sign in sign up
<<
Home , Colorado tick fever, Councilman body, Delicate deer mouse, Dengue fever, Flaviviridae, Japanese encephalitis

University of Northern Scholarship & Creative Works @ Digital UNC

Dissertations Student Research

5-2020

Comparative Response and Regulation in Deer Mice and Syrian Golden Hamsters During With Modoc

Tyler Sherman

Follow this and additional works at: https://digscholarship.unco.edu/dissertations

Recommended Citation Sherman, Tyler, "Comparative Host Response and Gene Regulation in Deer Mice and Syrian Golden Hamsters During Acute Infection With " (2020). Dissertations. 687. https://digscholarship.unco.edu/dissertations/687

This Text is brought to you for free and open access by the Student Research at Scholarship & Creative Works @ Digital UNC. It has been accepted for inclusion in Dissertations by an authorized administrator of Scholarship & Creative Works @ Digital UNC. For more information, please contact [email protected].

UNIVERSITY OF NORTHERN COLORADO

Greeley, Colorado

The Graduate School

COMPARATIVE HOST RESPONSE AND GENE REGULATION IN DEER MICE AND SYRIAN GOLDEN HAMSTERS DURING ACUTE INFECTION WITH MODOC VIRUS

A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

Tyler Sherman

College of Natural and Health Sciences School of Biological Sciences Biological Education

May 2020

This Dissertation by: Tyler Sherman

Entitled: Comparative Host Response and Gene Regulation in Deer Mice and Syrian Golden Hamsters During Acute Infection with Modoc Virus has been approved as meeting the requirement for the Degree of Doctor of Philosophy in College of Natural and Health Sciences, in School of Biological Sciences, Program of Biological Education

Accepted by the Doctoral Committee

______Ann C. Hawkinson, Ph.D., Research Advisor

______Patrick D. Burns, Ph.D., Committee Member

______Susan M. Keenan, Ph.D., Committee Member

______Stephen P. Mackessy, Ph.D., Committee Member

______David Hydock, Ph.D., Faculty Representative

Date of Dissertation Defense _____March 27, 2020______

Accepted by the Graduate School

______Cindy Wesley, Ph.D. Interim Associate Provost and Dean Graduate School and International Admissions

ABSTRACT

Sherman, Tyler. Comparative Host Response and Gene Regulation in Deer Mice and Syrian Golden Hamsters During Acute Infection with Modoc Virus. Published Doctor of Philosophy Dissertation, University of Northern Colorado, 2020

Flaviviruses (family , ) comprise a diverse group of globally-distributed that may specifically infect , invertebrates, or both. Of those that can replicate in both invertebrates (such as and mosquitoes) as well as vertebrates (such as ), many represent important . Examples of these include

(DENV), which may cause severe hemorrhagic , (WNV),

Japanese virus (JEV), and (ZIKV), all of which may cause febrile illness with neurologic involvement, and virus (YFV), which may induce hepatic dysfunction. Given that these viruses (among others) are carried and transmitted by ubiquitous whose range is increasingly expanding, hundreds of thousands of individuals worldwide are annually infected and succumb to either or . Further, despite monumental strides in molecular research and medical practice over the preceding decades, are limited, and treatment for those infected relies exclusively on supportive care.

While there are many obstacles to overcome in developing effective vaccines and antiviral therapeutics, one particular challenge during flaviviral infection lies in differentiating which aspects of the innate lead to viral

iii clearance or persistence and which may induce various immunopathologies. To glean further understanding in this respect, the research outlined in this dissertation utilizes two unique models to investigate properties of and immune gene expression during the early stages of infection with a rodent-borne flavivirus (Modoc virus; MODV). By first examining these features in the deer ( maniculatus), which serves as the putative reservoir host of the virus in nature, it was discovered that viral RNA could be detected from all of the tissues and fluids examined throughout the experimental duration while infectious virus could be only be recovered from the kidneys, , , and at later timepoints. Concomitant with this, the profile of select immune (IFN-γ, TNFα, IL-6, IL-10, and TGFβ) were shown to be differentially expressed in the brain as compared to all other tissues. Likewise, expression profiles significantly differed at a collection timepoint midway through the experiment (day 6) compared to the first (day 1) and last (day 10) timepoint of collection, suggesting that deer mice respond to MODV infection through a tightly-regulated expression of innate immune genes that is both tissue- and time-dependent. These features were then compared to those observed in Syrian golden hamsters (Mesocricetus auratus), which, unlike the deer mice, develop illness with neurologic involvement reminiscent of some flaviviral in . In hamsters, the patterns of viral replication different from deer mice in that the levels of MODV RNA were lower in all tissues at the first collection timepoint but then surpassed the levels observed in deer mice by day 4 and generally remained elevated throughout the remainder of the experiment.

iv

Although infectious virus could only be recovered from the kidneys and of infected hamsters, successful isolation was evident at all sampling timepoints suggesting that hamsters are less able to control the production of viral progeny in these tissues. In regard to gene expression, multivariate analysis depicts a differential expression of the important antiviral IFN-γ between deer mice and hamsters, a response which is especially evident in the brain tissue beginning at day 6 post-infection. This finding may represent a potential mechanism for immunopathology, which allows for increased viral dissemination into the CNS following the breakdown of vascular integrity after elevated proinflammatory signaling. All in all, there is hope that this research may contribute to the greater scientific community’s collective understanding of flaviviral and provide a novel avenue to pursue in the development of antiviral therapeutics.

v

ACKNOWLEDGEMENTS

To properly acknowledge all of those who have helped and supported me throughout this journey would require an entire dissertation of its own. For now, I wish to simply extend the maximum amount of gratitude that can be conveyed through words to the numerous individuals who were there to encourage, teach, mentor, guide, and challenge me throughout my doctoral research.

To my colleagues, both undergraduate and graduate alike, I thank you for your presence in this academic community. We no doubt have kept each other in check throughout the years, both emotionally as well as intellectually. For those of you that I have taught, thank you for providing a sense of purpose in my life as a blossoming educator. I hope I was able to reciprocate this positivity for you in some way. For those of you that have assisted me in my research, namely Bryce

Fuchs, Mykee Cain, James Butts, Brandon Clark, Jessica Pardue, and Jordan

Davenport, I thank you for your time and efforts; you are all great individuals to work with and wonderful people to get to know. I hope your experience with me as a research mentor was equally as enjoyable. To my fellow graduates, I thank you for your efforts in contributing toward the maintenance of a research community built on integrity. Your endless hours of hard work do not go unnoticed, especially by those working alongside you.

vi

Likewise, I would like to thank all of the faculty in the School of Biological

Sciences whose work ethic and commitment to delivering high-quality education has set the precedence for me as an individual. During my time in the program, you have demonstrated your tenacity and perseverance time and time again, and

I am certain that I have you to thank for setting such important examples. In particular, I would like to thank Drs. Holt, Fisher, Reinsvold, and Higgins for your aid in developing my knowledge and skills as an educator. Further, I would like to thank my committee members, Drs. Hydock, Keenan, Mackessy, and Burns for your guidance, advice, and patience throughout the design, implementation, and analysis of my research project, as well as the countless questions that accompanied. As well, thank you so much, Cindy and Elizabeth, for all of your help and friendship over the years.

Specifically, I would like to thank my lab mate Doug Petty, as well as my research advisor, Ann Hawkinson. You have both taught me so much over the years, not only in regard to academia, but likewise in various other walks of life.

You have both helped me, mentally as well as physically, to achieve my goals.

Most importantly, you each continuously voiced your confidence in me. Without you, and the roles that you’ve played in the my life over the last five and a half years, I know for certain that I wouldn’t be where I am today. I could never thank you enough.

Finally, I would like to extend the upmost thanks and gratitude to my closest friends and family for your unwavering support - Ali, Dad, Mom, Oma,

Tony, Tyler, Laura, Drew, Christine, Eric, Andrew, Zach, Anna, Brittney, Klint,

vii

Olivia, and Cara, among so many others. I consider my life enriched simply by your presence, let alone the energy that you have allocated to interacting with me throughout this roller coaster of an experience. Thank you.

viii

TABLE OF CONTENTS

CHAPTER I. INTRODUCTION ...... 14

II. REVIEW OF THE FLAVIVIRUS LITERATURE ...... 18

Abstract ...... 18 ...... 20 Flavivirus , Structure and Replication ...... 21 Flavivirus Phylogeny, Ecology and Distribution ...... 31 Modoc Flavivirus ...... 40 Recognition and Antiviral Signaling ...... 46 Flavivirus Immune Subversion and Host Pathology ...... 56 and Treatment Against Flaviviral Disease ...... 62 The Use of Models in Scientific Research ...... 67 Deer Mice ...... 70 Syrian Golden Hamsters ...... 73

III. HOW TO HOST UNINVITED GUESTS: CLUES ABOUT THE ANTIVIRAL RESPONSE IN A RESERVOIR MODEL OF ACUTE FLAVIVIRAL INFECTION ...... 77

Abstract ...... 77 Introduction ...... 79 Materials and Methods ...... 82 Results ...... 91 Discussion ...... 102

IV. COMPARATIVE HOST RESPONSE AND GENE REGULATION IN DEER MICE AND SYRIAN GOLDEN HAMSTERS DURING ACUTE INFECTION WITH MODOC VIRUS ...... 110

Abstract ...... 110 Introduction ...... 113 Materials and Methods ...... 117 Results ...... 127 Discussion ...... 138

V. CONCLUSIONS ...... 151

ix

VI. REFERENCES...... 157

APPENDIX A MELT CURVE ANALYSES FOR MODV- SPECIFIC AND β-ACTIN PRIMERS ...... 183

APPENDIX B GRAPHS OF RELATIVE GENE EXPRESSION WITH BIOREPLICATES AT EACH TIMEPOINT REPRESENTED (DEER MICE) ...... 185

APPENDIX C GRAPHS OF RELATIVE GENE EXPRESSION WITH BIOREPLICATES AT EACH TIMEPOINT REPRESENTED (HAMSTERS) ...... 188

APPENDIX D EXPERIMENTAL INFECTION OF DEER MICE AND HAMSTERS WITH MODOC VIRUS: IACUC MEMORANDUM APPROVAL ...... 191

x

LIST OF TABLES

Table:

2.1 Assignment of Flaviviruses to Clades and Clusters ...... 23

3.1 Sequence Data and Amplification Efficiency for Deer Mouse Primers ...... 91

4.1 Sequence Data and Amplification Efficiency for Deer Mouse Primers ..... 126

4.2 Sequence Data and Amplification Efficiency for Hamster Primers ...... 127

xi

LIST OF FIGURES

Figure:

2.1 Representation of a Typical Flavivirus Virion ...... 25

2.2 Representation of Organization of the Flavivirus Genus ...... 28

2.3 Overview of the General Flavivirus Replication Cycle ...... 31

2.4 Phylogenetic Representation of the genus Flavivirus ...... 33

2.5 Global Distribution of Medically-Important Flaviviruses ...... 38

2.6 Locations in the US of the aegypti ...... 40

2.7 Replication Kinetics of WNV, Wild-type MODV, and Chimeric MODV ...... 45

2.8 Simplified Schematic of Type I Through III IFN Signaling Pathways ...... 51

2.9 Simplified Phylogeny of the Muridae and Rodent Families ...... 71

3.1 Viral RNA Loads in Deer Mouse Urine and Sera ...... 94

3.2 Viral Loads and Successive Virus Isolations in Deer Mouse Tissues ...... 95

3.3 PCA of Gene Expression by Tissue ...... 97

3.4 PCA of Gene Expression by Collection Timepoint ...... 98

3.5 Gene Expression in Select Tissues of MODV-Infected Deer Mice ...... 100

3.6 Expression of Oas1b in Select Tissues of MODV-Infected Deer Mice ..... 101

4.1 Weight Gain or Loss in Deer Mice and Golden Hamsters ...... 129

4.2 Signs of Gross Pathology in Infected Hamsters ...... 130

4.3 Viral RNA Loads in Tissues and Fluids of Deer Mice and Hamsters ...... 133

4.4 Cytokine Gene Expression Levels In Deer Mice and Hamsters ...... 137

xii

4.5 Expression levels of Oas1b in Deer Mouse and Hamster Tissues ...... 138

5.1 Increased Prevalence of -Borne in the US...... 152

xiii

CHAPTER I

INTRODUCTION

Viruses such as West Nile virus (WNV), virus

(JEV), dengue virus (DENV), and Zika virus (ZIKV), along with more than 70 additional representatives, comprise a group of clinically-relevant infectious diseases. Although previously concentrated in the of , sub-

Saharan , and South Eastern , recent years have witnessed a migration of these viruses into areas outside of their historically range

(1). To exemplify this, the Center for Disease Control and Prevention (CDC) has reported 1,594 cases of WNV neuroinvasive disease in the US during 2018 alone

(2). Concerningly, mortality subsequent to WNV infection rose to nearly 12% during this period. Fortunately, laboratory-confirmed cases of symptomatic ZIKV infection have witnessed a decline since the notable 2016 outbreak; however, this virus continues to threaten equatorial countries around the globe and is of particular concern for pregnant women (3). DENV is now considered to be the most prevalent -borne virus in the world, with an estimated annual of severe disease and mortality in approximately 500,000 and 22,000 individuals, respectively (4, 5).

Despite numerous decades of research in an attempt to elucidate the underpinnings of infectious disease, many characteristics of remain incompletely understood. In large part, this challenge can be attributed to

14 the myriad of interactions which take place between a particular strain of virus and the infected host. Indeed, it is known that the unique genetic landscapes of each biologic entity strongly impact one another and ultimately influence the emergence of clinical manifestations. What remains to be further clarified, however, are the questions of which genetic elements are the most important and how much variance can be tolerated before leads to overt pathology.

Modoc virus (MODV) is a rodent-borne flavivirus initially isolated from deer mice in Modoc County, . While infection with MODV is not typically considered a notable threat to the general public, there has been at least one documented case in which MODV led to neurologic disease in a young boy (6).

Notwithstanding the general lack of human disease induced by this virus, there exists many genetic similarities between MODV and other, more virulent flaviviruses. This considered, research with MODV offers a unique avenue to supplement our current understanding of flaviviral infection and disease. When coupled with the use of comparative rodent models, it is possible to examine features of the disease process and subsequent immune response in a more complete, holistic manner. This research aims to accomplish the aforementioned goal by conducting experimental infections with MODV in deer mice, which resist disease from the virus, as well as Syrian golden hamsters, which recapitulate human-like illness once infected (7, 8). During an acute experimental infection period, tissues and excrement samples are collected and analyzed to examine two fundamental aspects of virological research: a) where and how quickly does

15 the virus replicate and b) on the molecular level, how does the host respond?

Answering these questions, especially in the context of comparative model systems, may provide novel information in regard to the greater scientific community’s understanding of flavivirus pathology. Ultimately, it is the hope that examining these viral infections through a unique lens will help to bolster the future development of vaccines or antiviral therapeutics.

A1 Review the flavivirus literature to define present framework of knowledge and attempt to identify gaps in current understanding of flaviviral pathology.

H1 As most research conducted to date has utilized inbred rodent models and serially-passaged isolates of lab-kept flaviviruses, there may be value in the development of unique and comparative model systems. Reviewing the literature and synthesizing an in-depth framework of understanding will help to reveal areas where research with flaviviruses is lacking, thereby providing the impetus for the development of novel research approaches.

A2 Examine aspects of host pathology, characterize viral replication kinetics, and investigate the expression of select immune genes in deer mice during acute experimental infection with Modoc virus (MODV).

H2 As the putative reservoirs of MODV, it is hypothesized that deer mice will demonstrate no signs of clinical pathology throughout the experimental duration. Likewise, despite the absence of any overt , infectious virus and viral RNA should be detectable in various tissues at multiple timepoints; however, quantities of detectable virus will be sustained at relatively low levels. Finally, it is hypothesized that the relative expression values of immune genes will vary across tissues and over time, thereby eliciting a tissue- and time-specific protective effect against both viral burdens as well as immunopathologies that may accompany a sustained proinflammatory response.

16

A3 Compare aspects of host pathology, viral replication kinetics, and immune gene expression between deer mice and Syrian golden hamsters at multiple timepoints during acute experimental infection with MODV.

H3 It is hypothesized that Syrian golden hamsters will demonstrate clinical signs of pathology when infected with MODV, thereby constituting a comparative pathogenic model. As such, viral loads should differ significantly in various tissues over time when compared to deer mice. Finally, the relative expression values of immune genes should differ significantly between hamsters and deer mice in various tissues at distinct timepoints throughout the experimental duration, highlighting the different modalities of immune regulation elicited by each rodent model even when pathogen dose and route of infection are standardized.

17

CHAPTER II

REVIEW OF THE FLAVIVIRUS LITERATURE

Abstract

Flaviviruses (family Flaviviridae, genus Flavivirus) are one of the oldest lineages of zoonotic viruses studied to date. Most medically-important examples, such as dengue virus (DENV) or West Nile virus (WNV), are vectored by mosquitoes, although others, such as -borne encephalitis virus (TBEV), are transmitted by ticks. Further, there exists a small intrageneric clade of non- vectored representatives, such as Modoc virus (MODV) or Rio Bravo virus

(RBV), that are transmitted directly from reservoir hosts like and bats, respectively. Indeed, flaviviruses have been isolated from nearly every continent across the globe and comprise more than 40 distinct that have been documented to cause disease in humans and domesticated . Although the majority of infections result in only sub-clinical or mild disease, various flaviviruses have periodically caused notable regional outbreaks. During these , infected populations may exhibit a broad range of clinical pathologies, the most severe of which include hemorrhagic fevers or illness with neurologic involvement. Despite many years of laboratory and clinical research utilizing a multitude of model systems, the exact mechanisms of most flaviviral pathologies remain incompletely understood. For example, while it is known that genetic polymorphisms, found in both the virus as well as the host, have the potential to

18 strongly influence the outcome of clinical disease, the observed variance that exists among these interactions make it difficult to pinpoint which polymorphisms are the most impactful. As well, it is now clear that the itself may contribute to disease under the context of many flaviviral infections; however, the degree to which this occurs, as well as the precise alterations in the immune response that lead to either viral clearance or disease exacerbation, appear to be extremely virus-host specific and therefore present a significant challenge in the development of antiviral therapeutics aimed at protecting large populations. In light of these obstacles, it appears that there is a need to continue developing diverse model systems that allow for the extrapolation of research data which accurately reflects the genetic diversity present in human populations. One way to accomplish this aim might be to implement the use of non-traditional rodent models (e.g., deer mice, golden hamsters, etc.) that are maintained in relatively outbred stocks as compared to the frequently-used inbred strains of the domestic laboratory mouse (Mus musculus). Finally, there could be value in the use of divergent flaviviruses, such as those belonging to the non-vectored clade, as comparative models that help to distinguish any subtle but universal discrepancies in the flaviviral genome or replication cycle that most significantly impact the outcome of disease.

19

Flaviviruses

Shortly following the widespread acceptance of germ theory in the late

1800s came the conscious awareness that certain diseases may be carried and transmitted to new hosts through intermediate vectors. Consequently, this realization occurred nearly concomitant with the onset of as a research discipline, arguably pioneered by the U.S. Army physician (among others) during his work in . Due to the prevalence of hemorrhagic fevers that continuously claimed the lives of canal workers, Reed sought to elaborate on the hypotheses of , a Cuban doctor who postulated that mosquitoes

(rather than direct contact) might, in fact, facilitate the spread of disease. In time,

Reed’s investigations led to the discovery of virus (YFV), the prototype virus that would later establish one of the most pervasive groups of viruses on the entire planet (9, 10).

As of 2017, a total of 89 distinct species have been described within the family Flaviviridae (11). This taxonomic collective of viruses has been further divided into four discrete genera on the basis of serologic cross-reactivity, unique biologic interactions, and minor discrepancies in viral genome organization and replication strategy (11, 12). Like the prototypic YFV, after which the family was named (the Latin flavus referring to the jaundiced appearance of diseased individuals), most representatives become recognized as medically-relevant when horizontal to susceptible hosts (e.g., humans and other mammals) occurs through the feeding of hematophagous such as ticks and mosquitoes. For this reason, flaviviruses are sometimes referenced

20 along with other, non-related viruses (e.g., togaviruses) as the

(arthropod-borne viruses) (13).

Fortunately, the vast majority of flaviviral infections are speculated to self- resolve following only mild or subclinical illness; however, instances of severe disease are not uncommon and, depending on the specific virus, may result in multiple-organ failure, autoimmune disorders, malformations in fetal development, hemorrhagic fevers with vascular leak , or meningioencephalitic ailments with the potential for permanent neurologic sequelae (11). Notable representatives of the Flavivirus genus include dengue virus (DENV), West Nile virus (WNV), tick-borne encephalitis virus (TBEV), and

Japanese encephalitis virus (JEV), as well as the increasingly recognized Zika virus (ZIKV). Other members of the Flaviviridae family include C virus

(HCV) of the genus and GBV-A virus of the genus, both of which are transmitted to new hosts exclusively through the intermixing of bodily fluids (14). Lastly, members of the genus (e.g., bovine viral virus, BVDV) are best known for their ability to elicit economically significant disease in ruminant animals with only rare instances of spillover into novel mammalian hosts (15, 16). The remainder of this literature review will focus solely on the current knowledge of the Flavivirus genus unless otherwise noted.

Flavivirus Taxonomy, Structure and Replication

Various members contained within the Flavivirus genus exhibit astonishing dissimilarity in regard to sequence and viral ecology. In one of the most extensive studies conducted to date on the organization of these viruses,

21

Kuno et al. describe a schematic in which the genus can be further categorized into the descending hierarchy of clusters (determined by 95% bootstrap cutoffs and vector-host associations), clades (groups which share ≥ 69% (whole- genome) sequence identity among one another) and species (groups with share

> 84% (whole-genome) sequence identity) (Table 2.1) (17). Finally, Kuno et al.

(17) denote viral strains as isolates from distinct geographic areas where a particular species is known to be endemic. As can be expected from RNA viruses, even strains within a species may exhibit remarkable variation (e.g. a maximum of 14% nucleotide dissimilarity between strains of YFV) (17).

Nevertheless, there are numerous conserved features that unite the taxon as a whole.

22

Table 1 Table 2.1: Assignment of flaviviruses to clusters and clades. Note. Modified table from Kuno et al. demonstrating the various clades, vectors, and distributions of distinguished representatives from the genus Flavivirus. Of particular note are the updated distributions of each representative based on current surveillance records available at the time of writing this review (17).

Virus Clade Traditional Vector a Distribution Non-vector Cluster San Perlita I N (rodent) Modoc II N (rodent) Western US Rio Bravo III N (bat) US, Mexico

Tick-borne Cluster Powassan IV Tick, mosquito , Tick-borne encephalitis IV Tick ( spp.) Russia, , Asia

Mosquito-borne Cluster Yellow fever VII Mosquito (Aedes spp.) Tropical regions Dengue-1 to -4 VIII Mosquito (Aedes spp.) Africa, , Zika X Mosquito (Aedes spp.) Asia, Africa, Americas West Nile XIV Mosquito ( spp.) Global (Temperate to Tropics) Japanese encephalitis XIV Mosquito (Culex spp.) Asian and parts of the Pacific a In vitro and in vivo evidence suggests that some MBFVs can be recovered from non-traditional vectors.

Universally, flaviviruses are distinguished as spherical, lipid-enveloped virions which contain a positive-sense RNA genome within a virally-encoded nucleocapsid (C) (11) (Figure 2.1). Immediately external to the are the virally-encoded and tightly-associated matrix (M) and envelope (E) , the combination of which encompasses a mature virion that ranges in size from approximately 40 to 60 nm. The flaviviral genome (generally 10-11 kb in length) encodes a single polyprotein within an (ORF) that is co- and post-translationally cleaved by host and viral to form the three structural proteins (C, M, and E) and at least seven non-structural proteins

(NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) (Figure 2.2). Flanking the

ORF on the 5’ and 3’ ends of the genome are untranslated regions (UTRs) of approximately 100 and 400-700 , respectively. Contained within the 5’

23

UTR is a conserved, Y-shaped, stem-loop A structure (SLA) believed to aid in viral replication by binding to complementary sequences in the 3’ end and circularizing the viral genome (18). Beyond the SLA, the 5’ end is covered by a type I methyl cap that closely resembles those of host cell transcripts.

Importantly, this cap serves the multipurpose role of preventing cytosolic degradation by passive exonucleases, inhibiting viral recognition within the cell and subsequent degradation by host antiviral proteins, and by recruiting host cell translational machinery. On the opposite end, the 3’ UTR contains a stem-loop

(SL) structure, which acts as the initiation site for negative-sense RNA synthesis.

Finally, most flaviviruses contain a non-coding RNA downstream of the SL that is likewise multifunctional (e.g. sfRNA; see section titled “Flavivirus Immune

Subversion and Host Pathology”) followed by a 3’ hydroxyl tail. Interestingly, while the presence of SLA and SL structures remain conserved across all examined flaviviruses, minor discrepancies in their sequences (as well as additional structures) in the 5’ and 3’ UTRs between representatives have been posited to elicited some of the variable pathogenesis witnessed during clinical infection (18, 19).

24

1.12.1 Figure. 2.1: Schematic representation of a typical virion from the genus Flavivirus. (A) Depicts RNA genome contained within proteins layered beneath the lipid membrane. Immature virions (left) maintain prM “caps” over E proteins to prevent preemptive fusion of viral progeny within host cell secretory pathways. In contrast, E proteins in mature virions (right) are anchored to the lipid membrane by prM-derived M proteins following cleavage by host proteases. “Anchor” and “stem” indicate discrete domains within the E protein as characterized by soluble forms of E (sE). (B) Computer-generated model of complete (mature) DENV 2 virion in which different shades indicate each of the three E protein domains. Modified from Vratskikh et al., 2013 (20).

As depicted in Figure 2.1, mature virions utilize a unique architecture in which E proteins lay parallel to the lipid bilayer in a “herringbone” configuration.

These proteins form ninety head-to-tail homodimers that undergo rotational rearrangement during final virion maturation to create environmentally-stable particles with icosahedral symmetry (21). Each E protein itself contains three structural domains: ED-I, ED-II, and ED-III. Virions are believed to accomplish fusion within pre-lysosomal endocytic vesicles via a fusion loop contained within the ED-II domain, while the ED-III domain acts as the primary antigenic epitope recognized by neutralizing (22). ED-I represents the central structural domain of the E protein and contains sites (determined by specific

25 hosts and viral strains) that are speculated to influence virion pH sensitivity and neuroinvasiveness (22). In addition to the E protein, virions display a superficial membrane (M) protein created by proteolytic cleavage of the pre-membrane

(prM) protein during viral maturation along the cellular exocytic pathway.

While it has been established that the E protein facilitates virus binding to susceptible host cells, the exact process by which virions are internalized is incompletely understood. In contrast to other well-studied viruses for which bona fide cognate receptors have been clearly elucidated, it is suggested that many flaviviruses may instead utilize multiple binding partners to aggregate on a cell’s surface before undergoing -mediated in either a - dependent or -independent manner (23, 24). Among these putative binding receptors are glycosaminoglycans (GAGs), such as heparan-sulfate and syndecans, which are expressed in high concentrations on nearly all cell surfaces and may, therefore, provide a rationale for the broad spectrum of tissue tropism by flaviviruses (23). Virion aggregation in this fashion is most likely a consequence of electrostatic attraction between positive and negative residues on the lateral surface of the E protein and sulfate groups on the polysaccharides of GAGs, respectively (25). However, while GAGs may contribute to the low- affinity virion aggregation on the cellular surface, their inability to initiate receptor- mediated endocytosis suggests that alternative receptors are required for . Additional co-receptors known to facilitate flaviviral attachment include the

C-type lectin receptors (CLRs), a class of pattern-recognition receptors (PRRs) expressed on myeloid-derived cells that recognize carbohydrate motifs found on

26 some pathogenic organisms (23). Of these PRRs, mannose receptors (MRs), -specific intercellular adhesion molecule-3-grabbing non-

(DC-SIGNs), and liver/-specific ICAM-3 grabbing non-integrins (L-

SIGNs) have all been implicated during in vitro infection of mammalian cells lines

(23, 26). Lastly, binding to αvβ3 integrins has been demonstrated for WNV (27); however, along with the aforementioned CLRs, productive infection of cell lines lacking these receptors further supports the notion that additional receptors or activation factors are needed to initiate endocytosis following viral adsorption (28,

29). In any case, it is important to note that the ability for some flaviviruses to utilize various receptors conserved throughout multiple phyla (e.g. invertebrates and vertebrates) serves as an important contribution to the viral ecology of these relatively promiscuous pathogens.

Once internalized, virions are trafficked through the endocytic pathway, where a decrease in pH induces rearrangement of E protein homodimers and facilitates the fusion of viral and host-cell lipid membranes (Figure 2.3) (30).

Following the dissolution of the nucleocapsid and release of viral nucleic acids into the host , viral RNA is targeted to the (ER), where the ~3400 amino acid polyprotein weaves through the ER membrane during (Figure 2.2). For most members, initiation of translation occurs via ribosomal scanning in a cap-dependent manner; however, the finding that

DENV may circumvent cap dependency suggests that other flaviviruses may be able to as well (19). Further, it has been demonstrated with DENV that a small

27 hairpin loop in the C gene facilitates ribosomal pausing over authentic start codons (i.e. AUG) in lieu of the typical eukaryotic Kozak motif (31).

A

B

C

3 Figure 2.2: Schematic representation of genome organization and polyprotein translation for members of the Flavivirus genus. (A) Depicts genomic polyprotein with 5’ UTR, cap and SLA on N-terminal end (left) along with 3’ UTR containing the SL, sfRNA, and hydroxyl tail on C-terminal end (right). (B) Organization of structural (C, prM/M, and E) and nonstructural (1, 2A, 2B, 3, 4A, 4B, and 5) proteins along with respective cleavage sites from host and viral proteases. (C) Depicts orientation of translated polyprotein throughout ER membrane along with respective cleavage sites; top of image represents ER lumen while the bottom represents the host cell cytoplasm. Modified from Fields et al., 2007 (32).

Upon translation and proteolytic cleavage, flaviviral nonstructural proteins contribute to the production of progeny virus through the induction of cellular remodeling and host immune subversion. Namely, the formation of vesicle packets (VPs) within convoluted membranes of the host ER is thought to provide a “place of refuge” for viral intermediates such as dsRNA (Figure 2.3 inset) (33,

34). Electron tomography studies conducted with DENV have revealed that VPs fully encompass viral components save for small cytoplasmic pores situated in

28 close proximity to distinct regions of viral budding from ER membranes (35).

Indeed, it has been suggested that these sites of co-localization increase viral replication efficiency by concentrating the NS proteins and other required components (e.g. host factors); however, further speculation has led to the working hypothesis that cellular remodeling in this manner may be a convergent strategy of diverse RNA viruses in which cytoplasmic replication drives the need to protect viral intermediates against passive immune surveillance (33). Lastly, the co-opted use and differential expression of -synthesizing such as hydroxy-methylglutaryl-CoA reductase (HMGCR) during flavivirus- induced cellular remodeling have been demonstrated to contribute to viral pathology by interfering with the host cell’s ability to respond to antiviral signaling

(36). As cholesterol is relocated away from surface membranes, lipid rafts (i.e. distinct regions of conglomerate, membrane-bound lipids, and proteins) lose functionality as sites of signal transduction in response to antiviral such as the (see the section titled “Pathogen Recognition and Antiviral

Signaling”). Aside from cellular remodeling, additional knowledge on the direct and indirect methods of immune subversion by flaviviruses are detailed later in this review.

