 200 μL/well of ice cold 80 % acetone (Univar, USA) in PBS. Acetone was flicked out of the microplate and left to dry overnight.

The following day, 200 μL/well of 10 % hydrogen peroxide (stock concentration of 30 % H2O2) (Univar) in methanol (Univar) was added to each well, and the plate incubated in the dark at room temperature for 20 min.

The microplate was then washed twice with PBS containing 1 % bovine serum albumin (BSA) (Sigma-Aldrich, Australia). Wells were blocked for 20 min with

200 μL/well of PBS containing 1 % BSA (Sigma-Aldrich). Cells were incubated for 1 h in mouse monoclonal anti-DenV-2 antibody (Tropbio, Australia) used at a 1:500 ratio diluted in PBS + 1 % BSA. The plate was washed twice with

PBS containing 1 % BSA followed by incubation in peroxidase-conjugated rabbit anti-mouse Ig-G-specific antibody, (Sigma-Aldrich, Australia) diluted

1:1000 with PBS containing 0.1 % BSA. The plate was washed twice with

PBS containing 1 % BSA. Virus infected cells were visualised using 100

μL/well of Sigmafast 3, 3’-diaminobenzidine tetrahydrochloride with metal enhancer (DAB) (Sigma-Aldrich, Australia) and incubating in the dark, at room temperature, for 40 minutes. Plate was rinsed with distilled water to

63 stop the reaction and air dried. Plates were scanned on an Epson perfection

2450 photo Scanner (Epson Australia, Australia) and the ‘plaques’ counted.

It is important to note, that although traditionally a plaque is regarded as an absence of cells on a cell monolayer, for the purpose of this DenV titration method a plaque refers to the stained area of infected cells.

Purification of RRV and BFV

RRV and BFV were purified by ultracentrifugation in a Himac CP90wx using a fixed angle rotor, P50AT-0091(Hitachi, Japan). In a hard sided ultracentrifuge tube, 0.7 mL of PBS + 20 % sucrose was used to form a sucrose cushion and layered with 2.5 mL virus suspension. Virus was then centrifuged at 30300 rpm for 5 hours at 4 °C. Supernatant was discarded and pellets were resuspended in PBS.

Purification of DenV

Purification of DenV by ultra-centrifugation

DenV was purified by ultracentrifugation in a Himac CP90wx using a swinging rotor, (Hitachi, Japan). In a polymer ultracentrifuge tube 1 mL of

PBS + 20 % sucrose was used to form a sucrose cushion and layered with 9 mL virus suspension. Virus was then centrifuged for 17 hours at 50,000 g at

4 °C. Pellet was resuspended in a minimal amount of PBS.

Purification of DenV by polyethylene glycol (PEG) precipitation

Sodium chloride, NaCl (Univar, USA) was added to the virus suspension to obtain a final concentration of 0.4 M. Then a 30 % PEG 6000 solution made with the addition of sterile PBS was added until the final strength was

7 %. Virus was precipitated overnight at 4 °C on a rotary mixer before incubation at 4 °C for 2 hours. Virus precipitate was collected by centrifugation at 10000 g for 30 min at 4 °C. Supernatant was discarded and pellet was incubated at 4 °C with 500 µL of PBS to soften pellet before being resuspended.

Virus kinetics studies

Infection of adherent cells

24 well plates were seeded with 1 x 105 cells per well (2.5 x 105 cells/mL) and incubated under normal growth conditions for 24 hours (Table 2.3). Cells were washed twice with PBS and 200 L of virus inoculation at multiplicity of infection (MOI) 1 or 0.1 was added per well and incubated for 1 hour (RRV and BFV) or 2 hours (DenV) at normal growth temperature. MOI indicates the ratio of viral plaque-forming units to the number of cells in a culture and was calculated using the following equation.

푣푖푟푢푠 푡푖푡푟푒 (푝푓푢⁄푚퐿) 푥 푣표푙푢푚푒 표푓 푣푖푟푢푠 (푚퐿) 푀푂퐼 = 푛푢푚푏푒푟 표푓 푐푒푙푙푠

After incubation, 1 mL of normal growth media was added. This was considered the zero time point. At designated time points (2, 4, 6, 12 and 24 hours) samples were taken. Either supernatant was removed and snap frozen at -80 ºC for later titration or RNA samples were taken by scraping cells and collecting the resulting suspension and adding it to 500 μL of TRI-reagent

(Ambion, USA). RNA samples were frozen at -80 ºC until extraction could be performed.

Infection of non-adherent cells

Cells were prepared in suspension and added to a centrifuge tube at 1 x 105 cells per well. Once aliquoted the tubes were centrifuged for 5 min at

315 x g. The cell pellet was washed in PBS and centrifugation repeated. Cells were infected at MOI 1 or 0.1 and the resuspended cell pellet incubated at 37

°C with 5 % CO2 for 1 hour (RRV and BFV) or 2 hours (DenV). The mock tube contained only SF Opti-MEM. After the 2 hour incubation period, media was added to make the cell suspension to 1.25 x 105 cells/mL and 1.2 mL of cell suspension added to each well of a 24 well tissue culture plate. This constituted the zero time point. Plates were then incubated at 37 ºC with 5

% CO2 and supernatant taken at 2, 4, 6, 12 and 24 hours post infection and snap frozen at -80 °C for later titration.

Mouse Studies

Mouse strains and breeding

Mice used in this study were 18-20 days old at time of infection. Wild type C57BL/6J mice were bred at the University of Canberra animal facility from breeding pairs purchased from the Animal Resources Facility (Western

Australia, Australia). In addition, Type I interferon receptor knockout mice

(IFNAR-/-) with a C57BL/6J background were kindly supplied by Dr Michael

Frese, University of Canberra, and bred in house. All work was conducted with the approval of the Committee for Ethics in Animal Experimentation at the University of Canberra.

Mouse model of disease.

C57BL/6J mice were injected subcutaneously in the axillary region at

18-20 days of age with 104 PFU (50µL) of Mam-RRV or Mos-RRV using a 29

G needle. Mice were then monitored daily for clinical score and weight gain for up to 30 days. The clinical score used was adapted from (Morrison et al.,

2006) and is presented in Table 2.6. A PBS control group was always included in which mice were inoculated with 50 µL of PBS. To assess cytokine profiles and tissue inflammation samples were collected at 6, 12 and 24 hours and 7 days post infection. At these time points, mice were euthanized using isofluorane and blood taken via cardiac puncture with a 29 G needle and snap frozen at -80 for further processing. Spleen, quadriceps, ankles and brain were also taken for further analysis. Samples to be later used for real-time

PCR were collected in sterile 2 mL microcentrifuge tubes (Quantum Scientific,

Australia) containing 5 mm stainless steel balls (Procureit, Australia) and 1 mL of TRI-reagent (Ambion). Samples for enzyme-linked immunosorbent assay (ELISA) or viral titrations were collected in sterile tubes containing stainless steel balls and 1 mL PBS. Samples were homogenised at a frequency of 30 i/s for two minutes using a Tissue lyser II (Qiagen, Netherlands) and stored at -80 for further processing. With all samples, tubes were weighed before and after the addition of tissue to enable calculations per gram of tissue. Samples taken for histology were stored in sterile tubes containing 10

% phosphate buffered formalin. These samples were then sent to Australian

National University where they were processed and stained with hematoxylin and eosin (H&E) for further examination by microscopy.

Table 2.6 - Clinical score of RRV disease in mice Score Clinical observations 0 No disease observed. 1 Ruffled fur and some lethargy. 2 Very mild hind limb weakness, but still able to grip using hind limbs. Will have reduced speed of movement. 3 Mild hind limb weakness characterised by resistance to griping with hind limbs and lethargy. 4 Moderate hind limb weakness characterised by very low hind limb grip strength, slow movement and lethargy 5 Severe hind limb weakness, characterised by resistance to movement and severe lethargy. No hind limb grip strength. Will often move both hind limbs together and will sit hunched with ruffled fur. Often with drop in body temperature. 6 Loss of function in the hind limbs, extremely resistant to movement, dragging of hind limbs can be seen as well as extreme lethargy and hunched position with ruffled fur and drop in body temperature. 7 Moribund 8 Death

Expression studies

RNA extraction from cultured cells

Samples were removed from storage at -80 °C and thawed on ice. RNA extraction was performed as recommended by the manufacturer (Ambion).

Throughout the RNA extraction procedure, RNase-free plastic tubes were used

(Molecular Bioproducts, USA) and unless stated otherwise were kept on ice.

Tubes containing cells in 500 μL of TRI-reagent (Ambion) were incubated at room temperature for 5 min. To each tube, 100 μL of chloroform (Sigma-

Aldrich) was added and incubated at room temperature for 15 min before being centrifuged for 15 min at 4 ºC (12 000 x g). The aqueous phase was collected and 250 μL of isopropanol (Sigma-Aldrich) was added followed by incubation at room temperature for 10 min and centrifugation for 10 min at

4 ºC (12 000 x g). Supernatant was discarded and the RNA pellet washed in

500 μL of 75 % ethanol in DEPC water with 5 min centrifugation at 4 ºC (12

000 x g) and the supernatant removed. The RNA pellet was air dried and resuspended in 21 μL of DEPC water. The RNA concentration and quality was determined using NanoDrop spectrophotometer ND-1000 (Thermo

Scientific, USA).

RNA extraction from mouse tissue

Samples taken for RNA extraction for mice followed the same extraction process detailed above in 2.6.1. , however, due to an increase in tissue present liquid volumes were doubled. As a result samples were in 1mL of TRI-reagent

(Ambion), 200 µL of chloroform was used. Additionally the isopropanol volume was 500 µL and the ethanol was 500 µL. The RNA pellet was air dried and resuspended in 21 μL of DEPC water.

Preparation of complementary DNA (cDNA)

The RNA samples were removed from storage and thawed on ice. In a new polymerase chain reaction (PCR), RNase free tube 1 µL of 10mM dNTP mix, 1 µL of random primer and 1-5 µg of RNA template and DEPC water were added to make 10 µL of total volume. Samples were then incubated in a thermocycler (Eppendorf, Australia) at 70 °C for 10 min after which they were placed on ice and 10 µL of the Reverse transcriptase master mix was added

(Table 2.7).

Table 2.7 - Reverse transcriptase master mix reagent Volume (μL) Reagent Supplier 2 10x M-MLV Reverse Transcriptase Sigma –Aldrich, Australia Buffer 1 M-MLV Reverse Transcriptase Sigma –Aldrich, Australia 0.5 Rnase Inhibitor Sigma –Aldrich, Australia 6.5 DEPC treated water Refer to Table 2.3

The RNA sample was further incubated in theromocycler; initially sample was warmed for 10 min at 25 °C to help prevent damage to RNA.

Amplification occurred at 37 °C for a period of 50 minute, denaturing occurred at 90 °C for 10 min and then returned to 4 °C. cDNA was diluted by adding

20 μL of DEPC water and concentration was determined using the Nanodrop

(Thermo Scientific). Samples were stored at -20 ºC for long term storage or at

4 °C for short term storage prior to being used for real time PCR analysis.

Real time PCR

Real time PCR was conducted using the Bio-Rad CFX96 Real-Time system with C1000 Thermal cycler (Bio-Rad, USA). A single master mix was made using the reagents below (Table 2.8). Reagent amounts are per sample; this was multiplied by the number of samples plus 1 to calculate the required volume.

Table 2.8 - Real-time PCR master mix reagent Volume (μL) Reagent Supplier 12.5 Sybr green mix Qiagen, Australia 2.5 Target Primer Qiagen, Australia 8 DEPC water Refer to Table 2.3

Samples were run in triplicate, in a Bio-Rad Multiplate PCR plate

(MLL9601). In each well 23 µL of master mix was added and then 2 µL of the cDNA template was added at a concentration of 100-200 ng/µL. Plate was then sealed with Bio-Rad PCR Sealers Microseal ‘B’ Film (MSB1001). While working, all reagents and samples were kept on ice. Plates were centrifuged briefly to ensure all liquid was collected in the bottom of the wells before being place in the Bio-Rad real time PCR machine and run as per the settings set out in Table 2.9.

Table 2.9 - Real time PCR parameters Cycle stage Temperature (˚C) Time (seconds) Denature 95 600 95 15 Cycling (40 cycles) 55 30 72 30 50 30 Melt curve Increase by 1 (repeat to 99) 5 99 50

A melt cycle was completed at the end of the PCR reaction to ensure that only one product was being amplified. In addition to the use of negative control and a house keeping gene, samples were run with an internal calibration sample to ensure any samples that were run on different plates were comparable.

Real time PCR calculation of relative expression

The comparative cycle threshold (CT) method previously documented

(Arcari et al., 1984; Godfrey et al., 2000) was used to calculate relative expression of messenger RNA (mRNA) species. Results were calculated relative to a negative control (PBS sample group) and quantity of RNA input was controlled by the inclusion of the house keeping gene, hypoxantine phsophoribosyltransferase (HPRT). In order to normalise samples the difference between the CT values for the target gene expression and HPRT gene expression had to be calculated (ΔCT). To normalise data against the negative control, ΔCNC was calculated for each of the negative control samples. Relative expression of mRNA was then calculated by using the amplification efficiency as determined by the real time PCR machine. Calculations were completed in the software package, Bio-Rad CFX Manager version 3.1. 72

Statistical analysis

Data has been presented, where possible, in graph form as the mean + standard error of the mean (SEM). GraphpadPrism 6 statistical software was used to perform statistical analysis and prepare graphical representations of data.

Characterisation of T48 Ross River virus replication in tissue culture

Introduction

RRV mammalian infection is initiated when an infected mosquito bites the vertebrate host during a blood meal; as a result, dermal dendritic cells that reside in the skin are likely encountered very early and have been suggested as target cells. However, mounting evidence implicates macrophages as not only an early target cell but also as a key contributor to the development of RRV disease through expression of pro-inflammatory cytokines, chemokines and chemoattractants (Herrero et al., 2011; Morrison et al., 2006; Shabman et al., 2008).

Type I IFNs, IFNβ and IFNα; have been identified as critical factors in

RRV disease. IFNβ is produced by many cells including macrophages, plasmacytoid dendritic cells and fibroblasts (Baccala et al., 2005; Taniguchi and Takaoka, 2001). Its role in the innate immune system is antiviral, and

IFNβ amplifies the expression of IFNα through a positive feedback loop. Both cytokines then act together to recruit and activate macrophages and NK cells

(Chadha et al., 2004). In addition, IFNβ provides protection to uninfected cells from subsequent infection, allowing time for the adaptive immune system to mount a response, as well as having anti-inflammatory effects

(Frolov et al., 2012). Of particular interest is a study by Holten (2004), where

IFNβ treatment was given at the onset of symptoms in mice with collagen-

75 induced arthritis, this resulted in significantly less bone and cartridge destruction compared to untreated mice as well as a reduction in inflammatory markers. This eludes to a possible role for IFNβ in down regulation of disease symptoms which result from tissue destruction as part of the host’s inflammatory processes.

In the course of natural RRV infection in the vertebrate host, the immune system first encounters Mos-RRV. However, once the virus has replicated in host cells, all subsequent virus particles are then Mam-RRV.

Whether this first contact can have an effect on the resulting disease, several days later, has not been clearly elucidated and understanding of this initial contact with the immune system is important in infection by arboviruses and vital to further developing the understanding of how arboviruses cause disease.

Structural differences between Mam-RRV and Mos-RRV have been identified at the virions N-linked glycosylation sites, with Mam-RRV shown to have complex oligosaccharides at E1 141 and E2 262, as well as a hybrid form at E2 200, whilst Mos RRV has high mannose oligosaccharides at E2 200 and

262 and paucimannose oligosaccharides at E1 141. These differences in glycosylation have been linked with the capacity to induce IFNβ and may result in different attachment mechanisms of the virus to target cells

(Shabman et al., 2008). Furthermore, Mos-RRV has been shown to have higher infectivity compared to Mam-RRV in myeloid dendritic cells. Shabman et al. (2008) found that deletion of all three glycosylation sites on Mam-RRV

76 reduced plaque size on BHK-21 cells, indicating viral growth was restricted by these mutations in this mammalian cell line.

Interestingly, deletion of all glycosylation sites reduced infectivity of both Mam-RRV and Mos-RRV, however in comparison Mos-RRV still had higher levels of infectivity than Mam-RRV, suggesting that glycosylation is not the only variable at play (Shabman et al., 2007). A study using SINV, an alphavirus related to RRV, found that a mutation in nsP2 made SINV a potent inducer of IFNβ and reduced plaque size in IFNβ competent cells (Cruz et al.,

2010).

This chapter aims to investigate the replication kinetics of Mam-RRV-

T48 and Mos-RRV-T48 in several cell lines to see if significant differences can be identified and also to confirm if IFNβ induction differs for Mam-RRV-T48 and Mos-RRV-T48 in both a macrophage and a dendritic cell line.

Results

Plaque morphology of Mam-RRV-T48 and Mos-RRV-T48

The first step in investigating the effect of cell line of origin on viral fitness was to compare the plaque morphology produced by Mam-RRV-T48 and Mos-RRV-T48 in Vero cells. Confluent washed Vero cells were incubated with virus for one hour before being overlaid with agarose and incubated for

48 hours at 37 C. After staining with crystal violet, plaque differences were assessed visually (Figure 3.1). The phenotypical plaque of Mam-RRV-T48 was

77 found to be larger with more diffuse edges, in contrast to Mos-RRV-T48 which generated small, ‘pin-prick’ plaques.

Figure 3.1 - Differential plaque morphology of Mam-RRV-T48 and Mos-RRV-T48

Mammalian derived RRV (top row) and Mosquito derived RRV (bottom row) were plaqued on Vero cells and incubated for 48 hours before being stained with crystal violet. Plaque morphology was then visually compared.

Multistep replication kinetics of Mam-RRV-T48 and Mos-

RRV-T48 in Vero, C6/36, Raw 264.7 and Jaws II cells

The next step in comparing Mam-RRV-T48 and Mos-RRV-T48 was to examine the growth kinetics of Mam-RRV-T48 and Mos-RRV-T48 in various cell lines. This was initially done in two cell lines commonly used to propagate arboviruses in a laboratory setting, Vero, a mammalian line, and C6/36, a mosquito cell line. When tested in the mammalian interferon deficient system

(Vero cells), Mam-RRV-T48 titres were significantly higher than Mos-RRV-T48 at four, 12, 24 and 48 hours post infection (Figure 3.2A). In contrast, in the

78 mosquito cell system, Mam-RRV-T48 titres were only significantly higher than

Mos-RRV-T48 titres at 24 hours, with Mos-RRV-T48 titres significantly higher than Mam-RRV-T48 titres at 6, 12 and 48 hours post infection (Figure 3.2B).

Both Mam-RRV-T48 and Mos-RRV-T48 reached higher titres when grown in

C6/36 cells compared to Vero cells.

As both Vero and C6/36 cells are not relevant to immune function in a mammalian host, the growth kinetics of Mam-RRV-T48 and Mos-RRV-T48 in immune cells was investigated. A macrophage cell line, Raw 267.4, (Figure

3.2C) and a dendritic cell line, Jaws II, (Figure 3.2D) were chosen as not only have these cell types been implicated as target cells but also are major players in disease development. Compared to overall titres obtained in Vero and C636 cells, titres were significantly reduced in RAW 264.7 and JAWS II cells by several logs. At 24 hours post infection titres in Vero cells were ~7.5 and ~6.5 log pfu/mL for Mam-RRV-T48 and Mos-RRV-T48 respectively. In C6/36 cells titres obtained at 24 hours post infections were ~11.5 and ~10.5 log pfu/mL for Mam-RRV and Mos-RRV respectively. In contrast, titres obtained in Raw

264.7 and in Jaws II cells were between 2.5 and 4 log pfu/mL for Mam-RRV-

T48 and Mos-RRV-T48. In addition, Mos-RRV-T48 titres were significantly higher than Mam-RRV-T48 titres in both Raw 264.7 and Jaws II cells from four hours post infection onwards.

A B 12 12 * *

)

L 10 * L 10

m m *

u 8 * u 8

g 6 g 6 *

l * l

(

e 4 e 4

i * i

T 2 T 2 0 0 0 20 40 0 20 40 Hours post infection Hours post infection

C D 5 5

) ) * * L 4 L 4 *

m * m

/ * /

u u f * f *

p 3 p 3

g g

o o

l l

( 2 ( 2 * *

e e

r r

t t i 1 i 1

T T 0 0 0 20 40 0 20 40 Hours post infection Hours post infection

1 2 Figure 3.2 - Growth kinetics of mammalian and mosquito derived RRV in various cell lines 1 2 M a m T 4 8 1 0 Mam-RRV-T48 ( ) andM a Mosm T-4RRV8 –T48 ( ) wereM o useds T to4 8 infect A) Vero, B) C6/36, C) Raw 1 0 8 264.7 and D) Jaws IIM o cellss T 4 at8 an MOI of 0.01 to obtain multistep growth curve kinetics. 8 Samples were taken at 0, 2, 4, 6, 12, 24 and 48 hours post infection and titrated by plaque 6 assay. Each data point represents the mean of three separate experiments and error bars 6 depict as SEM. Statistical significance (*, p<0.05) was determined by two-way ANOVA with 4 4 Tukey’s test. 2 2

0 0 0 2 0 4 0 0 2 0 4 0

Induction of IFNβ by Mam-RRV-T48 and Mos-RRV-T48 in

Raw 264.7 and Jaws II cells

IFNβ has been shown to be important in arboviral disease development and Mam-RRV-T48 and Mos-RRV-T48 have previously been shown to cause differential induction profiles in primary myeloid dendritic cells (Shabman et al., 2007). As such, it was important to investigate whether differential induction by Mam-RRV-T48 and Mos-RRV-T48 of IFNβ occurred in other immune cells. When assessed by real time PCR, gene expression of IFNβ was induced to higher levels by Mam-RRV-T48 in both Raw 264.7 and Jaws II cells compared to expression levels in cells infected with Mos-RRV-T48 and negative controls (Figure 3.3).

Expression of IFNβ by Raw 264.7 cells infected with Mam-RRV-T48 was over 2 fold higher than cells infected with Mos-RRV-T48 at all time points and was over 5 fold higher than negative controls at all time points. The gene expression of IFNβ by Raw 264.7 cells infected with Mos-RRV was over 2 fold higher than negative controls at four hours, was down regulated by 3.5 fold at six hours and was not significantly different at 12 hours post infection.

In Jaws II cells, infection with Mam-RRV-T48 resulted in significantly higher levels of IFNβ gene transcription than both Mos-RRV-T48 infected cells and negative controls at four hours six and 12 hours post infection.

A 10

o i * * s * *

p * *

x 6

s 4

a *

N 2 *

1 0 0 4 6 12 1 0 1 0 Hours post infection

8 B 10 * * * *

8 * * 8 * *

n o

i 8 s

s * * 6* * * * n * *

* n * e

n i

o s

x s

e e

r s * * * * 6

6 r

p r

l 6

x p

e o

e d

l o

s 4 * *

i o

a d

d e

i 4 m

l e

* s

r i

a 4

s 4

l a

a N

r m

o m

* r * r

o N N 2 2

2 * 2 * 0 * 4 6 12 Hours post infection 0 4 6 1 2 0 0 H o u r s p o s t i n f e c ti o n 4 6 1 2 4 6 1 2 H o u r s p o s t i n f e c ti o n Figure 3.3 - IFNβ induction by Mam-RRV-T48H o uandr s p Moso s t i-nRRVf e c ti-oT48n M am Raw 264.7M acellsm (A) andM o Jawss II cells (B) were infected with Mam-RRV-T48 ( ),M Mosam -RRV- T48 ( ) Moro sPBS ( ) PatB San MOI of 0.01. Samples were collected at 4, 6 and 12 Mhourso s post P B S infection.P RNAB S was extracted, transcribed to cDNA and used to determine gene expression of IFNβ by real time PCR. Data was normalised to the house keeping gene HPRT1 and expressed as relative expression to PBS inoculated mice. Results are presented as the mean (n=3) and error calculated as ±SEM. Statistical significance (*, p<0.05) was determined by two way ANOVA with Tukey’s post-test.

Discussion

Arboviruses are a significant burden to human health and have evolved many ways in which to evade host immune responses including utilising components of the mosquitos saliva that are injected into the host with the virus (McCracken et al., 2014; Schneider et al., 2004). In addition, both virally encoded mechanisms and N-linked glycosylations affect IFNβ production and modulate host responses (Cruz et al., 2010; Munoz-Jordan et al., 2005;

Shabman et al., 2008). When the virus replicates in the mosquito system, different glycans are linked at the N-linked glycosylation sites to those produced in mammalian cells. It is possible that these changes may have an effect on how the virus is then able to attach, replicate and disseminate in the mammalian host. Understanding the initial contact of the virus with a mammalian host and factors that may give the virus an advantage are vital in our understanding of these viruses and for the future development of treatment and prevention strategies.

Differential plaque morphology of mammalian and mosquito derived RRV

Plaque morphology is often used as a phenotypical indicator of a virus and can be used as a guide for the replication fitness in a cell line. In the case of RRV, Mam-RRV-T48 clearly has large, diffuse plaques when compared to the “pin prick” plaques of Mos-RRV-T48. Larger, more diffuse plaques indicate ease of growth and spread in the cell line, resulting in higher titres.

This was confirmed by the replication kinetics study in Vero cells, where

Mam-RRV-T48 produced higher titres in Vero cells than Mos-RRV-T48.

Interestingly, a recent study found that a small plaque Mam-RRV-T48 variant missing all three N-linked glycosylation sites resulted in a small plaque phenotype and reduced titre on BHK-21 cells, as well as reduced IFNβ induction in myeloid DCs, suggesting a potential link between the pin-prick plaques of the Mos-RRV-T48, glycosylation of the RRV envelope proteins and

IFNβ induction (Lidbury et al., 2011).

Differential replication kinetics of mammalian and mosquito derived RRV in Vero and C6/36 cells

Vero and C6/36 cell lines are commonly used to propagate arboviruses in the research environment and were the two lines used to propagate Mam-

RRV-T48 and Mos-RRV-T48. Both cell lines are deficient in immune functions, allowing viruses to reach high titres. Vero cells are not able to produce type I IFN, however they are able to respond to IFN if it is introduced into the culture (Mosca and Pitha, 1986). C6/36 cells have an analogous deficiency in that they do not have an antiviral RNA interference response which is the equivalent system in the insect immune response (Brackney et al., 2010). Mam-RRV-T48 produced significantly higher titres than Mos-RRV-

T48 at most time points when grown in Vero cells, with the highest titres obtained at 48 hours post infection. In contrast, when grown in C6/36 cells,

Mam-RRV-T48 was only significantly higher than Mos-RRV-T48 at 24 hours post infection, which was the highest titre reached during the time, after which Mam-RRV-T48 titres dropped below that of Mos-RRV-T48.

Comparing RRV infection in the two cell lines it is interesting to note that whilst the highest titre of Mam-RRV-T48 was obtained at 24 hours in

C6/36 cells, Mam-RRV-T48 was significantly higher than Mos-RRV-T48 in

Vero cells at most time points. Additionally Mos-RRV-T48 was significantly higher than Mam-RRV-T48 in C6/36 cells at most time points. This may indicate adaptation to the cell line to increase viral production in either mammalian or mosquito cell lines. A study by Fayzulin and Frolov (2004), investigated the capacity for adaptation between mammalian and mosquito viruses using SINV and found that SINV was capable of adaptation to either mammalian or mosquito cell lines within one passage. They also identified a

51 nucleotide conserved sequence element located in the NsP1 encoding region which is highly conserved in alphaviruses and required for efficient replication in mosquito cells but not mammalian cells, indicating that factors in the viral genome have an effect on viral replication in different hosts.

Differential replication kinetics of mammalian and mosquito derived RRV in Raw 264.7 and Jaws II cells

Raw 264.7, a macrophage cell line, and Jaws II, a dendritic cell line, are both representative of cells that are important mediators of the innate immune response, have been suggested to encounter RRV early in natural infection and are able to produce type I IFN. As a result, these cell lines were chosen to compare Mam-RRV-T48 and Mos-RRV-T48 replication kinetics.

Multistep replication kinetics were performed to mimic the replication occurring in the host, taking into account that only the initial viral inoculum is either Mam-RRV-T48 or Mos-RRV-T48, with all subsequently replicated

85 particles being Mam-RRV-T48 as generated by the cell line in use. As is expected in cells with a functional immune system, titres were much lower in both cell lines compared to infections carried out in Vero and C6/36 cell lines where immune response in dysfunctional. Mos-RRV-T48 had significantly higher titres than Mam-RRV-T48 in both Raw 264.7 and Jaws II cells from 6 hours post infection, indicating that in macrophage and DCs, Mos-RRV-T48 has a growth advantage over Mam-RRV-T48. As the infection was performed without the addition of mosquito saliva, proteins present in saliva cannot be the sole contributor to enhanced infection and other factors, including envelope glycosylation, are likely contributing to the effect (McCracken et al.,

2014).

Previous studies have shown that Mam-RRV-T48 and Mos-RRV-T48 surface glycoproteins have different glycosylation patterns, with Mam-RRV-

T48 glycoproteins having mainly complex and hybrid glycans and Mos-RRV-

T48 glycoproteins having high mannose and paucimannose glycans. Similar differences in glycan structure have been seen in DenV where high affinity

DC-SIGN binding increases cell infectivity and requires high mannose glycans present on Mos-DenV but not Mam-DenV (Lozach et al., 2005; Mondotte et al., 2007). Further, it has been demonstrated that enhanced infection of DCs occurs due to interactions between the high mannose glycans on both Mos-

SinV, an alphavirus, and Mos-WNV, a flavivirus, via DC-SIGN (Martina et al.,

2008; Morizono et al., 2010). Taken together, this information strongly suggests that different glycans present on mammalian and mosquito derived arboviruses play a role in the improved infectivity of Mos-RRV-T48 in immune

86 cells with DC-SIGN. However, arboviruses are quick to adapt and mutate and although it is unlikely that significant genomic changes have occurred in a single passage, a genetic difference cannot be ruled out. The initial virus-cell interaction of Mos-RRV with the host system is sufficient to give Mos-RRV a fitness advantage. This advantage is likely due to an immune evasion or down regulation strategy employed upon initial contact which gives Mos-RRV additional time to establish an infection. It is likely that Mam-RRV-T48 and

Mos-RRV-T48 would have differential cytokine expression profiles as a result.

Differential induction of IFNβ by mammalian and mosquito derived RRV in Raw 264.7 and Jaws II cells

Differential induction of IFNβ has been previously shown to occur in response to infection by Mam-RRV-T48 and Mos-RRV-T48, however these studies were performed in primary mouse myeloid dendritic cells (Shabman et al., 2007). It was important to determine not only if this differential induction could be replicated in other immune cells but also if this could be linked to the replication kinetics observed in the respective cell lines. In both

Raw 264.7 and Jaws II cells, Mam-RRV-T48 infection induced higher levels of

IFNβ gene expression from four hours post infection. Taken together with the replication kinetics, this strongly suggests that there is a difference in the way in which Mam-RRV-T48 and Mos-RRV-T48 interact with cells and evoke immune responses. In support of the theory that this difference is due to differential N-linked glycosylation, is the finding that RRV generated in a cell line capable of only complex glycan additions induced more IFNβ than RRV propagated in cells which only have the capacity to produce mannose glycans

(Shabman et al., 2007; Shabman et al., 2008). Although unlikely, it is important to note that genetic changes cannot be ruled out and future studies that sequence and compare the genomes of Mam-RRV-T48 and Mos-RRV-T48 may be required to determine the importance of genetic changes. Mutations occurring in the NsP2 region of SinV, a close relative of RRV, made the virus a potent inducer of type I IFN at early time points (Frolov et al., 2012).

Additionally a small plaque RRV variant (RRV-PERS) was found to have mutations in the NsP region, along with a point mutation in the E2 gene.

RRV-PERS grew to higher titres and induced less IFNβ in RAW 264.7 cells, as well as causing a more severe form of disease in mice (Lidbury et al., 2011).

Taken together differential induction of IFNβ in Raw 264.7 and Jaws II cells by Mam-RRV-T48 and Mos-RRV-T48 indicates that it is likely that a difference in disease progression will occur in the mouse model between the two viral preparations.

Conclusion

Differential growth kinetics and IFNβ induction of Mam-RRV-T48 and

Mos-RRV-T48 in cell culture indicates that there is a difference in how Mam-

RRV-T48 and Mos-RRV-T48 interact with, and infect, mammalian immune cells. Although many studies have been performed on immune interaction later in RRV infection, few have examined the initial contact with the host and the implications that this has for further disease. This may not only reveal novel ways of decreasing disease outcomes in the host but also has important medical research implications.

In a research setting, arboviruses such as RRV are commonly propagated in a mammalian cell line due to convenience and this is not thought to have any long term implication on research; if this fundamental assumption is incorrect then not all subsequent research may be valid to disease modelling. This has obvious research implications and should not be overlooked when examining immune response to arbovirus infection. More investigation is required to see if there is any downstream effect on disease progression at later time points but until the exact mechanism behind these differences is understood, they cannot be discounted.

Characterisation of Ross River virus field isolates in tissue culture

Introduction

RNA viruses, unlike DNA viruses, do not have effective proof reading mechanisms to stop mutations occurring during replication, resulting in high level of genetic variation and rapid evolution (Steinhauer and Holland, 1987).

However, although arboviruses are RNA viruses, alphaviruses such as Murray

Valley Encephalitis (MVE), Kunjin, SinV and EEEV in particular tend to have little variation between isolates, with less than two percent genetic variation

(Poidinger et al., 1997; White et al., 2011). These viruses have a transmission cycle of mosquito to bird and it has been hypothesised that this cycling between arthropod and avian species limits the ability of the virus to mutate, as it cannot select against either host. In contrast, RRV isolates have been reported to have a higher level of genetic diversity (Lindsay et al., 1996).

Previous studies found that in RRV areas of polymorphism occur at several regions of the E2 and NsP3 proteins and that two distinct genetic RRV variants circulate in eastern and Western Australia (WA) with greater than 6

% nucleotide diversity (Lindsay et al., 1993b; Liu et al., 2011b).

RRV is believed to cycle through mosquito and marsupial hosts, in particular kangaroos, however studies have shown that RRV can also infect the brush tail possum, flying foxes and domestic horses, dogs and cats (Boyd et al., 2001; Doherty et al., 1971; Kay et al., 2007). Another study also

91 confirmed the ability of vertical transmission of RRV in WA (Lindsay et al.,

1993a). This means that mosquitos may be born already infected and able to transmit RRV. In addition, RRV has a wide range of mosquito vectors that it can infect, including several of the Aedes species. It is possible that the wider genetic variation that occurs with RRV is due to different isolates performing better in certain hosts or vectors as well as different environmental pressures depending on the geographic region. In fact early findings suggested that RRV evolved in independent geographically isolated enzootic foci which could account for some of the variation (Sammels et al., 1995). Interestingly, BFV is similar to other alphaviruses in that it has a high degree of sequence homology between isolates although its primary reservoir is yet to be identified.

Investigations have revealed that the brush tail possum, a reservoir for RRV in suburban areas, was not susceptible to BFV infection (Boyd et al., 2001).

This chapter aims to investigate if the differential cell culture profiles of

Mam-RRV-T48 and Mos-RRV-T48 reflect the replication fitness of RRV field isolates as detailed in section 2.3.1.2. . Due to the fairly conserved genome of RRV, it is thought that differences in kinetics should be minimal between isolates

Results

Multistep replication kinetics of Mam-RRV and Mos-RRV field isolates in Vero cells

Examination of the multistep growth kinetics of Mam-RRV isolates in

Vero cells shows that titres of all isolates increased steadily over time up to

48 hours with the highest titre (13.15 ± 0.1283 log pfu/mL) occurring at 48 hours in cells infected with Mam-RRV-I (Figure 4.1A). The most variation in titres occurred at 24 hours post infection with some isolates e.g. Mam-RRV-

R and Mam-RRV-O having high titres of 10.71 ± 0.06367 log pfu/mL and

10.60 ± 0.08089 log pfu/mL respectively whilst infection of Vero cells with other isolates e.g. Mam-RRV-L and RRV-B resulted in titres of 4.660 ± 0.1224 and 5.896 ± 0.03563 log pfu/mL respectively. Mam-RRV-D was the only isolate to substantially deviate from the general growth curve in Vero cells compared to the other isolates. This isolate reached a peak titre of 10.81 ±

0.05239 log pfu/mL at 12 hours post infection, much earlier than any other isolate and then stabilised for the remaining time course.

Mos-RRV isolates increased in titre at each time point over the 48 hour period studied with the highest titre yield occurring at 48 hours in Vero cells infected with Mos-RRV-Q and was 9.649 ± 0.05112 log pfu/ml, whilst the lowest titre obtained at 48 hours was 8.4690 ± 0.02574 (Figure 4.1B). The only exception to this was Mos-RRV-D which had no significant difference in titres at 24 and 48 hours post infection. In contrast to other Mos-RRV isolates, Mos-RRV-M recorded no viral production until 12 hours post

93 infection. Additionally, Mos-RRV-L was significantly lower than all other isolates at 12 and 24 hours post infection with titres of 2.459 ± 0.05756 and

4.411 ± 0.02637 respectively.

The mean titre of Mam-RRV isolates reached at the 48 hour peak was log 11.53 ± 0.09468 whilst the mean titre of Mos-RRV isolates in Vero cells was significantly lower at a mean of 8.953 ± 0.04972.

Mam-RRV and Mos-RRV field isolate trends in Vero cells

When directly comparing mammalian and mosquito derived virus of the same RRV isolate (Appendix 1), three trends were apparent.

The first trend, represented in Figure 4.2A includes four isolates (RRV-

A,-M,-N,-R). In this group both Mam-RRV and Mos-RRV titres obtained in

Vero cells follow the same general curve but Mos-RRV titres are significantly lower than Mam-RRV titres from early time points.

The second trend, seen in Figure 4.2B, is characterised by significantly higher titres in Vero cells infected with Mos-RRV compared to Mam-RRV at early time points but at later time points viral titres are significantly higher in

Vero cells infected with Mam-RRV compared to Mos-RRV. Ten isolates fit within this group (RRV-B,-C,-E,-G,-I,-J,-K, O,-P,-Q).

The final trend, containing four isolates (RRV-D,-F,-H,-L) (Figure 4.2C), is defined by a lack of significant difference between titres obtained in Vero cells by Mam-RRV and Mos-RRV typically until 24 hours post infection.

However, at 48 hours post infection, significantly higher titres of Mam-RRV was obtained in Vero cells compared with Mos-RRV virus infection.

A 14

)

f 8

( 6

T 4

0 0 6 12 18 24 30 36 42 48 Hours post infection

B14

)

f 8

( 6

T 4

0 0 6 12 18 24 30 36 42 48 Hours post infection

Figure 4.1 - RRV field isolate multistep kinetics in mammalian cells

Mammalian derived (A) and mosquito derived (B) RRV field isolates were used to infect Vero cells at 0.001 MOI. Samples were taken at 2, 4, 6, 12, 24 and 48 hours post infection and titrated on Vero cells to determine viral titre. Data is represented as the mean (n=3) ± SEM.

