Gene Delivery of Adenoviral-TMBIM6 Vector Protects the Neonatal Brain After Hypoxic-Ischemic Injury Desislava Doycheva

Loma Linda University TheScholarsRepository@LLU: Digital Archive of Research, Scholarship & Creative Works

Loma Linda University Electronic Theses, Dissertations & Projects

6-2018 Gene Delivery of Adenoviral-TMBIM6 Vector Protects the Neonatal Brain after Hypoxic-Ischemic Injury Desislava Doycheva

Follow this and additional works at: http://scholarsrepository.llu.edu/etd Part of the Medical Physiology Commons

Recommended Citation Doycheva, Desislava, "Gene Delivery of Adenoviral-TMBIM6 Vector Protects the Neonatal Brain after Hypoxic-Ischemic Injury" (2018). Loma Linda University Electronic Theses, Dissertations & Projects. 509. http://scholarsrepository.llu.edu/etd/509

This Dissertation is brought to you for free and open access by TheScholarsRepository@LLU: Digital Archive of Research, Scholarship & Creative Works. It has been accepted for inclusion in Loma Linda University Electronic Theses, Dissertations & Projects by an authorized administrator of TheScholarsRepository@LLU: Digital Archive of Research, Scholarship & Creative Works. For more information, please contact [email protected].

LOMA LINDA UNIVERSITY School of Medicine in conjunction with the Faculty of Graduate Studies

______

Gene Delivery of Adenoviral-TMBIM6 Vector Protects the Neonatal Brain after Hypoxic-Ischemic Injury

Desislava Doycheva

______

A Dissertation submitted in partial satisfaction of the requirements for the degree Doctor of Philosophy in Physiology

______

June 2018

Each person whose signature appears below certifies that this dissertation in his/her opinion is adequate, in scope and quality, as a dissertation for the degree Doctor of Philosophy.

, Chairperson John H. Zhang, Professor of Physiology

Danilyn Angeles, Professor of Physiology

Eileen Brantely, Assistant Professor of Pharmacology

Jiping Tang, Professor of Physiology

Lubo Zhang, Professor of Pharmacology

iii

ACKNOWLEDGEMENTS

First and foremost, I would like to express my deepest gratitude to Dr. Zhang and

Dr. Tang for giving me this opportunity and for all their time invested in my academic pursuit. Words cannot describe how thankful I am for all your mentorship and guidance, patience, support, and for encouraging me on this journey. I am grateful for the numerous opportunities they have given me to not only advance in the scientific field but also to gain leadership and managerial skills. I would also like to extend my thanks to all the laboratory members of Dr. Zhang and Dr. Tang, I am very appreciative of all your help and collaborations, all the hard work and good times we have had. Thank you for sharing all your knowledge with me and for helping me be a better scientist, in particular Prativa

Sherchan.

To my committee members, Dr. Danilyn Angeles, Dr. Eileen Brantley and Dr.

Lubo Zhang, I am grateful for their guidance, support and direction. It has been such a great honor to have had them work with me and I thank them for their contributions.

To my friends and family, I am forever grateful for all their encouragements and for being there for me in both good and difficult times. Most importantly thank you to my husband, Jesse Lichtenfeld, and to my parents, Maria Nelson and Peder Nielsen, for all your love, support, sacrifice and inspiration to help me stay on course and make this possible.

CONTENT

Approval Page ...... iii

Acknowledgements ...... iv

Table of Contents ...... v

List of Figures ...... ix

List of Abbreviations ...... x

Abstract ...... xiii

Chapter

1. Introduction ...... 1

Neonatal Hypoxia Ischemia ...... 1

Epidemiology ...... 2 Pathophysiology ...... 3 Apoptotic Mechanisms ...... 4 Inflammatory Mechanisms ...... 7 Oxidative Stress ...... 9 Clinical Presentation and Therapeutic Strategies ...... 10

Hypoxic Ischemic Injury and Endoplasmic Reticulum Stress ...... 11

Apoptosis ...... 13 Oxidative Stress ...... 14 Inflammation ...... 16 Endoplasmic Reticulum Stress Suppressors ...... 17

The Transmembrane Bax Inhibitor-1 (TMBIM) Protein Super Family ...... 17

Bax Inhibitor-1 (BI-1/TMBIM6) Protein Family ...... 18 Bax Inhibitor-1 Like Proteins ...... 19

TMBIM1 or Lifeguard 3 ...... 19 TMBIM2 or Lifeguard 2 ...... 19 TMBIM3 or Lifeguard 1 ...... 20 TMBIM4 or Lifeguard 4 ...... 20 Lifeguard 5 ...... 20

Bcl-2/Bcl-xl...... 21 MicroRNAs ...... 21

Bax Inhibitor-1 and Endoplasmic Reticulum Stress ...... 21

BI-1 as a Regulator of Apoptosis ...... 21 BI-1 as a Regulator of Oxidative Stress and Inflammation ...... 23

Specific Aims ...... 24

Aim 1 ...... 26 Aim 2 ...... 27

References ...... 30

2. Gene Delivery of Adenoviral-TMBIM6 Vector Restores ER stress Induced Apoptosis in a Neonatal HI Rat Model ...... 48

Abstract ...... 49 Introduction ...... 51 Materials and Methods ...... 53

In Vitro Experiments ...... 53

Oxygen Glucose Deprivation Model ...... 53 Cell Death Assay...... 53 Western Blotting of Cells...... 54 Calculation of MOI ...... 54 Transfections ...... 54

In Vivo Experiments ...... 55

Hypoxic Ischemic Rat Model ...... 56 Drug Administration ...... 57 Infarction Area Calculation...... 57 Behavioral Analysis ...... 57 Western Blotting of Brain Tissue ...... 58 Histology ...... 59

Statistical Analysis ...... 61

Results ...... 61

To Establish Optimum Time for OGD and Optimal MOI Needed of Ad-TMBIM6 Vector to Successfully Transfect PC12 cells ...... 61

To Determine the Effect of Ad-TMBIM6 on Percent Cell Viability in an in vitro OGD Model ...... 62 To Establish BI-1’s Anti-Apoptotic Effects via the IRE1α Signaling Pathway in an in vitro OGD Model ...... 64 Expression Levels of Endogenous BI-1, IRE1α, XBP1 and CHOP in Time Dependent Manner post HI ...... 67 Low Dose Ad-TMBIM6 Administered at 48h pre-HI Reduced Infarct Area at 72h post HI ...... 67 Histological Staining at 72h post ...... 69 Ad-TMBIM6 Improved Long-Term Neurological Function at 4 weeks post HI ...... 71 BI-1 siRNA and IRE1α CRISPR Activation Plasmid Reversed Ad- TMBIM6 Protective Effects at 72h post HI ...... 73 BI-1 Exerted its Neuroprotective Effects via the IRE1α/XBP1/CHOP Pathway at 72h post HI ...... 75 Overexpression of BI-1 Suppressed Pro-Apoptotic Protein Markers and Attenuated Neuronal Apoptosis at 72h post HI ...... 78

Discussion ...... 81 References ...... 87

3. TMBIM6 Gene Protects the Neonatal Brain from Oxidative Stress and Inflammation via Disruption of NPR-CYP Complex Coupled with Upregulation of Nrf-2 in a HI Rat Model ...... 94

Abstract ...... 95 Introduction ...... 97 Materials and Methods ...... 100

In Vivo Experiments ...... 100

Hypoxic Ischemic Rat Model ...... 100 Time Course Evaluation of Proteins ...... 101 Viral Administration ...... 101 RNAi Administration ...... 102 Infarct Area Measurements ...... 102 Western Blotting for Brain Tissue ...... 103 Histology ...... 104

In Vitro Experiments ...... 105

Oxygen Glucose Deprivation Model ...... 105 Cell Death Assay...... 106 Western Blotting for Cells ...... 106 Calculation of MOI ...... 107 siRNA Transfection ...... 107

vii

Statistical Analysis ...... 108

Results ...... 108

Time Course Expression of Endogenous Proteins post HI ...... 108 To Determine Amount of Ad-TMBIM6 Viral Vector Present in the Brain and its Effects on Infarction at 72h post HI ...... 110 Overexpression of BI-1 Disrupted the NPR-CYP Complex at 72h post HI ...... 112 BI-1 Upregulated pNrf-2 Expression and Induced Anti-Oxidant Enzyme Production at 72h post HI ...... 114 BI-1 Overexpression Attenuated ROS Production and Inflammation at 72h post HI ...... 116 To Investigate BI-1’s Effects on NPR-CYP Complex in an in vitro OGD Model ...... 120

Discussion ...... 123 References ...... 130

4. Discussion and Conclusion ...... 136

Clinical Significance ...... 143

Can BI-1 Be a Potential therapeutic Target? ...... 143

Conclusion ...... 144 Future Directions ...... 145 References ...... 147

viii

FIGURES

Figures Page

1.1: Schematic Representation of the Central Hypothesis and Specific Aims ...... 29

2.1: Percent Cell Viability Significantly Improved after Transfection of cells with 100MOI of Ad-TMBIM6 in OGD Model...... 63

2.2: BI-1 Exerted its Anti-Apoptotic Effects via Inhibition of IRE1α in OGD Model ...... 66

2.3: To Measure Expression Levels of Endogenous Proteins and Determine Optimum Time and Dose of Ad-TMBIM6 Administration ...... 68

2.4: Histological Staining Showed Co-Localization of BI-1 on Neurons but not on Astrocytes ...... 70

2.5: Ad-TMBIM6 Improved Long-Term Neurological Deficits ...... 72

2.6: The Effects of BI-1 siRNA and IRE1α CRISPR Activation Plasmid on Infarct Area at 72h post HI ...... 74

2.7: BI-1 Exerted its Neuroprotective Effects via the IRE1α/XBP1/CHOP Pathway at 72h post HI...... 76

2.8: Overexpression of BI-1 Attenuated Apoptosis at 72h post HI ...... 79

3.1: Expression Levels of Endogenous Proteins ...... 109

3.2: Inhibition of BI-1 or Nrf-2 Increased Percent Infarcted Area at 72h post HI ...... 111

3.3: The Effects of BI-1 Inhibition on the NPR-CYP Complex at 72h post HI ...... 113

3.4: BI-1’s Role in the Nrf-2 Signaling Pathway at 72h post HI...... 115

3.5: Overexpression of BI-1 Attenuated ROS Production and Inflammation at 72h post HI ...... 118

3.6: Fluorescent Staining Demonstrated BI-1’s Ability to Reduce ROS Production and Attenuate Inflammation at 72h post HI...... 119

3.7: BI-1 Exerted its Anti-Inflammatory Effects via Inhibition of P4502E1 and Upregulation of Nrf-2 in an in vitro OGD Model...... 121

ABBREVIATIONS

Ad-TMBIM6 Adenoviral-Transmembrane Bax inhibitor Motif Containing 6

ATF4 Activating transcription factor-4

ATF6 Activating transcription factor-6

Bcl-2 B-cell lymphoma 2

BI-1 Bax Inhibitor-1

CBF Cerebral Blood Flow

CC3 Cleaved Caspase 3

CHOP CCAAT/enhancer-binding homologous protein

CNS Central Nervous System

CRISPR AP CRISPR Activation Plasmid

CSF Cerebrospinal Fluid

CYP Cytochrome P450

DHE Dihydroethidium eIf2α Eukaryotic Translation Initiation Factor 2 subunit α

ER Endoplasmic Reticulum

FJC Flouro Jade-C

HI Hypoxia Ischemia

HIE Hypoxic Ischemic Encephalopathy

HO-1 Heme Oxygenase-1

ICV Intracerebroventricular

IHC Immunohistochemistry

IL-1β Interleukin-1beta

IL-6 Interleukin-6

IL-8 Interleukin-8

ILs Interleukins

IRE Inositol-Requiring Enzyme 1

JNK C-Jun N-terminal Kinase

Lfg Lifeguard

MCAO Middle Cerebral Artery Occlusion

MMO Microsomal Monooxygenase

MOI Multiplicity of Infection

MPO Myeloperoxidase

NF-kβ Nuclear Factor-kappaβ

Nox4 NADPH oxidase

NPR NADPH-P450 Reductase

Nrf-2 Nuclear Factor Erythroid 2-Related Factor 2

OGD Oxygen Glucose Deprivation

PDI Protein Disulfide Isomerases

PERK RNA-Dependent Protein Kinase-like ER Kinase

ROMO1 Reactive Oxygen Species Modulator 1

ROS Reactive Oxygen Species

SAH Subarachnoid Hemorrhage

TBI Traumatic Brain Injury

TEGT Testis Enhanced Gene Transcript

TNFα Tumor Necrosis Factor alpha

TNFR Tumor Necrosis Factor Receptor

TRAF2 TNF receptor-associated factor 2

UPR Unfolded Protein Response

WB Western Blot

XBP1 X-Box-Binding Protein 1

xii

ABSTRACT OF THE DISSERTATION

Gene Delivery of Adenoviral-TMBIM6 Vector Protects the Neonatal Brain after Hypoxic-Ischemic Injury

Desislava Doycheva Doctor of Philosophy, Graduate Program in Physiology Loma Linda University, June 2018 Dr. John H. Zhang, Chairperson

Neonatal hypoxia ischemia (HI) is an injury caused to the immature brain due to reduced cerebral blood flow which is associated with life-long neurological impairments.

HI causes oxidative protein folding in the endoplasmic reticulum (ER), which results in

ER stress. Generation of reactive oxygen species (ROS), from cytochrome P450 members (CYP) and NADPH-P450 reductases (NPR), in combination with activation of the unfolded protein response (UPR) are two major consequences of ER stress that cause oxidative damage and cell death.

Herein we identified the role of Bax Inhibitor-1 (BI-1), an evolutionary conserved protein encoded by the Transmembrane Bax inhibitor Motif Containing 6 (TMBIM6) gene, in protection from ER stress after HI injury in the neonatal rat. As BI-1 has multimodal properties that can target a wide array of pathophysiological consequences after injury, our main objective was to evaluate BI-1’s protective mechanisms by overexpressing it, using viral-mediated gene delivery of human adenoviral-TMBIM6

(Ad-TMBIM6) vector, in an in vitro and in vivo model of HI.

Following the Rice-Vannuci model, ten-day old (P10) rat pups underwent right- unilateral carotid artery ligation followed by 1.5h of hypoxia. Our results showed that overexpression of BI-1 ameliorated short and long-term deficits following HI. There was

xiii

a significant reduction in percent infarcted area which was linked with attenuation of apoptosis, reduction of ROS accumulation and inhibition of pro-inflammatory mediators.

BI-1 mediated protection was observed to be via both inhibition of IRE1α signaling, a stress sensor protein part of the UPR response, and P4502E1 activity, a major contributor of ROS generation. Attenuation of apoptosis was associated with a decrease in IRE1α-

XBP1 levels while decreased levels of P4502E1 coupled with upregulation of Nrf-2 and

HO-1, anti-oxidant enzymes, were correlated with a reduction in ROS generation and inflammation.

In conclusion we established that overexpression of BI-1 attenuated the morphological and neurological consequences post HI via inhibition of ER stress induced pathways. This new finding may help to provide a basis for BI-1 as a potential therapeutic target.

xiv

CHAPTER ONE

INTRODUCTION

Neonatal Hypoxia Ischemia

Neonatal Hypoxic-ischemic (HI) brain injury, caused by asphyxia, remains the leading cause of neonatal morbidity and mortality (Volpe 1976). The predominant pathogenic mechanisms of asphyxia are, hypoxemia, which is reduced amount of oxygen in the blood supply, or reduced cerebral blood flow (ischemia), which is a diminished amount of blood perfusing the brain. The term asphyxia is clinically used to describe an impairment in placental or pulmonary gas exchange before, during or soon after birth and is diagnosed as hypoxic ischemic encephalopathy (HIE) (Volpe 2001).

Impaired gas exchange may be from the fetus, the mother or uteroplacental disruption. Fetal causes may result from umbilical cord complications, fetomaternal hemorrhage and cardiac arrhythmia. Maternal causes may result from cardiac arrest, hypovolemic shock, asphyxiation, status epilepticus or severe anaphylactoid reaction.

Uteroplacental causes are due to placental abruption or infection, cod prolapse, uterine rupture or hyperstimulation (Rootwelt et al. , 1995). Generally, the newborn can compensate for brief periods of depleted oxygen, however, if the asphyxia lasts too long brain tissue is destroyed. These adverse events that occur to the newborn immature brain often lead to severe impairments such as pulmonary immaturity, respiratory distress syndrome, hypercapnia, hypo perfusion, mental retardation, epilepsy, cerebral palsy, seizures and long-term cognitive and behavioral deficits.

HIE is increasingly becoming burdensome not only for the patients and their families but also for the US healthcare with lifelong costs projected for cerebral palsy

alone at $11.5 billion (Wirrell et al. , 2001; Bracewell et al. , 2002; Ferriero 2004; Bartha et al. , 2007; Derrick et al. , 2007). Therefore, it is imperative to investigate and understand HIE pathophysiology and signaling pathways to help us establish successful treatment strategies for the management of these patients.

Epidemiology

HIE is a major burden of disease to not only patients but also their families with more than a million children who survive birth asphyxia develop long-term complications such as cerebral palsy, mental retardation and cognitive and behavioral deficits (Ferriero

2004; Zanelli et al. , 2011). The prevalence of HIE in developing countries is reported to be 10 times higher compared to technologically advanced countries. In the United states alone, the incidence is 1-4 cases per 1000 births. According to the World Health

Organization, 23% of neonates worldwide have birth asphyxia (920,000 neonatal deaths) while causing another 1.1 million intrapartum stillbirths (Volpe 2001; Cerio et al. , 2013).

From the patients that experience moderate to severe HIE, mortality rates are reported at

23-50% prior to discharge, and between 37-38% at a 18-22month follow up (Gluckman et al. , 2005; Shankaran et al. , 2005). Moreover, 80% of the cases that experience severe encephalopathy develop serious long-term complications and approximately 30-50% of those that survive with moderate to severe injury will develop long-term complications

(Vannucci et al. , 1999; Zanelli et al. , 2011). Some cases in newborn infants that do not present neurological deficits at the initial period, may later develop long-term functional impairments. In a clinical study of children 7+ years old that had a history of moderately

severe HIE, 15-20% of them developed learning difficulties, despite no obvious signs of brain injury (Robertson et al. , 2006; van Handel et al. , 2007).

Pathophysiology

Injury to the neonatal brain as a result of reduced cerebral blood flow and oxygen supply are the primary biological processes that lead to hypoxic-ischemic encephalopathy

(HIE) (Grow et al. , 2002; Ferriero 2004; Perlman 2004). The amount of damage that the newborn sustains depends on the duration of insult, the age and apoptosis rate (Zanelli et al. , 2011).

In the initial stages of depleted oxygen and blood supply to the brain, the neonate compensates for it by increasing cerebral blood flow (CBF) and redistributing the blood to the vital organs such as the brain, heart and adrenal glands. However, once this initial compensatory mechanism fails, CBF falls, blood pressure falls, and brain injury occurs.

This leads to intracellular energy failure (Vannucci et al. , 2005; Hagberg et al. , 2016).

In the primary energy failure stage, which occurs minutes after the insult, the sodium potassium ATP pump will begin to fail which will result in accumulation of sodium and calcium ions. This will induce neuronal depolarization and release of excitatory amino acids, such as glutamate, that can trigger further pathological cascade activation. Accumulation of calcium will recruit several enzymes, phospholipases, endonucleases and proteases, which results in cytoskeletal and DNA damage. At this stage apoptotic and necrotic cascades are activated which will cause cells to die (Hagberg et al. , 2016).

Just before secondary energy failure occurs, there is an increase in free radical

production and calcium overload which causes oxidative stress, excitotoxicity, inflammation and apoptosis. The neonatal brain is highly susceptible to oxidative stress due to the low concentrations of antioxidants present and high consumption of oxygen.

Cells will then begin to release pro-inflammatory cytokines which initiates an inflammatory response (Ferriero 2004). The second energy failure occurs 24-48h after injury and is characterized by mitochondrial dysfunction which leads to the ‘delayed phase of neuronal injury’ (Fatemi et al. , 2009; Lai et al. , 2011; Hagberg et al. , 2016), which is correlated with adverse neurodevelopmental outcomes.

This window between primary and secondary energy failure is critical, as it is the optimal time for intervention strategies to ameliorate cell death. As apoptosis and necrosis progress at a high rate, it is crucial to rescue cells within hours of the insult otherwise the damage will become irreversible.

Apoptotic Mechanisms

Programmed cell death was first introduced in 1964 by Lockshin and colleagues.

They suggested that the cell death occurring during development may not be a spontaneous or accidental event but instead, an intricate sequence of controlled processes leading to self-destruction (Lockshin et al. , 1965). The term apoptosis was then first brought in 1972 by Kerr and colleagues, where they used to describe the morphological processes leading to self-destruction (Kerr et al. , 1972). Apoptosis is an important process during development and is involved in differentiation, proliferation/homoeostasis, regulation and function of the immune system as well as removal of harmful cells (Fadel et al. , 1999).

The morphological features of an apoptotic cell are shrinkage of the cell, chromatin condensation, blebbing and finally fragmentation into small structures called,

‘apoptotic bodies’. Necrosis, on the other hand, results in the cell bursting and releasing its contents into the cells’ environment thus inducing a strong inflammatory response

(Leist et al. , 2001). Apoptosis is more prominent in less severe insults and is a delayed event associated with activation of caspase cascades, thus being a potential therapeutic target, whereas necrosis is more prevalent in severe HIE (Bonfoco et al. , 1995). In 2009,

Fatemi and colleagues (Fatemi et al. , 2009) showed that activation of caspase 3 in a neonatal rodent model of HIE is prolonged and that it can persist for at least 7 days after the insult.

Apoptotic pathways are intricate signaling mechanisms that have a variety of players. A cell can undergo apoptosis via two major pathways: the extrinsic or the intrinsic pathway. The extrinsic pathway, mainly mediated by death receptors, belongs to the tumor necrosis factor receptor (TNFR) gene family. These receptors are located on the cell surface and are triggered by their respective ligands to initiate apoptotic signaling. However, a signal coming from an activated receptor may not generate a strong enough signal to activate caspase cascades which are responsible for cell death.

Therefore, this signal may be amplified via the mitochondrial-dependent apoptotic pathways by stimulating the pro-apoptotic Bcl-2 family members, Bax and Bak, which eventually result in caspase 3 activation leading to cell death (Fatemi et al. , 2009).

The intrinsic signaling pathway mainly acts on the inside of the cell by disruption of mitochondrial membrane potential causing the release of pro-apoptotic mediators into the cytoplasm and is triggered by DNA damage and oxidative stress (Bernardi et al. ,

1999; Loeffler et al. , 2000; Wang et al. , 2001) which triggers the caspase cascade. In addition to the release of mitochondrial factors, cell homeostasis is disrupted which results in activation of redox molecules such as NADH, NADPH and accumulation of reactive oxygen species (ROS) triggering inflammatory responses. Although necrosis may prevail in severe HIE cases, apoptosis is a good target for therapeutics due to its delayed response and the potential target window that it presents.

Many studies of animal models of HIE have shown that during the period after the injury many neurons die via activation of the pro-apoptotic protein, caspase-3. The expression of caspase 3 was shown to be very prolonged, which indicates a prolonged role of apoptosis after injury (Nakajima et al. , 2000). During this time, neurons can either die or survive over the next few days to weeks which gives ideal time for possible intervention strategies to preserve these neurons. Clinical studies that analyzed the cingulate gyrus of infants who either died from severe birth asphyxia with HIE or from sudden intrauterine death showed a significant number of apoptotic cells (Edwards et al. ,

1999). Caspases are proenzymes containing 3 subunits that cleave proteins in a specific way that represent the morphological features of an apoptotic cell (Banasiak et al. ,

2000). Caspase 3 is a widely studied caspase that plays a major role in neuronal death

(Banasiak et al. , 2000). In newborn piglets subjected to HIE, apoptotic cell death was detected morphologically and via DNA fragmentation 48h after the insult (Mehmet et al,

1994). In a rat model, apoptotic cell injury began between 6-24h after HI (Chen et al. ,

2003). In addition, brains from postmortem infants that suffered HIE were examined and discovered high levels of caspase 3 (Rossiter et al. , 2002). Up to date, intervening at the

caspase level thus inhibiting caspase 3 has been the most common intervention following ischemia.

Inflammatory Mechanisms

Hypoxia ischemia triggers an inflammatory reaction that progresses for days to weeks and plays a role in secondary brain injury. Inflammation is complex process associated with the production and release of hundreds of signaling molecules such as cytokines, chemokines and ROS which are highly induced after ischemia in the neonatal brain (Szaflarski et al. , 1995; Hagberg et al. , 1996; Ivacko et al. , 1997; Barone et al. ,

1999; Cowell et al. , 2002). Cytokines have been implicated as important players in the pathogenesis of HIE and may represent a final common pathway of brain injury. This has been shown in animal studies, where levels of IL-1β were significantly upregulated after

HI injury (Aly et al. , 2006; Liu et al. , 2013; Li et al. , 2014). However, details regarding the mechanisms or involvement of other cytokines in this process remains unclear.

Additionally, there are several other factors that are induced after ischemia such as adhesion molecules (Frijns et al. , 2002), nitric oxide synthase (Forster et al. , 1999) and cyclooxygenase-1 (Nogawa et al. , 1997).

Interleukins (ILs) such as TNFα, IL-6, IL-1β, IL-8 and tumor necrosis factors released after an HI insult are major pro-inflammatory cytokines that play a role in inflammatory processes and contribute to HI damage (Grow et al. , 2002). The inflammatory genes are evident as early as 8h after insult and further increased from 24 to 72h post HI. Several studies have shown a correlation between injury and elevated levels of cytokine/inflammatory response. Patients with HIE have been shown to have

elevated levels of inflammatory cytokines and are associated with later development of cerebral palsy (Dammann et al. , 2008). Elevated levels of IL-6 and IL-8 have been observed in the CSF of newborns which were correlated with exacerbated HIE and poor neurodevelopmental outcome (Savman et al. , 1998). It has also been reported that HIE patients have a significantly higher expression of cytokine concentrations (Shalak et al. ,

2002). A cohort study of full term infants that experienced perinatal asphyxia revealed increased levels of IL-1β, IL-6 and TNFα when compared to normal infants (Foster-

Barber et al. , 2001). Those levels were observed in infants that either died at 1 year or went on to develop cerebral palsy. Another clinical study, using blood collected from the umbilical cord of infants that experienced asphyxia, revealed a 375-fold higher concentration of IL-6 in those patients that developed HIE. Thus, indicating that the immature brain is highly susceptible to the consequences of an inflammatory response

(Anthony et al. , 1997; Bona et al. , 1999) making inflammation a prominent player in the pathogenesis of neonatal brain injury.

In mice animal studies, that underwent unilateral carotid artery ligation, it was shown a significant increase in inflammatory genes in the injured hemisphere after the insult (Bona et al. , 1999). IL-1β and TNFα are early response cytokines that are synthesized and released by microglia, astrocytes and neurons. Some of their features include the release of other cytokines, induction of leukocytes, influence of glial expression and stimulation of trophic factors (Szaflarski et al. , 1995). A study showed that when recombinant IL-1β receptor antagonist was administered it attenuated brain damage caused by HI in rats (Martin et al. , 1994; Hagberg et al. , 1996) which is an indicator that IL-1β plays a role in brain damage after HI. In 1995, Szaflarski and

colleagues demonstrated for the first time that ischemic insult stimulates cytokine production in the developing brain (Szaflarski et al. , 1995). In an HI rat model, IL-1β mRNA levels were significantly increased at 4h after HI. Furthermore, IL-1β triggers production of IL-6 which then acts together in response to HI. Another pro-inflammatory inhibitor, simvastatin was used as pretreatment for HIE and showed a reduction in expression of IL-1β and TNFα (Balduini et al. , 2004). Microglial activation, due to injury caused to the brain, secretes pro-inflammatory cytokines, extracellular proteases, and oxidative species, which damage surrounding tissue. This, in turn, stimulates a further release of pro-inflammatory mediators thus exacerbating the damage after HIE

(Ivacko et al. , 1997; Yang et al. , 2015). Hence, targeting inflammatory response is critical to ameliorate HI induced injury.

