Molecular Basis of the Dexh-Box RNA Helicase RNA Helicase a (RHA/DHX9) in Eukaryotic Protein Synthesis

Molecular basis of the DExH-box RNA helicase RNA helicase A (RHA/DHX9) in eukaryotic protein synthesis

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Sarah Elizabeth Fritz, B.S.

Graduate Program in Integrated Biomedical Science Program

The Ohio State University

2015

Dissertation Committee:

Dr. Kathleen Boris-Lawrie, Advisor

Dr. Joanna Groden

Dr. Jeffrey Parvin

Dr. Dawn Chandler

Sarah Elizabeth Fritz

2015

Abstract

Protein synthesis is a fundamental molecular event of cell biology. Its enactment requires the intricate choreographing of RNA-protein assemblies. These ribonucleoprotein (RNP) complexes are the performers of translation that function to regulate ribosome dynamics upon a target transcript. The outcome is distinct regulation of protein production. Deregulated translation activity is associated with cancer, neurological diseases and disorders, neurodegeneration, growth defects, and innate immune disorders. This dissertation aimed to enhance our understanding of the eukaryotic translation process by elucidating the molecular role of RNA helicase A

(RHA/DHX9) in protein synthesis.

RHA is a cellular protein and member of the DExH/D-box RNA helicase family that is necessary for the translation of pathogenic retroviral transcripts and the cellular proto-oncogene junD. The known characteristic of RHA-mediated protein synthesis is the select recognition and association of RHA with the distinguishing 5' termini of retroviral and junD transcripts. These mRNAs harbor the distinct 5' RNA motif known as the posttranscriptional control element (PCE). The PCE functions in cis to stimulate translation activity via the canonical cap-dependent scanning mechanism that defines eukaryotic protein synthesis. The main gap in knowledge is how RHA, as the host effector of the PCE, engages a RNP complex that facilitates translation activity. ii

In this study, RHA was identified to engage a unique RNP that confers a novel role for its activity in cap-dependent translation during cell stress. Here RHA was demonstrated to selectively interact with the non-canonical CBP80/20 cap-binding protein in the cell cytoplasm and on polysomes. Notably, this interaction was maintained during serum deprivation, torin 1-mediated mTOR inhibition and HIV-1 expression, all mechanisms that evoke cell stress and suppression of the canonical eIF4E cap-dependent translation. This observed effect correlated with sustained interactions between RHA,

CBP80/20, and the target PCE transcript HIV-1 gag on polysomes. The outcome was maintained cap-dependent retroviral PCE translation. This novel molecular finding proposed a new paradigm for eukaryotic translation control in that the RHA-CBP80/20

RNP maintains selective cap-dependent protein synthesis during cell stress.

This dissertation also identified novel significance for RHA as a post-initiation effector of protein synthesis. DExH/D-box RNA helicases are known to mediate the initial translation events of 5' ribosome binding and scanning. Through a reverse genetics approach, the N-terminal double-stranded RNA binding domains of RHA were shown to be critical for these canonical family activities. Notably, a role for the C-terminal arginine-glycine-rich domain of RHA was identified in 80S ribosome stabilization. This post-initiation translation activity was fundamental to the engagement of RHA with polyribosomes and completion of the translation process. These outcomes proposed a new paradigm for the molecular act of DExH/D-box RNA helicases in translation control that encompass both initiation and post-initiation translation activity.

iii

A third major finding of this dissertation was the identification of RHA to engage multiple RNP states that regulate its translation activity. Here RHA was shown to self- associate, a homopolymeric binding event that impaired translation cofactor binding. This result indicated a role for RHA self-association in the regulation of its RNP formation that facilitates translation activity. In addition, a select interaction between RHA and related DExH-box RNA helicase DHX30 was identified. This heteropolymeric-binding event was distinct in that it characterized a RHA RNP engaged with polyribosomes. This result indicated a role for a RHA-DHX30 association in the direct regulation of RHA translation during active protein synthesis. Together, these findings proposed novel significance for RHA RNP dynamics in the control of its translation activity.

Collectively, the data obtained from this dissertation research provided novel and significant insight into the eukaryotic translation process. By elucidating the molecular basis of RHA in protein synthesis, several new paradigms were identified for the eukaryotic translation process. These include: a molecular basis for maintained cap- dependent translation during cell stress, significance of DExH/D-box RNA helicases in post-initiation translation control, and a role for distinct homo- and hetero-polymeric binding events in the regulation of RNP dynamics that control the translation process. We posit that these findings afford molecular significance for the association of RHA with breast, prostate and lung cancer, its identification as the major auto-antigen in systemic lupus erythematosus patients, and the role for RHA as a major stimulator of pathogenic viruses that infect and cause disease within animal and human hosts. Future studies are aimed to connect the molecular findings of this dissertation research with the clinical

iv significance of RHA. The objective is to provide a greater understanding of the relationship between RHA, eukaryotic protein synthesis, and cell biology that informs animal and human health and disease.

Dedication

Dedicated to my grandmother and forever kindred spirit, Mary Louise Fritz

Acknowledgments

To my advisor, Dr. Kathleen Boris-Lawrie, thank you for being my greatest advocate and supporter. You always believed in me when I doubted myself, a characteristic of a true mentor.

To all members of the Boris-Lawrie lab, both past and present, I appreciate all of the help, guidance, and support you provided. Each of you went above and beyond to offer assistance at any moment, and I very much appreciated it. Arnaz, Marcela, Amit,

Amy, Justin, Jon, Alaina, Ioana, Gati, Allison, Greeshma and all others that I have worked alongside, thank you for being incredible colleagues and friends.

To my committee members, Dr. Joanna Groden, Dr. Jeffrey Parvin, and Dr. Dawn

Chandler, thank you for working together with Dr. Boris-Lawrie and I to make this moment possible. Your support and encouragement made the entire Ph.D. journey a rewarding experience.

To all colleagues in the Center for Retrovirus Research, you are incredible scientists, people, and friends. It has been a pleasure working alongside each and every one of you. Amanda, Jacob, Krissy, Cory, Alice, Nathan, Corine, Suresh, Jenna, Fei Fei,

Becca, and Raj, keep being your amazing selves. And a special thanks to Tim Vojt for believing in our message and working very efficiently to help create the amazing figures that provided us that voice.

vii

To the Gettysburg College Biology Department, I am forever indebted to your continuous support and encouragement. Your passion for science and commitment to inspiring your students is, in itself, a true inspiration. I aspire to follow in your footsteps.

To the Columbus Running Company, thank you for making Columbus home.

Star, Stephanie, Caroline, Katie, Melissa, Paige, Liz T., Liz M., Beth, Anna, Mary, Kerri,

Megan, Tayler, and all the many other incredible individuals I have shared runs with over these past five years, you are the greatest of training partners, role models, and friends.

Together, you make every marathon possible.

To my fellow Buckeyes, Julia, Hans, Adam, and Christian, and the amazing people you brought into my life, Matt, Isabel, Angela, and Ashley, our days in Ohio will always be ones that I treasure. It has been a true honor to go through this Ph.D. process with you and the experience would not have been the same or as meaningful without all of you as a part of it.

To my former Gettysburg teammates, thank you for continuing to be the greatest of role models, supporters, and friends. Especially to my dear friend Kerrin, I would not be the person I am today without your friendship, support, and encouragement.

And lastly, to my parents, thank you for letting me become the woman I am today. You never question nor do you ever doubt any of my decisions. You simply encourage me to choose my own path. There is nothing more I could ever ask for. I am eternally grateful.

viii

Vita

2006...... Ledyard High School

2010 ...... B.S. Biochemistry and Molecular Biology,

Gettysburg College

2010 to present ...... Ph.D. candidate, Integrated Biomedical

Sciences, The Ohio State University

Fields of Study

Major Field: Integrated Biomedical Science Program

Translation control

RNA biology

Molecular virology

Table of Contents

Abstract ...... ii

Dedication ...... vi

Acknowledgments ...... vii

Vita ...... ix

List of Tables ...... xiii

List of Figures ...... xiv

Chapter 1: The trRNPs of Eukaryotic Translation Control ...... 1

Overview of Eukaryotic Translation ...... 2

Initiation, mRNA activation and ribosome recruitment ...... 2

Initiation, the significance of DExH/D-box RNA helicases ...... 3

Initiation, scanning and start codon recognition ...... 13

Initiation, 60S subunit joining and formation of elongation-competent 80S ribosomes

...... 16

Initiation, variations to the central theme ...... 18

Elongation ...... 22

Termination ...... 26

Regulation of eukaryotic translation ...... 30

CBC and eIF4E: core trRNPs of eukaryotic translation control...... 38

The CBC trRNP ...... 39

The eIF4E trRNP ...... 43

Regulation of the CBC and eIF4E trRNPs ...... 46

The RHA trRNP ...... 48

Chapter 2: Selective Cap-Dependent Translation During Cell Stress is Maintained by the RNA Helicase A-CBP80/20 Translation RNP ...... 63

ABSTRACT ...... 63

INTRODUCTION ...... 64

MATERIALS AND METHODS ...... 68

RESULTS...... 74

DISCUSSION ...... 90

Chapter 3: Conserved Domains of RNA Helicase A Contribute Genetically

Separable Roles to Its Function in Cap-Dependent Translation Control ...... 109

ABSTRACT ...... 109

INTRODUCTION ...... 110

MATERIALS AND METHODS ...... 113

RESULTS...... 119

DISCUSSION ...... 141

Chapter 4: The Terminal Domains of RNA Helicase A Coordinate Its Cellular

Multimerization and Select Interaction with Related DHX30: Implications for

Regulated Translation Activity ...... 156

ABSTRACT ...... 156

INTRODUCTION ...... 157

MATERIALS AND METHODS ...... 162

RESULTS...... 166

DISCUSSION ...... 179

Chapter 5: Perspectives ...... 195

Cap-dependent translation during cell stress ...... 196

DExH/D-box RNA helicases with significance in post-initiation translation control

...... 202

Protein synthesis is fundamental to the survival of all cells and viruses ...... 206

RNA helicase A is a host protein with clinical significance ...... 208

Conclusion ...... 212

References ...... 213

xii

List of Tables

Table 1.1 Known features of select DExH/D-box RNA helicase trRNPs ...... 62

xiii

List of Figures

Figure 1.1 Schematic of canonical eukaryotic translation initiation...... 52

Figure 1.2 Schematic of canonical eukaryotic translation elongation ...... 54

Figure 1.3 Schematic of canonical eukaryotic translation termination...... 56

Figure 1.4 Overview of mTOR-directed eukaryotic translation control...... 57

Figure 1.5 Regulation of 4E-BP1 phosphorylation events and its role in eukaryotic translation control ...... 58

Figure 1.6 Integration of stress response signals into the phosphorylation of eIF2α and its outcome on regulated eukaryotic protein synthesis ...... 59

Figure 1.7 Model of CBC and eIF4E trRNP dynamics in the control of eukaryotic translation ...... 61

Figure 2.1 Endogenous RHA selectively interacts with CBP80 in the cytoplasm ...... 97

Figure 2.2 Epitope-tagged RHA and CBP80 demonstrate a select cytoplasmic association

...... 98

Figure 2.3 RHA interacts with the MIF4G domain-containing protein CTIF ...... 99

Figure 2.4 RHA-CBP80/20 is an active trRNP ...... 100

Figure 2.5 RHA and CBP80 maintain a cytoplasmic interaction and polysome association during serum deprivation ...... 102

Figure 2.6. RHA and CBP80 maintain a cytoplasmic interaction and polysome association during torin 1-mediated mTOR inhibition and HIV-1 expression ...... 103

Figure 2.7 Cytoplasmic and polysome-associated RHA and CBP80 bind the HIV-1 gag mRNA during cell stress ...... 105 xiv

Figure 2.8 Retroviral PCE translation is maintained in a cap-dependent manner during cell stress...... 106

Figure 2.9 Model of the canonical and RHA-CBP80/20 trRNP pathways and their functional significance for cap-dependent translation during cell stress ...... 108

Figure 3.1 Depiction of RHA domain structure with critical amino acids indicated ...... 147

Figure 3.2 N-terminal lysine residue 236 is necessary for a RHA-CBP80 association and RHA trRNP formation ...... 148

Figure 3.3 Mutation of the C-terminal RGG domain impacts RHA polysome association in a manner distinct from compromised N-terminal function ...... 149

Figure 3.4 The C-terminal arginine residues of RHA are critical for stable sedimentation of the RHA trRNP through a sucrose gradient ...... 150

Figure 3.5 Ribosomal run-off assays support a role for RHA's C-terminal RGG domain in initiation complex stabilization ...... 152

Figure 3.6 In vivo translation initiation assays demonstrate significance for RHA in early and late-stage initiation events ...... 154

Figure 3.7 Model of the molecular role of RHA in eukaryotic translation...... 155

Figure 4.1 RHA multimerizes in cells ...... 186

Figure 4.2 RHA multimers are coordinated by N- and C-terminal interactions ...... 187

Figure 4.3 RHA multimers compete with translation cofactor binding...... 189

Figure 4.4 Depiction of known DExH/D-box RNA helicase domain structures with critical roles in targeted translation control ...... 190

Figure 4.5 RHA and DHX30 exhibit a select association in cells ...... 191

Figure 4.6 N-terminus of RHA is necessary and sufficient for a DHX30 association; helicase activity of DHX30 is dispensable for this interaction ...... 192

Figure 4.7 DHX30 is identified in polysome-associated RHA trRNP complexes ...... 193

Figure 4.8 Model for regulated RHA translation activity ...... 194

Chapter 1 : The trRNPs of Eukaryotic Translation Control

Protein synthesis is the essential cellular process of translating the genetic code into the major structural and functional biomolecule of the cell: the protein. Eukaryotic translation is a dynamic molecular process choreographed by translation ribonucleoprotein (trRNP) complexes that assemble upon a messenger RNA (mRNA) and regulate its interaction with the bimolecular catalyst of this event, the ribosome.

From their transcription, mRNAs assemble within dynamic RNA-protein structures that regulate their stability, processing, localization, and ultimately their translation (Moore,

2005; G. Singh, Pratt, Yeo, & Moore, 2015). Ribonucleoprotein (RNP) complex is a term used to broadly define these RNA-protein assemblies (Mitchell & Parker, 2014). trRNP complexes specifically relate to dynamic mRNA-protein structures that coordinate the translation process. trRNPs are at the core of eukaryotic translation control. They are endowed with the ability to regulate the localization, conformation, and activation of mRNAs, all features that regulate engagement with the ribosome and generation of protein product.

The introduction to this dissertation provides an overview of the trRNPs of eukaryotic translation control. Mechanisms known for regulating eukaryotic trRNP activity will be discussed with reference to their significance in cell biology. The

1 importance of distinct trRNPs in selective translation control will be highlighted with a specific focus on the DExH/D-box RNA helicase trRNPs. The outcome of this introductory chapter is to inform the work of this dissertation, which presents the identification and characterization of the RNA helicase A (RHA) trRNP as a novel trRNP that maintains cap-dependent translation of select mRNAs during cell stress.

Overview of Eukaryotic Translation

Eukaryotic translation occurs in three mechanistic stages: initiation, elongation, and termination. Each stage is coordinated by a set of translation factors, which are RNA binding and/or scaffolding proteins that assemble upon an mRNA and regulate its engagement with the ribosome. These RNA-protein complexes give rise to the trRNPs of eukaryotic translation control. The integration of distinct RNA binding proteins, such as

DExH/D-box RNA helicases, into this process creates selective trRNPs important for targeted protein synthesis.

Initiation, mRNA activation and ribosome recruitment

Initiation is the rate-limiting step of protein synthesis with numerous regulatory mechanisms identified to control its activation and efficiency. The defining principle of eukaryotic translation initiation is a cap-dependent scanning mechanism whereby the ribosome binds the 5' terminus of an mRNA and proceeds along the transcript inspecting, base by base, for an appropriate start codon to initiate polypeptide synthesis (Jackson,

Hellen, & Pestova, 2010). This process requires at least nine initiation factors (eIFs) and

2 begins with the activation of an mRNA and its association with the 43S pre-initiation ribosome complex (Jackson et al., 2010) (Figure 1.1).

Eukaryotic mRNAs are defined by a 5' 7-methyl guanosine cap, which is bound co-transcriptionally and throughout the lifespan of an mRNA by a cap-binding protein.

The cap-binding protein, CBP80/20 or eIF4E, and its associated scaffolding factor, CTIF and/or eIF4G, generate a trRNP that primes an mRNA for translation (discussed further in section, 'CBC and eIF4E: core tRNPs of eukaryotic translation control'). This activation encompasses circularization of the mRNA via an interaction between the 5' scaffolding factor and the 3'-bound poly-A binding protein (PABP) to result in recruitment of the 43S pre-initiation ribosome complex to the 5' terminus of a transcript

(Hinnebusch, 2014; Hinnebusch & Lorsch, 2012) (Figure 1.1). The 43S pre-initiation ribosome complex consists of the small (40S) ribosomal subunit, initiator methionyl- tRNA, and five core initiation factors (eIF1, 1A, 2, 3, and 5) that each has a distinct role in the recruitment and scanning process (Jackson et al., 2010). Interactions between eIF3 of the 43S pre-initiation ribosome complex and the 5' mRNA-bound scaffolding factor,

CTIF and/or eIF4G, drive association of the 43S pre-initiation ribosome complex with the activated mRNA (Choe et al., 2012; Hinnebusch, 2014; Hinnebusch & Lorsch, 2012)

(Figure 1.1).

Initiation, the significance of DExH/D-box RNA helicases

Intricate to mRNA activation and ribosome recruitment are DExH/D-box RNA helicases. DExH/D-box RNA helicases are both RNA binding proteins and enzymatic

3 catalysts that couple the free energy of nucleotide triphosphate hydrolysis to RNA unwinding and/or RNP remodeling (Cordin, Banroques, Tanner, & Linder, 2006;

Jankowsky & Bowers, 2006; Linder, 2006; Rocak & Linder, 2004). The outcome is dynamic structural rearrangements within the trRNP that regulate placement and then movement of the 43S pre-initiation ribosome complex along the 5' terminus of an mRNA

(Parsyan et al., 2011).

Over fifty DExH/D-box RNA helicases have been identified with critical roles in posttranscriptional gene control. They are recognized by the presence of a core helicase domain within their protein structure that harbors nine conserved motifs involved in RNA binding, nucleotide triphosphate association, and its hydrolysis (Cordin et al., 2006;

Jankowsky & Bowers, 2006; Linder, 2006; Rocak & Linder, 2004). Motif II is one of these nine conserved motifs that lies within the center of the helicase domain and is critical for association with the β- and γ- phosphates of a nucleotide triphosphate, interactions that are facilitated by core Asp-Glu-x-His/Asp (DExH/D) residues for which

DExH/D-box RNA helicases derive their name (Cordin et al., 2006; Jankowsky &

Bowers, 2006; Linder, 2006; Rocak & Linder, 2004). Although conservation of sequence and function are critical for the classification of DExH/D-box RNA helicases, variations in the defining helicase core along with differences in flanking domains results in distinct trRNP association and function for each DExH/D-box RNA helicase (Table 1.1).

eIF4A (DDX2) is the DEAD-box RNA helicase most understood for regulating translation of eukaryotic mRNAs (Parsyan et al., 2011). Functioning as an ATP- dependent RNA binding protein, eIF4A separates localized duplexed strands by coupling

ATP-driven enzymatic changes to RNA unwinding (Parsyan et al., 2011). Its associations with eIF4G and the initiation factors eIF4B and eIF4H coordinate this activity near the 5' cap of an mRNA (Parsyan et al., 2011). eIF4G, by virtue of its interaction with the cap- binding protein eIF4E and a target transcript, recruits eIF4A to the 5' terminus of an activated mRNA (Hinnebusch & Lorsch, 2012; Jackson et al., 2010). The interaction of eIF4G with the helicase domain of eIF4A stabilizes eIF4A in a closed, RNA-bound conformation that stimulates its helicase activity (Parsyan et al., 2011). This effect is also facilitated by eIF4B and eIF4H, which function to regulate the affinity of eIF4A for ATP and ADP during its ATPase-driven RNA remodeling as well as its association with a target mRNA (Parsyan et al., 2011). eIF4B and eIF4H also have critical roles in preventing mRNA refolding and for directing the movement of the 43S pre-initiation ribosome complex in a 5' to 3' direction (Parsyan et al., 2011).

The necessity of eIF4A for translation initiation of cellular mRNAs harboring 5' secondary structure, which characterizes many eukaryotic transcripts, together with the impact of its mutation on global translation indicated eIF4A as a universal effector of protein synthesis (Svitkin et al., 2001). This established eIF4A, together with eIF4G, as an integral member of the eIF4E trRNP complex regulating steady-state expression of the majority of cellular transcripts (discussed further in section, 'CBC and eIF4E: core tRNPs of eukaryotic translation control'). However, a recent application of ribosome profiling to the study of eIF4A-dependent translation control revealed paradigm-shifting specificity to eIF4A-mediated protein synthesis (Wolfe et al., 2014).

Ribosome profiling is a recently introduced technique that has already revolutionized the study and understanding of eukaryotic translation control by allowing for an in vivo, global characterization of translational landscapes, such as that of eIF4A- dependent translation control (Ingolia, 2014; Michel & Baranov, 2013). By combining polysome profiling with nuclease footprinting and deep sequencing, this experimental approach provides high-resolution data on both ribosome abundance and positional occupancy along mRNAs in live cells (Ingolia, 2014; Michel & Baranov, 2013). The outcome is two-fold: (1) the identification of mRNAs subjected to regulation within a translational landscape based upon a change in their number of ribosome footprints between control and experimental conditions and (2) a molecular basis for this effect based upon the accompanied accumulation in ribosome footprint reads at particular positions along the target mRNAs (Ingolia, 2014). Distinct patterns that emerge from these results inform subsequent analyses that allow for identification of molecular signatures defining a particular translation landscape.

In the case of eIF4A, treatment of cells with silvestrol, a direct inhibitor of eIF4A helicase activity (Bhat et al., 2015), revealed that eIF4A-sensitive transcripts are specifically defined by a 12 or 9-nucleotide (CGG)4 motif within their 5' termini (Wolfe et al., 2014). This (CGG)4 motif forms a stable G-quadruplex secondary structure, sensitizing mRNAs to eIF4A helicase activity and identifying the eIF4A translational landscape (Wolfe et al., 2014). This eIF4A translational landscape encompasses many prominent transcriptional regulators, including several super-enhancers (Wolfe et al.,

2014), providing a molecular basis for the vast expression reprogramming observed upon

6 mutation of eIF4A even given its newly identified specificity in translation control. Thus, the significance of eIF4A in 43S pre-initiation ribosome complex recruitment and scanning is specific for a subset of mRNAs harboring its cognate cis-acting RNA element, the (CGG)4 motif.

The repertoire of cis-acting structural elements found within the 5' termini of eukaryotic mRNAs is diverse and extends beyond G-quadruplex structures to include various stem and internal loops, bulges, junctions, and pseudoknots. The DExH/D-box

RNA helicases eIF4AIII (DDX48), DDX3, DHX29, and RNA helicase A (RHA/DHX9) are intricate to the molecular basis by which these distinct cis-acting structural elements mediate 43S pre-initiation ribosome complex recruitment and scanning. Similar to eIF4A, the selectivity by which eIF4AIII, DDX3, DHX29, or RHA exert translational control is governed by distinct interactions between a defining cis-acting RNA element within the

5' termini of an mRNA and its cognate DExH/D-box RNA helicase. Together these select

RNA-protein interactions generate distinct trRNPs important for targeted protein synthesis (Table 1.1). As discussed below, the discrete composition and architecture of each DExH/D-box RNA helicase trRNP confers targeted roles for eIF4AIII, DDX3,

DHX29, and RHA in ribosome recruitment and scanning that extends beyond basic mRNA unwinding to encompass dynamic trRNP remodeling as well.

eIF4AIII is a known component of exon junction complexes, which are dynamic protein assemblies that organize upon the coding sequence of a mRNA with the conclusion of a splicing event to facilitate critical posttranscriptional activities. A recent study, however, revealed significance for eIF4AIII as a 5' interacting factor necessary for

7 translation initiation of mRNAs harboring stable stem loop structures within their 5'

UTRs (Choe, Ryu, Park, Park, et al., 2014b). The recruitment of eIF4AIII to the 5' terminus of a target transcript is facilitated by its direct interaction with the 5' scaffolding factor CTIF and is independent from its structural and functional integration within exon junction complexes (Choe, Ryu, Park, Park, et al., 2014b). The ATPase/helicase function of eIF4AIII is required for its stimulatory effect on translation initiation, indicating a role for this DEAD-box RNA helicase in resolving the stem-loop secondary structures to facilitate 43S pre-initiation ribosome complex association and scanning on target transcripts (Choe, Ryu, Park, Park, et al., 2014b). However, the mRNA features and/or protein cofactors that contribute to the specificity of the eIF4AIII trRNP in targeted 5'- cap-dependent translation control remain to be elucidated.

For DDX3, its functions in translation initiation are diverse and transcript- specific. In the case of cellular mRNAs with long and highly structured 5' UTRs, DDX3 is critical for translation activity in a manner that requires its helicase function and interaction with eIF4A and PABP of an activated mRNA as well as eIF2 and eIF3 of the

43S pre-initiation ribosome complex (Lai, Chang, Shieh, & Tarn, 2010; Lai, Lee, & Tarn,

2008; C.-S. Lee et al., 2008). These identified cellular interactions indicate significance for DDX3 in functioning synergistically with eIF4A to resolve extensive secondary structure in the 5' terminus, allowing for ribosome association and scanning.

Similarly, DDX3 is necessary for resolving the complex secondary structure directly near the 5' cap of the human immunodeficiency virus type 1 (HIV-1) genomic

RNA (Soto-Rifo et al., 2012). The outcome is productive translation and generation of

8 the major viral structural protein Gag (Soto-Rifo et al., 2012). Distinctly, however, this requirement of DDX3 for targeted HIV-1 translation control is independent of its helicase activity (Soto-Rifo et al., 2012). Instead, it involves the association of DDX3 with eIF4G,

PABP, and the viral protein Tat to generate a DDX3 trRNP that activates the target HIV-

1 genomic RNA for 43S pre-initiation ribosome complex attachment and subsequent protein synthesis (Lai et al., 2013; Soto-Rifo et al., 2012; Soto-Rifo, Rubilar, & Ohlmann,

2013).

DDX3 also functions in late-stage initiation with targeted 80S ribosome formation

(Geissler, Golbik, & Behrens, 2012). Here DDX3 coordinates a protein bridge or conformational rearrangement within the 60S ribosomal subunit that facilitates its recruitment to and association with the 40S ribosomal subunit at the start AUG to generate an elongation-competent 80S ribosome (Geissler et al., 2012). This effect is critical for expression of hepatitis C virus in a manner that is independent of cap- dependent scanning activity (Geissler et al., 2012).

Experimental data also indicate a role for select DDX3 trRNPs in inhibiting cellular cap-dependent initiation by functioning as a competitive eIF4E-binding protein and translational inhibitor (Shih, Tsai, Chao, & Wu Lee, 2008). DDX3 demonstrates a strong cellular affinity for eIF4E, which impedes its association with eIF4G (Shih et al.,

2008). The outcome is impaired eIF4E trRNP formation and repressed translation activity

(Shih et al., 2008). This effect is independent of DDX3's ATPase or helicase activity

(Shih et al., 2008). However, the cellular association of DDX3 with eIF4E and its significance for regulating protein synthesis remains controversial (C.-S. Lee et al., 2008;

Soto-Rifo et al., 2012; 2013). Likewise, the mRNA signatures and/or protein factors that dictate the specificity, both in composition and function, of the DDX3 trRNP in targeted translation control remain ambiguous.

The DExH-box RNA helicase DHX29 is critical for generating a trRNP that facilitates expression of cellular mRNAs harboring stable stem-loop structures within their 5' termini (Pisareva, Pisarev, Komar, Hellen, & Pestova, 2008). DHX29 associates with the 40S ribosomal subunit within a 43S pre-initiation complex (Hashem et al.,

2013). It is positioned near the mRNA entry channel latch where it stimulates its opening to capture single-stranded bases that have been released from resolution of nearby stem- loop structure (Dhote, Sweeney, Kim, Hellen, & Pestova, 2012; Hashem et al., 2013;

Pisareva et al., 2008). This activity of DHX29 in conjunction with the known roles of eIF4A in mRNA unwinding collaboratively facilitate association and scanning of the 43S pre-initiation ribosome complex along a target mRNA (Pisareva et al., 2008). The outcome is ensured fidelity of base inspection and start codon recognition (Abaeva,

Marintchev, Pisareva, Hellen, & Pestova, 2011; Pisareva et al., 2008). Thus, DHX29 drives critical trRNP remodeling rather than traditional mRNA unwinding to facilitate efficient and appropriate translation initiation (Abaeva et al., 2011; Hashem et al., 2013).

The significance of this translation activity is seen in the silencing of DHX29 expression, which results in suppression of cellular translation by approximately fifty percent

(Parsyan et al., 2009). However, identification of the molecular signatures, both mRNA sequence motifs and protein cofactors, which characterize the specific DHX29 trRNP and its translational landscape, have yet to be defined.

DHX9/RNA helicase A (RHA) is critical for facilitating the expression of retroviral transcripts, including the human pathogenic retroviruses HIV-1 and HTLV-1, as well as the cellular proto-oncogene junD (Bolinger, Sharma, Singh, Yu, & Boris-

Lawrie, 2010; Bolinger et al., 2007; Hartman et al., 2006; Ranji, Shkriabai,

Kvaratskhelia, Musier-Forsyth, & Boris-Lawrie, 2011). Here this DEIH-box helicase selectively associates with the 5' cis-acting posttranscriptional control element (PCE) that defines these target mRNAs (Bolinger et al., 2007; 2010; Hartman et al., 2006; Ranji et al., 2011). The PCE is a G-C-rich, dual stem-loop structure that specifically regulates cap-dependent translation of harboring transcripts (Bolinger et al., 2007; Butsch, Hull,

Wang, Roberts, & Boris-Lawrie, 1999; Roberts & Boris-Lawrie, 2000; 2003). Notably, the PCE functions as a positive cis-acting translational regulator (Butsch et al., 1999).

This effect is mediated through its association with the trans-acting host factor RHA

(Hartman et al., 2006). Suppression of cellular RHA expression reduces PCE translation activity, the association of target mRNAs with ribosomes, and subsequent protein production (Bolinger et al., 2007; 2010; Hartman et al., 2006). The outcome, as studied thus far, is impaired infectivity of HIV-1 progeny virions (Bolinger et al., 2010).

The role of RHA in cap-dependent translation control requires its association with the 5' terminal PCE as well as its ATPase activity, indicating canonical helicase function in its molecular basis of protein synthesis (Bolinger et al., 2010; Ranji et al., 2011).

However, RHA can also function in translation control independent of PCE binding. Here

RHA is recruited to a target mRNA through association with a cognate RNA binding protein. Such is the case for regulated expression of the pluripotency factor Oct4 (Jin et

11 al., 2011). The RNA binding protein Lin28 recruits RHA to the Oct4 mRNA, which in turn facilitates ribosome association of the trRNP and subsequent protein expression (Jin et al., 2011). A similar molecular basis for RHA-mediated translation control is critical for regulated expression of the type I collagen mRNAs (Manojlovic & Stefanovic, 2012).

The RNA binding protein La ribonucleoprotein domain family member 6 specifically recognizes and binds the defining 5' stem-loop structure of type I collagen mRNAs to bridge an association with RHA, stimulating translation activity (Manojlovic &

Stefanovic, 2012). Likewise, nuclear factor 90 bound to the 5' terminus of p53 recruits

RHA to this target mRNA and facilitates its expression in response to DNA damage

(“Translational Control Protein 80 Stimulates IRES-Mediated Translation of p53 mRNA in Response to DNA Damage,” n.d.). It is proposed that the recruitment of RHA to a target transcript, whether it be through direct PCE RNA binding or a protein bridge, facilitates trRNP remodeling in a manner that allows for 43S pre-initiation ribosome complex recruitment, scanning, and ultimately target protein production (Hartman et al.,

2006; Jin et al., 2011; Manojlovic & Stefanovic, 2012). An unresolved issue, however, is a complete characterization of the RHA trRNP and how this informs its role in targeted translation control (discussed further in section, 'The RHA trRNP').

Collectively, the DExH/D-box RNA helicase family harbors several members with critical roles in 43S pre-initiation ribosome complex recruitment. Yet, as discussed, the functional significances for each DExH/D-box RNA helicase in this process are distinct (mRNA unwinding versus trRNP remodeling) to result in select trRNPs, like the

RHA trRNP, that afford the cell novel and diverse mechanisms for controlling the

12 initiation of protein synthesis. Irrespective, an outstanding question across all mechanisms of 43S pre-initiation ribosome complex recruitment is the manner by which the mRNA becomes positioned within the binding channel of the 40S ribosomal subunit and the location at which scanning-dependent base inspection begins (Jackson et al.,

2010). Is the mRNA threaded through its binding channel in the 40S ribosomal subunit or does the 43S pre-initiation ribosome complex undergo direct positioning near the 5' terminus (Jackson et al., 2010)? How close to the 5' cap does the 43S pre-initiation ribosome complex need to bind in order to engage appropriate scanning in the correct reading frame (Jackson et al., 2010)? It is likely that the molecular bases for these effects are distinct for each class of transcripts and mediated by the particular mode of action of the associated DExH/D-box RNA helicase.

Initiation, scanning and start codon recognition

Once associated with a target mRNA, the 43S pre-initiation ribosome complex begins the scanning process in search of an appropriate start codon to initiate polypeptide synthesis. eIFs 1 and 1A are critical for facilitating this process. The positioning of eIFs 1 and 1A within the 43S pre-initiation ribosome complex confers structural arrangements that stabilize an 'open' conformation of the mRNA entry channel latch within the 40S ribosomal subunit (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012; Jackson et al., 2010)

(Figure 1.1). This places the 40S ribosomal subunit and associated initiator methionyl- tRNA in a conformation that is conducive for movement along a mRNA and base-by- base inspection (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012; Jackson et al., 2010).

Appropriate anticodon:codon base pairing between the initiator methionyl-tRNA and a start AUG triggers a conformational rearrangement in the 43S pre-initiation ribosome complex that causes the mRNA entry channel latch to close, stopping the scanning process (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012; Jackson et al., 2010) (Figure

1.1). This repositions the initiator methionyl-tRNA and 40S ribosomal subunit in a favorable context for joining of the large (60S) ribosomal subunit and formation of elongation-competent 80S ribosomes (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012;

Jackson et al., 2010).

Interactions between mRNA sequence, eIF2, and the 18S rRNA facilitate favorable anticodon:codon base pairing that halts ribosome scanning (Hinnebusch, 2014;

Hinnebusch & Lorsch, 2012; Jackson et al., 2010). Contacts between eIF2 and the 18S rRNA with distinct purine nucleotides at positions -3 and +1 relative to the adenine nucleotide within an identified AUG codon defines an optimal sequence context to end ribosome scanning and initiate polypeptide synthesis (Hinnebusch, 2014; Hinnebusch &

Lorsch, 2012; Jackson et al., 2010). These associations induce conformational changes within the 43S pre-initiation ribosome complex that tighten the interaction of eIF1A with the 40S ribosomal subunit, displacing eIF1 and resulting in a 'closed' conformation of the mRNA entry channel latch with the initiator methionyl-tRNA oriented in an inward, locked position primed for subsequent peptide bond formation (Hinnebusch, 2014;

Hinnebusch & Lorsch, 2012; Jackson et al., 2010). Ejection of eIF1 releases an inhibitory block on both ribosome conformation and eIF5-regulated eIF2 GTPase activity that

14 results in phosphate release, initiation factor dissociation, and 60S ribosomal subunit joining (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012; Jackson et al., 2010).

Outstanding questions remain in regards to the mechanism(s) underlying net 5' to

3' movement of the 43S pre-initiation ribosome complex as well as the fate of the initial recruitment connections during the scanning process. Studies of dicistronic polypeptide production from eukaryotic viral transcripts indicate an oscillating 43S pre-initiation ribosome complex with both forward and reverse movements (Matsuda & Dreher, 2006).

This fluctuating motion is critical to sequence surveillance and selection of an AUG for initiation of polypeptide synthesis (Matsuda & Dreher, 2006). Structural analyses support these molecular findings by presenting the ribosome as a 'processive Brownian motor'

(Frank & Gonzalez, 2010). This model proposes that movement of the ribosome along a transcript is directed by interactions with the nearby environment, such as collisions with mRNA secondary structure or associations with protein cofactors. The conformational changes that result permit both forward and reverse movements of the ribosome that are intricate to the selection of an AUG for initiation of polypeptide synthesis. A sum of these localized fluctuations results in a net outcome of observed 5' to 3' directionality in ribosome scanning.

The initial attachment of a 43S pre-initiation ribosome complex to an activated mRNA required interactions between eIF3 of the 43S pre-initiation ribosome complex and the 5'-bound scaffolding factor eIF4G or CTIF (Choe et al., 2012; Hinnebusch, 2014;

Hinnebusch & Lorsch, 2012). eIF4G also directly associates with the cap-binding protein eIF4E, regulating the stable association of eIF4E with the 5' cap (Yanagiya et al., 2009).

The additional RNA binding activity of eIF4G facilitates this effect (Yanagiya et al.,

2009). Furthermore, eIF4G directly associates with the 3' poly-A binding protein (PABP) to generate a circularized mRNA that is effective for 43S pre-initiation ribosome complex recruitment (Hinnebusch & Lorsch, 2012; Prévôt, Darlix, & Ohlmann, 2003).

Collectively, these eIF4G-driven associations link the 5' cap of a target mRNA with the

43S pre-initiation ribosome complex and are intricate to its recruitment and scanning activity. However, it remains to be determined whether these associations persist during ribosome scanning of the 5' terminus. Their maintenance proposes a 'loop-out' effect of the 5' untranslated region whereby the scanned mRNA bulges as the ribosome continues forward due to its maintained connection with the 5' cap (Jackson et al., 2010). This outcome would effectively allow only one scanning ribosome at a time (Jackson et al.,

2010). Alternatively the connections between the 43S pre-initiation ribosome complex and the 5' cap could be broken, allowing several ribosomes to translocate a 5' terminus simultaneously (Jackson et al., 2010). Advances in single molecule imaging technology and their application to understanding translation control will provide clarification on this fate of the initial recruitment connections during the scanning process.

Initiation, 60S subunit joining and formation of elongation-competent 80S ribosomes

The final step in the initiation stage of translation is joining of the large (60S) ribosomal subunit with the small (40S) ribosomal subunit at the start codon to generate an elongation-competent 80S ribosome (Figure 1.1). The translation initiation factor eIF5B mediates this effect. The presence of eIF5B and the 60S ribosomal subunit stimulates

16 complete release of eIF2-GDP and eIF5 from the stopped 43S pre-initiation ribosome complex at the start AUG (Hinnebusch & Lorsch, 2012). This dissociation allows for the interaction of eIF5B in a GTP-bound form with eIF1A, which remains temporarily associated with the 40S ribosomal subunit (Hinnebusch & Lorsch, 2012). Binding of the

60S ribosomal subunit then follows, stimulating the GTPase activity of eIF5B

(Hinnebusch & Lorsch, 2012). The hydrolysis of GTP to GDP weakens the affinity of eIF5B for the ribosome, resulting in its release (Hinnebusch & Lorsch, 2012). eIF1A subsequently dissociates to leave an 80S ribosome primed for elongation (Hinnebusch &

Lorsch, 2012).

Several mechanistic aspects remain to be defined for this final stage of translation initiation. First, it is unknown how eIF5B and the 60S ribosome are recruited to a 43S pre-initiation ribosome complex (Hinnebusch & Lorsch, 2012). Is eIF5B a component of this multifactor complex that then recruits the 60S ribosomal subunit when in an optimal conformation? Alternatively, does the recruitment of eIF5B to the 43S pre-initiation ribosome complex draw in the 60S ribosome or simply increase the likelihood of interactions between the small and large ribosomal subunits? Second, it is unclear as to what the molecular significance of the eIF5B-eIF1A interaction is in facilitating 60S ribosomal subunit joining (Hinnebusch & Lorsch, 2012). Does the interaction of eIF5B with eIF1A stimulate conformational rearrangements in the 40S ribosomal subunit that favor appropriate 60S binding? Or does an eIF5B-eIF1A association engage more in active recruitment? Ongoing structural studies are aimed at providing greater mechanistic

17 understanding into the significance of eIF5B and eIF1A in the generation of elongation- competent 80S ribosomes (2014 Cold Spring Harbor Translation Control Meeting).

Initiation, variations to the central theme

The cap-dependent scanning mechanism of translation initiation was identified as the defining feature for eukaryotic protein synthesis (Kozak, 1978). The study of RNA viruses, however, soon challenged this central canon with the observation of several alternative mechanisms for translation initiation. These include but are not limited to: internal ribosome entry-site (IRES)-mediated translation initiation, ribosome shunt, ribosomal frameshifting, leaky scanning, non-AUG initiation, and reinitiation (Firth &

Brierley, 2012). Variations to the central theme of 5'-end-dependent translation were also soon realized amongst classes of cellular transcripts. This was particularly observed during times of cell stress and division, resulting in non-canonical translation initiation becoming a hallmark feature of the cellular stress response and mitosis (B. Liu & Qian,

2014; Pyronnet & Sonenberg, 2001; X. Qin & Sarnow, 2004; Spriggs, Stoneley, Bushell,

& Willis, 2008). Recent application of ribosome profiling to the study of eukaryotic translation control has extended our realization of alternative initiation mechanisms in cells (Ingolia, 2014). The data from these studies support pervasive engagement of non-

AUG codons and upstream open reading frames in the steady-state regulation of cellular protein synthesis (Ingolia, 2014). Thus, eukaryotic translation initiation encompasses a diverse array of mechanisms to regulate the start of protein synthesis and intricately control gene expression.

Internal ribosome entry-site (IRES)-mediated translation initiation is by far the most studied mechanism of alternative engagement into protein synthesis. An IRES is an

RNA element capable of directly recruiting the small ribosomal subunit to a start codon without the need for interactions with the 5' cap or cap-associated factors (Firth &

Brierley, 2012; Jackson et al., 2010; Thompson, 2012). Although pervasively observed to control the expression of many RNA viruses and stress-response transcripts, there is no unifying mechanism to describe IRES-mediated translation control (Firth & Brierley,

2012; Jackson et al., 2010; Thompson, 2012). This is because an IRES can only be identified experimentally due to the lack of sequence and structure conservation

(Thompson, 2012). The 'gold-standard' approach is a bicistronic reporter assay whereby a putative IRES element is cloned to regulate the expression of a second, internal cistron within a plasmid harboring two adjacent open reading frames (typically Renilla and firefly luciferase) (Thompson, 2012). Expression of the first cistron is cap-dependent whereas protein production from the second cistron is regulated by the putative IRES element (Thompson, 2012). A relative increase in second cistron protein production, when compared to a construct cloned with a random sequence in its place, indicates IRES activity (Thompson, 2012). A major caveat of this approach, however, is the challenge of discriminating true IRES activity from cryptic promoter functions (Gilbert, 2010).

Additionally, the manner by which a putative IRES sequence is cloned out from its original context and into the bicistronic reporter can have significant consequences on the classification of a particular RNA element as IRES or not (Gilbert, 2010). Thus, in many

19 instances, such as that for the translational regulation of HIV-1, IRES activity remains controversial.

Yet the principle idea of end-independent translation initiation is still well founded with unifying themes observed across instances of accepted IRES activity. This is particularly evident for IRES-mediated translation control of RNA viruses. Type I and

II IRESes are identified by their maintained use of eIF4G and eIF4A to bridge contacts with eIFs 2 and 3 of the 43S pre-initiation ribosome complex, which facilitates its recruitment to a target mRNA (Firth & Brierley, 2012; Jackson et al., 2010). Instead of associating with eIF4E at the 5' cap, as seen in canonical cap-dependent initiation, eIF4G and eIF4A in these instances bind particular RNA elements within the IRES itself

(Jackson et al., 2010). A type I IRES, as exemplified in poliovirus, is mechanistically distinguished from a type II, such as the encephalomyocarditis virus IRES, by the fact that decent scanning of the 43S pre-initiation ribosome complex to the start codon is still involved upon mRNA association (Firth & Brierley, 2012; Jackson et al., 2010). On the contrary, type II IRESes exhibit recruitment of the 43S pre-initiation complex in close proximity to the start codon with minimal scanning observed (Firth & Brierley, 2012;

Jackson et al., 2010). A third class of IRES, type III, is identified by the requirement of only eIFs 2, 3, and 5 to facilitate direct positioning of the 43S pre-initiation ribosome complex at the start codon (Firth & Brierley, 2012; Jackson et al., 2010). This is observed in the case of hepatitis C virus-mediated translation control (Firth & Brierley, 2012;

Jackson et al., 2010). Type IV IRESes, like the cricket paralysis virus, interact with the ribosome independently of any eIF and initiate translation via a structural mimic of the

20 initiator methionyl-tRNA conferred by the IRES itself (Firth & Brierley, 2012; Jackson et al., 2010).