A 14

)

L 12 *

/ 10

u *

l 6 *

(

r 4

t *

T 2 * 0 0 10 20 30 40 50 Hours post infection B 14 *

)

L 12

/ 10

o * l 6

(

r 4 *

t *

T 2 0 0 10 20 30 40 50 Hours post infection C 14

)

L 12 *

/ 10

l 6

(

r 4

T 2 0 0 10 20 30 40 50 Hours post infection

1 4 Figure 4.2 - RRV field isolateC 6 /3 6growth R kinetic trends in mammalian cells 1 2 1 4 Mammalian* derived ( ) and mosquito derived ( ) RRV field isolates were used to infect Vero * V e ro R C 6 /3 6 R 1 2 cells at 0.001 MOI. Samples were taken* at 2, 4,6,12, 24 and 48 hours post infection and ) 1 0 V e ro R

) * g g 1 0

o titrated on Vero cells to determine viral titre (pfu/mL). Data is representative of trends seen

l l

( 8

(

8 in the isolates examined (A) isolate RRV-M is representative of four isolates, (B) isolate RRV- L * L

m 6 Q is representative of ten isolates and (C) isolate RRV-H is representative of four isolates. m

/ 6 *

/ u

u Graph is the mean (n=3) ± SEM. f * 4 f p 4 * p 98 * * 2 2 0 0 0 1 0 2 0 0 3 0 1 0 4 0 2 0 5 0 3 0 4 0 5 0 H o u r s p o s t i n f e c ti o n H o u r s p o s t i n f e c ti o n Multistep replication kinetics of Mam-RRV and Mos-RRV field isolates in a mosquito cell line

Examination of the multistep growth kinetics of Mam-RRV isolates in

C6/36 cells shows that all isolate titres increase over time up to 24 hours

Figure 4.3A. Between 24 and 48 hours post infection, titres remained stable with all changes being within one log. At 24 hours post infection Mam-RRV-

M titres in C6/36 cells were 100 fold higher than any other isolate titre at that time point and reached a titre of 12.52 ± 0.1865 log pfu/mL, the highest titre obtained over the time course. However, the titre of Mam-RRV-M in C6/36 cells dropped at the 48 hour time point and was no longer distinct from titres obtained in other isolates.

In general, when C6/36 cells were infected with Mos-RRV isolates, titres rose steadily up to 24 hour post infection after which point they either stabilised or continued to rise Figure 4.3B. However, Mos-RRV-B and D titres have more pronounce increases in titre between time points when compared to other isolates. As in the infection of C6/36 cells with Mam-RRV-M, at 24 hours post infection Mos-RRV-M titres were significantly higher than other isolate titres at this time point with a titre of 13.19 ± 0.07383 log pfu/mL and over 1000 fold increase in titre compared to the next highest titre, obtained in

Mos-RRV-R of 10.46 0.06413 log pfu/mL. Mos-RRV-M titres decreased by 48 hours post infection and were not significantly different to the titres of other

Mos-RRV isolates at this time point in C6/36 cells. At the 24 hour time point, infection of C6/36 cells with Mos-RRV produced a high titre group with titres ranging between 11 and 12 log pfu/mL and low titre group with titres ranging

99 between 9.5 and 10.5 log pfu/mL (Figure 4.3B). These groups were not easily distinguished at the 48 hour time point but all isolates in the lower titre group maintained a lower titre than isolates that were in the high titre group.

The mean titre of Mam-RRV isolates reached at the 48 hour time point was 10.57 ± 0.1135 log pfu/mL whilst the mean titre of Mos-RRV isolates in

C6/36 cells was significantly higher at a mean titre of 12.58 ± 0.1345 log pfu/mL.

Mam-RRV and Mos-RRV field isolate trends in a mosquito cell line.

When directly comparing the titres of Mam-RRV and Mos-RRV in C6/36 cells for the same viral isolate (Appendix 2), the trends seen in isolate growth kinetics could be grouped into three trends.

The first trend depicted in Figure 4.4A, shows clear distinction between

Mam-RRV and Mos-RRV titres throughout the time course, with Mos-RRV titres consistently significantly higher than Mam-RRV titres. There were seven isolates in this group, (RRV-A,-E,-F,-J,-M,-N,-O)

The second trend, Figure 4.4B, has less difference between Mam-RRV and Mos-RRV titres, with several time points in the middle of the time course showing no significant difference between the two viral preparations.

However, Mos-RRV titres were significantly higher than Mam-RRV titres at some early time points and were significantly higher by between 100-1000 fold at 48 hours post infection. Five isolates were identified as belonging to this trend (RRV-B,-D,-H,-K,-L).

100

The third trend, Figure 4.4C, had titres of Mos-RRV initially being higher than Mam-RRV titres, but at a point during the infection Mam-RRV titres rose to above those of Mos-RRV titres. At 48 hours post infection Mam-

RRV titres dropped below their peak and Mos-RRV titres were significantly higher than Mam-RRV titres. There were six isolates in this group (RRV-C,-

G,-I,-P,-Q,-R).

101

A 14

)

f 8

( 6

T 4

0 0 6 12 18 24 30 36 42 48 Hours post infection

B 14

)

f 8

( 6

T 4

0 0 6 12 18 24 30 36 42 48 Hours post infection

102

Figure 4.3 - RRV field isolate multistep kinetics in mosquito cells

Mammalian derived (A) and mosquito derived (B) RRV field isolates were used to infect C6/36 cells at 0.001 MOI. Samples were taken at 2, 4,6,12, 24 and 48 hours post infection and titrated on C6/36 cells to determine viral titre. Data is represented as the mean (n=3) ± SEM.

103

A 14 * ) *

L 12

m / 10

p *

l 6

(

r 4

t i * T 2 0 0 10 20 30 40 50 Hours post infection B 14 *

) *

L 12

m / 10

l 6

(

r 4

T 2 * 0 0 10 20 30 40 50 Hours post infection C 14 * ) *

L 12

m / 10

g *

l 6

(

r 4

t i * *

T 2 0 0 10 20 30 40 50 Hours post infection

1 4 Figure 4.4 - RRV field isolateC 6 /3 6growth R kinetic trends in mosquito cells 1 2 1 4 Mammalian * derived ( ) and mosquito derived ( ) RRV field isolates were used to infect * V e ro R C 6 /3 6 R 1 2 C6/36 cells at 0.001 MOI. Samples were* taken at 2, 4,6,12, 24 and 48 hours post infection

1 0 * V e ro R )

) 1 0 and titrated on C6/36 cells to determine viral titre (pfu/mL). Data is representative of trends

g g

o 8

o l

( seen in the isolates examined (A) isolate RRV-N is representative of seven isolates, (B) isolate (

6 * L RRV-H is representative of five isolates and (C) isolate RRV-Q is representative of six isolates. m *

m 6

/ /

u Graph is the mean (n=3) ± SEM. u f * 4 f

p 4 * p

2 * 2 * 104

0 0 0 1 0 2 0 3 00 4 01 0 5 02 0 3 0 4 0 5 0 H o u r s p o s t i n f e c ti o n H o u r s p o s t i n f e c ti o n Multistep replication kinetics of Mam-RRV and Mos-RRV field isolates in Raw 264.7 cells

Titres obtained from infection of the macrophage cell line,Raw 264.7, with Mam-RRV and Mos-RRV isolates were significantly lower than titres obtained in either Vero or C6/36 cell infection.

Titres from infection of Raw 264.7 cell with Mam-RRV resulted in growth curves showing a rise in titre up to 12 hours before titres stabilised with little increase between 12 and 48 hours Figure 4.5A. In contrast to the general trend, Mam-RRV-B titres decreased from 12 hours post infection and titres of Mam-RRV-I dropped dramatically from 24 hours to 48 hours to below the level of detection, making it significantly lower than other isolates at this time point.

When examining the titres of Mos-RRV isolates obtained after infection of Raw 264.7 cells, in general isolate titres rose between six and 12 hours post infection after which time they remain fairly stable, Figure 4.5B. Isolates that did not follow this trend include Mos-RRV-R which continued to rise across the time course and at 48 hours was significantly higher than the next highest isolate by over 10 fold. Additionally, Mos-RRV-A did not start to rise in titre until 24 hours post infection which is significantly later than other isolates and remained at a significantly lower titre compared to other isolates at this time point. Also, Mos-RRV-N and Mos-RRV-P titres decreased between 24 and 48 hours post infection to below the level of detection making them significantly lower than other isolate titres.

105

The mean titre of Mam-RRV isolates reached at the 48 hour peak was

3.441 ± 0.1936 log pfu/mL, the mean titre of Mos-RRV isolates in Raw 264.7 cells was determined to be not significantly different at a titre of 3.323 ±

0.2106 log pfu/mL.

Mam-RRV and Mos-RRV field isolate trends in Raw 264.7 cells

When comparing the growth kinetics of Mam-RRV and Mos-RRV of the same isolate in Raw 264.7 cells (Appendix 3), growth kinetics patterns of all isolates could be represented by four trends.

In the first trend, Figure 4.6A, both Mam-RRV and Mos-RRV titres increase at a similar rate with no significant difference until later in the time course when either Mam-RRV or Mos-RRV titres increase compared to each other. For all isolates except RRV-M, there was a significant difference between Mam-RRV and Mos-RRV at 48 hours post infection. Seven isolates were found to fit within this trend (RRV-B, -D, -F, -J, -M, -N -R).

The second trend, Figure 4.6B, contains only two isolates (RRV-P and

RRV-Q) and is distinguished by either Mam-RRV or Mos-RRV titres being significantly higher than the other by 12 hours post infection, followed by a reduction in titres occurring by 48 hours post infection making the previously higher titre viral preparation now the significantly lower viral preparation. For example, in the case of RRV-Q, at 12 hours post infection Mam-RRV-Q titres were significantly higher than Mos-RRV-Q titres; however, by 48 hours post

106 infection Mos-RRV-Q titres were significantly higher than Mam-RRV-Q titres in Raw 264.7 cells.

Isolates grouped into the third trend, Figure 4.6C, are characterised by significantly higher titres resulting from Mam-RRV infection in Raw 264.7 compared to titres obtained from Mos-RRV infection for a majority of the time course. The five isolates that fall within this trend are RRV-A, RRV-C, RRV-

E, RRV-G and RRV-O.

The fourth trend shown in Figure 4.6C, is characterised by Mos-RRV infection of Raw 264.7 cells resulting in significantly higher titres than Mam-

RRV infection for a majority of the time course. The isolates in this group are

RRV-H, RRV-I, RRV-K, and RRV-L.

107

A 14

)

f 8

( 6

T 4

0 0 6 12 18 24 30 36 42 48 Hours post infection

B14

)

f 8

( 6

T 4

0 0 6 12 18 24 30 36 42 48 Hours post infection

108

Figure 4.5 - RRV field isolate multistep kinetics in mammalian macrophage cells

Mammalian derived (A) and mosquito derived (B) RRV field isolates were used to infect Raw 264.7 cells at 0.001 MOI. Samples were taken at 2, 4,6,12, 24 and 48 hours post infection and titrated on Raw 264.7 cells to determine viral titre. Data is represented as the mean (n=3) ± SEM.

109

A B 6 6

) )

L 5 L 5

m m

/ /

u 4 * u 4 * f f *

p p *

g 3 g 3

o o *

l l

( ( 2 2

e e

r r

t t i 1 i 1

T T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

C D 6 6 * * *

) * ) L 5 L 5 * * *

m m

/ /

u 4 u 4

f f

p p

g 3 g 3 *

o o

l * l

( ( 2 2

e e

r r

t t i 1 i 1 *

T T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

1 4 Figure 4.6 - RRV field isolateC 6 /3 6 growthR kinetic trends in mammalian macrophages 1 2 1 4 Mammalian* derived ( ) and mosquito derived ( ) RRV field isolates were used to infect Raw * V e ro R C 6 /3 6 R 1 2 264.7 cells at 0.001 MOI. Samples were* taken at 2, 4,6,12, 24 and 48 hours post infection

1 0 * V e ro R )

) 1 0 and titrated on Raw 264.7 cells to determine viral titre. Data is representative of trends seen

g g

o 8

o l

( in the isolates examined (A) isolate RRV-J is representative of six isolates, (B) isolate RRV-Q (

6 * L is representative of two isolates, (C) isolate RRV-O is representative of five isolates and (D) m *

m 6

/ /

u isolate RRV-L is representative of four isolates. Graph is the mean (n=3) ± SEM. u f * 4 f

p 4 * p

2 * 2 *

0 0 0 1 0 2 0 300 4100 5200 3 0 4 0 5 0 H o u r s p o s t i n f e c ti o n H o u r s p o s t i n f e c ti o n

110

Discussion

Understanding the similarities and differences between viral isolates is important for in-depth understanding of factors that contribute to viral virulence, pathogenesis, transmission and ultimately affect the efficiency of vaccine strategies. Given the sequence conservation between RRV isolates, it would be expected that all isolates replicate in a similar manner. Additionally, investigation into the growth kinetics of Mam-RRV and Mos-RRV isolates will help to further elucidate differences between mammalian and mosquito derived viruses which were documented for the laboratory strain T48 (Chapter

3).

Differential replication kinetics in Vero and C6/36 cells

When comparing titres of Mam-RRV and Mos-RRV resulting from infection of Vero cells, isolates were found to fit into one of three trends as depicted in Figure 4.2. Interestingly, in all cases, at 48 hours post infection

Mam-RRV titres were higher than Mos-RRV titres demonstrating a clear advantage for Mam-RRV in Vero cells. This data correlates to previous findings in studies examining SinV infection that indicate an ability to adapt between cell lines in one passage (Fayzulin and Frolov, 2004). In addition to this, studies involving DenV showed that compared to alternate passaging of

DenV between human and mosquito cells, repeated passage of DenV in either cell line exclusively resulted in fitness gains for the virus in that cell line and a decrease in fitness in the other cell line (Vasilakis et al., 2009). Interestingly, in this study, the trends observed in the multistep replication kinetics of RRV

111 isolates in Vero cells were also observed when multistep replication kinetics studies were performed in C6/36 cells. However, in C6/36 cell infection, infection with Mos-RRV resulted in significantly higher titres compared to

Mam-RRV infection at 48 hours post infection. It is important to note, however, that isolates did not necessarily fall within the same trend for both

Vero and C6/36 growth kinetics. These results are supported by the ‘trade- off’ hypothesis; this theory hypothesises that the cycle of alternating passage between vertebrate and arthropod host restricts the evolution of arboviruses, as any fitness gains in one host would result in a fitness loss in the alternate host (Vasilakis et al., 2009). This theory has been used to account for the fact that despite being RNA viruses, with traditionally high mutation rates, the evolution rate of arboviruses is 10 fold slower than other RNA viruses (Jenkins et al., 2002).

Differential replication kinetics in Raw 264.7 cells.

The growth curves that resulted from the multistep kinetics in RAW

264.7 cells were not as expected. In line with previous studies in DCs, it was expected that Mos-RRV isolate titres would be significantly higher than Mam-

RRV isolate titres due to the interaction of glycosylated proteins with DC-SIGN

(Shabman et al., 2007). Results reveal that the Mos-RRV isolate titres were significantly higher at the 48 hour time point in only half of the isolates tested.

In the remaining isolates it was the Mam-RRV isolate titres that was significantly higher at this time point. Thus it is possible to group the viral replication of RRV isolates in RAW 264.7 cells by either the comparison of

Mam-RRV and Mos-RRV titres over the time course or simply by dividing them

112 into two groups, one in which Mam-RRV titres are significantly higher than

Mos-RRV titres and the other in which Mos-RRV titres are significantly higher than Mam-RRV titres at 48 hours post infection. Previous explanations for higher titres for mosquito cell derived virus in immune cells, includes glycosylation differences between Mam-virus and Mos-virus which may cause differential interactions between DC-SIGN and the virus E proteins (Kato et al., 2005; Klimstra et al., 2003). Future studies would need to look at the glycosylation of isolates that resulted in low titres for Mos-RRV to determine if this theory still holds true for RRV. Sequence data for all isolates would also add valuable information and should be included in future work.

When looking at the results of the multistep growth kinetics for each isolate across all groups, it is interesting to note that isolates within trend

4.2A for growth kinetics in Vero cells had growth curves in Raw 264.7 cells in which Mos-RRV titre was significantly lower than Mam-RRV at 48 hours post infection. Additionally all isolates that were grouped in trend 4.2C for growth kinetics in Vero cells had growth curves in Raw 264.7 cells in which Mos-RRV titre was significantly higher at 48 hours post infection. When comparing trends of isolates in C6/36 cells and Raw 264.7 cells, it was found that with the exception of RRV-F all isolates that were grouped into 4.4A for growth kinetics in C6/36 cells also had growth curves in which Mam-RRV was significantly higher at 48 hours whilst for isolates that were grouped with trend 4.4B, Mos-RRV produced significantly higher titres in Raw 264.7 cells at 48 hours post infection.

113

Conclusion

Study of the multistep growth curve kinetics of viruses isolated in the field in different cell lines gives researchers important information on the way in which viruses interact in different isolated systems. This study has shown that passaging RRV in a single cell line results in a fitness advantage in that cell line and repeated passage in Vero cells reduces fitness in C6/36 cells and vice versa. It is not clear whether passage in C6/36 cells affects replication fitness in Raw 264.7 as inconsistent results were obtained across isolates and it is likely that other factors contribute to the infection of this cell line.

114

Characterisation of a Ross River virus infection mouse model

Introduction

The first mouse model for RRV disease was characterised by Lidbury et al. (2000), a model that has since been further developed over a number of years with the implementation of several modifications (Morrison et al., 2006;

Rulli et al., 2009). The current mouse model of RRV disease has been established in 18-24 day old C57Bl/6J mice, and results in severe muscle pathology, hind limb weakness and dragging in infected mice. Inflammatory infiltrates, in particular monocytes, and tissue damage is observed in histological sections of the hind limb and virus can be isolated from these areas within 24 hours of infection. Disease in the mouse model is scored on the level of hind limb function along with general signs of illness such as ruffled fur (Chapter 2 - Table 2.5) and closely mimics several facets of human disease (Herrero et al., 2011).

Due to a lack of substantive evidence, CNS involvement has been deemed irrelevant to the mouse model and currently its role remains under research (Jupille et al., 2011). However, mouse studies with RRV in the early

1980s revealed clear encephalomyelitis with focal necrosis, presence of histopathological lesions containing macrophage and granulocyte infiltrates, as well as signs of demyelination. Furthermore, maximum viral titres in

115 serum and brain peaked at day two post infection, cell infiltrates were found on day five and demyelination was extensive by day eight, with partial remyelination occurring by day 13 (Seay and Wolinsky, 1982). These findings established a disease time course which mimics the time line of disease progression in the current mouse model and thus warrants further investigation.

The current mouse model of RRV disease uses both male and female mice, it is possible that differences between the immune response in males and females may affect the outcome and reproducibility of results obtained using the current RRV mouse model, although this has not yet been evaluated. Sex hormones have been found to have an effect on both the adaptive and innate immune systems, with females generally having better outcomes when challenged with infectious agents than males in animal studies (Marriott and Huet-Hudson, 2006; Suffredini, 2007). In addition, mounting evidence suggests that the inflammatory response differs between males and females, with males tending to mount a more aggressive inflammatory response when challenged with viral infection while females have a more protective immune response (Curiel et al., 1993; Huber, 2005; Li et al., 2009; Marriott and Huet-Hudson, 2006; McAbee et al., 2012; Nagayama et al., 2006; Quach et al., 2003; Villacres et al., 2004).

It has been suggested that male mice do not produce effective levels of type I IFN in early stages of infection making them more susceptible to viral infection; the inference being that gender dimorphic response to viral infection may occur (Ellermann-Eriksen et al., 1986; Pozzetto and Gresser, 1985;

116

Zawatzky et al., 1982). Type I IFN is known to play a critical role in the development of disease in alphavirus infection and has been shown to control alphavirus replication. In vitro studies have shown that several alphaviruses grown in mammalian cells induce a higher type I IFN response compared to alphaviruses grown in mosquito cells, however, this has yet to be demonstrated in vivo (Shabman et al., 2007). This differential induction is attributed to differences in structure of viral surface glycosylation and their interaction with host cell surface receptors, but is not yet fully understood

(Shabman et al., 2008). Type I IFN responses are initiated in DCs after exposure to UV inactivated SFV in vitro. UV-Mam-SFV was found to induce a type I IFN response at four hours post infection comparable to IFN induction by Wt Mam-SFV but had significantly reduced by over 50 % at six hours post infection. The induction capacity of UV-Mam SFV was logically attributed to initial attachment and fusion events and not replication indicating that replicating virus may not be required for type I IFN induction at early stages of disease (Hidmark et al., 2005).

In support of a role for sexual dimorphic infectivity, female BALB/c mice produce lower viral titres and have reduced viral spread when inoculated with vesicular stomatitis virus compared to male mice (Barna et al., 1996).

Additionally, type I IFN receptor signalling has been shown to be activated differentially in viral infection depending on sex in C57Bl/6J mice (Geurs et al., 2012). Investigations into gender dimorphic immunity have found that testosterone has an immunosuppressive effect whilst oestrogen has an immunoprotective and anti-inflammatory effect (Angele and Chaudry, 2005;

117

Choudhry et al., 2006; Marriott and Huet-Hudson, 2006; Pfeilschifter et al.,

2002). Additionally, oestrogen directly suppresses bone resorption as well as a reduction in the cytokines that activate osteoclasts and result in bone destruction; an important pathology observed in the RRV mouse model (Lam et al., 2000; Wong et al., 2006).

Castrated male C57Bl/6 mice treated with oestrogen were able to suppress arthritis and immune cell infiltrates and had decreased concentration levels of IL6 and TNFα, whereas untreated mice showed severe arthritis and cartilage destruction with increased IL6 and TNFα concentration

(Geurs et al., 2012). Oestrogen has also been shown to regulate the differentiation and function of DCs (a target cell in RRV infection) in vitro and in vivo (Geurs et al., 2012). In addition, a study by Inoue et al. (2013) examining rheumatoid arthritis in SKG/Jcl mice that had received an ovariectomy, found that mice given oestrogen treatment suppressed arthritic symptoms, accompanied by a reduction in cell infiltrates and decreased levels of IL6 and TNFα. This study also found that progesterone treatment resulted in abolishment of cell infiltrates and decreased IL6 levels in serum. Taken together, the two female hormones appear to work together in the amelioration of arthritis in female SKG/Jcl mice. Although this mechanism has not yet been elucidated it is likely that the effect is through the regulation of TH1 and TH17 cell differentiation by oestrogen as well as the suppression of

TNFα and IL6, whilst progesterone helps to reduce immune cell infiltration by further supressing IL6 (Inoue et al., 2013).

118

To confirm IFNβ was important in the development of RRV disease, viral replication and disease progression was examined in mice lacking IFNAR

(IFNAR -/- C57Bl/6 mice). Next, the differential effect of Mam-RRV-T48 and

Mos-RRV-T48 on RRV replication, disease progression and cytokine induction was examined using the current RRV mouse model. UV inactivated virus (UV-

Mam-T48-RRV or UV-Mos-T48-RRV) was also used to determine if any observed effect on disease pathology was the result of physical determinants or required a replicating virus. To assess the role of gender dimorphism in the RRV mouse model, viral titres, weight gain and clinical disease symptoms in C57Bl/6 male and female mice were compared following infection with both

Mam-RRV-T48 and Mos-RRV-T48. In addition to these factors, the gene expression of a range of pro-inflammatory and anti-inflammatory cytokines was examined to see if a differential induction occurred after exposure to either Mam-RRV-T48 or Mos-RRV-T48.

Results

Disease profile of IFNAR -/- mice in RRV infection

The importance of type I IFN in the development of RRV disease was investigated using C57Bl/6J Wt and IFNAR -/- mice. Type I IFN involvement was assessed by comparing viral titres present in whole blood, along with mouse weights and clinical scores in Wt and IFNAR -/- mice thus allowing a comparison between a standard in-vivo model and a type I IFN non-responsive model.

119

Wt and IFNAR -/- mixed sex mice were inoculated with 104 pfu/mouse of Mam-RRV-T48 or 50 µL PBS. PBS inoculated (negative control) Wt and

IFNAR -/- mice did not have any viral titres (data not shown) and gained weight steadily over the 30 day time course (Figure 5.1B and E) with a persistent clinical score of 0 (Figure 5.1C and D) and no fatalities. In contrast,

Wt Mam-RRV-T48 inoculated mice developed viremia (3.423 ± 0.2692 log pfu/mL) detectable at 24 hours post infection (Figure 5.1A), with no fatalities observed. Mam-RRV-T48 inoculated Wt mice gained significantly less weight

(Figure 5.1B) compared to Wt negative control mice from six days post infection. In addition, hind limb symptoms in accordance with the clinical score were observed in Mam-RRV-T48 inoculated Wt mice (Figure 5.1C) from day two, and peaked at eight days post infection, with a clinical score of 4.6±

0.24. RRV disease in T48-RRV inoculated Wt mice was resolved by 14 days post infection. As expected, Mam-RRV-T48 inoculated IFNAR -/- mice (104 pfu/mouse) developed systemic viremia (8.619 ± 0.5922 log pfu/mL) (Figure

5.1A) resulting in significantly higher RRV viral loads compared to Mam-RRV-

T48 inoculated Wt mice. Dramatic weight loss and death within 48 hours post infection occurred in all Mam-RRV-T48 inoculated IFNAR -/- mice

(Figure 5.1B).

In order to examine if disease progression could occur without fatal viremia in IFNAR -/- mice, viral inoculum was reduced to 101 pfu/mouse. As a comparison, one group of Wt mice was inoculated with Mam-RRV-T48 at

101 pfu/mouse and one with Mam-RRV-T48 V at 104 pfu/mouse. Wt and

IFNAR -/- negative controls were inoculated with PBS and no viral titres were

120 recovered (data not shown). As expected Wt mice inoculated with T48-RRV at

101 pfu/mouse did not have any detectable viral titres at 24 hours post infection, gained weight at the same rate as negative controls and had a clinical score of 0 for the duration of the experiment (data not shown). Viral titres of Wt mice inoculated with Mam-RRV-T48 at 104 pfu and IFNAR -/- mice inoculated with Mam-RRV-T48 at 101 pfu/mouse were significantly different to negative controls but interestingly were not significantly different to each other with viral titres of 3.889 ± 0.1066 and 3.534 ± 0.1049 log pfu/mL respectively (Figure 5.1D). Both Wt mice inoculated with Mam-RRV-

T48 at 104 pfu and IFNAR -/- mice inoculated with Mam-RRV-T48 at 101 pfu/mouse gained significantly less weight compared to negative controls, however again there was no statistical difference between weights of Mam-

RRV-T48 inoculated Wt (104 pfu/mouse) and IFNAR -/- (101 pfu/mouse) mice

(Figure 5.1E). This similarity between T48-RRV inoculated Wt (104 pfu/mouse) and IFNAR -/- (101 pfu/mouse) mice was not mirrored in the clinical disease scores where IFNAR -/- mice developed only a mild reaction from T48-RRV infection from day two to four after which they showed no disease symptoms as seen in Figure 5.1F. In contrast, the Mam-RRV-T48 inoculated Wt mice developed a typical disease score profile, with symptoms appearing on day two, peaking on day eight and resolving by 15 days post infection. The difference between the two disease profiles of Mam-RRV-T48 inoculated mice was statistically significant from day two until day 15.

121

A D

* ) 5

)

10 L

u 4

u 8

g 3

g 6 o

(

e 2

e 4 r

2 V

R 0 0 # IF NAR - /- Wt IF NAR - /- Wt

B E 3 * 200 200

)

(

a 150 a 150

e e 2

W 1 100 W 100 * 2 * 1 * * 0 5 10 15 0 5 10 15 Days post infection Days post infection

C F 10 2 10 * 1 *

e 8 e 8 2

r r *

c 1

S 6 * S 6

i 4 i 4

C 2 C 2

0 0 0 5 10 15 0 5 10 15 Days post infection Days post infection

122

3 * 2 0 0 3 P B S - IF N A R -/- ) * *

L 2 0 0 Figure 5.1 - Type I interferon is important in development of clinical disease in mouse ) P B S - W t 1 0 ) P B S - IF N A R -/-

model of Ross river virus% infection

m )

( *IF N AP BRS --/ W- t M T 4 8 1 0 ^ 4 - W t

% 1 0

(

C57Bl/6J/ wild type (Wt) ( or ) andM T 4 IFNAR8 1 0 ^ 4 knockout - W t (-/-) ( orI F N) 18MAT Rday4 8 - 1/old-0 ^ 1mice - IF N A R -/-

f 1 5 0 W t a

8 n

g p

M T 4 8 1 0 ^1 - IF N A R -/- f a of mixed sex were inoculated subcutaneously in the pectoral region with Mam-RRV-T48 at 1 5 0

t W t

g 8

4 4

t 3

10 pfu/mouse (A, B, C). g Experiment was repeated with Wt inoculated with 10 pfu/mouse

i o 6 h

g * 3

l e

g 2 0 0 1

i (

and IFNARo -/- 10 pfu/mouse (D, E, F). Mice were monitored for daily for weight gain and

6 * P B S - IF N A R 2-/-

l e W 1 0 0

2 0( 0 2 * e ) P B S - W t 1

W disease score. Negative controls (Wt and IFNAR -/- ) were injected with 50 µL of PBS, 4 1 0 0 * P B S - IF N A R -/- r *

% 1

e t

( M T 4 8 1 0 ^ 4 - W t i B,C,E,F). A,* D) At 24 hours post infection, blood was collected by cardiac puncture and

r 4 P B S - W t

)

i 0 M5T 4 8 1 0 ^ 1 -1 I0F N A R -/- 1 5

a 1 5 0 titrated on Vero cells. Statistical analysis was performed using a two tailed non-parametric 0 t 5 1 0 1 5

2 % g

D a y s pos t infe c tionM T 4 8 1 0 ^ 4 - W t

(

t DManna y s Whitneyp2os t in fte test.c tio n B, E). Weights were taken daily and graphed as percent of starting

h n

i M T 4 8 1 0 ^ 1 - IF N A R -/- g

1 5R 0 i

R weight. Data is represented as an average weight per group (n=5) ± SEM. Statistical a

0 e R g 2

0 W significance was determined using a two way ANOVA with Tukey’s test. C, F) Mice were

IF N A R - /- 1 0 0 t W t * 1

examinedh for signsIF*N ofA diseaseR - / -and ranked accordingW t to the clinical score. Data is represented

g i

0 ase the average5 clinical1 0 score per1 5 group (n=5) ± SEM. Statistical significance was determined 2 usingW 1 D0 a 0 twoy s p wayos t i ANOVAnfe c tio withn Tukey* ’s test. p<0.05 (*). *1 denotes statistically significant 1 difference between Mam-RRV* -T48 inoculated Wt mice and PBS controls, *2 denotes a significant difference between Mam-RRV-T48 inoculated IFNAR -/- mice and PBS controls 0 5 1 0 1 5 3 and * denotesD aa significanty s pos t difference infe c ti obetwn een Wt and IFNAR -/- PBS inoculated mice.

123

Inactivated RRV in a mouse model of disease

To assess the requirement of replicating virus to cause an immune response in a mouse model, UV inactivated virus propagated in either mammalian or mosquito cell lines was used to inoculate mixed sex Wt mice.

Viral titres were collected at 24 hours post infection and mice were monitored daily for weight change and clinical score. Wt mice were inoculated with either

UV-Mam-RRV-T48, Mam-RRV-T48, UV-Mos-RRV-T48 or Mos-RRV-T48 at 104 pfu/mouse or PBS (negative control).

As expected, negative control mice were negative for viral titres and clinical scores (data not shown) and showed progressive weight gain over the

30 day trial period. No viral titres were recovered from either of the mouse groups inoculated with UV-inactivated virus whilst titres from mice inoculated with active virus reached a combined average titre of 7.397 ± 0.0852 log pfu/mL (data not shown). During the first three days post infection, Wt mice inoculated with UV-Mam-RRV-T48 and UV-Mos-RRV-T48 gained less weight than PBS controls (Figure 5.2A). The difference between the UV inactivated virus and the PBS control was not statistically significant, however it is supported by the clinical score Figure 5.2B. Interestingly, after 15-20 days post infection, mice inoculated with UV inactivated virus consistently gained more weight than PBS controls, although again this difference was not deemed to be statistically significant. Mild disease symptoms were apparent at days two and three before reducing over the next two days. At day one and three post infection, clinical scores were comparable between UV inactivated and active virus groups. It was noted that mice inoculated with UV-RRV-T48

124 were more lethargic than PBS mice for up to 17 days post infection but did not meet the criteria for a clinical score of 1.

Mam-RRV-T48 and Mos-RRV-T48 in a mouse model of disease

The ability of Mam-RRV-T48 and Mos-RRV-T48 to produce differential disease profiles was examined in the mouse model. Mixed sex Wt mice were inoculated with either UV-Mam-RRV-T48, Mam-RRV-T48, UV-Mos-RRV-T48 or Mos-RRV-T48 at 104 pfu/mouse or PBS (negative control). Mice were monitored daily for disease symptoms and weight gain.

Wt mice inoculated with PBS, as expected, did not develop disease symptoms (Figure 5.2B). PBS inoculated mice gained weight at a steady rate throughout the 30 day trial, shown in Figure 5.2A. Mam-RRV-T48 mice gained more weight than the Mos-RRV-T48 mice, but not as much as PBS inoculated mice or mice inoculated with UV-inactivated virus. The difference between weights of Mam-RRV-T48 mice was significantly different to all other groups from day 10 to 19 and the clinical score was significantly different to

Mos-RRV inoculated mice from day 3 to 20 with the exclusion of day 5.

Clinical scores for Mos-RRV-T48 were significantly higher than all other groups from day two to day 20, with the exception of day five (Figure 5.2B).

The highest average score reached by this group was 5 on day 12; however the highest individual score was 8. A score of 8 indicates a fatality, this was the only death in this experiment and as a result after day 12 the number of mice in the Mos-RRV-T48 group was reduced to 4.

125

A 2 * 3 300 *

)

(

n i 200

W 100 *1 0 10 20 30 Days post infection B 8

* 6

l 4

c *

i *

l C 2

0 0 2 10 20 30 * 3 0 0 Days3 post infection 2 * P B S * 3 0 0 3 * P B S In a c tiv e v e ro T 4 8 2 Figure 5.2 - Disease progression in mice inoculatedIn a c t iwithv e v e RRV2ro T 4 8 In a c tiv e C 6 /3 6 T 4 8 * 3 * 3 3 0 0 ) 3 0 0 * P B S In a c tiv e C 6 /3 6 T 4*8 V e ro T 4 8 P B S

% C57Bl/6J wild type 18 day old mice of mixed sex were injected subcutaneously in the pectoral ( 2 4 In a c tiv e v e ro T 4 8 ) region with 10 pfu/mouse of Mam-RRV-T48 ( ), VMose ro- RRVT 4 8 -T48 ( ), UV-Mam-RRV-T48 ( In a c tiv e v e ro T 4 8 n 2 0 0 C 6 /3 6 T 4 8 i *

3 0 0 3

% a ( ), UV-Mos-RRV-T48 *( ) oInr awithc tiv e50 C 6µL/3 6PBS T 4 8( ). P MiceB S were monitored daily for A) weights In a c tiv e C 6 /3 6 T 4 8 C 6 /3 6 T 4 8

2 0 0 g

)

n i

t and B) disease progression.V e Diseasero T 4 8 progression wasIn a scoredc tiv e v eaccordingro T 4 8 to the clinical disease a

% V e ro T 4 8

)

(

)

% In a c tiv e C 6 /3 6 T 4 8

i score in Chapter 2 (TableC 2.6)6 /3 6and T 4 8data is represented as the mean of each group (n=5) ±

t n

2 0 0 ( C 6 /3 6 T 4 8

e %

h 2 0 0

( a n V e ro T 4 8

g SEM. Statistical significance was determined by two way ANOVA with Tukey’s test. Data is

i t

2 0 0 representative of four separatelyg run experiments. C 6 /3 6 T 4 8

h W

1 0 0 t

t e 1 0 0 i

h 1 126

e W

g * i

1 e 1 0 0 0 *1 0 2 0 3 0 W 1 0 0 1 0 0 0 1 1 0 D a y2s0 pos t infe c3t0ion * 1 D a y s pos t infe c tion * 1 0 1 0 2 0 3 0 * D a y s pos t infe c tion 0 1 0 2 0 3 0 0 1 0 2 0 3 0 D a y s pos t infe c tion D a y s pos t infe c tion Cytokine expression of active and inactive Mam-RRV-T48 or Mos-RRV-T48 in a mouse model of disease.

To examine the cytokines expressed in vivo in response to active and UV inactivated RRV, mixed sex Wt mice were inoculated with 104 pfu/mouse of

Mam-RRV-T48, Mos-RRV-T48, UV-Mam-RRV-T48, UV-Mos-RRV-T48 or 50

µL PBS and sacrificed at 12 and 24 hours post infection for tissue collection.

Tissue samples were then processed as per Chapter 2, and examined for gene expression of CCL2, IFNβ, IFNγ and MIF by rt-PCR. When examining the cytokine profiles it was discovered that within each group there appeared to be two subgroups which had differential expression. This lead to large standard deviations and thus, a reduction in the ability to determine significant differences between groups.

Cytokine expression in the quadriceps muscle of mice following RRV infection

At 12 hours post infection, gene expression of CCL2 in the quadriceps muscle was significantly higher than all other groups in mice inoculated with

Mam-RRV-T48, with an 81 fold increase in gene expression compared to PBS controls (Figure 5.3). Mice inoculated with Mos-RRV-T48 had an increase in

CCL2 gene expression by 24 fold, whilst UV-RRV-T48 inoculated mice had comparable gene expression levels to PBS inoculated mice. At 24 hours post infection, CCL2 gene expression was decreased across all groups with both

Mam-RRV-T48 and Mos-RRV-T48 inoculated mice showing up-regulated gene expression of CCL2 by less than fourfold compared to PBS inoculated mice.

Mice inoculated with UV-RRV-T48 had similar gene expression levels to PBS

127 inoculated mice at 24 hours post infection. IFNβ gene expression at 12 hours post infection was highest in mice inoculated with Mam-RRV-T48, being 2.6 fold greater than PBS inoculated mice. This was found to be significantly higher than the UV-RRV-T48 and PBS inoculated mice. At 24 hours post infection gene expression of IFNβ was similar to PBS inoculated mice in all

RRV inoculated mice groups.

High levels of IFNγ gene expression were found in Mam-RRV-T48 inoculated mice at 12 hours post infection with 577 fold increase compared to PBS inoculated mice. This was found to be significantly higher than both

UV-RRV-T48 and PBS inoculated mice. Mice inoculated with Mos-RRV-T48 also had a higher levels of gene expression of IFNγ than UV-RRV-T48 or PBS inoculated mice a (370 fold increase compared to PBS inoculated mice). At

24 hours post infection the gene expression levels of IFNγ were much lower than 12 hours post infection, with expression levels of 2.8 and 0.8 fold in

Mam-RRV-T48 and Mos-RRV-T48 inoculated mice respectively when compared to PBS inoculated mice. Gene expression of IFNγ in UV-RRV-T48 mice was not significantly different to PBS inoculated mice. MIF expression in all groups at 12 hours post infection was similar to PBS inoculated mice.

At 24 hours post infection a decrease in gene expression of MIF in mice inoculated with UV-Mos-RRV-T48 to 0.002 fold expression compared to PBS inoculated mice, was found to be significantly different to the expression of

MIF in active RRV-T48 inoculated mice and the PBS control.

