View Red Α-HA Affinity Matrix (Sigma), and Diluted to a Final

CHARACTERIZATION OF THE URE2 IRES ELEMENT AND THE

ROLE OF eIF2A IN ITS REGULATION

LUCAS C. REINEKE

Submitted in partial fulfillment of the requirements

for the degree of Doctor of Philosophy

Thesis Advisor: Dr. William Merrick

Department of Biochemistry

CASE WESTERN RESERVE UNIVERSITY

May, 2009 CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

______

candidate for the ______degree *.

(signed)______(chair of the committee)

______

(date) ______

*We also certify that written approval has been obtained for any proprietary material contained therein. Table of contents

LIST OF FIGURES vi

LIST OF TABLES ix

ACKNOWLEDGEMENTS x

LIST OF ABBREVIATIONS xii

ABSTRACT 1

CHAPTER I: Introduction 3

Section A, Regulation of gene expression and function 3

Section B, Translation initiation and factors 5

Section C, Cap-dependent translation initiation 6

Section D, Cap-independent translation initiation 10

Section E, Internal initiation in yeast 14

Section F, The URE2 IRES element 16

Section G, Eukaryotic initiation factor 2A and its role in internal initiation 20

Section H, Research plan 25

CHAPTER II: Materials and methods 29

Section A, Yeast strains and culture conditions 29

Section B, Plasmids and cloning 30

Section C, Reporter assays 32

Section D, RT-PCR analysis 32

Section E, Structure probing 33

Section F, Identification of eIF2A interacting partners 36

ii Section G, GST-pull-downs 37

Section H, Immunoprecipitations 38

Section I, Western blotting 39

CHAPTER III: Stability of the stem located in the minimal URE2 41

IRES sequence is not important for internal initiation

Section A, The coding region 3’ of the internal AUG contains multiple 41

regulatory sequences

Section B, Upstream sequences are capable of modulating IRES-mediated 50

expression of URE2.

Section C, Internal initiation on mRNAs with combined truncations 52

reflect changes observed with single truncations.

Section D, Structure probing analysis of the URE2 minimal IRES reveals 61

a stem loop structure that contains the initiator AUG.

Section E, The inhibitory element between nucleotides 98-205 does not 66

change the structure of the minimal IRES element.

Section F, Nucleotides within the stem region directly adjacent to the apical 72

loop are important for IRES function

Section H, Conclusions 84

CHAPTER IV: Elucidating the role of eIF2A and eEF1A in cap- 92

independent initiation of the URE2 IRES.

Section A, IRES elements lacking a poly(A) stretch are not targets of 92

eIF2A-mediated suppression

Section B, The eIF2 and eIF2A-mediated pathways compete for delivery 93

iii of initiator met-tRNAi to the ribosome during translation initiation on

the URE2 IRES element

Section C, Several proteins interact with eIF2A 95

Section D, Interactions between eIF2A and eEF1A or Ssb2p are RNA- 103

indpendent.

Section E, Eukaryotic initiation factor1A interacts with eIF2A under 104

physiological conditions

Section F, Truncation analysis indicates that eIF2A interacts with domain 107

3 of eEF1A

Section G, The C-terminus of eIF2A is important for the interaction with 114

eEF1A

Section H, Amino acids of eIF2A important for the interaction with eEF1A 117

are important for IRES-mediated repression

Section I, Conclusions 117

CHAPTER V: Future Directions 129

Section A, What sequences or structures within the URE2 mRNA regulatory 129

elements are important for modulating levels of internal initiation?

Section B, Does eEF1A interact directly with the URE2 IRES element? 130

Section C, What is the role, if any, of the interaction between the URE2 131

IRES element and eEF1A?

Section D, Does PABP promote recruitment of any IRES element to 40S 132

ribosomal subunits?

Section E, Does the presence of eEF1A specify the pathway for met-tRNAi 132

iv delivery (eIF2 or eIF2A)?

Section F, What factors are required for formation of stable 80S preinitiation 133

complexes on the URE2 IRES?

Section G, Are there unidentified proteins that interact with the URE2 IRES 135

element to alter levels of cap-independent translation? How might

those proteins fit into the model for eIf2A-mediated internal initiation?

Section H, Is the interaction between eIF2A and eEF1A important for 137

internal initiation in vivo?

Section I, What specific regions of eIF2A are important for suppression 138

of internal initiation?

Section J, Which regions of eIF2A are responsible for the characteristics 139

that have been previously observed?

Section K, Where is eIF2A located within the cell? 139

Section L, Summary 141

REFERENCES 142

v List of tables

Table 1: Proteins isolated in complex with HA-eIF2A. 101

vi List of figures

Figure 1: Proposed mechanism for cap-dependent initiation of protein 8

synthesis.

Figure 2: Model for IRES-mediated recruitment of the 40S ribosomal 12

subunit.

Figure 3: Two forms of Ure2p exist within the cell. 19

Figure 4: Schematic representation of the p281-4 construct used in these 21

studies.

Figure 5: Internal initiation of translation from the URE2 IRES element in 22

the eIF4Ets strain.

Figure 6: Comparison of eIF2 and eIF2A-mediated delivery of met-tRNAi 24

to the 40S ribosomal subunit for initiation of protein synthesis.

Figure 7: Role of eIF2A in modulation of internal initiation from the URE2 26

IRES element.

Figure 8: URE2 truncations indicate enhancer and inhibitory sequences 44

downstream of the internal AUG codon.

Figure 9: Abundance of mRNA is not responsible for the different levels 48

of -galactosidase activity observed in functional assays of 3’ truncations.

Figure 10: URE2 Truncations indicate an inhibitory sequence upstream of 51

the internal AUG codon.

Figure 11: Abundance of mRNA is not responsible for the different levels 53

of -galactosidase activity observed in functional assays of 5’ truncations.

vii Figure 12: The URE2 minimal IRES element is located between 56

nucleotides 205-309, and the upstream inhibitory element can be

defined to a 25 nucleotide region upstream of the minimal IRES.

Figure 13: Abundance of mRNA is not responsible for levels of - 58

galactosidase observed in functional assays of combined truncations.

Figure 14: The minimal IRES element contains a stem loop structure. 63

Figure 15: The inhibited IRES element does not significantly differ in 68

secondary structure in the region of the minimal IRES element.

Figure 16: Phylogenetic comparison of URE2 IRES sequence reveals little 73

covariation.

Figure 17: Effects of different parts of the URE2 IRES structure on cap- 77

independent translation initiation.

Figure 18: Mutations that stabilize the double stranded region adjacent to 82

the loop and destabilize the region around the internal AUG codon

do not affect activity.

Figure 19: Stability of the URE2 IRES element is not a critical determinant 85

for efficient internal initiation on the URE2 IRES element.

Figure 20: Internal ribosome entry sites containing poly(A) stretches are not 94

subject to eIF2A-mediated suppression.

Figure 21: Eukaryotic initiation factor 2 phosphorylation reveals 97

competition between eIF2 and eIF2A-mediated translation initation.

Figure 22: Several proteins exist in complex with eIF2A. 100

Figure 23: The 2A-HA yeast strain contains an integrated cassette 102

viii containing an HA-tag and a geneticin resistance marker, and

the eIF2A-HA protein is expressed.

Figure 24: Interaction between eIF2A and several proteins can be 105

verified in reciprocal pull-down experiments.

Figure 25: Interactions between eIF2A and either eEF1A or Ssb2p are 108

RNA-independent.

Figure 26: Eukaryotic initiation factor 2A and eEF1A interact under 111

physiological concentrations of the proteins.

Figure 27: Domain three of eEF1A is responsible for the interaction with 112

eIF2A. A, A domain map of eEF1A with color coded domains.

Figure 28: The region between amino acids 460 and 571 of eIF2A is 115

important for the interaction with eEF1A.

Figure 29: Functional analysis of C-terminal eIF2A deletion yeast strains 118

suggest that the region between amino acids 460 and 571 is important

for suppression of URE2 IRES-mediated translation initiation.

Figure 30: Model of competitive pathways for met-tRNAi delivery during 127

IRES-mediated initiation of protein synthesis..

ix Acknowledgements

I would like to thank my wife, Erin, for all the help she has given me during my graduate

studies. She has helped keep me organized and motivated throughout the process. Erin has also continually reminded me that work can be done more efficiently if you take

breaks to do recreational things. Finally, Erin has been a model student throughout her

studies, which has inspired me to try to do good science and use strong analytical thinking to solve problems.

I would also like to thank my parents for helping me learn what it means to be dedicated.

They have pushed me to do my best all the time, and supported me despite bad decisions.

They have also encouraged me to think independently, and maintain personal integrity during every endeavor.

I would like to thank my friends and lab members at Case Western Reserve University for long discussions and suggestions that contributed to my work. I really appreciate

Nasheed Hossain, a Case medical student, who has been a great help by providing me

with an extra pair of hands. I would like to thank all my friends for the great times

outside of the lab that maintained my morale and confidence for when I was in the lab.

I would like to express my appreciation for my boss, Bill Merrick, the father of

translation, was instrumental in my success. He has taught me how to think and write scientifically, and also how to keep having fun in a field that has a lot of bad news. I would also like to thank my committee (Dr. Jeff Coller, Dr. Mark Caprara, Dr. Eckhard

x Jankowski, and Dr. Anton Komar) for valuable discussions and criticism that have

resulted in better work and experimental design. I could not have done it without all your help.

This work was supported by National Institutes of Health Grants GM-68079 and T32

GM-08056.

xi List of abreviations mRNA Messenger RNA eIF2A Eukaryotic initiation factor 2A

IRES Internal Ribosome Entry Site eEF1A Eukaryotic elongation factor 1A

PABP Poly(A) binding protein mTOR Mammalian target of rapamycin m7G 7-methyl guanosine eIF4F Eukaryotic initiation factor 4F (containing eIF4E, eIF4G,

and eIF4A) eIF4E Eukaryotic initiation factor 4E (Part of eIF4F) eIF4G Eukaryotic initiation factor 4G (Part of eIF4F) eIF2 Eukaryotic initiation factor 2 met-tRNAi Initiator methionyl-tRNA

GTP Guanine triphosphate eIF3 Eukaryotic initiation factor eIF4E-BP Eukaryotic initiation factor

GCN2 General control non-derepressible-2 kinase

HRI Heme-regulated inhibitor kinase

PERK PKR-like endoplasmic reticulum kinase

PKR Double-stranded RNA protein kinase eIF4B Eukaryotic initiation factor

HCV Hepatitis C virus

xii CrPV Cricket paralysis virus

ITAF IRES trans-acting factor

USER Untranslated sequence elements for regulation

UTR Untranslated region poly(A) Poly adenine

SD Synthetic dropout

2A-HA yeast Yeast strain containing a C-terminally HA-tagged eIF2A

gene

YPD Yeast peptone dextrose

PCR Polymerase chain reaction

GST Glutathione-S-transferase

DEPC Diethylpyrocarbonate

SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis

xiii Characterization of the URE2 IRES Element and the Role of eIF2A in its Regulation

Abstract

LUCAS C. REINEKE

Translational control is an extremely important way to modulate gene expression in

response to different environmental cues. Gene expression is regulated at the level of translation in response to a variety of stimuli including hypoxia, apoptosis, viral infection, and glucose starvation. Understanding the factors and pathways that feed into the translational apparatus to up- or down-regulate expression from a set of mRNAs may have therapeutic implications in the future. Our laboratory has identified a protein

(eukaryotic initiation factor 2A; eIF2A) that is responsible for repressing translation initiated on the URE2 Internal Ribosome Entry Site (IRES). The URE2 IRES element is located within the URE2 mRNA and directs synthesis of an N-terminal truncation of a longer protein product involved in nitrogen assimilation in the yeast Saccharomyces cerevisiae. This thesis is focused on understanding the mechanism of IRES-mediated initiation by investigating the RNA requirements for IRES-mediated translation on the

URE2 IRES element and the role of eIF2A in this process. The data indicate that the

URE2 minimal IRES element folds into a small stem loop structure. Several nucleotides

within the minimal IRES element are important for efficient IRES-mediated translation,

and the stability of the stem is not a critical factor in modulating translation initiation.

Furthermore, eIF2A interacts with several proteins including eukaryotic elongation factor

1 1A (eEF1A). Despite the important function of eEF1A in the elongation phase of protein synthesis, it may play an important role as a mediator of the repressive effect of eIF2A.

The data suggest that eIF2A may repress IRES-mediated translation by trapping mRNAs containing IRES elements in preinitiation complexes. I propose that at least two pathways exist for delivery of initiator methionyl-tRNA to the ribosome for IRES- mediated initiation of protein synthesis. One pathway involves eIF2A and eEF1A, while the other pathway involves eIF2. These two pathways may compete under normal conditions, while environmental conditions that promotes turnover of eIF2A could lead to higher production of proteins from IRES elements because mRNAs containing those

IRES elements are no longer trapped in preinitiation complexes. This competition is a unique and intricate mechanism for regulation of IRES-mediated protein synthesis in yeast.

2 I. Introduction

A. Regulation of gene expression and function

Gene regulation is an important biological problem for cellular homeostasis.

Misregulation of gene expression and function can have drastic consequences for the cell.

Often inappropriate regulation of protein expression of function results in genetic instability, which can lead to cancer or developmental defects. The Central dogma of molecular biology states that DNA is transcribed into RNA, which is then used as a template for protein synthesis. The process protein synthesis is known as translation.

Several distinct mechanisms of regulation exist at each nodal point of the Central dogma.

Regulation of protein expression and function can be achieved during transcription in the following ways: modulation of the strength of transcription and alterations in promoter usage. Protein expression and function can also be altered after transcription in the following ways: alternative splicing, adjustments in the intensity of translation events, alternative translational start sites, and RNA turnover. Finally, once the protein is produced, activity can be adjusted based on the needs within the cell by posttranslational modifications and varying the rate at which the protein is degraded. It is clear that there are many different ways to adjust the levels and activity of a given protein within the cell.

One example where gene expression is regulated at all levels is evident in the case of the tumor suppressor p53. The p53 protein facilitates transcription of genes that promote

DNA repair, or stimulates cell death if the damage is too severe for the cell to recover.

The activity of this protein is abrogated in approximately 50% of human cancers, which indicates the importance of this protein in maintaining the integrity of the human genome.

3 This p53 gene can be regulated at many different steps. There are at least three different

promoters that are used during transcription, which produce at least two different proteins

(1). Furthermore, levels of p53 transcription can be negatively regulated by members of

the AP-1 family of transcription factors in response to serum, proinflammatory cytokines,

and UV irradiation (2).

After transcription, the C-terminus of the protein is dictated by variable splicing of the

mRNA (1). It is known that the p53 mRNA is stabilized by a protein that promotes

polyadenylation, thereby increasing translation of the p53 mRNA (3). Furthermore, a

RNA element that directs translation of a truncated form of the protein is important in

cell cycle-dependent regulation of protein production (4). After the protein is produced,

many interactions and posttranslational modifications are known to regulate the function

of the p53 protein. For example, interaction with a protein known as MDM2 promotes

p53 protein degradation (5). Addition of the ubiquitin-like molecules SUMO1 and

NEDD8 have both positive and negative effects on p53 function, respectively.

Furthermore, p53 protein can be phosphorylated, acetylated, and methylated with varying

effects on p53 function (6). Finally, intracellular localization is a key aspect of control of

p53 function. Increased nuclear concentrations of p53 are promoted in response to DNA

damage under normal cellular conditions. In contrast, many malignant tumors employ

the strategy of abrogating p53-dependent transcriptional activity by expressing proteins

that that promote cytoplasmic localization by either sequestration or hyperactive nuclear

export (7). Nearly every aspect of regulation of gene expression and function occurs on

the p53 gene, which makes it a good example for the control a cell has over the

4 expression level and activity of a protein. It also highlights the need for many layers of

gene regulation to rapidly and effectively “turn on” a pathway that promotes the upkeep

of an organism’s genome.

B. Translation Initiation and Factors

Translation, which is production of a polypeptide from an mRNA molecule, is an

important mode of regulation of gene expression. Regulation of mRNA translation has

added layers to the control a cell has over gene expression. Translational regulation of many genes involved in cancer, viral infection, and nutritional responses have been described. Several signaling pathways have been shown to feed directly into components of the translational apparatus, and affect the ability of subsets of mRNA molecules to be regulated. For example, the mammalian target of rapamycin (mTOR) is a protein that upregulates translation in response to growth factors and amino acid availability, and whose activity is inhibited by the anti-cancer drug rapamycin. The regulatory effect relies on the ability of this protein to act as a kinase, which phoshorylates other protein targets. Mammalian target of rapamycin has been shown to differentially regulate expression of subsets of mRNAs based on the proportion of each mTOR target that is phosphorylated (8). During viral infection, the use of mRNA structures to circumvent some types of translational regulation is important for viral replication. Some viruses

encode proteases that disrupt normal translation by cleaving factors important for

translation (9,10). While factor cleavage is important for some types of translation, it

does not significantly affect translation of many viruses. Finally, many microbes secrete

small molecules that are known to inhibit translation. This is typically done during

5 nutrient starvation, which induces production of a small molecule that has been shown to

bind and inhibit function of factors involved in translation (11).

It is clear that translation is an important mode of regulating gene expression. Additional

methods of regulating translation will be discussed in other sections of this thesis. The

translation cycle is generally separated into three phases: initiation, elongation, and

termination. Of those, initiation is the best studied, and accepted as the most regulated of

the three phases.

C. Cap-dependent Translation Initiation

There are two general types of translation initiation: cap-dependent and cap-independent

initiation. Cap-dependent initiation relies on a 7-methyl guanosine (m7G) residue at the

5’ end of the mRNA molecule. This structure is recognized by a protein complex known

as eukaryotic initiation factor 4F (eIF4F), a heterotrimeric complex consisting of: eukaryotic initiation factor 4E (eIF4E), which is responsible for cap recognition;

eukaryotic initiation factor 4G (eIF4G), which is a scaffolding protein that bridges the

cap structure with the 40S ribosomal subunit via its interaction with eIF4E; and eukaryotic initiation factor 4A, which is a RNA helicase. After recognition of the m7G

cap complex by eIF4F, the 40S ribosomal subunit, in association with a complex

containing eukaryotic initiation factor 2 (eIF2; α, , and  subunits), the initiating

methionyl tRNA molecule (met-tRNAi), and guanosine triphosphate (GTP; collectively known as the ternary complex), is recruited to the mRNA via an interaction between eIF4G and a ribosome-bound initiation factor known as eukaryotic initiation factor 3

6 (eIF3). Subsequently, the 40S subunit scans to an initiating AUG codon where the 60S

ribosomal subunit joins the poised 40S subunit, GTP is hydrolyzed causing eIF2 to leave

the complex, and the translation cycle enters the elongation phase. While there are

variations of cap-dependent initiation, the common feature is that they all rely on the

recognition of the m7G cap structure (Figure 1; (12,13)).

Cap-dependent initiation is regulated at many different steps. Both mRNA-specific and global regulatory mechanisms have been described. One step that is tightly regulated is the recruitment of the 40S ribosomal subunit to the mRNA (Figure 1; step 3 highlighted in yellow). This is regulated by the binding of eIF4E-binding proteins (eIF4E-BPs) to eIF4E at the same site that is recognized by eIF4G. By blocking eIF4G recruitment, 40S ribosomal subunit loading is prevented (14,15). Eukaryotic initiation factor 4E binding proteins have been described to regulate mRNAs in both global and mRNA-specific ways

(16). Another mechanism for regulation of 40S ribosomal subunit recruitment is by posttranslational modification, which can alter the affinity of regulatory factors for their targets and therefore modulate cap-dependent initiation. One such modification is the phosphorylation of eIF4E, which is known to increase its affinity for the m7G cap, and

thus thought to enhance translation initiation (Figure 1; step 3 highlighted in yellow;

(17,18)). This modification may enhance translation of subsets of mRNAs, while others

remain unaffected.

Figure 1: Proposed mechanism for cap-dependent initiation of protein synthesis. In this scheme, the 40S ribosomal subunit is recruited to mRNAs that have the eIF4F complex

8 prebound to the 5’ m7G cap structure. Subsequently the 40S ribosomal subunit scans to an appropriate AUG codon, and the 60S ribosomal subunit then joins to form an 80S subunit capable of entering the elongation phase of protein synthesis. Eukaryotic initiation factor 1A is thought to facilitate proper selection of the AUG codon.

Eukaryotic initiation factor 2B is a guanine nucleotide exchange factor for eIF2, which facilitates recycling of GDP for GTP for another round of initiation. Finally, eukaryotic initiation factor 5B is thought to enable joining of the 40 and 60S ribosomal subunits during translation initiation. Steps that are known to be highly regulated are indicated with yellow highlighting (this figure was adapted from (12)).

9 One type of global down regulation of cap-dependent translation initiation, which effects

efficiency of delivery of met-tRNAi, is the phosphorylation of the alpha subunit of eIF2

(Figure 1; step 7 highlighted in yellow). This phosphorylation prevents the recycling of

GDP for GTP resulting after AUG codon recognition. As a result, there is a reduced concentration of ternary complex available for initiation events, thus initiation on all mRNA molecules is inhibited proportionally (19,20). The phosphorylation of eIF2α is

carried out by a number of different kinases that are activated in response to many

different environmental stresses including: amino acid starvation (GCN2), heme

deficiency (HRI), accumulation of unfolded proteins in the endoplasmic reticulum

(PERK), and buildup of double stranded RNA molecules (PKR)(21,22). In another mode

of regulation of cap-dependent initiation, eukaryotic initiation factor 4B (eIF4B), which

is a cofactor that greatly stimulates the RNA helicase activity of eIF4A, is a target of

phosphorylation. Phosphorylation of eIF4B results in increased translation initiation. It

is believed that this modification has a global effect on translation initiation. Eukaryotic

initiation factor 4B phosphorylation can be triggered by a variety of stimuli including

serum, insulin, and phorbol esters (23). Both mTOR and protein kinase B have been

implicated in the phosphorylation of this protein, which indicates a role for mitogenic and

nutritional cues in its regulation (24). Both of these global modes of regulation enable

the cell to reduce protein synthesis during unfavorable cellular conditions, and conserve

resources for other processes or for the continuation of protein synthesis when necessary.

