Discovery and Characterization of Novel Inhibitors of Protein Kinase R

by Unkyung Shin

Department of Biochemistry McGill University, Montreal

August 2013

A thesis submitted to McGill University in partial fulfillment of the requirements of the degree of Master of Science

 Unkyung Shin 2013

TABLE OF CONTENTS

TABLE OF CONTENTS ...... iii LIST OF FIGURES ...... v ABSTRACT ...... vi RÉ SUMÉ ...... vii ACKNOWLEDGEMENTS ...... viii PREFACE ...... ix CONTRIBUTION OF AUTHORS ...... x ORIGINAL CONTRIBUTION TO KNOWLEDGE ...... xi CHAPTER 1: INTRODUCTION ...... 12 1.1 Protein Synthesis ...... 12 1.2 Eukaryotic Translation Initiation ...... 13 1.2.1 Cap-dependent initiation...... 13 1.2.2 IRES-mediated initiation ...... 15 1.3 Regulation of Translation Initiation in Eukaryotes ...... 17 1.3.1 eIF4F complex formation and eIF4E availability ...... 17 1.3.2 Ternary complex formation and eIF2 phosphorylation ...... 19 1.4 Kinases Mediating eIF2 Phosphorylation ...... 20 1.4.1 Heme-regulated inhibitor (HRI or EIF2AK1) ...... 20 1.4.2 dsRNA-dependent protein kinase (PKR or EIF2AK2) ...... 22 1.4.3 PKR-like endoplasmic reticulum kinase (PERK or EIF2AK3) ...... 25 1.4.4 General control non-depressible 2 (GCN2 or EIF2AK4) ...... 26 1.5 Pathological Implications of eIF2α Kinases ...... 28 1.6 Small Molecule Inhibitors of eIF2α Kinases ...... 30 1.6.1 Quercetin – inhibitor of HRI ...... 31 1.6.2 2-Aminopurine – inhibitor of PKR ...... 31 1.6.3 C16 – inhibitor of PKR ...... 32 1.6.4 GSK PERK inhibitor – inhibitor of PERK ...... 32 1.6.5 Syk inhibitor – inhibitor of GCN2 ...... 33

iii

1.7 Compound Screening Strategy ...... 33 1.8 Overview and Rationale for the Thesis ...... 37 CHAPTER 2: RESULTS ...... 38 2.1 Identification of Hymenialdisine and Isohymenialdisine as Stimulators of Translation ...... 38 2.2 Isohymenialdisine Inhibits PKR-mediated Phosphorylation of eIF2...... 49 CHAPTER 3: DISCUSSION ...... 55 CHAPTER 4: MATERIALS & METHODS ...... 61 4.1 Materials and General Methods ...... 61 4.2 Isolation of Hymenialdisine and Its Derivatives from Stylissa massa ...... 61 4.3 In vitro Transcription and Translation Reactions...... 62 4.4 In vitro Kinase Assays...... 63 4.5 Computational Mapping of the PKR Structure ...... 64 4.6 Docking of Hymenialdisine and Isohymenialdisine ...... 65 CHAPTER 5: CONCLUSION...... 66 ACKNOWLEDGEMENTS ...... 67

REFERENCES ...... 68

LIST OF FIGURES

CHAPTER 1

Figure 1.1 Overview of eukaryotic cap-dependent translation initiation...... 14 Figure 1.2 Key regulatory mechanisms of eukaryotic translation initiation...... 18 Figure 1.3 Overview of eIF2α phosphorylation...... 21 Figure 1.4 Chemical structures of eIF2α kinase inhibitors...... 31 Figure 1.5 Outline of the compound screening strategy ...... 35

CHAPTER 2

Figure 2.1 Effects of Hymenialdisine, Debromohymenialdisine, Isohymeni- aldisine on translation reactions in rabbit reticulocyte lysates (RRL)...... 39 Figure 2.2 Effects of Hymenialdisine Debromohymenialdisine, Isohymeni- aldisine on translation from the HCV and CrPV IRESes in RRL...... 42 Figure 2.3 Effects of Cmpds 1 - 3 on translation in Krebs, Wheat Germ, and E. coli S30 extracts...... 44 Figure 2.4 Effect of adding Cmpds 1 - 3 to actively translating extracts...... 47 Figure 2.5 Inhibitory effect of Isohymenialdisine on GST-PKR-KD-mediated eIF2α phosphorylation and GST-PKR-KD autophosphorylation...... 51

ABSTRACT

Double-stranded RNA-dependent protein kinase (PKR) is an interferon- inducible serine/threonine kinase that plays a crucial role in innate immunity against viral infections. Activated PKR downregulates global protein synthesis through phosphorylation of the α subunit of eukaryotic initiation factor 2 (eIF2) and induces transcriptional output of cellular mRNAs implicated in limiting viral spread. Accumulating evidence suggests that PKR is involved in a multitude of other cellular processes including signal transduction, apoptosis, cell growth and differentiation, as well as the pathogenesis of several neurodegenerative diseases.

In this context, small molecules targeting PKR would be useful as tools to help understand the intricacies of such biological pathways. In screening a library of natural and synthetic products for eukaryotic translation modulators, we identified two small molecules, isohymenialdisine and hymenialdisine, that exhibit stimulatory effects on translation. Characterization of the mode of action of isohymenialdisine revealed that it directly acts on PKR by inhibiting its autophosphorylation activity and perturbs the PKR-eIF2α phosphorylation axis.

RÉ SUMÉ

La protéine kinase dépendante de l’ARN double brin (PKR) est une kinase à sérine/threonine don’t l’activité est induite par l’interféron qui joue une rôle clé dans la défense immunitaire contre les infections virales. Lorsque PKR est active, elle régule négativement la synthèse protéique via la phosphorylation de la sous-unité α du facteur d’initiation de la traduction 2 (eIF2) et induit la transcription de certains ARN messagers cellulaires qui codent pour des gènes impliqués dans le contrôle de la propagation virale. PKR participe aussi à une multitude d’autres processus cellulaires qui incluent le contrôle de l’activité certaines voies de signalisation, l’apoptose, la croissance cellulaire, la différentiation ainsi que dans la pathogenèse de plusieurs maladies dégénératives.

À cet égard, l’identification de molécules inhibitrices ciblant dette protéine pourrait faciliter son étude et la compréhension de ses fonctions biologiques. Lors d’un criblage d’une librairie de composés naturels pouvant moduler la synthèse protéique eucaryote, nous avons identifié deux molécules, l’isohymenialdisine et l’hymenialdisine, qui présentent un effet stimulateur sur la traduction. La caractérisation du mécanisme d’action de l’isohymenialdisine à démontré qu’elle agit directement sur PKR en bloquant son autophosphorylation et, par conséquent, celle d’eIF2α.

vii

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my supervisor Jerry Pelletier for his guidance, support, and insightful discussions throughout my graduate studies.

His passion for science and contagious enthusiasm for new discoveries have been a constant source of inspiration. I am also grateful for the members of my

Research Advisory Committee, Dr. David Y. Thomas and Dr. Martin Schmeing for their helpful comments and suggestions.

My appreciation extends to the current and previous members of the lab

(Aouod Agenor, Regina Cencic, Jennifer Chu, Gabriela Galicia-Vázquez,

Alexandra Katigbak, Amina Kreps, Teresa Lee, Rayelle Maiga, Abba Malina,

John Mills, Sophia Moraitis, Zeina Nasr, Francis Robert, Patrick Sénéchal and

Tanner Shpiruk) for their help and friendship in addition to making the working environment stimulating and enjoyable. Special thanks go to Francis for translating the abstract in French and Jennifer for proofreading this thesis.

I also wish to express my gratitude to the collaborators, David Williams and Raymond Andersen at UBC, and Dima Kozakov, David Hall, Dmitri Beglov and Sandor Vadja at Boston University for their remarkable work and effort.

As always, I am forever indebted to my parents, brother and close friends for their unwavering support, encouragement, and faith in me throughout, for which my mere expression of thanks likewise does not suffice.

viii

PREFACE

This is a manuscript-based thesis of which Chapters 2, 3 and 4 are part of a research article submitted for publication for which I am the first author.

Chapter 2, 3 and 4 :

Unkyung Shin, David E. Williams, Dima Kozakov, David Hall, Dmitri Beglov, Sandor Vadja, Raymond J. Andersen, and Jerry Pelletier (2013). Isohymenialdisine and Hymenialdisine: Novel Inhibitor of Protein Kinase R- mediated Phosphorylation of eIF-2. Manuscript submitted.

CONTRIBUTION OF AUTHORS

Chapter 2 :

I performed all experiments within the manuscript except: isolation of hymenialdisine, debromohymenialdisine and isohymenialdisine from Axinella carteri was performed by David E. Williams under the supervision of Raymond J.

Andersen; strucutral modeling analysis of the PKR ATP binding pocket with the compounds was performed by Dima Kozakov, David Hall, Dmitri Beglov and

Sandor Vadja. Peter Beal’s laboratory kindly provided the GST-PKR-KD expression vector. The work in Chapter 2 was completed under the supervision of

Jerry Pelletier.

ORIGINAL CONTRIBUTION TO KNOWLEDGE

Chapter 2 :

 Discovered Isohymenialdisine and Hymenialdisine as stimulators of

translation in vitro

 Determined that the mechanism of action of Isohymenialdisine is through

inhibition of PKR autophosphorylaion activity, which in turn reduces eIF2α

phosphorylation

CHAPTER 1: INTRODUCTION

1.1 Protein Synthesis

Protein synthesis is a fundamental, yet energetically demanding process requiring 30-40% of cellular ATP (or GTP) (1), as to fulfill protein requirements for nearly all structural, catalytic and regulatory functions in cells. Dysregulation of protein synthesis can contribute to a diverse range of pathological conditions including tumorigenesis and neurological syndromes (2, 3). In view of the above, tight regulation of protein synthesis is essential to precisely define the proteome and maintain homeostasis while matching the needs of a cell with the supply of energy and other resources such as amino acids. Translational control of protein synthesis provides effective means to rapidly and reversibly adapt to changes in the normal physiological state, and to regulate gene expression patterns both spatially and temporally (4).

Translation occurs in three distinct phases – initiation, elongation and termination. The process initiates by the assembly of small and large ribosomal subunits on an mRNA transcript at the appropriate start codon. Assembled ribosome proceeds to the elongation phase to catalyze iterative cycles of peptide bond formation between two amino acids which are selected based on the mRNA sequence. Translation terminates upon ribosomal recognition of a stop codon

(UAA, UAG or UGA) that triggers release of a fully synthesized polypeptide from the translation machinery (5).

1.2 Eukaryotic Translation Initiation

In eukaryotes, initiation is rate-limiting and generally the most highly regulated phase during translation (6-9). The majority of cellular mRNAs initiate translation by canonical cap-dependent mechanism that involves recognition of a cap structure (m7GpppN where N is any nucleotide) at the 5’ end of transcript, followed by ribosome scanning to locate an appropriate start codon. Alternatively, numerous viral RNAs and certain cellular mRNAs initiate translation in a cap- independent manner by recruiting the translation machinery to a structural motif in the 5’-untranslated region (UTR) called the internal ribosome entry site (IRES)

(10).

1.2.1 Cap-dependent initiation

Cap-dependent initiation begins with the formation of a ternary complex

(TC) that comprises eukaryotic initiation factor 2 coupled to GTP (eIF2•GTP) and

Met initiator methionyl-tRNA (Met-tRNAi ) (Figure 1.1, Step 1). The TC binds the

40S ribosomal subunit with the help of eIFs 1, 1A, 3 and 5, forming the 43S pre- initiation complex (PIC) (Step 2). The mRNA transcript is activated by a group of factors, the eIF4F complex, eIF4B and poly(A)-binding protein (PABP), to become a messenger ribonucleoprotein (mRNP) (Step 3). The eIF4F complex consists of: (i) eIF4E, a cap-binding protein, (ii) eIF4A, an ATP-dependent

DEAD-box RNA helicase that unwinds secondary structure in the 5’-UTR for the incoming PIC, and (iii) eIF4G, a scaffold protein that bridges the 43S complex

13 and transcript. eIF4B functions to enhance the helicase activity of eIF4A. PABP is believed to stimulate translation by promoting mRNA circularization through interacting with 3’ poly(A) tail and eIF4G (4, 11-14).

Figure 1.1 Overview of eukaryotic cap-dependent translation initiation. See text for details.

Upon recruitment to the activated mRNP (Step 4), the PIC scans the 5’-

UTR in 5’-to-3’ direction to locate the appropriate AUG start codon (Step 5).

Following base pairing between the start codon and the anticodon of Met-

Met tRNAi , the 48S initiation complex is formed, and eIF2•GTP converts to eIF2•GDP via hydrolysis triggered by GTPase-activating protein eIF5 (Step 6). A

GTPase factor eIF5B, in conjunction with eIF1A, promotes joining of the 60S subunit and dissociation of several eIFs including eIF2•GDP. Upon hydrolysis of eIF5B•GTP, eIF5B•GDP and eIF1A dissociate from the translation machinery, and the 80S initiation complex forms and enters the elongation phase (4, 11, 13,

14).

1.2.2 IRES-mediated initiation

Alternative to the conventional cap-dependent scanning mode, translation initiation can occur without the cap structure by a highly structured RNA element called the internal ribosome entry site (IRES). Many viral RNAs harbor IRESes which can functionally substitute for the cap structure and many initiation factors

(in some cases, all), and facilitate hijacking of the host translational machinery

(15). In addition, a subset of cellular mRNAs, generally estimated as 5-10%, contains IRESes (5), and many of them have been identified to encode transcription factors, translation initiation factors, survival factors, and proteins involved in cell cycle progression and stress response. IRESes of such mRNAs serve as regulatory elements to initiate translation in a cap-independent manner

15 under conditions when global translation is compromised, such as hypoxia, apoptosis, and starvation (16).

IRES elements were first discovered in the RNA genomes of the picornaviruses, poliovirus (PV) (17) and encephalomyocarditis virus (EMCV)

(18). Since then, various other viral IRESes have been identified and classified into four different types based on their predicted secondary structures, locations relative to the start codon and factor requirements. Type 1 and 2 IRESes, such as

PV and EMCV IRESes respectively, require eIF4A and eIF4G but not eIF4E in

Met order to recruit the PIC (eIF1, eIF1A, eIF3, eIF5 and eIF2/GTP/Met-tRNAi ).

