Investigations Into Antiviral Microglia Activation During West

INVESTIGATIONS INTO ANTIVIRAL MICROGLIA ACTIVATION DURING WEST

NILE VIRUS INFECTION OF THE CENTRAL NERVOUS SYSTEM

EAMON DRAKE QUICK

B.S, Brown University, 2007

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirements for the degree of

Doctor of Philosophy

Neuroscience

2017

This thesis for the Doctor of Philosophy degree by

Eamon Drake Quick

has been approved for the

Neuroscience Program

Wendy Macklin, Chair

Kenneth L. Tyler, Advisor

Diego Restrepo

J. David Beckham

Laurel Lenz

Penny Clarke, Co-Advisor

Date: December 15th, 2017

ii Quick, Eamon Drake (PhD, Neuroscience)

Investigations into Microglia Activation During West Nile Virus Infection

Thesis directed by Kenneth L. Tyler

ABSTRACT

Microglia are the resident innate immune cells of the central nervous system (CNS),

and are thought to have important roles in the detection of noxious stimuli and the regulation

of inflammatory events within the CNS parenchyma. For certain diseases, the evidence for

microglia contributions to pathogenesis and inflammatory responses are lacking, and this is

especially true for neuroinvasive West Nile virus (WNV) infections. WNV infections are the

leading cause of acute viral encephalitis epidemics and related illnesses in North America,

with a significant rate of mortality observed in clinical cases. Studies of experimental WNV

infection in mice have established the key role played by both innate and adaptive immune

responses in clearing virus from the CNS and limiting neuronal death. Neuroinflammation of

peripheral immune cells is a vital component of WNV clearance, but any role for microglia

to initiate immune responses or to act as effector immune cells is unclear.

The present thesis details work towards the establishment and utilization of unique

model systems to investigate microglia activation in response to WNV infection, and the

potential consequences of this activation. An ex vivo slice culture model system for studying

WNV infection was used to determine the potential for innate immune activation by microglia, with many examples of morphological changes, chemotaxis, and phagocytosis.

Minocycline was administered to WNV-infected slice cultures to inhibit the pro- inflammatory activation of microglia, with notable reductions in relevant pro-inflammatory cytokines/chemokines and elevations in anti-inflammatory cytokines/chemokines that could

iii be attributed to microglia production. Finally, an in vivo microglia depletion model for

WNV infection was used through the administration of the CSF1R kinase inhibitor

PLX5622. WNV infection in mice treated with PLX5622 lead to decreased survival and significantly increased WNV infection of the CNS compared to control mice, with deficits in gene expression for CCL5 and TREM2 highlighting the ability to compare the results between the slice culture and in vivo model systems. Overall, this study has made significant gains in the understanding of how microglia can act as effector immune cells during WNV infection and their importance for necessary antiviral responses.

The form and content of this abstract are approved. I recommend its publication.

Approved: Kenneth L. Tyler

Dedicated to my parents

v ACKNOWLEDGEMENTS

I would like to thank the Kenneth L. Tyler lab for allowing me the space, time, and

resources to address experiments that had little to no precedent in the manner I thought best.

The experience has been extremely educational. Many thanks to the staff of the animal

facilities for maintaining healthy mice for our experiments, allowing us to work with

dangerous viruses on the premises, and monitoring the day-to-day work that allowed us the ability to do our research. Special thanks to Plexxikon for providing us with PLX5622 per the material transfer agreement.

This research would not have been possible without funding from several sources.

The Tyler lab was funded by the National Institute of Health (NIH) grants R01 NS076512 and R21/R33 AI101064, as well as a VA merit grant. Additional funding support for myself came from T32 HD041697-12 (2012-2013) and T32 AI052066-13 (2015-2016). The Tyler lab IACUC protocol number was B-34716(03)1E.

A final thanks goes to S. Rock Levinson, who taught me much about sample and antibody preparation that contributed greatly to the work in this dissertation. Many hours were spent on his Nikon confocal scope which he allowed me to use as my personal scope for several years and allowed me to take amazing images of microglial phagocytosis.

vi TABLE OF CONTENTS

CHAPTER

I. INTRODUCTION

Microglia During Development and Homeostasis…….…………………1

Microglia During Disease and Viral Infections………….………………2

West Nile Virus (WNV) Neuroinvasive Disease…………..…………….6

Microglia and WNV: Project Outline……………………………………9

II. AIM 1: MICROGLIA ACTIVATION IN A SLICE CULTURE MODEL

SYSTEM OF WEST NILE VIRUS INFECTION OF THE CENTRAL NERVOUS

SYSTEM

Introduction……………………………………………….…………….13

Materials and Methods………………………………………………….15

Results…………………………………………………………………..22

Discussion...………………………………………………………….…42

III. AIM 2: MINOCYCLINE ADMINISTRATION AND MICROGLIA

INHIBITION IN A SLICE CULTURE MODEL SYSTEM OF WEST NILE VIRUS

INFECTION OF THE CENTRAL NERVOUS SYSTEM

Introduction………………………………………………………….…50

Materials and Methods…………………………………………………52

Results………………………………………………………………….54

Discussion...……………………………………………………………63

vii IV. AIM 3: PLX5622 ADMINISTRATION AND MICROGLIA DEPLETION

DURING WEST NILE VIRUS INFECTION IN VIVO

Introduction…………………………………………………...…….…71

Materials and Methods……………………………………...…………75

Results…………………………………………………...…………….78

Discussion...……………………………………………………...……88

V. CONCLUSIONS AND FUTURE DIRECTIONS.

Microglia Phagocytosis and Antiviral Activity During WNV Infection of the

Central Nervous System….…….…………………………………...…92

Future Directions: Slice Culture Model System….……………...…….94

Future Directions: PLX5622 Model System of WNV Infection…..…..95

REFERENCES…………………………………………………….…………..99

APPENDIX……………………………………………………………………115

viii CHAPTER I

INTRODUCTION

Microglia During Development and Homeostasis

Within the mammalian central nervous system (CNS) reside the cells that control the body as a whole. The primary functional cells of this control mechanism are neurons, which are a largely post-mitotic population that require high levels of specialized support both from their environment and the support cells around them. The environment of neurons is maintained by the blood-brain barrier (BBB), which primarily serves to protect the neurons from infection, cytotoxic inflammatory cells, and noxious components in the blood. Within the BBB is the parenchyma of the CNS, which includes neurons and the support cells that assist neuronal function, generally known as glia.

The three major glia cell types in the CNS are astrocytes, oligodendrocytes, and microglia. In addition to providing trophic support, each of these cell types perform specific functions to create a proper environment for neuronal function. Astrocytes maintain the blood brain barrier and aid in the clearance and reuptake of neurotransmitters from the cerebrospinal fluid (CSF). Oligodendrocytes myelinate axons and provide structural support for neurons. Microglia provide innate immune surveillance and are the primary phagocyte of the CNS.

When the nervous system begins development, the precursor cells of the neuroectoderm that give rise to neurons also are the ancestral source of astrocytes and oligodendrocytes. Microglia originate from macrophage precursors in the yolk sac (Ginhoux

2010, Schulz 2012, Alliot 1999, Prinz and Mildner 2011), and then later migrate to the developing CNS to provide support via clearance of apoptotic cells (Marin-Teva 2004,

1 Wakselman 2008, Paolicelli 2011). Subsequent to the arrival of microglia to the developing

CNS, the blood-brain barrier begins to form and eventually encompasses the CNS parenchyma, closing it off from the rest of the body. Microglia form a self-sustaining population that is not renewed by peripheral macrophages or other blood-borne cells (Hoeffel

2012). This distinction makes microglia particularly different from peripherally-derived macrophage populations, including not only other tissue resident macrophages (e.g.,

Langerhans cells in the skin) but also from other CNS-related macrophages, including perivascular and meningeal macrophages (Jung and Schwartz 2012, Neumann and Wekerle

2013, Bessis 2007). These considerations highlight the fact that microglia are highly attuned to function within the CNS environment, and have a range of responsibilities that differ substantially from their peripheral counterparts.

Under homeostatic conditions, microglia have a much different set of responsibilities compared to peripheral macrophages and resident tissue innate immune cells. Microglia maintain their own niche spaces within the CNS and are uniformly distributed (Rezaie and

Male 2002). Thin cellular processes extend out from microglia in all directions and survey the microenvironment in a highly dynamic manner (Nimmerjahn 2005, Davalos 2005). In addition to their surveillance functions, microglia have been shown to perform a process known as synaptic pruning, where synapses are removed from the dendrites of neurons and thought to be a part of the broader paradigm of neuronal plasticity (Schafer 2012, Tremblay

2010).

Microglia During Disease and Viral Infections

As the resident innate immune cells of the CNS, microglia have the responsibility to detect noxious stimuli that pose a threat to the nervous system. While the catalogue of all the

2 potential insults that could apply to the CNS is large, they can broadly be categorized as traumatic, ischemic, neurodegenerative, neoplastic, immunological, or infectious. For every one of these types of insult, microglia are observed to respond with specific, stereotyped morphological changes that include the retraction of their thin surveillance processes, an increase in cell size, and a more amoeboid-like appearance; this is broadly (and perhaps crudely) deemed “microglia activation.” Another well-characterized component of the microglia activation profile is the expression of inflammatory cytokines and chemokines, reactive oxygen species, and/or trophic factors (Streit 1999, Rossum and Hanisch 2004, Wei

2013), with the specific set of factors produced being dependent upon the type of insult the microglia encounter (Ransohoff and Perry 2009, Kettenmann 2011).

The change in morphology that characterizes microglia activation harkens to their macrophage ontogeny, as does microglial expression of the protein ionized calcium-binding adapter molecule-1 (Iba1), which was originally discovered in macrophages (the gene name for Iba1 is AIF1, or allograft inhibitory factor 1) (Ito 1998, Ito 2001, Deininger 2002). Iba1 has become one of the most prominent markers for microglia immunostaining, and it functions by aiding in actin cytoskeletal rearrangements (Hirasawa 2005); when microglia activation occurs and the cell begins to dramatically alter its physical structure and size (and potentially migrate towards sites of injury/infection), Iba1 expression is increased, and this metric is often used as an indication that microglia have detected a problem and have begun innate immune processes.

The predominant function ascribed to microglia when their morphology is altered and

Iba1 expression increased is the production of pro-inflammatory cytokines and chemokines, across the spectrum of disease models (Block 2007, Husemann 2002, Gao 2011, Kettenmann

3 2011, Ransohoff and Perry 2009, Burguillos 2011, Neher 2013, Cunningham 2005, Butovsky

2005, Schwartz 2006, Pais 2008, Sivagnanam 2010, Lambersten 2009). Microglia themselves have cytotoxic capabilities given their ability to produce iNOS and reactive oxygen species (Newell 2007, Ueyama 2004, Barger 2007, Qin 2004).

Given these abilities, microglia are thought to be important mediators of neuroinflammation and capable of pro-inflammatory cytotoxic functions. However, neurons can be significantly damaged by these very cytotoxic mechanisms, and prolonged or chronic pro-inflammatory activation within the CNS has been directly related to underlying pathology that comprises many aspects of different CNS diseases (Ekdahl 2003, Liu 2007,

Hoehn 2005, Monje 2003). One line of thinking posits that microglia, as specialized innate immune cells that live among neurons, are more attuned to provide neuroprotective support and produce anti-inflammatory factors (in addition to their pro-inflammatory functions) compared to peripheral immune cells that have invaded the CNS from the periphery (Graeber and Streit 2010, Sierra 2013, Butovsky 2001). While evidence has been mounting on the anti-inflammatory abilities of microglia in certain disease models (Kumar Jha 2015, Orihuela

2015, Schwartz 2006, Streit 2002), it remains to be fully appreciated how pro- and anti- inflammatory events are regulated during CNS diseases, and if microglia do in fact have a regulatory role in these events. Much concerning their overall contribution for several CNS diseases remains a topic of intense debate (Sierra 2013, Loane and Kumar 2015, Wei 2013,

Martinez and Gordon 2014, Ransohoff 2016).

While many gains have been made in the understanding of how microglia activation relates to pathogenesis, inflammatory regulation, and neuroprotection, most of this knowledge has come from research pertaining to non-infectious models of CNS disease (e.g.

4 trauma, ischemia, neurodegeneration, etc.). The innate immune mechanisms employed for detection of different CNS insults, as well as the specific course of pathology determined by those insults, influence how microglia respond and what their activation means towards alleviating the problem.

For infectious diseases of the CNS, little is known concerning microglia activation and innate immunity. This is further compounded by the fact that innate immune mechanisms for bacterial infections (small cells that invade the CNS) are not necessarily the same as those for viral infections (small genomes and proteins that invade CNS cells themselves) as detection will occur through different pattern recognition receptors (Olson and Miller 2004, Hadas 2010, Ribes 2010, Peppoloni 2011, Kochan 2012, Collin 2013).

Given that the majority of what is known about microglia activation during infections has been found through experiments using lipopolysaccharide (LPS), a Gram-negative bacterial cell wall component, means that there is still much to learn about microglia responses to viral infections.

To compound further the lack of knowledge concerning microglia and viral infections, viruses have a wide range of genomic forms, tropisms (targets of infection), and consequences for their host cells (cell death, latency, etc.), so extrapolating the role of microglia for one type of viral infection may have no relevance for another. Taking all of these considerations together, and the field of microglia activation for specific CNS viral infections becomes esoteric. Therefore, there is much to learn about how microglia activation influences pathogenesis and antiviral mechanisms for many neuroinvasive viral infections.

5 West Nile Virus Neuroinvasive Disease

West Nile virus (WNV) is a member of the flavivirus family and has a single- stranded, positive-sense RNA genome that is approximately 10.3kb long, and encodes a single polyprotein that is processed/cleaved to eventually produce three structural and seven nonstructural proteins. WNV has the ability to replicate in cells of the mosquito (the primary transmission hosts) and certain birds (the primary amplification hosts), as well as humans and horses (incidental or “dead end” hosts) and potentially many other mammals (Hubalek and

Halouzka 1999). The transmission between mosquitoes and birds represent the enzootic cycle, and the movements of these vectors enable the geographic spread of WNV.

WNV was first discovered in a sick individual in the West Nile region of Uganda in

1937 (Smithburn 1940), and for the majority of time since then was not a major cause of disease until the late 1990s when several large outbreaks occurred with significant severity across several geographic regions, including the United States when the virus was introduced to North America via the New York City area in 1999 (MacKenzie 2004). Within a few years, WNV had spread across the contiguous United States, and in 2002 and 2003 a sharp increase in WNV neuroinvasive disease occurred with over 2800 cases reported across the country in each of those years (O’Leary 2004, Petersen and Hayes 2004), in part due to a different strain taking prevalence that had several amino acid substitutions which acted to increase its viral transmission in mosquitoes (Kilpatrick 2008, Davis 2005). Since then the annual number of WNV neuroinvasive cases has fluctuated, but a mortality rate of approximately 10% has remained constant and WNV continues to be the leading cause of epidemic acute viral encephalitis in the United States (ArboNET, Center for Disease Control and Prevention).

6 For most of the general population, WNV infection does not cause any problems and

the immune system can adequately deal with the virus. For some people, WNV infection can

cause flu-like symptoms (including headache, diarrhea, nausea, vomiting, etc.) that may

require hospitalization, but rarely does death occur when there is a lack of a neurological

infection component (Zou 2010). Neuroinvasive WNV disease is most likely to occur in

older, immunocompromised individuals (Lindsey 2010, Lindsey 2012, Carson 2012),

especially those who have undergone organ transplantation (Nett 2012). WNV

neuroinvasive disease includes West Nile meningitis, encephalitis, acute flaccid paralysis, or

combinations of these three. Clinical manifestations can vary greatly depending on the exact

source of WNV infection in or near the CNS, but some of the most common elements

include headache, ataxia or movement disorders, and altered mental status (Debiasi and Tyler

2006, Hayes 2005). Recovery for survivors of WNV neuroinvasive disease is often poor

(Pepperell 2002, Emig 2004, Klee 2004), highlighting the debilitating nature of these

infections on the CNS.

Wild-type and genetic knock-out mouse models of WNV infection have been

extensively utilized to better understand WNV pathogenesis, and have been useful for

establishing WNV tissue tropism (Kramer and Bernard 2001, Byrne 2001, Lim 2011,

Schneider 2008) and pathology associated with WNV neuroinvasion (Shrestha 2003).

Typically mice are infected via intramuscular inoculation to reflect the mosquito-borne route of viral entry into the body, with WNV actively infecting lymph nodes within the first few days post-infection (Shrestha 2006b). While much concerning the specific attributes of infection can be dependent upon many factors including virus strain and quantity (Donadieu

2013) and the genetic background of mice (Graham 2015), WNV viremia peaks around 3-4

7 days post-infection and infects a range of susceptible organs, including the liver, spleen, and kidneys (Shrestha 2004). Usually between 5-8 days post-infection, WNV peripheral infection and viremia drops as antiviral immune responses clear virus while neurological infection takes hold (Shrestha 2006b).

It has been well established that limiting peripheral WNV infection is the first step to reducing WNV neuroinvasive disease. Experimental animal models that have shown increased WNV viremia strongly correlate to increased likelihood of WNV neuroinvasion and high WNV titers in the CNS; this includes models lacking proper antibody production

(Diamond 2003a, Diamond 2003b) or factors related to antiviral interferon pathways

(Shrestha 2006a, Samuel 2005, Brien 2011, Suthar 2010, Cho 2013). IL-10 genetic knockout mice had increased survival after WNV challenge, and part of this was due to their having reduced WNV viremia (Bai 2009).

Several possible routes of WNV entry to the nervous system have been proposed

(Cho and Diamond 2012); a hematogenous route is thought to be the most likely. In WNV- infected TLR3 knockout mice, mortality was reduced due to less WNV entry to the CNS, despite increased WNV viremia and reduced pro-inflammatory cytokine production in the periphery compared to wild-type controls (Wang 2004). This effect was shown to be mediated by signaling pathways through TNFa receptor 1 (Wang 2004). Other possible routes might be likely to play a role in different circumstances, include infected macrophages or other immune cells crossing the BBB to facilitate viral CNS entry (known as the “Trojan horse” mechanism) (Paul 2017), infection of vascular endothelial cells of the BBB (Verma

2010), or retrograde axonal transport (Hunsperger and Roehrig 2006). Regardless of entry mechanism, what is known is that once WNV reaches the CNS the primary target cells for

8 infection among CNS parenchymal cells are neurons (Shrestha 2003, Cheeran 2005).

Neuronal death is the primary means for the deleterious effects of WNV infection of the

CNS, and caspase-3 cleavage and apoptotic mechanisms are a large contributor towards this; caspase-3 genetic knockout mice have improved survival due to lack of neuronal death

(Samuel 2007).

Once WNV is in the CNS, the importance of proper neuroinflammation for viral clearance has been extensively characterized. Proper recruitment and activation of CD8 T cells to the WNV-infected CNS for mediating viral clearance is essential (Wang 2003,

Shrestha 2004, Shrestha 2008, Brehin 2008), and limiting their cytolytic capabilities allows for unrestrained viral growth in the CNS and increased mortality (Shrestha 2006b, Shrestha

2007, Shrestha 2012). The impairment or deletion of most mediators for innate antiviral responses can cause further deleterious effects to proper pro-inflammatory neuroinflammation and activation of these responses within the infected CNS (Lazear 2013,

Kumar 2013, Durrant 2013, Szretter 2010). Beyond the necessity of these neuroinflammatory responses, however, little is known about how regulation for these factors occurs within the CNS or if the CNS itself has a role in regulating WNV clearance or initiating inflammation.

