SENSING of PICORNAVIRUS INFECTIONS Esther Francisco

SENSING OF PICORNAVIRUS INFECTIONS

Esther Francisco

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Sciences

COLUMBIA UNIVERSITY

2012

ABSTRACT

Sensing of Picornavirus Infections

Esther Francisco

The sensing of viral infection through pattern recognition receptors is necessary to trigger an immune response through the expression of interferons (IFNs) and cytokines.

For RNA viruses, which replicate in the cytoplasm, the cytoplasmic RIG-I like receptors are responsible for this sensing. The contributions of these receptors to sensing viruses of the Picornaviridae family were investigated. Using bone marrow derived macrophages from MDA-5 and RIG-I null mice we showed that encephalomyocarditis virus (EMCV) and coxsackievirus B3 (CVB3), picornaviruses of the cardiovirus and the enterovirus genus respectively, are sensed by both MDA-5 and RIG-I. Sensing of these viruses in macrophages leads to the expression of type I IFN early after infection with IFNβ expression reduced in the absence of each sensor, and IFNα expression reduced in the absence of MDA-5. However, in macrophages EMCV and CVB3 do not grow and we find that the sensing of viruses differs in murine embryonic fibroblasts (MEFs) and HeLa cells in which there is efficient replication. In MEFs RIG-I was found to be essential for the expression of type I IFNs but contribute to increases in viral titers in response to

CVB3 infections. MDA-5 inhibited CVB3 replication but in an IFN independent manner. In HeLa cells MDA-5 was found to have no effect on the sensing of EMCV but to be important for the sensing of poliovirus. However mutants of these viruses changed the involvement of MDA-5 in sensing infection, indicating picornaviruses can evade sensing through different pathways.

Table of Contents

List of Figures...... ii

List of Tables ...... vi

Acknowledgements ...... vii

Chapter I: Introduction...... 1

Chapter 2: Material and Methods...... 45

Chapter 3: Sensing of EMCV ...... 52

Chapter 4: Sensing of Coxsackievirus...... 70

Chapter 5: Sensing of Poliovirus ...... 87

Chapter 6: Discussion...... 97

References...... 116

i List of Figures

Figure 1. Picornaviridae genomes 5

Figure 2. The infectious cycle of picornaviruses 7

Figure 3. Picornavirus proteins 9

Figure 4. Intracellular forms of picornavirus RNA 11

Figure 5. Type I and Type III interferon expression 15

Figure 6. Interferon signaling 18

Figure 7. Toll like receptor signaling 23

Figure 8. RIG-I like receptor structures 25

Figure 9. RIG-I like receptor signaling 27

Figure 10. RIG-I like receptor ligands 31

Figure 11. RIG-I activation 33

ii Figure 12. Knockdown of MDA-5 in HeLa cells 54

Figure 13. Effect of MDA-5 knockdown on EMCV replication 56

Figure 14. Effects of IFNα pretreatment on EMCV replication 58

Figure 15. Effect of MDA-5 over expression on EMCV replication 59

Figure 16. Role of MDA-5 in inducing an antiviral state in HeLa cells infected with

EMCV 61

Figure 17. Basal levels of RIG-I and MDA-5 expression in macrophages 63

Figure 18. Role of MDA-5 and RIG-I in the induction of type I IFN in macrophages infected with EMCV 65

Figure 19. Role of MDA-5 and RIG-I in the induction of an antiviral state in MEFs infected with EMCV 67

Figure 20. Role of EMCV replication on the induction of an MDA-5 dependent antiviral state in MEFs 69

iii Figure 21. Role of MDA-5 and RIG-I in the induction of type I IFN in macrophages infected with CVB3 73

Figure 22. Role of RIG-I in the induction of an antiviral state in MEFs infected with

CVB3 75

Figure 23. Role of MDA-5 in the induction of an antiviral state in MEFs infected with

CVB3 77

Figure 24. Role of CVB3 replication on the induction of an antiviral state in MEFs in

MDA-5 and matched wild type MEFs 79

Figure 25. Effect of MDA-5 and IFNAR on CVB3 replication 81

Figure 26. Resistance of CVB3 growth to IFNα pretreatment 83

Figure 27. Role of MDA-5 and IFNAR on type III IFN expression in MEFs infected with

CVB3 86

Figure 28. Role of MDA-5 on IFNβ expression in HeLa cells infected with poliovirus

iv Figure 29. Role of MDA-5 on IFNβ expression in HeLa cells infected with PV-Y88L

Figure 30. Role of MDA-5 on IFNβ expression in HeLa cells infected with EMCV mutants. 93

Figure 31. Knockdown of RIG-I and IPS-1 in HeLa cells 95

Figure 32. Effect of apoptosis on poliovirus induced IFNβ expression 97

Figure 33. Coxsackievirus infection of MEFs 104

Figure 34. Expected response to viral sensing 111

Figure 35. Sensing in HeLa cells 114

List of Tables

Table 1. Picornaviruses 4

Table 2. RIG-I and MDA-5 sensed viruses 29

Table 3. qPCR primers 49

Table 4. Antibodies 51

Table 5. Summary of data from EMCV and CVB3 infections 107

Acknowledgements

I would like to thank first and foremost Vincent Racaniello, my advisor, for all his encouragement and guidance through out this process. I greatly appreciate the full support he's given to every idea and endeavor I've ever had and providing me with so many scientific opportunities and experiences. It has been a great privilege and joy to work in the Racaniello lab.

Christian Schindler, the chair of my committee, has also provided much support and guidance which I am very thankful for. His critical eye and open willingness for discussion has always been a source of new ideas.

With out the support of Michael Gale Jr. this work would not have been possible.

I cannot express how grateful I am for being able to have worked in his lab and have him be on my committee. His generosity with his lab, reagents and time is something that I am so thankful for.

Kang Liu has provided a great source of energy and optimism during in the last few years of my work in addition to helpful scientific guidance in and out of committee and lab meetings. I'd also like to thank her for the reading of my thesis.

I'd like to thank Stephen Goff for serving on my committee and his many helpful contributions through out my whole graduate career.

All current and past members of the 13th floor have contributed to this work through reagents, guidance, and friendship. The whole Racaniello lab, Saul Silverstein and the Silverstein lab, the Schindler lab, the Goff lab and the Liu Lab especially aidied

vii in this work. I'd also like to acknowledge the whole Gale Lab for their willingness to take me in and share their reagents, time, ideas and friendship with me.

I would like to thank all my friends and family for their support and encouragement. And last but not least the entire department of microbiology and immunology for the opportunity to perform this work and all their support through out it.

viii 1

Chapter I: Introduction

1. Picornaviruses

The Picornaviridae family is comprised of small RNA viruses that cause a variety of diseases in mammals with a diverse range of host tropism, severity and pathology.

Within the family there are twelve genera and twenty-eight species (184)(Table1). Of importance to humans are the members of the Hepatovirus, Enterovirus, Kobuvirus, and

Parechovirus genera, which contain viruses whose natural host is humans. Hepatitis A virus of the Hepatovirus genus is the etiological agent of hepatitis A, an acute hepatitis.

Rhinoviruses are species of the Enterovirus genus and are a major causative agent of the common cold. Within the same genus the more serious human disease, paralytic poliomyelitis is caused by the poliovirus, a member of the enterovirus C species. In 1948 another enterovirus was found to cause paralysis in Coxsackie, NY and named coxsackievirus. There are two groups of coxsackieviruses, A and B, which belong to the enterovirus A and enterovirus B species respectively. Seneca valley virus, whose natural host is pigs, is a notable picornavirus due to its selective ability to infect and lyse tumor cells in humans and has been through phase I trials for clinical use (67).

Two important animal viruses are also contained in the Picornaviridae family, encepohalomyocarditis virus (EMCV) and foot and mouth disease virus (FMDV), which belong to the Cardiovirus and Apthovirus genera respectively. FMDV is a highly infectious virus that can cause persistent infections of cloven-hoofed ruminants.

Although infections result in low mortality there is high degree of morbidity, which results in significant weight loss, and decreases in milk production and draught power in

2 livestock. EMCV infects a wide number of animal species to cause encephalitis and myocarditis, causing problems for both farm and zoo animals. Both of these picornaviruses are distinct from the more common human viruses of the Enterovirus genus in their coding for an extra non-structural protein, leader protein, their containing of a tract of over 200 cytosines (polyC tract) in their 5’ non-coding region, and the nature of the IRES (Figure1). The leader protein in FMDV is a protease that inhibits an antiviral response by cleavage of NF-κB, while in EMCV it lacks protease activity but interferes with an antiviral response by preventing IRF3 dimerization (41). The function of the polyC tract is unclear but shortening has been shown to decrease virulence. The use of a type II IRES by EMCV and FMDV rather than the type I IRES used by enteroviruses allows for infection of a broader tissue range (25).

Accordingly differences in functional properties of picornavirus IRESs correlate with differences in the structural properties of IRESs. The type I IRES used by enteroviruses contains six stem loop domains termed stem loops I-VI. The type II IRES used by cardioviruses and aphthoviruses contains twelve stem loop domains referred to as stem loops A-L. The nucleotide sequences between IRESs of the same type may vary considerably but the predicted secondary structures are highly conserved because theses structures are thought to be important in the formation of higher order structures through

RNA-RNA and RNA-protein interactions, which differ between IRES types. Proper higher order structures are necessary for the proper recognition of the IRES by translational machinery. In the case of type I IRESs the 40S ribosomal subunit is recruited to the IRES in a manner where it must scan the region for a start site. Type II

IRESs facilitate 40S ribosomal binding at the or near the start site, requiring no scanning

(25, 29, 133).

Table 1. Picornaviridae Family. Listing of all genera with in the picornavirus family and the number of species with in each genus in parenthesis. Species with notable animal and human pathogens are shown with examples of viral strain/types with in the species, host and diseases caused. Adapted from 29 and 183.

Figure 1. Picornavirus genomes. (A) Full genome sequences for isolates of representative genera and species are shown to scale, colored according to coding features, and aligned relative to the junction of 1D/2A. Missing 5’ terminal data (No seq) are estimated in each case as 100 bases. The 3’ polyA tail (40-100 bases) is not shown. (B) Key features of representative 5’ and 3’ UTRs are drawn to scale. Known RNA structure elements include terminal clover leafs (CL), terminal stems (S), type I psuedoknots (y), polyC tracts(C), oligo pyrimidine tracts (Y), spacer segments (SP), IRESs (type I, II, III or IV), ORF initiation codons (AUG), and ORF termination codons (circle with slash). (C) Structure of different picornavirus IRESs. Adapted from 25.

Infectious cycle

All members of the Picornaviridae family are single stranded positive sense RNA viruses with genomes in the range of 7-8Kb. They are non-enveloped viruses with an icosahedral capsid. The binding of this capsid to its receptor begins the infectious cycle

(Figure 2). For poliovirus this receptor is human CD155, also known as poliovirus receptor (PVR). Murine cells are permissive to poliovirus but not susceptible due to the insufficient homology between the murine and human CD155 preventing poliovirus entry into murine cells. The human coxsackievirus can use both human and murine coxsackievirus and adenovirus receptor (CAR) to gain entry into a host cell. The receptor for EMCV is not fully characterized. Vascular adhesion molecule 1 (VCAM-1) has been shown to be a receptor for EMCV on murine endothelial cells (47) but it is known that EMCV can also infect cells which do not express VCAM-1 such as HeLa and

K562 cells. In these cell lines a 70kDa sialoglycoprotein was found to act as a receptor for EMCV (53), the full identity of this protein however is still unknown.

Figure 2. The infectious cycle of picornaviruses. (1) Virus binds to its receptor and (2) releases its RNA genome into the cytoplasm. (3) The 5’ VPg protein is removed and cellular ribosomes bind the IRES to (4) translate the viral polyprotein. (5) The polyprotein is cleaved during and after synthesis to produce individual viral proteins (only the initial cleavage is depicted). (6) The nonstructural proteins encoded in the P2 and P3 regions of the polyprotein aid in the replication of the viral genome on membrane vesicles. (7) A switch is made to identify RNA as a template for replication and not translation. (8) The (+) strand RNA is copied into (-) strand RNA using a VPg primer that remains on the 5’ end. (9) The (-) strand RNA is used as a template to produce multiple (+) strand genomic RNA also using a VPg primer which remains on the 5’ end. (10) Some of the newly synthesized (+) strand RNA is used as a template for translation. (11) The P1 region of the polyprotein is cleaved to form the structural proteins that form the capsid which then (12) associate with (+) strand RNA to assemble progeny virus that is then (13) released. Adapted from 30.

Once bound the viral capsid undergoes conformational changes to release the

RNA genome into the cytoplasm. Since the genome is a positive sense strand of RNA with a polyA tail, cellular ribosomes directly bind the internal ribosomal entry site (IRES) to initiate cap independent translation and produce the viral poly protein (Figure 3). This protein is then cleaved into 7-8 mature and 3 functional precursor non-structural proteins as well as the 4 structural proteins that form the capsid. The non-structural (2A, 2B, 2C,

3A, 3B, 3C, 3D, 2BC, 3AB, 3CD, L) proteins play a role in altering the cell to evade immune responses and redirect the cellular machinery for the production of more viral proteins and genomes (13, 25, 29, 133).

Figure 3. Picornavirus proteins. Schematic of the proteins produced by picornaviruses. A single polyprotein is translated off of genomic RNA. 2A and 3C of enteroviruses are both proteases, 2Apro and 3Cpro, which cleave the polyprotein in cis to produce the P1, P2, and P3 fragments. These are further processed by cleavage to produce the 11-12 mature viral proteins as well as 3 precursors 2BC, 3AB, and 3CD; which have functions independent of their mature forms. Functions of proteins are indicated. “Pro” superscript indicates protease activity; “pol” super script indicates polymerase activity. Pink proteins contribute to protein processing. L protein is not present in all picornaviruses. Adapted from 13.

The production of viral genomes occurs on the surface of membrane vesicles induced by viral activation of the autophagy pathway by the 3A and 2BC proteins (64,

142, 149) and is performed by a viral RNA dependent RNA polymerase, 3Dpol. The polymerase 3Dpol is recruited to the viral RNA by interactions with protein complexes that form on the viral RNA at structured RNA sites in the 5’non-coding region, 3’ non- coding region and the internal cis-regulatory element (CRE). At the CRE 3Dpol interacts with a short viral protein 20-24aa long, 3B, and uses adenosines within a loop region of the CRE structure as a template to link uracil nucleotides to a tyrosine residue of 3B.

This viral 3B protein linked with uracil nucleotides then becomes a primer for the synthesis of negative sense strand viral RNA when bound to the positive strand polyA tail. The final product of negative strand synthesis is a fully duplexed double stranded

RNA consisting of a positive strand and negative strand RNA which can be detected by gel electrophoresis in infected cells as the replicative form (RF). (25, 29, 133)

Figure 4. Intracellular forms of picornavirus RNA. Following cell entry and uncoating, the picornavirus genomic RNA is altered by the cleavage of VPg from its 5’ terminus by an unidentified host cellular enzyme. Genomic picornavirus RNA molecules lacking VPg are the templates for translation. These templates for translation also serve as templates for (-) strand RNA synthesis, which results in a duplex of template and newly synthesized product RNA termed RF. Negative- strand RNA molecules act as template for positive-strand RNA synthesis in RI complexes. The RI complexes have multiple (+) strand RNAs synthesized from a single (-) strand template, resulting in asymmetric levels of positive versus negative strand viral RNAs in the infected cell. The (+) strand viral RNA molecules can serve as the templates for additional rounds of translation or (-) strand RNA synthesis, or are packaged into virions for subsequent infection of other host cells. Adapted from 25.

This newly synthesized negative strand of RNA is never found free but is used as the template for the production of multiple positive strand RNA genomes. During this process viral RNA is found as a replicative intermediate (RI) which consists of multiple strands of positive RNA partially duplexed to negative RNA (106, 127). Similar to negative strand synthesis the synthesis of positive strand RNA is performed by 3Dpol.

The uracil linked 3B protein is used as a primer resulting in positive sense viral genomes with the 3B linked to the 5’ end, also known as VPg (viral protein genome-linked). This protein is not necessary for viral infection but is a remnant of replication and is cleaved when viral genomes are released into the cytoplasm at the beginning of infection resulting in viral RNA with a pUp 5’end that is translated and replicated (4, 44, 104).

After viral genomes are produced viral capsid proteins are recruited to sites of replication by the viral 2C protein and RNA genomes are packaged to form progeny viruses (82,

105, 155). These newly formed infectious particles are then released by both non-lytic and lytic paths (25, 29, 30, 64, 133)(Figure 2).

2. Viral immunity

Defense against viral infections is comprised of both innate and adaptive immune responses. The adaptive response consists of lymphoid cells, such as B and T cells, that collectively harbor a large repertoire of antigen specific receptors, with each lymphocyte specifically recognizing one antigen. Mice deficient in these cells have shown their general importance for the clearance of many viral infections and survival of the host.

Without cytotoxic T-cells to eliminate virus-infected cells or B-cell to produce neutralizing antibodies viral infections can become persistent and more fatal (60).

The innate immune response is a non-specific first line of defense that serves to contain the initial infection, aid in tissue repair, alert the host of pathogen infection, and steer the nature of the immune response. The cellular mediators of these functions are monocytes, dendritic cells, natural killer cells, and polymorphonuclear leukocytes (60).

Of great importance to the function as well as activation of these cells are a group of small signaling molecules called cytokines. Various cytokines are initially released by infected cells, influencing the response of cells of the innate and in turn the adaptive immune response.

Interferons

A key group of cytokines in viral defense are interferons. Interferon was first identified 1957 as a secreted factor against heat inactivated influenza that “interfered” with infections by untreated influenza (49). By the 1970s it was determined that different cells could produce antigenically distinct species of interferon (42, 78). Fibroblasts produced one type, “F interferon”, while Leukocytes produced two, a minor amount of

“F interferon” but mainly “Le interferon” (42). More interferons have since been discovered and classified into three types based on their structure, receptor usage and function.

Type I interferons consist of the leukocyte and fibroblast interferons now known as IFNα and IFNβ respectively. This group also contains IFNδ, IFNω, IFNε, IFNκ, and

IFNτ. The IFNα, IFNβ, and IFNε type are found in mice and in humans. In addition to these three IFNs, humans also encode for IFNω and IFNκ. IFNδ and IFNτ are produced by the trophectoderm of pigs and ruminants respectively. Each class of type I interferons

14 contains just one species with the exception of IFNα of which there are 13 species of

IFNα in humans and 14 in mice. All type I IFNs signal through a ubiquitously expressed heterodimeric receptor composed of the transmembrane proteins IFNAR1 and IFNAR2

(16, 40, 116, 134, 154) (Figure 6). Expression of these IFNs is dependent on kinases that activate specific transcription factors, which bind to positive regulatory domains (PRDs) within IFN promoters (Figure 5). The kinases TBK1 and IKKε phosporylate IRF3 and

IRF7. This phosphorylation results in their dimerization, nuclear translocation and binding to PRDI and PRDIII. IKK (IκB kinase) phosporylates the IκB protein, releasing

NFκB to bind PRDII binding sites on the IFNβ promoter. The kinases p38 and JNK (c-

Jun N-terminal kinase) phosporylate ATF-2 and c-Jun. This results in their heterodimerization and binding to PRDIV with in the IFNβ promoter. The cooperative binding of these transcription factors leads to the expression of IFNβ. The IFNα promoters have similar PRD like elements (PRD-LE), which bind various IRFs to promote their expression. The differential expression and activation of various IRFs and other transcription factors such as NF-κB and ATF-2/c-Jun allow for differential expression of type I IFNs (28, 45, 134, 146).

Figure 5. Type I and Type III interferon expression. Sensing of viral infection leads to the activation of several transcription factors involved in inducing the expression of type I and type III interferons. Activation of IRFs (yellow/orange), NFkB (green), and AP1 (pink) leads to nuclear localization and the recognition of their binding sites (color coded) with in the promoters of different IFN genes. IFNβ and IFNλ1 promoters can be activated by IRF3 homodimers while IFNa and IFNλ2/3 promoters require IRF7. Adapted from 68.

Type II interferon consists of only IFNγ which signals through a different heterodimeric receptor consisting of IFNGR1 and IFNGR2 (Figure 6). Unlike type I

IFNs, IFNγ binds as a dimer and forms a complex that consists of two heterodimeric receptors. IFNGR1 is ubiquitously expressed while IFNGR2 expression is tightly regulated. This regulation controls the population IFNγ responsive cells, making IFNγ more of an immunomodulator than an antiviral cytokine. The expression of IFNγ is also different from Type I IFNs since it requires the transcription factor NFAT (nuclear factor activated T-cells) in addition to NFκB and AP-1 and is therefore only expressed by T- cells and NK cells. Additionally expression of INFγ is not induced by direct viral infection of NK or T-cells but in response to the cytokines IL-12 and

IL-18 produced by activated macrophages (27, 28).

