The Transcription Factor ATF5 Protects Cortical Neurons from Toxicity Induced by Prostaglandin J2

City University of New York (CUNY) CUNY Academic Works

All Dissertations, Theses, and Capstone Projects Dissertations, Theses, and Capstone Projects

2-2020

Fanny Cheung The Graduate Center, City University of New York

How does access to this work benefit ou?y Let us know!

More information about this work at: https://academicworks.cuny.edu/gc_etds/3616 Discover additional works at: https://academicworks.cuny.edu

This work is made publicly available by the City University of New York (CUNY). Contact: [email protected]

THE TRANSCRIPTION FACTOR ATF5

PROTECTS CORTICAL NEURONS FROM

TOXICITY INDUCED BY PROSTAGLANDIN

Fanny Cheung

A dissertation submitted to the Graduate Faculty in Biology in partial fulfillment of the

requirements for the degree of Doctor of Philosophy, The City University of New York

2020

Fanny Cheung

ii The transcription factor ATF5 protects cortical neurons from toxicity induced by prostaglandin J2.

Fanny Cheung

This manuscript has been read and accepted for the Graduate Faculty in Biology in satisfaction of the dissertation requirement for the degree of Doctor of Philosophy.

Chair of Examining Committee:

______Date Dr. Patricia Rockwell (Hunter College, CUNY)

Executive Officer:

______Date Dr. Christine Li (City College, CUNY)

Supervisory Committee:

______Date Dr. Maria E. Figueiredo-Pereira (Hunter College, CUNY)

______Date Dr. Peter Serrano (Hunter College, CUNY)

______Date Dr. Dan McCloskey (College of Staten Island, CUNY)

______Date Dr. Luena Papa (Icahn School of Medicine at Mount Sinai)

THE CITY UNIVERSITY OF NEW YORK

iii ABSTRACT

The transcription factor ATF5 protects cortical neurons from toxicity induced by prostaglandin J2.

Fanny Cheung

Advisor: Dr. Patricia Rockwell

Mitochondrial dysfunction has been recognized to play a central role in the pathogenesis of neurodegenerative disorders, including Parkinson’s and Alzheimer’s disease. Transcription factor ATF5 has been shown to mediate an endogenous stress pathway known as the UPRmt, a prosurvival mechanism amidst malfunctioning mitochondria. However, the ATF5/UPRmt protective pathway associated with mitochondrial impairment in neurons remains to be elucidated. To decipher the regulatory mechanisms involved with mediating neuroprotection, the effect of ATF5 was examined on the mitochondrial dysfunction induced by the proinflammatory prostaglandin PGJ2 in rat cortical neurons. ATF5 overexpression protected the effects induced by PGJ2 by attenuating HO-1 induction, Caspase-3 cleavage, cell loss, ROS production, and a collapse of the mitochondrial membrane potential ΔΨ . With the exception of Capsase-3 m activation, silencing ATF5 escalated PGJ2 induced toxicity thereby attenuating the prosurvival effects of ATF5 overexpression. Detecting protein expression only in the presence of PGJ2 treatment along with constant levels of ATF5 transcript, whether PGJ2 stress is present or not, proposes PGJ2 stress induces protein expression by stabilizing ATF5 transcript for translation.

PGJ2 stress elicited ATF5 trafficking between the mitochondria and nucleus suggests organelle

iv partitioning accompanies the protection promoted by ATF5. Inhibition of the MEK/ERK signaling pathway with the selective inhibitor U0126 revealed yet another level of anti-apoptotic regulation by ATF5. A blockade of ERK activation prevented the protection by ATF5 by increasing the loss in cell viability induced by PGJ2. These findings suggest that ERK activation plays a role in AFT5 mediated protection in cortical neurons. In assessing mitochondrial function and glycolysis respiration, our findings suggest ATF5 protection against PGJ2 toxicity involves modulating mitochondrial functions. Reduction of cellular respiration demand by ATF5 observed in PGJ2 stressed neurons may relieve damaged mitochondria from the pressure of maintaining unsustainable energy production levels. Glycolysis contribution towards cellular energy demands remained the same regardless of ATF5 overexpression or silencing supporting the ability of ATF5 to alter overall cellular expenditure requirements through mitochondrial respiration. Mediating ATF5 protection by improving mitochondrial efficacy provides a possible pathway towards re-establishing homeostasis and recovery of the mitochondria.

v ACKNOWLEDGEMENTS

All the work and dedication committed to earning this doctoral dissertation have culminated into a monumental achievement that I would not have been able to accomplish without the continuous support, guidance and patience from my mentor, Dr. Patricia Rockwell.

Her insightful thoughts and encouragement throughout my Ph.D studies has inspired and motivated me to pursue science from a refreshing perspective.

I am truly grateful to my current and past lab members: Tianfeng Hao, Qin Cao, Junkui

Chen and Rangon Islam. I would also like to thank members from Dr. Maria Figueiredo-

Pereira’s laboratory, Dr. Peter Serrano’s Laboratory, and Dr. Cole Haynes’ laboratory, especially, Chuhyon Corwin, Teneka Jean-Louis, Magdalena Kiprowska, Anna Stoll, Hu Wang,

Jorge Avila, and Christopher Fiorese for all their assistance. Additionally I would like to express my appreciation to Sonia Acevedo and Barbara Wolin from Hunter College Animal facility for all their countless last minute miraculous rescues.

I would like to extend my gratitude to my advisory committee members, Dr. Maria

Figueiredo-Pereira, Dr. Peter Serrano, Dr. Cole Haynes, and Dr. Dan McCloskey for their invaluable time and insightful advice. I am especially grateful to Dr. Maria Figueiredo-Pereira and Dr. Cole Haynes for introducing me to ATF5 without which my thesis would not have taken off.

Lastly, I wish to give a special thanks to my family and my friends at the City University of New York for their boundless support and encouragement throughout my doctoral studies.

vi Table of Contents

List of Figures…………………………………………………………………………………….x

List of Abbreviations…………………………………………………………………………...xii

Chapter I – ATF5 protects primary cortical neurons against PGJ2 induced toxicity…………………………………………………………………………………………….1

1.1 Introduction……………………………………………………………………………2

1.2 Results…………………………………………………………………………………7

1.2.1 Cell cytotoxicity induced by PGJ2……………………………………..…7

1.2.2 ATF5 overexpression protects against PGJ2 toxicity……………………..9

1.2.3 ATF5 traffics to both the mitochondria and nucleus ……………………11

1.2.4 PGJ2 stress induces ATF5 mRNA expression…………………………...13

1.2.5 ATF5 silencing exacerbates PGJ2 toxicity………………………………15

1.2.6 Silencing ATF5 abrogates mitochondrial and nuclear trafficking ……....17

1.2.7 Inhibition of ATF5 by siRNA reduces ATF5 expression levels…………19

1.3 Discussion……………………………………………………………………………20

Chapter II - ERK pathway regulates ATF5 protection against PGJ2 toxicity in primary cortical neurons………………………………………………………………………………....24

2.1 Introduction…………………………………………………………………………..25

2.2 Results………………………………………………………………………………..27

2.2.1 ERK plays a prosurvival role against PGJ2 toxicity……………………..27

2.2.2 ERK regulates apoptosis and oxidative stress…………………………...29

vii 2.2.3 ERK activation necessary for ATF5 overexpression to restore cell viability

against PGJ2 toxicity …………………………………………………....31

2.2.4 ATF5 expression elicited by ERK mediates survival against PGJ2 toxicity

independent of oxidative stress and apoptosis induction………………...33

2.3 Discussion……………………………………………………………………………35

Chapter III – ATF5 mediates cellular respiration to protect against PGJ2 associated mitochondrial dysfunction……………………………………………………………………..37

3.1 Introduction…………………………………………………………………………..38

3.2 Results………………………………………………………………………………..43

3.2.1 ATF5 regulates mitochondrial respiration in PGJ2 stress conditions……43

3.2.2 ATF5 protects against ROS production induced by PGJ2 toxicity…..….48

3.2.3 ATF5 maintains mitochondrial membrane potential against PGJ2

toxicity…………………………………………………………………...50

3.2.4 Glycolysis contribution maintained under PGJ2 stress …………………52

3.3 Discussion……………………………………………………………………………56

Chapter IV – Conclusion……………………………………………………………………….59

Chapter VII – Model.…………………………………………………………………………..61

Chapter V - Methodology………………………………………………………………………63

viii Chapter VI – Future Directions………………………………………………………………..70

Reference List…………………………………………………………………………………...74

ix Table of Figures

Figure 1: Potential mechanism by which J2 prostaglandins promote

Neurodegeneration……………………………………………………………….3

Figure 2: J2 prostaglandins target the ubiquitin proteasome pathway (UPP) and

mitochondria………………………………………………………...... 4

Figure 3: Cell Cytotoxicity induced by PGJ2 in primary cortical neurons……………..8

Figure 4: ATF5 overexpression protects against PGJ2 toxicity………………………...10

Figure 5: ATF5 traffics to both the mitochondria and nucleus………………………...12

Figure 6: PGJ2 stress induces ATF5 mRNA expression………………………...... 14

Figure 7: ATF5 silencing exacerbates PGJ2 toxicity……………………………………16

Figure 8: Silencing ATF5 abrogates mitochondrial and nuclear trafficking in PGJ2

stressed cortical neurons………………………………………...... 18

Figure 9: Inhibition of ATF5 by siRNA reduces ATF5 expression levels……...... 19

Figure 10: ERK plays a prosurvival role against PGJ2 toxicity…………………………28

Figure 11: ERK regulates apoptosis and oxidative stress………………………………...30

Figure 12: ERK activation necessary for ATF5 overexpression to restore cell viability

against PGJ2 toxicity…………………………………………………………...32

Figure 13: ATF5 expression elicited by ERK mediates survival against PGJ2 toxicity

independent of oxidative stress and apoptosis induction…………………….34

Figure 14: Schematic representation of Agilent Seahorse XF Mito Stress assay……….40

Figure 15: Complexes of ETC and target of action of all compounds in the Seahorse XF

Cell Mito Stress Test Kit……………………………………………………….40

x Figure 16: Schematic representation of the Agilent Seahorse XF Glycolytic Rate

assay……………………………………………………………………………..42

Figure 17: ATF5 regulates mitochondrial respiration in PGJ2 stress conditions…..….46

Figure 18: ATF5 protects against ROS production induced by PGJ2 toxicity…………49

Figure 19: ATF5 maintains mitochondrial membrane potential against PGJ2

toxicity………………………………………………………………………..….51

Figure 20: Glycolysis contribution maintained under PGJ2 stress……………...... 54

Figure 21: Model of ATF5 neuroprotection against PGJ2 induced toxicity…………….62

xi List of Abbreviations

2-DG: 2-deoxy-D-glucose FCCP: Carbonyl cyanide-4

ATF5: Activating Transcription Factor 5 (trifluoromethoxy) phenylhydrazone

ATFS-1: Activating Transcription GRA: Agilent Seahorse Glycolytic Rate

Factor associated with Stress-1 Assay

ATP: adenosine triphosphate HO-1: Heme oxygenase-1

ΔΨ : mitochondrial membrane HSP70: Heat shock protein 70 m

potential IMM: inner mitochondrial membrane

CICD: Caspase-independent cell death ISR: integrated stress response

CMS: Agilent Seahorse Cell Mito Stress MAPK: mitogen-activated protein

Test kinase

COX-2: Cyclooxygenase-2 MOMP: mitochondrial outer membrane

DMEM: Dulbecco’s Modified Eagle’s permeabilization

Minimal Essential Medium MTS: Mitochondrial targeting sequence

DMSO: Dimethyl Sulfoxide NLS: nuclear localization sequence

E18 : Embryonic day 18 NMD: nonsense mediated decay

ECAR: extracellular acidification rate OCR: Oxygen consumption rate OMM: outer mitochondrial membrane eIF2a: eukaryotic initiation factor-2 OXPHOS: Oxidative phosphorylation ERK1/2: Extracellular signal-regulated PER: Proton Efflux Rate protein kinases 1 and 2 PGJ2: Prostaglandin J2 ETC: mitochondrial electron transport

chain

xii RT qPCR: quantitative reverse transcription PCR

ROS: Reactive oxygen species

SEM: Standard error of the mean

U0126: MEK1/2 inhibitor uORF: upstream open reading frame

UPP: ubiquitin-proteasome pathway

UPRmt: Mitochondrial unfolded protein response

xiii

CHAPTER I

INTRODUCTION

ATF5 protects primary cortical neurons against PGJ2 induced toxicity

1 1.1. Introduction

Mitochondrial dysfunction has been recognized to play an important role in neurodegenerative disorders, including Parkinson’s and Alzheimer’s disease. The human brain accounts for approximately 20% of the body’s energy demand, more energy than any other organ

(Rolfe 1997). Essential for cellular function, mitochondria are tiny powerhouses with roles in

ATP production, calcium homeostasis, and apoptotic signaling (Celardo 2014). As high energy- using cells, neurons rely on the mitochondria for their supply of energy (Weissman 2007).

Therefore, understanding mitochondrial dysfunction in stressed neurons and strategies for recovery can provide novel insights towards therapeutic treatments for neurodegenerative disorders.

Chronic neuroinflammation is implicated in the pathogenesis of neurodegenerative disorders including Parkinson (PD), Alzheimer (AD), and Huntington diseases. The cascade of events leading to neuronal degeneration may possibly be triggered by a self-perpetuating cycle of inflammation processes that is activated upon injury to the CNS (Hirsch 2009). High levels of prostaglandins are released by the microglia and astrocytes that are activated by inflammatory mechanisms. Cyclooxygenases (COX-1 and COX-2) are key enzymes that contribute to the biosynthesis of prostaglandins in inflammation (Figueiredo-Pereira 2015). Evoked by numerous extrinsic and intrinsic stimuli, the cyclooxygenase pathway converts arachidonic acid to bioactive prostanoids. Transitioning from acute to chronic inflammation is a poorly understood mechanism. Animal studies have provided evidence that prostaglandins not only mediate acute inflammation but also function in the transition to and maintenance of chronic inflammation

(Pierre 2009; Shiver 2014). Derived from PDG2, the highly toxic endogenous J2 prostanglandin

2 levels are significantly increased when traumatic brain injuries are encountered. PGJ2 released from microglia and astrocytes perpetuate its synthesis via an upregulation of COX-2 to activate a

positive feed-back loop of neuroinflammation (Figure 1). The

ubiquitin-proteasome pathway (UPP) and mitochondrial

function are targets of J2 prostaglandins and impairment of

these mechanisms play key roles in the neurodegenerative

process. It has been demonstrated that J2 prostaglandins

inhibit the proteasome by (1) impairing the 26S proteasome

(2) inhibiting de-ubiquitinating enzymes (DUBs) and (3)

covalently modifying specific active site cysteines on UPP

components. Moreover, the inhibition of mitochondrial

function by J2 prostaglandins leads to oxidative stress and FIGURE 1 | Potential mechanism by which J2 prostaglandins promote apoptosis through (1) the inhibition of complex I (2) a neurodegenration. During neuroinflammation PGJ2 are decrease in the mitochondrial membrane potential (MMP) (3) released from activated microglia and astrocytes. PGJ2 increase the levels of ROS overproduction and (4) a blockade of mitochondrial COX-2, thus activating a positive feedback loop that division (Figure 2). Detecting stress-induced heme could mediate the transition from acute to chronic oxygenase-1 (HO-1) enzyme indicates the presence of inflammation. (Figueiredo- Pereira et al 2015) oxidative stress (Keyse & Tyrrell 1989). In the event of proapoptotic stimuli leading to apoptosis, activation of executioner Caspase-3 by cleavage follows after cytochrome c releases from the mitochondrial intermembrane into the cytosol to combine with Apaf-1 and Caspase-9 (Li 1997). Impairment of both the mitochondria and UPP are hallmarks of neurodegenerative disorders. As a model, utilizing PGJ2 to induce neuroinflammation mimics various key features, such as mitochondrial and UPP dysfunction

3 implicated with neurodegenerative diseases.