10 D. Cap-independent Translation Initiation

Unlike cap-dependent initiation, cap-independent initiation (also known as internal

initiation or IRES-mediated initiation) typically relies on an mRNA sequence or structure

that functions to recruit the 40S ribosomal subunit for initiation. Internal ribosome entry

sites range in size from about 22 nt to around 700 nt (25-28). They are typically located in the 5’ untranslated region of the mRNA directly upstream of the initiating AUG codon.

The sequences required for efficient cap-independent initiation of translation typically fold into stable elements containing significant secondary structure (28-30). Tertiary

RNA interactions are important for efficient internal initiation of some IRES elements such as the c-myc cellular and Hepatitis C Virus (HCV) IRES elements (26,27). The role of sequence downstream of the initiating AUG codon is not frequently examined for the ability to modulate IRES-mediated translation, but there are examples where the downstream sequence is important (e.g. the Giardiavirus and Human Immunodeficiency

Virus-2 IRES elements) (31,32). The IRES element functions in lieu of the 5’ terminal

cap structure to recruit the small ribosomal subunit (Figure 2).

The protein factors that are involved in IRES-mediated initiation vary depending on the

IRES element in question. Some IRES elements require none of the canonical initiation

factors (e.g. the Cricket Paralysis Virus IRES (CrPV) and the Taura Shrimp Virus), while

others require various combinations of the different factors. For example, the HCV IRES

requires eIF2 ternary complex and eIF3 for efficiency initiation, while the

Encephalomyocarditis Virus IRES requires eIF2, 3, eIF4A and eIF4G (33,34).

Figure 2: Model for IRES-mediated recruitment of the 40S ribosomal subunit. In this model the 40S ribosomal subunit and associated factors are recruited directly to the AUG codon. An RNA structure acts in lieu of a cap to bring together factors that are important for internal initiation. This IRES requires all of the standard initiation factors, but many

IRESs do not (this figure is compliments or Dr. Anton Komar at Cleveland State

University).

12 The first IRES element was discovered in the genome of poliovirus (35). The function of the IRES is to efficiently promote protein synthesis during conditions of viral infection in which global protein synthesis has been down regulated (36). Since then, many examples of IRES elements in viral genomes have been described. Additionally, many cellular IRES elements have been identified, although they have typically been significantly less active than the viral IRES elements (13). One dissimilar feature of viral and cellular IRES elements is that the cellular IRES elements tend to form much less

stable secondary structures, although the active form of most IRES elements is not well

understood in vivo (4,27,28). Furthermore, the structure(s) required for recruitment of

relevant trans-acting factors, the structure(s) required for ribosome binding, and those

required for initiation to occur may vary.

Regulation of IRES-mediated translation is different for each IRES element. However,

one mode of regulation common among many IRES elements is the expression of

different trans-acting factors that promote IRES-mediated expression (IRES trans-acting

factors; ITAFs) under various conditions of cellular stress (37,38). This contrasts the

preponderance of posttranslational modifications prominent for cap-dependent translation

regulation. In agreement with the idea that differential expression of ITAFs are

expressed under different cellular conditions, different mRNAs remain associated with

polyribosomes during down regulation of cap-dependent translation (37,39-41). Each

cellular condition or stress would promote expression of different ITAFs that enhance

expression of diverse IRES elements. The types of stresses that are known to activate

IRES-mediated translation of some IRES elements, presumably through expression of

13 different ITAFs, include: viral infection, apoptosis, hypoxia, and glucose starvation

(38,39).

E. Internal Initiation in Yeast

In yeast, IRES-mediated translation has been less well studied as compared to the

mammalian system. In fact, there have been few IRES elements identified, and many are

tainted by data suggesting the presence of cryptic transcription promoters (42-44).

Constructs used to test putative IRES elements usually contain both a hairpin to block

cap-dependent scanning of the 40S ribosomal subunit and a reporter gene. The putative

IRES element is inserted between the hairpin and in frame with the reporter gene.

Cryptic transcription promoters can result in truncated versions of the reporter mRNA that lack the hairpin structure, which results in production of the reporter from cap- dependent initiation of translation. Since IRES-mediated translation is much less efficient than cap-dependent translation, low levels of reporter protein resulting from cryptic promoter activity can be comparable to what would have been produced from an

IRES. As such, controls for experiments investigating the possibility of new IRES elements are very important to omit the possibility of cryptic promoter activity. Thus, it is unclear whether or not IRES elements exist in the mRNAs in which they have been described.

There are some examples of mammalian cellular or viral IRES elements being active in living yeast cells. One example involves the CrPV IRES element, which was shown to be active under conditions of decreased ternary complex concentration (45). Another

14 example involves the CrPV IRES element and two mammalian cellular IRES elements, which were activated when the yeast ribosomal protein S5 was replaced with the human

S5 protein. This switch was hypothesized to increase the size of a pocket on the ribosome (46). This would hypothetically result in an increased affinity of the ribosome for the IRES element, which would position the mRNA for delivery of the initiator methionyl tRNA, and promote conformational changes in the ribosome that would enable the hydrolysis of GTP in the ternary complex.

There are several yeast IRES elements that are well accepted as bona fide IRES elements.

Several of those elements are active during glucose starvation. To date, this is the only example of a group of IRES elements directing initiation of polypeptides involved in the

same pathway. However, the idea of coordinate regulation of protein expression for a

group of genes (posttranscriptional operon) in response to cellular conditions has been

previously described (47). This type of synchronous regulation depends on the presence

of cis-acting sequences (Untranslated sequence elements for regulation; USER codes)

that can bind trans-acting factors, which promote expression under specific cellular

conditions (47). An example of a posttranscriptional operon is the mRNAs responsible

for production of chemokines and receptors that attract leukocytes during inflammation

(48). This posttranscriptional operon relies on a protein known as L13a, which binds a

USER code in the 3’untranslated region (UTR) of mRNAs to shut off translation. The

purpose of this operon has been hypothesized to limit the deleterious effects of

inflammation (48). Likewise, polyadenosine (poly(A)) stretches in the 5’UTR of genes

involved in invasive growth act as the USER code to promote production of proteins

15 necessary during glucose starvation. Production of invasive growth proteins allows the colony to alter its shape to obtain more surface area, thus allowing the cell to maximize

glucose uptake when environmental concentrations are low.

It is thought that regulation of the invasive growth IRES elements relies on the presence

of the poly(A) stretches, which are thought to function by binding poly(A) binding

protein. Following binding to the poly(A) stretches, PABP is suspected to act in lieu of

eIF4E to recruit the 40S ribosomal subunit through eIF4G (39). The idea of initiation

occurring in the presence of poly(A) stretches upstream of the AUG has since been

demonstrated in vitro in the absence of eIF3 and eIF4F (49). However, PABP was not

present in the reactions so it is unclear how the 40S subunit could be recruited to mRNAs

containing poly(A) stretches in the absence of a protein capable of specifically interacting

with polyadenylate. Hence, while there has been much work aimed at identifying yeast

IRES elements, few are well accepted, and the mechanism of internal initiation remains

unclear.

F. The URE2 IRES Element

Another example of a well accepted yeast IRES element exists in the URE2 gene (50).

The protein produced from the URE2 gene, Ure2p, regulates the cellular response to

conditions of nitrogen deprivation. This ability results from interaction of Ure2p with the

transcription factor Gln3p in the cytoplasm. During conditions of low nitrogen, Ure2p

dissociates from Gln3p, which allows Gln3p to enter the nucleus and initiate transcription

of a subset of genes that promote the uptake of nitrogen into the cell and the conservation

16 of nitrogen stores already available (51). Additionally, this protein can form prion

aggregates that are dependent on the N-terminus of the polypeptide, and overexpression

of N-terminally truncated Ure2p can cure yeast of the prion phenotype (52).

Furthermore, prion aggregates of Ure2p are incapable of exerting control of Gln3p, which

indicates the potential for an additional layer of control on the process of nitrogen assimilation (53).

Another feature of Ure2p is that it has structural similarities to glutathione transferase proteins, but it does not act on substrates typically used for assessing glutathione transferase activity (53). However, deletions of URE2 show sensitivity to heavy metals and oxidants including hydrogen peroxide (54). Furthermore, measurement of in vitro peroxidase activity with N-terminally truncated Ure2p (lacking the prion forming domain), fibrillar Ure2p, and wild type protein revealed similar kinetics (55). Finally, a crystal structure of N-terminally truncated Ure2p reveal a ligand binding site, and dimerization properties characteristic of glutathione transferase proteins (52). Since the function in nitrogen assimilation, and not peroxidase activity, is altered when Ure2p forms prion aggregates, it is plausible that the biological advantage is derived from the role of Ure2p in regulating nitrogen assimilation. Therefore, the IRES element that controls production of the N-terminally truncated form of Ure2p may be important in regulating the ability of the cell to respond to fluctuations in environmental nitrogen.

The URE2 IRES element was identified due to the expression of two forms of the protein in the cell (Figure 3; (50)). It was known that there were no gene duplication events that

17 could allow for the production of the two forms. Minimal splicing exists in yeast,

decreasing the chances that the alternate protein product results from an alternative RNA

species. The presence of cryptic promoter activity was ruled out using in vitro translation

extracts where no transcription occurs. Finally, the possibility that the N-terminally

truncated form of Ure2p is a proteolytic product of the full-length protein was dismissed

because mutation of the initiating AUG codon for the full-length protein does not abolish

expression of the truncated protein. These pieces of information all support the

hypothesis that the N-terminally truncated Ure2p results from IRES-mediated translation.

The presence of an IRES element was verified using several different constructs. One

such construct harbors the GAL 1/10 inducible transcription promoter with a stable stem loop upstream of a fusion between the URE2 coding region and the LACZ gene (Figure 4; termed p281-4). The stable stem loop inhibits 40S scanning characteristic of cap- dependent initiation mechanisms. Therefore, -galactosidase is only expressed when translation is initiated from an IRES element. Furthermore, this construct allows in vivo assessment of the ability of the putative IRES element to function. Another method of validating the IRES element utilized a yeast strain harboring a temperature sensitive mutation in eIF4E (eIF4Ets), which abolishes m7G binding at the nonpermissive temperature. As expected, in yeast shifted to the nonpermissive temperature, the p281-4 construct containing URE2 yielded 10-fold more -galactosidase activity as compared to

the same construct at the permissive temperature (Figure 5; (50)). Concomitantly, cap-

dependent translation decreased as expected when m7G cap-binding by eIF4E was

abolished by shifting to the nonpermissive temperature. The enhanced IRES-mediated

Figure 3: Two forms of Ure2p exist within the cell. Gel electrophoresis followed by

Western blotting was used to identify the protein products that arise from the URE2 mRNA. It is clear that both a long and a short form are generated, and it is known that the shorter version results from internal initiation of translation. The short form is an N- terminal truncation of the longer protein which lacks the prion forming domain important in spontaneous aggregation of the protein. It is the product of initiation of translation at codon 94. The full-length protein is 354 amino acids long, and the short form is 260 amino acids in length (this figure is compliments or Dr. Anton Komar at Cleveland State

University).

19 translation is thought to result from the plethora of translational machinery available when cap-dependent translation is reduced. In the original publication, the AUG codon responsible for internal initiation was determined by mutating the predicted AUG codon to a noncoding CTT codon. Upon doing this, IRES activity was completely abolished

(Figure 5; (50)). In total, there are thirteen AUG codons in this gene. Two AUG codons encode the first two amino acids in the longer protein product. The next AUG codon is the ninety-fourth codon in the protein, which is followed by seven more. The ninety- fourth AUG codon is the internally initiated codon during IRES-mediated translation.

Other than these observations, there has been little other data published on this IRES element, and regulation of URE2 IRES-mediated translation is unknown.

G. Eukaryotic initiation factor 2A and its role in internal initiation

Several proteins are known to enhance levels of internal initiation. Some hnRNP proteins alter expression from IRES elements including c-myc, HCV and XIAP (56-58).

Furthermore, polypyrimidine tract binding protein modulates activity of several IRES elements including HCV, Hypoxia Inducible Factor 1α, and the Upstream of N-ras (58-

60). These proteins are thought to modulate IRES activity by either facilitating recruitment of the IRES to the ribosome via protein:protein interactions or by rearranging the structure of the IRES element so that ribosome binding is more favorable.

One protein that was found to suppress the URE2 IRES element is known as eIF2A, which is not to be confused with eIF2α. This protein is conserved from yeast to humans, and contains a single identifiable domain, known as a seven bladed β-propeller. This

AUG280 lacZ GAL

URE2

Figure 4: Schematic representation of the p281-4 construct used in these studies. This

construct is characterized by the presence of a GAL1/10 UAS, a stable hairpin previously

shown to block cap-dependent scanning (61), and an URE2-LACZ fusion that enables

measurement of internal initiation using -galactosidase activity. The AUG280 (dubbed

AUG280 because the adenosine of this codon is the 280th nucleotide in the coding region

of the longer Ure2p; also named AUG94 because it is the ninety-fourth codon in the full-

length Ure2p protein) refers to the AUG codon that is used for internal initiation of

translation. The blue and orange highlight the coding regions for Ure2p and -

galactosidase, respectively.

Figure 5: Internal initiation of translation from the URE2 IRES element in the eIF4Ets strain. The p281-4 plasmid was transformed into the eIF4Ets strain, and -galactosidase activity was measured. As controls, the internal AUG codon was mutated to a noncoding

CTT, and a cap-dependent reporter lacking the stable stem loop and the URE2 coding region was monitored (62).

22 domain is known to facilitate protein:protein interactions, although its function in the

context of eIF2A is unclear (62). Eukaryotic initiation factor 2A was originally

discovered for its ability to stimulate in vitro binding of the initiator methionyl tRNA to

the ribosome in an AUG codon-dependent manner (63). This contrasts the same function

of eIF2, which conducts the same reaction in a GTP-dependent manner (Figure 6; (62).

Eukaryotic initiation factor 2A was discovered based on similarity with the bacterial

system, which delivers the met-tRNA in an AUG-dependent manner. Although

functional conservation was anticipated, it was later determined that the mechanism

involving eIF2 and GTP-dependent delivery of met-tRNAi is favored in eukaryotes (62).

In fact, it has been determined that eIF2 is both more efficient in stimulating cap-

dependent translation, and more abundant in rabbit reticulocyte cell lysate. Since the discovery of eIF2A, the protein has not been studied extensively, and the role of its met- tRNAi delivery properties in biology remains unknown.

Since the role of eIF2A has remained elusive, it was tested for its ability to stimulate

IRES-mediated translation in yeast. Contrary to the prediction, it was found that eIF2A suppresses internal initiation of the URE2 IRES element, and the effects of eIF2A are not as profound for cap-dependent reporter mRNAs (Figure 7A). Interestingly, a fraction of the eIF2A protein in the cell is associated with 40S ribosomal subunits although most is found in 80S complexes (Figure 7B; (62)). One hypothesis about the origin of the

Figure 6: Comparison of eIF2 and eIF2A-mediated delivery of met-tRNAi to the 40S

ribosomal subunit for initiation of protein synthesis. Left, Pathway for delivery of met-

tRNAi by eIF2. Note that this protein is composed of 3 subunits, as addressed in the text, which delivers the initiator tRNA molecule to the ribosome in a GTP-dependent manner.

Right, Pathway for delivery of met-tRNAi by eIF2A. Unlike eIF2, this single

polypeptide delivers the initiator tRNA to the ribosome in an AUG-dependent way,

similar to the bacterial system. Each pathway culminates in the formation of the same

80S preinitiation complex (this figure was adapted from (62)).

24 suppressive effect is that the URE2 mRNA molecule is trapped in early initiation

complexes. Therefore, in eIF2A deletion cells, the inability to trap mRNAs in

preinitiation complexes allows the mRNA to be translated. However, this hypothesis has

not been directly tested. Based on the lack of recognizable RNA binding motifs, it is not

expected that the suppressive effect results from direct binding of eIF2A to the mRNA molecules.

Saccharomyces cerevisiae have been accepted as a pliable genetic organism within the scientific community because of the development of relatively simple genetic

manipulations. One of the shortfalls of yeast for the translation field has been the lack of

IRES elements, and the subsequent inability to characterize the process. Now that a

system exists in which internal initiation can be studied, yeast can be used to delineate the

process in much more detail. Furthermore, the role of eIF2A in internal initiation and

actin polymerization can be studied because of the viability of knockout strains.

H. Research Plan

There are two sets of questions that I sought to address in this thesis. The first has

focused on the characteristics of the URE2 IRES element that facilitate internal initiation.

The second centers around identification of factors that may be responsible for mediating

internal initiation in yeast. In studying this set of questions, I uncovered many things

about the process of internal initiation and the role of eIF2A.

25 A. B.

Figure 7: Role of eIF2A in modulation of internal initiation from the URE2 IRES element. A, The p281-4 construct was used to measure internal initiation in both wild-

type and eIF2A cells. As controls, the internal AUG codon was mutated to a noncoding

CTT codon, and a cap-dependent reporter lacking the stable stem loop and the URE2

sequence was used. B, Sucrose gradient ultracentrifugation was used to separate translation complexes based on size in wild-type cells transformed with a plasmid

directing overexpression of HA-tagged eIF2A protein. After fractionation of the sucrose

gradient of cell lysates, fractions were Western blotted with antibodies against eIF2A.

The fractions containing the 40S, 60S, and 80S ribosomes are highlighted (62).

26 To address the first set of questions, I began by investigating the minimal sequence and

structure of the URE2 mRNA that was required for efficient internal initiation. This was

done using the aforesaid p281-4 construct. Both a truncation analysis and structure

probing were employed to characterize the minimal URE2 IRES element. I also defined the structure of the minimal IRES element in the context of an inhibitory element directly upstream. Subsequently, site-directed mutagenesis and additional deletions allowed delineation of the nucleotides that participate in IRES-mediated translation in more detail.

I also conducted a limited deletion analysis of the upstream inhibitory element that is capable of modulating internal initiation of the minimal URE2 IRES element.

The second set of questions was addressed by investigating the proteins that interact with eIF2A; the rationale was that the proteins interacting with eIF2A may also effect internal initiation. As such, I identified several interacting proteins that had not previously been discovered during high-throughput proteomic screens. Several of these proteins also interact in an RNA-independent manner. Subsequently, I focused on eEF1A, and delineated the regions of eEF1A and eIF2A responsible for the interactions. I then attempted to use various deletions and mutants to understand the functional relevance of the interaction on internal initiation.

Previous reports suggested a PABP-dependent pathway for internal initiation, but results of these studies indicate that there are at least two distinct pathways (39). This suggestion originates from the analysis of PABP-dependent IRESS elements in functional assays in both wild type and eIF2A cells. Based on the results that suggest eEF1A and

27 eIF2A interact and those suggesting two separate pathways exist, I proposed a model

wherein the eIF2-dependent (and possibly PABP-dependent) pathway competes with the

eIF2A-dependent pathway (along with eEF1A). To test this hypothesis, I examined the

effect of reductions in ternary complex using either GCN2 mutants that render the protein

constitutively active (which results in eIF2 phosphorylation and inactivation) or yeast strains deleted for two of the four copies of initiator met-tRNA genes.

28 II. Materials and methods

A. Yeast Strains and Culture Conditions

Yeast used for these studies were of the BY4741 background (Research Genetics). For

all reporter assay experiments, wild type BY4741 (MATa, his3-1, leu2-0, met15-0, ura3-

0), ΔeIF2A knockout strains (MATa, his3-1, leu2-0, met15-0, ura3-0, ygr054::KanMX),

or full-length eIF2A-HA and C-terminal eIF2A truncations (MATα, his3-1, leu2-0,

met15-0, ygr054w::YGR054W-HA KANMX6) were inoculated from fresh synthetic

dropout (SD) plates and grown overnight at 30o C in 5 ml SD medium containing histidine, methionine, leucine and 2% glucose. After 20-22 h of growth, yeast were spun down at 3,000 rpm for 5 min and washed with 10 ml sterile deionized water. Yeast were then resuspended in 5 ml SD medium containing the aforementioned amino acids and 2% galactose for induction of mRNA expression. Yeast were incubated with galactose at 30o

C for 18-20 h before harvesting.

Culture conditions for identification of eIF2A-interacting partners were as follows

(Methods, Section F). Cultures of eIF2A knockout yeast transformed with either empty

pTB328 or pTB328_y_2A plasmid (64) were grown overnight in SD medium containing

histidine, methionine, uracil (SD-leu) and 2% glucose at 30oC. The cells were then

washed with water and resuspended in SD-leu medium supplemented with 2% galactose,

and cells were grown overnight at 30oC. Subsequently, cells were harvested at an OD

600 of approximately 2.5 and processed prior to analysis for eIF2A interacting partners.

For yeast whole cell lysates used in pull-down experiments (Methods, Section G and H),

a 50 ml starter culture of either eIF2A-HA wild type (2A-HA yeast) or C-terminal

29 deletion yeast was grown overnight in yeast peptone dextrose (YPD) medium at 30oC.

The starter culture was diluted to approximately 5X106 cells/ml in 1 L of YPD medium

and cells were grown at 30oC until the OD600 reached approximately 2.0 (about 4 h).

Subsequently, cells were pelleted and lysates were prepared.

B. Plasmids and Cloning

The yeast plasmid p281-4 harboring a GAL1/10 promoter, a stable stem loop, and a β-

galactosidase reporter gene (61) was utilized for URE2 truncation analyses and

experiments comparing the invasive growth IRES elements described by Gilbert et al.

(39). Each URE2 truncation or invasive growth IRES element was amplified from

pBIISK+(URE2) plasmid or pT7-stem_Fluc(pA)62 plasmids containing the IRES (39,50),

respectively, by polymerase chain reaction (PCR). Each PCR fragment was then

subcloned into the XhoI and EcoRI sites of the p281-4 vector using standard cloning

procedures. Approximately 400 ng of sequenced p281-4 plasmids were then transformed

into both BY4741 and BY4741 ∆eIF2A yeast strains for β-galactosidase measurements

using standard lithium acetate methods (65). For structure probing, NdeI, BamHI sites and the T7 promoter were added flanking the URE2 coding region during PCR and subcloned into the NdeI and BamHI sites of pUC19.