128

12 hour

n 24 hour

i i 150 6

p p * * * *

e e 100 4

o C C L 2 o

e e 50 2

o o 0 0

N N Mam Mos Mam Mos PBS Mam Mos Mam Mos PBS T48 UV- T48 T48 UV- T48

n n

o o

i 4 i 2.0

s s

s * * * s

e e

r r

p 3 p 1.5

d d

l l

o IFN 2 o 1.0

f f

i 1 i 0.5

l l

a a

r r

o 0 o 0.0

N Mam Mos Mam Mos PBS N Mam Mos Mam Mos PBS T48 UV- T48 T48 UV- T48

n n

o o

i 800 * * * i 5

s s

e e r r 4 p 600 * * * p

e 3

d d

l l

o IFN 400 o

f f

d d 2

i 200 i l l 1

a a

r r

o 0 o 0

N Mam Mos Mam Mos PBS N Mam Mos Mam Mos PBS T48 UV- T48 T48 UV- T48

n n

o o

i 1.5 i 1.5

s s

e e * * *

r r

p p

e 1.0 e 1.0

d d

l l

o MIF o

f f

e 0.5 e 0.5

i i

l l

a a

r r

o 0.0 o 0.0

N Mam Mos Mam Mos PBS N Mam Mos Mam Mos PBS T48 UV- T48 T48 UV- T48

130

i n

Figure 5.3s - Inflammatory cytokine profiles differ between mammalian and mosquito

i n

n 2derived. 5 RRV and also between active and inactive RRV in the quadriceps muscle.

i r

i M am

p e

s 2 . 5 s r C57Bl/6J wild type 18 day old mice of mixed sex were injected subcutaneously in the pectoral x M o s e M am 2 . 5 e 2 . 0

p * * r

2 . 5 e r

n 2 . 5 M am

x 4

p p

o regionM witham 2 10. 0 pfu/mouse of Mam-RRV-T48 ( ),M MosamM-aRRVm -T48 ( ),M UVo s-Mam-RRV- d

i * * e

x * * x

l M o s s 2 . 0

e * * M o s M am

e o

2 . 0 d

s 1T48. 5 *M o s * * * M o s

l 2 . 0 f M am

* d e * * * * d 1 . 5 M am

l * M o s o r *

l * * *

f d

o 1 . 5 ( ),M UVam-Mos-RRV-T48 ( ) Moro swith 50 µL PBS ( ).P BMiceS were sacrificed at 12 hours

p o

1 . 5

* f * * M o s P B S e

f *

x * * d 1 . 0 P B S

s 1 . 0 d

e 1 . 5 P B S i

d M o s

e and 24 hours post infection and the quadriceps muscle removed. Total RNA from the

1 . 0 l e

* * s 1 . 0

i s

l P B S i

l quadriceps 0muscle. 5 was isolated and analysed for mRNA expression by real time PCR. Data

a m

a 0 . 5

f 0 . 5 a 1 . 0

0 . 5 r

m m

d was normalised to the house keeping gene HPRT1 and expressed as relative expression to

r r

e 0 . 0

o N

s 0 . 0 o

i 0 . 0 0PBS. 0 inoculated mice.M a m ResultsM o ares representedM a m M asos the P meanB S of each group (n=5) ± SEM.

l N

0 . 5 N

M a m M os MMN aamm MMosos P BMSa m M os P B S a StatisticalM a m significanceM os M (*T ap<0.4m8 05)M wasos deUPterminedBVS- T 4 8 by one way ANOVA with Tukey’s test.

m T 48 U V -TT4488 U V - T 4 8 r o 0 . 0 T 48 U V - T 4 8

N M a m M os M a m M os P B S

T 48 U V - T 4 8

131

Cytokine expression in the ankle joint of mice following RRV infection

At 12 hour post infection the gene expression of CCL2 in the ankle joints was increased in mice inoculated with Mam-RRV-T48 and Mos-RRV-T48 by

10 and 5 fold respectively when compared to PBS inoculated mice, however this difference was not statistically significant due to large variations in expression levels in individual mice within each group as shown in Figure 5.4.

At 24 hours post infection CCL2 expression was lower than that seen at 12 hours post infection, with normalised relative expression levels of 1 and 1.6 for Mam-RRV-T48 and Mos-RRV-T48 respectively. Interestingly, the expression levels of CCL2 in Mos-RRV-T48 inoculated mice was found to be significantly higher than expression levels in mice groups inoculated with UV-

Mam-RRV-T48 and UV-Mos-RRV-T48, both of which had non-significant decreases in expression compared to PBS control mice. At 12 hours post infection the gene expression of IFNβ in mice inoculated with Mam-RRV-T48 was found to be significantly up-regulated compared to the gene expression in UV-Mam-RRV inoculated mice. All other groups were found to have similar expression levels to PBS inoculated mice. Expression levels of IFNβ in mice inoculated with Mam-RRV-T48 dropped from 3 fold at 12 hours post infection to 1 fold at 24 hours post infection when compared to PBS inoculated mice.

Mice inoculated with Mos-RRV-T48 however, maintained a fold change of 1.6 across the two time points and this was significantly higher than the expression of IFNβ in UV-RRV-T48 at 24 hours post infection.

132

The gene expression of IFNγ at 12 hours post infection was increased by

52 and 42 fold for Mam-RRV-T48 and Mos-RRV-T48 respectively when compared to PBS inoculated mice. This difference was not statistically significant due to a large variation in individual mice within the groups.

Expression levels of IFNγ in UV-RRV-T48 inoculated mice were comparable to

PBS inoculated mice. At 24 hours post infection gene expression of IFNγ was similar to PBS inoculated mice for all RRV inoculated mice. Gene expression of MIF at 12 hours was significantly reduced in mice inoculated with Mam-

RRV-T48 compared to mice inoculated with UV-Mos-RRV-T48 or PBS. At 24 hours post infection, the expression levels of Mam-RRV-T48 and Mos-RRV-

T48 inoculated mice were not significantly different to PBS inoculated mice.

However, mice inoculated with UV-Mam-RRV-T48 and UV-Mos-RRV-T48 were significantly lower than Mam-RRV-T48, Mos-RRV-T48 and PBS inoculated mice.

133

12 hour 24 hour

o *

i i 20 2.5

r r 2.0

p p 15

1.5

o o 10

C C L 2

d 1.0

5 l l 0.5

o o 0 0.0

N N Mam Mos Mam Mos PBS Mam Mos Mam Mos PBS T48 UV- T48 T48 UV- T48

n n

o o * i 5 i 2.5 * s * s

s s

e e r 4 r 2.0

p p

x x

e e 3 1.5

d d

l l

o o

f f

IFN

d 2 d 1.0

e e

s s

i i l 1 l 0.5

a a

m m

r r

o 0 o 0.0 N Mam Mos Mam Mos PBS N Mam Mos Mam Mos PBS T48 UV- T48 T48 UV- T48

n n

o o

i 100 i 8

s s

e e

r 80 r p p 6

x x

e e 60

d d

l l

o o 4

f f

IFN

d 40 d

e e

s s

i i 2 l 20 l

a a

m m

r r

o 0 o 0 N Mam Mos Mam Mos PBS N Mam Mos Mam Mos PBS T48 UV- T48 T48 UV- T48

n n

o * * o i 1.5 i 2.5

s s

s * s

e e * * r r 2.0

p p

x x * *

e 1.0 e 1.5

d d

l l * *

o o

f f

MIF

d d 1.0

e 0.5 e

s s

i i l l 0.5

a a

m m

r r

o 0.0 o 0.0 N Mam Mos Mam Mos PBS N Mam Mos Mam Mos PBS T48 UV- T48 T48 UV- T48

134

o i

i s

Figure 5.4s - Inflammatory cytokine profiles differ between mammalian and mosquito

e s

s 2 . 5 r

x derived RRV and also between active and inactive RRV in the ankle joint.

e M am

p r

2 . 5 e

p M am M o s p

d 2 . 0 2 *. 5 * e

C57Bl/6J wildx type 18 day old mice of mixed sex were injected subcutaneously in the pectoral 2 . 5 l

M am x M o s M am

2 . 5 e M am o 2 . 0 n *

d * * *

l o

region with 104 pfu/mouse of Mam-RRV-T48 ( ),M Mosam -RRV-T48 ( ),M UVo s-Mam-RRV-

i 2 . 0 1 . 5 d M o s * * M am M o s

2 . 0 o * l

d * * * s

d * * f

l M am

s o 1 . 5 M am * M o s * M o s P B S e T48 ( ), UV-Mos-RRV-T48 ( ) or with 50 µL PBS ( ). Mice were sacrificed at 12 o 2 . 0 * * f *

e * * *

f s

1 . 0 1 . 5 M o s

r i 1 . 5 M o s P B S e * *

d M am l * * p hours and 24 hours post infection and the ankle joint removed. Total RNA from the ankle d * *

s 1 . 0 P B S

e a

i P B S

l e

10..55 s 1 . 0 M o s

i s

1 . 0 m joint was isolated and analysed for mRNA expression by real time PCR. Data was normalised

a l

i * *

r l

0 . 5 l

a P B S

o a o to the house0 .keeping5 gene HPRT1 and expressed as relative expression to PBS inoculated

r 0 . 0 f

0 . 5 m

1 . 0

m o M a mr M os M a m M os P B S 0 . 0 d

r mice. Results are represented as the mean of each group (n=5) ± SEM. Statistical significance

e N o M a m M os M a m M o0s. 0 P B S

0 . 0 s T 48 U V - T 4 8

i l

N M a m M os M a m M os P B S M a m M os M a m M os P B S 0 . (*5 p<0.05) was determined by one way ANOVA with Tukey’s test. T 48 a U V - T 4 8

m T 48 U V - T 4 8

T 48 U V - T 4 8 r

o 0 . 0

N M a m M os M a m M os P B S T 48 U V - T 4 8

135

Comparison of gender in mouse model of RRV infection

The effect of gender on disease progression of RRV in a mouse model of disease was investigated due to inconsistent results obtained in the previous experiments in a mixed sex model. Male or female mice were inoculated with

104 pfu/mouse of Mam-RRV-T48, Mos-RRV-T48 or 50 µL PBS. Mice were then monitored for weight gain and disease symptoms as outlined in the clinical score for a period of 30 days.

Mice of both sexes treated with either Mam-RRV-T48 or Mos-RRV-T48 gained significantly less weight compared to PBS inoculated mice from day nine to day 18 post infection as seen in Figure 5.5A. Additionally, male mice inoculated with Mos-RRV-T48 were significantly different from PBS inoculated mice from day five to day 25 post infection. This group also had significantly less weight gain than male mice inoculated with Mam-RRV-T48 from day nine to 30 days post infection. A statistical difference between male and female mice inoculated with Mam-RRV-T48 was not seen until after 24 days post infection. Additionally, there was no significant difference in weight gain between female mice inoculated with either Mam-RRV-T48 or Mos-RRV-

T48.

Clinical scores of female mice inoculated with Mam-RRV-T48 or Mos-

RRV-T48 were comparable across the course of disease as seen in Figure

5.5B. The only exceptions occurred at days six and ten when mice inoculated with Mos-RRV-T48 had significantly higher clinical scores than Mam-RRV-

T48 inoculated mice. In contrast, male mice inoculated with Mos-RRV-T48 had significantly higher clinical scores than male mice inoculated with Mam-

136

RRV-T48 from day two to 17 post infection. When comparing mice inoculated with Mam-RRV-T48, male and female mice had comparable clinical scores across the time course with the exception of day three and four where the average clinical score was higher in male mice. In contrast, male mice inoculated with Mos-RRV-T48 had significantly higher clinical scores compared to female mice inoculated with Mos-RRV-T48 from two-18 days post infection.

The male mouse model of RRV yields reliable, repeatable statistically significant differences between Mam-RRV-T48 and Mos-RRV-T48. Mice inoculated with Mos-RRV-T48 consistently gain less weight than mice inoculated with Mam-RRV-T48 which is significant from around day 5 post infection and persists up to 30 days post infection, shown in Figure 5.6A. Mice inoculated with Mos-RRV-T48 also produce significantly higher clinical scores than mice inoculated with Mam-RRV-T48 which is apparent from around day three and continues for an average of 10 days post infection (Figure 5.6B). As a result of these finding all subsequent experimentation is completed in a male only mouse model.

137

A 300

) t 250

w 3

l *

200

(

h 150

100 * 2 * 1 5 10 15 20 25 30 Days post infection

C * B 6 6 *

c 4

S * S

l 2 l 2

0 0 0 5 10 15 20 0 5 10 15 20 Days post infection Days post infection

M a le a n d fe m a le m ic e in fe c te d w ith m o s q u ito a n d m a m m a lia n d e riv e d R R V

) ) M a le a n d fe m a le m ic e in fe c te d w ith m o s q u it ) o a n d m a m m a lia n d e riv e d R R V M t a le a n d fe m a le m ic e in fe c te d w ith m o s q u ito a n d m a m m a lia n d e riv e d R R V

t M a le a n d fe m a le m ic e in fe c te d w ith m o s q u ito a n d m a m m a lia n d e riv e d R R V t

) 3 0 0

h t h 3 0 0 3 0 0

h 3 0 0

g h

g Figure 5.5 - Male C57Bl/6 mice infected with mosquito derived T48 RRV lose more

g i

e weight and have higher clinical scores than female C57Bl/6 mice.

w P B S l P B S

C57Bl/6J wild type 18 day old male and femalew mice were injected subcutaneously in the

P B S a

F Mam l F Mam

l P B S

i a

t F Mam

a a i M mam 2 4

i pectoral region with 10 pfu/mouse of Mam-RRV-T48 (male orM femalemam ),F Mos Mam-RRV-T48

i i

t *

t n F Mos M mam t

3 0 0 i 3 F Mos i

i M mam

n (male orM femaleMos ), For M withos * 50 µL PBS (mixed sex )..P MiceB S were monitored daily for A) n

f M Mos n

i F Mos

o M Mos

2 0 0 f M Mos weights and B, C) disease progression. Diseasef progressionIn a c t i wasv e v e scoredro T 4 8 according to the

2 0 0 f

o %

2 0 0 o

) ( In 2a0c0tiv e C 6 /3 6 T 4 8

clinical disease score data and is represented as the mean of each group (n=5) ± SEM.

(

( i

( Statistical significance (* p<0.05) was determined by two wayV e rANOVAo T 4 8 with Tukey’s test. Data

(

n g i C 6 /3 6 T 4 8

is representative of two separately run experiments. n

2 0 0 i

e t

g 1 0 0

g e

i 1381 0 0

i W

e 0 0 0

1 2 3 e 1 0 0 W 1 0 0

1 0 0 W D a y s p o s t i n f e c t i o n W 0 0 0 1 2 3 1 0 0 0 * 1 2 30 0 0 D a y s p o s1 t i n f e c t i o n 2 3 0 1 0 2 0 D a y s p o3 0s t i n f e c t i o n D a y s p o s t i n f e c t i o n D a y s pos t infe c tion A 250 * 2

)

w 200

( 150

W 100 * 1

5 10 15 20 25 30 Days post infection

B 8

* 6

a 4

0 2 10 20 30 * Days post infection 3 0 0 3 2 * P B S * 3 0 0 3 * P B S In a c tiv e v e ro T 4 8 Figure 5.6 - Difference in disease and weights of male C57Bl/6 mice inoculated with

T48 RRV is highly reproducible. In a c tiv e v e ro T 4 8 In a c tiv e C 6 /3 6 T 4 8 ) In a c tiv e C 6 /3 6 T 4 8 V e ro T 4 8

% Male C57Bl/6J wild type 18 day old mice were injected subcutaneously in the pectoral region ( 2 4

) with 10 pfu/mouse of Mam-RRV-T48 ( ),V Mose ro -TRRV4 8 -T48 ( ), or with 50 µL PBS n 2 0 0 C 6 /3 6 T 4 8 * i

3 0 0 3

% a

* ( ( ). P B MiceS were monitored daily for A) weights and B) disease progression. Disease

C 6 /3 6 T 4 8

2 0 0 g

n i

t progressionIn a c tiv wase v escoredro T 4 8according to the clinical disease score data is represented as the mean

g ) In a c tiv e C 6 /3 6 T 4 8

i of each group (n=5) ± SEM. Statistical significance (* p<0.05) was determined by two way

% h

( V e ro T 4 8

g ANOVA with Tukey’s test. Data is representative of five separately run experiments.

n e

i 2 0 0 C 6 /3 6 T 4 8 a

W 1 0 0 g

139 t 1 0 0 h 1

g * i

1 e

0 * 1 0 2 0 3 0 W 1 0 0 0 1 0 D 2a0y s pos t inf3e0c tion D a y s pos t infe c tion 1 * 0 1 0 2 0 3 0 D a y s pos t infe c tion Viral titres in mice inoculated with Mam-RRV-T48 or Mos-

RRV-T48 in mice

C57Bl/6 mice were inoculated with 104 pfu/mouse of Mam-RRV-T48,

Mos-RRV-T48 or 50 µL of PBS, samples from blood, quadriceps muscle, ankle joint and brain were then taken at 24 hours post infection and at peak disease

(seven days) and homogenised in PBS. After clarification, samples were titrated on Vero cells to enable viral titre comparison between Mam-RRV-T48 and Mos-RRV-T48 inoculated mice.

At 24 hour post infection blood titres of Mam-RRV-T48 and Mos-RRV-

T48 in mice were comparable in blood samples with a combined mean of 7.397

± 0.0523 log pfu/mL as seen in Figure 5.7A. At peak disease titres between

Mam-RRV-T48 and Mos-RRV-T48 inoculated mice were still similar however the mean titre had reduced to 1.947 ± 0.1329 log pfu/mL (Figure 5.7E). RRV titres in the both the quadriceps and ankle joint were significantly higher in

Mam-RRV-T48 inoculated mice compared to Mos-RRV-T48 inoculated mice at 24 hours as shown in Figure 5.7B and D, however titres recovered from

Mos-RRV-T48 mice had risen by peak disease to be significantly higher than

Mam-RRV-T48 inoculated mice in both tissues as seen in Figure 5.7F and G.

In brain tissue there was no difference in RRV titres in mice inoculated with either Mam-RRV-T48 or Mos-RRV-T48 at 24 hours post infection (Figure

5.7D). At peak disease RRV titres were not recoverable in brain tissue (Figure

5.7H).

140

Blood 24 hours Blood peak disease A E 8 8

) 7 )

L 7

6 m / 6

f 5 f

p 5

g 4 g 4

(

t 2

T 1 T 1 0 0 Mam Mos Mam Mos T 48 T 48 T 48 T 48

Quadriceps peak disease B Quadriceps 24 hours F 8 * 8 *

) ) 7 7

/ / 6 6

o o 4 4

(

3 3

i i 2 2

T T 1 1 0 0 Mam Mos Mam Mos T 48 T 48 T 48 T 48

Ankle 24 hours Ankle peak disease C G 8 * 8 *

) 7 ) 7

g / 6 / 6

p p

5 5

g g

o 4 o 4

( ( 3 3

e e

i 2 i 2

T 1 T 1 0 0 Mam Mos Mam Mos T 48 T 48 T 48 T 48

Brain 24 hours Brain peak disease D H 8 8

) 7 ) 7

g g / 6 / 6

u u

f f

p p

5 5

g g

o 4 o 4

l l

( ( 3 3

e e

r r

t t

i 2 i 2 T 1 T 1 # # 0 0 Mam Mos Mam Mos T 48 T 48 T 48 T 48

142

o s

i Figure 5.7 - Blood and tissue virus titres.

s p e 2 . 5 r Male C57Bl/6J wild type 18 day old mice were injected subcutaneously in the pectoral region

x M am p

2 . 5 e

x 4 with 102 . pfu/mouse0 of Mam-RRV-T48 ( ) M ora m Mos-RRV-T48 ( ).M oA,s E) Blood, B, F)

d * *

e l M o s M am

2 . 0 o

d * quadriceps,* C, G) ankle and D, H* ) brain were* harvested at 24 hours and 7 days post infection.

l f

1 . 5 M am M o s

o * * * * f Tissuesd were homogenised and clarified before being titrated on Vero cells. Virus titre is 1 . 5 M o s P B S

e * * d

s 1 . 0 P B S i

e represented as the mean titre for each group (n=5) ± SEM. Statistical significance (* p<0.05) l

s 1 . 0

i a

l was determined0 . 5 by Mann-Whitney t test. # denotes that value was under the limit of detection

a m

0 . 5 r

m (0.6 log pfu/g). o

r 0 . 0 o 0 . 0 N M a m M os M a m M os P B S

N M a m M os M a m M os P B S T 48 U V - T 4 8 T 48 U V - T 4 8

143

Histology of mice infected with Mam-RRV-T48 or Mos-

RRV-T48

C57Bl/6J mice were inoculated with 104 pfu/mouse of either Mam-RRV-

T48, Mos-RRV-T48 or 50 µL of PBS. Mice were sacrificed at 24 hour post infection and at peak disease (seven days post infection), as determined by clinical score and weight data. Samples of the quadriceps muscle, ankle joint and brain were taken and fixed in buffered formalin and prepared for microscopic examination as outlined in Chapter 2. Histological sections were then examined to not only confirm differences from PBS controls but also to see if differences were apparent between Mam-RRV-T48 and Mos-RRV-T48 samples.

In general, mild histological changes could be detected in quadriceps and ankle tissue samples at 24 hours post infection; by peak disease changes in tissue due to inflammation had increased with inflammatory cells present and areas of necrotic tissue. Histological findings in the brain were mild but present with indications of encephalitis in RRV infected mice.

Quadriceps muscle histology of RRV infected mice

Sections taken from the quadriceps muscle of RRV infected mice at 24 hours post infection had an increase in mononuclear cells when compared to

PBS controls (Figure 5.8). In addition, necrotic fibres were present in mice inoculated with Mos-RRV-T48. At seven days post infection, there was an increase in mononucleated inflammatory cells in both Mam-RRV-T48 and

Mos-RRV-T48 inoculated mice compared with respective sections at 24 hours post infection and to PBS controls (Figure 5.9).

144

Figure 5.8 - Histology of quadriceps muscle in mice infected with RRV at 24 hours post infection.

C57Bl/6J wild type 18 day old mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of Mam-RRV-T48, Mos-RRV-T48 or 50 µL PBS. Samples were taken from quadriceps muscle at 24 hours post infection and fixed in formalin. Tissue was paraffin embedded, sectioned and stained with haematoxylin and eosin. Data is representative of each group (n=4) ± SEM.

145

Figure 5.9 - Histology of quadriceps muscle in mice infected with RRV at peak disease.

C57Bl/6J wild type 18 day old mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of Mam-RRV-T48, Mos-RRV-T48 or 50 µL PBS. Quadriceps muscle samples were taken at peak disease as indicated by clinical score and weight and fixed in formalin. Tissue was paraffin embedded, sectioned and stained with haematoxylin and eosin Data is representative of each group (n=4) ± SEM.

146

Sections from Mam-RRV-T48 inoculated mice had increased extracellular material compared to control (PBS) as well as some irregularity and swelling of fibre shape. Degeneration of muscle cells is evident but signs of cell regeneration are also present as indicated by central nucleation in a small number of cells. In the Mos-RRV-T48 sample, there was a large increase in extracellular material, including fibrotic tissue morphology. Muscle fibre size was more variable with several areas of atrophic fibres. Fibrosis was visible as were degenerated fibres.

Ankle joint histology of RRV infected mice

At 24 hours post infection ankle sections stained with H&E for both

Mam-RRV-T48 and Mos-RRV-T48 inoculated mice had minimal thickening of the synovial membrane but otherwise were comparable to PBS (Figure 5.10).

At peak disease, tissue and bone destruction are present in both Mam-RRV-

T48 and Mos-RRV-T48 inoculated mice, however the sections from mice inoculated with Mos-RRV-T48 were more severely affected (Figure 5.11).

There was evidence of hyperplasia in the Achilles tendon as well as degeneration of the tendon itself in RRV infected mice; additionally, the spongy bone sections had evidence of degradation and inflammatory infiltrates around the area of bone and cartilage. There was an increase in mononuclear cells and areas of fibrosis.

147

148

Figure 5.10 – Histology of ankle joint in mice infected with RRV at 24 hours post infection.

C57Bl/6J wild type 18 day old mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of Mam-RRV-T48, Mos-RRV-T48 or 50 µL PBS. Samples were taken from the ankle joint at 24 hours post infection and fixed in formalin. Tissue was paraffin embedded, sectioned and stained with haematoxylin and eosin. B) bone, C) cartridge, G) growth plate, M) muscle, P) periosteum, S) synovia/synovial space. Data is representative of each group (n=4) ± SEM.

149

150

Figure 5.11 – Histology of ankle joint in mice infected with RRV at peak disease.

C57Bl/6J wild type 18 day old mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of Mam-RRV-T48, Mos-RRV-T48 or 50 µL PBS. Samples were taken from the ankle joint at peak disease and fixed in formalin. Tissue was paraffin embedded, sectioned and stained with haematoxylin and eosin. B) bone, C) cartridge, G) growth plate, M) muscle, P) periosteum, S) synovia/synovial space. Data is representative of each group (n=4) ± SEM.

151

Brain histology of RRV infected mice

Examination of brain tissue sections from RRV infected mice contained signs of inflammation at 24 hours in Mos-RRV-T48 inoculated mice (Figure

5.12). There was an increase in mononucleated cells as well as necrotic neurons present which are indicators of encephalitis. Perivascular cuffing was seen in both Mam-RRV-T48 and Mos-RRV-T48 samples but was most predominant in Mos-RRV-T48 samples. At peak disease there was evidence of prior inflammation with several necrotic neurons undergoing phagocytosis.

In Mos-RRV-T48 samples, eosinophilic vacuoles and embolus could be observed in the brain tissue.

152

Figure 5.12 - Histology of brain tissue in mice infected with RRV at 24 hours and peak disease.

C57Bl/6J wild type 18 day old mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of Mam-RRV-T48, Mos-RRV-T48 or 50 µL PBS. Samples were taken from the brain at 24 hours post infection and at peak disease as determined by clinical scores and weight gain data and fixed in formalin. Tissue was paraffin embedded, sectioned and stained with haematoxylin and eosin. Data is representative of each group (n=4) ± SEM

153

Cytokine profiles of mice inoculated with Mam-RRV-T48 or Mos-RRV-T48

Gene expression of pro-inflammatory and anti-inflammatory cytokines in male mice after infection with Mam-RRV-T48 and Mos-RRV-T48 was examined in the quadriceps muscle and ankle joint by real time PCR.

Samples were taken at 6, 12 and 24 hours post infection and at peak disease, determined as seven days, by information taken from mouse weights and clinical scores. All results were normalised against the HPRT1 gene and are relative to expression levels in PBS inoculated mice.

Cytokine profiles in quadriceps muscle

At six hours post infection, mice infected with Mam-RRV-T48 had significantly higher gene expression levels of IL4 and IL12 compared to mice inoculated with either Mos-RRVT48 or PBS as seen in Figure 5.13A. In contrast, Mos-RRV-T48 inoculated mice had significantly higher gene expression of CCL2 and IFNβ compared to both Mam-RRV-T48 and PBS inoculated mice and significantly higher gene expression of IFNα compared to

Mam-RRV-T48 inoculated mice.

At 12 hours post infection gene expression of IFNβ in Mos-RRV-T48 inoculated mice was increased 3 fold compared to PBS inoculated mice, shown in Figure 5.13B. Additionally gene expression levels of IFNα were still significantly higher in Mos-RRV-T48 inoculated mice compared to Mam-RRV-

T48 mice at 12 hours but at a lower level than that at 6 hours at 2.6 and 4.9 fold respectively. In both Mam-RRV-T48 and Mos-RRV-T48 inoculated mice

154

IL12 gene expression was reduced compared to expression in PBS inoculated mice.

Interestingly, at 24 hours post infection IFNβ gene expression was significantly reduced in Mam-RRV-T48 inoculated mice compared to both

Mos-RRV-T48 and PBS inoculated mice (Figure 5.13C). Gene expression of

IFNγ was also reduced in Mam-RRV-T48 inoculated mice at 24 hours when compared to PBS inoculated mice. Both Mam-RRV-T48 and Mos-RRV-T48 inoculated mice were found to be significantly up-regulated for IL6 gene expression compared to PBS inoculated mice, whilst Mos-RRV-T48 inoculated mice had reduced expression levels of CXCR3 compared to PBS controls.

At peak disease expression levels of IFNβ, IFNγ and IL12 were significantly up-regulated in Mam-RRV-T48 inoculated mice compared to both Mos-RRV-T48 and PBS inoculated mice as shown in Figure 5.13D.

Additionally, Mam-RRV-T48 inoculated mice also had increased expression levels of IL4 and IL6 compared to PBS inoculated mice.

155

A 100 * *

l 50

m * *

* * * * * * 0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL6 IL 10 IL12 MIF T NF 

B 4 *

* *

i 3

l 2

n * * e *

e 1

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL6 IL 10 IL12 MIF T NF 

156

C 6

s * *

r 4

v 2

a l * * *

R *

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL6 IL 10 IL12 MIF T NF 

D20

* *

i 15

l 10

a * *

o * *

i *

a *

e 5

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL12 MIF T NF 

157

n o

s s

p o

i 2 . 5

x n

Figures 5.13 - Cytokine profiles in the quadriceps muscle of mice inoculated with M am

s i

mammalian and mosquito derived RRV. M o s

s e

d 2 . 0 * *

l r

s M am

o p e * *

2 . 5 f

r C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region x

1 . 5 M am M o s

p e

2 . 5 d * * x with 104 pfu/mouse of Mam-RRV-T48 ( ),M Mosam-RRV-T48 ( ),M oro swith 50 µL PBS ( P B S 2 . 0 e

d * *

s l

1 . 0

i M o s M am

2 . 0 o l

d * ). Mice* were sacrificed at A*) 6 hours,* B) 12 hours, C) 24 hours post infection and at D) peak

f a

1 . 5 M am M o s

o * * 0 . 5 * *

f d

disease and the quadriceps m muscle was removed. Total RNA from the quadriceps muscle was

1 . 5 M o s P B S e

* * r d

s 1 . 0

o P B S i

e isolated and analysed for mRNA0 . 0 expression by real time PCR. Data was normalised to the l

s 1 . 0

N i a M a m M os M a m M os P B S

l house0 . 5 keeping gene HPRT1 and expressed as relative expression to PBS inoculated mice. a m T 48 U V - T 4 8 0 . 5 r

m Results are represented as the mean of each group (n=5) ± SEM. Statistical significance (* o

r 0 . 0 o 0 . 0 p<0.05)N wasM adeterminedm M os by oneM a waym ANOVAM os withP BTukeyS ’s test. N M a m M os M a m M os P B S T 48 U V - T 4 8 T 48 U V - T 4 8

158

Cytokine profiles in ankle joint

At six hours post infection, gene expression levels of most cytokines in the ankle joint of Mam-RRV-T48, Mos-RRV-T48 and PBS inoculated mice were similar, as shown in Figure 5.14A. The only exception to this was expression levels of IL10 which were elevated in both Mam-RRV-T48 and Mos-

RRV-T48 inoculate mice, however, only Mos-RRV-T48 inoculated mice were found to have a significant difference compared to PBS inoculated mice.

Gene expression levels of IL10 were similar at 6 and 12 hours post infection and Mos-RRV-T48 inoculated mice were again significantly up- regulated compared to PBS inoculated mice. Mice inoculated with Mos-RRV-

T48 had up-regulated gene expression of both IFNα and IFNβ compared to

Mam-RRV-T48 and PBS inoculated mice at 12 hours post infection (Figure

5.14B). Additionally, both Mam-RRV-T48 and Mos-RRV-T48 inoculated mice had significantly down-regulated levels of IL4 gene expression when compared to PBS inoculated mice at 12 hours post infection.

At 24 hours post infection both Mam-RRV-T48 and Mos-RRV-T48 inoculated mice had significantly higher gene expression of IL10 when compared to PBS inoculated mice, and this was also the case for TNFα gene expression (Figure 5.14C). Mam-RRV-T48 inoculated mice expressed significantly higher levels of IL4 and significantly lower levels of CCL2 than

PBS inoculated mice at 24 hours post infection.

At peak disease, seven days post infection, IFNβ expression was significantly higher in Mam-RRV-T48 inoculated mice when compared to Mos-

RRV-T48 and PBS inoculated mice, as seen in Figure 5.14D. In addition,

159

Mam-RRV-T48 inoculated mice had significantly higher expression levels of both CXCR3 and IL6 compared to PBS inoculated mice at peak disease. IFNγ and IL12 gene expression levels were found to be significantly higher in Mos-

RRV-T48 inoculated mice compared to controls at peak disease.

160

A 5

* 4

o 2

e R 1

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL6 IL 10 IL12 MIF T NF 

B 5

* *

4 * *

x *

o 2

a l * *

e R 1

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL6 IL 10 IL12 MIF T NF 

161

C 15

o i * *

r 10

v 5

a l * * e * *

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL6 IL 10 IL12 MIF T NF 

D 60 * *

r 40

i *

v 20

a *

* * *

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL12 MIF T NF 

162

n 2 . 5 o

i M am s

Figure 5.14 - Cytokine profiless in the ankle joint of mice inoculated with mammalianM o s

e 2 . 0 * *

and mosquito derived RRV. r p * * M am

2 . 5 x

n e

C57Bl/6Jo wild type 18 day old male1 . 5mice were injected subcutaneouslyM am in the pectoral regionM o s

i *

2 . 5 d *

s n

4 l s

o with 10 pfu/mouse of Mam-RRV-T48 ( ),M Mosam -RRV-T48 ( ),M oro s with 50 µL PBS ( ).P B S o i 2 . 0 *

e *

r 1 . 0

s M o s M am d 2 . 0 Mice werep * sacrificed at A) 6 hours, B)* 12 hours, C) 24 hours post infection and at D) peak

e * *

r s e 1 . 5 M am p M o s disease and the ankle joint was removed.i Total RNA from the ankle was isolated and analysed

* * l

x 0 . 5 * *

a l e 1 . 5 M o s P B S

for mRNA expression by real time PCR. Data was normalised to the house keeping gene o

* * m

r l

1 . 0 P B S

o o

HPRT1 andd expressed as relative expression0 . 0 to PBS inoculated mice. Results are represented

f e

1 . 0

N s

d M a m M os M a m M os P B S

as the meani of each group (n=5) ± SEM. Statistical significance (* p<0.05) was determined l

e 0 . 5

a s i T 48

l 0 . 5 by one way ANOVA with Tukey’s test. U V - T 4 8

r o

m 0 . 0

r N o 0 . 0 M a m M os M a m M os P B S

N M a m M os M a m M os P B S T 48 U V - T 4 8 T 48 U V - T 4 8

163

Discussion

Whether there are differences in disease profiles of mice infected with mammalian or mosquito derived viruses has been a controversial issue. As only the initial viral inoculum in the human host is mosquito derived, it has been suggested that this would have no effect on disease progression as all replication rounds of the virus in the host would be mammalian derived.

However, differences in the induction of cytokines has been reported in a cell culture environment (Hidmark et al., 2005; Shabman et al., 2007). Of particular interest is the report that Mam-RRV-T48 and not Mos-RRV-T48 is a potent inducer of IFNβ. It has been suggested that IFNβ is important in the development of RRV symptoms in response to viral infection. We sought to confirm that IFNβ was required for disease development in a model of RRV disease prior to testing if this was differentially induced by Mam-RRV-T48 and

Mos-RRV-T48 in vivo.

IFNβ is critical for development of disease symptoms

Wt and IFNAR -/- mice were inoculated with 104 pfu virus. In Wt mice, this caused the expected disease profile comparable to previously published work using this particular model. In IFNAR -/- mice, the virus caused acute viremia which overloaded the system resulting in complete fatality in the group within 48 hours. This reaction was expected as the knockout mice are missing a critical part of the innate immune system that responds to invading pathogens. To overcome this, the viral load in the knockout mice was reduced to 101 pfu with interesting results. Whilst this amount would not be sufficient

164 to cause any significant difference in either the percentage of weight gain or the clinical score in the Wt mice, it did significantly affect the weight gain of the IFNAR -/- mouse. In addition to this, the viral titre in the blood of infected mice was comparable between Wt mice inoculated with 104 pfu RRV and

IFNAR -/- mice inoculated with 101 pfu. Even more interesting, is that despite these similarities IFNAR -/- mice did not develop any arthritic disease symptoms. This strongly indicates a role for IFNβ in the development of inflammation in the joints and resulting tissue destruction that is responsible for many of the symptoms of RRV disease. IFNβ has traditionally been associated with a reduction of inflammation and has been inferred as a treatment option for rheumatoid arthritis. However, recent studies have linked IFNβ treatment with the development of psoriatic arthritis and the up- regulation of the TH17 response in human patients (Toussirot et al., 2014;

Triantaphyllopoulos et al., 1999). It is likely that IFNβ plays a dual role in prevention or initiation of arthritic symptoms depending on timing and the magnitude of the response. Traditionally, IFNβ is associated with the reduction of pro-inflammatory cytokines such as IL12 and TNFα (Smeets et al., 2000). It would therefore have been expected that these mice would develop a higher level of disease without the ability to moderate the inflammatory reaction. The results from this study indicated that further investigation was warranted and as such, the cytokine profiles were examined to get a better understanding of the inflammatory responses that may underlie the observed effect.

165

UV inactivated virus does not cause arthritic disease symptoms in the RRV mouse model of disease.

To ensure that appropriate controls were utilised in all experimental work and to confirm that any effects seen required replicating virus, UV inactivated virus was tested in the mouse model. Findings show that whilst there is a mild initial reaction from the introduction of UV inactivated virus this only lasts briefly, confirming that replication of virus is required for disease symptoms and that simple interactions with the different glycosylations present on Mam-RRV-T48 and Mos-RRV-T48 are not enough to initiate a differential immune response.

Interestingly, whilst there were no arthritic disease symptoms seen in mice inoculated with UV inactivated virus, the animals did consistently gain more weight than the PBS control by the end of the 30 day period, however this was significant in only one instance. Mice inoculated with UV inactivated virus did appear to be lethargic in comparison to PBS controls, however, this was a subjective observation and difficult to measure in the absence of other indicators such as ruffled fur. It is hypothesised that although the introduction of the UV inactivated virus did not cause any arthritic disease symptoms it was enough to initiate some immune response which, in turn, made the mice feel unwell with general viral symptoms such as lethargy resulting in less movement in these mice. However, as they were not movement restricted, due to arthritic limbs, as in the case with the mice inoculated with active RRV, they continued to consume food and water at the normal rate. C57Bl/6J mice are often used for obesity studies as they will

166 eat more than their requirement and are prone to becoming overweight.

Taken together these factors would explain the consistently higher weights in this group. If the effects of host contact to inactivated virus were of interest this may be a useful measure to include in future animal models and may warrant further investigation, however they were not the focus of this study.

Mosquito derived RRV causes a decrease in percentage weight gain and an increase in clinical score in the RRV mouse model of disease.

The mouse model was used to investigate if differences in disease occurred in mice injected with Mam-RRV-T48 when compared to Mos-RRV-

T48. It was discovered that differences were apparent in both percentage of weight gained and in clinical score. This was generally reproducible but in one instance, the differences were not statistically significant, mostly due to high error bars caused by variations in individual mice within each group. It was observed in the course of running these experiments that male mice appeared to gain less weight and had higher clinical scores than female mice and this was further investigated to rule out a potentially overlooked variable.

Cytokine profiles of RRV disease depend on viral cell origin and differ from inactivated virus.

Cytokine profiles for CCL2, IFNβ, IFNγ and MIF were tested to determine if the inflammatory response may be different between Mam-RRV-T48, Mos-

RRV-T48, UV-Mam-RRV-T48, UV-Mos-RRV-T48 and PBS inoculated mice.

167

Differences between active and inactive virus were seen in every group, except

MIF at 12 hours post infection, but were not always found to be significant.

This indicated that replicating virus is required for a full immune response to be initiated by the host and that these factors are not dependent on binding alone. Of the findings that were significant, IFNβ was significantly up- regulated at 12 hours in both quadriceps and ankle samples in Mam-RRV-

T48 inoculated mice, and both CCL2 and IFNγ were up-regulated in the ankle in Mam-RRV-T48 inoculated mice.

The up-regulation of these cytokines is somewhat contradictory with

IFNγ being associated with a TH1 response and CCL2 being attributed to driving a TH2 response, as well as the infiltration of monocytes (Antonelli et al., 2009). However, recent studies into psoriatic arthritis have shown an increase in TH1 cytokines along with high circulating levels of CCL2 and interferon γ-inducible protein (IP-10) in early stages of disease with a further increase in CCL2 and a decrease in IP-10 occurring later in disease and indicating a shift to the TH2 response for chronic conditions (Devito, 2014).

Differences in gene expression of cytokines between Mam-RRV-T48 and Mos-

RRV-T48 infected mice were not clear, primarily due to large variation occurring within each group. However, this variation was in part suspected to be due to differences in cytokine expression between male and female mice in response to infection and thus warranted further investigation on the effect of Mam-RRV-T48 and Mos-RRV-T48 infection in male and female mice.

168

Results gained from the RRV mouse model are sexually dimorphic.

Mice used for the RRV mouse model of disease are inoculated with virus at between 18 and 21 days of age which is before they become sexually active.