Several studies have achieved neuroprotection against ischemia by intervening with inflammatory mediator’s function. For example, inhibition of either IL-1, TNFα or adhesion molecules conferred neuroprotection after ischemia (Relton et al. , 1992;

Loddick et al. , 1996; Connolly et al. , 1997). Interleukins seem to play a major role in

HIE progression as mice deficient in IL-18 or inhibition of IL-1 decreased CNS vulnerability to HI (Martin et al. , 1994; Hagberg et al. , 1996). Moreover, administration of the anti-inflammatory cytokine, IL-10 reduced HI insult (Mesples et al. , 2003).

Oxidative Stress

Before birth, the fetus is in a low oxygen environment however, in the first few minutes after birth, there is an abrupt increase in O2 levels (80-90%) creating pro-oxidant conditions to facilitate metabolic processes. In the event of pathological conditions, such

as after hypoxic ischemic (HI) insult, a serious of deleterious events occur such as excess calcium influx leading to oxidative stress (Allen et al. , 2011; Cao et al. , 2014; Zeeshan et al. , 2016). In animal models, it was reported to be a significant accumulation of H2O2 in neonatal mice after HI but not in adult (Lafemina et al. , 2006). Decrease in oxygen in the newborn triggers a serious of events and the release of oxidases, NOS while simultaneously reducing anti-oxidant enzymes triggers a massive upregulation of ROS. A major target of ROS is the mitochondria, and the immature brain is especially susceptible to that because of its poor scavenging system. When the ROS exceeds the mitochondrial capacity, the resulting oxidative stress ruptures the mitochondrial membrane allowing the release of proteins with the potential to trigger inflammatory cascades that lead to apoptosis (Blomgren et al. , 2006; Nakai 2007).

Clinical Presentation and Therapeutic Strategies

Certain signs may appear shortly after birth such as organ dysfunction mainly the heart, lungs, kidneys and liver which is a good indicator of possible HIE. In severe cases,

60% of the patients will develop seizures in the first 24h (Vasquez-Vivar et al. , 2017).

Other evident symptoms include arched back, breathlessness, lethargy and some nausea and vomiting. Other symptoms that may be presented by the infant are difficulties suckling, eating, swallowing and excessive drooling. In other cases, symptoms might not be evident at the time of birth but might become apparent later in life at the toddler ages.

A child with birth injury may present symptoms later in life such as delay in motor skills, lack of muscle control, speech problems, learning difficulties and general loss of motor and learning abilities (Millar et al. , 2017).

Current treatment modalities for HIE include anticonvulsants, hypothermic treatments and fluid and electrolyte management, but despite the success of some of these, about 30-

50% of the patients will still develop serious long-term complications (Zanelli et al. ,

2011). The serious lack of understanding HI pathology, timing of therapeutic window and lack of mechanistic studies has made it crucial to investigate novel therapeutic approaches. It is imperative to understand the physiological cascades triggered by HI.

Hypoxic Ischemic Injury and Endoplasmic Reticulum Stress

The endoplasmic reticulum (ER), located in the cytoplasm, serves multiple functions in maintaining cell homeostasis. One of its major functions is protein synthesis, correct folding, modification and transportation of proteins. The ER lumen is an oxidative environment crucial for proper folding of proteins and formation of disulfide bonds. As its main role in protein folding and transport, it also contains chaperones responsible for the control of these proteins. Many disturbances and stresses, such as ischemic events, oxidative stress, calcium disturbances and inhibition of protein glycosylation, can alter

ER’s functions causing an accumulation of unfolded proteins. When the amount of unfolded proteins exceeds the number of chaperones needed to refold these proteins it causes ER stress (Xu et al. , 2005; Rissanen et al. , 2006). For cells to cope with this stress they activate an evolutionary conserved response known as the unfolded protein response (UPR). The initial intent of the UPR is to restore correct protein folding and function and re-establish ER homeostasis. However, under prolonged stress, when these adaptations fail, the ER initiates pro-apoptotic pathways thus triggering cell suicide

(Oyadomari et al. , 2004). The three major ER stress triggered apoptotic proteins are the

inositol-requiring enzyme 1 (IRE1), the PKR-like ER kinase (PERK) and the activating transcription factor-6 (ATF-6) (Korennykh et al. , 2009; Korennykh et al. , 2012).

The IRE1 contains both serine/threonine kinase domains and an endoribonuclease domain. In mammals, IRE has been characterized to have two genes, IRE1α and IRE1β.

IRE1α is ubiquitously expressed in all tissues, whereas the IRE1β is only expressed selectively in intestinal and respiratory epithelial cells (Martino et al. , 2013). The IRE1α plays a critical role in splicing the X-box-binding protein 1 (XBP1) and has been demonstrated in deletion studies to play a crucial role in murine embryonic development, differentiation, function, and survival of many cell types which secrete large amounts of protein (Wu et al. , 2015). Accumulation of unfolded proteins stimulates activation of

IRE1α and XBP1 splicing through its ribonuclease domain, which then translocates to the nucleus and acts as a potent transcription factor of pro-apoptotic mediators (Hetz et al. ,

2011). In addition, it can also stimulate activation of Apoptotic signaling kinase-1

(ASK1) which causes downstream activation of stress kinases that promote apoptosis

(Ron et al. , 2008). Moreover, IRE1α also triggers pro-inflammatory responses in the ER through direct binding with adaptor protein tumor necrosis factor alpha (TNFα) receptor- associated factor 2 (TRAF2) and subsequent activation of the nuclear factor-kappaβ (NF-

κβ) and c-Jun N-terminal kinase (JNK) pathways (Xu et al. , 2005; Cao et al. , 2014).

The second effector protein, PERK, once activated induces its autophosphorylation and activates the kinase domain responsible for phosphorylating and inactivating the eukaryotic translation initiation factor 2 subunit α (eIf2α) (Donnelly et al.

, 2013), thus shutting down the mRNA translation. PERK activation and phosphorylation of eIF2a is linked with an increased translation of the mRNA encoding the initiation of

activating transcription factor 4 (ATF4). The ATF4 is a transcription factor which induces several genes involved in acid metabolism, antioxidant stress and protein secretion (Rouschop et al. , 2010; Rzymski et al. , 2010). ATF4 can also induce the pro- apoptotic effector CCAAT/enhancer-binding homologous protein (CHOP), that activates apoptosis (Oyadomari et al. , 2004; Jager et al. , 2012).

Finally, activation of ATF6 triggers a slightly different mechanism compared with

IRE1α and PERK. Accumulation of unfolded proteins activates ATF6 which causes it to translocate from the ER membrane to the Golgi apparatus where proteases cleave it, thus releasing it into the cytosol where it begins to regulate gene expressions causing an increase in XBP1 levels (Yamamoto et al. , 2007).

Among the three UPR proteins, IRE1α is the key molecule responsible for cell fate as there is evidence that it is the only protein that can directly bind to the misfolded proteins and trigger a signaling response. Numerous studies show evidence that supports this concept, including structural studies, protein interaction studies and peptide interaction studies (Gardner et al. , 2011).

Apoptosis

The UPR is an important adaptive mechanism to protein misfolding, however, prolonged ER stress leads to the activation of pro-apoptotic UPR signaling which plays a critical role in pathological conditions. One of the major contributing proteins to apoptosis that is activated as a result of ER stress is IREα. The IRE1α-XBP1 pathway is required for normal development and function, however, increasing evidence suggests that IRE1α may contribute to apoptotic cell death (Cao et al. , 2014). It can do so via

several different ways. First, it can activate the JNK pathway through direct interaction with TRAF which inhibits Bcl-2 and hence inhibits its anti-apoptotic effects (Cao et al. ,

2014). Second, IRE1α may bind Bax and Bak on the ER membrane and initiate a mitochondrial dependent cascade (Cao et al. , 2014). IRE1α has also been linked to activation of two pro-apoptotic molecules, PUMA and BH3 (Cao et al. , 2014). More recently, IRE1α has also been shown to degrade specific microRNAs that target caspase

2 and hence boosting caspase dependent apoptosis (Cao et al. , 2014).

Oxidative Stress

There are several ways the ER may cause reactive oxygen species (ROS) production. Firstly, in the ER lumen the correct folding of proteins requires formation of disulfide bonds to be able to stabilize tertiary and quaternary structures. Therefore, the lumen has a highly oxidized environment which is needed for bond formation (Cao et al.

, 2014). Evidence suggests that oxidative protein folding is an important source of ROS production (Cao et al. , 2014). ROS can be generated as a byproduct of protein folding, mitochondrial respiration and detoxification.

Protein folding is catalyzed by oxidoreductases such as protein disulfide isomerases (PDI). During bond formation, PDI transfers electrons to an acceptor to begin another cycle of bond formation. The acceptor then transfers these electrons to molecular oxygen and produces H2O2, which is major ROS produced in the ER lumen. Another inducer of ROS production is NADPH oxidase 4 (Nox4). Nox 4 can physically interact with PDI causing an increase in ROS generation. Studies have shown that under stress

Nox4 levels are increased and at the same time ROS generation is also increased

(Zeeshan et al. , 2016).

It has been reported that ROS production results mainly from three ER-electron coupling systems such as PDI and ERO1α, intra ER-GSSG/GSH and the NADPH-450

Reductase – Cytochrome P450 (NPR-CYP) complex. However, the majority ROS produced in the ER is from the microsomal monooxygenase (MMO) system which is composed of the NPR-CYP complex. It is responsible for the production of superoxide anion radicals and H2O2. The leakage of electrons during electron transfer between NPR to CYP plays an important role in ROS generation. Increased P4502E1, a member of the

CYP family, protein activity and expression has been linked with ROS generation and reported in several animal models (Kim et al. , 2009; Henke et al. , 2011). Inhibitors of

P4502E1 have shown protection against ethanol-induced hepatic injury (Gouillon et al. ,

2000). As a result, P4502E1 activation has led to increased levels of ROS. Further research is required to examine its mechanisms, but it is evident that liver injury and insulin resistance both involve P4502E1. Another potent suppressor of P4502E1 expression was shown to be Bax Inhibitor-1 (BI-1). Kim and colleagues showed that in the presence of BI-1, the levels of P4502E1 were significantly reduced thus ameliorating

ER stress induced ROS accumulation in in vitro experiments using Neo cells (Kim et al. ,

2009).

ROS may also be generated via release of calcium from the ER. This causes an increase in mitochondrial calcium which alters metabolism and eventually causes ROS production (Gorlach et al. , 2006). Increase in ROS overloads the mitochondria and stimulates it to work faster thus consuming more oxygen. An increase in oxygen

consumption will activate nitric oxide synthase which will then inhibit the activity of complex IV further increasing ROS production (Xu et al. , 2004; Murphy 2009; Meares et al. , 2011).

Inflammation

ER stress can contribute to a wide variety of inflammatory processes. In the CNS,

ER stress may induce inflammation which can produce interferon-gamma and apoptosis can occur as a direct result (Zhang et al. , 2008). ER stress is also observed in the lungs where it is associated with increase in CHOP and subsequent ROS accumulation and cell death (Endo et al. , 2006). During ER stress, there is also a release of calcium which activates inflammasomes through an inositol triphosphate (IP3)-mediated mechanism that releases the calcium stored in the ER, resulting in accumulation of ROS (Zeeshan et al. ,

2016). In addition, the accumulation of misfolded proteins has been linked with inflammatory processes in bowel disease and many autoimmune diseases (Garg et al. ,

2012). Hence, targeting inflammation with anti-inflammatory drugs has shown promise against ER stress induced injury (Yamazaki et al. , 2006).

The UPR may also initiate inflammatory cascades thus triggering pro- inflammatory protein signaling. These include NF-kβ and MAPKs. In studies, they were triggered from eIF2 phosphorylation from PERK signaling, as depressing mRNA translation caused decreased levels of NF-kβ in cytosol (Lin et al. , 2007; Lin et al. ,

2013; Lin et al. , 2014). Another UPR protein involved in inflammatory responses is

IRE1α. It can recruit the inflammatory marker, TRAF2, which can then couple IRE1α activation with various inflammatory pathways (Cao et al. , 2014).

In addition to inflammatory responses from UPR proteins, there are other signaling pathways that respond to ER stress and induce inflammation. ER stress and inflammation are connected via several pathways such as ROS, release of calcium, activation of NF-kβ and MAPK. It is well known that ROS are important mediators to inflammation (Raha et al. , 2000).

Correct protein folding in the ER is an energy consuming mechanism and highly oxidative process. During formation of disulfide bonds, to manufacture tertiary and quaternary protein structures, there is a large amount of electron flow and utilization of oxygen, which produces ROS (Tu et al. , 2002). Therefore, an increase in protein folding in the ER is corelated with ROS production and subsequent inflammation. Moreover, during ER stress, there is a release of calcium into the cell which causes depolarization of mitochondrial membrane, thus disrupting electron transport resulting in ROS. This cycle of calcium release, ROS production and accumulation of misfolded proteins work in conjunction to activate pathways that lead to inflammation (Malhotra et al. , 2007; Zhang et al. , 2008).

Endoplasmic Reticulum Stress Suppressors

The Transmembrane Bax Inhibitor-1 (TMBIM) Protein Super Family

The Bax Inhibitor-1 (BI-1) is an evolutionary conserved protein which is part of the TMBIM protein family involved in cytoprotection. Members of this protein family are hydrophobic in nature and hence reside on the ER membrane, Golgi or plasma membrane. The TMBIM family are composed of five highly conserved orthologues of

BI-1 which are referred to as TMBIM family of proteins, BI-1 family or Lifeguard

family: GRINA/TMBIM3 (glutamate receptor ionotropic NMDA protein 1), BI-

1/TMBIM6, TMBIM2/Lfg/FAIM2, GHITM/TMBIM5 (growth hormone-inducible transmembrane protein), RECS1/TMBIM1 (responsive to centrifugal force and shear stress gene 1 protein) and GAAP/TMBIM4 (Golgi antiapoptotic-associated protein)

(Zhou et al. , 2008; Rojas-Rivera et al. , 2015).

Among these proteins, it has been reported that BI-1/TMBIM6, GRINA/Lfg-1,

Lfg-2/FAIM2 and Lfg-4/GAAP play a role in apoptosis (Hu et al. , 2009), however,

TMBIM6/BI-1 is the best described member of the family and has several reports of its anti-apoptotic activities (Hu et al. , 2009; Rojas-Rivera et al. , 2015).

Bax Inhibitor-1 (BI-1/TMBIM6) Protein Family

The BI-1 is an evolutionary conserved protein first discovered in yeast and seen to be widely expressed in tissues such as the heart, lungs, liver, stomach, colon, kidney, brain, placenta, skeletal muscle and pancreas (Xu et al. , 1998; Grzmil et al. , 2006). It has been observed that BI-1’s levels vary in these organs during various stages of life.

For example, BI-1 was shown to peak during times of high apoptosis (Jean et al. , 1999;

Chae et al. , 2004), thus indicating a possible link between BI-1 and cell death.

BI-1 contains six alpha-helix transmembrane domains with both of its N and C terminus exposed to the cytosol. Its protective role has been mapped to its C terminus, as deletion studies of the C terminus abrogated BI-1’s ability to suppress Bax induced cell death in yeast (Kawai-Yamada et al. , 2004).

It was named as Bax Inhibitor-1 due to its functional effect to suppress cell death via inhibition of Bax gene but not because of a direct physical interaction with Bax (Xu et

al. , 1998). There are several studies that show evidence of BI-1’s ability to regulate the intrinsic pathways but not the Fas and TNFα induced extrinsic pathways. It was reported that overexpression of BI-1 in interleukin-3 dependent pro-B lymphocytes attenuated apoptosis induced from intrinsic pathway inhibitors (Xu et al. , 1998). Transfection of BI-

1’s antisense induced apoptosis in human embryonic kidney cells regardless of the stressors (Xu et al. , 1998). Additionally, it was reported that BI-1’s effects are caspase independent, as administering tunicamycin to BI-1 overexpressing cells in the presence of a caspase inhibitor did not cause translocation of Bax to mitochondria (Chae et al. ,

2004). However, when BI-1 overexpressing cells were treated with staurosporine with the caspase inhibitor, it showed translocation of Bax to mitochondria (Chae et al. , 2004).

This indicates that BI-1’s anti-apoptotic activity is more sensitive to ER stress inducing agents (Robinson et al. , 2011).

Bax Inhibitor-1 Like Proteins

TMBIM1 or Lifeguard 3. The TMBIM1 is also known as Lifeguard-3 (Lfg-3) or

RECS (Responsive to centrifugal force and shear). It is a seven helix transmembrane protein localized to endosomal and lysosomal membranes and expressed in almost all tissues (Hu et al. , 2009). It was shown to play a role in vascular remodeling as there were no abnormalities in Lfg-3 knockout mice (Zhao et al. , 2006), however in aged animals it caused aortic dilatation (Zhao et al. , 2006).

TMBIM2 or Lifeguard 2. TMBIM2 is also predicted to contain seven helix transmembrane domain and is located in Golgi and ER. It is predominantly expressed in nervous system (Schweitzer et al. , 1998), and specifically, in dendritic processes and

synapses. In contrast to BI-1, TMBIM2 protects from cell death induced by the extrinsic pathway, Fas, but not TNFα (Somia et al. , 1999; Fernandez et al. , 2007). It was reported that inhibition of TMBIM6 with shRNA increased sensitivity of neural cells to Fas induced cell death (Fernandez et al. , 2007).

TMBIM3 or Lifeguard 1. It was recently demonstrated that, like BI-1, TMBIM3 resides on the ER membrane and co-immunoprecipitation experiments revealed its physical interaction with BI-1 (Rojas-Rivera et al. , 2015). Studies showed that knocking down TMBIM-3 in BI-1 wild type cells did not cause spontaneous cell death, but in contrast, knocking down TMBIM-3 in BI-1 knockout cells led to appearance of cell death features (Rojas-Rivera et al. , 2012). ER stress may trigger upregulation of

TMBIM3/Lfg1 through PERK dependent mechanism (Hetz et al. , 2011). In conclusion,

TMBIM3 and TMBIM6 may have synergistic function for anti-apoptotic activity however there is very little information regarding this protein and its signaling mechanism is unclear.

TMBIM4 or Lifeguard 4. TMBIM4 is identified as a Golgi anti-apoptotic protein and housekeeping gene (Lee et al. , 2007). It is widely expressed in tissues and has shown close resemblance to BI-1. TMBIM4 can inhibit apoptosis triggered by intrinsic and extrinsic stimuli (Gubser et al. , 2007) and can reduce calcium stores (de

Mattia et al. , 2009). It was shown that inhibition of Lfg-4 caused apoptosis and was seen to also be upregulated in human breast cancer cells (van 't Veer et al. , 2002).

Lifeguard 5. Found predominantly in eutherian mammals and not in invertebrates, Lifeguard 5 is predicted to have eight transmembrane helices (Hu et al. ,

2009). It is reported as TMBIM1 because of its great resemblance. In mouse models, it is expressed only in the adult testis (Hu et al. , 2009).

Bcl-2/Bcl-Xl

These proteins are also localized to the ER membrane and are protective against

ER stress. The main mechanism via which Bcl-2 works is believed to be through a decrease in ER levels of calcium via IP3Rs.

MicroRNAs

New emerging data has shown that microRNAs may either modulate ER stress or be activated by ER stress (Sano et al. , 2013). The PERK mediated activation of miR30c is stimulated upon ER stress which is correlated with downregulation of XBP1 (Byrd et al. , 2012). miR211 has also been identified as a PERK target that promotes cell survival.

It was reported that miR221 is associated with CHOP mediated apoptosis during ER stress (Behrman et al. , 2011). As such, ER stress apoptosis was reported to be enhanced by miR221mimics and attenuated by miR221 inhibitors. Additionally, it was illustrated that CHOP regulates miR708 (Upton et al. , 2012) and spliced XBP1 was necessary for stress associated miR346 induction which regulate immune responses.

Bax Inhibitor-1 and Endoplasmic Reticulum Stress

BI-1 as a Regulator of Apoptosis

As the ER is responsible for correct protein folding and maturation of proteins as well as a regulation of calcium, any unfavorable environment may disrupt the ER

function inducing ER stress and the UPR. The main function of the UPR is to allow cells to adapt to different stresses. In the event that this is not resolved, and the stress is prolonged it will result in calcium release, promoting apoptosis and pro-apoptotic ER signaling.

Reports show evidence of BI-1’s protective effects in cell culture experiments where, cells overexpressing BI-1 that were challenged with ER stress inducers, or during

OGD and ischemic reperfusion, did not show any characteristics of apoptotic changes

(Chae et al. , 2004). There was no caspase activation, no blebbing, no formation of apoptotic bodies. Inversely, cells lacking BI-1 exhibited sensitivity to the same ER stresses (Chae et al. , 2004). Thus BI-1 may aid to inhibit ER stress signaling pathways

(Chae et al. , 2004; Bailly-Maitre et al. , 2006).

BI-1 was shown to interfere with all three ER signaling pathways, in particular, it was shown to be able to physically bind to IRE1α thereby blocking its RNase activity

(Lisbona et al. , 2009). BI-1 overexpressing cells showed a decrease in levels of IRE1α as well as its downstream factor XBP1 (Lee et al. , 2007). BI-1 was reported to suppress

IRE1α activity by forming a complex in both fly and mouse models of ER stress (Sano et al. , 2013). Knocking out TMBIM6 gene, responsible for BI-1 production, caused hyperactivity of the IRE1α pathway, while cells overexpressing BI-1 showed to inhibit

IRE1α kinase and endoribonuclease activity. In addition, deletion of the C terminus of

BI-1 failed to bind to IRE1α and hence did not suppress IRE1α activity (Sano et al. ,

2013). Additionally, BI-1 deficiency also increased XBP1, a major downstream effector of IRE1α. In mice knockout studies, it was shown that mice were viable and developed normally despite a deficiency in TMBIM6/BI-1 (Sano et al. , 2013; Urra et al. , 2013).

However, they were more susceptible to tissue injury induced by stroke, ischemia- reperfusion and experimental ER stress in brain, liver and kidney (Sano et al. , 2013).

Although BI-1 was not shown to interact with PERK or ATF6, it still affected their signaling pathways (Lee et al. , 2007). Lee and colleagues reported that eIF2α levels decreased when BI-1 is overexpressed, indicating that it may block PERK pathway via unknown mechanisms (Lee et al. , 2007). In addition, BI-1 was also shown to decrease

ATF6 cleavage thus also blocking its activity (Lee et al. , 2007).

BI-1 as a Regulator of Oxidative Stress and Inflammation

During times of stress, such as after HI insult, ROS production and accumulation is significantly increased in the cell thus culminating in oxidative stress and the release of

ROS. This release of ROS surrounds tissue and causes damage, potentially promoting inflammatory processes (Duval et al. , 2003; Kopprasch et al. , 2003).

The main source of ROS production in the ER is from the microsomal monooxygenase (MMO) system composed of NADPH-P450 reductase (NPR) and the cytochrome P450 (CYP) members (Nieto et al. , 2002; Robinson et al. , 2011).

Specifically, the CYP member, P4502E1, is known to produce significant amount of

ROS such as H2O2 (Nieto et al. , 2002).

It is now known that BI-1 may regulate ER stress induced ROS production via two mechanisms. Firstly, it can influence Nuclear Factor Erythroid 2-Related Factor 2 (Nrf-2) upregulation, which is a redox sensitive transcription factor that controls expression of different anti-oxidant enzymes, such as heme oxygenase-1 (HO-1). Cells overexpressing

BI-1 were shown to increase Nrf-2 levels, and subsequently HO-1 levels, which blocked

ROS (Lee et al. , 2007). Secondly, BI-1 can directly interact with NPR though its C terminus, thus altering the electron flow between NPR-CYP and destabilizing this complex resulting in blockage of ROS (Kim et al. , 2009). Co-immunoprecipitation studies have revealed that BI-1 is in close proximity and has high affinity for NPR in vivo but to a lesser extend for P4502E1 (Kim et al. , 2009; Watanabe et al. , 2009). However, BI-1 can regulate

P4502E1 activity, by binding to NPR and disrupting the flow of electrons between NPR and CYP hence lowering P4502E1 activity and its ability to produce ROS (Kim et al. ,

2009; Yang et al. , 2010; Henke et al. , 2011). The C terminus of BI-1 is required for BI-1 to bind to NPR and hence to disrupt this complex. Adding an antibody raised against the

BI-1’s C-terminus or in a C terminus BI-1 mutant abrogated the inhibitory effect of BI-1

(Watanabe et al. , 2009; Li et al. , 2014). Other studies also showed that deletion of C terminus of BI-1 had less inhibition of P4502E1 induced ROS production and there was no NPR binding (Kim et al. , 2009; Watanabe et al. , 2009; Li et al. , 2014). Despite the clear functional connection between BI-1 and P4502E1, a direct interaction of the two proteins was not shown.

Specific Aims

Hypoxic-ischemic (HI) brain injury remains a leading cause of mortality and morbidity for infants, affecting between 2-4 from 1000 full-term births, and nearly 60% of premature infants (Vannucci et al. , 1997; Volpe 2001). Among survivors, 20-40% develop significant neurological impairments, associated with life-long medical, social, and emotional difficulties (Bracewell et al. , 2002; Ferriero 2004; van Handel et al. ,

2007; Zanelli et al. , 2011). Numerous attempts to reduce HI-induced consequences have

failed in the clinical setting. In order to improve future outcomes, it is imperative to explore different therapeutic targets to either amplify or replace current therapeutic strategies.

Endoplasmic reticulum stress (ER) is a major pathology encountered after HI, associated with dysregulation of protein folding with downstream effects on apoptosis and inflammation which have been implicated in a variety of disease processes (Kaser et al. , 2008; Li et al. , 2012; Chaudhari et al. , 2014; Li et al. , 2015; Yang et al. , 2015). HI induced ER stress up-regulates several stress sensors’ signaling pathways, primarily, the inositol requiring enzyme-1 alpha (IRE1α) signaling pathway (Madeo et al. , 2009; Sano et al. , 2013). ER stress is also associated with reactive oxygen species (ROS) accumulation, mainly from the NADPH-dependent cytochrome P450 reductase (NPR) and P4502E1 (CYP) complex. CYP has been linked as one of the main players in ROS generation (Kim et al. , 2009).

Bax-inhibitor 1 (BI-1) protein, mainly expressed on ER membrane, has been shown to play a major role in inhibiting ER stress related apoptosis (Xu et al. , 1998;

Watanabe et al. , 2009) and inflammation. BI-1 is encoded by the TMBIM6 gene, part of the Bcl-2 family of proteins that can directly bind to Bcl-2 and form a complex, resulting in inhibition of pro-apoptotic molecules (Xu et al. , 1998; Watanabe et al. , 2009; Iwata et al. , 2011; Robinson et al. , 2011). Evidence from previous in vitro studies supports BI-

1’s ability to directly bind to IRE1α and inhibit this pro-apoptotic signaling pathway induced after ER stress (Chae et al. , 2004; Lisbona et al. , 2009; Bailly-Maitre et al. ,

2010). Furthermore, it has been shown that BI-1 can reduce ROS accumulation by

disrupting the electron transport between NPR and CYP and by simultaneously activating

Nrf-2 (Lee et al. , 2007; Kim et al. , 2009; Henke et al. , 2011).

This proposal will be the first to show BI-1’s anti-apoptotic and anti- inflammatory effects in an animal stroke model, by overexpression of the BI-1 protein with intracerebroventricular (icv) injection of human adenoviral-TMBIM6 (Ad-

TMBIM6) vector and elucidate the pathways associated with BI-1’s protective effects.

