The Detection of Ribonuclease Cleavage Sites

THE DETECTION OF RIBONUCLEASE CLEAVAGE SITES

IN HOST AND VIRAL RNAS DURING VIRAL INFECTIONS

by

DAPHNE A. COOPER

B.S., California Polytechnic State University, San Luis Obispo

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirements for the degree of

Doctor of Philosophy

Microbiology Program

2015

ii

This thesis for the Doctor of Philosophy degree by

Daphne A. Cooper

has been approved for the

Microbiology Program

by

Thomas E. Morrison, Chair

David J. Barton, Advisor

Linda F. van Dyk

Mario L. Santiago

Hugo R. Rosen

Jay R. Hesselberth

Date 5/06/2015

iii

Cooper, Daphne A (Ph.D., Microbiology)

The Detection of Ribonuclease Cleavage Sites in Host and Viral RNAs during Viral Infections

Thesis directed by Professor David J. Barton.

ABSTRACT

Ribonucleases are critical components for cellular responses to viruses. RNase L

is an endoribonuclease activated during many types of viral infections, yet the RNAs

targeted by RNase L to promote its antiviral activities are poorly characterized.

To better understand the RNAs cleaved by RNase L during viral infections, I

optimized and validated cDNA synthesis and deep sequencing methods that enrich for

RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates. RNase L, in addition to many

other ribonucleases, produces RNA cleavage fragments with terminal 2ˊ, 3ˊ-cyclic phosphates, making 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing methods broadly applicable for the study of many types of ribonucleases. Using these methods, I characterized RNA cleavage from cells infected with poliovirus, influenza A virus, hepatitis C virus, and from liver tissue from a hepatitis C virus-infected patient.

Deep sequencing of 2ˊ, 3ˊ-cyclic phosphate cDNA libraries revealed that RNase

L activation during poliovirus and influenza A virus infections provoked cleavage of host and viral RNAs. Two sites in 18S rRNA, UU541 and UU743, were among the most

frequently detected RNase L-dependent cleavage sites. Cleavage at these sites

potentially inhibits protein synthesis and promotes apoptotic cell death of virally-infected

cells. Although these studies focused on RNase L, I also detected RNA cleavage by other ribonucleases. Usb1, a nuclear 3ˊ→5ˊexoribonuclease, post-transcriptionally modifies the 3ˊ-end of U6 snRNA to generate a 2ˊ, 3ˊ-cyclic phosphate frequently detected in 2ˊ, 3ˊ-cyclic phosphate cDNA libraries. 2ˊ, 3ˊ-cyclic phosphates were also iv

detected at the end of 5S rRNA, indicating that 5S rRNA might be post-transcriptionally

modified in a manner similar to U6 snRNA. Finally, cleavage sites in tRNAs, rRNAs, and

U3 snoRNA, consistent with the activity of angiogenin and other RNase A family

endoribonucleases, were detected in hepatitis C virus-infected cells and tissues.

To conclude, the data described in this thesis define some of the RNA substrates

of RNase L and other ribonucleases during viral infections. 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing reveal the frequency, location, and specificity of ribonuclease cleavage sites in RNAs within cells, revealing, in part, how these enzymes contribute to health and disease.

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

Approved: David J. Barton

v

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to my advisor, Dave Barton, for his guidance, support, and encouragement during my training period within his lab. I am very grateful to have had the opportunity to work with Dave, where under his guidance I have developed critical skills to help me during my professional career.

This thesis research required a great deal of bioinformatic data processing, and almost everything I know in that regard is because of Jay Hesselberth. I am extremely grateful for all of his help and guidance during the course of my thesis work. Jay also authored several programs that I used to analyze data presented in this thesis.

I am also very thankful for my thesis committee, who provided me with encouragement and invaluable ideas and suggestions over the years to keep my research progress on the right track.

I would also like to recognize Shuvojit Banerjee from Robert Silverman’s Lab at

Cleveland Clinic, Katelyn Leahy, formerly of Hugo Rosen’s Lab at University of Colorado

Anschutz Medical Campus, Andrew Firth of University of Cambridge, and Ann

Palmenberg of University of Wisconsin, Madison, for their contributions to the data presented in this thesis.

I would also like to thank Brian Kempf of the Barton Lab. Without Brian, the lab would have been a very lonely place. I also want to thank Brian for his feedback and suggestions while I wrote this thesis.

Finally, I want to thank my boyfriend, Jon Mann, and my family and friends for their support and encouragement over the past six years.

vi

TABLE OF CONTENTS

CHAPTER

I. INTRODUCTION…………………………………………...... 1

2ˊ-5ˊ Oligoadenylate Synthetase (OAS)/RNase L system...... 3

Angiogenin and other Ribonuclease A Family Enzymes…………...... …21

IRE1 and the Unfolded Protein Response...... 29

RNase T2, An Enzyme Involved in rRNA Processing………………...... 32

Usb1, A 3ˊ→5ˊ exonuclease Important for Regulating the Length of U6 snRNA…………………………………………...... …33

Mechanisms of RNA Cleavage………………………………...... …… 35

Scope of Thesis………………………...... …38

II. METHODS TO DETECT RIBONUCLEASE L CLEAVAGE IN RNA…...... 40

Introduction………………………………………….……………………...... 40

Materials and Methods…………………………………….……………...... … 43

Results……………………………………………..…………………………...... 49

Discussion…………………………………………………………………...... … 70

III. POLIOVIRUS INFECTION: RIBONUCLEASE CLEAVAGE SITES IN HOST AND VIRAL RNAS……………………..………………...... … 76

Introduction………………………………...... 76

Materials and Methods……………………...... 79

Results…………………………………………..…...... 81

Discussion……………………………….……...... 99

IV. INFLUENZA A VIRUS INFECTION: RIBONUCLEASE CLEAVAGE SITES IN HOST AND VIRAL RNAs……...………………………………...... 107

Introduction…..………………………………….……………………………...... 107

Materials and Methods………………………………………………………...... 110

Results………………………..……………….………………………………...... 114 vii

Discussion…………………………...... 127

V. HEPATITIS C VIRUS INFECTION: RIBONUCLEASE CLEAVAGE SITES IN HOST AND VIRAL RNAS……………………………………….….....…135

Introduction……..………………...... …135

Materials and Methods………………...... 138

Results……………………...... 140

Discussion……..………………....…………...... 155

VI. SUMMARY AND FUTURE DIRECTIONS………………………...... 163

Characteristics of RNA Fragments Produced by RNase L Cleavage………...... 165

Impact of RNA Cleavage by RNase L……………………...... 166

Terminal 2ˊ, 3ˊ-Cyclic Phosphates on 5S rRNA…………...... 172

Angiogenin Stress Response………………………………...... 175

The Intersection of RNase L and Angiogenin…..…………...... 177

REFERENCES………………………………………………………...... 180 viii

LIST OF TABLES

TABLE

1.1 The function and specificity of some of the RNases expressed by humans that produce RNA fragments with terminal 2ˊ, 3ˊ -cyclic phosphate...... 6

1.2 Viral countermeasures to RNase L...... 18

2.1 TOPO-TA cloning of 2ˊ, 3ˊ-cyclic phosphate cDNA libraries from HCV RNA cleaved by RNase L…...... 53

2.2. 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing: HCV RNA...... 59

2.3 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing: PV RNA...... 60

3.1 Frequency of 2ˊ, 3ˊ-cyclic phosphates in host and viral RNAs from M25 HeLa cells...... 83

3.2 Frequency of 2ˊ, 3ˊ-cyclic phosphates in host and viral RNAs from W12 HeLa cells……...... 83

3.3 RNase L -dependent and -independent cleavage sites in 28S rRNA from W12 and M25 HeLa cells...... 88

3.4 RNase L -dependent and -independent cleavage sites in 18S rRNA from W12 and M25 HeLa cells...... 91

4.1 Ribonuclease cleavage of IAV RNAs...... 117

4.2 RNase L-dependent and -independent cleavage in 28S rRNA from A549 cells...... 124

4.3 RNase L -dependent and -independent cleavage in 18S rRNA from A549 cells………...... 125 ix

LIST OF FIGURES

FIGURE

1.1 2-5A and RNase L……………………...... 4

1.2 Viral dsRNA-induced type I IFN signaling………………...... 11

1.3 Mechanisms of RNA cleavage by ribonucleases...... 37

2.1 RNase L generates RNA cleavage fragments with terminal 2ˊ, 3ˊ-cyclic phosphates…………………...... 50

2.2 RNase L cleavage sites detected from TOPO-TA cloning of 2ˊ, 3ˊ-cyclic phosphate cDNA libraries...... 55

2.3 Viral RNA fragments produced by RNase L and RNase A...... 57

2.4 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods...... 58

2.5 Endoribonuclease cleavage sites in HCV RNA ………...... 61

2.6 Endoribonuclease cleavage sites in PV RNA……...... 62

2.7 Specificity of RNase L and RNase A…...... 63

2.8 HCV RNA secondary structures associated with RNase L and RNase A cleavage sites...... 65

2.9 svRNA3 is liberated by RNase L cleavage...... 67

2.10 Endoribonuclease susceptible regions of PV RNA...... 68

2.11 Antibody neutralization escape mutations within and near endonuclease susceptible regions of PV RNA…...... 69

3.1 PV infection of in W12 and M25 HeLa cells…………...... 82

3.2 Frequency of ribonuclease cleavage sites in host and viral RNAs from M25 and W12 HeLa cells…………...... 83

3.3 Ribonuclease cleavage sites in PV RNA……...... 85

3.4 Ribonuclease cleavage sites in 28S rRNA……...... 89

3.5 Ribonuclease cleavage sites in 18S rRNA……...... 92

3.6 Ribonuclease cleavage sites in 5.8S rRNA……...... 94

3.7 Ribonuclease cleavage sites in 5S rRNA...... 95 x

3.8 Location of RNase L-dependent cleavage sites in 80S ribosomes……….….... 96

3.9 Ribonuclease cleavage sites in U6 snRNA...... 97

3.10 Ribonuclease cleavage sites in FSCN1 mRNA………...... 98

3.11 Proximity of ribosomal protein rpS7 to the RNase L-dependent cleavage site, UU743…………………………………………………..…………....102

4.1 NS1 prevents RNase L activation during IAV infections of A549 cells………………………………………………………………………....115

4.2 Frequency of ribonuclease cleavage sites in host and viral RNAs from A549 cells……..……………………………………………………………...... 116

4.3 Dinucleotides at ribonuclease cleavage sites in IAV RNAs...... 118

4.4 Dinucleotides at ribonuclease cleavage sites in IAV RNAs (genome-wide)…………………………………………………...... 119

4.5 Frequency and location of ribonuclease cleavage sites in IAV RNAs in relation to synonymous site conservation………………...... 121

4.6 Prominent cleavage sites in M (+) strand RNA………………….……………... 122

4.7 RNase L-dependent cleavage sites in ribosomal RNA and 80S ribosomes……...... 126

4.8 RNase L-dependent cleavage sites in U3 snoRNA…………...... 127

5.1 Accumulation of HCV genomic RNA in HCV-infected Huh7.5.1 cells………………………………….…………………………………....141

5.2 Frequency of ribonuclease cleavage sites in host and viral RNAs from Huh7.5.1 cells ……………………………………………………………...... 141

5.3 Anticodon loop cleavage in tRNAs from HCV-infected Huh7.5.1 cells…...... 143

5.4 Ribonuclease cleavage sites in tRNA-Glu-CUC………...... 143

5.5 Ribonuclease cleavage sites in JFH1 HCV2A RNA ……...... 145

5.6 Ribonuclease cleavage sites in 28S rRNA …………………………………...... 147

5.7 Ribonuclease cleavage sites in 18S rRNA …………………………………...... 148

5.8 Ribonuclease cleavage sites in 5.8S and 5S rRNAs…………………………....149

5.9 Location of ribonuclease cleavage sites in 80S ribosomes ………...... 150 xi

5.10 Ribonuclease cleavage sites in U3 snoRNA …………...... 152

5.11 Anticodon loop cleavage in tRNAs from HCV-infected liver………...... 153

5.12 Ribonuclease cleavage sites in 28S rRNA, 18S rRNA, 5S rRNA, and U3 snoRNA from liver of HCV1A-infected patient ……………………...... 154 xii

LIST OF ABBREVIATIONS

2-5A 2ˊ-5ˊ Oligoadenylate

A. th. Arabidopsis thaliana

ALS Amyotrophic Lateral Sclerosis

CARD Caspase Activation and Recruitment Domain ciRNA Competitive Inhibitor RNA for RNase L

DN Dominant Negative dsRNA Double-Stranded RNA

EMCV Encephalomyocarditis Virus

HBV Hepatitis B Virus

HCV Hepatitis C Virus

IAV Influenza A Virus

IFN Interferon

IRE1 Inositol-Requiring Enzyme 1

ISG Interferon-Stimulated Gene

ISRE Interferon-Stimulated Response Element

JNK c-Jun N-Terminal Kinase

KEN Kinase Extension Nuclease

LGP2 Laboratory of Genetics and Physiology 2

MAVS Mitochondrial Antiviral Signaling Protein

MDA5 Melanoma Differentiation-Associated Gene 5 mRNA Messenger RNA

NS1 IAV Nonstructural Protein 1

OAS 2ˊ-5ˊ Oligoadenylate Synthetase

P:C:I Phenol:Chloroform:Isoamyl Alcohol

PAMP Pathogen Associated Molecular Pattern xiii

PKR dsRNA-Activated Protein Kinase

PN Poikiloderma with Neutropenia

Poly(I:C) Polyinosinic-Polycytidylic Acid

PRR Pattern Recognition Receptor

PV Poliovirus rDNA Ribosomal DNA

RIDD Regulated IRE1-Dependent Decay

RIG-I Retinoic Acid Inducible Gene-I

RLR RIG-I-Like Receptor

RNase Ribonuclease

RNP Ribonucleoprotein rRNA Ribosomal RNA

RSV Respiratory Syncytial Virus snoRNA Small Nucleolar RNA snRNA Small Nuclear RNA ssRNA Single-Stranded RNA svRNA3 Suppressor of Virus RNA 3

T4 PNK T4 Polynucleotide Kinase tiRNA tRNA-Derived Stress-Induced RNA

Tpt1 Yeast 2ˊ Phosphotransferase

UPR Unfolded Protein Response

WT Wild-Type

1

CHAPTER I

INTRODUCTION

Ribonucleases are protein or nucleic acid-based enzymes that cleave the

phosphodiester bond between each ribonucleotide in RNA. Cleavage of RNA is

essential for cellular functions like DNA transcription, mRNA translation, response to environmental stressors and defense against invading pathogens. Ribonuclease L

(RNase L) is often referred to as the antiviral ribonuclease. Double-stranded RNA

(dsRNA), produced during many types of viral infections, activate the interferon(IFN)- inducible oligoadenylate synthetase enzymes (OAS), which in turn produce 2ˊ-5ˊ linked oligoadenylates (2-5A) that bind to and activate RNase L. Activated RNase L cleaves

viral and host RNAs to: 1) promote apoptosis of the virally infected cell (Castelli et al.,

1998; Le Roy et al., 2007; Zhou et al., 1997); 2) produce RNA fragments that activate

RIG-I and MDA5 to induce expression of type I IFN (Malathi et al., 2007, 2010); and 3)

cleave host and viral RNA to inhibit protein synthesis (Clemens and Williams, 1978;

Hovanessian et al., 1977; Kerr and Brown, 1978; Li et al., 1998). In addition to its roles in

antiviral defense, RNase L is also important in antibacterial defense, prostate cancer,

chronic fatigue syndrome, and aging and senescence (Andersen et al., 2007; Demettre

et al., 2002; Li et al., 2008; Silverman, 2003). Despite its extensive characterization and

importance in health and disease, the RNAs cleaved by RNase L are poorly defined.

Information on the specificity and RNA substrates of other cellular ribonucleases

are also sparse, despite greater evidence that these RNA degrading enzymes have

important functions in health. Angiogenin, for example, is an RNase A family enzyme

highly expressed in many types of cancers (Hisai et al., 2003; Tello-Montoliu et al.,

2006). Angiogenin promotes blood vessel growth (angiogenesis), where its nuclease

activity is required. Mutations that inactivate the ribonuclease ability of angiogenin are

associated with amyotrophic lateral sclerosis (ALS) (Shapiro and Vallee, 1989; Shapiro 2

et al., 1989; Wu et al., 2007). Despite the importance of RNA cleavage by angiogenin,

the only defined RNA substrates of angiogenin are tRNAs, which angiogenin targets

during cellular stress (Fu et al., 2009a; Yamasaki et al., 2009). Another ribonuclease,

RNase T2, is implicated in familial cystic leukoencephalopathy, where loss of function

mutations in the enzyme lead to the accumulation of ribosomal RNA (rRNA) in

lysosomes within neurons (Haud et al., 2011). Very little is understood about the

specificity of RNase T2, and whether it targets other cellular RNAs other than rRNA.

Initially, the purpose of the study was to develop better methods to detect the

activity of RNase L during viral infections. To this end, I have adapted and optimized a

cDNA synthesis and deep sequencing method that detects RNAs with terminal 2ˊ, 3ˊ-

cyclic phosphate modifications (referred to as 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing) (Schutz et al., 2010). cDNA libraries from cells and tissues can be generated to detect the location, frequency, and specificity of cleavage sites in RNAs

targeted by specific ribonucleases. Although the majority of this thesis focuses on RNAs

cleaved by RNase L during viral infections, many ribonucleases cleave RNA to produce

RNA fragments with 2ˊ, 3ˊ-cyclic phosphate modifications (Table 1.1), making these

methods broadly applicable for studying many types of ribonucleases from different

experimental conditions.

This thesis addresses some of the aforementioned knowledge gaps of the RNA

targets of RNase L and other ribonucleases. The remainder of this chapter is a review of

several ribonucleases known to be important for health and disease, and the pathways

they are involved in. In the first data chapter, Chapter II, I describe the optimization of

2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and sequencing methods for the purpose of

detecting ribonuclease cleavage sites in RNA. Using the methods described in Chapter

II, I identified RNase L-dependent and independent cleavage sites in host and viral

RNAs during poliovirus (PV)-infection of HeLa cells (Chapter III). The RNase L- 3

dependent and independent cleavage sites detected in PV-infection of HeLa cells were

directly comparable with RNase L-dependent and independent cleavage sites detected

in influenza A virus (IAV)-infected A549 cells (Chapter IV), reinforcing the validity of these methods. In Chapter V, I use these methods to detect RNA cleavage consistent

with an angiogenin stress response in hepatitis C virus (HCV)-infected Huh7.5.1 cells and in HCV-infected patient liver. Finally in Chapter VI, I summarize the discoveries made using 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing methods, and discuss the broader impact of these studies and future directions to further define the importance of ribonucleases in regulating health and disease.

2′, 5′-Oligoadenylate Synthetase (OAS)/RNase L System

rRNA fragmentation and the inhibition of protein synthesis in dsRNA-treated cell

free lysates from IFN-treated cells led to the discovery of RNase L and the IFN-inducible

oligoadenylate synthetase (OAS) enzymes that synthesize the 2-5A activator of RNase L

(Figure 1.1A) (Clemens and Williams, 1978; Hovanessian et al., 1977; Kerr and Brown,

1978; Roberts et al., 1976; Wreschner et al., 1981a). The gene for RNase L is located

on chromosome 1, and encodes a protein of 741 amino acids in length. RNase L is

divided into an N-terminal regulatory domain with nine ankyrin repeats, a kinase-like

domain, and the endonuclease domain (Figure 1.1B) (Dong and Silverman, 1999). 2-5A

binding occurs within the ankyrin repeat region of RNase L (Figure 1.1B and C;

monomer) to induce conformational changes that facilitate dimerization, and potentially

higher-order homo-oligomers that activate RNase L (Figure 1.1C; dimer) (Cole et al.,

1996; Dong and Silverman, 1995; Han et al., 2012, 2014; Huang et al., 2014; Han et al.,

2014). Activated RNase L cleaves single-stranded RNA (ssRNA) 3ˊ of UN

dinucleotides, with a strong preference for UA and UU dinucleotides (Table 1.1) (Floyd-

4

A. B.

C.

Figure 1.1. 2-5A and RNase L. A. 2-5A activator produced by OAS enzymes in response to dsRNA. B. Diagram of RNase L (adapted from Chakrabarti et al., 2011). RNase L, a protein of 741 amino acids in length, has nine ankyrin repeats, a kinase- like domain, and an RNase domain. C. Structure of monomer unit of porcine RNase L bound to 2-5A (red). Coloring of domains are consistent with the diagram in (B). AMP-PNP (pink) is bound to the pseudokinase domain. Dimer conformation of RNase L results from 155° rotation across the ankyrin and protein kinase-like domain. One chain of RNase L is colored gold, and the other blue. Dimer structure is rotated relative to the monomer to highlight the interaction between the two chains of RNase L to form the dimer. Porcine RNase L, PDB 4O10 (Huang et al., 2014). Structures represented using PyMOL (DeLano Scientific, San Carlos, California, USA; http://www.pymol.org).

Smith et al., 1981; Wreschner et al., 1981b). rRNA, cellular mRNA, and viral RNA are all

reported targets of RNase L (Al-Ahmadi et al., 2009; Girardi et al., 2013; Han et al.,

2004; Li et al., 1998, 2000; Malathi et al., 2010; Nilsen and Baglioni, 1979; Silverman et

al., 1983; Wreschner et al., 1981a). RNase L is important for the initiation of apoptosis, inhibition of protein synthesis, and the generation of ligands for pattern recognition

receptors (PRRs) retinoic acid inducible gene-I (RIG-I) and melanoma differentiation-

associated protein 5 (MDA5) to induce and maintain the expression of type I IFN

(Clemens and Williams, 1978; Kerr and Brown, 1978; Malathi et al., 2007, 2010; Le Roy

et al., 2007; Rusch et al., 2000; Zhou et al., 1997). 5

Although most well-known for its roles in antiviral defense, RNase L is also important for cancer, aging and senescence, chronic fatigue syndrome, and antibacterial defenses (Andersen et al., 2007; Demettre et al., 2002; Li et al., 2008; Silverman, 2003;

Suhadolnik et al., 1997) . A mutation in RNase L, R462Q, results in reduced ribonuclease activity and is associated with susceptibility to hereditary prostate cancer

(Casey et al., 2002; Rökman et al., 2002; Silverman, 2003; Xiang et al., 2003). In patients with chronic fatigue syndrome, a proteolytically digested variant of RNase L is often found in peripheral blood mononuclear cells, and at one time was under consideration as a diagnostic indicator of the disease (Frémont et al., 2005). RNase L is also involved in aging. Mice deficient in RNase L lived 32% longer than their strain- matched counterparts expressing RNase L (Andersen et al., 2007). In bacterial infections, RNase L is important for the induction of proinflammatory cytokines necessary for control of infections caused by Escherichia coli and Bacillus anthracis (Li et al., 2008). In an experimental colitis model in mice, RNase-L-dependent production of

IFN-β and other proinflammatory cytokines in response to bacterial RNA was suggested to be important for protection against colitis (Long et al., 2013).

RNase L is constitutively expressed in many different tissue types, with high levels detected in lymph nodes, colon, skeletal muscle, and liver (Squire et al., 1994;

Zhou et al., 1993b, 2005). Because the promoter for RNASEL does not contain an interferon stimulated response element (ISRE), RNase L is not considered an interferon stimulated gene (ISG); however, the expression of RNase L is enhanced by IFNα and

IFNγ (in some cases), and various stress-inducing agents like hydrogen peroxide, tumor necrosis factor (TNF), and ribotoxins that damage rRNA (Chase et al., 2003; Floyd-

Smith, 1988; Li and Pestka, 2008; Pandey et al., 2004; Zhou et al., 1997).

6

Table 1.1. The function and specificity of some of the RNases expressed by humans that produce RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates

Cleavage Name Function Specificity References Products

ssRNA 2ˊ, 3ˊ>P Chakrabarti et al., 2011; IFN-mediated RNase L (Endo) 3ˊ of UA & UU > 3ˊ-P Floyd-Smith et al., 1981; antiviral defenses UG dinucleotides 5ˊ-OH Wreschner et al., 1981b

Antibacterial ssRNA, dsRNA 2ˊ, 3ˊ>P RNase 1 (Endo) Landré et al., 2002; defense; Dendritic 3ˊ of pyrimidines 3ˊ-P or 2ˊ-P (Pancreatic RNase) Sorrentino et al., 2003 cell maturation C > U 5ˊ-OH

ssRNA Domachowske et al., 2ˊ, 3ˊ>P RNase 2/EDN (Endo) Antiviral defense 3ˊ of pyrimidines 1998; Rugeles et al., 5ˊ-OH U > C 2003; Yang et al., 2004

ssRNA Domachowske et al., 2ˊ, 3ˊ>P RNase 3/ECP (Endo) Antiviral defense 3ˊ of pyrimidines 1998; Sorrentino and 5ˊ-OH U > C Glitz, 1991

ssRNA 2ˊ, 3ˊ>P Hofsteenge et al., 1998; Neuronal RNase 4 (Endo) 3ˊ of pyrimidines 3ˊ-P or 2ˊ-P Li et al., 2013; Shapiro et protection U > C 5ˊ-OH al., 1986

Curran et al., 1993; Fett ssRNA RNase 5/Angiogenin Angiogenesis; Cell 2ˊ, 3ˊ>P et al., 1985; Fu et al., 3ˊ of pyrimidines (Endo) survival 5ˊ-OH 2009; Yamasaki et al., C > U 2009

Antibacterial RNase 6 (Endo) N.D. N.D. Becknell et al., 2014 defense

Antibacterial Harder and Schröder, RNase 7 (Endo) N.D. N.D. defense 2002; Zhang, 2003

Antibacterial Rudolph et al., 2006; RNase 8 (Endo) N.D. N.D. defense Zhang et al., 2002

Hollien et al., 2009; ssRNA 2ˊ, 3ˊ>P IRE1α (Endo) UPR; RIDD Sidrauski and Walter, CNGNNGN 5ˊ-OH 1997

Mucin production; ssRNA 2ˊ, 3ˊ>P Iwawaki et al., 2001; IRE1β (Endo) Translation CNGNNGN 5ˊ-OH Martino et al., 2012 regulation ˊ ssRNA 2ˊ, 3ˊ>P Haud et al., 2011; RNase T2 (Endo) rRNA turnover poly(U); poly(A) 5ˊ-OH Thorn et al., 2012

U6 snRNA ssRNA Mroczek et al., 2012; hUsb1/hMpn1 (Exo) 2ˊ, 3ˊ>P maturation poly(U); poly(A) Shchepachev et al., 2012

ssRNA Likely 2ˊ, 3ˊ>P Economopoulou et al., RNase κ (Endo) N.D. 3ˊ of A > U 5ˊ-OH 2007; Kiritsi et al., 2012

ssRNA 2ˊ, 3ˊ>P Laneve et al., 2008; Poe EndoU/PP11 (Endo) B cell maturation poly(U) 5ˊ-OH et al., 2014 Endo:Endoribonuclease, Exo:Exoribonuclease, IFN:Interferon, UPR:Unfolded Protein Response, RIDD:Regulated IRE1-Dependent Decay of mRNA, rRNA:Ribosomal RNA, snRNA:Small Nuclear RNA, ssRNA:Single-Stranded RNA, dsRNA:Double-Stranded RNA, 2ˊ, 3ˊ>P: 2ˊ, 3ˊ -Cyclic Phosphate, 3ˊ-P:3ˊ- Phosphate, 2ˊ-P:2ˊ-Phosphate, 5ˊ-OH:5ˊ-Hydroxyl, N.D.:No Data 7

RNase L activation is dependent on the activity of the IFN-inducible OAS enzymes. In response to dsRNA, OAS enzymes synthesize 2ˊ-5ˊ-linked oligoadenylates of the formula n>1 (2-5A) (Figures 1.1A) (Hovanessian and Justesen, 2007). OAS

enzymes are related to other nucleotidyl transferases including CCA-adding enzymes,

poly-A polymerase, and cyclic GMP-AMP synthase (Kjaer et al., 2009). Prior to the

discovery of cyclic GMP/AMP synthase, an enzyme that synthesizes c[G(2ˊ-5ˊ)A(3ˊ-5ˊ)p] in response to cytoplasmic double-stranded DNA (Gao et al., 2013), OAS was the only known enzyme to polymerize nucleotides in the unusual 2ˊ-5ˊ orientation.

Humans express four forms of OAS: OAS1, 2, 3, and a catalytically inactive

OAS-like (OASL) protein. The genes encoding OAS1, 2, 3, and OASL are located at the

OAS-locus on chromosome 12, controlled by a promoter containing an ISRE, making

them ISGs (Benech et al., 1987; Hovnanian et al., 1998; Justesen et al., 2000). OAS1

has one OAS domain, whereas OAS2 and OAS3 have two and three OAS domains,

respectively (Benech et al., 1985; Marie and Hovanessian, 1992; Rebouillat et al., 1998).

It is important to note that in both OAS2 and OAS3, only the C-terminal OAS domain is

catalytically active (Donovan et al., 2015; Ibsen et al., 2014). OAS3 has higher affinity for

dsRNA than OAS1 and 2 due to the multivalent nature of repeated OAS domains, and is

reported to support binding of longer dsRNA due to the interactions of the RNA substrate

with the noncatalytic OAS domains (Donovan et al., 2015; Ibsen et al., 2014). Different

isoforms of OAS1 and OAS2 occur through alternative splicing, and certain SNPs that

influence splice variants of the OAS1 gene are associated with susceptibility to West

Nile virus and Dengue virus infections, responses to rubella vaccination, susceptibility to

diabetes, as well as outcomes of HCV infection and success of IFN-based therapies for

HCV infection (El Awady et al., 2011; Field et al., 2005; Haralambieva et al., 2010;

Knapp et al., 2003; Li et al., 2009; Lim et al., 2009; Lin et al., 2009). 8

Structural studies of OAS1 indicate that a minimum dsRNA duplex length of 17

bps is necessary to span the dsRNA binding surface of OAS1 (Donovan et al., 2013).

The preference of OAS1 for a specific sequence motif, NNWWNNNNNNNNNWGN-3ˊ,

was determined through in vitro studies using chemically synthesized dsRNAs and

purified OAS1 (Kodym et al., 2009). The G residue near the 3ˊ-end of the duplex RNA is

important for direct interactions with OAS1 (Donovan et al., 2013; Kodym et al., 2009).

Recently, features of dsRNA that enhance the activity of OAS1 have been discerned and

include pyrimidine overhangs 3ˊ of the preferred sequence motif (Vachon et al., 2014).

Conformational changes in the OAS enzymes in response to dsRNA binding likely

facilitate their downstream production of 2-5A (Donovan et al., 2013).

Once bound to and activated by dsRNA, OAS1 forms tetramers, OAS2 dimers,

and OAS3 monomers (Marie et al., 1990). OAS enzymes are non-processive, and it is

suggested that oligomerization of OAS1 and OAS2 juxtapose catalytic cores to

overcome the non-processive nature of 2-5A polymerization (Justesen et al., 1980). The

function of 2-5A is to activate RNase L, although RNase L activation requires that the 2-

5A molecules retain their 5ˊ-phosphate groups and contain at least three adenylates (2-

5A trimer) (Figure 1.1A) (Dong and Silverman, 1995; Dong et al., 1994). When RNase L is activated by 2-5A, it cleaves host and viral RNAs to limit viral infections. However,

OAS enzymes have RNase L-independent antiviral activities. Serum levels of OAS from patients undergoing IFNα-based treatment for HCV correlate with treatment success

(Kim et al., 2006; Shindo et al., 2008). Recombinant exogenous porcine OAS1 is taken up by IFN-deficient Vero cells to inhibit encephalomyocarditis virus (EMCV) by an unknown mechanism independent of IFN and RNase L (Kristiansen et al., 2010). In addition, an OAS1 splice variant, p40, inhibits HCV replication in tissue culture by inhibiting viral mRNA translation in an RNase L-independent manner (Erickson, 2008). 9

The 2ˊ-5ˊ OAS/RNase L pathway is important for induction of type I IFN as well as mediating the antiproliferative and antiviral effects of type I IFN. The relationship between IFN and OAS, 2-5A and RNase L were first observed in 1971 when cell-free lysates were inoculated with heat-treated cytoplasmic extracts from PV- infected cells

(Ehrenfeld and Hunt, 1971). Heat inactivated PV extracts provoked the production of a substance in the cell-free system that inhibited protein synthesis. Further investigation revealed that a low molecular weight inhibitor of protein synthesis formed when extracts from IFN-treated cells were incubated with dsRNA (Roberts et al., 1976), indicating an

IFN-inducible factor was synthesizing the low-weight molecular inhibitor in response to dsRNA and ATP. Using synthetic dsRNA, polyinosinic:polycytidylic acid (polyI:C), bound to sepharose, the IFN-inducible factor was purified from IFN-treated cells (Hovanessian et al., 1977). Incubation of the purified enzyme with ATP resulted in synthesis of the low molecular weight inhibitor, and when added to cell-free lysates, the in vitro synthesized inhibitor repressed protein synthesis from EMCV RNA (Hovanessian et al., 1977).

Digestion of the low molecular weight inhibitor with various nucleases indicated that it was not linked by 3ˊ-5ˊ phosphodiester bonds found in most oligonucleotides. Thin-layer chromatography of the products resulting from enzymatic digestion and periodate and β- elimination of the inhibitor confirmed that these low molecular weight inhibitors of protein synthesis were 2ˊ, 5ˊ-linked oligoadenylates (Kerr and Brown, 1978).

2′, 5′-OAS/RNase L System: Induction of Interferon

Type I IFN was discovered as a substance produced by incubating heat inactivated influenza virus with chick chorioallantoic membranes. This substance

“interfered” with virus infection of uninfected membranes (Isaacs and Lindenmann,

1957). Type I IFNs include 13 IFNα isoforms, IFNβ, IFNδ, IFNε, IFNκ, IFNτ, and IFNω, and are secreted from virally infected cells where they bind to and activate the type I IFN 10 receptors (IFNAR1 and IFNAR2). The activated IFNAR receptor complex signals through the Janus kinase (JAK)/signal transducer and activation of transcription (STAT) pathway to stimulate the expression of hundreds of ISGs (Der et al., 1998; Platanias,

2005). ISGs encode proteins important for mediating the antiviral effects of IFN. OAS enzymes and dsRNA-dependent protein kinase (PKR) are two ISGs that contribute to inhibition of protein synthesis in IFN treated cell lysates (Borden et al., 2007;

Hovanessian et al., 1977; Platanias, 2005; Wreschner et al., 1981a).

Like type I IFNs, type III IFNs are also induced by viral infection. Type III IFNs include IFNλ1-4, and stimulate the expression of largely the same set of ISGs as type I

IFN (Levy et al., 2011; Zhou et al., 2007). Although type I and type III IFN lead to very similar responses in cells, type III IFN mediates its functions through a different cell receptor, the IFN lambda receptor (IFNλR), composed of IL28Rα and IL10Rβ (Odendall and Kagan, 2015; Platanias, 2005). Studies that differentiate between the effects of type

I and type III IFN indicate that while IFNAR is expressed on nearly all cell types, IFNλR expression is restricted to epithelial and liver cells, suggesting that type III IFNs are important for the regulation of antiviral responses at mucosal surfaces and within the liver (Ank et al., 2006; Okabayashi et al., 2011; Sommereyns et al., 2008; Zhou et al.,

2007). In addition to type I and type III IFNs, type II IFN (IFNγ) is an important cytokine that has some functional overlap with type I and type III IFNs. IFNγ is predominantly produced by immune cells and is important for increasing the expression of MHC class I and II, among other functions, to bridge innate immune responses with adaptive immune responses (Der et al., 1998; Schroder et al., 2004).

Induction of IFN expression during viral infections is triggered by sensing of

pathogen associated molecular patterns (PAMPs) by host cell PRRs (Figure 1.2). Viral

nucleic acids, especially dsRNA, are potent PAMPs produced during many types of viral

infections. While some viruses have genomes composed of dsRNA, positive-strand RNA 11

Figure 1.2. Viral dsRNA-induced type I IFN signaling. Viral dsRNA activates cytoplasmic RIG-I/MDA5, and endosomal TLR3. OAS enzymes recognize dsRNA to produce 2-5A that activates RNase L. RNase L can cleave RNA to generate PAMPs recognized by RIG-I/MDA5. Activated RIG-I/MDA5 interacts with MAVS. TLR3- activated TRIF and MAVS induce signaling pathways that result in IRF3 and NFkB activation. NFkB and IRF3 translocate to the nucleus to activate expression of type I IFN (IFNα/β). IFNα/β is secreted from the infected cell, and binds to its receptor (IFNAR) on a neighboring cell, but can also bind autocrine IFNAR. Janus-activated kinase (JAK) and tyrosine kinase (TYK2) dock onto the cytoplasmic tails of the activated IFNAR subunits and recruit and phosphorylate STAT-1 and STAT-2 to form heterodimers. STAT-1:STAT-2 heterodimers associate with IRF9 to form ISGF3. ISGF3 binds to ISREs present in the promoter regions ISGs that include OAS, PKR, RIG-I, MDA5, TLR3, and MAVS, among hundreds of others. The increased expression of these ISGs provokes an antiviral state, so that the cell can readily respond to and restrict viral infections. viruses replicate though dsRNA intermediates in the cytoplasm and can harbor dsRNA structure. In addition, some DNA viruses encode bidirectional genes that when co- expressed can lead to the formation of dsRNA (Weber et al., 2006). Other viral PAMPs that induce type I IFN expression include single-stranded RNA sensed by endosomal toll-like receptors (TLR) 7 and 8; unmethylated CpG in single-stranded DNA and 12

RNA:DNA hybrids both sensed by endosomal TLR9; and cytoplasmic double-stranded

DNA sensed by cyclic GMP-AMP synthase (cGAS) (Cervantes et al., 2012; Diebold et al., 2004; Gao et al., 2013; Lund et al., 2004; Rigby et al., 2014; Rutz et al., 2004; Tanji et al., 2015). PRRs that recognize dsRNA to induce type I IFNs include endosomal

TLR3, and the cytosolic RIG-I-like receptors (RLRs), RIG-I and MDA5 (Sen and Sarkar,

2005; Yoneyama et al., 2004). While dsRNA-activated TLR3, RIG-I, and MDA5 all induce signaling pathways that result in NFκB- and IRF3-stimulated expression of type I

IFN (Kawai et al., 2005; Sen and Sarkar, 2005), these PRRs rely on different adaptor

molecules for the initiation of downstream signaling cascades (Figure 1.2). TLR3

initiates signaling through Toll/IL-1R domain containing adaptor inducing IFNβ (TRIF),

while RIG-I and MDA5 initiate signaling through mitochondrial antiviral signaling protein

(MAVS), also known as IFN β promoter stimulator-1 (IPS-1), to mediate the

phosphorylation and dimerization of IRF3, and the degradation of IκB to release active

NFκB (Borden et al., 2007; Fitzgerald et al., 2003; Häcker and Karin, 2006; Hiscott,

2007; Sankar et al., 2006; Sen and Sarkar, 2005; Seth et al., 2005).

The 2ˊ-5ˊ OAS/RNase L pathway is important in induction of type I IFN

expression by generating RNA fragments that function as ligands for RIG-I, but less so

for MDA5 (Figure 1.2) (Malathi et al., 2007, 2010). RIG-I and MDA5 are structurally

similar, containing two caspase activation and recruitment domains (CARD), a central

DExD/H-box helicase domain, and a C-terminal regulatory domain that binds PAMP

RNA. Laboratory of Genetics and Physiology 2 (LGP2) protein, another RLR, has similar

organization to RIG-I and MDA5, but does not contain CARD domains. To facilitate the

downstream signaling required for IFN production, the CARD domains of RIG-I and

MDA5 each interact with the CARD domain of the mitochondrial-membrane associated

signaling adaptor, MAVS, (Horner et al., 2011; Kawai et al., 2005; Meylan et al., 2005;

Yoneyama et al., 2015). Because LGP2 does not have a CARD domain, it cannot 13 directly activate MAVS; however it can bind RNA ligands and synergize with MDA5 to activate MAVS signaling (Bruns et al., 2013; Satoh et al., 2010).

Natural ligands for RIG-I, MDA5, and LGP2 are not well characterized, and continue to be a point of interest and dispute among researchers in the field. Compelling data indicate that RIG-I recognizes short dsRNA (minimum of 10 bps) with blunt ends, and a 5ˊ-triphosphate (Lu et al., 2011; Luo et al., 2011). The C-terminal regulatory domain of RIG-I completely surrounds the blunt end of its RNA substrates, strongly

interacting with the α and β phosphates, but not the γ phosphate of the 5ˊ-triphosphate

on PAMP RNA (Wang et al., 2010). Despite convincing data for dsRNA with 5ˊ

triphosphates as RIG-I ligands, other forms of RNA appear to function as RIG-I ligands

as well. A recent study demonstrated that 5ˊ-diphosphates, like those found on reovirus genomic segments, can also activate RIG-I (Goubau et al., 2014). Other RIG-I ligands derived from viral RNAs include the HCV PAMP derived from the polyU/C tract in the context of 5ˊ-triphosphate (Saito et al., 2008), and defective-interfering RNA associated with Sendai virus infections (Martínez-Gil et al., 2013), among other viral RNAs with dsRNA structure and 5ˊ-triphosphates (Kell and Gale, 2015). Whether RIG-I recognizes single-stranded RNAs with 5ˊ-triphosphate modifications remains unclear. IAV genomic

RNA segments have 5ˊ-triphosphates and activate RIG-I despite not containing detectable dsRNA (Weber et al., 2006); however, a recent study revealed that the panhandle structure of incoming genomic IAV nucleoprotein complexes activate RIG-I

(Pichlmair et al., 2006; Weber et al., 2013).

Unlike RIG-I, the C-terminal regulatory domain of MDA5 does not require end- recognition of PAMP RNA (Wu et al., 2013), and as a result can cooperatively bind to the PAMP RNA, forming filaments along long RNA substrates (Peisley et al., 2011; Wu et al., 2013). MDA5 recognizes longer dsRNAs of 0.5 – 7 kb, independent of 5ˊ- phosphate modifications (Kato et al., 2008). Recognition of RNAs from picornaviral RNA 14 is an important function of MDA5 (Gitlin et al., 2006; Kato et al., 2006). Picornavirus

RNAs have no RNA species with 5ˊ-triphosphate or diphosphate, but rather have a

protein, VPg, covalently linked to the 5ˊ end of its genomic RNA. VPg is cleaved from PV

RNAs that serve as templates for translation resulting in viral RNA bearing a 5 ˊ-

monophosphate (Flanegan et al., 1977; Lee et al., 1977; Ambros et al., 1978). Although

end-recognition is reportedly not required for MDA5 (Wu et al., 2013), another study that

focused on MDA5 recognition of coronavirus RNAs lacking 2ˊ-O-methyl group

modifications on their 5ˊ-cap structures (Züst et al., 2011), suggested that MDA5 can

recognize terminal modifications on PAMP RNA. MDA5 has also been reported to

induce expression of type I IFN in the context of LGP2. LGP2 enhances MDA5-

dependent induction of type I IFN. Due to high basal levels of ATP hydrolysis, LGP2 can

recognize a diverse set of RNA ligands and synergize with MDA5 to induce IFNβ

expression (Bruns et al., 2013; Satoh et al., 2010).

RNase L can produce RNA fragments that promote the expression of type I IFN.

When small RNA fragments (< 200 nts) generated by RNase L cleavage were

transfected into tissue culture cells, IFNβ was produced in a RIG-I-dependent, and to a

lesser extent MDA5-dependent manner (Malathi et al., 2007). Intraperitoneal injection of

2-5A into WT mice induced the expression of IFNβ, whereas injection of 2-5A into

RNase L-deficient mice does not induce expression of IFNβ. These studies indicate that

RNase L contributes to the production of type I IFN (Malathi et al., 2007). The RNA

fragments produced by RNase L have 5ˊ-hydroxyl and 3ˊ-phosphate termini (Carroll et

al., 1996; Wreschner et al., 1981b), but have not been carefully defined in the same

manner as in vitro synthesized RNA ligands for RIG-I and MDA5. However, studies of a

RIG-I ligand produced by RNase L cleavage of HCV RNA (svRNA3), indicated that a 3ˊ-

phosphate found on this RNA was important for its ability to activate RIG-I (Malathi et al.,

2010). The focus on the requirement of RIG-I ligands to have 5ˊ-triphosphate and – 15 diphosphate groups may impede the discovery of new ligands that have non-canonical features, like 3ˊ-phosphates that contribute to RIG-I recognition. Careful studies are

needed to assess the diversity of RNA PAMPs for both RIG-I and MDA5.

Several viruses have mechanisms to counteract the induction of type I IFN

through RIG-I and MDA5. Cleavage of MAVS from the mitochondrial membrane is a

common strategy used by viruses to inhibit the induction of type I IFN. HCV NS3/4A

protease, hepatitis A virus 3ABC precursor to 3C protease, coxsackie B 3C protease

and human rhinovirus 2A protease and 3C protease all cleave MAVS from the

mitochondrial membrane (Drahos and Racaniello, 2009; Li et al., 2005b; Mukherjee et

al., 2011; Yang et al., 2007). Viruses that inhibit type I IFN expression by directly

targeting RIG-I include PV, which directly degrades RIG-I through its 3C protease, and

IAV, which inhibits TRIM25-dependent ubiquitin-ligation of the RIG-I CARD to prevent

downstream activation of MAVS (Barral et al., 2009; Gack et al., 2007, 2009).

2ˊ-5ˊ OAS/RNase L System: RNA Cleavage

rRNA cleavage is considered the hallmark of RNase L activation, yet the

significance of rRNA cleavage has not been carefully studied (Silverman et al., 1982,

1983; Wreschner et al., 1981a). Initial experiments that characterized the effects of

dsRNA on cell-free extracts from IFN-treated HeLa cells resulted in inhibition of protein synthesis (Roberts et al., 1976), likely due to contributions from both PKR and the

OAS/RNase L pathway. To elucidate the effects of 2-5A on protein synthesis, Clemens

and Williams treated rabbit reticulocyte lysates (RRL) with 2-5A, which would have

activated RNase L. The results of these studies indicated that in the presence of 2-5A,

polysomes disaggregated into monomeric 80S ribosomes, and individual 40S and 60S

subunits. 2-5A did not prevent the association of initiator methionine-tRNA with 40S

subunits, yet the resulting initiation complexes were not incorporated into actively 16 translating polysomes. The authors concluded that 2-5A likely influenced 40S ribosomes, but the inhibition of protein synthesis observed in 2-5A-treated lysates was most likely due to the induction of a nuclease, later determined to be RNase L, that degraded mRNA substrates (Clemens and Williams, 1978).

The specific sites of RNase L cleavage in rRNA have been partially characterized using primer extension of poly(I:C)-treated HeLa cells. These sites, CU3999 and UG4000 were identified in 28S rRNA, and mapped to regions of the L1 protuberance, predicted to be important for tRNA release from the ribosome (Ben-Shem et al., 2010; Iordanov et al., 2000). Cleavage at these sites would likely inhibit protein synthesis. In experiments described in Chapters III and IV of this thesis, I describe RNase L-dependent cleavage

of rRNAs. RNase L frequently cleaves two sites in 18S rRNA. Although I was able to

detect the same sites reported by Iordanov et al., according to my data, I was unable to

directly attribute cleavage at these sites in 28S rRNA to RNase L activation.

As discussed earlier, RNase L cleaves viral RNAs and cellular mRNAs; however,

in vivo targets of RNase L are poorly defined. Studies that suggest RNase L directly

cleaves viral RNA include the selective reduction of EMCV RNA from IFN-treated cells

expressing WT RNase L (Li et al., 1998). Although the accumulation of viral RNA was

influenced by RNase L activation, global levels of cellular mRNAs remained unaffected.

The reduction of EMCV RNA was suggested to occur prior to the accumulation of high

levels of 2-5A that provoke RNase L-mediated cleavage of rRNA (Li et al., 1998).

Selective reduction of viral RNA by RNase L would suggest localized activation of

RNase L in specific cellular compartments, like viral replication complexes. In studies

from the Barton Lab, incubation of HCV RNA in HeLa cell lysates results in degradation

of HCV RNA by RNase L (Han and Barton, 2002; Han et al., 2004), and electroporation

of full-length HCV RNA into Huh7.5 cells with subsequent IFNβ treatment and

transfection of poly(I:C) or 2-5A resulted in the accumulation of the svRNA3 RIG-I PAMP 17

(Malathi et al., 2010). In HEK293 cells, RNase L-dependent cleavage of Sindbis virus

RNA resulted in the accumulation of small RNAs about 20-30 nucleotides in length,

called SvsRNAs (Girardi et al., 2013). The significance of these SvsRNAs has yet to be

assessed. Cellular mRNAs are also reported to be targeted by RNase L. These mRNAs

have been studied using Northern blots, and include mRNA encoding ISG15, ISG43,

MyoD, and mitochondrial mRNAs (Bisbal et al., 2000; Li et al., 2000; Le Roy et al.,

2007). Due to the limitations of Northern blots, whether RNase L regulates the stability of

these cellular mRNAs by direct cleavage remains unclear.

Although the in vivo host and viral RNA targets of RNase L have not been well

defined, many viruses have strategies to inhibit the OAS/RNase L pathway (Table 1.2).

PV, used in experiments described in Chapter III to provoke RNase L activation in HeLa

cells, encodes a structure within its RNA that acts as a competitive inhibitor of RNase L

(ciRNA). In molar excess, the ciRNA inhibits RNase L. Although this structure is

conserved in group C enteroviruses, its significance is not completely understood.

Infection of HeLa cells expressing WT RNase L with PV encoding a mutant ciRNA

provoked rRNA fragmentation earlier than infection with WT PV; however, the premature

rRNA fragmentation did not inhibit replication of the mutant PV when compared to

replication of WT PV (Han et al., 2007; Townsend et al., 2008a, 2008b). Another virus

that inhibits the activation of the OAS/RNase L pathway is IAV. The IAV NS1 protein is

important for preventing the induction of type I IFN, as well as counteracting the antiviral

effects of IFN (Hale et al., 2008; Krug, 2015; Mibayashi et al., 2007). Although IAV NS1

inhibits activation of RIG-I and PKR, its predominant role is to prevent activation of

OAS/RNase L, presumably by binding and sequestering dsRNA through its dsRNA

binding domain (Min and Krug, 2006; Min et al., 2007). In Chapter IV of this thesis, I

identified RNase L cleavage sites in host and viral RNAs from cells infected with

IAVΔNS1 to determine whether RNase L directly cleaves IAV RNA to mediate viral 18 restriction. Finally, in HCV, genotypes associated with favorable outcomes after IFN- based therapies (HCV2 and HCV3) tend to have a greater frequency of UA and UU dinucleotides in their genomes, whereas HCV genotype 1, affiliated with poor treatment outcomes for IFN-based therapies, has fewer UA and UU dinucleotides. The relationship between UA and UU dinucleotide frequency and success of IFN-based therapies suggests RNase L exerts selective pressures on HCV (Washenberger et al., 2007).

Other strategies employed by viruses to prevent the activation of RNase L, are listed in

Table 1.2.

Table 1.2. Viral countermeasures to RNase L Han et al., 2002 Reduced frequency of RNase L cleavage sites in HCV RNAs Hepatitis C Virus Han et al., 2004 correlates with IFN treatment failure in patients. Washenberger et al., 2008 Han et al., 2007 Poliovirus A viral RNA structure in group C enteroviruses inhibits RNase L. Townsend et al., 2008a Townsend et al., 2008b A viral dsRNA-binding protein (NS1) increases influenza virus Influenza Min and Krug 2006 replication (100 fold) in RNase L +/+ cells.

Coronaviruses Viral 2ˊ, 5ˊ phosphodiesterase destroys 2-5A (MHV ns2 protein) Zhao et al., 2012

Rotaviruses Viral 2ˊ, 5ˊ phosphodiesterase destroys 2-5A (RVA VP3 protein) Zhang et al., 2013

Sorgeloos et al., 2013 Theiler’s Virus Viral L* protein inhibits murine RNase L

2ˊ-5ˊ OAS/RNase L System: Importance in Apoptosis

RNase L is important for the induction of apoptosis in virally-infected cells

(Castelli et al., 1998; Li et al., 2004; A. Zhou et al., 1997). Apoptosis, or programmed cell death, is designed to eliminate cells without causing inflammation, and in turn many viruses have developed mechanisms to co-opt the apoptotic pathway. Viruses can maximize progeny virus production or transition into a state of persistent or chronic infection by inhibition of apoptosis, whereas promotion of apoptosis can facilitate viral

release during lytic infections.

Apoptosis is executed through two pathways: the extrinsic/death receptor pathway, and the intrinsic/mitochondrial pathway (Bratton and Salvesen, 2010; Elmore, 19

2007). The extrinsic pathway is initiated by transmembrane receptors of the TNF family

(TNF, Fas, TRAIL) that bind their respective proapoptotic ligands. In response to this binding, Fas-associated death domain (FADD) associates with the cytoplasmic tails of the receptor and recruits procaspase 8 to form a complex known as the death-inducing signaling complex (DISC) that serves to recruit additional procaspase 8. In the context of

DISC, procaspase 8 autoactivates to caspase 8, inducing the caspase cascade that results in apoptosis.

In the intrinsic/mitochondrial apoptotic pathway, apoptotic stimuli lead to the release of cytochrome c and other pro-apoptotic factors from the mitochondrial membrane. After its release into the cytoplasm, cytochrome c interacts with apoptotic protease factor-1 (Apaf-1) and procaspase 9 to form a complex called the apoptosome.

The apoptosome recruits additional procaspase 9 that autoactivates and leads to

activation of effector caspases 3 and 7. The activation of caspases 3 and 7 result in

downstream DNA fragmentation, chromatin condensation, and apoptotic cell death

(Bratton and Salvesen, 2010; Elmore, 2007; Janicke, 1998; Porter and Jänicke, 1999).

