Role of Mrna Surveillance Pathways During Oxidative Stress in Saccharomyces Cerevisiae a Thesis Submitted to the University of M

Role of mRNA surveillance pathways during oxidative

stress in Saccharomyces Cerevisiae

A thesis submitted to The University of Manchester

for the degree of

DOCTOR OF PHILOSOPHY

in the Faculty of Biology, Medicine and Health

2017

NUR HIDAYAH JAMAR

SCHOOL OF BIOLOGICAL SCIENCES

Table of Content

Table of Content 2

List of Figures 9

List of Tables 12

Declaration 13

Communications 14

Publication 14

Contributor’s acknowledgment 15

Acknowledgments 16

List of abbreviations 17

Abstract 21

1.0 introduction 23

1.1 Generation of reactive oxygen species (ROS) 23

1.2 Sources of ROS and commonly used ROS compounds 25

1.3 What happens when cells cannot handle oxidative stress? 26 1.3.1 Lipid peroxidation 27 1.3.2 Protein oxidation 29 1.3.3 Oxidatively damaged nucleic acids (DNA and RNA) 30

1.4.Transcriptional responses of S. cerevisiae during oxidative stress 31 conditions 1.4.1 Regulation of gene expression by Yap1 32

1.4.2 Modulation of the general stress response by MSN2/MSN4 33

1.5 Translational responses of S. cerevisiae to oxidative stress conditions 34 1.5.1 Overview of protein synthesis 34 1.5.1.1 Translation initiation 34 1.5.1.2 Translation elongation 37 1.5.1.3 Translation termination 39 1.5.2 Regulation of translation initiation during oxidative stress in S. 39 cerevisiae 1.5.2.1 Regulation of TC by eIF2α 41 1.5.2.2 Regulation of mRNA-specific translational control via Gcn4 43

1.6 Cytoplasmic mRNA degradation in S. cerevisiae 43 1.6.1 Normal mRNA degradation 44 1.6.2 Specialized mRNA quality control mechanisms 47 1.6.2.1 Nonsense-mediated decay (NMD) 48 1.6.2.2 Nonstop mRNA decay (NSD) 51 1.6.3.3 No-go mRNA decay (NGD) 53

1.7 Protein folding, misfolding, and aggregation 55

1.7.1 Amyloid aggregates 57 1.7.2 Amorphous aggregates 58 1.7.3 Protein aggregation related diseases 59

1.8 Prions 60 1.8.1 Mammalian prions 61 1.8.2 Yeast prions 64 1.8.2.1 [PSI+] 65 1.8.2.2 [PIN+] 67

1.9 Thesis objectives 68

2.0 Materials and Methods 69

2.1 Yeast strains and plasmids 69 2.1.1 Yeast strains 69 2.1.2 Plasmids 70

2.2 Strain construction and verification 71 2.2.1 Oligonucleotides 71 2.2.2 DNA amplification by polymerase chain reaction (PCR) 73 2.2.3 Yeast genomic DNA extraction 73

2.3 DNA/RNA manipulation and analysis 74 2.3.1 Plasmid extraction 74 2.3.2 Yeast transformation 75 2.3.3 Agarose gel electrophoresis 76 2.3.4 Quantitative Reverse Transcriptase PCR (qRT-PCR) 76 2.3.4.1 RNA extraction 77 2.3.4.2 RNA washes 77 2.3.4.3 q-RT PCR analysis 77 2.3.5 Quantitation of DNA/ RNA concentrations using a spectrophotometer 78

2.4 Media and growth conditions 78 2.4.1 S. cerevisiae 78 2.4.2 E. coli 79

2.5 Yeast growth and stress analysis 79 2.5.1 Spot tests for oxidant-stress sensitivity 79 2.5.2 Oxidant-growth sensitivity 79 2.5.3 Cell viability assay 80

2.6 Protein Analysis 80 2.6.1 Polysome analysis 80

2.6.1.1 Preparation of yeast cell extracts for polysome analysis 80 2.6.1.2 Preparation of sucrose gradients 81 2.6.1.3 Sedimentation of polyribosomes 82 2.6.2 35S cysteine/methionine radiolabelling 83 2.6.3 β-galactosidase reporter assays to measure stop codon readthrough 84 2.6.3.1 Preparation of whole cell extracts 84 2.6.3.2 Assay 84 2.6.4 Dual-luciferase reporter assays to measure stop codon readthrough 85 2.6.4.1 Preparation of whole cell extracts 85 2.6.4.2 Assay 86 2.6.5 Western blot analysis 86 2.6.5.1 Preparation of yeast whole cell extracts 86

2.6.5.2 Sodium dodecyl sulphate-polyacrylamide gel electrophoresis 87 (SDS-PAGE) and western blotting

2.7 Analysis of protein aggregation 89 2.7.1 Preparation of yeast cell cultures 89 2.7.2 Silver-staining of SDS-PAGE gels 90 2.7.3 Mass spectrometry 90 2.7.4 Bioinformatic analysis of aggregated proteins 90

2.8 Live-cell fluorescence microscopy 91 2.8.1 Preparation of yeast samples 91 2.8.2 DeltaVision fluorescence microscopy 91

2.9 Analysis of prion formation 91 2.9.1 Determination of de novo [PSI+] prion formation 92 2.9.2 Determination of de novo [PIN+] Prion Formation 93

3.0 Importance of mRNA surveillance pathways during 94 oxidative stress conditions

3.1 Introduction 94

3.2 Results 96 3.2.1 Construction of mRNA surveillance mutants 96 3.2.2 Phenotypic assays of [psi-] and [PSI+] strains 99 3.2.3 Requirement for mRNA surveillance pathways during oxidative stress 99 3.2.4 Growth and viability analysis of mRNA surveillance mutants during 100 oxidative stress conditions 3.2.4.1 The oxidant sensitivity of ski8 mutant depends on its [PSI+] 102 status 3.2.4.2 Loss of NMD may act to improve oxidant tolerance and this 104 phenotype is further observed in [PSI+] background 3.2.4.3 Mutants in the NGD pathway are modestly sensitive to 104 oxidative stress 3.2.5 Analysis of translational activity in mRNA surveillance mutants 106 during oxidative stress conditions 3.2.6 Analysis of protein synthesis in mRNA surveillance mutants during 108 oxidative stress conditions

3.3 Discussion 114

4.0 Requirement for nonstop decay during oxidative stress 116 conditions 4.1 Introduction 116

4.2 Results 118

4.2.1 Phenotypic assays of ski mutants 118

4.2.2 The SKI complex is required for oxidant tolerance 118

4.2.3 Analysis of NSD in ski complex and ski7 mutants 122

4.2.4 Overlapping requirements for Ski7 and Dom34/Hbs1 during oxidative 124 stress conditions 4.2.5 Oxidative stress Sup35 aggregation in dose-dependent manner and 128 potentially generates NSD substrate 4.2.6 Overexpression of Sup35 rescues oxidant sensitivity in WT and SKI 131 complex mutant strains 4.2.7 Overexpression of Sup35 decreases stop codon readthrough during 133 oxidative stress conditions in WT and ski2 mutant strains

4.3 Discussion 137

5.0 Importance of mRNA quality control for proteostasis 143

5.1 Introduction 143

5.2 Results 145 5.2.1 Loss of mRNA surveillance mutants causes widespread protein 145 aggregation 5.2.2 More proteins are susceptible to aggregation following loss of mRNA 147 surveillance mutant strains 5.2.3 Differences in subcellular localization of proteins isolated from 148 aggregates in mRNA surveillance mutants 5.2.4 Enrichment of functional categories within protein aggregates isolated 150 from mRNA surveillance mutants 5.2.5 Analysis of the physicochemical properties of aggregated proteins 152 identified in mRNA surveillance mutants 5.2.5.1 Aggregated proteins are enriched for abundant and highly 153 expressed proteins 5.2.5.2 Aggregated proteins are more hydrophobic than the 155 unaggregated ones 5.2.6 Analysis of selected amino acid composition in mRNA surveillance 157 mutant aggregates 7

5.2.7 Molecular chaperones are present within protein aggregates 159 5.2.8 An increased frequency of de novo prion formation occurs in NMD 161 mutants 5.2.9 The frequency of induced [PSI+] formation is increased in NMD 164 mutants 5.2.10 The [PSI+] status of strains improves cell viability upon exposure to 166 various stress conditions 5.3 Discussion 170

6.0 General discussion 177

6.1 Possible interaction between NGD complex and Gcn2 kinase 181 6.2 Aggregation of Sup35 upon oxidative stress: where does it take us? 182 6.3 Amorphous aggregates may potentially be more toxic than amyloid 183 aggregates 6.4 The presence of 22 PTCs in strain 74D-694 may actually contribute to 184 increased tolerance to various stresses particularly in NMD mutants

7.0 Bibliography 188

8.0 Appendix 201

Appendix 1 List of protein aggregates present within WT strain 201

Appendix 2 List of protein aggregates present within ski7 mutant 202

Appendix 3 List of protein aggregates present within ski8 mutant 204

Appendix 4 List of protein aggregates present within upf2 mutant 206

Appendix 5 List of protein aggregates present within hbs1 mutant 208

List of Figures

Figure 1.1 Mechanism of ROS production and conversion 24

Figure 1.2 Mechanism of lipid peroxidation 28

Figure 1.3 Simplified overview of the translation initiation process in 35 eukaryotes

Figure 1.4 Mechanism of translation elongation 38

Figure 1.5 Mechanism of translation termination in yeasts 40

Figure 1.6 Regulation of translation initiation by inhibiting TC formation 4 2

Figure 1.7 Normal mRNA decay mechanisms in eukaryotes 45

Figure 1.8 Recognition of NMD substrates 49

Figure 1.9 Recognition of NSD substrates 53

Figure 1.10 Recognition of NGD substrates 55

Figure 1.11 Comparison of the mechanisms between mammalian and yeast 62 [PSI+] prion formations

Figure 2.1 An example of a normal, untreated polysome trace from actively 83 translating yeast cells

Figure 3.1 Construction of mRNA surveillance mutants in strain 74D-694 97 using HIS3 deletion cassette

Figure 3.2 Construction of mRNA surveillance mutants in strain 74D-694 98 using HIS3 deletion cassette

Figure 3.3 Oxidative stress-sensitivity of mRNA surveillance mutants 101

Figure 3.4 ski8 mutant is hypersensitive to oxidative stress conditions but 103 this sensitivity is lost in the [PSI+] version

Figure 3.5 The oxidant tolerance of NMD mutants depends on their [PSI+] 105 status

Figure 3.6 Mutants in NGD pathway are modestly sensitive to oxidative 107

stress

Figure 3.7 Differences in oxidant sensitivity do not arise due to translational 109 activity in NSD mutants

Figure 3.8 NMD mutants exhibit moderate effect of inhibition of translation 110 initiation

Figure 3.9 Inhibition of translation initiation was observed in NGD mutants 111 even in unstressed condition

Figure 3.10 Hydrogen peroxide causes inhibition of protein synthesis all 113 strains following 1 hour treatment

Figure 4.1 Mutants in the Ski complex are sensitive to oxidative stress 119

Figure 4.2 Mutants in the Ski complex are sensitive to oxidative stress 121

Figure 4.3 Analysis of NSD in ski7 and ski complex mutants 123

Figure 4.4 Overlapping requirements for Ski7 and Dom34/Hbs1 during 125 oxidative stress conditions

Figure 4.5 Overlapping requirements for Ski7 and Dom34/Hbs1 during 127 oxidative stress conditions

Figure 4.6 Aggregation of Sup35 occurs in response to hydrogen peroxide 130 stress

Figure 4.7 Overexpression of SUP35 rescues oxidant sensitivity in WT and 132 SKI complex mutant strains

Figure 4.8 Overexpression of SUP35 rescues stop codon readthrough in WT 134 and SKI complex mutant strains

Figure 4.9 Overexpression of SUP35 rescues codon readthrough in WT and 136 ski2 mutant strains following hydrogen peroxide treatment

Figure 5.1 Strains lacking components of mRNA surveillance pathways have 146 higher levels of protein aggregation

Figure 5.2 Overlap of identified proteins within protein aggregates isolated 148

from WT and mRNA surveillance mutant strains

Figure 5.3 Subcellular localization prediction of aggregated proteins identified 150 in WT and mRNA surveillance mutant strains

Figure 5.4 MIPS functional categorization of aggregated proteins identified in 152 WT and mRNA surveillance mutant strains

Figure 5.5 Proteins within aggregates are abundant and highly expressed 155

Figure 5.6 Analysis of physicochemical properties within protein aggregates 157

Figure 5.7 Analysis of selected amino acid composition within protein 159 aggregates

Figure 5.8 Overlap of identified protein chaperones within protein aggregates 161 isolated from WT and mRNA surveillance mutant strains

Figure 5.9 Loss of NMD factors (UPF1 and UPF2) result in an increased 164

frequency of [PSI+] and [PIN+] prion formation

Figure 5.10 Induction of [PSI+] prion formation is uniquely observed in NMD 166 mutant strains

Figure 5.11 Survivability of WT and NMD mutant strains during stress 168

conditions are improved when they are in [PSI+] form

Figure 6.1 Summary of main findings and proposed models in this Study 180

List of Tables

Table 1.1 Comparison of main characteristics between mammalian and yeast 63 prions

Table 2.1 Yeast strains used in this study 69

Table 2.2 Plasmids used in this study 70

Table 2.3 Primers and sequences used for strain construction and 71 sequencing

Table 2.4 Primers used for qRT-PCR 76

Table 2.5 Preparation of pure sucrose gradients (15-50%) 82

Table 2.6 Primary antibodies used in this study 88

Table 2.7 Secondary antibodies used in this study 88

Declaration No portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning.

Copyright statement i. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the “Copyright”) and s/he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes. ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. iii. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the “Intellectual Property”) and any reproductions of copyright works in the thesis, for example graphs and tables (“Reproductions”), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions. iv. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=24420), in any relevant Thesis restriction declarations deposited in the University Library, The University Library’s regulations (see http://www.library.manchester.ac.uk/about/regulations/) and in The University’s policy on Presentation of Theses.

Communications (Poster)

British Yeast Group Meeting, The University of Manchester, UK; March, 2015

Nur Hidayah Jamar and Chris Grant. Role of mRNA surveillance pathways in the oxidative response of S. cerevisiae.

Summer Post-Graduate Research symposium, The University of Manchester,

UK; July, 2015

Nur Hidayah Jamar and Chris Grant. Role of mRNA surveillance pathways in the oxidative response of S. cerevisiae.

BioProNET meeting: Molecular chaperones: structure, function and application in bioprocessing, University of Kent, UK; December, 2016

Nur Hidayah Jamar and Chris Grant. Role of nonstop decay during oxidative stress in S. cerevisiae.

Publication

Jamar, N. H., Kritsiligkou, P., & Grant, C. (2017). The nonstop decay mRNA surveillance pathway is required for oxidative stress tolerance. Nucleic Acids Research, 45(11), 6881-6893.

Contributor’s acknowledgment

Besides working closely with Prof Chris Grant for the work described in this study, some of the work has been supported by Dr Paraskevi Kritsiligkou (The

University of Manchester, UK). As a contributor, Dr Paraskevi helped to perform the experiments shown in Figure 4.1C and the negative controls for Figure 4.6D. The contributor also helped in editing (Fig. 5.2) and generating several figures in Chapter

5: Figure 5.3, Figure 5.4, Figure 5.5, and Figure 5.6.

Acknowledgements In the name of Allah, the Most Gracious, the Most Merciful.

During the 2nd week of May 2014, I decided to change my lab and supervisor. I never regretted my decision to enter this lab and it is because of these people that I made it to where I am now-so here it goes.

I am most grateful to my supervisor, Prof Chris Grant, for his continuous support, guidance and feedback throughout my PhD and thesis writing. I thank my advisor, Prof Mark Ashe for his constructive feedbacks especially during lab meetings. I am largely indebted to Prof Graham Pavitt for his constructive and meaningful guidance as well as emotional support throughout my PhD life.

I would like to express my gratitude to the past and present members in Chris’ lab: Dr Arunkumar, Dr Shaun, Fadilah, Dr Irina, Dr Pari, Sarah, Dr Vicky D, and Jana. I would also like to thank both members within Mark’s and Graham’s lab particularly Ebele, Dr Chris K, Dr Martin, Henry, Dr Jess, Dr Jenny, Fabian, Tawni and Benga. Despite my zero background in this research, you guys have put up with me and taught me many things including friendship and I am truly thankful for it.

I am most thankful to my parents and sisters for putting up with me for all these years. Spending more than 10 years away from all of you had made me realized that family is there for a reason. Thank you for putting up with my selfishness, and for supporting me. Thanks to my parents in-law too for your wonderful support and for accepting me and the situation that I am into.

To Sallehin, I could have never made this far without you. Our life is so crazy especially for the past few months where you had to work more than 12 hours a day to support me for close to a year. I am sorry for a lot of reasons and I am more than thankful to you for putting up with everything.

To You, thank you for being there. My life is colourful thanks to You.

Finally, I would like to thank both of my sponsors- The Ministry Education of Malaysia as well as Universiti Kebangsaan Malaysia for providing scholarship for the past 3 years. I am more than grateful to receive a 6 months bursary award in terms of tuition fees as well as living allowance from The University of Manchester.

I realize that doing PhD is a very selfish thing to do and that a lot of people have to adjust for me. Hence, I promise to be even more humble in everything that I do from now on and be a better student, teacher, researcher, and as a human being. Thank you from the bottom of my heart.

List of abbreviations

¼ YPD quarter-YPD 3D three-dimensional 4EBP 4E-binding protein 4-HNE 4-hydroxynonenal 8oxodG 8-hydroxy-deoxyguanosine 8oxoG 8-hydroxy-guanosine aa-tRNA amino-acyl tRNA uORF upstream open reading frame UPF upstream frameshifting UTR untranslated region AD Alzheimer’s disease ADE adenine ALS amyotrophic lateral sclerosis AP-1 activating protein-1 ATP adenosine triphosphate bp basepair BSE bovine spongiform encelophaties bZIP basic leucine zipper domain cAMP cyclic adenosine monophosphate CJD Creutzfeldt -Jakob diseases CRD cysteine rich domains C-terminal carboxyl-terminal DSE downstream element DNA deoxyribonucleic acid E. coli Escherichia coli eEF eukaryotic elongation factor eIF eukaryotic initiation factor 17

eRF eukaryotic release factor ER endoplasmic reticulum for forward GCN general control nonrepressed GdnHCl guanosine hydrochloride GDP guanosine diphosphate GEF guanosine nucleotide exchange factor GFP green fluorescent protein GSH glutathione GTP guanosine triphosphate

H2O water

H2O2 hydrogen peroxide

- HO2 perhydroxyl HD Huntington’s disease HIS histidine Hsp heat shock protein LOO. lipid peroxyl radical LOOH lipid hydroperoxide MDA malondialdehyde

Met Met-tRNA i initiator methionyl-tRNA mRNA messenger RNA MSN multiple suppressor of SNF1 mutation mTORC1 mammalian target of rapamycin complex 1 NES nuclear export sequence NGD no-go decay NLS nuclear localization signal nm nanometer NMD nonsense-mediated decay NRF2 nuclear factor (erythroid-derived 2)-like 2 18

NSD nonstop decay nt nucleotide

O2 oxygen

.- O2 superoxide anion OH. hydroxyl radical ORF open reading frame [PSI+] the prion form of Sup35 [PIN+] [PSI+] inducibility element Pab1 poly (A)-binding protein PCR polymerase chain reaction PD Parkinson’s disease PE phosphatidylethanolamine PFD prion forming domain Pgk1 phosphoglycerate kinase PKA protein kinase A p/m polysome to monosome ratio PRNP prion protein gene Prp prion protein PTC premature termination codon PUFA polyunsaturated fatty acid PQC protein quality control rev reverse RNA ribonucleic acid ROS reactive oxygen species S.cerevisiae Saccharomyces cerevisiae SCD synthetic complete defined SOH sulfenic acid

SO2H sulfinic acid

SO3H sulfonic acid 19

SKI superkiller SOD superoxide dismutase STRE stress response element TC ternary complex TPR tetratricopeptide tRNA transfer RNA TRX thioredoxin gene uORF upstream open reading frame UPF upstream frame-shifting WT WT WD tryptophan-aspartic acid (beta-transducer repeat) YAP yeast activator protein YPD yeast extract peptone dextrose

Word count: 47,014 words 20

Abstract

Oxidative stress can result in oxidative damage to most cellular macromolecules including DNA, RNA and protein, and this damage has been implicated in ageing and cell death. Not surprisingly therefore, eukaryotic cells contain quality control systems which monitor mRNAs for errors that might cause the production of aberrant proteins. This study focuses on three translation-associated mRNA surveillance pathways which reduce the production of potentially toxic proteins in Saccharomyces cerevisiae: nonsense-mediated decay (NMD), no-go decay (NGD) and nonstop decay (NSD).

The data presented in Chapters 3 and 4 indicate that factors required for the recognition of NSD substrates and components of the SKI complex are required for oxidant tolerance. An overlapping requirement for Ski7, which bridges the interaction between the SKI complex and the exosome, and NGD components (Dom34/Hbs1) which have been shown to function in both NSD and NGD was also observed.

Additionally, both ski7 dom34 and ski7 hbs1 double mutants are sensitive to hydrogen peroxide stress and accumulate an NSD substrate. A model is presented to explain the generation of NSD substrates where oxidative stress causes aggregation of the Sup35 translation termination factor, which increases stop codon readthrough allowing ribosomes to translate into the 3’-end of mRNAs. Consistent with this model, overexpression of Sup35 was observed to decrease stop codon readthrough and rescued oxidant tolerance in both WT and ski2 mutant strains. In general, our data revealed an unanticipated requirement for the NSD pathway during oxidative stress conditions which reduces the production of aberrant proteins from

NSD mRNAs.

The data presented in Chapter 5 further examines protein aggregation in mRNA quality control mutants, since aggregation has been implicated in many neurodegenerative diseases including Alzheimer’s. Protein aggregation can be categorized as either disordered amorphous aggregation or highly organized amyloid aggregation. Loss of any mRNA surveillance pathway results in increased widespread protein aggregation and bioinformatic analysis indicates that increased aggregation of aggregation-prone proteins occurs in mRNA surveillance mutants.

Moreover, increased [PSI+] and [PIN+] prion formations were uniquely observed in

NMD mutants. In Chapter 3, the [PSI+] status improved viability when strains were exposed to hydrogen peroxide but this phenotype is not just limited to oxidative stress but includes other stress conditions such as heat or osmotic shocks as expanded in Chapter 5. Taken together, these data suggest that mRNA surveillance pathways are important for maintaining the integrity of protein production during both normal and stress conditions.

1.0 Introduction

The optimal growth and cellular functioning of all cell types takes place under a well-defined and controlled internal environment known as homeostasis. When an organism is exposed to chemical or physical conditions that disrupt this internal environment, a series of intricate stress response mechanisms are activated. These mechanisms include activation of sensing and signaling cascades which eventually lead to adjustment of the gene expression program to repair possible damage, to adjust metabolic activities so that cells can perform their normal routines, and to ensure survival. In a more general context, cell stress refers to any environmental conditions or factors that might activate a stress response mechanism within cells.

Stress-generating conditions such as increases in temperature, depletion of nutrients, changes in osmolarity and pH, presence of oxidative elements and/or radicals are all types of cell stress that must be maintained within the tolerance range of the organism so that homeostasis is achieved.

Stress responses have been extensively studied in the yeast S. cerevisiae.

We now know that cells can induce both stress-specific responses and also activate a collective response to multiple types of stress which is commonly referred as a general stress response. A key stress that cells are commonly exposed to is oxidative stress and this shall be explored further in the following section.

1.1 Generation of reactive oxygen species (ROS)

ROS are ubiquitous molecules formed as a result of singlet oxygen which reacts indiscriminately by attacking electrons from other molecules including oxygen ions, free radicals and peroxides (Fig. 1.1). ROS are unavoidable in living cells as

Figure 1.1: Mechanism of ROS production and conversion. There are a number of stimuli that can trigger the generation of ROS species either exogenously or endogenously. This includes UV irradiation, chemical agents, and during events of normal aerobic metabolism. Several antioxidants such as superoxide dismutases, glutathione and catalases act as defense mechanisms to convert ROS into less - dangerous, harmful substrates. For instance, the superoxide anion (O2 ) is first reduced into H2O2 by the cooperation of superoxide dismutases (encoded by SOD1 and SOD2 in yeast) and final conversion into H2O by the action of glutathione (GSH) and catalases (encoded by CTT1 and CTA1). Disruption of any of these intermediate pathways will result in accumulation of ROS via Fenton and Haber-Weiss reactions.

they are generated as mitochondrial by-products following aerobic metabolism. ROS are also generated during inflammation processes by neutrophils and macrophages as an important component of immunological defences against pathogens. Under physiological condition, the rate of generation of ROS appears to be balanced with the status of the antioxidant defense systems. Despite these defense systems, an oxidative stress is said to occur when there is an imbalance between radical production and stress protection leading to accumulation of ROS.

1.2 Sources of ROS and compounds commonly used to generate ROS

H2O2 is a ubiquitous molecule, which as well as being both freely diffusible and reactive, must be removed from cells by defense mechanisms to avoid Fenton and Haber-Weiss reactions (Fig. 1.1) leading to the formation of highly reactive hydroxyl radicals that may attack other molecules and produce more reactive species (Gutteridge, 1993; Halliwell and Gutteridge, 1989). However, during normal physiological conditions, H2O2 can serve as both an intra- and inter-cellular messenger in regulating many biological processes, including; apoptosis; gene expression and many cell signalling cascades (Veal et al., 2007). Diamide is one of the reactive thiol compounds that is commonly used to induce oxidative stress in yeast by reacting with glutathione (GSH). This is achieved by shifting the redox state of glutathione into an oxidised state (GSSG) via disulphide formation, thereby suppressing its role as a cofactor for many antioxidant enzymes in neutralising ROS

(Kosower and Kosower, 1995; Muller, 1996).

Metal ions such as iron, copper and cadmium have also been shown to be involved in ROS production. Iron (Fe2+) and Copper (Cu2+) are redox active metal

ions and therefore able to directly induce ROS production through Fenton and

Haber-Weiss reactions (De Freitas et al., 2003). Cadmium on the other hand is a heavy metal that indirectly induces oxidative stress by several mechanisms, including: 1) depletion of antioxidant defense mechanism such as GSH and sulfhydryl groups; 2) displacement of Fenton-active metals by freeing bound- iron/copper from membrane proteins for example and therefore able to participate in

Fenton and Haber-Weiss reactions; and 3) inhibition of mitochondrial transport chain

(Brennan and Schiestl, 1996).

Given the potential toxicity of ROS, it is not surprising therefore that most organisms (if not all) have intricate antioxidant defense systems to quickly convert

ROS into less harmful substrates and prevent damages to cellular components that are critical for their survival (Fig. 1.1). It is known that eukaryotes including S. cerevisiae have multiple antioxidant defense systems and further categorized into

- enzymatic and non-enzymatic antioxidants. For example, the superoxide anion (O2 ) is reduced into water (H2O) by the cooperation of two main antioxidant enzymes: superoxide dismutases (encoded by SOD1 and SOD2 in yeast) and catalases

(encoded by CTT1 and CTA1). Additionally, a number of non-enzymatic antioxidants, such as GSH can also bind and detoxify ROS. However, if these antioxidants are overwhelmed, and there is an imbalance towards the pro-oxidative state, an oxidative stress is said to occur. Thus, all aerobic life forms must maintain a balance between the production and removal of ROS.

1.3 What happens when cells cannot handle oxidative stress?

The formation of ROS is unavoidable in living cells since they are generated as by-products of normal aerobic metabolism. During oxidative stress conditions

however, levels of ROS increase that if left unattended may cause damage to macromolecules including proteins, lipids, and nucleic acids (DNA and RNA) leading to cell death or mutagenesis. This section focuses on the effect of oxidative stress on these macromolecules and how it is associated with many acute and chronic disorders including: cancer, diabetes, cardiovascular, and neurodegenerative diseases as discussed below.

1.3.1 Lipid peroxidation

Lipid peroxidation is a natural metabolic process and is best described as a chain termination process divided into three stages: initiation, propagation and termination (Fig. 1.2). The initiation step is a rate-limiting process and begins when a

ROS (e.g. OH.) attacks lipids containing carbon-carbon double bond(s), especially polyunsaturated fatty acids (PUFA) within cell membranes and removes the especially reactive hydrogen atom to form a carbon-centered lipid radical. During the propagation step, an oxygen is inserted to form lipid peroxyl radicals (LOO.) and hydroperoxides (LOOH), and finally, the reaction is terminated once the two radicals form paired stable electrons and produces a non-reactive species (Henkel, 2011).

. . Both hydroxyl radical (OH ) and hydroperoxyl (OH2 ) are the two most prevalent ROS that can directly oxidize lipids particularly PUFA’s during lipid peroxidation reactions resulting in changes to cell membrane integrity, impairing cellular functioning and finally, cell rupture (Ayala et al., 2014). While OH. is able to form lipid peroxyl radicals which are highly reactive and able to propagate the chain reaction, the hydroperoxyl radical on the other hand forms lipid hydroperoxides which in turn generates new ROS species (Ayala et al., 2014).

Figure 1.2: Mechanism of lipid peroxidation. Cell membranes are susceptible to oxidative stress due to high content of polyunsaturated fatty acids (PUFA). During initiation process, hydroxyl radical (OH.) or any other radicals takes a hydrogen atom from PUFA/polyunsaturated lipid (LH), thereby forming a lipid radical (L.). The free radical chain reaction is propagated by reaction with oxygen (O2) forming lipid the peroxyl radical (LOO.) which continues to attack PUFA’s and generates both lipid hyroperoxyl and lipid radicals. The chain reaction is terminated once two radical species donate single electrons and produce a non-radical species. Adapted from (Henkel, 2011).

There are several secondary products of lipid peroxidation which are known to significantly affect biological processes including reactive aldehydes such as malondialdehyde (MDA), 4-hydroxyalkenals and 4-hydroxynonenal (4-HNE) (Ayala et al., 2014; Gutteridge, 1981). MDA for instance has been shown to be both mutagenic and carcinogenic as it readily reacts with components of DNA to form

DNA adducts. On the other hand, 4-HNE is deemed as the most toxic species due to its ability to attack other covalent macromolecules and is sometimes described as a

‘second toxic messenger of free radicals’ due to its involvement with several transcription factors such as nuclear factor (erythroid-derived 2)-like 2 (Nrf2) and activating protein-1 (AP-1). Additionally, due to their stability and its diffusive ability, both MDA and 4-HNE have been extensively used as biomarkers for oxidative stress in tissues.

1.3.2 Protein oxidation

Proteins are a main target of oxidative stress due to their cysteine residues which are often found at catalytic and regulatory sites in proteins and enzymes

(Stadtman and Levine, 2000). Upon oxidative stress, cysteine residues are readily oxidized forming reversible [sulfenic acid (SOH), disulfide bonds (S—S)] and irreversible oxidative states [sulfinic (SO2H) and sulfonic acids (SO3H)] and this may lead to breakage between amino acid residues (Eaton, 2006). During such events, protein quality control (PQC) mechanisms including the ubiquitin-proteasome system may be activated to degrade any oxidized proteins that may have accumulated within the cytoplasm, nucleus, and endoplasmic reticulum (ER) (Hohn et al., 2014).

Whilst mildly oxidized proteins are readily degraded by PQC, severely oxidized

proteins including cross-linked proteins, proteins that are modified by glycation or metal-ion catalysed oxidation of proteins are found to be poor substrates for proteases (Hohn et al., 2014). Therefore, the removal of mildly oxidized protein before they are severely cross-linked is essential to prevent the accumulation of aggregated, toxic proteins which may contribute to the development of certain pathological diseases, particularly those that are related to ageing processes such as Alzheimer’s disease (AD), Parkinson’s disease (PD), and Huntington’s disease

(HD) (Thomas and Mallis, 2001).

1.3.3 Oxidatively damaged nucleic acids (DNA and RNA)

RNA and DNA are major macromolecules which essential for all living organisms. They are mainly composed of sequence of nucleotides (nitrogenous bases, 5-carbon sugar and one or more phosphate groups) held together by phosphodiester bonds. Every component within the DNA/RNA structure is highly susceptible to ROS attack and this may lead to a multitude of DNA/RNA-damaged products. For example, exposure of DNA/RNA to hydroxyl radicals (OH.) can either cause addition of hydrogen atoms to its nitrogenous base or removal of hydrogen atoms from the backbone. To date, more than 20 different types of hydroxyl-radical induced modifications of DNA/RNA structures including alterations in DNA/RNA bases, double-strand breaks and DNA cross-linking have been described (Valko et al., 2007).

Due to its low oxidation potential among other bases, guanine is readily attacked by ROS leading to the production of 8-hydroxy-deoxyguanosine (8oxodG) for DNA (also known as 8oxoG for RNA). This is a major mutagenic oxidative lesion in the genome used as a biomarker of oxidative DNA/RNA damage (Valko et al.,

2007). Other frequently detected lesions are 8-hydroxyadenine, 2,6-diamino-4- hydroxy-5-formamidoguanine, 4,6-diamino-5-formamidoadenine, and cytosine that can adversely affect the cell cycle and potentially lead to genetic defects (Fimognari,

2015; Fleming and Burrows, 2016). It is therefore not surprising that oxidative

DNA/RNA damage has been implicated in many human diseases including cancer and many neurodegenerative diseases such as AD, PD, and amyotrophic lateral sclerosis (ALS) (Fimognari, 2015).

Over the past decade, studies on oxidatively damaged nucleic acid have focused more on RNA than DNA for a number of reasons: 1) unlike DNA, RNA is mostly single-stranded and therefore its bases are not protected by hydrogen bonds that makes them more accessible to ROS; 2) RNA has lesser association with protective proteins than DNA; 3) RNAs mostly consist of cytoplasmic RNA, where mitochondrial ROS are directed; 4) more importantly, unlike several repair mechanisms found for damaged DNA, no repair mechanisms are known for oxidatively-damaged RNA (Li et al., 2006). Indeed, several studies have demonstrated that higher levels of oxidative damage are present in RNA than in

DNA for rat liver, human leukocytes and lung epithelial cells (Wurtmann and Wolin,

2009). It is therefore necessary to remove oxidized RNA before it can significantly impair other cellular molecules/functions including protein synthesis or other functional RNAs.

1.4. Transcriptional responses of S. cerevisiae during oxidative stress conditions

There is no doubt that the cell’s ability to sense and respond to alterations to achieve homeostasis is vital for cell survival. The type of stressor defines how the

stress is sensed by the cell, what type of signal transduction is initiated, which transcriptional factor(s) control(s) the expression level of the genes to be induced or repressed, and so forth. The characteristics of the stress responses mediated by the

Yap1 and Msn2/Msn4 transcription factors described in the following sections are just a small part of a much more elaborate response mechanism initiated by the cell to respond to a wide set of conditions that threaten its internal homeostasis.

1.4.1 Regulation of gene expression by Yap1

In the yeast S. cerevisiae, oxidative stress is mostly sensed by the yeast activator protein-1 (Yap1) transcription factor, a redox-sensitive signalling protein.

Yap1 contains six cysteine residues which can be equally divided into two main domains: amino (n)-terminal (n-CRD) and a carboxyl (c)-terminal (c-CRD) cysteine rich domains. During oxidative stress, Yap1 localizes to the nucleus where it activates transcription of many stress response genes including GSH and thioredoxin (Herrero et al., 2008; Morano et al., 2011). Early studies revealed two known mechanisms in which Yap1 may be activated in the presence of different oxidants. While thiol reactive compounds such as diamide only require the c-CRD of

Yap1 to activate TRX2 gene expression, both CRDs of Yap1 are required to activate

TRX2 upon hydrogen peroxide stress (Azevedo et al., 2003; Kuge and Jones, 1994;

Morano et al., 2011). Additionally, activation of both of these domains require additional factors including Gpx3 as well as Ybp1, a Yap1- binding protein required for the in vivo folding of Yap1 (Morano et al., 2011). Taken together, Yap1 is sensitive to specific oxidants such as hydrogen peroxide and diamide due to different arrangements between its CRDs which in turn, enables it to induce selective antioxidant genes.

1.4.2 Modulation of the general stress response by MSN2/MSN4

Msn2 and Msn4 (Msn2/4) are transcription factors that play important roles in regulating gene expression in response to several stresses, including heat shock, osmotic shock, oxidative stress, low pH, glucose starvation, sorbic acid and high ethanol concentrations (Hasan et al., 2002; Martinez-Pastor et al., 1996). Both

MSN2 and MSN4 genes encode partially redundant zinc finger transcription factors with several domains: a transcription-activating domain located in its N-terminal region, a nuclear export sequence (NES) domain, a nuclear localization signal domain (NLS), and a zinc-finger binding domain located at its C-terminal domain that binds to the stress response element (STRE) (5’-CCCCT-3’/5’-GGGGA-3’) in promoters of stress-response genes (Martinez-Pastor et al., 1996).

Similar to Yap1, both Msn2/4 are localized in the cytoplasm under normal physiological conditions. In the presence of stressors however, Msn2/4 are hyperphosphorylated and translocate to the nucleus via the exportin Msn5 where they bind to the STRE and activate STRE-dependent gene expression (Alepuz et al.,

1999; Görner et al., 2002). Both Msn2/4 display oscillatory patterns as they shuttle between the cytoplasm and nucleus and this process is negatively controlled via the cAMP/PKA pathway, the major glucose metabolism pathway in cells (Jacquet et al.,

2003). Several studies have highlighted the difference in oscillatory pattern between

Msn2 and Msn4 as Msn4 continues to oscillate while Msn2 is permanently localized in the nucleus when PKA is absent (Jacquet et al., 2003). Nevertheless, activation of these transcription factors leads to the regulation of transient change in the expression level of around 900 genes in S. cerevisiae genome (Causton et al.,

2001; Gasch et al., 2000).

1.5 Translational responses of S. cerevisiae to oxidative stress conditions

1.5.1 Overview of protein synthesis

Translation of the mRNA into protein takes place on the ribosome, a large nucleoprotein assembly found in the cytoplasm of all cells. The ribosome consists of a large 60S subunit and a small 40S subunit, which performs peptidyl transfer and decoding, respectively. The process of translation can be divided into three stages:

Met (1) the initiation phase which places an initiator tRNA (Met-tRNA i) at the start codon of the mRNA in the 80S initiation complex with the help of at least 12 eukaryotic initiation factors (eIFs), (2) the elongation phase which is a repeated cycle of aminoacyl-tRNA (aa-tRNA) delivery and peptide bond formation; and finally, (3) the termination phase which is mainly carried out by eukaryotic release factors

(eRFs) once the acceptor site on the ribosome meets one of the three stop codon(s)

(UAA, UAG, UGA).