Assembly of nascent virions likewise occurs within VPs as the structural proteins start to accumulate within luminal membranes, and NS proteins (e.g.

NS2A and NS3) begin to coordinate the shift from viral replication to packaging

(37). Through the use of cryogenic electron microscopy, the working explanation of virion morphogenesis suggests that cytoplasmic C protein dimers associate

29 with viral RNA in regions near E-prM heterodimers (at the time still bound to ER membranes via C-terminal anchor sequences) (38, 39). Budding of loosely- structured nucleocapsids into the ER lumen allows for the acquisition of a ER- derived lipid envelope containing a dense arrangement of pr-E heterodimers. As a point of scholarly interest, viral assembly in this manner is not uncommon; however, while the process is extremely rapid, theoretically producing up to hundreds of thousands of nascent virions within minutes, viruses must also contend with the inefficiency that accompanies packaging at such speeds. For example, premature assembly of nucleocapsids may lead to the production of virus-like particles containing truncated or no RNA at all. While some of these “empty shells” are simply immunogenic, those which contain sub-genomic portions (termed defective interfering (DI) particles) may actually enhance the infectivity of their wild type counterparts (40, 41). Nevertheless, budding virions are shuttled to the Golgi and through a secretory pathway equivalent to that used by newly synthesized host proteins destined for the plasma membrane (42-44).

Along the way, E proteins are occasionally glycosylated, a modification that has been shown to alter viral tropism and virulence in certain contexts (45). Upon entering the later stages of transport, the host catalyzes the structural maturation of immature virions by cleaving the pr domain from the M protein (46). At this time, it is believed that conservation of the pr-E heterodimers until just prior to release is necessary to prevent the acid-catalyzed rearrangement of E homodimers found on mature virions during trafficking through acid compartments along the secretory pathway (47).

30

4 Figure 2.3: Schematic overview of the general replication cycle for members of the genus Flavivirus (48). Of particular note are the vesicle packets (VP) formed along the perinuclear region of the ER membrane during viral replication and translation. In addition to shielding viral components from passive immune surveillance, VPs concentrate nonstructural proteins to form a replisome complex (lower right inset) (49). As well, the presence of immature virions and soluble pr proteins in the extracellular environment can be seen leading to the production of non-neutralizing antibodies and subsequent disease through -dependent enhancement (ADE). Images modified from Pierson, 2012, and Shi, 2014.

Flavivirus Phylogeny, Ecology and Distribution

As alluded to above, flaviviruses offer an interesting example of the dynamic interplay between and the environment. Overall, the genetic blueprint of the genus remains conserved enough that representatives can be distinguished from one another when coding regions are subjected to 31 phylogenetic analysis by either maximum-likelihood or neighbor-joining methods

(18, 50). Observing these analyses reveals consistent trends that indicate - specific (ISFV), no known vector (NKV), and tick-borne (TBFV) clusters all represent monophyletic groups while the mosquito-borne (MBFV) cluster remains less resolved (Figure 2.4). Despite this, clades within the MBFV cluster retain a certain level of stability that reflects a corresponding vector at the genus level

(e.g. Aedes vs. Culex spp.). Perhaps unsurprisingly, examining the entire nucleotide sequence of select representatives (17) does not appear to improve the resolution of these clades any further; however, the persistence of phylogenetic grouping that adheres to both sequence identity as well as vector specificity adds support to the hypothesis that conserved elements (e.g. SLA and sfRNAs) in non-coding regions may additionally contribute to important features of flaviviral ecology (18, 51, 52). Likewise, cross-examinational studies both in vivo and in vitro have clarified that while diverse mosquito vectors may be susceptible to MBFVs outside of their historical range (e.g. DENV in Culex spp. or WNV in Aedes spp.), viral titers recovered from these alternative vectors are typically lower and therefore suggest a reduced competency for transmissive ability concomitant with support for the coevolution of MBFV with their traditional vectors (53, 54).

32

5 Figure 2.4: Phylogenetic representation of the genus Flavivirus based on maximum-likelihood analysis from coding regions. Predicted amino acid substitutions per site are indicated by the scale bar. Of particular note are the monophyletic clades comprised of insect-specific, no known vector, and tick- borne representatives compared to the relatively unresolved mosquito-borne clade. Despite the poor resolution, however, there appears to be some level of phenotypic selection that favors the maintenance of various mosquito-borne members within specific vectors at the genus level. Modified from Bolling et al., 2015 (50).

33

Due to their medical relevance, MBFVs and TBFVs have been studied extensively and are known to exhibit varying degrees of replication efficiency and host pathogenesis in both and invertebrate tissues and/or cell lines.

Alternating between such disparate hosts highlights a remarkable adaptation of these viruses given their ability to utilize divergent cellular machinery while also subverting host-specific immune responses. Perhaps the most extraordinary example of this comes from WNV, which, despite a predominant circulation in various avian species, has been documented to additionally infect humans, equines, and at least in one instance, a captive orca (55). Due to low transmissive probability, humans and equines are viewed as “dead-end” hosts when infected with WNV; however, may act either as reservoirs or amplifying hosts of WNV, providing either a persistence or high-titer source of virus for ornithophilic mosquitoes, respectively (56). Although rare, WNV has even been recovered from multiple tick lineages while feeding on wild birds (57,

58).

Similar to WNV, there exists a collective agreement that birds (along with pigs) may act as reservoirs or amplifying hosts of JEV (59). In contrast, other

MBFVs such as YFV appear to abide by a more restrictive host range in which non-human (NHPs) serve as the primary reservoir. In this instance, the movement of virus between mosquitoes and NHPs denotes a transmission cycle identified as “jungle yellow fever,” while spillover and subsequent circulation in humans are sometimes referred to as “urban yellow fever” (60). Similarly, NHPs serve as the ostensible reservoirs of DENV and ZIKV (4, 61). Distinguishing the

34 two from one another, however, is lack of evidence supporting that vertebrates other than NHPs can be infected by DENV in nature while antibodies against

ZIKV have been recovered from domestic sheep and goats, horses, cows, ducks, rodents, bats, orangutans, and carabaos (61). Nonetheless, a failure to establish persistent viral infection along with an inherent ability to resist disease from

DENV and ZIKV further indicates that rodents are unlikely reservoirs of these

MBFVs (61, 62).

Apart from the aforementioned considerations which govern the viral ecology of MBFVs and TBFVs in diverse hosts, additional mechanisms that perpetuate the existence of these viruses in nature include instances of vertical transmission, non-vectored , and although uncommon, infection through contact with blood products, aerosols, or following consumption of unpasteurized human and animal milk (11, 63). For arthropods, vertical transmission may occur transovarial (from parent to offspring) or transstadial

(from one life stage to the next), both of which are important for ISFVs given their failure to replicate in vertebrates and thus, in contrast to MBFVs and TBFVs, an inability to rely on transmission to uninfected insect hosts through vertebrate intermediates (50). Intriguingly, the selective pressure for vertical transmission does not appear to hold true for NKVs despite a similar paradigm in which only vertebrate tissues and cell lines are permissive to infection (64). In the case of

NKVs, experimental infection of rodent mothers either failed to produce sufficient virus or antibody titers in offspring or, in one case, led to fetal pathogenesis, which contradicts support for vertical transmission in nature (64, 65). Rather,

35

NKVs appear to favor non-vectored horizontal transmission in which virus is spread through either direct or indirect contact. Support for these observations is primarily derived from studies involving the cohabitation of experimentally infected and non-infected individuals. In support of direct contact (i.e. grooming and/or fighting), virus was recovered from the lung tissue and salivary glands of infected rodents; indirect contact (e.g., fomites and aerosols), on the other hand, is corroborated by the finding that virus can be shed in the urine as well as saliva

(7, 8, 66). That being considered, it is curious to note that one of the most well- studied NKVs, Modoc virus (MODV), was initially isolated from the mammary glands both experimentally- and naturally-infected deer mice (8, 67) and that at least one investigation has demonstrated the possibility of vertical transmission through the consumption of breast milk by suckling pups (64).

Transmission through contact with animal products such as milk is less convoluted in the case of infection by TBEV, an occurrence which has led to a number of small outbreaks following the consumption of unpasteurized dairy products from domestic ruminants in parts of Europe (68-70). Moreover, WNV transmission has been documented to occur through human in at least one case from a mother that received a transfusion of donated blood following infant delivery (71). These examples not only highlight the ability for some flaviviruses to utilize atypical routes of transmission (e.g., intragastric) but also emphasize the need to execute surveillance and precautionary measures when handling blood products from subclinically-infected individuals. To this end, many studies have now verified the prevalence of WNV-positive samples from

36 otherwise healthy blood donors (72, 73). Finally, there have been increasing reports of intrauterine, intrapartum, or sexual transmission; while most of these cases involve infection with ZIKV, the involvement of reproductive organs or developing fetuses is not unprecedented for flaviviruses and has been additionally observed for DENV, WNV, and JEV as well as some of the NKVs

(74-77). Needless to say, the apparent diversity of transmissive potential for members of the Flavivirus genus represents a complicated, albeit fascinating, paradigm of viral ecology for which there is much still to learn.

In regard to distribution, recent years have witnessed a frightening migration of clinically-important flaviviruses into areas outside of their historically endemic ranges (1). Here, YFV may serve again as the prototype example; however, in this context, the impact of human movement around the globe underscores the potential consequence of inadvertently transporting deadly pathogens. Believed to have originated in Africa, it is now understood that the slave trade was likely responsible for introducing (and perhaps contributing to the dissemination of) YFV to the New World (60). Currently, YFV is considered endemic in the vast majority of South American and Central African countries

(78). Indeed, a threat remains even from the flaviviruses for which a exists. This is highlighted by a lingering outbreak of YFV beginning in 2016, which required the administration of more than 28 million vaccine doses, thereby exhausting the global YFV vaccine supply and prompting the Eliminate Yellow

Fever Epidemics (EYE) initiative by the World Health Organization (WHO) (79).

37

6 Figure 2.5: Global distribution of medically-important flaviviruses as reported by 2017. (A) Depicts overlapping distribution (orange) of TBEV (pink) and JEV (yellow). (B) Depicts widespread distribution of WNV. (C) Depicts overlapping distribution (dark green) of ZIKV (light green) and DENV (blue). Modified from Collins and Metz, 2017 (10).

Similarly, the infection incidence of TBEV, for which the risk of exposure extends throughout the entirety of Russia, Mongolia, , and , continues to grow despite the availability of a licensed vaccine (Figure 2.5) (10).

JEV overlaps these regions of East Asia (excluding Mongolia but including the

Korean peninsula), extending down through (including ) and into Northern . In one estimate, it was proposed that JEV is responsible for lasting neurologic sequelae or death in approximately 34,000 cases annually, primarily affecting children under the age of 15 (80). These numbers are particularly disturbing given the inclusion of a JEV vaccine administered as part of routine pediatric in countries like China (81). Likewise, the global distribution of WNV has expanded to include both Eastern and Western hemispheres, reaching from into and covering nearly all of Europe, Africa, Oceania, and some parts of South America. To exemplify this, the Center for Disease Control and Prevention (CDC) has reported 46,086 cases of clinical WNV disease in the U.S. from 1999 to 2016, a staggering incidence given the absence of this virus in North America prior to the turn of the 21st

38 century (82). Locally-acquired cases of ZIKV infection in Florida and Texas during 2017 further illustrate the continuing spread of flaviviruses into novel environments (3). At present, DENV and ZIKA share similar distributions as

WNV, excluding Europe but including practically all of South America (10). With nearly 40% of the global population at risk, DENV is considered the most prevalent vector-borne virus in the world, infecting more than 400 million individuals annually (4, 5).

Causes for the widespread distribution and expanding range of flaviviruses across the world are multifold and synergistic. As alluded to above, the increasing ease and frequency for which humans now transverse the globe has led to the unfortunate propagation of pathogens into regions previously devoid of these disease-causing agents. This piece of the epidemiologic puzzle does not end with the movement of individuals themselves; however, as with many human activities, may indirectly contribute to the prevalence of flaviviral disease through anthropogenic influence on the environment. For example, multiple investigations have found that international shipments of used automotive tires may harbor the and larvae of viral-infected mosquitoes, thereby facilitating the invasion of flaviviruses into new areas (83, 84). In line with this, it could be argued that the prevalence of water-bearing plastic waste in virtually all environments offers endless breeding opportunities for mosquitoes unable to procreate in otherwise arid locations. This occurrence is compounded by increasing global temperatures and subsequent variations in regional precipitation, which further permit the range expansion of arthropods (Figure 2.6)

39

(85-88). Finally, population growth, cohabitation with domestic livestock, and the unrelenting encroachment (i.e. ) of humans into environments like the equatorial tropics, which harbor the highest diversity of zoonotic pathogens

(i.e. disease agents which come from animals), all ensure that novel viruses will continue to emerge or re-emerge (89, 90).

7 Figure 2.6: Locations in the US of present-day and predicted future suitability for the mosquito. Regions for present-day were created using surveillance reports from 1950 to 2000, while predicted future habitat was inferred based on the Representative Concentration Pathway 8.5 (RCP8.5) model of . Modified from Monaghan et al., 2016 (91).

Modoc Flavivirus

Given the use of Modoc virus (MODV) as the model representative for this study, it seems fitting to include a short review of the history and current knowledge surrounding this particular flavivirus. MODV was first isolated from the mammary gland tissue of a wild deer mouse (Peromyscus maniculatus, mouse

M544) trapped in Hackamore forest, Modoc County, California, in July of 1958

40

(92). Since that time, successive isolates have been recovered exclusively from deer mice in Colorado, , and , as well as Alberta, Canada (6, 93).

Preliminary serologic evidence clarified MODV as a member of the group B arboviruses (i.e. flaviviruses) along with representatives such as YFV and DENV

(67). Unlike the latter, though, MODV cannot replicate in invertebrate cells and therefore belongs to clade II of the no known vector (NKV) cluster within the genus Flavivirus (Table 2.1) (17). At present, the complete genome of MODV has been sequenced and used to confirm its previous taxonomic assignment as well as distinguishing features that characterize this virus apart from other representatives both with and without arthropod vectors (94).

Coincidentally, the initial discovery of MODV occurred by happenstance during field studies in which wild-caught rodents were being examined for

Colorado tick fever virus (CTV) (67). Subsequent to preliminary tests which placed MODV in group B of the arboviruses, early laboratory work aimed to elucidate the ecology of this virus by first (unsuccessfully) attempting to infect mosquito cell lines and later by examining viral tissue tropism to gain clues about the transmission of this virus in nature (67). In brief, tissue tropism appears highly dependent on the exact rodent model utilized. For deer mice, which are presumed to be the natural reservoirs of MODV, virus preferentially replicates to higher and more sustained titers in the lungs, spleen, and salivary/submaxillary glands following experimental challenge through either the intranasal (8) or intramuscular route (unpublished results). As well, infectious virus has been recovered from the lactating mammary glands of both naturally- and

41 experimentally-infected deer mice, although transplacental infection in pregnant mothers does not seem common (8, 67). Finally, while MODV can sometimes be recovered from the kidneys and urine of infected deer mice, mean viral titers and prevalence of successful isolations among replicants are lower than the previously referenced organs (8, 92). Taken together, these findings suggest that the natural circulation of MODV among wild deer mice likely occurs at low levels through close direct contact when performing activities such as grooming, eating, or nursing.

Other investigations have conducted in vitro and in vivo experimental infections with MODV in Syrian golden hamsters and SCID or NMRI strains of M. musculus (65, 92, 95). Similar to studies with deer mice, tropism of MODV toward the salivary glands occurs in both golden hamsters and SCID mice; however, while tropism toward kidney and lung tissues is apparent in golden hamsters but relatively lacking in SCID mice, the opposite is true for tropism toward the spleen in SCID mice as compared to golden hamsters when inoculation dosage and routes are standardized. For hamsters, MODV has been recovered from kidney tissues up to 8 months post-infection with varying durations of accompanying viruria ranging from 5 to 150 days post-inoculation (7,

66). Further, inoculation of SCID and NMRI mice, as well as golden hamsters, reveals a pathogenic neurotropism of MODV that does not seem to occur in deer mice (8, 66). Indeed, a separate study has likewise shown that infection of golden hamsters with MODV leads to CNS involvement reminiscent of the one reported case of human infection as well as disease caused by other neurotropic

42 flaviviruses (e.g., WNV, JEV, and ZIKV) (96). Nevertheless, it is important to note that the route of inoculation appears to influence disease outcome. For SCID mice, which lack the ability to mount an adaptive immune response, inoculation through either intracerebral, intraperitoneal, or intranasal routes leads to by days 6, 11, and 12 days post-infection, respectively, and mortality by days 9, 13, and 14, respectively. In contrast, immunocompetent NMRI mice are paralyzed by days 6 and 12 when inoculated intracerebrally or intranasally, respectively (death occurs by days 9 and 13, respectively) but do not succumb to disease subsequent to intraperitoneal challenge (66). Finally, a separate investigation found that infection of SCID and NMRI mice with cyclophosphamide-induced immunodeficiency exhibited similar encephalitic disease which, when coupled with positive staining for MODV in , suggests that MODV may directly infect and damage neuronal tissue leading to disease in lieu of immunopathology caused by inflammatory infiltrates (95).

More recent studies with MODV have used the virus as a comparative model to better understand the unique features and subtle discrepancies which determine the range of host infectivity, replication efficiency, and disease outcome among various flaviviruses. For example, replacement of the prM-E structural genes of MODV with homologous gene sequences from WNV generates an infectious chimera that can still replicate in vertebrate (Vero) cells but does not gain the ability to replicate in mosquito (C6/36) cells (52). Moreover, it is interesting that the replication kinetics of both the wild-type and chimeric

MODV variants remain similar while wild-type WNV demonstrates a much less

43 linear, more rapid, and higher-titer replicative strategy (Figure 2.7). Conversely, the construction of viral recombinants containing either a DENV-2 or YFV backbone but with prM-E genes from MODV yields an infectious virus that replicates in both vertebrate and mosquito cell lines (97). Taken together, these findings suggest that the inability of MODV to replicate in invertebrate cells is not caused by a failure to infect susceptible cell lines but rather is limited by specific intracellular host factors or post-entry events. Lastly, aside from exploring the restricted capacity for MODV to infect invertebrate cells, alterations to tissue tropism and immune subversion have been identified by additional studies in which the genes encoding pr-M proteins or first 10 amino acids comprising the

NS5 protein, respectively, of MODV have replaced these components on a YFV backbone. In the former study, it was found that the chimeric virus exhibited properties of neuroinvasion that were absent from the wild-type strain of YFV when SCID mice were infected by the intraperitoneal route (98). Recovery of these chimeras from revealed N-linked glycosylation of the E protein, a discovery which corroborates separate research demonstrating variable neuroinvasion by distinct lineages of WNV with and without glycosylation motifs

(45). In the latter study, manipulating the upstream portion of YFV NS5 to match that of MODV’s NS5 sequence abrogates the ability for YFV to antagonize IFN-I signaling by binding to STAT2 transcription factors (99). Indeed, similar investigations have been carried out to examine the non-coding genetic elements in both the 5’ and 3’ UTRs of MODV and other flaviviruses, although the results of these studies are not reviewed here (94, 100).

44

Figure 8 Figure 2.7: Replication kinetics of WNV, wild-type MODV, and chimeric MODV in Vero cells. Vero cells were inoculated at a multiplicity of infection of 0.1 with either wild-type WNV (closed circles), wild-type MODV (squares), or a chimeric virus in which the prM and E structural proteins from a full-length MODV infectious cDNA clone were replaced with the same structural genes from WNV (diamonds). Modified from Saiyasombat et al., 2014 (52).

Typically, NKV flaviviruses present very little threat to general . For MODV, only one case of human disease has been documented in which a young boy developed aseptic after allegedly playing with a wild mouse (101). Regular subclinical infection does not appear uncommon, though, as antibodies against MODV have been recovered from humans in at least one investigation (6). While there is no indication that MODV may become an immediate concern, it is interesting to note that other flaviviruses (e.g., ZIKV) had circulated for many decades before random mutations led to increased virulence and pathogenesis in humans (102). Nevertheless, MODV offers an appealing alternative for flaviviral research given its ease for access (BLS-2), permissibility for laboratory culture and manipulation, and the ability to recapitulate neurologic disease reminiscent of other, more virulent flaviviruses

45

(64). Of all the NKVs characterized at present, MODV certainly has been the most thoroughly examined and therefore offers the most complete background of foundational knowledge to aid in the construction of novel research pursuits.

Pathogen Recognition and Antiviral Signaling

The outcome of any given viral infection is subject to a variety of non- mutually exclusive factors. In most cases, the route of viral inoculation, as well as the titer (i.e., dosage) of viral exposure will dictate the resulting host pathology

(103). However, it has been demonstrated that the exact host species can influence the outcome of clinical disease even when the parameters of viral infection are experimentally standardized (95, 104). In any case, the dichotomy between severe host pathology or disease resistance can essentially be accredited to the precarious balance between the immune system’s ability to restrict direct damage caused by viral replication and the propensity to inflict collateral damage through a process termed immunopathology (105, 106).

Regardless of the above considerations, successful resolution (i.e. clearance) of a viral infection in a vertebrate host begins with pathogen recognition and downstream activation of the innate immune response. Although often described as “non-specific,” surveilling cells of the innate response do, in fact, recognize non-self in the form of conserved, pathogen-associated molecular patterns (PAMPs) indicative of invading . Through the use of the membrane-bound major histocompatibility complex I (MHC-I), all nucleated cells possess the ability to present endogenous protein samples that serve to provide insight into intracellular environment (107). Specialize cells such

46 as CD8+ T lymphocytes (T cells) interact with protein samples displayed on

MHC-I to “taste” whether they constitute the material of foreign or self-origin. If receptors (TCRs) on CD8+ T cells recognize samples displayed on MHC-I as non-self, the newly activated cytotoxic T cell (Tc) may induce in the bound cell through the release of membrane-degrading granzymes and perforins.

Importantly, however, is that this death signal is only delivered when signaling from receptor-antigen complexes are accompanied by co-stimulation from either antigen-presenting cells (APCs) or via the presence of signaling molecules known as cytokines. Cytokines encompass a diverse group of small, secreted proteins that facilitate communication, proliferation, and effector functions between local and systemic cells during an active immune response. In the context of the antiviral response, cytokine representatives of great importance include the various interferons (IFNs), interleukins (ILs), and

(chemotactic cytokines; CCL or CXCLs).

Complementary to MHC-I presentation is the ability for some cells to present extracellularly-acquired protein samples on major histocompatibility complex II (MHC-II). These “professional” APCs include B lymphocytes (B cells) as well as myeloid-derived and dendritic cells (DCs). Through , , or receptor-mediated endocytosis, molecules encountered in the extracellular environment are internalized and trafficked to degradative wherein protein samples can be randomly selected for display on membrane-bound MHC-II. When acting as roaming sentinel cells, macrophages and dendritic cells that encounter antigen in the extracellular

47 environment may migrate back to secondary lymphoid tissues to either activate or “request clearance” from naïve CD4+ T cells. If deemed appropriate, T cell- dependent activation of these innate sentinels leads to the production of cytokines that coordinate a systemic immune response (e.g., global upregulation of MHC-I) (108). Similarly, B cells that encounter exogenous antigen complementary to B cell receptor (BCR) immunoglobulins may internalize the receptor-antigen complex for processing and presentation on MHC-II (109).

Subsequent T cell-dependent activation of B cells presenting in this manner allows for the release of antibodies that neutralize extracellular pathogens or flag them for destruction via phagocytosis.

Despite the perceived benefits of random molecular sampling and presentation on MHC-I and II, these receptors are limited by their need to display antigens of strictly protein origin. Toward this end, additional surveillance is performed by PRRs to detected non-proteinaceous molecular PAMPs. Namely,

PRRs of particular interest include the membrane-bound and endosomal Toll-like receptors (TLRs) as well as the cytosolic RIG-I-like receptors (RLRs). Expressed principally on sentinel cells such as DCs and macrophages, TLRs that appear to play a relevant role in viral sensing include TLRs 3, 7, 8 (humans only), and 9, each of which is understood to emphasize detection of unique motifs

(dsRNA, ssRNA, or unmethylated CpG repeats in DNA, respectively) (110). In contrast, viral sensing by RLRs utilizes ubiquitously expressed intracellular receptors, which appear to be more or less moiety specific. The most well- studied of these include the retinoic-acid-inducible gene I (RIG-I, after which the

48 receptor family is named) and melanoma differentiation-associated gene 5

(MDA5); while RIG-I is believed to specialize in the detection of short, dsRNA and ssRNA with 5’-triphosphate groups, MDA5 is speculated to provide recognition of longer dsRNA. For all of these receptors (i.e. TLRs and RLRs alike), stimulation secondary to PAMP binding results in either receptor dimerization and activation of kinase activity or the exposure of second messenger binding domains that trigger downstream signaling events. Through the use of multiple adaptor proteins (e.g., TRAF3, TBK1, TANK, etc.), phosphorylation cascades ultimately coalesce with the activation of transcription factors such as the -regulatory factors (IRFs) 3 and 7 as well as nuclear factor-κB (NF-κB). Activation of these transcriptional regulators and subsequent translocation into the nucleus then allows for the expression of antiviral signaling molecules such as the type I interferons (IFN-I).

The use of IFNs and their related machinery to combat intracellular pathogens is likely an ancient ancestral trait and is, in fact, conserved even at the phylum level (111, 112). Once expressed within an infected cell, IFNs are secreted as small peptides into the extracellular environment to elicit an antiviral response through both autocrine and paracrine signaling (113). As of 2012, there have been nine mammalian IFN-I subtypes identified: IFN-α (which includes 13 distinct isoforms), in addition to a single known representative of IFN-β, IFN-ε,

IFN-κ, IFN-ω, IFN-δ, IFN-τ, IFN-ν, IFN-ζ (114). Of these, IFN-α and IFN-β are the most well understood in the context of viral infection. In brief, binding of type I

IFNs to the heterodimeric transmembrane receptors IFNAR1 and IFNAR2

49 activates receptor-associated Janus kinases (JAK) Tyk2 and JAK1 (Figure 2.8).

Kinase activity from Tyk2 and JAK1 then auto- and transphosphorylate their associated receptors to expose binding domains for the constitutively-expressed signal transducer and activator of transcription (STAT) proteins. In the case of type I IFN stimulus, phosphorylation of recruited STATs 1 and 2, along with phosphorylated IRF9, generate the multimeric IFN-stimulated gene factor 3

(ISGF3) transcriptional complex. Canonical activation of ISGF3 following the aforementioned transduction pathway is known to regulate the expression of interferon-stimulated genes (ISGs) downstream of IFN-stimulated response element (ISRE) promoters. Some examples of ISGs expressed during the antiviral response include the intrinsic effectors proteins ISG15, 2’5’-OAS, PKR, and Mx; however, over 100 unique ISGs have been characterized and demonstrated to participate in various aspects of cellular defense or interference of the viral replication cycle (the eponym for which IFNs are named) (115).

50

Figure 9 Figure 2.8: Simplified schematic demonstrating the canonical signaling pathways of types I through III interferons (IFN). Binding of type I (primarily IFN-α/β) and type III (IFN-λ) IFNs with their cognate receptors leads to receptor dimerization, auto- and transphosphorylation of TYK2 and JAK1, phosphorylation and heterodimerization of STATs 1 and 2, the formation of the ISGF3 complex with IRF9, and translocation of the complex into the nucleus to induce expression of IFN-stimulated genes (ISGs) downstream of the IFN-stimulated response element (IRES). Activation of type II (IFN-γ) IFN system is similar; however, two additional membrane-bound receptor chains lead to the auto- and transphosphorylation of JAKs 1 and 2 followed by the formation of a STAT1 homodimer. The gamma-interferon activation factor (GAF) complex then translocates to the nuclear whereby antiviral genes (among others) can be expressed downstream of a gamma-interferon activation site (GAS). Modified from Hoffmann et al., 2015 (116).

As alluded to above, additional interferons exist that are functionally and structurally separate from the type I IFNs. These include the type III IFNs (IFN-

λ1, IFN-λ2, and IFN-λ3, previously denoted as IL-29, IL-28A, and IL-28B, 51 respectively) as well as a singular type II IFN (IFN-γ). Interestingly, all three types of IFNs share a conserved helical-bundle architecture and are thus grouped together in class II alpha-helical cytokine family along with representatives such as IL-10 and IL-22 (117). Despite commonalities in their larger secondary structure, however, amino-acid identity and structural homology between specific

IFNs of all types exhibit a range from 20 to 100% (114). This variability is reflected in the functional aspects for which each of the IFNs contributes to the induction of an antiviral response, as well as for additional roles that some IFNs serve. Like the type I IFNS, type III IFN signaling following viral stimulus utilizes

JAK1 and Tyk2 to activate the ISGF3 transcriptional regulator and express ISGs downstream of the ISRE promoter region. Contributing to their functional separation from type I IFNs, however, is the use of the receptors IFNLR1 and

IL10RB to initiate signal transduction; this suggests that type III IFNs may contribute to antiviral defenses in an independent and/or synergistic manner to type I IFNs. While IL10RB is expressed on a variety of diverse mammalian cells,

IFNLR1 expression is minimal or lacking on macrophages and some lymphoid- derived effector cells (114). Type II IFNs likewise utilize distinct membrane-bound receptors (IFNGR1 and 2) to elicit intracellular signaling; unique, however, is the activation of JAKs 1 and 2 to induce the formation of a dual-phosphorylated

STAT1 homodimer. Phospho-STAT1 homodimers translocate to the nucleus upon where they bind to a regulator promoter known as the gamma activation site (GAS). Genes expressed downstream of GAS elements include an assortment of molecules, some of which participate directly in the antiviral

52 response (e.g., ISGs), and some of which appear to coordinate actions between effector cells (see more below). Curiously, whereas a large variety of cell types may produce IFN-β, secretion of various IFN-α isoforms appears to be more selectively regulated by leukocytes such as macrophages and plasmacytoid dendritic cells (pDCs) (113). IFN-γ expression is similarly restricted to immune effectors cells, in this case, TH cells, macrophages, and natural killer cells (NK cells; innate effector cells of lymphoid origin). While it is known that experimental infection with an assortment of viruses leads to the expression of IFN-λ1, 2, or 3 in epithelial and -derived DCs, further research is still required to fully elucidate the origin and influence of these IFNs as they pertain to the antiviral response (118).

As a viral infection progresses beyond the initial stages of pathogen recognition and control, the innate immune response must ultimately attempt to coordinate the adaptive arm of the immune system. Initiation of this highly- specific response is accomplished in part by the presentation of captured antigens displayed on MHC II to naïve CD4+ T cells residing in germinal centers of secondary lymphatic tissues. Along with the aid of upregulated co-stimulatory molecules expressed on the surface of both APCs and T cells (e.g., CD40 and

CD40L, respectively), complete activation of naïve CD4+ T cells results in the production of T helper cell precursors (TH0) that undergo proliferative clonal expansion facilitated by autocrine signaling of IL-2. Once activated, the resident cytokine milieu influences a “polarization” (i.e. differentiation) of TH0 cells toward population phenotypes that favor either cell-mediated or humoral (119).