Whereas Type 1 IRESes are present upstream of the start codon, Type 2 IRESes are near the start codon allowing initiation at the site of ribosome binding. Type 3

IRESes, for example hepatitis C virus (HCV) IRES, recruit the 40S subunit directly at the start codon utilizing a few factors, eIF2 and eIF3. Lastly, Type 4

IRESes, like cricket paralysis virus (CrPV) IRES, require neither eIFs nor Met-

Met tRNAi , and initiate translation at a non-AUG codon. Due to the different factor requirements compared to the canonical initiation, IRES-driven initiation may not be subject to certain types of translational control, such as eIF4E sequestration (all types of IRESes) or eIF2 phosphorylation (Type 4 IRESes) (4, 19, 20). Unlike viral IRESes, cellular mRNA IRESes are not readily classifiable as they show little similarity in their structures. Nonetheless, they are thought to function as

Type 1 and 2 IRESes that require recruitment of the eIFA-eIFG complex (21).

1.3 Regulation of Translation Initiation in Eukaryotes

As previously described, translation initiation in eukaryotes is a dynamic process accomplished by the coordinated action of at least nine initiation factors

(4). By modulating activities or availabilities of the eIFs, the rate of translation can be effectively regulated. The two important regulatory points that are well characterized in the literature are: (i) formation of eIF4F complexes which relies on the availability of eIF4E (Figure 1.2A), and (ii) formation of ternary complexes which depends on the phosphorylation status of eIF2 (Figure 1.2B) (5,

11, 22).

1.3.1 eIF4F complex formation and eIF4E availability

The multi-subunit eIF4F complex (eIF4A, eIF4E and eIF4G) is a key translational machinery component that bridges the 43S complex and mRNA.

One of its subunits eIF4E, which is essential for binding the cap structure at the 5’ end of most cellular mRNAs, appears to be limiting, and thus its availability often determines translation rates (23, 24). eIF4E is regulated by a family of its repressor proteins called eIF4E binding proteins (4E-BPs) (Figure 2.1A). Hypo- phosphorylated 4E-BP binds eIF4E through the eIF4E-binding motif Tyr-X-X-X-

X-Leu-Φ (where X is any amino acid and Φ is a hydrophobic amino acid), where eIF4G can also binds (25). As a result, it prevents the eIF4E-eIF4G interaction and assembly of an active eIF4F complex, and this leads to attenuation of translation. In response to extracellular stimuli, such as growth factors, hormones,

17 and cytokines (26, 27), 4E-BP becomes hyperphosphorylated at multiple serine/threonine residues by kinases that are regulated by the rapamycin-sensitive kinase mTOR (mammalian target of rapamycin) (28, 29) or rapamyin-insensitive kinase Pim-2 (30). Consequently, eIF4E is disengaged from 4E-BP and binds eIF4G to stimulate initiation. Phosphatases targeting hyperphosphorylated 4E-BP may also be regulated by mTOR, Pim-2 or other signaling pathways (31).

Figure 1.2 Key regulatory mechanisms of eukaryotic translation initiation. The regulatory mechanisms of (A) the eIF4F complex formation, and (B) the ternary complex formation. See text for details.

1.3.2 Ternary complex formation and eIF2 phosphorylation

Another well-characterized target for modulating translation is the ternary

Met complex (eIF2•GTP-Met-tRNAi ) formation step (Figure 1.2B). This step is

Met crucial for delivering Met-tRNAi to the 40S subunit and is predominantly dependent on the phosphorylation status of eIF2. During initiation, upon base

Met paring between the start codon and the anticodon of Met-tRNAi , eIF2 is converted from its active GTP-bound state (eIF2•GTP) to an inactive GDP-bound state (eIF2•GDP) which is then released from the ribosome as a stable binary complex. In order to reconstitute a functional TC for a new round of initiation, eIF2•GDP needs to undergo a GTP-exchange reaction to regenerate eIF2•GTP, which is catalyzed by eIF2B (22, 32).

Under various environmental stimuli and cellular stresses, the exchange of

GDP/GTP is regulated via eIF2 phosphorylation. eIF2 is a trimeric complex consisting of α, β and γ subunits, and phosphorylation of the α subunit at Ser 51 increases the affinity of eIF2•GDP for eIF2B, leading to sequestration of eIF2•GDP-eIF2B complexes. Since eIF2 is present in excess of eIF2B, small changes in the phosphorylation status of eIF2 results in a significant inhibitory effect on global translation (33, 34). Conversely, eIF2 dephosphorylation can occur with eIF2α phosphatases, namely a constitutive phosphatase PP1/CReP and a stress-induced phosphatase PP1/GADD34, to recover protein synthesis (35, 36).

In mammals, four eIF2α kinases have been previously identified: double- stranded RNA-dependent protein kinase (PKR), general control non-depressible 2

(GCN2), PKR-like endoplasmic reticulum kinase (PERK) and heme-regulated inhibitor (HRI) (Figure 1.3). They share a conserved kinase domain but have distinct regulatory domains that enable them to act as “sentinels” against diverse stress signals (37). In the following sections, each member of the eIF2α kinase family as well as the associated chemical inhibitor(s) will be described in detail, which will be of direct relevance to the topic of this thesis.

1.4 Kinases Mediating eIF2 Phosphorylation

1.4.1 Heme-regulated inhibitor (HRI or EIF2AK1)

Heme-regulated inhibitor (HRI) or eIF2α kinase 1 (EIF2AK1) was the first discovered member of the eIF2α kinase family, and it was found in heme- deficient reticulocytes that exhibited protein synthesis inhibition (38). The well- established role of HRI is to coordinate the synthesis of heme-binding proteins called globins with heme availability for hemoglobin production in erythroid tissues. When HRI senses a low level of heme (or iron that leads to decrease in heme), it phosphorylates eIF2α, which in turn reduces global translation. Being the most abundantly expressed peptides in reticulocytes, the levels of α- and β- globin proteins are severely downregulated. (39, 40). As a result, the production of globin in excess of heme is prevented, or otherwise excessive globins will precipitate and form inclusions that may lead to proteotoxicity (41).

Figure 1.3 Overview of eIF2α phosphorylation. (A) Domain structure of the four mammalian eIF2α kinases. Heme-binding sites (H), kinase domains (KD), dsRNA-binding domains (dsRBD1 and 2), ER-luminal domain, signal sequence (SS), stress sensor, transmembrane domain (TM), cytosolic domain, RWD domain (RWD), pseudokinase domain (ΨKD), histidyl- tRNA synthetase-related domain (HisRS), and C-terminal domain (CTD) are shown. Domains involved in sensing stress signals or activation are in green, kinase domains in red, and other domains in grey. Domains are drawn to scale. Figure modified from Donnelly et al. (2013). (B) Control of eIF2α phosphorylation. Note that each eIF2α kinase can be activated by other stimuli or stresses. See text for details.

HRI is a multidomain protein composed of an amino-terminal domain

(NTD) and a kinase domain (KD). It contains two heme-binding sites: one at the

NTD that appears to bind heme constitutively, and another in the insert region of the KD which binds heme reversibly and regulates the kinase activity (42) (Figure

1.3A). HRI exists as a homodimer and heme-induced repression of HRI is thought to be associated with intermolecular disulfide bonds between subunits within the dimer (43). HRI activation induces a conformational change within the NTD that alters its inhibitory effect on the KD (44).

Although HRI is primarily reported to function in red blood cell precursors, it is expressed in almost all tissues including brain, lung, heart, liver, kidney and pancreas suggesting the possibility of its role in other tissues (45).

Also, HRI has been reported to respond to other stress conditions such as arsenite treatment, heat shock, osmotic stress and oxidative stress (44, 46).

1.4.2 dsRNA-dependent protein kinase (PKR or EIF2AK2)

Double-stranded (ds)RNA-dependent protein kinase (PKR) or eIF2α kinase 2 (EIF2AK2) was discovered after the observation that dsRNA inhibits translation in cell extracts prepared from interferon (IFN)-treated cells (47, 48).

This was found to be mediated by an interferon-induced, dsRNA-activated protein factor with kinase activity that is now known as PKR (49). PKR has also been referred as DAI, dsI, dsRNA-PK, P1 kinase, p68 or protein kinase R (50-56).

PKR plays a crucial role in innate immunity against viral infections. Its activation upon binding to viral dsRNAs mediates eIF2α phosphorylation, which leads to global translation arrest (57). Moreover, studies have shown other functions of PKR in signal transduction, apoptosis, cell growth and differentiation.

PKR is engaged in a number of signal transduction pathways induced by dsRNA, pro-inflammatory stimuli, cytokines, growth factors, oxidative stress and other extracellular stimuli that lead to regulation of transcription factors such as NF-κB,

IRF-1, STAT, p53 and ATF-3 (33, 58-63). For example, PKR mediates IFN or dsRNA signaling cascades by binding to STAT transcription factors and modulating their activities (60, 61); PKR also mediates inflammatory or immune signaling pathways by promoting nuclear translocation of a transcription factor

NF-κB through activation of IκB kinase complex (IKK complex) (64, 65). PKR is additionally involved in apoptosis induced by different stimuli including TNF-α, lipopolysaccharide, an ER-stress inducer tunicamycin, serum starvation, and viral infection (33, 66, 67). PKR-mediated apoptosis is particularly implicated in neuronal cell death and neurodegenerative disorders (68-72). A potential role of

PKR in cell growth control has been evidenced by several studies. Growth was suppressed after expression of human PKR in yeast, insect and mammalian cells

(73-75). Malignant transformation in NIH 3T3 cells and tumor growth in nude mice were observed upon expression of catalytically inactive mutants of human

PKR (75, 76). The implication of PKR in diverse biological processes is further remarked by the regulatory functions of PKR in cell differentiation that are

23 exemplified in mammalian keratinocytes, osteoblasts, osteoclasts, hematopoietic stem cells and skeletal muscle cells (77-81).

The structure of PKR consists of a dsRNA-binding domain (dsRBD) at the

N-terminus that contains two dsRNA-binding motifs, dsRBM1 and dsRBM2, and a kinase domain (KD) at the C-terminus (Figure 1.3A). PKR exists as a monomeric latent state, and the autoinhibitory effect of the dsRBD regulates the kinase domain. PKR can be activated by highly structured RNAs of cellular or synthetic (such as poly (I:C)) origin in addition to viral dsRNAs (33, 82). The minimal and optimal dsRNA lengths required for PKR activation appear to be about 30bp and 85bp respectively (83). Binding of dsRNA to the dsRBD, promotes dimerization of PKR and enables autophosphorylation and activation

(33). A striking feature of PKR activation is the “bell-shaped curve” observed for its dependence on dsRNA concentration. PKR activation begins to occur at very low dsRNA concentrations, but the activation disappears at high dsRNA concentrations. This suggests that binding of two PKR monomers to the same dsRNA molecule is required for efficient dimerization (i.e. 2 PKR monomers : 1 dsRNA), whereas high dsRNA concentrations cause the dimers to dissociate into inactive monomers which bind dsRNA molecules separately (i.e. 1 PKR monomer : 1 dsRNA) (83, 84). PKR can also be activated in a dsRNA- independent manner by a polyanionic heparin or cellular protein activator of PKR known as PACT. Both have been shown to activate PKR by interacting with the kinase domain (85, 86).

1.4.3 PKR-like endoplasmic reticulum kinase (PERK or EIF2AK3)

PKR-like endoplasmic reticulum kinase (PERK) or eIF2α kinase 3

(EIF2AK3) is an ER-localized type I transmembrane protein that senses the protein-folding capacity in the ER lumen. Accumulation of unfolded or misfolded proteins causes “ER stress” which triggers several signaling pathways collectively termed as “unfolded protein response” (UPR) to restore the ER homeostasis (87,

88). In metazoans, PERK is one of the three major transducers of the UPR (the others are IRE1 and ATF6) (89). Upon ER stress, it reduces the influx of new polypeptides into the ER by attenuating global protein synthesis through eIF2α phosphorylation. Paradoxically, certain mRNAs containing short upstream open reading frames (uORFs) in their 5’-UTR are preferentially translated. One of them encodes a basic leucine zipper transcription factor ATF4 (activating transcription factor 4) that subsequently induces CHOP (C/EBP-homologous protein) and

GADD34 (growth arrest and DNA damage-inducible 34) (90, 91). CHOP is a transcription factor that regulates genes involved in apoptosis. Thus, PERK activation at mild ER stress is protective under pro-survival signaling pathways but can potentiate pro-apoptotic signaling pathways, for instance, upon prolonged

ER stress (92). GADD34 is a protein phosphatase 1 (PP1)-interacting protein that triggers PP1 to dephosphorylate eIF2α, thereby reverse the global translational block and provide feedback control of the UPR (93).

Structural analyses of PERK revealed four distinct elements, which are signal sequence (SS) for its translocation into the ER, ER-luminal stress sensor domain, transmembrane domain (TM), and cytosolic kinase domain (KD) (37, 94)

(Figure 1.3A). ER stress sensed by the luminal sensors triggers dimerization, oligomerization and autophosphorylation of PERK which leads to eIF2α phosphorylation by the cytosolic kinases (89). Several studies have postulated models for regulatory mechanisms of the luminal sensor domain (88, 90). Early work suggested a competition model in which the luminal sensor domain competes with unfolded proteins for binding to a chaperone protein BiP (B-cell immunoglobulin binding protein), and dissociation of the luminal sensor domain from BiP triggers oligomerization and activation (95, 96). Subsequent work proposed a ligand binding model in which direct binding of unfolded protein to the luminal sensor domain promotes oligomerization and activation. It postulated that BiP rather serves to buffer the luminal sensor and thereby ensures that only the appropriate levels of stress trigger the response pathway (97, 98).

1.4.4 General control non-depressible 2 (GCN2 or EIF2AK4)

General control non-depressible 2 (GCN2) or eIF2α kinase 4 (EIF2AK4) primarily responds to amino acid (AA) deprivation by monitoring levels of uncharged tRNA. During nutritional stress, deficiency in one more AA results in accumulation of the corresponding uncharged tRNAs, which in turn is sensed by

GCN2 and triggers activation of GCN2 through dimerization and autophosphorylation. Subsequent eIF2α phosphorylation by GCN2 leads to attenuated global protein synthesis allowing cells to conserve resources, as well as preferential translation of select mRNAs (99-101). GCN4 and ATF4 in yeast and mammals respectively are such upregulated mRNAs due to the presence of

26 multiple uORFs in their 5’-UTR. They are both transcription factors that GCN4 is involved in amino acid and purine metabolism, and ATF4 is involved in amino acid metabolism and transport (102-104). In addition to amino acid starvation, studies have shown a number of GCN2 activators in yeast or mammalian cells, including UV irradiation, glucose or purine deprivation, high salinity and oxidative stress (105-109). Under various stresses, GCN2 has also been associated with lipid metabolism, feeding behavior, synaptic plasticity and memory (110-

112).

GCN2 has a complex architecture composed of: (i) a RWD domain

(present in RING finger and WD repeat containing proteins and DEXDc-like helicases) that interacts with the activator protein GCN1, (ii) a pseudokinase domain lacking key catalytic residues (ΨKD), (iii) a kinase domain (KD), (iv) a histidyl-tRNA synthetase-related domain (HisRS) that binds uncharged tRNAs, and (v) a C-terminal domain (CTD) that facilitates GCN2 dimerization, tRNA binding and ribosome binding (37, 113, 114) (Figure 1.3A). Interdomain interactions among the HisRS, CTD and kinase domains appear to regulate the kinase activity. GCN2 retains its inactive state by the interaction between the kinase domain and the flanking regions of HisRS and CTD. Under stress conditions, accumulating uncharged tRNAs bind HisRS, which in turn releases and activates the kinase domain (115).