Type III interferons were discovered in 2003 by two groups through sequence analysis of the human genome for cytokines and cytokine receptors similar to those for interferons and IL-10 (69, 136). This family was found to consist of three proteins

IFNλ1, IFNλ2 and IFNλ3 (also known as IL-29, IL-28A, and IL-28B respectively). In mice the gene for IFNλ1 is a pseudogene while IFNλ2 and IFNλ3 proteins are expressed

(74). In general Type III interferons seem to be quite similar to Type I interferons; they are induced by the same triggers of viral infection and TLR stimulation, regulated by the same transcriptions factors (Figure 5) and create an antiviral state after signaling through their receptors (6, 108, 112). A key difference between the type I and type III interferons are their receptor usage, the IFNλs signal through a heterodimeric receptor composed of

IFN-λR1 (IL-28RA) and IL-10R2 (IL-10Rβ) (69). In vivo studies have shown that the differential expression of IFN-λR1 compared to type I IFN receptors give these

17 interferons a more distinct role in viral infection, protecting mainly infections of the respiratory, gastro-intestinal tract and reproductive tract (5, 99, 100, 107, 123, 140).

Figure 6. Interferon signaling. Type I, II and III interferons are characterized by the use of different receptors complexes on the cell surface. Binding of the soluble IFN to receptor complexes activates kinases associated with the intracellular domains of the receptor chains. This leads to the phosphorylation and dimerization of cytoplasmic STAT proteins. STAT1 homodimers bind GAS sequences and lead to the induction of IFNγ induced genes such as ip-10, mig and irf-1. STAT1 and STAT2 heterodimers along with IRF9 form the transcription factor ISGF3 which binds ISRE sequences to induce the expression of the ISGs mx1, oas, irf-7, rig-I and mda-5. Adapted from 23.

Interferon signaling

Although the different IFNs signal through different receptors the end result of creating an antiviral state within a cell that interferes with viral replication is similar. The intracellular domains of the heterodimeric receptors associate with Janus protein tyrosine kinases. In the case of Type I and type III interferons it is the Jak1 and tyk2 kinases and for type II interferons the Jak1 and Jak2 kinases (Figure 6). Upon ligand binding these kinases are activated and phosporylate signal transducers and activators of transcription

(STAT) proteins, mainly STAT1 and STAT2. Once phosporylated these proteins dimerize to form two transcriptional activators. IFNα-activated factor (AAF), also known as IFNγ-activated factor (GAF), consists of a phosporylated STAT1 homodimer that translocates to the nucleus and binds IFNγ-activated sequences (GAS). The second transcriptional activator formed is IFN stimulated gene factor 3 (ISGF3), which consists of a phosphorylated heterodimer of STAT1 and STAT2 and the interferon regulatory factor 9 (IRF9) which binds to interferon stimulated regulatory elements (ISREs).

Activation of the type I and type III IFN pathways primarily result in the formation of

ISGF3 while activation of the type II IFN path in the formation of GAF. The GAS and

ISRE sequences these transcription factors bind are contained with in the promoters of hundreds of interferon-stimulated genes (ISGs) which are expressed after IFN signaling

(16, 23, 28, 45, 68, 146). It is the expression of these ISGs which result in the antiviral state of cells produced by interferons as many of these ISGs encode for proteins with antiviral activities. The viral sensors MDA-5 and RIG-I are an example of some ISGs induced by IFN (55, 56, 141).

2'-5' oligoadenylate synthetase (OAS) is another ISG which is particularly relevant

20 to picornavirus infections. OAS works in conjunction with RNaseL by converting ATP into 2’-5’-oligoadenylates when activated by dsRNA. These 2’-5’-oligoadenylates activate the latent nuclease RNaseL which then dimerizes and degrades RNA molecules

(22). It has been shown that OAS and RNase L expression is able to reduce replication of

EMCV (17, 179). Additionally mice deficient in RNaseL have an increased mortality

(180).

3. Viral sensing – Pattern recognition receptors

In order for IFNs to be expressed, antiviral states mounted and the cellular immune system activated, the host must first realize that it has been infected with by a viral pathogen. This recognition occurs through receptors within cells which recognize molecules characteristic of viruses – namely the unique genomes or genomic intermediates of viruses. There are two well-known classes of receptors that sense infection of RNA viruses, the toll-like receptors (TLRs) and the RIG-I like receptors

(RLRs).

Recently a set of cytoplasmic helicases (DDX1, DDX21 and DDX36) related to

RIG-I, but not with in the RLR family, has also been shown to cooperatively act in the sensing of RNA in myeloid dendritic cells (178). DDX1 was found to bind polyI:C through it’s helicase domain while DDX21 and DDX36 bound the adaptor molecule

TRIF (TIR-domain-containing adaptor inducing IFNβ) through its TIR (Toll/Interleukin-

1 receptor) domain. The RNA bound DDX1 was bridged to TRIF and DDX36 through interactions with DDX21. This DDX1-DDX21-DDX36-TRIF complex was found to translocate to the mitochondria after binding both long and short poly I:C to induce IFNα

21 and IFNβ expression. Knockdown of any of the helicases in this complex has been shown to reduce the levels of IFNβ expressed after infection with Influenza A and

Reovirus in DCs, but further experiments need to be done to characterize other viruses and RNA structures this complex may sense in cells other than DCs.

Toll Like Receptors

There are currently thirteen mammalian TLRs, ten human TLRs and twelve murine TLRs. TLR 1-9 are shared between humans and mice, TLR10 is expressed in humans but is a pseudogene in mice, and TLR 11-13 are only present in mice (2, 3, 61,

76). The ligands have been determined for all TLRs but the human TLR10, and the murine TLR12 and TLR13 (76).

The TLR family of proteins are type I transmembrane proteins which function as homo or heterodimers and can reside either on the plasma membrane or endosomal membranes (Figure 7B). They contain an extracellular or endosomal domain comprised of leucine rich repeats which bind to pattern associated molecular patterns (PAMPs) and a cytoplasmic Toll/IL-1 receptor homology domain (TIR) which associates with adaptor proteins to transmit signals. With the exception of TLR3 all TLRs associate with the adaptor protein MyD88 (Myeloid differentiation primary response gene 88) which signals through TRAF6 (TNF associated factor 6) to activate the transcription factors NFκB and

AP-1 and initiate the production of inflammatory cytokines. TLR3 uses the adaptor protein TRIF (TIR-domain-containing adaptor inducing IFNβ) which utilizes TRAF3 to activate IRFs in addition to NFκB and AP-1 (2, 61, 147, 185) (Figure 7B).

Activation of IRFs is crucial for the expression of IFNs and an antiviral response.

The TLRs involved in an antiviral response are TLR3, TLR4, TLR7/8 and TLR9 (Figure

7A). TLR3 recognizes dsRNA, TLR4 recognizes viral envelope proteins, TLR7/8 recognizes ssRNA and TLR9 recognizes DNA with GpG motifs (164). TLR4 activates

IRFs through interactions with TRIF as well as MyD88, however TLR 7/8 and TLR9 only signal through MyD88. These TLRs have been found to activate IRFs through

MyD88 dependent associations with TRAF6, IRAK1 (interleukin-1 receptor-associated kinase), and IRAK4, but how this MyD88-TRAF6-IRAK1-IRAK4- signaling differs from the MyD88-TRAF6-IRAK1-IRAK4 signaling of TLR1, TLR2, TLR5 and TLR6 which do not result in IRF activation is unknown. It has been proposed that the location of TLR7/8 and TLR9 within endosomes rather than the plasma membrane is involved in the determination of IRF activation by this signaling cascade (71).

Figure 7. Toll like receptor signaling. (A) Toll like receptors and their ligands. The main TLRs involved in sensing viral pattern recognition molecules are TLR7and TLR8 which recognize viral ssRNA and TLR3 which recognizes viral dsRNA. TLR 9 which detects viral DNA and TLR4 which can detect viral envelope proteins (B) Schematic of TLR signaling. TLR 3, 7/8 and 9 are found in the endosomes while other TLRs are localized on the plasma membrane. Most TLRs use the adaptor protein MyD88 to activate the transcription factors NFkB, AP-1 and IRFs through TRAF6. TLR 3 uses the adaptor protein TRIF and TRAF3 to activate IRFs. These TLRs along with TLR 9 reside in the endosomes. Adapted from 163 and 184.

RIG-I Like Receptors

The RLR family of receptors contains three members. RIG-I and MDA-5 are the two full length members of the family that consist of two N-terminal CARD domains, a

DExD/H Box RNA Helicase domain and a C terminal Domain CTD. Lgp-2 is similar in structure to RIG-I except that it lacks the N-terminal CARD domains (171, 173) (Figure

8A). This domain transmits the signal of infection through homotypic interaction with the card domain on the adaptor protein IPS-1. The Helicase domain is involved in the binding of RNA which results in conformational changes and ATP hydrolysis to allow for exposure of the CARD domains. The CTD contains a basic loop that binds RNA in

RIG-I and Lgp-2, but is clipped in MDA-5 to form an open flat surface that does not bind

RNA (144) (Figure 8B). The CTD of RIG-I and Lgp-2 have also been shown to have repressor functions which MDA-5 does not have.

Expression of these genes are detectable by microarray in most healthy human tissues (137, 169). Quantitative PCR studies have shown thelevels of expression to be comparable among most tissues to levels in the spleen in humans, while expression in mice is greatest in the kidney and spleen (75). Although detectable, the basal levels of

RLR message are low compared to GAPDH but can be induced upon stimulation with type I IFNs (52, 141).

Figure 8. RIG-I like receptor structure. The RLR family consists of three members: RIG-I, MDA-5 and LGP-2. (A) RIG-I and MDA-5 contain a CARD domain, helicase domain and C terminal domain while LGP2 only contains the helicase and C terminal domain. The C terminal domain of RIG-I and LGP-2 are also repressor domains (RD). Amino acid identity of the CARD domains and helicase/C terminal domain are indicated. (B) NMR based structure of the C terminal domain of each RLR member with a schematic representation beneath. RIG-I and LGP- 2 structures are shown bound to RNA while MDA-5 cannot bind RNA due to the flat open structure of the basic loop responsible for RNA binding. Adapted from 172 and 143

Immuno-precipitation and FRET studies have shown that signaling through these receptors results in a transmembrane dependent dimerization of IPS-1 allowing it to then signal through TBK1, FADD and RIP1 to lead to the activation of IRF3 and NF-κB to transcribe IFNs and other cytokines (9, 62, 79, 85, 93, 135, 146, 148, 168). It has also been shown that IPS-1 localizes to peroxisomal membranes as well as mitochondrial membranes. Signaling through peroxisomal IPS-1 can directly induce the early expression of ISGs under the control of IRF1 in the absence of IFNs (21, 26) (Figure 9).

Figure 9. RIG-I like receptor signaling. Infection with different viruses produces different forms of RNA which are recognized by different RLRs. RNA binding to RIG-I leads to conformational changes which result in RIG-I ubiquitnination and dimerization to allow signaling. This is promoted by SUMOlaytion and inhibited by phosphorylation and ISGylation with ISG15. MDA-5 also undergoes a conformational change upon RNA binding and is positively regulated by SUMOlaytion. Signaling can occur through the adaptor protein IPS-1 which resides on peroxisomes and mitochondria. Signaling through peroxisomal IPS-1 results in an early response that activates IRF1 and IRF3 which can directly induce the expression of ISGs. Signaling through the mitochondrial IPS-1 leads to the activation of IRF3, IRF7 and NFkB which induce the expression of Type I IFN and cytokines that must then signal through their receptors to turn on ISG expression create a delayed antiviral response. Adapted from 26

RIG-I

RIG-I was the first RLR to be identified as a sensor through a functional screen for proteins that could induce IFN by dsRNA stimulation, using the dsRNA mimic pI:C

(172). Later studies over-expressing, knocking down or knocking out RIG-I expression have shown its importance in sensing viruses from the Paramyxoviridae, Rhabdoviridae,

Orthomyxoviridae, Filoviridae, Arenaviridae, Bunyaviridae and Flaviviridae families

(37, 46, 57, 59, 83, 84, 88, 122, 143, 171, 181) (Table 2). The majority of these virus families contain negative single stranded RNA genomes which do not produce detectable amounts of dsRNA during replication (119, 163). This observation lead to the discovery that RIG-I sensing is independent of viral replication and senses ssRNA with a

5’triphosphate (119, 143, 145, 163). Recent studies involving pull-downs of RIG-I from influenza infected cells and analysis of bound RNA indicate that panhandle structures from defective-interfering (DI) genomes could be the ligand for RIG-I (12, 126) (Figure

10).

Table 2. Virus sensed by RIG-I and MDA-5. Viruses within indicated families sensed by RIG-I, MDA-5 or both. Nature of viral genomes within each family are indicated as (-) or (+) sense, single or double strand RNA, and segmented (S) or non-segmented (NS). Citations for each virus appear in the right column. Adapted from 84.

A more thorough study on RIG-I RNA interactions performed by Takahasi et. al. looked at the structural requirements of RNA and the ATPase and helicase requirements of RIG-I for binding and signaling. This study showed that recombinant RIG-I could bind 5’triphosphate ssRNA and dsRNA with a blunt end, 5’ overhang or 3’ overhang and pI:C independent of ATP but not 5’monophosphate ssRNA. ATPase activity was essential for unwinding which occurred with dsRNA with a blunt or 3’overhang ends

(145). Helicase activity was negatively correlated with IRF activation (8). IRF dimerization was found to occur only with dsRNA with at least one strand monophosphorylated and ssRNA with a 5’ tri-phosphate. Binding of these species of RNA was also found to induce ATPase activity (145). Although a 5’triphosphate is necessary for ssRNA recognition it is not sufficient for the sensing of HCV genomes which also requires polyU/UC region in the 3’NTR of the HCV genome (130, 153). Additionally even self RNA can also become a ligand for RIG-I upon RNase L cleavage which produces RNA with a 5’hydroxyl and 3’ mono-phosphate (89, 90) (Figure 10).

Figure 10. RIG-I like receptor ligands. MDA-5 has been shown to induce IFNβ promoter activity upon stimulation with undigested polyI:C greater than 2Kb in length. RIG-I stimulation occurs with shorter polyI:C ligands of less than 300bp in length. RIG-I treatment with dsRNA with at lease one strand with a monophosphate, ssRNA with a 5’ triphosphate, RNase L cleaved RNA with a 5’ hydroxyl and 3’ monophosphate and the double stranded panhandle structure of ssRNA DI particles have also been shown to induce IFN promoter activity.

In addition to the ligand requirements of RIG-I conformational requirements for signaling have been elucidated. Protease treatment of recombinant RIG-I in the presence of dsRNA or 5’ triphosphate ssRNA results in the formation of a protease resistant polypeptide of 30kDa and in the presence of pI:C a resistant polypeptide of 60kDa, indicating conformational changes upon ligand binding controlling signaling (131, 145).

In vivo studies using native gel electrophoresis show the formation of a multimeric RIG-I complex upon infection and in vitro studies using gel filtration indicate stable dimer formation upon ligand binding (19, 131). A recent crystal structure of the helicase and C- terminal domain of RIG-I suggest that in the absence of RNA RIG-I is in an inactive state with interactions between the helicase domain, CTD and CARD domains preventing the availability the CARD domains to bind IPS. Upon RNA binding the CARD domains are displaced and the helicase and CTD domains wrap around the RNA strand, extending the

CARD domains away from the RNA (87) (Figure 11).

Figure 11. RIG-I activation. (A) Linear schematic of RIG-I domains. Helicase consists of helicase domain1 (HEL1), helicase domain 2 (HEL2) and Helicase domain 2 insertion (HEL2i) followed by a pincer region (P) and the C terminal domain (CTD). (B) Crystal structure of boxed region in (A) bound to RNA. (C) Schematic of conformational changes during RIG-I activation. RIG-I is in a closed inactive form with CARD domains buried in interactions with helicase domains. Binding of RNA (pink circle) to helicase domains and CTD results in the exposure of CARD domains. ATP hydrolysis occurs and RIG-I molecules multermerize along a strand of RNA to activate MAVS. Adapted from 87

Post-translational modifications regulating RIG-I signaling have also been identified. The K63 poly-ubiquitination of the CARD domains by TRIM25 was the first modification found to be necessary for RIG-I to bind IPS-1 and initiate signaling (34).

Further ubiqutination of the N-terminal domain and C-terminal domain by REUL and

RIPLET/RNF135 respectively also stimulates RIG-I activity (35, 109, 111). Addition of the small ubiquitin-like modifier 1 (SUMO-1) onto RIG-I has been shown to increase its ubiquitination, interactions with IPS-1 and signaling (94). Phosphorylation of the CARD domains at S8 or T170 can block TRIM25 binding and ubiquitination, and inhibit RIG-I signaling (33, 103). However, if direct ubiquitination of the CARD domains is prevented by mutation it has been shown that in a reconstituted in vitro system the CARD domains can also bind free K63 poly-ubiquitin chains to mediate signaling (176). It has not been determined whether phosphorylation inhibits binding of free poly-ubiquitin chains to reduce signaling. Conjugation of ISG15, a ubiquitin-like protein, to RIG-I has been shown to reduce the levels of IFNβ promoter activity after stimulation (63). This

ISGylation of RIG-I is thought to work as a negative feedback loop to regulate the amount of IFN produced after the initial sensing. Additionally poly-ubiquitination of

RIG-I by RNF125, an E3 ligase expressed upon IFNα or pI:C treatment leads to proteasomal degradation of RIG-I as another negative feed back loop controlling RIG-I sensing (7) (Figure 9). Aside from posttranslational modification, RIG-I signaling was shown to be positively regulated through direct binding with ZAPS (Zinc Finger

Activated Protein Short) and DDX60. (43, 95).

A knockout mouse for RIG-I was created to study the in vivo role of RIG-I during viral infection and was found to be mainly embryonic lethal between embryonic days

12.5 and 14.0, with only a few mice reaching birth. These mice all died with in 3 weeks of birth, highlighting a role for RIG-I aside from an antiviral response (57). RIG-I was first identified as an induced gene after retinoic acid treatment of leukemia cells indicating a role in carcinogenesis, apoptosis, senescence and differentiation (81). A second knockout mouse targeting another region of RIG-I resulted in viable and fertile mice but developed a colitis-like phenotype associated with an increase in effector and decrease in naïve T-cells (159). This mouse also possessed an increase in granulocytes, showing a role for RIG-I as a negative regulator in the development and differentiation of

T-cells and granulocytes (177). Unexpectedly it was also shown that RIG-I plays a role in phagocytosis using this mouse since they were found to be more susceptible to E. coli infection and RIG-I-/- macrophages showed a significant reduction in the ability to phagocytose GFP-E. coli (66).

MDA-5

MDA-5 was first identified as a gene that was up regulated in melanoma cells induced to terminal differentiation by IFNβ and appropriately named Melanoma

Differentiation Associated protein -5 (52). After the full-length human gene was cloned, over-expression experiments showed that it was able to suppress the growth of melanoma cells and contribute to apoptosis (55, 56). Cloning of the murine MDA-5 showed its ability to increase the rate of apoptosis through a caspase cleavage between the card and helicase domains. This cleavage of the full-length cytoplasmic protein allowed for the helicase domain to translocate into the nucleus and accelerate DNA fragmentation – a hallmark of apoptosis (70).

Although MDA-5 was known to be induced by IFNβ its antiviral role was not known until searches for mammalian DExD/H helicases similar to RIG-I identified it along with LGP2 as a related protein (171) (Figure 8A). MDA-5 was then shown to activate IFNβ reporters upon pI:C treatment through IPS interactions, similar to RIG-I

(62, 72, 171).

MDA-5 signaling is generally similar to that of RIG-I. It exists in an inhibited form under physiological conditions, but un-like RIG-I this inhibited form is controlled by interactions with dihydroxyacetone kinase (DAK) through its NTD (20). Upon viral infection DAK disassociates with MDA-5. Recent structural studies on MDA-5 support a model where anti-parallel MDA-5 dimers cooperatively bind to dsRNA through the helicase domain to form a filament of MDA-5 visible by EM (15, 115). ATP hydrolysis is not necessary for binding but does promote dissociation and is inversely proportional to filament length, indicating a possible mechanisms for the discrimination of dsRNA length in sensing (115). This oligomerized MDA-5 can then bind to and promote IPS-1 oligomerization and signaling (8, 62, 72). Similar to RIG-I signaling, MDA-5 is positively regulated by SUMOlaytion and negatively regulated by RNF125 ubiquitination

(7, 32).