53

Historically denoted as “type 1 immunity,” cell-mediated immunity is chiefly characterized by the production of cytokines, which stimulate and further boost phagocytic activity. TH cells representing cell-mediated immunity

(referred to as TH1 cells) are driven to differentiation by expression of the transcription factor T-bet and secrete cytokines, which include, but are not limited to, IL-2, IL-12, and IFN-γ. Further, TH1 cells are characterized by a profile of receptors that facilitate the migration of these cells to areas of active inflammation (e.g., CXCR3 and CCR5) (120). In contrast, humoral (i.e. “type 2”) immunity can be identified by the production of cytokines IL-4, -5, and -13 by TH2 cell subtypes expressing high quantities of the transcription factor GATA-3.

Among various other functions, these cytokines lead to increased B cell activation (including clonal expansion and affinity maturation), antibody production by B cell-differentiated plasma cells, and antibody class-switching.

Chemokine receptors expressed on TH2 subtypes include, but are not limited to,

CCR3, -4, and -8.

In the context of viral infection, it is believed that the balance between TH phenotypes of the adaptive response, as well as the timing of the transition between the innate and adaptive responses, will likely play a significant role in the clinical outcome of disease (120). For this reason, it is imperative that all cells participating in the immune response be able to rapidly and appropriately respond given the origin of antigen stimulus so as to harmonize an amalgamation of the immune system as a whole. Further, it is crucial that both effector and non- effector cells possess the ability to mollify and suppress fulminant inflammation

54 prior to the development of immunopathology that may result during an overly- robust or chronically-sustained antiviral response. In the intracellular environment, this is accomplished by the upregulation of protein mediators that act to regulate activated signaling cascades. For example, suppressors of cytokine signaling (SOCS) proteins are encoded downstream of ISGs and thus are expressed subsequent to STAT-mediated transcription; once expressed,

SOCS proteins interfere with the functionality of JAK proteins in a manner that represents a negative feedback loop (112). Beyond this ability for individual cells to quell internal signaling cascades, there exists evidence for a phenotype separate from the previous two, which aids dampen the inflammatory response (121). Finally, cells participating in the immune response receive external instruction by an additional subset of T cells known as regulatory T cells

(TReg). TReg cells are typified by the expression of the forkhead box P3 (FOXP3) transcription factor and the secretion of immunomodulatory cytokines such as IL-

10 and transforming growth factor-β (TGFβ). Although all of the functional mechanisms of TReg cell-induced immunosuppression have yet to be fully elucidated, it is now clear that these cells utilize multiple strategies to direct immune homeostasis both during and immediately following the antiviral response (122). However, as described in the following section, viral interference of these suppressive mechanisms may be one of many strategies employed by viruses to subvert a successful immune response.

55

Flavivirus Immune Subversion and Host Pathology

As was very briefly described above, the vertebrate immune response has evolved to offer a complex and tightly regulated orchestration of cells proficient in detecting, neutralizing, and degrading viral pathogens. In light of this highly- effective protective system, but driven by the obligation to hijack cellular machinery to undergo replication, viruses have likewise evolved a slew of strategic mechanisms to overcome host recognition and removal. This paradigm is no less accurate for clinically relevant representatives of the Flavivirus genus, all of which act to suppress the innate immune response in some form or another

(123).

Upon initial infection and cellular entry, most viral pathogens are recognized by PRRs expressed either in endosomes (TLRs) or in the

(RLRs). For many of the well-studied flaviviruses, it appears that RLRs play a more pivotal role than TLRs, a premise corroborated by the finding that RIG-I- deficient cells infected with JEV fail to express type I IFNs indicative of the early antiviral response (110, 124). It appears that interactions like these are variable, however, as IRF3 activation is not diminished but simply delayed during infection of RIG-I null mouse fibroblasts with the New York strain of WNV (WNV-NY)

(125). Indeed, the inclusion of a 5’ methyl cap on flaviviral RNA presumably prevents immediate recognition by RIG-I sensors and further disguises the viral genome as a potential host transcript. Additionally, though, delayed recognition of flaviviral nucleic acids and replication intermediates by cytosolic PRRs is further preserved by the formation of vesicle packets (VPs) created by virus-

56 induced remodeling of the ER membrane (125). Not only do these membrane invaginations act to shield viral components from detection, but a redistribution of cholesterol within the cell to create VPs may abrogate cellular signaling cascades

(e.g., JAK/STAT) that initiate in lipid microdomains at the cell’s surface (26, 36).

Finally, some flaviviruses (e.g., WNV and DENV) may instigate an upregulation in the expression of MHC-I complexes, and antigen processing, a strategy that inhibits the cytotoxic protectivity of surveilling NK cells (126).

As is the case for many simplified RNA viruses, most flaviviral proteins

(both structural and nonstructural) serve the dual role of aiding in replication as well as antagonizing the antiviral immune response. In particular, the NS5 protein, whose primary function is to act as an RNA-dependent RNA polymerase

(RdRp), has been implicated in a variety of distinct IFN-suppressive roles depending on the exact host cell and flavivirus under investigation. Examples of such events include the proteasomal degradation of STAT2 proteins by DENV and ZIKV NS5, suppression of ISG transcription by NS5 from YFV and

Spondweni (SPOV; a close relative of ZIKV), activation of protein tyrosine phosphatases by JEV NS5, or suppression of IFNAR1 maturation and membrane trafficking by WNV, TBEV, and LGTV NS5 (99, 127-130). NS4B is another nonstructural protein required for viral replication but believed (in the case of

DENV infection) to downregulate RNA interference machinery (RNAi) used by host cells to degrade dsRNA (131). Nonstructural proteins NS2A and NS4A, both necessary for viral replication and packaging, impair IFN signaling by inhibiting

STAT1 phosphorylation and translocation to the nucleus. Even the 3’ UTR of

57 many flaviviruses participates in immune antagonism through the generation of a subgenomic flaviviral RNA (sfRNA) produced during the attempted degradation of viral transcripts by the host exoribonuclease XRN1 (132). As a consequence, not only do nascent copies of the sfRNA then directly inhibit further XRN1 activity, but sfRNAs additionally act to suppress the type I IFN response through mechanisms that remain to be completely understood.

Although it is highly unusual for any virus to encode and secrete viral proteins into the extracellular environment, flaviviral NS1 has been identified as a soluble, hexameric circulating in the blood or accumulating in the dimeric form on the surface of infected cell membranes (133, 134). Within the cell, NS1 performs a vital role in the replication cycle by acting as a for the RdRp activity of NS5 (135). There is evidence to suggest, however, that secreted NS1 circulating in the blood of vertebrate hosts may contribute to an increase in viral uptake for invertebrate vectors such as mosquitoes (136).

Furthermore, it has been shown that soluble NS1 may bind to and inactivate components of the (a branch of the innate immune response that utilizes soluble proteins to, among other functions, opsonize circulating viral pathogens) (133, 137). Indeed, crosstalk between complement components

(e.g., C1q) and secreted NS1 has provided the impetus for further investigation as to whether abnormally high levels of liver-derived acute-phase proteins during infection with DENV may contribute to the vascular leak syndromes associated with instances of hemorrhage disease (138).

58

In regard to the virion itself, alterations in the structure and antigenicity of the E protein caused by low-fidelity transcription of the RdRp may influence successful immune evasion as well as tissue tropism (i.e. preference) and pathogenic virulence (126). As base pair mutations in the genome of progeny virus accumulate, amino acid variations may give way to the production of escape mutants that circumvent neutralizing antibodies in a manner similar to seasonal strains of virus (139). In contrast, as has been demonstrated with WNV, differential glycosylation of the E protein by various host cell and tissue types may significantly alter neuroinvasiveness and neurovirulence even when variations in nucleotide and amino acid sequences remain minimal (140).

Finally, there exists a prominent concern for the ability of cross-reactive antibodies (generated either during previous flaviviral infection or against the exposed epitopes of immature virions) to promote increased viral replication when poorly neutralizing IgG facilities the uptake of virions into innate effector cells expressing Fc receptors (141, 142). While not yet fully understood, infection of , DCs, and macrophages in this manner could lead to an overproduction of proinflammatory cytokines in a process called antibody- dependent enhancement (ADE; Figure 2.3). Likewise, dysregulation during infection of these important immune cells may skew the timely secretion of immunosuppressive cytokines like IL-10, which help to control collateral damage inflicted by an unchecked host antiviral response (126, 143).

In any case, there is sufficient justification to postulate that the exact mechanisms of flaviviral immune evasion and subsequent virulence culminate as

59 the result of many subtle but significant interactions between polymorphic phenotypes in both the host as well as the infecting virus (123, 144, 145).

Fortunately, infection with most flaviviral isolates results in relatively mild, flu-like that spontaneously self-resolve in one to two weeks. In the event that immune subversion or antagonism is sufficient to induce severe clinical disease, however, flaviviral pathologies can be broadly divided into those which cause hemorrhagic fevers, hepatic or neurologic disease, and as of late, defects in fetal development. For hemorrhagic fevers, of which DENV is the most prominent and well-studied etiologic flaviviral agent, disease is largely immunopathogenic and results primarily from a loss in vascular integrity secondary to excessive activation of the complement system, cross-reactive TH cells, and proinflammatory DCs and macrophages (the latter of which act as the primary targets of DENV replication) (126, 146). Although manifestations of disease caused by DENV were previously categorized as either

(DF), dengue hemorrhagic fever (DHF; grades I-IV), or dengue

(DSS; specific to grades III and IV of DHF), classification was revised by the

WHO in 2009 due to the clinical and physiologic overlap between categories; as such, current diagnoses include only “dengue with or without warning signs”

(e.g., , persistent , mucosal , lethargy, mild and increased concurrent with ) or

“severe dengue” (e.g., plasma leakage leading to hypovolemic shock, fluid accumulation with respiratory distress and severe organ involvement) (147).

60

In a similar fashion, disease associated with representatives such as

WNV, JEV and TBEV appears to be induced immunopathologically; the difference in clinical presentation, therefore, results from a predilection of each virus toward neuronal cells in the and spinal cord (148, 149). As viral replication in the central (CNS) ensues, expression of proinflammatory cytokines by neurons and activated glial cells encourages the recruitment of cytotoxic effectors such as neutrophils and TH1 cells, both of which secrete mediators (e.g., TNFα) that lead to collateral damage of neurons and supportive cells in the local environment (148, 150). Ongoing studies with ZIKV have likewise indicated neuronal and glial tropism during acute infection; however, select isolates of ZIKV differently infect a broad range of tissues, including stem and progenitor cells in the CNS and retina, each of which may contribute to significant birth defects when mothers are exposed during early (151, 152). Further, the detection of infectious virus shed in the of adult males for up to three months following exposure has established

ZIKV as the only flavivirus implicated in direct human-to-human transmission

(153).

While disease caused by YFV may include hemorrhagic shock syndromes similar to those induced by DENV, YFV is uniquely but significantly hepatotropic

(154). As such, liver dysfunction over the course of infection leads to a jaundiced appearance in patients, a clinical feature from which the “yellow fever” virus derives its name. Further, it is curious to note the relative lack of inflammatory infiltrates (e.g., monocytes) in hepatic tissues analyzed from infected individuals

61

(155). Although this finding provides evidence to suggest that the observed damage to is, therefore, a consequence of direct, viral cytopathic effect (CPE), other reports highlight the apoptotic role of secreted TGF-β, a mechanism which may indicate that clinical disease caused by YFV is likewise primarily immunopathologic in origin (156).

Vaccination and Treatment Against Flaviviral Disease

Unlike most bacterial infections, disease caused by the majority of viral pathogens cannot be treated with the use of post-exposure therapeutics. This reality is no less applicable for the flaviviruses given that there are currently no licensed antiviral drugs that may be called upon to alleviate severe clinical illness. It is, therefore, essential that preventative practices shape the foundation upon which humans (and their animal companions) attempt to avoid morbidity and premature death caused by these viruses. Indeed, efforts to control vector populations and exposure have arguably provided the most effective avenue for thwarting infection; however, many scenarios exist in which these practices are infeasible or simply insufficient. These shortcomings are further compounded by the range expansion of invertebrate vectors into historically non-endemic

(see Flaviviral Ecology and Distribution). Coupled with the ever-increasing prevalence and frequency of global travel, an upraise in the incidence of flaviviral infection and disease has led to a subsequent urgency to continue pursuing vaccine development and implementation.

Astoundingly, the first successful arboviral vaccine occurred during an era when infectious disease research as a whole, was still very much in its infancy.

62

Through continuing the promising work of Louis Pasteur, a French isolate of YFV was passaged over 200 times in murine hosts, producing a distinct viral phenotype that exhibited substantial attenuation compared to the wild-type derivative (157). Inoculation with this live-attenuated virus provided sufficient protection to significantly decrease the global incidence of YF disease throughout the 1940s and 50s, prompting Collins et al. to label the vaccine as “the original breakthrough” against flaviviral disease (10). While this particular vaccine was eventually discontinued due to an unforeseen neurotropism acquired during serial passaging (and subsequent vaccine-induced encephalitis in children), it was immediately superseded by a comparable live-attenuated strain originally isolated from an infected patient (Mr. Asibi) in Ghana. Referred to now as the

17D vaccine strain, the Asibi vaccine is used worldwide to produce lifelong immunity by inducing neutralizing antibodies in 99% of adults and 90% of children (158). Further, aside from representing one of the oldest vaccines administered to date, the genomic architecture of the 17D strain has been utilized to create a “flexible backbone” in which structural genes from other flaviviruses can be exchanged with those of YFV to produce a recombinant virus capable of safely exposing patients to antigens from representatives that resist immunity elicited by traditional attenuation.

In addition to YFV, a successful live-attenuated vaccine has been developed for JEV (strain SA14-14-2). Unlike 17D, however, which is selectively administered pro re nata (e.g., military or research personnel or during times of endemic outbreaks), JEV SA14-14-2 is included in a series of routine pediatric

63 immunizations for children living in most regions of China (81). Widespread administration of this immunogenic vaccine has been paramount in the prevention of flaviviral encephalopathies induced not only by JEV but potentially by other neurotropic flaviviruses in light of the few studies which have observed protective immunity through cross-reactive neutralizing antibodies (159, 160). At present, the armamentarium against JEV includes a total of three vaccine formulations: the live-attenuated strain referenced above, an inactivated vaccine derived from one of four viral strains, or a (live-attenuated) 17D chimera in which the prM and E genes of YFV have been replaced by those of JEV (81).

Unlike YFV and JEV, flaviviruses such as WNV, ZIKV, and until very recently, DENV have all eluded traditional approaches to effective vaccination in human populations. For WNV, several vaccine candidates have utilized novel molecular technologies (e.g., DNA plasmids) to provide formulations that appear protective in rodent and equine models (161). As well, ongoing work with a 17D candidate (ChimeraVax) has elicited high titers of WNV-specific antibodies and is currently under investigation in Phase II clinical trials (162). Despite the promising outlook, factors such as long-term vaccine efficacy, stability, patient tolerability (i.e. allergenicity), and cost of production all contribute to the paucity of a WNV vaccine currently fit for public distribution. Similar obstacles have arisen in the attempt to develop a vaccine against DENV; however, researchers must also contend with the fact that DENV comprises four antigenically distinct that cocirculate in endemic regions (10). While some evidence suggests that infection with one may confer transient protective

64 immunity against symptomatic infection by diverse serotypes, contrasting data has indicated that cross-reactive, non-neutralizing antibodies may actually enhance disease during secondary infections (i.e. ADE) (163). The latter recognition has prompted the desire to construct a formulation that offers protection against all serotypes, given the concern that vaccination against only one may predispose individuals to develop more severe disease upon subsequent natural infection. This objective has proven difficult as evidenced by the late Phase III clinical trials of Dengvaxia (a 17D derivative) in 2011 when recipients demonstrated variable protectivity against each serotype and an alleged increased of vaccine-induced dengue fever in certain Asian demographics (164, 165). Fortunately, successive iterations of Dengvaxia have remedied any apparent adverse reactions and, as of 2019, was approved by the

FDA as the first licensed vaccine for the prevention of DENV in endemic regions

(166).

While there are no vaccines that have been approved for ZIKV at the time of writing this review, there exist multiple diverse and promising candidates currently in various phases of development and clinical testing. Surprisingly, both chemically-inactivated and live-attenuated approaches have demonstrated protectivity against lethal challenge in rodent models as well as non-human primates (167-169). Amazingly, there is even evidence to suggest that adenoviral vectors expressing ZIKV E proteins fused to fibritin (a bacteriophage structural protein) may protect mouse pups against viral challenge following birth (170).

65

Induced immunity to this extend marks enormous success given the ability of

ZIKV to transverse placental barriers and cause disease in developing fetuses.

In terms of protection against pan-flaviviral threats, there is reason to believe that increasingly novel vaccination methodologies will continue to provide support in the battle against disease. For example, the use of lipid nanoparticles to deliver mRNA transcripts encoding the prM and E proteins have induced neutralizing antibodies for ZIKV in addition to multiple tick-borne flaviviruses such as LGTV and (POWV) (171, 172). Broadening the scope of attack even further, some researchers have investigated the potential for reducing the severity of clinical disease by vaccinating against mosquito salivary proteins that appear to contribute to host immune suppression during flaviviral infection (173). Indeed, Bidet et al. have further postulated in recent studies that emulating “immune signatures” of natural exposure (through the use of targeted adjuvants) may provide a lucrative avenue in an attempt to strengthen the immunogenic response elicited by subunit vaccines (174). Lastly, there is still hope for anti-flaviviral therapeutics, which may provide an avenue for post- exposure intervention. To this end, ongoing research is exploring the potential for producing monoclonal antibodies (mAb) in plants, which can be glycoengineered to circumvent the induction of ADE when poorly neutralizing antibodies from cross-reactive flaviviruses interact (175). Likewise, extensive libraries of previously FDA-approved small molecules are continuously tested for repurposing as competitive or allosteric inhibitors of viral enzymes (176-179).

66

As a closing note, it is important to consider that the successful development of vaccines and therapeutics does not in itself guarantee a widespread decrease in . Rather, interventions must also be distributed and administered to the populations that need them most, an unfortunate conundrum given that the majority of highly at-risk individuals reside in impoverished areas with poor infrastructure (180). This issue is particularly problematic for live-attenuated vaccines, which typically demand higher production costs and must abide by cold-chain transportation. Indeed, the paradigm of ineffectual vaccine distribution was highlighted in 2003 by an estimate suggesting that nearly 3 million children succumb to preventable annually as a consequence of poor vaccine accessibility or implementation (181). Although this realization has sparked the construction of organizations such as the Global Alliance for Vaccines and (GAVI), factors such as population growth (in both affluent and impoverished demographics), frequent international travel, human encroachment in diverse habitats, and changes to global climate have all contributed to the continued need for advancements in the ongoing battle against viral disease.

The Use of Animal Models in Scientific Research

The use of non-human organisms to aid scientific research can be traced back as early as the 6th century BCE when canines were used to demonstrate that the brain is responsible for processing sensory input and coordinating complex behaviors (182). From this time forward, researchers have utilized a broad array of plant, bacterial, fungal, and animal models to better understand

67 the biotic world from which we are all derived. Eventually, “model organisms” were established that provided researchers with representatives that were well- studied, cost-effective to acquire and maintain, and easy to handle or manipulate

(183). Examples of these organisms include the small, flowering thale cress

(Arabidopsis thaliana), the bacterium Escherichia coli, the unicellular yeast

Saccharomyces cerevisiae, eukaryotic invertebrates such as the nematode

(Caenorhabditis elegans) and the fruit (Drosophila melanogaster), and the common house mouse, Mus musculus. Decades of research with these organisms have cast the foundation for our current knowledge of core disciplines such as genetics, developmental biology, anatomy and physiology, aging, metabolism, and infectious disease, among many others.

In particular, rodent models have served a pivotal role in pioneering the field for biomedical research (184). Although recognizably divergent from humans, homologous genes and conserved signaling pathways can be studied in rodents to better understand the functions and failures of these systems in humans. Likewise, research with these models in the context of infectious disease has been instrumental in bolstering the scientific community’s understanding of disease pathology, therapeutic targeting and treatment, and vaccine development and testing. Evidence for the importance of rodent models becomes increasingly apparent when considering the investigation of zoonoses, diseases that are acquired either directly or indirectly from non-human hosts

(185). Accounting for over 60% of emerging infectious diseases (EIDs) worldwide, zoonoses can be found on nearly every continent and infect hundreds

68 of millions of individuals annually (186, 187). Unfortunately, despite advances in scientific research and medical practice, there is little indication to believe that infection incidence from zoonoses will decline given the environmental changes and anthropogenic practices that encourage peridomestic arrangement with the animals that carry infectious disease (188). Due to this prevalent and on-going threat to global public health, rodent models are therefore employed to study zoonotic disease either directly, as is the case for rodent-borne diseases (e.g., hantaviruses), or indirectly, as when rodent models are used to recapitulate human-like disease (e.g., guinea pigs infected with Lujo ) (189, 190).

The potential for genetic manipulation and production for transgenic variants have also promoted Mus musculus (or M. domesticus) and, to a less extent, the brown (Rattus rattus or R. norvegicus) as the primary rodent models of choice (191, 192) for research in immunology and related fields. Over time, as the use of these models grew, litters were selectively inbred to produce colonies that expressed less genetic variation and, in some cases, exhibited peculiar phenotypes ideal for testing experimental hypotheses (e.g., severe combined immunodeficiency (SCID) mice used to examine the adaptive immune response). Currently, nearly 50,000 unique strains of M. musculus and around

2,900 strains of R. norvegicus have been identified and registered with the

Mouse and Rat Genome Databases, respectively (193, 194). Distinct phenotypes from various strains have provided models for research on, but not limited to, behavior, genetic heredity, aberrant metabolic function and disorders (e.g., nutrient acquisition and obesity), aging, , and neurologic disease (182). It

69 is important to note, however, that while selective inbreeding may reduce genetic variation, thereby allowing for high experimental reproducibility along with a reduction in background error, the caveat may be that results from these studies cannot be functionally extrapolated to genetically heterogeneous populations

(e.g., animals in nature) (195, 196). For this reason, among others, some researchers have taken up the use of non-traditional rodent models as a strategy to approach familiar questions from a new point of view.

Deer Mice

Despite their colloquial nomenclature, deer mice (genus Peromyscus) represent a phylogenetically distinct group of rodents situated within the family

Cricetidae (subfamily ) along with voles and Old World hamsters

(197). It is likely that these small (8 to 10 cm, 20-30 g) rodents were initially denoted as “mice” given their similar appearance to members of the Muridae family (198). In reality, however, peromyscines diverged from Mus and other murid representatives nearly 25 million years ago as determined through the comparative sequencing of four highly-conserved nuclear genes totaling approximately 6,400 base pairs (Figure 2.9) (197). This stands in contrast to the divergence between Peromyscus and the cricetids, which occurred only 16-19 million years ago when using the same phylogenetic approach (197). Currently,

56 biologically-distinct species have been recognized, including the prototypic deer mouse (P. maniculatus) as well as the equally renown white-footed mouse

(P. leucopus) (199). Representatives of this genus can be found in virtually every habitat of North America, extending from Canada and Alaska to the Southern

70 border of Panama, thereby making them the most abundant in the region (199-201).

Figure 10 Figure 2.9: Simplified phylogeny representing divergence of the Muridae and Cricetidae rodent families approximately 25 million years ago (199). Reconstruction of this simplified phylogram from data originally published by Steppan et al. (197).

Use of deer mice in the laboratory setting is not entirely novel, as these rodents have been implemented for research on animal ecology, behavior, reproduction, aging, evolution, and physiological adaptation as far back as the early 1900s (199, 202-204). Despite this, commercially-available reagents (e.g., antibodies) and public-access databases skewed toward a historical use of Mus and Rattus spp. has led to a paucity of data and resources for molecular and biomedical research with Peromyscus (205). On the other hand, recent decades have witnessed a renewed interest in the use of deer mice as alternative models for research in a variety of both homeostatic and infectious diseases (188, 204,

205). In part, it is intriguing that some species may exceed the average lifespan of M. musculus by nearly four times despite having a roughly equivalent body

71 mass (205). Nevertheless, colonies of Peromyscus must be maintained and utilized in relatively outbred stocks as compared to many inbred strains of Mus and Rattus (204). Granted, the purpose of implementing inbred strains for laboratory investigations posits that a decrease in genetic diversity should mitigate extraneous error and increase the reproducibility of experimental results; however, the caveat to this approach is that experimental findings may not be truly reflective of the biological outcomes in a natural setting where genetic diversity provides the foundation for adaptive evolution (205). Likewise, selective inbreeding may lead to the unintentional loss of important gene variants, which are polymorphic in nature but under no selective pressure to be maintained under laboratory husbandry (204-206). Thus, it is paramount to select an animal model that sufficiently reflects the inherent genetic diversity among species populations while still minimizing background error due to differences in individual genomes. This is especially true of investigations which hope to extrapolate data from research with animal models for use in human studies, for which deer mice may offer a novel approach.

In the context of infectious disease, however, perhaps the most enticing aspect for the use of peromyscines is their ability to host a variety of virulent pathogens without any apparent disease or decrease in fitness (188). In particular, the 1993 endemic outbreak of Sin Nombre hantavirus (SNV) in the

Four Corners region has sparked curiosity for the mechanisms which allow deer mice to serve as reservoirs for agents that otherwise lead to debilitating illness in humans (see more on reservoirs below) (207). Apart from SNV, deer mice have

72 been implicated as hosts for additional viruses such as POWV and

(DTV), MODV, and Sal Vieja virus (SVV) (64, 208, 209). Moreover, a significant number of bacterial pathogens have been isolated from deer mice such as

Anaplasma phagocytophilum, Ehrlichia muris, Coxiella burnetii, Bartonella spp., and various species of Borrelia, including B. burgdorferi, the causative agent of

Lyme disease (210-214). Although debated, Peromyscus spp. have even been suggested as important reservoirs for the enzootic maintenance of plague

(caused by the bacteria ) within the (215, 216).

Finally, there are a number of medically-relevant protozoan parasites that have been isolated from deer mice, including , Schistosomatium douthitti, and Trypanosoma spp., the causative agent of African sleeping sickness (217-219). Indeed, the ability for deer mice to host such a broad array of pathogens without succumbing to disease begs the question of what molecular underpinnings can be discovered by studying these immunologically tolerant rodents.

Syrian Golden Hamsters

Like the deer mouse, Syrian golden hamsters (Mesocricetus auratus) are taxonomically-nested within the Cricetidae family of rodents having diverged from murids approximately 25 million years ago (Figure 2.9) (197). Unlike the former, golden hamsters have a lifespan of only 3 to 4 years in captivity and grow comparatively larger than peromyscines, reaching lengths up to 18 cm with an approximate range in body mass from 110 to 120 g. Moreover, wild golden hamsters are far less ubiquitous than deer mice and have, in fact, been

73 registered as vulnerable by the International Union for Conservation of Nature

(IUCN) due to habitat loss within their limited geographic range northern Syria and southern Turkey (220). While the exact ecology and distribution of M. auratus have not been entirely elucidated, it is understood that nearly all of the individuals presently kept as pets or used in laboratory research are descended from a single litter of only five siblings bred in 1930 (221). Following field research conducted near Aleppo in 1997 through 1999, a total of 19 additional individuals were recovered and used to supplement breeding stocks housed in

Germany before dissemination across the globe for various research pursuits

(222).

Research with golden hamsters shares a similar history as deer mice in that their use in the laboratory setting is not entirely novel but has been restricted by a lack of commercially-available reagents and available data in public repositories. Nevertheless, early studies with these rodents utilized unique aspects of their physiology and behavior to explore aspects of rodent domestication, circadian rhythm, hibernation, and cancer. (223-227). Intriguingly, many studies have used golden hamsters as models for the development of dental caries, for which they appear to be naturally susceptible (228-230). Lastly, the extremely short gestation period of golden hamsters (roughly 16 days) has provided the impetus to further investigate the reproductive behavior and physiology of these rodents with questions ranging from how food intake is regulated during estrus to how social stress impacts a developing corpus luteum

(231-233).

74

Indeed, golden hamsters have become yet another increasingly attractive rodent model for studying infectious disease. In contrast to deer mice, however, which offer the appeal of a model system that can tolerate infection by a variety of diverse pathogens, golden hamsters are utilized given their ability to recapitulate human-like illness (234). Infection models such as these ideally mimic not only the clinical signs of disease but are also reminiscent of the molecular (i.e. immune) response to pathogen exposure, as witnessed in humans that become infected. Further, while these outcomes can sometimes be achieved in Mus models with the aid of immunosuppressants or genetic manipulation (e.g., ZIKV infection of STAT2 knockout mice), it is preferable to investigate the immune response in the absence of these modulations to better understand the interactions between pathogen and host under a more natural context (235). In light of this, viral infection models using golden hamsters include, but are not limited to, viral hemorrhagic fevers induced by viruses, , and viscerotropic flaviviruses, cardiopulmonary syndromes such as those elicited by hantaviruses, and encephalopathies caused by neurotropic flaviviruses (96, 236-240). In regard to bacteria, experimental infection with

Leptospira interrogans offers an informative model to better understand the cytokine profile and subsequent immune response following the development of (a concern for both humans as well as wild and domestic animals)

(241). Finally, golden hamsters as models for infectious disease are perhaps most well-renowned for their application in the study of parasitic eukaryotes such as Leishmania (trypanosomes) and Taenia (tapeworms), both of which have

75 eluded the use of more traditional rodent models (242-244). Due to their susceptibility toward infection and disease, Syrian golden hamsters may, therefore, offer an innovative approach for comparative studies aiming to elucidate more clearly aspects of the immune response that result in host pathologies.

76

CHAPTER III

HOW TO HOST UNINVITED GUESTS: CLUES ABOUT THE ANTIVIRAL RESPONSE IN A RESERVOIR MODEL OF ACUTE FLAVIVIRAL INFECTION

Author: Tyler Sherman

Contributions: Developed and implemented the study design. Generated and analyzed data. Wrote first draft of the manuscript.

Co-Author: Dr. Ann C. Hawkinson

Contributions: Helped conceive the study topic. Provided guidance on study design. Provided feedback on data interpretation and manuscript preparation.

Abstract

To better understand the clinical outcomes of viral disease, it is necessary to elucidate the interactions that take place between a virus and its host. In particular, important clues may be gleaned from investigating the acute phase of infection to better understand the early events which shape the subsequent immune response and influence the long-term outcome of exposure (i.e. viral clearance, persistence, or host death). In this study, we utilize the deer mouse

(Peromyscus maniculatus) to explore aspects of the innate immune response during infection with a rodent-borne flavivirus. As Modoc virus (MODV) has only ever been isolated from peromyscines in the wild, it is presumed that these rodents act as the primary reservoir hosts. Deer mice in this study were intraperitoneally challenged with MODV M544 and serially euthanized in order to collect tissue samples from the kidneys, spleen, liver, and brain over a 10-day

77 experimental period. Cardiac exsanguinations and bladder contents were additionally collected at the time of necropsy. Quantifiable levels of MODV RNA could be detected by RT-qPCR from all tissues and fluids at 24 hours post- infection and persisted throughout the entire duration of the experiment.