Also critical to the mechanisms by which IRESes initiate translation are distinct

RNA binding proteins that bind an IRES directly or nearby to regulate its efficiency

(King, Cobbold, & Willis, 2010; Komar & Hatzoglou, 2011). Known as IRES trans- acting factors (ITAFs), these RNA binding proteins are critical to the diversification observed in IRES activity, both in mechanism and occurrence (King et al., 2010; Komar

& Hatzoglou, 2011). They can function in RNA conformation control or in RNP remodeling by affecting the stability of RNA secondary structure or serving as bridging cofactors with the ribosome, respectively (King et al., 2010; Komar & Hatzoglou, 2011).

Of the ITAFs identified thus far, many are members of the heterogeneous nuclear ribonucleoprotein (hnRNP) family and include PTB (hnRNP1), hnRNPA1, hnRNPE2, hnRNPE1, hnRNPC, and hnRNPL (King et al., 2010; Komar & Hatzoglou, 2011). This significance is linked with the nuclear-cytoplasmic shuttling activity and additional posttranscriptional functions of hnRNPs and appears distinct for each IRES target

(Komar & Hatzoglou, 2011). Thus, the formation of select trRNPs to regulate eukaryotic translation is fundamental to IRES-mediated initiation.

The importance of trRNPs in translation control is also seen for the other mechanisms of end-independent initiation. Ribosome shunt, ribosomal frameshifting, leaky scanning, non-AUG initiation, and reinitiation are observed due to trRNP elements directing the ribosome to move in an alternative manner as their names imply (Firth &

Brierley, 2012). Although mechanistically characterized by variations in RNA sequence

21 and/or structure impacting ribosomal movement, it can be envisioned that distinct RNA binding proteins have critical roles in these processes as well.

Elongation

Extension of a polypeptide chain to produce a full-length protein product occurs by three steps: amino acid incorporation, peptide bond formation, and ribosome translocation. Two elongation factors (eEFs) coordinate these activities, the dynamics of which are influenced by associated RNA binding proteins. Thus, trRNP biology is fundamental to targeted translation control during the elongation stage of protein synthesis.

Polypeptide extension begins with the incorporation of a charged transfer RNA

(tRNA) into the unoccupied 3' aminoacyl-site (A site) of an 80S ribosome poised to elongate from the start codon (Dever & Green, 2012) (Figure 1.2). This activity is facilitated by the multi-subunit complex eEF1 in its GTP-bound form (Dever & Green,

2012). Identification of appropriate aminoacylated tRNA incorporation is driven by anticodon:codon base pairing (Dever & Green, 2012). Only correct matches allow for conformational arrangements of the tRNA that are tolerated by the ribosome (Dever &

Green, 2012). This recognition triggers GTP hydrolysis, eEF1 release, and full accommodation of the accepted aminoacyl-tRNA in the A site (Dever & Green, 2012).

Next, the A-site aminoacyl-tRNA is linked to the P-site aminoacyl-tRNA, which harbors the initiator methionine, via peptide bond formation (Dever & Green, 2012)

(Figure 1.2). This reaction is catalyzed by the peptidyl transferase center of the large

22 ribosomal subunit (Dever & Green, 2012). The peptidyl transferase center is a region of highly conserved ribosomal RNA that functions in orienting the acceptor stems of the A- and P-site tRNAs, which are the regions bound to the cognate amino acids, to drive the two-step reaction that results in peptide bond formation (Polacek & Mankin, 2005). This amino acid linkage between the A- and P-site tRNAs triggers the ribosomal subunits to move in a ratchet-like motion that results in a hybrid tRNA state prior to complete translocation (Dever & Green, 2012). Here the acceptor stems of the A- and P-site tRNAs are positioned within their adjacent E- and P-sites, respectively, while their anticodon stem loops remain in their corresponding A- and P-sites (Dever & Green, 2012).

Complete translocation of the tRNA-ribosome complex to the subsequent codon is driven by eEF2 association, GTP-hydrolysis, and phosphate release (Dever & Green, 2012)

(Figure 1.2). This results in a deacylated tRNA (a tRNA without its charged amino acid or associated polypeptide chain) within the 5' E-site, a P-site tRNA with a dipeptide bound to its acceptor stem, and an unoccupied A-site that is primed for acceptance of the next aminoacyl-tRNA based upon anticodon:codon base pairing. Repetition of these three steps along an open reading frame grows the encoded polypeptide and occurs until a stop codon is reached, which triggers termination and release of the protein product.

Intricate to the control of translation elongation are select RNA binding proteins, which associate with target transcripts and form distinct trRNPs that regulate the dynamics of polypeptide synthesis. Four major elongation effectors have been identified and mechanistically studied. They are: the fragile X mental retardation protein (FMRP), heat shock protein 70 (Hsp70), pumilio and argonaute (PUF-Ago), and heterogenous

23 ribonucleoprotein E1 (hnRNP E1) (Chen, Sharma, Shi, Agrawal, & Joseph, 2014; Darnell et al., 2011; Friend et al., 2012; Hussey et al., 2011; Shalgi et al., 2013). Although their targets and mechanisms for regulation are distinct, the unifying principle among all four effectors is that their formation of distinct trRNPs is intricate to the molecular basis by which they regulate the dynamics of polypeptide synthesis.

FMRP binds the mRNA coding sequence of pre- and postsynaptic transcripts and induces elongation pausing by competing with P-site tRNA binding (Chen et al., 2014;

Darnell et al., 2011). The result is impacted ribosome-tRNA dynamics that compromise the elongation process on target transcripts (Chen et al., 2014; Darnell et al., 2011). This directed translation control is critical for the spatiotemporal regulation of neuronal protein expression, which is essential for proper nervous system function (Chen et al.,

2014; Darnell et al., 2011). The specificity for FMRP-mediated control of translation elongation is driven by its affinity for particular RNA sequence elements, GAC,

ACUG/U, and A/UGGA, which have predicted G-quadruplex secondary structure (Suhl,

Chopra, Anderson, Bassell, & Warren, 2014). Mutational studies indicate that the ability of FMRP to bind RNA is sufficient for its induced elongation pausing effect (Chen et al.,

2014; Darnell et al., 2011).

Hsp70 is a critical molecular chaperone that facilitates protein production by associating with nascent polypeptide chains during the elongation process and coordinating their correct folding into a functional protein product (Shalgi et al., 2013).

This activity of Hsp70 involves its association with ribosomal proteins of the peptide exit tunnel, particularly the large ribosomal subunits RPL4 and RPL22, and elongation factor

24 eEF1A (Shalgi et al., 2013). These interactions function to facilitate efficient movement of the nascent polypeptide chain through the peptide exit tunnel and proper elongation kinetics (Shalgi et al., 2013). Heat shock, however, alters the interaction of Hsp70 with the ribosome and eEF1A such that the peptide exit tunnel becomes constricted and the nascent polypeptide chain exposed in a manner that compromises elongation kinetics to cause global ribosome pausing at codon 65 of most mRNAs (Shalgi et al., 2013). Codon

65 correlates with a sixty-five amino acid polypeptide, which is long enough to traverse the peptide exit tunnel and become exposed so that it is affected by altered interactions of

Hsp70 during heat shock. The outcome is regulated translation that is fundamental to the heat shock response (Shalgi et al., 2013).

PUF-Ago and hnRNP E1 control translation elongation by regulated interactions with elongation factor eEF1A. eEF1A is the subunit of eEF1 that binds GTP and is responsible for delivering aminoacyl-tRNAs to the A site of the ribosome (Sasikumar,

Perez, & Kinzy, 2012). In the case of PUF-Ago, these RNA binding proteins associate with eEF1A in a manner that inhibits its GTPase activity (Friend et al., 2012). The effect is stalled elongating ribosome complexes within the open reading frames of transcripts, hindering protein production (Friend et al., 2012). The influence of PUF-Ago on eEF1A

GTPase activity appears specific for elongating complexes that have traversed the mRNA to the point of nascent polypeptide chain emersion from the ribosomal exit tunnel (Friend et al., 2012). The factors directing this specificity and its significance for miRNA- mediated translation control remain to be elucidated. hnRNP E1, on the other hand, regulates eEF1A function by binding to and inhibiting its dissociation from ribosomes

(Hussey et al., 2011). The outcome is impaired ribosome translocation and protein production (Hussey et al., 2011). This effect is critical for regulated expression of epithelial-mesenchymal transition transcripts that are significant in development and cancer (Hussey et al., 2011).

Termination

The translation process concludes with termination of polypeptide elongation, nascent chain release, and ribosome recycling. Translation termination is triggered by the recognition of a stop codon (UAA, UAG, or UGA) within the A-site of an elongating ribosome (Jackson, Hellen, & Pestova, 2012) (Figure 1.3). This recognition occurs in eukaryotes by the eukaryotic termination factor 1 (eRF1), which structurally exists as a tRNA mimic and associates with the stop codon-containing A site (Jackson et al., 2012). eRF1 harbors a conserved GGQ motif that positions within the peptidyl transferase center and induces conformational changes that allow access of a water molecule to this active site (Jackson et al., 2012) (Figure 1.3). The result is breakage of the ester bond that holds the polypeptide chain to the P site tRNA, releasing the protein product (Jackson et al.,

2012). This function of eRF1 is stimulated by its association with the second termination factor, eRF3 (Jackson et al., 2012). eRF3 is a GTPase whose affinity for GTP is increased by association with eRF1 (Jackson et al., 2012). eRF1 and eRF3 are found within stable complexes and their ribosome association triggers GTP hydrolysis, movement of the

GGQ motif into the peptidyl transferase center, and eRF1-induced polypeptide release

(Jackson et al., 2012).

The ATP-binding cassette protein ABCE1 is responsible for ribosome recycling upon nascent chain release (Jackson et al., 2012) (Figure 1.3). The retention of eRF1 on post-termination complexes recruits ABCE1 and stimulates its ATPase activity (Jackson et al., 2012). Hydrolysis of ATP induces a conformational switch in ABCE1 from a closed ATP-bound state to an open ADP-bound state (Jackson et al., 2012). This movement of ABCE1 causes the 60S ribosomal subunit to split away from its 40S counterpart (Jackson et al., 2012). Initiation factors eIF1, 1A, and 3 facilitate subsequent deacylated tRNA and mRNA release from the 40S ribosomal subunit via competitive binding interactions (Jackson et al., 2012). The outcome is recycling of all factors required for subsequent rounds of protein synthesis. Interactions between the 3' poly-A binding protein (PABP) and the 5' scaffolding factor eIF4G, which as previously discussed function to create a circularized mRNA, allow for efficient reinitiation and subsequent rounds of protein synthesis upon termination (Hinnebusch, 2014; Hinnebusch

& Lorsch, 2012; Jackson et al., 2010).

Two main exceptions to this mechanistic paradigm of eukaryotic translation termination are known. They occur during the phenomena of reinitiation and nonsense- mediated mRNA decay. Reinitiation is when the ribosome fails to dissociate from a transcript upon termination of polypeptide synthesis (Jackson et al., 2010). Instead it resumes scanning to a downstream AUG where a second initiation event is engaged

(Jackson et al., 2010). This is observed among cellular transcripts that harbor short upstream open reading frames and is critical for their targeted translation control (Jackson et al., 2012). A principle of the eukaryotic scanning model of protein synthesis is that a

43S pre-initiation ribosome complex scans an mRNA until the first AUG is detected

(Jackson et al., 2012). It will then engage in 80S formation and polypeptide production irrespective of the length of the open reading frame (Jackson et al., 2012). Consequently, for mRNAs that harbor several open reading frames, a competition arises between upstream initiation events and downstream polypeptide production (Jackson et al., 2012).

Often it is only when translation activity is impaired, such as during cell stress, does compromised initiation at the upstream open reading frame allow for effective expression of the downstream protein product (see section below, 'Regulation of eukaryotic translation'). This results in a means for effective translation control, as the downstream open reading frame often encodes the functional cellular protein.

Reinitiation is also intricate to the molecular basis by which many infectious viral proteins are expressed (Firth & Brierley, 2012; Jackson et al., 2012). Although the mechanisms regulating reinitiation are diverse, a common theme is the role of distinct trRNPs in mediating this effect. Such is the case, for example, in the expression of critical structural proteins for mammalian caliciviruses, which are responsible for several life- threatening diseases in animals. An RNA element termed 'termination codon upstream ribosome-binding site' (TURBS) is critical to reinitiation and expression of infectious downstream viral protein products on caliciviral transcripts (Jackson et al., 2012). The

TURBS consists of two motifs that harbor sequences of the 3' terminus of the upstream open reading frame and the region between the first stop and second start site (Jackson et al., 2012). These motifs are necessary for reinitiation events by functioning to capture

40S ribosomal subunits upon termination of upstream translation, effects that occur by

28 the TURBS element forming a critical trRNP that associates with eIF3 and mimics 18S rRNA (Jackson et al., 2012). The outcome is effective expression of downstream viral protein products in measured amounts that are critical to caliciviral infection (Jackson et al., 2012).

Nonsense-mediated mRNA decay is the essential cellular process of mRNA surveillance and quality control. It is responsible for the resolution of premature termination events and subsequent degradation of the effected mRNA so as to prevent production and accumulation of rogue protein products. Premature termination results from the recognition of premature termination codons by eRF1 and eRF3 (Schoenberg &

Maquat, 2012). Premature termination codons are stop codons that are positioned upstream of the normal stop codon within an open reading frame (Schoenberg & Maquat,

2012). Although eRF1 and eRF3 are recruited to premature termination codons, the subsequent termination events of ABCE1 recruitment and ribosome recycling do not occur. Instead, an alternative trRNP forms that halts the translation process, inhibits subsequent initiation events, and induces mRNA decay (Schoenberg & Maquat, 2012).

This alternative trRNP consists of a complex series of interactions between the cap- binding protein CBP80/20, a nearby exon junction complex, and recruited factors from the up-frameshift (UPF) and suppressor of morphogenetic effect on genitalia (SMG) protein families (Schoenberg & Maquat, 2012). Collectively, these interactions and the progression in their association regulate trRNP dynamics that result in altered termination fates to control protein expression.

Regulation of eukaryotic translation

Global and targeted regulation of eukaryotic translation is observed at all three stages of protein synthesis. These mechanisms can be constitutive or induced upon changes in the cellular environment. In either case, distinct trRNP formation and activity is fundamental to each effect.

43S pre-initiation ribosome complex formation, its recruitment and scanning, and

60S ribosomal subunit joining are all examples of targetable steps for regulated gene expression at the stage of translation initiation. As previously discussed, DExH/D-box

RNA helicases and the select trRNPs they create are instrumental for coordinating 43S pre-initiation ribosome complex recruitment and scanning on target transcripts. Other trRNPs, formed by alternative RNA binding proteins, contribute a similar effect. Two classic examples are the regulated expression of ferritin mRNA in response to iron homeostasis and male-specific lethal-2 (msl-2) translation in Drosophila X-chromosome dosage compensation. During iron deprivation, the RNA binding protein iron regulatory protein (IRP) binds with high affinity to its cognate cis-acting RNA element, the iron- responsive element (IRE), which defines the 5' terminus of ferritin mRNA (Gray &

Hentze, 1994). The IRE is a stem-loop structure that when bound by IRP effectively impedes 43S pre-initiation ribosome complex association and expression of the ferritin mRNA (Gray & Hentze, 1994; Muckenthaler, Gray, & Hentze, 1998). Elevated iron levels reduce the affinity of IRP for IRE, effectively lifting its block on translation initiation and allowing for expression of ferritin (Gray & Hentze, 1994).

Inhibition of msl-2 expression in female flies is fundamental to dosage compensation and survival. This effect is mediated by the female-specific RNA binding protein Sex-lethal (SXL), which binds distinct poly-uridine tracts in the 5' and 3' termini of msl-2 and inhibits translation (Beckmann, Grskovic, Gebauer, & Hentze, 2005). The molecular basis for this effect is two-pronged. At the 3' terminus, SXL impairs 43S pre- initiation ribosome complex recruitment by associating with the 3'-bound poly-A binding protein (PABP) to interfere with dynamics at the 5' cap that allow for appropriate initiation (Beckmann et al., 2005; Duncan, Strein, & Hentze, 2009). This coordination between 3' effectors and a 5' outcome is mediated by PABP-induced mRNA looping

(Duncan et al., 2009). At the 5' terminus, SXL stalls scanning 43S pre-initiation ribosome complexes that were able to circumvent the 3'-mediated block, effectively reducing their affinity for msl-2 and causing their dissociation (Beckmann et al., 2005). The outcome is repressed expression of msl-2 protein and effective dosage compensation.

Blockade of 60S ribosomal subunit joining is another effective mechanism employed by distinct trRNPs to coordinate targeted translation regulation. Such is the case for controlled expression of erythroid 15 lipoxygenase (LOX) mRNA during erythroid differentiation. LOX mRNA encodes a critical enzyme important for internal membrane reorganization during late stages of red blood cell maturation (Ostareck,

Ostareck-Lederer, Shatsky, & Hentze, 2001; Ostareck et al., 1997). Temporal restriction of its expression is mediated by the association of the RNA binding proteins hnRNP K and hnRNP E1 with the cis-acting differentiation control element (DICE) in the 3' terminus of the LOX mRNA (Ostareck et al., 1997; 2001). DICE is a repeated CU-rich

31 motif that when bound by hnRNP K and hnRNP E1 inhibits 60S ribosomal subunit joining by interfering with initiation factor activity that mediates this effect (Ostareck et al., 1997; 2001). A similar molecular basis is seen in the spatiotemporal regulation of β- actin expression whereby the RNA binding protein Zipcode binding protein 1 (ZBP1) binds the 3' cis-acting zipcode RNA element in the β-actin mRNA and impedes 60S ribosomal subunit joining to effectively suppress translation activity (Hüttelmaier et al.,

2005). This blockade is uplifted by phosphorylation of ZBP1, which reduces its affinity for the β-actin mRNA and allows translation to proceed (Hüttelmaier et al., 2005).

Regulated elongation is third effective means by which protein synthesis can be controlled. As previously discussed, there are known instances for targeted elongation regulation in which specific RNA binding proteins influence trRNP dynamics to control elongating ribosome activity. Global regulation of elongation is also observed and occurs with the phosphorylation of the elongation factor eEF2. This posttranslational modification interferes with eEF2-GTP complex formation, which inhibits the association of eEF2 with the ribosome and effectively impedes translocation (Hizli et al., 2013).

Phosphorylation of eEF2 occurs by the eEF2 kinase (Hizli et al., 2013).

Mammalian eEF2 kinase activity is controlled by the cell's central commander of translation: the mammalian target of rapamycin (mTOR) (Figure 1.4). mTOR is a serine/threonine kinase that directs protein synthesis in reflection of the cellular environment by acting upon signals received from almost all major cell-receptor signaling pathways (Laplante & Sabitini, 2012). These include but are not limited to the

PI3K/AKT and Ras/ERK signaling cascades (Ma & Blenis, 2009). One downstream

32 effector of mTOR is the ribosomal protein S6 kinase (S6K) (Figure 1.4). During normal growth conditions or in response to mitogenic stimuli, mTOR is activated and phosphorylates S6K (Ma & Blenis, 2009). S6K, in turn, phosphorylates the eEF2 kinase, which results in eEF2 kinase inactivation, eEF2-GTP association, and the promotion of translation elongation (Hizli et al., 2013; Ruvinsky & Meyuhas, 2006). However, when mTOR is inactivated by cellular stress the inhibitory block of S6K on eEF2 kinase activity is relieved, resulting in phosphorylation of eEF2 and impaired translation elongation activity (Hizli et al., 2013; Ruvinsky & Meyuhas, 2006). The outcome is suppression of global translation.

Regulated protein synthesis in response to the environment is also observed at the stage of translation initiation. Here mTOR and its downstream effector S6K once again serve as a central line of command to influence trRNP dynamics and protein production

(Figure 1.4). In this instance, mTOR-directed S6K activation targets the translation initiation factor eIF4B and the regulatory protein programmed cell death 4 (PDCD4) (Ma

& Blenis, 2009; Ruvinsky & Meyuhas, 2006). Phosphorylation of eIF4B by S6K enhances eIF4B's stimulation of eIF4A helicase activity and protein production (Ma &

Blenis, 2009; Ruvinsky & Meyuhas, 2006). Likewise, phosphorylation of PDCD4 results in eIF4A-mediated translational activation by causing its subsequent ubiquitylation and degradation (Ma & Blenis, 2009; Ruvinsky & Meyuhas, 2006). PDCD4 is an inhibitor of eIF4A helicase activity; thus, its removal relieves an inhibitory block on eIF4A- dependent translation initiation (Ma & Blenis, 2009; Ruvinsky & Meyuhas, 2006). The

33 outcome of both effects is activation of eIF4A and effective translation initiation on target mRNAs.

Several additional downstream targets of S6K are also known and their activity has significant implications for regulated translation initiation (Figure 1.4). These S6K targets are: the 40S ribosomal subunit protein S6 (rpS6), the S6K1 Aly/REF-like target

(SKAR), the cAMP-responsive element modulator τ (CREM τ), CBP80 of the CBP80/20 cap-binding protein complex, the proapoptotic protein BAD, insulin receptor substrate

IRS (IRS), and mTOR itself (Ruvinsky & Meyuhas, 2006). However, in many of these studied instances, such as rpS6 and CBP80 phosphorylation, it remains controversial about the exact effects of this posttranslational modification on their role in translation control and the influence of S6K in these outcomes (Ruvinsky & Meyuhas, 2006).

A second line of command used by mTOR to control translation initiation activity is the eIF4E inhibitory protein 4E-binding protein 1 (4E-BP1) (Ma & Blenis, 2009;

Ruvinsky & Meyuhas, 2006) (Figure 1.4 and 1.5). 4E-BP1 influences cap-dependent translation initiation by competing with eIF4G for eIF4E association (Haghighat, Mader,

Pause, & Sonenberg, 1995; Mader, Lee, Pause, & Sonenberg, 1995; Marcotrigiano,

Gingras, Sonenberg, & Burley, 1999). A 4E-BP1:eIF4E interaction effectively disrupts cap-associated trRNP dynamics to suppress translation initiation (Haghighat et al., 1995;

Mader et al., 1995; Marcotrigiano et al., 1999). The association of 4E-BP1 with eIF4E is regulated by its phosphorylation status. In response to mitogenic stimuli, activation of mTOR results in its hyperphosphorylation of 4E-BP1 (Ma & Blenis, 2009; Pause et al.,

1994). In this posttranslational state, 4E-BP1 has reduced affinity for eIF4E (Pause et al.,

1994). This allows the eIF4E cap-binding protein to interact with eIF4G and promote cap-dependent translation (Pause et al., 1994). On the contrary, stress-induced suppression of mTOR activity results in hypophosphorylation of 4E-BP1, a strong 4E-

BP1:eIF4E association, and consequent suppression of eIF4E-dependent translation initiation (Ma & Blenis, 2009; Pause et al., 1994).

Besides directing mTOR-mediated translational control, the cellular environment influences protein synthesis by regulating the function of eIF2 in the earliest stage of initiation, 43S pre-initiation ribosome complex formation. The very initial association of an initiator methionyl-tRNA with a 40S ribosomal subunit is facilitated by eIF2 in its

GTP-bound form (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012). eIF2 consists of 3 subunits: α, β, and γ (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012) (Figure 1.6). The α subunit together with β serves as a critical allosteric effector of direct GTP and initiator methionyl-tRNA binding to the γ subunit of eIF2 (Hinnebusch, 2014; Hinnebusch &

Lorsch, 2012). GTP association occurs first and is rate-limiting for initiator methionyl- tRNA binding (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012). Phosphorylation of the

α subunit at serine residue 51 inhibits the exchange of GDP for GTP on eIF2, an effect mediated by the guanine nucleotide exchange factor eIF2B (Hinnebusch, 2014;

Hinnebusch & Lorsch, 2012) (Figure 1.6). The outcome is impaired efficiency of ternary complex formation (initiator methionyl-tRNA and eIF2 bound to the 40S ribosomal subunit), which results in reduced 43S pre-initiation ribosome complex formation, its recruitment to mRNAs, and subsequent translation (Hinnebusch, 2014; Hinnebusch &

Lorsch, 2012).

Four distinct protein kinases have been identified that phosphorylate eIF2α: haem-regulated inhibitor kinase (HRI), protein kinase R (PKR), PKR-like endoplasmic reticulum kinase (PERK), and general control non-derepressible-2 (GCN2) (Figure 1.6).

Notably, their activation occurs in response to a variety of environmental stressors as the cell attempts to rapidly adjust to a change in its homeostasis. Iron deficiency and osmotic or heat shock activate HRI; double-stranded RNA triggers PKR activity; ER-stress and hypoxia activate PERK; amino-acid deprivation and UV irradiation stimulate GCN2

(Holcik & Sonenberg, 2005). The outcome in all instances is phosphorylation of eIF2α and suppression of general translation, as the 43S pre-initiation complex is fundamental to protein synthesis. However, distinct transcripts, like mammalian activating transcription-factor-4 (ATF4) and the yeast transcriptional activator GCN4, circumvent this block and exhibit efficient expression even in the advent of eIF2α phosphorylation.

This targeted translation activity is specific for critical stress response proteins and is fundamental to the cell's ability to regain homeostasis and avoid apoptosis and cell death.

The ability of specific mRNAs to engage in translation in the advent of eIF2α phosphorylation is dictated by their distinct trRNPs. In the case of ATF4, its 5' terminus is characterized by two upstream open reading frames (Holcik & Sonenberg, 2005). In cellular states when eIF2α phosphorylation is low and ternary complex is abundant, upstream open reading frame 2 preferentially engages 43S pre-initiation ribosome complexes in recruitment and scanning over the downstream ATF4 open reading frame

(Holcik & Sonenberg, 2005). The outcome is inhibited expression of ATF4 (Holcik &

Sonenberg, 2005). However, in ER stress when misfolded proteins accumulate and

36 activate PERK, phosphorylation of eIF2α causes the limiting pool of 43S pre-initiation ribosome complexes to scan through upstream open reading frame 2 and initiate translation at the downstream ATF4 open reading frame (Holcik & Sonenberg, 2005).

The result is efficient synthesis of ATF4, which is a major transcription factor that facilitates expression of genes critical to resolving ER stress (Holcik & Sonenberg, 2005).

A similar molecular basis explains the translational activation of yeast GCN4 in response to amino acid starvation (Holcik & Sonenberg, 2005).

Distinct mechanisms of translation regulation, both global and specific, are also observed in virus-infected cells. Here a complex interplay between translation suppression and activation and canonical and non-canonical protein synthesis is essential for viral replication and the host innate defense. Although the specific mechanisms characterizing this translational reprogramming event are diverse among each virus infection, the central theme of trRNP-mediated translation control is fundamental to explaining each observed outcome. Human immunodeficiency virus type 1 (HIV-1) infection is a model example to demonstrate how trRNP biology is at that core of translational reprogramming that characterizes viral infections.

A hallmark feature of HIV-1 infection is global suppression of host cell translation (A. Sharma, Yilmaz, Marsh, Cochrane, & Boris-Lawrie, 2012). This effect is mediated by HIV-1-induced activation of 4E-BP1 (A. Sharma et al., 2012). 4E-BP1 is the eIF4E inhibitory protein that binds the translation initiation factor eIF4E and disrupts associated trRNP dynamics to impede canonical cap-dependent protein synthesis. Since the eIF4E trRNP is responsible for the bulk of cellular steady-state protein synthesis (see

37 below), the outcome is observed global suppression of host cell translation (A. Sharma et al., 2012). Yet during this effect, HIV-1 maintains expression of its critical structural proteins (A. Sharma et al., 2012). As an obligate parasite, HIV-1 requires the host cell translation machinery for expression of its encoded viral proteins. Furthermore, it does so in an arguably cap-dependent scanning manner (Bolinger et al., 2007; 2010). Yet, how is this possible with its suppression of eIF4E trRNP activity? The answer is in the ability of

HIV-1 transcripts to engage within an alternative trRNP that harbors the CBP80/20 cap- binding protein complex (CBC) (A. Sharma et al., 2012). CBC functions in translation initiation like eIF4E, however, its prominence is during the pioneer rounds of translation

(see below). Notably, CBC activity is independent of regulation by 4E-BP1 and is known to be functional during states of cell stress (see below). Thus, the CBC trRNP provides an alternative mechanism for HIV-1 viral protein expression that is cap-dependent but independent of eIF4E tRNP activity (A. Sharma et al., 2012). Therefore, the translation dynamics during HIV-1 infection demonstrate how distinct trRNP activity choreographs targeted protein synthesis.

CBC and eIF4E: core trRNPs of eukaryotic translation control

Two core trRNPs characterize general eukaryotic cap-dependent translation.

These are: the CBC trRNP and the eIF4E trRNP. As previously discussed, initiation of eukaryotic translation is defined by a cap-dependent scanning mechanism whereby the ribosome binds the 5' terminus of an mRNA and proceeds along the transcript inspecting, base by base, for an appropriate start codon to initiate polypeptide synthesis. This effect

38 is directed by trRNP complexes and in particular the defining cap-binding proteins,

CBP80/20 (CBC) or eIF4E. Although CBC and eIF4E direct the interaction of an mRNA with the ribosome in a similar manner, the molecular basis by which each mediates this effect is distinct (Chiu, Lejeune, Ranganathan, & Maquat, 2004). Consequently, the CBC trRNP and the eIF4E trRNP have distinct functional significances in the choreographing of eukaryotic translation (Figure 1.7).

The CBC trRNP

CBC is a heterodimeric protein complex of two subunits: CBP80 and CBP20.

CBP20 directly binds the 7-methyl guanosine cap of eukaryotic mRNAs while CBP80 regulates this interaction (Izaurralde et al., 1994). The association of CBC with a 5' cap occurs co-transcriptionally when a nascent transcript emerges from the RNA polymerase

II holoenzyme (Visa, Izaurralde, Ferreira, Daneholt, & Mattaj, 1996) (Figure 1.7). This affinity is facilitated by the abundant steady-state localization of CBC within the nucleus, an effect driven by a bipartite nuclear localization sequence within CBP80 and its association with the nuclear import factor importin-α1 (Görlich et al., 1996; Izaurralde,

McGuigan, & Mattaj, 1995b). Here CBC is critical for the major posttranscriptional events of 3' end processing and splicing that mature an mRNA (Flaherty, Fortes,

Izaurralde, Mattaj, & Gilmartin, 1997; Izaurralde et al., 1994; Lewis & Izaurralde, 1997).

CBC then remains bound to the 5' cap and facilitates mRNA nuclear export (Görlich et al., 1996; Izaurralde et al., 1995a; Izaurralde, Stepinski, Darzynkiewicz, & Mattaj, 1992;

Lewis & Izaurralde, 1997; Visa et al., 1996).

Once in the cytoplasm, CBC engages trRNP activity that coordinates the so-called pioneer round(s) of translation (Fortes et al., 2000; Ishigaki, Li, Serin, & Maquat, 2001)

(Figure 1.7). These encompass the initial interactions of a ribosome with an mRNA to generate protein product. Defining this CBC trRNP is CBC itself, remained bound to the

5' cap, and its direct association with the translation initiation factor eIF4G or CTIF (K.

M. Kim et al., 2009; Lejeune, Ranganathan, & Maquat, 2004). A CBC-eIF4G/CTIF interaction is important for generating a molecular bridge with the translation initiation factor eIF3, which recruits the 43S pre-initiation ribosome complex to the 5' terminus and initiates scanning (Choe et al., 2012; K. M. Kim et al., 2009; Lejeune et al., 2004).

Pioneer CBC trRNP complexes are also defined by the presence of exon junction complexes and the nuclear poly-A binding protein (PABPN) (Ishigaki et al., 2001;

Lejeune, Ishigaki, Li, & Maquat, 2002) (Figure 1.7). Exon junction complexes are dynamic protein assemblies that organize upon the coding sequence of an mRNA with the conclusion of a splicing event and facilitate critical posttranscriptional activities. One of these functions is to enhance the translation of spliced mRNAs by providing a molecular bridge with the 43S pre-initiation ribosome complex (Diem, Chan, Younis, &

Dreyfuss, 2007). This occurs via a direct association between the EJC interacting factor

PYM and the 40S ribosomal subunit (Diem et al., 2007). PABPN is a 3' associated factor that is bound to the poly-A tail of mRNAs and functions in the earlier posttranscriptional event of 3' end processing (Kühn & Wahle, 2004).

The significance of the CBC trRNP, both in composition and function, is for the mRNA surveillance and quality control process of nonsense-mediated mRNA decay

(NMD) (Chiu et al., 2004; Ishigaki et al., 2001). NMD is intricately linked to the pioneer rounds of translation whereby it assesses initial mRNA integrity for appropriate full- length protein production (Schoenberg & Maquat, 2012). This activity involves the recognition and resolution of premature termination events so as to prevent production and accumulation of rogue protein products (Schoenberg & Maquat, 2012). Critical interactions between CBP80 and the NMD effector up-frameshift protein 1 (UPF1) mediate this effect (Hwang, Sato, Tang, Matsuda, & Maquat, 2010). The associated exon junction complexes of the CBC trRNP are also necessary for coordinating recognition of premature termination events with subsequent translation inhibition and directed mRNA decay (Schoenberg & Maquat, 2012). Thus, the long-standing model has been that the

CBC trRNP is distinct for the pioneer round(s) of translation in order to identify targets of NMD.

Recent studies, however, have provided significance for the CBC trRNP beyond the pioneer round(s) of translation and NMD. Such is the case for the expression of antigenic peptides of the MHC class I pathway (Apcher et al., 2011). The ability to distinguish self from non-self is critical to appropriate immune function and recognition of invading pathogens. The MHC class I pathway functions in both T-cell education and activation of the immune system. Critical to this effect is the generation of antigenic peptides. Cap-dependent translation affords a molecular basis for regulated peptide expression. It was demonstrated that inhibition of eIF4E trRNP dynamics impaired the production of full-length protein products without consequence on the generation of antigenic peptides (Apcher et al., 2011). Furthermore, the temporal regulation of

41 antigenic peptide production was shown to coincide with prominent CBC trRNP activity

(Apcher et al., 2011). Thus, these findings indicate a role for the CBC trRNP in the innate immune response.

The CBC trRNP is also critical for the regulated expression of the core histone proteins (Choe et al., 2013). Genome integrity is dependent upon the effective packaging of DNA into appropriate chromatin structure. This effect is mediated by the coordination of histone protein synthesis with DNA replication. Robust histone protein production is observed during the S phase of the cell cycle when DNA replication occurs. Completion of S phase triggers the rapid degradation of histone mRNAs, which effectively inhibits their expression and coordinates histone production with DNA replication. The molecular basis by which histone translation is linked to its mRNA degradation is the CBC trRNP

(Choe et al., 2013). Histone mRNAs exhibit preferential association with the CBC trRNP during steady-state translation due to a direct interaction between CTIF and the identifying 3' stem-loop binding protein (SLBP) of histone mRNAs (Choe et al., 2013).

This SLBP-CTIF interaction generates a trRNP that facilitates efficient translation of histone mRNAs during S phase and primes for their rapid degradation upon the completion of DNA synthesis (Choe, Ahn, & Kim, 2014a). The rapid degradation of histone mRNAs is driven by S phase-dependent phosphorylation of UPF1 and its competition with CTIF for SLBP association (Choe, Ahn, & Kim, 2014a). The outcome is a dynamic rearrangement in the CBC trRNP that facilitates mRNA degradation (Choe,

Ahn, & Kim, 2014a). Collectively, these findings indicate significance for the CBC trRNP in the molecular basis of genome integrity.

An additional function for the CBC trRNP beyond the pioneer round(s) of translation is cap-dependent gene expression of HIV-1 (A. Sharma et al., 2012). As previously discussed, a hallmark feature of HIV-1 infection is global suppression of host cell translation (A. Sharma et al., 2012). This effect is due to HIV-1-induced activation of

4E-BP1 and consequent suppression of eIF4E trRNP activity (A. Sharma et al., 2012).

However, as an obligate parasite HIV-1 requires host cell translation machinery for protein expression. Furthermore, it does so in an arguably cap-dependent manner

(Bolinger et al., 2007; 2010). This seemingly paradoxical conundrum is resolved by the virus selectively engaging the CBC trRNP for expression of its critical structural proteins

(A. Sharma et al., 2012). Unlike eIF4E, CBC is independent of regulation by 4E-BP1

(see below). Furthermore, regulation of CBC trRNP activity has yet to be identified (see below). Therefore, this finding implies significance for the CBC trRNP in maintained cap-dependent translation during cell stress.

The eIF4E trRNP

The eIF4E trRNP dynamically assembles upon an mRNA subsequent to the CBC trRNP and facilitates steady-state protein synthesis (Chiu et al., 2004; Sato & Maquat,

2009) (Figure 1.7). The eIF4E trRNP is compositionally distinguished from the CBC trRNP by the presence of the eIF4E cap-binding protein bound to the 5' cap of an mRNA in place of CBC. eIF4E is selective for an association with eIF4G and does not exhibit an interaction with CTIF, making eIF4G another defining member of the eIF4E trRNP (K.

M. Kim et al., 2009). Furthermore, eIF4E selectively interacts with the DEAD-box RNA

43 helicase eIF4A such that eIF4A is also a defining member of eIF4E trRNP (Chiu et al.,

2004). The eIF4E trRNP is further distinguished from the CBC trRNP by its absence of associated exon junction complexes and its exclusive interaction with the cytoplasmic poly-A binding protein (PABPC) (Chiu et al., 2004; Lejeune et al., 2002).

These characteristic differences between the eIF4E trRNP and the CBC trRNP, both in composition and temporal association with an mRNA, are driven by the dynamic trRNP remodeling events that occur following the pioneer round(s) of translation (Figure

1.7). CBC has a high affinity for the nuclear import factor importin α1, an association that is driven by the classic nuclear localization sequence in CBP80 (Görlich et al., 1996).

Notably, the interaction of CBC with importin α1 is resistant to high salt and observed throughout the nuclear-cytoplasmic shuttling activity of CBC (Görlich et al., 1996).

Typical nuclear-cytoplasmic shuttling proteins only demonstrate a robust interaction with importin α1 in the cytoplasm, as their binding is rapidly dissociated upon nuclear import by the Ran-GTP gradient and interactions with the nuclear export factor exportin 2 (also known as CAS) (Cook, Bono, Jinek, & Conti, 2007). The unique affinity of CBC for importin α1 primes CBC for an interaction with importin β1 in the cytoplasm (Sato &

Maquat, 2009). Importin β1 is the critical karyopherin that drives importin α1-mediated nuclear import by serving as a molecular bridge between importin α1-cargo complexes and the nuclear pore complex (Cook et al., 2007). A cytoplasmic CBC-importin α1- importin β1 interaction destabilizes CBC from the 5' cap, allowing eIF4E to bind and for

CBC to recycle to the nucleus (Görlich et al., 1996; Sato & Maquat, 2009). Impaired importin α1-importin β1 dynamics reduces the exchange of CBC for eIF4E on mRNAs

(Sato & Maquat, 2009). Additionally, direct cofactor interactions can retain CBC on target transcripts by functioning as a molecular clasp that latches CBC onto the 5' cap and prevents eIF4E association (Choe et al., 2013). Notably, importin-driven CBC dissociation from the 5' cap is critical for the molecular basis of trRNP remodeling as eIF4E exhibits a lower affinity for 7-methyl-guanosine relative to CBC (Worch et al.,

2005). Thus, affinity competition is not sufficient to drive eIF4E association with the 5' cap even with eIF4E's abundant cytoplasmic localization.

The binding of eIF4E to the 5' cap draws stabilized associations with eIF4G that recruits eIF4A and completes formation of the eIF4E trRNP at the 5' terminus. Removal of exon junction complexes and the exchange of PABPN for PABPC occur with the pioneer rounds of translation (Sato & Maquat, 2009). The exact remodel mechanisms governing these effects remain to be elucidated (Sato & Maquat, 2009). It is important to emphasize here a critical distinction between trRNP remodeling at the 5' terminus from that at the 3' terminus in the exchange of the CBC trRNP for the eIF4E trRNP. As previously discussed, exchange of CBC for eIF4E at the 5' cap is driven by a translation- independent association of CBC with the nuclear import factors importin α1 and importin

β1 (Sato & Maquat, 2009). A CBC-importin α1-importin β1 complex formation destabilizes CBC from the 5' cap, allowing eIF4E to bind (Sato & Maquat, 2009).

Remodeling at the 3' terminus, however, with the removal of exon junction complexes and the exchange of PABPN for PABPC is dependent upon active ribosome scanning and translation (Sato & Maquat, 2009). These mechanistic distinctions provide opportunities for a transcript to undergo retention of CBC at the 5' cap but exhibit efficient remodel at

45 the 3' terminus. In most instances, however, complete trRNP exchange occurs and an mRNA becomes fully engaged within the eIF4E trRNP to undergo steady-state translation (Sato & Maquat, 2009).

Regulation of the CBC and eIF4E trRNPs

Both the CBC trRNP and the eIF4E trRNP are subjected to regulation by changes in the core translation machinery. This includes phosphorylation of eIF2α. CBC and eIF4E each interact with eIF2 and eIF3, implying similar associations with the 43S pre- initiation ribosome complex (Chiu et al., 2004). Furthermore, introduction of a phosphomimetic mutant of eIF2α significantly impairs nonsense mediated mRNA decay to indicate that the pioneer round of translation requires functional eIF2α like steady-state protein synthesis (Chiu et al., 2004).

A distinction is made between the regulation of the CBC trRNP and that of the eIF4E trRNP when 4E-BP1-mediated translation control is considered. As previously discussed, 4E-BP1 is a primary target and effector of mTOR activation. mTOR is the major cellular commander that translates extracellular stimuli into coordinated effects on protein synthesis. In an mTOR-induced hyperphosphorylated state, 4E-BP1 exhibits reduced affinity for eIF4E (Haghighat et al., 1995; Mader et al., 1995; Marcotrigiano et al., 1999). This allows an association between eIF4E and eIF4G that facilitates a composed eIF4E trRNP to promote steady-state translation (Haghighat et al., 1995;

Mader et al., 1995; Marcotrigiano et al., 1999). These trRNP dynamics are observed in response to mitogenic stimuli (Ma & Blenis, 2009; Pause et al., 1994). However, in cell

46 stress, such as amino acid deprivation or HIV-1 infection, mTOR activation is reduced

(Ma & Blenis, 2009; A. Sharma et al., 2012). This results in hypophosphorylation of 4E-

BP1 (Ma & Blenis, 2009; Pause et al., 1994; A. Sharma et al., 2012).

Hypophosphorylated 4E-BP1 exhibits a high affinity for eIF4E that competes with eIF4G for an association (Pause et al., 1994). The outcome is a 4E-BP1:eIF4E interaction that impairs eIF4E trRNP complex formation and results in suppression of eIF4E-mediated translation (Pause et al., 1994).

On the other hand, the CBC trRNP is insensitive to regulation by 4E-BP1 (Oh,

Kim, Cho, Choe, & Kim, 2007a; Oh, Kim, Choe, & Kim, 2007b; A. Sharma et al., 2012).

This results in its maintained translation activity during cell stress events, like hypoxia, serum starvation, and HIV-1 infection, which evoke suppression of mTOR activity and

4E-BP1 activation (Oh, Kim, Cho, Choe, & Kim, 2007a; Oh, Kim, Choe, & Kim, 2007b;

A. Sharma et al., 2012). In fact, no regulators of cytoplasmic CBC function have been identified. The nuclear activity of CBC in pre-mRNA splicing is sensitive to several extracellular stimuli, such as growth factors and UV irradiation, however the mechanisms governing this effect and its significance for downstream CBC trRNP function remain to be elucidated (Wilson et al., 1999). Likewise, CBC was identified as a phosphorylation target of the ribosomal protein S6 kinase (S6K) (Ruvinsky & Meyuhas, 2006). Yet the actual occurrence of CBC phosphorylation in cells and its significance for regulating

CBC trRNP activity remains ambiguous (Ruvinsky & Meyuhas, 2006).