How does RNase L contribute to apoptotic pathways? Mice unable to express

RNase L have an enlarged thymus, and thymocytes isolated from these RNase L-

deficient mice are resistant to apoptosis after treatment with anti-Fas, staurosporine,

anti-CD3, and TNFα (Zhou et al., 1997). These results suggest that RNase L is

important for the development of thymocytes through its influence on apoptosis. 2-5A

transfection of IFNα-treated MEFs provokes apoptotic cell death in about 15% of cells in

an RNase L-dependent manner (Zhou et al., 1997). Activation of RNase L by 2-5A

transfection causes release of cytochrome c from the mitochondrial membrane, and

subsequent caspase-3 dependent apoptotic cell death (Rusch et al., 2000). The release

of cytochrome c from the mitochondrial membrane is thought to be triggered by

activation of c-Jun N-terminal kinase (JNK) (Li et al., 2004). RNase L enhances the 20 activation of JNK in virally-infected cells (Li et al., 2004). JNK promotes apoptosis by mediating the proteosomal degradation of the transcriptional co-repressor CtBP (Wang et al., 2006) to activate the expression of pro-apoptotic genes (Grooteclaes et al., 2003).

JNK also regulates the Bcl2-related proteins that facilitate the release of cytochrome c from the mitochondrial membrane (Tournier et al., 2000; Davis, 2000). Although RNase

L-enhanced JNK activation is important for cytochrome c release and apoptosis, how

RNase L promotes JNK activation is not well understood. RNase L-induced rRNA cleavage is suspected to activate JNK, similar to JNK activation resulting from damage to 28S rRNA by ribotoxins or UV radiation (Tournier et al., 2000; Iordanov et al., 1997,

2000; Zhou et al., 2014). Treatment of cells with poly(I:C) to induce RNase L-mediated

rRNA cleavage leads to activation of JNK (Iordanov et al., 2000; Li et al., 2004).

Functional ribonuclease activity of RNase L is not only essential for cleavage of rRNA, but it is also important for the induction of apoptosis (Han et al., 2007; Li et al., 2004). A ribonuclease-deficient mutant of RNase L (R667A) was ineffective at inducing cleavage of poly(A) ADP-ribose polymerase (a substrate cleaved during the apoptotic caspase cascade) after transfection of 2-5A (Li et al., 2004). The contributions of RNase L to apoptotic pathways in the cell highlight its importance in cancer and mediating the anti- proliferative effects of IFN.

2ˊ-5ˊ OAS/RNase L System: Regulation

The OAS/RNase L pathway is regulated by a few mechanisms in cells. An IFN-

induced 2ˊ-5ˊ phosphodiesterase (2ˊ-5ˊ PDE) regulates the levels of 2-5A during virus

infections (Kubota et al., 2004; Poulsen et al., 2011; Schmidt et al., 1979). When this 2ˊ-

5ˊ PDE was identified in mouse L cell lysates, it was observed to degrade 2-5A as well as the terminal CCA modification on tRNAs (Schmidt et al., 1979). The 2ˊ-5ˊ PDE is ubiquitously expressed in different cell types to temper the antiviral and anti-proliferative 21 activity of RNase L (Kubota et al., 2004). A 2ˊ-5ˊ PDE was also found in the mitochondria, which may have implications for RNase L activation within in this organelle

(Kjær et al., 2014). 2ˊ-5ˊ PDEs are also encoded by coronaviruses and rotaviruses as strategies to counteract RNase L activation (Table 1.2) (Zhang et al., 2013; Zhao et al.,

2012). Another regulatory component to the 2ˊ-5ˊ-OAS/RNase L pathway is the RNase L inhibitor, which antagonizes 2-5A binding and ribonuclease activity of RNase L. The

RNase L inhibitor co-immunoprecipitates with RNase L, and its overexpression in HeLa cells inhibits the antiviral effects of IFN on EMCV infection (Bisbal et al., 1995).

Angiogenin and other Ribonuclease A Family Enzymes

Humans encode eight members of the RNase A superfamily (RNase 1-8), the genes of which are all located on chromosome 14 (Cho et al., 2005). A summary of these enzymes and their specificity are presented in Table 1.1. All members of the

RNase A superfamily are characterized by a catalytic triad composed of two histidines and a lysine that confers specificity for cleavage 3ˊ of pyrimidine bases; however some members of this family have preference for cleavage after cytidine over uridine and vice versa. All RNase A enzymes cleave RNA through a transphosphorylation reaction that produces RNA fragments with a 2ˊ, 3ˊ-cyclic phosphate and 5ˊ-hydroxyl termini; however certain RNase A superfamily members are reported to hydrolyze 2ˊ, 3ˊ-cyclic mononucleotides more efficiently than others, suggesting that some members of the

RNase A family of enzymes have the potential to hydrolyze the terminal 2ˊ, 3ˊ-cyclic phosphate found on RNA products to terminal 2ˊ or 3ˊ-monophosphates (Sorrentino,

2010). The eight RNase A family enzymes (Table 1.1) share 27 - 78% amino acid sequence identity with each other (Becknell et al., 2014; Futami et al., 1997; Rudolph et al., 2006; Sorrentino, 2010). All members of the RNase A family demonstrate antimicrobial activity and Northern blot detection of mRNA from RNase 1, RNase 2, 22

RNase 4, and RNase 5 (angiogenin) suggest that these ribonucleases are expressed throughout many tissue types (Futami et al., 1997).

Angiogenin was the first angiogenic factor purified from the conditioned media of tumor cell line HT-29 (Fett et al., 1985). The purpose of these studies was to identify the factors required for sustained and rapid growth of tumor cells. Purified angiogenin induced new blood vessel growth (angiogenesis) in a chick chorioallantoic membrane assay, and stimulated angiogenesis when implanted into rabbit corneas (Fett et al.,

1985). Angiogenin is the fifth member of the RNase A superfamily (Sorrentino, 2010), and has functional domains involved in receptor binding, RNA catalysis, and nuclear localization. The ribonuclease activity of angiogenin is crucial for angiogenic activity, but

the relationship between RNA cleavage by angiogenin and stimulation of angiogenesis

is not yet understood (Shapiro and Vallee, 1989; Shapiro et al., 1989).

Extracellular angiogenin is taken up by target cells through receptor-mediated

endocytosis and is transported to the nucleus where it accumulates in the nucleolus (Hu

et al., 1997; Moroianu and Riordan, 1994). Within the nucleolus, angiogenin binds to the

promoter region of the 47S rDNA gene to stimulate expression of 18S, 5.8S, and 28S

rRNAs (Hu et al., 2000; Li and Hu, 2010; Tsuji et al., 2005; Xu et al., 2003, 2002).

Knockdown of angiogenin in human umbilical cord endothelial vein cells resulted in

reduced 47S rDNA transcription, angiogenesis, and cell proliferation (Kishimoto et al.,

2005; Tsuji et al., 2005). Many types of cancers highly express angiogenin (Tello-

Montoliu et al., 2006), where angiogenin-stimulated 47S rDNA synthesis and ribosome biogenesis are important for the high levels of protein synthesis required in the continuously proliferating cancer cells, as well as angiogenesis to supply the rapidly growing cells with oxygen and nutrients (Nishida et al., 2006; Ruggero and Pandolfi,

2003; Derenzini et al., 2000). Angiogenin is regulated by the placental ribonuclease inhibitor (RNH1), which abolishes its ribonuclease activity and ability to provoke 23 angiogenesis (Shapiro and Vallee, 1987). Mutations in angiogenin that disrupt its ribonuclease activity or nuclear localization signal are associated with susceptibility to

ALS (Padhi et al., 2013; Wu et al., 2007).

Recently, angiogenin was identified as the ribonuclease that cleaves tRNAs in response to various types of cellular stress and is important for maintaining cell survival during times of stress. These tRNA fragments, derived from anticodon loop cleavage of mature tRNAs, accumulate in cells in response to oxidative stress, nutrient deprivation, radiation, hypothermia, hyperosmotic stress, sodium-arsenite treatment, and viral infection, and have been referred to as tRNA-derived stress-induced small RNAs

(tiRNAs), tRNA-derived RNA fragments (tRFs), or as tRNA halves (tRHs) (Fu et al.,

2009a; Lee and Collins, 2005; Thompson et al., 2008; Wang et al., 2013; Yamasaki et

al., 2009; Selitsky et al., 2015). From here on, angiogenin-mediated tRNA fragments will

be referred to as tiRNAs, where the 5ˊ half of the tRNA is the 5ˊ-tiRNA and the 3ˊ half of

the tRNA is the 3ˊ-tiRNA.

Anticodon loop cleavage of tRNAs is a conserved response to stress, and has been described in bacteria as a response to bacteriophage infection and nutrient deprivation (Amitsur et al., 1987; Haiser et al., 2008), in Tetrahymena as a response to starvation (Lee and Collins, 2005), and in yeast in response in oxidative stress

(Thompson and Parker, 2009). RNase T2, an endoribonuclease in the acid hydrolase family, is responsible for tRNA anticodon loop cleavage in yeast (Thompson and Parker,

2009). In mammalian cells, knockdown of RNase T2 had no impact on tRNA cleavage; rather, knockdown of angiogenin reduced the accumulation of tiRNAs in mammalian cells, whereas its overexpression resulted in increased accumulation tiRNAs (Fu et al.,

2009a; Wang et al., 2013; Yamasaki et al., 2009). Overexpression of the placental ribonuclease/angiogenin inhibitor, RNH1, reduced the accumulation of tiRNAs, as would be expected (Yamasaki et al., 2009). 24

Transfection of purified 5ˊ- tiRNAs, the 5ˊ half of the tRNA, into cells resulted in

~20 % reduction of protein synthesis (Emara et al., 2010; Ivanov et al., 2011; Yamasaki et al., 2009). Synthetic 5ˊ-tiRNAs encoding the anticodon loop for alanine and cysteine

(5’-tiRNA-Ala and 5’-tiRNA-Cys) have a greater ability to inhibit protein synthesis than other tiRNAs. Transfection of 5ˊ-tiRNA-Ala displaces the eIF4F complex (eIF4E, eIF4A, and eIF4G) from capped and uncapped mRNAs to inhibit their translation; however 5ˊ- tiRNA-Ala did not inhibit translation of an EMCV IRES-driven transcript (Ivanov et al.,

2011). Many of the cellular IRES-driven transcripts currently described encode anti- apoptotic genes (Li and Hu, 2012). YB-1, a protein that inhibits translation from capped and uncapped mRNA transcripts, was found to interact with 5ˊ-tiRNA-Ala, and was an important factor for translational repression induced by 5ˊ-tiRNA-Ala and 5ˊ-tiRNA-Cys, and for the production of stress granules (Ivanov et al., 2011). Stress granules are transient cytoplasmic foci that arise during cell stress and contain stalled pre-initiation complexes, 40S rRNA, mRNA, and various translation initiation factors (Kedersha and

Anderson, 2007). The formation of stress granules as a result of tiRNAs occured independent of phosphorylation of eIF2α (Emara et al., 2010; Ivanov et al., 2011), suggesting that tiRNAs repress translation by mechanisms independent of eIF2α. The oligo-guanine motif (a stretch of four to five guanines), found at the 5ˊ-end of tRNA-Ala and tRNA-Cys is important for the ability of these tiRNAs to inhibit translation. 5ˊ- tiRNA-

Met is normally unable to inhibit translation; however the addition of a synthetic oligoguanine motif to its 5ˊ-end enables 5ˊ-tiRNA-Met to inhibit protein synthesis (Ivanov et al., 2011, 2014). Angiogenin-dependent tiRNAs can also inhibit hyperosmotic-induced apoptosis by forming complexes with cytochrome c after its release from the mitochondria (Saikia et al., 2014), preventing cytochrome c-dependent induction of apoptotic cell death. Moreover, siRNA knockdown of angiogenin in a human bladder cancer cell line led to apoptotic cell death of the cancerous cells (Shu et al., 2015). 25

These findings suggest that angiogenin mediates cell survival by inhibition of protein synthesis to preserve cellular resources, by repression of apoptosis, and by potentially favoring the expression of IRES-containing transcripts which may encode for anti- apoptotic factors.

Some viral infections can also induce angiogenin-mediated tiRNA accumulation; although the accumulation of tiRNAs has only been studied in the context of respiratory syncytial virus (RSV), HCV, and hepatitis B virus (HBV) infections (Wang et al., 2013;

Selitsky et al., 2015). Deep sequencing of small RNA libraries from RSV-infected A549 cells identified many tiRNA species, but 5ˊ-tiRNA-Glu-CUC was frequently detected in

RSV-infected cells when compared to uninfected cells. Knockdown of the 5ˊ-tiRNA-Glu-

CUC fragment inhibited RSV replication, indicating that 5ˊ-tiRNA-Glu-CUC potentially stimulates RSV replication. RSV, a negative strand virus, produces capped and polyadenylated mRNAs (Cowton et al., 2006) that might be subject to translational repression through the actions of 5ˊ-tiRNA-Cys. Interestingly, the oligoguanine-motif containing 5ˊ-tiRNA-Cys-CGA was also frequently detected in RSV infected cells (Wang et al., 2013); however, its impact on RSV infection was not investigated. Deep sequencing of small RNA cDNA libraries from the livers of patients chronically infected with HCV or HBV revealed that 5ˊ-tiRNA-Gly and 5ˊ-tiRNA-Glu unexpectedly made up a significant proportion of the cDNA libraries (Selitsky et al., 2015); however, the significance of these tiRNAs in chronic HCV/HBV infection is not understood.

tRNAs are currently the only known target of angiogenin (Fu et al., 2009a;

Saxena et al., 1992; Yamasaki et al., 2009), but angiogenin likely cleaves other RNAs during cellular stress and to stimulate 47S rDNA transcription. Addition of angiogenin to rabbit reticulocyte lysates results in cleavage of rRNA and inhibition of protein synthesis, but the cleavage of rRNA and the inhibition of protein synthesis by this mechanism have not been verified in vivo (St Clair et al., 1987, 1988). In studies that highlighted the 26 importance of angiogenin for neuroprotection during stress, primary motor-neurons were

serum-starved to provoke secretion of angiogenin. Secreted angiogenin was taken up by

nearby astroglia where it induced RNA fragmentation in the astroglia; however, the RNA

fragments were not consistent in size with tiRNA fragments. tiRNA fragments typically

range from 30 to 40 nts in length (Skorupa et al., 2012), suggesting that other RNAs may

be targeted by angiogenin to promote survival during cell stress. In Chapter V, I

describe RNA cleavage in HCV-infected cells consistent with the activity of angiogenin,

and likely other RNase A-family ribonucleases. I detected cleavage within the anticodon

loop of tRNA-Glu-CUC, in addition to cleavage in viral and other host RNAs, potentially

identifying additional RNA targets of angiogenin.

Although angiogenin has demonstrated roles in cancer and cell survival, the

importance of other human RNase A family enzymes have not been as well-studied. The

remainder of this section describes what is known about their function (Table 1.1).

Pancreatic RNase (RNase 1), is the prototypic RNase A member and shares 70% amino

acid conservation with bovine pancreatic RNase (Beintema et al., 1984). As its name

suggests, pancreatic RNase/RNase 1 is abundantly expressed in the pancreas among a

wide variety of other tissue types (Futami et al., 1997). RNase 1 is heavily secreted by

human endothelial cells and is found in the bloodstream, urine, and seminal plasma

(Iwama et al., 1981; Landré et al., 2002; De Prisco et al., 1984). Pancreatic RNase from

ruminants is proposed to function in digestion and the degradation of large amounts of

RNA from food and gut microorganisms (Beintema et al., 1985), but the function of

RNase 1 in humans is unclear. Recombinant RNase 1 can activate dendritic cells, and

mRNA expressing RNase 1 is detected in human monocytes (Egesten et al., 1997; Yang

et al., 2004), suggesting that RNase 1 may have stimulatory effects on immune cells.

Compared to bovine RNase A, human RNase 1 can degrade dsRNA more readily due to

additional basic residues that destabilize dsRNA (Libonati and Sorrentino, 2001; De 27

Prisco et al., 1984; Yakovlev et al., 1997). The ability to disrupt dsRNA structure is critical because metal-ion-independent endoribonucleases like RNase A are single- strand specific enzymes.

RNase 2 and RNase 3 are closely related enzymes that share 70% amino acid identity (Sorrentino and Glitz, 1991). Both were discovered as factors secreted from the granules of activated eosinophils, and are also known as eosinophilic derived neurotoxin

(RNase 2) and eosinophilic cationic protein (RNase 3). These two proteins are considered neurotoxic when assayed by the “Gordon Phenomenon” diagnostic test for

Hodgkin Disease (Durack et al., 1979; Gleich et al., 1986). In addition to eosinophils,

RNase 2 is highly expressed by liver cells (Frank and Levy, 1976) and its mRNA can be

detected in lung, spleen, thymus, and human monocytes (Egesten et al., 1997; Futami et

al., 1997). RNase 2 and RNase 3 both prefer to cleave single-stranded poly(U)

substrates over poly(C) substrates (Sorrentino and Glitz, 1991; Sorrentino et al., 1992).

Like RNase 1, the functions of RNase 2 and RNase 3 are not well understood. RNase 2

and RNase 3, released from eosinophils in RSV virus suspensions, reduced the

infectivity of RSV by four-fold. This reduction was specific to the two RNase A-family

enzymes since addition of RNH1 reversed this antiviral effect (Domachowske et al.,

1998). Recombinant RNase 2 mixed with suspensions of RSV or parainfluenza virus-3

reduced viral infectivity (up to 40-fold) in a concentration-dependent manner (0 – 50 nM).

This reduction in viral infectivity was dependent on the ribonuclease activity of RNase 2,

since a catalytically inactive mutant did not affect RSV or parainfluenza virus infectivity

(Domachowske et al., 1998). RNase 2 is an alloantigen-stimulated factor secreted from

peripheral blood mononuclear cells during mixed lymphocyte reactions, where RNase 2

had antiviral activity towards human immunodeficiency virus and was suggested to

contribute to inhibition of HIV vertical transmission (Rugeles et al., 2003). 28

RNase 4 and RNase 5 (angiogenin) were both isolated from the conditioned medium of HT-29 colon carcinoma cells (Fett et al., 1985; Shapiro et al., 1986a). They share the same promoters (Dyer and Rosenberg, 2005; Futami et al., 1997), and share

38% amino acid identity (Futami et al., 1997). Both angiogenin and RNase 4 are expressed in a wide range of tissues, and highly expressed in liver tissue (Futami et al.,

1997). Angiogenin, in general, cleaves after pyrimidines, but has a preference for

cleavage 3ˊ of cytidines (Curran et al., 1993; Rybak and Vallee, 1988; Shapiro et al.,

1986b), while RNase 4 has a preference for cleavage 3ˊ of uridines (Hofsteenge et al.,

1998; Terzyan et al., 1999). Both angiogenin and RNase 4 are important for motor

neuron health, where they are associated with angiogenesis, neurogenesis, and

neuroprotection (Li et al., 2013). While angiogenin and RNase 4 have overlapping

effects in motor neuron cells, RNase 4 does not appear to be involved in tRNA cleavage

in response to cell stress. Treatment of cells with recombinant angiogenin resulted in the

accumulation of tiRNAs, whereas treatment of cells with recombinant RNase 4 did not

result in tiRNA accumulation (Yamasaki et al., 2009); however, exogenous RNase 4 was

taken up by human umbilical vein endothelial cells through an unknown mechanism to

induce angiogenesis and neurite outgrowth (Li et al., 2013). Like angiogenin, the

ribonuclease activity of RNase 4 is also important for angiogenesis (Li et al., 2013). Both

angiogenin and RNase 4 demonstrate antiviral activity, and were among the soluble

factors released by T cells that have antiviral activity against HIV strain X4 (Cocchi et al.,

2012).

RNase 6, RNase 7, and RNase 8 are all closely related and demonstrate

antimicrobial activity against Gram positive and Gram negative bacteria. RNase 7 was

first identified from skin as an antimicrobial peptide (Harder and Schröder, 2002);

however, it is also found in human urine, where it has roles in maintaining urinary tract

sterility (Spencer et al., 2011). While RNase 7 is consistently secreted into urine by 29 parenchymal cells in the kidney, RNase 6 is secreted into urine by recruited leukocytes during uropathogenic E. coli infection (Becknell et al., 2014). While RNase 6 and RNase

7 share 56% amino acid identity, RNase 7 shares 78% amino acid identity with RNase 8

(Becknell et al., 2014; Rudolph et al., 2006). Although RNase 8 has similar antimicrobial activity to RNase 7 (Rudolph et al., 2006), Northern blot analysis of RNASE8 mRNA from diverse tissue types indicate that RNase 8 is only expressed in placental tissue

(Zhang et al., 2002). While these three RNases have demonstrated ribonucleolytic activity, their specificity has yet to be characterized (Zhang et al., 2002).

In Chapter V, I describe data consistent with an angiogenin-dependent stress

response in HCV-infected Huh7.5.1 cells. The specificity of RNase A family members for

cleavage 3ˊ of pyrimidines and the reported specificity of angiogenin for cleavage within

the anticodon loop of tRNA will be important in the interpretation of data from HCV-

infected Huh7.5.1 cells.

IRE1 and the Unfolded Protein Response

The ER lumen is the site of protein modifications and protein folding. To ensure

that proteins are properly folded, the ER lumen contains transmembrane proteins ATF6,

PERK, and IRE1, that respond to the accumulation of unfolded proteins in the ER lumen

by initiating the Unfolded Protein Response (UPR). The purpose of the UPR is to

ameliorate the problem of unfolded proteins in the ER lumen by increasing protein

folding capacity, reducing translation, expanding the ER network, and increasing ER-

associated degradation of unfolded proteins. When the burden of unfolded proteins

overwhelms these homeostatic mechanisms, the UPR can provoke apoptotic cell death

(Lin et al., 2007).

ATF6 is a transmembrane protein that contains a bZIP transcription factor at its

amino terminus. Once ATF6 detects unfolded proteins, it relocalizes to the Golgi 30 apparatus where it is proteolytically cleaved to release the bZIP transcription factor. The liberated bZIP transcription factor translocates to the nucleus to induce the expression of genes involved in the UPR (Okada et al., 2002; Shen et al., 2002). PERK is a transmembrane kinase that phosphorylates the α-subunit of eIF2, the same substrate as

PKR, to inhibit protein synthesis. Inositol requiring enzyme 1 (IRE1) is a conserved transmembrane protein that senses unfolded proteins through its N-terminal luminal domain, and in turn oligomerizes and transphosphorylates to activate its C-terminal cytoplasmic endoribonuclease domain. Activated IRE1 cleaves XBP1unspliced mRNA at

two stem-loop structures, excising a 26 base-long intron. The two exons generated from

this unconventional cytoplasmic splicing reaction are then ligated back together to

generate XBP1spliced, which encodes a bZIP transcription factor that regulates the

expression of genes encoding chaperone proteins and components of ER-associated

degradation of unfolded proteins. While tRNA ligase (Trl1) has long been known as the

ligase responsible for ligating the exons of Hac1p mRNA, the yeast homolog of XBP1

mRNA, the mammalian ligase was recently identified as RtcB (Jurkin et al., 2014; Lu et

al., 2014; Kosmaczewski et al., 2014). In addition to cleavage of XBP1 mRNA during

UPR, IRE1 also targets several ER-associated mRNAs presumably to reduce stress on

translational machinery (Hollien et al., 2009). While the IRE1α isoform is associated with

XBP1 mRNA splicing, both IRE1α and the IRE1β isoform, participate in regulated IRE1-

dependent mRNA decay (RIDD) (Hollien et al., 2009; Maurel et al., 2014).

Although IRE1α is ubiquitously expressed, the expression of IRE1β is limited to

epithelial cells of the airway and gastrointestinal tract (Martino et al., 2012). Among the

RNA substrates cleaved by IRE1β is 28S rRNA to inhibit protein synthesis. Iwawaki et al.

identified IRE1β-dependent cleavage sites in 28S rRNA (CU4031, UG4032, GC4033, and

GC4131; converted to NR_003287.2 28S rRNA numbering) in HeLa cells overexpressing

IRE1β (Iwawaki et al., 2001). Cleavage of 28S rRNA at overlapping sites (CU4031 and 31

UG4032; converted to NR_003287.2 28S rRNA numbering) was attributed to RNase L activity in poly(I:C) treated HeLa cells (Iordanov et al., 2000). The various RNAs targeted by IRE1α and IRE1β and the RNA sequence motifs they recognize are reviewed by

Maurel et al., 2014. IRE1β is also reported to regulate mucin production in airway epithelial cells (Martino et al., 2012). As described in Chapters III-V, my data clearly distinguish between RNase L-dependent and RNase L-independent cleavage sites in

28S rRNA during viral infection. Thus, I will discuss the nature of cleavage sites in rRNAs and their attribution to individual ribounucleases in Chapters III-V.

IRE1 shares many similarities with RNase L. While IRE1 has an N-terminal domain that senses unfolded proteins to initiate the conformational changes activate its

C-terminal endonuclease, RNase L has an N-terminal regulatory domain that binds 2-5A to initiate the conformational changes required for dimerization/oligomerization and subsequent nuclease activation (Figure 1.1). RNase L and IRE1 share considerable homology within the kinase extension nuclease (KEN) domain (Bork and Sander, 1993;

Dong et al., 2001). While RNase L has a relatively relaxed specificity for single-stranded

UA and UU dinucleotides (Table 1.1) (Carroll et al., 1996; Floyd-smith et al., 2011;

Wreschner et al., 1981b), the specificity of IRE1 is more restrictive, preferring to cleave after a G within the single-stranded loop of an RNA stem-loop structure. The two single- stranded loop sequences cleaved by IRE1 in yeast Hac1 mRNA include 5ˊ-CAG\CCG-3ˊ and 5ˊ-CCG\AAG-3ˊ (Gonzalez et al., 1999; Sidrauski and Walter, 1997). The “\” indicates the IRE1 cleavage site after the G residue. In mammalian cells, IRE1 cleaves the loop structures of XBP1 mRNA at 5ˊ-CCG\CAG-3ˊ and 5ˊ-CUG\CAG-3ˊ. Like RNase

L, IRE1 and the UPR are important in viral infections, and many viruses encode strategies to manipulate the UPR. In HCV replicon systems, XBP1 mRNA is spliced, but

the encoded bZIP transcription factor is ineffective at inducing the expression of factors

important for UPR, resulting in an increase in HCV-IRES translation (Tardif et al., 2004). 32

For IAV, treatment of cells with a chemical chaperone that alleviates ER stress, or an

IRE1 inhibitor (3,5-dibromosalicylaldehyde), repressed IAV protein synthesis (Hassan et al., 2012).

RNase T2, An Enzyme Involved in rRNA Processing

Conserved RNase T2 enzymes are found in all eukaryotic genomes (Hillwig et al., 2009). RNase T2 activity is optimal at acidic pH, consistent with their cellular distribution in lysosomes and late endosomes (Campomenosi et al., 2006; Haud et al.,

2011). RNase T2 enzymes cleave ssRNA to produce RNA fragments with terminal 2ˊ,

3ˊ-cyclic phosphates and 5ˊ-hydroxyls (Table 1.1); however there is debate as to the base specificity of this enzyme. RNase T2 enzymes are widely reported to cleave indiscriminately or to have some preference for cleavage after adenosine or guanosine residues in RNA (Irie, 1999; Luhtala and Parker, 2010). Studies of the specificity of recombinant human RNase T2 revealed that poly(A) and poly(U) substrate was cleaved while no cleavage was detected in poly(C) or poly(G) (Campomenosi et al., 2006). The substrate specificity of RNase T2 has yet to be determined with a heteropolymeric sequence that might identify sequence motifs preferentially targeted by the enzyme.

Human RNase T2 is encoded by one gene located on the long arm of chromosome 6 in 6q27 (Trubia et al., 1997). Loss of function mutations in the gene encoding RNase T2 are involved in cystic leukoencephalopathy, a autosomal recessive disease that causes cortical cysts and white matter abnormalities in the brain, and resembles congenital cytomegalovirus brain infection (Henneke et al., 2009). Functional studies of RNase T2 in a zebrafish model indicated that a truncated version of RNase T2

resulted in the accumulation of undigested rRNA in the lysosomes of brain neurons

(Haud et al., 2011). Undigested rRNA also accumulated in vacuoles within A. th. mutants

lacking RNase T2, suggesting that RNase T2 has conserved roles in rRNA degradation

and ribophagy (Hillwig et al., 2011). Although RNase T2 knockdown did not lead to the 33 accumulation of rRNA in lysosomes of HEK293 cells, a human embryonic kidney cell line (Haud et al., 2011), the role of RNase T2 has yet to be studied in human neurons. In addition to cystic leukoencephalopathy, deletions or mutations in the gene encoding

RNase T2 are associated with many types of cancers, and recombinant RNase T2 has anti-tumorigenic and anti-angiogenic properties (Smirnoff et al., 2006).

In Chapters III-V, I describe ribonuclease cleavage of rRNAs. Although specific cleavage sites in rRNA can be attributed to RNase L, the source of most of the cleavage observed in rRNA is unknown. Because RNase T2 is involved in rRNA degradation and would leave 2ˊ, 3ˊ-cyclic phosphates on the RNA fragments it produces, RNase T2 might potentially contribute to the cleavage I detected within rRNAs from the experiments described in Chapters III-V.

Usb1, A 3ˊ→5ˊ exonuclease Important for Regulating the Length of U6 snRNA

Usb1, also known as Mpn1, is a 3ˊ→ 5ˊ exoribonuclease important for maturation of U6 small nuclear RNA (snRNA). U6 small nuclear ribonucleoprotein (snRNP) is a component of the spliceosome along with U1, U2, U4, U5 snRNAs, and around 200 proteins to catalyze the removal of introns from pre-mRNAs (Wahl et al., 2009). While the genes that encode U1, U2, U4, and U5 snRNAs are transcribed by pol II, the genes that encode U6 snRNA are transcribed by pol III, and thus terminate with five to six thymines, the pol III termination signal (Orioli et al., 2012). The majority of U6 snRNA in mammalian cells contain four templated uridines followed by one non-templated uridine with a terminal 2ˊ, 3ˊ-cyclic phosphate (Lund and Dahlberg, 1992).Terminal uridyltransferase (TUTase) post-transcriptionally modifies the 3ˊ-end of U6 snRNA by adding terminal uridylates (Trippe et al., 1998, 2003). The poly(U) tails are subsequently trimmed back by the 3ˊ→5ˊ exoribonuclease, Usb1, to generate a mature U6 snRNA with a terminal 2ˊ, 3ˊ-cyclic phosphate (Lund and Dahlberg, 1992; Mroczek et al., 2012; 34

Shchepachev et al., 2012). The terminal 2ˊ, 3ˊ-cyclic phosphate found at the end of mature U6 snRNA mediates stability and promotes binding of the Lsm2-8 complex,

which is important for interactions with U4 snRNA, nuclear retention, and recruitment of

recycling factors (Achsel et al., 1999; Licht et al., 2008; Mroczek et al., 2012; Spiller et

al., 2007).

Mutations in the gene encoding Usb1 (C16orf57) are associated with the rare

autosomal recessive disease poikiloderma with neutropenia (PN). PN is characterized

by hyperpigmentation on the chest and neck, finger and toenail dystrophy, neutropenia,

pulmonary infections, and predisposition to acute myeloid leukemia (Colombo et al.,

2012; Hilcenko et al., 2013; Shchepachev and Azzalin, 2013). PN-associated mutations in C16orf57 are predicted to result in truncated non-functional Usb1 (Colombo et al.,

2012). U6 snRNA from patients with PN have elongated non-templated poly(U) tails as well as non-templated 3ˊ-adenylates (Hilcenko et al., 2013). In budding yeast and

Drosophila, U6 snRNA is 3ˊ-adenylated by TRAMP-associated poly(A) polymerase and targeted for degradation. In vitro, Usb1 removes both poly(U) and poly(A), and perhaps the inability of mutant Usb1 to remove adenylates from U6 snRNA target these RNAs to the nuclear exosome for degradation (Hilcenko et al., 2013). Although patients with PN have U6 snRNA of abnormal lengths, deep sequencing of cDNA from a PN-patient derived lymphoblast cell line did not reveal any obvious splicing deficits in the transcriptome (Hilcenko et al., 2013). Splicing defects may not be apparent in the cell type used, or Usb1 potentially has other functions and RNA targets that contribute to the underlying mechanism behind the pathologies of PN.

In Chapters III, IV, and V, the 2ˊ, 3ˊ-cyclic phosphate produced by Usb1 was frequently detected at the end of U6 snRNA from 2ˊ, 3ˊ-cyclic phosphate cDNA libraries and was used as a quality control indicator that RNAs with terminal 2ˊ, 3ˊ-cyclic phosphates were detected from complex populations of host and viral RNAs. 35

Mechanisms of RNA Cleavage

There are two positions in a phosphodiester bond that can be cleaved by ribonucleases: 5ˊ of the scissile phosphate (the phosphate of the bond to be cleaved), to

produce RNA fragments with terminal 5ˊ-phosphate groups and 3ˊ-hydroxyl groups, or 3ˊ of the scissile phosphate, to produce RNA fragments with terminal 5ˊ-hydroxyl groups and 2ˊ, 3ˊ-cyclic phosphates. Cleavage of the phosphodiester bond occurs through acid- base catalysis, where 1) a general base activates the nucleophile through deprotonation,

2) the nucleophile attacks the scissile phosphate to form a pentacovalent intermediate and the phosphodiester bond is broken, and 3) a general acid donates a proton to facilitate the leaving group (Figure 1.3). Nucleophiles used for cleavage of RNA typically

include deprotonated water or deprotonated 2ˊ- or 3ˊ- hydroxyl groups on RNA.

The requirement of divalent metal ions for catalysis determines whether RNA

cleavage results in fragments with terminal 5ˊ-phosphate and 3ˊ-hydroxyl groups, or

fragments with terminal 3ˊ-phosphate and 5ˊ-hydroxyl groups. Divalent metal ions,

typically Mg2+, are used by some ribonucleases to facilitate the formation of the nucleophile (usually a deprotonated water molecule), and to position the ribonuclease in relation to the scissile phosphate (Figure 1.3A). The influence of the divalent metal ions ensures that catalysis results in RNA fragments with terminal 5ˊ-phosphate and 3ˊ- hydroxyl groups (Steitz and Steitz, 1993). Usually, two metal ions interact with the scissile phosphate through the non-bridging oxygen and the conserved aspartic acid of the ribonuclease; however, in the context of metal-ion-dependent ribozymes, the metal ions and non-bridging oxygen can coordinate with a phosphate group of the ribozyme

(Yang et al., 2006). Ribonucleases that depend on divalent metal ions for catalysis have varied RNA substrates: RNase Z and RNase P, tRNA processing endoribonucleases, cleave ssRNA; Dicer and Drosha, involved in miRNA generation, cleave dsRNA; and members of the RNase H superfamily cleave RNA in DNA:RNA hybrids (Yang, 2011). 36

In contrast to ribonucleases that depend on divalent metal ions for cleavage, metal-ion-independent ribonucleases typically deprotonate the 2ˊ-hydroxyl of RNA to act as the nucleophile (Figure 1.3B). The deprotonated 2ˊ hydroxyl can function as the nucleophile to attack the phosphodiester bond in single-stranded RNA; however, dsRNA is too inflexible for this attack and consequently metal-ion-independent ribonucleases target single-stranded regions of RNA (Yang, 2011). The only exception, RNase 1, can overcome this strict definition by destabilizing dsRNA, which in effect makes the RNA single-stranded and suitable for nucleophilic attack by the mechanisms outlined here

(Sorrentino et al., 2003; Yakovlev et al., 1997). These reactions result in terminal 2ˊ, 3ˊ-

cyclic phosphates on the 5ˊ RNA fragment, and a 5ˊ-hydroxyl on the 3ˊ- RNA fragment

(Figure 1.3B). Examples of metal-ion-independent ribonucleases include the RNase A

superfamily members, IRE1, tRNA splicing endonuclease, RNase T2 enzymes, RNase

L, and Usb1 exoribonuclease (Table 1.1) (Abelson et al., 1998; Carroll et al., 1996;

Gonzalez et al., 1999; Mroczek et al., 2012; Peebles et al., 1983; Schutz et al., 2010;

Shchepachev et al., 2012; Wreschner et al., 1981b).

Terminal 2ˊ, 3ˊ-cyclic phosphates and 5ˊ-hydroxyls are recognized by certain

RNA ligases involved in tRNA splicing and the UPR (Abelson et al., 1998; Schutz et al.,

2010; Tanaka and Shuman, 2011; Tanaka et al., 2011). The 2ˊ, 3ˊ-cyclic phosphate found at the end of U6 snRNA impacts RNA stability and protein interactions (Mroczek et al., 2012; Spiller et al., 2007). While a 2ˊ, 3ˊ-cyclic phosphate is an important characteristic of some RNAs, it is unclear whether the 2ˊ, 3ˊ-cyclic phosphate is an intermediate or the product of RNA cleavage by metal-ion-independent ribonucleases.

Studies of the cleavage mechanism of RNase A initially suggested that two steps were involved: 1) the transphosphorylation that produces the 2ˊ, 3ˊ-cyclic phosphate on the 5ˊ- fragment and 5ˊ-hydroxyl on the 3ˊ-RNA fragment, and 2) hydrolysis of the 2ˊ, 3ˊ-cyclic phosphate to either a 2ˊ-phosphate or a 3ˊ-phosphate on the 5ˊ-RNA fragment 37

A.

B.

Figure 1.3. Mechanisms of RNA cleavage by ribonucleases. RNA cleavage

typically occurs through an acid-base catalysis mechanism, where a general base facilitates the formation of the nucleophile. The nucleophile attacks the scissile phosphate, hydrolyzing the P—O bond, and a general acid protonates the leaving group. A. Divalent-metal-ions like Mg2+ are important for coordinating RNA cleavage by many ribonucleases. Divalent-metal-ions coordinate the nucleophile formation (in this case a hydroxide ion), and stabilize interactions between the ribonuclease (depicted as an aspartic acid residue), the nucleophile, and the scissile phosphate. Divalent-metal-ion dependent ribonuclease cleavage results in RNA fragments with terminal 3ˊ-hydroxyl and 5ˊ-phosphate groups. B. Some ribonucleases cleave RNA independent of divalent metals, and typically make use of the 2ˊ-OH of the ribose as the nucleophile. A general base deprotonates the 2ˊ-OH to form the nucleophile that attacks the scissile phosphate, forming a 2ˊ, 3ˊ-cyclic phosphate on the 5ˊ-fragment. A general acid protonates the leaving group, resulting in a 5ˊ-hydroxyl on the 3ˊ- fragment. Hydroxide ions can lead to further hydrolysis of the 2ˊ, 3ˊ-cyclic phosphate to either a 2ˊ- or 3ˊ-phosphate on the 5ˊ-fragment. (Adapted from Wang, 2011).

(Markham and Smith, 1952). However, studies using 31P NMR to the measure the kinetics of reaction product formation during RNA catalysis with bovine pancreatic

RNase indicated that the RNA substrate, poly(U), was almost completely degraded to

U>p (uridine 2ˊ, 3ˊ-cyclic phosphate) before hydrolysis of the 2ˊ, 3ˊ-cyclic phosphates to 38 monophosphates, at a point where there was limiting amounts of phosphodiester bonds left to cleave (Cuchillo et al., 1993, 2011; Thompson et al., 1994). Other single-stranded specific endoribonucleases like RNase T1, barnase, and RNase I produced similar results with the respective RNA substrates tested (Thompson et al., 1994). In contrast, the RNA fragments generated by staphylococcal nuclease cleavage, which does not generate 2ˊ, 3ˊ-cyclic phosphates, or hydroxide ion or imidazole resulted in the generation of uridine 2ˊ or 3ˊ-monophosphates without the accumulation of 2ˊ, 3ˊ-cyclic phosphates (Thompson et al., 1994). These results suggest that RNA fragments with 2ˊ,

3ˊ-cyclic phosphates are the product, rather than the intermediate of RNase A cleavage, and that RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates might be the reaction product for many other metal-ion-independent ribonucleases that have been reported to produce RNA fragments with terminal 3ˊ-monophosphates and 5ˊ-hydroxyls.

Scope of Thesis

When I joined the Barton Lab in 2010, it was clear that RNase L restricted virus replication and that viruses had evolved strategies to evade the antiviral activities of

RNase L. So far, the antiviral mechanisms of RNase L include: 1) cleavage of viral RNA to inhibit viral gene expression and replication; 2) cleavage of rRNA to inhibit host and viral mRNA translation; 3) induction of apoptosis; and 4) generation of RNA cleavage fragments that activate type I IFN production by RIG-I/MAVS signaling pathways.

Despite over 30 years of study that led to the breadth of information about RNase

L, the precise host and viral RNA targets of RNase L remained poorly defined.

Identification of the host and viral RNAs targeted by RNase L during viral infections provides insight into the antiviral activities of RNase L. To define the host and viral RNAs targeted by RNase L during viral infections, I adapted and optimized 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing methods that rely on the specificity of 39

Arabidopsis thaliana (A. th.) tRNA ligase to enrich for RNA fragments with 2ˊ, 3ˊ-cyclic

phosphates (Schutz et al., 2010) generated by metal-ion-independent ribonucleases,

thus identifying RNA fragments produced by ribonuclease cleavage from complex

populations of host and viral RNAs.

Using these deep sequencing methods, I determined that the RNA fragments

produced by RNase L were recognized by A. th. tRNA ligase as ligation substrates, indicating that the RNA fragments produced by RNase L cleavage had terminal 2 ˊ, 3ˊ- cyclic phosphates. Illumina deep sequencing of cDNA libraries generated using A. th. tRNA ligase and viral RNAs cleaved by purified RNase L revealed the frequency, location, and specificity of RNase L cleavage sites in viral RNAs (Chapter II). As

described in Chapters III, IV and V, I used 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and

Illumina sequencing methods to identify ribonuclease cleavage sites in host and viral

RNAs from PV-infected HeLa cells (Chapter III), IAV-infected A549 cells (Chapter IV)

and HCV-infected cells and human liver tissue (Chapter V).

In these studies, I established the utility of these new methods and I made a

series of expected and unexpected discoveries. I defined a constellation of RNase L-

dependent and RNase L-independent cleavage sites in host and viral RNAs. Taken

together, my data reveal the precise nature of cellular ribonucleases. Cellular

ribonucleases have precise targets in distinct RNAs. In Chapter VI, I discuss the nature

of these targets and their biological significance as they relate to antiviral defense,

ribosome biogenesis, ribosome function, apoptosis, and cancer.

40

CHAPTER II

METHODS TO DETECT RIBONUCLEASE L CLEAVAGE IN RNA1

Introduction

RNase L is a single-strand specific endoribonuclease and the antiviral effector of the 2ˊ-5ˊ OAS/RNase L pathway (reviewed in Chapter I). The antiviral activity of RNase

L can manifest by several mechanisms: 1) cleavage of viral RNA inhibits viral gene expression and replication; 2) cleavage of rRNAs is predicted to inhibit host and viral mRNA translation; 3) RNase L activity provokes apoptosis, which can be antiviral if virus-infected cells die before virus replication is complete; and 4) RNA fragments produced by RNase L can enhance the type I IFN response by activating RLRs, RIG-I and MDA5 (Malathi et al., 2007, 2010).

HCV RNA is detected and destroyed by the OAS/RNase L pathway in HeLa cell lysates (Han and Barton, 2002; Han et al., 2004). RNase L cleaved HCV RNA predominantly at single-stranded UA and UU dinucleotides in these experimental conditions, as detected by primer extension. svRNA3, one of the HCV RNA fragments produced by RNase L cleavage, activated RIG-I and induced the downstream expression of IFNβ (Malathi et al., 2010). The ability of svRNA3 to activate RIG-I was dependent on its 3ˊ-terminal phosphate, although the specific nature of the 3ˊ-terminal phosphate (cyclic phosphate or 3ˊ phosphate) was unknown at the time. The studies characterizing svRNA3 in vitro were the first to implicate a specific RNase L-dependent

RNA fragment as a RIG-I ligand. While RNase L and OAS are expressed in HCV- infected patient liver samples (Malathi et al., 2010), it is unknown whether svRNA3 is produced in HCV-infected patients to influence HCV infections and treatment outcomes.

1 Portions of this chapter were used with permissions from Cooper, D.A., Jha, B.K., Silverman, R.H., Hesselberth, J.R., and Barton, D.J. (2014). Ribonuclease L and metal- ion-independent endoribonuclease cleavage sites in host and viral RNAs. Nucleic Acids Res. 42, 5202-5216. 41

To detect the activity of RNase L in cells and tissues, a better understanding of the features of RNA fragments produced by RNase L was needed.

RNase L prefers to cleave ssRNA 3ˊ of UA, UU, and UG dinucleotides (Floyd-

Smith et al., 1981; Wreschner et al., 1981b). Based on similar electrophoretic mobilities

of RNA fragments produced by RNase L, other ribonucleases, and chemical treatments,

it was concluded that RNase L produced RNA fragments with 5ˊ-hydroxyl and 3ˊ-

terminal phosphate groups (Carroll et al., 1996; Wreschner et al., 1981b). RNA

fragments with terminal 3ˊ- phosphate groups are produced by metal-ion-independent

ribonucleases, which include RNase A, RNase T1 and T2, IRE1, and RNase L, which all target single-stranded RNA (Table 1.1; reviewed in Yang, 2011). The specificity of

metal-ion-independent ribonucleases for single-stranded RNA is due to the mechanism

of RNA cleavage. The ribose 2ˊ-hydroxyl adjacent to the scissile phosphate (the

phosphate forming the phosphodiester bond to be cleaved) is deprotonated to activate

the nucleophile, usually by a basic residue of the RNase. Through nucleophilic attack on

the scissile phosphate, the phosphodiester bond is cleaved to form a 5ˊ fragment with a

terminal 2ˊ, 3ˊ-cyclic phosphate and a 3ˊ fragment with a 5ˊ hydroxyl group (Figure 1.3)

(Yang, 2011). 2ˊ, 3ˊ-cyclic phosphates can be hydrolyzed to linear phosphates, as is assumed to occur with the 2ˊ, 3ˊ-cyclic phosphates found on RNA fragments produced by RNase L cleavage. However, homology shared between RNase L and IRE1 suggest that RNase L, like IRE1, cleaves RNA to produce fragments with terminal 5ˊ-hydroxyl

and 2ˊ, 3ˊ-cyclic phosphates (Bork and Sander, 1993; Dong et al., 2001; Gonzalez et al.,

1999). While the RNA fragments produced by bovine RNase A cleavage are reported to

terminate with 3ˊ-phosphates, studies of RNA fragments produced by bovine RNase A

indicated that 2ˊ, 3ˊ-cyclic phosphates are the products of RNase L cleavage rather than

the intermediates (Cuchillo et al., 1993, 2011; Thompson et al., 1994). Perhaps like 42

RNase A and IRE1, RNase L also produces RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates.

To determine whether RNA fragments produced by RNase L terminate with 2ˊ,

3ˊ-cyclic phosphates, I took advantage of the specificity of Arabidopsis thaliana (A. th.) tRNA ligase which recognizes RNA substrates with 2ˊ, 3ˊ-cyclic phosphates (Schutz et

al., 2010). tRNA ligases join the 5ˊ and 3ˊ-pretRNA halves generated during tRNA

splicing of intron-containing tRNA genes (Abelson et al., 1998; Knapp et al., 1979), and

for ligating the HAC1 exons generated by the IRE1 endonuclease to produce the spliced

form of HAC1 mRNA important for regulation of the unfolded protein response in

Saccharomyces cerevisiae (Sidrauski et al., 1996). During ligation, tRNA ligase opens

the 2ˊ, 3ˊ-cyclic phosphate found at the terminus of the 5ˊ-pretRNA half to generate a 2ˊ- phosphate. Next, tRNA ligase transfers a γ-phosphate from ATP or GTP to the 5ˊ-end of the 3ˊ-exon and adenylates the 5ˊ-phosphate. Finally, ligation occurs between the 3ˊ- hydroxyl of the 5ˊ-exon and the adenylated 5ˊ-end of the 3ˊ-exon (Abelson et al., 1998).

The 2ˊ-phosphate that remains at the splice junction is removed through NAD+- dependent 2ˊ-phosphate-specific phosphotransferase (Culver et al., 1993).

For this chapter I explored the ability of A. th. tRNA ligase to recognize RNA fragments produced by RNase L. I found that RNA fragments produced by RNase L were recognized as ligation substrates by A. th. tRNA ligase. Based on these data, I concluded that RNase L produces RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates.

I then used A. th. tRNA ligase to generate cDNA libraries from viral RNAs cleaved by purified RNase L and RNase A. Illumina sequencing of these cDNA libraries revealed the frequency, location, and specificity of endoribonuclease cleavage sites in the viral

RNAs. RNA fragments produced by RNase L terminated with UA, UU, and UG dinucleotides, while those produced by RNase A terminated with pyrimidines (U and C), consistent with the reported specificity of these enzymes (Floyd-Smith et al., 1981; Han 43 et al., 2004; Raines, 2004; Wreschner et al., 1981b). As expected, the RNase L and

RNase A cleavage sites detected by 2ˊ, 3ˊ-cyclic phosphate cDNA sequencing mapped predominantly to single-stranded regions of HCV and PV RNAs, consistent with the specificity of metal-ion-independent endoribonucleases (Yang, 2011). I discuss these findings as they relate to structural and functional features in the viral RNAs. Based on the validation of the 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and sequencing methods in this chapter, I concluded that cyclic phosphate cDNA synthesis and Illumina sequencing methods are suitable for identifying and quantifying endoribonuclease cleavage sites.

Materials and Methods

Viral RNAs

svRNA3+, a 400 base long RNA from the NS5B region of HCV1A that contains the svRNA3 (Malathi et al., 2010), was generated by T7 in vitro transcription (Epicentre).

To generate svRNA3+, PCR using an infectious clone of HCV1A (P90/HCVFLlong pU)

template along with primers complementary to the NS5B region [8947_HCVT7_F (5ˊ-

GAA ATT AAT ACG ACT CAC TAT AGG AGC CTG CTA CTC CAT AGA ACC-3ˊ) and

9417_HCV_R ( 5ˊ-GGA AAT GGC CTA AGA GG -3ˊ]). 4 μL of T7-promoter tagged PCR

product was used as template for the in vitro transcription reaction (Epicentre) to make

svRNA3+ RNA. Full-length HCV1A RNA (AF009606.1) and PV type 1 Mahoney

(V01149.1) RNA were synthesized from linearized P90/HCVFLlong pU, and pMOA80

using T7 polymerase (Epicentre), as previously described (Han et al., 2004, 2007).

Endoribonucleolytic Cleavage and Ligation of the svRNA3+ Transcript

Recombinant human RNase L and trimer 2-5A [p35ˊA(2ˊp5ˊA)2] were purified as previously described (Han et al., 2007; Rusch et al., 2001; Silverman et al., 2000). For cleavage of svRNA3+, 50 μL reactions containing 200 nM svRNA3+, 40 nM RNase L 44 with or without 40 nM 2-5A, and 1 X cleavage buffer (25 mM Tris-HCl [pH 7.4], 100 mM

KCl, 10 mM MgCl2 40 µM ATP, 7 mM βME) were incubated for 20 minutes at 32°C.

Cleavage reactions were terminated with 150 μL of 0.5% SDS buffer (0.5% SDS, 10 mM

Tris-HCl [pH 7.5], 1.25 mM EDTA [pH 8], 100 mM NaCl) and RNAs were purified using phenol:chloroform:isoamyl alcohol (P:C:I) extraction and ethanol precipitation. RNA was pelleted and resuspended in either A. th. tRNA ligase reactions (10 μL total volume) containing 50 pmol RNA linker [5ˊ/Phos/CAGU-AGA UCG GAA GAG CGU CGU GUA

GGG AAA GAG/Amino Modification/-3ˊ], 1 X tRNA ligase buffer (50 mM Tris-HCl [pH

7.5], 40 mM NaCl, 5 mM MgCl2, 0.03 mM ATP, and 1 mM DTT), and 1 μM of A. th. tRNA ligase (purified as previously described in (Schutz et al., 2010)), or the RNA was resuspended in T4 RNA ligase I reactions (10 μL total volume) containing 50 pmol RNA linker, 20 U T4 RNA ligase I (New England Biolabs; M0204L) and 1 X T4 RNA ligase buffer (Provided by manufacturer: 50 mM Tris-HCl, 10 mM MgCl2, 1 mM DTT, and 1 mM

ATP), or the RNA was treated with 10 U of T4 polynucleotide kinase (T4 PNK) (New

England Biolabs; M0201S) for 20 minutes at 37°C in reactions containing 100 mM imidazole (pH 6.0), 10 mM MgCl2, 10 mM βME, 0.1 mg/mL BSA (Zhu et al., 2007), and subsequently ligated with either A. th. tRNA ligase or T4 RNA ligase 1. Ligation of RNAs with A. th. tRNA ligase or T4 RNA ligase I occurred for 1 hour at 37°C. Ligated RNAs were purified with P:C:I extraction and ethanol precipitation with 1 μL Glycoblue carrier

(Ambion; AM9515), and fractionated on 6% denaturing polyacrylamide gel.

Viral RNAs Cleaved by Ribonuclease L and Ribonuclease A

Viral RNAs (100 nM) were incubated at 30º C in reactions (50 µl volume) containing cleavage buffer (25 mM Tris-HCl [pH 7.4], 100 mM KCl, 10 mM MgCl2 40 µM

ATP, 7 mM βME), purified bovine pancreatic RNase A (Ambion) (0.5 nM), or RNase L

(25 nM) and 2-5A (25 nM). Increased amounts of RNase L (50 nM) and 2-5A (50 nM) 45 were used in reactions containing PV RNA to overcome the ciRNA (Han et al., 2007;

Townsend et al., 2008a). Reactions were terminated after the indicated periods of time

(0, 2.5, 5, 10, and 20 minutes) with the addition of 150 µL of 0.5% SDS buffer. After

P.C.I. extraction and ethanol precipitation with Glycoblue, 2.5 µg of RNA from each reaction was fractioned by electrophoresis in 1.2% agarose (MOPS-formaldehyde).