Microglia and WNV: Project Outline

Despite a number of studies of neuroinflammatory mechanisms of WNV neuroinvasive disease, little to nothing has been definitively attributed to microglia. In vitro cultures of microglia exposed to WNV showed that as a cell population they have little infectivity compared to neurons and astrocytes (Cheeran 2005). In IL-34 genetic knockout mice, which have reduced microglia trafficking to the developing CNS and therefore lack

9 normal microglia development, modest increases in cell death were observed after intracranial infection of attenuated WNV (Wang 2012); however, the IL-34 knockout mice have not proved to be a sufficient model for microglia depletion generally, and intracranial inoculation with attenuated WNV does little to recapitulate actual pathogenesis mechanisms of WNV neuroinvasive disease. In a model of WNV recovery and memory impairment with an attenuated WNV strain, microglia were shown to mediate synapse loss and impaired behavior (Vasek 2016), but these studies did not examine acute pathogenesis for WNV neuroinvasive disease.

Within the WNV-infected CNS, the presence of peripherally-derived neuroinflammatory cells confounds to what extent any measured antiviral responses are mediated by innate microglia; generally when microglia could potentially be behind a certain antiviral response, so too could macrophages that have entered the CNS; these responses are generally categorized and/or attributed to “microglia/macrophages.” In theory, microglia are the first responders to WNV infection in the CNS, and would likely have important roles in the recruitment and activation of peripheral immune cells via pro-inflammatory cytokine/chemokine expression.

Given the importance of restricting WNV from reaching high titers in the CNS and the necessity of robust pro-inflammatory neuroinflammation to clear virus and control virus- induced injury, increasing the efficiency of early inflammatory responses could have a dramatic impact on viral clearance and mortality. Furthermore, while microglia are known to phagocytose apoptotic cells, little is known about their ability to phagocytose virally-infected cells, and this could be another potential antiviral mechanism of microglia activation.

Besides limited imaging evidence of WNV phagocytosis within the CNS by

10 “microglia/macrophages,” this aspect has not previously been explored in detail. In vitro evidence for phagocytosis during WNV infection has been shown for peritoneal macrophages which were exposed to specially-prepared cell lines (Youn 2010, Chung 2007), and these phagocytic mechanisms required the presence of antibodies and opsonization.

Whether microglia could perform these functions as well, or without adaptive immune mechanisms, has not previously been addressed.

To the extent to which microglia provide protection beyond pro-inflammatory antiviral mechanisms (either via pro-survival signals to neurons and other glia, or anti- inflammatory mechanisms to resolve cytotoxicity) is unclear. Given their role to promote homeostasis and protect the neuronal environment, proper neuroinflammatory regulation by microglia would include not only efficient, early pro-inflammatory responses but also the quick resolution of those responses to limit cytotoxicity to neurons. Again, while anti- inflammatory events in WNV have not been investigated extensively in general, no putative role for microglia in related mechanisms has been broached.

The goal of this project was to gain an understanding of the antiviral response capabilities of microglia and the consequences of this antiviral activation to the pathogenesis of WNV infections of the CNS. The hypothesis that drove this research was that microglia have the ability to perform functions that belie its role as an innate immune sentinel cell, including not only stereotypical “microglia activation” morphology, but also phagocytosis of

WNV-infected cells and relevant cytokine/chemokine expression that influences pro- inflammatory (and potentially anti-inflammatory) mechanisms.

To test how microglia activate in response to WNV infection, three main aims were established. First, an ex vivo CNS slice culture model system of WNV infection was

11 established to definitively verify and explore aspects and mechanisms of microglia activation in an environment lacking peripheral immune cell involvement (Aim 1). Second, within this slice culture model system, the ability of microglia to produce relevant pro- and anti- inflammatory cytokines and chemokines was determined to gauge how microglia can potentially influence neuroinflammatory events in the CNS (Aim 2); to achieve this aim, there was extensive utilization of the microglia modulatory reagent, minocycline. The final aim of this project was to assess mortality and viral restriction/clearance during WNV infection in live mice that had significantly depleted microglia, and to explore if any findings in the slice culture model system could be extrapolated to the in vivo setting (Aim 3); the usage of the CSF1R kinase inhibitor PLX5622 to deplete microglia was an important component to achieving this aim. Taken together, addressing these aims would add significantly to the field of knowledge concerning antiviral microglia activation and its ability to influence relevant inflammatory processes, as well as establishing unique model systems for future research into WNV pathogenesis.

12 CHAPTER II

AIM 1: MICROGLIA ACTIVATION IN A SLICE CULTURE MODEL SYSTEM OF

WEST NILE VIRUS INFECTION OF THE CENTRAL NERVOUS SYSTEM

Introduction

Model Systems for Researching Microglia and WNV

To properly investigate the abilities of microglia during WNV infection of the CNS in vivo, it is important to be able to identify their specific activity. However, the ability to distinguish microglia from peripherally-derived macrophages during neuroinflammatory events is not a straightforward endeavor (Sierra 2013, Biber 2014). Microglia and macrophages can both perform similar innate immune functions and express a similar repertoire of identifying surface markers, making definitive identification difficult (Jung and

Schwartz 2012, Walker and Lue 2015). Furthermore, the intrinsic abilities of microglia can be difficult to assess once peripherally-derived macrophages invade the CNS, making the contributions of microglia towards antiviral responses difficult to gauge.

To avoid issues concerning how macrophages or other peripheral immune cells are altering microglia abilities or identification, a culture model system of WNV infection of

CNS tissue would be ideal, but standard in vitro methodology is not satisfactory in this regard. Microglia activation and its subsequent potential for cytotoxicity is held in check by inhibitory cell-cell contacts with neurons, astrocytes, and other cells (Kierdorf 2013, Mott

2004), including the ligand-receptor axes of CD200-CD200R (Broderick 2002, Lyons 2007,

Hoek 2000), CX3CL1-CX3CR1 (Cardona 2006, Sierra 2013), among others (Biber 2014,

Biber 2007, Wolf 2013). This feature of microglia behavior makes standard cell separation techniques suboptimal, even taking mixed culture environments into account, as the basal

13 state of microglia activation is increased by removing these inhibitory inputs, essentially

priming the activation state of these cells before any modeling has occurred and not

accurately reflecting in vivo conditions (Hellwig 2013). Given that microglia should be in as quiescent a state as possible before modeling their response to WNV infection, other culture methods should be explored.

Slice Culture Model System for Researching Microglia and WNV

The utilization of ex vivo slice cultures of brain and/or spinal cord tissue presents itself as an ideal model system for investigating microglia responses to WNV infection. The concept of maintaining the endogenous cytoarchitecture of nervous tissue allows for the majority of cell-cell connectivity to remain in place, which suppresses unnecessary microglia activation as described in the previous section, while giving a better representation of natural cellular interactions compared to mixed cell culture systems.

There is limited precedent for the usage of slice cultures of CNS tissue to investigate mechanisms of virus-induced pathology. For the specific slice culture system employed in this thesis dissertation (as in the specific protocol, techniques, and reagents), previous work was performed on brain and spinal cord slice cultures for research into infections with an experimental neurotropic reovirus (Dionne 2011, Schittone 2012). Results in these studies observed that pathological mechanisms seen in the slice cultures reflected those observed for reovirus infections in vivo, including viral tropism, neuronal and tissue death, and the cleavage of caspase-3 as a major contributor to observed cell death.

Goals

Despite a lack of extensive utilization of slice culture models to investigate viral infections of the CNS, for the specific goal of researching microglia activation during WNV

14 infection, this model system provides outstanding advantages. The goals for this aim of the

project were primarily two-fold: the first was to establish that WNV infections in slice

cultures replicate features of in vivo WNV pathology, including infectivity that is predominantly neurotropic, increased neuronal death, and dependence on caspase-3 cleavage as an underlying mechanism of cell death (Samuel 2007). The second goal was to observe how microglia can activate in response to WNV infection, as determined by stereotypical morphological changes that reflect innate immune mechanisms, including chemotaxis and/or phagocytosis. Data from both brain and spinal cord slice cultures (BSC and SCSC, respectively) will be reported, with the bulk of the data primarily focused on SCSC. The reason for this was to have one consistent model system, and to create more individual cultures per tissue type (25-30 SCSC per spinal cord compared to 4-6 BSC per brain).

Materials and Methods

Brain and Spinal Cord Slice Cultures

All of the following methods used NIH Swiss Webster mice and were performed in

compliance with IACUC protocols and institutional guidelines, under sterile to semi-sterile

conditions.

Spinal cord slice cultures (SCSC). Mice pups (5-6 days old) were euthanized by decapitation and the base of the tail was removed to expose the caudal opening of the spinal column. A 27-29 gauge needle and syringe containing sterile phosphate buffered saline

(PBS) were inserted into the caudal spinal column opening and the spinal cord was extruded forcefully and firmly through the rostral opening of the spinal column into a Petri dish containing either sterile PBS or slicing media (Dulbecco’s modified Eagle’s medium

[DMEM], 10mM Tris, 28mM D-glucose, pH 7.2, equilibrated with 95% O2 and 5% CO2).

15 Once the desired amount of spinal cords were collected (typically 4-8), the spinal cords were embedded in 2% agarose in slicing media and supported with forceps to remain perpendicular while the agarose hardened (placing the plastic embedding mould into a Petri dish filled with ice water could increase the hardening process). Once firm, the agarose cube was removed from the embedding mould and excess agarose around the periphery of the cube was removed with a razor, in a tapered shape with a wide base (cervical ends of spinal cord on the bottom) and thin apex (lumbar/sacral ends of spinal cord on the top). The resultant agarose with perpendicular spinal cords was then glued with wide end on the bottom (cervical end of the spinal cords) to a Vibratome platform (Leica VT1000S). Within the well of the Vibratome, ice cold slicing media was poured in to reach the top of the agarose mount. Transverse sections of spinal cords in agarose were cut on the Vibratome at

400um thickness (speed setting 6-7, frequency setting 7) and collected into a Petri dish with ice cold slicing media. Thoracic and lumbar sections of the spinal cords were the majority of the spinal cord sections collected. Once the Vibratome reached the bottom of the mount platform, typically 25-30 spinal cord sections were collected per spinal cord within the agarose. Spinal cord sections were then removed from the agarose with a disposable plastic pipette and collected into individual wells of a 6-well (35mm) culture plate with ice cold slicing media to aid in counting slices and removing from remnant agarose. Spinal cord slices were then placed onto 30mm, 0.4um pore-sized cell culture membrane inserts

(Millipore) using a disposable plastic pipette, with as much as possible slicing media removed with a Pasteur pipette and bulb. Each insert could have anywhere from 20-120 spinal cord slices, dependent on the experimental requirements.

16 Brain slice cultures (BSC). Mice pups (2-3 days old) were euthanized by decapitation and the brain was dissected out of the skull. The brain was placed into an empty Petri dish and the cerebellum was removed via a coronal cut with a razor blade, creating a flat surface to glue the brain onto a Vibratome platform (flat, caudal end on bottom, rostral end on top).

Within the well of the Vibratome, ice cold slicing media was poured in to reach the top of the brain. Coronal sections of brain were cut on the Vibratome at 400um thickness (speed setting 5, frequency setting 9) and collected into individual wells of a 6-well (35mm) culture plate with ice cold slicing media. Brain slices were then placed onto 30mm, 0.4um pore- sized cell culture membrane inserts (Millipore), with as much as possible slicing media removed with a Pasteur pipette and bulb. Each insert could have anywhere from 2-4 brain slices, dependent on the experimental requirements.

Slice culture conditions and media. Membrane inserts with brain or spinal cord slices were each placed into individual wells within a 6-well (35mm) culture plate with 1.1ml slice culture media (Neurobasal A medium, 10mM HEPES, 400uM L-glutamine, 600uM

Glutamax, 1x B-27, 60ug/ml streptomycin, 60U/ml penicillin, 6U/ml nystatin) supplemented with 10% fetal bovine serum (FBS). The culture plate was placed into a cell culture incubator, then after one day the slice culture media was replaced with fresh media supplemented with 5% FBS and incubated for two days prior to WNV inoculation.

WNV and Slice Culture Infection

West Nile virus stocks were procured from clone-derived strain 382-99 (NY99) as previously described (Brault 2007). By the third day of the slices being in culture (one day with 10% FBS supplement, following two days with 5% FBS supplement), slice culture media was again changed, with no FBS present. Culture plates were transported to the BSL3

17 facility and via their apical surfaces were inoculated with WNV NY99 strain at a range of

104-105 PFU/slice within 20ul-50ul of slice culture vehicle (dependent on amount of tissue present). After 12 hours, slices were washed with culture media which was withdrawn with a

Pasteur pipette and bulb to remove excess virus from the tissue. Every two days subsequent to the day of infection, slice culture media was replaced with fresh media until tissue collection.

Experimental Methods

Fluorescent immunohistochemistry (IHC). Membrane inserts with slice cultures were moved to a new 6-well (35mm) plate and washed in PBS for 30 minutes (below and above membrane; total volume approximately 2-3ml) before fixation with 10% neutral buffered formalin (NBF) for at least one hour, but up to several days if further time points were to be collected and processed together (also below and above membrane). Fixed slice cultures were washed in PBS for at least 30 minutes before the inserts were drained and the membrane containing the slice cultures cut out from the inserts with a scalpel. Slice cultures with attached membrane were then immersed in permeabilization/blocking solution (PBS,

4% normal goat serum, 2% bovine serum albumin, 0.3% Triton-X 100) in individual wells of a 24-well (16mm) culture plate for at least one hour. Slice cultures were then incubated overnight at room temperature in primary antibodies diluted in permeabilization/blocking solution. Primary antibodies used were mouse anti-WNV E protein (1:200, ATCC), rabbit anti-Map2 (1:100, Millipore), rabbit anti-Iba1 (1:500), and rabbit anti-GFAP (1:1000,

Abcam). The next day slice cultures were washed in PBS three times (approximately 10 minutes per wash with light agitation/rotation) before incubation for two to three hours in secondary antibodies, also diluted in permeabilization/blocking solution. Secondary

18 antibodies used were AlexaFluor goat anti-mouse 568 and AlexaFluor goat anti-rabbit 488

(1:1000, Invitrogen). Slice cultures were again washed in PBS three times in a similar fashion before lightly being rinsed with water and lightly dried by removing excess fluid with a Pasteur pipette and bulb and tissue paper. Slice cultures (with the membrane) were mounted onto microscope slides with Prolong Gold or Diamond antifade reagent (Molecular

Probes) and coverslipped. Slides were imaged using a Nikon PCM-2000 laser scanning confocal microscope, and image procurement and processing was done with SimplePCI v4.6 software (Compix). Oil immersion objectives (with numerical aperture) used were 40x/NA

1.3, 60x/NA 1.4, and 100x/NA 1.45.

WNV PCR. Slice cultures were removed from insert membranes and homogenized in

RLT buffer (Qiagen) containing 1% B-mercaptoethanol, then loaded into RNeasy spin columns (Qiagen) for RNA purification. Purified RNA was collected from the spin columns following manufacturer protocols. RNA integrity was assessed using an Agilent 2100 bioanalyzer (Agilent) and required an A260/280 value greater than 2.0 to be high quality.

RNA concentrations were equalized and used to make cDNA using either SuperScript Vilo kit (Invitrogen) or iScript (Bio-Rad) following manufacturer directions. cDNA was mixed with 1ul/well of primers for WNV E protein and 2x iTaq mastermix (Bio-Rad) for a total volume of 20ul. Samples were loaded in triplicate into 96 well PCR plates and PCR amplification was performed with a CFX96 thermocycler (Bio-Rad). Equivalent plaque- forming units (ePFU) were quantified by taking WNV-infected brain lysates of varying levels of infection and calibrating the specific PFU titer for each to create a standard curve ranging from 104 to 108 which was used to determine the WNV RNA load of SCSC samples.

19 Cleaved caspase-3 activity assay. Slice cultures were removed from insert membranes and both homogenized and sonicated in RIPA lysis buffer (Cell Signaling). Slice culture lysates were prepared and diluted following manufacturer instructions for a caspase-3 fluorometric assay (B&D Bioscience) and loaded into a 96-well plate. Caspase-3 cleavage was assessed with a synthetic peptide (Ac-DEVD-AFC) that is converted to the fluorescent reporter molecule (AFC) via cleavage by activated caspase-3, which was added to lysates, and the reaction product was measured on a Cytofluor 4000 spectrometer (Applied

Biosystems) at an excitation wavelength of 450nm and emission wavelength of 530nm.

Western blot. Slice cultures were collected in PBS and then centrifuges at 2000rpm to obtain a pellet of tissue which was titurated with RIPA lysis buffer (Cell Signaling) containing protease inhibitor cocktail (Active Motif) and DTT. After homogenization and sonication, lysates were mixed with 5x Laemmli buffer and boiled for 2 minutes before loading onto 12% Tris-polyacrylamide gels which ran overnight at 65V. Proteins were transferred onto nitrocellulose blots and blocked for 1 hour with 5% dry milk in 1% TBS-

Tween (TBST). After blocking the blots were put into primary antibody solutions and incubated overnight at 4 degrees. Antibodies were diluted with 2% dry milk in TBST.

Primary antibodies used were rabbit anti-cleaved caspase-3 (1:1000, Cell Signaling) and rabbit anti-beta actin (1:3000, Cell Signaling). Following 3 washes (10 minutes each) in

TBST, blots were incubated in secondary antibody solutions for 2 hours at room temperature.

Secondary antibodies used were HRP-conjugated goat anti-rabbit IgG (1:10000, Jackson

Labs). After 3 washes in TBST (10 minutes each) the blots were developed using

SuperSignal West Pico chemiluminescent substrate (Thermo Scientific) and imaged on a

20 FluorochemQ MultiImage III workstation. Image analysis and processing was performed

with Alphaview v3.0 software (Alpha Innotech).

MTT assay and densitometry. MTT (3-[4,5-dimethylthiazol-2-yl]-2,5- diphenyltetrazolium bromine) is a yellow tetrazolium salt that is converted to a purple formazan crystal via mitochondrial mechanisms in metabolically-active cells. This allows

for qualitative, colorimetric assessment of tissue health. 0.5mg/ml MTT (Roche) was added

to fresh slice culture media and incubated for 30 minutes at 42 degrees, then the reaction was

stopped via fixation with 10% NBF. Imaging was performed with a ScanMaker 8700

scanner (Microtek) and image processing was performed with ScanWizard Pro v7.10

(Microtek). MTT images were analyzed using ImageJ software (NIH). Color images were

converted 16-bit grayscale formats and inverted. The free-hand selection tool was used to

trace the outline of individual slice culture samples, then the mean gray value was measured

for each sample (n=8 per condition). Background was determined by measuring regions not

containing any sample, and this value was subtracted from the overall mean gray value

measures for each sample.

Q-VD-OPh treatment. Q-VD-OPh (Biovision) is a pan-caspase inhibitor comprised of a carboxyl terminal phenoxy group conjugated to the amino acids valine and aspartate, and potently blocks apoptotic signaling through caspase-9/3, caspase8/10, and caspase-12 mechanisms. Q-VD-OPh was diluted to 500ug/ml in DMSO vehicle and applied directly to the apical surface of slice cultures at the time of infection and every media change.

Quantification of cell infectivity. Fluorescent IHC images were compiled from infected slice cultures imaged with 20x and 60x objectives. Individual cells that were labeled with antibodies against Map2 (neurons), GFAP (astrocytes), or Iba1 (microglia) were labeled

21 as either positive or negative for WNV envelope protein antibody staining using Microsoft

Paint software. These cells were counted and the values were recorded and tabulated as a

percentage of infected cells per image field of vision (FOV). Each FOV was taken from

similar anatomical regions across multiple slice culture samples for each condition.

