A Dissertation Entitled Restriction of Tick-Borne Flaviviruses in the White

A Dissertation

entitled

Restriction of tick-borne flaviviruses in the white-footed mouse

Adaeze Izuogu

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the

Doctor of Philosophy Degree in Biomedical Sciences:

Medical Microbiology and Immunology

______Dr. R. Travis Taylor, Committee Chair

______Dr. Kevin Pan, Committee Member

______Dr. Mark Wooten, Committee Member

______Dr. Randall Worth, Committee Member

______Dr. Saurabh Chattopadhyay, Committee Member

______Dr. Amanda Bryant-Friedrich, Dean College of Graduate Studies

The University of Toledo

August 2017

This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author. An Abstract of

Restriction of Tick-borne Flaviviruses in The White-footed Mouse

Adaeze Izuogu

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Doctor of Philosophy Degree in Biomedical Science: Medical Microbiology and Immunology

The University of Toledo

August 2017

Tick-borne flaviviruses (TBFVs), including Powassan virus and tick-borne encephalitis virus cause encephalitis or hemorrhagic fevers in humans with case-fatality rates ranging from 1-30%. Despite severe disease in human hosts, TBFV infection of natural rodent hosts has little noticeable effect. Currently, the basis for this resistance to disease is not known. We hypothesize that the coevolution of flaviviruses with their respective hosts has shaped the evolution of potent antiviral factors that suppress virus replication and protect the host from lethal infection. In the current study, we compared virus infection between reservoir host cells and related susceptible species. Infection of primary fibroblasts from the white-footed mouse (Peromyscus leucopus, a representative host) with multiple TBFVs showed up to a 10,000-fold reduction in virus titer compared to control Mus musculus cells. However, replication of the unrelated vesicular stomatitis virus was equivalent in P. leucopus and M. musculus cells suggesting that restriction was virus-specific. Step-wise comparison of the virus infection cycle revealed a significant block to viral RNA replication, but not virus entry, in P. leucopus cells. To understand

iii the role of the type I interferon (IFN) response in virus restriction, we depleted signal transducer and activator of transcription 1 (STAT1) or the type I IFN receptor (IFNAR1) by RNA interference. Loss of IFNAR1 or STAT1 significantly relieved the block in virus replication in P. leucopus cells. The major IFN antagonist encoded by TBFV, nonstructural protein 5, was functional in P. leucopus cells, thus ruling out ineffective viral antagonism of the host IFN response and increased virus susceptibility to intrinsic cellular responses. Additionally, Kunjin virus, a member of the mosquito-borne group of flaviviruses was similarly restricted in P. leucopus cells suggesting that virus restriction is specific to the flavivirus family. Collectively, this work demonstrates that the IFN response of P. leucopus imparts a strong barrier to flavivirus replication. Future identification of the IFN-stimulated genes responsible for virus restriction specifically in

P. leucopus will yield mechanistic insight into efficient control of virus replication and may inform the development of antiviral therapeutics.

This work is dedicated to my parents and brothers. Your love shapes me, your unwavering support strengthens me.

Acknowledgements

My heartfelt gratitude goes to my advisor Dr. Travis Taylor who has been an amazing guide, teacher and all-round excellent mentor. I am truly grateful for the opportunity I have had to work with him and learn from him. To my committee members, Dr. Kevin

Pan, Dr. Mark Wooten, Dr. Randall Worth, Dr. Saurabh Chattopadhyay, I say a big thank you for the wealth of wisdom and positivity in charting this project and helping my scientific progress. I also thank Dr. Isabel Novella for getting me more excited about virology and for serving on my committee during her time here. Special thanks go to past and present members of the Taylor lab particularly John Presloid, Brian Youseff and

Heather Brown, it has been an immense pleasure to work with you all, thank you for the lessons, laughs and building a positive work environment.

I thank all the faculty, staff and students of the department of Medical Microbiology and

Immunology for being supportive and contributing to my learning process. This department has re-enforced an age-old saying that “it takes a village to raise a child”.

I appreciate our collaborators Dr. Sonja Best, Dr. Kristen McNally, Dr. Munshi-South,

Dr. Stephen Harris and Dr. Christopher Burlak for their valuable contribution to this work. My graduate school experience was made so much better by the love of my friends and life-given sisters Nneka, Akor, Oluchi, Jennifer, Anita and Brianne. Thank you ladies for being pillars of support. Thanks to my teacher, friend and sister EJ, for helping me to choose science and pursue an advanced degree.

The highest praise and glory go to my father in heaven, the center of my joy. With Him, my life and dreams are made possible.

Table of Contents

Abstract ...... iii

Acknowledgements ...... v

Table of Contents ...... vi

List of Tables ...... xi

List of Figures ...... xii

List of Abbreviations ...... xiii

List of Symbols ...... xviii

1 Literature Review ...... 1

1.1 Flaviviruses ...... 1

1.1.1 Virion structure and organization ...... 1

1.1.1.1 Viral structural proteins ...... 3

1.1.1.2 Viral non-structural proteins ...... 4

1.1.2 Flavivirus replication cycle ...... 5

1.1.3 Antigenic structure of flaviviruses ...... 12

1.1.4 Flavivirus pathogenesis and clinical manifestation ...... 12

1.1.5 Arthropod-borne flaviviruses ...... 16

1.1.6 Epidemiology of flavivirus disease ...... 21

1.1.7 Tick-borne encephalitis virus serocomplex ...... 22

1.2 Innate Immune response to flaviviruses ...... 25

1.2.1 Interferon induction pathway ...... 25

1.2.1.1 RIG-I-like receptors ...... 26

1.2.1.2 Toll-like receptors ...... 28

1.2.2 Interferon response pathway ...... 30

1.2.2.1 Jak-STAT signaling ...... 29

1.2.2.2 Interferon stimulated genes ...... 30

1.3 Flavivirus Restriction factors ...... 32

1.3.1. Identification of restriction factors ...... 33

1.3.2. Key restriction factor groups ...... 38

1.3.2.1 STATs ...... 38

1.3.2.2 TRIMs ...... 39

1.3.2.3 OASs ...... 42

1.3.2.4 IFITS and IFITMs ...... 44

1.3.2.5 IRFs ...... 45

1.3.3 Probing for restriction factors in relevant hosts ...... 46

1.3.4 Restriction factors and drug development ...... 47

1.4 Viral Antagonism of the interferon response ...... 50

1.4.1 Flavivirus immune evasion strategies ...... 50

1.4.2 Flavivirus NS5 antagonist mechanism ...... 52

1.4.3 Role of antagonism in species tropism ...... 53

1.5 Viral modulation of reactive oxygen species ...... 54

1.6 Reservoir species ...... 56

1.6.1 Viral maintenance in nature ...... 57

1.6.2 Disease resistance in rodent hosts ...... 59

1.6.3 Peromyscus leucopus as reservoir species of Powassan virus ...... 61

2. Materials and Methods ...... 64

vii

2.1 Cell culture and reagents ...... 64

2.2 Plasmids and vectors ...... 64

2.3 Antibodies ...... 65

2.4 Viruses and infections ...... 65

2.5 Virus titration by immunofocus assay and plaque assay ...... 66

2.6 Immunofluorescence confocal microscopy ...... 67

2.7 Western blotting ...... 68

2.8 Novel gene sequencing ...... 68

2.9 Quantitative real time polymerase chain reaction ...... 69

2.10 Antiviral assay ...... 70

2.11 Viral entry assays ...... 71

2.12 Infectious clone technology ...... 71

2.13 Generation of lentiviruses ...... 72

2.14 Generation of cell lines ...... 73

2.15 Construction of ectopic expression vectors ...... 73

2.16 Cell transfection for immunoassays ...... 74

2.17 Transfected cell lines ...... 75

2.18 Degradation assays ...... 75

2.19 Co-affinity purification ...... 76

2.20 Ubiquitination assays ...... 76

2.21 Co-localization assays ...... 77

2.22 Assessing total cellular ROS ...... 77

2.23 Assessing mitochondrial ROS ...... 78

viii

2.24 Testing the role of ROS in LGTV infection ...... 78

2.25 Statistical analysis ...... 78

3. Restriction of tick-borne flaviviruses in a cell culture model of P. leucopus ...... 79

3.1 Introduction ...... 79

3.2 Results ...... 80

3.2.1 Langat virus is restricted in P. leucopus ...... 80

3.2.2 Virulent TBFV strains are restricted in P. leucopus ...... 82

3.2.3 Restriction in P. leucopus is specific to flaviviruses ...... 82

3.2.4 Virus entry is not inhibited in P. leucopus cells ...... 83

3.2.5 Viral RNA is translated in P. leucopus cells ...... 85

3.2.6 Viral RNA replication is inhibited in P. leucopus cells ...... 86

3.3 Chapter 3 discussion ...... 87

Data figures ...... 91

4. Investigating the role of interferon signaling in P. leucopus ...... 98

4.1 Introduction ...... 98

4.2 Results ...... 100

4.2.1 Interferon impacts LGTV replication in P. leucopus ...... 100

4.2.2 LGTV NS5 functions as an IFN antagonist in P. leucopus ...... 101

4.2.3 Homologs of antiviral signaling genes identified in P. leucopus ..103

4.2.4 Targeting STAT1 relieves LGTV restriction in P. leucopus ...... 104

4.2.5 Signaling through IFNAR1 is necessary for LGTV restriction .....105

4.2.6 Characterizing plTRIM79 as a potential restriction factor ...... 106

4.2.7 plTRIM79 is a virus and interferon-stimulated gene ...... 108

4.2.8 plTRIM79 interacts with LGTV NS5 ...... 109

4.2.9 plTRIM79 impacts ubiquitination of LGTV NS5 ...... 110

4.3 Chapter 4 discussion ...... 111

Data figures ...... 117

5. Preliminary studies on the role of oxidative stress in P. leucopus resistance ...... 128

5.1 Introduction ...... 128

5.2 Results ...... 129

5.2.1 Cellular ROS in cells with LGTV infection ...... 129

5.2.2 Mitochondrial ROS in response to LGTV infection ...... 130

5.2.3 Inducing ROS in P. leucopus increases LGTV replication ...... 130

5.3 Chapter 5 discussion ...... 131

Data figures ...... 134

6. Discussion and Summary ...... 137

References ...... 146

List of Tables

4-1 Comparison of antiviral genes in P. leucopus and M. musculus ...... 127

List of Figures

1-1 Schematic of flavivirus infection cycle ...... 10 1-2 Flavivirus replication complex and RNA synthesis ...... 11 1-3 Viral maintenance in ticks ...... 20 1-4 Phylogeny tree and spread of flaviviruses ...... 24 1-5 Overview of interferon signaling ...... 32 1-6 Challenges to large ISG screens ...... 49 3-1 LGTV is restricted in P. leucopus cells ...... 91 3-2 Virulent TBFVs are restricted in P. leucopus cells ...... 92 3-3 Restriction in P. leucopus cells is specific to flaviviruses ...... 93 3-4 Virus entry is not restricted in P. leucopus ...... 94 3-5 Viral RNA can be translated in P. leucopus ...... 95 3-6 The viral RNA synthesis step is blocked in P. leucopus ...... 96 3-7 Hypothetical model of virus restriction ...... 97 4-1 IFN protects P. leucopus from virus infection ...... 117 4-2 LGTV NS5 antagonizes IFN signaling in P. leucopus ...... 118 4-3 Homologs of antiviral genes are expressed in P. leucopus ...... 119 4-4 STAT1 KD relives virus restriction in P. leucopus ...... 121 4-5 IFNAR1 KD relieves virus restriction in P. leucopus ...... 123 4-7 plTRIM79 is regulated by the proteasome ...... 124 4-8 plTRIM79 interacts with LGTV NS5 ...... 125 4-9 plTRIM79 impacts the ubiquitination level of LGTV NS5 ...... 126 5-1 Cellular ROS is increased by LGTV in susceptible cells ...... 134 5-2 Mitochondrial ROS is unchanged with LGTV infection ...... 135 5-3 ROS induction increases virus replication in P. leucopus ...... 136

xii

List of Abbreviations

C ...... Capsid CARD ...... Caspase Recruitment Domain cDNA ...... Complementary DNA CNS ...... Central Nervous System CO-AP ...... Co-Affinity purification CO-IP ...... Co-Immunoprecipitation CTD ...... C-terminal Domain

DCs ...... Dendritic cells DC-SIGN ...... Dendritic Cell-Specific Intercellular Adhesion Molecule-3-Grabbing Non-Integrin DENV ...... Dengue Virus DHF ...... Dengue Hemorrhagic Fever DMEM ...... Dulbecco's Modified Eagle Medium dsRNA ...... Double-stranded RNA

E ...... Envelop ELF ...... E74-Like Factor FE-TBEV ...... Far Eastern Tick-Borne Encephalitic Virus

HSP ...... Heat shock protein

ICAM ...... Intercellular Adhesion Molecule IFIT ...... Interferon Induced Proteins with Tetratricopeptide Repeats IFITM ...... Interferon-Induced Transmembrane Protein IFN Interferon IFNAR ...... Interferon- alpha/beta receptor IPS-1 ...... Interferon-Beta Promoter Stimulator IRF ...... Interferon Regulatory Factor ISG ...... Interferon Stimulated Gene ISRE ...... Interferon-Stimulated Response Element

JAK ...... Janus Kinase JEV ...... Japanese Encephalitis Virus

KD ...... Knockdown KUNV ...... Kunjin Virus

xiii

LAMP ...... Lysosome-Associated Membrane Protein LGP2 ...... Laboratory of Genetics and Physiology 2 LGTV ...... Langat Virus

M ...... Membrane MAVS ...... Mitochondrial Antiviral Signaling Protein MBFV ...... Mosquito-borne Flavivirus MDA5 ...... Melanoma Differentiation-Associated protein 5 MEF ...... Mouse Embryonic Fibroblasts MX ...... Myxovirus Resistance GTPase

NKV ...... No known vector NKκB ...... Nuclear Factor Kappa-Light-Chain- Enhancer of Activated B Cells NS ...... Non-Structural

OAS ...... Oligoadenylate Synthetase

PAMPs ...... Pathogen-Associated Molecular Pattern PEF ...... Peromyscus Embryonic Fibroblasts PGSC ...... Peromyscus Genetic Stock Center POWV ...... Powassan Virus PrM ...... Pre-Membrane PRRs ...... Pattern Recognition Receptors

RD ...... Repressor Domain RdRp ...... RNA-dependent RNA Polymerase RF ...... Restriction RIG-I ...... Retinoic Acid Inducible Gene I RING ...... Really Interesting New Gene RLR ...... Rig-I-like Receptor RNA ...... Ribonucleic Acid

SGE ...... Salivary Gland Extract siRNA ...... Short Interfering Rna SLA ...... Stem Loop A SOCS ...... Suppressor of Cytokine Signaling ssRNA ...... Single-stranded RNA STAT ...... Signal Transducer and Activator of Transcription STING ...... Stimulator of Interferon Genes

TANK ...... TRAF Family Member Associated NFKB Activator TBD ...... Tick-Borne Disease

xiv

TBEV ...... Tick-Borne Encephalitis Virus TBFV ...... Tick-Borne Flavivirus TLR ...... Toll-Like Receptor TRAF ...... Tumor necrosis factor receptor-associated factor TRIM ...... Tripartite motif TYK ...... Tyrosine Kinase

USP ...... Ubiquitin-Specific Peptidase UTR ...... Untranslated Region

WE-TBEV ...... Western Tick-Borne encephalitic virus WNV ...... West Nile Virus

Y2H ...... Yeast 2-Hybrid YFV ...... Fever Virus ZIKV ...... Zika Virus

List of Symbols

α ...... Alpha β ...... Beta ɛ ...... Epsilon κ ...... Kappa µ ...... Mu, micro

xvi

Chapter 1

Introduction and Literature Review

1.1 Flaviviruses

This genus of viruses belongs to the family Flaviviridae with pestiviruses and hepaciviruses being the other two genera in the family. The nomenclature derives from the root latin word ‘flavus’ meaning ‘yellow’ to represent one of the earlier identified members of the genus – the yellow fever virus (YFV). Members of the flavivirus genus have been associated with human disease which can result in mortality and prolonged clinical impacts. Some notable disease-causing flaviviruses include West Nile virus (WNV), dengue virus (DENV), Powassan virus (POWV),

Zika virus (ZIKV), tick-borne encephalitis virus (TBEV) and Japanese encephalitis virus (1).

Diseases caused by flaviviruses are broadly categorized into two main clinical presentations – encephalitis and hemorrhagic fevers. Indeed, flaviviruses contribute a substantial percentage of the viral encephalitic cases reported in the clinic. Despite the distinct disease symptoms that occur during infection, flaviviruses have a conserved genome structure that serves to propagate the virus during infection of a host cell.

1.1.1 Virion structure and organization

Flaviviruses are generally small, icosahedral particles of about 50nm bearing an outer envelope glycoprotein anchored in a host-derived membrane lipid bilayer (2). The virus encodes an 11kb single stranded RNA genome of positive polarity which is enclosed within the capsid. Together, the envelop (E), membrane (M), and capsid (C) proteins comprise the viral structural proteins which form part of the virion particle and are expressed when the virus gets into a cell. The

composition of the viral genome and capsid protein together form the nucleocapsid. In addition to the 3 structural proteins that make up the virus particle during host cell infection, the virus expresses 7 non-structural (NS) proteins: NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5 within the host cell making a total of 10 viral proteins to mediate the infection cycle (2,3)

1.1.1.1 Viral structural proteins

The structural proteins generally function in entry, maturation, assembly, egress and overall promote viral pathogenesis.

The E protein is a 53 kDa protein highly conserved among flaviviruses which has been elucidated as a type II viral fusion protein having 3 domains namely E-DI, E-DII and

E-DIII linked together by short hinges (4). There are 180 copies of E present on a virion particle and its main functions include mediating virus attachment, host cell entry and membrane fusion to release new virus particles from infected cells. Characterization and atomic resolution of the envelop protein domains have provided some insight into virus entry, antibody binding regions and glycosylation sites within the virus envelop. The DIII domain interacts with cellular receptors while the ‘finger’- like structures on the D-II domain are implicated in dimerization. Domain I primarily serves as a central domain linking domains II and III (5). The three domains are in rotation to allow for structural/conformational changes to occur in the process of virus maturation and fusion.

X-ray crystallography and cryo-electron microscopy studies have shown that the exposed portion of the E protein can induce neutralizing antibodies. Importantly, antibody neutralization by flaviviruses requires numerous antibody molecules to bind the virus particle (6).

The C protein is a 14kDA basic protein having affinity for both nucleic acids and lipid membranes (7,8). Its main function in the virus is assembly of the nucleocapsid bearing a copy of the viral genome. Experiments performed by nuclear magnetic resonance show that there is an internal hydrophobic domain within the C protein and additional α-helices which can associate with the internal hydrophobic domain to form dimers (9,10). Functionally, the positively charged surface of this protein aids to stabilize it with viral RNA in a dephosphorylation-dependent manner and potentially promote association with lipid droplets (11). Protein localization has been deciphered in the ER membrane, the nucleolus and the cytoplasm within lipid droplets (7).

The prM protein is located downstream of the C protein, and serves as the membrane protein localizing to the ER during translation of the polyprotein from a signal sequence. Its main function is to promote chaperone-mediated folding of the E protein.

Importantly, the ‘pr’ peptide specifically conceals the fusion loop of the E protein to avoid premature fusion with host cell membranes during virus egress. The presence of prM on a virion particle connotes it as an immature particle, therefore, cleavage of prM is required to yield the M protein which comprises the mature form of the virus (5). Cleavage occurs at the trans-Golgi network by host furin enzyme and requires a low pH environment to promote structural re-organization which then exposes the cleavage site and allows pr to remain associated with the E protein. It is important to note that this cleavage process is not completely efficient since virions can be released immaturely or partially mature

(12,13).

1.1.1.2 Viral non-structural proteins

These proteins are translated within the cell, they are not present on the incoming virion particle and their main tasks involve replicating the viral genome and evading host immune responses as described in later sections.

NS1 is an immunogenic glycoprotein found secreted on the cell surface in a dimeric form (14,15). It associates with NS4A to promote virus replication within vesicular compartments and also antagonizes the host complement pathways (2).

NS2A is recruited to the site of replication and associates with other NS proteins to mediate replication. It has been implicated in mediating neuroinvasiveness in some virus strains (16).

NS2B and NS3 form the viral protease NS2B/3 with NS2B functioning as the cofactor of NS3. Together with host proteases, this viral protease cleaves the viral polyprotein at specific sites. Furthermore, NS3 performs additional functions as the viral helicase, associating with NS5 to mediate replication and efficiently antagonizing host responses (17).

NS4A is a transmembrane protein which contributes anchorage of the viral replication components to the ER and also modifies host membranes for replication (18).

NS4B is a membrane protein which co-localizes with the replicative form of the viral

RNA and also contributes to immune antagonism strategies of the virus (2,19).

NS5 is considered a major virulence factor owing to its large size, conservation and multiple functions during infection. It provides methyltransferase (MTase), guanylyltransferase (GTase) and immune antagonism activity for the virus. Importantly, in

a study by Yoshii et al., 2014, NS5 was implicated as a major determinant of neurovirulence in the murine host (20).

1.1.2 Flavivirus replication cycle

Flavivirus particles first attach to cells by the use of attachment factors and next initiate the process of cell entry via interactions between viral glycoproteins on the surface of the virus and host cell receptors. The specific receptors for flavivirus entry have not been clearly defined, however, putative factors have been studied and a major player in the viral attachment process include sulfated glycosaminoglycans like host heparin sulfate which ultimately allow virus concentrations to increase on the cell surface (4). Additionally, a disaccharide on the glycosphingolipid, neolactotetraosylceramide has been identified as an attachment factor of some flaviviruses such as DENV. On the other hand, a potential flavivirus receptor is the dendritic cell-specific intercellular adhesion molecule-3

(ICAM3)-Grabbing Non-integrin DC-SIGN which potentially requires other attachment factors to mediate virus entry (21,22). Ultimately, variable usage of cell attachment factors and receptors depend on the strain of virus and more studies are needed to clearly elucidate the flavivirus receptors on various cell types. Virus internalization occurs by receptor-mediated endocytosis and the virus is delivered in clathrin-coated pits into a pre- lysosomal endocytic compartment. The early endomsomes mature into late endosomes and the low acidic pH within the endosomal compartment allows a conformational change in the E glycoprotein thereby promoting fusion of the viral membrane and host cell endosomal membrane (22,23). This membrane fusion releases the nucleocapsid into the

cytoplasm of the host cell where capsid disintegration further causes the release of viral

RNA by the nucleocapsid uncoating (24,25). The viral RNA encodes one polyprotein which is cleaved to generate individual viral proteins as described above. Specifically, regions C-prM, prM-E, E-NS1 and NS4A c-terminus are cleaved by the host signal peptidase. On the other hand, the viral serine protease cleaves regions NS21/B,

NS2B/NS3, NS3/NS4A, NS4A/NS4B and NS4B/NS5. It is hypothesized that viral translation and replication occur sequentially, rather than simultaneously as seen with other viruses, particularly since translation occurs in the 5’-3’ orientation while replication proceeds in the opposite direction (26).

Ultimately, the process of RNA replication, capping and methylation is mediated by a replication complex made up of viral NS proteins, the viral genomic RNA and other host factors that are currently unidentified (2). Results of pull-down assays suggest that all the flavivirus NS proteins are indeed present in the replication complex and perform distinct roles to ensure viral genome replication. Replication occurs on modified ER membranes and electron microscopy techniques have revealed that spherical vesicles are formed in infected cells by invagination of the ER membrane. These vesicles are single- membrane structures expressing the viral NS proteins and viral genomic RNA therefore implicating them as the sites of virus replication. Additionally, a pore on the vesicle is speculated to permit exchange of nucleotides between the vesicle and the cytoplasm

(27,28). Anchorage of the replication complex and modifications on these ER membranes are likely performed by the viral membrane proteins NS2A, NS2B, NS4A and NS4B

(14,26). Although NS3 lacks a transmembrane region, its interaction with the protease cofactor NS2B allows it to localize to the membrane. Similarly, while NS5 is not a transmembrane protein, its interaction with NS3 leads to its localization at the membrane.

Overall, these replication processes require adequate localization of the viral proteins to mediate interaction with viral and host cell factors (29).

The actual process of RNA synthesis is primarily driven by the enzymes NS3 and

NS5. Although the exact mechanism of coordination between both proteins is unknown, the interaction point has been mapped to the C-terminus of NS3 and the RdRp region of

NS5 (26,30). The synthesis of new viral RNA occurs in a semi-conservative and asymmetric manner involving distinct steps with varying levels of efficiency; the first step is the synthesis of a negative strand from the positive strand RNA that enters the cell with the virus particle and this step depends on the cyclization of RNA and is mediated by NS5.

The polymerase recognizes a structure within the 5’ UTR of RNA known as step loop A

(SLA), and this serves as a promoter to start RNA synthesis at the 3’ end of the genome.

The negative strand RNA remains base-paired with the positive strand template to form a double strand intermediate (31). After synthesis of the negative strand, the next step is synthesis of a new positive strand using the negative strand present on the double strand

RNA as a template. This step is initiated by NS3 which performs helicase function to unwind the double-stranded RNA into the respective positive and negative strands. Next, the NS5 RdRp region binds the 3’ end of the negative strand RNA via its template-binding channel and uses this as a template to generate new positive strand after which the double-

strand RNA now made up of the previous negative strand and a new positive strand will be unbound from the RdRp domain of NS5 and re-used for synthesis of more positive strand RNA copies, of which up to 10 new RNA strands are synthesized on an RF template and the overall process forms more positive sense RNA strands than negative sense (26). These RNA populations have been shown by radiolabeling and sedimentation experiments which identified three distinct populations; the first being full/genome length dsRNA, the second being a combination of double-strand regions with single-strand side likely undergoing synthesis of the nascent strand and the third being single-stranded RNA identical to the viral genome (2).

The next step after RNA synthesis is capping and methylation of the free positive strand RNA displaced from the double-strand intermediate by a nascent strand. A type 1 cap is added to the 5’ end of the RNA by an initial transfer of a guanosine mono- phosphate (GMP) moiety to the 5’ diphosphate of the positive strand followed by methylation at the N7 position of the guanine cap and the 2’-OH. Importantly, the SLA in the 5’ UTR is required for the N7 cap methylation (32). After RNA synthesis, the next stage in the infection cycle is the packaging of new virion particles by formation of the nucleocapsid and this requires two main processes: 1.) Dimerization of the C protein and its interaction with the ER membrane and lipid droplets through its hydrophobic face and simultaneous binding to the viral RNA in a non-specific manner (33) and 2.)

Heterodimerization of prM and E which is primarily mediated by prM and required for complete folding of E and formation of its antigenic conformation. This

heterodimerization allows prM/E to become arranged as spikes on the ER membrane opposite the RNA replication sites (33). This leads to new virus particles budding into the

ER lumen. Unfortunately, the kinetics of the virus budding process are very rapid and make it difficult to visualize this step. However, virus assembly primarily occurs at the ER except for a single account of budding at the plasma membrane obeserved in a strain of

WNV (7).

Overall, an immature virion is formed in the ER having an icosahedral arrangement. Virions are then translocated within vesicles to the Golgi (27). While moving through the Golgi, the prM is cleaved by host furin molecules and a neutral pH eventually causes dissociation of the pr peptide to yield a mature form of the virus (7). The mature virus expresses 90 E homodimers on its surface to mediate entry into another host cell. Interestingly, in addition to mature virions, relatively smaller flavivirus particles are released from infected cells and they possess E and M proteins but lack nucleocapsids and are hence, non-infectious (2).

Figure 1-1 Schematic representation of the flavivirus infection cycle initiated by viral attachment and endocytosis through to the egress of new virion particles. Pierson TC, Diamond MS. Degrees of maturity: the complex structure and biology of flaviviruses. Current Opinion in Virology. 2012 Apr;2(2):168–75. Copyright permission received.

Figure 1-2 A. Components of the flavivirus replication complex assembled on the ER membrane B. Schematic representation of the RNA synthesis (left) and capping (right) phases during flavivirus replication. Klema VJ, Padmanabhan R, Choi KH. Flaviviral Replication Complex: Coordination between RNA Synthesis and 5'-RNA Capping. Viruses. 2015 Aug;7(8):4640–56. Open access.

1.1.3 Antigenic structure of flaviviruses

Flaviviruses are phylogenetically grouped based on information from genome sequences and amino acid sequences of the viral proteins. With regard to antigenic structure, flaviviruses were previously grouped based on their cross-reactivity determined by using polyclonal sera in hemagglutinin inhibition assays (34). However, virus neutralization defines closely-related flaviviruses by grouping them into serocomplexes (35) such that viruses in different serocomplexes have only about 80% identity in the internal regions of their E proteins. Nonetheless, there exist antigenic variation even within respective serocomplexes (e.g. the E protein of the four DENV serotypes vary by up to 37%). Indeed, the antigenic reactivity of flaviviruses is greatly determined by the amino acid composition of the E protein given that this is the outermost exposed portion of the virus particle. These antigenic variations are particularly crucial in determining the population of cross-reactive neutralizing antibodies in the context of viral infection or re-infection. A good example of this phenomenon is demonstrated by all 4 serotypes of DENV being able to infect the same individual (24) likely owing to the differences in their antigenic profile.

On the other hand, up to 80% identity exists between JEV E and WNV E such that JEV vaccines conferred protection from WNV disease in a series of animal studies (36).

1.1.4 Flavivirus pathogenesis and clinical manifestation

Infection with flaviviruses occurs by the bite of an infected arthropod although there have been reports of transmission by the drinking of unpasteurized milk. Dendritic cells (DCs), keratinocytes and fibroblasts are the initial infected cells. Virus is transported within the

infected cells into the lymphatic system where viremia occurs and promotes the spread of virus to various secondary organs and to uninfected arthropods (37). Flaviviruses are broadly classified based on pathogenicity as being encephalitic or hemorrhagic. The encephalitic flaviviruses are able to transcend the blood-brain barrier (BBB) and proceed to infect neuronal cells in the central nervous system (CNS) through mechanisms that are not yet fully characterized. Various hypotheses have been explored to explain BBB crossing and these include entry via immune cells (38), transcytosis by vascular endothelial cells, inflammation, transport within monocytes (Trojan-horse process) and entry via the olfactory lobe (39). Using these different routes of entry, the virus arrives at the CNS and causes damage to neurons. Apoptotic cell death has been identified in vitro as a hallmark of cell damage to neurons infected with WNV. This phenotype was dependent on virus replication because infection with UV-treated WNV did not result in neuronal cell death (40). Interestingly, expression of WNV NS3 or C protein alone could induce apoptosis (41). Additionally, damage to neural cells occurs through by-stander effects from the release of TNF-alpha and reactive oxygen species (ROS). Studies to identify viral determinants of neurological disease have implicated the E protein in flaviviruses such as JEV for receptor interaction on target cells (39). Neuropathology and neuronal dysfunction resulting from CNS infection manifest as muscle weakness, tremors, paralysis, cognitive impairment, and in some cases, death (42).