1.5.1.1 Translation initiation

Translation is first initiated via interaction between the cap-binding protein complex eIF4F and the m7GpppN-cap structure located at the 5’ end of mRNAs

(Kapp and Lorsch, 2004; Preiss and Hentze, 2003). eIF4F comprises of three components: eIF4E which acts as the cap-binding protein, eIF4A which is a RNA helicase, and a scaffolding protein eIF4G that bridges the mRNA and the ribosome via eIF3 (Fig. 1.3). With the help of eIF1, eIF1A, and eIF3, the ternary complex (TC) which consists of eIF2, initiator-Met-tRNA (tRNAMeti) and GTP, binds to a 40S ribosomal subunit to form a 43S pre-initiation complex and subsequently binds with mRNA. mRNA activation is then achieved by binding to the eIF4E site of the eIF4F complex resulting in the formation of a 48S initiation complex. Once the 48S complex

Figure 1.3: A simplified overview of the translation initiation process in eukaryotes. (1) Formation of the 43S pre-initiation complex comprising the TC (eIF2–GTP-tRNAMeti) and a 40S ribosomal subunit associated with eIF1, eIF1A, eIF3, and eIF5. (2) mRNA activation by the eIF4F complex and attachment to the 43S complex. The mRNA cap structure binds to eIF4E, where it unwinds the secondary structure of the 5’-UTR in an ATP-dependent manner. Both eIF4B and Pab1 are recruited to the 43S complex. (3) Scanning in a 5′ to 3′ direction for the start codon by the 43S complex. Once the initiation codon is recognised, eIF2-GTP is hydrolysed and phosphate (Pi) is released to form the 48S initiation complex. (4) A 60S ribosomal subunit is joined to the 48S complex via eIF5B-GTP. The rest of the initiation factors including eIF2-GDP, eIF1, eIF3, eIF4B, eIF4F and eIF5 are subsequently released. (5) Hydrolysis of eIF5B-GTP subsequently releases eIF1A to form an 80S initiation complex which is competent for translation elongation. Adapted from (Jackson et al., 2010).

recognizes the start codon (AUG), the eIF2-GTP is hydrolysed by a GTPase- activating protein, eIF5. The eukaryotic initiation factors (eIF1, eIF3, eIF4B, eIF4F and eIF5) are then released and the 60S ribosomal subunit is recruited to the 48S complex by eIF5B-GTP. GTP hydrolysis by eIF5B further releases eIF1A and guanosine diphosphate (GDP)-bound eIF5B. Further dissociation releases the rest of the bound initiation factors and eIF5B to make the 80S initiation complex and start translation. The initiator-Met-tRNA is now bound to the P site of the ribosome and this marks the beginning of the elongation phase.

1.5.1.2 Translation elongation

Ribosomes contain three tRNA binding sites: 1) the A-site (aminoacyl) where the aminoacyl-tRNA (aa-tRNA) binds; 2) the P-site where formation of the peptide bond occurs and; 3) the E-site where deacylated tRNAs exits the ribosome (Kapp and Lorsch 2004). The elongation process starts with eukaryotic elongation factor 1A

(eEF1A)-GTP bound with aa-tRNA entering the A-site. Once the anticodon of the aa- tRNA matches with the codon of the mRNA, eIF1A gets hydrolysed to form a peptide bond with the amino acid located on the P-site, resulting in: 1) a deacylated tRNA on the P-site and a newly formed peptidyl-tRNA on the A-site and; 2) eEF1A-GDP is recycled back to (eEF1A)-GTP by eukaryotic elongation factor 1B (eEF1B). The ribosome then moves the codon on the mRNA with the help of eEF2 leading to the movement of deacylated tRNA into the E-site via eEF3 and peptidyl-tRNA to the P- site. This elongation process then proceeds until a stop codon is encountered at the

A-site and the completed polypeptide chain exits the ribosome (Fig. 1.4). Actively translating mRNAs have many ribosomes bound which are referred to as polyribosomes (polysomes).

Figure 1.4: Mechanism of translation elongation. During the elongation process, an aminoacyl tRNA (Aa-tRNA) binds to the empty aminoacyl-site (A-site) of the ribosome. This process is controlled by both eEF1A-GTP and eEF1B. When the correct codon-anticodon interaction happens, GTP is hydrolysed and a peptide bond is formed between the aa-tRNA and the amino acid in the A-site. The ribosome then translocates along the mRNA via eEF2. The deacylated tRNA is displaced to the E- site and subsequently dissociates from the ribosome with the help of eEF3. eEF1A is released in its GDP-bound (inactive form) and recycled back to eEF1A-GTP bound (active form) by eEF1B. Adapted from (Grant, 2011) with modifications.

1.5.1.3 Translation termination

There are three stop codons in eukaryotic cells (UAA, UAG or UGA) which do not code for any amino acids. Thus, when the ribosome encounters a stop codon, the aa-tRNA by itself cannot be translocated into the A-site of the ribosome. This process is facilitated by two eukaryotic release factors, namely eRF1/eRF3, which are also known as Sup45/Sup35 in yeast. Once eRF1 recognizes either one of the three stop codons and binds to the mRNA, the ester bond between the completed peptide and the P-site of tRNAs is broken with the help of eRF3 (Rospert et al.,

2005; von der Haar and Tuite, 2007). Subsequently, the nascent peptide is released from the ribosome, marking the end of termination process (Fig. 1.5). After the termination phase, the ribosome is dissociated from eRF1 by eRF3-GTP and finally, both mRNA and the deacylated tRNA in the P-site are recycled in preparation for another round of translation.

1.5.2 Regulation of translation initiation during oxidative stress in S. cerevisiae

Translational regulation is important since it is often found that the cellular levels of proteins do not correlate with their corresponding mRNA levels (Dever and

Green, 2012). However, it is also known that oxidative stress causes a rapid and reversible inhibition of translation, and hence the relationships among these dynamic regulatory changes are not clear (Clemens, 2001; Proud, 2005). Translation initiation is the key rate-limiting step in eukaryotic translation and is highly conserved from yeast to mammalian cells, suggesting that it plays major roles in eukaryotic gene expression (Grant, 2011; Jackson et al., 2010) . Since it is both rapid and requires the least amount of energy to be regulated, the initiation phase is thought to be the main target of translational regulation. There are two main types of initiation

Figure 1.5: Mechanism of translation termination in eukaryotes. During translation process, ribosome continues to translate until it eventually encounters a stop codon on the A-site of the ribosome which does not code for any amino acid. The stop codon is recognized by eRF1(Sup45) which recruits the GTPase eRF3 (Sup35) and couples with ribosome to form translation termination complex. Sup45 activates the hydrolysis of the ester bond between the polypeptides and the tRNA on the P-site, releasing the now completed polypeptides with the help of Sup35.

regulation, which are known to be activated in the presence of H2O2: (1) global translational control resulting from changes in the phosphorylation state of eIF2; and

(2) the regulation of initiation factors binding to the mRNA cap by eIF4E. Additionally, oxidative stress is also known to regulate translation via mRNA-specific mechanisms, which has been particularly well-studied for the Gcn4 transcription factor in S. cerevisiae (Mascarenhas et al., 2008).

1.5.2.1 Regulation of TC by eIF2α

eIF2 is a heterotrimeric complex consisting of α, β and γ subunits and it is one of the eukaryotic initiation factors (eIFs) required during the translation initiation process (Fig. 1.6). The assembly of the TC (eIF2-tRNAMeti-GTP) which is bound by the 40S ribosomal subunit, is regulated via eIF2B. During translation initiation, the

GTP bound to eIF2 is hydrolysed to GDP. eIF2 must exchange GDP for GTP to regain affinity for initiator-met-tRNA and to regenerate TC for subsequent rounds of initiation, and this requires the guanine nucleotide exchange factor (GEF) eIF2B to catalyse the swap. However, phosphorylation of eIF2-α at serine 51 converts it to an inhibitor of eIF2B (Pavitt et al., 1998). In yeast for example, cellular stresses including amino acid starvation and oxidative stress activate Gcn2, the sole kinase that phosphorylates eIF2 (Harding et al., 2000; Shenton et al., 2006). Since eIF2 is present in excess over eIF2B, even small changes in the phosphorylation of eIF2 may significantly affect the formation of TC. Thus, eIF2 phosphorylation reduces translation of most mRNAs by preventing the recognition of 5’-cap structure and reducing global translation initiation rates (Hinnebusch, 2000).

Figure 1.6: Regulation of translation initiation by inhibiting TC formation. There are four mammalian kinases currently known: (1) GCN2 is activated by oxidative stress and nutrient availability; (2) PKR is activated by viral RNA infection, HRI is activated by the haem limitation, and PERK is activated by ER stress or during hypoxia. During normal condition, eIF2 which is composed of and  complex exists in its GDP-bound form. To begin the translation initiation process, eIF2 needs to be in its GTP-bound form and this process is catalysed by eIF2B. Upon stress conditions however, GCN2 phosphorylates eIF2- at serine 51 resulting in a complex that inhibits eIF2B activity. The resulting decrease in a guanidine nucleotide exchange factor activity by eIF2B decreases formation of eIF2-GTP, TC complex and subsequently translation initiation rates. 42

1.5.2.2 Regulation of mRNA-specific translational control via Gcn4

As the name implies, mRNA-specific translational control regulates the expression of particular mRNAs during stressful conditions such as oxidative stress.

For example, upstream open reading frames (uORFs) are short open reading frames found in the 5'-untranslated region of approximately 13% of yeast ORFs (Lawless et al., 2009). They typically act as translational repressors, but in some cases like Gcn4 can be activated in response to limited TC levels (Hinnebusch, 2005). GCN4 is a basic leucine zipper domain (bZIP) transcription factor and is essential for a global transcriptional response induced by phosphorylation of eIF2α when S. cerevisiae is challenged with amino acid starvation (Hinnebusch, 2005) or upon oxidative stress

(Mascarenhas et al., 2008). However, transcriptional profiling showed that Gcn4 is required for a more limited set of genes (about ~10%) in response to hydrogen peroxide as opposed to amino acid starvation, indicating that there are other limiting factors which may specifically regulate Gcn4 output during oxidative stress conditions (Lawless et al., 2009; Mascarenhas et al., 2008).

1.6 Cytoplasmic mRNA degradation in S. cerevisiae

The process of mRNA decay is often recognized as a major contributor to the regulation of gene expression as well as maintaining regulatory responses crucial for cellular homeostasis. In addition, mRNA turnover plays a critical role in assessing the accuracy of mRNA biogenesis and degradation of aberrant transcripts. These aberrant transcripts are recognized and targeted for rapid degradation by mRNA surveillance mechanisms. Such processes block the expression of defective mRNAs, which can have deleterious consequences for the cell (Fasken and Corbett,

2005; Hilleren and Parker, 1999; Maquat and Carmichael, 2001). Consistently,

misregulation of mRNA stability has been implicated in disease states, including immune and inflammatory diseases, AD and cancer (Kiledjian and Maquat, 2008).

Several distinct mechanisms by which eukaryotic mRNAs can be degraded have been defined. These mechanisms include general mRNA decay and aberrant quality control pathways such NMD, NSD and NGD occurring in the cytoplasm. However, the relative contribution of each pathway may vary depending on the organism, either in general or for specific transcripts and this shall be explored further in the following sections.

1.6.1 Normal mRNA degradation

The mRNA has several important features including the 5’-cap and the 3’- poly(A) tail that serve as protective mechanisms from ribonucleases as well as controlling translation and turnover processes (Fig. 1.7). Like all eukaryotes, general cytoplasmic mRNA degradation is first initiated by a rate-limiting process known as deadenylation or shortening of the 3’-poly(A) tail. The poly(A) tail is protected from ribonuclease attack at the 3’-end by poly(A) binding protein (Pab1) which must be removed to trigger deadenylation by recruiting several deadenylases including the

Pan2-Pan3 complex which trims most of the poly(A) tail (from ~70 nt to ~25 nt) and the Ccr4-Pop2-Not complex which removes the rest of the poly (A) residues until ~10 nt, respectively (Parker, 2012). Following deadenylation, the removal of the 5’-cap structure is triggered with the help of Dcp1-Dcp2 decapping enzymes. These deadenylated and uncapped mRNAs are then subjected to rapid degradation either from a 5’-to-3’ direction or from 3’-to-5’ direction which employs the exonuclease

Xrn1 or cytoplasmic exosome, respectively.

Figure 1.7: Normal mRNA decay mechanisms in eukaryotes. The mRNA body is protected by a 5’-cap structure and 3’-poly(A) tail from ribonuclease attack. The mRNA degradation pathway is first initiated by removal of the poly(A) tail (deadenylation) from the 3’-end. Deadenylated mRNA is then decapped by decapping enzymes before rapidly being degraded by 5′-to-3′ decay pathway via Xrn1, or alternatively, by a 3′-to-5′ decay mechanism which involves the cytoplasmic exosome, a conserved multiprotein complex.

The 3’-to-5’ degradation pathway involves the cytoplasmic exosome along with several cofactors including the Ski complex. The exosome is a large protein complex with exoribonuclease activity that contains ten essential core subunits in yeast (Allmang et al., 1999; Mitchell et al., 1997). Specifically, the nine subunits of this exosome form a barrel-like structure with a central cavity of tightly packed six- ring subunit. The tenth subunit that makes up exosome, Rrp44 (also known as Dis3), is the only detectable active exonuclease in the complex (Dziembowski et al., 2007).

Defects in any of these core subunits generally lead to similar phenotypes in exosome-dependent events, indicating that these core subunits function as a whole and not as individuals (Allmang et al., 1999; Anderson and Parker, 1998; van Hoof et al., 2000). The exosome complex then associates with the Ski complex composed of

Ski2, Ski3, Ski8 by interacting with Ski7 to facilitate the 3’-to-5’ mRNA degradation

(Araki et al., 2001).

Several studies have suggested that the 5’-to-3’ degradation pathway is the major mechanism of mRNA turnover due to several important reasons, including: 1) the 3’-to-5’ decay pathway does not affect mRNA half-lives in a WT strain (Anderson and Parker, 1998), 2) the 5’-to-3’ pathway degrades mRNAs more rapidly than the

3’-to-5’ degradation pathway since it is found that an uncapped mRNA is an excellent substrate for Xrn1 (Anderson and Parker, 1998; Muhlrad et al., 1995).

However, other studies using mutants defective in either 5’-to-3’ degradation or 3’-to-

5’ decay have found minimal effects on global mRNA decay rates/mRNA abundance, suggesting that each pathway may compensate for the other when necessary (He et al., 2003; Houalla et al., 2006). Environmental factors may also influence the requirement for these decay pathways. During amino acid starvation for instance, 3’-to-5’ decay is more active than 5’-to-3’ degradation in S. cerevisiae

(Benard, 2004). Nevertheless, simultaneous inactivation of both decay pathways result in synthetic lethality while conditional double mutants have been shown to stabilize mRNAs (Anderson and Parker, 1998; Soto and Estrada, 2008), further highlighting the importance of both pathways in mediating normal mRNA degradation in yeast. Ultimately, it is difficult to determine which pathway is deemed more important than the other as the pathways between the 5’-to-3’ and 3’-to-5’ degradation routes may depend on many factors including environmental conditions, nature of the transcript, or organisms that are involved.

1.6.2 Specialized mRNA quality control mechanisms

It is common for cells to make mistakes as many types of errors may arise during each step of gene expression leading to production of aberrant transcripts.

Indeed, there are many specialized control pathways that have evolved to protect cells from the potentially toxic effects of accumulated defective transcripts during translation. This current study focuses on three cytoplasmic mRNA quality control mechanisms- NMD, NSD, and NGD that may be activated to degrade certain classes of aberrant mRNAs so that only correctly processed mRNAs are released.

All of these specialized pathways mostly use the same decay enzymes responsible for degrading normal transcripts; however, they differ in terms of their mode in recognizing and targeting defective mRNAs for degradation which will be described in the following sections.

1.6.2.1 Nonsense-mediated decay (NMD)

NMD was originally thought to only degrade mRNAs containing premature termination codons (PTC) (Losson and Lacroute, 1979). Since then, NMD has been

found to target a variety of aberrant mRNAs including: 1) mRNAs containing nonsense or frameshift mutations; 2) unspliced introns containing stop codons; 3) abnormally long 3’-UTRs; 4) WT transcripts containing short uORFs in their 5’-UTR prior to their true coding region; and 5) mRNAs which potentially promote out-of- frame translation initiation (Parker 2012). Interestingly, NMD is also activated to degrade a subset of “normal” mRNAs particularly those involved in cell-surface dynamics and chromosome structure (Parker, 2012). As shown in Fig. 1.8, a distinctive feature of the NMD pathway is that NMD substrates are rapidly decapped and degraded via Xrn1 from the 5’-end without prior deadenylation (Muhlrad et al.,

1994). The most direct evidence for this comes from studies using WT and nonsense

-containing transcripts that were stabilized upon XRN1 deletion (Muhlrad et al.,

1994). Moreover, both of these strains lacked a 5’-cap structure, indicating that it was processed by the Dcp1-Dcp2 complex (Hatfield et al., 1996). Taken together, these findings indicate that NMD utilizes normal mRNA turnover pathways once a specific transcript has been targeted for degradation. However, how these mRNAs are subjected to degradation by NMD remained poorly understood.

Several genes including UPF1/NAM7/MOF4, UPF2/NMD2/SUA1,

UPF3/SUA6, PRT1, HRP1, MOF2, MOF5, MOF8 and DBP2, have been identified which mediate NMD in S. cerevisiae (Parker, 2012). However, research has mainly focused on the three upstream frameshifting (Upf) proteins for a few important reasons: 1) the Upf proteins are regarded as the core components of the NMD pathway since they are highly conserved throughout eukaryotes (however, the key mechanistic details are different among different organisms); 2) disruption of any of these UPF genes result in stabilization of nonsense-containing transcripts (Parker,

2012). It is noteworthy to point out that all of the Upf proteins are shown to interact

Figure 1.8: Recognition of NMD substrates. During the normal termination process, eRF3 (Sup35) binds with eRF1 (Sup45) to form a termination complex within the ribosomal A-site, provided that eRF3 must be in close proximity with Pab1, a highly conserved protein that stabilizes poly(A) tail. NMD is activated when eRF3 is not in close proximity of Pab1, thus recruiting Upf1 to directly interact with mRNAs containing premature termination codons (PTC) and subsequent degradation via the 5’-to-3’ pathway without prior deadenylation. Both Upf2 and Upf3 are shown to mediate Upf1’s activity.

with both Sup45 and Sup35 translation termination factors during NMD and therefore, may provide a direct link between a PTC recognition event and mRNA decay (Parker, 2012).

Upf1 is an ATP-dependent RNA superfamily 1 (SF1) helicase that acts as the main regulator of the NMD pathway (Parker, 2012). The N-terminal domain of Upf1 is rich with cysteine and histidine (CH) residues, while its C-terminal domain consists of a helicase region (Chakrabarti et al., 2011; Clerici et al., 2009). The second core

NMD factor, Upf2, is an acidic protein that acts as a scaffold between Upf1 and Upf3

(Chakrabarti et al., 2011). Upf3 is the least conserved Upf proteins that serves as a shuttling protein, but predominantly exists in the cytoplasm during steady state conditions (Shirley et al., 1998). Both Upf2 and Upf3 were found to mediate NMD by regulating Upf1’s function (Parker, 2012). An immunoprecipitation study, for example revealed that Upf1 physically interacts with both the eRF1 and eRF3 termination factors (Czaplinski et al., 1998) while in vitro studies showed that both Upf2 and Upf3 interact with eRF3 only (Wang et al., 2001). During the NMD process, Upf1 is thought to form a complex with eRF1 and eRF3 translation termination factors bound at the ribosomal A-site to mediate peptide release (Wang et al., 2001). Once the peptide is released, the complex is disassembled so that Upf1 may form another complex with Upf2-Upf3 to recruit decapping enzymes to degrade these nonsense transcripts (Wang et al., 2001).

Several studies have found that NMD is intimately linked to the translation process. Studies using sucrose density gradients showed that all Upf proteins co- distributed with 80S ribosomes and polyribosomes as well as nonsense containing transcripts, indicating that these NMD factors depend on the translation process

(Parker, 2012). Furthermore, additional experiments revealed that nonsense

transcripts are stabilized in the presence of cycloheximide, a drug used to inhibit the translation process (Parker, 2012). Finally, a different study showed that when monitoring the sensitivity of yeast strains against the toxic arginine analogue, canavanine, the upf1 mutant showed the highest nonsense suppression of the CAN1 allele compared with both UPF2/UPF3, indicating that Upf1 is most critical in providing a direct link in regulating translation termination and mRNA degradation

(Maderazo et al., 2000).

1.6.2.2 Nonstop mRNA decay (NSD)

NSD is a specialized decay process which occurs when mRNA transcripts lack a stop codon or during conditions where ribosomes continue to translate through the 3’ UTR and poly(A) tail and eventually stall at the 3’ end of the mRNA

(Frischmeyer et al., 2002; van Hoof et al., 2002). Nonstop transcripts are typically caused by the presence of cryptic poly(A) addition sites that leads to premature polyadenylation of transcripts, through frameshift mutations in the DNA causing stop codon disruption or any event that may cause transcription processes to be aborted

(reviewed by (Klauer and van Hoof, 2012). Several factors including Ski7, the Ski complex (Ski2-Ski3-Ski8), and the cytoplasmic exosome have been identified as the core components of the NSD process (Klauer and van Hoof 2012). Additionally, both

Ski7 and the Ski complex were originally discovered to have a superkiller phenotype due to their increased production of K1 (Killer) toxin in viral RNAs (Klauer and van

Hoof 2012). Apart from its function in mRNA degradation, Ski8 has also been shown to participate during the meiotic recombination process by initiating the breakage of double-stranded DNA to allow genetic diversification (Arora et al., 2004; Neale and

Keeney, 2006).

The NSD process is thought to occur when the Ski7 C-terminus interacts with the empty A-site of the stalled ribosome and subsequently binds with the Ski complex while its N-terminal region recruits the exosome, resulting in exosome- mediated 3’-to-5’ degradation (van Hoof et al. 2002). Interestingly, several studies have reported that in the absence of Ski7, degradation of nonstop mRNAs occurs via accelerated decapping followed by 5’-to-3’ degradation via Xrn1 (Frischmeyer et al.,

2002; Inada and Aiba, 2005). This is perhaps due to the displacement of Pab1 during translation through the poly(A) tail resulting in the loss of the 5’-m7G cap structure (Inada and Aiba, 2005). Therefore, it is possible that both pathways function in parallel to degrade NSD substrates as briefly outlined in Fig. 1.9.

Studies on domain interactions between the Ski complex and Ski7 revealed a dynamic interaction between Ski2 and Ski8 which requires Ski3 to act as a scaffold

(Wang et al., 2005). Moreover, in-depth studies on the Ski complex revealed its existence as a heterotetramer of a RNA helicase protein Ski2, a tetratricopeptide

(TPR) repeat protein Ski3 and a WD repeat protein Ski8 with a ratio of 1:1:2, respectively (Klauer and van Hoof 2012). On the other hand, co-immunoprecipitation experiments had shown that the N-terminal domain of Ski7 (1-96 amino acids) interacts directly with only Ski3 while the rest of the Ski7 N-terminus (80-264 amino acids) mediates binding to the cytoplasmic exosome (Araki et al. 2001; Wang et al.

2005). Finally, structural studies revealed that the C-terminal domain of Ski7 closely resembles that of eEF1A and eRF3 of translational elongation and termination factors, respectively (Klauer and van Hoof 2012).

There are several important features which set Ski7 apart from the Ski complex. For instance, unlike the Ski complex which is conserved in all organisms examined thus far, Ski7 only exists in a small subset of yeasts including S.

Figure 1.9: Recognition of NSD substrates. In the activation of NSD process, ribosomes continue to translate through the poly (A) tail and stall due to the lack of in-frame termination codons. This causes Ski7 to be recruited which in turn, recruits the Ski complex as well as the exosome to degrade the nonstop transcript in a 3’-to- 5’ manner. In the absence of Ski7 however, loss of Pab1 allows for rapid decapping to degrade nonstop transcript in a 5’-to-3’ manner. Both pathways do not require deadenylation which sets them apart from general mRNA decay.

cerevisiae (Atkinson et al., 2008). Instead, a paralog of Ski7, Hbs1, one of the key components in NGD, is thought to perform Ski7’s function in other fungi/eukaryotes

(reviewed in (Hoshino, 2012). Moreover, while the NSD process specifically requires both the Ski complex and the whole Ski7 to be activated, several studies have demonstrated that the presence of the N-terminal domain of Ski7 alone is sufficient during exosome-mediated mRNA decay (Araki et al. 2001).

1.6.3.3 No-go mRNA decay (NGD)

In 2006, Doma and Parker found that two protein factors termed Dom34 and

Hbs1 are involved in the NGD pathway. Dom34 (also known as Pelota in mammals) is structurally similar to eRF1 while the GTPase Hbs1 closely resembles those of eRF3 and eEF1A, indicating their involvement during the translation process (Parker

2012). NGD substrates are thought to be mostly caused by inhibitory mRNA structures such as: 1) artificial, stable stem-loop structure, 2) pseudoknots, or 3) rare codons e.g. those containing GC rich sequences or damaged RNA bases (Doma and Parker, 2006; Gandhi et al., 2008). The current model of NGD suggests that once an elongating ribosome gets stalled due to the presence of strong, secondary structures, Dom34 binds to the empty A-site of the ribosome and recruits Hbs1 to form a stable ternary complex similar to the eRF1-eRF3 translation termination factors.

There are several important features which sets NGD apart from NMD and

NSD. First and foremost, NGD has been discovered to recognize stalled ribosomes during the translation elongation process and not during the termination event (Doma and Parker, 2006). Secondly, unlike eRF1, Dom34 does not have a GCQ motif required to releases the peptide from the P-site of tRNA; the stalled ribosome

Figure 1.10: Recognition of NGD substrates. The NGD complex recognizes a stalled ribosome that may arise from several mechanisms during translation elongation eg. inhibitory mRNA secondary structures such as stem loops, peptide- mediated targets, or poly (A) mediated targets. Regardless of where these stalled ribosomes came from, all are subjected to endonucleotytic digestion to generate secondary targets, which can be further degraded either by the exosome or an exonuclease. Following cleavage, these mRNA fragments can be further degraded either by exonuclease Xrn1 in 5’-dependent manner or by the cytoplasmic exosome from the 3’-end, respectively. 55

therefore remains bounded with tRNA upon NGD process which is not observed in both NMD and NSD processes (Lee et al., 2007; Shoemaker et al., 2010).

1.7 Protein folding, misfolding, and aggregation

Proteins are amino acid polymers that are synthesized during translation based on the genetic code. As the nascent proteins make their exit from ribosomes, they simultaneously undergo a series of rapid conformational changes mainly to prevent the exposure of hydrophobic residues and to acquire the most thermodynamically stable conformation known as the native state (Linding et al.,

2004; Routledge et al., 2009). Depending on its size, the folding process usually takes few seconds at most so that proteins can become biologically active and functional (Eaton et al., 2000). In cells, protein folding usually takes place within the cytoplasm or in the secretory pathway and is usually assisted with a number of molecular chaperones to ensure that the proteins are folded properly (Dobson,

2003).

During certain conditions that may influence protein stability however, hydrophobic residues may become exposed resulting in inter-molecular contacts with other proteins, generating larger, often insoluble, non-native conformations

(Chiti and Dobson, 2006). Despite being thermodynamically less stable than the native state, these misfolded/partially folded states are thought to favour the aggregation state due to several reasons including: 1) during heat or oxidative stress, or 2) during folding process where they can be kinetically trapped, or 3) during alterations of optimal conditions such as pH, or 4) presence of solutes such as salt that may disrupt proper folding process (Sherman and Goldberg, 2001). This induces the association of misfolded oligomers to become aggregation prone

proteins which can be categorized as either amorphous aggregates or highly ordered amyloid fibrils.

1.7.1 Amyloid aggregates

The best characterized form of protein aggregates are highly structured amyloid fibrils. They mostly consist of repeated β-strands that run perpendicular to the fiber axis assembling into stacked cross-β sheets (Tycko, 2006). Amyloids are generally assembled into filamentous aggregates with diameters ranging from 2-20 nm with relatively short lengths ranging from hundreds of nm to few m as seen either under atomic force microscopy (ATM) or transmission electron microscopy

(TEM) (Caughey and Lansbury, 2003).These very stable amyloid “cores” are highly solvent protected and also protease resistant (Caughey and Lansbury, 2003).

Amyloids can also bind specifically to histological dyes such as congo red and thioflavin T so that the amyloid fibrils can either appear as red or have fluorescence enhancement, respectively (Biancalana and Koide, 2010).

1.7.2 Amorphous aggregates

Unlike amyloid fibrils, the mechanism of amorphous aggregation is less understood. Amorphous aggregates are often present in a granular form consisting of random assembly of monomers, or through intermediates like oligomers or protofibrils (Yoshimura et al., 2012). They are often deemed as a transition step towards amyloid formation. For instance, amyloid plaques in bacterial inclusion bodies are found to be a mixture of amyloid fibrils and amorphous aggregates

(Yoshimura et al., 2012). To date, various amorphous aggregates have been identified to occur either in vitro or in vivo with different structural properties and

toxicity levels but hard to classify due to high structural flexibility and heterogeneity

(Yoshimura et al., 2012).

In vitro aggregates are frequently detected during protein purification procedures e.g. following aluminium sulphate precipitation when coupled with protein denaturation steps. They are also observed when the native proteins are either salted-out, isoelectrically precipitated, self-associate, or are over-produced (although it is still diluted compared to in vivo conditions) (Wang et al., 2011). Meanwhile, in vivo amorphous aggregates are usually observed upon obstruction of PQC machineries such as molecular chaperones and proteases as well as loss of essential proteins leading to accumulation of unwanted or toxic polypeptides

(Hingorani and Gierasch, 2014). It is noteworthy to point out that there are major differences between in vitro and in vivo aggregates due to the complexity of protein folding in vivo. For instance, the unfolded state of in vitro samples are often observed in a controlled, thermodynamically stable environment while it is difficult to estimate the time that it takes for nascent polypeptides to unfold within cells (Hingorani and

Gierasch, 2014). Moreover, the crowded environment within cells is hard to mimic in vitro as many protein folding experiments done outside the cells were diluted

(Hingorani and Gierasch, 2014). Therefore, although recent advancements have been observed in protein folding studies due to in vitro experiments, there are still many challenges in order to fully understand the process of protein folding and consequently protein aggregation in vivo.

1.7.3 Protein aggregation related diseases

As described previously, protein aggregation is caused by misfolding of proteins resulting in insoluble deposits of either amorphous or highly ordered amyloid aggregates. These aggregates generally result in either loss-of-function or gain-of- function attributing to cellular toxicity. Several loss-of-function diseases are due to the degradation of the misfolded proteins by the proteasome such as in cystic fibrosis and α1-antitrypsin (α-AT) deficiency in humans (Soto and Estrada, 2008). On the other hand, gain-of-function diseases are usually more devastating and can sometimes be caused by: 1) the aggregation of misfolded proteins associated with conformational changes or, 2) the deposition of the aggregates outside the cell (Soto and Estrada, 2008). These include neurodegenerative diseases (AD, PD and HD), type II diabetes, prion-related diseases (bovine spongiform encephalopathy (BSE) and Creutzfeldt-Jakob disease (CJD)), cataracts, and amyloidosis (Soto and

Estrada, 2008).

However, aggregation is not only associated with pathological conditions but can be beneficial as well as several organisms have been shown to exploit aggregation for function. In E.coli, structural amyloids such as curli can mediate host protein binding and play a key role during biofilm formation (Chapman et al., 2002).

Chaplins, one of the structural amyloids found in gram-positive bacteria, play a critical role in aerial hyphae growth for spore formation (Elliot et al., 2003). Chorion proteins are another type of structural amyloid found in fish and insect eggshells to stabilize the eggshell’s structure and protect the larvae against harmful situations

(Fowler et al., 2007). Amyloid fibrils are also discovered in human Pme17 that is important for melanin production in humans (Berson et al., 2003). The yeast prions

[PSI+] and [URE3] for example, have been identified as functional amyloids that

facilitate phenotypic variation particularly during stressful conditions (True et al.,

2004; True and Lindquist, 2000; Tyedmers et al., 2008). Interestingly however, the mammalian prions are often linked to neurodegenerative diseases and therefore the difference between mammalian and yeast prions shall be highlighted in the following sections.

1.8 Prions

Known as proteinaceous infectious particles, prions were originally thought to represent an epigenetic switch solely determined by changes in protein structure; this heritable variation is not based on alterations of the genetic material given that it appeared to be resistant to radiation (Prusiner, 1982). Some lines of evidences supporting this hypothesis has come from in vitro experiments showing that injection of the infectious prion protein (PrPsc) form into WT mice causes the development of prions (Castilla et al., 2005; Deleault et al., 2007). Others had claimed to successfully generate prions using recombinant PrPsc alone (Wang et al., 2010).

However, many studies have refuted this ‘protein-only hypothesis’. For example, several studies have shown that they could only generate infectious prions from recombinant PrPsc with additional factors including phospholipid and RNA

(Legname et al., 2004; Makarava et al., 2010) thereby contradicting the previous result (Wang et al., 2010). In addition, the membrane lipid phosphatidylethanolamine

(PE) is also thought to be a very important cofactor for maintaining and facilitating the infectious state of prions, further suggesting that additional factors other than proteins may be required during prion formation (Deleault et al., 2012). Furthermore, mutation in the PRNP (prion protein) gene have been identified in every inherited prion disease which further supporting this hypothesis (Mastrianni, 2010).

Prions have been discovered in both mammals and yeast, though the protein determinants of mammalian and yeast prions are not the same. Unlike mammalian prions however, yeast prions are not toxic to the cell; instead they are suggested to be beneficial and play functional role within cells (True et al., 2004; True and

Lindquist, 2000; Tyedmers et al., 2008). However, mammalian and yeast prions do share several features including their amyloid fold structure, the mechanism of formation, and their ability to exist in several distinct conformations commonly known as prion strains for mammals or prion variants for yeasts, respectively (Table 1.1)

(Bradley et al., 2002; Stahl et al., 1993).The ability of prions to convert non-prion protein into the prion conformation highly suggests that the prion traits are dominant and are inherited in a non-Mendelian inheritance manner (Bradley et al., 2002; Stahl et al., 1993). Despite the differences observed between mammal and yeast prions, investigating yeast prions have proven to be beneficial in understanding mammalian prions due to their similar mechanism of replication as illustrated in Fig. 1.11.

1.8.1 Mammalian prions

The mammalian prion protein, PrP, is a glycosylphosphatidyl-inositol-linked cell surface protein that is expressed mainly in the central nervous system and can exist in two isoforms- the soluble PrPC state and the insoluble, aggregate PrPsc state

(Bolton et al. 1982). The soluble PrPC is mostly α-helical while the aggregated PrPsc is rich in β-sheets (Pan et al. 1993; Prusiner et al. 1983). The infectious PrPsc may act as a template to recruit soluble PrPC and convert it to the PrPsc form. Normally, this spontaneous conversion to PrPsc is prevented due to the high activation energy barrier involved in the conformational change; however, direct interaction with

Figure 1.11: Comparison of the mechanisms of mammalian prion and yeast [PSI+] prion formation. (a) In the mammalian prion, PrPC is likely to be converted to PrPsc by the presence of lipids and/or other co-factors thereby elongating the PrPsc fibril. These PrPsc aggregates get fragmented by an as yet unidentified fragmentation process. (b) Soluble Sup35 can spontaneously misfold into a partially folded intermediate which then forms [PSI+] depending on interaction with the actin cytoskeleton and/or cross-seeding by other yeast prions e.g. [PIN+]. Fragmentation of yeast prion aggregates is accomplished by the yeast protein chaperones Hsp104 and Hsp70/ Hsp40 into small oligomers (propagons) which then acts as a template and convert another soluble Sup35 to [PSI+]. Adapted from (Hofmann et al., 2012). 62

Table 1.1 Comparison of the main characteristics between mammalian and yeast [PSI+] prions Characteristics Mammalian prion Yeast [PSI+] prion Implicated protein Prion protein (Prp) Sup35 Cellular localization of GPI-anchored on plasma Cytosol native isoform membrane Predicted site of prion Plasma membrane and/or IPOD (insoluble protein formation/ propagation endocytic pathway deposit) Prion domain (PrD) Not well-defined (amino N-domain acid 90-231) Sequence similarity: -repeat region -Yes -Yes -Q/N rich -No -Yes Aggregate fragmentation None defined Chaperone Hsp104 in factors conjunction with Hsp70/Hsp40 Infectivity Yes (cell to cell, Yes intraspecies and interspecies (transmission barrier) Mitotic stability Yes (in cell culture) Yes Strains Yes Yes Protein-only evidence Not well-defined Yes preformed PrPsc can cause PrPC to undergo an induced conformational change and yield PrPsc instead (Colby and Prusiner 2011).

Although the exact mechanisms remained unclear, mammalian prions cause neurodegenerative diseases by aggregating extracellularly in the central nervous system to form amyloids that cause impairment of brain function (Bolton et al. 1982).

Some examples of these fatal prion diseases are grouped together as Transmissible

Spongiform Encelopathies (TSE) and this includes CJD, BSE, and kuru (Collinge,

2010; Hofmann et al., 2012). Among the amyloid-forming proteins, mammalian prions are considered as unique as they can be transmitted between organisms, and in some cases, between species as evidenced by the transmission of BSEs (or Mad

Cow disease) to humans in the late 1990’s (Collinge, 2010; Hofmann et al., 2012).

1.8.2 Yeast prions

Yeast prions share similar characteristics with mammalian prions e.g. they are insoluble and non-crystalline in their aggregated forms, but none of them have sequence homology with PrPC (Wickner et al. 2001). It has been suggested that yeast prions are only recognized as prions once they pass these requirements: 1) a prion should be reversibly cured by some growth condition or chemical treatment e.g. using guanine hydrochloride (GdnHcl), 2) overexpressing the prion protein should increase the frequency of the de novo formation of the prion and 3) the phenotype produced by a particular prion generally resembles that of the same protein when it is in the inactive form (Wickner et al. 2006).

The disaggregase chaperone Hsp104 is thought to play a key role by fragmenting large amyloids into small oligomers known as propagons so that the prion conformation is transmitted into new daughter cells (Fig. 1.11). GdnHCl is often used to inhibit the ATPase activity of Hsp104 in most of the yeast prions (Ferreira et al., 2001; Jung and Masison, 2001; Tuite et al., 1981). In addition, other chaperones including Hsp70 and Hsp40 are also known to influence yeast prion propagation

(Glover and Lindquist, 1998; Sondheimer et al. 2001; Lopez et al. 2003, Aron et al.