Quantification of microglia morphological characteristics. Fluorescent IHC images

were compiled from mock-infected and WNV-infected slice cultures images with 60x

objective for cell size measurements, 40x objective for Iba1 pixel intensity, and both 40x and

60x objectives for cell state (e.g., amoeboid, quiescent, etc.) measures. ImageJ software

(NIH) was used to trace cell perimeters to measure pixel intensity and cell area. Microsoft

Paint was used to tabulate amoeboid cells.

Statistical analyses. All statistics were calculated using InStat and Prism software

(GraphPad). The Mann-Whitney t-test was used for MTT densitometry analysis, microglia

activation characterization, analysis of infectivity percentages, and caspase-3 assay.

Results

WNV Growth, Tropism, and Cell Death

To characterize WNV infection of slice cultures, spinal cord slice cultures (SCSC)

were infected with 105 PFU/slice, with individual SCSC samples collected at 12 hours post-

infection and then daily up to seven days post-infection (7dpi). Viral growth was determined

using RT-PCR detection of WNV RNA (Figure 2-1). By 2dpi the viral load had increased

1000-fold from the 12 hour time point, indicating growth within the cultures. From 3dpi to

7dpi the viral load remained relatively constant between 107 and 108 PFU equivalents

(ePFU), indicating that WNV infection had established itself in the slice cultures.

Figure 2-1. WNV RNA detected in single SCSC over 7 day time course, ran in triplicate. From Quick 2014, J Virol 88(22) (Fig.1A).

23 To further characterize WNV infection in slice cultures, the tropism of WNV infection was determined in order to understand which specific cell types were infected.

SCSC were analyzed via IHC at 3dpi and 6dpi for the presence of cytoplasmic WNV envelope protein (WNV-E) antibodies within different CNS cell populations. Neurons, astrocytes, and microglia were identified using antibodies against Map2 (microtubule associated protein 2), GFAP (glial fibrillary acidic protein), and Iba1 (ionized calcium- binding adapter molecule 1), respectively.

WNV-E antigen was most commonly observed in neurons (Figure 2-2). At 3dpi, approximately 27% of neurons were positive for WNV-E protein, and this value increased to over 90% by 6dpi. The increase of WNV infection from 3dpi to 6dpi in neurons was very statistically significant (p < 0.001, n = 416 neurons assessed). Astrocyte infection by WNV was less common, with approximately 11% of astrocytes positive for WNV-E protein at 3dpi and 29% WNV-E+ at 6dpi (n = 211). Microglia proved to have little to no cytoplasmic

WNV-E staining at either time point; 6% of microglia displayed some WNV-E protein at

3dpi, and less than 1% at 6dpi (n = 365).

Cytoplasmic distribution of WNV was readily apparent in WNV-infected neurons, with the nuclei absent of any WNV-E (Figure 2-3). At later time points in the slice cultures

(typically around 5-7dpi) neuronal loss due to WNV infection was substantial. In cortical regions of BSC imaged at 3dpi, 5dpi, and 7dpi, neuronal WNV infection can be observed to increase between 3dpi to 5dpi; by 7dpi, Map2 expression is virtually undetectable, indicating neuron death (Figure 2-4). Similar loss of Map2+ neurons was also observed in SCSC infected with WNV at 5dpi when compared to mock-infected samples (Figure 2-5). In this example, individual SCSC were imaged at low-magnification, with the mock-infected sample

Figure 2-2. The percentage of cells infected for populations of neurons (n = 416 per time point), astrocytes (n = 211 per time point), and microglia (n = 365 per time point), at 3dpi and 6dpi, per field of vision (FOV). Asterisks indicate statistical significance (***, P<0.001; unpaired student t-test). From Quick 2014, J Virol 88(22) (Fig.1B).

Figure 2-3. Neurons labeled with MAP2 (green) are shown at 60x original magnification in mock-infected and WNV-infected samples. Cytoplasmic distribution of WNV-E (red) in infected samples is seen in each neuron in the field of view. Scale bars: 30um. From Quick 2014, J Virol 88(22) (Fig.1C,D).

Figure 2-4. Neuronal death from WNV infection in BSC from 3dpi to 7dpi. MAP2 (green) and WNV-E (red) in the cortex region. Scale bars: 60um. Most panels from Clarke 2014, J Virol 88(2) (Fig. 3A,C).

Figure 2-5. Neuronal loss in SCSC at 6dpi. MAP2 (green) and WNV-E (red) within single SCSC samples. Mock-infected SCSC (left panel) show a wide distribution of neurons, with stronger signal in the dorsal horn regions (top). In WNV-infected SCSC sample there is a clear loss of neurons in the top half of the slice culture, and remaining neuronal cell bodies are apparent in the middle of the slice. Scale bars: 500um.

27 displaying full coverage with Map2 staining; the WNV-infected SCSC shows nearly half of

its tissue lacking Map2 signal, indicative of significant neuronal loss.

While neurons were clearly the most infectable cell type in BSC and SCSC,

astrocytes would sometimes display cytoplasmic WNV-E and could be classified as infected

(Figure 2-6). However, the most prevalent feature of astrocytes in WNV-infected slice cultures was thickened cellular processes and increased GFAP expression, the hallmarks of astrogliosis (Figure 2-7). Like neurons, astrocytes were also prone to substantial levels of cell death during WNV infection in slice cultures. In the cortex region of BSC at 3dpi, when

WNV infection has not begun to cause apparent damage, astrocytes do not appear much different compared to astrocytes in mock-infected samples, but by 5dpi considerable astrogliosis and astrocyte loss is observed (Figure 2-8). This was true in other regions of

BSC, including the hippocampus (Figure 2-9).

To assess if the cell death observed in slice cultures was coincident with caspase-3 cleavage as previously reported (Samuel 2007, Michaelis 2007), Western blots and cleaved caspase-3 activity assays were performed on lysates from WNV-infected SCSC at 6-7dpi and compared to mock-infected SCSC at the same time points. Western blots from SCSC collected at 7dpi show a qualitative increase in detectable cleaved caspase-3 (Figure 2-10), with a similar result with cleaved caspase-3 activity assay at SCSC samples collected at 6dpi

(Figure 2-11). The difference in the activation of cleaved caspase-3 between mock and

WNV-infected SCSC was significant (p = 0.003).

To gauge the contribution of caspase-3 cleavage to WNV-induced cell death, SCSC infected with WNV were treated with the pan-caspase inhibitor Q-VD-OPh and then their tissue death was compared to vehicle-treated, WNV-infected SCSC via MTT assay at 6dpi

Figure 2-6. Astrocyte infection by WNV in SCSC. Left panel: Z-stack of an infected astrocyte. Right panel: a single plane from the z-stack showing cytoplasmic distribution of WNV-E (red) within GFAP (green) staining. Scale bars: 30um.

Figure 2-7. Astrocytes labeled with GFAP (green) are shown at 60x original magnification in mock-infected and WNV-infected samples. Astrogliosis is observed in WNV-infected samples. Scale bars: 30um. From Quick 2014, J Virol 88(22) (Fig1E,F).

Figure 2-8. Astrocyte loss in cortex region of BSC, 3dpi to 5dpi. GFAP (green) appears similar at 3dpi for mock-infected and WNV-infected BSC. At 5dpi, WNV-E (red) has increased in infected samples, with a loss of astrocytes apparent. Scale bars: 60um.

Figure 2-9. Astrocyte cell death during WNV in BSC hippocampus at different magnifications. Scale bars: 120um (10x panels), 60um (20x panels), 20um (60x panels).

Figure 2-10. Western blots were used to determine the amount of cleaved caspase-3 present in lysates prepared from mock-infected and WNV-infected SCSCs at 7dpi. β-actin (42KDa) is present as a loading control. WNV-infected SCSC have increased cleaved caspase-3 staining at 17KDa and 19KDa compared to mock-infected samples. From Quick 2014, J Virol 88(22) (Fig.2B).

Figure 2-11. A fluorogenic caspase-3 activity assay was used to assess apoptosis in SCSC lysates at 6dpi, comparing mock-infected and WNV-infected samples. WNV-infected SCSCs had increased caspase-3 activity compared to mock-infected SCSCs. Asterisks indicate statistical significance (**, P=0.003; unpaired student t-test). From Quick 2014, J Virol 88(22) (Fig.2A).

33 (Figure 2-12). MTT staining in mock-infected SCSC showed that the entirety of SCSC were metabolically active, while in vehicle-treated SCSC infected with WNV there was a significant loss of MTT staining, indicating cell death. However, MTT staining in WNV- infected SCSC treated with Q-VD-OPh had a significant reduction in cell death compared to their vehicle-treated WNV-infected counterpart samples (p = 0.015).

Microglia Morphology and Phagocytosis During WNV Infection

To assess microglia activation during WNV infection, observations of microglia morphology were conducted utilizing IHC against the Iba1 and WNV-E antibodies, primarily at 6dpi when WNV infection of neurons was very high (Figure 2-13). In mock infected BSC and SCSC, microglia displayed low-level expression of Iba1 and had thin cellular processes extending out radially, indicative of quiescent status. Microglia in WNV-infected BSC and

SCSC displayed higher levels of Iba1 staining intensity with distinct large amoeboid cellular processes. Amoeboid microglia were characterized by large cellular processes directed against WNV antigen and displayed cell shapes that strongly suggest chemotactic motility, including lamellipodia and stretched cell bodies. Microglia were often able to produce high amounts of filopodial structures and stretch lamellipodial structures relatively long distances within BSC and SCSC tissue, and Iba1 staining intensity was often highest within cells in filopodial and lamellipodial structures, indicative of active cytoskeletal rearrangements

(Hirasawa 2005, Lee 2007). Metrics of microglia morphological activation in WNV-infected

SCSC were quantified (Figure 2-14) and observed to be significantly elevated compared to microglia in mock infected SCSC, including larger cell size (p < 0.0001), increased Iba1 pixel intensity (p = 0.0002), and a higher percentage of cells in an amoeboid state (p = 0.02).

Figure 2-12. Tissue health was determined with MTT staining in mock-infected and WNV- infected SCSCs at 7dpi, with significant cell death occurring in WNV-infected samples. Caspase inhibition rescued WNV-infected SCSCs from considerable tissue death. Vehicle = DMSO; Pan-caspase inhibitor = Q-VD-OPh. Densitometry via pixel intensity measures (ImageJ) of MTT images show significant tissue death from WNV infection and rescue from tissue death with caspase inhibition. For each condition, n = 8. Asterisks indicate statistical significant (***, P < 0.001; *, P < 0.05; unpaired student t-test). From Quick 2014, J Virol 88(22) (Fig.2C,D).

Figure 2-13. Mock-infected and WNV-infected SCSC were collected at 6dpi and processed for immunohistochemical staining. Iba1 staining (green) of mock-infected SCSC reveals quiescent microglia and low Iba1 signal. Microglia in WNV-infected SCSC show enlarged cellular projections (lamellipodia) directed towards WNV-E (red) cells and debris. Scale bar: 60um. From Quick 2014, J Virol 88(22) (Fig.3A).

Figure 2-14. Quantifications of microglia activation-associated via morphological characteristics at 6dpi. Pixel intensity of Iba1 staining shows increased Iba1 expression in WNV-infected SCSC compared to mock-infected controls (left panel; n=20). Microglia cell size is increased for WNV-infected SCSC compared to mock-infected samples (middle panel; n=49). Microglia cells displaying amoeboid characteristics (motile functions, large cellular projections, abnormal shape) were more prevalent in WNV-infected SCSC compared to mock-infected samples (right panel; n=69). Error bars represent the standard error of the mean. Asterisks indicate statistical significant (***, P < 0.001; *, P < 0.05; unpaired student t-test). From Quick 2014, J Virol 88(22) (Fig.3B).

36 A wide range of microglia phagocytic abilities were apparent at 6dpi in WNV- infected BSC and SCSC. Several stages of WNV phagocytosis were be observed within

WNV-infected SCSC (Figure 2-15), including the process of engulfment of a WNV+ cell with a large portion of a microglia cell body (short, thin arrow), the formation of a phagosome structure around another WNV+ cell by a different microglia cell (long, thin arrow), and a WNV+ cell that has been completely phagocytosed by yet another large microglia cell (short, fat arrow). Large vacuolar structures within the microglia can also be observed in proximity to WNV+ cells that are undergoing phagocytosis, which may be lysosomal in nature and part of the assembly towards phagolysosomal structures (Sierra

2013). Subcellular phagocytosis also observed against WNV-E+ debris (Figure 2-16), highlighting how microglia can maintain their environment by removing even very small amounts of WNV-E+ antigenic debris.

Many distinct microglia cell processes related to cell motility and phagocytosis of

WNV-infected cells and debris in BSC and SCSC were observed at a resolution not previously described (Figure 2-17). Microglia were depicted performing a wide range of activity; some microglia cells attempt to surround multiple cells (Figure 2-17E), while others phagocytosis subcellular WNV+ debris (Figure 2-17C). Striking examples of phagocytic structures were also observed, including the “phagocytic cup” (Figure 2-17B) and a phagosome with a WNV+ cell enclosed (2-17F).

Filopodial structures were often observed in contact or near contact with WNV-E+ cells and debris (Figure 2-17A). Sometimes microglia could produce a striking amount of filopodia as they assessed and searched for WNV-E cells (Figure 2-18), and could produce them for seemingly independent detection processes while appearing to be chemotaxing to a

Figure 2-15. Mock- and WNV-infected samples were collected at 6dpi and stained for Iba1 (green) and WNV-E (red). Mock-infected SCSC (left) display low levels of Iba1 staining. WNV-infected SCSC (right) show increased expression of Iba1 as well as phagocytic functions performed by microglia. Examples of microglial phagocytic activity include partial engulfment of a WNV-E+ cell (short, thin arrow), the formation of a phagosomal compartment with fully engulfed WNV-E+ cell (long, thin arrow), and a WNV-E+ cell within a microglia cell (short, thick arrow). Scale bar: 30um. From Quick 2014, J Virol 88(22) (Fig.4).

Figure 2-16. Microglia phagocytosis of WNV+ debris in BSC, with merged and individual color channel panels for Iba1 (green) and WNV-E (red). Scale bar: 10um. Top panel from Clarke 2014, J Virol 88(2) (Fig. 3E).

Figure 2-17. High-magnification IHC of Iba1 (green) and WNV-E (red). Scale bars: 12um (panel A), 6um (panel B), 5um (panel C), 9um (panel D), 20um (panel E), 8um (panel F). From Quick 2014, J Virol 88(22) (Fig. 5).

Figure 2-18. Filopodial structures on microglia directed towards WNV-E+ cells in BSC, with increased zoom (bottom panel) on these structures. Scale bars: 40um (top), 20um (bottom).

41 different location (Figure 2-19). Microglia often were able to stretch their cell bodies into

many different directions and across great lengths (Figures 2-20 and 2-21). Iba1 staining

intensity was often notable at locations where microglia were in close proximity to WNV-E+

bodies (Figure 2-22), or the leading edges of lamellipodia (Figure 2-23). Microglia were able

to multitask while appearing to be in motion, with motility-related functions occurring while

actively phagocytosing (Figure 2-23).

Finally, microglia were able to phagocytose many WNV-E+ bodies at once, including

more than one infected cell at a time (Figure 2-24). Despite all the internalized WNV

microglia took up, they rarely appeared to be actually infected; most often many small bodies

of WNV were seen in single microglia, and often in small WNV-E+ puncta within the

membranes of small vesicular structures (Figure 2-25), possibly indicating the post-

phagocytic processing of lysed debris or cells.

Discussion

Modeling Neurotropic WNV Infection in Slice Cultures

The first goal of this aim was to establish a model system that adequately reflects

important components of WNV pathology, and the results described here indicate that this

goal was reached with high success. The slice culture system reflected WNV infectivity

observed in vitro (Cheeran 2005) (Figure 2-26) as well as significant neuronal death and the verification of the contribution of caspase-3 cleavage towards WNV-induced cell death previously described in vivo (Samuel 2007). Aspects of astrogliosis and astrocyte death, with

a notable lack of microglia infection, were additionally promising in that this model system

allowed for observation of microglia being the predominant antiviral cell type within the

tissue.

Figure 2-19. Filopodia from long microglia cell towards WNV infected cell in SCSC, with increased zoom (right panel) on filopodial contacts with WNV-E+ cell. Scale bar: 35um.

Figure 2-20. Examples of long cellular extensions from microglia cells. Scale bars: 25um.

Figure 2-21. Examples of rod-shaped and extended microglia cell bodies. Scale bars: 60um.

Figure 2-22. Microglia detection of infection cell and initiation of surrounding, with increased zoom (right panel) of cellular contact by microglia with infected cell. Left panel: note the two projections to the lower left and lower right of the focused projection, creating a “three prong” formation for eventual surrounding of the infected cell. Scale bars: 25um (left), 10um (right).

Figure 2-23. Microglia leading lamellipodial foot and internalization of WNV+ debris with increased zoom (right panel) of WNV-E+ debris endocytosis. Scale bars: 25um (left), 6um (right).

Figure 2-24. Multiple cells phagocytosed by single microglia cells. Scale bars: 15um (top), 25um (bottom).

Figure 2-25. WNV puncta within vesicular compartments of microglia. Scale bars: 20um (top), 8um (bottom).

Figure 2-26. WNV growth in human fetal neurons (A), astrocytes (B), and microglia (C). Note the different x-axis legends for time passed for each panel. From Cheeran 2005, J Neurovirol 11(6) (Fig.1).

Figure 2-2 (repeated). For comparison to Figure 2-26.

48 Microglia Activation and Phagocytosis During WNV Infection

Having the ability to distinguish intrinsic microglial functions and responses to WNV infection has not previously been investigated. The utilization of the slice culture model system has enabled the study of innate immune activity performed by microglia during WNV infection. By using a system that maintains the cytoarchitecture of the CNS while devoid of peripheral immune responses, the dynamic morphological activation of microglia displayed here highlights the potential role of microglia towards WNV clearance in vivo.

The ability to clearly label WNV-infected cells and antigenic debris allows for a unique model system to investigate mechanisms of phagocytosis. For example, in addition to these results being the first definitive evidence of microglia phagocytosis of WNV-infected cells, it is also the first report of microglial phagocytosis of subcellular WNV+ objects, and highlights the ability of microglia to not only phagocytose infected cells, but potentially infectious subcellular debris.

49 CHAPTER III

AIM 2: MINOCYCLINE ADMINISTRATION AND MICROGLIA INHIBITION IN A

SLICE CULTURE MODEL SYSTEM OF WEST NILE VIRUS INFECTION OF THE

CENTRAL NERVOUS SYSTEM

Introduction

Pharmacological Modulation of Microglia Activation

As described in Chapter I: Introduction, microglia have the ability to produce factors

that can influence pro- and anti-inflammatory events within the CNS across disease models.

Concerning WNV infections, little is known if microglia perform in a similar capacity. In

theory, microglia could be potential regulators of neuroinflammation by expressing

chemokines and cytokines to attract and activate peripheral immune cells extravasating the blood-brain barrier; microglia could also be mediators of neurosurvival by promoting the resolution of cytotoxic inflammation and returning the CNS to a state of homeostasis. If microglia do perform these functions during WNV infection, they could be targeted for therapeutic intervention via modulation of their role of regulators of the CNS inflammatory state.