1.5 Pathological Implications of eIF2α Kinases

In response to a variety of cellular and environmental stresses, translational output and transcriptional profiles that govern complex biological processes must be precisely and adequately altered to ensure efficient adaptation.

Given that phosphorylation status of eIF2α is critical for appropriately mediating various stress response pathways, it is not surprising that aberrant activities or deficiency of any of the eIF2α kinases may lead to diverse pathological conditions (37, 116, 117).

HRI, which is predominantly in erythroid cells and thus closely associated with heme metabolism and hemoglobin synthesis, has been implicated with red blood cell (RBC) disorders. Homozygous HRI-null (HRI-/-) mice under heme deficiency suffer from aggregation of globins devoid of heme, resulting in a hyperchromic and normocytic anemia, compensatory erythroid hyperplasia, and accelerated apoptosis in bone marrow and spleen. This suggested the regulatory roles of HRI for globin synthesis and survival of erythroid precursors during heme deprivation (39). A subsequent study with HRI-/- mice bearing inherited

RBC defects demonstrated potential protective roles of HRI during the disease states. HRI-/- mice suffering erythropoietic protoporphyria (EPP) were much more severely affected, and HRI-/- mice suffering β-thalassemia became embryonically lethal (118).

Elevated PKR activity has been involved in multiple pathological situations including neurodegeneration, muscle atrophy, obesity, diabetes,

28 hematological toxicity and cancer. Extensive studies have reported significant increases in phosphorylated PKR (p-PKR) in brain tissues of patients with

Alzheimer’s disease (AD) (119), Huntington’s disease (HD), Parkinson’s diseases (PD) (71, 120) and Creutzfeldt-Jakob disease (CJD) (121), and in spinal cord tissues of patients with amyotrophic lateral sclerosis (ALS) (122). Increased levels of p-PKR were also observed in transgenic mice that expressed high levels of amyloid β peptides (Aβ) in neurons, which is a hallmark of AD (123).

Activated PKR has been linked to skeletal muscle atrophy as PKR-dependent depression of protein synthesis and induction of protein degradation in response to cachectic factors were observed in murine myotubes (124). In addition, a striking increase in p-PKR was observed in white adipose tissues from genetically obese mice and liver tissues from insulin resistant mice, suggesting a possible contribution of higher PKR activity to obesity and type 2 diabetes (125).

Phosphorylation levels of PKR was also higher in bone marrow mononuclear cells of patients with myelodysplastic syndromes (MDS), which is a term for a group of hematological disorders, compared to that of healthy donors (126).

Other studies have demonstrated increased levels of PKR expression and activity in melanoma, breast carcinoma, hepatocellular carcinoma and colon adeno- carcinoma (127-131).

Missense or truncation mutation within the kinase domain of PEKR is linked to a rare human genetic disorder called Wolcott-Rallison syndrome

(WRS). WRS is characterized by neonatal onset of diabetes with a severe loss of pancreatic β-cells, which later develops into skeletal dysplasia, growth

29 retardation and multiple systematic dysfunctions (116, 132). PERK-/- mice that were born with a normal pancreas developed early onset diabetes with a loss of

β-cells and later recapitulated many of the symptoms of WRS in human (133,

134).

GCN2 as an amino acid deprivation sensor is a key participant of metabolic adaption to starvation and is involved in amino acid and lipid metabolism. When any of the essential amino acids is absent, not only the rate of global protein synthesis is repressed, but also the expression of fatty acid synthase (FAS), a key factor in lipid synthesis, is downregulated (135). In liver tissues of GCN2-/- mice, lipid production occurred unabated during prolonged leucine-deficient diet, resulting in an excessive build-up of fat known as liver steatosis (110).

1.6 Small Molecule Inhibitors of eIF2α Kinases

Several small molecule inhibitors targeting different eIF2α kinases have been identified. Application of some of these compounds has enabled chemical dissection of the relevant pathways and provided new insights into the interplay between pathways and associated pathological states. The following eIF2α kinases inhibitors will be described below in detail: Quercetin, 2-aminopurine,

C16, GSK PERK inhibitor and Syk inhibitor.

Figure 1.4 Chemical structures of eIF2α kinase inhibitors.

1.6.1 Quercetin – inhibitor of HRI

Quercetin (3,3’,4’,5,7-pentahydroxyflavone) is a polyphenolic, plant- derived flavonoid that has been known to inhibit human HRI in an ATP- competitive manner (136). It was further shown to have similar effectiveness for canine and rodent HRI reflecting the high degree of homology in their ATP- binding sites, which suggested that among mammalian HRIs, the ATP-binding and catalysis have been conserved (137)

1.6.2 2-Aminopurine – inhibitor of PKR

2-aminopurine (2-AP) is an adenine analog that has been shown to inhibit

PKR in vitro and in vivo. Kinetic analyses in vitro indicated that 2-AP acts as a

31 competitive inhibitor with respect to ATP. In subsequent in vivo experiments,

HeLa cells were treated with poly (I:C) to activate PKR, and the activation was inhibited by the addition of 2-AP. 2-AP inhibits PKR activity at millimolar concentrations both in vitro and in vivo (138). Unfortunately, it inhibits other kinases at these relatively high concentrations (139).

1.6.3 C16 – inhibitor of PKR

C16 (8-(imidazol-4-ylmethylene)-6H-azolidino[5,4-g]benzothiazol-7-one) is an imidazolo-oxindole derivative which was discovered as an inhibitor of PKR by screening a library of ATP-binding site directed inhibitors. Whereas 2-AP inhibited PKR activity at relatively high millimolar concentrations, C16 effectively inhibited PKR at micromolar concentrations in rabbit reticulocyte lysates (140). C16 has been utilized in several studies to explore PKR-mediated pathways in response to different stress conditions (80, 141-145) (Details will be discussed in Chapter 3).

1.6.4 GSK PERK inhibitor – inhibitor of PERK

GSK PERK inhibitor (1-[5-(4-amino-7-methyl-7H-pyrrolo[2,3-d]pyrimidin-5-yl)-2,3-dihydro-1H-indol-1-yl]-2-[3-fluoro-5-(trifluoromethyl)phenyl]etha- none; 3-Fluoro-GSK2606414) was discovered by a group at GlaxoSmithKline

(GSK) through screening of a kinase inhibitor library, lead optimization and structure-activity relationship (SAR) studies. Its close structural analog

GSK2606414 has been shown to potently and selectively inhibit PERK autophosphorylation activity in A459 cells in response to an ER stress inducer thapsigargin (146).

1.6.5 Syk inhibitor – inhibitor of GCN2

During the process of screening a library of kinase inhibitors and ATP- analogs for inhibitors of GCN2, an oxindole derivative compound (2-(2- aminoethylamino)-4-(3-trifluoromethylanilino)-pyrimidine-5-carboxamide, dihy- drochloride, dihydrate), which has been previously identified as a spleen tyrosine kinase (Syk) inhibitor (147), exhibited inhibitory effect on GCN2-mediated eIF2α phosphorylation upon UV irradiation. Structural modeling analysis suggested that

Syk inhibitor is suitably accommodated in the ATP binding pocket of GCN2

(148).

1.7 Compound Screening Strategy

Small molecules are often defined as carbon-based low molecular weight compounds (<800 Daltons) that have biological activities (149). The use of small molecules to perturb a cellular process is a powerful approach to dissect its function in complex biological systems.

In an effort to identify small molecules that modulate eukaryotic protein synthesis, a multiplex screen based on in vitro translation of a bicistronic mRNA reporter has been developed (150) (Figure 1.5). The bicistronic mRNA contains

33 an upstream firefly luciferase (FF) cistron that initiates translation in a cap- dependent manner, followed by the HCV IRES and renilla luciferase (Ren) cistron that initiates translation via IRES-mediated manner. The production of FF and Ren in mouse Krebs-2 ascites extracts is quantitatively assessed to identify small molecules that: (i) modulate cap-dependent translation and thereby affect

FF production (Figure 1.5A, Step 1), (ii) bind the stem-loop structure formed by

CAG repeats in the 5’-UTR, perturb ribosome scanning and affect FF production

(Step 2), (iii) modulate elongation (Step 3) or termination (Step 4) and affect both

FF and Ren, and (iv) modify IRES-mediated initiation and affect Ren production

(Step5). In this pilot screen, compounds that show more than 3-fold inhibition of translation (relative light units or RLU < 0.33) are considered as positive “hits.”

Selected ligands are tested in the counterscreen by incubating with FF or Ren proteins to eliminate inhibitors of luciferase enzyme activity. Compounds that have not been eliminated by the counterscreen are re-tested in a secondary assay that involves in vitro translation using [35S]methionine followed by SDS-PAGE analysis to monitor protein levels of FF and Ren.

Compounds that pass the counterscreen and secondary assays are then tested for their effects on different IRESes such as the EMCV IRES and CrPV

IRES. Because the factor requirement of each IRES is different for ribosome recruitment, the sensitivity of each IRES to a compound may provide clues to the potential compound target(s).

Figure 1.5 Outline of the compound screening strategy (A) Schematic diagram of FF/HCV/Ren mRNA. A small molecule can have an effect on in vitro translation at one of five sites: by modulating cap-dependent initiation (Step 1), by binding the RNA motif in the 5’-UTR and perturbing ribosome scanning (Step 2), by modulating elongation (Step 3) or termination (Step 4), or by modulating IRES-mediated initiation (Step 5). (B) Flow chart for identifying novel modulators of protein synthesis in eukaryotes.

Additionally, candidate compounds are tested in other cell-free translation systems including rabbit reticulocyte lysate (RRL), wheat germ and E.coli S30 extracts. This can inform whether the target is conserved in mammalians or plants, or in eukaryotes or prokaryotes. In addition, the degree of cap-dependency of translation varies among different extracts – translation in Krebs and wheat germ extracts is generally cap-dependent, whereas translation in RRL is much less so

(151).

Subsequently, compounds are subject to various assays to further characterize the modes of action. Examples of these assays include ribosome- binding assays, kinetic studies analyzing actively translating extracts in the presence of a compound, and computer modeling analyses.

The screening strategy described above has been a powerful tool to discover modulators of protein synthesis. It has been internally validated by the detection of several known inhibitors including cycloheximide and puromycin

(150). As well, it has newly identified hippuristanol as an initiation inhibitor (152), cytotrienine A as an elongation inhibitor (153), and ethidium bromide and acriflavine, which are nucleic acid intercalators, as inhibitors of HCV IRES- mediated initiation (154).

1.8 Overview and Rationale for the Thesis

As explained in the previous sections, protein synthesis is a tightly and complicatedly regulated process. Despite the advances in understanding the molecular mechanisms of protein synthesis, which have been noteworthy, putative or known pathways that regulate translation need to be explored further for a better understanding of physiological and pathological implications.

Classical genetic methods to investigate the above aspects have limitations especially in case of studying overlapping pathways (155). In this context, identification of compounds that selectively target different components of overlapping pathways can be beneficial, and the use of these compounds in chemical genetic studies can complement the classical genetic approaches.

In this work, a forward chemical genetic screen of natural and synthetic products was conducted in efforts to identify novel modulators of protein synthesis. The thesis highlights the characterization of the positive “hits” from the screen by unveiling their mechanisms of action.

CHAPTER 2: RESULTS

2.1 Identification of Hymenialdisine and Isohymenialdisine as

Stimulators of Translation In Vitro.

During the course of screening natural and synthetic products using in vitro translation assays for modulators of protein synthesis, we noticed a novel stimulatory activity during bioassay-guided fractionation of one of the extracts.

Hymenialdisine (1), debromohymenialdisine (2) and isohymenialdisine (3) were purified from Stylissa massa extracts and individually characterized in rabbit reticulocyte lysates (RRL) programmed with FF/EMCV/Ren (Figures 2.1A, B).

Anisomycin and hippuristanol, inhibitors of elongation and of initiation respectively, blocked production of both firefly (FF) and renilla (Ren) luciferase, as expected (Figures 2.1C, D). Titration of hymenialdisine and isohymenialdisine stimulated production of both FF and Ren, with a more pronounced effect observed on EMCV-driven Ren production (Figures 2.1C, D).

Debromohymenialdisine had no effect on the translational output of

FF/EMCV/Ren (Figure 2.1C, compare lanes 7 – 9 to 1). Neither hymenialdisine nor isohymenialdisine stimulated luciferase enzyme activity per se when incubated in the presence of FF or Ren proteins, indicating that the results are not due to specific stimulation of luciferase enzyme activity (U. Shin, data not shown).

Figure 2.1 Effects of Hymenialdisine, Debromohymenialdisine, Isohymeni- aldisine on translation reactions in rabbit reticulocyte lysates (RRL). See figure legend on the following page.

Figure 2.1 Effects of Hymenialdisine, Debromohymenialdisine, Isohymeni- aldisine on translation reactions in rabbit reticulocyte lysates (RRL).

(A) Chemical structures of Hymenialdisine (1), Debromohymenialdisine (2) and Isohymenialdisine (3). (B) Schematic diagram of pKS/FF/EMCV/Ren. The plasmid was linearized with BamHI and used in in vitro transcriptions to generate mRNA. (C) Representative autoradiogram demonstrating the effects of compounds (Cmpds) 1 – 3 on translation of FF/EMCV/Ren mRNA. RRL was programmed with FF/EMCV/Ren mRNA (24 ng/μl) in the presence of [35S]methionine and the indicated amounts of compound. Following incubation at 30oC for 1 hr, samples were fractionated by electrophoreses in 10% SDS- polyacrylamide gels. Gels were treated with EN3HANCE, dried and exposed to Blue Ray film (VWR). The position of migration of firefly (FF) and renilla (Ren) luciferase is denoted. (D) Hymenialdisine and Isohymenialdisine stimulate production of firefly and renilla luciferase from FF/EMCV/Ren mRNA. Summary of translations performed in RRL. The relative luciferase values are set relative to vehicle (0.25% DMSO) controls (which were set at 1). On average, the firefly and renilla luciferase values in the vehicle controls were ~400,000 and 1,500,000 RLUs. n=3; Bars denote the error of the mean.

To determine if hymenialdisine and isohymenialdisine could also stimulate translation directed from other IRESes, we titrated compounds (Cmpds)

1 – 3 in RRL that had been programmed with FF/HCV/Ren or Ren/CrPV/FF

(Figure 2.2A). In the case of FF/HCV/Ren, hymenialdisine and isohymenialdisine stimulated both FF and Ren production (Figure 2.2B). In contrast, the CrPV IRES was recalcitrant to stimulation by any of the compounds used in this study, whereas production of Ren luciferase was increased in the presence of hymenialdisine and isohymenialdisine (Figure 2.2C). We attribute these observations to differential factor requirements of IRES-mediated initiation.