Viruses sensed by MDA-5 and/or RIG-I

Two groups made MDA-5 knockout mice in 129x1/SvJ and C57BL/6 backgrounds to study the effects of this sensor during different viral infections (36, 59).

Unlike RIG-I knockout mice these mice were viable and showed no defects in any lymphoid cell populations indicating that MDA-5 does not have a role in lymphoid

37 development as RIG-I does (36). IFNβ ELISAs of RIG-I-/- and MDA-5-/- murine embryonic fibroblasts (MEFs) infected with VSV, SeV, NDA, Influenza, JEV and

EMCV showed that RIG-I was important for the sensing of all these viruses except for

EMCV, which was sensed by MDA-5 (59). Additionally MDA-5 deficient mice infected with EMCV were more susceptible to infection than WT mice leading to the belief that

RIG-I and MDA-5 are responsible for sensing different families of viruses. RIG-I senses viruses of the Rhabdoviridae, Paramyxoviridae, Orthomyxoviridae and Flavivridae families and MDA-5 senses the synthetic pI:C and members of the Picornaviridae family. Additionally studies comparing Norovirus infections in MDA-5 and TLR3 deficient mice show that a virus of the Caliciviridae family is sensed by MDA-5 (91).

Although mouse hepatitis virus (MHV), a member of the Coronaviridae family, induces minimal amounts of type I IFN in most tissues in brain microglia cells from WT, RIG-I,

TLR3 and MyD88 deficient mice, IFNβ expression is detectable but reduced in MDA-5-/- microglia cells indicating that MDA-5 senses Coronaviridae infections (128). MHV and a human coronavirus were shown to induce more IFN when the 2’-O-methyltransferase was mutated leading to 5’cap structures lacking 2’-O-methylation. These viruses were shown to be sensed well by MDA-5 (183). Recently the retrovirus SIV was shown to produce dsRNA sensed by MDA-5 in brains (18).

Later studies infecting RIG-I-/- and MDA-/- MEFs showed that for the Flaviviridae

Dengue Virus and West Nile Virus both RIG-I and MDA-5 are individually dispensable for ISG expression, showing that some flaviviruses are sensed by both MDA-5 and RIG-I

(31, 83). Over expression of MDA-5 can also potentiate the sensing of measles virus and influenza virus, showing that viruses of the Paramyxoviridae and Orthomyxovirdae

38 families can be sensed by MDA-5 as well as RIG-I (14, 138). However a study on the sensing of DI particles of Sendai virus shows that this sensing is done through MDA-5, making it unclear if MDA-5 is sensing of paramyxoviruses is of the replicative viruses or only the byproduct of infection, defective interfering particles (174) (Table 2).

The factors that determine if a virus is sensed by MDA-5, RIG-I or both has therefore become a topic of interest. One of the main determinants of MDA-5 vs. RIG-I sensing elucidated so far is dsRNA length. By treating pI:C with RNase III for varying amounts of time various lengths of dsRNA were produced. When MDA-5-/- MEFs were treated with these RNAs the relative induction of IFNβ produced decreased with increased time of RNase treatment, indicating that sensing through RIG-I increased as pI:C length decreased. The inverse was seen in RIG-I-/- MEFs indicating that sensing through MDA-5 decreased as pI:C length decreased (58). Additionally treatment of these

MEFs with RNA isolated from mock or Reovirus infected cells showed that Reovirus was sensed by both RIG-I and MDA-5. The Reovirus genome consists of three dsRNA segments of different lengths: S (1.2-1.4kbp) M (2.2-2.3kbp) and L (3.9kbp). When

MEFs were treated with purified RNA of each segment it was shown that sensing of the S segment was dependent on RIG-I, sensing of M mainly on RIG-I but also MDA-5 and sensing of L dependent on both equally (58). However, immuno-precipitation of dsRNA from EMCV infected cells, which was shown to induce IFN through MDA-5, revealed a prominent dsRNA band of >11kbp which was not able to stimulate IFNβ production, indicating that long dsRNA alone cannot be sensed by MDA-5. Only the very high molecular weight complex containing dsRNA and ssRNA that would be stuck in the wells of an agarose gel was able to stimulate IFNβ production through MDA-5,

39 proposing that high molecular weight RNA complexes are necessary to stimulate signaling through MDA-5 (120). While many different RNA structures have been identified for the activation of RIG-I, the only structural characteristic for RNA to activate MDA-5 is size, however, the size of picornavirus genomes is one of the smallest

(Figure 10).

LGP2

The third member of the RLR family, LGP2, can bind RNA through its C terminal domain but lacks the CARD domains which facilitate the transmission of a signal through interactions with the CARD domain of the RLR adaptor protein IPS1 (51,

65, 145). Over expression of LGP2 was shown to reduce the expression of ISRE, NFκB and PRDI reporters by infection with SeV and NDV, while knockdown of LGP2 resulted in an increase of reporter activity (129, 171) leading to the belief that the absence of

CARD domains gave LGP2 a role as a negative regulator of RIG-I and MDA-5 signaling.

Studies have also shown that LGP2 has a greater affinity for RNA than RIG-I and the RD domain of Lpg2 can interact with that of RIG-I and disrupt its ability to dimerize, leading to possible mechanisms of LGP2 negative regulation of RLR signaling (80, 121, 131).

However Lgp2 knockout mice have been made by two groups reporting that in vivo infections of Lgp2 deficient mice by the picornavirus EMCV results in an increased lethality and a decrease in cytokine production arguing a case where LGP2 acts a positive regulator of RLR signaling (132, 156). Although data on the role of LGP2 in EMCV infection agrees between these two groups they show contradicting data on the role of

LGP2 as a negative or positive regulator of other RLR ligands such as VSV and pI:C which has yet to be clarified (132, 156).

5. Viral sensing and picornaviruses

Initially a study by Kato et. al. showed that there was an increase in mortality in mice deficient in MDA-5, MyD88 and IFNAR when infected with EMCV but not in mice deficient in RIG-I and TLR3 (59). In vitro experiments infecting both MEFs and

BMDCs from RIG-I knockout mice at varying multiplicities of infection showed no decrease in IFNβ expression at a low MOI and only a slight decrease in IFNβ expression at a high MOI. EMCV infections of MDA-5-/- MEFs and bone marrow derived dendritic cells (BMDC) showed a complete ablation of IFNβ expression at various MOIs while bone marrow derived macrophages (BMDM) and peritoneal macrophage infections show a significant decrease in IFNα, IL-6 and MCP-1 at higher MOIs which disappears at lower MOIs (36). These data led to the conclusion that MDA-5 is the main sensor of picornaviruses.

However, later studies infecting TLR3 deficient mice with either a myocarditic variant of EMCV or a diabetes inducing variant (EMCV-D) showed an increase in mortality, disease and viral titers in TLR3-/- mice compared to wild type mice indicating a role for TLR3 as a sensor for picornavirus infections (39, 91). Consistent with the role of a sensor the loss of TLR3 resulted in an early decrease in IFN during EMCV-D infections, although this decrease was transient. On the contrary, infections with the myocarditic EMCV resulted in a transient increase in IFNβ message in cardiac tissue.

These studies show how a lack of increased mortality does not correspond with a lack of

41 viral sensing and immune response. One study has shown a lack of TLR3 involvement in

EMCV infection but other studies have shown that with different viral strains and infection conditions TLR3 does play a crucial role in viral infection. It is therefore reasonable to investigate whether RIG-I, which has been previously shown to not be involved in sensing EMCV, has a role in picornavirus infections.

Studies using a picornavirus from the Enterovirus genus, Coxsackievirus, highlight the importance of mouse strains during in vivo studies. Using mice deficient in

MDA-5 in either a C57BL/6 background or a 129x1/SvJ background, infections with

CVB3 at 105 pfu/mouse showed mice in the B6 background lacking MDA-5 had a significantly greater mortality than their wild type counterparts. Meanwhile mice in a

129 background, even those lacking MDA-5, showed no mortality at all and no difference in health was noticed between the control and MDA-5-/- mice (48). This makes it even more difficult to conclude that RIG-I does not play a role in the sensing of picornaviruses since the MDA-5-/- mice and RIG-I-/- mice were made in different murine backgrounds, whose effects could be masking the role of RIG-I.

Additionally in vivo studies on the model picornavirus poliovirus further exemplify the importance of background strain as well as underscore the importance of infection conditions. Mice deficient in TLR3 and its adaptor protein TRIF both show increased lethality compared to wild type controls after intravenous infection with poliovirus ranging from 102 pfu to 104 pfu. This correlates nicely with an increase in viral load and a decrease in IFNα, OAS and IRF-7 expression, indicating TLR3 as an important sensor for PV infections. However intravenous infections of MDA-5 knockout mice in an ICR background also show an increased lethality compared to their wild type

42 counter parts, but only when infected with 104 pfu. Interestingly MDA-5 knockout mice in a B6 background infected with the same amount of virus actually have a greater survival rate than their wild type controls whether infected intravenously or intracerebrally. Increasing the infectious dose to 105 pfu intravenously in the B6 mice reversed the phenotype of a loss of MDA-5 resulting in an increase in survival to a decrease in survival (1).

The myriad of cell types involved in in vivo infections could be contributing to this variability in results. Infection of mice with EMCV-D show an increase in lethality in both MDA-5-/- and TLR3-/- mice compared to a wild type control. By reconstituting γ- irradiated MDA-5-/- or WT mice with bone marrow from WT or MDA-5-/- mice respectively bone marrow chimeric mice were made. Infection of WT mice reconstituted with MDA-5-/- bone marrow (MDA5-/-  WT) showed complete resistance to EMCV-D infections, while those containing WT bone marrow in an MDA-5-/- mouse (WT 

MDA-5-/-) were susceptible to infection, similar to MDA-5-/- mice. This indicates that the role of MDA-5 in sensing is important in non-hematopoietic, γ-irradiation resistant compartments. Meanwhile similar bone marrow chimeras made with TLR3-/- mice showed that TLR3 protects against infection through the hematopoietic compartment

(92).

In natural infections virus first replicates at the site of infection before progeny virus are released into the bloodstream to have direct access to multiple tissues. The difference in sensors used in various tissues therefore leads to questions regarding the interpretation of in vivo models where systemic infections are directly established through intraperitoneal and intravenous injection of virus. The demonstrated complexity

43 of the immune response in vivo makes the relative importance of sensors in these studies difficult to interpret on their own.

From current studies we have come to know that MDA-5 plays a role in the sensing of EMCV (36, 59, 92), Theilers (54), Coxsackie (48, 158) and poliovirus (1).

Additionally we know through in vivo work that TLRs are also involved in the sensing of

EMCV (39, 92), CVB3 (102), and PV (1). This has lead to the in vitro examination and confirmation for the role of MDA-5 and TLRs in the sensing of rhinoviruses (139, 152,

161, 162). However, little work has been done looking at the role of RIG-I in the sensing of picornaviruses. Papon et. al. has shown that during infections with EMCV RIG-I is degraded by both the picornaviral 3C protease as well as caspases, which can inhibit

IFNβ expression through RIG-I. (114) Our lab has also shown that the cleavage of RIG-I by EMCV, as well as the picornaviruses poliovirus, rhinovirus 16 and echovirus (11).

Additionally it has been shown that the 3C of enterovirus 71 can inhibit sensing through

RIG-I in a cleavage independent manner (77).

This would seem to argue that RIG-I is irrelevant in picornavirus infections due to their ability to inhibit signaling through RIG-I. However, it has also been shown that a number of these picornaviruses are also able to inhibit sensing by MDA-5 and TLRs.

Our lab has also shown that MDA-5 is cleaved upon infections with poliovirus, EMCV, and rhinovirus 1a (10). Rhinovirus1a, Coxsackie and Hepatitis A infections have all been shown to result in the cleavage of IPS-1, the adaptor molecule for RLRs (24, 86, 102,

170). The picornaviral 3C protease of Coxsackie and Hepatitis A virus has also been shown to cleave TRIF, the adaptor molecule for TLR3 (102, 124). Despite the ability of

44 picornaviruses to dampen signals through these pathways it has been shown in many cases that these pathways can still confer protection to picornavirus infections.

We therefore set out to study the role of the cytoplasmic sensors RIG-I and MDA-

5 in vitro with various cell types and picornaviruses viruses to discern the role theses sensors play in sensing virus in these cells and their affect on viral replication. Using mice that were deficient in RIG-I and MDA-5 we were able to look at the role of RIG-I and MDA-5 in EMCV and CVB3 infections as model picornaviruses of the cardiovirus and enterovirus genera capable of infecting murine cells. We were able to investigate the differences of sensors in the immune cells such as macrophages and “resident tissue cells”, fibroblasts. Additionally using human HeLa cells KD for MDA-5 we could study its effect on the important human picornavirus, poliovirus.

Chapter 2: Material and Methods

Constructs and cell lines

HeLa S3 and 293A cells were maintained in Dulbecco’s modified essential medium (DMEM; Invitrogen, Carlsbad, CA) supplemented with 10% bovine calf serum

(Hyclone, Logan, UH) and 1% penicillin-streptomycin solution (Invitrogen). Over expression of MDA-5 was achieved by transfection of C terminal flagged MDA-5 in pcDNA3 using FuGene6 (Roche, Indianapolis, IN) according to manufacture protocol.

Knockdown cell lines were produced by infection with retrovirus expressing miR30 based short hairpin RNAs (shRNAs) designed against target genes. Short hairpin RNA sequence for MDA-5 (5’TGCTGTTGACAGTGAGCGCACTGACATAAG

AATCAATAAATAGTGAAGCCACAGATGTATTTATTGATTCTTATGTCAGTTTG

CCTACTGCCTCGGA3’) and RIG-I (5’TGCTGTTGACAGTGAGCGCTCAGCTACA

GGGAATGAGTAATAGTGAAGCCACAGATGTATTACTCATTCCCTGTAGCTGA

ATGCCTACTGCCTCGGA3’) targets the 3’untranslated region of the cellular mRNA and will not affect the expression of exogenously expressed constructs. The IPS-1 shRNA sequence (5’TGCTGTTGACAGTGAGCGATACCACCTTGATGCCTGTGAA

TAGTGAAGCCACAGATGTATTCACAGGCATCAAGGTGGTAGTGCCTACTGCC

TCGGA3’) targets exon 5 of IPS-1 which is shared among IPS-1 variant. Short hairpin

RNA targeted against Luciferase was used as a control and previously published (24)

Retroviruses were constructed by co-transfection of psPAX2, pMD.G and pAPM containing the shRNA sequence in 293A. Two days after transfection supernatants were harvested and used to infect target HeLa cells for knockdown. Infected HeLa cells were selected by puromycin treatment.

MDA-5-/- and matched wild type MEFs in a 129x1/SvJ background were a gift from Marco Collona. Washington University. IFNAR-/- MEFs were also obtained on a

129x1/SvJ background as a gift from David Levy, New York University. RIG-I-/- and their matched wild type controls in a mixed ICR/B6/129 background were a gift from

Michael Gale, University of Washington. MEFs were maintained in DMEM supplemented with 15% Fetal Bovine Serum (Atlanta Biologicals, Lawrenceville, GA),

1mM sodium pyruvate (Invitrogen), 2mM L-glutamine (Invitrogen), 1% penicillin- streptomycin, 0.1mM Non-essential amino acids (Invitrogen).

MDA-5-/- and wild type control mice in a 129x1/SvJ background were a gift from

Paula Longhi, Rockefeller University. RIG-I-/- and wild type control mice in a mixed

ICR/B6/129 background were a gift from Michael Gale, University of Washington. Bone marrow from mice was harvested and cultured in DMEM supplemented with 20% L-cell conditioned media, 10% Fetal Bovine Serum, and 1% penicillin/streptomycin for 5-10 days for 6-8days.

Viruses

T7 polymerase (Fermentas) In vitro transcribed full length viral genomes were produced from plasmids containing DNA copies of PV, poliovirus type 1 Mahoney

(pT7M); PV-Y88L, a poliovirus mutant containing a mutation from a tyrosine to leucine at residue 88 of 2A (pT7M-2A-Y88L); EMCV (pEC4); EP4, a EMCV-PV chimeric virus containing the poliovirus 2A protein with in the EMCV genome at the P1/P2 junction

(pEC4-EP4), or CVB3 (pSVCVB3) (38, 101, 125, 151). Viral stocks were produced by transfection of transcribed RNA into HeLa cells using DEAE-dextran (113). The

47 chimeric EP4C4 virus was isolated from plaque-purified clones of EP4 selected for their ability to replicate well in IFNα pretreated cells.

Infections

Infections were performed in 35mm dishes. Cells were counted at the time of infection and viral stocks were diluted to for infections at the appropriate MOI in

PBS+0.01%BSA. Two hundred microliters of diluted virus was used to infect each plate at 37oC for 45min with rocking. Cells were then washed twice with PBS and overlaid with 1mL of appropriate media.

qPCR

Total RNA was harvested using a RNeasy Kit (Qiagen, Valencia, CA) with

DNase I treatment (Qiagen) according to manufacturer protocol. Complementary DNA was generated using iScript cDNA synthesis kit (BioRad, Hercules, CA) using both random hexamer and polyT primers according to manufacturer protocol. Complementary cDNA was diluted 1:10 and qPCR done on individual genes using Syber Green PCR

Master Mix (Applied Biosystems, Foster City, CA) and appropriate primers found on

Table 3. Quantitative PCR was performed in triplicate using an ABI 7500 real time PCR

Ct machine. Data was analyzed using the 2ΔΔ method as previously described using β-actin as the calibrator gene for murine qPCR and GAPDH as the calibrator for human qPCR

(117). Primers used are as listed in Table 3.

Table 3. qPCR primers

Western Blot Analysis

Cells were lysed in RIPA buffer and protein concentrations determined using

Quick Start Bradford Dye Reagent 1x (BioRad). Equal amounts of protein were run on

10% SDS polyacrylimide gel and transferred overnight onto nitrocelluose membranes

(EMD Millipore, Billerica, MA) and blocked for an hour in PBST+5% nonfat dry milk.

Membranes were probed overnight with primary unconjugated antibody diluted in blocking buffer and washed 5 times with PBST before probing with secondary HRP conjugated antibody. Membranes were incubated with appropriate HRP conjugated antibodies for one hour and washed 5 times with PBST again. Blots were developed using Amersham ECL plus (GE Life sciences, Piscataway, NJ). Antibodies and dilutions used are as listed in Table 4.

Table 4. Antibodies

Plaque assays

Total virus samples were harvested by collecting supernatants from infected cells and supernatant after three cycles of freezing and thawing followed by centrifugation at

5000xg for 5 min. All titrations were done on monolayers of HeLa cells in six well plates. Ten fold serial dilutions were made of virus in PBS+0.001% BSA (sigma, St.

Louis, MO) and 0.1mL of viral dilutions was added per well. Plates were incubated at

37oC for 45min with rocking and then overlay added. For PV and CVB3 a single agar overlay of 0.8% Bacto-agar (sigma) DMEM, 5%BCS, 0.05%NaHCO3, was used. For

EMCV a two-overlay system was used. Overlay 1 was an agar overlay with a final composition of DMEM, 0.8% noble agar, 1%BCS, 0.2% NaHCO3, 50mM MgCl2, 1% non-essential amino acids. Overlay 2 was a liquid overlay with a final composition of

DMEM, 0.1%BSA, 40mM MgCl2, 0.2% glucose, 2mM sodium pyruvate, 4mM L- glutamine, 4mM oxaloacetic acid (Sigma), and 0.2% NaHCO3. PV, CVB3 and EMCV plaque assays were incubated for 48hrs at 37oC. Cells were then fixed with 10%TCA

(Sigma) and stained with 0.1% crystal violet (Sigma) in 20% ethanol.

Chapter 3: Sensing of EMCV

Knock down of MDA-5 does not affect sensing of EMCV in HeLa cells

All previous literature has shown that the picornavirus EMCV is sensed through

MDA-5 (36, 59, 91). Therefore our first step in investigating the sensing of picornaviruses was to create a HeLa cell stably knocked down for MDA-5 which would be susceptible and permissive to human picornaviruses and the pan-tropic EMCV. Basal expression of MDA-5 was knocked down in one cell line to undetectable levels by western blot (Figure 12A). Upon stimulation of MDA-5 expression by IFNα or pI:C pretreatment MDA-5 protein could be detected in the knockdown cell line, however expression was greatly reduced compared to a control cell line containing a miRNA directed against luciferase, a gene not expressed in HeLa cells (Figure 12).

Figure 12. Knockdown of MDA-5 in HeLa cells. HeLa cells were infected with lentivirus targeted to knockdown MDA-5 expression and infected cells selected by puromyocin treatment. (A) Indicated cells lines were mock treated or treated with 20ug/mL of polyI:C for 16 hrs and cell extracts were analyzed by western blot. (B) Indicated cell lines were mock treated or treated with 1000U/mL of IFNa for 16 hrs. Cell extracts were harvested and analyzed by Western blot.