The ex vivo slice culture system has the advantage of being able to easily apply

reagents for modulating microglia function, without the traditional concerns regarding drugs

passing through the blood-brain barrier for in vivo modeling of microglia activation. This ability can be exploited to use reagents that alter the activation profile of microglia, including potential expression of cytokines/chemokines. The results of Aim 1 showed that the slice culture system can replicate important features of WNV neuropathology and that microglia

50 can robustly activate in response to WNV infection, so if microglia are producing

inflammatory factors it should be detectable in this model.

Minocycline: Inhibitor of Pro-Inflammatory Microglia Activation

Minocycline is a synthetic tetracycline derivative originally developed as an antibiotic

(Chopra and Roberts 2001) that has been shown to have the ability to generally reduce pro-

inflammatory neuroinflammation (Liao 2013, Li 2013, Garrido-Mesa 2013, Moller 2016,

Fagan 2010, Hayakawa 2008, Hewlett and Corbett 2006) and microglia pro-inflammatory

activity (Kobayashi 2013, Zhu 2014, Stokes 2017, Markovic 2011, Peng 2014). Among

reagents used to modulate microglia function in vitro, minocycline is one of the most well-

characterized (Zemke and Majid 2004, Moller 2016).

The use of minocycline as an antiviral agent has precedent, albeit with mixed results

depending on the type of virus (Jackson 2012, Nagarakanti and Bishburg 2015). Treatment

with minocycline has had the most success as an antiviral with animal models for Japanese

encephalitis virus (JEV) infections, where it has been reported to greatly reduce mortality and

decrease associated pathologic mechanisms (Mishra 2008, Das 2011, Dutta 2010). For

WNV, there is limited knowledge about the effects of minocycline on WNV neuropathology

and related mechanisms; in human neuronal cell lines minocycline reduced WNV replication

(Michaelis 2007), but in the slice culture model the effect of minocycline can be assessed in a

more complicated environment.

Goals

The slice culture model system can be used to determine the extent of inflammatory signaling (i.e., the production of chemokines and cytokines) by CNS parenchymal cells without the assistance of peripheral immunity, while inhibiting pro-inflammatory microglia

51 activation with minocycline administration can inform what specific contributions microglia

have in inflammatory signaling. The goals of this aim were three-fold. The first goal was to

verify the ability of minocycline to reduce microglia activation, as measured by changes to

Iba1 expression and reduced morphological activation. The second goal was to assess pro-

and anti-inflammatory signaling in WNV-infected slice cultures, and then use minocycline to

observe any changes in this inflammatory expression profile. The final goal of this aim was

to check if minocycline treatment (and pro-inflammatory microglia activation) altered the

health status of CNS tissue during WNV infection.

Materials and Methods

Minocycline Administration

Minocycline HCl (Sigma) was prepared at a stock concentration of 1mM in slice

culture media, then diluted to a final concentration of 20uM in slice culture media to be used

for media changes. Minocycline treatment began at the time of WNV infection (0dpi) and

was present for all subsequent media changes until time of sample collection. Minocycline

was present only in the slice culture media (it was not directly applied to the apical surface of

slice cultures).

Experimental Methods

Please refer to Chapter II: Materials and Methods for previously mentioned experimental details; below are methods and materials not previously described.

Fluorescent immunohistochemistry (IHC). For iNOS detection, slides were imaged using a Zeiss Axiocam on a Nikon Eclipse E800 epifluorescent microscope, with image procurement conducted with Axiovision v4.8 software (Zeiss). The antibody used was rabbit anti-iNOS (1:300, Abcam).

52 RT-qPCR. Slice cultures were removed from insert membranes and homogenized in

RLT buffer (Qiagen) containing 1% beta-mercaptoethanol, then loaded into RNeasy spin columns (Qiagen) for RNA purification. Purified RNA was collected from the spin columns following manufacturer protocols. RNA integrity was assessed using an Agilent 2100 bioanalyzer (Agilent) and required an A260/280 value greater than 2.0 to be high quality.

RNA concentrations were equalized and used to make cDNA using iScript (Bio-Rad) following manufacturer directions. cDNA was mixed with 0.7ul/well of primers against genes of choice and 2x Sybr Green Mastermix (SABiosciences) for a total volume of 20ul.

Samples were loaded in triplicate into 96 well PCR plates and PCR amplification was performed with a CFX96 thermocycler (Bio-Rad). Relative gene expression was determined via threshold cycle analysis using Bio-Rad CFX Manager software. Beta actin served as the reference gene to compare expression across samples, while the 1dpi mock vehicle or 1dpi mock minocycline sample was the normalization set-point for individual gene analyses across samples.

Enzyme-linked immunosorbent assay (ELISA). Slice cultures were collected in lysis

buffer (R&D) to create lysate to screen with a custom Mouse Mix & Match Cytokine

ELISArray strip kit (Signosis). After homogenization and sonication, lysate was added to

ELISArray plate wells for 2 hour binding incubation, washed 3 times, then incubated for 1

hour with streptavidin-bound detection antibodies. After 3 additional washes, streptavidin-

HRP solution was incubated for 45 minutes. Detection solution was then added to each well

for 30 minutes before stop solution was applied, turning the solution yellow, depending on

the amount of bound detection antibody. The strength of the yellow color was quantified

colorimetrically at 450nm with an Emax spectrometer (Molecular Devices).

53 Results

Microglia Activation During WNV Infection with Minocycline Treatment

To determine if minocycline treatment had the ability to reduce microglia activation during WNV infection, the genetic expression of Iba1 was measured at 1, 3, 5, and 7dpi in

WNV-infected SCSC with or without minocycline treatment (as well as time-matched mock- infected SCSC for both conditions) (Figure 3-1A). Iba1 expression increased over the time course of WNV infection in vehicle-treated slice cultures, and this increased expression was significantly reduced in samples treated with minocycline. To gauge whether minocycline was specifically inhibiting microglial function, the genetic expression of GFAP was measured in the same samples. GFAP expression was elevated in WNV-infected SCSC when compared to mock samples, reflecting WNV-induced astrogliosis, but minocycline treatment had no significant reductive effect on this expression; GFAP expression actually increased with minocycline treatment, but not significantly.

Given the broad range of morphological activity by microglia during WNV infection as described in the previous chapter, qualitative assessment of WNV-induced microglia morphology changes were made in SCSC treated with minocycline at 5dpi (Figure 3-1B). In vehicle-treated SCSC infected with WNV, stereotypical morphology of microglia was observed throughout the tissue, including long cellular processes directed towards WNV+ cells and subcellular debris. These amoeboid structures were broadly unapparent in minocycline-treated SCSC with WNV infection; while microglia still appeared ruffled and large compared to microglia in mock-infected slice cultures, signs of chemotaxis and WNV- directed filopodia/lamellipodia were not observed with high frequency.

Figure 3-1. (A) RT-qPCR analyses were performed for Iba1 and GFAP expression in SCSC taken at 1dpi, 3dpi, 5dpi, and 7dpi. Changes in gene expression levels are indicated as fold increase over mock, with 1dpi mock-infected vehicle-treated SCSC used as the normalized control sample (expression set as 1) and beta-actin used as the normalized control gene. Iba1 gene expression (left) increases in WNV-infected (gray bars), compared to mock-infected (open bars), vehicle-treated SCSC in a time-dependent manner, indicating microglia activation. Minocycline treatment of WNV-infected SCSC (gray dashed bars) caused significant decreases in Iba1 expression when compared to WNV-infected, vehicle-treated counterparts (gray bars). GFAP gene expression (right) increases in WNV-infected, vehicle treated SCSC (gray bars) compared to mock-infected, vehicle treated controls (open bars), but not as dramatically nor as highly as Iba1. There was no decrease in WNV-induced GFAP expression following minocycline treatment. Asterisks indicate statistically significant differences in Iba 1 expression in minocycline-treated verses vehicle-treated, WNV-infected SCSC (*, P<0.05; **, P<0.01; unpaired student t-test). Error bars signify the standard error of the mean. 50 SCSC (n=2 mice) were used per experimental condition. (B) Immunohistochemistry of Iba1 (green) and WNV envelope protein (WNV-E, red) imaged with 60x objective from mock-infected (top row) and WNV-infected (bottom row) SCSC at 6dpi. Iba1 expression is increased in WNV-infected SCSC compared to mock-infected samples, but the amoeboid and enlarged microglia cells seen in vehicle-treated SCSC are absent from minocycline-treated SCSC. Scale bars: 60um. From Quick 2017, J Virol 91(22) (Fig 1).

55 WNV-Induced Inflammatory Signaling with Minocycline Treatment

To assess the expression of WNV-relevant pro-inflammatory factors in slice cultures,

CC-motif chemokine gene and protein expression was measured with and without minocycline treatment in slice cultures. Several CC-motif chemokines have been shown to be important pro-inflammatory factors of WNV neuropathology (Bardina 2015, Getts 2008,

Kumar 2013), and microglia are reported to be major producers of them, including CCL2,

CCL3, and CCL5 (Piotrowska 2016, Li 2009, Hanisch 2004, Rock 2004).

The genetic expression of CCL2, CCL3, CCL5, and CCL7 were assessed via RT- qPCR in SCSC that were collected at 1, 3, 5, and 7dpi (Figure 3-2A) for the full set of conditions used for establishing Iba1 inhibition. During WNV infection, each of the examined chemokines was broadly upregulated over the infection time course; again, minocycline had an inhibitory effect on this WNV-induced increase in expression. To verify that these observed genetic measures translated into related changes in protein expression,

SCSC were collected at 5dpi and protein measures were made for CCL2, CCL3, and CCL5 via ELISA (Figure 3-2B). Each of these chemokines were elevated in vehicle-treated WNV- infected SCSC compared to mock-infected counterparts, and for each chemokine minocycline treatment lead to a significant reduction in WNV-induced expression.

To gauge how microglia activation contributed to pro-inflammatory signaling more broadly, a repertoire of pro-inflammatory cytokines and chemokines with relevance to WNV infection was quantified by RT-qPCR (Kumar 2016, Klein 2005, Glass 2005, Glass 2006,

Shrestha 2008). Using the same panel of SCSC samples collected at 1, 3, 5, and 7dpi, gene expression was measured for CXCL10, IFNa, IL-1b, IL-6, IRF1, and TNFa (Figure 3-3).

Each of these genes was broadly upregulated over the time course of WNV infection. For

Figure 3-2. RT-qPCR analyses were performed for CC-motif chemokine gene expression in SCSC taken at 1dpi, 3dpi, 5dpi, and 7dpi. Each chemokine examined (CCL2, CCL3, CCL5, CCL7) rose with WNV infection in a time-dependent manner, with significant reductions in expression with minocycline-treatment (gray dashed bars), compared to vehicle-treated (gray bars), WNV-infected SCSC. 50 SCSC (n=2 mice) were used per experimental condition. (B) ELISA analyses for CCL2, CCL3, CCL5, and CCL7 from SCSC lysates taken at 5dpi. Each chemokine was elevated in WNV-infected vehicle-treated SCSC (gray bars), compared to mock-infected controls (open bars), with significant reductions following minocycline treatment (gray dashed bars) for CCL2, CCL3, and CCL5. Asterisks indicate statistically significant differences in minocycline treated, verses vehicle-treated, WNV-infected SCSC (*, P<0.05; **, P<0.01; ***, P<0.001; unpaired student t-test). Error bars signify the standard error of the mean. 100 SCSC (n=4 mice) were used per experimental condition. From Quick 2017, J Virol 91(22) (Fig 2).

Figure 3-3. RT-qPCR analyses were performed for pro-inflammatory gene expression in SCSC taken at 1dpi, 3dpi, 5dpi, and 7dpi. The expression of pro-inflammatory genes rose with WNV infection in a time-dependent manner, with varying levels of minocycline- dependent decreases (or no effect) in expression. Asterisks indicate statistically significant differences (*, P<0.05; **, P<0.01; ***, P<0.001; unpaired student t-test). Error bars signify the standard error of the mean. 50 SCSC (n=2 mice) were used per experimental condition. From Quick 2017, J Virol 91(22) (Fig 3A).

58 some genes, minocycline treatment had no significant effect on gene expression, including

CXCL10 and IFNa. Other genes had significant reductions at only one or two time points, and were otherwise generally reduced following minocycline treatment.

To understand if anti-inflammatory cytokines were produced in WNV-infected slice cultures, and if microglia had a potential role in their expression, further RT-qPCR analysis on the previously mentioned samples was conducted on a selection of canonical anti- inflammatory factors from other disease models (Orihuela 2016, Bergold 2016, Kumar Jha

2016, Zhang and Gensel 2014); this strategy was employed as anti-inflammatory signaling during WNV infection does not have an established body of evidence to choose genes of interest besides a few studies (Qian 2014, Bai 2009, Nelson 2015). The selected genes were

FIZZ1, IL-4, IL-7, IL-10, IL-13, and TREM2 (Figure 3-4). During WNV infection in vehicle-treated SCSC, each of these genes was upregulated compared to mock-infected samples. Following minocycline treatment, most of these genes were seen to increase in their expression, with some of these elevations reaching significance at one or two time points per gene.

WNV Growth and Tissue Health with Minocycline Treatment

To determine if minocycline had an effect on WNV growth in slice cultures, RT-PCR with WNV RNA was used to determine viral load in SCSC samples at 3, 5, and 7dpi (Figure

3-5). Significantly lower amounts of WNV were detected in minocycline-treated SCSC infected with WNV compared to their vehicle-treated counterparts; however, the overall reduction in WNV load in minocycline-treated SCSC was less than one order of magnitude.

Next the expression of iNOS was assessed via IHC at 6dpi in SCSC infected with

WNV, with and without minocycline treatment (Figure 3-6). Mock-infected SCSC showed

Figure 3-4. RT-qPCR analyses were performed for anti-inflammatory gene expression in SCSC taken at 1dpi, 3dpi, 5dpi, and 7dpi. The expression of anti-inflammatory genes rose with WNV infection in a time-dependent manner, with broad minocycline-dependent increases in expression at different time points. Asterisks indicate statistically significant differences (*, P<0.05; **, P<0.01; ***, P<0.001; unpaired student t-test). Error bars signify the standard error of the mean. 50 SCSC (n=2 mice) were used per experimental condition. From Quick 2017, J Virol 91(22) (Fig 3B).

Figure 3-5. WNV-infected SCSC samples were treated with either minocycline or vehicle. Samples were collected at 1dpi, 3dpi, 5dpi, and 7dpi (n=3) and mRNA was isolated for RT- qPCR analysis. Viral load (equivalent PFUs) was determined using a standard curve created from RNA from WNV-infected brain tissue. The viral load is measured on a log10 scale. Error bars represent the standard error of the mean. Asterisks indicate values that are statistically significant (***, P<0.001; **, P<0.01; unpaired student t-test). From Quick 2014, J Virol 88(22) (Fig. 8B).

Figure 3-6. Immunohistochemistry of iNOS (green) from mock-infected and WNV-infected SCSC at 6dpi (right column). Clusters of iNOS+ cells are delineated in vehicle-treated SCSC with WNV infection, whereas no similar cells are seen in minocycline-treated SCSC with WNV infection or any mock-infected SCSC. DAPI (left column) imaging shows equivalent amounts of cells for each field of vision. Scale bars: 100um. From Quick 2017, J Virol 91(22) (Fig 6).

62 no sign of iNOS expression, while iNOS expression was observed throughout tissue in

WNV-infected SCSC; this expression was markedly diminished in minocycline-treated

SCSC infected with WNV.

To determine the effect of minocycline treatment on tissue health, cell loss was measured via IHC cell counts for Map2 (neurons), GFAP (astrocytes), and Iba1 (microglia) at 6dpi in WNV-infected SCSC treated with vehicle or minocycline. For neurons in WNV- infected SCSC, neuronal loss was evident when compared to the number of neurons present in mock-infected SCSC (Figure 3-7); when vehicle-treated and minocycline-treated WNV- infected SCSC were compared, a significant yet modest increase in neuron numbers was observed. The same pattern held for microglia cell counts (Figure 3-8).

Astrocyte health was characterized as normal, dead, or in a state of astrogliosis

(Figure 3-9), with mock-infected SCSC displaying mostly normal astrocytes with slightly less astrogliosis observed in samples treated with minocycline. The amount of astrocytes displaying signs of astrogliosis and cell death in WNV-infected SCSC was elevated, with significantly increased healthy astrocytes and significantly decreased dead astrocytes observed in minocycline-treated samples; the level of astrogliosis in both appeared to be equivalent. The difference between astrogliosis and astrocyte loss was often difficult to determine, but regions that had high levels of punctate GFAP signal were often deemed regions of astrocyte death (Figure 3-10).

Discussion

WNV-Induced Microglia Activation with Minocycline

Iba1 expression upregulation is a widely-used metric to describe pro-inflammatory microglia activation (Gao 2017, Go 2016, Ito 2001, Mori 2000), and its expression was

Figure 3-7. Immunohistochemistry of Map2 (green) and WNV-E (red) from mock-infected (top row) and WNV-infected (bottom row) SCSC at 6dpi. Mock-infected SCSC show disperse Map2 expression throughout tissue, while WNV-infected SCSC have greatly reduced Map2 expression and cell bodies. Minocycline-treated SCSC with WNV infection have noticeably more Map2+ cells than vehicle-treated counterpart. Quantification of Map2+ cell bodies imaged with 10x objective. Across multiple fields of vision (FOV), there is a large loss of Map2+ cells in WNV-infected SCSC compared to mock-infected SCSC, but a statistical difference between minocycline- and vehicle-treated SCSC with WNV infection. For mock-infected, n = 8 SCSC; for WNV-infected, n = 12 SCSC. For each condition, n = 8 SCSC. Asterisks indicate statistically significant differences (*, P<0.05; unpaired student t- test). Error bars signify the standard error of the mean. Scale bars: 100um. From Quick 2017, J Virol 91(22) (Fig 5A, B).

Figure 3-8. Immunohistochemistry of Iba1 (green) and WNV-E (red) from mock-infected (top row) and WNV-infected (bottom row) SCSC at 6dpi. WNV infection causes substantial microglia loss in vehicle-treated SCSC, with less microglia loss observed with minocycline- treatment. Quantification of Iba1+ cell bodies imaged with 10x objective. Large loss of Iba1+ cell bodies is observed for WNV-infected SCSC, but cell loss is more prominent for vehicle-treated SCSC compared to minocycline-treated SCSC. For each condition, n = 12 SCSC. Asterisks indicate statistically significant differences (**, P<0.01; unpaired student t- test). Error bars signify the standard error of the mean. Scale bars: 100um. From Quick 2017, J Virol 91(22) (Fig 5C, D).

Figure 3-9. Immunohistochemistry of GFAP (green) and WNV-E (red) from mock-infected (top row) and WNV-infected (bottom row) SCSC at 6dpi. WNV infection causes significant astrogliosis (increased GFAP expression, thickened processes) compared to mock-infected SCSC. Increased astrocyte cell death is largely apparent in vehicle-treated WNV-infected SCSC (small GFAP+ particles) but not as widespread in minocycline-treated WNV-infected SCSC. Quantification for the percent of each field of vision (FOV) in which astrocytes are normal, astrogliotic, or dead in images taken with 20x objective. Dead astrocytes were observed in WNV-infected SCSC, but to a larger extent in vehicle-treated versus minocycline-treated SCSC. Both WNV-infected SCSCs had similar levels of astrogliosis. For each condition, n = 8 SCSC. Asterisks indicate statistically significant differences (*, P<0.05; **, P<0.01; unpaired student t-test). Error bars signify the standard error of the mean. Scale bars: 100um. From Quick 2017, J Virol 91(22) (Fig 5E,F).