For point mutants of URE2, mutations were generated using the pBIISK+(URE2) plasmid as a template in PCR with Pfu-turbo. PCR reactions were then digested for 1 h with DpnI, and transformed directly into E. coli DH5 cells. Each clone was then sequenced for the presence of the respective mutations. Following positive identification

30 of mutant plasmids, mutant URE2 sequences were amplified and cloned into pUC19 as

described above.

For p281-4-Firefly constructs, URE2 and Firefly luciferase were amplified from pBIISK+(URE2) and pGL2-Basic plasmids, respectively, and gel purified. Homologous ends were added to the primers for both the URE2 and luciferase reactions. The p281-4

plasmid was digested with XhoI and BssHII, which cuts within the -galactosidase

coding region, and gel purified. Approximately 60 ng of the Firefly luciferase and URE2

truncation PCR products were then transformed with 36 ng of digested p281-4 plasmid

into BY4741 yeast. Plasmids were recovered from the yeast and sequenced prior to

conducting luciferase assays.

For glutathione-S-transferase (GST)-fusion constructs, the coding region of DED1,

NPL3, SSB2, and TEF1 (eEF1A gene) were amplified from yeast genomic DNA by PCR.

For TEF1, SSB2, and NPL3, BamHI and XhoI restrictions sites were added during PCR, while EcoRI and XhoI sites were added to DED1. Deoxynucleic acid fragments were digested with the enzymes cleaving at the aforementioned sites and cloned into pGEX-

6p-1 in frame with the N-terminal glutathione-S-transferase protein to produce GST-

DED1, GST-NPL3, GST-SSB2, and GST-TEF1. Each Construct was verified both by sequencing and checking expression levels in BL(21)-DE3 E. coli. Mutants of eEF1A were either subcloned from plasmids containing mutant eEF1A, as described previously

(66,67), or generated de novo using site-directed mutagenesis (described above) using

pGEX-6p-1 GST-TEF1 as a template.

C. Reporter Assays

Spectrophotometric β-galactosidase assays were performed as described previously with the following modifications (50): two colonies were selected from each transformation and assayed in duplicate. In addition, optical densities at 420 and 600 nm were analyzed in flat, clear bottom 96 well plates (Corning) using a Spectramax M2 Plate Reader

(Molecular Devices). These experiments were repeated at least three times.

Preparation of yeast cell lysates for firefly luciferase measurements was performed as described previously (68). Briefly, 1.5 ml culture was washed in PBS and resuspended in

150 l reporter lysis buffer (Promega). 50-100 l glass beads were added to each sample and vortexed three times for 1 min each. Cell debris were removed and the supernatant was utilized for activity measurements. Firefly luciferase activity was measured using a

Lmax luminometer (Molecular Devices).

D. RT-PCR Analysis

Total RNA was isolated after induction with galactose for each construct using the

MasterPure Yeast RNA Purification Kit (Epicenter) and quantified by UV spectrophotometry. Subsequently, RNA was analyzed using primers that amplified a region of the mRNA encompassing a 1500 nt region of LACZ with PGK1 as an internal control. RT-PCR was conducted using a one-step RT-PCR protocol in which an enzyme mix of Superscript II RT and Platinum Taq Polymerase (Invitrogen) were mixed to a final concentration of 12.5 U/l and 2.5 U/l, respectively, in dilution buffer (20 mM Tris-

32 HCl, pH 7.5, 1 mM DTT, 0.01% NP40, 0.1 mM EDTA, 0.1 M NaCl and 50% glycerol).

The reactions were then assembled in 2X reaction buffer (50 mM Tris-HCl, pH8.35, 1

mM MgCl2, 60 mM KCl and 0.002 mM DTT), 1 l enzyme mix, 50 pmol of each primer

and 100 ng total RNA. Each sample was analyzed in triplicate with primers targeting

LACZ and once with PGK1 primers. Reactions containing LACZ or PGK1 primers were run for 25 or 20 cycles, respectively. After amplification, gel electrophoresis was conducted with 1% agarose gels to resolve the products and stained for 15 min in the presence of ethidium bromide followed by two washes in deionized water for 5 min.

Data was quantified using ImageQuant 5.0. These experiments were repeated at least

twice.

To analyze RNA content of GST-pulldown reactions after incubation with RNaseA

(Methods, Section G), 50 µl of supernatant from each reaction was removed after the 1

hour incubation with RNaseA and treated with 50 U DNaseI for 30 min at 37oC. After

DNaseI treatment, reactions were extracted with an equal volume of phenol:chloroform

and ethanol precipitated according to standard procedures. After resuspending the

nucleic acid pellet, the RNA solution was analyzed by multiplex RT-PCR (described

above) using primers directed against both the 18S and 25S rRNA. Twenty cycles were used to amplify a 100 nt region of 18S rRNA and 200 nt region of 25S RNA. Products of

RT-PCR were analyzed by 1% agarose gel electrophoresis.

33 E. Structure Probing

RNA was in vitro transcribed using T7 RNA polymerase in the presence of 1 mM CTP,

UTP and ATP, 0.5 mM GTP and 1.32 mM guanosine in a final volume of 100 l for 1.5

h at 37o C. The reaction was then phenol:chloroform extracted and run through a

Sephadex G50 column to desalt the sample. RNA was precipitated with ethanol and

resuspended in 11 l RNase-free H2O. T4 polynucleotide kinase was then used in the

presence of 7000 Ci/mmol [γ-32P]ATP to end label the RNA in a final volume of 20 l

o for 45 min at 37 C. Thirty microliters of RNase-free H2O were then added to each

reaction and the reactions were extracted with phenol:chloroform. Extracted samples

were desalted with a Sephadex G50 column and run on a 7 M urea 6% 19:1

acrylamide:bis-acrylamide gel. Bands corresponding to the correct molecular weight of

RNA were then excised from the gel and eluted overnight in 520 l ATE buffer (300 mM

o NaCH3CO2, pH5.4, 10 mM Tris-HCl, pH 7.5 and 1 mM EDTA, pH 8.0) at 4 C. Four

hundred fifty microliters of ATE buffer was phenol:chloroform extracted and the RNA

was precipitated with ethanol in dry ice for 1 h. The precipitate was pelleted at 4o C at

13,000 rpm for 15 min. The pellet was dissolved in 100 l ATE buffer and precipitated again in dry ice for 1 h followed by another spin at 4o C at 13,000 rpm for 15 min. The

pellet was then dried and dissolved in 50 l nuclease-free H2O. Radioactivity was then

measured by scintillation spectroscopy and RNA concentrations were determined using a

spectrophotometer.

Prior to structure probing, RNA solutions were heated to 90o C and quick cooled on ice.

Labeled RNA (2x105 cpm) was then placed in reaction buffer (10 mM HEPES-KOH, pH

34 7.5, 100 mM KCl) minus and plus magnesium chloride (2.5 mM MgCl2) and allowed to

fold at 30o C for 10 min prior to adding RNases. After folding, 2 l RNase dilutions of

T1 (Ambion 1 U/l), and V1 (Pierce 911 U/ml or Ambion 100 U/ml) were added to the reaction and allowed to digest for 1 min. Reactions were stopped with 3 l of 5 mg/ml aurin tricarboxylic acid. In parallel, controls lacking RNase were conducted using 2 l

RNase-free H2O instead of RNases. One microliter of 0.5 M EDTA, pH 8.0 was added to

each sample followed by 19 l 2X colorless urea loading dye. Seven microliters of each

reaction were then loaded onto 7 M urea 6% 19:1 acrylamide:bis-acrylamide sequencing

gels to resolve fragments. Data was then compared to RNA structures predicted with the

MFOLD server (http://frontend.bioinfo.rpi.edu/applications/mfold/cgi-bin/rna-form1-

2.3.cgi) at 30o C and double stranded regions of the structure prediction that did not agree

with the data were broken apart (69,70).

Diethyl pyrocarbonate (DEPC) probing was performed as described above, but with the

following modifications. Instead of adding RNases, 2 l DEPC (Acros Organics 97%)

was added to reactions only in the presence of magnesium and mixed with a vortex

mixer. In parallel, reactions without DEPC were conducted with RNase-free H2O. After a 10 min incubation at 30o C, 1 l 10 mg/ml yeast tRNA was added to the reactions.

RNA was then phenol:chloroform extracted after adding 1 l 0.5 M EDTA, pH 8.0 and precipitated. RNA was then dissolved in 20 l aniline-acetate solution (10 l aniline

(Sigma 99%), 83 l RNase-free H2O and 6 l glacial acetic acid). Reactions were

incubated at 60o C in the dark for 15 min followed by ethanol precipitation. Pellets were

35 dissolved in 10 l 1X colorless urea loading dye and 3 l were utilized for gel

electrophoresis. Each structure probing experiment was repeated at least twice.

F. Identification of eIF2A interacting partners

For identification of eIF2A-interacting proteins, cells (Methods, Section A) were washed

in NETK Buffer (20 mM Tris-HCl, pH8.0, 100 mM KCl, 1 mM EDTA, 10% glycerol, 1

mM DTT, and 0.1% NP-40) and resuspended in 0.5 ml NETK Buffer containing protease

inhibitors (1 µg/ml pepstatin, 0.2 mM phenylmethylsulphonyl floride, and 10 µg/ml of

each aproptinin and leupeptin). Acid washed glass beads (~250 l) were then added to

the cell resuspension and vortexed for three cycles of 3 min. with 3 min. on ice in between. Cell debris was then pelleted at 4750 rpm at 4oC. The supernatant was

precleared with 50 µl volume of NETK-equilibrated protein A Sepharose 4 fastflow (GE

Healthcare) for 1 h at 4oC with rotating. Subsequently, the protein A sepharose beads

were pelleted at 3,000 rpm, and the protein concentration of the supernatant was assessed

using a Bradford assay. Five hundred micrograms of total protein was added to 25 µl

NETK-equilibrated EZview red α-HA affinity matrix (Sigma), and diluted to a final

volume of 500 µl prior to overnight incubation at 4oC with rotation. Subsequently, the α-

HA affinity matrix was washed three times with NETK buffer for 5 min at 4oC. Finally,

HA-tagged protein was eluted twice from the α-HA affinity matrix by incubating for 1 h

at room temperature in 0.2 mg/ml HA peptide in radioimmunoprecipitation assay buffer

(1X phosphate-buffered saline, 1% NP-40, 0.5% sodium deoxycholate, and 0.1% SDS) .

The protein obtained in each eluate was precipitated by trichloroacetic acid precipitation

prior to resolving by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis

36 (SDS-PAGE). Bands specific to the reactions containing HA-eIF2A protein were

subjected to trypsin digestion followed by LC-tandem mass spectrometry using a

ThermoFisher LTQ ion trap mass spectrometer (Conducted by Dr. Michael Kinter at the

Cleveland Clinic Foundation).

G. GST-pull-downs

BL(21)DE3 E. coli containing various GST-fusion constructs were grown in 250 ml LB

broth. After reaching an OD 600 of 0.6, GST-fusion protein expression was induced by

adding isopropyl--D-thiogalactopyranoside to a final concentration of 1 mM. Proteins were expressed for 3 h at 30oC and pelleted at 14,000 rpm for 30 min. Cell pellets were then dissolved in 5 ml solution A (50 mM Tris-HCl, pH 8.0, 100 mM KCl, 1 mM DTT, 5 mg lysozyme, and protease inhibitor mix) and incubated 30 min. on ice in the presence of

5 mg lysozyme. Cell resuspensions were then sonicated, cell debris were pelleted, and the supernatant was added to approximately 100 µl of NETK-equilibrated glutathione sepharose 4B (GE Healthcare). After an incubation of approximately 2 h at 4oC with

rotation, GST-fusion protein-conjugated resin was washed three times with NETK and

stored at 4oC for future use.

Pull-down experiments were conducted using either wild type eIF2A-HA or C-terminal

eIF2A-HA deletion yeast cell lysates (culture conditions described in Methods, Section

A). Yeast cell lysates were produced by washing cells in 10 ml NETK buffer once and

then resuspending the cells in 25 ml NETK buffer containing protease inhibitors.

Subsequently approximately 7 ml of acid washed glass beads (Sigma) were added to the

37 resuspension and lysed in three cycles of 3 min. of vortexing followed by 3 min. on ice.

Cell debris was pelleted at 3,000 rpm at 4oC, and the protein concentration in the supernatant was quantified using Bradford assays. Finally, cell lysates were aliquoted and stored at -80oC for later use.

The GST-pull-down reactions were performed as previously described (71). Briefly, 10

µl volume of GST-fusion-conjugated resin was incubated with 140 µg total protein for approximately 2 h in a reaction volume of 500 µl at 4oC. Reactions were then washed three times with NETK buffer before adding 10 µl Laemmli sample buffer. Reactions were then analyzed by 12% SDS-PAGE followed by Western blotting (Methods, Section

I). Experiments examining the RNA-dependence of interactions were done as described above, except 1 ml reactions were prepared and incubated 3 h at 4oC with rotation.

Subsequently, reactions were split in half, and 10 µg RNaseA was added to one of two reactions. Reactions were incubated for an additional hour at 4oC before analysis by

SDS-PAGE.

H. Immunoprecipitations

For endogenous pull-downs of eIF2A, 500 µg protein from 2A-HA yeast lysate in a 500

µl reaction volume (Methods, Section G) was pre-cleared with 30 µl of protein A sepharose resin for 1 h at 4oC with rotating. Subsequently, the supernatant was removed, and 5 µl mouse monoclonal α-HA antibody (Sigma) or 10 µl α-3-phosphoglycerate kinase (Invitrogen) was used to immunoprecipitate eIF2A-HA or 3-phosphoglycerate kinase, respectively, for 2 h at 4oC. After this incubation period, NETK-pre-equilibrated

38 protein A Sepharose resin was added to each sample and incubated for an additional 2 h at 4oC with rotation. The resin was then pelleted and washed in 0.5 ml NETK buffer

three times followed by addition of 10 µl Laemmli buffer. Immunopellets were then

analyzed by 10% SDS-PAGE followed by Western blotting (Methods, Section I).

Reciprocal pull-down experiments of eEF1A were conducted with 500 µg 2A-HA yeast

lysate in a 500 µl reaction volume, which was precleared with protein A Sepharose resin,

as described above. Two microliters of either α-eEF1A antiserum or preimmune serum

were added to each reaction and reactions were incubated overnight at 4oC with rotation.

Ten microliters of protein A Sepharose resin was then added to each reaction followed by incubation for 2 h at 4oC with shaking. Reactions were then washed three times with

NETK buffer and 10 µl Laemmli buffer was added to the resin. Protein complexes were

then analyzed by 10% SDS-PAGE and Western blotting (Methods, Section I).

Finally, pull-down experiments examining the interaction between Kem1p and eIF2A were conducted as described above except only 140 µg 2A-HA yeast lysate was used.

Reactions were analyzed by 10% SDS-PAGE followed by Western blotting (Methods,

Section I).

I. Western blotting

Western blotting was conducted according to standard procedures. Membrane-bound eIF2A-HA protein was detected after SDS-PAGE (Methods, Section G and H) using a

1:1000 dilution of α-HA-horseradish peroxidase-conjugated antibody (α-HA-HRP;

Roche). Western blot detection of eEF1A protein (Methods, Section H) was conducted

39 using rabbit α-eEF1A antiserum (Compliments of Terri Kinzy at UMDNJ Robert Wood

Johnson Medical School) at 1:5000 dilution. Membranes from Kem1p immunoprecipitation experiments (Methods, Section H) were probed with either 1:2000 of α-Kem1p antiserum (Compliments of Arlen Johnson at UT Austin) or 1:1000 α-HA-

HRP antibody. Secondary detection was achieved using either goat -rabbit-HRP or goat

-mouse-HRP-conjugated antibody (Santa Cruz) and chemiluminescent ECL reagents as described by the manufacturer (GE Healthcare).

40 III. Stability of the stem located in the minimal URE2 IRES sequence is not important for internal initiation

A part of this chapter has been published in the Journal of Biological Chemistry (72).

In order to understand the mechanism of internal initiation on the URE2 IRES element, it is important to know the minimal sequence and secondary structure within it. These pieces of information are critical to understanding the trans-acting factors and interactions with the ribosome necessary for efficient internal initiation. Furthermore, cis-sequences that are integral for modulation of internal initiation on the URE2 IRES element are an interesting feature of IRES-mediated translation that have recently been recognized as a new way of regulating cap-independent translation. These studies will provide a foundation for identification of new factors and IRES elements using the powerful genetic manipulations that have been developed for yeast; thus permitting a more comprehensive picture of the process of internal initiation. In this chapter, I sought to determine the minimal sequence, and the importance of RNA structure in IRES- mediated initiation of translation on the URE2 IRES element. This was done using a mutational analysis and structure probing, as mentioned in chapter one. These data were used to produce a model for the minimal URE2 IRES structure. I tested this model by making site-directed deletions and mutations to assess the importance of different nucleotides for internal initiation of translation.

41 A. The coding region 3’ of the internal AUG contains multiple regulatory sequences

Upon identification of the first yeast IRES element (50), I sought to determine the minimal sequence within the URE2 mRNA that maintains the ability to promote internal initiation. This was done using the p281-4 plasmid described (Figure 4). By inserting

URE2 truncations in frame between the stable stem loop and the LACZ coding region, the only protein produced results from internal initiation, as shown previously (50,62). Thus, the use of in frame truncations allowed us to elucidate the minimal IRES sequence by measuring -galactosidase activity in yeast lysates. In parallel, the eIF2A-mediated suppression of internal initiation was examined by assaying each truncation in wild type and eIF2A yeast. Since a small number of other yeast IRES elements have been identified, and this is one of the few examples of an IRES residing within a coding region of a protein, there were not many common motifs or minimal IRES lengths that could be used for comparison. For that reason, I began by making truncations of regions 3’ to the internal AUG.

The results obtained revealed that there may be several elements that are capable of modulating internal initiation of URE2 in cis. For example, a putative enhancer element exists between nucleotides 812 and 1061 of the URE2 coding region. This enhancer sequence was identified because deletion of the region between 812 and 1061 results in low levels of -galactosidase activity (Figure 8A; compare 7-1061 to 7-812). Another potential cis element exists between nucleotides 309 and 497, which is capable of inhibiting URE2 IRES-mediated expression (Figure 8A).

42 Since each truncation resulted in a different -galactosidase fusion protein, it was possible that differences in the specific activity of -galactosidase were responsible for the variations in activity observed (Figure 8A). Therefore, I attempted to use Western blotting to determine the relative levels of the fusion proteins. As has been observed by others, I was unable to sensitively and reliably detect -galactosidase fusion proteins from yeast cell extracts (73). I then attempted to generate constructs in which the - galactosidase coding region was replaced with firefly luciferase for each 3’ URE2 truncation. The rationale for producing these constructs was that if the URE2 fragment was affecting the specific activity of -galactosidase, then it would not be expected to affect the specific activity of luciferase in the same way due to the structural differences between the two reporter proteins. When the luciferase activity obtained for each 3’ truncation was compared with the -galactosidase activity, it was evident that there were no differences that could be attributed to changes in the specific activity of - galactosidase (Figure 8B). The activity for each truncation mirrored the activity obtained for each -galactosidase measurement, suggesting that cis-acting sequences are responsible for modulating levels of internal initiation on the URE2 IRES and not the specific activity of the reporter gene.

43 A.

44 B.

Figure 8: URE2 truncations indicate enhancer and inhibitory sequences downstream of

the internal AUG codon. A, -galactosidase assays using the p281-4 reporter construct,

which harbors a stable stem loop designed to inhibit cap-dependent scanning of the 40S

ribosomal subunit, were used to determine the importance of the deleted URE2 regions.

Left, nucleotide numbers and schematic representations of the different URE2 mRNA truncations. Right, activity measurements of the respective URE2 mRNA truncations in both BY4741 (wild type; gray bars) and BY4684 (eIF2A; hatched bars) yeast cells.

Activity measurements for each truncation containing a CTT mutation in place of the internal AUG are shown for wild type (black bars) and eIF2A cells (dotted bars). B, A new p281-4 construct (designated p281-4-Firefly) in which the LACZ gene from p281-4 was replaced by the Firefly luciferase gene was created by homologous recombination.

Luciferase activity was then measured for each 3’ truncation in wild type and eIF2A

45 cells. -galactosidase (wild type; gray bars and eIF2A; black bars) and luciferase (wild type; stripped bars and eIF2A grid pattern) activity are plotted as percent of full length

URE2 measured in eIF2A cells.

46 Since no truncation resulted in activity in eIF2A cells that was equal to the levels

observed in wild type cells, it was clear that there were no regions 3’ of the initiating

AUG that were responsible for the eIF2A-mediated repression. Based on the data presented above showing that activity was maintained when the coding region was truncated at nucleotide 309, I will henceforth consider this the 3’ boundary of the URE2

IRES element.

For each truncation construct used in this study, the internal AUG was mutated to CTT to

determine the possible contribution of initiation at alternative AUG codons to the

observed activity. In each case, mutation of the internal AUG to CTT resulted in basal

levels of internal initiation, suggesting that the contribution from other AUG codons is

insignificant (Figure 8A). Although too little activity was observed to accurately assess

the influence of eIF2A on expression from constructs containing the CTT mutation, most

constructs displayed slightly greater activity in the eIF2A knockout strain.

To evaluate the possibility that gross changes in mRNA abundance resulted in the

changes in -galactosidase activity, levels of LACZ reporter mRNA in each cell type

were examined by RT-PCR. This was done using primers directed against the LACZ

region that result in a 1500 bp fragment (Figure 9A). These experiments were performed

for each mRNA in both wild type and ΔeIF2A yeast strains and with the CTT mutants of

each truncation. Small changes in RNA abundance may account for changes in -

galactosidase activity; therefore, I normalized each sample to levels of endogenous PGK1

47 A.

Figure 9: Abundance of mRNA is not responsible for the different levels of - galactosidase activity observed in functional assays of 3’ truncations. RT-PCR analysis of levels of reporter mRNAs was conducted in both wild type and eIF2A knockout cells using primers directed against the LACZ region of the mRNA. A, 1% agarose gel

48 showing the results of RT-PCR analysis of AUG constructs in each cell type. The

molecular weight standard (lane 1), full length URE2 (lanes 2-3), URE2 nucleotides 7-

812 (lanes 4-5), URE2 nucleotides 7-497 (lanes 6-7) and URE2 nucleotides 7-309 (lanes

8-9) are shown. Lanes 2, 4, 6 and 8 represent levels of mRNAs in wild type cells, while lanes 3, 5, 7 and 9 represent levels of mRNA in eIF2A knockout cells. B, Bar graph showing levels of reporter mRNA relative to PGK1 mRNA for each truncation (with

AUG and CTT at the position of the internal codon) in both cell types. Values were obtained by quantitation of RT-PCR products resolved by agarose gel electrophoresis.