The specific objective of this proposal is to establish BI-1’s anti-apoptotic and anti-inflammatory effects in an in vitro oxygen glucose deprivation (OGD) model, as well as in an in vivo neonatal hypoxia ischemia (HI) rat model, followed by subsequent clarification of the mechanisms via which it confers its protective properties. Our central hypothesis is that (1) transfection of PC-12 cells with Ad-TMBIM6 vector will improve cell viability after OGD (2) in vitro OGD model will help determine the major pathways involved in BI-1’s protective effects (3) overexpression of the BI-1 protein in the brain, via Ad-TMBIM6 injection will improve recovery after neonatal HI by reducing ER stress induced (a) neuronal apoptosis via inhibition of IRE1α signaling pathway and (b) oxidative stress induced inflammation via dissociation of the NPR-CYP complex and subsequent inhibition of ROS. This hypothesis will be tested with the following specific aims:

Aim 1

To determine whether BI-1 upregulation exerts its anti-apoptotic effects via the

IRE1α signaling pathway in an in vitro and in vivo neonatal HI rat model. This aim will examine the effects of BI-1 overexpression on IRE1α and PERK signaling, as well as

examine BI-1’s long-term effects on functional deficits after HI. ER stress, as a result of

HI injury, is reported to trigger three main signaling pathways: IRE1α, PERK and ATF6, which under pathological conditions lead to apoptosis. Studies have shown that overexpression of BI-1 in cells in response to ER stress inducers, or during OGD and I/R injury, can attenuate ER stress induced apoptotic phenotypes. BI-1 knockout cells showed the opposite effects, indicating that BI-1 aids to inhibit ER stress signaling pathways (Chae et al. , 2004). BI-1 can also physically bind to and inhibit IRE1α (Bailly-

Maitre et al. , 2010) and its downstream molecules such as XBP1 (Lisbona et al. , 2009;

Bailly-Maitre et al. , 2010). Several in vitro studies have been done in TBI, MCAO and hepatic ischemia/reperfusion models which present data in support of BI-1’s anti- apoptotic effects (Bailly-Maitre et al. , 2006; Krajewska et al. , 2011; Ma et al. , 2014). It is unclear whether BI-1 exerts its anti-apoptotic effect mainly via IRE1α inhibition or whether it also plays a role in inhibiting PERK. This aim will help determine whether

IRE1α is the major pathway via which BI-1 exerts its neuroprotective role and translate this in vitro finding to our in vivo studies.

Aim 2

To investigate the anti-oxidative effects of BI-1 overexpression and the signalling pathways involved in an in vitro and in vivo neonatal HI rat model. This aim will examine BI-1’s role in oxidative stress, focusing on BI-1’s ability to dissociate the NPR-

CYP complex and activate Nrf-2 transcription factor resulting in attenuation of ER stress induced ROS accumulation and subsequent inflammation.

The major source of ER stress induced ROS accumulation is from the microsomal monooxygenase (MMO) system which is composed of cytochrome P4502E1 (CYP),

NADPH-P450 reductase (NPR) and some phospholipids (Kim et al. , 2009). The leakage of the electron transfer between NPR and CYP results in the accumulation of ROS. Our specific hypothesis is that BI-1 will reduce ER stress induced ROS accumulation and subsequent inflammation via: (1) coupling with NPR and disrupting the NPR-CYP complex; (2) upregulating the anti-oxidant transcription factor, Nrf-2, which leads to increase in HO-1 production. HO-1 is a rate-limiting enzyme shown to be able to attenuate ROS accumulation. Overexpression of BI-1 should increase HO-1, which should subsequently decrease the UPR by limiting oxidative dysregulation of the ER.

Figure 1.1 Schematic Representation of the central hypothesis and specific aims Aim 1 will investigate the effects of overexpression of BI-1on the stress sensor protein, IRE1α in HI model. The signaling pathway molecules will be measured as well as the apoptotic markers in the brain and assess infarct area. Aim 2 will study the effects of overexpression of BI-1 on the NPR-CYP complex, responsible for production of ROS, and also measure the levels of anti-oxidant enzymes and inflammatory mediators.

References

Allen, K. A. and D. H. Brandon (2011). "Hypoxic Ischemic Encephalopathy:

Pathophysiology and Experimental Treatments." Newborn Infant Nurs Rev 11(3):

125-133.

Aly, H., M. T. Khashaba, M. El-Ayouty, O. El-Sayed and B. M. Hasanein (2006). "IL-

1beta, IL-6 and TNF-alpha and outcomes of neonatal hypoxic ischemic

encephalopathy." Brain Dev 28(3): 178-182.

Anthony, D. C., S. J. Bolton, S. Fearn and V. H. Perry (1997). "Age-related effects of

interleukin-1 beta on polymorphonuclear neutrophil-dependent increases in

blood-brain barrier permeability in rats." Brain 120 ( Pt 3): 435-444.

Bailly-Maitre, B., B. F. Belgardt, S. D. Jordan, B. Coornaert, M. J. von Freyend, A.

Kleinridders, et al. (2010). "Hepatic Bax inhibitor-1 inhibits IRE1alpha and

protects from obesity-associated insulin resistance and glucose intolerance." J

Biol Chem 285(9): 6198-6207.

Bailly-Maitre, B., C. Fondevila, F. Kaldas, N. Droin, F. Luciano, J. E. Ricci, et al.

(2006). "Cytoprotective gene bi-1 is required for intrinsic protection from

endoplasmic reticulum stress and ischemia-reperfusion injury." Proc Natl Acad

Sci U S A 103(8): 2809-2814.

Balduini, W., S. Carloni, E. Mazzoni and M. Cimino (2004). "New therapeutic strategies

in perinatal stroke." Curr Drug Targets CNS Neurol Disord 3(4): 315-323.

Banasiak, K. J., Y. Xia and G. G. Haddad (2000). "Mechanisms underlying hypoxia-

induced neuronal apoptosis." Prog Neurobiol 62(3): 215-249.

Barone, F. C. and G. Z. Feuerstein (1999). "Inflammatory mediators and stroke: new

opportunities for novel therapeutics." J Cereb Blood Flow Metab 19(8): 819-834.

Bartha, A. I., J. Shen, K. H. Katz, R. E. Mischel, K. R. Yap, J. A. Ivacko, et al. (2007).

"Neonatal seizures: multicenter variability in current treatment practices." Pediatr

Neurol 37(2): 85-90.

Behrman, S., D. Acosta-Alvear and P. Walter (2011). "A CHOP-regulated microRNA

controls rhodopsin expression." The Journal of Cell Biology 192(6): 919-927.

Bernardi, P., L. Scorrano, R. Colonna, V. Petronilli and F. Di Lisa (1999). "Mitochondria

and cell death. Mechanistic aspects and methodological issues." Eur J Biochem

264(3): 687-701.

Blomgren, K. and H. Hagberg (2006). "Free radicals, mitochondria, and hypoxia-

ischemia in the developing brain." Free Radic Biol Med 40(3): 388-397.

Bona, E., A. L. Andersson, K. Blomgren, E. Gilland, M. Puka-Sundvall, K. Gustafson, et

al. (1999). "Chemokine and inflammatory cell response to hypoxia-ischemia in

immature rats." Pediatr Res 45(4 Pt 1): 500-509.

Bonfoco, E., D. Krainc, M. Ankarcrona, P. Nicotera and S. A. Lipton (1995). "Apoptosis

and necrosis: two distinct events induced, respectively, by mild and intense insults

with N-methyl-D-aspartate or nitric oxide/superoxide in cortical cell cultures."

Proc Natl Acad Sci U S A 92(16): 7162-7166.

Bracewell, M. and N. Marlow (2002). "Patterns of motor disability in very preterm

children." Ment Retard Dev Disabil Res Rev 8(4): 241-248.

Byrd, A. E., I. V. Aragon and J. W. Brewer (2012). "MicroRNA-30c-2* limits expression

of proadaptive factor XBP1 in the unfolded protein response." The Journal of Cell

Biology 196(6): 689-698.

Cao, S. S. and R. J. Kaufman (2014). "Endoplasmic reticulum stress and oxidative stress

in cell fate decision and human disease." Antioxid Redox Signal 21(3): 396-413.

Cerio, F. G., I. Lara-Celador, A. Alvarez and E. Hilario (2013). "Neuroprotective

therapies after perinatal hypoxic-ischemic brain injury." Brain Sci 3(1): 191-214.

Chae, H. J., H. R. Kim, C. Xu, B. Bailly-Maitre, M. Krajewska, S. Krajewski, et al.

(2004). "BI-1 regulates an apoptosis pathway linked to endoplasmic reticulum

stress." Mol Cell 15(3): 355-366.

Chaudhari, N., P. Talwar, A. Parimisetty, C. Lefebvre d'Hellencourt and P. Ravanan

(2014). "A molecular web: endoplasmic reticulum stress, inflammation, and

oxidative stress." Front Cell Neurosci 8: 213.

Chen, Q., M. Crosby and A. Almasan (2003). "Redox Regulation of Apoptosis before

and after Cytochrome C Release." Korean J Biol Sci 7(1): 1-9.

Connolly, E. S., Jr., C. J. Winfree, C. J. Prestigiacomo, S. C. Kim, T. F. Choudhri, B. L.

Hoh, et al. (1997). "Exacerbation of cerebral injury in mice that express the P-

selectin gene: identification of P-selectin blockade as a new target for the

treatment of stroke." Circ Res 81(3): 304-310.

Cowell, R. M., H. Xu, J. M. Galasso and F. S. Silverstein (2002). "Hypoxic-ischemic

injury induces macrophage inflammatory protein-1alpha expression in immature

rat brain." Stroke 33(3): 795-801.

Dammann, O. and T. M. O'Shea (2008). "Cytokines and perinatal brain damage." Clin

Perinatol 35(4): 643-663, v. de Mattia, F., C. Gubser, M. M. T. van Dommelen, H. J. Visch, F. Distelmaier, A.

Postigo, et al. (2009). "Human Golgi Antiapoptotic Protein Modulates

Intracellular Calcium Fluxes." Molecular Biology of the Cell 20(16): 3638-3645.

Derrick, M., A. Drobyshevsky, X. Ji and S. Tan (2007). "A model of cerebral palsy from

fetal hypoxia-ischemia." Stroke 38(2 Suppl): 731-735.

Donnelly, N., A. M. Gorman, S. Gupta and A. Samali (2013). "The eIF2alpha kinases:

their structures and functions." Cell Mol Life Sci 70(19): 3493-3511.

Duval, C., A. V. Cantero, N. Auge, L. Mabile, J. C. Thiers, A. Negre-Salvayre, et al.

(2003). "Proliferation and wound healing of vascular cells trigger the generation

of extracellular reactive oxygen species and LDL oxidation." Free Radic Biol

Med 35(12): 1589-1598.

Edwards, S., M. Garman, A. Hughes, C. Letts and I. Sinka (1999). "Clinical Forum

Assessing the comprehension and production of language in young children: an

account of the Reynell Developmental Language Scales III." International Journal

of Language & Communication Disorders 34(2): 151-171.

Endo, M., M. Mori, S. Akira and T. Gotoh (2006). "C/EBP homologous protein (CHOP)

is crucial for the induction of caspase-11 and the pathogenesis of

lipopolysaccharide-induced inflammation." J Immunol 176(10): 6245-6253.

Fadel, J., M. Sarter and J. P. Bruno (1999). "Age-related attenuation of stimulated

cortical acetylcholine release in basal forebrain-lesioned rats." Neuroscience

90(3): 793-802.

Fatemi, A., M. A. Wilson and M. V. Johnston (2009). "Hypoxic-ischemic

encephalopathy in the term infant." Clin Perinatol 36(4): 835-858, vii.

Fernandez, M., M. F. Segura, C. Sole, A. Colino, J. X. Comella and V. Cena (2007).

"Lifeguard/neuronal membrane protein 35 regulates Fas ligand-mediated

apoptosis in neurons via microdomain recruitment." Journal of Neurochemistry

103(1): 190-203.

Ferriero, D. M. (2004). "Neonatal brain injury." N Engl J Med 351(19): 1985-1995.

Forster, C., H. B. Clark, M. E. Ross and C. Iadecola (1999). "Inducible nitric oxide

synthase expression in human cerebral infarcts." Acta Neuropathol 97(3): 215-

220.

Foster-Barber, A., B. Dickens and D. M. Ferriero (2001). "Human perinatal asphyxia:

correlation of neonatal cytokines with MRI and outcome." Dev Neurosci 23(3):

213-218.

Frijns, C. J. and L. J. Kappelle (2002). "Inflammatory cell adhesion molecules in

ischemic cerebrovascular disease." Stroke 33(8): 2115-2122.

Gardner, B. M. and P. Walter (2011). "Unfolded proteins are Ire1-activating ligands that

directly induce the unfolded protein response." Science 333(6051): 1891-1894.

Garg, A. D., A. Kaczmarek, O. Krysko, P. Vandenabeele, D. V. Krysko and P. Agostinis

(2012). "ER stress-induced inflammation: does it aid or impede disease

progression?" Trends Mol Med 18(10): 589-598.

Gluckman, P. D., J. S. Wyatt, D. Azzopardi, R. Ballard, A. D. Edwards, D. M. Ferriero,

et al. (2005). "Selective head cooling with mild systemic hypothermia after

neonatal encephalopathy: multicentre randomised trial." Lancet 365(9460): 663-

670.

Gorlach, A., P. Klappa and T. Kietzmann (2006). "The endoplasmic reticulum: folding,

calcium homeostasis, signaling, and redox control." Antioxid Redox Signal 8(9-

10): 1391-1418.

Gouillon, Z., D. Lucas, J. Li, A. L. Hagbjork, B. A. French, P. Fu, et al. (2000).

"Inhibition of ethanol-induced liver disease in the intragastric feeding rat model

by chlormethiazole." Proc Soc Exp Biol Med 224(4): 302-308.

Grow, J. and J. D. Barks (2002). "Pathogenesis of hypoxic-ischemic cerebral injury in the

term infant: current concepts." Clin Perinatol 29(4): 585-602, v.

Grzmil, M., S. Kaulfuss, P. Thelen, B. Hemmerlein, S. Schweyer, S. Obenauer, et al.

(2006). "Expression and functional analysis of Bax inhibitor-1 in human breast

cancer cells." J Pathol 208(3): 340-349.

Gubser, C., D. Bergamaschi, M. Hollinshead, X. Lu, F. J. M. van Kuppeveld and G. L.

Smith (2007). "A new inhibitor of apoptosis from vaccinia virus and eukaryotes."

Plos Pathogens 3(2): 246-259.

Hagberg, H., A. David Edwards and F. Groenendaal (2016). "Perinatal brain damage:

The term infant." Neurobiol Dis 92(Pt A): 102-112.

Hagberg, H., E. Gilland, E. Bona, L. A. Hanson, M. Hahin-Zoric, M. Blennow, et al.

(1996). "Enhanced expression of interleukin (IL)-1 and IL-6 messenger RNA and

bioactive protein after hypoxia-ischemia in neonatal rats." Pediatr Res 40(4): 603-

609.

Henke, N., D. A. Lisak, L. Schneider, J. Habicht, M. Pergande and A. Methner (2011).

"The ancient cell death suppressor BAX inhibitor-1." Cell Calcium 50(3): 251-

260.

Hetz, C., F. Martinon, D. Rodriguez and L. H. Glimcher (2011). "The unfolded protein

response: integrating stress signals through the stress sensor IRE1alpha." Physiol

Rev 91(4): 1219-1243.

Hu, L., T. F. Smith and G. Goldberger (2009). "LFG: a candidate apoptosis regulatory

gene family." Apoptosis 14(11): 1255-1265.

Ivacko, J., J. Szaflarski, C. Malinak, C. Flory, J. S. Warren and F. S. Silverstein (1997).

"Hypoxic-ischemic injury induces monocyte chemoattractant protein-1 expression

in neonatal rat brain." J Cereb Blood Flow Metab 17(7): 759-770.

Iwata, A., R. A. de Claro, V. L. Morgan-Stevenson, J. C. Tupper, B. R. Schwartz, L. Liu,

et al. (2011). "Extracellular administration of BCL2 protein reduces apoptosis and

improves survival in a murine model of sepsis." PLoS One 6(2): e14729.

Jager, R., M. J. Bertrand, A. M. Gorman, P. Vandenabeele and A. Samali (2012). "The

unfolded protein response at the crossroads of cellular life and death during

endoplasmic reticulum stress." Biol Cell 104(5): 259-270.

Jean, J. C., S. M. Oakes and M. Joyce-Brady (1999). "The Bax inhibitor-1 gene is

differentially regulated in adult testis and developing lung by two alternative

TATA-less promoters." Genomics 57(2): 201-208.

Kaser, A., A. H. Lee, A. Franke, J. N. Glickman, S. Zeissig, H. Tilg, et al. (2008). "XBP1

links ER stress to intestinal inflammation and confers genetic risk for human

inflammatory bowel disease." Cell 134(5): 743-756.

Kawai-Yamada, M., Y. Ohori and H. Uchimiya (2004). "Dissection of Arabidopsis Bax

inhibitor-1 suppressing Bax-, hydrogen peroxide-, and salicylic acid-induced cell

death." Plant Cell 16(1): 21-32.

Kerr, J. F., A. H. Wyllie and A. R. Currie (1972). "Apoptosis: a basic biological

phenomenon with wide-ranging implications in tissue kinetics." Br J Cancer

26(4): 239-257.

Kim, H. R., G. H. Lee, E. Y. Cho, S. W. Chae, T. Ahn and H. J. Chae (2009). "Bax

inhibitor 1 regulates ER-stress-induced ROS accumulation through the regulation

of cytochrome P450 2E1." J Cell Sci 122(Pt 8): 1126-1133.

Kopprasch, S., J. Pietzsch and J. Graessler (2003). "Validation of different

chemilumigenic substrates for detecting extracellular generation of reactive

oxygen species by phagocytes and endothelial cells." Luminescence 18(5): 268-

273.

Korennykh, A. and P. Walter (2012). "Structural basis of the unfolded protein response."

Annu Rev Cell Dev Biol 28: 251-277.

Korennykh, A. V., P. F. Egea, A. A. Korostelev, J. Finer-Moore, C. Zhang, K. M. Shokat,

et al. (2009). "The unfolded protein response signals through high-order assembly

of Ire1." Nature 457(7230): 687-693.

Krajewska, M., L. Xu, W. Xu, S. Krajewski, C. L. Kress, J. Cui, et al. (2011).

"Endoplasmic reticulum protein BI-1 modulates unfolded protein response

signaling and protects against stroke and traumatic brain injury." Brain Res 1370:

227-237.

Lafemina, M. J., R. A. Sheldon and D. M. Ferriero (2006). "Acute hypoxia-ischemia

results in hydrogen peroxide accumulation in neonatal but not adult mouse brain."

Pediatr Res 59(5): 680-683.

Lai, M. C. and S. N. Yang (2011). "Perinatal hypoxic-ischemic encephalopathy." J

Biomed Biotechnol 2011: 609813.

Lee, G. H., H. K. Kim, S. W. Chae, D. S. Kim, K. C. Ha, M. Cuddy, et al. (2007). "Bax

inhibitor-1 regulates endoplasmic reticulum stress-associated reactive oxygen

species and heme oxygenase-1 expression." J Biol Chem 282(30): 21618-21628.

Leist, M. and M. Jaattela (2001). "Triggering of apoptosis by cathepsins." Cell Death

Differ 8(4): 324-326.

Li, B., J. C. Reed, H. R. Kim and H. J. Chae (2012). "Proteomic profiling of differentially

expressed proteins from Bax inhibitor-1 knockout and wild type mice." Mol Cells

34(1): 15-23.

Li, B., R. K. Yadav, G. S. Jeong, H. R. Kim and H. J. Chae (2014). "The characteristics

of Bax inhibitor-1 and its related diseases." Curr Mol Med 14(5): 603-615.

Li, S. J., W. Liu, J. L. Wang, Y. Zhang, D. J. Zhao, T. J. Wang, et al. (2014). "The role of

TNF-alpha, IL-6, IL-10, and GDNF in neuronal apoptosis in neonatal rat with

hypoxic-ischemic encephalopathy." Eur Rev Med Pharmacol Sci 18(6): 905-909.

Li, X., Y. Wang, H. Wang, C. Huang, Y. Huang and J. Li (2015). "Endoplasmic

reticulum stress is the crossroads of autophagy, inflammation, and apoptosis

signaling pathways and participates in liver fibrosis." Inflamm Res 64(1): 1-7.

Lin, W., S. L. Bailey, H. Ho, H. P. Harding, D. Ron, S. D. Miller, et al. (2007). "The

integrated stress response prevents demyelination by protecting oligodendrocytes

against immune-mediated damage." J Clin Invest 117(2): 448-456.

Lin, W., Y. Lin, J. Li, A. G. Fenstermaker, S. W. Way, B. Clayton, et al. (2013).

"Oligodendrocyte-specific activation of PERK signaling protects mice against

experimental autoimmune encephalomyelitis." J Neurosci 33(14): 5980-5991.

Lin, Y., X. Pang, G. Huang, S. Jamison, J. Fang, H. P. Harding, et al. (2014). "Impaired

eukaryotic translation initiation factor 2B activity specifically in oligodendrocytes

reproduces the pathology of vanishing white matter disease in mice." J Neurosci

34(36): 12182-12191.

Lisbona, F., D. Rojas-Rivera, P. Thielen, S. Zamorano, D. Todd, F. Martinon, et al.

(2009). "BAX inhibitor-1 is a negative regulator of the ER stress sensor

IRE1alpha." Mol Cell 33(6): 679-691.

Liu, F. and L. D. McCullough (2013). "Inflammatory responses in hypoxic ischemic

encephalopathy." Acta Pharmacol Sin 34(9): 1121-1130.

Lockshin, R. A. and C. M. William (1965). "Programmed Cell Death. 3. Neural Control

of the Breakdown of the Intersegmental Muscles of Silkmoths." J Insect Physiol

11: 601-610.

Loddick, S. A. and N. J. Rothwell (1996). "Neuroprotective effects of human

recombinant interleukin-1 receptor antagonist in focal cerebral ischaemia in the

rat." J Cereb Blood Flow Metab 16(5): 932-940.

Loeffler, M. and G. Kroemer (2000). "The mitochondrion in cell death control:

certainties and incognita." Exp Cell Res 256(1): 19-26.

Ma, Y., X. Xia, J. M. Cheng, Y. Q. Kuang, T. Yang, L. B. Yang, et al. (2014). "Emodin

inhibits inducible nitric oxide synthase in a rat model of craniocerebral explosive

injury." Neurochem Res 39(9): 1809-1816.

Madeo, F. and G. Kroemer (2009). "Intricate links between ER stress and apoptosis."

Mol Cell 33(6): 669-670.

Malhotra, J. D. and R. J. Kaufman (2007). "Endoplasmic reticulum stress and oxidative

stress: a vicious cycle or a double-edged sword?" Antioxid Redox Signal 9(12):

2277-2293.

Martin, D., N. Chinookoswong and G. Miller (1994). "The interleukin-1 receptor

antagonist (rhIL-1ra) protects against cerebral infarction in a rat model of

hypoxia-ischemia." Exp Neurol 130(2): 362-367.

Martino, M. B., L. Jones, B. Brighton, C. Ehre, L. Abdulah, C. W. Davis, et al. (2013).

"The ER stress transducer IRE1beta is required for airway epithelial mucin

production." Mucosal Immunol 6(3): 639-654.

Meares, G. P., K. J. Hughes, A. Naatz, F. R. Papa, F. Urano, P. A. Hansen, et al. (2011).

"IRE1-dependent activation of AMPK in response to nitric oxide." Mol Cell Biol

31(21): 4286-4297.

Mesples, B., F. Plaisant and P. Gressens (2003). "Effects of interleukin-10 on neonatal

excitotoxic brain lesions in mice." Brain Res Dev Brain Res 141(1-2): 25-32.

Millar, L. J., L. Shi, A. Hoerder-Suabedissen and Z. Molnar (2017). "Neonatal Hypoxia

Ischaemia: Mechanisms, Models, and Therapeutic Challenges." Front Cell

Neurosci 11: 78.

Murphy, M. P. (2009). "How mitochondria produce reactive oxygen species." Biochem J

417(1): 1-13.

Nakai, A. (2007). "Role of mitochondrial permeability transition in the immature brain

following intrauterine ischemia." J Nippon Med Sch 74(3): 190-201.

Nakajima, W., A. Ishida, M. S. Lange, K. L. Gabrielson, M. A. Wilson, L. J. Martin, et

al. (2000). "Apoptosis Has a Prolonged Role in the Neurodegeneration after

Hypoxic Ischemia in the Newborn Rat." The Journal of Neuroscience 20(21):

7994-8004.

Nieto, N., S. L. Friedman and A. I. Cederbaum (2002). "Cytochrome P450 2E1-derived

reactive oxygen species mediate paracrine stimulation of collagen I protein

synthesis by hepatic stellate cells." J Biol Chem 277(12): 9853-9864.

Nogawa, S., F. Zhang, M. E. Ross and C. Iadecola (1997). "Cyclo-oxygenase-2 gene

expression in neurons contributes to ischemic brain damage." J Neurosci 17(8):

2746-2755.

Oyadomari, S. and M. Mori (2004). "Roles of CHOP/GADD153 in endoplasmic

reticulum stress." Cell Death Differ 11(4): 381-389.

Perlman, J. M. (2004). "Brain injury in the term infant." Semin Perinatol 28(6): 415-424.

Raha, S. and B. H. Robinson (2000). "Mitochondria, oxygen free radicals, disease and

ageing." Trends Biochem Sci 25(10): 502-508.

Relton, J. K. and N. J. Rothwell (1992). "Interleukin-1 receptor antagonist inhibits

ischaemic and excitotoxic neuronal damage in the rat." Brain Res Bull 29(2): 243-

246.

Rissanen, A., J. Sivenius and J. Jolkkonen (2006). "Prolonged bihemispheric alterations

in unfolded protein response related gene expression after experimental stroke."

Brain Res 1087(1): 60-66.

Robertson, C. M. and M. Perlman (2006). "Follow-up of the term infant after hypoxic-

ischemic encephalopathy." Paediatr Child Health 11(5): 278-282.

Robinson, K. S., A. Clements, A. C. Williams, C. N. Berger and G. Frankel (2011). "Bax

inhibitor 1 in apoptosis and disease." Oncogene 30(21): 2391-2400.

Rojas-Rivera, D., R. Armisen, A. Colombo, G. Martinez, A. L. Eguiguren, A. Diaz, et al.

(2012). "TMBIM3/GRINA is a novel unfolded protein response (UPR) target

gene that controls apoptosis through the modulation of ER calcium homeostasis."

Cell Death Differ 19(6): 1013-1026.

Rojas-Rivera, D. and C. Hetz (2015). "TMBIM protein family: ancestral regulators of

cell death." Oncogene 34(3): 269-280.

Ron, D. and S. R. Hubbard (2008). "How IRE1 reacts to ER stress." Cell 132(1): 24-26.

Rootwelt, T., R. Almaas, S. Oyasaeter, A. Moen and O. D. Saugstad (1995). "Release of

xanthine oxidase to the systemic circulation during resuscitation from severe

hypoxemia in newborn pigs." Acta Paediatr 84(5): 507-511.

Rossiter, J., L. Anderson, F. Yang and G. Cole (2002). "Caspase-3 activation and

caspase-like proteolytic activity in human perinatal hypoxic-ischemic brain

injury." Acta Neuropathologica 103(1): 66-73.

Rouschop, K. M., T. van den Beucken, L. Dubois, H. Niessen, J. Bussink, K. Savelkouls,

et al. (2010). "The unfolded protein response protects human tumor cells during

hypoxia through regulation of the autophagy genes MAP1LC3B and ATG5." J

Clin Invest 120(1): 127-141.

Rzymski, T., M. Milani, L. Pike, F. Buffa, H. R. Mellor, L. Winchester, et al. (2010).