On the other hand, hemorrhagic flaviviruses are primarily restricted to other peripheral organs including the liver, kidney, spleen etc. and cause multiple organ failure,

vascular permeability and severe bleeding. DENV has been widely studied as a hemorrhagic flavivirus with extensive damage to secondary organs when infection proceeds to severe stages. There are 4 antigenically distinct serotypes of DENV known in circulation (DENV1-4) (43) and no correlation has been drawn between serotype and disease severity (44). Survival of infection by one serotype confers lifelong immunity to that serotype but only partial immunity to other serotypes (21). Studies suggest that severe clinical cases of DENV pathogenesis are less common with primary infections and more prevalent during secondary infections with non-exposed serotypes (44). Infection with

DENV may be asymptomatic or may lead to a range of clinical presentations.

Symptomatic infection can result in dengue fever (DF), which is a brief ailment presenting with flu-like symptoms. In severe infection cases, the disease can progress to dengue hemorrhagic fever (DHF) which is associated with higher vascular permeability, spontaneous bleeding, thrombocytopenia, and plasma leakage. The most severe disease state of DENV infection is dengue shock syndrome (DSS) characterized by cold skin, undetectable blood pressure and ultimately, disease can be fatal (43). There exists some overlap in the tissue tropism between hemorrhagic and encephalitic flaviviruses because similar to encephalitic viruses, DENV is able to cross the blood-brain barrier and infect neurons (neurotropic) however, cases of DENV encephalitis (direct neuronal damage by the viruses) are relatively rare and have been reported in only 1-6% of dengue infection cases (45). Overall, infection with hemorrhagic flaviviruses is characterized by a productive infection of peripheral organs including the liver, spleen and kidney and

symptoms range from mild flu-like manifestations to spontaneous bleeding, effusion of fluid in abdominal cavities, thrombocytopenia and death (43).

A major similarity in the disease course of all pathogenic flaviviruses characterized till date is the biphasic nature of disease. The initial phase involves an influenza-like illness characterized by headache, fatigue, pyrexia and other non-specific symptoms, which occur anywhere between 3-28 days of exposure to the virus. This phase resolves to an asymptomatic phase lasting 2-7 days where the patient appears to have recovered; this can mark the end of clinical symptoms. However, in a subset of patients, a severe phase develops and the symptoms in this phase depend largely on the infecting agent. One pioneer member of the flavivirus group – YFV causes jaundice in the severe phase of infection. Additional severe symptoms include renal failure, melena, bleeding, hematemesis and extended shock. Death can occur by the 10th day of severe disease (46).

In contrast, severe infection with JEV is distinctly associated with a Parkinson’s disease- like syndrome in addition to seizures, cranial nerve palsies and motor dysfunction

(Solomon, 2002). WNV infection is characterized by a rash, meningitis, myocarditis and severe neurological symptoms (Solomon, 2002; (46). More recently, Zika virus infection has been associated with severe congenital deficits by virus interaction with neural progenitor cells resulting in impaired neurological development presenting as microcephaly. ZIKV infection also results in severe eye inflammation, potential blindness and fetal loss (47). Importantly, these clinical presentations with ZIKV are novel and have only recently been elucidated during the current outbreak compared to the mild flu-like disease previously associated with the virus. This further highlights the evolving nature of

flavivirus and emphasizes the need to understand determinants of disease and tissue tropism. Indeed, infection with members of the tick-borne encephalitis virus serocomplex follows a similar biphasic pattern and is further reviewed below. Only a subset of individuals progress to the severe phases of flavivirus infection, and the underlying determinants for disease severity have not been clearly defined (48).

1.1.5 Arthropod-borne flaviviruses

Flaviviruses are distinguished from other genera within the Flaviviridae family by their requirement for an arthropod vector for transmission. While there are some flaviviruses with no known vector (NKV), the vast majority of pathogenic flaviviruses characterized till date are either tick-borne flaviviruses (TBFV) or mosquito-borne flaviviruses (MBFV).

Phylogenetic analysis suggest evolutionary distinctions based on mutational changes whereby MBFV have twice the mutation rate of TBFV (49). Both the TBFV and MBFV groups contain viruses that cause encephalitis or hemorrhagic fever and are globally spread based on the biology of the transmitting vector. In nature, these arboviruses are maintained in an enzootic cycle involving the tick or mosquito vector in which they persist, and an intermediate reservoir host (50-52). It is estimated that 25% of the known flaviviruses are transmitted by ticks, particularly Ixodid ticks – I. scapularis and I. ricinus

(17). MBFV depend on mosquitoes in the Culex or Aedes genera for transmission wherein viruses transmitted by Culex mosquitoes utilize avian species as intermediate hosts (53).

The main difference in the spread and emergence of arthropod-borne flaviviruses can be linked to vector ecology, human migration, socio-economic activities and climatic

influences. Most MBFV have recorded a more aggressive global spread and disease incidence likely due to the involvement of birds as natural reservoirs for MBFVs in nature.

It is theorized that because birds can travel a longer distance, this increases the propensity for disease to spread (46). Additionally, human agricultural practices such as irrigation increase the density of mosquitoes breeding in warm, agricultural fields. While the interactions occurring between the virus and arthropod vector at the molecular levels are not fully understood, studies with DENV suggest that the capability of an arthropod vector to spread disease in a given location is influenced by the virus genotype (54). MBFV vectored by Aedes aegypti include DENV and YFV; however, despite the close genetic and vector similarities between both viruses, DENV demonstrates a wider spread likely due to higher replication in the mosquito vector. JEV and WNV are members of the

MBFV group transmitted by culex mosquitoes and other species that feed on birds, reptiles, horses and pigs (46). Initial isolation of WNV was from mosquitoes infecting horses and similar to humans, the horses are dead-end hosts of MBFV infection (55).

TBFV are primarily transmitted by ticks of Ixodes spp., particularly I. ricinus and I. scapularis. Viruses have also been detected in Dermacentor spp. including D. marginatus and D. silvarum; however human attacks and infectivity are lower with these species, hence Ixodid transmission remains widespread. The tick vector is often resident in moist forest regions and infection cases are higher in areas where the cities overlap forests at their boundaries (46). Small mammals such as rodents serve as virus reservoirs in nature wherein the virus is maintained and amplified. Larger mammals including deer and cattle are also infected by TBFV while humans are accidental hosts (56). Chances of human

infection are higher between March to November and up to 20% of ticks harbor viruses in endemic regions (56). Once infected, the tick harbors the virus for the rest of its life and virus can be transmitted within ticks transstadially, transovarially and between co-feeding ticks in close proximity on the same mammalian host (56,57). Tick infection can occur at any stage of the tick’s life cycle and the larval and nymph stages feed on small mammals while the adult ticks feed on other large animals (58). Within the tick, the virus is localized to various parts of the tick including the salivary gland from where it can be transmitted to the host. Studies have been performed to determine the role of tick salivary gland extracts (SGE) in transmission of tick-borne flaviviruses and results suggest that the

SGE contain bioactive molecules which in addition to promoting arthropod feeding, create a conducive environment at the site of the tick bite to allow viruses to survive and persist; this process is termed saliva-assisted transmission (SAT) (59). Additionally, SGE has been shown to modulate the host immune response particularly in dendritic cells, thereby allowing the virus to establish an infection (60,61) and virus can be detected in tick saliva for up to 120 days (59). Interestingly, recent animal studies by Hermance and

Thangamani, 2015 showed that the effect of SGE on virus infection is dose-dependent such that with higher infection doses of POWV, SGE had no impact on mouse survival, disease severity etc. However, at lower doses of infection, the presence of SGE promoted virus transmission and the outcome of disease. In this study, low viral titers were associated with absence of overt disease signs highlighting the relevance of virus accumulation to disease pathogenesis (62). During tick feeding, the mammal acquires the virus through the bite from the tick or contamination of an old bite wound with tick feces.

Virus spreads throughout the mammal thereby making it possible for virus to be transmitted horizontally to another feeding tick without the requirement for viremia, rather by taking advantage of the anti-inflammatory responses and immune cell recruitment to the site of the bite (57,63,64).

Figure 1-3 Overview of the enzootic cycle maintaining tick-borne encephalitic viruses in nature and the initial pathogenesis in humans at the site of arthropod contact. Růžek D, Dobler G, Donoso-Mantke O. Tick-borne encephalitis: pathogenesis and clinical implications. Travel Medicine and Infectious Disease. 2010 Jul;8(4):223–32. Copyright permission received.

1.1.6 Epidemiology of flaviviruses

Flaviviruses are present on every continent of the world and virus distribution in different regions is closely-linked to the prevalence of the transmitting arthropod vector. Many pathogenic flaviviruses are considered emerging pathogens owing to their occurrence in new regions where they were not previously associated with disease. The MBFV group include the viruses with the widest recorded spread of which JEV causes the highest number of encephalitis cases in central and south-east Asia. Annually, up to 50,000 infection cases are reported with about 20-30% of severe cases resulting in death (39).

Although there have been successful vaccine schemes to prevent JEV, proper implementation of the vaccination programs is impeded by financial constraints in low income regions. DENV is a widespread hemorrhagic flavivirus and the most important arbovirus with an estimate of 400 million infections occurring annually (39,65) About 2.5 billion people are at risk of DENV infection and majority of this population is in Asia

(37). Up to 500,000 cases of DHF occur annually (39) and there is currently no treatment for DENV infections. WNV is another emerging arbovirus which has been associated with outbreaks in the past, particularly related to its appearance in New York during the summer of 1999 resulting in numerous cases and recorded death of humans and livestock.

Historically, WNV has been endemic in Africa, Europe and Asia prior to its entrance and spread in North America (66,67). Infection is more prevalent in elderly individuals; about

30,000 cases have been reported in the US till date and 4% (1200) of those have been fatal. There is no vaccine to prevent WNV infection and like many other flaviviruses, treatment is limited to supportive care (39). TBEV is mostly endemic in Russia, Germany,

United Kingdom, and Australia. It is considered an emerging pathogen owing to its occurrence in new regions such as the Austrian Alps (57). Despite the availability of a vaccine, up to 2500 cases occur annually in Europe (56). A primary hypothesis to explain virus spread is the impact of climate change on the tick vector leading to lower infection incidence during milder winters, however, this notion is under debate (56)

1.1.7 Tick-borne encephalitis virus serocomplex

The TBEV serocomplex group includes closely antigenic related flaviviruses transmitted by ticks and have varying impacts and outcome of infection. The degree of virulence is demonstrated in the biosafety level (BSL) required for laboratory research and of relevance to the study described herein. Members of the serocomplex include TBEV

(BSL4), POWV I (BSL3), POWV II (BSL3), LGTV (BSL2) and LIV (BSL2). TBEV is the most-studied representative of this group and most of the progress made in understanding this pathogen will likely prove beneficial in elucidating the biology of other members in the serocomplex. Pathogenesis of TBEV is highly similar to other flaviviruses beginning with the bite of an infected arthropod (tick), although about 30% of patients do not recall a tick bite hence these largely remain unreported (56). The initial febrile illness can resolve or transcend to a severe phase and the proportion of severe infections varies with the infecting subtype. Up to 60% of severe infections are fatal with the far eastern subtype (FE-TBEV). Reports show that severity increases with the age of the patient and typically involves severe neurological complications. In understanding encephalitic

TB(FVs), two main concepts apply; 1.) Neuroinvasiveness which is the invasion of the

CNS by the infecting pathogen and 2.) Neurovirulence which is a direct infection of neural cells in the CNS and high virus titers in peripheral secondary organs is crucial for crossing the BBB. Of the hypothesis to elucidate the mechanism of BBB crossing, TBEV likely transcend the BBB by transcytosis through cells of the vascular endothelium. Once infection progresses to severe phases, various clinical manifestations occur including encephalitis characterized by decreased consciousness, stupor, hyperkinesia of facial muscles, ataxia, delirium, psychosis, convulsion, possible coma and death (56,68).

Infection can have prolonged neurological impact on patients and likely impacts the quality of life. Similarly, POWV infection is associated with meningioencephalitis, encephalitis, acute meningitis and prolonged neurological impact in patients who survive disease including memory deficits and severe headaches (69). POWV is a more recently characterized member of this group and the only tick-borne flavivirus primarily endemic in North America (70). Langat virus is a naturally attenuated member of the TBEV serocomplex and was previously tested as a vaccine candidate but found to cause encephalitis in 1 out of 12,000 vaccine recipients (3,71).

Figure 1-4 Phylogeny tree of related flaviviruses with host and geographical spread Gould EA, Solomon T. Pathogenic flaviviruses. The Lancet. 2008 Feb;371(9611):500–9 Copyright permission received.

1.2 Innate Immune response to flaviviruses

Flavivirus infection of a cell initiates a cascade of responses that constitute an attempt by the host cell to block virus progression in the infection cycle. These responses are rapid and closely linked to the route of virus infection and is executed as an initial wave of gene stimulation which is further amplified to alert bystander cells and establish an antiviral state. As outlined above, entry of flaviviruses into a cell occurs by receptor-mediated endocytosis (3) and virus particles are internalized into the cell within clathrin-coated pits.

Ultimately, membrane fusion allows dissociation of the nucleocapsid and subsequent release of viral RNA into the host cytoplasm (2,72). The exposed viral RNA serves as a pathogen-associated molecular pattern (PAMP) that can be recognized and responded to by cellular pattern recognition receptors (PRRs). These PRRs are ubiquitously expressed cell factors which identify conserved molecular signatures shared by a group of pathogens to elicit downstream signaling and also serve as a bridge between innate and adaptive immunity by activating DCs. There exist a number of PRR families within the cell but the families of Toll-like receptors - TLRs (73) and retinoic acid-inducible gene I (RIG-I)- like receptors –RLRs (74) particularly elicit signaling in response to flavivirus infection.

1.2.1 Interferon induction pathway

Interferon (IFN) signaling is the first line of defense launched against invading flaviviruses and ultimately serves to protect cells from further virus amplification. The

IFN signaling cascade is ideally primed very early in infection to modify gene expression and create an antiviral environment refractory to virus replication (75). The importance of

IFN has been elucidated in increased susceptibility of IFN-incompetent cells to flavivirus infection. The IFN signaling cascade occurs in two main phases wherein the first phase involves signaling events that lead to transcriptionally-mediated release of IFN and the second phase involves signaling by IFN to upregulate other antiviral genes (76). The IFN induction pathway involves the sum-total of events that occur in the cell in response to external stimulus resulting in the production of type 1 IFNs in the stimulated cell. IFN- beta (IFNb) is initially produced during the first wave of induction while subsequent signaling produces higher levels of IFN-alpha (IFNa) (77). This first round of signaling in response to flavivirus infection is primarily mediated by the RLR and TLR family of pattern recognition receptors within the infecting cell.

1.2.1.1 RIG-I-like receptors

The RLR family comprises three members namely, RIG-I the pioneer member, MDA5 and LGP2. They are DExD/H box RNA helicases localized to the cytoplasm of various cell types where they are expressed at low levels under unstimulated conditions, and expression increases in response to virus infection or IFN treatment (78). They are shown to recognize non-self RNA and different members of this group are preferentially activated by specific RNA signatures. In vivo studies performed by Kato et al., 2006 to elucidate the functions of RIG-I and MDA5 showed that mice lacking RIG-I are able to induce IFNa and IFNb when challenged with poly(I:C) while MDA-5- deficient mice did not produce IFNs when exposed to poly(I:C) suggesting that MDA5 expression is necessary for the cellular recognition of poly(I:C). On the other hand, challenging cells

with dsRNA synthesized in vitro required RIG-I expression for the induction of IFNb while MDA-5 deficiency had no impact on IFNb production in response to the synthesized RNA (79). These data indicated the relative specificity in the RNA populations identified by these two proteins. While RIG-I and MDA5 function in parallel and are relatively specific for their viral substrates (77), they are both required for the recognition of PAMPs associated with flaviviruses during infection (78,80,81) except in the case of JEV which RIG-I expression is sufficient to produce IFNb (79). Structurally,

RIG-I and MDA5 are similar in having three domains namely, the N terminal region containing capsize activation and recruitment domains (CARD), the center portion containing the RNA helicase domain and the C-terminal region which bears the repressor domain (RD) in RIG-I, while MDA5 has a CTD domain similar to the RD (78). LGP2 lacks a CARD domain and is hypothesized to serve as a regulator of RIG-I and MDA5

(82). The RD domain of RIG-I particularly serves a repressor function to self-regulate the protein such that in resting state, RIG-I assumes a closed conformation whereby the

CARD domains are not exposed. However, upon recognition of short, blunt non-self dsRNA fragments, or ssRNA within the cell, RIG-I becomes activated and the CARD domains are released to interact with the substrate –RIG-I specifically binds RNA with terminal 5’ triphosphate groups (5’ ppp) although a single phosphate is sufficient for activation, triphosphate is necessary for full signaling (78,83). RIG-I also recognizes other secondary PAMPS such as polyuridine as non-self features to induce IFN signaling.

Crystal structure resolutions have shown that RIG-I interacts with RNA through its CTD while it can bind host proteins via CARD-CARD interactions. It is important to note that

for optimal activity of RIG-I, it must be ubiquitinated by the host E3 ubiquitin ligases

TRIM25 and Riplet which help to stabilize it (84,85). In their active state, RLRs multimerize and activate the adaptor mitochondrial antiviral signaling protein (MAVS) with which they form a complex through the CARD-CARD interactions (82). MAVS is a membrane protein found in association with the mitochondria therefore, complex formation of RLRs with MAVS helps to recruit them to the mitochondria (73,86,87)and is required for IFN signaling by bridging RNA sensing of RIG-I with downstream signaling cascades. Additionally, MAVS activation causes it to dimerize and recruit other host proteins to form the signalosome including TRAF2, TRAF3, TRAF6 and TANK which further activate TBK1 and IKKe and these in turn phosphorylate members of the IRF family IRF3 and IRF7 that homodimerize, translocate to the nucleus and promote IFN signaling. On the other hand, MAVS mediates activation of IKKa which further phosphorylates IKKb to allow the release of NFkB which then translocates to the nucleus and mediate upregulation of type 1 interferons. Recent studies showed that altered MAVS expression is associated with increased neuroinvasion by TBEV (88,89) thereby highlighting the relevance of MAVS in the RLR signaling pathway.

1.2.1.2 Toll-like receptors

These were the initial PRRs identified and have been shown to recognize molecular signatures of a wide range of pathogens. There are 10 and 12 members of the TLR family in humans and mice respectively and these are differentially localized to various parts of the cell (73). TLRs that recognize nucleic acids are typically localized within endosomal

compartments and other intracellular vesicles such as within lysosome, the ER, and endolysosomes (90,91).This localization puts them in an important position to recognize viral RNA released within these compartments and also keeps them from cytoplasmic locations where “self” nucleic acids are likely present (90). TLRs recognizing nucleic acids include TLR3, 7, 8 and 9. Structurally, TLRs consist of a transmembrane domain, an ectodomain and toll-IL-1 (TIR) domain on the cytoplasmic side. In order to become active, TLRs are first processed by cellular cathepsins (91). The signaling response elicited by TLRs is highly dependent on the infecting pathogen and upon virus recognition, signaling is ultimately mediated by the adaptor proteins recruited of which

MyD88 is a key adaptor protein utilized by many TLRs although TLR3 and 4 utilize the adaptor protein TRIF. Indeed, TLRs represent one of the first lines of defense against flaviviruses and TLR3 primarily recognizes WNV and DENV. Signaling of TLR3 through the adaptor protein TRIF involves the binding of TRIF to TRAF3 and TRAF6 while

TRAF3 signaling ultimately leading to the phosphorylation of IRF3, TRAF6 signaling initiates dephosphorylation of IKB through AP-1 and eventual activation of NFkB.

Together, IRF3, AP-1 and NFkB are translocated to the nucleus to activate IFN signaling.

Importantly, the synergistic function between RIG-I, MDA5 and TLR3 further limits flavivirus infection in a cell culture model (92).

1.2.2 Interferon response pathway

Overall, the events occurring in the IFN induction pathway are aimed at releasing IFN out of the infected cell to prime neighboring cells and amplify the response through the IFN

response pathway. The IFN response pathway is initiated from the production of IFN in the induction pathway and involves signaling cascades that occur through the Type 1 IFN receptor (76). Secreted IFN binds to the Type 1 IFN receptor (IFNAR1) on the same host cell (autocrine) or on adjacent host cells (paracrine) to elicit signaling by the Janus kinases

(Jak) and Signal transducers and activators of interferon (STATs).

1.2.2.1 Jak-STAT signaling

This constitutes the cascade of downstream events triggered from IFN binding to the type

1 IFN receptors on the surface of a cell. The major players in this pathway belong to the

Jak and STAT family of proteins. The Jak family comprises 3 members, Jak 1-3 and Tyk2 of which Jak1 and Tyke 2 particularly function in the antiviral signaling pathway through the Type 1 IFN receptors staying in stable association with IFNAR2 and IFNAR1 respectively, and becoming activated by a phosphorylation event (93). On the other hand,

STATs exist in an inactive state in an unstimulated cell, and this family comprises about 9 members in mammals (93). STAT1 and STAT2 particularly mediate the antiviral response in the IFN induction pathway. Activation occurs by phosphorylation on a carboxyl- terminal tyrosine residue leading to their homo- or heterodimerization and nuclear translocation (94). In the nucleus, STAT proteins are able to bind specific sequences in the promoter region of their target genes. The recognition sites could either be gamma activated sites (GAS) or the IFN-stimulated response element (ISRE). Generally, STAT proteins demonstrate distinct activation modes in response to cell stimulus and this determines the nature of the response. For instance, during IFN-gamma signaling, STAT1

forms a homodimer to mediate this response. However, IFN-beta signaling leads to the formation of a heterodimer by STAT1 and STAT2. This heterodimeric complex is further recruited to DNA by a third protein, IRF9 to form the ISGF3 transcription factor, which leads to the transcriptional control of multiple target genes (94).

1.2.2.2 Interferon stimulated genes

JAK-STAT signaling ultimately leads to the induction of over 200 interferon-stimulated genes (ISGs) in response to virus infection. The produced ISGs are the main effectors of the antiviral response by performing a range of functions that render the host cell environment inhospitable for virus replication (81,95). It is worth noting that a subset of

ISGs can be expressed in a cell independently of IFN production and this occurs through activation of proteins belonging to the IFN regulatory factor (IRF) group (94). When virus is detected by TLR signaling, IRF-3/7 can mediate the production of ISGs whereas the

IFN response pathway uses IRF-9 to upregulate ISGs. The specific antiviral function of many ISGs remain to be determined in great mechanistic detail; however, many ISGs have been identified as viral restriction factors that specifically restrict viral proliferation and pathogenesis (93).

Figure 1-5 Overview of the IFN signaling pathway in response to dsRNA intermediates. IFN released further activates signaling via the JAK-STAT pathway to produce ISGs and amplify the IFN response. Haller O, Kochs G, Weber F. The interferon response circuit: induction and suppression by pathogenic viruses. Virology. 2006 Jan;344(1):119–30. Open access.

1.3. Flavivirus restriction factors

There is a myriad of defenses elicited against flaviviruses during cellular infection.

Various signaling pathways trigger the expression of antiviral genes which encode for proteins to disrupt the virus infection cycle. Viruses likely exert a selective pressure on the host resulting in evolution of proteins with multiple functions including the restriction of virus infection. Restriction factors (RFs) are defined as fast evolving cellular proteins that function as part of the innate immune system to prevent virus replication (96). While

many restriction factors are multipurpose to a cell by having other endogenous homeostatic functions, their primary biological roles involve antiviral activity. Antiviral genes utilize various means including post-translational modifications to increase coding potential, which determines the function and specificity of host factors (97). The action of host restriction factors against flaviviruses is closely linked to their route and mechanism of infection (98,99).

The relevance of further studying restriction factors to specific viral entities cannot be overemphasized. Understanding the role of host antiviral factors in recognizing and preventing virus replication will provide insight on correlates of host protection.

Moreover, the nature of antiviral responses likely shapes the viral counter-response strategies used to promote virus pathogenesis (96,100).

1.3.1. Identification of restriction factors

Several studies have probed for host restriction factors through a variety of screens and pull-down assays. Human and murine siRNA libraries have proven valuable for the discovery of host factors pertinent to the virus replication process. Viral protein interactome studies have also helped to show specific factors that directly bind viral proteins and serve as RF candidates. In many cases, further mechanistic studies have been used to show the impact of a given host factor on virus infection by performing knockdown, overexpression and biochemical studies. Common trends shown by well- characterized RFs suggest that for a gene product to be identified as a viral restriction factor, three main criteria should hold true:

1. Gene expression of restriction factors is likely increased during virus infection

– Various screens have identified differentially expressed genes during infection

with flaviviruses. In 2007, Fink et al, performed microarray studies in human cell

lines infected with DENV-2 as well as human patient samples and identified that

the genes with highest expression belonged to 3 main clusters of genes involved in

the NFkB signaling pathway, the IFN response pathway, and the proteasome-

ubiquitin pathway. The most prominent genes in this study were IFIT1, 2 and 3,

OAS1, 2 and 3, OASL, STAT1 and 2, IP-10, MX-1, viperin and USP15 (101).This

study demonstrates how gene expression studies represent the first step towards

identifying antiviral factors.

2. Virus replication is decreased in the presence of RFs; inversely, when the

restriction factor is absent, there is higher virus replication. Many studies have

followed-up on the gene expression studies by selecting the genes with the highest

expression and testing their impact on virus replication. Vast progress in the field

of restriction factors has come from molecular experiments to deplete or

overexpress specific genes and determine their role in virus replication. For

example, in the study by Fink et al., the authors went further to test the role of

viperin by performing overexpression studies and demonstrated that when viperin

is highly expressed in human cells, there is significantly less DENV replication –

thus highlighting viperin as a potential flavivirus restriction factor (101).

Additionally, the use of high throughput overexpression studies has proven very

essential to identifying restriction factors to flaviviruses. An overexpression study

by Schoggins et al in 2011 particularly shed some light on potential flavivirus

restriction factors. In this study, the authors collated microarray data from the

literature and used the information to test the impact of 380 human ISGs on

diverse viruses including WNV and YFV both belonging to the flavivirus group.

This screen used a lentiviral-based overexpression system and identified antiviral

genes against flaviviruses having broad and narrow effector functions (102). The

main anti-flavivirus genes identified in this study were IFITM3, IRF1, IFI6 and

RTP4, which inhibited YFV replication in a dose-dependent manner. IRF1, HPSE

and MDA5 expression also had an adverse effect on WNV replication.

Importantly, the authors tested the combinatorial effect of ISGs and demonstrated

an additive effect on virus restriction when ISGs were co-expressed. On the other

hand, Yasunaga et al, (2014) performed a high throughput genome-wide RNAi

screen in Drosophila to identify restriction factors of WNV on the premise that the

same genes might restrict a broader group of arthropod-borne viruses. Indeed, the

study identified 50 genes whose abrogation led to higher WNV replication.

Seventeen of these genes also had antiviral activity against other (flavi)viruses

including WNV-KUNV and DENV (103). Importantly, 86% of the identified

drosophila genes have human orthologs and are all conserved in mosquitoes,

which are the arthropod vectors for the flaviviruses tested (103).

3. The restriction factor specifically interacts with the viral proteins or genome –

In addition to being virus-induced and preventing virus replication, a specific

restriction factor would ideally bind to a viral protein or genomic material. Several

proteomic screens have probed the interactions between host and flavivirus

proteins. Overall, interaction proteomics are relevant for determining factors that

prevent optimal progression of the virus through the infection cycle and restriction

mechanisms that occur at the molecular level (104). Le Breton et al performed a

Yeast 2 hybrid (Y2H) screen in 2011 using the NS3 and NS5 proteins from Kunjin

virus (KUNV), Japanese encephalitis virus (JEV), DENV, WNV, and TBEV. The

screen identified 108 human proteins that bound to flavivirus NS5 or NS3. Some

interacting proteins identified in this study include, but are not limited to STAT2,

TRAF4, TRIM21, IFNAR2, TYK2 and HSP90AB1 (105). Most interacting

proteins were overrepresented in the human interactome. In the same year, Khadka

et al., screened all 10 DENV proteins for their interacting partners in the human

liver and identified 139 interactions between DENV proteins and human proteins.

Forty-five of these interacting proteins have been linked with the proliferation of

other viruses and belong to known RF families (106). Examples include –

TRIM62, STAT2 and serpin. More recently, Miariang et al, performed another

Y2H screen in 2011 to identify interactions of DENV proteins with both mosquito

and human proteins. This study again identified interactions between DENV

proteins and members of the TRIM, HSP, STAT and Serpin group of proteins.

Many orthologous interactions were also identified in mosquitoes suggesting that

the interactions are conserved across species. The authors went a step further to

confirm the interactions by co-affinity purification (Co-AP) assays (106). In 2011,

Taylor et al., identified TRIM79 as a TBEV-specific restriction factor through a

Y2H screen. This study identified an interaction between TBEV NS5 and the

murine protein TRIM79. Co-AP studies also confirmed the interaction hence

implicating TRIM79 as a potential restriction factor (107). As described below, our

group has identified a homolog of TRIM79 in Peromyscus leucopus, the North-

American natural reservoir of TBFV. TRIM79 from P. leucopus retains the ability

to co-localize and interact with TBFV NS5; thereby suggesting that the restriction

factor function of TRIM79 is conserved across species.

Overall, we have gained clearer understanding of how restriction factors function against flaviviruses in recent years. However, many questions remain unanswered regarding how information gathered from restriction factor studies could be applied into the rational design of therapeutics and further understanding of viral pathogenesis pertaining to host and tissue tropism. These remain the subject of ongoing studies and focus in the field. It is noteworthy that large-scale approaches to identifying relevant ISGs has certain pitfalls and challenges which were recently highlighted in a review by Lazear and Diamond (2015).

The main points bordered on the lack of necessary host interactions when ISGs are expressed individually, the variation in ISG population of a given cell type and the use of virulent strains in screens which are capable of antagonizing the IFN response and masking certain ISGs. These limitations can likely explain the narrow overlaps observed in the results of various screens performed over time to identify anti-flavivirus factors

(108). Furthermore, while many ISGs and immune signaling proteins have been associated with a protective role during flavivirus infection, there have been other reports of a

proviral effect observed with host factors otherwise expected to restrict replication. This is potentially due to host factors being co-opted by viral proteins. Manocha et al., 2014, using a human cell model, identified TRIM21 to be a negative regulator of IFN-beta signaling during JEV infection (109). Our group has also identified a component of the

TLR signaling pathway - TRAF6 to function as a proviral factor for TBEV, and this occurs through interaction with the viral NS3 protease (Youseff et al., unpublished data).

Therefore, this possibility that some immune signaling genes could perform proviral roles is important in interpreting the results of virus interaction screens.

1.3.2. Key flavivirus restriction factors

Based on the various screens performed to identify anti-flavivirus factors, certain genes have been recurrent over the years and have been studied in detail for their impact on flavivirus infection. In this section, we will provide an overview of the key players in the antiviral response to flaviviruses by discussing their mode of action, specificity and their potential role in species tropism. It is worth noting that this is not an exhaustive list but a snapshot of the main antiviral families.