2007). To date, there are at least ten prion proteins described in S. cerevisiae (Harbi and Harrison, 2014; Harbi et al., 2012). However, only two prions, [PSI+] and [PIN+],

which are well-characterized prions used in this current study will be described in the following sections.

1.8.2.1 [PSI+]

The best studied yeast prion is [PSI+], which is the prion form of Sup35 (Cox,

1965). Sup35 is a GTPase that interacts with Sup45 and together with ribosomes, forms the translation termination complex during translation (Stansfield et al. 1995;

Zhouravleva et al. 1995). When Sup35 is converted to [PSI+], translation termination efficiency is compromised by limiting its accessibility to Sup45 and ultimately, this leads to nonsense suppression due to stop codon readthrough (Paushkin et al.

1997). As described previously, most yeast prions including [PSI+] can be cured by inhibiting/eliminating the disaggregase Hsp104; uniquely, overexpression of Hsp104 has also been shown to cure [PSI+] (Chernoff et al. 1995; Derkatch et al. 1997;

Moriyama et al. 2000; Du et al. 2008; Patel et al. 2009).

Analogous to mammalian prions, some of the yeast prions including [PSI+] can exist in more than one conformation and exhibit distinct phenotypes/variants.

[PSI+] variants have been identified as either weak or strong [PSI+] depending on the levels of nonsense suppression. For instance, strong [PSI+] variants are associated with high levels of nonsense suppression and this correlates with a very low amount of soluble Sup35, and vice versa (Derkatch et al. 1996). Moreover, structural studies of [PSI+] variants in vitro have shown that strong [PSI+] variants tend to have a shorter amyloid core (~40 amino acids), are less thermostable, and have faster kinetics of aggregation when introduced into [psi-] cells (Tanaka et al. 2008). As in the case of other prions, mating between strong and weak variants of [PSI+] results

in strong [PSI+] progeny, suggesting that strong [PSI+] may propagate more efficiently than weak [PSI+] strains (Bradley et al. 2002).

Sup35 has three major domains- N (amino), M (middle), and C (carboxyl)- terminal domains. Known as the prion forming domain (PFD), the N-terminal domain contains five and a half imperfect oligopeptide repeats and a glutamine and asparagine (NQ)-rich region and is required for the induction and propagation of

[PSI+] (Ter-avanesyan et al. 1993; Paushkin et al. 1997). The M-domain (124-253 amino acids) provides structural stability for the N-domain and also contains Hsp104 binding sites (Liu et al. 2002). The essential C-terminal domain (254-685 amino acids) is mostly in -helical form and is solely responsible for translation termination activity (Ter-avanesyan et al. 1993; Zhouravleva et al. 1995).

Several studies have investigated how [PSI+] prion formation occurs de novo and found that: 1) A [PIN+] prion is required for the induction of [PSI+] but not for its propagation (Derkatch et al. 2001); 2) environmental factors such as heat shock and oxidative stress have been shown to trigger the de novo appearance of [PSI+]

(Tyedmers et al. 2008); 3) the actin cytoskeleton is required during localization to insoluble protein deposits (IPOD), an amyloid deposition site where prion protein is believed to undergo fragmentation and seeding process, hence triggering de novo

[PSI+] formation (Spokoini et al., 2012; Treusch and Lindquist, 2012); 4) overexpression of Hsp104 chaperone will also increase the frequency of de novo

[PSI+] (Chernova et al., 2014); and finally, 5) the de novo appearance of [PSI+] is increased when either PFD of Sup35 is being overexpressed or when Sup35 is being overexpressed due to the higher number of protein molecules to convert soluble Sup35 to misfold and form [PSI+] (Chernoff et al. 1993; Ter-avanesyan et al.

1993; Derkatch et al. 1996; Derkatch et al. 2001).

[PSI+] formation can be detected based on the suppression of a premature termination codon (UGA) in the ADE1 gene (Chernoff et al., 1995; Serio and

Lindquist, 1999). The soluble Sup35 [psi-] cells become adenine-auxotrophic and therefore cannot grow on media lacking adenine. On rich media, [psi-] cells appear red in color as a result of an intermediate product of adenine biosynthesis. In contrast, the [PSI+] cells are capable of growing on media lacking adenine due to stop codon readthrough and appear white in color on rich media.

1.8.2.2 [PIN+]

The [PIN+]/[RNQ+] prion is the aggregated form of the Rnq1 protein, whose functional role in cells is yet to be determined. Also known as the [PSI+] Inducible element, [PIN+] was initially identified as a prion due to its unmistakable requirement in the de novo appearance of [PSI+] (Sondheimer and Lindquist, 2000). To date, the presence of [PIN+] has been shown to facilitate the de novo appearance of other yeast prions, including [URE3] (Liebman and Chernoff, 2012), [SWI+] (Du et al.

2008), and [OCT+] (Patel et al. 2009). As Rnq1 has no known function, much evidence on [PIN+] existence comes from many indirect studies instead. For instance, the PFD of Sup35 can be successfully replaced with the PFD of Rnq1 in transgenic yeasts to indirectly monitor [PIN+] via the [PSI+] phenotype (Derkatch et al. 2008). The most definitive proof comes from the in vitro study of recombinant

Rnq1 which has been shown to form amyloid fibers and that these fibers were capable of inducing [PIN+] formation when transformed into [pin-] cells (Sondheimer and Lindquist, 2000; Patel et al. 2007; Wickner et al. 2008).

1.9 Thesis objectives

The overall aim of this thesis was to examine the requirement for mRNA surveillance pathways including NMD, NSD and NGD during oxidative stress conditions. Oxidative stress is known to elicit complex translational reprogramming and we tested the hypothesis that mRNA surveillance pathways act as quality control systems protecting gene expression during stress. A second objective was to determine whether loss of NSD, NMD or NGD, would have any effects on proteostasis by examining protein aggregation in strains defective in mRNA surveillance pathways. The final aim was to examine the overlap between mRNA surveillance pathways and [PSI+], which is the prion form of the Sup35 translation termination factor. Therefore, this study was undertaken to evaluate whether mutants deficient in mRNA surveillance pathways show increased frequencies of [PSI+] prion formation. Additionally, [PSI+] prion-mediated alterations in gene expression was identified to influence survival during various stress conditions such as during oxidative stress and osmotic shock.

2.0 Materials and Methods

2.1 Yeast strains and plasmids

2.1.1 Yeast strains

The [PIN+][psi-] and [PIN+][PSI+] versions of yeast strain 74D-694 (MATa ade1-14 trp1-289 his3-200 ura3-52 leu2-3,112) were used throughout this study.

[pin-][psi-] versions were made by curing [PIN+][PSI+] strains on YPD agar plates containing 4 mM guanidine hydrochloride (GdnHCl) for at least five consecutive rounds. Mutant strains containing gene deletions were constructed using HIS3 and/or LEU2 selection cassettes as listed in Table 2.1. Additionally, ski2, ski3, ski7, and ski8 mutants were used in the BY4741 (MATa his3-1 leu2-0 met15-0 ura3-0) strain background for comparison.

Table 2.1 Yeast strains used in this study Genotype Source 74D-694 [PIN+][PSI+] MATa ade1-14 trp1-289 his3-200 Lab strain ura3-52 leu2-3,112 74D-694 [PIN+][psi-] MATa ade1-14 trp1-289 his3-200 Lab strain ura3-52 leu2-3,112 74D-694 [PIN+][psi-] ski2::HIS3 This study 74D-694 [PIN+][psi-] ski3::HIS3 This study 74D-694 [PIN+][psi-] ski7::HIS3 This study 74D-694 [PIN+][psi-] ski8::HIS3 This study 74D-694 [PIN+][psi-] upf1::HIS3 This study 74D-694 [PIN+][psi-] upf2::HIS3 This study 74D-694 [PIN+][psi-] dom34::HIS3 This study 74D-694 [PIN+][psi-] hbs1::HIS3 This study 74D-694 [PIN+][psi-] ski7::HIS3, dom34::LEU2 This study 74D-694 [PIN+][psi-] ski7::HIS3, hbs1::LEU2 This study 74D-694 [PIN+][PSI+] ski7::HIS3 This study

74D-694 [PIN+][PSI+] ski8::HIS3 This study 74D-694 [PIN+][PSI+] upf1::HIS3 This study 74D-694 [PIN+][PSI+] upf2::HIS3 This study 74D-694 [PIN+][PSI+] dom34::HIS3 This study 74D-694 [PIN+][PSI+] hbs1::HIS3 This study BY4741 MATa his3-1 leu2-0 met15-0 ura3-0 Lab strain BY4741 ski2::KanMX Euroscarf BY4741 ski3::KanMX Euroscarf BY4741 ski7::KanMX Euroscarf BY4741 ski8::KanMX Euroscarf BY4741 xrn1::KanMX Euroscarf BY4741 SUP35-GFP Life Technologies

2.1.2 Plasmids

Table 2.2 Plasmids used in this study Description Source pRS413 Lab strain pRS415 Lab strain pRS416 Lab strain pRS425 (multi-copy of LEU2) Lab strain pUKC815 Stansfield et al. 1995 pUKC819 Stansfield et al. 1995 CUP1 Sup35NM-GFP::URA3 Patino et al. 1995 Hsp104-mRFP::URA3 Malinovska et al. 2012 GAL1 Sup35::LEU2 Josse et al. 2012 pAV184 (hereafter refer as nonstop protein A) Wilson et al. 2007 pAV185 (hereafter refer as stop protein A) Wilson et al. 2007 p469 (hereafter refer as nonstop luciferase) Keeling et al. 2004 p477 (UGA luciferase) Keeling et al. 2004

2.2 Strain construction and verification

2.2.1 Oligonucleotides

Oligonucleotides were synthesised by Sigma-Aldrich as listed in Table 2.3 and designated as: for = forward primer; rev = reverse primer; Cfor = forward verification primer; Crev = reverse verification primer; Ffor = Flag forward primer;

Frev = Flag reverse primer; Mfor = Middle forward primer; Orev = Outside reverse primer together with the flanking sequences as follows

a) Forward primer: ~50 nt from 5’ end of the start codon (AUG) of the ATG of

interest+ GTACGCTGCAGGTCGAC (homologous to HIS3/LEU2 cassette on

the pRS413/ pRS415 vector).

b) Reverse primer: reverse complement of ~50 nt after stop codon of the ATG of

interest + ATCGATGAATTCGAGCTCG (homologous to HIS3/LEU2 cassette

on the pRS413/ pRS415 vector).

Table 2.3 Primers and sequences used for strain construction and sequencing Gene Oligo Sequence (5’-to-3’ direction) SKI2 For CCGATATGAACAACCTAACTCACAAAATTTACTGTACTAAT ACTAATTTAT rev AGAGCTCATTTATTCTCAATGTGAGTCATGTTTATAAAAGT TAATGTTTTCG Cfor GATATCACGACGGACGAAG Crev GCAACGGGCGGACTATTGTAA SKI3 For GTCAAGAAAGACACTAAGAACACAGAAAAGAAACACGAA GAGCAGAGG rev AAATTTGGATTCGAGATAGTCAATCAAACCTTAATGTAACA TTAG Cfor GTTTGCTCAATAAACGATTCG Crev GCTGTATAGAGTAATCATGCCATTTGG 71

SKI7 For TCCGTTTGCATCTTTACATTGCCTTTGGCATATTTGGCAGT AGTA rev TCAACGCCTATGCCTACTTTATCCAACAATGCGGATGATTT CTCTGCAG Cfor TACTCGTTGTATACTTTGTTAGCATCA Crev TTAAACAAGGTAATAAAACAGATTGAAAAA SKI8 For GGGGTTGATTTTAACATCGTCCACCTTGATTCTTAACTTTT CACTCATTTTC rev CAAGTTCTCCACATATACATATGTAAGAATTGAGTCTATCT CTTTCTAATA Cfor ATCATTTCCGGATCATCATATCTTCT Crev AGGATCATCTTTCTAATCTATCATTTT UPF1 For GTTTTAACGCACACTAACAGAAGACTCTATTTCTCTTGTCA GCCAACA rev GTATGTGATAAAGGGGCATGGACTTGATATCCTAGCCTAC TAATCTCTT Cfor GGGAAATAAGAAAAACAAAAAGAAAATATA Crev CATAGTTCACACTTTTATCTCCTTAGTTTG UPF2 For CGTTTCTATGATCACTACGGGATATTATGATATTGTTAGGG GGTTATATTG rev CACGTAGCCATTTAATAGAGCCAAACATAAGTGGGCATAT ATAGTAA Cfor GAATGGTTAAGCTAAAATAGACTCTGTATG Crev ATTATTATATTGGCTAAAGATGGTGTA DOM34 For GCATTCGTTGCTGCATCGTTGTCATTTTGTTCAATTATCGC ATTCCTATCAT rev CCGCAATTAATTGAGTTGTTGTGAAATGGTTCTAGTTAGTC ACGTGCAGC Cfor TCTGGATTAGAATTAAGACTTGTATATCTTATTCTTTTTCCC TGTGG Crev GTCGATATACTACGTCAAGGTGTAACTAA HBS1 For GTGGTACCAGGCGTGAATGCTATATAAATATAATAACAATC GTGTA

rev ATAGTTTTTTAGATATATTCAGAGGCAGGTATTCTTTTGTAT TGTCGCCCT Cfor TGACAATCCTCAAAAGTATACACG Crev CATTATCTTCATCATTCTTATCATTTTC HIS3 rev CGACAACTGCGTACGGCCTG LEU2 rev CTTTGCACTTCTGGAACGGTG

2.2.2 DNA amplification by polymerase chain reaction (PCR)

PCR reactions were used to amplify plasmid and genomic DNA, and to verify deleted strains. A typical 50 µL PCR reaction mixture in deionised water contained

25 µL of MyTaq Red Mix (2x) (Bioline), 1 µg for each primer as wells as ~50 ng of diluted DNA/ plasmid. The PCR conditions used were generally: initial denaturation

(94°C for 5 min), followed by 3 steps of 30/ 35 cycles of (1) 94°C for 1 min

(denature), (2) 55- 60°C for 1 min (annealing), (3) 72°C for 3 min (extension), and a final extension (72°C for 5 min).

2.2.3 Yeast genomic DNA extraction

Genomic DNA was purified from candidate yeast colonies as described previously (Hoffman and Winston, 1987). Genomic DNA was used in PCR reactions to confirm correct gene deletions. Individual colonies were inoculated into 10 mL of

YPD media and grown overnight at 30°C (180 rpm) to reach an OD600 of 0.7-0.8.

Cells were centrifuged at 4000 rpm for 5 min, washed in 0.5 mL of deionised water and transferred to a 1.5 mL screw-top eppendorf tube. The cell pellet was resuspended in 0.3 mL of solution C (2% v/v Triton X-100, 1% w/v SDS, 100 mM

NaCl, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA) and kept on ice. 0.3 mL of phenol:chloroform:iso-amylalchohol (25:24:1) was added into the suspension,

followed by ~200 μL of glass beads. Cells were lysed for 2 x 45 sec using a mini- bead beater (Biospec Scientific, Bartlesville, OK) at 4°C. 0.2 mL of TE buffer (pH 8.0) was added and the mixture was vortexed before being centrifuged at 13,000 rpm for

5 min (4°C). The supernatant was collected and transferred to a new eppendorf tube containing 1 mL of 96% (v/v) ethanol and kept at -20°C for 30 min. The solution was centrifuged in a microfuge for 10 min and the supernatant was carefully removed, leaving the DNA on the walls of the tube. 0.4 mL of TE buffer and 3 µL of RNAse A

(10 mg/mL) was added and the mixture was incubated at 37°C for 15 min. Next, 10

µl of 3 mM sodium acetate (pH 5.2) was added to deactivate RNAse A, followed by 1 mL of 96% (v/v) ethanol and mixed by inverting the tubes several times. The mixture was centrifuged for 10 min and the cell pellet was air dried and washed with 70%

(v/v) ethanol, air dried and resuspended in 50 µL of deionised water.

2.3 DNA/RNA manipulation and analysis

2.3.1 Plasmid extraction

Plasmid DNA was extracted using the Qiagen mini-prep kitTM. Bacterial cells were harvested by centrifugation at 4,000 rpm for 5 min at 4°C. Cells were resuspended in buffer P1 (cold) and transferred into a 1.5 mL microcentrifuge tube.

250 µL of buffer P2 was added into the tube followed by 350 µL of buffer N3 and the tube was thoroughly mixed prior to centrifugation at 13,000 rpm (maximum speed) for 10 min. The supernatant was pipetted into QIAprep spin column and quickly centrifuged for 1 min at maximum speed. The flow-through was discarded and the column was washed twice with 0.5 mL buffer PB and 0.75 mL buffer PE, respectively. The flow-through was discarded and the column was additionally spun for 1 min at maximum speed to remove the remaining residual wash buffer. The

column was transferred to a new 1.5 mL eppendorf tube and 50 µL of buffer EB (10 mM Tris-Cl, pH 8.5) was pipetted to the centre of the spin column. The column was left to stand for 5 minutes before a final spin at maximum speed for 1 min. The eluted plasmid DNA was stored at -20°C for further analysis.

2.3.2 Yeast transformation

Plasmids and PCR products were transformed into S.cerevisiae using the lithium acetate transformation method (Gietz et al., 1995). Briefly, 5 mL of overnight cultures were grown in YPD media at 30°C to reach late exponential phase (OD600 of

0.7-0.8). Cells were centrifuged at 4000 rpm for 5 min, resuspended in lithium acetate/TE buffer (0.2 M lithium acetate; 10 mM Tris base; 1.0 mM EDTA (pH 7.5)) and transferred into a new eppendorf tube. Cells were centrifuged again at 10,000 rpm for 30 sec and resuspended in 500 µL of lithium acetate/TE buffer to permeabilize the yeast cell wall. 100 µL of cell suspension was then mixed with 8 µL of sheared salmon sperm DNA (10 mg/mL; Invitrogen) and 500 µL of 40% PEG 4000

(in Lithium acetate/TE buffer) and incubated at 30°C for 45 min to help stabilizing yeast cells and thereby increasing their recovery. The mixture was immediately heat shocked for 15 min at 42°C which creates pores on the plasma membrane to allow the plasmid DNA to easily enter into the yeast cell wall. Cells were immediately centrifuged and resuspended in 100 µL deionised water and immediately plated on

SCD media lacking histidine or leucine for plasmid selection, and incubated at 30°C for 5-7 days to obtain colonies.

2.3.3 Agarose gel electrophoresis

DNA samples were separated according to size using 1% (w/v) agarose gels.

Gels were prepared and electrophoresed in 1 x TAE buffer (40 mM Tris base, 20 mM acetic acid, 1 mM EDTA, pH 8.0) containing 8 µL of safeview nucleic acid stain (NBS

Biologicals). Molecular weight markers (Hyperladder 1kb, Bioline) were used to estimate the size of DNA between 200 bp and 10 kbp. Electrophoresis was performed at 110 volts for 50 min, and DNA bands were visualised using UV illumination (254 nm) and photographed using a gel documentation system.

2.3.4 Quantitative Reverse Transcriptase PCR (qRT-PCR)

For all RNA work it was essential to maintain an RNAse free environment at all times during the procedure by spraying RNAse Zap™ to work areas and utensils.

All of the reagents as well as the extraction procedure was performed at cold temperature. New, unopened filtered pipette tips, eppendorf tubes and fresh gloves were used while handling the samples. PCR primers were designed using standard

PCR parameters and purchased from Sigma as listed in Table 2.4 below:

Table 2.4 Primers used for qRT-PCR Protein Oligo Sequences (5’-to-3’ direction) Protein A For CCCAAGCCAAAGCGCTAACC

rev ACCTGGCAATTCCTTACCGGAT

Sup35 For TGGGTACCAACAAGGTGGCT

rev ACCGCCTTGAGGATTGTACTGT

Actin For AGAGATTTGACTGACTACTTGATG

rev GAAGATTGAGCAGCGGTTT

2.3.4.1 RNA extraction

Cell cultures were centrifuged at 4000 rpm for 3 min before resuspension in 1 mL of Trizol® (Ambion). Samples were quickly transferred to a pre-chilled eppendorf tube containing 200 μm acid-washed beads (Sigma) and the tubes were vortexed for

6 x 20 sec each with at least 40 sec intervals on ice. For a separation phase procedure, 750 μL of homogenised cells were transferred to pre-chilled eppendorf tube containing 150 μL of chloroform (Fisher) and vortexed for 20 sec prior to a final centrifugation at 13,000 rpm for 15 min. In the final step, only the upper layer of the supernatant was transferred to new, pre-chilled eppendorf tubes containing 1 μL of glycoBlue™ (Ambion) and 350μL of isopropanol (Fisher) to precipitate the RNA.

Samples were quickly vortexed for 20 sec and stored at -80°C until further analysis.

2.3.4.2 RNA washes

Samples were thawed on ice and centrifuged at 13,000 rpm for 15 min. The supernatant was removed and the cell pellet should be gray/blue in colour. To this, 1 mL of 75% ethanol was added and RNA were spun at 10,000 rpm for 5 min, twice.

RNA pellets were then allowed to air-dry, prior to adding 20 μL of pre-warmed nuclease-free water (Gibco). Finally, the OD260 of the RNA extracts were quantified using a Nanodrop-8000 spectrophotometer and stored at -80 °C until further use.

2.3.4.3 q-RT PCR analysis

To assess the levels of Protein A and actin mRNA transcripts, a two-step qRT-PCR method was performed. Firstly, reverse transcription to generate the cDNAs was carried out using a ProtoScript® First Strand cDNA Synthesis Kit (New

England Biolabs) according to the manufacturer’s instructions. The q-PCR analysis was performed using IQ SYBR Green (Biorad) as per the manufacturer’s instructions from at least 3 independent experiments and expressed as fold difference of nonstop

Protein A-nonstop mRNA levels relative to actin.

2.3.5 Quantitation of DNA/ RNA concentrations using a spectrophotometer

The ratio of absorbance at 260 nm and 280 nm was used to assess the purity of DNA and RNA, respectively. DNA or RNA concentrations were quantified using a

NanoDrop 8000 spectrophotometer (Thermo Scientific), which reads the absorbance at 260 nm and converts it into ng/µL DNA. An absorbance of 1 unit at 260 nm corresponds to 50 ng/µL of DNA and 40 ng/µL of RNA, respectively. Pure DNA generally has an OD260/OD280 ratio of ~1.8; pure RNA has an OD260/OD280 ratio of

~2.0. The ratio of 260/230 was used to determine the purity of nucleic acid with an expected value of 2.0-2.2. Any ratio that was significantly different indicates potential phenol or protein contamination which absorb strongly at or near 230 and 280 nm, respectively.

2.4 Media and growth conditions

2.4.1 S. cerevisiae

Strains were grown at 30°C (180 rpm) in YPD media (2% w/v glucose, 2% w/v peptone, 1% w/v yeast extract) or minimal SCD medium (2% w/v glucose, 0.17% w/v yeast nitrogen base supplemented with Synthetic Complete (SC) Kaiser amino acid mixes (Formedium, England). SRaf and SGal media contained 2% raffinose or 1% raffinose/ galactose in place of glucose, respectively. Additionally, ¼ YPD media (4

% w/v glucose, 1 % w/v peptone, 0.25 % w/v yeast extract) was used to distinguish

the phenotypes of [psi-]/[PSI+] strains based on their red/white colour appearance, respectively. Solidification of media was achieved by adding 2 % (w/v) agar.

2.4.2 E. Coli

E. coli were grown at 37°C and 220 rpm in Luria-Bertani (LB) media (0.5%

(w/v) NaCl, 0.5% (w/v) yeast extract and 1% (w/v) tryptone). When required, LB media was solidified through the addition of 2% (w/v) agar. LB media was supplemented with 100 µg/ml ampicillin to select for plasmids conferring ampicillin resistance.

2.5 Yeast growth and stress analysis

2.5.1 Spot tests for oxidant-stress sensitivity.

Stress sensitivity was determined by growing cells to stationary phase in YPD media at 30°C/180 rpm and spotting cells onto SCD plates containing different concentrations of oxidants including hydrogen peroxide, diamide, copper (II) sulphate and cadmium sulphate all supplied by Sigma. Cultures were first diluted from 5 mL of starting culture so that the final OD600 is 1.0, and then serially diluted in a ratio of 1:1, 1:10, 1:100, and 1:1000 in deionised water. Next, 8 μL of each of the diluted suspensions were spotted onto freshly prepared SCD agar plates containing oxidants and grown for a minimum of 2 days at 30°C.

2.5.2 Oxidant-growth sensitivity

To analyse growth in the presence of oxidants, cultures were first grown overnight to early exponential phase (OD600~0.2) in SCD media at 30°C. Next, 30 mL aliquots of cultures were separated into individual flasks and treated in the

absence of presence of H2O2 (1 mM). Cultures were incubated at 30°C /180 rpm and growth monitored by measuring OD600 every hour over a period of six or eight hours.

2.5.3 Cell viability assay

Cells were grown to exponential phase (OD600 0.7-0.8) in SCD media at 30°C and exposed to various stress conditions including 2 mM H2O2, 6 mM diamide, 16%

(v/v) ethanol (Fisher), a temperature shift to 45°C and 3 mM NaCl. All stress conditions were imposed for one hour except for NaCl stress which was imposed for four hours. To monitor cell viability, 100 μL of cultures were serially diluted in deionised water and 100 μL of the diluted suspensions were spread on YPD plates and incubated for 3 days growth at 30°C. Cell colonies were counted and cell viability expressed as a percentage of the untreated cultures (100%).

2.6 Protein Analysis

2.6.1 Polysome analysis

2.6.1.1 Preparation of yeast cell extracts for polysome analysis

Polysomes are ribosomes that are actively translating mRNAs and can be separated using sucrose gradients. All of the reagents were prepared in autoclaved diethyl pyrocarbonate (DEPC)-treated water (200 μL of DEPC in 1L of deionised water, stirred overnight before being autoclaved to denature the DEPC) and were kept cold at all times throughout the procedure. Extracts were prepared in the presence of cycloheximide (Sigma) as described (Shenton et al., 2006). Briefly, 40 mL of overnight cultures were grown in SCD media to exponential phase (OD600 =

0.5-0.6) at 30°C/ 180 rpm prior to H2O2 treatment (0, 0.5, and 1.0 mM) for 1 hr. Cells were then transferred to pre-chilled tubes containing 400 μL of 10 mg/mL

cycloheximide (Sigma) to quickly stop protein synthesis and incubated for 30 min.

Cells were harvested by centrifugation at 4000 rpm for 5 min (4°C), washed in 20 mL of lysis buffer (20 mM HEPES at pH 7.4, 2 mM magnesium acetate, 100 mM potassium acetate, 0.5 mM dithiotreitol (DTT), 1mg/mL cycloheximide) and centrifuged again at 4000 rpm for 5 min (4°C). Cell pellets were resuspended in 750

μL lysis buffer and transferred to a pre-chilled microfuge tube. Cells were harvested at 10,000 rpm for 30 sec at 4°C and resuspended in 200 μL of lysis buffer.

Approximately 200 μL of acid-washed glass beads (Sigma) were added and the tubes were vortexed for 6 x 20 sec with at least 40 sec intervals on ice. Samples were centrifuged at 10,000 rpm for 5 min (4°C) and supernatants were transferred to pre-chilled eppendorf tubes for a final centrifugation at 10,000 rpm for 15 min (4°C).

In the final step, supernatants were collected and transferred to new, pre-chilled eppendorf tubes and the OD260 of the RNA extracts were quantified using a

Nanodrop-8000 spectrophotometer and stored at -80°C until further analysis.

2.6.1.2 Preparation of sucrose gradients

Sucrose gradients were prepared as previously described (Luthe, 1983).

Briefly, 15-50% sucrose solutions were individually prepared from a 60% sucrose solution (pure sucrose made with autoclaved DEPC-treated water) made in 10x polysome gradient buffer (100 mM tris acetate (pH 7.4), 700 mM ammonium acetate;

40 mM magnesium acetate). They were dispensed in L870 ultracentrifuge tubes

(Beckman Instruments) in liquid nitrogen starting with the 50% solution at the bottom.

Once the 50% sucrose step had frozen, the next layer was added consecutively with

15% sucrose as the uppermost layer. Gradients were then stored at -80°C and

allowed to thaw for at least 16 hours (4°C) prior to use. Each fraction was made in a total volume of 50 mL as described in Table 2.5 below:

Table 2.5 Preparation of pure sucrose gradients (15-50%) Sucrose concentration (%) 50 42 33 24 15 60% of filtered sucrose (mL) 41.6 35.0 27.3 20.0 12.5 10X polysome gradient buffer (mL) 5.0 5.0 5.0 5.0 5.0 Autoclaved DEPC-treated water (mL) 3.4 10.0 17.7 25.0 32.5

2.6.1.3 Sedimentation of polyribosomes

A volume equivalent to 2.5 OD260 of RNA was loaded onto sucrose gradients and spun in a SW41 rotor (Beckman Instruments) at 40,000 rpm (4°C). After 2.5 hr, gradients were collected and the absorbance was continuously measured at 254 nm to generate polysome profiles using a UV/Vis detector (ISCO UA-6) set to a chart speed of 60 cm/hr and peristaltic pump set to a speed of 20 (arbitrary units). As a chase solution, 60% (w/v) non-pure sucrose (non-pure sucrose dissolved in deionised water) was used. Finally, the analysis of ribosome distribution on sucrose density gradients was performed according to (Ashe et al., 2000). Image J software

(http://rsb.info.nih.gov/ij/) was used to calculate polysome to monosome (p/m) ratios by quantifying the areas under the first three polysome peaks divided by the cumulative area under the 80S monosome, 60S and 40S peaks. An example of a polysome trace is shown in Fig. 2.1 below:

Sedimentation of ribosomes in a 15-50%

pure sucrose gradients

254 A

Figure 2.1: An example of a normal, untreated polysome trace from actively translating yeast cells. Normal profiles exhibit several peaks corresponding to 40S and 60S ribosomal subunits, monosomes (80S ribosomes), and polysomes. Polysomes are ribosomes that are actively translating mRNAs. An inhibited trace will show redistribution from the area under the polysome peaks to the 80S monosome peak.

2.6.2 35S cysteine/methionine radiolabelling

The rate of protein synthesis was analysed as described previously (Shenton and Grant, 2003). 30 mL of exponentially growing cells (OD600 0.5-0.6) were grown in SCD media (without methionine) prior to hydrogen peroxide treatment (0, 0.5, 0.8 and 1.0 mM) for 1 hr. For the last 5 min of treatment, cells were pulse-labelled with

85 μM L-[35S] cysteine/methionine (MP Biomedicals) and kept on ice. 2 mL of prechilled 20% TCA was then added and samples were heated to 95°C for 10 min.

Samples were then allowed to cool for 20 min on ice and collected on Whatman

GF/C glass microfiber filters under vacuum. Filters were first rinsed with 2 mL of 20%

TCA followed by 2 mL of acetone to remove the acid. All filters were transferred to scintillation tubes, dried at 60°C for 30 min and finally suspended in 5 mL of scintillation fluid. Protein synthesis was calculated using a Packard tri-card 2100 TR

Liquid Scintillation Analyser programmed for [35S] expressed as counts per minute

(c.pm.). All experiments were repeated in triplicate.

2.6.3 β-galactosidase reporter assays to measure stop codon readthrough

2.6.3.1 Preparation of whole cell extracts

Cell cultures were harvested by centrifugation at 4000 rpm for 6 min. Cell pellets were resuspended in 1 mL of breaking buffer (100 mM Tris-HCl (pH 8), 1 mM

2-mercaptoethanol, 20% (w/w) glycerol) and transferred to a microfuge tube. Cells were centrifuged at 10,000 rpm for 1 min and resuspended in 250 μL of breaking with 12.5 μL of phenylmethylsulfonyl fluoride (PMSF, 40 mM in 100% isopropanol), a protease inhibitor. 200 μL of acid-washed glass beads were added and the cells were lysed for 40 sec using a mini-bead beater twice. Then, another 200 μL of breaking buffer was added and the mixture vortexed well. Samples were then centrifuged at 13,000 rpm for 15 min and the supernatant removed to a new eppendorf tube. Protein concentrations of the cell extracts were quantified using a

Nanodrop-8000 spectrophotometer at OD280 and stored at -20°C until further analysis.

2.6.3.2 Assay

The β-galactosidase assay employed was based on that of (Finkelstein and

Strausberg, 1983) with slight modifications. 200 μL of the clarified extract was added to 800 μL Z buffer (0.06 M Na2HPO4-7 H2O, 0.04 M NaH2PO4-H2O, 0.075% (w/w)

KCl, 0.025% (w/w) MgSO4, 0.27% of 2-mercaptoethanol, pH 7) in glass tubes (12x75 mm, VWR) and incubated at 28°C for 5 min to equilibrate the temperature. Since there is no stop codon in pUKC815 and thus the reaction is predicted to proceed at a much faster rate, the cell extract was diluted in a ratio of 1:10 before it was added to the Z buffer. The hydrolysis reaction was initiated by the addition of 200 μL onitrophenyl-β-D-galactoside (ONPG, 4 mg/mL in Z buffer) then incubated at 28°C until the samples developed a pale yellow colour. At this point, the reaction was terminated by the addition of 500 μL of 1M Na2CO3 which denatures β-galactosidase and approximately doubles the intensity of the yellow colour. The optical density was measured at 420 nm (Beckman DU-64 Spectrophotometer) as well as the reaction time was calculated based on the starting time when ONPG was added and the final time when Na2CO3 were recorded to calculate specific activity (SA) using the formula below:

Specific Activity = OD420 X total volume (mL)

(mole/min/mg protein) 0.0045 x protein (mg/mL) x extract (mL) x reaction time

2.6.4 Dual-luciferase reporter assays to measure stop codon readthrough

2.6.4.1 Preparation of whole cell extracts

Cell cultures were harvested by centrifugation at 4000 rpm for 6 min. Cell pellets were resuspended in 1 mL of prechilled 1X PBS buffer containing 1X

Complete Mini protease cocktail (50 mL) and transferred to microfuge tubes. Cells were centrifuged at 13,000 rpm for 1 min and resuspended in 200 μL of prechilled 1X

PBS buffer. 200 μL of acid-washed glass beads were added and the cells were lysed for 40 sec using a mini-bead beater twice. Samples were then centrifuged at 13,000 rpm for 15 min and the supernatant removed to a new eppendorf tube. Protein

concentrations of the cell extracts were quantified using a Nanodrop-8000 spectrophotometer at OD280.

2.6.4.2 Assay

The luciferase assay used a Dual-Glo® luciferase Assay System (Promega).

Both Luciferase Assay Reagent (LARII) and Stop and Glo Reagent (S/GR) were prepared beforehand according to the manufacturer’s instructions. 10 μg of lysate was first made up to 20 μL in 1X PBS buffer and kept at room temperature for 5 min.

Then, 50 μL of LARII reagent was added to measure Relative Luminescence Units

(RLUs) produced by firefly luciferase activity. The mixture was quickly vortexed and quantified using a Glomax 20/20 luminometer. Next, 50 μL of S/GR reagent was quickly added to quench the firefly activity and activate the Renilla luciferase activity.

Again, the mixture was quickly vortexed and quantified using the luminometer. The total elapsed time for both readings should be kept maximal at 30 sec. Finally, the ratio of luminescence of the firefly to the Renilla control was calculated. Negative controls that contained all the reaction components except cell lysates were used to determine the background for each luciferase reaction and were subtracted from the experimental values obtained.

2.6.5 Western blot analysis

2.6.5.1 Preparation of yeast whole cell extracts

Cells equivalent to 10 OD600 units were collected by centrifuging at 4000 rpm at 4°C for 5 min. Cell pellets were resuspended in 200 μL of prechilled 10% trichloro- acetic acid (TCA) and acid-washed glass beads were added prior to bead beating in a mini bead beater twice (40 sec each, medium speed). Another 200 μL of chilled

10% (TCA) solution was added and samples were then centrifuged at 13,000 rpm,

4°C for 10 min. The supernatant was then transferred to a new microcentrifuge tube, centrifuged at 13,000 rpm, 4°C for 5 min and the resulting pellet washed twice with

400 μL of acetone. Samples were then heated at 37°C for at least 5 min to evaporate the acetone prior to incubating with 150 μL of solubilisation buffer (100 mM Tris (pH 8), 1% SDS, 1 mM EDTA, 1 mM PMSF and 1X Complete Mini protease cocktail [Roche]) at 37°C for 1 hr. Finally, proteins were denatured by heating up at

95°C for 5 min in 4X Bolt LDS loading buffer (62.5 mM Tris-HCl (pH 6.8), 10% glycerol, 1% LDS, 0.005% bromophenol blue; Invitrogen) as well as 1M dithiothreitol

(DTT) (dissolve 1.55g of DTT powder in 10 mL of deionized water; Invitrogen) before proceeding to SDS-PAGE analysis.

2.6.5.2 Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-

PAGE) and western blotting.

SDS-PAGE analysis was performed using Novex NuPAGE or Bolt SDS-

PAGE gels (Life technologies). 10 μL of protein samples were loaded onto SDS-

PAGE gels (10 well 10% NuPAGE Bis-Tris; 10 well 4–12% Bolt Bis Tris Pre-cast gels, Life technologies) and electrophoresed in running buffer according to the manufacturer’s instructions. The gel, the blotting papers and pads were equilibrated in transfer buffer (29 mM glycine, 58 mM Tris-HCl pH 7.5, 0.0375% (w/v) SDS and

20% (v/v) methanol). The polyvinylidene difluoride (PVDF) membrane (GE

Healthcare) was quickly activated in methanol and then immersed in transfer buffer.

Proteins were then electroblotted in transfer buffer using the wet transfer process.

The transfer was carried out using XCell II Blot Module (Life technologies) for 1 hr at

30 V. The membrane was incubated for 1 hr at room temperature with blocking

solution (1 mM NaPO4 (pH 7.4), 0.09% NaCl and 0.01% tween, 5% (w/v) skimmed milk powder). The membrane was first incubated with the primary antibody diluted in blocking solution (1:5000) overnight at 4°C. After incubation, the membrane was washed for 3 x 15 min with PBS-Tween (1 mM NaPO4 (pH 7.4), 0.09% NaCl and

0.01% tween) buffer followed with a second incubation for 1 hr in the dark with specified secondary antibody (1:5000). The membrane was then washed for 3x15 min with PBS-Tween buffer in the dark. Visualisation and quantification of the protein expression levels were accomplished using WesternSure® Chemiluminescent

Reagents (LI-COR) and C-DiGit® Blot Scanner (LI-COR) from at least 3 independent experiments. Table 2.5 and Table 2.6 described the list of primary and secondary antibodies used in this study.