Infectious virus could be recovered from multiple individuals from all tissues at various timepoints but not from the sera or urine. Fold change expression values of the cytokine genes IFN-γ, TNFα, IL-6, IL-10, and TGFβ, as well as the antiviral effector protein Oas1b, were quantified using the ΔΔCt method. Using a

MANOVA, it was discovered that both tissue type and collection timepoint had a significant impact on cytokine gene expression (p = 0.01), suggesting that the sampled immune genes are indeed expressed in a tissue- and time-specific manner in deer mice infected with MODV. Corrected models of between-subject tests depicted that IFN-γ, TNFα, and TGFβ were all significantly differently expressed across tissues over time; however, Tukey HSD pos hoc analyses revealed that IFN-γ alone was differentially expressed in the brain as compared to all other tissues examined (p ≤ 0.01). Finally, a two-factor ANOVA demonstrated a significant difference in the expression levels of Oas1b over time but not between tissue types. Taken together, these findings suggest that deer mice respond to MODV infection with a customized profile of cytokine genes that are both tissue- and time-dependent. Additionally, as indicated by the expression of the type I interferon (IFN)-stimulated gene (ISG) Oas1b, MODV does not appear to antagonize the type I IFN response in a manner similar to other flaviviruses.

78

Introduction

Modoc virus (MODV) is a rodent-borne flavivirus that has been isolated exclusively from wild deer mice in multiple states across the Western U.S. and

Canada (6, 67, 245). Morphologically, an enveloped virion encapsulates a positive-sense, ssRNA genome of approximately 10.6 kb which is translated into a single polyprotein that is co- and post-translationally cleaved to form three structural (envelope, E; capsid, C; pre-membrane/membrane, prM) and seven non-structural (NS1, NS2A, NS2B, NS3, NS4A, NS4B, NS5) proteins. In addition, the 5’ end of the viral genome contains a methyl cap that mimics host transcripts while the 3’ end encompasses various critical non-coding elements (e.g., subgenomic flaviviral RNA, sfRNA). As such, MODV belongs to the genus

Flavivirus (family Flaviviridae), along with other important representatives such as West Nile virus (WNV), dengue virus (DENV), tick-borne encephalitis virus

(TBEV), and Zika virus (ZIKV), all of which are carried and transmitted through arthropod vectors such as ticks and mosquitoes. Unlike these latter, however,

MODV cannot replicate within invertebrate cells and thus belongs to the intrageneric cluster of flaviviruses with no known vector (NKV) (17).

Although the incident rate and disease severity of human infection with

NKV flaviviruses are far less concerning compared to their arthropod-borne counterparts, it is not unprecedented. In at least two cases, rodent-borne NKVs

(MODV and Apoi, APOIV) have been indicated as the etiologic cause of illness with neurologic involvement (i.e. encephalitis and meningitis, respectively) (64,

101). Similarly, bat-borne NKVs such as Dakar bat virus (DBV) and Rio Bravo

79 virus (RBV) have caused isolated cases of disease in researchers working with experimentally-infected individuals (64, 75). Aside from MODV and RBV, other

NKVs that have been isolated from wild-caught animals in North America include the rodent-borne San Perlita (SPV), Sal Vieja (SVV), and Cowbone Ridge viruses (CRV) as well as the bat-borne Montana myotis leukoencephalitis virus

(17, 64, 94). While these other NKVs have not yet been documented to cause prominent clinical disease, antibodies against MODV and other MODV-like viruses have been recovered from humans and small mammals in at least one investigation suggesting that exposure to NKVs is not uncommon and that replication in novel vertebrate hosts is possible (6).

To date, relatively few studies have utilized deer mice (Peromyscus maniculatus) as rodent models for molecular or biomedical research (204).

Despite the extreme ubiquity of these rodents, most commercially available reagents (e.g., antibodies) have been optimized for work with the common house mouse (Mus musculus) and show poor cross-reactivity with peromyscines that have undergone considerable evolutionary divergence from Mus (201). As such, only a small subset of publications have described aspects of immune gene regulation in deer mice during pathogenic infection (104, 190, 246, 247).

Fortunately, techniques such as reverse-transcription quantitative PCR (RT- qPCR) have now become standard practice and provide the opportunity to examine gene regulation and mechanisms of the subsequent immune response in nearly any model of interest. In particular, deer mice pose an intriguing avenue for infectious disease research given their ability to act as reservoir hosts for a

80 variety of pathogens such as Powassan (POWV), deer tick (DTV), MODV, and

SVV flaviviruses, hantaviruses such as Sin Nombre (SNV) and New York-1 virus

(NYV), as well as a handful of bacteria and protozoa, most of which have been indicated to cause moderate to severe disease in humans (64, 209-212, 217,

248-251). In cases such as these, examining the immune response of a natural host may provide insight for the balanced immune response which prevents pathology from both the infectious agent as well as the immune system itself

(185).

Currently, publicly available vaccines against medically relevant flaviviruses are limited in both variety and accessibility. Likewise, there are no

FDA-approved antiviral drugs that can be enlisted to treat acutely infected individuals. Given these deficiencies, coupled with the ever-increasing spread of flaviviral distributions worldwide, the need for continued research to better understand the mechanisms of flaviviral pathology has never been more vital.

Thus, despite the relatively low threat of clinical disease from NKVs, we feel confident that important clues can nevertheless be obtained by utilizing MODV as a unique model for further investigating universal aspects of flaviviral replication kinetics, tissue tropism, and subsequent immune gene regulation during acute infection. As the putative reservoirs of MODV, it is hypothesized that deer mice will demonstrate no signs of clinical pathology throughout the experimental duration. Likewise, despite the absence of any overt pathologies, infectious virus and viral RNA should be detectable in various tissues at multiple timepoints; however, quantities of detectable virus will be sustained at relatively low levels.

81

Finally, it is hypothesized that relative expression values of the immune genes

IFN-γ, TNFα, IL-6, IL-10, and TGFβ will vary across tissues and over time, thereby eliciting a tissue- and time-specific protective effect against both viral burdens as well as immunopathologies that may accompany a sustained proinflammatory response.

Materials and Methods

Ethics Statement and Animal Husbandry

All of the methods conducted in this study were approved by the

University of Northern Colorado (UNCO) Institutional Animal Care and Use

Committee (IACUC; Protocol No. 1603C-AH-DmH-19) in accordance with the

Animal Welfare Act. Live deer mice (Peromyscus maniculatus) of both genders

(9 male and 9 female), aged 10-16 weeks old, were supplied from an established breeding colony at UNCO.

Modoc Virus

Modoc virus (MODV; strain M544) obtained from ATCC (Manassas, VA,

USA) was passaged once in Vero E6 cells (ATCC) for seven days prior to titration using the Reed-Muench 50% endpoint dilution assay.

Experimental Infections

Deer mice were inoculated intraperitoneally with 0.3 mL of MODV at an

4.2 infectious titer of approximately 10 TCID50/100mL (virus diluted in sterile 1x

DPBS; pH = 7.3). Likewise, negative control animals (6 of each gender) were inoculated intraperitoneally with 0.3 mL of 1x DPBS (pH = 7.3). Following the administration of virus or saline, all animals were housed individually and

82 monitored daily for any signs of distress or morbidity (e.g., tachypnea, altered gait, abnormal behavior, etc.). Experimentally infected animals were serially euthanized at the following predetermined timepoints: 1, 2, 4, 6, 8, and 10 days post-infection. Negative control animals were euthanized at 1, 6, and 10 days post-infection. For all animals, euthanasia was performed by means of cervical dislocation and cardiac exsanguination following anesthesia with isoflurane vapors.

Sample Collection and Preparation for the Examination of Replication Kinetics

Necropsies were performed immediately following euthanasia and included the collection of urine (extracted directly from the bladder using a tuberculin syringe), blood sera (collected at the time of cardiac exsanguination and separated by incubating at room temperature for 10 min, chilling on ice for

60 min, and then centrifuging at 4000 x g for 10 min), kidneys, liver, spleen, and brain. All collected tissues were frozen immediately in liquid nitrogen baths and stored at -80°C until further use. To prepare kidney, liver, spleen, and brain tissues for RNA extraction and virus isolation, tissue samples of approximately

30 mg were first homogenized using a Mini-Beadbeater tissue disrupter

(Biospec, Bartlesville, OK, USA) with zirconia/silica beads in 600 uL of sterile 1x

PBS. Homogenates were then centrifuged through a QiaShredder column

(Qiagen, Germantown, MD, USA) at 14,000 x g for 8 min at 4°C to further mechanically lyse cells and pellet cellular debris. Following centrifugation, 150 uL

83 of supernatant were aseptically removed and used for viral RNA extractiµon while the remaining 450 uL were used for virus isolation. 10 uL of urine was first removed from all samples for subsequent virus isolation, and remaining volumes comprised of less than 140 uL were brought up to this volume with 1x PBS for use in RNA extraction. In contrast, 140 µL of blood sera were first separated from collected samples for use in RNA extraction; all remaining blood sera were then used to perform viral isolation.

Extraction of Viral RNA and Synthesis of Viral cDNA

RNA was extracted from tissue homogenates, urine, and blood sera samples using a QIAmp Viral RNA Mini kit (Qiagen, Germantown, MD, USA) according to the manufacturer’s protocol. To reverse-transcribe viral RNA, 2 µL of sample extract were first mixed with 5.2 µL of nuclease-free water and 2.8 µL of MODV-specific forward and reverse primers (Eurofins, Louisville, KY, USA) at a final concentration of 280 nM (66). RNA and primers were incubated at 94°C for 1 min and then placed in an ice bath for 3 min before the addition of a mix containing 0.5 µL of an Episcript RT and 2 µL of 10x RT buffer (Lucigen,

Radnor, PA, USA), 2 µL of DTT, 1 µL of dNTPs at 10 mM, 0.5 µL of a recombinant RNase inhibitor (New England BioLabs, Ipswich, MA, USA), and 4

µL of nuclease-free water for a total final volume of 20 µL per reaction. First- strand cDNA synthesis was carried out by incubating the entire reaction mixture at 45°C for 1.5 hours, followed by 5 min at 85°C to halt enzymatic activity.

Samples were selected at random along with positive controls (obtained from

84 stock suspensions of MODV) to assess for successful RNA extraction and cDNA synthesis by conducting endpoint PCR using the manufacturer’s protocol for an

AccuStart II PCR SuperMix kit (Quantabio, Beverly, MA, USA), the same MODV- specific primers at a final concentration of 300 nM, and the following cycling parameters: 5 min at 95°C followed by 40 cycles of 15 s at 95°C and 1 min at

60°C. Amplification of the expected gene product was observed using agarose gel electrophoresis.

Quantification of Viral Loads Using RT-qPCR

Absolute quantities of viral cDNA collected from select tissues were assessed using RT-qPCR and MODV-specific primers that recognize a region contained within the NS5 gene of the viral genome (66). To begin, the amplification efficiencies were assessed by performing qPCR on 1:10 dilutions of reverse-transcribed viral RNA extracted from stock MODV suspensions.

Following amplification efficiency optimization, qPCR analysis was conducted by combining 2 µL of viral cDNA template with 10 µL of PerfeCTa SYBR FastMix

(Quantabio, Beverly, MA, USA), 2 µL of forward and reverse primers each at 3

µM, and 6 µL of nuclease-free water for a total final reaction volume of 20 µL.

Thermocycling was carried out on a CFX384 Touch (BioRad, Hercules,

California, USA) with the following conditions for both amplification efficiency tests as well as analysis of collected samples: initial denaturation step at 95°C for

10 min followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. The generation of a melt curve clarified the production of a single amplicon. Finally, 1:10 serial dilutions of stock MODV cDNA were used to generate four standard curves from

85 which the means were used to represent the tissue culture infective dose 50

(TCID50) “equivalent titer” of recovered viral RNA.

Virus Isolation from Select Tissues

Following homogenization, tissue sample supernatants were prepared for virus isolation by first creating suspensions containing 30% volume/volume (v/v) fetal bovine (FBS; Corning, Corning, NY, USA). All samples were then plated in triplicate on subconfluent layers of Vero E6 cells and incubated for 2.5 hours at 37°C and 5% CO2. Positive controls contained 1:10 dilutions of stock

MODV in 1x PBS and 30% v/v FBS while negative controls contained 30% v/v

FBS in 1x PBS alone. Following incubation, all samples and control inoculums were removed, and cell monolayers were washed once with 1x PBS. Complete

DMEM culture media containing 2% v/v FBS was added to all wells, and plates were incubated at 37°C and 5% CO2 for a maximum of 14 days or until negative controls demonstrated any signs of senescence. Samples were scored as positive for virus isolation if at least two of three wells indicated clear signs of

CPE. The same procedure as described above was utilized to detect infectious virus in blood sera and urine (examined at a 1:10 dilution) samples given sample availability for all infected animals in this study.

Sample Collection and Preparation for Examination of Immune Genes

Liver, spleen, kidney, and brain tissues were collected from infected

(biological triplicates) and mock-infected (biological replicates) deer mice at days

86

1, 6, and 10 post-infection and to examine immune gene expression. All specimens collected at the time of necropsy were immediately flash-frozen in liquid nitrogen and held at -80°C until further use. Using a Mini-Beadbeater tissue disrupter (Biospec, Bartlesville, OK, USA), approximately 30 mg of frozen tissue were added to a 1.5 mL snap-cap tube containing TRK lysis buffer (E.Z.N.A Total

RNA kit, Omega Bio-Tek, Norcross, GA, USA), zirconia/silica beads, and 20

µL/mL of β-mercaptoethanol. Following tissue homogenization, entire suspensions were transferred to a QiaShredder column (Qiagen, Germantown,

MD, USA) and centrifuged at 14,000 x g for 8 min at 4°C to further mechanically lyse cells and pellet cellular debris. After spinning, supernatants were carefully removed and utilized for RNA extraction, as described below.

RNA Extraction and cDNA Synthesis

Total cellular RNA was extracted from tissue homogenate supernatants using an E.Z.N.A Total RNA kit (Omega Bio-Tek, Norcross, GA, USA) according to the manufacturer’s protocol. RNA extracts (eluded in 50 µL of nuclease-free water) were assessed for concentration and purity using a Nanodrop 2000 spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA). For any samples with 260/280 or 260/230 absorbance ratios below 1.8, an RNA cleanup protocol was utilized in which samples were first q.s.’d to 50 µL with nuclease- free water before adding 5 µL of 3M sodium acetate (pH = 5.5) and 50.0 µL of room temperature isopropanol. Samples were incubated at room temperature in the dark for 20 min before centrifuging at 4°C for 10 min at 14,000 x g to pellet

RNA. Pellets were washed three times with ice-cold 70% ethanol, tubes were left

87 open in an RNA hood for any remaining alcohol to evaporate, and finally, pellets were resuspended in 50 µL of fresh nuclease-free water and reanalyzed for purity and concentration. To determine RNA integrity, random samples were selected and run through agarose gel electrophoresis. Prior to cDNA synthesis, all samples were treated for genomic DNA contamination by using a PerfeCTa

DNase I removal kit (Quantabio, Beverly, MA, USA) according to the manufacturer’s protocol. In brief, 1 ug of total RNA was mixed with 0.5 µL (1 U) of recombinant DNase I, 1 µL of reaction buffer, and sufficient nuclease-free water to constitute a final reaction volume of 10 µL. Samples were incubated for 30 min at 37°C and then returned immediately to ice. 1 µL of DNase Stop Buffer

(provided in the same kit) was added to each sample and reactions were incubated at 65°C for 10 min to halt the enzymatic activity of the DNase. Finally, first-strand cDNA was synthesized by mixing the entire DNase-treated reaction with 4 µL of qScript cDNA SuperMix (Quantabio, Beverly, MA, USA) and 5 µL of nuclease-free water followed by incubation at 25°C for 10 min, 42°C for 30 min, and then 85°C for 5 min.

Immune Gene Expression Analysis

Prior to immune gene expression analysis, all of the primers used for this study were first analyzed for amplification specificity and efficiency using endpoint PCR and RT-qPCR, respectively. For the former, a minimum of two experimental samples were utilized to conduct endpoint PCR using 2 µL of cDNA template and an AccuStart II PCR SuperMix kit (Quantabio, Beverly, MA, USA) according to the manufacturer’s protocol. Amplification of a single gene product

88 at the expected amplicon size was observed using agarose gel electrophoresis.

Primers were then assessed for amplification efficiency by conducting RT-qPCR on 1:10 dilutions of the same template cDNA. All primers used for this study were either recovered from published literature or manually designed by T. Sherman with the aid of PrimerQuest and OligoAnalyzer (IDT). Details of all primer sequences, amplification efficiencies, and gene accessions can be found in

Table 3.1.

Following satisfactory internal validation measures, analysis of immune gene RT-qPCR was conducted by combining 2 µL of template cDNA (diluted to

10 ng/µL) with 5 µL of SsoAdvanced SYBR Supermix (BioRad, Hercules,

California, USA), 2 µL of forward and reverse primers each at 2 µM, and 1 µL of nuclease-free water for a final total reaction volume of 10 µL. All samples were run on a CFX384 Touch thermocycler (BioRad, Hercules, California, USA) with the following conditions: initial denaturation step at 95°C for 10 min followed by

40 cycles at 95°C for 15 s and 60°C for 1 min. Additionally, melt curves were generated for each sample, as described above. Relative gene expression was calculated using the comparative Ct method (i.e., ΔΔCt) as proposed by Livak and Schmittgen whereby immune genes were first normalized against the constitutively-expressed β-actin gene (ΔCt) and then calibrated against the ΔCt of mock-infected controls (ΔΔCt) (252). Finally, relative gene expression was depicted as the mean fold change (2-ΔΔCt) ± standard error of the mean (SEM) between infected and mock-infected animals at each of the previously described

89 collection timepoints. Samples that did not generate a Ct value were given a default value of 40.

Statistical Analysis

Log10 transformations were first applied to all relative gene expression values prior to any downstream statistical analysis. Using SPSS v25.0 (IMB,

Armonk, New York, USA), separate MANOVAs were utilized to assess for significant differences in the gene expression levels a) between all tissues and collection timepoints and b) between collection timepoints within brain tissues alone. Additionally, a two-factor ANOVA was used to examine the relative expression levels of Oas1b between tissues at each collection timepoint. For all analyses, significant differences between groups were evaluated using corrected models and an alpha significance of 0.05.

90

Table 3.1: Sequence data and amplification efficiencies for deer mouse primers.

Target Primer Sequence (5’ to 3’) Size R2 Eff. (%) Gene Accession Reference

β-Actin Mallarino et 123 1.000 102.0 XM_006998174.2 Fw GCTACAGCTTCACCACCACA al., 2016 (253) Rv TCTCCAGGGAGGAAGAGGAT Oas1b Fw CAGTATGCCCTGGAGCTGC 111 0.998 98.3 XM_006970848.1 This study Rv GTACTTGGTGACCAGTTCC IFN-γ Fw ACAGCAGTGAGGAGAAACGG 115 0.970 94.7 AY289494.1 This study Rv GACAGGCGGTACATCACTCC TNFα Fw GGGCTGTACCTCGTCTACTC 121 0.999 100.4 XM_006995235.2 This study Rv ACAGGAGGTTGACTTTGTCC TGFβ Schountz et Fw CGTGGAACTCTACCAGAAATACAGC 96 0.999 95.6 XM_006988036.2 al., 2012 (190) Rv TCAAAAGACAACCACTCAGGCG IL-6 Schountz et Fw CCATCCAACTCATCCTGAAAGC 101 0.999 96.1 AY256518.1 al., 2012 (190) Rv CCACAGATTGGTACACATAGGCAC IL-10 Fw CAGACCTACACGCTTCGAG 128 0.999 111.2 XM_006995328.2 This study Rv CCCAGGTAACCCTTAAAGTCC MODV-NS5 Leyssen et al., Fw CCAGGACAAGTCATGTGGTAGC 101 0.998 107.5 NC_003635.1 2001 (66) Rv TCCCAAAGATGTTCCTCACCTT Table 2

Results

Modoc Virus Replicates to Low Levels in Deer Mouse Tissues

Viral RNA in select deer mouse tissues (spleen, kidneys, liver, and brain) and bodily fluids (blood sera and urine) could be quantified by RT-qPCR as early as 24 hours post-intraperitoneal infection and remained detectable throughout the entire 10-day experimental period (Figure 3.1 and 3.2). Interestingly, the examination of MODV RNA in these tissues revealed a relatively consistent trend in which viral loads were maintained at an equilibrium roughly one order of magnitude less than that of the experimentally-administered infectious dose

3.0 (mean loads among all tissues = 10 TCID50 infective dose equivalent, SEM =

91

102.2). Likewise, viral RNA loads in the urine and sera remained comparatively low throughout the study duration (mean loads in urine = 102.4, SEM = 101.9 and sera = 102.7, SEM = 102.3).

In regard to viral replication kinetics in individual tissues and fluids, the results of this study predominantly corroborate published data from the limited number of previous in vivo investigations of deer mice experimentally infected with MODV. Specifically, minimal viral RNA shed in the urine with a peak quantity detected at day 6 post-infection supports findings by Davis et al. in which viruria was observed on days 4 to 6 post-intranasal inoculation (8). Similarly, higher loads of viral RNA in the sera early during infection (days 1 to 4) correlate with detected during the same timeframe in two separate studies following both intranasal and intraperitoneal challenge (8, 254). Detection of viral loads in liver, for which mean values were the lowest among all tissues throughout the experiment (mean = 102.6, SEM = 102.0), matches findings by Davis et al. in which virus isolation was barely quantifiable throughout a 10-day infection period.

Further, both this study as well as the work conducted by Davis et al. demonstrate selective replication of MODV in the spleen (mean = 103.2, SEM =

102.7) with a peak titer at day 6 post-infection (both studies) even when routes of inoculation differ (intranasal versus intraperitoneal). Levels of MODV RNA detected in deer mouse kidneys were reminiscent of patterns observed by Davis et al. in which viral loads appear to peak and then declined every 2 to 4 days but remain relatively low throughout the study period (mean = 102.7, SEM = 102.1).

Finally, in contrast to the aforementioned tissues and fluids, quantification of

92

MODV RNA in the brain (mean = 103.0, SEM = 102.0) of infected deer mice did not parallel the results of virus isolation in the Davis et al. investigation; however, this could be explained by the decreased sensitivity for detection in virus isolation methods which utilize cell culture as compared to the detection of viral nucleic acids via RT-qPCR.

To glean additional clues about the tropism of MODV in deer mouse tissues, this study also conducted virus isolation utilizing cell culture methods, which allow for the detection of infectious virus present at the time of tissue collection. In brief, no infectious virus was detected in the urine (n = 18) throughout the entire experimental period or in the sera (n = 5) of any individuals past day 8 post-infection. In all tissues tested, infectious virus was recovered from at least 2 of 3 bioreplicants on days 8 and 10 post-infection with additional isolations in at least 1 of 3 replicants on day 1 (liver), day 2 (spleen), day 4

(kidneys and brain) and day 6 (brain) post-infection. Interestingly, these findings suggest an increased tropism of MODV toward the brain and liver compared to a higher frequency of successful isolations in the spleen as demonstrated by Davis et al. Granted, it is unfortunate to report that samples from multiple replicants at mid-collection timepoints from all tissues in the present study were lost due to fungal contamination during plate incubation and thus could not be included in this examination. Due to this shortcoming, a more exact illustration of tissue tropism during acute infection could not be constructed.

93

Figure 11

Viral Loads In Urine Viral Loads In Sera 10000 10000

1000 1000

/100mL) 100 100 50

10 10 (TCID

Infective Dose Equivalent Dose Equivalent Infective 1 1 12345678910 12345678910

Days Post-Infection Figure 3.1: Viral RNA loads detected in the urine and blood sera of infected deer mice over a 10-day infection period. Mean (log10) viral loads were quantified using two-step RT-qPCR and translated into TCID50 “equivalent titers.” A limit of detection (dotted grey line) was created by conducting RT-qPCR on MODV RNA template that had not undergone reverse-transcription.

94

Figure 12 Virus Isolation and Loads In Spleen Virus Isolation and Loads In Kidneys

10000 3 10000 3

1000 1000 2 2

/100uL) 100 100 50 1 1

(TCID 10 10

Infective Dose Equivalent Dose Equivalent Infective 1 0 1 0 Successive Isolations Successive Virus 12345678910 12345678910

Virus Isolation and Loads In Liver Virus Isolation and Loads In Brain

10000 3 10000 3

1000 1000 2 2 100 100 1 1 10 10 (TCID50/100uL)

1 0 1 0 Isolations Successive Virus Infective Dose Equivalent Dose Equivalent Infective 12345678910 12345678910

Days Post-Infection

Figure 3.2: Recovery of infectious virus and viral RNA in select tissues collected from infected deer mice over a 10-day infection period. Mean (log10) viral loads were quantified using two-step RT-qPCR and translated into TCID50 “equivalent titers” (primary vertical axis). Isolation of infectious virus from biotriplicates at each collection timepoint was deemed positive if at least two of three wells indicated clear signs of virally-induced CPE when homogenized tissues samples were incubated with Vero cells (secondary axis).

Immune Genes Are Differentially Expressed in the Brain of Deer Mice During Acute Infection with Modoc Virus

The transcript expression levels of five immune genes (IFN-γ, TNFα, IL-6,

IL-10, and TGFβ) were analyzed using RT-qPCR following in vivo experimental infections of deer mice with MODV. At three distinct timepoints (1, 6, and 10 d.p.i.), animals were euthanized and necropsied for the collection of kidney,

95 spleen, liver, and brain tissues. Given the predicted variance that may occur between gene expression levels, a log10 transformation was applied to all relative expression values prior to any downstream statistical analysis. Once transformed, a Shapiro-Wilk test for normality confirmed that all expression values were normally distributed (p ≥ 0.11) except for one instance (TGFβ in liver tissues; p = 0.04).

Mapping the expression levels of these genes across all tissues and timepoints (n = 3 per gene) in principle component analyses (PCA) revealed clustering which suggests that there are patterns of differential expression in the brain compared to the kidney, spleen, and liver tissues (Figure 3.3) as well as between collection timepoints (Figure 3.4).

96

Figure 13 Figure 3.3: Principle component analysis (PCA) of gene expression in MODV-infected deer mice. Plots were created in PC-ORD (V7.0) using log10 transformations of gene expression data (randomized tests, 999 runs). When grouped by tissue, gene expression values cluster differentially in brain tissues (save for one outlier) as compared to kidneys, spleen, and liver tissues.

97

Figure 14 Figure 3.4: Principle component analysis (PCA) of gene expression in MODV-infected deer mice. Plots were created in PC-ORD (V7.0) using log10 transformations of gene expression data (randomized tests, 999 runs). When grouped by collection timepoint, gene expression values appear to cluster more tightly at the first and last collection timepoints as compared to the mid-collection timepoint.

To further clarify whether the patterns observed in the principle component analyses were statistically meaningful, a multivariate analysis of variance

(MANOVA) was enlisted to identify the significance of all gene expression levels when both tissue and collection timepoint are treated as fixed factors. To this end, it was discovered that both interactions were significant (p < 0.00), suggesting that these immune genes are indeed expressed in a tissue- and time-

98 specific manner in deer mice infected with MODV. Moreover, corrected models of between-subject tests revealed that IFN-γ (p < 0.00), TNFα (p < 0.00), and TGFβ

(p < 0.00), were all significantly differently expressed, while IL-10 (p = 0.06) demonstrated marginally significant differential expression. In contrast, IL-6 (p =

0.14) does not appear to be significantly differential expressed between the selected tissues or at the timepoint collected. Tukey HSD pos hoc analyses revealed that individual genes may be significantly differentially expressed in separate tissues; however, IFN-γ was the only gene found to be differentially expressed in one particular tissue (brain) as compared to all other tissues examined (p ≤ 0.01).

Lastly, a separate MANOVA was conducted using samples collected from brain tissues alone to further elucidate any significant variation in the expression level of all genes between collection timepoints. For this analysis, brain tissue was chosen for scrupulous examination given the finding that the expression of

IFN-γ was significantly different as compared to all other tissues and because

MODV has been documented to cause neurologic disease in certain rodent models and one instance of human infection. To this end, corrected models for between-subjects effects revealed that only TNFα was significantly differentially expressed between all timepoints (p < 0.00). However, it may be noteworthy to highlight that IFN-γ displays a unimodal expression pattern while all other genes

(except for IL-10, which could not be detected at 10 days post-infection) display gradually increasing levels of expression throughout the experimental duration

(although the variance in expression for this gene at day 6 is high) (Figure 3.5).

99

Along with this, Tukey HSD post hoc analysis shows that expression of TNFα differed significantly between days 1 and 6 (p = 0.01) as well as days 1 and 10 (p

< 0.00) but not between days 6 and 10 post-infection (p = 0.53).

Figure 15

Figure 3.5: Gene expression levels in select tissues of MODV-infected deer mice. Relative expression levels were determined using the ΔΔCt method (n = 3 per timepoint) as described by Livak and Schmittgen (252). DPI; days post- infection.

The Flavivirus Resistance Gene Oas1b is Significantly Upregulated During Early Infection with Modoc Virus

A sixth immune gene, Oas1b, was chosen as an indicator of the intracellular antiviral immune response following experimental infection. To

100 assess for significant differences in Oas1b gene expression between tissues and over time, a two-factor univariate analysis of variance (ANOVA) was used in which tissue and collection timepoint were treated as fixed factors. Results from this ANOVA indicated that there was a significant difference in expression levels between collection timepoints (p ≤ 0.00) but not between the tissues examined (p

= 0.25). Further, Tukey HSD pos hoc analysis suggested that the expression of

Oas1b across all tissues did not differ significantly between days 1 and 6 post- infection (p = 0.30), but rather, demonstrated a consistent pattern of early upregulation that was maintained until at least day 6 post-infection followed by a significant decrease (p < 0.00) in expression sometime before day 10 post- infection (Figure 3.6).

Figure 16

Figure 3.6: Expression levels of Oas1b captured in selected tissues of MODV-infected deer mice. Relative expression levels were determined using the ΔΔCt method (n = 3 per timepoint) as described by Livak and Schmittgen (252). DPI; days post-infection.

101

Discussion

In regard to overt disease, no deer mice infected with MODV demonstrated any indications of distinguishable morbidity (characterized by ruffled fur or piloerection, fluid discharge from the nose, mouth or eyes, hematuria, muscle , tremor, , or general lethargy and changes in behavior) throughout the entire experimental duration. Despite this, viral RNA could be detected in multiple tissues and fluids from 24 hours post- infection until the end of the experiment at 10 days post-infection. Moreover, infectious virus could be recovered from multiple tissues in multiple individuals at various collection timepoints. Taken together, these observations corroborate previous investigations with MODV in deer mice and add further support to the proposition that these rodents may function as the hosts of the virus in nature. Finally, by mimicking the inoculation dose and route of MODV used in earlier studies, which demonstrate paralytic syndromes in Syrian golden hamsters and SCID mice around 9 days post-infection, the research presented here confirms the of use of deer mice as a comparatively non-pathogenic model during acute infection with MODV (66, 96).