Whereas both EMCV and HCV IRES require at least eIF2 and eIF3, CrPV IRES

Met requires neither eIFs nor Met-tRNAi to initiate translation (4, 19, 20).

To assess if these effects would also be observed in other mammalian translation systems, we tested extracts prepared from Krebs-2 cells and programmed with FF/EMCV/Ren (Figure 2.3A). We observed a stimulatory effect on production of FF and Ren in the presence of 50 μM isohymenialdisine, but not at similar concentrations of hymenialdisine or debromohymenialdisine.

None of the compounds had any effect on FF production when translations were performed in wheat germ extracts (Figure 2.3B) or in E. coli coupled transcription-translation extracts (Figure 2.3C). Taken together, these results indicate that the effects of isohymenialdisine on increasing the yield of FF and

Ren from FF/EMCV/Ren are not unique to RRL but extend to Krebs extracts.

Isohymenialdisine is however not effective in plant or bacterial translation systems.

Figure 2.2 Effects of Hymenialdisine Debromohymenialdisine, Isohymeni- aldisine on translation from the HCV and CrPV IRESes in RRL. See figure legend on the following page.

Figure 2.2 Effects of Hymenialdisine Debromohymenialdisine, Isohymeni- aldisine on translation from the HCV and CrPV IRESes in RRL.

(A) Schematic diagram of bicistronic reporters harboring the HCV or CrPV IRESes. Each plasmid was linearized with BamHI for in vitro transcription reactions. (B) Effects of Cmpds 1 – 3 on translation from FF/HCV/Ren mRNA in RRL. Left: Representative autoradiogram demonstrating the relative translation efficiencies of FF/HCV/Ren mRNA. Following in vitro translations, samples were fractionated in 10% SDS-polyacrylamide gels. Gels were treated with EN3HANCE, dried and exposed to Blue Ray film (VWR). The position of migration of FF and Ren luciferase is denoted by arrows to the right. Right: Quantitation of effects of Cmpds 1 – 3 on translation from FF/HCV/Ren mRNA in RRL. Luciferase activities were normalized to the activity obtained in vehicle (0.25% DMSO) controls. On average, the firefly and renilla luciferase values in the vehicle controls were ~2,000,000 (FF) and ~150,000 (Ren) RLUs for FF/HCV/Ren. n=3; Bars denote the error of the mean. (C) Effects of Cmpds 1 – 3 on translation from Ren/CrPV/FF mRNA in RRL. Left: Representative autorad demonstrating the relative translation efficiencies of Ren/CrPV/FF mRNA. Following in vitro translations, samples were fractionated in 10% SDS- polyacrylamide gels. Gels were treated with EN3HANCE, dried and exposed to Blue Ray film (VWR). The position of migration of FF and Ren luciferase is denoted by arrows to the right. Right: Quantitation of effects of Cmpds 1 – 3 on translation from Ren/CrPV/FF mRNA in RRL. Luciferase activities were normalized to the activity obtained in vehicle (0.25% DMSO) controls. On average, the luciferase values in the vehicle controls were ~200,000 (Ren) and ~800,000 (FF) RLUs for Ren/CrPV/FF. n=3; Bars denote the error of the mean.

Figure 2.3 Effects of Cmpds 1 - 3 on translation in Krebs, Wheat Germ, and E. coli S30 extracts. See figure legend on the following page.

Figure 2.3 Effects of Cmpds 1 - 3 on translation in Krebs, Wheat Germ, and E. coli S30 extracts.

(A) Effect of compounds on translation in Krebs extracts programmed with FF/EMCV/Ren mRNA (24 ng/μl). Luciferase values were normalized to vehicle (0.25% DMSO) controls. On average, the firefly and renilla luciferase values in the vehicle controls were ~70,000 and ~2,500,000 RLUs. n =3; Bars denote the error of the mean. (B) Effect of compounds (50 μM) on translation in wheat germ extracts. Wheat germ extracts were programmed with FF/HCV/Ren mRNA (24 ng/μl) and firefly luciferase activity monitored and normalized to the activity obtained in the control translation reactions containing 0.25% DMSO. Renilla luciferase production was not monitored since the HCV IRES is not functional in wheat germ extracts (154). On average, the firefly luciferase values in the vehicle controls were ~350,000 RLUs. n=3; Bars denote the error of the mean. (C) Effect of compounds (50 μM) on coupled transcription-translation reactions in E. coli S30 extracts. Extracts were programmed with pBESTluc DNA (50 ng/μl) and the obtained firefly luciferase activity normalized to vehicle (0.25% DMSO) controls. On average, the firefly luciferase values in the vehicle controls were ~500,000 RLUs. n=3; Bars denote the error of the mean.

To obtain insight into the kinetics of stimulation exerted by Cmpds 1 – 3, we documented the kinetics of stimulation upon addition of compounds to actively translating extracts (Figure 2.4). Here, RRLs were programmed with FF

Luc mRNA (Figure 2.4A) and compounds added 15 mins after start of the translation reactions. In this situation, addition of cycloheximide results in a rapid cessation of translation elongation with little protein production occurring over the ensuing 75 minutes (Figure 2.4B). Homoharringtonine, allows ribosome run- off since it has low affinity for translating polysomes but inhibits the first step of elongation (156). When added to translating RRLs, protein synthesis continues unabated for ~5 – 7 mins while ribosomes run-off pre-loaded mRNA template, and then stops (Figure 2.4C). Addition of vehicle (DMSO) was accompanied by an increase in FF Luc production from 15 – 45 min followed by a plateauing and little change in FF Luc output beyond 60 min. Similar results were obtained when debromohymenialdisine was added to translating RRLs (Figure 2.4B). In contrast, addition of isohymenialdisine caused a significant increase in FF Luc output that was linear for a much longer period of time than observed from the vehicle control extracts (Figure 2.4B). Addition of hymenialdisine also stimulated production of FF Luc, but to a lower degree than obtained with isohymenialdisine

(Figure 2.4C).

Figure 2.4 Effect of adding Cmpds 1 - 3 to actively translating extracts. See figure legend on the following page.

Figure 2.4 Effect of adding Cmpds 1 - 3 to actively translating extracts.

(A) Schematic diagram of T3LucpA+. The template was linearized with BamHI and used for in vitro transcription reactions. (B-C) Translations in RRL were initiated by addition of T3LucpA+ mRNA (4 ng/μl). Aliquots for luciferase activity were removed at various time points. Compounds were added at the times indicated by a downward arrow (15 min). Hymenialdisine and derivatives were tested at final concentration of 25 μM. DMSO was present at 0.25%. (D) Western blot documenting eIF2α phosphorylation status from RRL lacking or containing exogenously supplied FF/HCV/Ren mRNA (4 ng/μl). Samples for analysis were removed at the indicated times and probed with antibodies directed to the proteins shown to the right of the blots. (E-F) Inhibition of PKR prolongs translational output in RRL. Translation extracts were pre-incubated with the indicated compounds (5 μM) for 20 min, at which point translations were initiated by the addition of FF/HCV/Ren mRNA (4 ng/μl). Aliquots were taken at the indicated times and cap-dependent firefly luciferase activity measured. (G) Western blot illustrating phosphorylation status of eIF2α from RRL programmed with FF/HCV/Ren mRNA (4 ng/μl) in the presence of the indicated compounds at 5 μM. Samples were taken 5 and 90 mins after the start of translation (s.e., short exposure; l.e., long exposure).

2.2 Isohymenialdisine Inhibits PKR-mediated Phosphorylation of eIF2.

One mechanism that can limit the translational output of mammalian in vitro translation extracts is phosphorylation of eIF2α (see Introduction). Indeed, in vitro transcribed RNA using phage polymerases often contains double stranded product due to a snap-back phenomenon which can lead to activation of PKR in translation extracts (157). To directly test this, we probed RRL that had been incubated in the absence or presence of in vitro transcribed RNA (Figure 2.4D).

The results clearly show phosphorylation of eIF2α occurring as a function of time upon addition of in vitro transcribed RNA (Figure 2.4D). We therefore tested whether inhibitors to the four known eIF2α kinases could mimic the results obtained with isohymenialdisine in vitro. Addition of small molecule inhibitors targeting GCN2 (Figure 2.4E), PERK (Figure 2.4F) and HRI (Figure 2.4F) (Syk inhibitor, GSK PERK inhibitor and Quercetin respectively) had little influence on

FF production when added to RRLs programmed with FF/HCV/Ren mRNA. In contrast, addition of a PKR inhibitor (C16 or PKRi) increased FF production to levels that resembled those obtained with isohymenialdisine (Figure 2.4E).

Western blot analysis revealed that PKRi and isohymenialdisine (3) diminished the status of phospho-eIF2α in RRL, 5 and 90 mins after their addition to translating RRL extracts (Figure 2.4G). These results suggest that isohymenialdisine may be stimulating translation in RRL by blocking PKR- mediated phosphorylation of eIF2α. The effect of hymenialdisine was not as pronounced as that of isohymenialdisine, and paralleled the relative potencies of

49 these two compounds for stimulation of translation (Figure 2.4C). We thus focused our attention on the characterization of isohymenialdisine.

To test if PKR was a direct target of isohymenialdisine (3), we expressed and purified recombinant His6-eIF2α and GST-PKR-kinase domain (KD). GST-

PKR-KD has been previously used to characterize small molecule inhibitors of

PKR (140) and is capable of phosphorylating His6-eIF2α in vitro

(Figure 2.5A, compare lanes 4 and 5 to 1-3). PKRi and isohymenialdisine (3) inhibited eIF2α phosphorylation in vitro consistent with isohymenialdisine being a direct PKR inhibitor. To extend these results and demonstrate a direct effect of isohymenialdisine (3) on PKR kinase activity, we incubated GST-PKR-KD with

[γ-32P]ATP in the presence of vehicle (Figure 2.5B, lane 3), PKRi (lanes 4-5) or isohymenialdisine (lanes 6-7). In this scenario, the heterologous GST dimerization domain functionally substitutes for double stranded RNA and enables dimerization and activation of the kinase domain (158). We found that

PKRi and isohymenialdisine (3) inhibited autophosphorylation of PKR-KD

(compare lanes 4-7 to lanes 2 and 3) consistent with isohymenialdisine being able to directly inhibit PKR activation.

To obtain insight into the interactions between hymenialdisine and isohymenialdisine with PKR, we mapped the kinase domain of PKR for druggable ‘hot spots’ (159) and docked the two compounds to PKR based on the mapping results (160). The PKR structure was derived from the PKR kinase domain-eIF2α-AMP-PNP complex by stripping the eIF2α and AMP-PNP domains (161).

Figure 2.5 Inhibitory effect of Isohymenialdisine on GST-PKR-KD-mediated eIF2α phosphorylation and GST-PKR-KD autophosphorylation. See figure legend on the following page.

Figure 2.5 Inhibitory effect of Isohymenialdisine on GST-PKR-KD-mediated eIF2α phosphorylation and GST-PKR-KD autophosphorylation.

(A) Western blot probing the phosphorylation status of recombinant eIF2α following pre-incubation with GST-PKR-KD and the indicated small molecules. (B) Autoradiogram demonstrating the autophosphorylation activity of GST-PKR- KD in the presence of the indicated small molecules. Kinase reactions were performed as described in the Materials and Methods. (C) Location of the largest hot spot of PKR. Results are shown for the PKR structure bound to eIF2α•AMP- PNP (PDB code 2A19). Expanded view shows the position of the probe cluster site (cyan lines) relative to AMP-PNP (shown as yellow sticks). (D) Detailed top- down view of the three consensus probe clusters that define the major binding hot spot (CS1 cyan, CS4 blue and CS8 green). PKR side chains close to the hot spot are also shown (red – oxygen, blue – nitrogen). (E) Overlay of AMP-PNP (shown in stick representation, carbon atoms colored yellow) with the probe clusters. Hydrogen bonds are shown as black dashed lines. (F) Overlay of hymenialdisine (in stick representation, carbon and polar hydrogen atoms shown in white) with the probe clusters, showing the five hydrogen bonds formed. (G) Overlay of isohymenialdisine with the probe clusters.

We found that the main hot spot (i.e., the energetically most important region of the binding site, shown as cyan lines) is in a deep crevice within the

ATP site (Figure 2.5C). The bound AMP-PNP is superimposed (shown as yellow sticks) for reference and highlights that the main hot spot overlaps with the adenylate ester moiety (see expanded view on Figure 2.5C). A top-down view of the main hot spot revealed three probe cluster sites: (i) a large consensus site CS1

(cyan, 28 probe clusters), (ii) consensus site CS4 (blue, 7 probe clusters) and (iii) consensus site CS8 (green, 3 probe clusters) (Figure 2.5D). Superimposition of

AMP-PNP onto the probe cluster sites showed that the ligand overlaps with the central ring in the major hot spot CS1, but does not occupy two hot spot regions that expand from the ring (Figure 2.5E). The best docked pose of hymenialdisine fits better into these additional hot spots (Figure 2.5F), in spite of the fact that we mapped the PKR structure from its complex with AMP-PNP. It was shown previously that better fits into hot spots implies better ligand efficiency and higher affinity (162). In addition, hymenialdisine forms five hydrogen bonds with PKR.

Three hydrogen bonds are on the pyrroloazepine double ring, between the N1 atom of the pyrrole ring and the carbonyl oxygen of C369, between the O1 carbonyl oxygen of the azepine ring and the backbone NH group of C369, and between the N2 amide of the azepine ring and the carbonyl oxygen of E367

(Figure 2.5F). The other two hydrogen bonds are on the guanidine ring system, between the N5 amino group and the carbonyl oxygen of S418, and between the

O2 of the guanidine ring and the side chain of K296 (Figure 2.5F). We note that the same atoms of hymenialdisine form hydrogen bonds with CDK2, with the

53 only difference that O2 of the guanidine ring participates in a water-mediated interaction. In contrast, the original ligand in the X-ray structure, AMP-PNP, forms only two hydrogen bonds with PKR (Figure 2.5E). The bromine atom of hymenialdisine is predicted to interact with the amide CO group of the Q376 side chain (Figure 2.5F). Such halogen bonds are known to be favorable (163), which may explain why debromohymenialdisine is a weaker binder than hymenialdisine.

Isohymenialdisine also fits into the binding hot spot very well. Four of the hydrogen bonds are the same as for hymenialdisine (Figure 2.5G). Rather than directly interacting, the N5 amino group of the guanidine ring and the carbonyl oxygen of S418 appear to form a water-mediated hydrogen bond, which is not modeled in our calculations and hence not shown in Figure 2.5G. The bromine atom is moved into a hydrophobic environment and forms an interaction with the side chain of M366, which is expected to be more favorable than its interaction with the amide CO of Q376. This model data offers a potential mechanism by which isohymenialdisine and hymenialdisine inhibit PKR function.