Figure 3-10. Astrocyte death in vehicle-treated WNV-infected SCSC is increased compared to minocycline-treated counterparts. GFAP (green) appears more punctate and is not associated with cells in regions with astrocyte death, which is more apparent in vehicle- treated SCSC. Scale bars: 80um.

67 significantly reduced following minocycline treatment. While GFAP expression in

astrocytes does not necessarily reflect the same mechanism of antiviral cell activation as Iba1

does for microglia, the lack of significant changes to its expression following minocycline

treatment does lend credence to the notion that minocycline was specifically targeting

microglia activity. Furthermore, following minocycline treatment in WNV-infected SCSC

the morphology of microglia was qualitatively less antiviral, giving additional support to the

ability of minocycline to inhibit microglia antiviral responses.

WNV-Induced Inflammatory Signaling with Minocycline

The data presented in this aim was the first attempt to characterize microglia

contributions to pro- and anti-inflammatory signaling mechanisms during WNV infection of

the CNS. A dichotomous influence of minocycline administration was observed on WNV-

induced inflammatory gene expression: pro-inflammatory genes were broadly and

significantly downregulated following minocycline treatment, whereas anti-inflammatory

genes were broadly and significantly upregulated in the same samples (Figure 3-11). While

the results are suggestive to the idea that microglia are directly producing these inflammatory cytokines/chemokines, ultimately this can only be extrapolated. Still, it can be deduced that microglia likely play an important part in regulating the expression of these factors, given the concurrent significant reductions in Iba1 gene expression and morphological activation profile following minocycline treatment.

As previous studies have deduced that minocycline heavily inhibits microglia pro- inflammatory activation and increases anti-inflammatory activation, the evidence presented here also suggests that microglia are driving the observed inflammatory phenotype.

Nonetheless, minocycline can effect on non-microglial cells in the CNS, including neurons

Figure 3-11. RT-qPCR analysis data was graphed to show the percent change in expression with minocycline treatment at each time point for each gene. When categorized as M1 or M2, a clear pattern is observed with minocycline treatment, with large-scale reductions in M1 gene expression (red bars) and increases in M2 gene expression (green bars). Iba1 and GFAP expression are also indicated (black bars) to show minocycline effect on glial activation. Asterisks indicate statistically significant differences (*, P<0.05; **, P<0.01; ***, P<0.001; unpaired student t-test).

69 infected with WNV (Michaelis 2007), and the ability of these cell types to be expressing inflammatory factors and to contribute to the minocycline-dependent changes in their expression cannot be ruled out. Astrocyte function following minocycline treatment is not as extensively characterized as microglia, and current literature suggests that when astrocyte function and/or reactivity is altered in certain inflammatory environments, it is a consequence of altered microglia function (Stokes 2017, Liddelow 2017).

WNV-Induced Tissue Injury and Viral Growth with Minocycline

The data here reflects previous studies that have shown a reductive effect on WNV replication in the presence of minocycline (Michaelis 2007). It is unclear how the modest reduction in WNV load is related to the modest reduction of cell death observed in minocycline-treated SCSC infected with WNV; this is further confounded by the minocycline-dependent changes to pro- and anti-inflammatory cytokine/chemokine expression. What is again clear is that without adequate viral clearance, WNV infection produces significant tissue injury and cell death, despite alterations to the microglia activation profile.

70 CHAPTER IV

AIM 3: PLX5622 ADMINISTRATION AND MICROGLIA DEPLETION DURING

WEST NILE VIRUS INFECTION IN VIVO

Introduction

In Vivo Microglia Depletion Model

The slice culture model has many advantages for specific avenues of research,

particularly the investigation of intrinsic microglia activation abilities without confounding

peripheral immune responses. However, to fully understand how microglia contribute to

WNV neuropathology, including aspects related to virus restriction/clearance and

neuroinflammatory regulation, an in vivo model system needs to be used.

The ability to deplete microglia from mice would be an ideal method to evaluate their

importance to WNV pathogenesis. Recently a line of CSF1R kinase inhibitors was developed by the pharmaceutical company Plexxikon, and these drugs have been shown to have highly reproducible effects on microglia depletion in brains of adult mice (Elmore 2014,

De 2014). Microglia are the only cells in the CNS that express CSF1R (Erblich 2011,

Ginhoux 2010), and once mice have been treated for several days with CSF1R kinase inhibitors, depletion levels can reach over 90% (Elmore 2014, Dagher 2015). These reports have highlighted the ease with which the drugs can be used (the drug is incorporated into mouse chow), the lack of apparent behavioral distortions of mice taking the drug, and the speed with which the drug can enact its ablation effects (typically within several days for significant effects to a few weeks for maximal effects).

The vast majority of the studies conducted using these drugs have focused on non- viral models of CNS disease, particularly for glioblastoma (Antonios 2017, Prada 2013) and

71 neurodegeneration (Spangenberg 2016, Dagher 2015). No studies using these drugs have yet

been reported for flaviviral infections, and considerable utility for investigations into

microglia contributions to antiviral responses can be made by using these CSF1R kinase

inhibitors to ablate microglia in infected mice.

Plexxikon generously provided PLX5622, one of the characterized drugs that can

adequately ablate microglia (Spangenberg 2016, Feng 2016, Acharya 2016, Ebneter 2017),

for use in studying the role of microglia in WNV pathogenesis. Generally, the effects on

microglia ablation for PLX5622 have been observed to be similar to other Plexxikon-

generated CSF1R kinase inhibitors, particularly the better characterized PLX3397 (De 2014,

Elmore 2014, Prada 2013).

Swiss Webster Mouse Model of WNV Infection

The Swiss Webster (SW) mouse strain that was the source of tissue for the slice

culture experiments is not used nearly as widely as the C57/Bl6 line, which is the

predominant strain for genetic manipulations in mouse model systems. Consequently, SW

mouse models for WNV infection are not commonly encountered in the literature. Mice of

different experimental strains are known to have varying degrees of similarities and

differences concerning their reactivity to noxious stimuli across disease models, and the same appears to be true for WNV infections (Graham 2015).

Previous studies using SW mice infected with WNV NY99 are not only rare, but often performed with intraperitoneal and intracranial inoculations (Davis 2004, Beasley

2002) which will have a different pathogenesis compared to intramuscular/subcutaneous inoculation models. Several experiments with SW mice in our lab with inoculation via the intramuscular injection protocol generally have a mortality rate around 40% (Figure 4-1).

Figure 4-1. Swiss-Webster (SW) WNV survival model. Top panel: survival curves for five independent experiments with intramuscular NY99 infection (1000PFU) of SW mice, with the numbers of mice per experiment noted in the legend (right). Bottom panel: combined survival curve for NY99 infections in SW mouse model.

73 For the purposes of investigating the effect of microglia depletion on WNV neuropathology,

having a baseline mortality rate of approximately 40% is helpful; the hypothesis that

microglia contribute towards important antiviral responses and neuroinflammatory regulation

would mean that microglia depletion should have a deleterious effect on survival and

antiviral responses.

One last component to consider for usage of the SW mouse model is that for the

majority of the CSF1R kinase inhibitor studies utilizing experimental mouse models, again

the C57/Bl6 mouse is the most widely utilized strain. No previous studies have employed the

SW mouse model, which requires that PLX5622 will work in accordance to what has

previously been reported for different strains of mice.

Goals

In order to maintain consistency to be more able to extrapolate findings from the slice culture model into results gathered from the in vivo WNV infection model, the SW mouse model was utilized for this aim. In doing so, this report is the first to utilize this mouse strain with any of the Plexxikon line of CSF1R kinase inhibitors. The first goal of this aim was to verify that microglia depletion occurred in SW mice in a similar manner to results seen in other experimental mouse strains; namely, that the majority of microglia were depleted within a time course spanning several days of treatment with PLX5622.

Following the hypothetical thinking that microglia depletion would lead to reduced antiviral surveillance and increased mortality, the next goal of this aim was to determine the extent to which microglia depletion effected survival and WNV load with the CNS of

PLX5622-treated mice. Across the WNV animal model literature, arguably the most important indicators of increased mortality were a lack of restriction to WNV reaching the

74 CNS and a lack of WNV clearance from the CNS via neuroinflammatory activated peripheral

immune cells. In assessing mortality and viral load in this model system, important

conclusions can be made concerning the role of microglia in WNV neuroinvasive disease.

The final goal of this aim is to investigate inflammatory cytokine and chemokine

levels in the CNS of WNV-infected mice to observe if any factors that were linked to

microglia expression in slice cultures could be associated with results in the PLX5622-treated

mice.

Materials and Methods

Swiss Webster Mice

All methods for proper housing and treatment of mice were in accordance with the

University of Colorado – Denver and Institutional Animal Care and Use Committee guidelines.

PLX5622 Administration

In accordance with the Material Transfer Agreement with Plexxikon, PLX5622 chow

and control chow were provided by Research Diets Incorporated at a concentration of

1200mg/kg in AIN-76A chow. Mice were given unlimited access to PLX5622 or control

chow, with access to either chow type started at the same time and throughout the duration of

all experiments. Verification of microglia depletion experiments used pretreatment time

courses of 4 and 10 days; all WNV infection experiments used pretreatment time courses of

14 days.

WNV Infection Model in SW Mice

After having access to PLX5622 or control chow for 14 days, 12-week old mice were

inoculated with 1000 PFU of WNV NY99 via footpad intramuscular injection. Mice were

75 assessed and weighed daily, while having similar access to their food for the duration of the

infection time course.

Experimental Methods

Survival and body weight curves. Mouse mortality was checked daily and each mouse weighed daily to assess changes in body weight over time. Individual mice were identified via tail markings.

Fluorescent immunohistochemistry. Mice were perfusion fixed with 20ml of 4% paraformaldehyde (PFA) followed by 20ml of PBS-EDTA. Brains were dissected out of the skull and stored in 10% NBF. Brains were then processed for platform mounting by removing the cerebellum and flattening the caudal end of the brain with a razor cut. Brains were immersed in 3% agarose, and after hardening and shaving off excess agarose the mounts were glued to a Vibratome platform (Leica VT1000S) and the well filled with PBS to the top of brain (the rostral tip). Coronal slice sections were collected at 100um thickness

(speed setting 1, frequency setting 9) into a Petri dish with PBS. Sections were placed into were placed into individual wells of a 24-well (16mm) culture plate with permeabilization/blocking solution (PBS, 4% normal goat serum, 2% bovine serum albumin,

0.3% Triton-X 100) for at least one hour. Brain sections were then incubated overnight at room temperature in primary antibodies diluted in permeabilization/blocking solution.

Primary antibodies used were mouse anti-WNV E protein (1:200, ATCC), rabbit anti-NeuN

(1:100, Abcam), and rabbit anti-Iba1 (1:500). The next day brain sections were washed in

PBS three times (approximately 10 minutes per wash with light agitation/rotation) before incubation for two to three hours in secondary antibodies, also diluted in permeabilization/blocking solution. Secondary antibodies used were AlexaFluor goat anti-

76 mouse 568 and AlexaFluor goat anti-rabbit 488 (1:1000, Invitrogen). Brain sections were again washed in PBS three times in a similar fashion before lightly being rinsed with water and placed onto microscope slides, with excess fluid removed with tissue paper. Microscope slides were mounted with Prolong Gold or Diamond antifade reagent (Molecular Probes) and coverslipped. Slides were imaged using either 1) a Zeiss Axiocam on a Nikon Eclipse E800 epifluorescent microscope, with image procurement conducted with Axiovision v4.8 software (Zeiss), or 2) a Leica TCS SP8x laser scanning confocal microscope, with image procurement conducted with LAS X software (Leica).

WNV RT-PCR. Mice were euthanized and the brains quickly dissected out of the skull and homogenized with 1ml PBS in a 2ml douncer. 200ul of homogenate was mixed with 1ml RLT buffer (Qiagen) with 1% B-mercaptoethanol and then quickly frozen. At the time of RNA purification, 150ul of thawed homogenate in RLT was combined with 150ul fresh RLT with 1% B-mercaptoethanol, vortexed and homogenized further with 29g needle and syringe before being put into RNeasy spin columns (Qiagen). Purified RNA was collected from the spin columns following manufacturer protocols. RNA integrity was assessed using an Agilent 2100 bioanalyzer (Agilent) and required an A260/280 value greater than 2.0 to be high quality. RNA concentrations were equalized and used to make cDNA using either SuperScript Vilo kit (Invitrogen) or iScript (Bio-Rad) following manufacturer directions. cDNA was mixed with 1ul/well of primers for WNV E protein and 2x iTaq mastermix (Bio-Rad) for a total volume of 20ul. Samples were loaded in sextuplicate into 96 well PCR plates and PCR amplification was performed with a CFX96 thermocycler (Bio-

Rad). Equivalent plaque-forming units (ePFU) were quantified by taking WNV-infected brain lysates of varying levels of infection and calibrating the specific PFU titer for each to

77 create a standard curve ranging from 104 to 108 which was used to determine the WNV RNA

load of brain lysate samples.

RT-qPCR. Mice were euthanized and the brains quickly dissected out of the skull and homogenized with 1ml PBS in a 2ml douncer. 200ul of homogenate was mixed with 1ml

RLT buffer (Qiagen) with 1% B-mercaptoethanol and then quickly frozen. At the time of

RNA purification, 150ul of thawed homogenate in RLT was combined with 150ul fresh RLT with 1% B-mercaptoethanol, vortexed and homogenized further with 29g needle and syringe before being put into RNeasy spin columns (Qiagen). Purified RNA was collected from the spin columns following manufacturer protocols. RNA integrity was assessed using an

Agilent 2100 bioanalyzer (Agilent) and required an A260/280 value greater than 2.0 to be high quality. RNA concentrations were equalized and used to make cDNA using iScript

(Bio-Rad) following manufacturer directions. cDNA was mixed with 0.7ul/well of primers against genes of choice and 2x Sybr Green Mastermix (SABiosciences) for a total volume of

20ul. Samples were loaded in sextuplicate into 96 well PCR plates and PCR amplification was performed with a CFX96 thermocycler (Bio-Rad). Relative gene expression was determined via threshold cycle analysis using Bio-Rad CFX Manager software. Beta actin served as the reference gene to compare expression across samples, while samples from uninfected PLX5622 or uninfected control brains were used for the normalization set-point for individual gene analyses across samples.

Results

PLX5622 and Microglia Depletion in SW Mice

To verify that PLX5622 could deplete microglia in SW mice, Iba1+ cells were

imaged via IHC in the brains of mice treated with PLX5622 for 10 days, with their Iba1+

78 staining compared to the brains of mice on control chow for an equivalent duration (Figure 4-

2). In the brains of mice treated with PLX5622 for 10 days, significantly reduced Iba1+ cells

were observed across all brain regions; when quantified across multiple fields of vision

(FOV) from multiple mice (Figure 4-3), approximately 85% reduction was seen and was very

significant for both the cortex and striatum (p < 0.0001).

PLX5622 and WNV Mortality

Mice treated with PLX5622 for 14 days were infected with WNV to assess the

consequences of microglia depletion on mortality and morbidity. Out of six control mice

infected with WNV, one died at 11dpi and another at 15dpi, reflective of a 66% survival rate,

which is close to the normal survival rate for the SW mouse model of WNV infection (Figure

4-4A). Out of six PLX5622-treated mice infected with WNV, none survived past 14dpi (-2

at 9dpi, -1 at 10dpi, -1 at 11dpi, -1 at 13dpi, -1 at 14dpi), which is a 0% survival rate.

To assess the timing of illness in these mice, the daily body weight loss for each

mouse was calculated, and these values were averaged for all surviving mice by the day and

graphed according to their conditional grouping (control or PLX5622) (Figure 4-4B). Both sets of mice began to collectively lose weight around the 7dpi time point with a steady decline daily for several days. While the PLX5622-treated mice had a slightly increased amount of average body weight loss, the overall trajectory for both sets of mice is similar; at

10dpi, the three surviving PLX5622-treated mice had an average body weight loss that was identical to the six control mice.

PLX5622 and WNV Load in the Brain

To determine the amount of WNV infection within the CNS of PLX5622-treated mice, RT-PCR was used to determine WNV RNA levels in the brains of mice at 6dpi (six

Figure 4-2. Microglia are depleted from the brains of PLX5622-treated mice. Iba1 (green) is widely distributed in the brain regions of control mice, whereas Iba1+ cells are notably missing in PLX5622 treated mice brains. Scale bars: 200um (10x), 100um (20x).

Figure 4-3. Microglia counts in the brains at 10d. For both cortex and striatum regions of the brain, Iba1+ cells were significantly depleted in PLX5622-treated mice. For each region, n=9 (n=3 for an individual mouse, 3 mice total). Asterisks indicate statistical significance (***, P<0.001; unpaired student t-test).

Figure 4-4. PLX5622 treatment greatly increases WNV mortality in SW mice. Top panel (A): survival curve for control and PLX5622-treated groups of SW mice infected with WNV. Both sets of mice were on their respective food for 14 days prior to infection, and remained on their food for the duration of the infection time course. For each group, n=6. Bottom panel (B): the average percent change in body weight for all mice in either group.

82 mice per condition) and 9dpi (9 mice per condition) (Figure 4-5). For both time points, significantly increased amounts of WNV were detected in PLX5622-treated mice brains compared to control mice.

To further verify the extent of WNV infection in the brains on PLX5622-treated mice,

IHC was performed for antibodies against WNV-E at 9dpi and cells counted to quantify the extent of infection (Figure 4-6). The increased amount of WNV infection in the brains of

PLX5622-treated mice was dramatically apparent; hundreds of WNV-infected cell bodies were observed in nearly every brain section assessed, predominantly in thalamic and striatal regions but also in the cortex (to a lesser extent). In the brains of control mice, there were never more than a dozen (if any) WNV-infected cell bodies observed within CNS nervous tissue. However, considerable WNV was often still observed in brain sections from control mice, but often restricted to cells of the lateral ventricles and blood-brain barrier (Appendix).

To determine if the WNV-E+ cell bodies within the brain were neurotropic infections,

WNV-E staining was assessed for colocalization with antibody staining against NeuN, the nuclear neuronal marker (Figure 4-7). WNV-E was often seen surrounding NeuN staining, indicating that the WNV infection present within the brain was infecting neurons. Given that in slice cultures similarly overwhelming WNV coincided with microglial phagocytosis, Iba1 staining was performed with WNV-E staining to determine if phagocytosis of WNV-E+ cells could be detected in vivo (Figures 4-8 and 4-9). Instances of phagocytosis being carried out by Iba1+ cells were observed in heavily-infected regions.

PLX5622 and Inflammatory Gene Expression in the Brain

To assess the expression of inflammatory cytokines and chemokines in the brains of

WNV-infected PLX5622-treated mice, RT-qPCR was performed on the brains of mice

Figure 4-5. WNV infection is higher in the brains of PLX5622-treated mice. At 6dpi, WNV was undetectable the threshold of 103 PFU for control mouse brains, while there was a significant increase in WNV for PLX5622-treated mice (P=0.0152; Mann-Whitney t-test). At 9dpi, while control mice had more WNV compared to 6dpi, the amount of WNV detected in PLX5622-treated brains was again significantly higher (P=0.0479; Mann-Whitney t-test). Each data point is the average of a single brain lysate from an individual brain, ran in sextuplicate.

Figure 4-6. WNV cell counts within the striatum of WNV-infected mice. Over multiple FOV from different mice, only several WNV-E+ cells were observed in the striatum of control mice. In the striatum of PLX5622-treated mice, WNV-E+ cells were widespread and significantly higher than in control mice (P=0.0064; unpaired t-test).