1.3.2.1 STATs

This group of cytoplasmic transcription factors form an important part of the antiviral signaling cascade as well as other cellular signaling pathways necessary for optimal cell homeostasis. STAT proteins exist in a latent state prior to cell stimulation and when stimulated by virus infection and IFN signaling, can mediate a potent host response. As indicated, the IFNβ signaling response to flaviviruses is crucial for controlling virus

infection and pathogenesis and STAT proteins play a central role in this signaling cascade.

Although less dominant than STAT1 and STAT2, studies have shown that STAT3 and

STAT4 also become phosphorylated during type 1 IFN signaling (94). Studies using gene targeting tools and transgenic mice have highlighted the role of IFN-dependent JAK-

STAT signaling pathways in the control of virus infection. STAT1 signaling is particularly key to the IFN response due to its ability to form complexes that bind both GAS and ISRE elements thus making it important for the induction of other downstream transcription factors such as the IFN regulatory factor 1 (IRF1). STAT1 functions in a positive feedback loop to amplify production of type 1 IFN during virus infection and mutant STAT1 mouse models have shown greater susceptibility to virus infection (94). Lazear et al., 2015 demonstrated that IFN-gamma signaling diminished the movement of WNV across the blood-brain barrier (BBB) by making controlling tight junction proteins in a manner that was dependent on STAT1 (110). The importance of STATs as restriction factors is emphasized by the fact that STAT proteins are targeted by different flaviviruses as part of their pathogenic strategy (reviewed in later sections) thus, the ability of flaviviruses to antagonize IFN signaling by targeting STATs greatly affects cell permissiveness to infection.

1.3.2.2 TRIMs

Tripartite motif (TRIM) proteins are so named because contain an RBCC domain made up of three parts including a RING domain, a B Box and a coiled-coil domain at the N terminus (111). The B-box domain typically contains a zinc finger motif and is

structurally similar to the RING domain. Most TRIM proteins also have a C-terminal domain which is variable and functions in multimerization and binding of substrates (112).

The RING domain functions in protein-protein interactions and typically confers E3 ubiquitin-ligase activity on the TRIM protein. In humans, the TRIM family comprises about 77 members, which have been increasingly recognized as a component of the innate immune response to viruses. In addition to being E3 ubiquitin ligases, TRIM proteins can also specifically interact with various cellular and viral proteins to regulate transcription, intracellular trafficking, membrane repair, etc. (113). Many members of the TRIM family were initially found in screens to identify interferon-induced genes. The role of TRIM proteins as antiviral restriction factors was first highlighted with TRIM5-alpha which was recognized as an anti-retroviral factor, preventing uncoating of the human immunodeficiency virus (HIV). Since the discovery and extensive characterization of

TRIM5, more attention has been paid to TRIM proteins and their potential roles as antiviral effectors against other groups of DNA and RNA viruses (113). Evidence for RF function in some TRIM proteins is supported by their IFN-dependent induction during virus infection, trends of positive selection identified by genetic analysis and their direct interaction with viral proteins. (114-116). Many TRIM proteins have a broad target range, for example, TRIM5 has since been shown to have antiviral effector functions against other Retroviruses (96,117). Studies have shown that members of the TRIM family regulate PRR signaling pathways through K63 (non-proteolytic) linked ubiquitination.

This exemplifies broad-spectrum effects of TRIM proteins on the cell (85). In 2007, Gack et al., demonstrated that RIG-I is ubiquitinated at its N-terminus by TRIM25 through a

K63 linkage. In this study, the authors reported that upon virus infection, TRIM25 binds to

RIG-I, facilitates its K63-linked ubiquitination and thereby allows it to interact with

MAVS to promote PRR signaling and induce type 1 interferons (84,118). On the other hand, TRIM proteins can regulate the pro-inflammatory response by the addition of K48- linked (proteolytic) ubiquitin chains to substrate proteins (118,119). In addition to generally modulating the immune response, several TRIM proteins function as antiviral restriction factors by their direct interaction with viral proteins to inhibit virus infection.

This restriction function occurs at various stages of virus infection and typically requires a direct interaction between the TRIM protein and a specific viral protein. Due to the need for direct binding, the binding sites of restriction factors and viral proteins show evidence of positive selection (118,120). Various TRIM proteins have been identified to perform specific antiviral function against flaviviruses. Liu et al., in 2014 (121) demonstrated that

TRIM56 exerts an antiviral impact on YFV and DENV, both members of the flavivirus group. The effect of TRIM56 was mediated by its E3 ubiquitin ligase activity and required an intact C-terminal domain. The authors demonstrated that TRIM56 restriction did not occur through its impact on overall IFN signaling, rather, TRIM56 negatively affected the accumulation of flavivirus RNA thus leading to reduced virus release (121).

As indicated, our group identified another TRIM protein (TRIM79) through a Y2H screen with TBEV NS5. TRIM79 is an ISG, analogous to TRIM30D and functions as a

TBEV restriction factor by directly binding to the viral polymerase and targeting it for degradation through the lysosome. Interestingly, TRIM79 did not have the same effect on

the closely related WNV thereby highlighting the tight specificity of certain restriction factors (107). Several other TRIM proteins have been observed in screens assessing interaction between flavivirus and host proteins. However, many of them remain to be characterized and their roles could be very broad, encompassing many cellular processes or very specific by targeting particular viruses and viral products .

1.3.2.3 OASs

The 2, 5-oligoadenylate synthetase (OAS) is an antiviral protein induced by IFN signaling and plays a role in establishing the antiviral state in response to viral ssRNA or dsDNA.

The OAS family includes OAS1, OAS2, OAS3, and OAS-like genes in humans, whereas eight small Oas1 (Oas1a-h), one Oas2, one Oas3, and two Oas-like (OasL1 and OasL2) genes are expressed in mice (122). Of the eight murine Oas1 gene isoforms, Oas1b has been implicated as a flavivirus resistance determinant wherein a truncation in the OAS1b protein arising from a mutant oas1b gene serves as a determinant for flavivirus susceptibility (123). Interestingly, the majority of inbred laboratory mice strains do not express a functional Oas1b gene (124). OAS enzymes catalyze the oligomerization of

ATP to 2,5-linked oligoadenylate (2-5A) which further converts the downstream effector

RNase L from a latent form into an enzymatically active form. RNase L functions as an antiviral factor by degrading viral and host cell RNA. Studies by Samuel et al., 2006 showed that RNase L restricted the spread and replication of WNV in the periphery and

CNS respectively (125). However, Oas1b can exert antiviral impact on WNV in the

absence of RNase L as shown in mouse RNaseL knockdown cells (126). In cell culture, expression of WT Oas1b results in impaired replication of WNV (123).

Yoshii et al., showed that differential susceptibility of two Far-Eastern TBEV strains to

Oas1b resulted in variable neurovirulence of the virus strains in congenic mice expressing a functional Oas1b gene – Infection with the Oshima strain led to minimal symptoms whereas the Sofjin-HO strain led to severe neurological disease and death in the mice.

These data suggest that even in the presence of a functional restriction factor, certain viruses can overcome restriction by antagonism and/or alternative mechanisms of pathogenesis (reviewed in later sections). On the other hand, the role of Oas genes in flavivirus disease in humans has not been clearly elucidated. Studies performed in human cell model demonstrated that while donors expressing two copies of a less functional

OAS1 variant showed higher rate of virus infection, the presence of symptoms and disease severity were not influenced (127). Cellular localization of host factors likely influences their antiviral effects by influencing their binding partners. Courtney et al., 2012 showed for the first time that full length Oas1b contains a C-terminal transmembrane domain, which localizes it to the ER membrane. The authors further performed a Y2H screen using an Oas1b transmembrane mutant and identified two major binding partners of Oas1b. This study characterized ABCF3 as an Oas1b-interacting protein, which likely mediates the previously elucidated anti-flavivirus effects of Oas1b. Targeting ABCF3 by RNAi resulted in higher replication of WNV. Interestingly, this restrictive impact was only seen in cells expressing Oas1b suggesting that ABCF3 requires Oas1b for its activity (128). These

studies demonstrate how the inquiry into the mechanistic function and binding partners of one restriction factor can lead to the identification of other relevant restriction factors.

1.3.2.4 IFITs and IFITMs

The IFN-induced protein with tetratricopeptide repeats (IFIT) family and the IFN-induced transmembrane protein family (IFITM) family have been studied as restriction factors to flaviviruses particularly WNV and DENV. The IFIT and IFITM gene families are unrelated in humans but are similar in being induced by IFN signaling and virus infection.

The human IFIT family comprises four members: IFIT1 (ISG56), IFIT2 (ISG54), IFIT3

(ISG60 or IFIT4), and IFIT5 (ISG58) while the murine ifit family includes three members

- Ifit1 (Isg56), Ifit2 (Isg54), and Ifit3 (Isg49) (129).

IFITs are constitutively expressed at basal levels within a cell however, they can be induced by stimuli from RNA viruses which are recognized by TLRs or RLRs to initiate a downstream signaling cascade. Activated transcription factors further translocate to the nucleus and bind to the ISRE in the IFIT promoters (130). IFIT genes are also induced directly by IFN treatment independent of viral infection. The IFIT proteins are localized to the cytoplasm and are not enzymatically active. Although they all contain a helix-turn-helix motif known as the tetratricopeptide repeat (TPR), there is limited sequence conservation between the different IFIT members and this could explain their functional variability. IFIT proteins have been recognized as important restriction factors to flaviviruses. When IFIT2 expression is abrogated in dendritic cells and cerebellar

neurons, there is a significantly higher increase in WNV replication compared to cells expressing IFIT2 (131).

Functionally, IFIT1 and IFIT2 inhibit mRNA translation by binding to the eukaryotic translation factor 3 (elf3). Both proteins can recognize and bind to mRNA lacking 2’-O methylation in their 5’ cap – this is a distinguishing factor in some viral mRNAs since cellular mRNAs are 2’-O methylated. Binding of the IFIT protein thus outcompetes for elf4e binding and prevents translation of viral RNA (95,132).

The IFITM family consists of five members in humans and six members in mice.

IFITM1 is mainly expressed at the surface of cells while IFITM2 and 3 are expressed intracellularly. These proteins are structurally related but have variable antiviral targets and have only recently been appreciated for their role in flavivirus restriction(133-135).

Overall, IFITMs are induced by IFN and restrict flaviviruses at the entry stages of the viral infection cycle, particularly preventing virion fusion at the early (WNV) and late (DENV) endosomal compartments (134). One proposed mechanism of function is that IFITMs decrease the endosomal fluidity thereby reducing the fusion potential and subsequently impacting the pH of the endosome to prevent membrane fusion (136,137). Additionally, In vivo genetic manipulation studies have demonstrated the importance of IFITM proteins

(138).

1.3.2.5 IRFs

IFN regulatory factors are a group of transcription factors that have been widely studied for their role in modulating both the innate and adaptive immune response to pathogens

(139). A notable member of this group, IRF-3 is a cytosolic protein ubiquitously expressed in all cells thus making it unique from other IRF proteins, It has been described as a master transcription factor necessary to activate the immune response to pathogens (140).

Phosphorylation of IRF-3 is triggered by dsRNA structures occurring during replication of

RNA viruses, or the accumulation of viral proteins. Following phosphorylation, IRF-3 dimerizes and interacts with its co-activator to induce expression of specific genes, one of them being IFN-b. As stated above, secreted IFN-b can further activate the JAK-STAT signaling pathway to induce the expression of ISGs. Therefore, IRF-3 is essential for virus-induced IFN signaling through its complex with other proteins including NFkB and c-Jun (141). Daffis, 2007 demonstrated the inhibitory role of IRF-3 in WNV replication and showed that this occurred by both dependence and independence on IFN signaling

(140). Importantly, TBEV antagonism of IRF-1 lends credence to its function as a virus restriction factor (142).

1.3.2 Probing restriction factors in relevant hosts

Many of the known antiviral factors have been identified and studied in human and murine models of infection; however, these species are largely susceptible to infection despite the expression of restriction factors. On the other hand, many intermediate mammalian reservoir hosts of flaviviruses become infected in the wild but remain relatively disease-free. It is quite possible therefore, that there are inherent differences in the structural and effector properties of restriction factors between the resistant and susceptible species. Hence, uncovering these differences will further provide insight into

the flavivirus restriction pathways. A classic example of variable susceptibility among strains can be found with the discovery that alleles of the Oas1b gene (originally named

Flv) controlled susceptibility to disease by flaviviruses. Perelygin 2002, demonstrated that

Inbred laboratory mouse strains are susceptible to flaviviruses (Flvs) compared to their wild counterparts (Flvr). The resistance determinant was narrowed down to the Oas1b gene being expressed as a truncated form in the susceptible species (128,143,144). This study highlights the importance of comparing the molecular differences between resistant and susceptible species as a means to identify the genetic determinants of disease. Given the close genetic similarity between susceptible and resistant species, it is possible that similar antiviral genes are expressed; however, the task lies in identifying subtle sequence differences that could result in functional variation and more potent restriction in the resistant host. Cell culture models to recapitulate resistance of natural hosts and probe for the genetic determinants of disease are essential in this regard.

1.3.3 Restriction factors and drug development

Theoretically, the lessons learned from the study of restriction factors can be applied into the design of therapeutics to mimic the mechanism of action in blocking virus proliferation. The use of peptide mimetics would be particularly beneficial to bind and prevent the function of viral proteins. Furthermore, the design of mutant viruses lacking or harboring binding sites for proviral and antiviral factors respectively will greatly contribute to the vaccination options available to prevent flavivirus infections.

Additionally, a case is to be made for tick-borne flaviviruses and the need for further

restriction factor studies in this group of flaviviruses. There is a current focus on mosquito-borne flaviviruses and this is largely due to the great burden of disease contributed by this group and the increasing epidemiological concern. However, there has been a steady increase in the incidence of tick-borne flaviviruses in recent years (145).

The emerging nature of these viruses calls for closer attention to ensure preparedness and availability of information that can applied to the design of therapeutics in the event that

TBFV incidence continues to increase.

Figure 1-6 Summary of obstacles to identifying new restriction factors by performing large screens. More targeted approaches will be beneficial to defining functional restriction factors. Lazear, H.M and Diamond, M.S. (2015) New insights into innate immune restriction of West Nile virus infection. Current Opinion in Virology, 11:1-6 (100). Copyright permission received.

1.4. Viral Antagonism of the IFN response

Under ideal conditions, all of the immune response processes launched by the host against invading flaviviruses would eliminate infection and combat any instance of disease.

However, as described above infection with flaviviruses can cause severe disease and potentially lead to death. Multiple lines of evidence have shown that viruses devise numerous means to circumvent host immune responses and this notion was first described in Influenza virus. Immune antagonism involves evasion of host detection pathways, direct inhibition of detection or direct antagonism of other secondary host restriction mechanisms.

1.4.1 Flavivirus immune evasion strategies

Immune evasion by flaviviruses is initiated early on in infection whereby the virus hides from host detection by remaining within vesicles and thus delaying IFN induction

(146).The virus replicates within these vesicles surrounded by cellular membranes known as viroplasm-like structures (VLS) to hide the dsRNA from host RLRs and retard the production of IFN (147). Indeed, infection with WNV is associated with delayed PRR activation thereby allowing the virus more time to establish high numbers at early time points post infection (148). A similar phenotype has also been observed with JEV indicating that members of MBFV and TBFV group utilize this technique to evade host responses. Furthermore, flaviviruses are able to modify their RNA to avoid expressing the specific molecular signatures identified by host PRRs. Many flaviviruses encode 2’ O methyltransferases which modify their RNA cap to resemble the host cap structure and

hence evade detection and restriction by antiviral factors. Indeed, blocking this 2 ‘O methyltransferase activity in DENV and JEV gave rise to attenuated viruses (147). In addition to escaping detection by PRRs, flaviviruses also inhibit PRR function through various mechanisms. Expression of WNV NS1 resulted in failure of NFkB to translocate to the nucleus thereby blocking IFN induction in a cell-type dependent manner (148).

Flavivirus sera which is derived from the 3’ UTR of the virus promotes viral pathogenesis and blocks RIG-I activation by preventing TRIM25-mediated ubiquitination (147,149).

DENV NS2B/3 also prevents RIG-I translocation to the mitochondria by competitively binding the chaperone 14-3-3e (150). Another strategy used by flaviviruses is to inhibit the interacting partners of PRR molecules for example, the E protein of WNV blocks ubiquitination of RIP-1 which is required for signaling in the TLR and RLR pathways.

The IFN induction pathway is also antagonized downstream of PRR activation for example, TBK1 activation is blocked by DENV NS4A while NS2B/3 blocked the enzymatic activity of IKKe (151). As highlighted above, signaling by IFN is triggered via binding to the cognate receptors to elicit signaling and upregulate antiviral gene. This pathway is also antagonized by flaviviruses by blocking the post-translational modifications necessary for signaling or by blocking expression of key signaling factors.

The NS4B protein efficiently blocks phosphorylation of JAK1 and Tyk2 which is necessary for their activation thereby consequently inhibiting STAT1 activation.

Conversely, WNV inhibits IFNAR signaling by upregulating the inhibitors of this pathway namely suppressors of cytokine signaling (SOCS1 and SOCS 3) (148). Most of the work characterizing IFN antagonism in flaviviruses has been completed using

members of the mosquito-borne group; while many similarities likely exist in the mechanism of host response antagonism, further work is necessary to clearly elucidate the specific processes used by TBFVs to evade immune responses.

1.4.2. Flavivirus NS5 antagonist mechanism

The NS5 protein is the largest and most-conserved protein among the flaviviruses which performs multiple functions to promote virus pathogenesis. Central to this study is the function of NS5 and a potent antagonist of host IFN responses, particularly in the IFN response pathway occurring through IFNAR signaling. The mechanism of antagonism by flavivirus NS5 varies from one strain to another but the overall outcome is diminished IFN signaling in infected cells to promote virus replication. LGTV NS5 was the first flavivirus

NS5 to be characterized as an antagonist of the IFN cascade by preventing phosphorylation of STAT1, its nuclear localization and role in gene expression; this is achieved by indirectly blocking the kinases Jak1 and Tyk2 through diminished expression of IFNAR1 on the surface of infected cells (71,152). Further mechanistic studies revealed that this function of LGTV NS5 was mediated through interaction with a host protein prolidase (153). WNV NS5 functions similar to TBEV NS5 as infection with WNV but not JEV, decreases IFNAR1 expression on the surface of infected cells. Interestingly,

Kunjin virus (KUNV) the less virulent variant of WNV was unable to effectively inhibit

IFN signaling (154). Furthermore, DENV NS5 degrades STAT2 in human cells and its close relative ZIKV has recently been elucidated to perform a similar function. Ectopic expression of DENV NS5 results in degradation of STAT1 and suppression of ISG

expression. Interestingly, when NS5 was expressed by itself, it only blocked the phosphorylation of STAT2; however, when expressed in the context of a polyprotein requiring an initial cleavage step at the N terminus of NS5 within cells, it led to STAT2 degradation, importantly, this function of NS5 was mediated by UBR4 (155-157).

1.4.3. Role of antagonism in species tropism

Given the importance of viral antagonism to pathogenesis and the finding that viruses with impaired immune antagonist function are antagonized in the host, the species spectrum is likely impacted by the ability of an infecting strain to circumvent the immune response and successfully establish an infection. A classic example is observed with the DENV

NS5 antagonism described above. While DENV NS5 is a potent antagonist of IFN signaling in humans via its adverse effect on STAT2 signaling, the viral protein is unable to perform this function in mouse cells thereby making immunocompetent mice largely resistant to virus infection. This phenomenon has contributed to the lack of a complete animal model for the study of DENV infection and could partially explain why humans are the primary reservoirs for DENV in nature. Similarly, the DENV NS2B/3 protease antagonizes IFN signaling in humans by cleaving the host adaptor protein, stimulator of the IFN gene (STING) which normally functions as part of the RLR pathway by interacting with RIG-I and MAVS to induce IFN signaling (158). Cleavage of STING by the viral protease is essential for virus replication and expression of STING proteins lacking this cleavage site impaired virus replication in human cells (158). However,

DENV NS3B/3 is unable to cleave the mouse STING protein thereby contributing to the

inability of DENV to successfully replicate and cause disease in mice (158,159). Further studies will be necessary to elucidate the range of virus restriction in different mammalian species to distinguish between lack of proviral and antiviral mechanisms of disease resistance.

1.5. Viral control of reactive oxygen species

During normal cellular processes or pathogenic invasion, partial reduction of oxygen occurs, leading to the formation of highly active by-products named reactive oxygen species (ROS) including superoxide and nitric oxide. When present at low levels, these

ROS are beneficial to the cell and mediate signaling cascades to regulate transcription, cytokine expression, immune responses and apoptosis (160). However, when produced in excess, ROS can be detrimental to the cell. Owing to their signaling potential, increasing evidence shows that viruses can modulate the level of ROS during infection for various benefits such that antioxidant treatment diminishes virus infection even though the same

ROS molecules are known to negatively impact other groups of pathogens, such as bacteria (161). This phenomenon has been characterized in diverse virus families since the initial observation was made with Sendai virus as an inducer of oxidative stress in mouse cells (160,162). Infection with the respiratory syncytial virus (RSV) could be relieved by treatment with antioxidants (163) and this was similar to observations made with hepatitis

C virus whereby treatment with antioxidant pine extract Pycnogenol led to decreased infection (164). Of relevance to our research focus is the finding that distinct flaviviruses have been implicated as manipulators of cellular ROS to induce oxidative stress. Example

include DENV which was shown to induce oxidative stress and cell death during infection such that treatment with an oxidase inhibitor led to decreased hemorrhage severity (Yen,

2008). This control varies with the age of the cell source such that monocyte cultures derived from neonates and elderly individuals show higher oxidative stress in response to

DENV infection compared to adult-derived cells (165). Another flavivirus JEV has been demonstrated manipulate redox levels during infection and antioxidants are likely antiviral candidates to combat JEV (166). More recently, Kuzmenko et al., 2016 showed that

TBEV NS1 protein activates the antioxidant pathway and is associated with altered oxidative responses in human cells thus suggesting that the virus attempts to manipulate

ROS levels within a cell (167). An interesting study by Gullberg, 2015 showed that

KUNV virus infection increases ROS levels in cells to induce oxidative stress and blocking ROS during infection led to lower virus release from the cell. The authors further showed that treating virus-infected cells with antioxidant led to a change in the RNA strand accumulation (positive-to-negative strand ratio) and impaired flavivirus RNA capping thereby linking oxidative stress to RNA capping and replication (168).

The rationale for viruses to manipulate oxidation levels within a cell was unclear for some time particularly because ROS were not appreciated as signaling molecules due to their perceived lack of specificity. However, further studies have revealed that ROS can indeed activate multiple cellular pathways; unlike the typical ligand-receptor interactions which occur at the macromolecular level to activate signaling pathways, ROS signal at the atomic level via chemical reactions triggered by sensing of redox signatures (169). In

addition to generation of ROS through biological processes particularly from the mitochondria, cells also generate ROS specifically as an enzymatic defense mechanism within phagosomes in response to pathogenic insults. When ROS is produced in excess, the cell triggers cell death pathways in an attempt to eliminate the oxidative stress and also activates the NFκB pathway at various signaling points (170). This interaction can influence the ability of NFκB to activate downstream signaling pathways. Hence, viruses appear to manipulate ROS signaling in an attempt to co-opt NFκB for their replication goal (160). Additionally, reducing the antioxidant levels during infection likely impairs the function of immune cells which require higher amounts of antioxidants (160) and this concept has been worked out in an HIV model of infection where disease progression and immune cell depletion are associated with lower antioxidant levels (171). During DENV infection, oxidative stress controls the immune response and apoptosis, hence contributing to possible rationale for ROS control by viruses (172).

1.6. Reservoir species

Vector-borne flaviviruses are zoonotic and harbored in animals for extended periods of time where they are continuously cycled between mammals and arthropod vectors. TBFV are primarily maintained in rodents and other small mammals such as hares. Studies have shown virus persistence in cell culture using eukaryotic cells and cells of the arthropod vector. Virus can also be detected in animals within endemic areas (17). There has been evidence of viral persistence in humans for long periods and this is observed to a large extent with TBFV infections (173,174).

1.6.1 Viral maintenance in nature

As highlighted in the clinical presentation of virus infections, patients recovering from severe acute illness suffer for extended periods of time in a post-encephalitic syndrome

(175). Additionally, viruses have persisted in symptomatic humans long enough to be transmitted to infected patients through blood transfusion (176,177) In the tick vector, surveillance studies have shown persistence of TBFV in diverse species and varying degrees of tick infections have been identified at various locations. The prevalence of

TBFV surveyed in Russia shows that up to 40% of ticks were infected while prevalence is about 5% in Europe (178). Within the tick vector, virus disseminate from the gut where it is initially harbored for extended periods within the blood meal. Blood meal is digested intracellularly and insect tissues are perfused in hemolymph within the hemocoel allowing the spread of viruses to various organs within a tick vector (17,179). Overall, the tick remains infected with the virus for the rest of its life, thus contributing to persistence.

Similarly, MBFV is disseminated in the mosquito vector and virus titers increase within the midgut of the mosquito (180). Persistence of virus in the mosquito vector is higher doing low temperatures and flaviviruses have persisted in Culex mosquitoes for up to hundred days during the winter season (181). Importantly, MBFVs apply diverse mechanisms to circumvent the mosquito immune response and promote resistance in the vector, thereby increasing the propensity of transmission to secondary hosts (182).

Zoonotic flaviviruses also persist in vertebrate hosts including small mammals and birds while larger animals and humans remain inadvertent hosts (17,183-185). Species of rodents which serve as reservoirs for tick-borne viruses include bank voles (Modes

glarcolus), striped skunks (Mephitis mephitis), yellow-necked mouse (Apodemus flavicolis), white-footed mouse (Peromyscus leucopus) and red-backed voles (Myodes rotilus) in which viruses have been detected through various surveillance studies (17).

Furthermore, transmission of TBFV through consumption of milk suggests that the virus can persist in large animals including goats and sheep for extended periods (186-188).

Essentially, infection of these vertebrate hosts allows virus to remain available for transmission to ticks during a blood meal; indeed the lack of viremia requirement for transmission allows a broader range of hosts to serve as reservoirs irrespective of whether viremia occurs during infection (189). While TBFVs primarily persist in rodent hosts, some MBFVs persist in birds and pigs. For example, antibodies to JEV have been recovered in pigeons over one year after infection while WNV was found hundred days after infection (190). It is important to note that many species of birds do succumb to disease following infection with some MBFV such as WNV. However, within the short time of infection, the infected birds can travel wide distances thereby promoting virus dissemination and increasing transmission to other mosquito vectors. Mechanistically, it is hypothesized that persistence is accomplished by viral evasion and antagonism of host responses and persistence likely varies in different posts based on protein interactions as outlined in previous sections. Also, mathematical modeling by Rosa, 2003 suggest a potential role of various transmission pathways in establishing persistent infection of TBD including non-viremic transmission rate of host encounter etc. (191).

1.6.2. Disease resistance in reservoir species

A reservoir has been defined as a population wherein the pathogens can be maintained for distinct periods and from which infection is transmitted to susceptible populations (192).

The terms tolerance and resistance have been distinctly used to describe the mechanisms underlying the absence of disease such that resistance involves an active inhibition of disease pathogenesis while tolerance refers to the passive ability of the host to withstand the negative impacts of infection (193,194). As described above, viruses persist in natural hosts for a long period of time however, many natural hosts do not demonstrate overt signs of disease despite being infected with potentially lethal viruses. Although there have been reports of symptomatology in natural hosts of MBFV, the vast majority of TBFV reservoirs harbor viruses without developing disease. The main underlying mechanisms of resistance in natural hosts have not been uncovered and this concept poses a major puzzle in the overall study of emerging zoonotic viruses. Several hypothesis have attempted to explain the absence of disease in reservoir species and these include lack of cytopathic effect induced by virus in the natural reservoir, differential tissue tropism, distinct protein- protein or nucleic acid interactions, symbiotic nature of the virus-host relationship with the reservoir hosts, increased potency of the immune response and disease tolerance in the reservoir species (194). As described, TBEV persists in mammals and it is estimated that about 100 animals are infected with TBEV (195)). Several studies have detected TBEV in wild rodents across various continents such as the 2008 study by Kim et al., which assayed lungs and spleens of rodents Apodemus agarius and wild boars by qRT-PCR and detected the western subtype of TBEV in South Korea, even though TBEV infections had

not been reported in that region at this time. (196). Interestingly, the TBEV strains isolated from neighboring countries Japan and China were the far eastern TBEV subtype such as the detection of FE-TBEV in the spleens of rodents Apodemus speciosus and

Clethronymys rufocanus in Japan by Takeda et al, in 1999 (197). This demonstrates reservoir strain preferences by viral subtypes at various geographical locations.

Additionally, Frey et al., 2013 sequenced the genome of a TBEV strain isolated from the brain of the yellow-necked mouse (Apodemus flavicollis) in Austria (198). POWV antibodies have also been detected in groundhogs, opossums and striped skunks through isolation of feeding Ixodid ticks thereby implicating them as potential intermediate hosts

(199). It is important to note that virus detection in wild hosts has been from relatively healthy animals, thus suggesting that these animals withstand infection and remain alive despite the lethal potential of infecting viruses. Additionally, feeding patterns of infected ticks in endemic regions have helped to identify potential reservoir species of TBFV.

While most of the information regarding clinical resistance in reservoir hosts has been mostly anecdotal, few laboratory studies have been performed to corroborate these indications and demonstrated that virus infection of true reservoir hosts does not lead to disease symptoms. Egyed, 2016 reported that infection of Apodemus agarius with resulted in an asymptomatic infection although the virus could replicate within the brain and the rodents showed seroconversion (200). Similarly, A 1997 study by Telford et al., reported asymptomatic infection in P. leucopus inoculated via the intracranial route with POWV

(201). However, this study did not examine in-depth viral pathogenesis occurring in P. leucopus rather, a recent study has further characterized the outcome of P. leucopus

infection with POWV and shown that virus infection in these species did not result in lethality while control C57BL/6 and BALB/C mice mostly succumbed to infection. In this study, virus was restricted to specific regions of the brain and did not diffuse as seen in the susceptible species (202). Further mechanistic studies are required to elucidate the underlying processes of disease resistance likely occurring due to molecular interactions between the host and virus. Till date, there is no specific treatment to resolve infection with flaviviruses, yet some reservoir hosts appear to withstand infection effectively therefore, identifying the genetic determinants of disease in these hosts will likely provide insight for the design of therapeutics.

1.6.3. Peromyscus leucopus as reservoir species of Powassan virus

Peromyscus spp comprises the largest rodent group in North America of about 56 members dispersed in every habitat along the coast of the USA and parts of Canada with the most abundant members being P. leucopus and the closely-related P. maniculatus

(203). Although physically similar to members of the Mus genus, they differ genetically and evolutionarily. Owing to their distinct breeding and ecological features (204).