CHAPTER 3: DISCUSSION

Growing evidence has implicated dysregulation of mRNA translation in a broad spectrum of pathological states ranging from neurological disorders, tumor progression and metastasis (3, 164). In an effort to investigate unexplored biochemical aspects of mRNA translation, relatively new small molecules as probes have been successfully employed in addition to classical genetics. Small molecules have unraveled novel molecular mechanisms of translation as well as signaling pathways that regulate it. Furthermore, several small molecule inhibitors have advanced to clinical trials with potential therapeutic benefits (165). Herein, we report on the identification of two small molecules, isohymenialdisine and hymenialdisine as inhibitors of PKR, a kinase targeting eIF2, identified during the course of functional compound screening.

We undertook a screening protocol based on in vitro translation of a bicistronic mRNA reporter that allows identification of compounds capable of modulating eukaryotic translation (150). We identified a stimulatory activity during bioassay-guided fractionation of a crude extract from the marine sponge

Stylissa massa. Hymenialdisine (1), debromohymenialdisine (2), and isohymenialdisine (3) were isolated (Figure 2.1A) and characterized in in vitro translation reactions using rabbit reticulocyte lysate (RRL) programmed with different types of bicistronic mRNA reporters. Whereas debromohymenialdisine did not show any noticeable effect on in vitro translation, both hymenialdisine and isohymenialdisine stimulated cap-dependent translation, and EMCV IRES- and

HCV IRES-mediated translation, with a more striking effect on IRES-mediated outputs (Figures 2.1C, D and 2.2B). In contrast, CrPV IRES, which requires neither eIFs nor initiator tRNA, appeared to be insensitive to hymenialdisine or isohymenialdisine, resulting in undisturbed translational outputs (Figure 2.1C).

Isohyenialdisine among the three compounds demonstrated a stimulatory effect on both cap-dependent and EMCV IRES-driven translation in another type of mammalian extracts prepared from mouse Krebs-2 cells (Figure 2.3A). None of the compounds at 50 μM had any effect on translation in wheat germ extracts, or on coupled transcription-translation in E.coli S30 extracts (Figure 2.3B, C).

In vitro translation extracts are powerful systems that have been useful in dissecting and characterizing various steps of protein synthesis and in generating proteins for functional studies. Mammalian extracts prepared from RRL, HeLa cells, or Krebs-2 cells have been extensively employed to delineate mRNA structural features required for efficient recruitment of 43S pre-initiation complexes to mRNA templates (151). One major mechanism that limits the translation rates of current mammalian extracts is phosphorylation of eIF2α (166).

Indeed, heme is typically supplemented to RRL in order to prevent heme- regulated inhibitor (HRI) from phosphorylating eIF2α and rapidly inactivating translation initiation (167). More recently, phosphorylation of eIF2α in cell-free systems has been characterized by Mikami et al. (168). They attributed eIF2α phosphorylation to PERK, PKR, and GCN2, and showed that the induced translation inhibition could be overcome by addition of GADD34 (which stimulated eIF2α dephosphorylation) or the vaccinia virus K3L eIF2α

56 pseudosubstrate (168). As well, extracts prepared from murine embryonic fibroblasts harboring an eIF2α homozygous knock-in mutation (Ser51 to Ala change that no longer is a substrate for the eIF2α kinases) have a translational efficiency that is 30-fold higher than extracts made from wild-type cells. These studies document that a current limitation of in vitro mammalian translation extracts occurs when eIF2α phosphorylated.

In vitro transcribed RNA, commonly used to program translation extracts are often contaminated with double stranded RNA products that can activate PKR

(157). In the experiments reported herein, we suspect that our in vitro transcribed mRNA was responsible for in vitro phosphorylation of eIF2α – an effect that could be blocked by a small molecule inhibitor of PKR (Figure 2.4D-F). Because of this, we were fortunate to detect a translation stimulatory activity in Stylissa massa extracts, and could attribute this to isohymenialdisine (and to a lower extent, hymenialdisine).

Hymenialdisine has been characterized as an inhibitor of cyclin dependent kinase 1 (CDK1; IC50 = 22 nM and CDK2; IC50 = 40 nM), glycogen synthase kinase 3β (IC50 = 10 nM), and casein kinase 1 (IC50 = 35 nM) and has been shown to function as a competitive inhibitor of ATP (169). Hymenialdisine is also a potent inhibitor of NF-κB in U937 cells (170) and of mitogen-activated protein kinase kinase-1 (MEK-1) (171). Previous studies aimed at identifying additional targets of hymenialdisine using affinity chromatography approaches (172) and computation approaches (173) identified a number of additional kinase targets, although not PKR. We find that hymenialdisine also inhibits eIF2α

57 phosphorylation (Figure 2.4G, compare lane 3 to 1 of p-eIF2α blot). Although little has been reported in the literature on isohymenialdisine, it appears to be more potent at stimulating translation (Figures 2.1, 2.2 and 2.4) and inhibiting eIF2α phosphorylation than hymenialdisine (Figure 2.4G, compare lane 10 to 8).

To further characterize isomenialdisine for its inhibitory effect on PKR-mediated eIF2α phosphorylation, we conducted in vitro assays with recombinant His6- eIF2α and GST-PKR-KD. Western blot analysis probing for eIF2α demonstrated that Isohymenialdisine is capable of inhibiting PKR-directed eIF2α phosphorylation (Figure 2.5A). SDS-PAGE analysis of 32P-phosphorylated PKR confirmed that isohymenialdisine directly acts on PKR by inhibiting its autophosphorylation activity (Figure 2.5B). Our modeling data is consistent with the ability of isohymenialdisine (Figure 2.5G) and hymenialdisine (Figure 2.5F) to bind to the ATP binding pocket of PKR.

Hymenialdisine has been reported to be cytotoxic against L5178y mouse lymphoma cells (174), block G2/M progression in 786-O renal cancer cells (<40%

G2/M arrest at 100 μM) (172), and inhibit LoVo (171) and CME T leukemia (175) cell proliferation. In these assays, debromohymenialdisine was reported to be as active as hymenialdisine (174). It is therefore unlikely that the anti-proliferative activity of debromohymenialdisine in these reports is due to PKR inhibition, since debromohymenialdisine does not inhibit PKR in vitro.

Augmented PKR activity has been implicated in a number of pathological situations including neurodegenerative disorders (71, 119-123), muscle atrophy

(124), obesity, diabetes (125), hematological toxicity (126) and cancer (127-131)

58 as described in Chapter 2. Surprisingly, there is little published literature on targeting PKR in drug discovery campaigns. A few inhibitors have been reported; high (mM) concentrations of 2-aminopurine (2-AP) (Figure 1.4) can block PKR activity (138), whereas Jammi et al. identified distinct classes of ATP inhibitors that interfere with PKR activity (140). These tool compounds have highlighted the value of targeting PKR in a number of settings. For example, PKR involvement in signal transduction was first suggested based on the observation that chemical inhibition of PKR by 2-AP blocked IFN gene expression and IFN-induced expression of various genes (176-178). Later, 2-AP was used in studies to identify

PKR-mediated pathways such as the signaling pathway involving NF-κB (179), a transcription factor that controls genes involved in various cellular processes including immunity, inflammation, apoptosis and stress response (33, 179).

Recently, a number of studies have utilized an imidazolo-oxindole compound C16 (Figure 1.4), which is one of the ATP-competitive inhibitors identified by that Jammi et al (140). They have investigated the pathological role of PKR in neurodegeneration that can be prevented by inhibiting PKR activity.

Compound C16 has been shown to prevent the death of human neuroblastoma cells (SH-SY5Y) induced by tunicamycin (an ER-stress inducer) (141). Similar neuroprotection was observed in cultured cerebellar granule neurons against apoptosis induced by potassium deprivation (180) and by amprolium, an agent which causes thiamine deficiency (181). Moreover, administration of C16 in chemically-induced mouse models of Huntington’s disease is capable of preventing neurodegeneartion and improve behavioral outcome (144). Another in

59 vivo study reported that C16 specifically inhibited the apoptotic PKR/eIF2α pathway in rat models displaying highly activated brain PKR, without stimulating the mTOR/p70S6K pathway, a consequence that could lead to proliferation and tumorigenesis (143). These results raise the interesting possibility that chemical inhibition of PKR may have therapeutic value in neurodegenerative disorders.

Compound C16 was also used to evaluate PKR as a potential target for attenuating muscle atrophy in cancer cachexia. Injection of C16 into mouse models of cachexia bearing MAC 16 tumors resulted in a reduction of phospho-

PKR in muscle and attenuated both muscle wasting and tumor growth (142). In a recent study, a potential role of PKR in the negative regulation of hematopoiesis as well as bone marrow (BM) failure has been recognized (80). In transgenic mice expressing human PKR in hematopoietic cells, C16 showed the ability to rescue reduced hematopoietic colony formation as a result of apoptosis-inducing stress.

This study suggested PKR inhibition as a potential therapeutic approach in BM failure or method to accelerate hematopoietic reconstitution in BM transplantation or therapy-related BM suppression.

As evidenced by a growing body of research above, chemical genetic approaches using PKR inhibitors as tool compounds have led to a better understanding of PKR-mediated pathways and the implicated pathologies, as well as raising the possibilities of novel therapeutic interventions. Future efforts on isohymenialdisine should determine if potency and selectivity against PKR can be improved upon structure-activity relationship (SAR) analysis and modeled structural information.

CHAPTER 4: MATERIALS & METHODS

4.1 Materials and General Methods

Restriction endonucleases, SP6 and T7 RNA polymerase were purchased from New England Biolabs, and T3 RNA polymerase was purchased from

Fermentas. [35S]methionine (>1000 Ci/mmol) and [γ-32P]ATP (6000 Ci/mmol) were purchased from PerkinElmer. Cycloheximide (CHX) (Bishop) was resuspended in H2O, whereas, homoharringtonine (HHT) (Sigma-Aldrich), four eIF2α kinase inhibitors, namely, C16 (PKRi) (Calbiochem) (140), SyK inhibitor

(GCN2i) (Calbiochem) (147), GSK PERKi (Toronto Research Chemicals Inc.)

(146, 182) and Quercetin (HRIi) (Sigma-Aldrich) (136, 137) were resuspended in

100% DMSO. Antibodies were as follows: anti-phospho-eIF2α (#3597), anti- eIF2α (#9722) and anti-eEF2 (#2332) were from Cell Signaling Technology.

Preparation of plasmid DNA, restriction digestion, agarose gel electrophoresis of

DNA, and bacterial transformations were carried out using standard methods

(183).

4.2 Isolation of Hymenialdisine and Its Derivatives from

Stylissa massa

The sponge sample was collected under contract with the Pfizer research laboratories (St. Louis). When Pfizer disbanded its natural products program, they generously donated their marine invertebrate collection to the Andersen

61 group at UBC. Specimens were collected by hand using SCUBA on a coral reef slope at 50 feet off Manado, Indonesia in November 1994 (N 01° 31’33”, E 124°

50’07”). A voucher sample has been deposited at the Netherlands Centre for

Biodiversity Naturalis in Leiden, The Netherlands.

Freshly collected sponge was frozen on site and transported frozen.

Lyophilized sponge (5 g) was cut into small pieces, immersed in and subsequently extracted repeatedly with MeOH (3 x 50 mL) at room temperature. The combined methanolic extracts were concentrated in vacuo, and the resultant extract was then partitioned between EtOAc (3 x 5 mL) and H2O (15 mL). The resulting aqueous extract was extracted with n-BuOH (3 x 5 mL). The combined n-BuOH extracts were evaporated to dryness, and the resulting oil was chromatographed on

Sephadex LH-20 with MeOH as eluent. The late eluting active fraction was purified via C18 reversed-phase HPLC using a CSC-Inertsil 150A/ODS2, 5 µm 25 x 0.94 cm column, with 9:1 (0.05% TFA/H2O)/MeCN as eluent to give pure samples of hymenialdisine (1) (6.6 mg), debromohymenialdisine (2) (2.4 mg) and isohymenialdisine (3) (3.8 mg). The structures of 1, 2 and 3 were established by standard 1 and 2D NMR spectroscopic and HRESIMS analyses.

4.3 In vitro Transcription and Translation Reactions.

In vitro transcriptions were preformed as previously described (150).

Plasmids pKS/FF/EMCV/Ren, pSP/(CAG)33/FF/HCV/Ren, pGL3/Ren/CrPV/FF and T3LucpA+ were linearized with BamHI and transcribed in vitro to generate

62 mRNA transcripts. In vitro translations using [35S]methionine were performed in rabbit reticulocyte lysate (RRL) as recommended by the manufacturer (Promega).

For luciferase assays, firefly and renilla lucifearase activities were measured on a

Berthold Lumat LB9507 luminometer. For gel analysis, translation products were separated on 10% SDS-polyacrylamide gels that were subsequently treated with

EN3HANCE, dried and exposed to Blue Ray film (VWR). In vitro translations in

Krebs extracts were performed as previously described (150). Translations in wheat germ extracts and Escherichia coli S30 extracts were performed as directed by the manufacturer (Promega).

4.4 In vitro Kinase Assays

Plasmids pGEX-5X-1-KD and pET-15b/eIF2α were transformed into

Escherichia coli BL21(DE3)/pLysS. GST-PKR-KD and His6-eIF2α were expressed and purified as previously described (140, 148). The eIF2α kinase assay and PKR autophosphorylation assay were performed as reported with slight modifications (140, 148). For eIF2α kinase assays, 0.5 μg of GST-PKR-KD and

His6-eIF2α were incubated with 50 μM ATP in kinase buffer (20 mM Tris-HCl

[pH 7.6], 10 mM KCl, 10 mM MgCl2 and 10% glycerol) at 30°C for 30 min.

Phosphorylation of eIF2α was assessed by Western blotting. For PKR autophosphorylation assays, 0.5 μg of GST-PKR-KD was incubated with 50 μM

[γ-32P]ATP in kinase buffer at 30°C for 30 min. The reaction was quenched with

3x SDS-PAGE loading buffer, boiled and resolved in 10% SDS-polyacrylamide gels. The gel was dried and exposed to Blue Ray film (VWR).

4.5 Computational Mapping of the PKR Structure

Mapping was performed on the PKR kinase domain-eIF2α-AMP-PNP complex (PDB ID 2A19). Chain A representing eIF2α, the ligand AMP-PNP and all crystallographic water molecules were removed prior to calculations. One

PKR monomer (Chain B) was globally mapped using the FTMap server

(http://ftmap.bu.edu/). FTMap places molecular probes, small organic molecules containing various functional groups, on a dense grid defined around the entire protein without any restriction, finds favorable positions using empirical energy functions, further refines the selected poses by energy minimization, clusters the low energy conformations, calculates the partition function for each cluster, and ranks the clusters on the basis of their probabilities (159). The consensus sites are then defined as those where clusters of different probes overlap; such consensus sites identify the binding hot spots, i.e., regions of interactions that substantially contribute to the binding free energy and hence are of prime importance to ligand binding and to drug design .