A distinction between CBC trRNP regulation and that of eIF4E trRNP control is significant for eukaryotic translation in three main ways. First, it affords the cell a

47 mechanism to separate mRNA surveillance translation activities from that of steady-state protein synthesis. This allows for the maintenance of critical quality control functions in the advent of suppressed steady-state protein synthesis. Second, it provides two distinct molecular bases for cap-dependent translation control. This feature is critical for enacting targeted protein synthesis, especially during states of cellular stress. Third, it presents an opportunity for multiple integration mechanisms of invading viral pathogens into the host translation process. Collectively these features are critical to informing the work of this dissertation, which further investigates the significance of trRNP biology in eukaryotic translation control by assessing the molecular basis for the select RHA trRNP in targeted protein synthesis.

The RHA trRNP

RNA helicase A (RHA/DHX9) is a cellular DExH/D-box RNA helicase family member with critical roles in posttranscriptional gene control. Fundamental to its effect is the ability of RHA to engage distinct RNPs that have targeted outcomes on gene expression regulation. Consequently, RHA is a host protein with strong clinical significance. It is necessary for development, deregulated in breast, prostate, and lung cancer, a major autoantigen in systemic Lupus, and a critical stimulator of viruses that infect animal and human hosts. Thus, the study of RHA biology, and in particular its role in translation control, serves as a model system to investigate fundamental cellular processes and their significance for both animal and human health and disease.

As previously introduced, it is known that RHA is the host factor critical for the translation of retroviral transcripts and the cellular proto-oncogene junD (Bolinger et al.,

2007; 2010; Hartman et al., 2006; Ranji et al., 2011). This effect is brought about by its distinct association with the cis-acting posttranscriptional control element (PCE) that defines the 5' termini of these target transcripts (Bolinger et al., 2007; 2010; Hartman et al., 2006; Ranji et al., 2011). The PCE is a G-C-rich, dual stem-loop structure that functions as a positive cis-acting translational regulator (Bolinger et al., 2007; Butsch et al., 1999; Roberts & Boris-Lawrie, 2000; 2003). This effect is mediated through its association with the trans-acting host factor RHA (Bolinger et al., 2007; 2010; Hartman et al., 2006). Together RHA-PCE forms an RNA-protein interaction that gives rise to the distinct RHA trRNP important for targeted eukaryotic translation control. A main gap in knowledge is the molecular basis by which the RHA trRNP exerts its translational effect and the significance of this activity on the outcome of targeted protein synthesis.

To understand the significance of the RHA trRNP in coordinating dynamics of protein synthesis, two critical features of its activity must be considered. First, the transcripts it regulates are prominently expressed during cell stress when vast translational reprogramming occurs (Hernandez, Floyd, Weilbaecher, Green, & Boris-

Lawrie, 2008; A. Sharma et al., 2012). This implies functionality of the RHA trRNP during the cellular stress response. Second, the PCE does not function as an internal ribosome entry site (IRES) element; rather its effect is on cap-dependent translation control (Bolinger et al., 2007; Butsch et al., 1999). This indicates that the RHA trRNP regulates protein synthesis in a cap-dependent scanning manner. Thus, two seemingly

49 paradoxical features define the molecular basis for the RHA trRNP in translation control: cap-dependent initiation during cell stress.

This dissertation elucidates the molecular basis for the RHA trRNP in translation control by providing evidence that the select association of RHA with the non-canonical

CBP80/20 cap-binding protein complex (CBC) affords the RHA trRNP distinct functional activity during cell stress. By doing so it presents a new paradigm for eukaryotic protein synthesis in which cap-dependent gene expression is maintained during cell stress. Chapter 2 is the molecular characterization of the RHA trRNP composition and activity. Here a select association between RHA and CBC is demonstrated and shown to be functional in translation control during cell stress. Chapter

3 is an analysis of RHA's distinct role in influencing ribosome dynamics during the translation process. Here critical RHA residues are defined to have important functions in

43S pre-initiation ribosome complex recruitment, scanning, and 80S stabilization.

Chapter 4 is an investigation into the regulatory mechanisms controlling RHA trRNP activity. Here RHA is demonstrated to self-associate and interact with the structurally related DExH-box RNA helicase DHX30 in a manner that implies functional regulation.

The final chapter, Perspectives, discusses the significance of these findings in regards to our understanding of the molecular basis of eukaryotic translation control as well as their implications for our understanding of animal and human health and disease.

Figure 1.1 Schematic of canonical eukaryotic translation initiation. Canonical eukaryotic translation initiation is characterized by a cap-dependent scanning mechanism whereby the ribosome binds the 5' terminus of an mRNA and proceeds along the transcript inspecting, base-by-base, for an appropriate start codon to initiate polypeptide synthesis. This process begins with the activation of an mRNA and its association with the 43S pre-initiation ribosome complex (43S PIC). Interactions between the 5' associated scaffolding factor, CTIF or eIF4G, with the 3'-bound poly-A-binding protein (PABP) circularizes an mRNA to generate a translation ribonucleoprotein (trRNP) complex that engages with the 43S PIC. This occurs via stimulated interactions between the scaffolding factor and eIF3. 5' to 3' scanning follows, as the 43S PIC the searches for an optimal start codon (AUG) to initiate polypeptide synthesis. Directed movement is facilitated by initiation factors eIF1 and eIF1A. Their positioning within the 43S PIC stabilizes an 'open' conformation of the mRNA entry channel latch and orients the initiator methionyl-tRNA in an outward position that is conducive for scanning. Start codon recognition occurs when an AUG is reached within an optimal sequence context and appropriate anticodon:codon base pairing occurs. This induces a conformational rearrangement in the 43S PIC that causes the mRNA entry channel latch to close and the scanning process to stop. Repositioning of eIF1A stimulates eIF1 release and rotation of the initiator methionyl-tRNA into an inward position that is favorable for 60S ribosomal subunit joining. Joining of the large (60S) ribosomal subunit is facilitated by eIF5B and results in the formation of an elongation-competent 80S ribosome primed for polypeptide synthesis.

Figure 1.1 Schematic of canonical eukaryotic translation initiation. 52

Figure 1.2 Schematic of canonical eukaryotic translation elongation. Canonical eukaryotic translation elongation proceeds in three steps: amino acid incorporation, peptide bond formation, and ribosome translocation. Upon formation of an elongation- competent 80S ribosome at the start codon (AUG), the elongation factor eEF1 facilitates association of a charged tRNA with an available codon in the adjacent, 3' A-site. Recognition of appropriate anticodon:codon base pairing stimulates GTP hydrolysis and release of eEF1. Peptide bond formation is subsequently catalyzed by the peptidyl transferase center of the large ribosomal subunit. This results in a polypeptide chain of n+1 positioned with the A site of the ribosome. Ribosomal translocation to the next 3' codon is stimulated by peptide bond formation and completed with the association of elongation factor eEF2 and GTP hydrolysis. Repetition of these three steps along an open reading frame grows the encoded polypeptide.

Figure 1.2 Schematic of canonical eukaryotic translation elongation.

Figure 1.3 Schematic of canonical eukaryotic translation termination. Canonical eukaryotic translation termination is triggered by the recognition of a stop codon (e.g. UAA) within the A-site of an elongating ribosome. This recognition occurs by the eukaryotic translation termination factor eRF1 in complex with eukaryotic translation termination factor eRF3 and GTP. eRF1 harbors a conserved GGQ motif that positions within the peptidyl transferase center upon stop codon recognition and GTP hydrolysis. This movement of the GGQ motif into the peptidyl transferase center induces a conformational rearrangement in the large ribosomal subunit that causes hydrolysis of the ester bond linking the polypeptide chain to the P site tRNA, releasing the protein product. The ATP-binding cassette protein ABCE1 resolves the post-termination complex by splitting away the large (60S) ribosomal subunit from the now vacant 80S complexes. Initiation factors eIF1, 1A, and 3 facilitate subsequent separation and recycling of the deacylated tRNA, mRNA, and small (40S) ribosomal subunit.

Figure 1.3 Schematic of canonical eukaryotic translation termination.

Figure 1.4 Overview of mTOR-directed eukaryotic translation control. Mammalian target of rapamycin (mTOR) is a serine/threonine kinase that serves as the central commander of eukaryotic translation control. It directs protein synthesis in reflection of the cellular environment by acting upon two downstream effectors: ribosomal protein S6 kinase (S6K) and eIF4E binding protein 1 (4E-BP1). S6K targets several factors critical in the regulation of translation. These include: the 40S ribosomal subunit protein S6 (rpS6), eEF2 kinase (eEF2K), CBP80 of the CBP80/20 cap-binding complex, the S6K1 Aly/REF-like target (SKAR), regulatory protein programmed cell death 4 (PDCD4), and translation initiation factor eIF4B. The phosphorylation of these translation factors by mTOR-S6K signaling alters their function in a manner that facilitates translation activity. For eEF2K, its phosphorylation by S6K results in its inactivation. This allows elongation factor eEF2 to effectively associate with elongating ribosomes and facilitate their translocation. Similarly, the phosphorylation of PDCD4 results in its inactivation to facilitate steady-state protein synthesis. Here S6K-mediated phosphorylation of PDCD4 induces its ubiquitinylation and subsequent degradation. This removes PDCD4 as an inhibitor of eIF4A helicase activity, allowing translation initiation to proceed. Likewise, the phosphorylation of eIF4B by S6K also facilitates effective steady-state protein synthesis by stimulating eIF4A helicase activity. In this instance, phosphorylated eIF4B exhibits enhanced association for eIF4A, which effectively stimulates translation activity. The direct effects of mTOR/S6K-induced phosphorylation on rpS6 and CBP80 translation activity remain ambiguous, however, data support a stimulatory outcome on initiation. For the second line of command, mTOR-directed, 4E-BP1 signaling controls translation by regulating trRNP dynamics. Here phosphorylation of 4E-BP1 by mTOR reduces its affinity for the eIF4E cap-binding protein. This allows eIF4E to effectively associate with the translation initiation factor eIF4G and create a trRNP that is productive for steady-state protein synthesis.

Figure 1.5 Regulation of 4E-BP1 phosphorylation events and its role in eukaryotic translation control. The eIF4E binding protein 1 (4E-BP1) is a critical effector of translation activity. Its function is dependent upon upstream signaling events that converge on the regulation of its major commander, the mammalian target of rapamycin (mTOR). In the presence of mitogenic stimuli (e.g. growth factors), mTOR is activated and results in the hyperphosphorylation of 4E-BP1. The hyperphosphorylation of 4E-BP1 is a hierarchical event with phosphorylation at threonine (Thr) residues 37 and 46 priming for phosphorylation at Thr70, which then allows for phosphorylation at serine (Ser) residue 65. Phosphorylation at Ser65 is the critical effector posttranslational modification that regulates the functional effects of 4E-BP1 in translation control. In its presence, 4E-BP1 exhibits reduced affinity for the eIF4E cap-binding protein. This allows eIF4E to associate with the translation initiation factor and scaffolding protein eIF4G. An eIF4E-eIF4G association allows for effective trRNP formation that facilitates translation activity. In the advent of stress, however, mTOR activity is reduced to result in hypophosphorylation at Ser65. This change in the posttranslational modification status of 4E-BP1 increases its affinity for eIF4E. A 4E-BP1-eIF4E association effectively inhibits an interaction between eIF4E and eIF4G. This results in impaired trRNP formation and inhibited translation activity. It has also been shown that phosphorylation at Ser101 of 4E-BP1 regulates phosphorylation at Ser65 and that phosphorylation at Ser112 facilitates release of 4E-BP1 from eIF4E.

Figure 1.6 Integration of stress response signals into the phosphorylation of eIF2α and its outcome on regulated eukaryotic protein synthesis. Integral to the function of translation initiation factor eIF2 is its ability to function as a GTPase. eIF2-GTP drives ternary complex formation and is necessary for 43S pre-initiation ribosome complex assembly, recruitment, and scanning. Start codon recognition triggers GTP hydrolysis, an outcome that induces conformational rearrangements that halt the scanning process and engage 80S ribosome complex formation. The exchange of GDP for GTP on eIF2 is necessary for subsequent re-engagement of eIF2-driven translation initiation. This requires the guanine nucleotide exchange factor eIF2B, an effect that is highly regulated by the cellular environment. eIF2 consists of 3 subunits: α, β, and γ. The α subunit is the critical allosteric effector of direct GTP binding. Phosphorylation of the α subunit at serine (Ser) residue 51 inhibits the exchange of GDP for GTP on eIF2. This effectively impedes eIF2-mediated initiation events, resulting in compromised translation activity. Phosphorylation of eIF2α occurs by four distinct kinases, each of which is stimulated by distinct cellular stressors. Haem-regulated inhibitor kinase (HRI) is activated by iron deficiency and osmotic or heat shock, protein kinase R (PKR) is stimulated by double- stranded RNA, PKR-like endoplasmic reticulum kinase (PERK) responds to ER-stress and hypoxia, and general control non-derepressible-2 (GCN2) is activated by amino-acid deprivation and UV irradiation. The outcome in all instances is phosphorylation of eIF2α, which impairs the exchange of GDP for GTP to result in suppression of general translation.

Figure 1.7 Model of CBC and eIF4E trRNP dynamics in the control of eukaryotic translation. Canonical eukaryotic cap-dependent translation is governed by a complex interplay between the CBC trRNP and the eIF4E trRNP. Beginning at transcription, the CBP80/20 cap-binding protein complex (CBC) directly binds the 5' 7-methyl-guanosine cap of a nascent transcript (1). This association is driven by the high affinity of CBC for 7-methyl-guanosine and its abundant nuclear localization. In this RNP association, CBC facilitates mRNA maturation and its nuclear export. Upon entry into the cytoplasm, CBC engages trRNP formation that facilitates the initial round(s) of translation (2). This involves the direct association of CBP80 with either CTIF or eIF4G, which provides a molecular bridge to eIF3 binding and 43S pre-initiation ribosome complex recruitment. This CBC trRNP also harbors nuclear and cytoplasmic poly-A binding protein (PABP) at its 3' terminus and exon junction complexes (EJC) along the open reading frame. These features are critical for the CBC trRNP in engaging translation activity by facilitating interactions between the target transcript and the 43S pre-initiation ribosome complex. Remodeling of a transcript from a CBC trRNP to an eIF4E trRNP follows the pioneer round(s) of translation and is critical for regulated steady-state protein synthesis (3). This trRNP remodel is facilitated in two ways. At the 5' terminus, exchange of CBC for eIF4E is governed by a translation-independent mechanism whereby the strong affinity of CBP80 for the nuclear import factor importin α1 (IMP α1) drives its association with the nuclear pore-associated karyopherin importin β1 (IMP β1). This CBC-IMP α1-IMP β1 complex formation results in CBC destabilization from the 5' cap and its nuclear recycling. The association between CBC and IMP α1 is driven by the canonical nuclear localization sequence (NLS) found within the N-terminus of CBP80. IMP α1 recruits IMP β1 via its IMP β1-binding domain (IBB). Displacement of CBC from the 5' cap allows eIF4E to bind. eIF4E then draws stabilized associations with eIF4G that recruit eIF4A and complete formation of the eIF4E trRNP at the 5' terminus. Removal of exon junction complexes and the complete exchange of PABPN for PABPC occur in a translation-dependent manner, although, the exact molecular mechanisms remain unknown.

Figure 1.7 Model of CBC and eIF4E trRNP dynamics in the control of eukaryotic translation.

RNA helicase A DDX2 (eIF4A) DDX48 (eIF4AIII) DDX3 DHX29 (RHA/DHX9) 1. Cognate cis- Posttranscriptional acting mRNA (CGG)4 motif/ Long, structured, Long, structured, control element element G-quadruplex Stable stem loop and G-C rich and G-C rich (PCE) 2. Cap-dependent Enhancer OR translation Enhancer Enhancer repressor Enhancer Enhancer CBC trRNP + + - ? eIF4E trRNP + - + + ? mRNA unwinding AND mRNP mRNA unwinding to mRNA unwinding to remodeling to mRNP remodeling to promote 43S PIC promote 43S PIC facilitate 80S facilitate 48S Mechanism scanning scanning formation formation ? Enhancer of select type II; Interferes Inferred not 3. IRES-mediated Enhancer of type I Enhancer of HCV with select type III influential because translation and II (type III) and type IV PCE inactive Table 1.1 Known features of select DExH/D-box RNA helicase trRNPs.

Chapter 2 : Selective Cap-Dependent Translation During Cell Stress is Maintained

by the RNA Helicase A-CBP80/20 Translation RNP

ABSTRACT

Cell stress is a major effector of eukaryotic translation control. A hallmark feature is suppression of cap-dependent protein synthesis in order to facilitate cell survival and recovery. Yet targeted translation activity is fundamental to the cellular stress response and pathogenic HIV-1 infection. HIV-1 structural transcripts specifically maintain cap- dependent translation activity to promote their expression. This effect is brought about by their engagement in a distinct translation ribonucleoprotein (trRNP) complex that is defined by the host factor RNA helicase A (RHA). RHA selectively recognizes and binds the cis-acting posttranscriptional control element (PCE) that characterizes the 5' termini of HIV-1 structural mRNAs. Together RHA-PCE forms a distinct RHA trRNP functional in cap-dependent translation. Yet to be defined, however, is a complete characterization of the RHA trRNP and its significance for informing the role of RHA in maintained cap- dependent translation during cell stress. Herein the RHA trRNP is characterized to distinctly include the non-canonical CBP80/20 (CBC) cap-binding protein complex.

RHA and CBC are shown to interact in the cytoplasm and on polysomes, characteristics that exemplify an active trRNP. Notably, the cytoplasmic and polysome association of

RHA with CBC is maintained during serum-deprivation, torin-1 mediated mTOR inhibition, and HIV-1 expression, all mechanisms that evoke cell stress and suppression of canonical eIF4E cap-dependent translation. This observed effect correlates with sustained interactions between RHA, CBC, and the HIV-1 gag mRNA on polysomes and, most notably, maintained cap-dependent PCE translation. We therefore propose a new paradigm for eukaryotic translation control in that selective cap-dependent protein synthesis during cell stress is maintained by the RHA-CBC trRNP. This affords a targeted molecular basis for expression of pathogenic HIV-1 with critical implications for the regulated expression of stress response proteins.

INTRODUCTION

Protein synthesis is fundamental to the survival of all cells and viruses.

Deregulated translation activity is associated with cancer, neurological diseases and disorders, neurodegeneration, growth defects, and innate immune disorders (Carpenter,

Ricci, Mercier, Moore, & Fitzgerald, 2014; Ruggero, 2013; Scheper, van der Knaap, &

Proud, 2007; Silvera, Formenti, & Schneider, 2010). The cellular act of protein synthesis is a dynamic molecular process choreographed by ribonucleoprotein (RNP) complexes that assemble upon an mRNA and coordinate progression through the three mechanical stages: initiation, elongation, and termination. Translation initiation is the well-known rate-limiting step with tight regulation occurring by both intracellular and extracellular signaling pathways. In eukaryotes, translation initiation is a cap-dependent scanning mechanism whereby critical translation RNP (trRNP) complexes form and control

64 ribosome recruitment and movement along the 5' terminus to initiate protein synthesis at an appropriate start codon(Jackson et al., 2010).

Two core trRNP complexes characterize eukaryotic translation initiation: the

CBC trRNP and the eIF4E trRNP. The CBC trRNP is defined by its distinctive heterodimeric CBP80/20 cap-binding protein complex (CBC) and takes precedence in translation initiation during the so-called pioneer rounds. Here the CBC trRNP is essential for ribosome-mediated mRNA surveillance and quality control (Maquat, Tarn,

& Isken, 2010). The eIF4E trRNP is distinguished from the CBC trRNP by its defining eIF4E cap-binding protein and dynamically assembles upon an mRNA subsequent to

CBC to facilitate steady-state protein synthesis (Maquat et al., 2010). Despite these differing compositions and significances in translation control, both the CBC trRNP and the eIF4E trRNP functionally interact with the ribosome in a similar manner to regulate its recruitment and activity (Maquat et al., 2010). But it is their distinct purposes in doing so that results in altered mechanisms of targeted regulation. The eIF4E trRNP is highly sensitive to cellular stress, especially environmental changes and viral infections, as the cell aims to minimize steady-state protein synthesis and allocate critical energy and resources for recovery (Holcik & Sonenberg, 2005; B. Liu & Qian, 2014). Conversely, the CBC trRNP is functionally active during the cellular stress response in order to sustain critical mRNA surveillance and quality control activity (Oh, Kim, Cho, Choe, &

Kim, 2007a; Oh, Kim, Choe, & Kim, 2007b).

Although the CBC trRNP and the eIF4E trRNP characterize general eukaryotic translation control, there is a growing list of mechanisms for targeted translational

65 regulation whereby mRNAs are placed within distinct trRNPs to control their expression in a specified manner. These distinct trRNPs arise from the association of specific RNA binding proteins with cognate cis-acting RNA elements that are intrinsic to particular transcripts. The outcomes of these associations are novel mechanisms for translational control that fine-tune gene expression in a selective manner.

The cellular stress response exemplifies targeted translational control. Here the cell engages distinct trRNPs to facilitate expression of critical stress response proteins in the advent of suppressed eIF4E trRNP activity. These distinct trRNPs are often composed of RNA binding proteins known as IRES trans-acting factors (ITAFs) that bind intrinsic cis-acting RNA elements and facilitate ribosome recruitment and activity independent of the need for functional cap-binding proteins (Spriggs et al., 2008). More recent evidence, however, supports adapted CBC trRNPs as novel mechanisms for targeted translational control (Choe et al., 2013; A. Sharma et al., 2012). In these instances, the CBC trRNP is recruited for steady-state expression of target mRNAs beyond the pioneer rounds, relieving their dependence on eIF4E for robust protein production (Choe et al., 2013; A.

Sharma et al., 2012). In the case of replication-dependent histone mRNAs, this functional modification of the CBC trRNP is facilitated by the RNA binding protein SLBP, which specifically recognizes and binds the distinct 3' stem loop structure that defines these mRNAs (Choe et al., 2013). Together SLBP and CBC generate a distinct trRNP important for targeted regulation of histone mRNA expression that is intricately linked to the cell cycle (Choe et al., 2013; Choe, Ahn, & Kim, 2014a). In the case of HIV-1

66 structural mRNAs, the molecular basis by which the CBC trRNP is adaptively engaged for robust viral protein production has yet to be defined (A. Sharma et al., 2012).

We previously established that cap-dependent translation of HIV-1 structural mRNAs and the cellular proto-oncogene junD are regulated by the trans-acting host factor RNA helicase A (RHA) and its cis-acting 5' RNA element, the posttranscriptional control element (PCE) (Bolinger et al., 2007; 2010; Hartman et al., 2006; Ranji et al.,

2011). The PCE does not function as an internal ribosome entry site (IRES) element; rather, its effect is on cap-dependent translation control (Bolinger et al., 2007; Butsch et al., 1999; Roberts & Boris-Lawrie, 2000). Thus, critical to the mechanism of RHA-PCE- dependent translational control is the recognition and association of RHA with distinct structural features of the PCE to generate a cap-associated trRNP productive for robust protein synthesis (Bolinger et al., 2007; 2010; Hartman et al., 2006; Ranji et al., 2011).

An unresolved issue, however, is a complete characterization of the RHA trRNP and how this distinct trRNP informs the activity of RHA in targeted cap-dependent translation control.

Herein the RHA trRNP is characterized to selectively include the non-canonical

CBC cap-binding protein complex. RHA is shown to selectively associate with CBC in the cytoplasm and on polysomes, features that support an active RHA-CBC trRNP.

Notably, we show that a functional interaction between RHA and CBC is maintained upon environmental, drug, or viral-induced activation of the eIF4E trRNP repressor protein 4E-BP1 and that this RHA-CBC trRNP is functional in selective protein synthesis during the cell stress event. Taken together, our data indicate the RHA trRNP as a

67 specialized adaptation of the CBC trRNP that functions to maintain selective cap- dependent translation during cell stress.

MATERIALS AND METHODS

Cell culture and transfections

HEK293 cells were grown in Eagle's Minimum Essential Medium (ATCC) supplemented with 10% FBS (Gibco) and 1% antibiotic (Gibco). One day prior to transfection, cells were seeded in 10 cm plates at a density of 1.5×106 cells/plate. A total of 2 plates were prepared for each gradient experiment and 1 plate for each cellular IP assay. Each plate of cells was transfected with 2 μg of designated FLAG-RHA plasmid

(see Plasmids for more details) or 3 μg of HIV-1 pr- using the X-tremeGene HP DNA

Transfection Reagent (Roche) at a ratio of 1:3. Media was changed 24 hr post- transfection and cells harvested at forty-eight. CEMx174 cells were grown in RPMI supplemented with 10% FBS and 1% antibiotic.

For luciferase reporter assays, HEK293 cells were seeded in 6 well plates at a density of 3×105 cells/well. Each well was transfected with 0.5 μg of designated luciferase reporter construct using the X-tremeGene HP DNA Transfection Reagent

(Roche) at a ratio of 1:3. Serum deprivation was conducted at 24 hr post-transfection.

Serum deprivation experiments were conducted by washing plated HEK293 cells three times with 1x PBS and then culturing in complete media containing 0.5% FBS for

24 hr. Torin 1 treatment of HEK293 cells was performed by culturing with 250 nM torin

1 (Tocris 4247) for 1 hr (Thoreen et al., 2009).

SDS-PAGE and Western Blotting

Protein samples were run on 4-15% gradient SDS-polyacrylamide gels (BIO-

RAD) at 200 V for 45 minutes and transferred to nitrocellulose membranes (BIO-RAD) via a wet transfer at 80 V for 1 hour by standard methods. Membranes were incubated in

5% non-fat dry milk in PBS-T with the addition of designated primary antibody overnight at 4 degrees. Blots were washed 3 X 5 min with PBS-T after antibody incubation. HRP-conjugated secondary antibodies (GE Healthcare) were used at 1:5000 or 1:10,000 in 5% milk/PBS-T for 1 hour at RT, washed as before, and HRP signal detected by Amersham ECL Prime Western Blotting Detection Reagent (GE Healthcare).

For endogenous IPs, the Clean-Blot IP Detection Reagent (ThermoScientific) was used in place of HRP-conjugated secondary antibodies. The Fuji Imaging System (FUJIFILM) and MultiGauge Software Program (FUJIFILM) were used to develop and assess protein band intensity.

Antibodies

The following primary antibodies were used for Western blot analysis: anti-FLAG

(Sigma, F7425) at 1:5000, anti-HA (Covance, MMS-101R) at 1:1000, anti-RHA

(Vaxron, PA-001) at 1:7500, anti-DDX3 (Bethyl, A300-474A) at 1:1000, anti-

NCBP1/CBP80 (Bethyl, A301-793A) at 1:1000, anti-phospho-Ser209 eIF4E (Cell

Signaling, #9741S), anti-eIF4E (Cell Signaling, #9742S) at 1:1000, anti-eIF4G (Cell

Signaling, #2498S) at 1:1000, anti-eIF4A1 (Cell Signaling, #2490S) at 1:1000, anti-

PABPC (Abcam, ab6125) at 1:1000, anti-S6 ribosomal protein (Cell Signaling, #2217) at

1:1000, anti-L5 ribosomal protein (Bethyl, A303-933A) at 1:1000, anti-eIF2α (Cell

Signaling, #9722) at 1:1000, anti-phospho-Ser51 eIF2α (Cell Signaling, #3597S) at

1:1000, anti-4EBP1 (Cell Signaling, #9542) at 1:1000, anti-phospho-Ser65 4EBP1 (Cell

Signaling, #9451) at 1:1000, anti-α-tubulin (SantaCruz, sc-23948) at 1:1000, anti-CTIF

(gift from Yoon Ki Kim) at 1:1000, and anti-HIV-1 p24 (NIH) at 1:1000.

For FLAG and HA IPs, anti-FLAG M2 affinity gel (Sigma, A2220) or anti-HA agarose (Sigma, A2095) was used in the IP step of the FLAG- or HA-conjugated constructs, respectively. For endogenous IPs, anti-RHA (Bethyl, A300-855A), anti-

DDX3 (Bethyl, A300-474A), anti-NCBP1/CBP80 (Bethyl, A301-794A), or anti-eIF4E

(Santa Cruz, sc-9976AC) was used according to manufacture recommendations based upon the concentration of input cell lysate.

Plasmids

The FLAG-RHA and luciferase reporter constructs were previously described

(Bolinger et al., 2010; Hartman et al., 2006; Ranji et al., 2011; Roberts & Boris-Lawrie,

2000). HA-CBP80 and 2xFLAG-eIF4E were gifts from Alan Cochrane. FLAG-CTIF was a gift from Yoon Ki Kim. HIV-1 pr- was a gift from Lawrence Kleinman.

Polysome Profiling

HEK293 cells were treated with 0.1 mg/mL cyclohexamide (CHX; prepared in

DMSO, Sigma C1988) for 20 minutes, then washed twice with ice cold PBS with CHX, and scraped off the plates in 5 mL of ice cold PBS with CHX. Resuspended cells from

70 both plates were combined in 15 mL conical tubes and then pelleted at 1500 rpm for 4 min at 4C. Supernatant was removed and cell pellet resuspended in 0.75 mL of low salt buffer (20 mM Tris-HCl, pH 7.5, 3 mM MgCl2, 10 mM NaCl, 2 mM DTT, 1X protease inhibitor cocktail EDTA-free (Roche 11 873 580 001), 5 μL/mL RNase Out). Cells were allowed to swell on ice for 5 min and then lysed by the addition of 0.25 mL lysis buffer

(0.2 M sucrose and 1.2% Triton X-100 prepared in low salt buffer) followed by 10 strokes with a dounce homogenizer (Kimble Chase). Lysate was purified by spinning at top speed (16,000 x g) for 1 minute at 4C, then layered onto a 10-50% sucrose gradient

(10 mM Tris-HCL, pH 7.5, 5 mM MgCl2, 100 mM KCl, 2 mM DTT, 0.1 mg/mL CHX,

10/50% sucrose), and spun in an ultracentrifuge for 2 hr 40 min at 35,000 RPM at 4C.

Loading equivalent amounts of RNA OD units across samples for each experiment normalized the profiles. Polysome profiles were generated by continuous monitoring of

RNA absorbance at 254 nm by the ISCO UA-6 Absorbance Detector unit and fractionated into 22 equivalent volume (0.5 mL) fractions using the ISCO Foxy R1 fraction collector. Brandel Peak Trace Software was used to generate the corresponding profile traces.

For assays with harringtonine, transfected cells were treated with 2 μg/mL harringtonine (Abcam 141941) for 2 min then immediately incubated with 0.1 mg/mL

CHX for 1 min and lysed. Puromycin treatment was performed by treating transfected cells with 400 nM puromycin (Sigma P7255) for 1 hr and then incubating with 0.1 mg/mL CHX for 10 min.

Protein Association with Polysome Fractions

Samples from representative fractions across each sucrose gradient profile were combined with 1 mg/mL BSA and brought to a total volume of 1 mL by the addition of polysome gradient buffer (10 mM Tris-HCL, pH 7.5, 5 mM MgCl2, 100 mM KCl). Ice- cold TCA was added to a final concentration of 20% and the fractions were incubated overnight at -20C. Samples were then spun at 12,000 rpm for 20 min at 4C and supernatant decanted. Two washes with 0.5 mL ice-cold acetone followed. The protein pellet was resuspended in low salt buffer (20 mM Tris-HCl, pH 7.5, 3 mM MgCl2, 10 mM NaCl) and equivalent volumes of precipitated protein from each fraction were loaded directly onto the gels after the addition of loading buffer. 1% of the initial cleared lysate was diluted with low salt buffer and run as INPUT for each gradient.

Immunoprecipitations

For FLAG or HA IPs, anti-FLAG- or HA-conjugated agarose beads (Sigma) were pre-blocked for 1 hr in 0.5% BSA, washed, and then incubated with harvested cell lysate for 3 hr at 4C with end-over-end rotation. Total volume of each IP was brought to 1 mL with 1x wash buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl). Immunoprecipitated complexes were washed three times with NETN-150 wash buffer (20 mM Tris-HCl, pH

7.4, 150 mM NaCl, 0.1 mM EDTA, 0.5% NP40) and twice with 1x wash buffer. Loading buffer was added to isolated bead complexes, heated, and loaded directly onto the gel.

2% of the initial cleared lysate was reserved and diluted with 1x wash buffer and run as

INPUT for each IP. 2% of the flow-through following IP was reserved and run as FT for each IP.

For endogenous IPs, designated primary antibody was incubated with magnetic

Dynabeads Protein A (Invitrogen) for 1 hr at RT with end-over-end rotation. Antibody- bead complexes were washed with 1x wash buffer and then incubated with harvested cell lysate for 3 hr at 4C with end-over-end rotation. Total volume of each IP was brought to

1 mL with 1x wash buffer. Immunoprecipitated complexes were then washed three times with NETN-150 wash buffer. Samples were prepared and loaded directly onto the gel as described.

For polysome IPs, alternating sucrose gradient fractions across each polysome profile were diluted with 0.5 mL of 1x wash buffer and then indicated complexes were immunoprecipitated as described.

RNA Isolation and Analysis

For RNA immunoprecipitations, RNA was extracted from isolated IP complexes using the Trizol LS Reagent (Invitrogen) and purified using the RNeasy Clean-Up Kit

(QIAGEN) according to manufacture instructions. Contaminating DNA was removed using TURBO DNase (Ambion) according to manufacture instructions. cDNA was prepared with equal volumes of purified RNA from each IP using random hexamers

(Invitrogen) and Omniscript Reverse Transcriptase (QIAGEN). 1/20th of the RT reaction was used as input for PCR. HIV-1 gag PCR was performed as previously described (A.

Sharma et al., 2012).

For analysis of cytoplasmic RNA, RNA was extracted and purified as described above. cDNA was prepared with 100 ng of purified RNA as input. Gene-specific PCR was performed as previously described (S. Qian et al., 2009; A. Sharma et al., 2012).

Luciferase Reporter Assay

Transfected HEK293 cells were washed once with PBS and then lysed in 100 μL passive lysis buffer (Promega). Assessment of firefly luciferase activity was performed with 5 μL of lysate and the Luciferase Assay System (Promega) according to manufacture instructions.

RESULTS

RHA selectively interacts with the CBP80/20 cap-binding protein complex (CBC) in the cytoplasm.

Given the CBC and eIF4E trRNPs coordinate eukaryotic cap-dependent translation (Maquat et al., 2010) and that RHA selectively facilitates this process

(Bolinger et al., 2007; 2010; Hartman et al., 2006; Ranji et al., 2011), we hypothesized that the RHA trRNP is a distinct adaptation of either the CBC trRNP or the eIF4E trRNP and that these specified interactions are critical to informing the role of RHA in translation control (Figure 2.1a).

To investigate these possible cellular interactions, we immunoprecipitated endogenous RHA from cytoplasmic cellular extract and performed immunoblot analysis to assess for the co-precipitation of the two cap-binding proteins that define either the

CBC trRNP or the eIF4E trRNP: the CBP80/20 cap complex (CBC) or the eIF4E cap- binding protein, respectively (Figure 2.1a). As shown in Figure 2.1b, endogenous RHA selectively co-precipitated CBP80 of the CBC. In comparison, related DEAD-box RNA helicase DDX3 selectively co-precipitated eIF4E (Figure 2.1b). DDX3 is the host factor important for targeted translation control of long or structured cellular transcripts and the hepatitis C virus (Lai et al., 2008). Intricate to the molecular basis of the DDX3 trRNP is its select association with eIF4E (Shih et al., 2008; 2012). Our result corroborated these previous findings and notably distinguished RHA by its select interaction with CBC.

To further characterize the RHA trRNP, we assessed the cytoplasmic interaction of RHA with additional defining components of the CBC trRNP and the eIF4E trRNP by co-immunoprecipitation analysis (Figure 2.1a and c). When RHA selectively co- precipitated CBP80 from cytoplasmic cellular extract it also co-precipitated the CBC- associated scaffolding protein eIF4G but at reduced levels compared to CBP80 (Figure

2.1c). An interaction between RHA and the alternative CBC-associated MIF4G domain- containing protein CTIF was observed upon chemical crosslinking or exogenous expression of an epitope-tagged FLAG-CTIF (Figure 2.3). Distinctly, when RHA co- precipitated CBP80 from the cytoplasm it also co-precipitated the cytoplasmic poly-A binding protein (PABPC), which is a critical 3' end-binding RNA binding protein that characterizes active trRNP complexes (Figure 2.1a and c). The select interaction of RHA with CBP80, eIF4G, and PABPC was additionally observed following exogenous expression and immunoprecipitation of FLAG-RHA (Figure 2.2a). Neither endogenous

RHA nor FLAG-RHA demonstrated an observable co-precipitation of the eIF4E cap-

75 binding protein or the eIF4A1 RNA helicase of the eIF4E trRNP (Figure 2.1a and c and

Figure 2.2a). The specific interaction of RHA with CBP80 over eIF4E was seen both in non-crosslinked and chemically crosslinked cells (Figure 2.3). These results indicate a select interaction of RHA with the CBC trRNP.

We next performed reciprocal co-immunoprecipitations to support these findings.

Endogenous CBP80 selectively co-precipitated RHA from cytoplasmic cellular extract whereas eIF4E-RHA complexes were not observed (Figure 2.1d). Consistent with the findings in Figure 2.1c, the isolated CBP80-RHA complexes also contained eIF4G and

PABPC to support the isolation of active CBC trRNP complexes (Figure 2.1d). A select association between CBP80 and RHA was further observed upon exogenous expression and immunoprecipitation of a HA-CBP80 (Figure 2.2b). An interaction between RHA and FLAG-eIF4E was not identified (Figure 2.2b). Taken together, these cellular interaction studies demonstrate that the RHA trRNP selectively includes the CBP80/20 cap-binding protein (CBC).

Cytoplasmic RHA-CBC complexes exhibit characteristics of an active trRNP.

To assess the identified RHA-CBC cytoplasmic complexes as functional trRNPs, we analyzed the co-distribution and interaction of FLAG-RHA with CBP80 across a polysome profile. To this end, cytoplasmic cell lysate was layered upon a 10-50% sucrose gradient and separated by ultracentrifugation. The distribution or enrichment of

FLAG-RHA protein complexes was then analyzed in designated sub-polysome or polysome fractions based upon the traced absorbance of the ribosomal RNA. We

76 observed that FLAG-RHA was efficiently detected and immunoprecipitated across all analyzed gradient fractions with particular prominence in the 80S and early polysome peaks (Figure 2.4a). This distribution reflected previous observations for endogenous

RHA distribution across a polysome profile to support a role for RHA in translation control that can be studied with an exogenously expressed FLAG-tagged construct

(Hartman et al., 2006). Notably, we observed that FLAG-RHA selectively co-precipitated

CBP80 and the small ribosomal S6 subunit from the 80S and early polysome peaks

(Figure 2.4a). This finding indicated that cytoplasmic RHA-CBP80 complexes are active trRNPs, as they co-sediment with the ribosome in functionally characterized translation fractions of a sucrose gradient. Consistent with the findings in Figure 2.2, FLAG-RHA did not co-precipitate eIF4E across the polysome profile (Figure 2.4a).

To demonstrate that the identified RHA-CBP80 polysome complexes are active in translation, we assessed their sensitivity to general translation inhibition with the initiation inhibitor harringtonine. Harringtonine traps initiated 80S complexes at the start codon by blocking recruitment of subsequent amino-acyl transfer RNAs and interfering with peptidyl-transferase activity (Fresno, Jiménez, & Vázquez, 1977). This effectively results in the completion of active translation without new initiation (Fresno et al., 1977;

Ingolia, Lareau, & Weissman, 2011). The outcome in a profile analysis is a decrease in total cellular polysomes and an accumulation of 80S monosomes, as ribosomes run-off transcripts without being able to re-initiate (Fresno et al., 1977; Ingolia et al., 2011).

Sensitivity of trRNP complexes to harringtonine, as evident by a shift in their distribution and interaction towards sub-polysome gradient fractions, supports an active role in

77 coordinated translation control. To this end, cells transfected with FLAG-RHA were treated with harringtonine for two minutes (Ingolia et al., 2011; Ingolia, Brar, Rouskin,

McGeachy, & Weissman, 2012), processed, and then layered upon a 10-50% sucrose gradient and separated by ultracentrifugation. As conducted in panel a, the distribution or enrichment of FLAG-RHA protein complexes was then analyzed in designated sub- polysome or polysome fractions based upon the traced absorbance of the ribosomal RNA.

A two-minute harringtonine treatment resulted in a significant decrease in total cellular polysomes and an accumulation of 80S monosomes within FLAG-RHA transfected cells, as assessed by the traced absorbance of the ribosomal RNA (Figure 2.4, compare polysome and sub-polysome peak heights from the ribosomal RNA profile trace in b to that in a). This characteristic change in the profile supported effective drug- induced translation inhibition and ribosome run-off, as previously reported (Fresno et al.,

1977; Ingolia et al., 2011). Correspondingly, FLAG-RHA exhibited a significant shift in its distribution towards sub-polysome fractions in the presence of harringtonine (Figure

2.4, compare FLAG-RHA panels in b to those in a). This sensitivity of FLAG-RHA to general inhibition of translation initiation corroborated its established role in translation control (Hartman et al., 2006). Notably, we observed a shift in the distribution of co- immunoprecipitated FLAG-RHA-CBP80 complexes from the actively translating 80S and early polysome fractions to that of the sub-polysome fractions in the presence of harringtonine (Figure 2.4, compare IP panels in b to those in a). A similar shift in the distribution of FLAG-RHA-rpS6 complexes was also observed to further indicate the sensitivity of identified RHA trRNP complexes to inhibition of translation initiation

(Figure 2.4, compare IP panels in b to those in a). Harringtonine treatment did not affect the interaction between FLAG-RHA with eIF4E (Figure 2.4b).

To further demonstrate that FLAG-RHA-CBP80 polysome complexes exhibit functional characteristics of an active trRNP, we assessed their sensitivity to general translation inhibition with the elongation inhibitor puromycin. Puromycin is an aminoacyl-transfer RNA analog that binds the ribosome A-site and causes premature chain termination and ribosome release by accepting nascent polypeptide chains (Blobel &

Sabatini, 1971; Cannon, 1968). The outcome in a profile analysis is a decrease in total cellular polysomes and an accumulation of sub-polysome complexes, as elongating ribosomes are prematurely terminated and dissociated (Blobel & Sabatini, 1971; Cannon,

1968). Here sensitivity to puromycin is indicated by the accumulation of trRNP complexes in sub-polysome gradient fractions upon drug treatment.

A one-hour puromycin treatment resulted in a significant decrease in total cellular polysomes and an accumulation of sub-polysome complexes within FLAG-RHA transfected cells, as expected and assessed by the traced absorbance of the ribosomal

RNA (Figure 2.4c). Correspondingly, FLAG-RHA exhibited a significant shift in its distribution towards sub-polysome fractions in the presence of puromycin (Figure 2.4c).

This sensitivity of FLAG-RHA to general inhibition of translation elongation reflected our previously reported outcome for endogenous RHA gradient distribution in the presence of puromycin (Hartman et al., 2006). Notably, we observed a shift in the distribution of co-immunoprecipitated FLAG-RHA-CBP80 complexes from the actively translating 80S and early polysome fractions to that of the sub-polysome fractions in the

79 presence of puromycin (Figure 2.4c). Likewise, a similar shift in the distribution of

FLAG-RHA-rpS6 complexes was also observed to further indicate the sensitivity of identified RHA trRNP complexes to inhibition of translation elongation (Figure 2.4c). As with harringtonine, puromycin treatment did not affect the interaction between FLAG-

RHA and eIF4E (Figure 2.4c). Taken together, these polysome analyses indicate cytoplasmic RHA-CBC complexes as active trRNPs.

RHA and CBC maintain a trRNP interaction and polysome association during stress-induced activation of 4E-BP1.