Ethidium bromide and UV light revealed the location of viral RNA and viral RNA fragments in the gel. For HCV RNA fragments used for TOPO-TA cloning and sequencing reactions, 40 nM RNase L and 2-5A were used in reactions containing 100

nM HCV RNA. These reactions proceeded for one hour at 32°C.

2ˊ, 3ˊ-Cyclic Phosphate cDNA Synthesis and Illumina Sequencing

A. th. tRNA ligase and yeast 2ˊ-phosphotransferase (Tpt1) were purified as

previously described (Schutz et al., 2010). An RNA linker with a 4-base barcode

(5ˊ/Phos/CAGU-AGA UCG GAA GAG CGU CGU GUA GGG AAA GAG/Ammo/-3ˊ) was

used to make cDNA libraries for TOPO-TA cloning experiments. An RNA linker with an

8-base unique molecular identifier (UMI) (5ˊ/Phos/NNNNNNNN-AGA UCG GAA GAG

CGU CGU GUA GGG AAA GAG-Amino Modification-3ˊ [NNNNNNNN=UMI] [Integrated

DNA Technologies]) was used to make cDNA libraries for Illumina sequencing. The RNA

linker (20 µM) was attached to RNA (2 µg) in reactions (20 µL volume) containing buffer

(50 mM Tris-HCl [pH 7.5], 40 mM NaCl, 5 mM MgCl2, 0.03 mM ATP, and 1 mM DTT), A.

th. tRNA ligase (0.75 µM), and 1 U Riboguard (Epicentre; RG90925). For ligation

reactions used for TOPO-TA cloning, Riboguard was not included. After incubation at

37ºC for one hour, RNAs were purified using P:C:I extraction and ethanol precipitation

with 1 μL Glycoblue carrier. All ethanol precipitations from here on include Glycoblue

carrier. The pelleted RNAs were resuspended for in 1 X Fragmentation Reagents

(Ambion; AM8740), and incubated for 10 minutes at 70ºC, followed by addition of the 46

Stop Solution and an ethanol precipitation to pellet the fragmented RNAs. Because a 2ˊ- phosphate persists at the ligation junction (Culver et al., 1993; Schutz et al., 2010), fragmented RNAs were dephosphorylated at 30ºC for one hour in reactions (20 µL volume) containing 1 X Tpt1 buffer (20 mM Tris-HCl [pH 7.5], 5 mM MgCl2, 0.1 mM DTT, and 0.4% Triton X-100), NAD+ (10 mM), and Tpt1 (4.5 µM). These RNAs were P:C:I extracted and ethanol precipitated. Following ligation and Tpt1 reactions, RNAs were resuspended in denaturing sample buffer (95% formamide, 18 mM EDTA, 0.025% SDS, and 0.025% each bromophenol blue and xylene cyanol), heat denatured (10 minutes at

65ºC), and fractionated on a 6% urea-polyacrylamide gel. RNAs in the gel were visualized using SYBR-Gold (Invitrogen; S-11494) and blue light transillumination. RNAs

100-500 nts in length were excised from the gel, crushed into a slurry with nuclease-free water, and incubated at 42°C for 2 hours. RNAs were purified from polyacrylamide by running samples through DTR cartridges (EdgeBio; 98780) followed by ethanol precipitation.

Superscript III (Invitrogen; 18080-044) and a DNA primer (5ˊ-AAT GAT ACG

GCG ACC ACC GAG ATC TAC ACT CTT TCC CTA CAC GAC GCT C-3ˊ) were incubated with the gel purified RNAs to make cDNA (per the manufacturer’s instructions). Exonuclease I (0.5 U) and recombinant shrimp alkaline phosphatase (0.25

U) (USB/Affymetrix; Exo1, 70073Z, rSAP, 78390) were added to each reaction, followed by additional incubation at 37ºC for 30 minutes to remove excess DNA primer and dNTPs. Reactions were heated to 95ºC to inactivate the enzymes. RNA remaining in the reverse transcription reactions was hydrolyzed using 1 N NaOH treatment at 100ºC for

10 minutes, and neutralized with 1 N HCl. cDNA was P:C:I extracted and ethanol precipitated.

A DNA linker was attached to the 3ˊ end of the cDNA in reactions (10 µL volume) containing 1 µM miRNA cloning linker 1 (5ˊ-rApp-CTG TAG GCA CCA TCA AT-ddC-3ˊ 47

[Integrated DNA Technologies]), 10 U of T4 RNA ligase I, and T4 RNA ligase buffer (50

mM Tris-HCl [pH 7.5], 10 mM MgCl2, and 1 mM DTT). Linked cDNAs were P:C:I

extracted and ethanol precipitated. The cDNAs were resuspended in formamide sample

buffer and fractionated on a 6% urea-polyacrylamide gel. cDNAs greater than 150 bases

long were gel purified and resuspended in 30 µL of water.

PCR was used to amplify cDNA libraries before Illumina sequencing. PCR

reactions (50 µL) contained 1 U Phusion High Fidelity DNA polymerase (New England

Biolabs), HF Phusion buffer, 0.2 mM dNTPs, 5 – 8 µL of cDNA template, 0.2 µM forward

indexed primer (5ˊ-CAA GCA GAA GAC GGC ATA CGA GAT XXX XXX GTG ACT GGA

GTT CAT TGA TGG TGC CTA CAG-3ˊ where XXX XXX = a distinct 6-base index

sequence used to distinguish one cDNA library from another when mixing multiple

libraries into one Illumina sequencing run), and 0.2 µM reverse primer (5ˊ-AAT GAT

ACG GCG ACC ACC GAG ATC TAC ACT CTT TCC CTA CAC GAC GCT CTT CCG

ATC T-3ˊ). Thermal cycling conditions were an initial 98°C for 30 seconds, 28 – 30

cycles of [98°C for 5 seconds, 59°C for 20 seconds, and 72°C for 30 seconds], a final

extension at 72°C for 10 minutes, and a 4°C hold. For Illumina sequencing, PCR

products 250 to 1000 bps in length were purified from a non-denaturing polyacrylamide

gel and resuspended in 10 mM Tris-HCl [pH 8.5] with 0.1% Tween-20. The amounts of

DNA in each library were determined using the Qubit fluorometer (Invitrogen). Equal

volumes of each cDNA library were combined for multiplexed sequencing for a

combined total of 10 nM of DNA for the Illumina sequencing run. A multiplexing index

read primer (5ˊ-CTG TAG GCA CCA TCA ATG AAC TCC AGT CAC-3ˊ) was used to

identify the library after sequencing on the MiSeq or GAIIx (Illumina, Inc). For TOPO-TA

cloning reactions, PCR products of 150 to 250 and 250 to 500 bases in length were gel

purified to determine optimal purification size for sequencing.

48

Bioinformatic Analyses

FASTQ data files were aligned to poliovirus type 1 Mahoney (V01149.1 ) or hepatitis C virus 1A (AF009606.1) using Bowtie (Langmead et al., 2009) with –m 1 - - best - -strata options, which discards any read with more than one reportable alignment, and reports alignments with the least number of mismatches. After alignment, the original number of RNA molecules in the library was determined by counting the number of UMI sequences at the 5ˊ-end of each aligned read using a custom python program

(J.R. Hesselberth, 2012). cDNA sequences with distinct UMIs were counted, while cDNA sequences with repetitive UMIs (same UMI sequence seen more than once on any particular RNA fragment) were discarded from further analyses. UMI-corrected cDNA analyses eliminate PCR bias (where one cDNA molecule is amplified and sequenced multiple times). To determine the position and frequency of endonuclease cleavage sites in an RNA, the UMI-corrected alignment data were converted to Bedgraph format using

Bedtools (Quinlan and Hall, 2010) (bedtools genomecov -bg -5 to report the end of the read/cleavage position). The frequency of each cleavage site was plotted against the nucleotide position of each RNA using R (R Core Team, 2013). For RNase L-cleaved

HCV, and RNase A-cleaved HCV and PV RNA data, the signal from the “no 2-5A” and

“no RNase A” was subtracted from the 0, 2.5, 5, 10, and 20 minute datasets. To determine the 3ˊ dinucleotide of each read, the last two nucleotides of each read was quantified and the sum of each of the 16 possible dinucleotides was divided by the total number of aligned reads and multiplied by 100 to get a percentage.

Data Access

The data discussed in this chapter have been deposited in NCBI's Gene

Expression Omnibus (Barrett et al., 2013) and are accessible through GEO Series accession number GSE52489 (Cooper et al., 2014). 49

Results

RNase L Cleaves RNA to Produce Fragments with Terminal 2ˊ, 3ˊ-Cyclic Phosphates

RNA fragments produced by RNase L have been reported to possess terminal

5ˊ-hydroxyl and 3ˊ-phosphate groups (Carroll et al., 1996; Wreschner et al., 1981b).

However, its metal-ion-independence and homology to IRE1 (Bork and Sander, 1993;

Yang, 2011) suggest that RNA fragments produced by RNase L terminate with 2ˊ, 3ˊ-

cyclic phosphates. To determine whether RNA fragments produced by RNase L

terminate with a 2ˊ, 3ˊ-cyclic phosphate, I used purified RNase L and 2-5A to cleave the

svRNA3+ substrate, a 400-nt long HCV RNA fragment encompassing svRNA3 RIG-I

PAMP (Malathi et al., 2010).

The mobility of svRNA3+ and its RNase L-dependent RNA fragments were

examined by urea-polyacrylamide gel electrophoresis (Figure 2.1A). The ssRNA

molecular weight marker provides one indication of resolution within the gel (Figure

2.1A, M.W. Marker, left lane of gel). The svRNA3+ transcript migrated just below the

500-nucleotide marker (Figure 2.1A, lane 10). In the absence of 2-5A, the svRNA3+

remained intact (Figure 2.1A, lanes 5-8 and lane 10), while the addition of 2-5A

provoked RNase L to cleave svRNA3+ (Figure 2.1A, lanes 1-4).

RNA fragments generated by RNase L cleavage were ligated to an RNA linker

using either A. th. tRNA ligase (Figure 2.1A, lanes 1 and 2) or T4 RNA ligase 1 (Figure

2.1A, lanes 3 and 4) with or without prior treatment with T4 PNK. In addition to kinase

activities, T4 PNK also removes 2ˊ, 3ˊ-cyclic phosphates to generate a 3ˊ-hydroxyl

(Cameron and Uhlenbeck, 1977; Das and Shuman, 2013; Zhu et al., 2007), a ligation substrate for T4 RNA ligase 1.

Incubation of the RNA fragments with A. th. tRNA ligase and an RNA linker

resulted in ligation of the RNA fragments to the linker (Figure 2.1A; compare lanes 1 to

9; note decreased mobility of RNA fragments following ligation), indicating that the RNA 50

A.

B.

Figure 2.1. RNase L generates RNA cleavage fragments with terminal 2ˊ, 3ˊ- cyclic phosphates. A. After a 20 minute incubation with 200 uM svRNA3+ substrate and 40 uM RNase L & 2-5A (Lane 9) or no 2-5A (Lane 10), RNA was further incubated in various conditions and fractionated on a 6% urea-polyacrylamide gel visualied with SYBR Gold and UV light. Lanes 1 – 4: After incubation with RNase L and 2-5A, RNA was incubated with A.th. tRNA ligase and an RNA linker (Lane 1), with A.th. tRNA ligase and RNA linker after treatment with T4 PNK (Lane 2), with T4 RNA ligase I and RNA linker (Lane 3), or withT4 RNA ligase and RNA linker after treatment withT4 PNK (Lane 4). Lanes 5-8: After incubation with RNase L without 2- 5A, RNA was incubated with A. th. tRNA ligase and RNA linker (Lane 5), with A. th. tRNA ligase and RNA linker following T4 PNK treatment (Lane 6), with T4 RNA ligase I and RNA linker (Lane 7), or with T4 RNA ligase 1 and RNA linker following T4 PNK treatment (Lane 8). B. Summary of (A). Red bars are UA and UU RNase L cleavage sites within the intact svRNA3+ substrate. The blue bar is the RNA linker. RNase L produces RNA fragments with terminal 5ˊ-OH and 2ˊ, 3ˊ-cyclic phosphates. These fragments are directly ligated by A. th. tRNA ligase. Treatment with T4 PNK removes the 2ˊ, 3ˊ-cyclic phosphate to generate ligation substrates for T4 RNA ligase.

51 fragments produced by RNase L have terminal 2ˊ, 3ˊ-cyclic phosphates. Intact svRNA3+

(incubated with RNase L without 2-5A; Figure 2.1A, lanes 5-8) was ligated by T4 RNA ligase 1 and not by A. th. tRNA ligase (Figure 2.1A; compare lanes 5 and 6 with lanes 7 and 8; note decreased mobility of svRNA3+ following incubation with T4 RNA ligase 1).

A proportion of RNA remains uncleaved even when 2-5A is added to reactions (Figure

2.1A; Lane 9, largest band). When T4 RNA ligase I is incubated with RNAs from reactions containing RNase L and 2-5A, only the intact RNA (which terminates with hydroxyl groups) is ligated while the RNA fragments produced by RNase L cleavage remain unchanged (Figure 2.1A; compare lanes 3 and 7 to lanes 9 and 10). A. th. tRNA ligase does not recognize RNA fragments treated with T4 PNK as ligation substrates,

(Figure 2.1A, compare lanes 2 and 6 to 9 and 10), but small amounts of ligation are observed when T4 RNA ligase 1 is used (Figure 2.1A; Compare lanes 4 and 9; shifts in fragment mobility observed between 200 and 300 nts, and just below full-length RNA).

Several conclusions can be drawn about RNA fragments produced by RNase L from these data, which are also diagrammed in Figure 2.1B: 1) RNA fragments produced by RNase L resist ligation by T4 RNA ligase, indicating that they do not have

3ˊ-hydroxyl groups; 2) T4 PNK treatment of RNase L-dependent RNA fragments increases their susceptibility to ligation by T4 RNA ligase but makes them unsuitable for ligation by A. th. tRNA ligase, indicating that RNase L leaves a terminal phosphate group on cleavage fragments produced by RNase L; and 3) A. th. tRNA ligase recognizes RNA fragments produced by RNase L as substrates for ligation, indicating that RNase L produces RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates. While experiments with

RNase A and other ssRNA-specific ribonucleases indicate that terminal 2ˊ, 3ˊ-cyclic phosphates are the product of cleavage, 2ˊ, 3ˊ-cyclic phosphates can slowly hydrolyze to

2ˊ or 3ˊ-phosphates (Thompson et al., 1994). A. th. tRNA ligase can recognize 2ˊ- phosphates in addition to 2ˊ, 3ˊ-cyclic phosphates, thus, it is possible that a small 52 proportion of RNA fragments with terminal 3ˊ-phosphates are missed by A. th. tRNA ligase; however, at least under in vitro conditions, nearly all RNA fragments produced by

RNase L cleavage appear to undergo ligation (Figure 2.1; compare lanes 1 to 9).

A. th. tRNA Ligase Detects RNA Fragments Produced by RNase L

RNase L prefers to cleave RNA after single-stranded UA and UU dinucleotides

(Floyd-Smith et al., 1981; Han et al., 2004; Wreschner et al., 1981). After digesting in vitro transcribed HCV RNA with RNase L and 2-5A, the RNA fragments were ligated to an RNA linker using A. th. tRNA ligase, and a cDNA library was constructed from the ligated RNAs. After PCR amplification of the cDNA library, the PCR products were gel- purified in two size groups to determine the optimal size range for sequencing of cDNAs.

One group contained PCR products in the size range of 150 – 250 bps, and the other group contained PCR products in the size range of 250 – 500 bps.

Gel-purified cDNAs from the two size ranges were TOPO-TA cloned, PCR screened with M13 primers, and 33 clones were sequenced to characterize the nature of each cDNA insert (Table 2.1). Out of the 33 clones sequenced, one was discarded for

poor quality, ten contained only linker sequence, and the remaining 22 contained HCV

sequence. The majority of clones generated from the 150 – 250 bps size-range of the

cDNA library contained only linker sequence, while 18 of the 20 clones generated from

the 250 – 500 bps size range of the cDNA library contained HCV sequence. Three of the

clones containing HCV sequence inserts represented two disparate HCV sequences

ligated together, indicating the importance of saturating RNA linker conditions to favor

ligation of the RNA linker to RNA fragments instead of ligation of RNA fragments to each

other. Only the sequence ligated directly to the RNA linker was considered for the

analysis.

53

Table 2.1. TOPO-TA cloning of 2ˊ, 3ˊ -cyclic phosphate cDNA libraries from HCV RNA cleaved by RNase L Sequence Cleavage Size Insert ID (bps) Site DAC-46 150 – 250 Linker - DAC-47 150 – 250 Linker - DAC-48 150 – 250 Linker - DAC-49 150 – 250 Linker - DAC-50 150 – 250 N/A - DAC-51 150 – 250 Linker - DAC-52 150 – 250 HCV1A UG1632 DAC-53 150 – 250 Linker - DAC-54 150 – 250 HCV1A UA1641 DAC-55 150 – 250 HCV1A UA1755 DAC-56 150 – 250 HCV1A UG1632 DAC-57 150 – 250 Linker - DAC-58 150 – 250 Linker - DAC-59 250 – 500 HCV1A UA4047 DAC-60 250 – 500 Linker - DAC-61 250 – 500 HCV1A CA7020 DAC-62 250 – 500 HCV1A UA1669 DAC-63 250 – 500 HCV1A UA9358 DAC-64 250 – 500 HCV1A GA1633 DAC-65 250 – 500 HCV1A UU2150 DAC-66 250 – 500 HCV1A UA1669 DAC-67 250 – 500 HCV1A UA8688 DAC-68 250 – 500 HCV1A UA1669 DAC-69 250 – 500 HCV1A UU1734 DAC-70 250 – 500 Linker - DAC-71 250 – 500 HCV1A UU8687 DAC-72 250 – 500 HCV1A UA4903 DAC-73 250 – 500 HCV1A UA4924 DAC-74 250 – 500 HCV1A CC5699 DAC-75 250 – 500 HCV1A UG4410 DAC-76 250 – 500 HCV1A UA4903 DAC-77 250 – 500 HCV1A UU7611 250 - 500 HCV1A UC7701 DAC-78 Sequences were manually trimmed of linker sequence and aligned to HCV1A (AF009606.1) using Clustal Omega (McWilliam et al., 2013). 54

The RNA linker used with A. th. tRNA ligase has a four-base barcode (ACTG) at its 5ˊ end. Ideally, the terminal 2ˊ, 3ˊ-cyclic phosphate of the HCV RNA fragment will be ligated directly to the barcode sequence, but six of the 22 inserts contained RNA linker without a barcode, or an incomplete linker sequence. This may be due to endoribonucleases contaminating the ligation reactions and degrading the RNA linker.

For future studies, RNase inhibitors will be used to inactivate contaminating ribonucleases.

Of the 22 sequences containing HCV inserts, 11 (50%) corresponded to an RNA fragment with a terminal UA dinucleotide, four with a terminal UU dinucleotide, three with a terminal UG dinucleotide, and four remaining sequences that terminated in UC, GA,

CA, and CC dinucleotides (Figure 2.2A; 2B; Table 2.1). The preference of RNase L to cleave after UA, UU, and UG dinucleotides was evident in the TOPO-TA cloned cDNA library (Table 2.1). The 22 inserts sequenced from the TOPO-TA clones identified 18 cleavage sites throughout HCV RNA. Ten of the 22 inserts were derived from cleavage sites within the HCV E2 envelope glycoprotein gene (Figure 2.2A), five from the NS3 gene, one from the NS4B gene, one from the NS5A gene, and five from the NS5B gene.

No inserts derived from the 5ˊ and 3ˊ non-translated regions (NTR) or genes encoding for core, E1 envelope glycoprotein, P7, NS2, or NS4A (Figure 2.2A). The 5ˊ and 3ˊ NTR of HCV RNA are highly structured features among the viral RNA, although the 3ˊ-NTR contains the polyU/C tract, which if unstructured, would contain many favorable RNase L cleavage sites.

When mapped to the secondary structure of a portion of the E2 gene (nts 1627 –

1674; Figure 2.2C), cleavage sites UG1632, GA1633, UA1641, and UA1669 map in and

around single-stranded bulges and loops. After the E2 gene, the NS3 gene was most frequently cleaved. When mapped to the secondary structure of a portion of the NS3

55

Figure 2.2. RNase L cleavage sites detected from TOPO-TA cloning of 2ˊ, 3ˊ- cyclic phosphate cDNA libraries. A. Frequency, location, and specificity of cleavage sites detected in HCV RNA. B. Quantification of 3ˊ dinucleotide from RNase L cleavage sites detected from TOPO-TA clones. C. Cleavage sites mapped onto the secondary structure of a portion of the HCV E2 gene (nts 1624 – 1675) and in a portion of the NS3 gene (nts 4899-4944). Mfold structures from sections of the E2 and NS3 genes were taken from the full-length HCV1A Mfold (NC_004102.1) (Zuker, 2003) provided by A.C. Palmenberg.

gene (bases 4899-4944; Figure 2.2D), cleavage sites UA4903 and UA4924 were located at

single stranded bulges and loops. To conclude, cDNA libraries constructed using A.th tRNA ligase reveal cleavage sites with the expected specificity of RNase L: ssRNA containing UA, UU, and UG dinucleotides. I also determined that gel-purifying PCR amplified cDNA libraries from 250 bps and above will reduce linker contamination in the

Illumina sequencing datasets.

56

Cleavage of Viral RNAs using Purified RNase L and RNase A

After using TOPO-TA cloning methods to verify that RNA fragments produced by

RNase L were specifically detected by A. th. tRNA ligase, I generated 2ˊ, 3ˊ-cyclic

phosphate cDNA libraries using viral RNAs cleaved with RNase L or RNase A (Figure

2.3). HCV and PV RNAs (Figure 2.3A) were incubated for 0, 2.5, 5, 10, and 20 minutes

in reactions containing RNase L with 2-5A (Figure 2.3B) or RNase A (Figure 2.3C).

RNase L and RNase A generated viral RNA fragments ranging from 100 to several

thousand bases in length (Figure 2.3B and C). In contrast to cleavage by RNase A

where all full-length RNA is digested, a small portion of full-length PV and HCV RNA

remain intact in reactions containing RNase L. Full-length svRNA3+ also remained in

reactions containing RNase L and 2-5A (Figure 2.1; Lanes 9 and 11). Specific cleavage

intermediates were generated in a consistent pattern, indicating that RNase L and

RNase A do not randomly target RNA (Figure 2.3B and C).

2ˊ, 3ˊ-Cyclic Phosphate cDNA Synthesis and Illumina Sequencing Methods

After optimizing and validating various experimental conditions, I adopted the

following strategy for cDNA synthesis and deep sequencing (Figure 2.4). The RNA

fragments with 2ˊ, 3ˊ-cyclic phosphates are attached to defined RNA linkers containing an eight-base long unique molecular identifier (UMI) sequence (Kivioja et al., 2011) by A. th .tRNA ligase. After ligation, the UMI sequence at the 5ˊ end of the RNA linker is

juxtaposed next to the endoribonuclease cleavage site at the end of the RNA fragment

as illustrated by the terminal UA dinucleotide in the RNA fragment. A DNA primer

complementary to the RNA linker is used to make cDNA, followed by the addition of a

DNA linker to the 3ˊ end of the cDNA. PCR amplification produces cDNA libraries

suitable for Illumina sequencing. The use of PCR primers containing defined index

57

A.

B. C.

Figure 2.3. Viral RNA fragments produced by RNase L and RNase A. Hepatitis C virus (HCV) and poliovirus (PV) RNAs were incubated with RNase L and RNase A to produce RNA fragments for 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and sequencing. Agarose gel electrophoresis and EtBr staining revealed the size of viral RNA fragments. A. Diagram of HCV and PV RNAs. HCV RNA is 9,648 bases long. PV RNA is 7,500 bases long. B. Viral RNAs incubated with RNase L. HCV and PV RNAs were incubated with RNase L for 20 minutes in the absence of 2-5A (Lanes 2 and 8), or with RNase L and 2-5A for 0 min, 2.5 min, 5 min, 10 min, and 20 min (Lanes 3-7 for HCV and lanes 9-13 for PV). C. Viral RNAs incubated with RNase A. HCV and PV RNAs were incubated for 20 minutes in the absence of RNase A (Lanes 2 and 8), and the presence of RNase A for 0 min, 2.5 min, 5 min, 10 min, and 20 min (Lanes 3- 7 for HCV and lanes 9-13 for PV).

sequences allows for the combination of several cDNA libraries into each Illumina

sequencing reaction. cDNA sequencing and bioinformatic analyses of the data reveal

the frequency and location of endoribonuclease cleavage sites in host and viral RNAs.

Frequency, Location, and Specificity of Endoribonuclease Cleavage Sites in Viral RNAs

lllumina sequencing of 2ˊ, 3ˊ-cyclic phosphate cDNA libraries generated from the viral RNA fragments from Figure 2.3, revealed the frequency, location, and specificities of RNase L and RNase A. The number of reads corresponding to cleavage sites in the 58

Figure 2.4. 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods. Diagram outlining 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods (Schutz et al., 2010). A. th. tRNA ligase recognizes RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates. The RNA fragments with 2ˊ, 3ˊ- cyclic phosphates are covalently attached to defined RNA linkers containing an 8- base-long unique molecular identifier (UMI) sequence (Kivioja et al., 2011). After ligation, the 8 base UMI sequence at the 5ˊ end of the RNA linker is juxtaposed next to the endoribonuclease cleavage site at the end of the RNA fragment as illustrated by the UA dinucleotide in the RNA fragment. A DNA primer complementary to the RNA linker is used to make cDNA, followed by the addition of a DNA linker to the 3ˊ end of the cDNA. PCR amplification produces cDNA libraries suitable for Illumina sequencing. Defined index sequences in PCR primers allow for the combination of several cDNA libraries into each Illumina sequencing reaction. cDNA sequencing and bioinformatic analyses of the data reveal the frequency and location of endoribonuclease cleavage sites in host and viral RNAs.

59 viral RNAs increase from 0 to 20 minutes (Tables 2.2 and 2.3). RNase L and RNase A cleavage sites in the viral RNAs were reproducible throughout each time point. RNase L cleaved HCV and PV RNAs predominantly at UpNp dinucleotides (UA and UU > UG), with prominent cleavage at distinct locations in the viral RNAs (Figures 2.5A and 2.6A).

Pyrimidines were the most common nucleotides at the end of viral RNA fragments

produced by RNase A (Figure 2.5B and 2.6B), consistent with the known specificity of

RNase A (Raines, 2004; Rushizky et al., 1961).

Table 2.2. 2’, 3’-cyclic phosphate cDNA synthesis and Illumina sequencing: HCV RNA

60

Table 2.3. 2’, 3’-cyclic phosphate cDNA synthesis and Illumina sequencing: PV RNA

To address potential bias in A. th. tRNA ligase, or potential unrecognized substrate specificities for RNase L and RNase A, I examined the sequence five bases upstream and five bases downstream from the cleavage sites using MEME (Figure 2.7)

(Bailey et al., 2009). As expected, RNase L demonstrated a bias toward uridine one base upstream (position 4) from the RNase L cleavage site (Figure 2.7 A and B; cleavage is after position 5 and is indicated with an arrow), while at position 5, RNase L had a preference toward uridine and adenosine, consistent with reported sequence specificity of RNase L (UA and UU dinucleotides). Downstream of the cleavage site at position 6 in HCV and 6, 7, and 8 with PV RNA, there was a bias toward uridine and adenosine. This likely reflects some of the sequence differences in HCV and PV RNA

(HCV is more GC-rich and PV is more AU-rich), and may indicate that RNase L 61

The HCV RNAs in Figure 2.5. Endoribonuclease cleavage sites in HCV RNA. Figure 2.3 were used for 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing. Location and frequency of RNase L cleavage sites in HCV RNA. X- A. axis: Nucleotide position in HCV RNA. Y-axis: Number of distinct UMI linkers detected at each cleavage site. Dinucleotide specificity of RNase L cleavage sites B. in HCV RNA. X-axis: Dinucleotide at the 3ˊ end of HCV RNA fragments. Y-axis: Percent of total cDNA reads. Location and frequency of RNase A cleavage sites in C. HCV RNA. D. Dinucleotide specificity of RNase A cleavage sites in HCV RNA.

62

Figure 2.6. Endoribonuclease cleavage sites in PV RNA. The PV RNAs shown in Figure 1 were used for 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing. A.Location and frequency of RNase L cleavage sites in PV RNA. X-axis: Nucleotide position in PV RNA. Y-axis: Number of distinct UMI linkers detected at each cleavage site. B. Dinucleotide specificity of RNase L cleavage sites in PV RNA. X-axis: Dinucleotide at the 3ˊ end of PV RNA fragments Y-axis: Percent of total cDNA reads. C. Location and frequency of RNase A cleavage sites in PV RNA. X- axis: Nucleotide position in PV RNA. Y-axis: Number of distinct UMI linkers detected at each cleavage site. D. Dinucleotide specificity of RNase A cleavage sites in PV RNA. X-axis: Dinucleotide at the 3ˊ end of PV RNA fragments. Y-axis: Percent of total cDNA reads.

63 cleavage sites tend to be in UA-rich stretches. For RNase A, the expected pyrimidine bias at position 5 was detected; however, I also observed extreme bias for an adenosine at position 6 (Figure 2.7C and D). In early studies characterizing bovine RNase A,

Witzel and Barnard observed that the rate of RNase A cleavage was dependent on the base 3ˊ to the cleavage site, and adenosines influenced this rate at the greatest magnitude (Witzel and Barnard, 1962). No obvious sequence bias common to all four

data sets was detected upstream of the cleavage site, indicating that A. th. tRNA ligase

likely does not have a substantial preference for specific bases.

Figure 2.7. Specificity of RNase L and RNase A. To assess enzyme specificity and potential bias, the bases -5 and +5 from the cleavage site (between position 5 and 6, X-axis) were analyzed by MEME (Bailey et al., 2009) to assess enzyme bias. The top 10% of sites most frequently cleaved in A. PV RNA by RNase L, B. HCV RNA by RNase L, C. PV RNA by RNase A, and D. HCV RNA by RNase A, were assessed. Logo 1 is represented for each dataset.

64

Both RNase L and RNase A targeted single-stranded regions of HCV RNA, a highly structured RNA (Figure 2.8A), especially within regions of the E2 gene (Figure

2.8B and C). Frequent amounts of RNase L cleavage within the E2 gene were detected at several sites: UU1727, UU1734 and UA1755, and frequent amounts of RNase A cleavage were detected at UC1671, AC1676, UC1682 and CU1754. In addition, RNase L targeted a

distinct region in the HCV NS3 gene, UU4422 and UU4465. I mapped RNase L cleavage

sites from the 10 minute time point onto three sections of HCV RNA: two regions of the

E2 gene: Structure 1 (nucleotides 1600-1703; Figure 2.8B) Structure2 (nucleotides

1744 – 1818; Figure 2.8C) and a portion of the NS3 gene: Structure 3 (nucleotides 4415

– 4503; Figure 2.8D). Frequently cleaved sites (>=99th percentile of cDNA reads;

darkest blue) map to single stranded regions of the stem-loop structures (UA1755, UA4422,

and UU4465). RNase L also targeted the regions surrounding the frequently targeted sites. This was best demonstrated by the descending frequency of cleavage that radiated around the frequently cleaved site at UU4465 (Figure 2.8D). UU4465 is located within a loop structure that contains other favorable RNase L cleavage sites: UU4466,

UU4467, and UA4468. These sites were reduced in frequency compared to UU4465. This

may suggest a limitation to the detection of cleavage sites in RNA through linker ligation

and sequencing methods, where favorable substrates are potentially cleaved so

efficiently that the fragments are too small to detect.

Sites targeted by RNase A (>=99th percentile of cDNA reads) are circled in black

1755 in Figures 2.8B-D. The single-stranded bulge where RNase L frequently cleaved UA

was also efficiently targeted by RNase A (Figure 2.8C). In addition, while the stem-loop

structure in Figure 2.8C does not contain any favorable RNase L sites, it does contain

several sites targeted by RNase A: CU1780, UC178, AC1784. RNase A cleavage sites were also detected in the NS3 structure depicted in Figure 2.8D, each 3ˊ to a pyrimidine

65

Figure 2.8. HCV RNA secondary structures associated with RNase L and RNase A cleavage sites. A. HCV genotype 1A RNA secondary structure (NC_004102.1) determined by Mfold (Zuker., 2003). The MinE structure of HCV is represented. Colors of the Mfold structure represent low (red), medium (yellow) and high (purple) P-num values, as described previously (Palmenberg and Sgro, 1997). HCV image and Mfold were provided by A.C. Palmenberg. RNase L (from 10 min) and RNase A (from 2.5 min) were plotted onto HCV RNA structures (B., C., and D.). For RNase L, 99th percentile refers to sites where the number of reads exceeds 249, 95th percentile: >= 53 reads, and 90th percentile: >= 24 reads. For RNase A, 99th percentile: >= 478 reads, 95th percentile: >= 102 reads, and 90th percentile: >= 43 reads. B. HCV RNA structure 1 (E2 gene, nts 1600 – 1703). C. HCV RNA structure 2 (E2 gene, nts 4437 – 4477). D. HCV RNA structure 3 (NS3 gene, nts 4415 – 4503).

base . UU4467 was targeted by both RNase L and RNase A. Among the most frequently targeted regions of HCV RNA by RNase A was a region from within the HCV E2 structure (nts 1600-1713; Figure 2.8B). A stretch of RNA from nts 1667 – 1691 includes

three single-stranded bulges in succession. This region of RNA was readily targeted by

RNase A, especially at UC1671, AC1676, and UC1682, three of the most frequently cleaved sites in the dataset. RNase L (cleavage sites highlighted in blue) also targeted several 66 sites within this region, including a site at UU1681. The overlap in regions targeted by

RNase L and RNase A suggest that the HCV RNA structure contains single-stranded regions that are susceptible to cleavage by single-strand specific endoribonucleases.

svRNA3, an RNA ligand for RIG-I, is liberated when RNase L cleaves the HCV

NS5B gene at UA9191 and a UU9281 UU9282 or UA9283 (Figure 2.9; (J.-Q. Han and Barton,

2002; J.-Q. Han et al., 2004; Malathi et al., 2010)). To determine whether the svRNA3 was liberated by RNase L in my experiments, I mapped the frequency and location of cleavage sites from nts 9000 to 9400 (Figure 2.9). Cleavage was detected at UA9191,

UU9281, UU9282, and UA9283 by the 5 minute time point; however, the frequency was very

low (less than 10 cDNA reads for each cleavage site). By the 20 minute time point,

there were 50 cDNA reads that corresponded to UA9191 and 47 cDNA reads that

corresponded to UU9282. Cleavage at sites that liberate the svRNA3 may be dependent

on cleavage of other sites UU9016 and UA9388, which are detected at the 2.5 min time

point. Perhaps cleavage at UU9016 and UA9388 liberate a fragment from the highly structured region of the HCV NS5B gene that then becomes susceptible to cleavage at

UA9191 and UU8281, UU9282, or UA9283 to produce the svRNA3. Another cleavage site,

UC9209, downstream of UA9191, was also detected; however, cleavage at UC9209 probably

does not impact the overall structure of svRNA3 substantially, and may also produce a

ligand for RIG-I.

PV RNA, like HCV RNA, was targeted by RNase L and RNase A in similar

regions (Figure 2.6). Three independent regions of cleavage, deemed the endoribonuclease susceptible sites, were evident in PV RNA: region 1 (nts 884 to 1042); region 2 (nts1705 to 1827); and region 3 (nts3088 to 3221) (Figure 2.10). Mapping

RNase L cleavage sites onto the secondary structure of PV RNA demonstrated that frequently targeted sites tend to be within and around single-stranded regions of RNA;

67

Figure 2.9. svRNA3 is liberated by RNase L cleavage. A. Schematic of the release of svRNA3 by RNase L cleavage at UA9191 and UU9281, UU9282, or UA9283 (Adapted from Malathi et al., 2010). B. Frequency, location, and specificity of cleavage sites from nts 9000 – 9400. Gray shading highlights the location of svRNA3.

68

Figure 2.10. Endonuclease susceptible regions of PV RNA. A. Secondary structure of PV RNA. Endonuclease susceptible regions 1, 2, and 3 (green). Mfold data provided by A.C. Palmenberg and imaged using R2R (Weinberg & Breaker, 2011). Data from 10 minute time point of RNase L data highlighted for each

susceptible region in B, C, and D with number of cDNA reads detected in each time point for most targeted (>=1000 cDNA reads) sites. 99th percentile >=365 cDNA reads; 95th percentile >=76 cDNA reads; 90th percentile >=31 cDNA reads.

however, there were some frequently cleaved sites within predicted double-stranded regions with poly(U) stretches. Unlike HCV RNA, which is highly structured (Figure

2.8A), PV RNA is structurally flexible aside from the structured 5ˊ and 3ˊ NTRs, and

thus, the Mfold structure of PV RNA provides a potential structure within a population of

very dynamic RNA molecules (Palmenberg and Sgro, 1997). As highlighted in Figure 69

2.10, UA928, UA1016, and UA961, were three of the most frequently cleaved sites within region 1, each corresponding to more than 1000 cDNA reads. UA1803, UA1738, UA1755, and UU1737 were the four most frequently cleaved sites in region 2, each corresponding

to more than 2000 cDNA reads. UU3102 and UU3103 were the most frequent RNase L cleavage sites within region 3, corresponding to more than 2000 cDNA reads at each site. The frequency of cleavage at these highly targeted sites increased during the time course (Figure 2.10; tables below B, C, and D). Interestingly, some overlap occurred with the endoribonuclease susceptible regions of PV RNA and regions that encoded portions of the capsid proteins recognized by neutralizing antibodies (Figure 2.11).

Figure 2.11. Antibody neutralization escape mutations within and near endonuclease susceptible regions of PV RNA. A. Locations of antibody neutralization escape mutations in poliovirus RNA (pink). Capsid protein epitopes and corresponding neutralization escape mutations reported by Page et al., 1988. B. Region 1 (epitopes in VP2 are adjacent to endonuclease susceptible region 1 of PV RNA). C. Region 2 [an epitope in VP2 (Arg270 / nts 1756 – 1758) maps within endonuclease susceptible region 2 of PV RNA]. D. Region 3 (neutralization epitopes in VP1 map within endonuclease susceptible region 3 of PV RNA).

70

Discussion

Features of RNA Fragments Produced by RNase L

Several lines of evidence suggest that RNase L cleavage produces RNA

fragments with 5ˊ-hydroxyl and 2ˊ, 3ˊ-cyclic phosphate termini: 1) Early studies suggested that RNase L terminated with 3ˊ-phosphate groups (Carroll et al., 1996;

Wreschner et al., 1981b). 3ˊ-phosphate groups on RNA fragments produced by ribonuclease cleavage indicate that cleavage occurred through transphosphorylation, where a 2ˊ, 3ˊ-cyclic phosphate is produced, or is generated as an intermediate that is later hydrolyzed to 2ˊ- or 3ˊ-phosphates (Raines, 2004; Yang, 2011); 2) RNase L shares homology with IRE1, an endoribonuclease involved in the unfolded protein response that produces RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates (Bork and Sander, 1993;

Gonzalez et al., 1999); and 3) Like IRE1, RNase A and RNase T1, RNase L is a metal- ion-independent endoribonuclease that uses the 2ˊ-hydroxyl adjacent to the scissile phosphate as a nucleophile to mediate cleavage and form a 2ˊ, 3ˊ-cyclic phosphate

(Yang, 2011).

As I have shown in this chapter, a detectable portion of the 3ˊ-phosphate found on RNA fragments produced by RNase L and another metal-ion endoribonuclease,

RNase A, is in the form of a 2ˊ, 3ˊ-cyclic phosphate. Using A. th. tRNA ligase that specifically recognizes RNAs with terminal 2ˊ, 3ˊ-cyclic phosphates/2ˊ-phosphates, I generated cDNA libraries that captured the specificities of RNase L and RNase A, with the goal of developing methods to detect RNA fragments produced by RNase L from virus-infected cells, and from HCV-infected patients to determine whether RNase L is active to produce RIG-I ligands like the svRNA3 (Figure 2.9) (Malathi et al., 2010).

71

Detection of Endoribonuclease Cleavage Sites in Host and Viral RNAs

The Barton Lab’s original interest in methods to detect RNase L cleavage sites in

RNA stemmed from previous studies using RNase L and HCV RNA. Primer extension of

HCV RNA cleaved by purified RNase L revealed 297 cleavage sites in HCV RNA (Han

et al., 2004), but this method suffered from several limitations including the use of

multiple primers (30 primers across the HCV RNA genome), variable hybridization

efficiencies for each primer, and decreased sensitivity of cleavage site detection as the

distance increased from the site of primer hybridization. In addition, while primer

extension is a suitable method for detecting cleavage sites in known RNAs, it is not

suitable for identification of unknown RNA targets of RNase L, like those that might be

found in diverse pools of cellular RNA. To overcome the limitations of primer extension, I

optimized 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods

(Schutz et al., 2010) to detect the viral RNA fragments produced by cleavage with purified RNase L and RNase A.

2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing efficiently detected RNase L cleavage sites in HCV RNA (Figure 2.4). The 8-base UMI sequence in RNA linkers, which corresponds to 65,536 unique molecules, allowed for precise quantification of cleavage sites (Kivioja et al., 2011). I detected 265 of the 297 RNase L cleavage sites in HCV RNA originally reported by primer extension (Han et al., 2004a).

While both primer extension methods and 2ˊ, 3ˊ-cyclic phosphate cDNA libraries detected the prominent cleavage site at UA1755, sites at UU1734 and UU1727 were missed by primer extension, highlighting the sensitivity of the 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and sequencing methods. Using these methods, I found that RNase L cleaves

HCV RNA predominantly within and around single-stranded regions at UA and UU

dinucleotides (Figures 2.5A and 2.8). 72

RNase L cleavage of HCV RNA liberates a small RNA, the svRNA3, derived

from the NS5B gene that refolds into a PAMP for RIG-I (Malathi et al., 2010). Whether

RNase L is activated in HCV-infected patients has yet to be determined; however, 2ˊ, 3ˊ- cyclic phosphate cDNA libraries from HCV RNA cleaved with purified RNase L indicated that RNase L can target these regions of highly structured HCV RNA (Figure 2.8A) to liberate the svRNA3 (Figure 2.9), but only after substantial RNA cleavage within other regions of HCV had already occurred, suggesting that the release of svRNA3 is dependent on primary cleavage sites that expose sites UA9191, UU9281, UU9282, and

UA9283 that liberate the svRNA3 (Figures 2.5A and 2.9). Whether the svRNA3 is

produced in patients with HCV infection remains unanswered. HCV encodes the NS3/4A

protease that cleaves adaptor molecule MAVS from the mitochondrial membrane to

inhibit RIG-I and MDA5-dependent induction of type I IFN (Li et al., 2005a, 2005b), and

therefore HCV can potentially inhibit the induction of type I IFN in response to svRNA3

activation of RIG-I. However, sequence analysis of distinct genotypes of HCV suggests

that RNase L exerts selective pressures on HCV. HCV RNAs from distinct genotypes

have variably reduced frequencies of UA and UU dinucleotides (Washenberger et al.,

2007). HCV genotypes associated with unfavorable treatment outcomes (when

treatment consisted of pegylated IFNα and ribavirin), tended to have reduced

frequencies of UA and UU dinucleotides within their genomes. The potential role(s) of

RNase L in HCV infections and IFN-based antiviral therapies deserves further

examination now that a method has been developed to detect and quantify RNase L

cleavage sites in host and viral RNAs.

I also observed that RNase L, as well as RNase A, targeted HCV RNA within the

E2 envelope gene more frequently than other regions (Figure 2.5). E2 and E1 envelope

glycoproteins form a heterodimer complex that mediates viral entry (De Beeck et al.,

2000; Lindenbach and Rice, 2013), and the hypervariable region 1 (HVR1) located 73 within the first 12 amino acids of the E2 gene is the most variable region in the HCV genome (Cortés et al., 2014; Kato et al., 1992). Hypervariable regions within the E2 gene contribute to viral quasispecies and antibody escape mutations and have made the development of vaccines against HCV challenging. Although the HVR within the E2 gene do not directly overlap with regions targeted by RNase L (data not shown), the

Mfold structure and p-num values of full-length HCV RNA (Figure 2.8A) demonstrate that regions within the E2 gene tend to be less structured. The sequence plasticity of the

HVR in the E2 gene potentially influences the RNA structure in and around those regions, and may result in increased susceptibility to endoriboucleases in these adjacent regions. Once determined, the tertiary interactions of HCV RNA might explain why some regions of HCV RNA are more susceptible to endoribonuclease cleavage than others.

Endoribonuclease cleavage sites across PV RNA genomes have not been previously described. The striking distribution of RNase L and RNase A cleavage sites in

PV RNA (Figure 2.6), with prominent amounts of cleavage in discrete regions of the genome, was not expected. PV RNA, like HCV RNA, may assume secondary and tertiary structures that render some portions of the genome particularly susceptible to single-strand specific endoribonucleases while rendering other regions resistant (Figure

2.10). Antigenic variation in the capsid proteins, as a consequence of antibody selection,

could impact the structure of the capsid genes and render these regions more sensitive

to cleavage by single-strand specific endoribonucleases, due to the counter-selective

forces of antigenic variation and resistance to endoribonucleases (Figure 2.11). In addition, it is also a possibility that higher order oligomers of RNase L (Han et al., 2012) and RNase A (Gotte and Libonati, 2004) could anchor activated nucleases on particular

regions of viral RNA, leading to cleavage within localized regions. However, the

predicted secondary structures of HCV and PV RNAs do provide insight into the

particular regions targeted by RNase L and RNase A (Palmenberg and Sgro, 1997). 74

Frequently targeted regions of HCV and PV RNA tended to localize in and around single-stranded regions of the viral genomes, consistent with the specificity of metal-ion- independent endoribonucleases (Figures 2.5 and 2.6). In addition, atomic force microscopy of HCV and PV RNA genomes reveals globular structures, where some regions of the viral RNA genomes might be more exposed to endoribonucleases than others (Davis et al., 2008; Kuznetsov et al., 2005). Intriguingly, antibody neutralization escape mutations are found within and near endonuclease susceptible regions of PV

RNA (Fiore et al., 1997; Page et al., 1988); however, the peaks of endonucleolytic cleavage in the PV capsid genes do not overlap exclusively with the location of RNA changes associated with antigenic variation (Figure 2.11). A complete understanding of

the discrete distribution of endoribonuclease cleavage sites in PV RNA awaits further

investigation.

Summary

The experiments described in this chapter demonstrate that 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods efficiently detect RNase L

and other metal-ion-independent endoribonuclease cleavage sites in RNA. In Chapters

III – V, I use these same methods to detect ribonuclease cleavage from virally-infected

cells to determine the in vivo targets of RNase L. Particular RNAs of interest are viral

RNAs and host rRNAs. Because many viruses, including PV encode mechanisms to

counteract RNase L, RNase L may directly cleave viral RNA to limit viral replication.

While these ideas have been postulated for years, very little direct evidence exists to

demonstrate that RNase L directly cleaves viral RNA in the context of infections. rRNA is

characteristically fragmented when RNase L is activated and has long been used as an

indicator of RNase L activation, yet the cleavage sites in rRNAs have not been carefully

defined. 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing of RNA from cells 75 where RNase L is activated will identify the precise cleavage sites in rRNAs, important

markers of RNase L activation. 76

CHAPTER III

POLIOVIRUS INFECTION: RIBONUCLEASE CLEAVAGE SITES

IN HOST AND VIRAL RNAS2

Introduction

In Chapter II, I described a 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina

deep sequencing method that identified endoribonuclease cleavage sites in RNA.

Because the precise RNA targets of RNase L are not well characterized, I applied these

methods to RNA isolated from HeLa cells where RNase L was activated by poliovirus

(PV) infection to induce cleavage of rRNAs, the characteristic signature of RNase L

activation.

PV is the prototypic member of the Picornaviridae family of viruses and is the

cause of poliomyelitis in humans. PV encodes a single positive-strand RNA genome of

~7500 nts in length, encapsidated in naked icosahedral capsids (Hogle et al., 1985).

After infection and uncoating, the PV (+) strand IRES-containing genome is directly translated into a polyprotein by cellular translation machinery. The polyprotein is proteolytically cleaved by viral proteases into its functional units composing structural and non-structural proteins. To favor synthesis of its own viral proteins, PV mediates host cell translational shut off by viral 2A-protease-dependent cleavage of eIF4G, an initiation factor important for cap-dependent translation, and by viral 3C-protease- dependent cleavage of poly(A)-binding protein which is suggested to inhibit ribosome recycling (Bovee et al., 1998; Gradi et al., 1998; Kräusslich et al., 1987; Kuyumcu-

Martinez et al., 2004). Replication to produce progeny virus genomes occurs within

2 Portions of this chapter were used with permissions from Cooper, D.A., Jha, B.K., Silverman, R.H., Hesselberth, J.R., and Barton, D.J. (2014). Ribonuclease L and metal- ion-independent endoribonuclease cleavage sites in host and viral RNAs. Nucleic Acids Res. 42, 5202-5216. 77 enclosed membrane structures to prevent the detection of viral dsRNA replication

intermediates (Ahlquist, 2006; Richards et al., 2014).

PV is notoriously resistant to IFNα treatment (Morrison and Racaniello, 2009). To

prevent initiation of type I IFN and counteract the antiviral affects of IFN, viral 3C

protease cleaves RIG-I, and its viral 2A protease counteracts the antiviral effects of IFNα

through unknown mechanisms (Barral et al., 2009; Morrison and Racaniello, 2009). PV

also encodes an RNA structure that acts as a competitive inhibitor of RNase L (Han et

al., 2007; Townsend et al., 2008a). The RNase L ciRNA is evident in PV and related

group C enteroviruses (Townsend et al., 2008b). Despite the presence of the ciRNA,

RNase L is activated during the course of PV infection in HeLa cells, and under

experimental conditions using low MOIs and plaque assays, PV spreads more readily

from cell-to-cell in the presence of RNase L activity than in the absence of RNase L

activity, producing larger plaques than those from cells expressing dominant negative

RNase L (R667A mutation) (Han et al., 2007). Thus, PV may be able to both evade and

co-opt RNase L activity in virus-infected cells.

When I entered the Barton lab, I became interested in the targets of RNase L and

the mechanisms by which a virus might evade and/or co-opt RNase L activity. How can

PV RNA escape the antiviral activity of RNase L? Is PV RNA cleaved by RNase L in

virus-infected cells? When I started thesis work in the Barton lab, there were no

precisely characterized RNase L-dependent cleavage sites defined in host or viral RNA

from infected cells, apart from the unspecified cleavage sites in rRNAs which serve as a

characteristic signature of RNase L activity in virus-infected cells (Silverman et al., 1983;

Wreschner et al., 1981a), where both 28S and 18S rRNA are cleaved in an RNase L-

dependent fashion (Silverman et al., 1983). Although RNase L activation is frequently

associated with rRNA cleavage, certain viral infections induce RNase L-independent

cleavage of rRNAs (Banerjee et al., 2000). While two RNase L-induced cleavage site in 78

28S rRNA as been mapped through primer extension (Iordanov et al., 2000), a comprehensive analysis of RNase L-induced changes in host and viral RNAs has yet to be accomplished.

As described in Chapter II, I optimized and validated 2ˊ, 3ˊ-cyclic phosphate

cDNA synthesis and sequencing methods to identify and quantify endoribonuclease

cleavage sites in RNA. I used this method to characterize RNase L-dependent and

RNase A-dependent cleavage sites in HCV and PV RNA in vitro. In this chapter, I deployed the cyclic phosphate cDNA synthesis methods to characterize the host and viral RNAs cleaved by ribonucleases in PV-infected cells. I generated and sequenced

2ˊ, 3ˊ-cyclic phosphate cDNA libraries using RNA from uninfected- and PV-infected

HeLa cells under conditions where RNase L was activated during the course of the virus

infection (Han et al., 2007). Using these methods, I identified a constellation of RNase L-

dependent and RNase L-independent cleavage sites in host and viral RNAs. I detected

an RNase L-dependent cleavage site within PV RNA and several RNase L-induced

cleavage sites in rRNAs. I mapped rRNA cleavage sites onto the secondary and tertiary

structures of the human 80S ribosome (Anger et al., 2013), revealing that RNase L-

dependent cleavage sites are surface exposed, where they would be accessible to

RNase L. In addition, I detected 2ˊ, 3ˊ-cyclic phosphates at the end of U6 snRNA,

consistent with the enzymatic activity of a 3ˊ→5ˊ exoribonuclease, hUSB1 (Mroczek et

al., 2012; Shchepachev and Azzalin, 2013; Shchepachev et al., 2012). Lastly, and

unexpectedly, my data indicate that a 2ˊ, 3ˊ-cyclic phosphate is present at the end of 5S

rRNA. I discuss these findings as they relate to RNase L-dependent and RNase L-

independent cleavage sites in host and viral RNAs.