As such, sex has not been considered to be an issue in the development of disease in these mice; however, due to the reproducible observation of males in the lowest weight group and displaying the most severe clinical symptoms in this study, it was important to rule out sex as a variable in the mouse model. The results showed minimal difference between female mice inoculated with either Mam-RRV-T48 or Mos-RRV-T48 but a significant difference in male mice. This difference was seen in both the percentage of weight gained and the clinical score over the 30 day observation period with male mice inoculated with Mos-RRV-T48 gaining less weight and having higher clinical scores than male mice inoculated with Mam-RRV-T48.

Many studies have shown that sexual dimorphism exists in relation to inflammation and response to viral infection (Barna et al., 1996; Choudhry et al., 2006; Geurs et al., 2012; Marriott and Huet-Hudson, 2006; Pfeilschifter et al., 2002). This has mostly been linked to the oestrus cycle and differences in hormonal levels in females but some studies are elucidating a role for other immune response factors such as cytokines (Marriott and Huet-Hudson,

2006; Pfeilschifter et al., 2002). Differences in cytokine expression, as well as hormonal factors, are likely at play in the development of RRV in the mouse model and offers some level of protection to female mice against the factors in

Mos-RRV-T48 which cause more severe disease in males. Future studies need

169 to examine the cytokine and hormone profiles of female and male mice inoculated with Mos-RRV-T48 with the possibility of increasing the oestrogen in male mice to examine the effect on disease progression.

Male mice show greater reproducible variation when inoculated with Mam-RRV or Mos-RRV.

The male mouse model of RRV disease was repeated several times to ensure reproducibility and consistency of results. Each time the experiment was run, it produced similar results with a significant difference between the weight gain and clinical scores of Mam-RRV-T48 and Mos-RRV-T48 inoculated mice. As the aim of this study was to investigate differences between mammalian and mosquito derived virus the male model of disease was chosen, however as mentioned previously, investigation into why these differences occur in male and female mice would make an interesting future study. All further experimentation completed in the mouse model in this thesis was done using only male mice to reduce variability and to examine possible causes for the differential effect between mammalian and mosquito derived virus.

Viral titres are higher in quadriceps muscle and ankle joint of Mos-RRV-T48 inoculated mice at peak disease

RRV titres recovered from blood and brain samples of Mam-RRV-T48 and Mos-RRV-T48 inoculated mice were comparable indicating a similar level of infection. Interestingly, however, higher titres were recovered in Mam-RRV-

170

T48 inoculated mice in quadriceps and ankle joint at 24 hours post infection compared to Mos-RRV-T48 inoculated mice. This was unexpected as it was thought that increased disease severity would correlate with an increase in early viral titres. However, at peak disease, titres recoverable from the tissue of Mos-RRV-T48 inoculated mice had increased and were significantly higher than Mam-RRV-T48 inoculated mice. This indicates that Mos-RRV-T48 has a replication advantage over Mam-RRV-T48 and this must be initiated by early immune interactions. One possibility is that differential cytokine profiles activated by either Mam-RRV-T48 or Mos-RRV-T48 result in differences in disease progression due to reduced clearance of Mos-RRV-T48 from inoculated mice.

Mice have different histology when inoculated with Mam-

RRV or Mos-RRV

Changes in histology of mice inoculated with RRV could be visualised as early as 24 hours which was not expected. Both Mam-RRV-T48 and Mos-

RRV-T48 infected mice show signs of immune response in muscle and ankle, whilst evidence of immune response in the brain is most evident in Mos-RRV.

At peak disease, RRV infected tissue can be clearly differentiated from uninfected mouse control tissue and it is also possible to see that the level of tissue destruction and inflammation is greater in the Mos-RRV-T48 mice. At both time points there is an increase in mononucleated cells and it is likely that the activation states of these cells plays and important role in the cytokine profile in each tissue.

171

Mice have different cytokine profiles when inoculated with Mam-RRV or Mos-RRV

Examination of the cytokine profiles of C57Bl/6J mice inoculated with either Mam-RRV-T48 or Mos-RRV-T48 clearly demonstrated that different gene expression levels are induced depending on the cell line of origin of the virus. At six hours post infection CCL2 was expressed to high levels in Mos-

RRV-T48 inoculated mice but unexpectedly IL4 levels, which are often up- regulated in the presence of CCL2, were not up-regulated in Mos-RRV-T48 inoculated mice but were up-regulated in Mam-RRV-T48 inoculated mice.

There was also an up-regulation of type I IFN at both 6 and 12 hours post infection in Mos-RRV-T48 inoculated mice. This was not expected as in vitro studies have shown that Mam-RRV-T48 and not Mos-RRV-T48 is a potent inducer of type I IFN. However, at peak disease the levels of type I IFN had returned to expression levels similar to that in PBS inoculated mice; there was an increase in the gene expression of IFNβ in Mam-RRV-T48. It is possible that IFNβ is expressed at earlier time points in Mos-RRV-T48 inoculated mice and at later time points in Mam-RRV-T48 inoculated mice.

This is supported by the cytokine studies completed in the ankle joint.

At 12 hours post infection gene expression of IFNα and IFNβ were up- regulated in Mos-RRV-T48 inoculated mice, returning to normal levels by 24 hours post infection. At peak disease, IFNβ was up-regulated in Mam-RRV-

T48 inoculated mice compared to Mos-RRV-T48 inoculated mice.

Interestingly, IL10 was up-regulated in ankle joint in both Mam-RRV-T48 and

Mos-RRV-T48 inoculated mice but was not up-regulated in quadriceps

172 muscle. This, along with other differences in cytokines indicates that the cytokine response is tissue specific in response to RRV infection as has been proposed by previous studies (Stoermer Burrack et al., 2014).

Conclusion

The mouse model of RRV has been re-evaluated and refined to a single sex male mouse model; this reduces error whilst increasing robustness and reproducibility which is an important consideration in animal experiments.

The data reported in this chapter has also shown that active virus is required for the progression of infection to disease symptoms and the up-regulation of

TH1 cytokines. It was also shown that Mam-RRV-T48 and Mos-RRV-T48 infected mice have different levels of weight gain post inoculation and also different clinical score, histology and cytokine profiles. A deeper understanding of this difference in host response is vital for understanding

RRV pathogenesis.

173

Characterisation of Ross River virus isolates in a mouse model of disease

Introduction

The in-vivo immune system is a complex balance between resistance to foreign entities and maintenance of self. As a result, studies in isolated cell systems alone do not allow all possibilities to be taken into consideration. For this reason when looking at effects on disease progression that occur downstream from initial infection, investigations in a mouse model offers invaluable insights into the multitude of factors at play in a host after viral infection. Another important consideration is the potential for different host immune reactions in response to different viral isolates within the same species. Genomic variation between isolates has been shown to be an important factor in disease outcome of alphaviruses (Herrero et al., 2014;

Jupille et al., 2011; Stoermer Burrack et al., 2014).

This is illustrated by the mouse avirulent RRV isolate DC5692, isolated from mosquitos in WA. Studies found that replacing the NsP1 region of RRV

T48 with that of DC5692 resulted in attenuated disease (Jupille et al., 2011).

Additional studies expanded on this using RRV T48 and RRV-T48-nsP16M, a

T48 mutant carrying the six non-synonymous DC5692 nucleotide differences in the NsP1 region. When mice were inoculated with equal amounts of these viruses, in addition to attenuation of disease for mice inoculated with RRV-

T48-NsP16M, viral load in the skeletal muscle of these mice was found to be

175 significantly lower than in RRV-T48 mice at 5 days post infection. It was also found that these mutations made RRV-T48-NsP16M more sensitive to type I

IFN in vitro. It is possible that type I IFN is able to control RRV infection in a tissue specific manner (Stoermer Burrack et al., 2014). Additionally, studies using BFV, another Alphavirus, showed that different isolates resulted in significant differences in disease, with mortality ranging from 100 % to 19.2

% in C57Bl/6J mice inoculated with 104 pfu/mouse for a range of BFV isolates. This illustrates that mutations occurring in the natural population can have an effect on disease outcomes, supporting the idea of differences in immune reaction within the host to different isolates of the same virus.

In this study, RRV isolates were obtained from circulating mosquito populations in WA and are an example of current variations which exist in nature. It is possible that variation is an evolution strategy to enhance isolates fitness in available local hosts, for example the infections of macropods in rural areas and domestic animals in urban areas. Studies have shown that repeated passaging of alphavirus in tissue culture makes it more virulent in that cell type with a trade off in virulence in other cell types(Dubuisson et al., 1997; Greene et al., 2005; Heil et al., 2001).

Additionally, passaging in avian cells has been shown to cause changes at the

E2 glycosylation site, resulting in modified growth patterns in mice for two of the three variants tested. The specific mutations, E2 substitutions at either residue 4 or 218, could not be replicated by passage in human or mosquito cell lines (Kerr et al., 1993). The effect of mutations after passaging was also reported by Lidbury et al., (2011), where a variant, RRVPERS, was able to

176 establish a persistent infection in the macrophage cell line RAW 264.7.

RRVPERS had a small plaque phenotype, was resistant to IFNβ, could attenuate expression of IFNβ and exhibited a significantly enhanced disease in mice.

RRVPERS was found to have several nucleotide differences in the NsP region and also a point mutation in the E2 gene.

No studies have so far addressed a correlation between tissue culture kinetics and in-vivo mouse disease kinetics for RRV field isolates and a better understanding of the link between tissue culture results and those seen in vivo is important in order to minimise animal experimentation.

Results

Weight gain in mice inoculated with Mam-RRV-T48 or

Mos-RRV-T48 field isolates

Eighteen field isolates of RRV were obtained from WA and propagated once in either C6/36 (mosquito) cells or Vero (mammalian) cells. These viral preparations were then used to inoculate 18 day old C57Bl/TJ male mice to allow comparative analysis of the differences between isolates in disease progression of infected mice. This also enabled us to compare differences within each individual RRV isolate on the effect of Mam-RRV and Mos-RRV on disease progression in inoculated mice.

Mice inoculated with Mam-RRV-isolates experienced minimal changes in weight during the first few days post infection; however, by day six all inoculated mice weighed more than at day zero as shown in Figure 6.1A. The

177 only exception to this occurred in mice inoculated with Mam-RRV-B where a significant drop in the weight of mice occurred on day five, however this was quickly regained. Analysis of mice inoculated with Mam-RRV-isolates reveals two distinct growth curves as in Figure 6.1A. The first growth curve, containing a majority of the RRV isolate groups showed a slight reduction in weights over the first few days post inoculation and then a gradually increase in weight surpassing day zero weights by day six post infection.

The second growth curve consisted of mice inoculated with Mam-RRV-

M, -N and –R. Mice inoculated with these isolates did not experience the initial weight loss seen in the mice in the previous trend and gained weight steadily over the time course. This resulted in the weights of these mice being significantly higher by 30 days post infection, compared to the weights of mice inoculated with isolates in the other growth curve. The heaviest group at the end of the 30 day time course consisted of mice inoculated with Mam-RRV-N with a final mean weight of 327 % of initial body weight ± 8 %. Mice inoculated with Mos-RRV isolates lost weight in the five days post inoculation before weight stabilised or began to rise with several of the inoculated mouse groups taking longer than ten days to recover to weights recorded prior to inoculation

(Figure 6.1B). Mice inoculated with Mos-RRV-M had significantly lower weights than all other groups from day 14 to 27, and had the lowest mean weight at the end of the time course at an average of 190 % of initial body weight ± 11 %.

178

A )

% 3 5 0

(

h g

i 3 0 0

y 2 5 0

l 2 0 0

i g

i 1 5 0

n 1 0 0

c r

e 5 0 P 0 5 1 0 1 5 2 0 2 5 3 0 D a y s p o s t i n f e c t i o n B

% 3 5 0

h g

i 3 0 0 300

y 2 5 0

o b

200

l 2 0 0

g i

r 1 5 0

t 100 n

e 1 0 0

r e

P 5 0 0 0 5 1 0 1 5 2 0 2 5 3 0 0 5 10 15 20 25 D a y s p o s t i n f e c t i o n

RRV- Isolate legend A B C D E F G H I

J K L M N O P R Q

180

Figure 6.1 - Weight gain of C57Bl/6J mice inoculated with Mam-RRV and Mos-RRV field isolates.

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of A) Mam-RRV-isolate, B) Mos-RRV-isolate or 50 µL PBS (data not shown). Mice were monitored daily for weight gain which is shown as a percentage of the initial weight of the mice immediately prior to inoculation. Different isolates are represented by different symbols as depicted in the accompanying legend. Data is represented as the mean of each group (n=5) ± SEM.

181

Clinical scores of mice inoculated with Mam-RRV or Mos-

RRV field isolates.

Male mice were inoculated with either the mammalian or mosquito preparation of one of 18 isolates and observed daily for disease symptoms and rated as per the defined clinical disease score (Table 2.5) to determine differences between isolates and also between mammalian and mosquito derived preparations of the same isolate. Comparison of the clinical scores of mice inoculated with a Mam-RRV isolate did not reveal any discreet trends, with a lot of variation existing between groups (Figure 6.2A). The highest clinical score attained by mice inoculated with a Mam-RRV isolate was reached on day five by mice inoculated with Mam-RRV-H at 4 ± 0. For all mice inoculated with Mam-RRV isolates, clinical disease was resolved by 20 days post inoculation.

For mice inoculated with Mos-RRV-isolates, the highest average clinical score was 6 ± 0, reached on day five by mice inoculated with Mos-RRV-L and on day 14 by mice inoculated with Mos-RRV-G (Figure 6.2B). In contrast to all other isolates, mice inoculated with Mos-RRV-G reached peak clinical disease score at day 14 post infection, several days later than other isolates.

All mice inoculated with a Mos-RRV isolate produced a disease score ≥1 between days three and ten. Disease symptoms resulting in a clinical score were resolved in all mice inoculated with Mos-RRV isolates by day 25.

182

A 8

a 4

l C 2

0 0 6 1 2 1 8 2 4 3 0 D a y s p o s t i n f e c t i o n B 8

6 300

a 4 200

l C 2 100

0 0 0 6 1 2 1 8 2 4 3 0 0 D 5a y s p o s10t i n f e c t15i o n 20 25

RRV- Isolate legend A B C D E F G H I

J K L M N O P R Q

184

Figure 6.2 - Clinical score of C57Bl/6J mice inoculated with Mam-RRV and Mos-RRV isolates.

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of A) Mam-RRV-isolate, B) Mos-RRV-isolate or with 50 µL PBS (data not shown). Mice were monitored daily for disease progression which was scored according to the established clinical disease score as described in Table 2.5, briefly, 0 – no disease, 1 – ruffled fur, 2 – Very mild hind limb weakness, 3 – mild hind limb weakness, 4 – Moderate hind limb weakness, 5 – Severe hind limb weakness, 6 – Loss of function. Different isolates are represented by different symbols as depicted in the accompanying legend. Data is represented as the mean of each group (n=5) ± SEM.

185

Disease progression in mice inoculated with Mam-RRV or

Mos-RRV field isolates

Eighteen isolates were used to examine the weight gain and clinical disease profile of male C57BL/6J mice inoculated with Mam-RRV and Mos-RRV over a 30 day period as shown previously in Figure 6.1 and Figure 6.2. This information was then collated for each isolate and mice inoculated with Mam-

RRV were compared to their Mos-RRV counterpart. The disease profiles in mice could be grouped into four trends.

In the first trend depicted in Figure 6.3A, there was less than three points of significant difference in weight gain and clinical score when comparing

Mam-RRV and Mos-RRV inoculated mice. This group contained mice inoculated with isolates RRV-D, RRV-F, RRV-H and RRV-L.

The second trend was represent by mice inoculated with four RRV isolates

(RRV-B, -I, -K, -Q) and is shown in Figure 6.3B. Mice inoculated with the

Mam-RRV and Mos-RRV had similar weight gains but significantly different clinical scores at later time points. Mam-RRV isolate inoculated mice returned to a clinical score of zero, on average five days earlier than mice inoculated with Mos-RRV. A return to zero clinical score occurred in all groups by day 13, however for mice inoculated with Mos-RRV-I there was an additional increase in the clinical scores on day 13, 14 and 16, raising them from zero but remaining under two.

The third trend is depicted in Figure 6.3C. In this group, there was a significant difference in weight gain and clinical scores at later time points between Mam-RRV and Mos-RRV inoculated mice. Mice inoculated with

186

Mam-RRV isolates had an early peak disease prior to 10 days post infection, but clinical scores remain under 4. In addition, disease symptoms resolved on average four days earlier than mice inoculated with the Mos-RRV. On the other hand, mice inoculated with Mos-RRV isolates had late peak in disease symptoms with the peak occurring after 10 days post infection. This group contained mice inoculated with the isolates RRV-C, -E, -G, -J, -O and -P.

The final trend shown in Figure 6.3D is characterised by a significant difference between the weight gain and clinical scores of mice inoculated with

Mam-RRV and Mos-RRV. In this trend, mice inoculated with Mos-RRV isolates gained significantly less weight compared to mice inoculated with

Mam-RRV isolates. In the mice inoculated with RRV-A, a significant difference in weights was observed for mice inoculated with Mam-RRV-A compared to Mos-RRV-A only after 12 days post infection. In contrast, mice inoculated with the other isolates in this trend (RRV-M, -N, -R), exhibited a significant change in weights from day one post infection. Clinical scores were significantly different between Mam-RRV and Mos-RRV inoculated mice by day three post infection for all isolate groups tested.

187

) 350 8

(

h 300

e * * * 6

w e

250 r

200 a 4

i *

r 150

t 2

e 100

e P 50 0 0 10 20 30 0 10 20 30 Day Days post infection

) 350 8

(

h 300

e 6

w e

250 r

200 a 4

i * * n *

r 150

o * * t 2

e 100 *

e P 50 0 0 10 20 30 0 10 20 30 Days post infection Days post infection

) 350 8

(

h 300

e 6

w e

250 r

y * * o

d * * * * * c

* l * *

l * 200 * a 4

* c

* i * n * *

i * *

r 150

o * * *

t 2

e 100 * * *

e P 50 0 0 10 20 30 0 10 20 30 Days post infection Days post infection

188

) 350 * * 8

% * ( * * t * * h 300 * g *

i * * * * * * e * 6 *

w * e

250 * r * *

y *

* c * *

o * S

* b * * l

l * * 200 * * a 4

n *

i * n

l * i *

r 150 * * C

o * * * * * * * t * 2

e 100

e P 50 0 0 10 20 30 0 10 20 30 Days post infection Days post infection

M a m H M a m H M o s H

) Figure 6.3 - Disease profile trends for mice inoculated with Mam-RRV and Mos-RRV.

) M o s H P B S

)

) (

% 3 0 0 C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region )

P B S M a m m a lia n d e riv e d T 4 8 -R R V (

t M a m H

% %

3 0 0 ( % M a m H

( with 10 pfu/mouse Mam-RRV-T48 ( ), M Mosa m -mRRVa lia n- T48d e riv e( d )T Mam4M8o-Rs qR-uVRRVito d e-risolateiv e d T 4 8(-R),R V Mos-RRV-

h t

M a m H M o s H

( t

M o s H

t g h * M o s q u it*o d*e riv e d T 4 8 -R R V t M o s H

i isolate ( ) or 50 µL PBS ( ). P MiceB S were monitored daily for weight gain, which is shown as a

h h g P B S

h * e 3 0 0 i

g P B S M a m* m*a lia n d e riv e d T 4 8 -R R V g

i percentage3 0 0 of the initial weight of the mice immediately prior to inoculation M anda m m diseasea lia n d e r iv e d T 4 8 -R R V

g i

3 0 0 e i

w M a m m a lia n d e riv e d TM4o8s-qRuRitoV d e riv e d T 4 8 -R R V

e e

M o s q u ito d e riv e d T 4 8 -R R V e w progression,M o s qasu it oscored* d e*riv e d according T 4 8 -R R V to the established clinical disease score. Data is shown as

w * * y * w *

* w

* *

d y

representative trends and are illustrated by the results gained from mice inoculated with a

d y

2 0 0 d

b o

d representative isolate from each trend, A) RRV-H, B) RRV-Q, C) RRV-O and D) RRV-N. Data

o 2 0 0

is represented as the mean of each group (n=5) ± SEM.

2 0 0

l 2 0 0

a a

2 0 0 i

g g

t t

t 1 0 0

n n

1 0 0 e 1 0 0

c e

e 1 0 0

c e

1 0 0 r

P P P 0 1 0 1200 2 03 0 3 0 0 0 1 0 0 1 0 2 0 1 0 2 0 3 0 2 0 D a3y0 3 0 D a y D a y D a y D a y

189

Viral titres in mice inoculated with Mam-RRV or Mos-RRV field isolates

Male mice were inoculated with either Mam-RRV-isolate or Mos-RRV- isolate or 50 µL of PBS. Samples of blood, quadriceps muscle, ankle joint and brain were taken 24 hours post infection. These were homogenised in PBS and titrated on Vero cells. To enable a comparison between the titres recovered from RRV isolates and those recovered from RRV-T48 the data previously presented in Figure 5.7 is included again in Figure 6.4. No viral titres were recovered from any sample in mice inoculated with PBS (data not shown).

No difference was found at 24 hours post infection between the blood titres of Mam-RRV and Mos-RRV inoculated mice of the same RRV isolate

(Figure 6.4A). In contrast, there was a significant difference between titres recovered from mice inoculated with Mam-RRV-T48, Mam-RRV-M and Mam-

RRV-R. Additionally, Mos-RRV-T48, Mos-RRV-M and Mos-RRV-R inoculated mice also had significantly different viral blood titres compared to one another.

In the quadriceps muscle, (Figure 6.4B) there was a significant difference in titres obtained from mice inoculated with the Mam-RRV isolate compared to the Mos-RRV for each RRV isolate at 24 hours post infection.

In the ankle joint, (Figure 6.4C) titres were again significantly different between mice inoculated with Mam-RRV compared to Mos-RRV for both, RRV-

T48 and RRV-R; however, there was no difference between Mam-RRV and

Mos-RRV for the isolate RRV-M. In addition, titres in the ankle joint of mice

190 inoculated with Mam-RRV-R were significantly lower than mice inoculated with Mam-RRV-T48. Viral titres recovered from the brain tissue at 24 hours post infection were found to be similar for all isolates and were between 3.01

± 0.326 and 4.04 ± and 0.317 log pfu/g as seen in Figure 6.4D.

Peak disease blood titres were significantly lower (Figure 6.4E) in comparison to titres obtained at 24 hours post infection (Figure 6.4A). There was no significant difference between mice inoculated with different RRV isolates and all titres were between 1.78 ± 0.205 and 2.32 ± 0.0209 log pfu/ml.

In contrast, most titres recovered at peak disease from quadriceps muscle

(Figure 6.4F) and ankle joint (Figure 6.4G) increased compared to titres recovered at 24 hours post infection (Figure 6.4B and C). The only exception to this was for titres recovered from mice inoculated with Mam-RRV-T48 which decreased slightly between 24 hours post infection and peak disease in both quadriceps muscle and ankle joint. Additionally, mice inoculated with

Mam-RRV-T48 had significantly lower viral titres in quadriceps muscle at peak disease when compared to the titres recovered from Mos-RRV-T48 and

Mam-RRV-M inoculated mice.

In the ankle joint at peak disease, a significant difference was found between the titres of Mam-RRV-T48 and Mos-RRV-T48 inoculated mice and also between Mam-RRV-R and Mos-RRV-R inoculated mice (Figure 6.4G).

Additionally, mice inoculated with Mam-RRV-T48 had significantly lower viral titres in ankle joint tissue compared to mice inoculated with Mam-RRV-R. No viral titres were detected for any of the mice inoculated with RRV in brain tissue at peak disease (Figure 6.4H).

191

Blood 24 hours Blood peak disease A * E 8 * 8

) 7 ) 7

L L

m 6 m 6

/ /

u u

f 5 f 5

p p

g 4 g 4

o o

l l

( 3 ( 3

e e

r 2 r 2

t t

i i

T 1 T 1 0 0 Mam Mos Mam Mos Mam Mos Mam Mos Mam Mos Mam Mos T 48 T 48 M M R R T 48 T 48 M M R R

B Quadriceps 24 hours F Quadriceps peak disease 8 8 * 7 7 *

) )

g * g

/ 6 / 6

u u f 5 * f 5 p * p

g 4 g 4

o o

l l

( ( 3 3

e e

r r

t t

i 2 i 2 T 1 T 1 0 0 Mam Mos Mam Mos Mam Mos Mam Mos Mam Mos Mam Mos T 48 T 48 M M R R T 48 T 48 M M R R

Ankle 24 hours Ankle peak disease C G * 8 * 8 * * 7 * 7

) )

g g

/ 6 / 6

u u f 5 f 5 p * p

g 4 g 4

o o

l l

( ( 3 3

e e

r r

t t

i 2 i 2 T 1 T 1 0 0 Mam Mos Mam Mos Mam Mos Mam Mos Mam Mos Mam Mos T 48 T 48 M M R R T 48 T 48 M M R R

Brain 24 hours Brain peak disease D H 8 8 7 7

) )

g g

/ 6 / 6

u u

f f 5 5

g g 4 4

(

3 3

i i 2 2

T T # 1 1 0 0 Mam Mos Mam Mos Mam Mos Mam Mos Mam Mos Mam Mos T 48 T 48 M M R R T 48 T 48 M M R R

192

B lood 2 4 hour s * B lood 2 4 hour8s * M am * 7 T 48 8 * 6 M am M o s B lood 2 4 7hour s 5 T 48 T 48 B lood 2 4 hour s

* ) 6 l

8 m M o s M am * * / 4

u M am B lood 2 4 ho8ur s 5 f T 48 *

7 p M (

) M am l T 438 *

e 7 m

Figurer 6.4 - Viral titres recovered from blood, quadriceps muscle, ankle joint and brain

/ M am 6 4 t M o s T 48 i 8 u M o s * f 2 of miceT inoculatedB lood with2 4 ho uMamr s -RRV and Mos6 -RRV field isolates. p M M M am

( M o s

5 T 48

) 3 7 l

e 1 * T 48 r

m 5 T 48 )

t M o s M am

l / 4 i C57Bl/6JM8 am wild type 18 day old male mice were injected subcutaneously in the pectoral region

u 2 *

T m

f 6

/ M o s p 0 M MRa4m M am

M u

( 4 f

3 with 107 pfu/mouse of Mam-RRV-T48 ( ), Mos-RRV-T48 ( ), Mam-RRV-M ( ), Mos-RRV- p e 1 M a m M o s M a m M o s M a m M o s T 48 M

5 ( T 48

M o s ) M am M3 o s i

T 4 8 T 4 8 M l M R R e

2 6 r

T M ( ), Mam-RRV-R ( ) or Mos-RRV-R ( ). Samples were taken at 24 hours post infection t

m M o s M o s i 0 M / 4R R M am

u 2 T

1 M a m M o s M a m M o s 5 M a m M o s f T 48 M ) and at peak disease frop m A, E) blood, B, F) quadriceps muscle,M C, G) ankle joint and D, H)

l M am ( M o s T 4 8 T 4 8 M M R R 3

m 1 e / M am M am

0 R4 r u

brain. Samples were homogenised,t R clarified and titrated on VeroM o s cells. Data is represented f M a m M o s M a m M o s M a m M o s i p 2

T M 0 R ( T 4 8 T 4 8 M M R R M3 o s M

as thee mean of each group (n=4) ± SEM. # denotesM a m thatM o s valueM a m Mwaso s Mundera m M o thes limit of detection r

t M o s M o s i R 1 2 T 4 8 T 4 8 M M R R T M am (0.6 log pfu/g). M R 1 0 R M a m M o s M a m M Mo sam M a m M o s 0 T 4 8 T 4 8 M MR R R M o s M a m M o s M a m M o s M a m M o s R T 4 8 T 4 8 M M R R M o s R

193

Quadriceps muscle histology of mice inoculated with

Mam-RRV or Mos-RRV field isolates at 24 hours post infection

To examine quadriceps pathology at 24 hours post inoculation, male mice were inoculated with Mam-RRV-M, Mos-RRV-M, Mam-RRV-R, Mos-RRV-R or

Mam-RRV inoculated mice and mild/moderate signs of inflammation evident in the sections taken from mice inoculated with Mos-RRV.

The quadriceps transverse and longitudinal sections from mice inoculated with Mam-RRV-M show evidence of mild inflammation. However, Mos-RRV-

M inoculation of mice resulted in moderate inflammation present in the quadriceps muscle as evident by a widening of the perimysium. Mos-RRV-M inoculated mice also have necrotic fibres visible in the transverse sections of the quadriceps muscle.

Apart from a widening in the extracellular space, transverse quadriceps sections for mice inoculated with Mam-RRV-R are comparable to PBS controls in terms of negligible structural and cellular changes. Only, a slight increase in adipose tissue can be seen in longitudinal quadriceps muscle sections. In contrast, the evidence of inflammation is the most prominent in quadriceps of mice inoculated with Mos-RRV-R. The transverse and longitudinal sections show an increase in inflammatory infiltrates, tissue necrosis as well as fibrosis and increase in adipose tissue.

194

196

Figure 6.5 - Histology of quadriceps muscle in mice infected with RRV at 24 hours post infection

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of Mam-RRV-M, Mam-RRV-R, Mos-RRV-M, Mos-RRV-R or 50 µL of PBS. Samples were taken from quadriceps muscle and fixed in formalin. Tissues were paraffin embedded, sections were cut and stained with haematoxylin and eosin. Images are representative of each group (n=4).

197

Ankle joint histology of mice inoculated with Mam-RRV or

Mos-RRV field isolates at 24 hours post infection

To examine ankle pathology at 24 hours post inoculation, male mice were inoculated with Mam-RRV-M, Mos-RRV-M, Mam-RRV-R, Mos-RRV-R or

PBS. At 24 hours post inoculation mice were sacrificed and tissues harvested fixed in formalin. They were then imbedded in paraffin and stained with H&E before examination by microscopy of four chosen sections to allow evaluation over the whole ankle area. All ankle sections of RRV inoculated mice show evidence of inflammation however, the mice inoculated with RRV-R show more severe tissue destruction than that of mice inoculated with RRV-M.

Mam-RRV-M inoculated mice show signs of inflammation in the ankle joint as evident by a thickening of the synovial membrane. Inflammation is also evident in the Achilles tendon, with an increase of adipose tissue and some evidence of haemorrhage. Similar findings can be seen in the ankle sections of mice inoculated with Mos-RRV-M with the addition of inflammatory percussions, contributing to severe disruption of the synovial membrane, increased infiltrates fibrosis and pannus formations.

In the ankle section of mice inoculated with Mam-RRV-R and Mos-RRV-

R, a moderate to severe amount of tissue damage can be seen which is likely due to the inflammatory process. This is evident by disruption to the synovial membrane, and increase in inflammatory infiltrates, fibrosis and pannus formation as well as the increase in adipose tissue in the ligaments and tendons. The resulting image shows that there is major disruption to the normal architecture of the joint.

198

199

200

Figure 6.6 - Histology of ankle joint of mice infected with RRV at 24 hours post infection

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of Mam-RRV-M, Mos-RRV-M, Mam-RRV-R, Mos-RRV-R or 50 µL PBS. Samples were taken from the ankle joint at 24 hours post infection and fixed in formalin. Tissue was paraffin embedded, sectioned and stained with haematoxylin and eosin. B) bone, C) cartridge, G) growth plate, M) muscle, P) periosteum, S) synovia/synovial space. Data is representative of each group (n=4) ± SEM.

201

Brain histology of mice inoculated with Mam-RRV and

Mos-RRV field isolates at 24 hours post infection

To examine the pathology occurring in brain tissue at 24 hours post inoculation, male mice were inoculated with Mam-RRV-M, Mos-RRV-M, Mam-

RRV-R, Mos-RRV-R or PBS. At 24 hours post inoculation mice were sacrificed and tissues harvested fixed in formalin. They were then imbedded in paraffin and stained with H&E before examination by microscopy.

In mice inoculated with Mam-RRV-M, sections of brain tissue reveal that some inflammation has occurred. There is evidence of necrotic neurons and vacuolisation of the cytoplasm, collagen fibres and other protein rich structures. The brain sections from mice inoculated with Mos-RRV-M contain clusters of inflammatory infiltrates as well as dark neurons and necrotic neurons with shrinkage of the nucleus and vacuolisation of the cytoplasm.

Brain sections for mice inoculated with Mam-RRV-R and Mos-RRV-R show similar signs of mild inflammation with the presence of dark neurons and necrotic neurons along with inflammatory infiltrates and some vacuolisation.

202

Figure 6.7 - Histology of brain tissue in mice infected with RRV at 24 hours post infection

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of mammalian derived RRV isolates M and R, mosquito derived RRV isolate M and R or 50 µL PBS. Samples were taken from the brain and fixed in formalin. Tissues were paraffin embedded, sections were cut and stained with Haematoxylin and eosin. Data is represented as the mean of each group (n=4) ± SEM.

203

Quadriceps muscle histology of mice inoculated with

Mam-RRV and Mos-RRV field isolates at peak disease

To examine tissue pathology at the peak of disease symptoms, male mice were inoculated with Mam-RRV-M, Mos-RRV-M, Mam-RRV-R, Mos-

RRV-R or PBS. At when at peak symptomatic disease (7 days post infection), as guided from previous data (6.2.1. and 6.2.2. ), mice were sacrificed and tissues harvested fixed in formalin. They were then imbedded in paraffin and stained with H&E before examination by microscopy.

Mice inoculated with Mam-RRV-M show signs of mild/moderate inflammation of the quadriceps muscle (Figure 6.8). The transverse and longitudinal sections show muscle being replaced by adipose tissue, myofiber necrosis with myophagocytosis and presence of eosinophil and lymphocyte infiltrates. Inflammation is clearly visible in the longitudinal section with areas of endomesysial lymphocytic inflammatory infiltrate. The muscle sections from the Mos-RRV-M group show widening of the perimysium and the endomysium cell fibres are swollen and irregular indicating myositis. Cell infiltrates are present and neutrophils, eosinophils and lymphocytes can be readily identified with areas of focal invasion of muscle fibres by inflammatory cells. There is also evidence of late stage necrosis.

204

Quadriceps muscle sections for mice inoculated with Mam-RRV R seen in Figure 6.8 show similar evidence of cellular infiltrates, the perimesium is widened and fibre splitting is present as is fibre necrosis. Mos-RRV-R inoculated mice show dramatic changes due to myositis with fibres becoming small and rounded, an increase in extracellular space, areas of fibrosis as well as degenerating fibres, areas of necrosis and cytoplasmic vacuoles are all present in the sections. Inflammatory infiltrates such as neutrophils and eosinophils can be identified. Additionally, muscle fibres are being replaced and infiltrated by adipose tissue.

For both isolate groups myositis and myopathy is most severe in mice inoculated with Mos-RRV compared to the mice inoculated with Mam-RRV.

Additionally, the myopathies were more prominent in mice inoculated with

RRV-R isolate compared to mice inoculated with RRV-M isolate.

205

206

Figure 6.8 - Histology of quadriceps muscle in mice infected with RRV at peak disease

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of mammalian derived RRV isolate M and R, mosquito derived RRV isolate M and R or 50 µL PBS. Samples were taken from, quadriceps muscle, ankle joint and brain fixed in formalin. Tissues were paraffin embedded, sections were cut and stained with Haematoxylin and eosin. Data is represented as the mean of each group (n=4) ± SEM.

207

Ankle joint histology of mice inoculated with Mam-RRV or

Mos-RRV field isolates at peak disease

All mice inoculated with RRV showed evidence of inflammation and associated tissue damage in ankle sections at peak disease when compared with the PBS inoculated control mice. In all cases, infection resulted in damage to the tissue and disruption of the normal architecture of the joint however, this was more severe in Mos-RRV inoculated mice compared to

Mam-RRV inoculated mice. Additionally, the most tissue damage was seen in mice inoculated with Mos-RRV-R (Figure 6.9).

In mice inoculated with Mam-RRV-M, some synovial hyperplasia and fibrosis is evident around the navicular ligament with inflammatory infiltrates. Additionally, there is evidence of muscle fibre necrosis in the

Achilles tendon with fibrosis/inflammatory pannus. In contrast, the histology of mice inoculated with Mos-RRV-M shows severe damage to the area around the navicular bone, causing loss of normal architecture. Cellular infiltrates, cartridge degeneration and bone erosion are present as well as areas of synovial hyperplasia and pannus formation.

Mice inoculated with Mam-RRV-R showed inflammatory infiltrates, mild cartilage destruction and bone erosion with mild synovial hyperplasia.

The ligaments show evidence of fibrosis and inflammatory infiltrate as well as some muscle necrosis. In Mos-RRV-R inoculated mice, inflammation and associated tissue damage is severe with the presence of inflammatory infiltrates, destruction of cartridge and bone erosion. Fibrosis, inflammatory pannus and synovial hyperplasia of the foot has disrupted normal

208 architecture. The navicular ligament has reduced in size and adipose infiltration is evident as is muscle degeneration and fibrosis in the Achilles tendon.

209

210

211

Figure 6.9 - Histology of ankle joint in mice infected with RRV at peak disease

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of Mam-RRV-M, Mos-RRV-M, Mam-RRV-R, Mos-RRV-R or 50 µL PBS. Samples were taken from the ankle joint at peak disease and fixed in formalin. Tissue was paraffin embedded, sectioned and stained with haematoxylin and eosin. B) bone, C) cartridge, G) growth plate, M) muscle, P) periosteum, S) synovia/synovial space. Data is representative of each group (n=4) ± SEM.

212

Brain histology of mice inoculated with Mam-RRV or Mos-

RRV at peak disease

Examination of the brain tissue at peak disease showed evidence of encephalitic involvement. This was most prominent in mice inoculated with

Mos-RRV, however minor inflammatory signs were also present in mice inoculated with Mam-RRV. The most severe encephalitic presentation was seen in the Mos-RRV-R group as seen in Figure 6.10.

For mice inoculated with Mam-RRV-M there is some evidence of vacuolation and thrombus with minimal infiltrates. Necrotic neuron damage can also be identified, as well as embolus. In mice inoculated with Mos-RRV-

M, inflammatory cells are present, in addition, dark neurons are present indicating irreversible neuron injury. Vacuolation and an increase in protein rich fibres are also present in the white matter.

In mice inoculated with Mam-RRV-R, some inflammatory infiltrates are present along with low levels of necrotic neurons. Evidence of a minor embolus can be seen but was not an occurrence in all samples. In contrast, multiple embolus are present in mice inoculated with Mos-RRV-R.

Inflammatory cells, swollen astrocytes, necrotic neurons and inclusions are present in the cellular make up. Signs of early neuron injury are present and inflammatory tissue damage is extensive in areas of the white matter.

213

Figure 6.10 - Histology of brain tissue in mice infected with RRV at peak disease

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region with 104 pfu/mouse of mammalian derived RRV isolate M and R, mosquito derived RRV isolate M and R or 50 µL PBS. Samples were taken from the brain and fixed in formalin. Tissues were paraffin embedded, sections were cut and stained with Haematoxylin and eosin. Data is represented as the mean of each group (n=4) ± SEM.

214

Cytokine profiles of mice inoculated with Mam-RRV and

Mos-RRV field isolates

Gene expression of inflammatory cytokines were examined by real-time

PCR for quadriceps muscle and ankle joints for male mice inoculated with

Mam-RRV-M, Mos-RRV-M, Mam-RRV-R and Mos-RRV-R. Samples were taken at 6, 12 and 24 hours as well as at peak symptomatic disease as guided by previous data in this chapter. All results are normalised against the HPRT1 house keeping geneand expressed as fold change compared to levels in PBS control mice.

Cytokine profiles in quadriceps muscle of RRV inoculated mice at six hours post infection

At six hours post infection the gene expression of mice for most cytokines were not significantly different to either the other RRV inoculated mouse groups or PBS inoculated mice. Two notable exceptions were gene expression of CCL2 and IL4 (Figure 6.11A). CCL2 expression in mice inoculated with Mos-RRV-R was significantly higher than that for both Mam-

RRV-R and PBS inoculated mice. Expression of CCL2 in mice inoculated with

Mos-RRV-R was found to be 24 fold higher than PBS inoculated mice and 2.5 fold higher than Mam-RRV-R inoculated mice. Additionally, IL4 expression in Mos-RRV-M inoculated mice was found to be 4 fold higher than PBS inoculated mice.