The observed levels of MODV in select tissues and fluids, as determined by the quantification of viral RNA using RT-qPCR, likewise indicates that infection in deer mice is non-pathogenic and may in fact be under tight regulatory control given that viral loads are maintained at consistently low levels even when exposure to the virus is unnaturally high (as in the case of experimental infection). In the sera, viral loads from this study coincide with previous flaviviral

102 investigations, which describe that the highest mean titers are detected early post-infection and begin to wane by days 4-6 (8, 254, 255). As compared to

MBFVs and TBFVs, though, MODV does not appear to approach the higher titers of viremia required to infect an invertebrate vector (e.g., 105.0 PFU/mL threshold to infect Cx. pipiens and Cx. quinquefasciatus with WNV) (256). Granted, this study utilized a technique for viral detection (RT-qPCR), which does not necessarily equate to functional levels of infectious virus; however, it would be highly improbable that the titers of infectious virus exceed those of corresponding

RNA loads given that viral RNA may be detected from both infectious and non- infectious particles. In any case, there were no successful viral isolations from any of the deer mouse sera samples tested (in total, 2/3 individuals were tested on day 8, and 3/3 individuals were tested on day 10 post-infection).

Patterns of viral replication in the spleen and liver grossly adhere to prior investigations that depict more or less tropism toward these tissues, respectively

(8). In part, this may be explained by the fact that many flaviviruses tend to preferentially infect monocyte-derived cells (of which many can be found in the murine spleen compared to the liver) (257-259). Indeed, while hepatic tropism is apparent in the case of YFV, the alternative tropism toward splenic tissues has been demonstrated in at least two other MBFVs, WNV and ZIKV (260, 261). In contrast, low viral loads detected in the kidneys and urine (with no infectious virus isolated from 1:10 dilutions of urine for any animals) support the assertion that MODV in deer mice is less trophic toward renal tissues; in combination with previous studies demonstrating higher titers detected in the lungs, salivary-

103 submaxillary glands, or mammary tissues, there is a strong argument for horizontal transmission to occur through direct contact as opposed to excremental shedding (8, 92). Lastly, detection of viral RNA in the brain was unsurprising given the neurotrophic nature of MODV depicted in earlier studies

(66, 95). Nevertheless, a trending decrease in viral loads in the brain beginning at day 8 post-infection suggest that deer mice are able to control virus replication in brain tissues to an extent which prevents CNS-like disease (although the observed decrease is not yet significant by day 10).

The immune genes examined in this study represent a very small subset of a much larger conglomerate of molecules which act either in synchrony or antagonistically to elicit a productive immune response against any given pathogen. Known more specifically as cytokines, these molecules may serve as the “alarm bells” of pathogen infection (e.g., IFN-γ), as proinflammatory mediators (e.g., TNFα and IL-6), or as immunomodulatory regulators (e.g., IL-10 and TGFβ). Importantly, it is believed that the level of expression for each, as well as the timing of expression, may ultimately lead to the coordination of an immune response which either leads to a) pathogen clearance, b) pathogen tolerance, or c) disease as a result of uncontrolled pathogen replication or collateral damage from the immune response itself (or both).

In particular, transcripts for interferon-gamma (IFN-γ or type II IFN), which serves multiple important functions in controlling microbial invasion and physiologic homeostasis, were discovered to be significantly differentially expressed in the brain as compared to all other tissues (Tukey HSD, p ≤ 0.01).

104

This finding is in accordance with the transcriptome profiling data of deer mice infected with Powassan virus (POWV), a tick-borne, neurotrophic flavivirus, demonstrating that the differential expression of IFN-γ is significant in the brain from days 3 to 15 post-infection (247). Further, the importance of IFN-γ functionality has been reported in the context of infection with WNV, another flavivirus with neurotrophic potential (262). By comparing viral replication and the host response in wild-type mice to that of either IFN-γ-/- or IFN-γR-/- knockouts, it was found that individuals which lacked the ability to produce or respond to IFN-γ signals succumbed to lethal infection faster and more frequently under the burden of higher viremia and viral replication in the lymphoid tissues (the latter of which is believed to influence viral invasion into the , or

CNS). At the same time, separate studies have shown that the subsequent downregulation of IFN-γ following initial virally-induced expression is of equal importance to avoid disintegration of the blood-brain barrier (BBB) (263). This considered, and in light of the perceived necessity for the presence and timing of

IFN-γ expression in previous studies with neurotrophic flaviviruses, it appears fitting that deer mice respond to MODV infection in the brain via the significant but unimodal expression of IFN-γ which may correlate with the decline in viral loads witnessed at the final collection timepoint (Figure 3.5).

While none of the other cytokine genes examined display the same statistical significance in the brain as IFN-γ, it is interesting that TNFα and IL-6 display increasing levels throughout the experimental duration with expression values that differ significantly compared to uninfected controls by days 6 and 10

105 post-infection. Tumor factor alpha (TNFα), like IFN-γ, serves myriad roles in antimicrobial immunity as well as normal physiologic homeostasis (264).

As an antiviral, TNFα elicits fever, stimulates the expression of acute-phase liver proteins, and may induce apoptosis in infected endothelial cells, among other functions. Similarly, IL-6 helps to orchestrate the aforementioned immune armaments in addition to promoting inflammation, which aids in leukocyte migration to areas of viral infection. Indeed, both of these cytokines have been correlated with positive clinical outcomes of viral disease; however, atypically high levels of each have been found in instances of severe flaviviral illness under both the context of CNS involvement (JEV) as well as hemorrhagic fevers

(DENV) (143, 265, 266). It is therefore tempting to postulate that deer mice exhibit an expression pattern that productively controls viral replication in the brain without the consequence of immunopathology which may accompany a fulminant inflammatory response.

Perhaps most interesting is the observed expression of Oas1b, an antiviral effector protein induced by IFN-α/β (i.e. type I IFN) signaling. Although we chose not to examine type I IFN transcripts (primarily due to the absence of any verified

IFN-α/β transcript sequences for peromyscines in the NCBI database), it is well known that many flaviviruses antagonize this vital antiviral signaling system with a plethora of strategies (123, 126, 267, 268). Moreover, there is evidence that some of these strategies are conserved among diverse flaviviruses (e.g., WNV vs. JEV), implicating that research examining counter-antagonism using less pathogenic representatives (e.g., MODV) may still lead to novel

106 applicable to more virulent members (66, 123). Therefore, as an alternative to examining transcript levels of the type I IFNs themselves, we choose to examine an IFN-stimulated gene (ISG) as a quantifiable downstream indicator of productive type I IFN signaling (269). In Mus, it is known that the Oas1b isoform of the Oas1 protein is particularly efficacious toward controlling infection of WNV

(270-273). In one particular study examining transcriptional regulation in Mus models during infection with WNV, expression of Oas1b was observed to be increased in the spleen during infection from seven distinct strains of the virus

(274). In the present study, we found that Oas1b is significantly expressed in all examined tissues as compared to uninfected controls within the first 24 hours of infection and remains significantly expressed to a high degree until some point after the collection period at 6 days post-infection. Indeed, the near-immediate and extreme expression of these transcripts imply that MODV recognition in deer mice is rapid and leads to the activation of antiviral ISGs through productive type

I IFN signaling. This observation lends promise for continued future investigations, given the results of separate studies depicting the importance of virus-specific ISGs and their role in limiting tissue-specific infections (247, 275,

276).

Conclusions and Future Directions

While it has been established that deer mice are capable of hosting a variety of bacterial, protozoic, and viral pathogens, relatively little molecular work has been conducted to understand the ability of these rodents to act as reservoir hosts, which avoid overt disease. Historically, studies examining cytokine gene

107 expression in the peromyscine immune response involved models of infection by members of the genus (family ) (190, 246). In these examples, it appears as though higher expression of transforming growth factor beta (TGFβ) at later timepoints of infection (between 15 and 45 days post- infection) leads to immune suppression that fails to clear viral infection but prevents immunopathology caused by a sustained proinflammatory response.

More recently, models have been constructed using deer mice to examine flaviviral pathology; however, these models either utilize in vitro methods that may fail to capture the whole-animal response or exploit TBFVs (e.g., POWV) that must be manipulated solely in specialized -3 (BSL-3) facilities

(104, 247). To complement these studies and contribute further to the increasing knowledge of virus-reservoir-host relationships, as well as the intricate discrepancies between various flavivirus replication strategies and disease pathologies, the research presented here, therefore, takes advantage of a rodent-borne flavivirus to conduct an in vivo investigation of the immune response during acute infection in the virus’s putative reservoir host in a BSL-2 laboratory.

Future investigations related to the research presented here may wish to extend the duration of experimental infection to better capture ongoing changes to viral replication kinetics, tissue tropism, and the host response potentially missed by only examining the first 10 days following infection. Likewise, it could be interesting to include an examination of the lungs and salivary-submaxillary glands from which Davis et al. recovered the highest titers of MODV from all

108 examined tissues in infected deer mice (8). Increasing the sample size of replicates per collection timepoint and limiting the animals used in the study to a single sex may help to alleviate some of the variance observed with gene expression analyses (Appendix B); however, this variation may nevertheless persist given the relatively outbred nature of deer mice as compared to other laboratory rodents (190). Finally, proteomic studies should be used to compliment gene expression analyses such that post-transcriptional events can be resolved and interactions between cellular and viral proteins can be elucidated.

109

CHAPTER IV

COMPARATIVE HOST RESPONSE AND GENE REGULATION IN DEER MICE AND SYRIAN GOLDEN HAMSTERS DURING ACUTE INFECTION WITH MODOC VIRUS

Author: Tyler Sherman

Contributions: Developed and implemented the study design. Generated and analyzed data. Wrote first draft of the manuscript.

Co-Author: Dr. Ann C. Hawkinson

Contributions: Helped conceive the study topic. Provided guidance on study design. Provided feedback data interpretation and manuscript preparation.

Abstract

Despite their pervasive distribution and global burden on public health, the factors which influence the outcome of clinical disease upon infection with flaviviruses (family Flaviviridae, genus Flavivirus) remain incompletely understood. While it is known that differences in the host innate immune response will subsequently influence aspects of viral tropism, dissemination, and long-term persistence, it is challenging to delineate which factors are protective and which may contribute to immunopathogenesis. In an attempt to shed light on this matter, we have chosen to examine the patterns of viral replication and immune gene expression in two unique rodent models during acute infection with

Modoc (MODV), a rodent-borne flavivirus. Deer mice (Peromyscus maniculatus) and Syrian golden hamsters (Mesocricetus auratus) were challenged

110 intraperitoneally and serially euthanized over a 10-day period. Using RT-qPCR, it was determined that viral RNA could be detected from 24 hours post-infection in the kidneys, spleen, liver, brain, urine, and sera. In hamsters, the levels of viral

RNA generally began lower than in deer mice, increased in every tissue and fluid by day 4, and then remained generally higher in all tissues except for the spleen and liver. Infectious virus could be recovered from all tissues in the deer mice but only at later timepoints. In contrast, infectious virus could only be recovered from the kidneys and liver of hamsters, although virus isolation was successful at all timepoints examined. Neither viremia nor viruria was evident for either animal during the experimental period. The expression of five important cytokine genes

(IFN-γ, TNFα, IL-6, IL-10, and TGFβ) as well as an anti-flaviviral effector protein

Oas1b were also examine using RT-qPCR and the ΔΔCt method. Statistical analysis with a MANOVA suggests significance in the complete expression profile of cytokine genes when examined across tissues and over time between rodent species (p = 0.02). Further, between-subject effects with corrected models revealed a significant difference specific to the expression IFN-γ (p = 0.01). In support of this, the relative expression of IFN-γ in the of infected hamsters was significantly higher than that of deer mice by day 6 post-infection and persistent until the end of the experiment. As over-expression of IFN-γ has occasionally been shown to correlate with losses in vascular integrity, the findings presented here may highlight a potential mechanism for viral dissemination into the CNS following fulminate activation of cells from the innate response (e.g., NK cells). This could explain the increase in viral RNA detected in

111 the brains of hamsters followed by the paralytic syndromes, which often develop in these animals (neither of which appears to occur in the deer mice).

112

Introduction

Flaviviruses represent a taxonomic collective of globally-distributed, predominantly vector-borne pathogens that may elicit a vast range of clinical pathologies. Fortunately, the majority of human infections appear to self-resolve following only mild or subclinical illness; however, instances of severe disease are not uncommon and may result in multiple-organ failures, autoimmune disorders, malformations in fetal development, hemorrhagic fevers with vascular leak syndromes, or meningioencephalitic ailments with the potential for permanent neurologic sequelae (11, 277). Moreover, a significant increase in the spread and incidence of flaviviral disease has provoked renewed concern for the burden of these viruses on global public health (61, 180, 278). For perspective, dengue virus (DENV), now considered to be the most common vector-borne virus in the world, infects more than 390 million individuals annually (95% confidence interval) and leads to clinical disease in approximately 96 million (10,

279). Other notable representatives such as West Nile virus (WNV), Japanese encephalitis virus (JEV), yellow fever virus (YFV), and the newly emergent Zika virus (ZIKV) continue to threaten individuals worldwide, with a conservative estimate of annual infection approaching nearly 4 billion (3, 4, 59, 78, 82). The issue of exposure is further compounded by anthropogenic impacts on the environment (e.g., global warming and plastic waste production), leading to range expansion for vectors like ticks and mosquitoes which carry and transmit flaviviruses (among others) (5, 83, 85, 88, 279). Combined with a relative lack of flaviviral vaccine availability and FDA-approved antiviral therapies for those at

113 risk or acutely infected, the need for continued research with these pervasive pathogens is apparent (10).

As a member of the genus Flavivirus, Modoc virus (MODV) is characterized by an enveloped, (+)ssRNA genome that encodes a total of three structural (envelope, E; capsid, C; pre-membrane/membrane, prM) and seven non-structural (NS1, NS2A, NS2B, NS3, NS4A, NS4B, NS5) proteins. Further, the 5’ end of the viral genome contains a methyl cap that mimics host transcripts while the 3’ end encompasses various non-coding elements (e.g., subgenomic flaviviral RNA, sfRNA), all of which have been universally speculated to contributed to flaviviral infectivity and transmission, replication, and host pathogenesis (51, 94, 280, 281). In contrast to the aforementioned flaviviruses, however, isolation of infectious MODV has occurred exclusively from wild deer mice (Peromyscus spp.) in nature and cannot be recovered from passage in both invertebrate as well as vertebrate cell lines for unknown reasons (52, 67, 92,

100). Given this restriction in vector/host range, MODV belongs to a group of other related flaviviruses known as the no known vector (NKV) cluster (17). To date, only a few instances of clinical human disease caused by NKVs have been officially documented (75, 101). Still, NKVs like MODV may provide an attractive avenue for continued flaviviral research given their ability for manipulation at biosafety level-2 (BSL-2) facilities and capacity to recapitulate examples of human-like flaviviral disease in rodent models (e.g., mild febrile illness with the potential for reproductive or CNS involvement (96, 282).

114

As the putative reservoir hosts of MODV in nature, as well as a variety of other viral, bacterial, and eukaryotic pathogens, infectious disease research with peromyscines has questioned the immunoregulatory processes at play which prevent overt disease in these rodents (188, 190, 246). On the opposite end of the spectrum, Syrian golden hamsters (Mesocricetus auratus) can be utilized as models which develop disease reminiscent of human infections from many pathogens without the need for genetic manipulation or chemical immunocompromise necessary in other rodent models such as Mus musculus

(66, 234, 283, 284). Indeed, both animals have been enlisted to study facets of

MODV ecology and infection in the past and have revealed clues about the potential tropism and subsequent pathology of MODV in vivo (7, 8, 65, 92, 95,

254). In particular, it is known that MODV will induce a poliomyelitis-like syndrome in hamsters that appears absent from deer mice when experimentally infected through the same route (separate studies) (66, 254). Granted, no studies thus far have included an examination of any immune genes expressed during acute infection.

Due to their medical relevance, mosquito-borne (MBFVs) and tick-borne flaviviruses (TBFVs) have been studied extensively over the preceding decades.

Indeed, advances in molecular techniques have created many opportunities for refining the scientific and medical communities’ current understanding of flaviviral ecology and disease. Despite this, many questions still remain with regard to the precise details which influence flaviviral transmission, replication, tissue tropism, immune subversion strategies, and subsequent disease pathologies.

115

Nevertheless, there appears to be sufficient justification to postulate that the exact mechanisms leading to the various observed outcomes of infection culminate as the result of a host individual’s unique immune response (45, 124,

147, 148, 150, 268, 285, 286). As such, investigations that directly compare the immune response in terms of cytokine gene expression between different models of disease may offer a novel approach to elucidating the underlying mechanisms which lead to either viral clearance/persistence or pathology. Here we present one such approach which utilizes MODV, the most well-characterized NKV, to compare and contrast the discrepancies in viral replication kinetics, tissue tropism, and immune gene expression between deer mice and Syrian golden hamsters during an acute duration of experimental infection. Therefore, it is hypothesized that Syrian golden hamsters will demonstrate clinical signs of pathology when infected with MODV, thereby constituting a comparative pathogenic model. As such, viral loads should differ significantly in various tissues over time when compared to deer mice. Finally, the relative expression values of immune genes should differ significantly between hamsters and deer mice in various tissues at distinct timepoints throughout the experimental duration, highlighting the different modalities of immune regulation elicited by each rodent model even when pathogen dose and route of infection are standardized.

116

Materials and Methods

Ethics Statement and Animal Husbandry

All of the methods conducted in this study were approved by the

University of Northern Colorado (UNCO) Institutional Animal Care and Use

Committee (IACUC; Protocol No. 1603C-AH-DmH-19) in accordance with the

Animal Welfare Act. Live deer mice (Peromyscus maniculatus) of both genders

(9 male and 9 female), aged 10-16 weeks old, were supplied from an established breeding colony at UNCO while Syrian golden hamsters (Mesocricetus auratus) were purchased from Envigo Laboratories (Denver, CO, USA).

Modoc Virus

Modoc virus (MODV; strain M544) obtained from ATCC (Manassas, VA,

USA) was passaged once in Vero E6 cells (ATCC) for seven days prior to titration using the Reed-Muench 50% endpoint dilution assay.

Experimental Infections

Eighteen deer mice and eighteen hamsters (nine of each gender for each species) ranging in age from 10-16 weeks old were inoculated intraperitoneally

4.2 with 0.3 mL of MODV at an infectious titer of approximately 10 TCID50 (virus diluted in sterile 1x DPBS; pH = 7.3). Likewise, negative control animals (12 of each species, 6 of each gender) were inoculated intraperitoneally with 0.3 mL of

1x DPBS (pH = 7.3). Following administration of virus or saline, all animals were housed individually with free access to food and water and monitored daily for any signs of distress or morbidity (e.g., tachypnea, altered gait, abnormal behavior, etc.). As well, all animals were weighed daily beginning at 24 hours

117 post-infection and subsequently throughout the entire experimental period.

Experimentally infected animals were serially euthanized at the following predetermined timepoints: 1, 2, 4, 6, 8, and 10 days post-infection. Negative control animals were euthanized at 1, 6, and 10 days post-infection. For all animals, euthanasia was performed by means of cervical dislocation and cardiac exsanguination following anesthesia with isoflurane vapors.

Sample Collection and Preparation for the Examination of Replication Kinetics

Necropsies were performed immediately following euthanasia and included the collection of urine (extracted directly from the bladder using a tuberculin syringe), blood sera (collected at the time of cardiac exsanguination and separated by incubating at room temperature for 10 min, chilling on ice for

60 min, and then centrifuging at 4000 x g for 10 min), kidneys, liver, spleen, and brain. All collected tissues were frozen immediately in liquid nitrogen baths and stored at -80°C until further use. To prepare kidney, liver, spleen, and brain tissues for RNA extraction and virus isolation, tissue samples of approximately

30 mg were first homogenized using a Mini-Beadbeater tissue disrupter

(Biospec, Bartlesville, OK, USA) with zirconia/silica beads in 600 µL of sterile 1x

PBS. Homogenates were then centrifuged through a QiaShredder column

(Qiagen, Germantown, MD, USA) at 14,000 x g for 8 min at 4°C to further mechanically lyse cells and pellet cellular debris. Following centrifugation, 150 µL of supernatant were aseptically removed and used for viral RNA extraction while

118 the remaining 450 µL were used for virus isolation. 10 µL of urine was first removed from all samples for subsequent virus isolation and remaining volumes comprised of less than 140 µL were brought up to this volume with 1x PBS for use in RNA extraction. In contrast, 140 µL of blood sera were first separated from collected samples for use in RNA extraction; all remaining blood sera were then used to perform viral isolation.

Extraction of Viral RNA and Synthesis of Viral cDNA

RNA was extracted from tissue homogenates, urine, and blood sera samples using a QIAmp Viral RNA Mini kit (Qiagen, Germantown, MD, USA) according to the manufacturer’s protocol. To reverse-transcribe viral RNA, 2 µL of sample extract were first mixed with 5.2 µL of nuclease-free water and 2.8 µL of MODV-specific forward and reverse primers (Eurofins, Louisville, KY, USA) at a final concentration of 280 nM (66). RNA and primers were incubated at 94°C for 1 min and then placed in an ice bath for 3 min before the addition of a mix containing 0.5 µL of an Episcript RT enzyme and 2 µL of 10x RT buffer (Lucigen,

Radnor, PA, USA), 2 µL of DTT, 1 µL of dNTPs at 10 mM, 0.5 µL of a recombinant RNase inhibitor (New England BioLabs, Ipswich, MA, USA), and 4

µL of nuclease-free water for a total final volume of 20 µL per reaction. First- strand cDNA synthesis was carried out by incubating the entire reaction mixture at 45°C for 1.5 hours, followed by 5 min at 85°C to halt enzymatic activity.

Samples were selected at random along with positive controls (obtained from stock suspensions of MODV) to assess for successful RNA extraction and cDNA synthesis by conducting endpoint PCR using the manufacturer’s protocol for an

119

AccuStart II PCR SuperMix kit (Quantabio, Beverly, MA, USA), the same MODV- specific primers at a final concentration of 300 nM, and the following cycling parameters: 5 min at 95°C followed by 40 cycles of 15 s at 95°C and 1 min at

60°C. Amplification of the expected gene product was observed using agarose gel electrophoresis.

Quantification of Viral Loads Using RT-qPCR

Absolute quantities of viral cDNA collected from select tissues were assessed using RT-qPCR and MODV-specific primers that recognize a region contained within the NS5 gene of the viral genome (66). To begin, the amplification efficiency of these primers were assessed by performing qPCR on

1:10 dilutions of reverse-transcribed viral RNA extracted from stock MODV suspensions. Following amplification efficiency optimization, qPCR analysis was conducted by combining 2 µL of viral cDNA template with 10 µL of PerfeCTa

SYBR FastMix (Quantabio, Beverly, MA, USA), 2 µL of forward and reverse primers each at 3 µM, and 6 µL of nuclease-free water for a total final reaction volume of 20 µL. Thermocycling was carried out on a CFX384 Touch (BioRad,

Hercules, California, USA) with the following conditions for both amplification efficiency tests as well as analysis of collected samples: initial denaturation step at 95°C for 10 min followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. The generation of a melt curve clarified the production of a single amplicon. Finally,

1:10 serial dilutions of stock MODV cDNA were used to generate four standard curves from which the means were used to represent the tissue culture infective dose 50 (TCID50) “equivalent titer” of recovered viral RNA.

120

Virus Isolation from Select Tissues

Following homogenization, tissue sample supernatants were prepared for virus isolation by first creating suspensions containing 30% volume/volume (v/v) fetal bovine serum (FBS; Corning, Corning, NY, USA). All samples were then plated in triplicate on subconfluent layers of Vero E6 cells and incubated for 2.5 hours at 37°C and 5% CO2. Positive controls contained 1:10 dilutions of stock

MODV in 1x PBS and 30% v/v FBS while negative controls contained 30% v/v

FBS in 1x PBS alone. Following incubation, all samples and control inoculums were removed, and cell monolayers were washed once with 1x PBS. Complete

DMEM culture media containing 2% v/v FBS was added to all wells, and plates were incubated at 37°C and 5% CO2 for a maximum of 14 days or until negative controls demonstrated any signs of senescence. Samples were scored as positive for virus isolation if at least two of three wells indicated clear signs of

CPE. The same procedure, as described above, was utilized to detect infectious virus in blood sera and urine (examined at a 1:10 dilution) samples given sample availability for all infected animals in this study.

Sample Collection and Preparation for Examination of Immune Genes

Liver, spleen, kidney, and brain tissues were collected from infected

(biological triplicates) and mock-infected (biological replicates) deer mice and hamsters at days 1, 6, and 10 post-infection and to examine immune gene expression. In one instance, a hamster was euthanized early (day 7 instead of

121 day 10) due to apparent morbidity secondary to experimental infection; data collected for this hamster was still grouped with others from the day 10 collection timepoint. All specimens collected at the time of necropsy were immediately flash frozen in liquid nitrogen and held at -80°C until further use. Using a Mini-

Beadbeater tissue disrupter (Biospec, Bartlesville, OK, USA), approximately 30 mg of frozen tissue were added to a 1.5 mL snap-cap tube containing TRK lysis buffer (E.Z.N.A Total RNA kit, Omega Bio-Tek, Norcross, GA, USA), zirconia/silica beads, and 20 µL/mL of β-mercaptoethanol. Following tissue homogenization, entire suspensions were transferred to a QiaShredder column

(Qiagen, Germantown, MD, USA) and centrifuged at 14,000 x g for 8 min at 4°C to further mechanically lyse cells and pellet cellular debris. After spinning, supernatants were carefully removed and utilized for RNA extraction, as described below.

RNA Extraction and cDNA Synthesis

Total cellular RNA was extracted from tissue homogenate supernatants using an E.Z.N.A Total RNA kit (Omega Bio-Tek, Norcross, GA, USA) according to the manufacturer’s protocol. RNA extracts (eluded in 50 µL of nuclease-free water) were assessed for concentration and purity using a Nanodrop 2000 spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA). For any samples with 260/280 or 260/230 absorbance ratios below 1.8, an RNA cleanup protocol was utilized in which samples were first q.s.’d to 50 µL with nuclease- free water before adding 5 µL of 3M sodium acetate (pH = 5.5) and 50.0 µL of room temperature isopropanol. Samples were incubated at room temperature in

122 the dark for 20 min before centrifuging at 4°C for 10 min at 14,000 x g to pellet

RNA. Pellets were washed three times with ice-cold 70% ethanol, tubes were left open in an RNA hood for any remaining alcohol to evaporate, and finally, pellets were resuspended in 50 µL of fresh nuclease-free water and reanalyzed for purity and concentration. To determine RNA integrity, random samples were selected and run through agarose gel electrophoresis. Prior to cDNA synthesis, all samples were treated for genomic DNA contamination by using a PerfeCTa

DNase I removal kit (Quantabio, Beverly, MA, USA) according to the manufacturer’s protocol. In brief, 1 ug of total RNA was mixed with 0.5 µL (1 U) of recombinant DNase I, 1 µL of reaction buffer, and sufficient nuclease-free water to constitute a final reaction volume of 10 µL. Samples were incubated for 30 min at 37°C and then returned immediately to ice. 1 µL of DNase Stop Buffer

(provided in the same kit) was added to each sample, and reactions were incubated at 65°C for 10 min to halt enzymatic activity of the DNase. Finally, first- strand cDNA was synthesized by mixing the entire DNase-treated reaction with 4

µL of qScript cDNA SuperMix (Quantabio, Beverly, MA, USA) and 5 µL of nuclease-free water followed by incubation at 25°C for 10 min, 42°C for 30 min, and then 85°C for 5 min.

Immune Gene Expression Analysis

Prior to immune gene expression analysis, all of the primers used for this study were first analyzed for amplification specificity and efficiency using endpoint PCR and RT-qPCR, respectively. For the former, a minimum of two experimental samples were utilized to conduct endpoint PCR using 2 µL of cDNA

123 template and an AccuStart II PCR SuperMix kit (Quantabio, Beverly, MA, USA) according to the manufacturer’s protocol. Amplification of a single gene product at the expected amplicon size was observed using agarose gel electrophoresis.

Primers were then assessed for amplification efficiency by conducting RT-qPCR on 1:10 dilutions of the same template cDNA. All primers used for this study were either recovered from published literature or manually designed by T. Sherman with the aid of PrimerQuest and OligoAnalyzer (IDT). Details of all primer sequences, amplification efficiencies, and gene accessions can be found in

Tables 4.1 and 4.2.

Following satisfactory internal validation measures, analysis of immune gene RT-qPCR was conducted by combining 2 µL of template cDNA (diluted to

10 ng/µL) with 5 µL of SsoAdvanced SYBR Supermix (BioRad, Hercules,

California, USA), 2 µL of forward and reverse primers each at 2 µM, and 1 µL of nuclease-free water for a final total reaction volume of 10 µL. All primer pairs were utilized at a final concentration of 200 nM except for the hamster primers for

IL-6 and IFN-γ, which were utilized at final concentrations of 400 nM and 100 nM, respectively (water volumes adjusted accordingly to retain a final total reaction volume of 10 µL). All samples were run on a CFX384 Touch thermocycler

(BioRad, Hercules, California, USA) with the following conditions: initial denaturation step at 95°C for 10 min followed by 40 cycles at 95°C for 15 s and

60°C for 1 min. Additionally, melt curves were generated for each sample, as described above. Relative gene expression was calculated using the comparative Ct method (i.e. ΔΔCt) as proposed by Livak and Schmittgen

124 whereby immune genes were first normalized against the constitutively- expressed β-actin gene (ΔCt) and then calibrated against the ΔCt of mock- infected controls (ΔΔCt) (252). Finally, relative gene expression was depicted as the mean fold change (2-ΔΔCt) ± standard error of the mean (SEM) between infected and mock-infected animals at each of the previously described collection timepoints. Samples that did not generate a Ct value were given a default value of 40.

Statistical Analysis

Log10 transformations were first applied to all relative expression values prior to any downstream statistical analysis. The statistical software SPSS v25.0

(IMB, Armonk, New York, USA) was used to conduct a MANOVA examining significant differences in the relative expression values of cytokine genes between species among select tissues at distinct collection timepoints. As well, separate MANOVA tests were run for each species of animal to examine differences in gene expression profiles within groups when tissues and collection timepoints are held as fixed factors. Independent samples t-tests were conducted using PRISM v8.3.1 (Graphpad, LaJolla, CA, USA) to determine whether the expression of particular genes differed significantly between species within a particular collection tissue and timepoint. Finally, Mann-Whitney tests were conducted in PRISM to determine whether viral loads differed significantly between species in a particular tissue over time. For all analyses, significant differences between groups were evaluated with an alpha significance of 0.05.

125

Table 4.1: Sequence data and amplification efficiencies for deer mouse primers.