Mitochondria are critical regulators of cell survival and death, such that their dysfunction

has been suggested to play a central role

in the pathogenesis of neurodegeneration.

Maintaining mitochondrial protein

homeostasis, consisting of proper folding

and assembly of newly translated

polypeptides along with efficient

trafficking and turnover of those proteins

that fail to fold correctly, allows for

proper function and biogenesis (Haynes

2013; Lionaki 2015). Thirteen of the

components of the electron transport

chain (ETC) and the tRNAs and rRNAs

required for their synthesis are encoded

by mitochondrial DNA. Therefore, a

majority of the 1100 proteins that FIGURE 2 | J2prostaglandins target the ubiquitin proteasome pathway (UPP) and mitochondria. J2 comprises the mitochondrial proteome is prostaglandins affect the UPP by: (1) impairing the 26S proteasome by inducing oxidation of proteasome subunits, or promoting its disassembly,(2) inhibiting encoded by nuclear genes where they are de-ubiquitinating enzymes (DUBs),and (3) covalently modifying specific active site cysteines on UPP translated in the cytosol and then components such as E1 activating enzymes, E2 conjugating enzymes, and some E3ligases. J2 prostaglandins can also inhibit mitochondrial function imported into each organelle. Nucleus- by: (1) inhibiting NADH-ubiquinone reductase in complex I, (2) reducing membrane potential, (3 ) encoded proteins synthesized in the blocking fission,and (4) inducing the generation of reactive oxygen species (ROS) and apoptosis. cytosol that are targeted to the (Figueiredo-Pereira et al 2015)

4 mitochondria typically have an N-terminal mitochondrial targeting sequence (MTS) (Jovaisaite

2014; Haynes 2010). An endogenous stress pathway known as the UPRmt, that promotes survival during mitochondrial dysfunction, has been identified in C. elegans (Nargund 2012). Under conditions of mitochondrial stress, cell monitoring or protein import efficiency by the mitochondria signals accumulation. This event leads to induction of UPRmt and promotes mitochondrial protective genes to re-establish protein homeostasis. Evidence indicates the requirement for Activating Transcription Factor associated with Stress-1 (ATFS-1) in UPRmt induction. ATFS-1 is believed to function as a sensor of mitochondrial import efficiency by possessing the ability to localize to both the mitochondria and nucleus due to an N-terminal MTS region and a nuclear localization sequence (NLS) in the leucine-zipper domain, respectively.

When ATFS-1 is directed to the mitochondria in healthy cells it undergoes rapid degeneration.

However, a reduction in mitochondria import efficiency during mitochondrial stress allows

AFTS-1 to accumulate in the cytoplasm. Because ATFS-1 also has a nuclear localization sequence (NLS), it then traffics to the nucleus where it mediates a protective transcriptional program specific for restoring mitochondrial homeostasis. The trafficking of ATFS-1 either to the mitochondria in non-stress cells or to the nucleus to activate the protective UPRmt under mitochondrial stress conditions provides ATFS-1/UPRmt pathway as a mechanism for cells to monitor mitochondrial homeostasis. ATF5, the mammalian ortholog of AFTS-1, has been revealed by studies to be highly expressed as a anti-apoptotic protein in a variety of human cancer cell lines and to function in suppressing the differentiation of neuroprogenitor cells into neurons or glia (Greene 2009). In addition, studies have demonstrated that adult neurons of a mouse model of status epilepticus express ATF5 and that they increase its levels upon endoplasmic reticulum stress as a pro-survival mechanism, suggesting a neuroprotective role for

5 ATF5 (Torres-Peraza 2013; Hetz 2013). Evidence also revealed a decrease and sequestering of the neuroprotective transcription factor ATF5 in Huntington’s disease (Hernandez 2017). The 5’- untranslated region regulates the mRNA stability of stress-response transcription factor ATF5

(Hatano 2013). Phosphorylation of eIF2 has been shown to direct ATF5 translation in environmental stress conditions by a mechanism delaying translation reinitiation in the 5’- untranslated region (Zhou 2008). The binding of HSP70 contributes to the stability of ATF5 by blocking protein degradation to promote survival (Li 2011). However, it is unknown whether this protection is linked to mitochondrial impairment in neurons through UPRmt.

The objective of this study is to evaluate the impact ATF5 overexpression will have on mitochondrial dysfunction induced by chronic neuroinflammation. As highly neurotoxic endogenous products of inflammation, J2 prostaglandins (products of cyclooxygenases) lead to oxidative stress and apoptosis by inhibiting mitochondrial function. Utilizing an established in vitro model of neuroinflammation induced by prostaglandin J2 (PGJ2) to activate the

ATF5/UPRmt pathway we demonstrated ATF5 to be a protective transcription factor against

PGJ2 induced toxicity in rat cortical neurons.

6 1.2. RESULTS

1.2.1. Cell cytotoxicity induced by PGJ2

ATFS-1 activates the protective UPRmt pathway during mitochondrial stress in C. elegans

(Nargund 2012). To study ATF5, the mammalian ortholog of AFTS-1, under conditions of cell cytotoxicity, proteins extracted from PGJ2 treated rat cortical neurons were analyzed by western blotting. Cellular damage was assessed by probing for the presence of HO-1, a biomarker of oxidative stress. Additionally, cleaved Caspase-3 levels were examined to ascertain execution activities that lead to apoptosis. Significant increases of cleaved Caspase-3 and HO-1 levels were observed with PGJ2 treatment suggesting oxidative stress and apoptosis are induced under the stress of PGJ2 (Figure 3A compare lane 1 and 2). Results also revealed an upregulation of ATF5 when cortical neurons were treated with PGJ2 (Figure 3A, compare lane 1 and 2). Performing an

MTS assay to evaluate cell viability showed a cell loss of approximately 20% with PGJ2 treatment (Figure 3E). The loss of cell viability further implicates PGJ2 induces cytotoxicity demonstrated by the increase in both oxidative stress and caspase activation.

Figure 3. Cell cytotoxicity induced by PGJ2 in primary cortical neurons. (A) E18 cortical neurons were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Lysates were analyzed by western blotting probing with antibodies for Caspase-3, HO-1, and ATF5. Actin was probed as a loading control. (B,C,D) Blots were quantified using ImageJ. ****(P<0.0001) indicates a significant difference between PGJ2 vs. control in DMEM for Caspase-3, HO-1, and ATF5. (E) Cell viability was determined using an MTS assay as described in the Materials and Methods section. Results represent the percent cell viability relative to the vehicle control + S.E.M. ****(P<0.0001) indicates a significant difference between PGJ2 vs. control.

8 1.2.2. ATF5 overexpression protects against PGJ2 toxicity

An increase in ATF5 is associated with PGJ2 induced toxicity (Figure 6). To investigate the role of ATF5 under a PGJ2 stressed environment cortical neurons were transfected with an

ATF5 construct to promote overexpression. Western blot analyses of extracted proteins probing for ATF5 indicated an increase in ATF5 protein levels when treated with PGJ2 (Figure 4A, compare lane 2 and 4). Elevated ATF5 levels with PGJ2 treatment alone manifested under induction conditions. Conversely, ATF5 overexpression diminished HO-1 and Caspase-3 activation identified in PGJ2 treated neurons (Figure 4A, compare lane 2 and 4). The lowering of

HO-1 and cleaved Caspase-3 levels suggest a protective role for ATF5. According to MTS assays, overexpressing ATF5 in PGJ2 treated cells improved cell viability by approximately 10% demonstrating its ability to insulate against PGJ2 toxicity (Figure 4E).

9 Figure 4. ATF5 overexpression protects against PGJ2 toxicity. (A) E18 cortical neurons transfected with an ATF5 construct using a magnetofection technique as described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Lysates were analyzed by western blotting probing with antibodies for Caspase-3, HO-1, and ATF5. Actin was probed as a loading control. (B,C,D) Blots were quantified using ImageJ. ****P<0.0001 for comparisons to Control in DMEM for Caspase-3, HO-1, and ATF5. ####P<0.0001 for comparisons to ATF5/PGJ2 in DMEM for Caspase-3, HO-1, and ATF5. ++++P<0.0001 for comparisons to ATF5/Control in DMEM for Caspase-3, HO-1, and ATF5. (E) Cell viability was determined using an MTS assay as described in the Materials and Methods section. Results represent the percent cell viability relative to the vehicle control + S.E.M. ****P<0.0001, **(P<0.01) for comparisons to Control. ####P<0.0001 for comparisons to ATF5/PGJ2. +(P<0.05) for comparisons to ATF5/Control.

10 1.2.3. ATF5 traffics to both the mitochondria and nucleus

Evidence has established the trafficking of ATFS-1 to the nucleus under mitochondrial stress conditions activates the protective UPRmt (Nargund 2012). Identifying the nuclear localization sequence (NLS) in ATF5 along with its ability to regulate mitochondrial chaperone

HSP60 and mtHSP70 suggests ATF5 mimics the ATFS-1 mediated UPRmt in C. elegans (Fiorese

2016). To explore a similar mechanism possibly mediating ATF5 protection in cortical neurons cell fractionations of PGJ2 treated cells were utilized to determine the localization of ATF5.

Analyzing proteins extracted from each fraction by western blotting revealed increased levels of

ATF5 in the nuclear fractions (Figure 5A, compare lanes 1,3,5 to 2,4,6) when treated with PGJ2.

Examining neurons transfected with ATF5 in combination with PGJ2 treatment showed elevated amounts of ATF5 in both the mitochondrial and nuclear fractions (Figure 5A, compare lanes

7,9,11 to 8,10,12). Similar to results observed in Fig 4, a decrease in cleaved Caspase-3 levels in the cytosolic fractions of PGJ2 treated neurons transfected with ATF5 (Figure 5A, compare lanes

2 and 8) suggest the drop in initiating the caspase cascade leads to lower occurrences of apoptosis. Localization of ATF5 in PGJ2 stressed cortical neurons to both the mitochondria and nucleus combined with an increase in cell viability when ATF5 is overexpressed proposes trafficking induced by PGJ2 stress conditions promote ATF5-mediated protection.

B C

Figure 5. ATF5 traffics to both the mitochondria and nucleus. (A) E18 cortical neurons transfected with an ATF5 construct using a magnetofection technique as described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Lysates cell fractionated as described in the Materials and Methods section were analyzed by western blotting probing with antibodies for Caspase-3 and ATF5. Cytochrome C was probed as a fractionation control. Tubulin was probed as a loading control. (B,C) Blots were quantified using ImageJ. ##P<0.01 for comparisons to Mit/control in DMEM for ATF5. +P=0.005, +++P<0.001 for comparisons to Nuc/control in DMEM for ATF5. ****P<0.0001 for comparisons to Cyt/control in DMEM for Caspase3.

12 1.2.4. PGJ2 stress induces ATF5 mRNA expression

Elevated ATF5 protein levels can be attributed to the upregulation of endogenous ATF5 and/or expression of the ATF5 construct. Examining expression levels of ATF5 by quantitative RT PCR will elucidate the source of the increase. After confirming the ATF5 construct successfully increased gene expression (data not shown), focus was directed towards determining whether PGJ2 induces ATF5 at the transcriptional level independent of the ATF5 construct. Analyzing RNA extracted from cortical neurons treated with and without PGJ2

(10μM) by RT qPCR revealed similar steady rates of ATF5 expression levels for both Primer 1

(Figure 6A, compare Control vs PGJ2) and Primer 2 (Figure 6B, compare Control vs PGJ2).

Negligible differences between PGJ2 treated and non-stressed neurons demonstrate a constant transcription of ATF5 mRNA independent of PGJ2 stress. Amino acid limitation has been shown to induce ATF5 mRNA expression in HeLaS3 cells partly as a result of increasing the half-life of

ATF5 mRNA transcript (Watatani 2007). Identifying uORFs in the 5’-UTRs of mRNA splice variants ATF5a and ATF5b revealed 5’-UTRa regulates mRNA stability via NMD in response to stress (Hatano 2013). Additionally, studies have illustrated ATF5 degradation by the ubiquitin-proteasome pathway (UPP) is prevented by cadmium treatment and DNA-damaging agent cisplatin (Wei 2008; Uekusa 2009). Observing an increase in ATF5 protein levels under

PGJ2 treatment conditions suggest PGJ2 stress induces the stabilization of ATF5 mRNA and together with its inhibition of the proteasome (Wang 2006) obstructs the degradation of ATF5, thus enabling gene expression.

ATF5 Primer 1

1.5

1.0

0.5

0.0

Relative mRNA expression PGJ2 - +

B ATF5 Primer 2

2.0

1.5

1.0 ** 0.5

0.0 Relative mRNA expression PGJ2 - +

Figure 6. PGJ2 stress induces ATF5 mRNA expression. (A) Quantitative RT PCR was performed on RNA extracted from E18 cortical neurons transfected with an ATF5 construct using a magnetofection technique as described in the Materials and Methods section after incubation of 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Results generated from Primer 1 and primer 2 were quantified using ImageJ.

14 1.2.5. ATF5 silencing exacerbates PGJ2 toxicity

Investigating the effects of silencing ATF5 in PGJ2 stressed cortical neurons by siRNA provides additional evidence to support the protective role of ATF5 (Figure 4). Western blot analyses of extracted proteins probing for ATF5 indicated the increase in ATF5 under PGJ2 stress conditions (Figure 7A, compare lane 1 and 3) diminishes when ATF5 is silenced (Figure

7A, compare lane 3 and 4). Additionally, a higher oxidative stress presence, demonstrated by a rise in HO-1, was observed in ATF5 suppressed neurons treated within PGJ2 (Figure 7A, compare lane 3 and 4). However, cleaved Caspase-3 levels did not exhibit a significant change when ATF5 is silenced under PGJ2 stressed conditions (Figure 7A, compare lane 3 and 4).