There was no significant difference in gene expression of CXCR3, IFNα,

IFNβ, IFNγ IL6, IL10, IL12, MIF and TNFα between RRV and PBS inoculated mice at six hours post infection in the quadriceps muscle.

215

Cytokine profiles in quadriceps muscle of RRV inoculated mice at 12 hours post infection

At 12 hours post infection the expression of CCL2 gene expression in the quadriceps muscle was significantly higher in mice inoculated with Mam-

RRV-R showing a 15 fold increase compared to PBS inoculated mice (Figure

6.11B). CXCR3 gene expression was elevated for both Mam-RRV-R and Mos-

RRV-R inoculated mice when compared to PBS mice at an 8 and 10 fold increase respectively. Gene expression of IFNα was significantly elevated between 15-26 fold for all mice inoculated with RRV, with the exception of

Mam-RRV-M, when compared to PBS inoculated mice. Additionally, gene expression of IFNα in mice inoculated with Mam-RRV-M was found to be significantly different to gene expression of mice inoculated with Mos-RRV-M.

This pattern was similar for IFNβ gene expression, however the expression levels of Mos-RRV-M, Mam-RRV-R and Mos-RRV-R were between 67-84 fold higher than the gene expression of IFNβ in PBS inoculated mice. The gene expression in mice inoculated with Mam-RRV-M was down regulated by over

2.5 fold compared to PBS however this was not statistically significant, but the difference between Mam-RRV-M and Mos-RRV-M was found to be statistically significant with a 6.5 fold change. The gene expressions levels of

IFNγ in all RRV inoculated mice was not significantly different compared to expression in PBS controls. However, mice inoculated with Mam-RRV-M showed downregulation of IFNγ expression which was significantly different by 6.5 fold when compared to expression of IFNγ by Mos-RRV-M inoculated mice.

216

Mice inoculated with Mos-RRV-M, Mam-RRV-R or Mos-RRV-R all had significantly elevated gene expression of IL4 at 12 hours post infection when compared to PBS inoculated mice, with 123-247 fold difference. Additionally

Mam-RRV-M was found to be significantly different to Mos-RRV-M with a 24.5 fold difference between the two. A significant difference between mice inoculated with Mam-RRV-M or Mos-RRV-M was also found in regards to IL10 gene expression in Mos-RRV-M inoculated mice 3 fold higher than Mam-RRV-

M inoculated mice, however none of the RRV inoculated mouse groups were found to be significantly different to PBS inoculated mice for IL10 gene expression. Mice inoculated with Mos-RRV-M and Mam-RRV-R were found to have elevated expression of IL2 at 3.5 fold change compared with PBS inoculated mice. In addition, expression of IL12 in Mos-RRV-M inoculated mice was 6.5 fold higher than expression in Mam-RRV-M inoculated mice.

There was no significant difference in gene expression of IL6, MIF and

TNFα between RRV and PBS inoculated mice at 12 hours post infection in the quadriceps muscle.

217

A 40

* *

i 30

l 20

e 10

R *

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL 10 IL 12 MIF T NF 

B 300 * * *

r 200

e *

m * *

r *

n *

v 100

* * * * * * * * * * * * 0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL 10 IL 12 MIF T NF 

218

C 15

e * * * r 10

v 5 i * *

* * * * * * * * * 0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL 10 IL 12 MIF T NF 

D 40

* * *

i 30

e * *

s *

l 20

n * * *

e * v *

e 10

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL 12 MIF T NF 

219

Figure 6.11 - Cytokine profiles in the quadriceps muscle of mice inoculated with mammalian and mosquito derived RRV field isolates.

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region

4 with 10 pfu/mouse of Mam-RRV-M ( ), Mos-RRV-M ( ), Mam-RRV-R ( ), Mos-RRV- R ( ) or with 50 µL PBS ( ). Mice were sacrificed at A) 6 hours, B) 12 hours, C) 24 hours post infection and at D) peak disease and the quadriceps muscle was removed. Total RNA from the quadriceps muscle was isolated and analysed for mRNA expression by real time PCR. Data was normalised to the house keeping gene HPRT1 and expressed as relative expression to PBS inoculated mice. Results are represented as the mean of each group (n=5) ± SEM. Statistical significance (* p<0.05) was determined by one way ANOVA with Tukey’s test.

220

Cytokine profiles in quadriceps muscle of RRV inoculated mice at 24 hours post infection.

CCL2 gene expression was significantly elevated 6-8 fold for mice inoculated with Mos-RRV-M, Mam-RRV-R or Mos-RRV-R when compared to mice inoculated with PBS (Figure 6.11C). Mice inoculated with Mos-RRV-M and Mam-RRV-R had significantly higher levels of CXCR3 gene expression at this time point compared to control mice with a fold change of between 2.5 and 3.5. A significant down regulation in the gene expression of IFNβ was found in mice inoculated with Mos-RRV-M compared to PBS inoculated mice, no significant differences were found in the other groups. However, for IFNγ gene expression, mice treated with Mos-RRV-M were the only group not significantly down regulated for gene expression compared to PBS with the other groups having a negative fold change of between 6 and 11. Mice inoculated with Mam-RRV-M were found to have higher expression of IL6 by over 3.5 fold compared to control mice and also had a significant fold change difference of 3.5 when compared to Mos-RRV-M inoculated mice. The gene expression for MIF was found to be significantly down regulated in Mos-RRV-

M and Mam-RRV-R by 18.5 and 3 fold respectively. Additionally, a significant

12.5 fold difference was found between Mam-RRV-M and Mos-RRV-M. There was no significant difference in gene expression of IFNα, IL4, IL10, IL12 and

TNFα between RRV and PBS inoculated mice at 24 hours post infection in the quadriceps muscle.

221

Cytokine profiles in quadriceps muscle of RRV inoculated mice at peak disease post infection.

IFNβ gene expression was significantly elevated between 11 and 12 fold for mice inoculated with Mam-RRV-M or Mam-RRV-R (Figure 6.11D).

Additionally mice inoculated with Mam-RRV had significantly different IFNβ expression to mice inoculated with Mos-RRV with fold changes of 11 for RRV-

M and 5 for RRV-R inoculated mice. IFNγ gene expression was significantly up-regulated in Mam-RRV inoculated mice with fold changes of 27 and 14 for

Mam-RRV-M and Mam-RRV-R inoculated mice respectively. Additionally, mice inoculated with Mam-RRV-M had a significant fold change of over 4.5 compared to Mos-RRV-M inoculated mice. Both Mam-RRV-M and Mam-RRV-

R inoculated mice were significantly elevated for IL12 gene expression with fold changes of 13 and 10 respectively compared to control mice. The gene expression of TNFα was significantly elevated for mice inoculated with both

Mos-RRV-M and Mam-RRV-R with 15 and 18 fold change compared to PBS inoculated mice and a significant fold change of 11 found between Mam-RRV-

M and Mos-RRV-M inoculated mice. There was no significant difference in gene expression of CCL2, CXCR3, IFNα, IL4, IL6 and MIF between RRV and

PBS inoculated mice at peak disease post infection in the quadriceps muscle.

Cytokine profiles in ankle joint of RRV inoculated mice at 6 hours post infection.

The only cytokine to have gene expression level differences at 6 hours post infection in RRV inoculated mice was CCL2 as seen in (Figure 6.12A).

Both Mos-RRV-M and Mos-RRV-R inoculated mice were found to have

222 elevated levels of gene expression for CCL2 (6.5 and 13 fold change respectively) at six hours post infection in the ankle joint. Additionally, expression in Mos-RRV-M inoculated mice was found to be 2 fold higher than

Mam-RRV-M inoculated mice and Mos-RRV-R inoculated mice had 5.5 fold higher expression of CCL2 than Mam-RRV-R inoculated mice.

Cytokine profiles in ankle joints of RRV inoculated mice at

12 hours post infection.

Gene expression of IFNα was significantly up-regulated by 11 fold in

Mos-RRV-M inoculated mice compared to control mice (Figure 6.12B). Mos-

RRV-M inoculated mice also had significantly higher IFNα gene expression profiles than Mam-RRV-M with a fold difference of 7. IFNβ gene expression was significantly elevated in Mos-RRV-M inoculated mice with a fold difference of 5 compared to PBS inoculated mice and 7.5 fold change compared to Mam-RRV-M. Mice inoculated with Mos-RRV-R had elevated levels of IL10 compared to both Mam-RRV-R and PBS inoculated mice with fold changes of 2.5 and 5.5 respectively. Mos-RRV-R inoculated mice also had 3 fold elevated gene expression of IL12 compared to PBS inoculated mice.

There was no significant differences in gene expression of CCL2, CXCR3, IFNγ,

IL4, IL6, MIF and TNFα between RRV and PBS inoculated mice at 12 hours post infection in the ankle joint.

223

A 20 * * *

i 15

s *

i l 10 *

e 5

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL 10 IL 12 MIF T NF 

B 15

* *

r 10

i l * *

o * *

v 5

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL 10 IL 12 MIF T NF 

224

C 20 * * * *

i 15

l 10

a l * * * * e 5

* * * *

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL 10 IL 12 MIF T NF 

D 100 *

m r * * o 40

* *

e * R 20 * * *

0 CCL 2 CXCR3 IFN IFN IFN IL 4 IL 6 IL 12 MIF T NF 

225

Figure 6.12 - Cytokine profiles in the ankle joint of mice inoculated with mammalian and mosquito derived RRV field isolates.

C57Bl/6J wild type 18 day old male mice were injected subcutaneously in the pectoral region

4 with 10 pfu/mouse of Mam-RRV-M ( ), Mos-RRV-M ( ), Mam-RRV-R ( ), Mos-RRV- R ( ) or with 50 µL PBS ( ). Mice were sacrificed at A) 6 hours, B) 12 hours, C) 24 hours post infection and at D) peak disease and the ankle joint was removed. Total RNA from the ankle joint was isolated and analysed for mRNA expression by real time PCR. Data was normalised to the house keeping gene HPRT1 and expressed as relative expression to PBS inoculated mice. Results are represented as the mean of each group (n=5) ± SEM. Statistical significance (* p<0.05) was determined by one way ANOVA with Tukey’s test.

226

Cytokine profiles in ankle joints of RRV inoculated mice at

24 hours post infection.

CCL2 gene expression levels in the ankle joints of RRV inoculated mice was significantly down regulated for Mos-RRV-M and Mam-RRV-M inoculated mice by 4.7 fold compared to PBS inoculated mice (Figure 6.12C). Mice inoculated with Mos-RRV-M had a significant 3 fold decrease in CCL2 gene expression compared to Mam-RRV-M inoculated mice and Mam-RRV-R inoculated mice had a significant 5 fold decrease in gene expression compared

Mos-RRV-R inoculated mice. Although no RRV inoculated mice were found to be significantly different to PBS inoculated mice for CXCR3 gene expression levels, Mam-RRV-M inoculated mice were found to have a significant 7 fold increase in gene expression compared to Mos-RRV-M inoculated mice. Mam-

RRV-M inoculated mice had significantly higher (4 fold) IFNα and IFNβ levels compared to PBS inoculated mice. Additionally, mice inoculated with Mam-

RRV-M were found to have significantly higher gene expression levels to mice inoculated with Mos-RRV-M for IFNα and IFNβ gene expression with 2.5 and

5 fold difference respectively. All RRV inoculated mice had elevated levels of

IL10 at 24 hours post infection with fold change differences of between 9.5-

15.5 when compared to PBS inoculated mice. There was no significant differences in gene expression of IFNγ, IL4, IL6, IL12, MIF and TNFα between

RRV and PBS inoculated mice at 24 hours post infection in the ankle joint.

227

Cytokine profiles in ankle joints of RRV inoculated mice at peak disease post infection.

Mice inoculated with Mam-RRV-M had significantly higher gene expression levels of CXCR3 by 2 fold in ankle joints when compared to both

Mos-RRV-M and PBS inoculated mice (Figure 6.12D). Both Mam-RRV-M and

Mam-RRV-R inoculated mice had significantly elevated levels of IFNβ gene expression with 31-32 fold change difference compared to PBS inoculated mice. Additionally Mam-RRV-M inoculated mice had a 5.5 fold increase in gene expression of IFNβ when compared to Mos-RRV-M and Mam-RRV-R inoculated mice had a 4.5 fold increase when compared to Mos-RRV-R inoculated mice. Gene expression of IFNγ was elevated for Mos-RRV-M inoculated mice with a 67.5 fold difference compared to PBS inoculated mice.

Mam-RRV-M inoculated mice had elevated IL6 gene expression by 17 fold and

Mos-RRV-M inoculated mice were up-regulated in regards to IL12 expression by 8 fold compared to PBS inoculated mice. There was no significant differences in gene expression of CCL2, IFNα, IL4, IL10, MIF and TNFα between RRV and PBS inoculated mice at peak disease in the ankle joint.

Discussion

In the previous chapter, it was found that differences in the disease profiles of Mam-RRV-T48 and Mos-RRV-T48 occurred in a male model of disease. To further investigate if this was unique to the laboratory strain of

RRV or if this was typical of wild circulating RRV strains, several field isolates

228 taken from mosquitos in WA were obtained and used in the mouse model of

RRV disease.

Weight gain and clinical score trends in mice inoculated with mammalian and mosquito derived RRV isolates.

The reduced weight gain of mice inoculated with RRV is due to disease pathology which limits the mobility of the mice, likely due to malaise. Mice typically rest and resist any movement or exercise in the peak stages of severe disease; they access less food, resulting in the observed weight loss. It is also possible that in some instances there is a reduced urge to eat as well as decreased movement but this is difficult to quantify.

Of the 18 isolates tested in the mouse model, all isolates resulted in significant differences in either percentage weight gain or clinical score in RRV inoculated mice when compared to PBS inoculated mice. Additionally, inoculation of mice with 14 (79 %) of the isolates tested produced a significant difference between Mam-RRV and Mos-RRV inoculated mice in either weight gain and/or clinical score. In the isolates that resulted in differential disease profiles in mice, disease scores were always more severe in mice inoculated with Mos-RRV, indicating that viral replication in mosquito cells has an impact on disease progression in a majority of cases for RRV in the isolates tested. There was not one instance of Mam-RRV resulting in a more severe clinical score in mice than the mosquito derived counterpart. It is worth noting that due to the animal numbers required these disease profile studies were only completed once however five mice per group were used to ensure

229 that rigorous statistical analysis could be completed and with the statistical tests used it is more likely that a false negative would occur than a false positive.

It is possible that the four isolates that have been grouped into Figure

6.3A simply have smaller differences that are more difficult to determine. In fact all four of these isolates do have points in the clinical disease score that are significantly different and in all cases it is again the mice inoculated with the Mos-RRV isolate preparation that have higher scores than the Mam-RRV isolate inoculated mice. Smaller differences may be further elucidated in the isolates grouped into Figure 6.3A if the disease time course could be repeated and additionally if further investigation was completed such as histology and cytokine studies. It is interesting to note that although Mos-RRV inoculated mice consistently had higher clinical scores than Mam-RRV inoculated mice, there were several instances where Mos-RRV inoculated mice had higher weights than Mam-RRV inoculated mice. Isolates where this occurred were

RRV-B, -O and –P. One explanation for the increase in weight gain in Mos-

RRV inoculated mice is that a minimal impact on mobility results in exercise impairment but not food seeking behaviours. C57Bl/6J mice are prone to weight gain, depression and overeating and are often used for obesity studies and these factors would have a greater effect on this strain than in other mice strains (Guo et al., 2009; Iniguez et al., 2014). For this thesis, two isolates were chosen from group D that produced the greatest variation in weights and clinical scores in inoculated mice, isolates M and R. These were used to further explore these differences and the consequences on the immune

230 response and pathology of resulting disease. For these two isolates the experiment was repeated (data not shown) to ensure reproducibility and results were consistent with the first experiment.

Investigation of differential viral titres in tissue of mice inoculated with RRV field isolates.

There was no difference in the viral titres present in blood for RRV-M and RRV-R inoculated mice at 24 hours post infection or at peak disease.

Interestingly Mam-RRV-T48 was found to be significantly higher than Mam-

RRV-R and Mos-RRV-T48 was found to be significantly higher than Mos-RRV-

R at 24 hours post infection. This difference may be due to immune response differences which impede viral replication. Mice inoculated with RRV-R had lower viral titres in the quadriceps muscle and ankle joint compared to RRV-

T48 at 24 hours but only the difference between Mam-RRV-T48 and Mam-

RRV-R in ankle joints was found to be statistically significant. By peak disease, there was no significant difference in viral titres recovered for mice inoculated with RRV-T48 and RRV-R in blood or quadriceps. Mam-RRV-T48 and Mam-RRV-R were still significantly different in ankle joints however they were reversed with Mam-RRV-R being significantly higher than Mam-RRV-

T48. It is important to note that at 24 hours post infection Mam-RRV and

Mos-RRV were deemed significantly different to each other for all groups in quadriceps and for RRV-T48 and RRV-R in ankle joint. As blood titres were comparable, this difference is indicative of a difference in immune response occurring within different tissues resulting in different levels of viral replication or survival. Surprisingly, for both RRV-T48 and RRV-M viral titres

231 for Mam-RRV were higher than Mos-RRV but this was reversed for RRV-R.

When comparing RRV-R to the other isolates, in each case Mos-RRV-R is similar to other Mos-RRV isolates but Mam-RRV-R is lower.

Taken together this indicates that Mam-RRV-R is not as efficient at establishing early infection in the mouse host as the other viral preparations but may be slower in doing so as the viral loads were not reduced at peak disease. Interestingly a study by Stoermer Burrack (2014) has also shown that the innate immune response to RRV is tissue specific causing variations in viral titres in different tissues. It is interesting to note that all mice inoculated with RRV isolates had recoverable virus in brain tissue. Previously brain involvement has been discounted as unimportant in the mouse model of RRV disease. However, the fact that virus is found in the brain indicates that it may have some impact on disease progression in the mouse model.

Differential tissue pathology in mice inoculated with RRV field isolates at 24 hours post infection

At 24 hours post inoculation pathological changes in tissue were not expected as it was considered to be too early to see changes in tissue, however changes can be seen in all three areas examined indicating that the inflammatory response is quick and in some cases, severe. Signs of mild inflammation were present in mice inoculated with Mam-RRV isolates at 24 hours post infection but were more clearly seen in mice inoculated with Mos-

RRV isolates. Both Mos-RRV-M and Mos-RRV-R inoculated mice showed evidence of inflammatory infiltrates, fibre necrosis and fibrosis in quadriceps

232 sections. The fast acting effect of infection on the muscle is supported by the clinical scores. In addition, it is interesting to note that the clinical scores of mice inoculated with Mos-RRV-M were highest and the quadriceps sections show more signs of inflammation induced pathology than those of mice inoculated with Mos-RRV-R. However, this is not reproduced in the context of ankle histology. In the ankle, mice inoculated with RRV-R show greater disruption to the ankle architecture than mice inoculated with RRV-M. All isolates showed evidence of moderate inflammatory response in brain tissue.

Inflammation in the brain is supported by the observation of visible cerebral oedema resulting in a frontal cranial dome on RRV inoculated mice. This was normally visible several days into infection rather than at 24 hours post infection and was more prominent in mice inoculated with Mos-RRV isolates but did occasionally occur in Mam-RRV inoculated mice. It is possible that mice are also suffering encephalitic symptoms along with arthritic symptoms in this instance. These symptoms would include headaches, muscle soreness, nausea and light sensitivity and this may contribute to the severity of the weight loss and clinical score measured. As inflammation and disease is clearly seen in these mice, brain involvement in the process should not be ruled out in the mouse model as it previously has and should be considered as a possibility when examining clinical scores. However, strong evidence supports the linking of symptoms to hind limb inflammatory pathology and there is no reason to believe that the current scoring system is not sound in its measure of inflammatory pathologies in the hind limb of RRV infected

C57Bl/6J mice.

233

Differential tissue pathology in tissue of mice inoculated with RRV field isolates at peak disease post infection

Although signs of inflammatory tissue damage are present at peak disease in both Mam-RRV inoculated mice tissue samples they are most prevalent in the Mos-RRV inoculated mice tissue sections. Large numbers of inflammatory infiltrates are present as well as destruction of muscle fibres.

This is seen again in the ankle sections where severe inflammatory damage and bone destruction has occurred in Mos-RRV inoculated mice and milder inflammatory pathology has occurred in the Mam-RRV inoculated mice sections. For the two isolates studied, results indicate that an aggressive inflammatory response occurs in response to viral infection in the muscle and ankle of inoculated mice, which is worse if the virus used for inoculation is propagated in a mosquito cell line. This infection results in tissue damage and affects the use of the hind limbs in agreement with previous finding using other RRV strains. Inflammation has previously been shown to persist past the resolution of the viral infection and it is the immune inflammatory response that results in the majority of tissue damage rather than any effect of the virus itself. Interestingly in both cases, the most severe levels of inflammation were seen in mice infected with RRV-R indicating some difference in the way the immune system responds to the different isolates.

Bone loss during RRV infection is supported by a recent study which found that not only did bone loss occur but that it could be prevented by IL6 inhibition (Chen et al., 2014b).

234

Despite the lack of detectable virus at peak disease in the brain, histology sections reveal signs of inflammation. This indicates that encephalitis may have some impact on disease progression such as weight loss. Additionally it is supported by the cranial encephalitis that was observed over the course of the experiment. Interestingly this was observed predominantly in Mos-RRV inoculated mice and was worse at times of peak disease. While the first examination of brain histology in RRV mouse studies occurred in 1973, there has been little further investigations into this as a contributing factor to disease (Mims et al., 1973). These previous studies identified general neurological lesions at early time points which developed into general monocyte infiltration by day 13; it was deemed that this was not the source behind the hind limb paralysis seen. However, what effect this brain involvement does have on the disease has not been investigated and may provide valuable insights. As many of the alphaviruses are encephalitic viruses, and even those deemed arthritic in nature have been shown to be able to cause encephalitis such as ChikV, rejection of brain infection as a factor of disease appear premature (Chandak et al., 2009; Zacks and Paessler,

2010). C57Bl6 mice infected with mouse passaged ChikV were found to have brain inflammation and necrosis and displayed symptoms such as ruffled fur, lethargy and hunched posture and died by day 8 post infection (Wang et al.,

2008). Additionally, recent studies of BFV in a mouse model have shown that virus is able to reach high titres in the brain, also that this virus does not cause the same arthritic disease seen in RRV with hind limb dysfunction, but does induce lethargy, ruffled fur and a hunched posture (Herrero et al., 2014).

It is possible that these symptoms may in part be attributed to involvement

235 of the brain by the virus in both cases, and thus it is possible that infection of the brain in RRV mouse model leads to a mild encephalitic condition compounding the arthritic symptoms caused by inflammatory pathology in the hind limbs. Future studies could look at the link between encephalitic involvement and severity of disease.

Cytokine profiles in quadriceps muscle of mice inoculated with RRV field isolates

CCL2 is primarily involved in the recruitment of immune cells such as monocytes, basophils, eosinophils and dendritic cells. It has been linked to an overexpression of IL4, resulting in defects in cell-mediated immunity and also affects adaptive immunity by skewing T helper cell polarisation to the TH2 response (Gu et al., 2000). At six hours post infection CCL2 was significantly higher in Mos-RRV-R inoculated mice compared to Mam-RRV-R and PBS controls. At the same time point there was a statistically significant rise in

IL4 expression for Mos-RRV-M inoculated mice; in addition, at 12 hours post infection there was a huge gain in the expression levels of IL4 for Mos-RRV-

M, Mam-RRV-R and Mos-RRV-R inoculated mice, indicating a polarisation of

T helper cells to a TH2 response. It is of interest to note that signs of disease were least severe in Mam-RRV-M inoculated mice for clinical score and histology and that these mice also did not have significantly higher levels of

CCL2 in muscle at any time. This is in support of the findings by Rulli et al.

(2009) who found that elimination of CCL2 in mice through bindirit treatment, an anti-inflammatory drug, resulted in a mild attenuated disease presentation in mice. By 24 hours post infection raised IL4 levels had resolved but

236 expression of CCL2 was still elevated. By peak disease, both CCL2 and IL4 expression levels were similar to controls indicating that these cytokines have a role in the early stages of pathology in the quadriceps muscle. Interestingly, although a predominant TH2 response appears to be emerging in the first 24 hours post infection, by peak disease Mam-RRV have higher expression levels of TH1 cytokines such as IFNγ and IL12 as well as having higher levels of IFNβ than controls or Mos-RRV inoculated mice. Interestingly Mos-RRV M and

Mam-RRV-R inoculated mice had significantly higher expression levels of

TNFα than controls. In addition, Mos-RRV-M was significantly higher than to

Mam-RRV-M. There are a few instances where the cytokine profiles differ between RRV-M and RRV-R inoculated mice, indicating that immune function may be quite variable between isolates, especially considering that for this project isolates that had similar disease profiles were chosen for further study.

In addition, at 24 hours post infection there was down regulation of TH1 and inflammation as seen by the low expression levels of CXCR3, IFNγ and MIF.

IL12 expression levels were also low, however none of the mouse groups were found to be significantly different to PBS controls. IL6, a TH2 cytokine, was up-regulated in Mam-RRV-M inoculated mice. The results seen at peak disease were contradictory, in that IFNβ has been shown to reduce TH1 responses through inhibition of IL12 production by dendritic cells and also to inhibit IFNγ secretion by T cells yet these were all significantly higher at peak disease (Bartholome et al., 1999). Additionally while IFNβ has been shown to inhibit TNFα production by 88 % (Jungo et al., 2001), there are several reports of both being induced together in a synergistic fashion in response to viral infection (Goldfeld and Maniatis, 1989). There are currently three hypotheses

237 around the interactions of IFNβ and TNFα; the first is that the cytokines act as two opposing forces both inducing opposite effects, where equal amounts result in an equilibrium leading to protective immunity and that disruption of this equilibrium past certain thresholds results in pathological states of autoimmunity, allergy or inflammation (Ivashkiv, 2003). Alternatively, type I

IFN and TNFα actively influence each other’s levels and when balance is lost, pathology occurs. Another theory states that type I IFN is involved in the initiation of autoimmunity and that separately TNFα levels increase in the secondary inflammation phase. To complicate this, many studies give conflicting results in regards to the regulation of these two cytokines (Cantaert et al., 2010).

Cytokine profiles in ankle joints of mice inoculated with

RRV field isolates

Gene expression of CCL2 was significantly up-regulated in the ankle joints of Mos-RRV inoculated mice at six hours post infection. In addition,

Mos-RRV inoculated mice were found to be significantly different to their

Mam-RRV inoculated mouse counterparts. Increased gene expression of

CCL2 in the ankle joint would be associated with an increase in inflammatory cells and is found at the sites of bone degradation. These findings fit with the histological images presented earlier. It is however interesting to note that this is an early stage effect as the levels of CCL2 are not significantly up- regulated at any other point in the ankle joint and are in fact significantly down-regulated at 24 hours indicating that infiltration of inflammatory cells occurs very early in infection in mouse ankle joints. At 12 hours post

238 infection, the gene expression of IL10 in Mos-RRV-R inoculated mice is significantly up-regulated and IL12 is appropriately down-regulated, indicating a TH2 response is occurring. By 24 hours post infection IL10 gene expression levels are significantly up-regulated in all RRV inoculated mouse groups, indicating again that early response to infection is a TH2 response.

Also at 12 hours post infection, type I IFN was significantly up-regulated in Mos-RRV-M but by 24 hours this had changed to being up-regulated in

Mam-RRV-M and by peak disease, both Mam-RRV isolates had significantly up-regulated gene expression of IFNβ but not IFNα. IFNβ, as well as being antiviral when expressed, is also able to supress immunity and prevent pathological inflammatory conditions. A recent study has identified an IFN- stimulated gene which encodes for the enzyme cholesterol 25-hydroxylase

(Ch25h) which results in the production of 25-HC which in turn reduces IL1b transcription and inflammasome activity. Mam-RRV inoculated mice had less inflammatory pathology and tissue/bone destruction than Mos-RRV inoculated mice and the increased levels in Mam-RRV inoculated mice may, at peak disease, be one explanation for this.

Conclusion

Mam-RRV and Mos-RRV infected mice have different levels of weight gain post inoculation and different clinical scores in a majority of the field isolates tested. Additionally Mos-RRV is always the most severe in clinical score or failure to gain weight when compared to Mam-RRV field isolate infection of mice. Of the chosen isolates, Mos-RRV histology of quadriceps

239 and ankle showed a higher influx of inflammatory cells and bone destruction at peak disease when compared to Mam-RRV. Cytokines revealed that whilst there was an up-regulation of type I IFN expression in Mos-RRV inoculated mice at early time points, this was altered at peak disease with Mam-RRV inoculated mice having high expression levels of IFNβ. These differences confirm that differential disease profiles can be induced by Mam-RRV when compared to Mos-RRV and that Mos-RRV is more pathological in a model of disease.

240

Mammalian and mosquito derived arbovirus phenotypes

Introduction

Investigations into the different behaviour of Mam-RRV and Mos-RRV in this thesis have shown that differences exist in the way that Mam-RRV and

Mos-RRV interact with the host system. Further investigations are required to assess if this differential induction between Mam-RRV and Mos-RRV is limited to RRV or if this is common among Alphaviruses. In addition, these findings may be relevant to arboviruses in general. Shabman et al., (2007) found that the infectivity of mammalian derived RRV and VEEV was reduced in DC's, but that the induction of IFNβ was increased compared to IFNβ induction by mosquito derived virus. In contrast, wild type BFV, lacking E2 glycans, was used to assess IFNβ induction and no significant difference was found between Mam-BFV and Mos-BFV. However when using the 10E101SC strain of BFV, which contains N-linked glycans on its E2 glycoprotein, Mam-

BFV-10E101SC induced significantly higher induction of IFNβ in DCs than

Mos-BFV-10E101SC (Shabman et al., 2007). Further investigations were conducted on the role of E2 and E1 glycosylations in RRV infection and the effect on virus host interaction. It was found that whilst Mam-RRV had hybrid and complex oligosaccharides at the glycosylation points on E2 and E1, Mos-

RRV had high mannose and paucimannose oligosaccharides. It was later

241 shown that the complex glycans present on Mam-RRV were responsible for the strong IFNβ response in DC’s (Shabman et al., 2008).

Infectivity studies using the Alphavirus SinV, found that propagation in mammalian cells resulted in the production of heterogeneous particles with two different particle populations identified according to virion density; heavy and light, however, virus propagation in a mosquito cell line resulted in a homogenous particle population. Interestingly infectivity was increased for virus propagated in mosquito cells and for the heavier mammalian derived

SinV, compared to lighter mammalian derived SinV particles, with enhanced viral translation and RNA synthesis early in infection. In addition when passaged in mammalian cells, there were more infective light particles produced than heavy particles as determined by plaque assay (Sokoloski et al., 2013). Research into the propagation and isolation of DenV also found that virus production was greatly increased in mammalian cells if the virus was previously passaged in the mosquito C6/36 cells (Phanthanawiboon et al., 2014).

Mam-SinV light particles resulted in significantly higher production of

IFNβ in L929 cells than either the mammalian derived heavy particles or the mosquito derived particles (Sokoloski et al., 2013). This indicates that there may be several factors, rather than glycosylation alone, that affect host response to viral particles propagated in either mammalian or mosquito cell lines. Hence, it is important to identify which effects are shared among

Alphaviruses and also more generally among arboviruses.

242

Results

Plaque morphology of Mam-BFV and Mos-BFV

Titration of Mam-BFV and Mos-BFV on Vero cells revealed different plaque phenotypes for the two viral preparations. A majority of Mam-BFV plaques were larger and more diffuse than those of Mos-BFV, although it is noted that there were several smaller plaques present. The Mos-BFV plaques were smaller and more defined than Mam-BFV plaques (Figure 7.1).

Figure 7.1 - Differential plaque morphology of mammalian and mosquito derived Barmah Forest virus

Mammalian derived Barmah Forest virus and Mosquito derived Barmah Forest virus were plaqued on Vero cells and incubated for 48 hours before being stained with crystal violet. Plaque morphology was then visually compared.

Multistep replication kinetics of Mam-BFV and Mos-BFV in Raw 264.7 cells

The growth kinetics of Mam-BFV and Mos-BFV virus in Raw 267.4 cells was within 10 fold of each other at most early time points. However, Mam-

BFV virus titres declined after the peak at 12 hours post infection significantly more rapidly than Mos-BFV virus. This rapid decline by Mam-BFV virus

243 resulted in a difference of 100 fold between the two viral preparations at the final 24 hour time point.

) 6 L *

/ 5

p 4 *

( 3

t 2

i T 1 0 0 10 20 30 Hours post infection

Figure 7.2 - Growth kinetics of mammalian and mosquito derived Barmah forest virus in Raw 264.7 cells.

Mammalian derived Barmah Forest virus ( ) and mosquito derived Barmah Forest virus ( ) were used to infect Raw 264.7 cells at an MOI of 0.01 to obtain multistep growth curve kinetics. Samples were taken at 0, 2, 4, 6, 12, and 24 hours post infection and titrated by plaque assay. Each data point represents the mean of three separate experiments and error is calculated as ± SEM. Statistical significance (*, p<0.05) was determined by two-way ANOVA with Tukey’s test.

Differential IFNβ induction by mammalian and mosquito derived Barmah Forest virus in Raw 264.7 cells

Examination of the gene expression for IFNβ showed different expression profiles for Mam-BFV and Mos-BFV. Infection with Mam-BFV

244

resulted in significantly higher IFNβ expression than Mos-BFV and PBS at

both four and six hours post infection. It is interesting to note that expression

was highest in cells infected with Mam-BFV at the earlier four hour time point

where it was 9 fold higher than the PBS group compared to 2 fold greater than

PBS at six hours post infection. Mos-BFV infected cells had expression levels

that were comparable to PBS at both time points and no significant difference

was found.

15000 *

e 10000

n o

d 5000 i 0 . 0 0 1 5 *

s s

i *

n e

l n

0 . 0 0 1 5 r o

i *

p s

i 0 . 0 0 1 5 x

m s

s * e

r 0 . 0 0 1 0

r v

p t 0 e

x 0 . 0 0 1 0

0 . 0 0 1 0 v 4 hours 6 hours

a i

d 0 . 0 0 0 5

t e

e *

i l

e 0 . 0 0 0 5

r a

e *

s m

d 0 . 0 0 0 5

r l

e *

a o

s 0 . 0 0 0 0

N l

Figure 7.3 - IFNβ inductionr by mammalian and mosquito derived Barmah forest virus 4 H o u r s 6 H o u r s a

o 0 . 0 0 0 0

m N r 4 H o u r s 6 H o u r s

o 0 . 0 0 0 0 Raw 264.7 cells were infected with mammalian derived Barmah Forest virus ( ) Mora m -D e n M o s -D e n P B S N 4 H o u r s mosquito6 H o u rderiveds Barmah MForesta m -D evirusn ( ) Mato ans -D MOIe n of 0.01 orP BoverlayedS with 100µL of M a m -D e n M o s -D e n PBS ( ). P SamplesB S were taken at four and six hours post infection. Cells and supernatant from infected wells was removed, total RNA isolated and analysed for mRNA expression by real time PCR. Data was normalised to the house keeping gene HPRT1 and expressed as relative expression to PBS inoculated mice. Results are represented as the mean of each group (n=5) ± SEM. Statistical significance (* p<0.05) was determined by one way ANOVA with Tukey’s test.

245

Plaque morphology of Mam-DenV and Mos-DenV

DenV plaque phenotype was examined by using an immuno-focus assay in Vero cells outlined in section 2.3.6. Mam-DenV generated large diffuse plaques in contrast to smaller defined pin-prick plaques created by

Mos-DenV.

Figure 7.4 - Differential plaque morphology of mammalian and mosquito derived dengue virus

Mammalian derived dengue virus and mosquito derived dengue virus was visualised on Vero cells by immune-focus assay. Virus was overlayed on Vero cells and incubated for 4 days before being stained with antibodies in order to visualise plaques as outlined in section 2.3.6.

Multistep replication kinetics of Mam-DenV and Mos-

DenV in Jaws II cells

Growth kinetic curves for Mam-DenV and Mos-DenV were found to be different, with significantly different titres seen at each time point. Mam-

DenV rose in titre until day two at which point it remained stable for the remainder of the time course. In contrast, Mos-DenV rose nearly to its highest titre on day one with only a small rise from day one to day two, before a steady decline that continued over the time course.

246

6 *

) 5

u 4

g 3

(

e 2

T 1 0 0 2 4 6 Days post infection

Figure 7.5 - Growth kinetics of mammalian and mosquito derived dengue virus in various Jaws II cells

Mammalian derived dengue virus ( ) and mosquito derived dengue virus ( ) were used to infect Jaws II cells at an MOI of 0.01 to obtain multistep growth curve kinetics. Samples were taken at 0, 2, 4, 6, 12, and 24 hours post infection and titrated by plaque assay. Each data point represents the mean of three separate experiments and error is calculated as ±SEM. Statistical significance (*, p<0.05) was determined by two-way ANOVA with Tukey’s test.

Differential IFNβ induction by DenV in Jaws II cells

Examination of the gene expression profile of IFNβ in Jaws II cells revealed different expression profiles for Mam-DenV and Mos-DenV. Mam-

DenV induced significantly higher IFN expression at both time points compared to Mos-DenV and PBS. Additionally, Mam-DenV induced IFNβ was highest at the four hour time point, 63 fold higher than PBS, compared to 6

247

hours post infection when gene expression had dropped to 15 fold higher than

PBS.

0.0015

n *

e 0.0010

a n

l o

i 0 . 0 0 1 5

n * s

r o

i 0 . 0 0 1 5 n

* e

s r

o 0 . 0 0 1 5 0.0005 s

i *

s e

s * x

r s

r 0 . 0 0 1 0

p e

r 0 . 0 0 1 0

e e

0 . 0 0 1 0

e i

r a

t 0.0000

l a

d 0 . 0 0 0 5

r e

e 4 Hours 6 Hours *

d 0 . 0 0 0 5

0 . 0 0 0 5 l d

e * a

e *

i s

a a

o 0 . 0 0 0 0

m N

Figure 7.6 - IFNβ induction ofr mammalian and mosquito derived dengue virus

r 4 H o u r s 6 H o u r s o

o 0 . 0 0 0 0 0 . 0 0 0 0

N N Jaws II cells4 H wereo u r s infected with 6 mammalian H o u r s derived4 H o u r s dengue virus ( 6 H),o Mu mosquitorasm -D e n derived M o s -D e n P B S dengueM a m -D evirusn ( ) Mato ans -D MOIe n of 0.01M a morP- DoverlayedB Se n withM 100µLo s -D eofn PBS ( ). P SaB Smples were then taken at four hours and six hours post infection. Cells and supernatant from infected wells was removed and total RNA was isolated and analysed for mRNA expression by real time PCR. Data was normalised to the house keeping gene HPRT1 and expressed as relative expression to PBS inoculated mice. Results are represented as the mean of each group (n=5) ± SEM. Statistical significance (* p<0.05) was determined by one way ANOVA with Tukey’s test.

Discussion

The investigations covered in this thesis have revealed many differences

between Mam-RRV and Mos-RRV in regards to plaque morphology, in vitro

kinetics, in vivo disease progression and gene expression. Continuing from

these studies, the next step was to see if these finding were applicable only to

248

RRV or could be applied to other arboviruses. There have been some initial reports of differences in other arboviruses such as WNV, SinV, VEEV, BFV and DenV (Frolov et al., 2012; Hacker et al., 2009; Klimstra et al., 2003;

Shabman et al., 2007). Some of these include plaque morphologies and IFNβ induction, but each study focuses on a separate virus rather than including several arboviral genera and families. For the purposes of this study, BFV was chosen from the same family as RRV, and DenV, a Flavivirus, was chosen as a representative from a different arboviral family. Initial investigations examined the plaque morphology, the growth kinetics in the primary target immune cell and the induction of IFNβ.