The above results show that a select association with CBC characterizes the RHA trRNP. A distinguishing feature of CBC is its maintained translation activity during cell stress (Maquat et al., 2010). Cell stress is characterized by vast translational reprogramming, an effect required for cell recovery and survival. Often this molecular event involves suppression of the serine/threonine kinase mammalian target of rapamycin

(mTOR) (Ma & Blenis, 2009). mTOR regulates the bulk of cellular translation by controlling the activation of the eIF4E repressor protein 4E-binding protein 1 (4E-BP1)

(Ma & Blenis, 2009; Ruvinsky & Meyuhas, 2006). During cell stress, suppression of mTOR activity results in hypophosphorylated 4E-BP1(Ma & Blenis, 2009; Pause et al.,

1994). Hypophosphorylated 4E-BP1 binds with strong affinity to eIF4E (Ma & Blenis,

2009; Pause et al., 1994). The outcome is impaired eIF4E trRNP dynamics and compromised translation activity (Ma & Blenis, 2009; Pause et al., 1994). Yet targeted synthesis of critical stress response proteins is essential to the cell recovery process and

80 for resolution of the stress event. This translational activation is often attributed to an internal ribosome entry site-mediated mechanism of protein synthesis. Yet the known functionality of CBC during cell stress offers an alternative molecular basis for this effect

(Maquat et al., 2010; A. Sharma et al., 2012). Furthermore, RHA is known to facilitate the expression of HIV-1 in a cap-dependent manner (Bolinger et al., 2010). A defining feature of HIV-1 infection is activation of 4E-BP1(A. Sharma et al., 2012). Our identification herein of a select RHA-CBC association indicates a role for the RHA trRNP in maintained cap-dependent translation during cell stress.

To assess the functional significance of a select association between RHA and

CBC on informing a role for the RHA trRNP in cap-dependent translation during cell stress, we first investigated cellular RHA trRNP composition and activity in response to a twenty-four hour culture in serum-deprived media. Serum deprivation is an example of nutrient stress that triggers hypophosphorylation of 4E-BP1 and eIF4E trRNP-dependent translation suppression (Oh, Kim, Cho, Choe, & Kim, 2007a; A. Sharma et al., 2012). To this end, cells were cultured in media containing either 10% or 0.5% serum for twenty- four hours and then processed. As expected, a twenty-four culture in 0.5% serum resulted in hypophosphorylation of 4E-BP1 without impact on total 4E-BP1 or eIF4E protein levels (Figure 2.5a) (Oh, Kim, Cho, Choe, & Kim, 2007a) (A. Sharma et al., 2012).

Notably, serum deprivation did not induce hyperphosphorylation or a change in the total abundance of the general translation initiation factor eIF2 (Figure 2.5a). Phosphorylation of eIF2 is known to occur under targeted instances of cellular stress and results in suppression of translation by impairing 43S pre-initiation complex formation (B. Liu &

Qian, 2014). Serum-deprivation, however, is an example of a cellular stress that does not induce phosphorylation of eIF2 and instead is characterized by the activation of 4E-BP1

(Oh, Kim, Cho, Choe, & Kim, 2007a) (Maquat et al., 2010). In addition, the steady-state levels and migration patterns of CBP80 and RHA did not change following a twenty-four hour culture in serum-deprived media to indicate that serum deprivation did not impact their expression or posttranslational modification (Figure 2.5a). It has been reported that

CBP80 is phosphorylated in a regulated manner by the ribosomal protein S6 kinase; however, the experimental evidence and functional significance for this posttranslational modification is controversial (Ruvinsky & Meyuhas, 2006). A posttranslational modification regulating RHA translation activity has not been reported.

To characterize the RHA trRNP in response to serum deprivation, we immunoprecipitated endogenous RHA from the cytoplasmic extract of cells cultured for twenty-four hours in media containing either 10% or 0.5% serum. Shown in Figure 2.5b, endogenous RHA co-immunoprecipitated CBP80, the small ribosomal subunit S6, and the large ribosomal subunit L5 in the presence of 10% serum, as previously demonstrated

(Figure 2.1). Notably, endogenous RHA maintained a robust co-immunoprecipitation of both CBP80 and the ribosomal subunits from the cytoplasmic cell lysate of serum- deprived cells (Figure 2.5b). This result indicates that the RHA trRNP sustained its composition as well as its interaction with CBP80 and the ribosome even in the advent of nutrient stress and activated 4E-BP1.

To investigate the functionality of the RHA trRNP during nutrient stress and activated 4E-BP1, we assessed the co-distribution of RHA and CBP80 across a polysome

82 profile following a twenty-four hour culture of cells in media containing either 10% or

0.5% serum. Twenty-four hour serum deprivation resulted in a significant decrease in total cellular polysomes as well as initiating 48S complexes, as assessed by the traced absorbance of the ribosomal RNA (Figure 2.5c, compare peak heights in the profile traces between the 10% and 0.5% serum panels). This characteristic change in the profile trace supported effective suppression of cellular translation upon serum deprivation, as anticipated. Correspondingly, the polysome distribution of eIF4E was significantly impacted upon serum deprivation with the distribution of eIF4E shifting to sub-polysome fractions upon a twenty-four culture in 0.5% serum (Figure 2.5c). This result reflects the expected impact of nutrient stress on impairing eIF4E trRNP formation and activity due to activation of 4E-BP1 (B. Liu & Qian, 2014). Remarkably, the polysome distributions of RHA and CBP80 were not significantly impacted upon serum deprivation with the majority of both proteins maintaining a polysome association upon a twenty-four culture in 0.5% serum (Figure 2.5c). The small ribosomal subunit S6 and the large ribosomal subunit L5 also demonstrated sustained polysome association to support maintained translation activity (Figure 2.5c). These results indicate that the RHA trRNP exhibits sustained functionality during nutrient stress.

To assess the conservation of RHA trRNP composition and functionality in translation during cellular stress and activated 4E-BP1, we investigated RHA-CBC interaction and polysome association following treatment with the mTOR inhibitor torin

1 or forty-eight hour HIV-1 expression. Mammalian target of rapamycin (mTOR) is the serine/threonine kinase responsible for phosphorylating and inactivating 4E-BP1

(Laplante & Sabatini, 2012). Competitive inhibition of mTOR by torin 1 results in hypophosphorylation of 4E-BP1, disrupted eIF4E trRNP activity, and translation suppression (Thoreen et al., 2009; 2012). HIV-1 expression also induces hypophosphorylation of 4E-BP1 and eIF4E-dependent translation suppression; however, the molecular basis by which HIV-1 mediates its effect on mTOR activity has yet to be defined (A. Sharma et al., 2012). To this end, cells were treated for one hour with torin 1 or cultured for forty-eight hours in the presence of HIV-1 expression and then processed.

As expected, both torin 1 treatment and HIV-1 expression resulted in the hypophosphorylation of 4E-BP1 without impact on total 4E-BP1 or eIF4E protein levels

(Figure 2.6a) (A. Sharma et al., 2012; Thoreen et al., 2009; 2012). This effect was also observed when the treatment conditions were combined (Figure 2.6a). Also as expected, torin 1 treatment and HIV-1 expression did not induce hyperphosphorylation or a change in the total abundance of eIF2 (A. Sharma et al., 2012; Thoreen et al., 2012) (Figure

2.6a). Thus, both torin 1 and HIV-1 mediate their effects on cellular translation in a 4E-

BP1-dependent manner. Likewise, the steady-state levels and migration patterns of

CBP80 and RHA did not change in the presence of torin 1 or HIV-1 expression to indicate that their expression or posttranslational modification was not altered in either treatment condition (Figure 2.6a).

To assess the cellular interaction between RHA and CBC in response to direct mTOR inhibition or HIV-1 expression, we immunoprecipitated endogenous RHA from the cytoplasmic extract of cells that were cultured for one hour with torin 1 or forty-eight hours in the presence of HIV-1 expression. Shown in Figure 2.6b, endogenous RHA co-

84 immunoprecipitated CBP80, the small ribosomal subunit S6, and the large ribosomal L5 under mock conditions, as previously demonstrated (Figure 2.1). Notably, endogenous

RHA maintained a robust co-immunoprecipitation of both CBP80 and the ribosomal subunits from the cytoplasmic cell lysate of torin 1-treated cells (Figure 2.6b). A similar observation was made in the presence of HIV-1 expression and when the treatment conditions were combined (Figure 2.6b). These results indicate that the RHA trRNP sustained its composition as well as its interaction with CBP80 and the ribosome upon drug or viral-induced activation of 4E-BP1.

To investigate the functionality of the RHA trRNP in response to direct mTOR inhibition or HIV-1 expression, we assessed the co-distribution of RHA and CBP80 across a polysome profile following a one-hour culture of cells with torin 1 or forty-eight hour HIV-1 expression. Both stress conditions resulted in a significant decrease in total cellular polysomes, as assessed by the traced absorbance of the ribosomal RNA (Figure

2.6c, compare peak heights in the profile traces of the treatment groups to the mock control). This characteristic change in the profile supported effective suppression of cellular translation upon drug or viral-induced activation of 4E-BP1, as expected (A.

Sharma et al., 2012; Thoreen et al., 2012). Correspondingly, the polysome distribution of eIF4E was significantly impacted in both stress conditions with the distribution of eIF4E shifting to sub-polysome fractions upon a one hour culture of cells with torin 1 or forty- eight hour HIV-1 expression (Figure 2.6c). This effect was also observed when the treatment conditions were combined. These results reflect the expected impact of direct mTOR inhibition and HIV-1 expression on impairing eIF4E trRNP formation and

85 activity due to activation of 4E-BP1 (A. Sharma et al., 2012; Thoreen et al., 2009; 2012).

Remarkably, the polysome distribution of RHA and CBP80 was not significantly impacted upon torin 1 treatment or HIV-1 expression with the majority of both proteins maintaining a polysome association (Figure 2.6c). The small ribosomal subunit S6 and the large ribosomal subunit L5 also demonstrated sustained polysome association in both stress conditions to support maintained functional translation activity (Figure 2.6c).

Furthermore, the cellular steady-state levels of HIV-1 Gag expression were not impacted by torin 1 treatment (Figure 2.6d). Taken together, these cellular interaction studies show that the RHA trRNP is compositionally maintained and functionally conserved across stress conditions that induce activation of 4E-BP1.

The RHA-CBC trRNP sustains selective cap-dependent translation during cell stress.

The above results indicate the RHA-CBC trRNP as a molecular basis for selective cap-dependent translation during cell stress. Synthesis of the HIV-1 structural protein

Gag is a model example of a protein expressed during cell stress (A. Sharma et al., 2012).

Notably, its translation requires RHA and CBC (Bolinger et al., 2010; A. Sharma et al.,

2012). The data from this study indicate this effect is due to the engagement of the HIV-1 gag mRNA within a RHA-CBC trRNP. To assess this possibility, we first conducted

RNA immunoprecipitations from cytoplasmic cell lysate and analyzed for the direct association of HIV-1 gag mRNA with RHA and CBC. To this end, FLAG-RHA or endogenous CBP80 were immunoprecipitated from cytoplasmic cell lysate following a

86 forty-eight expression of HIV-1. To determine the relative co-precipitation of HIV-1 gag mRNA, the isolated FLAG-RHA and CBP80 complexes were then subjected to RNA extraction and RT-PCR analyses. Protein A and IgG isotype control immunoprecipitations were conducted in parallel to demonstrate the specificity of detected mRNA-protein interactions. A ten-fold serial dilution of the RT reaction from total cytoplasmic RNA (before IP) was conducted to show PCR sensitivity and to serve as a baseline for determining the relative enrichment of HIV-1 gag mRNA within the isolated RHA and CBC RNP complexes.

Shown in Figure 2.7a, both FLAG-RHA and CBP80 were effectively immunoprecipitated from cytoplasmic cell lysate under these conditions. Notably, a strong co-precipitation of RHA with CBP80 was observed (Figure 2.7a). This result corroborates the findings of a maintained RHA-CBP80 cytoplasmic association during

HIV-1 expression (Figure 2.6b). Of significance was the observed enrichment of HIV-1 gag mRNA within the immunoprecipitated FLAG-RHA and CBP80 complexes (Figure

2.7b, compare the gag mRNA signal within the FLAG-RHA and CBP80 IP lanes to that in the input lanes before IP). This interaction was select, as HIV-1 gag mRNA was not detected in the protein A and IgG isotype control immunoprecipitations (Figure 2.7b).

Notably it was also robust, as the HIV-1 gag mRNA signal within the FLAG-RHA and

CBP80 IP was equivalent to that detected in the input with limited dilution (Figure 2.7b).

These findings of a robust and select interaction between HIV-1 gag mRNA, RHA, and

CBP80 indicate the engagement of the HIV-1 gag mRNA within a RHA-CBC trRNP that facilitates its cap-dependent translation activity.

To assess the association of HIV-1 gag mRNA with RHA-CBC trRNPs that are active in translation, we extended the above RNA immunoprecipitation analyses to the polysome fractions of a HIV-1 sucrose gradient (Figure 2.7c and d). Shown in Figure

2.7c, both RHA and CBP80 were effectively immunoprecipitated from the polysome fractions of a HIV-1 sucrose gradient. This result corroborates the findings of a maintained RHA and CBP80 polysome association during HIV-1 expression (Figure

2.6c). Notably, HIV-1 gag mRNA was robustly detected within these identified RHA and

CBP80 polysome complexes (Figure 2.7d). This interaction was select, as HIV-1 gag mRNA was not detected in the IgG isotype control immunoprecipitation (Figure 2.7d).

Notably it was also robust, as the HIV-1 gag mRNA signal within the RHA and CBP80

IP was equivalent to that detected in the input with limited dilution (Figure 2.7d). These findings of a robust and select interaction between HIV-1 gag mRNA, RHA, and CBP80 on polysomes show the engagement of the HIV-1 gag mRNA within RHA-CBC trRNPs that are active in translation. Of significance is that these identified associations together with the known effect of HIV-1 expression on 4E-BP1 activation indicate a role for RHA and CBC in targeted cap-dependent translation during cell stress.

To directly assess RHA-CBC-mediated cap-dependent translation during cell stress, we employed our previous PCE-luciferase reporter constructs to measure cap- dependent translation in response to 4E-BP1 activation (Roberts & Boris-Lawrie, 2000).

The posttranscriptional control element (PCE) is the cis-acting 5' RNA element that selectively engages RHA translation activity and is characteristic of retroviral transcripts

(Bolinger et al., 2007; 2010; Butsch et al., 1999; Hartman et al., 2006). Notably, the PCE

88 does not function as an IRES; rather, it stimulates target protein synthesis in a cap- dependent manner (Bolinger et al., 2007; Butsch et al., 1999; Roberts & Boris-Lawrie,

2000). Thus, PCE activity is a measure of productive RHA-dependent protein synthesis that is distinctly cap-dependent.

Here we assessed PCE activity in response to serum deprivation (Figure 2.8).

Serum deprivation is a stress event that results in the activation of 4E-BP1 (Oh, Kim,

Cho, Choe, & Kim, 2007a; A. Sharma et al., 2012). Notably, RHA and CBC exhibited maintained polysome association in response to serum deprivation (Figure 2.5c). Given the above identification of a select association between the PCE-containing HIV-1 gag transcript, RHA, and CBC during activated 4E-BP1, we expected PCE activity to be maintained in response to serum deprivation. In comparison, eIF4E-dependent translation activity would be significantly compromised. To assess this possibility, cells were transfected with a firefly luciferase reporter construct harboring either the luciferase open reading frame under the 5' control of the SV40 promoter (SV40-Luc) or the luciferase open reading frame under the 5' control of the spleen necrosis virus PCE (PCE-Luc)

(Roberts & Boris-Lawrie, 2000). Twenty-four hours post-transfection, cells were placed in media containing either 10% or 0.5% serum, incubated for an additional twenty-four hours, and then processed.

Shown in Figure 2.8, a twenty-four hour culture in 0.5% serum resulted in a significant, two-fold decrease in luciferase activity from the SV40-Luc construct. This result indicates suppression of canonical cap-dependent translation, as the SV40-Luc construct does not harbor a 5' element (PCE or IRES) to sustain translation in the advent

89 of 4E-BP1 activation. In comparison, luciferase activity from the IRES-Luc construct was not significantly impacted following a twenty-four culture in 0.5% serum (Figure

2.8). This result was expected given the known functional activity of IRES elements during cell stress. Remarkably, the PCE-Luc construct also exhibited sustained luciferase activity following a twenty-four culture in 0.5% serum (Figure 2.8). Given the PCE mediates its translation effect through a cap-dependent manner that requires RHA and that RHA selectively interacts with CBC, these results show that the RHA-CBC trRNP exhibits sustained cap-dependent protein synthesis during cell stress.

DISCUSSION

Translational reprogramming characterizes the cellular stress response.

Fundamental to this effect are translation RNP (trRNP) complexes, which function to both regulate and activate targeted gene expression. The RHA trRNP is a distinct trRNP that facilitates cap-dependent expression of pathogenic HIV-1 during states of cell stress

(Bolinger et al., 2010; Hartman et al., 2006; Hernandez et al., 2008; A. Sharma et al.,

2012). RNA helicase A (RHA) selectively recognizes and binds the cis-acting posttranscriptional control element (PCE) of HIV-1 and junD mRNAs to generate the

RHA trRNP that is productive in protein synthesis (Bolinger et al., 2010; Hartman et al.,

2006; Ranji et al., 2011). Notably, RHA-PCE coordinates translation in a cap-dependent manner independent of IRES activity (Bolinger et al., 2007; Butsch et al., 1999). An unresolved issue, however, was a complete characterization of the RHA trRNP and how this informs a role for RHA in cap-dependent translational control during the cellular

90 stress response. Here we characterize the RHA trRNP to distinctly include the heterodimeric CBP80/20 cap-binding protein complex (CBC). The select association of

RHA with CBC distinguishes the RHA trRNP from canonical cap-dependent trRNP complexes, which are characterized by the eIF4E cap-binding protein. Notably, we show that the select association of RHA with CBC informs a role for the RHA trRNP in facilitating cap-dependent translation during cell stress. Taken together, our results indicate significance for the RHA-CBC trRNP in maintaining cap-dependent gene expression during the cellular stress response. Figure 2.9 presents a model summarizing these findings.

Translational reprogramming is fundamental to cell survival and recovery in response to stress. Activation of the eIF4E repressor 4E-BP1 is a signature molecular event that results in disruption of the eIF4E trRNP complex and consequent suppression of the bulk of steady-state translation. Yet targeted expression of critical stress response proteins is also observed and required for resolution of the stress event. This activity requires distinct trRNP function that is insensitive to the activation of 4E-BP1. CBC and

RHA offer a molecular basis for this effect. In comparison to eIF4E, a functional CBC repressor protein has yet to be characterized. Likewise, a stress-effector of RHA translation activity remains to be identified. Furthermore, cellular and viral transcripts that are insensitive to 4E-BP1 translational suppression demonstrate a favored association with CBC and RHA (Bolinger et al., 2010; A. Sharma et al., 2012) (and results presented herein). Our identification herein of a select association between RHA and CBC affords a molecular basis for maintained cap-dependent translation during the cellular stress

91 response. This effect is due to RHA and CBC establishing a distinct trRNP that is insensitive to 4E-BP1.

Translational reprogramming is also fundamental to virus biology. Modulation of host cell translation is a defining characteristic of viral infection and critical to disease biology. Yet the dependence of viruses on the host translation machinery for their propagation requires them to engage distinct trRNPs for their targeted expression. In the case of HIV-1, viral infection induces the activation of 4E-BP1 and consequent suppression of host cell translation via disruption of the eIF4E trRNP (A. Sharma et al.,

2012). Yet cap-dependent expression of the HIV-1 gag mRNA is maintained via its select association with the non-canonical CBC (A. Sharma et al., 2012). Also intricate to this molecular basis of HIV-1 cap-dependent protein synthesis is RHA. HIV-1 structural mRNAs harbor a cis-acting PCE within their 5' leaders that is selectively recognized and bound by RHA (Bolinger et al., 2010; Ranji et al., 2011). This RNA-protein interaction facilitates productive trRNP formation that engages robust protein synthesis in a cap- dependent manner (Bolinger et al., 2010; Hartman et al., 2006; Ranji et al., 2011). How

RHA translation activity coordinates with CBC function in regulating HIV-1 protein synthesis remained to be established.

The data presented in this study provide a molecular bridge between the fundamental roles of CBC and RHA in HIV-1 biology. Our cellular interaction and functional studies establish a molecular association between RHA and CBC that defines the RHA trRNP and informs a role for this distinct trRNP in retroviral translation control.

Notably, the select association between RHA and CBC identifies a distinct trRNP for

92 cap-dependent expression of retroviruses (Bolinger et al., 2007; 2010; Hartman et al.,

2006). In comparison, hepatitis C virus associates with the host factor and related DEAD- box helicase DDX3 (Geissler et al., 2012; Shih et al., 2008). Here DDX3 forms a distinct

DDX3 trRNP that facilitates hepatitis C viral protein expression in a cap-independent manner (Geissler et al., 2012; Shih et al., 2008). We show herein that DDX3 selectively associates with eIF4E (Figure 2.1b). This characteristic distinguishes our identified RHA-

CBC interaction and provides a molecular basis for the independent functions of the

RHA and DDX3 trRNPs in translation control. It has been reported that DDX3 can regulate expression of HIV-1 in a CBC and eIF4E-independent manner (Soto-Rifo et al.,

2012; 2013); however, we have not identified a role for DDX3 in HIV-1 translation control (Hartman et al., 2006). Additionally, we repeatedly observe a strong cellular interaction between DDX3 and eIF4E in both normal and cellular stress states (Figure

2.1b and data not shown). It is plausible that a RHA trRNP and a DDX3 trRNP both regulate HIV-1 gene expression but in a spatiotemporal manner that is mutually exclusive.

Likewise, the host factor and related DExH-box helicase DHX29 also forms a distinct trRNP critical for targeted viral gene expression. Here DHX29 facilitates Sindbis virus and Aichivirus protein synthesis in a cap-independent manner (Skabkin et al., 2010;

Y. Yu et al., 2011). This effect involves its coordination with additional host RNA binding proteins to form a distinct DHX29 trRNP important for targeted translation control (Skabkin et al., 2010; Y. Yu et al., 2011). We also identify DHX29 to selectively associate with eIF4E (S.F. and K.B.L., data not shown). This characteristic further

93 distinguishes the RHA trRNP from the DHX29 trRNP and indicates significance for a

RHA-CBC association in mediating selective cap-dependent translation of retroviral transcripts.

Our study therefore presents a new paradigm for translation control in which cap- dependent gene expression is maintained during cell stress by the RHA-CBC trRNP. This indicates significance for cap-dependent translation beyond steady-state protein synthesis to encompass targeted gene expression during the cellular stress response. Thus, translational reprogramming during cell stress extends beyond IRES-mediated activation to include RHA-CBC-dependent protein expression. Future studies aimed at assessing the relationship between the RHA-CBC trRNP and specialized ribosomes in mediating this effect are interesting to consider.

A final consideration the data presented herein elicits is the constitutive nature of the RHA-CBC association and its implications for destined cap-dependent translation during cell stress. A select cytoplasmic and polysome association between RHA and

CBC was consistently observed even in the absence of cell stress (Figure 2.1-2.4, 2.5b,

2.6b). These results indicate that select mRNAs engage the RHA-CBC trRNP to facilitate their translation even under steady-state conditions. Furthermore, the select cytoplasmic and polysome association between RHA and CBC was repeatedly observed in the absence of its HIV-1 target transcript gag (Figure 2.1-2.4, 2.5b, 2.6b). This finding indicates that the RHA-CBC trRNP is significant for the translational regulation of distinct cellular transcripts. Indeed, RHA is known to selectively bind and facilitate the translation of the cellular proto-oncogene junD (Hartman et al., 2006). Recent

94 applications of RIP-seq and polysome-CHIP to the study of RHA-dependent translation control uncovered additional cellular transcripts as targets for RHA-mediated protein synthesis (S.F., A.Y., M.H., and K.B.L., unpublished data). Several of these identified mRNAs are known to encode critical stress response proteins, such as human antigen R

(huR) (S.F., A.Y., and K.B.L., unpublished data). Notably, the translation efficiency of huR was not significantly impacted by torin 1 treatment in a recent ribosome profiling study (Thoreen et al., 2012). Furthermore, the mechanistic engagement of huR in translation is cap-dependent, as the huR 5' terminus does not exhibit functional IRES activity (D.S., S.F., and K.B.L., unpublished data). Collectively, these data indicate that select stress response proteins engage the RHA-CBC trRNP to facilitate their steady-state protein synthesis. This characteristic endows them with the ability to maintain their cap- dependent expression during cell stress. Thus, in addition to IRES-mediated translation control, the RHA-CBC trRNP affords a molecular basis for selective protein synthesis during the cellular stress response.

The mechanistic features governing retention of CBC on RHA-bound transcripts remain ambiguous. mRNAs engaged within the canonical trRNP pathway are subjected to cap exchange following the pioneer rounds of translation (Figure 2.9). This effect occurs in a translation-independent manner that is driven by the association of the nuclear import factors importin α1/β1 with CBC (Sato & Maquat, 2009). CBC exhibits a strong and stable association with importin α1 (Görlich et al., 1996). This effect is conferred by the classic nuclear localization sequence in CBP80 (Görlich et al., 1996). The unique affinity of CBC for importin α1 primes CBC for an interaction with importin β1 in the

95 cytoplasm (Sato & Maquat, 2009). A cytoplasmic CBC-importin α1-importin β1 interaction destabilizes CBC from the 5' cap, allowing eIF4E to bind (Sato & Maquat,

2009).

Impaired cap-exchange is observed in the case of replication-dependent histone mRNAs when a direct interaction between their identifying RNA binding protein, the stem-loop binding protein (SLBP), and the CBC-associated scaffolding factor CTIF occurs (Choe et al., 2013). Together, a SLBP-CTIF interaction stabilizes CBC on the 5' cap, blocking exchange with eIF4E (Choe et al., 2013). In the case of RHA and CBC, their association is sensitive to RNase-treatment (S.F. and K.B.L. data not shown). This indicates that an alternative feature of the RHA-CBC trRNP is critical for stabilizing

CBC on the 5' cap of RHA-bound transcripts. One hypothesis is that RHA-CBC trRNPs harbor CBCs that lack a strong affinity for importin α1 or have been outcompeted for importin β1 binding so as to effectively hinder exchange with eIF4E. Preliminary data also indicate that the nuclear export pathway is influential on CBC retention. Exchange from a CRM1- to a TAP-mediated nuclear export pathway reduces CBC stabilization on the HIV-1 gag mRNA, facilitating its exchange for eIF4E (A.S. and K.B.L., unpublished data). Future studies are aimed at further elucidating the molecular basis for the retained association of RHA with CBC and its significance for RHA-CBC trRNP biology in translation control during steady state and cell stress.

Figure 2.1 Endogenous RHA selectively interacts with CBP80 in the cytoplasm. (a) Schematic diagram of the CBC and eIF4E trRNPs that define eukaryotic cap-dependent regulation and the hypothesized integration of the RHA trRNP. (b) Endogenous RHA or DDX3 was immunoprecipitated from HEK293 cytoplasmic cell lysate with anti-RHA (Bethyl), anti-DDX3 (Bethyl), or isotype control IgG antibody and probed for the coprecipitation of either the CBP80 or eIF4E cap-binding protein. Shown is representative Western blot images. Input, 2% of cell lysate. (c) Endogenous RHA was immunoprecipitated from CEMx174 cytoplasmic cell lysate with anti-RHA (Vaxron) or isotype control IgG antibody and probed for the coprecipitation of the indicated proteins by Western blot. Shown is representative Western blot images. Input, 2% of cell lysate. IP, immunoprecipitated complex. FT, 2% of flow-through following IP. (d) Co- immunoprecipitations as in (c), with endogenous CBP80 (Bethyl) and eIF4E (Santa Cruz).

Figure 2.2 Epitope-tagged RHA and CBP80 demonstrate a select cytoplasmic association. (a) FLAG-RHA was immunoprecipitated from HEK293 cytoplasmic cell lysate with anti-FLAG-conjugated agarose beads (Sigma) and probed for the coprecipitation of the indicated proteins by Western blot. Shown is representative Western blot images. Input, 2% of cell lysate. IP, immunoprecipitated complex. FT, 2% of flow-through following IP. (b) Co-immunoprecipitations as in (a), with HA-CBP80 and 2xFLAG-eIF4E from input (left panels) or flow-through after FLAG IP (right panels, of FLAG IP FT).

Figure 2.3 RHA interacts with the MIF4G domain-containing protein CTIF. (a) FLAG-RHA was immunoprecipitated from HEK293 cell lysate in the absence or presence of formaldehyde crosslinking with anti-FLAG-conjugated agarose beads (Sigma) and probed for the coprecipitation of the indicated proteins by Western blot. Show is representative Western blot images. Input, 2% of cell lysate. IP, immunoprecipitated complex. FT, 2% of flow-through following IP. (b) Co- immunoprecipitations as in (a), with FLAG-CTIF.

Figure 2.4 RHA-CBP80/20 is an active trRNP. (a) Polysome profile analysis of the CBP80/20-RHA trRNP. A254 trace of total RNA distribution is shown. Top panels: representative Western blots of indicated protein distribution across the profile following TCA precipitation. Input, 2% of cell lysate. Bottom Panels: FLAG-RHA was immunoprecipitated across the profile with anti-FLAG-conjugated agarose beads (Sigma) and probed for the coprecipitation of the indicated proteins by Western blot. Shown is representative Western blot images. (b) Polysome profile analysis as in (a), following 2 min treatment with the general translation initiation inhibitor harringtonine. (c) Polysome profile analysis as in (a), following 1 hr treatment with the general translation elongation inhibitor puromycin. 100

Figure 2.5 RHA and CBP80 maintain a cytoplasmic interaction and polysome association during serum deprivation. (a) Steady-state levels and phosphorylation status of designated translation initiation factors in HEK293 cytoplasmic cell lysate following a 24 hr culture in media containing 10% or 0.5% serum. Lanes 1x, 50 μg total protein loaded. Lanes 0.5x, 25 μg total protein loaded. (b) Endogenous RHA was immunoprecipitated from indicated HEK293 cytoplasmic cell lysate with anti-RHA (Bethyl) or isotype control IgG antibody and probed for the co-precipitation of indicated proteins by Western blot. Show is representative Western blot images. (c) Polysome profile analysis of RHA and CBP80 in response to 24 hr serum deprivation. A254 traces of total RNA distribution for each gradient are shown. Representative Western blots of endogenous protein distribution across the profile are presented.

101

Figure 2.5 RHA and CBP80 maintain a cytoplasmic interaction and polysome association during serum deprivation. 102

Figure 2.6 RHA and CBP80 maintain a cytoplasmic interaction and polysome association during torin 1-mediated mTOR inhibition and HIV-1 expression. (a) Steady-state levels and phosphorylation status of designated translation initiation factors in HEK293 cytoplasmic cell lysate following 1 hr treatment with torin 1 or 48 hr expression of HIV-1. Lanes 1x, 50 μg total protein loaded. Lanes 0.5x, 25 μg total protein loaded. (b) Endogenous RHA was immunoprecipitated from indicated HEK293 cytoplasmic cell lysate with anti-RHA (Bethyl) or isotype control IgG antibody and probed for the co-precipitation of indicated proteins by Western blot. Shown is representative Western blot images. (c) Polysome profile analysis of RHA and CBP80 in response to torin1 treatment or HIV-1 expression. A254 traces of total RNA distribution for each gradient are shown. Representative Western blots of endogenous protein distribution across the profile are presented. (d) Steady-state cellular HIV-1 Gag levels as detected with the anti-p24 antibody. Continued 103

Figure 2.6 continued d

104

Figure 2.7 Cytoplasmic and polysome-associated RHA and CBP80 bind the HIV-1 gag mRNA during cell stress. (a) FLAG-RHA or endogenous CBP80 were immunoprecipitated from HEK293 cytoplasmic cell lysate with anti-FLAG-conjugated agarose beads (Sigma) or anti-CBP80 antibody (Bethyl), respectively. Equivalent aliquots of each immunoprecipitated complex were assessed by Western blot for IP efficiency. Titrating the total protein loaded from input lysate indicated Western blot sensitivity. Specificity of the IP was verified using unconjugated Protein A beads of isotype control IgG antibody, respectively. (b) RT-PCR of co-immunoprecipitated HIV-1 gag mRNA from (a). ND, no dilution. - RT, without reverse transcriptase. + RT, with reverse transcriptase. (c) Endogenous RHA or CBP80 were immunoprecipitated from indicated polysome samples of Figure 2.6 with anti-RHA (Bethyl), anti-CBP80 (Bethyl), or isotype control IgG antibody. Equivalent aliquots of each immunoprecipitated complex were assessed by Western blot for IP efficiency. (d) RT-PCR of co- immunoprecipitated HIV-1 gag mRNA from (c).

105

Figure 2.8 Retroviral PCE translation is maintained in a cap-dependent manner during cell stress. HEK293 cells were transfected with indicated luciferase reporter constructs in triplicate and then cultured for 24 hr in media containing either 10% or 0.5% serum. Cytoplasmic lysates were harvested and assayed for luciferase activity. *** indicates p < 0.05.

106

Figure 2.9 Model of the canonical and RHA-CBP80/20 trRNP pathways and their functional significance for cap-dependent translation during cell stress. The canonical trRNP pathway models the majority of eukaryotic translational control. Here an mRNA engages first within the CBC trRNP to facilitate its participation in the initial rounds of protein synthesis. Subsequently, an mRNA remodels into the eIF4E trRNP in which engages in steady-state protein synthesis. Cell stress effectively targets the eIF4E trRNP by activating the eIF4E repressor 4E-BP1. In a hypophosphorylated state, 4E-BP1 binds with high affinity to eIF4E. This effectively compromises the interaction between eIF4E and eIF4G, resulting in disrupted eIF4E trRNP dynamics and impaired translation activity. In the case of the RHA-CBP80/20 trRNP pathway, RHA selectively recognizes and binds the 5' posttranscriptional control element (PCE) that defines target transcripts. This RNA-protein interaction engages RHA trRNP formation to facilitate productive cap- dependent translation. For the RHA trRNP, this activation stably includes the CBP80/20 (CBC) cap-binding protein complex. The select association of the RHA trRNP with CBC distinguishes the RHA trRNP from the canonical trRNP pathway regulating the bulk of cellular translation. It also endows the RHA trRNP with the ability to maintain cap- dependent protein synthesis upon stress-induced activation of 4E-BP1. Thus, the RHA- CBP80/20 trRNP provides a targeted mechanism for maintained cap-dependent protein synthesis during cell stress.

107

Figure 2.9 Model of the canonical and RHA-CBP80/20 trRNP pathways and their functional significance for cap-dependent translation during cell stress.

108

Chapter 3 : Conserved Domains of RNA Helicase A Contribute Genetically

Separable Roles to Its Function in Cap-Dependent Translation Control

ABSTRACT

Cellular DExH/D-box RNA helicases are critical regulators of eukaryotic translation. They perform essential RNA unwinding and/or RNP remodeling activities that enable the characteristic cap-dependent scanning mechanism of eukaryotic protein synthesis. RNA helicase A (RHA) is the cellular DEIH-box RNA helicase required for cap-dependent translation of pathogenic retroviral transcripts and the cellular proto- oncogene junD. Its two N-terminal double-stranded RNA binding domains (dsRBDs) direct interaction with its cognate cis-acting RNA element, the posttranscriptional control element (PCE), which identifies target mRNAs. The association of RHA with the PCE engages a translation RNP (trRNP) complex that facilitates productive cap-dependent translation. Yet to be defined, however, is the molecular significance of RHA for particular steps of the eukaryotic translation process. Herein we employed extended genetic and molecular techniques to show functional importance for RHA in initiation and post-initiation translation activity. We identified significance for conserved lysine residues of the N-terminal dsRBDs in coordinating RHA trRNP formation. This activity was fundamental to the engagement of RHA within the translation process. Mutation of

109 the conserved dsRBD lysine residues resulted in an impaired association of RHA with

43S pre-initiation ribosome complexes, hindering its polysome engagement and targeted product generation. We also identified molecular significance for RHA in 80S ribosome stabilization. This function was attributed to its C-terminal arginine-glycine (RGG)-rich domain. Mutation of the arginine residues of the RGG-domain did not impact RHA trRNP formation or its 43S pre-initiation ribosome complex engagement. It did, however, significantly compromise its 80S association and polysome identification. This effect was not attributable to altered interactions with elongation factors; rather, ribosomal run-off assays supported significance for the C-terminal domain in ribosome stabilization. Taken together, our findings indicate novel functional activity for RHA as a DExH/D-box RNA helicase with significance in initiation and post-initiation translation activities. We propose this feature is critical to the selective recruitment of RHA by complex retroviral and cellular transcripts for their regulated protein synthesis.

INTRODUCTION

RNA helicase A (RHA/DHX9) is a DEIH-box RNA helicase with a breadth of cellular activity: transcriptional regulation, posttranscriptional gene control, and innate immunity. These diverse biological functions are conferred upon RHA by its multi- domain architecture and nucleo-cytoplasmic shuttling activity. A 142 kilo-dalton protein,

RHA harbors two N-terminal double-stranded RNA binding domains (dsRBDs), a central catalytic and DEIH helicase core, a helicase-associated 2 domain, and a C-terminal domain of unknown function and arginine-glycine-rich (RGG) domain (Ranji & Boris-

110

Lawrie, 2010) (Figure 3.1). The C-terminus of RHA also contains an active bidirectional nuclear transport domain that allows for its rapid nucleo-cytoplasmic shuttling (H. Tang,

McDonald, Middlesworth, Hope, & Wong-Staal, 1999). Predominantly nuclear at steady state, cytoplasmic RHA is identified with polyribosomes and functions as a critical translation regulator of retroviral transcripts and the cellular proto-oncogene junD

(Bolinger et al., 2007; 2010; Hartman et al., 2006; Ranji et al., 2011). The specificity of this activity is controlled by its N-terminal dsRBDs, which bind target mRNAs via a select association of surface exposed lysine residues with its cognate RNA element (Ranji et al., 2011). Overexpressed N-terminus inhibits RHA-dependent translation activity, emphasizing the significance of the dsRBDs in regulating formation of a productive translation ribonucleoprotein (trRNP) complex that allows for protein synthesis (Ranji et al., 2011).

The central catalytic core of RHA is also critical for its translation activity, as mutation of an essential ATP-binding residue significantly reduces target HIV-1 Gag production (Bolinger et al., 2010). This result implies importance for RHA in engaging

ATP-dependent nucleic acid unwinding and/or mRNP remodeling activity to facilitate protein synthesis. Roles for the C-terminal domains of RHA in translation control have yet to be investigated; however, the importance of the conserved RGG domain in the functioning of other translation regulators indicates significance (Blackwell & Ceman,

2011; Chen et al., 2014; Dresios et al., 2005; Rajyaguru, She, & Parker, 2012;

Thandapani, O'Connor, Bailey, & Richard, 2013).

111

The mechanistic process of eukaryotic protein synthesis is divided into three phases: initiation, elongation, and termination. Each step is coordinated by designated cofactors that intricately facilitate ribosome dynamics and polypeptide synthesis.

Initiation is the rate-limiting step, requiring at least nine core factors (Jackson et al.,

2010). Its efficiency is highly regulated by the cellular environment (Jackson et al., 2010;

B. Liu & Qian, 2014). It is also the stage historically most influenced by specialized RNA binding proteins, particularly DExD/H-box RNA helicases, recruited for targeted translation control (Parsyan et al., 2011). However, elongation is now being realized as equally important in the regulation of protein synthesis with numerous examples of specialized RNA binding proteins and ribosome associated factors influencing the process of polypeptide extension (Chen et al., 2014; Darnell et al., 2011; Friend et al.,

2012; Hussey et al., 2011; Nottrott, Simard, & Richter, 2006; Petersen, Bordeleau,

Pelletier, & Sharp, 2006; Ricci et al., 2014; Shalgi et al., 2013).

The initial identification of RHA as the specialized RNA binding protein necessary for cap-dependent translation of retroviral mRNAs and the cellular proto- oncogene junD, and its observed association with a distinct 5' RNA element, termed the posttranscriptional control element (PCE), that regulates these transcripts' cap-dependent translation efficiency, favored a model for RHA in translation initiation (Hartman et al.,

2006). This was further supported by the stable stem-loop structure of the PCE and the importance of RHA's ATP-binding domain in its translation activity (Bolinger et al.,

2010; Roberts & Boris-Lawrie, 2003). Together this data afforded a model for RHA in the resolution of the PCE to promote efficient ribosome scanning during the early stages

112 of initiation (Bolinger et al., 2010; Hartman et al., 2006). However, the additional sedimentation of RHA within heavy polysome fractions of a sucrose gradient and the sensitivity of these complexes to puromycin suggested that RHA also has a critical role in late-stage initiation or even post-initiation translation events (Hartman et al., 2006).

The objective of this study was to elucidate the molecular significance of RHA in the eukaryotic translation process. Herein we extended established genetic and molecular techniques to analyze the role of RHA in influencing ribosome dynamics during the three stages of protein synthesis. We show that critical residues within the conserved N- terminal dsRBDs, DEIH helicase core, and C-terminal RGG domain confer genetically separable roles for RHA in 43S pre-initiation ribosome complex recruitment, scanning, and 80S stabilization. Together these findings provide a molecular basis for RHA in the translation process and highlight the significance of a specialized RNA binding protein functioning at multiple stages to promote protein synthesis.

MATERIALS AND METHODS

Cell culture and transfections

(see Plasmid Construction for more details) using the X-tremeGene HP DNA

113

Transfection Reagent (Roche) at a ratio of 1:3. Media was changed 24 hr post- transfection and cells harvested at forty-eight.

SDS-PAGE and Western Blotting

Protein samples were run on 4-15% gradient SDS-polyacrylamide gels (BIO-

RAD) at 200 V for 45 minutes and transferred to nitrocellulose membranes (BIO-RAD) via a wet transfer at 80 V for 1 hour by standard methods. Membranes were incubated in

5% non-fat dry milk in PBS-T with the addition of designated primary antibody overnight at 4 degrees. Blots were washed 3 × 5 min with PBS-T after antibody incubation. HRP-conjugated secondary antibodies (GE Healthcare) were used at 1:5000 or 1:10,000 in 5% milk/PBS-T for 1 hour at RT, washed as before, and HRP signal detected by Amersham ECL Prime Western Blotting Detection Reagent (GE Healthcare).

The Fuji Imaging System (FUJIFILM) and MultiGauge Software Program (FUJIFILM) were used to develop and assess protein band intensity.

Antibodies

The following primary antibodies were used for Western blot analysis: anti-FLAG

(Sigma, F7425) at 1:5000, anti-RHA (Vaxron, PA-001) at 1:7500, anti-NCBP1/CBP80

(Bethyl, A301-793A) at 1:1000, anti-eIF4E (Cell Signaling, #9742S) at 1:1000, anti-S6 ribosomal protein (Cell Signaling, #2217) at 1:1000, anti-L5 ribosomal protein (Bethyl,

A303-933A) at 1:1000, anti-eEF1A (Cell Signaling, #3586) at 1:1000, and anti-eEF2

(Cell Signaling, #2332) at 1:1000.

114

For FLAG IPs, anti-FLAG M2 affinity gel (Sigma, A2220) was used in the IP step of the FLAG-conjugated constructs.

Plasmid Construction

The N-term RHA, FLAG-RHA, and FLAG-RHA K417R constructs used in this study were previously described (Bolinger et al., 2010; Hartman et al., 2006; Ranji et al.,

2011). FLAG-RHA K54A/55A was created by site-directed mutagenesis on FLAG-RHA with primers KB1676 and KB1677; FLAG-RHA K236E with primers KB1440 and

KB1441; and FLAG-RHA RA9 with primers KB1393 and KB1394, KB1395 and

KB1396, KB1418 and KB1413, KB1414 and KB1415, and KB1416 with KB1417.

Cloning of all three constructs was performed previously by C.B. FLAG-RHA

K54A/55A/236E was created by digesting FLAG-RHA K236E with HindIII and HpaI and ligating insert into FLAG-RHA K54A/55A cut with HindIII and HpaI (cloning conducted by A.R.).

KB1676: 5' ccaccacccgggtaccatgatcgatccacttactgatactcctgacac 3'

KB1677: 5' gagtcaggagtatcagtaagtggatcgatcatggtacccgggtggtgg 3'

KB1440: 5' ggatcaaataaggaattggcagcacagtcc 3'

KB1441: 5' ggactgtgctgccaattccttatttgatcc 3'

KB1393: 5' gaggctatgcaggagtttccgcaggtggctttgcaggcaac 3'

KB1394: 5' gttgcctgcaaagccacctgcggaaactcctgcatagcctc 3'

KB1395: 5' gagactacgcagggcctagtggaggct acgcaggatctgg 3'

115

KB1396: 5' ccagatcctgcgtagcctccactaggccctgcgtagtctc 3'

KB1418: 5' ggatatagagcgggaggttctagt 3'

KB1413: 5' actagaacctcccgctctatatcc 3'

KB1414: 5' gattccaggcaggaggtggtgcgggggcctatggaact 3'

KB1415: 5' agttccataggcccccgcaccacctcctgcctggaatc 3'

KB1416: 5' ttggacagggagcaggaggtggcggctat 3'

KB1417: 5' atagccgccacctcctgctccctgtccaa 3'

Polysome Profiling

Transfected HEK293 cells were treated with 0.1 mg/mL cyclohexamide (CHX; prepared in DMSO, Sigma C1988) for 20 minutes, then washed twice with ice cold PBS with CHX, and scraped off the plates in 5 mL of ice cold PBS with CHX. Resuspended cells from both plates were combined in 15 mL conical tubes and then pelleted at 1500 rpm for 4 min at 4C. Supernatant was removed and cell pellet resuspended in 0.75 mL of low salt buffer (20 mM Tris-HCl, pH 7.5, 3 mM MgCl2, 10 mM NaCl, 2 mM DTT, 1X protease inhibitor cocktail EDTA-free (Roche 11 873 580 001), 5 μL/mL RNase Out).