Table 2.5 Primary antibodies used in this study Primary antibody Dilution factor Source Sup35 (rabbit) 1:5000 Mick Tuite Protein A (rabbit) 1:5000 Sigma Phosphoglycerol kinase-1 (Pgk1, rabbit) 1:5000 Sigma Phosphoglycerol kinase-1 (Pgk1, mouse) 1:5000 Thermo Scientific Hsp104 (rabbit) 1:5000 Sigma M2 Flag (mouse) 1:5000 Sigma -Actin (mouse) 1:5000 Thermo Scientific

Table 2.6 Secondary antibodies used in this study Secondary antibody Dilution factor Source Goat Anti-Rabbit IgG Horseradish 1:5000 Sigma Peroxidase Conjugate Goat Anti-Mouse IgG Horseradish 1:5000 Promega Peroxidase Conjugate

2.7 Analysis of protein aggregation

2.7.1 Preparation of yeast cell cultures

Isolation of insoluble protein aggregates were based on (Jacobson et al.,

2012; Rand and Grant, 2006; Weids and Grant, 2014). Cell cultures were normalized to 10 OD600 units and cells were harvested at 4000 rpm for 3 min at 4°C and resuspended in 1 mL of aggregate lysis buffer (ALB – 50 mM potassium phosphate,

1 mM EDTA, 5% (v/v) glycerol, 1 mM PMSF and 1X Complete Mini protease cocktail

[Roche]) at 13000 rpm (maximum speed). Following another centrifugation step (1 min), cells were resuspended in 300 µL of ALB and 100 µL of 1 mg/mL lyticase was added prior to 30 min incubation at 30°C/180 rpm for 30 min to digest the yeast cell wall. Cell breakage was achieved by sonication (Sonifier 150, Branson; 8 x 5 sec,

Level 4) followed by centrifugation at max speed for 15 min (4°C). Samples were adjusted to equal protein concentrations (A280) before centrifugation at maximum speed for 20 min. The supernatant was removed (deemed as Total protein (T)) and the pellet was resuspended by sonication (4 x 5 sec, Level 4) in a mixture of ALB

(300 µL) and a non-ionic detergent to solubilize membrane proteins without denaturing them (100 µL of 10% (v/v) Igepal CA-630). The sample was centrifuged at maximum speed for 20 min and the supernatant was removed (Soluble protein

(S)) before repeating the procedure once more. To remove residual detergent, the sample was washed twice with ALB (without detergent), resuspending the pellet by sonication (4 x 5 sec, Level 4) and centrifuging at maximum speed for 20 min each time. The final washed pellet (Insoluble protein (I)) was resuspended in 100 µL of

ALB as well as in Nupage-loading buffer (4X) and DTT (10X) prior to protein denaturation at 95°C for 5 min. Samples were loaded onto 10% Nupage pre-cast gels and analyzed by silver staining and western blotting (see Section 2.6.5.2).

2.7.2 Silver-staining of SDS-PAGE gels

Silver staining of SDS-PAGE gels was performed using a commercial kit

(Silver Stain Plus, Bio-rad) according to the manufacturer’s instructions.

2.7.3 Mass spectrometry

Aggregated proteins were identified by mass spectrometry (performed by the

Biomolecular Analysis Core Facility, Faculty of Biology, Medicine and Health, the

University of Manchester) in triplicate for each condition. For protein identification, protein samples were run a short distance into SDS-PAGE gels and stained using

Brilliant Coomassie Blue (Sigma) for 1 hr followed by rinsing with sterilized water for

1 hr. Total proteins were excised, trypsin digested and identified using liquid chromatography-mass spectrometry (LC-MS).

2.7.4 Bioinformatic analysis of aggregated proteins

Proteins were identified using the Mascot mass fingerprinting programme

(www.matrixscience.com) to search the NCBInr and Swissprot databases. Final datasets for each condition were determined by selecting proteins that were identified in at least two of the three replicates. The various aggregated protein sets were initially compared using Venn diagrams generated using Venny

(http://bioinfogp.cnb.csic.es/tools/venny/). Protein datasets were assessed for enrichment of functional group categories with p-values <0.01 (MIPS database) using FunCat (http://mips.helmholtzmuenchen.de/en/ibis/). Finally, Mann–Whiney U- test was used to assess statistical differences in: 1) the codon adaptation index

(CAI); 2) isolelectric point (pI); 3) GRAVY score (Hydrophobicity); 4) protein stability by measuring its half-lives; 5) protein translation rates; and 6) protein abundance.

2.8 Live-cell fluorescence microscopy

2.8.1 Preparation of yeast samples

Following desired treatments of choice, 100 µL of cells were taken and briefly spun at 3,000 rpm for 30 sec, resuspended in the same media and immobilised on

10% poly-l-lysine pre-coated slides (Sigma). The number of cells containing aggregates deemed as puncta was quantified from at least 100 cells counted and was calculated based on the mean of at least three independent biological repeat experiments.

2.8.2 DeltaVision fluorescence microscopy

All images were acquired on a DeltaVision (Applied Precision) restoration microscope using a 100x/ 1.40 Plan Apo objective and FITC from the Sedat filter set

(Chroma) and MetaVue software (Bioimaging Facility, Faculty of Biomedicine and

Health, the University of Manchester. Raw images were then deconvolved using the

Softworx software (Applied Precision) and maximum intensity projections of these deconvolved images were analysed using Image J and shown in the results.

2.9 Analysis of prion formation

The 74D-694 strain used for these studies contains the ade1-14 nonsense

(UGA) mutant allele, which confers adenine autotrophy. The [psi−] cells made truncated Ade1 protein and therefore accumulate a red pigment, while suppression of the ade1-14 nonsense mutation in the [PSI+] gives rise to white/pink

Ade+ colonies. This phenotypic change was used for the following assays described below.

2.9.1 Determination of de novo [PSI+] prion formation

Briefly, the frequency of [PSI+] prion formation was scored by plating cells on

SC-Ade + 1% YPD (Sideri et al., 2011). Starting from an OD600 0.025, cells were grown for 3 days in 30 mL YPD media at 30°C/180 rpm. To determine cell viability, appropriate dilutions were then made using distilled water prior to plating on freshly prepared YPD plates as well as SC-Ade + 1% YPD plates and grown at 30°C for 3 days and 7 days, respectively. After counting the colonies, at least 200 colonies grown from SC-Ade + 1% YPD plates were restreaked onto new SC-Ade + 4 mM guanidine hydrochloride (GdnHCl), while plates containing SC-Ade + 1% YPD only acted as a control. GdnHCl inhibits the Hsp104 chaperone which is required during yeast prion fibre fragmentation and propagation and therefore can eliminate/cure yeast prions (Ferreira et al., 2001; Ness et al., 2002). Therefore, GdnHCl is used to distinguish between nuclear mutation and [PSI+] colonies; colonies that grew on SC-

Ade + 1% YPD plates but not on SC-Ade + 4mM GdnHCl were indeed [PSI+] colonies rather than nuclear mutation colonies. Finally, the frequency of de novo

[PSI+] prion formation as well as frequency of nuclear mutations was scored using the detailed formula as follows:

Frequency of de novo/ = Number of true [PSI+] colonies at 10x dilution for spontaneous [PSI+] prion a particular strain (on SC-Ade plates) formation Total number of colonies at 10 6+x dilution for a particular strains (on YPD plates) (x:dilution factor eg. 0, -2)

2.9.2 Determination of de novo [PIN+] Prion Formation

This assay has been previously described (Sideri et al., 2011). Strains were first transformed with the CUP1-SUP35NM-GFP plasmid and grown in the presence of 4 mM GdnHCl to cure the [PIN+] prion for at least 5 consecutive generations. 96

[pin−] transformed colonies were then inoculated in 150 μl of SC-Ura media containing 25 μM CuSO4 in 96-well plates and incubated overnight at 30°C/180 rpm.

5 μl from each of the 96 cultures were spotted onto ¼ YPD, ¼ YPD + 4 mM GdnHCl, or SC-Ade (containing 1% YPD to improve growth) plates for 7 days at 30°C.

Overexpression of Sup35NM-GFP generates [PSI+] colonies, which were scored as white/pink colonies and by their ability to grow in the absence of adenine, with both phenotypes being curable in the presence of GdnHCl. The rate of [PIN+] formation was inferred from the rate of [PSI+] formation, which is entirely dependent on cells being [PIN+] and expressed in triplicates.

3.0 Importance of mRNA surveillance pathways during oxidative stress conditions

3.1 Introduction

Cells typically respond to stress conditions by invoking complex regulatory mechanisms which act to reprogram translation during any stress conditions. For example, oxidative stress causes a global inhibition of translation which prevents continued global gene expression during potentially error-prone conditions as well as allowing for the turnover of existing mRNAs and proteins whilst gene expression is reprogrammed to deal with the stress (Proud, 2005). The initiation phase is the main target of regulation and represents a key control point for eukaryotic gene expression

(Shenton and Grant, 2003; Shenton et al., 2006). Oxidative stress also causes post- initiation inhibition of translation, increasing the average ribosomal transit time on mRNAs and causing codon-specific pausing (Pelechano et al., 2015; Shenton et al.,

2006). Additionally, oxidative stress conditions affect other stages of translation and for example, have been shown to promote misreading including translational readthrough of stop codons and frameshifting (Gerashchenko et al., 2012).

Moreover, other studies have also highlighted the importance of mRNA turnover as part of the coordinated response to oxidative stress (Henkel, 2011; Molina-Navarro et al., 2008). Taken together, it is clear that oxidative stress conditions impact protein synthesis at multiple levels and cells must be able to regulate their translational machinery to prevent the production of aberrant proteins.

The studies described in this Chapter have examined protein synthesis during oxidative stress conditions, focusing on the roles of mRNA surveillance pathways.

These are quality control systems which might act to prevent oxidative damage to mRNAs, or errors in translation, from altering the proteomic output, but has not been

systematically examined. The main focus of this Chapter therefore was to determine whether cytoplasmic mRNA surveillance pathways are required for resistance to oxidative stress conditions. They include NMD which targets mRNAs containing

PTC, NGD which targets mRNAs containing a sequence that stalls translating ribosomes during elongation and NSD which targets mRNAs lacking a stop codon or when ribosomes do not terminate translation at their normal stop codons causing them to translate into the 3’-UTR.

There are few examples of translational regulatory mechanisms which affect translation termination factor efficiency. One such example is provided by the eRF3

(Sup35) from Saccharomyces cerevisiae which is able to form prion aggregates known as [PSI+]. A further aim of this Chapter was to investigate the role mRNA surveillance pathways during the response to oxidative stress conditions, and whether this is influenced by the [PSI+] prion.

3.2 Results

3.2.1 Construction of mRNA surveillance mutants

Yeast mutants were constructed lacking various components of mRNA surveillance pathways including UPF1, UPF2 (NMD), DOM34, HBS1 (NGD) and

SKI7, SKI8 (NSD) in the 74D-694 yeast strain background which contains a suppressible ade1-14 nonsense mutation (UGA) to detect the presence of the [PSI+] prion (Chernoff et al., 1995; Tyedmers et al., 2008). Mutants were constructed using a standard PCR-based gene deletion strategy as detailed in Fig. 3.1. Briefly, primers have ~50-nucleotides (nt) ends for priming upstream and downstream of the ATG sites flanking the HIS gene in pRS413 and ~20-nt 5’ ends homologous to the upstream and downstream chromosomal sequences to the gene of interest.

Transformants were selected by plating onto SCD media lacking histidine. Genomic

DNA was extracted from candidate transformants and correct deletions identified using a second round of PCR to verify the correct integration of the HIS3 cassette.

Representative images are shown confirming the deletions made for SKI8, UPF2 and HBS1 in [psi-] and [PSI+] versions of strain 74D-694 (Fig. 3.2). In each case, forward (for) and reverse (rev) oligonucleotides were used to amplify the gene of interest. Additionally, PCR reactions were performed using for and rev primers corresponding to an internal region of the HIS3 gene. The expected sizes confirming successful gene deletions are detailed in the legend for Fig. 3.2:

Figure 3.1: Construction of mRNA surveillance mutants in strain 74D-694 using a HIS3 cassette. The HIS3 gene was amplified in a first round of PCR using oligonucleotides which contained 20 bp homology regions flanking immediately upstream (for) and downstream (rev) of the yeast ORF of interest. This was introduced into the genomic DNA of S. cerevisiae by transformation and homologous recombination. Genomic DNA was then extracted from candidate mutant strains which were screened using a second round of PCR. The same protocol was used to generate LEU2 deletion cassettes using different template DNA.

Figure 3.2: Construction of mRNA surveillance mutants in strain 74D-694 using a HIS3 cassette. (A) Agarose gels (1%) are shown for HIS3 deletion cassettes which were generated by amplifying the HIS3 gene using oligonucleotides which included 50 bp regions flanking upstream (for) and downstream (rev) of the gene of interest. Verification PCR reactions were used to confirm that the HIS3 cassette was systematically replaced for SKI8 (B), UPF2 (C) and HBS1 (D). DNA from the WT strain was used in lanes 1 and 5 as positive controls and expected sizes are 2394 bp with SKI8 primers, 3783 bp with UPF2 primers and 2236 bp with HBS1 primers. Lanes 5 and 6 also used DNA from the WT strain but with a HIS3 rev primer. Hence, there should not be any band from a WT strain and therefore acts as negative control. Lanes 3 and 7 used for and rev primers flanking the gene of interest hence a successful HIS3 gene deletion should produce a band of 2200 bp. Lanes 4 and 8 used a for primer upstream of the gene of interest with a HIS3 primer (HIS3 rev). Thus, a single band of 1100 bp confirming successful gene deletion is expected.

3.2.2 Phenotypic assays of [psi-] and [PSI+] strains

Unless otherwise stated, all strains were made in the 74D-694 background containing a premature UGA stop codon in the ADE1 gene. Therefore, the [psi–] strains were unable to grow in the absence of exogenous adenine due to terminating translation at the PTC and accumulate an intermediate in the adenine biosynthetic pathway that causes the colonies to be red (Serio and Lindquist, 1999). The ¼ YPD media is used because it allows the colonies appearing red in colour to be more easily observed. All strains grew normally on a control plate. However, it was observed that the ski8 mutant is slightly pink in colour compared to the red colour observed in the other strains (Fig. 3.3A).

3.2.3 Requirement for mRNA surveillance pathways during oxidative stress

The sensitivity of mRNA surveillance pathway mutants to oxidants was examined to determine the requirement for mRNA surveillance pathways during oxidative stress conditions. Sensitivity to oxidative stress conditions was first examined using spot test assays. The WT and mutant strains were grown to stationary phase in SCD media and serial dilutions spotted onto SCD plates containing various concentrations of H2O2, diamide, copper (CuSO4), and cadmium

(CdSO4). All strains appeared to grow comparably under non-stressed condition.

Mutants lacking UPF1 or UPF2 were relatively unaffected in oxidant sensitivity, and showed modest resistance to copper stress, suggesting that NMD is not required for oxidant tolerance (Fig. 3.3B). In contrast, strains deleted for DOM34 or HBS1 showed modest sensitivity to H2O2, diamide and cadmium, and were very sensitive to copper stress. The strongest phenotype was observed with a mutant strain deleted for SKI8 which was hypersensitive to all toxicants (Fig. 3.3B). However,

mutant lacking SKI7 was unaffected in sensitivity to oxidants compared with the ski8 mutant. This difference in sensitivity between ski7 and ski8 mutants was unexpected and is further examined in Chapter 4. In summary, some mRNA surveillance mutants appeared to be sensitive to all stressors (ski8 mutant), or were stress-specific (NGD mutants) while others remained unaffected when exposed to these ROS-generating compounds (NMD mutants). For the sake of simplicity, we have focused on H2O2 from this point forward to further understand the role of mRNA surveillance mechanisms during oxidative stress conditions.

3.2.4 Growth and viability analysis of mRNA surveillance mutants during oxidative stress conditions

To verify the results obtained using spot test assays, two additional assays were used to compare oxidant sensitivity. First, we examined whether the loss of mRNA surveillance pathways alters growth kinetics during oxidative stress conditions. Cultures were grown to early exponential phase (OD600 ~0.2) and then exposed to 1 mM H2O2 before monitoring growth hourly over six hours. Growth sensitivity might arise due to inhibiting cell division or causing cell death. In a second assay we therefore examined cell viability following exposure to H2O2. Late- exponential phase cells (OD600 ~0.7) were treated with 2 mM H2O2 for 1 hour.

Appropriate dilutions were then made before plating on YPD media and determining cell survival. Both assays were performed using [psi-] and [PSI+] versions of mRNA surveillance mutants to determine whether the [PSI+] status influences oxidant sensitivity.

100

Figure 3.3: Oxidative stress-sensitivity of mRNA surveillance mutants. (A) The WT and mutant strains were spotted onto media lacking adenine and on ¼ YPD plates to confirm that they are [psi-]. (B) Sensitivity to oxidative stress was determined by spotting strains onto SCD media containing various concentrations of

H2O2, diamide, copper (II) sulphate and cadmium sulphate. Results are shown for the WT, ski7, ski8, upf1, upf2, dom34 and hbs1 mutant strains grown to stationary phase, and the A600 was adjusted to 1, 0.1, 0.01, or 0.001 before spotting. Growth was monitored after three days incubation at 30°C.

101

3.2.4.1 The oxidant sensitivity of ski8 mutant depends on its [PSI+] status

Little difference in growth rate was observed between the [psi-] versions of the

WT, ski7 and ski8 mutants in the absence of stress (Fig. 3.4A). Exposure to 1 mM

H2O2 significantly inhibited the growth of all strains, but the ski8 mutant showed the highest sensitivity compared to the WT and ski7 mutant strains. For the viability assay, a final concentration of 2 mM of H2O2 was used since this dose is relatively toxic to yeast cells. Approximately 90% loss of viability was observed in the WT and ski7 mutants within one hour of treatment (Fig. 3.4B). In comparison, the ski8 mutant was more sensitive to H2O2 confirming that H2O2 causes a greater loss of viability

(~98%) in the ski8 mutant compared with WT and the ski7 mutant strains (Fig. 3.4B).

Interestingly, no differences in oxidant sensitivity were observed in the [PSI+] versions of the WT, ski7 and ski8 mutant strains. All three strains showed a similar inhibition of growth following exposure to 1 mM H2O2 (Fig. 3.4C). The WT, ski7 and ski8 mutant strains also showed a similar loss of viability following exposure to 2 mM

+ H2O2 (Fig. 3.4D). The [PSI ] strains were generally more oxidant tolerant than the

[psi-] strains suggesting that [PSI+] improves oxidant tolerance such that differences in oxidant sensitivity are no longer observed between ski7 and ski8 mutants.

102

Figure 3.4: ski8 mutant is hypersensitive to oxidative stress conditions but this sensitivity is lost in the [PSI+] version. Growth curves are shown for the WT, ski7 and ski8 mutant strains treated with 1 mM H2O2 for six hours in A) and C). Growth was monitored by measuring absorbance at 600 nm. Filled symbols denote growth in the absence of oxidant and open symbols denote growth in the presence of oxidant and compared between [psi-] strains (A) and the [PSI+] strains (C). Viability analysis is shown for the same strains grown to exponential phase in minimal media and treated with 2 mM H2O2 for one hour in B) and D). Cells were diluted and plated in triplicate onto YPD medium to monitor cell viability. Percent survival is expressed relative to the untreated control cultures and compared between the [psi-] strains (B) and the [PSI+] strains (D).

103

3.2.4.2 Loss of NMD may act to improve oxidan1t tolerance and this phenotype is further observed in [PSI+] background

Growth inhibitory assay revealed that the [psi-] NMD mutants (upf1, upf2) had better oxidant tolerance compared with the WT strain (Fig. 3.5A). However, viability analysis revealed that the upf1 and upf2 mutants were more sensitive to H2O2 compared with the WT strain (Fig. 3.5B). Thus, although NMD mutants grow at a faster rate than the WT in the presence of H2O2, cell death is higher which may account for the lack of sensitivity to H2O2 observed using spot test analysis (Fig. 3.3).

Interestingly however, the [PSI+] versions of both upf1 and upf2 mutants were significantly more oxidant tolerant than the [PSI+] WT strain measured using growth rate (Fig. 3.5C) and viability (Fig. 3.5D) analyses. This suggests that the [PSI+] status of cells has synergistic effects in combination with NMD mutants resulting in improved viability and shall be explored further in Chapter 5.

3.2.4.3 Mutants in the NGD pathway are modestly sensitive to oxidative stress

Similar to the spot tests shown in Fig. 3.3, the [psi-] NGD mutants (dom34, hbs1) showed modest sensitivity to H2O2 as measured using growth rate (Fig. 3.6A) and viability (Fig. 3.6B) analyses. Again, the [PSI+] strains were generally more oxidant tolerant than their [psi-] counterparts. Additionally, no oxidant sensitivity was observed in [PSI+] versions of NGD mutants compared with the [PSI+] WT strain as measured using growth rate (Fig. 3.6C) and viability (Fig. 3.6D) analyses.

104

Figure 3.5: The oxidant tolerance of NMD mutants depend on their [PSI+] status. Growth curves are shown for the WT, upf1 and upf2 mutant strains treated with 1 mM H2O2 for six hours in A) and C). Growth was monitored by measuring absorbance at 600 nm. Filled symbols denote growth in the absence of oxidant and open symbols denote growth in the presence of oxidant and compared between [psi-] strains (A) and the [PSI+] strains (C). Viability analysis is shown for the same strains grown to exponential phase in minimal media and treated with 2 mM H2O2 for one hour in B) and D). Cells were diluted and plated in triplicate onto YPD medium to monitor cell viability. Percent survival is expressed relative to the untreated control cultures and compared between the [psi-] strains (B) and the [PSI+] strains (D).

105

3.2.5 Analysis of translational activity in mRNA surveillance mutants during oxidative stress conditions

H2O2 stress is known to cause a global inhibition of translation in yeast, predominantly occurring at the initiation phase (Shenton and Grant, 2003; Shenton et al., 2006). We therefore used polysome analysis to compare translational activity in mRNA surveillance mutants to determine whether any differences in translational regulation might account for oxidant sensitivity. Polysomes are ribosomes which are actively translating mRNAs. They can be separated on sucrose density gradients and quantified by measuring absorbance at 254nm. Strains were grown to exponential phase (OD600 ~0.5) and treated with increasing concentrations of hydrogen peroxide (0-1 mM) for 1 hr. Extracts were prepared in the presence of cycloheximide to quickly arrest protein synthesis prior to analysis. Polysome profiles were analysed using [psi-] and [PSI+] versions of mRNA surveillance mutants to determine whether the [PSI+] status influences translational regulation during oxidative stress conditions.

Previous studies have shown that hydrogen peroxide causes a dose-dependent inhibition of translation initiation (Shenton et al., 2006). A similar translational response was observed in our [psi-] WT strain with a shift of ribosomes from the polysomal region into the monosome or 80S peak indicating a block of translation initiation (Fig. 3.7). Treatment with 1 mM H2O2 caused a stronger inhibition of translation initiation compared with 0.5 mM H2O2 as indicated by the decreased polysome to monosome (p/m) ratio. This inhibition of translation was comparable in

WT and NSD mutant strains (ski7, ski8) suggesting that alterations in translational activity do not account for the oxidant sensitivity of the ski8 mutant.

106

Figure 3.6: Mutants in the NGD pathway are modestly sensitive to oxidative stress. Growth curves are shown for the WT, dom34 and hbs1 mutant strains treated with 1 mM H2O2 for six hours in (A) and (C). Growth was monitored by measuring absorbance at 600 nm. Filled symbols denote growth in the absence of oxidant and open symbols denote growth in the presence of oxidant and compared between [psi-] strains (A) and the [PSI+] strains (C). Viability analysis is shown for the same strains grown to exponential phase in minimal media and treated with 2 mM

H2O2 for one hour in B) and D). Cells were diluted and plated in triplicate onto YPD medium to monitor cell viability. Percent survival is expressed relative to the untreated control cultures and compared between the [psi-] strains (B) and the [PSI+] strains (D).

107

Polysome profiles were harder to interpret for NMD (Fig. 3.8) and NGD (Fig.

3.9) mutants. This was because the p/m values were already decreased in upf1, upf2, dom34 and hbs1 mutants compared with a WT strain in the absence of stress.

It is unclear why this arises but it may indicate that there is a slower rate of translation initiation in NMD and NGD mutants. Alternatively, increased concentrations of mRNA pools in these mutants may result in increased monosome and decreased polysome levels due to limiting ribosome availability. Nevertheless, inhibition of translation initiation was observed in NMD (Fig. 3.8) and NGD (Fig. 3.9) mutants in response to oxidative stress since the p/m values were decreased compared with the untreated strains. The [PSI+] status of all strains did not significantly affect translation initiation as similar polysome profiles were observed in

[psi-] and [PSI+] strains. Taken together, these data suggest that inhibition of translation initiation does not account for sensitivity to oxidative stress in mRNA surveillance mutants and this was investigated further in the next section by examining the rate of protein synthesis.

3.2.6 Analysis of protein synthesis in mRNA surveillance mutants during oxidative stress conditions

Since there were no significant differences in polysome profiles from the pairs of mutants examined for each pathway we initially focused on ski8 (NSD), upf2

(NMD) and hbs1 (NGD) mutants. Protein synthesis was measured by growing cells to early exponential phase, treating with various concentrations of H2O2 and pulse labelling with [35S]-cysteine/methionine for the final five minutes of the one hour treatment. Treatment with 0.5 mM H2O2 modestly inhibited protein synthesis in a

108

Figure 3.7: Differences in oxidant sensitivity do not arise due to translational activity in NSD mutants. Polyribosome traces are shown for the indicated strains treated with increasing concentrations of H2O2 for one hour. A block in initiation is indicated by an accumulation of 80S monosomes and a reduction in polysomes. Numbers in brackets are the polysome:monosome (p:m) values determined by the ratio between the area under the monosome to the polysome peaks. Data are expressed relative to the untreated controls. Representative data are shown from at least 3 independent experiments.

109

Figure 3.8: NMD mutants exhibit moderate effect of inhibition of translation initiation. Polyribosome traces are shown for the indicated strains treated with increasing concentrations of H2O2 for one hour. A block in initiation is indicated by an accumulation of 80S monosomes and a reduction in polysomes. Numbers in brackets are the polysome:monosome (p:m) values determined by the ratio between the area under the monosome to the polysome peaks. Data are expressed relative to the untreated controls. Representative data are shown from at least 3 independent experiments.

110

Figure 3.9: Inhibition of translation initiation was observed in NGD mutants even in unstressed condition. Polyribosome traces are shown for the indicated strains treated with increasing concentrations of H2O2 for one hour. A block in initiation is indicated by an accumulation of 80S monosomes and a reduction in polysomes. Numbers in brackets are the polysome:monosome (p:m) values determined by the ratio between the area under the monosome to the polysome peaks. Data are expressed relative to the untreated controls. Representative data are shown from at least 3 independent experiments.

111

WT strain compared with 0.8 mM and 1 mM H2O2 treatments which inhibited protein synthesis by approximately 60% and 90%, respectively (Fig. 3.10A).

A similar inhibition of protein synthesis was observed in response to H2O2 in the WT and ski8 mutant strains (Fig. 3.10A) in agreement with the polysome analysis

(Fig. 3.7). However, the [PSI+] ski8 mutant showed a greater inhibition of protein

- synthesis following treatment with 0.5 and 0.8 mM H2O2 compared with the [psi ] ski8 mutant. The upf2 mutant was somewhat more resistant to H2O2 compared with the

WT strain. Finally, loss of HBS1 resulted in increased inhibition of protein synthesis

+ in response to H2O2 compared with the WT strain (Fig. 3.10A). However, the [PSI ] status of the hbs1 mutant did not appear to strongly affect protein synthesis.

Additionally, since translation is known to be inhibited at a post-initiation phase in response to H2O2 stress (Shenton et al., 2006), it is therefore possible that other phases of translation eg. elongation phase is more sensitive to oxidative stress as evident in the hbs1 mutant.

We decided to make a direct comparison between ski7 and ski8 mutants in terms of protein synthesis rate. In the absence of H2O2, all strains appeared to have similar rate of protein synthesis (Fig. 3.10B). Treatment with H2O2 markedly reduced the rate of protein synthesis of the ski7 and ski8 mutants particularly at 1 mM, albeit to a similar pattern observed in the wild type strain, thus confirming the results observed in polysome analysis earlier (Fig. 3.10B). Our polysome analysis revealed that regulation of translation initiation could not explain the differences in the oxidant sensitivity observed between the ski7 and the ski8 mutants.

112

Figure 3.10: Hydrogen peroxide causes inhibition of protein synthesis all strains following 1 hour treatment. (A) Following H2O2 treatment, hbs1 mutant is most sensitive followed by ski8, WT and upf2. (B) No difference in the rate of protein synthesis between the NSD mutants and the WT strain. Mid-exponential cells (OD600

~0.5) were treated with increasing concentrations of H2O2 (0-1 mM) for 1 hour. Protein synthesis was measured by pulse labelling with [35S] cysteine/methionine during the last 5 min of the treatment. Data are shown for untreated cultures (100%) and relative percentage for the treated samples. Each concentration was analysed in triplicate ± SD.

113

3.3 Discussion

All organisms are exposed to ROS during the course of normal aerobic metabolism resulting in translational defects and aberrant protein production.

Therefore, mRNA surveillance pathways which target aberrant mRNAs for degradation may be activated to prevent the production of aberrant proteins. In this study, we examined the requirement for mRNA surveillance pathways including

NSD, NMD and NGD pathways during oxidative stress conditions. Two striking phenotypes were observed: 1) mutant lacking SKI8, one of the core components of

NSD, was found to be hypersensitive to oxidative stress; 2) mutant lacking SKI7, another member of NSD however appeared unaffected in sensitivity to all toxicants, indicating that different requirements exist between the ski7 and ski8 mutants during oxidative stress conditions.

The second aim of this study was to examine whether the [PSI+] status of mRNA surveillance pathway mutants influences oxidant tolerance. Although the consequences of [PSI+] status on readthrough and evolution have previously been considered, little attention has focused in the context of stress conditions. Studies on prion mechanisms have revealed that the shift to the [PSI+] state for some mutants or strains may provide a mechanism for generating heritable phenotypic diversity by allowing cells to reprogram gene expression such that they may become more tolerant of specific stress conditions (Eaglestone et al., 1999, True and Lindquist,

2000; Tyedmers et al., 2008). Additionally, the [PSI+] prion previously shown in our lab provided yeast cells with an adaptive advantage under oxidative stress conditions, since elimination of the [PSI+] status from tsa1 tsa2 mutants rendered

+ cells hypersensitive to H2O2 (Sideri et al., 2011). Consistently, the [PSI ] versions of

74D-694 strain background used in this study were also found to be more resistant

114

to oxidative stress which might be due to stop codon readthrough that produces genetic diversity and possibly alters translation of specific mRNAs to promote oxidant tolerance.

The most dramatic phenotypes were found with mutants lacking SKI8 which were hypersensitive to all oxidants tested suggesting that the recognition and degradation of NSD substrates is particularly required for oxidant tolerance.

However, a ski7 mutant was unaffected in oxidant sensitivity. Growth and viability analysis revealed that the oxidant sensitivity of the ski8 mutant arises due to effects on both growth and cell viability. The hypersensitivity of the ski8 mutant to oxidative stress was abrogated in the [PSI+] ski8 mutant suggesting that the [PSI+] state can improve defense mechanisms during oxidative stress. It does not appear that this improvement in oxidant tolerance is specific to the ski8 mutant since a similar increase in oxidant resistance was observed in the WT strain, too.

We reasoned that differences in oxidant sensitivity and the effect of the [PSI+] prion might arise from alterations in translational regulation as part of the stress response. However, our current findings ruled out the possibility that differences in oxidant sensitivity between ski7 and the ski8 mutants arise as a result of differences in translational activity of these mutants. Similar inhibition of translation initiation, as assessed using polysome analyses, and inhibition of protein synthesis, as assessed using radiolabeling, was observed in the WT, ski7 and ski8 mutant strains. SKI8 is a part of the Ski complex composed of SKI2, SKI3 and SKI8, and together with the cytoplasmic exosome, is recruited by SKI7 during NSD. Therefore, we reasoned that difference in the functional roles of Ski7 and Ski8 might account for such phenotypic differences during oxidative stress conditions, and this shall be examined in more detail in Chapter 4.

115

4.0 Requirement for nonstop decay during oxidative stress conditions

4.1 Introduction

mRNA degradation is a crucial part of the regulation of gene expression.

Cytoplasmic mRNAs are degraded via two different pathways that have been conserved throughout eukaryotes and are first initiated by the removal of the poly(A) tail (deadenylation). This is then followed by removal of the 5’-cap and subsequent degradation of the mRNA from the 5’-end by the Xrn1 exonuclease (Siwaszek et al.,

2014). Xrn1 has also been shown to promote general 5’-3’ co-translational mRNA decay following the last translating ribosome (Pelechano et al., 2015). Alternatively, mRNAs can be degraded from the 3’-end by the action of the exosome, which is a conserved multiprotein complex that degrades RNAs in the 3’-to-5’ direction in the cytoplasm of eukaryotic cells (Lebreton and Seraphin, 2008).

As described in Section 1.6, mRNA decay is tightly linked to translation and there are three translation-associated mRNA surveillance pathways which prevent the production of potentially toxic proteins: NMD, NGD and NSD. These quality control mechanisms all ultimately target mRNAs for degradation (Shoemaker and

Green, 2012). Aberrant mRNAs might arise from mistakes during transcription or translation events, or due to environmental factors causing mRNA damage. A previous study for example has shown that yeast mutants deficient in the exoribonuclease activity of the exosome are sensitive to oxidative stress (Tsanova et al., 2014). Another study has shown that 8-oxoguanosine (8-oxoG) accumulates in mRNAs isolated from mutants deficient in NGD and the formation of 8-oxoG on mRNAs can stall the translational machinery (Simms et al., 2014). 8-oxoG is a major nucleotide oxidation product and these data reinforce the idea that mRNA

116

surveillance mechanisms are important to protect against the consequences of decoding damaged RNA.

In Chapter 3, yeast mutants defective in various components of mRNA surveillance pathways were tested for their sensitivity to oxidative stress. Our data indicate that mutants lacking components of NMD and NGD pathways are unaffected in oxidant sensitivity suggesting that the recognition of ribosomal pauses

(NGD) or PTCs (NMD) is not generally required for yeast tolerance to oxidative stress conditions. Interestingly, we found that the ski8 mutant is particularly sensitive to oxidative stress. Surprisingly however, the ski7 mutant was not found to be oxidant sensitive. Ski8 is a component of the Ski complex, which is an evolutionarily conserved complex of three proteins that is functionally and physically associated with the exosome. It comprises the DExH RNA helicase Ski2, the tricopeptide repeat protein Ski3 and two copies of the WD repeat protein Ski8 (Halbach et al., 2013; van

Hoof et al., 2002). In yeast, Ski7 (an eRF3 family member) bridges the interaction between the SKI complex and the exosome. Ski7 is thought to bind to ribosomes stalled at the 3′-end of mRNAs where it recruits the exosome to trigger 3′-to-5′ degradation of NSD substrates. We therefore tested the hypothesis that the oxidant sensitivity of ski8 mutant is a common feature of ski complex mutants by examining the oxidant sensitivity of mutants deleted for SKI2 and SKI3.

117

4.2 Results

4.2.1 Phenotypic assays of ski mutants

The ski2 and ski3 mutants were constructed in yeast strain 74D-694 as described in Section 3.2. The growth of the ski2 and ski2 mutants was first compared with WT, ski7, ski8 and mutant strains using spot test analyses. All strains appeared to grow similarly on a control SCD plate (Fig. 4.1A). However, the ski complex mutants consisting of ski2, ski3 and ski8 grew pink in colour on ¼ YPD media as opposed to the red colour observed on the WT and ski7 mutant strains. This is perhaps due to low reduced-glutathione levels which are needed in transporting the red intermediate compound (P-ribosylaminoimidazole carboxylayte) into vacuoles and develop the red color (Bharathi et al., 2016). Additionally, some small white colonies grew in the ski3, ski7 and ski8 spots on –Ade plate after seven days growth at 30°C suggesting the formation of Ade+ colonies. These colonies were tested for

[PSI+] formation by restreaking onto YPD plates with or without 4 mM GdnHCl to test for curing. However, the pale red colour and growth on –Ade plates was unaffected by curing suggesting that they probably arise due to nuclear mutations in adenine biosynthetic genes.

4.2.2 The Ski complex is required for oxidant tolerance

Sensitivity to oxidative stress was first determined using spot test assays. The

WT and ski mutants were grown to stationary phase in SCD media and spotted onto

SCD plates containing various concentrations of H2O2, diamide, CuSO4 and CdSO4.

Mutants deleted for SKI2 or SKI3 were found to show strong sensitivity to H2O2 similar to the ski8 mutant (Fig. 4.1B). The sensitivity to H2O2 appears to be a general

118

Figure 4.1: Mutants in the Ski complex are sensitive to oxidative stress. (A) The WT and ski mutant strains were spotted onto media lacking adenine and on ¼ YPD plates to confirm that they are [psi-]. (B) Sensitivity to oxidative stress was determined by spotting strains onto SCD media containing various concentrations of hydrogen peroxide, diamide, copper (II) sulphate and cadmium sulphate. Results are shown for the WT, ski2, ski3, ski7 and ski8 mutant strains following three days growth at 30°C. (C) The experiment was repeated but this time using NSD mutants as well as xrn1 mutant in the BY4741 background. Sensitivity to oxidative stress was determined using 2.5 mM, 3.25 mM and 3.5 mM H2O2. Results are shown for the WT, ski2, ski3, ski7, ski8, and xrn1 mutant strains following three days growth at 30°C.

119

sensitivity to oxidative stress conditions since the ski2, ski3 and ski8 mutant strains were also sensitive to diamide, copper and cadmium stress (Fig. 4.1B).

To check whether this pattern of H2O2 sensitivity is specific to the 74D-694 yeast strain background, the same ski mutants were tested in the BY4741 strain background. The xrn1 mutant from the BY4741 strain background was also spotted to determine whether the cytoplasmic exosome is required for oxidant tolerance, since a previous study had shown that mutants deficient in the exoribonuclease activity of the exosome are sensitive to oxidative stress conditions (Tsanova et al.