We hypothesized that in the absence of MDA-5 type I IFN would not be expressed upon EMCV infection and at a low MOI, which allows for multiple rounds of replication, we would therefore see an increase in viral titers since EMCV is known to be sensitive to type I IFN (17, 101). However, infection of control and MDA-5 knockdown

HeLa cell lines showed no difference in viral titers. (Figure 13)

Figure 13.Effect of MDA-5 knockdown on EMCV replication. MDA-5 and control knockdown cell lines were infected with EMCV at an MOI of 0.1. Samples were harvested at indicated time points and viral titers determined by plaque assay. Representative data from one experiment of three independent experiments done in duplicate.

Because basal levels of MDA-5 are low, knockdown cell lines were pre-treated with 0, 10, or 100 Units/mL of IFNα to increase the difference of MDA-5 present at the time of infection between MDA-5 and control knockdown cell lines, therefore increasing the difference in viral replication at low a low MOI. IFNα pretreatment resulted in a decrease in viral titers in both cell lines with an increase in concentration of IFNα with only a very slight increase in viral titers in the MDA-5 knockdown line compared to control (Figure 14). To ensure that the expression of other ISGs induced by IFNα treatment, which caused the overall decrease in viral titers in both the control and MDA-5 knockdown cell lines, were not masking the effects of MDA-5 on viral replication we also over expressed MDA-5 in both cell lines and looked for differences in viral titers after infection at a low MOI, but again saw no difference (Figure 15).

Figure 14. Effects of IFNα pretreatment on EMCV replication. Control and MDA-5 knockdown cell lines were treated with IFNa at indicated concentrations for 18 hours and then infected with EMCV at an MOI of 0.1. At various points post infection total virus was harvested and titers determined by plaque assay. Representative data from one experiment of two independent experiments done in duplicate.

Figure 15. Effect of MDA-5 over expression on EMCV replication. Control and MDA-5 knockdown cell lines were stably transfected with a plasmid expressing MDA-5 or empty vector. (A) Lysates from transfected cells were harvested and western blot analysis performed against MDA-5 and GAPDH (loading control). (B) Cell lines were infected with EMCV at an MOI of 0.1 and total virus harvested at various time points post infection. Titers were determined by plaque assay. Representative data from one experiment of two independent experiments done in duplicate.

Quantitative PCR was performed to examine the expression of type I IFN and downstream ISGs after EMCV infection. Only a slight decrease in IFNβ induction was seen in MDA-5 knockdown cell lines compared to control cells (Figure 16B). The induction of IFNβ did indeed produce an antiviral state within the cells as expression of the ISGs RIG-I, OAS and PKR were increased after infection in both cell lines, however there was little to no difference in the amount of ISG expression between cell lines

(Figure 16 B, C). This lack of difference in an antiviral state between cell lines however is likely not the cause for the lack of difference in viral titers since the expression of both

IFNβ and ISGs occurs late, nine to twenty four hours post infection. Viral replication however peaks at 9 hours post infection, before the induction of IFN and ISGs (Figure

16A). Therefore the lack of difference in viral titers is more likely due to the lack of any effective response early in infection.

Figure 16. Role of MDA-5 in inducing an antiviral state in HeLa cells infected with EMCV. MDA-5 and control knockdown cell lines were infected at an MOI=0.1 with EMCV. At indicated times post infection (A) total virus was harvested and titers assayed by plaque assay. (B) Total RNA was harvested and gene expression determined by qPCR for (B) IFNβ (C) indicated ISGs. Values were normalized to β- actin levels and fold change calculated relative to mock (M) infected samples. Representative data from one experiment of two independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

RIG-I is important and MDA-5 is essential for sensing EMCV infection of macrophages

To confirm that the lack of difference in EMCV titers in control and MDA-5 knockdown HeLa cells was not due to an incomplete knockdown of MDA-5 expression or sensing through an alternative path we tested EMCV infection in cells knocked out for

MDA-5 or RIG-I. Bone marrow derived macrophages from MDA-5-/- mice in a

129x1/SvJ background (36) and RIG-I -/- mice in a mixed ICR background (59) and matched wild type controls were infected with EMCV at a low MOI. Type I IFN induction was observed in WT cells from both backgrounds after EMCV infection, however with different kinetics. In a mixed background there was a strong and early induction of IFNβ, peaking at 3 hours post infection, while in a 129x1/SvJ background the induction was more gradual peaking at 6 hours post infection (Figure 18). To investigate this, the basal level of sensors was examined by qPCR using plasmids encoding for murine MDA-5 and RIG-I as standards. The levels of RIG-I and MDA-5 expression are both higher in macrophages derived from mice of a mixed ICR background than those in the 129x1/SvJ background. Within each background however the expression of RIG-I is greater than MDA-5, and the ablation of one of the sensors does not lead to a change in the expression of the other (Figure 17A). In both backgrounds this difference between RIG-I and MDA-5 expression is the same with about 2 copies of RIG-I to every copy of MDA-5 message (Figure 17B).

Figure 17. Basal levels of RIG-I and MDA-5 expression in macrophages. Total RNA from uninfected bone marrow derived macrophages of different mice was harvested and copies of RIG-I and MDA-5 mRNA calculated from qPCR using a plasmid standard. (A) Copy numbers were normalized to β-Actin Ct values. (B) Ratio of relative RIG-I to MDA-5 expression was calculated. n=5 biological replicates, qPCR performed in triplicate. Error bars show standard deviation between samples.

Comparing each sensor knockout to its matched wild type control, the absence of

MDA-5 shows the loss of induction of IFNβ as well as IFNα is lost. In the absence of

RIG-I the induction of IFNβ is greatly reduced while the induction of IFNα is not, but is actually induced to a greater extent than in wild type cells (Figure 18 A and B). Enzyme linked immunosorbent assays for IFNα confirm this difference in the expression of IFNα is seen at the protein level as well as the RNA level (Figure 18 C). The effects of these differences in type I IFN expression on viral replication are unknown since growth curves show limited viral replication and only a decrease in infectious particles, which occurs at the same rate both RIG-I-/- macrophages and MDA-5-/- macrophages and their wild type controls (Figure 18 D). This indicates that EMCV does not productively infect macrophages, and during these non-productive infections MDA-5 is crucial for IFNα and

IFNβ expression. RIG-I, although not essential, is important for a robust IFNβ response, but not for an IFNα response.

Figure 18. Role of MDA-5 and RIG-I in the induction of type I IFN in macrophages infected with EMCV. Bone marrow from knockout MDA- 5 or RIG-I mice and their matched wild type controls were harvested and cultured in LCM conditioned media and infected with EMCV at an MOI=1. Total RNA was harvested at various times posit infection and expression of (A) IFNβ and (B) IFNα examined by qPCR. (C) Supernatants were used to measure IFNα protein levels by ELISA. (D) Viral titers during infection were determined by plaque assay. Representative data from one experiment of three independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

RIG-I and MDA-5 contribute similarly to sensing EMCV infection of MEFs

The role of MDA-5 and RIG-I in the sensing of EMCV infection was examined in the context of productive replication in MEFs. In this cell type EMCV was found to replicate well (Figure 19A). Type I IFN was induced in response to infection with kinetics similar to those seen in HeLa cells. Peak expression of type I IFNs was seen at

24 hours post infection with significant levels only appearing at 9 hours post infection.

This resulted in a late induction of expression of the ISG OAS. The absence of both

MDA-5 and RIG-I resulted in a marked decrease in the amounts of IFNβ, IFNα and OAS

(Figure 19 B-D). These decreases in antiviral response had no effect on viral replication since notable amounts of type I IFN in wild type cells were not produced until after the peak of viral replication. However, the magnitude of the decrease in type I IFNs in the absence of MDA-5 and RIG-I indicate that both proteins are involved in the sensing of

EMCV infection in MEFs.

Figure 19. Role of MDA-5 and RIG-I in the induction of an antiviral state in MEFs infected with EMCV. MDA-5 and RIG-I deficient MEFs and their matched wild type controls were infected with EMCV at an MOI=1. At various times post infection total RNA and total virus was harvested. (A) Viral titers during infection were determined by plaque assay and (B) IFNβ (C) IFNa and (D) OAS expression determined by qPCR. Representative data from one experiment of three independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

Sensing of EMCV in MEFs in replication dependent

Because the roles of MDA-5 and RIG-I in sensing EMCV infection were different in MEFs, where viral replication was robust, than in macrophages, where viral replication was very poor, the role of replication in viral sensing was examined in MEFs. MDA-5-/-,

RIG-I-/- and matched WT control MEFs were treated with one pfu or UV inactivated pfu per cell. Total virus was harvested from both conditions 24 hours post treatment and plaque assays were performed to ensure that no viral replication occurred in MEFs treated with UV inactivated virus. Replicating virus was sensed in wild type, MDA-5-/- and RIG-I-/- MEFs to lead to the induction of IFNβ and IFNα 24 hours post treatment.

However treatment with UV inactivated EMCV resulted in no notable induction of IFNβ or IFNα in either wild type, MDA-5-/- or RIG-I-/- cells indicating that sensing in MEFs is replication dependent (Figure 20).

Figure 20. Role of EMCV replication on the induction of an MDA-5 dependent antiviral state in MEFs. (A,B) MDA-5 and (C,D) RIG-I deficient MEFs and their matched WT controls were treated with EMCV or UV inactivated EMCV at concentration of 1pfu or UV inactivated pfu per cell. Total RNA was collected at 24hrs p.i. from mock (M) and EMCV (24C) treated cells as controls and at varying times post UV inactivated EMCV treatment and expression of (A,C) IFNβ and (B,D) IFNa determined by qPCR. Representative data from one experiment of two independent experiments, qCR performed in triplicate. Error bars show standard deviation within experiment.

In summary we have shown that EMCV is sensed by both MDA-5 and RIG-I in murine cells. In murine macrophages, MDA-5 is essential for IFNβ and IFNα expression while RIG-I is important for maximal levels of IFNβ expression but not IFNα expression. In MEFs Both MDA-5 and RIG-I are needed for a robust IFNβ and IFNα response. This response in MEFs occurs late in infection, after the peak of viral growth and therefore does not affect viral titers. In macrophages the expression of if type I IFNs occurs very early after infection, but does not have an effect on viral replication since replication in macrophages is poor to begin with. In human HeLa cells EMCV replicates well similar to MEFs, but MDA-5 does not contribute greatly to the expression of IFNβ or the induction of ISGs.

Chapter 4: Sensing of Coxsackievirus

RIG-I is important and MDA-5 is essential for the sensing of CVB3 infection of macrophages

Most picornaviruses belong to the Enterovirus genus, yet most of the work done on sensing of picornaviruses has been done using EMCV, one of two viruses within the

Cardiovirus genus. Experiments showing the importance of MDA-5 in sensing EMCV have lead to the investigation of MDA-5 in the sensing of other picornaviruses.

However, in the previous chapter it was show that RIG-I also senses the infections with

EMCV. Therefore to investigate the role of RIG-I in sensing of picornaviruses for the

Enterovirus genus, the human Coxsackievirus B3 (CVB3), which can also infect mice, was used to infect cells derived from RIG-I and MDA-5 knockout mice.

Infection of murine cells at a low MOI did not result in the induction of type I

IFN, ISGs or cytopathic effect probably due to a different infectivity of CVB3 in murine cells than cells derived from its natural host, humans. Therefore, infections of murine cells were done at high MOIs of 10-50. At an MOI of 10 CVB3 infection of bone marrow derived macrophages elicited a Type I IFN response, but of a greatly reduced magnitude than that elicited by EMCV. Similar to the infection of macrophages by

EMCV the IFNβ response to CVB3 infection was very early in the mixed ICR background, peaking at 3 hours post infection and more gradual and delayed in a

129x1/SvJ background not peaking until 12hours post infection (Figure 21A). In the

129x1/SvJ background CVB3 infection resulted in the induction of both IFNβ and IFNα in wild type macrophages which is lost when MDA-5 is knocked out. In an ICR mixed background there is only the induction of IFNβ in wild type macrophages which is

71 reduced early in the absence of RIG-I early in infection, at the peak of IFNβ expression.

Although IFNα was not induced in WT macrophages of a mixed ICR background, expression was induced in the absence of RIG-I (Figure 21 A and B). A similar negative regulation of IFNα by RIG-I was seen previously in EMCV infected macrophages

(Figure 18 B). This difference between the induction IFNα message is also seen in IFNα protein levels (Figure 21C). The effects of these observed differences in Type I IFN expression does not result in any differences in viral replication as CVB3 does not replicate in macrophages (Figure 21D). Therefore, similar to EMCV, CVB3 infection of macrophages results in a non-productive infection that is sensed by MDA-5 to induce a

IFNβ and IFNα response. RIG-I contributes to positively to the induction of IFNβ but negatively to the induction of IFNα.

Figure 21. Role of MDA-5 and RIG-I in the induction of type I IFN in macrophages infected with CVB3. Bone marrow from knock out MDA-5 or RIG-I mice and their matched wild type controls were harvested and cultured in LCM conditioned media and infected with infected with CVB3 at an MOI=10. Total RNA was harvested at various times posit infection and expression of (A) IFNβ and (B) IFNα examined by qPCR. (C) Supernatants were used to measure IFNα by ELISA. (D) Viral titers during infection were determined by plaque assay. Representative data from one experiment of three independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

RIG-I is essential for the production type I IFN during CVB3 infection of MEFs

CVB3 is able to replicate more robustly in MEFs than macrophages. At an MOI of 1 the induction of IFNβ can be detected in wild type cells of the mixed ICR background, although magnitude of IFNβ induction is still notably lower than that of

EMCV infection of these cells. Increasing the MOI of infection results in an increase in induction of IFNβ although at a very high MOI of 50 this induction is not seen until 24 hours post infection, similar to the induction of IFNα. The increase of type I IFN after infection also results in the increase of ISGs shown by increases in ISG49 mRNA and

ISG56 protein, and to a lesser extent OAS mRNA. In the absence of RIG-I the increase in type I IFN, ISG49, ISG56 and the small increase in OAS are completely ablated, indicating that RIG-I is essential for the sensing of CVB3 infection and the induction of type I IFNs (Figure 22A-E). However, it appears that the loss of Type I IFN induction does not correlate with a loss of an effective antiviral state since viral replication appears to decrease in the absence of RIG-I, indicating that the type I IFN response is not protective and that RIG-I may be inhibiting a response that is protective (Figure 22F).

Figure 22.Role or RIG-I in the induction of an antiviral state in MEFs infected with CVB3. RIG-I deficient MEFs and their matched wild type controls were infected with CVB3 at an MOI of (A-B) 1, (C-D) 10, (E-F) 50. At various points post infection total RNA, virus and protein lysates were collected. (A,C,E) RNA was used to measure gene expression for indicated genes by qPCR (B,D) Lysates were analyzed by western blot for ISG56 and actin (control). (F) Virus titers were measured by plaque assay. Representative data from one experiment of three independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

MDA-5 has a type I IFN independent role in inhibiting CVB3 replication in MEFs

Infection of 129x1/SvJ MEFs with CVB3 at an MOI of 50 resulted in a very modest increase of IFNβ and IFNα peaking at 6 hours post infection. In the absence of

MDA-5 an increase in IFNβ and IFNα was still observed and peaked at 24 hours post infection indicating that MDA-5 is not essential for the induction of type I IFNs (Figure

23B, C). The increase in Type I IFN message did not however lead to an increase in expression of the ISG OAS in either the wild type or the MDA-5 deficient cells (Figure

23D). Interestingly, despite the absence of an antiviral state detectable by OAS induction and the absence of a notable difference in type I IFN production between wild type and

MDA-5 deficient MEFs, there is a difference in the replication of CVB3 in the two cell lines. CVB3 replicates to titers ten fold higher in the absence of MDA-5 (Figure 23A).

This indicates that MDA-5, while not essential for the induction of type I IFN is important for the induction of some antiviral mechanism.

Figure 23. Role of MDA-5 in the induction of an antiviral state in MEFs infected with CVB3. MDA-5 deficient MEFs and their matched wild type controls were infected with EMCV at an MOI=50. At various times post infection total RNA and total virus was harvested. (A) Viral titers during infection were determined by plaque assay and (B) IFNβ (C) IFNa and (D) OAS expression determined by qPCR. Representative data from one experiment of three independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

Additionally, treatment of wild type and MDA-5 deficient MEFs with UV inactivated CVB3 showed that the modest amounts of type I IFN induced is dependent on viral replication. (Figure 24A, B) However, in the absence of viral replication and of a type I IFN response a modest induction of the ISG OAS was detected, but only in MDA-

5 knockout MEFs. (Figure 24C) This may indicate the ability for MDA-5 to inhibit the sensing of non-replicating virus through a type I IFN independent manner. UV exposure causes adjacent pyrimidines to form dimers through covalent bonds. This damage prevents or stalls translations by ribosomes preventing the expression of viral proteins and replication of genomes (167). Therefore the viral ligands being sensed are either viral capsid protein or damaged genomic RNA.

Figure 24. Role of CVB3 replication on the induction of an antiviral state in MDA-5 and matched wild type MEFs. Wild type and MDA-5 deficient MEFs were treated with UV inactivated CVB3 at concentration of 50 UV inactivated pfu per cell, EMCV at an MOI=1 or mock treated. Total RNA was collected at 24hrs p.i. from mock (M) and EMCV (24C) treated cells as controls and at varying times post UV inactivated CVB3 treatment and expression of (A) IFNβ (B) IFNa and (C) OAS determined by qPCR. Representative data from one experiment of two independent experiments, qPCR performed in triplicate. Error bars show standard deviation within experiment.

To ensure that the antiviral state produced by MDA-5 was independent of type I

IFN, MEFs unable to respond to type I IFNs due to the knockout of the interferon alpha receptor 1 (IFNAR-/-) were obtained in a 129x1/SvJ background. These MEFs were infected along with wild type and MDA-5-/- MEFs and viral titers measured at varying times post infection. As seen previously CVB3 titers increased modestly in wild type

MEFs. In the absence of MDA-5 viral growth was about ten fold greater. This was seen both at a high MOI of 50 where the level of some type I IFN induction could be detected as well as at a lower MOI of 1 where no type I IFN induction was detectable by qPCR.

In the absence of IFNAR viral growth was slightly greater than it was in wild type MEFs, but not as great at in MDA-5-/- MEFs (Figure 25). This further indicates that sensing through MDA-5 produces some antiviral response that is independent of type I IFNs.

Figure 25. Effect of MDA-5 and IFNAR on CVB3 replication. MDA-5-/- and IFNAR-/- MEFs were infected with CVB3 at (A) MOI=1 or (B) MOI=50. At various times post infection samples were harvested for total virus and titer determined by plaque assay. Representative data from one experiment of two independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

It was previously shown in our lab that the Cardiovirus EMCV is sensitive to type

I IFN, being completely unable to replicate in cells pre-treated with IFNα. The

Enterovirus poliovirus however, was able to replicate in cells pre-treated with IFNα and described as an interferon resistant virus (101). To determine if type I IFNs could have an effect on CVB3 replication, HeLa cells mock or pre-treated with IFNα at the very high dose of 1000 units/mL to establish an antiviral state cells were infected with CVB3.

IFNα pre-treatment reduced the levels of CVB3 growth in HeLa cells similar to the reduction of poliovirus growth in IFNα pretreated cells indicating that CVB3 is an interferon resistant virus (Figure 26A). The poliovirus 2A protease was shown to be responsible for the ability of poliovirus to evade type I IFN pre-treatment. This ability was dependent on the presence of a tyrosine at residue 88 of 2Apro which is though to be involved in the discernment of protease substrates (101). An alignment of the poliovirus and coxsackievirus 2A proteases shows the presence of a tyrosine at coxsackie residue which corresponds to the tyrosine at position 88 of poliovirus 2A, suggesting that the ability of CVB3 to replicate in the presence of type I IFN is similar to that of poliovirus: it is able to cleave some ISGs which are inhibitory to its replication. However, CVB3 must not be able to inhibit all inhibitory ISGs since some decrease in viral titer is observed in IFNα pretreated cells.

Figure 26. Resistance of CVB3 growth to IFNa pretreatment. (A) HeLa cells were treated with or without 1000U/mL of IFNa for 16 hours and infected with CVB3 at an MOI=1. At various points post infection total virus was harvested and titers determined by plaque assay. (B) Alignment of poliovirus and coxsackie B3 virus 2A proteases. The asterisk marks the conserved tyrosine at amino acid 88. Representative data from one experiment of two independent experiments. Plaque assays performed in duplicate.