Cells were allowed to swell on ice for 5 min and then lysed by the addition of 0.25 mL lysis buffer (0.2 M sucrose and 1.2% Triton X-100 prepared in low salt buffer) followed by 10 strokes with a dounce homogenizer (Kimble Chase). Lysate was purified by spinning at top speed (16,000 x g) for 1 minute at 4C, then layered onto a 10-50% sucrose gradient (10 mM Tris-HCL, pH 7.5, 5 mM MgCl2, 100 mM KCl, 2 mM DTT, 0.1 mg/mL CHX, 10/50% sucrose), and spun in an ultracentrifuge for 2 hr 40 min at 35,000

116

RPM at 4C. Loading equivalent amounts of RNA OD units across samples for each experiment normalized the profiles. Polysome profiles were generated by continuous monitoring of RNA absorbance at 254 nm by the ISCO UA-6 Absorbance Detector unit and fractionated into 22 equivalent volume (0.5 mL) fractions using the ISCO Foxy R1 fraction collector. Brandel Peak Trace Software was used to generate the corresponding profile traces.

Ribosomal Run-off Assays

Forty-eight hours post-transfection of the designated FLAG-RHA construct,

HEK293 cells were treated with 2 μ g/mL harringtonine (Abcam 141941) for two minutes then immediately incubated with

0.1 mg/mL CHX for one minute and lysed. Polysome profiling was conducted as described. For ribosomal run-off assays with puromycin, transfected cells were treated with 400 nM puromycin (Sigma P7255) for 1 hr and then incubated with 0.1 mg/mL

CHX for 10 min.

In vivo translation initiation Assays

Twenty-four hours post-transfection of the designated FLAG-RHA construct,

HEK293 cells were washed three times with 1x PBS and then cultured in complete media containing 0.5% FBS for 24 hr. Media was removed and replaced with complete media containing 10% FBS for either 1 hr or 2 hr upon which cells were treated with 0.1 mg/mL

CHX for 20 min and then lysed. Lysates were separated on a 10-30% sucrose gradient by

117 ultracentrifugation at 35,000 rpm for 3 hr at 4C and then fractionated into 22 equivalent volume (0.5 mL) fractions.

Protein Association with Polysome Fractions

Immunoprecipitations

For FLAG IPs, anti-FLAG-conjugated agarose beads (Sigma) were pre-blocked for 1 hr in 0.5% BSA, washed, and then incubated with harvested cell lysate for 3 hr at

4C with end-over-end rotation. Total volume of each IP was brought to 1 mL with 1x wash buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl). Immunoprecipitated complexes were washed three times with NETN-150 wash buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1 mM EDTA, 0.5% NP40) and twice with 1x wash buffer. Loading buffer

118 was added to isolated bead complexes, heated, and loaded directly onto the gel. 2% of the initial cleared lysate was reserved and diluted with 1x wash buffer and run as INPUT for each IP. 2% of the flow-through following IP was reserved and run as FT for each IP.

For polysome IPs, alternating sucrose gradient fractions across each polysome profile were diluted with 0.5 mL of 1x wash buffer and then indicated complexes were immunoprecipitated as described.

RESULTS

Conserved N-terminal lysine residues of RHA coordinate formation of its distinct trRNP.

Translation ribonucleoprotein (trRNP) complexes are the core effectors of eukaryotic translation. Their formation is fundamental to the activation of an mRNA and engagement within the translation process. The RHA trRNP is defined by RHA and its cognate cis-acting RNA element, the posttranscriptional control element (PCE) (Hartman et al., 2006). A stable association with the non-canonical CBP80/20 cap-binding protein complex (CBC) also characterizes the RHA trRNP (Chapter 2). This distinct association with CBC endows the RHA trRNP with a role in cap-dependent translation control during cell stress (Chapter 2). To understand the molecular significance of RHA in the translation process, we sought to first identify the residues of RHA necessary for trRNP formation. In its N-terminal dsRBDs, RHA harbors critical surface-exposed lysine residues that are necessary for specific interactions with its cognate PCE and productive translation activity (Ranji et al., 2011). Mutation of lysine residues 54, 55, or 236 within

119 the dsRBDs lessens target RNA binding but with a significant impact on RHA-dependent translation activity (Ranji et al., 2011). The molecular basis for this effect has yet to be characterized.

To determine the significance of the dsRBDs, and in particular lysine residues 54,

55, and 236, in coordinating RHA trRNP formation, we assessed the cellular interaction of N-term RHA with CBC by co-immunoprecipitation analysis. N-term RHA, harboring the first 300 amino acids of RHA, was expressed in cells, immunoprecipitated from cell lysate, and assessed for the co-precipitation of CBP80 by immunoblot analysis. The results show that N-term RHA was sufficient for interaction with CBP80; it did not co- precipitate the eIF4E cap-binding protein (Figure 3.2a). This select association between

N-term RHA and CBC was expected given the distinct interaction identified between endogenous RHA and CBC (Chapter 2). When critical substitution mutations were introduced at lysine residues 54, 55, and/or 236 the select interaction between N-term

RHA and CBP80 was significantly impaired (Figure 3.2a). Extension of this analysis to

FLAG-RHA showed that FLAG-RHA K236E exhibited reduced cytoplasmic association with CBP80 relative to FLAG-RHA (Figure 3.2b). In comparison, FLAG-RHA

K54A/55A maintained an interaction with CBP80 (Figure 3.2b). This result indicates that lysine residue 236 is necessary for a stable association with CBP80. Of significance is the result that mutation of lysine residue 236 also compromised the interaction of RHA with the small and large ribosomal subunits rpS6 and rpL5, respectively (Figure 3.2b).

Alternatively, FLAG-RHA demonstrated a robust co-precipitation of rpS6 and rpL5

(Figure 3.2b). These results indicate that lysine residue 236 of the N-terminal dsRBDII is

120 critical for engaging a stable CBC association that generates a productive RHA trRNP.

Notably, FLAG-RHA K54A/55A also demonstrated impaired interactions with rpS6 and rpL5 even though it effectively co-precipitated CBP80 (Figure 3.b). This result indicates that lysine residues 54 and 55 of the N-terminal dsRBDI are also necessary for productive

RHA trRNP formation; however, their molecular significance in this process is distinct from that of lysine residue 236 and its engagement of a CBC association.

We previously demonstrated that lysine residue 417 of the DEIH helicase core is necessary for RHA-dependent translation activity (Bolinger et al., 2010). This amino acid facilitates ATP-dependent helicase activity of RHA and its mutation to arginine significantly impairs target HIV-1 Gag production (Bolinger et al., 2010). To investigate the molecular significance of lysine residue 417 in RHA trRNP formation, we assessed the ability of the FLAG-RHA K417R mutant to interact with CBP80, rpS6, and rpL5 in cells via a co-immunoprecipitation analysis (Figure 3.2c). Here the co- immunoprecipitations were conducted both in the absence and presence of serum deprivation given the significance of the RHA trRNP in promoting cap-dependent translation during cell stress (Chapter 2). In addition, we assessed the ability of a C- terminal FLAG-RHA RA9 mutant to interact with CBP80, rpS6, and rpL5. RGG domains are known to influence trRNP dynamics and preliminary data indicated significance for the C-terminal RGG domain of RHA in facilitating productive translation activity (C.B. and K.B.L., data not shown). Shown in Figure 3.2c, both FLAG-RHA

K417R and FLAG-RHA RA9 effectively co-precipitated CBP80, rpS6, and rpL5 like

FLAG-RHA. Alternatively, FLAG-RHA K54A/55A/236E exhibited a significant

121 impairment in its association with CBP80, rpS6, and rpL5 relative to FLAG-RHA (Figure

3.2c). These results were observed in both culture conditions (10% or 5% serum) to indicate conserved significance of the N-terminal dsRBDs in facilitating productive RHA trRNP formation. On the other hand, the data indicate that the DEIH helicase core and C- terminal RGG domains of RHA are significant for its role in the translation process beyond trRNP formation.

RHA N- and C-terminal domains exhibit genetically separable roles in polysome association.

The above results show that conserved lysine residues within the N-terminal dsRBDs of RHA are critical for productive RHA trRNP formation. This finding indicates significance for RHA in the early stages of initiation that promote mRNA activation to facilitate ribosome recruitment. Yet the inability of the RHA helicase and C-terminal mutants to engage in productive translation activity but maintain appropriate trRNP formation signifies additional roles for RHA in the translation process. To achieve further insight into the molecular significance of RHA and its conserved domains in the translation process, we assessed the distribution of FLAG-RHA, FLAG-RHA

K54A/55A/236E, FLAG-RHA K417R, and FLAG-RHA RA9 across a polysome profile.

Polysome profiling is a molecular technique that enables direct monitoring of the translation process by allowing the fate of a translation factor and/or target mRNA to be followed as it engages with the ribosome and proceeds through the three stages of polypeptide synthesis. Our application here of directed RHA mutagenesis allowed us to

122 specifically analyze the molecular role of RHA in the translation process. Notably, this approach provided an opportunity to assess the mechanistic significance of critical domain residues in engaging RHA with the ribosome and enabling translation activity beyond mRNA activation and ribosome recruitment. Complexes identified within the first two peaks of a polysome profile trace (40S and 60S, 'sub-polysome') represent trRNPs engaged in the initial stages of ribosome recruitment, scanning, and start codon recognition. Those isolated from the third peak reflect trRNPs engaged with elongation- competent 80S ribosomes or free 80S ribosome complexes. Translation factors and mRNAs identified within the heavy, polysome peaks of a profile trace indicate complexes robustly engaged with several initiating and/or elongating 80S ribosomes productive in polypeptide synthesis. For our assessment, we focused on the FLAG-RHA

K54A/55A/236E, FLAG-RHA K417R, and FLAG-RHA RA9 mutants given their distinct phenotypes in RHA trRNP formation (Figure 3.2) and known functional significance in RHA-dependent translation activity (Bolinger et al., 2010; Ranji et al.,

2011) (C.B. and K.B.L., data not shown).

To this end, cells were transfected with FLAG-RHA or indicated mutants, processed, and then layered upon a 10-50% sucrose gradient and separated by ultracentrifugation. The distribution of FLAG-RHA or indicated mutants was then analyzed in designated sub-polysome or polysome fractions based upon the traced absorbance of the ribosomal RNA for each profile. FLAG-RHA was efficiently detected across all analyzed gradient fractions with particular prominence in the 80S and early polysome peaks (Figure 3.3, top panel of first Western blot set). This distribution

123 reflected our previous observations for FLAG-RHA as well as endogenous RHA distribution across a polysome profile (Hartman et al., 2006) (Figure 3.3, compare top and middle panels of first Western blot set). This result supports a role for RHA in translation control that can be studied with an exogenously expressed FLAG-tagged construct.

In comparison, FLAG-RHA K54A/55A/236E distribution was limited to the early sub-polysome fractions (Figure 3.3, compare top panels of first and second Western blot sets). This impaired polysome distribution of FLAG-RHA K54A/55A/236E corresponded with an altered profile trace, having reduced polysome peak height relative to that of FLAG-RHA (Figure 3.3, compare indicated profile traces at top). This result indicates an impact of FLAG-RHA K54A/55A/236E expression on target RHA translation activity. Notably, this effect was observed even with maintained endogenous

RHA across the polysome profile to indicate a dominant effect of FLAG-RHA

K54A/55A/236E on RHA-dependent translation activity (Figure 3.3, compare top and middle panels of second Western blot set). This restriction of FLAG-RHA

K54A/55A/236E to the early sub-polysome fractions corresponded with its inability to appropriately associate with CBC and the ribosome (Figure 3.2). Collectively, these results indicate a direct correlation between RHA trRNP formation, target mRNA activation, and engagement of RHA within translation initiation.

In contrast to FLAG-RHA K54A/55A/236E, FLAG-RHA RA9 engaged beyond the sub-polysome fractions to the monosome peak of a polysome profile (Figure 3.3, compare top panels of second and third Western blot sets). However, like FLAG-RHA

124

K54A/55A/236E, FLAG-RHA RA9 demonstrated an altered heavy polysome association relative to that of FLAG-RHA (Figure 3.3, compare top panels of first and third Western blot sets). This altered polysome distribution of FLAG-RHA RA9 corresponded with an altered profile trace, having increased sub-polysome and polysome peak heights relative to that of FLAG-RHA (Figure 3.3, compare indicated profile traces at top). This result indicates an impact of FLAG-RHA RA9 expression on target RHA translation activity.

As with FLAG-RHA K54A/55A/236E, these effects were observed even with maintained endogenous RHA across the polysome profile to indicate a dominant effect of FLAG-

RHA RA9 on RHA-dependent translation activity (Figure 3.3, compare top and middle panels of third Western blot set).

Of significance was the observed ability of FLAG-RHA RA9 to engage at least one elongating ribosome (Figure 3.3, top panel of third Western blot set). This polysome profile distribution of FLAG-RHA RA9 reflects its observed ability to associate with

CBP80, rpS6, and rpL5 (Figure 3.2c), indicating a role for RHA in translation beyond ribosome recruitment once an appropriate trRNP is established. However, the inability of

FLAG-RHA RA9 to engage heavy polysome complexes like FLAG-RHA (Figure 3.3, compare top panels of first and third Western blot sets) and generate productive protein product (C.B. and K.B.L., data not shown) signifies importance for the C-terminal RGG domain in early post-initiation activities that promote 80S stabilization and effective elongation.

We also assessed the distribution of the FLAG-RHA K417R mutant across a polysome profile. FLAG-RHA K417R exhibited reduced polysome association relative to

125

FLAG-RHA (Figure 3.3, compare top panels of first and fourth Western blot sets). This effect, however, was not as significant as seen with FLAG-RHA K54A/55A/236E or

FLAG-RHA RA9 (Figure 3.3, compare top panels in second, third, and fourth Western blot sets). FLAG-RHA K417R expression also altered the profile trace with increased sub-polysome and polysome peak heights relative to that seen with expression of FLAG-

RHA (Figure 3.3, compare indicated profile traces at top). This result indicates an impact of FLAG-RHA K417R expression on target RHA translation activity. Like FLAG-RHA

RA9, FLAG-RHA K417R was able to associate with CBC and form a RHA trRNP

(Figure 3.2c). However, its restriction from productive polysome association and inability to generate detectable protein product indicates significance for RHA ATPase activity in the ribosome recruitment and scanning process.

The C-terminal RGG domain is critical for initiation complex stabilization.

The results presented indicate a role for RHA in early translation initiation events that encompass productive trRNP formation, target mRNA activation, and engagement with initiating ribosome complexes (Figure 3.2 and 3.3). This effect is coordinated by the

N-terminal dsRBDs of RHA (Figure 3.2 and 3.3). The data, however, also indicate a role for RHA in early post-initiation events (Figure 3.3). These activities require functional

ATPase association and an intact C-terminal RGG domain (Figure 3.3). It is known that scanning ribosome intermediates readily dissociate from an mRNA when trRNP stability is compromised (Gebauer, Grskovic, & Hentze, 2003; Pestova, Borukhov, & Hellen,

1998). This effect manifests in a sucrose gradient with a shift in cofactor sedimentation

126 from polysome to sub-polysome fractions and destabilization of trRNP complexes. Thus, an alternative model to consider is that mutation of RHA's helicase core or C-terminal

RGG domain compromises its trRNP stability in manner that induces scanning ribosome dissociation from a target transcript. This initiation defect also offers a molecular explanation for the reduced polysome association of FLAG-RHA K417R and FLAG-

RHA RA9.

To further investigate the molecular significance of RHA in initiation or post- initiation translation events, we assessed RHA trRNP complex stability across the sucrose gradients analyzed in Figure 3.3. To this end, FLAG-RHA or designated FLAG-RHA mutants were immunoprecipitated from indicated fractions across their respective polysome profiles and then assessed for the co-precipitation of CBP80 and rpL5 by immunoblot analysis. Shown in Figure 3.4a (top panel of first Western blot sets), FLAG-

RHA was effectively immunoprecipitated across the polysome profile in a manner that reflected its observed distribution when assessed directly from each fraction (Figure 3.3).

Furthermore, when FLAG-RHA was immunoprecipitated from the sub-polysome and

80S fractions it also co-precipitated CBP80 and rpL5. An identified association of

FLAG-RHA with CBP80 and rpL5 in these polysome fractions indicates: (1) stable RHA trRNP formation that is effectively engaged in translation initiation and (2) significance for the RHA trRNP in promoting translation events through 80S formation. Thus, FLAG-

RHA is observed across a polysome profile like endogenous RHA because it can effectively engage stable trRNP activity that promotes productive translation.

127

In contrast, immunoprecipitation of FLAG-RHA K54A/55A/236E was limited to the early sub-polysome fractions of its sucrose gradient (Figure 3.4a, top panel of second

Western blot sets). This immunoprecipitation profile of FLAG-RHA K54A/55A/236E reflected its observed distribution when assessed directly from each fraction (Figure 3.3).

Furthermore, when FLAG-RHA K54A/55A/236E was immunoprecipitated from the early sub-polysome fractions it did not co-precipitate either CBP80 or rpL5 (Figure 3.4a, second Western blot panel sets). This result indicates that FLAG-RHA K54A/55A/236E compromises stable trRNP formation, which reflects the impaired association observed between FLAG-RHA K54A/55A/236E, CBP80, and rpL5 in cytoplasmic lysate prior to separation on a sucrose gradient (Figure 3.2c). Thus, FLAG-RHA K54A/55A/236E is only observed in sub-polysome fractions of a sucrose gradient because of its impaired trRNP formation.

Of significance is the observation that FLAG-RHA RA9 did not co-precipitate either CBP80 or rpL5 even when it was effectively immunoprecipitated from sub- polysome fractions like FLAG-RHA (Figure 3.4a, compare first and third Western blot panel sets). In contrast to FLAG-RHA K54A/55A/236E, FLAG-RHA RA9 demonstrated effective associations with CBP80 and rpL5 in cytoplasmic lysate prior to separation on a sucrose gradient (Figure 3.2c). Given unstable trRNP complexes engaged in translation initiation events readily dissociate when centrifuged through a sucrose gradient (Gebauer et al., 2003; Pestova et al., 1998), this result indicates that FLAG-RHA RA9 impairs

RHA translation activity by destabilizing its trRNP during scanning initiation events.

128

Thus, FLAG-RHA RA9 is only weakly observed in 80S and very early polysome fractions of a sucrose gradient because of its impaired trRNP stabilization.

FLAG-RHA K417R, on the other hand, maintained a co-precipitation of CBP80 and rpL5 when it was effectively immunoprecipitated from sub-polysome fractions like

FLAG-RHA (Figure 3.4a, compare first and fourth Western blot panel sets). These associations reflect the demonstrated interactions between FLAG-RHA K417R, CBP80, and rpL5 in cytoplasmic lysate prior to separation on a sucrose gradient (Figure 3.2c).

Notably, this result distinguishes the mechanistic significance of RHA's ATPase activity from the role of the C-terminal RGG domain in coordinating RHA-dependent translation activity. It demonstrates that RHA's ATPase activity is not required for trRNP stabilization during initial scanning events; rather, it signifies importance for ATP- dependent trRNP remodeling activities that promote 80S ribosome activities, as FLAG-

RHA K417R exhibited comparable sub-polysome association but reduced polysome accumulation relative to FLAG-RHA (Figure 3.3 and 3.4a).

A limitation of co-immunoprecipitation analyses, especially those conducted from sucrose gradient fractions when complexes are diffused within a medium, is that the sensitivity of detection of cofactor interactions is dependent upon the efficiency of target factor immunoprecipitation. To validate the above findings, we performed similar co- immunoprecipitation analyses of FLAG-RHA and designated mutants from sucrose gradient fractions across a polysome profile. Except, here, sub-polysome and polysome fractions were pooled to enhance the efficiency of FLAG-RHA immunoprecipitation.

129

With this approach, FLAG-RHA was efficiently immunoprecipitated from the pooled sub-polysome fractions of a sucrose gradient and less efficiently immunoprecipitated from the corresponding polysome fractions (Figure 3.4b, top panel of the left Western blot set). This immunoprecipitation profile of FLAG-RHA reflected its observed distribution when assessed directly from each fraction or individually immunoprecipitated (Figure 3.3 and 3.4a). Thus, combining related gradient fractions to conduct more robust co-immunoprecipitation analyses does not alter the representation of factors across a polysome profile.

Notably, when FLAG-RHA was immunoprecipitated from the pooled sub- polysome fractions it also co-precipitated CBP80, rpS6, and rpL5 (Figure 3.4b, left

Western blot panel set). This result reflects the observed interactions between FLAG-

RHA and CBP80, rpS6, and rpL5 in cytoplasmic lysate prior to its separation on a sucrose gradient or when FLAG-RHA is individually immunoprecipitated (Figure 3.2 and 3.4a). Furthermore, FLAG-RHA was observed to co-precipitate the translation initiation factor eIF2 from pooled sub-polysome gradient fractions (Figure 3.4b, left

Western blot panel set). eIF2 is a critical component of the 43S pre-initiation ribosome complex that functions in initial ternary complex formation, scanning, and start codon recognition (Hinnebusch, 2014; Hinnebusch & Lorsch, 2012; Jackson et al., 2010). This detected interaction between FLAG-RHA and eIF2 supports the engagement of FLAG-

RHA within stable trRNP complexes that actively interact with the 43S pre-initiation ribosome complex. Notably, an interaction between FLAG-RHA and eIF2 is only efficiently observed when co-immunoprecipitations are conducted from pooled sucrose

130 gradient fractions (S.F. and K.B.L., data not shown). This finding validates the importance of immunoprecipitation efficiency on detecting cofactor interactions and supports the use pooled gradient fractions to investigate cofactor interactions that are critical for translation activity. Collectively, these results confirm the above findings that engagement of FLAG-RHA within stable trRNP complexes facilitates critical translation initiation events.

In contrast, immunoprecipitation of FLAG-RHA K54A/55A/236E was limited to the early sub-polysome fractions of its sucrose gradient even when samples were pooled

(Figure 3.4b, compare top panels of the left and middle Western blot sets). This immunoprecipitation profile of FLAG-RHA K54A/55A/236E reflected its observed distribution when assessed directly from each fraction or individually immunoprecipitated (Figure 3.3 and 3.4a). Notably, when FLAG-RHA K54A/55A/236E was immunoprecipitated from the pooled sub-polysome fractions it did not co-precipitate

CBP80, rpS6, rpL5, or eIF2 (Figure 3.4b, middle Western blot panel set). This result reflects the impaired trRNP formation of FLAG-RHA K54A/55A/236E seen in cytoplasmic lysate prior to separation on a sucrose gradient and when FLAG-RHA

K54A/55A/236E is individually immunoprecipitated across a polysome profile (Figure

3.2 and 3.4a). This result confirms the above findings that the N-terminus of RHA is critical for coordinating productive trRNP formation that engages interactions with ribosome to promote translation activity.

Of significance is the observation that FLAG-RHA RA9 did not co-precipitate

CBP80, rpS6, rpL5, or eIF2 even when it was effectively immunoprecipitated from

131 pooled sub-polysome fractions like FLAG-RHA (Figure 3.4b, compare left and right

Western blot panel sets). Given FLAG-RHA RA9 effectively co-precipitates CBP80, rpS6, and rpL5 from cytoplasmic lysate prior to its separation on a sucrose gradient

(Figure 3.2c), this result confirms the above findings that the C-terminal RGG domain of

RHA is critical for stabilizing trRNP activity that promotes translation initiation events.

Mutation of the arginine residues within the RGG domain destabilizes the RHA trRNP, resulting in impaired polysome engagement.

Collectively, the co-immunoprecipitations and polysome profiles presented indicate a prominent role for RHA in translation initiation. This activity requires N- terminal-coordinated trRNP formation and C-terminal-mediated trRNP stability. Yet the presence of RHA within heavy polysome fractions and the ability of the FLAG-RHA

RA9 mutant to engage up through the first monosome fraction indicate a role for RHA in post-initiation translation activity. This effect could be linked to its established significance in trRNP stability or involve interactions of RHA with elongating ribosomes and/or elongation factors that promote polypeptide synthesis. Thus, the defect in the polysome profile distribution of FLAG-RHA RA9 could indicate impaired trRNP stability that results in 80S ribosome destabilization (a phenomenon known as ribosome drop-off in prokaryotes) or signify a block by the RA9 mutant in the actual steps of the elongation process. To discriminate between these potential mechanistic explanations for the post-initiation significance of RHA in the translation process, we performed ribosomal run-off assays with the translation inhibitors harringtonine and puromycin.

132

Ribosomal run-off assays allow for the identification of specific elongation defects by assessing the sensitivity of ribosome complexes and their associated factors to stalls in the translation process. Here cells are treated with a specific translation inhibitor to block polypeptide synthesis at a particular stage of the translation process. The fate of elongating ribosome complexes is then monitored by polysome profile analysis. An inhibitor commonly used in this assay is the translation initiation inhibitor harringtonine.

Harringtonine traps initiated 80S complexes at the start codon by blocking recruitment of subsequent amino-acyl transfer RNAs and interfering with peptidyl-transferase activity

(Fresno et al., 1977). This effectively inhibits new initiation events while allowing previously engaged ribosomes to complete the translation process (Ingolia et al., 2011).

Thus, complexes productively engaged in translation elongation are able complete a translation event but are unable to reinitiate. This outcome is evident in a polysome profile analysis with an observed shift in complex distribution from polysome to sub- polysome fractions upon drug treatment. Complexes with particular sensitivity to harringtonine indicate unstable elongating ribosomes, as translation inhibition promotes their already prevalent drop-off effect. This outcome is evident in a polysome profile analysis with a rapid shift from polysome to sub-polysome fractions upon a short (2 minute or less) harringtonine treatment (Ingolia et al., 2011; Petersen et al., 2006). In contrast, complexes stalled within the elongation process exhibit a delay in their completion of the allowable translation event. This outcome manifests in a delayed shift from polysome to sub-polysome fractions within a polysome profile analysis (Darnell et al., 2011; Shalgi et al., 2013).

133

To this end, cells transfected with FLAG-RHA or FLAG-RHA RA9 were treated with harringtonine for two minutes (Ingolia et al., 2011; 2012), processed, and then layered upon a 10-50% sucrose gradient and separated by ultracentrifugation. The distribution of FLAG-RHA or FLAG-RHA RA9 was then analyzed in designated sub- polysome or polysome fractions based upon the traced absorbance of the ribosomal RNA.

Monitoring of ribosomal protein S6 distribution across the polysome profile served as an internal control for effective treatment conditions.

A two-minute harringtonine treatment resulted in a significant decrease in total cellular polysomes and an accumulation of 80S monosomes within FLAG-RHA and

FLAG-RHA RA9 transfected cells, as assessed by the traced absorbance of the ribosomal

RNA (Figure 3.5a, compare polysome and sub-polysome peak heights in the ribosomal

RNA profile traces of control and harringtonine-treated cells). This characteristic change in the profile supported effective drug-induced translation inhibition and ribosome run- off, as previously reported (Fresno et al., 1977; Ingolia et al., 2011). Correspondingly,

FLAG-RHA exhibited a significant shift in its distribution towards sub-polysome fractions in the presence of harringtonine (Figure 3.5a, compare left FLAG-RHA

Western blot panels of control and drug-treated cells). This sensitivity of FLAG-RHA to general inhibition of translation initiation corroborated its established role in translation control (Hartman et al., 2006) and the findings of this study. Ribosomal protein S6 also demonstrated a significant shift in its distribution towards sub-polysome fractions with a prominent accumulation in the 80S peak (Figure 3.5a, compare left rpS6 Western blot panels of control and drug-treated cells). This outcome further supported effective

134 experimental conditions and that the sensitivity of FLAG-RHA to harringtonine indicates its productive stimulation of elongation.

Notably, we observed a rapid shift in the distribution of FLAG-RHA RA9 to sub- polysome fractions upon a two-minute harringtonine treatment (Figure 3.5a, compare right FLAG-RHA RA9 Western blot panels of control and drug-treated cells). This altered distribution of FLAG-RHA RA9 occurred with a corresponding shift of the ribosomal protein S6 to indicate a direct effect of harringtonine treatment (Figure 3.5a, compare right rpS6 Western blot panels of control and drug-treated cells). This result indicates reduced stability of 80S ribosome complexes associated with the RA9 mutant such that harringtonine treatment augments their drop-off effect.

To validate these findings, we performed a second ribosomal run-off assay with the translation elongation inhibitor puromycin. Puromycin is an amino-acyl-transfer RNA analog that binds the ribosome A-site and causes premature chain termination and ribosome release by accepting nascent polypeptide chains (Blobel & Sabatini, 1971;

Cannon, 1968). It is often used in ribosomal run-off assays for two purposes: (1) to assess the fate of elongating ribosome complexes in response to translation inhibition and (2) to demonstrate activate polypeptide synthesis (Nottrott et al., 2006). As with harringtonine, sensitivity to puromycin indicates functional elongation activity and is observed in a polysome profile analysis as a shift from polysome to sub-polysome fractions. Enhanced sensitivity to puromycin indicates unstable elongating 80S ribosome complexes, which are further promoted to displace within sub-polysome fractions upon drug treatment. In

135 contrast, stalled elongating 80S ribosome complexes exhibit reduced sensitivity to puromycin and corresponding delay in sub-polysome accumulation.

A one-hour puromycin treatment resulted in a significant decrease in total cellular polysomes and an accumulation of 80S monosomes within FLAG-RHA and FLAG-RHA

RA9 transfected cells, as assessed by the traced absorbance of the ribosomal RNA

(Figure 3.5b, compare polysome and sub-polysome peak heights in the ribosomal RNA profile traces of control and harringtonine-treated cells). This characteristic change in the profile supported effective drug-induced translation inhibition and ribosome run-off.

Correspondingly, FLAG-RHA exhibited a significant shift in its distribution towards sub- polysome fractions in the presence of puromycin (Figure 3.5b, compare left FLAG-RHA

Western blot panels of control and drug-treated cells). This sensitivity of FLAG-RHA to general inhibition of translation initiation corroborated its established role in translation control (Hartman et al., 2006) and the findings of this study. Ribosomal protein S6 also demonstrated a significant shift in its distribution towards sub-polysome fractions with a prominent accumulation in the 80S peak (Figure 3.5b, compare left rpS6 Western blot panels of control and drug-treated cells). This outcome further supported effective experimental conditions and that the sensitivity of FLAG-RHA to puromycin indicates its productive stimulation of elongation.

Notably, we observed a rapid shift in the distribution of FLAG-RHA RA9 to sub- polysome fractions upon puromycin treatment (Figure 3.5b, compare right FLAG-RHA

RA9 Western blot panels of control and drug-treated cells). This altered distribution of

FLAG-RHA RA9 occurred with a corresponding shift of the ribosomal protein S6 to

136 indicate a direct effect of puromycin treatment (Figure 3.5b, compare right rpS6 Western blot panels of control and drug-treated cells). Given puromycin targets 80S ribosomes competent for elongation, this result indicates that FLAG-RHA RA9 complexes reduce

80S ribosome stability such that puromycin treatment augments their drop-off effect.

Collectively, the data from these ribosomal run-off assays demonstrate a role for

RHA in promoting trRNP stability that facilitates effective post-initiation translation events. This conclusion is further supported by co-immunoprecipitations that investigated the association of FLAG-RHA and FLAG-RHA RA9 with elongation factors 1 and 2

(eEF1 and eEF2, respectively). eEF1 is critical for delivering amino-acylated tRNAs to elongating 80S ribosomes while eEF2 promotes their translocation (Dever & Green,

2012). FLAG-RHA did not co-precipitate either eEF1 or eEF2 from cytoplasmic cell lysate (Figure 3.5c, left Western blot panel set). It did, however, robustly co-precipitate

CBP80 to indicate effective experimental conditions that allowed for detectable cellular interactions (Figure 3.5c, left Western blot panel set). Likewise, FLAG-RHA RA9 also did not co-precipitate eEF1 or eEF2 from cytoplasmic lysate but did co-precipitate

CBP80 (Figure 3.5c, right Western blot panel set). Elongation pausing by RNA binding proteins and ribosome-associated factors has been attributed to altered elongation factor interactions (Friend et al., 2012; Hussey et al., 2011; Shalgi et al., 2013). The absence of an association between FLAG-RHA and FLAG-RHA RA9 with eEF1 and eEF2 as well as their lack of a detectable interaction with translation termination factors (S.F. and

K.B.L., data not shown) indicates significance to the findings of the ribosomal run-off

137 assays in that RHA is critical for trRNP stability that promotes 80S stabilization and productive elongation.

RHA exhibits functional prominence in early and late-stage initiation events.

The data show significance for RHA in initiation and post-initiation translation events. The molecular basis is N-terminal-coordinated trRNP formation and C-terminal- mediated 80S stabilization. A gap in knowledge, however, is the exact role for RHA in facilitating post-initiation translation activity. The above ribosomal run-off assays and co- immunoprecipitations indicate this effect is independent of coordinating elongation factor activity. Instead, the data support an earlier initiation event that drives productive 80S formation and stabilization. To directly test this hypothesis, we performed in vivo translation initiation assays with FLAG-RHA and designated mutants. Translation initiation assays are similar to polysome profile experiments except that cytoplasmic cell lysate is layered upon a 10-30% sucrose gradient. This narrower gradient range allows for enhanced resolution of initiation complexes and monitoring of translation activity during this stage of the process. Typically, translation initiation assays are conducted in vitro with synthesized RNA and purified translation components. However, our limitations on effectively producing purified RHA had us develop an in vivo system. Here cultured mammalian cells are serum starved for twenty-four hours to arrest and synchronize the translation process and then stimulated with serum for indicated time intervals to monitor the effects on translation activity once protein synthesis has been re-engaged.

138

To this end, cells transfected with FLAG-RHA were serum starved for twenty- four hours and then stimulated with serum for one or two hours, processed, and then layered upon a 10-30% sucrose gradient and separated by ultracentrifugation. The distribution of FLAG-RHA was then analyzed in designated peak fractions based upon the traced absorbance of the ribosomal RNA. Here four distinct complexes are observed:

(1) mRNP (non-translating trRNP complexes), (2) 48S (43S pre-initiation ribosome complexes engaged with mRNA), (3) 80S (elongation-competent 80S ribosomes), and (4) heavy polysome complexes (Figure 3.6b and d). Monitoring of ribosomal protein S6 and

L5 distribution across the profile served as an internal control for effective initiation complex separation and analysis.

Following one-hour serum stimulation, FLAG-RHA was robustly detected within the 48S and 80S peak complexes (Figure 3.6b). The corresponding distribution of rpS6 within both peaks as well as the distinct sedimentation of rpL5 in the 80S indicates effective initiation complex separation and analysis (Figure 3.6b). This result indicates significance for RHA in engaging early initiation events that involve 43S pre-initiation ribosome complex recruitment and scanning as well as late-stage initiation events associated with 80S ribosomal complex formation.

When cells transfected with FLAG-RHA were stimulated for two hours by serum,

FLAG-RHA was now detected within the heavy polysome peak as well as the top mRNP complexes (Figure 3.6b). The increase in peak heights across the profile as well as the maintained steady-state levels of FLAG-RHA signify an effect due to enhance translation activity of FLAG-RHA and not an alteration in its steady-state production or stability

139

(Figure 3.6a and b). Yet the maintained prominence of FLAG-RHA within the 48S and

80S peak complexes indicates functional significance for RHA in early and late-stage initiation events that drive productive post-initiation activity (Figure 3.6b).

We also performed the in vivo translation initiation analyses with FLAG-RHA

K54A/55A/236E and FLAG-RHA RA9 to investigate their impact on RHA translation initiation activity. In contrast to FLAG-RHA, FLAG-RHA K54A/55A/236E failed to engage in initiation complex formation following one-hour serum stimulation (Figure

3.6d). Instead, it retained within the non-translating mRNP peak fractions (Figure 3.6d).

This result shows impaired translation initiation engagement of an N-terminal RHA mutant and corroborates the earlier findings of this study of compromised trRNP formation for FLAG-RHA K54A/55A/236E (Figure 3.2). Notably, FLAG-RHA RA9 was observed in the 48S and 80S peak complexes (Figure 3.d). However, this engagement was not as robust as FLAG-RHA (Figure 3.6, compare FLAG-RHA Western blot panel in b to corresponding FLAG-RHA RA9 Western blot panel in d). Furthermore, the majority of FLAG-RHA RA9 was detected within the non-translating mRNP peak fractions like FLAG-RHA K54A/55A/236E (Figure 3.6d). Given FLAG-RHA

K54A/55A/236E and FLAG-RHA RA9 exhibited similar steady-state expression levels and profile traces (Figure 3.6c and d), these results confirm the previous findings of the study that the N-terminus of RHA coordinates trRNP formation to engage RHA in translation while the C-terminus mediates initiation complex stabilization to promote 80S formation and productive elongation.

140

DISCUSSION

In this study we elucidated the molecular significance of RNA helicase A (RHA) in the eukaryotic translation process. This activity includes functional prominence in early and late-stage initiation events that promote productive post-initiation translation activity. Notably, we identified molecular significance for the conserved N-terminal dsRBD lysine residues in engaging appropriate RHA trRNP formation that allows for

RHA's introduction into the translation process. Furthermore, we demonstrated a role for the C-terminal RGG domain of RHA in ribosome complex stabilization that promotes associated 80S formation and effective elongation. The outcome of these coordinated events is productive generation of target protein products. Figure 3.7 presents a model summarizing these findings.

The canonical dsRBD is known for its role in binding duplexed RNA (St

Johnston, Brown, Gall, & Jantsch, 1992). Yet, structural features of the dsRBD allow its importance to extend beyond this nucleic acid activity and involve protein interactions and ribosome-related functions (Gleghorn & Maquat, 2014; Luo, Duchaîne, &

DesGroseillers, 2002; Ricci et al., 2014). RHA harbors two adjacent dsRBDs within its

N-terminus that are important for target mRNA binding and subsequent protein synthesis

(Ranji et al., 2011). Herein we showed that the N-terminal dsRBDs of RHA were critical for its role in initiation by generating a productive trRNP complex. This activity extended beyond target mRNA binding to include stable CBC cap-binding association and ribosome association (Figure 3.2). Mutation of conserved, surface-exposed lysine residues within the dsRBDs disrupted these molecular activities and impeded the ability

141 of RHA to engage within initiation complex formation (Figure 3.3 and 3.6). Thus, the significance of the dsRBDs for RHA in its translation activity extends beyond target mRNA binding to include protein interactions and ribosome-related function. How the dsRBDs of RHA might coordinate these various molecular activities remains to be determined. A recent cryo-EM structure of the related DExH-box helicase DHX29 in complex with the 43S pre-initiation complex provided critical insight into the mechanism by which DHX29 controls translation initiation on its target mRNAs (Hashem et al.,

2013). Here DHX29 was identified near the mRNA entry channel latch of the 43S pre- initiation ribosome complex and was positioned in a manner to promote trRNP remodeling that facilitates efficient ribosome scanning (Hashem et al., 2013). This structure identification supported previous biochemical studies that indicated significance for the N-terminal dsRBD of DHX29 in ribosome binding and 48S complex formation

(Dhote et al., 2012). A similar study with RHA would be valuable for elucidating how the dsRBDs coordinate appropriate initiation complex formation on target viral and cellular transcripts.

The central catalytic core of RHA was also significant for its translation activity.

However, this molecular functioning was mutually exclusive from that of the dsRBDs, as an ATPase mutant still formed appropriate RHA trRNP complexes but exhibited impaired polysome engagement. These findings address an important consideration about the mechanism by which RHA, as a DEIH-box RNA helicase, regulates translation: mRNA unwinding or mRNP remodeling? The cognate RNA element to which RHA binds is a highly structured dual stem-loop that is both necessary for cap-dependent

142 translation while serving as a structural barrier to ribosome scanning (Hartman et al.,

2006) (and references therein). Given this feature of RHA-dependent translation control and the initial identification of the importance of the ATP-dependent helicase function of

RHA in target HIV-1 translation, a role was favored for RHA in mRNA unwinding whereby binding of the dsRBDs to the cognate RNA element positioned the helicase domain in manner to resolve the secondary structure (Bolinger et al., 2010; Hartman et al., 2006; Ranji et al., 2011). However, the molecular findings from this study, in addition to a previous report showing RHA to be relatively inefficient at unwinding short RNA duplexes (S. Zhang & Grosse, 1997), support a role for RHA in mRNP remodeling.

Functional and structural studies of related DHX29 and DDX3 indicate a similar mechanistic significance for these DExH-box RNA helicases in translation control

(Dhote et al., 2012; Geissler et al., 2012; Hashem et al., 2013; Pisareva et al., 2008).

Thus, the findings from this study contribute evidence to the complexity of DExH-box

RNA helicases in translation control in a manner that extends beyond simple mRNA unwinding.

A third major finding from this study was the identification of a post-initiation role for the C-terminal RGG domain of RHA. Like the dsRBD, RGG/RG motifs function in nucleic-acid binding, protein interactions, and ribosome-related functions critical for posttranscriptional gene control (Thandapani et al., 2013). We previously demonstrated that the C-terminal RGG domain of RHA was not significant in target mRNA binding

(Ranji et al., 2011). However, mutation of the basic arginine residues impaired RHA- dependent translation activity, supporting a role for the RGG domain in facilitating

143 protein synthesis (C.B. and K.B.L., data not shown). The polysome profiling and ribosomal run-off assays conducted in this study revealed that a C-terminal RA9 mutant of RHA failed to allow extended polysome formation even though it exhibited productive trRNP formation, an effect attributed to ribosome drop-off (Figure 3.2, 3.3, and 3.5).

Furthermore, in vivo translation initiation assays showed that the C-terminal RA9 mutant was able to engage in 43S and 80S ribosome complex formation (Figure 3.6). However, its reduced stabilization resulted in the majority dissociating from productive translation and accumulating in non-translating mRNP fractions (Figure 3.6). This data implied significance for the C-terminal RGG domain in 80S ribosome stabilization that promotes efficient elongation.

Thus far, the majority of RGG-domain containing proteins have been implicated in translation repression (Chen et al., 2014; Darnell et al., 2011; De Leeuw et al., 2007;

Rajyaguru et al., 2012; Solomon et al., 2007). Although the mechanisms for most have yet to be elucidated, the RGG motif of Scd6 was shown to repress translation by binding the initiation factor eIF4G and inhibiting 48S complex formation (Rajyaguru et al.,

2012). Here we provide evidence for a positive role of an RGG domain in facilitating post-initiation translation activity. How the RGG domain of RHA mechanistically stabilizes ribosomes to facilitate elongation and the number of rounds this is critical for remains to be determined. RGG domains and related basic C-terminal tails lack defined structure and engage specific conformations upon substrate-induced binding (Mallam et al., 2011; Phan et al., 2011). It is possible that the flexible conformation of RHA's C- terminal RGG domain allows for its embrace of translationally active 80S ribosomes in a

144 manner that stabilizes the complex to allow for efficient elongation kinetics. Mutation of the basic arginine residues could induce rigidity to the RHA-ribosome complex that impairs this stabilization and stimulates ribosome dissociation. A related concept was proposed for the mechanism by which the yeast Ltn1 protein functions in ribosome quality control (Lykke-Andersen & Bennett, 2014). Ltn1 exhibits an extended flexible structure, which was proposed to confer its ability to contact the ribosome in a manner that facilitates its function in influencing ribosome structure dynamics (Lykke-Andersen

& Bennett, 2014; Lyumkis et al., 2013). Maintenance of RHA's flexible C-terminal domain could similarly allow its interaction with the ribosome; however, here it is in a manner that facilitates ribosome stabilization.