Findings from MTS assays showing an additional drop of approximately10-20% in cell viability

(Figure 7E) coincide with the increase of oxidative stress and a decrease of ATF5 levels. To account for the increased loss in cell viability alternative forms of cell death pathways besides caspase-dependent apoptosis should be considered, such as caspase-independent cell death involving autophagy, necrosis, or MOMP-inducing stimuli (Tait and Green 2008). Together these results suggest silencing ATF5 attenuates protection against PGJ2 toxicity.

15 A B

D E

Figure 7. ATF5 silencing exacerbates PGJ2 toxicity. (A) ATF5 silenced E18 cortical neurons using an siRNA technique described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Lysates were analyzed by western blotting probing with antibodies for Caspase-3, HO-1, and ATF5. Actin was probed as a loading control. (B,C,D) Blots were quantified using ImageJ. ****P<0.0001, ***P<0.001 for comparisons to Control in DMEM for HO-1, ATF5, and Caspase-3. ***P<0.001 for comparisons to Control siRNA in DMEM for Caspase3. ####P<0.0001 for comparisons to Control siRNA in DMEM for HO-1 and ATF5. ++++(P<0.0001), +++P<0.001, +P<0.05 for comparisons to PGJ2 siRNA in DMEM for ATF5, HO-1, and Caspase-3. (E) Cell viability was determined using an MTS assay as described in the Materials and Methods section. Results represent the percent cell viability relative to the vehicle control + S.E.M. ****P<0.0001, ***P<0.001 for comparisons to Control. #P<0.05 for comparisons to PGJ2.

16 1.2.6. Silencing ATF5 abrogates mitochondrial and nuclear trafficking

To further substantiate the proposal that PGJ2 stress stabilization of ATF5 mRNA promotes trafficking cell fractionations of ATF5 silenced cortical neurons treated with and without PGJ2 were examined to identify the localization of ATF5. Probing each fraction for

ATF5 by western blotting revealed augmented levels of ATF5 protein in the nuclear fractions treated with PGJ2 (Figure 8A, compare lanes 1,3,5 to 2,4,6) but the silencing of ATF5 eliminated those increases (Figure 8A, compare lanes 4,6 to 10,12). In contrast, cleaved Caspase-3 levels in cytosolic fractions remained high and are visibly present in the mitochondria of ATF5 silenced neurons that received PGJ2 treatment (Figure 8A, compare lanes 2,4 to 8,10). Progression towards apoptosis by the activation of the executioner caspase reflects the decrease observed in cell viability when ATF5 is silenced in PGJ2 stressed conditions (Figure 7). The loss of ATF5 protein levels, in both the nuclear and mitochondrial fractions, resulting from siRNA targeted degradation inhibits ATF5 transcript stability preventing protection against PGJ2 toxicity.

B C

Figure 8. Silencing ATF5 abrogates mitochondrial and nuclear trafficking in PGJ2 stressed cortical neurons. (A) ATF5 silenced E18 cortical neurons using an siRNA technique described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Lysates cell fractionated as described in the Materials and Methods section were analyzed by western blotting probing with antibodies for Caspase-3 and ATF5. Cytochrome C was probed as a fractionation control. Tubulin was probed as a loading control. (B,C) Blots were quantified using ImageJ. ****P<0.0001, *P<0.05 for comparisons to Cyt/control in DMEM for ATF5 and Caspase- 3. ####P<0.0001, #P<0.05 for comparisons to Mit/control in DMEM for ATF5 and Caspase-3. ++++P<0.0001 for comparisons to Nuc/control in DMEM for ATF5.

18 1.2.7. Inhibition of ATF5 by siRNA reduces ATF5 expression levels

Investigating expression levels under ATF5 repression allows exploration of siRNA effectiveness for ATF5. RNA extracted from cortical neurons with ATF5 silencing treated with and without PGJ2 (10μM) analyzed by RT qPCR showed a significant decrease in ATF5 expression levels compared to those without siRNA for both Primer 1 (Figure 9A, compare

Control vs Control siATF5 and PGJ2 vs PGJ2 siATF5) and Primer 2 (Figure 9B, compare

Control vs Control siATF5 and PGJ2 vs PGJ2 siATF5). Demonstrating proficient suppression of

ATF5 expression by siRNA establishes that siRNA diminished ATF5 protection against PGJ2 stress, reflected in the exacerbation of PGJ2 toxicity (Figure 7) and abrogation of ATF5 trafficking between the mitochondria and nucleus (Figure 8).

A BB ATF5 Primer 2 1.5

1.0 + + *

0.5

0.0 Relative mRNA expression siATF5siATF5 - + - + PGJ2PGJ2 - - + +

Figure 9. Inhibition of ATF5 by siRNA reduces ATF5 expression levels. (A) Quantitative RT PCR was performed on RNA extracted from ATF5 silenced E18 cortical neurons using an siRNA technique described in the Materials and Methods section after incubation of 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Results generated from Primer 1 and primer 2 were quantified using ImageJ. ****P<0.0001, **P<0.01, *P<0.05 for comparisons to Control in DMEM for ATF5 Primer 1 and Primer 2. ++++P<0.0001, ++P<0.01, +P<0.05 for comparisons to PGJ2 in DMEM for ATF5 Primer 1 and Primer 2.

19 1.3. Discussion

Accounting for approximately 20% of the body’s energy demand (Rolfe 1997), the human brain relies on the mitochondria as a major source of energy (Weissman 2007).

Recognizing the central role mitochondrial dysfunction plays in the pathogenesis of neurodegeneration necessitates insight into understanding recovery from mitochondrial impairment. As a highly neurotoxic endogenous product of inflammation, J2 prostaglandins lead to oxidative stress and apoptosis by inhibiting mitochondrial function (Figueiredo-Pereira 2015).

Utilizing an established in vitro model of neuroinflammation induced by PGJ2 permitted investigating ATF5, the mammalian ortholog of UPRmt activator ATFS-1, under conditions of mitochondrial dysfunction. Results from our study affirm that ATF5 protects against PGJ2 induced toxicity in rat E18 primary cortical neurons. The upregulation of ATF5 protein expression ameliorates HO-1 induction, cleaved Caspase-3 activation, and cell loss invoked by

PGJ2 stress (Figure 4). Additionally, exacerbating cell loss viability with the silencing of ATF5 demonstrates the protection endogenous ATF5 provides against PGJ2 toxicity (Figure 7), whereby upregulation enhances this protection. Improvements to the 10% cell loss recovery with

ATF5 upregulation can further illustrate the potency of ATF5 protection. This may be achievable by increasing PGJ2 stress with an increase in treatment concentration but for a shorter period of time or decreasing PGJ2 toxicity with a decrease in treatment concentration but increase exposure time to longer than 16 hrs.

The protective ATFS-1/UPRmt pathway is identified in C. elegans as a mechanism for cells to monitor mitochondrial homeostasis (Nargund 2012). Predicted to contain both an N- terminal mitochondrial targeting sequence (MTS) and a nuclear localization signal (Claros

20 1996), ATFS-1 possesses the ability to migrate to either the mitochondria or the nucleus. When directed to the mitochondria in healthy cells it undergoes rapid degeneration, while trafficking to the nucleus during mitochondrial stress activates the protective UPRmt (Nargund 2012). The association between the mitochondria and nucleus in UPRmt signaling is implicated for ATF5 in fractionation experiments showing ATF5 presence in the mitochondria whereas accumulation in the nucleus occurs only under PGJ2 stress conditions (Figure 5). This finding is further demonstrated by ATF5 overexpression where protein levels are increased in both the nucleus and mitochondria, suggesting UPRmt involvement. Apoptosis associated with mitochondrial dysfunction via the intrinsic pathway involves the release of cytochrome c into the cytosol (Tait

2008). The lack of cytochrome c detection in the cytosol fractions (Figure 5) may simply be an antibody sensitivity issue. This can be resolved by using an alternative cytochrome c antibody with greater sensitivity in western blot analyses. Stress evoked induction of mitochondrial protective transcripts in an ATF5-dependent manner identifies ATF5 regulation of a UPRmt in mammalian cells, analogous to the response regulated by ATFS-1 in C.elegans (Fiorese 2016).

Affiliating ATF5 with a UPRmt in cortical neurons presents the possibility that molecular chaperones are also activated to restore mitochondrial homeostasis.

ATF5 is a stress-response transcription factor regulated by the 5’-untranslated region stabilizing its mRNA (Hatano 2013). This ability to stabilize mRNA along with PGJ2 induced proteasome inhibition (Wang 2006) interfering with ATF5 degradation provides an explanation for the detection of ATF5 protein expression uniquely under PGJ2 stress conditions (Figure 3) even though transcript levels remained constant in the absence and presence of PGJ2 treatment

(Figure 6). Furthermore, silencing ATF5 reduced mRNA in scenarios with and without PGJ2 treatment (Figure 9), curtailing ATF5 protection against PGJ2 toxicity (Figure 7 and 8).

21 Collectively, these findings suggest that PGJ2 stress stabilizes ATF5 mRNA for translation into protein and prevents proteasome degradation of ATF5 allowing it to protect against PGJ2 toxicity.

Translocations of active Caspase-3 from the cytosol to the mitochondria (Chandra 2003) evoked by Status Epilepticus in neuronal death suggest the maneuver facilitates apoptosis (Kim

2018). A trafficking pattern of enhanced Caspase-3 activation in both the cytosol and mitochondria under ATF5 silencing and PGJ2 treatment (Figure 8) demonstrate a similar rationale applies to our scenario. The accompanying reduction in cell viability (Figure 7E) fortifies the reasoning that active Caspase-3 in the mitochondria contributes to the apoptotic process by possibly participating in the degrading of mitochondria proteins, activation of resident procaspases, and dismantling of the mitochondria.

An in vitro model of mammalian cell death studying trophic factor-deprivation-induced death of rat sympathetic neurons implicates the mitochondrion as an important regulator of caspase-dependent and caspase-independent cell death (CICD) (Chang 2002). Findings speculate that cytochrome c release together with activation of caspases govern cell death (Putcha 1999;

Deshmukh 1996). In the absence of caspase, extending the commitment to neuronal death beyond cytochrome c release ends at the point of mitochondrial depolarization (Deshmukh

2000), while inhibition with permeability transition pore (PTP) inhibitor Cyclosporin A in sympathetic neurons reveals PTP opening is a critical event in CICD (Chang 2002). Therefore, it has been proposed that the consequences from damaging the mitochondria entail killing cells in two ways: one is quick and dependent on caspase activation whereas the other is slower, repercussions of a major dysregulation of mitochondrial functions (Ricci 2003). The additional loss in cell viability observed in ATF5 silenced neurons receiving PGJ2 treatment, where

22 Caspase-3 cleavage levels were not exacerbated but an escalation in HO-1 indicated oxidative stress (Figure 7), can potentially be explained by CICD. Mitochondria-derived ROS, where overproduction results in oxidative stress, have been suggested to orchestrate caspase-dependent and -independent neuronal deaths in sympathetic neurons (McManus 2014). CICD was accompanied by the loss of mitochondrial membrane potential and the generation of reactive oxygen species in mouse embryonic fibroblasts (MEFs) (May 2007). Further support was exhibited by the involvement of ROS in death ligand FasL-induced CICD and inflammatory responses in MEFs (Chen 2009). Human embryonic kidney epithelial cells (293T cells) also demonstrate caspase-independent cell death results from the progressive loss of mitochondria function, rather than the release of cytotoxic protein (Lartigue 2009). ROS production was revealed to be involved in inducing autophagy, a contributor to caspase-independent macrophage cell death (Xu-2006). Silencing ATF5 impedes protection against PGJ2 toxicity to create an environment of chronic oxidative stress (Figure 7D), allowing CICD to possibly develop.

Overall, our study demonstrates the neuroprotective role ATF5 plays against PGJ2 induced toxicity in cortical neurons is mediated by transcript stabilization. Insight provided from our findings about mitochondrial impairment and strategies for recovery by overexpressing

ATF5 may lead to the development of therapeutic treatments for neurodegenerative disorders.

CHAPTER II

ERK pathway regulates ATF5 protection against PGJ2 toxicity in

primary cortical neurons

24 2.1. Introduction

The extracellular signal-regulated kinase (ERK) signaling cascade is involved in a variety of cellular functions including migration, differentiation, proliferation and death. As a member of the mitogen-activated protein kinase (MAPK) family, ERK is expressed as a protein kinase intracellular signaling molecule that participates in a communication cascade from the cell membrane to the nucleus when activated by phosphorylation. ERK illustrates its role in cell survival and death by regulating neuronal apoptosis (Li 2014). Maintaining survival proteins at high levels via the ERK-dependent pathway has been demonstrated in chick retinal neurons

(Desire 2000). The ERK survival pathway is shown to regulate apoptosis in hippocampal neurons (Perkins 2003).

A chain reaction of caspase activation is associated with the two principal pathways that lead to apoptosis (Cory 2002). The cell-extrinsic death pathway involves death receptors while the Bcl-2 protein family regulates the intracellular stress induced cell-intrinsic death pathway

(Cory 2002). Consisting of anti-apoptotic proteins and two pro-apoptotic protein groups, the Bax and BH3-only families, interactions between the Bcl-2 family proteins control permeabilization of the mitochondrial outer membrane (MOM) (Puthalakath 2002). MOMP (mitochondrial outer membrane permeabilization) engenders the release of intermembrane space proteins (i.e.

Cytochrome c), located between the outer (OMM) and inner (IMM) mitochondrial membranes, into the cytosol triggering cell death (Chipuk 2006). Functionally, anti-apoptotic proteins including Bcl-2 and Bcl-xl inhibit apoptosis, cellular stress sensors BH3 proteins activate

MOMP, and executioner proteins Bax or Bak permeabilize MOM by oligomerizing (Shamas-Din

2013). Regulation of the cell-intrinsic death pathway has been linked to the ERK1/2 signaling

25 pathway. Expression of BH3-only protein, Bim is repressed by phosphorylation and proteasome- dependent degradation mediated under ERK1/2 pathway activation (Weston 2003, Ley 2003).

In addition to mediating cell fate, findings implicate ERK1/2 pathway activation of playing a role in neuroprotection. Phospho-activation of MEK and ERK inactivates pro-apoptotic

Bcl-2-associated death protein, BAD, to protect hypoxic neuronal injury in a mouse model (Jin

2002). Protecting rat cortical neurons from oxidant injury is mediated by ERK activation and

Bcl-2 induction (Sanchez 2012). MEK/ERK pathway is involved in vascular endothelial growth factor (VEGF) protection of cortical neurons against mechanical trauma injury induced apoptosis

(Ma 2011).