MDA-5 sensing of CVB3 infection of MEFs is independent of type III IFN

Type III IFNs have been shown to be under the control of the transcription factors

NFκB, AP-1 and IRFs similar to type I IFNs (108). Therefore type III IFNs are frequently induced in response to the same triggers that lead to type I IFN , however, differences in the expression of type I IFN and type III IFN have been noted.

Macrophages infected with HSV have been shown to induce type I IFN but not type III

IFN and epithelial cells infected with influenza A produce higher levels of type III IFNs than type I IFN (160). These differences are likely due to difference in the amount of dependence each IFN has for the various transcription factors. Type I IFNs are strongly dependent on IRFs while type III IFNs have been found to be more dependent on NFκB

(50, 150). Additionally there are some differences in the signaling of type III IFNs from the signaling of type I IFNs since type I IFNs produce a positive feedback loop in which they up-regulate the expression of more type I and type III IFN. Type III IFN signaling however does not result in an increase in type I or type III IFN expression (5). We therefore hypothesized that type III IFNs would be a likely candidate for a type I IFN independent inhibition of CVB3. Although no notable differences in induction of type I

IFNs was found between wild type and MDA-5 deficient MEFs, the loss of MDA-5 may result in a loss of type III IFN expression. Additionally, since type III IFNs use a different receptor for signaling they may preferentially activate a different assortment of

STATs to induce a different array of ISGs than type I IFNs. It could be possible that these ISGs would be once which CVB3 does not cleave to allow replication in IFNα activated cells.

Therefore total RNA was collected from wild type, MDA-5 and IFNAR MEFs infected with CVB3 at an MOI of 1 and 50. Induction of the type III IFNs IFNλ2 and

IFNλ3 detected by the same primer set was examined by qPCR. If type III IFNs were responsible for the antiviral effects of MDA-5 we expected the induction of IFNλ2/3 to be lower in MDA-5 deficient cells but instead found that IFNλ expression was not induced at all in wild type MEFs but was induced in the absence of MDA-5. The induction of IFNλ2/3 in MDA-5-/- MEFs was even detectable in MEFs infected at the low

MOI of 1, in which no induction of type I IFNs was found (Figure 27). This therefore shows that inhibitory effect of MDA-5 on CVB3 replication is independent of type III

IFN as well as type I IFN. It also suggests that MDA-5 negatively regulates type III IFN expression.

Figure 27. Role if MDA-5 and IFNAR on type III IFN expression in MEFs infected with CVB3. Wild type, MDA-5 and IFNAR deficient MEFs were infected with CVB3 at (A) MOI=1 and (B) MOI=50. Total RNA was harvested and qPCR for murine IFNλ2/3 performed. Representative data from one experiment of three independent experiments, qPCR performed in triplicate. Error bars show standard deviation within experiment.

In summary, similar to EMCV, CVB3 replicated poorly in murine macrophages, but was able to be sensed. MDA-5 was essential for this sensing and RIG-I contributed to the full expression of IFNβ but not IFNα, which RIG-I expression had a negative effect on. In MEFs CVB3 was able to replicate, MDA-5 inhibited replication but did not contribute to type I IFN expression. Meanwhile RIG-I enhanced viral replication but sensed infection to produce a type I IFN response. The MDA-5 inhibition of CVB3 replication was independent of type III IFN as well as type I IFN. MDA-5 was also found to contribute negatively to the expression of type III IFN.

Chapter 5: Sensing of Poliovirus

Poliovirus infection is sensed by MDA-5

Poliovirus is the best-studied and model picornavirus because it is the etiological agent of paralytic poliomyelitis – one of the most significant diseases caused by picornaviruses. However, little has been done to examine the sensing of this picornavirus. Two studies crossed transgenic mice expressing the human poliovirus receptor (PVR Tg) to mice lacking various sensors and shown that in vivo, TLR3 is important for the sensing of poliovirus in mice since there is a consistent increase in mortality, viral load and decrease in type I IFNs and ISGs in certain tissues after infection of PVR Tg/TRIF-/- mice (1, 110). However in vivo infections of PVR Tg/MDA-5-/-, PVR

Tg/RIG-I-/-, and PVR-Tg/IPS-/- mice do exhibit an increased mortality under various conditions. The importance of these sensors in poliovirus infections may be obscured by the complexity of various cell types involved in in vivo experiments, especially during unnatural routes of infection such as intraperitoneal and intravenous injections that override natural immune barriers and provide immediate access to many tissues.

Therefore the sensing of poliovirus infection was studied in vitro using human

HeLa cells knocked down for the sensor MDA-5 (Figure 12). Infection of control and

MDA-5 knockdown cells with poliovirus showed a great decrease in the induction of

IFNβ in cells with reduced levels of MDA-5. However, no change was seen in viral titers since the induction of IFNβ occurred late in infection, at 8-12 hours post infection and viral replication peaked by 6 hours post infection (Figure 28).

Figure 28. Role of MDA-5 on IFNβ expression in HeLa cells infected with poliovirus. Control and MDA-5 knockdown cell lines were infected with poliovirus at an MOI of 0.01. (A) Total virus was collected at various times post infection and titer determined by plaque assay. (B) RNA was harvested and expression of IFNβ determined by qPCR. Representative data from one experiment of two independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

This is quite different from the response seen for EMCV infection of HeLa cells in which there was little change in IFNβ and ISG expression in the absence of MDA-5

(Figure 15). One of the differences between EMCV and poliovirus is the 2A viral protein, of which poliovirus 2A has protease activity and EMCV 2A does not (182). The protease activity of poliovirus is important for cleavage of the viral polyprotein into mature viral proteins as well as the cleavage of various host proteins. A mutation of a tyrosine at residue 88 of poliovirus 2A to a leucine changes the substrate binding loop in a manner which still allows for cis cleavage of the polyprotein but not trans cleavage of host proteins such as the translation factor eIF4G(175). To determine if the 2A protease plays a role in the difference in dependence on MDA-5 for sensing of poliovirus and

EMCV infections, control and MDA-5 knock down cell lines were infected with a poliovirus mutant containing the mutation of tyrosine at residue 88 of 2A to leucine (PV-

Y88L). Similar to infections with poliovirus, the PV-Y88L mutant induced IFNβ, but after the peak of viral replication. Therefore the expression of IFNβ had no effect on viral titers. However, unlike infections with poliovirus the PV-Y88L mutant induced an

IFNβ response in MDA-5 knockdown cell lines to levels greater than those in the control cell line (Figure 29), suggesting that the poliovirus 2A protease has a positive effect on the ability for MDA-5 to sense poliovirus.

Figure 29. Role of MDA-5 on IFNβ expression in HeLa cells infected with PV-Y88L. Control and MDA-5 knockdown cell lines were infected with the poliovirus mutant PV-Y88L at an MOI of 0.01. (A) Total virus was collected at various times post infection and titer determined by plaque assay. (B) RNA was harvested and expression of IFNβ determined by qPCR. Representative data from one experiment of two independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

Previously our lab had shown that the 2A protease activity of poliovirus is necessary but not sufficient to allow for viral replication in IFNα pretreated cells.

Absence of the protease activity in the PV-Y88L mutant prevented the replication of poliovirus in IFNα pretreated cells but addition of the 2A protease into EMCV, which is unable to replicate in IFNα pretreated cells only allowed the EMCV mutant (EP4) to replicate in cells pretreated with a low concentration of IFNα but was not sufficient to allow for EMCV replication in cells pretreated with high concentration of IFNα. To see if the protease activity of poliovirus 2A is capable of inducing the ability for EMCV to be sensed by MDA-5, control and MDA-5 knockdown cells were infected with the EMCV mutant EP4 which contains the poliovirus 2A protease within the EMCV genome. EP4 was able to induce an IFNβ response in both control and MDA-5 knockdown cells with little difference in the amount of IFNβ induced, similar to EMCV (Figure 30 A, B).

However a clone of EP4, EP4C4 (which was originally isolated based on its greater ability to replicate in cells pretreated with high amounts of IFNα) did show a greater dependence on MDA-5 for sensing indicated by a reduced amount of IFNβ produced in the absence of MDA-5 (Figure 30 C, D). This indicates that additional changes were required for the EMCV to be sensed by MDA-5 in HeLa cells. Sequencing of EP4C4 has shown 3 residue changes from EP4 in the L, 2B and 3C proteins. Interestingly 3C is a viral protease that has been suggested to be involved in evasion of viral sensing for some picornaviruses by cleavage of RIG-I, IPS or TRIF (11, 102, 114, 124).

Figure 30. Role of MDA-5 on IFNβ expression in HeLa cells infected with EMCV mutants. Control and MDA-5 knockdown cell lines were infected with (A-B) EP4 or (C-D) EP4C4. (A, C) Total virus was collected at various times post infection and titer determined by plaque assay. (B, D) RNA was harvested at various times post infection and expression of IFNβ determined by qPCR. Representative data from one experiment of two independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

It is possible that poliovirus and EMCV can be sensed through multiple pathways but certain pathways are evaded by the ability of the viral proteases to cleave proteins involved in each signaling pathway. The inhibition of sensing through one pathway then leads to sensing through another pathway, which in some cases may be more or less robust in activating IFNβ transcription. To investigate alternate sensing pathways a

HeLa cells were knocked down for RIG-I, the other cytoplasmic RNA sensor, and IPS-1 down stream adaptor molecule for both RIG-I and MDA-5. The RIG-I knockdown would allow for determination of the role of sensing through an alternate cytoplasmic receptor pathway and the knockdown of IPS would allow for the determination of the role of non-cytoplasmic receptor sensing, i.e. TLR sensing. The knock down of RIG-I resulted in non-detectable basal levels of RIG-I by western blot. The induction of RIG-I expression by IFNα or pI:C treatment however increased levels of RIG-I in the RIG-I knockdown cell line to levels which are detectable by western blot. These levels however were greatly reduced in comparison to those in IFNα and pI:C treated control knockdown cells (Figure 31A). Two variants of IPS-1 are produced in human cells giving rise to proteins detected at 80kDa and 60kDa. The IPS-1 knockdown cell line reduced the level of both these forms to undetectable levels by western blot analysis

(Figure 31B). These cell lines will be used in future studies to investigate the ability for poliovirus and EMCV to activate other sensing pathways.

Figure 31. Knockdown of RIG-I and IPS-1 in HeLa cells. (A) Control or RIG-I knockdown HeLa cell lines were mock treated or treated with 1000U/mL of IFNa or 20ug/mL of pI:C for 16hrs and cell extracts were analyzed by western blot for RIG-I. (B) Extracts of control and IPS-1 knockdown HeLa cell lines were analyzed by western blot for IPS-1. Arrows indicate 80KDa and 60KDa forms of IPS-1. Representative data from one experiment of two independent experiments.

Because poliovirus is a cytopathic virus and IFNβ responses were induced late in infection, after the peak of viral replication when cells appeared to be in late stages of apoptosis, the contributions of apoptosis to IFNβ induction was tested by the induction of apoptosis by staurosporine treatment. Staurosporine is a kinase inhibitor which induces apoptosis. Western blot analysis for poly ADP ribose polymerase (PARP), a substrate of caspases was performed on lysates at various times post treatment to ensure the occurrence of apoptosis and qPCR done to determine the induction of IFNβ. Caspase activation was confirmed but no amplification of IFNβ was seen by qPCR although amplification of the standard curve and the internal control, GAPDH, were observed

(Data not shown). Additionally the activation of caspases, the executioners of apoptosis, were inhibited during poliovirus infections using the pan-caspase inhibitor QVD. In cells treated with the solvent DMSO caspase activation, indicated by PARP cleavage was detected at 9 hours after infection, while no cleavage was detected in QVD treated cells

(Figure 32A and B). The inhibition of caspase activation however had no effect on viral growth and minimal effect on IFNβ expression (Figure 32 C and D). This indicates that it is not the process of apoptosis but viral infection that is inducing IFNβ expression.

Figure 32. Effect of apoptosis on poliovirus induced IFNβ expression. HeLa cells were mock treated with QVD or DMSO solvent as a control and infected with poliovirus at an MOI=1. At various times post infection (A-B) protein lysates (C) total virus and (D) RNA were collected. Apoptosis during infection in the presence of (A) DMSO and (B) QVD was detected by western blot analysis for PARP cleavage. (C) Total virus titers were determined by plaque assay. (D) Expression of IFNb was determined by qPCR. Representative data from one experiment of two independent experiments. Plaque assays performed in duplicate and qPCR performed in triplicate. Error bars show standard deviation within experiment.

Chapter 6: Discussion

The cytoplasmic receptors RIG-I and MDA-5 were initially thought to sense distinct sets of viruses. It has been commonly cited that RIG-I recognizes infections by most families of RNA viruses while MDA-5 senses viruses of the Picornaviridae family

(2, 71, 118, 147, 165). This distinction was made from a set of experiments infecting

MEFs, BMDCs and mice with an array of different RNA viruses and looking for crude differences in type I IFN expression and animal survival as indications of sensing. While theses initial studies were very informative, more recent work has begun to elaborate on the generalized conclusions made from them. One conclusion made was that RIG-I was the sensor for flaviviruses based on experiments using Japanese Encephalitis virus (59).

Although the flaviviruses Japanese Encephalitis virus, Hepatitis C virus, West Nile virus and Dengue virus are all sensed by RIG-I, Dengue virus and West Nile virus have also been shown to be sensed by MDA-5. It was therefore concluded that genomes of these viruses contained structures that were ligands for both MDA-5 and RIG-I . We therefore were interested in further developing the understanding of sensing using viruses of the picornavirus family and found that the discernment between sensors is dependent on more than just structural features of the viral genome.

RIG-I senses picornaviruses

It is already well known that picornaviruses are sensed by MDA-5 but few studies have looked at the sensing of picornaviruses by RIG-I. It has been suggested that sensing does not occur through RIG-I due to the lack of structural features such as 5’ tri-

98 phosphates as well as the ability for picornaviruses to evade RIG-I signaling through the viral 3C protease. The 3C protease of Enterovirus 71 has been shown to inhibit IFNβ promoter activity stimulated by Sendai virus or constitutively active RIG-I, independent of its protease function . Poliovirus, echovirus, rhinovirus 16 and EMCV however have been shown to cleave RIG-I during infection using the 3C protease . In MDA-5 deficient

MEFs the IFNβ promoter is not activated upon EMCV infection, however activity can be detected in infection if RIG-I is over expressed, showing that RIG-I is capable of sensing

EMCV infection.

Through our use of RIG-I knock out MEFs and macrophages we have shown that not only can RIG-I sense picornavirus infections but also that RIG-I does sense picornavirus infection. CVB3 infection of macrophages deficient in RIG-I resulted in a decreased level of IFNβ expression while infection of RIG-I deficient MEFs resulted in a complete loss of IFNβ expression. Therefore, for CVB3 RIG-I is necessary for the maximal induction of IFNβ in macrophages and is essential for the production of type I

IFNs in MEFs (Figure 21 and 22). EMCV infections of RIG-I deficient macrophages and

MEFs resulted in decreased levels of IFNβ expression, and in MEFs a decrease in IFNα expression as well. This indicates that during EMCV infections RIG-I is necessary for the maximal production of IFNβ in macrophages and IFNβ and IFNα in MEFs (Figure

18 and 19).

No previous studies had been done to examine the role of RIG-I in CVB3 infection but for EMCV it has been shown that the absence of RIG-I does not produce a significant effect on Type I IFN expression in MEFs and BMDC . The differences in our data and published data within leukocytes could be in the difference in leukocytes

99 examined since our studies used BMDM as opposed to BMDC. Additionally our studies in leukocytes showed differences the induction of type I IFNs occurred early after infection peaking at three to six hours post infection in RIG-I-/- and matched wild type control cells (Figure 18 and 21) while other studies only looked at one time point at 24 hours after infection. At this point we found that Type I IFN expression was low and levels between RIG-I-/- and wild type cells were comparable.

The differences between published data and our data in MEFs, where we found

IFN induction to be late with peak expression at 24 hours after infection, are likely due to background differences. Our RIG-I-/- MEFs were derived from viable mice obtained by the crossing of heterozygous RIG-I mice in a mixed 129x1/SvJ and C57BL/6 background with outbred ICR mice. While other studies used MEFs in a 129Sv X C57BL/6 background in which both RIG-I-/- and MDA-5-/- mice were lethal at embryonic day 12.5

. Although this background allows for the direct comparison of the roles of MDA-5 and

RIG-I in infection it cannot be assumed that their roles would be as clearly defined in a viable mouse.

This new finding of RIG-I contributing to picornavirus sensing leads to the conclusion that picornavirus infections produce an RNA ligand recognizable by RIG-I.

Because the ends of RNA molecules recognized by RIG are thought to be important and the ends of picornavirus genomes are obstructed due to the requirement of a protein primer, VPg, to initiate viral RNA polymerization how RIG-I senses picornavirus RNA is an interesting question (Figure 4). One possible explanation is that RIG-I is sensing the viral RNAs that do not have VPg attached. It has been shown that once viral RNA has entered a cell VPg is removed by “unlinkase” before translation leaving a 5’

100 monophosphate end (4, 44, 104). This unlinked RNA with an unobstructed 5’ end may now be recognized by RIG-I, or more likely the unlinked RNA bound to a negative sense strand of viral RNA (RF RNA) since it has been shown that RIG-I can recognize double stranded RNA with one strand monophosphorylated(145). To investigate this isolated RF

RNA could be transfected into cells and sensing measured. Additionally localization studies should indicate the presence of RIG-I at replication complexes where RF RNA is found.

Because the levels of RF RNA with in a cell are low it is also possible that RIG-I is sensing a more abundant RNA structure such as the IRES. These highly structured

RNA domains contain many double stranded regions which may be ligands for RIG-I since RIG-I can also bind double stranded RNA. To investigate this gel shift assays can be performed with RIG-I and IRES structures to show binding.

Coxsackievirus B3 replication is inhibited by an IFN independent mechanism

The absence of MDA-5 or RIG-I in macrophages infected with EMCV or CVB3 results in decreased antiviral state but does not result in an increase in viral growth since both EMCV and CVB3 replicate poorly in macrophages (Figure 18 and 21).

Both viruses do however replicate well in MEFs, but the absence of RIG-I or MDA-5 in

MEFs still has no effect on viral growth of EMCV even though it has an effect on the expression of type I IFNs and ISGs since the induction of theses genes does not occur until after the peak of total viral replication (Figure 19).

The absence of RIG-I during CVB3 infection however appeared to inhibit the formation of an antiviral state by preventing the induction of type I IFN as well as ISGs

101 such as ISG 49, ISG 56 and OAS. Interestingly this did not result in an increase in viral replication compared to replication in wild type cells but in a decrease in viral replication

(Figure 22 and 33). The absence of MDA-5 during CVB3 infection showed no appreciable difference in the induction of type I IFNs, which was low enough to not result in an increase in the expression of the ISG OAS. However, MDA-5 deficiency did result in an increase in viral titers (Figure 23 and 33). This inhibitory effect of MDA-5 on CVB3 replication was apparent not only at the high MOI of 50 which was needed to see an induction of type I IFN, but also at a lower MOI of 1 at which no increase in type I

IFNs could be detected (Figure 24). The increase in viral titer in the absence of MDA-5 despite the lack of decrease in type I IFN, along with the decrease in viral titers in the absence of RIG-I which does result in a decrease in type I IFN, indicates that there must be factors independent of type I IFN expression that are inhibiting CVB3 replication.

This supports in vivo work where infections of MDA-5-/- mice show no significant increase in IFNα serum levels or IFNβ or OAS expression in tissues, yet a transient early increase in CVB3 titers, which lead to increased tissue damage and disease is observed

(48). Infection of IFNAR-/- MEFs confirmed this type I IFN independent inhibition of

CVB3 replication since viral titers in these cells were lower than those in MDA-5 MEFs at both an MOI of 1 and an MOI of 50 (Figure 24). Titers in IFNAR-/- MEFs were slightly higher than those of wild type MEFs indicating that type I IFN does have a slight role in inhibiting CVB3 replication but is not fully responsible for the MDA-5 dependent inhibition of CVB3 replication. Additionally CVB3 was found to be able to replicate in the presence of an antiviral state induced by type I IFN, further supporting the idea that the MDA-5 response is independent of type I IFN (Figure 25 and 33).

102

Type III IFNs have been found to also be induced in response to viral infections and to be capable of producing an antiviral state after signaling through their receptors.

Quantitative PCR for IFNλ2/3 expression also did not show a decreased induction in the absence of MDA-5. Interferon λ2/3 expression was instead induced in the absence of

MDA-5. The kinetics of this induction was late, similar to that of type I IFNs, beginning at nine hours post infection, and therefore likely to not have a causative role in the increase in viral replication in MDA-5 deficient MEFs, which begins earlier in infection

(Figure 26 and 33).