79

Materials and Methods

RNA from Uninfected and PV-Infected W12 and M25 HeLa Cells

HeLa cells transfected with pcDNA3 vectors expressing wild-type RNase L (W12

HeLa cells) or a dominant-negative R667A mutant form of RNase L (M25 HeLa cells) were grown as previously described (Han et al., 2007) in Dulbecco’s Modified Eagle’s medium (DMEM; Gibco) containing 10% fetal bovine serum, 250 µg per mL G418, 100 U per mL penicillin, and 100 µg per mL streptomycin (Hyclone). Cells were infected with 10

PFU per cell of poliovirus diluted in PBS (~1.2 X 106 cells per 35 mm well). Following one hour of virus adsorption at room temperature, the inoculum was removed and replaced with 2 mL of DMEM containing fetal bovine serum, pen-strep, and G418. Mock infections were performed using PBS without virus. Mock-infected and PV-infected cells were incubated at 37ºC. At the indicated times (0, 2, 4, 6 and 8 hours post-adsorption),

RNA was isolated from the cells using guanidine thiocyanate disruption (4 M guanidine thiocyanate, 25 mM sodium citrate, 0.5% N-laurylsarcosine, and 0.1 M βME), acid phenol-chloroform extraction, and ethanol precipitation. Host and viral RNAs in each sample were separated by denaturing agarose gel electrophoresis and visualized using ethidium bromide and UV light. PV was isolated from a parallel set of cells and quantified by plaque assays to monitor the magnitude and kinetics of virus replication (Han et al.,

2007).

2ˊ, 3ˊ-Cyclic Phosphate cDNA Synthesis and Illumina Sequencing

Cyclic phosphate cDNA libraries were prepared using 5 μg of total cellular RNA and were generated and sequenced as described in Chapter II (Figure 2.4).

80

Bioinformatic Analysis

Raw FASTQ data were trimmed of poor quality bases and contaminating linker

using Cutadapt (code.google.com/p/cutadapt/). The 8-base unique molecular identifiers

were trimmed from the sequences and reported in the FASTQ record name

(github.com/brwnj/umitools). FASTQ data were aligned using Bowtie2 with default

parameters or local mode with the - -un option in a sequential manner to the following

sequences: poliovirus type 1 Mahoney (V01149); 45S pre-rRNA transcript

(NR_046235.1) containing 18S, 5.8S, and 28S rRNA sequences; 5S rRNA

(NR_023371.1) using - -local; U6 snRNA (NR_046491.1), U1 snRNA (NR_004430.2),

U3 snoRNA (snoRNA-LBME) and SRP RNA (hg18 chr14:49123048-49123346) using --

local; hg18 tRNAs (http://gtrnadb.ucsc.edu/) using - -local; and Homo sapiens

hg18/NCBI36. The aligned data were UMI-corrected and filtered so that only uniquely

mapped reads with MAPQ scores 10 and higher were used in the analysis (Kivioja et al.,

2011). For reads reported using local mode alignment, non-templated reads were not tolerated at the end of the read that reported the cleavage site. For alignments to hg18, reads with more than one reported alignment were tolerated as long as the MAPQ score was 10 or above. Because of the repetitive nature of tRNA genes, all reported alignments were considered for the analysis. Alignment bam files were converted to bedgraph using Bedtools (Quinlan and Hall, 2010) and data were normalized to % total

cDNA reads (# UMI reads at a position / total # of UMI reads in library)*100. Plots were

generated using (R Core Team, 2013) and Microsoft Excel. To determine whether a

cleavage site was RNase L-dependent or RNase L-independent, I compared the relative

frequencies of cleavage sites in PV-infected W12 HeLa cells at 8 hpa to Mock-infected

W12 HeLa cells at 8 hpa. RNase L-dependent signal was considered as 5-fold greater

(or unique) in the PV-infected W12 at 8 hpa cells compared to the Mock-infected 8 hpa 81 sample. RNase L-dependent sites were further filtered by comparisons to the frequencies detected in M25 HeLa cells, and by comparisons with earlier time points.

It is important to note that the bioinformatic methods used here are slightly different from those reported in (Cooper et al., 2014). After publishing these data in

2014, I determined that data normalization for each cDNA library provided a more precise analysis of cleavage sites within host and viral RNAs. Data normalization minimizes the variation observed between libraries, allowing for direct comparison of cleavage in one cDNA library with that in another cDNA library.

Results

RNA from Uninfected and PV-Infected HeLa Cells

After optimizing and validating the 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and

Illumina sequencing methods using purified endoribonucleases (Chapter II), I used the

2ˊ, 3ˊ-cyclic phosphate cDNA synthesis methods to characterize RNA from uninfected and PV-infected HeLa cells (Figure 3.1). W12 HeLa cells express wildtype RNase L

whereas M25 HeLa cells express a dominant-negative mutant form of RNase L (R667A

mutation). The Barton Lab previously established that RNase L activity, which requires

2-5A from dsRNA-activated OAS, is provoked during the course of PV infection in W12

HeLa cells, whereas RNase L activity remains undetectable during the course of PV

infection in M25 HeLa cell (Han et al., 2007). The kinetics and magnitudes of PV

replication in W12 and M25 HeLa cells are very similar (Figure 3.1A). When RNAs from

the PV-infected cells were analyzed by agarose gel electrophoresis, the accumulation of

viral RNA was evident at 4 to 8 hours post adsorption (hpa) in both M25 and W12 HeLa

cells (Figure 3.1B, PV RNA is evident above 28S rRNA). rRNA fragments characteristic of RNase L activity were evident in RNAs from PV-infected W12 HeLa cells at 6 and 8 hpa, but these rRNA fragments were not detected in PV-infected M25 HeLa cells 82

A. B.

Figure 3.1. PV infection of W12 and M25 HeLa cells. RNA was isolated from mock-infected and PV-infected HeLa cells for 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing A. W12 and M25 HeLa cells were infected with PV using 10 PFUs per cell. PV titers determined by plaque assay and plotted versus time (hpa). B. RNA isolated from infected cells was fractionated by agarose gel electrophoresis and visualized using EtBr and UV light. *Indicates rRNA fragments in the 6 and 8 hpa time points.

(Figure 3.1B, asterisks indicate the location of rRNA fragments characteristic of RNase

L activity).

2ˊ, 3ˊ-cyclic phosphate cDNA libraries were prepared and sequenced using the

RNAs from mock-infected and PV-infected HeLa cells (Figure 3.2A and B). The

normalized amounts of cDNA reads corresponding to PV RNA increased in frequency

from 0-8 hpa. In both M25 and W12 HeLa cells, we capture most PV-specific reads at

the 6 hpa time points (Figure 3.2A : 28.2% of the library in M25 HeLa cells / Figure

3.2B: 5.3% of the library in W12 HeLa cells; Tables 3.1 and 3.2). Despite over-

expressing a dominant-negative form of RNase L, PV cDNA reads were more abundant

in M25 HeLa cells than W12 HeLa cells. Abundant amounts of cDNA in each sample

aligned to ribosomal RNAs (28S/18S/5.8S/5S rRNAs) and U6 snRNA (Figure 3.2; Table

3.1 and 3.2). In the PV-infected W12 HeLa cells at 2 through 8 hpa, the frequency of

cDNA reads corresponding to 18S rRNA exceeded the amount corresponding to 28S

rRNA (Figure 3.2B). Variable amounts of cDNA reads corresponding to cellular RNAs 83

Figure 3.2. Frequency of ribonuclease cleave in host and viral RNAs from M25 and W12 HeLa cells. The RNAs shown in Figure 3.1B were used for 2ˊ, 3ˊ- cyclic phosphate cDNA synthesis and Illumina sequencing (Methods described in Chapter II; Figure 2.4). Data from M25 (A) and W12 (B) HeLa cells are plotted from Tables 3.1 and 3.2 and represent the relative frequency of RNAs with terminal 2ˊ, 3ˊ-cyclic phosphates detected in cDNA libraries.

Table 3.1. Frequency of 2’, 3’-cyclic phosphates in host and viral RNAs from M25 HeLa cells

Table 3.2. Frequency of 2’, 3’-cyclic phosphates in host and viral RNAs from W12 HeLa cells

84

(mRNAs, tRNAs, ribosomal-like RNAs and other noncoding RNAs) were detected in each sample.

Endoribonuclease Cleavage Sites in PV RNAs Isolated from HeLa cells

2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and sequencing of RNAs from PV-

infected M25 and W12 HeLa cells revealed the frequency and location of cleavage sites

across PV RNA (Figure 3.3). In both M25 and W12 HeLa cells, reads in PV RNA

become most apparent at 4 hpa, coincident in time with the initial detection of PV RNA in

agarose gels (Figure 3.1B). Cleavage sites in M25 HeLa cells were more numerous and

more widely distributed across the PV RNA compared to the cleavage sites in PV RNA

from W12 HeLa cells. For example, in the 4 hpa time point, there were 1,021 cleavage

sites in PV RNA from M25 HeLa cells compared to only 97 from W12 HeLa cells (Figure

3.3). Data from in vitro experiments showed that purified RNase L targeted discrete

regions of PV RNA (Chapter II: Figure 2.6 and 2.10). I compared the RNase L cleavage

sites in PV RNA identified in vitro (Figure 2.6) with those found in viral RNA from

infected HeLa cells. The regions of PV RNA targeted by endoribonucleases were similar

in vitro and in vivo. UA1715 was consistently detected among the top 99th percentile of

RNase L cleavage sites in PV RNA in vitro, and was in the top five frequently cleaved

sites from the 2.5, 5, 10, and 20 minute time points (Figures 2.6 and 2.10). Consistent

with these in vitro data, UA1715 was the most frequent RNase L-dependent cleavage site

detected in PV RNA at 6 and 8 hpa in W12 HeLa cells. While UA1715 was also cleaved in

M25 HeLa cells, the frequency of cleavage at this site were much lower in M25 HeLa cells than in W12 HeLa cells, and by the 8 hpa time point, UA1715 was cleaved 50 times more frequently in W12 HeLa cells than in M25 HeLa cells (UA1715 , Figure 3.3C). Based

on these in vitro and in vivo data, UA1715 is clearly one of the most frequent targets of

RNase L in poliovirus RNA. 85

Figure 3.3. Ribonuclease cleavage sites in PV RNA. RNAs from PV-infected HeLa cells (Figure 3.1B) were used for 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing. A. Frequency, location, and specificity of cleavage sites in PV RNA from M25 HeLa cells. X-axis: Nucleotide position in PV RNA. Y-axis: % of total cDNA library detected at each cleavage site. B. Frequency, location, and specificity of 1715 cleavage sites in PV RNA from W12 HeLa cells. C. Frequency of cleavage at UA , UA946, UU1701, UU3102, UA11803, UA1708 in W12 (red) and M25 (gray) HeLa cells.

86

Other RNase L cleavage sites in PV RNA defined in vitro include UA1803, UU3102, and UA1738. However, the frequency of cleavage at UU3102 was similar between W12 and M25 HeLa cells (Figure 3.3C), indicating that this site in PV RNA can be cleaved by endoribonucleases other than RNase L in HeLa cells. Likewise, UA1803 and UA1738 were

cleaved more frequently in M25 cells than in W12 HeLa cells (Figure3.3C), presumably

due to ribonucleases other than RNase L. Other RNase L-dependent cleavage sites

were found in PV RNA from W12 HeLa cells, and included UA946 and UU1701 (Figure

3.3C). UA946 was identified in the top 95th percentile of RNase L cleavage sites in vitro

(Figure 2.6A) and UU1701 was identified in the top 90th percentile of RNase L cleavage

sites in vitro at the 10 and 20 minute time points; however their cleavage frequency in

virus-infected W12 HeLa cells were relatively low compared to the frequency of cleavage

at UA1715. Thus, based on the cyclic phosphate cDNA libraries from PV-infected HeLa

cells, only three sites in PV RNA appear to be targeted by RNase L in vivo : UA946,

UU1701 and UA1715. There were no other obvious RNase L-dependent cleavage sites in

PV RNA isolated from W12 HeLa cells. One key difference between in vitro and in vivo

experiments is the presence of cellular RNAs and cellular ribonucleases other than

RNase L. Exoribonucleases may degrade RNA fragments produced by RNase L and

other endoribonucleases, confounding their detection in vivo, and the high abundance of rRNAs with 2ˊ, 3ˊ-cyclic phosphates impacts the detection of less abundant RNAs.

Nonetheless, the cleavage sites in PV RNA from HeLa cells mapped to the same regions of PV RNA as the cleavage sites found using purified RNase L and RNase A

(Figure 2.6), with notable peaks of cleavage in the 5ˊ half of the open-reading frame

(compare Figure 2.6 with Figures 3.3). Remarkably, PV cDNA reads from HeLa cells corresponded to cleavage sites almost exclusively within positive-strand RNA; no cDNA reads corresponding to cleavage sites within PV negative-strand RNA were found in

HeLa cells. 87

Ribonuclease Cleavage Sites in rRNAs Isolated from HeLa Cells

The frequency and location of endoribonuclease cleavage sites in rRNAs was

largely consistent across all of the RNA samples from uninfected and PV-infected HeLa cells (Figures 3.4 through 3.7). By comparing the cleavage in 28S rRNA in both W12 and M25 HeLa cells, it was clear that 28S rRNA was cleaved in an RNase L- independent fashion at specific sites; most frequently at AG409, GU432, CA1699, UG2055,

GC2083, GA2093, GU2097, AA2396, UG2427, AU4512, and UA4727 (Table 3.3; Figure 3.4A and

B). Unlike cleavage in 18S rRNA, no obvious changes were observed in the cleavage

patterns of 28S rRNA from the W12 and M25 HeLa cells, even when rRNA

fragmentation characteristic of RNase L activity was observed specifically in W12 HeLa

cells (Figure 3.1B). Because rRNA fragmentation is most pronounced in the 8 hpa time

point from W12 HeLa cells, I took a closer look at the top 20 most frequently cleaved

sites in 28S rRNA from PV-infected W12 HeLa cells at 8 hpa (Table 3.3). I compared the

frequencies of cleavage at these sites in mock- and PV-infected W12 and M25 HeLa

cells at 8 hpa. Cleavage sites enriched ≥ 5-fold in PV-infected W12 HeLa cells at 6 and 8

hpa as compared to mock-infected W12 HeLa cells, and at least 2.5-fold greater than

PV-infected M25 HeLa cells were considered RNase L-dependent cleavage sites

(Figure 3.4B, RNase L-dependent cleavage sites highlighted in red). I examined the

frequency of RNase L-dependent cleavage in 28S rRNA during the course of the PV

infection (Figure 3.4C). The frequency of cleavage increased at three sites coincident in

time with the fragmentation of rRNA in W12 HeLa cells; UA1637, UG2486 and CA1443 (Table

3.3; bold red; Figure 3.4C). In contrast, the frequency of cleavage at AG1700 and CA1939 were inconsistent through the time course (Table 3.3; italicized red; Figure 3.4C).

Interestingly UG4032, identified as an RNase L–dependent cleavage site by primer extension of RNA from poly(I:C)-treated HeLa cells (Iordanov et al., 2000), was detected at the 6 hpa time point from W12 HeLa cells and increased in frequency relative to M25 88

Table 3.3. RNase L-dependent and -independent cleavage sites in 28S rRNA from W12 and M25 HeLa cells

W12 HeLa Cells M25 HeLa Cells

28S Mock- Mock- PV-Infected PV-Infected rRNA Infected Infected (% Library) (% Library) position (% Library) (% Library) CA1699 0.53 1.39 0.24 0.12 AU4512 0.85 1.02 1.60 1.17 AG409 1.41 0.84 2.33 1.96 GC2083 0.58 0.79 0.49 0.57 UG2055 0.32 0.55 0.83 0.32 UG2427 0.28 0.54 0.08 0.30 GC1698 0.12 0.45 0.01 0.04 UA1637 0.06 0.43 0.00 0.01 UA4729 0.23 0.37 0.44 0.10 UG2486 0.03 0.30 0.00 0.01 GU2097 0.22 0.26 0.55 0.19 CG1697 0.10 0.23 0.03 0.03 AA2396 0.16 0.23 0.14 0.09 AG1700 0.04 0.22 0.00 0.04 GA2093 0.09 0.19 0.11 0.06 CA1443 0.02 0.15 0.00 0.00 CU1636 0.05 0.14 0.00 0.01 AU1930 0.08 0.14 0.06 0.11 CA1939 0.00 0.14 0.03 0.04 GG2475 0.02 0.14 0.14 0.18

20 most frequent cleavage sites (benchmarked to W12 PV 8hpa). % Total cDNA reads at individual sites in 28S rRNA. Red: RNase L-dependent cleavage sites. > 5-fold increased cleavage in W12 PV 8hpa as compared to mock-infected W12 8hpa and mock-infected M25 8hpa, and > 2.5-fold increased over M25 PV 8hpa. Italicized:Site did not increase in an infection-dependent manner. 89

Figure 3.4. Ribonuclease cleavage sites in 28S rRNA. RNAs from mock-infected and PV-infected M25 (A) and W12 (B) HeLa cells (Figure 3.1B) were used for 2ˊ, 3ˊ- cyclic phosphate cDNA synthesis and Illumina sequencing. The frequency, location, and specificity of cleavage sites in 28S rRNA are plotted. RNase L cleavage sites are highlighted in red. C. The frequency of RNase L-dependent cleavage sites, as highlighted in Table 3.3, are plotted.

90

HeLa cells. Its cleavage frequency, however, decreased by the 8 hpa time point, and therefore was not included as an RNase L-dependent cleavage site by the metric applied here (cleavage sites enriched ≥ 5-fold in PV-infected W12 HeLa cells at 8 hpa as compared to mock-infected W12 HeLa cells, and at least 2.5-fold greater than PV-

infected M25 HeLa cells). Thus, based on these data, I concluded that 28S rRNA was

cleaved in an RNase L-dependent manner in PV-infected cells at UA1637, UG2486 and

CA1443 (Table 3.3; bold red) (Figure 3.4C).

Like 28S rRNA, 18S rRNA was cleaved reproducibly at specific sites (Table 3.4;

Figure 3.5A and B). Cleavage sites commonly identified in 18S rRNA from M25 and

W12 HeLa cells include CG130, AG225, CC502, AC1403, UA1506, and UU1721 (Table 3.4;

Figure 3.5A and B). While the cleavage patterns of 28S rRNA from W12 and M25 HeLa

cells looked similar, there were striking differences between the two cell lines in 18S

rRNA. For example, cleavage sites located in the 1400 – 1500 nt range were more

frequent in W12 HeLa cells compared to M25 HeLa cells. The cleavage sites within this

stretch appear to increase in frequency as a function of PV-infection in the W12

background, but do not correspond to fragmentation of rRNA that can be observed

through RNA fractionation on agarose gels (Figure 3.1). To assess RNase L-dependent

cleavage sites, I examined the top 20 cleavage sites in 18S rRNA from PV-infected W12

HeLa cells (Table 3.4), and used the same metric as that applied to the data for 28S

rRNA noted above. Three cleavage sites in 18S rRNA were enriched ≥ 5-fold in PV-

infected W12 HeLa cells at 8 hpa as compared to mock-infected W12 HeLa cells, and at

least 2.5-fold greater than PV-infected M25 HeLa cells: UU743, UU541, UG1536 (Table 3.4,

red and bold; Figure 3.5C). Among the other frequent cleavage sites in 18S rRNA were

UA1506 and GU1407; however, these sites exhibited increased cleavage well before rRNA

fragmentation became apparent in the virus-infected W12 HeLa cells (Figure 3.5C;

Table 3.4, red and italicized). 91

Table 3.4. RNase L-dependent and -independent cleavage sites in 18S rRNA from W12 and M25 HeLa cells

W12 HeLa Cells M25 HeLa Cells

18S Mock- Mock- PV-Infected PV-Infected rRNA Infected Infected (% Library) (% Library) position (% Library) (% Library) UA1506 0.04 2.08 0.21 0.13 UU743 0.04 1.66 0.03 0.01 AC1403 0.62 1.51 0.13 0.13 UC1433 0.51 1.19 0.01 0.05 UU1463 0.49 1.15 0.09 0.02 AU1477 0.31 0.82 0.04 0.05 AG1507 0.25 0.63 0.04 0.09 GU1462 0.25 0.61 0.05 0.03 GU1432 0.20 0.55 0.03 0.02 CC502 0.35 0.54 0.26 0.11 UU1721 0.22 0.51 0.26 0.16 CC1419 0.14 0.42 0.04 0.02 AG225 0.35 0.38 0.65 0.50 UU541 0.02 0.34 0.00 0.00 CG130 0.93 0.32 1.16 2.04 UG1536 0.01 0.23 0.00 0.00 CA1454 0.09 0.22 0.00 0.01 UC417 0.04 0.22 0.05 0.05 AU1397 0.13 0.21 0.04 0.03 GU1407 0.03 0.21 0.04 0.03

20 most frequent cleavage sites (benchmarked to W12 PV 8hpa). % Total cDNA reads at individual sites in 18S rRNA. Red: RNase L-dependent cleavage sites. > 5-fold increased cleavage in W12 PV 8hpa as compared to mock-infected W12 8hpa and mock-infected M25 8hpa, and > 2.5-fold increased over M25 PV 8hpa. Italicized:Site did not increase in an infection-dependent manner.

92

A. B.

Figure 3.5. Ribonuclease cleavage sites in 18S rRNA. RNAs from mock-infected and PV-infected M25 (A) and W12 (B) HeLa cells (Figure 3.1B) were used for 2ˊ, 3ˊ- cyclic phosphate cDNA synthesis and Illumina sequencing. The frequency, location, and specificity of cleavage sites in 18S rRNA are plotted. RNase L cleavage sites are highlighted in red. C. The frequency of RNase L-dependent cleavage sites, as highlighted in Table 3.4, are plotted.

93

5.8S rRNA, a component of the large ribosomal subunit, had few 2 ˊ, 3ˊ-cyclic phosphates relative to other rRNAs (Figures 3.2, 3.6), likely due to its small size (156 nts). A consistent pattern of RNase L-independent cleavage sites was detected within

5.8S rRNA in uninfected- and PV-infected M25 and W12 HeLa cells (Figure 3.6). The most frequent cleavage sites identified in 5.8S rRNA include CU109, CU141, and CU155 among others (Figure 3.6). While the mature 3ˊ-end of 5.8S is reported to end with

CUU156, I detected a terminal 2ˊ, 3ˊ-cyclic phosphate 1 base upstream at CU155 on a

small portion of 5.8S rRNA molecules. While the cleavage detected in 5.8S rRNA

remains relatively constant during the PV-infection in M25 HeLa cells, the relative

amounts of cyclic phosphates decreased in frequency in PV-infected W12 HeLa cells

(Figure 3.6).

5S rRNA, another component of the large ribosomal subunit, consistently had a

high frequency of RNase L-independent 2ˊ, 3ˊ-cyclic phosphate termini (Figure 3.2; 3.7).

10-20% of the cyclic phosphates found in HeLa cells were present at the 3 ˊ end of 5S

rRNA. This finding was striking and unexpected, as cyclic phosphates have never been

reported at the 3ˊ-terminus of 5S rRNA. 2ˊ, 3ˊ-cyclic phosphates were detected

120 predominantly after CUU in both M25 and W12 HeLa cells (Figure 3.7).

The location of RNase L-dependent and RNase L-independent cleavage sites were mapped onto the secondary structures of rRNAs, as well as 80S ribosomes

(Figure 3.8). On the secondary structure of the large subunit, UG2486 is located in a

predicted double-stranded region of an expansion segment, ES19L (Figure 3.8; inset

#1).CA1443 is located on H30 of ES9L (Figure 3.8; inset #2), a single-stranded region

that interacts with a single-stranded portion of ES15L (these bases are circled in back).

UA1637 (not pictured) is located in a single-stranded region between H33 and H36, and is

buried in the 80S structure. Finally, UG4032 is located on a single-stranded region of

ES30L on H78 (Figure 3.8; inset #3). 94

Mock 0 hpa Mock 0 hpa

Mock 8 hpa Mock 8 hpa

PV 0 hpa PV 0 hpa

PV 2 hpa PV 2 hpa

PV 4 hpa PV 4 hpa

PV 6 hpa PV 6 hpa

PV 8 hpa PV 8 hpa

Figure 3.6. Ribonuclease cleavage sites in 5.8S rRNA. RNAs from mock- infected and PV-infected M25 and W12 HeLa cells (Figure 3.1B) were used for 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing. The frequency, location, and specificity of cleavage sites in 5.8S rRNA are plotted.

On the secondary structure of the small subunit, UG1536 (Figure 3.8; inset #4) mapped to a single-stranded region of h41. UU541 is located in the single-stranded loop of h16 (Figure 3.8; inset #5). Finally UU743, the most frequently detected RNase L-

dependent cleavage site detected in rRNA, is located in a single-stranded region of

expansion segment E6S (Figure 3.8; inset #6). With the exception of UA1637, all RNase

L-dependent cleavage sites are located on the exterior of the ribosome, which would be accessible to RNase L in the cytosol of cells. The 2ˊ, 3ˊ-cyclic phosphates near the 3ˊ- ends of 5.8S and 5S rRNA (Figure 3.6; 3.7; 3.8, gold) are surface exposed on the 60S

95

Mock 0 hpa Mock 0 hpa

Mock 8 hpa Mock 8 hpa

PV 0 hpa PV 0 hpa

PV 2 hpa PV 2 hpa

PV 4 hpa PV 4 hpa

PV 6 hpa PV 6 hpa

PV 8 hpa PV 8 hpa

Figure 3.7. Ribonuclease cleavage sites in 5S rRNA. RNAs from mock-infected and PV -infected M25 and W12 HeLa cells (Figure 3.1B) were used for 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing. The frequency, location, and specificity of cleavage sites in 5S rRNA are plotted.

subunit where they could potentially be targeted by cellular nucleases, resulting in the terminal 2ˊ, 3ˊ-cyclic phosphates (Figures 3.6; 3.7; 3.8).

2ˊ, 3ˊ-Cyclic Phosphates at the End of U6 snRNA

U6 snRNA accounted for a substantial amount of cDNA reads from the HeLa

cells, ranging between10% to 20% of total cDNA reads in M25 HeLa cells (Figure 3.2A)

and uninfected W12 HeLa cells (Figure 3.2B). cDNA reads corresponding to U6 snRNA

accounted for less than 10% of total cDNA reads in PV-infected W12 HeLa cells (Figure 96

Figure 3.8. Location of RNase L-dependent cleavage sites in 80S ribosomes. RNase L-dependent cleavage sites were mapped onto secondary structures (Anger et al., 2013) and the structure of the human 80S ribosome (Anger et al., 2013). Cleavage sites are numbered on the 80S ribosome structure and the respective secondary structure is provided as an inset. PDB files 3J3A, 3J3B, 3J3D, and 3J3F (Anger et al., 2013) were visualized using PyMOL (DeLano Scientific, San Carlos, California, USA; http://www.pymol.org).

97

3.2B). The U6 snRNA cDNA reads in our libraries align almost exclusively to the 3ˊ end of mature U6 snRNA (Figure 3.9). U6 snRNA, an RNA polymerase III (Pol III) transcript, is post-transcriptionally modified by a nuclease encoded by gene C16orf57 (yeast) / hUSB1 (humans), which leaves a 2ˊ, 3ˊ-cyclic phosphate at the end of mature U6 snRNA (Mroczek et al., 2012; Shchepachev and Azzalin, 2013; Shchepachev et al.,

2012). Thus, cyclic phosphates are expected to be present at the 3ˊ end of U6 snRNA.

The abundant amounts of cDNA in our libraries corresponding to this particular modification of U6 snRNA further validate the reliability of the 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods.

Mock 0 hpa Mock 0 hpa

Mock 8 hpa Mock 8 hpa

PV 0 hpa PV 0 hpa

PV 2 hpa PV 2 hpa

PV 4 hpa PV 4 hpa

PV 6 hpa PV 6 hpa

PV 8 hpa PV 8 hpa

Figure 3.9. Ribonuclease cleavage sites in U6 snRNA. RNAs from mock-infected and PV-infected M25 and W12 HeLa cells (Figure 3.1B) were used for 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing. The frequency, location, and specificity of cleavage sites in U6 snRNA are plotted.

98

Endoribonuclease Cleavage Sites in mRNAs and Other Host RNAs

I detected endoribonuclease cleavage sites in human RNAs after alignment to

Homo sapiens Build 36 (hg18). Sites frequently cleaved tended to map to mitochondrial

16S rRNA and other non-coding RNAs, although at a much lower frequency than 28S and 18S,rRNA (data not shown). In addition to other non-coding RNAs, I detected signal in cellular mRNAs. The most frequently targeted mRNA encoded fascin1 (FSCN1)

(Figure 3.10). The gene for fascin1 is located on chromosome7 from bases 5,598,980 –

5,612,812. Fascin 1 is a scaffolding protein involved in bundling of actin filaments to mediate cell motility (Kureishy et al., 2002).

Figure 3.10. Ribonuclease cleavage sites in FSCN1 mRNA. The frequency, location, and specificity of cleavage sites in the last coding exon extending into the 3ˊUTR of fascin1 mRNA from M25 HeLa cells and W12 HeLa cells. Cleavage sites are labeled with dinucleotide and the position (as hundreds) on the chr7. For example, position 5,611,500 would be labeled as 500.

99

Our data indicate that the last coding exon extending into the 3ˊ-UTR of the transcript was frequently targeted by ribonucleases in both M25 and W12 HeLa cells (Figure

3.10); however there was inconsistency in its detection in M25 HeLa cells. In PV-infected

M25 HeLa cells from 0, 2, and 8 hpa, low amounts of signal in FSCN1 mRNA was detected. This pattern of inconsistency was not observed with other RNAs analyzed from cyclic phosphate cDNA libraries. Although the pattern of cleavage in FSCN1 mRNA was variable throughout the PV-infection time course, no RNase L-dependent cleavage sites

were detected in FSCN1 mRNA.

Discussion

For this chapter, I generated 2ˊ, 3ˊ-cyclic phosphate cDNA libraries from RNA

isolated from PV-infected HeLa cells for the purpose of identifying host and viral RNAs

cleaved by RNase L. Although many RNase L-independent cleavage sites were

detected in host and viral RNAs (5.8S and 5S rRNA, U6 snRNA, and FSCN1 mRNA),

RNase L-dependent cleavage sites were detected in 18S rRNA, 28S rRNA, and PV

RNA. This discussion addresses the potential significance of RNase L-dependent and

independent cleavage of host and viral RNAs.

PV RNA

Cleavage sites in PV RNA were detected in both M25 and W12 HeLa cells.

RNase L-specific cleavage sites, UA1715, UA946, and UU1701, were detected both in vitro

using purified RNase L (Figure 2.6) and in vivo (Figure 3.3). UA1715, the most robustly

detected RNase L-dependent site in PV RNA from W12 HeLa cells, is located in the VP3 capsid gene for PV. The detection of UA1715 in both the 6 and 8 hpa time points was

coincident with RNase L-induced rRNA cleavage (Figure 3.1B), suggesting that RNase

L cleaves PV RNA at this site in vivo. 100

An intriguing observation was the absence of cDNA reads corresponding to PV

minus-strand RNA in cyclic phosphate libraries. PV, a positive-strand RNA virus,

replicates within membrane-anchored RNA replication complexes in the cytoplasm of infected cells. Viral dsRNA replication intermediates are sequestered within these replication complexes (Richards et al., 2014). PV minus-strand RNA may be protected from cleavage by RNase L, due in part to its compartmentalization in replication complexes. In addition, an RNA structure within the open reading frame of PV RNA competitively inhibits RNase L (the so-called RNase L ciRNA) (Han et al., 2007;

Townsend et al., 2008a, 2008b). Within replication complexes, the ciRNA would be in

molar excess to RNase L, and would prevent RNase L-mediated cleavage of PV plus-

and minus-strand RNAs. rRNA depletion before cDNA library preparation may allow for

enhanced detection of potential cleavage sites in viral minus-strand RNA.

Although WT RNase L is activated during PV infection in W12 HeLa cells, (as

indicated by rRNA fragmentation (Figure 3.1A), more PV RNA fragments were detected

in cDNA libraries from M25 HeLa cells, where PV infection does not provoke rRNA

fragmentation. Despite similar replication kinetics (Figure 3.1B), 28% of the cDNA

library from M25 HeLa cells at 6 hpa corresponded to PV-specific cDNA reads, while

only 6% of the cDNA library from W12 HeLa cells corresponded to PV-specific reads

(Tables 3.1 and 3.2). It is unclear why more cleavage sites PV RNA were detected in

M25 HeLa cells; however in studies of Sindbis virus replication in RNase L-deficient

MEFs, it was suggested that RNase L might have roles in stabilizing replication

complexes (Sawicki et al., 2003). Perhaps the nuclease-defective RNase L used in

these experiments cannot stabilize PV replication complexes in the M25 HeLa cells,

leading to their premature turnover that exposes PV RNA degradation by cellular

ribonucleases. This would potentially explain why more PV RNA was detected from M25

HeLa cells, but if this were the case, more PV cDNA reads corresponding to minus 101

strands should also be detected in cyclic phosphate cDNA libraries. As previously

mentioned, cDNAs corresponding to PV minus strands were largely absent from

libraries.

Ribosomal RNAs

2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing revealed a

constellation of RNase L-dependent and RNase L-independent cleavage sites in rRNAs

(Tables 3.3 and 3.4;Figures 3.4 -3.8). Because rRNA is the most abundant RNA in the cell, many studies actively exclude rRNAs from deep sequencing experiments to enrich for other RNA populations. Because the characteristic signature of RNase L activation is rRNA fragmentation, I intentionally retained rRNAs in this study to identify RNase L- dependent cleavage sites. Several RNase L-dependent cleavage sites in 28S and 18S rRNA were detected in single-stranded regions of rRNA within surface exposed areas of the ribosome (Figure 3.8). Two of the most frequently RNase L-dependent cleavage sites were UU743 and UU541 in 18S rRNA (Figures 3.5 and 3.8). Cleavage of the 1,869

base-long 18S rRNA at these sites would produce RNA fragments consistent with the

characteristic rRNA fragments observed by agarose gel electrophoresis in this study

(Figure 3.1B), and in previous studies by the Barton Lab and others (Han et al., 2007;

Silverman et al., 1983; Wreschner et al., 1981a). As expected for RNase L, UU541 and

UU743 were associated with single-stranded regions of 18S rRNA on the surface of 80S

ribosomes (Figure 3.8). UU541 is located in the helix 16 region of the small subunit

(Figure 3.8) (Anger et al., 2013). Helices 16, 18, and 34 are involved in forming the

mRNA entry channel. Binding of initiation factors induces conformational changes in the

40S subunit to enable mRNA scanning for the initiator codon. The conformational

changes of the 40S subunit are likely due to interactions between helix 16 and rpS3,

which encourage an open conformation for scanning (Ben-Shem et al., 2010; Passmore 102

et al., 2007). In addition, UU541 is located in a region that directly interacts with scanning factor DHX29 (Hashem et al., 2013), a helicase important for unwinding structured mRNAs for mRNA scanning during initiation. Cleavage of UU541 could potentially inhibit

protein synthesis by inhibition of 43S-mRNA scanning. The more frequently detected

UA743 is located in between helix A and B of expansion segment 6S of the small subunit

(Figure 3.8). The terminal loop of helix E from expansion segment 6S interacts with helix

B of expansion segment 3S (Alkemar and Nygård, 2003, 2004; Anger et al., 2013). The

region of ES6S and ES3S influence binding of translation initiation factors eIF3 and

eIF4G (Siridechadilok et al., 2005; Srivastava et al., 1992; Yu et al., 2011). Cleavage at

UU743 may disrupt the interaction between ES6S and ES3S, preventing the binding of

eIF3 and eIF4G, thus inhibiting protein synthesis. In addition, ribosomal protein S7

(rpS7) binds in proximity to UU743 (Anger et al., 2013) (Figure 3.11). During cellular

stress, rpS7 binds to MDM2 to prevent MDM2-mediated degradation of p53, thereby

Figure 3.11. Proximity of ribosomal protein rpS7 to the RNase L-dependent cleavage site, UU743. UU743 (red) is located at the base of helix A in expansion segment ES6S in 18S rRNA (gray). rpS7 (green) binds in proximity to the base of helix A as well.

103

provoking p53-induced cell-cycle inhibition and apoptosis (Chen et al., 2007; Zhu et al.,

2009). Cleavage at UU743 might prevent binding of rpS7 or may liberate rpS7 from the

ribosome, inhibiting translation and MDM2-mediated degradation of p53.

Although RNase L-induced cleavage of 18S rRNA cleavage has not been

characterized, RNase L-induced cleavage in 28S rRNA has been mapped by primer

extension methods (Iordanov et al., 2000). CU3999 and UG4000 in 28S rRNA were

identified as RNase L-induced cleavage sites from poly(I:C)-treated HeLa cells. These sites were detected in 2ˊ, 3ˊ-cyclic phosphate cDNA libraries, and correspond to CU4031 and UG4032 on the 28S rRNA sequence used for these analyses; however they are detected most frequently in the 6 hpa timepoint from PV-infected W12 HeLa cells

(Figure 3.4C; panel labeled UG4032). These same sites were identified as IRE1β

cleavage sites in HeLa cells (Iwawaki et al., 2001). IRE1β is a homolog of IRE1α, and

both enzymes possess endoribonuclease activity and are localized to the endoplasmic

reticulum where they sense unfolded proteins. While IRE1α is ubiquitously expressed in

different tissue types, IRE1β has limited expression in epithelial cells where it regulates

mucin production (Martino et al., 2012). Whether cleavage at CU4031 and UG4032 is can be attributed to IRE1β or RNase L requires further testing.

A number of RNase L-independent cleavage sites were identified consistently in rRNAs (Figures 3.4 – 3.7; Tables 3.3 and 3.4). These sites could have arisen during the

RNA linker ligation to RNAs with 2ˊ, 3ˊ-cyclic phosphates through stray nucleases,

divalent metal-ion-cleavage, or geometry-induced backbone cleavage (Forconi and

Herschlag, 2009; Soukup and Breaker, 1999). These sites could also arise as

degradation intermediates of rRNA turnover in cells. RNase T2, for example, is important

for rRNA turnover and would leave 2ˊ, 3ˊ-cyclic phosphates on the RNA fragments it

produces (Andersen and Collins, 2012; Haud et al., 2011; Hillwig et al., 2011). AG409 in

28S rRNA, one of the most prominent RNase L-independent cleavage sites in both 104

uninfected and PV-infected HeLa cells (Figure 3.4; Table 3.3), is analogous to an

oxidative stress-induced rRNA cleavage site identified in yeast (Mroczek and Kufel,

2008).

The 2ˊ, 3ˊ-cyclic phosphates detected near the ends of 5S rRNAs were

unexpected. Cyclic phosphates have been described at the 3ˊ end of U6, 7SK, and MRP

RNAs, but not at the 3ˊ end of 5S rRNA (Gu et al., 1997). Nonetheless, 5S rRNA was

among the most common sequences detected in cyclic phosphate cDNA libraries

(Figures 3.2 and 3.7; Tables 3.1 and 3.2). Nearly all of the 5S rRNA signal was affiliated

with the 3ˊ end of mature 5S rRNA (Figure 3.7). The 3ˊ ends of 5.8S and 5S rRNAs are

on the surface of 60S ribosomal subunits (Figure 3.8; ends of each RNA highlighted

with gold). The surface exposed 3ˊ-termini might be susceptible to cellular ribonucleases

that could modify these termini. Rex1p, a 3ˊ to 5ˊ exonuclease in the RNase D family, is

involved with modifying the 3ˊ ends of both 5S and 5.8S rRNA in yeast; however, RNase

D family members are metal-ion-dependent, thus it is unlikely that a human homolog of

Rex1p is responsible for the 2ˊ, 3ˊ-cyclic phosphate found at the end of these rRNAs

(van Hoof et al., 2000; Piper et al., 1983). The terminal cyclic phosphate detected after

GCU155 in 5.8S rRNA (Figure 3.6) does not represent a large proportion of 5.8S rRNA if

compared to the frequencies of 2ˊ, 3ˊ-cyclic phosphates observed in other rRNAs (18S

and 28S rRNA), since these are processed from the same precursor transcript and

should be present at a 1:1:1 molar ratio. Studies of the 3ˊ-terminus of 5S rRNA from

human KB cells indicate that two 3ˊ-termini exist, CUU120 and CUUU121, both with

terminal hydroxyl groups (Forget and Weissman, 1968, 1969). Perhaps the terminal cyclic phosphates detected in 5S rRNA represent a degradation intermediate of

119 CUUU121 to CUU120. In Euglena, a protist, 5S rRNA terminates with CUUp (Kumazaki

et al., 1982). Perhaps a population of 5S rRNA exists in human cells where a terminal cyclic phosphate has a functional role. While the identities of the ribonucleases 105 responsible for RNase L-independent cleavage sites in rRNA remain to be determined, the cyclic phosphates found near the ends of 5S rRNA maybe consistent with the enzymatic activity of the enzyme associated with U6 snRNA maturation, hUsb1. Both U6 snRNA and 5S rRNA are pol III transcripts, and there may be overlapping pathways leading to their maturation.

FSCN1 mRNA

Fascins are evolutionarily conserved proteins involved in cellular motility and adhesion (Kureishy et al., 2002). Fascin1 bundles actin filaments to form sensory organelles called filopodia (Li et al., 2010b). High expression of fascin1 is associated with poor prognosis and metastasis of many types of cancers (Fu et al., 2009b; Park et al., 2014; Vignjevic et al., 2007; Wu et al., 2010). HeLa cells are derived from an adenocarcinoma of cervical epithelial cells (ATCC). While the expression of fascin1 in the HeLa cells used in this study has not been assessed, purification of fascin1 was initially achieved by mixing HeLa cell homogenates with F-actin (Yamashiro-Matsumura and Matsumura, 1985), thus, HeLa cells express fascin1. The mRNA encoding fascin1 was the most cleaved mRNA detected in 2ˊ, 3ˊ-cyclic phosphate cDNA libraries from

HeLa cells. Cleavage sites within FSCN1 mRNA abruptly begin near the middle of the last coding exon (around AU5611523 [chr7 position]), trailing into the 3ˊUTR (Figure 3.10).

Perhaps a mechanism of regulating FSCN1 mRNA involves 3ˊ→5ˊ exoribonucleases.

The abrupt stop near the middle of the last coding exon implicates potential structure that causes the exoribonuclease to stall. Regardless, understanding how fascin1 is regulated at the RNA level could provide important insights into its role in cancers and the development of therapeutics.

106

Summary

2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods

efficiently detect RNase L and other metal-ion-independent endoribonuclease cleavage

sites in host and viral RNA from cells. In this chapter, I have described RNase L-

dependent and RNase L-independent cleavage sites in host and viral RNAs from PV-

infected HeLa cells. Two robust RNase L-dependent cleavage sites in 18S rRNA, UU541 and UU743, were detected in cyclic phosphate cDNA libraries coincident with RNase L- induced rRNA fragmentation observed on agarose gels. In the next chapter, 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and sequencing methods are used to determine whether

RNase L directly cleaves viral RNA to restrict IAV. The constellation of RNase L- dependent and –independent cleavage sites detected in host RNAs described in this chapter can be compared directly to those detected from IAV-infected A549 cells.

107

CHAPTER IV

INFLUENZA A VIRUS INFECTION: RIBONUCLEASE CLEAVAGE SITES

IN HOST AND VIRAL RNAs3

Introduction

In Chapter III, I described RNase L-dependent and independent cleavage sites

in host and viral RNAs from PV-infected HeLa cells using the 2ˊ, 3ˊ-cyclic phosphate

cDNA synthesis and deep sequencing methods described in Chapter II. These same

methods are used in this chapter to identify RNase L-dependent cleavage in host and

viral RNAs from influenza A virus (IAV)-infected A549 cells to determine how RNase L

contributes to attenuation of NS1-deficient IAV virus infections.

IAV is a significant human pathogen that causes substantial morbidity and

mortality despite the availability of vaccines and antiviral drugs. In the 2013-2014

season, IAV was widespread in all 50 US States, accounting for deaths in children and

adults (Arriola et al., 2014). Due to these notable disease burdens, IAV continues to be

studied intensely (Medina and García-Sastre, 2011). Recent discoveries of new IAV

mRNAs and proteins highlight progress in our understanding of viral replication and

pathogenesis (Jagger et al., 2012; Wise et al., 2012). Despite these advances, it is not completely understood how new pandemic IAV strains arise and develop sustained transmission in human populations. Reassortment of segmented IAV RNA genomes is clearly involved in the generation of new IAV strains, but the precise conditions and

constraints associated with this process are not completely understood (Ma et al., 2014).

IAV belongs to the Orthomyxoviridae family of viruses, and has eight negative-

sense genomic RNA segments (PB2, PB1, PA, HA, NA, NP, M and NS RNA segments),

3 Portions of this chapter were used with permissions from Cooper, D.A.*, Banerjee, S.*, Chakrabarti, A., Garcia-Sastre, A., Hesselberth, J.R., Silverman, R.H., and Barton, D.J. (2015). RNase L targets distinct sites in influenza A virus RNAs. J. Virol. 89, 2764-2776. 108 each encoding one or more viral protein that contribute to virus replication and pathogenesis (Horimoto and Kawaoka, 2005). The IAV genomic RNAs, which are present as ribonucleoprotein (RNP) complexes containing nucleoprotein (NP) and viral polymerase (PB2, PB1 and PA proteins) (Arranz et al., 2012; Portela and Digard, 2002), function as templates for viral transcription and RNA replication in the nucleus of infected cells. During RNA replication, both genomic and complementary anti-genomic IAV RNAs are maintained in RNPs (Arranz et al., 2012; Portela and Digard, 2002), preventing the formation of large double-stranded RNA (dsRNA) intermediates that might activate cellular sensors of dsRNA (Weber et al., 2006). Nonetheless, IAV encodes NS1 protein, a multifunctional protein primarily involved with antagonizing the innate antiviral response through multiple interactions with innate antiviral PRRs (Hale et al., 2008;

Krug, 2015). IAV encoding WT NS1 protein is resistant to the antiviral effects of IFN, whereas substantial deletions of the NS1 gene renders the virus incapable of growing in

IFN-competent cells (Egorov et al., 1998; García-Sastre et al., 1998; Min and Krug,

2006). Studies have shown that the NS1 protein impairs RIG-I signaling by preventing

TRIM25-dependent ubiquitin-ligation of the RIG-I N-terminal CARD to inhibit RIG-I- dependent induction of type I IFN (Gack et al., 2009; Guo et al., 2007; Kato et al., 2005;

Mibayashi et al., 2007). Direct interactions between PKR and NS1 prevent conformation changes required for activation of PKR in response to dsRNA or the cellular protein

PACT, preventing PKR-dependent translation inhibition (Li et al., 2006; Min et al., 2007).

Mutation or deletion to the dsRNA binding domain of NS1 renders IAV sensitive to the antiviral effects of IFN, predominantly through the IFN-regulated OAS/RNase L pathway

(Egorov et al., 1998; García-Sastre et al., 1998; Min and Krug, 2006). RNase L is especially important in the restriction of IAV when NS1 protein is disabled (Min and Krug,

2006). An experimental antiviral drug targeting NS1 exploits these pathways, inhibiting virus replication in an RNase L-dependent manner (Engel, 2013; Walkiewicz et al., 109

2011). RNase L also contributes to IAV-induced immunopathology in the lung (Zhou et

al., 2013) .

Despite the impact of RNase L during IAV infections, the exact mechanisms by

which RNase L restricts IAV replication are not yet understood. In this study, I sought to

determine whether RNase L cleaves IAV RNAs within infected cells, or whether RNase

L-mediated cleavage of rRNA is sufficient to inhibit viral gene expression and replication.

Furthermore, it is unknown whether IAV genomes, which are maintained within helical

nucleoprotein complexes (RNPs) (Arranz et al., 2012), are susceptible to cleavage by

RNase L.

To address these issues, I used 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing methods described in Chapter II and III (Cooper et al., 2014) to detect

RNase L-dependent cleavage sites in RNAs from IAV∆NS1-infected A549 cells. Analysis of 2ˊ, 3ˊ-cyclic phosphate cDNA libraries indicated endoribonuclease cleavage sites in each segment of IAV RNA from both genomic and anti-genomic strands; however, cleavage was most abundant within genomic segments encoding the polymerase subunits and the anti-genomic/mRNA of the segment encoding the M proteins. I observed frequent cleavage within and around some regions of increased synonymous site conservation within IAV RNAs. These regions pertain to RNA sequences associated with overlapping alternate reading frames (Chen et al., 2001; Jagger et al., 2012; Wise et al., 2012), alternative splicing (Wise et al., 2012) and virion RNA packaging (Gog et al., 2007; Wise et al., 2012). Robust RNase L-dependent cleavage sites were also detected in 18S rRNA at UU541 and UU743, consistent with RNase L activity detected in

PV-infected HeLa cells (Chapter III). I will discuss the significance of these findings as

they relate to the evolution of influenza viruses and the antiviral activities of RNase L.

110

Materials and Methods

Cell and Virus Culture

Canine kidney epithelium (MDCK) and human lung epithelium (A549) cell lines

were purchased from ATCC. MDCK-NS1-GFP cells were described previously (Gack et

al., 2009). A549 cells were maintained in RPMI-1640 medium supplemented with 10%

fetal bovine serum (FBS) (Invitrogen, Carlsbad, CA) and antibiotics. MDCK cells were

maintained in DMEM medium supplemented with 10% FBS. MDCK-NS1-GFP cells were

grown in DMEM supplemented with 10% FBS, 1% penicillin–streptomycin and 0.5 mg/ml

hygromycin (Invitrogen). The mouse adapted H1N1 strain A/PR/8/34 of influenza A virus

(IAV) and its mutant deleted for the NS1 gene, IAVΔNS1 (García-Sastre et al., 1998) was

used for these experiments. The wild type IAV was prepared as a stock grown in 11-day

embryonated chicken eggs as previously described (García-Sastre et al., 1998).

IAVΔNS1 was grown in MDCK-NS1 cells using DMEM supplemented with 0.3 % bovine

albumin (Sigma-Aldrich), 1% penicillin streptomycin, and 1 µg/ml of L-(tosylamido-2-

phenyl) ethylchloromethyl ketone (TPCK)-treated trypsin (Sigma-Aldrich) (Gack et al.,

2009). Cell and virus culture were performed by Shuvojit Banerjee from Robert

Silverman’s Lab at Cleveland Clinic.

Stable Depletion of RNase L in A549 Cells

Lentiviruses expressing either control non-mammalian shRNA or shRNA targeting

the 3′-UTR of RNase L (Sigma-Aldrich: products SHC002 and TRCN0000226437, respectively) along with lentivirus packaging plasmids were transfected into HEK293T cells using Lipofectamine 2000 (Life Technologies). Virus-containing medium was collected 48 hours after transfection. A549 cells were infected with the lentiviruses in the presence of 8 μg/ml polybrene (Life Technologies). A549 cells stably expressing control shRNA or RNase L shRNAs were selected with puromycin (2.5 µg/ml). Twenty clones 111 expressing RNase L shRNA were screened as follows. RNase L activity in the intact cells was determined by transfection of poly(rI):poly(rC) (poly I:C) (Sigma-Aldrich) at 2 µg/ml for

5 hours followed by isolation of total RNA and analysis of rRNA for RNase L-specific cleavage products with RNA chips (with an Agilent 2100 Bioanalyzer) (Xiang et al., 2003).

RNase L expression was detected by immunoblotting with a monoclonal antibody (Dong and Silverman, 1995). Three clones in which RNase L expression was depleted with shRNA were combined into a single culture for IAV infections. Stable depletion of RNase L was performed by Shuvojit Banerjee from Robert Silverman’s Lab at Cleveland Clinic.

IAV Infections

For RNA analysis, A549 cells expressing control shRNA or RNase L shRNA (2 X

106 per 10 cm dish) were washed twice in phosphate buffered saline (PBS) and placed in serum-free RPMI 1640 medium. Cells were infected with IAV or IAVΔNS1 at an MOI of 10 and incubated for 1 hour with gentle agitation every 10 minutes. The media was removed, cells were washed twice with PBS, media containing 10% FBS was added and cells were incubated for 24 hours. The cells were washed with PBS (5 times), trypsinized and re-suspended in 1 ml PBS per dish. Cell suspensions from 10 identical biological replicates (control shRNA or RNase L shRNA) were mixed together. Total RNA was isolated with Trizol® (Life Technologies). For viral plaque assays, A549 cells (2.5 X 106 per 10 cm dish) expressing control shRNA or RNase L shRNA were infected with IAV or

IAV∆NS1 at MOI of 1 for 24 hours. IAV infections were performed by Shuvojit Banerjee from Robert Silverman’s Lab at Cleveland Clinic.

112

Plaque Assays

IAV and IAV∆NS1 titers were determined on MDCK cells or MDCK-NS1-GFP cells,

respectively, in six-well plates (106 cells/well). Cells were washed twice with PBS and

incubated with virus (diluted in DMEM) for 1 hour at 37°C with frequent shaking (Dong and

Silverman, 1995). After incubation, the virus inoculum was removed and cells were

overlaid with DMEM containing a 0.6% oxoid agar and 1 μg/mL TPCK-treated trypsin

(both from Sigma-Aldrich). After 72 hours the plaques were visualized by staining with

crystal violet. Plaque assays to determine IAV viral titers were performed by Shuvojit

Banerjee from Robert Silverman’s Lab at Cleveland Clinic.

rRNA Cleavage Assays

rRNA cleavage was monitored on RNA chips using an Agilent 2100 Bioanalyzer

as described previously (Xiang et al., 2003). rRNA cleavage assays were performed by

Shuvojit Banerjee from Robert Silverman’s Lab at Cleveland Clinic.

Immunoblotting

Cell lysates (30 μg protein) were fractionated by 10% polyacrylamide gel

electrophoresis (SDS-PAGE) and proteins were transferred to polyvinylidene difluoride

membranes (0.45 μm) (BioRad) and probed with RNase L (Dong and Silverman, 1995)

or β-actin antibodies (Sigma-Aldrich). Immunoblotting was performed by Shuvojit

Banerjee from Robert Silverman’s Lab at Cleveland Clinic.

2ˊ, 3ˊ-Cyclic Phosphate cDNA Synthesis and Illumina Sequencing

2ˊ, 3ˊ-cyclic phosphate cDNA libraries and Illumina sequencing were performed as described in Chapters II and III (Cooper et al., 2014). For sequencing on the Illumina

MiSeq platform, 2.5 nM of each cDNA library (uninfected + control shRNA / uninfected + 113

RNase L shRNA / IAV∆NS1-infected + control shRNA/ IAV∆NS1-infected + RNase L

shRNA) was pooled into one run.