Figure 4-7. WNV infection of neuronal cells in the striatum of PLX5622-treated mice. NeuN (green) and WNV-E (red) was seen to colocalize in PLX5622-treated mice brains, while no WNV-E was apparent in the brains of control mice. Scale bars: 100um.

Figure 4-8. WNV infection leads to phagocytosis of WNV-E+ (red) cells by Iba1+ (green) cells in the striatum of PLX5622-treated mice. Control mice had normal distribution and appearance of Iba1+ cells, while widespread WNV infection and phagocytosis was apparent. Scale bars: 100um.

Figure 4-9. Cropped image of high-mag phagocytosis by Iba1+ (green) cells of WNV-E+ (red) cells in PLX5622-treated mouse brain (striatum). Scale bar: 25um.

87 collected at 6dpi (6 mice per condition) and 9dpi (9 mice per condition) (Figure 4-10). The genetic expression of several important inflammatory cytokines/chemokines with relevance to WNV (CCL2, CCL3, CCL5, CCL7, CXCL9, CXCL10), as well as genes related to pro- and anti-inflammatory signaling (IFNg and TREM2, respectively).

Most chemokines examined were elevated in PLX5622 brains across the 6dpi and

9dpi time points, with significant increased at 6dpi for CCL2 and CXCL10 and at 9dpi for

CCL2, CCL7, CXCL9, CXCL10, and IFNg. Notably, CCL3 and CCL5 had no significant changes in expression. Meanwhile, TREM2 was significantly decreased at both 6dpi and

9dpi.

Discussion

Microglia and CNS WNV Load In Vivo

The results of this aim show that microglia depletion correlates strongly with increased WNV in the CNS, and as expected this coincides with increased mortality. Just how WNV infection gets so high in the brains of PLX5622-treated mice is unclear.

Microglia may potentially play a role in restricting WNV from entering the CNS, or initiating antiviral neuroinflammatory mechanisms to promote viral clearance, or a combination of the two.

A potential caveat to the observed results is that CSF1R inhibition has a dampening effect on peripheral immune responses that lead to greater WNV peripheral infection, which in turn could make WNV infection of the CNS more likely. However, if this were the case then the PLX5622 mice would be expected to display overt illness at much earlier time points compared to control mice infected with WNV. This does not appear to be the case when the average body weight losses began at accelerate at similar time points for both

Figure 4-10. Gene expression in the brains of WNV-infected control (white circles) and PLX5622-treated (black squares) mice at 6dpi and 9dpi. Y-axis is fold increase expression compared to mock-infected control or PLX5622-treated mice, respectively. Each data point is the average of a single brain lysate from an individual brain, ran in sextuplicate. Asterisks indicate statistical significance (**, P<0.01; *, P<0.05; unpaired student t-test).

89 PLX5622-treated and control mice. Furthermore, in an attempt to detect WNV in the livers

of the same mice used for the 6dpi/9dpi brain collections (the liver is susceptible to WNV

infection), no detectable virus was observed; if peripheral immune responses were indeed

dampened and peripheral WNV increased, some level of WNV would have been expected to

be detected.

Microglia and Inflammatory Signaling in the WNV-Infected Brain

Significant differences in pro-inflammatory cytokine and chemokine expression were

observed in the brains of PLX5622 mice compared to the brains of control mice. Most of

this can be attributed to the fact that PLX5622 mice have overwhelming levels of WNV

infection in the brain, and clearly the Iba1+ cells present are highly active in carrying out

phagocytic and antiviral mechanisms (and are also the clearest depiction of phagocytosis

during WNV CNS infection in vivo); however, the interesting result is which inflammatory mediators were not increased: CCL3 and (in particular) CCL5. Both of these chemokines were observed to be elevated during WNV infection in slice cultures, especially CCL5, which not only was repeatedly one of the highest-increased pro-inflammatory factors produced in slice cultures, but is known to be an important factor for in vivo responses

(Kumar 2016, Glass 2005, Glass 2006). The fact that other factors were significantly elevated in PLX5622 brains but CCL3 and CCL5 were not are a strong indication that microglia are a major source of their production during WNV infection in vivo. The lack of

CCL5 expression in PLX5622-treated mice brains has also been noted in a cranial irradiation model, which also saw most other pro-inflammatory genes upregulate while CCL5 did not

(Acharya 2016). Another study utilizing PLX5622 noted a general reduction in pro-

90 inflammatory factors and increased anti-inflammatory factors in a sub-acute disease model for alcohol withdrawal (Walter 2017).

The same reasoning for assuming microglia production of CCL5 due to its conspicuous lack of upregulation can also be attributed to the expression of TREM2, which was significantly decreased at both 6dpi and 9dpi; in slice cultures TREM2 was increased with minocycline expression as one of the putative anti-inflammatory genes, and has previously been described as a microglia-produced neuroprotective factor (Takahashi 2005,

Takahashi 2007, Fu 2014, Sierra 2013). While the cytokine/chemokine profiling highlighted in this aim was by no means exhaustive, it nonetheless was able to recapitulate some of the conclusions brought about in the slice cultures.

91 CHAPTER V

CONCLUSIONS AND FUTURE DIRECTIONS

Microglia Phagocytosis and Antiviral Activity During WNV Infection of the Central

Nervous System

This project encompasses the first specific study into the mechanisms of how microglia, as the resident innate immune cells of the CNS, respond to WNV infection and the consequences of their activity towards antiviral responses. To address how microglia perform these antiviral activities, unique model systems were established to gain significant insight into innate immune functions of microglia. In the slice culture model, unparalleled observations of phagocytosis of WNV-infected cells and antigenically-positive debris were made, and represent arguably the most striking examples of microglia phagocytosis in any culture model system of CNS disease. With the use of minocycline in the slice culture model, a role for microglia production of pro- and anti-inflammatory cytokines and chemokines was observed, highlighting the ability to use this model system for assessing potentially therapeutic reagents while attempting to gauge specific mechanisms of microglia function in regards to inflammatory regulation. Finally, a novel microglia depletion model of

WNV infection in vivo allowed for the observation that microglia are a crucial component of

WNV restriction from the CNS.

Concerning the slice culture model system, a potential caveat is that the usage of neonatal brain and spinal cord tissue may not accurately reflect innate immune responses that occur in the adult CNS. While this could potentially have some truth, the slice culture model system is still an improvement upon most other culture systems that utilize cells taken from even earlier time points (e.g., embryonic CNS tissue) or from cancerous cell lines. By

92 maintaining the cell-cell contacts among microglia, neurons, and other cells, a more accurate

representation is made between infected cells, uninfected bystander cells, and the microglia

cells that must interact with them all. Nonetheless, improvements could be made to some

aspects pertaining to the collection and culture of brain and spinal cord slices, including the

potential use of trophic supplements (e.g., BDNF, NGF, etc.) or other improvements to

culture media, varying the length of the time and/or conditions of acclimatization to the

culture environment, and/or varying the metrics of WNV infection (assessing early time

point, using less virus, attenuated vs. pathological strains, etc.). Attempts were made to

create an adult slice culture model akin to that used for this project, but these cultures were

not viable for any productive length of time.

Support for maintaining the slice culture system as it is currently manifested is the

link of CCL5 and TREM2 expression changes with minocycline treatment when compared to

the PLX5622-treated mice; microglia are associated with the production of CCL5

(Piotrowska 2016, Li 2009, Hanisch 2004, Rock 2004) and TREM2 (Linnartz 2010,

Takahashi 2007), and both were found to be significantly lacking in expression with

minocycline treatment ex vivo and PLX5622 treatment in vivo during WNV infection.

Further studies could determine a wider range of inflammatory, antiviral, and/or neuroprotective factors differentially modulated in either model and use the other to verify whether or not microglia are a putative source of production. Cross-over of these types of findings could lead to further insights as to how microglia regulate antiviral responses and contribute towards the restriction of WNV that was observed in the PLX5622 model.

93 Future Directions: Slice Culture Model System

The slice culture model system proved to be an excellent avenue for investigating the

intrinsic capabilities of microglia activation during WNV infection. While microglia

activation is a broad categorization across disease models that can encompass many disparate

cellular functions, questions remain as to what specific signals received by microglia can

activate different aspects of activation; for example, what receptors mediate the expression of

pro- versus anti-inflammatory cytokines and chemokines, and how different stimulation

paradigms (e.g. ischemic, neurodegenerative, viral, etc.) alter the activation profile of

microglia. The use of slice cultures for a range of CNS disease models could yield

interesting findings that would not be available in more traditional culture model systems, for

reasons previously discussed.

The observations made in the slice culture model system while using minocycline

during WNV infection provided useful insight not only into potential roles for microglia

function, but more broadly for minocycline itself as a potent anti-inflammatory reagent.

Despite its widespread usage and utility, the actual underlying mechanism for how

minocycline works is unknown. This model system represents a future direction to

investigate the problem of minocycline mechanism, as it provides a more stable state for

microglia (given the retained cell-cell contacts to restrict unnecessary inflammatory

stimulation) and other cells, as well as a more representative depiction of the in situ cellular

environment. With the same reasoning, the use of other drugs and reagents could also be

investigated, with the ability to compare numerous metrics of antiviral responses to different therapeutics and assess their efficacy and mechanisms.

94 Slice cultures could also be used to address the functions of other cell types, including astrocytes and oligodendrocytes, during CNS viral infection. For WNV and other neurotropic viral infections, specific contributions by astrocytes and oligodendrocytes to viral detection and innate immune mechanisms remains underexplored. Investigations into how viruses with different tropisms (e.g., TMEV and oligodendrocytes, HIV and microglia, etc.) could prove to be another useful pursuit as a means to catalogue pan-viral infection profiles of the CNS parenchymal cells towards antiviral and inflammatory responses.

For practitioners of research into the mechanisms that regulate phagocytosis, many have alluded to the necessity of better model systems to explore how cells perform a range of functions related to phagocytosis, including the involved receptors, signaling cascades and intracellular cytoskeletal rearrangements, and the steps that govern phagolysosome formation and degradation (Fu 2014, Sierra 2013, Hellwig 2013). From the evidence reported here, this model system could be a lucrative avenue of research. Combining the slice culture model system with advanced and more refined techniques, further gains in knowledge related to phagocytic mechanisms could be achieved.

Future Directions: PLX5622 Model of WNV Infection

PLX5622 treatment during WNV infection in mice had a clear effect on survival and

WNV load in the brain, but many questions remain regarding how this effect comes about.

Microglia depletion is a major factor, but it is difficult to say with confidence that this is the only factor. Macrophages express CSF1R and it regulates their proliferation and differentiation (Patel and Player 2009, Li 2006), and the ability of PLX3397 to inhibit aspects of macrophage function is well-characterized (Chitu 2012, Coniglio 2012, DeNardo 2011, He

2012); therefore, some degree of peripheral immune cell activation mediated by macrophages

95 would be expected to be at least moderately reduced, but if there is any difference between

PLX3397 and PLX5622 in this regard had not been determined. While the fact that the

overall trajectory of body weight loss between control and PLX5622-treated mice did not

differ significantly during WNV infection provides a clue that peripheral infection was not

grossly exaggerated, future studies should be performed at earlier time points to assess WNV

growth in peripheral organs and viremia. Measuring activation profiles of macrophages

themselves during peripheral WNV infection in PLX5622 treated mice could also be a useful

pursuit in determining to what extent the observed phenotype seen during WNV infection

can be ascribed to microglia depletion.

Taking into account the potential effects of PLX5622 treatment in the periphery,

microglia depletion had a profound effect on the ability of WNV to infect neurons. How

exactly microglia contribute to the restriction and/or clearance of WNV from the CNS should

be the goal of subsequent studies with this model. Microglia depletion could have an effect

on the permeability of the BBB, thereby allowing WNV an easier route to the CNS compared

to controls; or, WNV is reaching the brain in a similar fashion for both PLX5622 and control

mice, but the lack of proper viral detection and inflammatory signaling allows WNV to grow

unchecked in the microglia-depleted mice; perhaps a combination of the two occurs.

Addressing this would go a long way in determining how not only how microglia provide

antiviral support, but could give support to previous studies that champion the permeability

of the BBB as a means of WNV entry into the CNS.

A final aspect of in vivo WNV infection that requires mention and has applicability to future PLX5622 studies is the role of the BBB itself in mediating WNV infection of the

CNS. In vivo studies on WNV neuroinvasive disease over the last decade and a half have

96 focused primarily on genetic knockout mouse models with data metrics largely comprised of viral titers, flow cytometry, and measures of cytokine/chemokine expression. When microscopy is used to assess WNV infection of the CNS, it is often done sparingly or to prove a point beyond simple characterization of WNV infection, while traditional microscopy sample preparations utilize brain and spinal cord slices that normally measure less than 5-7um.

In the microscopy analyses reported here, brains were sliced at a thickness of 100um, allowing for relatively thick regions of brain tissue to be assessed, which is likely the first time such analysis has occurred. Combined with an excellent antibody for WNV-E, it was observed that across both control and PLX5622-treated mice infected with WNV, as well as at a variety of time points, the blood vessels of the CNS were often infected with WNV to a considerable degree (Appendix). This was observed nearly in equal measure for both control and PLX5622-treated mice. More often than not, vascular networks were positive for WNV-

E somewhere within most brain sections examined. No direct relationship between the location of these WNV-E+ vascular networks and WNV-E+ CNS parenchymal cells was readily apparent, but clearly the ability of WNV to infect the vasculature of the CNS should be considered a likely mechanism of WNV transfer to the CNS (versus the idea of “passive transfer”), while still not ruling out other potential mechanisms of WNV entry.

The extent to which CNS vasculature is infectable with WNV raises important considerations for WNV neuroinvasive disease at large. Much of what is described clinically as WNV neuroinvasive disease often manifests predominantly as meningitis, though making clear distinctions between meningitis versus encephalitis is often difficult. Therefore, WNV neuroinvasive disease is often termed WNV meningoencephalitis. Given the findings

97 described here, it is possible that WNV infection of the CNS vasculature could be a large component of these types of infections. Future studies with PLX5622 should include a component concerning the extent to which the vasculature of the CNS is infected with WNV and if microglia depletion has any effect in this becoming worse or if WNV infectivity of the vasculature in particular regions can be connected to later infection in surrounding CNS parenchymal areas. Even future studies not focused on microglia depletion could make use of this observation.

98 REFERENCES

Acharya MM, Green KN, Allen BD, Najafi AR, Syage A, Minasyan H, Le MT, Kawashita T, Giedzinski E, Parihar VK, West BL, Baulch JE, Limoli CL. 2016. Elimination of microglia improves cognitive function following cranial irradiation. Sci Rep 6:31545.

Alliot F, Godin I, Pessac B. 1999. Microglia derive from progenitors, originating in the yolk sac, and which proliferate in the brain. Brain Res Dev Brain Res 117(2):145-52.

Antonios JP, Soto H, Everson RG, Moughon D, Orpilla JR, Shin NP, Sedighim S, Treger J, Odesa S, Tucker A, Yong WH, Li G, Cloughesy TF, Liau LM, Prins RM. 2017. Immunosuppressive tumor-infiltrating myeloid cells mediate adaptive immune resistance via a PD-1/PD-L1 mechanism in glioblastoma. Neuro Oncol 19(6):796-807.

Bai F, Town T, Qian F, Wang P, Kamanaka M, Connolly TM, Gate D, Montgomery RR, Flavell RA, Fikrig E. 2009. IL-10 signaling blockade controls murine West Nile virus infection. PLoS Pathog 5(10):e1000610.

Bardina SV, Michlmayr D, Hoffman KW, Obara CJ, Sum J, Charo IF, Lu W, Pletnev AG, Lim JK. 2015. Differential roles of chemokines CCL2 and CCL7 in monocytosis and leukocyte migration during West Nile virus infection. J Immunol 195:4306-4318.

Barger SW, Goodwin ME, Porter MM, Beggs ML. 2007. Glutamate release from activated microglia requires the oxidative burst and lipid peroxidation. J Neurochem 101(5):1205-13.

Beasley DW, Li L, Suderman MT, Barrett AD. 2002. Mouse neuroinvasive phenotype of West Nile virus strains varies depending upon virus genotype. Virology 296(1):17-23.

Bergold PJ. 2016. Treatment of traumatic brain injury with anti-inflammatory drugs. Exp Neurol 3:367-380.

Bessis A, Bechade C, Bernard D, Roumier A. 2007. Microglial control of neuronal death and synaptic properties. Glia 55(3):233-8.

Biber K, Neumann H, Inoue K, Boddeke HW. 2007. Neuronal ‘On’ and ‘Off’ signals control microglia. Trends Neurosci. 30:596-602.

Biber K, Owens T, Boddeke E. 2014. What is microglia neurotoxicity (not)? Glia 62:841- 854.

Block ML, Zecca L, Hong JS. 2007. Microglia-mediated neurotoxicity: uncovering the molecular mechanisms. Nat Rev Neurosci 8(1):57-69.

Brault AC, Huang CY, Langevin SA, Kinney RM, Bowen RA, Ramey WN, Panella NA, Holmes EC, Powers AM, Miller BR. 2007. A single positively selected West Nile viral mutation confers increased virogenesis in American crows. Nat Genet 39:1162-1166.

Bréhin AC, Mouriés J, Frenkiel MP, Dadaglio G, Després P, Lafon M, Couderc T. 2008. Dynamics of immune cell recruitment during West Nile encephalitis and identification of a new CD19+B220-BST-2+ leukocyte population. J Immunol. 180:6760-6767.

Brien JD, Daffis S, Lazear HM, Cho H, Suthar MS, Gale M Jr, Diamond MS. 2011. Interferon regulatory factor-1 (IRF-1) shapes both innate and CD8+ T cell immune responses against West Nile virus infection. PLoS Pathog. 7:e1002230. doi: 10.1371/journal.ppat.1002230.

Broderick C, Hoek RM, Forrester JV, Liversidge J, Sedgwick JD, Dick AD. 2002. Constitutive retinal CD200 expression regulates resident microglia and activation state of inflammatory cells during experimental autoimmune uveoretinitis. Am J Pathol 161(5):1669- 77.

Burguillos MA, Deierborg T, Kavanagh E, Persson A, Hajji N, Garcia-Quintanilla A, Cano J, Brundin P, Englund E, Venero JL, Joseph B. 2011. Caspase signaling controls microglia activation and neurotoxicity. Nature 472(7343):319-24.

Butovsky O, Hauben E, Schwartz M. 2001. Morphological aspects of spinal cord autoimmune neuroprotection: colocalization of T cells with B7—2 (CD86) and prevention of cyst formation. FASEB 15(6):1065-7.

Butovsky O, Talpalar AE, Ben-Yaakov K, Schwartz M. 2005. Activation of microglia by aggregated beta-amyloid or lipopolysaccharide impairs MHC-II expression and renders them cytotoxic whereas IFN-gamma and IL-4 render them protective. Mol Cel Neurosci 29(3):381-93.

Byrne SN, Halliday GM, Johnston LJ, King NJ. 2001. Interleukin-1beta but not tumor necrosis factor is involved in West Nile virus-induced Langerhans cell migration from the skin in C57BL/6 mice. J Invest Dermatol 117(3):702-9.

Cardona AE, Pioro EP, Sasse ME, Kostenko V, Cardona SM, Dijkstra IM, Huang D, Kidd G, Dombrowski S, Dutta R, Lee JC, Cook DN, Jung S, Lira SA, Littman DR, Ransohoff RM. 2006. Control of microglial neurotoxicity by the fractalkine receptor. Nat Neurosci 9(7):917- 24.