Although RHA was not identified to interact with either elongation factor to possibly explain its molecular significance in post-initiation translation (Figure 3.5c), it does associate with the cytoplasmic poly-A binding protein (PABP) in a manner that requires its C-terminal RGG domain (S.F. and K.B.L., data not shown). Given the position of RHA's cognate RNA element near the 5' cap of target mRNAs, it is possible that the interaction of 5'-bound RHA with 3'-bound PABP is critical for mRNA circularization, which facilitates efficient post-initiation activity. Disruption of RHA-

PABP interaction via mutation of the RGG domain could impede important trRNP stability that results in decreased post-initiation efficiency and ribosome drop-off.

The RGG domain of RHA is subjected to methylation by protein-arginine methyltransferase 1 and this posttranslational modification regulates its nuclear localization (Smith, Schurter, Wong-Staal, & David, 2004). Although the FLAG-RHA

145

RA9 mutant does not exhibit impaired cellular distribution at steady state (A.R. and

K.B.L., data not shown), it is possible that its real-time nucleo-cytoplasmic shuttling was impacted in a manner that impaired its ability to engage in appropriate trRNP complex formation resulting in inefficient protein synthesis and ribosome dissociation.

Given the data presented herein, it can be envisioned that RHA associates with its target transcript via direct binding between its N-terminal dsRBDs and 5' cognate RNA element (Figure 3.7). When appropriately localized and directed by the cellular environment, this association nucleates active trRNP complex formation via recruitment of and association with the 43S pre-initiation complex by the dsRBDs (Figure 3.7). The

ATP-dependent helicase function of RHA drives mRNP remodel activity to allow for appropriate ribosome scanning and start codon recognition (Figure 3.7). The intricate role of the dsRBDs to initiation complex formation maintains significance in 60S ribosomal subunit joining. The tethered C-terminal RGG domain is stimulated to interact with the initiating trRNP complexes in a manner that facilitates efficient progression into and potentially extension of polypeptide formation (Figure 3.7). Thus, the DEIH-box RNA helicase RHA/DHX9 functions in both initiation and post-initiation translation control via the intricate molecular functioning of its conserved protein domains.

146

Figure 3.1 Depiction of RHA domain structure with critical amino acids indicated. RHA harbors two N-terminal double stranded RNA binding domains (dsRBD I and II), a central DEIH helicase core, a helicase-associated 2 domain (HA-2), and a C-terminal domain of unknown function (DUF) and arginine-glycine-rich (RGG) domain. Critical domain residues and the corresponding amino acid substitutions analyzed in this study for functional significance in the translation process are indicated. Figure was adapted from (Ranji & Boris-Lawrie, 2010).

147

Figure 3.2 N-terminal lysine residue 236 is necessary for a RHA-CBP80 association and RHA trRNP formation. (a) N-term RHA and associated mutants were immunoprecipitated from HEK293 cell lysate with anti-FLAG-conjugated agarose beads (Sigma) and probed for the coprecipitation of the indicated proteins by Western blot. Shown is representative Western blot images. Input, 2% of cell lysate. IP, immunoprecipitated complex. FT, 2% of flow-through following IP. (b) Co- immunoprecipitations as in (a) with FLAG-RHA or associated full-length mutants. (c) Co-immunoprecipitations as in (a), with FLAG-RHA or associated full-length mutants from HEK293 cytoplasmic cell lysate following a 24 hr culture in media containing either 10% serum or 0.5% serum.

148

Figure 3.3 Mutation of the C-terminal RGG domain impacts RHA polysome association in a manner distinct from compromised N-terminal function. Polysome profile analysis of FLAG-RHA or designated mutants across a 10-50% sucrose gradient. A254 traces of total RNA distribution are shown. Representative Western blots of indicated protein distribution across the profiles are presented.

149

Figure 3.4 The C-terminal arginine residues of RHA are critical for stable sedimentation of the RHA trRNP through a sucrose gradient. (a) Polysome profiles from Figure 3.3 analyzed by immunoprecipitation of FLAG-RHA or designated mutants across the profile. Representative Western blot images are shown for the co-precipitation of indicated proteins. (b) Co-immunoprecipitations as in (a), with gradient fractions combined from indicated regions of the polysome profile trace. Representative Western

blot images are shown for the co-precipitation of indicated proteins. Continued

150

Figure 3.4 continued

151

Figure 3.5 Ribosomal run-off assays support a role for RHA's C-terminal RGG domain in initiation complex stabilization. (a) Ribosomal run-off of FLAG-RHA (left panels) or the associated FLAG-RHA RA9 mutant (right panels) in the presence of the translation initiation inhibitor harringtonine. Overlaid A254 traces of total RNA distribution for each gradient are shown. Representative Western blot images of FLAG- RHA distribution across the profile are presented. The small ribosomal S6 protein (rpS6) was used to assess ribosome abundance in each gradient fraction. (b) Ribosomal run-off assays as in (a), with the translation elongation inhibitor puromycin. (c) FLAG-RHA (left panels) or the associated FLAG-RHA RA9 mutant (right panels) were immunoprecipitated from cytoplasmic HEK293 cell lysate with anti-FLAG-conjugated agarose beads (Sigma) and probed for the coprecipitation of the indicated proteins by Western blot. Shown is representative Western blot images. Input, 2% of cell lysate. IP,

immunoprecipitated complex. FT, 2% of flow-through following IP. Continued

152

Figure 3.5 continued

153

Figure 3.6 In vivo translation initiation assays demonstrate significance for RHA in early and late-stage initiation events. (a) Steady-state expression of FLAG-RHA following 1 or 2 hr stimulation with 10% FBS after a 24 hr culture in serum-deprived (0.5% FBS) media. (b) Analysis of FLAG-RHA across a 10-30% sucrose gradient following 1 or 2 hr stimulation with 10% FBS after a 24 hr culture in serum-deprived (0.5% FBS) media. (c) Steady-state expression of designated FLAG-RHA mutants following 1 hr stimulation with 10% FBS after a 24 hr culture in serum-deprived (0.5% FBS) media. (d) Analysis of designated FLAG-RHA mutants across a 10-30% sucrose gradient following 1 hr stimulation with 10% FBS after a 24 hr culture in serum-deprived (0.5% FBS) media.

154

Figure 3.7 Model of the molecular role of RHA in eukaryotic translation. RHA engages in translation activity by first binding a target transcript. This association occurs with the recognition of its cognate cis-acting RNA element, the posttranscriptional control element (PCE). Surface exposed lysine residues within the N-terminus of RHA govern PCE recognition and binding. Upon mRNA association, RHA engages trRNP formation that facilitates 43S pre-initiation ribosome complex (43S PIC) recruitment and scanning. This activity requires ATP-dependent mRNA unwinding and/or trRNP remodeling by RHA. Distinctly, RHA remains associated with initiated 43S PICs at the start AUG to facilitate stable 80S ribosome formation and engagement in elongation. This late-stage initiation activity requires the C-terminal RGG domain of RHA. Thus, RHA is a DExH-box RNA helicase import for targeted translation activity by facilitating both early and late-state initiation events. Notably, this molecular effect is achieved by genetically separable roles of RHA's conserved terminal domains in the translation process.

155

Chapter 4 : The Terminal Domains of RNA Helicase A Coordinate Its Cellular

Multimerization and Select Interaction with Related DHX30: Implications for

Regulated Translation Activity

ABSTRACT

Ribonucleoprotein (RNP) complexes are the fundamental effectors of posttranscriptional gene control. Their compositions and arrangements regulate RNA subcellular localization, conformation, and activation, all features that control mRNA maturation, nuclear export, and translation. RNA helicase A (RHA) is the critical RNA binding protein and DEIH-box RNA helicase that identifies the RNP fundamental to targeted gene expression of retroviral transcripts and the cellular proto-oncogene junD.

The association of RHA with its target mRNA engages an RNP productive for translation activity. This RHA translation RNP (trRNP) is distinctly defined by an association with the non-canonical CBP80/20 cap-binding protein complex and exhibits a role in 43S pre- initiation ribosome complex recruitment, scanning, and 80S stabilization. Yet to be defined, however, are the molecular interactions regulating RHA trRNP formation and its subsequent activation in protein synthesis. Herein we identify distinct homodimeric and heterodimeric RHA associations that exhibit critical implications for regulated RHA trRNP formation and translation activity. RHA is demonstrated to self-associate within

156 cells, an intermolecular binding event mediated by its N- and C-terminal domains.

Notably, isolated RHA multimers exhibit impaired associations with CBP80 and the cytoplasmic poly-A binding protein. This feature of homodimeric RHA complexes indicates significance for an inherent molecular basis of regulated RHA trRNP formation and translation activity. Furthermore, a select interaction between RHA and the related

DEVH-box RNA helicase DHX30 is identified. This association also requires the N- and

C-terminal domains of RHA; however, the residues necessary for this heterodimeric binding event are distinct from those required in RHA multimerization. Notably, the association of RHA with DHX30 coincides with its CBP80 association and is observed on polysomes in a manner that is sensitive to targeted disruption of RHA trRNP formation or general translation inhibition. This finding indicates a role for a RHA-

DHX30 association in coordinating RHA trRNP formation that facilitates functional translation activity. We therefore propose that three distinct cellular RHA RNPs exist to regulate its translation activity: a RHA multimer RNP, a RHA trRNP, and a translationally arrested RHA trRNP. Collectively, these RHA RNPs coordinate the molecular basis of RHA in eukaryotic protein synthesis.

INTRODUCTION

Ribonucleoprotein (RNP) complexes exhibit significance in eukaryotic protein synthesis in ways that extend beyond engagement of ribosome activity to encompass critical regulatory functions as well. A defining RNP property that can exert this effect is cellular multimerization (Bleichert & Baserga, 2010). Related RNA molecules, by virtue

157 of base or structure complimentary, can interact in a manner that unites two similar RNP complexes. The outcome is distinct functional effects on inherent translation activity.

Such is observed for regulated translation of the HIV-1 unspliced RNA. The HIV-1 unspliced RNA serves two essential purposes: (1) as an mRNA template for the synthesis of viral Gag and Gag-pol proteins and (2) as the genomic RNA that is packaged into assembling virions. Packaging of the unspliced viral RNA is characterized by its dimerization, an intermolecular binding event mediated by associations between two cis- acting packaging signals (Nikolaitchik et al., 2013). Although translation of the unspliced

RNA is not required for its packaging (Butsch & Boris-Lawrie, 2000), differing RNP structures and compositions are fundamental to distinguishing a translating HIV-1 unspliced RNA from a packaging HIV-1 unspliced RNA (I.B. and K.B.L., unpublished data). This characteristic implies significance to RNP biology in the coordination of viral

RNA translation versus packaging that is fundamental to HIV-1 replication.

Likewise, RNA binding proteins of RNPs can oligomerize and unite related RNP complexes with distinct functional consequences on inherent translation activity.

Intermolecular self-association of the RNA binding protein Staufen1, for example, is fundamental to its role in targeted translation and the translation-associated process of

Staufen1-mediated mRNA decay (Gleghorn, Gong, Kielkopf, & Maquat, 2013; Martel et al., 2010). Staufen1 homodimerization is mediated by distinct protein-protein interactions between two-adjacent C-terminal domains: the Staufen-swapping motif (SSM) and the dsRNA-binding domain 5 (RBD5) (Gleghorn et al., 2013). The two alpha helices of the

SSM of one Staufen1 molecule interact with a hydrophobic patch of the RBD5 of a

158 second Staufen1 molecule in a manner that promotes a homodimer complex capable of enhanced binding to the RNA helicase UPF1(Gleghorn et al., 2013). A Staufen1-UPF1 association is a fundamental trigger for activation of Staufen1-mediated mRNA decay

(Gleghorn et al., 2013; E. Park & Maquat, 2013). Disruption of Staufen1 homodimerization by overexpression of competing peptides impaired RNP dynamics so as to result in inhibited Staufen1-mediated mRNA decay (Gleghorn et al., 2013). It has also been demonstrated that related Staufen2 can self-associate or heterodimerize with

Staufen1 in a manner that generates a primed RNP complex for Staufen1-mediated mRNA decay (E. Park, Gleghorn, & Maquat, 2013). This feature indicates that cellular multimerization is fundamental to Staufen1-mediated translation and the translation- associated process of Staufen1-mediated mRNA decay.

Numerous additional examples exist of RNA binding proteins forming dimeric complexes with significant functional consequences on associated RNP activity.

Intermolecular interactions of double-stranded RNA binding domains (dsRBDs) often mediates this effect (Cole, 2007; Daher et al., 2001; Peters, Hartmann, Qin, & Sen, 2001;

Valente & Nishikura, 2007). Yet the known diversity and prevalence of dsRBDs throughout eukaryotic RNA binding proteins implies directed and regulated associations

(Gleghorn & Maquat, 2014).

Furthermore, RNA helicases are known to engage dimer formation for regulated

RNP activities. Such is seen for the virally encoded DExH-box RNA helicase NS3 of the hepatitis C virus and the DEAD-box RNA helicase Hera of Thermus thermophilus (Khu et al., 2001; Klostermeier & Rudolph, 2009). NS3 self-associates via interactions

159 between N-terminal regions of its helicase core (Khu et al., 2001). This association requires nucleic acid binding and is essential for demonstrated NS3 helicase activity (Khu et al., 2001). Hepatitis C virus is a RNA virus that does not demonstrate a DNA intermediate. Thus, RNA-based activities, like NS3-driven RNP remodeling, are fundamental to it survival and propagation. This effect is intricately linked to and regulated by NS3's dimerization ability.

Hera self-associates via intermolecular interactions between its C-terminal alpha- helical dimerization motif (Klostermeier & Rudolph, 2009). Crystal structures of the

Hera dimer showed that in this self-associated conformation its adjacent RNA binding domains become juxtaposed in a manner that supports coordinated RNA binding

(Klostermeier & Rudolph, 2009). This feature is proposed to stimulate helicase activity

(Klostermeier & Rudolph, 2009). Furthermore, formation of a Hera dimer does not impose rigidity onto its structure (Klostermeier & Rudolph, 2009). Instead, its RecA helicase domains are afforded maintained flexibility (Klostermeier & Rudolph, 2009).

This data supports a model whereby Hera dimerization facilitates its helicase-driven RNP remodeling effects that are fundamental to its associated biological activities

(Klostermeier & Rudolph, 2009).

RNA helicase A (RHA/DHX9) is a DEIH-box RNA helicase significant for the regulated expression of retroviral transcripts and the cellular proto-oncogene junD

(Bolinger et al., 2007; 2010; Hartman et al., 2006; Ranji et al., 2011). The data presented in this dissertation thus far has demonstrated that RHA mediates its translation effect by engaging a distinct translation RNP (trRNP) complex with the non-canonical CBP80/20

160 cap-binding protein and functioning in 43S pre-initiation ribosome complex recruitment, scanning, and 80S stabilization (Chapters 2 and 3, respectively). Yet to be defined, however, are the features of the RHA trRNP regulating its translation activity. RHA multimerization, either self-association or oligomerization with related DExH/D-box

RNA helicases, provides a molecular basis for this effect.

RHA is both a RNA binding protein and RNA helicase. The auxiliary domains flanking its defining helicase core promote this functional diversity. These include two

N-terminal double-stranded RNA binding domains (dsRBDs), a helicase-associated 2 domain, domain unknown function, and C-terminal arginine-glycine (RGG)-rich domain

(Ranji & Boris-Lawrie, 2010) (Figure 3.1 and 4.4). Notably, this multi-domain architecture of RHA provides opportunities for self-association. Indeed preliminary data indicated a role for N- and C-terminal interactions in bridging two RHA molecules in cells (W.J. and K.B.L., unpublished data). Furthermore, the prevalence of these and related auxiliary domains among other members of the DExH/D-box RNA helicase family imply possibilities for heterodimeric interactions that regulate RHA translation activity. Indeed an association between RHA and the DEAD-box RNA helicase DDX3 has been identified in relation to regulated nuclear-export activity (Lai et al., 2008). It remains to be determined whether a RHA-DDX3 association or the interaction of RHA with other DExH/D-box RNA helicases is significant to regulated translation activity.

This chapter presents preliminary data demonstrating the presence of RHA multimers and RHA-DHX30 complexes in cells. The N-terminal dsRBDs and C-terminal

RGG domain of RHA are shown to coordinate these associations. Notably, RHA

161 multimers and RHA-DHX30 complexes are identified as distinct RNPs with differing implications for RHA-mediated translation activity. RHA multimers interact in a manner that competes for translation cofactor binding, affording an inherent RNP complex regulating RHA's engagement in translation. Alternatively, a RHA-DHX30 interaction was identified in polysome-associated RHA trRNP complexes. This finding indicates significance for a DHX30 association in mediating RHA translation activity. Taken together, the data presented in this chapter provide critical insight into molecular bases for regulated RHA translation activity.

MATERIALS AND METHODS

Cell culture and transfections

HEK293 cells were grown in Eagle's Minimum Essential Medium (ATCC) supplemented with 10% FBS (Gibco) and 1% antibiotic (Gibco). One day prior to transfection, cells were seeded in 10 cm plates at a density of 1.5×106 cells/plate. Each plate of cells was transfected with 2 μg of designated FLAG-RHA, HA-RHA, or HA-

DHX30 plasmid (see Plasmid Construction for more details) using the X-tremeGene HP

DNA Transfection Reagent (Roche) at a ratio of 1:3. Media was changed 24 hr post- transfection and cells harvested at forty-eight.

SDS-PAGE and Western Blotting

Protein samples were run on 4-15% gradient SDS-polyacrylamide gels (BIO-

RAD) at 200 V for 45 minutes and transferred to nitrocellulose membranes (BIO-RAD)

162 via a wet transfer at 80 V for 1 hour by standard methods. Membranes were incubated in

The Fuji Imaging System (FUJIFILM) and MultiGauge Software Program (FUJIFILM) were used to develop and assess protein band intensity.

Antibodies

The following primary antibodies were used for Western blot analysis: anti-FLAG

(Sigma, F7425) at 1:5000, anti-HA (Covance, MMS-101R) at 1:1000, anti-RHA

(Vaxron, PA-001) at 1:7500, anti-DHX30 (Bethyl, A302-218A1) at 1:1000, anti-DHX29

(Bethyl, A300-751A) at 1:1000, anti-DDX3 (Bethyl, A300-474A) at 1:1000, anti-

NCBP1/CBP80 (Bethyl, A301-793A) at 1:1000, anti-eIF4E (Cell Signaling, #9742S) at

1:1000, anti-PABPC (Abcam, ab6125) at 1:1000, and anti-GFP (Abcam, 290) at 1:5000.

DHX30 (Bethyl, A302-218A1), anti-DHX29 (Bethyl, A300-751A), or anti-DDX3

(Bethyl, A300-474A) were used according to manufacture recommendations based upon the concentration of input cell lysate.

163

Plasmid Construction

The FLAG-RHA and GST-RHA constructs were previously described (Hartman et al., 2006; Jin et al., 2011; Ranji et al., 2011). Cloning of designated FLAG-RHA mutants was described in Chapter 3. YFP-RHA was generated by digesting FLAG-RHA with EcoRI and ligating the insert into pEYFPC1 cut with EcoRI (performed by W.J.).

HA-DHX30 and the associated helicase mutant were gifts from Chen Liang.

GST pull-down assays

GST pull-downs were conducted as previously described (Jin et al., 2011).

Immunoprecipitations

7.4, 150 mM NaCl, 0.1 mM EDTA, 0.5% NP40) and twice with 1x wash buffer. Loading buffer was added to isolated bead complexes, heated, and loaded directly onto the gel.

2% of the initial cleared lysate was reserved and diluted with 1x wash buffer and run as

INPUT for each IP. 2% of the flow-through following IP was reserved and run as FT for

164 each IP. When indicated, cells were lysed in the presence of 100 ug of RNase A prior to

IP.

For tandem IPs, FLAG IPs were conducted as above. Isolated FLAG complexes were then incubated with 100 μg of competing FLAG peptide (Sigma) and then incubated with end-over-end rotation for one hour. Samples were spun to collect eluted FLAG complexes and the eluent was subjected to HA IP as described above.

For endogenous IPs, designated primary antibody was incubated with magnetic

1 mL with 1x wash buffer. Immunoprecipitated complexes were then washed three times with NETN-150 wash buffer. Samples were prepared and loaded directly onto the gel as described. When indicated, cells were lysed in the presence of 100U RNase 1 prior to IP.

Polysome Profiling

For 10-50% sucrose gradients, experiments previously described and analyzed in

Chapters 2 and 3 were re-probed for DHX30.

For the 5-25% sucrose gradient, HEK293 cells were treated with 0.1 mg/mL cyclohexamide (CHX; prepared in DMSO, Sigma C1988) for 20 minutes, then washed twice with ice cold PBS with CHX, and scraped off the plates in 5 mL of ice cold PBS with CHX. Resuspended cells were pelleted by spinning at 1500 rpm for 4 min at 4C.

Supernatant was removed and cell pellet resuspended in 0.75 mL of low salt buffer (20

165 mM Tris-HCl, pH 7.5, 3 mM MgCl2, 10 mM NaCl, 2 mM DTT, 1X protease inhibitor cocktail EDTA-free (Roche 11 873 580 001), 5 μL/mL RNase Out). Cells were allowed to swell on ice for 5 min and then lysed by the addition of 0.25 mL lysis buffer (0.2 M sucrose and 1.2% Triton X-100 prepared in low salt buffer) followed by 10 strokes with a dounce homogenizer (Kimble Chase). Lysate was purified by spinning at top speed

(16,000 x g) for 1 minute at 4C, then layered onto a 5-25% sucrose gradient (10 mM

Tris-HCL, pH 7.5, 5 mM MgCl2, 100 mM KCl, 2 mM DTT, 0.1 mg/mL CHX, 10/50% sucrose), and spun in an ultracentrifuge for 2 hr 40 min at 38,000 RPM at 4C. The polysome profile was generated by continuous monitoring of RNA absorbance at 254 nm by the ISCO UA-6 Absorbance Detector unit and fractionated into 22 equivalent volume

(0.5 mL) fractions using the ISCO Foxy R1 fraction collector. Brandel Peak Trace

Software was used to generate the corresponding profile trace.

RESULTS

RHA exhibits self-association within cells; an effect coordinated by interactions between its N- and C-terminal domains.

Preliminary GST pull-down data indicated a select affinity between the N- terminal dsRBDs of RHA and its C-terminal RGG domain (W.J. and K.B.L., data not shown). Extension of this analysis to full-length RHA showed that GST-RGG was sufficient for interacting with FLAG-RHA and FLAG-RHA ΔRGG from mammalian cell lysate (W.J. and K.B.L., data not shown). Deletion of the dsRBDs, however, impaired the association between FLAG-RHA and GST-RGG to support a select interaction between

166 the N- and C-terminal domains of RHA (W.J. and K.B.L., data not shown). These results implied significance for intra- or intermolecular associations of RHA molecules in regulating its cellular activity. Indeed a preliminary analysis identified intermolecular associations between FLAG epitope-tagged RHA and HA epitope-tagged RHA in cells

(W.J. and K.B.L., data not shown).

To validate the ability of RHA to engage in cellular multimerization, tandem immunoprecipitations were performed with transfected HEK293 cell lysate and analyzed for the isolation of FLAG-RHA:HA-RHA complexes. Figure 4.1a is a schematic overview of the tandem immunoprecipitation process. HEK293 cells transfected with a

FLAG epitope-tagged RHA and a HA epitope-tagged RHA expression construct were harvested, lysed, and incubated with sepharose beads pre-conjugated to anti-FLAG antibody. This first immunoprecipitation isolated all FLAG-RHA RNP complexes in the cell. After extensive washing, the isolated FLAG-RHA RNP complexes were eluted from the FLAG sepharose beads by incubation with competing FLAG peptide. Addition of the eluent, which contained all effectively eluted FLAG-RHA RNP complexes, to sepharose beads pre-conjugated with anti-HA antibody allowed for the exclusive isolation of

FLAG-RHA:HA-RHA RNP complexes. Assessment of the outcomes of the tandem immunoprecipitation was performed by Western blot analysis. Figure 4.1b presents representative results.

First it was noted that both FLAG-RHA and HA-RHA were efficiently and equivalently expressed in HEK293 cells using the methods described in Materials and

Methods (Figure 4.1b, left input lanes). This validated a robust experimental set-up to

167 assess intermolecular RHA associations. Analysis of the flow-through following the first

FLAG immunoprecipitation showed detection of HA-RHA but not FLAG-RHA molecules (Figure 4.1b, FT after FLAG IP lane). This result indicated effective immunoprecipitation of all FLAG-RHA complexes from the HEK293 cell lysate and, notably, that not all expressed HA-RHA was engaged with FLAG-RHA. Analysis of the

FLAG sepharose beads after FLAG elution was performed to assess elution efficiency.

Figure 4.1b shows that a portion of FLAG-RHA complexes remained bound to the FLAG sepharose beads even after elution with the competing FLAG peptide (Beads after FLAG elution lane). This result is significant for considering the implications of the second HA immunoprecipitation given the detection of cellular interactions by an immunoprecipitation approach is limited by sensitivity. Notably, analysis of the HA IP lane showed a strong detection of both FLAG-RHA and HA-RHA (Figure 4.1b). This result identified FLAG-RHA:HA-RHA cellular complexes and confirmed the previous findings of RHA multimers within cells. FLAG-RHA was distinctly detected in the flow- through after the second HA immunoprecipitation to indicate that select FLAG-RHA

RNP complexes eluted from the FLAG sepharose beads were not engaged in self- association (Figure 4.1b, FT after HA IP lane). These results were observed irrespective of whether FLAG-RHA and HA-RHA were co-expressed or if their respective expression lysates were mixed, signifying a strong affinity between RHA molecules (S.F. and K.B.L, data not shown).

This FLAG-RHA, HA-RHA tandem immunoprecipitation was further performed in the presence of RNase to determine the significance of RNA-binding in mediating

168

FLAG-RHA:HA-RHA cellular interactions. Figure 4.1c shows effective identification of both FLAG-RHA and HA-RHA within the HA IP lane even in the presence of RNase.

This result indicates that a molecular bridge through RNA is not required for RHA self- association. Collectively, these data validate our previous findings of RHA multimers existing within cells and extends our understanding of the molecular associations mediating this event by demonstrating that RNA is not required to bridge RHA molecules together.

To confirm the domains necessary for RHA self-association, we performed GST pull-down assays using HEK293 cell lysate as a source of full-length endogenous RHA and probing for its interaction with indicated GST-tagged RHA domains isolated from bacterial lysate (Figure 4.2a and b). The GST-tagged RHA domains selected for assessment with endogenous RHA represented all conserved domains across the RHA protein (Figure 3.1 and 4.4) and were cloned to maintain appropriate conformational folding. The left Coomassie panels of Figures 4.2a and b show effective expression and pull-down of each GST-tagged RHA domain. The assessment of GST alone served as the control for non-specific interactions. Notably when these GST-RHA pull-downs were assessed by Western blot analysis, both the N-terminal 300 residues of RHA (GST-N300) and its C-terminal RGG domain were sufficient for interacting with endogenous RHA of mammalian cell lysate (Figures 4.2a and b, right Western blot panels). This outcome was observed both in the absence and presence of RNase treatment to corroborate the findings identified above that a RNA bridge is not required for RHA self-association (Figures 4.2a and b, right Western blot panels, ± RNase). Furthermore, when substitution mutations

169 were introduced into the GST-tagged N- and C-terminal domains only mutation of the C- terminal arginine residues (GST-RA9) compromised association with endogenous RHA of mammalian cell lysate (Figure 4.2b, right Western blot panels). Mutation of critical surface exposed lysine residues of the N-terminus, which were previously demonstrated as significant for RHA trRNP formation and translation engagement (Chapter 3), did not impact RHA self-association (Figure 4.2b, right Western blot panels). This result indicates a molecular distinction between the roles of the N- and C-termini in RHA trRNP formation from that of their significance in mediating RHA self-association.

Cellular co-immunoprecipitations validated the above in vitro GST pull-down findings by demonstrating that FLAG-RHA effectively co-precipitates an YFP epitope- tagged RHA construct in mammalian cell lysate (Figure 4.2c, left Western blot panel set).

Expression and precipitation of a FLAG-RHA K54A/55A/236E mutant, in which the indicated N-terminal lysine residues of RHA were mutated from lysine to alanine, did not impact the ability of RHA to self-associate with an YFP-RHA counterpart (Figure 4.2c, middle Western blot panel set). Added mutation of the C-terminal arginine residues, however, did compromise the co-precipitation between FLAG-RHA K54A/55A/236E and YFP-RHA (Figure 4.2c, right Western blot panel set). This result corroborates the findings above that N- and C-terminal interactions are critical for mediating RHA self- association in a manner that is distinct from their roles in RHA trRNP formation.

170

RHA multimers compete for translation cofactor binding, indicating significance for regulated RHA translation activity.

The results presented thus far indicate the prevalence of RHA multimers in cells and that this self-association is mediated by intermolecular interactions between RHA's

N- and C-terminal domains. Notably, the data indicate a molecular distinction between the roles of RHA's N- and C-terminal domains in RHA self-association from that of their significance in RHA trRNP formation. While N-terminal lysine residues 54, 55, and 236 are dispensable for RHA multimer formation, they are fundamental to RHA trRNP formation (compare findings in Figure 4.2 to that in Figure 3.2). In contrast, C-terminal arginine residues are critical for RHA multimer formation but are dispensable for RHA trRNP formation (although, they are significant for trRNP stabilization) (compare findings in Figure 4.2 to that in Figure 3.2). This data signifies possible functional differences between RHA multimer RNPs and RHA trRNPs.

To assess the functional relationship between RHA multimer RNPs and RHA trRNPs, we re-examined the tandem immunoprecipitation blots from Figure 4.1b for interactions of isolated FLAG-RHA:HA-RHA complexes with translation cofactors that define the RHA trRNP. Specifically, we focused on the co-precipitation of CBP80 and

PABPC within the HA IP lane. CBP80 is a component of the non-canonical CBP80/20 cap-binding protein complex that was identified as a distinguishing component of active

RHA trRNPs (Chapters 2 and 3). PABPC is the cytoplasmic poly-A binding protein that was also identified as an active component of RHA trRNPs (Chapter 2). Figure 4.3 shows that when the tandem immunoprecipitation blots were re-probed for CBP80 and PABPC,

171 a co-precipitation of either protein was not detected in the HA IP lane. In comparison, a single FLAG-RHA immunoprecipitation showed robust co-precipitation of CBP80 and

PABPC (Figure 4.3, right Western blot panel set). This result indicates a functional distinction between RHA multimer RNPs and RHA trRNPs and signifies a possible role for RHA multimerization in regulating RHA translation activity.

RHA selectively associates with related DEVH-box RNA helicase DHX30.

In addition to self-association, heterodimerization with related DExH/D-box RNA helicases affords another molecular basis for regulated RHA translation activity. Three distinct candidates include: DHX30, DHX29, and DDX3. DHX30 is structurally similar to RHA with N-terminal dsRBD(s), a central helicase core, and a flanking helicase- associated (HA2) domain and domain of unknown function (DUF) (Figure 4.4). Notably,

DHX30 lacks a C-terminal RGG domain yet like RHA is significant for replication of

HIV-1 (Y. Zhou et al., 2008). Here DHX30 stimulates transcriptional activity from the viral promoter in a manner that is independent of its helicase function (Y. Zhou et al.,

2008). The outcome is enhanced viral gene expression but with significant implications for viral RNA packaging (Y. Zhou et al., 2008). Thus DHX30 is fundamental to the coordination of critical RNA events, functions that overlap with the significance of RHA in posttranscriptional gene control. Furthermore, DHX30 is predominantly observed within the nucleus at steady-state but does demonstrate cytoplasmic localization (Y. Zhou et al., 2008). This subcellular localization pattern reflects that of RHA (Hartman et al.,

172

2006) (A.R., S.P., and K.B.L., unpublished data); however, its significance for DHX30 activity in cell biology remains to be elucidated.

DHX29 and DDX3 structurally differ from RHA (Figure 4.4) and exhibit greater cytoplasmic steady-state abundance (Parsyan et al., 2009; Schröder, 2010) yet functionally relate to RHA in their significance for targeted translation activity (Lai et al.,

2008; 2010; Pisareva et al., 2008; Shih et al., 2008). These characteristics afford possible coordinated and regulated functions with RHA.

To assess the cellular interaction of RHA with DHX30, DHX29, and/or DDX3, we performed reciprocal co-immunoprecipitations among the four DExH/D-box RNA helicases from mammalian cytoplasmic cell lysate. Shown in Figure 4.5a, immunoprecipitation of RHA resulted in the select co-precipitation of DHX30 (first

Western blot panel set). It did not enrich a detectable interaction between RHA and

DHX29 or DDX3 to indicate a distinct cellular association of RHA with DHX30 (Figure

4.5a, first Western blot panel set). Likewise, immunoprecipitation of DHX30 resulted in a select co-precipitation of RHA (Figure 4.5a, second Western blot panel set). An interaction between DHX30 and DHX29 or DDX3 was not detected (Figure 4.5a, second

Western blot panel set). It followed then that enrichment of DHX29 cellular complexes by DHX29 immunoprecipitation did not result in the identification of a DHX29-

RHA/DHX30/DDX3 association (Figure 4.5a, third Western blot panel set). Remarkably, however, immunoprecipitation of DDX3 resulted in a strong co-precipitation of RHA,

DHX30, and DHX29 (Figure 4.5a, fourth Western blot panel set). This result follows previous reports of a DDX3-RHA association in nuclear-export RNP complexes (Lai et

173 al., 2008); however, it contradicts the findings in the related reciprocal co- immunoprecipitations presented here. It is possible that the association of DDX3 with

RHA, DHX30, and DHX29 represent minor RNP complexes that become enriched for detection when DDX3 is immunoprecipitated. Notably, the data clearly indicate a robust and distinct interaction between RHA and DHX30.

Further reciprocal co-immunoprecipitations were performed to confirm the identified RHA-DHX30 cellular interaction. Here the analyses were conducted both in the absence and presence of RNase to determine the significance of RNA in mediating a

RHA-DHX30 cellular association. Shown in Figure 4.5b, immunoprecipitation of RHA demonstrated a robust co-precipitation of DHX30 both in the absence and presence of

RNase (top Western blot panel set). This result indicates a direct association between

RHA and DHX30 that does not require RNA as a molecular bridge. Reciprocal DHX30 co-immunoprecipitations confirmed these findings by showing isolated DHX30-RHA cellular complexes both in the absence and presence of RNase (Figure 4.5b, bottom

Western blot panel set). Collectively, these cellular interaction studies demonstrate a select association between RHA and DHX30 in the cell cytoplasm that indicates functional significance for regulated RHA translation activity.

Given domain-mapping studies provided critical insight into the molecular significance of RHA multimers in regulated RHA translation activity, we performed GST pull-downs to identify the domains of RHA necessary for a DHX30 association. As conducted in Figure 4.2, HEK293 cell lysate was the source of full-length, endogenous

DHX30 and indicated GST-tagged RHA domains were isolated from bacterial lysate. The

174 left Coomassie stain shows effective expression and pull-down of all GST-tagged RHA domains analyzed (Figure 4.6a). Notably, the N-terminal 300 residues of RHA (GST-

N300) and its C-terminal RGG domain were distinctly sufficient to interact with DHX30

(Figure 4.6a, right Western blot panel). These associations were abolished upon mutation of the conserved N-terminal lysine residues 54, 55, and 236 or mutation of the C-terminal arginine residues (Figure 4.6a, right Western blot panel). This interaction phenotype mimics that observed for identified RHA-CBP80 associations but is distinct from that of

RHA-RHA multimerization. Thus, these domain-mapping studies indicate a possible relationship between RHA-DHX30 association and RHA trRNP formation.

To confirm the above in vitro GST pull-down findings in cells, we performed cellular co-immunoprecipitations with FLAG-RHA and indicated FLAG-RHA mutants

(Figure 4.6b). As shown in Figure 4.6b, immunoprecipitated FLAG-RHA robustly co- precipitated DHX30 both in the absence and presence of RNase. This result recapitulates the direct association observed between endogenous RHA and DHX30 (Figure 4.6b) to support the use of FLAG-RHA in the study of RHA cellular activity. Notably, the cellular interaction of FLAG-RHA with DHX30 was significantly compromised by mutation of its N-terminal lysine residues 54, 55, and 236 (Figure 4.6b). This impaired RHA-DHX30 association correlated with compromised RHA-CBP80 interaction (Figure 4.6b). In contrast, mutation of RHA's critical ATPase residue 417 or its C-terminal arginine residues did not impact detectable FLAG-RHA-DHX30 cellular associations (Figure

4.6b). It follows that the FLAG-RHA K417R and RA9 mutants also did not exhibit impaired interactions with CBP80 (Figure 4.6b). Collectively, these cellular interaction

175 and domain-mapping studies demonstrate a select interaction between RHA and DHX30 that has potential for functional significance in regulated RHA trRNP formation and translation activity.

To achieve further insight into the molecular architecture of RHA-DHX30 complexes, we next performed GST pull-down assays to determine the domain of

DHX30 critical for associating with RHA (Figure 4.6c). Here GST-N-term RHA, harboring the first 300 amino acids of RHA, was assessed for its ability to interact with

HA-DHX30 and an associated HA-DHX30 AAVH helicase mutant. Figure 4.6c shows that effective GST-N-term RHA pull-down resulted in a detectable interaction between the N-terminus of RHA and HA-DHX30. This association was not compromised by the

AAVH helicase mutation to indicate that the catalytic core of DHX30 is dispensable for its association with RHA (Figure 4.6c). Future experiments are aimed at assessing the significance of the auxiliary domains of DHX30 in mediating its interaction with RHA.

DHX30 is a component of polysome-associated RHA trRNP complexes; implications for stimulated RHA translation activity.

The above results demonstrate a select association between RHA and DHX30 that implies DHX30 as a component of the RHA trRNP. To assess the relationship between a

RHA-DHX30 association and RHA trRNP polysome engagement, we re-examined previous polysome profiling experiments for an interaction between DHX30 and polysome-associated RHA (Figure 4.7). Shown in Figure 4.7a, a re-probe of the Western blot panels from Figure 3.4a revealed a select association between DHX30 and

176 polysome-associated FLAG-RHA (top Western blot panel set). These DHX30-RHA polysome complexes were significantly impaired upon mutation of the N-terminal lysine residues 54, 55, and 236 of RHA and followed the altered polysome distribution of

FLAG-RHA K54A/55A/236E (Figure 4.7a, middle Western blot panel set). Given the above results of a demonstrated significance for the N-terminal lysine residues of RHA in initial DHX30 associations and that mutation of amino acids 54, 55, and 236 of RHA also disrupt RHA-CBP80 interaction and RHA trRNP formation (Chapter 3), these findings indicate DHX30 as a component of active RHA trRNP complexes. Notably, mutation of

RHA's C-terminal RGG domain also impacted the association of DHX30 with polysome- associated RHA (Figure 4.7a, bottom Western blot panel set). This compromised

DHX30-RHA polysome interaction followed the altered profile distribution of FLAG-

RHA RA9 and mimicked previous results demonstrating significance for the C-terminal

RGG domain of RHA in promoting RHA trRNP complex stability (Figure 3.4 and 3.5).

Collectively, these polysome profile re-assessments demonstrate coordinated RHA and

DHX30 polysome distributions that indicate DHX30 as a component of active RHA trRNPs.

To confirm the above findings, we also re-assessed polysome-associated FLAG-

RHA complexes for DHX30 interactions in the presence of the translation inhibitors harringtonine and puromycin (Figure 4.7b and c). Harringtonine and puromycin effectively disrupt polysome engagement of trRNP complexes by blocking re-initiation or prematurely terminating elongation, respectively. The outcome in a polysome profile analysis is a shift in the sedimentation of trRNP complexes from heavy polysome

177 fractions to lighter, sub-polysome peaks. We previously showed that treatment of cells with harringtonine or puromycin resulted in a shift of identified RHA-CBP80 complexes from polysome fractions to sub-polysome peaks (Figure 2.4b and c). This result validated

RHA-CBP80 as an active trRNP. Here we re-probed these Western blot panels to determine the fate of DHX30. Shown in Figure 4.7b, treatment of cells with harringtonine resulted in a shift of identified FLAG-RHA-DHX30 complexes from heavy polysome fractions to sub-polysome peaks (compare Western blot panels to those of FLAG-RHA in

Figure 4.7a). A similar effect was observed with puromycin treatment (Figure 4.7c). This sensitivity of polysome-associated RHA-DHX30 complexes to general translation inhibition mimicked that observed of RHA-CBP80 complexes to indicate DHX30 as a component of active RHA trRNPs.

To validate the identified significance of DHX30 in coordinated RHA translation activity, we assessed the distribution of DHX30 across a 5-25% sucrose gradient. We previously showed functional significance for RHA in 43S pre-initiation ribosome complex recruitment, scanning, and 80S stabilization (Chapter 3). Separation of RHA complexes in a narrower sucrose gradient allowed for this enhanced resolution of RHA's translation initiation activities. Given the select interaction identified herein between

RHA and DHX30 and their indicated association within an active trRNP complex, we hypothesized that the distribution of DHX30 within a 5-25% initiation gradient should mimic that of RHA. To this end, HEK293 cytoplasmic cell lysate was layered upon a 5-

25% sucrose gradient and separated by ultracentrifugation. Representative fractions across the profile were analyzed for RHA-DHX30 co-sedimentation. These samples

178 corresponded to distinct functional peaks of translation initiation: 48S, 80S, and heavy polysome complexes. Shown in Figure 4.7d, RHA was detected in peaks 2, 3, and 4 of a

5-25% sucrose gradient, as previously observed (Figure 3.6b). These peaks correspond with 48S, 80S, and heavy polysome complexes, respectively. Notably, RHA exhibited greatest prominence in peak 3 to support its identified role in 80S stabilization (Figure

4.7d and Chapter 3). Remarkably, DHX30 exhibited a distribution profile identical to that of RHA (Figure 4.7d). A difference was observed when cells were lysed in the presence of 25 mM EDTA, which effectively dissociates 80S ribosomes into 40S and 60S subunit counterparts. Here RHA exhibited a prominent sedimentation with the 40S ribosomal subunit while DHX30 equally sedimented with the 60S ribosomal subunit (S.F. and

K.B.L., data not shown). Collectively, these results support a cooperative relationship between RHA and DHX30 in targeted translation activity. Notably, the preferential association of DHX30 with the large ribosomal subunit indicates molecular significance for RHA-mediated 80S stabilization.

DISCUSSION

Herein we demonstrated cellular RHA multimerization and identified a distinct interaction between RHA and its related DEVH-box RNA helicase DHX30. Both identified RHA RNP complexes were shown to be coordinated through the N- and C- terminal domains of RHA. Notably, however, the significances of each RHA domain in mediating these interactions were distinct. This afforded distinguishing compositional phenotypes for each RNP that implied differing roles in the regulation of RHA translation

179 activity. RHA multimers lacked detectable interaction with the CBP80/20 cap-binding protein complex and the cytoplasmic poly-A binding protein, translation factors that identify a functional RHA translation RNP (trRNP). This characteristic of the RHA multimer RNP indicated a role for its activity in the regulation of RHA translation.

Alternatively, RHA-DHX30 complexes demonstrated a maintained interaction with

CBP80/20 and were identified on polysomes in a manner that was sensitive to targeted disruption of the RHA trRNP or general translation inhibition. These features of the

RHA-DHX30 interaction signified importance for DHX30 as a component of the RHA trRNP that facilitates RHA translation activity. Collectively, the results obtained in this study indicate the existence of three distinct cellular RHA RNP complexes that function to regulate its translation activity: the RHA multimer RNP, the RHA trRNP, and a translationally arrested RHA trRNP. Figure 4.8 presents a model summarizing these findings.

The existence of multiple RNPs functioning to regulate the cellular activity of

RHA is further supported by recent gel filtration studies. Here RHA was identified within four discrete RNP complexes (A.S., G.S., and K.B.L., unpublished data). These analyses were conducted in the presence of HIV-1 expression and showed that a population of

RHA distinctly co-eluted with translation factors (i.e. CBP80/20 and the small and large ribosomal subunits) in a large RNP complex while a second population of RHA distinctly co-eluted with indicators of HIV-1 viral particle assembly (i.e. the viral structural protein

Gag) (A.S., G.S., and K.B.L., unpublished data). These findings indicated the engagement of RHA within two discrete RNP complexes that exhibit inherent functional

180 differences: a RHA trRNP focused on facilitating target transcript expression and an

RHA assembly RNP centered upon viral RNA packaging and virion assembly. The two additional RHA RNP complexes further identified in this study were characterized to represent non-translating RHA trRNPs and non-productive RHA RNPs, respectively, based upon their complex size and co-eluting factors (A.S. and K.B.L., unpublished data). These findings corroborate the data presented in this study and indicate possible significance for RHA multimers in the identification and function of the fourth, non- productive RHA RNP complex.