2014). Again, the ski2, ski3 and ski8 mutants were more sensitive to 3.25 mM H2O2 compared with the WT and ski7 mutant strains (Fig. 4.1C). The xrn1 mutant showed moderate sensitivity to 3.5 mM H2O2 suggesting that this cytoplasmic exosome may also be required during oxidative stress conditions (Fig. 4.1C).

Growth rate analysis (Fig. 4.2A) and viability measurements (Fig. 4.2B) were used to verify the oxidant sensitivity of the ski mutants. For growth analysis, cultures were initially grown overnight to early exponential phase (A600~0.2) in SCD media before adding H2O2 to a final concentration of 1 mM. Growth was monitored by measuring A600 every hour for a total of eight hours. For viability analysis, mid- exponential phase cultures were grown in SCD media and treated with 2 mM H2O2 for one hour. Cultures were serially diluted into YPD media and plated onto YPD plates. Viable counts were recorded and expressed as a percentage of untreated cultures following three days growth at 30°C.

During normal, unstressed conditions, all strains appeared to grow comparably. Following H2O2 treatment however, the ski complex mutants are significantly inhibited (Fig. 4.2A). Again, greater loss of cell viability was observed in the ski2, ski3 and ski8 mutants compared with the WT and ski7 mutant strains

120

Figure 4.2: Mutants in the ski complex are sensitive to oxidative stress. (A)

Growth curves are shown for the same strains treated with 1 mM H2O2 for eight hours. Growth was monitored by measuring absorbance at 600 nm. Filled symbols denote growth in the absence of oxidant and open symbols denote growth in the presence of oxidant. (B) Viability analysis is shown for the same strains grown to exponential phase in minimal media and treated with 2 mM H2O2 for one hour. Cells were diluted and plated in triplicate onto YPD medium for three days at 30°C to monitor cell viability. Percent survival is expressed relative to the untreated control cultures.

121

(Fig. 4.2B). Taken together, these data indicate that the SKI complex is required for oxidant tolerance, but there does not appear to be a similar requirement for Ski7.

4.2.3 Analysis of NSD in ski complex and ski7 mutants

Given the differences in oxidant sensitivity observed in ski2, ski3 and ski8 mutants compared with the ski7 mutant, we next examined whether they show any differences in the recognition and degradation of an NSD substrate. To monitor NSD in ski mutants, we used a Protein A-nonstop reporter construct which contains the

GAL1 promoter, the Protein A coding region and the PGK1 3’UTR (Wilson et al.,

2007). This reporter lacks any stop codon and so its mRNA is a target for NSD (Fig.

4.3A). Cells were grown to exponential phase (A600 ~0.3-0.5) in SRaf media prior to adding 1% galactose (in the presence or absence of 0.5 mM H2O2) to induce the expression of the GAL1 promoter for six hours. NSD can be monitored by analysing

Protein A levels which are minimal in WT strain, whereas Protein A should be produced in a NSD mutant due to stabilization of the Protein A nonstop mRNA

(Wilson et al., 2007).

During unstressed conditions, high levels of Protein A were detected in ski complex mutants compared with the WT strain (Fig. 4.3B). Interestingly, far more

Protein A was produced in the ski complex mutants compared with the ski7 mutant

(Fig. 4.3B). We also examined whether oxidative stress caused by H2O2 exposure affected the production of the NSD nonstop Protein A reporter. Oxidative stress did not alter the high levels of Protein A in the ski complex mutants, whereas there was an approximately 50% decrease in Protein A production in the ski7 mutant

(Fig.4.3B). This decrease in the production of an NSD protein correlates with the

122

14 A. C. ) Control

aU 12 H2O2

10 NSD ( NSD 8 6 4

0 WTWT ski2 ski3 ski7 ski8

WT ski2 ski3 WT ski7 ski8

kDa - + - + - + - + - + - +

control control 225 120 80 70 50 40 30 NSD NSD 25 15

actin actin

Figure 4.3: Analysis of NSD in ski7 and ski complex mutants. (A) A Protein A- nonstop reporter construct was used to monitor NSD in WT and ski family mutant strains. This construct contains the GAL1 promoter, the Protein A coding region and the PGK1 3’UTR (Wilson et al., 2007). (B) Protein extracts were isolated from WT and the ski mutant strains expressing Protein A-nonstop protein and analyzed by

Western blotting in the presence or absence of 0.5 mM H2O2 for six hours. Blots were probed with a Protein A antibody (NSD) and an actin antibody as a loading control. Control denotes a WT strain containing an empty vector. (C) Quantification of Protein A production as shown in (B) from three independent experiments ± SD.

123

increased oxidant tolerance of the ski7 mutant compared with other ski mutants.

Similar levels of actin were detected in the WT and ski mutant strains. Overall, these data suggest that Ski7 is not solely responsible for recognizing NSD substrates in S. cerevisiae.

4.2.4 Overlapping requirements for Ski7 and Dom34/Hbs1 during oxidative stress conditions

Given the role of SKI7 in recruiting the Ski complex and the cytoplasmic exosome to degrade NSD substrate, it was somewhat unexpected that the ski7 mutant is unaffected in oxidative stress tolerance when mutants lacking components of the Ski complex are hypersensitive to oxidative stress. Ski7 is closely related to

Hbs1 which is required for NGD and it was demonstrated that Hbs1 can function in both NSD and NGD in mammalian cells (Shoemaker and Green, 2012). We therefore investigated the overlap between NSD (Ski7) and NGD (Dom34/Hbs1) components to examine whether they play a redundant role during oxidative stress conditions. Mutants were constructed lacking both SKI7 (HIS3) and DOM34 (LEU2) or SKI7 (HIS3) and HBS1 (LEU2) using standard PCR gene deletion methodology described in Section 3.1 and examined for their sensitivity to oxidative stress.

Our spot test assays revealed that deletion of DOM34 or HBS1 increased the oxidant sensitivity of a ski7 mutant to H2O2, although the double mutants were not as sensitive as a ski complex mutant (Fig. 4.4A). We further confirmed the oxidant sensitivity of ski7 dom34 and ski7 hbs1 mutants using growth kinetics (Fig. 4.4B) and cell viability analysis (Fig. 4.4C) following H2O2 treatment. These analyses showed that the ski7 dom34 and ski7 hbs1 mutants were more sensitive to H2O2 compared with the respective single mutant strains.

124

Figure 4.4: Overlapping requirements for Ski7 and Dom34/Hbs1 during oxidative stress conditions. (A) Sensitivity to oxidative stress was determined by spotting strains onto SCD media containing various concentrations of H2O2. Results are shown for the WT, ski7, dom34, hbs1, ski7 dom34, ski7 hbs1 and ski2 mutant

strains following three days growth on 2.5 mM H2O2. (B) For growth analysis, cultures were initially grown overnight to early exponential phase (A600~0.2) in SCD media before adding H2O2 to a final concentration of 1 mM. Growth was monitored by measuring A600 every hour for a total period of eight hours. (C) For viability analysis, mid-exponential phase (A600~0.4) cells grown in SCD media were treated with 2 mM H2O2 for one hour. Cultures were serially diluted into YPD media and plated onto YPD plates. Viable counts were recorded and expressed as a percentage of untreated cultures following three days growth at 30°C.

125

We used the Protein A-nonstop reporter construct to determine whether loss of DOM34 or HBS1 affects the recognition of a nonstop mRNA substrate in a ski7 mutant. The Protein A-nonstop reporter construct was present at similar concentrations in dom34 and hbs1 mutants when compared with the ski7 mutant; this is consistent with the idea that the Hbs1-Dom34 complex can function in NSD

(Fig. 4.5A). Furthermore, there was a significant increase in Protein A levels in the ski7 dom34 and ski7 hbs1 mutants suggesting that Ski7 and the Hbs1-Dom34 complex may play a redundant role in the recognition of NSD substrates. However, the levels of Protein A in the ski7 dom34 and ski7 hbs1 mutants were approximately

50% of those detected in the ski2 mutant suggesting that other factors are also required for the Ski complex to recognize NSD substrates.

We used qRT-PCR to measure any changes in Protein A-nonstop mRNA levels for the WT, ski7, dom34, hbs1, ski7 dom34, ski7 hbs1 and ski2 mutant strains.

Cells were grown to exponential phase (A600 ~0.3-0.5) in SRaf media before adding

1% galactose to induce expression of the GAL1 promoter for six hours. One of the housekeeping genes, -actin was used to control for differences in the mRNA levels between samples. We found that the levels of Protein A-nonstop mRNA were higher for all the mutant strains compared with the WT strain. Furthermore, both ski7 dom34 and ski7 hbs1 mutant strains exhibit greater Protein A-nonstop mRNA levels than the single mutant strains alone. This confirmed that the relative concentrations of Protein A-nonstop protein levels observed in Fig. 4.5A correlate with changes in mRNA levels (Fig. 4.5B).

126

Figure 4.5: Overlapping requirements for Ski7 and Dom34/Hbs1 during oxidative stress conditions. (A) Protein was isolated from the WT, ski7, dom34, hbs1, ski7 dom34, ski7 hbs1 and ski2 mutant strains expressing Protein A-nonstop protein and analysed by Western blotting. Blots were probed with a Protein A antibody (NSD) and an actin antibody as a loading control. Quantification is shown for Protein A concentrations relative to actin. Data are the means of three independent biological repeats ± SD. (B) Quantitative RT-PCR of Protein A-nonstop mRNA levels. Quantification is shown for Protein A mRNA concentrations relative to actin mRNA. Data are the means of three independent biological repeats ± SD. 127

4.2.5 Oxidative stress increases widespread aggregation of Sup35 in dose- dependent manner and potentially generates NSD substrate

NSD recognizes mRNAs where ribosomes translate into the 3’-poly(A) tail.

This can arise due to processing errors introducing a premature polyadenylation signal within mRNA coding regions or due to mutations that alter normal stop codons and their recognition (Frischmeyer et al., 2002; Klauer and van Hoof, 2012; van Hoof et al., 2002). There is no evidence at present to suggest that oxidative stress promotes premature polyadenylation. It is not thought that mutations in stop codons would routinely generate NSD substrates since frequent in-frame stop codons are found in the 3’UTRs of eukaryotic mRNAs (Frischmeyer et al., 2002). We therefore considered that conditions which promote nonsense suppression might cause readthrough of multiple mRNA stop codons effectively generating NSD substrates.

Oxidative stress conditions are known to alter translation efficiency and have been shown to promote misreading including translational readthrough of stop codons (Gerashchenko et al., 2012; Katz et al., 2016; Schosserer et al., 2015). The possible mechanisms underlying this increase in readthrough during oxidative stress conditions are largely unknown. One possibility is that oxidative stress conditions target the translation termination machinery, since reducing its efficiency would increase the frequency of stop codon readthrough. The eRF3 (Sup35) from S. cerevisiae is well known for its ability to form prion aggregates known as [PSI+]

(Wickner, 1994). Formation of Sup35 amyloid aggregates sequesters it away from its normal function in translation termination and elevated readthrough of termination codons is therefore a well-known phenotype of the yeast [PSI+] prion. However,

[PSI+] formation is an extremely rare event occurring at a frequency of approximately

-5 5 x 10 during normal growth, with a 10-fold increase observed in response to

128

oxidative stress conditions (Doronina et al., 2015). Thus, rare prion formation would not cause a measurable increase in stop codon readthrough in a population of yeast cells. We therefore asked whether aggregation of Sup35 occurs in response to oxidative stress conditions, which would deplete Sup35 from the soluble fraction, increasing nonsense suppression, and therefore potentially generating NSD substrates.

Cellular protein aggregation was analysed using a biochemical approach which separates insoluble proteins from soluble proteins by differential centrifugation, and removes any contaminating membrane proteins using detergent washes (Jang et al., 2004; Koplin et al., 2010; Rand and Grant, 2006; Tomoyasu et al., 2001). In order to show protein aggregation of insoluble fractions comparably with both total and soluble fractions in one gel, as much as four-times of the insoluble fraction as loaded relative to the total or soluble fractions to aid visualization in Fig. 4.6A and Fig. 4.6B. The global levels of protein aggregation were low during control, non-stress conditions, but increased in response to treatments with 0.5 mM or 1 mM H2O2 (Fig. 4.6A). The glycolytic enzyme Pgk1 was used as a negative control and minimal amounts of Pgk1 were detected in the insoluble fraction during normal or stress conditions (Fig. 4.6B). Western blot analysis of Sup35 revealed that a small fraction of Sup35 was detected in the insoluble fractions during non-stress conditions. This proportion significantly increased in response to H2O2 treatment such that approximately 20% of Sup35 was present in the insoluble fractions (Fig. 4.6C). As an alternative means to confirm Sup35 aggregation in response to oxidative stress, we examined the distribution of Sup35-GFP using fluorescence microscopy (Fig. 4.6D).

129

Figure 4.6: Aggregation of Sup35 in response to hydrogen peroxide stress. (A) Protein aggregates were isolated from the WT strain grown to exponential phase

(A600 ~0.4-0.6) prior to H2O2 treatment of 0.5 mM (+) or 1 mM (++) for 6 hours and analyzed by SDS-PAGE and silver staining. T, total cell extracts; S, soluble fraction; I, insoluble aggregate fraction. Four-times as much of the insoluble fraction was loaded relative to the total or soluble fractions to aid visualization. (B) The same protein samples as for panel A were analysed by Western blotting using antibodies that recognize Sup35 and Pgk1 as a loading control. (C) Quantification of Sup35 aggregation (percentage of total) as shown in panel B from triplicate experiments ± SD. (D) Sup35-GFP or cells expressing GFP alone was visualized in the WT strain exposed to 1 mM H2O2 for six hours. Examples of cells containing visible Sup35- GFP puncta are shown. The percentage of cells containing visible puncta was quantified from three independent biological repeat experiments ± SD.

130

Diffuse cytoplasmic Sup35-GFP fluorescence was observed in all cells during non- stress conditions. However, following a 1 mM H2O2 treatment, approximately 8% of cells were found to contain visible fluorescent puncta consistent with Sup35 aggregation. Finally, a yeast strain expressing GFP alone was used as a negative control to confirm that aggregation is due to Sup35 rather than GFP itself following 6 hours of H2O2 treatment (Fig 4.6D).

4.2.6 Overexpression of Sup35 rescues oxidant sensitivity in WT and SKI complex mutant strains.

Given that Sup35 aggregates during oxidative stress conditions, we next examined whether increasing the cellular concentration of Sup35 could rescue oxidant sensitivity. Cells were grown for 3 days in SRaf media and spotted onto media containing various concentrations of hydrogen peroxide supplied with 1% raffinose/galactose or 2% glucose as a control. For these experiments, Sup35 was expressed under the control of the GAL1 promoter while a shuttle vector pRS416 was used as a negative control. The oxidant sensitivity of WT and ski mutant strains were examined using spot test assays.

In the absence of hydrogen peroxide, the WT and ski complex mutant strains appeared to grow similarly in SRaf/SGal media. Overexpression of Sup35 was found to improve the H2O2 tolerance of both WT and ski2 mutant strains (Fig. 4.7A).

Similarly, overexpression of Sup35 rescued the oxidant tolerance of ski3 and ski8 mutants (Fig. 4.7B). Taken together, these data indicate that loss of Sup35 activity may cause oxidant sensitivity in WT and the ski complex mutants.

131

A. Control 3 mM H2O2 V WT glucose Sup35

V ski2 Sup35

V galactose WT Sup35

ski2 Sup35

B. Control 2.25 mM H2O2

V WT Sup35 glucose V ski3 Sup35

V ski8 Sup35

V WT Sup35 galactose

V ski3 Sup35

V ski8 Sup35

Figure 4.7: Overexpression of SUP35 rescues oxidant sensitivity in WT and SKI complex mutant strains. Sensitivity to oxidative stress was determined by spotting strains onto SGlu or SGal media containing various concentrations of hydrogen peroxide. Results are shown after three days growth for (A) WT and ski2 mutant strains containing a vector that has no SUP35 overexpression or by overexpressing SUP35 at 3 mM H2O2 and (B) WT, ski3 and ski8 mutant strains containing an empty vector or overexpressing SUP35 at 2.25 mM H2O2. 132

4.2.7 Overexpression of Sup35 decreases stop codon readthrough during oxidative stress conditions

Three different assays were used to examine whether oxidative stress increases the readthrough of stop codons, and if so, whether it can be rescued by the overexpression of Sup35. The levels of termination codon readthrough were first measured using a -galactosidase reporter system (Stansfield et al., 1995). Stop codon readthrough was quantified using plasmid pUKC819 (which contains the lacZ gene with a premature UGA termination codon) and expressed as a proportion of control -galactosidase levels, measured in transformants carrying the control plasmid pUKC815 (which contains the WT lacZ gene).

Firstly, we observed higher level of stop codon readthrough in response to

H2O2 stress in both strains (Fig. 4.8 A). This is consistent with previous findings that showed an elevation of readthrough in antioxidant mutants following H2O2 treatment

(Sideri et al., 2010). An increase of stop codon readthrough of approximately 2-fold was observed in the ski2 mutant strain relative to the WT strain, while overexpression of Sup35 significantly lowered UGA termination codon readthrough in both the WT and ski2 mutant strains during both normal and oxidative stress conditions (Fig. 4.8A).

Possible concerns have been suggested when using this assay system since it uses two separate reporter plasmids to quantify readthrough. For example, any factor that specifically has a direct impact on mRNA stability and/or translation initiation of the PTC-mRNA would possibly create artefacts. This is especially worrying when working with factors known to impact mRNA stability such as SKI2 and in conditions supposed to impact translation initiation such as oxidative stress.

133

Figure 4.8: Overexpression of SUP35 rescues stop codon readthrough in WT and ski2 mutant strains following hydrogen peroxide treatment. Strains were grown to exponential phase (A600 ~0.4-0.6) in 1% of SRaf/Gal media and treated with

0.5 mM H2O2 for 6 hours. Readthrough was quantified using a plasmid that carries: (A) the lacZ gene with a premature UGA termination codon (Stansfield et al., 1995). Values shown are means ± SD from three independent determinations. Readthrough is expressed as a percentage of control -galactosidase levels measured using WT lacZ gene; or (B) tandem Renilla and firefly luciferase genes separated by a single UGA stop codon (Keeling et al., 2004). Values shown are means ± SD from three independent determinations. Readthrough is expressed as a percentage of Firefly/renilla activity.

134

Therefore, a dual reporter construct which contains tandem Renilla and firefly luciferase genes separated by a single UGA stop codon was used to confirm our results (Keeling et al., 2004). Unlike the -galactosidase reporter plasmids that uses two plasmids and therefore may be subject to fluctuations, the dual luciferase system is thought to more accurately represent the efficiency of translation termination since both Renilla and luciferase genes have the same AUG codon to initiate translation.

Nevertheless, similar trends were observed with the dual luciferase systems; readthrough was elevated in both WT and ski2 mutant strains during oxidative stress conditions (Fig. 4.8B). Overall, both results support the hypothesis that overexpression of Sup35 rescues oxidant sensitivity by suppressing stop codon readthrough.

As an alternative approach, we used the Protein A construct containing an in- frame termination codon (Fig. 4.3A). We reasoned that if readthrough of the sole stop codon in this construct was increased in response to H2O2 stress, then ribosomes would translate through to the 3’-end of the Protein A mRNA resulting in the generation of an NSD substrate. In agreement with this idea, Protein A production was decreased in the WT strain in response to oxidative stress, whereas

Protein A production was unaffected in a ski2 mutant consistent with NSD accounting for the loss of Protein A following oxidant exposure (Fig. 4.9).

Overexpression of Sup35 prevented the oxidant-induced decrease in Protein A production observed in the WT strain (Fig. 4.9). Western blot analysis confirmed that

Sup35 was similarly overexpressed in the WT and ski2 mutant strains (Fig. 4.9). This in consistent with the idea that loss of Sup35 activity during oxidative stress conditions results in stop codon readthrough and the generation of NSD substrates.

135

Figure 4.9: Overexpression of SUP35 rescues hydrogen peroxide stop codon readthrough. Protein was isolated from both WT and ski2 mutant strains expressing the Protein A-stop construct and containing either a vector that has no SUP35 overexpression or a vector that is overexpressing SUP35. Strains were grown in

SRaf media until exponential phase (A600 ~0.4-0.6). 1% of fresh galactose was then added and strains were treated in the presence or absence of 0.5 mM H2O2 for 6 hours. Western blots were probed with a Protein A antibody and a Pgk1 antibody as a loading control. Blots probed with an anti-Sup35 antibody to confirm overexpression of Sup35. Quantification is shown for Protein A concentrations relative to Pgk1. The data are representative of three independent experiments ± SD.

136

4.3 Discussion

In this Chapter, we examined the requirement for the nonstop decay pathway during oxidative stress conditions and found that mutants lacking components of the

SKI complex or factors required for recognition of NSD substrates are required for oxidant tolerance. We found an overlapping requirement for Ski7, which bridges the interaction between the Ski complex and the exosome, and NGD components

(Dom34/Hbs1) which have been shown to function in both NSD and NGD. We showed that ski7 dom34 and ski7 hbs1 mutants are sensitive to hydrogen peroxide stress and accumulate an NSD substrate. We further showed that NSD substrates are generated during ROS exposure as a result of aggregation of the Sup35 translation termination factor, which increases stop codon readthrough allowing ribosomes to translate into the 3’-end of mRNAs. This suggests that the recognition and degradation of NSD substrates is required for oxidant tolerance. The exosome is the main cellular nuclease which catalyses 3’-5’ mRNA degradation and previous studies have shown that mutants deficient in its exoribonuclease activity are sensitive to oxidative stress conditions (Tsanova et al., 2014). These data therefore suggest that the inability to recognize and degrade NSD substrates causes sensitivity to oxidative stress, presumably as a result of the translation and production of aberrant proteins from NSD mRNAs.

The mechanism of NSD is conserved among eukaryotes with the exception of

Ski7 which is only found in a small subset of yeasts and not in higher eukaryotes

(Atkinson et al., 2008). Ski7 bridges the interaction between the Ski complex and the exosome (van Hoof et al., 2002). Ski7 binds to ribosomes stalled at the 3′-end of mRNAs where it recruits the exosome to trigger 3′-to-5′ degradation of NSD substrates. Ski7 is closely related to Hbs1, which is another conserved Sup35 family

137

member required for NGD and it is thought that Hbs1 can function in both NSD and

NGD (Shoemaker and Green, 2012). The Dom34-Hbs1 complex binds to the ribosomal A site and promotes the dissociation of subunits on stalled ribosomes

(Anderson and Parker, 1998; Dziembowski et al., 2007). This is important for RNA quality control in NGD targeting aberrant mRNAs for degradation by the exosome

(Doma and Parker, 2006). The Dom34-Hbs1 complex is also thought to play a role in

NSD by facilitating the degradation of mRNAs where ribosomes are stalled at the 3′ end of mRNAs lacking a termination codon (Tsuboi et al., 2012). This is because the translation of poly(A) tails into poly-lysines following stop codon readthrough can cause stalling analogous to NGD (Ito-Harashima et al., 2007). The Hbs1-Dom34 complex has also been directly shown to function in NSD in mammalian cells (Saito et al., 2013). We used a Protein A-nonstop reporter construct (Wilson et al., 2007) to monitor NSD and found that Protein A production was higher in dom34 and hbs1 mutants, compared with a WT strain, consistent with a role for the Hbs1-Dom34 complex in the turnover of NSD substrates. Furthermore, Protein A production was increased in ski7 hbs1 and ski7 dom34 mutants compared with the single parent mutants in agreement with the idea that Ski7 and Hbs1-Dom34 play a redundant role in the recognition of NSD substrates.

Most studies on NSD have used artificial reporter constructs similar to the

Protein A nonstop reporter construct (Wilson et al., 2007) used in our current study since relatively little is known regarding physiological NSD substrates. The most likely source of NSD substrates is considered to arise from 3’-end processing signals occurring within gene coding regions causing premature 3’-end cleavage and processing (Frischmeyer et al., 2002). Whilst this would occur during normal growth conditions, it is also possible that oxidative stress conditions might cause cleavage

138

and mRNA truncation events similar to defective 3’-end processing events. Stop codon readthrough has also been considered as a potential source of NSD substrates, but the prevalence of in-frame stop codons in the 3’UTRs of most mRNAs means that it was considered unlikely as a major source of NSD substrates

(Frischmeyer et al., 2002). For example, in-frame stop codons are significantly over represented downstream of normal ORF stop codons (Atkinson et al., 2008; Williams et al., 2004). We hypothesised that stress conditions, such as oxidative stress, which reduce the fidelity of stop codon recognition, might generate NSD substrates by causing ribosomes to readthrough stop codons into the 3’-end of mRNAs. Although stop codon readthrough is a relatively rare event, the production of even small amounts of aberrant proteins due to stabilization of carboxyl-terminal extended proteins may be sufficient to cause toxicity. The addition of C-terminal amino acids to just one key protein could potentially alter its biological function potentially resulting in toxicity. For example, previous studies have identified cases where readthrough into 3’UTRs generates aberrant and aggregated proteins which can cause toxicity

(Atkinson et al., 2008; Hoshino, 2012; Inada and Aiba, 2005). NSD may therefore be particularly important to protect against protein production under oxidative stress conditions since oxidative stress can promote stop codon readthrough and the resulting translation into the 3’UTR of key mRNAs may produce toxic and aberrant proteins.

When a stop codon is translocated into the ribosomal A-site, it is recognised by eukaryotic release factor 1 (eRF1) (von der Haar and Tuite, 2007). eRF1 activates hydrolysis of the ester bond between the completed polypeptide chain and the tRNA in the ribosomal P-site. eRF3 is a GTPase that associates with eRF1 and is essential for the termination reaction. Yeast Sup35 is well known for its ability to

139

form prion aggregates known as [PSI+], which sequesters it away from its normal function in translation termination and elevated readthrough of termination codons is therefore a well-known phenotype of the yeast [PSI+] prion (Wickner, 1994).

However, [PSI+] formation is an extremely rare event occurring at a frequency of approximately 5 x 10-4 following exposure to hydrogen peroxide (Doronina et al.,

2015). Hence, the rare formation of [PSI+] in response to oxidative stress cannot account for aggregation and significant loss of Sup35 activity. Instead, we found that oxidative stress promotes aggregation of Sup35, which analogous to Sup35 amyloid formation, titrates Sup35 away from its normal function in translation termination.

This was a relatively frequent event since approximately 20% of Sup35 was sequestered into an insoluble aggregated form in response to hydrogen peroxide stress. A number of growth conditions can cause protein misfolding and aggregation including advanced age and environmental stresses such as oxidative stress

(O'Connell et al., 2012; Vendruscolo, 2012). Highly abundant proteins such as

Sup35 are prone to aggregation and it is thought that stress conditions such as oxidative stress cause aggregation by lowering the threshold for aggregate formation

(Ibstedt et al., 2014; Weids et al., 2016). Hence, alterations in the fidelity of translation termination due to protein aggregation of Sup35 may be an unanticipated consequence of ageing and various stress conditions

We found that overexpressing Sup35 restored termination efficiency in agreement with the idea that aggregation of Sup35 titrates it from its normal function in translation termination. Increased cellular concentrations of Sup35 also rescued the oxidant sensitivity of ski complex mutants. Taken together, our data indicate that

NSD acts to protect against protein production from mRNAs where Sup35- aggregation promotes a decrease in the fidelity of stop codon recognition. This

140

global increase in stop codon readthrough appears to be detrimental to cells, based on the increased cellular sensitivity to oxidative stress conditions that it causes.

However, it has previously been suggested that the shift to the [PSI+] prion state may provide a mechanism for generating heritable phenotypic diversity by allowing cells to reprogram gene expression such that new genetic traits become uncovered which aid survival during altered or unfavourable conditions (Eaglestone et al., 1999, True and Lindquist, 2000; Tyedmers et al., 2008). Most phenotypic alterations are thought to arise due to changes in translation termination efficiency and not other properties such as protein aggregation (True et al., 2004), or to produce polypeptides with C- terminal extensions e.g. polyamines (Namy et al., 2008). It is therefore possible that protein aggregation drives similar phenotypic plasticity and mRNA-specific alterations in stop codon readthrough may be beneficial during particular growth conditions.

The evolutionarily conserved Ski complex functions in many cytoplasmic exosome-mediated pathways including 3’-5’-mRNA degradation, NSD and NMD

(Araki et al., 2001; He et al., 2003; Houalla et al., 2006; van Hoof et al., 2002; van

Hoof et al., 2000). Loss of Hbs1 or Dom34 in a ski7 mutant increased oxidant sensitivity suggesting that an overlapping role for Ski7 and the Dom34-Hbs1 complex in recognising NSD substrates is required for oxidant tolerance. However, hbs1 ski7 and dom34 ski7 mutants were not as sensitive as a ski2 mutant to hydrogen peroxide stress suggesting that oxidant sensitivity might not solely arise in a ski complex mutant due to the inability to recognize and degrade NSD substrates.

It is also known that mRNA turnover is regulated in response to oxidative stress conditions and hence disrupting this process may also account for some oxidant sensitivity (Molina-Navarro et al., 2008; Marguerat et al., 2014). Upf1, a key regulator

141

of NMD, has also been shown to be required for the transcriptional induction of many oxidative stress-regulated genes in the fission yeast Schizosaccharomyces pombe and upf1 strains are sensitive to H2O2 stress (Rodriguez-Gabriel et al., 2006).

Further work will be required to understand the hypersensitivity of SKI complex mutants to oxidative stress. This will be important since RNA metabolism has many disease links and mutations in human exosome subunit genes have been linked with childhood-onset neurological diseases (Muller et al., 2015).

142

5.0 Importance of mRNA quality control for proteostasis

5.1 Introduction

The primary structure of proteins consists of a linear sequence of amino acid residues known as polypeptide chains. However, most proteins must adopt a three- dimensional (3D) structure in a process known as protein folding in order to be functional and participate in virtually all process within cells. Cells therefore have several mechanisms including PQC machineries, usually assisted by a number of molecular chaperones, that have evolved to maintain correct folding as well as degradation pathways that protect cells from aggregated proteins (Pechmann et al.,

2013). However, accurate protein folding is often regarded as a challenging task as mistakes may arise on several occasions, including: 1) wrong incorporation of amino acid during protein synthesis; 2) post-translational modification errors including wrong assembly of protein complexes; 3) changes that affect cellular conditions such as pH, oxidative stress, or temperature; or 4) a decline in effectiveness of PQC machineries due to ageing (Hartl et al., 2011; Stefani and Dobson, 2003; Tyedmers et al., 2010). Such errors may greatly affect protein homeostasis (proteostasis) leading to accumulation of misfolded and aggregated proteins that are often linked to conformational diseases (Chiti and Dobson, 2006). These aggregates are characterized either as a highly ordered structured amyloids enriched with -sheets or protein aggregates that unlike amyloids, do not have any defined structure to date.

Understanding protein folding mechanisms remains one of the most fervently pursued goals for researchers. Much of the interest in solving the protein folding problem stems from the fact that many diseases result from misfolded proteins

(Valastyan and Lindquist, 2014). A common example from this category includes cystic fibrosis, a loss-of-function type mutation which results in an altered protein

143

structure. The other main class of folding disorders involves toxic gain-of-function that has been implicated in several catastrophic diseases including neurodegenerative diseases (AD, PD, HD), prion diseases, ALS, and type 2 diabetes, which has been a focus of ongoing research for several decades

(Valastyan and Lindquist, 2014). As outlined in Section 1.7 however, many studies have reported that protein aggregation not only occurs in higher eukaryotes, but also exists in lower organisms such as unicellular fungi or bacteria (Morell et al., 2008;

Ventura and Villaverde, 2006). Interestingly, the aggregates formed in all of these organisms are classified together as amyloids (Carrió et al., 2005; Morell et al., 2008;

Wang et al., 2008), indicating that this very organized structure predominantly enriched in β-sheet protein is considered as a threat in maintaining correct folding, regardless of whether it is proving to be beneficial or toxic to the host (Dobson, 2001;

Rousseau et al., 2006).

As described previously, mRNA quality control mechanisms recognize stalled ribosomes on abnormal mRNAs resulting from nonsense, nonstop, or mRNAs impaired during translation elongation. However, the importance that mRNA surveillance pathways might have in regulating proteostasis is currently not well- understood. The aim of the studies reported in this Chapter was to determine whether loss of NSD, NMD or NGD, would have any effects on proteostasis by examining both amorphous and amyloid aggregation. We hypothesised that any defects in mRNA quality control systems may lead to stabilization of aberrant transcripts which in turn could lead to production of aberrant proteins that tend to misfold and form aggregates.

144

5.2 Results

5.2.1 Loss of mRNA surveillance mutants causes widespread protein aggregation

To examine whether protein aggregation occurs following loss of mRNA surveillance pathways, we used an established biochemical approach which separates insoluble proteins from soluble proteins by differential centrifugation, and removes any contaminating membrane proteins using detergent washes (Jang et al.,

2004; Koplin et al., 2010; Rand and Grant, 2006; Tomoyasu et al., 2001). Aggregates were prepared from cultures of the WT strain and mutants deficient in NSD (ski7, ski8), NMD (upf1, upf2), and NGD (dom34, hbs1) grown from OD600 0.01 for 24 hours prior to protein purification. Proteins in aggregate preparations were separated using SDS-PAGE and visualized by silver-staining. Aggregates were prepared from three repeated experiments and a representative gel is shown in Fig. 5.1. As indicated by the increased intensity of the bands, elevated protein aggregation was observed in all of the mRNA surveillance pathways compared with the WT strain

(Fig. 5.1).

Next, we monitored sites of protein aggregation in mRNA surveillance mutants using fluorescently-tagged chaperone Hsp104 protein disaggregase (Erjavec et al.,

2007; Lee et al., 2010).Hsp104-RFP is normally observed as diffuse cytoplasmic fluorescence in cells, but its accumulation at the sites of protein aggregation means that it can be used as an in situ marker to visualize protein aggregation (Lee et al.,

2010; Lum et al., 2004). We found that diffuse cytoplasmic Hsp104-RFP fluorescence was detected in the majority of WT cells examined and only 3% of cells contained visible puncta marking sites of protein aggregation

145

Figure 5.1: Strains lacking components of mRNA surveillance pathways have higher levels of protein aggregation. (A) Examples of protein aggregates isolated from the WT and mRNA surveillance mutant strains involved in NSD (ski7, ski8), NMD (upf1, upf2), and NGD (dom34, hbs1) followed by SDS-PAGE and visualized by silver staining. (B) Hsp104-RFP was visualized in (A).The percentage of cells containing visible Hsp104-RFP puncta is quantified for each strain. Data shown are the means of three repeated experiments ± SD. Unpaired t-test was used to determine statistical significance between WT and mRNA surveillance mutant strains (*p < 0.05, **p < 0.01, p*** < 0.001). (C) Protein aggregates were isolated from the same strains as shown in panel A and analyzed by Western blot using an Hsp104 antibody while the Pgk1 antibody was used as a loading control.

146

(Fig. 5.1B). In contrast, more puncta were observed in all of the mRNA surveillance mutant strains (Fig. 5.1B). The number of Hsp104-RFP puncta detected was elevated approximately 3-fold in NSD (ski7, ski8) and NGD (dom34, hbs1) mutant strains while a much larger increase was observed in NMD (upf1, upf2) mutants, where approximately 16-18% of cells examined contained visible Hsp104 puncta.

Finally, western blot analysis was used to confirm that similar protein levels of

Hsp104 were observed in all of the mRNA surveillance mutants and the WT strains, while the glycolytic enzyme Pgk1 was used as a loading control (Fig. 5.1C).

5.2.2 More proteins are susceptible to aggregation following loss of mRNA surveillance mutant strains

The proteins that aggregate following loss of mRNA surveillance pathways were identified using mass spectrometry. We focused on five out of the initial seven strains given that a similar pattern of increased aggregation was observed in all of the mRNA surveillance mutant strains (Fig. 5.1). This included a single mutant from each of the NSD, NMD and NGD pathways to assess their protein composition. We also included the ski8 mutant to further elucidate the difference between Ski7 and

Ski8 as previously described in Chapter 3 and 4. Similarly, HBS1 was chosen due to its functional redundancy with SKI7 as previously observed in Chapter 4. Therefore, proteins that aggregate from the WT, ski7, ski8, upf2, and hbs1 mutant strains were identified using liquid chromatography-mass spectrometry (LC-MS) and only considered as significant if they were identified in at least two out of three independent experiments.

147

Figure 5.2: Overlap of identified proteins within protein aggregates isolated from WT and mRNA surveillance mutant strains. Proteins within insoluble aggregate fractions as seen in Fig. 5.1 were identified by LC-MS. (A) Venn diagrams showing a comparison of proteins in aggregate fractions from the WT (blue) and ski7 (red), ski8 (pink), upf2 (green), and hbs1 (orange) mutant strains, respectively. (B) The Venn diagram shows the overlaps between the proteins aggregating in all five strains identified as common aggregates (198) as well as aggregates specific to each strain.

148

Our mass spectrometry analysis identified 246, 428, 399, 429, and 349 of proteins which aggregate in the WT, ski7, ski8, upf2, and hbs1 mutant strains, respectively (Appendix tables 1-5; Fig. 5.2A). The aggregated proteins from each mutant were individually compared with the WT strain which showed that there is generally a large overlap in the proteins that aggregate in both the WT and mutant strains. Additionally, only 3-9% of protein aggregates were specifically found in the

WT strain while 35-45% of protein aggregates were specific to each of these mRNA surveillance mutants (ski7=45%; ski8=42%; upf2=41%; hbs1=35%; Fig. 5.2A).

Similarly, the number of protein aggregates uniquely observed in each mutant strain was also increased in comparison to the WT strains (WT, n=5; ski7, n=42; ski8, n=36; upf2, n=42; hbs1, n=11; Fig. 5.2B). However, there was a large overlap in the proteins which aggregate in all five strains (198 proteins) indicating that the majority of proteins do not aggregate in a mutant-specific manner (Fig. 5.2B). Taken together, the pronounced increase observed in the number of proteins that aggregate in the ski7, ski8, upf2, and hbs1 mutant relative to the WT strain suggests that loss of mRNA surveillance mutants disrupts the integrity of proteostasis. A direct comparison across all five strains indicates that as well as mutant-specific aggregation, a number of common proteins tend to aggregate in all strains suggesting that loss of mRNA surveillance pathways increases the abundance of these proteins in aggregate fractions compared with the WT strain.

5.2.3 Differences in subcellular localization of proteins isolated from aggregates in mRNA surveillance mutants

Given that the function of a protein is generally related to its subcellular localization, we next examined the subcellular localization of the aggregated proteins

149

Figure 5.3: Subcellular localization prediction of aggregated proteins identified in WT and mRNA surveillance mutant strains. Histogram showing the relative localisation of the proteins present in the aggregates isolated from the WT (246), ski7 (428), ski8 (399), upf2 (429) and hbs1 (349) mutant strains. Proteins with annotated localisation for each cellular component were compared to the total proteins with annotated localisation in each data set.