103

Figure 33. Coxsackievirus infection of MEFs. An infecting virus replicates its genome by first producing a complementary negative strand of RNA followed by multiple positive sense strands of RNA off of the negative strand. New genomic RNAs are used to form new infectious particles. The sensors MDA-5 and RIG-I bind some form of viral RNA as the virus replicates. Binding of RNA to RIG-I results in the expression of type I IFNs and ISGs, but not the inhibition of viral replication. The presence of RIG-I actually results in an increase in viral replication. Binding of RNA to MDA-5 does not result in the expression of type I IFNs but results in a decrease in viral replication and the absence of a type III IFN response.

104

It is known that viral sensing results in the induction of many cytokines under the control of NFκB in addition to IFNs. One of these other cytokines may be contributing to decreases in CVB3 replication.

However a more appealing explanation is that MDA-5 is signaling through IPS-1 on the peroxisome. It has been found that an IFN independent antiviral state may be induced by RIG-I through signaling using peroxisomal IPS-1 rather than mitochondrial

IPS-1 (21). This signaling was found to induce expression of ISGs directly through the activation of IRF1 and to do so before the induction of type I IFNs. If MDA-5 can also signal through peroxisomal IPS it could inducing an effective early antiviral response that results in a decrease in CVB3 titers. Since differences in viral titers are seen early in infection, this seems like a plausible explanation. The expression of ISGs under control of the IRF1 transcription factor can be examined to determine this. It would also be expected that these ISGs would be the ones which are responsible for the slight decrease in CVB3 replication in HeLa cells pretreated with IFNα. Therefore if IRF1 dependent

ISGs are up-regulated in wildtype but not MDA-5-/- MEFs over expression of these ISGs should be able to reduce viral replication.

If RIG-I signaling in this context olny occurs through mitochondrial IPS-1 this could also explain why the induction of type I IFNs and ISGs by RIG-I does not limit viral replication but promotes it. The inability for type I IFN signaling to have an effect on viral replication is due to the lateness in the response after viral replication is complete. Additionally the sensing of viral RNA by RIG-I may compete with the ability for MDA-5 to bind and sense viral RNA. Therefore in the presence of RIG-I the early, effective, peroxisomal response through MDA-5 is dampened and viral replication is

105 greater than in the absence of RIG-I. Immunofluorescence staining of RIG-I, mitochondrial and peroxisomal markers after infection would verify this hypothesis.

Involvement of RIG-I and MDA-5 in sensing varies based on host, cell type, and genetic background

Although RIG-I was not essential for the sensing of EMCV in either macrophages or MEFs, MDA-5 was found to be essential for the production of type I IFNs in macrophages but not in MEFs where there was still an induction of IFNβ and IFNα message, albeit greatly reduced (Figure 18 and 19). MDA-5 was also found to be essential in the induction of IFNβ and IFNα expression in response to CVB3 infection in macrophages. However, in MDA-5-/- MEFs type I IFNs were produced to levels equal to or greater than those in wild type MEFs (Figure 21 and 23). The role of RIG-I in CVB3 infection also varied depending on the cell type infected, where it was essential for the sensing of CVB3 in MEFs but only contributed to the induction of maximal levels of

IFNβ in macrophages (Figure 21 and 22). These differences highlight the importance of cell type in sensing since the involvement of RIG-I and MDA-5 changes for the same virus in different cells from the same mice (Table 5).

106

Table 5. Summary of data from EMCV and Coxsackievirus infections. Results from infections of cell lines deficient in indicated genes by indicated viruses is described in respect to the effect in control cells. (+) values greater than control (-) values less than control (--) values less than control to the extent of an absent response (±) values the same as control.

107

In addition to the cell types resulting in differences in sensing through different

RLRs, different cell types also contribute to differences in viral growth, which may contribute to the differences in sensing. For both EMCV and CVB3 viral replication in macrophages was poor while in MEFs both viruses were able to grow (Figures 18, 19, 21,

22 and 23). This defect is not due to the early type I IFN antiviral response observed after infection since the absence of MDA-5 results in the loss of this response but not an increase in replication. This defect may be to the expression more proteins with intrinsic immune functions in macrophages, protecting them from active viral infections by many viruses (96-98, 166). One such protein is SAMHD1 which was recently found to restrict the infection of dendritic and other myeloid cells from HIV by keeping the level of intracellular dNTPs too low for the polymerization of viral DNA in the cytoplasm (73)

(Table 5).

We were not able to compare the roles of MDA-5 or RIG-I within different strain backgrounds to determine the role of murine strain on sensor usage, although we did notice differences between the induction of type I IFNs in wild type macrophages of different backgrounds. Macrophages in a mixed ICR background induced a robust type I

IFN response early after infection with levels peaking at 3-6 hours post infection when infected with EMCV or CVB3. Meanwhile infections of macrophages in a 129x1/SvJ background produced a more gradual and delayed induction of type I IFNs which peaked at six hour after infections with EMCV and 12 to 14 hours after infection with CVB3

(Figure 18 and 21). One possible explanation for this difference is different basal levels of sensors within macrophages of different backgrounds. Quantitative PCR was used to calculate numbers of RIG-I and MDA-5 message relative to β-actin Ct values and it was

108 found that although the ratio of RIG-I copes to MDA-5 copies in each background were similar, the basal levels of both RIG-I and MDA-5 are twice as high in macrophages in a mixed ICR background than those in a 129x1/SvJ background. Also, the deletion of one sensor had no effect on the expression of the other sensor (Figure 17). This may also explain why the role of RIG-I in vivo was not detected since the mixed ICR strain may be able to use an alternate sensing pathway more easily than other strains leading to a resistant strain in general, masking the effects of any sensor. It has been previously shown that strain background can mask the effect of sensors using CVB3 where intraperitoneal infections of C57BL/6 mice with and without MDA-5 resulted in a much greater lethality in the MDA-5 deficient mice. However intraperitoneal infections of wild type and MDA-5 deficient mice with the sane amount of virus in a 129x1/SvJ background showed no difference as both mice were completely resistant to infection

(48).

With differences in cell type and strain background it is not surprising that differences in host organism also affects the role of sensing which is seen in the difference in EMCV infection of HeLa cells and murine cells. In HeLa cells MDA-5 appears to not be important in infection as levels of IFNβ in MDA-5 knockdown cells are limited and differences in levels of ISGs are inappreciable (Figure 16). However, in

MEFs MDA-5 deficiency results in a striking decrease in type I IFN and ISG production and in macrophages is essential in triggering a type I IFN response and ISG induction

(Figure 18 and 19). Taken together this all shows that the discernment between sensor proteins for a specific virus is very dependent upon the context of infection.

109

Sensors can positively and negatively regulate antiviral responses

The sensing of viral infections is key to initiating an immune response to limit and clear an infection. Type I IFNs have been known to play a key role in this since their signaling leads to the induction of hundreds of ISGs, which create an intracellular antiviral state to limit viral replication but also aid in extracellular defenses against viral infection through the expression of chemokines which aid in the recruitment of leukocytes. Therefore the induction of type I IFNs is a standard readout for the sensing of viral infections. The expression of these IFNs often correlates well with an increase in

ISGs, a decrease in viral titers and a decrease in disease (Figure 34). Therefore the absence of a sensor is assumed to result in a decrease in IFNs and ISGs and an increase in viral titers and pathogenesis.

110

Figure 34. Expected responses to viral sensing. An infecting virus replicates its genome by first producing a complementary negative strand of RNA followed by multiple positive sense strands of RNA off of the negative strand. New genomic RNAs are used to form new infectious particles. A sensor recognizes some form of viral RNA to initiate a signaling cascade that results in the expression of cytokines including Type I and type III IFNs. Interferon signaling results in the expression of ISGs. Proteins encoded by ISGs function in limiting/inhibiting viral replication.

111

This however, is not always the case. For example, in the absence of RIG-I IFNβ expression after infection with EMCV and CVB3 is reduced, indicating that RIG-I is a sensor. However the expression of IFNα, another type I IFN does not decrease in the absence of RIG-I but increases, indicating that RIG-I does not contribute to a protective immune response. Since EMCV and CVB3 do not grow in macrophages changes in viral growth are not a good indicator of which type I IFN response, the decreased IFNβ response or increased IFNα response, is more indicative of viral sensing (Figure 18 and

21). Infections of MEFs deficient in RIG-I with CVB3 show a clear absence of IFNβ,

IFNα and ISG expression indicating that RIG-I is sensing CVB3 infections in MEFs to induce an antiviral state. However, this antiviral state is ineffective since it was found that titers of CVB3 actually decrease in the absence of RIG-I (Figure 22). Additionally the growth of CVB3 in the absence of MDA-5 is reduced indicating that MDA-5 is involved in the triggering of an antiviral state. However type I IFNs are not reduced in the absence f MDA-5 and type III IFNs expression is actually increased (Figure 23 and

27).

These data exemplify the complexity of inducing an antiviral response, even in vitro. A sensor may positively regulate one antiviral response but negatively regulate another.

Viral 2A protease has a role in “modulating” sensing

In HeLa cells MDA-5 was found to be dispensable for the induction of IFNβ,

IFNα and ISGs during infections with EMCV (Figure 16). The knockdown of MDA-5 however resulted in a notable decrease in IFNβ expression during poliovirus infections

112

(Figure 28). Although EMCV and poliovirus are similar enough to belong to the same family they are different enough to belong to separate genera. Differences in both the proteins and RNA genomes of the Cardiovirius and Enterovirus genera exist which could account for this difference in sensing. The 5’ untranslated region of cardioviruses contain pseudoknots and a poly C tract which is absent in enteroviruses while enteroviruses have a 5’ clover leaf structure which is replaced by a simple stem loop in cardioviruses.

Additionally the highly structured IRES region of picornavirus genomes which allows for cap-independent translation differs between the viruses (Figure 1). These different RNA structures could act as different ligands for different sensors, however since MDA-5 is able to sense infections of EMCV in murine cells, and human and murine MDA-5 are

80% identical, it is unlikely that this difference in structural features of the viral genome explain why EMCV is not sensed by MDA-5 in HeLa cells.

Key differences in EMCV and poliovirus are the presence of an extra non- structural leader protein, L, encoded in the EMCV genome which poliovirus lacks and the absence of protease activity in the EMCV 2A protein which is present in poliovirus

2A. Using a poliovirus mutant, PV-Y88L, which contains a mutation in the 2A protease to allow it to cleave the viral poly protein but not host cell proteins the ability for 2A to control sensing by MDA-5 in a protease dependent manner was tested. The absence of the ability for the 2A protease to cleave host factors resulted in the ability for PV-Y88L to be able to be sensed in the absence of MDA-5 (Figure 29).

113

Figure 35. Sensing in HeLa cells. (A) Poliovirus infections are capable of being sensed by more than one sensor but sensing through proteins other than MDA-5 are inhibited by the protease function of 2A. (B) EMCV infection is sensed by multiple sensors. Addition of the poliovirus 2A protease and mutation of the L, 2B and 3C protein inhibits the ability for EMCV to be sensed by sensors other than MDA-5. Only one or multiple of these changes may be needed to inhibit sensing through alternate paths. Alternate paths of sensing could be through RIG-I, TLRs or an unidentified sensor.

114

One explanation for this is that protease activity of 2A is involved in inhibiting an alternate sensing pathway such as sensing through RIG-I or TLR3 (Figure 35). The expression of the poliovirus 2A protease during infections with EMCV should be able to inhibit the sensing of EMCV through these alternate pathways, resulting a dependence on

MDA-5 for the sensing of EMCV. The infection of control and MDA-5 knockdown cell lines with EP4, a mutant EMCV virus carrying the poliovirus 2A protease did not show a strong dependence on MDA-5 for sensing. However a plaque purified clone of EP4,

EP4C4, was more dependent on MDA-5 for sensing (Figure 30). The recent sequencing of this clone has shown additional mutations in the L, 2B and 3C proteins. EMCV mutants containing these individual mutations need to be made to determine which of these proteins in EMCV contributes to changing the dependence on MDA-5 for sensing.

However changes in the 3C protein are interesting since it has been shown that the 3C protein is capable of inhibiting the RIG-I and TLR sensing pathway. Hepatitis A and

CVB3 3C proteases have been found to cleave TRIF and inhibit signaling (102, 124).

EMCV 3C protease has been shown to be able to dampen RIG-I induced activation of the

IFNβ promoter through by cleavage of RIG-I (114) and enterovirus 71 3C has been shown to bind RIG-I and inhibit its ability to activate the IFNβ promoter in a protease independent manner (77). This mutation in 3C may confer or enhance the ability for

EMCV to inhibit signaling through these pathways, leading EP4C4’s dependence on

MDA-5 for IFNβ induction (Figure 35). To investigate this idea we would like to examine the contribution of other sensing pathways to the sensing of poliovirus, PV-

Y88L, EMCV and EP4C4. The creation of RIG-I and IPS-1 knockdown HeLa cell lines will help to determine the contributions of RIG-I and non-RLR signaling pathways to

115 sensing. We would hope to see sensing not occurring through RIG-I or any alternative pathway in infections with PV or EP4C4, but sensing through an alternative pathway in infections with PV-Y88L and EMCV.

Alternatively it has been shown that other 2Apro mutations at tyrosine 88 have effects on viral RNA replication (157). Therefore another possible cause for the change in dependence on MDA-5 for sensing may be due to a change in RNA replication, either a decrease in total viral RNA or a specific intracellular form of viral RNA (Figure 4) could also be factors affecting the ability of virus to be sensed by an alternate pathway.

A correlation between kinetics of viral growth and sensing has not been observed (Figure

28, 29 and 30) but the amount of infectious virus produced is not necessarily indicative of the amount of RNA produced. Therefore further investigations should be done to examine the levels of viral RNA during infections through qPCR or northern blot analysis of RNA.

116

References

1. Abe Y., K. Fujii, N. Nagata, O. Takeuchi, S. Akira, H. Oshiumi, M. Matsumoto, T. Seya, and S. Koike. 2012. The toll-like receptor 3-mediated antiviral response is important for protection against poliovirus infection in poliovirus receptor transgenic mice. Journal of Virology 86:185–194.

2. Akira S. 2009. Innate immunity to pathogens: diversity in receptors for microbial recognition. Immunol. Rev. 227:5–8.

3. Akira S., S. Uematsu, and O. Takeuchi. 2006. Pathogen recognition and innate immunity. Cell 124:783–801.

4. Ambros V., and D. Baltimore. 1980. Purification and properties of a HeLa cell enzyme able to remove the 5'-terminal protein from poliovirus RNA. J. Biol. Chem. 255:6739–6744.

5. Ank N., M. B. Iversen, C. Bartholdy, P. Staeheli, R. Hartmann, U. B. Jensen, F. Dagnaes-Hansen, A. R. Thomsen, Z. Chen, H. Haugen, K. Klucher, and S. R. Paludan. 2008. An important role for type III interferon (IFN-lambda/IL- 28) in TLR-induced antiviral activity. J Immunol 180:2474–2485.

6. Ank N., H. West, C. Bartholdy, K. Eriksson, A. R. Thomsen, and S. R. Paludan. 2006. Lambda interferon (IFN-lambda), a type III IFN, is induced by viruses and IFNs and displays potent antiviral activity against select virus infections in vivo. Journal of Virology 80:4501–4509.

7. Arimoto K.-I., H. Takahashi, T. Hishiki, H. Konishi, T. Fujita, and K. Shimotohno. 2007. Negative regulation of the RIG-I signaling by the ubiquitin ligase RNF125. Proc Natl Acad Sci USA 104:7500–7505.

8. Bamming D., and C. M. Horvath. 2009. Regulation of signal transduction by enzymatically inactive antiviral RNA helicase proteins MDA5, RIG-I, and LGP2. J. Biol. Chem. 284:9700–9712.

9. Baril M., M.-E. Racine, F. Penin, and D. Lamarre. 2009. MAVS dimer is a crucial signaling component of innate immunity and the target of hepatitis C virus NS3/4A protease. Journal of Virology 83:1299–1311.

10. Barral P. M., J. M. Morrison, J. Drahos, P. Gupta, D. Sarkar, P. B. Fisher, and V. R. Racaniello. 2007. MDA-5 is cleaved in poliovirus-infected cells. Journal of Virology 81:3677–3684.

11. Barral P. M., D. Sarkar, P. B. Fisher, and V. R. Racaniello. 2009. RIG-I is cleaved during picornavirus infection. Virology 391:171–176.

117

12. Baum A., R. Sachidanandam, and A. García-Sastre. 2010. Preference of RIG- I for short viral RNA molecules in infected cells revealed by next-generation sequencing. Proc Natl Acad Sci USA 107:16303–16308.

13. Bedard K. M., and B. L. Semler. 2004. Regulation of picornavirus gene expression. Microbes Infect. 6:702–713.

14. Berghäll H., J. Sirén, D. Sarkar, I. Julkunen, P. B. Fisher, R. Vainionpää, and S. Matikainen. 2006. The interferon-inducible RNA helicase, mda-5, is involved in measles virus-induced expression of antiviral cytokines. Microbes Infect. 8:2138–2144.

15. Berke I. C., and Y. Modis. 2012. MDA5 cooperatively forms dimers and ATP- sensitive filaments upon binding double-stranded RNA. EMBO J.

16. Borden E. C., G. C. Sen, G. Uze, R. H. Silverman, R. M. Ransohoff, G. R. Foster, and G. R. Stark. 2007. Interferons at age 50: past, current and future impact on biomedicine. Nat Rev Drug Discov 6:975–990.

17. Chebath J., P. Benech, M. Revel, and M. Vigneron. 1987. Constitutive expression of (2“-5”) oligo A synthetase confers resistance to picornavirus infection. Nature 330:587–588.

18. Co J. G., K. W. Witwer, L. Gama, M. C. Zink, and J. E. Clements. 2011. Induction of innate immune responses by SIV in vivo and in vitro: differential expression and function of RIG-I and MDA5. J. Infect. Dis. 204:1104–1114.

19. Cui S., K. Eisenächer, A. Kirchhofer, K. Brzózka, A. Lammens, K. Lammens, T. Fujita, K.-K. Conzelmann, A. Krug, and K.-P. Hopfner. 2008. The C-terminal regulatory domain is the RNA 5'-triphosphate sensor of RIG-I. Mol. Cell 29:169–179.

20. Diao F., S. Li, Y. Tian, M. Zhang, L.-G. Xu, Y. Zhang, R.-P. Wang, D. Chen, Z. Zhai, B. Zhong, P. Tien, and H.-B. Shu. 2007. Negative regulation of MDA5- but not RIG-I-mediated innate antiviral signaling by the dihydroxyacetone kinase. Proc Natl Acad Sci USA 104:11706–11711.

21. Dixit E., S. Boulant, Y. Zhang, A. S. Y. Lee, C. Odendall, B. Shum, N. Hacohen, Z. J. Chen, S. P. Whelan, M. Fransen, M. L. Nibert, G. Superti- Furga, and J. C. Kagan. 2010. Peroxisomes are signaling platforms for antiviral innate immunity. Cell 141:668–681.

22. Dong B., L. Xu, A. Zhou, B. A. Hassel, X. Lee, P. F. Torrence, and R. H. Silverman. 1994. Intrinsic molecular activities of the interferon-induced 2-5A- dependent RNase. J. Biol. Chem. 269:14153–14158.

23. Donnelly R. P., and S. V. Kotenko. 2010. Interferon-lambda: a new addition to an old family. J. Interferon Cytokine Res. 30:555–564.

118

24. Drahos J., and V. R. Racaniello. 2009. Cleavage of IPS-1 in cells infected with human rhinovirus. Journal of Virology 83:11581–11587.

25. Ehrenfeld E., and A. S. F. Microbiology. 2010. The Picornaviruses, 1st ed. Amer Society for Microbiology.

26. Eisenächer K., and A. Krug. 2011. Regulation of RLR-mediated innate immune signaling - It is all about keeping the balance. European journal of cell biology.

27. Farrar M. A., and R. D. Schreiber. 1993. The molecular cell biology of interferon-gamma and its receptor. Annu. Rev. Immunol. 11:571–611.

28. Fensterl V., and G. C. Sen. 2009. Interferons and viral infections. Biofactors 35:14–20.

29. Fields B. N., D. M. Knipe, and P. M. Howley. 1996. Fields virology. Lippincott Williams & Wilkins.

30. Flint S. J., L. W. Enquist, and V. R. Racaniello. 2009. Principles of Virology (2 Volume Set), 3rd ed. ASM Press.

31. Fredericksen B. L., B. C. Keller, J. Fornek, M. G. Katze, and M. Gale. 2008. Establishment and maintenance of the innate antiviral response to West Nile Virus involves both RIG-I and MDA5 signaling through IPS-1. Journal of Virology 82:609–616.