"Regulation of autophagy by ATF4 in response to severe hypoxia." Oncogene

29(31): 4424-4435.

Sano, R. and J. C. Reed (2013). "ER stress-induced cell death mechanisms." Biochim

Biophys Acta 1833(12): 3460-3470.

Savman, K., M. Blennow, K. Gustafson, E. Tarkowski and H. Hagberg (1998). "Cytokine

response in cerebrospinal fluid after birth asphyxia." Pediatr Res 43(6): 746-751.

Schweitzer, B., V. Taylor, A. A. Welcher, M. McClelland and U. Suter (1998). "Neural

membrane protein 35 (NMP35): A novel member of a gene family which is

highly expressed in the adult nervous system." Molecular and Cellular

Neuroscience 11(5-6): 260-273.

Shalak, L. F., A. R. Laptook, H. S. Jafri, O. Ramilo and J. M. Perlman (2002). "Clinical

chorioamnionitis, elevated cytokines, and brain injury in term infants." Pediatrics

110(4): 673-680.

Shankaran, S., A. R. Laptook, R. A. Ehrenkranz, J. E. Tyson, S. A. McDonald, E. F.

Donovan, et al. (2005). "Whole-body hypothermia for neonates with hypoxic-

ischemic encephalopathy." N Engl J Med 353(15): 1574-1584.

Somia, N. V., M. J. Schmitt, D. E. Vetter, D. Van Antwerp, S. F. Heinemann and I. M.

Verma (1999). "LFG: An anti-apoptotic gene that provides protection from Fas-

mediated cell death." Proceedings of the National Academy of Sciences of the

United States of America 96(22): 12667-12672.

Szaflarski, J., D. Burtrum and F. S. Silverstein (1995). "Cerebral hypoxia-ischemia

stimulates cytokine gene expression in perinatal rats." Stroke 26(6): 1093-1100.

Tu, B. P. and J. S. Weissman (2002). "The FAD- and O(2)-dependent reaction cycle of

Ero1-mediated oxidative protein folding in the endoplasmic reticulum." Mol Cell

10(5): 983-994.

Upton, J.-P., L. Wang, D. Han, E. S. Wang, N. E. Huskey, L. Lim, et al. (2012). "IRE1α

Cleaves Select microRNAs During ER Stress to Derepress Translation of

Proapoptotic Caspase-2." Science 338(6108): 818-822.

Urra, H., E. Dufey, F. Lisbona, D. Rojas-Rivera and C. Hetz (2013). "When ER stress

reaches a dead end." Biochim Biophys Acta 1833(12): 3507-3517. van 't Veer, L. J., H. Dai, M. J. van de Vijver, Y. D. He, A. A. Hart, M. Mao, et al.

(2002). "Gene expression profiling predicts clinical outcome of breast cancer."

Nature 415(6871): 530-536. van Handel, M., H. Swaab, L. S. de Vries and M. J. Jongmans (2007). "Long-term

cognitive and behavioral consequences of neonatal encephalopathy following

perinatal asphyxia: a review." Eur J Pediatr 166(7): 645-654.

Vannucci, R. C., J. R. Connor, D. T. Mauger, C. Palmer, M. B. Smith, J. Towfighi, et al.

(1999). "Rat model of perinatal hypoxic-ischemic brain damage." J Neurosci Res

55(2): 158-163.

Vannucci, R. C. and S. J. Vannucci (1997). "A model of perinatal hypoxic-ischemic brain

damage." Ann N Y Acad Sci 835: 234-249.

Vannucci, R. C. and S. J. Vannucci (2005). "Perinatal hypoxic-ischemic brain damage:

evolution of an animal model." Dev Neurosci 27(2-4): 81-86.

Vasquez-Vivar, J., Z. Shi, K. Luo, K. Thirugnanam and S. Tan (2017).

"Tetrahydrobiopterin in antenatal brain hypoxia-ischemia-induced motor

impairments and cerebral palsy." Redox Biol 13: 594-599.

Volpe, J. J. (1976). "Perinatal hypoxic-ischemic brain injury." Pediatr Clin North Am

23(3): 383-397.

Volpe, J. J. (2001). "Perinatal brain injury: from pathogenesis to neuroprotection." Ment

Retard Dev Disabil Res Rev 7(1): 56-64.

Wang, G. Q., B. R. Gastman, E. Wieckowski, L. A. Goldstein, A. Rabinovitz, X. M. Yin,

et al. (2001). "Apoptosis-resistant mitochondria in T cells selected for resistance

to Fas signaling." J Biol Chem 276(5): 3610-3619.

Watanabe, N. and E. Lam (2009). "Bax Inhibitor-1, a conserved cell death suppressor, is

a key molecular switch downstream from a variety of biotic and abiotic stress

signals in plants." Int J Mol Sci 10(7): 3149-3167.

Wirrell, E. C., E. A. Armstrong, L. D. Osman and J. Y. Yager (2001). "Prolonged

seizures exacerbate perinatal hypoxic-ischemic brain damage." Pediatr Res 50(4):

445-454.

Wu, R., Q. H. Zhang, Y. J. Lu, K. Ren and G. H. Yi (2015). "Involvement of the

IRE1alpha-XBP1 pathway and XBP1s-dependent transcriptional reprogramming

in metabolic diseases." DNA Cell Biol 34(1): 6-18.

Xu, C., B. Bailly-Maitre and J. C. Reed (2005). "Endoplasmic reticulum stress: cell life

and death decisions." J Clin Invest 115(10): 2656-2664.

Xu, Q. and J. C. Reed (1998). "Bax inhibitor-1, a mammalian apoptosis suppressor

identified by functional screening in yeast." Mol Cell 1(3): 337-346.

Xu, W., L. Liu, I. G. Charles and S. Moncada (2004). "Nitric oxide induces coupling of

mitochondrial signalling with the endoplasmic reticulum stress response." Nat

Cell Biol 6(11): 1129-1134.

Yamamoto, K., T. Sato, T. Matsui, M. Sato, T. Okada, H. Yoshida, et al. (2007).

"Transcriptional induction of mammalian ER quality control proteins is mediated

by single or combined action of ATF6alpha and XBP1." Dev Cell 13(3): 365-376.

Yamazaki, T., M. Muramoto, T. Oe, N. Morikawa, O. Okitsu, T. Nagashima, et al.

(2006). "Diclofenac, a non-steroidal anti-inflammatory drug, suppresses apoptosis

induced by endoplasmic reticulum stresses by inhibiting caspase signaling."

Neuropharmacology 50(5): 558-567.

Yang, H., Y. Nie, Y. Li and Y. J. Wan (2010). "Histone modification-mediated CYP2E1

gene expression and apoptosis of HepG2 cells." Exp Biol Med (Maywood)

235(1): 32-39.

Yang, L., D. Zhao, J. Ren and J. Yang (2015). "Endoplasmic reticulum stress and protein

quality control in diabetic cardiomyopathy." Biochim Biophys Acta 1852(2): 209-

218.

Zanelli, S., M. Buck and K. Fairchild (2011). "Physiologic and pharmacologic

considerations for hypothermia therapy in neonates." J Perinatol 31(6): 377-386.

Zeeshan, H. M., G. H. Lee, H. R. Kim and H. J. Chae (2016). "Endoplasmic Reticulum

Stress and Associated ROS." Int J Mol Sci 17(3): 327.

Zhang, K. and R. J. Kaufman (2008). "From endoplasmic-reticulum stress to the

inflammatory response." Nature 454(7203): 455-462.

Zhao, H., A. Ito, S. H. Kimura, N. Yabuta, N. Sakai, M. Ikawa, et al. (2006). "RECS1

deficiency in mice induces susceptibility to cystic medial degeneration." Genes

Genet Syst 81(1): 41-50.

Zhao, H., A. Ito, N. Sakai, Y. Matsuzawa, S. Yamashita and H. Nojima (2006). "RECS1

is a negative regulator of matrix metalloproteinase-9 production and aged RECS1

knockout mice are prone to aortic dilation." Circ J 70(5): 615-624.

Zhou, J. H., T. Zhu, C. X. Hu, H. Y. Li, G. Chen, G. Xu, et al. (2008). "Comparative

genomics and function analysis on BI1 family." Computational Biology and

Chemistry 32(3): 159-162.

CHAPTER TWO

GENE DELIVERY OF ADENOVIRAL-TMBIM6 VECTOR RESTORES ER

STRESS INDUCED APOPTOSIS IN A NEONATAL HI RAT MODEL

Desislava Doycheva, PhD1; Ningbo Xu, PhD2; Harpreet Kaur, MD1; Jay Malaguit, BS1;

Jiping Tang, MD1; John Zhang, MD, PhD1,2 *

1Department of Physiology and Pharmacology, Basic Sciences, School of Medicine, Loma

Linda University, Loma Linda, CA 92354, USA

2Departments of Anesthesiology, Neurosurgery and Neurology, Loma Linda University

School of Medicine, Loma Linda, CA 92354, USA

Abstract

A wide variety of conditions may contribute to oxidative protein folding processes in the endoplasmic reticulum (ER) which cause ER stress. Three major pathways, referred to as the unfolded protein response (UPR), transmit information regarding protein folding status in the ER lumen to help cells cope with altered protein synthesis demands. However, after hypoxic ischemic (HI) injury, prolonged activation of the UPR triggers pro-apoptotic signaling causing cells to undergo apoptosis.

Here we identified Bax Inhibitor-1 (BI-1), an evolutionary protein encoded by the

Transmembrane Bax inhibitor Motif Containing 6 (TMBIM6) gene, as a novel modulator of ER stress induced apoptosis after HI brain injury in a neonatal rat pup. BI-1 may modulate the UPR response by a direct physical interaction with the stress sensor, IRE1α, and via indirect inhibition of the PERK receptor. The main objective of our study is to overexpress BI-1 using viral-mediated gene delivery of human adenoviral-TMBIM6 (Ad-

TMBIM6) vector, to investigate its anti-apoptotic effects as well as elucidate its signaling pathways in an in vitro oxygen glucose deprivation (OGD) model and in an in vivo neonatal HI rat model.

Cells transfected with Ad-TMBIM6 vector improved cell viability, while manipulation of both IRE1α and PERK determined IRE1α as the predominant pathway via which BI-1 confers neuroprotection after OGD. Our in vivo data showed similar results in ten-day old unsexed Sprague-Daley rat pups that underwent right common carotid artery ligation followed by 1.5h of hypoxia. Rat pups injected with Ad-TMBIM6 vector showed a reduction in percent infarcted area and improved long-term neurological

outcomes, as well as attenuated neuronal degeneration and reduced ER stress induced neuronal apoptosis via inhibition of IRE1α signaling pathway.

This study showed a novel role for BI-1 and ER stress in the pathophysiology of

HI and provided a basis for BI-1 as a potential therapeutic target.

Introduction

Hypoxic-ischemic (HI) brain injury is a result of hypoxemia or reduced cerebral blood flow (Ferriero 2004) causing pulmonary immaturity, respiratory distress syndrome, hypercapnia, hypo perfusion, seizures and long-term cognitive and behavioral deficits

(Wirrell et al. , 2001; Bracewell et al. , 2002; Ferriero 2004; Bartha et al. , 2007; Yang et al. , 2015). HI affects 1-4 cases per 1000 births (Vannucci 2000) with morbidity rates ranging between 10-80% based on the severity of injury (Kaser et al. , 2008; Chaudhari et al. , 2014). Current treatments include hypothermia, anticonvulsants, fluid and electrolytes management, and drugs such as atropine and epinephrine but (Zanelli et al. ,

2011; Jacobs et al. , 2013) despite the efficacy of some of the treatments, 40-50% of treated patients still have persistent neurodevelopmental deficits (Barrett et al. , 2007;

Zanelli et al. , 2011).

The Endoplasmic reticulum (ER) is responsible for correct protein folding and protein function (Zhang et al. , 2014). Protein folding is crucial for cell survival.

However, under pathological conditions, such as HI, there is an accumulation of unfolded proteins beyond the capacity of the chaperons to fold the protein causing ER stress (Xu et al. , 2005; Barrett et al. , 2007; Zhang et al. , 2014). This condition triggers the unfolded protein response, resulting in the activation of pro-apoptotic cascades (Madeo et al. ,

2009; Sano et al. , 2013). The three transmembrane receptor pathways activated as a result of ER stress are (1) Inositol-requiring protein 1 alpha (IRE1α), (2) RNA-dependent protein kinase-like ER kinase (PERK) and (3) activating transcription factor 6 (ATF6)

(Sano et al. , 2013). The main function of these receptors is to alleviate stress and restore normal cell function by reducing misfolded proteins. However, under prolonged ER

stress, these pathways become over activated resulting in incorrect signalling, leading to apoptosis (Tabas et al. , 2011). This is further exacerbated by subsequent reactive oxygen species (ROS) accumulation.

The Bax-inhibitor 1 (BI-1) protein mainly resides in the ER membrane and is encoded by the TMBIM6 gene (Xu et al. , 1998; Li et al. , 2014; Rojas-Rivera et al. ,

2015). BI-1 is part of the Bcl-2 family of proteins that regulates pro-apoptotic molecules such as Bax by directly interacting with Bcl-2 as shown in in vivo experiments (Xu et al. ,

1998; Watanabe et al. , 2009; Iwata et al. , 2011). Studies have shown that cells overexpressing BI-1, that have been under ER stress inducers, or during OGD and I/R injury, did not show apoptotic phenotypes; whereas, BI-1 knockout cells showed the opposite effects, indicating that BI-1 aids to inhibit ER stress signaling pathways (Chae et al. , 2004). BI-1 can also physically bind to and inhibit IRE1α (Bailly-Maitre et al. ,

2010) and its downstream molecules such as XBP1 (Lisbona et al. , 2009; Bailly-Maitre et al. , 2010). Several in vitro studies have been done in TBI, MCAO and hepatic ischemia/reperfusion models which present data in support of BI-1’s anti-apoptotic effects via inhibition of ER stress and the subsequent pro-apoptotic pathways (Bailly-

Maitre et al. , 2006; Krajewska et al. , 2011; Ma et al. , 2014).

It is unclear whether BI-1 exerts its anti-apoptotic effects mainly via IRE1α inhibition or whether it also plays a role in inhibiting the PERK receptor; Using an in vitro oxygen glucose deprivation (OGD) model, we sought to determine whether IRE1α is BI-1’s major pathway which we can then translate to our in vivo model. Our specific objective is to establish that overexpression of BI-1 protein, via administration of human adenoviral-

TMBIM6 (Ad-TMBIM6) vector, will attenuate the morphological and neurological

consequences post neonatal HI by attenuation of ER stress induced pathways. Elucidating

BI-1’s mechanism in vitro and translating it to our in vivo HI model is crucial and will help to provide a basis for BI-1 as a potential therapeutic target in clinics.

Materials and methods

In Vitro Experiments

Rat pheochromocytoma cells, PC-12 cells, were used at passage 6 through 9. Cells were grown as previously described(Dickson et al. , 1986).

Oxygen Glucose Deprivation Model

Cells were replaced with glucose deprived media after which they were placed in hypoxic chamber and flushed with 1% Oxygen for 1h, 1.5h, 3h, 5h and 6h. Complete growth media was then added, and cells allowed to recover for 18h after which were prepared for cell death assay and western blotting(Agani et al. , 2002; Souvenir et al. ,

2014).

Cell Death Assay

Assessed with trypan exclusion as previously described(Souvenir et al. , 2014).

Briefly, cells were scraped and then centrifuged for 5min after which they were re- suspended in 10ml complete growth media. 20µl from cells was added to 20µl Trypan blue and well mixed and let to sit for 3min. 10µl from mixture was placed on cell counter slides, slides were then read using an automated cell counter. The average of 6 counts was used (Souvenir et al. , 2011).

Western Blotting for Cells

Cells were collected 18h after OGD and stored for western blotting as previously described (Dominguez et al. , 2008; Souvenir et al. , 2014). Briefly, cells were centrifuged and then re-suspended in 1ml PBS, after which they were centrifuged one more time at 14000rpm for 5min. The supernatant was removed, leaving only the pellet.

RIPA was added with protease inhibitor and pipetted thoroughly until pellet was fully broken down. It was then allowed to sit on ice for 20min following which it was centrifuged for 30min. The supernatant was harvested, and the pellet was discarded.

Protein concentration was measured using Bradford assay and SD-PAGE electrophoresis was performed as previously described(Dominguez et al. , 2008). Primary antibodies: BI-

1 antibody (1:100, Abcam), anti-IRE1α antibody (1:100, Abcam), Anti-PERK antibody

(1:100, Abcam) and Casp3/cleaved caspase 3 antibodies (1:400, cell signaling technology).

Calculation of MOI

To determine the total amount of infectious particles needed to infect one cell,

Multiplicity of Infection (MOI), was calculated as follows: #cells * desired MOI= total

PFU (or Plaque Forming Units) needed; (total PFU needed) / (PFU/ml) = total ml of virus needed to reach desired dose.

Transfections

siRNA transfection: Prior to transfection, cells were differentiated and allowed to reach 80% confluency in poly D-lysine coated 6 well plates. siRNA was prepared

according to manufacturer’s protocol (Sigma-Aldrich); a stock of 10uM siRNA was prepared, 4ul from siRNA stock were mixed with 125µl Opti-MEM. In a separate tube,

7µl liofectamine3000 was mixed with 125µl Opti-Mem. The two tubes were combined and left to sit in room temperature for 15-45min. 250µl mixture/well was applied to cells and an additional 1ml Opti-MEM was added. Cells were then placed in an incubator for

5-7h, following which media was removed and replaced with normal growth media.

CRISPR transfection: 50,000 cells were plated and allowed to grow and differentiate following which cells were transfected first with Ad-TMBIM6, then 700µl

CRISPR was added 48h later. Plasmid was diluted in 200µl sterile DNase free water

(provided) giving a concentration of 0.1µg/µl (Santa Cruz). Stock solution was further diluted with transfection medium to bring to desired concentration of 2µg-8µg per

500,000 to 1 million cells. After transfection cells were placed in incubator for 5-7h following which media was removed and replaced with normal media and prepared for

OGD.

STF-083010 (IRE1α antagonist), CCT020312 (PERK agonist) and APY-29 (IRE1α agonist) transfection: dissolved and prepared according to manufacturer recommendations. STF-083010 was prepared to final concentration at 20µM or 60µM,

CCT020312 was tested at 10µM and APY-29 was tested at 20µM and 60µM. Best dose was chosen at 20µM for both drugs (cell viability quantification data not shown).

In Vivo Experiments

The Institutional Animal Care and Use Committee of Loma Linda University has approved all protocols. The animals were cared for and all studies conducted in

accordance with the United States Public Health Service's Policy on Humane Care and

Use of Laboratory Animals. Sprague Dawley rat mothers, with litters of 10~12 pups, were purchased from Harlan Labs (Livermore, CA). The following have been applied to ensure robust and unbiased experimental design and analysis of data, (A) Randomization

& Blinding: All animals were randomly assigned to experimental groups via generation of random number assignment. Furthermore, all investigators were blinded to animal groups and the interventions that animals received. (B) Control Groups: Appropriate control groups were included for each intervention group. Furthermore, for siRNA experiments, animals were injected only siRNA or scramble, without treatment (Ad-

TMBIM6) as control. No animals were excluded apart from the normal 12% mortality rates for this model.

Hypoxic Ischemic Rat Model

HI surgery was performed using the established Rice Vannucci model as previously described (Vannucci et al. , 1997; Vannucci et al. , 2004). Briefly, P10 unsexed Sprague-Daley rat pups were placed into a temperature-controlled chamber for induction of general anesthesia. The animals were anesthetized with 3% isoflurane gas in air and maintained at 2.5% isoflurane in air during surgery. Throughout the surgical and postoperative period, temperature was controlled with heating blankets and incubators.

After induction of anesthesia the neck of the rats was prepared and draped using standard sterile techniques. Next a small midline neck incision on the anterior neck was made with a No. 11 blade surgical knife (approximately 3-5 mm in length). Using gentle blunt dissection, the right carotid artery was isolated and gently separated from surrounding

structures. The carotid artery was then ligated with 5-O surgical suture. After the carotid artery had been ligated, the skin was sutured. Rats were then allowed to recover for 1h on a heated blanket. Thereafter, they were placed in a 500-ml airtight jar partially submerged in a 37 °C water bath to maintain a constant thermal environment. Rat pups were exposed to a gas mixture of 8% Oxygen and 92% Nitrogen for 90min. Thereafter, the animals were returned to their mothers and monitored daily.

Drug Administration

Human adenoviral TMBIM6 vector (Ad-TMBIM6) (Vector Biolabs) was injected intracerebroventricularly (icv) at 2µl containing 1.6x1011 PFU/ml or 1.7x1011 PFU/ml per injection(Bailly-Maitre et al. , 2010) at 72h, 48h and 24h pre-HI, as well as 1h post HI.

BI-1 siRNA (Sigma-Aldrich) and IREα CRISPR activation plasmid (AP) (Santa Cruz) were administered icv at 48h pre-HI.

Infection Area Calculation

Animals were euthanized at 72h post HI and infarct area was assessed using

2,3,5-Triphenyltetrazolium Chloride Monohydrate (TTC) staining. Briefly, brains were sectioned into 2mm slices and placed in TTC stain as previously described(Zhou et al. ,

2009; Li et al. , 2012). Red is indicative of live tissue.

Behavioral Analysis

Neurobehavior test were performed at 4 weeks using rotarod, water maze and foot fault. We chose these neurobehavioral tests because they effectively cover a range of

neurological functioning (sensation, motor function, and cognitive functioning). In addition, these tests have a good predictive value of impaired neurological function. The water maze test has been closely correlated with long-term functional deficits. Animals were then euthanized, and brain samples weighed and collected for histopathological examination.

Foot Fault: Animals were placed on a horizontal grid floor (square size 28 x 3 cm, wire diameter 0.4 cm) for 2 min. Foot-fault is defined as when the animal inaccurately places a hindlimb, and it falls through one of the openings in the grid. The numbers of foot-faults for each limb of each animal was recorded (Hartman et al. , 2005;

Bouet et al. , 2010).

Rotarod: The rotarod assesses motor impairment. Animals were placed on a spinning bar to test their motor function. Rats underwent stationary, constant and 5rpm acceleration which was recorded by (Hartman et al. , 2005; Bouet et al. , 2010).

Water Maze: Morris water maze test assesses learning, memory, and visual functions (Hartman et al. , 2005; Bouet et al. , 2010). Activity of animal’s swim paths were recorded and measured for quantification of distance, latency, and swimming speed for 5 days using the Video Tracking System SMART-2000 (San Diego Instruments Inc.,

CA). Animals were then sacrificed, and brains collected for histological analysis.

Western Blotting of Brain Tissue

Western blotting was performed as previously described (Li et al. , 2015). Equal amounts of protein (30μg) was loaded on an 8%~12% SDS-PAGE gel and electrophoresed, it was then transferred to a nitrocellulose membrane (0.2μm). The

membrane was blocked for 1h and then incubated with primary antibodies overnight at

4°C: BI-1 antibody (1:100, Abcam), anti-IRE1α/anti-pIRE1α antibodies (1:100, Abcam),

Anti-XBP1 antibody (1:100, Abcam), anti-Chop antibody (1:500, Proteintech),

Casp3/cleaved caspase 3 antibodies (1:400, cell signaling technology), anti-Bcl2 (1:500,

Abcam), anti-bax (1:500, Abcam) and anti-ROMO1 antibody (1:100, Abcam). The membranes were then incubated at room temperature for 1h with horseradish peroxidase- conjugated secondary antibodies (Santa Cruz Biotechnology) followed by chemiluminescence detection (GE Healthcare). The densitometric quantification of the bands was performed with image J software.

Histology

Brains were post-fixed in formalin, and then sectioned into 10μm thick coronal slices by cryostat (CM3050S; Leica Microsystems) for immunohistochemistry, Flouro-

Jade C and into 16-20μm thickness for Nissl’s staining.

Immunofluorescence staining: Rats were anesthetized at 72h post HI and transcardially perfused with PBS and formalin. Brain sections were then post fixated with formalin for 1 day and dehydrated in 30% sucrose solution. Immunofluorescence staining was performed as previously described(Shi et al. , 2017). The tissue mounted on the slides were washed with 0.1M PBS three times for 5min then incubated in 0.3% Triton

X-100 in 0.1M PBS for 30min at room temperature. This was then washed with 0.1M

PBS for 5min, 3 times, and primary antibodies were applied overnight at 4°C: anti-BI-1 antibody (1:100, Abcam), anti-IRE1α antibody (1:100, Abcam), NeuN ((1:100, Abcam), and CC3 (1:100, Abcam) respectively. After washing with PBS, sections were incubated

with appropriate secondary antibodies: anti-rabbit IgG-TR, anti-mouse IgG-FITC, anti- goat IgG-FITC, anti-rabbit IgG-FITC (1:200) for 1h @ RT. Finally, slides were covered with DAPI and visualized under a fluorescent microscope and a Magna Fire SP system

(Olympus) was used to analyze microphotographs.

Fluoro-Jade C: 10μm brain sections were cut with cryostat (Leica LM3050S) for

Fluoro-Jade C as previously described(Xie et al. , 2017). Slides were immersed in 1% sodium hydroxide dissolved in 80% ethanol for 5min, followed by being rinsed for 2min in 70% ethanol, then 2min in distilled water. Slides were then incubated in 0.06% potassium permanganate solution for 10 min, followed by 2min in distilled water, and then transferred into 0.0001% solution of Fluoro-Jade C (Millipore, USA) which was dissolved in 0.1% acetic acid. Slides were then rinsed 3 times with distilled water for

1min. After the water was drained and slides dried, they were cleared in xylene for 1min, and finally cover slipped with DPX (Sigma-Aldrich, USA).

Nissl’s staining and evaluation of brain tissue loss and ventricular area: Pups were euthanized at 4weeks post HI and brains were sectioned. The prepared slides (16 μm) were dehydrated in 95% and 70% ethanol for 1min respectively, rinsed in tap water and distilled water for 30sec. The slides were then stained with 0.5% cresyl violet (Sigma-

Aldrich, USA) for 2min, and washed in distilled water for 10 and 30sec. 100% ethanol and xylene were used to dehydrate the tissue sections 2 times for 1.5min before placing a coverslip with permount on the slide (Shi et al. , 2017). Image J was used to measure percent tissue loss.

Statistical Analysis

Rats were randomly assigned to each group and the investigators were blinded from previous procedures. All values will be presented as mean ± standard deviation (SD), and analyzed by one-way analysis of variances (ANOVA), followed by Tukey test or

Holm-Sidak or Dunnett’s post hoc analysis. A P-value less than 5% will be considered significant. Sample sizes were determined by power of 0.8, α=0.05, and a 20% standard deviation from the preliminary results. The estimated sample size was 4 to 9/group.

Results

To Establish the Optimum Time for OGD and Optimal Multiplicity of Infection

(MOI) Needed of Ad-TMBIM6 Vector to Successfully Transfect PC12 Cells

Data showed a time dependent decrease in percent cell viability in cultured PC12 cells when exposed to variable periods of OGD. Cells were exposed to 5 different OGD time periods ranging from 1h to 6h. Percent cell viability was significantly reduced when compared to control at 3h, 5h and 6h (Figure 2.1A). Optimal time for OGD exposure was chosen at the 3h time point to best represent severity of our in vivo HI model.

Optimal Multiplicity of Infection (MOI) was shown to be 100MOI as it significantly improved cell viability compared to vehicle and to 200/1000MOI after 3h OGD (Figure

2.1B).

To Determine the Effect of Ad-TMBIM6 on Percent Cell Viability in an in vitro

OGD Model

Cells exposed to 3h of OGD (Vehicle) showed a significantly higher percent of cell death when compared to control and OGD+ Ad-TMBIM6 group (Figure 2.1C-D). BI-1 siRNA (a BI-1 inhibitor), APY-29 (20µM; IRE1α agonist) or CCT020312 (PERK agonist) reversed BI-1’s protective effects as seen from the significantly lower percentage of cell viability (Figure 2.1D). Ad-TMBIM6+PERK CRISPR(knockdown) did not improve cell viability when compared to OGD group. On the contrary, adding STF-

083010 (IRE1α antagonist; 20µM) to Ad-TMBIM6 showed to significantly improved cell viability (~85% living cells) when compared to PERK CRISPR or OGD (~40% living cells) group.