Bioinformatic Analysis

Deep sequencing data were analyzed in the same manner as data from PV-

infected W12 and M25 HeLa cells (Chapter III), except - -local alignment mode was

used for the Influenza A virus segments:(A/Puerto Rico/8/34/Mount Sinai(H1N1))

segment 1 (AF389115.1), segment 2 (AF389116.1), segment 3 (AF389117.1), segment

4 (AF389118.1), segment 5 (AF389119.1), segment 6 (AF389120.1), segment 7

(AF389121.1), and segment 8 (AF389122.1) with nucleotides 57 – 528 deleted.

Because Bowtie2 - - local alignment “soft-trims” non-templated bases at the ends of

reads, potential host-derived mRNA sequences attached to 5ˊ-IAV mRNAs from viral

cap-snatching strategies can be detected. Raw sequencing data for these experiments

were deposited in NCBI’s Gene Expression Omnibus (Barrett et al., 2013); GEO Series

accession number GSE60581 (Cooper et al., 2015).

Synonymous Site Conservation

Synonymous site conservation, using synplot2 (Firth, 2014) was performed using

large numbers of IAV RNA sequences (from 960 to 1600, depending on the segment).

These IAV RNA sequences were selected via BLASTClust to sample the diversity of all available sequences. These analyses were performed by the Andrew Firth Lab at

University of Cambridge.

114

Results

RNase L Restricts IAV Replication and Cleaves rRNA when NS1 Protein is Deleted

A cell culture system using human A549 lung carcinoma cells was used to analyze RNase L-mediated rRNA cleavage products during IAV infections (Figure 4.1). shRNAs were used to reduce the expression of endogenous RNase L, as compared to control shRNA-treated cells (Figure 4.1A). RNase L, when activated in virus-infected cells, produces characteristic rRNA fragments (Silverman et al., 1983; Wreschner et al.,

1981a). RNase L-mediated rRNA fragments were evident in IAV∆NS1-infected A549 cells and in cells transfected with poly I:C, a synthetic dsRNA (Figure 4.1B, control shRNA-treated cells, mobility of rRNA fragments indicated by arrowheads). rRNA fragments were absent in both uninfected cells and in WT IAV-infected cells, where NS1 was expressed normally (Figure 4.1B). Furthermore, knockdown of RNase L expression dramatically diminished the cleavage of rRNA in IAV∆NS1-infected cells and in poly I:C transfected cells, as compared to control shRNA-treated cells (Figure 4.1B). The magnitudes of IAV replication were consistent with these observations. Wild type IAV replicated equally well in both control and RNase L shRNA-treated cells, whereas

IAV∆NS1 replication was restricted in the control shRNA-treated cells (Figure 4.1C).

Knockdown of RNase L in A549 cells increased IAV∆NS1 replication 17-fold (Figure

4.1C). These results, which show that RNase L is activated during infection when IAV

NS1 protein is deleted, are consistent with those reported by others (Min and Krug,

2006; Walkiewicz et al., 2011).

Ribonuclease Cleavage Sites in Host and Viral RNAs

Cyclic phosphate cDNA synthesis and Illumina sequencing reveal the frequency and location of cleavage sites within host and viral RNAs (Chapter II; Figure 2.4)

115

Figure 4.1. NS1 prevents RNase L activation during IAV infections of A549 cells. A. RNase L levels determined by immunoblotting in control and RNase L shRNA-treated A549 cells. Results of biological triplicates are shown. B. RNase L activity by rRNA cleavage was determined on RNA chips (Agilent) with RNA isolated from control or RNase L shRNA-treated cells infected with IAV or IAVΔNS1 at an MOI of 10 for 24 hours or transfected with poly I:C (2 μg/ml) for 5 hours. Arrows (to the right) indicate positions of RNase L-mediated degradation products of rRNA. The RNAs used for 2ˊ,3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing are indicated by checkmarks. C. IAV yields following infection (MOI of 1 / 24 hpi) (Manicassamy et al., 2010). Results are shown as the mean ± S.D. from three biological replicates. Two-tailed t tests were done. ns, non significant; *, p<0.05.

(Cooper et al., 2014). These methods exploit a specific chemical signature (2ˊ, 3ˊ-cyclic phosphates) found at the 3ˊ end of RNA fragments produced by RNase L and other metal-ion-independent ribonucleases (Yang, 2011). Cyclic phosphate cDNA libraries were made using RNA from uninfected and IAV∆NS1-infected A549 cells (Figure 4.1B,

checkmarks). We compared data from four cDNA libraries: 1) uninfected control shRNA-

treated cells, 2) uninfected RNase L shRNA-treated cells, 3) IAV∆NS1-infected control

shRNA-treated cells, and 4) IAV∆NS1-infected RNase L shRNA-treated cells (Figure

4.2). Ribonuclease cleavage sites were found in host and viral RNAs (Figure 4.2A).

Ribosomal RNAs (28S, 18S, 5.8S, and 5S rRNAs) and U6 snRNA accounted for a large

portion of cleavage sites in all four samples (Figure 4.2A). More than 10% of all

cleavage events were in IAV RNAs in control shRNA-treated cells (Figure 4.2A, IAV,

). Cleavage within IAV RNAs was reduced by RNase L shRNA (Figure 4.2A, IAV, ).

It is important to note that IAV PA and PA-X endonucleases are metal-ion-dependent enzymes that do not produce cyclic phosphates at RNA cleavage sites 116

Figure 4.2. Frequency of ribonuclease cleavage sites in host and viral RNAs from A549 cells. 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing were used to detect ribonuclease cleavage of host and viral RNAs. A. Amounts of cleavage in host and viral RNAs detected in cDNA libraries from uninfected and IAV∆NS1-infected A549 cells treated with either control or RNase L shRNA. Frequency of cleavage in individual IAV RNA segments: genomic (B) and anti- genomic/mRNA (C).

(Crépin et al., 2010; Yuan et al., 2009). Consequently, the cyclic phosphate cDNA libraries that I use to identify RNase L cleavage sites do not contain confounding PA endonuclease cleavage sites.

Ribonuclease cleavage sites were detected in each segment of IAV RNA, in both

(-) and (+) strands (Figures 4.2B, 4.2C and Table 4.1). PB2, PB1 and PA genomic (-) strand RNAs were cleaved more frequently than other IAV genomic RNAs (Figure 4.2B, 117

Table 4.1. Ribonuclease cleavage of IAV RNAs

and Table 4.1). Among IAV (+) strand RNAs, M was cleaved most frequently (Figure

4.2C and Table 4.1). As noted above, RNase L knockdown reduced the amounts of cleavage in each IAV RNA segment (Figures 4.2B, 4.2C, and Table 4.1). Dinucleotides frequently targeted in genomic strands include UU > UG > GU > CU > AU > UA

(Figures 4.3 and 4.4). In antigenomic RNAs/mRNAs, dinucleotides frequently targeted include UU > UG > UC > AU/CU > UA (Figures 4.3 and 4.4). Knockdown of RNase L

reduced cleavage predominantly at UA and UU dinucleotides, especially within IAV (+) strand RNAs (Figures 4.3 and 4.4). RNase L knockdown reduced cleavage at UA and

UU dinucleotides in IAV (+) strands by nearly 10-fold and 14-fold, respectively (Figure 118

Figure 4.3. Dinucleotides at ribonuclease cleavage sites in IAV RNAs. (IAV   RNAs 1-8). IAV∆NS1-infected A549 cells: Ctrl shRNA ( ) & RNase L shRNA ( ) treated cells. % Total cDNA Reads at each of 16 dinucleotides in IAV genomic RNA segments (left side) and anti-genomic / mRNA segments (right side).

119

Figure 4.4. Dinucleotides at ribonuclease cleavage sites in IAV RNAs (genome-   wide). IAV∆NS1-infected A549 cells: Ctrl shRNA ( ) & RNase L shRNA ( ) treated cells. A. IAV (-) strand RNAs. B. IAV (+) strand RNAs. Upper panel: % Total cDNA Reads at each of 16 dinucleotides in IAV RNAs. Lower panel: Fold reduction in cleavage due to RNase L knockdown (Ratio of % Total cDNA Reads in Ctrl:RNase L shRNA-treated samples). Gray line: ratio of 1 (no change). Red: Average reduction for all dinucleotides.

4.4). In genomic strands, knockdown of RNase L reduced cleavage at UA dinucleotides

by 9-fold, while cleavage at UU dinucleotides in IAV (-) strands was not substantially impacted by RNase L knockdown. Significant amounts of cleavage were evident 3ˊ of uridines (UU, AU, GU, and CU dinucleotides), most notably within IAV genomic RNAs

(Figures 4.3 and 4.4). Cleavage 3ˊ of uridines is characteristic of some RNase A family

enzymes (Hofsteenge et al., 1998; Shapiro et al., 1986a, 1986b). 120

Cleavage Sites Mapped to Discrete Regions within IAV RNAs

Discrete regions of PB2, PB1, PA, NP and M RNA segments accounted for the

majority of cleavage events within IAV RNAs. Notably, significant amounts of cleavage

were observed within and adjacent to areas of increased synonymous site conservation

(Figure 4.5). The observed-to-expected ratio (obs/exp) and corresponding p-values for synonymous site conservation are shown (Figure 4.5). The obs/exp values for synonymous site conservation allow for a direct comparison across different IAV RNA segments (Figure 4.5). The IAV NS RNA segment was excluded from these analyses due to the deletion in this RNA and the low number of cleavage events detected in the residual ∆NS1 RNA (Figures 4.2B; 4.2C; and Table 4.1). Prominent cleavage sites within and near regions of increased synonymous site conservation in M RNA segments were mapped to base-by-base sequences (Figure 4.6).

PB2 (-) strand RNA was cleaved in a discrete region spanning nucleotides 250 to

650, sequences without increased synonymous site conservation (Figure 4.5). PB1

RNA was cleaved frequently in both the (-) and (+) strands, with substantial amounts of

cleavage at the 3ˊ end of the (-) strand and the 5ˊ end of the (+) strand (Figure 4.5). The

endonuclease cleavage sites in PB1 coincide with areas of increased synonymous site

conservation, including the areas associated with the PB1-F2 ARF (Figure 4.5). PA

RNA was cleaved at discrete locations in the (-) strand, within and adjacent to areas of

increased synonymous site conservation (Figure 4.5). HA and NA RNAs, which have

relatively low amounts of synonymous site conservation, had limited amounts of

cleavage compared to other IAV RNA segments (Figure 4.5; Table 4.1). NP and M

RNAs were cleaved predominantly in (+) strands, within and adjacent to areas of

increased synonymous site conservation (Figure 4.5). Cleavage sites within M (+)

strands were predominantly 3ˊ of UpN dinucleotides, and cleavage at these sites were

121

Figure 4.5. Frequency and location of ribonuclease cleavage sites in IAV RNAs in relation to synonymous site conservation. Cleavage sites (# of cDNA reads) in IAV negative- (Blue bars) and positive-strand (Red bars) RNAs plotted alongside synonymous site conservation (SSC). SSC (grey lines): observed # of substitutions in a 15-codon sliding window / expected # of substitutions along the window. Dashed line indicates SSC obs/exp ratio of 1. Cleavage sites from IAV∆NS1-infected control shRNA-treated A549 cells.

122

Figure 4.6. Prominent cleavage sites in M (+) strand RNA. IAV∆NS1-infected A549 cells: Ctrl shRNA () and RNase L shRNA () treated cells. % Total cDNA Reads at indicated positions in M (+) strand RNA from nts 100 - 259. Annotations include locations of M1 ORF, M42 ARF, M4 intron, M4 splice-donor sequence (m4SD), and the RNA structure described by Moss et al., 2011.

almost completely eliminated by RNase L knockdown (Figures 4.3 and 4.6). The IAV

RNA sequence most frequently cleaved by RNase L was present in M (+) strand RNA; a

15 base sequence 200AUUUUAGGAUUUGUG214 with more than 200 cDNA reads predominantly at UU, UA, and UG dinucleotides (Figure 4.6). In contrast, RNase L knockdown reduced the number of cDNA reads in this region to 13 (Figure 4.6). These

RNase L cleavage sites lie within an area of increased synonymous site conservation

(Figure 4.5, M RNA), 50 bases downstream from the M4 splice-donor sequence (Figure

4.6), within an RNA structure of unknown function (Moss et al., 2011). 123

RNase L-Dependent Cleavage sites Within rRNAs

RNase L-dependent and RNase L-independent cleavage sites within rRNAs were clearly apparent in the cDNA libraries (Figure 4.7; Tables 4.2 and 4.3). 18S rRNA

was cleaved most frequently by RNase L at UU743 and UU541 (Figures 4.7; Table 4.3),

consistent with the experimental data from PV-infected HeLa cells in Chapter III (Cooper

et al., 2014). 18S rRNA was cleaved less frequently by RNase L at UC136, UC199, CU540, and UU1297 (Figure 4.7; Table 4.3). 28S rRNA was cleaved in an RNase L-dependent

manner at UG2486, UU2741 and UG2762 (Figure 4.7; Table 4.2). Two cleavage sites in 28S

rRNA, CU4031 and UG4032, previously reported to be RNase L-dependent (Iordanov et al.,

2000), were evident in IAV∆NS1-infected cells; however, cleavage at these sites was not diminished by RNase L knockdown, preventing attribution to RNase L activity. RNase L- dependent cleavage sites in 28S and 18S rRNAs were on the surface of ribosomes

(Figure 4.7).

RNase L-Dependent Cleavage in U3 snoRNA

U3 small nucleolar RNA (snoRNA) is a C/D box containing non-coding RNA predominantly located in the nucleolus where it is involved with rRNA processing and biogenesis. Human U3 snoRNA is 217 nts in length, and is important for guiding site- directed cleavage of 18S rRNA from the 47S transcript (Beltrame and Tollervey, 1995;

Borovjagin and Gerbi, 1999; Dragon et al., 2002). RNase L-dependent cleavage occurred in single-stranded regions of stem loop II at UU122, CU134, UU135, UG136, GU179,

and UA180 (Figure 4.8).

124

Table 4.2. RNase L-dependent and -independent cleavage sites in 28S rRNA from A549 cells

20 most frequent cleavage sites (benchmarked to IAVΔNS1-infected control shRNA-treated A549 cells). % Total cDNA reads at individual sites in 28S rRNA. Red: RNase L-dependent cleavage sites. > 5-fold increased cleavage in IAVΔNS1-infected control shRNA-treated A549 cells as compared to uninfected control shRNA-treated cells and uninfected RNase L shRNA-treated cells, and > 2.5-fold increased over IAVΔNS1-infected

125

Table 4.3. RNase L-dependent and -independent cleavage sites in 18S rRNA from A549 cells

20 most frequent cleavage sites (benchmarked to IAVΔNS1-infected control shRNA-treated A549 cells). % Total cDNA reads at individual sites in 18S rRNA. Red: RNase L-dependent cleavage sites. > 5-fold increased cleavage in IAVΔNS1-infected control shRNA-treated A549 cells as compared to uninfected control shRNA-treated cells and uninfected RNase L shRNA-treated cells, and > 2.5-fold increased over IAVΔNS1-infected RNase L shRNA-treated cells.

126

Figure 4.7. RNase L-dependent cleavage sites in ribosomal RNA and 80S ribosomes. Human ribosome structures from Anger et al., 2013. Frequency and location of cleavage sites in 28S (A) and 18S (B) rRNAs. RNase-dependent sites in red. C. Location of RNase L cleavage sites in 80S ribosomes. PDB files 3J3A, 3J3B, 3J3D, and 3J3F (Anger et al., 2013) were visualized using PyMOL (DeLano Scientific, San Carlos, California, USA; http://www.pymol.org).

127

Figure 4.8. RNase L-dependent cleavage sites in U3 snoRNA. A. Frequency and

location of cleavage sites in U3 snoRNA. B. RNase L-dependent cleavage sites mapped onto the secondary structure of U3 snoRNA. Human U3 snoRNA (www- snorna.biotoul.fr) Mfold structure re-drawn using R2R (Weinberg and Breaker, 2011; Zuker, 2003).

Discussion

RNase L is a key effecter of type I interferon-regulated antiviral responses

(Silverman, 2007). In recent years, viral countermeasures to RNase L have been described, reinforcing the important nature of this host defense mechanism for picornaviruses (Han et al., 2007; Sorgeloos et al., 2013), coronaviruses (Zhao et al.,

2012), rotaviruses (Zhang et al., 2013) and influenza viruses (Min and Krug, 2006)

(Table 1.2). Nonetheless, the antiviral mechanisms of RNase L are not completely understood, in part because the host and viral RNAs targeted by RNase L to promote its antiviral effects are not well-defined. To provide insight to the antiviral activity of RNase L and the importance of IAV NS1 protein, I used 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing methods (Chapter II and III (Cooper et al., 2014)) to identify the targets of RNase L in IAV∆NS1-infected A549 cells. 128

Viral RNAs, NS1, and RNase L

It is well established that IAV NS1 prevents the activation of RNase L (Min and

Krug, 2006), among other functions (Engel, 2013; Hale et al., 2008; Krug, 2015). When

NS1 is disabled, rRNAs were cleaved in an RNase L-dependent manner (Figure 4.1B), and viral replication was impaired (Figure 4.1C). In comparison to control shRNA-

treated cells, cells treated with shRNA specific for RNase L did not exhibit rRNA

fragmentation in response to IAV infection, and these cells supported increased

magnitudes of IAV∆NS1 replication (Figure 4.1C). Although RNase L shRNA-treated

A549 cells supported greater viral yields, less cleavage if IAV RNAs were detected from

RNase L shRNA treated cells compared to the cells treated with control shRNA (Figure

4.2B and C).

In cells treated with control shRNA, infection with IAVΔNS1 resulted in RNase L

activation and cleavage of viral RNA segments predominantly at UA, UU, and UG

dinucleotides (Figures 4.3 and 4.4). These results indicate that RNase L limits IAV

infections, in part, by directly cleaving viral RNAs when NS1 is disabled. Certain IAV

RNAs were targeted more frequently than others by RNase L and other

endoribonucleases, and included genomic segments encoding PB2, PB1, and PA

polymerase subunits, and the antigenomic segments/mRNAs encoding PB1, NP, and M.

Why are certain IAV segments targeted over others? Does the absence of NS1 In

infections with IAV impact the stability of certain IAV RNAs? Previous studies to assess

the role of NS1 on viral protein synthesis revealed that NS1 was important for the

expression of the IAV M1 protein. M1 protein is a major structural component of the

virion, and in contrast to NS1 protein, which is expressed early in infection, M1 protein is

expressed later in infection (Shapiro et al., 1987; Vester et al., 2010). In IFN-signaling

competent cells infected with virus containing deletions or mutations in NS1, the

expression of M1 protein is severely debilitated (Egorov et al., 1998; Khaperskyy et al., 129

2012; Salvatore et al., 2002). These studies suggest that NS1 might be an important factor in the stability of M1 RNA. In the studies described in this chapter, substantial cleavage is observed in the M1 coding sequence, or an intron derived from alternative splicing to produce other M segment derived proteins (Figures 4.5 and 4.6). In another study, NS1 was found to stimulate translation of viral RNA segments that contain the sequence motif, “GGUAGAUA”, found in the 5ˊ-NTR of both M and NP genomic segments (Enami et al., 1994). The translational stimulation of NS1 on M and NP was independent of their transcription. It is possible that without functional NS1 protein, M and NP mRNA are degraded in the cytosol by RNase L late in infection (24hpi). While the relationship between NS1 and the RNA for M and NP is intriguing, it does not correspond to the RNA for PB1. The “GGUAGAUA” sequence is not found in PB1 RNA, and expression of a construct containing the 5ˊ-NTR sequence of PB1 was not influenced by NS1 (Enami et al., 1994), thus, this scenario involving “GGUAGAUA” sequences does not explain the sensitivity of PB1 (+) strands to RNase L. Future studies that monitor the synthesis and abundance of the three viral RNA species: (-) strand genomic segments, (+) strand antigenomic segment, and mRNAs during infections might provide insight into how NS1 influences viral RNA stability and susceptibility to RNase L cleavage (Kawakami et al., 2011; Vester et al., 2010).

2ˊ, 3ˊ-cyclic phosphate cDNA libraries from IAVΔNS1-infected A549 cells also revealed that RNase L and other ribonucleases targeted regions of some IAV RNAs within and adjacent to areas of increased synonymous site conservation (Figure 4.5).

These RNAs included PB1 (+) and (-) strands, PA (-) strands, and NP and M (+) strands

(Figure 4.5). In a reciprocal manner, HA and NA RNAs, which by comparison have lower amounts of synonymous site conservation, exhibited little cleavage by RNase L or other ribonucleases (Figure 4.5). Based on the patterns of synonymous site conservation and ribonuclease cleavage, it might be important to consider the potential 130 roles of ribonucleases in vRNA packaging and reassortment. Because HA and NA RNA segments are more refractory to riboonucleases, they might reassort more successfully than other RNA segments.

In Chapter III, I found that PV RNA was cleaved by RNase L and other metal- ion-independent endoribonucleases; however, only the positive-strand of PV RNA was cleaved in infected cells (Chapter III; (Cooper et al., 2014)). Ribonuclease cleavage sites were conspicuously absent in PV negative-strands. In experiments described here, cleavage sites were abundant in both negative strand and positive strand IAV RNAs.

While I found that knockdown of RNase L in A549 cells produced fewer cDNA reads corresponding to IAV RNA, increased amounts of PV-specific reads were detected in

HeLa cells expressing a dominant-negative form of RNase L as compared to HeLa cells expressing WT RNase L. These paradoxical observations might be attributed to the nature of membranous viral replication complexes involved in positive-strand RNA virus replication (Ahlquist, 2006). The membranous replication complexes of positive-strand viruses sequester replicating viral RNA and provide a compartment where the PV- encoded ciRNA can overwhelm small amounts of RNase L, preventing the cleavage of viral negative-strand templates and nascent positive-strands (Ahlquist, 2006; Han et al.,

2007; Townsend et al., 2008a) In contrast to PV, IAV RNAs are localized intermittently in both the nucleus and the cytoplasm, and they are not sequestered within membranous replication complexes. In the current study, we found that IAV RNAs were cleaved by

RNase L and other metal-ion-independent ribonucleases in both negative and positive strands (Figures 4.2-4.4 and Table 4.1). In fact, PB2, PB1 and PA negative strand genomic segments were among the most frequently cleaved IAV RNAs (Figures 4.2,

4.3 and Table 4.1), suggesting that the nucleocapsid that interacts with viral RNAs does not protect IAV RNAs from cleavage, consistent with structural studies of RNPs (Arranz et al., 2012) 131

Frequent cleavage after uridines within IAV RNAs (Figures 4.3 and 4.4) is

consistent with the specificity of RNase L (UU) which can be found in the cytoplasm and

nucleus (Bayard and Gabrion, 1993; Al-Ahmadi et al., 2009), and of an RNase A family member, RNase 4 (Hofsteenge et al., 1998; Li et al., 2013). RNase 4 is taken up by cells

where it traffics to the nucleus (Li et al., 2013). The cleavage detected after uridines in

genomic segments was not reduced significantly by RNase L shRNA (Figure 4.4).

Genomic RNA segments, especially at 24 hpi, would be present in both the cytoplasm

and the nucleus. Thus, based on the nature of the cleavage sites (uridines) and their

resistance to RNase L knock down, it is reasonable to implicate RNase 4 and potentially

other RNase A family enzymes in the cleavage of IAV RNAs.

RNase L-Dependent Cleavage Sites in rRNA and U3 snoRNA

rRNAs are obvious targets of RNase L (Figure 4.1B). In this study RNase L-

dependent cleavage sites in 18S and 28S rRNAs were observed (Figure 4.8, Tables 4.2

and 4.3). Compared to uninfected cells, particular sites in rRNA from IAVΔNS1-infected

A549 cells demonstrated > 5-fold increased cleavage. RNase L knockdown reduced the

amounts of cleavage specifically at these RNase L-dependent sites (Tables 4.2 and

4.3). 18S rRNA UU743 and UU541 were the most frequent targets of RNase L among the host and viral RNAs from A549 cells (Figure 4.7B, Table 4.3). These are the same

RNase L-dependent 18S rRNA cleavage sites identified in PV-infected HeLa cells

(Chapter III; Figures 3.5 and 3.8; (Cooper et al., 2014)). The antiviral activity of RNase

L likely involves cleavage at these sites in 18S rRNA. Cleavage at these sites could inhibit protein synthesis, thereby contributing to the inhibition of viral gene expression.

Like studies with PV-infected HeLa cells, cleavage sites in 28S rRNAs were also

discovered. Two of these sites CU4031 and UG4032, were heavily induced by IAVΔNS1

infection, but were not reduced by RNase L knockdown (Table 4.2). Studies where 132

IRE1β was overexpressed in HeLa cells revealed fragmentation of 28S rRNA. CU4031 and UG4032 were identified as IRE1β-dependent cleavage sites in 28S rRNA through primer extension (Iwawaki et al., 2001).

As I discussed in Chapter III, cleavage at UU743 and UU541 in 18S rRNA may have potential implications for ribosome function. Cleavage at UU743 may disrupt

interactions between ES6S and ES3S, important for initiation factor binding (Alkemar

and Nygård, 2003, 2004; Anger et al., 2013), and might disrupt the binding of rpS7, a

protein that inhibits MDM2-mediated degradation of p53 (Chen et al., 2007; Zhu et al.,

2009). Cleavage at UU541, located within helix 16 of 18S rRNA, could inhibit mRNA scanning by the 40S ribosome to prevent mRNA translation (Hashem et al., 2013; Ben-

Shem et al., 2010; Passmore et al., 2007).

The cleavage of UU743 and UU541 should also be considered in the context of

other antiviral factors likely activated during IAV infections when NS1 is deleted. In

addition to the OAS/RNase L pathway, IAV also inhibits PKR. PKR might facilitate

RNase L-dependent rRNA cleavage by inducing the formation of stress granules,

cytoplasmic foci that contain translationally stalled mRNA, 18S rRNA, and various

proteins (Kedersha et al., 2000, 2002,1999; Onomoto et al., 2014). RNase L, OAS, RIG-

I, and MDA5 localize to stress granules (Langereis et al., 2013; Onomoto et al., 2012),

where RNase L might target 18S rRNA. Stress granules form in HeLa cells infected with

IAVΔNS1, but do not form during infections with WT IAV (Onomoto et al., 2012). Does

RNase L cleave 18S rRNA in the context of stress granules? And, would this explain

why only 18S rRNA, and not 28S rRNA, is frequently and consistently cleaved by RNase

L? Determining the functional significance of RNase L-dependent cleavage sites in 18S

rRNA, and the context of where they occur in the cell, warrants future investigation.

Another unexpected finding was the cleavage of U3 snoRNA (Figure 4.8).

Cleavage of U3 snoRNA by RNase L has implications for the biogenesis and processing 133

of rRNA and downstream activation of p53. Base pairing between U3 snoRNA and the

47S pre-rRNA transcript mediate site-specific cleavage to liberate 18S rRNA. The 47S

transcript contains 18S, 5.8S, and 28S rRNA in addition to internal and external

transcribed sequences that must be removed for proper processing and assembly of the

small and large ribosomal subunits (Beltrame and Tollervey, 1995; Borovjagin and

Gerbi, 1999; Dragon et al., 2002). Defects in 18S or 28S processing induce the ribotoxic

stress response, where ribosomal proteins are released into the nucleoplasm to stabilize

and activate p53, resulting in cell cycle arrest and/or apoptosis (Hölzel et al., 2010; Kim

et al., 2014; Yuan et al., 2005). Potentially, RNase L-induced cleavage of U3 snoRNA

could inhibit 18S rRNA processing to induce p53-mediated apoptosis. Further

investigation is needed to determine the functional impact of U3 snoRNA cleavage.

U3 snoRNA also associates with an IFN-induced, single-strand specific, 3ˊ-5ˊ exonuclease, ISG20 (Espert et al., 2003; Nguyen et al., 2001). ISG20 is metal-ion- dependent (Nguyen et al., 2001), and thus would not leave 2ˊ, 3ˊ-cyclic phosphates on degraded RNA intermediates detectable by 2ˊ, 3ˊ-cyclic phosphate cDNA methods. IFN- treatment of cells enhances the accumulation of ISG20 within cajal bodies, where ISG20 potentially has roles in regulation of snoRNA and snRNA biogenesis or maturation

(Espert et al., 2003). The significance of ISG20 and its interactions with snoRNAs and snRNAs has yet to be determined, but suggests interesting and unexplored roles of snoRNAs in IFN-mediated defenses against viruses.

Summary

In this chapter, I have described how RNase L contributes to inhibition of IAV

when NS1 is disabled. 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing

methods detected RNase L-dependent cleavage in IAV viral RNAs, rRNAs, and U3

snoRNA. Direct cleavage of IAV RNA contributes to IAV restriction, cleavage of rRNA 134

potentially inhibits viral protein synthesis, and cleavage of U3 snoRNA might inhibit

rRNA biogenesis and induce pro-apoptotic pathways in the infected cell. In the next

chapter, I apply 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods to RNAs from HCV-infected tissue culture cells and a human liver sample to

characterize RNA cleavage provoked during HCV-infections. 135

CHAPTER V

HEPATITIS C VIRUS INFECTION: RIBONUCLEASE CLEAVAGE SITES

IN HOST AND VIRAL RNAS

Introduction

In Chapters III and IV, I described ribonuclease cleavage sites in host and viral

RNAs from PV-infected HeLa cells and IAV-infected A549 cells using 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and high-throughput sequencing methods. These same methods are used in this chapter to identify ribonuclease cleavage sites in host and viral

RNA from hepatitis C virus (HCV)-infected Huh7.5.1 cells and liver tissue.

HCV, a major global health problem, infects approximately 130 – 150 million people worldwide, amounting to 2.2 – 3% of the global population (World Health

Organization, 2014). Transmission of HCV is primarily through injection drug use, improperly screened blood products, and improperly sterilized hospital equipment. Less commonly, HCV can be sexually transmitted. Acute HCV infections tend to go unnoticed since they are typically asymptomatic. However, more than 50% of acute HCV infections will transition into chronic infections that can then progress to liver failure and liver cancer. Liver failure and liver cancer attributed to chronic HCV infections account for nearly 350,000 deaths annually (World Health Organization, 2014) and HCV is a leading cause of chronic hepatitis, cirrhosis, liver cancer, and indication for liver transplantation in the Western world (Rosen, 2011). There are six genotypes (HCV1-6) of HCV, each with multiple subtypes (i.e. HCV1A, HCV1B) (Smith et al., 2014). Like PV, HCV is a (+) strand RNA virus that replicates within the cytoplasm of infected cells.

The mechanisms by which innate and acquired immune responses impact acute

HCV infections are only partially defined. Likewise, the host-virus interactions that occur during chronic infections, and their impact on disease progression, are poorly understood. The Barton lab is interested in the mechanisms by which HCV establishes 136 chronic infections and provokes liver disease. In particular, we seek to understand the roles of endoribonucleases during acute and chronic HCV infections.

Endoribonucleolytic host responses are predicted to influence the outcomes of HCV infections and IFN-based treatment regimens (Han and Barton, 2002; Malathi et al.,

2010; Mihm et al., 2009, 2010; Washenberger et al., 2007). In addition, endoribonucleases provoke cytopathic effects that likely contribute to HCV-associated cirrhosis and hepatocellular carcinoma (Selitsky et al., 2015; Zhou et al., 2013).

Increased expression of angiogenin, an RNase A superfamily member involved in angiogenesis (Fett et al., 1985), is correlated with tumor vascularization and poor prognosis for heptocellular carcinomas (Hisai et al., 2003). Angiogenin is also upregulated in many other types of cancers (Li and Hu, 2010), where it is secreted, taken up by neighboring cells, and transported to the nucleus (Hu et al., 1997, 2000;

Tsuji et al., 2005). Once in the nucleus, angiogenin promotes 47S rDNA transcription necessary for the continued replication of cancerous cells (Tsuji et al., 2005; Xu et al.,

2002). In addition to promotion of angiogenesis, angiogenin cleaves tRNAs within their anticodon loop to produce 5ˊ and 3ˊ- tRNA halves, or tRNA-derived- stress induced small RNAs (tiRNAs), in response to cellular stress in humans (Fu et al., 2009a;

Yamasaki et al., 2009). tRNAs are currently the only defined substrate of angiogenin; however it is suspected that angiogenin cleaves other RNA targets in the cell since its ribonuclease activity is required for angiogenin-dependent stimulation of angiogenesis

(Shapiro and Vallee, 1989; Shapiro et al., 1989).

The accumulation of tiRNAs have been studied using deep sequencing of small

RNA libraries from RSV-infected A549 cells (Wang et al., 2013). These libraries revealed that tiRNAs were generated in a manner specific to RSV-infection, since they did not accumulate in 293 cells infected with human metapneumovirus. A particular tiRNA, 5ˊ- tiRNA-Glu-CUC was abundant during RSV infection, and its accumulation was reduced 137 when angiogenin expression was repressed using siRNA. RSV replication in the presence of a synthetic 5ˊ-tiRNA-Glu-CUC enhanced RSV replication, suggesting that

5ˊ-tiRNA-Glu-CUC was beneficial for the virus (Wang et al., 2013).

Several lines of evidence suggest that HCV infection provokes an angiogenin-like stress response. Infection of Huh7 cells with J6/JFH1, an HCV2A chimeric virus resulted in six-fold increased expression of angiogenin over uninfected cells (Presser et al.,

2011). Deep sequencing of small RNA libraries generated from liver biopsies of uninfected patients and of patients chronically infected with HCV or hepatitis B virus

(HBV) demonstrated that a large proportion of cDNA reads were derived from tiRNAs in patients with chronic HCV or HBV infection compared to those from uninfected liver

(Selitsky et al., 2015). Oxidative stress, provoked by HCV infections, may also serve to activate angiogenin during infection (Koike and Miyoshi, 2006).

RNA fragments produced by RNase A family members terminate with 2ˊ, 3ˊ- cyclic phosphates and 5ˊ-hydroxyl groups (Rybak and Vallee, 1988; Sorrentino, 2010;

Table1.1). The RNA fragments produced by RNase A enzymes, like angiogenin, would be detected using by 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods that were previously used in Chapters III- IV to reveal a consistent pattern of ribonuclease cleavage sites in host and viral RNAs from PV-infected HeLa cells and

IAV-infected A549 cells (Cooper et al., 2014, 2015). As described in this chapter, I sought to determine whether JFH1-infection of Huh7.5.1 cells provoked the generation of tiRNAs consistent with the angiogenin stress response and in turn identify other RNA substrates cleaved during this stress response. Using cyclic phosphate cDNA synthesis and Illumina sequencing methods, I found that the tRNA-Glu-CUC was cleaved within its anticodon loop at 4 and 5 days post infection (dpi) in JFH1-infected Huh7.5.1 cells.

Coincident with tRNA cleavage, I also detected a novel series of cleavage sites in 28S,

18S, and 5S rRNAs, and U3 snoRNA. Whether cleavage in 28S rRNA, 18S rRNA, 5S 138

rRNA, and U3 snoRNA can be attributed to angiogenin is unclear, since the only defined

substrate for angiogenin is tRNA. Other RNase A family members (Table 1.1), like

RNase 1, 2, and 4, possibly contribute to RNA cleavage in Huh7.5.1 cells, since these

enzymes are expressed in the liver (Futami et al., 1997; Sorrentino et al., 1988). RNA

cleavage detected in HCV-infected Huh7.5.1 cells occurred 3ˊ of pyrimidine bases,

consistent with the specificity of angiogenin and other RNase A family enzymes (Curran

et al., 1993; Rybak and Vallee, 1988; Sorrentino, 2010). I also generated cyclic

phosphate cDNA libraries from a liver section isolated from a patient chronically infected

with HCV1A, and again found an enrichment of tiRNAs corresponding to 5 ˊ-tiRNA-Glu-

CUC, in addition to 5ˊ-tiRNA-Gly-GCC and 5ˊ-tiRNA-Val-CAC, and detected cleavage in

U3 snoRNA and 18S rRNA. I discuss the potential impact of cleavage of these RNAs on

HCV infections and HCV-induced pathologies.

Materials and Methods

HCV-Infected Huh7.5.1 Cells

Huh 7.5.1 hepatoma cells were obtained from Francis Chisari (Scripps Research

Institute, La Jolla, CA). Huh7.5.1 cells were cultured and infected with HCV2A JFH1

(MOI of 0.1 FFU per cell) as previously described (Lindenbach et al., 2005; Wakita et al.,

2005). Mock and JFH1-infected Huh7.5.1 cells were resuspended in Qiagen lysis buffer

(Buffer ATL) and run through a Qiagen Qiashredder column for tissue homogenization

(Preparation of JFH1-infected Huh7.5.1 cells was performed by Katelyn Leahy of the

Hugo Rosen Lab). RNA was isolated from uninfected cells at the time of infection (0 dpi)

and HCV-infected cells at 1, 2, 3, 4 and 5 days post-infection (dpi). The amounts of HCV

RNA within each sample were determined by qRT-PCR as previously described

(Golden-Mason et al., 2012): JFH-1 primers (HCV-forward: 5ˊ-CGA CAC TCC GCC ATC 139

AAT CAC-3ˊ; HCV-reverse: 5ˊ-CAC TCG CAA GCG CCC TAT CA-3ˊ; and TaqMan

HCV-probe: 6FAM-AGG CCT TTC GCA ACC CAA CGC TAC T-TAMRA).

HCV1A-Infected Liver Section

A de-identified liver section from a patient with chronic HCV1A infection was

provided by Hugo Rosen (COMIRB protocol 06-0566 from the Colorado Multiple

Institutional Review Board). RNA was isolated from the liver biopsy to examine endoribonuclease cleavage sites in host and viral RNAs (COMIRB protocol 10-1555 from the Colorado Multiple Institutional Review Board). The liver section was incubated with 4M GTC buffer supplemented with proteinase K for one hour at 37°C, and manually homogenized with a micropestel. RNA was extracted using acid phenol chloroform methods.

2ˊ, 3ˊ-Cyclic Phosphate cDNA Synthesis and Illumina Sequencing

Cyclic phosphate cDNA synthesis and Illumina sequencing were performed as

previously described for RNA from patient liver (Figure 2.4; Chapters II, III and IV).

Because I sequenced libraries from the JFH-1-infected Huh7.5.1 cells on the Illumina

HiSeq platform, I used the following DNA linker: 5ˊ-/rApp/GAT CGG AAG AGC ACA

CGT CTG AAC TCC AGT CAC/ddC/-3ˊ for the ligation of cDNAs.

Bioinformatic Data Processing

Data were trimmed, aligned, and processed using methods consistent with those

described in Chapter III and IV. JFH1 (AB047639.1) was used as the reference

sequence for alignments of cDNA libraries from Huh7.5.1 cells to viral RNA. Due to the

variability of HCV subtypes, 582 HCV1A genomic sequences downloaded from the ViPR

database (Pickett et al., 2011) were used for alignments of the cDNA library from 140

HCV1A-infected liver. To map cleavage sites onto individual tRNAs, reads that aligned to hg18-tRNAs (gtrnadb.ucsc.edu), were re-aligned to individual tRNAs. tRNAs with anticodon loop cleavage were parsed from total mapped tRNA reads by searching for tRNAs with cleavage spanning nucleotide positions 30 to 40.

Results

Cellular stress provoked by certain chemical treatments, hyperosmotic stress and nutrient deprivation result in activation of the angiogenin-dependent tRNA stress response in human cells (Fu et al., 2009a; Yamasaki et al., 2009). Others have reported that RSV-infections also provoke cleavage of tRNAs by angiogenin (Wang et al., 2013).

To determine whether JFH1 (HCV2A)-infection of Huh7.5.1 cells provokes an angiogenin-like stress response, I generated 2ˊ, 3ˊ-cyclic phosphate cDNA libraries using RNA from JFH1-infected Huh7.5.1 cells.

During JFH1-infection over a span of 1-5 days, HCV genomes increased in abundance within infected cells (Figure 5.1). Relatively small amounts of HCV RNA were present in cells at 1 dpi, with increasing amounts accumulating at 2-5 dpi (Figure

5.1). These data are consistent with those reported by others (Lindenbach et al., 2005;

Wakita et al., 2005). Endoribonuclease cleavage sites were evident in cellular RNAs from both uninfected and HCV-infected cells (Figure 5.2A). The relative amounts of cleavage in HCV RNAs (0.01 to 0.05 % of Total cDNA Reads at 2 to 5 dpi) were small in comparison to the relative amounts of cleavage in cellular RNAs (Figure 5.2B, note y- axis scale). Host rRNAs (28S, 18S, 5.8S and 5S rRNAs) accounted for a large portion of the library (Figure 5.2A). Reads corresponding to 28S rRNA accounted for 60-80% of cleavage sites (Figure 5.2A), and their frequency increased during the course of the

HCV infection whereas the relative frequency of cleavage in other cellular RNAs tended to decrease during the course of the HCV infection. 141

Figure 5.1. Accumulation of HCV genomic RNA in HCV-infected Huh7.5.1 cells. Huh7.5.1 cells were infected with HCV JFH1 at an MOI of 0.1 FFU per cell. RNA was isolated from uninfected cells at the time of infection and HCV-infected cells at 1, 2, 3, 4 and 5 dpi. The amount of HCV RNA in samples was determined by qRT-PCR.

Figure 5.2. Frequency of ribonuclease cleavage sites in host and viral RNAs from Huh7.5.1 cells. 2ˊ, 3ˊ-cyclic phosphate cDNA libraries were prepared using RNA samples from HCV-infected Huh7.5.1 cells and were sequenced on the Illumina HiSeq. A. The relative amounts of cDNA detected in host and viral RNAs are indicated as a % of Total cDNA Reads. B. The frequency of HCV positive (teal)- and negative-strand (gray) cDNA reads. Numbers within the bars indicate the # of cDNA reads corresponding to HCV positive and negative strands.

142

Cleavage of Anticodon Loop of tRNAs

The cleavage of tRNAs within their anticodon loops is mediated by the activity of

angiogenin, a member of the RNase A superfamily (Sorrentino, 2010). Angiogenin

cleaves single-stranded RNA after pyrimidines to generate fragments with terminal 2ˊ,

3ˊ-cyclic phosphates (Rybak and Vallee, 1988). Cleavage sites in tRNAs made up 0.4% to 2.3% of the cyclic phosphate libraries (Figure 5.2A; tRNAs), and some of the cleavage sites in tRNAs corresponded to anticodon loop cleavage (Figure 5.3).

Cleavage was detected in several tRNAs containing anticodon loops specific for aspartic acid, glutamine, glutamic acid, glycine, histidine, leucine, lysine, tyrosine, and valine

(Figure 5.3); however, cleavage in the anticodon loop for the tRNA-Glu increased by 26-

fold from the uninfected sample to the 5 dpi sample (Figure 5.4). The majority of

anticodon loop cleavage in tRNA-Glu corresponded to tRNA-Glu-CUC rather than tRNA-

Glu-CUU. Because tRNA genes are highly repetitive throughout the human genome, I

aligned to a single tRNA gene for glutamic acid-CUC to determine the precise positions

of cleavage throughout the anticodon loop (Figure 5.4A). I observed cleavage directly in

the anticodon CUC sequence as well as after a cytidine just 5ˊ-of the anticodon.

Cleavage 3ˊ of pyrimidines, especially after cytidine, is consistent with the specificity of

angiogenin (Curran et al., 1993; Rybak and Vallee, 1988; Shapiro et al., 1986b). While I detected cleavage directly within the anticodon of tRNA-Glu-CUC, another group

reported RSV infection-induced cleavage of tRNA-Glu-CUC at positions slightly

upstream of the anticodon (Wang et al., 2013). These differences may reflect variables

in cDNA library preparation methods. Because A. th. tRNA ligase specifically links an

RNA linker to RNAs with 2ˊ, 3ˊ-cyclic phosphates, the native ends of 5ˊ-tiRNAs may be

more efficiently detected, providing more reliable data than that obtained with other

143

Figure 5.3. Anticodon loop cleavage in tRNAs from HCV-infected Huh7.5.1 cells. The frequency of tRNA-specific cDNA reads derived from tRNA anticodon loop cleavage (cleavage within nts 30-40) were plotted.

Nucleotide

Figure 5.4. Ribonuclease cleavage sites in tRNA-Glu-CUC. A. The frequency of cDNA reads were mapped to the base-by-base sequence of tRNA59-Glu-CUC (gtrnadb.ucsc.edu; hg18 tRNAs) for uninfected and the HCV-infected samples (1-5 dpi). B. Cleavage sites within the anticodon loop of tRNA-Glu-CUC. Circled bases represent the CUC anticodon, and bases highlighted in blue are cleavage sites C32, 34 35 36 C , U , and C .

144

methods. Since I detected cleavage within the anticodon loop of tRNA-Glu-CUC, consistent with angiogenin activity (Fu et al., 2009a; Wang et al., 2013), I suspected that other viral and cellular RNAs might be cleaved by angiogenin and/or other RNases in the

HCV-infected cells.

Frequency, Location, and Specificity of Ribonuclease Cleavage Sites in HCV RNA

HCV RNA fragments with cyclic phosphates were detected in all samples, with

increasing amounts at 2, 3, 4 and 5 dpi (Figure 5.2B), consistent with the increasing amounts of HCV RNA detected through qRT-PCR (Figure 5.1). Both HCV negative- and positive-strand RNAs were cleaved throughout the infection (Figure 5.2B). The amounts

of HCV-specific positive-strand cDNA reads increased throughout the HCV infection: 19

reads at 1 dpi, 76 at 2 dpi, 80 at 3 dpi, 250 at 4 dpi and 348 at 5 dpi (Figure 5.2B,

number of positive-strand reads annotated in dark blue bars). Likewise, the number of

cleavage sites in HCV negative-strand RNAs increased from 4 reads at 1 dpi to 141

reads at 5 dpi (Figure 5.2B, number of negative-strand UMI reads annotated in light

gray colored bars). The ratio of viral negative strand to positive strand RNA in HCV

replication complexes is reported to be around one negative strand RNA to two-ten

positive strand RNAs (Quinkert et al., 2005). The ratio of cleavage sites detected in

negative strand to positive strand HCV RNAs range from one cleavage site in negative

strand RNA to three to eight cleavage sites in positive strand RNA are consistent with

the expected ratios of negative to positive strand HCV RNA in infected cells

These data indicate that both HCV positive-strand and negative strand RNAs were

cleaved by ribonucleases during HCV infection of Huh7.5.1 cells (Figure 5.1).

145

Figure 5.5. Ribonuclease cleavage sites in JFH1 HCV2A RNA. A. The frequency, location, and specificity of cleavage sites in HCV positive-strand RNA. B. Dinucleotide specificity of cleavage sites in HCV positive-strand RNA.

I examined the frequency, location and specificity of cleavage sites to gain some insight into the nature of the ribonucleases responsible for the cleavage of HCV RNA

(Figure 5.5). The specificity of cleavage sites detected in HCV RNA from 1-3 dpi was ambiguous; however, a pattern characteristic of RNase A family enzymes (cleavage 3ˊ of pyrimidines) became apparent at 4 and 5 dpi (Figure 5.5B). UU406 in the HCV core gene, and AU8738 and GU9008 in the 3ˊ end of NS5B were consistently cleaved in the 4 and 5 dpi samples (Figure 5.5A). The RNase A superfamily includes RNases 1 – 8.

Members of the RNase A superfamily, in general, share sequence and structural homology and cleave 3ˊ of pyrimidines in ssRNA to produce RNA fragments with 2ˊ, 3ˊ- cyclic phosphate termini (Cuchillo et al., 1993; Libonati and Sorrentino, 2001; Sorrentino, 146

2010; Thompson et al., 1994). RNase 4 and RNase 2 prefer to cleave 3ˊ of uridines

(Hofsteenge et al., 1998; Zhou and Strydom, 1993; Sorrentino, 2010) while RNase 5

(angiogenin) prefers to cleave 3ˊ of cytidines (Curran et al., 1993) (Table 1.1). 81% of cleavage sites detected in JFH1 RNA at 5 dpi were after a pyrimidine, and of these, there seemed to be a slight preference for cleavage 3ˊ of uridine over cytidine. RNase 4 and angiogenin share the same promoters and are often co-expressed (Futami et al.,

1997; Li et al., 2013). Based on these observations, I suspect that RNase A enzymes that include angiogenin, RNase 4, or RNase 2, cleave HCV RNA in Huh7.5.1 cells at 4 and 5 dpi. However, studies that assess the expression and function of these ribonucleases during HCV infections are needed to determine whether these RNase A- family enzymes contribute to HCV RNA degradation during HCV infections.

Frequency, Location and Specificity of Ribonuclease Cleavage Sites in rRNAs

rRNAs are the most abundant RNAs within cells and have the most abundant amounts of cyclic phosphates in cDNA libraries (Figure 5.2A). Cyclic phosphates are abundant within rRNAs from both uninfected and HCV-infected cells (Figure 5.2A). The frequency, location and specificity of cleavage sites within 28S, 18S, 5.8S and 5S rRNAs revealed a temporal change in ribonuclease activity during the course of HCV infection

(Figures 5.6-5.8).

To clearly show the temporal change in ribonuclease activity, I subtracted the signal in 28S and 18S rRNAs detected in uninfected cells from the 28S and 18S signal detected in HCV-infected cells (Figures 5.6 and 5.7). Because the cleavage patterns of

28S rRNA detected at 1 and 2 dpi were very similar to the cleavage of 28S from uninfected cells, no distinct cleavage patterns were observed in 28S rRNA for those samples (Figure 5.6, note absence of novel cleavage sites at 1 and 2 dpi). However, by

147

A. B.

Figure 5.6. Ribonuclease cleavage sites in 28S rRNA. A. Frequency and location of cleavage sites in 28S rRNA. B. Dinucleotide specificity of cleavage sites in 28S rRNA.Signal from the uninfected sample was subtracted from HCV-infected samples to highlight temporal changes in RNA cleavage.

3 dpi, several cleavage sites within 28S rRNA increased in frequency relative to

uninfected cells, and continued to increase in frequency during the course of the JFH1

infection (GC133, GC170, CU171, CC998, CC999 and CU4965, among others) (Figure 5.6A).

The dinucleotide specificity of cleavage sites in 28S rRNA reflect the temporal changes observed throughout the infection (Figure 5.6B). Cleavage after NpU and NpC sites were more frequent in 28S rRNA at 3-5 dpi as compared to earlier times during the course of the HCV infection (Figure 5.5B, note cleavage at C and U nucleotides). This activity is consistent with RNase A family endoribonucleases. Almost 96% of the cleavage sites in 28S rRNA were at pyrimidines at 4 and 5 dpi. Of these sites, most are after cytidines. 148

A. B.

Figure 5.7. Ribonuclease cleavage sites in 18S rRNA. A. Frequency and location of cleavage sites in 18S rRNA. B. Dinucleotide specificity of cleavage sites in 18S rRNA. Signal from the uninfected sample was subtracted from HCV-infected samples to highlight temporal changes in RNA cleavage.

Temporal changes in 18S rRNA cleavage were also evident (Figure 5.7). 18S

rRNA was cleaved frequently at several sites in HCV-infected cells (Figure 5.7A: AC69,

GC72, CC73, GU76, UC243, CC324, CC325, UU1112, CC1419, CC1554, CU1555, GC1765, CC1766,

CC1767). Cleavage after pyrimidines, especially after CC and GC dinucleotides, were more frequent in 18S rRNA at 4 and 5 dpi than at earlier times during the course of HCV infections (Figure 5.7B).

5.8S and 5S rRNAs have characteristic cleavage patterns with prominent amounts of cyclic phosphate near their 3ˊ termini (Figure 5.8). The 5.8S and 5S rRNA patterns from uninfected Huh7.5.1 cells are typical of healthy cells (Chapter III; Figure

3.7; (Cooper et al., 2014)). The relative amounts of normal signal near the 3 ˊ ends of 149

Uninfected Uninfected

1 dpi 1 dpi

2 dpi 2 dpi

3 dpi 3 dpi

4 dpi 4 dpi

5 dpi 5 dpi

Figure 5.8. Ribonuclease cleavage sites in 5.8S and 5S rRNAs. Frequency and location of cleavage sites in 5.8S rRNA (A) and 5S rRNA (B).

mature 5.8S and 5S rRNAs decreased during the course of the HCV infection whereas

cleavage at GU72 and UU73 arose at 4 and 5 dpi in 5S rRNA (Figure 5.8B).

I mapped rRNA cleavage sites onto human 80S rRNA structures (Figure 5.8A).

The majority of prominent cleavage sites highlighted in 28S rRNA (Figure 5.5), 18S rRNA (Figure 5.6) and 5S rRNA (Figure 5.7B) were located on the surface of 80S ribosomes (Figure 5.9), accessible to cellular RNases within HCV-infected cells. The pyrimidine specificity of the cleavage sites in ribosomal RNA, again, suggests one or

150

Ends of 5.8S rRNA, 5S rRNA, and tRNA are labeled in gold.

Figure 5.9. Location of ribonuclease cleavage sites in 80S ribosomes. Cleavage sites that increase in frequency during the course of HCV infection were mapped onto the structure of the 80S ribosome (Anger et al., 2013) and are highlighted with red spheres. PDB files 3J3A, 3J3B, 3J3D, and 3J3F (Anger et al., 2013) were visualized using PyMOL (DeLano Scientific, San Carlos, California, USA; http://www.pymol.org). another RNase A family member; but without further studies, I cannot be certain which

RNase A family member enzyme(s) are responsible for cleavage of the host and viral

RNAs.