Differential plaque phenotypes of arboviruses

Both Mam-BFV and Mam-DenV resulted in large diffuse plaques compared to Mos-BFV and Mos-DenV respectively. This is similar to the finding with Mam-RRV-T48 and Mos-RRV-T48 (Chapter 3). This indicates a growth advantage for the mammalian derived viruses in a mammalian cell line in the absence of immune functions and type I IFN, with surrounding cells succumbing easily to infection and thus leading to development of cytopathic effect. Mosquito derived virus appears to have growth restrictions in Vero cells contributing to the pin prick, defined plaque morphology of arboviruses in this study. This is an interesting initial screening process that may help to identify if differences are likely to occur between mammalian and mosquito derived viruses.

249

Differential growth kinetics of arboviruses in immune cell lines

Macrophages have been identified as primary targets and critical disease mediators for RRV as well as permitting persistent infection. The primary target cell for BFV has not yet been elucidated. Previous studies have established that in DC’s Mam-BFV induces higher levels of IFNβ than Mos-

BFV, but only for strains containing the N-linked glycans on the E2 glycosylation site (Linn et al., 1998). A role for macrophages has not yet been investigated and as BFV is closely related to RRV, macrophages are likely important for infection and pathogenesis. On the other hand, DC’s are generally accepted as the primary target cell in early DenV infection (Libraty et al., 2001; Marovich et al., 2001; Palmer et al., 2005). Interestingly, for both

BFV and DenV infections in vitro, mosquito derived preparations grew to significantly higher titres at early time points. This indicates some advantage to mosquito derived virus, whether this is an attachment or entry advantage due to glycosylations, an immune evasion mechanism such as reduced induction of IFNβ or an unknown mechanism, will require further investigation to elucidate. Additionally, it is interesting to note that in both cases the highest titres recorded were for the mosquito derived preparation of the virus. This again relates well to the data presented in Chapter three, where Mos-RRV-T48 had significantly higher titres compared to Mam-RRV-

T48 from six hours post infection in both Raw 264.7 cells and Jaws II cells.

The differential growth kinetics for DenV is surprising as it is in contrast to a previous study which reported no significant difference between Mam-DenV

250 and Mos-DenV when comparing percent of infected primary human DC

(Hacker et al., 2009). Whilst the number of infected cells between the two preparations was not significantly different, it does not preclude resulting overall titres of infectious particles being different, but titres were not measured in the study. Additionally, the DC’s used in the study conducted by Hacker et al., (2009) were differentiated from monocytes that had been extracted from whole blood and there is no mention of the success rate in obtaining human DC’s, which would need to be clarified to remove the possibility of other cells contributing to the resulting titres.

IFNβ induction by mammalian and mosquito derived arboviruses

The induction of IFNβ was highest in the mammalian derived preparation for both BFV and DenV. Contrary to expectations, both Mam-

BFV and Mam-DenV had the highest levels of IFNβ gene expression at the earlier four hour time point. Given that DenV has a longer propagation time it would have been expected that the induction of IFNβ would also be delayed.

However, this very early induction supports a mechanism that occurs on attachment to cells rather than any process that occurs after entry or during replication. This is in agreement with previous studies showing that the N- linked glycosylations on the E protein play a critical role in IFNβ induction

(Shabman et al., 2008). Additionally, studies into the N-linked glycans of

DenV have found that Mam-DenV contains a mix of high mannose and complex glycans, whilst Mos-DenV has a mix of high mannose and paucimannose glycans (Hacker et al., 2009). Interestingly, the DenV protease

251

NsP2B3 from Mos-DenV has also been found to inhibit the production of type

I IFN in human DC’s at 5 days post infection through cleavage of stimulator of interferon genes (STING), although it is unable to cleave mouse STING

(Aguirre et al., 2012). NsP2B3 is a viral protease which cleaves the DenV polyprotein synthesised in the first steps of the viral cycle. This requires cell entry and the beginning of viral replication to occur, thus this effect would not come into play at the early time points being investigated in this thesis but may be an additional mechanism that is important at later points in the infection process.

Conclusion

Differences between mammalian and mosquito cell derived virus are not limited to RRV or the Alphavirus family. The related Alphavirus, BFV, and the

Flavivirus, DenV, both display similar differences between mammalian and mosquito derived preparations. Mammalian cell derived virus results in a larger diffuse plaque when using Vero cells, indicating a growth advantage of the virus in these cells compared to mosquito derived virus. When grown in the target immune cell, mosquito derived virus produces higher titres early in infection and reaches higher peak titres than mammalian derived virus. This indicates that there are mechanisms that enable mosquito derived virus to gain an advantage in the early stages of infection. Examination of the IFNβ gene expression reveals early advantages for mosquito derived arbovirus with a lack of induction of IFNβ. In contrast, mammalian derived virus induces high levels of IFNβ gene expression at early time points, initiating host immune response before the virus has a chance to establish a solid infection.

252

Taken together, this indicates that at very early stages mosquito derived virus has an advantage in mammalian cells over mammalian derived virus. As this occurs in very early stages, prior to replication, this difference can only be due to differences in the particles and/or cell recognition between mammalian and mosquito derived virus. For WNV, SinV and DenV, the presence of cell surface

DC-SIGN distinguishes between mammalian and mosquito derived virus, resulting in a more virulent infection by the mosquito derived virus in each case. In the case of WNV this has been attributed to the lectin carbohydrate recognition domains, whilst for DenV this is attributed to the terminal mannose residues present only on Mos-DenV (Davis et al., 2006; Hacker et al., 2009; Klimstra et al., 2003). There may be many differences occurring during translation and assembly of the virus particles in mammalian or mosquito cells that cause host cells to react differently and this area requires further investigation.

253

Discussion The studies undertaken in this thesis have shown that there is a difference in the disease pathology in mice infected with mammalian derived virus compared to mosquito derived virus. Thesis studies show that Mos-RRV causes a more severe disease than Mam-RRV. Whilst the mechanisms behind this difference is yet to be elucidated, valuable steps forward have been made in understanding how this difference occurs.

Arobovirus growth kinetics in tissue culture

The origin of the cell line in which RRV-T48 is propagated, has been shown to have an effect on replication kinetics in myeloid DCs (Shabman et al., 2007). This effect was also observed following RRV-T48 infection of Vero,

C6/36, Raw 264.7 and Jaws II cells as discussed in Chapter 3. Both Vero and C6/36 cells are immune deficient as these cells are unable to produce antiviral factors in response to viral infection (Brackney et al., 2010; Mosca and Pitha, 1986). Repeat passage in either cell line provided RRV-T48 with an advantage during a subsequent infection of the same cell line. This is indicative of RRV’s ability to quickly adapt to the chosen host. These findings are in agreement with the theory that mutations in arboviruses are kept in check through alternating replication cycles in different hosts. This alternating replication pattern restricts the ability of the virus to mutate without causing a fitness loss in the alternate host. Furthermore, previous findings have shown that SinV is capable of adapting to different host cell lines within one passage through mutations in the 5’ end of the genome

255 increasing fitness in that cell line (Fayzulin and Frolov, 2004). Although evidence for genetic mutation exists, literature also supports the possibility of a non-genetic adaptation method through differences in glycosylation of the

E protein. In addition, one passage through the alternate cell line will revert the virus back to its alternate profile (Shabman et al., 2007). However, RRV has been shown to have more genetic variation than other Alphaviruses and it is not possible to rule out mutations (Lindsay et al., 1993a). As the natural reservoir for RRV is thought to be marsupials, in particular kangaroos, it would be important for future studies to examine cell lines from these species and see what effect, if any, this has on viral adaptation and fitness. These findings would contribute to better understanding of the mechanism behind

RRV viral adaptation.

Differences between Mam-RRV-T48 and Mos-RRV-T48 plaque size and replication kinetics were similar for both, BFV and DenV as previously discussed in Chapter 7. Viral propagation of mammalian derived virus in

Vero cells was characterised by a large diffuse plaque and an increase in viral titres. However, viral propagation of mosquito derived virus in Vero cells was characterised by a smaller pin prick plaque and reduced titres. RRV isolates behaved similarly. Infection of Vero cells with Mam-RRV isolates resulted in higher titre compared to Mos-RRV isolate infection by 48 hours post infection.

Additionally, Mos-RRV infection of C6/36 cells resulted in higher titres than

Mam-RRV isolates infection of C6/36 cells, as previously discussed in

Chapter 4. These findings, support the theory that RRV is able to quickly adapt to the host cell line and confer a fitness advantage in the new host.

256

Whilst this ability may mean trade-offs in the initial round of replication, it is indicative of its ability to adapt to the specific host and increase its infection potential.

When examining arboviral infection in an immune competent cell line such as the macrophage cell line (Raw 264.7), it is not only the capacity of the virus to adapt to a different cell line that is of importance but also the viral interaction with host immunity. As expected, titres recovered from Raw 264.7 cell infections were significantly lower than those in either Vero or C6/36 cells for all viruses tested, suggesting that these cells are able to mount some defence against infection but are not resistant to infection.

In addition, Mos-RRV-T48 replicated to higher titres than Mam-RRV-

T48 at all-time points in macrophage (Raw 264.7) and DC (Jaws II) lines. This was also true for the replication kinetics of BFV in a macrophage cell line. In contrast, Mos-DenV replicated to higher titres for the first 3 days post infection in a DC line compared to Mam-DenV; however, Mos-DenV viral titres were lower than titres of Mam-DenV for days four and five post infection. As

RRV and BFV replication cycles are more rapid than those of DenV, this experiment was extended for several days to accommodate for this delay; however, it is possible that the drop in viral titres observed for Mos-DenV may have also occurred in RRV and BFV infections if their respective time courses were extended as well. From these results it appears that mosquito derived arbovirus has a fitness advantage in the early stages of infection. Indeed previous studies found that enhanced infection of DCs by Mos-DenV compared to Mam-DenV in primary human DC’s was due to an increase in

257

DC-SIGN binding affinity for the high mannose and paucimannose glycans present on Mos-DenV E asparagine-67 glycosylation site (Lozach et al., 2005;

Mondotte et al., 2007). Additionally, it has been shown that both Mos-SinV and Mos-WNV also benefit from enhanced infection of DCs through increased affinity with DC-SIGN to high mannose glycans present on the mosquito derived virus (Martina et al., 2008; Morizono et al., 2010). This information provides strong evidence that high mannose glycosylations present on mosquito derived virus enhance viral infectivity and give a fitness advantage to the virus in the mammalian host.

However, RRV field isolates showed mixed results in terms of infectivity for Mam-RRV and Mos-RRV in Raw 264.7 cells, with Mos-RRV isolates reaching higher titres in only around half of the isolates tested. Previous studies have shown that Mam-RRV-T48 and Mos-RRV-T48 infection of DC-

SIGN expressing cells results in differences in IFN induction and replication kinetics. This has been largely attributed to different N-linked glycans on the

E protein of mammalian and mosquito derived arboviruses factors (Shabman et al., 2007; Shabman et al., 2008). The multistep replication kinetics of different RRV isolates in Raw 264.7 demonstrates that Mos-RRV does not always reproduce to higher titres in cells expressing DC-SIGN and indicates that although glycosylation differences are important, other factors may also be at play. This is supported by reports that a mutation in the NsP1-NsP2 domain of both, SinV and RRV results in not only an increase in IFNβ production, but also a decrease in viral titres (Cruz et al., 2010). Taken together, glycosylation differences and viral genome variations together are

258 most likely responsible for the effects on viral replication that occur in immune cells by arboviruses.

Gender dimorphism in the RRV mouse model of disease

In order to further characterise the differences between mosquito and mammalian derived RRV, we examined whether the effect was relevant in the mouse model of disease using T48 as a pilot study (Chapter 5). During this study, it was discovered that whilst a differential disease profile did appear for Mam-RRV and Mos-RRV in a mixed sex mouse model, this was more prominent in a male only mouse model. Female mice did not develop as much difference in weight gain and clinical scores when inoculated with Mam-RRV or Mos-RRV and as examination of this difference was the focus of our study, females were found not suitable for the viral mouse model. Many studies involving animals limit the study to the use of one sex with the use of male animals usually preferred. This is to eliminate potential sources of variation, such as differences between the sexes, but also due to the oestrus cycle of female animals that leads to hormonal fluxes which can cause result inconsistencies (Marriott and Huet-Hudson, 2006; Suffredini, 2007). Studies have shown that concentrations of pro-inflammatory cytokines such as IL6 are higher in men compared to women and that CCL2 and soluble vascular cell adhesion molecule-1 (sVCAM1) are decreased in the presence of oestrogen

(Chan et al., 2002; Register et al., 2005; Sperry et al., 2008). In addition, females have a higher incidence of apoptosis in inflammatory periodontal

259 disease which removes inflammatory cells and promotes healing, leading to better health outcomes (Johnson and Wikle, 2014). This is supported by a study which examined periapical inflammation and associated bone loss in mice and found that male mice had higher levels of inflammatory infiltrates which correlated to an increase in bone loss (McAbee et al., 2012).

Interestingly, in human subjects females are usually deemed at higher risk of the inflammatory disease, rheumatoid arthritis and have been shown to have lower innate immunity responses (Kovacs and Olsen, 2011; Moxley et al.,

2002). Differences in bone densities between males and females has been widely documented, with female bone densities being significantly lower than males at all comparative stages of growth. This is of importance as women are approximately twice as likely to have a fracture in their lifetimes (Bonnick,

2006; Nguyen et al., 2007). Bone densities may have important implications for severity of disease pathology associated with RRV if bone tissue comes under attack.

Taken together with the results shown in this thesis there is mounting evidence to support the idea that gender dimorphism may play an important role in inflammatory pathology and disease outcomes of RRV in the mouse model and future studies could compare these differences in more detail.

260

Mammalian and mosquito derived RRV results in differential disease profiles in a mouse model of disease

In the RRV disease mouse model, RRV-T48 infection resulted in a difference in disease pathology between Mam-RRV and Mos-RRV inoculated mice. Mos-RRV inoculated mice gained less weight and had higher clinical scores when compared to Mam-RRV infected mice. This is supported by previous studies, which found that IFNβ is differentially inducted by Mam-

RRV and Mos-RRV in tissue culture, with potent induction by Mam-RRV in

DCs (Shabman et al., 2007; Shabman et al., 2008). The differential induction capacity is indicative of differences in the early innate immune response and in turn, can influence disease development as demonstrated in our study.

This finding is also supported by the recent finding suggesting that the first few hours after Alphavirus infection are the most critical for determining disease outcome (Frolov et al., 2012). Infection of mouse embryonic fibroblasts (MEF) with SinV showed that infected cells were unable to respond to IFNβ by 4 hours post infection (Frolov et al., 2012). Additionally, low levels of IFNβ insufficient to initiate cellular antiviral responses, have been found to enhance the cells ability to express IFNβ and IFN stimulated genes on subsequent exposure to SinV (Frolov et al., 2012). Taken together, this suggests that during the first few hours after Alphavirus infection, there is a balance between viral replication and induction of host cell immune responses such as type I IFN that ultimately affects disease progression several days later.

261

The differential disease profile of Mam-RRV and Mos-RRV was significant in approximately 80 % of the RRV isolates tested in the mouse model as shown in Chapter 6. This finding is similar to findings in tissue culture and indicates that although the propagation in cell line plays an important role in RRV infection and subsequent disease pathology, it is not the only factor at play.

Trends in infection and disease caused by RRV field isolates in a male mouse model

Isolates can be grouped into four groups defined by the in vitro kinetic profiles discussed in Chapter 4 and the in vivo disease profile as discussed in

Chapter 6. For kinetics in Vero and C6/36 cells, the groups represented in

Table 8.1 are the same as those presented in Chapter 4. However, the relationship between viral titres in Raw 264.7 cells was more complex and for the purposes of this table has been simplified as to whether Mos-RRV produced significantly higher or lower titres by the end of the time course.

262

Table 8.1 - Correlation between in vivo and in vitro profiles of RRV isolates

Kinetics in Kinetics in Vero Kinetics in C6/36 RRV disease RRV Group Raw 264.7 cells cells model in mice isolates cells Mos-RRV titres No difference in titres significantly higher for Mos-RRV until late in time course majority of time course No difference in weights produced higher A when Mam-RRV titres but at one or more points or clinical score (Fig. H,L,F,D titres (Fig 4.6A were significantly Mam-RRV titres are 6.3A) and D) higher. (Fig 4.2C) significantly higher (Fig 4.4C)

Mos-RRV titres is 2 significantly higher at No difference in titres No difference in Mos-RRV early time points and until late in time course weights. Higher clinical produced higher

B Mam-RRV titres are when Mos-RRV titres score for Mos-RRV at Q,I,K,B 63 titres (Fig 4.6 A, significantly higher at were significantly higher later points in disease B and D) later time points (Fig (Fig 4.4B) time course (Fig 6.3B) 4.3B) Mos-RRV titres is Significant differences significantly higher at No difference in titres Mos-RRV in weights and higher early time points and until late in time course produced lower clinical score for Mos- C Mam-RRV titres are when Mos-RRV titres O,G,C,J,E,P titres (Fig 4.6A, B RRV at later points in significantly higher at were significantly higher and C) disease time course (Fig later time points (Fig (Fig 4.4B) 6.3C) 4.3B) Significant differences Mam-RRV titres Mos-RRV titres Mos-RRV in weights and higher significantly higher for significantly higher for produced lower clinical score for Mos- D A,R,M,N majority of the time majority of time course titres (Fig 4.6A RRV early, and course (Fig 4.3A) (Fig 4.4A) and C) continuing through time course. (Fig 6.3D)

The isolates in group A (Table 8.1) showed no difference between Mam-

RRV and Mos-RRV titres in Vero cells until late time points where Mam-RRV titres were significantly higher. In C6/36 cells, group A were the only isolates that had higher titres for Mos-RRV isolates for the majority of time points however, there is at least one time point at which Mam-RRV had significantly higher titres in C6/36 cells compared to the Mos-RRV isolate (Figure 4.4C).

In Raw 264.7, group A Mos-RRV isolates produced higher titres than Mam-

RRV. Infection of mice with Mam-RRV and Mos-RRV isolates in this group showed no significant difference in terms of mouse weight gain or clinical score. From this information, it appears that group A isolates did not exhibit an early fitness advantage when used to infect the same cells they were propagated in. Conversely, it appears that there is an advantage conferred for Mos-RRV in the infection of Raw 264.7, cells. Interestingly, this results in no differences seen in the mouse model of disease between infection with

Mam-RRV isolates compared to Mos-RR isolates. One possible explanation for this is that although there is an increased ability for Mos-RRV to preferentially infect immune cells (as shown by kinetics in Raw 264.7 cells), a reduction in adaptation ability, displayed by kinetics in Vero and C6/36 cells, resulted in a lag that negated any advantage given by increased infectivity alone.

In groups B and C (Table 8.1) isolates displayed characteristics shown in

Figure 4.2B for replication kinetics of Mos-RRV and Mam-RRV in Vero cells.

Mos-RRV isolates produced significantly higher viral titres at early time points while Mam-RRV isolates induced higher titres at later time points in Vero

264 cells. Isolates from both of these groups also shared the same pattern of replication kinetics in C6/36 cells, with no difference observed at early time points; however, Mos-RRV showed significantly higher titres at later time points (Figure 4.4B) However, differences were seen between the two groups for both in vitro replication kinetics in macrophage (Raw 264.7) cells and mice weight gain following in vivo infection. For group B, infection of Raw 264.7 cells resulted in higher titres for Mos-RRV isolates when compare to Mam-

RRV isolates. When the group B isolates were used in the mouse model, there was no significant difference in the weight gain of mice inoculated with either

Mam-RRV or Mos-RRV. In contrast, significant differences in symptomatic disease occurred at later time points with Mos-RRV isolate inoculation resulting in higher clinical scores compared to Mam-RRV isolates.

In Raw 264.7 cells, infection with Mos-RRV isolates from group C produced lower titres than Mam-RRV isolates. Additionally, mice inoculated with isolates from group C showed a significant difference in weight gain, with mice inoculated with Mam-RRV gaining more weight than mice inoculated with Mos-RRV. Clinical scores of mice inoculated with Mos-RRV were higher at later time points than mice inoculated with Mam-RRV.

In both groups B and C, Mos-RRV isolates appear to have an advantage in mammalian cells, which is of great importance, as this would assist in establishing of the initial infection. However, as Mam-RRV has a disadvantage in mammalian cells at early time points compared to Mos-RRV, thus following the initial infection, replication will be retarded. In Raw 264.7 cells infection, group B isolates resulted in Mos-RRV obtaining higher titres

265 than Mam-RRV indicating that in a functioning immune cell, Mam-RRV was at a disadvantage. This suggests that a more effective immune response is mounted against mammalian derived virus and thus after the initial round of replication, following infection by Mos-RRV, Mam-RRV is produced and restricted by the body. In contrast to group B, group C Mam-RRV isolates had an advantage in the host after the initial round of replication resulting in an increase in disease pathology in the mouse model. This would support the idea that a higher titre in Raw 264.7 cells by Mam-RRV increases disease pathology.

Group D Mam-RRV isolates produced significantly higher titres in Vero cells for the majority of time points, whilst in C6/36 cells Mos-RRV produced significantly higher titres for the majority of the time points. In Raw 264.7 cells, Mos-RRV isolates produced significantly lower titres compared to Mam-

RRV isolates. In vivo, group D isolates resulted in a significant difference between Mam-RRV and Mos-RRV inoculated mice for both weight gain and clinical score. This group displayed the most symptoms of disease pathology, both in weight gain and in clinical scores. Not only does Mam-RRV have an advantage in Raw 264.7 cells but there is also evidence that the virus can adapt quickly from one host to the other. This is shown by the fact that Mam-

RRV titres are higher in mammalian cells (Vero) and Mos-RRV titres are higher in mosquito cells (C6/36). This is likely due to genetic variation of these isolates and not just the glycosylation, as the latter would affect all isolates in a similar way. Therefore, these results suggest that both, glycosylation and genetic factors play an important role in disease development as well as in

266 speed of adaptation to alternate hosts. These factors can act together to increase disease pathology, giving the virus an advantage in infection of the host that would otherwise take several rounds of replication to achieve.

Interestingly, RRV-T48 did not fit into any of the groups defined in Table

8.1; however, it shared characteristics with group A and D. Mam-RRV-T48 produced higher titres than Mos-RRV-T48 in Vero cell infection whilst Mos-

RRV-T48 reached higher titres than Mam-RRV-T48 in C6/36 cells.

Additionally, infection of Raw 264.7 with Mam-RRV-T48 produced higher viral titres than infection with Mos-RRV-T48. There were significant differences in both the weight gain and clinical score of mice inoculated with Mam-RRV-T48 and Mos-RRV-T48, with mice inoculated with Mos-RRV-T48 gaining less weight and achieving higher clinical scores. In contrast to the RRV isolate studies, although Mam-RRV-T48 had an advantage in mammalian cells such as Vero, it did not have an advantage in the Raw 264.7 cells. This may be due to cytokine involvement rather than just genetic or glycosylation differences.

Unexpectedly, high titres of Mos-RRV field isolates in Raw 264.7 cells correlated with a lower disease profile in vivo. However, this is not true for

RRV-T48 in which Mos-RRV-T48 reached higher titres compared to Mam-

RRV-T48 in Raw 264.7 cells, but also caused more severe disease in vivo.

Previous studies have reported differential infectivity of myeloid DCs, with evidence that this enhanced viral fitness was due to N-linked glycans that occurred after passage through mosquito cells. Additionally, genetic mutations were discounted as a single passage of Mos-RRV through

267 mammalian cells reverted the virus to its previous infection profile (Shabman et al., 2007; Shabman et al., 2008). Taken together, it was expected that Mos-

RRV isolates would also have higher titres in the macrophage cell line. The data in this thesis both supports and contradicts previous studies. As a result, it is hypothesised that while glycosylation plays an important role in fitness of RRV to infect immune cells, other factors are also at play. A likely candidate is mutations occurring either in the NsP1-2 domain or on E2.

Previous studies of RRV have shown that single mutations in these areas effect viral titres and disease pathology and it is likely this may be a contributing factor in the differences observed among the various isolates

(Cruz et al., 2010; Jupille et al., 2011; Stoermer Burrack et al., 2014). Future studies should involve the sequencing of the isolates to further elucidate any genetic differences that may contribute to the infectivity of the viral isolates and resulting disease pathology.

Inflammatory responses to RRV in vivo

At six hours post infection, the most notable result is the up-regulation of CCL2 expression in infected mice. This up-regulation was seen in quadriceps muscle from mice infected with either Mos-RRV-T48 or Mos-RRV-

R and in the ankle joint from mice infected with Mos-RRV-T48, Mos-RRV-M and Mos-RRV-R at 6 hours post infection. No CCL2 up-regulation was seen in mice infected with Mam-RRV for any of the isolates tested at this time point.

CCL2, also known as MCP-1, is expressed by osteoclasts, osteoblasts and monocytes. It recruits monocytes, other inflammatory and immune cells to the site of infection. CCL2 release has been linked with bone degradation in

268 several studies and has been specifically shown to impact ChikV and RRV pathogenesis (Chen et al., 2015; Guglielmotti et al., 2002; Rulli et al., 2009).

It has also been shown that CCL2 inhibition results in a reduction in inflammatory infiltrates and lower production of NF-kB and TNFα, contributing to a reduction in tissue destruction (Rulli et al., 2009). The expression of CCL2 at such early time points induced by Mos-RRV and not

Mam-RRV is likely an important factor in the development of the more severe disease seen in mice infected with Mos-RRV. By 12 hours post infection,

CCL2 had returned to normal levels in all mice, except for mice infected with

Mam-RRV-R, which experienced an up-regulation of CCL2 even at this time point.

At 12 hours post infection, changes were seen in the expression level of type I IFN; however, the results for both quadriceps and ankle were at times contradictory. While type I IFN expression was up-regulated in the quadriceps muscle for mice infected with Mos-RRV-T48, Mos-RRV-M and Mos-RRV-R, it was also up-regulated for mice infected with Mam-RRV-R. This appears to contradict previous reports which have shown that Mam-RRV is a potent inducer of type I IFN in tissue culture while Mos-RRV fails to induce its expression (Shabman et al., 2007). Mam-RRV-T48 infection induced type I

IFN up-regulation in the ankle at 12 hours post infection, which was in line with expectations. However, among all the isolates, only Mos-RRV-M was found to induce up-regulation at this time point. It is possible that the up- regulation of type I IFN’s at this early time point contributes to disease pathology. An up-regulation of class I MHC molecules and recruitment of

269 macrophages in conjunction with cytotoxic cells would result in a high level of apoptosis and inflammatory infiltrates. It is possible that this unbalanced immune response causes tissue degradation and pathology. This is supported by, the highly elevated expression levels of IL4 in mice infected with

Mam-RRV-R, Mos-RRV-M and Mos-RRV-R in quadriceps and the downregulation of IL4 expression in mice infected with Mos-RRV-T48 in ankle at the same time. IL4 has a role in reduction of inflammation and is induced by

Mam-RRV-R 1.5 fold higher than Mos-RRV-R. This higher level of expression would offset the induction of type I IFN to some extent in mice infected with

Mam-RRV-R compared to those infected with Mos-RRV-R. Additionally, as type I IFN is not up-regulated by Mam-RRV-M, neither is IL4 as there is no need to combat the inflammatory TH1 process. This relationship is not mirrored in the ankle joint where IL4 expression levels are similar to PBS or below.

IL10 is upregulated by Mos-RRV-T48 and Mos-RRV-R in the ankle but not by Mos-RRV-M. Even though IL10 is an anti-inflammatory cytokine, high expression levels may not indicate a resolution of pathology but rather an increase in cytotoxic cells, indicating that the inflammatory process is already underway. In support of this is the finding that by 24 hours post infection in the ankle, but not quadriceps, all RRV conditions tested in mice result in significant up-regulation of expression of IL10, indicating that there is a response to the inflammatory processes occurring in ankle tissue.

Interestingly, CCL2 is again significantly up-regulated in quadriceps muscles for mice infected with Mos-RRV-M, Mam-RRV-R, Mos-RRV-R and Mos-RRV-

270

T48, indicating a second influx of inflammatory infiltrates which would contribute to inflammatory pathogenesis. Additionally, IL6 is upregulated by

Mam-RRV-T48, Mos-RRV-T48 and Mam-RRV-M in quadriceps muscle. Levels of IL6 expression were found to be 2.5 fold higher than that seen for mice infected with PBS, Mam-RRV-R and Mos-RRV-R, but this was not found to be statistically significant. The high levels of IL6 are indicative of the acute phase response and innate immune response to pathogens which includes fever, tiredness and loss of appetite. In this scenario, IL6 would be working to induce inflammation in an effort to remove the invading virus in the quadriceps muscle. The gene expression of IL6 in the ankle was not significantly upregulated, however, similar to the findings in quadriceps muscle, they were found to be 2.5 fold higher than PBS controls for mice infected with Mam-RRV-M, Mam-RRV-R and Mos-RRV-R.

At 24 hours post infection type I IFN gene expression was only significantly up-regulated in mice infected with Mam-RRV-M in the ankle, however by peak disease, IFNβ but not IFNα gene expression was upregulated in all Mam-RRV inoculated mouse groups. The high levels of IFNβ induction at this stage of disease in all Mam-RRV infected mice correlate with previous findings in tissue culture that that Mam-RRV but not Mos-RRV is a potent induced of IFNβ and that the induction of IFNβ plays a role in reducing disease in pathology (Lidbury et al., 2011; Shabman et al., 2007). In addition, recent studies in a tissue culture system have suggested that a delayed IFNβ induction is associated with an enhanced susceptibility to viral infection and replication (Chen et al., 2014a). It is possible that if another time point was

271 taken past peak disease, we would see an associated rise in IFNβ and this would be something to consider for future work.

At peak disease IFNγ was upregulated between 7 and 65 fold across all groups analysed in quadriceps and ankle joint. IFNγ increases osteoclast apoptosis, activates macrophages and other immune cells; promotes TH1 differentiation and is associated with auto-inflammatory disease. An increase in IFNγ expression is in line with previous studies, highlighting this as an important factor in pathogenesis of alphaviral disease (Vollmer-Conna et al.,

2008; Way et al., 2002). Supporting this data is the finding that all samples that were examined and had significantly up-regulated levels of IFNγ also had significantly up-regulated gene expression of IL12, which is known to up- regulate the expression of IFNγ and induce a TH1 response.

At peak disease, IL6, which has been previously linked with RRV disease pathology, was up-regulated in most instances (Chen et al., 2014b).

IL6 was significantly upregulated by Mam-RRV-T48 in both quadriceps muscle and ankle joint and by Mam-RRV-M in ankle samples at peak disease.

However, even though not significant, IL6 was up-regulated in all instances in ankle by over 6 fold and in quadriceps in all samples by 2 fold except for

RRV-M inoculated mice samples. As previously discussed in section 5.1, IL6 is associated with an inflammatory response and increased disease pathology in RRV infection. Surprisingly, TNFα was only found to be significantly up- regulated in Mos-RRV-M and Mam-RRV-R infected mice at peak disease. This result is unexpected, as several studies have shown TNFα to be upregulated in RRV infection (Chen et al., 2014a; Way et al., 2002). However, this

272 discrepancy may have been due to missing the peak up-regulation of gene expression in the mouse model as at peak disease only one time point was chosen across a 24 hour interval. TNFα is associated with apoptotic cell death, inflammation and tissue destruction and so it would be appropriate for this cytokine to be upregulated. Future studies may include time points taken across the period of peak disease to further elucidate events occurring around that period.

Conclusions and future directions

In conclusion, this study has directly compared the effect of mammalian and mosquito derived RRV on infection and replication kinetics in tissue culture and in a mouse model of disease. The results show that a clear difference in infectivity, kinetics and immune response occurs between mammalian and mosquito derived RRV. Additionally, the mouse model of

RRV has been refined, as this thesis has provided evidence of a sexual dimorphic response to RRV infection. It has also been shown that whilst effects such as glycosylation, covered in previous studies, obviously plays a role in the infectivity and pathogenesis of the virus, it is likely that glycosylation and genetic factors are interlinked. Hence, both are required to improve or reduce viral fitness, as shown by the variety of responses obtained using RRV field isolates. Future studies should expand on this work by examining the genetic sequences of the field isolates used and combining this information to further elucidate the different roles of genetic mutation and cell line of propagation. Additionally it was shown that pilot studies in other related and unrelated viruses may share some of the same dimorphic patterns

273 between mammalian and mosquito derived virus. It would be of great importance for future studies to expand on this study and further define common characteristics of arboviruses.

This thesis has contributed to the knowledge of how arboviruses establish and induce pathogenic disease and has created pathways for future studies that will add valuable knowledge to the field of mosquito borne disease.

274

Place a graph here: Place a graph here: Appendix Double-click, 1 or– Mammalian Double-click, and or  Drag a graph from  Drag a graph from mosquito the navigator derived Ross theriver navigator virus field isolate multistep kinetics in Vero cells

RRV field isolate A RRV field isolate B 14 14 * ) ) *

L 12 L 12

m m

/ 10 / 10

u u

f f

p p

8 8 *

g g

o o

l 6 l 6

( (

e e

r 4 r 4

t t

i i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate C RRV field isolate D 14 14 * ) * )

L 12 L 12

m m

/ 10 * / 10

u u

f f

p p

8 8

g g

o o

l 6 l 6

( (

e e

r 4 r 4

t t

i * i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

275

RRV field isolate E RRV field isolate F 14 14

) * )

L 12 L 12 *

m m / 10 * / 10

u u

f f

p p

8 8 g g *

o o l 6 l 6

( ( *

e e

r 4 r 4

t t i * * i T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate G RRV field isolate H 14 14

) * ) L 12 L 12 *

m m

/ 10 * / 10

u u

f f

p p

8 8

g g

o o l 6 l 6

( (

e e

r 4 r 4

t t i * i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate I RRV field isolate J 14 * 14 *

) )

L 12 L 12

m m

/ 10 / 10

u u

f f

p p

8 8 g * g

o o l 6 l 6

( (

e e

r 4 r 4

t t

i i

T 2 T 2 * 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

276

RRV field isolate K RRV field isolate L 14 14

) * )

L 12 L 12 m m * / 10 / 10

u u

f f

p p

8 8

g g

o o l 6 l 6

( (

e e

r 4 r 4

t t i * * i T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate M RRV field isolate N 14 14 * ) * )

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8

g g

o o l 6 l 6

( (

e e

r 4 r 4

t t

i i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate O RRV field isolate P 14 14

) ) * L 12 L 12 * *

m m / 10 / 10

u u

f f

p p

8 8

g g

o o l 6 * l 6 *

( (

e e

r 4 r 4

t t i i * *

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

277

RRV field isolate Q RRV field isolate R 14 * 14

) ) *

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8

g * g

o o

l 6 l 6

( (

e e

r 4 r 4

t t

i i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

1 4 Appendix 1 - Mammalian and mosquito derived Ross River field isolate kinetics in Vero cells C 6 /3 6 R 1 2 1 4 Mammalian* derived ( ) and mosquito derived ( ) RRV field isolates were used to infect Vero * V e ro R C 6 /3 6 R 1 2 1 0 cells at 0.001 MOI. Samples were taken* at 2, 4,6,12,V e r o24 R and 48 hours post infection and

) *

) g

g 1 0 titrated on Vero cells to determine viral titre (log pfu/mL). Each data point represents the

o o

l 8

( ( mean of three separate experiments and error is calculated as SEM. Statistical significance

8 L

6 * L (*, p<0.05) was determined by two way ANOVA with Tukey’s test.

m m

/ 6 *

u u f * 4 f

p 4 * p

2 * 2 *

0 0 0 1 0 2 0 03 0 1 04 0 2 05 0 3 0 4 0 5 0 H o u r s p o s t i n f e c ti o n H o u r s p o s t i n f e c ti o n

278

Place a graph here: Place a graph here: Appendix Double-click, 2 or– Mammalian Double-click, and or  Drag a graph from  Drag a graph from mosquito the navigator derived Ross theriver navigator virus field isolate multistep kinetics in C6/36 cells

RRV field isolate A RRV field isolate B * 14 * 14

) )

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8

g g

o o

l 6 l 6

( (

e e

r 4 r 4

t t *

i i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate C RRV field isolate D 14 * 14 *

) )

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8 *

g g

o o

l 6 l 6

( ( *

e e

r 4 r 4

t * t

i i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

279

RRV field isolate E RRV field isolate F * 14 * 14

) )

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8

g g

o o l 6 l 6

( (

e e

r 4 r 4

t t

i * i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate G RRV field isolate H 14 14 *

) ) *

L 12 * L 12

m m / 10 / 10

u u

f f

p p

8 8

g g

o o l 6 l 6

( (

e e

r 4 r 4

t * t i i * T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate I RRV field isolate J 14 14 * *

) )

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8

g g

o o l 6 l 6

( (

e e

r 4 r 4

i i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

280

RRV field isolate K RRV field isolate L 14 14 *

) * )

L L 12 12

m m

/ / 10 10

p 8 8 *

g * g

l l 6 6

(

e e

r r 4 4

t t i * i *

T T 2 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate M RRV field isolate N * 14 14 *

) )

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8

g g

o o l 6 l 6

( (

e e

r 4 r 4

t t

i i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate O RRV field isolate P 14 14 * *

) )

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8

g g

o o l 6 l 6

( (

e e

r 4 r 4

t t

i i

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

281

RRV field isolate Q RRV field isolate R 14 * 14

) ) *

L 12 L 12

m m / 10 / 10

u u

f f

p p

8 8

g g

o o *

l 6 l 6

( (

e e

r 4 r 4

t t

i i *

T 2 T 2 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

1 4 Appendix 1 - Mammalian and mosquito derived Ross River field isolate kinetics in C6/36 cells C 6 /3 6 R 1 2 1 4 Mammalian* derived ( ) and mosquito derived ( ) RRV field isolates were used to infect * V e ro R C 6 /3 6 R 1 2 1 0 C6/36 cells at 0.001 MOI. Samples were* taken at 2, 4,6,12,V e ro R 24 and 48 hours post infection

) *

) g

g 1 0 and titrated on Vero cells to determine viral titre (log pfu/mL). Each data point represents

o o

l 8

( ( the mean of three separate experiments and error is calculated as SEM. Statistical

8 L

6 * L significance (*, p<0.05) was determined by two way ANOVA with Tukey’s test.

m m

/ 6 *

u u f * 4 f

p 4 * p

2 * 2 *

0 0 0 1 0 2 0 30 0 1400 2500 3 0 4 0 5 0 H o u r s p o s t i n f e c ti o n H o u r s p o s t i n f e c ti o n

282

Place a graph here: Place a graph here: Appendix Double-click, 3 or– Mammalian Double-click, and or  Drag a graph from  Drag a graph from mosquito the navigator derived Ross theriver navigator virus field isolate multistep kinetics in Raw 264.7 cells

RRV field isolate A RRV field isolate B 6 6 *

) ) L 5 L 5 * *

m m

/ /

u 4 u 4

f f

p p

g 3 g 3

o o

l l

( ( 2 2

e e

r r

t t i 1 i 1

T T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate C RRV field isolate D 6 * 6

) ) L 5 L 5 * m m *

/ /

u 4 u 4

f f

p p

g 3 g 3

o o

l l

( ( 2 2

e e

r r

t t i 1 i 1

T T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

283

RRV field isolate E RRV field isolate F * 6 6

) ) L 5 L 5 *

m m *

/ u 4 u 4

p p

g 3 g 3

l l

( (

2 2

e e

i i 1 1

T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate G RRV field isolate H 6 6

) )

L 5 L 5

m m * / * / u 4 u 4 *

f f

p p

g 3 g 3 *

o o

l l

( ( 2 2

e e

r r

t t i 1 i 1

T T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate I RRV field isolate J 6 6

) )

L 5 L 5

m m

/ / u 4 * u 4 f f *

p p

g 3 g 3

o o

l l

( ( 2 2

e e

r r

t t i 1 i 1

T T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

284

RRV field isolate K RRV field isolate L 6 6

)

L L 5 5 *

/ / *

u u 4 4

g 3 g 3

(

2 2

i 1 i 1

T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate M RRV field isolate N 6 6

)

L 5 L 5

/ * /

u 4 u 4

g 3 g 3

o o *

( ( *

2 2

i 1 i 1

T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

RRV field isolate O RRV field isolate P 6 * 6

)

L 5 L 5

u 4 u 4

f f *

g 3 g 3 * *

(

2 2

i 1 i 1

T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

285

RRV field isolate Q RRV field isolate R 6 6 *

) )

L 5 L 5

m m

/ * /

u 4 u 4

f f

p p

g g 3 3

o o

l l

( (

2 2

r r

i i 1 1

T 0 0 0 10 20 30 40 50 0 10 20 30 40 50 Hours post infection Hours post infection

1 4 Appendix 3 – Mammalian and mosquito derived Ross River field isolate kinetics in Raw 264.7 cells C 6 /3 6 R 1 2 1 4 Mammalian* derived ( ) and mosquito derived ( ) RRV field isolates were used to infect Vero * V e ro R C 6 /3 6 R 1 2 1 0 cells at 0.001 MOI. Samples were taken* at 2, 4,6,12,V e r o24 R and 48 hours post infection and

) *

) g

g 1 0 titrated on Vero cells to determine viral titre (log pfu/mL). Each data point represents the

o o

l 8

( ( mean of three separate experiments and error is calculated as SEM. Statistical significance

8 L

6 * L (*, p<0.05) was determined by two way ANOVA with Tukey’s test.

m m

/ 6 *

u u f * 4 f

p 4 * p

2 * 2 *

0 0 0 1 0 2 0 03 0 1 04 0 2 05 0 3 0 4 0 5 0 H o u r s p o s t i n f e c ti o n H o u r s p o s t i n f e c ti o n

286

References

Aaskov, J.G., Dalglish, D.A., Harper, J.J., Douglas, J.F., Donaldson, M.D., Hertzog, P.J., 1987. Natural killer cells in viral arthritis. Clinical and experimental immunology 68, 23-32. Aguirre, S., Maestre, A.M., Pagni, S., Patel, J.R., Savage, T., Gutman, D., Maringer, K., Bernal-Rubio, D., Shabman, R.S., Simon, V., Rodriguez-Madoz, J.R., Mulder, L.C., Barber, G.N., Fernandez-Sesma, A., 2012. DENV inhibits type I IFN production in infected cells by cleaving human STING. PLoS pathogens 8, e1002934. Ahmad, A., Alvarez, F., 2004. Role of NK and NKT cells in the immunopathogenesis of HCV-induced hepatitis. Journal of leukocyte biology 76, 743-759. Aichinger, G., Ehrlich, H.J., Aaskov, J.G., Fritsch, S., Thomasser, C., Draxler, W., Wolzt, M., Muller, M., Pinl, F., Van Damme, P., Hens, A., Levy, J., Portsmouth, D., Holzer, G., Kistner, O., Kreil, T.R., Barrett, P.N., 2011. Safety and immunogenicity of an inactivated whole virus Vero cell-derived Ross River virus vaccine: a randomized trial. Vaccine 29, 9376-9384. Alejandria, M.M., 2004. Dengue Fever. Clinical Evidence 12, 1062-1071. Alen, M.M., Dallmeier, K., Balzarini, J., Neyts, J., Schols, D., 2012. Crucial role of the N-glycans on the viral E-envelope glycoprotein in DC-SIGN- mediated dengue virus infection. Antiviral research 96, 280-287. Angele, M.K., Chaudry, I.H., 2005. Surgical trauma and immunosuppression: pathophysiology and potential immunomodulatory approaches. Langenbeck's archives of surgery / Deutsche Gesellschaft fur Chirurgie 390, 333-341. Ank, N., Iversen, M.B., Bartholdy, C., Staeheli, P., Hartmann, R., Jensen, U.B., Dagnaes-Hansen, F., Thomsen, A.R., Chen, Z., Haugen, H., Klucher, K., Paludan, S.R., 2008. An important role for type III interferon (IFN-lambda/IL- 28) in TLR-induced antiviral activity. J Immunol 180, 2474-2485. Ank, N., West, H., Bartholdy, C., Eriksson, K., Thomsen, A.R., Paludan, S.R., 2006. Lambda interferon (IFN-lambda), a type III IFN, is induced by viruses and IFNs and displays potent antiviral activity against select virus infections in vivo. J Virol 80, 4501-4509. Antonelli, A., Fallahi, P., Delle Sedie, A., Ferrari, S.M., Maccheroni, M., Bombardieri, S., Riente, L., Ferrannini, E., 2009. High values of Th1 (CXCL10) and Th2 (CCL2) chemokines in patients with psoriatic arthtritis. Clinical and experimental rheumatology 27, 22-27. Arcari, P., Martinelli, R., Salvatore, F., 1984. The complete sequence of a full length cDNA for human liver glyceraldehyde-3-phosphate dehydrogenase: evidence for multiple mRNA species. Nucleic Acids Res 12, 9179-9189. Arjona, A., Ledizet, M., Anthony, K., Bonafe, N., Modis, Y., Town, T., Fikrig, E., 2007. West Nile virus envelope protein inhibits dsRNA-induced innate immune responses. J Immunol 179, 8403-8409. Artpradit, C., Robinson, L.N., Gavrilov, B.K., Rurak, T.T., Ruchirawat, M., Sasisekharan, R., 2013. Recognition of heparan sulfate by clinical strains of dengue virus serotype 1 using recombinant subviral particles. Virus Res 176, 69-77.