Stress-inducible HSP70 is a molecular chaperone that assists in cellular protein folding processes (Mayer 2005). Associating protective ERK pathway activation with HSP70 protein folding, eIF2a protein synthesis, and protein modification activities suggest a possible pathway for ATF5 protection. Activation of the MAPK/ERK pathway mediates HSP70 protection against oxygen-glucose deprivation injury (Yu 2015). ERK signaling increases HSP70 expression to protect against oxidative stress-induced neuronal apoptosis (Qi 2016). Amino acid stress- activated pathway requires MEK signaling to phosphorylate eIF2a (Thiaville 2008). Cell proliferation and survival is reliant on p300-dependent ATF5 acetylation to mediate ERK/MAPK activation of transcription factor early growth response factor-1 (Egr-1) gene (Liu 2011).

Examination of the ERK signaling pathway under PGJ2 stress in our studies allowed elucidation of the ATF5 neuroprotective pathway. Results demonstrated ERK activation mediated ATF5 protection in the presence of PGJ2 toxicity.

26 2.2. RESULTS

2.2.1. ERK plays a prosurvival role against PGJ2 toxicity

The dual role of extracellular signal-regulated kinase (ERK) in neurons has been shown to be both anti-apoptotic as well as pro-apoptotic (Li 2014). Exploring ERK under PGJ2 stress will elucidate its participation in the mechanism recruited for ATF5 neuroprotection. Proteins extracted from PGJ2 treated rat cortical neurons, in the absence or presence of MEK 1/2 inhibitor

U0126, were analyzed by western blotting. MEK is the upstream kinase that activates

(phosphorylates) ERK1/2. Probing for pERK revealed significant activation of ERK with PGJ2 treatment alone (Figure 10A compare lane 1 and 2; Figure 10C). Introducing U0126 to PGJ2 treated neurons resulted in an immense reduction in ERK activation (Figure 10A compare lane 2 and 4; Figure 10C). MTS assay results exhibited a more noticeable decrease in cell viability for

U0126 treated neurons with or without PGJ2 compared to PGJ2 treatment alone (Figure 10B).

Associating greater ERK activation and lower cell loss with PGJ2 stress conditions proposes an anti-apoptotic role for ERK.

B C

Figure 10. ERK plays a prosurvival role against PGJ2 toxicity. (A) E18 cortical neurons were incubated for 16hrs in DMEM and treated with and without PGJ2 (10μM) in the presence of U0126 (10µM) for 16 h. Lysates were analyzed by western blotting probing with antibodies for pERK and Total ERK. Total ERK was probed as a loading control. (B) Cell viability was determined using an MTS assay as described in the Materials and Methods section. Results represent the percent cell viability relative to the vehicle control + S.E.M. ****P<0.0001 for comparisons to Control. ####P<0.0001, ###P<0.001 for comparisons to PGJ2. (C) Blots were quantified using ImageJ. ****P<0.0001 for comparisons to Control in DMEM for pERK. ####P<0.0001 for comparisons to PGJ2 in DMEM for pERK. ++++P<0.0001 for comparisons to U0126 in DMEM for pERK.

28 2.2.2. ERK regulates apoptosis and oxidative stress

To further study the impact ERK protection has against PGJ2 toxicity proteins extracted from PGJ2 treated rat cortical neurons in the absence or presence of U0126 were analyzed by western blotting. Significant increases of cleaved Caspase-3 and HO-1 activation were detected in neurons treated with PGJ2 alone or in combination with U0126 (Figure 11A compare lane

1,2,3 and 4). Results also identified an upregulation of ATF5 protein levels when cortical neurons were treated with PGJ2 or in combination with U0126 (Figure 11A compare lane 1,2,3 and 4). Lower levels of ATF5, HO-1 and Caspase-3 are observed when ERK activation is inhibited in PGJ2 treated cells (Figure 11A compare lane 2 and 4). These results suggest that

ERK regulates in part apoptosis and oxidative stress induced by PGJ2. The effects of ERK inhibition are twofold. ERK regulates cell viability, Caspase-3 activation and HO-1 induction in

PGJ2 treated cells. Whereas ERK inhibition exacerbates the loss in cell viability induced by

PGJ2 (Figure 10B), a blockade of ERK activation decreases Caspase-3 activation and HO-1 induction in PGJ2 treated cells. These findings suggest that reducing apoptosis or oxidative stress cannot relieve the loss in cell viability induced by PGJ2 when ERK activation is suppressed.

Therefore, ERK is essential for maintaining cell viability in PGJ2 treated cells via mechanisms independent of Caspase-3 activation and HO-1 induction.

Figure 11. ERK regulates apoptosis and oxidative stress. (A) E18 cortical neurons were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) in the presence of U0126 (10µM) for 16 h. Lysates were analyzed by western blotting probing with antibodies for Caspase-3, HO-1, and ATF5. Actin was probed as a loading control. (B,C,D) Blots were quantified using ImageJ. ****P<0.0001, *P<0.05 for comparisons to PGJ2 in DMEM for ATF5. ###P<0.001, ##P<0.01 for comparisons to U0126 + PGJ2 in DMEM for ATF5. ****P<0.0001, ***P<0.001, **P<0.01 for comparisons to Control in DMEM for Caspase-3 and HO-1. ####P<0.0001, ###P<0.001 for comparisons to PGJ2 in DMEM for Caspase-3 and HO-1. ++(P=0.0014) indicates a significant difference between U0126 vs. U0126 + PGJ2 in DMEM for HO-1. +P<0.05 for comparisons to U0126 in DMEM for HO-1.

30 2.2.3. ERK activation necessary for ATF5 overexpression to restore cell viability against

PGJ2 toxicity

Overexpressing ATF5 in the presence of ERK inhibition will allow additional understanding about the protective pathway instituted by ERK under PGJ2 stressed conditions.

Proteins extracted from PGJ2 treated rat cortical neurons overexpressing ATF5 in the absence or presence of U0126 were analyzed by western blotting. Probing for pERK revealed noticeable

ERK stimulation with PGJ2 treatment alone compared to Control (Figure 12A compare lane 1 and 2; Figure 12B). However, inhibiting with U0126 significantly decreased the observed PGJ2 induced ERK activation (Figure 12A compare lane 2 and 4; Figure 12B). Enhancing ATF5 culminated in a similar pattern of ERK activation (Figure 12A compare lane 5,6,7 and 8; Figure

12B), although at notably diminished levels. MTS assay results (Figure 12C) illustrate a significant increase in cell viability for either U0126 or PGJ2 treated neurons when ATF5 is enriched (Figure 12C). In these experiments, overexpression of ATF5 rescues neurons from the loss in cell viability induced by PGJ2, or U0126 alone but not PGJ2/U0126. These findings reinforce the notion that ERK is required to maintain viability in PGJ2 treated cells.

B C

Figure 12. ERK activation necessary for ATF5 overexpression to restore cell viability against PGJ2 toxicity. (A) E18 cortical neurons transfected with an ATF5 construct using a magnetofection technique as described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) in the presence of U0126 (10µM) for 16 h. Lysates were analyzed by western blotting probing with antibodies for pERK and Total ERK. Total ERK was probed as a loading control. (B) Blots were quantified using ImageJ. ****P<0.0001 for comparisons to PGJ2 in DMEM for pERK. #P<0.05 for comparisons to U0126 + PGJ2 in DMEM for pERK. (C) Cell viability was determined using an MTS assay as described in the Materials and Methods section. Results represent the percent cell viability relative to the vehicle control + S.E.M. ***P<0.001 for comparisons to PGJ2. ####P<0.0001 for comparisons to U0126.

32 2.2.4. ATF5 expression elicited by ERK mediates survival against PGJ2 toxicity independent of oxidative stress and apoptosis induction

Examining ATF5 under ERK inhibition will provide insight into the interplay between these two components of a prospective neuroprotective regulatory mechanism against PGJ2 toxicity. Western blots consisting of proteins extracted from PGJ2 treated rat cortical neurons overexpressing ATF5 in the absence or presence of U0126 were probed for Caspase-3, HO-1, and ATF5. Although an escalation in ATF5 is observed in conjunction with a decrease in cleaved

Caspase-3 and HO-1 activation in PGJ2 stressed neurons overexpressing ATF5 (Figure 13A compare lane 2 and 6), diminished ATF5 levels along with a reduction in Caspase-3 activation and HO-1 induction associated with ERK inhibition in PGJ2 treated cells remain unaltered under

ATF5 enhancement (Figure 13A compare lane 4 and 8. In these experiments, overexpression of

ATF5 does not significantly relieve the loss in cell viability in PGJ2/U0126 treated cells (Figure

12C). Results suggest the protective role of ATF5 is regulated by the presence of ERK managing

ATF5 protein levels. Besides demonstrating the capability to bring about oxidative stress and apoptosis, ERK also exhibits a prosurvival role independent of these inductions.

Figure 13. ATF5 expression elicited by ERK mediates survival against PGJ2 toxicity independent of oxidative stress and apoptosis induction. (A) E18 cortical neurons were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) in the presence of U0126 (10µM) for 16 h. Lysates were analyzed by western blotting probing with antibodies for Caspase-3, HO-1, and ATF5. Actin was probed as a loading control. (B,C,D) Blots were quantified using ImageJ. ****P<0.0001 for comparisons to PGJ2 in DMEM for Caspase-3, HO-1, and ATF5. ##P<0.05 for comparisons to Control in DMEM for Caspase-3.

34 2.3. Discussion

The extracellular signal-regulated kinase (ERK) signaling cascade plays a role in the regulation of neuronal apoptosis (Li 2014). Under amino acid limitation in Hepatoma cells MEK signaling phosphorylates eIF2a (Thiaville 2008) to activate the integrated stress-response (ISR), a diverse cellular stress response pathway (Pakos-Zebrucka 2016). ERK signaling has been shown to increase HSP70 expression to protect against oxidative stress-induced neuronal apoptosis (Qi 2016). Additionally, it has been demonstrated in Zebrafish that HSP70 expression is dependent on ERK activation (Keller 2008). HSP70 binding contributes to the stability of

ATF5 by blocking protein degradation to promote survival (Li 2011). Reflective of literature suggesting a protective role for ERK, our studies examining the ERK signaling pathway under

PGJ2 stress elucidated a possible regulatory mechanism mediating ATF5 neuroprotection. Both

ATF5 protein levels and ERK phosphorylation increased under PGJ2 stress conditions (Figure

12B and 13C). ATF5 overexpression rescues PGJ2 or U0126 treated cells from a loss in cell viability (Figure 12C). However, inhibition of ERK with U0126 in PGJ2 treated cells eliminates the recovery in cell viability elicited by the enhancement of ATF5 (Figure 12C). Results demonstrate ERK is essential for maintaining cell viability in PGJ2 treated cells. ERK signaling cascades have been implicated in maintaining cell viability in glioblastoma via Nrf2, a transcription factor NF-E2-related factor 2 that controls adaptive defense responses (Cong 2013), and in mouse progenitor B cell line BaF3 following heat shock insult (Ng 2000).

Protection exhibited by the overexpression of ATF5 in earlier findings were incapable of mitigating the reduction in cell viability from PGJ2 stress when MEK inhibitor U0126 attenuated

ERK activation (Figure 12C). Previous studies associate the MAPK/ERK pathway with

35 mediating HSP70, a stabilizer of ATF5 in glioma cells (Li 2011). In conditions of ERK inhibition a reduction in HSP70 expression would expose ATF5 to degradation. A considerable decline in ATF5 protein expression in cells experiencing PGJ2 stress and ERK inhibition (Figure

13C), even though it was being upregulated, endorses this possibility of ERK having a hand in regulating ATF5.

A hypothesis proposes chronic neuroinflammation involvement in the pathogenesis of neurodegeneration results from the up-regulation of cyclooxygenase COX-2 by prostaglandins

(Figueiredo-Pereira 2015). Prostaglandins of the J2 series have been shown to induce intracellular oxidative stress (Kondo 2001). Moreover, a metabolite of PGJ2, 15d-PGJ2 induces mitochondrial-dependent apoptosis in renal proximal epithelial cells (Lee 2009). 15d-PGJ2 utilizes MAPK activation involving ROS to induce COX-2 expression in human osteosarcoma cells (Kitz 2011). Supportive of these existing observations, our results also indicate ERK activation induces apoptosis and oxidative stress under PGJ2 toxicity (Figure 13B and D).

Independent of eliciting oxidative stress and apoptosis, also exhibiting prosurvival capabilities illustrate the dual roles played by ERK (Figure 12).

Collectively, results from our study implicate ERK as a mediator of ATF5 protection against PGJ2 toxicity. The phosphorylation of ERK coinciding with enhanced ATF5 expression under PGJ2 stress suggests a mechanistic pathway to stabilize ATF5 in order to induce a protective cascade, such as UPRmt.

CHAPTER III

ATF5 mediates cellular respiration to protect against PGJ2 associated

mitochondrial dysfunction

37 3.1. Introduction

The high-energy demands of brain metabolism necessitate consumption of nearly 20% of a body’s daily energy resources yet it constitutes only 2% of the body’s mass (Rolfe 1997). As a major supplier of ATP, the mitochondrion plays an important role in cellular bioenergetics supporting cell function and cell survival/death pathways (Nicholls 2000; Bernardi 1999).

Mitochondria are double-membrane bound organelles consisting of an outer mitochondrial membrane (OMM) containing protein-based pores allowing the passage of small molecules, an inner mitochondrial membrane (IMM) composed of proteins involved in electron transport that drives oxidative phosphorylation (OXPHOS) to generate ATP, and an intermembrane space between the OMM and IMM (The Cell: A Molecular Approach 2nd edition). The process of

OXPHOS entails protons traveling from protein complexes I, II, III, and IV within the electron transport chain that ultimately reduces oxygen to form water and synthesize ATP (Papa 2012).

Accumulating evidence suggests mitochondrial dysfunction contributes to the pathogenesis of neurodegenerative diseases (Schapira 2012; Reddy 2009; Beal 2005).

Mitochondrial gene expression time-course of aging mice implicates mitochondrial dysfunction, oxidative damage, and cytochrome c in age-related neurodegenerative diseases (Manczak 2005).

Exacerbation of oxidative stress by mitochondrial dysfunction is proposed to contribute causally to neurodegenerative pathology (Lin 2006). Robust removal of H2O2 in the brain demonstrates

ROS detoxifiying capabilities by the mitochondria (Zoccarato 2004). ROS removal system failure induces a net ROS production by mitochondria resulting in oxidative stress (Andreyez

2005).