Control groups, OGD+Ad-TMBIM6+scramble/or DMSO/or control CRISPR showed to significantly improve cell viability when compared to Vehicle group/or APY-29/or

CCT020312/or PERK CRISPR but did not restore cell viability up to control or Ad-

TMBIM6 treated group levels.

Figure 2.1 Percent cell viability significantly improved after transfection of cells with 100MOI of Ad-TMBIM6 in an in vitro OGD Model Time dependent decrease was seen in percent cell viability in cultured PC12 cells when exposed to 1h, 1.5, 3h, 5h and 6h in OGD. Percent cell viability was significantly reduced when compared to control at 3h, 5h and 6h (A). Cells transfected with 100MOI significantly improved percent cell viability when compared to vehicle or to 200MOI and 1000MOI at 3h of OGD (B). Representative picture showing cell morphology and density of cells after OGD in all groups (C). Cell death assay showed that Ad-TMBIM6 treatment group as well as the control groups (Ad-TMBIM6+scramble siRNA; Ad- TMBIM6+DMSO and Ad-TMBIM6+control CRISPR) and Ad-TMBIM6+STF-083010 significantly improved cell viability when compared to vehicle, BI-1 siRNA, APY-29, CCT020312 and PERK CRISPR knockdown groups (D). (Data expressed as mean +/− SD; *p<0.05 vs control; #p<0.05 vs vehicle; @p<0.05 vs Ad-TMBIM6 and scramble siRNA or vs 100MOI; ε p<0.05 vs Ad-TMBIM6 only; $p<0.05 vs BI-1 siRNA; &p<0.05 vs APY-29; +p<0.05 vs DMSO; %p<0.05 vs CCT020312; ^p<0.05 vs STF-083010 and Control CRISPR. n=5-6/group using one-way ANOVA followed by Tukey multiple- comparison post hoc analysis).

To Establish BI-1’s Anti-Apoptotic Effects via the IRE1α Signaling Pathway in

an in vitro OGD Model

To test the underlying mechanism via which BI-1 attenuated apoptosis both IRE1α agonist (APY-29) and antagonist (STF-083010) were used as well as PERK agonist

(CCT020312) and antagonist (PERK CRISPR knockdown). These helped to manipulate both pathways to determine the dominant one. Quantification data showed that BI-1 expression levels were significantly increased in the Ad-TMBIM6 treated group and in the control groups: scramble siRNA, DMSO and Control CRISPR; while BI-1 siRNA and APY-29 showed to reverse those effects (Figure 2.2A-B). CCT020312 and PERK

CRISPR as well as STF-083010 showed similar expression levels of BI-1 to Ad-

TMBIM6 treatment group (Figure 2.2A-B).

Intervening at the PERK receptor did not seem to alter BI-1 expression levels as

CCT020312 did not show to affect BI-1 levels whereas intervening at IRE1α with APY-

29 significantly reduced BI-1’s levels (Figure 2.2B). Ad-TMBIM6, scramble siRNA +

Ad-TMBIM6, DMSO + Ad-TMBIM6 and Control CRISPR significantly reduced pIRE1α, pPERK and cleaved caspase 3 (CC3) expression levels (Figure 2.2C-E) compared to vehicle. Meanwhile, BI-1 siRNA, APY-29, CCT020312 and PERK CRISPR showed to increase pIRE1α expression levels (Figure 2.2C). Expression levels of pPERK were significantly upregulated after OGD in vehicle group, BI-1 siRNA and CCT020312 when compared to the rest of the groups (Figure 2.2D). While Ad-TMBIM6 and control groups (scramble siRNA, DMSO and Control CRISPR) significantly reduced CC3 expression, activation of PERK with CCT020312 did not significantly upregulate CC3 when compared to treatment or control groups (Figure 2.2E), in fact it showed to

significantly decrease CC3 expression levels. Furthermore, APY-29 showed to significantly upregulate CC3 when compared to CCT020312. In addition, STF-083010 showed to be more effective at reducing CC3 compared to PERK CRISPR (Figure 2.2E).

Figure 2.2 BI-1 exerted its anti-apoptotic effects via inhibition of IRE1α in an in vitro OGD model Representative picture of western blot bands showing protein expression levels (A). Quantification data of bands showed that BI-1 expression levels were significantly increased in the Ad-TMBIM6 treatment group and in the control groups (scramble siRNA, DMSO and Control CRISPR), while BI-1 siRNA and APY-29 groups reversed those effects (B). Ad-TMBIM6 and control groups significantly reduced pIRE1α, pPERK and CC3 expression levels (C-E). BI-1 siRNA, APY-29, CCT020312 and PERK CRISPR increased pIRE1α expression levels (C). Expression levels of pPERK were significantly upregulated in vehicle, BI-1 siRNA and CCT020312 compared to the rest of the groups. PERK CRISPR effectively knocked down pPERK expression levels (D). Ad-TMBIM6 and control groups significantly reduced CC3 expression levels, activation of PERK with CCT020312 did not significantly upregulate CC3 when compared to treatment or control groups (E). APY-29 showed to significantly upregulate CC3 when compared to CCT020312. In addition, STF-083010 was more effective at reducing CC3 compared to PERK CRISPR (E). (Data represent +/− SD; *p<0.05 vs control; #p<0.05 vs vehicle; @p<0.05 vs Ad-TMBIM6 and Ad-TMBIM6+scramble siRNA; $p<0.05 vs BI-1 siRNA; &p<0.05 vs APY-29; +p<0.05 DMSO; %p<0.05 vs CCT020312; ^p<0.05 vs STF-083010 and Control CRISPR. n=5-6/group using one-way ANOVA followed by Tukey multiple- comparison post hoc analysis).

Expression Levels of Endogenous BI-1, IRE1α, XBP1 and CHOP in Time

Dependent Manner post HI

BI-1 expression levels increased in a time dependent manner, which peaked at 24h and then significantly decreased to sham levels at 72h post HI (Figure 2.3A-B). IRE1α and XBP1 expression levels significantly increased at 6h post HI and remained elevated until 72h post HI (Figure 2.3A, C-D). CHOP levels significantly increased at 24h post HI when compared to sham (Figure 2.3A, E).

Low Dose Ad-TMBIM6 Administered at 48h pre-HI Reduced Infarct Area at

72h post HI

To test the best time of administration of Ad-TMBIM6 vector we tested 4 time points.

Ad-TMBIM6 significantly reduced infarct area compared to vehicle treated group when it was administered at 48h pre-HI (Figure 2.3F). Furthermore, to evaluate dose response of Ad-TMBIM6 viral vector on infarct area two doses were used (1) low dose of

1.6x1011PFU/ml and (2) a high dose of 1.7x1011PFU/ml. TTC staining results showed that the optimal dose for administration of the viral vector was the low dose

(1.6x1011PFU/ml) as it proved to be more effective as it significantly reduced infarct area when compared to vehicle or high dose vector (Figure 2.3G).

Figure 2.3 To Measure expression levels of endogenous BI-1, IRE1α, XBP1 and CHOP and determine optimum time and dose of Ad-TMBIM6 administration post HI Representative pictures of Western blot bands (A). Data showed that BI-1 expression levels significantly increased in a time-dependent manner from 6h to 24h, and then dropped to sham levels at 72h post HI (B). IRE1α inceased significantly post HI compared to sham and remained elevated untill 72h post HI (C). The expression of XBP1 increased post HI and remained elevated from 6h through 72h post HI (D). CHOP levels increased from 6h to 72h, peaking at 24h post HI (E). Data expressed as mean +/- SD; *p<0.05 vs sham; # p<0.05 vs 6h, @ p < 0.05 vs 24h; n=4, using one-way ANOVA followed by Tukey multiple-comparison post hoc analysis. Optimal time of injection was at 48h pre-HI, at this time significantly reduced the percentage of infarcted area versus vehicle, or the other time points (F). Low dose of Ad-TMBIM6 (1.6x1011 PFU) viral vector significantly improved the percentage of infarcted area when compared to vehicle or high dose (1.7x1011 PFU) vector (G). (Data expressed as mean +/- SD; *p<0.05 vs sham; # p<0.05 vs vehicle, @ p < 0.05 vs Ad-TMBIM6 (1.6x1011 PFU), n=6, using Unpaired T-Test for (F) and one-way ANOVA followed by Tukey multiple-comparison post hoc analysis for (G)). Blots were cropped to improve clarity.

Histological Staining at 72h post HI

Immunofluorescence staining showed a significantly higher expression of BI-1 on neurons in the Ad-TMBIM6 treatment group when compared to vehicle (Figure 2.4A).

Furthermore, IRE1α was significantly decreased in the Ad-TMBIM6 treatment group when compared to vehicle, but also showed co-localization with neurons (Figure 2.4B).

In addition, staining for BI-1 on astrocytes demonstrated minimal colocalization in all groups (Figure 2.4C). (Green was for neuronal/astrocyte staining, Red was for BI-

1/IRE1α staining and Blue was for DAPI. Merge showed the colocalization of BI-1 or

IRE1α on neurons/astrocytes. Scale bar = 50 μm; n=1/group).

Figure 2.4 Histological staining showed co-localization of BI-1 on neurons but not with astrocytes at 72h post HI Immunofluorescence staining demonstrated a significantly higher expression of BI-1 on neurons in Ad-TMBIM6 treatment group compared to sham or vehicle (A). Fluorescent levels of IRE1α was reduced in Ad-TMBIM6 treatment group when compared to vehicle (B). (Green was for neuronal staining, Red was for BI-1 or IRE1α staining and Blue was for DAPI). Merge showed the colocalization of BI-1 or IRE1α on neurons (A-B). (Scale bar = 50 μm; n=1/group). There was minimal co-localization of BI-1 with astrocytes (C). (Green was for astrocyte staining, Red was for BI-1 and Blue was for DAPI). Merge showed the colocalization of BI-1 on astrocytes (C). (Scale bar = 50 μm; n=1/group).

Ad-TMBIM6 Improved Long-term Neurological Function at 4weeks post HI

To test the effects of Ad-TMBIM6 treatment on the long-term neurological impairments induced by neonatal HI, neurological function was assessed by foot fault, rotarod and water maze at 4weeks post HI. In all three behavioral tests, vehicle group performed significantly worse compared to sham. Ad-TMBIM6 treatment group significantly improved sensorimotor coordination as assessed by the fewer number of foot faults (Figure 2.5A) and longer distance covered on the rotarod test (Figure 2.5B) versus vehicle. In the water maze test, compared to sham group, vehicle group demonstrated substantial loss of memory and learning abilities in terms of the time it took to reach the platform (Figure 2.5C) as well as the distance travelled to find the platform

(Figure 2.5D).

However, Ad-TMBIM6 showed to significantly improve both memory and learning function when compared to vehicle (Figure 2.5C-D). Administration of Ad-TMBIM6 significantly reduced brain atrophy as seen from the percentage of brain weights (Figure

2.5E). Nissl’s staining demonstrated that Ad-TMBIM6 can attenuate percent brain tissue loss (Figure 2.5F).

Figure 2.5 Ad-TMBIM6 improved long term neurological deficits when assessed at 4 weeks post HI Ad-TMBIM6 improved long-term motor deficits as seen from the significantly lower number of foot faults (A) and the longer distance traveled on the rotating rod (B) in the Ad-TMBIM6 treatment group when compared to vehicle. Long-term learning and memory were significantly improved in Ad-TMBIM6 group compared to vehicle (C-D). Percent of ipsilateral/contralateral brain weights were significantly improved in Ad-TMBIM6 group when compared to vehicle (E). Nissl staining of brains demonstrated a significant lower percentage of tissue loss in Ad-TMBIM6 group compared to vehicle (F). (Data expressed as mean +/- SD; *p＜0.05 vs sham; #p＜0.05 vs vehicle; n=6-9 using one-way ANOVA followed by Tukey multiple-comparison/or Holm-Sidak post hoc analysis).

BI-1 siRNA and IRE1α CRISPR Activation Plasmid Reversed Ad-TMBIM6

Protective Effects at 72h post HI

To evaluate whether pathway interventions can affect apoptosis and increase infarcted area, we used a BI-1 inhibitor (BI-1 siRNA) and a IRE1α activator (IRE1α CRISPR

Activation Plasmid (AP)). Results demonstrated that Ad-TMBIM6+BI-1siRNA group significantly reversed BI-1’s protective effects, as did Ad-TMBIM6+IRE1α CRISPR AP group, as seen from the significant increase in infarcted area when compared to Ad-

TMBIM6 only group (Figure 2.6A-B). Both control groups, scramble siRNA and control

CRISPR AP, did not exacerbate the damage when compared to Ad-TMBIM6 treated group. Furthermore, BI-1 siRNA only group (without treatment) further exacerbated the injuries when compared to Ad-TMBIM6 treatment group.

Figure 2.6 The effects of BI-1 siRNA and IRE1α CRISPR activation plasmid on infarct area at 72h Post HI The percentage of infarcted area was significantly increased in vehicle, BI-1siRNA and IRE1α CRISPR activation group when compared to Ad-TMBIM6 and their respective control groups (A-B). Representative schematic for mechanism study (C). (Data expressed as mean +/− SD; *p<0.05 vs sham; #p<0.05 vs vehicle; @p<0.05 vs Ad-TMBIM6; $p<0.05 vs BI-1 siRNA; +p<0.05 vs Ad-TMBIM6+scramble siRNA; &p<0.05 vs IRE1α CRISPR AP; %p<0.05 vs control CRISPR. n=6/group using one-way ANOVA followed by Tukey multiple-comparison post hoc analysis).

BI-1 Exerted its Neuroprotective Effects via the IRE1α/XBP1/CHOP pathway

at 72h post HI

To test the underlying mechanism via which BI-1 attenuated apoptosis the following groups were used: sham, vehicle, Ad-TMBIM6, Ad-TMBIM6+BI-1siRNA, Ad-

TMBIM6+scramble siRNA, Ad-TMBIM6+IRE1α CRISPR AP, AD-TMBIM6+ control

CRISPR, BI-1siRNA only and scramble siRNA only groups. Quantification data demonstrated that BI-1 expression levels were significantly increased in the Ad-TMBIM6 treatment group, Ad-TMBIM6+scramble siRNA and in the Ad-TMBIM6+control

CRISPR groups; while BI-1 siRNA and IRE1α CRISPR AP groups reversed those effects

(Figure 2.7A-B). BI-1 siRNA only and scramble siRNA only groups showed similar expression levels of BI-1 as vehicle. Ad-TMBIM6 significantly reduced expression of pIRE1α, XBP1, CHOP and ROMO1 (a modulator that induces production of ROS); while BI-1-siRNA and IRE1α CRISPR AP reversed BI-1’s protective effects which demonstrated elevated expression of pro-apoptotic pathway proteins (Figure 2.7A, C-F).

No significant difference was observed in total IRE1α expression levels

(quantification data not shown). Immunohistochemistry demonstrated that siRNA reduced expression of BI-1 in all three cell types (Figure 2.7G).

Figure 2.7 BI-1 exerted its protective effects via the IRE1α/XBP1/CHOP pathway at 72h post HI Representative picture of western blot data (A). Quantification of bands demonstrated that Ad-TMBIM6 as well as control groups (Ad-TMBIM6+scramble siRNA and Ad- TMBIM6+CRISPR Control) significantly increased BI-1 expression levels in the brain when compared to vehicle or sham. While, BI-1 siRNA and IRE1α CRISPR AP significantly reversed those effects. Furthermore, BI-1 siRNA only group also reduced BI- 1’s expression levels (B). Data showed that pIRE1α, XBP1, CHOP and ROMO1 expression levels were significantly elevated in vehicle, BI-1 siRNA and IRE1α CRISPR AP groups whereas Ad-TMBIM6 and control groups reversed those effects (C-F) at 72h post HI. (Data expressed as mean +/− SD; *p<0.05 vs sham; #p<0.05 vs vehicle; @p<0.05 vs Ad-TMBIM6; $p<0.05 vs BI-1 siRNA; +p<0.05 vs Ad-TMBIM6+scramble siRNA; &p<0.05 vs IRE1α CRISPR AP; %p<0.05 vs control CRISPR; ^p<0.05 vs Ad-TMBIM6+ BI-1 siRNA; ε p<0.05 vs BI-1 siRNA only group. n=6/group using one-way ANOVA followed by Tukey multiple-comparison post hoc analysis). Immunofluorescent staining shows the effects of siRNA on neurons, astrocytes and microglia (G) (Scale bar = 50 μm; n=1/group).

Overexpression of BI-1 Suppressed Pro-Apoptotic Protein Markers and

Attenuated Neuronal Apoptosis at 72h post HI

To evaluate whether over expression of BI-1 with Ad-TMBIM6 could attenuate apoptosis, western blot was done to quantify Bcl-2, Bax, Cleaved Caspase 3 (CC3) and

Caspase 3.The data demonstrated that Ad-TMBIM6 treatment group can significantly upregulate anti-apoptotic marker, Bcl-2 (Figure 2.8A-B), while it attenuated expression of pro-apoptotic markers, Bax and CC3 (Figure 2.8A, C-D). In addition, administration of BI-1 si-RNA and IRE1α CRISPR AP reversed Ad-TMBIM6’s protective effects; it significantly upregulated pro-apoptotic markers (Bax and CC3) and downregulated the anti-apoptotic marker, Bcl-2 (Figure 2.8A-D). There was no significant difference in total

Caspase 3 expression levels (quantification data not shown).

Since HI results in neuronal degeneration and apoptosis, we stained for CC3 and used

Flouro-Jade C staining to test Ad-TMBIM6’s neuroprotective effects. Immunoflourescent staining showed higher expression and co-localization of CC3 on neurons in vehicle group compared to sham or Ad-TMBIM6 groups. (Figure 2.8E; Scale bar 100µm). Green

Flouro-Jade C staining showed more positively stained neurons undergoing apoptosis in vehicle group compared to sham or Ad-TMBIM6 treatment group (Figure 2.8F; Scale bar

100µm).

Figure 2.8 Overexpression of BI-1 attenuated apoptosis at 72h post HI Representative picture of western blot data showing bands of the expression levels of Bcl- 2, Bax and CC3 (A). Western blot data quantification of bands showed that Ad-TMBIM6 significantly increased Bcl-2 expression levels compared to vehicle or sham (B). Ad- TMBIM6 significantly reduced Bax and CC3 at 72h post HI compared to vehicle while BI- 1 siRNA and IRE1α CRISPR AP intervention groups reversed those effects (C-D). (Data expressed as mean +/− SD; *p<0.05 vs sham; #p<0.05 vs vehicle; @p<0.05 vs Ad- TMBIM6; $p<0.05 vs BI-1 siRNA; +p<0.05 vs Ad-TMBIM6+scramble siRNA; &p<0.05 vs IRE1α CRISPR AP; %p<0.05 vs control CRISPR. n=6/group using one-way ANOVA followed by Tukey multiple-comparison/Holm-Sidak/Dunnett’s post hoc analysis). Fluorescent staining showed more positively CC3 stained neurons in vehicle group compared to sham or Ad-TMBIM6 treated group (E). (Scale bar 100µm; n=1/group). Fluoro-Jade C staining showed more positively stained neurons undergoing apoptosis in vehicle group compared to sham or Ad-TMBIM6 treatment group (F). (Scale bar 100µm; n=1/group).

Discussion

Understanding cellular and molecular mechanisms can lead to a more effective therapies needed to ameliorate the long-term deficits after Hypoxic-Ischemic (HI) injury

(Volpe 2001; Bracewell et al. , 2002; van Handel et al. , 2007). Bax Inhibitor-1 (BI-1) is a promising target for therapeutic use, however little is known about its signaling mechanism which limits its clinical translation. Given the lack of effective treatment options for neonatal HI injury, this study focused on identifying a novel role for BI-1 protein and Endoplasmic Reticulum (ER) stress in the pathophysiology of neonatal HI.

Hopefully, this new understanding can be leveraged to design interventions that positively affect the outcome of HI patients.

One of the three main pathways triggered after ER stress, main pathology encountered post HI injury, is IRE1α; a key molecule responsible for cell fate, as it can directly bind to the misfolded proteins, unlike PERK or ATF6 (Gardner et al. , 2011;

Sano et al. , 2013). Accumulation of unfolded proteins stimulates autophosphorylation of

IRE1α’s kinase domain, which then activates the RNase to splice X-box-binding protein

1 (XBP1). This acts as a potent transcription factor that triggers pro-apoptotic signaling mechanisms. PERK also contributes to ER stress by attenuating mRNA translation thus inhibiting protein synthesis (Sano et al. , 2013). We focused on IRE1α and PERK, as they are the major receptors that contribute to ER stress induced cell death.

BI-1 has been characterized in various insults as a pro-survival protein in ER stress (Chae et al. , 2004; Bailly-Maitre et al. , 2006; Lee et al. , 2007; Lisbona et al. ,

2009; Bailly-Maitre et al. , 2010; Castillo et al. , 2011; Robinson et al. , 2011). BI-1’s anti-apoptotic effects were first described in cultured cells and in vivo mice model (Chae

et al. , 2004; Bailly-Maitre et al. , 2006; Krajewska et al. , 2011; Ma et al. , 2014). It has been reported to directly bind to and inhibit IRE1α (Lisbona et al. , 2009) by abolishing its endoribonuclease activity (Castillo et al. , 2011). BI-1’s protective effects were further reported in db/db mice where BI-1 protected mice from obesity associated insulin resistance due to suppression of IRE1α (Bailly-Maitre et al. , 2010). Some evidence showed that BI-1 may also indirectly inhibit PERK signaling pathway (Lee et al. , 2007), however, other reports have shown that eIF2α phosphorylation, downstream of PERK, was unaltered with BI-1 knockout mice (Bailly-Maitre et al. , 2006).

BI-1 has been found to be a potent inhibitor of ER stress induced pathways, however the exact mechanism via which it may exert its pro-survival signaling remains unknown. In this study we investigated the role of BI-1 on ER stress-induced apoptosis, and we further elucidated BI-1’s main signaling pathway. As BI-1 has been shown to inhibit IRE1α and to some extend PERK (Lee et al. , 2007; Robinson et al. , 2011), we sought to investigate their involvement and to identify the dominant factor by using an in vitro oxygen glucose deprivation (OGD) model. PC12 cells were transfected with (1) BI-

1 siRNA, IRE1α agonist (APY-29) and antagonist (STF-083010) in combination with human adenoviral TMBIM6 (Ad-TMBIM6; gene responsible for expressing BI-1)

(Bailly-Maitre et al. , 2010) vector to test the role of the IRE1α signaling pathway and (2)

PERK CRIPSR knockdown and the CCT020312 (PERK agonist) together with Ad-

TMBIM6 to test the involvement of the PERK signaling pathway. Cell viability data indicated that Ad-TMBIM6 increased the percentage of viable cells versus cells that underwent 3h OGD without Ad-TMBIM6 (Vehicle) (Figure 2.1C-D). Furthermore, BI-1 siRNA and APY-29 significantly reversed Ad-TMBIM6s protective effects (Figure

2.1D). Activation of PERK with CCT020312 decreased the percentage of viable cells compared to Ad-TMBIM6. However, double inhibition of PERK with Ad-

TMBIM6+PERK CRISPR yielded a low cell viability count (~40%), which may suggest that PERK may not play a major role in BI-1’s protective effects. On the contrary, double inhibition of IRE1α with Ad-TMBIM6+STF-083010 improved cell viability to ~85% versus ~40% in Vehicle and the PERK CRISPR group (Figure 2.1D). This data indicated that manipulation of IRE1α may play a greater role than PERK signaling in BI-1’s protective effects.

In support of our cell viability data, western blot results indicated that Ad-

TMBIM6 or double inhibition of IRE1α with Ad-TMBIM6+ STF-083010, successfully upregulated BI-1’s expression levels in PC12 cells and inhibited pIRE1α (Figure 2.2A-

C). However, inhibiting BI-1 with siRNA and activating IRE1α with APY-29, significantly reversed those effects as seen from the decreased expression level of BI-1

(Figure 2.2B). Activation of IRE1α with APY-29 significantly reduced BI-1 and increased pIRE1α, however, either activation of PERK with CCT020312 or inhibition with PERK CRISPR did not alter BI-1’s nor pIRE1α expression levels (Figure 2.2B-C), indicating that BI-1 is more influenced by IRE1α as opposed to PERK. While, APY-29 significantly increased the pro-apoptotic marker, CC3, CCT020312 did not show significant CC3 upregulation (Figure 2.2E). Furthermore, double inhibition of IRE1α with Ad-TMBIM6+STF-080310 further reduced CC3 versus double inhibition of PERK with Ad-TMBIM6+PERK CRISPR, indicating that BI-1 may more potently attenuate apoptosis via inhibition of IRE1α.

Summary of our in vitro findings showed that either activating or inhibiting

PERK did not change BI-1’s expression levels versus Ad-TMBIM6, whereas manipulation of the IRE1α receptor showed changes in BI-1. In addition, Ad-TMBIM6 successfully attenuated pIRE1α expression while inhibition of BI-1 reversed those effects

(Figure 2.2C). Based on these findings, we can conclude that despite BI-1’s inhibitory effects on pPERK expression (Figure 2.2D), the effect isn’t prominent enough to confer neuroprotection as seen from the CC3 levels (Figure 2.2E). Taken all the above together, we believe that although BI-1 may have some inhibitory effects on PERK, its major pathway operates via inhibition of IRE1α signaling.

We then hypothesized that intracerebroventricular injection of Ad-TMBIM6 vector would overexpress BI-1 and provide protection in an in vivo neonatal HI model via inhibition of IRE1α singling. This hypothesis stems from our in vitro data as well as from previous reports on BI-1’s anti-apoptotic effects via inhibition of IRE1α, causing disruption of the receptors interaction with downstream pro-apoptotic molecules (XBP1 and CHOP) (Bailly-Maitre et al. , 2006; Lisbona et al. , 2009; Bailly-Maitre et al. ,

2010).

Expression of BI-1 may begin within 24-72h of injection, therefore we administered Ad-TMBIM6 vector at multiple time points and sacrificed animals at 72h post HI to determine the optimal time needed for the virus to replicate and begin overexpressing BI-1(Bailly-Maitre et al. , 2010). Our results indicated that administration of Ad-TMBIM6 vector at 48h pre-HI significantly reduced the percentage of infarcted area at 72h post HI and was chosen as the optimal time (Figure 2.3F). In addition, a low

dose viral vector (1.6x1011 PFU) was more protective then the higher dosage (1.7x1011

PFU) (Figure 2.3G).

To evaluate whether Ad-TMBIM6 attenuated neuronal apoptosis we performed

Flouro-Jade C (FJC) staining, which stains for degenerating neurons (Figure 2.8F). Ad-

TMBIM6 vector reduced the number of FJC positively stained neurons, indicating attenuation of apoptosis. To further confirm BI-1’s anti-apoptotic effects, we stained for

CC3 on neurons and saw that Ad-TMBIM6 was able to attenuate the expression of CC3 as well as reduce co-localization with neurons which is indicative of less neuronal death

(Figure 2.8E).

To determine a delay in acquisition of neurobehavioral milestones and the effects of Ad-TMBIM6 treatment post HI, neurocognitive outcomes were measured at 4 weeks using the Morris water maze, rotarod and foot fault tests. Results showed that Ad-

TMBIM6 significantly improved motor and learning deficits (Figure 2.5A-D).

Furthermore, brain weight measurement and Nissl staining showed a significantly lower percentage of tissue loss in the Ad-TMBIM6 group compared to vehicle (Figure 2.5E-F), which is representative of our behavior findings.