32. Fu J., Y. Xiong, Y. Xu, G. Cheng, and H. Tang. 2011. MDA5 is SUMOylated by PIAS2β in the upregulation of type I interferon signaling. Mol Immunol 48:415–422.

33. Gack M. U., E. Nistal-Villán, K.-S. Inn, A. García-Sastre, and J. U. Jung. 2010. Phosphorylation-mediated negative regulation of RIG-I antiviral activity. Journal of Virology 84:3220–3229.

34. Gack M. U., Y. C. Shin, C.-H. Joo, T. Urano, C. Liang, L. Sun, O. Takeuchi, S. Akira, Z. Chen, S. Inoue, and J. U. Jung. 2007. TRIM25 RING-finger E3 ubiquitin ligase is essential for RIG-I-mediated antiviral activity. Nature 446:916–920.

35. Gao D., Y.-K. Yang, R.-P. Wang, X. Zhou, F.-C. Diao, M.-D. Li, Z.-H. Zhai, Z.-F. Jiang, and D.-Y. Chen. 2009. REUL is a novel E3 ubiquitin ligase and stimulator of retinoic-acid-inducible gene-I. PLoS ONE 4:e5760.

36. Gitlin L., W. Barchet, S. Gilfillan, M. Cella, B. Beutler, R. A. Flavell, M. S. Diamond, and M. Colonna. 2006. Essential role of mda-5 in type I IFN responses to polyriboinosinic:polyribocytidylic acid and encephalomyocarditis picornavirus. Proc Natl Acad Sci USA 103:8459–8464.

119

37. Habjan M., I. Andersson, J. Klingström, M. Schümann, A. Martin, P. Zimmermann, V. Wagner, A. Pichlmair, U. Schneider, E. Mühlberger, A. Mirazimi, and F. Weber. 2008. Processing of genome 5' termini as a strategy of negative-strand RNA viruses to avoid RIG-I-dependent interferon induction. PLoS ONE 3:e2032.

38. Hahn H., and A. C. Palmenberg. 1995. Encephalomyocarditis viruses with short poly(C) tracts are more virulent than their mengovirus counterparts. Journal of Virology 69:2697–2699.

39. Hardarson H. S., J. S. Baker, Z. Yang, E. Purevjav, C.-H. Huang, L. Alexopoulou, N. Li, R. A. Flavell, N. E. Bowles, and J. G. Vallejo. 2007. Toll- like receptor 3 is an essential component of the innate stress response in virus- induced cardiac injury. Am. J. Physiol. Heart Circ. Physiol. 292:H251–8.

40. Hardy M. P., C. M. Owczarek, L. S. Jermiin, M. Ejdebäck, and P. J. Hertzog. 2004. Characterization of the type I interferon locus and identification of novel genes. Genomics 84:331–345.

41. Hato S. V., C. Ricour, B. M. Schulte, K. H. W. Lanke, M. de Bruijni, J. Zoll, W. J. G. Melchers, T. Michiels, and F. J. M. van Kuppeveld. 2007. The mengovirus leader protein blocks interferon-alpha/beta gene transcription and inhibits activation of interferon regulatory factor 3. Cell. Microbiol. 9:2921– 2930.

42. Havell E. A., B. Berman, C. A. Ogburn, K. Berg, K. Paucker, and J. Vilcek. 1975. Two antigenically distinct species of human interferon. Proc Natl Acad Sci USA 72:2185–2187.

43. Hayakawa S., S. Shiratori, H. Yamato, T. Kameyama, C. Kitatsuji, F. Kashigi, S. Goto, S. Kameoka, D. Fujikura, T. Yamada, T. Mizutani, M. Kazumata, M. Sato, J. Tanaka, M. Asaka, Y. Ohba, T. Miyazaki, M. Imamura, and A. Takaoka. 2011. ZAPS is a potent stimulator of signaling mediated by the RNA helicase RIG-I during antiviral responses. Nat. Immunol. 12:37–44.

44. Hewlett M. J., J. K. Rose, and D. Baltimore. 1976. 5'-terminal structure of poliovirus polyribosomal RNA is pUp. Proc Natl Acad Sci USA 73:327–330.

45. Honda K., and T. Taniguchi. 2006. IRFs: master regulators of signalling by Toll-like receptors and cytosolic pattern-recognition receptors. Nat. Rev. Immunol. 6:644–658.

46. Hornung V., J. Ellegast, S. Kim, K. Brzózka, A. Jung, H. Kato, H. Poeck, S. Akira, K.-K. Conzelmann, M. Schlee, S. Endres, and G. Hartmann. 2006. 5'- Triphosphate RNA is the ligand for RIG-I. Science 314:994–997.

47. Huber S. A. 1994. VCAM-1 is a receptor for encephalomyocarditis virus on

120

murine vascular endothelial cells. Journal of Virology 68:3453–3458.

48. Hühn M. H., S. A. Mccartney, K. Lind, E. Svedin, M. Colonna, and M. Flodström-Tullberg. 2010. Melanoma differentiation-associated protein-5 (MDA-5) limits early viral replication but is not essential for the induction of type 1 interferons after Coxsackievirus infection. Virology 401:42–48.

49. ISAACS A., and J. LINDENMANN. 1957. Virus interference. I. The interferon. Proc. R. Soc. Lond., B, Biol. Sci. 147:258–267.

50. Iversen M. B., N. Ank, J. Melchjorsen, and S. R. Paludan. 2010. Expression of type III interferon (IFN) in the vaginal mucosa is mediated primarily by dendritic cells and displays stronger dependence on NF-kappaB than type I IFNs. Journal of Virology 84:4579–4586.

51. Jiang F., A. Ramanathan, M. T. Miller, G.-Q. Tang, M. Gale, S. S. Patel, and J. Marcotrigiano. 2011. Structural basis of RNA recognition and activation by innate immune receptor RIG-I. Nature 479:423–427.

52. Jiang H., and P. B. Fisher. 1993. Use of a sensitive and efficient subtraction hybridization protocol for the identification of genes differentially regulated during the induction of differentiation in human melanoma cells. Mol Cell Different 1:285–299.

53. Jin Y. M., I. U. Pardoe, A. T. Burness, and T. I. Michalak. 1994. Identification and characterization of the cell surface 70-kilodalton sialoglycoprotein(s) as a candidate receptor for encephalomyocarditis virus on human nucleated cells. Journal of Virology 68:7308–7319.

54. Jin Y.-H., S. J. Kim, E. Y. So, L. Meng, M. Colonna, and B. S. Kim. 2012. Melanoma differentiation-associated gene 5 is critical for protection against Theiler's virus-induced demyelinating disease. Journal of Virology 86:1531– 1543.

55. Kang D.-C., R. V. Gopalkrishnan, L. Lin, A. Randolph, K. Valerie, S. Pestka, and P. B. Fisher. 2004. Expression analysis and genomic characterization of human melanoma differentiation associated gene-5, mda-5: a novel type I interferon-responsive apoptosis-inducing gene. Oncogene 23:1789– 1800.

56. Kang D.-C., R. V. Gopalkrishnan, Q. Wu, E. Jankowsky, A. M. Pyle, and P. B. Fisher. 2002. mda-5: An interferon-inducible putative RNA helicase with double-stranded RNA-dependent ATPase activity and melanoma growth- suppressive properties. Proc Natl Acad Sci USA 99:637–642.

57. Kato H., S. Sato, M. Yoneyama, M. Yamamoto, S. Uematsu, K. Matsui, T. Tsujimura, K. Takeda, T. Fujita, O. Takeuchi, and S. Akira. 2005. Cell type- specific involvement of RIG-I in antiviral response. Immunity 23:19–28.

121

58. Kato H., O. Takeuchi, E. Mikamo-Satoh, R. Hirai, T. Kawai, K. Matsushita, A. Hiiragi, T. S. Dermody, T. Fujita, and S. Akira. 2008. Length-dependent recognition of double-stranded ribonucleic acids by retinoic acid-inducible gene- I and melanoma differentiation-associated gene 5. J. Exp. Med. 205:1601–1610.

59. Kato H., O. Takeuchi, S. Sato, M. Yoneyama, M. Yamamoto, K. Matsui, S. Uematsu, A. Jung, T. Kawai, K. J. Ishii, O. Yamaguchi, K. Otsu, T. Tsujimura, C.-S. Koh, C. Reis E Sousa, Y. Matsuura, T. Fujita, and S. Akira. 2006. Differential roles of MDA5 and RIG-I helicases in the recognition of RNA viruses. Nature 441:101–105.

60. Kaufmann S. H. E., B. T. Rouse, and D. L. Sacks. 2010. The immune response to infection. Amer Society for Microbiology.

61. Kawai T., and S. Akira. 2009. The roles of TLRs, RLRs and NLRs in pathogen recognition. Int. Immunol. 21:317–337.

62. Kawai T., K. Takahashi, S. Sato, C. Coban, H. Kumar, H. Kato, K. J. Ishii, O. Takeuchi, and S. Akira. 2005. IPS-1, an adaptor triggering RIG-I- and Mda5-mediated type I interferon induction. Nat. Immunol. 6:981–988.

63. Kim M.-J., S.-Y. Hwang, T. Imaizumi, and J.-Y. Yoo. 2008. Negative feedback regulation of RIG-I-mediated antiviral signaling by interferon-induced ISG15 conjugation. Journal of Virology 82:1474–1483.

64. Kirkegaard K., and W. T. Jackson. 2005. Topology of double-membraned vesicles and the opportunity for non-lytic release of cytoplasm. Autophagy 1:182–184.

65. Komuro A., D. Bamming, and C. M. Horvath. 2008. Negative regulation of cytoplasmic RNA-mediated antiviral signaling. Cytokine 43:350–358.

66. Kong L., L. Sun, H. Zhang, Q. Liu, Y. Liu, L. Qin, G. Shi, J.-H. Hu, A. Xu, Y.-P. Sun, D. Li, Y.-F. Shi, J.-W. Zang, J. Zhu, Z. Chen, Z.-G. Wang, and B.-X. Ge. 2009. An essential role for RIG-I in toll-like receptor-stimulated phagocytosis. Cell Host Microbe 6:150–161.

67. Koppers-Lalic D., and R. C. Hoeben. 2011. Non-human viruses developed as therapeutic agent for use in humans. Rev. Med. Virol. 21:227–239.

68. Kotenko S. V. 2011. IFN-λs. Current Opinion in Immunology 23:583–590.

69. Kotenko S. V., G. Gallagher, V. V. Baurin, A. Lewis-Antes, M. Shen, N. K. Shah, J. A. Langer, F. Sheikh, H. Dickensheets, and R. P. Donnelly. 2003. IFN-lambdas mediate antiviral protection through a distinct class II cytokine receptor complex. Nat. Immunol. 4:69–77.

70. Kovacsovics M., F. Martinon, O. Micheau, J. L. Bodmer, K. Hofmann, and

122

J. Tschopp. 2002. Overexpression of Helicard, a CARD-containing helicase cleaved during apoptosis, accelerates DNA degradation. Curr. Biol. 12:838–843.

71. Kumar H., T. Kawai, and S. Akira. 2009. Pathogen recognition in the innate immune response. Biochem. J. 420:1–16.

72. Kumar H., T. Kawai, H. Kato, S. Sato, K. Takahashi, C. Coban, M. Yamamoto, S. Uematsu, K. J. Ishii, O. Takeuchi, and S. Akira. 2006. Essential role of IPS-1 in innate immune responses against RNA viruses. J. Exp. Med. 203:1795–1803.

73. Lahouassa H., W. Daddacha, H. Hofmann, D. Ayinde, E. C. Logue, L. Dragin, N. Bloch, C. Maudet, M. Bertrand, T. Gramberg, G. Pancino, S. Priet, B. Canard, N. Laguette, M. Benkirane, C. Transy, N. R. Landau, B. Kim, and F. Margottin-Goguet. 2012. SAMHD1 restricts the replication of human immunodeficiency virus type 1 by depleting the intracellular pool of deoxynucleoside triphosphates. Nat. Immunol. 13:223–228.

74. Lasfar A. 2006. Characterization of the Mouse IFN- Ligand-Receptor System: IFN- s Exhibit Antitumor Activity against B16 Melanoma. Cancer Res. 66:4468– 4477.

75. Lech M., A. Avila-Ferrufino, V. Skuginna, H. E. Susanti, and H.-J. Anders. 2010. Quantitative expression of RIG-like helicase, NOD-like receptor and inflammasome-related mRNAs in humans and mice. Int. Immunol. 22:717–728.

76. Lee C. C., A. M. Avalos, and H. L. Ploegh. 2012. Accessory molecules for Toll-like receptors and their function. Nat. Rev. Immunol. 12:168–179.

77. Lei X., X. Liu, Y. Ma, Z. Sun, Y. Yang, Q. Jin, B. He, and J. Wang. 2010. The 3C protein of enterovirus 71 inhibits retinoid acid-inducible gene I-mediated interferon regulatory factor 3 activation and type I interferon responses. Journal of Virology 84:8051–8061.

78. Levy-Koenig R., and R. Golgher. 1970. Immunology of Interferons. The Journal of ….

79. Li X.-D., L. Sun, R. B. Seth, G. Pineda, and Z. J. Chen. 2005. Hepatitis C virus protease NS3/4A cleaves mitochondrial antiviral signaling protein off the mitochondria to evade innate immunity. Proc Natl Acad Sci USA 102:17717– 17722.

80. Li X., C. T. Ranjith-Kumar, M. T. Brooks, S. Dharmaiah, A. B. Herr, C. Kao, and P. Li. 2009. The RIG-I-like receptor LGP2 recognizes the termini of double-stranded RNA. J. Biol. Chem. 284:13881–13891.

81. Liu T. X., J. W. Zhang, J. Tao, R. B. Zhang, Q. H. Zhang, C. J. Zhao, J. H. Tong, M. Lanotte, S. Waxman, S. J. Chen, M. Mao, G. X. Hu, L. Zhu, and

123

Z. Chen. 2000. Gene expression networks underlying retinoic acid-induced differentiation of acute promyelocytic leukemia cells. Blood 96:1496–1504.

82. Liu Y., C. Wang, S. Mueller, A. V. Paul, E. Wimmer, and P. Jiang. 2010. Direct interaction between two viral proteins, the nonstructural protein 2C and the capsid protein VP3, is required for enterovirus morphogenesis. PLoS Pathogens 6:e1001066.

83. Loo Y.-M., J. Fornek, N. Crochet, G. Bajwa, O. Perwitasari, L. Martinez- Sobrido, S. Akira, M. A. Gill, A. García-Sastre, M. G. Katze, and M. Gale. 2008. Distinct RIG-I and MDA5 signaling by RNA viruses in innate immunity. Journal of Virology 82:335–345.

84. Loo Y.-M., and M. Gale. 2011. Immune signaling by RIG-I-like receptors. Immunity 34:680–692.

85. Loo Y.-M., D. M. Owen, K. Li, A. K. Erickson, C. L. Johnson, P. M. Fish, D. S. Carney, T. Wang, H. Ishida, M. Yoneyama, T. Fujita, T. Saito, W. M. Lee, C. H. Hagedorn, D. T.-Y. Lau, S. A. Weinman, S. M. Lemon, and M. Gale. 2006. Viral and therapeutic control of IFN-beta promoter stimulator 1 during hepatitis C virus infection. Proc Natl Acad Sci USA 103:6001–6006.

86. Lu J., L. Yi, J. Zhao, J. Yu, Y. Chen, M. C. Lin, H.-F. Kung, and M.-L. He. 2012. Enterovirus 71 Disrupts Interferon Signaling by reducing the Interferon Receptor I. Journal of Virology.

87. Luo D., S. C. Ding, A. Vela, A. Kohlway, B. D. Lindenbach, and A. M. Pyle. 2011. Structural insights into RNA recognition by RIG-I. Cell 147:409–422.

88. Luthra P., D. Sun, R. H. Silverman, and B. He. 2011. Activation of IFN- β expression by a viral mRNA through RNase L and MDA5. Proc Natl Acad Sci USA 108:2118–2123.

89. Malathi K., B. Dong, M. Gale, and R. H. Silverman. 2007. Small self-RNA generated by RNase L amplifies antiviral innate immunity. Nature 448:816–819.

90. Malathi K., T. Saito, N. Crochet, D. J. Barton, M. Gale, and R. H. Silverman. 2010. RNase L releases a small RNA from HCV RNA that refolds into a potent PAMP. RNA 16:2108–2119.

91. Mccartney S. A., L. B. Thackray, L. Gitlin, S. Gilfillan, H. W. Virgin Iv, and M. Colonna. 2008. MDA-5 Recognition of a Murine Norovirus. PLoS Pathogens 4:e1000108.

92. Mccartney S. A., W. Vermi, S. Lonardi, C. Rossini, K. Otero, B. Calderon, S. Gilfillan, M. S. Diamond, E. R. Unanue, and M. Colonna. 2011. RNA sensor-induced type I IFN prevents diabetes caused by a β cell-tropic virus in mice. J. Clin. Invest. 121:1497–1507.

124

93. Meylan E., J. Curran, K. Hofmann, D. Moradpour, M. Binder, R. Bartenschlager, and J. Tschopp. 2005. Cardif is an adaptor protein in the RIG-I antiviral pathway and is targeted by hepatitis C virus. Nature 437:1167–1172.

94. Mi Z., J. Fu, Y. Xiong, and H. Tang. 2010. SUMOylation of RIG-I positively regulates the type I interferon signaling. Protein Cell 1:275–283.

95. Miyashita M., H. Oshiumi, M. Matsumoto, and T. Seya. 2011. DDX60, a DEXD/H box helicase, is a novel antiviral factor promoting RIG-I-like receptor- mediated signaling. Mol. Cell. Biol. 31:3802–3819.

96. Mogensen S. C. 1985. Genetic aspects of macrophage involvement in natural resistance to virus infections. Immunol. Lett. 11:219–224.

97. Mogensen S. C. 1979. Role of macrophages in natural resistance to virus infections. Microbiol. Rev. 43:1–26.

98. Morahan P. S., J. R. Connor, and K. R. Leary. 1985. Viruses and the versatile macrophage. Br. Med. Bull. 41:15–21.

99. Mordstein M., G. Kochs, L. Dumoutier, J.-C. Renauld, S. R. Paludan, K. Klucher, and P. Staeheli. 2008. Interferon-lambda contributes to innate immunity of mice against influenza A virus but not against hepatotropic viruses. PLoS Pathogens 4:e1000151.

100. Mordstein M., E. Neugebauer, V. Ditt, B. Jessen, T. Rieger, V. Falcone, F. Sorgeloos, S. Ehl, D. Mayer, G. Kochs, M. Schwemmle, S. Günther, C. Drosten, T. Michiels, and P. Staeheli. 2010. Lambda interferon renders epithelial cells of the respiratory and gastrointestinal tracts resistant to viral infections. Journal of Virology 84:5670–5677.

101. Morrison J. M., and V. R. Racaniello. 2009. Proteinase 2Apro is essential for enterovirus replication in type I interferon-treated cells. Journal of Virology 83:4412–4422.

102. Mukherjee A., S. A. Morosky, E. Delorme-Axford, N. Dybdahl-Sissoko, M. S. Oberste, T. Wang, and C. B. Coyne. 2011. The coxsackievirus B 3C protease cleaves MAVS and TRIF to attenuate host type I interferon and apoptotic signaling. PLoS Pathogens 7:e1001311.

103. Nistal-Villán E., M. U. Gack, G. Martínez-Delgado, N. P. Maharaj, K.-S. Inn, H. Yang, R. Wang, A. K. Aggarwal, J. U. Jung, and A. García-Sastre. 2010. Negative role of RIG-I serine 8 phosphorylation in the regulation of interferon-beta production. J. Biol. Chem. 285:20252–20261.

104. Nomoto A., N. Kitamura, F. Golini, and E. Wimmer. 1977. The 5'-terminal structures of poliovirion RNA and poliovirus mRNA differ only in the genome- linked protein VPg. Proc Natl Acad Sci USA 74:5345–5349.

125

105. Nugent C. I., K. L. Johnson, P. Sarnow, and K. Kirkegaard. 1999. Functional coupling between replication and packaging of poliovirus replicon RNA. Journal of Virology 73:427–435.

106. Oberg B., and L. Philipson. 1969. Replication of poliovirus RNA studied by gel filtration and electrophoresis. Eur. J. Biochem. 11:305–315.

107. Okabayashi T., T. Kojima, T. Masaki, S.-I. Yokota, T. Imaizumi, H. Tsutsumi, T. Himi, N. Fujii, and N. Sawada. 2011. Type-III interferon, not type-I, is the predominant interferon induced by respiratory viruses in nasal epithelial cells. Virus Res. 160:360–366.