An open issue presented by the model proposed in Figure 4.8 is the molecular events coordinating the directed engagement of RHA in each RNP complex. RNP complexes are known dynamic subcellular structures; thus, it is most likely that the engagement of RHA in each RNP complex is transitional and based upon the cellular environment. A possible molecular basis for directed RHA RNP engagement is the posttranslational modification of RHA. RHA harbors several predicted residues for posttranslational modification events, particularly within its N-terminus (R.L. and K.B.L., unpublished data). These include phosphorylation, ubiquitinylation, and sumolyation

(R.L. and K.B.L., unpublished data). It is well known that posttranslational modification is a fundamental effector of protein subcellular localization, cofactor interaction, and

RNA binding, all features that regulate inherent cellular activity. Indeed it has been shown that methylation of RHA's C-terminus is a required posttranslational modification event for its nuclear trafficking (Smith et al., 2004). In the nucleus, RHA serves as a fundamental regulator of RNA polymerase II transcriptional activity (S. F. Anderson,

181

Schlegel, Nakajima, Wolpin, & Parvin, 1998; Aratani et al., 2001; Nakajima et al., 1997;

K. Zhou et al., 2003). It is thus conceivable that C-terminal methylation of RHA redirects its engagement from a trRNP to one, possibly the multimer RNP, which facilitates nuclear import.

In addition to methylation, RHA is phosphorylated in a manner that regulates its cellular activity. In the case of a human-derived leukemia cell line, hyperphosphorylated

RHA was constitutively identified (Zhong & Safa, 2007). This posttranslational modification event occurred due to the stable association of RHA with the DNA- dependent protein kinase (Zhong & Safa, 2007). The outcome was enhanced functional activity of RHA at the multi-drug resistant promoter, which resulted in an increased expression of the drug efflux transporter p-glycoprotein and the multi-drug resistant phenotype of these cancer cells (Zhong & Safa, 2007). Activated phosphorylation of

RHA was also demonstrated to occur with significant functional consequence on RHA transcriptional activity. Here stimulation by polyinosinic-polycytidylic acid (pIC) resulted in the activation of protein kinase R and targeted phosphorylation of RHA within its first 300 residues (Sadler, Latchoumanin, Hawkes, Mak, & Williams, 2009). This posttranslational modification event resulted in reduced binding of RHA to double- stranded RNA and subsequent impairment of RHA-mediated transcriptional activation

(Sadler et al., 2009). The findings presented within this study together with the data presented above indicate that RHA transcriptional activity may be intricately coordinated with its RNP identity. Thus, it is conceivable to envision that phosphorylation of RHA

182 regulates its RNP engagement so as to promote or impair nuclear localization and transcriptional activities depending upon the cellular state and effector kinase.

In addition to the general mechanism coordinating RHA RNP engagement, several outstanding questions remain in regards to the distinct properties of the RHA multimer RNP and RHA-DHX30 assemblies that confer their effects on RHA translation activity. In regards to RHA multimer RNPs, their preferential subcellular localization remains to be determined. If predominantly nuclear, this characteristic affords a molecular basis for distinguished RHA transcriptional activity from that of cytoplasmic translation functions. Alternatively, if predominantly cytoplasmic, this feature would propose a direct competition model between RHA multimer RNPs and RHA trRNPs in the regulation of RHA translation activity.

A second characteristic of the RHA multimer RNP that remains to be defined is the significance of the cellular state for its formation and stability. In other words, are

RHA multimer RNPs a constitutive phenomenon or is their formation induced upon cell activation or stress? We demonstrated that the association of RHA with the non-canonical

CBP80/20 cap-binding protein complex was a constitutive event that informed a role for

RHA in maintained cap-dependent translation during cell stress. It is conceivable that

RHA multimer RNPs exhibit a similar characteristic to regulate corresponding RHA translation activity. In support of this model, the tandem immunoprecipitation data presented herein was achieved from studies conducted at steady state and showed robust

RHA multimer formation.

183

A third open issue about the features of the RHA multimer RNP is the number of

RHA molecules that self-associate to create this complex. Preliminary in vitro studies indicate that several (n = 2-6) RHA N-terminal molecules can self-associate (X.H., I.B., and K.B.L., unpublished data). Notably, this number is influenced by the bound RNA structure (X.H., I.B., and K.B.L., unpublished data). This feature indicates that the composition of the RHA multimer RNP can be diverse and dependent upon the RNA engaged within. Although these studies have yet to be performed in cells, it is conceivable that cellular RHA multimer RNPs range in number from dimeric to hexameric complexes with each exhibiting a distinct functional outcome on RHA translation activity.

In regards to characteristics of the identified RHA-DHX30 association, the domains of DHX30 necessary for mediating this binding event remain to be defined. The similarities in structure between RHA and DHX30 imply significance for DHX30's N- and C-terminal domains in coordinating an association with RHA. This model follows the demonstrated significance for RHA's N-terminal dsRBDs and C-terminal RGG domain in mediating its self-association. However, the absence of a specified RGG domain in the C-terminus of DHX30 signifies an altered model. Here it is conceivable that DHX30's N-terminal dsRBD(s) solely drive association with RHA. This binding event could occur through direct interactions with RHA's N-terminal dsRBDs or its C- terminal RGG domain. It is known that dsRBDs are strong effectors of dimerization events and that RHA oligomerizes in vitro via N-N or N-C terminal interactions (W.J. and K.B.L., unpublished data). The association of DHX30's N-terminus with either

184

RHA's N- or C-terminal domains would afford a molecular basis for stabilization of the

RHA trRNP that is instrumental to its role in 43S pre-initiation ribosome complex recruitment, scanning, and 80S formation.

A second open issue about the features of the RHA-DHX30 association is the timing of their engagement within the biogenesis of the RHA trRNP. Both RHA and

DHX30 exhibit a prominent nuclear localization at steady state (Hartman et al., 2006; Y.

Zhou et al., 2008). Yet each is also detected within in the cytoplasm to suggest an association that could nucleate in either subcellular compartment. It is known that activation of a RHA-cognate transcript for translation requires an initial nuclear localization event (Dangel, Hull, Roberts, & Boris-Lawrie, 2002). This feature could be significant for engaging either an RHA or a RHA-DHX30 association. If the association of DHX30 with a destined RHA trRNP occurs in the nucleus, this characteristic would propose significance for DHX30 in priming an RNA trRNP complex for translation activity upon cytoplasmic localization. Alternatively, if the association of DHX30 with a destined RHA trRNP occurs upon its cytoplasmic entry, this feature would indicate significance for DHX30 in directly activating the RHA trRNP for translation. Future studies aimed at further defining the characteristics of the RHA multimer RNP and the

RHA-DHX30 associations that determine their distinct roles in RHA-mediated translation will be fundamental to elucidating the molecular basis of regulated RHA translation activity. This knowledge will contribute to our fundamental understanding of eukaryotic translation control and in particular the significance of RNP complexes in coordinating distinct translation outcomes.

185

Figure 4.1 RHA multimerizes in cells. (a) Schematic outlining RHA tandem immunoprecipitation protocol. (b) FLAG-RHA was immunoprecipitated from HEK293 cell lysate with anti-FLAG-conjugated agarose beads (Sigma). Isolated FLAG-RHA complexes were then eluted by incubation with competing FLAG peptide. Eluent was subjected to HA-RHA immunoprecipitation with an-HA-conjugated agarose beads (Sigma) and then probed for the co-precipitation of FLAG-RHA by Western blot. Shown is representative Western blot images. Input, 2% of cell lysate. IP, immunoprecipitated complex. FT, 2% of flow-through following IP. (c) RHA tandem immunoprecipitations as in (b), with RNase treatment of cell lysate prior to IP.

186

Figure 4.2 RHA multimers are coordinated by N- and C-terminal interactions. (a) HEK293 cell lysate was incubated with bacterial lysate containing indicated recombinant RHA domains or GST and then subjected to GST pull-down. Left panels, Coomassie stain and Western blot showing effective GST pull-down and detection of cellular RHA, respectively. Right panels, representative GST pull-down results as assessed by Western blot analysis. Experiment was conducted both in the absence and presence of RNase, as indicated. (b) GST pull-down assays as in (a), with the addition of indicated mutant recombinant RHA domains. (c) FLAG-RHA or indicated mutants were immunoprecipitated from HEK293 cell lysate with anti-FLAG-conjugated agarose beads (Sigma) and then probed for the co-precipitation of YFP-RHA by Western blot. Shown is representative Western blot images. Input, 2% of cell lysate. IP, immunoprecipitated complex. FT, 2% of flow-through following IP.

Continued

187

Figure 4.2 continued

188

Figure 4.3 RHA multimers compete with translation cofactor binding. Right panels, RHA tandem immunoprecipitations from Figure 4.1b were re-probed for indicated translation cofactor association. Left panels, FLAG-RHA was immunoprecipitated from HEK293 cell lysate with anti-FLAG-conjugated agarose beads (Sigma) and then probed for the co-precipitation of indicated proteins by Western blot. Shown is representative Western blot images. Input, 2% of cell lysate. IP, immunoprecipitated complex. FT, 2% of flow-through following IP.

189

Figure 4.4 Depiction of known DExH/D-box RNA helicase domain structures with critical roles in targeted translation control. Figure was adapted from (Ranji & Boris- Lawrie, 2010). dsRBD, double-stranded RNA binding domain. HA2, helicase-associated 2. DUF, domain of unknown function. RGG, arginine-glycine (RGG)-rich. WH, winged- helix. OB, oligonucleotide/oligosaccharide-binding. RS/GYR, arginine-serine/glycine- tyrosine-rich.

190

Figure 4.5 RHA and DHX30 exhibit a select association in cells. (a) Endogenous RHA, DHX30, DHX29 or DDX3 was immunoprecipitated from HEK293 cytoplasmic cell lysate with anti-RHA (Bethyl), anti-DHX30 (Bethyl), anti-DHX29 (Bethyl), anti- DDX3 (Bethyl), or isotype control IgG antibody and probed for the coprecipitation of related DExH/D-box RNA helicase. Shown is representative Western blot images. Input, 2% of cell lysate. (b) Co-immunoprecipitations as in (a), with with RNase treatment of cell lysate prior to IP.

191

Figure 4.6 N-terminus of RHA is necessary and sufficient for a DHX30 association; helicase activity of DHX30 is dispensable for this interaction. (a) HEK293 cell lysate was incubated with bacterial lysate containing indicated recombinant RHA domains or GST and then subjected to GST pull-down. Left panels, Coomassie stain and Western blot showing effective GST pull-down and detection of cellular DHX30, respectively. Right panels, representative GST pull-down results as assessed by Western blot analysis. Experiment was conducted both in the absence and presence of RNase, as indicated. (b) FLAG-RHA or indicated mutants were immunoprecipitated from HEK293 cell lysate with anti-FLAG-conjugated agarose beads (Sigma) and then probed for the co- precipitation of indicated proteins by Western blot. Shown is representative Western blot images. Input, 2% of cell lysate. Experiment was conducted both in the absence and presence of RNase, as indicated. (c) HEK293 cell lysate harboring exogenously expressed HA-DHX30 or the HA-DHX30 AAVH helicase mutant were incubated with bacterial lysate containing GST-N-term RHA and then subjected to GST pull-down. Input, 2% of cell lysate. PD, GST pull-down complex. FT, 2% of flow-through following pull-down.

192

Figure 4.7 DHX30 is identified in polysome-associated RHA trRNP complexes. (a) Polysome profiles from Figure 3.4a, re-probed for DHX30. (b) Polysome profile from Figure 2.4b, reprobed for DHX30. (c) Polysome profile from Figure 2.4c, reprobed for DHX30. (d) Analysis of endogenous RHA or DHX30 distribution across a 5-25% sucrose gradient. Left Western blot panels, steady-state expression of RHA and DHX30 in HEK293 cytoplasmic cell lysate prior to separation through the sucrose gradient. Right Western blot panels, RHA and DHX30 co-sedimentation in indicated peak fractions. Shown is representative Western blot images.

193

Figure 4.8 Model for regulated RHA translation activity. Three distinct RHA ribonucleoprotein (RNP) complexes exist with significant implications for regulated RHA translation activity. RHA RNP complex 1 identifies RHA molecules engaged in intermolecular self-associations. This RHA RNP complex forms due to affinities between available N- and C-terminal domains of adjacent RHA molecules. The outcome is impaired translation activity due to reduced affinity of RHA RNP complex 1 for translation factors (i.e. CBP80/20 (CBC). In contrast, RHA RNP complex 2 identifies RHA molecules actively associated with CBP80/20 (CBC) and DHX30 in a manner that facilitates engaged translation activity. Mutation of RHA disrupts RNP complex 2 formations to result in a third RHA RNP complex that exhibits translational stalling. The spatiotemporal relationship of these three RHA RNP complexes and its significance for regulated RHA translation activity remains to be determined.

194

Chapter 5: Perspectives

The aim of the dissertation research presented herein was to elucidate the molecular basis of the DExH-box RNA helicase RNA helicase A (RHA/DHX9) in eukaryotic protein synthesis. This objective was accomplished in three main ways: (1) by characterizing the RHA translation ribonucleoprotein complex, (2) by identifying the molecular significance of RHA in the translation process, and (3) by determining the cellular features that regulate RHA translation activity. The outcome has been the identification of novel significance for RHA in eukaryotic protein synthesis.

In Chapter 2, RHA was shown to associate with the non-canonical CBP80/20 cap- binding protein complex. This molecular interaction distinguished the RHA translation ribonucleoprotein complex and, notably, informed a role for RHA in maintained cap- dependent protein synthesis during cell stress. In Chapter 3, RHA was identified to exhibit molecular significance in both initiation and post-initiation translation activities.

These functional effects required RHA-mediated ribosome recruitment and stabilization, activities that extend beyond the basic ATP-dependent mRNA unwinding or RNP remodeling functions traditionally associated with DExH/D-box RNA helicases. In

Chapter 4, RHA was recognized to exist in multiple ribonucleoprotein complexes, two of which were defined by RHA multimerization and a select RHA-DHX30 association. This finding indicated a complex cellular network regulating the activity of RHA in translation

195 control. Collectively, these data afforded a molecular basis for RHA in eukaryotic protein synthesis.

The final chapter of this dissertation serves to discuss the significance of these findings by addressing their implications for our fundamental understanding of the eukaryotic translation process. The significance of this work will also be explored in regards to the role of protein synthesis and RHA in animal and human health and disease.

The outcome of this final chapter is to inform the work of future studies focused on further elucidating the molecular significance and mechanism of RHA in gene expression control.

Cap-dependent translation during cell stress

Eukaryotic protein synthesis is defined by a cap-dependent scanning mechanism whereby the ribosome binds the 5' terminus of an mRNA and proceeds along the transcript inspecting, base-by-base, for an appropriate start codon to initiate polypeptide synthesis (Jackson et al., 2010). In the advent of cell stress, it is well accepted that the canonical cap-dependent scanning mechanism of eukaryotic protein synthesis is compromised (Holcik & Sonenberg, 2005; B. Liu & Qian, 2014; S. Yamasaki &

Anderson, 2008). This is due to the known functional inactivation of the cap-binding protein eIF4E, which is the fundamental orchestrator of initial scanning events (Ma &

Blenis, 2009; Pause et al., 1994). As a result, cell stress is characterized by significant downregulation of protein synthesis.

196

It is also known, however, that targeted translation is maintained during cell stress. This is critical for the expression of stress response proteins that serve in cell survival and recovery (Holcik & Sonenberg, 2005; B. Liu & Qian, 2014; S. Yamasaki &

Anderson, 2008). It is also a fundamental feature that facilitates pathogenic viral replication upon infection (Firth & Brierley, 2012). The molecular basis for translation during cell stress is attributed to an alterative mechanism of ribosome recruitment whereby 5' RNA elements directly attract initiating 43S ribosome complexes to the start codon (Spriggs et al., 2008). This internal ribosome entry site (IRES)-mediated initiation of protein synthesis bypasses the need of cap-associated factors to engage protein synthesis (Jackson et al., 2010; Spriggs et al., 2008). Consequently, it circumvents the block on eIF4E translation activity. Thus mRNAs, both cellular and viral, that harbor

IRES elements maintain their translation during cell stress.

A major finding of this dissertation is a paradigm-shifting concept that cap- dependent translation is maintained during cell stress. This is due to the alternative

CBP80/20 cap-binding protein complex (CBC) and RNA helicase A (RHA). Together

CBC and RHA form a distinct translation ribonucleoprotein (trRNP) capable of cap- dependent translation during cell stress (Chapter 2). Unlike eIF4E, CBC is immune to the major cell signaling pathways that affect eIF4E activity (Maquat et al., 2010).

Consequently, CBC is able to maintain its translation function during cell stress (Oh,

Kim, Cho, Choe, & Kim, 2007a; A. Sharma et al., 2012). Furthermore, it is known that

CBC functionally interacts with the ribosome in a similar manner to eIF4E (Chiu et al.,

2004; Maquat et al., 2010). Thus, the defining mechanisms of eukaryotic cap-dependent

197 translation are preserved when coordinated by CBC. Collectively these features rationalize a role for CBC in maintained cap-dependent translation during cell stress.

Our identification herein of a select association between CBC and RHA offers a molecular basis for targeted cap-dependent translation during cell stress. We specifically demonstrated significance for CBC and RHA in maintained polysome association of

HIV-1 gag mRNA during HIV-1 expression (Chapter 2). A hallmark feature of HIV-1 expression is the activation of 4E-BP1(A. Sharma et al., 2012). Activated 4E-BP1 binds with high affinity to eIF4E, resulting in the suppression of its translation activity (Ma &

Blenis, 2009; Pause et al., 1994). However, the engagement of the HIV-1 gag transcript within a RHA-CBC trRNP allows it to circumvent this block and facilitate its maintained translation activity in a cap-dependent manner (Chapter 2).

An open issue presented by our proposed model is the relationship between maintained cap-dependent translation during cell stress and IRES-mediated initiation.

Although IRES elements have been well characterized in the regulation of viral protein synthesis, their understanding in the control of host translation remains ambiguous (Firth

& Brierley, 2012; Jackson et al., 2010; Thompson, 2012). This is, in part, due to the lack of sequence or structure conservation across known cellular IRES elements (Firth &

Brierley, 2012; Jackson et al., 2010; Thompson, 2012). Thus, identification of IRES activity is limited to experimental bicistronic reporter assays (Thompson, 2012). This affords opportunities for false positive results depending upon the extent of quality control measures taken. Such is seen in the study of HIV-1 translation control. Numerous studies have reported expression of the major HIV-1 structural protein Gag to be driven

198 by IRES-mediated translation. An equal number have countered with data to support cap- dependent protein synthesis. Is one correct?

A recent perspective was presented exploring a possible synergistic relationship between IRES- and cap-mediated translation (Gilbert, 2010). Here an IRES element was viewed as a translational enhancer that regulates cap-associated translation activities

(Gilbert, 2010). This model was presented based upon the known inefficiency of IRES- mediated translation compared cap-driven protein synthesis. In fact, most cellular IRES activity is only identified when canonical eIF4E-driven protein synthesis is impaired.

Furthermore, this model was based upon the observations that IRES elements mediate their translational effects via the select RNA binding proteins that associate with them. It also argues a role for the 5' terminus in the active regulation of the translation process rather than serving as a passive bystander during initial ribosome scanning events. So is

RHA-CBC-driven translation during cell stress a manifestation of this model in which

CBC affords the cap-dependent translation activity and RHA-PCE is the 'IRES' feature that directs the targeted maintenance of CBC activity on a transcript?

It is likely that all scenarios are possible and actively engaged by the cell in a spatiotemporal manner that depends upon the exact mode of cell stress. In instances of targeted downregulation of eIF4E activity, such as those examined in our study presented herein, CBC-driven translation, and in particular RHA-CBC-driven translation, may afford the primary mechanism for targeted maintenance of protein synthesis. However, in situations when the stress event targets other factors of translation control, like eIF2, that are required of both eIF4E and CBC for cap-mediated protein synthesis IRES translation

199 may predominate. This situation might also capitalize on mechanisms whereby 'IRES' and cap-mediated translation synergize to facilitate targeted protein synthesis.

Future studies aimed at further dissecting the relationship between maintained cap-dependent protein synthesis and/or IRES-mediated translation during cell stress are interesting to consider, especially in regards to their significance for protein synthesis in disease. A hallmark feature of cancer, pathogenic viral infection, neurodegeneration, and innate immunity is deregulated translation activity (Carpenter et al., 2014; Ruggero,

2013; Scheper et al., 2007; Silvera et al., 2010). Hence significant research is invested towards understanding the fundamental principles of protein synthesis in both normal and disease states. The outcome has been the identification of several therapeutic inhibitors, many of which have shown promise in the clinic (Bhat et al., 2015). However, the complexity of translational reprogramming that occurs with the transition of a cell from a healthy to a disease state has hindered the overall success for many of these therapeutics.

Our identification of a novel role for RHA and CBC in maintained cap-dependent translation during cell stress demonstrates an example of this complexity that challenges effective therapeutic outcomes. In particular, our findings are significant given the focus towards eIF4E trRNP inhibitors (Bhat et al., 2015). Instead our data warrants consideration for RHA and CBC in the development of therapeutics that aim to reprogram cellular translation in favor of disease resolution.

In order to therapeutically target the RHA-CBC trRNP, several fundamental questions need to be addressed. First is the characterization of the RHA-CBC translatome. In this study, a candidate approach showed significance for RHA and CBC

200 in the regulated expression of HIV-1 during cell stress. However, our data in addition to others indicate significance for RHA and CBC in the translation control of critical stress response proteins. In regards to RHA, this includes huR and p53 (“Translational Control

Protein 80 Stimulates IRES-Mediated Translation of p53 mRNA in Response to DNA

Damage,” n.d.) (S.F., D.S., and K.B.L., data not shown). Polysome RIP-seq is a robust tool that would allow for the characterization of the RHA-CBC translatome. Here polysome-associated RHA-CBC trRNP complexes would be isolated by immunoprecipitation and then subjected to high-throughput sequencing. When conducted under both normal and cellular stress states (e.g. retroviral infection) an identification of those transcripts maintained on polysomes due to their association with RHA and CBC would be achieved. The outcome from this analysis would be the identification of transcripts that exhibit a strong disease association and are resistant to current eIF4E inhibitors. In addition, critical insight would be achieved into likely candidates whose expression is linked to pro-tumorigenic and/or anti-viral activity during disease stress.

A second outstanding question that needs to be addressed is the significance of

RHA-CBC translation activity in acute versus long-term stress. In this study, we showed significance for RHA and CBC in maintained translation activity following short-term

(forty-eight hours or less) stress events. It would be of importance to assess those transcripts bound by RHA and CBC and that are expressed following several days of serum deprivation, HIV-1 infection, or in a tumor sample. A similar approach was instrumental in the identification of the fragile-X mental retardation translatome and its significance for translation control in Fragile X syndrome and related autistic disorders

201

(Darnell et al., 2011). Together, the outcomes from these future directions would provide greater insight into the significance of RHA and CBC in eukaryotic translation control and, notably, their role in disease. This information will be critical for directing the next- generation of translation inhibitors that can serve as effective anticancer or anti-viral therapeutics.

DExH/D-box RNA helicases with significance in post-initiation translation control

The DExH/D-box RNA helicase family harbors many known effectors of eukaryotic protein synthesis. These include eIF4A (DDX2), eIF4AIII (DDX48), DHX29, and DDX3 in addition to RHA (DHX9). The well-accepted model for their functional significance in the translation process is ATP-dependent mRNA unwinding. This allows for the resolution of 5' RNA secondary structures that serve as barriers to efficient ribosome scanning. More recent studies, however, have indicated significance for

DExH/D-box RNA helicases in the translation process beyond this implied canonical mRNA unwinding activity. These critical functions include trRNP remodeling that facilitates 43S ribosome complex engagement and 80S ribosome formation. In all cases, however, DExH/D-box RNA helicases are implied to strictly function within translation initiation events. This model is further supported by experimental evidence showing that

DExH/D-box RNA helicases (eIF4A, eIF4AIII, DHX29, and DDX3) are restricted to the early initiating peaks within a polysome profile (Lai et al., 2008; Parsyan et al., 2009).

This dissertation presents RHA as a DExH-box RNA helicase with significance in post-initiation translation control. A molecular importance for RHA beyond canonical

202 initiation was initially supported by the observation of its distribution within the polysome fractions of a sucrose gradient, a sedimentation effect that was sensitive to translation inhibition by puromycin (Hartman et al., 2006). This polysome profile phenotype distinguished RHA from related DExH/D-box RNA helicases, which are often restricted to the early initiating fractions of a sucrose gradient (Lai et al., 2008; Parsyan et al., 2009). The data presented in this dissertation corroborated these initial findings and furthered our understanding of the significance for RHA in post-initiation translation control. Herein we demonstrated a role for RHA's C-terminal RGG domain in maintaining 80S ribosome stabilization that promotes early elongation events (Chapter

3). Mutation of RHA's C-terminal domain induced a phenomenon known as ribosomal drop-off in prokaryotes (Chapter 3).

An open issue presented by our model is the significance of RHA's post-initiation role for our understanding of the molecular basis of eukaryotic translation control. It is well known that RNA binding proteins can moderate the act and efficiency of polypeptide synthesis. Often this activity is directed to particular transcripts by virtue of cis-acting RNA elements that specifically recruit a cognate RNA binding protein effector.

In the case of post-initiation translation control, an RNA binding protein effector is directly recruited to the coding sequence or elongating ribosome where it mediates its effect on polypeptide synthesis (Darnell et al., 2011; Friend et al., 2012; Hussey et al.,

2011; Shalgi et al., 2013). Is RHA an example of this well-known translation control phenomenon?

203

RHA's cognate cis-acting RNA element, the posttranscriptional control element

(PCE), was initially identified as a 5' feature of retroviral transcripts and the cellular proto-oncogene junD (Bolinger et al., 2007; 2010; Butsch et al., 1999; Hartman et al.,

2006; Hull & Boris-Lawrie, 2002). Notably, in this position, the PCE served as a critical regulator of translation efficiency (Bolinger et al., 2007; 2010; Butsch et al., 1999;

Dangel et al., 2002; Hartman et al., 2006; Roberts & Boris-Lawrie, 2000). Given this data and the identification of RHA as the host effector and binding partner of the PCE, a model was proposed that featured RHA as a 5'-binding regulator of protein synthesis

(Hartman et al., 2006).

Yet our recent application of PAR-CLIP to the study of RHA biology identified additional RHA binding sites in the 3' terminus of the PCE-containing HIV-1 gag transcript (K.B.L. in collaboration with S.K. and P.B., unpublished data). PAR-CLIP, photoactivatable-ribonucleoside-enhanced crosslinking and immunoprecipitation, is a molecular technique that extends the basic identification of RNA-protein interactions to the mapping of these binding events along a target transcript. With this technique, we found RHA to bind elements of the PCE within the 5' terminus of HIV-1 gag (K.B.L. in collaboration with S.K. and P.B., unpublished data). However, we also identified binding interactions between RHA and the 3' rev-response element (RRE) (K.B.L. in collaboration with S.K. and P.B., unpublished data). The RRE is a critical cis-acting posttranscriptional regulator of HIV-1 gene expression. A dual association of RHA with a target transcript affords a molecular basis for its identified roles in both initiation and post-initiation translation control.

204

Future studies focused on the interaction of RHA with the ribosome will provide further molecular insight into how this DExH-box RNA helicase exhibits roles in both initiation and post-initiation translation control. A cryo-electron microscopy (cryo-EM) structure of RHA in complex with the 43S ribosome will be instrumental to resolving the interactions of RHA that facilitate ribosome recruitment, scanning, and stabilization.

Such was the case for the related DExH-box RNA helicase DHX29. Here the cryo-EM structure showed DHX29 positioned near the mRNA entry-channel latch of the small

(40S) ribosomal subunit (Hashem et al., 2013). The contact residues identified indicated a role for DHX29 in the remodeling of the entry channel latch that would promote the scanning process (Hashem et al., 2013). These findings supported previous biochemical data that suggested significance for DHX29 in initial mRNP remodeling events (Pisareva et al., 2008). It can be envisioned that a similar study with RHA would provide critical insight into its mechanistic significance for translation initiation, particularly the role of its N-terminal dsRBDs in facilitating this process.

A similar cryo-EM study conducted with the fragile X mental retardation protein

(FMRP) and the 80S ribosome was instrumental in identifying the significance for this

RNA binding protein in post-initiation translation control (Chen et al., 2014). Here the identified interaction sites showed FMRP binding two large (60S) ribosomal subunits, L5 and L18, which are important for peptidyl-tRNA stabilization (Chen et al., 2014). This finding provided a molecular basis for the previously identified significance of FMRP in elongation stalling and, notably, afforded a model by which FMRP competes with peptidyl-site tRNA binding to result in compromised elongation and impaired protein

205 production (Chen et al., 2014; Darnell et al., 2011). Preliminary molecular data indicates a strong interaction between RHA and the L5 subunit (S.F. and K.B.L., data not shown).

It is interesting to consider the possibility that RHA is also important for peptidyl-tRNA stablization. A future cryo-EM study of RHA in complex with the 80S ribosome would provide critical mechanistic insight into its significance for post-initiation translation control, particularly the role of its C-terminal RGG domain in facilitating 80S stabilization.

Protein synthesis is fundamental to the survival of all cells and viruses

Proteins are essential macromolecules necessary for cell structure and function; they are critical to virion formation, viral propagation and infection. Deregulated protein synthesis, structure, and/or function are at the core of animal and human neurological and neurodegenerative diseases and disorders, autoimmunity, and cancer. It is also significant for plant sustainability. Thus, a fundamental understanding of translation is critical to elucidating the molecular basis of plant, animal, and human diseases. This knowledge will allow for more effective and targeted therapies to be developed and applied.

This dissertation focused on further elucidating our understanding of one mechanism of translation control. The direct implications are significant given the targets of RHA-regulated protein synthesis. Gag is the major structural protein of simple and complex retroviruses. Compromise of its production significantly impairs fundamental processes that are necessary for continued propagation and infectivity. RHA is the critical regulator of Gag protein synthesis. The Boris-Lawrie laboratory has shown that this

206 mechanism is conserved across the Retroviridae family. This includes HIV-1 as well as the human leukemic retrovirus HTLV-1 and the pathogenic animal retroviruses spleen necrosis virus (SNV), feline leukemia virus (FeLV), bovine leukemia virus (BLV), and

Mason Pfizer monkey virus (MPMV) (Bolinger et al., 2007; 2010; Butsch et al., 1999;

Hartman et al., 2006; Hull & Boris-Lawrie, 2002). Notably, the laboratory has shown that a reduction in RHA expression causes impaired Gag production (Bolinger et al., 2007;

2010; Hartman et al., 2006). In the case of HIV-1, this results in impaired infectivity in a cell culture system (Bolinger et al., 2010). Given RHA is a host-encoded protein that is ubiquitously expressed across tissue types, an understanding of its molecular functioning is critical for developing targeted anti-retroviral therapies against the main structural and functional protein Gag.

RHA is also the critical regulator of cellular junD and huR protein synthesis

(Hartman et al., 2006) (D.S., S.F., and K.B.L., unpublished data). junD is a member of the AP-1 transcription factor family whose activity is intricately linked with several cancer-associated cellular events, including tumor angiogenesis, cell proliferation, and apoptosis (Hernandez et al., 2008). Human antigen R (huR) is an RNA binding protein with critical roles in mRNA stability and translation, which are also highly significant to cancer development and progression (Z. Yuan, Sanders, Ye, & Jiang, 2010). The significance of RHA for regulated expression of junD and huR implies a functional role for RHA in cancer as well. Thus, an understanding of its molecular basis in protein synthesis is critical for developing novel anti-cancer therapies that target the expression

207 of key protein regulators in the cellular processes underlying tumor development, growth, and metastasis.

The implications of this dissertation for our fundamental understanding of translation control extend beyond the immediate significance of RHA in regulated protein synthesis. DExH/D-box RNA helicases are conserved throughout the eukaryotic and prokaryotic kingdoms. In all studied cases thus far, they are fundamental effectors of protein synthesis via direct or indirect roles in the translation process. Given this significance, extensive research is still ongoing to further understand the molecular basis of DExH/D-box RNA helicases in posttranscriptional gene control. Notably, we demonstrated herein that although many DExH/D-box RNA helicases exhibit similar structure, function, and/or cellular characteristics their molecular acts in the translation process are distinct. Furthermore, the cofactors they engage to mediate this effect are diverse. This knowledge is significant for future studies aimed at targeting DExH/D-box

RNA helicase function in disease therapies.

RNA helicase A is a host protein with clinical significance

In addition to functioning as a major stimulator of pathogenic viruses that infect and cause disease within animal and human hosts, RHA is deregulated in breast, prostate, and lung cancer and is the major auto-antigen identified in systemic lupus erythematosus patients (Guénard et al., 2009; C. G. Lee et al., 1999; Takeda et al., 1999; Vázquez-Del

Mercado et al., 2010; Wei et al., 2004; Y. Yamasaki et al., 2007). In a familial study of breast cancer, RHA sequence variants were identified as high-risk predictors for the

208 development of non-BRCA1/BRCA2 breast cancer (Guénard et al., 2009). These mutations lie within the protein coding sequence of RHA and are found within the linker regions between its N-terminal double-stranded RNA binding domains (dsRBDs), helicase core, and C-terminal helicase-associated 2 (HA2) domain (Guénard et al., 2009).

Two non-synonymous sequence variants, one within the linker between dsRBD1 and dsRBD2 and the second within the helicase core, were predicted to significantly impair

RHA's structure and function (Guénard et al., 2009). Likewise, recent analysis of RHA mutations in the Cancer Genome Atlas showed several identified missense mutations within RHA's dsRBD linker region and helicase core across numerous cancer studies

(G.A. and K.B.L., unpublished data). These findings signify importance for RHA and, notably, its conserved domains in appropriate cell growth and activity; its mutation is highly correlated with cell transformation and cancer.

Although the specific mutations identified in the cancer studies above were not directly tested herein, the findings of this dissertation indicate significance for RHA's conserved domains in appropriate cellular functioning. Notably, the results show that single amino acid substitutions can dramatically impair the interaction of RHA with critical translation cofactors. The outcome is altered RHA cellular activity with significant implications for regulated protein expression and cell homeostasis. Such was seen in the mutation of lysine residue 236 within the second dsRBD of RHA (Chapter 3).

Here its substitution to an alanine resulted in impaired CBP80 binding (Chapter 3). The outcome was ineffective engagement of RHA within the translation process (Chapter 3).

209

Future studies aimed at assessing the effect of RHA mutations on overall cell biology

(growth, division, morphogenesis, migration, etc) are interesting to consider.

The significance of RHA as a serological marker for systemic lupus erythematosus (SLE) is unknown. Yet three independent studies identified autoantibodies to RHA as hallmark characteristics of patients with SLE (Takeda et al., 1999; Vázquez-

Del Mercado et al., 2010; Y. Yamasaki et al., 2007). Mechanistically, apoptosis-induced proteolytic cleavage of RHA was proposed as the molecular basis that generates antigenic

RHA peptides (Takeda et al., 1999). However the outstanding question still remains, why

RHA?

Besides being an abundant, ubiquitously expressed protein, the large, multi- domain architecture of RHA and its significance for diverse cellular processes most likely contributes to its identity as an autoantigen in disease states. One model to consider is a possible significance for RHA in antigenic peptide production. In this dissertation, a novel relationship was identified between RHA-mediated translation control and

CBP80/20-directed protein synthesis (Chapter 2). In addition to serving critical mRNA surveillance functions, CBP80/20 is significant for the production of MHC class I antigenic peptides (Apcher et al., 2011). Given the select interaction identified herein between RHA and CBP80/20, an interesting hypothesis to consider is a role for RHA in antigenic peptide production. Our recent RIP-seq, RIP-CHIP, and PAR-CLIP studies show that the RHA transcript is a target for RHA protein binding and translation regulation (S.F., A.Y., M.H., and K.B.L. in collaboration with S.K. and P.B., data not shown). A possibility is that the RHA-CBP80/20 trRNP, identified in this dissertation to

210 exhibit maintained functionality during cell stress, critically regulates RHA protein production. In stress conditions, the maintained activation of RHA-CBP80/20 facilitates robust RHA protein production. This results in the generation of either full-length or truncated RHA protein fragments, which become abundant sources for antigenic peptides. When disproportionately selected or acted upon, autoimmunity to RHA develops.

A role for RHA in the innate immune response has been indicated based upon several findings. These include: significance for RHA in sensing viral dsRNA in myeloid dendritic cells, an identified interaction between RHA and the toll-like receptor protein

MyD88 in response to microbial DNA, and the localization of a related RHA molecule, leukophysin, in cytotoxic T cell granules (Abdelhaleem, Hameed, Klassen, & Greenberg,

1996; T. Kim et al., 2010; Z. Zhang, Yuan, Lu, Facchinetti, & Liu, 2011). Notably, the interaction of RHA with MyD88 required its C-terminal domain of unknown function

(DUF) (T. Kim et al., 2010). Likewise, leukophysin is characterized as the C-terminus of

RHA, including its RGG domain (Abdelhaleem et al., 1996). In this dissertation, significance for RHA's C-terminal RGG domain in 80S ribosome stabilization was identified. Whether this molecular role is conserved in directing a stable association with

MyD88 or is influential in targeting RHA into microvesicles, like cytotoxic T cell granules, remains to be determined. However, preliminary data supports a role of RHA's

C-terminal DUF and RGG domains in harboring membrane targeting residues that direct its packaging into secreted microvesicles (S.P. and K.B.L., unpublished data). It is interesting to consider a mechanistic link between the dynamics of RHA with the 80S

211 ribosome and those that regulate its membrane targeting. This would afford a unique molecular connection between a DExH-box RNA helicase functioning in translation and its role in the innate immune response.

Conclusion

The DExH-box RNA helicase RNA helicase A is an essential host protein fundamental to cell biology. One of its critical functions is the targeted regulation of viral and cellular protein synthesis. This dissertation aimed to enhance our understanding of

RHA and cell biology by elucidating the molecular basis of RHA in eukaryotic translation control. The outcomes have been novel mechanistic insight into the eukaryotic translation process and added significance for the clinical importance of RHA in animal and human health and disease. As technology continues to develop and scientists probe deeper and more complex biological questions, it will be interesting to learn how this work informs future studies and scientific discoveries.

212

References

Abaeva, I. S., Marintchev, A., Pisareva, V. P., Hellen, C. U. T., & Pestova, T. V. (2011). Bypassing of stems versus linear base-by-base inspection of mammalian mRNAs during ribosomal scanning. The EMBO Journal, 30(1), 115–129. http://doi.org/10.1038/emboj.2010.302

Abdelhaleem, M. M., Hameed, S., Klassen, D., & Greenberg, A. H. (1996). Leukophysin: an RNA helicase A-related molecule identified in cytotoxic T cell granules and vesicles. Journal of Immunology (Baltimore, Md. : 1950), 156(6), 2026–2035.

Anderson, S. F., Schlegel, B. P., Nakajima, T., Wolpin, E. S., & Parvin, J. D. (1998). BRCA1 protein is linked to the RNA polymerase II holoenzyme complex via RNA helicase A. Nature Genetics, 19(3), 254–256. http://doi.org/10.1038/930

Apcher, S., Daskalogianni, C., Lejeune, F., Manoury, B., Imhoos, G., Heslop, L., & Fåhraeus, R. (2011). Major source of antigenic peptides for the MHC class I pathway is produced during the pioneer round of mRNA translation. Proceedings of the National Academy of Sciences of the United States of America, 108(28), 11572– 11577. http://doi.org/10.1073/pnas.1104104108

Aratani, S., Fujii, R., Oishi, T., Fujita, H., Amano, T., Ohshima, T., et al. (2001). Dual roles of RNA helicase A in CREB-dependent transcription. Molecular and Cellular Biology, 21(14), 4460–4469. http://doi.org/10.1128/MCB.21.14.4460-4469.2001

Beckmann, K., Grskovic, M., Gebauer, F., & Hentze, M. W. (2005). A dual inhibitory mechanism restricts msl-2 mRNA translation for dosage compensation in Drosophila. Cell, 122(4), 529–540. http://doi.org/10.1016/j.cell.2005.06.011

Bhat, M., Robichaud, N., Hulea, L., Sonenberg, N., Pelletier, J., & Topisirovic, I. (2015). Targeting the translation machinery in cancer. Nature Reviews. Drug Discovery, 14(4), 261–278. http://doi.org/10.1038/nrd4505

Blackwell, E., & Ceman, S. (2011). A new regulatory function of the region proximal to the RGG box in the fragile X mental retardation protein. Journal of Cell Science, 124(Pt 18), 3060–3065. http://doi.org/10.1242/jcs.086751

213

Bleichert, F., & Baserga, S. J. (2010). Ribonucleoprotein multimers and their functions. Critical Reviews in Biochemistry and Molecular Biology, 45(5), 331–350. http://doi.org/10.3109/10409238.2010.496772

Blobel, G., & Sabatini, D. (1971). Dissociation of mammalian polyribosomes into subunits by puromycin. Proceedings of the National Academy of Sciences of the United States of America, 68(2), 390–394.

Bolinger, C., Sharma, A., Singh, D., Yu, L., & Boris-Lawrie, K. (2010). RNA helicase A modulates translation of HIV-1 and infectivity of progeny virions. Nucleic Acids Research, 38(5), 1686–1696. http://doi.org/10.1093/nar/gkp1075

Bolinger, C., Yilmaz, A., Hartman, T. R., Kovacic, M. B., Fernandez, S., Ye, J., et al. (2007). RNA helicase A interacts with divergent lymphotropic retroviruses and promotes translation of human T-cell leukemia virus type 1. Nucleic Acids Research, 35(8), 2629–2642. http://doi.org/10.1093/nar/gkm124

Butsch, M., & Boris-Lawrie, K. (2000). Translation is not required To generate virion precursor RNA in human immunodeficiency virus type 1-infected T cells. Journal of Virology, 74(24), 11531–11537.

Butsch, M., Hull, S., Wang, Y., Roberts, T. M., & Boris-Lawrie, K. (1999). The 5' RNA terminus of spleen necrosis virus contains a novel posttranscriptional control element that facilitates human immunodeficiency virus Rev/RRE-independent Gag production. Journal of Virology, 73(6), 4847–4855.

Cannon, M. (1968). The puromycin reaction and its inhibition by chloramphenicol. European Journal of Biochemistry / FEBS, 7(1), 137–145.

Carpenter, S., Ricci, E. P., Mercier, B. C., Moore, M. J., & Fitzgerald, K. A. (2014). Post- transcriptional regulation of gene expression in innate immunity. Nature Reviews. Immunology, 14(6), 361–376. http://doi.org/10.1038/nri3682

Chen, E., Sharma, M. R., Shi, X., Agrawal, R. K., & Joseph, S. (2014). Fragile X mental retardation protein regulates translation by binding directly to the ribosome. Molecular Cell, 54(3), 407–417. http://doi.org/10.1016/j.molcel.2014.03.023

Chiu, S.-Y., Lejeune, F., Ranganathan, A. C., & Maquat, L. E. (2004). The pioneer translation initiation complex is functionally distinct from but structurally overlaps with the steady-state translation initiation complex. Genes & Development, 18(7), 745–754. http://doi.org/10.1101/gad.1170204

214

Choe, J., Ahn, S. H., & Kim, Y. K. (2014a). The mRNP remodeling mediated by UPF1 promotes rapid degradation of replication-dependent histone mRNA. Nucleic Acids Research, 42(14), 9334–9349. http://doi.org/10.1093/nar/gku610

Choe, J., Kim, K. M., Park, S., Lee, Y. K., Song, O.-K., Kim, M. K., et al. (2013). Rapid degradation of replication-dependent histone mRNAs largely occurs on mRNAs bound by nuclear cap-binding proteins 80 and 20. Nucleic Acids Research, 41(2), 1307–1318. http://doi.org/10.1093/nar/gks1196

Choe, J., Oh, N., Park, S., Lee, Y. K., Song, O.-K., Locker, N., et al. (2012). Translation initiation on mRNAs bound by nuclear cap-binding protein complex CBP80/20 requires interaction between CBP80/20-dependent translation initiation factor and eukaryotic translation initiation factor 3g. The Journal of Biological Chemistry, 287(22), 18500–18509. http://doi.org/10.1074/jbc.M111.327528

Choe, J., Ryu, I., Park, O. H., Park, J., Cho, H., Yoo, J. S., et al. (2014b). eIF4AIII enhances translation of nuclear cap-binding complex-bound mRNAs by promoting disruption of secondary structures in 5'UTR. Proceedings of the National Academy of Sciences of the United States of America, 111(43), E4577–86. http://doi.org/10.1073/pnas.1409695111

Cole, J. L. (2007). Activation of PKR: an open and shut case? Trends in Biochemical Sciences, 32(2), 57–62. http://doi.org/10.1016/j.tibs.2006.12.003

Cook, A., Bono, F., Jinek, M., & Conti, E. (2007). Structural biology of nucleocytoplasmic transport. Annual Review of Biochemistry, 76(1), 647–671. http://doi.org/10.1146/annurev.biochem.76.052705.161529

Cordin, O., Banroques, J., Tanner, N. K., & Linder, P. (2006). The DEAD-box protein family of RNA helicases. Gene, 367, 17–37. http://doi.org/10.1016/j.gene.2005.10.019

Daher, A., Longuet, M., Dorin, D., Bois, F., Segeral, E., Bannwarth, S., et al. (2001). Two dimerization domains in the trans-activation response RNA-binding protein (TRBP) individually reverse the protein kinase R inhibition of HIV-1 long terminal repeat expression. The Journal of Biological Chemistry, 276(36), 33899–33905. http://doi.org/10.1074/jbc.M103584200

Dangel, A. W., Hull, S., Roberts, T. M., & Boris-Lawrie, K. (2002). Nuclear interactions are necessary for translational enhancement by spleen necrosis virus RU5. Journal of Virology, 76(7), 3292–3300.