Gray bars represent AUG-containing truncations in wild type cells, hatched bars represent AUG-containing constructs in eIF2A cells, black bars represent CTT- containing constructs in wild type cells and dotted bars represent relative mRNA levels of

CTT-containing constructs in eIF2A cells.

49 mRNA (Figure 9B). From these analyses, it is clear that the levels of LACZ mRNA do

not change in such a way that can account for the observed levels of activity. In AUG-

containing constructs, levels of mRNA are roughly equal in all cases examined (Figure

9B). The only exception is for the RNA isolated from wild type cells containing the 7-

812 URE2 truncation, but this construct showed low levels of -galactosidase activity in

both the wild type and eIF2A cells (Figure 9A).

B. Upstream sequences are capable of modulating IRES-mediated expression of URE2.

In order to assess the 5’ border of the URE2 minimal IRES element, I made truncations

upstream of the internal AUG codon similar to those made to determine the 3’ boundary

in Figure 8. It became clear that an inhibitory element is located between nucleotides 98

and 205 (henceforth termed “inhibited IRES”; Figure 10). Deletion of this inhibitory

element results in an increase in internal initiation to levels approximately five-fold

higher than the full length URE2 coding region. Interestingly, none of the truncations

resulted in a loss of eIF2A-mediated suppression, suggesting that the inhibitory effect is

not due to direct binding of eIF2A to the mRNA in regions other than that corresponding

to the minimal IRES element (Figure 10). Another interesting feature is that truncations

past nucleotide 205 resulted in a complete loss of internal initiation (Figure 10; truncation

at nucleotide 246). Therefore, from this point forward I will refer to nucleotide 205 as

the 5’ boundary of the URE2 IRES.

Figure 10: URE2 Truncations indicate an inhibitory sequence upstream of the internal

AUG codon. The p281-4 reporter, described in the legend for Figure 4, was used to determine the importance of 5’ URE2 deletions. Left, nucleotide numbers and schematic representations of different URE2 truncations. Right, activity measurements of the respective URE2 truncations in both BY4741 (wild type; gray bars) and BY4684

(eIF2A; hatched bars) yeast cells. Activity measurements for each truncation containing a CTT mutation in place of the internal AUG are shown for wild type (black bars) and eIF2A cells (dotted bars).

51 As with the 3’ truncations, in each construct, the internal AUG was mutated to a CTT to

exclude the possibility of alternative AUG codons acting as initiation codons. In each

case where this mutation was made, levels of internal initiation dropped to basal levels.

Therefore, the levels of internal initiation observed in Figure 10 are due to authentic

initiation events at the AUG located between nucleotides 280 and 282.

RNA analysis for each truncation was conducted as described above. Levels of mRNA

in each construct are roughly equal (Figure 11A), which suggests that changes in mRNA

levels are not responsible for the changes in activity observed in experiments presented in

Figure 10. When the levels of reporter mRNA are normalized to a PGK1 control, it is

clear that the RNA abundance is within 2-fold for all the AUG-containing constructs, and

there is no correlation between those changes and changes in reporter activity (Figure

11B). Furthermore, those that have higher levels of mRNA are in constructs that do not yield much reporter activity. Interestingly, in the 205-1061 URE2 truncation transformed

into eIF2A, levels of mRNA are the lowest amongst all the 5’ truncations, and yet, the

levels of activity are higher (Figure 11B). This suggests a higher efficiency of translation

initiation with this mRNA as compared to the other 5’ truncations.

C. Internal initiation on mRNAs with combined truncations reflect changes observed with

single truncations.

Important 5’ truncations were combined with the 3’ truncation at nucleotide 309 to affirm

that the minimal IRES could be isolated from the rest of the URE2 coding sequence

52 A.

Figure 11: Abundance of mRNA is not responsible for the different levels of - galactosidase activity observed in functional assays of 5’ truncations. RT-PCR analysis of levels of reporter mRNAs was conducted as described in Methods. A, 1% agarose gel showing the results of RT-PCR analysis of AUG constructs in each cell type. The molecular weight standard (lane 1), full length URE2 (lanes 2-3), URE2 nucleotides 98-

53 1061 (lanes 4-5), URE2 nucleotides 205-1061 (lanes 6-7) and URE2 nucleotides 246-

1061 (lanes 8-9) are shown. Lanes 2, 4, 6 and 8 represent levels of mRNAs in wild type

cells, while lanes 3, 5, 7 and 9 represent levels of mRNA in eIF2A knockout cells. B,

Bar graph showing levels of reporter mRNA relative to PGK1 mRNA for each truncation

(with AUG and CTT at the position of the internal codon) in both cell types. Gray bars represent AUG-containing truncations in wild type cells, thatched bars represent AUG- containing constructs in eIF2A cells, black bars represent CTT-containing constructs in

wild type cells and dotted bars represent relative mRNA levels of CTT-containing

constructs in eIF2A cells.

54 without abrogating internal initiation. In other words, I sought to determine whether or not the putative minimal sequence could act independent of enhancer or inhibitory sequences located up or downstream of nucleotides 205 or 309, respectively. It was evident that the levels of internal initiation observed in constructs containing truncations on both sides of the initiating AUG were representative of the levels observed when single truncations were analyzed. For example, the inhibitory element between 98 and

205 maintained the inhibitory effect when it was in the context of double truncations

(Figure 12; 98-309 truncation as compared to the 205-309 truncation). Additionally, losing the enhancer element located between nucleotides 812 and 1061 resulted in a partial loss of activity in the minimal IRES (Figure 12; compare 205-1061 truncation with 205-309 truncation).

In an effort to more precisely define the inhibitory element between nucleotides 98 and

205, further 5’ truncations were produced in combination with the 3’ IRES boundary at nucleotide 309. This analysis shows that the inhibitory effect can be reduced to a 25 nucleotide region between nucleotide 180 and 205. This element can be considered a bona fide regulatory sequence because the Ure2p region of the fusion protein did not change in the constructs that led to the discovery of this element. The small size of this region in combination with secondary structure prediction data (see below) suggest that either a protein may bind to this region or a small structural element may interact with the minimal IRES to evoke an inhibitory effect. As with single truncations that suggest no

Figure 12: The URE2 minimal IRES element is located between nucleotides 205-309,

and the upstream inhibitory element can be defined to a 25 nucleotide region upstream of

the minimal IRES. The p281-4 reporter, described in the legend for Figure 4, was used to

determine the importance of 5’ URE2 deletions. Left, nucleotide numbers and schematic representations of different URE2 truncations. Right, activity measurements of the respective URE2 truncations in both BY4741 (wild type; gray bars) and BY4684

56 region of the RNA is responsible for the eIF2A-mediated suppression, the increased

activity of the minimal IRES in eIF2A cells as compared with wild type cells is not

changed (Figure 12).

When the AUG initiation codon is mutated to a CTT codon in constructs containing

combined truncations, there is complete loss of the ability to internally initiate translation.

This suggests that there is no change in cis-sequences at the junction between the URE2

and LACZ that promotes initiation at alterative AUG codons. RNA analysis of these

constructs supports the conclusion that gross changes in RNA abundance are not

responsible for changes in internal initiation because there are no substantial changes in

mRNA levels for AUG-containing constructs that would explain the trend of activity

observed in functional assays (Figure 13A). This result was expected because the -

galactosidase activity for combined truncations mimicked the trends observed with single

truncations and the RNA analysis for single truncations indicated that RNA abundance

was not responsible for the different levels of activity (Figures 10 and 12). Normalizing

the levels of LACZ-containing mRNA to levels of PGK1 mRNA supported the idea that

the changes in steady state RNA levels are not responsible for changes in reporter activity

(Figure 13B). At most there are 3-fold differences in RNA levels, but there was no

correlation between mRNA abundance and relative activity between constructs. In fact,

several samples showed higher levels of mRNA in AUG-containing constructs (such as the 118-309 URE2 truncation in wild type and eIF2A cells), but most of these constructs showed little or no activity in functional assays. In contrast, the clones

57 A.

58 C.

Figure 13: Abundance of mRNA is not responsible for levels of -galactosidase observed in functional assays of combined truncations. RT-PCR analysis of levels of reporter mRNAs was conducted as described in Methods. A, 1% agarose gel showing the results of RT-PCR analysis of AUG constructs in each cell type. The molecular weight standard (lane 1), full length URE2 (lanes 2-3), URE2 nucleotides 205-1061 (lanes 4-5),

URE2 nucleotides 98-309 (lanes 6-7), URE2 nucleotides 118-309 (lanes 8-9), URE2 nucleotides 140-309 (lanes 10-11), URE2 nucleotides 180-309 (lanes 12-13) and URE2 nucleotides 205-309 (lanes 14-15) are shown. Lanes 2, 4, 6, 8, 10, 12 and 14 represent levels of mRNAs in wild type cells, while lanes 3, 5, 7, 9, 11, 13 and 15 represent levels of mRNA in eIF2A knockout cells. B, Bar graph showing levels of reporter mRNA relative to PGK1 for each truncation (with AUG and CTT at the position of the internal codon) in both cell types. Gray bars represent AUG-containing truncations in wild type cells, thatched bars represent AUG-containing constructs in eIF2A cells, black bars represent CTT-containing constructs in wild type cells and dotted bars represent relative mRNA levels of CTT-containing constructs in eIF2A cells. C, Schematic

59 representation of the inhibited IRES secondary structure as determined by the MFOLD server. Each Roman numeral corresponds to the stem loop depicted directly above it.

60 containing the 205-309 URE2 truncation, which I define as the minimal IRES element,

have levels of mRNA comparable with mRNA levels expressed from the construct

containing nucleotides 7-1061 of URE2. This illustrates that the 205-309 and 7-1061

URE2 constructs can be compared directly for the ability to direct internal initiation.

Secondary structure predictions of the inhibited IRES element using the MFOLD server indicate that four stem loop structures can form within this region of the RNA (Figure

13C) (69,70). I have labeled each stem I-IV in consecutive order, with stem I being at the

most 5’ end of the RNA. Stem loop IV corresponds to the region defined as the minimal

IRES, which is capable of efficiently promoting internal initiation.

D. Structure probing analysis of the URE2 minimal IRES reveals a stem loop structure that contains the initiator AUG.

After defining the minimal sequence responsible for internal initiation, it was important to understand what structural elements exist within that region. Therefore, I tested the reliability of the secondary structure prediction generated with MFOLD (Figure 13C).

RNase T1, RNase V1 and DEPC were used to analyze the structure within the URE2

IRES RNA. RNase T1 is a small enzyme (~13 kDa) with specificity for single stranded guanosine residues. RNase V1 is another small enzyme (~10 kDa) capable of cleaving nucleotides participating in hydrogen bonding interactions, and can less efficiently react with stacked single stranded nucleotides. Diethylpyrocarbonate is a small molecule probe (~300 Da) that modifies the N7 position of single stranded adenosine residues (74).

Diethylpyrocarbonate is a very specific probe that is used because of its small size, which

61 typically results in higher resolution in the mapping of single stranded regions.

Considering the URE2 IRES element is approximately 43% adenosine residues, DEPC

was a useful tool in elucidating regions with single-stranded character.

After comparing the structure probing data to structure predictions obtained using the

MFOLD server, it was clear that the minimal IRES element folds into the predicted stem

loop structure (stem IV), and the AUG is involved in base-pairing interactions (Figure

14A and B). Another interesting feature of the IRES is that there is a relatively large

apical loop that would be predicted to destabilize stem loop IV (Figure 14B) (75).

While some of the data seemed to clearly support the structure predicted from MFOLD

(i.e. DEPC and T1), I was not as confident that the RNase V1 data supported that

prediction because of the low intensity of V1 cleavage. Therefore, I made a single point

mutation (A238G) in the minimal IRES element (stem IV). The rationale for making this

mutation was that if the MFOLD prediction was correct, the mutation would stabilize the

helix in a predictable way that would allow us to more accurately assign cleavage events

in the wild type structure. If the MFOLD prediction was not correct, the mutation would

be expected to disrupt or alter the structure that was forming and presumably change the

digestion pattern that I observe in an unpredictable way. After probing the mutant

62 A. B.

63 D.

Figure 14: The minimal IRES element contains a stem loop structure. A, 6%

acrylamide/7 M Urea gel analysis of RNase T1 (lanes 1 and 2), RNase V1 (lane 3) and

DEPC-mediated strand scission analysis (lane 10). No RNase (lanes 4 and 5) and no

DEPC reactions (lane 11) are shown. Lanes 2, 3, 5, 10 and 11 contain reactions

conducted in the presence of 2.5 mM MgCl2. Additionally, base hydrolysis (lanes 6 and

9) and RNase T1 ladders (lanes 7 and 8) are shown. B, A composite structure that highlights the location of cleavages resulting from different probes is shown. Green arrows signify locations of RNase T1 cleavages, blue dots signify locations of RNase V1 cleavages and red triangles show locations of DEPC-mediated cleavages. Lower efficiency cleavages are denoted by light colored shapes. Outlined shapes show locations of Mg2+-dependent changes in cleavage efficiency. The internal AUG codon is

transposed on a yellow background. C, Parallel comparison of minimal IRES RNase

64 probing and an A238G point mutant within the minimal IRES (Stem IV). Arrows illustrate that RNase T1 cleavage efficiency at nucleotides 267, 268, 276 and 277 are higher in the wild type minimal IRES sequence, while red brackets illustrate increased

RNase V1 cleavages at nucleotides 237-243 in the A238G mutant of the minimal IRES.

D, A composite structure that highlights the differences in cleavage efficiency for RNases

T1 (arrows) and RNase V1 (dots) between the wild type and A238G mutant form of the

URE2 minimal IRES. Black shapes illustrate increased cleavage in the mutant as compared to the wild type, while grey shapes illustrate decreased cleavages in the mutant as compared to the wild type. There were no obvious differences between the wild type and mutant with DEPC-probing. The internal AUG is highlighted in yellow and the mutated nucleotide is denoted in red.

65 minimal IRES element, I determined that the digestion pattern was changing in a way

that indicated the MFOLD prediction was correct. For example, RNase V1 was able to cleave much more efficiently between nucleotides 236-241, while RNase T1 was able to cleave less efficiently at guanines 267-268 and 276-277 in stem IV (Figure 14C).

Consistent with the idea that the IRES was stabilized in a predictable way, guanine residue 214 was more efficiently cleaved by RNase T1 which suggests there were less options available to the enzyme in the mutant situation as compared to the wild type

IRES (Figure 14D). Another feature that is consistent with the formation of the stem is that DEPC probing did not seem to change upon stabilization of stem IV with the A238G mutation (Figure 14D). Furthermore, I did not detect the presence of tertiary interactions in these experiments. One region of the minimal IRES (nucleotides 209-217) was efficiently cleaved by RNase V1 (Figure 14A and B). However, sequence alignment of this region with other regions within the minimal IRES element did not predict the formation of base pairing indicative of tertiary interactions (data not shown). As mentioned above, it is known that RNase V1 can cleave nucleotides in stacked single stranded regions. Due to the high frequency of adenosines in this region, RNase V1 may be reacting with stacked residues.

E. The inhibitory element between nucleotides 98-205 does not change the structure of the minimal IRES element.

I have shown that the region directly upstream of the minimal IRES element is capable of repressing the URE2 IRES activity. Therefore, I addressed the question of whether or not there was a change in the structure of the minimal IRES element (stem IV) in the

66 presence of stems I-III that might lead to the inhibitory effect. To examine this

possibility, I probed the structure of the inhibited IRES in the same way that I did for the

minimal IRES described above. Interestingly, the structure of the minimal element did

not change in the presence of the upstream inhibitory sequence (Figure 15A and B).

Considering our data that further defines the inhibitory effect to the region between

nucleotides 180 and 205, it is interesting to note that stem loop III exists within this

region of the RNA and may suppress IRES activity (Figure 15B).

As with the minimal IRES element, I decided to investigate the validity of the structures by determining if introduction of point mutations would change the structure in a

predictable way. Therefore, I introduced three point mutations (A148U in stem II,

A202C in stem III, and A238G in stem IV) in the inhibited IRES. These mutations were

selected to stabilize each of the stems predicted to form by MFOLD, with the exception

of the most 5’ stem. The mutation in stem IV enabled RNase V1 to cleave more

efficiently in the region of the minimal IRES element between nucleotides 230-240

(Figure 15C). Concordantly, RNase T1 was able to cleave less efficiently at guanosine

residues 267-268 and 275-276, which is the same result observed with the isolated

minimal IRES element (stem IV). These data help to support the structure of the minimal

IRES because the enzymes responded in the same way with the inhibited IRES as they

did with the minimal IRES element. Unlike the situation with the isolated minimal IRES

element, a change in DEPC modification at several nucleotides was observed in the triple

mutant of the inhibited IRES. Specifically, several nucleotides in the region of stem IV

67 A. B.

68 D.

Figure 15: The inhibited IRES element does not significantly differ in secondary

structure in the region of the minimal IRES element. A, 6% acrylamide/7 M Urea gel

analysis of base hydrolysis (lanes 2 and 9) and RNase T1 ladders (lanes 1 and 8) are

shown. RNase T1 (lanes 6 and 7), RNase V1 (lane 5) and DEPC-mediated strand

scission analysis (lane 10). No RNase (lanes 3 and 4) and no DEPC reactions (lane 11)

are shown. Lanes 3, 5, 6, 10 and 11 contain reactions conducted in the presence of 2.5

mM MgCl2. B, A composite structure that highlights the location of cleavages resulting

from different nucleolytic probes is shown. Color and shape identities corresponding to

intensity and probes are as depicted in 15B. The internal AUG codon is highlighted in

yellow. C, Parallel comparison of inhibited IRES RNase probing and the A148U, A202C

and A238G triple mutant. Arrows illustrate increased RNase T1 cleavage efficiency in

the wild type as compared to the mutant, which occur at nucleotides 153 and 156 and at nucleotide 255 in stem IV that corresponds to the minimal IRES region. Red brackets

69 illustrate increased RNase V1 cleavages at nucleotides 147-151 in the triple mutant of the inhibited IRES. D, A composite structure that highlights the differences in cleavage efficiency for between the wild type and A148U, A202C and A238G triple mutant form of the URE2 inhibited IRES. Shapes coincide with the labeling depicted in 15D, except that triangles are added to denote changes in DEPC modification in the mutant as compared to the wild type. Shape colors are as depicted in 15D where black indicates increased cleavage efficiency and grey indicates decreased cleavage efficiency in the mutant relative to the wild type. The internal AUG is highlighted in yellow and mutated nucleotides are shown in red letters.

70 were cleaved less efficiently in the mutant as compared to the wild type (Figure 15D).

There are also a number of nucleotides in stem loop II that were cleaved less efficiently

with DEPC in the mutant (Figure 15C). These data were very useful in elucidating the

structure of the region located between nucleotides 100-128 and 144-173 (corresponding

to the regions of stem I and II). In the wild type inhibited IRES, the data suggests these

regions are flexible because they are able to be modified with DEPC and cleaved by

RNase V1 (Figure 15B). However, in the mutant inhibited IRES, they more exclusively

form the stem loops predicted with MFOLD. In contrast, the region corresponding to stem III, which has been shown to be responsible for the inhibitory effect, is capable of

forming the predicted helical structure. While introduction of a stabilizing mutation in

stem III is evident by structure probing (Figure 15D), it did not change the pattern of

nuclease digestion in the region of stem III like it did with stem II. This suggests that

stem loop III is forming in this region of the RNA in the wild type inhibited IRES, though

the importance of this structure has not been further tested. I was unable to find evidence

of tertiary interactions with this region of the RNA that may have been responsible for

the inhibitory effect. The lack of significant change in secondary structure in stem IV in

the presence of the inhibitory region and the lack of efficient cleavages by V1 in the

loops of stems III and IV provides confidence that there are no tertiary interactions

between the inhibitory region and the minimal IRES detectable in these structure probing

analyses.

71 F. Nucleotides within the stem region directly adjacent to the apical loop are important

for IRES function

Since it was determined that the URE2 minimal IRES element (stem IV) can fold into a

small stem loop structure in vitro, I was interested to know the importance of different

nucleotides in the structure for internal initiation. In order to identify nucleotides that

may be important for IRES structure and function, I first examined the sequence of the

IRES in different Saccharomyces species (Figure 16A). The URE2 gene is not well

conserved outside of the Saccharomyces genus making phylogenetic comparisons impossible. Unfortunately, while the gene is well conserved within Saccharomyces, it

lacks considerable covariation of nucleotides that would provide clues about the

importance of different nucleotides for IRES function (Figure 16B; block arrows

highlight variant nucleotides that maintain base pairing potential). As expected, much of

the variation that does occur appears in the wobble position of the codons (Figure 16A

and B).

To begin a functional analysis of the important nucleotides for internal initiation of

translation, I began by investigating the importance of the large apical loop using several

deletions and mutations (Figure 17A). Deletion of either 6 or 12 nt from the loop region

resulted in no change in activity. However, deletion of 18 nt from the loop reduced IRES

activity to background levels (Figure 17B). This suggests that some portion of the loop is important for activity. Since these data do not address the question of whether it is the sequence or the structure of the loop that is important, I also utilized several adenine to guanine mutants in the apical loop of the minimal URE2 IRES element (Figure 17A).