Target Primer Sequence (5’ to 3’) Size R2 Eff. (%) Gene Accession Reference

β-Actin Mallarino et 123 1.000 102.0 XM_006998174.2 Fw GCTACAGCTTCACCACCACA al., 2016 (253) Rv TCTCCAGGGAGGAAGAGGAT Oas1b Fw CAGTATGCCCTGGAGCTGC 111 0.998 98.3 XM_006970848.1 This study Rv GTACTTGGTGACCAGTTCC IFN-γ Fw ACAGCAGTGAGGAGAAACGG 115 0.970 94.7 AY289494.1 This study Rv GACAGGCGGTACATCACTCC TNFα Fw GGGCTGTACCTCGTCTACTC 121 0.999 100.4 XM_006995235.2 This study Rv ACAGGAGGTTGACTTTGTCC TGFβ Schountz et Fw CGTGGAACTCTACCAGAAATACAGC 96 0.999 95.6 XM_006988036.2 al., 2012 (190) Rv TCAAAAGACAACCACTCAGGCG IL-6 Schountz et Fw CCATCCAACTCATCCTGAAAGC 101 0.999 96.1 AY256518.1 al., 2012 (190) Rv CCACAGATTGGTACACATAGGCAC IL-10 Fw CAGACCTACACGCTTCGAG 128 0.999 111.2 XM_006995328.2 This study Rv CCCAGGTAACCCTTAAAGTCC MODV-NS5 Leyssen et al., Fw CCAGGACAAGTCATGTGGTAGC 101 0.998 107.5 NC_003635.1 2001 (66) Rv TCCCAAAGATGTTCCTCACCTT able 3

126

Table 4.2: Sequence data and amplification efficiencies for hamster primers.

Target Primer Sequence (5’ to 3’) Size R2 Eff. (%) Gene Accession Reference

β-Actin Mallarino et al., Fw GCTACAGCTTCACCACCACA 123 1.000 102.4 XM_013120404.2 Rv TCTCCAGGGAGGAAGAGGAT 2016 (253) Oas1b Fw CAGTATGCCCTGGAGCTGC 111 0.999 103.5 XM_013119795.2 This study Rv GTACTTGGTGACCAGTTCC IFN-γ Espitia et al., Fw TGTTGCTCTGCCTCACTCAGG 130 1.000 104.2 AF034482.1 2014 (243) Rv AAGACGAGGTCCCCTCCATTC TNFα Espitia et al., Fw TGAGCCATCGTGCCAATG 79 0.998 97.7 XM_005086799.3 2014 Rv AGCCCGTCTGCTGGTATCAC TGFβ Zivcec et al., Fw TGTGTGCGGCAGCTGTACA 63 1.000 100.7 XM_013125593.2 2011 (234) Rv TGGGCTCGTGAATCCACTTC IL-6 Zivcec et al., Fw CCTGAAAGCACTTGAAGAATTCC 78 1.000 112.4 XM_005087110.2 2011 Rv GGTATGCTAAGGCACAGCACACT IL-10 Espitia et al., Fw GGTTGCCAAACCTTATCAGAAATG 194 1.000 99.5 XM_021232886.1 2014 Rv TTCACCTGTTCCACAGCCTTG MODV-NS5 Leyssen et al., Fw CCAGGACAAGTCATGTGGTAGC 101 0.998 107.5 NC_003635.1 2001 (66) Rv TCCCAAAGATGTTCCTCACCTT Table 4

Results

Deer Mice and Hamsters Exhibit Differential Modoc Virus- Induced Clinical Outcomes

Both deer mice and hamsters demonstrated susceptibility to intraperitoneal challenge with MODV although the clinical outcomes of infection differed. Over the 10 day experimental period, none of the deer mice demonstrated any distinguishable signs of morbidity as characterized by ruffled fur or piloerection, fluid discharge from the nose, mouth or eyes, hematuria, muscle weakness, tremor, or flaccid paralysis, general lethargy and changes in behavior or significant weight loss (Figure 4.1). In contrast, hamsters began to

127 exhibit gross organ pathologies beginning at day 4 post-infection in the form of and epididymitis/ in one of the three individuals euthanized at that timepoint (Figure 4.2). On day 7, one female (H16) was euthanized three days ahead of the predetermined schedule of serial euthanasia due to morbidity in the form of total limb paralysis, obvious tachypnea, crusted blood in one nare, and apparent incontinence. Two more hamsters developed similar signs (i.e. extreme lethargy, weepy, squinting eyes with negligibly intact corneal reflexes and quivering front limbs) of disease by day 8 and were euthanized immediately

(which happened to coincide with the predetermined schedule). Finally, by day

10, the last two remaining hamsters presented with extreme lethargy, slow/weak reflexes in the front limbs, splenomegaly, orchitis, and extreme weight loss. Thus,

6/18 (33.3%) infected hamsters demonstrated some form of gross anatomical pathology or clinical manifestation of disease within the 10 day experimental period. For both the deer mice and hamsters, none of the uninfected controls presented with any signs of illness throughout the investigation.

128

Figure 17

Figure 4.1: Weight gain or loss in deer mice (left) and golden hamsters (right) over the experimental period. Each data point represents the weight for a particular infected individual at that timepoint or the mean (+/- SEM) weight of uninfected individuals at the same timepoint (n = 2). Of particular note is the rapid loss of weight in one hamster (H16) which corresponds with clinical signs of disease and subsequent early termination.

129

Figure 18

Figure 4.2: Signs of gross pathology in infected hamsters. Instances of epididymitis (left; day 4 post-infection) and orchitis (right; day 8 post-infection) observed in hamsters infected with MODV.

Modoc Virus Replication Kinetics Differ Among Various Tissues in Deer Mice and Hamsters

To assess for differences in viral replication kinetics between deer mice and hamsters, RT-qPCR was utilized to quantify the levels of MODV RNA expressed in select tissues and fluids over the experimental duration (Figure

4.3). To this end, it was discovered that MODV RNA is generally recovered at higher levels in the spleen and liver of infected deer mice as compared to the higher levels recovered from the sera, urine, brain, and kidneys of infected hamsters. Further, it was curious to find that in all of the tissues and fluids examined levels of MODV RNA began to increase by day 4 in hamsters while the same timepoint generally marked the beginning of decreased levels in deer mice

(save for the spleen and liver, which depicted a slight increase of viral RNA on days 4 and 6 post-infection, respectively).

More specifically, levels in the sera were similar at the first timepoint of sampling but began to diverge around day 4 post-infection. By day 6, peak levels

130 of MODV RNA were reached in hamster sera, which differed significantly from those found in deer mice. Interestingly, levels then begin to decline in hamsters around day 8 concomitant with a very slight increase in deer mice. In the urine, levels differed significantly at the first timepoint of collection and then increased in hamsters around day 4 while remaining relatively lower in deer mice (save for one individual on day 6). Likewise, RNA levels recovered from brain homogenates in deer mice were significantly higher at the first timepoint of collection; however, quantities of MODV RNA then spiked in the brains of two out of three individuals at both day 6 and 8 post-infection and remained higher than deer mice in all three hamster bioreplicates at day 10 post-infection. Infectious virus was recovered from at least two of three bioreplicates in the brains of infected deer mice on days 6, 8, and 10 post-infection but not from the sera (in total, 2/3 individuals on day 8 and 3/3 individuals on day 10) or urine (all individuals) of any deer mice. In the hamsters, no infectious virus was recovered from the brain or sera of any individuals, while infectious virus was recovered only from one individual on day 10 post-infection.

Following similar trends in the urine and brain, levels of viral RNA in the kidneys where initially significantly higher in the deer mice. By day 4, though, levels began to increase in the hamsters and reached a significant peak on day 8 before exhibiting a rapid decline. At all collection timepoints infectious virus could be recovered from at least two of three bioreplicates in hamster kidney homogenates (the same was not true of deer mice kidney homogenates, for which successful virus isolation occurred in only two of three bioreplicates on

131 days 8 and 10 post-infection). In the spleen, one of three hamster bioreplicates demonstrated high levels of MODV RNA on day 4. Besides this, levels remained marginally higher in deer mouse spleens with a peak on day 6 followed by a decline which remained significantly higher than the hamsters on day 10.

Infectious virus was recovered from at least two of three deer mouse bioreplicates on days 8 and 10 but not from any hamsters. Finally, levels of viral

RNA in the liver of deer mice remained consistently higher through the experimental period and displayed a significant difference on day 10 post- infection; however, while infectious virus could only be recovered from two of three deer mouse bioreplicates on days 8 and 10, viral isolation was successful for at least two of three hamster bioreplicates at all timepoints sampled. The above considered, it was found that the levels of MODV RNA ultimately only differed significantly (p ≤ 0.05) between species over time only in the liver and sera. It may be, though, that the high variance experienced between bioreplicates at nearly every timepoint could have contributed to a loss of significance that might otherwise be found were the sample size increased.

132

Figure 19

Figure 4.3: Viral RNA loads detected in select tissues and fluids of deer mice (squares) and hamsters (circles) over the experimental period. Viral RNA was quantified using two-step RT-qPCR and translated into TCID50 “equivalent titers.” Each data point represents the mean (+/- SEM) recovered from infected replicates at each timepoint (n = 3). Asterisks represent a significant difference in RNA loads between deer mice and hamsters at a particular timepoint as determined by Mann-Whitney tests. Data points with red squares or dots represent instances in which infectious virus was successfully isolated for at least two of three bioreplicates.

133

Expression of the Antiviral Cytokine Interferon-γ Differs Between Deer Mice and Hamsters

Given the predicted variance that may occur between gene expression levels, a log10 transformation was first applied to all relative expression values prior to any downstream statistical analysis. Once transformed, a Shapiro-Wilk test for normality confirmed that all relative expression values were normally distributed in deer mice (p ≥ 0.11) except for one instance (TGFβ in liver tissues; p = 0.04). Likewise, all relative expression values in hamsters were normally distributed following transformation (p ≥ 0.09) save for IFN-γ expression in the spleen (p = 0.02) and brain (p = 0.21) as well as the expression of IL-10 in the brain (p = 0.04).

To determine whether the observed profiles of IFN-γ, TNFα, IL-6, IL-10, and TGFβ expression were statistically meaningful, multivariate analysis of variance (MANOVA) tests were used to first examine gene expression in the deer mice, then in the hamsters, and finally in both species together. As well, independent samples t-tests were used to compare the mean levels of cytokine gene expression at specific timepoints between each species; however, multiple corrections for these t-tests were not included (Figure 4.4). For each species, it was discovered that there was indeed a significant difference in the profile of cytokine genes expressed between tissues over time (deer mice, p = 0.01; hamsters, p = 0.01). In the deer mice, between-subjects effects with corrected models revealed that IFN-γ, TNFα, and TGFβ were significantly differentially

134 expressed when both tissue and collection timepoint are treated as cofactors.

Curiously, the same analysis conducted with data collected from the hamsters demonstrated that IFN-γ, TNFα, IL-6, and IL-10 were significantly differentially regulated, but TGFβ was not. When a single MANOVA was used to compare the profiles of expressed genes between species, collection timepoints, and tissues, it was again found that there was a significant difference in the relative expression of all genes over time between tissues and species (p = 0.02). In this case, however, only the expression of IFN-γ differed significantly (p = 0.01).

When comparing the expression of individual genes between species, it was interesting to find that in some cases, the patterns of expression between tissues corresponded with the levels of MODV RNA detected. For example, expression of IFN-γ was initially higher for deer mice in the kidneys, spleen, and liver, although not the brain. Along with this, IFN-γ expression remained at an elevated state in deer mice relative to hamsters in the spleen and liver, both of which were found to contain marginally higher levels of viral RNA. In a similar fashion, higher relative expression of IFN-γ in the kidneys (past day 1) and brains of infected hamsters matches the elevated levels of viral RNA recovered from those same tissues. Indeed, it seems intuitive that antiviral cytokines such as

IFN-γ would be highly expressed in instances of higher viral loads; granted, these trends may also be highly tissue-dependent as suggested by examining one particular hamster which demonstrated much higher levels of both viral RNA and

IFN-γ on day 6 in the kidneys but only high levels of viral RNA (and the lowest expression of IFN-γ among biotriplicates) in the brain at the same timepoint.

135

Aside from IFN-γ, all other genes were found to be significantly differentially expressed between species; however, these cases occurred only in certain tissues and at particular collection timepoints. For example, expression of

TNFα differed significantly between deer mice and hamsters on day 1 in the kidneys and brain as well as day 6 in the kidneys and spleen but at no other timepoints in any tissues. IL-6 only differed significantly at day 10 in the kidneys.

Finally, IL-10 differed significantly in the brain on days 6 and 10. These results may implicate that while only IFN-γ was shown to be significantly differently expressed among all tissues between species over time, the differential expression of individuals genes is likewise a potentially important factor in determining the outcome of viral infection.

136

Figure 20

Figure 4.4: Cytokine gene expression levels as detected by RT-qPCR in select tissues from both deer mice and hamsters at distinct collection timepoints. Bars with asterisks denote a significant difference (p ≤ 0.05) in the expression levels between species at that particular timepoint of collection (uncorrected for multiple tests). Peromyscus maniculatus is represented by “PM” (black bars) while Mesocricetus auratus is represented by “MA” (grey bars).

137

The Interferon-Stimulated Gene Oas1b is Significantly Upregulated in Hamsters

Using a three-factor analysis of variance (ANOVA), it was discovered that the interferon-stimulated gene (ISG) Oas1b was significantly differentially expressed between deer mice and hamsters in select tissues at distinct collection timepoints (p = 0.03). Namely, relative expression values in the hamster kidney tissues at all timepoints, as well as in brain tissues at days 6 and 10 post- infection, were much higher than those of deer mice at the same collection timepoints. In contrast, relative expression values in deer mouse liver tissues was significantly higher at days 1 and 6 but not at day 10.

Figure 21

Figure 4.5: Expression levels of Oas1b in deer mouse and hamster tissues. Expression levels of the antiviral effector Oas1b as detected by RT-qPCR in select tissues from both deer mice and hamsters at distinct collection timepoints. Bars with asterisks denote a significant difference (p ≤ 0.05) in expression levels between species at that particular timepoint of collection (uncorrected for multiple tests). Peromyscus maniculatus is represented by “PM” (black bars) while Mesocricetus auratus is represented by “MA” (grey bars).

Discussion

The research presented here attempts to take a comparative approach to elucidate the profile of cytokine genes expressed in previously described pathogenic and non-pathogenic models of flaviviral encephalitis (7, 8, 96). Using

138

MODV, a BSL-2, rodent-borne flavivirus, deer mice and hamsters were intraperitoneally challenged and serially euthanized to capture the expression of five cytokine genes (interferon-gamma, IFN-γ; alpha, TNFα; interleukin-6, IL-6; interleukin-10, IL-10; and transforming growth factor beta,

TGFβ; Figure 4.4) and one antiviral effector gene (2’-5’ oligoadenylate isoform

1b, Oas1b; Figure 4.5) across four tissues (kidneys, spleen, liver, and brain) at three distinct timepoints (1, 6, and 10 days post-infection). Further, levels of

MODV RNA were quantified within the same tissues at each of the aforementioned timepoints in addition to 2, 4, and 8 days post-infection (Figure

4.3).

With regard to clinical presentation, no deer mice infected with MODV demonstrated any signs of gross anatomical pathology or overt disease throughout the entire experimental duration. These findings both corroborate and contribute to the very limited handful of previous investigations of deer mice experimentally infected with MODV by demonstrating that even at the higher-titer challenge, the deer mice are able to resist pathology (8, 254). In contrast, 6/18

(33.3%) infected hamsters demonstrated some form of pathology similar to the findings of previous studies (7, 66, 96). In one such case, Adams et al. describe an experiment in which two hamsters developed bilateral hindlimb paralysis or complete paralysis with intact corneal reflexes by days 6 and 7 post-infection, respectively (7). During a separate investigation, Leyssen et al. discovered that approximately half of MODV-infected hamsters developed (clinically-confirmed) signs of neurologic disease reminiscent of other encephalitic flaviviruses (e.g.,

139

WNV and JEV) within 10 days post-infection (66). Taken together, these studies and the research presented here support the notion of utilizing hamsters and deer mice infected with MODV as comparative models of flaviviral encephalitic disease or resistance, respectively, and provide the impetus to explore the differential expression of immune response genes to help elucidate these opposing outcomes.

Past investigations suggest that MODV may differentially replicate in the spleen, lungs, and salivary-submaxillary glands of deer mice while appearing more trophic toward the kidneys and brain of infected hamsters (7, 8, 65, 66, 96,

254). Indeed, both viremia and viruria are possible to varying degrees in each species; however, no studies thus far have directly compared the viral loads in each model using quantitative methods when both dosage route and infectious titer are standardized. Having this information may help to further clarify the tissue tropism and replication kinetics of MODV in these animals while further providing clues as to how the regional cytokine milieu may influence the outcome of local or systemic pathologies. One of the aims of this research was therefore to complement previous studies by resolving this gap in knowledge.

To this end, we found that MODV RNA was generally higher in the spleen and liver of deer mice as compared to hamsters. In both organs, these levels correspond with higher expression of IFN-γ and TNFα in deer mice compared to hamsters at all collection timepoints. Indeed, it is expected that higher levels of antiviral, proinflammatory cytokines such as these would be associated with tissues that contain higher burdens of replicating virus. However, it is interesting

140 that the levels of IL-6, another proinflammatory cytokine, do not likewise follow these trends. In contrast, mean expression levels of IL-6 remain lower in the spleen and of infected deer mice at 1 and 6 days post-infection but are then higher than the expression levels observed in hamsters for both organs by day 10. In any case, infectious virus was only recovered from at least two of three bioreplicates on days 8 and 10 in the spleen and liver of infected deer and from two of three bioreplicates at all collection timepoints in the livers of hamsters. No successful virus isolations were obtained from the spleens of infected hamsters.

Generally, higher levels of viral RNA were recovered from the kidneys of infected hamsters, a finding that is consistent with previous research demonstrating that MODV displays a trophism toward renal tissues in hamsters and may persist for long periods of time (e.g., up to 14 months post-infection) (7,

65, 92). As well, infectious virus could be recovered from at least two of three hamster bioreplicates at all timepoints examine but only on days 8 and 10 with the deer mice. Unlike gene expression trends observed in the spleen and livers of deer mice, significantly higher relative expression of TNFα is evident by day 1, whereas higher levels of IFN-γ do not occur until day 6. Moreover, while IFN-γ is maintained at higher levels than in deer mice until at least day 10, expression of

TNFα drops significantly by day 6 before increasing slightly by day 10. IL-6 follows a similar trend as TNFα whereby expression is first higher than that of deer mice, decreases by day 6, and is then significantly higher by day 10. The implications that these findings in the kidney have on subsequent viruria are

141 unclear. Similar to the levels of MODV RNA observed in the kidneys, though, viral loads are initially higher in the deer mice before falling below that of the hamsters (save for one individual on day 6) for the remainder of the experiment.

Nevertheless, the fact that infectious virus was only recovered from one individual (hamster) on day 10 post-infection could be due to an initial replication eclipse period that needs to be overcome before sufficient infectious virus is recoverable; this notion is supported by previous investigations with MODV in which infectious virus is not successfully isolated from the urine until past day 7 following infection via subcutaneous or intranasal route (7, 65)

Expression of IFN-γ in the brains of hamsters was elevated above that of deer mice on all days. In particular, extreme differences were observed by day 6 and were significant by day 10. Despite the finding that levels of viral RNA were initially significantly higher in deer mice, levels of viral RNA in hamsters begin to rise rapidly by day 2, surpassing levels in the deer mice by day 4 and remaining higher for the remainder of the experiment. This observation matches previous studies that demonstrate a neurotropism of MODV in hamsters and supports the high levels of IFN-γ expression in these tissues (7, 66, 96). It is curious, though, that the levels of MODV RNA recovered from the brains of hamsters displayed extreme variance concomitant with the levels of variance observed between expressions of IFN-γ. For example, on day 6, viral RNA was most highly expressed in one particular hamster, which also expressed the fewest transcript copies of IFN-γ. This stands in contrast to the data collected from the same individual in kidney tissues, whereby both the levels of IFN-γ and viral RNA were

142 the highest among the three bioreplicates examined. Expression of TNFα was likewise higher on days 1 and 6 but not by day 10, whereas IL-6 was more highly expressed in hamsters than deer mice on day 6 but not days 1 and 10. Finally, the levels of IL-10 expressed in the brains of infected hamsters notably exceeded those observed in the deer mice at all timepoint and were significantly higher on days 6 and 10.

It has been well established that many flaviviruses antagonize the type I interferon (IFN) antiviral signaling system at multiple levels and may subsequently inhibit the expression of intracellular, antiviral effector proteins

(123, 126, 267, 268). However, due to the absence of available sequences for deer mouse and hamster type I IFN transcripts, we choose to examine expression of Oas1b as a representative IFN-stimulated gene (ISG) to quantify downstream type I IFN signaling (269). The Oas (2’-5’ oligoadenylate) family of proteins generally encode enzymes capable of activating latent cytosolic ribonuclease L (RNAse L), which acts to degrade viral and host transcripts.

Although Oas1b lacks this enzymatic activity and therefore exerts antiviral activity through other, incompletely understood mechanisms, it has nevertheless been demonstrated to be protective against flaviviral infections such as WNV (270-

273). For this study, primers for the Oas1b gene were designed using homologous sequences found in regions of the Oas1b coding transcripts from deer mice, golden hamsters, and Mus musculus. Using these primers, it was interesting to find that significantly higher expression of Oas1b appeared in the livers of infected deer mice as opposed to the kidneys and brains of infected

143 hamsters (Figure 4.5). These patterns roughly coincide with the apparent tropism of MODV toward these tissues as determined by the quantities of viral transcripts recovered from each species, although the same trends were not observed in the spleens of infected animals. Given these outcomes it would appear that MODV does not antagonize the type I IFN response in the same manner that is characteristic of some other flaviviruses. Further, the high expression levels of

Oas1b by 24 hours post-infection confirms that both species are capable of recognizing MODV viral antigens and initiating expression of ISGs within a timeframe similar to that induced by other flaviviruses such as WNV (125).

Despite these similarities, it is curious that the expression levels in the kidneys and brains of hamsters far exceed those in deer mice at any timepoint except day 1 post-infection.

IFN-γ (also known as type II IFN) is a multifunctional, pleiotropic cytokine that is important in the fight against viruses among various other pathogens

[citations]. It has been cited as one of the most potent activators of macrophages

(particularly those with the classical, or M1 phenotype) and may also induce nitric oxide (NO) production contaminant with the upregulation of major histocompatibility complexes (MHC) I and II, all of which aid in the response against intracellular pathogens (287-289). Indeed, the polarization of immune effectors toward a phenotype which better targets intracellular invaders

(historically referred to as the TH1 phenotype) is in large part accomplished by the expression of IFN-γ from activated cells such as helper T lymphocytes (TH) and natural killer (NK) cells (287, 288). Similarly, TNFα serves myriad roles in

144 antimicrobial immunity as well as normal physiologic homeostasis (264). As an antiviral secreted from activated T cells and macrophages, TNFα elicits fever, stimulates the expression of acute-phase liver proteins, and may induce apoptosis in infected endothelial cells (264). Finally, like IFN-γ and TNFα, IL-6 is another important proinflammatory mediator which may be released from many cell types, including macrophages and T cells upon stimulation from cytokines like TNFα (290). As an antiviral, IL-6 further contributes to the febrile response and may recruit cytotoxic neutrophils to sites of active infection as well as promoting B cell differentiation and subsequent antibody production (291, 292).

In contrast to the above three cytokines, IL-10 and TGFβ have traditionally been viewed as immunomodulatory or even immunosuppressive, the expression of each presumed to mollify sustained inflammatory responses prior to the onset of immunopathogenic collateral damage (293). The mechanisms through which these two cytokines function are diverse and incompletely understood, though examples of putative roles include the suppression of IFN signaling (IL-10) or the induction of T regulatory (Treg) cell-mediated tolerance (293, 294).

The above considered, the profile of cytokines expressed between deer mice and hamsters across tissues and over time during infection with MODV paints an intriguing picture with regard to the outcomes of clinical disease. As was determined by statistical analysis, IFN-γ appears to exert a particularly important role in coordinating the early innate immune response as this was the sole cytokine to demonstrate significance (p = 0.01) when tissue, species, and timepoint of collection are all held as fixed factors. Moreover, the sheer quantity

145 of IFN-γ transcripts detected between deer mice and hamsters (namely in the kidneys and brain) elicit curiosity for the differential mechanisms which control the expression of cytokines between these two species. Although the exact antiviral mechanisms of IFN-γ have only recently been more clearly elucidated

(largely due to the redundant and overlapping downstream effects induced by other IFNs), it has been demonstrated repeatedly that the protective levels of

IFN-γ exist within a range such that over- or under-expression may lead to deleterious clinical outcomes (295). For example, it has been shown that intact

IFN-γ signaling is required to promote the protective roles of innate γδ T cells and primary dendritic cells against infection with WNV (262). On the other hand, overexpression of IFN-γ may lead to the upregulation of adhesion molecules such as intercellular adhesion molecule-1 (ICAM-1), a putative binding receptor for many flaviviruses (126, 146, 296). However, the consequences of this induced upregulation of ICAM-1 may be flavivirus-specific, as infection by some flaviviruses (e.g., WNV) appear to upregulate these receptors in a manner independent of IFN-γ expression (297). In any case, it could be that discrepancies in the activation of innate pattern recognition receptors (PRRs) by

MODV antigens in deer mice and hamsters leads to differential cytokine responses, an event that has been documented in other flaviviral systems (26,

174, 298).

The expression of IL-10 in the brains of infected hamsters may be a point of interest for further investigation given its increased expression at all timepoints relative to deer mice, the magnitude of expression on days 6 and 10, and

146 previous research which has correlated this particular cytokine in both protective and adverse outcomes of flaviviral infection (namely DENV) (299, 300). Like IFN-

γ, though, regulation of this cytokine may be highly virus (i.e. serotype) specific, as some research has shown IL-10 to be elevated only for DENV-2 infection while apparently being downregulated during infection by DENV-1, 3, and 4

(301). Nevertheless, multiple studies have suggested that excessive production of IL-10 (as may be witnessed here) could inhibit the protective effects of T cells, alternatively favoring viral persistence (302-304). To this end, the expression levels of IL-10 in this study are significant when only tissue and species are treated as fixed factors in a multivariate analysis (p ≤ 0.00). This may add support to the notion that regulation of this cytokine does indeed play an antagonistic role in the productive control of MODV in brains of infected hamsters.

It is unclear at this time how the patterns of TNFα, IL-6, and TGFβ influence MODV replication in either deer mouse or hamster tissues. For all of these, the relative lack of variation in expression suggests that their role under the parameters of this experiment is less important; however, as there were some instances of statistical significance, further investigation may be warranted.

As well, assays that examine the circulating protein levels of these cytokines could be telling, particularly for TGFβ, which is less transcriptionally regulated in mammalian cells (293, 305).

147

Conclusions and Future Directions

This study utilizes two unique rodent models to directly compare the host response, viral tropism, and cytokine gene expression profiles in select tissues following infection with a NKV flavivirus. As the putative reservoir hosts of MODV, deer mice are presumed to share a unique relationship with their viral counterparts in which a delicate balance has developed between the host immune response and viral infection (185). This relationship between a virus and its reservoir holds valuable information when attempting to further understand the molecular mechanisms in place that prevent illness in the reservoir but lead to disease when non-reservoir species (such as humans) become infected.

Alternatively, hamsters offer a susceptible and pathogenic model in which the molecular underpinnings of disease development can theoretically bed elucidated when the immune response is compared to that of a non-pathogenic model, and the parameters of experimental infection are standardized.

It is known that the cytokine response is critical to the outcome of flaviviral clearance, persistence, or clinical pathogenesis (150, 265, 306-309). This study superficially demonstrates that importance by highlighting the involvement of one such cytokine, IFN-γ, which was shown to be significantly differentially expressed across tissues and over time even when variation between individuals was high.

Findings like these may help to provide a link between the early innate immune response and the subsequent outcome of viral infection, both in terms of controlling immediate viral burdens as well as eliciting a productive adaptive response.

148

It is our hope that this study will further provide the impetus to continue work with MODV as a BSL-2 model of flaviviral encephalitis given the contributions depicted here in accordance with previous research conducted. For example, Adams et al. demonstrated that MODV-infected hamsters displayed lesions in the central nervous system (CNS) similar to those described for WNV- infected hamsters (7). However, work by Adams et al. did not include an investigation of any cytokines expressed during acute infection, a factor which may strongly influence viral tropism; in this way we hope that our research will complement what has been conducted in the past.

Another example of the potential to utilize MODV in comparative analyses against other, more medically-pertinent representatives involves our finding that

MODV may display a tropism toward reproductive tissues (Figure 4.2). This could be useful given that the recently acclaimed ZIKV has been shown to infect reproductive tissues in otherwise healthy adults and developing fetuses (153). As well, using MODV may provide an feasible avenue for further understanding differences in host sex, which may influence the outcome of viral disease. In our study, the individuals which demonstrated neurologic involvement (via limb paralysis) were consistently female. As it has been additionally demonstrated that Norway infected with Seoul virus (SEOV) demonstrate differential innate responses dependent upon host sex, it could be that continued investigation of these instances could help better guide the development of antiviral therapies administered to either human males or females (310).

149

Finally, future research may benefit from using MODV to further investigate the aspects between host and viral, which lead to persistence following the initiation of the innate response. For example, previous studies with

MODV have demonstrated that the virus will persist in the kidneys of infected hamsters for long periods of time in a manner similar to that of WNV and St.

Louis encephalitis virus (SLEV) (311, 312). Neither of these latter studies, however, included the examination of cytokines, which may allow for viral persistence, something which we hope could be supplemented by studies like ours.

150

CHAPTER V

CONCLUSIONS

Flaviviruses (family Flaviviridae) comprise a taxonomic collective of globally-distributed viruses that may elicit a vast range of clinical pathologies. In particular, members of the genus Flavivirus constitute a significant portion of all arthropod-borne viruses (i.e. arboviruses), viruses that are carried and transmitted to new susceptible hosts by insects such as ticks and mosquitoes.

For perspective, dengue virus (DENV), now considered to be the most common in the world, infects more than 390 million individuals annually (95% confidence interval) across 129 countries and leads to clinical disease in approximately 96 million and death in nearly 40,000 (10, 279, 313). Other flaviviruses of medical relevance include Japanese encephalitis virus (JEV), yellow fever virus (YFV), Zika virus (ZIKV), and West Nile virus (WNV), all of which are vectored by the prevalent Culex or Aedes genera of mosquitoes. Tick- borne encephalitis virus (TBEV) and Powassan virus (POWV) are additionally of concern and are transmitted by ticks such as those of the Iodexes complex.

Unfortunately, there are no FDA-licensed antivirals for any of the known flaviviruses. While vaccines exist for some (e.g., YFV and JEV), they are often expensive or entirely inaccessible to the populations that need them most (those in tropical, underdeveloped nations). Moreover, ongoing changes to global climate have led to the range expansion of ticks and mosquitoes, a factor which

151 is compounded by other anthropogenic practices such as waste plastic and tire production (1, 84, 85, 314). Indeed, the Centers for Disease Control and

Prevention (CDC) have estimated that diseases caused by arthropods in the US have tripled in the last 13 years (Figure 5.1) (315). Likewise, although not yet endemic in the US, it has been proposed that global cases of DENV infection have increased 30-fold in the last 30 years (316). In light of this, there exists a pertinent need for continued research with flaviviruses to better inform the science behind disease prevention and mitigation.