Angiogenin, RNase 1, RNase 4, and RNase 2 (liver specific endonuclease / eosinophilic derived neurotoxin) are highly expressed in liver (Futami et al., 1997), but aside from the tRNA targets of angiogenin, the in vivo targets have not been characterized for any of these riboonucleases. However, angiogenin has been described to inhibit protein synthesis by cleavage of rRNAs in rabbit reticulocyte lysates (St Clair et al., 1987, 1988). 151

Ribonuclease Cleavage Sites in U3 snoRNA

U3 snoRNA is normally found in the nucleolus of cells where it facilitates rRNA

processing and ribosome biogenesis (Beltrame and Tollervey, 1995; Borovjagin and

Gerbi, 1999; Dragon et al., 2002). In uninfected cells, little cleavage was detected in U3

snoRNA (Figure 5.10A, 0 dpi). At 2 dpi, U3 snoRNA cleavage at GU179 increased above

background levels, followed by further increases at 3, 4 and 5 dpi (Figure 5.10A).

Cleavage at other sites within U3 snoRNA were apparent at 4 and 5 dpi (Figure 5.10A:

CU105 and UU122, among other sites). The cleavage sites within U3 snoRNA mapped predominantly to single-stranded pyrimidines (Figure 5.10B).

Ribonuclease Cleavage in Host RNA from HCV-Infected Liver

tiRNAs were abundant in small RNA cDNA libraries generated from HCV and

HBV infected liver samples (Selitsky et al., 2015). To determine whether cyclic

phosphate cDNA sequencing captures these same tRNA fragments detected in liver

tissue from a patient chronically infected with HCV, as well as detect the same rRNA and

U3 snoRNA fragments identified from JFH1-infected Huh7.5.1 cells, I made cDNA

libraries using RNA from a liver section of a patient with chronic HCV1A infection.

The tiRNAs most abundant in liver samples from patients with chronic HCV or

HBV infections included 5ˊ-tiRNA-Gly-[C/G]CC and 5ˊ-tiRNA-Val-[A/C]AC (Selitsky et al.,

2015). Data from cyclic phosphate cDNA libraries from HCV1A-infected liver tissue

revealed anticodon loop cleavage of many tRNAs, but 5ˊ-tiRNA-Val, 5ˊ-tiRNA-Gly, and

5ˊ-tiRNA-Glu were most frequently detected (Figure 5.11A). These data revealed a

preference for specific anticodons: anticodon GCC for tRNA-Gly, was cleaved more

frequently than anticodons CCC or UCC (data not shown); anticodon CUC for tRNA-Glu,

was cleaved more frequently than anticodon CUU (data not shown); and anticodon CAC

152

Figure 5.10. Ribonuclease cleavage sites in U3 snoRNA. A. Frequency and location of cleavage sites in U3 snoRNA. B. Secondary structure of U3 snoRNA highlighting locations of cleavage. Human U3 snoRNA (www-snorna.biotoul.fr) mfold structure re-drawn using R2R (Weinberg and Breaker, 2011; Zuker, 2003).. Blue circles highlight the location of the most prominent cleavage sites: > 0.025% of Total cDNA reads from the 4 dpi sample. for tRNA-Val was more frequently cleaved than UAC and AAC (data not shown).

Cleavage in each of these tRNAs occurred within the anticodon (Figure 5.11B).

Cleavage sites were evident in 28S rRNA from human liver (Figure 5.12A). The pattern of cleavage in 28S rRNA was consistent with those observed in HeLa cells

(Chapter III; Figure 3.4) and in A549 cells (Chapter IV; Figure 4.7A). However, only one cleavage site in 28S rRNA from HCV-infected liver was also identified in HCV- infected Huh7.5.1 cells (UC2526; bold and red, Figure 5.12A). Intriguingly, many 153

Nucleotide

Figure 5.11. Anticodon loop cleavage in tRNAs from HCV-infected liver. After alignment to hg18 tRNAs (gtrnadb.ucsc.edu/Hsapi/hg18-tRNAs.fa), the frequency of reads that mapped between 30 and 40 nts for each tRNA were plotted (A). The frequency, location, and specificity of 5ˊ-tiRNAs commonly detected in liver. The anticodon is underlined (B).

154

A.

B.

C.

D.

Figure 5.12. Ribonuclease cleavage sites in 28S rRNA, 18S rRNA, 5S rRNA, and U3 snoRNA from liver of HCV1A-infected patient. Frequency and location of cleavage sites in 28S rRNA (A), 18S rRNA (B), 5S rRNA (C), and U3 snoRNA (D). Cleavage sites with bold and red labels indicate sites also detected in JFH1-infected Huh7.5.1 cells.

cleavage sites in 18S rRNA from JFH1-infected Huh7.5.1 cells were also detected in

18S rRNA from HCV1A-infected liver (compare Figures 5.7B and 5.12B, bold and red

sites). Sites in18S rRNA from HCV-infected Huh7.5.1 cells shared with HCV-infected liver included GC72, CC73, GU76, UC243, CC324, UU1112, CC1419, CU1555 and CC1767 (Figure

5.12B; Bold and red sites). In JFH1-infected Huh7.5.1 cells, cleavage was detected in 155

5S rRNA at CU71 and UU72 at 4 and 5 dpi; however, these sites were not detected in 5S

rRNA from HCV-infected liver. Consistent with data from HeLa cells and A549 cells, a 2ˊ,

3ˊ-cyclic phosphate was detected at the 3ˊ-end of 5S rRNA from both liver and Huh7.5.1

cells (Figure 5.8B and 5.12C). Cleavage in U3 snoRNA was detected in the HCV-

infected liver sample. The most frequent cleavage site in U3 snoRNA from HCV1A-

infected liver tissue was GU179, which was also the most frequent site detected in JFH1- infected Huh7.5.1 cells. UU122 was another cleavage site detected in both HCV1A- infected liver and JFH1-infected Huh7.5.1 cells, especially at 4 and 5 dpi. UU135 was

frequently cleaved in U3 snoRNA from liver, but was not detected in JFH1-infected

Huh7.5.1 cells. I did not detect any cleavage sites in HCV RNA from the HCV-infected

liver tissue.

Discussion

Hypothermia, heat shock, hypoxia, nutritional deficiency, and oxidative stress

induce an angiogenin-dependent stress response in cells that result in angiogenin

cleavage of tRNA anticodon loops (Fu et al., 2009a; Yamasaki et al., 2009). Certain viral

infections also trigger the production of tiRNAs in an angiogenin-dependent manner

(Wang et al., 2013). RSV-infection of A549 cells resulted in the accumulation of tiRNAs.

In particular, 5ˊ-tiRNA-Glu-CUC was frequently detected in RSV-infected cells compared to uninfected cells (Wang et al., 2013). In another study, tiRNAs, predominantly from tRNA-Gly-[G/C]CC and tRNA-Val-[A/C]AC were highly abundant in liver tissue from

patients with chronic HCV or HBV infection (Selitsky et al., 2015). To determine whether

HCV-infection provoked an angiogenin-stress response, I made 2ˊ, 3ˊ-cyclic phosphate

cDNA libraries from JFH1 (HCV2A)-infected Huh7.5.1 cells, and from HCV1A-infected

human liver tissue. Not only did I detect 5ˊ-tiRNAs in JFH1-infected Huh7.5.1 cells, but I

also discovered cleavage within 28S rRNA, 18S rRNA, 5S rRNA, and U3 snoRNA that 156

became more frequent as the HCV infection progressed in Huh7.5.1 cells. In HCV-

infected liver tissue, I detected 5ˊ-tiRNAs corresponding to tRNA-Gly-GCC, tRNA-Glu-

CUC, and tRNA-Val-CAC, along with cleavage in 18S rRNA and U3 snoRNA.

Potential Function of Anticodon Loop Cleavage in tRNAs

The cleavage evident in the anticodon loop of tRNA-Glu-CUC is characteristic of

angiogenin-dependent stress responses, where 5ˊ- and 3ˊ-tiRNAs accumulate in cells

(Fu et al., 2009a; Selitsky et al., 2015; Wang et al., 2013). However, in other studies

using nutrient deprivation to induce the angiogenin stress response, a broader group of

tiRNAs derived from many different tRNAs were detected (Fu et al., 2009a). In addition to the accumulation of 5ˊ-tiRNA-Glu-CUC, 5ˊ-tiRNA-Cys-GCA, 5ˊ-tiRNA-Gly-GCC, and

5ˊ-tiRNA-Lys-CUU also accumulated in response to RSV infection (Wang et al., 2013).

In contrast to previous studies by other groups, I did not detect a diverse group of 5 ˊ- tiRNAs, rather, I only detected frequent amounts of 5 ˊ- tiRNA-Glu-CUC in cyclic phosphate cDNA libraries from HCV-infected Huh7.5.1 cells. Regardless, nutrient

deprivation (Fu et al., 2009a), RSV-infection (Wang et al., 2013), and HCV-infection may

provoke alternate stress responses that result in different tRNAs targeted for cleavage.

The impact of cell culture conditions on the expression of angiogenin and activation of angiogenin, as well as other RNase A family members, must be investigated to determine whether changes in cell confluency or medium conditions activate an angiogenin-dependent stress response independent of viral infection.

The functional impact of tRNA cleavage is uncertain; however, tiRNA fragments produced by angiogenin have been found to induce the formation of stress granules

(Emara et al., 2010), and certain 5ˊ-tiRNA fragments containing a 5ˊ-oligoguanine motif

(5ˊ-tiRNA-Ala and 5ˊ-tiRNA-Cys) inhibit translation of capped and uncapped mRNA through displacement of eIF4G/A (Ivanov et al., 2011, 2014). Interestingly, IRES 157 containing transcripts are resistant to the inhibitory effects of the 5ˊ- tiRNA-Ala and 5ˊ- tiRNA-Cys fragments (Ivanov et al., 2011). Furthermore, IRESes are frequently detected in mRNAs encoding pro-survival factors (Li and Hu, 2012), suggesting that the angiogenin stress response might remodel the host translation network to favor expression of host mRNAs encoding pro-survival factors to promote survival during cell stress. In another mechanism to promote survival during cell stress, tiRNAs form complexes with cytochrome c released from the mitochondrial membrane during hyperosmotic stress to prevent cytochrome c-mediated apoptotic cell death (Saikia et al.,

2014). These findings strongly suggest that tiRNAs generated by angiogenin during cellular stress function to promote cell survival. However, different tiRNAs may be generated in response to different types of cellular stress, and these different tiRNAs may have variable functions. In JFH1-infected Huh7.5.1 cells, I do not detect anticodon loop cleavage in tRNAs containing oligoguanine motifs associated with inhibition of capped and uncapped mRNA translation, but I did increasingly detect 5ˊ- tiRNA-Glu-

CUC during the course of the HCV infection. Whether 5 ˊ-tiRNA-Glu-CUC has the capacity to inhibit apoptosis, inhibit protein synthesis, and enhance HCV replication is yet to be determined. A synthetic 5ˊ-tiRNA-Glu-CUC could be used to treat cells prior to

HCV infection to determine whether this tRNA fragment enhances HCV infection. The effects on protein synthesis and induction of apoptosis could also be assessed by in the context of cells treated with synthetic 5ˊ-tiRNA-Glu-CUC.

Although I have yet to assess the expression of angiogenin in the uninfected and

JFH1-infected Huh7.5.1 cells, another study indicated that angiogenin expression increased six-fold in J6/JFH1-infected Huh7 cells when compared to uninfected Huh7 cells (Presser et al., 2011). In addition, cleavage of tRNA-Glu-CUC within its anticodon loop is consistent with what others have reported and occurs predominantly after single- stranded cytidines, consistent with the specificity of angiogenin (Curran et al., 1993; 158

Rybak and Vallee, 1988; Shapiro et al., 1986b). While I am fairly confident that the

anticodon loop cleavage detected in tRNA-Glu-CUC is due to an angiogenin-dependent

stress response in HCV-infected Huh7.5.1 cells based on the detection of tiRNAs in liver

samples from patients chronically infected with HCV (Selitsky et al., 2015), these results

need further validation. siRNA-directed knockdown of angiogenin in Huh7.5.1 cells will

provide further validation to the observations I’ve made in JFH1-infected Huh7.5.1 cells.

If HCV-infection of Huh7.5.1 cells provokes cellular stress that activates angiogenin,

tiRNAs should be detected in control siRNA treated cells, whereas a significant reduction

in tiRNAs accumulation should be observed in Huh7.5.1 cells treated with angiogenin- specific siRNA. These experiments would also provide insight into whether the anti- apoptotic actions of angiogenin and the tiRNAs it generates enhance HCV replication.

Cleavage of Viral RNA, rRNAs, and U3 snoRNA

Despite the anti-apoptotic and pro-survival functions reported for tiRNAs (Ivanov

et al., 2011, 2014; Saikia et al., 2014), JFH1-infection of Huh7.5.1 cells ultimately leads

to death of infected cells. Perhaps some of the other cellular and viral RNA targets or

angiogenin and likely other RNase A-family enzymes counter the anti-apoptotic effects

of angiogenin-induced tRNA fragments.

As HCV RNA accumulated in JFH1-infected Huh7.5.1 cells, I detected increased

amounts of cleavage in HCV RNA (Figure 5.1). The unique permissivity of Huh7.5.1

cells to HCV infection is attributed to specific defects in innate immunity. Huh7 cells do

not express the endosomal dsRNA RNA sensor, TLR3 (Li et al., 2005a). Huh7.5 cells do

not express robust levels of RNase L (Kwon et al., 2013) and contain an inactivating

point mutation in the dsRNA sensor RIG-I, rendering it defective for induction of type I

IFN (Sumpter et al., 2005). Cleavage of HCV RNA was typically after pyrimidines,

consistent with the specificity of RNase A family enzymes (Table 1.1, Sorrentino, 2010). 159

I also detected angiogenin and/or other RNase A family-like cleavage in rRNAs

(28S, 18S, and 5S; Figures 5.6, 5.7, 5.8 and 5.9). Angiogenin is reported to inhibit translation through cleavage of 40S ribosomes when added to rabbit reticulocyte lysates

(St Clair et al., 1987, 1988). On that same note, I detected RNase L-dependent cleavage of 18S rRNA in PV-infected HeLa cells overexpressing wild-type RNase L (Chapter III,

Figure 3.5; (Cooper et al., 2014)), and in A549 cells infected with IAV unable to express

NS1 protein (which inhibits the activation of the OAS/RNase L pathway) (Chapter IV,

Figure 4.7; (Cooper et al., 2015)). Unlike RNase L-induced 18S rRNA cleavage in HeLa and A549 cells, I detected increasing amounts of riboonuclease cleavage in both 18S and 28S rRNA during the course of HCV infections (Figure 5.6 and 5.7). The enzyme/enzymes responsible for cleavage of 18S and 28S rRNAs exhibited specificity for cytidines over uridines, consistent with angiogenin or RNase 1 (Table 1.1; (Curran et al., 1993; Sorrentino et al., 1988)), while the cleavage sites in 5S rRNA were both 3ˊ of uridines, consistent with the specificity of RNase 4 or RNase 2 (Hofsteenge et al., 1998;

Sorrentino, 2010). The functional impact cleavage detected in 28S, 18S, and 5S rRNAs may be to inhibit protein synthesis in JFH1-infected Huh7.5.1 cells, provoking apoptosis of infected cells.

Cleavage in U3 snoRNA also increased in frequency from 2 to 5 dpi (Figure

5.10). The sites frequently cleaved in U3 snoRNA terminate with a uridine, likely indicative of an RNase A enzyme other than angiogenin (Curran et al., 1993; Hofsteenge et al., 1998; Sorrentino et al., 1988). U3 snoRNA is normally found in the nucleolus of cells where it facilitates rRNA processing and ribosome biogenesis (Beltrame and

Tollervey, 1995; Borovjagin and Gerbi, 1999; Dragon et al., 2002). I detected RNase L- dependent cleavage of U3 snoRNA in IAV-infected A549 cells when NS1 was debilitated

(Chapter IV, Figure 4.10; (Cooper et al., 2015)). Defects in 18S or 28S processing caused by siRNA knockdown of proteins important for 18S and 28S rRNA processing 160 result in induction of the ribotoxic stress response and stabilization of p53 (Hölzel et al.,

2010). Ribotoxic stress mediates the release of ribosomal proteins into the nucleoplasm, where they can inhibit MDM2 complexes that regulate p53 stability to result in cell cycle arrest and/or apoptosis (Kim et al., 2014; Yuan et al., 2005). Because U3 snoRNA is required for 18S rRNA processing (Beltrame and Tollervey, 1995), I would expect cleavage of U3 snoRNA to induce the ribotoxic stress response to promote apoptosis.

The cleavage detected in rRNA and U3 snoRNA may serve to counteract the potential pro-survival effects of angiogenin-induced tRNA fragments to eventually lead to apoptotic cell death of JFH1-infected Huh7.5.1 cells.

HCV-Infected Liver Tissue

HCV-infections frequently progress to chronic infections that can lead to liver cancer (Rosen, 2011). To study RNA cleavage in HCV-infected patient liver, I generated cyclic phosphate cDNA libraries from a liver sample of an HCV1A-infected patient provided by the Rosen Lab. Although a larger sample size is needed along with controls from uninfected liver samples, the RNA cleavage detected from the single HCV-infected liver provided interesting data consistent with recent reports about tRNA fragments discovered by sequencing of small RNA libraries from HCV and HBV-infected liver samples (Selitsky et al., 2015). 5ˊ-tiRNA frequently detected in HCV and HBV-infected liver samples included 5ˊ-tiRNA-Gly-[G/C]CC and 5ˊ-tiRNA-Val-[A/C]AC. Cyclic phosphate cDNA libraries from the HCV1A-infected liver sample revealed frequent detection of 5ˊ-tiRNA-Gly-GCC, 5ˊ-tiRNA-Val-CAC, and 5ˊ-tiRNA-Gly-CUC (Figure

5.11). The function of these 5ˊ-tiRNAs in HCV infection and their potential roles in chronic HCV infections are not yet understood. tRNA fragments derived from tRNA-Glu-

CUC potentially stimulate HCV-replication (Wang et al., 2013). Anti-apoptotic functions of tiRNAs may prevent apoptotic cell death of HCV-infected cells, potentiating continued 161

inflammation that contributes to fibrosis and the eventual development of liver cancer in

some patients with chronic HCV-infections (Ivanov et al., 2011; Saikia et al., 2014).

Liver cancers expressing high levels of angiogenin are typically associated with

poor or moderate histopathology grades, and high serum angiogenin levels were

correlated with poor 5-year prognoses (Hisai et al., 2003). A recent study investigating

the relationship between hepatocellular carcinoma and hepatic stellate cells indicated

that angiogenin, secreted from HepG2 cells, was taken up by hepatic stellate cells (LX2

cells), transforming them from a quiescent state into a myofibroblast-type cell involved in

remodeling and fibrosis (Bárcena et al., 2015). Whether the angiogenin secreted from

HepG2 cells is taken up by LX2 cells to mediate tRNA cleavage has not been studied,

but may contribute to the microenvironmental conditions that influence the development

of liver cancer.

In addition to tRNA anticodon loop cleavage, I also detected stress-induced

cleavage in 18S rRNA and U3 snoRNA. U3 snoRNA from HCV1A-infected liver

contained cleavage sites that I identified from both JFH1-infected Huh7.5.1 cells, and

from IAVΔNS1-infected A549 cells where RNase L was activated. Cleavage at GU179 was the most prominent cleavage site in U3 snoRNA from JFH1-infected Huh7.5.1 cells

(Figure 5.10), whereas UU122 (Figure 5.10 and 4.10) was found in U3 snoRNA from

both JFH1-infected Huh7.5.1 cells and IAVΔNS1-infected A549 cells (Figure 5.10 and

4.10). Cleavage in U3 snoRNA at UU135 was found only in the IAVΔNS1-A549 cells

expressing control shRNA (Figure 4.10). While Huh7.5.1 cells express low levels of

RNase L (Kwon et al., 2013), liver cells express relatively large amounts of RNase L and

OAS (Malathi et al., 2010; Zhou et al., 2005). Because HCV RNA was undetectable

cyclic phosphate libraries from HCV-infected liver, questions remain as to whether

RNase L is activated in HCV-infected hepatocytes to produce immunostimulatory RIG-I 162 ligands (Malathi et al., 2010), and whether 18S rRNA cleavage (Cooper et al., 2014,

2015) and U3 snoRNA cleavage impact the outcomes of HCV infections.

Summary

In this chapter, I have provided evidence consistent with an angiogenin- dependent stress response during HCV infection of Huh7.5.1 cells. However, the specificity of the detected cleavage sites suggest that other RNase A-family members in addition to angiogenin were likely involved in RNA degradation during HCV infection.

Cleavage sites were detected in HCV RNA, rRNAs, U3 snoRNA and tRNA-Glu-CUC from HCV-infected Huh7.5.1 cells. Because angiogenin produces RNA fragments with terminal 2ˊ, 3ˊ-cyclic phosphates, I detected only the 5ˊ-half of tRNA-Glu-CUC. 3ˊ-tRNA halves also have potential anti-apoptotic functions (Ivanov et al., 2011; Saikia et al.,

2014), and a comprehensive analysis of the generation of tiRNAs would include both 5ˊ and 3ˊ-tiRNAs. Another RNA ligase, RtcB, would be able to recognize both 5ˊ and 3ˊ- tiRNAs since RtcB can recognize the 2ˊ, 3ˊ-cyclic phosphate of 5ˊ-tiRNAs and the 3ˊ- hydroxyl on 3ˊ-tiRNAs (Tanaka et al., 2011). Deep sequencing methods that precisely identify the tiRNAs produced by angiogenin cleavage, among other RNA targets of angiogenin and RNase A family member enzymes, will provide insight into how the angiogenin and other RNase A family member enzymes impact HCV-infected patients and treatment outcomes.

163

CHAPTER VI

SUMMARY AND FUTURE DIRECTIONS

The studies I described within this thesis identify the frequency and location of

ribonuclease cleavage sites in cellular and viral RNAs targeted during the course of viral

infections, providing insight into the actions of these ribonucleases during viral infections.

The motivation for pursuing this work was due to the lack of sensitive methods to detect

RNA fragments produced by RNase L from cells. Viral countermeasures to prevent

RNase L activation (Table 1.2) underscore the importance of this enzyme in mediating

the antiviral response. In studies by Malathi et al., RNase L cleaved full-length HCV RNA

to produce an RNA fragment (svRNA3) that activated RIG-I. However, whether or not

this RNA is produced in patients to influence disease progression or treatment outcomes

is not known. The svRNA3 is a perfect example of the importance of sensitive methods

to detect ribonuclease cleavage in RNA. Most methods that have been used to identify

RNA cleavage from cells rely on primer extension or Northern blots. Unfortunately, both

of these methods require a probe specific for an RNA of interest; however, the RNA

substrates for many ribonucleases are uncharacterized, and methods that rely on

sequence specific probes cannot be used to identify these novel targets of

ribonucleases.

To address these issues, I adapted and optimized 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and Illumina sequencing methods that rely on A. th. tRNA ligase, to identify

host and viral RNAs cleaved by ribonucleases in cells (Schutz et al., 2010). Many

ribonucleases, including RNase L, RNase A family members, IRE1α and IRE1β, and

RNase T2, cleave RNA to produce fragments with terminal 2ˊ, 3ˊ-cyclic phosphates

(Table 1.1) that are detected by these methods. These methods are widely applicable

and can be used to map cleavage sites in RNAs in vitro (Chapter II) and in vivo

(Chapters III – V). 164

Using these methods, I made several novel discoveries: 1) RNase L, previously

reported to leave non-specific 3ˊ-phosphoryl groups and 3ˊ-phosphate groups on RNAs

(Malathi et al., 2010; Wreschner et al., 1981b), was revealed to produce RNA fragments

recognized by A. th. tRNA ligase, indicating that cleavage fragments produced by RNase

L terminate with 2ˊ, 3ˊ-cyclic phosphates or 2ˊ-phosphates (Schutz et al., 2010)

(Chapter II, Figure 2.1); 2) Purified RNase L and RNase A targeted similar regions of

PV and HCV RNA, suggesting that RNA structural constraints exposes specific regions

of RNA that can be targeted by ribonucleases (Chapter II, Figure 2.8 and Figure 2.10).

3) While rRNA cleavage has long been an indicator of RNase L activation, the precise

locations of cleavage have not been well characterized. The experiments described

within this thesis demonstrate that when activated during PV or IAV infections, RNase L

consistently targets two prominent sites in 18S rRNA, UU541 and UU743 (Chapter III,

Figure 3.5; Figure IV, Figure 4.7); 4) RNase L is activated to cleave IAV RNA when

NS1 protein is disabled, contributing to RNase L-mediated restriction of IAV when NS1

protein is disabled. (Chapter IV, Figures 4.1, Figure 4.6, Table 4.1); 5) tiRNAs derived

from anticodon loop cleavage of tRNA-Glu-CUC, and RNA cleavage fragments from U3 snoRNA and rRNAs accumulated in HCV-infected Huh7.5.1 cells, consistent with the activity of angiogenin and other RNase A family members (Chapter V). 6) 5S rRNA was consistently detected with a terminal 2ˊ, 3ˊ-cyclic phosphate (Chapter III, Figure 3.7;

Chapter V, Figure 5.8).

Although the cleavage sites detected in cellular and viral RNAs during viral infections provide information about the action of RNase L, angiogenin, and others, questions remain about how these cleavage events impact the cell’s response to viral infections. For example, how stable are 2ˊ, 3ˊ-cyclic phosphates? Do cleavage sites in rRNA impact protein synthesis? Does rRNA cleavage serve as the signal to activate 165

apoptosis? What RNA fragments generated by RNase L cleavage serve as ligands for

RIG-I and MDA5 (Malathi et al., 2007)? These questions and others are addressed

herein.

Characteristics of RNA Fragments Produced by RNase L Cleavage

While the experiments described within this thesis demonstrate that RNase L

cleaves RNA through transphosphorylation, resulting in fragments with terminal 5ˊ-

hydroxyl and 2ˊ, 3ˊ-cyclic phosphates/2ˊ-phosphates detectable using A. th. tRNA

ligase, these experiments do not address the stability of the 2ˊ, 3ˊ-cyclic phosphate, or whether RNase L hydrolyzes these 2ˊ, 3ˊ-cyclic phosphates into 2ˊ- or 3ˊ- phosphates at some rate. A. th. tRNA ligase does not differentiate between a 2ˊ, 3ˊ-cyclic phosphate and a 2ˊ-phosphate, but does not recognize RNA with terminal 3ˊ-phosphates for ligation, consequently, a subset of RNAs are potentially missed by 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing methods.

tRNA ligase opens 2ˊ, 3ˊ-cyclic phosphates to 2ˊ-phosphates prior to ligation, thus, RNAs with terminal 2ˊ, 3ˊ-cyclic phosphate and 2ˊ-phosphates are substrates for tRNA ligases (Abelson et al., 1998). The enzymatic mechanisms of RNase L have not been studied as extensively as those of bovine pancreatic RNase A. For RNase A cleavage, step one involves transphosphorylation, producing the 2ˊ, 3ˊ-cyclic phosphate” intermediate, and step 2 involves resolving the 2ˊ, 3ˊ-cyclic phosphate into a 2ˊ- or 3ˊ- phosphate; however several studies indicated there was a lag in the formation of 2ˊ- or

3ˊ- monophosphates (Cuchillo et al., 1993). Whether the second step occurred sequentially after the first step to produce the 2ˊ or 3ˊ-monophosphate, or whether the

2ˊ, 3ˊ-cyclic phosphate was the product, was assessed through studies of RNA product formation during RNase A cleavage using 31P NMR to demonstrate that RNase A

cleavage produced RNA fragments with 2ˊ, 3ˊ-cyclic phosphates. 2ˊ- or 3ˊ- 166

monophosphates only formed after nearly all of the RNA substrate had been cleaved to

fragments with 2ˊ, 3ˊ-cyclic phosphates, and their formation was so slow that the authors concluded that 2ˊ, 3ˊ-cyclic phosphates were the products of RNase A cleavage rather than the intermediate (Thompson et al., 1994). Experiments of this nature could easily be completed to study the kinetics of product formation by RNase L to determine whether RNase L, like RNase A, forms stable 2ˊ, 3ˊ-cyclic phosphates on the ends of

RNA fragments. If 2ˊ, 3ˊ-cyclic phosphates are rapidly hydrolyzed to 2ˊ- or 3ˊ- monophosphates, enzymatic treatments with RNA cyclases can convert 3ˊ- monophosphates to 2ˊ, 3ˊ-cyclic phosphates to increase the population of RNA fragments susceptible to ligation by A. th. tRNA ligase.

Impact of RNA Cleavage by RNase L

Cleavage of Viral RNA in Cells

Does RNase L-dependent cleavage in viral RNAs impact viral replication and

gene expression? The answer to this question likely depends on the virus and the cell

type used for experiments. For example, reovirus infections rely on both RNase L and

PKR to mediate host translational shut-off, and the virus replicates better in the presence

of RNase L (Smith et al., 2005). Despite the detection of a few RNase L-dependent

cleavage sites in PV RNA from HeLa cells overexpressing WT RNase L, the replication

kinetics of PV in these cells was largely unaffected by RNase L activation and rRNA

fragmentation (Figure 3.1B). PV, a virus notoriously resistant to the antiviral effects of

IFN, may not be susceptible to RNase L, and like reovirus, may co-opt the functions of

RNase L for its own benefit. However, HeLa cells may not be the best infection model for

PV, since they are highly susceptible to PV infection. Perhaps experiments making use

of a less permissive cell type with type I IFN-pretreatment would lead to a more robust

RNase L-dependent phenotype during PV infections. WT PV also encodes a competitive 167

inhibitor of RNase L (ciRNA) within its open reading frame, and in molar excess the

ciRNA inhibits RNase L (Han et al., 2007; Keel et al., 2012; Townsend et al., 2008a,

2008b). The ciRNA may have inhibited RNase L activity during the course of PV

infection in HeLa cells, delaying RNase L activation (Townsend et al., 2008b). In future

studies, the use of PV encoding a mutant ciRNA unable to inhibit RNase L coupled with

a less permissive cell type may help reveal more RNase L targets in the context of PV

infections, and might provide insight into the importance of the ciRNA during PV

infections.

Studies on the antiviral effects of RNase L during EMCV virus infection indicated

that RNase L exerted its antiviral effects prior to rRNA fragmentation (Li et al., 1998). In the experiments described in Chapter III, RNase L-dependent cleavage sites in PV RNA were observed at 6 and 8 hpa, coincident with rRNA fragmentation (Figure 3.1B and

Figure 3.3). To focus on RNase L-dependent cleavage of PV RNA during infections, future studies using 2ˊ, 3ˊ-cyclic phosphate cDNA library methods should include depletion of rRNAs to increase sensitivity for viral RNAs. Another manner of enhancing the detection of viral fragments might rely on inhibition of XRN1 or the SKIV2L RNA exosome (Eckard et al., 2014; Girardi et al., 2013).

PV, a positive-strand RNA virus, replicates within complexes in cytoplasm,

sequestering its dsRNA replication intermediates. In contrast to PV, IAV is a negative-

strand virus that replicates in the nucleus without a protective replication complex. Like

PV, IAV has evolved strategies to evade RNase L through expression of its NS1 protein.

NS1 is predicted to sequester dsRNA to prevent the activation of OAS, thus preventing

RNase L activation (Min and Krug, 2006). In IAVΔNS1-infected A549 cells, RNase L-

dependent cleavage of the viral RNA segments impact the magnitudes of viral

replication (Figure 4.1C). When RNase L expression was reduced through RNase L-

specific shRNA, IAVΔNS1 replicated to similar magnitudes as WT IAV, and significantly 168

less viral RNA cleavage was detected. Certain viral RNA segments were susceptible to

RNase L-dependent cleavage, and included the genomic strands of PB2, PB1, and PA,

and the antigenomic/mRNAs of M, NP, and PB1. Why certain segments were targeted

while others were largely untouched remains unanswered. However, abundant cellular

RNAs, like rRNAs, are frequently detected in cDNA libraries. Perhaps the viral segments

frequently targeted by RNase L were more abundant in the A549 cells at that point in

time. Monitoring the synthesis of each viral species (-) strand genomes, (+) strand

antigenomes, and mRNAs would provide insight into the abundance of each viral RNA,

and whether PB2, PB1, and PA genomic strands, and M, NP, and PB1

antigenomic/mRNAs are more abundant than others. Regardless, it is clear that RNase

L cleaves viral RNAs to limit IAV replication; however it is important to note that NS1

inhibits other pathways including PKR and RIG-I (Gack et al., 2009; Min et al., 2007).

Perhaps studies using WT IAV, where RNase L is activated by transfection of 2-5A to bypass NS1-mediated inhibition of OAS, will reveal whether RNase L cleaves viral RNA in the context of WT infections.

Inhibition of Protein Synthesis

rRNA cleavage is considered the indicator of RNase L activation, yet the

significance of rRNA cleavage has yet to be determined (Silverman et al., 1982, 1983;

Wreschner et al., 1981a). Does RNase L-induced cleavage of 18S rRNA at UU541 and

UU753 impact protein synthesis? Would the highly structured and complex ribosome even

be affected by cleavage at these sites? Inhibition of protein synthesis was observed in

the initial experiments that characterized the effects of dsRNA on cell-free extracts from

IFN-treated HeLa cells (Roberts et al., 1976). Studies on the effects of 2-5A on protein

synthesis of rabbit reticulocyte lysates indicated that in the presence of 2-5A, polysomes

disaggregated into monomeric 80S ribosomes, and individual 40 and 60S subunits. 2-5A 169

did not inhibit the association of initiator met-tRNA with 40S subunits, yet these initiation

complexes were not incorporated into active polysomes. The authors concluded that 2-

5A likely influenced 40S ribosomes, but the inhibition of protein synthesis in 2-5A-treated

lysates was most likely due to its induction of a nuclease to degrade mRNA substrates

(Clemens and Williams, 1978).

There is evidence to suggest that cleavage of rRNA inhibits protein synthesis.

The effects of rRNA cleavage on protein synthesis were examined to some degree in

experiments using rabbit reticulocyte lysates and Staphylococcus aureus nuclease

(micrococcal nuclease), a Ca2+-dependent single-stranded RNA-targeting

endoribonuclease (Holmberg and Nygård, 1997). Micrococcal nuclease cleavage sites in

rRNAs were identified using primer extension of ribosomes purified from unfractionated

lysates, isolated subunits, and polysomes. Primer extension revealed that micrococcal

nuclease cleaved both 18S and 28S rRNA; however, in 18S rRNA, one of the most

frequently targeted sites by micrococcal nuclease was within a region of helix16, also

targeted by RNase L (Chapters III, Figures 3.5 and 3.8; Chapter IV, Figure 4.7).

Micrococcal nuclease targeted the single-stranded loop of helix 16 within 540UUU542.

RNase L activation in virally-infected cells resulted in cleavage after UU541. In 28S rRNA micrococcal nuclease frequently cleaved the phosphodiester bonds within 1562GAAA1565

in helix34, among several other cleavage sites. As greater amounts of micrococcal

nuclease were added to rabbit reticulocyte lysates, new sites emerged, yet the decoding

channel of 18S, and the peptidyl-transferase region of 28S rRNA, remained uncleaved.

These same regions were uncleaved in cells where RNase L was activated by viral

infections (Chapter III, Figure 3.8; Chapter IV, Figure 4.7). Treatment of rabbit

reticulolcyte lysates with increasing concentrations of micrococcal nuclease resulted in a

reduction of 14C-labeled leucine incorporation into the newly synthesized proteins, indicating that increasing amounts of rRNA cleavage was related to reduction in protein 170

synthesis. Analysis of 80S ribosomes from rabbit reticulocyte lysates programmed with

globin mRNA revealed that micrococcal nuclease-cleaved ribosomes were components

of polysomes, suggesting that cleavage of rRNA does not prevent these damaged

ribosomes from participation in actively translating polysomes. The authors suggested

that the reduction of protein synthesis by micrococcal nuclease cleavage of rRNA may

influence the flexibility of these ribosomes where translation efficiency is impacted

(Holmberg and Nygård, 1997). However, these experiments do not address what fraction of polysomes are composed of 80S ribosomes with intact rRNA versus cleaved rRNA. Polysomes containing damaged 80S ribosomes may act in an inhibitory manner, potentially stalling on mRNA, inhibiting elongation of 80S ribosomes with intact rRNA.

The experiments by Holmberg and Nygard highlight the importance of the

location of nuclease cleavage in rRNA. Micrococcal nuclease predominantly targeted

rRNA in exterior locations on the ribosome, rather than the buried core regions, or

regions involved in catalysis. The cleavage sites in rRNA detected in 2ˊ, 3ˊ-cyclic phosphate cDNA libraries were also located on the exterior of the ribosome, where they might impact ribosome function and protein synthesis in a similar manner as the micrococcal nuclease-cleaved ribosomes (Holmberg and Nygård, 1997). Experiments

like those used to assess the translational capacity of micrococcal-nuclease treated

rabbit reticulocyte lysates could be used to assess the translational capacity of 2-5A

treated HeLa cell lysates. These types of experiments could be used to determine

whether RNase L –dependent rRNA cleavage inhibits protein synthesis and whether

polysomes contain damaged 80S ribosomes. Capped, uncapped, and IRES-containing

transcripts can be programmed into cell free reactions to determine whether RNase L

cleavage of rRNA impacts the expression of one type of RNA over another.

171

Induction of Apoptosis

While cleavage of rRNA likely impacts protein synthesis, it may also trigger

apoptotic pathways. RNase L activation is important for apoptosis, and it has been

proposed that RNase L cleavage of rRNA serves as the signal to activate JNK signaling

that leads to the eventual release of cytochrome c from the mitochondria (Li et al., 2004;

Rusch et al., 2000). Treatment of cells with poly(I:C) to induce RNase L-mediated rRNA

cleavage leads to activation of JNKs (Iordanov et al., 2000; Li et al., 2004). Damage to

28S rRNA by ribotoxins ricin or sarcin induce activation of JNK (Iordanov et al., 1997),

and inhibition of protein synthesis also serves as a trigger for apoptosis (Iordanov et al.,

2000). Thus, it is reasonable that RNase L-induced cleavage of rRNA might be the

signal to provoke apoptotic pathways. However, the most frequently detected RNase L-

induced cleavage site in cellular RNAs included UU743 in 18S rRNA (Chapter III, Figure

3.5; Chapter IV, Figure 4.7). This site is located in a single-stranded region of a stem- loop structure near the binding site of rpS7. rpS7 mediates p53 stabilization by inhibiting

MDM2-directed ubiquitin ligation and degradation of p53. Could the cleavage at UU743

prevent binding of rpS7, or liberate rpS7 from 40S ribosomes, resulting in p53-mediated

cell cycle arrest/apoptosis? In this case, ribosomes, purified using sucrose density

gradients of 2-5A transfected cell lysates could be assayed for the presence of rpS7.

Decrease of the association of rpS7 with ribosomes should be correlated to increase of

rpS7:MDM2 complexes to coordinate stability and activation of p53-mediated anti-

proliferative functions.

Ligands for RIG-I

RNase L degradation of cellular RNA purified from cells produces RNA

fragments with 2ˊ, 3ˊ-cyclic phosphates. Transfection of these RNase L-cleaved cellular

RNAs provokes robust RIG-I activation that leads to IFNβ production (Malathi et al., 172

2007). Ribosomal RNA is the most abundant RNA in cells, and therefore might make up

a large component of the cellular RNA fragments with RIG-I immunostimulatory potential

when transfected into cells. Does RNase L-induced cleavage of 18S rRNA at UU541 and

UU743 produce RIG-I ligands? The svRNA3, an RNA fragment liberated by RNase L

cleavage, has a 5ˊ-hydroxyl and 3ˊ-phosphate/2ˊ,3ˊ-cyclic phosphate, and the ability of this RNA to activate RIG-I is dependent on its 3ˊ-phosphate group (Malathi et al., 2010).

It is important to note that the in vitro synthesized svRNA3 with a terminal 3ˊ-phosphate was used to test the ability of the svRNA3 to activate RIG-I. The data discussed in this thesis would suggest that RIG-I PAMP RNAs generated by RIG-I would have a terminal

2ˊ,3ˊ-cyclic phosphate; however, it is entirely possible that the termini of RNA fragments produced by RNase L are further processed by cellular enzymes to have terminal modifications that important in RIG-I activation. Cleavage at UU743 in 18S rRNA is

located in a single stranded region at the base of a stem loop structure and cleavage at

UU541 is located at the apical region of helix 16, another stem-loop structure. Would the

liberation of this 202-nt long RNA fragment activate RIG-I? Transfection of this RNA

fragment into RIG-I-/-, MDA5-/-, and IPS1/MAVS-/- MEFs with subsequent measurement of

IFNβ expression would determine whether this rRNA fragment liberated by RNase L has immunostimulatory potential.

Terminal 2ˊ, 3ˊ-Cyclic Phosphates on 5S rRNA

Among the discoveries made from 2ˊ, 3ˊ-cyclic phosphate cDNA libraries was the unexpected 2ˊ, 3ˊ-cyclic phosphate at the end of 5S rRNA. Mature 5S in human KB cells is reported as a 121-nt or 120-nt long RNA that terminates with a 3ˊ-hydroxyl on either

U120 or U121 (Forget et al., 1968). The 2ˊ, 3ˊ-cyclic phosphate on 5S rRNA detected in cyclic phosphate cDNA libraries from HeLa, A549, and Huh7.5.1 cells was detected at 173

U120. 3ˊ-phosphate groups have been reported on 5S rRNA isolated from Euglena 80S ribosomes, but the significance of the terminal 3ˊ-phosphate on Euglena 5S rRNA is unknown.

An interesting comparison can be made between U6 snRNA and 5S rRNA. Both

U6 snRNA and 5S rRNA are pol III transcripts; however, U6 snRNA is post- transcriptionally modified by both a U6 snRNA-specific TUTase and a 3ˊ→5ˊ exonuclease, hUSB1, that trims back the poly(U) tail to produce the terminal 2ˊ, 3ˊ-cyclic phosphate detected in 2ˊ, 3ˊ-cyclic phosphate cDNA libraries (Lund and Dahlberg, 1992;

Mroczek et al., 2012; Shchepachev et al., 2012; Trippe et al., 1998). In HeLa cell extracts incubated with αP32-UTP, 5S rRNA was labeled by a cellular TUTase (Trippe et

al., 1998). This “non-specific” TUTase was distinct from the highly specific TUTase that recognized U6 snRNA as its substrate RNA. Whether hUSB1 or a similar 3ˊ→5ˊ exoribonuclease modifies 5S rRNA in the same manner as U6 snRNA has yet to be assessed, but experiments using siRNA-mediated knockdown of hUsb1, could be used

to determine whether hUsb1 also influences the maturation of 5S rRNA to produce

terminal 2ˊ, 3ˊ-cyclic phosphates.

Because a terminal 2ˊ, 3ˊ-cyclic phosphate on 5S rRNA was a novel discovery from 2ˊ, 3ˊ-cyclic phosphate cDNA libraries, confirming its presence using other methods is important. The 2ˊ, 3ˊ-cyclic phosphate found at the end of U6 snRNA was originally described through two-dimensional thin layer chromatography methods and RNase T1 fingerprint methods. RNase T1 fingerprint analysis of U6 snRNA treated with acid, calf- intestinal phosphatase (CIP), both acid and CIP, or sodium periodate (Lund and

Dahlberg, 1992). 2ˊ, 3ˊ-cyclic phosphates are resistant to removal by CIP, but pre- treatment with acid opens the 2ˊ, 3ˊ-cyclic phosphate to 2ˊ- or 3ˊ-phosphates that can be removed by CIP. Sodium periodate and treatment of RNA with oxidation removes unmodified nucleotides with 2ˊ and 3ˊ-hydroxyl groups. Tracking the movement of the 174

RNase T1 digest product corresponding to the 3ˊ end of U6 snRNA in response to the

mentioned treatments confirmed the 2ˊ, 3ˊ-cyclic phosphate at the end of U6 snRNA

(Lund and Dahlberg, 1992). While these experiments can be used to confirm the

terminal 2ˊ, 3ˊ-cyclic phosphate on 5S rRNA, simpler methods that rely on high resolution gel electrophoresis to distinguish between the different species of 5S rRNA after the same chemical treatments used by Lund and Dahlberg could also be designed

to study the terminal modifications on 5S rRNA.

5S rRNA is a component of the large ribosomal subunit, but it is also reported to be complexed with the ribosomal L5 protein independent of 60S ribosomes, and with

MDMX, a regulator of p53 stability (Li and Gu, 2011; Steitz et al., 1988). In addition, 5S rRNA has yet-to-be identified functions in mitochondria (Smirnov et al., 2008). While 5S

isolated from ribosomes is reported to terminate with a 3ˊ-hydroxyl (Anger et al., 2013;

Forget and Weissman, 1968), 5S rRNA may be modified by a nuclease in the context of

MDMX complexes, or within mitochondria. In addition, the 3ˊ-end of 5S rRNA is surface exposed on the 80S ribosome. Perhaps a portion of 80S ribosomes contain a 5S rRNA with terminal 2ˊ, 3ˊ-cyclic phosphates. In U6 snRNA, the 2ˊ, 3ˊ-terminal cyclic phosphate is important for its stability. The stability of synthetic 32P-labeled 5S rRNA with 3ˊ- hydroxyl or 2ˊ, 3ˊ-cyclic phosphate termini could be assessed in HeLa cell lysates. The

2ˊ, 3ˊ-cyclic phosphate at the end of U6 snRNA promotes the binding of the Lsm2-8 complex, important for interactions with U4 snRNA, nuclear retention, and recruitment of recycling factors (Achsel et al., 1999; Licht et al., 2008; Mroczek et al., 2012; Spiller et al., 2007). Does the terminal 2ˊ, 3ˊ-cyclic phosphate detected at the end of 5S rRNA also promote the binding of Lsm proteins for the purpose of nuclear retention? Isolation and analysis of 5S rRNA from cytoplasmic and nuclear fractions might reveal whether a cyclic phosphate and associated Lsm proteins are found on 5S rRNA.

175

Angiogenin Stress Response

In HCV-infected Huh7.5.1 cells, tRNA fragments corresponding to anticodon loop

cleavage of 5ˊ-tRNA-Glu-CUC (5ˊ-tiRNA-Glu-CUC) accumulated in cells and were detected at 4 and 5 dpi. How does 5ˊ-tiRNA-Glu-CUC impact HCV infections? tiRNAs produced by angiogenin have functions in protein inhibition and repression of apoptosis

(Ivanov et al., 2011a; Saikia et al., 2014). 5ˊ-tiRNA-Glu-CUC was found to accumulate in

RSV-infected cells, and this fragment was beneficial for RSV replication (Wang et al.,

2013). While the angiogenin-dependent tRNA stress response has been best characterized using chemical treatments, nutrient starvation, and hypoxia and hypothermia that induce cellular stress, these experiments provide insight into the triggers of the angiogenin stress response during viral infections, especially HCV.

Sodium arsenite treatment of cells produces reactive oxygen species that induce the angiogenin stress response, as demonstrated by the accumulation of tiRNAs (Saikia et al., 2012; Yamasaki et al., 2009). Reactive oxygen species are also produced during

HCV infections. When expressed in cells, the HCV viral core and NS5A proteins induce the formation of reactive oxygen species (Gong et al., 2001; Okuda et al., 2002). In the context of in vivo liver infections by HCV, reactive oxygen species are generated in response to pro-inflammatory cytokines like TNFα and IL-1β released by infiltrating immune cells (Koike and Miyoshi, 2006). Perhaps angiogenin in response to reactive oxygen species produced through HCV infection of Huh7.5.1 cells preferentially targets tRNA-Glu-CUC.

In addition to the accumulation of 5ˊ-tiRNA-Glu-CUC, I also found that many other cellular RNAs were cleaved during the viral infection. These included 28S, 18S, and 5S rRNAs, and U3 snoRNA. Angiogenin is reported to have a preference for cleavage after cytidines (Chapter I, Table 1.1), yet U3 snoRNA and 5S rRNA were clearly targeted by a uridine-preferring enzyme (Chapter V, Figures 5.8 and 5.10). Both 176

RNase 4 and RNase 2, two RNase A family enzymes, have preference for cleavage 3ˊ of uridines over cytidines, and are highly expressed in the liver (Sorrentino et al., 1988;

Futami et al., 1997). Other RNase A family member enzymes, in addition to angiogenin, likely contribute to the RNA cleavage detected in tRNA, rRNAs, and U3 snoRNA from

HCV-infected Huh7.5.1 cells. It would be worth testing the importance of other RNase A family enzymes to determine whether they have roles during viral infection. These studies can be initiated by simply testing for the expression of these ribonucleases during JFH1-infection of Huh7.5.1. cells. Both serum and intracellular fractions should be assayed for the expression of these RNases, since many RNase A family enzymes are secreted from the cell. RNase A members whose expression is increased during HCV infection are likely candidates of the ribonucleases that responsible for the RNA cleavage observed in HCV-infected Huh7.5.1 cells. Angiogenin and related enzymes may have important consequences for HCV infections, and it will be important to determine whether angiogenin is beneficial for HCV by promoting cell-survival, or whether it is inhibitory to HCV infections. These sorts of experiments can make use of siRNA knockdown of angiogenin and other RNase A family members to determine their impact on HCV-infection. Perhaps knocking down RNase 4, a ribonuclease that potentially cleaves the U3 snoRNA, will tip the balance toward cell survival, with JFH1 infections transitioning into a long term persistent infection. Knockdown of angiogenin might tip the balance to a pro-apoptotic state, where the pro-survival functions of angiogenin, like induction of rRNA transcription and inhibition of apoptosis, lead to rapid apoptotic cell death of infected cells (Saikia et al., 2014; Shu et al., 2015; Xu et al., 2003,

2002).

177

The Intersection of RNase L and Angiogenin

Although I did not detect RNA fragments consistent with the angiogenin stress

response in experiments where RNase L was activated by viral infections, it is

reasonable to assume there are instances where RNA regulation by these two

ribonucleases intersects, especially since both of these enzymes are widely expressed

(Futami et al., 1997; Zhou et al., 2005), and both have roles in apoptosis and cancer.

RNase L is generally thought of as pro-apoptotic and the missense mutation in RNASEL associated with susceptibility to prostate cancer (R462Q) yields an enzyme with reduced ability to respond to 2-5A and induce apoptotic cell death of prostate cancer cells (Rusch et al., 2000; Xiang et al., 2003; Zhou et al., 1997). Angiogenin, on the contrary, is associated with inhibition of apoptosis and is highly upregulated in cancer (Ivanov et al.,

2011a; Li and Hu, 2010, 2012; Saikia et al., 2014).

Perhaps in the case of HCV infections, HCV forms dsRNA intermediates that lead to RNase L activation through 2-5A, and forms reactive oxygen species that activate the angiogenin stress response. RNase L cleaves rRNA to inhibit protein synthesis and potentiate apoptotic pathways, while angiogenin cleaves tRNAs to inhibit translation and inhibit apoptosis. The tRNA fragments, or tiRNAs, also induce the formation of stress granules. Both RNase L and angiogenin localize to stress granules

(Onomoto et al., 2012, 2014; Pizzo et al., 2013), where RNase L might cleave 18S rRNA maintain inhibit protein synthesis inhibition, or potentially release the rpS7 to activate p53. In the context of stress granules, angiogenin may cleave rRNA in addition to tRNAs. The rRNA cleavage observed during HCV-infection of Huh7.5.1 is consistent with the specificity of angiogenin (very cytidine-specific). Perhaps angiogenin, like

RNase L cleaves rRNA to inhibit protein synthesis during cellular stress. However, the tiRNAs that are also produced by angiogenin ensure that inhibition of protein synthesis does not result in apoptotic cell death (Saikia et al., 2014). Survival during stress is 178

critical in certain situations and cell types, like motor neurons for example, where

angiogenin is important for protection against excitotoxic cell stress (Skorupa et al.,

2012). However, pro-survival pathways can potentially progress to chronic infections and

cancer, and may serve to explain why angiogenin is upregulated in hepatocellular

carcinoma and other types of cancer (Hisai et al., 2003; Li and Hu, 2010).

Future studies that tease apart the intricacies of these counter-acting pathways might lay the groundwork for new targeted anti-cancer therapies and treatments for chronic infections. Neamine, a drug that prevents angiogenin entry into the nucleus, repressed xenograft growth of human prostate cancer cells in mice, and repressed rRNA transcription to eventually lead to apoptosis of the cancerous cells (Ibaragi et al., 2009).

Susceptibility to prostate cancer is associated with the R462Q mutation in RNase L. This variant of RNase L has reduced endonuclease ability, and a reduced ability to induce apoptosis in prostate cancer cell lines (Xiang et al., 2003). Does angiogenin play a role in prostate cancers associated with RNase L mutations? Perhaps patients with the

R462Q mutation are more susceptible to the pro-survival and proliferative effects of angiogenin when the function of RNase L is debilitated. Angiogenin is most heavily produced by the liver (Weiner et al., 1987), and thus may impact the liver’s susceptibility to HCV infection. Perhaps neamine, in concert with IFN-based therapies, would enhance the effectiveness of IFN-based therapies for chronic HCV infections.

Studies that bridge cell stress and apoptosis might be important in determining whether RNase L and angiogenin counteract each other. One such study might include pre-treatment of W12 HeLa cells, which overexpress WT RNase L, and M25 HeLa cells, which overexpress dominant negative RNase L, with exogenous angiogenin.

Exogenously added angiogenin promotes the formation of tiRNAs (Yamasaki et al.,

2009). Subsequent transfection of 2-5A to activate RNase L would provide insight as to whether tiRNA-mediated protein synthesis inhibition and repression of apoptosis inhibit 179

the anti-proliferative effects of RNase L. Synthetic tiRNAs, like the 5ˊ-tiRNA-Glu-CUC

produced during HCV infections, would also be interesting to study in the context of

RNase L. This tiRNA fragment had a beneficial effect on RSV replication (Wang et al.,

2013). The mechanism behind this viral stimulation is not known. Does this tiRNA

fragment directly benefit the virus, or does it inhibit the function of antiviral factors like

RNase L?