72 A.

205 60

S. cerevisiae S. douglassi S. paradoxus S. bayanus S. mikatae 61 309

S. cerevisiae * S. douglassi S. paradoxus S. bayanus S. mikatae

73 G C G A U C A A 255 C A G G A C A A C A G A G A C 245 G A A A A C G G U U A AT C G 265 C G A C GC U A G C A U GGG 235 A U A G U C A U A C 275 T C G G A U A G A U CC U A G A U A G C 225 U A A G A U 285 C AA A G C S.douglassi/S.paradoxus G A A G C G TC G S. bayanus CA AA A UA A UG A UA A UG GA G U A U UC C A GA A U U A CA CA C A A A S. mikatae C 205 215 295 305

Figure 16: Phylogenetic comparison of URE2 IRES sequence reveals little covariation.

A, Sequence alignment of several Saccharomyces species indicates that the URE2 minimal IRES sequence is relatively well conserved. The nucleotides that vary primarily reside in the wobble position of each codon (denoted by pink dots). The internal AUG codon is highlighted by a bar and the star. Numbers correspond to nucleotide numbers.

B, Sequence variation plotted on a map of the structure. Saccharomyces douglasi/paradoxus, bayanus, and mikatae nucleotide variation are in red, blue and green letters, respectively. Points of whole or partial covariation are marked with block arrows.

74 The wobble codon position is denoted by pink dots, the internal AUG codon is highlighted in yellow, and numbers correspond to nucleotide numbers.

75 None of these mutants resulted in a change in IRES-mediated translation suggesting the structure, but not the sequence is important for internal initiation (Figure 17B). Each deletion and mutant was tested in both the wild type and eIF2A cells to examine if any altered the ability of eIF2A to act as a suppressor of internal initiation. In each case, the ability of eIF2A to suppress internal initiation on the minimal URE2 IRES was present

(Figure 17B).

I then attempted to determine whether internal initiation could be modulated by altering

the number of base pairs with the internal AUG codon. This was done by mutating

nucleotides participating in base pair interactions with the AUG codon (Figure 17C; nucleotides 226-228). I hypothesized that if the AUG codon was not participating in base

pair interactions, it would be available to interact with the incoming met-tRNAi. Thus,

when additional base pairs with the internal AUG codon are present, it would be less available to interact with incoming met-tRNAi, resulting in reduced activity. The A226C

mutant (which is predicted to add a base pair with the AUG codon) showed levels of

activity which mirrored the levels observed for the wild type minimal URE2 IRES

(Figure 17C). When either an A227C or an U228C mutant (each of which are predicted to abolish base pairs with the AUG codon) was tested, the levels of activity resembled

what was observed with the minimal IRES element (Figure 17C). Since the data

indicated that the base pairing status of the internal AUG codon is not important for

activity, I sought to investigate if the positioning of the AUG codon is a determinant for

efficient internal initiation. This was done by switching the AUG codon with either the

Figure 17: Effects of different parts of the URE2 IRES structure on cap-independent

translation initiation. A, Schematic representation of loop deletions and mutations.

Dashed lines represent nucleotides that were deleted relative to wild type within the loop

78 region. Positions of stem IV mutations are denoted by red arrows. The internal AUG codon is highlighted in grey. B, The monocistronic -galactosidase reporter plasmid, p281-4 described in the legend for Figure 4, was used to assess the effect of deletions within the apical loop region of the URE2 minimal IRES element. Mutations within the loop were utilized to delineate between the importance of the structure and the sequence for IRES activity. C, Activity measurements for mutations that alter the base pairing status of the internal AUG codon in stem IV using the p281-4 reporter construct.

AUGGAU and AUGAGU represent mutations of the URE2 IRES element in which the position of the internal AUG codon was swapped with the codon directly before or after it, respectively, to test the importance of internal AUG position for internal initiation. These mutations result in production the short from Ure2p that vary in length by one amino acid. D, Activity measurements for mutations and compensatory mutations in the stem region directly adjacent to the apical loop illustrated in A. In B-D, measurements for wild type (grey bars) and ΔeIF2A (hatched bars) yeast strains are depicted.

79 preceding codon or the codon immediately following, resulting in production of β- galactosidase fusion proteins containing either one additional or one less amino acid, respectively. Neither construct led to abrogated levels of internal initiation (Figure 17C).

These data suggest that neither the stability of the stem region surrounding the internal

AUG codon nor the positioning of the AUG codon is important for IRES-mediated initiation on the URE2 IRES element. Furthermore, suppression of cap-independent translation initiation mediated by eIF2A is not affected by the base pairing status or the positioning of the internal AUG codon (Figure 17C).

I then wanted to understand the importance of the double stranded region between the apical loop and the AUG codon. This question was addressed using the A238G mutation described for structure probing (Figure 17D). The A238G mutation showed approximately 8-fold reduced activity. Restoration of the bulge to the stem using the double A238G/C269A mutant did not result in recovery of activity (Figure 17D). This suggests that either the sequence or structure of this region of the IRES element is important for activity. Next, the U272G mutation, which breaks a base pair in the stem, was used to reduce the stability of the stem. This base pair was also restored using a double mutant (A235C/U272G; Figure 17D). The introduction of the U272G mutation resulted in a 20-fold reduction in activity, which was not restored with the compensatory mutation (Figure 17D). These data suggest that the sequence of this stem is important for

URE2 IRES activity because neither compensatory mutation restores IRES activity to wild type levels.

80 I then tested if increasing the size of the double stranded region (and the stability) would

affect internal initiation while maintaining the A238 and U272 positions. This was done

by producing an A264C/C265A mutant, which extends the stranded region by utilizing

nucleotides within the loop region (Figure 18A; compare the stem size in Figure 18A to

that of 14B (5 versus 3 base pairs, respectively)). The reduction in loop size by four

nucleotides was not considered to be an issue here because previous experiments

suggested it is possible to delete 6 nt from the loop without perturbing internal initiation

(Figure 17B). Interestingly, the A264C/C265A mutant allowed internal initiation at

levels comparable to the wild type minimal IRES element (Figure 18B). When this

mutant is combined with an A227C/U228A mutant, thus abolishing base pairing with the

internal AUG codon, levels of internal initiation were not abrogated (Figure 18A and

18B; predicted structure and mutant IRES activity, respectively). The rationale for making the A227C/U228A/ A264C/C265A quadruple mutant was that if both the size of

the double stranded region and the ability of the AUG codon to base pair with the met-

tRNAi were important for internal initiation, it may be possible to produce an IRES

element that is substantially more active. The A227C/U228A mutant alone did not result

in a change in IRES-mediated translation of the URE2 IRES element as expected based

on results with single mutants (Figure 18B). These data verify that base pair interactions

with the internal AUG codon do not effect levels of internal initiation on the URE2 IRES, and support the idea that the sequence of the stem between the apical loop and the internal AUG codon is important.

Figure 18: Mutations that stabilize the double stranded region adjacent to the loop and destabilize the region around the internal AUG codon do not affect activity. A, The

82 predicted structure that results from a quadruple mutant of the URE2 minimal IRES element. Mutations in the loop that extend the double stranded region by two base pairs

(A264U/C265A), and mutations that disrupt two base pair interactions with the internal

AUG codon (A227U/U228A) are shown in red. B, Activity measurements for double mutants that extend the stem, perturb base pair interactions with the internal AUG, and the quadruple mutant shown in A. IRES activity in wild type (grey bars) and eIF2A

(hatched bars) are shown.

83 While both the structure and sequence have been shown to be important for internal initiation (from data examining the apical loop and stem between the apical loop and internal AUG, respectively), the importance of stem stability remains unclear. To directly assess the contribution of stem stability to cap-independent translation initiation on the URE2 IRES element, I predicted the change in free energy of each mutation or deletion tested in this thesis using the MFOLD RNA folding algorithm (69,70). When the activity of each mutant was plotted against the structure stabilities, there was no correlation between activity and stability for the URE2 IRES mutants and deletions

(Figure 19). This observation is true when examining URE2 IRES mutant activity in both the wild type and ΔeIF2A strains. Despite the importance of the single stranded apical loop, these data strongly suggest that that the stability of the stem is not an important feature for IRES activity. Furthermore, the results reveal that the sequence of the double stranded region directly below the apical loop is important for IRES activity, and confirm that the base pairing status of the internal AUG codon is not an important determinant for IRES-mediated translation.

G. Conclusions

This work established that the minimal IRES element for the URE2 IRES is located between nucleotides 205-309 using a reporter assay designed to minimize the effect of cap-dependent initiation of translation. In these experiments, several controls were utilized to affirm that changes in activity measurements were resulting from differences in translation initiation. First, the internal AUG was mutated to CTT for each of the

Figure 19: Stability of the URE2 IRES element is not a critical determinant for efficient

internal initiation on the URE2 IRES element. Stability of different mutants and

deletions of the minimal URE2 IRES element were predicted with the MFOLD server.

Activity measurements for each deletion or mutation of the URE2 IRES element obtained

using the p281-4 construct was normalized to activity of the wild type minimal IRES

element. Internal ribosome entry site activity was normalized to the wild type minimal

URE2 IRES element in both wild type and eIF2A cells. Normalized IRES activity was

then plotted against secondary structure stability for both the wild type (red diamonds) and eIF2A (green squares) cells.

85 constructs to confirm that activity was not resulting from initiation at alternative AUG codons. Second, mRNA abundance was measured using RT-PCR to be sure that gross changes in levels of mRNA were not the cause of changes in IRES-mediated expression.

Finally, primers were designed in such a way that the codon at the junction between the

URE2 sequence and the start of LACZ did not change between URE2 truncations, which eliminates the possibility of a rare codon causing ribosome stalling and artificially low levels of -galactosidase activity.

It is difficult to make comparisons between the URE2 IRES element and other yeast

IRES elements because evidence suggests that there are cryptic promoters in many that have been previously characterized (42-44). However, it is possible to make some comparisons with the invasive growth IRES elements described by Gilbert et al. (39).

Many of the invasive growth IRES elements and the URE2 IRES element contain an A- rich stretch. This stretch is important for the activity of the invasive growth IRESs whereas mutations that break up the A-rich nature of the URE2 IRES stretch do not affect internal initiation (Figure 16). Since the mutations show that the A-rich characteristic is not important for URE2 IRES function, it was interesting to find that IRES activity is abolished when 18 nt are deleted from the loop (Figure 17). It is not expected that this result is related to increased stem stability because stabilizing the stem by adding base pairs does not result in a decrease in activity (Figure 18).

Considering there are few accepted yeast IRES elements, it is of interest to compare the

URE2 IRES element with elements from other organisms. The URE2 IRES element is

86 104 nucleotides long, which is in the range of IRES element sizes previously observed.

As mentioned above, IRES elements have been shown to be as small as 22 nucleotides,

and as large as 700 nt (13,25). There is also a report of a yeast element that interacts with

ribosomal RNA to recruit the ribosome which is only 9 nucleotides in length (76). In

terms of stability, the URE2 IRES element has a G of approximately -15 kcal/mol at 30o

C, while most cellular and viral IRES elements are at least -50 kcal/mol (27,28). While

URE2 IRES function is not thought to be dependent on the low level of stem stability, it

is thought that the low stem stability is an important feature for translation of the long

form of Ure2p. If the stability of the URE2 IRES element were too high, it may cause reduced levels of the long form of Ure2p due to frameshifting, which is known to result from the presence of stable pseudoknot structures in the coding region of viral mRNAs

(77,78).

The stability of the URE2 IRES stem is not thought to be critical for internal initiation.

When the change in free energy is plotted against the activity for every mutant used in this study, there is no linear correlation that suggests that the stability is important for internal initiation (Figure 19). If stability of the URE2 IRES element were an important factor in modulating internal initiation, a decrease in activity would be observed as additional base pairs were introduced by mutagenesis. Taken together, the data indicate that the sequence, particularly at positions A238 and U272, are important for activity.

Mutations at these positions may perturb binding sites for a trans-acting factor that is important in recruitment of the 40S ribosomal subunit to the initiating AUG codon for translation initiation. The single stranded structure of the loop may be important in an

87 initiation step, which may differ from the step involving nucleotides A238 and U272.

This could explain the reduction in activity when 18 nt are deleted from the loop since the

loop is presumably too small at this point to enable cap-independent initiation (Figure

17B).

The determination of the secondary structure within the URE2 IRES element reveals some striking characteristics. The presence of base pairing interactions between nucleotides within the internal AUG suggests that the unwinding of this structure, at least

in part, is an important step in internal initiation. If the structure were not at least partly

unwound, the tRNA anticodon would not be able to recognize the start site because of its

involvement in base pairing interactions. To my knowledge, this is a unique aspect of the

URE2 IRES element. Probing experiments of Apaf-1, Baf-1, Cat-1, and c-myc IRES

elements suggest that the internal AUG codons are not imbedded in higher order RNA

structures (27,29,30,79). In those experiments, the possibility of a buried initiation codon

may not have been apparent because the internal AUG was at the 3’ terminus of the

synthetic RNA being probed. The data suggest that the base pairing status of the internal

AUG codon in the minimal URE2 IRES is not a determinant of IRES function because

mutations that abolish base pairing with the internal AUG codon do not result in a

reduction in translation initiation (Figure 17C and 19). However, it is possible that the

base pairing status of the internal AUG codon is important in events downstream of IRES

recognition. If trans-acting factors recognize and rearrange the structure of the URE2

minimal IRES element, then the newly configured IRES element may have a higher

affinity for the ribosome. Furthermore, the structure of the IRES element with trans-

88 factors bound may have the internal AUG codon exposed and available for base pairing

with the incoming met-tRNAi anticodon.

A unique aspect of the URE2 IRES element is the sequences surrounding the minimal

URE2 IRES element that are capable of modulating its activity. As mentioned previously, several viral IRES elements require some downstream sequence for activity

(31,32); however, to our knowledge, there are no cellular IRES elements that utilize downstream elements to modulate levels of internal initiation. Usually downstream

sequences are not examined because the IRES element is located in the 5’ UTR of the

mRNA. The fact that the URE2 IRES element resides in the coding region of full length

Ure2p mRNA triggered the investigation into the role of downstream sequences in

controlling internal initiation. These experiments suggest that there may be downstream

sequences that are capable of modulating internal initiation of other IRES elements. I

was able to verify the possible contribution of these regions of RNA by replacing the -

galactosidase coding region with firefly luciferase. This allowed us to show that the

different Ure2p fragments were not responsible for differentially altering the specific

activity of -galactosidase protein (Figure 8B). Therefore, the observed changes in

activity are reflective of cis-acting RNA sequences within the URE2 mRNA. Given the

magnitude of the effect imparted by the enhancer sequence located between nucleotides

812 and 1061 of the URE2 IRES, downstream regulatory elements could have profound

effects on the activity of other IRES elements.

89 To elucidate the effect of the upstream inhibitory region, I examined changes in the

structure of the minimal IRES element using RNA structure probing. Interestingly, there were no obvious changes in the structure of the minimal IRES region when the sequence was probed in the presence of the inhibitory sequence immediately upstream. In fact, no tertiary interactions were detectable when the URE2 minimal IRES element was probed either in the presence or absence of the inhibitory sequence. The small stem loop region between nucleotides 180-205 may play a role in the inhibitory effect by interacting with a protein that is capable of rearranging the RNA into an inactive form of the IRES or by altering the positioning of the initiating AUG on the 40S subunit (i.e. perhaps not directly to the P-site).

At the start of these studies, there were no cellular examples of the architecture of the

URE2 IRES, wherein the minimal sequence is located within the coding region of a protein. Interestingly, a cellular IRES that resembles the format of the URE2 IRES has been identified in mammalian p53 mRNA, but it is unknown whether there are sequences up or downstream of the internal AUG that are capable of modulating internal initiation

(4). This IRES element has a much higher predicted stability as compared to the URE2

IRES (-92 versus -15 kcal/mol, respectively), and none of the predicted structures within this region resemble the URE2 IRES. Furthermore, the p53 IRES is regulated in a cell cycle-dependent manner. While cell cycle-dependent regulation of the URE2 IRES cannot be excluded, it is more likely that this IRES is regulated in response to nutritional cues because of the role of Ure2p in nitrogen assimilation.

90 These studies are in agreement with our view of the role of eIF2A in internal initiation.

None of the truncations or mutations examined altered the ability of eIF2A to suppress internal initiation. Based on how this protein was identified, it was not surprising that I

determined that the protein is most likely acting on the process of internal initiation

without direct binding to the IRES mRNA. In the following chapter of this thesis, I will

present evidence that there are at least two pathways for regulation of internal initiation in

yeast, an eIF2A-dependent and an eIF2A-independent pathway.

91 IV. Elucidating the role of eIF2A and eEF1A in cap-independent initiation of the

URE2 IRES.

As mentioned above, yeast is a valuable genetic model for dissecting complex biological

pathways. Until recently, yeast have not been extensively used to understand the

mechanism of IRES-mediated initiation of translation because few yeast IRES elements

had been identified. The previous chapter focused on the questions oriented around the

nucleic acid requirements for translation from the URE2 IRES element. This chapter will

focus on the role of eIF2A and other possible trans-acting factors that may influence

internal initiation in yeast.

A. IRES elements lacking a poly(A) stretch are not targets of eIF2A-mediated

suppression

After Gilbert et al. identified the invasive growth IRES elements, it was important to

examine whether translation of those IRES elements is suppressed by the eIF2A protein

(39). To address this question, several IRES elements described by Gilbert et al. were

placed in the construct which I used for analysis of URE2 (p281-4). To begin to

understand the importance of PABP in internal initiation, invasive growth IRES elements

used in this study were selected for the presence and absence of poly(A) stretches.

Internal ribosome entry sites containing poly(A) stretches tested include the FLO8,

TIF4632, YMR181c, and the NCE102, while those lacking this stretch are the URE2 and

GIC1 elements. Each IRES element was examined for activity in both the wild type and

ΔeIF2A yeast strains. Interestingly, the data show that the eIF2A-mediated suppressive

92 effect is specific to the IRES elements lacking poly(A) stretches (Figure 20). The URE2

IRES element, which lacks a poly(A) stretch, acts as a positive control for the suppressive

effect of eIF2A. These data indicate that there are at least two pathways that influence

internal initiation in yeast: one dependent on eIF2A and another independent of eIF2A .

B. The eIF2 and eIF2A-mediated pathways compete for delivery of initiator met-tRNAi to

the ribosome during translation initiation on the URE2 IRES element

Due to the existence of both eIF2 and eIF2A in yeast, and the redundant roles in delivery

of met-tRNAi during initiation, it is intriguing to hypothesize that eIF2 and eIF2A can

compete for delivery of the initiator tRNA on some IRES elements. If true, this

hypothesis could explain the increase in IRES activity in ΔeIF2A yeast cells because of

the removal of one of two competing pathways. If the eIF2 pathway is kinetically faster,

than the eIF2A-mediated pathway, IRES activity would be expected to increase in the

absence of eIF2A. This is true because the faster eIF2-mediated initiation would account

for all of the initiation events on the URE2 IRES. This hypothesis is difficult to assess by removal of the eIF2-mediated pathway because each component of eIF2 is essential for cell viability. However, it is possible to abrogate eIF2-mediated delivery of met-tRNAi by inactivating eIF2 by phosphorylation. Inactivation of eIF2-mediated delivery of met-tRNAi can be achieved using a protein known as Gcn2p, which phosphorylates the

alpha subunit of eIF2 rendering the pathway inactive because eIF2 bound GDP can no

longer be exchanged for GTP. The decrease in overall protein synthesis is further

amplified because the guanine nucleotide exchange factor (eukaryotic initiation factor

Figure 20: Internal ribosome entry sites containing poly(A) stretches are not subject to eIF2A-mediated suppression. A collection of invasive growth IRES elements with or without poly(A) stretches were cloned into the p281-4 reporter plasmid and tested for levels of internal initiation using -galactosidase assays. Each IRES element was tested in both wild type (grey bars) and eIF2A (hatched bars) yeast strains. The identity of the

IRES element and the poly(A) status is listed on the x-axis.

94 2B), which is present in limiting quantities in the cell, is sequestered by phosphorylated eIF2α (20). For these experiments, Gcn2p is mutated at positions E601K and E1606G,

which alter the protein kinase and ribosome-binding and dimerization domains,

respectively, to produce constitutively active protein (Gcn2pc) (45,80).

To investigate the possibility of competition between the eIF2 and eIF2A-mediated

pathways, a Gcn2pc mutant was expressed in both wild type and ΔeIF2A yeast strains in

the presence of a URE2 cap-independent reporter of translation initiation (Figure 4). In these experiments, the Gcn2pc allele was expressed from a plasmid, so vector alone

controls were used. When IRES activity is compared between cells expressing mutant

Gcn2pc, it is clear the there is a greater reduction in URE2 IRES-mediated translation in

ΔeIF2A cells than in wild type cells (Figure 21A). In contrast, cap-dependent initiation

of translation is comparable in both wild type and ΔeIF2A cells, and the reduction in

initiation in the presence of Gcn2pc is not greater in ΔeIF2A yeast (Figure 21B).

Reductions in both cap-dependent initiation of translation (Figure 21B) and overall cell

growth (data not shown) act as controls for eIF2α phosphorylation in these experiments.

These experiments indicate that the two pathways are in competition as suggested in

experiments with eIF2A yeast and the URE2 IRES elements. However, Western blot

analyses of levels of eIF2α phosphorylation should be conducted to more directly control

for the level of inactivation of eIF2-mediated translation initiation.

95 C. Several proteins interact with eIF2A

Given that both eIF2A-dependent and eIF2A-independent pathways appear to exist and likely compete for internal initiation in yeast, it is of interest to investigate the differences in the eIF2A-mediated pathway that cause it to specifically repress cap-independent translation initiation. In order to understand the differences in the eIF2A-mediated pathway, a better understanding of the function of eIF2A in translation is needed. To this end, I searched for proteins that interact with eIF2A expecting that a few of these may influence internal initiation by acting with eIF2A to suppress cap-independent initiation of translation. To identify eIF2A-interacting partners, HA-tagged eIF2A was overexpressed in eIF2A cells under the control of an inducible GAL 1/10 promoter.