38 An essential participant in the pathogenesis of neurodegenerative diseases, ascertaining mitochondrial health in cells allows evaluation of its impairment under stress conditions.

Utilizing the Agilent Seahorse XF Cell Mito Stress Test (CMS) to measure key parameters of mitochondrial function enables an assessment. This assay can directly measure cellular oxygen consumption rate (OCR) by applying sequential compound injections of oligomycin, carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP), and a mix of rotenone and antimycin A.

Calculations utilizing these parameters along with basal respiration will yield data about proton leak and spare respiratory capacity (Figure 14). As an inhibitor of ATP synthase (complex V), the oligomycin injection will naturally result in the decrease of OCR to demonstrate the level of

ATP production that comes from mitochondria. Maximum capability of the mitochondria is examined with the injection of FCCP, an uncoupling agent that collapses the proton gradient and disrupts the mitochondrial membrane potential consequently allowing oxygen consumption by complex IV. The final injection combination of rotenone, a complex I inhibitor, and antimycin

A, a complex III inhibitor disables mitochondrial respiration and facilitates the determination of nonmitochondrial respiration executed by processes besides the mitochondria (Figure 15).

Essentially, serially injected compound oligomycin measures ATP production, FCCP measures maximal respiration, and a mix of rotenone and antimycin A measures non-mitochondrial respiration.

Considering the brain depends on glucose as its primary energy source, it is suitable to focus on glycolysis as an alternative metabolic pathway to compensate for the energy deficiencies associated with mitochondrial dysfunction. Findings suggest neurons have the ability to increase glycolysis in response to stimulation (Yellen 2018).

Mitochondrial dysfunction is associated with an increase of glycolysis in blood cells

39 Figure 14. Schematic representation of the Agilent Seahorse XF Mito Stress assay. The Agilent SeahorseXFp analyzer measures OCR at basal and after injection of oligomycin, FCCP, and antimycin A/rotenone for three measurement cycles at each step. Key parameters of mitochondrial function such as basal respiration, ATP production, proton leak, maximal respiration, spare respiratory capacity, and nonmitochondrial respiration can be interpreted as shown in the figure. Time points of different drug injections are shown with arrows. (Agilent Technologies)

Figure 15. Complexes of ETC and Target of Action of all compounds in the Seahorse XF Cell Mito Stress Test Kit. Oligomycin inhibits ATP synthase (complex V), FCCP uncouples oxygen consumption from ATP production, and rotenone and antimycin A inhibits complexes I and III, respectively. (Agilent Technologies)

40 preceding and early Parkinson’s (Smith 2018). Energy contributions attributed to glycolysis can be determined by utilizing the Agilent Seahorse XF Glycolytic Rate Assay (GRA).

Measurements of glycolytic rates for basal conditions and compensatory glycolysis are ascertained by inhibiting the mitochondria with rotenone/antimycin A. Administering rotenone and antimycin A to inhibit mitochondrial complex I and III, respectively, effectively obstructs

OXPHOS and stimulates the cell to use glycolysis as a compensatory alternative to satisfy the cells’ energy demands. The expulsion of protons, generated during the conversion process of glucose to lactose or CO2 to water in the mitochondria, acidifies the extracellular medium.

Calculating mitochondrial CO2 contribution to extracellular acidification allows for the determination of the Proton Efflux Rate (PER), the number of protons exported by cells into the assay medium over time, strictly attributed to glycolysis. The injection of 2-deoxy-D-glucose (2-

DG) certifies that the PER generated before its injection is primarily from glycolysis. Glucose analog 2-DG inhibits glycolysis through competitive binding to glucose hexokinase, the initial enzyme in the glycolytic pathway (Figure 16).

Employing both Agilent Seahorse Cell Mito Stress Test and Glycolytic Rate Assay in the study of ATF5 protection under PGJ2 stress allowed further understanding of the protection provided against mitochondrial dysfunction. Associating an ATF5 response with a rise in mitochondrial activity and a corresponding reduction in glycolysis suggests protective capability against PGJ2 toxicity.

Figure 16. Schematic representation of the Agilent Seahorse XF Glycolytic Rate assay. The Agilent SeahorseXFp analyzer measures ECAR and OCR at basal and after injection of antimycin A/rotenone and 2-deoxy-D-glucose (glucose analog which inhibits glycolysis) for three measurement cycles at each step. Key parameters of glycolysis such as mitochondrial CO2 contribution to acidification, basal glycolysis, and compensatory glycolysis can be interpreted as shown in the figure. Time points of different drug injections are shown with arrows.

42 3.2. RESULTS

3.2.1. ATF5 regulates mitochondrial respiration in PGJ2 stress conditions

Based on findings presented thus far, it has been demonstrated that ATF5 plays a neuroprotective role against PGJ2 toxicity. Since mitochondrial stress induces ATFS-1 activation of the protective UPRmt pathway in C. elegans, ATF5 was explored as a factor that can protect against mitochondrial dysfunction associated with PGJ2 toxicity. Conducting the Agilent

Seahorse XF Cell Mito Stress (CMS) Test on rat cortical neurons treated with PGJ2 alone,

U0126 alone, or in combination generated basal oxygen consumption rates (OCR) (Figure 17A) for the Control that are consistent with a healthy functioning mitochondria while levels decreased approximately 25% from the Control for PGJ2 stressed cells (Figure 17A). Treatment with U1026 in the presence or absence of PGJ2 diminished OCR levels even more significantly to below 75% of control (Figure 17A). This suggests ERK contributes to mitochondrial function.

The fact that viability is reduced in PGJ2/U0126 treated cells (Figure 10) may reflect this result.

Basal levels are indicative of energy demands of a cell where healthy neurons require enormous amounts of energy. Following the oligomycin injection, which inhibits complex V, a considerable drop in OCR for the Control revealed that the mitochondria contributes significantly to the ATP demands of the cell (Figure 17A). As for U0126 or U0126 + PGJ2 treated neurons, the persistence of low OCR after the oligomycin injection, reflected by low ATP production by the mitochondria (Figure 17A), in conjunction with a modest proton leak implies low mitochondrial respiration. The examining of basal respiration not attributed to ATP production in PGJ2 treated cells, revealed significant proton leak and low ATP contribution from respiration suggesting the presence of mitochondrial damage (Figure 17A). Lastly, analyzing the

43 maximum capability of the mitochondria showed OCR increased back to similar baseline levels for control cells after the FCCP injection demonstrating mitochondria in control cells are capable of responding to greater ATP demands (Figure 17A). Conversely, cortical neurons treated with either PGJ2 or U0126 or in combination did not exhibit this increase of OCR after the FCCP injection proposing a large portion of respiration maximum was utilized leaving minimal reserves available to satisfy higher energy requests. This again suggest that ERK may regulate mitochondrial function as mentioned above.

To understand the impact ATF5 has on mitochondrial function Cell Mito Stress (CMS)

Test assessments, made by measuring key parameters of mitochondrial function, were performed on ATF5 transfected rat cortical neurons treated with PGJ2 alone, U0126 alone, or in combination. Basal OCR for Control cells enhancing ATF5 (Figure 17B), albeit at slightly lower levels than Controls not overexpressing ATF5 (Figure 17A), along with ATP production levels being at least 75% of basal respiration levels demonstrating the mitochondria contributed significantly to the ATP demands of Control cells (Figure 17B), indicate a normal functioning mitochondria (Figure 17B). Dramatic reduction in basal OCR along with low OCR in ATP production and minimal proton leak for ATF5 upregulated cells receiving PGJ2 treatment implies minimal mitochondrial respiration activity (Figure 17B). Cortical neurons treated with

U1026 or a combination of U0126 and PGJ2 displayed an OCR pattern similar to PGJ2 exposed cells (Figure 17B), revealing low mitochondrial respiration under these stress conditions while overepressing ATF5. Additionally, relatively low spare capacity for all conditions indicate the inability of the cell to respond to additional energetic demands (Figure 17B). In the case of control cells transfected with ATF5, high basal OCR and ATP production coordinated with low proton leak and spare capacity suggest healthy mitochondria are working at their maximum

44 capacities. For ATF5 enhanced cells treated with PGJ2, U0126, or in combination the low basal

OCR, ATP production, proton leak, and spare capacity indicate mitochondria of distressed cells working at their apex are only capable of generating low amounts of ATP (Figure 17B). ATF5 appears to have an effect on the mitochondrial respiration of transfected Control and PGJ2 treated cells (Figure 17B) compared to untransfected cells (Figure 17A), proposing a reduction in basal OCR minimizes proton leak and enables mitochondria working at maximum capacity to satisfy ATP needs, correlating with improved cell viability results (Figure 4E).

Further exploration of ATF5 capabilities to mediate mitochondrial function was accomplished by performing a CMS Test on ATF5 silenced rat cortical neurons treated with

PGJ2 alone, U0126 alone, or in combination. Results revealed ATF5 silenced controls (Figure

17C) produced basal OCR, ATP production, proton leak, and spare capacity readings similar to those observed in unaltered Controls (Figure 17A). Basal respiration in ATF5 silenced cells is representative of healthy mitochondria since the inability to induce protection by ATF5 was unnecessary (Figure 17C). Correspondingly, the detection of extremely low OCR in all areas of mitochondrial function for silenced cells receiving treatment exhibit an analysis comparable to treated cells overexpressing ATF5 (Figure 17B) suggesting an analogous scenario where maximizing mitochondria efforts provide meager amounts of ATP to low respiration energy demanding cells (Figure 17C). Silencing ATF5 in the presence of unhealthy mitochondria and mitochondrial inhibition appears to diminish mitochondrial activity to barely existent levels.

Overall, investigations conducted with CMS tests unveiled the ability of ATF5 to modulate mitochondrial functions to alleviate PGJ2 toxicity by maximizing respiration in healthy mitochondria to fulfill energetic demand.

45 A DMSO PGJ2 & U0126 PGJ2 125 U0126 U0126 + PGJ2 100

75 * *** 50 ** *** ++++ **** **** ++ 25 ++++ ++ **** +++ +++ *

OCR (pmol/min) 0 Basal ATP Proton Spare Production Leak Capacity

B DMSO ATF5 PGJ2 & U0126 (ATF5) PGJ2 ATF5 125 U0126 ATF5 U0126 + PGJ2 ATF5 100

25 * * ** * ** ** * OCR (pmol/min) 0 Basal ATP Proton Spare Production Leak Capacity

C DMSO siATF5 PGJ2 & U0126 (siATF5) PGJ2 siATF5 125 U0126 siATF5 U0126 + PGJ2 siATF5 100

25 **** **** *** * ** OCR (pmol/min) 0 Basal ATP Proton Spare Production Leak Capacity

Figure 17. ATF5 regulates mitochondrial respiration in PGJ2 stress conditions. The OCR was measured in (A) E18 cortical neurons incubated for 16hrs in DMEM and treated with and without PGJ2 (2.5µM) in the absence or presence of U0126 (2.5µM) for 16 h, (B) E18 cortical neurons transfected with an ATF5 construct using a magnetofection technique as described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (2.5µM) in the presence of U0126 (2.5µM) for 16 h, (C) ATF5 silenced E18 cortical neurons using an siRNA technique described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (2.5µM) in the presence of U0126 (2.5µM) for 16 h. Quantification of the mean OCR shown for respiration under basal conditions (basal), basal respiration rate minus the respiration rate recorded with oligomycin (ATP Production), Oligomycin response minus rotenone/antimycin A response (proton leak), and respiration rate obtained with FCCP minus the basal respiration rate (spare capacity). *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001 for comparisons to Control with or without ATF5 overexpression/silencing. ++P<0.01, +++P<0.001, ++++P<0.0001 for comparisons to PGJ2 with or without ATF5 overexpression/silencing.

47 3.2.2. ATF5 protects against ROS production induced by PGJ2 toxicity

Studies performed so far show oxidative stress increasing under conditions of PGJ2 toxicity. Oxidative stress ensues when there is an imbalance between reactive oxygen species

(ROS) production and the cellular antioxidant response (Ray-2012). Considering excessive ROS is associated with oxidative stress detecting ROS by staining live cortical neurons with Carboxy-

H2DCFCA will allow an assessment of oxidative damage. Fluorescent images illustrate significant ROS generation after a16hr treatment with PGJ2 (Figure 18, compare A2 to B2). Also affirming earlier findings, diminishing ROS in ATF5 overexpressing cells treated with PGJ2 demonstrate ATF5 protects against PGJ2 stress (Figure 18, compare B2 and D2). An inverse phenomenon of elevated ROS production was observed in ATF5 silenced neurons with or without PGJ2 treatment (Figure 18, compare B2, E2, and F2). Created during mitochondrial respiration, ROS experiences reductions or escalations depending on ATF5 expression, reinforcing the Cell Mitochondrial Stress Test analyses suggesting ATF5 protects against PGJ2 toxicity by mediating mitochondrial function.

Figure 18. ATF5 protects against ROS production induced by PGJ2 toxicity. (A) E18 cortical neurons incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h, (B) E18 cortical neurons transfected with an ATF5 construct using a magnetofection technique as described in the Materials and Methods section that were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) (C) ATF5 silenced E18 cortical neurons using an siRNA technique described in the Materials and Methods section that were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Cells were labeled with ROS (green) as described in Materials and Methods section. Nuclei were counterstained with DAPI. Data are representative of experiments repeated at least three times.

49 3.2.3. ATF5 maintains mitochondrial membrane potential against PGJ2 toxicity

Mitochondrial dysfunction indicated by Cell Mito Stress Tests (Figure 17) coincides with

HO-1 induction and Caspase-3 cleavage observed under conditions of PGJ2 toxicity (Figure 3).

Visualizing ROS production further associates mitochondrial impairment with PGJ2 stress

(Figure 19B2). Evaluating the ΔΨ of cortical neurons by labeling with TMRE, a positive m charged dye retained in healthy mitochondria due to high ΔΨ , revealed increased episodes of m

ΔΨ collapse with PGJ2 treatment (Figure 19A, compare A2 and B2). By overexpressing ATF5, m the depletion of ΔΨ associated with PGJ2 toxicity is reduced (Figure 19B, compare B2 and C2). m

In contrast, silencing ATF5 achieves the opposite effect, an increase in dissipating ΔΨ under m

PGJ2 stress conditions is observed (Figure 19C, compare B2 and F2). Improving TMRE retention in PGJ2 treated cells by elevating ATF5 expression demonstrates ATF5 restoration of mitochondrial health combats PGJ2 toxicity.

Figure 19. ATF5 maintains mitochondrial membrane potential against PGJ2 toxicity. (A) E18 cortical neurons incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h, (B) E18 cortical neurons transfected with an ATF5 construct using a magnetofection technique as described in the Materials and Methods section that were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) (C) ATF5 silenced E18 cortical neurons using an siRNA technique described in the Materials and Methods section that were incubated for 16hrs in DMEM and treated with and without PGJ2 (10µM) for 16 h. Cells were labeled with TMRE (green) as described in Materials and Methods section. Nuclei were counterstained with DAPI. Data are representative of experiments repeated at least three times.