150

in the WT and mRNA surveillance mutant strains. No major differences were observed in the proportion of aggregated proteins localized to different compartments in the different strains. The major fraction of proteins examined were predicted to localise in the cytoplasm of all strains examined (Fig. 5.3). Proteins were also predicted to localise to the nucleus, mitochondrion, ER, Golgi vacuole, cytoskeleton and transport vesicles.

5.2.4 Enrichment of functional categories within protein aggregates isolated from mRNA surveillance mutants

We next analyzed the datasets for any enrichment of functional categories to determine whether particular biological processes may be affected by protein aggregation following loss of mRNA surveillance mutants. Significant functional enrichment (false discovery rate (FDR) < 5%) within the datasets was determined using the MIPS Functional Catalogue (Ruepp et al., 2004). Proteins present in the aggregates isolated from the WT (246), ski7 (428), ski8 (399), upf2 (429) and hbs1

(349) mutant strains were compared (Fig. 5.4).

Our analysis revealed that the proteins within these aggregates could generally be grouped into two major overarching categories: protein synthesis and proteins with binding function or cofactor requirement (Fig. 5.4). Meanwhile, additional enrichments were observed in the functional categories within protein aggregates isolated from ski7, ski8, upf2 and hbs1 mutant strains. In particular, stress categories including the heat shock response and the unfolded protein response were enriched in the mRNA surveillance mutant strains but not in the WT dataset.

151

Figure 5.4: MIPS functional categorization of aggregated proteins identified in WT and mRNA surveillance mutant strains. Significantly enriched functional categories within the data-sets were determined using FunCat (FDR < 5%). Results are ordered based on MIPS category classification numbers and overarching categories are in capitals. Confidence of each classification category is shown as Bonferroni corrected p-values.

152

Taken together, these data indicate that the aggregates isolated mRNA surveillance mutants are enriched for several categories that are similarly enriched in the WT-set.

However, a number of functional categories were identified which are previously absent from WT dataset, suggesting that some functional categories may be affected by aggregation in mRNA surveillance-mutant specific manner.

5.2.5 Analysis of the physicochemical properties of aggregated proteins identified in mRNA surveillance mutants

We next assessed the physicochemical properties of the aggregated proteins within our datasets to determine whether they possess particular properties which make them aggregation-prone including: abundance measured by the number of molecules/cell, protein expression levels as indicated by the codon adaptation index

(CAI), protein translation rate (the number of growing proteins/second), protein size as indicated by its molecular weight (kDa), hydrophobicity (measured by the GRAVY score- a computer program used to calculate the average hydropathy of a protein based on its amino acid sequence), isoelectric point (pI) and protein stability as indicated by its half-life (measured in minutes).

For comparison, a list of yeast proteins detected by mass spectrometry in logarithmically growing cells, referred to as the MS-set, was used to represent the properties of unaggregated proteins (Washburn et al., 2001). For this analysis, the

Common-set consisting of 198 proteins which aggregate in all strains were compared with the proteins which were found to specifically aggregate in the mRNA surveillance mutant strains (ski7, n=199; ski8, n=173; upf2, n=198; hbs1, n=133)

(Fig. 5.2A).

153

The one-way analysis of variance (ANOVA) was used to determine whether there are any statistically significant difference between the means of the Common- set with the rest of mRNA surveillance mutant datasets using box-and-whisker plot where the whiskers were drawn between the smallest (5th) to largest (95th) percentiles while individual values below and above the whiskers were drawn as individual dots superimposed on the graph. The box plot was set from the lowest

(25th) to highest (75th) percentiles which are called the hinges of the plot while the middle of the box tells the median of the plot.

5.2.5.1 Aggregated proteins are enriched for abundant and highly expressed proteins

Analysis of aggregated proteins in the Common-set revealed that they are significantly more abundant and highly expressed when compared with the MS-set

(Fig. 5.5A-B). Similarly, highly expressed and abundant proteins were also significantly enriched in the aggregate fractions within mRNA surveillance mutant strains (Fig. 5.5A-B). The mRNA translation rate of aggregated proteins was next compared with a global estimation of translation rates database generated from quantitative microarray analysis (Arava et al., 2003). The translation rates of proteins in the Common-set were elevated compared with the MS-set correlating with their high protein abundances (Fig. 5.5C). The translation rates of the aggregated proteins identified in the upf2 mutant was not significantly different to the MS-set although their translation rates appeared to be somewhat shifted to a faster rate (Fig. 5.5C).

Finally, the stability of proteins within our datasets were compared with a list of values obtained under basal conditions in S. cerevisiae (Christiano et al., 2014).

154

Figure 5.5: Proteins within aggregates are abundant, stable, and highly expressed. Aggregated proteins within our proteins were compared with a list of unaggregated proteins identified by mass-spectrometry and referred to as MS-set (Washburn et al. 2001). (A-D) The following properties were compared within each dataset -protein abundance (molecules/cell), gene expression level as indicated by CAI, translation rates per protein generated from ribosome profiling data (Arava et al. 2003), and protein stability (Christiano et al. 2014). The Mann–Whitney U-tests were used to assess the statistical significance of observed differences: *p < 0.05, **p < 0.01, p*** < 0.001.

155

In general, proteins within the Common-set as well as aggregates isolated from the mRNA surveillance mutant strains were enriched for stable proteins compared with the MS-set (Fig. 5.5D). Taken together, these data suggest that aggregated proteins within the Common-set tend to be highly abundant, highly expressed as well as more stable than the unaggregated proteins. Although similar observation is also observed in mRNA surveillance mutant datasets, their abundance, translation rates and protein stability are somewhat reduced when compared to the Common-set.

5.2.5.2 Aggregated proteins are more hydrophobic than the unaggregated ones

To further define the properties of the aggregated proteins identified in mRNA surveillance mutants we examined a number of physicochemical proteins including protein size (kDa), hydrophobicity based on the GRAVY score and isoelectric point

(pI). No significant differences were observed in terms of the molecular sizes

(molecular weight) (Fig. 5.6A) as well as the pI’s (Fig. 5.6C) in the Common-set when compared with the MS-set. In contrast, proteins within the Common-set were significantly more hydrophobic (Fig. 5.6B) than the MS-set.

A number of differences were observed in the physicochemical properties of the aggregated proteins identified in mRNA surveillance mutants compared with the

Common and MS-datasets. For example, protein aggregates isolated from both ski7 and ski8 mutants were enriched for proteins which were more hydrophobic (Fig.

5.6A) and had lower pI’s (Fig 5.6B) compared with the MS-set while protein aggregates in both upf2 and hbs1 mutants were only significantly enriched for hydrophobicity (Fig. 5.6A). Taken together, proteins identified in aggregates from mRNA surveillance mutant strains share a number of common biophysical

156

Figure 5.6: Analysis of physicochemical properties within protein aggregates.

Aggregated proteins within our proteins were compared with a list of unaggregated proteins identified by mass-spectrometry and referred to as MS-set (Washburn et al.

2001). (A-C) The following physicochemical properties were compared within each dataset- molecular weight, hydrophobicity based on GRAVY score, and isoelectric point (pI). The Mann–Whitney U-tests were used to assess the statistical significance of observed differences: *p < 0.05, **p < 0.01, p*** < 0.001.

157

properties, although some specific differences were observed in particular mRNA surveillance mutants.

5.2.6: Analysis of selected amino acid composition in mRNA surveillance mutant aggregates

It is well-established that certain amino acids are influential in promoting aggregation. Most notably, glutamine (N) and asparagine (Q) enriched regions are important in driving aggregation of prions and other human disease related and amyloid forming proteins including HD (Michelitsch and Weissman, 2000).

Additionally, previous biochemical studies have noted that proteins with iterated positively-charged residues such as poly(lysine) and poly(arginine) have slower translation rates presumably due to stalled ribosomes (Lu and Deutsch, 2008). We therefore asked whether lysine, glutamine and asparagine residues are enriched in the protein aggregates isolated from mRNA surveillance mutants.

The relative amino acid composition of glutamine was somewhat higher in the

Common-set and mRNA surveillance mutant compared with the MS-set, although this was only statistically significant in the ski8 and upf2 mutants (Fig. 5.7A). This is consistent with similar study which found no enrichment for asparagine (Q)/ glutamine (N) content in the WT strain under basal and stress conditions (Weids et al. 2016). Similarly, no enrichment of asparagine was observed in any of our datasets compared with the MS-set (Fig. 5.7B). Finally, only the proteins identified in the Common-set were enriched for the positively charged amino acid lysine (Fig.

5.7C), which correlates with the enrichment for proteins with high pI-values observed within the Common-set (Fig. 5.6C).

158

Figure 5.7: Analysis of selected amino acid composition within protein aggregates. Aggregated proteins within our proteins were compared with a list of unaggregated proteins identified by mass-spectrometry and referred to as MS-set (Washburn et al. 2001). The following amino acid contents were compared: (A) glutamine (%), (B) asparagine (%), and (C) lysine (%). The Mann–Whitney U-tests were used to assess the statistical significance of observed differences: *p < 0.05, **p < 0.01, p*** < 0.001.

159

5.2.7 Molecular chaperones are present within protein aggregates

We next examined whether chaperones were isolated in our aggregate fractions, although it should be emphasized that our analysis does not distinguish between chaperones which are functional within aggregates, compared with chaperones which are aggregation-prone themselves. Based on the 63 known chaperones in S. cerevisiae (Gong et al., 2009), we identified a total of 23 chaperones distributed between all the datasets: one Hsp90, nine Hsp70s, four

Hsp40s, one AAA+ family member and eight chaperonin (Hsp60) subunits (Fig.

5.8). Out of these, 12 molecular chaperones spanning all five chaperone categories were identified in all strains: Hsp104, Hsc82, Cct2, Cct4, Ssa1, Ssa2,

Kar2, Ssb1, Ssc1, Sse1, Ssz1, and Ydj1. The biggest category of chaperones belong to Hsp70 family members (Ssa1, Ssa2, Kar2, Ssb1, Ssc1, Sse1, Ssz1) which are important in assisting a variety of protein folding processes and maintain cellular functions (Saibil, 2013). Overall, we observed that the total number of chaperones present in aggregates from mRNA surveillance mutants was generally higher than the wild type strain (WT, n= 14; ski7, n=21; ski8, n=18; upf2, n=22; hbs1, n=19; Fig. 5.8).

A total of seven chaperones belonging to chaperonins (Cct3, Cct5, Cct6,

Cct7, Cct8, Cct1) and Hsp70 (Sse2) were observed in the ski7-set. The chaperones present within upf2-set were similar to the ski7-set; the only exception was Sec63, a Hsp40 family member belonging to the upf2-set (Fig. 5.8). Sec63 is a part of the Sec61 protein complex which is thought to be the key component of post-translational protein translocation into the ER (Deshaies et al., 1991; Panzner et al., 1995). Finally, seven chaperones were observed in the hbs1-set; five of

160

Figure 5.8: Overlap of identified protein chaperones within protein aggregates isolated from WT and mRNA surveillance mutant strains. Proteins within insoluble aggregate fractions as seen in Fig. 5.1 were identified by LC-MS. The Venn diagram shows the comparison of chaperones within aggregates observed between the WT (blue) and ski7 (red), ski8 (brown), upf2 (yellow), and hbs1 (green) mutant strains, respectively.

161

these (Cct3, Cct5, Cct7, Cct8, Cct1) belong to the chaperonins family while the other two (Ssa3 and Sse2) belong to the Hsp70 family proteins (Fig. 5.8).

Interestingly, Ssa3 is uniquely present in the hbs1-set but it is currently unclear how this particular chaperone relates to the function of HBS1 itself (Fig. 5.8).

Taken together, chaperones which generally aggregate in the WT-set also appeared to aggregate in the mRNA surveillance mutant strains; however, there are several chaperones that specifically aggregate in the absence of mRNA surveillance pathways, highlighting their importance in mediating protein folding processes in mRNA surveillance mutant strains.

5.2.8 An increased frequency of de novo prion formation occurs in NMD mutants

Given the observed increase in protein aggregation detected in mRNA surveillance mutants, we next examined whether frequencies of prions, a type of amyloid aggregates, is similarly elevated in these mutants. The 74D-694 strain used for these studies contains a mutation in the adenine biosynthetic pathway which can be used to differentiate between [psi-] and [PSI+] strains (Chernoff et al., 1995; Serio and Lindquist, 1999). The 74D-694 strain contains an ade1-14 mutation and [psi−] strains are auxotrophic for adenine and appear red due to the accumulation red intermediate while in the [PSI+] strains, suppression of the ade1-14 nonsense mutation gives rise to white/pink Ade+ colonies due to the production of functional

Ade1 protein (Chernoff et al., 1995; Serio and Lindquist, 1999). The frequency of de novo [PSI+] prion formation was scored by counting white/pink Ade+ colonies following three days growth in minimal media. [PSI+] strains were differentiated from nuclear mutations by their irreversible elimination following growth in the presence of

162

GdnHCl (Tuite et al., 1981). GdnHCl effectively blocks the propagation of yeast prions by inhibiting the key ATPase activity of Hsp104, a molecular chaperone that functions as a disaggregase (Ferreira et al., 2001; Jung and Masison, 2001).

The frequency of [PSI+] prion formation detected in the WT strain was approximately 7x10-6, comparable to previous frequencies ranging between 10-5 and

10-7 (Fig. 5.9A) (Doronina et al., 2015; Lancaster et al., 2010; Lund and Cox, 1981;

Speldewinde et al., 2015). The frequency of de novo [PSI+] formation was significantly elevated by 4-fold (upf1) and 7-fold (upf2) in NMD mutants (Fig. 5.9A).

Interestingly, the frequency of de novo [PSI+] formation was significantly reduced following loss of NSD or NGD pathways, compared to the WT strain (Fig. 5.9A).

Next, the nuclear mutation rate in mRNA surveillance mutants was quantified by counting the formation of Ade+ colonies that are not curable with GdnHCl

(Fig. 5.10B). A previous study demonstrated that antioxidant mutants which showed high frequencies of de novo [PSI+] prion were often also observed to have higher nuclear mutation rates (Doronina et al., 2015). Similarly, we also observed significant increase in the frequency of nuclear mutations in NMD mutants (Fig. 5.9B).

Interestingly, the frequency of nuclear mutations was different in NSD mutants; the ski7 mutant showed an increased frequency comparable to NMD mutants, whereas, the ski8 mutant had the lowest frequency nuclear mutations (Fig. 5.9B). Finally, although nuclear mutations were formed at significantly lower frequencies

(approximately 10-fold) as opposed to the de novo [PSI+] prion formation, we concluded that no correlation was observed between these two (Fig. 5.9B).

Given the increased frequency of [PSI+] prion formation in NMD mutants, we next examined the de novo formation of the Rnq1/[PIN+] prion which is not related in

163

Figure 5.9: Loss of NMD factors (UPF1 and UPF2) result in an increased frequency of [PSI+] and [PIN+] prion formation. (A) The de novo [PSI+] prion formation was quantified in the WT, NSD, NMD, and NGD mutant strains by growth on medium lacking adenine, indicative of decreased translational termination efficiency. Data shown are the means of three repeated experiments expressed as the mean formation of [PSI+] per 106 cells. (B) True [PSI+] formation was differentiated from nuclear gene mutations that give rise to adenine prototrophy by their irreversible elimination in GdnHCl. Data shown are the means of three independent biological repeat experiments per 107 ± SD. (C) [PIN+] prion formation was scored in the WT, NSD, NMD, and NGD mutant strains and is expressed as the percentage of [PIN+] colonies formed per 96 colonies examined. Data shown are the means of three independent biological repeat experiments ± SD. (A-C) Unpaired t- test was used to determine statistical significance between WT and mRNA surveillance mutant strains (*p < 0.05, **p < 0.01, p*** < 0.001).

164

sequence to the Sup35/[PSI+] prion. The presence of a [PIN+] prion is absolutely required for the de novo formation of most yeast prions including [PSI+] (Derkatch et al., 2001; Sideri et al., 2011). To quantify [PIN+] formation, we used the Sup35NM-

GFP plasmid in [pin-][psi-] versions of all strains and scored the frequency of de novo

[PSI+] colonies based on the number of white/pink Ade+ colonies that are curable by

GdnHCl (Sideri et al., 2011). The frequency of [PIN+] formation was approximately

3% in the WT [pin-] strain (Fig. 5.9C), comparable to previous studies (Sideri et al.,

2011; Speldewinde et al., 2015). Consistent with the de novo [PSI+] results, an increased frequency of [PIN+] formation was also observed in NMD mutants, but not in NSD and NGD mutants (Fig. 5.9C).

5.2.9 The frequency of induced [PSI+] formation is increased in NMD mutants

Previous studies have shown that [PSI+] prion formation can be induced by the overexpression of SUP35NM-GFP in [PIN+][psi-] strains since the excess Sup35 increases the probability for prion seed formation (Wickner, 1994). We therefore tested whether induced [PSI+] prion formation is also elevated in NMD mutants. Cells with an initial OD600 of 0.01 were grown for 24 hours in the presence of copper to induce SUP35NM-GFP expression prior to visualization by microscopy.

Overexpression of SUP35NM-GFP resulted in detectable protein aggregates in 4.8% of WT cells examined (Fig. 5.10A), similar to previous studies (Speldewinde et al.,

2015). When the same experiment was repeated in the mRNA surveillance mutants, the number of visible aggregates comparable to the WT strain was either significantly increased (upf1, upf2), unaffected (ski7) or significantly decreased (ski8, hbs1, dom34) (Fig. 5.10A). Western blots were used to confirm that similar levels of Sup35 were expressed in each strain before and after copper addition (Fig. 5.10B),

165

Figure 5.10: Induction of [PSI+] prion formation is uniquely observed in NMD mutants. (A) Cells containing SUP35NM-GFP plasmid was visualized in WT, NSD,

NMD, and NGD mutant strains. Cultures were initially grown for 24 hours from OD600 0.01 in the presence of copper and representative images containing visible puncta are shown from an average of 100 cells counted in all strains. The percentage of cells containing visible SUP35NM-GFP puncta is quantified for each strain. (B) Western blot analysis of the WT and mRNA surveillance mutant strains using Sup35 or Pgk-1 antibodies as a loading control following 24 hours of copper induction. (C) [PSI+] prion formation was quantified in the WT and NMD mutant strains containing the Sup35NM-GFP plasmid following 24 hours of copper induction. (A, C) Unpaired t-test was used to determine statistical significance between WT and mRNA surveillance mutant strains (*p < 0.05, **p < 0.01, p*** < 0.001). Data shown are the means of three independent biological repeat experiments ± SD. 166

eliminating the possibility that the increased number of aggregates observed in NMD mutants might arise due to difference in the expression levels of Sup35.

Given that the increased formation of Sup35 aggregates was only observed in the NMD mutants, we next asked whether there is a correlation between the Sup35 aggregates and induced [PSI+] prion formation using the ade1-14 mutant allele assay as previously described. As expected, overexpression of SUP35NM-GFP significantly induced [PSI+] formation, with an approximate 10-fold increase in frequency observed in the WT strain (compare Fig. 5.9A with Fig. 5.10C). This induction of [PSI+] formation was increased to approximately 30-fold in NMD mutants, confirming that the frequency of both spontaneous and overexpression-induced [PSI+] formation is elevated in NMD mutants (Fig. 5.10C).

5.2.10 The [PSI+] status of strains improves cell viability upon exposure to various stress conditions

Our results indicate that the frequency of [PSI+] prion formation is increased in

NMD mutants but not in other mRNA surveillance pathways. This may mean that

NMD normally acts to suppress the de novo formation of [PSI+], although it is not clear how a defense system which normally recognizes and degrades mRNAs containing premature stop codons might influence prion formation. Alternatively, there may be a selective pressure to form the [PSI+] prion, if [PSI+]-status provides some growth advantage to NMD mutants. Previously in Chapter 3, we observed improved viability in the [PSI+] versions of all strains following hydrogen peroxide stress (compare Fig. 3.5B with Fig. 3.5D). However, this effect was even more pronounced in NMD mutants with [PSI+] versions showing strongly increased

167

168

Figure 5.11: Survival of the WT and NMD mutant strains during stress conditions are improved when they are in [PSI+] form. Viability analysis is shown for the WT and NMD mutant strains grown to exponential phase (OD600 0.4-0.7) and challenged with: A) 6 mM diamide for one hour; B) 16% ethanol for one hour; C) 45 ºC heat shock treatment for one hour; and D) 3 mM NaCl for four hours. Cells were diluted and plated in triplicate onto YEPD medium to monitor cell viability. Percent survival is expressed relative to the untreated control cultures and compared between the [psi-] and its isogenic [PSI+] cells, respectively.

169

resistance to hydrogen peroxide compared with their [psi−] versions (compare Fig.

3.5B with Fig. 3.5D). This suggests that the [PSI+] status of NMD mutants may be particularly beneficial to NMD mutants during hydrogen peroxide stress. We therefore examined whether this is also true for other oxidants and stress conditions.

The general stress sensitivity of [PSI+] and [psi−] versions of WT and NMD mutants were compared using viability tests. We first examined sensitivity to oxidative stress caused by the thiol oxidant diamide. The [psi−] versions of both NMD mutants showed a modest increase in resistance to diamide compared with the WT strain (Fig 5.11A, left). In comparison, the [PSI+] versions of all strains were more resistant to diamide compared with their isogenic [psi−] versions (Fig 5.11A, right).

Thus, although the result is less pronounced than with hydrogen peroxide stress, these data further indicate that the [PSI+] NMD mutant strains show increased oxidant tolerance.

We next asked whether these differences in stress tolerance are specific for oxidative stress or are also true for other types of stress. We repeated our viability experiments by treating exponentially growing cells (OD600 0.4-0.7) with several different stress conditions including: 16% ethanol for one hour to measure ethanol tolerance (Fig. 5.11B), 45 ºC heat for one hour to induce heat shock (Fig. 5.11C) and

3 mM NaCl for four hours to induce an osmotic shock (Fig. 5.11D). Similar to diamide treatment, the [psi−] versions of both NMD mutants showed a modest increase in resistance to ethanol (Fig 5.11B, left), heat shock (Fig. 5.11C, left) and osmotic shock (Fig. 5.11D, left) compared with the [psi−] WT. Similarly, the [PSI+] versions of all strains were improved in terms of viability when subject to ethanol (Fig. 5.11B, right), heat shock (Fig. 5.11C, right) and osmotic shock (Fig. 5.11D, right) treatments.

[PSI+] versions of NMD mutants behaved somewhat differently following heat shock

170

since the upf2 mutant was more resistance than both the WT and upf1 mutant strains. In summary, [psi−] versions of NMD mutants were generally more stress resistant than the corresponding WT strain. The [PSI+] status of all strains increased their stress tolerance compared with [psi−] strains, and the pattern of increased resistance in NMD mutants compared with a WT strain was still maintained.

5.3 Discussion

For decades, huge interest in understanding protein depositions has been attributed to its serious implications in many pathological diseases as diverse as eye disorders (inherited cataracts) to several neurodegenerative diseases (e.g. AD, PD and HD) (Chiti and Dobson, 2006; Invernizzi et al., 2012). Despite the large number of proteins known to aggregate, only a few globular proteins have known aggregation mechanisms including prions, a type of protein aggregate enriched in β-sheet structures. To attempt to address the underlying mechanism of protein aggregation, we therefore identified and characterized the range of proteins which aggregate following loss of NSD, NMD, and NGD pathways. These quality control mechanisms exist to recognize aberrant mRNAs including those identified as nonstop mRNAs,

PTC-containing mRNAs, or mRNAs that stalled ribosomes due to the presence of stable, secondary structure, respectively. Failure to degrade these aberrant mRNAs might result in the production of aberrant proteins which could form aggregates. In general, we found that loss of mRNA surveillance pathways results in an accumulation of aggregation, and that they are enriched in some different functional categories compared with the WT strain.

Our first important observation is that all of the mutants had elevated protein aggregation detected both biochemically, and using the Hsp104 chaperone as a

171

marker of aggregates, indicating that loss of mRNA surveillance pathways disturbs the proteostasis network. Differences in the normal intracellular localization of proteins within aggregates were found between the WT and ski7, ski8, upf2 and hbs1 mutant strains. In particular, a higher number of proteins were found that localize in the cell wall, cytoskeleton and golgi apparatus in all of the mutant strains. The cytoskeleton provides shape and rigidity to the cell and facilitates transport of cell organelles through three main components: microfilaments, intermediate filaments, and microtubules. Interestingly, several studies have linked loss of cytoskeletal function in a number of neurodegenerative diseases. For example, higher numbers of aggregates were observed due to loss of microtubules in nerve axons and dendrites in AD patients (Irvine et al., 2008). Another study showed that aggregates containing cytoskeleton components and vesicular transport motors were observed in SOD mutants and this may trigger the onset of ALS disease (Wang et al., 2009).

Therefore, our data suggest a mechanism by which disruption of mRNA surveillance pathways initiate a cascade of events including defects in cytoskeleton and this may contribute to neurodegeneration.

We next assessed the influence of several properties of aggregated proteins including physicochemical properties identified in mRNA surveillance pathways, since a protein’s thermodynamic stability, structural class and amino acid sequences are all found to govern aggregation propensity (Agostini et al., 2012; Conchillo-Solé et al., 2007; Tartaglia and Vendruscolo, 2010). For instance, enrichment for certain amino acids can be associated with amyloid diseases including HD (glutamine) and prions (glutamine and asparagine). We found that glutamine was enriched in both ski8 and upf2 mutants while asparagine was under-represented in all strains, suggesting that the amino acid asparagine has little influence in driving protein

172

aggregation. This result is in agreement with a recent study published from our lab which also observed no significant enrichment of this amino acid (Weids et al., 2016).

Moreover, the fact that there are other, well-known amyloid-related diseases including AD and type 2 diabetes that do not have specific enrichment of amino acid sequences (Michelitsch and Weissman, 2000; Perutz, 1999) highly suggest that it is not the key contributor for protein aggregation.

Which aggregation propensities then, acts as the driving force of protein aggregation? Many studies have suggested hydrophobicity as the main driving force; the burial of hydrophobic segments in a native protein and in surfaces highly stabilize both folding and aggregation processes (Linding et al., 2004; Routledge et al., 2009).

In agreement with the idea, our current data and previous work done in our lab

(Weids et al., 2016) had observed that the aggregates were found to be significantly more hydrophobic than the MS-set during basal conditions. Meanwhile, several studies had attempted to correlate aggregation with other physicochemical properties including protein abundance and mRNA expression levels with conflicting results. For instance, it was observed that aggregation prone proteins in S. cerevisiae are generally low in abundance and are lowly expressed presumably due to the tight translational regulation resulting in increased protein turnover (Gsponer and Babu,

2012; Tartaglia et al., 2007). In contrast, other studies including our data have found that aggregation-prone proteins are generally more abundant as well as highly expressed compare with unaggregated ones (Ciryam et al., 2013; Tartaglia and

Vendruscolo, 2009; Weids et al., 2016). They reasoned that the threat of aggregation is unavoidable in highly concentrated proteins as they pose a bigger challenge on the proteostasis network or when the network is being disturbed. Indeed, estimates have suggested that the aggregation tendency of globular proteins may be as high as 20%

173

in diverse proteomes (Rousseau et al., 2006). This is possibly due to a very crowded and complex environment that nascent proteins may have encountered resulting in greater chance of specific and non-specific molecular interactions and thus, aggregation is more likely to occur (Gierasch and Gershenson, 2009). Taken together, we suggest that abundance, translation rates, and hydrophobicity are good indicators of physicochemical properties of protein aggregation in vivo.

The increased risk of aggregation caused by macromolecular crowding in cells further emphasizes the critical roles played by molecular chaperones, which aid folding, as well as PQC systems such as the proteasome (Gershenson and

Gierasch, 2011; Vendruscolo, 2012). Both chaperones and proteases have been shown to be vital in maintaining proteostasis during stress conditions (Hartl et al.,

2011; Richter et al., 2010) and also during physiological/unstressed conditions

(Powers and Balch, 2013). In agreement with the idea, we determined that molecular chaperones are detected within our aggregates. As expected, we observed that several chaperones are present in aggregates including those isolated from the WT strain similar to previous study (Weids et al., 2016). We generally observed an increase in the number of chaperones associated with aggregates isolated in the mRNA surveillance mutants compared to the WT strain. Moreover, the majority of molecular chaperones that were detected belong to Hsp70s, the largest class of family members that bind to short hydrophobic segments in proteins and therefore function to protect unfolded proteins and prevent aggregation (Saibil, 2013). While

Hsp70s mainly function to block aggregation from happening, another important class of molecular chaperones that is also majorly present within our aggregates, chaperonins, serves to provide an isolated environment within cells such that nascent proteins are able to avoid improper interaction and properly fold (Saibil,

174

2013). We also observed that sHsps (small heat shock proteins) chaperone members are not present within our aggregates. Given that sHsps are mainly up- regulated in response to various stress conditions (Bakhtisaran et al. 2014), our study correlates with a previous study (Weids et al., 2016) which showed the presence of sHsps in aggregates formed in response to hydrogen peroxide and arsenite stress but not during non-stressed conditions.

While an increase in protein aggregation in general were observed in all of the

NSD, NMD, and NGD mutant strains, increased amyloid aggregates either by de novo formation of [PSI+], or by overexpressing SUP35 to induce the [PSI+] prion formation, was only observed in NMD mutants (compare Fig. 5.9 and Fig. 5.10). In addition, we tested the possibility that the [PSI+] prion is formed at a higher frequency in NMD mutants since it provides a selective growth advantage. We performed viability assays using several stress conditions including oxidative stress, heat shock, ethanol tolerance and osmolarity that are known to affect regulation of gene expression in yeast cells and examined the effect of [PSI+] status in the WT and

NMD mutants. We found that the [PSI+] status improved viability in all stress conditions examined thus far. Similarly, previous studies have indicated that the

[PSI+] status may be beneficial to cells as it allows cells to reprogram gene expression such that new traits may be uncovered which may aid cells during certain stressful conditions including heat shock (Eaglestone et al., 1999; True and

Lindquist, 2000; True et al. 2004; Tyedmers et al., 2008). In addition, recent findings published in our lab have found that increased [PSI+] prion formation is observed during chronological ageing, and that [PSI+] formation influenced cell longevity such that cells had increased lifespan during ageing in the WT strain (Speldewinde and

Grant 2017). Therefore, although the exact reason as to why such a tendency of

175

forming [PSI+] prion is only observed in NMD mutants is unclear, it is possible that switching to [PSI+] is necessary for cells to survive in adverse situations, and that this effect may not necessarily due to specific changes in gene expression but through other mechanisms that require further investigation.

176

6.0 General Discussion

Organisms are exposed to ROS during the course of normal aerobic metabolism which can cause wide-ranging damage to macromolecules including cellular mRNAs resulting in translational defects and this has been implicated in many neurodegenerative diseases (Gutteridge, 1993). We studied the requirement for three mRNA quality control mechanisms which monitor mRNAs for translational errors during oxidative stress: NMD, NSD and NGD. Using mainly hydrogen peroxide to induce oxidative stress, we found that a mutant deleted for SKI8 is particularly sensitive to oxidative stress conditions in Chapter 3. Ski8 is a part of the

Ski complex composed of Ski2, Ski3, and Ski8 and we further showed in Chapter 4 that mutants lacking components of the Ski complex are generally sensitive to oxidative stress conditions. Interestingly however, the same requirement did not exist for Ski7, indicating that Ski7 and Ski8 are not similarly required during oxidative stress conditions. Ski7 is thought to function as an adaptor protein which bridges the interaction between the Ski complex and the exosome (Frischmeyer et al., 2002; van

Hoof et al., 2002). However, Ski7 has only been found in a small subset of yeasts including S. cerevisiae (Atkinson et al., 2008), and it has been demonstrated that

NGD components (Dom34/Hbs1) also function during NSD in mammals (Tsuboi et al., 2012). Furthermore, our data showed that these NGD factors may also compensate for the loss of Ski7 during oxidative stress conditions in S. cerevisiae; this may explain why loss of Ski7 alone does not cause sensitivity to oxidative stress due to functional redundancy with the NGD components.

177

regulator of NMD, has also been shown to be required for the transcriptional induction of many oxidative stress-regulated genes in the fission yeast

Schizosaccharomyces pombe and upf1 strains are sensitive to H2O2 stress

(Rodriguez-Gabriel et al., 2006). Further work will be required to understand the hypersensitivity of ski complex mutants to oxidative stress. This will be important since RNA metabolism has many disease links and mutations in human exosome subunit genes have been linked with the childhood-onset neurological diseases

(Muller et al., 2015).

The process of mRNA decay is recognized as a major contributor to the regulation of gene expression as well as maintaining regulatory responses crucial for cellular homeostasis. mRNA turnover plays critical role in assessing the accuracy of mRNA biogenesis and in the degradation of aberrant transcripts. The importance that mRNA surveillance pathways might have in regulating proteostasis is currently not well understood. The aim of Chapter 5 therefore was to determine whether loss of NSD, NMD or NGD, would have any effects on proteostasis by examining protein aggregation in strains defective in mRNA surveillance pathways. We showed that defects in mRNA quality control systems result in the production of aberrant proteins that tend to misfold and form aggregates by isolating and identifying the proteins that aggregate. Indeed, our bioinformatic analysis indicates that increased aggregation of aggregation-prone proteins predominantly occurs in mRNA surveillance mutants, rather than pathway-specific aggregation. The proteins that aggregate in mRNA surveillance mutants tend to be more highly expressed, more abundant and more stable proteins compared with the wider proteome. There is also a strong correlation with the proteins that aggregate in response to nascent protein misfolding and an enrichment for proteins that are substrates of ribosome-associated Hsp70

178

chaperones consistent with susceptibility for aggregation primarily occurring during translation/folding. Finally, we also specifically examined the effect of [PSI+] status in the WT and NMD mutant strains and in general, we observed enhance viability in both WT and NMD mutant strains across a range of stressful conditions. Fig. 6.1 summarizes the main findings in this thesis as well as how these mRNA surveillance pathways could contribute to cellular responses such as oxidative stress in general.

179

Result: Termination Normal ROS level efficiency is compromised due to Sup35 aggregation Antioxidant mechanisms Oxidative stress (Chapter 4 and Chapter 5). Model: -any additional roles that Increased ROS level may be compromised too? (See Section 6.2) - Sup35 aggregation: amorphous vs amyloid: which one is potentially more toxic to cells? (See Section 6.3) Oxidation Normal level of Increased level of mRNA damage mRNA damage

Activation of mRNA surveillance pathways

Normal Normal

When compromised

NGD NMD Result: The NGD complex Result: not required during is required during several stress conditions NSD translation initiation including oxidative stress, Result: The Ski complex (Chapter 3). heat shock or osmotic is required for oxidant How: Loss of NGD shock (Chapter 3 and tolerance (Chapter 3 complex results in Chapter 5). and Chapter 4). increased inhibition of How: Improved cell viability How: Hypersensitivity to translation initiation as in variety of stress oxidative stress is observed in polysome conditions, and its PSI+ observed upon loss of profiles even in the status further enhanced Ski complex, but this absence of oxidative this condition. phenotype is not shared stress Model: Due to the by Ski7. Model: Is this inhibition presence of 22 PTCs in the Model: Functional due to the action of Gcn2 strain background redundancy: Ski7 versus kinase or the NGD including MSN4? (See NGD complex (See complex or both? (See Section 6.4) Chapter 4) Section 6.1)

Figure 6.1: Summary of main findings and proposed models in this study

180

6.1 Possible interaction between NGD complex and Gcn2 kinase

It is noteworthy to mention that loss of the NGD complex had the most dramatic effect on translation activity as translation initiation appeared inhibited even in unstressed condition as compared to the WT strain. Given that the NGD complex participates in ribosome recycling by dissociating ribosomes into 40S and 60S subunits so that they become available for subsequent rounds of initiation, any disruption of this NGD complex may potentially result in attenuation of translation initiation, consistent with previous findings (Bhattacharya et al., 2010; Carr-Schmid et al., 2002; den Elzen et al., 2014). Additionally, given that translation initiation is known to be inhibited by the action of the Gcn2 kinase in response to oxidative stress (Shenton et al., 2006), it will be interesting in future experiments to examine whether the attenuation of translation initiation observed in NGD mutants depends on Gcn2. Deleting GCN2 in dom34 and hbs1 mutants and comparing their polysome profiles in the presence and absence of hydrogen peroxide stress would be the first step to address this issue. If a double knockout of GCN2/DOM34 or GCN2/HBS1 results in attenuation of translation initiation even in the absence of hydrogen peroxide, this means that the initiation block observed in the NGD mutants (Fig. 3.9) does not depend on Gcn2 kinase. Further experiments can be carried out by assessing growth rates and performing viability assays on all of these mutants. If the hypothesis holds true, disruption of Gcn2 should result in a high resistance phenotypes to hydrogen peroxide in NGD mutants, as NGD mutants showed modest inhibition to hydrogen peroxide as discussed in Chapter 3.

181

6.2 Aggregation of Sup35 upon oxidative stress: where does it take us?

We tested one potential mechanism to explain why NSD is required during conditions of oxidative stress. Our data indicate that oxidative stress conditions target the translation termination machinery, causing aggregation of the Sup35 termination factor. This sequesters Sup35 away from its normal function in translation termination and potentially generates NSD substrates by allowing ribosomes to readthrough stop codons into the 3’-end of mRNAs. However, several previous studies have described additional functional roles for Sup35 even in the absence of oxidative stress. For example, a genome-wide analysis of the [PSI+] prion revealed that the effects of [PSI+] are not just limited to stop codon readthrough, but may also influence translation quality without affecting stress response mechanisms

(Baudin-Baillieu et al., 2014). Furthermore, there is evidence that release factors can discriminate against sense codons under normal conditions and Sup45–Sup35 may be capable of binding to stalled ribosomal complexes that contain a sense codon in the ribosomal A-site (Janzen et al., 2002; Doronina et al., 2008; Chiabudini et al.,

2014). Both findings suggest that Sup45-Sup35 together with the canonical release factors could serve as regulatory mechanism for cell survival under unfavourable conditions without activating translation-stress signalling as degrading misfolded peptides and rescuing stalled ribosomes is a lengthy and continuous process.