Tagged eIF2A protein was then isolated from cell lysates using HA-affinity chromatography, and coimmunoprecipitated proteins were resolved by 10% SDS-PAGE

(Figure 22). Cell lysates transformed with empty vector were analyzed in parallel to control for proteins that bound the resin non-specifically. There were many proteins isolated that were specific for associating with eIF2A (Figure 22; compare vector alone eluate to HA-eIF2A eluate). Bands specific to the HA-eIF2A eluate were excised from the gel, digested with trypsin, and identified by mass spectrometry (Table 1). As expected, based on previous data that indicated that eIF2A binds to 40S ribosomal subunits, a number of ribosomal proteins were identified. These proteins are useful as a positive control. Many of the ribosomal proteins identified are part of the large ribosomal subunit, despite the idea that eIF2A interacts with the small ribosomal subunit.

This observation is not surprising because eIF2A primarily appears with 80S ribosomes

96 A.

97 Figure 21: Eukaryotic initiation factor 2 phosphorylation reveals competition between eIF2 and eIF2A-mediated translation initation. A, Wild type (grey bars) and eIF2A

(hatched bars) yeast were transformed with either vector or vector expressing a constitutively active mutant of Gcn2p in combination with the minimal URE2 IRES element in the p281-4 reporter of internal initiation (Figure 4). Reporter mRNA was induced in the presence of galactose, and -galactosidase activity was measured.

Minimal URE2 IRES activity in the presence or absence of Gcn2pc is indicated on the x-

axis. B, A cap-dependent initiation reporter lacking the stable stem loop and URE2

region in the p281-4 construct (Figure 4) was transformed into wild type and eIF2A yeast (bar colors are as indicated in A) in the presence of vector alone or vector expressing Gcn2pc. Levels of cap-dependent expression were measured using β- galactosidase assays, as described in A.

98 in polysome profile experiments (Figure 7B). The other proteins that were identified are

involved in many different processes in yeast including 5’ to 3’ RNA turnover,

nucleocytoplasmic shuttling of RNA, and facilitating proper protein folding. While it is

possible to obtain false positives from an experiment of this type, the binding proteins

that I focused on have all been shown by others to interact with eIF2A in global

proteomic studies (Table 1) further supporting their positive interaction with eIF2A.

Several eIF2A-interacting partners were of particular interest based on previous evidence

that they function in RNA processing or translation. As a result, I focused on further

analyzing the interaction of eIF2A with Kem1p, eEF1A, Ssb2p, Npl3p and Ded1p. It was important to verify these interactions using an independent method. Since the initial

immunoprecipitation relied on exogenously expressed eIF2A, the interactions should be

confirmed with eIF2A at an endogenous level of expression. To do this, I generated a

yeast strain expressing eIF2A-HA under the control of its endogenous promoter

(henceforth termed 2A-HA yeast), which should result in natural cellular concentrations

of the eIF2A-HA protein. This strain was produced by homologous recombination of a

cassette containing the HA-tag and a geneticin resistance marker at the eIF2A locus,

which resulted in production of C-terminally HA-tagged eIF2A protein (81). This strain

was verified by both PCR analysis of genomic DNA (Figure 23A; top panel) and

Western blotting (Figure 23B; bottom panel) to confirm the expression of an HA-tagged

protein of the proper molecular weight.

Figure 22: Several proteins exist in complex with eIF2A. Hemaglutanine epitope

tagged eIF2A protein was overexpressed in eIF2A cells, and immunoprecipitated from

yeast whole cell lysates with -HA antibodies. Protein complexes were eluted from -

HA sepharose with HA peptides, concentrated by TCA precipitation, and analyzed by

10% SDS-PAGE. In parallel, the same procedure was carried out with lysates from yeast containing empty vector. After the elution, -HA sepharose beads from each reaction were boiled and the remaining protein was analyzed on the same gel as the eluates from the -HA sepharose. The resulting gel was stained with Coomassie, and bands specific to the reaction containing HA-eIF2A were identified by LC mass spectrometry. The gel used for mass spectrometry is displayed above with HA-eIF2A eluates, vector control eluates, and beads for each reaction after elution as indicated. Numbered bands represent those excised for mass spectrometry, which were specific for the yeast expressing HA- eIF2A protein.

100

101

Figure 23: The 2A-HA yeast strain contains an integrated cassette containing an HA-tag and a geneticin resistance marker, and the eIF2A-HA protein is expressed. After

selecting for yeast containing an integrated geneticin resistance marker, genomic DNA

from both wild type and the 2A-HA yeast were collected and subjected to PCR to verify

the presence of the HA-tagging cassette at the eIF2A locus. The presence of a band at

1750 bp indicates successful integration of the HA cassette at the eIF2A locus.

Subsequently, lysates were prepared for the wild type and 2A-HA yeast and analyzed by

SDS-PAGE followed by Western blotting with -HA-HRP antibody. A band

representing the C-terminally tagged eIF2A protein is visible at approximately 75 kDa,

which is the expected molecular weight for this protein.

102 Yeast lysates from 2A-HA yeast could then be utilized in GST-pull-down experiments using glutathione-S-transferase (GST) fusion proteins to verify the interactions between eIF2A and eEF1A, Ssb2p, Npl3p and Ded1p. These fusion proteins were expressed in E. coli, and partially purified using glutathione-sepharose beads. In parallel, GST alone was isolated for use as a negative control. The partially purified GST-fusion proteins were then utilized in pull-down experiments using 2A-HA yeast extract. The amount of eIF2A-HA in the reaction was also analyzed to assess the fraction of the eIF2A interacting with the GST-fusion proteins (Figure 24A; compare reactions containing each

GST-fusion protein with the 8% input control). Each of the aforementioned protein interactions were verified by this method except Kem1p because it is lethal to E. coli cells. As such, I verified the Kem1p:eIF2A interaction under endogenous concentrations of both proteins by isolating eIF2A-HA from 2A-HA yeast, analyzing proteins with SDS-

PAGE, and detecting levels of Kem1p with -Kem1p antibody (Figure 24B).

D. Interactions between eIF2A and eEF1A or Ssb2p are RNA-independent.

Each of the candidate proteins are known to either bind RNA directly or interact with the ribosome, so it is possible that the observed interactions are indirect, and are visible because the ribosome or other mRNP complexes were being immunoprecipitated.

Another feature of the candidate proteins is the presence of RNA recognition motifs that would be capable of mediating an RNA-dependent interaction with eIF2A. To assess the

RNA-dependence of these interactions, I conducted the GST-pull-down experiments described above in the presence and absence of a high concentration of RNase A. If the interaction is RNA-dependent, reduced amounts of protein would be visible in the lanes

103 containing reactions done in the presence of RNase A. For Kem1p, the importance of

RNA was analyzed under endogenous conditions, as described above. Interestingly, the

data indicates that only eEF1A and Ssb2p interact with eIF2A in an RNA-independent

manner (Figure 25A). As a control, levels of ribosomal RNA at the end of the incubation

period were assessed by RT-PCR to verify that the RNA was being degraded (Figure

25B). Ribosomal RNA was analyzed for the control because it is the most abundant

RNA in the cell. If ribosomal RNA appears to be degraded by RT-PCR, it suggests that

there are little other intact RNAs remaining in the reaction. From this point forward I

will focus on the relevance of the interaction between eEF1A and eIF2A because of the

implications on what is known about eEF1A function. Considering the role of eEF1A in

elongation, it is of interest to understand the importance of eEF1A in regulation of

eIF2A-mediated translation initiation. The Ssb2p protein would be of particular interest

for future studies both because the interaction with eIF2A is not RNA-dependent, and

Ssb2p is involved in protein folding, a function others have suggested eIF2A may

regulate (82). The other proteins, which may be important for eIF2A-mediated internal

initiation despite evidence that the interactions are RNA-dependent, may be focused on in

the future.

E. Eukaryotic initiation factor1A interacts with eIF2A under physiological conditions

To understand whether the interaction between eIF2A and eEF1A could be occurring

under endogenous concentrations of the proteins within the cell, I first

immunoprecipitated eIF2A-HA from 2A-HA yeast lysate using anti-HA antibodies. The

104

Figure 24: Interaction between eIF2A and several proteins can be verified in reciprocal pull-down experiments. A, The TEF1 (eEF1A gene), SSB2, DED1 and NPL3 genes were

fused to GST, and partially purified from E. coli. GST-fusion proteins were incubated in the presence of HA-2A yeast lysate, and protein complexes were analyzed by 10% SDS-

PAGE followed by Western blotting with -HA antibodies. The Western blot image is shown in the upper panel, while the Coomassie stained membrane is shown in the lower panel to show the amount of GST-fusion protein in the reaction. The GST-fusion protein used in each lane of the Western blot is indicated. B, Eukaryotic initiation factor 2A was

105 immunoprecipitated with -HA antibodies from 2A-HA yeast and immunoprecipitated complexes were assessed by Western blotting with -Kem1p antibodies. Controls for total eIF2A put in the reaction are indicated.

106 immunoprecipitates were analyzed by SDS-PAGE followed by a Western blotting using

anti-eEF1A (Figure 26A). As a negative control for the specificity of the -HA antibody,

anti-Pgk1p (phosphoglycerate kinase; an enzyme that adds a phosphate to 3-

phosphoglycerate during glycolysis) was used in parallel. When comparing the amount of eEF1A between the eIF2A-HA and the Pgk1p complexes, it is apparent that significantly more protein is present with eIF2A (Figure 26A), indicating a positive interaction between these proteins. I also conducted the reciprocal experiment in which

-eEF1A antibody was used to immunoprecipitate eEF1A from 2A-HA yeast lysate

(Figure 26B). In this experiment, preimmune serum was used as a negative control for the specificity of the antibody. Preimmune serum acted as a negative control in this

experiment because the -eEF1A antibody was not purified, so other antibodies may be

present that could recognize proteins that have a molecular weight similar to that of

eEF1A. In theory, preimmune serum contains all the proteins and antibodies in the -

eEF1A serum with the exception of the antibodies that formed against eEF1A.

Significantly more eIF2A protein was observed in the reaction containing the -eEF1A antibody as compared to the reaction containing the preimmune serum (Figure 26B).

These data confirm the previous results and indicate that eEF1A and eIF2A interact in yeast.

F. Truncation analysis indicates that eIF2A interacts with domain 3 of eEF1A

The next step in understanding the nature and relevance of the interaction between eIF2A and eEF1A is to delineate which domain of eEF1A is required for the interaction. This was achieved by making various deletions of the protein and expressing these as GST-

107

Figure 25: Interactions between eIF2A and either eEF1A or Ssb2p are RNA-

independent. A, GST-fusion proteins were isolated from E. coli and incubated with 2A-

HA yeast lysate in the presence and absence of RNase A. Protein complexes were

108 analyzed for the presence of eIF2A-HA protein using -HA antibodies by Western

blotting (upper panels). Coomassie stained membrane controls are also shown (lower

panels). Both GST-PIN1 (a mammalian peptidyl-prolyl isomerase) and GST alone act as

negative controls in this experiment. Ribonucleic acid dependence of the interaction between eIF2A and Kem1p was analyzed as in Figure 24, but in the presence and absence of RNase A. B, The supernatant from each reaction shown in A was removed after

RNase A incubation, and RNA was isolated. Integrity of ribosomal RNA was then

assessed in RT-PCR experiments with primers that amplify 100 and 200 nt regions of

each rRNA. Products of RT-PCR experiments are shown, with labels indicating from

which reaction the RNA was isolated.

109 fusion proteins. Eukaryotic elongation factor 1A contains three domains that have been

implicated in the delivery of aminoacyl-tRNA molecules to the ribosome. The key

regions of the protein thought to be important in this function are domains one and two

(Figure 27A). Domain one binds guanosine nucleotide triphosphate and Mg2+, while

domain two interacts with the aminoacyl-tRNA (83). Domain three has been described to

be involved in interacting with actin, the yeast specific translation elongation factor eEF3,

and a component of the proteosome (Rpt1p; Figure 27A) (84-86). For GST-fusion

proteins, each domain was expressed by itself and in combinations with the other two

domains. The fusion proteins were partially purified from E.coli using glutathione-

sepharose beads and incubated with 2A-HA yeast cell lysate. Bound proteins were

analyzed by SDS-PAGE followed by Western blotting. These data indicate that domain

three of eEF1A is critical for the interaction with eIF2A because eIF2A only interacts

with GST-eEF1A when domain three is present (Figure 27B; compare reactions

containing full length eEF1A, domains 2 and 3, and domain 3 GST-eEF1A fusion proteins to any other lane). As a negative control, GST alone was used in parallel

reactions and did not interact with eIF2A. Interestingly, the region associated with

binding aminoacyl-tRNA is not capable by itself of mediating the interaction with eIF2A

(Figure 27B; compare any lane containing GST-eEF1A truncations to the last lane

containing GST-eEF1A domain 3).

It is possible that truncation of large regions of eEF1A could effect global protein folding

and therefore prevent binding to eIF2A. To rule out this possibility and further extend

110

Figure 26: Eukaryotic initiation factor 2A and eEF1A interact under physiological concentrations of the proteins. A, Hemaglutanin tagged eIF2A and PGK-1 protein were immunoprecipitated from 2A-HA yeast lysate, and Western blots were conducted with -

HA antibody or -eEF1A antiserum to detect eIF2A-HA or eEF1A protein, respectively.

B, Eukaryotic elongation factor 1A was immunoprecipitated with -eEF1A antiserum from 2A-HA yeast lysate, and a Western blot was performed with -HA antibody to detect levels of eIF2A that bound. Preimmune serum was used as a negative control for the specificity of the antibodies in the -eEF1A antiserum. One percent of the total protein in the immunoprecipitation reactions are shown in both A and B.

111

112 Figure 27: Domain three of eEF1A is responsible for the interaction with eIF2A. A, A domain map of eEF1A with color coded domains. Functional role of different regions of

the protein are indicated in brackets, and mutants used in later experiments are

highlighted with arrows. Amino acid numbers delineate domain boundaries and the wild

type and mutant amino acids are indicated as single letter codes before and after the

amino acid number, respectively. B, Different combinations of eEF1A domains were expressed as GST-fusion proteins in E. coli and partially purified. GST-fusion proteins were incubated with yeast lysate from 2A-HA yeast and protein complexes were analyzed by Western blotting with -HA antibodies (upper panel). The Coomassie stained membrane is used as a loading control (lower panel). The domains of eEF1A present in each reaction are represented, and colors correspond to those shown in A. C,

Full length eEF1A was expressed as a GST-fusion protein with different mutants indicated as in A. Precipitation reactions were conducted as described above in the presence of 2A-HA lysate and eIF2A protein was detected by Western blotting with -

HA antibodies. Glutathione-S-transferase alone was used as a negative control for both B and C.

113 these results, I analyzed the effects of different mutations of eEF1A on the interaction

between eEF1A and eIF2A using GST-pull-down experiments. Eukaryotic elongation

factor 1A mutants in domain two and three were examined for binding to eIF2A in 2A-

HA yeast lysate (Figure 27A). As expected, eIF2A did not co-purify with eEF1A when

domain three mutants were utilized (Figure 27C; S405P and F422A mutants). This

verifies the importance of domain three in eEF1A for the interaction with eIF2A.

G. The C-terminus of eIF2A is important for the interaction with eEF1A

It is also important to analyze which region of eIF2A is important for the interaction with

eEF1A. There are no obvious structural domains within the eIF2A protein with the

exception of a WD-repeat, which is thought to fold into a seven bladed -propeller

(Figure 28A). This domain is hypothesized to mediate binding of the eIF2A protein to

the 40S ribosomal subunit. This is based on its general use as a protein:protein

interaction domain, and its presence in another protein, RACK1, which is known to bind

the 40S ribosomal subunit (87); however, the idea that the seven bladed β-propeller

mediates the interaction with the ribosome has not been tested. To address the question

of the importance of regions of eIF2A for interaction with eEF1A, various C-terminal

eIF2A deletion yeast strains were produced by homologous recombination within the

yeast chromosome, as described for the 2A-HA strain. Full length GST-eEF1A fusion

protein was then incubated with lysate from yeast expressing different eIF2A deletions.

Western blotting indicates that the truncated proteins are expressed at similar levels in

each strain (Figure 28B). Furthermore, the ability of full length eEF1A to interact with

114

Figure 28: The region between amino acids 460 and 571 of eIF2A is important for the

interaction with eEF1A. A, A putative domain map of eIF2A listing color coded

locations of eIF2A deletions used in experiments assessing the importance of different

regions for the interaction with eEF1A. The WD-repeat domain, which is predicted to fold into a seven bladed -propeller, is annotated. B, Full length eEF1A was expressed as a fusion protein with GST and incubated with lysates from yeast strains expressing different C-terminal deletions of eIF2A. Protein complexes were analyzed by Western blotting for eIF2A protein using -HA antibodies. Western blots (upper panel) and

115 Coomassie stained membrane (lower panel) controls are shown for each truncation, which are listed in colors that correspond to the information depicted in A.

116 eIF2A is abolished upon removal of the region between amino acids 460-571 (Figure

28B; compare reactions from eIF2A 1-460 to 1-571 lysates). These data also suggest that other regions of the protein are important for interaction with eEF1A, though to a lesser extent, possibly by mediating proper folding of eIF2A protein. This feature is demonstrated upon deletion of the region between 413-460 amino acids in eIF2A, which results in partial recovery of the interaction with eEF1A (Figure 28B). Interestingly, the seven bladed β-propeller domain, which is recognized as a protein:protein interaction domain, is not required for the interaction between eEF1A and eIF2A.

H. Amino acids of eIF2A important for the interaction with eEF1A are important for

IRES-mediated repression

To begin to elucidate the functional relevance of this interaction, I conducted IRES activity measurements using the C-terminal eIF2A deletion strains described above.

These experiments were conducted using the p281-4 construct containing the minimal

URE2 IRES element. In agreement with the protein:protein interaction data, the region of eIF2A important for interacting with eEF1A is also important for repression of IRES- mediated translation initiation of the URE2 minimal IRES element (Figure 29; compare

IRES activity in the 1-460AA strain to the 1-571AA strain). Though more experiments are needed to identify the key residues in eIF2A important for interacting with eEF1A, and the specific role of eEF1A in eIF2A-mediated repression, these data suggest there is an important role for eEF1A in IRES-mediated translation initiation.

117

Figure 29: Functional analysis of C-terminal eIF2A deletion yeast strains suggest that the region between amino acids 460 and 571 is important for suppression of URE2 IRES- mediated translation initiation. The p281-4 construct containing the minimal URE2 IRES element (inset; four hairpins represent secondary structure in the 5’ UTR designed to prevent scanning associated with cap-dependent initiation, light blue represents the minimal URE2 IRES element, and orange represents the LACZ gene) was transformed into each C-terminal eIF2A deletion strain, and IRES activity was measured using the - galactosidase reporter (Figure 4). Constructs containing noncoding CTT mutations in the place of the internal AUG codon were used as negative controls in these experiments.

118 I. Conclusions

These studies demonstrate that two poly(A) lacking IRES elements, from URE2 and

GIC1 mRNAs, are regulated by eIF2A, while others that contain poly(A) stretches are not. It is possible that comparing the sequence and structure within the URE2 and GIC1 elements may provide insight into the mechanism of eIF2A-mediated regulation.

MFOLD predictions of secondary structure within the GIC1 5’ untranslated region of the mRNA can fold the UTR into several structural elements. Based on its location, one structural element that may mediate internal initiation is a large stem structure that spans the region 70 nt upstream and 10 nt downstream of the AUG codon. Predictions of this stem indicate that it folds around the internal AUG codon, which is a feature similar to the URE2 IRES element. However, the impact of base pairing with the internal AUG codon is not thought to be important for the regulation of activity based on the aforementioned studies on the URE2 IRES element. With the exception of a low stability for the GIC1 stem loop (ΔG of -12.20 kcal/mol versus -15 kcal/mol for the URE2 IRES)

and the lack of an adenosine stretch, there is little similarity to the structure of the URE2

minimal IRES element. The GIC1 stem loop contains significantly more base pairs,

several internal bulges, and the apical loop is substantially smaller (e.g. 6 nt versus 24 nt

for the GIC1 and URE2 IRES elements, respectively). Despite the presence of more base

pairs in the putative GIC1 IRES stem, the low level of stability arises from a higher

adenosine and uracil content as compared to the URE2 IRES element (73% versus 64%,

respectively).

119 The dissimilarity between the sequences and the predicted structure of the GIC1 stem and

the minimal URE2 IRES element suggests that the key feature in regulation by eIF2A is the absence of the poly(A) stretch. Studies by Gilbert et al. indicate that the poly(A)- containing IRES element in the 5’ UTR of YMR181c contains very little secondary structure and little sequence similarity to other yeast IRES elements, which highlights the importance of the poly(A)-stretch in expression from this IRES (39). Since some of the mRNAs containing invasive growth IRES elements are not transcribed until after the cell encounters glucose starvation (Wendy Gilbert, personal communication), the role of eIF2A in regulating these IRES elements may not be biologically relevant if eIF2A is not present during glucose starvation. Our laboratory has not examined whether eIF2A is present during glucose starvation, but many other stresses, including heat shock and sorbitol treatment, do result in disappearance of eIF2A mRNA (unpublished data). It is of interest to know which IRES elements are not expressed until cells are starved for glucose, and if those IRES elements rely on poly(A)-stretches and PABP for efficient internal initiation.

These studies identify novel RNA-independent binding partners of eIF2A. It is likely these partners may influence internal initiation of translation of poly(A)-lacking IRES elements through modulation of eIF2A activity. Both eEF1A and Ssb2p are proteins that could potentially influence translation initiation via protein:protein interactions with eIF2A. Since Ded1p, Kem1p, and Npl3p associate with eIF2A in an RNA-dependent manner, it is possible that these proteins may influence internal initiation without directly interacting with eIF2A, but may still act in pathways requiring the presence of eIF2A.

120 The interaction with Ssb2p and the RNA-dependent interacting partners must be

investigated more fully in order to understand their biological importance. However, the data suggests that the interaction between eIF2A and eEF1A is important for internal

initiation of translation. Interestingly, mutants in eEF1A have been previously shown to

affect protein synthesis at the initiation step (67). The studies described in this thesis

provide a possible explanation of how an elongation factor, eEF1A, could be affecting

translation initiation.

Domain three of eEF1A is important for the interaction with eIF2A. This domain has

been shown to play roles in other aspects of biology. For example, domain three has also

been implicated in interactions with Rpt1p, a component of the 19S proteosome (85).

This interaction is thought to mediate co-translational degradation of damaged proteins.