Carson PJ, Borchardt SM, Custer B, Prince HE, Dunn-Williams J, Winkelman V, Tobler L, Biggerstaff BJ, Lanciotti R, Petersen LR, Busch MP. 2012. Neuroinvasive disease and West Nile virus infection, North Dakota, USA, 1999-2008. Emerg Infect Dis 18(4):684-6.

Cheeran MCJ, Shuxian H, Sheng WS, Rashid A, Peterson PK, Lokensgard JR. 2005. Differential responses of human brain cells to West Nile virus infection. J Neurovirol. 11:512-524.

100 Chitu V, Nacu V, Charles JF, Henne WM, McMahon HT, Nandi S, Ketchum H, Harris R, Nakamura MC, Stanley ER. 2012. PSTPIP2 deficiency in mice causes osteopenia and increased differentiation of multipotent myeloid precursors into osteoclasts. Blood 120(15):3126-35.

Cho H, Shrestha B, Sen GC, Diamond MS. 2013. A role for Ifit2 in restricting West Nile virus infection in the brain. J Virol 87(15):8363-71.

Cho H, Diamond MS. 2012. Immune responses to West Nile virus infection in the central nervous system. Viruses 4(12):3812-30.

Chopra I, Roberts M. 2001. Tetracycline antibiotics: mode of action, applications, molecular biology, and epidemiology of bacterial responses. Microbiol Mol Biol Rev 65(2):232-60.

Chung KM, Thompson BS, Fremont DH, Diamond MS. 2007. Antibody recognition of cell surface-associated NS1 triggers Fc-gamma receptor-mediated phagocytosis and clearance of West Nile Virus infected cells. J Virol. 81:9551-9555.

Collin M, McGavern N, Haniffa M. 2013. Human dendritic cell subsets. Immunology 140(1):22-30.

Coniglio SJ, Eugenin E, Dobrenis K, Stanley ER, West BL, Symons MH, Segall JE. 2012. Microglial stimulation of glioblastoma invasion involves epidermal growth factor receptor (EGFR) and colony stimulating factor 1 receptor (CSF-1R) signaling. Mol Med 18:519-27.

Cunningham C, Wilcockson DC, Campion S, Lunnon K, Perry VH. 2005. Central and systemic endotoxin challenges exacerbate the local inflammatory response and increase neuronal death during chronic neurodegeneration. J Neurosci 25(40):9275-84.

Dagher NN, Najafi AR, Kayala KM, Elmore MR, White TE, Medeiros R, West BL, Green KN. 2015. Colony-stimulating factor 1 receptor inhibition prevents microglial plaque association and improves cognition in 3xTg-AD mice. J Neuroinflammation 12:139.

Das S, Dutta K, Kumawat KL, Ghoshal A, Adhya D, Basu A. 2011. Abrogated inflammatory response promotes neurogenesis in a murine model of Japanese encephalitis. PLoS One 6:e17225.

Davalos D, Grutzendler J, Yang G, Kim JV, Zuo Y, Jung S, Littman DR, Dustin ML, Gan WB. 2005. ATP mediates rapid microglial response to local brain injury in vivo. Nat Neurosci 8(6):752-8.

Davis CT, Beasley DW, Guzman H, Siirin M, Parsons RE, Tesh RB, Barrett AD. 2004. Emergence of attenuated West Nile virus variants in Texas, 2003. Virology 330(1):342-50.

101 Davis CT, Ebel GD, Lanciotti RS, Brault AC, Guzman H, Siirin M, Lambert A, Parsons RE, Beasley DW, Novak RJ, Elizondo-Quiroga D, Green EN, Young DS, Stark LM, Drebot MA, Artsob H, Tesh RB, Kramer LD, Barrett AD. 2005. Phylogenetic analysis of North American West Nile virus isolates, 2001-2004: evidence for the emergence of a dominant phenotype. Virology 342(2):252-65.

De I, Nikodemova M, Steffen MD, Sokn E, Maklakova VI, Watters JJ, Collier LS. 2014. CSF1 overexpression has pleiotropic effects on microglia in vivo. Glia 62(12):1955-67.

DeBiasi R, Tyler KL. 2006. West Nile virus meningoencephalitis. Nat Clin Pract Neurol 2:264-275.

Deininger MH, Meyermann R, Schluesener HJ. 2002. The allograft inflammatory factor-1 family of proteins. FEBS Lett. 514:115-121.

DeNardo DG, Brennan DJ, Rexhepaj E, Ruffell B, Shiao SL, Madden SF, Gallagher WM, Wadhwani N, Keil SD, Junaid SA, Rugo HS, Hwang ES, Jirström K, West BL, Coussens LM. 2011. Leukocyte complexity predicts breast cancer survival and functionally regulates response to chemotherapy. Cancer Discov 1(1):54-67.

Diamond MS, Shrestha B, Marri A, Mahan D, Engle M. 2003a. B cells and antibody play critical roles in the immediate defense of disseminated infection by West Nile encephalitis virus. J Virol 77(4):2578-86.

Diamond MS, Sitati EM, Friend LD, Higgs S, Shrestha B, Engle M. 2003b. A critical role for induced IgM in the protection against West Nile virus infection. J Exp Med 198(12):1853-62.

Dionne KR, Leser JS, Lorenzen KA, Beckham JD, Tyler KL. 2011. A brain slice culture model of viral encephalitis reveals an innate CNS cytokine response profile and the therapeutic potential of caspase inhibition. Exp Neurol. 228:222-231.

Donadieu E, Lowenski S, Servely JL, Laloy E, Lilin T, Nowotny N, Richardson J, Zientara S, Lecollinet S, Coulpier M. 2013. Comparison of the neuropathology induced by two West Nile virus strains. PLoS One 8(12):e84473.

Durrant DM, Robinette ML, Klein RS. 2013. IL-1R1 is required for dendritic cell-mediated T cell reactivation within the CNS during West Nile virus encephalitis. J Exp Med 210(3):503-16.

Dutta K, Kumawat KL, Nazmi A, Mishra MK, Basu A. 2010. Minocycline differentially modulates viral infection and persistence in an experimental model of Japanese encephalitis. J Neuroimmune Pharmacol 5:553-565.

Ebneter A, Kokona D, Jovanovic J, Zinkernagel MS. 2017. Dramatic effect of oral CSF-1R kinase inhibitor on retinal microglia revealed by in vivo scanning laser ophthalmoscopy. Transl Vis Sci Technol 6(2):10.

102

Ekdahl CT, Claasen JH, Bonde S, Kokaia Z, Lindvall O. 2003. Inflammation is detrimental for neurogenesis in adult brain. PNAS 100(23):13632-7.

Elmore MR, Najafi AR, Koike MA, Dagher NN, Spangenberg EE, Rice RA, Kitazawa M, Matusow B, Nguyen H, West BL, Green KN. 2014. Colony-stimulating factor 1 receptor signaling is necessary for microglia viability, unmasking a microglia progenitor cell in the adult brain. Neuron 82(2):380-97.

Emig M, Apple DJ. 2004. Severe West Nile virus disease in healthy adults. Clin Infect Dis 38(2):289-92.

Erblich B, Zhu L, Etgen AM, Dobrenis K, Pollard JW. 2011. Absence of colony-stimulating factor 1 receptor results in loss of microglia, disrupted brain development and olfactory deficits. PLoS One 6(10):e26317.

Fagan SC, Waller JL, Nichols FT, Edwards DJ, Pettigrew LC, Clark WM, Hall CE, Switzer JA, Ergul A, Hess DC. 2010. Minocycline to improve neurologic outcome in stroke (MINOS): a dose-finding study. Stroke 41(10):2283-7.

Feng X, Jopson TD, Paladini MS, Liu S, West BL, Gupta N, Rosi S. 2016. Colony- stimulating factor 1 receptor blockade prevents fractionated whole-brain irradiation-induced memory deficits. J Neuroinflammation 13(1):215.

Fu R, Shen Q, Xu P, Luo JJ, Tang Y. 2014. Phagocytosis of microglia in the central nervous system diseases. Mol Neurobiol. 49:1422-1434.

Gao HM, Zhou H, Zhang F, Wilson BC, Kam W, Hong JS. 2011. HMGB1 acts on microglia Mac1 to mediate chronic neuroinflammation that drives progressive neurodegeneration. J Neurosci 31(3):1081-92.

Gao M, Dong Q, Yao H, Zhang Y, Yang Y, Dang Y, Zhang H, Yang Z, Xu M, Xu R. 2017. Induced neural stem cells modulate microglia activation states via CXCL12/CXCR4 signaling. Brain Behav Immun 59:288-299.

Garrido-Mesa N, Zarzuelo A, Galvez J. 2013. Minocycline: far beyond an antibiotic. Br J Pharmacol 169:337-352.

Getts DR, Terry RL, Getts MT, Muller M, Rana S, Strestha B, Radford J, van Rooijen N, Campbell IL, King NJ. 2008. Ly6c+ “inflammatory monocytes” are microglial precursors recruited in a pathogenic manner in West Nile virus encephalitis. J Exp Med 205:2319-2337.

Ginhoux F, Greter M, Leboeuf M, Nandi S, See P, Gokhan S, Mehler MF, Conway SJ, Ng LG, Stanley ER, Samokhvalov IM, Merad M. 2010. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science 330(6005)841-5.

103 Glass WG, Lim JK, Cholera R, Pletnev AG, Gao JL, Murphy PM. 2005. Chemokine receptor CCR5 promotes leukocyte trafficking to the brain and survival in West Nile virus infection. J Exp Med. 202:1087-1098.

Glass WG, McDermott DH, Lim JK, Lekhong S, Yu SF, Frank WA, Pape J, Cheshier RC, Murphy PM. 2006. CCR5 deficiency increases risk of symptomatic West Nile virus infection. J Exp Med. 203:35-40.

Go M, Kou J, Lim JE, Yang J, Fukuchi KI. 2016. Microglial response to LPS increases in wild-type mice during aging but diminishes in an Alzheimer’s mouse model: implication of TLR4 signaling in disease progression. Biochem Biophys Res Commun 479:331-337.

Graeber MB, Streit WJ. 2010. Microglia: biology and pathology. Acta Neuropathol 119(1):89-105.

Graham JB, Thomas S, Swarts J, McMillan AA, Ferris MT, Suthar MS, Treuting PM, Ireton R, Gale M Jr, Lund JM. 2015. Genetic diversity in the collaborative cross model recapitulates human West Nile virus disease outcomes. MBio 6(3):e00493-15.

Hadas S, Reichert F, Rotshenker S. 2010. Dissimilar and similar functional properties of complement receptor-3 in microglia and macrophages in combating yeast pathogens by phagocytosis. Glia 58(7):823-30.

Hanisch UK, van Rossum D, Xie Y, Gast K, Misselwitz R, Auriola S, Goldsteins G, Koistinaho J, Kettenmann H, Moller T. 2004. The microglia-activating potential of thrombin: the protease is not involved in the induction of proinflammatory cytokines and chemokines. J Biol Chem 279:51880-51887.

Hayakawa K, Mishima K, Nozako M, Hazekawa M, Mishima S, Fujioka M, Orito K, Egashira N, Iwasaki K, Fujiwara M. 2008. Delayed treatment with minocycline ameliorates neurologic impairment through activated microglia expressiong a high-mobility group box1- inhibiting mechanism. Stroke 39(3):951-8.

Hayes EB, Sejvar JJ, Zaki SR, Lanciotti RS, Bode AV, Campbell GL. 2005. Virology, pathology, and clinical manifestations of West Nile virus disease. Emerg Infect Dis 11(8): 1174-9.

He Y, Rhodes SD, Chen S, Wu X, Yuan J, Yang X, Jiang L, Li X, Takahashi N, Xu M, Mohammad KS, Guise TA, Yang FC. 2012. c-FMS signaling mediates neurofibromatosis Type-1 osteoclast gain-in-functions. PLoS One 7(11):e46900.

Hellwig S, Heinrich A, Biber K. 2013. The brain’s best friend: microglial neurotoxicity revisited. Front Cell Neurosci 7:71.

104 Hewlett KA, Corbett D. 2006. Delayed minocycline treatment reduces long-term functional deficits and histological injury in a rodent model of focal ischemia. Neuroscience 141(1):27- 33.

Hirasawa T, Ohsawa K, Imai Y, Ondo Y, Akazawa C, Uchino S, Kohsaka S. 2005. Visualization of microglia in living tissue using Iba1-EGFP transgenic mice. J Neurosci Res. 81:357-362.

Hoeffel G, Wang Y, Greter M, See P, Teo P, Malleret B, Leboeuf M, Low D, Oller G, Almeida F, Choy SH, Grisotto M, Renia L, Conway SJ, Stanley ER, Chan JK, Ng LG, Samokhvalov IM, Merad M, Ginhoux F. 2012. Adult Langerhans cells derive predominantly from embryonic fetal liver monocytes with a minor contribution of yolk sac-derived macrophages. J Exp Med 209(6):1167-81.

Hoehn BD, Palmer TD, Steinberg GK. 2005. Neurogenesis in rats after focal cerebral ischemia is enhanced by indomethacin. Stroke 36(12):2718-24.

Hoek RM, Ruuls SR, Murphy CA, Wright GJ, Goddard R, Zurawski SM, Blom B, Homola ME, Streit WJ, Brown MH, Barclay AN, Sedgwick JD. 2000. Down-regulation of the macrophage lineage through interaction with OX2 (CD200). Science 290(5497):1768-71.

Hubalek Z, Halouzka J. 1999. West Nile fever – a reemerging mosquito-borne viral disease in Europe. Emerg Infect Dis 5(5):643-50.

Hunsperger EA, Roehrig JT. 2006. Temporal analyses of the neuropathogenesis of a West Nile virus infection in mice. J Neurovirol 12(2):129-39.

Husemann J, Loike JD, Anankov R, Febbraio M, Silverstein SC. 2002. Scavenger receptors in neurobiology and neuropathology: their role on microglia and other cells of the nervous system. Glia. 40:195-205.

Ito D, Imai Y, Ohsawa K, Nakajima K, Fukuuchi Y, Kohsaka S. 1998. Microglia-specific localization of a novel calcium binding protein, Iba1. Brain Res. 57:1-9.

Ito D, Tanaka K, Suzuki S, Dembo T, Fukuuchi Y. 2001. Enchanced expression of Iba1, ionized calcium-binding adapter molecule 1, after transient focal cerebral ischemia in rat brain. Stroke 32:1208-1215.

Jackson AC. 2012. Is minocycline useful for therapy of acute viral encephalitis? Antiviral Res 95(3):242-4.

Jung S, Schwartz M. 2012. Non-identical twins – microglia and monocyte-derived macrophages in acute injury and autoimmune inflammation. Front Immunol 3:89.

Kettenmann H, Hanisch UK, Noda M, Verkhratsky A. 2011. Physiology of microglia. Physiol Rev. 91:461-553.

105

Kierdorf K, Prinz M. 2013. Factors regulating microglia activation. Front Cell Neurosci. 7:44. doi:10.3389/fncel.2013.00044.

Kilpatrick AM, Meola MA, Moudy RM, Kramer LD. 2008. Temperature, viral genetics, and the transmission of West Nile virus by Culex pipiens mosquitoes. PLoS Pathog 4(6):e1000092.

Klee AL, Maidin B, Edwin B, Poshni I, Mostashari F, Fine A, Layton M, Nash D. 2004. Long-term prognosis for clinical West Nile virus infection. Emerg Infect Dis 10(8):1405-11.

Klein RS, Lin E, Zhang B, Luster AD, Tollett J, Samuel MA, Engle M, Diamond MS. 2005. Neuronal CXCL10 directs CD8+ T-cell recruitment and control of West Nile virus encephalitis. J Virol 79:11457-11466.

Kobayashi K, Imagama S, Ohgomori T, Hirano K, Uchimura K, Sakamoto K, Hirakawa A, Takeuchi H, Suzumura A, Ishiguro N, Kadomatsu K. 2013. Minocycline selectively inhibits M1 polarization of microglia. Cell Death Dis 4:e525. Kochan 2012

Kramer LD, Bernard KA. 2001. West Nile virus infection in birds and mammals. Ann NY Acad Sci 951:84-93.

Kumar M, Roe K, Orillo B, Muruve DA, Nerurkar VR, Gale M Jr, Verma S. 2013. Inflammasome adaptor protein apoptosis-associated speck-like protein containind CARD (ASC) is critical for the immune response and survival in West Nile virus encephalitis. J Virol 87:3655-3667.

Kumar M, Belcaid M, Nerurkar VR. 2016. Identification of host genes leading to West Nile virus encephalitis in mice brain using RNA-seq analysis. Sci Rep 6:26350.

Kumar Jha M, Lee WH, Suk K. 2016. Functional polarization of neuroglia: implications in neuroinflammation and neurological disorders. Biochem Pharmacol 1:103-116.

Lambertsen KL, Clausen BH, Babcock AA, Gregersen R, Fenger C, Nielsen HH, Haugaard LS, Wirenfeldt M, Nielsen M, Dagnaes-Hansen F, Bluethmann H, Faergeman NJ, Meldgaard M, Deierborg T, Finsen B. 2009. Microglia protect neurons against ischemia by synthesis of tumor necrosis factor. J Neurosci 29(5):1319-30.

Lazear HM, Pinto AK, Ramos HJ, Vick SC, Shrestha B, Suthar MS, Gale M Jr, Diamond MS. 2013. Pattern-recognition receptor MDA5 modulates CD8+ T cell-dependent clearance of West Nile virus from the central nervous system. J Virol 87(21):11401-15.

Li J, Chen K, Zhu L, Pollard JW. 2006. Conditional deletion of the colony-stimulating factor-1 receptor (c-fms proto-oncogene) in mice. Genesis 44(7):328-35.

106 Li X, Hanson C, Cmarik JL, Ruscetti S. 2009. Neurodegeneration induced by PVC-211 murine leukemia virus is associated with increased levels of vascular endothelial growth factor and macrophage inflammatory protein 1a and is inhibited by blocking activation of microglia. J Virol 83:4912-4922.

Li C, Yuan K, Schluesener H. 2013. Impact of minocycline on neurodegenerative disease in rodents: a meta-analysis. Rev Neurosci 24:553-562.

Liao TV, Forehand CC, Hess DC, Fagan SC. 2013. Minocycline repurposing in critical illness: focus on stroke. Curr Top Med Chem 13:2283-2290.

Liddelow SA, Guttenplan KA, Clarke LE, Bennett FC, Bohlen CJ, Schirmer L, Bennett ML, Munch AE, Chung WS, Peterson TC, Wilton DK, Frouin A, Napier BA, Panicker N, Kumar M, Buckwalter MS, Rowitch DH, Dawson VL, Dawson TM, Stevens B, Barres BA. 2017. Neurotoxic reactive astrocytes are induced by activated microglia. Nature 541:481-487.

Liu Z, Fan Y, Won SJ, Neumann M, Hu D, Zhou L, Weinstein PR, Liu J. 2007. Chronic treatment with minocycline preserves adult new neurons and reduces functional impairment after focal cerebral ischemia. Stroke 38(1):146-52.

Linnartz B, Wang Y, Neumann H. 2010. Microglial immunoreceptor tyrosine-based activation and inhibition motif signaling in neuroinflammation. Int J Alzheimers Dis pii:587463.

Lim PY, Behr MJ, Chadwick CM, Shi PY, Bernard KA. 2011. Keratinocytes are cell targets of West Nile virus in vivo. J Virol 85(10):5197-201.