Sup35 is also involved in other cellular processes including mRNA decay through deadenylation, chromosome segregation and cytoskeleton organization (Lee et al., 2007; Linding et al., 2004; Shoemaker et al., 2010). Sup35 was shown to directly interact with Pab1 in the process of deadenylation (Hoshino et al., 1999;

Hosoda et al., 2003). Pab1 functions in protecting poly(A) tails to give stability to mRNA structures and has been shown to control the length of poly(A) in vitro

182

(Amrani et al., 1997; Hector et al., 2002). Furthermore, Pab1 plays important roles during mRNA loop formation via eIF4G to facilitate the translation re-initiation process (Tarun et al., 1997). It has been shown that deletion of the Sup35 N- terminus results in longer poly(A) tail length and that this decreases mRNA turnover rates (Hosoda et al., 2003). Therefore, any conditions that abrogate the function of the Sup35 N-terminus, such as oxidative stress-induced Sup35 aggregation, may also disrupt the normal interaction between Sup35 and Pab1 which protects mRNA from normal turnover. It would therefore be interesting in future studies to examine the competition between Sup35 and Pab1 during oxidative stress conditions and to test how this influences the 3’-poly(A) tail.

6.3 Widespread protein aggregation may potentially be more toxic than amyloid aggregates

In chapter 4 we found that exposure to acute stresses such as oxidative stress may produce aberrant polypeptides due to Sup35 aggregation, which is even more severe or pronounced in the absence of NSD. This may present cells with a challenge to maintain protein integrity viability during such stress conditions. A number of PQC systems including folding catalysts, molecular chaperones and proteases exist to protect against this. This is important because the formation of potentially toxic protein aggregates is thought to be an underlying cause of several neurodegenerative diseases such as AD and HD. Indeed, recent studies have highlighted the importance of ribosome quality control (RQC) targeting terminally misfolded proteins for degradation via the ubiquitin proteasome system (Choe et al.

2016; Yonashiro et al. 2016). The roles of several other factors including mRNA surveillance pathways which are required to monitor the quality of mRNAs are clearly

183

implicated in mRNA decay, but their roles in proteostasis remains unclear. We therefore proceeded to investigated whether loss of these mRNA surveillance pathways may also affect proteostasis leading to protein aggregation as described in

Chapter 5. An increase in widespread aggregation was observed following loss of all of the mRNA surveillance pathways, whereas, increased amyloid aggregation, as measured by [PSI+] prion formation, was only observed in NMD mutants.

The reason for the high frequency of [PSI+] prion formation in NMD mutants is currently unclear. As briefly outlined in Section 1.7, amorphous aggregates share a lot of structural similarity with intermediate oligomers and protofibrils, which were found to be the most toxic species in human amyloidosis. For example, the oligomers of amyloid proteins including A (Lesne et al. 2006; Walsh et al. 2002), - synuclein (Conway et al. 1998; Winner et al. 2011), polyQ (Sanchez et al. 2003;

Nagai et al. 2007), and PrP (Lasmezas et al. 1997), were found to cause cellular toxicity in disease pathogenesis when introduced into cell and animal models. In addition, the presence of amyloid fibers in the brain of healthy individuals further challenged the hypothesis that amyloid aggregates actually contribute to disease pathology (Broersen et al. 2010). Future work will be required to establish whether the amyloid aggregates e.g. [PSI+] prions may actually sequester the toxic oligomers that are more abundant during amorphous state and therefore act as defense mechanism for cells in the absence of NMD mutants in particular.

6.4 The presence of 22 PTCs in strain 74D-694 may actually contribute to increased tolerance to various stresses particularly in NMD mutants.

Whilst PTCs are unlikely to be introduced into mRNAs as a result of oxidative stress, many mRNAs contain naturally occurring PTCs. Conditions which promote

184

termination codon readthrough such as oxidative stress may therefore cause the production of altered and potentially toxic proteins from such mRNAs. For example, whole genome sequencing of the yeast strain used in our study has shown that twenty-two genes encode potential premature stop codons which might affect the production of proteins involved in a variety of cellular processes (Fitzpatrick et al.,

2011). Hence, the [PSI+] status in our strain background might result in nonsense suppression of these 22 PTC-containing genes and altered protein production. One of these genes encodes for Msn4, a transcription factor that together with Msn2, activate the general stress response (ESR) as briefly described in Section 1.4.2. One possible model therefore, is that in [PSI+] cells, a functional Msn4 is produced leading to up-regulation of stress-responsive genes which are regulated by Msn4.

This effect may be further increased in NMD mutants resulting in the stabilization of

PTC-containing mRNAs. If this model holds true, then other strain backgrounds such as W303/BY4741/BY4742 that have functional Msn4 may be more stress tolerance than the 74D-694 strain background. Indeed, when we compared our nonstop mutants between two different strain backgrounds (74D-694 and BY4741), we found that much higher concentrations of hydrogen peroxide are required to inhibit the growth of the BY4741 background compared with the 74D-694 background (Figure

4.1).

Future work should investigate the role of PTCs in mediating [PSI+]- dependent stress tolerance. For example, it would be interesting to substitute a functional MSN4 gene (instead of the truncated version) into the 74D-694 strain background to determine whether it improves stress tolerance in a [psi-] background, relative to its isogenic [PSI+] version. The switch to [PSI+] may have an additive on readthrough in NMD mutants which would essentially select for [PSI+] formation in

185

NMD mutants if it provides a growth advantage. The presence of a functional Sup45-

Sup35 termination complex is required in order for NMD to function properly, so it would be interesting to find out whether the [PSI+] effects in NMD mutants may exhibit an additive effect resulting in more readthrough of Msn4. Such an effect may also influence NMD mutants to form [PSI+] since it is advantageous during various stress conditions (Fig.5.11). These experiments should also include a strain background that has functional Msn4 e.g. in the BY4741 background so as to eliminate the possibility of artefacts caused by mRNA stability. Western blots should also be used to monitor the expression levels of Msn4 in all strains; more Msn4 will be produced in the [PSI+] WT as opposed to its isogenic [psi-] background. We should also expect higher Msn4 is produced in the [PSI+] NMD mutants due to higher readthrough of Msn4 compared to the [PSI+] WT.

Interestingly, NMD has been shown to play either protective or deleterious role due to the presence of premature stop mutations in a large number of known inherited human diseases and cancers such as -thalassemia, Marfan syndrome, muscular dystrophy, cystic fibrosis, and many more (Dietz 1997; Valentine 1998). - thalassemia for instance is caused by PTCs in the HBB genes which encodes for the

-globin protein. Therefore, a patient that has homozygous -globin alleles may suffer severe anemia due to deficiency of tetrameric haemoglobin which requires both -globin and -globin to function (Peixeiro et al. 2011). On the other hand, a patient that has Marfan syndrome may benefit from the protective effects of NMD.

Marfan syndrome is an inherited disorder of connective tissue caused by nonsense mutations in the FBN1 gene (codes for fibrillin-1 protein). When NMD is activated, a truncated version of Fbn1 is produced leading to a milder phenotype of the disease while a severe phenotype is observed when the NMD is deactivated causing the 186

truncated version to accumulate and becomes toxic to the patient (Dietz et al. 1993).

Thus, given the important role of NMD in a significant portion of human diseases, a better understanding of the intricate molecular mechanisms of NMD will provide a more solid foundation for the treatment of NMD-related human diseases in future.

187

7.0 BIBLIOGRAPHY

Agostini, F., Vendruscolo, M., and Tartaglia, G.G. (2012). Sequence-based prediction of protein solubility. Journal of Molecular Biology 421, 237-241. Alepuz, P.M., Matheos, D., Cunningham, K.W., and Estruch, F. (1999). The Saccharomyces cerevisiae RanGTP-binding protein msn5p is involved in different signal transduction pathways. Genetics 153, 1219-1231. Allmang, C., Kufel, J., Chanfreau, G., Mitchell, P., Petfalski, E., and Tollervey, D. (1999). Functions of the exosome in rRNA, snoRNA and snRNA synthesis. The EMBO Journal 18, 5399-5410. Amrani, N., Minet, M., Le Gouar, M., Lacroute, F., and Wyers, F. (1997). Yeast Pab1 interacts with Rna15 and participates in the control of the poly(A) tail length in vitro. Molecular and Cellular Biology 17, 3694-3701. Anderson, J.S., and Parker, R.P. (1998). The 3' to 5' degradation of yeast mRNAs is a general mechanism for mRNA turnover that requires the SKI2 DEVH box protein and 3' to 5' exonucleases of the exosome complex. The EMBO Journal 17, 1497-1506. Araki, Y., Takahashi, S., Kobayashi, T., Kajiho, H., Hoshino, S.-i., and Katada, T. (2001). Ski7p G protein interacts with the exosome and the Ski complex for 3′-to-5′ mRNA decay in yeast. The EMBO Journal 20, 4684-4693. Arava, Y., Wang, Y., Storey, J.D., Liu, C.L., Brown, P.O., and Herschlag, D. (2003). Genome-wide analysis of mRNA translation profiles in Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences of the United States of America 100, 3889- 3894. Arora, C., Kee, K., Maleki, S., and Keeney, S. (2004). Antiviral protein Ski8 is a direct partner of Spo11 in meiotic DNA break formation, independent of its cytoplasmic role in RNA metabolism. Molecular Cell 13, 549-559. Ashe, M.P., De Long, S.K., and Sachs, A.B. (2000). Glucose depletion rapidly inhibits translation initiation in yeast. Molecular Biology of the Cell 11, 833-848. Atkinson, G.C., Baldauf, S.L., and Hauryliuk, V. (2008). Evolution of nonstop, no-go and nonsense-mediated mRNA decay and their termination factor-derived components. BMC Evolutionary Biology 8, 290-290. Ayala, A., Mu, #xf1, oz, M.F., Arg, #xfc, and elles, S. (2014). Lipid Peroxidation: production, metabolism, and signaling mechanisms of Malondialdehyde and 4-Hydroxy-2- Nonenal. Oxidative Medicine and Cellular Longevity 2014, 31. Azevedo, D., Tacnet, F., Delaunay, A., Rodrigues-Pousada, C., and Toledano, M.B. (2003). Two redox centers within Yap1 for H2O2 and thiol-reactive chemicals signaling. Free Radical Biology and Medicine 35, 889-900. Baudin-Baillieu, A., Legendre, R., Kuchly, C., Hatin, I., Demais, S., Mestdagh, C., Gautheret, D., and Namy, O. (2014). Genome-wide translational changes induced by the Prion [PSI+]. Cell Reports 8, 439-448. Benard, L. (2004). Inhibition of 5′ to 3′ mRNA degradation under stress conditions in Saccharomyces cerevisiae: from GCN4 to MET16. RNA 10, 458-468. Berson, J.F., Theos, A.C., Harper, D.C., Tenza, D., Raposo, G., and Marks, M.S. (2003). Proprotein convertase cleavage liberates a fibrillogenic fragment of a resident glycoprotein to initiate melanosome biogenesis. The Journal of Cell Biology 161, 521-533. Bharathi, V., Girdhar, A., Prasad, A., Verma, M., Taneja, V., and Patel, B.K. (2016). Use of ade1 and ade2 mutations for development of a versatile red/white colour assay of amyloid- induced oxidative stress in Saccharomyces cerivisiae. Yeast 33, 607-620.

188

Bhattacharya, A., McIntosh, K.B., Willis, I.M., and Warner, J.R. (2010). Why Dom34 stimulates growth of cells with defects of 40S ribosomal subunit biosynthesis. Molecular and Cellular Biology 30, 5562-5571. Biancalana, M., and Koide, S. (2010). Molecular mechanism of thioflavin-T binding to Aamyloid fibrils. Biochimica et biophysica acta 1804, 1405-1412. Bradley, M.E., Edskes, H.K., Hong, J.Y., Wickner, R.B., and Liebman, S.W. (2002). Interactions among prions and prion "strains" in yeast. Proceedings of the National Academy of Sciences of the United Science of America. 99 Suppl 4, 16392-16399. Brennan, R.J., and Schiestl, R.H. (1996). Cadmium is an inducer of oxidative stress in yeast. Mutat Res 356, 171-178. Carr-Schmid, A., Pfund, C., Craig, E.A., and Goss Kinzy, T. (2002). Novel G-protein complex whose requirement is linked to the translational status of the cell. Molecular and Cellular Biology 22, 2564-2574. Carrió, M., González-Montalbán, N., Vera, A., Villaverde, A., and Ventura, S. (2005). Amyloid-like properties of bacterial inclusion bodies. Journal of Molecular Biology 347, 1025-1037. Castilla, J., Saá, P., Hetz, C., and Soto, C. (2005). In vitro generation of infectious Scrapie prions. Cell 121, 195-206. Caughey, B., and Lansbury, P.T. (2003). Protofibrils, pores, fibrils and neurodegeneration: Separating the Responsible Protein Aggregates from The Innocent Bystanders. Annual Review of Neuroscience 26, 267-298. Causton, H.C., Ren, B., Koh, S.S., Harbison, C.T., Kanin, E., Jennings, E.G., Lee, T.I., True, H.L., Lander, E.S., and Young, R.A. (2001). Remodeling of yeast genome expression in response to environmental changes. Molecular Biology of the Cell 12, 323-337. Chakrabarti, S., Jayachandran, U., Bonneau, F., Fiorini, F., Basquin, C., Domcke, S., Le Hir, H., and Conti, E. (2011). Molecular mechanisms for the RNA-dependent ATPase activity of Upf1 and its Regulation by Upf2. Molecular Cell 41, 693-703. Chapman, M.R., Robinson, L.S., Pinkner, J.S., Roth, R., Heuser, J., Hammar, M., Normark, S., and Hultgren, S.J. (2002). Role of Escherichia coli Curli Operons in directing amyloid fiber formation. Science 295, 851-855. Chernoff, Y., Lindquist, S., Ono, B., Inge-Vechtomov, S., and Liebman, S. (1995). Role of the chaperone protein Hsp104 in propagation of the yeast prion-like factor [PSI+]. Science 268, 880-884. Chernova, T.A., Wilkinson, K.D., and Chernoff, Y.O. (2014). Physiological and environmental control of yeast prions. FEMS Microbiology Reviews 38, 326-344. Chiabudini, M., Tais, A., Zhang, Y., Hayashi, S., Wolfe, T., Fitzke, E., and Rospert, S. (2014). Release factor eRF3 mediates premature translation termination on polylysine-stalled ribosomes in S. cerevisiae. Molecular and Cellular Biology 34, 4062-4076. Chiti, F., and Dobson, C.M. (2006). Protein misfolding, functional amyloid, and human disease. Annual Review of Biochemistry 75, 333-366. Christiano, R., Nagaraj, N., Frohlich, F., and Walther, T.C. (2014). Global proteome turnover analyses of the yeasts S. cerevisiae and S. pombe. Cell Reports 9, 1959-1965. Ciryam, P., Tartaglia, G.G., Morimoto, R.I., Dobson, C.M., and Vendruscolo, M. (2013). Widespread aggregation and neurodegenerative diseases are associated with supersaturated proteins. Cell Reports 5, 781-790. Clemens, M.J. (2001). Initiation factor eIF2 alpha phosphorylation in stress responses and apoptosis. Prog Mol Subcell Biol 27, 57-89.

189

Clerici, M., Mourão, A., Gutsche, I., Gehring, N.H., Hentze, M.W., Kulozik, A., Kadlec, J., Sattler, M., and Cusack, S. (2009). Unusual bipartite mode of interaction between the nonsense-mediated decay factors, UPF1 and UPF2. The EMBO Journal 28, 2293-2306. Collinge, J. (2010). Prion strain mutation and selection. Science 328, 1111. Conchillo-Solé, O., de Groot, N.S., Avilés, F.X., Vendrell, J., Daura, X., and Ventura, S. (2007). AGGRESCAN: a server for the prediction and evaluation of "hot spots" of aggregation in polypeptides. BMC Bioinformatics 8, 65-65. Czaplinski, K., Ruiz-Echevarria, M.J., Paushkin, S.V., Han, X., Weng, Y., Perlick, H.A., Dietz, H.C., Ter-Avanesyan, M.D., and Peltz, S.W. (1998). The surveillance complex interacts with the translation release factors to enhance termination and degrade aberrant mRNAs. Genes & Development 12, 1665-1677. De Freitas, J., Wintz, H., Kim, J.H., Poynton, H., Fox, T., and Vulpe, C. (2003). Yeast, a model organism for iron and copper metabolism studies. Biometals 16, 185-197. Deleault, N.R., Harris, B.T., Rees, J.R., and Supattapone, S. (2007). Formation of native prions from minimal components in vitro. Proceedings of the National Academy of Sciences of the United States of America 104, 9741-9746. Deleault, N.R., Walsh, D.J., Piro, J.R., Wang, F., Wang, X., Ma, J., Rees, J.R., and Supattapone, S. (2012). Cofactor molecules maintain infectious conformation and restrict strain properties in purified prions. Proceedings of the National Academy of Sciences of the United States of America 109, E1938-E1946. den Elzen, A.M.G., Schuller, A., Green, R., and Séraphin, B. (2014). Dom34-Hbs1 mediated dissociation of inactive 80S ribosomes promotes restart of translation after stress. The EMBO Journal 33, 265-276. Derkatch, I.L., Bradley, M.E., Hong, J.Y., and Liebman, S.W. (2001). Prions affect the appearance of other prions: the story of [PIN(+)]. Cell 106, 171-182. Deshaies, R.J., Sanders, S.L., Feldheim, D.A., and Schekman, R. (1991). Assembly of yeast Sec proteins involved in translocation into the endoplasmic reticulum into a membrane-bound multisubunit complex. Nature 349, 806-808. Dever, T.E., and Green, R. (2012). The elongation, termination and recycling phases of translation in eukaryotes. Cold Spring Harbor perspectives in biology 4, a013706-a013706. Dobson, C.M. (2001). The structural basis of protein folding and its links with human disease. Philosophical Transactions of the Royal Society of London Series B 356, 133-145. Dobson, C.M. (2003). Protein folding and misfolding. Nature 426, 884-890. Doma, M.K., and Parker, R. (2006). Endonucleolytic cleavage of eukaryotic mRNAs with stalls in translation elongation. Nature 440, 561-564. Doronina, V.A., Staniforth, G.L., Speldewinde, S.H., Tuite, M.F., and Grant, C.M. (2015). Oxidative stress conditions increase the frequency of de novo formation of the yeast [PSI(+) ] prion. Molecular Microbiology 96, 163-174. Doronina, V.A., Wu, C., de Felipe, P., Sachs, M.S., Ryan, M.D., and Brown, J.D. (2008). Site-specific release of nascent chains from ribosomes at sense codon. Molecular and Cellular Biology 28, 4227-4239. Dziembowski, A., Lorentzen, E., Conti, E., and Seraphin, B. (2007). A single subunit, Dis3, is essentially responsible for yeast exosome core activity. Nature Structural and Molecular Biology 14, 15-22. Eaglestone, S.S., Cox, B.S., and Tuite, M.F. (1999). Translation termination efficiency can be regulated in Saccharomyces cerevisiae by environmental stress through a prion-mediated mechanism. The EMBO Journal 18, 1974-81.

190

Eaton, P. (2006). Protein thiol oxidation in health and disease: Techniques for measuring disulfides and related modifications in complex protein mixtures. Free Radical Biology and Medicine 40, 1889-1899. Eaton, W.A., Muñoz, V., Hagen, S.J., Jas, G.S., Lapidus, L.J., Henry, E.R., and Hofrichter, J. (2000). Fast Kinetics and Mechanisms in Protein Folding. Annual Review of Biophysics and Biomolecular Structure 29, 327-359. Elliot, M.A., Karoonuthaisiri, N., Huang, J., Bibb, M.J., Cohen, S.N., Kao, C.M., and Buttner, M.J. (2003). The chaplins: a family of hydrophobic cell-surface proteins involved in aerial mycelium formation in Streptomyces coelicolor. Genes & Development 17, 1727- 1740. Erjavec, N., Larsson, L., Grantham, J., and Nystrom, T. (2007). Accelerated aging and failure to segregate damaged proteins in Sir2 mutants can be suppressed by overproducing the protein aggregation-remodeling factor Hsp104p. Genes & Development 21, 2410-2421. Fasken, M.B., and Corbett, A.H. (2005). Process or perish: quality control in mRNA biogenesis. Nature Structural Molecular Biology 12, 482-488. Ferreira, P.C., Ness, F., Edwards, S.R., Cox, B.S., and Tuite, M.F. (2001). The elimination of the yeast [PSI+] prion by guanidine hydrochloride is the result of Hsp104 inactivation. Molecular Microbiology 40, 1357-1369. Fimognari, C. (2015). Role of oxidative RNA damage in chronic-degenerative diseases. Oxidative Medicine and Cellular Longevity 2015, 8. Finkelstein, D.B., and Strausberg, S. (1983). Identification and expression of a cloned yeast heat shock gene. Journal of Biological Chemistry 258, 1908-1913. Fitzpatrick, D.A., O'Brien, J., Moran, C., Hasin, N., Kenny, E., Cormican, P., Gates, A., Morris, D.W., and Jones, G.W. (2011). Assessment of inactivating stop codon mutations in forty Saccharomyces cerevisiae strains: implications for [PSI] prion- mediated phenotypes. PLoS One 6, e28684. Fleming, A.M., and Burrows, C.J. (2016). Formation and processing of DNA damage substrates for the hNEIL enzymes. Free Radical Biology and Medicine. Fowler, D.M., Koulov, A.V., Balch, W.E., and Kelly, J.W. (2007). Functional amyloid – from bacteria to humans. Trends in Biochemical Sciences 32, 217-224. Frischmeyer, P.A., van Hoof, A., O'Donnell, K., Guerrerio, A.L., Parker, R., and Dietz, H.C. (2002). An mRNA surveillance mechanism that eliminates transcripts lacking termination codons. Science 295, 2258-2261. Gandhi, R., Manzoor, M., and Hudak, K.A. (2008). Depurination of brome mosaic virus RNA3 in vivo results in translation-dependent accelerated degradation of the viral RNA. Journal of Biological Chemistry 283, 32218-32228. Gasch, A.P., Spellman, P.T., Kao, C.M., Carmel-Harel, O., Eisen, M.B., Storz, G., Botstein, D., and Brown, P.O. (2000). Genomic expression programs in the response of yeast cells to environmental changes. Molecular Biology of the Cell 11, 4241-4257. Gerashchenko, M.V., Lobanov, A.V., and Gladyshev, V.N. (2012). Genome-wide ribosome profiling reveals complex translational regulation in response to oxidative stress. Proceedings of the National Academy of Sciences of the United Science of America. Gershenson, A., and Gierasch, L.M. (2011). Protein Folding in the Cell: Challenges and Progress. Current Opinion in Structural Biology 21, 32-41. Gierasch, L.M., and Gershenson, A. (2009). Post-reductionist protein science, or putting Humpty Dumpty back together again. Nature Chemical Biology 5, 774-777. Gietz, R.D., Schiestl, R.H., Willems, A.R., and Woods, R.A. (1995). Studies on the transformation of intact yeast cells by the LiAc/SS-DNA/PEG procedure. Yeast 11, 355-360.

191

Gong, Y., Kakihara, Y., Krogan, N., Greenblatt, J., Emili, A., Zhang, Z., and Houry, W.A. (2009). An atlas of chaperone-protein interactions in Saccharomyces cerevisiae: implications to protein folding pathways in the cell. Molecular Systems Biology 5, 275. Görner, W., Durchschlag, E., Wolf, J., Brown, E.L., Ammerer, G., Ruis, H., and Schüller, C. (2002). Acute glucose starvation activates the nuclear localization signal of a stress-specific yeast transcription factor. The EMBO Journal 21, 135-144. Grant, C.M. (2011). Regulation of translation by hydrogen peroxide. Antioxidants Redox Signal 15, 191-203. Gsponer, J., and Babu, M.M. (2012). Cellular strategies for regulating functional and nonfunctional protein aggregation. Cell Reports 2, 1425-1437. Gutteridge, B. (1981). Aspects to consider when detecting and measuring lipid peroxidation. Free Radical Research Communications 1, 173-184. Gutteridge, J.M.C. (1993). Free radicals in disease processes: a compilation of cause and consequence. Free Radical Research Communications 19, 141-158. Halbach, F., Reichelt, P., Rode, M., and Conti, E. (2013). The yeast ski complex: crystal structure and RNA channeling to the exosome complex. Cell 154, 814-826. Halliwell, B., and Gutteridge, J.M.C. (1989). Free Radicals in Biology and Medicine, 2nd edn (Oxford, Oxford University Press). Harbi, D., and Harrison, P.M. (2014). Interaction networks of prion, prionogenic and prion- like proteins in budding yeast, and their role in gene regulation. PLoS One 9, e100615. Harbi, D., Parthiban, M., Gendoo, D.M.A., Ehsani, S., Kumar, M., Schmitt-Ulms, G., Sowdhamini, R., and Harrison, P.M. (2012). PrionHome: A database of prions and other sequences relevant to prion phenomena. PLoS One 7, e31785. Harding, H.P., Novoa, I., Zhang, Y., Zeng, H., Wek, R., Schapira, M., and Ron, D. (2000). Regulated translation initiation controls stress-induced gene expression in mammalian cells. Molecular Cell 6, 1099-1108. Hartl, F.U., Bracher, A., and Hayer-Hartl, M. (2011). Molecular chaperones in protein folding and proteostasis. Nature 475, 324-332. Hasan, R., Leroy, C., Isnard, A., D. , Labarre, J., Boy-Marcotte, E., and Toledano, M.B. (2002). The control of the yeast H2O2 response by the Msn2/4 transcription factors. Molecular Microbiology 45, 233-241. Hatfield, L., Beelman, C.A., Stevens, A., and Parker, R. (1996). Mutations in trans-acting factors affecting mRNA decapping in Saccharomyces cerevisiae. Molecular and Cellular Biology 16, 5830-5838. He, F., Li, X., Spatrick, P., Casillo, R., Dong, S., and Jacobson, A. (2003). Genome-wide analysis of mRNAs regulated by the nonsense-mediated and 5′ to 3′ mRNA decay pathways in yeast. Molecular Cell 12, 1439-1452. Hector, R.E., Nykamp, K.R., Dheur, S., Anderson, J.T., Non, P.J., Urbinati, C.R., Wilson, S.M., Minvielle-Sebastia, L., and Swanson, M.S. (2002). Dual requirement for yeast hnRNP Nab2p in mRNA poly(A) tail length control and nuclear export. The EMBO Journal 21, 1800-1810. Henkel, R.R. (2011). Leukocytes and oxidative stress: dilemma for sperm function and male fertility. Asian Journal of Andrology 13, 43-52. Herrero, E., Ros, J., Belli, G., and Cabiscol, E. (2008). Redox control and oxidative stress in yeast cells. Biochim Biophys Acta 1780, 1217-1235. Hilleren, P., and Parker, R. (1999). Mechanisms of mRNA surveillance in eukaryotes. Annual Review of Genetics 33, 229-260. Hingorani, K.S., and Gierasch, L.M. (2014). Comparing protein folding in vitro and in vivo: foldability meets the fitness challenge. Current opinion in structural biology 0, 81-90.

192

Hinnebusch, A.G. (2000). Mechanism and regulation of initiator methionyl-tRNA binding to ribosomes. In Translational Control of Gene rxpression, N. Sonemberg, J.W.B. Hershey, and M.B. Mathews, eds. (New York, Cold Spring Harbor Laboratory Press), pp. 185-243. Hinnebusch, A.G. (2005). Translational regulation of gcn4 and the general amino acid control of yeast. Annual Review of Microbiology 59, 407-450. Hoffman, C.S., and Winston, F. (1987). A ten-minute DNA preparation from yeast efficiently releases autonomous plasmids for transformation of Escherichia coli. Gene 57, 267-272. Hofmann, J., Wolf, H., Grassmann, A., Arndt, V., Graham, J., and Vorberg, I. (2012). Creutzfeldt-Jakob disease and mad cows: lessons learnt from yeast cells. Swiss Medical Weekly. Hohn, A., Jung, T., and Grune, T. (2014). Pathophysiological importance of aggregated damaged proteins. Free Radical Biology and Medicine 71, 70-89. Hoshino, S.-i. (2012). Mechanism of the initiation of mRNA decay: role of eRF3 family G proteins. Wiley Interdisciplinary Reviews: RNA 3, 743-757. Hoshino, S.-i., Imai, M., Kobayashi, T., Uchida, N., and Katada, T. (1999). The eukaryotic polypeptide chain releasing factor (eRF3/GSPT) carrying the translation termination signal to the 3′-Poly(A) tail of mRNA: direct association of eRF3/GSPT with polyadenylate-binding protein. The Journal of Biological Chemistry 274, 16677-16680. Hosoda, N., Kobayashi, T., Uchida, N., Funakoshi, Y., Kikuchi, Y., Hoshino, S., and Katada, T. (2003). Translation termination factor eRF3 mediates mRNA decay through the regulation of deadenylation. The Journal of Biological Chemistry 278, 38287-38291. Houalla, R., Devaux, F., Fatica, A., Kufel, J., Barrass, D., Torchet, C., and Tollervey, D. (2006). Microarray detection of novel nuclear RNA substrates for the exosome. Yeast 23, 439-454. Ibstedt, S., Sideri, T.C., Grant, C.M., and Tamas, M.J. (2014). Global analysis of protein aggregation in yeast during physiological conditions and arsenite stress. Biology open. Inada, T., and Aiba, H. (2005). Translation of aberrant mRNAs lacking a termination codon or with a shortened 3′-UTR is repressed after initiation in yeast. The EMBO Journal 24, 1584-1595. Invernizzi, G., Papaleo, E., Sabate, R., and Ventura, S. (2012). Protein aggregation: mechanisms and functional consequences. The International Journal of Biochemistry & Cell Biology 44, 1541-1554. Irvine, G.B., El-Agnaf, O.M., Shankar, G.M., and Walsh, D.M. (2008). Protein aggregation in the brain: the molecular basis for Alzheimer’s and Parkinson’s diseases. Molecular Medicine 14, 451-464. Ito-Harashima, S., Kuroha, K., Tatematsu, T., and Inada, T. (2007). Translation of the poly(A) tail plays crucial roles in nonstop mRNA surveillance via translation repression and protein destabilization by proteasome in yeast. Genes & Development 21, 519-524. Jackson, R.J., Hellen, C.U., and Pestova, T.V. (2010). The mechanism of eukaryotic translation initiation and principles of its regulation. Nature Reviews Molecular Cell Biology 11, 113-127. Jacobson, T., Navarrete, C., Sharma, S.K., Sideri, T.C., Ibstedt, S., Priya, S., Grant, C.M., Christen, P., Goloubinoff, P., and Tamas, M.J. (2012). Arsenite interferes with protein folding and triggers formation of protein aggregates in yeast. Journal of Cell Science 125, 5073-5083. Jacquet, M., Renault, G., Lallet, S., De Mey, J., and Goldbeter, A. (2003). Oscillatory nucleocytoplasmic shuttling of the general stress response transcriptional activators Msn2 and Msn4 in Saccharomyces cerevisiae. The Journal of Cell Biology 161, 497-505.

193

Jang, H.H., Lee, K.O., Chi, Y.H., Jung, B.G., Park, S.K., Park, J.H., Lee, J.R., Lee, S.S., Moon, J.C., Yun, J.W., et al. (2004). Two enzymes in one; two yeast peroxiredoxins display oxidative stress-dependent switching from a peroxidase to a molecular chaperone function. Cell 117, 625-635. Janzen, D.M., Frolova, L., and Geballe, A.P. (2002). Inhibition of translation termination mediated by an interaction of eukaryotic release factor 1 with a nascent peptidyl-tRNA. Molecular and Cellular Biology 22, 8562-8570. Jung, G., and Masison, D.C. (2001). Guanidine hydrochloride inhibits Hsp104 activity in vivo: a possible explanation for its effect in curing yeast prions. Current Microbiology 43, 7- 10. Kapp, L.D., and Lorsch, J.R. (2004). The molecular mechanics of eukaryotic translation. Annual Review of Biochemistry 73, 657-704. Katz, M.J., Gandara, L., De Lella Ezcurra, A.L., and Wappner, P. (2016). Hydroxylation and translational adaptation to stress: some answers lie beyond the STOP codon. Cellular and Molecular of Life Sciences 73, 1881-1893. Keeling, K.M., Lanier, J., Du, M., Salas-Marco, J., Gao, L., Kaenjak-Angeletti, A. and Bedwell, D.M. (2004). Leaky termination at premature stop codons antagonizes nonsense- mediated mRNA decay in S. cerevisiae. RNA, 10, 691-703. Kiledjian, M., and Maquat, L. (2008). Preface. In Methods in Enzymology (Academic Press), pp. xxi-xxii. Klauer, A.A., and van Hoof, A. (2012). Degradation of mRNAs that lack a stop codon: a decade of nonstop progress. Wiley Interdiscip Rev RNA 3, 649-660. Koplin, A., Preissler, S., Ilina, Y., Koch, M., Scior, A., Erhardt, M., and Deuerling, E. (2010). A dual function for chaperones SSB-RAC and the NAC nascent polypeptide-associated complex on ribosomes. Journal of Cell Biology 189, 57-68. Kosower, N.S., and Kosower, E.M. (1995). Diamide: an oxidant probe for thiols. Methods Enzymology 251, 123-133. Kuge, S., and Jones, N. (1994). YAP1 dependent activation of TRX2 is essential for the response of Saccharomyces cerevisiae to oxidative stress by hydroperoxides. The EMBO Journal 13, 655-664. Lancaster, A.K., Bardill, J.P., True, H.L., and Masel, J. (2010). The spontaneous appearance rate of the yeast prion [PSI+] and its implications for the evolution of the evolvability properties of the [PSI+] system. Genetics 184, 393-400. Lawless, C., Pearson, R.D., Selley, J.N., Smirnova, J.B., Grant, C.M., Ashe, M.P., Pavitt, G.D., and Hubbard, S.J. (2009). Upstream sequence elements direct post-transcriptional regulation of gene expression under stress conditions in yeast. BMC Genomics 10 1-20. Lebreton, A., and Seraphin, B. (2008). Exosome-mediated quality control: substrate recruitment and molecular activity. Biochim Biophys Acta 1779, 558-565. Lee, H.H., Kim, Y.-S., Kim, K.H., Heo, I., Kim, S.K., Kim, O., Kim, H.K., Yoon, J.Y., Kim, H.S., Kim, D.J., et al. (2007). Structural and functional insights into Dom34, a key component of No-Go mRNA Decay. Molecular Cell 27, 938-950. Lee, R.E., Brunette, S., Puente, L.G., and Megeney, L.A. (2010). Metacaspase Yca1 is required for clearance of insoluble protein aggregates. Proceedings of the National Academy of Sciences of the United States of America 107, 13348-13353. Legname, G., Baskakov, I.V., Nguyen, H.-O.B., Riesner, D., Cohen, F.E., DeArmond, S.J., and Prusiner, S.B. (2004). Synthetic Mammalian Prions. Science 305, 673. Li, Z., Wu, J., and DeLeo, C.J. (2006). RNA damage and surveillance under oxidative stress. IUBMB Life 58, 581-588.

194

Linding, R., Schymkowitz, J., Rousseau, F., Diella, F., and Serrano, L. (2004). A comparative study of the relationship between protein structure and β-aggregation in globular and intrinsically disordered proteins. Journal of Molecular Biology 342, 345-353. Losson, R., and Lacroute, F. (1979). Interference of nonsense mutations with eukaryotic messenger RNA stability. Proceedings of the National Academy of Sciences of the United States of America 76, 5134-5137. Lu, J., and Deutsch, C. (2008). Electrostatics in the Ribosomal Tunnel Modulate Chain Elongation Rates. Journal of Molecular Biology 384, 73-86. Lum, R., Tkach, J.M., Vierling, E., and Glover, J.R. (2004). Evidence for an unfolding/threading mechanism for protein disaggregation by Saccharomyces cerevisiae Hsp104. Journal of Biological Chemistry 279, 29139-29146. Lund, P.M., and Cox, B.S. (1981). Reversion analysis of [psi-] mutations in Saccharomyces cerevisiae. Genetics Research 37, 173-182. Luthe, D.S. (1983). A simple technique for the preparation and storage of sucrose gradients. Analytical Biochemistry 135, 230-232. Maderazo, A.B., He, F., Mangus, D.A., and Jacobson, A. (2000). Upf1p control of nonsense mRNA translation is regulated by Nmd2p and Upf3p. Molecular and Cellular Biology 20, 4591-4603. Makarava, N., Kovacs, G.G., Bocharova, O., Savtchenko, R., Alexeeva, I., Budka, H., Rohwer, R.G., and Baskakov, I.V. (2010). Recombinant prion protein induces a new transmissible prion disease in WT animals. Acta Neuropathologica 119, 177-187. Maquat, L.E., and Carmichael, G.G. (2001). Quality Control of mRNA Function. Cell 104, 173-176. Martinez-Pastor, M.T., Marchler, G., Schuller, C., Marchler-Bauer, A., Ruis, H., and Estruch, F. (1996). The Saccharomyces cerevisae zinc finger proteins Msn2p and Msn4p are required for transcriptional induction through the stress-response element. The EMBO Journal 15, 2227-2235. Mascarenhas, M., Edwards-Ingram, L.C., Zeef, L., Shenton, L., Ashe, M.P., and Grant, C.M. (2008). Gcn4 is required for the response to peroxide stress in the yeast Saccharomyces cerevisiae. Molecular Biology of the Cell 19, 2995-3007. Mastrianni, J.A. (2010). The genetics of prion diseases. Genetics Medicine 12, 187-195. Michelitsch, M.D., and Weissman, J.S. (2000). A census of glutamine/asparagine-rich regions: Implications for their conserved function and the prediction of novel prions. Proceedings of the National Academy of Sciences of the United States of America 97, 11910- 11915. Mitchell, P., Petfalski, E., Shevchenko, A., Mann, M., and Tollervey, D. (1997). The Exosome: A Conserved Eukaryotic RNA Processing Complex Containing Multiple 3′→5′ Exoribonucleases. Cell 91, 457-466. Molina-Navarro, M.M., Castells-Roca, L., Belli, G., Garcia-Martinez, J., Marin-Navarro, J., Moreno, J., Perez-Ortin, J.E., and Herrero, E. (2008). Comprehensive transcriptional analysis of the oxidative response in yeast. Journal of Biological Chemistry 283, 17908-17918. Morano, K.A., Grant, C.M., and Moye-Rowley, W.S. (2011). The Response to Heat Shock and Oxidative Stress in Saccharomyces cerevisiae. Genetics 190, 1157-1195. Morell, M., Czihal, P., Hoffmann, R., Otvos, L., Avilés, F.X., and Ventura, S. (2008). Monitoring the interference of protein-protein interactions in vivo by bimolecular fluorescence complementation: the DnaK case. Proteomics 8, 3433-3442. Muhlrad, D., Decker, C.J., and Parker, R. (1994). Deadenylation of the unstable mRNA encoded by the yeast MFA2 gene leads to decapping followed by 5'-->3' digestion of the transcript. Genes & Development 8, 855-866.