Members of this genus have been extensively used as non-traditional models to study metabolism, aging, reproduction, behavioral evolution, epigenetics, tumorigenesis, and pathogenesis of infectious disease (203,205,206). A particularly attractive feature is that they are adaptable to laboratory conditions and their genetic diversity recapitulate that of their wild counterparts making them ideal representatives for the study of human diseases

(204,207,208). P. leucopus is naturally infected with a variety of bacterial pathogens

(Borrelia, Anaplasma, and Rickettsia and other viruses (hantaviruses) without succumbing to infection (209). Of relevance to this study is that fact that P. leucopus is also infected with POWV and in-depth pathogenesis studies have established resistance to neurological effects of POWV infection (202,210-212). Thus, P. leucopus likely functions as a reservoir host for POWV wherein the virus is unable to cause disease but is simply maintained at low levels, enough to be transmitted to ticks during a blood meal. This resistant phenotype indeed satisfies the criteria for the definition of a host as a reservoir specie, since it must remain relatively fit and available for tick feeding (209). Peromyscus mice differ significantly from both the Mus and Rattus genera, having diverged an estimated 22 million years ago (213,214). Laboratory mice (Mus musculus) are useful in the modeling of flavivirus encephalitis, including POWV, as they are susceptible to infection and recapitulate many aspects of human disease, including the identity of infected cells and organs, induction of inflammatory responses and damage of the central nervous system (215). This is in contrast to wild-caught and outbred P. leucopus mice experimentally infected with POWV showing no signs of encephalitis or death [18] and this dichotomy of infection outcomes between M. musculus and P. leucopus infected with

POWV provides an ideal opportunity to examine the mechanism of restriction of TBFVs in P. leucopus hosts. Broadly, the use extensive use of Peromyscus species as models of human disease has been hampered by paucity of genetic information and laboratory reagents (216). Specific molecular studies to delineate the genetic determinants of reservoir host protection from POWV also do not exist because of the absence of a well- defined cell culture model of TBFV restriction.

Overall, the natural resistance of P. leucopus to TBFV infection forms the basis of the study described herein. The goal of this study is to elucidate the molecular determinants of virus resistance in reservoir species using P. leucopus as a model. As highlighted above, this rodent species harbors numerous pathogens in the wild, yet the mechanism of resistance is not understood. This dissertation project therefore takes a systematic approach to identify the point of virus restriction during the infection cycle, explores viral antagonism in the reservoir species and assesses the relative contribution of the host innate response to TBFV restriction in P. leucopus cells. Findings of this study form the basis for further probe into the immune system of P. leucopus and offers a potential blue-print for the study of other pathogen-reservoir pairs to elucidate resistance mechanisms.

Chapter 2

Materials and Methods

2.1 Cell culture and reagents

Human embryonic kidney (HEK) 293, HEK 293T, Vero, P. leucopus adult skin fibroblast cells (AG22353, Peromyscus Genetic Stock Center (PCSC), University of South Carolina) and M. musculus (C57BL/6) murine embryonic fibroblast (MEF) cells (217) were grown in Dulbecco’s modified enrichment medium (DMEM) containing 10% (vol/vol) fetal bovine serum (FBS, Gibco) 100 units/ml penicillin, and 100 mg/ml streptomycin (Thermo

Fisher Scientific) in an atmosphere of 5% CO2 at 37°C. Peromyscus embryonic fibroblasts (PEFs) were generated using our previous described protocol (217). Tissues for cell isolation were purchased from the PCSC and produced under IACUC protocol #2162-

100829-0.

2.2 Plasmids and vectors

Entry vectors used in this study were pENTR-D-TOPO (Thermo Fisher Scientific) and pENTR-U6 (Thermo Fisher Scientific). Langat virus (LGTV) nonstructural (NS) protein 5

(NS5) was derived by PCR amplification using the LGTV E5 infectious cDNA clone as template (provided by Dr. A. Pletnev, NIAID, NIH) as previously described (218). The

PCR product was directionally cloned into the entry vector by gateway cloning and further recombined into the lentiviral expression plasmid pLVU-GFP (Addgene #24177) with the vector backbone pLenti6/UbC/V5-Dest. To design lentiviral plasmids for gene

knockdown, plenti-X2-hygro-DEST (Thermo Fisher Scientific) was used as a destination vector.

2.3 Antibodies

The following antibodies were used: a-actin (A5441, Sigma); a-tubulin (sc-12462, Santa

Cruz); a-GFP (JL-8, 632381, Clontech); a-dsRed (632496, Clontech); a-V5 (R960-25,

Invitrogen); a-STAT1 (9172S, Cell Signaling Technology); a-STAT1-P (9167S, Cell

Signaling Technology); a-LGTV E and NS1 (provided by Dr. C. Schmaljohn,

USAMRIID); a-LGTV NS3) (217); a-IFIT2 and a-IFIT3 (provided by Dr. S.

Chattopadhyay, University of Toledo).

2.4 Viruses and infections

The following viruses were used in this study: LGTV (TP21 strain, provided by Dr. A.

Pletnev, NIAID, NIH), vesicular stomatitis virus (VSV, strain Indiana, provided by Dr. I.

Novella, University of Toledo), tick-borne encephalitis virus (TBEV, strain Sofjin, provided by Dr. M. Holbrook, NIAID, NIH), Kunjin virus (KUNV), POWV lineage I

(POWV-I, strain LB) and POWV lineage II (POWV-II, strain Spooner) provided by Dr. S.

M. Best, NIAID, NIH, and Sendai virus (SeV, strain Cantell; Charles River Laboratories).

Virus working stocks were propagated and titrated by immunofocus assay (LGTV) or plaque assay (TBEV, POWV, and VSV) on Vero cells. All procedures with POWV were performed under biosafety level-3 (BSL-3) conditions; procedures with TBEV were

performed under BSL-4 conditions at the Rocky Mountain Laboratories Integrated

Research Facility (Hamilton, MT). Multiplicity of infection (MOI) is represented as focus forming units (FFU) or plaque forming units (PFU) per cell.

2.5 Virus titration by immunofocus assay and plaque assay

For flavivirus quantitation, test cells were set up at 1 X 105 per well in 24-well dishes and infected with the indicated MOI. Viral supernatant was collected at various time points post infection. The virus was titrated by performing 10-fold dilutions of the supernatants and infecting Vero cells (2 X 105/well) with 250 µl of diluted virus stocks. After 1h adsorption period, the inoculum was removed and the cells were overlaid with growth medium containing 0.8% methylcellulose (w/v) and 2% (vol/vol) FBS (Gibco). At 4-days post infection, the infected Vero cells were washed twice with PBS and fixed with 100% methanol for 30 min. at room temperature (RT). Plates were washed twice with PBS and then blocked with OptiMEM for 30 min. at RT. Cells were then incubated with virus- specific antibodies (a-LGTV E for LGTV) for 1 h at 37˚C. The plates were washed twice with PBS and incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (Dako) in OptiMEM for 1 h at 37˚C. Cells were washed twice with PBS and the FFU were visualized with freshly prepared peroxidase solution containing 0.4 mg/ml

3,3’-diaminobenzidine (Sigma) and 0.0135% hydrogen peroxide in PBS. Plaque assays to quantitate VSV were performed as with the immunofocus assays with the exception that following a 48 h incubation, the cells were washed twice in PBS and fixed with crystal violet (0.8% in ethanol). To quantitate POWV and TBEV, modified plaque assays were

performed using 1.5% carboxymethylcellulose sodium salt (Sigma Aldrich) as an overlay and plates were fixed at 4 days’ post-titration with 10% formaldehyde and further stained with crystal violet (0.8% in ethanol)

2.6 Immunofluorescence confocal microscopy

Cells were plated in Lab-Tek II chamber slides (Thermo Fisher Scientific) and prepared by washing twice with PBS before fixing with 4% paraformaldehyde for 20 min. Cells were washed 3 times with PBST (PBS, 0.5% Tween-20) and then incubated with permeabilization buffer (0.1% Triton X-100, 0.1% sodium citrate) for 5 min. followed by incubation with blocking solution (PBS, 0.5% BSA, 1% goat serum) for 1h at RT. Cells were then incubated with primary antibody for 1h at RT depending on the gene of interest.

After incubation, with primary antibody, the cells were washed 3 times with PBST and incubated with secondary antibodies conjugated to Alexa-Fluor 488 or Alexa-Fluor 594

(Thermo Fisher Scientific) for 1 h. Secondary antibodies used corresponded to the primary antibody origin. To counterstain already-fluorescent cells, similar experiments were performed using anti-GFP primary antibodies. Slides were washed 3 times with

PBST in the dark and overlaid on glass coverslips using Prolong Gold + DAPI mounting media (Thermo Fisher Scientific). Stained cells were visualized using an Olympus confocal microscope (Olympus Fluoview FV1000) and images were analyzed on the

Fluoview software.

2.7 Western blotting

Cells were washed twice in PBS and lysed in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl, 150 mM NaCl, 0.1 % SDS, 1% NP-40, 0.5% Na-deoxycholate) with complete protease inhibitor cocktail (Roche). Cell lysates were treated with Turbo

DNase (Thermo Fisher Scientific) and cellular debris was removed by centrifugation

(2,000 x g for 5 min) while the supernatant was reserved. Equal amounts (10-30 µg) of protein were loaded on a 10% polyacrylamide gel and resolved by electrophoresis.

Protein was transferred on to a nitrocellulose membrane using the iBlot 2 Gel Transfer

Device (Thermo Fisher Scientific). Membranes were probed with specific primary antibodies overnight at 4˚C followed by a secondary incubation with goat anti-mouse IgG or goat anti-rabbit IgG antibodies (Thermo Fisher Scientific) for 1h at RT.

Immunoreactive proteins were detected using ECL Plus Western chemiluminescent system (Thermo Fisher Scientific) and exposed to film.

2.8 P. leucopus gene sequencing

A search for the genes of interest was performed from a RNAseq database of P. leucopus cells from Dr. J. Munshi-South (219). All the hits obtained for each gene sequence were aligned with the sequence of the corresponding P. maniculatus gene (STAT1-

XM_006974860, RIGI-I-XM_006975324, IFNAR1-XM_006983579, MAVS-

XM_006984557, IRF1-XM_006995899, TRIM79-XM_006507549.2). The consensus sequence from each alignment was used as a template for the design of cloning primers.

Further, cDNA was purified from P. leucopus cells and probed using the gene-specific

primers (available upon request). The PCR product was resolved on an agarose gel and further cloned into the entry vector pENTR-D-TOPO. DNA sequencing was performed on entry clones and PCR products using internal cloning and sequencing primers for complete resolution.

2.9 Quantitative reverse transcriptase polymerase chain reaction (RT-qPCR)

Total RNA was purified from cultured cells using the RNeasy Mini kit (Qiagen) with

DNase 1 digestion step (Qiagen). Resultant RNA was reversed transcribed using the

QuantiTect reverse transcription kit (Qiagen). The cDNA was used as a template for

SYBR green-based qPCR using the FastStart essential DNA GreenMaster (Roche) according to the manufacturer’s protocol. IFNAR1 Fwd –

CTGGAGACCACTCGGATAAATG, Rev- CTCGTACCCGGAGAAAGAAAG ; STAT1

Fwd – GAGAGAAACTTCTGGGTCCTAAC, Rev – GATCCAAGGCCAGAAGGAAA ;

MAVS Fwd- GTCTTCCTCTTCCACTGGATTG, Rev –

GTCACAGAATTGGTGGGTACTT ; RIG-I Fwd – GGTTCTGAAACTTGCTTTGGAG,

Rev - GCAGCTTTACTTTCAACCCTTT ; IRF1 Fwd –

CAGCACCAGCGATCTGTATAA, Rev – TTCCTTCCTCGTCCTCATCT; b-actin

(Fwd-CACACTGTGCCATCTATGA, Rev- GGATCTTCATGAGGTAGTCTGTC). All reactions were performed in triplicate in the Roche LightCycler® 96 instrument and analyzed with the LightCycler® 480 Software, Version 1.5. Results were normalized to mRNA levels of b-actin. For relative quantification in stimulated cells, results are expressed as a fold change relative to RNA samples from mock-infected, unstimulated

cells using the comparative threshold cycle method. To assess viral RNA in virus-infected cells, absolute quantifications were performed using the relative standard curve method generated from 10-fold serial dilutions (109-100 genome copies) of the LGTV E5 infectious clone cDNA. Positive strand RNA of LGTV was assessed with primers previously described (220) using forward primer LGTV911F

(GGATTGTTGCCCAGGATTCTC) and reverse primer LGTV991R

(TTCCAGGTGGGTGCATCTC) and normalized to mRNA levels of b-actin from either

P. leucopus or M. musculus (Fwd-GCAAGCAGGAGTACGATGAG, Rev-

CCATGCCAATGTTGTCTT). Similarly, negative sense RNA of LGTV was quantified using forward primer-GTCTCCGGTTGCAGGACTGT and reverse primer–

CTCGGTCAGTAGGATGGTGTTG [67]. To test viral RNA release after infection, viral supernatants were collected at various time points post infection. Viral RNA was purified from the supernatant using the QIAamp Viral RNA Mini Kit (Qiagen). The resultant RNA was normalized and used as template for cDNA synthesis. The cDNA obtained was further analyzed by qPCR using primers specific for the LGTV positive strand RNA as described above.

2.10 Antiviral assay

The test cells were infected with LGTV (MOI 1 and 10) or SeV (600 hemagglutinin (HA) units) for the indicated period of time. Viral supernatant was collected and added to fresh cells in 18 2-fold serial dilutions. As a control, cells were treated with 18 2-fold dilutions of mouse IFN-b (mIFN-b, PBL Assay Science). Treated cells were infected at 16 h post-

treatment with VSV (MOI 0.1). At 36 h post-infection, cells were washed twice with PBS and stained with crystal violet solution. Plates were evaluated by measuring the dilution at which 50% virus inhibition occurred and measured based on the mIFN-b standard dilution plate to extrapolate the level of IFN responsiveness in the test cells.

2.11 Viral entry assays

Cells were plated in Lab-Tek II chamber slides (Thermo Fisher Scientific) or culture flasks at 1 X 105 cells and infected with LGTV at varying MOIs. Cells were incubated at

4˚C for 1 h and then shifted to 37˚C for an additional hour. The cells were then washed 3 times in ice cold PBS followed by an acid glycine wash (137 mM NaCl, 5 mM KCl, 0.49 mM MgCl2.6H2O, 0.68 mM CaCl2.2H2O, 99.84 mM glycine, pH 2.0) to remove unbound virus. Slides were fixed and stained according to the immunofluorescence protocol.

Assays were quantitated by visually counting the number of virus-positive cells stained with a-LGTV E. Cells grown in culture flasks were harvested for purification of total

RNA and subsequent analysis by RT-qPCR.

2.12 Infectious clone technology

Full-length plasmid DNA p61-E5 corresponding to the LGTV strain E5 (provided by Dr.

A. Pletnev, NIAID, NIH) was digested for 4 h at 37 ˚C. The linearized cDNA was cleaned up by incubating with 3M sodium acetate (NaCO3CO2) and 100% ethanol. Clean DNA pellet was recovered into solution and used as a template for in vitro transcription by incubating for 2 h at 40˚C with the transcription cocktail (18.75 mM rATP, 18.75 mM

rCTP, 18.75 mM rUTP, 3.75 mM rGTP, 15 mM m7G cap analog, SP6 polymerase,

RNAse OUT, 100 mM DTT) and 1 µg of DNA. The resultant RNA was treated with

DNAse to remove DNA contamination. Cells were transfected with viral RNA using

Lipofectamine 3000 (Thermo Fisher Scientific) according to the manufacturer’s recommendation. Transfected cells were fixed and probed for viral protein staining using a-LGTV E and visualized by immunofluorescence confocal microscopy.

2.13 Generation of lentiviruses for gene knock-down or overexpression

Short hairpin RNA (shRNA) for targeted gene knockdown was generated using the Block-

IT RNAi Knockdown System (Thermo Fisher Scientific). Oligonucleotides targeting various regions of the genes of interest were annealed and the double stranded oligonucleotides were cloned into the pENTR-U6 entry vector and subsequently sequenced. Primer sequences will be available upon request. The entry vectors were further recombined into the pLenti-X2-hygro-DEST using the Gateway LR enzyme

(Thermo Fisher Scientific). To recover infectious lentiviruses, the shRNA-containing vectors were transfected into HEK 293T cells along with the ViraPower Lentiviral

Packaging Mix (Thermo Fisher Scientific). At 24 h post-transfection, cells were treated with 10 mM sodium butyrate and incubated for additional 48 h before lentiviruses were harvested from the cell supernatant and cell debris was removed by centrifugation (2,000 x g for 5 min). Lentivirus production was confirmed and semi-quantitated using the Lenti-

X GoStix (Clontech). To make lentiviruses for overexpression, the gene of interest (LGTV

NS5) was cloned into the lentiviral expression plasmid (pLVU-GFP) and the resultant

expression vector was transfected into HEK 293T cells for virus packaging and generation of lentiviruses.

2.14 Generation of cell lines

P. leucopus fibroblasts were transduced with lentiviruses expressing the gene of interest, control marker gene, shRNA for targeted gene knockdown or control shRNA. Cells were concurrently treated with polybrene at 6 µg/ml. At 24 h post-transduction, cells were re- transduced for an additional 6 h before replacing inoculum with complete culture medium.

At 48 h post-transduction, cells were incubated in culture medium containing the appropriate drug selection: 10 µg/ml blasticidin (pLVU-GFP) or 400 µg/ml hygromycin B

(pLENTI-X2-DEST) (Thermo Fisher Scientific). After passaging in selection for an additional 7-10 days, cells were plated for isolation in 48 well cluster dishes. Single clones were then expanded into individual culture flasks and tested for the gene of interest/knockdown phenotype by RT-qPCR and western blotting.

2.15 Construction of ectopic expression vectors

TRIM79 was detected in IFN-treated P. leucopus cells and PCR-amplified as described above. The PCR product with the most abundant band by agarose gel was cloned into the pENTR/SD/D-TOPO (Invitrogen) entry vector to generate entry clones which were further sequenced to determine sequence accuracy. The entry clone was directionally recombined into various gateway destination vectors including pcDNA-3.2/cap-TEV-

NT/V5-DEST (N-terminal V5/AP tag); pcDNA-3.2/cap-TEV- N-EmGFP (N- terminal

emerald GFP tag and pcDNA6.2-mCherry-C-DEST (C-terminal mCherry tag, from Dr. S.

Grieshaber, University of Florida). Conversely, LGTV NS5 was derived from PCR amplification using the LGTV E5 molecular cDNA clone as template (from Dr. A.

Pletnev, NIAID, NIH). Each gene was PCR amplified and directionally cloned into the

Gateway entry vector pENTR/SD/D-TOPO (Invitrogen) as previously described (221).

The NS5 entry vectors were directionally cloned into C-terminal gateway destination vectors: pcDNA-3.2/capTEV-CT/V5-DEST (C-terminal V5/AP tag) and pDS_GFP-XB

(C-terminal GFP tag, ATCC).

2.16 Cell transfections for immunoassays

Transfections were performed in HEK 293 cells seeded in 6 well culture plates at 1 X 106 cells/well or 8-well labtek dishes at 5 X 104 cells/well for co-affinity purification or co- localization assays respectively. Cells were allowed to attach for 18 hours prior to transfection. DNA of interest was measured with a Nanodrop machine or an aliquot (1uL) was mixed with Quibit BR DNA reagent (Invitrogen) and quantified using a Quibit reader.

The Lipofectamine LTX reagent was used for the co-affinity purification assays according to the manufacturers’ protocol. For degradation assays, Lipofectamine LTX or

Lipofectamine 3000 reagents were used with similar results. Transfected cells were allowed at least 24 hours to express the gene of interest.

2.17 Transfected cell lines

To generate a P. leucopus cell line by transfection, P. leucopus fibroblasts were set up in a

6-well culture plate at 1 X 106 and transfected with 8ug DNA of plTRIM79-EmGFP. The cell culture media was replaced at 24 hours post-transfection. Furthermore, selection media containing 600ug blasticidin was added to transfected cells at 48 hours post- transfection. On the 4th day, the cells were trypsinized and transferred into a T75 culture flask. With sufficient growth the cells were sub-cloned into 48-well dishes to achieve 1 cell per well. Clonal cells expressing the emerald GFP to high levels were further transferred into 6 well dishes and subsequently into T75 flasks to obtain a cell line. For cell treatment experiments, cells were set up to equal density and treated with MG132 prior to imaging.

2.18 Degradation assays

The genes of interest were co-transfected into varying amounts in HEK 293 cells.

Specifically, the amount of LGTV NS5 transfected into cells was kept constant while

TRIM79 from P. leucopus cells was transfected in increasing amounts. At 24 hours post- transfection, the cells were washed and harvested in RIPA buffer. Samples were further processed according to the immunoblotting protocol described above and probed using antibodies to the fusion tags of the respective expression vectors.

2.19 Co-affinity Purification

For Co-AP experiments, HEK-293 cells were transfected with at least one expression vectors fused to a V5/AP tag (pcDNA-3.2/capTEV-NT or CT/V5-DEST) which contains a

V5 sequence that allows biotinylation of the target protein after transfection in eukaryotic cells. At 24 hours post-transfection, cells were then washed twice with DPBS and harvested in an IP wash buffer (50 mM Tris-HCl pH 7.5, 150mM NaCl, 1% NP-40, and

0.5% Na-deoxycholate) with protease inhibitor cocktail (Roche). Cells were lysed by freeze-thawing three times in liquid nitrogen and 37ºC water bath respectively and treated with TURBO DNase (Thermo Fisher Scientific) to eliminate DNA contamination. The samples were centrifuged for 10 minutes to remove any cellular debris. The supernatant was incubated with uncoated agarose control beads for 3 hours at 4ºC with rotation as a pre-clear step. After pre-clearing, the samples were added to streptavidin-coated beads and incubated overnight at 4ºC with rotation. The beads were washed 3 times in RIPA buffer and beads were eluted by incubation at 95ºC in 50ul sample buffer (62.5 mM TRIS pH

6.8, 10% glycerol, 15 mM EDTA, 4% 2-ME, 2% SDS, and bromophenol blue). The eluted samples were assayed by immunoblotting as described above. Blots were probed with anti-V5 antibody at 1:4000 dilution.

2.20 Ubiquitination Assays

The expression plasmids corresponding to the genes of interest were co-transfected into

HEK 293 cells along with HA-tagged Ub plasmid: pRK5-HA-ubiquitin (Ub)-WT

(Addgene plasmids 17608 and 17603; from Dr. T. Dawson, Johns Hopkins University). At

24 hours post-transfection, samples were harvested and processed similar to the co-AP assay. The exception was the addition of NEM to the IP lysis buffer (This prevents the activity of deubiquitinase enzymes). Additionally, samples were heated in 1% SDS at

95ºC for 5 minutes before affinity purification to remove interacting proteins. Samples were assayed by immunoblot as described above.

2.21 Co-immunolocalization assays

Expression vectors corresponding to the genes of interest were co-transfected into HEK-

293 cells. Transfected cells were fixed according to the immunofluorescence microscopy assay described above with the exception that primary antibody incubation was performed overnight at 4ºC. Additionally, antibodies were used to target the specific tag fused to the genes of interest including anti-mCherry (living colours) and anti-GFP (Invitrogen) antibodies. To assess lysosomal localization, the lysosomal compartment was stained using an anti-LAMP1 antibody to specifically target the lysosomal marker LAMP1.

2.22 Assessing total cellular ROS

Cells were set up in LABTEK slides at 5 X 104 cells per well. 18 hours later, the cells were infected with LGTV at MOI 10. At 1 hour post-infection, the culture media was changed to a phenol-free medium. After another hour (2 hours post-infection) the cells were washed and loaded with 10µM H2DCFDA dye (ThermoFisher scientific). Loaded cells were incubated for 30 minutes at 37ºC to allow cell permeation of the dye. For detection, the cells were overlaid with fresh phenol-free media and transferred in the dark

to the confocal microscope. Controls were unloaded cells and uninfected cells. A similar approach was used for the longer term infections and culture media was removed just before loading the cells with the CM- H2DCFDA dye.

2.23 Assessing mitochondrial ROS

Cells were set up in labtek slides at 1 X 105 cells/well overnight. Attached cells were infected with LGTV at MOI 1, 10 or 100. At 2 hours post-infection, the cells were washed and loaded with 10µM of the MitoSOX detection reagent (Invitrogen). The cells were incubated with the dye for 10 minutes at 37ºC and further washed twice with PBS. The slides were overlaid with mounting medium – ProLong Gold with DAPI stain and immediately transferred to the confocal microscope for imaging.

2.24 Testing the role of ROS on LGTV replication

P. leucopus cells were set up in 24-well culture plates or 8-well labtek slides at 1 X 105 cells/well. Attached cells were treated with 20uM ceramide, 30mM glucose (labteks) or

40uM ceramide (culture plates). At 8 hours post-treatment, the cells were infected with

LGTV at MOI 10 and viral supernatants were collected (culture plates) or the cells were fixed (labtek). Viral supernatant was titrated according to the immunofocus assay protocol described above comparing the level of virus replication in treated or untreated cells.

2.25 Statistical analysis

Data were analyzed by an unpaired t test or Mann-Whitney U test using GraphPad Prism 6 software.

Chapter 3 Restriction of Tick-Borne Flaviviruses in a Cell Culture

Model of Peromyscus leucopus Infection

3.1 Introduction

Viral reservoirs in nature typically remain refractory to infection and exhibit limited symptoms of disease. TBFVs particularly utilize rodents as intermediate hosts of infection and can persistently infect the reservoir hosts thus allowing virus transmission during tick feeding as described above (17). Studies suggest that various TBFVs use distinct reservoir species and the virus-host pair differs in endemic regions, likely controlled by host biology, ecological factors, and intrinsic physiological characteristics. POWV virus infection of the rodent host P. leucopus is associated with disease resistance as shown by a study performed in 1997 by Telford et al. This study demonstrated that infecting P. leucopus mice intracranially with POWV did not lead to neurological disease as is observed in susceptible species thereby suggesting that P. leucopus are rodent reservoirs of POWV (201). However, direct in vivo comparisons of POWV replication between P. leucopus and susceptible counterparts were only performed recently by Mlera et al., 2017.

This study systematically demonstrated that interperitoneal and intracranial infection of P. leucopus mice leads to low viral load, limited viral spread in the CNS, and overall survival of infection while control C57BL/6 and BALB/c mice largely had higher viral load, increased viral spread, demonstrated neurological symptoms and ultimately died from

infection (202). These studies further implicate P. leucopus as a natural reservoir host of

TBFV infection. Ultimately, the goal of studying reservoir species is to identify the mechanism of resistance against virus infection with the aim of applying the information to therapeutic design. While previous studies have confirmed disease resistance in P. leucopus at the whole organism level, the genetic determinants of infection remain to be identified. Based on flavivirus biology, virus restriction is likely to occur at the molecular level whereby the virus infection cycle could be impaired at one or more stages through the action of antiviral factors. Alternatively, the host could lack specific host factors thereby hindering progress of the infection cycle. Taking these factors into account, the primary aim of this project was to elucidate molecular factors contributing to TBFV restriction in P. leucopus. The first step involved designing a cell culture model to recapitulate the natural resistance and test the in vivo findings at the cellular level based on the rationale that a representative model will provide a basis to study the intricacies of the virus life cycle and make molecular modifications to identify the exact mechanism of restriction.

3.2 Results

3.2.1 LGTV is restricted in cells of the reservoir host P. leucopus

To determine whether the resistance of P. leucopus to TBFs can be recapitulated in vitro, primary skin fibroblasts of P. leucopus were infected with LGTV - a non-virulent member of the tick-borne flavivirus serocomplex. Virus infection was compared to fibroblasts from a closely-related susceptible mouse strain Mus musculus – C57BL/6J. After 1-hour

absorption period, the inoculum was removed and replaced with fresh culture medium.

The number of infectious virions released into the culture medium was titrated at various time points post-infection by immunofocus assays on Vero cells using a primary antibody to detect the virus E glycoprotein. There were significantly less infectious virions released from P. leucopus fibroblasts compared to the control M. musculus cells (Figure 3-1A).

Additionally, virus infection was visualized by an immunofluorescence assay. P. leucopus adult skin fibroblasts and M. musculus embryonic fibroblasts (MEFs) were infected in

Labtek slides at increasing multiplicities of infection (MOI) 1, 10 and 100 for 72 hours after which the slides were fixed and stained with antibodies against the E glycoprotein and viral protease NS3. The viral protein staining showed markedly lower expression of viral protein in the P. leucopus cells such that even at MOI 100 – the highest MOI tested, there were less infected P. leucopus cells than the M. musculus cells infected at an MOI of

1 (Figure 3-1B). In order to determine if the restriction phenotype in P. leucopus is dependent on the origin of the primary fibroblasts – i.e. adult vs embryonic, we performed a control experiment using P. leucopus embryonic fibroblasts (PEFs). In a conserved trend to the adult fibroblasts, virus titration by an immunofocus assay showed significantly lower virus replication at various time points in the PEFs compared to the MEFs.

Similarly, the immunofluorescence staining for viral proteins E and NS3 showed markedly lower virus infection in the PEFs showing that the restriction of LGTV in P. leucopus occurs irrespective of the fibroblast origin. (Figure 3-1C-D).

3.2.2 Virulent TBFVs are restricted in P. leucopus cells

The possibility exists, that the in vitro model shows restriction to LGTV because this is an attenuated strain with limited capability to promote pathogenesis and replicate in vitro

(88). Additionally, the ultimate goal of this study is to identify restriction factors that prevent disease by highly virulent TBFVs. Therefore, our next step was to test the restriction of virulent TBFV strains in P. leucopus fibroblasts. The viruses used included

POWV lineage 1 (BSL3), deer tick virus (BSL3) and TBEV- Sofjin strain (BSL4). For these experiments, P. leucopus and M. musculus fibroblasts were infected with the respective viruses and virus supernatant was harvested at various time points post infection. Virus was titrated by a plaque assay and results show that all the TBFV strains were restricted similar to LGTV such that there was significantly less virus replication in the P. leucopus cells compared to M. musculus cells (Figure 3-2A-C). These data suggest that LGTV is a suitable representative model to study the restriction of TBFVs in P. leucopus cell culture system.

3.2.3 Restriction in P. leucopus cells is specific to flaviviruses

To determine if the observed restriction is specific to the flavivirus family, we performed growth curve studies using KUNV a BSL-2 strain of the mosquito-borne WNV sharing about 98% amino acid homology with the virulent strain (222-224). Virus titration in P. leucopus and M. musculus cells showed lower virus replication of KUNV in P. leucopus cells (Figure 3-3A) suggesting that the restriction in P. leucopus cells occurred with a

MBFV. As a virus family control, we performed growth curve experiments using

vesicular stomatitis virus (VSV) a member of the Rhabdoviridae family of viruses.