Domain three of eEF1A is also thought to interact with a yeast specific translation

elongation factor, eEF3, which catalyzes the dissociation of deacylated tRNAs from the

E-site of the ribosome (84). Amino acid sequence BLAST comparison of eIF2A with

either Rpt1p or eEF3 reveals little similarity that might confirm a role for eIF2A in either

proteosome-dependent protein degradation or release of tRNA from the E-site of the

ribosome. However, eIF2A may be involved in these processes by virtue of the presence

of structural similarities to Rpt1p or eEF3 that are not predictable from sequence

homology, and cannot be accurately predicted without structural information on eIF2A.

Domain three of eEF1A has also been shown to interact with the actin cytoskeleton.

Domain three has been implicated in the bundling of actin in Tetrahymena and

121 Saccharomyces cerevisiae (66,86). Several domain three mutants have been described

that abolish this interaction and abrogate translation in yeast (67). Basic local alignment

of the eIF2A and actin amino acid sequences reveal identity of 8/24 amino acids between

amino acids 516-539 of eIF2A, and homology in 12/24 amino acids in this region.

Interestingly, this region of eIF2A is important for the interaction with eEF1A, and

functionally involved in eIF2A-mediated repression of internal initiation on the URE2

IRES element. This implies that actin may be involved in regulating internal initiation of

translation, or eIF2A may be involved in regulation of cytoskeletal organization. Perhaps

competition between eIF2A and actin for eEF1A binding regulates one or both of these

processes.

Eukaryotic initiation factor 2A has been shown to affect cell cycle progression and actin

cytoskeletal organization (62). When eIF2A is deleted in cells carrying temperature

sensitive mutations in eIF4E, the cells accumulate in the G2 phase of the cell cycle. The

eIF4Ets/eIF2A yeast strain also exhibits pronounced morphological changes

characterized by a larger proportion of round cells. The changes in cell morphology are

thought to relate to the concurrent accumulation of actin (by approximately 50%) (62).

There have been genetic interactions described between eIF2A and several members of

the prefoldin complex, which is known to mediate proper protein folding by transferring

nascent peptides to a chaperone known as chaperonin. The prefoldin complex is known

to interact with actin as it is being translated, suggesting a functional link between eIF2A

and the actin cytoskeleton (88). Double knockout strains lacking eIF2A and each of the

following prefoldin complex members have severe slow growth phenotypes: GIM3,

122 GIM5, and PAC10 (82). Therefore, it is likely that eIF2A may play an undefined role in the regulation of actin polymer assembly.

The role for eIF2A in directing met-tRNAi to the 40S ribosomal subunit for initiation of protein synthesis has previously only been demonstrated in vitro (89); however, this data

strongly suggests in vivo binding of met-tRNAi by eIF2A during URE2 IRES-mediated translation. Experiments employing Gcn2pc mutant constructs, which result in eIF2

phosphorylation and inactivation of eIF2-mediated delivery of met-tRNAi (45,80),

indicate that the eIF2 and eIF2A-mediated pathways may compete for delivery of the

met-tRNAi to the ribosome for internal initiation on the URE2 IRES element (Figure 30).

This is the case because there is a greater reduction in IRES-mediated initiation in the

strain lacking eIF2A as compared to the wild type strain when the eIF2-mediated

pathway is knocked down by phosphorylation of eIF2α (Figure 21A).

The data presented in this thesis, in combination with previous results, suggest a model

for eIF2A-mediated cap-independent translation regulation on the URE2 IRES element

(Figure 30). In this model, I propose that eEF1A directs binding of the IRES element to

the 40S ribosomal subunit, a possibility that is derived from preliminary evidence

suggesting eEF1A interacts directly with the URE2 IRES element (data not shown). This

possibility would presumably rely on the similarity in the double stranded region of the

URE2 IRES element to double stranded regions within tRNA molecules since eEF1A

binds to tRNA during elongation. The requirement for eEF1A to recognize sequences

within double stranded regions has been documented for recognition of tRNA molecules

123 by elongation factor TU (EFTU), the bacterial homolog of eEF1A, which differentially

affects the affinity of tRNAs for the protein (90). However, the possibility that binding

of the URE2 IRES element is independent of eEF1A is also illustrated in this model. It

should be noted that while I have included eEF1A in this model, many aspects of eEF1A involvement are unknown, and direct evidence for the interaction of eEF1A with the 40S

ribosomal subunit is lacking. None the less, after binding of the URE2 IRES element

(and possibly other IRES elements lacking poly(A) stretches), eIF2A is thought to direct binding of the met-tRNAi (63,91). The data in this thesis indicate that recruitment of the

eIF2A●met-tRNAi complex may be mediated by domain 3 of eEF1A (Figure 27B and

C). Recruitment of the eIF2A●met-tRNAi complex may be specified because of the

presence of eEF1A, which would enable the eIF2A-mediated pathway to compete with

the eIF2-mediated pathway. After met-tRNAi delivery by eIF2A, eIF5B is thought to

mediate subunit joining, based on evidence from cap-dependent initiation, to form an 80S

initiation complex (92,93). It is then thought that eIF2A leaves the 80S complex prior to

transition to the elongation phase of protein synthesis. This possibility is supported by

previous work that indicates that eIF2A is present in 80S complexes, but not in

polyribosomal mRNPs (62,64). Since the sequence of eEF1A binding is unknown, eEF1A is depicted as dissociating from 80S ribosomes with eIF2A.

The pathway for eIF2A-mediated delivery of met-tRNAi is thought to be kinetically

much slower than the eIF2-mediated pathway. This is expected because of the increase in IRES activity observed in ΔeIF2A yeast, and in vitro data hinting at this possibility

(91). In fact, the physical presence of eIF2A on 80S ribosomes may prevent progression

124 to elongating ribosomes that mimics what is observed with chemical inhibitors of protein

synthesis and other proteins (67,94). This feature may result in trapping of URE2 IRES

elements in 80S initiation complexes. The proportion of IRES elements trapped in 80S

complexes could then be expected to increase each time the IRES element reaches the

bifurcation between the eIF2- and eIF2A-mediated pathways for met-tRNAi delivery. In

other words, after several rounds of initiation, mRNAs that had initiated translation

through the eIF2-mediated pathway during the first couple of rounds may be trapped in

the eIF2A pathway during later rounds. Eventually, little or no protein would be

produced from a given IRES element because all of the mRNAs are sequestered in 80S

initiation complexes containing eIF2A. However, during stresses that promote

degradation or inactivation of eIF2A, the eIF2-mediated pathway for cap-independent

initiation would predominate, which would result in production of proteins that are not produced under homeostatic conditions. As mentioned above, several stresses are known to result in turnover of eIF2A mRNA ((62); unpublished data). Furthermore, our laboratory has observed turnover or modification of eIF2A protein during various cell stresses (unpublished data).

This raises the question of the role of PABP during internal initiation of translation. It is known that PABP is important for internal initiation of poly(A)-containing IRES elements (39). The results in this thesis indicate that PABP may be involved in the eIF2- mediated pathway for met-tRNAi delivery because IRES elements containing poly(A)

stretches are not translated more efficiently in ΔeIF2A cells (Figure 20). Poly(A) binding

protein may act to recruit the 40S ribosomal subunit to the mRNA through eIF4G, which

125 is suggested by data from Gilbert et al. (39). Alternatively, there may be a third pathway

used for some IRES elements, which contains some different factors and/or a different

sequence of events resulting in internal initiation in a PABP-dependent manner. In this

case, it is still expected that eIF2 would be utilized for delivery if met-tRNAi based on the aforementioned results in the ΔeIF2A yeast cells. As mentioned above, regulation by eIF2A may not be biologically relevant for the poly(A)-rich IRES elements depending on when those mRNAs are expressed. If IRES-containing mRNAs are expressed after the cells are challenged by the stress, the eIF2A pathway may have been inactivated rendering the effects of eIF2A on those mRNAs nonphysiological.

126

Figure 30: Model of competitive pathways for met-tRNAi delivery during IRES-

mediated initiation of protein synthesis. An eIF2A (top) and eIF2-mediated (bottom)

pathway for IRES-mediated translation initiation are illustrated. Domains 1-3 of eEF1A

are shown in red, dark blue and green circles, respectively, in the eIF2A-mediated

pathway for internal initiation. Eukaryotic initiation factor 2A is represented as a light

green circle in complex with met-tRNAi. Eukaryotic initiation factor 2 (purple circle) is shown in complex with met-tRNAi and GTP (yellow circle). The ribosome (small and

large brown circles, which represent the 40S and 60S ribosomal subunits, respectively)

and the URE2 IRES element (light blue oval) are components common to both pathways.

References for the data supporting each numbered step follows. 1-Preliminary studies

during this thesis work involving biotinylated URE2 minimal IRES element mRNA

indicate that eEF1A may bind the URE2 IRES element directly, which raises the possibility that eEF1A mediates delivery of the IRES element to the 40S ribosomal

127 subunit. It is also possible that the URE2 IRES element binds the 40S ribosomal subunit independent of eEF1A. 2-Eukaryotic initiation factor 2A is known to direct AUG codon-

dependent met-tRNAi binding to 40S ribosomal subunits (63,91). 3-Work described in this thesis indicates that eEF1A and eIF2A interact through domain three of eEF1A

(green circle). I hypothesize that this interaction is a mechanism to recruit the eIF2Amet-tRNAi complex to the 40S ribosomal subunit prior to subunit joining. 4-

These components are known to be involved with 80S ribosomal subunit joining. It has

been previously shown that eIF5B is important for eIF2A-mediated synthesis of met-

puromycin in a purified system (91). Furthermore, eIF5B is known to be involved in 60S

subunit joining during cap-dependent initiation, which suggests it may be involved with cap-independent initiation of translation (92,93). 5-Eukaryotic initiation factor 2A

resides on 80S ribosomes as analyzed by sucrose density ultracentrifugation (62). The

sequence of release of eEF1A is not known, so it is depicted as bound to the

80SeIF2Amet-tRNAi complex. 6-It is anticipated that the presence of eIF2A on 80S ribosomes physically blocks the first step of translation elongation based on observations with other proteins and chemical inhibitors of protein synthesis (i.e. expression of eEF1A mutants and puromycin treatment, respectively (67,94). The fact that eIF2A accumulates in 80S-containing fractions in sucrose gradients indicates that dissociation is a very slow step in this pathway. Furthermore, it is known that eIF2A is not located in polyribosomal fractions, which is why it is depicted as dissociating from ribosomes at this step in the pathway (62). Again, the point at which eEF1A dissociates is unknown. As such, it is depicted as dissociating from ribosomal complexes with eIF2A. 7-The pathway for eIF2A-mediated met-tRNAi delivery is expected to be slower than the pathway for eIF2-

128 mediated translation initiation based on previous studies examining conversion to puromycin-sensitive 80S complexes (91), and the observations in this thesis (and by others) that demonstrates increased initiation in vivo in eIF2A yeast (62). 8-Although it is known that ternary complexes bind the ribosome prior to mRNA binding for cap- dependent initiation, it is not known whether this sequence is true for IRES-mediated translation (12). As such, the possibility that either sequence of events could be used is depicted in this model for the eIF2-mediated pathway. Subsequent steps are represented based on numerous previous studies of the eIF2 pathway for translation initiation.

129 V. Future Directions

There are several questions that are raised by these studies. These questions can be

divided into those that relate to the structure or function of the URE2 IRES element and

those that relate to eIF2A and proteins that interact with eIF2A. Since some future

studies require genetics and manipulation of the proteins integral for cell survival, yeast are a sensible organism to productively address these questions. Answering these questions will add to our knowledge of the process of internal initiation, and may indirectly permit development of novel treatments for human diseases.

A. What sequences or structures within the URE2 mRNA regulatory elements are important for modulating levels of internal initiation?

One question stemming from the studies with the URE2 minimal IRES element is the importance and relevance of the enhancer and inhibitor sequences downstream of the minimal IRES element. Thus far I have investigated the importance of the upstream inhibitory element in some detail, but information regarding downstream elements is lacking. It would be informative to investigate the boundaries and distance dependence of downstream cis-regulatory elements. This could be done using deletions of URE2 in the p281-4 reporter construct described in Figure 4. The secondary structure within the downstream cis-regulatory elements could then be delineated using structure probing experiments in combination with MFOLD predictions. These data would be useful in predicting tertiary interactions with the minimal URE2 IRES element that may result in regulation of internal initiation. Tertiary interactions could be tested by making mutations that abolish the interactions, and compensatory mutations to restore the tertiary

130 interactions. Tertiary interactions may alter the structure of the minimal URE2 IRES element, which could result in secondary structure that can more easily explain the mutant results obtained in this thesis. Therefore, identifying any tertiary interactions between the minimal URE2 IRES element and the downstream cis-regulatory elements is an important step to understanding the mechanism behind internal initiation for this IRES element.

B. Does eEF1A interact directly with the URE2 IRES element?

Another way mRNA sequences (either the URE2 IRES element or cis-elements) may regulate internal initiation is through interactions with other proteins or with the ribosome itself. One protein that may mediate internal initiation on the URE2 IRES element is eEF1A. If this protein binds the URE2 IRES element itself, which is indicated by

preliminary data, it may mediate internal initiation by facilitating delivery of the IRES to

the ribosome (Figure 30). An important experiment is to test the association between

eEF1A protein and the URE2 IRES element using RNA affinity chromatography. This

could be accomplished using biotinylated URE2 RNA in combination with yeast whole

cell lysates. The presence of eEF1A protein would be examined in the eluate of the RNA

affinity column.

It is possible that eEF1A binds the URE2 IRES element by virtue of similarities between

the IRES element and aminoacyl-tRNAs. The E286K mutant of eEF1A would be useful

in delineating the properties of the URE2 IRES element that contribute to eEF1A binding.

The reason the E286K mutant would be useful is that it is thought to have a lower affinity

131 for aminoacyl-tRNAs relative to the wild type protein. Since all of the eEF1A mutants

examined in this thesis are viable, with the exception of the F422A mutation, it is

possible to utilize lysates from yeast expressing the E286K mutant as the only form of

eEF1A in RNA affinity chromatography experiments. Comparing URE2 IRES binding

between the wild type and the E286K mutant of eEF1A would address the question of

whether eEF1A binds because of similarities between the URE2 IRES element and

aminoacyl-tRNAs. If there is a difference between the wild type and E286K mutant of

eEF1A, it would indicate that the binding is due to the similarity between the sequence or

structure of the URE2 IRES element and aminoacyl-tRNAs. As a follow up study,

purified proteins could be used in filter binding assays to examine the difference in

affinity that is imparted by abrogating the ability of eEF1A to bind aminoacyl-tRNA.

C. What is the role, if any, of the interaction between the URE2 IRES element and

eEF1A?

If eEF1A protein binds the URE2 IRES element, the functional relevance could be

addressed using in vitro sucrose density ultracentrifugation experiments. This would be

done by purifying 40S ribosomal subunits (95), and incubating them with radiolabeled

URE2 IRES RNA in the presence and absence of purified eEF1A. Subsequent sucrose

density ultracentrifugation could then be used to fractionate complexes with different molecular weights to examine co-sedimentation of the 40S ribosomal subunit and the

URE2 IRES element. If the 40S ribosomal subunit and the URE2 IRES element co-

sediment in the presence of eEF1A, but not in its absence, it will suggest that eEF1A protein delivers the URE2 IRES element to the ribosome. This could be done in the

132 presence and absence of GTP and purified eIF2A. Experiments in the presence and

absence of GTP and eIF2A will help to delineate their role in the process of internal

initiation on the URE2 IRES element (Figure 30).

D. Does PABP promote recruitment of any IRES element to 40S ribosomal subunits?

Since data from this thesis and the publication by Gilbert et al. have only inferred that translation of some IRES elements are not PABP-dependent (39), it is important to more directly rule out the possibility that PABP is important for cap-independent initiation on both URE2 and the invasive growth IRES elements lacking poly(A) stretches. As such, experiments similar to those listed above (Future Directions, Section C) could be

performed to examine PABP-dependent association of the both the URE2 IRES and invasive growth IRES elements with the 40S ribosomal subunits. Invasive growth IRES elements would be selected based on the presence and absence of poly(A) stretches.

Mutants of poly(A)-containing IRES elements lacking the poly(A) stretch would also be evaluated. Experiments assessing the importance of PABP may require the addition of other factors such as eIF2, eIF4G, and eIF3, which are thought to mediate recruitment of the 40S ribosomal subunit during cap-dependent initiation of translation (33,96). These experiments will be conducted in the presence and absence of eIF2A. Depending on the results of the above experiments involving eEF1A and PABP, these studies will provide valuable insight into the pathways for yeast internal initiation.

133 E. Does the presence of eEF1A specify the pathway for met-tRNAi delivery (eIF2 or

eIF2A)?

Based on the interest in the pathway of eIF2A-mediated initiation of cap-independent

translation, I would focus subsequent experiments on dissecting this pathway. As such, if

the results of sucrose density experiments suggest that eEF1A promotes binding of the

URE2 IRES element to the ribosome and specifies delivery of the met-tRNAi by eIF2A,

the affect of eEF1A on use of eIF2 ternary complex could be examined. This could be

done by preforming 40SeEF1AURE2 IRES complexes, and adding either purified

eIF2A, eIF2, or both in the presence of GTP and labeled met-tRNAi. Labeling of the

URE2 IRES element and the met-tRNAi with different isotopes would allow delineation

of whether the URE2 IRES and the met-tRNAi are co-sedimenting after sucrose gradient

ultracentrifugation. Western blotting for protein components of the complexes would

indicate whether the 40S and eIF2A or eIF2 were sedimenting in the same fractions as

the URE2 IRES element and the met-tRNAi. If eIF2A were the only protein mediating met-tRNAi delivery to the ribosome in the presence of eEF1A, it would support the idea

wherein the presence of eEF1A dictates the requirement for eIF2A-mediated delivery of

met-tRNAi (Figure 30). This result would also begin to assess the role of the competition

between eIF2A and eIF2-mediated pathways for delivery of the met-tRNAi to

ribosomeIRES complexes.

134 F. What factors are required for formation of stable 80S preinitiation complexes on the

URE2 IRES?

Identification of proteins that mediate formation of 80S preinitiation complexes on the

URE2 IRES element would be extremely valuable in understanding the mechanism for

internal initiation in yeast. Investigating the proteins required for formation of

preinitiation complexes at different points in the pathway would help to dissect the role of

components in the pathway for eIF2A-mediated cap-independent initiation. Presumably

the formation of stable 80S preinitiation complexes would require all the proteins necessary for the formation of URE2 IRES40S complexes, as analyzed in sucrose

density experiments above. However, it is expected that either additional factors would

be required for 80S complex formation (such as eIF1A, eIF5A, and/or eIF5B based on mammalian work), or there would be a rearrangement of factors already in the URE2

IRES40S complexes. These events would promote recruitment of the 60S ribosomal subunit, thus forming the 80S complex.

The formation of stable 80S preinitiation complexes on the URE2 IRES element would be assessed with various combinations of factors using toeprinting experiments. Factors of interest would be eEF1A, eIF2A, eIF2, and PABP for reasons discussed above.

Complexes would be formed using purified components in vitro. Toeprinting experiments rely on the ability of 80S preinitiation complexes to inhibit primer extension from radiolabeled primers. If the 80S complex is efficiently forming on the URE2 IRES element, primer extension will be abrogated when the polymerase reaches the internal

AUG because of steric interactions with the ribosome. Analysis of reactions in which

135 80S preinitiation complexes are efficiently formed will reveal the presence of a DNA fragment of a characteristic molecular weight. If 80S complexes are not efficiently formed, no DNA fragments corresponding to the number of nucleotides between the primer and the internal AUG codon will be observed. Toeprinting reactions could be done in the presence and absence of the aforementioned proteins as well as other canonical initiation factors, to investigate their role in formation of 80S preinitiation complexes at the internal AUG codon in the URE2 IRES element. Since these experiments assess the role of proteins in a complex closer to the formation of the first peptide bond (80S preinitiation complexes), the use of inactive mutants of the URE2

IRES element as negative controls would uncover the point in the pathway of eIF2A- mediated internal initiation that is important for recognition of the URE2 nucleotides important for activity.

G. Are there unidentified proteins that interact with the URE2 IRES element to alter levels of cap-independent translation? How might those proteins fit into the model for eIf2A-mediated internal initiation?

Identification of other proteins that may regulate internal initiation in yeast could be accomplished using yeast-3-hybrid experiments with a yeast cDNA library to define the proteins that interact with either the minimal URE2 IRES element itself or the cis-acting elements. An alternative method for identifying proteins that interact with either the minimal URE2 IRES element or cis-regulatory elements would be to conduct RNA affinity chromatography, isolate proteins that interact with the RNA, and then subject those proteins to SDS-PAGE followed by mass spectrometry. Inactive mutants of the

136 URE2 IRES element would be useful as negative controls in experiments designed to

identify proteins interacting specifically with the URE2 IRES element.

Assuming that proteins are identified during yeast-3-hybrid screening, it would be of

interest to examine how these proteins affect internal initiation in yeast. The first

experiments would be to conduct RNA protection assays to address where the proteins

bind the URE2 IRES element. Subsequently, changes in the structure of the URE2 IRES

element due to protein binding could be examined using structure probing in the presence

and absence of the candidate protein. Structure probing experiments may help to explain

changes in the structure of the URE2 minimal IRES element that promote or inhibit

internal initiation. These experiments would be important to understand the effect of

eEF1A binding on the URE2 IRES depending on the results of binding experiments. The

functional consequences of binding of candidate proteins to the URE2 IRES element

could then be investigated using deletion strains in combination with IRES reporter

assays. Since deletion of many proteins involved in translation are not viable, techniques

like DAmP knockdown (Decreased Abundance by mRNA Perturbation; (97)), temperature sensitive mutants, or diploid strains with haploinsufficiency may need to be utilized. Furthermore, the role of candidate proteins in eIF2A-mediated translation initiation could be assessed by making double knockout strains of eIF2A and the candidate proteins. To test whether these proteins are also involved with eIF2-mediated internal initiation, the Gcn2pc mutant could be utilized to “knockdown” the pathway by

phosphorylation of eIF2, as demonstrated above.