287

Assuncao-Miranda, I., Bozza, M.T., Da Poian, A.T., 2010. Pro-inflammatory response resulting from sindbis virus infection of human macrophages: implications for the pathogenesis of viral arthritis. Journal of medical virology 82, 164-174. Atkins, G.J., 2013. The Pathogenesis of Alphaviruses. ISRN Virology 2013, 22. Avirutnan, P., Hauhart, R.E., Marovich, M.A., Garred, P., Atkinson, J.P., Diamond, M.S., 2011. Complement-mediated neutralization of dengue virus requires mannose-binding lectin. mBio 2. Azeredo, E.L., De Oliveira-Pinto, L.M., Zagne, S.M., Cerqueira, D.I., Nogueira, R.M., Kubelka, C.F., 2006. NK cells, displaying early activation, cytotoxicity and adhesion molecules, are associated with mild dengue disease. Clinical and experimental immunology 143, 345-356. Baccala, R., Kono, D.H., Theofilopoulos, A.N., 2005. Interferons as pathogenic effectors in autoimmunity. Immunological reviews 204, 9-26. Bai, F., Town, T., Qian, F., Wang, P., Kamanaka, M., Connolly, T.M., Gate, D., Montgomery, R.R., Flavell, R.A., Fikrig, E., 2009. IL-10 signaling blockade controls murine West Nile virus infection. PLoS pathogens 5, e1000610. Barber, B., Denholm, J.T., Spelman, D., 2009. Ross River virus. Australian family physician 38, 586-589. Barkhordarian, A., Thames, A.D., Du, A.M., Jan, A.L., Nahcivan, M., Nguyen, M.T., Sama, N., Chiappelli, F., 2015. Viral immune surveillance: Toward a TH17/TH9 gate to the central nervous system. Bioinformation 11, 47-54. Barna, M., Komatsu, T., Bi, Z., Reiss, C.S., 1996. Sex differences in susceptibility to viral infection of the central nervous system. Journal of neuroimmunology 67, 31-39. Bartholome, E.J., Willems, F., Crusiaux, A., Thielemans, K., Schandene, L., Goldman, M., 1999. Interferon-beta inhibits Th1 responses at the dendritic cell level. Relevance to multiple sclerosis. Acta neurologica Belgica 99, 44-52. Basler, C.F., Garcia-Sastre, A., 2002. Viruses and the type I interferon antiviral system: induction and evasion. International Reviews of Immunology 21, 305-337. Bekisz, J., Schmeisser, H., Hernandez, J., Goldman, N.D., Zoon, K.C., 2004. Human interferons alpha, beta and omega. Growth Factors 22, 243-251. Bente, D.A., Melkus, M.W., Garcia, J.V., Rico-Hesse, R., 2005. Dengue fever in humanized NOD/SCID mice. J Virol 79, 13797-13799. Best, S.M., Morris, K.L., Shannon, J.G., Robertson, S.J., Mitzel, D.N., Park, G.S., Boer, E., Wolfinbarger, J.B., Bloom, M.E., 2005. Inhibition of interferon- stimulated JAK-STAT signaling by a tick-borne flavivirus and identification of NS5 as an interferon antagonist. J Virol 79, 12828-12839. Bonnick, S.L., 2006. Osteoporosis in men and women. Clinical cornerstone 8, 28-39. Boyd, A.M., Hall, R.A., Gemmell, R.T., Kay, B.H., 2001. Experimental infection of Australian brushtail possums, Trichosurus vulpecula (Phalangeridae: Marsupialia), with Ross River and Barmah Forest viruses by use of a natural mosquito vector system. Am J Trop Med Hyg 65, 777-782. Brackney, D.E., Scott, J.C., Sagawa, F., Woodward, J.E., Miller, N.A., Schilkey, F.D., Mudge, J., Wilusz, J., Olson, K.E., Blair, C.D., Ebel, G.D.,

288

2010. C6/36 Aedes albopictus cells have a dysfunctional antiviral RNA interference response. PLoS Negl Trop Dis 4, e856. Briant, L., Despres, P., Choumet, V., Misse, D., 2014. Role of skin immune cells on the host susceptibility to mosquito-borne viruses. Virology 464-465, 26-32. Byrnes, A.P., Griffin, D.E., 1998. Binding of Sindbis virus to cell surface heparan sulfate. J Virol 72, 7349-7356. Byrnes, A.P., Griffin, D.E., 2000. Large-plaque mutants of Sindbis virus show reduced binding to heparan sulfate, heightened viremia, and slower clearance from the circulation. J Virol 74, 644-651. Calisher, C.H., 2005. A very brief history of arbovirology, focusing on contributions by workers of the Rockefeller foundation. Vector Borne Zoonotic Dis 5, 202-211. Cantaert, T., Baeten, D., Tak, P.P., van Baarsen, L.G., 2010. Type I IFN and TNFalpha cross-regulation in immune-mediated inflammatory disease: basic concepts and clinical relevance. Arthritis research & therapy 12, 219. Carver, S., Bestall, A., Jardine, A., Ostfeld, R.S., 2009. Influence of hosts on the ecology of arboviral transmission: potential mechanisms influencing dengue, Murray Valley encephalitis, and Ross River virus in Australia. Vector Borne Zoonotic Dis 9, 51-64. CDC, 2005. Dengue Fever Fact Sheet. Center for Disease Control (CDC). CDC, 2011. La Crosse Encephalitis, in: diseases, D.o.V.-B. (Ed.). Centres for Disease Control and Prevention. Chaaitanya, I.K., Muruganandam, N., Sundaram, S.G., Kawalekar, O., Sugunan, A.P., Manimunda, S.P., Ghosal, S.R., Muthumani, K., Vijayachari, P., 2011. Role of proinflammatory cytokines and chemokines in chronic arthropathy in CHIKV infection. Viral immunology 24, 265-271. Chadha, K.C., Ambrus, J.L., Jr., Dembinski, W., Ambrus, J.L., Sr., 2004. Interferons and interferon inhibitory activity in disease and therapy. Experimental Biology and Medicine (Maywood, N.J) 229, 285-290. Chamberlain, R.W., 1982. Arbovirology--then and now. The American Journal of Tropical Medicine and Hygiene 31, 430-437. Chambers, T.J., Hahn, C.S., Galler, R., Rice, C.M., 1990. Flavivirus genome organization, expression, and replication. Annual Reviews of Microbiology 44, 649-688. Chan, J.C., Cheung, J.C., Stehouwer, C.D., Emeis, J.J., Tong, P.C., Ko, G.T., Yudkin, J.S., 2002. The central roles of obesity-associated dyslipidaemia, endothelial activation and cytokines in the Metabolic Syndrome--an analysis by structural equation modelling. International journal of obesity and related metabolic disorders : journal of the International Association for the Study of Obesity 26, 994-1008. Chandak, N.H., Kashyap, R.S., Kabra, D., Karandikar, P., Saha, S.S., Morey, S.H., Purohit, H.J., Taori, G.M., Daginawala, H.F., 2009. Neurological complications of Chikungunya virus infection. Neurology India 57, 177-180. Chaturvedi, U.C., Elbishbishi, E.A., Agarwal, R., Raghupathy, R., Nagar, R., Tandon, R., Pacsa, A.S., Younis, O.I., Azizieh, F., 1999a. Sequential production of cytokines by dengue virus-infected human peripheral blood leukocyte cultures. Journal of medical virology 59, 335-340.

289

Chaturvedi, U.C., Raghupathy, R., Pasca, A.S., Elbishbishi, E.A., Agarwal, R., Nagar, R., Misra, A., Kapoor, S., Mukerjee, R., Mathur, A., Khan, M.A.Y., Azizieh, F., 1999b. Shift from a Th1-type response to Th2-type in dengue haemorrhagic fever. Current Science 76, 63-69. Chen, H.C., Hofman, F.M., Kung, J.T., Lin, Y.D., Wu-Hsieh, B.A., 2007. Both virus and tumor necrosis factor alpha are critical for endothelium damage in a mouse model of dengue virus-induced hemorrhage. J Virol 81, 5518-5526. Chen, R.F., Liu, J.W., Yeh, W.T., Wang, L., Chang, J.C., Yu, H.R., Cheng, J.T., Yang, K.D., 2005. Altered T helper 1 reaction but not increase of virus load in patients with dengue hemorrhagic fever. FEMS Immunology and Medical Microbiology 44, 43-50. Chen, W., Foo, S.S., Li, R.W., Smith, P.N., Mahalingam, S., 2014a. Osteoblasts from osteoarthritis patients show enhanced susceptibility to Ross River virus infection associated with delayed type I interferon responses. Virol J 11. Chen, W., Foo, S.S., Rulli, N.E., Taylor, A., Sheng, K.C., Herrero, L.J., Herring, B.L., Lidbury, B.A., Li, R.W., Walsh, N.C., Sims, N.A., Smith, P.N., Mahalingam, S., 2014b. Arthritogenic alphaviral infection perturbs osteoblast function and triggers pathologic bone loss. Proc Natl Acad Sci U S A 111, 6040-6045. Chen, W., Foo, S.S., Taylor, A., Lulla, A., Merits, A., Hueston, L., Forwood, M.R., Walsh, N.C., Sims, N.A., Herrero, L.J., Mahalingam, S., 2015. Bindarit, an inhibitor of monocyte chemotactic protein synthesis, protects against bone loss induced by chikungunya virus infection. J Virol 89, 581-593. Chen, Y.C., Wang, S.Y., 2002. Activation of terminally differentiated human monocytes/macrophages by dengue virus: productive infection, hierarchical production of innate cytokines and chemokines, and the synergistic effect of lipopolysaccharide. Journal of Virology 76, 9877-9887. Chen, Z., Liu, L.M., Gao, N., Xu, X.F., Zhang, J.L., Wang, J.L., An, J., 2009. Passive protection assay of monoclonal antibodies against dengue virus in suckling mice. Current microbiology 58, 326-331. Cheng, R.H., Kuhn, R.J., Olson, N.H., Rossmann, M.G., Choi, H.K., Smith, T.J., Baker, T.S., 1995. Nucleocapsid and glycoprotein organization in an enveloped virus. Cell 80, 621-630. Chinetti-Gbaguidi, G., Staels, B., 2011. Macrophage polarization in metabolic disorders: functions and regulation. Current opinion in lipidology 22, 365- 372. Choudhry, M.A., Bland, K.I., Chaudry, I.H., 2006. Gender and susceptibility to sepsis following trauma. Endocrine, metabolic & immune disorders drug targets 6, 127-135. Chu, J.J., Leong, P.W., Ng, M.L., 2006. Analysis of the endocytic pathway mediating the infectious entry of mosquito-borne flavivirus West Nile into Aedes albopictus mosquito (C6/36) cells. Virology 349, 463-475. Ciota, A.T., Kramer, L.D., 2010. Insights into arbovirus evolution and adaptation from experimental studies. Viruses 2, 2594-2617. Conway, M.J., Watson, A.M., Colpitts, T.M., Dragovic, S.M., Li, Z., Wang, P., Feitosa, F., Shepherd, D.T., Ryman, K.D., Klimstra, W.B., Anderson, J.F., Fikrig, E., 2014. Mosquito saliva serine protease enhances dissemination of dengue virus into the mammalian host. J Virol 88, 164-175.

290

Cruz, C.C., Suthar, M.S., Montgomery, S.A., Shabman, R., Simmons, J., Johnston, R.E., Morrison, T.E., Heise, M.T., 2010. Modulation of type I IFN induction by a virulence determinant within the alphavirus nsP1 protein. Virology 399, 1-10. Curiel, R.E., Miller, M.H., Ishikawa, R., Thomas, D.C., Bigley, N.J., 1993. Does the gender difference in interferon production seen in picornavirus- infected spleen cell cultures from ICR Swiss mice have any in vivo significance? Journal of interferon research 13, 387-395. Dalbeth, N., Callan, M.F., 2002. A subset of natural killer cells is greatly expanded within inflamed joints. Arthritis and rheumatism 46, 1763-1772. Dalrymple, N., Mackow, E.R., 2011. Productive dengue virus infection of human endothelial cells is directed by heparan sulfate-containing proteoglycan receptors. J Virol 85, 9478-9485. Davis, C.W., Nguyen, H.Y., Hanna, S.L., Sanchez, M.D., Doms, R.W., Pierson, T.C., 2006. West Nile virus discriminates between DC-SIGN and DC-SIGNR for cellular attachment and infection. Journal of Virology 80, 1290-1301. de Curtis, I., Simons, K., 1988. Dissection of Semliki Forest virus glycoprotein delivery from the trans-Golgi network to the cell surface in permeabilized BHK cells. Proc Natl Acad Sci U S A 85, 8052-8056. de Vries, R.P., Smit, C.H., de Bruin, E., Rigter, A., de Vries, E., Cornelissen, L.A., Eggink, D., Chung, N.P., Moore, J.P., Sanders, R.W., Hokke, C.H., Koopmans, M., Rottier, P.J., de Haan, C.A., 2012. Glycan-dependent immunogenicity of recombinant soluble trimeric hemagglutinin. J Virol 86, 11735-11744. Devito, A., 2014. Interferon gamma -induced chemokines in psoriatic arthritis. La Clinica terapeutica 165, e442-446. Di Rosa, F., Pabst, R., 2005. The bone marrow: a nest for migratory memory T cells. Trends in immunology 26, 360-366. Diamond, M.S., Roberts, T.G., Edgil, D., Lu, B., Ernst, J., Harris, E., 2000. Modulation of Dengue virus infection in human cells by alpha, beta, and gamma interferons. Journal of Virology 74, 4957-4966. Doherty, R.L., Standfast, H.A., Domrow, R., Wetters, E.J., Whitehead, R.H., Carley, J.G., 1971. Studies of the epidemiology of arthropod-borne virus infections at Mitchell River Mission, Cape York Peninsula, North Queensland. IV. Arbovirus infections of mosquitoes and mammals, 1967-1969. Trans R Soc Trop Med Hyg 65, 504-513. Doherty, R.L., Whitehead, R. H., Gorman, B. M., 1963. The isolation of a third group A arbovirus in Australia, with preliminary observations on its relationship to epidemic polyarthritis. The Australian Journal of Science 26, 183-184. Dubuisson, J., Lustig, S., Ruggli, N., Akov, Y., Rice, C.M., 1997. Genetic determinants of Sindbis virus neuroinvasiveness. J Virol 71, 2636-2646. Ehlers, M.R., 2000. CR3: a general purpose adhesion-recognition receptor essential for innate immunity. Microbes and infection / Institut Pasteur 2, 289-294. Ellermann-Eriksen, S., Liberto, M.C., Iannello, D., Mogensen, S.C., 1986. X- linkage of the early in vitro alpha/beta interferon response of mouse peritoneal macrophages to herpes simplex virus type 2. J Gen Virol 67 ( Pt 6), 1025-1033.

291

Fayzulin, R., Frolov, I., 2004. Changes of the secondary structure of the 5' end of the Sindbis virus genome inhibit virus growth in mosquito cells and lead to accumulation of adaptive mutations. J Virol 78, 4953-4964. Feinberg, H., Mitchell, D.A., Drickamer, K., Weis, W.I., 2001. Structural basis for selective recognition of oligosaccharides by DC-SIGN and DC-SIGNR. Science 294, 2163-2166. Firth, A.E., Chung, B.Y., Fleeton, M.N., Atkins, J.F., 2008. Discovery of frameshifting in Alphavirus 6K resolves a 20-year enigma. Virol J 5, 108. Forsell, K., Xing, L., Kozlovska, T., Cheng, R.H., Garoff, H., 2000. Membrane proteins organize a symmetrical virus. The EMBO journal 19, 5081-5091. Foy, N.J., Akhrymuk, M., Akhrymuk, I., Atasheva, S., Bopda-Waffo, A., Frolov, I., Frolova, E.I., 2013a. Hypervariable domains of nsP3 proteins of New World and Old World alphaviruses mediate formation of distinct, virus-specific protein complexes. J Virol 87, 1997-2010. Foy, N.J., Akhrymuk, M., Shustov, A.V., Frolova, E.I., Frolov, I., 2013b. Hypervariable Domain of Nonstructural Protein nsP3 of Venezuelan Equine Encephalitis Virus Determines Cell-Specific Mode of Virus Replication. J Virol 87, 7569-7584. Fragkoudis, R., Breakwell, L., McKimmie, C., Boyd, A., Barry, G., Kohl, A., Merits, A., Fazakerley, J.K., 2007. The type I interferon system protects mice from Semliki Forest virus by preventing widespread virus dissemination in extraneural tissues, but does not mediate the restricted replication of avirulent virus in central nervous system neurons. J Gen Virol 88, 3373- 3384. Fredericksen, B.L., Gale, M., Jr., 2006. West Nile virus evades activation of interferon regulatory factor 3 through RIG-I-dependent and -independent pathways without antagonizing host defense signaling. J Virol 80, 2913-2923. Frolov, I., Akhrymuk, M., Akhrymuk, I., Atasheva, S., Frolova, E.I., 2012. Early events in alphavirus replication determine the outcome of infection. J Virol 86, 5055-5066. Fros, J.J., Domeradzka, N.E., Baggen, J., Geertsema, C., Flipse, J., Vlak, J.M., Pijlman, G.P., 2012. Chikungunya virus nsP3 blocks stress granule assembly by recruitment of G3BP into cytoplasmic foci. J Virol 86, 10873- 10879. Fuchs, A., Lin, T.Y., Beasley, D.W., Stover, C.M., Schwaeble, W.J., Pierson, T.C., Diamond, M.S., 2010. Direct complement restriction of flavivirus infection requires glycan recognition by mannose-binding lectin. Cell host & microbe 8, 186-195. Gardner, C.L., Burke, C.W., Tesfay, M.Z., Glass, P.J., Klimstra, W.B., Ryman, K.D., 2008. Eastern and Venezuelan equine encephalitis viruses differ in their ability to infect dendritic cells and macrophages: impact of altered cell tropism on pathogenesis. J Virol 82, 10634-10646. Gardner, C.L., Ebel, G.D., Ryman, K.D., Klimstra, W.B., 2011. Heparan sulfate binding by natural eastern equine encephalitis viruses promotes neurovirulence. Proc Natl Acad Sci U S A 108, 16026-16031. Gardner, C.L., Yin, J., Burke, C.W., Klimstra, W.B., Ryman, K.D., 2009. Type I interferon induction is correlated with attenuation of a South American eastern equine encephalitis virus strain in mice. Virology 390, 338-347.

292

Geijtenbeek, T.B., Torensma, R., van Vliet, S.J., van Duijnhoven, G.C., Adema, G.J., van Kooyk, Y., Figdor, C.G., 2000. Identification of DC-SIGN, a novel dendritic cell-specific ICAM-3 receptor that supports primary immune responses. Cell 100, 575-585. Geurs, T.L., Hill, E.B., Lippold, D.M., French, A.R., 2012. Sex differences in murine susceptibility to systemic viral infections. Journal of autoimmunity 38, J245-253. Gibbons, D.L., Erk, I., Reilly, B., Navaza, J., Kielian, M., Rey, F.A., Lepault, J., 2003. Visualization of the target-membrane-inserted fusion protein of Semliki Forest virus by combined electron microscopy and crystallography. Cell 114, 573-583. Godfrey, T.E., Kim, S.H., Chavira, M., Ruff, D.W., Warren, R.S., Gray, J.W., Jensen, R.H., 2000. Quantitative mRNA expression analysis from formalin- fixed, paraffin-embedded tissues using 5' nuclease quantitative reverse transcription-polymerase chain reaction. The Journal of Molecular Diagnostics 2, 84-91. Goldfeld, A.E., Maniatis, T., 1989. Coordinate viral induction of tumor necrosis factor alpha and interferon beta in human B cells and monocytes. Proc Natl Acad Sci U S A 86, 1490-1494. Goodbourn, S., Didcock, L., Randall, R.E., 2000. Interferons: cell signalling, immune modulation, antiviral response and virus countermeasures. Journal of General Virology 81, 2341-2364. Gordon, S., Martinez, F.O., 2010. Alternative activation of macrophages: mechanism and functions. Immunity 32, 593-604. Green, M.R., Kennell, A.S., Larche, M.J., Seifert, M.H., Isenberg, D.A., Salaman, M.R., 2005. Natural killer cell activity in families of patients with systemic lupus erythematosus: demonstration of a killing defect in patients. Clinical and experimental immunology 141, 165-173. Greene, I.P., Wang, E., Deardorff, E.R., Milleron, R., Domingo, E., Weaver, S.C., 2005. Effect of Alternating Passage on Adaptation of Sindbis Virus to Vertebrate and Invertebrate Cells. J Virol 79, 14253-14260. Grieder, F.B., Davis, B.K., Zhou, X.D., Chen, S.J., Finkelman, F.D., Gause, W.C., 1997. Kinetics of cytokine expression and regulation of host protection following infection with molecularly cloned Venezuelan equine encephalitis virus. Virology 233, 302-312. Groat-Carmona, A.M., Orozco, S., Friebe, P., Payne, A., Kramer, L., Harris, E., 2012. A novel coding-region RNA element modulates infectious dengue virus particle production in both mammalian and mosquito cells and regulates viral replication in Aedes aegypti mosquitoes. Virology 432, 511-526. Gu, L., Tseng, S., Horner, R.M., Tam, C., Loda, M., Rollins, B.J., 2000. Control of TH2 polarization by the chemokine monocyte chemoattractant protein-1. Nature 404, 407-411. Guglielmotti, A., D'Onofrio, E., Coletta, I., Aquilini, L., Milanese, C., Pinza, M., 2002. Amelioration of rat adjuvant arthritis by therapeutic treatment with bindarit, an inhibitor of MCP-1 and TNF-alpha production. Inflammation research : official journal of the European Histamine Research Society ... [et al.] 51, 252-258. Gunn, B.M., Morrison, T.E., Whitmore, A.C., Blevins, L.K., Hueston, L., Fraser, R.J., Herrero, L.J., Ramirez, R., Smith, P.N., Mahalingam, S., Heise,

293

M.T., 2012. Mannose binding lectin is required for alphavirus-induced arthritis/myositis. PLoS pathogens 8, e1002586. Guo, J., Jou, W., Gavrilova, O., Hall, K.D., 2009. Persistent diet-induced obesity in male C57BL/6 mice resulting from temporary obesigenic diets. PLoS One 4, e5370. Guu, T.S., Zheng, W., Tao, Y.J., 2012. Bunyavirus: structure and replication. Adv Exp Med Biol 726, 245-266. Hacker, K., White, L., de Silva, A.M., 2009. N-linked glycans on dengue viruses grown in mammalian and insect cells. J Gen Virol 90, 2097-2106. Halstead, S.B., 2002. Dengue. Current Opinion in Infectious Diseases 15, 471-476. Hanley, K.A., Manlucu, L.R., Gilmore, L.E., Blaney, J.E., Jr., Hanson, C.T., Murphy, B.R., Whitehead, S.S., 2003. A trade-off in replication in mosquito versus mammalian systems conferred by a point mutation in the NS4B protein of dengue virus type 4. Virology 312, 222-232. Hardy, W.R., Strauss, J.H., 1989. Processing the nonstructural polyproteins of sindbis virus: nonstructural proteinase is in the C-terminal half of nsP2 and functions both in cis and in trans. J Virol 63, 4653-4664. Harley, D., Sleigh, A., Ritchie, S., 2001. Ross River virus transmission, infection, and disease: a cross-disciplinary review. Clin Microbiol Rev 14, 909- 932, table of contents. Health, D.o., 2010. Table of arboviruses of importance to Australia, https://www.health.gov.au/internet/main/publishing.nsf/Content/health- arbovirus-important.htm. Health, D.o., 2015. Notifications of a selected diseases by State and Territiory and Year, National Notifiable Diseases Survellance System. Australian Government. Heil, M.L., Albee, A., Strauss, J.H., Kuhn, R.J., 2001. An amino acid substitution in the coding region of the E2 glycoprotein adapts Ross River virus to utilize heparan sulfate as an attachment moiety. J Virol 75, 6303- 6309. Hernandez, R., Brown, D.T., Paredes, A., 2014. Structural differences observed in arboviruses of the alphavirus and flavivirus genera. Advances in virology 2014, 259382. Hernandez, R., Ferreira, D., Sinodis, C., Litton, K., Brown, D.T., 2005. Single amino acid insertions at the junction of the sindbis virus E2 transmembrane domain and endodomain disrupt virus envelopment and alter infectivity. J Virol 79, 7682-7697. Hernandez, R., Luo, T., Brown, D.T., 2001. Exposure to low pH is not required for penetration of mosquito cells by Sindbis virus. J Virol 75, 2010-2013. Herrero, L.J., Lidbury, B.A., Bettadapura, J., Jian, P., Herring, B.L., Hey- Cunningham, W.J., Sheng, K.C., Zakhary, A., Mahalingam, S., 2014. Characterization of Barmah Forest virus pathogenesis in a mouse model. J Gen Virol 95, 2146-2154. Herrero, L.J., Nelson, M., Srikiatkhachorn, A., Gu, R., Anantapreecha, S., Fingerle-Rowson, G., Bucala, R., Morand, E., Santos, L.L., Mahalingam, S., 2011. Critical role for macrophage migration inhibitory factor (MIF) in Ross River virus-induced arthritis and myositis. Proc Natl Acad Sci U S A 108, 12048-12053.

294

Hickie, I., Davenport, T., Wakefield, D., Vollmer-Conna, U., Cameron, B., Vernon, S.D., Reeves, W.C., Lloyd, A., 2006. Post-infective and chronic fatigue syndromes precipitated by viral and non-viral pathogens: prospective cohort study. BMJ (Clinical research ed.) 333, 575. Hidmark, A.S., McInerney, G.M., Nordstrom, E.K., Douagi, I., Werner, K.M., Liljestrom, P., Karlsson Hedestam, G.B., 2005. Early alpha/beta interferon production by myeloid dendritic cells in response to UV-inactivated virus requires viral entry and interferon regulatory factor 3 but not MyD88. J Virol 79, 10376-10385. Hollidge, B.S., Gonzalez-Scarano, F., Soldan, S.S., 2010. Arboviral encephalitides: transmission, emergence, and pathogenesis. Journal of neuroimmune pharmacology : the official journal of the Society on NeuroImmune Pharmacology 5, 428-442. Hollidge, B.S., Nedelsky, N.B., Salzano, M.V., Fraser, J.W., Gonzalez-Scarano, F., Soldan, S.S., 2012. Orthobunyavirus entry into neurons and other mammalian cells occurs via clathrin-mediated endocytosis and requires trafficking into early endosomes. J Virol 86, 7988-8001. Holzer, G.W., Coulibaly, S., Aichinger, G., Savidis-Dacho, H., Mayrhofer, J., Brunner, S., Schmid, K., Kistner, O., Aaskov, J.G., Falkner, F.G., Ehrlich, H., Barrett, P.N., Kreil, T.R., 2011. Evaluation of an inactivated Ross River virus vaccine in active and passive mouse immunization models and establishment of a correlate of protection. Vaccine 29, 4132-4141. Hsieh, P., Robbins, P.W., 1984. Regulation of asparagine-linked oligosaccharide processing. Oligosaccharide processing in Aedes albopictus mosquito cells. J Biol Chem 259, 2375-2382. Hu, W., Mengersen, K., Dale, P., Tong, S., 2010. Difference in mosquito species (Diptera: Culicidae) and the transmission of Ross River virus between coastline and inland areas in Brisbane, Australia. Environmental entomology 39, 88-97. Huang, Y.H., Lei, H.Y., Liu, H.S., Lin, Y.S., Liu, C.C., Yeh, T.M., 2000. Dengue virus infects human endothelial cells and induces IL-6 and IL-8 production. The American Journal of Tropical Medicine and Hygiene 63, 71-75. Huber, S.A., 2005. Increased susceptibility of male BALB/c mice to coxsackievirus B3-induced myocarditis: role for CD1d. Medical microbiology and immunology 194, 121-127. Ikegami, T., Makino, S., 2011. The Pathogenesis of Rift Valley Fever. Viruses 3, 493-519. Iniguez, S.D., Riggs, L.M., Nieto, S.J., Dayrit, G., Zamora, N.N., Shawhan, K.L., Cruz, B., Warren, B.L., 2014. Social defeat stress induces a depression- like phenotype in adolescent male c57BL/6 mice. Stress (Amsterdam, Netherlands) 17, 247-255. Inoue, K., Inoue, E., Imai, Y., 2013. Female sex hormones ameliorate arthritis in SKG mice. Biochemical and biophysical research communications 434, 740-745. Ivashkiv, L.B., 2003. Type I interferon modulation of cellular responses to cytokines and infectious pathogens: potential role in SLE pathogenesis. Autoimmunity 36, 473-479. Jacups, S.P., Whelan, P.I., Currie, B.J., 2008. Ross River virus and Barmah Forest virus infections: a review of history, ecology, and predictive models,

295 with implications for tropical northern Australia. Vector Borne Zoonotic Dis 8, 283-297. Jenkins, G.M., Rambaut, A., Pybus, O.G., Holmes, E.C., 2002. Rates of molecular evolution in RNA viruses: a quantitative phylogenetic analysis. Journal of molecular evolution 54, 156-165. Johnson, R.B., Wikle, J.C., 2014. Sex differences in inflammatory and apoptotic signaling molecules in normal and diseased human gingiva. Journal of periodontology 85, 1612-1619. Johnston, L.J., Halliday, G.M., King, N.J., 2000. Langerhans cells migrate to local lymph nodes following cutaneous infection with an arbovirus. The Journal of investigative dermatology 114, 560-568. Jones, A., Lowry, K., Aaskov, J., Holmes, E.C., Kitchen, A., 2010. Molecular evolutionary dynamics of Ross River virus and implications for vaccine efficacy. J Gen Virol 91, 182-188. Jones, M., Davidson, A., Hibbert, L., Gruenwald, P., Schlaak, J., Ball, S., Foster, G.R., Jacobs, M., 2005. Dengue virus inhibits alpha interferon signaling by reducing STAT2 expression. Journal of Virology 79, 5414-5420. Jones, P.H., Maric, M., Madison, M.N., Maury, W., Roller, R.J., Okeoma, C.M., 2013. BST-2/tetherin-mediated restriction of chikungunya (CHIKV) VLP budding is counteracted by CHIKV non-structural protein 1 (nsP1). Virology 438, 37-49. Jovanovic, D.V., Di Battista, J.A., Martel-Pelletier, J., Jolicoeur, F.C., He, Y., Zhang, M., Mineau, F., Pelletier, J.P., 1998. IL-17 stimulates the production and expression of proinflammatory cytokines, IL-beta and TNF-alpha, by human macrophages. J Immunol 160, 3513-3521. Jungo, F., Dayer, J.M., Modoux, C., Hyka, N., Burger, D., 2001. IFN-beta inhibits the ability of T lymphocytes to induce TNF-alpha and IL-1beta production in monocytes upon direct cell-cell contact. Cytokine 14, 272-282. Jupille, H.J., Medina-Rivera, M., Hawman, D.W., Oko, L., Morrison, T.E., 2013. A tyrosine-to-histidine switch at position 18 of the Ross River virus E2 glycoprotein is a determinant of virus fitness in disparate hosts. J Virol 87, 5970-5984. Jupille, H.J., Oko, L., Stoermer, K.A., Heise, M.T., Mahalingam, S., Gunn, B.M., Morrison, T.E., 2011. Mutations in nsP1 and PE2 are critical determinants of Ross River virus-induced musculoskeletal inflammatory disease in a mouse model. Virology 410, 216-227. Kato, H., Sato, S., Yoneyama, M., Yamamoto, M., Uematsu, S., Matsui, K., Tsujimura, T., Takeda, K., Fujita, T., Takeuchi, O., Akira, S., 2005. Cell type- specific involvement of RIG-I in antiviral response. Immunity 23, 19-28. Kay, B.H., Boyd, A.M., Ryan, P.A., Hall, R.A., 2007. Mosquito feeding patterns and natural infection of vertebrates with Ross River and Barmah Forest viruses in Brisbane, Australia. Am J Trop Med Hyg 76, 417-423. Kay, B.H., Fanning, I.D., Carley, J.G., 1982. Vector competence of Culex pipiens quinquefasciatus for Murray Valley encephalitis, Kunjin, and Ross River viruses from Australia. Am J Trop Med Hyg 31, 844-848. Kerr, P.J., Weir, R.C., Dalgarno, L., 1993. Ross River virus variants selected during passage in chick embryo fibroblasts: serological, genetic, and biological changes. Virology 193, 446-449.