Other than RNase L, other ribonucleases might potentially counteract the anti-

apoptotic functions of angiogenin. In Chapter V, U3 snoRNA was cleaved during the

course of viral infections, very specifically at uridines. While uridine specificity is

characteristic of RNase L activity, there were no indications of RNase L activity in the

JFH1-infected Huh7.5.1 cells. RNase 4 (Table 1.1) is co-expressed with angiogenin, and has a preference for RNA substrates containing uridine. When incubated with human umbilical vein cells, RNase 4 is taken up by these cells and transported to the nucleus

(Li et al., 2013). Perhaps during the angiogenin stress response, RNase 4 is activated in the nucleus to cleave substrates like U3 snoRNA which would likely activate p53- mediated apoptosis in response to ribotoxic stress.

Understanding how these pathways intersect and regulate each other will undoubtedly be important for understanding biology and disease, and studying these pathways using 2ˊ, 3ˊ-cyclic phosphate cDNA synthesis and deep sequencing methods will facilitate progress in our understanding of the complex relationship between ribonucleases, RNA, and disease.

180

REFERENCES

Abelson, J., Trotta, C.R., and Li, H. (1998). tRNA splicing. J. Biol. Chem. 273, 12685– 12688.

Achsel, T., Brahms, H., Kastner, B., Bachi, A., Wilm, M., and Lührmann, R. (1999). A doughnut-shaped heteromer of human Sm-like proteins binds to the 3ˊ-end of U6 snRNA, thereby facilitating U4/U6 duplex formation in vitro. EMBO J. 18, 5789–5802.

Ahlquist, P. (2006). Parallels among positive-strand RNA viruses, reverse-transcribing viruses and double-stranded RNA viruses. Nat. Rev. Microbiol. 4, 371–382.

Al-Ahmadi, W., Al-Haj, L., Al-Mohanna, F.A., Silverman, R.H., and Khabar, K.S. a (2009). RNase L downmodulation of the RNA-binding protein, HuR, and cellular growth. Oncogene 28, 1782–1791.

Ambros, V., Pettersson, R.F., and Baltimore, D. (1978). An enzymatic activity in uninfected cells that cleaves the linkage between poliovirion RNA and the 5ˊ terminal protein. Cell 15, 1439-1446.

Alkemar, G., and Nygård, O. (2003). A possible tertiary rRNA interaction between expansion segments ES3 and ES6 in eukaryotic 40S ribosomal subunits. RNA 9, 20–24.

Alkemar, G., and Nygård, O. (2004). Secondary structure of two regions in expansion segments ES3 and ES6 with the potential of forming a tertiary interaction in eukaryotic 40S ribosomal subunits. RNA 10, 403–411.

Amitsur, M., Levitz, R., and Kaufmann, G. (1987). Bacteriophage T4 anticodon nuclease, polynucleotide kinase and RNA ligase reprocess the host lysine tRNA. EMBO J. 6, 2499–2503.

Andersen, K.L., and Collins, K. (2012). Several RNase T2 enzymes function in induced tRNA and rRNA turnover in the ciliate Tetrahymena. Mol. Biol. Cell 23, 36–44.

Andersen, J.B., Li, X.L., Judge, C.S., Zhou, A., Jha, B.K., Shelby, S., Zhou, L., Silverman, R.H., and Hassel, B. a (2007). Role of 2-5A-dependent RNase-L in senescence and longevity. Oncogene 26, 3081–3088.

Anger, A.M., Armache, J.-P., Berninghausen, O., Habeck, M., Subklewe, M., Wilson, D.N., and Beckmann, R. (2013). Structures of the human and Drosophila 80S ribosome. Nature 497, 80–85.

Ank, N., West, H., Bartholdy, C., Thomsen, A.R., Paludan, S.R., Ank, N., West, H., Bartholdy, C., Eriksson, K., Thomsen, A.R., and Paludan, S.R. (2006). Lambda interferon (IFN- λ), a type III IFN, is induced by viruses and IFNs and displays potent antiviral activity against select virus infections. J. Virol. 80, 4501–4509.

Arranz, R., Coloma, R., Chichon, F.J., Conesa, J.H., Carrascosa, J.L., Valpuesta, J.M., Ortin, J., and Martin-Benito, J. (2012). The Structure of native influenza virion ribonucleoproteins. Science 338, 1634–1638.

181

Arriola, C.S., Brammer, L., Epperson, S., Blanton, L., Kniss, K., Mustaquim, D., Steffens, C., Dhara, R., Leon, M., Perez, A., Chaves, S.S., Katz, J., Wallis, T., Villanueva, J., Xu, X., Elal, A.I.A., Gubareva, L., Cox, N., Finelli, L., Bresee, J., and Jhung, M. (2014). Update: influenza activity - United States, September 29, 2013 - February 8, 2014. MMWR. Morb. Mortal. Wkly. Rep. 63, 148–154.

Bailey, T.L., Boden, M., Buske, F.A., Frith, M., Grant, C.E., Clementi, L., Ren, J., Li, W.W., and Noble, W.S. (2009). MEME Suite: Tools for motif discovery and searching. Nucleic Acids Res. 37, 202–208.

Banerjee, S., An, S., Zhou, A., Silverman, R.H., and Makino, S. (2000). RNase L- independent specific 28S rRNA cleavage in murine coronavirus-infected cells. J. Virol. 74, 8793–8802.

Bárcena, C., Stefanovic, M., Tutusaus, A., Martinez-Nieto, G. A., Martinez, L., García- Ruiz, C., de Mingo, A., Caballeria, J., Fernandez-Checa, J.C., Marí, M., and Morales, A. (2015). Angiogenin secretion from hepatoma cells activates hepatic stellate cells to amplify a self-sustained cycle promoting liver cancer. Sci. Rep. 5, 7916.

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

Barrett, T., Wilhite, S.E., Ledoux, P., Evangelista, C., Kim, I.F., Tomashevsky, M., Marshall, K.A., Phillippy, K.H., Sherman, P.M., Holko, M., Yefanov, A., Lee, H., Zhang, N., Robertson, C.L., Serova, N., Davis, S., and Soboleva, A. (2013). NCBI GEO: archive for functional genomics data sets - Update. Nucleic Acids Res. 41, 991–995.

Bayard, B.A., and Gabrion, J.B. (1993). 2ˊ, 5ˊ-Oligoadenylate-dependent RNAase located in nuclei: biochemical characterization and subcellular distribution of the nuclease in human and murine cells. Biochem J. 295, 155-160.

Becknell, B., Eichler, T.E., Beceiro, S., Li, B., Easterling, R.S., Carpenter, A.R., James, C.L., McHugh, K.M., Hains, D.S., Partida-Sanchez, S., and Spencer, J.D. (2014). Ribonucleases 6 and 7 have antimicrobial function in the human and murine urinary tract. Kidney Int. 87, 151–161.

De Beeck, A.O., Montserret, R., Duvet, S., Cocquerel, L., Cacan, R., Barberot, B., Maire, M. Le, Penin, F., and Dubuisson, J. (2000). The transmembrane domains of hepatitis C virus envelope glycoproteins E1 and E2 play a major role in heterodimerization. J. Biol. Chem. 275, 31428–31437.

Beintema, J.J., Wietzes, P., Weickmann, J.L., and Glitz, D.G. (1984). The amino acid sequence of human pancreatic ribonuclease. Anal. Biochem. 136, 48–64.

Beintema, J.J., Fitch, W.M., and Carsana, A. (1985). The molecular evolution of pancreatic-type ribonuclease. Mol. Biol. Evol. 3, 262–275.

Beltrame, M., and Tollervey, D. (1995). Base pairing between U3 and the pre-ribosomal RNA is required for 18S rRNA synthesis. EMBO J. 14, 4350–4356.

182

Benech, P., Mory, Y., Revel, M., and Chebath, J. (1985). Structure of two forms of the interferon-induced (2ˊ-5ˊ) oligo A synthetase of human cells based on cDNAs and gene sequences. EMBO J. 4, 2249–2256.

Benech, P., Vigneron, M., Peretz, D., Revel, M., and Chebath, J. (1987). Interferon- responsive regulatory elements in the promoter of the human 2ˊ,5ˊ-oligo(A) synthetase gene. Mol. Cell. Biol. 7, 4498–4504.

Ben-Shem, A., Jenner, L., Yusupova, G., and Yusupov, M. (2010). Crystal structure of the eukaryotic ribosome. Science. 330, 1203–1209.

Bisbal, C., Martinand, C., Silhol, M., Lebleu, B., and Salehzada, T. (1995). Cloning and characterization of a RNase L inhibitor. A new component of the interferon-regulated 2- 5A pathway. J. Biol. Chem. 270, 13308–13317.

Bisbal, C., Silhol, M., Laubenthal, H., Kaluza, T., Carnac, G., Milligan, L., Le Roy, F., and Salehzada, T. (2000). The 2ˊ-5ˊ oligoadenylate/RNase L/RNase L inhibitor pathway regulates both MyoD mRNA stability and muscle cell differentiation. Mol. Cell. Biol. 20, 4959–4969.

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

Bork, P., and Sander, C. (1993). A hybrid protein kinase-RNase in an interferon-induced pathway? FEBS Lett. 334, 149–152.

Borovjagin, A.V., and Gerbi, S.A. (1999). U3 small nucleolar RNA is essential for cleavage at sites 1, 2 and 3 in pre-rRNA and determines which rRNA processing pathway is taken in Xenopus oocytes. J. Mol. Biol. 286, 1347–1363.

Bovee, M.L., Marissen, W.E., Zamora, M., and Lloyd, R.E. (1998). The predominant elF4G-specific cleavage activity in poliovirus-infected HeLa cells is distinct from 2A protease. Virology 245, 229–240.

Bratton, S.B., and Salvesen, G.S. (2010). Regulation of the Apaf-1-caspase-9 apoptosome. J. Cell Sci. 123, 3209–3214.

Bruns, A.M., Pollpeter, D., Hadizadeh, N., Myong, S., Marko, J.F., and Horvath, C.M. (2013). ATP hydrolysis enhances RNA recognition and antiviral signal transduction by the innate immune sensor, laboratory of genetics and physiology 2 (LGP2). J. Biol. Chem. 288, 938–946.

Cameron, V., and Uhlenbeck, O.C. (1977). 3ˊ-Phosphatase activity in T4 polynucleotide kinase. Biochemistry 16, 5120–5126.

Campomenosi, P., Salis, S., Lindqvist, C., Mariani, D., Nordström, T., Acquati, F., and Taramelli, R. (2006). Characterization of RNASET2, the first human member of the Rh/T2/S family of glycoproteins. Arch. Biochem. Biophys. 449, 17–26.

183

Carroll, S.S., Chen, E., Viscount, T., Geib, J., Sardana, M.K., Gehman, J., and Kuo, L.C. (1996). Cleavage of oligoribonucleotides by the 2ˊ,5ˊ-oligoadenylate- dependent ribonuclease L. J. Biol. Chem. 271, 4988–4992.

Casey, G., Neville, P.J., Plummer, S.J., Xiang, Y., Krumroy, L.M., Klein, E.A., Catalona, W.J., Nupponen, N., Carpten, J.D., Trent, J.M., Silverman, R.H., and Witte, J.S. (2002). RNASEL Arg462Gln variant is implicated in up to 13% of prostate cancer cases. Nat. Genet. 32, 581–583.

Castelli, J.C., Hassel, B.A., Maran, A., Paranjape, J., Hewitt, J.A., Li, X.L., Hsu, Y.T., Silverman, R.H., and Youle, R.J. (1998). The role of 2ˊ-5ˊ oligoadenylate-activated ribonuclease L in apoptosis. Cell Death Differ. 5, 313–320.

Cervantes, J.L., Weinerman, B., Basole, C., and Salazar, J.C. (2012). TLR8: the forgotten relative revindicated. Cell. Mol. Immunol. 9, 434–438.

Chase, B.I., Zhou, Y., Xiang, Y., Silverman, R.H., and Zhou, A. (2003). Proteasome- mediated degradation of RNase L in response to phorbol-12-myristate-13-acetate (PMA) treatment of mouse L929 Cells. J. Interf. Cytokine Res. 23, 565–573.

Chen, D., Zhang, Z., Li, M., Wang, W., Li, Y., Rayburn, E.R., Hill, D.L., Wang, H., and Zhang, R. (2007). Ribosomal protein S7 as a novel modulator of p53-MDM2 interaction: binding to MDM2, stabilization of p53 protein, and activation of p53 function. Oncogene 26, 5029–5037.

Chen, W., Calvo, P.A., Malide, D., Gibbs, J., Schubert, U., Bacik, I., Basta, S., O’Neill, R., Schickli, J., Palese, P., Henklein, P., Bennink, J.R., and Yewdell, J.W. (2001). A novel influenza A virus mitochondrial protein that induces cell death. Nat. Med. 7, 1306– 1312.

Cho, S., Beintema, J.J., and Zhang, J. (2005). The ribonuclease A superfamily of mammals and birds: Identifying new members and tracing evolutionary histories. Genomics 85, 208–220.

Clemens, M.J., and Williams, B.R. (1978). Inhibition of cell-free protein synthesis by pppA2ˊp5ˊA2ˊp5ˊA: a novel oligonucleotide synthesized by interferon-treated L cell extracts. Cell 13, 565–572.

Cocchi, F., DeVico, A.L., Lu, W., Popovic, M., Latinovic, O., Sajadi, M.M., Redfield, R.R., Lafferty, M.K., Galli, M., Garzino-Demo, A., Gallo, R.C. (2012). Soluble factors from T cells inhibiting X4 strains of HIV are a mixture of chemokines and RNases. Proc. Natl. Acad. Sci. U.S.A. 109, 5411–5416.

Cole, J.L., Carroll, S.S., and Kuo, L.C. (1996). Dimerization of ribonuclease L. Biochemistry 3979–3981.

Colombo, E.A., Elcioglu, N.H., Yucelten, D., Altunay, I., Cetincelik, U., Teti, A., Del Fattore, A., Luciani, M., Sullivan, S.K., Yan, A.C., Volpi, L., and Larizza, L. (2012). Novel C16orf57 mutations in patients with Poikiloderma with Neutropenia: bioinformatic analysis of the protein and predicted effects of all reported mutations. Orphanet J. Rare Dis. 7, 7. 184

Cooper, D.A., Jha, B.K., Silverman, R.H., Hesselberth, J.R., and Barton, D.J. (2014). Ribonuclease L and metal-ion-independent endoribonuclease cleavage sites in host and viral RNAs. Nucleic Acids Res. 42, 5202–5216.

Cooper, D.A., Banerjee, S., Chakrabarti, A., García-Sastre, A., Hesselberth, J.R., Silverman, R.H., and Barton, D.J. (2015). RNase L targets distinct sites in influenza A virus RNAs. J. Virol. 89, 2764–2776.

Cortés, K.C., Zagordi, O., Perlejewski, K., Laskus, T., Maroszek, K., Bukowska-O Ko, I., Pawe Czyk, A., P Oski, R., Berak, H., Horban, A., and Horban, A. (2014). Deep sequencing of hepatitis C virus hypervariable region 1 reveals no correlation between genetic heterogeneity and antiviral treatment outcome. BMC Infect. Dis. 14, 389.

Cowton, V.M., McGivern, D.R., and Fearns, R. (2006). Unravelling the complexities of respiratory syncytial virus RNA synthesis. J. Gen. Virol. 87, 1805–1821.

Crépin, T., Dias, A., Palencia, A., Swale, C., Cusack, S., and Ruigrok, R.W.H. (2010). Mutational and metal binding analysis of the endonuclease domain of the influenza virus polymerase PA subunit. J. Virol. 84, 9096–9104.

Cuchillo, C.M., Parés, X., Guasch, A., Barman, T., Travers, F., and Nogués, M. V (1993). The role of 2ˊ,3ˊ-cyclic phosphodiesters in the bovine pancreatic ribonuclease A catalysed cleavage of RNA: intermediates or products? FEBS Lett. 333, 207–210.

Cuchillo, C.M., Nogués, M.V., and Raines, R.T. (2011). Bovine pancreatic ribonuclease: Fifty years of the first enzymatic reaction mechanism. Biochemistry 50, 7835–7841.

Culver, G.M., McCraith, S.M., Zillmann, M., Kierzek, R., Michaud, N., LaReau, R.D., Turner, D.H., and Phizicky, E.M. (1993). An NAD derivative produced during transfer RNA splicing: ADP-ribose 1˝-2˝ cyclic phosphate. Science 261, 206–208.

Curran, T.P., Shapiro, R., and Riordan, J.F. (1993). Alteration of the enzymatic specificity of human angiogenin by site-directed mutagenesis. Biochemistry 32, 2307– 2313.

Das, U., and Shuman, S. (2013). Mechanism of RNA 2ˊ, 3ˊ-cyclic phosphate end healing by T4 polynucleotide kinase-phosphatase. Nucleic Acids Res. 41, 355–365.

Davis, R.J. (2000). Signal transduction by the JNK group of MAK kinases. Cell 103, 239- 252.

Davis, M., Sagan, S.M., Pezacki, J.P., Evans, D.J., and Simmonds, P. (2008). Bioinformatic and physical characterizations of genome-scale ordered RNA structure in mammalian RNA viruses. J. Virol. 82, 11824–11836.

Demettre, E., Bastide, L., D’Haese, A., De Smet, K., De Meirleir, K., Tiev, K.P., Englebienne, P., and Lebleu, B. (2002). Ribonuclease L proteolysis in peripheral blood mononuclear cells of chronic fatigue syndrome patients. J. Biol. Chem. 277, 35746– 35751.

185

De Prisco, R., Sorrentino, S., Leone, E., and Libonati, M. (1984). A ribonuclease from human seminal plasma active on double-stranded RNA. Biochim. Biophys. Acta 788, 356–363.

Der, S.D., Zhou, A., Williams, B.R., and Silverman, R.H. (1998). Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays. Proc. Natl. Acad. Sci. U.S.A. 95, 15623–15628.

Derenzini, M., Trere, D., Pession, A., Govoni, M., Sirri, V., and Chieco, P. (2000). Nucleolar size indicates the rapidity of cell proliferation in cancer tissues. J. Path. 191, 181-186.

Diebold, S.S., Kaisho, T., Hemmi, H., Akira, S., and Reis e Sousa, C. (2004). Innate antiviral responses by means of TLR7-mediated recognition of single-stranded RNA. Science 303, 1529–1531.

Domachowske, J.B., Dyer, K.D., Bonville, C.A., and Rosenberg, H.F. (1998). Recombinant human eosinophil-derived neurotoxin/RNase 2 functions as an effective antiviral agent against respiratory syncytial virus. J. Infect. Dis. 177, 1458–1464.

Dong, B., and Silverman, R.H. (1995). 2-5A-dependent RNase molecules dimerize during activation by 2-5A*. J. Biol. Chem. 270, 4133-4137.

Dong, B., and Silverman, R.H. (1999). Alternative function of a protein kinase homology domain in 2ˊ, 5ˊ-oligoadenylate dependent RNase L. Nucleic Acids Res. 27, 439–445.

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

Dong, B., Niwa, M., Walter, P., and Silverman, R.H. (2001). Basis for regulated RNA cleavage by functional analysis of RNase L and Ire1p. RNA 7, 361–373.

Donovan, J., Dufner, M., and Korennykh, A. (2013). Structural basis for cytosolic double- stranded RNA surveillance by human oligoadenylate synthetase 1. Proc. Natl. Acad. Sci. U.S.A. 110, 1652–1657.

Donovan, J., Whitney, G., Rath, S., and Korennykh, A. (2015). Structural mechanism of sensing long dsRNA via a noncatalytic domain in human oligoadenylate synthetase 3. Proc. Natl. Acad. Sci. U.S.A. 112, 3949-3954.

Dragon, F., Gallagher, J.E.G., Compagnone-Post, P.A., Mitchell, B.M., Porwancher, K.A., Wehner, K.A., Wormsley, S., Settlage, R.E., Shabanowitz, J., Osheim, Y., Beyer, A.L., Hunt, D.F., and Baserga, S.J. (2002). A large nucleolar U3 ribonucleoprotein required for 18S ribosomal RNA biogenesis. Nature 417, 967–970.

Drahos, J., and Racaniello, V.R. (2009). Cleavage of IPS-1 in cells infected with human rhinovirus. J. Virol. 83, 11581–11587.

Durack, D.T., Sumi, S.M., and Klebanoff, S.J. (1979). Neurotoxicity of human eosinophils. Proc. Natl. Acad. Sci. U.S.A. 76, 1443–1447. 186

Dyer, K.D., and Rosenberg, H.F. (2005). The mouse RNase 4 and RNase 5/ang 1 locus utilizes dual promoters for tissue-specific expression. Nucleic Acids Res. 33, 1077–1086.

Eckard, S.C., Rice, G.I., Fabre, A., Badens, C., Gray, E.E., Hartley, J.L., Crow, Y.J., and Stetson, D.B. (2014). The SKIV2L RNA exosome limits activation of the RIG-I-like receptors. Nat. Immunol. 15.

Egesten, A., Dyer, K.D., Batten, D., Domachowske, J.B., and Rosenberg, H.F. (1997). Ribonucleases and host defense: Identification, localization and gene expression in adherent monocytes in vitro. Biochim. Biophys. Acta - Mol. Cell Res. 1358, 255–260.

Egorov, A., Brandt, S., Sereinig, S., Romanova, J., Ferko, B., Katinger, D., Grassauer, A., Alexandrova, G., Katinger, H., and Muster, T. (1998). Transfectant influenza A viruses with long deletions in the NS1 protein grow efficiently in Vero cells. J. Virol. 72, 6437–6441.

Ehrenfeld, E., and Hunt, T. (1971). Double-stranded poliovirus RNA inhibits initiation of protein synthesis by reticulocyte lysates. Proc. Natl. Acad. Sci. U.S.A. 68, 1075–1078.

El Awady, M.K., Anany, M.A., Esmat, G., Zayed, N., Tabll, A.A., Helmy, A., El Zayady, A.R., Abdalla, M.S., Sharada, H.M., El Raziky, M., El Akel, W., Abdalla, S., and Bader El Din, N.G. (2011). Single nucleotide polymorphism at exon 7 splice acceptor site of OAS1 gene determines response of hepatitis C virus patients to interferon therapy. J. Gastroenterol. Hepatol. 26, 843–850.

Elmore, S. (2007). Apoptosis: A review of programmed cell death. Toxicol Pathol 35, 495–516.

Emara, M.M., Ivanov, P., Hickman, T., Dawra, N., Tisdale, S., Kedersha, N., Hu, G.F., and Anderson, P. (2010). Angiogenin-induced tRNA-derived stress-induced RNAs promote stress-induced stress granule assembly. J. Biol. Chem. 285, 10959–10968.

Enami, K., Sato, T.A., Nakada, S., and Enami, M. (1994). Influenza virus NS1 protein stimulates translation of the M1 protein. J. Virol. 68, 1432–1437. Engel, D.A. (2013). The influenza virus NS1 protein as a therapeutic target. Antiviral Res. 99, 409–416.

Erickson, A.K. (2008). The role of interferon stimulated genes in resistance and immunity to hepatitis C virus infection. Ph.D. Thesis. University of Texas Southwestern Medical Center at Dallas: U.S.A.

Espert, L., Degols, G., Gongora, C., Blondel, D., Williams, B.R., Silverman, R.H., and Mechti, N. (2003). ISG20, a new interferon-induced RNase specific for single-stranded RNA, defines an alternative antiviral pathway against RNA genomic viruses. J. Biol. Chem. 278, 16151–16158.

Fett, J.W., Strydom, D.J., Lobb, R.R., Alderman, E.M., Bethune, J.L., Riordan, J.F., and Vallee, B.L. (1985). Isolation and characterization of angiogenin, an angiogenic protein from human carcinoma cells. Biochemistry 24, 5480–5486.

187

Field, L.L., Bonnevie-Nielsen, V., Pociot, F., Lu, S., Nielsen, T.B., and Beck-Nielsen, H. (2005). OAS1 splice site polymorphism controlling antiviral enzyme activity influences susceptibility to type 1 diabetes. Diabetes 54, 1588–1591.

Fiore, L., Ridolfi, B., Genovese, D., Buttinelli, G., Lucioli, S., Lahm, A., and Ruggeri, F.M. (1997). Poliovirus Sabin type 1 neutralization epitopes recognized by immunoglobulin A monoclonal antibodies. J. Virol. 71, 6905–6912.

Firth, A.E. (2014). Mapping overlapping functional elements embedded within the protein-coding regions of RNA viruses. Nucleic Acids Res. 42, 12425–12439.

Fitzgerald, K.A., McWhirter, S.M., Faia, K.L., Rowe, D.C., Latz, E., Golenbock, D.T., Coyle, A.J., Liao, S.-M., and Maniatis, T. (2003). IKKepsilon and TBK1 are essential components of the IRF3 signaling pathway. Nat. Immunol. 4, 491–496.

Flanegan, J.B., Pettersson, R.F., Ambros, V., Hewlett, M.J., and Baltimore, D. (1977). Covalent linkage of a protein to a defined nucleotide sequence at the 5ˊ-terminus of virion and replicative intermediate RNAs of poliovirus. Proc. Natl. Acad. Sci. U.S.A. 74, 961-965.

Floyd-Smith, G. (1988). (2ˊ-5ˊ)An-dependent endoribonuclease: enzyme levels are regulated by IFNβ, IFNγ, and cell culture conditions. J. Cell. Biochem. 38, 13–21.

Floyd-Smith, G., Slattery, E., and Lengyel, P. (1981). Interferon action: RNA cleavage pattern of a (2ˊ-5ˊ)oligoadenylate--dependent endonuclease. Science 212, 1030–1032.

Forconi, M., and Herschlag, D. (2009). Metal ion-based RNA cleavage as a structural probe. Methods Enzymol. 468, 91–106.

Forget, B.G., and Weissman, S.M. (1968). Oligonucleotides produced by digestion of KB cell ribosomal 5S ribonucleic acid with specific nucleases. J. Biol. Chem. 243, 5709– 5723.

Forget, B.G., and Weissman, S.M. (1969). The nucleotide sequence of ribosomal 5 S ribonucleic acid from KB cells. J. Biol. Chem. 244, 3148–3165.

Frank, J.J., and Levy, C.C. (1976). Properties of a human liver ribonuclease. J. Biol. Chem. 251, 5745–5751.

Frémont, M., Vaeyens, F., Herst, C.V., De Meirleir, K., and Englebienne, P. (2005). 37- Kilodalton/83-kilodalton RNase L isoform ratio in peripheral blood mononuclear cells: analytical performance and relevance for chronic fatigue syndrome. Clin. Diagn. Lab. Immunol. 12, 1259–1260.

Fu, H., Feng, J., Liu, Q., Sun, F., Tie, Y., Zhu, J., Xing, R., Sun, Z., and Zheng, X. (2009a). Stress induces tRNA cleavage by angiogenin in mammalian cells. FEBS Lett. 583, 437–442.

Fu, H., Hu, Z., Wen, J., Wang, K., and Liu, Y. (2009b). TGF-β promotes invasion and metastasis of gastric cancer cells by increasing fascin1 expression via ERK and JNK signal pathways. Acta Biochim. Biophys. 41, 648–656. 188

Futami, J., Tsushima, Y., Murato, Y., Tada, H., Sasaki, J., Seno, M., and Yamada, H. (1997). Tissue-specific expression of pancreatic-type RNases and RNase inhibitor in humans. DNA Cell Biol. 16, 413–419.

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

Gack, M.U., Albrecht, R.A., Urano, T., Inn, K.S., Huang, I.C., Carnero, E., Farzan, M., Inoue, S., Jung, J.U., and García-Sastre, A. (2009). Influenza A virus NS1 targets the ubiquitin ligase TRIM25 to evade recognition by the host viral RNA sensor RIG-I. Cell Host Microbe 5, 439–449.

Gao, P., Ascano, M., Wu, Y., Barchet, W., Gaffney, B.L., Zillinger, T., Serganov, A.A., Liu, Y., Jones, R.A., Hartmann, G., Tuschi, T., and Patel, D.J. (2013). Cyclic [G(2ˊ,5ˊ)pA(3ˊ,5ˊ)p] is the metazoan second messenger produced by DNA- activated cyclic GMP-AMP synthase. Cell 153, 1094–1107.

García-Sastre, A., Egorov, A., Matassov, D., Brandt, S., Levy, D.E., Durbin, J.E., Palese, P., and Muster, T. (1998). Influenza A virus lacking the NS1 gene replicates in interferon-deficient systems. Virology 252, 324–330.

Girardi, E., Chane-Woon-Ming, B., Messmer, M., Kaukinen, P., and Pfeffer, S. (2013). Identification of RNase L-dependent, 3ˊ-end-modified, viral small RNAs in Sindbis virus- infected mammalian cells. MBio 4, 00698-13.

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

Gleich, G.J., Loegering, D.A., Bell, M.P., Checkel, J.L., Ackerman, S.J., and McKean, D.J. (1986). Biochemical and functional similarities between human eosinophil-derived neurotoxin and eosinophil cationic protein: homology with ribonuclease. Proc. Natl. Acad. Sci. U.S.A. 83, 3146–3150.

Gog, J.R., Dos Santos Afonso, E., Dalton, R.M., Leclercq, I., Tiley, L., Elton, D., von Kirchbach, J.C., Naffakh, N., Escriou, N., and Digard, P. (2007). Codon conservation in the influenza A virus genome defines RNA packaging signals. Nucleic Acids Res. 35, 1897–1907.

Golden-Mason, L., Stone, A.E.L., Bambha, K.M., Cheng, L., and Rosen, H.R. (2012). Race- and gender-related variation in NKp46 expression associated with differential anti- HCV immunity. Hepatology 56, 1214–1222.

Gong, G., Waris, G., Tanveer, R., and Siddiqui, A. (2001). Human hepatitis C virus NS5A protein alters intracellular calcium levels, induces oxidative stress, and activates STAT-3 and NF-kappa B. Proc. Natl. Acad. Sci. U.S.A. 98, 9599–9604.

189

Gonzalez, T.N., Sidrauski, C., Dörfler, S., and Walter, P. (1999). Mechanism of non- spliceosomal mRNA splicing in the unfolded protein response pathway. EMBO J. 18, 3119–3132.

Gotte, G., and Libonati, M. (2004). Oligomerization of ribonuclease A: two novel three- dimensional domain-swapped tetramers. J. Biol. Chem. 279, 36670–36679.

Goubau, D., Schlee, M., Deddouche, S., Pruijssers, A.J., Zillinger, T., Goldeck, M., Schuberth, C., Van der Veen, A.G., Fujimura, T., Rehwinkel, J., Iskarpatyoti, J.A., Barchet, W., Ludwig, J., Dermody, T.S., Hartmann, G., and Reis e Sousa, C. (2014). Antiviral immunity via RIG-I-mediated recognition of RNA bearing 5ˊ-diphosphates. Nature 514, 372–375.

Gradi, A., Svitkin, Y. V, Imataka, H., and Sonenberg, N. (1998). Proteolysis of human eukaryotic translation initiation factor eIF4GII, but not eIF4GI, coincides with the shutoff of host protein synthesis after poliovirus infection. Proc. Natl. Acad. Sci. U.S.A. 95, 11089–11094.

Grooteclaes, M., Deveraux, Q., Hildebrand, J., Zhang, Q., Goodman, R.H., and Frisch, S.M. (2003). C-terminal-binding protein corepresses epithelial and proapoptotic gene expression programs. Proc. Natl. Acad. Sci. U.S.A. 100, 4568–4573.

Gu, J., Shumyatsky, G., Makan, N., and Reddy, R. (1997). Formation of 2 ˊ,3ˊ-cyclic phosphates at the 3ˊ end of human U6 small nuclear RNA in vitro. J. Biol. Chem. 272, 21989–21993.

Guo, Z., Chen, L.M., Zeng, H., Gomez, J. A., Plowden, J., Fujita, T., Katz, J.M., Donis, R.O., and Sambhara, S. (2007). NS1 protein of influenza A virus inhibits the function of intracytoplasmic pathogen sensor, RIG-I. Am. J. Respir. Cell Mol. Biol. 36, 263–269.

Häcker, H., and Karin, M. (2006). Regulation and Function of IKK and IKK-Related Kinases. Sci. STKE. 357, re13.

Haiser, H.J., Karginov, F. V., Hannon, G.J., and Elliot, M.A. (2008). Developmentally regulated cleavage of tRNAs in the bacterium Streptomyces coelicolor. Nucleic Acids Res. 36, 732–741.

Hale, B.G., Randall, R.E., Ortin, J., and Jackson, D. (2008). The multifunctional NS1 protein of influenza A viruses. J. Gen. Virol. 89, 2359–2376.

Han, J.-Q., and Barton, D.J. (2002). Activation and evasion of the antiviral 2 ˊ-5ˊ oligoadenylate synthetase / ribonuclease L pathway by hepatitis C virus mRNA. RNA 8, 512–525.

Han, J.-Q., Wroblewski, G., Xu, Z., Silverman, R.H., and Barton, D.J. (2004). Sensitivity of hepatitis C virus RNA to the antiviral enzyme ribonuclease L is determined by a subset of efficient cleavage sites. J. Interferon Cytokine Res. 24, 664–676.

Han, J.-Q., Townsend, H.L., Jha, B.K., Paranjape, J.M., Silverman, R.H., and Barton, D.J. (2007). A phylogenetically conserved RNA structure in the poliovirus open reading frame inhibits the antiviral endoribonuclease RNase L. J. Virol. 81, 5561–5572. 190

Han, Y., Whitney, G., Donovan, J., and Korennykh, A. (2012). Innate immune messenger 2-5A tethers human RNase L into active high-order complexes. Cell Rep. 2, 902–913.

Han, Y., Donovan, J., Rath, S., Whitney, G., Chitrakar, A., and Korennykh, A. (2014). Structure of human RNase L reveals the basis for regulated RNA decay in the IFN response. Science 343, 1244–1248.

Haralambieva, I.H., Dhiman, N., Ovsyannikova, I.G., Vierkant, R.A., Pankratz, V.S., Jacobson, R.M., and Poland, G.A. (2010). 2’-5'-Oligoadenylate synthetase single- nucleotide polymorphisms and haplotypes are associated with variations in immune responses to rubella vaccine. Hum. Immunol. 71, 383–391.

Harder, J., and Schröder, J.M. (2002). RNase 7, a novel innate immune defense antimicrobial protein of healthy human skin. J. Biol. Chem. 277, 46779–46784.

Hashem, Y., Des Georges, A., Dhote, V., Langlois, R., Liao, H.Y., Grassucci, R. A., Hellen, C.U.T., Pestova, T. V., and Frank, J. (2013). XStructure of the mammalian ribosomal 43S preinitiation complex bound to the scanning factor DHX29. Cell 153, 1108–1119.

Hassan, I.H., Zhang, M.S., Powers, L.S., Shao, J.Q., Baltrusaitis, J., Rutkowski, D.T., Legge, K., and Monick, M.M. (2012). Influenza A viral replication is blocked by inhibition of the inositol-requiring enzyme 1 (IRE1) stress pathway. J. Biol. Chem. 287, 4679– 4689.

Haud, N., Kara, F., Diekmann, S., Henneke, M., Willer, J.R., Hillwig, M.S., Gregg, R.G., Macintosh, G.C., Gärtner, J., Alia, A., and Hurlstone, A.F. (2011). rnaset2 mutant zebrafish model familial cystic leukoencephalopathy and reveal a role for RNase T2 in degrading ribosomal RNA. Proc. Natl. Acad. Sci. U.S.A. 108, 1099–1103.

Henneke, M., Diekmann, S., Ohlenbusch, A., Kaiser, J., Engelbrecht, V., Kohlschütter, A., Krätzner, R., Madruga-Garrido, M., Mayer, M., Opitz, L., Rodriguez, D., Ruschendorf, F., Schumacher, J., Thiele, H., Thoms, S., Steinfeld, R., Nurnberg, P., and Gartner, J. (2009). RNASET2-deficient cystic leukoencephalopathy resembles congenital cytomegalovirus brain infection. Nat. Genet. 41, 773–775.

Hilcenko, C., Simpson, P.J., Finch, A.J., Bowler, F.R., Churcher, M.J., Jin, L., Packman, L.C., Shlien, A., Campbell, P., Kirwan, M., Dokal, I., and Warren, A.J. (2013). Aberrant 3ˊ oligoadenylation of spliceosomal U6 small nuclear RNA in poikiloderma with neutropenia.

Hillwig, M.S., Rizhsky, L., Wang, Y., Umanskaya, A., Essner, J.J., and MacIntosh, G.C. (2009). Zebrafish RNase T2 genes and the evolution of secretory ribonucleases in animals. BMC Evol. Biol. 9, doi: 10.1186/1471-2148-9-170.

Hillwig, M.S., Contento, A.L., Meyer, A., Ebany, D., Bassham, D.C., and Macintosh, G.C. (2011). RNS2, a conserved member of the RNase T2 family, is necessary for ribosomal RNA decay in plants. Proc. Natl. Acad. Sci. U.S.A. 108, 1093–1098.

191

Hisai, H., Kato, J., Kobune, M., Murakami, T., Miyanishi, K., Takahashi, M., Yoshizaki, N., Takimoto, R., Terui, T., and Niitsu, Y. (2003). Increased expression of angiogenin in hepatocellular carcinoma in correlation with tumor vascularity. Clin. Cancer Res. 9, 4852–4859.

Hiscott, J. (2007). Triggering the innate antiviral response through IRF-3 activation. J. Biol. Chem. 282, 15325–15329.

Hofsteenge, J., Moldow, C., Vicentini, a M., Zelenko, O., Jarai-Kote, Z., and Neumann, U. (1998). A single amino acid substitution changes ribonuclease 4 from a uridine- specific to a cytidine-specific enzyme. Biochemistry 37, 9250–9257.

Hogle, A.J.M., Chow, M., and Filman, D.J. (1985). Three-dimensional structure of poliovirus at 2 .9 Å resolution. Science 229, 1358–1365.

Hollien, J., Lin, J.H., Li, H., Stevens, N., Walter, P., and Weissman, J.S. (2009). Regulated Ire1-dependent decay of messenger RNAs in mammalian cells. J. Cell Biol. 186, 323–331.

Holmberg, L., and Nygård, O. (1997). Mapping of nuclease-sensitive sites in native reticulocyte ribosomes--an analysis of the accessibility of ribosomal RNA to enzymatic cleavage. Eur. J. Biochem. 247, 160–168.

Hölzel, M., Orban, M., Hochstatter, J., Rohrmoser, M., Harasim, T., Malamoussi, A., Kremmer, E., Längst, G., and Eick, D. (2010). Defects in 18S or 28S rRNA processing activate the p53 pathway. J. Biol. Chem. 285, 6364–6370.

Horimoto, T., and Kawaoka, Y. (2005). Influenza: lessons from past pandemics, warnings from current incidents. Nat. Rev. Microbiol. 3, 591–600.

Horner, S.M., Liu, H.M., Park, H.S., Briley, J., and Gale, M. (2011). Mitochondrial- associated endoplasmic reticulum membranes (MAM) form innate immune synapses and are targeted by hepatitis C virus. Proc. Natl. Acad. Sci. U.S.A. 108, 14590–14595.

Hovanessian, A.G., and Justesen, J. (2007). The human 2ˊ-5ˊoligoadenylate synthetase family: unique interferon-inducible enzymes catalyzing 2ˊ-5ˊ instead of 3ˊ-5ˊ phosphodiester bond formation. Biochimie 89, 779–788.

Hovanessian, A.G., Brown, R.E., and Kerr, I.M. (1977). Synthesis of low molecular weight inhibitor of protein synthesis with enzyme from interferon treated cells. Nature 268, 537–540.

Hovnanian, A., Rebouillat, D., Mattei, M.G., Levy, E.R., Marié, I., Monaco, A.P., and Hovanessian, A.G. (1998). The human 2ˊ,5ˊ-oligoadenylate synthetase locus is composed of three distinct genes clustered on chromosome 12q24.2 encoding the 100-, 69-, and 40-kDa forms. Genomics 52, 267–277.

Hu, G.F., Riordan, J.F., and Vallee, B.L. (1997). A putative angiogenin receptor in angiogenin-responsive human endothelial cells. Proc. Natl. Acad. Sci. U.S.A. 94, 2204– 2209.

192

Hu, G.F., Xu, C.J., and Riordan, J.F. (2000). Human angiogenin is rapidly translocated to the nucleus of human umbilical vein endothelial cells and binds to DNA. J. Cell. Biochem. 76, 452–462.

Huang, H., Zeqiraj, E., Dong, B., Jha, B.K., Duffy, N.M., Orlicky, S., Thevakumaran, N., Talukdar, M., Pillon, M.C., Ceccarelli, D.F., Wan, E.C., Juang, Y.C., Mao, D.Y., Gaughan, C., Brinton, M.A., Perelygin, A.A., Kourinov, I., Guarné , A., Silverman, R.H., and Sicheri, F. (2014). Dimeric structure of pseudokinase RNase L bound to 2-5A reveals a basis for interferon-induced antiviral activity. Mol. Cell 53, 221–234.

Ibaragi, S., Yoshioka, N., Li, S., Hu, M.G., Hirukawa, S., Sadow, P.M., and Hu, G.F. (2009). Neamine Lnhibits prostate cancer growth by suppressing angiogenin-mediated rRNA transcription. Clin. Cancer Res. 15, 1981–1988.

Ibsen, M.S., Gad, H.H., Thavachelvam, K., Boesen, T., Despres, P., and Hartmann, R. (2014). The 2ˊ-5ˊ-oligoadenylate synthetase 3 enzyme potently synthesizes the 2ˊ-5ˊ- oligoadenylates required for RNase L activation. J. Virol. 88, 14222–14231.

Iordanov, M.S., Pribnow, D., Magun, J.L., Dinh, T.H., Pearson, J.A., Chen, S.L., and Magun, B.E. (1997). Ribotoxic stress response: activation of the stress-activated protein kinase JNK1 by inhibitors of the peptidyl transferase reaction and by sequence-specific RNA damage to the alpha-sarcin/ricin loop in the 28S rRNA. Mol. Cell. Biol. 17, 3373– 3381.

Iordanov, M.S., Paranjape, J.M., Zhou, A., Wong, J., Williams, B.R., Meurs, E.F., Silverman, R.H., and Magun, B.E. (2000). Activation of p38 mitogen-activated protein kinase and c-Jun NH(2)-terminal kinase by double-stranded RNA and encephalomyocarditis virus: involvement of RNase L, protein kinase R, and alternative pathways. Mol. Cell. Biol. 20, 617–627.

Irie, M. (1999). Structure-function relationships of acid ribonucleases: lysosomal, vacuolar, and periplasmic enzymes. Pharmacol. Ther. 81, 77–89.

Isaacs, A., and Lindenmann, J. (1957). Virus interference. I. the interferon. Proc. R. Soc. London 147, 258–267.

Ivanov, P., Emara, M.M., Villen, J., Gygi, S.P., and Anderson, P. (2011). Angiogenin- Induced tRNA Fragments Inhibit Translation Initiation. Mol. Cell 43, 613–623.

Ivanov, P., O’Day, E., Emara, M.M., Wagner, G., Lieberman, J., and Anderson, P. (2014). G-quadruplex structures contribute to the neuroprotective effects of angiogenin- induced tRNA fragments. Proc. Natl. Acad. Sci. U.S.A. 111, 18201–18206.

Iwama, M., Kunihiro, M., Ohgi, K., and Irie, M. (1981). Purification and properties of human urine ribonuclueases. J. Biochem. 89, 1005–1016.

Iwawaki, T., Hosoda, A., Okuda, T., Kamigori, Y., Nomura-Furuwatari, C., Kimata, Y., Tsuru, A., and Kohno, K. (2001). Translational control by the ER transmembrane kinase/ribonuclease IRE1 under ER stress. Nat. Cell Biol. 3, 158–164.

193

Jagger, B.W., Wise, H.M., Kash, J.C., Walters, K.-A., Wills, N.M., Xiao, Y.-L., Dunfee, R.L., Schwartzman, L.M., Ozinsky, A., Bell, G.L., Dalton, R.M., Lo, A., Efstathiou, S., Atkins, J.F., Firth, A.E., Taubenberger, J.K., and Digard, P. (2012). An overlapping protein-coding region in influenza A virus segment 3 modulates the host response. Science 337, 199–204.

Janicke, R.U., Sprengart, M.L., Wati, M.R., and Porter, A.G. (1998). Caspase-3 is required for DNA fragmentation and morphological changes associated with apoptosis. J. Biol. Chem. 273, 9357–9360.

Jurkin, J., Henkel, T., Nielsen, A.F., Minnich, M., Popow, J., Kaufmann, T., Heindl, K., Hoffmann, T., Busslinger, M., and Martinez, J. (2014). The mammalian tRNA ligase complex mediates splicing of XBP 1 mRNA and controls antibody secretion in plasma cells. EMBO 33, 2922–2937.

Justesen, J., Ferbus, D., and Thang, M.N. (1980). Elongation mechanism and substrate specificity of 2ˊ,5ˊ-oligoadenylate synthetase. Proc. Natl. Acad. Sci. U.S.A. 77, 4618– 4622.

Justesen, J., Hartmann, R., and Kjeldgaard, N.O. (2000). Gene structure and function of the 2ˊ-5ˊ-oligoadenylate synthetase family. Cell. Mol. Life Sci. 57, 1593–1612.

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

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

Kato, H., Takeuchi, O., Mikamo-Satoh, E., Hirai, R., Kawai, T., Matsushita, K., Hiiragi, A., Dermody, T.S., Fujita, T., and Akira, S. (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.

Kato, N., Ootsuyama, Y., Ohkoshi, S., Nakazawa, T., Sekiya, H., Hijikata, M., and Shimotohno, K. (1992). Characterization of hypervariable regions in the putative envelope protein of hepatitis C virus. Biochem. Biophys. Res. Commun. 189, 119–127.

Kawai, T., Takahashi, K., Sato, S., Coban, C., Kumar, H., Kato, H., Ishii, K.J., Takeuchi, O., and Akira, S. (2005). IPS-1, an adaptor triggering RIG-I- and MDA5-mediated type I interferon induction. Nat. Immunol. 6, 981–988.

Kawakami, E., Watanabe, T., Fujii, K., Goto, H., Watanabe, S., Noda, T., and Kawaoka, Y. (2011). Strand-specific real-time RT-PCR for distinguishing influenza vRNA, cRNA, and mRNA. J. Virol. Methods 173, 1–6.

Kedersha, N., and Anderson, P. (2007). Mammalian stress granules and processing bodies. Methods Enzymol. 431, 61-81.

194

Kedersha, N., Cho, M.R., Li, W., Yacono, P.W., Chen, S., Gilks, N., Golan, D.E., and Anderson, P. (2000). Dynamic shuttling of TIA-1 accompanies the recruitment of mRNA to mammalian stress granules. J. Cell Biol. 151, 1257–1268.

Kedersha, N., Chen, S., Gilks, N., Li, W., Miller, I.J., Stahl, J., and Anderson, P. (2002). Evidence that ternary complex (eIF2-GTP-tRNAiMet)-deficient preinitation complexes are core constituents of mammalian stress granules. Mol. Biol. Cell 13, 195–210.

Kedersha, N.L., Gupta, M., Li, W., Miller, I., and Anderson, P. (1999). RNA-binding proteins TIA-1 and TIAR link the phosphorylation of eIF-2α to the assembly of mammalian stress granules. J. Cell Biol. 147, 1431–1441.

Keel, A.Y., Jha, B.K., and Kieft, J.S. (2012). Structural architecture of an RNA that competitively inhibits RNase L. RNA 18, 88–99.

Kell, A.M., and Gale Jr., M. (2015). RIG-I in RNA virus recognition. Virology doi: 10.1016/j.virol.2015.02.017

Kerr, I.M., and Brown, R.E. (1978). pppA2ˊp5ˊA2ˊp5ˊA: An inhibitor of protein synthesis synthesized with an enzyme fraction from interferon-treated cells. Proc. Natl. Acad. Sci. U.S.A. 75, 256–260.

Khaperskyy, D.A., Hatchette, T.F., and McCormick, C. (2012). Influenza A virus inhibits cytoplasmic stress granule formation. FASEB J. 26, 1629–1639.

Kim, K.-I., Kim, S.-R., Sasase, N., Taniguchi, M., Harada, S., Kinoshita, K., Kim, S.-H., Akimoto, Y., Shikata, M., Kimura, N., Izawa, S., Ohtani, A., Nakao, K., Motojima, M., Kinoshita, M., Hirai, M., Ohzu, M., Hirooka, T., Nabeshima, S., Ishii, F., Tanaka, K., and Hotta, H. (2006). 2ˊ-,5ˊ-Oligoadenylate synthetase response ratio predicting virological response to PEG-interferon-alpha2b plus ribavirin therapy in patients with chronic hepatitis C. J. Clin. Pharm. Ther. 31, 441–446.

Kim, T., Leslie, P., and Zhang, Y. (2014). Ribosomal proteins as unrevealed caretakers for cellular stress and genomic instability. Oncotarget 5, 860–871.

Kishimoto, K., Shumei, L., Tsuji, T., Olson, K.A., and Hu, G. (2005). Endogenous antigenic in endothelil cells is a general requirement for cell proliferation and angiogenesis. Oncogene 24, 445-456.

Kivioja, T., Vähärautio, A., Karlsson, K., Bonke, M., Enge, M., Linnarsson, S., and Taipale, J. (2011). Counting absolute numbers of molecules using unique molecular identifiers. Nat. Methods 9, 72–74.

Kjær, K., Pahus, J., Hansen, M., Poulsen, J., Christensen, E., Justesen, J., and Martensen, P. (2014). Mitochondrial localization of the OAS1 p46 isoform associated with a common single nucleotide polymorphism. BMC Cell Biol. 15, doi: 10.1186/1471- 2121-15-33. . Kjaer, K.H., Poulsen, J.B., Reintamm, T., Saby, E., Martensen, P.M., Kelve, M., and Justesen, J. (2009). Evolution of the 2ˊ-5ˊ-oligoadenylate synthetase family in eukaryotes and bacteria. J. Mol. Evol. 69, 612–624. 195

Knapp, G., Ogden, R.C., and Jolla, L. (1979). Splicing of yeast tRNA precursors : structure of the reaction intermediates. Cell 18, 37–45.

Knapp, S., Yee, L.J., Frodsham, a J., Hennig, B.J.W., Hellier, S., Zhang, L., Wright, M., Chiaramonte, M., Graves, M., Thomas, H.C., Hill, A.V.S., and Thursz, M.R. (2003). Polymorphisms in interferon-induced genes and the outcome of hepatitis C virus infection: roles of MxA, OAS-1 and PKR. Genes Immun. 4, 411–419.

Kodym, R., Kodym, E., and Story, M.D. (2009). 2ˊ-5ˊ-Oligoadenylate synthetase is activated by a specific RNA sequence motif. Biochem. Biophys. Res. Commun. 388, 317–322.

Koike, K., and Miyoshi, H. (2006). Oxidative stress and hepatitis C viral infection. Hepatol. Res. 34, 65–73.

Kosmaczewski, S.G., Edwards, T.J., Han, S.M., Eckwahl, M.J., Meyer, B.I., Peach, S., Hesselberth, J.R., Wolin, S.L., and Hammarlund, M. (2014). The RtcB ligase I an essential component of the metazoan unfolded protein response. EMBO Reports 15, 1278-1285.

Kräusslich, H.G., Nicklin, M.J., Toyoda, H., Etchison, D., and Wimmer, E. (1987). Poliovirus proteinase 2A induces cleavage of eucaryotic initiation factor 4F polypeptide p220. J. Virol. 61, 2711–2718.

Kristiansen, H., Scherer, C.A., McVean, M., Iadonato, S.P., Vends, S., Thavachelvam, K., Steffensen, T.B., Horan, K.A., Kuri, T., Weber, F., Paludan, S.R., and Hartmann, R. (2010). Extracellular 2ˊ-5ˊ oligoadenylate synthetase stimulates RNase L-independent antiviral activity: a novel mechanism of virus-induced innate immunity. J. Virol. 84, 11898-11904.

Krug, R.M. (2015). Functions of the influenza A virus NS1 protein in antiviral defense. Curr. Opin. Virol. 12, 1–6.

Kubota, K., Nakahara, K., Ohtsuka, T., Yoshida, S., Kawaguchi, J., Fujita, Y., Ozeki, Y., Hara, A., Yoshimura, C., Furukawa, H., Haruyama, H., Ichikawa, K., Yamashita, M., Matsuoka, T., and Iijima, Y. (2004). Identification of 2ˊ-phosphodiesterase, which plays a role in the 2-5A system regulated by interferon. J. Biol. Chem. 279, 37832–37841.

Kumazaki, T., Hori, H., and Osawa, S. (1982). Nucleotide sequence of cytoplasmic 5S ribosomal RNA from Euglena gracilis. J. Mol. Biol. 18, 293–296.

Kureishy, N., Sapountzi, V., Prag, S., Anilkumar, N., and Adams, J.C. (2002). Fascins, and their roles in cell structure and function. BioEssays 24, 350–361.

Kuyumcu-Martinez, N.M., Van Eden, M.E., Younan, P., and Lloyd, R.E. (2004). Cleavage of poly(A)-binding protein by poliovirus 3C protease inhibits host cell translation: a novel mechanism for host translation shutoff. Mol. Cell. Biol. 24, 1779– 1790.

Kuznetsov, Y.G., Daijogo, S., Zhou, J., Semler, B.L., and McPherson, a (2005). Atomic force microscopy analysis of icosahedral virus RNA. J. Mol. Biol. 347, 41–52. 196

Kwon, Y.-C., Kang, J.-I., Hwang, S.B., and Ahn, B.-Y. (2013). The ribonuclease L- dependent antiviral roles of human 2ˊ,5ˊ-oligoadenylate synthetase family members against hepatitis C virus. FEBS Lett. 587, 156–164.