4.6 Docking of Hymenialdisine and Isohymenialdisine

Based on the mapping results, a box with 4Å padding was created around the putative binding site. The docking was carried out using the standard settings of AutoDock Vina 1.1.2 (184), and a number of lowest energy binding modes were retained for each ligand. The selection of the most likely pose was based on the atom densities calculated from the mapping results (160). We considered each retained pose separately, for each atom summed the atomic densities on the grid points within a 1Å radius, and then added these values for all atoms. The poses were ranked on the basis of this overlap measure, and the pose with the best overlap was selected (160). The selected complexes were minimized using the

Charmm potential allowing for complete flexibility (185).

CHAPTER 5: CONCLUSION

Although the molecular mechanisms of protein synthesis have been elucidated in considerable detail, many intriguing questions, especially concerning signaling pathways that regulate protein synthesis or mechanisms underlying pathological consequences, remain unanswered. Chemical genetic approaches of using small molecule perturbations to dissect cellular pathways have provided valuable insights regarding these issues. In this work, we have identified two small molecules, Isohymenialdisine and Hymenialdisine as ATP- competitive inhibitors of PKR. Our encouraging results warrant future work on isohymenialdisine which will need to be improved, in terms of its potency and target selectivity, using structure-activity relationship (SAR) analysis and computer-aided molecular modeling.

ACKNOWLEDGEMENTS

We thank Dr. Peter Beal (UC Davis) for the kind gift of GST-PKR-KD expression vector and Dr. Nicole J. de Voogd (Netherlands Centre for

Biodiversity Naturalis, Leiden) for identification of the sponge Stylissa massa.

This work was supported by a grant from the Canadian Institutes of Health

Research (CIHR) (MOP-106530) to J.P. and a grant from the National Institute of

General Medical Sciences (GM064700) to S.V.

REFERENCES

1. Hands SL, Proud CG, and Wyttenbach A. mTOR's role in ageing: protein synthesis or autophagy? Aging. 2009;1(7):586-97. 2. Rosenwald IB. Deregulation of protein synthesis as a mechanism of neoplastic transformation. BioEssays : news and reviews in molecular, cellular and developmental biology. 1996;18(3):243-50. 3. Le Quesne JP, Spriggs KA, Bushell M, and Willis AE. Dysregulation of protein synthesis and disease. The Journal of pathology. 2010;220(2):140- 51. 4. Jackson RJ, Hellen CU, and Pestova TV. The mechanism of eukaryotic translation initiation and principles of its regulation. Nature reviews Molecular cell biology. 2010;11(2):113-27. 5. Hershey JW, Sonenberg N, and Mathews MB. Principles of translational control: an overview. Cold Spring Harbor perspectives in biology. 2012;4(12). 6. Frederickson RM, and Sonenberg N. Signal transduction and regulation of translation initiation. Seminars in cell biology. 1992;3(2):107-15. 7. Preiss T, and Hentze MW. From factors to mechanisms: translation and translational control in eukaryotes. Current opinion in genetics & development. 1999;9(5):515-21. 8. Londei P. eLS. John Wiley & Sons, Ltd; 2001. 9. Sonenberg N, and Hinnebusch AG. Regulation of translation initiation in eukaryotes: mechanisms and biological targets. Cell. 2009;136(4):731-45. 10. Fraser CS, and Doudna JA. Structural and mechanistic insights into hepatitis C viral translation initiation. Nature reviews Microbiology. 2007;5(1):29-38. 11. Gebauer F, and Hentze MW. Molecular mechanisms of translational control. Nature reviews Molecular cell biology. 2004;5(10):827-35. 12. Derry MC, Yanagiya A, Martineau Y, and Sonenberg N. Regulation of poly(A)-binding protein through PABP-interacting proteins. Cold Spring Harbor symposia on quantitative biology. 2006;71(537-43. 13. Aitken CE, and Lorsch JR. A mechanistic overview of translation initiation in eukaryotes. Nature structural & molecular biology. 2012;19(6):568-76. 14. Hinnebusch AG, and Lorsch JR. The mechanism of eukaryotic translation initiation: new insights and challenges. Cold Spring Harbor perspectives in biology. 2012;4(10). 15. Schuler M, Connell SR, Lescoute A, Giesebrecht J, Dabrowski M, Schroeer B, Mielke T, Penczek PA, Westhof E, and Spahn CM. Structure of the ribosome-bound cricket paralysis virus IRES RNA. Nature structural & molecular biology. 2006;13(12):1092-6. 16. Van Der Kelen K, Beyaert R, Inze D, and De Veylder L. Translational control of eukaryotic gene expression. Critical reviews in biochemistry and molecular biology. 2009;44(4):143-68.

17. Pelletier J, and Sonenberg N. Internal initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA. Nature. 1988;334(6180):320-5. 18. Jang SK, Krausslich HG, Nicklin MJ, Duke GM, Palmenberg AC, and Wimmer E. A segment of the 5' nontranslated region of encephalomyocarditis virus RNA directs internal entry of ribosomes during in vitro translation. Journal of virology. 1988;62(8):2636-43. 19. Kieft JS. Viral IRES RNA structures and ribosome interactions. Trends in biochemical sciences. 2008;33(6):274-83. 20. Thompson SR. Tricks an IRES uses to enslave ribosomes. Trends in microbiology. 2012;20(11):558-66. 21. Jackson RJ. The current status of vertebrate cellular mRNA IRESs. Cold Spring Harbor perspectives in biology. 2013;5(2). 22. Blagden SP, and Willis AE. The biological and therapeutic relevance of mRNA translation in cancer. Nature reviews Clinical oncology. 2011;8(5):280-91. 23. Hiremath LS, Webb NR, and Rhoads RE. Immunological detection of the messenger RNA cap-binding protein. The Journal of biological chemistry. 1985;260(13):7843-9. 24. Duncan R, Milburn SC, and Hershey JW. Regulated phosphorylation and low abundance of HeLa cell initiation factor eIF-4F suggest a role in translational control. Heat shock effects on eIF-4F. The Journal of biological chemistry. 1987;262(1):380-8. 25. Gosselin P, Oulhen N, Jam M, Ronzca J, Cormier P, Czjzek M, and Cosson B. The translational repressor 4E-BP called to order by eIF4E: new structural insights by SAXS. Nucleic acids research. 2011;39(8):3496-503. 26. Gingras AC, Gygi SP, Raught B, Polakiewicz RD, Abraham RT, Hoekstra MF, Aebersold R, and Sonenberg N. Regulation of 4E-BP1 phosphorylation: a novel two-step mechanism. Genes & development. 1999;13(11):1422-37. 27. Raught B, and Gingras AC. eIF4E activity is regulated at multiple levels. The international journal of biochemistry & cell biology. 1999;31(1):43- 57. 28. Gingras AC, Raught B, Gygi SP, Niedzwiecka A, Miron M, Burley SK, Polakiewicz RD, Wyslouch-Cieszynska A, Aebersold R, and Sonenberg N. Hierarchical phosphorylation of the translation inhibitor 4E-BP1. Genes & development. 2001;15(21):2852-64. 29. Raught B, Gingras AC, and Sonenberg N. The target of rapamycin (TOR) proteins. Proceedings of the National Academy of Sciences of the United States of America. 2001;98(13):7037-44. 30. Fox CJ, Hammerman PS, Cinalli RM, Master SR, Chodosh LA, and Thompson CB. The serine/threonine kinase Pim-2 is a transcriptionally regulated apoptotic inhibitor. Genes & development. 2003;17(15):1841-54. 31. Clemens MJ. Targets and mechanisms for the regulation of translation in malignant transformation. Oncogene. 2004;23(18):3180-8.

32. Sonenberg N, and Dever TE. Eukaryotic translation initiation factors and regulators. Current opinion in structural biology. 2003;13(1):56-63. 33. Garcia MA, Gil J, Ventoso I, Guerra S, Domingo E, Rivas C, and Esteban M. Impact of protein kinase PKR in cell biology: from antiviral to antiproliferative action. Microbiology and molecular biology reviews : MMBR. 2006;70(4):1032-60. 34. Krishnamoorthy T, Pavitt GD, Zhang F, Dever TE, and Hinnebusch AG. Tight binding of the phosphorylated alpha subunit of initiation factor 2 (eIF2alpha) to the regulatory subunits of guanine nucleotide exchange factor eIF2B is required for inhibition of translation initiation. Molecular and cellular biology. 2001;21(15):5018-30. 35. Boyce M, Bryant KF, Jousse C, Long K, Harding HP, Scheuner D, Kaufman RJ, Ma D, Coen DM, Ron D, et al. A selective inhibitor of eIF2alpha dephosphorylation protects cells from ER stress. Science (New York, NY). 2005;307(5711):935-9. 36. Dalton LE, Healey E, Irving J, and Marciniak SJ. Phosphoproteins in stress-induced disease. Progress in molecular biology and translational science. 2012;106(189-221. 37. Baird TD, and Wek RC. Eukaryotic initiation factor 2 phosphorylation and translational control in metabolism. Advances in nutrition (Bethesda, Md). 2012;3(3):307-21. 38. Chen JJ. Regulation of protein synthesis by the heme-regulated eIF2alpha kinase: relevance to anemias. Blood. 2007;109(7):2693-9. 39. Han AP, Yu C, Lu L, Fujiwara Y, Browne C, Chin G, Fleming M, Leboulch P, Orkin SH, and Chen JJ. Heme-regulated eIF2alpha kinase (HRI) is required for translational regulation and survival of erythroid precursors in iron deficiency. The EMBO journal. 2001;20(23):6909-18. 40. Suragani RN, Zachariah RS, Velazquez JG, Liu S, Sun CW, Townes TM, and Chen JJ. Heme-regulated eIF2alpha kinase activated Atf4 signaling pathway in oxidative stress and erythropoiesis. Blood. 2012;119(22):5276- 84. 41. Ron D. Translational control in the endoplasmic reticulum stress response. The Journal of clinical investigation. 2002;110(10):1383-8. 42. Dever TE, Dar AC, and Sicheri F. 12 The eIF2α Kinases. 2007. 43. Yang JM, London IM, and Chen JJ. Effects of hemin and porphyrin compounds on intersubunit disulfide formation of heme-regulated eIF-2 alpha kinase and the regulation of protein synthesis in reticulocyte lysates. The Journal of biological chemistry. 1992;267(28):20519-24. 44. Yun BG, Matts JA, and Matts RL. Interdomain interactions regulate the activation of the heme-regulated eIF 2 alpha kinase. Biochimica et biophysica acta. 2005;1725(2):174-81. 45. Miksanova M, Igarashi J, Minami M, Sagami I, Yamauchi S, Kurokawa H, and Shimizu T. Characterization of heme-regulated eIF2alpha kinase: roles of the N-terminal domain in the oligomeric state, heme binding, catalysis, and inhibition. Biochemistry. 2006;45(32):9894-905.

46. Lu L, Han AP, and Chen JJ. Translation initiation control by heme- regulated eukaryotic initiation factor 2alpha kinase in erythroid cells under cytoplasmic stresses. Molecular and cellular biology. 2001;21(23):7971- 80. 47. Metz DH, and Esteban M. Interferon inhibits viral protein synthesis in L cells infected with vaccinia virus. Nature. 1972;238(5364):385-8. 48. Friedman RM, Metz DH, Esteban RM, Tovell DR, Ball LA, and Kerr IM. Mechanism of interferon action: inhibition of viral messenger ribonucleic acid translation in L-cell extracts. Journal of virology. 1972;10(6):1184-98. 49. Roberts WK, Hovanessian A, Brown RE, Clemens MJ, and Kerr IM. Interferon-mediated protein kinase and low-molecular-weight inhibitor of protein synthesis. Nature. 1976;264(5585):477-80. 50. Clemens MJ, Hershey JW, Hovanessian AC, Jacobs BC, Katze MG, Kaufman RJ, Lengyel P, Samuel CE, Sen GC, and Williams BR. PKR: proposed nomenclature for the RNA-dependent protein kinase induced by interferon. Journal of interferon research. 1993;13(3):241. 51. Manche L, Green SR, Schmedt C, and Mathews MB. Interactions between double-stranded RNA regulators and the protein kinase DAI. Molecular and cellular biology. 1992;12(11):5238-48. 52. Petryshyn R, Chen JJ, and London IM. Detection of activated double- stranded RNA-dependent protein kinase in 3T3-F442A cells. Proceedings of the National Academy of Sciences of the United States of America. 1988;85(5):1427-31. 53. Hovanessian AG. The double stranded RNA-activated protein kinase induced by interferon: dsRNA-PK. Journal of interferon research. 1989;9(6):641-7. 54. Thomis DC, and Samuel CE. Mechanism of interferon action: autoregulation of RNA-dependent P1/eIF-2 alpha protein kinase (PKR) expression in transfected mammalian cells. Proceedings of the National Academy of Sciences of the United States of America. 1992;89(22):10837- 41. 55. Meurs EF, Watanabe Y, Kadereit S, Barber GN, Katze MG, Chong K, Williams BR, and Hovanessian AG. Constitutive expression of human double-stranded RNA-activated p68 kinase in murine cells mediates phosphorylation of eukaryotic initiation factor 2 and partial resistance to encephalomyocarditis virus growth. Journal of virology. 1992;66(10):5805-14. 56. Sadler AJ, and Williams BR. Structure and function of the protein kinase R. Current topics in microbiology and immunology. 2007;316(253-92. 57. Clemens MJ. PKR--a protein kinase regulated by double-stranded RNA. The international journal of biochemistry & cell biology. 1997;29(7):945- 9. 58. Zamanian-Daryoush M, Mogensen TH, DiDonato JA, and Williams BR. NF-kappaB activation by double-stranded-RNA-activated protein kinase (PKR) is mediated through NF-kappaB-inducing kinase and IkappaB kinase. Molecular and cellular biology. 2000;20(4):1278-90.