215

Darnell, J. C., Van Driesche, S. J., Zhang, C., Hung, K. Y. S., Mele, A., Fraser, C. E., et al. (2011). FMRP stalls ribosomal translocation on mRNAs linked to synaptic function and autism. Cell, 146(2), 247–261. http://doi.org/10.1016/j.cell.2011.06.013

De Leeuw, F., Zhang, T., Wauquier, C., Huez, G., Kruys, V., & Gueydan, C. (2007). The cold-inducible RNA-binding protein migrates from the nucleus to cytoplasmic stress granules by a methylation-dependent mechanism and acts as a translational repressor. Experimental Cell Research, 313(20), 4130–4144. http://doi.org/10.1016/j.yexcr.2007.09.017

Dever, T. E., & Green, R. (2012). The Elongation, Termination, and Recycling Phases of Translation in Eukaryotes. Cold Spring Harbor Perspectives in Biology, 4(7), a013706–a013706. http://doi.org/10.1101/cshperspect.a013706

Dhote, V., Sweeney, T. R., Kim, N., Hellen, C. U. T., & Pestova, T. V. (2012). Roles of individual domains in the function of DHX29, an essential factor required for translation of structured mammalian mRNAs. Proceedings of the National Academy of Sciences of the United States of America, 109(46), E3150–9. http://doi.org/10.1073/pnas.1208014109

Diem, M. D., Chan, C. C., Younis, I., & Dreyfuss, G. (2007). PYM binds the cytoplasmic exon-junction complex and ribosomes to enhance translation of spliced mRNAs. Nature Structural & Molecular Biology, 14(12), 1173–1179. http://doi.org/10.1038/nsmb1321

Dresios, J., Aschrafi, A., Owens, G. C., Vanderklish, P. W., Edelman, G. M., & Mauro, V. P. (2005). Cold stress-induced protein Rbm3 binds 60S ribosomal subunits, alters microRNA levels, and enhances global protein synthesis. Proceedings of the National Academy of Sciences of the United States of America, 102(6), 1865–1870. http://doi.org/10.1073/pnas.0409764102

Duncan, K. E., Strein, C., & Hentze, M. W. (2009). The SXL-UNR corepressor complex uses a PABP-mediated mechanism to inhibit ribosome recruitment to msl-2 mRNA. Molecular Cell, 36(4), 571–582. http://doi.org/10.1016/j.molcel.2009.09.042

Firth, A. E., & Brierley, I. (2012). Non-canonical translation in RNA viruses. The Journal of General Virology, 93(Pt 7), 1385–1409. http://doi.org/10.1099/vir.0.042499-0

Flaherty, S. M., Fortes, P., Izaurralde, E., Mattaj, I. W., & Gilmartin, G. M. (1997). Participation of the nuclear cap binding complex in pre-mRNA 3' processing. Proceedings of the National Academy of Sciences of the United States of America, 94(22), 11893–11898.

216

Fortes, P., Inada, T., Preiss, T., Hentze, M. W., Mattaj, I. W., & Sachs, A. B. (2000). The yeast nuclear cap binding complex can interact with translation factor eIF4G and mediate translation initiation. Molecular Cell, 6(1), 191–196.

Frank, J., & Gonzalez, R. L. (2010). Structure and dynamics of a processive Brownian motor: the translating ribosome. Annual Review of Biochemistry, 79(1), 381–412. http://doi.org/10.1146/annurev-biochem-060408-173330

Fresno, M., Jiménez, A., & Vázquez, D. (1977). Inhibition of translation in eukaryotic systems by harringtonine. European Journal of Biochemistry / FEBS, 72(2), 323– 330.

Friend, K., Campbell, Z. T., Cooke, A., Kroll-Conner, P., Wickens, M. P., & Kimble, J. (2012). A conserved PUF-Ago-eEF1A complex attenuates translation elongation. Nature Structural & Molecular Biology, 19(2), 176–183. http://doi.org/10.1038/nsmb.2214

Gebauer, F., Grskovic, M., & Hentze, M. W. (2003). Drosophila sex-lethal inhibits the stable association of the 40S ribosomal subunit with msl-2 mRNA. Molecular Cell, 11(5), 1397–1404.

Geissler, R., Golbik, R. P., & Behrens, S.-E. (2012). The DEAD-box helicase DDX3 supports the assembly of functional 80S ribosomes. Nucleic Acids Research, 40(11), 4998–5011. http://doi.org/10.1093/nar/gks070

Gilbert, W. V. (2010). Alternative ways to think about cellular internal ribosome entry. The Journal of Biological Chemistry, 285(38), 29033–29038. http://doi.org/10.1074/jbc.R110.150532

Gleghorn, M. L., & Maquat, L. E. (2014). 'Black sheep‘ that don’t leave the double- stranded RNA-binding domain fold. Trends in Biochemical Sciences, 39(7), 328– 340. http://doi.org/10.1016/j.tibs.2014.05.003

Gleghorn, M. L., Gong, C., Kielkopf, C. L., & Maquat, L. E. (2013). Staufen1 dimerizes through a conserved motif and a degenerate dsRNA-binding domain to promote mRNA decay. Nature Structural & Molecular Biology, 20(4), 515–524. http://doi.org/10.1038/nsmb.2528

Görlich, D., Kraft, R., Kostka, S., Vogel, F., Hartmann, E., Laskey, R. A., et al. (1996). Importin provides a link between nuclear protein import and U snRNA export. Cell, 87(1), 21–32.

217

Gray, N. K., & Hentze, M. W. (1994). Iron regulatory protein prevents binding of the 43S translation pre-initiation complex to ferritin and eALAS mRNAs. The EMBO Journal, 13(16), 3882–3891.

Guénard, F., Labrie, Y., Ouellette, G., Beauparlant, C. J., Durocher, F., INHERIT BRCAs. (2009). Genetic sequence variations of BRCA1-interacting genes AURKA, BAP1, BARD1 and DHX9 in French Canadian families with high risk of breast cancer. Journal of Human Genetics, 54(3), 152–161. http://doi.org/10.1038/jhg.2009.6

Haghighat, A., Mader, S., Pause, A., & Sonenberg, N. (1995). Repression of cap- dependent translation by 4E-binding protein 1: competition with p220 for binding to eukaryotic initiation factor-4E. The EMBO Journal, 14(22), 5701–5709.

Hartman, T. R., Qian, S., Bolinger, C., Fernandez, S., Schoenberg, D. R., & Boris- Lawrie, K. (2006). RNA helicase A is necessary for translation of selected messenger RNAs. Nature Structural & Molecular Biology, 13(6), 509–516. http://doi.org/10.1038/nsmb1092

Hashem, Y., Georges, des, A., Dhote, V., Langlois, R., Liao, H. Y., Grassucci, R. A., et al. (2013). Structure of the mammalian ribosomal 43S preinitiation complex bound to the scanning factor DHX29. Cell, 153(5), 1108–1119. http://doi.org/10.1016/j.cell.2013.04.036

Hernandez, J. M., Floyd, D. H., Weilbaecher, K. N., Green, P. L., & Boris-Lawrie, K. (2008). Multiple facets of junD gene expression are atypical among AP-1 family members. Oncogene, 27(35), 4757–4767. http://doi.org/10.1038/onc.2008.120

Hinnebusch, A. G. (2014). The scanning mechanism of eukaryotic translation initiation. Annual Review of Biochemistry, 83(1), 779–812. http://doi.org/10.1146/annurev- biochem-060713-035802

Hinnebusch, A. G., & Lorsch, J. R. (2012). The mechanism of eukaryotic translation initiation: new insights and challenges. Cold Spring Harbor Perspectives in Biology, 4(10), a011544–a011544. http://doi.org/10.1101/cshperspect.a011544

Hizli, A. A., Chi, Y., Swanger, J., Carter, J. H., Liao, Y., Welcker, M., et al. (2013). Phosphorylation of Eukaryotic Elongation Factor 2 (eEF2) by Cyclin A–Cyclin- Dependent Kinase 2 Regulates Its Inhibition by eEF2 Kinase. Molecular and Cellular Biology, 33(3), 596–604. http://doi.org/10.1128/MCB.01270-12

218

Holcik, M., & Sonenberg, N. (2005). Translational control in stress and apoptosis. Nature Reviews. Molecular Cell Biology, 6(4), 318–327. http://doi.org/10.1038/nrm1618

Hull, S., & Boris-Lawrie, K. (2002). RU5 of Mason-Pfizer monkey virus 5' long terminal repeat enhances cytoplasmic expression of human immunodeficiency virus type 1 gag-pol and nonviral reporter RNA. Journal of Virology, 76(20), 10211–10218.

Hussey, G. S., Chaudhury, A., Dawson, A. E., Lindner, D. J., Knudsen, C. R., Wilce, M. C. J., et al. (2011). Identification of an mRNP complex regulating tumorigenesis at the translational elongation step. Molecular Cell, 41(4), 419–431. http://doi.org/10.1016/j.molcel.2011.02.003

Hüttelmaier, S., Zenklusen, D., Lederer, M., Dictenberg, J., Lorenz, M., Meng, X., et al. (2005). Spatial regulation of beta-actin translation by Src-dependent phosphorylation of ZBP1. Nature, 438(7067), 512–515. http://doi.org/10.1038/nature04115

Hwang, J., Sato, H., Tang, Y., Matsuda, D., & Maquat, L. E. (2010). UPF1 association with the cap-binding protein, CBP80, promotes nonsense-mediated mRNA decay at two distinct steps. Molecular Cell, 39(3), 396–409. http://doi.org/10.1016/j.molcel.2010.07.004

Ingolia, N. T. (2014). Ribosome profiling: new views of translation, from single codons to genome scale. Nature Reviews. Genetics, 15(3), 205–213. http://doi.org/10.1038/nrg3645

Ingolia, N. T., Brar, G. A., Rouskin, S., McGeachy, A. M., & Weissman, J. S. (2012). The ribosome profiling strategy for monitoring translation in vivo by deep sequencing of ribosome-protected mRNA fragments. Nature Protocols, 7(8), 1534– 1550. http://doi.org/10.1038/nprot.2012.086

Ingolia, N. T., Lareau, L. F., & Weissman, J. S. (2011). Ribosome profiling of mouse embryonic stem cells reveals the complexity and dynamics of mammalian proteomes. Cell, 147(4), 789–802. http://doi.org/10.1016/j.cell.2011.10.002

Ishigaki, Y., Li, X., Serin, G., & Maquat, L. E. (2001). Evidence for a pioneer round of mRNA translation: mRNAs subject to nonsense-mediated decay in mammalian cells are bound by CBP80 and CBP20. Cell, 106(5), 607–617.

Izaurralde, E., Lewis, J., Gamberi, C., Jarmolowski, A., McGuigan, C., & Mattaj, I. W. (1995a). A cap-binding protein complex mediating U snRNA export. Nature, 376(6542), 709–712. http://doi.org/10.1038/376709a0

219

Izaurralde, E., Lewis, J., McGuigan, C., Jankowska, M., Darzynkiewicz, E., & Mattaj, I. W. (1994). A nuclear cap binding protein complex involved in pre-mRNA splicing. Cell, 78(4), 657–668.

Izaurralde, E., McGuigan, C., & Mattaj, I. W. (1995b). Nuclear localization of a cap- binding protein complex. Cold Spring Harbor Symposia on Quantitative Biology, 60, 669–675.

Izaurralde, E., Stepinski, J., Darzynkiewicz, E., & Mattaj, I. W. (1992). A cap binding protein that may mediate nuclear export of RNA polymerase II-transcribed RNAs. The Journal of Cell Biology, 118(6), 1287–1295.

Jackson, R. J., Hellen, C. U. T., & Pestova, T. V. (2010). The mechanism of eukaryotic translation initiation and principles of its regulation. Nature Reviews. Molecular Cell Biology, 11(2), 113–127. http://doi.org/10.1038/nrm2838 Jackson, R. J., Hellen, C. U. T., & Pestova, T. V. (2012). Termination and post- termination events in eukaryotic translation. Advances in Protein Chemistry and Structural Biology, 86, 45–93. http://doi.org/10.1016/B978-0-12-386497-0.00002-5

Jankowsky, E., & Bowers, H. (2006). Remodeling of ribonucleoprotein complexes with DExH/D RNA helicases. Nucleic Acids Research, 34(15), 4181–4188. http://doi.org/10.1093/nar/gkl410

Jin, J., Jing, W., Lei, X.-X., Feng, C., Peng, S., Boris-Lawrie, K., & Huang, Y. (2011). Evidence that Lin28 stimulates translation by recruiting RNA helicase A to polysomes. Nucleic Acids Research, 39(9), 3724–3734. http://doi.org/10.1093/nar/gkq1350

Khu, Y. L., Koh, E., Lim, S. P., Tan, Y. H., Brenner, S., Lim, S. G., et al. (2001). Mutations that affect dimer formation and helicase activity of the hepatitis C virus helicase. Journal of Virology, 75(1), 205–214. http://doi.org/10.1128/JVI.75.1.205- 214.2001

Kim, K. M., Cho, H., Choi, K., Kim, J., Kim, B.-W., Ko, Y.-G., et al. (2009). A new MIF4G domain-containing protein, CTIF, directs nuclear cap-binding protein CBP80/20-dependent translation. Genes & Development, 23(17), 2033–2045. http://doi.org/10.1101/gad.1823409

Kim, T., Pazhoor, S., Bao, M., Zhang, Z., Hanabuchi, S., Facchinetti, V., et al. (2010). Aspartate-glutamate-alanine-histidine box motif (DEAH)/RNA helicase A helicases sense microbial DNA in human plasmacytoid dendritic cells. Proceedings of the National Academy of Sciences of the United States of America, 107(34), 15181– 15186. http://doi.org/10.1073/pnas.1006539107

220

King, H. A., Cobbold, L. C., & Willis, A. E. (2010). The role of IRES trans-acting factors in regulating translation initiation. Biochemical Society Transactions, 38(6), 1581– 1586. http://doi.org/10.1042/BST0381581

Klostermeier, D., & Rudolph, M. G. (2009). A novel dimerization motif in the C-terminal domain of the Thermus thermophilus DEAD box helicase Hera confers substantial flexibility. Nucleic Acids Research, 37(2), 421–430. http://doi.org/10.1093/nar/gkn947

Komar, A. A., & Hatzoglou, M. (2011). Cellular IRES-mediated translation: the war of ITAFs in pathophysiological states. Cell Cycle (Georgetown, Tex.), 10(2), 229–240.

Kozak, M. (1978). How do eucaryotic ribosomes select initiation regions in messenger RNA? Cell, 15(4), 1109–1123.

Kühn, U., & Wahle, E. (2004). Structure and function of poly(A) binding proteins. Biochimica Et Biophysica Acta, 1678(2-3), 67–84. http://doi.org/10.1016/j.bbaexp.2004.03.008

Lai, M.-C., Chang, W.-C., Shieh, S.-Y., & Tarn, W.-Y. (2010). DDX3 regulates cell growth through translational control of cyclin E1. Molecular and Cellular Biology, 30(22), 5444–5453. http://doi.org/10.1128/MCB.00560-10

Lai, M.-C., Lee, Y.-H. W., & Tarn, W.-Y. (2008). The DEAD-box RNA helicase DDX3 associates with export messenger ribonucleoproteins as well as tip-associated protein and participates in translational control. Molecular Biology of the Cell, 19(9), 3847– 3858. http://doi.org/10.1091/mbc.E07-12-1264

Lai, M.-C., Wang, S.-W., Cheng, L., Tarn, W.-Y., Tsai, S.-J., & Sun, H. S. (2013). Human DDX3 interacts with the HIV-1 Tat protein to facilitate viral mRNA translation. PloS One, 8(7), e68665. http://doi.org/10.1371/journal.pone.0068665

Laplante, M., & Sabatini, D. M. (2012). mTOR signaling in growth control and disease. Cell, 149(2), 274–293. http://doi.org/10.1016/j.cell.2012.03.017

Lee, C. G., Eki, T., Okumura, K., Nogami, M., Soares, V. D. C., Murakami, Y., et al. (1999). The human RNA helicase A (DDX9) gene maps to the prostate cancer susceptibility locus at chromosome band 1q25 and its pseudogene (DDX9P) to 13q22, respectively. Somatic Cell and Molecular Genetics, 25(1), 33–39.

Lee, C.-S., Dias, A. P., Jedrychowski, M., Patel, A. H., Hsu, J. L., & Reed, R. (2008). Human DDX3 functions in translation and interacts with the translation initiation factor eIF3. Nucleic Acids Research, 36(14), 4708–4718. http://doi.org/10.1093/nar/gkn454 221

Lejeune, F., Ishigaki, Y., Li, X., & Maquat, L. E. (2002). The exon junction complex is detected on CBP80-bound but not eIF4E-bound mRNA in mammalian cells: dynamics of mRNP remodeling. The EMBO Journal, 21(13), 3536–3545. http://doi.org/10.1093/emboj/cdf345

Lejeune, F., Ranganathan, A. C., & Maquat, L. E. (2004). eIF4G is required for the pioneer round of translation in mammalian cells. Nature Structural & Molecular Biology, 11(10), 992–1000. http://doi.org/10.1038/nsmb824

Lewis, J. D., & Izaurralde, E. (1997). The role of the cap structure in RNA processing and nuclear export. European Journal of Biochemistry / FEBS, 247(2), 461–469.

Linder, P. (2006). Dead-box proteins: a family affair--active and passive players in RNP- remodeling. Nucleic Acids Research, 34(15), 4168–4180. http://doi.org/10.1093/nar/gkl468

Liu, B., & Qian, S.-B. (2014). Translational reprogramming in cellular stress response. Wiley Interdisciplinary Reviews. RNA, 5(3), 301–315. http://doi.org/10.1002/wrna.1212

Luo, M., Duchaîne, T. F., & DesGroseillers, L. (2002). Molecular mapping of the determinants involved in human Staufen-ribosome association. The Biochemical Journal, 365(Pt 3), 817–824.

Lykke-Andersen, J., & Bennett, E. J. (2014). Protecting the proteome: Eukaryotic cotranslational quality control pathways. The Journal of Cell Biology, 204(4), 467– 476. http://doi.org/10.1083/jcb.201311103

Lyumkis, D., Doamekpor, S. K., Bengtson, M. H., Lee, J.-W., Toro, T. B., Petroski, M. D., et al. (2013). Single-particle EM reveals extensive conformational variability of the Ltn1 E3 ligase. Proceedings of the National Academy of Sciences of the United States of America, 110(5), 1702–1707. http://doi.org/10.1073/pnas.1210041110

Ma, X. M., & Blenis, J. (2009). Molecular mechanisms of mTOR-mediated translational control. Nature Reviews. Molecular Cell Biology, 10(5), 307–318. http://doi.org/10.1038/nrm2672

Mader, S., Lee, H., Pause, A., & Sonenberg, N. (1995). The translation initiation factor eIF-4E binds to a common motif shared by the translation factor eIF-4 gamma and the translational repressors 4E-binding proteins. Molecular and Cellular Biology, 15(9), 4990–4997.

222

Mallam, A. L., Jarmoskaite, I., Tijerina, P., Del Campo, M., Seifert, S., Guo, L., et al. (2011). Solution structures of DEAD-box RNA chaperones reveal conformational changes and nucleic acid tethering by a basic tail. Proceedings of the National Academy of Sciences of the United States of America, 108(30), 12254–12259. http://doi.org/10.1073/pnas.1109566108

Manojlovic, Z., & Stefanovic, B. (2012). A novel role of RNA helicase A in regulation of translation of type I collagen mRNAs. RNA (New York, N.Y.), 18(2), 321–334. http://doi.org/10.1261/rna.030288.111

Maquat, L. E., Tarn, W.-Y., & Isken, O. (2010). The pioneer round of translation: features and functions. Cell, 142(3), 368–374. http://doi.org/10.1016/j.cell.2010.07.022

Marcotrigiano, J., Gingras, A. C., Sonenberg, N., & Burley, S. K. (1999). Cap-dependent translation initiation in eukaryotes is regulated by a molecular mimic of eIF4G. Molecular Cell, 3(6), 707–716.

Martel, C., Dugré-Brisson, S., Boulay, K., Breton, B., Lapointe, G., Armando, S., et al. (2010). Multimerization of Staufen1 in live cells. RNA (New York, N.Y.), 16(3), 585– 597. http://doi.org/10.1261/rna.1664210

Matsuda, D., & Dreher, T. W. (2006). Close spacing of AUG initiation codons confers dicistronic character on a eukaryotic mRNA. RNA (New York, N.Y.), 12(7), 1338– 1349. http://doi.org/10.1261/rna.67906

Michel, A. M., & Baranov, P. V. (2013). Ribosome profiling: a Hi-Def monitor for protein synthesis at the genome-wide scale. Wiley Interdisciplinary Reviews. RNA, 4(5), 473–490. http://doi.org/10.1002/wrna.1172

Mitchell, S. F., & Parker, R. (2014). Principles and properties of eukaryotic mRNPs. Molecular Cell, 54(4), 547–558. http://doi.org/10.1016/j.molcel.2014.04.033

Moore, M. J. (2005). From birth to death: the complex lives of eukaryotic mRNAs. Science (New York, N.Y.), 309(5740), 1514–1518. http://doi.org/10.1126/science.1111443

Muckenthaler, M., Gray, N. K., & Hentze, M. W. (1998). IRP-1 binding to ferritin mRNA prevents the recruitment of the small ribosomal subunit by the cap-binding complex eIF4F. Molecular Cell, 2(3), 383–388.

Nakajima, T., Uchida, C., Anderson, S. F., Lee, C. G., Hurwitz, J., Parvin, J. D., & Montminy, M. (1997). RNA helicase A mediates association of CBP with RNA polymerase II. Cell, 90(6), 1107–1112. 223

Nikolaitchik, O. A., Dilley, K. A., Fu, W., Gorelick, R. J., Tai, S.-H. S., Soheilian, F., et al. (2013). Dimeric RNA recognition regulates HIV-1 genome packaging. PLoS Pathogens, 9(3), e1003249. http://doi.org/10.1371/journal.ppat.1003249

Nottrott, S., Simard, M. J., & Richter, J. D. (2006). Human let-7a miRNA blocks protein production on actively translating polyribosomes. Nature Structural & Molecular Biology, 13(12), 1108–1114. http://doi.org/10.1038/nsmb1173

Oh, N., Kim, K. M., Cho, H., Choe, J., & Kim, Y. K. (2007a). Pioneer round of translation occurs during serum starvation. Biochemical and Biophysical Research Communications, 362(1), 145–151. http://doi.org/10.1016/j.bbrc.2007.07.169

Oh, N., Kim, K. M., Choe, J., & Kim, Y. K. (2007b). Pioneer round of translation mediated by nuclear cap-binding proteins CBP80/20 occurs during prolonged hypoxia. FEBS Letters, 581(26), 5158–5164. http://doi.org/10.1016/j.febslet.2007.10.002

Ostareck, D. H., Ostareck-Lederer, A., Shatsky, I. N., & Hentze, M. W. (2001). Lipoxygenase mRNA silencing in erythroid differentiation: The 3'UTR regulatory complex controls 60S ribosomal subunit joining. Cell, 104(2), 281–290.

Ostareck, D. H., Ostareck-Lederer, A., Wilm, M., Thiele, B. J., Mann, M., & Hentze, M. W. (1997). mRNA silencing in erythroid differentiation: hnRNP K and hnRNP E1 regulate 15-lipoxygenase translation from the 3' end. Cell, 89(4), 597–606.

Park, E., & Maquat, L. E. (2013). Staufen-mediated mRNA decay. Wiley Interdisciplinary Reviews. RNA, 4(4), 423–435. http://doi.org/10.1002/wrna.1168

Park, E., Gleghorn, M. L., & Maquat, L. E. (2013). Staufen2 functions in Staufen1- mediated mRNA decay by binding to itself and its paralog and promoting UPF1 helicase but not ATPase activity. Proceedings of the National Academy of Sciences of the United States of America, 110(2), 405–412. http://doi.org/10.1073/pnas.1213508110

Parsyan, A., Shahbazian, D., Martineau, Y., Petroulakis, E., Alain, T., Larsson, O., et al. (2009). The helicase protein DHX29 promotes translation initiation, cell proliferation, and tumorigenesis. Proceedings of the National Academy of Sciences of the United States of America, 106(52), 22217–22222. http://doi.org/10.1073/pnas.0909773106

Parsyan, A., Svitkin, Y., Shahbazian, D., Gkogkas, C., Lasko, P., Merrick, W. C., & Sonenberg, N. (2011). mRNA helicases: the tacticians of translational control. Nature Reviews. Molecular Cell Biology, 12(4), 235–245. http://doi.org/10.1038/nrm3083

224

Pause, A., Belsham, G. J., Gingras, A. C., Donzé, O., Lin, T. A., Lawrence, J. C., & Sonenberg, N. (1994). Insulin-dependent stimulation of protein synthesis by phosphorylation of a regulator of 5'-cap function. Nature, 371(6500), 762–767. http://doi.org/10.1038/371762a0

Pestova, T. V., Borukhov, S. I., & Hellen, C. U. (1998). Eukaryotic ribosomes require initiation factors 1 and 1A to locate initiation codons. Nature, 394(6696), 854–859. http://doi.org/10.1038/29703

Peters, G. A., Hartmann, R., Qin, J., & Sen, G. C. (2001). Modular structure of PACT: distinct domains for binding and activating PKR. Molecular and Cellular Biology, 21(6), 1908–1920. http://doi.org/10.1128/MCB.21.6.1908-1920.2001

Petersen, C. P., Bordeleau, M.-E., Pelletier, J., & Sharp, P. A. (2006). Short RNAs repress translation after initiation in mammalian cells. Molecular Cell, 21(4), 533– 542. http://doi.org/10.1016/j.molcel.2006.01.031

Phan, A. T., Kuryavyi, V., Darnell, J. C., Serganov, A., Majumdar, A., Ilin, S., et al. (2011). Structure-function studies of FMRP RGG peptide recognition of an RNA duplex-quadruplex junction. Nature Structural & Molecular Biology, 18(7), 796– 804. http://doi.org/10.1038/nsmb.2064

Pisareva, V. P., Pisarev, A. V., Komar, A. A., Hellen, C. U. T., & Pestova, T. V. (2008). Translation initiation on mammalian mRNAs with structured 5'UTRs requires DExH-box protein DHX29. Cell, 135(7), 1237–1250. http://doi.org/10.1016/j.cell.2008.10.037

Polacek, N., & Mankin, A. S. (2005). The ribosomal peptidyl transferase center: structure, function, evolution, inhibition. Critical Reviews in Biochemistry and Molecular Biology, 40(5), 285–311. http://doi.org/10.1080/10409230500326334

Prévôt, D., Darlix, J.-L., & Ohlmann, T. (2003). Conducting the initiation of protein synthesis: the role of eIF4G. Biology of the Cell / Under the Auspices of the European Cell Biology Organization, 95(3-4), 141–156.

Pyronnet, S., & Sonenberg, N. (2001). Cell-cycle-dependent translational control. Current Opinion in Genetics & Development, 11(1), 13–18.

Qian, S., Zhong, X., Yu, L., Ding, B., de Haan, P., & Boris-Lawrie, K. (2009). HIV-1 Tat RNA silencing suppressor activity is conserved across kingdoms and counteracts translational repression of HIV-1. Proceedings of the National Academy of Sciences of the United States of America, 106(2), 605–610. http://doi.org/10.1073/pnas.0806822106

225

Qin, X., & Sarnow, P. (2004). Preferential translation of internal ribosome entry site- containing mRNAs during the mitotic cycle in mammalian cells. The Journal of Biological Chemistry, 279(14), 13721–13728. http://doi.org/10.1074/jbc.M312854200

Rajyaguru, P., She, M., & Parker, R. (2012). Scd6 targets eIF4G to repress translation: RGG motif proteins as a class of eIF4G-binding proteins. Molecular Cell, 45(2), 244–254. http://doi.org/10.1016/j.molcel.2011.11.026

Ranji, A., & Boris-Lawrie, K. (2010). RNA helicases: emerging roles in viral replication and the host innate response. RNA Biology, 7(6), 775–787.

Ranji, A., Shkriabai, N., Kvaratskhelia, M., Musier-Forsyth, K., & Boris-Lawrie, K. (2011). Features of double-stranded RNA-binding domains of RNA helicase A are necessary for selective recognition and translation of complex mRNAs. The Journal of Biological Chemistry, 286(7), 5328–5337. http://doi.org/10.1074/jbc.M110.176339

Ricci, E. P., Kucukural, A., Cenik, C., Mercier, B. C., Singh, G., Heyer, E. E., et al. (2014). Staufen1 senses overall transcript secondary structure to regulate translation. Nature Structural & Molecular Biology, 21(1), 26–35. http://doi.org/10.1038/nsmb.2739

Roberts, T. M., & Boris-Lawrie, K. (2000). The 5' RNA terminus of spleen necrosis virus stimulates translation of nonviral mRNA. Journal of Virology, 74(17), 8111–8118.

Roberts, T. M., & Boris-Lawrie, K. (2003). Primary sequence and secondary structure motifs in spleen necrosis virus RU5 confer translational utilization of unspliced human immunodeficiency virus type 1 reporter RNA. Journal of Virology, 77(22), 11973–11984.

Rocak, S., & Linder, P. (2004). DEAD-box proteins: the driving forces behind RNA metabolism. Nature Reviews. Molecular Cell Biology, 5(3), 232–241. http://doi.org/10.1038/nrm1335

Ruggero, D. (2013). Translational control in cancer etiology. Cold Spring Harbor Perspectives in Biology, 5(2), a012336–a012336. http://doi.org/10.1101/cshperspect.a012336

Ruvinsky, I., & Meyuhas, O. (2006). Ribosomal protein S6 phosphorylation: from protein synthesis to cell size. Trends in Biochemical Sciences, 31(6), 342–348. http://doi.org/10.1016/j.tibs.2006.04.003

226

Sadler, A. J., Latchoumanin, O., Hawkes, D., Mak, J., & Williams, B. R. G. (2009). An antiviral response directed by PKR phosphorylation of the RNA helicase A. PLoS Pathogens, 5(2), e1000311. http://doi.org/10.1371/journal.ppat.1000311

Sasikumar, A. N., Perez, W. B., & Kinzy, T. G. (2012). The many roles of the eukaryotic elongation factor 1 complex. Wiley Interdisciplinary Reviews. RNA, 3(4), 543–555. http://doi.org/10.1002/wrna.1118

Sato, H., & Maquat, L. E. (2009). Remodeling of the pioneer translation initiation complex involves translation and the karyopherin importin beta. Genes & Development, 23(21), 2537–2550. http://doi.org/10.1101/gad.1817109

Scheper, G. C., van der Knaap, M. S., & Proud, C. G. (2007). Translation matters: protein synthesis defects in inherited disease. Nature Reviews. Genetics, 8(9), 711–723. http://doi.org/10.1038/nrg2142

Schoenberg, D. R., & Maquat, L. E. (2012). Regulation of cytoplasmic mRNA decay. Nature Reviews. Genetics, 13(4), 246–259. http://doi.org/10.1038/nrg3160

Schröder, M. (2010). Human DEAD-box protein 3 has multiple functions in gene regulation and cell cycle control and is a prime target for viral manipulation. Biochemical Pharmacology, 79(3), 297–306. http://doi.org/10.1016/j.bcp.2009.08.032

Shalgi, R., Hurt, J. A., Krykbaeva, I., Taipale, M., Lindquist, S., & Burge, C. B. (2013). Widespread regulation of translation by elongation pausing in heat shock. Molecular Cell, 49(3), 439–452. http://doi.org/10.1016/j.molcel.2012.11.028

Sharma, A., Yilmaz, A., Marsh, K., Cochrane, A., & Boris-Lawrie, K. (2012). Thriving under stress: selective translation of HIV-1 structural protein mRNA during Vpr- mediated impairment of eIF4E translation activity. PLoS Pathogens, 8(3), e1002612. http://doi.org/10.1371/journal.ppat.1002612

Shih, J.-W., Tsai, T.-Y., Chao, C.-H., & Wu Lee, Y.-H. (2008). Candidate tumor suppressor DDX3 RNA helicase specifically represses cap-dependent translation by acting as an eIF4E inhibitory protein. Oncogene, 27(5), 700–714. http://doi.org/10.1038/sj.onc.1210687

Shih, J.-W., Wang, W.-T., Tsai, T.-Y., Kuo, C.-Y., Li, H.-K., & Wu Lee, Y.-H. (2012). Critical roles of RNA helicase DDX3 and its interactions with eIF4E/PABP1 in stress granule assembly and stress response. The Biochemical Journal, 441(1), 119–129. http://doi.org/10.1042/BJ20110739

227

Silvera, D., Formenti, S. C., & Schneider, R. J. (2010). Translational control in cancer. Nature Reviews. Cancer, 10(4), 254–266. http://doi.org/10.1038/nrc2824

Singh, G., Pratt, G., Yeo, G. W., & Moore, M. J. (2015). The Clothes Make the mRNA: Past and Present Trends in mRNP Fashion. Annual Review of Biochemistry, 84(1), 150317182619002. http://doi.org/10.1146/annurev-biochem-080111-092106

Skabkin, M. A., Skabkina, O. V., Dhote, V., Komar, A. A., Hellen, C. U. T., & Pestova, T. V. (2010). Activities of Ligatin and MCT-1/DENR in eukaryotic translation initiation and ribosomal recycling. Genes & Development, 24(16), 1787–1801. http://doi.org/10.1101/gad.1957510

Smith, W. A., Schurter, B. T., Wong-Staal, F., & David, M. (2004). Arginine methylation of RNA helicase a determines its subcellular localization. The Journal of Biological Chemistry, 279(22), 22795–22798. http://doi.org/10.1074/jbc.C300512200

Solomon, S., Xu, Y., Wang, B., David, M. D., Schubert, P., Kennedy, D., & Schrader, J. W. (2007). Distinct structural features of caprin-1 mediate its interaction with G3BP- 1 and its induction of phosphorylation of eukaryotic translation initiation factor 2alpha, entry to cytoplasmic stress granules, and selective interaction with a subset of mRNAs. Molecular and Cellular Biology, 27(6), 2324–2342. http://doi.org/10.1128/MCB.02300-06

Soto-Rifo, R., Rubilar, P. S., & Ohlmann, T. (2013). The DEAD-box helicase DDX3 substitutes for the cap-binding protein eIF4E to promote compartmentalized translation initiation of the HIV-1 genomic RNA. Nucleic Acids Research, 41(12), 6286–6299. http://doi.org/10.1093/nar/gkt306

Soto-Rifo, R., Rubilar, P. S., Limousin, T., de Breyne, S., Décimo, D., & Ohlmann, T. (2012). DEAD-box protein DDX3 associates with eIF4F to promote translation of selected mRNAs. The EMBO Journal, 31(18), 3745–3756. http://doi.org/10.1038/emboj.2012.220

Spriggs, K. A., Stoneley, M., Bushell, M., & Willis, A. E. (2008). Re-programming of translation following cell stress allows IRES-mediated translation to predominate., 100(1), 27–38. http://doi.org/10.1042/BC20070098

St Johnston, D., Brown, N. H., Gall, J. G., & Jantsch, M. (1992). A conserved double- stranded RNA-binding domain. Proceedings of the National Academy of Sciences of the United States of America, 89(22), 10979–10983.

228

Suhl, J. A., Chopra, P., Anderson, B. R., Bassell, G. J., & Warren, S. T. (2014). Analysis of FMRP mRNA target datasets reveals highly associated mRNAs mediated by G- quadruplex structures formed via clustered WGGA sequences. Human Molecular Genetics, 23(20), 5479–5491. http://doi.org/10.1093/hmg/ddu272

Svitkin, Y. V., Pause, A., Haghighat, A., Pyronnet, S., Witherell, G., Belsham, G. J., & Sonenberg, N. (2001). The requirement for eukaryotic initiation factor 4A (elF4A) in translation is in direct proportion to the degree of mRNA 5' secondary structure. RNA (New York, N.Y.), 7(3), 382–394.

Takeda, Y., Caudell, P., Grady, G., Wang, G., Suwa, A., Sharp, G. C., et al. (1999). Human RNA helicase A is a lupus autoantigen that is cleaved during apoptosis. Journal of Immunology (Baltimore, Md. : 1950), 163(11), 6269–6274.

Tang, H., McDonald, D., Middlesworth, T., Hope, T. J., & Wong-Staal, F. (1999). The carboxyl terminus of RNA helicase A contains a bidirectional nuclear transport domain. Molecular and Cellular Biology, 19(5), 3540–3550.

Thandapani, P., O'Connor, T. R., Bailey, T. L., & Richard, S. (2013). Defining the RGG/RG motif. Molecular Cell, 50(5), 613–623. http://doi.org/10.1016/j.molcel.2013.05.021

Thompson, S. R. (2012). So you want to know if your message has an IRES? Wiley Interdisciplinary Reviews. RNA, 3(5), 697–705. http://doi.org/10.1002/wrna.1129

Thoreen, C. C., Chantranupong, L., Keys, H. R., Wang, T., Gray, N. S., & Sabatini, D. M. (2012). A unifying model for mTORC1-mediated regulation of mRNA translation. Nature, 485(7396), 109–113. http://doi.org/10.1038/nature11083

Thoreen, C. C., Kang, S. A., Chang, J. W., Liu, Q., Zhang, J., Gao, Y., et al. (2009). An ATP-competitive mammalian target of rapamycin inhibitor reveals rapamycin- resistant functions of mTORC1. The Journal of Biological Chemistry, 284(12), 8023–8032. http://doi.org/10.1074/jbc.M900301200

Translational Control Protein 80 Stimulates IRES-Mediated Translation of p53 mRNA in Response to DNA Damage. (n.d.). Translational Control Protein 80 Stimulates IRES- Mediated Translation of p53 mRNA in Response to DNA Damage. http://doi.org/10.1155/2015/708158

Valente, L., & Nishikura, K. (2007). RNA binding-independent dimerization of adenosine deaminases acting on RNA and dominant negative effects of nonfunctional subunits on dimer functions. The Journal of Biological Chemistry, 282(22), 16054– 16061. http://doi.org/10.1074/jbc.M611392200

229

Vázquez-Del Mercado, M., Palafox-Sánchez, C. A., Muñoz-Valle, J. F., Orozco-Barocio, G., Oregon-Romero, E., Navarro-Hernández, R. E., et al. (2010). High prevalence of autoantibodies to RNA helicase A in Mexican patients with systemic lupus erythematosus. Arthritis Research & Therapy, 12(1), R6. http://doi.org/10.1186/ar2905

Visa, N., Izaurralde, E., Ferreira, J., Daneholt, B., & Mattaj, I. W. (1996). A nuclear cap- binding complex binds Balbiani ring pre-mRNA cotranscriptionally and accompanies the ribonucleoprotein particle during nuclear export. The Journal of Cell Biology, 133(1), 5–14.

Wei, X., Pacyna-Gengelbach, M., Schlüns, K., An, Q., Gao, Y., Cheng, S., & Petersen, I. (2004). Analysis of the RNA helicase A gene in human lung cancer. Oncology Reports, 11(1), 253–258.

Wilson, K. F., Fortes, P., Singh, U. S., Ohno, M., Mattaj, I. W., & Cerione, R. A. (1999). The nuclear cap-binding complex is a novel target of growth factor receptor-coupled signal transduction. The Journal of Biological Chemistry, 274(7), 4166–4173.

Wolfe, A. L., Singh, K., Zhong, Y., Drewe, P., Rajasekhar, V. K., Sanghvi, V. R., et al. (2014). RNA G-quadruplexes cause eIF4A-dependent oncogene translation in cancer. Nature, 513(7516), 65–70. http://doi.org/10.1038/nature13485

Worch, R., Niedzwiecka, A., Stepinski, J., Mazza, C., Jankowska-Anyszka, M., Darzynkiewicz, E., et al. (2005). Specificity of recognition of mRNA 5' cap by human nuclear cap-binding complex. RNA (New York, N.Y.), 11(9), 1355–1363. http://doi.org/10.1261/rna.2850705

Yamasaki, S., & Anderson, P. (2008). Reprogramming mRNA translation during stress. Current Opinion in Cell Biology, 20(2), 222–226. http://doi.org/10.1016/j.ceb.2008.01.013

Yamasaki, Y., Narain, S., Yoshida, H., Hernandez, L., Barker, T., Hahn, P. C., et al. (2007). Autoantibodies to RNA helicase A: a new serologic marker of early lupus. Arthritis and Rheumatism, 56(2), 596–604. http://doi.org/10.1002/art.22329

Yanagiya, A., Svitkin, Y. V., Shibata, S., Mikami, S., Imataka, H., & Sonenberg, N. (2009). Requirement of RNA binding of mammalian eukaryotic translation initiation factor 4GI (eIF4GI) for efficient interaction of eIF4E with the mRNA cap. Molecular and Cellular Biology, 29(6), 1661–1669. http://doi.org/10.1128/MCB.01187-08

230

Yu, Y., Sweeney, T. R., Kafasla, P., Jackson, R. J., Pestova, T. V., & Hellen, C. U. (2011). The mechanism of translation initiation on Aichivirus RNA mediated by a novel type of picornavirus IRES. The EMBO Journal, 30(21), 4423–4436. http://doi.org/10.1038/emboj.2011.306

Yuan, Z., Sanders, A. J., Ye, L., & Jiang, W. G. (2010). HuR, a key post-transcriptional regulator, and its implication in progression of breast cancer. Histology and Histopathology, 25(10), 1331–1340.

Zhang, S., & Grosse, F. (1997). Domain structure of human nuclear DNA helicase II (RNA helicase A). The Journal of Biological Chemistry, 272(17), 11487–11494.

Zhang, Z., Yuan, B., Lu, N., Facchinetti, V., & Liu, Y.-J. (2011). DHX9 pairs with IPS-1 to sense double-stranded RNA in myeloid dendritic cells. Journal of Immunology (Baltimore, Md. : 1950), 187(9), 4501–4508. http://doi.org/10.4049/jimmunol.1101307

Zhong, X., & Safa, A. R. (2007). Phosphorylation of RNA helicase A by DNA-dependent protein kinase is indispensable for expression of the MDR1 gene product P- glycoprotein in multidrug-resistant human leukemia cells. Biochemistry, 46(19), 5766–5775. http://doi.org/10.1021/bi700063b

Zhou, K., Choe, K.-T., Zaidi, Z., Wang, Q., Mathews, M. B., & Lee, C.-G. (2003). RNA helicase A interacts with dsDNA and topoisomerase IIalpha. Nucleic Acids Research, 31(9), 2253–2260.

Zhou, Y., Ma, J., Bushan Roy, B., Wu, J. Y.-Y., Pan, Q., Rong, L., & Liang, C. (2008). The packaging of human immunodeficiency virus type 1 RNA is restricted by overexpression of an RNA helicase DHX30. Virology, 372(1), 97–106. http://doi.org/10.1016/j.virol.2007.10.027

231