Interestingly, VSV replication was not restricted in P. leucopus cells at the various time points (Figure 3-3B). Taken together, this set of experiments suggest that P. leucopus cells are refractory to infection by flaviviruses including highly virulent strains and this restriction is family-specific since VSV replication is not restricted. This in vitro model recapitulated TBFV restriction and offers a framework to closely study the virus infection cycle in the reservoir species. Given that the level of virus replication is correlated with the propensity of the host to exhibit symptoms of disease, these results are therefore consistent with the phenotype observed in nature and reports of in vivo studies whereby the reservoir species P. leucopus survives infection and harbors the virus for varying periods. In order to clearly elucidate the point of virus restriction, we put the cell culture model to use by systematically testing the steps involved in the flavivirus infection cycle with the aim of determining the restriction point.

3.2.4 Virus entry is not inhibited in P. leucopus cells

Entry of flaviviruses into host cells occurs by receptor-mediated endocytosis. The events after infection lead to delivery of the virus into early endosomes, which mature and fuse with the lysosome. The viral envelope protein further fuses with the host membrane to release the viral RNA for translation and subsequent replication. Based on the elucidated route of flavivirus entry, a lack of attachment factors, receptor, or other pertinent host factors in the reservoir host could inherently result in a failed infection and might potentially explain why TBFV infection of P. leucopus cells has shown consistently low

numbers. To determine if virus restriction in P. leucopus occurs through a block at the virus entry step or due to a lack of virus receptors, entry assays were performed to compare the number of infected P. leucopus cells early in infection compared to M. musculus cells. Briefly, both cell types were infected at varying MOIs and incubated or 1 hour at low temperature to allow for virus attachment and promote a synchronous infection. Next, the cells were transferred to 37ºC to allow virus entry. At 1hour post entry, the cells were subjected to an acid wash to ensure that only internalized virus would be quantified. We first visualized the virus by staining for the structural E glycoprotein and Figure 3-4A illustrates a positively infected cell at low (20X) and high (60X) magnification. Quantification of immunostained cells showed that there was no significant difference in the number of virus-positive P. leucopus cells compared to the susceptible

M. musculus control cells (Figure 3-4B). Treatment of both cell types with bafilomycin

A1,- an inhibitor of the endocytic trafficking pathway, served as a negative control and led to significantly lower number of infected cells (Figure 3-4B, right-most bars).

Furthermore, the RNA population that is predominant at early stages of infection (positive sense RNA) was measured by quantitative real-time PCR and similar to the immunofluorescence results, there was no significant difference in the abundance of viral positive sense RNA in P. leucopus and M. musculus cells following infection (Figure 3-

4C). Taken together, these data suggest that the route of virus entry is the same in cells of the resistant and susceptible cells and importantly, that restriction does not occur at the point of entry. Essentially, the entry receptors for flavivirus infection are likely present on these cells thereby making them permissible to infection. For reference, we observe that

the restriction likely occurs in a dose-dependent manner since increasing the viral MOI in

P. leucopus leads to higher virus replication (Figure 3-1A), thus hinting that host factors for replication might be present in these cells.

3.2.5 Viral RNA is successfully translated in P. leucopus

After virus entry, the positive strand RNA of flaviviruses can immediately serve as messenger RNA and is translated to protein in the host cell. The newly-synthesized viral proteins are essential to determine the fate of the virus infection as these are the main components required for virus RNA replication, packaging of new virion particles, antagonism of the host cell defense, and overall manipulation of the host cell environment.

Consequently, any restriction at the translation stage would be detrimental to the virus and can result in the release of significantly fewer virion particles from the host cells as observed in P. leucopus cells. First, we assessed the expression of viral protein in LGTV- infected P. leucopus and M. musculus cells. Viral protein was detected by immunoblotting in both cell types at various times post-infection, however there was higher accumulation of viral E and NS1 proteins in the M. musculus cells (Figure 3-5A), likely due to the higher rate of virus replication in these cells. To directly test the ability of viral RNA to get translated in P. leucopus cells, we utilized the infectious clone technology whereby an infectious cDNA clone of LGTV was used to transcribe viral RNA in vitro. Equal amount of in-vitro transcribed LGTV RNA was then transfected into P. leucopus and control M. musculus cells before probing for viral protein staining by immunofluorescence. In both cell types, we were able to detect viral protein including viral structural (E) and

nonstructural (NS3) viral proteins (Figure 3-5B). This suggests that delivering viral RNA led to detection of viral protein in the resistant and susceptible cells, indicating that restriction to the TBFV -LGTV in P. leucopus does not occur at the translation stage.

3.2.6. Replication of viral RNA is significantly inhibited in P. leucopus cells

After viral translation, the viral NS proteins cooperate within a replication complex to mediate genome proliferation. The replication of flavivirus RNA is initiated by synthesis of a complimentary negative strand, which serves as a template for the synthesis of more positive strand copies as reviewed above. Therefore, to test this next step in the virus infection cycle, abundance of the viral negative strand RNA was measured in P. leucopus cells as a marker of virus replication at various time points post-infection. There was significantly less virus replication in P. leucopus cells compared to the control (Figure 3-

6A). This is the first point of virus block identified during the infection cycle in the reservoir host cell and suggests that restriction of LGTV at least occurs at the replication step of infection. Furthermore, the number of infectious virions released from the cell correlated with the restriction at the replication phase at MOI 10 in both cell types (Figure

3-6B) thereby suggesting that a block to replication subsequently results in less virion particles being packaged and released from the host cell. In order to distinguish whether the lower titers in P. leucopus supernatants was due to an overall lower number of virions or reduced infectivity of the virus particles being released from these cells, we directly isolated viral RNA from supernatants derived from infected cells. The viral RNA was quantified by qRT-PCR to specifically measure the positive strand population by an

absolute quantification. We observed significantly lower amount of viral RNA in the supernatants obtained from P. leucopus cells compared to the M. musculus cells (Figure

3-6C). These data suggest that a block in the replication step leads to lower RNA accumulation and likely results in an overall lower number of virion particles being released from the reservoir host cells.

3.3 Chapter 3 Discussion

This section has elucidated for the first time, that the resistance of a natural TBFV reservoir can be recapitulated in host cells. Till date, molecular factors necessary for resistance in natural TBFV hosts have not been fully elucidated and this is likely due to the lack of a well-defined cell culture model. Although a number of studies have probed for host restriction factors against TBFVs and flaviviruses in general, most of these studies are performed in human and mouse models of infection. The main caveat to this approach is that these species are inherently susceptible to infection therefore, the restriction factors expressed may not be sufficiently potent to prevent disease. The cell culture model described here uses both primary embryonic and adult fibroblasts from P. leucopus cells and compares the rate of TBFV infection to that of M. musculus cells. The data show a clear restriction phenotype even with the highly virulent strains of TBF. It is worthy of note, that higher MOI in P. leucopus cells led to higher rate of virus replication suggesting that necessary host factors are likely present and restriction is not due to lack of cell permissiveness (Figures 3-1 A and 3-4B). Interestingly, KUNV which is a mosquito- borne flavivirus showed restriction in P. leucopus suggesting that there may be broad

antiviral factors that recognize and restrict conserved components of the flavivirus particle or similar proviral factors necessary for replication could be lacking in the natural host.

Indeed, several studies have shown broad-acting restriction factors that recognize more than one member of the flavivirus family (225-227). It is possible that the mechanisms of restriction in P. leucopus varies between the tick and mosquito-borne viruses such that restriction factors against TBFV have evolved over time during the vector-host interaction in nature while MBFV antiviral factors function differently within this host. Nonetheless, further surveillance studies will be required to completely rule out the possibility that cross-species infections have occurred in the wild to potentially expose MBFV to tick vectors and subsequently allow P. leucopus mice to encounter KUNV in the wild – This could offer an alternative explanation as to why MBFV are restricted in the known reservoir species of TBFV.

To identify the exact point of restriction, a step-wise study of the individual steps involved in the flavivirus life cycle was performed, and identified replication as a point of restriction. The rationale for these studies is based on previous work done with other viruses showing that the viral life cycle can be inhibited at various stages and typically impacts the subsequent stages of the viral infection course. For example, human immunodeficiency virus (HIV) is restricted at the early entry step by host TRIM5-alpha while the virus release stage is restricted by another host factor tetherin (228,229).

The restriction at the replication stage in our study was deciphered by significantly lower accumulation of viral negative strand RNA during infection in P. leucopus cells.

This suggests that one or more proteins within the viral replication complex is impaired in

these cells and is unable to adequately promote viral replication. The viral enzymes NS5 and NS3 are pertinent for viral RNA synthesis and could be potential candidates targeted in P. leucopus cells. The finding that these resistant cells are in fact, permissible to infection suggest that any impediment to viral pathogenesis occurs intracellularly and served as an initial basis for us to hypothesize that the host possesses necessary factors for virus replication. However, we remain cognizant of the fact that certain proviral factors might be lacking at the replication stage and could explain why there is significantly less viral RNA synthesis in P. leucopus cells. Based on the RNA quantification studies, we were able to distinguish the outcome of restriction in P. leucopus cell culture infections to be occurring due to a lower number of virion particles being released from the cells and not due to a decrease in infectivity of the released virions. Given that infectivity of viruses released from P. leucopus cells remains intact and flavivirus infectivity is mediated by viral structural proteins, we can infer from these data that the restriction of TBFV in P. leucopus likely occurs by targeting the viral non-structural proteins, although this remains to be determined by further biochemical studies.

Overall, findings of the infection cycle studies described herein present a hypothetical model whereby in the susceptible species, the virus enters the cell by endocytic trafficking and proceeds through the steps of the infectious cycle relatively uninhibited – replication leads to accumulation of viral RNA which can be packaged into virion particles and released from the cell. On the other hand, while virus can enter the reservoir host cells via the same endocytic trafficking pathway and express viral proteins successfully, there is a retardation to the replication process in these cells leading to

significantly less accumulation of viral RNA, likely lower number of virion particles to be packaged and released from the host cell (Figure 3-7). This set of data presents a simple model of infection whereby the resistance to TBFV is recapitulated in P. leucopus cells and a clear point of restriction could be identified therefore suggesting that this is a viable model to specifically identify the molecular mechanism(s) of restriction. Importantly, these experiments set an outline for the potential study of other pathogen-reservoir pairs.

All the data in this chapter have been published in the primary dissertation manuscript

(230).

Figure 3-1 LGTV is restricted in P. leucopus cells Adult and embryonic fibroblasts from P. leucopus and M. musculus were infected with LGTV 1 and 10 respectively and viral supernatants were quantified at the indicated time points (A), Infected cells were also assessed by immunofluorescence to detect viral E and NS3 proteins. Cell nuclei are stained in blue (DAPI). 20X magnification (B). LGTV MOI 10 infection was also quantified in PEFs and compared to MEFs by titration (C) or visualized by immunofluorescence microscopy (D). Mean ± SD, data are from three independent experiments performed in triplicate. Immunofocus assay data are presented as FFU/ml. Asterisks indicate: * = p < 0.05, ** = p <0.01, *** = p <0.0001

Figure 3-2 Virulent TBFV are restricted in P. leucopus cells P. leucopus and M. musculus fibroblasts were infected with: (A) POWV-I B) POWV-II C)TBEV strain Sofjin. Supernatants were collected at the indicated time points and quantified by plaque assay. Plaque assay data are presented as PFU/ml Mean ± SD, data are from three independent experiments performed in triplicate. Asterisks indicate: * = p < 0.05, ** = p <0.01, *** = p <0.0001.

Figure 3-3 Restriction in P. leucopus is specific to flaviviruses P. leucopus and M. musculus fibroblasts were infected with: A) VSV and B) KUNV at MOI 0.1 and 10 respectively. Supernatants were collected at the indicated time points and quantified by a plaque assay. Data are presented as PFU/ml. Mean ± SD, data are from three independent experiments performed in triplicate. Asterisks indicate: * = p < 0.05, ** = p <0.01, *** = p <0.0001.

Figure 3-4 The viral entry step is not restricted in P. leucopus cells Representative confocal 20X (top) and 60X (bottom) images showing P. leucopus cells infected with LGTV after 1 hpi at 37°C. The viral E protein is stained in green and nuclei are stained in blue (DAPI) (A). Quantified data are shown as the total number of infected cells at MOI 1, 10, and 100 respectively. Cells treated with bafilomycin A1 (BAF) at 15 min post-infection are shown as a negative control (B), Viral entry was assayed in P. leucopus and M. musculus fibroblasts infected with LGTV MOI 10 AT 1h post virus absorption. Data show abundance of LGTV positive (+ve) strand RNA assessed by qRT-PCR (C).

Figure 3-5 Viral RNA is translated in P. leucopus cells A) Immunoblot of LGTV NS1 and E proteins in P. leucopus and M. musculus fibroblasts at the indicated time points post infection. B) Confocal image showing P. leucopus and M. musculus cells expressing LGTV proteins. Cells were transfected with viral RNA for 5 days after which the cells were fixed and immunostained for LGTV E (green) and NS3 (red). Cell nuclei were stained with DAPI (blue) and visualized by confocal microscopy (20X magnification)

Figure 3-6 The RNA synthesis step is restricted in P. leucopus cells P. leucopus and C57BL/6 cells were infected at MOI 10. RNA lysates were collected at the indicated times post-infection and the resulting cDNA was used as template for RT-qPCR. The quantity of viral RNA was determined by the relative standard curve method to determine the abundance of: (A) LGTV negative strand (-) RNA from cell lysates (B) LGTV FFU/ml titrated from virus recovered in the cell culture supernatant and (C) LGTV positive strand (+) RNA from from virus released into the cell culture supernatants. Mean ± SD, data are from three independent experiments performed in triplicate. Asterisks indicate: * = p < 0.05, ** = p <0.01, *** = p <0.0001.

Figure 3-7 Hypothetical model of TBFV restriction in the reservoir host A hypothetical model based on the results of the infection cycle studies comparing virus progress in the reservoir and susceptible species. In both cell types, the viral receptors are expressed and permit comparable entry of virus into the cell, expression of viral protein and ability of the virus to antagonize host responses. During infection in the susceptible species, the virus is likely able accumulate high amounts of its RNA genome, this replication stage is significantly inhibited in the reservoir host likely leading to low availability of nascent RNA for packaging and ultimately lower numbers of virion particles released. The decrease in the total number of virions released, despite successful viral entry supports the notion that an intracellular block restricts viral progression in the susceptible species.

Chapter 4 Assessing the role of IFN signaling to virus restriction in

P. leucopus

4.1 Introduction

Based on flavivirus biology, we reasoned that the observed restriction to TBFV replication in P. leucopus is either due to a lack of proviral factors needed for viral proteins to function and mediate replication or due to the presence of potent antiviral factors that actively block the replication step. Since the flavivirus genome only encodes 10 proteins, the virus needs to maximize protein functionality by making post-translational modifications to individual proteins. The modifications ultimately impact the specific function(s) of a protein population. Additionally, viral proteins co-opt host proteins to successfully proceed through various stages of the infection cycle. Essentially, high level of virus replication is dependent on availability of host factors needed to promote viral protein function such as host ubiquitin ligases and kinases. On the other hand, the viral proteins have to actively antagonize the host immune response or to evade it altogether as reviewed above. Therefore, low virus replication in cells like P. leucopus fibroblasts could be due to inability of the viral proteins to co-opt host proteins and antagonize the P. leucopus immune response, lack of or variable expression of proviral factors or an active suppression of virus replication by host antiviral factors. While all 10 flavivirus proteins are important for virus infection, the NS5 protein is particularly important as the viral

polymerase. Additionally, the NS5 protein is implicated as the main virulence factor of flaviviruses due to its ability to antagonize the innate immune response by a variety of mechanisms. Flavivirus infection of host cells leads to the release of IFN which further signals through the IFN receptor to control the expression of antiviral genes which should ideally control virus infection. However, the NS5 protein of distinct flaviviruses efficiently blocks IFN signaling by targeting various players in the IFN signaling cascade

(71). It is important to note that the ability of flaviviruses to antagonize IFN signaling particularly through the function of NS5 is a major determinant virus species tropism. An example is seen with DENV which is not able to infect mice, although it successfully causes disease in humans. Studies have demonstrated that while DENV NS5 can degrade

STAT2 in human cells to antagonize IFN signaling (231), the viral protein is unable to degrade mouse STAT2 or antagonize mouse IFN responses therefore posing a major challenge to development of a mouse model of DENV infection (155). Of relevance to this study is the finding that NS5 of TBFV inhibits IFN signaling by preventing the phosphorylation of STAT1 during virus infection thus causing us to probe whether low virus replication in P. leucopus cells is due to inability of NS5 to function as an IFN antagonist in these cells. Essentially, the success of flavivirus infection in a given cell population will be determined by the nature of the IFN response, the ability of the virus to evade these responses and the host potent countermeasures via specific restriction factors.

In this section, we assessed the relative contribution of IFN signaling to TBFV restriction in P. leucopus cells by first testing potential reasons for restriction, assessing the functionality of NS5 in P. leucopus cells, identifying novel key antiviral signaling genes

and altering gene expression to relieve restriction. Furthermore, we present preliminary characterization of P.leucopus TRIM79 (plTRIM79) a homolog of a murine TBEV- specific restriction factor previously characterized by our group (217).

4.2 Results

4.2.1 IFN treatment impacts LGTV replication in P. leucopus

Having identified restriction to begin at the replication phase, we rationalized that, cellular mediators being released during infection could signal downstream to allow upregulation of antiviral genes and ultimately establish an antiviral state. We therefore wanted to test the effect of IFN treatment on virus infection in the reservoir species. For these experiments, we utilized mouse IFNβ (mIFNβ) since there is currently no commercially- available IFNβ from P. leucopus. First, we tested if mIFNβ could effectively stimulate antiviral signaling in P. leucopus cells by performing a functional antiviral assay using the interferon-sensitive virus VSV. Indeed, pre-treatment of cells with mIFNβ led to significantly lower virus replication and protection of cells from cytopathic effects (CPE) as shown in Figure 4-1A-B demonstrating that the mIFNβ works to elicit a protective response in P. leucopus cells and can be used as a tool to study IFN responses in these cells. We next anticipated that if IFN signaling contributes to LGTV restriction in P. leucopus cells, then addition of exogenous IFN will further decrease virus replication.

Indeed, pre-treatment of P. leucopus cells with 1000IU of mIFNβ prior to infection with

LGTV resulted in significantly less virus replication (Figure 4-1C). Next, we asked whether IFN is produced to excessively high levels in P. leucoupus cells to make the

100

cellular environment refractory to infection. To test this, we performed a functional antiviral assay by transferring supernatant from virus-stimulated cells to fresh cells on the premise that any soluble mediators present in the supernatant will confer protection from subsequent infection with an interferon-sensitive virus - VSV. Protection levels were compared to an IFN-b standard for each cell type which was created using serial dilutions of known IFN concentration. There was higher level of protection conferred by supernatants from P. leucopus cells treated with a potent IFN inducer –Sendai virus

(SENV) compared to supernatant from M. musculus cells. This suggests that when stimulated with an IFN-inducing virus, P. leucopus cells released correlates of protection to higher levels than M. musculus cells. Interestingly, the results suggest that endogenous

IFN is not produced to high levels in either P. leucopus or M. musculus cells following

LGTV infection (Figure 4-1D-E). This is not surprising as flaviviruses have been associated with low and delayed IFN induction during infection (232). Therefore, if IFN signaling is necessary for restriction in P. leucopus cells, it is likely occurring through the up-regulation of antiviral genes and not due to accumulation of IFN to high concentrations.

4.2.2 LGTV NS5 protein functions as an IFN antagonist in P. leucopus

Many groups have shown that the NS5 protein of TBFVs and MBFVs functions as an antagonist of Type 1 IFN signaling by preventing the phosphorylation of transcription factor STAT1(155,231). As highlighted in section 4.1, the antagonism can determine the success of the virus infection cycle and is crucial for species tropism. Indeed, if the NS5

101

protein fails to antagonize IFN signaling in the reservoir species, virus replication could be impaired and thus explain why we consistently observe significantly lower levels of virus replication in the reservoir host cells. We tested this hypothesis by assessing the levels of phosphorylated STAT1 in P. leucopus and M. musculus cells following virus infection.

There was visually lower pSTAT1 staining in both cell types although the decrease was less drastic in P. leucopus cells (Figure 4-2A). However, these cells have lower accumulation of viral protein due to reduced replication and this could in turn affect the level of NS5 available to alter pSTAT1 expression. We next tested NS5 function directly by making clonal cell lines of P. leucopus overexpressing the viral NS5 protein fused to a

GFP tag (Figure 4-2B). Functional assessment of these cells by an antiviral assay using

VSV for its IFN sensitivity, showed that in the presence of NS5, IFN treatment failed to protect the cells from VSV infection compared to the control GFP-only cells (Figure 4-

2C). Additionally, when the clonal cells were treated with exogenous mIFNβ, there was significantly lower induction of an ISG plTRIM79 in the presence of NS5 (Figure 4-2D).

Overall, these data suggest that the viral NS5 protein functions as an IFN antagonist in the reservoir species and this is likely not the mechanism of restriction. Importantly, these findings suggest that the virus has the necessary host factors needed to mediate replication and block IFN signaling. Despite this ability, virus infection in P. leucopus cells is consistently associated with significantly lower virus replication in our hands thereby suggesting that although the virus has the necessary proviral factors to antagonize IFN signaling, the reservoir host likely launches a more potent counter-response to override the

IFN antagonism and maintain virus replication at a low level.

102

4.2.3 Homologs of antiviral signaling genes are expressed in P. leucopus and upregulated during virus infection

The fact that virus replication is relatively low in P. leucopus cells despite IFN antagonism activity by NS5 suggests that the cells elicit a more potent response to overcome the viral antagonism and block efficient replication. While viral recognition and antiviral signaling factors have been widely studied in various species, little is known about the expression of those genes in reservoir hosts and how they function in response to infection.

Unfortunately, very minimal genome information has been available for P. leucopus until recent work performed by our lab and others (233). Using RNAseq data from an ecological P. leucopus study, we identified homologs of core components of the IFN signaling system in P. leucopus (Figure 4-3A). Our group has now resolved the sequence of these genes for the first time and the information was submitted to the NCBI database and assigned the following accession numbers – RIG-I (KY451964) MAVS (KY451965),

IRF1 (KY451966), STAT1 (KY451962) and IFNAR1 (KY451963). Additionally, we have identified a homolog of the only known TBEV-specific restriction factor TRIM79

(KY451967) belonging to the TRIM family of proteins (further discussed below). These genes are relatively identical to their counterparts in M. musculus, (Table 4-1) however; it remains to be tested whether the inherent differences could result in reduced replication of virus in the resistant species. To study the kinetics of gene expression primers were designed based on the new sequence information and, transcript levels were measured following IFN treatment (Figure 4-3B) and virus infection (Figure 4-3C). As expected,

103

there was high induction of STAT1, RIG-I and TRIM79 in response to IFN treatment while increases in IFNAR1, MAVS and IRF1 levels were modest as these are not typically

ISGs. Next, we assessed protein levels of key IFN-induced flavivirus restriction factors

IFIT2 and IFIT3 by immunoblotting protein lysate from virus-infected and interferon- treated cells. There was an increase in IFIT2 protein staining in both cells following IFN treatment while protein increase with virus infection was little to none (Figure 4-3 D).

While IFIT protein expression levels were generally lower in P. leucopus cells, we made an interesting observation that the size of the IFIT proteins tested varied between both cell types particularly IFIT2 suggesting that these proteins could be markedly different. On the other hand we assessed STAT1 expression in response to virus infection (Figure 4-3 E) and IFN treatment (Figure 4-3 F) and the data show consistently higher basal and induced levels of STAT1 in P. leucopus cells compared to M. musculus cells.

4.2.4 Targeting transcription factor STAT1 relieves LGTV restriction in P. leucopus

In order to fully delineate whether low replication of TBFV in P. leucopus occurs through the lack of proviral factors or expression of antiviral mediators, we hypothesized that if the latter were the case, then targeting expression of key antiviral signaling factors will relieve restriction. On the other hand if restriction occurs as a result of P. leucopus lacking proviral factors, then molecular targeting of the IFN pathway will have no impact on the restriction phenotype. We started by targeting STAT1, a major player in the IFN response pathway crucial for signaling and downstream activation of IFN signaling genes to establish an antiviral state. We confirmed STAT1 knockdown at the RNA, protein and

104

functional levels respectively (Figure 4-4A-C). In line with our hypothesis, knockdown of

STAT1 using a lentiviral vector in P. leucopus cells led to higher virus replication and greater accumulation of viral protein in the P. leucopus cells (Figure 4-4D-E).

Additionally, restriction was lost at the replication phase in STAT1 knockdown cells as demonstrated by significantly higher virus titers and higher accumulation of virus negative strand RNA (Figure 4-4F-G). This loss of restriction is the first indication that virus replication can occur to high levels in these cells and thereby suggests that the restriction observed in the wild-type P. leucopus cells is an active suppression of virus infection and that the virus replication can proceed efficiently in these cells if not otherwise inhibited.

4.2.5 Signaling through the type 1 IFN receptor IFNAR1 is necessary for LGTV restriction in P. leucopus

Following the IFN induction process in virus infected cells, IFN binds in an autocrine or paracrine manner to the IFN receptor to elicit downstream signaling and establish an antiviral state. Having identified STAT1 as a potential restriction factor in P. leucopus cells, we wanted to determine if the restriction phenotype was due to its general role in

IFN signaling or a more specific restriction cascade. For this purpose, we targeted IFN signaling on a broader scale by knocking down the type 1 IFN receptor (IFNAR1) and impairing IFN signaling Figure 4-5A, 4-4C). As expected, IFNAR1 knockdown led to a loss of restriction and more robust virus replication as visualized by immunofluorescence microscopy (Figure 4-5B). There was also higher number of virion particles released and accumulation of negative-strand RNA in the IFNAR1 KD cells compared to the non-

105

silencing control (Figure 4-5C-D). Interestingly, knockdown of MAVS showed only a modest effect on virus replication (not shown) suggesting that restriction is occurring mainly through the IFN response cascade and less through the IFN induction pathway.

4.2.6 Characterizing TRIM79 as a potential restriction factor of virus replication in

P. leucopus

Our lab has previously identified the only known TBFV-specific restriction in a murine model of infection –TRIM79. This host protein was detected via a yeast two hybrid screen whereby TRIM79 bound to the viral protein NS5. Expression studies showed significant induction of TRIM79 in mouse cells during virus infection and IFN treatment suggesting that the gene is regulated by these events. Furthermore, cellular distribution experiments showed that TRIM79 is highly expressed in organs that are associated with immune control such as the spleen (Taylor, 2011). On a molecular level, TRIM79 closely resembles the human TRIM5-α gene which has been widely studied for its role as a virus restriction factor. Mechanistically, TRIM79 was demonstrated to be a restriction factor which physically interacted with LGTV NS5 and targeted it for degradation by the lysosomal pathway. The degradation was a striking phenotype shown on immunoblots whereby exogenous co-expression of LGTV NS5 with increasing amounts of TRIM79 led to decreased NS5 expression. The lysosomal dependence was inferred because treatment of cells with the lysosomal inhibitor ammonium chloride interfered with this degradation phenotype and the protein co-localized with lysosomal markers. Proof-of-concept

106

experiments further showed that when TRIM79 expression was targeted using lentiviruses expressing shRNA against the gene, there was increased TBEV replication therefore demonstrating that TRIM79 expression helps to decrease virus replication in mouse cells

(217). This process also shows the importance of NS5 as a viral protein necessary for successful replication of the infecting virus. A major finding of the TRIM79 study was the specificity of action displayed by the host protein; while TRIM79 bound to NS5 from

TBEV, there was no interaction observed with NS5 from WNV which is a mosquito-borne member of the flavivirus family. These studies present a unique opportunity to make sequence comparisons between TBFV and MBFV NS5 proteins and identify regions of variation that affect restriction factor binding, with a view to providing new therapeutic design strategies. It is important to note that even with TRIM79 expression, the mouse cells are still more susceptible to infection than P. leucopus cells as has been demonstrated in this study. Based on our findings, the virus restriction in P. leucopus is mediated by a potent IFN response since IFNAR1 KD relieved the restriction phenotype. Ultimately, signaling through IFNAR1 will upregulate a myriad of genes and these would likely be the main players of the potent antiviral response. Indeed, mouse cells also have a potent

IFN signaling cascade in response to virus infection and numerous antiviral genes are upregulated to control infection. However, the disparity in the level of virus infection between wild type cells from the two species – P. leucopus and M. musculus suggests that the rate of IFN signaling might not be the determinant of virus replication, rather virus restriction is determined by the actual identity and potency of the genes upregulated by

IFN signaling during virus infection. Therefore, a major goal of this project is to identify

107

specific gene signatures induced by virus infection and IFN signaling in P. leucopus cells and determine their level of sequence similarity to the mouse counterparts, mechanism of action and functionality. We hypothesize that while IFN signaling might proceed by a similar mechanism, the gene signatures induced in P. leucopus cells are likely unique in action thereby leading to lower virus infection in these cells. As a starting point, we sought to identify a P. leucopus homolog of the TBEV restriction factor TRIM79 in P. leucopus cells.

4.2.7 plTRIM79 is a virus and IFN-stimulated gene in P. leucopus which is regulated by the proteasome

To clone out TRIM79 from P. leucopus, the host fibroblast cells were treated with mIFNβ,

RNA was harvested 4 h later and cDNA was obtained from the resultant RNA. The cDNA was amplified using primers that were designed from an EST library sequence hit obtained using the mouse TRIM79 sequence. Overall, we were able to specifically purify TRIM79 from P. leucopus (Figure 4-7A) and determine its sequence identity to the mouse

TRIM79 to be about 75%. Next we challenged P. leucopus cells with LGTV or mIFNb and assessed the gene fold induction relative to mock treatment. The results showed that plTRIM79 is upregulated up to three and eight fold following LGTV infection and IFN treatment respectively. In order to determine if plTRIM79 is regulated in the reservoir host cells similar to the M. musculus counterparts, we made P. leucopus cells overexpressing plTRIM79 by transfecting an expression plasmid of plTRIM79 fused to an emerald GFP tag (plTRIM79-EmGFP) into the host cells and expanding single clonal cells with

108

adequate GFP expression by sub-cloning techniques. Next, we treated the clonal cells (or not) with an inhibitor of the proteasomal pathway (MG132) for 12 hours and following the incubation period, we imaged the cells for GFP expression. Interestingly, there was higher accumulation of green cells in the presence of MG132 compared to the untreated cells showing that MG132 treatment increased the expression of plTRIM79 (Figure 4-7B).

This suggests that plTRIM79 is regulated by the proteasome similar to its murine counterpart and these data together demonstrate that although sequence differences exist, the TRIM79 homolog in P. leucopus exhibits similar cellular characteristics to the previously-characterized mouse gene in terms of induction and regulation.