Lindsey NP, Staples JE, Lehman JA, Fischer M; Centers for Disease Control and Prevention (CDC). 2010. Surveillance for human West Nile virus disease – United States, 1999-2008. MMWR Surveill Summ 59(2):1-17.

Lindsey NP, Staples JE, Lehman JA, Fischer M. 2012. Medical risk factors for severe West Nile virus disease, United States, 2008-2010. Am J Trop Med Hyg 87(1):179-84.

Loane DJ, Kumar A. 2016. Microglia in the TBI brain: the good, the bad, and the dysregulated. Exp Neurol 3:316-327.

Lyons A, Downer EJ, Crotty S, Nolan YM, Mills KH, Lynch MA. 2007. CD200 ligand receptor interaction modulates microglial activation in vivo and in vitro: a role for IL-4. J Neurosci 27(31):8309-13.

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.

107 Marín-Teva JL, Dusart I, Colin C, Gervais A, van Rooijen N, Mallat M. 2004. Microglia promote the death of developing Purkinje neurons. Neuron 41(4):535-47.

Markovic DS, Vinnakota K, van Rooijen N, Kiwit J, Synowitz M, Glass R, Kettenmann H. 2011. Minocycline reduces glioma expansion and invasion by attenuating microglial MT1- MMP expression. Brain Behav Immun 25:624-628.

Martinez FO, Gordon S. 2014. The M1 and M2 paradigm of macrophage activation: time for reassessment. F1000Prime Rep 6:13.

Michaelis M, Kleinschmidt MC, Doerr HW, Cinatl J Jr. 2007. Minocycline inhibits West Nile virus replication and apoptosis in human neural cells. J Antimicrob Chemother 60(5):981-6.

Mishra MK, Basu A. 2008. Minocycline neuroprotects, reduces microglial activation, inhibits caspase 3 induction, and viral replication following Japanese encephalitis. J Neurochem 105:1582-1595.

Möller T, Bard F, Bhattacharya A, Biber K, Campbell B, Dale E, Eder C, Gan L, Garden GA, Hughes ZA, Pearse DD, Staal RG, Sayed FA, Wes PD, Boddeke HW. 2016. Critical data-based re-evaluation of minocycline as a putative specific microglia inhibitor. Glia 64(10):1788-94.

Monje ML, Toda H, Palmer TD. 2003. Inflammatory blockade restores adult hippocampal neurogenesis. Science 302(5651):1760-5.

Mori I, Imai Y, Kohsaka S, Kimura Y. 2000. Upregulated expression of Iba1 molecules in the central nervous system of mice in response to neurovirulent influenza A virus infection. Microbiol Immunol 44:729-735.

Mott RT, Ait-Ghezala G, Town T, Mori T, Vendrame M, Zeng J, Ehrhart J, Mullan M, Tan J. 2004. Neuronal expression of CD22: novel mechanism for inhibiting microglial proinflammatory cytokine production. Glia 46(4):369-79.

Nagarakanti S, Bishburg E. 2015. Is minocycline an antiviral agent? A review of current literature. Basic Clin Pharmacol Toxicol 118(1):4-8.

Neher JJ, Emmrich JV, Fricker M, Mander PK, Théry C, Brown GC. 2013. Phagocytosis executes delayed neuronal death after focal brain ischemia. PNAS 110(43):e4098-107.

Nelson J, Roe K, Orillo B, Shi PY, Verma S. 2015. Combined treatment of adenosine nucleoside inhibitor NITD008 and histone deacetylase inhibitor vorinostat represents an immunotherapy strategy to ameliorate West Nile virus infection. Antiviral Res 122:39-45.

108 Nett RJ, Kuehnert MJ, Ison MG, Orlowski JP, Fischer M, Staples JE. 2012. Current practices and evaluation of screening solid organ donors for West Nile virus. Transpl Infect Dis 14(3):268-77.

Newell EW, Stanley EF, Schlichter LC. 2007. Reversed Na+/Ca2+ exchange contributes to Ca2+ influx and respiratory burst in microglia. Channels (Austin) 1(5):366-76.

Neumann H, Wekerle H. 2013. Brain microglia: watchdogs with pedigree. Nat Neurosci 16(3):253-5.

Nimmerjahn A, Kirchhoff F, Helmchen F. 2005. Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science 308(5726):1314-8.

O'Leary DR, Marfin AA, Montgomery SP, Kipp AM, Lehman JA, Biggerstaff BJ, Elko VL, Collins PD, Jones JE, Campbell GL. 2004. The epidemic of West Nile virus in the United States, 2002. Vector Borne Zoonotic Dis 4(1):61-70.

Olson JK, Miller SD. 2004. Microglia initiate central nervous system innate and adaptive immune responses through multiple TLRs. J Immunol 173(6):3916-24.

Orihuela R, McPherson CA, Harry GJ. 2016. Microglial M1/M2 polarization and metabolic states. Br J Pharmacol 173:649-665.

Pais TF, Figueiredo C, Peixoto R, Braz MH, Chatterjee S. 2008. Necrotic neurons enhance microglial neurotoxicity through induction of glutaminase by a MyD88-dependent pathway. J Neuroinflammation 5:43.

Paolicelli RC, Bolasco G, Pagani F, Maggi L, Scianni M, Panzanelli P, Giustetto M, Ferreira TA, Guiducci E, Dumas L, Ragozzino D, Gross CT. 2011. Synaptic pruning by microglia is necessary for normal brain development. Science 333(6048):1456-8.

Patel S, Player MR. 2009. Colony-stimulating factor-1 receptor inhibitors for the treatment of cancer and inflammatory disease. Curr Top Med Chem 9(7):599-610.

Paul AM, Acharya D, Duty L, Thompson EA, Le L, Stokic DS, Leis AA, Bai F. 2017. Osteopontin facilitates West Nile virus neuroinvasion via neutrophil “Trojan horse” transport. Sci Rep 7(1):4722.

Peng B, Xiao J, Wang K, So KF, Tipoe GL, Lin B. 2014. Suppression of microglial activation is neuroprotective in a mouse model of human retinitis pigmentosa. J Neurosci 34:8139-8150.

Pepperell C, Rau N, Krajden S, Kern R, Humar A, Mederski B, Simor A, Low DE, McGeer A, Mazzulli T, Burton J, Jaigobin C, Fearon M, Artsob H, Drebot MA, Halliday W, Brunton J. 2002. West Nile virus infection in 2002: morbidity and mortality among patients admitted to hospital in southcentral Ohio. CMAJ 168(11):1399-405.

109

Peppoloni S, Posteraro B, Colombari B, Manca L, Hartke A, Giard JC, Sanguinetti M, Fadda G, Blasi E. 2011. Role of (Mn)superoxide dismutase of Enterococcus faecalis in the in vitro interaction with microglia. Microbiology157(6):1816-22.

Petersen LR, Hayes EB. 2004. Westward ho? The spread of West Nile virus. NEJM 351(22):2257-9.

Piotrowska A, Kwiatkowski K, Rojewska E, Slusarczyk J, Makuch W, Basta-Kaim A, Przewlocka B, Mika J. 2016. Direct and indirect pharmacological modulation of CCL2/CCR2 pathway results in attenuation of neuropathic pain – in vivo and in vitro evidence. J Neuroimmunol 297:9-19.

Prada CE, Jousma E, Rizvi TA, Wu J, Dunn RS, Mayes DA, Cancelas JA, Dombi E, Kim MO, West BL, Bollag G, Ratner N. 2013. Neurofibroma-associated macrophages play roles in tumor growth and response to pharmacological inhibition. Acta Neuropathol 125(1):159- 68.

Prinz M, Mildner A. 2011. Microglia in the CNS: immigrants from another world. Glia. 59:177-187.

Qian F, Thakar J, Yuan X, Nolan M, Murray KO, Lee WT, Wong SJ, Meng H, Fikrig E, Kleinstein SH, Montgomery RR. 2014. Immune markers associated with host susceptibility to infection with West Nile virus. Viral Immunol 27:39-47.

Qin L, Liu Y, Wang T, Wei SJ, Block ML, Wilson B, Liu B, Hong JS. 2004. NADPH oxidase mediates lipopolysaccharide-induced neurotoxicity and proinflammatory gene expression in activated microglia. J Biol Chem 279(2):1415-21.

Ransohoff RM, Perry VH. 2009. Microglial physiology: unique stimuli, specialized responses. Annu Rev Immunol 27:119-45.

Ransohoff RM. 2016. A polarizing question: do M1 and M2 microglia exist? Nat Neurosci 19:987-991.

Rezaie P, Male D. 2002. Mesoglia & microglia – a historical review of the concept of mononuclear phagocytes within the central nervous system. J Hist Neurosci 11(4):325-74.

Ribes S, Ebert S, Regen T, Agarwal A, Tauber SC, Czesnik D, Spreer A, Bunkowski S, Eiffert H, Hanisch UK, Hammerschmidt S, Nau R. 2010. Toll-like receptor stimulation enhances phagocytosis and intracellular killing of nonencapsulated and encapsulated Streptococcus pneumoniae by murine microglia. Infect Immun 78(2):865-71.

Rock RB, Gekker G, Hu S, Sheng WS, Cheeran M, Lokensgard JR, Petersen PK. 2004. Role of microglia in central nervous system infections. Clin Microbiol Rev 17:942-964.

110 Rossum D, Hanisch UK. 2004. Microglia. Metab Brain Dis 19(3-4):393-411.

Samuel MA, Diamond MS. 2005. Alpha/beta interferon protects against lethal West Nile virus infection by restricting cellular tropism and enhancing neuronal survival. J Virol 79(21):13350-61.

Samuel MA, Morrey JD, Diamond MS. 2007. Caspase 3-dependent cell death of neurons contributes to the pathogenesis of West Nile virus encephalitis. J Virol. 81:2614-2623. Schafer 2012

Schittone SA, Dionne KR, Tyler KL, Clarke P. 2012. Activation of innate immune responses in the central nervous system during reovirus myelitis. J Virol. 86:8107-8118.

Schneider BS, Higgs S. 2008. The enhancement of arbovirus transmission and disease by mosquito saliva is associated with modulation of the host immune response. Trans R Soc Trop Med Hyg 102(5):400-8.

Schulz C, Gomez Perdiguero E, Chorro L, Szabo-Rogers H, Cagnard N, Kierdorf K, Prinz M, Wu B, Jacobsen SE, Pollard JW, Frampton J, Liu KJ, Geissmann F. 2012. A lineage of myeloid cells independent of Myb and hematopoietic stem cells. Science 336(6077):86-90.

Schwartz M, Butovsky O, Brück W, Hanisch UK. 2006. Microglial phenotype: is the commitment reversible? Trends Neurosci 29(2):68-74.

Shrestha B, Gottlieb D, Diamond MS. 2003. Infection and injury of neurons by West Nile encephalitis virus. J Virol 77(24):13203-13.

Shrestha B, Diamond MS. 2004. Role of CD8+ T cells in control of West Nile virus infection. J Virol. 78:8312-8321.

Shrestha B, Wang T, Samuel MA, Whitby K, Craft J, Fikrig E, Diamond MS. 2006a. Gamma interferon plays a crucial early antiviral role in protection against West Nile virus infection. J Virol 80:5338-5348.

Shrestha B, Samuel MA, Diamond MS. 2006b. CD8+ T cells require perforin to clear West Nile virus from infected neurons. J Virol 80:119-121.

Shrestha B, Diamond MS. 2007. Fas ligand interactions contribute to CD8+ T-cell-mediated control of West Nile virus infection in the central nervous system. J Virol 81:11749-11757.

Shrestha B, Pinto AK, Green S, Bosch I, Diamond MS. 2012. CD8+ T cells use TRAIL to restrict West Nile virus pathogenesis by controlling infection in neurons. J Virol 86:8937- 8948.

111 Shrestha B, Zhang B, Purtha WE, Klein RS, Diamond MS. 2008. Tumor necrosis factor alpha protects against lethal West Nile virus infection by promoting trafficking of mononuclear leukocytes into the central nervous system. J Virol. 82:8956-8964.

Sierra A, Abiega O, Shahraz A, Neumann H. 2013. Janus-faced microglia: beneficial and detrimental consequences of microglial phagocytosis. Front Cell Neurosci. 7:6. doi:10.3389/fncel.2013.00006.

Sivagnanam V, Zhu X, Schlichter LC. 2010. Dominance of E. coli phagocytosis over LPS in the inflammatory response of microglia. J Neuroimmunol 227(1-2):111-9.

Smithburn KC, Hughes TP, Burke AW, Paul JH. 1940. A neurotropic virus isolated from the blood of a native of Uganda. Am J Trop Med Hyg 20:471-3.

Spangenberg EE, Lee RJ, Najafi AR, Rice RA, Elmore MR, Blurton-Jones M, West BL, Green KN. 2016. Eliminating microglia in Alzheimer’s mice prevents neuronal loss without modulating amyloid-beta pathology. Brain 139(4):1265-81.

Stokes JA, Arbogast TE, Moya EA, Fu Z, Powell FL. 2017. Minocycline blocks glial cell activation and ventilatory acclimatization to hypoxia. J Neurophysiol jn.00525.2016.

Streit WJ, Walter SA, Pennell NA. 1999. Reactive microgliosis. Prog Neurobiol 57(6):563- 81.

Streit WJ. 2002. Microglia as neuroprotective, immunocompetent cells of the CNS. Glia 40(2):133-9.

Suthar MS, Ma DY, Thomas S, Lund JM, Zhang N, Daffis S, Rudensky AY, Bevan MJ, Clark EA, Kaja MK, Diamond MS, Gale M Jr. 2010. IPS-1 is essential for the control of West Nile virus infection and immunity. PLoS Pathog 6(2):e10000757.

Szretter KJ, Daffis S, Patel J, Suthar MS, Klein RS, Gale M Jr, Diamond MS. 2010. The innate immune adaptor molecule MyD88 restricts West Nile virus replication and spread in neurons of the central nervous system. J Virol. 84:12125-12138.

Takahashi K, Rochford CD, Neumann H. 2005. Clearance of apoptotic neurons without inflammation by microglial triggering receptor expressed on myeloid cells-2. J Exp Med 201(4):647-57.

Tremblay MÈ, Lowery RL, Majewska AK. 2010. Microglial interactions with synapses are modulated by visual experience. PLoS Biol 8(11):e10000527.

Ueyama T, Lennartz MR, Noda Y, Kobayashi T, Shirai Y, Rikitake K, Yamasaki T, Hayashi S, Sakai N, Seguchi H, Sawada M, Sumimoto H, Saito N. 2004. Superoxide production at phagosomal cup/phagosome through beta I protein kinase C during Fc gamma R-mediated phagocytosis in microglia. J Immunol 173(7):4582-9.

112

Vasek MJ, Garber C, Dorsey D, Durrant DM, Bollman B, Soung A, Yu J, Perez-Torres C, Frouin A, Wilton DK, Funk K, DeMasters BK, Jiang X, Bowen JR, Mennerick S, Robinson JK, Garbow JR, Tyler KL, Suthar MS, Schmidt RE, Stevens B, Klein RS. 2016. A complement-microglial axis drives synapse loss during virus-induced memory impairment. Nature 534(7608):538-43.

Verma S, Kumar M, Gurjav U, Lum S, Nerurkar VR. 2010. Reversal of West Nile virus- induced blood-brain barrier disruption and tight junction proteins degradation by matrix metalloproteinases inhibitor. Virology 397(1):130-8.

Wakselman S, Béchade C, Roumier A, Bernard D, Triller A, Bessis A. 2008. Developmental neuronal death in hippocampus requires the microglial CD11b integrin and DAP12 immunoreceptor. J Neurosci 28(32):8138-43.

Walker DG, Lue LF. 2015. Immune phenotypes of microglia in human neurodegenerative disease: challenges to detecting microglial polarization in human brains. Alzheimers Res Ther 7(1):56.

Walter TJ, Crews FT. 2017. Microglial depletion alters the brain neuroimmune response to acute binge ethanol withdrawal. J Neuroinflammation 14(1):86.

Wang Y, Lobigs M, Lee E, Müllbacher A. 2003. CD8+ T cells mediate recovery and immunopathology in West Nile virus encephalitis. J Virol 77(24):13323-34.

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

Wang Y, Szretter KJ, Vermi W, Gilfillan S, Rossini C, Cella M, Barrow AD, Diamond MS, Colonna M. 2012. IL-34 is a tissue-restricted ligand of CSF1R required for the development of Langerhans cells and microglia. Nat Immunol. 13:753-760.

Wei J, Gabrusiewicz K, Heimberger A. 2013. The controversial role of microglia in malignant gliomas. Clin Dev Immunol 2013:285246.

Wolf Y, Yona S, Kim KW, Jung S. 2013. Microglia, seen from the CX3CR1 angle. Front Cell Neurosci. 7:26. doi: 10.3389/fncel.2013.00026.

Youn S, Cho H, Fremont DH, Diamond MS. 2010. A short N-terminal peptide motif on flavivirus nonstructural protein NS1 modulates cellular targeting and immune recognition. J Virol. 84:9516-9532.

Zemke D, Majid A. 2004. The potential of minocycline for neuroprotection in human neurologic disease. Clin Neuropharmacol 27(6):293-8.

113 Zhang B, Gensel JC. 2014. Is neuroinflammation in the injured spinal cord different than in the brain? Examining intrinsic differences between the brain and spinal cord. Exp Neurol 258:112-120.

Zhu F, Zheng Y, Liu Y, Zhang X, Zhao J. 2014. Minocycline alleviates behavioral deficits and inhibits microglial activation in the offspring of pregnant mice after administration of polyriboinosinic-polyribocytidilic acid. Psychiatry Res 219:680-686.

Zou S, Foster GA, Dodd RY, Petersen LR, Stramer SL. 2010. West Nile fever characteristics among viremic persons identified through blood donor screening. J Infect Dis 202(9):1354- 61.

114 APPENDIX

WEST NILE VIRUS INFECTION OF THE BLOOD-BRAIN BARRIER

WNV-E staining in the vasculature of the brain was widespread, in both control mice and PLX5622-treated mice. Such staining could be detected as early as 3dpi, but became truly apparent at later time points (Figure A-1). High magnification of this staining pattern looked remarkably like blood vessels (Figure A-2, top). To more properly assess if this staining pattern was indicative of WNV in the blood vessels, CD31 antibody co-staining was performed (Figure A-2, bottom and Figure A-3). While the CD31 antibody was only average in its staining ability, clear overlaps were seen with CD31 and WNV-E. In most instances there did not appear to be a correlation to WNV in the vasculature and WNV in the CNS parenchyma, but in the PLX5622-treated mice there was an increase colocalization of WNV in the vessels and in the parenchyma; further analysis would have to be performed to determine if this is a real effect, given this data was taken from only a few mice.

115

Figure A-1. WNV infection of the blood-brain barrier in a control mouse brain. In the top panel, virtually the entire vasculature network of the right cortex is WNV-E+ at 10x. The bottom panel is at 40x. Scale bars: 200um (top), 40um (bottom).

116

Figure A-2. WNV infection of the blood-brain barrier. Top panel: 40x zoom of WNV-E+ antibody only. Bottom panel: co-stain with WNV-E (red) and CD31 (green). Scale bars: 40um.

117

Figure A-3. WNV infection of the blood-brain barrier. CD31 (green) and WNV-E (red) can be seen to overlap; in the control mouse brain, no WNV-E is seen beyond the staining of the vasculature, while in the PLX5622-treated mouse brain WNV-E can be seen both in the vasculature and parenchyma of the CNS. Scale bars: 40um.

118