Finally, to determine that BI-1 confers protection via inhibition of IRE1α, we used a BI-1 siRNA to inhibit BI-1 and an IRE1α CRISPR activation plasmid (AP) to counteract BI-1’s effects as depicted in the schematic in figure 2.6C. From our WB results, we demonstrated that inhibition of BI-1 with siRNA, or activation of IRE1α with

CRISPR AP, significantly reduced BI-1’s expression levels (Figure 2.7A-B) and increased pro-apoptotic markers (Figure 2.8 C-D). On the contrary, administration of Ad-

TMBIM6 significantly increased BI-1 levels in the brain, attenuated pIRE1α, XBP1

CHOP and reduced pro-apoptotic markers via inhibition of IRE1α (Figure 2.7 and 2.8).

To determine siRNA’s effects on different cell types we stained for neurons, astrocytes and microglia. Results demonstrated a reduction in BI-1 expression in all three cell types after administration of siRNA (Figure 2.7G) versus Ad-TMBIM6 treated groups (Figure

2.4A, C).

The ER dictates the fate of proteins, while the level of redox mediators modulates the level of reactive oxygen species (ROS) (Cao et al. , 2014; Zeeshan et al. , 2016).

Persistent ER stress causes accumulation of misfolded proteins, which then triggers

IRE1α activation and further upregulation of CHOP. Increased levels of CHOP further exacerbate cell death signaling through oxidative stress, which also contributes to ROS generation (Cao et al. , 2014; Zeeshan et al. , 2016). Herein, we used a ROS marker,

Reactive-Oxygen-Species-Modulator-1 (ROMO1), which is protein coding gene responsible for increasing ROS in cells. Western blot data demonstrated a significant reduction in ROMO1 levels in the brain after Ad-TMBIM6 (Figure 2.7F). These results were confirmed to be via BI-,1 as either BI-1 inhibition with siRNA, or over activation of

IRE1α with CRISPR AP in the presence of Ad-TMBIM6, reversed those effects.

In summary, overexpression of BI-1 attenuated apoptosis and improved neurological outcomes post HI. The main novelty lies in successfully identifying BI-1’s protective role and signaling pathway in ER stress induced apoptosis post neonatal HI.

This work is essential as it provides critical information that may be used to develop more beneficial clinical treatments for HI patients, as well as provide a foundation for future research in other diseases with similar pathologies.

References

Agani, F. H., M. Puchowicz, J. C. Chavez, P. Pichiule and J. LaManna (2002). "Role of

nitric oxide in the regulation of HIF-1alpha expression during hypoxia." Am J

Physiol Cell Physiol 283(1): C178-186.

Bailly-Maitre, B., B. F. Belgardt, S. D. Jordan, B. Coornaert, M. J. von Freyend, A.

Kleinridders, et al. (2010). "Hepatic Bax Inhibitor-1 Inhibits IRE1 alpha and

Protects from Obesity-associated Insulin Resistance and Glucose Intolerance."

Journal of Biological Chemistry 285(9): 6198-6207.

Bailly-Maitre, B., B. F. Belgardt, S. D. Jordan, B. Coornaert, M. J. von Freyend, A.

Kleinridders, et al. (2010). "Hepatic Bax inhibitor-1 inhibits IRE1alpha and

protects from obesity-associated insulin resistance and glucose intolerance." J

Biol Chem 285(9): 6198-6207.

Bailly-Maitre, B., C. Fondevila, F. Kaldas, N. Droin, F. Luciano, J. E. Ricci, et al.

(2006). "Cytoprotective gene bi-1 is required for intrinsic protection from

endoplasmic reticulum stress and ischemia-reperfusion injury." Proc Natl Acad

Sci U S A 103(8): 2809-2814.

Barrett, R. D., L. Bennet, J. Davidson, J. M. Dean, S. George, B. S. Emerald, et al.

(2007). "Destruction and reconstruction: hypoxia and the developing brain." Birth

Defects Res C Embryo Today 81(3): 163-176.

Bartha, A. I., J. Shen, K. H. Katz, R. E. Mischel, K. R. Yap, J. A. Ivacko, et al. (2007).

"Neonatal seizures: multicenter variability in current treatment practices." Pediatr

Neurol 37(2): 85-90.

Bouet, V., T. Freret, S. Ankri, M. Bezault, S. Renolleau, M. Boulouard, et al. (2010).

"Predicting sensorimotor and memory deficits after neonatal ischemic stroke with

reperfusion in the rat." Behav Brain Res 212(1): 56-63.

Bracewell, M. and N. Marlow (2002). "Patterns of motor disability in very preterm

children." Ment Retard Dev Disabil Res Rev 8(4): 241-248.

Cao, S. S. and R. J. Kaufman (2014). "Endoplasmic reticulum stress and oxidative stress

in cell fate decision and human disease." Antioxid Redox Signal 21(3): 396-413.

Castillo, K., D. Rojas-Rivera, F. Lisbona, B. Caballero, M. Nassif, F. A. Court, et al.

(2011). "BAX inhibitor-1 regulates autophagy by controlling the IRE1alpha

branch of the unfolded protein response." EMBO J 30(21): 4465-4478.

Chae, H. J., H. R. Kim, C. Xu, B. Bailly-Maitre, M. Krajewska, S. Krajewski, et al.

(2004). "BI-1 regulates an apoptosis pathway linked to endoplasmic reticulum

stress." Mol Cell 15(3): 355-366.

Chaudhari, N., P. Talwar, A. Parimisetty, C. Lefebvre d'Hellencourt and P. Ravanan

(2014). "A molecular web: endoplasmic reticulum stress, inflammation, and

oxidative stress." Front Cell Neurosci 8: 213.

Dickson, G., H. Prentice, J. P. Julien, G. Ferrari, A. Leon and F. S. Walsh (1986). "Nerve

growth factor activates Thy-1 and neurofilament gene transcription in rat PC12

cells." EMBO J 5(13): 3449-3453.

Dominguez, E., C. Rivat, B. Pommier, A. Mauborgne and M. Pohl (2008). "JAK/STAT3

pathway is activated in spinal cord microglia after peripheral nerve injury and

contributes to neuropathic pain development in rat." J Neurochem 107(1): 50-60.

Ferriero, D. M. (2004). "Neonatal brain injury." N Engl J Med 351(19): 1985-1995.

Gardner, B. M. and P. Walter (2011). "Unfolded proteins are Ire1-activating ligands that

directly induce the unfolded protein response." Science 333(6051): 1891-1894.

Hartman, M. and L. H. Warren (2005). "Explaining age differences in temporal working

memory." Psychol Aging 20(4): 645-656.

Iwata, A., R. A. de Claro, V. L. Morgan-Stevenson, J. C. Tupper, B. R. Schwartz, L. Liu,

et al. (2011). "Extracellular administration of BCL2 protein reduces apoptosis and

improves survival in a murine model of sepsis." PLoS One 6(2): e14729.

Jacobs, S. E., M. Berg, R. Hunt, W. O. Tarnow-Mordi, T. E. Inder and P. G. Davis

(2013). "Cooling for newborns with hypoxic ischaemic encephalopathy."

Cochrane Database Syst Rev 1: CD003311.

Kaser, A., A. H. Lee, A. Franke, J. N. Glickman, S. Zeissig, H. Tilg, et al. (2008). "XBP1

links ER stress to intestinal inflammation and confers genetic risk for human

inflammatory bowel disease." Cell 134(5): 743-756.

Krajewska, M., L. Xu, W. Xu, S. Krajewski, C. L. Kress, J. Cui, et al. (2011).

"Endoplasmic reticulum protein BI-1 modulates unfolded protein response

signaling and protects against stroke and traumatic brain injury." Brain Res 1370:

227-237.

Lee, G. H., H. K. Kim, S. W. Chae, D. S. Kim, K. C. Ha, M. Cuddy, et al. (2007). "Bax

inhibitor-1 regulates endoplasmic reticulum stress-associated reactive oxygen

species and heme oxygenase-1 expression." J Biol Chem 282(30): 21618-21628.

Li, B., R. K. Yadav, G. S. Jeong, H. R. Kim and H. J. Chae (2014). "The characteristics

of Bax inhibitor-1 and its related diseases." Curr Mol Med 14(5): 603-615.

Li, L., N. H. Khatibi, Q. Hu, J. Yan, C. Chen, J. Han, et al. (2012). "Transmembrane

protein 166 regulates autophagic and apoptotic activities following focal cerebral

ischemic injury in rats." Exp Neurol 234(1): 181-190.

Li, L., D. Klebe, D. Doycheva, D. W. McBride, P. R. Krafft, J. Flores, et al. (2015). "G-

CSF ameliorates neuronal apoptosis through GSK-3beta inhibition in neonatal

hypoxia-ischemia in rats." Exp Neurol 263: 141-149.

Lisbona, F., D. Rojas-Rivera, P. Thielen, S. Zamorano, D. Todd, F. Martinon, et al.

(2009). "BAX inhibitor-1 is a negative regulator of the ER stress sensor

IRE1alpha." Mol Cell 33(6): 679-691.

Ma, Y., X. Xia, J. M. Cheng, Y. Q. Kuang, T. Yang, L. B. Yang, et al. (2014). "Emodin

inhibits inducible nitric oxide synthase in a rat model of craniocerebral explosive

injury." Neurochem Res 39(9): 1809-1816.

Madeo, F. and G. Kroemer (2009). "Intricate links between ER stress and apoptosis."

Mol Cell 33(6): 669-670.

Robinson, K. S., A. Clements, A. C. Williams, C. N. Berger and G. Frankel (2011). "Bax

inhibitor 1 in apoptosis and disease." Oncogene 30(21): 2391-2400.

Rojas-Rivera, D. and C. Hetz (2015). "TMBIM protein family: ancestral regulators of

cell death." Oncogene 34(3): 269-280.

Sano, R. and J. C. Reed (2013). "ER stress-induced cell death mechanisms." Biochim

Biophys Acta 1833(12): 3460-3470.

Shi, X., L. Xu, D. M. Doycheva, J. Tang, M. Yan and J. H. Zhang (2017). "Sestrin2, as a

negative feedback regulator of mTOR, provides neuroprotection by activation

AMPK phosphorylation in neonatal hypoxic-ischemic encephalopathy in rat

pups." J Cereb Blood Flow Metab 37(4): 1447-1460.

Souvenir, R., N. Fathali, R. P. Ostrowski, T. Lekic, J. H. Zhang and J. Tang (2011).

"Tissue inhibitor of matrix metalloproteinase-1 mediates erythropoietin-induced

neuroprotection in hypoxia ischemia." Neurobiol Dis 44(1): 28-37.

Souvenir, R., J. J. Flores, R. P. Ostrowski, A. Manaenko, K. Duris and J. Tang (2014).

"Erythropoietin inhibits HIF-1alpha expression via upregulation of PHD-2

transcription and translation in an in vitro model of hypoxia-ischemia." Transl

Stroke Res 5(1): 118-127.

Tabas, I. and D. Ron (2011). "Integrating the mechanisms of apoptosis induced by

endoplasmic reticulum stress." Nat Cell Biol 13(3): 184-190. van Handel, M., H. Swaab, L. S. de Vries and M. J. Jongmans (2007). "Long-term

cognitive and behavioral consequences of neonatal encephalopathy following

perinatal asphyxia: a review." Eur J Pediatr 166(7): 645-654.

Vannucci, R. C. (2000). "Hypoxic-ischemic encephalopathy." Am J Perinatol 17(3): 113-

120.

Vannucci, R. C. and S. J. Vannucci (1997). "A model of perinatal hypoxic-ischemic brain

damage." Ann N Y Acad Sci 835: 234-249.

Vannucci, S. J. and H. Hagberg (2004). "Hypoxia-ischemia in the immature brain." J Exp

Biol 207(Pt 18): 3149-3154.

Volpe, J. J. (2001). "Perinatal brain injury: from pathogenesis to neuroprotection." Ment

Retard Dev Disabil Res Rev 7(1): 56-64.

Watanabe, N. and E. Lam (2009). "Bax Inhibitor-1, a conserved cell death suppressor, is

a key molecular switch downstream from a variety of biotic and abiotic stress

signals in plants." Int J Mol Sci 10(7): 3149-3167.

Wirrell, E. C., E. A. Armstrong, L. D. Osman and J. Y. Yager (2001). "Prolonged

seizures exacerbate perinatal hypoxic-ischemic brain damage." Pediatr Res 50(4):

445-454.

Xie, Z., L. Huang, B. Enkhjargal, C. Reis, W. Wan, J. Tang, et al. (2017). "Intranasal

administration of recombinant Netrin-1 attenuates neuronal apoptosis by

activating DCC/APPL-1/AKT signaling pathway after subarachnoid hemorrhage

in rats." Neuropharmacology 119: 123-133.

Xu, C., B. Bailly-Maitre and J. C. Reed (2005). "Endoplasmic reticulum stress: cell life

and death decisions." J Clin Invest 115(10): 2656-2664.

Xu, Q. and J. C. Reed (1998). "Bax inhibitor-1, a mammalian apoptosis suppressor

identified by functional screening in yeast." Mol Cell 1(3): 337-346.

Yang, L., D. Zhao, J. Ren and J. Yang (2015). "Endoplasmic reticulum stress and protein

quality control in diabetic cardiomyopathy." Biochim Biophys Acta 1852(2): 209-

218.

Zanelli, S., M. Buck and K. Fairchild (2011). "Physiologic and pharmacologic

considerations for hypothermia therapy in neonates." J Perinatol 31(6): 377-386.

Zeeshan, H. M., G. H. Lee, H. R. Kim and H. J. Chae (2016). "Endoplasmic Reticulum

Stress and Associated ROS." Int J Mol Sci 17(3): 327.

Zhang, H. Y., Z. G. Wang, X. H. Lu, X. X. Kong, F. Z. Wu, L. Lin, et al. (2014).

"Endoplasmic Reticulum Stress: Relevance and Therapeutics in Central Nervous

System Diseases." Mol Neurobiol.

Zhou, Y., N. Fathali, T. Lekic, J. Tang and J. H. Zhang (2009). "Glibenclamide improves

neurological function in neonatal hypoxia-ischemia in rats." Brain Res 1270: 131-

139.

CHAPTER THREE

TMBIM6 GENE PROTECTS THE NEONATAL BRAIN FROM OXIDATIVE

STRESS AND INFLAMMATION VIA DISRUPTION OF NPR-CYP COMPLEX

COUPLED WITH UPREGULATION OF NRF-2 IN A HI RAT MODEL

Desislava Doycheva, PhD1; Ningbo Xu, PhD2; Jiping Tang, MD1; John Zhang, MD,

PhD1,2 *

1Department of Physiology and Pharmacology, Basic Sciences, School of Medicine,

Loma Linda University, Loma Linda, CA 92354, USA

2Departments of Anesthesiology, Neurosurgery and Neurology, Loma Linda University

School of Medicine, Loma Linda, CA 92354, USA

Abstract

Oxidative stress, inflammation and Endoplasmic Reticulum (ER) stress play a major role in the pathogenesis of hypoxic ischemic (HI) injury in the neonate. ER stress results in accumulation of unfolded proteins that trigger the NADPH-P450 reductase

(NPR) and the microsomal monooxygenase system which is composed of cytochrome

P450 members (CYP) generating reactive oxygen species (ROS) as well as the release of inflammatory cytokines.

We explored the role of Bax Inhibitor-1 (BI-1) protein encoded by the

Transmembrane Bax inhibitor Motif Containing 6 (TMBIM6) gene, in protection from

ER stress after HI brain injury. BI-1 may regulate ER stress induced ROS production and release of inflammatory mediators via two main mechanisms. Firstly, BI-1 can directly bind to NPR, thus disrupting the NPR-CYP complex; and secondly, ROS production is also controlled by the redox sensitive transcription factor, Nrf-2, which controls expression of different anti-oxidant enzymes such as HO-1. BI-1 can upregulate Nrf-2, promoting an increase in anti-oxidant enzymes, thus reducing ROS. Therefore, in this study we evaluated BI-1’s inhibitory effects on ROS and inflammation by overexpressing

BI-1 in 10-day old unsexed Sprague-Dawley rat pups that underwent right carotid ligation, followed by 1.5h of hypoxia using adenoviral mediated gene delivery of

TMBIM6.

Overexpression of BI-1 significantly reduced the percentage of infarcted area; while silencing BI-1 reversed those effects. BI-1’s mediated protection was observed to be via inhibition of P4502E1, a major contributor to ROS generation, as seen from the decreased levels of P4502E1 and upregulation of Nrf-2 and HO-1, which correlated with

a decrease in ROS and inflammatory markers. This effect was reversed with BI-1 or Nrf-

2 silencing. This study showed a novel role for BI-1 to inhibit a major regulator of ROS production. Inhibition of P4502E1, together with activation of Nrf-2, demonstrates that

BI-1 has a great promise as a therapeutic target.

Introduction

Hypoxia Ischemia (HI) in the perinatal period is associated with long-term disabilities in children affecting 1-4 infants per 1000 births (Vannucci 2000). The most common cause of HI is intrauterine asphyxia, which may be brought on by several factors such as placental artery clotting, abruption or inflammatory processes (Fatemi et al. ,

2009). These result in diminished supply of blood and oxygen to the neonatal brain. In the event of prolonged abruption and extended period of HI the neonate develops hypoxic ischemic encephalopathy (HIE), causing irreversible brain injury (Fatemi et al. , 2009).

Despite the significant research progress identifying many molecular mechanisms for neonatal HIE, therapeutic interventions are limited (Ferriero 2004; Gunn et al. , 2008;

Shankaran 2008).

One of the main contributing injury mechanism post HI is the disruption of correct protein folding, that subsequently triggers reactive oxygen species (ROS) accumulation, followed by activation of microglia, ultimately causing inflammation

(Ferriero 2004; Khwaja et al. , 2008). Inflammatory cytokine release has been associated with HIE and has been shown to be significantly elevated in the full-term infant, which is linked with further exacerbating the damage and results in poor neurodevelopmental outcome (Fatemi et al. , 2009). Therefore, inhibition of inflammation post HI is an attractive target for new therapeutic strategies.

Correct folding of transmembrane proteins takes place in the endoplasmic reticulum (ER) which is composed of an elaborate system of chaperones and enzymes to aid in this process(Stahl et al. , 2013). However, under stressful conditions, such as after a hypoxic ischemic event, the number of proteins to be folded exceeds the capacity of the

chaperones, leading to an accumulation of unfolded proteins, causing ER stress (Sano et al. , 2013). The ER responds to this by activating the unfolded protein response (UPR), which activates sensor proteins which recognize and ameliorate ER stress (Xu et al. ,

2005; Sano et al. , 2013; Zhang et al. , 2014). However, persisting ER stress leads to the overactivation of the UPR, a highly redox dependent pathway, which is followed by an accumulation of reactive oxygen species (ROS) (Kim et al. , 2009). ROS are a natural byproduct of signaling pathways, however during stress, ROS production is increased significantly, which causes oxidative stress and damage to the cell (Kim et al. , 2009).

A major source of ROS production at the ER is from the microsomal monooxygenase (MMO) system which is composed of cytochrome P450 (CYP),

NADPH-P450 reductase (NPR) and some phospholipids (Kim et al. , 2009; Henke et al. ,

2011; Robinson et al. , 2011). Specifically, cytochrome P4502E1, a member of cytochrome P450, is associated with the production of large amounts of ROS. This has been shown to occur during leakage of electron transfer between P4502E1 and NPR, thus, indicating an important role of these cytochromes during ER stress (Kim et al. ,

2009; Robinson et al. , 2011).

BI-1 has been shown to regulate ER stress induced ROS production via two important mechanisms (Li et al. , 2014). Firstly, BI-1 can directly inhibit ROS production by disrupting the NPR-CYP complex, which is a major contributor to ROS generation.

BI-1 causes dissociation of the complex by altering the electron flow from P4502E1 to

NPR, causing destabilization of this complex, blocking electron transfer and attenuating

ROS accumulation (Lee et al. , 2007; Kim et al. , 2009; Henke et al. , 2011). The

P4502E1 member was found to be significantly reduced and its upregulation attenuated

after ER stress in BI-1 overexpressing cells (Kim et al. , 2009). Secondly, BI-1 can directly increase the production of anti-oxidant factors such as nuclear factor erythroid 2- related factor 2 (Nrf-2), a transcription factor the regulates ROS production (Lee et al. ,

2007; Li et al. , 2014). It does so by stimulating the production of the anti-oxidant enzyme, heme oxygenase-1 (HO-1), which in turn blocks ROS, thereby attenuating inflammation and promoting cell survival. G. H. Lee showed that cells overexpressing

BI-1 increased Nrf-2 activation and HO-1 expression thus blocking ROS. BI-1’s protective effects against ER stress induced cell death were abrogated after inhibition of

HO-1 (Lee et al. , 2007).

The specific objective of this study was to establish that overexpression of BI-1 protein, with Adenoviral-TMBIM6 vector, can attenuate the morphological and neurological consequences post neonatal HI via attenuation of ER stress induced ROS production and inflammation. We hypothesized that BI-1 will reduce ER stress induced

ROS accumulation and subsequent inflammation via coupling with NPR and hence disrupting the NPR-CYP complex. In addition, BI-1 can upregulate Nrf-2 which leads to an increase in HO-1 production. HO-1 is a rate-limiting enzyme shown to be able to attenuate ROS accumulation. Increased levels of HO-1 may limit oxidative dysregulation which causes misfolding of ER proteins thereby decreasing the unfolded protein response

(UPR).

With this study we hope to establish a novel role for BI-1 protein in regulating ER stress induced ROS production after neonatal HI.

Materials and Methods

In vivo Experiments

All protocols are approved by The Institutional Animal Care and Use Committee of

Loma Linda University and with NIH guide for The Care and Use of Laboratory

Animals. The animals were cared for and all studies conducted in accordance with the

United States Public Health Service's Policy on Humane Care and Use of Laboratory

Animals. Sprague Dawley rat mothers, with litters of 10~12 pups, were purchased from

Harlan Labs (Livermore, CA). The pups were kept in temperature-controlled environment with regular light/dark cycle until they were ready for surgery at 10days old post birth (P10), to allow them enough time to adjust.

A total of 100 unsexed rat pups weighing 15-20g were assigned to the following groups: sham, HI+Vehicle, HI+Ad-TMBIM6, HI+Ad-TMBIM6+BI-1siRNA, HI+Ad-

TMBIM6+Nrf-2siRNA, HI+Ad-TMBIM6+scramble siRNA, HI+BI-1siRNA, HI+Nrf-

2siRNA and HI+scramble siRNA.

Hypoxic Ischemic Rat Model

Hypoxic Ischemic (HI) injury was induced as previously described following the well-established Rice Vannucci model (Vannucci et al. , 1997; Vannucci et al. , 2004).

Briefly, P10 unsexed rats were anesthetized with 3% isoflurane gas in air and maintained at 2.5% isoflurane during surgery. Once the rat was fully anesthetized and unresponsive, the rats neck was prepared and draped using standard sterile techniques. A small midline neck incision on the anterior neck was made with a No. 11 blade surgical knife

(approximately 3-5 mm in length). The right carotid artery was dissected and isolated

100

from surrounding structures. The carotid artery was then double ligated with 5-O surgical suture and cut between the ligations. Animal’s skin was sutured back to close the incision. Throughout the surgical and postoperative period, temperature was controlled with heating blankets and incubators. Rats were then allowed to recover for

1h on a heated blanket, then placed in a 2L Erlenmeyer airtight flask, which was partially submerged in a 37 °C water bath to maintain a constant thermal environment. Rat pups were exposed to a gas mixture of 8% Oxygen and 92% Nitrogen for 90min. Thereafter, the animals were returned to their mothers and monitored daily.

Time Course Evaluation of Proteins

Time course expression of endogenous BI-1, P4502E1, NPR, pNrf-2 and HO-1 levels were measured at 6h, 12h, 24h and 72h post HI. Rats were randomly divided into the groups and brains were collected for western blotting. Sham animals underwent surgery; however, the artery was only isolated without being ligated or cut, and pups were euthanized at 24h post HI.

Viral Administration

Human adenoviral-TMBIM6 vector (Ad-TMBIM6) (Vector Biolabs) was injected intracerebroventricularly (icv) at 2µl containing 1.6x1011 PFU/ml per injection (Bailly-

Maitre et al. , 2010) at 48h pre-HI. The best dose and time of injection of virus was determined in our preliminary experiments (data not shown).

101

RNAi Administration

Rats were anesthetized and placed on a stereotactic frame. 2µl of BI-1 siRNA

(300pmol/µl, Sigma-Aldrich), Nrf-2 siRNA or scramble siRNA (300pmol/µl, Santa Cruz) were administered icv using a Hamilton syringe (10µl, Hamilton Co) into the right lateral ventricle (1.5mm posterior, 1.5mm lateral to bregma and 1.7mm down from surface of brain) at 48h pre-HI at a rate of 0.3µl/min (Shi et al. , 2017). To prevent backflow, needle was left in place for 10min after administration was completed and was then withdrawn slowly over 5min. The hole was sealed with bone wax and the skin was sutured.

Infarct Area Measurements

TTC (2,3,5-Triphenyltetrazolium Chloride Monohydrate) staining was performed at 72h post HI to determine percentage of infarcted area. Animals were anesthetized, the brains were removed and sectioned into 2mm slices. A total of 5-6 slices were cut per brain and were then immersed in 2% TTC solution until brains turned pink-red (~5min) at room temperature (Zhou et al. , 2009; Li et al. , 2012). Slices were then washed in PBS and fixed overnight in 10% formaldehyde solution and imaged. Red stained tissue indicated live tissue while white indicated the infarcted tissue. Image J software was used to calculate the percentage of infarcted area.

The area of each slice was calculated using the formula: ((Area of Contralateral

Hemisphere – Area of non-infarcted ipsilateral hemisphere) / 2* (Area of Contralateral

Hemisphere)) * 100 (Chen et al. , 2018; Xu et al. , 2018). The average of all 5 slices was taken as a representative of percentage of infarcted area for each animal. All experiments were performed in an unbiased blinded fashion.

102

Western Blotting for Brain Tissue

Western blotting was performed as previously described (Li et al. , 2015; Shi et al. , 2017; Xie et al. , 2017). Rats were anesthetized at 72h post HI and transcardially perfused with 100ml ice cold PBS (pH 7.4). The brain was isolated and divided into contralateral and ipsilateral hemispheres, then immediately snap frozen in liquid nitrogen and stored in -80°C for further use. To prepare samples for western blotting, the ipsilateral hemispheres were homogenized in RIPA lysis buffer (Santa Cruz

Biotechnology) for 15min followed by centrifugation at 14,000g at 4°C for 30min. The supernatant was collected, and protein concentration was measured using a detergent compatibility assay (Catalog number: 5000112, DC™ Protein Assay Kit II, Bio-Rad,

USA).

Equal amounts of protein (30μg, chosen as best loading amount based on previous studies (Thacker et al. , 2016; Moritz 2017; Shi et al. , 2017) were loaded on an 8%~12%

SDS-PAGE gel and electrophoresed and then transferred to a nitrocellulose membrane

(0.2μm). The membrane was blocked for 1h and then incubated with primary antibodies overnight at 4°C: BI-1 antibody (1:1000, Abcam), anti-P4502E1 (1:2500, Abcam), Anti-

NPR antibody (1:1000, Abcam), anti-pNrf-2/Nrf-2 antibody (1:1000, Abcam), anti-HO1 antibody (1:200, Santa Cruz Biotechnology), anti-ROMO1 antibody (1:1000, Abcam), anti-TNFα antibody (1:1000, Abcam), anti-IL-6 antibody (1:500, Abcam), anti-Il-1β antibody (1:500, Abcam), Anti-HA tag antibody (1:4000, Abcam) and actin (1:4000,

Santa Cruz Biotechnology). The following day, membranes were incubated at room temperature for 1h with horseradish peroxidase-conjugated secondary antibodies (Santa

Cruz Biotechnology) followed by chemiluminescence detection (ECL plus kit,

103

Amersham Bioscience). The densitometric quantification of the bands was performed with image J software and results were expressed as relative density to actin.