4.2.8 TRIM79 from P. leucopus co-localizes and binds to LGTV NS5

Based on the knowledge of murine TRIM79 function, we reasoned that for P. leucopus

TRIM79 to be implicated as a restriction factor of TBFV, there would be cellular co- localization and interaction with the virulence factor of the virus NS5. To test co- localization, we co-transfected plTRIM79 fused to an mCherry fluorescent tag

(plTRIM79-mCherry) with LGTV NS5 fused to a V5 (NS5-VS) into human embryonic kidney- HEK-293 cells in LABtek slides. In these experiments 293 cells were used for their amenability to transfection. The slides were fixed and stained with a counter fluorescent secondary antibody to the V5 tag to obtain fluorescent images. Confocal microscopy images show that TRIM79 from P. leucopus co-localized with the NS5 protein of LGTV (Figure 4-8A). Interestingly, the cellular distribution of plTRIM79 is similar to the mouse homolog of this gene demonstrated by a distinct punctate

109

morphology in the transfected cells thereby suggesting that these proteins have similar cellular distributions. Next, we tested for a physical interaction between plTRIM79 and

LGTV NS5 by performing a co-affinity purification assay. plTRIM79-mCherry was co- transfected into HEK-293 cells with LGTV NS5 bearing a modified V5 tag (CT tag) that allows it to become biotinylated inside mammalian cells. The protein samples harvested from the transfected cells were affinity purified using streptavidin beads and the samples were analyzed by immunoblotting. The results show that NS5 was able to pull down plTRIM79 in HEK 293 cells suggesting that both the host and viral protein interact physically. The reverse pull-down using plTRIM79 fused to an NT tag also showed an interaction between both proteins (not shown). Taken together, this set of data suggests that plTRIM79 is a host factor upregulated in response to virus infection and interacts with the viral protein NS5.

4.2.9 TRIM79 from P. leucopus impacts the ubiquitination status of LGTV NS5

Studies have shown the importance of post-translational modifications to viral proteins.

Particularly because the viruses encode only a few proteins, these modifications are necessary to bestow multiple functions on the viral proteins by broadening their range of interactions with other host proteins. One of the post-translational modifications identified by our group for LGTV NS5 is the addition of a ubiquitin moiety to form an ubiquitinated population of NS5. In order to test the impact of plTRIM79 on LGTV NS5 ubiquitination, we performed a ubiquitination assay by co-transfecting plTRIM79 and LGTV NS5 into

HEK 293 cells along with ubiquitin bearing an HA tag. Protein was harvested from

110

transfected cells and affinity purified prior to western blotting. The data suggests that when LGTV NS5 is expressed along with plTRIM79, there is decreased ubiquitination status compared to plTRIM79 expressed alone. (Figure 4-9A). Interestingly, there was decreased NS5 expression in the presence of plTRIM79 compared to NS5 alone suggesting that plTRIM79 might be degrading LGTV NS5 when co-expressed and might actually explain the decreased ubiquitination staining. To perform a direct degradation assay, we co-transfected increasing amounts of plTRIM79 with constant levels of LGTV

NS5 in HEK 293 cells and probed protein expression by immunoblotting. There was no visible decrease in the NS5 levels in this experiment (Figure 4-9B). This differs from the ubiquitination assay by the addition of ubiquitin in the former experiment but not the latter therefore, further details of the role of ubiquitination to NS5 degradation will highly benefit this line of enquiry. Additionally, we performed a co-localization experiment to determine if plTRIM79 and LGTV NS5 are localized to the lysosome. For this purpose we performed a similar experiment as described in figure 4-7A and added a probe for the lysosomal marker LAMP1. Here, we did not detect consistent localization to the lysosome

(Figure 4-9C), thereby suggesting that degradation by lysosomal degradation is likely not observable in the human cell line tested for this gene.

4.3 Chapter 4 Discussion

The work reported in this section involved broad characterization of the IFN response in

P. leucopus cells and the specific role of IFN signaling to the restriction of LGTV in these cells. Initial assessment of the host response with addition of exogenous IFN led to less

111

LGTV replication and suggested a role for IFN in protecting cells from infection. While interferon protection is generally expected in host cells, these studies were necessary as an initial step to implicating immune signaling as the underlying reason for restriction in P. leucopus. Interestingly, antiviral assays showed low induction of IFN mediators accompanying virus infection although correlates of host protection released in response to SENV infection were higher in P. leucopus cells than the control. Overall, the low levels of IFN observed with LGTV infection were not surprising due to previous reports that infection with TBFVs is associated with low IFN production likely due to the ability of the virus to ‘hide’ within enclosed compartments during infection (232,234). This observation indicated that viral measures of immune evasion could be occurring in a similar mechanism in both cell types. Nonetheless we performed a series of detailed experiments to directly test the ability of LGTV NS5 to function as an IFN antagonist in the reservoir species. Preliminary observation of pSTAT1 in LGTV infected cells showed decreased protein levels during infection although the decrease was less drastic in P. leucopus cells – however, this might be explained by higher pSTAT1 expression in these cells. Overexpression of NS5 in P. leucopus cells diminished the ability of these cells to respond to infection and upregulate an ISG. This suggests that NS5 functions in P. leucopus cells and restriction cannot be attributed to a non-functional IFN antagonism.

Indeed, these data further hint that restriction may be occurring through an impressive ability of the reservoir host cells to override the viral antagonism mechanisms and counter-respond to infection. The specific mechanism of overcoming NS5 antagonism is likely an elaborate one and could potentially involve host proteins targeting alternative

112

viral factors to halt the replication process. Based on the replication restriction data, it is possible that enzymatic functions to replicate the genome are potently targeted in response to IFN antagonism as an over-compensation attempt by the host cells to make up for diminished IFN signaling. Essentially, IFN antagonism could alert the cells of the reservoir host to efficiently block replication. Further studies will be required to test this hypothesis by using viral strains lacking the IFN antagonism function in NS5. It is also worthy to note that viral evasion and antagonism occurs via multiple routes through various viral proteins such as NS3; therefore, extensive studies will be beneficial to reveal whether other pertinent antagonism processes also occur in P. leucopus or if the virus is only able to utilize NS5 in these cells thereby having fewer antagonism options compared to the susceptible control cells.

The underlying inference from these studies suggest that proviral factors are available to mediate virus replication in these cells and restriction is likely due to counter- response. The ability of a species to demonstrate disease resistance could be linked to passive tolerance of virus infection or active suppression by the host. Till date, the mechanism of TBFV restriction had not been resolved and no studies had differentiated whether resistance occurs due to an active host response or lack of host factors. These studies overcame a major hurdle in the study of P. leucopus which is the brevity of genetic information and the fact that many immune signaling genes are undefined in this species.

We successfully cloned and sequenced relevant antiviral genes in P. leucopus for the first time and studied their response signatures when challenged with interferon treatment and virus infection. As expected, most of these genes were upregulated by IFN

113

treatment and STAT1 was significantly induced at the mRNA level thus suggesting that these genes have similar expression patterns and responses in the susceptible cells. Thus, any potential difference in functionality will likely be bestowed through distinct interactions or modifications occurring within P. leucopus cells. Interestingly, STAT1 protein level was consistently visuallhy observed to be higher and more induced in P. leucopus cells compared to the M. musculus control. This might potentially point to differences in gene responses between these cells. It is particularly interesting that IFIT2 protein size was visibly different between both cell types. The data from sequence analysis show that the antiviral genes assessed in this study have a high percent identity (ranging from 75-90%), however, further studies are needed to determine the functional implication of the sequence variations observed in P. leucopus.

Having the novel sequence data, proceeded to test the specific impact of IFN signaling to virus infection in P. leucopus. We accomplished this by targeting two key players in the IFN response pathway – STAT1 and IFNAR1, resulting in significantly higher virus replication suggesting that LGTV can adequately replicate in these cells but is hindered under normal conditions by the host IFN response. Importantly, this section distinctly contributes an explanation for LGTV restriction as occurring through an antiviral axis and not a proviral one. These findings provide a premise to probe into the discovery of specific ISGs that potentially restrict virus infection

This chapter also describes novel homolog identification and initial characterization of the only known TBEV-specific restriction factor in the naturally resistant host species P. leucopus. Sequence comparison showed only about 75% identity

114

between mouse and P. leucopus TRIM79, yet the protein retains its ability to bind the

LGTV protein NS5 suggesting that the identity between the gene in this species likely preserves function. Interestingly, we did not detect direct degradation of viral NS5 by plTRIM79 suggesting that potential restriction factor function occurs through alternate mechanisms or cell types. Results of the ubiquitination assay suggest that NS5 has a decreased ubiquitination status when co-expressed with plTRIM79 however, the impact of this finding remains to be further elucidated. Nonetheless, post-translational modifications are very essential to define the role of specific viral proteins in a given cell compartment and these functions are likely time-dependent, based on the stage of virus infection. Any impact of plTRIM79 on the functionality of NS5 without involving degradation could potentially be mediated by impacting the cellular localization of NS5, competitively binding an interacting partner of the viral protein or preventing its enzymatic function.

Moreover, impact on the post-translational modifications of NS5 would likely involve the action of other host proteins such as deubiquitinases. The identification of host restriction factors is hampered by the biology of host cells used in many probes (100) among other factors. Many degradation assays performed to test the role of plTRIM79 as a restriction factor were performed in HEK 293 cells which are of human origin and could potentially mask the function of plTRIM79 if the cells lack similar interacting partners as it would encounter in the native host cells. A rationale for using these cells is due to their ease of transfection and amenability to large amounts of ectopic nucleic acids, however direct molecular targeting in P. leucopus cells will likely provide definitive information on the role of plTRIM79 during virus infection in these species. Overexpression studies

115

performed by generating clonal cell lines of plTRIM79 in P. leucopus cells were hampered by tight regulation of this protein in the host cells. As demonstrated above,

MG132 treatment increased plTRIM79 expression showing that it is regulated by the proteasomal pathway. Growth curve experiments in these cells would require treatment of the clonal cells with MG132 to allow optimal overexpression of plTRIM79 however,

MG132 treatment alone had other impacts on the cells that interfere with interpretation of results. A direct approach to implicate plTRIM9 as a restriction factor in P. leucopus will involve gene knockdown to assess a potential increase in virus replication. Overall, this section highlights the first TRIM protein identified in P. leucopus showing similar expression pattern and host cell regulation in the reservoir species; interaction with the

TBFV virulence factor NS5 provides an initial indication it may function as a restriction factor in this species, pending further studies.

The data in sections 4.2.1-4.2.5 form part of a recently-published manuscript for this dissertation (230)

116

Figure 4-1 IFN treatment protects P. leucopus cells from virus infection P. leucopus cells were mock- or pre-treated with 10IU or 1000IU of mIFNβ prior to infection with VSV at MOI 0.1 and virus was quantified by a plaque assay at the indicated time points (A) and the cytopathic effect (CPE) of virus replication was imaged at 24 hpi comparing untreated cells to 1000IU mIFN-β treatment (B). Cells were further tested for mIFNβ protection from LGTV infection at MOI 50 and expressed as the % infection compared to mock-treated (C). An antiviral assay was performed to compare cellular protection by supernatants from mIFN-β, SENV or LGTV-challenged P. leucopus against further VSV replication (D) and the correlates of protection compared to an IFN β standard is quantified in (E).

117

Figure 4-2 The LGTV NS5 protein is an IFN antagonist in P. leucopus cells P. leucopus and M. musculus cells were challenged with LGTV infection and IFN treatment after which protein was harvested and probed for pSTAT1 expression (A). P. leucopus cells were transduced with lentiviruses over-expressing LGTV NS5-GFP fusion or GFP vector only. Cell clones were screened for GFP expression by immunoblot (B) after which NS5 #1 and GFP #2 were chosen for further experiments. Clonal NS5-GFP cells were pretreated (or not) with mIFN-b for 16 h prior to infection with VSV for an additional 24 h. Supernatants from virus-infected cells were titrated by a plaque assay (C). Further, cells were treated with mIFN-b (8 h) or infected with LGTV (24 h) and expression of plTRIM79 was assessed by RT-qPCR using gene-specific primers (D). Mean ± SD, data are from three independent experiments performed in triplicate. Asterisks indicate: * = p < 0.05, ** = p <0.01, *** = p <0.0001.

118

Figure 4-3 Homologs of antiviral genes are expressed in P. leucopus Homologs of antiviral signaling genes were cloned from P. leucopus and detected on an agarose gel (A). RT-qPCR data show relative fold induction of STAT1, RIG-I, IFNAR1, MAVS, IRF-1 and TRIM79 in P. leucopus cells that were treated with mIFN-b for 8 h and 24 h (B) or infected with LGTV for 24 h and 72 h respectively (C). Cellular cDNA was probed using gene-specific primers designed from gene sequences in P. leucopus. Data were quantified relative to mock- treated/infected cells, normalized to P. leucopus beta-actin levels and are representative of three independent experiments performed in triplicate. Asterisks indicate: * = p < 0.05, ** = p <0.01, *** = p <0.0001

119

Figure 4-3 continued Protein expression of ISGs was assessed in the reservoir host by immunoblotting. P. leucopus and M. musculus cells challenged with mIFNβ, LGTV, VSV or SENV for 8 h or 24 h respectively and assayed with gene-specific antibodies to IFIT2 and IFIT3 (D). Cells were also probed for STAT1 expressed in P. leucopus and M. musculus cells infected with LGTV for 24 h and 48 h (E), or challenged with mIFN-b (F).

120

Figure 4-4 Knocking down STAT1 relieves restriction in P. leucopus cells Newly-resolved sequence of STAT1 was targeted by shRNA technology. Cell lines were generated by lentiviral transduction and cell lysates were assessed for knockdown by qPCR using gene-specific primers (A) and by immunoblotting (B) showing expression of STAT1 and STAT1- P in knockdown cells targeted with various short hairpins (#1, #2 and ALL) compared to the non- silencing (NS) control. RT-qPCR of ISG plTRIM79 was performed in STAT1KD and NS control cells treated with mIFN-b (C). Data are shown as relative fold induction of plTRIM79 compared to mock-treated cells. STAT1KD cells and control (NS) cells of P. leucopus were infected with LGTV at MOI 100 for 72 h and immunostained for the viral E protein (green), NS3 protein (red) and nuclei were stained with DAPI (blue). Cells were visualized by confocal microscopy at 40X magnification (D).

121

Figure 4-4 continued LGTV-infected STAT1 KD and NS KD cells were also probed by immunoblot for the viral E protein (E). To quantify virus replication, the STAT1 KD and NS cells were infected with LGTV at MOI 10 and virus released was quantified by an immunofocus assay (F) while accumulation of viral –ve strand RNA was assessed by qRT-PCR (G). Mean ± SD, Data are from three independent experiments performed in triplicate. Asterisks indicate: ** = p<0.01, *** = p<0.001, **** = p<0.00001.

122

Figure 4-5 Knocking down IFNAR1 relieves restriction in P. leucopus cells IFNAR1 was targeted by shRNA technology using lentiviruses and cell lysates were assessed for knockdown by qPCR using gene-specific primers (A). IFNAR1 KD cells and control (NS) cells of P. leucopus were infected with LGTV for 72 h and immunostained for the viral E protein (green), NS3 protein (red) and nuclei were stained with DAPI (blue). Cells were visualized by confocal microscopy at 20X magnification (B). To assess viral replication, IFNARI KD and NS P. leucopus cells were infected with LGTV at MOI 10 and quantified by an immunofocus assay (C) while viral negative sense RNA was measured by qRT-PCR. Mean ± SD, Data are from three independent experiments performed in triplicate. Asterisks indicate: ** = p<0.01, *** = p<0.001, **** = p<0.00001.

123

Figure 4-7 plTRIM79 is regulated by the proteasome in P. leucopus Primers to detect TRIM79 in P. leucopus were designed using a consensus sequence from the Peromyscus EST library and the gene could be cloned out of IFN-treated P. leucopus cells and imaged on a transilluminator (A). To assess the gene regulation pathway, clonal cells were generated expressing a plTRIM79-GFP fusion or GFP vector only. The cells were treated (or not) with 10µm of the proteasomal inhibitor MG132 for 4h prior to imaging the cells (B).

124

Figure 4-8 plTRIM79 interacts with the LGTV NS5 protein HEK-293 cells were transfected with LGTV NS5-mCherry and plTRIM79-GFP fusions, fixed and assessed for co-localization by immunofluorescence microscopy (A). Similarly, cells expressing LGTV NS5-AP and plTRIM79-mCherry were probed by a co-affinity purification assay and blotted using antibodies against the respective tags (B) An empty GFP vector was used as filler DNA to ensure that equal amount of DNA was transfected.

125

Figure 4-9 plTRIM79 changes the ubiquitination level of LGTV NS5 Testing the impact of co-expressing plTRIM79 and LGTV NS5 by performing a ubiquitination assay with LGTV NS5-V5, plTRIM79-GFP and HA-tagged ubiquitin and immunobloted for protein expression (A), a degradation assay expressing steady amount of LGTV NS5 with increasing amounts of plTRIM79 and mouse TRIM79 for comparison (B) and an immunofluorescence assay staining for LGTV NS5-mCherry, plTRIM79-GFP and lysosomal marker LAMP1, imaged by confocal microscopy to detect co-localization (C.).

126

Table 4-1. Comparison of P. leucopus gene homologs to M. musculus genes

Summary of alignment data showing the percentage identity of newly resolved P. leucopus gene homologs to their M. musculus counterparts. Accession numbers of the gene references are shown on the far right of the table.

P. leucopus gene P. leucopus % Identity M. musculus accession number accession number STAT1 KY451962 90% NM_001205313.1 IFNAR1 KY451965 78% NM_010508.2 RIG-I KY451966 88% NM_172689.3 MAVS KY451963 82% NM_144888.2 IRF1 KY451964 89% NM_008390.2 TRIM79 KY451967 75% XM_006507549.2

127

Chapter 5 Preliminary assessment of the role of oxidative stress responses on virus replication in P. leucopus

5.1 Introduction

In addition to the classical innate antiviral response posing a challenge to viruses invading a host cell, variations in other biological processes can potentially impact the fate of virus infection. However, these are not likely to be through direct efforts of the signaling pathways to mitigate virus replication rather occurs as a by-product of intrinsic host homeostatic strategies. Indeed, it has been recognized that among multiple factors potentially defining the ability of a species to serve as virus reservoirs, one of the main reasons could be due to intrinsic disease tolerance stemming from the reservoir simply withstanding the deleterious effects of infection thereby surviving and maintaining virus at low levels (194). An intersection between biological homeostasis and virus infection has been appreciated in viruses particularly with regard to the expression of reactive oxygen species and induction of oxidative stress as described above

Although ROS is not detrimental in low amounts, excessive production can become pathological, therefore tight control is maintained via the action of antioxidants and this balance is essential for protection of nucleic acids and the cellular environment

(160). A key role for maintaining the oxidative balance in a host is defined in the hypothesis of aging whereby the ability of a given species to live long is linked to its intrinsic ability to withstand oxidative stress, maintain low levels of ROS and concurrently

128

possess high antioxidant activity. Based on the permissiveness of ROS-rich environments to flavivirus infection, it can thus be postulated that species with intrinsically low ROS levels and high antioxidant activities likely present an unconducive environment for virus replication.

Interestingly, much of the aging hypothesis has been elucidated using P. leucopus as a model and comparing physiological responses to M. musculus. Several studies have shown that P. leucopus mice not only live more than two fold longer than M. musculus mice, but also display oxidative stress resistance, high antioxidant activity, and low mitochondrial stress in vitro and ex vivo (206,235,236). These factors together have helped to explain the longer life span of P. leucopus mice according to the hypothesis of aging.

Based on the relationship between ROS levels and virus infection, the low ROS levels in

P. leucopus cells and concurrent low level of virus infection compared to the susceptible shorter lived M. musculus cells, we hypothesized that TBFV potentially prefers an ROS- rich host cell for replication and attempts to increase ROS levels during infection but is antagonized in P. leucopus cells due to high antioxidant activity and low oxidative stress potential in this species.

5.2 Results

5.2.1 LGTV infection increases ROS levels in M. musculus cells to higher levels than in P. leucopus cells

Previous studies have shown a higher basal level of cellular ROS activity in M. musculus cells than in P. leucopus cells including both cellular and mitochondrial ROS. In order to

129

test how these levels are influenced by virus infection, we infected both cell types with

LGTV and measured cellular ROS at early (Figure 5-1A) and late time points (Figure 5-

1B) post infection. The results show that while there is an increase of ROS at early times after infection, these levels remain steady in P. leucopus cells but progressively increase in the susceptible M. musculus cells. This suggests that the virus infection induces sustained

ROS responses in the susceptible cells but not in the reservoir species.

5.2.2 Mitochondrial ROS expression in M. musculus and P. leucopus cells infected with LGTV virus

To determine the impact of virus infection on mitochondrial ROS in the cells, we infected both cells with LGTV for 8 hours and loaded with a dye to detect mitochondrial ROS.

Cells were then washed and visualized for ROS expression. There wasn’t a drastic increase in the mitochondrial ROS levels following virus infection in both cell types and expression levels did not appear markedly different between the cell types either. This suggests that the cellular ROS, rather than mitochondrial ROS is modulated by virus infection. Interestingly, there appeared to be a distinct change in the staining localization in the susceptible cells compared to the resistant cell line (Figure 5-2).

5.2.3 ROS induction in P. leucopus cells increases LGTV replication

To directly test the impact of ROS induction on virus replication in the reservoir species, cells were treated with ROS inducing agents (high glucose concentration and ceramide) prior to virus infection. The infected cells were stained and quantified for the number of

130

virus-positive cells. The data shows that there was higher number of infected cells in the drug-treated cells compared to the mock-treated cells (Figure 5-3A). In another experiment, we pre-treated cells with a ceramide and further infected the treated (or untreated) cells with LGTV for 24-72 hours. Viral supernatants were assessed for LGTV replication and we observed significantly higher levels of LGTV replication in the ceramide-treated cells than the untreated cells Figure 5-3B) suggesting that ROS induction promoted virus replication in P. leucopus cells.

5.3 Chapter 5 discussion

This section presents preliminary data comparing the ROS response in P. leucopus cells and M. musculus cells of which previous studies have demonstrated the former to have low ROS levels and high resistance to the toxic effects of ROS accumulation compared to the latter. Our current studies examining virus infection in both species presents a similar model used in the aging pair studies. In our hands, TBFV and KUNV have shown significantly lower replication in P. leucopus cells and this restriction occurs at the actual replication phase whereby the negative strand RNA of LGTV is significantly less than in the susceptible control. Interestingly, Gullberg et al., 2015 showed that oxidative stress induction during KUNV infection is pertinent for RNA capping and replication such that treatment of KUNV-infected cells with antioxidant reduced the accumulation of negative strand RNA (168). On the other hand, Kuzmenko, 2016 showed that TBEV NS1 alters the redox balance during infection. We rationalize that the restriction of ROS-requiring viruses in an ROS-diminished species P. leucopus is not coincidental and that the low

131

ROS in this species likely impacts virus restriction (167). Furthermore, our findings show that the innate immune response in the reservoir species mediates virus restriction; based on the intersection of ROS and immune signaling, we hypothesize that virus is unable to co-opt ROS in P. leucopus cells thereby allowing a potent counter- immune response in this species despite NS5 antagonism. This redox distinction between the resistant and susceptible species might likely provide an explanation for why two species with similar virus entry, immune responses and viral antagonism demonstrate a disparity in viral susceptibility. The initial set of experiments shown here suggest that upon LGTV infection in both cell types, the ROS levels increase considerably over the course of infection in M. muscuuls cells while they remain relatively steady in P. leucopus cells suggesting that the antioxidant response in the reservoir species likely prevents continuous increase of ROS.

Furthermore, while there wasn’t a drastic difference in the mitochondrial ROS during infection, there appeared to be a re-distribution of mitochondrial ROS during infection of

M. musculus compared to P. leucopus cells although further studies are required to clearly elucidate the mitochondrial ROS response. To link ROS expression with virus infection in

P. leucopus cells, we have shown that treating the cells with inducers of ROS increases virus replication. Future studies intend to fully implicate ROS levels as determinants of infection in the reservoir species, mechanistic studies will reveal the distinct impact of

ROS induction and antioxidant signaling in these cells. It is possible that the naturally higher levels of antioxidant activity in the reservoir exerts a similar effect as observed by

Gullberg et al., leading to low accumulation of negative strand RNA. The direct link of this phenotype to the IFN response pathway will remain to be determined. If the IFN

132

signaling response intersects with antioxidant effects during virus infection in P. leucopus, then infecting the IFNARI KD cells should theoretically lead to higher ROS induction and lower antioxidant activity to potentially explain the higher levels of virus replication in the clonal cells. Similarly, treatment of P. leucopus IFNAR1 KD cells with antioxidants could potentially lead to significantly lower virus replication in the KD cells compared to the NS control cells – likely associated with a variable gene expression profile during infection.

Many questions remain to be answered in this aspect of the project but the preliminary findings presented herein form the basis for further probe into alternate biological processes likely impacting the host response to virus infection in P. leucopus cells.

133

Figure 5-1 LGTV increases cellular ROS to higher levels in M. musculus cells The level of ROS induction in response to LGTV infection as assessed to determine oxidative stress. P. leucopus and M. musculus cells were infected with LGTV (or not) for 2 h (A) or 24 and 72 h (B). At the indicated time points, the viral supernatant was replaced with phenol-free media before cells were loaded with H2CFDA or CM-H2CDFA dye respectively and immediately imaged by confocal microscopy (20X) magnification. Green fluorescence represents cellular ROS expression.

134

Figure 5-1 Mitochondrial ROS expression is unchanged in response to LGTV infection Mitochondrial ROS was visualized in response to LGTV infection in P. leucopus and M. musculus cells infected for 8 h. Viral supernatant was removed, cells were washed and loaded with the dye to probe mitochondrial ROS. The cells were counter-stained with DAPI and ™ﱞMitoSOX immediately visualized. Red fluorescence connotes areas of mitochondrial-derived ROS while blue represents the nucleus. Images were taken at 20X magnification.

135

Figure 5-3 Treating P. leucopus cells with ROS inducers increases virus replication The effect of ROS induction on virus replication was tested in P. leucopus cells by pre-treating with ROS-inducing agents prior to virus infection. A) Cells were treated with 32mM glucose or 20µM ceramide for 24 hours after which the cells were infected with LGTV at MOI 10 for 72 hours before cells were fixed and stained for the viral E protein. The number of virus-positive was quantified and compared to the untreated cells (A). Cells were also treated with 40µM ceramide and infected with LGTV at MOI 10 for 24, 48 and 72 hours. Viral supernatants were collected and quantified by an immunofocus assay (B).

136

Chapter 6 Discussion and Summary

Many of the notorious viruses associated with immense human suffering and global health burden are zoonotic pathogens that persist in the environment within animals before being transmitted to humans. Interestingly, many animal species remain relatively healthy despite being infected with lethal pathogens and ensure that virus is available for transmission directly or via secondary arthropod vectors. The underlying mechanisms allowing animal reservoirs to withstand infection remain enigmatic and few studies have directly compared virus replication and pathogenesis between susceptible and resistant hosts to tease apart the molecular basis of resistance. This paucity of extensive reservoir host studies likely comes from the lack of defined experimental models to accurately recapitulate the natural phenomenon of resistance and more importantly, the limited availability of research regents to study novel species at the molecular level. Emerging viruses are of particular concern to human health because we are yet to gain a clear understanding of the primary factors that promote the emerging nature of these viruses.

While various hypotheses have been presented over time and range from environmental to genetic factors, it remains logical that in order to protect the human population from outbreaks of viral disease, we must remain one step ahead of the viral pathogens by having a clear understanding of their pathogenesis, immune response gaps, viral evasion strategies and host counter-responses.

137

The flavivirus group contains some of the most clinically-important arboviruses and can be found on every continent. Over the years, a great deal of research efforts have focused on understanding viral pathogenesis and interaction of the virus with host cell factors in order to design therapeutics; these studies have truly provided a clearer clinical picture of severe infection cases, vaccine candidates against some viral strains, in-depth understanding of the virus infection cycle and relationships between the viral entities.

However, there remains no definitive cure for infection with flaviviruses and treatment of clinically recognized cases is limited to palliative care. Of major concern is the fact that through the years, the rate of global distribution and burden of disease contributed by flaviviruses has been on the increase; these viruses continue to emerge in new areas, manifest new clinical symptoms and pose a threat to many populations therefore emphasizing the need for our research approaches to be re-examined and our viral combat strategies reviewed. Many of the afore-mentioned studies to elucidate flavivirus pathogenesis and host responses to infection have been performed in murine and human cell models of infection. Through a variety of screens and mechanistic studies, several viral restriction factors have been identified to have broad or specific action in preventing flavivirus replication and these studies suggest that the host is able to launch a proper response against invading flaviviruses. It is therefore curious that although these hosts under study express the restriction factors identified so far, they remain susceptible to infection and this is likely because the host species used in the restriction factor studies are susceptible to disease thereby suggesting that under normal conditions, the restriction factors expressed are not sufficient to prevent disease pathogenesis. This emphasizes the

138

fact that more studies ought to be performed in naturally-resistant host species which likely express more potent restriction factors in order to maintain resistance. These studies will not only reveal ways to incapacitate the virus and inform the development of vaccine strains, but they will likely highlight more potent homologs of the already-known restriction factors or new restriction factors which will provide knowledge for the design of an efficient peptide mimetic against flaviviruses. Essentially, if we are looking for resistance determinants to enable us design flavivirus therapeutics, then we should be looking in hosts that are naturally resistant to infection in order to learn what their immune response does differently to prevent disease pathogenesis. The studies described in this project represent a first step towards studying reservoir species to determine the mechanism of flavivirus restriction.