Landré, J.B.P., Hewett, P.W., Olivot, J.M., Friedl, P., Ko, Y., Sachinidis, A., and Moenner, M. (2002). Human endothelial cells selectively express large amounts of pancreatic-type ribonuclease (RNase 1). J. Cell. Biochem. 86, 540–552.

Langereis, M.A., Feng, Q., and van Kuppeveld, F.J. (2013). MDA5 localizes to stress granules, but this localization is not required for the induction of type I interferon. J. Virol. 87, 6314–6325.

Langmead, B., Trapnell, C., Pop, M., and Salzberg, S.L. (2009). Ultrafast and memory- efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, R25.

Lee, Y.F., Nomoto, A., Detjen, B.M., and Wimmer, E. (1977). A protein covalently linked to poliovirus genome RNA. Proc. Natl. Acad. Sci. U.S.A. 74, 59-63.

Lee, S.R., and Collins, K. (2005). Starvation-induced cleavage of the tRNA anticodon loop in Tetrahymena thermophila. J. Biol. Chem. 280, 42744–42749.

Levy, D.E., Marié, I.J., and Durbin, J.E. (2011). Induction and function of type I and III interferon in response to viral infection. Curr. Opin. Virol. 1, 476–486.

Li, M., and Gu, W. (2011). A critical role for noncoding 5S rRNA in regulating Mdmx stability. Mol. Cell 43, 1023–1032.

Li, M., and Pestka, J.J. (2008). Comparative induction of 28S ribosomal RNA cleavage by ricin and the trichothecenes deoxynivalenol and T-2 toxin in the macrophage. Toxicol. Sci. 105, 67–78.

Li, S., and Hu, G.F. (2010a). Angiogenin-mediated rRNA transcription in cancer and neurodegeneration. Int. J. Biochem. Mol. Biol. 1, 26–35.

Li, A., Dawson, J.C., Forero-Vargas, M., Spence, H.J., Yu, X., König, I., Anderson, K., and Machesky, L.M. (2010b). The actin-bundling protein fascin stabilizes actin in invadopodia and potentiates protrusive invasion. Curr. Biol. 20, 339–345.

Li, S., and Hu, G.F. (2012). Emerging role of angiogenin in stress response and cell survival under adverse conditions. J. Cell. Physiol. 227, 2822–2826.

Li, C.Z., Kato, N., Chang, J.H., Muroyama, R., Shao, R.X., Dharel, N., Sermsathanasawadi, R., Kawabe, T., and Omata, M. (2009). Polymorphism of OAS-1 determines liver fibrosis progression in hepatitis C by reduced ability to inhibit viral replication. Liver Int. 29, 1413–1421.

Li, G., Xiang, Y., Sabapathy, K., and Silverman, R.H. (2004). An apoptotic signaling pathway in the interferon antiviral response mediated by RNase L and c-Jun NH2- terminal kinase. J. Biol. Chem. 279, 1123–1131.

197

Li, K., Foy, E., Ferreon, J.C., Nakamura, M., Ferreon, A.C.M., Ikeda, M., Ray, S.C., Gale Jr., M., and Lemon, S.M. (2005a). Immune evasion by hepatitis C virus NS3/4A protease-mediated cleavage of the Toll-like receptor 3 adaptor protein TRIF. Proc. Natl. Acad. Sci. U.S.A. 102, 2992–2997.

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

Li, S., Min, J.Y., Krug, R.M., and Sen, G.C. (2006). Binding of the influenza A virus NS1 protein to PKR mediates the inhibition of its activation by either PACT or double- stranded RNA. Virology 349, 13–21.

Li, S., Sheng, J., Hu, J.K., Yu, W., Kishikawa, H., Hu, M.G., Shima, K., Wu, D., Xu, Z., Xin, W., Sims, K.B., Landers, J.E., Brown, R.H., and Hu, G.F. (2013). Ribonuclease 4 protects neuron degeneration by promoting angiogenesis, neurogenesis, and neuronal survival under stress. Angiogenesis 16, 387–404.

Li, X.L., Blackford, J. A., and Hassel, B.A. (1998). RNase L mediates the antiviral effect of interferon through a selective reduction in viral RNA during encephalomyocarditis virus infection. J. Virol. 72, 2752–2759.

Li, X.L., Blackford, J.A., Judge, C.S., Liu, M., Xiao, W., Kalvakolanu, D. V, and Hassel, B.A. (2000). RNase-L-dependent destabilization of interferon-induced mRNAs. A role for the 2-5A system in attenuation of the interferon response. J. Biol. Chem. 275, 8880– 8888.

Li, X.-L., Ezelle, H.J., Kang, T.-J., Zhang, L., Shirey, K.A., Harro, J., Hasday, J.D., Mohapatra, S.K., Crasta, O.R., Vogel, S.N., Cross, A.S., and Hassel, B.A. (2008). An essential role for the antiviral endoribonuclease, RNase-L, in antibacterial immunity. Proc. Natl. Acad. Sci. U.S.A. 105, 20816–20821.

Libonati, M., and Gotte, G. (2004). Oligomerization of bovine ribonuclease A: structural and functional features of its multimers. Biochem. J. 380, 311–327.

Libonati, M., and Sorrentino, S. (2001). Degradation of double-stranded RNA by mammalian pancreatic-type ribonucleases. Methods Enzymol. 341, 234–248.

Licht, K., Medenbach, J., Lührmann, R., Kambach, C., and Bindereif, A. (2008). 3 ˊ-cyclic phosphorylation of U6 snRNA leads to recruitment of recycling factor p110 through LSm proteins. RNA 14, 1532–1538.

Lim, J.K., Lisco, A., McDermott, D.H., Huynh, L., Ward, J.M., Johnson, B., Johnson, H., Pape, J., Foster, G.A., Krysztof, D., Follmann, D., Stramer, S.L., Margolis, L.B., and Murphy, P.M. (2009). Genetic variation in OAS1 is a risk factor for initial infection with West Nile virus in man. PLoS Pathog. 5. e1000321.

Lin, J.H., Li, H., Yasumura, D., Cohen, H.R., Zhang, C., Panning, B., Shokat, K.M., Lavail, M.M., and Walter, P. (2007). IRE1 signaling affects cell fate during the unfolded protein response. Science 318, 944–949.

198

Lin, R.-J., Yu, H.-P., Chang, B.-L., Tang, W.-C., Liao, C.-L., and Lin, Y.-L. (2009). Distinct antiviral roles for human 2ˊ,5ˊ-oligoadenylate synthetase family members against dengue virus infection. J. Immunol. 183, 8035–8043.

Lindenbach, B.D., and Rice, C.M. (2013). The ins and outs of hepatitis C virus entry and assembly. Nat. Rev. Microbiol. 11, 688–700.

Lindenbach, B.D., Evans, M.J., Syder, A.J., Wölk, B., Timothy, L., Liu, C.C., Maruyama, T., Hynes, R.O., Burton, D.R., Mckeating, J.A., and Rice, C.M. (2005). Complete replication of hepatitis C virus in cell culture. Science 309, 623–626.

Long, T.M., Chakrabarti, A., Ezelle, H.J., Brennan-Laun, S.E., Raufman, J.-P., Polyakova, I., Silverman, R.H., and Hassel, B.A. (2013). RNase-L deficiency exacerbates experimental colitis and colitis-associated cancer. Inflamm. Bowel Dis. 19, 1295–1305.

Lu, C., Ranjith-Kumar, C.T., Hao, L., Kao, C.C., and Li, P. (2011). Crystal structure of RIG-I C-terminal domain bound to blunt-ended double-strand RNA without 5ˊ triphosphate. Nucleic Acids Res. 39, 1565–1575.

Lu, Y., Liang, F.X., and Wang, X. (2014). A synthetic biology approach identifies the mammalian UPR RNA ligase RtcB. Mol. Cell 55, 758–770.

Luhtala, N., and Parker, R. (2010). T2 family ribonucleases: ancient enzymes with diverse roles. Trends Biochem. Sci. 35, 253–259.

Lund, E., and Dahlberg, J.E. (1992). Cyclic 2ˊ,3'-phosphates and nontemplated nucleotides at the 3ˊ end of spliceosomal U6 small nuclear RNA's. Science 255, 327– 330.

Lund, J.M., Alexopoulou, L., Sato, A., Karow, M., Adams, N.C., Gale, N.W., Iwasaki, A., and Flavell, R.A. (2004). Recognition of single-stranded RNA viruses by Toll-like receptor 7. Proc. Natl. Acad. Sci. U.S.A. 101, 5598–5603.

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

Ma, W., Liu, Q., Qiao, C., del Real, G., García-Sastre, A., Webby, R.J., and Richt, J.A. (2014). North American triple reassortant and eurasian H1N1 swine influenza viruses do not readily reassort to generate a 2009 pandemic H1N1-like virus. MBio 5, e00919-13.

Malathi, K., Dong, B., Gale Jr, M., and Silverman, R.H. (2007). Small self-RNA generated by RNase L amplifies antiviral innate immunity. Nature 448, 816–820.

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

Marie, I., and Hovanessian, A.G. (1992). The 69-kDa 2-5A synthetase is composed of two homologous and adjacent functional domains. J. Biol. Chem. 267, 9933–9939.

199

Marie, I., Svab, J., Robert, N., Galabru, J., and Hovanessian, A.G. (1990). Differential expression and distinct structure of 69- and 100-kDa forms of 2-5A synthetase in human cells treated with interferon. J. Biol. Chem. 265, 18601–18607.

Markham, R., and Smith, J.D. (1952). The structure of ribonucleic acids. 1. Cyclic nucleotides produced by ribonuclease and by alkaline hydrolysis. Biochem. J. 52, 552– 557.

Martínez-Gil, L., Goff, P.H., Hai, R., García-Sastre, A., Shaw, M.L., and Palese, P. (2013). A Sendai virus-derived RNA agonist of RIG-I as a virus vaccine adjuvant. J. Virol. 87, 1290–1300.

Martino, M.B., Jones, L., Brighton, B., Ehre, C., Abdulah, L., Davis, C.W., Ron, D., O’Neal, W.K., and Ribeiro, C.M.P. (2012). The ER stress transducer IRE1β is required for airway epithelial mucin production. Mucosal Immunol. 6, 639–654.

Maurel, M., Chevet, E., Tavernier, J., and Gerlo, S. (2014). Getting RIDD of RNA: IRE1 in cell fate regulation. Trends Biochem. Sci. 39, 245–254.

McWilliam, H., Li, Weizhong, Uludag, M., Squizzato, S., Park, Y.M., Buso, N., Cowley, A.P., and Lopez, R. (2013). Analysis tool web services from the EMBL-EBI. Nucleic Acids Res. 41, W597-W600.

Medina, R.A., and García-Sastre, A. (2011). Influenza A viruses: new research developments. Nat. Rev. Microbiol. 9, 590–603.

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

Mibayashi, M., Martínez-Sobrido, L., Loo, Y.-M., Cárdenas, W.B., Gale, M., and García- Sastre, A. (2007). Inhibition of retinoic acid-inducible gene I-mediated induction of beta interferon by the NS1 protein of influenza A virus. J. Virol. 81, 514–524.

Mihm, U., Ackermann, O., Welsch, C., Herrmann, E., Hofmann, W.P., Grigorian, N., Welker, M.W., Lengauer, T., Zeuzem, S., and Sarrazin, C. (2009). Clinical relevance of the 2ˊ-5ˊ-oligoadenylate synthetase/RNase L system for treatment response in chronic hepatitis C. J. Hepatol. 50, 49–58.

Mihm, U., Hofmann, W.P., Welsch, C., Polta, A., Lengauer, T., Zeuzem, S., Sarrazin, C., and Herrmann, E. (2010). Effect of ribavirin on the frequency of RNase L cleavage sites within the hepatitis C viral genome. J. Viral Hepat. 17, 217–221.

Min, J.-Y., and Krug, R.M. (2006). The primary function of RNA binding by the influenza A virus NS1 protein in infected cells: Inhibiting the 2’-5' oligo (A) synthetase/RNase L pathway. Proc. Natl. Acad. Sci. U.S.A. 103, 7100–7105.

Min, J.Y., Li, S., Sen, G.C., and Krug, R.M. (2007). A site on the influenza A virus NS1 protein mediates both inhibition of PKR activation and temporal regulation of viral RNA synthesis. Virology 363, 236–243.

200

Moroianu, J., and Riordan, J.F. (1994). Nuclear translocation of angiogenin in proliferating endothelial cells is essential to its angiogenic activity. Proc. Natl. Acad. Sci. U.S.A. 91, 1677–1681.

Morrison, J.M., and Racaniello, V.R. (2009). Proteinase 2Apro is essential for enterovirus replication in type I interferon-treated cells. J. Virol. 83, 4412–4422.

Moss, W.N., Priore, S.F., and Turner, D.H. (2011). Identification of potential conserved RNA secondary structure throughout influenza A coding regions. RNA 17, 991–1011.

Mroczek, S., and Kufel, J. (2008). Apoptotic signals induce specific degradation of ribosomal RNA in yeast. Nucleic Acids Res. 36, 2874–2888.

Mroczek, S., Krwawicz, J., Kutner, J., Lazniewski, M., Kuciński, I., Ginalski, K., and Dziembowski, A. (2012). C16orf57, a gene mutated in poikiloderma with neutropenia, encodes a putative phosphodiesterase responsible for the U6 snRNA 3ˊ end modification. Genes Dev. 26, 1911–1925.

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

Nguyen, L.H., Espert, L., Mechti, N., and Wilson, D.M. (2001). The human interferon- and estrogen-regulated ISG20/HEM45 gene product degrades single-stranded RNA and DNA in vitro. Biochemistry 40, 7174–7179.

Nilsen, T.W., and Baglioni, C. (1979). Mechanism for discrimination between viral and host mRNA in interferon-treated cells. Proc. Natl. Acad. Sci. U.S.A. 76, 2600–2604.

Nishida, N., Yano, H., Nishida, T., Kamura, T., and Kojiro, M. (2006). Angiogenesis in cancer. Vasc Health Risk Manag. 2, 213-219.

Odendall, C., and Kagan, J.C. (2015). The unique regulation and functions of type III interferons in antiviral immunity. Curr. Opin. Virol. 12, 47–52.

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

Okada, T., Yoshida, H., Akazawa, R., Negishi, M., and Mori, K. (2002). Distinct roles of activating transcription factor 6 (ATF6) and double-stranded RNA-activated protein kinase-like endoplasmic reticulum kinase (PERK) in transcription during the mammalian unfolded protein response. Biochem. J. 366, 585–594.

Okuda, M., Li, K., Beard, M.R., Showalter, L.A., Scholle, F., Lemon, S.M., and Weinman, S.A. (2002). Mitochondrial injury, oxidative stress, and antioxidant gene expression are induced by hepatitis C virus core protein. Gastroenterology 122, 366–375.

201

Onomoto, K., Jogi, M., Yoo, J.S., Narita, R., Morimoto, S., Takemura, A., Sambhara, S., Kawaguchi, A., Osari, S., Nagata, K., Matsumiya, T., Namiki, H., Yoneyama, M., and Fujita, T. (2012). Critical role of an antiviral stress granule containing RIG-I and PKR in viral detection and innate immunity. PLoS One 7, e43031.

Onomoto, K., Yoneyama, M., Fung, G., Kato, H., and Fujita, T. (2014). Antiviral innate immunity and stress granule responses. Trends Immunol. 35, 420–428.

Orioli, A., Pascali, C., Pagano, A., Teichmann, M., and Dieci, G. (2012). RNA polymerase III transcription control elements: Themes and variations. Gene 493, 185– 194.

Padhi, A.K., Jayaram, B., and Gomes, J. (2013). Prediction of functional loss of human angiogenin mutants associated with ALS by molecular dynamics simulations. Sci. Rep. 3, 1225.

Page, G.S., Mosser, a G., Hogle, J.M., Filman, D.J., Rueckert, R.R., and Chow, M. (1988). Three-dimensional structure of poliovirus serotype 1 neutralizing determinants. J. Virol. 62, 1781–1794.

Palmenberg, A.C., and Sgro, J.-Y. (1997). Topological organization of picornaviral genomes: statistical prediction of RNA structural signals. Semin. Virol. 8, 231–241.

Pandey, M., Bajaj, G.D., and Rath, P.C. (2004). Induction of the interferon-inducible RNA-degrading enzyme, RNase L, by stress-inducing agents in the human cervical carcinoma cells. RNA Biol. 1, 21–27.

Park, S.H., Song, J.Y., Kim, Y.K., Heo, J.H., Kang, H., Kim, G., An, H.J., and Kim, T.H. (2014). Fascin1 expression in high-grade serous ovarian carcinoma is a prognostic marker and knockdown of fascin1 suppresses the proliferation of ovarian cancer cells. Int. J. Oncol. 44, 637–646.

Passmore, L.A., Schmeing, T.M., Maag, D., Applefield, D.J., Acker, M.G., Algire, M.A., Lorsch, J.R., and Ramakrishnan, V. (2007). The Eukaryotic translation Initiation Factors eIF1 and eIF1A Induce an open conformation of the 40S ribosome. Mol. Cell 26, 41–50.

Peebles, C.L., Gegenheimer, P., and Abelson, J. (1983). Precise excision of intervening sequences from precursor tRNAs by a yeast endonuclease. Cell 32, 525-536.

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

Pichlmair, A., Schulz, O., Ping Tan, C., Naslund, T.I., Liljestrom, P., Weber, F., and Reis e Sousa, C. (2006). RIG-I-mediated antiviral responses to single-stranded RNA bearing 5ˊ-phosphates. Science 314, 997–1002.

Piper, P.W., Bellatin, J.A., and Lockheart, A. (1983). Altered maturation of sequences at the 3ˊ terminus of 5S gene transcripts in a Saccharomyces cerevisiae mutant that lacks a RNA processing endonuclease. EMBO J. 2, 353–359.

202

Pizzo, E., Sarcinelli, C., Sheng, J., Fusco, S., Formiggini, F., Netti, P., Yu, W., D’Alessio, G., and Hu, G.-F. (2013). Ribonuclease/angiogenin inhibitor 1 regulates stress-induced subcellular localization of angiogenin to control growth and survival. J. Cell Sci. 126, 4308–4319.

Platanias, L.C. (2005). Mechanisms of type-I- and type-II-interferon-mediated signalling. Nat. Rev. Immunol. 5, 375–386.

Portela, A., and Digard, P. (2002). The influenza virus nucleoprotein: A multifunctional RNA-binding protein pivotal to virus replication. J. Gen. Virol. 83, 723–734.

Porter, A.G., and Jänicke, R.U. (1999). Emerging roles of caspase-3 in apoptosis. Cell Death Differ. 6, 99–104.

Poulsen, J.B., Andersen, K.R., Kjær, K.H., Durand, F., Faou, P., Vestergaard, A.L., Talbo, G.H., Hoogenraad, N., Brodersen, D.E., Justesen, J., and Martensen, P.M. (2011). Human 2ˊ-phosphodiesterase localizes to the mitochondrial matrix with a putative function in mitochondrial RNA turnover. Nucleic Acids Res. 39, 3754–3770.

Presser, L.D., Haskett, A., and Waris, G. (2011). Hepatitis C virus-induced furin and thrombospondin-1 activate TGF-β1: role of TGF-β1 in HCV replication. Virology 412, 284–296.

Quinkert, D., Bartenschlager, R., and Lohmann, V. (2005). Quantitative analysis of the hepatitis C virus replication complex. J. Virol. 79, 13594-13605.

Quinlan, A.R., and Hall, I.M. (2010). BEDTools: A flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842.

R Core Team (2013). R : A Language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.

Raines, R.T. (2004). Active site of ribonuclease A. Artif. Nucleases 13, 19–32.

Rebouillat, D., Marié, I., and Hovanessian, A.G. (1998). Molecular cloning and characterization of two related and interferon- induced 56-kDa and 30-kDa proteins highly similar to 2ˊ-5ˊ oligoadenylate synthetase. Eur. J. Biochem. 257, 319–330.

Richards, A.L., Soares-Martins, J.A.P., Riddell, G.T., and Jackson, W.T. (2014). Generation of unique poliovirus RNA replication organelles. MBio 5, e00833-13.

Rigby, R.E., Webb, L.M., Mackenzie, K.J., Li, Y., Leitch, A., Martin, A.M., Lundie, R.J., Revuelta, A., Davidson, D.J., Diebold, S., Modis, Y., MacDonald, A.S., and Jackson, A.P. (2014). RNA : DNA hybrids are a novel molecular pattern sensed by TLR 9. EMBO J. 33, 542-558.

Roberts, W.K., Clemens, M.J., and Kerr, I.M. (1976). Interferon-induced inhibition of protein synthesis in L-cell extracts: an ATP-dependent step in the activation of an inhibitor by double-stranded RNA. Proc. Natl. Acad. Sci. U.S.A. 73, 3136–3140.

203

Rökman, A., Ikonen, T., Seppälä, E.H., Nupponen, N., Autio, V., Mononen, N., Bailey- Wilson, J., Trent, J., Carpten, J., Matikainen, M.P., Koivisto, P.A., Tammela, T.L.J., Kallioniemi, O.-P., and Schleutker, J. (2002). Germline alterations of the RNASEL gene, a candidate HPC1 gene at 1q25, in patients and families with prostate cancer. Am. J. Hum. Genet. 70, 1299–1304.

Rosen, H.R. (2011). Chronic Hepatitis C Infection. N. Engl. J. Med. 364, 2429–2438.

Le Roy, F., Silhol, M., Salehzada, T., and Bisbal, C. (2007). Regulation of mitochondrial mRNA stability by RNase L is translation-dependent and controls IFNalpha-induced apoptosis. Cell Death Differ. 14, 1406–1413.

Rudolph, B., Podschun, R., Sahly, H., Schubert, S., Schröder, J.M., and Harder, J. (2006). Identification of RNase 8 as a novel human antimicrobial protein. Antimicrob. Agents Chemother. 50, 3194–3196.

Rugeles, M.T., Trubey, C.M., Bedoya, V.I., Pinto, L.A., Oppenheim, J.J., Rybak, S.M., and Shearer, G.M. (2003). Ribonuclease is partly responsible for the HIV-1 inhibitory effect activated by HLA alloantigen recognition. AIDS 17, 481–486.

Rusch, L., Zhou, A., and Silverman, R.H. (2000). Caspase-dependent apoptosis by 2ˊ,5ˊ-oligoadenylate activation of RNase L is enbanced by IFN-β. J. Interf. Cytokine Res. 20, 1091–1100.

Rusch, L., Dong, B., and Silverman, R.H. (2001). Monitoring activation of ribonuclease L by 2ˊ,5ˊ-oligoadenylates using purified recombinant enzyme and intact malignant glioma cells. Methods Enzymol. 342, 10-20.

Rushizky, G.W., Knight, C.A., and Sober, H.A. (1961). Studies on the preferential specificity of pancreatic ribonuclease as deduced from partial digests. J. Biol. Chem. 236, 2732–2737.

Rutz, M., Metzger, J., Gellert, T., Luppa, P., Lipford, G.B., Wagner, H., and Bauer, S. (2004). Toll-like receptor 9 binds single-stranded CpG-DNA in a sequence- and pH- dependent manner. Eur. J. Immunol. 34, 2541–2550.

Ruggero, D., and Pandolfi, P.P. (2003). Does the ribosome translate cancer? Nat. Rev. Cancer 3, 179-192.

Rybak, S.M., and Vallee, B.L. (1988). Base cleavage specificity of angiogenin with Saccharomyces cerevisiae and Escherichia coli 5S RNAs. Biochemistry 27, 2288–2294.

Saikia, M., Krokowski, D., Guan, B.J., Ivanov, P., Parisien, M., Hu, G.F., Anderson, P., Pan, T., and Hatzoglou, M. (2012). Genome-wide identification and quantitative analysis of cleaved tRNA fragments induced by cellular stress. J. Biol. Chem. 287, 42708–42725.

Saikia, M., Jobava, R., Parisien, M., Putnam, A., Krokowski, D., Gao, X.-H., Guan, B.-J., Yuan, Y., Jankowsky, E., Feng, Z., Hu, G.-F., Pusztai-Carey, M., Gorla, M., Sepuri, N.B.V., Pan, T., and Hatzoglou, M. (2014). Angiogenin-cleaved tRNA halves interact with cytochrome c, protecting cells from apoptosis during osmotic stress. Mol. Cell. Biol. 34, 2450–2463. 204

Saito, T., Owen, D.M., Jiang, F., Marcotrigiano, J., and Gale Jr., M. (2008). Innate immunity induced by composition-dependent RIG-I recognition of hepatitis C virus RNA. Nature 454, 523–527.

Salvatore, M., Basler, C.F., Parisien, J.-P., Horvath, C.M., Bourmakina, S., Zheng, H., Muster, T., Palese, P., and Garcia-Sastre, A. (2002). Effects of influenza A virus NS1 protein on protein expression : the NS1 protein enhances translation and is not required for shutoff of host protein synthesis. J. Virol. 76, 1206–1212.

Sankar, S., Chan, H., Romanow, W.J., Li, J., and Bates, R.J. (2006). IKK-i signals through IRF3 and NFκB to mediate the production of inflammatory cytokines. Cell. Signal. 18, 982–993.

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

Sawicki, D.L., Silverman, R.H., Williams, B.R., and Sawicki, S.G. (2003). Alphavirus minus-strand synthesis and persistence in mouse embryo fibroblasts derived from mice lacking RNase L and protein kinase R. J. Virol. 77, 1801–181

Saxena, S.K., Rybak, S.M., Davey, R.T., Youle, R.J., and Ackerman, E.J. (1992). Angiogenin is a cytotoxic, tRNA-specific ribonuclease in the RNase A superfamily. J. Biol. Chem. 267, 21982–21986.

Schmidt, A., Chernajovsky, Y., Shulman, L., Federman, P., Berissi, H., and Revel, M. (1979). An interferon-induced phosphodiesterase degrading ( 2ˊ-5ˊ) oligoisoadenylate and the C-C-A terminus of tRNA. Proc. Natl. Acad. Sci. U.S.A. 76, 4788–4792.

Schroder, K., Hertzog, P.J., Ravasi, T., and Hume, D.A. (2004). Interferon-γ: an overview of signals, mechanisms and functions. J. Leukoc. Biol. 75, 163–189.

Schutz, K., Hesselberth, J.R., and Fields, S. (2010). Capture and sequence analysis of RNAs with terminal 2ˊ, 3ˊ-cyclic phosphates. RNA 16, 621–631.

Selitsky, S.R., Baran-Gale, J., Honda, M., Yamane, D., Masaki, T., Fannin, E.E., Guerra, B., Shirasaki, T., Shimakami, T., Kaneko, S., Lanford, R.E., Lemon, S.M., and Sethupathy, P. (2015). Small tRNA-derived RNAs are increased and more abundant than microRNAs in chronic hepatitis B and C. Sci. Rep. 5, 7675.

Sen, G.C., and Sarkar, S.N. (2005). Transcriptional signaling by double-stranded RNA: role of TLR3. Cytokine Growth Factor Rev. 16, 1–14.

Seth, R.B., Sun, L., Ea, C.-K., and Chen, Z.J. (2005). Identification and characterization of MAVS, a mitochondrial antiviral signaling protein that activates NF-κB and IRF3. Cell 122, 669–682.

Shapiro, R., and Vallee, B.L. (1987). Human placental ribonuclease inhibitor abolishes both angiogenic and ribonucleolytic activities of angiogenin. Proc. Natl. Acad. Sci. U.S.A. 84, 2238–2241.

205

Shapiro, R., and Vallee, B.L. (1989). Site-directed mutagenesis of histidine-13 and histidine-114 of human angiogenin. Alanine derivatives inhibit angiogenin-induced angiogenesis. Biochemistry 28, 7401–7408.

Shapiro, G.I., Gurney, T., and Krug, R.M. (1987). Influenza virus gene expression: control mechanisms at early and late times of infection and nuclear-cytoplasmic transport of virus-specific RNAs. J. Virol. 61, 764–773.

Shapiro, R., Fett, J.W., Strydom, D.J., and Vallee, B.L. (1986a). Isolation and characterization of a human colon carcinoma-secreted enzyme with pancreatic ribonuclease-like activity. Biochemistry 25, 7255–7264.

Shapiro, R., Riordan, J.F., and Vallee, B.L. (1986b). Characteristic ribonucleolytic activity of human angiogenin. Biochemistry 25, 3527–3532.

Shapiro, R., Fox, E.A, and Riordan, J.F. (1989). Role of lysines in human angiogenin: chemical modification and site-directed mutagenesis. Biochemistry 28, 1726–1732.

Shchepachev, V., and Azzalin, C.M. (2013). The Mpn1 RNA exonuclease: cellular functions and implication in disease. FEBS Lett. 587, 1858–1862.

Shchepachev, V., Wischnewski, H., Missiaglia, E., Soneson, C., and Azzalin, C.M. (2012). Mpn1, Mutated in Poikiloderma with Neutropenia Protein 1, Is a Conserved 3ˊ-to- 5ˊ RNA Exonuclease Processing U6 Small Nuclear RNA. Cell Rep. 2, 855–865.

Shen, J., Chen, X., Hendershot, L., and Prywes, R. (2002). ER stress regulation of ATF6 localization by dissociation of BiP/GRP78 binding and unmasking of golgi localization signals. Dev. Cell 3, 99–111.

Shindo, M., Hamada, K., Morikawa, T., Harano, Y., Nakajima, T., and Okuno, T. (2008). In vivo interferon system assessed by 2ˊ-5ˊ oligoadenylate synthetase activity in chronic hepatitis C virus patients treated with pegylated interferon and ribavirin. Hepatol. Res. 38, 1213–1220.

Shu, J., Huang, M., Tian, Q., Shui, Q., Zhou, Y., and Chen, J. (2015). Downregulation of angiogenin inhibits the growth and induces apoptosis in human bladder cancer cells through regulating AKT/mTOR signaling pathway. J. Mol. Histol. 46, 157-171.

Sidrauski, C., and Walter, P. (1997). The transmembrane kinase Ire1p is a site-specific endonuclease that initiates mRNA splicing in the unfolded protein response. Cell 90, 1031–1039.

Sidrauski, C., Cox, J.S., and Walter, P. (1996). tRNA ligase is required for regulated mRNA splicing in the unfolded protein response. Cell 87, 405–413.

Silverman, R.H. (2003). Implications for RNase L in prostate cancer biology. Biochemistry 42, 1805–1812.

Silverman, R.H., Cayley, P.J., Knight, M., Gilbert, C.S., and Kerr, I.M. (1982). Control of the ppp(A2ˊp)nA system in HeLa cells. Eur. J. Biochem. 124, 131–138.

206

Silverman, R.H., Skehel, J.J., James, T.C., Wreschner, D.H., and Kerr, I.M. (1983). rRNA cleavage as an index of ppp(A2ˊp)nA Activity in interferon-treated encephalomyocarditis virus-infected cells. J. Virol. 46, 1051–1055.

Silverman, R.H., Dong, B., Maitra, R.K., Player, M.R., and Torrence, P.F. (2000). Selective RNA cleavage by isolated RNase L activated with 2-5A antisense chimeric oligonucleotides. Methods Enzymol. 313, 522–533.

Siridechadilok, B., Fraser, C.S., Hall, R.J., Doudna, J.A., and Nogales, E. (2005). Structural roles for human translation factor eIF3 in initiation of protein synthesis. Science 310, 1513–1515.

Skorupa, A., King, M. A., Aparicio, I.M., Dussmann, H., Coughlan, K., Breen, B., Kieran, D., Concannon, C.G., Marin, P., and Prehn, J.H.M. (2012). Motoneurons secrete angiogenin to induce RNA cleavage in astroglia. J. Neurosci. 32, 5024–5038.

Smirnoff, P., Roiz, L., Angelkovitch, B., Schwartz, B., and Shoseyov, O. (2006). A recombinant human RNASET2 glycoprotein with antitumorigenic and antiangiogenic characteristics: expression, purification, and characterization. Cancer 107, 2760–2769.

Smirnov, A., Tarassov, I., Mager-Heckel, A.-M., Letzelter, M., Martin, R.P., Krasheninnikov, I.A., and Entelis, N. (2008). Two distinct structural elements of 5S rRNA are needed for its import into human mitochondria. RNA 14, 749–759.

Smith, D.B., Bukh, J., Kuiken, C., Muerhoff, A.S., Rice, C.M., Stapleton, J.T., and Simmonds, P. (2014). Expanded classification of hepatitis C virus into 7 genotypes and 67 subtypes: updated criteria and genotype assignment web resource. Hepatology 59, 318–327.

Smith, J.A., Schmechel, S.C., Williams, B.R.G., Silverman, R.H., and Schiff, L.A. (2005). Involvement of the interferon-regulated antiviral proteins PKR and RNase L in reovirus- induced shutoff of cellular translation. J. Virol. 79, 2240–2250.

Sommereyns, C., Paul, S., Staeheli, P., and Michiels, T. (2008). IFN-lambda (IFN-λ) is expressed in a tissue-dependent fashion and primarily acts on epithelial cells in vivo. PLoS Pathog. 4, e1000017.

Sorgeloos, F., Jha, B.K., Silverman, R.H., and Michiels, T. (2013). Evasion of antiviral innate immunity by Theiler’s virus L* Protein through direct inhibition of RNase L. PLoS Pathog. 9, e1003474.

Sorrentino, S. (2010). The eight human “canonical” ribonucleases: Molecular diversity, catalytic properties, and special biological actions of the enzyme proteins. FEBS Lett. 584, 2194–2200.

Sorrentino, S., and Glitz, D.G. (1991). Ribonuclease activity and substrate preference of human eosinophil cationic protein (ECP). FEBS Lett. 288, 23–26.

Sorrentino, S., Tucker, G.K., and Glitz, D.G. (1988). Purification and characterization of a ribonuclease from human liver *. J. Biol. Chem. 263, 16125–16131.

207

Sorrentino, S., Glitz, D.G., Hamann, K.J., Loegering, D.A., Checkel, J.L., and Gleich, G.J. (1992). Eosinophil-derived neurotoxin and human liver ribonuclease. Identity of structure and linkage of neurotoxicity to nuclease activity. J. Biol. Chem. 267, 14859– 14865.

Sorrentino, S., Naddeo, M., Russo, A., and D’Alessio, G. (2003). Degradation of double- stranded RNA by human pancreatic ribonuclease: crucial role of noncatalytic basic amino acid residues. Biochemistry 42, 10182–10190.

Soukup, G.A., and Breaker, R.R. (1999). Relationship between internucleotide linkage geometry and the stability of RNA . RNA 5,1308–1325.

Spencer, J.D., Schwaderer, A.L., Dirosario, J.D., McHugh, K.M., McGillivary, G., Justice, S.S., Carpenter, A.R., Baker, P.B., Harder, J., and Hains, D.S. (2011). Ribonuclease 7 is a potent antimicrobial peptide within the human urinary tract. Kidney Int. 80, 174–180.

Spiller, M.P., Boon, K.L., Reijns, M. A.M., and Beggs, J.D. (2007). The Lsm2-8 complex determines nuclear localization of the spliceosomal U6 snRNA. Nucleic Acids Res. 35, 923–929.

Squire, J., Zhou, A., Hassel, B.A., Nie, H., and Silverman, R.H. (1994). Localization of the interferon-induced, 2-5A-dependent RNase gene (RNS4) to human chromosome 1q25. Genomics 19, 174–175.

Srivastava, S., Verschoor, A., and Frank, J. (1992). Eukaryotic initiation factor 3 does not prevent association through physical blockage of the ribosomal subunit-subunit interface. J. Mol. Biol. 226, 301–304.

St Clair, D.K., Rybak, S.M., Riordan, J.F., and Vallee, B.L. (1987). Angiogenin abolishes cell-free protein synthesis by specific ribonucleolytic inactivation of ribosomes. Proc. Natl. Acad. Sci. U.S.A. 84, 8330–8334.

St Clair, D.K., Rybak, S.M., Riordan, J.F., and Vallee, B.L. (1988). Angiogenin abolishes cell-free protein synthesis by specific ribonucleolytic inactivation of 40S ribosomes. Biochemistry 27, 7263–7269.

Steitz, T.A., and Steitz, J.A. (1993). A general two-metal-ion mechanism for catalytic RNA. Proc. Natl. Acad. Sci. U.S.A. 90, 6498–6502.

Steitz, J.A., Berg, C., Hendrick, J.P., La Branche-Chabot, H., Metspalu, A., Rinke, J., and Yario, T. (1988). A 5S rRNA/L5 complex is a precursor to ribosome assembly in mammalian cells. J. Cell Biol. 106, 545–556.

Suhadolnik, R.J., Peterson, D.L., Brien, K.O., Cheney, P.R., Herst, C.V.T., Reichenbach, N.L., Kon, N., Horvath, S.E., Iacono, K.T., Adelson, M.E., De Meirleir, K., De Becker, P., Charubala, R., and Pfleiderer, W. (1997). Biochemical evidence for a novel low molecular weight 2-5A-dependent RNase L in chronic fatigue syndrome. J. Interferon Cytokine Res. 17, 377–385.

208

Sumpter Jr., R., Loo, Y.-M., Foy, E., Li, K., Yoneyama, M., Fujita, T., Lemon, S.M., and Gale Jr., M., (2005). Regulating intracellular antiviral defense and permissiveness to hepatitis C virus RNA replication through a cellular RNA helicase, RIG-I. J. Virol. 79, 2689-2699.

Tanaka, N., Meineke, B., and Shuman, S. (2011). RtcB, a novel RNA ligase, can catalyze tRNA splicing and HAC1 mRNA splicing in vivo. J. Biol. Chem. 286, 30253– 30257.

Tanji, H., Ohto, U., Shibata, T., Taoka, M., Yamauchi, Y., Isobe, T., Miyake, K., and Shimizu, T. (2015). Toll-like receptor 8 senses degradation products of single-stranded RNA. Nat. Struct. Mol. Biol. 22, 109–115.

Tardif, K.D., Mori, K., Kaufman, R.J., and Siddiqui, A. (2004). Hepatitis C virus suppresses the IRE1-XBP1 pathway of the unfolded protein response. J. Biol. Chem. 279, 17158–17164.

Tello-Montoliu, A., Patel, J. V., and Lip, G.Y.H. (2006). Angiogenin: a review of the pathophysiology and potential clinical applications. J. Thromb. Haemost. 4, 1864–1874.

Terzyan, S.S., Peracaula, R., de Llorens, R., Tsushima, Y., Yamada, H., Seno, M., Gomis-Rüth, F.X., and Coll, M. (1999). The three-dimensional structure of human RNase 4, unliganded and complexed with d(Up), reveals the basis for its uridine selectivity. J. Mol. Biol. 285, 205–214.

Thompson, D.M., and Parker, R. (2009). The RNase Rny1p cleaves tRNAs and promotes cell death during oxidative stress in Saccharomyces cerevisiae. J. Cell Biol. 185, 43–50.

Thompson, D.M., Lu, C., Green, P.J., and Parker, R. (2008). tRNA cleavage is a conserved response to oxidative stress in eukaryotes. RNA 14, 2095–2103.

Thompson, J.E., Venegas, F.D., and Raines, R.T. (1994). Energetics of catalysis by ribonucleases: fate of the 2ˊ, 3ˊ-cyclic phosphodiester intermediate. Biochemistry 33, 7408–7414.

Tournier, C., Hess, P., Yang, D.D., Xu, J., Turner, T.K., Nimnual, A., Bar-sagi, D., Jones, S.N., Flavell, R.A., and Davis, R.J. (2000). Requirement of JNK for stress-induced activation of cytochrome c-mediated death pathway. Science 288, 870–874.

Townsend, H.L., Jha, B.K., Han, J.-Q., Maluf, N.K., Silverman, R.H., and Barton, D.J. (2008a). A viral RNA competitively inhibits the antiviral endoribonuclease domain of RNase L. RNA 14, 1026–1036.

Townsend, H.L., Jha, B.K., Silverman, R.H., and Barton, D.J. (2008b). A putative loop E motif and an H-H kissing loop interaction are conserved and functional features in a group C enterovirus RNA that inhibits ribonuclease L. RNA Biol. 5, 263–272.

Trippe, R., Sandrock, B., and Benecke, B.J. (1998). A highly specific terminal uridylyl transferase modifies the 3ˊ-end of U6 small nuclear RNA. Nucleic Acids Res. 26, 3119– 3126. 209

Trippe, R., Richly, H., and Benecke, B.J. (2003). Biochemical characterization of a U6 small nuclear RNA-specific terminal uridylyltransferase. Eur. J. Biochem. 270, 971–980.

Trubia, M., Sessa, L., and Taramelli, R. (1997). Mammalian Rh/T2/S-glycoprotein ribonuclease family genes: cloning of a human member located in a region of chromosome 6 (6q27) frequently deleted in human malignancies. Genomics 42, 342– 344.

Tsuji, T., Sun, Y., Kishimoto, K., Olsen, K.A., Liu, S., Hirukawa, S., and Hu, G. (2005). Angiogenin Is translocated to the nucleus of HeLa cells and is involved in ribosomal RNA transcription and cell proliferation involved in ribosomal RNA transcription and cell proliferation. Cancer Res. 65, 1352–1360.

Vachon, V.K., Calderon, B.M., and Conn, G.L. (2014). A novel RNA molecular signature for activation of 2ˊ-5ˊ oligoadenylate synthetase-1. Nucleic Acids Res. 43, 544–552. van Hoof, A., Lennertz, P., and Parker, R. (2000). Three conserved members of the RNase D family have unique and overlapping functions in the processing of 5S, 5.8S, U4, U5, RNase MRP and RNase P RNAs in yeast. EMBO J. 19, 1357–1365.

Vester, D., Lagoda, A., Hoffmann, D., Seitz, C., Heldt, S., Bettenbrock, K., Genzel, Y., and Reichl, U. (2010). Real-time RT-qPCR assay for the analysis of human influenza A virus transcription and replication dynamics. J. Virol. Methods 168, 63–71.

Vignjevic, D., Schoumacher, M., Gavert, N., Janssen, K.P., Jih, G., Laé, M., Louvard, D., Ben-Ze’ev, A., and Robine, S. (2007). Fascin, a novel target of β-catenin-TCF signaling, is expressed at the invasive front of human colon cancer. Cancer Res. 67, 6844–6853.

Wahl, M.C., Will, C.L., and Lührmann, R. (2009). The Spliceosome: Design Principles of a Dynamic RNP Machine. Cell 136, 701–718.

Wakita, T., Pietschmann, T., Kato, T., Date, T., Miyamoto, M., Zhao, Z., Murthy, K., Habermann, A., Kräusslich, H.-G., Mizokami, M., Bartenschlager, R., and Liang, T.J.. (2005). Production of infectious hepatitis C virus in tissue culture from a cloned viral genome. Nat. Med. 11, 791–796.

Walkiewicz, M.P., Basu, D., Jablonski, J.J., Geysen, H.M., and Engel, D.A. (2011). Novel inhibitor of influenza non-structural protein 1 blocks multi-cycle replication in an RNase L-dependent manner. J. Gen. Virol. 92, 60–70.

Wang, Q., Lee, I., Ren, J., Ajay, S.S., Lee, Y.S., and Bao, X. (2013). Identification and functional characterization of tRNA-derived RNA fragments (tRFs) in respiratory syncytial virus infection. Mol. Ther. 21, 368–379.

Wang, S.Y., Iordanov, M., and Zhang, Q. (2006). c-Jun NH2-terminal kinase promotes apoptosis by down-regulating the transcriptional co-repressor CtBP. J. Biol. Chem. 281, 34810–34815.

210

Wang, Y., Ludwig, J., Schuberth, C., Goldeck, M., Schlee, M., Li, H., Juranek, S., Sheng, G., Micura, R., Tuschl, T., Hartmann, G., and Patel, D.J. (2010). Structural and functional insights into 5ˊ-ppp RNA pattern recognition by the innate immune receptor RIG-I. Nat. Struct. Mol. Biol. 17, 781–787.

Washenberger, C.L., Han, J., Kechris, K.J., Jha, B.K., Silverman, R.H., and Barton, D.J. (2007). Hepatitis C virus RNA: dinucleotide frequencies and cleavage by RNase L. Virus Res. 130, 85–95.

Weber, F., Wagner, V., Simon, B., Hartmann, R., Paludan, S.R., and Rasmussen, S.B. (2006). Double-stranded RNA is produced by positive-strand RNA viruses and DNA viruses but not in detectable amounts by negative-strand RNA viruses. J. Virol. 80, 5059–5064.

Weber, M., Gawanbacht, A., Habjan, M., Rang, A., Borner, C., Schmidt, A.M., Veitinger, S., Jacob, R., Devignot, S., Kochs, G., Garcia-Sastre, A., and Weber, F. (2013). Incoming RNA virus nucleocapsids containing a 5ˊ-triphosphorylated genome activate RIG-I and antiviral signaling. Cell Host Microbe 13, 336–346.

Weiner, H.L., Weiner, L.H., and Swain, J.L. (1987). Tissue distribution and developmental expression of the messenger RNA encoding angiogenin. Science 237, 280–282.

Wise, H.M., Hutchinson, E.C., Jagger, B.W., Stuart, A.D., Kang, Z.H., Robb, N., Schwartzman, L.M., Kash, J.C., Fodor, E., Firth, A.E., Gog, J.R., Taubenberger, J.K., and Digard, P. (2012). Identification of a novel splice variant form of the Influenza A virus M2 ion channel with an antigenically distinct ectodomain. PLoS Pathog. 8, e1002998.

Witzel, H., and Barnard, E.A. (1962). Mechanism and binding sites in the ribonuclease reaction II. kinetic studies on the first step of the reaction. Biochem. Biophys. Res. Commun. 7, 295–299.

Wreschner, D.H., James, T.C., Silverman, R.H., and Kerr, I.M. (1981a). Ribosomal RNA cleavage, nuclease activation and 2-5A (ppp(A2ˊp)nA) in interferon-treated cells. Nucleic Acids Res. 9, 1571–1581.

Wreschner, D.H., McCauley, J.W., Skehel, J.J., and Kerr, I.M. (1981b). Interferon action- -sequence specificity of the ppp(A2ˊp)nA-dependent ribonuclease. Nature 289, 414–417.

Wu, B., Peisley, A., Richards, C., Yao, H., Zeng, X., Lin, C., Chu, F., Walz, T., and Hur, S. (2013). Structural basis for dsRNA recognition, filament formation, and antiviral signal activation by MDA5. Cell 152, 276–289.

Wu, D., Yu, W., Kishikawa, H., Folkerth, R.D., Iafrate, A.J., Shen, Y., Xin, W., Sims, K., and Hu, G.F. (2007). Angiogenin loss-of-function mutations in amyotrophic lateral sclerosis. Ann. Neurol. 62, 609–617.

Wu, D., Chen, L., Liao, W., Ding, Y., Zhang, Q., Li, Z., and Liu, L. (2010). Fascin1 expression predicts poor prognosis in patients with nasopharyngeal carcinoma and correlates with tumor invasion. Ann. Oncol. 21, 589–596.

211

Xiang, Y., Wang, Z., Murakami, J., Plummer, S., Klein, E.A., Carpten, J.D., Trent, J.M., Isaacs, W.B., Casey, G., and Silverman, R.H. (2003). Effects of RNase L mutations associated with prostate cancer on apoptosis induced by 2ˊ, 5ˊ -oligoadenylates. Cancer Res. 63, 6795–6801.

Xu, Z., Tsuji, T., Riordan, J.F., and Hu, G. (2003). Identification and characterization of an angiogenin-binding DNA sequence that stimulates luciferase reporter gene expression. Biochemistry 42, 121–128.

Xu, Z.P., Tsuji, T., Riordan, J.F., and Hua, G.F. (2002). The nuclear function of angiogenin in endothelial cells is related to rRNA production. Biochem. Biophys. Res. Commun. 294, 287–292.

Yakovlev, G., Moiseyev, G.P., Sorrentino, S., De Prisco, R., and Libonati, M. (1997). Single-strand-preferring RNases degrade double-stranded RNAs by destabilizing its secondary structure. J. Biomol. Struct. Dyn. 15, 243–250.

Yamasaki, S., Ivanov, P., Hu, G.F., and Anderson, P. (2009). Angiogenin cleaves tRNA and promotes stress-induced translational repression. J. Cell Biol. 185, 35–42.

Yamashiro-Matsumura, S., and Matsumura, F. (1985). Purification and characterization of an f-actin-bundling 55-kilodalton protein from HeLa cells. J. Biol. Chem. 260, 5087– 5097.

Yang, W. (2011). Nucleases: diversity of structure, function and mechanism. Q Rev Biophys. 44, 1-93.

Yang, D., Chen, Q., Rosenberg, H.F., Rybak, S.M., Newton, D.L., Wang, Z.Y., Fu, Q., Tchernev, V.T., Wang, M., Schweitzer, B., Kingsmore, S.F., Patel, D.D., Oppenheim, J.J., and Howard, O.M. (2004). Human ribonuclease A superfamily members, eosinophil-derived neurotoxin and pancreatic ribonuclease, induce dendritic cell maturation and activation. J. Immunol. 173, 6134–6142.

Yang, W., Lee, J.Y., and Nowotny, M. (2006). Making and breaking nucleic acids: two- Mg2+-ion catalysis and substrate specificity. Mol. Cell 22, 5–13.

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

Yoneyama, M., Kikuchi, M., Natsukawa, T., Shinobu, N., Imaizumi, T., Miyagishi, M., Taira, K., Akira, S., and Fujita, T. (2004). The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral responses. Nat. Immunol. 5, 730–737.

Yoneyama, M., Onomoto, K., Jogi, M., Akaboshi, T., and Fujita, T. (2015). Viral RNA detection by RIG-I-like receptors. Curr. Opin. Immunol. 32, 48–53.

Yu, Y., Abaeva, I.S., Marintchev, A., Pestova, T. V., and Hellen, C.U.T. (2011). Common conformational changes induced in type 2 picornavirus IRESs by cognate trans-acting factors. Nucleic Acids Res. 39, 4851–4865. 212

Yuan, P., Bartlam, M., Lou, Z., Chen, S., Zhou, J., He, X., Lv, Z., Ge, R., Li, X., Deng, T., Fodor, E., Rao, Z., and Liu, Y. (2009). Crystal structure of an avian influenza polymerase PAN reveals an endonuclease active site. Nature 458, 909–913.

Yuan, X., Zhou, Y., Casanova, E., Chai, M., Kiss, E., Gröne, H.J., Schütz, G., and Grummt, I. (2005). Genetic inactivation of the transcription factor TIF-IA leads to nucleolar disruption, cell cycle arrest, and p53-mediated apoptosis. Mol. Cell 19, 77–87.

Zhang, J., Dyer, K.D., and Rosenberg, H.F. (2002). RNase 8, a novel RNase A superfamily ribonuclease expressed uniquely in placenta. Nucleic Acids Res. 30, 1169– 1175.

Zhang, R., Jha, B.K., Ogden, K.M., Dong, B., Zhao, L., Elliott, R., Patton, J.T., Silverman, R.H., and Weiss, S.R. (2013). Homologous 2 ˊ,5ˊ-phosphodiesterases from disparate RNA viruses antagonize antiviral innate immunity. Proc. Natl. Acad. Sci. U.S.A. 110, 13114–13119.

Zhao, L., Jha, B.K., Wu, A., Elliott, R., Ziebuhr, J., Gorbalenya, A.E., Silverman, R.H., and Weiss, S.R. (2012). Antagonism of the interferon-induced OAS-RNase L pathway by murine coronavirus ns2 protein is required for virus replication and liver pathology. Cell Host Microbe 11, 607–616.

Zhou, H.M., and Strydom, D.J. (1993a). The amino acid sequence of human ribonuclease 4, a highly conserved bonuclease that cleaves specifically on the 3ˊ side of uridine. Eur. J. Biochem. 217, 401–410.

Zhou, A., Hassel, B.A., and Silverman, R.H. (1993b). Expression cloning of 2-5A- dependent RNAase: a uniquely regulated mediator of interferon action. Cell 72, 753– 765.

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

Zhou, A., Molinaro, R.J., Malathi, K., and Silverman, R.H. (2005). Mapping of the human RNASEL promoter and expression in cancer and normal cells. J. Interferon Cytokine Res. 25, 595–603.

Zhou, H.-R., He, K., Landgraf, J., Pan, X., and Pestka, J. (2014). Direct activation of ribosome-associated double-stranded RNA-dependent protein kinase (PKR) by deoxynivalenol, anisomycin and ricin: a new model for ribotoxic stress response induction. Toxins 6, 3406–3425.

Zhou, Y., Kang, M.-J., Jha, B.K., Silverman, R.H., Lee, C.G., and Elias, J.A. (2013). Role of ribonuclease L in viral pathogen-associated molecular pattern/influenza virus and cigarette smoke-induced inflammation and remodeling. J. Immunol. 191, 2637–2646.

213

Zhou, Z., Hamming, O.J., Ank, N., Paludan, S.R., Nielsen, A.L., and Hartmann, R. (2007). Type III interferon (IFN) induces a type I IFN-like response in a restricted subset of cells through signaling pathways involving both the Jak-STAT pathway and the mitogen-activated protein kinases. J. Virol. 81, 7749–7758.

Zhu, H., Smith, P., Wang, L.K., and Shuman, S. (2007). Structure-function analysis of the 3ˊ phosphatase component of T4 polynucleotide kinase/phosphatase. Virology 366, 126–136.

Zhu, Y., Poyurovsky, M. V., Li, Y., Biderman, L., Stahl, J., Jacq, X., and Prives, C. (2009). Ribosomal protein S7 is both a regulator and a substrate of MDM2. Mol. Cell 35, 316–326.

Züst, R., Cervantes-Barragan, L., Habjan, M., Maier, R., Neuman, B.W., Ziebuhr, J., Szretter, K.J., Baker, S.C., Barchet, W., Diamond, M.S., Siddell, S.G., Ludewig, B., and Thiel, V. (2011). Ribose 2ˊ-O-methylation provides a molecular signature for the distinction of self and non-self mRNA dependent on the RNA sensor Mda5. Nat. Immunol. 12, 137–143.