296

Kilpatrick, A.M., Pape, W.J., 2013. Predicting human West Nile virus infections with mosquito surveillance data. American journal of epidemiology 178, 829-835. Klimstra, W.B., Nangle, E.M., Smith, M.S., Yurochko, A.D., Ryman, K.D., 2003. DC-SIGN and L-SIGN can act as attachment receptors for alphaviruses and distinguish between mosquito cell- and mammalian cell-derived viruses. Journal of Virology 77, 12022-12032. Klimstra, W.B., Ryman, K.D., Johnston, R.E., 1998. Adaptation of Sindbis virus to BHK cells selects for use of heparan sulfate as an attachment receptor. J Virol 72, 7357-7366. Kong, Y.Y., Feige, U., Sarosi, I., Bolon, B., Tafuri, A., Morony, S., Capparelli, C., Li, J., Elliott, R., McCabe, S., Wong, T., Campagnuolo, G., Moran, E., Bogoch, E.R., Van, G., Nguyen, L.T., Ohashi, P.S., Lacey, D.L., Fish, E., Boyle, W.J., Penninger, J.M., 1999. Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304-309. Kovacs, W.J., Olsen, N.J., 2011. Sexual dimorphism of RA manifestations: genes, hormones and behavior. Nature reviews. Rheumatology 7, 307-310. Kreil, T.R., Eibl, M.M., 1996. Nitric oxide and viral infection: NO antiviral activity against a flavivirus in vitro, and evidence for contribution to pathogenesis in experimental infection in vivo. Virology 219, 304-306. Kuhn, R.J., Niesters, H.G., Hong, Z., Strauss, J.H., 1991. Infectious RNA transcripts from Ross River virus cDNA clones and the construction and characterization of defined chimeras with Sindbis virus. Virology 182, 430- 441. Kuruvilla, J.G., Troyer, R.M., Devi, S., Akkina, R., 2007. Dengue virus infection and immune response in humanized RAG2(-/-)gamma(c)(-/-) (RAG- hu) mice. Virology 369, 143-152. Labadie, K., Larcher, T., Joubert, C., Mannioui, A., Delache, B., Brochard, P., Guigand, L., Dubreil, L., Lebon, P., Verrier, B., de Lamballerie, X., Suhrbier, A., Cherel, Y., Le Grand, R., Roques, P., 2010. Chikungunya disease in nonhuman primates involves long-term viral persistence in macrophages. The Journal of clinical investigation 120, 894-906. Lam, J., Takeshita, S., Barker, J.E., Kanagawa, O., Ross, F.P., Teitelbaum, S.L., 2000. TNF-alpha induces osteoclastogenesis by direct stimulation of macrophages exposed to permissive levels of RANK ligand. The Journal of clinical investigation 106, 1481-1488. Laskin, D.L., 2009. Macrophages and inflammatory mediators in chemical toxicity: a battle of forces. Chemical research in toxicology 22, 1376-1385. Lemm, J.A., Rumenapf, T., Strauss, E.G., Strauss, J.H., Rice, C.M., 1994. Polypeptide requirements for assembly of functional Sindbis virus replication complexes: a model for the temporal regulation of minus- and plus-strand RNA synthesis. The EMBO journal 13, 2925-2934. Lescar, J., Roussel, A., Wien, M.W., Navaza, J., Fuller, S.D., Wengler, G., Wengler, G., Rey, F.A., 2001. The Fusion glycoprotein shell of Semliki Forest virus: an icosahedral assembly primed for fusogenic activation at endosomal pH. Cell 105, 137-148. Levy, A., Valero, N., Espina, L.M., Anez, G., Arias, J., Mosquera, J., 2010. Increment of interleukin 6, tumour necrosis factor alpha, nitric oxide, C-

297 reactive protein and apoptosis in dengue. Trans R Soc Trop Med Hyg 104, 16- 23. Levy, D.E., Marie, I.J., Durbin, J.E., 2011. Induction and function of type I and III interferon in response to viral infection. Current opinion in virology 1, 476-486. Li, K., Xu, W., Guo, Q., Jiang, Z., Wang, P., Yue, Y., Xiong, S., 2009. Differential macrophage polarization in male and female BALB/c mice infected with coxsackievirus B3 defines susceptibility to viral myocarditis. Circulation research 105, 353-364. Libraty, D.H., Pichyangkul, S., Ajariyakhajorn, C., Endy, T.P., Ennis, F.A., 2001. Human dendritic cells are activated by dengue virus infection: enhancement by gamma interferon and implications for disease pathogenesis. Journal of Virology 75, 3501-3508. Lidbury, B.A., Rulli, N.E., Musso, C.M., Cossetto, S.B., Zaid, A., Suhrbier, A., Rothenfluh, H.S., Rolph, M.S., Mahalingam, S., 2011. Identification and characterization of a ross river virus variant that grows persistently in macrophages, shows altered disease kinetics in a mouse model, and exhibits resistance to type I interferon. J Virol 85, 5651-5663. Lidbury, B.A., Rulli, N.E., Suhrbier, A., Smith, P.N., McColl, S.R., Cunningham, A.L., Tarkowski, A., van Rooijen, N., Fraser, R.J., Mahalingam, S., 2008. Macrophage-derived proinflammatory factors contribute to the development of arthritis and myositis after infection with an arthrogenic alphavirus. The Journal of infectious diseases 197, 1585-1593. Lidbury, B.A., Simeonovic, C., Maxwell, G.E., Marshall, I.D., Hapel, A.J., 2000. Macrophage-induced muscle pathology results in morbidity and mortality for Ross River virus-infected mice. The Journal of infectious diseases 181, 27-34. Liljestrom, P., Lusa, S., Huylebroeck, D., Garoff, H., 1991. In vitro mutagenesis of a full-length cDNA clone of Semliki Forest virus: the small 6,000-molecular-weight membrane protein modulates virus release. J Virol 65, 4107-4113. Lin, R.J., Chang, B.L., Yu, H.P., Liao, C.L., Lin, Y.L., 2006. Blocking of interferon-induced Jak-Stat signaling by Japanese encephalitis virus NS5 through a protein tyrosine phosphatase-mediated mechanism. J Virol 80, 5908-5918. Lindsay, M., Oliveira, N., Jasinska, E., Johansen, C., Harrington, S., Wright, A.E., Smith, D., 1996. An outbreak of Ross River virus disease in Southwestern Australia. Emerg Infect Dis 2, 117-120. Lindsay, M.D., Broom, A.K., Wright, A.E., Johansen, C.A., Mackenzie, J.S., 1993a. Ross River virus isolations from mosquitoes in arid regions of Western Australia: implication of vertical transmission as a means of persistence of the virus. Am J Trop Med Hyg 49, 686-696. Lindsay, M.D., Coelen, R.J., Mackenzie, J.S., 1993b. Genetic heterogeneity among isolates of Ross River virus from different geographical regions. J Virol 67, 3576-3585. Linn, M.L., Mateo, L., Gardner, J., Suhrbier, A., 1998. Alphavirus-specific cytotoxic T lymphocytes recognize a cross-reactive epitope from the capsid protein and can eliminate virus from persistently infected macrophages. J Virol 72, 5146-5153.

298

Liu, H., Gao, X., Liang, G., 2011a. Newly recognized mosquito-associated viruses in mainland China, in the last two decades. Virology Journal 8, 68. Liu, W.J., Rourke, M.F., Holmes, E.C., Aaskov, J.G., 2011b. Persistence of multiple genetic lineages within intrahost populations of Ross River virus. J Virol 85, 5674-5678. Loewy, A., Smyth, J., von Bonsdorff, C.H., Liljestrom, P., Schlesinger, M.J., 1995. The 6-kilodalton membrane protein of Semliki Forest virus is involved in the budding process. J Virol 69, 469-475. Lozach, P.Y., Burleigh, L., Staropoli, I., Navarro-Sanchez, E., Harriague, J., Virelizier, J.L., Rey, F.A., Despres, P., Arenzana-Seisdedos, F., Amara, A., 2005. Dendritic cell-specific intercellular adhesion molecule 3-grabbing non- integrin (DC-SIGN)-mediated enhancement of dengue virus infection is independent of DC-SIGN internalization signals. J Biol Chem 280, 23698- 23708. Lupi, O., 2011. Mosquito-borne hemorrhagic fevers. Dermatol Clin 29, 33-38. Ma, D.Y., Suthar, M.S., 2015. Mechanisms of innate immune evasion in reemerging RNA viruses. Current opinion in virology 12, 26-37. Mahalingam, S., Friedland, J.S., Heise, M.T., Rulli, N.E., Meanger, J., Lidbury, B.A., 2003. Chemokines and viruses: friends or foes? Trends in Microbiology 11, 383-391. Malmgaard, L., 2004. Induction and regulation of IFNs during viral infections. Journal of Interferon and Cytokine Research: The Official Journal of the International Society for Interferon and Cytokine Research 24, 439-454. Marovich, M., Grouard-Vogel, G., Louder, M., Eller, M., Sun, W., Wu, S.J., Putvatana, R., Murphy, G., Tassaneetrithep, B., Burgess, T., Birx, D., Hayes, C., Schlesinger-Frankel, S., Mascola, J., 2001. Human dendritic cells as targets of dengue virus infection. The journal of investigative dermatology. Symposium proceedings / the Society for Investigative Dermatology, Inc. [and] European Society for Dermatological Research 6, 219-224. Marriott, I., Huet-Hudson, Y.M., 2006. Sexual dimorphism in innate immune responses to infectious organisms. Immunologic research 34, 177-192. Martina, B.E., Koraka, P., van den Doel, P., Rimmelzwaan, G.F., Haagmans, B.L., Osterhaus, A.D., 2008. DC-SIGN enhances infection of cells with glycosylated West Nile virus in vitro and virus replication in human dendritic cells induces production of IFN-alpha and TNF-alpha. Virus Res 135, 64-71. Martins Sde, T., Silveira, G.F., Alves, L.R., Duarte dos Santos, C.N., Bordignon, J., 2012. Dendritic cell apoptosis and the pathogenesis of dengue. Viruses 4, 2736-2753. Mathieu-Rivet, E., Kiefer-Meyer, M.C., Vanier, G., Ovide, C., Burel, C., Lerouge, P., Bardor, M., 2014. Protein N-glycosylation in eukaryotic microalgae and its impact on the production of nuclear expressed biopharmaceuticals. Frontiers in plant science 5, 359. Mazzon, M., Jones, M., Davidson, A., Chain, B., Jacobs, M., 2009. Dengue virus NS5 inhibits interferon-alpha signaling by blocking signal transducer and activator of transcription 2 phosphorylation. The Journal of infectious diseases 200, 1261-1270. McAbee, J., Li, Q., Yu, H., Kirkwood, K.L., 2012. Sexual dimorphism in periapical inflammation and bone loss from mitogen-activated protein kinase phosphatase-1 deficient mice. Journal of endodontics 38, 1097-1100.

299

McCracken, M.K., Christofferson, R.C., Grasperge, B.J., Calvo, E., Chisenhall, D.M., Mores, C.N., 2014. Aedes aegypti salivary protein "aegyptin" co- inoculation modulates dengue virus infection in the vertebrate host. Virology 468-470C, 133-139. Melancon, P., Garoff, H., 1987. Processing of the Semliki Forest virus structural polyprotein: role of the capsid protease. J Virol 61, 1301-1309. Mellor, P.S., 2000. Replication of arboviruses in insect vectors. Journal of Comparative Pathology 123, 231-247. Melton, J.V., Ewart, G.D., Weir, R.C., Board, P.G., Lee, E., Gage, P.W., 2002. Alphavirus 6K proteins form ion channels. J Biol Chem 277, 46923-46931. Menon, B., Gullick, N.J., Walter, G.J., Rajasekhar, M., Garrood, T., Evans, H.G., Taams, L.S., Kirkham, B.W., 2014. Interleukin-17+CD8+ T Cells Are Enriched in the Joints of Patients With Psoriatic Arthritis and Correlate With Disease Activity and Joint Damage Progression. Arthritis & Rheumatology (Hoboken, N.j.) 66, 1272-1281. Mims, C.A., Murphy, F.A., Taylor, W.P., Marshall, I.D., 1973. Pathogenesis of Ross River virus infection in mice. I. Ependymal infection, cortical thinning, and hydrocephalus. The Journal of infectious diseases 127, 121-128. Monath, T.P., Heinz, F.X., 1996. Flaviviruses, in: Fields, B.N., Knipe, D.M., Howley, P.M. (Eds.), Fields Virology, 3rd ed. Lippincott-Raven Publishers, Philadelphia, pp. 962-963. Mondotte, J.A., Lozach, P.Y., Amara, A., Gamarnik, A.V., 2007. Essential role of dengue virus envelope protein N glycosylation at asparagine-67 during viral propagation. J Virol 81, 7136-7148. Moreno-Altamirano, M.M., Romano, M., Legorreta-Herrera, M., Sanchez- Garcia, F.J., Colston, M.J., 2004. Gene expression in human macrophages infected with dengue virus serotype-2. Scandinavian Journal of Immunology 60, 631-638. Morizono, K., Ku, A., Xie, Y., Harui, A., Kung, S.K., Roth, M.D., Lee, B., Chen, I.S., 2010. Redirecting lentiviral vectors pseudotyped with Sindbis virus- derived envelope proteins to DC-SIGN by modification of N-linked glycans of envelope proteins. J Virol 84, 6923-6934. Morrison, J., Aguirre, S., Fernandez-Sesma, A., 2012. Innate immunity evasion by Dengue virus. Viruses 4, 397-413. Morrison, T.E., Fraser, R.J., Smith, P.N., Mahalingam, S., Heise, M.T., 2007. Complement contributes to inflammatory tissue destruction in a mouse model of Ross River virus-induced disease. J Virol 81, 5132-5143. Morrison, T.E., Simmons, J.D., Heise, M.T., 2008. Complement receptor 3 promotes severe ross river virus-induced disease. J Virol 82, 11263-11272. Morrison, T.E., Whitmore, A.C., Shabman, R.S., Lidbury, B.A., Mahalingam, S., Heise, M.T., 2006. Characterization of Ross River virus tropism and virus- induced inflammation in a mouse model of viral arthritis and myositis. J Virol 80, 737-749. Mosca, J.D., Pitha, P.M., 1986. Transcriptional and posttranscriptional regulation of exogenous human beta interferon gene in simian cells defective in interferon synthesis. Molecular and cellular biology 6, 2279-2283. Mosquera, J.A., Hernandez, J.P., Valero, N., Espina, L.M., Anez, G.J., 2005. Ultrastructural studies on dengue virus type 2 infection of cultured human monocytes. Virology Journal 2, 26.

300

Moxley, G., Posthuma, D., Carlson, P., Estrada, E., Han, J., Benson, L.L., Neale, M.C., 2002. Sexual dimorphism in innate immunity. Arthritis and rheumatism 46, 250-258. Muller, U., Steinhoff, U., Reis, L.F., Hemmi, S., Pavlovic, J., Zinkernagel, R.M., Aguet, M., 1994. Functional role of type I and type II interferons in antiviral defense. Science 264, 1918-1921. Munoz-Jordan, J.L., Laurent-Rolle, M., Ashour, J., Martinez-Sobrido, L., Ashok, M., Lipkin, W.I., Garcia-Sastre, A., 2005. Inhibition of alpha/beta interferon signaling by the NS4B protein of flaviviruses. Journal of Virology 79, 8004-8013. Munoz-Jordan, J.L., Sanchez-Burgos, G.G., Laurent-Rolle, M., Garcia-Sastre, A., 2003. Inhibition of interferon signaling by dengue virus. Proc Natl Acad Sci U S A 100, 14333-14338. Nagayama, Y., Tsubaki, T., Nakayama, S., Sawada, K., Taguchi, K., Tateno, N., Toba, T., 2006. Gender analysis in acute bronchiolitis due to respiratory syncytial virus. Pediatric allergy and immunology : official publication of the European Society of Pediatric Allergy and Immunology 17, 29-36. Neuvonen, M., Kazlauskas, A., Martikainen, M., Hinkkanen, A., Ahola, T., Saksela, K., 2011. SH3 domain-mediated recruitment of host cell amphiphysins by alphavirus nsP3 promotes viral RNA replication. PLoS pathogens 7, e1002383. Nguyen, N.D., Ahlborg, H.G., Center, J.R., Eisman, J.A., Nguyen, T.V., 2007. Residual lifetime risk of fractures in women and men. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 22, 781-788. Noret, M., Herrero, L., Rulli, N., Rolph, M., Smith, P.N., Li, R.W., Roques, P., Gras, G., Mahalingam, S., 2012. Interleukin 6, RANKL, and osteoprotegerin expression by chikungunya virus-infected human osteoblasts. The Journal of infectious diseases 206, 455-457: 457-459. Orlowsky, E.W., Kraus, V.B., 2015. The Role of Innate Immunity in Osteoarthritis: When Our First Line of Defense Goes On the Offensive. The Journal of rheumatology 42, 363-371. Osterlund, P.I., Pietila, T.E., Veckman, V., Kotenko, S.V., Julkunen, I., 2007. IFN regulatory factor family members differentially regulate the expression of type III IFN (IFN-lambda) genes. J Immunol 179, 3434-3442. Pacsa, A.S., Agarwal, R., Elbishbishi, E.A., Chaturvedi, U.C., Nagar, R., Mustafa, A.S., 2000. Role of interleukin-12 in patients with dengue hemorrhagic fever. FEMS Immunology and Medical Microbiology 28, 151-155. Palmer, D.R., Sun, P., Celluzzi, C., Bisbing, J., Pang, S., Sun, W., Marovich, M.A., Burgess, T., 2005. Differential effects of dengue virus on infected and bystander dendritic cells. Journal of Virology 79, 2432-2439. Park, E., Griffin, D.E., 2009. The nsP3 macro domain is important for Sindbis virus replication in neurons and neurovirulence in mice. Virology 388, 305- 314. Parker, M.D., Buckley, M.J., Melanson, V.R., Glass, P.J., Norwood, D., Hart, M.K., 2010. Antibody to the E3 glycoprotein protects mice against lethal venezuelan equine encephalitis virus infection. J Virol 84, 12683-12690.

301

Parrott, M.M., Sitarski, S.A., Arnold, R.J., Picton, L.K., Hill, R.B., Mukhopadhyay, S., 2009. Role of conserved cysteines in the alphavirus E3 protein. J Virol 83, 2584-2591. Peng, T., Wang, J.L., Chen, W., Zhang, J.L., Gao, N., Chen, Z.T., Xu, X.F., Fan, D.Y., An, J., 2009. Entry of dengue virus serotype 2 into ECV304 cells depends on clathrin-dependent endocytosis, but not on caveolae-dependent endocytosis. Canadian journal of microbiology 55, 139-145. Peranen, J., Rikkonen, M., Liljestrom, P., Kaariainen, L., 1990. Nuclear localization of Semliki Forest virus-specific nonstructural protein nsP2. J Virol 64, 1888-1896. Petitdemange, C., Wauquier, N., Rey, J., Hervier, B., Leroy, E., Vieillard, V., 2014. Control of Acute Dengue Virus Infection by Natural Killer Cells. Frontiers in Immunology 5. Pfeilschifter, J., Koditz, R., Pfohl, M., Schatz, H., 2002. Changes in proinflammatory cytokine activity after menopause. Endocrine reviews 23, 90-119. Phanthanawiboon, S., A, A.N., Panngarm, N., Limkittikul, K., Ikuta, K., Anantapreecha, S., Kurosu, T., 2014. Isolation and propagation of Dengue virus in Vero and BHK-21 cells expressing human DC-SIGN stably. Journal of virological methods 209, 55-61. Pierson, T.C., Diamond, M.S., 2012. Degrees of maturity: the complex structure and biology of flaviviruses. Current opinion in virology 2, 168-175. Pohlmann, S., Soilleux, E.J., Baribaud, F., Leslie, G.J., Morris, L.S., Trowsdale, J., Lee, B., Coleman, N., Doms, R.W., 2001. DC-SIGNR, a DC- SIGN homologue expressed in endothelial cells, binds to human and simian immunodeficiency viruses and activates infection in trans. Proc Natl Acad Sci U S A 98, 2670-2675. Poidinger, M., Roy, S., Hall, R.A., Turley, P.J., Scherret, J.H., Lindsay, M.D., Broom, A.K., Mackenzie, J.S., 1997. Genetic stability among temporally and geographically diverse isolates of Barmah Forest virus. Am J Trop Med Hyg 57, 230-234. Pozzetto, B., Gresser, I., 1985. Role of sex and early interferon production in the susceptibility of mice to encephalomyocarditis virus. J Gen Virol 66 ( Pt 4), 701-709. Quach, C., Piche-Walker, L., Platt, R., Moore, D., 2003. Risk factors associated with severe influenza infections in childhood: implication for vaccine strategy. Pediatrics 112, e197-201. Raghupathy, R., Chaturvedi, U.C., Al-Sayer, H., Elbishbishi, E.A., Agarwal, R., Nagar, R., Kapoor, S., Misra, A., Mathur, A., Nusrat, H., Azizieh, F., Khan, M.A., Mustafa, A.S., 1998. Elevated levels of IL-8 in dengue hemorrhagic fever. Journal of medical virology 56, 280-285. Register, T.C., Cann, J.A., Kaplan, J.R., Williams, J.K., Adams, M.R., Morgan, T.M., Anthony, M.S., Blair, R.M., Wagner, J.D., Clarkson, T.B., 2005. Effects of soy isoflavones and conjugated equine estrogens on inflammatory markers in atherosclerotic, ovariectomized monkeys. The Journal of clinical endocrinology and metabolism 90, 1734-1740. Richter, M.K., da Silva Voorham, J.M., Torres Pedraza, S., Hoornweg, T.E., van de Pol, D.P., Rodenhuis-Zybert, I.A., Wilschut, J., Smit, J.M., 2014.

302

Immature dengue virus is infectious in human immature dendritic cells via interaction with the receptor molecule DC-SIGN. PLoS One 9, e98785. Rikkonen, M., 1996. Functional significance of the nuclear-targeting and NTP-binding motifs of Semliki Forest virus nonstructural protein nsP2. Virology 218, 352-361. Rikkonen, M., Peranen, J., Kaariainen, L., 1994. Nuclear targeting of Semliki Forest virus nsP2. Archives of virology. Supplementum 9, 369-377. Ross, G.D., 2000. Regulation of the adhesion versus cytotoxic functions of the Mac-1/CR3/alphaMbeta2-integrin glycoprotein. Critical reviews in immunology 20, 197-222. Rulli, N.E., Guglielmotti, A., Mangano, G., Rolph, M.S., Apicella, C., Zaid, A., Suhrbier, A., Mahalingam, S., 2009. Amelioration of alphavirus-induced arthritis and myositis in a mouse model by treatment with bindarit, an inhibitor of monocyte chemotactic proteins. Arthritis and rheumatism 60, 2513-2523. Rulli, N.E., Melton, J., Wilmes, A., Ewart, G., Mahalingam, S., 2007. The molecular and cellular aspects of arthritis due to alphavirus infections: lesson learned from Ross River virus. Ann N Y Acad Sci 1102, 96-108. Rupp, J.C., Jundt, N., Hardy, R.W., 2011. Requirement for the amino- terminal domain of sindbis virus nsP4 during virus infection. J Virol 85, 3449- 3460. Russell, R.C., 1998. Mosquito-borne arboviruses in Australia: the current scene and implications of climate change for human health. International Journal for Parasitology 28, 955-969. Ryman, K.D., Gardner, C.L., Burke, C.W., Meier, K.C., Thompson, J.M., Klimstra, W.B., 2007a. Heparan sulfate binding can contribute to the neurovirulence of neuroadapted and nonneuroadapted Sindbis viruses. J Virol 81, 3563-3573. Ryman, K.D., Gardner, C.L., Meier, K.C., Biron, C.A., Johnston, R.E., Klimstra, W.B., 2007b. Early restriction of alphavirus replication and dissemination contributes to age-dependent attenuation of systemic hyperinflammatory disease. J Gen Virol 88, 518-529. Ryman, K.D., Klimstra, W.B., 2008. Host responses to alphavirus infection. Immunological reviews 225, 27-45. Sammels, L.M., Coelen, R.J., Lindsay, M.D., Mackenzie, J.S., 1995. Geographic distribution and evolution of Ross River virus in Australia and the Pacific Islands. Virology 212, 20-29. Samsa, M.M., Mondotte, J.A., Caramelo, J.J., Gamarnik, A.V., 2012. Uncoupling cis-Acting RNA elements from coding sequences revealed a requirement of the N-terminal region of dengue virus capsid protein in virus particle formation. J Virol 86, 1046-1058. Sawicki, D.L., Perri, S., Polo, J.M., Sawicki, S.G., 2006. Role for nsP2 proteins in the cessation of alphavirus minus-strand synthesis by host cells. J Virol 80, 360-371. Schneider, B.S., Soong, L., Zeidner, N.S., Higgs, S., 2004. Aedes aegypti salivary gland extracts modulate anti-viral and TH1/TH2 cytokine responses to sindbis virus infection. Viral immunology 17, 565-573. Schwartz, O., Albert, M.L., 2010. Biology and pathogenesis of chikungunya virus. Nature reviews. Microbiology 8, 491-500.

303

Seay, A.R., Wolinsky, J.S., 1982. Ross River virus-induced demyelination: I. Pathogenesis and histopathology. Annals of neurology 12, 380-389. Seth, R.B., Sun, L., Chen, Z.J., 2006. Antiviral innate immunity pathways. Cell Research 16, 141-147. Shabman, R.S., Morrison, T.E., Moore, C., White, L., Suthar, M.S., Hueston, L., Rulli, N., Lidbury, B., Ting, J.P., Mahalingam, S., Heise, M.T., 2007. Differential induction of type I interferon responses in myeloid dendritic cells by mosquito and mammalian-cell-derived alphaviruses. J Virol 81, 237-247. Shabman, R.S., Rogers, K.M., Heise, M.T., 2008. Ross River virus envelope glycans contribute to type I interferon production in myeloid dendritic cells. J Virol 82, 12374-12383. Shi, X., Jarvis, D.L., 2007. Protein N-glycosylation in the baculovirus-insect cell system. Current drug targets 8, 1116-1125. Shope, R.E., 1994. The discovery of arbovirus diseases. Ann N Y Acad Sci 740, 138-145. Shresta, S., Sharar, K.L., Prigozhin, D.M., Beatty, P.R., Harris, E., 2006. Murine model for dengue virus-induced lethal disease with increased vascular permeability. J Virol 80, 10208-10217. Silva, M.C., Guerrero-Plata, A., Gilfoy, F.D., Garofalo, R.P., Mason, P.W., 2007. Differential activation of human monocyte-derived and plasmacytoid dendritic cells by West Nile virus generated in different host cells. J Virol 81, 13640-13648. Sjoberg, M., Lindqvist, B., Garoff, H., 2011. Activation of the alphavirus spike protein is suppressed by bound E3. J Virol 85, 5644-5650. Smeets, T.J., Dayer, J.M., Kraan, M.C., Versendaal, J., Chicheportiche, R., Breedveld, F.C., Tak, P.P., 2000. The effects of interferon-beta treatment of synovial inflammation and expression of metalloproteinases in patients with rheumatoid arthritis. Arthritis and rheumatism 43, 270-274. Smit, J.M., Klimstra, W.B., Ryman, K.D., Bittman, R., Johnston, R.E., Wilschut, J., 2001. PE2 cleavage mutants of Sindbis virus: correlation between viral infectivity and pH-dependent membrane fusion activation of the spike heterodimer. J Virol 75, 11196-11204. Smit, J.M., Waarts, B.L., Kimata, K., Klimstra, W.B., Bittman, R., Wilschut, J., 2002. Adaptation of alphaviruses to heparan sulfate: interaction of Sindbis and Semliki forest viruses with liposomes containing lipid-conjugated heparin. J Virol 76, 10128-10137. Smith, P.L., Lombardi, G., Foster, G.R., 2005. Type I interferons and the innate immune response--more than just antiviral cytokines. Molecular Immunology 42, 869-877. Snyder, A.J., Mukhopadhyay, S., 2012. The alphavirus E3 glycoprotein functions in a clade-specific manner. J Virol 86, 13609-13620. Snyder, A.J., Sokoloski, K.J., Mukhopadhyay, S., 2012. Mutating conserved cysteines in the alphavirus e2 glycoprotein causes virus-specific assembly defects. J Virol 86, 3100-3111. Snyder, J.E., Kulcsar, K.A., Schultz, K.L., Riley, C.P., Neary, J.T., Marr, S., Jose, J., Griffin, D.E., Kuhn, R.J., 2013. Functional Characterization of the Alphavirus TF Protein. J Virol 87, 8511-8523.

304

Sokoloski, K.J., Snyder, A.J., Liu, N.H., Hayes, C.A., Mukhopadhyay, S., Hardy, R.W., 2013. Encapsidation of host-derived factors correlates with enhanced infectivity of Sindbis virus. J Virol 87, 12216-12226. Sourisseau, M., Schilte, C., Casartelli, N., Trouillet, C., Guivel-Benhassine, F., Rudnicka, D., Sol-Foulon, N., Le Roux, K., Prevost, M.C., Fsihi, H., Frenkiel, M.P., Blanchet, F., Afonso, P.V., Ceccaldi, P.E., Ozden, S., Gessain, A., Schuffenecker, I., Verhasselt, B., Zamborlini, A., Saib, A., Rey, F.A., Arenzana-Seisdedos, F., Despres, P., Michault, A., Albert, M.L., Schwartz, O., 2007. Characterization of reemerging chikungunya virus. PLoS pathogens 3, e89. Sperry, J.L., Friese, R.S., Frankel, H.L., West, M.A., Cuschieri, J., Moore, E.E., Harbrecht, B.G., Peitzman, A.B., Billiar, T.R., Maier, R.V., Remick, D.G., Minei, J.P., 2008. Male gender is associated with excessive IL-6 expression following severe injury. The Journal of trauma 64, 572-578; discussion 578- 579. St. John, A.L., Rathore, A.P.S., Yap, H., Ng, M.L., Metcalfe, D.D., Vasudevan, S.G., Abraham, S.N., 2011. Immune surveillance by mast cells during dengue infection promotes natural killer (NK) and NKT-cell recruitment and viral clearance. Proc Natl Acad Sci U S A 108, 9190-9195. Steinhauer, D.A., Holland, J.J., 1987. Rapid evolution of RNA viruses. Annual review of microbiology 41, 409-433. Stephenson, J.R., 2005. Understanding dengue pathogenesis: implications for vaccine design. Bulletin of the World Health Organisation 83, 308-314. Stoermer Burrack, K.A., Hawman, D.W., Jupille, H.J., Oko, L., Minor, M., Shives, K.D., Gunn, B.M., Long, K.M., Morrison, T.E., 2014. Attenuating mutations in nsP1 reveal tissue-specific mechanisms for control of Ross River virus infection. J Virol 88, 3719-3732. Stoermer, K.A., Burrack, A., Oko, L., Montgomery, S.A., Borst, L.B., Gill, R.G., Morrison, T.E., 2012. Genetic ablation of arginase 1 in macrophages and neutrophils enhances clearance of an arthritogenic alphavirus. J Immunol 189, 4047-4059. Strauss, E.G., Levinson, R., Rice, C.M., Dalrymple, J., Strauss, J.H., 1988. Nonstructural proteins nsP3 and nsP4 of Ross River and O'Nyong-nyong viruses: sequence and comparison with those of other alphaviruses. Virology 164, 265-274. Strauss, J.H., Strauss, E.G., 1994. The alphaviruses: gene expression, replication, and evolution. Microbiological reviews 58, 491-562. Strauss, J.H., Strauss, E.G., 2001. Virus evolution: how does an enveloped virus make a regular structure? Cell 105, 5-8. Subak-Sharpe, I., Dyson, H., Fazakerley, J., 1993. In vivo depletion of CD8+ T cells prevents lesions of demyelination in Semliki Forest virus infection. J Virol 67, 7629-7633. Suffredini, A.F., 2007. Systemic inflammation and sexual dimorphism: more than meets the eye. Critical care medicine 35, 1610-1612. Suhrbier, A., La Linn, M., 2004. Clinical and pathologic aspects of arthritis due to Ross River virus and other alphaviruses. Current opinion in rheumatology 16, 374-379.

305

Sun, P., Kochel, T.J., 2013. The Battle between Infection and Host Immune Responses of Dengue Virus and Its Implication in Dengue Disease Pathogenesis. The Scientific World Journal 2013. Tamm, K., Merits, A., Sarand, I., 2008. Mutations in the nuclear localization signal of nsP2 influencing RNA synthesis, protein expression and cytotoxicity of Semliki Forest virus. J Gen Virol 89, 676-686. Taniguchi, T., Takaoka, A., 2001. A weak signal for strong responses: interferon-alpha/beta revisited. Nature Reviews. Molecular Cell Biology 2, 378-386. Tassaneetrithep, B., Burgess, T.H., Granelli-Piperno, A., Trumpfheller, C., Finke, J., Sun, W., Eller, M.A., Pattanapanyasat, K., Sarasombath, S., Birx, D.L., Steinman, R.M., Schlesinger, S., Marovich, M.A., 2003. DC-SIGN (CD209) mediates dengue virus infection of human dendritic cells. The Journal of Experimental Medicine 197, 823-829. Thangamani, S., Higgs, S., Ziegler, S., Vanlandingham, D., Tesh, R., Wikel, S., 2010. Host Immune Response to Mosquito-Transmitted Chikungunya Virus Differs from That Elicited by Needle Inoculated Virus. PLoS One 5. Tilg, H., Kaser, A., 2004. Type I interferons and their therapeutic role in Th2- regulated inflammatory disorders. Expert Opinion on Biological Therapy 4, 469-481. Tomerini, D.M., Dale, P.E., Sipe, N., 2011. Does mosquito control have an effect on mosquito-borne disease? The case of Ross River virus disease and mosquito management in Queensland, Australia. J Am Mosq Control Assoc 27, 39-44. Tong, S., Dale, P., Nicholls, N., Mackenzie, J.S., Wolff, R., McMichael, A.J., 2008. Climate variability, social and environmental factors, and ross river virus transmission: research development and future research needs. Environmental health perspectives 116, 1591-1597. Toussirot, E., Bereau, M., Bossert, M., Malkoun, I., Lohse, A., 2014. Occurrence of Psoriatic Arthritis during Interferon Beta 1a Treatment for Multiple Sclerosis. Case reports in rheumatology 2014, 949317. Triantaphyllopoulos, K.A., Williams, R.O., Tailor, H., Chernajovsky, Y., 1999. Amelioration of collagen-induced arthritis and suppression of interferon- gamma, interleukin-12, and tumor necrosis factor alpha production by interferon-beta gene therapy. Arthritis and rheumatism 42, 90-99. Tuittila, M., Hinkkanen, A.E., 2003. Amino acid mutations in the replicase protein nsP3 of Semliki Forest virus cumulatively affect neurovirulence. J Gen Virol 84, 1525-1533. van Holten, J., Reedquist, K., Sattonet-Roche, P., Smeets, T.J., Plater-Zyberk, C., Vervoordeldonk, M.J., Tak, P.P., 2004. Treatment with recombinant interferon-beta reduces inflammation and slows cartilage destruction in the collagen-induced arthritis model of rheumatoid arthritis. Arthritis research & therapy 6, R239-249. van Knippenberg, I., Elliott, R.M., 2015. Flexibility of bunyavirus genomes: creation of an orthobunyavirus with an ambisense S segment. J Virol. Vaney, M.C., Duquerroy, S., Rey, F.A., 2013. Alphavirus structure: activation for entry at the target cell surface. Current opinion in virology 3, 151-158. Vasilakis, N., Deardorff, E.R., Kenney, J.L., Rossi, S.L., Hanley, K.A., Weaver, S.C., 2009. Mosquitoes put the brake on arbovirus evolution: experimental

306 evolution reveals slower mutation accumulation in mosquito than vertebrate cells. PLoS pathogens 5, e1000467. Vasta, G.R., 2009. Roles of galectins in infection. Nature reviews. Microbiology 7, 424-438. Vasudevan, S.S., Lopes, N.H., Seshiah, P.N., Wang, T., Marsh, C.B., Kereiakes, D.J., Dong, C., Goldschmidt-Clermont, P.J., 2003. Mac-1 and Fas activities are concurrently required for execution of smooth muscle cell death by M-CSF-stimulated macrophages. Cardiovascular research 59, 723-733. Verreck, F.A., de Boer, T., Langenberg, D.M., Hoeve, M.A., Kramer, M., Vaisberg, E., Kastelein, R., Kolk, A., de Waal-Malefyt, R., Ottenhoff, T.H., 2004. Human IL-23-producing type 1 macrophages promote but IL-10- producing type 2 macrophages subvert immunity to (myco)bacteria. Proc Natl Acad Sci U S A 101, 4560-4565. Villacres, M.C., Longmate, J., Auge, C., Diamond, D.J., 2004. Predominant type 1 CMV-specific memory T-helper response in humans: evidence for gender differences in cytokine secretion. Human immunology 65, 476-485. Villordo, S.M., Gamarnik, A.V., 2013. Differential RNA sequence requirement for dengue virus replication in mosquito and Mammalian cells. J Virol 87, 9365-9372. Vollmer-Conna, U., Piraino, B.F., Cameron, B., Davenport, T., Hickie, I., Wakefield, D., Lloyd, A.R., 2008. Cytokine polymorphisms have a synergistic effect on severity of the acute sickness response to infection. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 47, 1418-1425. Waldock, J., Chandra, N.L., Lelieveld, J., Proestos, Y., Michael, E., Christophides, G., Parham, P.E., 2013. The role of environmental variables on Aedes albopictus biology and chikungunya epidemiology. Pathogens and global health 107, 224-241. Walter, C.T., Barr, J.N., 2011. Recent advances in the molecular and cellular biology of bunyaviruses. J Gen Virol 92, 2467-2484. Wang, E., Volkova, E., Adams, A.P., Forrester, N., Xiao, S.Y., Frolov, I., Weaver, S.C., 2008. CHIMERIC ALPHAVIRUS VACCINE CANDIDATES FOR CHIKUNGUNYA. Vaccine 26, 5030-5039. Wang, T., Town, T., Alexopoulou, L., Anderson, J.F., Fikrig, E., Flavell, R.A., 2004. Toll-like receptor 3 mediates West Nile virus entry into the brain causing lethal encephalitis. Nature medicine 10, 1366-1373. Wang, W.K., Sung, T.L., Tsai, Y.C., Kao, C.L., Chang, S.M., King, C.C., 2002. Detection of dengue virus replication in peripheral blood mononuclear cells from dengue virus type 2-infected patients by a reverse transcription-real- time PCR assay. Journal of clinical microbiology 40, 4472-4478. Wang, Y.F., Sawicki, S.G., Sawicki, D.L., 1991. Sindbis virus nsP1 functions in negative-strand RNA synthesis. J Virol 65, 985-988. Way, S.J., Lidbury, B.A., Banyer, J.L., 2002. Persistent Ross River virus infection of murine macrophages: an in vitro model for the study of viral relapse and immune modulation during long-term infection. Virology 301, 281-292. Wengler, G., Koschinski, A., Wengler, G., Dreyer, F., 2003. Entry of alphaviruses at the plasma membrane converts the viral surface proteins into

307 an ion-permeable pore that can be detected by electrophysiological analyses of whole-cell membrane currents. J Gen Virol 84, 173-181. White, G.S., Pickett, B.E., Lefkowitz, E.J., Johnson, A.G., Ottendorfer, C., Stark, L.M., Unnasch, T.R., 2011. Phylogenetic analysis of eastern equine encephalitis virus isolates from Florida. Am J Trop Med Hyg 84, 709-717. White, J., Helenius, A., 1980. pH-dependent fusion between the Semliki Forest virus membrane and liposomes. Proc Natl Acad Sci U S A 77, 3273- 3277. WHO, 2001. Impact of Dengue. World Health Organisation (WHO). Wichmann, O., Jelinek, T., 2004. Dengue in travelers: a review. Journal of Travel Medicine 11, 161-170. Wong, P.K., Quinn, J.M., Sims, N.A., van Nieuwenhuijze, A., Campbell, I.K., Wicks, I.P., 2006. Interleukin-6 modulates production of T lymphocyte- derived cytokines in antigen-induced arthritis and drives inflammation- induced osteoclastogenesis. Arthritis and rheumatism 54, 158-168. Wu, S.J., Grouard-Vogel, G., Sun, W., Mascola, J.R., Brachtel, E., Putvatana, R., Louder, M.K., Filgueira, L., Marovich, M.A., Wong, H.K., Blauvelt, A., Murphy, G.S., Robb, M.L., Innes, B.L., Birx, D.L., Hayes, C.G., Frankel, S.S., 2000. Human skin Langerhans cells are targets of dengue virus infection. Nature medicine 6, 816-820. Zacks, M.A., Paessler, S., 2010. Encephalitic alphaviruses. Veterinary microbiology 140, 281-286. Zaid, A., Rulli, N.E., Rolph, M.S., Suhrbier, A., Mahalingam, S., 2011. Disease exacerbation by etanercept in a mouse model of alphaviral arthritis and myositis. Arthritis and rheumatism 63, 488-491. Zawatzky, R., Kirchner, H., DeMaeyer-Guignard, J., DeMaeyer, E., 1982. An X-linked locus influences the amount of circulating interferon induced in the mouse by herpes simplex virus type 1. J Gen Virol 63, 325-332. Zhang, W., Heil, M., Kuhn, R.J., Baker, T.S., 2005. Heparin binding sites on Ross River virus revealed by electron cryo-microscopy. Virology 332, 511-518. Zhang, X., Fugere, M., Day, R., Kielian, M., 2003. Furin processing and proteolytic activation of Semliki Forest virus. J Virol 77, 2981-2989. Zhang, Y., Burke, C.W., Ryman, K.D., Klimstra, W.B., 2007. Identification and characterization of interferon-induced proteins that inhibit alphavirus replication. J Virol 81, 11246-11255. Zhou, Z., Hamming, O.J., Ank, N., Paludan, S.R., Nielsen, A.L., Hartmann, R., 2007. Type III interferon (IFN) induces a type I IFN-like response in a restricted subset of cells through signaling pathways involving both the Jak- STAT pathway and the mitogen-activated protein kinases. J Virol 81, 7749- 7758. Zhu, W., Wang, L., Yang, Y., Jia, J., Fu, S., Feng, Y., He, Y., Li, J.P., Liang, G., 2010. Interaction of E2 glycoprotein with heparan sulfate is crucial for cellular infection of Sindbis virus. PLoS One 5, e9656.

308