51 3.2.4. Glycolysis contribution maintained under PGJ2 stress

Considering mitochondrial respiration is a major source of energy for neurons, severe mitochondrial impairment induced by PGJ2 resulted in only a moderate 20% loss of cell viability. Cell recovery achievement by ATF5 optimizing mitochondrial function inadequately satisfies the energy demand needed for accomplishing the restoration observed. Probing for an alternative energy source to compensate for mitochondrial dysfunction energy limitations prompted the usage of the Agilent Seahorse XF Glycolytic Rate Assay (GRA) to measure glycolysis. Performing the GRA Test on rat cortical neurons untreated and treated with PGJ2 alone, U0126 alone, or in combination generated similar basal levels (Figure 20A) for all conditions. Lower compensatory levels compared to Control indicated cells treated with PGJ2,

U0126 or in combination engaged glycolysis to a higher extent (Figure 20A). This escalation, detected under treatment conditions, implies glycolysis may be compensating for the reduction observed in mitochondrial respiration (Figure 17A). Moreover, these results suggest inhibiting

ERK regulation of mitochondrial function (Figure 17A) influences a cell’s glycolytic expenditure similarly to that exhibited under mitochondrial dysfunction induced by PGJ2 stress.

Both Cortical neurons with enhanced ATF5 expression (Figure 20B) and cells experiencing

ATF5 silencing (Figure 20C) illustrated deficits in basal and compensatory levels compared to untransfected cells for all treated and untreated circumstances (Figure 20A). Mimicking the basal and compensatory pattern of untransfected cells (Figure 20A) also depicts higher glycolytic activity for upregulated and silenced cells receiving treatment (Figure 20B and 20C). Increasing glycolysis may have contributed to improvements observed in the cell viability of PGJ2 treated cells upregulating ATF5 (Figure 4). However, under ERK inhibition the overexpression of ATF5 does not appear to enhance glycolysis compensation in a manner that relieves the loss in cell

52 viability (Figure 12) induced by PGJ2 stress. In comparing the contribution glycolysis and mitochondrial respiration supplies to fulfill the energetic demands of neurons, glycolysis appears to maintain the same contribution percentage pattern across treatments between untransfected and ATF5 upregulated or silenced cells (Figure 20D). Decreases in respiration OCR and glycolysis, observed under circumstances of ATF5 overexpression or silencing, leads to the speculation that glycolytic activity adjusts to ATF5 modulations of mitochondrial function. The ability for cell loss recovery by healthy cells overexpressing ATF5 compared to exhibiting an escalation of cell loss in distressed ATF5 silenced cells in which both glycolysis and mitochondrial respiration is maximized in a lower cellular energetic demand environment suggests mitochondrial respiration efficacy as a conduit for the more favorable outcome of higher cell viability in PGJ2 stressed cells experiencing elevated ATF5 expression.

53 B DMSO ATF5 PGJ2 & U0126 (ATF5) PGJ2 ATF5 150 U0126 ATF5 U0126 + PGJ2 ATF5 125

100 75 ns ns 50 *** ** 25 * **** 0 glycoPER (pmol/min)

Basal Compensatory

C DMSO siATF5 PGJ2 & U0126 (siATF5) PGJ2 siATF5 150 U0126 siATF5 U0126 + PGJ2 siATF5 125

100

75 ns ns 50 ** 25 ** *** 0 glycoPER (pmol/min)

Basal Compensatory

D DMSO % PER (Basal) PGJ2 U0126 100 U0126 + PGJ2 80

0 % PER from Glycolysis Without ATF5 ATF5 siATF5

Figure 20. Glycolysis contribution maintained under PGJ2 stress. The OCR and ECAR was measured in (A) E18 cortical neurons incubated for 16hrs in DMEM and treated with and without PGJ2 (2.5µM) in the absence or presence of U0126 (2.5µM) for 16 h, (B) E18 cortical neurons transfected with an ATF5 construct using a magnetofection technique as described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (2.5µM) in the presence of U0126 (2.5µM) for 16 h, (C) ATF5 silenced E18 cortical neurons using an siRNA technique described in the Materials and Methods section were incubated for 16hrs in DMEM and treated with and without PGJ2 (2.5µM) in the presence of U0126 (2.5µM) for 16 h. Quantification of the mean glycoPER shown for glycolysis under basal conditions (Total ECAR – Mitochondrial-derived CO2 contribution to acidification), rotenone/antimycin A response minus basal glycolysis (compensatory glycolysis), and % basal glycolysis/Total ECARrespiration (% PER from glycolysis). *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001 for comparisons to Control.

55 3.3. Discussion

As a major supplier of ATP, the mitochondrion plays an important role in cellular bioenergetics supporting cell function and cell survival/death pathways (Nicholls 2000; Bernardi

1999). An essential participant in the pathogenesis of neurodegenerative diseases, ascertaining mitochondrial health in cells allows an evaluation of its impairment under stress conditions. In our studies, we demonstrate ATF5 has the ability to mediate cellular respiration to protect against PGJ2 induced mitochondrial dysfunction.

Results from CMS assays suggest ATF5 protection against PGJ2 toxicity involves modulating mitochondrial functions. Expectedly, an increase in basal respiration and ATP production would indicate improved mitochondrial health, as observed in Control cells (Figure

17A). Overexpressing ATF5 produced lower basal OCR and proton leak in PGJ2 treated cells

(Figure 17B) demonstrating a reduction in respiration diminishes the presence of damage.

Focusing on altering respiration as a way to minimize proton leak would enhance cell viability

(Figure 4D) and enable improvements in efficacy. Furthermore, the significant decline in basal

OCR together with employing respiration maximum (Figure 17A) and exhibiting a reduction in cell viability (Figure 10) under U0126 treatment suggest ERK regulates mitochondrial function.

Comparisons between cell viability and CMS test results illustrated the larger impact observed in cell viability was not reflected in CMS tests. Although the tetrazolium salt used in cell viability assays measure mitochondrial metabolic rate, mitochondria are not the sole source of MTS reduction. Additional sources outside of the mitochondria contributing to MTS reduction might explain why CMS tests do not demonstrate greater significance in mitochondrial respiration between control and treated neurons. Another explanation for the disparity observed in the two

56 assays may result from not normalizing the CMS test results to cell viability at the completion of the assay. By normalizing to cell viability, comparisons between treated and control conditions might yield a data pattern with greater significance.

ATF5 protection against PGJ2 induced ROS and loss of membrane potential (Figure 15 and16) demonstrates mitochondrial impairment is associated with PGJ2 toxicity. These results coincide with the previous analysis conducted on mitochondrial function indicating ATF5 protection against PGJ2 stress targets respiration to re-establish mitochondrial health in an attempt to recover cell viability.

The Pasteur Effect and Crabtree Effect both describe the phenomenon where increasing either oxygen availability or glucose consumption, respectively, will result in the reduction of the other metabolic pathway (Swerdlow 2013). Findings suggest neurons have the ability to increase glycolysis in response to stimulation (Yellen 2018). A reciprocal glycolysis-respiration relationship exists under conditions of fixed energy demand (Swerdlow 2013). We demonstrate in our studies of glycolysis the existence of a compensatory relationship between glycolysis and mitochondrial respiration. Engaging glycolysis more extensively (Figure 20) appears to compensate for the reduction in mitochondrial respiration (Figure 17) exhibited by treated cells.

While increasing glycolysis under ATF5 overexpression improves cell viability in PGJ2 treated cells (Figure 12), it does not relieve the loss in cell viability experienced by PGJ2 stressed cells inhibiting ERK (Figure 12). Comparing the contributions made by glycolysis vs respiration towards cellular energetic demands (Figure 20D) revealed maintaining the balance between the

ATP generating pathways in varying ATF5 expression conditions involved modulating overall energy requirements. For treated cells upregulating or silencing ATF5 observing a decline in glycolysis (Figure 20B and 20C) complemented similar adjustments illustrated in mitochondrial

57 respiration (Figure 17B and 17C), thus preserving the balance with a universal reduction in cellular ATP needs.

Taken together, ATF5 protects against PGJ2 induced mitochondrial dysfunction by reducing ROS, loss of membrane potential, and mitochondrial damage. Low basal OCR accompanied by a decline in mitochondrial damage in PGJ2 treated neurons overexpressing

ATF5 suggests modulating mitochondrial functions as a protective mechanism. Attempts at improving mitochondrial efficacy decreased both mitochondrial respiration and glycolysis demonstrating a reciprocal glycolysis-respiration relationship entails maintaining metabolic pathway balance. The improvement in cell viability in PGJ2 or U0126 treated cells upregulating

ATF5 occurs under the depreciation of both mitochondrial respiration and glycolysis, however this restoration in PGJ2 stressed cells is impaired by the inhibition ERK regulating mitochondrial function.

CHAPTER IV

CONCLUSION

59 Overall, our present study demonstrates multiple layers are involved in the regulatory mechanism mediating ATF5 protection against PGJ2 induced toxicity in rat cortical neurons.

ATF5 protects primary cortical neurons against mitochondrial dysfunction associated with PGJ2 inflammation by reducing apoptotic activator Caspase-3 and oxidative stress indicator

HO-1. Further demonstrating its capacity to alleviate oxidative damage, ATF5 upregulation diminished the dissipation of mitochondrial membrane potential and ROS production. Analogous to ATFS-1 studies, illustrating ATF5 expression in PGJ2 stressed cortical neurons suggests mRNA transcript stabilization elicits accumulation as well as trafficking between the mitochondria and nucleus supporting the possibility of UPRmt induction, thereby activating mitochondrial protective genes in order to re-establish protein homeostasis. Assessing the oxygen consumption rate of cortical neurons revealed ATF5 possesses the capability to modulate mitochondrial respiration in an effort to improve mitochondrial health, consequently promoting survival against PGJ2 stress. The escalation in glycolysis, exhibited by PGJ2 stressed cells overexpressing ATF5, coinciding with the reduction in mitochondrial respiration suggests glycolysis contributes to the recovery of cell viability. Additionally, while ERK activation in the presence of PGJ2 stress propagates HO-1 and Caspase-3, it also portrays a prosurvival role independent of these inductions through the regulation of ATF5 and mitochondrial function.

In conclusion, the mediation of ATF5 neuroprotection against PGJ2 toxicity entails a coordinated effort by a multitude of regulatory mechanisms that recover mitochondrial functionality, essential for high energy-using neurons. Mitochondrial restorative pathways provide an avenue of exploration for therapeutic treatments targeting neurodegenerative disorders.

CHAPTER VII

MODEL

61 PGJ2

U0126

ROS ↑

ΔΨ P m ↓ MEK CASPASE-3 ↑

HO-1 ↑

P ERK

Neuronal Cell Death

ATF5

Activation Inhibition Translocation Neuroprotection

Figure 21. Model of ATF5 neuroprotection against PGJ2 induced toxicity. ATF5 protects against PGJ2 induced stress by reducing Caspase-3 cleavage, HO-1 induction, ROS production, and the loss of ΔΨ . The phosphorylation of ERK is necessary for ATF5 to m restore cell viability loss induced by PGJ2 toxicity possibly through the regulation of ATF5.

CHAPTER V

METHODOLOGY

63 5.1 Materials – Prostanglandins J2 analog CAY10410 from Cayman Chemical (Ann Harbor,

MI). The inhibitor of MEK1/2 (U0126) was obtained from Promega Corporation (Madison, WI).

Primary Antibodies: Caspase-3 (Cell Signaling Technologies, Danvers, MA, USA), HO-1 (Santa

Cruz Biotechnology, Santa Cruz, CA, USA), Actin (Sigma, St. Louis, MO), pERK (Cell

Signaling Technologies, Danvers, MA, USA), Total ERK (Santa Cruz Biotechnology, Santa

Cruz, CA, USA), a-tubulin (Calbiochem, San Diego, CA, USA), ATF5 (a generous gift provided by Dr. Cole Haynes’ Lab, University of Massachusetts Medical Center, Worchester,

MA; produced according to Zhou, Palam et al 2008).

5.2 Cell Cultures – Primary cerebral cortical cultures: Dissociated cultures from Sprague

Dawley rat embryonic (E18, both sexes) cerebral cortical neurons were prepared as follows. The isolated cortices free of meninges were digested with papain (0.5 mg/ml from Worthington

0 biochemical) in Hibernate E without calcium (BrainBits LLC) at 37 C for 30 min in a humidified atmosphere containing 5% CO2. After removal of the enzymatic solution, the tissues were gently dissociated in Neurobasal (NB) medium (Invitrogen). Dissociated tissues were centrifuged at 300 x g for 2 min. The pellet was resuspended in Neurobasal medium without antibiotics and plated on 10-cm dishes precoated with 50 µg/ml poly-D-lysine (Sigma). Cells were plated at a density of 6 x 106 cells/10-cm dish. Cultures were maintained in Neurobasal medium supplemented with

0 B27 and L-GlutaMAX (all from Invitrogen) at 37 C in a humidified atmosphere containing 5%

CO2. Half of the medium was changed every 4 days. Experiments were performed at 7 DIV.

5.3 Culture treatments – On DIV 7, cells were changed to DMEM supplemented with L-

GlutaMAX and sodium pyruvate (all from invitrogen) in the absence of serum, and incubated for

64 16 hr prior to harvest for western blotting or subject to cell viability assay, ROS and TMRE detection. All treatments were added during this 16 hr period as follows when applicable: PGJ2

(10 µM) and U0126 (10 µM) were treated for 16 hr. DMSO was used as vehicle control.

5.4 Protein Extraction and Western Blotting - After treatment, cells were rinsed with PBS and harvested in a lysis buffer as described previously (Rockwell et al. 2004). Protein concentrations were determined using a bicinchoninic acid assay (BCA) according to manufacturer’s instructions (Pierce, Rockford, IL). Equal amounts of protein (30 µg) were resolved by SDS- polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. Blots were then

0 blocked and incubated overnight at 4 C with the following primary antibodies: HO-1 and

ERK1/2 from Santa Cruz; Caspase-3 and pERK from Cell Signaling. Actin (Sigma-Aldrich) was used as a loading control. Immunoreactive bands were detected with corresponding anti-mouse and anti-rabbit secondary antibodies conjugated to horseradish peroxidase and visualized with the SuperSignal West Pico Chemiluminescent Substrate (Pierce Endogen, Rockford, IL). Data were quantified using ImageJ from NIH.