Figure 22 Figure 5.1: Increased prevalence of vector-borne diseases in the US over the last 13 years. Recent years have witnessed an alarming increase in the infection incidence of diseases carried by arthropod vectors such as ticks, mosquitoes, and . Although there are many factors that may influence this trend, it is believed that warming global temperatures have led to increased ranges for the insects that vectors diseases such as ZIKV, WNV, and DENV (315).

152

Although Modoc virus (MODV) strictly infects vertebrate hosts and is therefore not considered an arbovirus, it has nevertheless been classified as a member of the Flavivirus genus in light of the virion structure, viral genome organization, and replication strategies it employs (17). As a biosafety level-2

(BSL-2) virus, MODV has been utilized in attempts to elucidate the discrete characteristics which promote or restrict replication in vertebrate or invertebrate hosts, respectively (52, 98, 100). While MODV is not typically considered a concern to general public health, both disease and subclinical infection have been documented in humans (6, 101). Further, a small handful of studies have used MODV to investigate aspects of neurologic disease in Syrian golden hamsters, a rodent model that recapitulates instances of flaviviral encephalitis such as those caused by JEV, WNV, or St. Louis encephalitis virus (SLEV) (95,

96). Finally, there has been interest in the use of deer mice to better understand the immunological underpinnings which allow for this rodent to act as a reservoir host for MODV as well as a slew of other human viral and bacterial pathogens without the development of any apparent disease (188, 190, 210, 215, 246, 317).

For these reasons, the research presented here attempts to create a novel model for the investigation of flaviviral disease by directly comparing the host response, viral replication kinetics, and subsequent immune gene expression in deer mice and golden hamsters acutely infected with MODV. This research is novel in that it is the first to investigate the profile of immunological signals (i.e. cytokine and effector gene transcripts) expressed during infection with MODV concomitant with the patterns of viral replication in select host tissues over time.

153

Results from this research corroborate previous studies demonstrating that experimental infection of deer mice with MODV is non-pathogenic while hamsters experience clinical signs of disease similar to human cases of flavivirus illness with neurologic involvement (e.g., rapid weight loss followed by paralytic syndromes). Analysis with RT-qPCR indicates that MODV RNA can be recovered from the kidneys, spleen, liver, brain, urine, and sera of all intraperitoneally-infected animals as early as 24 hours post-infection and persists until at least 10 days post-infection. While higher average levels of viral RNA were observed in the spleen and liver of deer mice, higher quantities were recovered from the kidneys, brain, sera, and urine of hamsters. Along with this, infectious virus could be recovered from all tissues (but not the sera or urine) of deer mice by day 8 post-infection; this finding is expected given that MODV may develop persistent and transferrable infections in deer mice under natural conditions as the putative reservoir hosts. On the other hand, infectious virus could be recovered from the kidneys and liver of multiple hamsters at all timepoints sampled, suggesting that acute infection in less controlled in these rodents as compared to deer mice.

The examination of select cytokine transcripts in the same tissues of infected animals depicts an inflammatory response characteristic of those anticipated by the induction of viral pathogens (e.g., elevation in cytokines such as IFN-γ which drive polarization of an antiviral TH1 response). In contrast to one another, however, the quantity and pattern of transcripts expressed between deer mice and hamsters varied radically, especially in brain tissues. Of note, IFN-

154

γ was shown to be significantly differentially expressed when tissue type, collection timepoint, and host species are treated as fixed factors. This may warrant further investigation given that IFN-γ has been implicated in both protective as well as deleterious outcomes of flaviviral infection (262, 263). As well, the increased expression of IL-10 at all timepoints in the brains of infected hamsters relative to deer mice is curious given that expression levels are statistically significant when only tissue type and host species are treated as fixed factors and that, like IFN-γ, this immunomodulatory cytokine has correlated with both positive and negative outcomes of viral disease (143, 300).

Future research utilizing the models presented here may wish to include a more complete examination of the cytokine (and chemokine) genes expressed during infection, perhaps through the use of RNA sequencing and transcriptomic analyses. Additionally, it would be important to include assays which examine the levels of translated proteins corresponding to expressed genes of interest; however, this may be difficult given that few reagents (e.g., antibodies) are commercially available for deer mouse and hamster proteins which are poorly cross-reactive with those of Mus and Rattus species (234). Alternatively, there is evidence to suggest that other non-proteinaceous inflammatory (or anti- inflammatory) mediators could influence the outcome of infectious disease. To this end, lower serum levels of the lipid messenger sphingosine 1-phosphate have been correlated with decreased vascular barrier integrity and increased disease presentation during infection with DENV (318). Likewise, mediators generated from polyunsaturated fatty acids (e.g., resolvins and protectins) have

155 been demonstrated to aid in mitigation of the inflammatory response; however, how these molecules function in the context of flaviviral disease is not yet clear

(319).

Taken together, the findings reported in this dissertation describe a system in which differences in the kinetics of viral replication, as well as the subsequent host immune response, can be directly compared in two unique rodent models of infection. In it of itself, a system like this may help to further clarify the immunological paradigms which influence the outcome of clinical disease during infection with representative zoonotic RNA viruses. Moreover, there is hope that research such as this may contribute to the greater scientific community’s understanding of flaviviral disease and the differential host factors which may either prevent or promote overt pathologies. In particular, the assessment of cytokine profiles during acute durations of infection may help to better define parameters of disease or aid in the development of vaccines which induce robust and durably protective immunity.

156

REFERENCES

1. Ross J (2016) What the rise of Zika (and other viruses) might tell us about our planet. in Harvard Health Publications (Harvard University, Online).

2. Centers for Disease C (2019) West Nile Virus Disease Cases by State 2018.

3. CDC (2018) Zika Virus Case Count in the US. (Centers for Disease Control and Prevention, Atlanta, Georgia).

4. Organization WH (2016) Dengue and severe dengue. (World Health Organization, Geneva, Switzerland).

5. Guzman MG & Harris E (2015) Dengue. Lancet 385(9966):453-465.

6. Zarnke RL & Yuill TM (1985) Modoc-like virus isolated from wild deer mice (Peromyscus maniculatus) in Alberta. J Wildl Dis 21(2):94-99.

7. Adams AP, Travassos da Rosa AP, Nunes MR, Xiao SY, & Tesh RB (2013) Pathogenesis of Modoc virus (Flaviviridae; Flavivirus) in persistently infected hamsters. Am J Trop Med Hyg 88(3):455-460.

8. Davis JW, Hardy JL, & Reeves WC (1974) Modoc viral infections in the deer mouse Peromyscus maniculatus. Infect Immun 10(6):1362-1369.

9. Staples JE & Monath TP (2008) Yellow fever: 100 years of discovery. JAMA 300(8):960-962.

10. Collins MH & Metz SW (2017) Progress and Works in Progress: Update on Flavivirus Vaccine Development. Clin Ther 39(8):1519-1536.

11. Simmonds P, et al. (2017) ICTV Virus Taxonomy Profile: Flaviviridae. J Gen Virol 98(1):2-3.

12. Westaway EG, et al. (1985) Flaviviridae. Intervirology 24(4):183-192.

13. Huang YJ, Higgs S, Horne KM, & Vanlandingham DL (2014) Flavivirus- mosquito interactions. Viruses 6(11):4703-4730.

157

14. Smith DB, et al. (2016) Proposed update to the taxonomy of the genera Hepacivirus and Pegivirus within the Flaviviridae family. J Gen Virol 97(11):2894-2907.

15. Tautz N, Tews BA, & Meyers G (2015) The Molecular Biology of . Adv Virus Res 93:47-160.

16. Firth C, et al. (2014) Detection of zoonotic pathogens and characterization of novel viruses carried by commensal Rattus norvegicus in . MBio 5(5):e01933-01914.

17. Kuno G, Chang GJ, Tsuchiya KR, Karabatsos N, & Cropp CB (1998) Phylogeny of the genus Flavivirus. J Virol 72(1):73-83.

18. Villordo SM, Carballeda JM, Filomatori CV, & Gamarnik AV (2016) RNA Structure Duplications and Flavivirus Host Adaptation. Trends Microbiol 24(4):270-283.

19. Edgil D, Diamond MS, Holden KL, Paranjape SM, & Harris E (2003) Translation efficiency determines differences in cellular infection among dengue virus type 2 strains. Virology 317(2):275-290.

20. Vratskikh O, et al. (2013) Dissection of antibody specificities induced by yellow fever vaccination. PLoS Pathog 9(6):e1003458.

21. Rey FA, Heinz FX, Mandl C, Kunz C, & Harrison SC (1995) The envelope glycoprotein from tick-borne encephalitis virus at 2 A resolution. Nature 375(6529):291-298.

22. Zhang X, et al. (2017) Structures and Functions of the Envelope Glycoprotein in Flavivirus Infections. Viruses 9(11).

23. Perera-Lecoin M, Meertens L, Carnec X, & Amara A (2013) Flavivirus entry receptors: an update. Viruses 6(1):69-88.

24. Laureti M, Narayanan D, Rodriguez-Andres J, Fazakerley JK, & Kedzierski L (2018) Flavivirus Receptors: Diversity, Identity, and Cell Entry. Front Immunol 9:2180.

25. Chen Y, et al. (1997) Dengue virus infectivity depends on envelope protein binding to target cell heparan sulfate. Nat Med 3(8):866-871.

26. Fernandez-Garcia MD, Mazzon M, Jacobs M, & Amara A (2009) Pathogenesis of flavivirus infections: using and abusing the host cell. Cell Host Microbe 5(4):318-328.

158

27. Chu JJ & Ng ML (2004) Interaction of West Nile virus with alpha v beta 3 mediates virus entry into cells. J Biol Chem 279(52):54533-54541.

28. Medigeshi GR, Hirsch AJ, Streblow DN, Nikolich-Zugich J, & Nelson JA (2008) West Nile virus entry requires cholesterol-rich membrane microdomains and is independent of alphavbeta3 integrin. J Virol 82(11):5212-5219.

29. Lozach PY, et al. (2005) Dendritic cell-specific intercellular adhesion molecule 3-grabbing non-integrin (DC-SIGN)-mediated enhancement of dengue virus infection is independent of DC-SIGN internalization signals. J Biol Chem 280(25):23698-23708.

30. Kaufmann B & Rossmann MG (2011) Molecular mechanisms involved in the early steps of flavivirus cell entry. Microbes Infect 13(1):1-9.

31. Clyde K & Harris E (2006) RNA secondary structure in the coding region of dengue virus type 2 directs translation start codon selection and is required for viral replication. J Virol 80(5):2170-2182.

32. Fields BN, Knipe DM, & Howley PM (2007) Fields virology (Wolters Kluwer Health/Lippincott Williams & Wilkins, ).

33. Westaway EG, Mackenzie JM, & Khromykh AA (2002) Replication and gene function in . Curr Top Microbiol Immunol 267:323-351.

34. Mackenzie JM, Jones MK, & Young PR (1996) Improved membrane preservation of flavivirus-infected cells with cryosectioning. J Virol Methods 56(1):67-75.

35. Welsch S, et al. (2009) Composition and three-dimensional architecture of the dengue virus replication and assembly sites. Cell Host Microbe 5(4):365-375.

36. Mackenzie JM, Khromykh AA, & Parton RG (2007) Cholesterol manipulation by West Nile virus perturbs the cellular immune response. Cell Host Microbe 2(4):229-239.

37. Khromykh AA, Varnavski AN, Sedlak PL, & Westaway EG (2001) Coupling between replication and packaging of flavivirus RNA: evidence derived from the use of DNA-based full-length cDNA clones of Kunjin virus. J Virol 75(10):4633-4640.

38. Kiermayr S, Kofler RM, Mandl CW, Messner P, & Heinz FX (2004) Isolation of capsid protein dimers from the tick-borne encephalitis flavivirus and in vitro assembly of capsid-like particles. J Virol 78(15):8078-8084.

159

39. Op De Beeck A, Rouille Y, Caron M, Duvet S, & Dubuisson J (2004) The transmembrane domains of the prM and E proteins of yellow fever virus are endoplasmic reticulum localization signals. J Virol 78(22):12591- 12602.

40. Li D & Aaskov J (2014) Sub-genomic RNA of defective interfering (D.I.) dengue viral particles is replicated in the same manner as full length genomes. Virology 468-470:248-255.

41. Li D, et al. (2011) Defective interfering viral particles in acute dengue infections. PLoS One 6(4):e19447.

42. Mackenzie JM & Westaway EG (2001) Assembly and maturation of the flavivirus Kunjin virus appear to occur in the rough endoplasmic reticulum and along the secretory pathway, respectively. J Virol 75(22):10787- 10799.

43. Hanna SL, et al. (2005) N-linked glycosylation of west nile virus envelope proteins influences particle assembly and infectivity. J Virol 79(21):13262- 13274.

44. Mason PW (1989) Maturation of Japanese encephalitis virus produced by infected mammalian and mosquito cells. Virology 169(2):354- 364.

45. Donadieu E, et al. (2013) Differential virulence and pathogenesis of West Nile viruses. Viruses 5(11):2856-2880.

46. Stadler K, Allison SL, Schalich J, & Heinz FX (1997) Proteolytic activation of tick-borne encephalitis virus by furin. J Virol 71(11):8475-8481.

47. Lindenbach B, Thiel HJ, & Rice CM (2007) Flaviviridae: The viruses and their replication. Fields Virology:1101-1151.

48. Pierson T (2012) Major Areas of Research. in Major Areas of Research (NIH, National Institute of Allergy and Infectious Disease).

49. Shi PY (2014) Structural biology. Unraveling a flavivirus enigma. Science 343(6173):849-850.

50. Bolling BG, Weaver SC, Tesh RB, & Vasilakis N (2015) Insect-Specific Virus Discovery: Significance for the Arbovirus Community. Viruses 7(9):4911-4928.

160

51. de Borba L, et al. (2019) RNA Structure Duplication in the Dengue Virus 3' UTR: Redundancy or Host Specificity? MBio 10(1).

52. Saiyasombat R, Carrillo-Tripp J, Miller WA, Bredenbeek PJ, & Blitvich BJ (2014) Substitution of the premembrane and envelope protein genes of Modoc virus with the homologous sequences of West Nile virus generates a chimeric virus that replicates in vertebrate but not mosquito cells. Virol J 11:150.

53. Goddard LB, Roth AE, Reisen WK, & Scott TW (2002) Vector competence of California mosquitoes for West Nile virus. Emerg Infect Dis 8(12):1385- 1391.

54. Kuwata R, et al. (2012) Establishment and characterization of a cell line from the mosquito (Diptera: Culicidae). In Vitro Cell Dev Biol Anim 48(6):369-376.

55. St Leger J, et al. (2011) West Nile virus infection in , Texas, USA, 2007. Emerg Infect Dis 17(8):1531-1533.

56. Strand TM, Lundkvist A, Olsen B, & Gustafsson L (2018) Breeding consequences of flavivirus infection in the collared flycatcher. BMC Evol Biol 18(1):13.

57. Hoogstraal H, Clifford CM, Keirans JE, Kaiser MN, & Evans DE (1976) The Ornithodoros (Alectorobius) capensis group (Acarina: Ixodoidea: Argasidae) of the palearctic and oriental regions. O. (A.) maritimus: identity, marine hosts, virus infections, and distribution in western Europe and northwestern Africa. J Parasitol 62(5):799-810.

58. L'Vov D K, et al. (2002) [Isolation of the virus from the great cormorant Phalacrocorax carbo, the crow corone, and Hyalomma marginatum ticks associated with them in natural and synanthroic biocenosis in the Volga delta (Astrakhan region, 2001)]. Vopr Virusol 47(5):7-12.

59. Organization WH (2015) Japanese encephalitis. (World Health Organization, Geneva, Switzerland).

60. Monath TP (2001) Yellow fever: an update. Lancet Infect Dis 1(1):11-20.

61. Vorou R (2016) Zika virus, vectors, reservoirs, amplifying hosts, and their potential to spread worldwide: what we know and what we should investigate urgently. Int J Infect Dis 48:85-90.

161

62. Back AT & Lundkvist A (2013) Dengue viruses - an overview. Infect Ecol Epidemiol 3.

63. Bolling BG, et al. (2015) Insect-specific viruses detected in laboratory mosquito colonies and their potential implications for experiments evaluating arbovirus vector competence. Am J Trop Med Hyg 92(2):422- 428.

64. Blitvich BJ & Firth AE (2017) A Review of Flaviviruses that Have No Known Arthropod Vector. Viruses 9(6).

65. Davis JW & Hardy JL (1974) Characterization of persistent Modoc viral infections in Syrian hamsters. Infect Immun 10(2):328-334.

66. Leyssen P, et al. (2001) A novel model for the study of the of flavivirus infections using the Modoc virus. Virology 279(1):27-37.

67. Johnson HN (1967) Ecological implications of antigenically related mammalian viruses for which arthropod vectors are unknown and avian associated soft tick viruses. Jpn J Med Sci Biol 20 Suppl:160-166.

68. Paulsen KM, et al. (2019) Tick-borne encephalitis virus in cows and unpasteurized cow milk from Norway. Zoonoses Public Health 66(2):216- 222.

69. Hudopisk N, et al. (2013) Tick-borne encephalitis associated with consumption of raw goat milk, Slovenia, 2012. Emerg Infect Dis 19(5):806- 808.

70. Balogh Z, et al. (2010) Tick-borne encephalitis outbreak in Hungary due to consumption of raw goat milk. J Virol Methods 163(2):481-485.

71. Prevention. CfDCa (2002) Possible West Nile virus transmission to an infant through breast-feeding. Morb Mortal Wkly Rep. 51(39):877-878.

72. Betsem E, et al. (2017) Correlation of West Nile Virus Incidence in Donated Blood with West Nile Neuroinvasive Disease Rates, United States, 2010-2012. Emerg Infect Dis 23(2):212-219.

73. Kolodziejek J, et al. (2015) West Nile virus positive and subsequent entomological investigation, Austria, 2014. PLoS One 10(5):e0126381.

74. Chye JK, et al. (1997) Vertical transmission of dengue. Clin Infect Dis 25(6):1374-1377.

162

75. Sulkin SE, Burns KF, Shelton DF, & Wallis C (1962) Bat virus: infections of man and . Tex Rep Biol Med 20:113-127.

76. Prevention CfDCa (2002) Intrauterine West Nile Virus infection - United States. in Morb Mortal Wkly Rep (CDC), pp 1135-1136.

77. Chaturvedi UC, et al. (1980) Transplacental infection with Japanese encephalitis virus. J Infect Dis 141(6):712-715.

78. Organization WH (2016) Yellow fever. (World Health Organization, Geneva, Switzerland).

79. WHO (2018) A global strategy to Eliminate Yellow fever Epidemics 2017- 2016. in World Health Organization, ed Organization WH (Geneva).

80. Campbell GL, et al. (2011) Estimated global incidence of Japanese encephalitis: a systematic review. Bull World Health Organ 89(10):766- 774, 774A-774E.

81. Hegde NR & Gore MM (2017) Japanese encephalitis vaccines: Immunogenicity, protective efficacy, effectiveness, and impact on the burden of disease. Hum Vaccin Immunother 13(6):1-18.

82. Centers for Disease C (2016) West Nile virus disease cases reported to CDC by state of residence, 1999-2016.

83. Benedict MQ, Levine RS, Hawley WA, & Lounibos LP (2007) Spread of the tiger: global risk of invasion by the mosquito . Vector Borne Zoonotic Dis 7(1):76-85.

84. Reiter P & Sprenger D (1987) The used tire trade: a mechanism for the worldwide dispersal of container breeding mosquitoes. J Am Mosq Control Assoc 3(3):494-501.

85. Monaghan AJ, et al. (2016) The potential impacts of climatic and population changes on human exposure to the virus vector mosquito Aedes aegypti. Climatic Change:1-14.

86. Rochlin I, Ninivaggi DV, Hutchinson ML, & Farajollahi A (2013) Climate change and range expansion of the Asian tiger mosquito (Aedes albopictus) in Northeastern USA: implications for public health practitioners. PLoS One 8(4):e60874.

87. Samy AM, et al. (2016) Climate Change Influences on the Global Potential Distribution of the Mosquito , Vector of West Nile Virus and Lymphatic Filariasis. PLoS One 11(10):e0163863.

163

88. Kraemer MUG, et al. (2019) Past and future spread of the arbovirus vectors Aedes aegypti and Aedes albopictus. Nat Microbiol 4(5):854-863.

89. Origin NRCUCoASGCfSaRtEDoZ (2009) Drivers of Zoonotic Diseases. Sustaining Global Surveillance and Response to Emerging Zoonotic Diseases, ed Keusch GT PM, Gonzalez MC, et al. (National Academies Press (US), (DC)).

90. Rochlin I, Faraji A, Ninivaggi DV, Barker CM, & Kilpatrick AM (2016) Anthropogenic impacts on mosquito populations in North America over the past century. Nat Commun 7:13604.

91. Monaghan A (2016) Global warming to expose more people to Zika- spreading mosquito Aedes aegypti. Environment + Energy.

92. Johnson HN (1970) Long-term persistence of Modoc virus in hamster- kidney cells. In vivo and in vitro demonstration. Am J Trop Med Hyg 19(3):537-539.

93. Berge TO (1975) International Catalogue of Arboviruses Including Certain Other Viruses of Vertebrates. ed Berge TO (U.S. Department of Health, Education, and Welfare, Public Health Service), 2 Ed Vol 75.

94. Leyssen P, et al. (2002) Complete genome sequence, taxonomic assignment, and comparative analysis of the untranslated regions of the Modoc virus, a flavivirus with no known vector. Virology 293(1):125-140.

95. Leyssen P, et al. (2003) Impact of direct virus-induced neuronal dysfunction and immunological damage on the progression of flavivirus (Modoc) encephalitis in a murine model. J Neurovirol 9(1):69-78.

96. Leyssen P, et al. (2003) Acute encephalitis, a poliomyelitis-like syndrome and neurological sequelae in a hamster model for flavivirus infections. Brain Pathol 13(3):279-290.

97. Charlier N, et al. (2010) Replication of not-known-vector flaviviruses in mosquito cells is restricted by intracellular host factors rather than by the proteins. J Gen Virol 91(Pt 7):1693-1697.

98. Charlier N, et al. (2004) Exchanging the yellow fever virus envelope proteins with Modoc virus prM and E proteins results in a chimeric virus that is neuroinvasive in SCID mice. J Virol 78(14):7418-7426.

164

99. Laurent-Rolle M, et al. (2014) The interferon signaling antagonist function of yellow fever virus NS5 protein is activated by type I interferon. Cell Host Microbe 16(3):314-327.

100. Tumban E, et al. (2013) Replacement of conserved or variable sequences of the mosquito-borne dengue virus 3' UTR with homologous sequences from Modoc virus does not change infectivity for mosquitoes. J Gen Virol 94(Pt 4):783-788.

101. Davis JW & Hardy JL (1973) In vitro studies with Modoc virus in Vero cells: plaque assay and kinetics of growth, neutralization, and thermal inactivation. Appl Microbiol 26(3):344-348.

102. Goo L, VanBlargan LA, Dowd KA, Diamond MS, & Pierson TC (2017) A single mutation in the envelope protein modulates flavivirus antigenicity, stability, and pathogenesis. PLoS Pathog 13(2):e1006178.

103. Baron S, Fons M, & Albrecht T (1996) Viral Pathogenesis. Medical Microbiology, eds th & Baron SGalveston (TX)).

104. Izuogu AO, et al. (2017) Interferon signaling in Peromyscus leucopus confers a potent and specific restriction to vector-borne flaviviruses. PLoS One 12(6):e0179781.

105. Ruzek D, et al. (2009) CD8+ T-cells mediate immunopathology in tick- borne encephalitis. Virology 384(1):1-6.

106. Li G, et al. (2008) Yellow fever virus infection in Syrian golden hamsters: relationship between cytokine expression and pathologic changes. Int J Clin Exp Pathol 1(2):169-179.

107. Neefjes J, Jongsma ML, Paul P, & Bakke O (2011) Towards a systems understanding of MHC class I and MHC class II antigen presentation. Nat Rev Immunol 11(12):823-836.

108. Johnson DR & Pober JS (1990) Tumor necrosis factor and immune interferon synergistically increase transcription of HLA class I heavy- and light-chain genes in vascular endothelium. Proc Natl Acad Sci U S A 87(13):5183-5187.

109. Rodriguez-Pinto D (2005) B cells as antigen presenting cells. Cell Immunol 238(2):67-75.

110. Stetson DB & Medzhitov R (2006) Type I interferons in host defense. Immunity 25(3):373-381.

165

111. Piehler J, Thomas C, Garcia KC, & Schreiber G (2012) Structural and dynamic determinants of type I interferon receptor assembly and their functional interpretation. Immunol Rev 250(1):317-334.

112. Rawlings JS, Rosler KM, & Harrison DA (2004) The JAK/STAT signaling pathway. J Cell Sci 117(Pt 8):1281-1283.

113. Ivashkiv LB & Donlin LT (2014) Regulation of type I interferon responses. Nat Rev Immunol 14(1):36-49.

114. de Weerd NA & Nguyen T (2012) The interferons and their receptors-- distribution and regulation. Immunol Cell Biol 90(5):483-491.

115. Schoggins JW, et al. (2011) A diverse range of gene products are effectors of the type I interferon antiviral response. Nature 472(7344):481- 485.

116. Hoffmann HH, Schneider WM, & Rice CM (2015) Interferons and viruses: an evolutionary arms race of molecular interactions. Trends Immunol 36(3):124-138.

117. Pestka S, Krause CD, & Walter MR (2004) Interferons, interferon-like cytokines, and their receptors. Immunol Rev 202:8-32.

118. Kotenko SV, et al. (2003) IFN-lambdas mediate antiviral protection through a distinct class II cytokine receptor complex. Nat Immunol 4(1):69- 77.

119. Spellberg B & Edwards JE, Jr. (2001) Type 1/Type 2 immunity in infectious diseases. Clin Infect Dis 32(1):76-102.

120. Annunziato F, Romagnani C, & Romagnani S (2015) The 3 major types of innate and adaptive cell-mediated effector immunity. J Allergy Clin Immunol 135(3):626-635.

121. Mosser DM (2003) The many faces of macrophage activation. J Leukoc Biol 73(2):209-212.

122. Vignali DA, Collison LW, & Workman CJ (2008) How regulatory T cells work. Nat Rev Immunol 8(7):523-532.

123. Best SM (2017) The Many Faces of the Flavivirus NS5 Protein in Antagonism of Type I Interferon Signaling. J Virol 91(3).

124. Suthar MS, Diamond MS, & Gale M, Jr. (2013) West Nile virus infection and immunity. Nat Rev Microbiol 11(2):115-128.

166

125. Fredericksen BL & Gale M, Jr. (2006) West Nile virus evades activation of interferon regulatory factor 3 through RIG-I-dependent and -independent pathways without antagonizing host defense signaling. J Virol 80(6):2913- 2923.

126. Ye J, Zhu B, Fu ZF, Chen H, & Cao S (2013) Immune evasion strategies of flaviviruses. Vaccine 31(3):461-471.

127. Lubick KJ, et al. (2015) Flavivirus Antagonism of Type I Interferon Signaling Reveals Prolidase as a Regulator of IFNAR1 Surface Expression. Cell Host Microbe 18(1):61-74.

128. Ashour J, Laurent-Rolle M, Shi PY, & Garcia-Sastre A (2009) NS5 of dengue virus mediates STAT2 binding and degradation. J Virol 83(11):5408-5418.

129. Grant A, et al. (2016) Zika Virus Targets Human STAT2 to Inhibit Type I Interferon Signaling. Cell Host Microbe 19(6):882-890.

130. Lin RJ, Chang BL, Yu HP, Liao CL, & Lin YL (2006) Blocking of interferon- induced Jak-Stat signaling by Japanese encephalitis virus NS5 through a protein tyrosine phosphatase-mediated mechanism. J Virol 80(12):5908- 5918.

131. Zmurko J, Neyts J, & Dallmeier K (2015) Flaviviral NS4b, chameleon and jack-in-the-box roles in viral replication and pathogenesis, and a molecular target for antiviral intervention. Rev Med Virol 25(4):205-223.

132. Roby JA, Pijlman GP, Wilusz J, & Khromykh AA (2014) Noncoding subgenomic flavivirus RNA: multiple functions in West Nile virus pathogenesis and modulation of host responses. Viruses 6(2):404-427.

133. Avirutnan P, et al. (2010) Antagonism of the complement component C4 by flavivirus nonstructural protein NS1. J Exp Med 207(4):793-806.

134. Rastogi M, Sharma N, & Singh SK (2016) Flavivirus NS1: a multifaceted enigmatic . Virol J 13:131.

135. Mackenzie JM, Jones MK, & Young PR (1996) Immunolocalization of the dengue virus nonstructural glycoprotein NS1 suggests a role in viral RNA replication. Virology 220(1):232-240.

136. Liu J, et al. (2016) Flavivirus NS1 protein in infected host sera enhances viral acquisition by mosquitoes. Nat Microbiol 1(9):16087.

167

137. Chung KM, et al. (2006) West Nile virus nonstructural protein NS1 inhibits complement activation by binding the regulatory protein factor H. Proc Natl Acad Sci U S A 103(50):19111-19116.

138. Huerta-Zepeda A, et al. (2008) Crosstalk between and inflammation during Dengue virus infection. Thromb Haemost 99(5):936- 943.

139. McAuley AJ, et al. (2016) Recovery of West Nile Virus Envelope Protein Domain III Chimeras with Altered Antigenicity and Mouse Virulence. J Virol 90(9):4757-4770.

140. Beasley DW, et al. (2005) Envelope protein glycosylation status influences mouse neuroinvasion phenotype of genetic lineage 1 West Nile virus strains. J Virol 79(13):8339-8347.

141. Bardina SV, et al. (2017) Enhancement of Zika virus pathogenesis by preexisting antiflavivirus immunity. Science 356(6334):175-180.

142. Dejnirattisai W, et al. (2010) Cross-reacting antibodies enhance dengue virus infection in humans. Science 328(5979):745-748.

143. Kuczera D, Assolini JP, Tomiotto-Pellissier F, Pavanelli WR, & Silveira GF (2018) Highlights for Dengue Immunopathogenesis: Antibody-Dependent Enhancement, , and Beyond. J Interferon Cytokine Res 38(2):69-80.

144. Le Breton M, et al. (2011) Flavivirus NS3 and NS5 proteins interaction network: a high-throughput yeast two-hybrid screen. BMC Microbiol 11:234.

145. Pabalan N, et al. (2018) Associations of DC-SIGN (CD209) promoter - 336G/A polymorphism (rs4804803) with dengue infection: A systematic review and meta-analysis. Acta Trop 177:186-193.

146. Clyde K, Kyle JL, & Harris E (2006) Recent advances in deciphering viral and host determinants of dengue virus replication and pathogenesis. J Virol 80(23):11418-11431.

147. Yacoub S, Wertheim H, Simmons CP, Screaton G, & Wills B (2014) Cardiovascular manifestations of the emerging dengue . Nat Rev Cardiol 11(6):335-345.

148. Bogovic P, et al. (2019) Inflammatory Immune Responses in the Pathogenesis of Tick-Borne Encephalitis. J Clin Med 8(5).

168

149. Gelpi E, et al. (2005) Visualization of Central European tick-borne encephalitis infection in fatal human cases. J Neuropathol Exp Neurol 64(6):506-512.