108. Onoguchi K., M. Yoneyama, A. Takemura, S. Akira, T. Taniguchi, H. Namiki, and T. Fujita. 2007. Viral infections activate types I and III interferon genes through a common mechanism. J. Biol. Chem. 282:7576–7581.

109. Oshiumi H., M. Matsumoto, S. Hatakeyama, and T. Seya. 2009. Riplet/RNF135, a RING finger protein, ubiquitinates RIG-I to promote interferon-beta induction during the early phase of viral infection. J. Biol. Chem. 284:807–817.

110. Oshiumi H., M. Okamoto, K. Fujii, T. Kawanishi, M. Matsumoto, S. Koike, and T. Seya. 2011. The TLR3/TICAM-1 pathway is mandatory for innate immune responses to poliovirus infection. The Journal of Immunology 187:5320–5327.

111. Oshiumi H., K. Sakai, M. Matsumoto, and T. Seya. 2010. DEAD/H BOX 3 (DDX3) helicase binds the RIG-I adaptor IPS-1 to up-regulate IFN-beta-inducing potential. Eur J Immunol 40:940–948.

112. Osterlund P., V. Veckman, J. Sirén, K. M. Klucher, J. Hiscott, S. Matikainen, and I. Julkunen. 2005. Gene expression and antiviral activity of alpha/beta interferons and interleukin-29 in virus-infected human myeloid dendritic cells. Journal of Virology 79:9608–9617.

113. Pagano J. S., and A. Vaheri. 1965. Enhancement of infectivity of poliovirus RNA with diethylaminoethyl-dextran (DEAE-D). Arch Gesamte Virusforsch 17:456–464.

114. Papon L., A. Oteiza, T. Imaizumi, H. Kato, E. Brocchi, T. G. Lawson, S. Akira, and N. Mechti. 2009. The viral RNA recognition sensor RIG-I is degraded during encephalomyocarditis virus (EMCV) infection. Virology 393:311–318.

115. Peisley A., C. Lin, B. Wu, M. Orme-Johnson, M. Liu, T. Walz, and S. Hur. 2011. Cooperative assembly and dynamic disassembly of MDA5 filaments for viral dsRNA recognition. Proc Natl Acad Sci USA 108:21010–21015.

126

116. Pestka S., C. D. Krause, and M. R. Walter. 2004. Interferons, interferon-like cytokines, and their receptors. Immunol. Rev. 202:8–32.

117. Pfaffl M. W. 2001. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 29:e45.

118. Pichlmair A., and C. Reis E Sousa. 2007. Innate recognition of viruses. Immunity 27:370–383.

119. Pichlmair A., O. Schulz, C. P. Tan, T. I. Näslund, P. Liljeström, F. Weber, and C. Reis E Sousa. 2006. RIG-I-mediated antiviral responses to single- stranded RNA bearing 5'-phosphates. Science 314:997–1001.

120. Pichlmair A., O. Schulz, C. P. Tan, J. Rehwinkel, H. Kato, O. Takeuchi, S. Akira, M. Way, G. Schiavo, and C. Reis E Sousa. 2009. Activation of MDA5 requires higher-order RNA structures generated during virus infection. Journal of Virology 83:10761–10769.

121. Pippig D. A., J. C. Hellmuth, S. Cui, A. Kirchhofer, K. Lammens, A. Lammens, A. Schmidt, S. Rothenfusser, and K.-P. Hopfner. 2009. The regulatory domain of the RIG-I family ATPase LGP2 senses double-stranded RNA. Nucleic Acids Res. 37:2014–2025.

122. Plumet S., F. Herschke, J.-M. Bourhis, H. Valentin, S. Longhi, and D. Gerlier. 2007. Cytosolic 5'-triphosphate ended viral leader transcript of measles virus as activator of the RIG I-mediated interferon response. PLoS ONE 2:e279.

123. Pott J., T. Mahlakõiv, M. Mordstein, C. U. Duerr, T. Michiels, S. Stockinger, P. Staeheli, and M. W. Hornef. 2011. IFN-lambda determines the intestinal epithelial antiviral host defense. Proc Natl Acad Sci USA 108:7944– 7949.

124. Qu L., Z. Feng, D. Yamane, Y. Liang, R. E. Lanford, K. Li, and S. M. Lemon. 2011. Disruption of TLR3 Signaling Due to Cleavage of TRIF by the Hepatitis A Virus Protease-Polymerase Processing Intermediate, 3CD. PLoS Pathogens 7:e1002169.

125. Racaniello V. R., and D. Baltimore. 1981. Cloned poliovirus complementary DNA is infectious in mammalian cells. Science 214:916–919.

126. Rehwinkel J., C. P. Tan, D. Goubau, O. Schulz, A. Pichlmair, K. Bier, N. Robb, F. Vreede, W. Barclay, E. Fodor, and C. Reis E Sousa. 2010. RIG-I detects viral genomic RNA during negative-strand RNA virus infection. Cell 140:397–408.

127. Richards O. C., S. C. Martin, H. G. Jense, and E. Ehrenfeld. 1984. Structure of poliovirus replicative intermediate RNA. Electron microscope analysis of RNA cross-linked in vivo with psoralen derivative. J. Mol. Biol. 173:325–340.

127

128. Roth-Cross J. K., S. J. Bender, and S. R. Weiss. 2008. Murine coronavirus mouse hepatitis virus is recognized by MDA5 and induces type I interferon in brain macrophages/microglia. Journal of Virology 82:9829–9838.

129. Rothenfusser S., N. Goutagny, G. DiPerna, M. Gong, B. G. Monks, A. Schoenemeyer, M. Yamamoto, S. Akira, and K. A. Fitzgerald. 2005. The RNA helicase Lgp2 inhibits TLR-independent sensing of viral replication by retinoic acid-inducible gene-I. J Immunol 175:5260–5268.

130. Saito T., and M. Gale. 2008. Differential recognition of double-stranded RNA by RIG-I-like receptors in antiviral immunity. J. Exp. Med. 205:1523–1527.

131. Saito T., R. Hirai, Y.-M. Loo, D. Owen, C. L. Johnson, S. C. Sinha, S. Akira, T. Fujita, and M. Gale. 2007. Regulation of innate antiviral defenses through a shared repressor domain in RIG-I and LGP2. Proc Natl Acad Sci USA 104:582– 587.

132. Satoh T., H. Kato, Y. Kumagai, M. Yoneyama, S. Sato, K. Matsushita, T. Tsujimura, T. Fujita, S. Akira, and O. Takeuchi. 2010. LGP2 is a positive regulator of RIG-I- and MDA5-mediated antiviral responses. Proc Natl Acad Sci USA 107:1512–1517.

133. Semler B. L., and E. Wimmer. 2002. Molecular biology of picornaviruses. Amer Society for Microbiology.

134. Sen G. C. 2001. Viruses and interferons. Annu. Rev. Microbiol. 55:255–281.

135. Seth R. B., L. Sun, C.-K. Ea, and Z. J. Chen. 2005. Identification and characterization of MAVS, a mitochondrial antiviral signaling protein that activates NF-kappaB and IRF 3. Cell 122:669–682.

136. Sheppard P., W. Kindsvogel, W. Xu, K. Henderson, S. Schlutsmeyer, T. E. Whitmore, R. Kuestner, U. Garrigues, C. Birks, J. Roraback, C. Ostrander, D. Dong, J. Shin, S. Presnell, B. Fox, B. Haldeman, E. Cooper, D. Taft, T. Gilbert, F. J. Grant, M. Tackett, W. Krivan, G. McKnight, C. Clegg, D. Foster, and K. M. Klucher. 2003. IL-28, IL-29 and their class II cytokine receptor IL-28R. Nat. Immunol. 4:63–68.

137. Shmueli O., S. Horn-Saban, V. Chalifa-Caspi, M. Shmoish, R. Ophir, H. Benjamin-Rodrig, M. Safran, E. Domany, and D. Lancet. 2003. GeneNote: whole genome expression profiles in normal human tissues. C. R. Biol. 326:1067–1072.

138. Sirén J., T. Imaizumi, D. Sarkar, T. Pietilä, D. L. Noah, R. Lin, J. Hiscott, R. M. Krug, P. B. Fisher, I. Julkunen, and S. Matikainen. 2006. Retinoic acid inducible gene-I and mda-5 are involved in influenza A virus-induced expression of antiviral cytokines. Microbes Infect. 8:2013–2020.

128

139. Slater L., N. W. Bartlett, J. J. Haas, J. Zhu, S. D. Message, R. P. Walton, A. Sykes, S. Dahdaleh, D. L. Clarke, M. G. Belvisi, O. M. Kon, T. Fujita, P. K. Jeffery, S. L. Johnston, and M. R. Edwards. 2010. Co-ordinated role of TLR3, RIG-I and MDA5 in the innate response to rhinovirus in bronchial epithelium. PLoS Pathogens 6:e1001178.

140. Sommereyns C., S. Paul, P. Staeheli, and T. Michiels. 2008. IFN-Lambda (IFN-λ) Is Expressed in a Tissue-Dependent Fashion and Primarily Acts on Epithelial Cells In Vivo. PLoS Pathogens 4:e1000017.

141. Su Z.-Z., D. Sarkar, L. Emdad, P. M. Barral, and P. B. Fisher. 2007. Central role of interferon regulatory factor-1 (IRF-1) in controlling retinoic acid inducible gene-I (RIG-I) expression. J. Cell. Physiol. 213:502–510.

142. Suhy D. A., T. H. Giddings, and K. Kirkegaard. 2000. Remodeling the endoplasmic reticulum by poliovirus infection and by individual viral proteins: an autophagy-like origin for virus-induced vesicles. Journal of Virology 74:8953–8965.

143. Sumpter R., Y.-M. Loo, E. Foy, K. Li, M. Yoneyama, T. Fujita, S. M. Lemon, and M. Gale. 2005. Regulating intracellular antiviral defense and permissiveness to hepatitis C virus RNA replication through a cellular RNA helicase, RIG-I. Journal of Virology 79:2689–2699.

144. Takahasi K., H. Kumeta, N. Tsuduki, R. Narita, T. Shigemoto, R. Hirai, M. Yoneyama, M. Horiuchi, K. Ogura, T. Fujita, and F. Inagaki. 2009. Solution structures of cytosolic RNA sensor MDA5 and LGP2 C-terminal domains: identification of the RNA recognition loop in RIG-I-like receptors. J. Biol. Chem. 284:17465–17474.

145. Takahasi K., M. Yoneyama, T. Nishihori, R. Hirai, H. Kumeta, R. Narita, M. Gale, F. Inagaki, and T. Fujita. 2008. Nonself RNA-sensing mechanism of RIG-I helicase and activation of antiviral immune responses. Mol. Cell 29:428– 440.

146. Takaoka A., and H. Yanai. 2006. Interferon signalling network in innate defence. Cell. Microbiol. 8:907–922.

147. Takeuchi O., and S. Akira. 2009. Innate immunity to virus infection. Immunol. Rev. 227:75–86.

148. Tang E. D., and C.-Y. Wang. 2009. MAVS self-association mediates antiviral innate immune signaling. Journal of Virology 83:3420–3428.

149. Taylor M. P., and K. Kirkegaard. 2008. Potential subversion of autophagosomal pathway by picornaviruses. Autophagy 4:286–289.

150. Thomson S. J. P., F. G. Goh, H. Banks, T. Krausgruber, S. V. Kotenko, B.

129

M. J. Foxwell, and I. A. Udalova. 2009. The role of transposable elements in the regulation of IFN-lambda1 gene expression. Proc Natl Acad Sci USA 106:11564–11569.

151. Tracy S., Z. Tu, N. Chapman, and G. Hufnagel. 1995. Genetics of coxsackievirus B3 cardiovirulence. Eur. Heart J. 16 Suppl O:15–17.

152. Triantafilou K., G. Orthopoulos, E. Vakakis, M. A. E. Ahmed, D. T. Golenbock, P. M. Lepper, and M. Triantafilou. 2005. Human cardiac inflammatory responses triggered by Coxsackie B viruses are mainly Toll-like receptor (TLR) 8-dependent. Cell. Microbiol. 7:1117–1126.

153. Uzri D., and L. Gehrke. 2009. Nucleotide sequences and modifications that determine RIG-I/RNA binding and signaling activities. Journal of Virology 83:4174–4184.

154. van Pesch V., H. Lanaya, J. C. Renauld, and T. Michiels. 2004. Characterization of the Murine Alpha Interferon Gene Family. Journal of Virology 78:8219–8228.

155. Vance L. M., N. Moscufo, M. Chow, and B. A. Heinz. 1997. Poliovirus 2C region functions during encapsidation of viral RNA. Journal of Virology 71:8759–8765.

156. Venkataraman T., M. Valdes, R. Elsby, S. Kakuta, G. Caceres, S. Saijo, Y. Iwakura, and G. N. Barber. 2007. Loss of DExD/H box RNA helicase LGP2 manifests disparate antiviral responses. J Immunol 178:6444–6455.

157. Ventoso I., and L. Carrasco. 1995. A poliovirus 2A(pro) mutant unable to cleave 3CD shows inefficient viral protein synthesis and transactivation defects. Journal of Virology 69:6280–6288.

158. Wang J. P., A. Cerny, D. R. Asher, E. A. Kurt-Jones, R. T. Bronson, and R. W. Finberg. 2010. MDA5 and MAVS Mediate Type I Interferon Responses to Coxsackie B Virus. Journal of Virology 84:254–260.

159. Wang J. P., D. R. Asher, M. Chan, E. A. Kurt-Jones, and R. W. Finberg. 2007. Cutting Edge: Antibody-mediated TLR7-dependent recognition of viral RNA. J Immunol 178:3363–3367.

160. Wang J., R. Oberley-Deegan, S. Wang, M. Nikrad, C. J. Funk, K. L. Hartshorn, and R. J. Mason. 2009. Differentiated human alveolar type II cells secrete antiviral IL-29 (IFN-lambda 1) in response to influenza A infection. The Journal of Immunology 182:1296–1304.

161. Wang Q., D. R. Nagarkar, E. R. Bowman, D. Schneider, B. Gosangi, J. Lei, Y. Zhao, C. L. Mchenry, R. V. Burgens, D. J. Miller, U. Sajjan, and M. B. Hershenson. 2009. Role of Double-Stranded RNA Pattern Recognition

130

Receptors in Rhinovirus-Induced Airway Epithelial Cell Responses. The Journal of Immunology 183:6989–6997.

162. Wang Q., D. J. Miller, E. R. Bowman, D. R. Nagarkar, D. Schneider, Y. Zhao, M. J. Linn, A. M. Goldsmith, J. K. Bentley, U. S. Sajjan, and M. B. Hershenson. 2011. MDA5 and TLR3 Initiate Pro-Inflammatory Signaling Pathways Leading to Rhinovirus-Induced Airways Inflammation and Hyperresponsiveness. PLoS Pathogens 7:e1002070.

163. Weber F., V. Wagner, S. B. Rasmussen, R. Hartmann, and S. R. Paludan. 2006. Double-stranded RNA is produced by positive-strand RNA viruses and DNA viruses but not in detectable amounts by negative-strand RNA viruses. Journal of Virology 80:5059–5064.

164. West A. P., A. A. Koblansky, and S. Ghosh. 2006. Recognition and Signaling by Toll-Like Receptors. Annu. Rev. Cell Dev. Biol. 22:409–437.

165. Wilkins C., and M. Gale. 2010. Recognition of viruses by cytoplasmic sensors. Current Opinion in Immunology 22:41–47.

166. Wu L., and P. S. Morahan. 1992. Macrophages and other nonspecific defenses: role in modulating resistance against herpes simplex virus. Curr. Top. Microbiol. Immunol. 179:89–110.

167. Wurtmann E. J., and S. L. Wolin. 2009. RNA under attack: cellular handling of RNA damage. Crit. Rev. Biochem. Mol. Biol. 44:34–49.

168. Xu L.-G., Y.-Y. Wang, K.-J. Han, L.-Y. Li, Z. Zhai, and H.-B. Shu. 2005. VISA is an adapter protein required for virus-triggered IFN-beta signaling. Mol. Cell 19:727–740.

169. Yanai I., H. Benjamin, M. Shmoish, V. Chalifa-Caspi, M. Shklar, R. Ophir, A. Bar-Even, S. Horn-Saban, M. Safran, E. Domany, D. Lancet, and O. Shmueli. 2005. Genome-wide midrange transcription profiles reveal expression level relationships in human tissue specification. Bioinformatics 21:650–659.

170. Yang Y., Y. Liang, L. Qu, Z. Chen, M. Yi, K. Li, and S. M. Lemon. 2007. Disruption of innate immunity due to mitochondrial targeting of a picornaviral protease precursor. Proc Natl Acad Sci USA 104:7253–7258.

171. Yoneyama M., M. Kikuchi, K. Matsumoto, T. Imaizumi, M. Miyagishi, K. Taira, E. Foy, Y.-M. Loo, M. Gale, S. Akira, S. Yonehara, A. Kato, and T. Fujita. 2005. Shared and unique functions of the DExD/H-box helicases RIG-I, MDA5, and LGP2 in antiviral innate immunity. J Immunol 175:2851–2858.

172. Yoneyama M., M. Kikuchi, T. Natsukawa, N. Shinobu, T. Imaizumi, M. Miyagishi, K. Taira, S. Akira, and T. Fujita. 2004. The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral

131

responses. Nat. Immunol. 5:730–737.

173. Yoneyama M., K. Onomoto, and T. Fujita. 2008. Cytoplasmic recognition of RNA. Adv Drug Deliv Rev 60:841–846.

174. Yount J. S., L. Gitlin, T. M. Moran, and C. B. López. 2008. MDA5 participates in the detection of paramyxovirus infection and is essential for the early activation of dendritic cells in response to Sendai Virus defective interfering particles. J Immunol 180:4910–4918.

175. Yu S. F., and R. E. Lloyd. 1991. Identification of essential amino acid residues in the functional activity of poliovirus 2A protease. Virology 182:615–625.

176. Zeng W., L. Sun, X. Jiang, X. Chen, F. Hou, A. Adhikari, M. Xu, and Z. J. Chen. 2010. Reconstitution of the RIG-I pathway reveals a signaling role of unanchored polyubiquitin chains in innate immunity. Cell 141:315–330.

177. Zhang M., X. Wu, A. J. Lee, W. Jin, M. Chang, A. Wright, T. Imaizumi, and S.-C. Sun. 2008. Regulation of IkappaB kinase-related kinases and antiviral responses by tumor suppressor CYLD. J. Biol. Chem. 283:18621–18626.

178. Zhang Z., T. Kim, M. Bao, V. Facchinetti, S. Y. Jung, A. A. Ghaffari, J. Qin, G. Cheng, and Y.-J. Liu. 2011. DDX1, DDX21, and DHX36 Helicases Form a Complex with the Adaptor Molecule TRIF to Sense dsRNA in Dendritic Cells. Immunity 34:866–878.

179. Zhou A., J. M. Paranjape, B. A. Hassel, H. Nie, S. Shah, B. Galinski, and R. H. Silverman. 1998. Impact of RNase L overexpression on viral and cellular growth and death. J. Interferon Cytokine Res. 18:953–961.

180. Zhou A., J. Paranjape, T. L. Brown, H. Nie, S. Naik, B. Dong, A. Chang, B. Trapp, R. Fairchild, C. Colmenares, and R. H. Silverman. 1997. Interferon action and apoptosis are defective in mice devoid of 2“,5-”oligoadenylate- dependent RNase L. EMBO J. 16:6355–6363.

181. Zhou S., A. M. Cerny, A. Zacharia, K. A. Fitzgerald, E. A. Kurt-Jones, and R. W. Finberg. 2010. Induction and inhibition of type I interferon responses by distinct components of lymphocytic choriomeningitis virus. Journal of Virology 84:9452–9462.

182. Zoll J., F. J. van Kuppeveld, J. M. Galama, and W. J. Melchers. 1998. Genetic analysis of mengovirus protein 2A: its function in polyprotein processing and virus reproduction. J. Gen. Virol. 79 ( Pt 1):17–25.

183. Züst R., L. Cervantes-Barragan, M. Habjan, R. Maier, B. W. Neuman, J. Ziebuhr, K. J. Szretter, S. C. Baker, W. Barchet, M. S. Diamond, S. G. Siddell, B. Ludewig, and V. Thiel. 2011. Ribose 2'-O-methylation provides a molecular signature for the distinction of self and non-self mRNA dependent on

132

the RNA sensor Mda5. Nat. Immunol. 12:137–143.

184. http://ictvonline.org/virusTaxonomy.asp?version=2009&bhcp=1.

185. http://www.clinsci.org/cs/121/0415/cs1210415f01.htm?resolution=HIGH