137 Another way to identify new proteins that regulate internal initiation would be to employ a URE2 reporter construct containing a nutritional marker downstream, but in frame with, the internal AUG codon in the URE2 IRES element. This construct could be transformed into the yeast knockout library and assayed for a gain or loss of function using growth as the readout. If clones grow more slowly than the wild type strain, on medium selecting for the nutritional marker, the gene knocked out may be involved in promoting internal initiation. If clones grow more rapidly than the wild type strain, that gene may be involved in suppressing internal initiation of translation. It would also be possible to look for other proteins that compensate for the effects of eIF2A by transferring the eIF2A deletion into the knockout library, and examining URE2 IRES-mediated translation in the double knockout strains. Genes identified in the above experiments could be useful in understanding the signals that promote IRES-mediated translation of the URE2 IRES element, and may help to elucidate the biological advantages to formation of Ure2p prion aggregates.

H. Is the interaction between eIF2A and eEF1A important for internal initiation in vivo?

The above approaches address internal initiation from the perspective of the URE2 IRES element, but there are many things that can be learned by approaching the problem from the perspective of eIF2A. The first set of experiments from this perspective would be to further examine the functional role of the interaction between eIF2A and eEF1A. There are several mutants of eEF1A that do not interact with eIF2A. Since eEF1A is an essential gene, the effect of these mutants on cap-independent initiation could be assessed using strains containing chromosomal deletions of eEF1A. The strains would survive on

138 plasmid-borne eEF1A genes. Plasmids containing wild type copies of the eEF1A gene could be swapped with mutant alleles by plasmid shuffling. Subsequent analysis of internal initiation on the URE2 IRES element in strains expressing mutant eEF1A protein

would provide an understanding of the importance of the interaction between eEF1A and

eIF2A for internal initiation.

To more directly assess the role of eIF2A in these experiments, the eIF2A gene could be

deleted in the yeast strain prior to introduction of mutant eEF1A alleles. In the presence of the eIF2A gene, strains expressing the S405P mutant of eEF1A should yield higher levels of internal initiation on the URE2 IRES element. In the absence of eIF2A, all the eEF1A mutants, including the wild type protein, would be expected to have increased

IRES-mediated expression. It is possible that the E286K eEF1A protein will also show reduced levels of internal initiation. This possibility would be in vivo evidence that eIF2A-mediated translation initiation relies on direct binding between the URE2 IRES element and eEF1A based on similarities between the URE2 IRES element and aminoacyl-tRNAs.

I. What specific regions of eIF2A are important for suppression of internal initiation?

The different regions of the eIF2A protein necessary for suppression of internal initiation are not clearly defined. To address this problem, smaller C-terminal deletions should be made in the wild type background using the method described above (81). Additionally,

N-terminal deletions should be made. The affect of these deletions on internal cap- independent initiation of the URE2 IRES element could then be assessed by transforming

139 the p281-4 construct containing the URE2 IRES element and performing β-galactosidase

assays to measure levels of cap-independent initiation. The amino acid sequence in

regions important for cap-independent initiation could then be compared to other phyla in

an attempt to understand important residues in the regulation of cap-independent

initiation, since it is known that eIF2A is conserved from yeast to humans (64). Residues

that appear to be highly conserved could be mutated to examine the affect on internal

initiation using β-galactosidase activity measurements.

J. Which regions of eIF2A are responsible for the characteristics that have been

previously observed?

It is possible that the suppressive effect on internal initiation results from the ability to

bind met-tRNAi and the ribosome. Therefore, each deletion or mutation in eIF2A should

be examined for its ability to bind both met-tRNAi and the ribosome, as done previously

(89). The met-tRNAi binding could be measured using filter binding experiments with purified eIF2A protein and labeled met-tRNAi. Comparisons with wild type protein

would be critical to assess the altered met-tRNAi binding properties of the mutant or

truncated protein. The ability of each mutant or truncation to bind the ribosome could be

tested using sucrose gradient ultracentrifugation. Western blotting of HA-tagged deletion

or mutation strains in each fraction of the gradient would tell us whether the protein is

still co-sedimenting with the 40S and 80S ribosomes, and suggest that binding is still

intact. Since the ability to bind the ribosome is thought to be related to the presence of

the seven bladed β-propeller domain, deletions and mutations that abrogate this structure

140 would be expected to cause the protein to sediment at the top of the gradient separate

from fractions containing either 40S or 80S ribosomal subunits.

K. Where is eIF2A located within the cell?

The localization of a protein in the cell can tell a lot about the function of the protein, and

the predominant role of the protein in biology. As such, I propose that localization studies of eIF2A should be done to gain insight into where eIF2A protein acts in the cell.

These studies can be done with Green Fluorescence Protein (GFP) fused to eIF2A.

Protein localization can be examined by fluorescence microscopy, and the localization

can be compared to other proteins that may interact with or colocalize with eIF2A.

Examples of other proteins that should be examined in parallel to eIF2A are eEF1A and

actin due to the link between eEF1A and actin (66,67,86). Additionally, the link

between eIF2A, actin, and members of the prefoldin complex present a strong case for

examining the co-localization of these components (62).

Other important proteins or complexes to examine for co-localization with eIF2A are decapping enzymes, other components of processing bodies (p-bodies), and protein components of the ribosome. The reason for examining proteins that co-localize with p- bodies is two-fold. First, it has been demonstrated that p-bodies are sites of translational repression, and eIF2A has been described based on its ability to repress cap-independent

initiation of translation (62,98). Secondly, I have shown that RNA-dependent

associations exist between eIF2A and components of yeast p-bodies including Ded1p and

Xrn1p (98-100). Finally, studies examining co-localization with the ribosome are

141 important to confirm that eIF2A associates with the ribosome. Furthermore, confirming

co-localization of eIF2A with the ribosome, but not with p-bodies would indicate a suppressive effect independent of p-body function. This study could be extended by assessing co-localization of different deletions of eIF2A and ribosomal components.

These would be important experiments to verify that the association with the ribosome is dependent on the presence of an intact seven bladed β-propeller domain.

L. Summary

There are still many things to learn about the URE2 IRES element and its regulation. The identification and characterization of a yeast IRES element also presents a valuable opportunity to use the genetic power of yeast to dissect the pathways involved in cap- independent initiation and their regulation. Further investigating the importance of eIF2A in IRES-mediated suppression would help us to characterize the pathways of internal initiation. Investigating PABP and invasive growth IRES elements in future studies helps to extend our knowledge of cap-independent initiation to regulation of other

IRES elements by assessing the competition between different internal initiation pathways. Internal ribosome entry site-mediated translation has been found to play roles in apoptosis, viral infection, hypoxia, and cancer, to name a few (38). Understanding the mechanisms behind cap-independent translation under both normal and disease circumstances can be used to develop treatments to target these mechanisms in disease.

For these reasons, it is very important that the above experiments be pursued.

142 References

1. Mills, A. A. (2005) Genes Dev 19, 2091-2099 2. Shaulian, E., and Karin, M. (2001) Oncogene 20, 2390-2400 3. Burns, D. M., and Richter, J. D. (2008) Genes Dev 22, 3449-3460 4. Ray, P. S., Grover, R., and Das, S. (2006) EMBO Rep 7, 404-410 5. Marine, J. C., Francoz, S., Maetens, M., Wahl, G., Toledo, F., and Lozano, G. (2006) Cell Death Differ 13, 927-934 6. Olsson, A., Manzl, C., Strasser, A., and Villunger, A. (2007) Cell Death Differ 14, 1561-1575 7. O'Brate, A., and Giannakakou, P. (2003) Drug Resist Updat 6, 313-322 8. Reiter, A. K., Bolster, D. R., Crozier, S. J., Kimball, S. R., and Jefferson, L. S. (2008) Biochem Biophys Res Commun 374, 345-350 9. de Breyne, S., Bonderoff, J. M., Chumakov, K. M., Lloyd, R. E., and Hellen, C. U. (2008) Virology 378, 118-122 10. Marissen, W. E., Gradi, A., Sonenberg, N., and Lloyd, R. E. (2000) Cell Death Differ 7, 1234-1243 11. Milon, P., Tischenko, E., Tomsic, J., Caserta, E., Folkers, G., La Teana, A., Rodnina, M. V., Pon, C. L., Boelens, R., and Gualerzi, C. O. (2006) Proc Natl Acad Sci U S A 103, 13962-13967 12. Merrick, W. C. (2004) Gene 332, 1-11 13. Komar, A. A., and Hatzoglou, M. (2005) J Biol Chem 280, 23425-23428 14. Lin, T. A., Kong, X., Haystead, T. A., Pause, A., Belsham, G., Sonenberg, N., and Lawrence, J. C., Jr. (1994) Science 266, 653-656 15. Haghighat, A., Mader, S., Pause, A., and Sonenberg, N. (1995) Embo J 14, 5701- 5709 16. Richter, J. D., and Sonenberg, N. (2005) Nature 433, 477-480 17. Sonenberg, N., and Gingras, A. C. (1998) Curr Opin Cell Biol 10, 268-275 18. Scheper, G. C., van Kollenburg, B., Hu, J., Luo, Y., Goss, D. J., and Proud, C. G. (2002) J Biol Chem 277, 3303-3309 19. Kimball, S. R. (1999) Int J Biochem Cell Biol 31, 25-29 20. Dever, T. E., Yang, W., Astrom, S., Bystrom, A. S., and Hinnebusch, A. G. (1995) Mol Cell Biol 15, 6351-6363 21. Chen, J. J., and London, I. M. (1995) Trends Biochem Sci 20, 105-108 22. Su, Q., Wang, S., Gao, H. Q., Kazemi, S., Harding, H. P., Ron, D., and Koromilas, A. E. (2008) J Biol Chem 283, 469-475 23. Duncan, R., and Hershey, J. W. (1985) J Biol Chem 260, 5486-5492 24. van Gorp, A. G., van der Vos, K. E., Brenkman, A. B., Bremer, A., van den Broek, N., Zwartkruis, F., Hershey, J. W., Burgering, B. M., Calkhoven, C. F., and Coffer, P. J. (2008) Oncogene 25. Chappell, S. A., and Mauro, V. P. (2003) J Biol Chem 278, 33793-33800 26. Kieft, J. S., Zhou, K., Jubin, R., Murray, M. G., Lau, J. Y., and Doudna, J. A. (1999) J Mol Biol 292, 513-529

143 27. Le Quesne, J. P., Stoneley, M., Fraser, G. A., and Willis, A. E. (2001) J Mol Biol 310, 111-126 28. Pilipenko, E. V., Blinov, V. M., Chernov, B. K., Dmitrieva, T. M., and Agol, V. I. (1989) Nucleic Acids Res 17, 5701-5711 29. Mitchell, S. A., Spriggs, K. A., Coldwell, M. J., Jackson, R. J., and Willis, A. E. (2003) Mol Cell 11, 757-771 30. Pickering, B. M., Mitchell, S. A., Spriggs, K. A., Stoneley, M., and Willis, A. E. (2004) Mol Cell Biol 24, 5595-5605 31. Garlapati, S., and Wang, C. C. (2004) J Biol Chem 279, 3389-3397 32. Herbreteau, C. H., Weill, L., Decimo, D., Prevot, D., Darlix, J. L., Sargueil, B., and Ohlmann, T. (2005) Nat Struct Mol Biol 12, 1001-1007 33. Pestova, T. V., Kolupaeva, V. G., Lomakin, I. B., Pilipenko, E. V., Shatsky, I. N., Agol, V. I., and Hellen, C. U. (2001) Proc Natl Acad Sci U S A 98, 7029-7036 34. Pisarev, A. V., Shirokikh, N. E., and Hellen, C. U. (2005) C R Biol 328, 589-605 35. Pelletier, J. a. S., N. (1988) Nature 334, 320-325 36. Sarnow, P., Cevallos, R. C., and Jan, E. (2005) Biochem Soc Trans 33, 1479-1482 37. Bushell, M., Stoneley, M., Kong, Y. W., Hamilton, T. L., Spriggs, K. A., Dobbyn, H. C., Qin, X., Sarnow, P., and Willis, A. E. (2006) Mol Cell 23, 401-412 38. Spriggs, K. A., Stoneley, M., Bushell, M., and Willis, A. E. (2008) Biol Cell 100, 27-38 39. Gilbert, W. V., Zhou, K., Butler, T. K., and Doudna, J. A. (2007) Science 317, 1224-1227 40. Johannes, G., Carter, M. S., Eisen, M. B., Brown, P. O., and Sarnow, P. (1999) Proc Natl Acad Sci U S A 96, 13118-13123 41. Thomas, J. D., and Johannes, G. J. (2007) Rna 13, 1116-1131 42. Han, B., and Zhang, J. T. (2002) Mol Cell Biol 22, 7372-7384 43. Hecht, K., Bailey, J. E., and Minas, W. (2002) FEMS Yeast Res 2, 215-224 44. Verge, V., Vonlanthen, M., Masson, J. M., Trachsel, H., and Altmann, M. (2004) Rna 10, 277-286 45. Thompson, S. R., Gulyas, K. D., and Sarnow, P. (2001) Proc Natl Acad Sci U S A 98, 12972-12977 46. Galkin, O., Bentley, A. A., Gupta, S., Compton, B. A., Mazumder, B., Kinzy, T. G., Merrick, W. C., Hatzoglou, M., Pestova, T. V., Hellen, C. U., and Komar, A. A. (2007) Rna 13, 2116-2128 47. Keene, J. D., and Tenenbaum, S. A. (2002) Mol Cell 9, 1161-1167 48. Vyas, K., Chaudhuri, S., Leaman, D. W., Komar, A. A., Musiyenko, A., Barik, S., and Mazumder, B. (2009) Mol Cell Biol 29, 458-470 49. Shirokikh, N. E., and Spirin, A. S. (2008) Proc Natl Acad Sci U S A 105, 10738- 10743 50. Komar, A. A., Lesnik, T., Cullin, C., Merrick, W. C., Trachsel, H., and Altmann, M. (2003) Embo J 22, 1199-1209 51. Cooper, T. G. (2002) FEMS Microbiol Rev 26, 223-238 52. Bousset, L., Belrhali, H., Janin, J., Melki, R., and Morera, S. (2001) Structure 9, 39-46 53. Lian, H. Y., Jiang, Y., Zhang, H., Jones, G. W., and Perrett, S. (2006) Biochim Biophys Acta 1764, 535-545

144 54. Rai, R., Tate, J. J., and Cooper, T. G. (2003) J Biol Chem 278, 12826-12833 55. Bai, M., Zhou, J. M., and Perrett, S. (2004) J Biol Chem 279, 50025-50030 56. Lewis, S. M., Veyrier, A., Hosszu Ungureanu, N., Bonnal, S., Vagner, S., and Holcik, M. (2007) Mol Biol Cell 18, 1302-1311 57. Paek, K. Y., Kim, C. S., Park, S. M., Kim, J. H., and Jang, S. K. (2008) J Virol 82, 12082-12093 58. Kim, J. H., Paek, K. Y., Choi, K., Kim, T. D., Hahm, B., Kim, K. T., and Jang, S. K. (2003) Mol Cell Biol 23, 708-720 59. Cornelis, S., Tinton, S. A., Schepens, B., Bruynooghe, Y., and Beyaert, R. (2005) Nucleic Acids Res 33, 3095-3108 60. Tischendorf, J. J., Beger, C., Korf, M., Manns, M. P., and Kruger, M. (2004) Arch Virol 149, 1955-1970 61. Altmann, M., Muller, P. P., Wittmer, B., Ruchti, F., Lanker, S., and Trachsel, H. (1993) Embo J 12, 3997-4003 62. Komar, A. A., Gross, S. R., Barth-Baus, D., Strachan, R., Hensold, J. O., Goss Kinzy, T., and Merrick, W. C. (2005) J Biol Chem 280, 15601-15611 63. Merrick, W. C., and Anderson, W. F. (1975) J Biol Chem 250, 1197-1206 64. Zoll, W. L., Horton, L. E., Komar, A. A., Hensold, J. O., and Merrick, W. C. (2002) J Biol Chem 277, 37079-37087 65. Ito, H., Fukuda, Y., Murata, K., and Kimura, A. (1983) J Bacteriol 153, 163-168 66. Gross, S. R., and Kinzy, T. G. (2005) Nat Struct Mol Biol 12, 772-778 67. Gross, S. R., and Kinzy, T. G. (2007) Mol Cell Biol 27, 1974-1989 68. Zhou, W., Edelman, G. M., and Mauro, V. P. (2001) Proc Natl Acad Sci U S A 98, 1531-1536 69. Mathews, D. H., Sabina, J., Zuker, M., and Turner, D. H. (1999) J Mol Biol 288, 911-940 70. Zuker, M. (2003) Nucleic Acids Res 31, 3406-3415 71. Reineke, E. L., Lam, M., Liu, Q., Liu, Y., Stanya, K. J., Chang, K. S., Means, A. R., and Kao, H. Y. (2008) Mol Cell Biol 28, 997-1006 72. Reineke, L. C., Komar, A. A., Caprara, M. G., and Merrick, W. C. (2008) J Biol Chem 283, 19011-19025 73. Silver, P. A., Keegan, L. P., and Ptashne, M. (1984) Proc Natl Acad Sci U S A 81, 5951-5955 74. Ehresmann, C., Baudin, F., Mougel, M., Romby, P., Ebel, J. P., and Ehresmann, B. (1987) Nucleic Acids Res 15, 9109-9128 75. Groebe, D. R., and Uhlenbeck, O. C. (1988) Nucleic Acids Res 16, 11725-11735 76. Chappell, S. A., Edelman, G. M., and Mauro, V. P. (2004) Proc Natl Acad Sci U S A 101, 9590-9594 77. Brierley, I., Gilbert, R. J., and Pennell, S. (2008) Biochem Soc Trans 36, 684-689 78. Green, L., Kim, C. H., Bustamante, C., and Tinoco, I., Jr. (2008) J Mol Biol 375, 511-528 79. Yaman, I., Fernandez, J., Liu, H., Caprara, M., Komar, A. A., Koromilas, A. E., Zhou, L., Snider, M. D., Scheuner, D., Kaufman, R. J., and Hatzoglou, M. (2003) Cell 113, 519-531 80. Ramirez, M., Wek, R. C., Vazquez de Aldana, C. R., Jackson, B. M., Freeman, B., and Hinnebusch, A. G. (1992) Mol Cell Biol 12, 5801-5815

145 81. Longtine, M. S., McKenzie, A., 3rd, Demarini, D. J., Shah, N. G., Wach, A., Brachat, A., Philippsen, P., and Pringle, J. R. (1998) Yeast 14, 953-961 82. Tong, A. H., Lesage, G., Bader, G. D., Ding, H., Xu, H., Xin, X., Young, J., Berriz, G. F., Brost, R. L., Chang, M., Chen, Y., Cheng, X., Chua, G., Friesen, H., Goldberg, D. S., Haynes, J., Humphries, C., He, G., Hussein, S., Ke, L., Krogan, N., Li, Z., Levinson, J. N., Lu, H., Menard, P., Munyana, C., Parsons, A. B., Ryan, O., Tonikian, R., Roberts, T., Sdicu, A. M., Shapiro, J., Sheikh, B., Suter, B., Wong, S. L., Zhang, L. V., Zhu, H., Burd, C. G., Munro, S., Sander, C., Rine, J., Greenblatt, J., Peter, M., Bretscher, A., Bell, G., Roth, F. P., Brown, G. W., Andrews, B., Bussey, H., and Boone, C. (2004) Science 303, 808-813 83. Ozturk, S. B., Vishnu, M. R., Olarewaju, O., Starita, L. M., Masison, D. C., and Kinzy, T. G. (2006) Genetics 174, 651-663 84. Anand, M., Balar, B., Ulloque, R., Gross, S. R., and Kinzy, T. G. (2006) J Biol Chem 281, 32318-32326 85. Chuang, S. M., Chen, L., Lambertson, D., Anand, M., Kinzy, T. G., and Madura, K. (2005) Mol Cell Biol 25, 403-413 86. Morita, K., Bunai, F., and Numata, O. (2008) Zoolog Sci 25, 22-29 87. Coyle, S. M., Gilbert, W. V., and Doudna, J. A. (2008) Mol Cell Biol 88. Vainberg, I. E., Lewis, S. A., Rommelaere, H., Ampe, C., Vandekerckhove, J., Klein, H. L., and Cowan, N. J. (1998) Cell 93, 863-873 89. Anderson, W. C. M. a. W. F. (1975) Journal of Biological Chemistry 250, 1107- 1206 90. Dale, T., and Uhlenbeck, O. C. (2005) Trends Biochem Sci 30, 659-665 91. Adams, S. L., Safer, B., Anderson, W. F., and Merrick, W. C. (1975) J Biol Chem 250, 9083-9089 92. Acker, M. G., Shin, B. S., Nanda, J. S., Saini, A. K., Dever, T. E., and Lorsch, J. R. (2009) J Mol Biol 385, 491-506 93. Pestova, T. V., Lomakin, I. B., Lee, J. H., Choi, S. K., Dever, T. E., and Hellen, C. U. (2000) Nature 403, 332-335 94. Petersen, C. P., Bordeleau, M. E., Pelletier, J., and Sharp, P. A. (2006) Mol Cell 21, 533-542 95. Pestova, T. V., Hellen, C. U., and Shatsky, I. N. (1996) Mol Cell Biol 16, 6859- 6869 96. Pestova, T. V., and Kolupaeva, V. G. (2002) Genes Dev 16, 2906-2922 97. Schuldiner, M., Collins, S. R., Thompson, N. J., Denic, V., Bhamidipati, A., Punna, T., Ihmels, J., Andrews, B., Boone, C., Greenblatt, J. F., Weissman, J. S., and Krogan, N. J. (2005) Cell 123, 507-519 98. Parker, R., and Sheth, U. (2007) Mol Cell 25, 635-646 99. Aizer, A., Brody, Y., Ler, L. W., Sonenberg, N., Singer, R. H., and Shav-Tal, Y. (2008) Mol Biol Cell 19, 4154-4166 100. Beckham, C., Hilliker, A., Cziko, A. M., Noueiry, A., Ramaswami, M., and Parker, R. (2008) Mol Biol Cell 19, 984-993

146