150. Kumar M, Verma S, & Nerurkar VR (2010) Pro-inflammatory cytokines derived from West Nile virus (WNV)-infected SK-N-SH cells mediate neuroinflammatory markers and neuronal death. J 7:73.

151. Bell TM, Field EJ, & Narang HK (1971) Zika virus infection of the central nervous system of mice. Arch Gesamte Virusforsch 35(2):183-193.

152. Miranda HA, 2nd, et al. (2016) Expanded Spectrum of Congenital Ocular Findings in with Presumed Zika Infection. Ophthalmology 123(8):1788-1794.

153. Garcia-Bujalance S, et al. (2017) Persistence and infectivity of Zika virus in semen after returning from endemic areas: Report of 5 cases. J Clin Virol 96:110-115.

154. Monath TP & Vasconcelos PF (2015) Yellow fever. J Clin Virol 64:160- 173.

155. Vieira WT, Gayotto LC, de Lima CP, & de Brito T (1983) of the human liver in yellow fever with special emphasis on the diagnostic role of the . Histopathology 7(2):195-208.

156. Quaresma JA, et al. (2006) Revisiting the liver in human yellow fever: virus-induced apoptosis in hepatocytes associated with TGF-beta, TNF- alpha and NK cells activity. Virology 345(1):22-30.

157. Barrett AD (1997) Yellow fever vaccines. Biologicals 25(1):17-25.

158. Belmusto-Worn VE, et al. (2005) Randomized, double-blind, phase III, pivotal field trial of the comparative immunogenicity, safety, and tolerability of two yellow fever 17D vaccines (Arilvax and YF-VAX) in healthy infants and children in . Am J Trop Med Hyg 72(2):189-197.

159. De Madrid AT & Porterfield JS (1974) The flaviviruses (group B arboviruses): a cross-neutralization study. J Gen Virol 23(1):91-96.

160. Mansfield KL, et al. (2011) Flavivirus-induced antibody cross-reactivity. J Gen Virol 92(Pt 12):2821-2829.

161. Ulbert S & Magnusson SE (2014) Technologies for the development of West Nile virus vaccines. Future Microbiol 9(10):1221-1232.

169

162. Biedenbender R, Bevilacqua J, Gregg AM, Watson M, & Dayan G (2011) Phase II, randomized, double-blind, placebo-controlled, multicenter study to investigate the immunogenicity and safety of a West Nile virus vaccine in healthy adults. J Infect Dis 203(1):75-84.

163. Wahala WM & Silva AM (2011) The human antibody response to dengue virus infection. Viruses 3(12):2374-2395.

164. Capeding MR, et al. (2014) Clinical efficacy and safety of a novel tetravalent in healthy children in Asia: a phase 3, randomised, observer-masked, placebo-controlled trial. Lancet 384(9951):1358-1365.

165. Villar L, et al. (2015) Efficacy of a tetravalent dengue vaccine in children in Latin America. N Engl J Med 372(2):113-123.

166. FDA (2019) Dengvaxia. (United States Government, U.S. Food & Drug Administration).

167. Abbink P, et al. (2016) Protective efficacy of multiple vaccine platforms against Zika virus challenge in rhesus monkeys. Science 353(6304):1129- 1132.

168. Larocca RA, et al. (2016) Vaccine protection against Zika virus from . Nature 536(7617):474-478.

169. Shan C, et al. (2017) A live-attenuated candidate induces sterilizing immunity in mouse models. Nat Med 23(6):763-767.

170. Kim E, et al. (2016) Preventative Vaccines for Zika Virus Outbreak: Preliminary Evaluation. EBioMedicine 13:315-320.

171. Richner JM, et al. (2017) Modified mRNA Vaccines Protect against Zika Virus Infection. Cell 169(1):176.

172. VanBlargan LA, et al. (2018) An mRNA Vaccine Protects Mice against Multiple Tick-Transmitted Flavivirus Infections. Cell Rep 25(12):3382-3392 e3383.

173. Pandey RK, Dahiya S, Mahita J, Sowdhamini R, & Prajapati VK (2019) Vaccination and immunization strategies to design Aedes aegypti salivary protein based subunit vaccine tackling Flavivirus infection. Int J Biol Macromol 122:1203-1211.

170

174. Bidet K, et al. (2019) Mimicking immune signatures of flavivirus infection with targeted adjuvants improves dengue subunit vaccine immunogenicity. NPJ Vaccines 4:27.

175. Sun H, Chen Q, & Lai H (2017) Development of Antibody Therapeutics against Flaviviruses. Int J Mol Sci 19(1).

176. Li Z, et al. (2018) Erythrosin B is a potent and broad-spectrum orthosteric inhibitor of the flavivirus NS2B-NS3 protease. Antiviral Res 150:217-225.

177. Ncube NB, Ramharack P, & Soliman MES (2018) An "All-In-One" Pharmacophoric Architecture for the Discovery of Potential Broad- Spectrum Anti-Flavivirus Drugs. Appl Biochem Biotechnol 185(3):799-814.

178. Khan A, et al. (2018) Allosteric ligands for the pharmacologically important Flavivirus target (NS5) from ZINC database based on pharmacophoric points, free energy calculations and dynamics correlation. J Mol Graph Model 82:37-47.

179. Brecher M, et al. (2015) Novel Broad Spectrum Inhibitors Targeting the Flavivirus . PLoS One 10(6):e0130062.

180. Mackenzie JS, Gubler DJ, & Petersen LR (2004) Emerging flaviviruses: the spread and resurgence of Japanese encephalitis, West Nile and dengue viruses. Nat Med 10(12 Suppl):S98-109.

181. Martin JF & Marshall J (2003) New tendencies and strategies in international immunisation: GAVI and The Vaccine Fund. Vaccine 21(7- 8):587-592.

182. Ericsson AC, Crim MJ, & Franklin CL (2013) A brief history of animal modeling. Mo Med 110(3):201-205.

183. Fields S & Johnston M (2005) Cell biology. Whither model organism research? Science 307(5717):1885-1886.

184. Vandamme TF (2014) Use of rodents as models of human diseases. J Pharm Bioallied Sci 6(1):2-9.

185. Mandl JN, et al. (2015) Reservoir host immune responses to emerging zoonotic viruses. Cell 160(1-2):20-35.

186. Jones KE, et al. (2008) Global trends in emerging infectious diseases. Nature 451(7181):990-993.

187. Cross AR, et al. (2018) Zoonoses under our noses. Microbes Infect.

171

188. Barbour AG (2017) Infection resistance and tolerance in Peromyscus spp., natural reservoirs of microbes that are virulent for humans. Semin Cell Dev Biol 61:115-122.

189. Bird BH, et al. (2012) Severe hemorrhagic fever in strain 13/N guinea pigs infected with Lujo virus. PLoS Negl Trop Dis 6(8):e1801.

190. Schountz T, et al. (2012) Kinetics of immune responses in deer mice experimentally infected with Sin Nombre virus. J Virol 86(18):10015- 10027.

191. Iannaccone PM & Jacob HJ (2009) Rats! Dis Model Mech 2(5-6):206-210.

192. Capecchi MR (2001) Generating mice with targeted mutations. Nat Med 7(10):1086-1090.

193. Smith C.L. BJA, Kadin J.A., Richardson J.E., Bult C.J. (2018) Mouse Genome Database (MGD)-2018: knowledgebase for the laboratory mouse. . (The Jackson Laboratory, Online).

194. Laulederkind SJ, et al. (2013) PhenoMiner: quantitative phenotype curation at the rat genome database. Database (Oxford) 2013:bat015.

195. Day CP, Merlino G, & Van Dyke T (2015) Preclinical mouse cancer models: a maze of opportunities and challenges. Cell 163(1):39-53.

196. Wade CM & Daly MJ (2005) Genetic variation in laboratory mice. Nat Genet 37(11):1175-1180.

197. Steppan S, Adkins R, & Anderson J (2004) Phylogeny and divergence- date estimates of rapid radiations in muroid rodents based on multiple nuclear genes. Syst Biol 53(4):533-553.

198. Hooper E (1968) Biology of Peromyscus (Rodentia) (American Society of Mammalogists).

199. Bedford NL & Hoekstra HE (2015) Peromyscus mice as a model for studying natural variation. Elife 4.

200. Hall ER (1981) The mammals of North America (Wiley, New York) 2d Ed.

201. Ramsdell CM, et al. (2008) 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 8:65.

172

202. Sumner FB (1917) Several Color "Mutations" in Mice of the Genus Peromyscus. Genetics 2(3):291-300.

203. Burger J & Gochfeld M (1992) Survival and reproduction in Peromyscus leucopus in the laboratory: viable model for aging studies. Growth Dev Aging 56(1):17-22.

204. Joyner CP, Myrick LC, Crossland JP, & Dawson WD (1998) Deer Mice As Laboratory Animals. ILAR J 39(4):322-330.

205. Havighorst A, Crossland J, & Kiaris H (2017) Peromyscus as a model of human disease. Semin Cell Dev Biol 61:150-155.

206. Lindenmann J (1964) Inheritance of Resistance to Influenza Virus in Mice. Proc Soc Exp Biol Med 116:506-509.

207. CDC (2014) Hantavirus. (Centers for Disease Control and Prevention, Online).

208. Deardorff ER, et al. (2013) Powassan virus in mammals, Alaska and , U.S.A., and Russia, 2004-2007. Emerg Infect Dis 19(12):2012- 2016.

209. Ebel GD, Campbell EN, Goethert HK, Spielman A, & Telford SR, 3rd (2000) Enzootic transmission of deer tick virus in New England and Wisconsin sites. Am J Trop Med Hyg 63(1-2):36-42.

210. Baggs EM, Stack SH, Finney-Crawley JR, & Simon NP (2011) Peromyscus maniculatus, a possible reservoir host of Borrelia garinii from the Gannet Islands off Newfoundland and Labrador. J Parasitol 97(5):792- 794.

211. Breitschwerdt EB & Kordick DL (2000) Bartonella infection in animals: carriership, reservoir potential, pathogenicity, and zoonotic potential for human infection. Clin Microbiol Rev 13(3):428-438.

212. Thompson M, Mykytczuk N, Gooderham K, & Schulte-Hostedde A (2012) Prevalence of the bacterium Coxiella burnetii in wild rodents from a Canadian natural environment park. Zoonoses Public Health 59(8):553- 560.

213. Zeidner NS, et al. (2000) Transmission of the agent of human granulocytic by Ixodes spinipalpis ticks: evidence of an enzootic cycle of dual infection with Borrelia burgdorferi in Northern Colorado. J Infect Dis 182(2):616-619.

173

214. Levin ML, Nicholson WL, Massung RF, Sumner JW, & Fish D (2002) Comparison of the reservoir competence of medium-sized mammals and Peromyscus leucopus for Anaplasma phagocytophilum in . Vector Borne Zoonotic Dis 2(3):125-136.

215. Danforth M, Tucker J, & Novak M (2018) The Deer Mouse (Peromyscus maniculatus) as an Enzootic Reservoir of Plague in California. Ecohealth 15(3):566-576.

216. Walsh M & Haseeb MA (2015) Modeling the ecologic niche of plague in sylvan and domestic animal hosts to delineate sources of human exposure in the western United States. PeerJ 3:e1493.

217. Spielman A, et al. (1981) Reservoir hosts of human on Nantucket Island. Am J Trop Med Hyg 30(3):560-565.

218. Schwanz LE (2006) Schistosome infection in deer mice (Peromyscus maniculatus): impacts on host physiology, behavior and energetics. J Exp Biol 209(Pt 24):5029-5037.

219. McKown RD, Upton SJ, Klemm RD, & Ridley RK (1990) New host and locality record for Trypanosoma peromysci. J Parasitol 76(2):281-283.

220. Yigit NK, B. (2008) Mesocricetus auratus. in e.T13219A3421173 (The IUCN Red List of Threatened Species 2008, Online).

221. Henwood C (2001) The Discovery of the Syrian (Golden) Hamster, Mesocricetus Auratus. 39.

222. Gattermann R, et al. (2001) Notes on the current distribution and the ecology of wild golden hamsters (Mesocricetus auratus). Journal of Zoology 254:359-365.

223. Gattermann R, et al. (2008) Golden hamsters are nocturnal in captivity but diurnal in nature. Biol Lett 4(3):253-255.

224. Gattermann R, Fritzsche P, Weinandy R, & Neumann K (2002) Comparative studies of body mass, body measurements and organ weights of wild-derived and laboratory golden hamsters (Mesocricetus auratus). Lab Anim 36(4):445-454.

225. Moravec J, Melichar I, Jansky L, & Vyskocil F (1973) Effect of hibernation and noradrenaline on the resting state of neuromuscular junction of golden hamster (Mesocricetus auratus). Pflugers Arch 345(2):93-106.

174

226. Steinberg B, Singh IJ, & Mitchell OG (1981) The effects of cold-stress. Hibernation, and prolonged inactivity on bone dynamics in the golden hamster, Mesocricetus auratus. J Morphol 167(1):43-51.

227. Mohr U (1979) The Syrian golden hamster as a model in cancer research. Prog Exp Tumor Res 24:245-252.

228. Gustafson G, Stelling E, Mornstad H, & Brunius E (1973) The effect of the calcium saccharose phosphate preparation Anticay on dental caries in the golden hamster. Odontol Revy 24(4):325-334.

229. Gustafson G, Stelling E, & Brunius E (1964) Dietary Calcium and Caries. Experimental Dental Caries in Golden Hamster. Xii. Acta Odontol Scand 22:477-485.

230. Mascres C, The Luu H, & Magnan R (1982) [Influence of pulp necrosis on experimental caries in the golden Syrian hamster]. Actual Odontostomatol (Paris) 36(140):565-576.

231. Diamond M & Yanagimachi R (1970) Reproductive development in the female golden hamster in relation to spontaneous estrus. Biol Reprod 2(2):223-229.

232. Fleming AS (1978) Food intake and body weight regulation during the reproductive cycle of the golden hamster (Mesocricetus auratus). Behav Biol 24(3):291-306.

233. Fritzsche P, Riek M, & Gattermann R (2000) Effects of social stress on behavior and corpus luteum in female golden hamsters (Mesocricetus auratus). Physiol Behav 68(5):625-630.

234. Zivcec M, Safronetz D, Haddock E, Feldmann H, & Ebihara H (2011) Validation of assays to monitor immune responses in the Syrian golden hamster (Mesocricetus auratus). J Immunol Methods 368(1-2):24-35.

235. Tripathi S, et al. (2017) A novel Zika virus mouse model reveals strain specific differences in virus pathogenesis and host inflammatory immune responses. PLoS Pathog 13(3):e1006258.

236. Gowen BB, et al. (2010) Assessing changes in vascular permeability in a hamster model of . Virol J 7:240.

237. Ebihara H, et al. (2013) A Syrian golden hamster model recapitulating ebola hemorrhagic fever. J Infect Dis 207(2):306-318.

175

238. Sbrana E, Xiao SY, Popov VL, Newman PC, & Tesh RB (2006) Experimental yellow fever virus infection in the golden hamster (Mesocricetus auratus) III. Clinical laboratory values. Am J Trop Med Hyg 74(6):1084-1089.

239. Safronetz D, et al. (2011) Pathogenesis and host response in Syrian hamsters following intranasal infection with Andes virus. PLoS Pathog 7(12):e1002426.

240. Xiao SY, Guzman H, Zhang H, Travassos da Rosa AP, & Tesh RB (2001) West Nile virus infection in the golden hamster (Mesocricetus auratus): a model for West Nile encephalitis. Emerg Infect Dis 7(4):714-721.

241. Vernel-Pauillac F & Merien F (2006) Proinflammatory and immunomodulatory cytokine mRNA time course profiles in hamsters infected with a virulent variant of interrogans. Infect Immun 74(7):4172-4179.

242. Melby PC, Tryon VV, Chandrasekar B, & Freeman GL (1998) Cloning of Syrian hamster (Mesocricetus auratus) cytokine cDNAs and analysis of cytokine mRNA expression in experimental visceral leishmaniasis. Infect Immun 66(5):2135-2142.

243. Espitia CM, et al. (2014) Transcriptional profiling of the spleen in progressive visceral leishmaniasis reveals mixed expression of type 1 and type 2 cytokine-responsive genes. BMC Immunol 15:38.

244. Mendlovic F, Cruz-Rivera M, Avila G, Vaughan G, & Flisser A (2015) Cytokine, antibody and proliferative cellular responses elicited by Taenia solium calreticulin upon experimental infection in hamsters. PLoS One 10(3):e0121321.

245. Karabatsos N (1984) International Catalogue of Arboviruses Including Certain Other Viruses of Vertebrates. American Society of and Hygiene, ed Berge TO (U.S. Department of Health, Education, and Welfare, Public Health Service, San Antonio, TX), 3 Ed, pp 691-692.

246. Schountz T, et al. (2007) Regulatory T cell-like responses in deer mice persistently infected with Sin Nombre virus. Proc Natl Acad Sci U S A 104(39):15496-15501.

247. Mlera L, et al. (2018) Peromyscus leucopus mouse brain transcriptome response to Powassan virus infection. J Neurovirol 24(1):75-87.

176

248. Elliott LH, et al. (1994) Isolation of the causative agent of hantavirus pulmonary syndrome. Am J Trop Med Hyg 51(1):102-108.

249. Pesko KN, Torres-Perez F, Hjelle BL, & Ebel GD (2010) Molecular epidemiology of Powassan virus in North America. J Gen Virol 91(Pt 11):2698-2705.

250. Song JW, et al. (1994) Isolation of pathogenic hantavirus from white- footed mouse (Peromyscus leucopus). Lancet 344(8937):1637.

251. Perz JF & Le Blancq SM (2001) Cryptosporidium parvum infection involving novel in wildlife from lower New York State. Appl Environ Microbiol 67(3):1154-1162.

252. Livak KJ & Schmittgen TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25(4):402-408.

253. Mallarino R, Linden TA, Linnen CR, & Hoekstra HE (2017) The role of isoforms in the evolution of cryptic coloration in Peromyscus mice. Mol Ecol 26(1):245-258.

254. Fairbrother A & Yuill TM (1987) Experimental infection and horizontal transmission of Modoc virus in deer mice (Peromyscus maniculatus). J Wildl Dis 23(2):179-185.

255. Kamiyama N, et al. (2017) inhibits Zika virus (ZIKV) replication in vitro and suppresses viremia in ZIKV-infected STAT1-deficient mice. Antiviral Res 146:1-11.

256. Allison AB, Mead DG, Gibbs SE, Hoffman DM, & Stallknecht DE (2004) West Nile virus viremia in wild rock pigeons. Emerg Infect Dis 10(12):2252-2255.

257. Hey YY & O'Neill HC (2012) Murine spleen contains a diversity of myeloid and dendritic cells distinct in antigen presenting function. J Cell Mol Med 16(11):2611-2619.

258. Swirski FK, et al. (2009) Identification of splenic reservoir monocytes and their deployment to inflammatory sites. Science 325(5940):612-616.

259. Kyle JL, Beatty PR, & Harris E (2007) Dengue virus infects macrophages and dendritic cells in a mouse model of infection. J Infect Dis 195(12):1808-1817.

177

260. Brown AN, Kent KA, Bennett CJ, & Bernard KA (2007) Tissue tropism and neuroinvasion of West Nile virus do not differ for two mouse strains with different survival rates. Virology 368(2):422-430.

261. Miner JJ & Diamond MS (2017) Zika Virus Pathogenesis and Tissue Tropism. Cell Host Microbe 21(2):134-142.

262. Shrestha B, et al. (2006) Gamma interferon plays a crucial early antiviral role in protection against West Nile virus infection. J Virol 80(11):5338- 5348.

263. Bonney S, et al. (2019) Gamma Interferon Alters Junctional Integrity via Rho Kinase, Resulting in Blood-Brain Barrier Leakage in Experimental . mBio 10(4).

264. Sedger LM & McDermott MF (2014) TNF and TNF-receptors: From mediators of cell death and inflammation to therapeutic giants - past, present and future. Cytokine Growth Factor Rev 25(4):453-472.

265. Winter PM, et al. (2004) Proinflammatory cytokines and chemokines in humans with Japanese encephalitis. J Infect Dis 190(9):1618-1626.

266. Chua JJ, Bhuvanakantham R, Chow VT, & Ng ML (2005) Recombinant non-structural 1 (NS1) protein of dengue-2 virus interacts with human STAT3beta protein. Virus Res 112(1-2):85-94.

267. Munoz-Jordan JL & Fredericksen BL (2010) How flaviviruses activate and suppress the interferon response. Viruses 2(2):676-691.

268. Manet C, et al. (2018) Host genetic control of mosquito-borne Flavivirus infections. Mamm Genome.

269. Liu SY, Sanchez DJ, Aliyari R, Lu S, & Cheng G (2012) Systematic identification of type I and type II interferon-induced antiviral factors. Proc Natl Acad Sci U S A 109(11):4239-4244.

270. Brinton MA (2014) Genetic resistance to flaviviruses. Monoclon Antib Immunodiagn Immunother 33(3):183-185.

271. Graham JB, et al. (2015) Genetic diversity in the collaborative cross model recapitulates human West Nile virus disease outcomes. mBio 6(3):e00493-00415.

272. Mashimo T, et al. (2002) A nonsense mutation in the gene encoding 2'-5'- oligoadenylate synthetase/L1 isoform is associated with West Nile virus

178

susceptibility in laboratory mice. Proc Natl Acad Sci U S A 99(17):11311- 11316.

273. Perelygin AA, et al. (2002) Positional cloning of the murine flavivirus resistance gene. Proc Natl Acad Sci U S A 99(14):9322-9327.

274. Venter M, et al. (2005) Gene expression in mice infected with West Nile virus strains of different neurovirulence. Virology 342(1):119-140.

275. Schoggins JW (2014) Interferon-stimulated genes: roles in viral pathogenesis. Curr Opin Virol 6:40-46.

276. Fuchs J, et al. (2017) Evolution and Antiviral Specificities of Interferon- Induced Mx Proteins of Bats against Ebola, Influenza, and Other RNA Viruses. J Virol 91(15).

277. Brett D. Lindenbach H-JT, Charles M. Rice (2007) Flaviviridae: The Viruses and Their Replication. Fields Virology, ed Howley DMKaPM (Lippincott-Raven, Philadelphia), 5th Ed.

278. Guzman A & Isturiz RE (2010) Update on the global spread of dengue. Int J Antimicrob Agents 36 Suppl 1:S40-42.

279. Bhatt S, et al. (2013) The global distribution and burden of dengue. Nature 496(7446):504-507.

280. Bidet K & Garcia-Blanco MA (2014) Flaviviral : weapons and targets in the war between virus and host. Biochem J 462(2):215-230.

281. Pijlman GP, et al. (2008) A highly structured, nuclease-resistant, noncoding RNA produced by flaviviruses is required for pathogenicity. Cell Host Microbe 4(6):579-591.

282. Charlier N, et al. (2002) Infection of SCID mice with Montana Myotis leukoencephalitis virus as a model for flavivirus encephalitis. J Gen Virol 83(Pt 8):1887-1896.

283. Yauch LE & Shresta S (2008) Mouse models of dengue virus infection and disease. Antiviral Res 80(2):87-93.

284. Wahl-Jensen V, et al. (2012) Use of the Syrian hamster as a new model of ebola virus disease and other viral hemorrhagic fevers. Viruses 4(12):3754-3784.

285. Sessions OM, et al. (2013) Host cell transcriptome profile during wild-type and attenuated dengue virus infection. PLoS Negl Trop Dis 7(3):e2107.

179

286. O'Neal JT, et al. (2019) West Nile Virus-Inclusive Single-Cell RNA Sequencing Reveals Heterogeneity in the Type I Interferon Response within Single Cells. J Virol 93(6).

287. Boehm U, Klamp T, Groot M, & Howard JC (1997) Cellular responses to interferon-gamma. Annu Rev Immunol 15:749-795.

288. Billiau A & Matthys P (2009) Interferon-gamma: a historical perspective. Cytokine Growth Factor Rev 20(2):97-113.

289. Morikawa A, et al. (2000) Augmentation of nitric oxide production by gamma interferon in a mouse vascular endothelial cell line and its modulation by tumor necrosis factor alpha and lipopolysaccharide. Infect Immun 68(11):6209-6214.

290. Rose-John S (2018) Interleukin-6 Family Cytokines. Cold Spring Harb Perspect Biol 10(2).

291. Dienz O, et al. (2012) Essential role of IL-6 in protection against H1N1 influenza virus by promoting neutrophil survival in the lung. Mucosal Immunol 5(3):258-266.

292. Dienz O, et al. (2009) The induction of antibody production by IL-6 is indirectly mediated by IL-21 produced by CD4+ T cells. J Exp Med 206(1):69-78.

293. Rouse BT & Sehrawat S (2010) Immunity and immunopathology to viruses: what decides the outcome? Nat Rev Immunol 10(7):514-526.

294. Li MO, Wan YY, Sanjabi S, Robertson AK, & Flavell RA (2006) Transforming growth factor-beta regulation of immune responses. Annu Rev Immunol 24:99-146.

295. Kang S, Brown HM, & Hwang S (2018) Direct Antiviral Mechanisms of Interferon-Gamma. Immune Netw 18(5):e33.

296. Garcia-Nicolas O, Lewandowska M, Ricklin ME, & Summerfield A (2019) Monocyte-Derived Dendritic Cells as Model to Evaluate Species Tropism of Mosquito-Borne Flaviviruses. Front Cell Infect Microbiol 9:5.

297. King NJ, Shrestha B, & Kesson AM (2003) Immune modulation by flaviviruses. Adv Virus Res 60:121-155.

298. Morrison J, Aguirre S, & Fernandez-Sesma A (2012) Innate immunity evasion by Dengue virus. Viruses 4(3):397-413.

180

299. Libraty DH, et al. (2002) High circulating levels of the dengue virus nonstructural protein NS1 early in dengue illness correlate with the development of dengue hemorrhagic fever. J Infect Dis 186(8):1165-1168.

300. Tsai TT, et al. (2013) An emerging role for the anti- interleukin-10 in dengue virus infection. J Biomed Sci 20:40.

301. Bhatt RS, Kothari ST, Gohil DJ, D'Souza M, & Chowdhary AS (2015) Novel evidence of microglial immune response in impairment of Dengue infection of CNS. Immunobiology 220(10):1170-1176.

302. Brooks DG, et al. (2006) Interleukin-10 determines viral clearance or persistence in vivo. Nat Med 12(11):1301-1309.

303. Brockman MA, et al. (2009) IL-10 is up-regulated in multiple cell types during viremic HIV infection and reversibly inhibits virus-specific T cells. Blood 114(2):346-356.

304. Cavalcanti YV, Brelaz MC, Neves JK, Ferraz JC, & Pereira VR (2012) Role of TNF-Alpha, IFN-Gamma, and IL-10 in the Development of Pulmonary Tuberculosis. Pulm Med 2012:745483.

305. Assoian RK, et al. (1987) Expression and secretion of type beta transforming growth factor by activated human macrophages. Proc Natl Acad Sci U S A 84(17):6020-6024.

306. Malavige GN, et al. (2012) Cellular and cytokine correlates of severe dengue infection. PLoS One 7(11):e50387.

307. Cruz Hernandez SI, et al. (2016) Primary dengue virus infections induce differential cytokine production in Mexican patients. Mem Inst Oswaldo Cruz 111(3):161-167.

308. McFadden G, Mohamed MR, Rahman MM, & Bartee E (2009) Cytokine determinants of viral tropism. Nat Rev Immunol 9(9):645-655.

309. Restrepo BN, et al. (2008) Serum levels of cytokines in two ethnic groups with dengue virus infection. Am J Trop Med Hyg 79(5):673-677.

310. Hannah MF, Bajic VB, & Klein SL (2008) Sex differences in the recognition of and innate antiviral responses to Seoul virus in Norway rats. Brain Behav Immun 22(4):503-516.

181

311. Tesh RB, et al. (2005) Persistent West Nile virus infection in the golden hamster: studies on its mechanism and possible implications for other flavivirus infections. J Infect Dis 192(2):287-295.

312. Siirin MT, et al. (2007) Chronic St. Louis encephalitis virus infection in the golden hamster (Mesocricetus auratus). Am J Trop Med Hyg 76(2):299- 306.

313. WHO (2020) Vector-Borne Diseases. Newsroom Fact Sheets.

314. Ostfeld RS & Brunner JL (2015) Climate change and Ixodes tick-borne diseases of humans. Philos Trans R Soc Lond B Biol Sci 370(1665).

315. CDC (2018) Illnesses from Mosquito, Tick, and Bites Increasing in the US. in Cases triple; better tools needed to fight mosquitoes, ticks, and fleas (CDC Newsroom Releases Online).

316. WHO (2020) Dengue and severe dengue. Newsroom Fact Sheets.

317. Lehmer EM, Lavengood K, Miller M, Rodgers J, & Fenster SD (2018) Evaluating the Impacts of on Immune System Function of the Deer Mouse ( Peromyscus Maniculatus) Using Sin Nombre Virus and Bartonella as Model Pathogen Systems. J Wildl Dis 54(1):66-75.

318. Gomes L, et al. (2014) Sphingosine 1-phosphate in acute dengue infection. PLoS One 9(11):e113394.

319. Kohli P & Levy BD (2009) Resolvins and protectins: mediating solutions to inflammation. Br J Pharmacol 158(4):960-971.

182

APPENDIX A

MELT CURVE ANALYSES FOR MODV-SPECIFIC AND β-ACTIN PRIMERS

183

Melt curve analyses for MODV-specific (top) and β-actin primers (bottom) used in this study. Melt curves for both analyses were generated by incrementally increasing the thermocycler temperature by 0.5°C from 65°C to 95°C every 5 seconds. The generation of melt curves helps to clarify the production of a single amplicon.

184

APPENDIX B

GRAPHS OF RELATIVE GENE EXPRESSION WITH BIOREPLICATES AT EACH TIMEPOINT REPRESENTED (DEER MICE)

185

186

Relative gene expression values with bioreplicates at each timepoint represented. These graphs are intended to demonstrate the variance between bioreplicates from which tissues were collected at each timepoint. Note that the experimental design of this study was cross-sectional; therefore, “DM1, 2, and 3” at each timepoint does not indicate repeated-measures sampling, but rather represents each of the three individuals euthanized at that particular collection timepoint. Data are displayed on a logarithmic scale in order to capture values from all individuals.

187

APPENDIX C

GRAPHS OF RELATIVE GENE EXPRESSION WITH BIOREPLICATES AT EACH TIMEPOINT REPRESENTED (HAMSTERS)

188

189

Relative gene expression values with bioreplicates at each timepoint represented. These graphs are intended to demonstrate the variance between bioreplicates from which tissues were collected at each timepoint. Note that the experimental design of this study was cross-sectional; therefore, “H1, 2, and 3” at each timepoint does not indicate repeated-measures sampling, but rather represents each of the three individuals euthanized at that particular collection timepoint. Data are displayed on a logarithmic scale in order to capture values from all individuals.

190

APPENDIX D

EXPERIMENTAL INFECTION OF DEER MICE AND HAMSTERS WITH MODOC VIRUS: IACUC MEMORANDUM APPROVAL

191

Protocol Approval: 1603C-AH-DmH-19

192