Histology

Brains were post-fixed in formalin, and then sectioned into 10μm thick coronal slices by cryostat (CM3050S; Leica Microsystems) for immunohistochemistry and

Dihydroethidium (DHE) staining.

Immunofluorescence staining: Rats were anesthetized at 72h post HI and transcardially perfused with PBS and 10% formalin. Brain sections were then post fixated with formalin overnight and dehydrated in 30% sucrose solution for 3-5d after which brains were frozen in OCT. Immunofluorescence staining was performed as previously described (Shi et al. , 2017). The tissue mounted on the slides was washed with 0.1M

PBS three times for 5min and then incubated in 0.3% Triton X-100 in 0.1M PBS for

30min at room temperature. The tissue was then washed once more by 0.1M PBS for

5min, 3 times, and primary antibodies were applied overnight at 4°C: anti-BI-1 antibody

(1:100, Abcam), anti-P4502E1 antibody (1:100, Abcam), anti-NPR (1:100, Abcam), anti-

Nrf-2 (1:100, Abcam), anti-HA tag (1:50, Abcam), anti-Iba-1 (1:100, Abcam), anti-Il-1β

(1:100, Abcam) and anti-MPO (1:100, Abcam). After washing with PBS, sections were incubated with appropriate secondary antibodies: anti-rabbit IgG-TR, anti-mouse IgG-

FITC, anti-goat IgG-FITC, anti-rabbit IgG-FITC (1:200) for 2h @ RT. Finally, slides were mounted using Vectashield Antifade with DAPI (Catalog Number: H-1200, Vector

Laboratories Inc., USA) and visualized under Fluorescent microscope (Leica DMi8,

Leica Microsystems) and Magna Fire SP system (Olympus) was used to analyze

104

microphotographs. Quantification of Il-1β and MPO positive cells was manually counted in the peri-ischemic regions. A total of 6 sections/brain were averaged and expressed as the ratio of positive cells to total cells (%).

DHE staining: To evaluate ROS production we stained brain slices with

Dihydroethidium (DHE) stain (Owusu-Ansah et al. , 2008). Slides were prepared as described above in Immunofluorescence staining methods. To prepare a stock solution, a total of 25g of Dihydroethidium (DHE) (Invitrogen) solid was diluted with 1ml methanol.

The working reagent was prepared by taking 25.3µl of stock solution and diluting it in

1ml methanol. After washing slides in PBS 1x10min, a working reagent of DHE was applied to each brain section and slides were let stand in a dark room for 30min. Slides were then washed in PBS 3x10min, dried off, then mounted using DAPI and coverslip. A

Leica DMi8 fluorescent microscope was used to visualize the slides.

In Vitro Experiments

Rat microglial immortalized cell line, HAPI (Millipore Sigma), was used at passage 6 through 9. Cells were grown as previously described (Dickson et al. , 1986).

Oxygen Glucose Deprivation Model

Cells were grown in full growth media which was replaced with glucose deprived media prior to being placed in a chamber. Cells were placed in a hypoxic chamber and flushed with 1% Oxygen for 3h. Exposure of cells to 3h of OGD was chosen to be best time from our preliminary results (data not shown). Media was removed, and complete growth media was re-introduced to the cells. Cells were left in an incubator to recover for

105

18h, after which, they were prepared for cell death assay and western blotting (Agani et al. , 2002; Souvenir et al. , 2014).

Cell Death Assay

To determine the percentage of viable cells, we used trypan exclusion as previously described (Souvenir et al. , 2014). Briefly, cells were scraped from the plates, centrifuged for 5min, and then re-suspended in 10ml complete growth media. Equal volumes of cell suspension were added to Trypan blue, cells were then vortexed and let stand for 3min. 10µl from mixture was placed on cell counter slides and slides were read using an automated cell counter using the average of 6 counts (Souvenir et al. , 2011).

Western Blotting for Cells

Cells were collected 18h after OGD and stored for western blotting as previously described (Dominguez et al. , 2008; Souvenir et al. , 2014). Briefly, cells were scraped, centrifuged and then re-suspended in 1ml PBS, after which they were centrifuged one more time at 14000rpm for 5min. The remaining supernatant was removed after centrifugation, leaving only the pellet. RIPA lysis buffer was added with protease inhibitor cocktail and pipetted thoroughly until the pellet was fully dissolved. This suspension was then left on ice for 20min following which it was centrifuged for 30min.

The supernatant was harvested while pellet was discarded. Protein concentration was measured using Bradford assay and SD-PAGE electrophoresis was performed as previously described (Dominguez et al. , 2008). Primary antibodies: BI-1 antibody

(1:200, Abcam), anti-P4502E1 antibody (1:1000, Abcam), anti-Nrf-2 antibody (1:1000,

106

Abcam), anti-TNFα antibody (1:1000, Abcam), anti-IL-6 antibody (1:500, Abcam), anti-

Il-1β antibody (1:500, Abcam) and actin (1-4000, Santa Cruz) were applied to the membrane and left overnight.

Calculation of MOI

The Multiplicity of Infection (MOI) was calculated to determine the total amount of infectious particles needed to infect one cell. MOI was calculated as follows: #cells * desired MOI= total PFU (or Plaque Forming Units) needed; (total PFU needed) /

(PFU/ml) = total ml of virus needed to reach your desired dose. A 100MOI of Ad-

TMBIM6 virus was chosen as the optimal amount needed to infect cells and provide protection from our preliminary results (data not shown).

siRNA Transfection

Cells were allowed to differentiate and reach 80% confluency in poly D-lysine coated

6 well plates prior to transfection. Both BI-1 and Nrf-2 siRNAs were prepared according to manufacturer’s protocol (Sigma-Aldrich); A stock solution of 10µM siRNA was prepared. From the siRNA stock,4µl was taken and mixed with 125µl Opti-MEM. In a separate tube, 7µl liofectamine3000 was mixed with 125µl Opti-MEM. The solution from both tubes was combined, mixed well, and left to sit at room temperature for 15-45min.

Growth media was removed from cells and replaced with 250µl/well siRNA transfection solution as prepared above. An additional 1ml Opti-MEM was added to each well and cells were then placed in an incubator for 5-7h, following which, the media was removed and replaced with normal growth media.

107

Statistical Analysis

Statistical analysis was performed using one-way analysis of variances (ANOVA), followed by Tukey test. All values were presented as mean ± standard deviation (SD), with a P-value less than 5% considered significant. Sample sizes were determined by power of 0.8, α=0.05, and a 20% standard deviation from the preliminary results. The estimated sample size was 6-8/group for in vivo studies and 4/group for cell culture studies. All rats were randomly assigned to each group and all investigators were blinded.

Results

Time Course Expression of Endogenous Proteins post HI

Endogenous expression of BI-1, NPR, P4502E1, pNrf-2 and HO-1 were measured at

6h, 12h, 24h and 72h post HI. BI-1 levels significantly increased at 24h post HI and returned to sham levels at 72h post HI (Figure 3.1A-B). While NPR expression levels showed no changes (Figure 3.1B), P4502E1 significantly increased in a time dependent manner, reaching peak at 72h post HI (Figure 3.1C). Nrf-2 levels significantly decreased from 6h to 72h, reaching significantly lower expression levels at 72h post HI compared to sham (Figure 3.1D). Lastly, HO-1 showed to peak at 24h post HI (Figure 3.1E).

108

Figure 3.1 Endogenous Expression levels of BI-1, NPR, P4502E1, Nrf-2 and HO-1 post HI Representative pictures of Western blot bands (A). BI-1 expression levels significantly increased in a time-dependent manner from 6h to 24h and returned to sham levels at 72h post HI (B). There were no changes in endogenous NPR levels among the groups (C). The expression of P4502E1 increased at 6h and remained elevated, reaching significance at 72h post HI (D). Nrf-2 levels decreased in a time-dependent manner (E). HO-1 levels increased at 24h and 72h post HI (Data expressed as mean +/- SD; *p<0.05 vs sham; # p ＜0.05 vs 6h, @ p < 0.05 vs 24h; n=4, using one-way ANOVA followed by Tukey multiple- comparison post hoc analysis).

109

To Determine Amount of Ad-TMBIM6 Viral Vector Present in the Brain and

its Effects on Infarction at 72h post HI

To determine amount of Ad-TMBIM6 present in the brain at 72h post HI, western blot was performed using anti-HA tag antibody. This specifically measures how much virus was in the brain when administered 48h pre-HI. Quantification data showed that there was significant amount of Ad-TMBIM6 present at 72h post HI versus nothing in sham or vehicle. To further confirm these findings, IHC staining results indicated a large expression of Ad-TMBIM6 in the brain when stained with anti-HA tagged antibody

(Figure 3.2A). There was no staining seen for Ad-TMBIM6 in sham or vehicle (data not shown).

To evaluate the effects of pathway interventions on the percentage of infarcted area, we used a BI-1 inhibitor (BI-1 siRNA) and a Nrf-2 inhibitor (Nrf-2 siRNA). Results showed that Ad-TMBIM6+BI-1siRNA significantly reversed BI-1’s protective effects as did Ad-TMBIM6+Nrf-2siRNA group. This is seen from the significant increase in percent infarcted area when compared to Ad-TMBIM6 only group (Figure 3.2B). Control group, Ad-TMBIM6+scramble siRNA, did not exacerbate the damage when compared to

Ad-TMBIM6 treated group. Furthermore, BI-1 and Nrf-2 siRNA only groups (without treatment) showed similar percent infarcted area as compared to HI+vehicle or

HI+scramble siRNA groups.

110

Figure 3.2 Inhibition of BI-1 or Nrf-2 increased percent infarcted area at 72h Post HI Anti-HA tagged antibody was used to measure amount of viral vector present in the brain at 72h post HI. Western blot data indicated a significant expression of Ad-TMBIM6 in the brain after intracerebroventricular administration (A) Percent infarct area was significantly increased in vehicle, BI-1siRNA and Nrf-2 siRNA groups when compared to Ad-TMBIM6 and the respective control group (B). (Data expressed as mean +/− SD; *p<0.05 vs sham; #p<0.05 vs vehicle; @p<0.05 vs Ad-TMBIM6 or scramble; n=6-8/group using one-way ANOVA followed by Tukey multiple-comparison post hoc analysis).

111

Overexpression of BI-1 Disrupted the NPR-CYP Complex at 72h post HI

To determine BI-1’s role on the NPR-CYP complex the following groups were used: sham, vehicle, Ad-TMBIM6, Ad-TMBIM6+BI-1siRNA, Ad-TMBIM6+Nrf-2 siRNA,

Ad-TMBIM6+scramble siRNA, HI+BI-1siRNA only, HI+Nrf-2siRNA only and

HI+scramble siRNA only groups.

Quantification data showed that BI-1 expression levels were significantly increased in the Ad-TMBIM6 treatment and in the Ad-TMBIM6+scramble siRNA groups, while BI-1 siRNA reversed those effects (Figure 3.3A-B). In addition, all three groups (without Ad-

TMBIM6) BI-1 siRNA, Nrf-2 siRNA and scramble, showed low expression levels of BI-

1 versus Ad-TMBIM6.

Ad-TMBIM6 significantly reduced expression of P4502E1 (Figure 3.3C), while silencing BI-1 reversed those effects, as seen from the elevated expression levels of

P4502E1 (Figure 3.3C).

To observe changes in BI-1 and P4502E1 expression levels, as well as co-localization with microglia, after either Ad-TMBIM6 or after silencing of BI-1 and Nrf-2, immunofluorescence staining was performed. Data showed increased expression of BI-1 and colocalization on microglia in Ad-TMBIM6 treatment group and scramble control group compared to sham or vehicle while BI-1 siRNA reversed those effects (Figure

3.3D). On the contrary, P4502E1 expression levels were elevated in vehicle, BI-1 siRNA and Nrf-2 siRNA groups but not in sham or Ad-TMBIM6 treated groups (Figure 3.3E).

There was also col-localization of P4502E1 on microglia.

112

Figure 3.3 The effects of BI-1 inhibition on the NPR-CYP complex at 72h post HI Representative picture of western blot data (A). Quantification of bands showed that Ad- TMBIM6 as well as control group (Ad-TMBIM6+scramble siRNA) significantly increased BI-1 expression levels in the brain when compared to vehicle or sham, while, BI-1 siRNA and Nrf-2 siRNA significantly reversed those effects. Furthermore, BI-1 and Nrf-2 siRNA only groups also reduced BI-1’s expression levels (B). Data showed that P4502E1 expression levels were significantly elevated in vehicle and BI-1 siRNA groups whereas Ad-TMBIM6 and control group reversed those effects (C) at 72h post HI. (Data expressed as mean +/− SD; *p<0.05 vs sham; #p<0.05 vs vehicle; @p<0.05 vs Ad-TMBIM6 or scramble; n=6/group using one-way ANOVA followed by Tukey multiple-comparison post hoc analysis). Immunofluorescence staining showed a significantly higher expression of BI-1 on microglia in Ad-TMBIM6 treatment and scramble groups compared to sham or vehicle, while BI-1 siRNA reversed those effects (D). Staining showed higher expression and colocalization of P4502E1 on microglia in vehicle, BI-1 siRNA and Nrf-2 siRNA groups compared to sham or Ad-TMBIM6 treated groups (E). (Green was for microglial staining, Red was for BI-1/ or P4502E1 staining and Blue was for DAPI). Merge showed the colocalization of BI-1 or P4502E1 on microglia (D-E). (Scale bar = 50 μm; n=3/group).

113

BI-1 Upregulated pNrf-2 Expression and Induced Anti-Oxidant Enzyme

Production at 72h post HI

To evaluate overexpression of BI-1 on Nrf-2 levels and its ability to upregulate anti- oxidant enzymes needed to block ROS, the following groups were employed: sham, vehicle, Ad-TMBIM6, Ad-TMBIM6+BI-1siRNA, Ad-TMBIM6+Nrf-2 siRNA, Ad-

TMBIM6+scramble siRNA, HI+BI-1siRNA only, HI+Nrf-2siRNA only and HI+scramble siRNA only groups. Quantification data of western blot bands indicated a significantly higher expression of pNrf-2 in Ad-TMBIM6 treated groups when compared to vehicle.

Either silencing BI-1 or Nrf-2 with siRNAs reversed those effects (Figure 3.4A, C). An increase in Nrf-2 levels was followed by an increase in the anti-oxidant enzyme HO-1 in the Ad-TMBIM6 group when compared to vehicle (Figure 3.4A, D). Furthermore, inhibition of BI-1 or Nrf-2 significantly decreased HO-1 levels and its ability to block

ROS.

In addition, to look at expressions of NPR and Nrf-2 and co-localization with microglia in the presence of Ad-TMBIM6 as well as with BI-1 and Nrf-2 inhibition, we performed immunohistochemical staining. Fluorescent staining showed that although

NPR co-localized with microglia, its expression levels were unchanged among groups

(Figure 3.4E). On the contrary, Ad-TMBIM6 increased Nrf-2 expression and colocalization with microglia compared to sham or vehicle, while Nrf-2 siRNA abolished that effect (Figure 3.4F).

114

Figure 3.4 BI-1’s role in the Nrf-2 signaling pathway at 72h post HI Representative picture of western blot data (A). No changes were observed in NPR expression levels among groups (B) Quantification of bands showed that Ad-TMBIM6 as well as control group (Ad-TMBIM6+scramble siRNA) significantly upregulated pNrf-2 levels versus vehicle, BI-1 siRNA and Nrf-2 siRNA groups (C) Ad-TMBIM6 significantly upregulated HO-1 expression levels while BI-1 siRNA and Nrf-2 siRNA reversed those effects (D). (Data expressed as mean +/− SD; *p<0.05 vs sham; #p<0.05 vs vehicle; @p<0.05 vs Ad-TMBIM6 or scramble; n=6/group using one-way ANOVA followed by Tukey multiple-comparison post hoc analysis). Immunofluorescence staining showed minimal changes in NPR expression among groups (E) while Nrf-2 levels were upregulated in Ad-TMBIM6. Those effects were abolished in the Nrf-2 siRNA group (F). (Green was for microglial staining, Red was for NPR / or Nrf- 2 staining and Blue was for DAPI. Merge showed the colocalization of NPR or Nrf-2 on microglia (E-F). (Scale bar = 50μm; n=3/group).

115

BI-1 Overexpression Attenuated ROS Production and Inflammation at

72h Post HI

To evaluate whether over expression of BI-1 with Ad-TMBIM6 could inhibit

ROS production and subsequently attenuate inflammation, western blot was done to quantify ROMO1 (a modulator that induces production of ROS), IL-6, TNFα and IL-1β expression levels. Data showed that Ad-TMBIM6 treatment group significantly reduced

ROMO1 levels (Figure 3.5A-B) and attenuated pro-inflammatory markers, IL-6, TNFα and IL-1β (Figure 3.5C-E). In addition, administration of BI-1 siRNA and Nrf-2 siRNA reversed Ad-TMBIM6’s protective effects as seen from the significantly higher upregulation of pro-inflammatory mediators and ROS.

Since HI results in induction of inflammation, we stained for IL-1β (a pro- apoptotic marker) with microglia. The counting of positively stained cells for IL-1β with microglia showed a higher percentage of IL-1β on microglia in vehicle, BI-1 siRNA and

Nrf-2 siRNA groups compared to sham, Ad-TMBIM6 or scramble group. (Figure 3.5F;

Scale bar 50µm).

To evaluate ROS production, we stained brain slices with Dihydroethidium

(DHE) dye. Cells with high ROS accumulation will have a red fluorescent light. Our data revealed a higher number of red positively stained cells in vehicle and siRNA groups versus sham, Ad-TMBIM6 and control group in cortex region (Figure 3.6A).

Furthermore, we also stained for ROS around the ventricles and observed trends similar to the cortex region (Figure 3.6B). In addition, to determine whether ROS accumulation is correlated with increased inflammation we stained for Myeloperoxidase

(MPO; a marker that shows the inflammatory activity) and detected significantly higher

116

expression of MPO positively stained cells in vehicle and siRNA groups compared to sham or Ad-TMBIM6 treated group (Figure 3.6C).Thus showing both a correlation between an increase in ROS production with an increase in inflammation, as well as BI-

1’s ability to inhibit ROS production and the increase in inflammation independently.

117

Figure 3.5 Overexpression of BI-1 attenuated ROS production and inflammation at 72h post HI Representative picture of western blot data showing bands of the expression levels of ROMO1, IL-6, TNFα and IL-1β (A). Western blot data quantification of bands showed that Ad-TMBIM6 significantly reduced ROMO1 (B), IL-6 (C), TNFα (D) and IL-1β (E) expression levels compared to vehicle, sham, BI-1 siRNA or Nrf-2 siRNA groups. (Data expressed as mean +/− SD; *p<0.05 vs sham; #p<0.05 vs vehicle; @p<0.05 vs Ad- TMBIM6 or scramble; n=6/group using one-way ANOVA followed by Tukey multiple- comparison post hoc analysis). Fluorescent staining detected significantly more positively IL-1β stained microglia in vehicle, BI-1 siRNA and Nrf-2 siRNA groups compared to sham or Ad-TMBIM6 treated groups (F). (Green was for microglial staining, Red was for IL-1β. Merge showed the colocalization of IL-1β on microglia (E-F). Scale bar 50µm; n=3/group).

118

Figure 3.6 Fluorescent staining demonstrated BI-1’s ability to reduce ROS production and attenuate inflammation at 72h post HI DHE staining for ROS demonstrated that Ad-TMBIM6 reduced the amount of positively stained cells when compared to vehicle, BI-1 siRNA or Nrf-2 siRNA groups in the cortex (A). DHE staining around ventricles showed more positively stained ROS cells in vehicle compared to sham or Ad-TMBIM6 group (B). MPO staining for inflammation showed a significant reduction in positively stained MPO cells in Ad-TMBIM6 and sham groups compared to vehicle and siRNA groups (C). (Red was for ROS or MPO staining and Blue was for DAPI. Merge shows the number of positively stained cells (E-F). (For panel A scale bar = 50μm; For panel B scale bar = 100μm; For panel C scale bar = 25 μm; n=3/group).

119

To Investigate BI-1’s Effects on NPR-CYP Complex in an in vitro OGD Model

As BI-1 has been previously reported to disrupt the NPR-CYP complex as well as activate Nrf-2 resulting in inhibition of ROS production and subsequent attenuation of inflammation, we employed an in vitro OGD model using a microglial cell line to test the specificity of BI-1 related attenuation of ROS induced inflammation in microglia. To test

BI-1’s pathway, we administered a BI-1 siRNA to knockdown BI-1 expression levels and a Nrf-2 siRNA to silence Nrf-2 expression. Data showed that administration of Ad-

TMBIM6 vector significantly upregulated BI-1 levels in cells, both in treated and scramble control group, while BI-1 siRNA group reversed those effects, Nrf-2 siRNA did not affect BI-1’s levels (Figure 3.7A-B). Intervening at either BI-1 level or Nrf-2 level with siRNAs significantly reversed BI-1’s protective effects as seen from the increased expression levels of inflammatory markers, IL-6, TNFα and IL-1β when compared to Ad-

TMBIM6 treated group, sham or scramble groups (Figure 3.7C-E). Ad-TMBIM6 inhibited P4502E1, the major source of ROS production, while simultaneously upregulating pNrf-2 levels (Figure 3.7F-G). Silencing BI-1 or Nrf-2 reversed those effects. Percent cell viability data showed that Ad-TMBIM6 was able to significantly improve the percent of viable cells while knockdown of BI-1 and Nrf-2 reversed those effects (Figure 3.7H).

120

121

Figure 3.7 BI-1 exerted its anti-inflammatory effects via inhibition of P4502E1 and upregulation of Nrf-2 in an in vitro OGD model Representative picture of western blot bands showing protein expression levels (A). Quantification data of bands showed that BI-1 expression levels were significantly increased in the Ad-TMBIM6 treatment group and in scramble control group, while BI-1 siRNA reversed that effect (B). Ad-TMBIM6 and scramble control group significantly reduced IL-6 (C), TNFα (D) and IL-1β (E) expression levels. Expression levels of P4502E1 were significantly upregulated in vehicle, BI-1 siRNA and Nrf-2 siRNA groups versus sham, Ad-TMBIM6 treated group or scramble control group (F). Ad-TMBIM6 and scramble control group significantly upregulated pNrf-2 levels which was reversed by BI- 1 siRNA and Nrf-2 siRNA (G). Cell viability data showed a significantly higher percentage of viable cells in Ad-TMBIM6 group and scramble control group versus vehicle, BI-1 siRNA or Nrf-2 siRNA groups (H). (Data represent +/− SD; *p<0.05 vs control; #p<0.05 vs vehicle; @p<0.05 vs Ad-TMBIM6 and scramble; $p<0.05 vs BI-1 siRNA; n=4-5/group using one-way ANOVA followed by Tukey multiple-comparison post hoc analysis).

122

Discussion

The Endoplasmic Reticulum (ER) is a major organelle that has an essential role in multiple cellular processes. One of these major roles is the control of correct protein folding and function (Xu et al. , 2005; Zhang et al. , 2014). A wide variety of external factors may contribute to a disruption in ER homeostasis, such as hypoxia ischemia, oxidative stress, calcium disturbances and inhibition of protein glycosylation, which all cause ER stress (Zhang et al. , 2014). Cells may respond to ER stress in various ways, activating several pathways, including promoting the ability of proteins to refold correctly, inhibiting protein translation, increasing protein degradation, stimulating the transcription of genes and enabling self-repair mechanisms (Zhang et al. , 2014). All these processes are referred to as the unfolded protein response (UPR) which under prolonged activation result in cell death (Xu et al. , 2005).

Bax Inhibitor-1 (BI-1) is an evolutionary conserved protein that resides on the intracellular membranes of the ER (Henke et al. , 2011). BI-1 was first named as Bax

Inhibitor due to its ability to suppress cell death in yeast (Xu et al. , 1998). More recently it is also referred to as TMBIM6 as it is part of the transmembrane Bax Inhibitor-1 containing motif 6 family (Li et al. , 2014). Overexpression of BI-1 has been demonstrated to play protective role in ER stress-induced cell death, and to a lesser extent in other ER related stresses, such as oxidative stress and inflammation.

ER stress is associated with the production of reactive oxygen species (ROS) through oxidative protein folding by the NADPH-P450 reductase (NPR), as well as the microsomal monooxygenase system, which is composed of cytochrome P450 (CYP) members such as P4502E1 (Nieto et al. , 2002; Nieto et al. , 2002; Robinson et al. ,

123

2011), which triggers the UPR. Study have shown that BI-1 overexpressing cells can regulate UPR induction and inhibit ROS accumulation under ER stress (Kim et al. ,

2009), thus making BI-1 a possibly key regulator of ROS inhibition. Cells respond to

ROS by activating genes to encode antioxidative stress enzymes. A key transcription factor activated is nuclear factor erythroid 2-related factor 2 (Nrf-2), which regulates the production of several cytoprotective enzymes, such as heme oxygenase-1 (HO-1). HO-1 is a rate limiting enzyme and a potent inhibitor of ROS production (Lee et al. , 2007; Kim et al. , 2009). Studies show, that in BI-1 overexpressing cells, the inhibition of HO-1 attenuated BI-1 mediated protection against ER stress (Lee et al. , 2007). However, the role of HO-1 in BI-1’s protective mechanism is largely unknow and debatable.

Some studies have shown that HO-1 was unaffected in BI-1 deficient mice embryonic fibroblast cells (Lisbona et al. , 2009). In contrast, BI-1 overexpressed cells showed significant suppression of ROS production in human embryonic kidney cells

(Kim et al. , 2012). Moreover, this was shown to be due to BI-1’s inhibitory effects on

P4502E1 expression(Kim et al. , 2009). The P4502E1 member of the cytochrome P450

ER heme proteins, is a major contributor to ROS production. It acts by metabolizing and activating substrates into more toxic products, thus not only increasing ROS production but also stimulating inflammatory cascades and ultimately worsening HI pathology.

P4502E1 has been shown to be upregulated during ER stress in several studies (Lee et al. , 2007; Kim et al. , 2009; Kim et al. , 2012; Lee et al. , 2012), thus playing an important role in HI pathology.

BI-1 may inhibit ROS accumulation by binding to NPR and disrupting the NPR-

CYP complex by decreasing electron uncoupling. It was found that BI-1 overexpressing

124

cells caused P4502E1 degradation, leading to ER stress suppression, and subsequently causing ROS reduction (Lee et al. , 2012). In addition, other studies showed that

P4502E1 induced oxidative stress played a role in translocation of Nrf-2 to the nucleus followed by upregulation of anti-oxidant gene HO-1. Moreover, increased HO-1 levels may be dependent on P4502E1 (Gong et al. , 2003), perhaps attempting to counteract

P4502E1 effects and act as survival signal. The same study showed that inhibition of

P4502E1 significantly reduced ROS generation in an acute kidney injury model (Wang et al. , 2014). The relationship between P4502E1 and HO-1 needs further research to fully understand the specifics covering the apparent connection. Since P4502E1 is a major regulator of ROS production and that Nrf-2 has a role in the response against P4502E1, inhibition of P4502E1 coupled with activation of Nrf-2 is a promising therapeutic target.

Therefore, in this study we investigated the role of overexpressing BI-1, by administering

Adenoviral-TMBIM6 vector intracerebroventricularly at 48h pre-HI, on ER stress induced accumulation of ROS and subsequent inflammation. We hypothesized that BI-1 may carry this protective effect by interacting with the CYP-NPR complex while also upregulating Nrf-2 levels.