The experimental model used a rodent species abundant in North America, and able to harbor the North-American TBFV- POWV. In addition to its relevance in the

POWV-endemic region, P. leucopus was also selected for its ability to harbor other pathogens while remaining relatively healthy – these factors make it an interesting candidate with which to investigate genetic determinants of resistance. On the other hand, this model included a susceptible control M. musculus widely used in flavivirus studies to examine viral pathogenesis. Experiments performed in vivo by other groups have clearly shown the ability of P. leucopus to withstand infection with POWV via various routes of inoculation while control mice succumbed to neurological disease. While these studies have served to reiterate the previously-known involvement of P. leucopus as a reservoir of

139

POWV, they have not demonstrated a mechanistic basis for these observations likely because a molecular approach is needed to distinctly identify the pathways of resistance in this species. This study therefore has set up an in vitro model which aptly recapitulates the in vivo phenomenon of not only the BSL2 strain LGTV, but the more virulent TBFVs

POWV I, POWV II, and TBEV where virus infection occurred to significantly lower levels in cells of the reservoir host compared to the susceptible control. Importantly, another member of the flavivirus family KUNV was restricted in P. leucopus cells while the rhabdovirus VSV was not restricted. This cell culture model therefore showed specificity for flavivirus restriction and presents an opportunity to test the potential mechanistic reasons for virus restriction. To date, no in-depth studies have been performed to delineate the progress of virus through its infection cycle in the reservoir species – our work presents a novel report of systematically comparing various stages of LGTV infection in P. leucopus and M. musculus cells. Initially observing significantly low virus release in P. leucopus cells raised questions as to whether the reservoir cells were at all permissive to virus infection or expressed a viral receptor. Our entry studies showed that entry does occur in these cells and that equal amounts of virus can enter the susceptible and resistant cells following infection. Entry was also inhibited in both cell types by blocking the endocytic trafficking pathway thereby suggesting that virus enters both cells via the same route. Although equal amount of virus is seen at the entry step, there was significantly less virus being released thereby suggesting that an intracellular factor blocks virus accumulation within the P. leucopus cells. We established that this was not due to an inability of viral RNA to become translated in these cells because direct delivery of viral

140

RNA led to expression of viral protein. Ultimately, the viral restriction was identified at the replication stage as seen by reduced accumulation of viral negative strand RNA thus culminating in an overall lower number of virions released from P. leucopus cells. These studies not only demonstrate that step-wise analysis of virus infection in these species can provide a proper focus point for further examining the viral restriction, they also highlighted to us that virus restriction was dose-dependent since infecting the reservoir cells at a higher MOI led to higher virus replication. Nonetheless, the distinction of whether restriction occurred due to a lack of proviral factors or high expression of antiviral factors remained in question at this point. Based on flavivirus biology and the immune response to infection, we reasoned that potent antiviral responses in the reservoir are likely mediated via the IFN signaling cascade and we therefore tested potential reasons for restriction. Interestingly, IFN was not produced to high amounts following infection although addition of exogenous IFN was able to protect the cells from virus replication thus suggesting that IFN does play a role in virus restriction. A major factor determining the fate of flavivirus replication is its ability to antagonize IFN signaling particularly through the NS5 protein – we rationalized that if this antagonism does not occur in P. leucopus cells due to a lack of/overabundance of proviral and antiviral host factors respectively, then that might explain restriction. Interestingly, our data show that NS5 functions as an antagonist in P. leucopus cells therefore offering fresh perspective that

NS5 functionality can be upheld in a resistant host compared to the current scenario with

DENV where murine resistance is associated with the inability of NS5 to block IFN signaling. These findings particularly suggest that the host launches a potent counter-

141

response to NS5 thereby maintaining virus at low levels even though NS5 functions in both the susceptible and reservoir species. Indeed, further studies can be performed to directly compare the extent of NS5 functionality in both cell types but the current model suggested to us that since NS5 has the proviral factors required to function in these cells, then an antiviral response is likely responsible for restriction in which case blunting the antiviral response should relieve restriction.

A major hurdle faced in gene manipulation was the brevity of genetic information available for P. leucopus cells. This project therefore pioneered the identification of P. leucopus homologs of antiviral signaling genes and tested their expression in response to infection. We anticipate that the sequence information generated from these studies will be beneficial for further probe aimed at understanding the immune response in the yet enigmatic reservoir host P. leucopus. We ourselves utilized the novel sequence information to design knockdown tools targeting the IFN response pathway in P. leucopus cells. Infection studies in the knockdown cells supported our hypothesis that restriction was mediated by an antiviral response since virus replicated to significantly higher numbers in the knockdown cells. Remarkably, visualizing viral protein staining in the P. leucopus IFNAR1 KD cells looked very similar to the WT M. musculus cells suggesting that the reservoir species can indeed support high levels of virus replication but the antiviral response suppresses the infection in these cells. It is worth stating that the novelty of this project does not merely lie in identifying that IFN signaling prevents flavivirus replication as this has been a known fact for many years. Rather, the innovative idea

142

represented here is demonstrating that IFN signaling confers protection in a resistant species suggesting that the outcome of signaling likely differs from what occurs in the susceptible species that have been used in prior studies. This difference is likely represented in the identity, modifications, interactions and mechanistic action of the ISGs induced during IFN signaling in both species. Indeed, IFN induction in P. leucopus cells offered twice the protection seen in the antiviral assays. Interestingly, low amount of IFN was induced following LGTV infection which was not surprising given viral evasion strategies however, it does suggest that whatever low amount of IFN are produced in these cells beyond the limit of our detection is likely sufficient to elicit a response against virus infection. The exact nature of this response is the next aim of this project – to accurately delineate the specific genes induced in response to virus infection which could potentially function differently than the M. musculus genes to restrict virus. Based on the finding that restriction occurs at the step of RNA synthesis, the restriction factors expressed in P. leucopus cells could potentially act by disrupting the conformation and/or localization of the viral replication complex through an interaction with one or more viral proteins in the complex. There could also be degradation of viral components to prevent the progress of virus replication. NS5 being the polymerase which functions with the helicase activity of

NS3 could be targeted in these cells via a block to its enzymatic action. Indeed, this restriction could be occurring in a variety of ways but the results presented by this project provide a focal point for further restriction factor studies. An initial step to characterizing restriction factors in P. leucopus cells was performed with TRIM79 a restriction factor identified in mouse cells to restrict TBEV infection. The P. leucopus homolog of this gene

143

retained co-localization and binding to LGTV NS5. Direct degradation assay in human cells did not show NS5 degradation by plTRIM79 but curiously, a ubiquitination assay suggested that NS5 expression was diminished when co-expressed with plTRIM79. While the mouse TRIM79 is not known to function by ubiquitination, the mechanism might be different with plTRIM79. Additionally, the P. leucopus gene might function in a cell-type dependent manner based on potentially necessary interactions with other host factors which might help explain the different infection outcome in both cells. Ultimately, further mechanistic studies and gene targeting processes are needed to unequivocally implicate plTRIM79 as a restriction factor in the reservoir host.

We strongly believe that multiple factors are likely at play in mediating the restriction phenotype in P. leucopus. The latter portion of this project involved preliminary characterization of alternative factors contributing to low virus replication in the reservoir host by taking known host biology into account. TBEV has recently been implicated as an inducer of oxidative stress via the NS1 protein while other flaviviruses have also been shown to thrive and replicate in high oxidative stress conditions during infection. On the other hand, P. leucopus has been characterized as having low oxidative stress potential and high antioxidant activity. We therefore made the correlation that the low ROS levels in P. leucopus cells might be hindering the agenda of flaviviruses during infection in P. leucopus cells. Indeed virus infection of P. leucopus cells was associated with lower ROS induction over the course of infection compared to M. musculus cells.

Also, treating the reservoir cells with ROS inducers led to higher virus replication thus

144

suggesting a link between ROS expression and virus replication. The ROS studies presented here are only initial findings but hold promise to delineating the sum-total of factors impacting virus replication in P. leucopus cells. Further mechanistic studies and identifying potential links of ROS induction and IFN signaling in P. leucopus cells will likely provide more insight into the determinants of virus pathogenesis.

Overall, these studies have set the basis for further investigation of reservoir host resistance to flavivirus infection. The cell culture model described herein will likely be beneficial to studying other viral entities such as hantaviruses and also provide a platform for extensive mechanistic studies to identify more potent restriction factors. There could also be far-reaching benefits to the study of other pathogens as P. leucopus is a reservoir for multiple pathogens Borellia spp. In addition to testing mouse reagents in P. leucopus, this project has also developed and several P. leucopus reagents and assay systems ranging from gene sequences, primers, shRNA-expressing lentiviruses and cell lines.

Importantly, the idea represented herein could also be extrapolated to setting up simple yet beneficial models for other pathogen-reservoir host pairs with the aim to determining the genetic and mechanistic basis of disease resistance in order to inform therapeutic design.

There are likely many answers to questions posed in infectious disease research embedded within other natural species, the task remains to take a closer look at these species and gain the answers we seek.

145

References

1. Mackenzie JS, Gubler DJ, Petersen LR. Emerging flaviviruses: the spread and resurgence of Japanese encephalitis, West Nile and dengue viruses. Nat Med. 2004 Nov 30;10(12s):S98–S109.

2. Lindenbach BD, Rice CM. Molecular biology of flaviviruses. Elsevier; 2003. pp. 23–61. (Advances in Virus Research; vol. 59).

3. Gritsun TS, Lashkevich VA, Gould EA. Tick-borne encephalitis. ANTIVIRAL RESEARCH. 2003 Jan;57(1-2):129–46.

4. Kaufmann B, Rossmann MG. Molecular mechanisms involved in the early steps of flavivirus cell entry. Microbes and Infection. Elsevier Masson SAS; 2011 Jan 1;13(1):1–9.

5. Kuhn RJ, Dowd KA, Post CB, Pierson TC. Shake, rattle, and roll_ Impact of the dynamics of flavivirus particles on their interactions with the host. Virology. Elsevier; 2015 May 1;479-480(C):508–17.

6. Pierson TC, Diamond MS. Molecular mechanisms of antibody-mediated neutralisation of flavivirus infection. Expert Rev Mol Med. Cambridge University Press; 2008 May 12;10:e12.

7. Roby JA, Hall RA, Setoh YX, Khromykh AA. Post-translational regulation and modifications of flavivirus structural proteins. Journal of General Virology. 2015 Jul 1;96(7):1551–69.

8. Markoff LEA. A Conserved Internal Hydrophobic Domain Mediates the Stable Membrane Integration of the Dengue Virus Capsid Protein. 1997 Jun 6;:1–13.

9. Dokland T, Walsh M, Mackenzie JM, Khromykh AA, Ee K-H, Wang S. West Nile Virus Core Protein. Structure. 2004 Jul;12(7):1157–63.

10. Bhuvanakantham R, Ng M-L. Analysis of self-association of West Nile virus capsid protein and the crucial role played by Trp 69 in homodimerization. Biochemical and Biophysical Research Communications. 2005 Apr;329(1):246– 55.

11. Martins IC, Gomes-Neto F, Faustino AF, Carvalho FA, Carneiro FA, Bozza PT, et al. The disordered N-terminal region of dengue virus capsid protein contains a lipid-droplet-binding motif. Biochem J. 2012 Jun 15;444(3):405–15.

146

12. Cherrier MV, Kaufmann B, Nybakken GE, Lok S-M, Warren JT, Chen BR, et al. Structural basis for the preferential recognition of immature flaviviruses by a fusion-loop antibody. EMBO J. 2009 Aug 27;28(20):3269–76.

13. Zhang Y, Kaufmann B, Chipman PR, Kuhn RJ, Rossmann MG. Structure of Immature West Nile Virus. Journal of Virology. 2007 May 11;81(11):6141–5.

14. Westaway EG MJKMMK, AA K. Ultrastructure of Kunjin Virus-Infected Cells: Colocalization of NS1 and NS3 with Double-Stranded RNA, and of NS2B with NS3, in Virus-Induced Membrane Structures. Journal of Virology. 71st ed. 1997 Aug 14;:1–12.

15. Watterson D, Modhiran N, Young PR. The many faces of the flavivirus NS1 protein offer a multitude of options for inhibitor design. ANTIVIRAL RESEARCH. Elsevier B.V; 2016 Jun 1;130(C):7–18.

16. Melian EB, Hinzman E, Nagasaki T, Firth AE, Wills NM, Nouwens AS, et al. NS1' of Flaviviruses in the Japanese Encephalitis Virus Serogroup Is a Product of Ribosomal Frameshifting and Plays a Role in Viral Neuroinvasiveness. Journal of Virology. 2010 Jan 11;84(3):1641–7.

17. Mlera L, Melik W, Bloom ME. The role of viral persistence in flavivirus biology. Pathogens Disease. 2014 May 12;71(2):137–63.

18. Mackenzie JMEA. Subcellular Localization and Some Biochemical Properties of the Flavivirus Kunjin Nonstructural Proteins NS2A and NS4A. 1998 May 27;:1– 13.

19. Zmurko J, Neyts J, Dallmeier K. Flaviviral NS4b, chameleon and jack-in-the-box roles in viral replication and pathogenesis, and a molecular target for antiviral intervention. Rev Med Virol. 2015 Apr 1;25(4):205–23.

20. Yoshii K, Sunden Y, Yokozawa K, Igarashi M, Kariwa H, Holbrook MR, et al. A Critical Determinant of Neurological Disease Associated with Highly Pathogenic Tick-Borne Flavivirus in Mice. Journal of Virology. 2014 Apr 22;88(10):5406– 20.

21. Acosta EG, Kumar A, Bartenschlager R. Revisiting Dengue Virus–Host Cell Interaction. Elsevier; 2014. pp. 1–109. (Advances in Virus Research; vol. 88).

22. Smit J, Moesker B, Rodenhuis-Zybert I, Wilschut J. Flavivirus Cell Entry and Membrane Fusion. Viruses. 2011 Dec;3(12):160–71.

23. Mukhopadhyay S, Kuhn RJ, Rossmann MG. A structural perspective of the

147

flavivirus life cycle. Nat Rev Microbiol. Nature Publishing Group; 2005 Jan;3(1):13–22.

24. Heinz FX, Stiasny K. Flaviviruses and their antigenic structure. Journal of Clinical Virology. Elsevier B.V; 2012 Dec 1;55(4):289–95.

25. Stiasny K, Fritz R, Pangerl K, Heinz FX. Molecular mechanisms of flavivirus membrane fusion. Amino Acids. 5 ed. 2009 Nov 1;41(5):1159–63.

26. Klema V, Padmanabhan R, Choi K. Flaviviral Replication Complex: Coordination between RNA Synthesis and 51-RNA Capping. Viruses. 2015 Aug;7(8):4640–56.

27. Welsch S, Miller S, Romero-Brey I, Merz A, Bleck CKE, Walther P, et al. Composition and Three-Dimensional Architecture of the Dengue Virus Replication and Assembly Sites. Cell Host and Microbe. Elsevier Ltd; 2009 Apr 23;5(4):365–75.

28. Miorin L, Romero-Brey I, Maiuri P, Hoppe S, Krijnse-Locker J, Bartenschlager R, et al. Three-Dimensional Architecture of Tick-Borne Encephalitis Virus Replication Sites and Trafficking of the Replicated RNA. Journal of Virology. 2013 May 7;87(11):6469–81.

29. Tay MYF, Smith K, Ng IHW, Chan KWK, Zhao Y, Ooi EE, et al. The C-terminal 18 Amino Acid Region of Dengue Virus NS5 Regulates its Subcellular Localization and Contains a Conserved Arginine Residue Essential for Infectious Virus Production. Rey FA, editor. PLoS Pathog. 2016 Sep 13;12(9):e1005886–35.

30. Moreland NJ, Tay MYF, Lim E, Rathore APS, Angeline P.C. Lim, Hanson BJ, et al. Monoclonal antibodies against dengue NS2B and NS3 proteins for the study of protein interactions in the flaviviral replication complex. Journal of Virological Methods. Elsevier B.V; 2012 Jan 1;179(1):97–103.

31. A Novel in Vitro Replication System for Dengue Virus. 1999 Nov 3;:1–10.

32. Garcia-Blanco MA, Vasudevan SG, Bradrick SS, Nicchitta C. Flavivirus RNA transactions from viral entry to genome replication. ANTIVIRAL RESEARCH. Elsevier B.V; 2016 Oct 1;134(C):244–9.

33. Brinton M. Replication Cycle and Molecular Biology of the West Nile Virus. Viruses. 2014 Jan;6(1):13–53.

34. Heinz FX, Stiasny K. Flaviviruses and their antigenic structure. Journal of Clinical Virology. Elsevier B.V; 2012 Dec 1;55(4):289–95.

148

35. Calisher CH, Karabatsos N, Dalrymple JM, Shope RE, Porterfield JS, Westaway EG, et al. Antigenic relationships between flaviviruses as determined by cross- neutralization tests with polyclonal antisera. Journal of General Virology. 1989 Jan;70 ( Pt 1):37–43.

36. Lobigs M, Diamond MS. Feasibility of cross-protective vaccination against flaviviruses of the Japanese encephalitis serocomplex. Expert Review of Vaccines. 2014 Jan 9;11(2):177–87.

37. Pastorino B, Nougairède A, Wurtz N, Gould E, de Lamballerie X. Role of host cell factors in flavivirus infection: Implications for pathogenesis and development of antiviral drugs. ANTIVIRAL RESEARCH. Elsevier B.V; 2010 Sep 1;87(3):281–94.

38. Wang T, Town T, Alexopoulou L, Anderson JF, Fikrig E, Flavell RA. Toll-like receptor 3 mediates West Nile virus entry into the brain causing lethal encephalitis. Nat Med. 2004 Nov 21;10(12):1366–73.

39. Sips GJ, Wilschut J, Smit JM. Neuroinvasive flavivirus infections. Rev Med Virol. 2011 Nov 16;22(2):69–87.

40. Parquet MC, Kumatori A, Hasebe F, Morita K, Igarashi A. West Nile virus- induced bax-dependent apoptosis. FEBS Lett. 2001 Jun;500(1-2):17–24.

41. Ramanathan MP, Chambers JA, Pankhong P, Chattergoon M, Attatippaholkun W, Dang K, et al. Host cell killing by the West Nile Virus NS2B–NS3 proteolytic complex: NS3 alone is sufficient to recruit caspase-8-based apoptotic pathway. Virology. 2006 Feb;345(1):56–72.

42. Donadieu E, Lowenski S, Servely J-L, Laloy E, Lilin T, Nowotny N, et al. Comparison of the Neuropathology Induced by Two West Nile Virus Strains. Thompson R, editor. PLoS ONE. 2013 Dec 18;8(12):e84473–15.

43. Martina BEE, Koraka P, Osterhaus ADME. Dengue Virus Pathogenesis: an Integrated View. Clinical Microbiology Reviews. 2009 Oct 12;22(4):564–81.

44. Herrero LJ, Zakhary A, Gahan ME, Nelson MA, Herring BL, Hapel AJ, et al. Dengue virus therapeutic intervention strategies based on viral, vector and host factors involved in disease pathogenesis. Pharmacology and Therapeutics. Elsevier B.V; 2013 Feb 1;137(2):266–82.

45. Hapuarachchi HC, Oh HML, Thein TL, Pok K-Y, Lai Y-L, Tan L-K, et al. Clinico-genetic characterisation of an encephalitic Dengue virus 4 associated with multi-organ involvement. Journal of Clinical Virology. Elsevier B.V; 2013 May

149

1;57(1):91–4.

46. Gould EA, Solomon T. Pathogenic flaviviruses. The Lancet. 2008 Feb;371(9611):500–9.

47. Miner JJ, Diamond MS. Zika Virus Pathogenesis and Tissue Tropism. Cell Host and Microbe. Elsevier Inc; 2017 Feb 8;21(2):134–42.

48. Fernandez-Garcia M-D, Mazzon M, Jacobs M, Amara A. Pathogenesis of Flavivirus Infections: Using and Abusing the Host Cell. Cell Host and Microbe. Elsevier Inc; 2009 Apr 23;5(4):318–28.

49. Zanotto PM, Gao GF, Gritsun T, Marin MS, Jiang WR, Venugopal K, et al. An arbovirus cline across the northern hemisphere. Virology. 1995 Jun;210(1):152–9.

50. Robertson SJ, Mitzel DN, Taylor RT, Best SM, Bloom ME. Tick-borne flaviviruses: dissecting host immune responses and virus countermeasures. Immunol Res. 2008 Oct 8;43(1-3):172–86.

51. Chambers TJ, Hahn CS, Galler R, Rice CM. Flavivirus genome organization, expression, and replication. Annu Rev Microbiol. 1990;44:649–88.

52. Lasala PR, Holbrook M. Tick-borne flaviviruses. Clinics in Laboratory Medicine. 2010 Mar;30(1):221–35.

53. Pradier S, Lecollinet S, Leblond A. West Nile virus epidemiology and factors triggering change in its distribution in Europe. Rev Sci Tech. 2012 Dec;31(3):829–44.

54. Anderson JR, Rico-Hesse R. Aedes aegypti vectorial capacity is determined by the infecting genotype of dengue virus. American Journal of Tropical Medicine and Hygiene. 2006 Nov;75(5):886–92.

55. Benjelloun A, Harrak El M, Belkadi B. West Nile Disease Epidemiology in North-West Africa: Bibliographical Review. Transbound Emerg Dis. 4 ed. 2015 Mar 6;63(6):e153–9.

56. Kaiser R. Tick-Borne Encephalitis. Infectious Disease Clinics of North America. 2008 Sep;22(3):561–75.

57. Karbowiak G, Biernat B. The role of particular tick developmental stages in the circulation of tick-borne pathogens affecting humans in Central Europe. 2. Tick- borne encephalitis virus. Ann Parasitol. 2016;62(1):3–9.

58. Review articles The role of particular tick developmental stages in the circulation

150

of tick-borne pathogens affecting humans in Central Europe. 2. Tick-borne encephalitis virus. 2016 Apr 2;:1–7.

59. McNally KL, Mitzel DN, Anderson JM, Ribeiro JMC, Valenzuela JG, Myers TG, et al. Differential salivary gland transcript expression profile in Ixodes scapularis nymphs upon feeding or flavivirus infection. Ticks and Tick-borne Diseases. Elsevier GmbH; 2012 Feb 1;3(1):18–26.

60. Labuda MEA. Importance of Localized Skin Infection in Tick-Borne Encephalitis Virus Transmission. 1996 May 1;:1–10.

61. Fialová A, Cimburek Z, Iezzi G, Kopecký J. Ixodes ricinus tick saliva modulates tick-borne encephalitis virus infection of dendritic cells. Microbes and Infection. Elsevier Masson SAS; 2010 Jul 1;12(7):580–5.

62. Hermance ME, Thangamani S. Proinflammatory Cytokines and Chemokines at the Skin Interface during Powassan Virus Transmission. Journal of Investigative Dermatology. Nature Publishing Group; 2014 Apr 17;134(8):2280–3.

63. Labuda M, Austyn JM, Zuffova E, Kozuch O, Fuchsberger N, Lysy J, et al. Importance of localized skin infection in tick-borne encephalitis virus transmission. Virology. 1996 May;219(2):357–66.

64. Labuda M, Nuttall PA, Kozuch O, Eleckova E, Williams T, Zuffova E, et al. Non- viraemic transmission of tick-borne encephalitis virus: a mechanism for arbovirus survival in nature. Experientia. 1993 Sep;49(9):802–5.

65. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW, Moyes CL, et al. The global distribution and burden of dengue. Nature. 2013 Apr 7;496(7446):504–7.

66. Kramer LD, Li J, Shi P-Y. West Nile virus. The Lancet Neurology. 2007 Feb;6(2):171–81.

67. Roth D, Henry B, Mak S, Fraser M, Taylor M, Li M, et al. West Nile Virus Range Expansion into British Columbia. Emerg Infect Dis. 2010 Aug;16(8):1251–8.

68. Bröker M, Kollaritsch H. After a tick bite in a tick-borne encephalitis virus endemic area: Current positions about post-exposure treatment. Vaccine. 2008 Feb;26(7):863–8.

69. Haglund M, Günther G. Tick-borne encephalitis—pathogenesis, clinical course and long-term follow-up. Vaccine. 2003 Apr;21:S11–8.

70. Piantadosi A, Rubin DB, McQuillen DP, Hsu L, Lederer PA, Ashbaugh CD, et al. Emerging Cases of Powassan Virus Encephalitis in New England: Clinical

151

Presentation, Imaging, and Review of the Literature. Clin Infect Dis. 2016 Mar 1;62(6):707–13.

71. Best SM. The Many Faces of the Flavivirus NS5 Protein in Antagonism of Type I Interferon Signaling. Pierson TC, editor. Journal of Virology. 2017 Jan 18;91(3):e01970–16–14.

72. Mandl CW. Steps of the tick-borne encephalitis virus replication cycle that affect neuropathogenesis. Virus Research. 2005 Aug;111(2):161–74.

73. Kawai T, Akira S. Toll-like Receptors and Their Crosstalk with Other Innate Receptors in Infection and Immunity. Immunity. Elsevier Inc; 2011 May 27;34(5):637–50.

74. Nazmi A, Dutta K, Basu A. RIG-I Mediates Innate Immune Response in Mouse Neurons Following Japanese Encephalitis Virus Infection. Schneider BS, editor. PLoS ONE. 2011 Jun 30;6(6):e21761–13.

75. Muñoz-Jordán JL, Fredericksen BL. How Flaviviruses Activate and Suppress the Interferon Response. Viruses. 2010 Feb;2(2):676–91.

76. Takaoka A, Yanai H. Interferon signalling network in innate defence. Cell Microbiol. 2006 Jun;8(6):907–22.

77. Haller O, Kochs G, Weber F. The interferon response circuit: Induction and suppression by pathogenic viruses. Virology. 2006 Jan;344(1):119–30.

78. Loo Y-M, Gale M Jr. Immune Signaling by RIG-I-like Receptors. Immunity. Elsevier Inc; 2011 May 27;34(5):680–92.

79. Kato H, Takeuchi O, Sato S, Yoneyama M, Yamamoto M, Matsui K, et al. Differential roles of MDA5 and RIG-I helicases in the recognition of RNA viruses. Nature. 2006 Apr 9;441(7089):101–5.

80. Fredericksen BL, Keller BC, Fornek J, Katze MG, Gale M. Establishment and Maintenance of the Innate Antiviral Response to West Nile Virus Involves both RIG-I and MDA5 Signaling through IPS-1. Journal of Virology. 2008 Jan 2;82(2):609–16.

81. Loo YM, Fornek J, Crochet N, Bajwa G, Perwitasari O, Martinez-Sobrido L, et al. Distinct RIG-I and MDA5 Signaling by RNA Viruses in Innate Immunity. Journal of Virology. 2007 Dec 10;82(1):335–45.

82. Nakhaei P, Genin P, Civas A, Hiscott J. RIG-I-like receptors: Sensing and responding to RNA virus infection. Seminars in Immunology. 2009

152

Aug;21(4):215–22.

83. Schmidt A, Rothenfusser S, Hopfner K-P. Sensing of viral nucleic acids by RIG-I: From translocation to translation. European Journal of Cell Biology. Elsevier GmbH; 2012 Jan 1;91(1):78–85.

84. Gack MU, Shin YC, Joo C-H, Urano T, Liang C, Sun L, et al. TRIM25 RING- finger E3 ubiquitin ligase is essential for RIG-I-mediated antiviral activity. Nature. 2007 Mar 28;446(7138):916–20.

85. Versteeg GA, Rajsbaum R, Sánchez-Aparicio MT, Maestre AM, Valdiviezo J, Shi M, et al. The E3-Ligase TRIM Family of Proteins Regulates Signaling Pathways Triggered by Innate Immune Pattern-Recognition Receptors. Immunity. Elsevier Inc; 2013 Feb 21;38(2):384–98.

86. Seth RB, Sun L, Ea C-K, Chen ZJ. Identification and Characterization of MAVS, a Mitochondrial Antiviral Signaling Protein that Activates NF-κB and IRF3. Cell. 2005 Sep;122(5):669–82.

87. Kawai T, Takahashi K, Sato S, Coban C, Kumar H, Kato H, et al. IPS-1, an adaptor triggering RIG-I- and Mda5-mediated type I interferon induction. Nat Immunol. 2005 Aug 28;6(10):981–8.

88. Weber E, Finsterbusch K, Lindquist R, Nair S, Lienenklaus S, Gekara NO, et al. Type I Interferon Protects Mice from Fatal Neurotropic Infection with Langat Virus by Systemic and Local Antiviral Responses. Journal of Virology. 2014 Oct 6;88(21):12202–12.

89. Kurhade C, Zegenhagen L, Weber E, Nair S, Michaelsen-Preusse K, Spanier J, et al. Type I Interferon response in olfactory bulb, the site of tick-borne flavivirus accumulation, is primarily regulated by IPS-1. Journal of Neuroinflammation. Journal of Neuroinflammation; 2016 Jan 26;:1–14.

90. Kawai T, Akira S. Pathogen recognition with Toll-like receptors. Current Opinion in Immunology. 2005 Aug;17(4):338–44.

91. Blasius AL, Beutler B. Intracellular Toll-like Receptors. Immunity. Elsevier Inc; 2010 Mar 26;32(3):305–15.

92. Nasirudeen AMA, Wong HH, Thien P, Xu S, Lam K-P, Liu DX. RIG-I, MDA5 and TLR3 Synergistically Play an Important Role in Restriction of Dengue Virus Infection. Harris E, editor. PLoS Negl Trop Dis. 2011 Jan 4;5(1):e926–12.

93. Schindler C, Plumlee C. Inteferons pen the JAK–STAT pathway. Seminars in

153

Cell and Developmental Biology. 2008 Aug;19(4):311–8.

94. Levy D. The virus battles: IFN induction of the antiviral state and mechanisms of viral evasion. Cytokine and Growth Factor Reviews. 2001 Jun;12(2-3):143–56.

95. Sadler AJ, Williams BRG. Interferon-inducible antiviral effectors. Nat Rev Immunol. 2008 Jul;8(7):559–68.

96. Duggal NK, Emerman M. Evolutionary conflicts between viruses and restriction factors shape immunity. Nat Rev Immunol. 2012 Sep 14;12(10):687–95.

97. Sauter D. Counteraction of the multifunctional restriction factor tetherin. 2014 Apr 8;:1–14.

98. Bruni D, Chazal M, Sinigaglia L, Chauveau L, Schwartz O, Despres P, et al. Viral entry route determines how human plasmacytoid dendritic cells produce type I interferons. Science signaling. American Association for the Advancement of Science; 2015 Mar 3;8(366):ra25–5.

99. Shu Q, Lennemann NJ, Sarkar SN, Sadovsky Y, Coyne CB. ADAP2 Is an Interferon Stimulated Gene That Restricts RNA Virus Entry. Gale M, editor. PLoS Pathog. 2015 Sep 15;11(9):e1005150–24.

100. Lazear HM, Diamond MS. ScienceDirect New insights into innate immune restriction of West Nile virus infection. Current Opinion in Virology. Elsevier B.V; 2015 Apr 1;11:1–6.

101. Fink J, Gu F, Ling L, Tolfvenstam T, Olfat F, Chin KC, et al. Host Gene Expression Profiling of Dengue Virus Infection in Cell Lines and Patients. Farrar J, editor. PLoS Negl Trop Dis. 2007 Nov 7;1(2):e86–11.

102. Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, Bieniasz P, et al. A diverse range of gene products are effectors of the type I interferon antiviral response. Nature. 2011 Apr 10;472(7344):481–5.

103. Yasunaga A, Hanna SL, Li J, Cho H, Rose PP, Spiridigliozzi A, et al. Genome- Wide RNAi Screen Identifies Broadly-Acting Host Factors That Inhibit Arbovirus Infection. Brass AL, editor. PLoS Pathog. 2014 Feb 13;10(2):e1003914–7.

104. Gerold G. Protein interactions during the flavivirus and hepacivirus life cycle. 2017 Jan 11;:1–55.

105. Le Breton M, Meyniel-Schicklin L, Deloire A, Coutard B, Canard B, de Lamballerie X, et al. Flavivirus NS3 and NS5 proteins interaction network: a high-throughput yeast two-hybrid screen. BMC Microbiology. BioMed Central

154

Ltd; 2011 Oct 20;11(1):234.

106. Khadka S, Vangeloff AD, Zhang C, Siddavatam P, Heaton NS, Wang L, et al. A Physical Interaction Network of Dengue Virus and Human Proteins. Molecular & Cellular Proteomics. 2011 Dec 6;10(12):M111.012187–7.

107. Taylor RT, Lubick KJ, Robertson SJ, Broughton JP, Bloom ME, Bresnahan WA, et al. TRIM79α, an Interferon-Stimulated Gene Product, Restricts Tick-Borne Encephalitis Virus Replication by Degrading the Viral RNA Polymerase. Cell Host and Microbe. 2011 Sep;10(3):185–96.