195

Muhlrad, D., Decker, C.J., and Parker, R. (1995). Turnover mechanisms of the stable yeast PGK1 mRNA. Molecular and Cellular Biology 15, 2145-2156. Muller, E.G.D. (1996). A glutathione reductase mutant of yeast accumulates high levels of oxidized glutathione and requires thioredoxin for growth. Molecular Biology of the Cell 7, 1805-1813. Namy, O., Galopier, A., Martini, C., Matsufuji, S., Fabret, C., and Rousset, J.P. (2008). Epigenetic control of polyamines by the prion [PSI+]. Nature Cell Biology 10, 1069-75. Neale, M.J., and Keeney, S. (2006). Clarifying the mechanics of DNA strand exchange in meiotic recombination. Nature 442, 153-158. Ness, F., Ferreira, P., Cox, B.S., and Tuite, M.F. (2002). Guanidine hydrochloride inhibits the generation of prion "seeds" but not prion protein aggregation in yeast. Molecular and Cellular Biology 22, 5593-5605. O'Connell, J.D., Zhao, A., Ellington, A.D., and Marcotte, E.M. (2012). Dynamic reorganization of metabolic enzymes into intracellular bodies. Annual Review of Cell Developmental Biology 28, 89-111. Panzner, S., Dreier, L., Hartmann, E., Kostka, S., and Rapoport, T.A. (1995). Posttranslational protein transport in yeast reconstituted with a purified complex of Sec proteins and Kar2p. Cell 81, 561-570. Parker, R. (2012). RNA Degradation in Saccharomyces cerevisae. Genetics 191, 671. Pavitt, G.D., Ramaiah, K.V., Kimball, S.R., and Hinnebusch, A.G. (1998). eIF2 independently binds two distinct eIF2B subcomplexes that catalyze and regulate guanine- nucleotide exchange. Genes & Development 12, 514-526. Pechmann, S., Willmund, F., and Frydman, J. (2013). The Ribosome as a Hub for Protein Quality Control. Molecular Cell 49, 411-421. Pelechano, V., Wei, W., and Steinmetz, L.M. (2015). Widespread Co-translational RNA Decay Reveals Ribosome Dynamics. Cell 161, 1400-1412. Perutz, M.F. (1999). Glutamine repeats and neurodegenerative diseases: molecular aspects. Trends in Biochemical Sciences 24, 58-63. Powers, E.T., and Balch, W.E. (2013). Diversity in the origins of proteostasis networks- a driver for protein function in evolution. Nature review Molecular Cell Biology 14, 237-248. Preiss, T.W., and Hentze, M. (2003). Starting the protein synthesis machine: eukaryotic translation initiation. Bioessays 25, 1201-1211. Proud, C.G. (2005). eIF2 and the control of cell physiology. Seminar of Cell Developmental Biology 16, 3-12. Prusiner, S.B. (1982). Novel proteinaceous infectious particles cause scrapie. Science 216, 136. Rand, J.D., and Grant, C.M. (2006). The thioredoxin system protects ribosomes against stress-induced aggregation. Molecular Biology of the Cell 17, 387-401. Richter, K., Haslbeck, M., and Buchner, J. (2010). The heat shock response: life on the verge of death. Molecular Cell 40, 253-266. Rodriguez-Gabriel, M.A., Watt, S., Bahler, J., and Russell, P. (2006). Upf1, an RNA helicase required for nonsense-mediated mRNA decay, modulates the transcriptional response to oxidative stress in fission yeast. Molecular and Cellular Biology 26. Rospert, S., Rakwalska, M., and Dubaquie, Y. (2005). Polypeptide chain termination and stop codon readthrough on eukaryotic ribosomes. Reviews Physiological Biochemistry Pharmacology 155, 1-30. Rousseau, F., Serrano, L., and Schymkowitz, J.W. (2006). How evolutionary pressure against protein aggregation shaped chaperone specificity. Journal of Molecular Biology 355, 1037- 1047.

196

Routledge, K.E., Tartaglia, G.G., Platt, G.W., Vendruscolo, M., and Radford, S.E. (2009). Competition between intramolecular and intermolecular interactions in an amyloid-forming protein. Journal of Molecular Biology 389, 776-786. Ruepp, A., Zollner, A., Maier, D., Albermann, K., Hani, J., Mokrejs, M., Tetko, I., Guldener, U., Mannhaupt, G., Munsterkotter, M., et al. (2004). The FunCat, a functional annotation scheme for systematic classification of proteins from whole genomes. Nucleic Acids Research 32, 5539-5545. Saibil, H. (2013). Chaperone machines for protein folding, unfolding and disaggregation. Nature reviews Molecular Cell Biology 14, 630-642. Saito, S., Hosoda, N., and Hoshino, S. (2013). The Hbs1-Dom34 protein complex functions in nonstop mRNA decay in mammalian cells. Journal of Biological Chemistry 288, 17832- 17843. Schosserer, M., Minois, N., Angerer, T.B., Amring, M., Dellago, H., Harreither, E., Calle- Perez, A., Pircher, A., Gerstl, M.P., Pfeifenberger, S. (2015). Methylation of ribosomal RNA by NSUN5 is a conserved mechanism modulating organismal lifespan. Nature Communications 6, 6158. Serio, T.R., and Lindquist, S.L. (1999). [PSI+]: An Epigenetic Modulator of Translation Termination Efficiency. Annual Review of Cell and Developmental Biology 15, 661-703. Shenton, D., and Grant, C.M. (2003). Protein S-thiolation targets glycolysis and protein synthesis in response to oxidative stress in the yeast Saccharomyces cerevisiae. Biochemical Journal 374, 513-519. Shenton, D., Smirnova, J.B., Selley, J.N., Carroll, K., Hubbard, S.J., Pavitt, G.D., Ashe, M.P., and Grant, C.M. (2006). Global translational responses to oxidative stress impact upon multiple levels of protein synthesis. Journal of Biological Chemistry 281, 29011-29021. Sherman, M.Y., and Goldberg, A.L. (2001). Cellular defenses against unfolded proteins: a cell biologist thinks about neurodegenerative diseases. Neuron 29, 15-32. Shirley, R.L., Lelivelt, M.J., Schenkman, L.R., Dahlseid, J.N., and Culbertson, M.R. (1998). A factor required for nonsense-mediated mRNA decay in yeast is exported from the nucleus to the cytoplasm by a nuclear export signal sequence. Journal of Cell Science 111, 3129. Shoemaker, C.J., Eyler, D.E., and Green, R. (2010). Dom34:Hbs1 promotes subunit dissociation and peptidyl-tRNA drop off to initiate no-go decay. Science 330, 369-372. Shoemaker, C.J., and Green, R. (2012). Translation drives mRNA quality control. Nature Structural Molecular Biology 19, 594-601. Sideri, T.C., Koloteva-Levine, N., Tuite, M.F., and Grant, C.M. (2011). Methionine oxidation of Sup35 protein induces formation of the [PSI+] prion in a yeast peroxiredoxin mutant. Journal of Biological Chemistry 286, 38924-38931. Simms, C.L., Hudson, B.H., Mosior, J.W., Rangwala, A.S., and Zaher, H.S. (2014). An active role for the ribosome in determining the fate of oxidized mRNA. Cell Reports 9, 1256- 1264. Siwaszek, A., Ukleja, M., and Dziembowski, A. (2014). Proteins involved in the degradation of cytoplasmic mRNA in the major eukaryotic model systems. RNA Biology 11, 1122-1136. Soto, C., and Estrada, L.D. (2008). Protein misfolding and neurodegeneration. Archives of Neurology 65, 184-189. Speldewinde, S.H., Doronina, V.A., and Grant, C.M. (2015). Autophagy protects against de novo formation of the [PSI+] prion in yeast. Molecular Biology of the Cell 26, 4541-4551. Spokoini, R., Moldavski, O., Nahmias, Y., England, J.L., Schuldiner, M., and Kaganovich, D. (2012). Confinement to organelle-associated inclusion structures mediates asymmetric inheritance of aggregated protein in budding yeast. Cell Reports 2, 738-747.

197

Stadtman, E.R., and Levine, R.L. (2000). Protein Oxidation. Annuals of the New York Academy of Sciences 899, 191-208. Stahl, N., Baldwin, M.A., Teplow, D.B., Hood, L., Gibson, B.W., Burlingame, A.L., and Prusiner, S.B. (1993). Structural studies of the scrapie prion protein using mass spectrometry and amino acid sequencing. Biochemistry 32, 1991-2002. Stansfield, I., Akhmaloka, and Tuite, M.F. (1995). A mutant allele of the SUP45 (SAL4) gene of Saccharomyces cerevisiae shows temperature-dependent allosuppressor and omnipotent suppressor phenotypes. Current Genetics 27, 417-426. Stefani, M., and Dobson, C.M. (2003). Protein aggregation and aggregate toxicity: new insights into protein folding, misfolding diseases and biological evolution. Journal of Molecular Medicine 81, 678-699. Tartaglia, G.G., Pechmann, S., Dobson, C.M., and Vendruscolo, M. (2007). Life on the edge: a link between gene expression levels and aggregation rates of human proteins. Trends Biochemistry Scientific 32, 204-206. Tartaglia, G.G., and Vendruscolo, M. (2009). Correlation between mRNA expression levels and protein aggregation propensities in subcellular localisations. Molecular Biosystems 5, 1873-1876. Tartaglia, G.G., and Vendruscolo, M. (2010). Proteome-level interplay between folding and aggregation propensities of proteins. Journal of Molecular Biology 402, 919-928. Tarun, S.Z., Wells, S.E., Deardorff, J.A., and Sachs, A.B. (1997). Translation initiation factor eIF4G mediates in vitro poly(A) tail-dependent translation. Proceedings of the National Academy of Sciences of the United States of America 94, 9046-9051. Thomas, J.A., and Mallis, R.J. (2001). Aging and oxidation of reactive protein sulfhydryls. Experimental Gerontology 36, 1519-1526. Tomoyasu, T., Mogk, A., Langen, H., Goloubinoff, P., and Bukau, B. (2001). Genetic dissection of the roles of chaperones and proteases in protein folding and degradation in the Escherichia coli cytosol. Molecular Microbiology 40, 397-413. Treusch, S., and Lindquist, S. (2012). An intrinsically disordered yeast prion arrests the cell cycle by sequestering a spindle pole body component. Journal Cell Biology 197, 369-379. True, H.L., Berlin, I., and Lindquist, S.L. (2004). Epigenetic regulation of translation revelas hidden genetic variation to produce complex traits. Nature 431, 184-187. True, H.L., and Lindquist, S.L. (2000). A yeast prion provides a mechanism for genetic variation and phenotypic diversity. Nature 407, 477-483. Tsanova, B., Spatrick, P., Jacobson, A., and van Hoof, A. (2014). The RNA exosome affects iron response and sensitivity to oxidative stress. RNA 20, 1057-1067. Tsuboi, T., Kuroha, K., Kudo, K., Makino, S., Inoue, E., Kashima, I., and Inada, T. (2012). Dom34:hbs1 plays a general role in quality-control systems by dissociation of a stalled ribosome at the 3' end of aberrant mRNA. Molecular Cell 46, 518-529. Tuite, M.F., Mundy, C.R., and Cox, B.S. (1981). Agents that cause a high frequency of genetic change from [PSI+] to [psi-] in Saccharomyces cerevisiae. Genetics 98, 691-711. Tycko, R. (2006). Molecular structure of amyloid fibrils: insights from solid-state NMR. Quarterly Reviews of Biophysics 39, 1-55. Tyedmers,J., Madariaga, M.L., and Lindquist, S. (2008). Prion switching in response to environmental stress. PloS Biology 6, e294. Tyedmers, J., Mogk, A., and Bukau, B. (2010). Cellular strategies for controlling protein aggregation. Nature Review Molecular Cell Biology 11, 777-788. Valastyan, J.S., and Lindquist, S. (2014). Mechanisms of protein-folding diseases at a glance. Disease Models & Mechanisms 7, 9.

198

Valko, M., Leibfritz, D., Moncol, J., Cronin, M.T.D., Mazur, M., and Telser, J. (2007). Free radicals and antioxidants in normal physiological functions and human disease. The International Journal of Biochemistry & Cell Biology 39, 44-84. van Hoof, A., Frischmeyer, P.A., Dietz, H.C., and Parker, R. (2002). Exosome-mediated recognition and degradation of mRNAs lacking a termination codon. Science 295, 2262- 2264. van Hoof, A., Lennertz, P., and Parker, R. (2000). Three conserved members of the RNase D family have unique and overlapping functions in the processing of 5S, 5.8S, U4, U5, RNase MRP and RNase P RNAs in yeast. The EMBO Journal 19, 1357-1365. Veal, E.A., Day, A.M., and Morgan, B.A. (2007). Hydrogen peroxide sensing and signaling. Molecular Cell 26, 1-14. Vendruscolo, M. (2012). Proteome folding and aggregation. Current Opinion of Structural Biology 22, 138-143. Ventura, S., and Villaverde, A. (2006). Protein quality in bacterial inclusion bodies. Trends in Biotechnology 24, 179-185. von der Haar, T., and Tuite, M.F. (2007). Regulated translational bypass of stop codons in yeast. Trends Microbiology 15, 78-86. Wang, F., Wang, X., Yuan, C.-G., and Ma, J. (2010). Generating a prion with bacterially expressed recombinant prion protein. Science 327, 1132-1135. Wang, J., Farr, G.W., Hall, D.H., Li, F., Furtak, K., Dreier, L., and Horwich, A.L. (2009). An ALS-linked mutant SOD1 produces a locomotor defect associated with aggregation and synaptic dysfunction when expressed in neurons of Caenorhabditis elegans. PLOS Genetics 5, e1000350. Wang, L., Lewis, M.S., and Johnson, A.W. (2005). Domain interactions within the Ski2/3/8 complex and between the Ski complex and Ski7p. RNA 11, 1291-1302. Wang, L., Maji, S.K., Sawaya, M.R., Eisenberg, D., and Riek, R. (2008). Bacterial inclusion bodies contain amyloid-like structure. PLoS Biology 6, e195. Wang, Q., Zhuravleva, A., and Gierasch, L.M. (2011). Exploring weak, transient protein- protein interactions in crowded in vivo environments by in-cell NMR spectroscopy. Biochemistry 50, 9225-9236. Wang, W., Czaplinski, K., Rao, Y., and Peltz, S.W. (2001). The role of Upf proteins in modulating the translation readthrough of nonsense-containing transcripts. The EMBO Journal 20, 880-890. Washburn, M.P., Wolters, D., and Yates, J.R.r. (2001). Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nature Biotechnology 19, 242-247. Weids, A.J., and Grant, C.M. (2014). The yeast peroxiredoxin Tsa1 protects against protein- aggregate-induced oxidative stress. Journal of Cell Science 127, 1327-1335. Weids, A.J., Ibstedt, S., Tamas, M.J., and Grant, C.M. (2016). Distinct stress conditions result in aggregation of proteins with similar properties. Scientific reports 6, 24554. Wickner, R.B. (1994). [URE3] as an altered URE2 protein: evidence for a prion analog in Saccharomyces cerevisiae. Science 264, 566-5699. Williams, I., Richardson, J., Starkey, A., and Stansfield, I. (2004). Genome-wide prediction of stop codon readthrough during translation in the yeast Saccharomyces cerevisiae. Nucleic Acids Res 32, 6605-6616. Wilson, M.A., Meaux, S., and van Hoof, A. (2007). A genomic screen in yeast reveals novel aspects of nonstop mRNA metabolism. Genetics 177, 773-784. Wurtmann, E.J., and Wolin, S.L. (2009). RNA under attack: Cellular handling of RNA damage. Critical Reviews in Biochemistry and Molecular Biology 44, 34-49.

199

Yoshimura, Y., Lin, Y., Yagi, H., Lee, Y.-H., Kitayama, H., Sakurai, K., So, M., Ogi, H., Naiki, H., and Goto, Y. (2012). Distinguishing crystal-like amyloid fibrils and glass-like amorphous aggregates from their kinetics of formation. Proceedings of the National Academy of Sciences of the United States of America 109, 14446-14451.

200

8.0 Appendix Appendix 1 List of genes encoding proteins found to aggregate in WT

TEF2 GUS1 SAC6 DED81 FUN12 FUR1 RPN6 PGK1 ILV5 ATP1 TDH2 DBP2 RPS23A SIS1 YEF3 NDE1 RPS0B ARG5 RPN2 NAT1 RPS27B YMLO39W PFK1 YDJ1 RPS19A ERG9 DBP3 GPM1 RPS3 THS1 FRM2 ERG6 CDC48 NIC96 SSA1 HSC82 PFK2 NOP56 SPF1 RPL19B CBR1 MET17 URA2 DED1 GLC7 LYS12 LEU1 NDI1 RPL3 RPS1B RPL27A ARF2 SEC18 MCK1 SSB1 GCN1 RLI1 RNA1 YKR018C YME2 CDC19 HXK2 SEC53 TRP5 SNQ2 RPL15A ACC1 FAS1 RPS31 AIF1 RPL33A RPS17B PET9 SCP160 ATP2 SEC23 VMA5 COX15 PMA1 CYS4 RPS15 IDH2 RPS25B BAT1 ENO2 TIF2 GSH1 AHA1 ECM29 SUB2 RPS4B ADH1 OAC1 RVB1 ASN1 CCT2 FBA1 RPG1 MKT1 SRP54 CLU1 RPL35B ACT1 HSP104 DPM1 RPS28B SAC1 BBC1 EFT2 RPL20A RPL30 NOG2 PHB1 LSP1 PDC1 ACS2 KES1 RPL8B AIM9 YNL134C RPP0 RPS2 SAM1 RPL10 SUP45 SRP72 RPS6B RPL1B IMD2 CHC1 RPL25 CRM1 RPS18A GFA1 ENO1 ARO1 UTP10 DHH1 RPL11B LYS20 SSA2 GRS1 NEW1 UBP6 SSE1 ARG1 RPL7B RPS20 XRN1 NFS1 RRP5 RPS24A RPL4B RPL6A MYO2 VTC2 RPS8A ALA1 FAS2 RPS22A GSP2 NOP58 GRE2 ERG20 ALD6 LEU4 CAM1 NMD3 ARB1 RPL18B RPL16B COP1 PHB2 DNM1 TEF4 YHR020W YDR341C RPL26B SPT16 GCD11 OLA1 NIP1 SAM2 CDC39 TIF4631 RPL31A RPS5 PRT1 RPS14B RPS16B DPS1 TRP2 VMA2 RPL9A ILS1 VAS1 RPN3 SRP68 AAD10 PIL1 ASN2 YCF1 RPL38 TCB2 HSP78 MIR1 VMA1 YKL071W CTR9 NUG1 ILV2 RPS7A KAR2 NOP1 HYP2 PDR12 RPS9A POR1 VPS1 RPL2A YML131W RHO1 PAB1 CCT4 IMD3 ZUO1 STE24 YHM2 RPS13 SSZ1 RPL12B RPL17A HAS1 ARF1 QCR2 RPL5 RPL14B FKS1 SEC21 CBF5 KAP123 RPL32 ACO1 RPS11B RPL24B SSC1

201

Appendix 2 List of genes encoding proteins found to aggregate in ski7 mutant

TEF2 ALD6 ARG1 CDC48 PHB2 RPL28 SSA2 ENO1 RPG1 RPS24A RPL23A RPS31 CCT2 RPL4B PGK1 HSP104 ALA1 RPL19B DPL1 RPL35B SAM2 YEF3 RPL20A COP1 RPL5 ATP2 IDH2 RPS14B YMLO39W CHC1 RPL26B DED81 HTS1 SER1 ILS1 RPS3 ACS2 ERG20 LEU1 SPT16 TCB3 LEU4 HSC82 ARO1 SSC1 COR1 TIF4631 LSP1 CDC39 URA2 RPS2 RPS16B FAA1 RPS15 GET3 TOM1 RPL3 RPS6B VAS1 RPL32 DPS1 SES1 PYC1 SSB1 RPS18A YCF1 SEC18 PSA1 KES1 RPL12B CDC19 RPL16B RPL18B YKR018C HYP2 RPL16A RPL14B ACC1 RPL11B YHR020W SAC6 VMA13 ATP4 ACO1 PET9 SSE1 KAR2 RPL33A RPL24B YPR1 ADE3 PMA1 RPS8A YKL071W VMA5 GSH1 SEY1 TDH2 ENO2 GRE2 NIP1 HIS4 TRP5 HIS5 ARG5 RPL8B ARB1 NOP1 MES1 RPS23A NFS1 SUP35 RPS4B TEF4 RPL2A RPS25B OAC1 CDC33 SPF1 FBA1 OLA1 PRT1 ECM29 CCT6 LYS1 RPL21A ACT1 RPS5 VPS1 RPS19A AIF1 HHF1 TAL1 EFT2 GRS1 ZUO1 ATP1 TCP1 NOP58 GUT2 PDC1 YDR341C RPL9A ERG6 RPT1 SAM1 ZWF1 RPP0 VMA2 SEC26 ASN1 RPL36B IDH1 SEC53 GUS1 RPS20 OYE2 CLU1 ILV1 GCD11 ARF2 ILV5 AAD10 GLN4 YDJ1 VPH1 GLY1 CAM1 NDE1 HSP78 RPL17A FRM2 NAT1 KGD2 RNA1 PFK1 ILV2 PIL1 SAC1 SEC23 IMD2 PIM1 THS1 RPS9A IMD3 RPT6 DBP3 KAP95 PRP4 PFK2 PAB1 MIR1 PHB1 GSF2 FAA4 RPN3 DED1 KRS1 RPS7A AIM9 MAE1 CPR6 IPP1 RPS1B ASN2 RPS11B ACO2 DPM1 PUF6 ARO3 GCN1 VMA1 CDC60 NOP56 MSC7 RVB2 YML131W HXK2 RPS13 POR1 SUP45 BFR1 MRT4 SSE2 RPL7B QCR2 FUN12 RPL25 NDI1 NUG1 STE24 FAS1 RPL6A CCT4 GLC7 MCK1 RBG2 FRS1 SCP160 KAP123 ADE5 UBA1 RPL15A CDC10 HAS1 CYS4 RPS22A DBP2 NEW1 RPS17B HTB2 SEC21 TIF2 GUA1 SSZ1 CCT8 SEC14 LSB3 FUR1 ADH1 RPL1B RPN2 RPL27A ARC1 YJU3 FRS2 RPL10 GFA1 ERG9 KGD1 BAT1 CYT1 SAH1 FAS2 LYS20 ADE6 GSP2 FCJ1 GTT2 RPN9

202

SSK2 PRX1 FRA1 CPR1 MKT1 TRP2 ILV6 TUM1 CPA2 RPT3 UBP6 MIS1 RSP5 MSS116 GPD1 TMA19 YME2 UFD2 KRE33 SIS1 UGP1 SRP68 SUI3 MDH3 SEC27 NSR1 DNM1 RNR2 GLN1 LYS9 NUP188 GPX2 ADE4 SDH1 SSD1 RPS10B RPL13B TIF34 MET10 ARX1 GLK1 ARO4 RPL31A PAA1 CCT3 YER156C ERG1 RPL8A NAM9 KAP104 RKR1 HOM3 RNR4 PRO2 YRA1 RPL6B YIL108W YPT32 NOP7 TSA1 RPN6 DOM34 SEC17 URA8 PDA1 RPS28B RVS161 CCT5 CRM1 TPD3 ARF1 YHM2 DHH1 MET6 RPL34B SSA1 HXK1 RPT4 OSH7 CCP1 DUG1 IRC24 PAT1 PRO1 CCT7 SEC61 ADO1 STT3 NMD3 RPN11 DBP5 YMR226C TUB2 CBF5 LYS12 GVP36 ATP5 RPF2 PHO88 YMR090W NCP1 TOM40 PGI1 RPS29A NOG1 RTK1 URA7 LSC1 TYS1 EMC1 AAD3 TDH3 RPS26B EMP24 CBR1 RPL43B PUB1 SUB2 IDP1 APA1 SEC24 UTP4 YHI9 AHA1 WRS1 RRP5 RVB1 AAP1 FKS1 RPL30 BAT2 SNQ2 RPS12 YNR021W RPS0B YNL134C ERG26 RLI1 VAC8 SHM2 RPL38 SEC4 PGM1 TOM70 RPN1 MVP1 TRM3 ARP2

203

Appendix 3 List of genes encoding proteins found to aggregate in ski8 mutant

TEF2 HSP104 ALA1 ASN1 RSP5 RPL26B TIF4631 ENO1 RPL20A ERG20 CLU1 ARC1 CDC39 RPS15 PGK1 CHC1 VAS1 RPS0B BAT1 RPS16B RPN3 YEF3 ACS2 YCF1 YDJ1 SUB2 PYC1 RPL38 YMLO39W ARO1 RPL18B FRM2 FCJ1 KAR2 DYN1 RPS3 RPS2 YHR020W RPT6 RPL28 VPS1 ARO3 HSC82 RPS6B YKL071W ACO2 CCT2 SEC26 VMA13 URA2 RPS18A NIP1 NOP56 IDH2 GLN4 STE24 RPL3 RPL16B NOP1 RPL25 SEC27 RPL17A FRS1 SSB1 RPL11B RPL2A GLC7 TOM70 IMD3 URA7 CDC19 SSE1 PRT1 LYS12 RPL30 RPS7A FUR1 ACC1 RRP5 ZUO1 PHO88 RNR4 RPL12B RPS23A PET9 RPS8A RPL9A NEW1 SRP72 UTP20 CCT6 PMA1 GRE2 OYE2 ZWF1 ILV6 CCT4 VPS13 ENO2 ARB1 PIL1 RPL27A UBP6 RPN2 ILV1 RPL8B TEF4 FKS1 RLI1 GPD1 ERG9 MKT1 RPS4B OLA1 MIR1 SEC53 CIT2 RPL14B NIC96 FBA1 RPS5 RPS11B KGD1 SAM1 MDN1 MAE1 ACT1 GRS1 CDC60 GSP2 KGD2 TDH2 DPM1 EFT2 YDR341C POR1 RPS31 SIS1 FAA1 MCK1 PDC1 VMA2 FUN12 SPT16 RNR2 RPL32 YME2 RPP0 RPS20 ADE5 DPS1 IMD2 SEC18 RPL15A GUS1 AAD10 DBP2 PSA1 RPN5 RPL33A SEC14 ILV5 HSP78 SSZ1 GND1 HXT3 HIS4 MDM38 NDE1 ILV2 ADE6 HYP2 PRX1 RPS19A SEC63 PFK1 RPS9A CDC48 YML131W PAA1 RPL21A RPL35B THS1 PAB1 RPL23A SSE2 RTN1 SAC1 GLN1 PFK2 KRS1 RPL19B SEC21 SEC17 DBP5 SEC24 DED1 ASN2 RPL5 GSH1 SSA1 PHB1 AHA1 GCN1 RPS13 DED81 TRP5 ARF1 AIM9 TCB3 HXK2 QCR2 COR1 OAC1 SSA2 GUT2 RPL13B FAS1 RPL6A YKR018C SAH1 RPS1B UBA1 LSP1 SCP160 KAP123 SAC6 AIF1 RPL7B CCT8 GET3 CYS4 RPS22A VMA5 RPT1 RPL4B MYO2 ISW1 TIF2 LEU4 MES1 RPL36B SAM2 ARF2 RPS12 ADH1 GUA1 RPS25B VPH1 RPS14B CAM1 CCT3 RPL10 RPL1B SUP35 DBP3 ILS1 PHB2 LYS2 FAS2 LYS20 ATP1 GSF2 VMA1 DPL1 KES1 ALD6 ARG1 ERG6 CBR1 GFA1 RNA1 RPL16A RPG1 RPS24A SPF1 NDI1 COP1 HTS1 CRM1

204

CDC53 ERG3 ARP2 YPR1 CSE1 GCD11 HXK1 MDR1 SRP68 SEY1 CSR1 TSA1 LCB2 AIM46 YMR226C DUG1 YDR476C VAR1 NFS1 RPL6B PRE3 CDC33 RBG2 FLR1 SNF2 LCB1 SSC1 SUI3 MRPS17 ACO1 VTC2 EMC2 ADE3 HHF1 CDC28 LEU1 EHT1 RHO1 ADO1 SEC4 MRP1 SUP45 RPN1 YRA1 TAL1 NMD3 DOM34 ATP2 ERG27 PDR1 PGI1 TUB2 RVS161 FRS2 ARP3 VTC4 RPO21 DNM1 ISD11 SEC23 TUM1 SDS23 RVB1 NOG1 CTP1 GLK1 RPS26B TRL1 SES1 NOP2 RPT4 YNL134C TMA19 UTP8 RPN6 YPK2 SAR1 NOP58 LSG1 SXM1 STI1 MDH3 TUB3 AAP1 NUP133 ERV25 CPR6 POL5 EMP24 TRP2 GPX2 PUB1 GPM1 YPR091C APA1 DLD3 RPS10B YPP1 ASC1 ARX1 AST1 TRR1 RVB2 MRPS18 RPS1A PMR1 SNQ2 YMR090W HOM3 ECM29

205

Appendix 4 List of genes encoding proteins found to aggregate in upf2 mutant

TEF2 ALD6 RPL26B SEC18 FRS1 VAC8 ERG9 PGK1 RPG1 ERG20 YKR018C SEC21 NFS1 RPL14B YEF3 HSP104 SSC1 SAC6 FUR1 CDC33 ACO1 YMLO39W RPL20A RPS16B RPL33A RPS23A VTC2 TDH2 RPS3 CHC1 VAS1 HIS4 RPB2 NOP58 ARG5 HSC82 ACS2 YCF1 MES1 OAC1 IDH1 VMA5 URA2 ARO1 RPL18B RPS25B RPL36B DNM1 SPF1 RPL3 RPS2 YHR020W SUP35 ILV1 GCD11 RPL21A SSB1 RPS6B KAR2 RPS19A MKT1 SSD1 RPT6 CDC19 RPS18A YKL071W ATP1 SEC23 MIS1 UTP10 ACC1 RPL16B NIP1 ERG6 DBP3 GLY1 ZWF1 PET9 RPL11B NOP1 ASN1 NIC96 KGD2 CCT8 PMA1 SSE1 RPL2A RPS0B GSF2 PDH1 ARF2 SSA2 RRP5 PRT1 YDJ1 MAE1 SIS1 PHB2 ENO2 RPS8A VPS1 FRM2 DPM1 MDH3 DPL1 RPL8B GRE2 ZUO1 SAC1 MSC7 RNR2 PIM1 RPS4B ARB1 RPL9A DBP5 BFR1 IMD2 HTS1 FBA1 TEF4 SEC26 PHB1 NDI1 YPR091C TIF4631 ACT1 OLA1 GLN4 AIM9 RSP5 RPS10B PRP4 EFT2 RPS5 RPL17A ACO2 MCK1 FAA4 RPN3 PDC1 GRS1 PIL1 NOP56 RPL15A RPL6B PSA1 RPP0 VMA2 MIR1 SUP45 RPS17B TSA1 RPL38 GUS1 RPS20 RPS7A RPL25 SEC14 CYT1 ARO3 ILV5 AAD10 RPS11B TAL1 SUB2 MAS1 STE24 NDE1 HSP78 UTP20 GLC7 FCJ1 ENO1 HAS1 PFK1 ILV2 POR1 LYS12 RPL35B RPL7B RPL24B THS1 RPS9A FUN12 UBA1 BBC1 RPL4B CCT6 PFK2 PAB1 ADE5 RPL27A IDH2 YDR341C SAH1 DED1 KRS1 DBP2 RLI1 SEC27 SAM2 VPS13 RPS1B VMA1 SSZ1 SEC53 GLN1 RPS14B TCP1 GCN1 RPS13 RPN2 KGD1 AHA1 ILS1 RPN9 HXK2 QCR2 ADE6 GSP2 RVB1 ASN2 OYE3 FAS1 RPL6A CDC48 CAM1 RPL30 GUA1 AAD3 SCP160 KAP123 RPL23A RPS31 GET3 CDC39 YME2 CYS4 RPS22A RPL19B RNA1 RPS12 OYE2 COX15 TIF2 LEU4 RPL5 ATP2 SES1 IMD3 MSC6 ADH1 RPL1B DED81 RPS15 KES1 RPL12B ARC1 RPL10 GFA1 LEU1 DPS1 YNL134C CDC60 BAT1 FAS2 LYS20 COR1 HYP2 SRP72 CCT4 SEC63 ALA1 ARG1 FAA1 VMA13 ILV6 COP1 SEC24

206

ATM1 EMW1 RSC8 NAM9 ADE4 YPK2 PDX3 ATP4 RPL13B LSG1 YMR090W UBP6 LSP1 BAT2 RFC2 AFG3 DCP2 YNR021W YCL057C-A SEC4 YCL019W CPR6 YPT7 SAM1 CCT3 MVP1 PHO86 YME1 RNR4 PRX1 SNQ2 CPR1 YIL108W PAA1 ECM29 QRI1 FRA1 PDR16 CLU1 TUM1 CRM1 LEU9 NEW1 RPS27B SER33 SLA2 SPT16 MRT4 NOP12 MET5 CTR9 TPA1 DHH1 SRP68 GSH1 CDC28 YDR061W ERG1 AIF1 YKT6 CDC53 LYS9 NAT1 FSF1 SPT6 SRP101 CCT2 SSA3 SRP54 ERG3 SPT5 TIF35 YPR1 TIF34 NUP157 GBP2 HXK1 NUG1 RRP12 CBF5 CIT2 ARO4 RPN1 CCR4 HIS5 SEC13 RPL31A RPS29A TRM1 AIM46 NUP133 OST3 LYS1 CMP2 KAP95 YNL040W SUI3 YER156C HXT3 YGR001C CCT7 YGR266W RVB2 SOL2 HHF1 LAT1 RPL8A EHT1 HTB2 UBP3 GDH2 GDE1 YDR476C AFG1 EMC2 RHO1 ARP2 GGC1 YHM2 ERG27 YDL124W SSA1 TUB2 GGA2 GPD2 ATP5 CCT5 RPS1A ARP3 SLT2 GTT2 NCP1 PMT2 OSH7 HOM6 MRP1 FLR1 PSD2 DOM34 PHO88

207

Appendix 5 List of genes encoding proteins found to aggregate in hbs1 mutant

TEF2 ACS2 RPS7A SUB2 NIP1 ACO2 RPT1 PGK1 RPS2 RPS11B FCJ1 NOP1 TAL1 RPL36B YEF3 RPS6B POR1 RPL28 RPL2A GLC7 ILV1 YMLO39W RPS18A FUN12 IDH2 ZUO1 PHO88 VPH1 RPS3 RPL11B ERG9 RVB1 OYE2 GUT2 NAT1 HSC82 SSE1 RPL23A YNL134C GLN4 UTP10 CPA2 URA2 RRP5 ACO1 ILV6 PIL1 ZWF1 SEC23 RPL3 RPS8A RPL5 AFG3 IMD3 CCT8 DBP3 SSB1 ARB1 DED81 CDC33 RPL12B RPL27A MAE1 CDC19 TEF4 RPL33A NOP58 CDC60 RLI1 MSC7 ACC1 OLA1 VMA5 SAM1 CCT4 SEC53 BFR1 PET9 RPS5 RPS19A NMD3 ADE5 KGD1 RSP5 PMA1 GRS1 ATP1 MDH3 DBP2 ARF2 YME2 ENO2 VMA2 CLU1 SHM2 SSZ1 CAM1 RPS17B RPL8B RPS20 RPS0B RPS28B RPN2 DPL1 COX15 RPS4B AAD10 FRM2 ENO1 ADE6 RNA1 SEC14 FBA1 ILV2 SAC1 SSA2 CDC48 PIM1 ARC1 ACT1 RPS9A PHB1 HXK2 RPL19B HTS1 BAT1 EFT2 PAB1 AIM9 RPL7B RPL14B SPT16 UGP1 PDC1 RPS13 NOP56 RPL4B ADE3 TIF4631 CCT2 RPP0 QCR2 SUP45 FAS2 LEU1 DPS1 RPL35B GUS1 RPL6A RPL25 RPG1 COR1 PRP4 SEC27 ILV5 KAP123 LYS12 ARO1 TDH2 RPN3 GLN1 NDE1 RPS22A NEW1 RPL16B FAA1 PSA1 SEC24 PFK1 LEU4 MYO2 GRE2 RPL32 IPP1 RPL13B THS1 GUA1 GSP2 YDR341C ARG5 RPL38 TOM70 PFK2 RPL1B ATP2 SAM2 SEC18 ARO3 LSP1 DED1 GFA1 RPS15 RPS14B SAC6 SSE2 GLK1 RPS1B LYS20 HYP2 HSP78 HIS4 STE24 RPS12 GCN1 ARG1 YML131W ILS1 MES1 FRS1 CCT3 FAS1 RPS24A VMA13 KRS1 RPS25B HAS1 SES1 SCP160 ERG20 URA7 ASN2 SUP35 SEC21 LYS2 CYS4 RPS16B GSH1 VMA1 ERG6 RPL24B KES1 TIF2 RPL18B OAC1 ALA1 SPF1 FUR1 RPL16A ADH1 YHR020W MKT1 COP1 ADO1 TRP5 SPT5 RPL10 PRT1 GSF2 RPL26B ASN1 RPS23A RHR2 ALD6 VPS1 CBR1 SSC1 YDJ1 RPB2 SER33 HSP104 RPL9A DPM1 VAS1 RPL21A CCT6 VAC8 RPL20A RPL17A MCK1 PYC1 DBP5 AIF1 MAS2 CHC1 MIR1 RPL15A KAR2 RPT6 TCP1 YDR061W

208

HXK1 YDL124W DUG1 URA8 UBP6 PRS3 GPD1 FSF1 CIT2 NOP7 TRM1 HFD1 CWH43 TPS1 LYS1 PDX3 CCT7 ERV25 HHF1 YCF1 EHT1 FKS1 ARP2 SNQ2 ERG27 BBC1 TUB2 GCD11 CPR1 RBG2 HMG1 TDH3 QRI1 YKL071W NCP1 SEC26 TYS1 AAD3 SSD1 DHH1 RPS26B SEC4 RPL43B LYS4 RPL31A PRX1 TMA19 GCD6 BAT2 SSA1 GCN20 SAR1 YNR021W KAP95 TUF1 PAA1 LEU9 SLA2 MET5 MRT4 ERG1 SRP101 ARO4 RPL6B LCB1 KAP104

209