59. Kirchhoff S, Koromilas AE, Schaper F, Grashoff M, Sonenberg N, and Hauser H. IRF-1 induced cell growth inhibition and interferon induction requires the activity of the protein kinase PKR. Oncogene. 1995;11(3):439-45. 60. Wong AH, Tam NW, Yang YL, Cuddihy AR, Li S, Kirchhoff S, Hauser H, Decker T, and Koromilas AE. Physical association between STAT1 and the interferon-inducible protein kinase PKR and implications for interferon and double-stranded RNA signaling pathways. The EMBO journal. 1997;16(6):1291-304. 61. Deb A, Zamanian-Daryoush M, Xu Z, Kadereit S, and Williams BR. Protein kinase PKR is required for platelet-derived growth factor signaling of c-fos gene expression via Erks and Stat3. The EMBO journal. 2001;20(10):2487-96. 62. Cuddihy AR, Wong AH, Tam NW, Li S, and Koromilas AE. The double- stranded RNA activated protein kinase PKR physically associates with the tumor suppressor p53 protein and phosphorylates human p53 on serine 392 in vitro. Oncogene. 1999;18(17):2690-702. 63. Guerra S, Lopez-Fernandez LA, Garcia MA, Zaballos A, and Esteban M. Human gene profiling in response to the active protein kinase, interferon- induced serine/threonine protein kinase (PKR), in infected cells. Involvement of the transcription factor ATF-3 IN PKR-induced apoptosis. The Journal of biological chemistry. 2006;281(27):18734-45. 64. Viatour P, Merville MP, Bours V, and Chariot A. Phosphorylation of NF- kappaB and IkappaB proteins: implications in cancer and inflammation. Trends in biochemical sciences. 2005;30(1):43-52. 65. Bonnet MC, Daurat C, Ottone C, and Meurs EF. The N-terminus of PKR is responsible for the activation of the NF-kappaB signaling pathway by interacting with the IKK complex. Cellular signalling. 2006;18(11):1865- 75. 66. Gil J, and Esteban M. Induction of apoptosis by the dsRNA-dependent protein kinase (PKR): mechanism of action. Apoptosis : an international journal on programmed cell death. 2000;5(2):107-14. 67. Garcia MA, Meurs EF, and Esteban M. The dsRNA protein kinase PKR: virus and cell control. Biochimie. 2007;89(6-7):799-811. 68. Chang RC, Suen KC, Ma CH, Elyaman W, Ng HK, and Hugon J. Involvement of double-stranded RNA-dependent protein kinase and phosphorylation of eukaryotic initiation factor-2alpha in neuronal degeneration. Journal of neurochemistry. 2002;83(5):1215-25. 69. Suen KC, Yu MS, So KF, Chang RC, and Hugon J. Upstream signaling pathways leading to the activation of double-stranded RNA-dependent serine/threonine protein kinase in beta-amyloid peptide neurotoxicity. The Journal of biological chemistry. 2003;278(50):49819-27. 70. Onuki R, Bando Y, Suyama E, Katayama T, Kawasaki H, Baba T, Tohyama M, and Taira K. An RNA-dependent protein kinase is involved in tunicamycin-induced apoptosis and Alzheimer's disease. The EMBO journal. 2004;23(4):959-68.

71. Bando Y, Onuki R, Katayama T, Manabe T, Kudo T, Taira K, and Tohyama M. Double-strand RNA dependent protein kinase (PKR) is involved in the extrastriatal degeneration in Parkinson's disease and Huntington's disease. Neurochemistry international. 2005;46(1):11-8. 72. Morel M, Couturier J, Lafay-Chebassier C, Paccalin M, and Page G. PKR, the double stranded RNA-dependent protein kinase as a critical target in Alzheimer's disease. Journal of cellular and molecular medicine. 2009;13(8a):1476-88. 73. Chong KL, Feng L, Schappert K, Meurs E, Donahue TF, Friesen JD, Hovanessian AG, and Williams BR. Human p68 kinase exhibits growth suppression in yeast and homology to the translational regulator GCN2. The EMBO journal. 1992;11(4):1553-62. 74. Barber GN, Tomita J, Garfinkel MS, Meurs E, Hovanessian A, and Katze MG. Detection of protein kinase homologues and viral RNA-binding domains utilizing polyclonal antiserum prepared against a baculovirus- expressed ds RNA-activated 68,000-Da protein kinase. Virology. 1992;191(2):670-9. 75. Koromilas AE, Roy S, Barber GN, Katze MG, and Sonenberg N. Malignant transformation by a mutant of the IFN-inducible dsRNA- dependent protein kinase. Science (New York, NY). 1992;257(5077):1685- 9. 76. Meurs EF, Galabru J, Barber GN, Katze MG, and Hovanessian AG. Tumor suppressor function of the interferon-induced double-stranded RNA-activated protein kinase. Proceedings of the National Academy of Sciences of the United States of America. 1993;90(1):232-6. 77. Kuyama M, Nakanishi G, Arata J, Iwatsuki K, and Fujimoto W. Expression of double-stranded RNA-activated protein kinase in keratinocytes and keratinocytic neoplasia. The Journal of dermatology. 2003;30(8):579-89. 78. Yoshida K, Okamura H, Amorim BR, Hinode D, Yoshida H, and Haneji T. PKR-mediated degradation of STAT1 regulates osteoblast differentiation. Experimental cell research. 2009;315(12):2105-14. 79. Teramachi J, Morimoto H, Baba R, Doi Y, Hirashima K, and Haneji T. Double stranded RNA-dependent protein kinase is involved in osteoclast differentiation of RAW264.7 cells in vitro. Experimental cell research. 2010;316(19):3254-62. 80. Liu X, Bennett RL, Cheng X, Byrne M, Reinhard MK, and May WS, Jr. PKR regulates proliferation, differentiation, and survival of murine hematopoietic stem/progenitor cells. Blood. 2013;121(17):3364-74. 81. Kronfeld-Kinar Y, Vilchik S, Hyman T, Leibkowicz F, and Salzberg S. Involvement of PKR in the regulation of myogenesis. Cell growth & differentiation : the molecular biology journal of the American Association for Cancer Research. 1999;10(3):201-12. 82. Anderson BR, Muramatsu H, Nallagatla SR, Bevilacqua PC, Sansing LH, Weissman D, and Kariko K. Incorporation of pseudouridine into mRNA

enhances translation by diminishing PKR activation. Nucleic acids research. 2010;38(17):5884-92. 83. Cole JL. Activation of PKR: an open and shut case? Trends in biochemical sciences. 2007;32(2):57-62. 84. Robertson HD, and Mathews MB. The regulation of the protein kinase PKR by RNA. Biochimie. 1996;78(11-12):909-14. 85. Patel CV, Handy I, Goldsmith T, and Patel RC. PACT, a stress-modulated cellular activator of interferon-induced double-stranded RNA-activated protein kinase, PKR. The Journal of biological chemistry. 2000;275(48):37993-8. 86. Anderson E, Pierre-Louis WS, Wong CJ, Lary JW, and Cole JL. Heparin activates PKR by inducing dimerization. Journal of molecular biology. 2011;413(5):973-84. 87. Liu CY, and Kaufman RJ. The unfolded protein response. Journal of cell science. 2003;116(Pt 10):1861-2. 88. Parmar VM, and Schroder M. Sensing endoplasmic reticulum stress. Advances in experimental medicine and biology. 2012;738(153-68. 89. Hetz C. The unfolded protein response: controlling cell fate decisions under ER stress and beyond. Nature reviews Molecular cell biology. 2012;13(2):89-102. 90. Walter P, and Ron D. The unfolded protein response: from stress pathway to homeostatic regulation. Science (New York, NY). 2011;334(6059):1081- 6. 91. Kraskiewicz H, and FitzGerald U. InterfERing with endoplasmic reticulum stress. Trends in pharmacological sciences. 2012;33(2):53-63. 92. Szegezdi E, Logue SE, Gorman AM, and Samali A. Mediators of endoplasmic reticulum stress-induced apoptosis. EMBO reports. 2006;7(9):880-5. 93. Brush MH, Weiser DC, and Shenolikar S. Growth arrest and DNA damage-inducible protein GADD34 targets protein phosphatase 1 alpha to the endoplasmic reticulum and promotes dephosphorylation of the alpha subunit of eukaryotic translation initiation factor 2. Molecular and cellular biology. 2003;23(4):1292-303. 94. Chakrabarti A, Chen AW, and Varner JD. A review of the mammalian unfolded protein response. Biotechnology and bioengineering. 2011;108(12):2777-93. 95. Bertolotti A, Zhang Y, Hendershot LM, Harding HP, and Ron D. Dynamic interaction of BiP and ER stress transducers in the unfolded-protein response. Nature cell biology. 2000;2(6):326-32. 96. Okamura K, Kimata Y, Higashio H, Tsuru A, and Kohno K. Dissociation of Kar2p/BiP from an ER sensory molecule, Ire1p, triggers the unfolded protein response in yeast. Biochemical and biophysical research communications. 2000;279(2):445-50. 97. Credle JJ, Finer-Moore JS, Papa FR, Stroud RM, and Walter P. On the mechanism of sensing unfolded protein in the endoplasmic reticulum.

Proceedings of the National Academy of Sciences of the United States of America. 2005;102(52):18773-84. 98. Pincus D, Chevalier MW, Aragon T, van Anken E, Vidal SE, El-Samad H, and Walter P. BiP binding to the ER-stress sensor Ire1 tunes the homeostatic behavior of the unfolded protein response. PLoS biology. 2010;8(7):e1000415. 99. Grallert B, and Boye E. The Gcn2 kinase as a cell cycle regulator. Cell cycle (Georgetown, Tex). 2007;6(22):2768-72. 100. Balasubramanian MN, Butterworth EA, and Kilberg MS. Asparagine synthetase: regulation by cell stress and involvement in tumor biology. American journal of physiology Endocrinology and metabolism. 2013;304(8):E789-99. 101. Gallinetti J, Harputlugil E, and Mitchell JR. Amino acid sensing in dietary-restriction-mediated longevity: roles of signal-transducing kinases GCN2 and TOR. The Biochemical journal. 2013;449(1):1-10. 102. Zhang P, McGrath BC, Reinert J, Olsen DS, Lei L, Gill S, Wek SA, Vattem KM, Wek RC, Kimball SR, et al. The GCN2 eIF2alpha kinase is required for adaptation to amino acid deprivation in mice. Molecular and cellular biology. 2002;22(19):6681-8. 103. Vattem KM, and Wek RC. Reinitiation involving upstream ORFs regulates ATF4 mRNA translation in mammalian cells. Proceedings of the National Academy of Sciences of the United States of America. 2004;101(31):11269-74. 104. Ye J, Kumanova M, Hart LS, Sloane K, Zhang H, De Panis DN, Bobrovnikova-Marjon E, Diehl JA, Ron D, and Koumenis C. The GCN2- ATF4 pathway is critical for tumour cell survival and proliferation in response to nutrient deprivation. The EMBO journal. 2010;29(12):2082-96. 105. Jiang HY, and Wek RC. GCN2 phosphorylation of eIF2alpha activates NF-kappaB in response to UV irradiation. The Biochemical journal. 2005;385(Pt 2):371-80. 106. Yang R, Wek SA, and Wek RC. Glucose limitation induces GCN4 translation by activation of Gcn2 protein kinase. Molecular and cellular biology. 2000;20(8):2706-17. 107. Rolfes RJ, and Hinnebusch AG. Translation of the yeast transcriptional activator GCN4 is stimulated by purine limitation: implications for activation of the protein kinase GCN2. Molecular and cellular biology. 1993;13(8):5099-111. 108. Cai Q, and Brooks HL. Phosphorylation of eIF2alpha via the general control kinase, GCN2, modulates the ability of renal medullary cells to survive high urea stress. American journal of physiology Renal physiology. 2011;301(6):F1202-7. 109. Chaveroux C, Lambert-Langlais S, Parry L, Carraro V, Jousse C, Maurin AC, Bruhat A, Marceau G, Sapin V, Averous J, et al. Identification of GCN2 as new redox regulator for oxidative stress prevention in vivo. Biochemical and biophysical research communications. 2011;415(1):120- 4.

110. Guo F, and Cavener DR. The GCN2 eIF2alpha kinase regulates fatty-acid homeostasis in the liver during deprivation of an essential amino acid. Cell metabolism. 2007;5(2):103-14. 111. Maurin AC, Jousse C, Averous J, Parry L, Bruhat A, Cherasse Y, Zeng H, Zhang Y, Harding HP, Ron D, et al. The GCN2 kinase biases feeding behavior to maintain amino acid homeostasis in omnivores. Cell metabolism. 2005;1(4):273-7. 112. Costa-Mattioli M, Gobert D, Harding H, Herdy B, Azzi M, Bruno M, Bidinosti M, Ben Mamou C, Marcinkiewicz E, Yoshida M, et al. Translational control of hippocampal synaptic plasticity and memory by the eIF2alpha kinase GCN2. Nature. 2005;436(7054):1166-73. 113. Thomas ED, Arvin CD, and Frank S. 12 The eIF2α Kinases. Cold Spring Harbor Monograph Archive; Volume 48 (2007): Translational Control in Biology and Medicine. 2007. 114. Donnelly N, Gorman AM, Gupta S, and Samali A. The eIF2alpha kinases: their structures and functions. Cellular and molecular life sciences : CMLS. 2013. 115. Qiu H, Dong J, Hu C, Francklyn CS, and Hinnebusch AG. The tRNA- binding moiety in GCN2 contains a dimerization domain that interacts with the kinase domain and is required for tRNA binding and kinase activation. The EMBO journal. 2001;20(6):1425-38. 116. Calkhoven CF, Muller C, and Leutz A. Translational control of gene expression and disease. Trends in molecular medicine. 2002;8(12):577-83. 117. Wek RC, Jiang HY, and Anthony TG. Coping with stress: eIF2 kinases and translational control. Biochemical Society transactions. 2006;34(Pt 1):7-11. 118. Han AP, Fleming MD, and Chen JJ. Heme-regulated eIF2alpha kinase modifies the phenotypic severity of murine models of erythropoietic protoporphyria and beta-thalassemia. The Journal of clinical investigation. 2005;115(6):1562-70. 119. Chang RC, Wong AK, Ng HK, and Hugon J. Phosphorylation of eukaryotic initiation factor-2alpha (eIF2alpha) is associated with neuronal degeneration in Alzheimer's disease. Neuroreport. 2002;13(18):2429-32. 120. Peel AL, Rao RV, Cottrell BA, Hayden MR, Ellerby LM, and Bredesen DE. Double-stranded RNA-dependent protein kinase, PKR, binds preferentially to Huntington's disease (HD) transcripts and is activated in HD tissue. Human molecular genetics. 2001;10(15):1531-8. 121. Paquet C, Bose A, Polivka M, Peoc'h K, Brouland JP, Keohane C, Hugon J, and Gray F. Neuronal phosphorylated RNA-dependent protein kinase in Creutzfeldt-Jakob disease. Journal of neuropathology and experimental neurology. 2009;68(2):190-8. 122. Hu JH, Zhang H, Wagey R, Krieger C, and Pelech SL. Protein kinase and protein phosphatase expression in amyotrophic lateral sclerosis spinal cord. Journal of neurochemistry. 2003;85(2):432-42.