108. Lazear HM, Diamond MS. ScienceDirect New insights into innate immune restriction of West Nile virus infection. Current Opinion in Virology. Elsevier B.V; 2015 Apr 1;11:1–6.

109. Manocha G, Mishra R, Sharma N, Kumawat K, Basu A, Singh SK. Regulatory role of TRIM21 in the type-I interferon pathway in Japanese encephalitis virus- infected human microglial cells. Journal of Neuroinflammation. 2014;11(1):24–5.

110. Lazear HM, Daniels BP, Pinto AK, Huang AC, Vick SC, Doyle SE, et al. Interferon- restricts West Nile virus neuroinvasion by tightening the blood-brain barrier. Science Translational Medicine. 2015 Apr 22;7(284):284ra59–9.

111. Towers GJ. Control of Viral Infectivity by Tripartite Motif Proteins. Human Gene Therapy. 2005 Oct;16(10):1125–32.

112. Kawai T, Akira S. Regulation of innate immune signalling pathways by the tripartite motif (TRIM) family proteins. EMBO Mol Med. 2011 Aug 9;3(9):513– 27.

113. Turrini F, Di Pietro A, Vicenzi E. Lentiviral Effector Pathways of TRIM Proteins. DNA and Cell Biology. 2014 Apr;33(4):191–7.

114. Sawyer SL, Emerman M, Malik HS. Discordant evolution of the adjacent antiretroviral genes TRIM22 and TRIM5 in mammals. PLoS Pathog. 2007 Dec;3(12):e197.

115. Sawyer SL, Wu LI, Emerman M, Malik HS. Positive selection of primate TRIM5alpha identifies a critical species-specific retroviral restriction domain. Proceedings of the National Academy of Sciences. 2005 Feb;102(8):2832–7.

116. Yap MW, Stoye JP. TRIM proteins and the innate immune response to viruses. Adv Exp Med Biol. 2012;770:93–104.

117. De Silva S, Wu L. TRIM5 Acts as More Than a Retroviral Restriction Factor.

155

Viruses. 2011 Dec;3(12):1204–9.

118. Rajsbaum R, García-Sastre A, Versteeg GA. TRIMmunity: The Roles of the TRIM E3-Ubiquitin Ligase Family in Innate Antiviral Immunity. Elsevier Ltd; 2014 Feb 20;:1–20.

119. Caroline. Anti-viral TRIMs – Friend or Foe in autoimmune disease? 2011 Apr 5;:1–30.

120. Ozato K, Shin D-M, Chang T-H, Morse HC. TRIM family proteins and their emerging roles in innate immunity. Nat Rev Immunol. 2008 Nov;8(11):849–60.

121. Liu S-Y, Sanchez DJ, Cheng G. New developments in the induction and antiviral effectors of type I interferon. Current Opinion in Immunology. Elsevier Ltd; 2011 Feb 1;23(1):57–64.

122. Chebath J, Philipe B, Ara H, Julien G, Revel M. Four Different Forms of Interferon-induced 2’,5’-01igo(A) Synthetase Identified by Immunoblotting in Human Cells*. biological chemistry. 262nd ed. 2001 Jul 12;:1–6.

123. Lucas M, Tomoji M, Marie-Pascale F, Dominique S-C, Xavier M, Pierre- Emmanuel C, et al. Infection of mouse neurones by West Nile virus is modulated by the interferon-inducible 2. Immunology and Cell Biology. 2011 May 17;:230– 6.

124. Yoshii K, Moritoh K, Nagata N, Yokozawa K, Sakai M, Sasaki N, et al. Susceptibility to flavivirus-specific antiviral response of Oas1b affects the neurovirulence of the Far-Eastern subtype of tick-borne encephalitis virus. Archives of Virology. 2012 Dec 25;158(5):1039–46.

125. Samuel MA, Diamond MS. Pathogenesis of West Nile Virus Infection: a Balance between Virulence, Innate and Adaptive Immunity, and Viral Evasion. Journal of Virology. 2006 Sep 13;80(19):9349–60.

126. Kajaste-Rudnitski A, Mashimo T, Frenkiel M-P, Guenet J-L, Lucas M, Despres P. The 2“,5-”oligoadenylate synthetase 1b is a potent inhibitor of West Nile virus replication inside infected cells. Journal of Biological Chemistry. 2006 Feb;281(8):4624–37.

127. Lim JK, Lisco A, McDermott DH, Huynh L, Ward JM, Johnson B, et al. Genetic Variation in OAS1 Is a Risk Factor for Initial Infection with West Nile Virus in Man. Rice CM, editor. PLoS Pathog. 2009 Feb 27;5(2):e1000321–12.

128. Courtney SC, Di H, Stockman BM, Liu H, Scherbik SV, Brinton MA.

156

Identification of Novel Host Cell Binding Partners of Oas1b, the Protein Conferring Resistance to Flavivirus-Induced Disease in Mice. Journal of Virology. 2012 Jul 11;86(15):7953–63.

129. Diamond MS, Farzan M. The broad-spectrum antiviral functions of IFIT and IFITM proteins. Nat Rev Immunol. Nature Publishing Group; 2012 Dec 14;13(1):46–57.

130. Schoggins JW, Wilson SJ, Panis M, Murphy MY, Jones CT, Bieniasz P, et al. A diverse range of gene products are effectors of the type I interferon antiviral response. Nature. 2011 Apr 10;472(7344):481–5.

131. Wacher C, Muller M, Hofer MJ, Getts DR, Zabaras R, Ousman SS, et al. Coordinated Regulation and Widespread Cellular Expression of Interferon- Stimulated Genes (ISG) ISG-49, ISG-54, and ISG-56 in the Central Nervous System after Infection with Distinct Viruses. Journal of Virology. 2006 Dec 27;81(2):860–71.

132. Schoggins JW, Rice CM. Interferon-stimulated genes and their antiviral effector functions. Current Opinion in Virology. 2011 Dec;1(6):519–25.

133. Schoggins JW. ScienceDirect Interferon-stimulated genes: roles in viral pathogenesis. Current Opinion in Virology. Elsevier B.V; 2014 Jun 1;6:40–6.

134. Smith SE, Weston S, Kellam P, Marsh M. ScienceDirect IFITM proteins — cellular inhibitors of viral entry. Current Opinion in Virology. Elsevier B.V; 2014 Feb 1;4:71–7.

135. Diamond MS, Farzan M. The broad-spectrum antiviral functions of IFIT and IFITM proteins. Nat Rev Immunol. Nature Publishing Group; 2012 Dec 14;13(1):46–57.

136. Jiang D, Weidner JM, Qing M, Pan XB, Guo H, Xu C, et al. Identification of Five Interferon-Induced Cellular Proteins That Inhibit West Nile Virus and Dengue Virus Infections. Journal of Virology. 2010 Jul 16;84(16):8332–41.

137. Chutiwitoonchai N, Hiyoshi M, Hiyoshi-Yoshidomi Y, Hashimoto M, Tokunaga K, Suzu S. Characteristics of IFITM, the newly identified IFN-inducible anti- HIV-1 family proteins. Microbes and Infection. Elsevier Masson SAS; 2013 Apr 1;15(4):280–90.

138. Brass AL, Huang I-C, Benita Y, John SP, Krishnan MN, Feeley EM, et al. The IFITM Proteins Mediate Cellular Resistance to Influenza A H1N1 Virus, West Nile Virus, and Dengue Virus. Cell. Elsevier Ltd; 2009 Dec 24;139(7):1243–54.

157

139. Ozato K, Tailor P, Kubota T. The Interferon Regulatory Factor Family in Host Defense: Mechanism of Action. Journal of Biological Chemistry. 2007 Jul 6;282(28):20065–9.

140. Daffis S, Samuel MA, Keller BC, Gale M, Diamond MS. Cell-Specific IRF-3 Responses Protect against West Nile Virus Infection by Interferon-Dependent and -Independent Mechanisms. PLoS Pathog. 2007;3(7):e106–11.

141. Fredericksen BL. The neuroimmune response to West Nile virus. J Neurovirol. 2013 Jul 11;20(2):113–21.

142. Robertson SJ, Lubick KJ, Freedman BA, Carmody AB, Best SM. Tick-Borne Flaviviruses Antagonize Both IRF-1 and Type I IFN Signaling To Inhibit Dendritic Cell Function. The Journal of Immunology. 2014 Mar 7;192(6):2744– 55.

143. Perelygin AA, Scherbik SV, Zhulin IB, Stockman BM, Li Y, Brinton MA. Positional cloning of the murine flavivirus resistance gene. Proceedings of the National Academy of Sciences. 2002 Jul;99(14):9322–7.

144. Perelygin AA, Zharkikh AA, Scherbik SV, Brinton MA. The mammalian 2“-5” oligoadenylate synthetase gene family: evidence for concerted evolution of paralogous Oas1 genes in Rodentia and Artiodactyla. J Mol Evol. 2006 Oct;63(4):562–76.

145. Mandl CW. Steps of the tick-borne encephalitis virus replication cycle that affect neuropathogenesis. Virus Research. 2005 Aug;111(2):161–74.

146. Overby AK, Popov VL, Niedrig M, Weber F. Tick-Borne Encephalitis Virus Delays Interferon Induction and Hides Its Double-Stranded RNA in Intracellular Membrane Vesicles. Journal of Virology. 2010 Aug 5;84(17):8470–83.

147. Gack MU, Diamond MS. ScienceDirect Innate immune escape by Dengue and West Nile viruses. Current Opinion in Virology. Elsevier B.V; 2016 Oct 1;20:119–28.

148. Ye J, Zhu B, Fu ZF, Chen H, Cao S. Immune evasion strategies of flaviviruses. Vaccine. Elsevier Ltd; 2013 Jan 7;31(3):461–71.

149. Manokaran G, Finol E, Wang C, Gunaratne J, Bahl J, Ong EZ, et al. Dengue subgenomic RNA binds TRIM25 to inhibit interferon expression for epidemiological fitness. Science. 2015 Oct 8;350(6257):217–21.

150. Chan YK, Gack MU. A phosphomimetic-based mechanism of dengue virus to

158

antagonize innate immunity. Nat Immunol. 2016 Mar 21;17(5):523–30.

151. Anglero-Rodriguez YI, Pantoja P, Sariol CA, Waters WR. Dengue Virus Subverts the Interferon Induction Pathway via NS2B/3 Protease-I B Kinase Interaction. Clinical and Vaccine Immunology. 2013 Dec 31;21(1):29–38.

152. Best SM, Morris KL, Shannon JG, Robertson SJ, Mitzel DN, Park GS, et al. Inhibition of Interferon-Stimulated JAK-STAT Signaling by a Tick-Borne Flavivirus and Identification of NS5 as an Interferon Antagonist. Journal of Virology. 2005 Sep 27;79(20):12828–39.

153. Lubick KJ, Robertson SJ, McNally KL, Freedman BA, Rasmussen AL, Taylor RT, et al. Flavivirus Antagonism of Type I Interferon Signaling Reveals Prolidase as a Regulator of IFNAR1 Surface Expression. Cell Host and Microbe. 2015 Jul;18(1):61–74.

154. Laurent-Rolle M, Boer EF, Lubick KJ, Wolfinbarger JB, Carmody AB, Rockx B, et al. The NS5 Protein of the Virulent West Nile Virus NY99 Strain Is a Potent Antagonist of Type I Interferon-Mediated JAK-STAT Signaling. Journal of Virology. 2010 Mar 8;84(7):3503–15.

155. Ashour J, Morrison J, Laurent-Rolle M, Belicha-Villanueva A, Plumlee CR, Bernal-Rubio D, et al. Mouse STAT2 Restricts Early Dengue Virus Replication. Cell Host and Microbe. Elsevier Inc; 2010 Nov 18;8(5):410–21.

156. Jones M, Davidson A, Hibbert L, Gruenwald P, Schlaak J, Ball S, et al. Dengue Virus Inhibits Alpha Interferon Signaling by Reducing STAT2 Expression. Journal of Virology. 2005 Apr 12;79(9):5414–20.

157. Morrison J, Laurent-Rolle M, Maestre AM, Rajsbaum R, Pisanelli G, Simon V, et al. Dengue Virus Co-opts UBR4 to Degrade STAT2 and Antagonize Type I Interferon Signaling. Diamond MS, editor. PLoS Pathog. 2013 Mar 28;9(3):e1003265–14.

158. Aguirre S, Maestre AM, Pagni S, Patel JR, Savage T, Gutman D, et al. DENV Inhibits Type I IFN Production in Infected Cells by Cleaving Human STING. Diamond MS, editor. PLoS Pathog. 2012 Oct 4;8(10):e1002934–14.

159. Best SM. The Many Faces of the Flavivirus NS5 Protein in Antagonism of Type I Interferon Signaling. Pierson TC, editor. Journal of Virology. 2017 Jan 18;91(3):e01970–16–14.

160. Camini FC, Caetano CCS, Almeida LT, Magalhães CLB. Implications of oxidative stress on viral pathogenesis. Archives of Virology. Springer Vienna;

159

2016 Dec 27;162(4):907–17.

161. Paiva CN, Bozza MT. Are Reactive Oxygen Species Always Detrimental to Pathogens? Antioxidants & Redox Signaling. 2014 Feb 20;20(6):1000–37.

162. Peterhans E. Sendai Virus Stimulates Chemiluminescence in Mouse Spleen cells. Biochemical and Biophysical Research Communications. 2003 Oct 29;91:1–10.

163. Castro SM, Guerrero-Plata A, Suarez-Real G, Adegboyega PA, Colasurdo GN, Khan AM, et al. Antioxidant Treatment Ameliorates Respiratory Syncytial Virus– induced Disease and Lung Inflammation. Am J Respir Crit Care Med. 2006 Dec 15;174(12):1361–9.

164. Ezzikouri S, Nishimura T, Kohara M, Benjelloun S, Kino Y, Inoue K, et al. Inhibitory effects of PycnogenolÂ® on hepatitis C virus replication. ANTIVIRAL RESEARCH. Elsevier B.V; 2015 Jan 1;113(C):93–102.

165. Valero N, Mosquera J, Añez G, Levy A, Marcucci R, de Mon MA. Differential Oxidative Stress Induced by Dengue Virus in Monocytes from Human Neonates, Adult and Elderly Individuals. Liu G, editor. PLoS ONE. 2013 Sep 17;8(9):e73221–10.

166. Zhang Y, Wang Z, Chen H, Chen Z, Tian Y. Antioxidants: potential antiviral agents for Japanese encephalitis virus infection. International Journal of Infectious Diseases. International Society for Infectious Diseases; 2014 Jul 1;24:30–6.

167. Kuzmenko YV, Smirnova OA, Ivanov AV, Starodubova ES, Karpov VL. Nonstructural Protein 1 of Tick-Borne Encephalitis Virus Induces Oxidative Stress and Activates Antioxidant Defense by the Nrf2/ARE Pathway. Intervirology. 2016 Dec 20;59(2):111–7.

168. Gullberg RC, Jordan Steel J, Moon SL, Soltani E, Geiss BJ. Oxidative stress influences positive strand RNA virus genome synthesis and capping. Virology. 2015 Jan;475:219–29.

169. D'Autréaux B, Toledano MB. ROS as signalling molecules: mechanisms that generate specificity in ROS homeostasis. Nat Rev Mol Cell Biol. 2007 Oct;8(10):813–24.

170. Morgan MJ, Liu Z-G. Morgan MJ, Liu Z-GCrosstalk of reactive oxygen species and NF-??B signaling. Cell Res 21:103- 115. Nature Publishing Group. Nature Publishing Group; 2010 Dec 28;21(1):103–15.

171. Kim SJ, Wong PKY. ROS upregulation during the early phase of retroviral

160

infection plays an important role in viral establishment in the host cell. Journal of General Virology. 2013 Sep 13;94(Pt_10):2309–17.

172. Olagnier D, Peri S, Steel C, van Montfoort N, Chiang C, Beljanski V, et al. Cellular Oxidative Stress Response Controls the Antiviral and Apoptotic Programs in Dengue Virus-Infected Dendritic Cells. Randall G, editor. PLoS Pathog. 2014 Dec 18;10(12):e1004566–18.

173. Ishikawa T, Yamanaka A, Konishi E. A review of successful flavivirus vaccines and the problems with those flaviviruses for which vaccines are not yet available. Vaccine. Elsevier Ltd; 2014 Mar 10;32(12):1326–37.

174. Diamond MS. Evasion of innate and adaptive immunity by flaviviruses. Immunology and Cell Biology. 2011 May 17;:196–206.

175. Gritsun TS, Lashkevich VA, Gould EA. Tick-borne encephalitis. ANTIVIRAL RESEARCH. 2003 Jan;57(1-2):129–46.

176. Pealer LN, Marfin AA, Petersen LR, Lanciotti RS, Page PL, Stramer SL, et al. Transmission of West Nile virus through blood transfusion in the United States in 2002. N Engl J Med. 2003 Sep;349(13):1236–45.

177. Montgomery SP, Brown JA, Kuehnert M, Smith TL, Crall N, Lanciotti RS, et al. Transfusion-associated transmission of West Nile virus, United States 2003 through 2005. Transfusion. 2006 Dec;46(12):2038–46.

178. Ustinova OI, Volechova GM, Deviatkov MI, Gusmanova AI. [The clinico- epidemiological characteristics of tick-borne encephalitis in Perm Province]. Zh Mikrobiol Epidemiol Immunobiol. 1997 May;(3):33–6.

179. Nuttall PA, Labuda M. Dynamics of infection in tick vectors and at the tick-host interface. Adv Virus Res. 2003;60:233–72.

180. Sanchez-Vargas I, Scott JC, Poole-Smith BK, Franz AWE, Barbosa-Solomieu V, Wilusz J, et al. Dengue virus type 2 infections of Aedes aegypti are modulated by the mosquito's RNA interference pathway. PLoS Pathog. 2009 Feb;5(2):e1000299.

181. Kramer LD, Li J, Shi P-Y. West Nile virus. The Lancet Neurology. 2007 Feb;6(2):171–81.

182. Mlera L, Melik W, Bloom ME. The role of viral persistence in flavivirus biology. Pathogens Disease. 2014 May 12;71(2):137–63.

183. Mackenzie JS, Barrett ADT, Deubel V. The Japanese encephalitis serological

161

group of flaviviruses: a brief introduction to the group. Curr Top Microbiol Immunol. 2002;267:1–10.

184. Jacobson ER, Ginn PE, Troutman JM, Farina L, Stark L, Klenk K, et al. West Nile virus infection in farmed American alligators (Alligator mississippiensis) in Florida. J Wildl Dis. 2005 Jan;41(1):96–106.

185. Mackenzie JS, Gubler DJ, Petersen LR. Emerging flaviviruses: the spread and resurgence of Japanese encephalitis, West Nile and dengue viruses. Nat Med. 2004 Nov 30;10(12s):S98–S109.

186. Offerdahl DK, Clancy NG, Bloom ME. Stability of a Tick-Borne Flavivirus in Milk. Front Bioeng Biotechnol. 2016 May 11;4:262–6.

187. Vereta LA, Skorobrekha VZ, Nikolaeva SP, Aleksandrov VI, Tolstonogova VI, Zakharycheva TA, et al. [The transmission of the tick-borne encephalitis virus via cow's milk]. Med Parazitol (Mosk). 1991 May;(3):54–6.

188. Hudopisk N, Korva M, Janet E, Simetinger M, Grgic-Vitek M, Gubensek J, et al. Tick-borne encephalitis associated with consumption of raw goat milk, Slovenia, 2012. Emerg Infect Dis. 2013 May;19(5):806–8.

189. Labuda M, Randolph SE. Survival strategy of tick-borne encephalitis virus: cellular basis and environmental determinants. Zentralbl Bakteriol. 1999 Dec;289(5-7):513–24.

190. Mlera L, Melik W, Bloom ME. The role of viral persistence in flavivirus biology. Pathogens Disease. 2014 May 12;71(2):137–63.

191. Rosà R, Pugliese A, Norman R, Hudson PJ. Thresholds for disease persistence in models for tick-borne infections including non-viraemic transmission, extended feeding and tick aggregation. Journal of Theoretical Biology. 2003 Oct;224(3):359–76.

192. Haydon DT, Cleaveland S, Taylor LH, Laurenson MK. Identifying reservoirs of infection: a conceptual and practical challenge. Emerg Infect Dis. 2002 Dec;8(12):1468–73.

193. Raberg L, Sim D, Read AF. Disentangling genetic variation for resistance and tolerance to infectious diseases in animals. Science. 2007 Nov;318(5851):812–4.

194. Mandl JN, Ahmed R, Barreiro LB, Daszak P, Epstein JH, Virgin HW, et al. Reservoir host immune responses to emerging zoonotic viruses. Cell. 2015 Jan;160(1-2):20–35.

162

195. Hrnjakovic Cvjetkovic I, Cvjetkovic D, Patic A, Radovanov J, Kovacevic G, Milosevic V. TICK-BORNE ENCEPHALITIS VIRUS INFECTION IN HUMANS. Med Pregl. 2016 Mar;69(3-4):93–8.

196. Kim S-Y, Yun S-M, Han MG, Lee IY, Lee NY, Jeong YE, et al. Isolation of Tick- Borne Encephalitis Viruses from Wild Rodents, South Korea. Vector-Borne and Zoonotic Diseases. 2008 Feb;8(1):7–14.

197. Takeda T, Ito T, Osada M, Takahashi K, Takashima I. Isolation of tick-borne encephalitis virus from wild rodents and a seroepizootiologic survey in Hokkaido, Japan. American Journal of Tropical Medicine and Hygiene. 1999 Feb;60(2):287– 91.

198. Frey S, Essbauer S, Zoller G, Klempa B, Weidmann M, Dobler G, et al. Complete Genome Sequence of Tick-Borne Encephalitis Virus Strain A104 Isolated from a Yellow-Necked Mouse (Apodemus flavicollis) in Austria. Genome Announcements. 2013 Jun 27;1(4):e00564–13–e00564–13.

199. Dupuis AP II, Peters RJ, Prusinski MA, Falco RC, Ostfeld RS, Kramer LD. Isolation of deer tick virus (Powassan virus, lineage II) from Ixodes scapularis and detection of antibody in vertebrate hosts sampled in the Hudson Valley, New York State. Parasites & Vectors. Parasites & Vectors; 2013 Jul 15;6(1):1–1.

200. Egyed L, Zöldi V, Szeredi L. Subclinical Tick-Borne Encephalitis Virus in Experimentally Infected Apodemus agrarius. Intervirology. 2016 May 28;58(6):369–72.

201. Telford SR3, Armstrong PM, Katavolos P, Foppa I, Garcia AS, Wilson ML, et al. A new tick-borne encephalitis-like virus infecting New England deer ticks, Ixodes dammini. Emerg Infect Dis. 1997 Apr;3(2):165–70.

202. Mlera L, Meade-White K, Saturday G, Scott D, Bloom ME. Modeling Powassan virus infection in Peromyscus leucopus, a natural host. Ebel GD, editor. PLoS Negl Trop Dis. 2017 Jan 31;11(1):e0005346–18.

203. Munshi-South J, Richardson JL. Peromyscus transcriptomics: Understanding adaptation and gene expression plasticity within and between species of deer mice. Elsevier Ltd; 2017 Jan 7;:1–9.

204. Kiaris H. Editorial. Seminars in Cell and Developmental Biology. 2017 Jan;61:80–1.

205. Vandegrift KJ, Critchlow JT, Kapoor A, Friedman DA, Hudson PJ. Peromyscus as a model system for human hepatitis C: An opportunity to advance our

163

understanding of a complex host parasite system. Elsevier Ltd; 2017 Jan 7;:1–8.

206. Labinskyy N, Mukhopadhyay P, Toth J, Szalai G, Veres M, Losonczy G, et al. Longevity is associated with increased vascular resistance to high glucose- induced oxidative stress and inflammatory gene expression in Peromyscus leucopus. AJP: Heart and Circulatory Physiology. 2009 Feb 6;296(4):H946–56.

207. Havighorst A, Crossland J, Kiaris H. Peromyscus as a model of human disease. Seminars in Cell and Developmental Biology. 2017 Jan;61:150–5.

208. Vandegrift KJ, Critchlow JT, Kapoor A, Friedman DA, Hudson PJ. Peromyscus as a model system for human hepatitis C: An opportunity to advance our understanding of a complex host parasite system. Seminars in Cell and Developmental Biology. 2017 Jan;61:123–30.

209. Barbour AG. Infection resistance and tolerance in Peromyscus spp., natural reservoirs of microbes that are virulent for humans. Seminars in Cell and Developmental Biology. Elsevier Ltd; 2016 Jul 2;:1–22.

210. Ebel GD. Update on Powassan Virus: Emergence of a North American Tick- Borne Flavivirus. Annu Rev Entomol. 2010 Jan;55(1):95–110.

211. Unknown. A New Tick-borne Encephalitis-like Virus Infecting New England Deer Ticks. 2009 Jan 7;:1–6.

212. Khoury El MY, Hull RC, Bryant PW, Escuyer KL, St George K, Wong SJ, et al. Diagnosis of Acute Deer Tick Virus Encephalitis. Clin Infect Dis. 2013 Jan 23;56(4):e40–7.

213. Steppan S, Adkins R, Anderson J. Phylogeny and divergence-date estimates of rapid radiations in muroid rodents based on multiple nuclear genes. Syst Biol. 2004 Aug;53(4):533–53.

214. Ramsdell CM, Lewandowski AA, Glenn J, Vrana PB, O'Neill RJ, Dewey MJ. Comparative genome mapping of the deer mouse (Peromyscus maniculatus) reveals greater similarity to rat (Rattus norvegicus) than to the lab mouse (Mus musculus). BMC Evol Biol. 2008;8(1):65–14.

215. Santos R, Hermance M, Gelman B, Thangamani S. Spinal Cord Ventral Horns and Lymphoid Organ Involvement in Powassan Virus Infection in a Mouse Model. Viruses. 2016 Aug;8(8):220–17.

216. Havighorst A, Crossland J, Kiaris H. Seminars in Cell & Developmental Biology. Seminars in Cell and Developmental Biology. Elsevier Ltd; 2017 Jan 1;61:150–5.

164

217. Taylor RT, Lubick KJ, Robertson SJ, Broughton JP, Bloom ME, Bresnahan WA, et al. TRIM79α, an Interferon-Stimulated Gene Product, Restricts Tick-Borne Encephalitis Virus Replication by Degrading the Viral RNA Polymerase. Cell Host and Microbe. 2011 Sep;10(3):185–96.

218. Pletnev AG. Infectious cDNA Clone of Attenuated Langat Tick-Borne Flavivirus (Strain E5) and a 3′ Deletion Mutant Constructed from It Exhibit Decreased Neuroinvasiveness in Immunodeficient Mice. Virology. 2001 Apr;282(2):288– 300.

219. Munshi-South J, Zolnik CP, Harris SE. Population genomics of the Anthropocene: urbanization is negatively associated with genome-wide variation in white-footed mouse populations. Evol Appl. 2016 Feb 11;9(4):546–64.

220. Mitzel DN, Best SM, Masnick MF, Porcella SF, Wolfinbarger JB, Bloom ME. Identification of genetic determinants of a tick-borne flavivirus associated with host-specific adaptation and pathogenicity. Virology. 2008 Nov;381(2):268–76.

221. Laurent-Rolle M, Boer EF, Lubick KJ, Wolfinbarger JB, Carmody AB, Rockx B, et al. The NS5 Protein of the Virulent West Nile Virus NY99 Strain Is a Potent Antagonist of Type I Interferon-Mediated JAK-STAT Signaling. Journal of Virology. 2010 Mar 8;84(7):3503–15.

222. Hall RA, Broom AK, Smith DW, Mackenzie JS. The ecology and epidemiology of Kunjin virus. Curr Top Microbiol Immunol. 2002;267:253–69.

223. Hoang-Le D, Smeenk L, Anraku I, Pijlman GP, Wang XJ, de Vrij J, et al. A Kunjin replicon vector encoding granulocyte macrophage colony-stimulating factor for intra-tumoral gene therapy. Gene Ther. 2008 Dec 18;16(2):190–9.

224. Audsley M, Edmonds J, Liu W, Mokhonov V, Mokhonova E, Melian EB, et al. Virulence determinants between New York 99 and Kunjin strains of West Nile virus. Virology. 2011 May;414(1):63–73.

225. Dai J, Pan W, Wang P. ISG15 facilitates cellular antiviral response to dengue and west nile virus infection in vitro. Virol J. 2011;8(1):468–7.

226. Duggal NK, Emerman M. Evolutionary conflicts between viruses and restriction factors shape immunity. Nat Rev Immunol. 2012 Sep 14;12(10):687–95.

227. Diamond MS, Farzan M. The broad-spectrum antiviral functions of IFIT and IFITM proteins. Nature Publishing Group. Nature Publishing Group; 2012 Dec 14;13(1):46–57.

165

228. Perez-Caballero D, Zang T, Ebrahimi A, McNatt MW, Gregory DA, Johnson MC, et al. Tetherin inhibits HIV-1 release by directly tethering virions to cells. Cell. 2009 Oct;139(3):499–511.

229. Black LR, Aiken C. TRIM5 Disrupts the Structure of Assembled HIV-1 Capsid Complexes In Vitro. Journal of Virology. 2010 Jun 8;84(13):6564–9.

230. Izuogu AO, McNally KL, Harris SE, Youseff BH, Presloid JB, Burlak C, et al. Interferon signaling in Peromyscus leucopus confers a potent and specific restriction to vector-borne flaviviruses. PLoS ONE. 2017;12(6):e0179781.

231. Ashour J, Laurent-Rolle M, Shi PY, Garcia-Sastre A. NS5 of Dengue Virus Mediates STAT2 Binding and Degradation. Journal of Virology. 2009 May 12;83(11):5408–18.

232. Overby AK, Popov VL, Niedrig M, Weber F. Tick-Borne Encephalitis Virus Delays Interferon Induction and Hides Its Double-Stranded RNA in Intracellular Membrane Vesicles. Journal of Virology. 2010 Aug 5;84(17):8470–83.

233. Harris SE, Munshi-South J, Obergfell C, O’Neill R. Signatures of Rapid Evolution in Urban and Rural Transcriptomes of White-Footed Mice (Peromyscus leucopus) in the New York Metropolitan Area. Johnson N, editor. PLoS ONE. 2013 Aug 28;8(8):e74938–19.

234. Ye J, Zhu B, Fu ZF, Chen H, Cao S. Immune evasion strategies of flaviviruses. Vaccine. Elsevier Ltd; 2013 Jan 7;31(3):461–71.

235. Shi Y, Pulliam DA, Liu Y, Hamilton RT, Jernigan AL, Bhattacharya A, et al. Reduced mitochondrial ROS, enhanced antioxidant defense, and distinct age- related changes in oxidative damage in muscles of long-lived Peromyscus leucopus. AJP: Regulatory, Integrative and Comparative Physiology. 2013 Mar 1;304(5):R343–55.

236. Csiszar A, Labinskyy N, Zhao X, Hu F, Serpillon S, Huang Z, et al. Vascular superoxide and hydrogen peroxide production and oxidative stress resistance in two closely related rodent species with disparate longevity. Aging Cell. 2007 Sep 14;6(6):783–97.

166