5.5 Cell Viability – Neurons were plated in pre-coated 96-well microtiter plates (2x104 cells/well) and maintained Neurobasal medium supplemented with B27 and L-GlutaMAX. On

DIV 7, cells were changed into DMEM as indicated for 16 hr prior to MTS assay. Prior to MTS assay measurement, all treatments were performed as follows when applicable: PGJ2 (10 µM) and U0126 (10 µM) were treated for 16 hr. On Div 8, cell viability was determined using colorimetric MTS assay (Promega Corp, Madison, WI) and quantified according to

65 manufacturer’s instruction. Survival measurements are expressed as the percent of the untreated control.

5.6 Oxidative Stress – ROS was measured using the fluorescent dye Carboxy-H2-DCFCA

(Invitrogen) following the manufacturer’s instruction. Briefly, neurons plated on precoated 8- well chamber slides were incubated in DMEM for 16 hr prior to adding H2-DCFCA. All treatments were performed as follows during this 16 hr if applicable: PGJ2 (10 µM) and U0126

(10 µM) were treated for 16 hr. THBP (100 µM) was used to induce ROS production as a positive control for 90 min. On DIV 8, 16 hr after in DMEM cells were incubated with 25 µM

H2-DCFCA following the manufacturer’s instructions and images were immediately taken using a Nikon Eclipse TE200 inverted epifluorescence scope. Data are representative of experiments that were repeated at least three times.

5.7 Mitochondrial Membrane Potential – Loss of mitochondrial membrane potential was assessed in living cells labeled with the fluorescent lipophilic cationic dye tetramethylrhodamine, ethyl ester (TMRE) (abcam). This dye incorporates into mitochondria with an intact transmembrane potential and its release serves as an indicator of a loss in inner membrane potential. Neurons plated on 8-well chamber slides were incubated in DMEM for 16 hr prior to adding TMRE. All treatments were performed as follows during this 16 hr if applicable: PGJ2

(10 µM) and U0126 (10 µM) were treated for 16 hr. FCCP (20 µM) was used for 30 min to disrupt the mitochondrial membrane potential as a positive control. On DIV 8, 16 hr after being

0 in DMEM, neurons were incubated with 200 nM TMRE for 30 min at 37 C, then washed with fresh medium. Fluorescence images were immediately taken using a Nikon Eclipse TE200

66 inverted epifluorescence scope. Data are representative of experiments that were repeated at least three times.

5.8 RNA Interference – On DIV 7, cultured neurons were transfected with 25nM of Atf5

SMARTpool® siRNA duplexes (Dharmacon RNA Technologies, Lafayette, CO) using

Dharmafect Transfection Reagents. After 32 hr, cells were changed to fresh DMEM for 16hrs.

All treatments were performed as follows during this 16 hr if applicable: PGJ2 (10 µM) and

U0126 (10 µM) were treated for 16 hr.

5.9 Transfection with magnetofection – On DIV 7, cultures were transfected by magnetofection using NeuroMag (Oz Biosciences) on a magnetic plate for 30 min at 37OC using a ratio of 1µg of plasmid DNA to 2ul NeuroMag according to manufacturer’s instructions. After

32 hr, cells were changed to fresh DMEM for 16hrs. All treatments were performed as follows during this 16 hr if applicable: PGJ2 (10 µM) and U0126 (10 µM) were treated for 16 hr. pCMVATF5 plasmid (a generous gift provided by Dr. Cole Haynes’ Lab, University of

Massachusetts Medical Center, Worchester, MA)

5.10 RT-PCR analysis – Total RNA was isolated from rat E18 cortical neurons with the

RNeasy® Mini Kit (Qiagen). Quantitative RT-PCR (qRT-PCR): Total RNA (1µg) was reverse transcribed to cDNA with the All-in-oneTM First-Strand cDNA Synthesis Kit (GeneCopoeia).

The quantitative analysis of the cDNAs were performed using the All-in-OneTM qPCR Mix Kit

(GeneCopoeia) in an Applied Biosystems ViiA 7 real-time PCR system with cycling parameters of 95OC for 10 min, 95OC for 10 sec, 58OC for 45 sec, and 72OC for 35 sec for 50 cycles.

67 Primer sequences used for qRT-PCR were rat-atf5-1: forward GTA TAA GGC ACG AAG CCA GA and reverse AGC TTG GTT CTC AGT TGC AC; rat-atf5-2: forward AGA GGG CAG AGT CAG TGG

AA and reverse GGA AGT GAA ATG GAG GGA CA; GAPDH: forward

CTTCTGTCTTGATGTTTGTCAACC and reverse ATGGTGAAGGTCGGTGTGAACG.

5.11 Cell Fractionation – To check the subcellular localization of ATF5, treated cells were separated to cytosol, mitochondria, and nucleus with cell fractionation kit (Abcam, ab1097919) following manufacturer’s instructions, except 2x Buffer A was used in 50% of volume recommended.

5.12 Respiration Analysis – Oxygen consumption rate (OCR) and extra acidification rate

(ECAR) were measured using Seahorse XFp Extracellular Flux Analyzer (Seahorse

Biosciences). Mitochondrial respiration assays were performed using a Seahorse XFp Cell Mito

Stress Test Kit and cellular glycolysis was measured using a Seahorse XFp Glycolytic Rate

Assay Kit, in accordance with manufacturer’s instructions. Briefly, primary cortical neurons were seeded at 40,000 cells per well in 8-well Seahorse cell culture miniplates. Following treatments, on DIV 8 or 9 culture media in each well was replaced with Seahorse assay media and incubated at 37OC in an incubator w/o CO2 for 1 hr. Then the plates were subjected to analysis using 1 µM oligomycin, 500 nM FCCP, and 500 nM rotenone/antimycin for Cell Mito

Stress Test Kit or 500 nM rotenone/antimycin and 50 mM 2-deoxy-D-glucose (2-DG) for

Glycolytic Rate Assay kit as indicated. The Seahorse assays were analyzed using XFp Wave software, according to manufacturer’s instructions.

68 5.13 Statistical Analysis- Data are expressed as the mean + SEM of experiments, all of which were replicated at least three times. Statistical significance was assessed by a one-way ANOVA

(Tukey or Bonferroni analysis) with the Prism 7 program. A difference resulting in P<0.05 is considered significant.

CHAPTER VI

FUTURE DIRECTIONS

70 6.1 Does the overexpression of ATF5 modulate the expression of mitochondrial protective genes?

Techniques such as reverse transcription polymerase chain reaction (RT-PCR) and microarray analysis can be utilized to identify which, if any mitochondrial protective genes are expressed when ATF5 is upregulated. Activation of the UPRmt has been associated with increased transcriptions of HSP60, mtHSP70, and LONP1. Their inductions would indicate attempts at re-establishing protein homeostasis. The effect of ATF5 on mitochondrial gene expression will be determined using Mouse Mitochondria PCR array (Qiagen) which profiles the expression of genes for mitochondrial biogenesis, energy production and function with controls by quantitative PCR. These results will show whether ATF5 mimics ATFS-1 in protecting against mitochondrial impairment by distinguishing genes upregulated in an ATFS-1 dependent manner.

6.2 Does the silencing ATF5 induce a Caspase independent cell death in cortical neurons?

PGJ2 treatment induces cell death accompanied by an increase in Capase-3 cleavage.

ATF5 silencing exacerbates cell viability but not Caspase-3 cleavage. To investigate further if both events are caspase-dependent, neurons receiving PGJ2 will be treated with the general

Caspase inhibitor z-VAD-fmk in the absence or presence of ATF5. Cell viability will be measured. A continued presence of cell loss would indicate another method of cell death, besides caspase-dependent, is associated with PGJ2 toxicity. Considering autophagy is also a possibility,

71 monitoring LC3-I conversion to LC3-II by immunoblotting would allow an assessment of this scenario. An increased amount of LC3-II reflects increased autophagosome formation as a result of autophagy induction.

6.3 Does ATF5 trafficking to the nucleus in PGJ2 stressed conditions mediate protection?

The ability of ATF5 to localize to both the mitochondria and the nucleus is due to possessing an N-terminal MTS region and a nuclear localization signal (NLS) in the leucine- zipper domain. Transfecting a pCMVATF5 plasmid without the NLS, preventing ATF5 localization to the nucleus of cortical neurons, would allow insight into protection mediated by nuclear trafficking. If protection against PGJ2 is maintained without the NLS present, then nuclear trafficking would be shown not to mediate protection. However, loss of ATF5 protection indicated by increased Caspase-3, HO-1, ROS induction, along with cell loss would show nuclear trafficking mediates protection.

6.4 Does ATF5 activate the UPRmt in PGJ2 stressed cortical neurons?

Induction of the mitochondrial chaperone hsp-60 promoter is associated with activation of the UPRmt. To investigate whether PGJ2 stress in cortical neurons induces the UPRmt, a GFP reporter containing the hsp-60 promoter would be generated and introduced to cortical neurons.

Viewing cultures under a fluorescence microscope would allow detection of hsp-60pr::gfp expression. Activation of hsp-60pr::gfp in a PGJ2 stress environment by cells overexpressing

72 mt ATF5 would illustrate ATF5 induces the UPR . The lack of hsp-60pr::gfp expression would instead indicate ATF5 does not activate the UPRmt.

6.5 How to achieve a more statistically accurate result of ROS production?

The method used to illustrate the presence of ROS in this work was not statistically analyzed. In order to accomplish that flow cytometry would be applied so that dead cells can be distinguished from live cells. Since ROS is generated in both dead and live cells, being able to distinguish the ROS in only live cells would enable an accurate assessment of PGJ2 toxicity vs a normal course of cell death that occurs in a culture of cells. The use of CellROX Green Flow

Cytometry Assay Kit to bind DNA will exhibit a strong fluorogenic signal upon oxidation that remains localized to the nucleus and cytoplasm. When used together with SYTOX red dead cell stain, ROS production strictly by live cells can be identified by flow cytometry.

REFERENCES

74 Andreyev, A., Kushnareva, Y., Starkov, A. (2005). Mitochondrial metabolism of reactive oxygen species. Biochemistry (Moscow) 70(2): 200-214

Beal, M. F. (2005). Mitochondria take center stage in aging and neurodegeneration. Annals of Neurology 58(4): 495-505

Bernardi, P., Scorrano, L., Colonna, R., Petronilli, V., Di Lisa, F. (1999). Mitochondria and cell death. European Journal of Biochemistry 264(3): 687-701

Celardo, I., Martins, L. M., Ghandi, S. (2014). Unraveling mitochondrial pathways to Parkinson’s disease. British Journal of Pharmacology 171(8): 1943-1957

Chandra, D., Tang, D. G. (2003). Mitochondrially localized active caspase-9 and caspase-3 result mostly from translocation from the cytosol and partly from caspase-mediated activation in the organelle. Lack of evidence for Apaf-1 mediated procaspase-9 activation in the mitochondria. The Journal of biological chemistry 278(19): 17408-17420

Chang, L. K., Johnson, E. M. (2002). Cyclosporin A inhibits caspase-independent death of NGF- deprived sympathetic neurons: a potential role for mitochondrial permeability transition. The Journal of cell biology 157(5): 771-781

Chang, L. K., Putcha, G. V., Deshmukh, M., Johnson, E. M. (2002). Mitochondrial involvement in the point of no return in neuronal apoptosis. Biochimie 84(2): 223-231 Chen, T., Chi, K., Wang, J., Chien, C., Lin, W. (2009). Reactive oxygen species are involved in Fas-Linduced caspase-independent cell death and inflammatory responses. Free Radical Biology and Medicine 46(5): 643-655

Chipuk, J. E., Bouchier-Hayes, L., Green, D. R. (2006). Mitochondrial outer membrane permeabilization during apoptosis: the innocent bystander scenario. Cell Death and Differentiation 13(8): 1396-1402

Claros, M. G. and Vincens, P. (1996). Computational method to predict mitochondrially imported proteins and their targeting sequences. European Journal of Biochemistry 241(3): 779- 786

Cong, Z., Wang, H., Wang, J., Zhou, Y., Pan, H., Zhang, D., Zhu, L. (2013). ERK and PI3K signaling cascades induce Nrf2 activation and regulate cell viability partly through Nrf2 in human glioblastoma cells. Oncology Reports 30: 715-722

Cooper, G. (2000). The Cell: A molecular approach, 2nd Edition, Chapter 10.

Cory, S., Adams, J. (2002). The Bcl2 family: regulators of the cellular life-or-death switch. Nature Reviews Cancer 2(9): 647-656

75 Deshmukh, M., Kuida, K., Johnson, E. M. (2000). Caspase inhibition extends the commitment to neuronal death beyond cytochrome c release to the point of mitochondrial depolarization. The Journal of cell biology 150(1): 131-143

Deshmukh, M., Vasilakos, J., Deckwerth, T. L., Lampe, P.A., Shivers, B. D., Johnson, E. M. (1996). Genetic and Metabolic Status of NGF-deprived Sympathetic Neurons Saved by an Inhibitor of ICE Family Proteases. The Journal of Cell Biology 135(5): 1341-1354

Desire, L., Courtois, Y., Jeanny, J. (2000). Endogenous and Exogenous Fibroblasts Growth Factor 2 Support Survival of Chick Retinal Neurons by Control of Neuronal bcl-xL and bcl-2 Expression Through a Fibroblast Growth Factor Receptor 1- and ERK-Dependent pathway. Journal of Neurochemistry 75(1): 151-163

Figueiredo-Pereira, M.E., Rockwell, P., Schmidt-Glenewinkel, T., Serrano, P. (2015). Neuroinflammation and J2 prostaglandins: linking impairment of the ubiquitin-proteasome pathway and mitochondria to neurodegeneration. Frontiers in Molecular Neuroscience 7(104): 1- 20

Fiorese, C. J., Schulz, A. M., Lin, Yi-Fan, Rosin, N., Pellegrino, M. W., Haynes, C. M. (2016). The Transcription Factor ATF5 Mediates a Mammalian Mitochondrial UPR. Current Biology 26(15): 2037-2043

Greene, L. A., Lee, H. Y., Angelastro, J. M. (2009). The transcription factor ATF5: role in neurodevelopment and neural tumors. Journal of Neurochemistry 108(1): 11-22

Hatano, M., Umemura, M., Kimura, N., Yamazaki, T., Takeda, H., Nakano, H., Takahashi, S., Takahashi, Y. (2013). The 5’-untranslated region regulates ATF5 mRNA stability via non-sense- mediated mRNA decay in response to environmental stress. The FEBS Journal 280(18): 4693- 4707

Haynes, C. M., K. Petrova, C. Benedetti, Y. Yang and D. Ron (2007). ClpP mediates activation of a mitochondrial unfolded protein response in C. elegans. Developmental Cell 13(4): 467-480.