123. Peel AL, and Bredesen DE. Activation of the cell stress kinase PKR in Alzheimer's disease and human amyloid precursor protein transgenic mice. Neurobiology of disease. 2003;14(1):52-62. 124. Eley HL, and Tisdale MJ. Skeletal muscle atrophy, a link between depression of protein synthesis and increase in degradation. The Journal of biological chemistry. 2007;282(10):7087-97. 125. Nakamura T, Furuhashi M, Li P, Cao H, Tuncman G, Sonenberg N, Gorgun CZ, and Hotamisligil GS. Double-stranded RNA-dependent protein kinase links pathogen sensing with stress and metabolic homeostasis. Cell. 2010;140(3):338-48. 126. Follo MY, Finelli C, Mongiorgi S, Clissa C, Bosi C, Martinelli G, Blalock WL, Cocco L, and Martelli AM. PKR is activated in MDS patients and its subcellular localization depends on disease severity. Leukemia. 2008;22(12):2267-9. 127. Kim SH, Gunnery S, Choe JK, and Mathews MB. Neoplastic progression in melanoma and colon cancer is associated with increased expression and activity of the interferon-inducible protein kinase, PKR. Oncogene. 2002;21(57):8741-8. 128. Kim SH, Forman AP, Mathews MB, and Gunnery S. Human breast cancer cells contain elevated levels and activity of the protein kinase, PKR. Oncogene. 2000;19(27):3086-94. 129. Nussbaum JM, Major M, and Gunnery S. Transcriptional upregulation of interferon-induced protein kinase, PKR, in breast cancer. Cancer letters. 2003;196(2):207-16. 130. Delhem N, Sabile A, Gajardo R, Podevin P, Abadie A, Blaton MA, Kremsdorf D, Beretta L, and Brechot C. Activation of the interferon- inducible protein kinase PKR by hepatocellular carcinoma derived- hepatitis C virus core protein. Oncogene. 2001;20(41):5836-45. 131. Pataer A, Swisher SG, Roth JA, Logothetis CJ, and Corn PG. Inhibition of RNA-dependent protein kinase (PKR) leads to cancer cell death and increases chemosensitivity. Cancer biology & therapy. 2009;8(3):245-52. 132. Delepine M, Nicolino M, Barrett T, Golamaully M, Lathrop GM, and Julier C. EIF2AK3, encoding translation initiation factor 2-alpha kinase 3, is mutated in patients with Wolcott-Rallison syndrome. Nature genetics. 2000;25(4):406-9. 133. Harding HP, Zeng H, Zhang Y, Jungries R, Chung P, Plesken H, Sabatini DD, and Ron D. Diabetes mellitus and exocrine pancreatic dysfunction in perk-/- mice reveals a role for translational control in secretory cell survival. Molecular cell. 2001;7(6):1153-63. 134. Zhang P, McGrath B, Li S, Frank A, Zambito F, Reinert J, Gannon M, Ma K, McNaughton K, and Cavener DR. The PERK eukaryotic initiation factor 2 alpha kinase is required for the development of the skeletal system, postnatal growth, and the function and viability of the pancreas. Molecular and cellular biology. 2002;22(11):3864-74.

135. Dudek SM, and Semenkovich CF. Essential amino acids regulate fatty acid synthase expression through an uncharged transfer RNA-dependent mechanism. The Journal of biological chemistry. 1995;270(49):29323-9. 136. Srivastava AK, and Chiasson JL. Effect of quercetin on serine/threonine and tyrosine protein kinases. Progress in clinical and biological research. 1986;213(315-8. 137. Kanelakis KC, Pyati J, Wagaman PC, Chuang JC, Yang Y, and Shankley NP. Functional Characterization of the Canine Heme-Regulated eIF2alpha Kinase: Regulation of Protein Synthesis. Advances in hematology. 2009;2009(251915. 138. Hu Y, and Conway TW. 2-Aminopurine inhibits the double-stranded RNA-dependent protein kinase both in vitro and in vivo. Journal of interferon research. 1993;13(5):323-8. 139. Posti K, Leinonen S, Tetri S, Kottari S, Viitala P, Pelkonen O, and Raunio H. Modulation of murine phenobarbital-inducible CYP2A5, CYP2B10 and CYP1A enzymes by inhibitors of protein kinases and phosphatases. European journal of biochemistry / FEBS. 1999;264(1):19-26. 140. Jammi NV, Whitby LR, and Beal PA. Small molecule inhibitors of the RNA-dependent protein kinase. Biochemical and biophysical research communications. 2003;308(1):50-7. 141. Shimazawa M, and Hara H. Inhibitor of double stranded RNA-dependent protein kinase protects against cell damage induced by ER stress. Neuroscience letters. 2006;409(3):192-5. 142. Eley HL, Russell ST, and Tisdale MJ. Attenuation of muscle atrophy in a murine model of cachexia by inhibition of the dsRNA-dependent protein kinase. British journal of cancer. 2007;96(8):1216-22. 143. Ingrand S, Barrier L, Lafay-Chebassier C, Fauconneau B, Page G, and Hugon J. The oxindole/imidazole derivative C16 reduces in vivo brain PKR activation. FEBS letters. 2007;581(23):4473-8. 144. Chen HM, Wang L, and D'Mello SR. A chemical compound commonly used to inhibit PKR, {8-(imidazol-4-ylmethylene)-6H-azolidino[5,4-g] benzothiazol-7-one}, protects neurons by inhibiting cyclin-dependent kinase. The European journal of neuroscience. 2008;28(10):2003-16. 145. Ill-Raga G, Palomer E, Wozniak MA, Ramos-Fernandez E, Bosch-Morato M, Tajes M, Guix FX, Galan JJ, Clarimon J, Antunez C, et al. Activation of PKR causes amyloid ss-peptide accumulation via de-repression of BACE1 expression. PloS one. 2011;6(6):e21456. 146. Axten JM, Medina JR, Feng Y, Shu A, Romeril SP, Grant SW, Li WH, Heerding DA, Minthorn E, Mencken T, et al. Discovery of 7-methyl-5-(1- {[3-(trifluoromethyl)phenyl]acetyl}-2,3-dihydro-1H-indol-5-yl)-7H-p yrrolo[2,3-d]pyrimidin-4-amine (GSK2606414), a potent and selective first-in-class inhibitor of protein kinase R (PKR)-like endoplasmic reticulum kinase (PERK). Journal of medicinal chemistry. 2012;55(16):7193-207. 147. Lai JY, Cox PJ, Patel R, Sadiq S, Aldous DJ, Thurairatnam S, Smith K, Wheeler D, Jagpal S, Parveen S, et al. Potent small molecule inhibitors of

spleen tyrosine kinase (Syk). Bioorganic & medicinal chemistry letters. 2003;13(18):3111-4. 148. Robert F, Williams C, Yan Y, Donohue E, Cencic R, Burley SK, and Pelletier J. Blocking UV-induced eIF2alpha phosphorylation with small molecule inhibitors of GCN2. Chemical biology & drug design. 2009;74(1):57-67. 149. Verkman AS. Drug discovery in academia. American journal of physiology Cell physiology. 2004;286(3):C465-74. 150. Novac O, Guenier AS, and Pelletier J. Inhibitors of protein synthesis identified by a high throughput multiplexed translation screen. Nucleic acids research. 2004;32(3):902-15. 151. Svitkin YV, Ovchinnikov LP, Dreyfuss G, and Sonenberg N. General RNA binding proteins render translation cap dependent. The EMBO journal. 1996;15(24):7147-55. 152. Bordeleau ME, Mori A, Oberer M, Lindqvist L, Chard LS, Higa T, Belsham GJ, Wagner G, Tanaka J, and Pelletier J. Functional characterization of IRESes by an inhibitor of the RNA helicase eIF4A. Nature chemical biology. 2006;2(4):213-20. 153. Lindqvist L, Robert F, Merrick W, Kakeya H, Fraser C, Osada H, and Pelletier J. Inhibition of translation by cytotrienin A--a member of the ansamycin family. RNA (New York, NY). 2010;16(12):2404-13. 154. Malina A, Khan S, Carlson CB, Svitkin Y, Harvey I, Sonenberg N, Beal PA, and Pelletier J. Inhibitory properties of nucleic acid-binding ligands on protein synthesis. FEBS letters. 2005;579(1):79-89. 155. Kawasumi M, and Nghiem P. Chemical genetics: elucidating biological systems with small-molecule compounds. The Journal of investigative dermatology. 2007;127(7):1577-84. 156. Chan J, Khan SN, Harvey I, Merrick W, and Pelletier J. Eukaryotic protein synthesis inhibitors identified by comparison of cytotoxicity profiles. Rna. 2004;10(3):528-43. 157. Gunnery S, Rice AP, Robertson HD, and Mathews MB. Tat-responsive region RNA of human immunodeficiency virus 1 can prevent activation of the double-stranded-RNA-activated protein kinase. Proc Natl Acad Sci U S A. 1990;87(22):8687-91. 158. Ung TL, Cao C, Lu J, Ozato K, and Dever TE. Heterologous dimerization domains functionally substitute for the double-stranded RNA binding domains of the kinase PKR. EMBO J. 2001;20(14):3728-37. 159. Brenke R, Kozakov D, Chuang GY, Beglov D, Hall D, Landon MR, Mattos C, and Vajda S. Fragment-based identification of druggable 'hot spots' of proteins using Fourier domain correlation techniques. Bioinformatics. 2009;25(5):621-7. 160. Kozakov D, Hall DR, Chuang GY, Cencic R, Brenke R, Grove LE, Beglov D, Pelletier J, Whitty A, and Vajda S. Structural conservation of druggable hot spots in protein-protein interfaces. Proc Natl Acad Sci U S A. 2011;108(33):13528-33.

161. Dar AC, Dever TE, and Sicheri F. Higher-order substrate recognition of eIF2alpha by the RNA-dependent protein kinase PKR. Cell. 2005;122(6):887-900. 162. Hall DR, Ngan CH, Zerbe BS, Kozakov D, and Vajda S. Hot spot analysis for driving the development of hits into leads in fragment-based drug discovery. J Chem Inf Model. 2012;52(1):199-209. 163. Bissantz C, Kuhn B, and Stahl M. A medicinal chemist's guide to molecular interactions. J Med Chem. 2010;53(14):5061-84. 164. Nasr Z, and Pelletier J. Tumor progression and metastasis: role of translational deregulation. Anticancer research. 2012;32(8):3077-84. 165. Pelletier J, and Peltz SW. 30 Therapeutic Opportunities in Translation. 2007. 166. Hershey JWB, and Merrick WC. Initiation of Protein Synthesis. Cold Spring Harbor: Cold Spring Habor Laboratory Press; 2000. 167. Chen JJ, and London IM. Regulation of protein synthesis by heme- regulated eIF-2 alpha kinase. Trends Biochem Sci. 1995;20(3):105-8. 168. Mikami S, Kobayashi T, Yokoyama S, and Imataka H. A hybridoma- based in vitro translation system that efficiently synthesizes glycoproteins. Journal of biotechnology. 2006;127(1):65-78. 169. Meijer L, Thunnissen AM, White AW, Garnier M, Nikolic M, Tsai LH, Walter J, Cleverley KE, Salinas PC, Wu YZ, et al. Inhibition of cyclin- dependent kinases, GSK-3beta and CK1 by hymenialdisine, a marine sponge constituent. Chemistry & biology. 2000;7(1):51-63. 170. Breton JJ, and Chabot-Fletcher MC. The natural product hymenialdisine inhibits interleukin-8 production in U937 cells by inhibition of nuclear factor-kappaB. J Pharmacol Exp Ther. 1997;282(1):459-66. 171. Tasdemir D, Mallon R, Greenstein M, Feldberg LR, Kim SC, Collins K, Wojciechowicz D, Mangalindan GC, Concepcion GP, Harper MK, et al. Aldisine alkaloids from the Philippine sponge Stylissa massa are potent inhibitors of mitogen-activated protein kinase kinase-1 (MEK-1). Journal of medicinal chemistry. 2002;45(2):529-32. 172. Wan Y, Hur W, Cho CY, Liu Y, Adrian FJ, Lozach O, Bach S, Mayer T, Fabbro D, Meijer L, et al. Synthesis and target identification of hymenialdisine analogs. Chemistry & biology. 2004;11(2):247-59. 173. Rockey WM, and Elcock AH. Rapid computational identification of the targets of protein kinase inhibitors. J Med Chem. 2005;48(12):4138-52. 174. Supriyono A, Schwarz B, Wray V, Witte L, Muller WE, van Soest R, Sumaryono W, and Proksch P. Bioactive alkaloids from the tropical marine sponge Axinella carteri. Zeitschrift fur Naturforschung C, Journal of biosciences. 1995;50(9-10):669-74. 175. Sharma V, Lansdell TA, Jin G, and Tepe JJ. Inhibition of cytokine production by hymenialdisine derivatives. J Med Chem. 2004;47(14):3700-3. 176. Tiwari RK, Kusari J, and Sen GC. Functional equivalents of interferon- mediated signals needed for induction of an mRNA can be generated by

double-stranded RNA and growth factors. The EMBO journal. 1987;6(11):3373-8. 177. Marcus PI, and Sekellick MJ. Interferon induction by viruses. XVI. 2- Aminopurine blocks selectively and reversibly an early stage in interferon induction. The Journal of general virology. 1988;69 ( Pt 7)(1637-45. 178. Tiwari RK, Kusari J, Kumar R, and Sen GC. Gene induction by interferons and double-stranded RNA: selective inhibition by 2- aminopurine. Molecular and cellular biology. 1988;8(10):4289-94. 179. Cheshire JL, Williams BR, and Baldwin AS, Jr. Involvement of double- stranded RNA-activated protein kinase in the synergistic activation of nuclear factor-kappaB by tumor necrosis factor-alpha and gamma- interferon in preneuronal cells. The Journal of biological chemistry. 1999;274(8):4801-6. 180. Chen HM, Wang L, and D'Mello SR. Inhibition of ATF-3 expression by B-Raf mediates the neuroprotective action of GW5074. Journal of neurochemistry. 2008;105(4):1300-12. 181. Wang X, Fan Z, Wang B, Luo J, and Ke ZJ. Activation of double-stranded RNA-activated protein kinase by mild impairment of oxidative metabolism in neurons. Journal of neurochemistry. 2007;103(6):2380-90. 182. Baltzis D, Pluquet O, Papadakis AI, Kazemi S, Qu LK, and Koromilas AE. The eIF2alpha kinases PERK and PKR activate glycogen synthase kinase 3 to promote the proteasomal degradation of p53. The Journal of biological chemistry. 2007;282(43):31675-87. 183. Sambrook J, and Russell DW. Molecular Cloning. A laboratory Manual. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press; 2001. 184. Trott O, and Olson AJ. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem. 2010;31(2):455-61. 185. Brooks BR, Bruccoleri RE, Olafson BD, States DJ, Swaminathan S, and Karplus M. Charmm - a Program for Macromolecular Energy, Minimization, and Dynamics Calculations. J Comput Chem. 1983;4(2):187-217.