Haynes, C. M., Fiorese, C.J., Lin, Y. (2013). Evaluating and responding to mitochondrial dysfunction: the mitochondrial unfolded-protein response and beyond. Trends in Cell Biology 23(7): 311-318

Hernandez, I., Torres-Peraza, J., Santos-Galindo, M., Ramos-Moron, E., Fernandez-Fernandez, M., Perez-Alvarez, M., Miranda-Vizuete, A., Lucas, J. (2017). The neuroprotective transcription factor ATF5 is decreased and sequestered into polyglutamine inclusions in Huntington’s disease. Acta Neuropathologica 134(6): 839-850

Hetz, C., Chevet, E., Harding, H. P. (2013). Targeting the unfolded protein response in disease. Nature Reviews Drug discovery 12(9): 703-19 Hirsch, E. C., Hunot, S. (2009). Neuroinflammation in Parkinson’s disease: a target for neuroprotection? Lancet Neurology 8(4): 382-397

Jin, K., Mao, X. O., Zhu, Y., Greenberg, D. A. (2002). MEK and ERK protect hypoxic cortical neurons via phosphorylation of Bad. Journal of Neurochemistry 80(1): 119-125

Jovaisaite, V., Mouchiroud, L., Auwerx, J. (2014). The mitochondrial unfolded protein response, a conserved stress response pathway with implications in health and disease. The Journal of Experimental Biology 217(pt 1): 137-143

Keller, J. M., Escara-Wilke, J. F., Keller, E. T. (2008). Heat stress-induced heat shock protein 70 expression is dependent on ERK activation in zebrafish (Danio rerio) cells. Comparative Biochemical and Physiology, Part A, 150(3): 307-314

Keyse, S. M., Tyrrell, R. M. (1989). Heme oxygenase is the major 32-kDa stress protein induced human skin fibroblasts by UVA radiation, hydrogen peroxide, and sodium arsenite. Proceedings of the National Academy of Sciences of the United States of America 86(1): 99-103

Kim, J., Kang, T. (2018). Differential Roles of Mitochondrial Translocation of Active Caspase-3 and HMGBI in Neuronal Death Induced by Status Epilepticus. Frontiers in Cellular Neuroscience 12(301): 1-17

Kitz, K., Windischhofer, W., Leis, H, Huber, E., Kollroser, M., Malle, E. (2011). 15-deoxy- D12,14-prostaglandin J2 induces Cox-2 expression in human osteosarcoma cells through MAPK and EGFR activation involving reactive oxygen species. Free Radical Biology and Medicine 50(7): 854-865

Kondo, M., Oya-Ito, T., Kumagai, T., Osawa, T., Uchida, K. (2001). Cyclopentenone prostaglandins as potential inducers of intracellular oxidative stress. The Journal of biological chemistry 276(15): 12076-12083

Lartigue, L., Kushnareva, Y., Seong, Y., Lin, H., Faustin, B., Newmeyer, D. (2009). Molecular biology of the cell 20(23): 4871-4884

Lee, D. R., Kwon, C. H., Park, J. Y., Kim, Y. K., Woo, J. S. (2009). 15-Deoxy-D 12,14- prostaglandin J2 induces mitochondrial-dependent apoptosis through inhibition of PKA/NF-kB in renal proximal epithelial cells. Toxicology 258(1): 17-24

Ley, R., Balmanno, K., Hadfield, K. Weston, C., Cook, S. J. (2003). Activation of the ERK1/2 signaling pathway promotes phosphorylation and proteasome-dependent degradation of the BH3-only protein, Bim. The Journal of biological chemistry 278(21): 18811-18816

Li, G., Xu, Y., Guan, D., Liu, Z., Liu, D. X. (2011). HSP70 protein promotes survival of C6 and U87 glioma cells by inhibition of ATF5 degradation. The Journal of biological chemistry 286(23): 20251-20259

77 Li, P., Nijhawan, D., Budihardjo, I., Srinivasula, S. M., Ahmad, M., Alnemri, E. S., Wang, X. (1997). Cytochrome c and dATP-Dependent Formation of Apaf-1/Caspase-9 Complex Initiates an Apoptotic Protease Cascade. Cell 91(4): 479-489

Li, Q., Chen, M., Liu, H., Yang, L., Yang, T., Yang, G., He, G. (2014). The dual role of ERK signaling in the apoptosis of neurons. Frontiers in Bioscience 19: 1411-1417

Lin, M. T., Beal, M. F. (2006). Mitochondrial dysfunction and oxidative stress in neurodegenerative diseases. Nature 443(7113): 787-795

Lionaki, E., Markaki, M., Palikaras, K., Tavernarakis, N. (2015). Mitochondria, autophagy and age-associated neurodegenerative diseases: new insights into a complex interplay. Biochimica et Biophysica Acta – Bioenergetics 1847(11): 1412-1423

Liu, D. X., Qian, D., Wang, B., Yang, J., Lu, Z. (2011). p300-Dependent ATF5 acetylation is essential for Egr-1 gene activation and cell proliferation and survival. Molecular and cellular biology 31(18): 3906-3916

Ma, Y., Liu, W., Wang, Y., Chao, X., Qu, Y., Wang, K., Fei, Z. (2011). VEGF protects rat cortical neurons from mechanical trauma injury induced apoptosis via the MEK/ERK pathway. Brain Research Bulletin 86(5-5): 441-446

Manczak, M., Jung, Y., Park, B. S., Partovi, D., Reddy, P. H. (2005). Time-course of mitochondrial gene expressions in mice brains: implications for mitochondrial dysfunction, oxidative damage, and cytochrome c in aging. Journal of Neurochemistry 92(3): 494-504

May, M. J., Madge, L. A. (2007). Caspase inhibition sensitizes inhibitor NF-kappaB kinase beta- deficient fibroblasts to caspase-independent cell death via the generation of reactive oxygen species. The Journal of biological chemistry 282(22): 16105-16116

Mayer, M., Bukau, B. (2005). Hsp70 chaperones: cellular functions and molecular mechanism. Cellular and Molecular Life Sciences 62(6): 670-684

McManus, M. J., Murphy, M. P., Franklin, J. L. (2014). Mitochondria-derived reactive oxygen species mediate caspase-dependent and -independent neuronal deaths. Molecular and Cellular Neuroscience 63: 13-23 Nargund, A. M., Pellegrino, M. W., Fiorese, C. J., Baker. B. M., Haynes, C. M. (2012). Mitochondrial import efficiency of ATFS-1 regulates mitochondrial UPR activation. Science 337(6094): 587-590

Ng, D. C. and Bogoyevitch, M. A. (2000). The mechanism of heat shock activation of ERK mitogen-activated protein kinases in the interleukin 3-dependent ProB cell line BaF3. The Journal of biological chemistry 275(52): 40856-40866

Nicholls, D., Budd, S. (2000). Mitochondria and neuronal survival. Physiological Reviews *0(1): 315-360

Pakos-Zebrucka, K., Koryga, I., Mnich, K., Ljujic, M., Samali, A., Gorman, A. M. (2016). The integrated stress response. EMBO reports 17(10): 1374-1395

Papa, S., Martino, P. L., Capitanio, G., Gaballo, A., De Rasmo, D., Signorile, A., Petruzzella, V. (2012). Advances in Experimental Medicine and Biology, Advances in Mitochondrial Medicine Chapter 1: 3-37

Perkins, D., Pereira, E. F. R., Aurelian, L. (2003). The Herpes Simplex Virus Type 2 R1 Protein Kinase (ICP10 PK) Functions as a Dominant Regulator of Apoptosis in Hippocampal Neurons Involving Activation of the ERK Survival Pathway and Upregulation of the Antiapoptotic Protein Bag-1. The Journal of Virology 77(2): 1292- 1305 Pierre, S. R., Lemmens, M. A., Figueiredo-Pereira, M.E. (2009). Subchronic infusion of the product of inflammation prostaglandin J2 models sporadic Parkinson’s disease in mice. Journal of Neuroinflammation 6(18): 1-12

Putcha, G. V., Deshmukh, M., Johnson, E. M. (1999). BAX Translocation is a Critical Event in Neuronal Apoptosis: Regulation by Neuroprotectants, BCL-2, and Caspases. The Journal of Neuroscience 19(17): 7476-7485

Puthalakath, H., Strasser, A. (2002). Keeping killers on a tight leash: transcriptional and post- transcriptional control of the pro-apoptotic activity of BH3-only proteins. Cell Death and Differentiation 9(5): 505-512

Qi, Z., Qi, S., Gui, L., Shen, L., Feng, Z. (2016). Daphnetin protects oxidative stress-induced neuronal apoptosis via regulation of MAPK signaling and HSP70 expression. (mitogen activated protein kinase and heat shock protein 70) (Report). Oncology Letters 12(3): 1959-1964

Ray, P. D., Huang, B., Tsuji, Y. (2012). Reactive oxygen species (ROS) homeostasis and redox regulation in cellular signaling. Cellular Signalling 24(5): 981-990

Reddy, R. H. (2009). Role of mitochondria in neurodegenerative disease: mitochondria as a therapeutic target in Alzheimer’s disease. CNS Spectrums 14(8 Suppl 7): 8-13

Ricci, J-E., Waterhouse, N., Green, D. R. (2003). Mitochondrial fucntions during cell death, a complex (I-V) dilemma. Cell Death and Differentiation 10(5): 488-492

Rockwell, P., Martinez, J., Papa, L., Gomes, E. (2004). Redox regulates COX-2 upregulation and cell death in the neuronal response to cadmium. Cellular Signalling 16(3): 343-353

Sanchez, A., Tripathy, D., Yin, X., Luo, J., Martinez, J., Grammas, P. (2012). Pigment epithelium-derived factor (PEDF) protects cortical neurons in vitro from oxidant injury by activation of extracellular signal-regulated kinase (ERK) ½ and induction of Bcl-2. Neuroscience Research 72(1):1-8

Schapira, A. H. V. (2012). Mitochondrial diseases. The Lancet 379(9828): 1825-1834

Shamas-Din, A., Kale, J., Leber, B., Andrews, D. W. (2013). Mechanisms of action of Bcl-2 family proteins. Cold Spring Harbor perspectives in biology 5(4): 1-21

Shiver, K. Y., Nikolopoulou, A., Machlovi, S. I., Vallabhajosula, S., Figueiredo-Pereira, M.E. (2014). PACAP27 prevents Parkinson-like neuronal loss and motor deficits but not microglia activation induced by prostaglandin J2. Biochimica et Biophysica Acta 1842(9): 1707- 1719

Smith, A. M., Depp, C., Ryan, B. J., Johnston, G. I., Alegre-Abarrategui, J., Evetts, S., Rolinski, M., Baig, F., Ruffmann, C., Simon, A. K., Hu, M. T. M., Wade-Martins, R. (2018). Mitochondrial dysfunction and increased glycolysis in prodromal and early Parkinson’s blood cells. Movement Disorders 33(10): 1580-1590

Swerdlow, R. H., E, Lezi., Aires, D., Lu, J. (2013). Glycolysis-respiration relationships in a neuroblastoma cell line. Biochimica et Biophysica Acta – General Subjects 1830(4): 2891-2898

Tait, S. W. and Green, D. R. (2008). Caspase-independent cell death: leaving the set without the final cut. Oncogene 27(50): 6452-6461

Thiaville, M. M., Pan, Y., Gjymishka, A., Zhong, C., Kaufman, R. J., Kilberg, M. S. (2008). MEK signaling is required for phosphorylation of eIF2alpha following amino acid limitation of HepG2 human hepatoma cells. The Journal of biological chemistry 283(16): 10848-10857

Torres-Peraza, J. F., Engel, T., Martin-Ibanez, R., Sanz-Rodriguez, A., Fernandez-Fernandez, M. R., Esgleas, M., Canals, J. M., Henshall, D. C., Lucas, J. (2013). Protective neuronal induction of ATF5 in endoplasmic reticulum stress induced by status epilepticus. Brain 136(4): 1161-76

Uekusa, H., Namimatsu, M., Hiwatashi, Y., Akimoto, T., Nishida, T., Takahashi, S., Takahashi, Y. (2009). Cadmium interferes with the degradation of ATF5 via a post-ubiquitination step of the proteasome degradation pathway. Biochemical and Biophysical Research Communications 380(3): 673-678

Wang, Z., Aris, V. M., Ogburn, K. D., Soteropoulos, P., Figueiredo-Pereira, M. E. (2006). Prostaglandin J2 alters pro-survival and pro-death gene expression patterns and 26S proteasome assembly in human neuroblastoma cells. The Journal of biological chemistry 281(30): 21377- 21386

Watatani, Y., Kimura, N., Shimizu, Y., Akiyama, I., Tonaki, D., Hirose, H., Takahashi, S., Takahashi, Y. (2007). Amino acid limitation expression of ATF5 mRNA at the post- transcriptional level. Life Sciences 80(9): 879-885

Wei, Y., Jiang, J., Liu, D., Zhou, J., Chen, X., Zhang, S., Zong, H., Yun, X., Gu, J. (2008). Cdc34-mediated degradation of ATF5 is blocked by cisplatin. The Journal of biological chemistry 283(27): 18773-18781

80 Weissman, L., de Souza-Pinto, N. C., Stevnsner, T., Bohr, V. A. (2007). DNA repair, Mitochondria, and neurodegeneration. Neuroscience 145(4): 1318-1329

Weston, C., Balmanno, K., Chalmers, C., Hadfield, K., Molton, S., Ley, R., Wagner, E., Cook, S. (2003). Activation of ERK1/2 by [delta]Raf-1: ER* represses Bim expression independently of the JNK or PI3K pathways. Oncogene 22(9): 1281-1293

Yellen, G. (2018). Fueling thought: Management of glycolysis and oxidative phosphorylation in neuronal metabolism. The Journal of cell biology 217(7): 2235-2246

Xu, Y., Kim, S. O., Li, Y., Han, J. (2006). Autophagy contributes to caspase-independent macrophage cell death. The Journal of biological chemistry 281(28): 19179-19187

Yu, J., Jiang, Z., Ning, L., Zhao, Z., Yang, N., Chen, L., Ma, H., Li, L., Fu, Y., Zhu, H., Qi, H. (2015). Protective HSP70 induction by Z-Ligustilide against oxygen glucose deprivation injury via activation of the MAPK Pathway but not of HSF1. Biological and Pharmaceutical Bulletin 38(10): 1564-72

Zhou, D., Palam, L. R., Jiang, L., Narasimhan, J., Staschke, K. A., Wek, R. C. (2008). Phosphorylation of eIF2 directs ATF5 translational control in response to diverse stress conditions. The Journal of biological chemistry 283(11): 7064-73

Zoccarato, F., Cavallini, L., Alexandre, A. (2004). Repiration-dependent removal of exogenous H2O2 in brain. The Journal of biological chemistry 279(6): 4166-4174