Although we are now beginning to understand the importance of BI-1 in the cell and in human physiology, its function and signaling mechanisms remain unknown. In view of the importance of P4502E1 and Nrf-2, finding a key molecule to regulate both is of great interest. Here we identified BI-1 as one such potential molecule which may protect the cell against ER stress induced ROS production as well as the subsequent increase in inflammatory response via two possible mechanisms. First, it may interact with NPR, thus destabilizing the NPR-CYP complex, which directly inhibits the

125

formation of ROS by reducing the activity of P4502E1 (Kim et al. , 2009). BI-1 may interact though its C-terminus with NPR and to lesser extend with P4502E1 (Kim et al. ,

2009). This interaction induces destabilization of the NPR-CYP complex, thus blocking electron transfer and ROS production (Robinson et al. , 2011). A recent study showed that

BI-1 was able to inhibit ER stress induced ROS production via a reduction in P4502E1 activity and protein expression (Kim et al. , 2009). We observed similar findings from our western blot results; overexpression of BI-1 in the neonatal rat significantly upregulated

BI-1 levels in the brain, while simultaneously reduced P4502E1 expression levels (Figure

3.3A-C). Inhibition of BI-1 with BI-1 siRNA reversed those effects, while inhibition of

Nrf-2 did not seem to significantly change P4502E1 levels, indicating its action as downstream. This data was further supported by our IHC staining that demonstrated across groups that: as BI-1 expression increased, P4502E1 decreased; they have converse expression patterns throughout (Figure 3.3D-E).

Secondly, BI-1 may upregulate Nrf-2 which in turn triggers HO-1 thus inhibiting

ROS production and attenuating inflammation. Lee et al 2007 showed that overexpression of BI-1 in cells increased Nrf-2 transcription factor, which then translocated to the nucleus where it stimulated production of anti-oxidant enzymes such as HO-1. HO-1 is known to block ROS production and accumulation thereby promoting cell survival (Lee et al. , 2007). Similar to Lee et al, we found that BI-1 overexpression, after neonatal HI, significantly upregulated Nrf-2 (Figure 3.4C) and HO-1 (Figure 3.4D) while silencing of either BI-1 or Nrf-2 with siRNAs reversed those effects. In addition, our IHC staining showed changes in Nrf-2 expression patters among groups which correlated with our western blot findings (Figure 3.4F). Furthermore, inhibition of

126

P4502E1 with activation of Nrf-2 and HO-1 was linked with reduction in ROS accumulation and the subsequent release of pro-inflammatory mediators. Here, we used a

ROS marker, Reactive Oxygen Species Modulator 1 (ROMO1), which is protein coding gene responsible for ROS generation. There was a significant reduction in ROMO1 expression levels (Figure 3.5A-B) as well as in IL-6, TNF-α and IL-β (Figure 3.5C-D).

Staining for IL-1β on microglia demonstrated a significantly higher expression and colocalization in vehicle, BI-1 siRNA and Nrf-2 siRNA groups compared to sham or Ad-

TMBIM6 treated group (Figure 3.5F). These results were confirmed to be via BI-1 induced inhibition of P4502E1 or activation of Nrf-2 as either inhibition of BI-1 or Nrf-2 reversed those effects. To further examine BI-1’s inhibitory role on ROS accumulation, we used a Dihydroethidium (DHE) dye was used to detect ROS production. DHE is possibly the most specific and least problematic dye for detection of superoxide radicals

(Owusu-Ansah et al. , 2008). Once cells are stained with DHE, the dye binds with the superoxide anions, thus illuminating a red fluorescent image which is an indication of

ROS presence. Our data indicated a higher amount of ROS accumulation in vehicle, BI-1 siRNA and Nrf-2 siRNA groups versus sham or Ad-TMBIM6 treated groups, in both cortex region and around ventricles, while overexpression of BI-1 showed a reduction in

ROS (Figure 3.6A-B). To detect whether a reduction in ROS was associated with a reduction in inflammation we performed immunofluorescence staining for

Myeloperoxidase (MPO). Similar pattern was observed in the MPO staining where Ad-

TMBIM6 significantly reduced MPO positively stained cells compared to vehicle, BI-1 siRNA or Nrf-2 siRNA groups (Figure 3.6C) thus indicating a correlation between a reduction in ROS and the subsequent attenuation of the inflammatory processes.

127

In summary, overexpression of BI-1 with Ad-TMBIM6, promotes cell survival by increasing production of anti-oxidant enzymes as well as destabilizing the complex responsible for ROS production. This was observed as inhibition either at the BI-1 or

Nrf-2 level; with siRNAs reversed BI-1’s protective effects by altering P4502E1, Nrf-2 and HO-1 expressions. This is indicative of BI-1’s ability to interact with the NPR-CYP complex and induce dissociation, thus disrupting the electron flow, as a physical association between NPR and CYP is required for production of ROS. Furthermore, BI-1 was shown to directly upregulate Nrf-2 thus triggering HO-1 to inhibit ROS. These two mechanisms play a major role in BI-1’s anti-inflammatory effects.

To ensure the above in vivo results were also specific to microglia, we employed an in vitro oxygen glucose deprivation (OGD) model in an immortalized microglial cell line. HAPI cells were transfected with (1) BI-1 siRNA in combination with human adenoviral TMBIM6 (Ad-TMBIM6; gene responsible for expressing BI-1 (Bailly-Maitre et al. , 2010) vector to test the role of the BI-1 and NPR-CYP complex; and (2) Nrf-2 siRNA together with Ad-TMBIM6 to test the involvement of Nrf-2 and HO-1 in BI-1’s effects. Cells were collected after OGD and used for cell viability and western blot analysis. Cell viability data indicated that Ad-TMBIM6 significantly increased the percentage of viable cells when compared to cells that underwent 3h OGD without Ad-

TMBIM6 (Vehicle) (Figure 3.7). Furthermore, BI-1 siRNA or Nrf-2 siRNA significantly reversed Ad-TMBIM6s protective effects (Figure 3.7).

Western Blot experiments further supported our cell viability data (Figure 3.7).

Results indicated that Ad-TMBIM6 successfully upregulated BI-1’s expression levels in

HAPI cells, while concomitantly reducing the pro-inflammatory markers (Figure 3.7).

128

Inhibition either with BI-1 siRNA or Nrf-2 siRNA increased the pro-inflammatory markers. In addition, similar trends were seen in our in vitro model for P4502E1 and Nrf-

2 levels in comparison to our in vivo model. Data showed that while BI-1 overexpressing cells successfully reduced P4502E1 levels while upregulating Nrf-2, either inhibiting BI-

1 or Nrf-2 reversed that trend.

This data showed a strong support of BI-1’s ability to inhibit P4502E1 thus disrupting the complex while upregulating the anti-oxidant factor Nrf-2, hence attenuating ROS production and inflammation after HI. These data expand the current evidence on BI-1 as a protective protein and advance the in vitro studies done by Kim et al and Lee et al to a in vivo stroke model. A major challenge with any design to inhibit either cell death pathways or inflammatory pathways caused by ER stress lies in the existence of the several parallel pathways that may act at the same time. Thus, focusing on only one may not be ideal to preserve cell survival. Based on the evidence, BI-1’s role as both an anti-inflammatory mediator, and as an anti-apoptotic protein with its ability to influence several major ER stress related pathways, makes BI-1 an ideal and promising candidate for potential therapeutic approach.

129

References

Agani, F. H., M. Puchowicz, J. C. Chavez, P. Pichiule and J. LaManna (2002). "Role of

nitric oxide in the regulation of HIF-1alpha expression during hypoxia." Am J

Physiol Cell Physiol 283(1): C178-186.

Bailly-Maitre, B., B. F. Belgardt, S. D. Jordan, B. Coornaert, M. J. von Freyend, A.

Kleinridders, et al. (2010). "Hepatic Bax Inhibitor-1 Inhibits IRE1 alpha and

Protects from Obesity-associated Insulin Resistance and Glucose Intolerance."

Journal of Biological Chemistry 285(9): 6198-6207.

Chen, D., B. J. Dixon, D. M. Doycheva, B. Li, Y. Zhang, Q. Hu, et al. (2018).

"IRE1alpha inhibition decreased TXNIP/NLRP3 inflammasome activation

through miR-17-5p after neonatal hypoxic-ischemic brain injury in rats." J

Neuroinflammation 15(1): 32.

Dickson, G., H. Prentice, J. P. Julien, G. Ferrari, A. Leon and F. S. Walsh (1986). "Nerve

growth factor activates Thy-1 and neurofilament gene transcription in rat PC12

cells." EMBO J 5(13): 3449-3453.

Dominguez, E., C. Rivat, B. Pommier, A. Mauborgne and M. Pohl (2008). "JAK/STAT3

pathway is activated in spinal cord microglia after peripheral nerve injury and

contributes to neuropathic pain development in rat." J Neurochem 107(1): 50-60.

Fatemi, A., M. A. Wilson and M. V. Johnston (2009). "Hypoxic-ischemic

encephalopathy in the term infant." Clin Perinatol 36(4): 835-858, vii.

Ferriero, D. M. (2004). "Neonatal brain injury." N Engl J Med 351(19): 1985-1995.

130

Gong, P., A. I. Cederbaum and N. Nieto (2003). "Increased expression of cytochrome

P450 2E1 induces heme oxygenase-1 through ERK MAPK pathway." J Biol

Chem 278(32): 29693-29700.

Gunn, A. J., T. Hoehn, G. Hansmann, C. Buhrer, G. Simbruner, J. Yager, et al. (2008).

"Hypothermia: an evolving treatment for neonatal hypoxic ischemic

encephalopathy." Pediatrics 121(3): 648-649; author reply 649-650.

Henke, N., D. A. Lisak, L. Schneider, J. Habicht, M. Pergande and A. Methner (2011).

"The ancient cell death suppressor BAX inhibitor-1." Cell Calcium 50(3): 251-

260.

Khwaja, O. and J. J. Volpe (2008). "Pathogenesis of cerebral white matter injury of

prematurity." Arch Dis Child Fetal Neonatal Ed 93(2): F153-161.

Kim, H. R., G. H. Lee, E. Y. Cho, S. W. Chae, T. Ahn and H. J. Chae (2009). "Bax

inhibitor 1 regulates ER-stress-induced ROS accumulation through the regulation

of cytochrome P450 2E1." J Cell Sci 122(Pt 8): 1126-1133.

Kim, J. H., E. R. Lee, K. Jeon, H. Y. Choi, H. Lim, S. J. Kim, et al. (2012). "Role of BI-1

(TEGT)-mediated ERK1/2 activation in mitochondria-mediated apoptosis and

splenomegaly in BI-1 transgenic mice." Biochim Biophys Acta 1823(4): 876-888.

Lee, G. H., H. K. Kim, S. W. Chae, D. S. Kim, K. C. Ha, M. Cuddy, et al. (2007). "Bax

inhibitor-1 regulates endoplasmic reticulum stress-associated reactive oxygen

species and heme oxygenase-1 expression." J Biol Chem 282(30): 21618-21628.

Lee, G. H., H. R. Kim and H. J. Chae (2012). "Bax inhibitor-1 regulates the expression of

P450 2E1 through enhanced lysosome activity." Int J Biochem Cell Biol 44(4):

600-611.

131

Li, B., R. K. Yadav, G. S. Jeong, H. R. Kim and H. J. Chae (2014). "The characteristics

of Bax inhibitor-1 and its related diseases." Curr Mol Med 14(5): 603-615.

Li, L., N. H. Khatibi, Q. Hu, J. Yan, C. Chen, J. Han, et al. (2012). "Transmembrane

protein 166 regulates autophagic and apoptotic activities following focal cerebral

ischemic injury in rats." Exp Neurol 234(1): 181-190.

Li, L., D. Klebe, D. Doycheva, D. W. McBride, P. R. Krafft, J. Flores, et al. (2015). "G-

CSF ameliorates neuronal apoptosis through GSK-3beta inhibition in neonatal

hypoxia-ischemia in rats." Exp Neurol 263: 141-149.

Lisbona, F., D. Rojas-Rivera, P. Thielen, S. Zamorano, D. Todd, F. Martinon, et al.

(2009). "BAX inhibitor-1 is a negative regulator of the ER stress sensor

IRE1alpha." Mol Cell 33(6): 679-691.

Moritz, C. P. (2017). "Tubulin or Not Tubulin: Heading Toward Total Protein Staining as

Loading Control in Western Blots." Proteomics 17(20).

Nieto, N., S. L. Friedman and A. I. Cederbaum (2002). "Cytochrome P450 2E1-derived

reactive oxygen species mediate paracrine stimulation of collagen I protein

synthesis by hepatic stellate cells." J Biol Chem 277(12): 9853-9864.

Nieto, N., S. L. Friedman and A. I. Cederbaum (2002). "Stimulation and proliferation of

primary rat hepatic stellate cells by cytochrome P450 2E1-derived reactive

oxygen species." Hepatology 35(1): 62-73.

Owusu-Ansah, E., A. Yavari and U. Banerjee (2008). "A protocol for _in vivo_ detection

of reactive oxygen species."

Robinson, K. S., A. Clements, A. C. Williams, C. N. Berger and G. Frankel (2011). "Bax

inhibitor 1 in apoptosis and disease." Oncogene 30(21): 2391-2400.

132

Sano, R. and J. C. Reed (2013). "ER stress-induced cell death mechanisms." Biochim

Biophys Acta 1833(12): 3460-3470.

Shankaran, S. (2008). "Prevention, diagnosis, and treatment of cerebral palsy in near-

term and term infants." Clin Obstet Gynecol 51(4): 829-839.

Shi, X., L. Xu, D. M. Doycheva, J. Tang, M. Yan and J. H. Zhang (2017). "Sestrin2, as a

negative feedback regulator of mTOR, provides neuroprotection by activation

AMPK phosphorylation in neonatal hypoxic-ischemic encephalopathy in rat

pups." J Cereb Blood Flow Metab 37(4): 1447-1460.

Souvenir, R., N. Fathali, R. P. Ostrowski, T. Lekic, J. H. Zhang and J. Tang (2011).

"Tissue inhibitor of matrix metalloproteinase-1 mediates erythropoietin-induced

neuroprotection in hypoxia ischemia." Neurobiol Dis 44(1): 28-37.

Souvenir, R., J. J. Flores, R. P. Ostrowski, A. Manaenko, K. Duris and J. Tang (2014).

"Erythropoietin inhibits HIF-1alpha expression via upregulation of PHD-2

transcription and translation in an in vitro model of hypoxia-ischemia." Transl

Stroke Res 5(1): 118-127.

Stahl, S., J. M. Burkhart, F. Hinte, B. Tirosh, H. Mohr, R. P. Zahedi, et al. (2013).

"Cytomegalovirus downregulates IRE1 to repress the unfolded protein response."

PLoS Pathog 9(8): e1003544.

Thacker, J. S., D. H. Yeung, W. R. Staines and J. G. Mielke (2016). "Total protein or

high-abundance protein: Which offers the best loading control for Western

blotting?" Anal Biochem 496: 76-78.

Vannucci, R. C. (2000). "Hypoxic-ischemic encephalopathy." Am J Perinatol 17(3): 113-

120.

133

Vannucci, R. C. and S. J. Vannucci (1997). "A model of perinatal hypoxic-ischemic brain

damage." Ann N Y Acad Sci 835: 234-249.

Vannucci, S. J. and H. Hagberg (2004). "Hypoxia-ischemia in the immature brain." J Exp

Biol 207(Pt 18): 3149-3154.

Wang, Z., S. V. Shah, H. Liu and R. Baliga (2014). "Inhibition of cytochrome P450 2E1

and activation of transcription factor Nrf2 are renoprotective in myoglobinuric

acute kidney injury." Kidney Int 86(2): 338-349.

Xie, Z., L. Huang, B. Enkhjargal, C. Reis, W. Wan, J. Tang, et al. (2017). "Intranasal

administration of recombinant Netrin-1 attenuates neuronal apoptosis by

activating DCC/APPL-1/AKT signaling pathway after subarachnoid hemorrhage

in rats." Neuropharmacology 119: 123-133.

Xu, C., B. Bailly-Maitre and J. C. Reed (2005). "Endoplasmic reticulum stress: cell life

and death decisions." J Clin Invest 115(10): 2656-2664.

Xu, N., Y. Zhang, D. M. Doycheva, Y. Ding, Y. Zhang, J. Tang, et al. (2018).

"Adiponectin attenuates neuronal apoptosis induced by hypoxia-ischemia via the

activation of AdipoR1/APPL1/LKB1/AMPK pathway in neonatal rats."

Neuropharmacology 133: 415-428.

Xu, Q. and J. C. Reed (1998). "Bax inhibitor-1, a mammalian apoptosis suppressor

identified by functional screening in yeast." Mol Cell 1(3): 337-346.

Zhang, H. Y., Z. G. Wang, X. H. Lu, X. X. Kong, F. Z. Wu, L. Lin, et al. (2014).

"Endoplasmic Reticulum Stress: Relevance and Therapeutics in Central Nervous

System Diseases." Mol Neurobiol.

134

Zhou, Y., N. Fathali, T. Lekic, J. Tang and J. H. Zhang (2009). "Glibenclamide improves

neurological function in neonatal hypoxia-ischemia in rats." Brain Res 1270: 131-

139.

135

CHAPTER FOUR

DISCUSSION AND CONCLUSION

The role of Bax Inhibitor-1 (BI-1) as a potential therapeutic target against

Endoplasmic Reticulum (ER) stress and its potential effects on pathophysiological outcomes after hypoxic ischemic (HI) insult has not been reported previously. Several studies have shown BI-1’s anti-apoptotic and anti-oxidative properties (Chae et al. ,

2004; Kim et al. , 2009; Bailly-Maitre et al. , 2010), however, little is known about its signaling mechanism which limits its clinical translation. In this study, we showed for the first time that overexpressing BI-1, with gene delivery of adenoviral-TMBIM6 (Ad-

TMBIM6) vector, attenuated short and long-term deficits after HI in a neonatal rat pup.

BI-1’s protection was mediated through IRE1α inhibition as well as through dissociation of the NPR-CYP complex coupled with upregulation of Nrf-2.

We employed both an in vitro, oxygen glucose deprivation (OGD) model, and in vivo neonatal hypoxic ischemic model in rat pup. The main findings of our study are: (1)

In vitro experiments using the OGD model showed that cells transfected with Ad-

TMBIM6 significantly improved cell viability, whereas silencing the TMBIM6 gene reversed those effects. (2) In our OGD experiments we elucidated at least one major pathway via which BI-1 attenuated apoptosis. Results indicated that BI-1 exerts its anti- apoptotic effects via inhibition of IRE1α and to a lesser extend via PERK. (3)

Administration of Ad-TMBIM6 in a neonatal rat subjected to HI, improved both short and long-term outcomes after HI. (4) At Least one of BI-1’s signaling pathways was confirmed to be via IRE1α inhibition, and its downstream molecules, thus attenuating apoptosis, in a rat HI model. (5) BI-1 disrupted the NPR-CYP complex, which was

136

coupled with upregulation of Nrf-2 and anti-oxidant enzymes, HO-1, and this attenuated

ROS generation and subsequent inflammation.

Historically, there has been a vast focus on anti-apoptotic and anti-inflammatory processes after hypoxic or ischemic events. Specifically, the Bcl-2 protein family has been extensively studied in this context, but reports have shown that most members of that family are poorly conserved in animals (Rojas-Rivera et al. , 2015). In contrast, the

TMBIM family are highly conserved evolutionary proteins with common ancestors in yeast (Xu et al. , 1998; Rojas-Rivera et al. , 2012). One member of the TMBIM family,

Bax Inhibitor-1 (BI-1) protein, has been well described as an anti-apoptotic protein conferring protection in cells and animal studies (Bailly-Maitre et al. , 2006; Lee et al. ,

2007; Kim et al. , 2009; Lisbona et al. , 2009). The BI-1 is a small, six-helix domain protein located primarily in the endoplasmic reticulum (ER) (Xu et al. , 1998). It was first identified in human screening to evaluate proteins that could inhibit Bax-mediated cell death (Xu et al. , 1998). It was first referred to as the testis enhanced gene transcript

(TEGT), as it was found to be in high abundance in rat and mouse testis (Walter et al. ,

1994; Walter et al. , 1995). The gene was also detected in the brain, liver, kidney, thalamus, lung, ovary and heart of rat (Walter et al. , 1994).

Since its initial discovery, BI-1’s anti-apoptotic effects have been studied extensively by many groups. BI-1protected yeast cells and HEK293 cells against overexpression of Bax (Xu et al. , 1998). It was protective in plants by inhibiting Bax and

H2O2 (Kawai et al. , 1999). BI-1 can even attenuate Chlamydia induced apoptosis in

HeLa cells (Perfettini et al. , 2002).

137

With more and more emerging studies it soon became evident that BI-1’s protection was most prominent in ER stress and ischemia-reperfusion injury. ER stress is a major pathology encountered due to the accumulation of misfolded proteins, after HI. A series of in vitro studies were employed that showed BI-1’s protection against various ER stress inducers like thapsigargin, and tunicamycin, which are both inhibitors of transport of proteins between ER and Golgi (Chae et al. , 2004). BI-1 also showed protection in

HEK293 cells and neuronal cells after oxygen glucose deprivation (OGD) (Bailly-Maitre et al. , 2006). Studies then moved on to culturing cells from transgenic animals overexpressing BI-1, this demonstrated that cells were protected when exposed to various stresses (Chae et al. , 2004). In contrast, cells that were cultured from BI-1 deficient models were much more sensitive to ER stress or ischemia reperfusion injury (Krajewska et al. , 2011). Similarly, knock down of BI-1 using shRNAs also demonstrated increased sensitivity to ER stress in HeLa cells (Xu 2008). In recent studies of in vivo models of ischemia-reperfusion, traumatic brain injury or SAH, BI-1 overexpression conferred protection by interfering with ER stress induced apoptotic signals (Chae et al. , 2004;

Krajewska et al. , 2011; Shi et al. , 2018).

BI-1 has also been described in diverse other functions. BI-1 acted as a negative regulator of ER stress sensors, such as IRE1α (Lisbona et al. , 2009), it can increase actin polymerization (Lee et al. , 2010), regulate the release of calcium from intracellular stores (Chae et al. , 2004) and reduce reactive oxygen species (ROS) generation by direct interaction with members of the microsomal monooxygenase system (MMO), cytochrome P450 (CYP) and NADPH-P450 reductase (NPR) (Kim et al. , 2009).

138

In vitro studies described BI-1’s protective functions to be against ER stress inducers. Bailly- Maitre and colleagues used BI-1 knockout samples, from liver and kidney, to analyze ER stress markers (Bailly-Maitre et al. , 2006). The most striking results were the evident increase of the spliced X-box-binding protein-1 (XBP1), which is a downstream regulator of IRE1α. This data was later confirmed by Lisbona et al, where a significant increase of XBP1 was observed in BI-1 knockout animals which was correlated with an increase in IRE1α activity (Lisbona et al. , 2009). There was speculation that BI-1 may have some direct relationship with IRE1α-XBP1, since wild type cells that were treated with ER stress inducers, like tunicamycin, showed an initial increase in XBP1 which then decreased over time. However, in BI-1 knockout cells, the levels of XBP1 remained elevated, thus suggesting inhibitory function of BI-1 on XBP1.

This observation was soon confirmed to be via BI-1’s inhibitory effect on IRE1α-XBP1 by direct interaction with IRE1α protein through its C-terminus (Lisbona et al. , 2009).

Co-immunoprecipitation experiments have shown this direct interaction between BI-1 and IRE1α. In a cell culture study where cells were BI-1 deficient, knockdown of either

IRE1α or XBP1 further exacerbated the damage caused by tunicamycin (Lisbona et al. ,

2009). BI-1’s role was also reported in obesity. It regulated glucose metabolism and reversed hyperglycemia in obese mice that were administered adenoviral-TMBIM6 vector, which caused them to overexpress BI-1 (Bailly-Maitre et al. , 2010). This was hypothesized to be via the inhibition of IRE1α and attenuation of XBP1 splicing in mice overexpressing BI-1 in liver. This gives a strong support to BI-1’s role on ER stress sensors, however, more work is needed to show this intricate relationship and its effects in stroke.

139

Some reports have shown that BI-1 may also indirectly inhibit another ER stress induced protein, PERK (Lee et al. , 2007). In contrast, another study reported that BI-1 may not have any effect on that pathway, as eIF2α phosphorylation levels, a downstream factor of PERK, were unchanged in BI-1 knockout mice (Bailly-Maitre et al. , 2006).

Therefore, due to the controversies of BI-1’s role in PERK signaling, we used an

OGD model to first establish BI-1’s role on the two branches of the UPR that are activated during ER stress, IRE1α and PERK. We intervened at both receptor levels to help us elucidate BI-1’s predominant pathway via which it attenuates ER stress induced apoptosis. Our data indicated that cells transfected with Ad-TMBIM6 significantly improved percent cell viability after 3h of OGD, while silencing BI-1 reversed those effects. In conjunction with above mentioned studies, we also observed an inhibition of

IRE1α and XBP1 in cells overexpressing BI-1, while either inhibiting BI-1 or using an

IRE1α agonist (APY-29) abrogated those effects and exacerbated the injury. This was seen from the increase in the pro-apoptotic marker, cleaved-caspase 3, levels.

Furthermore, we compared BI-1’s role on the IRE1α branch to the PERK branch and observed that cells overexpressing BI-1, that were transfected with either a PERK agonist, CCT020312, or inhibition of PERK with a CRISPR k/o, did not alter BI-1’s expression levels. Additionally, cleaved caspase-3 levels were lower in BI-1 overexpressing cells transfected with an IRE1α inhibitor versus that of cells transfected with PERK inhibitor. This is an indication that BI-1 may more potently attenuate apoptosis via IRE1α inhibition than via PERK inhibition. This result suggests that a direct interaction exists between BI-1 and IRE1α, but not between BI-1 and PERK, thus

BI-1 has a more potent effect on IRE1α (Lee et al. , 2007; Lisbona et al. , 2009). Our data

140

also supports that interaction, as we saw intervening at IRE1α levels caused changes in

BI-1 expression patterns whereas manipulation of PERK did not alter BI-1’s levels.

To date, BI-1’s role and signaling mechanisms have not been elucidated in a neonatal HI rat model. In this study, for the first time, we showed that overexpressing BI-

1, using an Ad-TMBIM6, administered intracerebroventricularly into the neonatal brain, conferred protection after HI. Ad-TMBIM6 significantly reduced percent infarcted area, attenuated neuronal apoptosis and long-term motor, learning and memory deficits.

Morphological analysis revealed lower percentage of tissue loss in animals overexpressing BI-1 versus animals in vehicle group. These protective effects were observed to be via inhibition of the IRE1α branch of the ER stress response signaling pathway.

In conjunction with activating the UPR and initiating pro-apoptotic signaling, ER stress has also been linked with increased reactive oxygen species (ROS) which leads to oxidative stress and inflammation (Cao et al. , 2014; Zeeshan et al. , 2016). HI can trigger a robust inflammatory response prompting the activation of inflammatory cells in the brain which exacerbates brain injury (Zhao et al. , 2016). ROS is generated due to activation of these inflammatory cells, which further contributes to the activation of pro- inflammatory mediators. Apart from BI-1’s role as an anti-apoptotic protein and its ability to attenuate the UPR response, it has also been implicated in its ability to reduce

ROS (Kawai-Yamada et al. , 2004; Kim et al. , 2009; Henke et al. , 2011).

Overexpression of BI-1 reduced ROS and attenuated oxidative damage (Kim et al. , 2009). A major contributor to ROS generation during ER stress is from the microsomal monooxygenase system (MMO). This occurs as a result of electron leakage

141

between NPR and the major component of the MMO system, P4502E1 (CYP).

Specifically, P4502E1 activity was increased in response to ER stress inducers, thapsigargin, while BI-1 overexpressing cells inhibited P4502E1 expression (Kim et al. ,

2009). Levels of P4502E1 were lower in BI-1 cells versus wildtype, which was consistent with lower levels of ROS accumulation. Although BI-1 attenuates P4502E1 expression and activity, a clear interaction between these two has not been shown. Instead, co- immunoprecipitation studies have shown a physical interaction to exist between BI-1 and

NPR (Kim et al. , 2009). It has been suggested that BI-1 may induce dissociation of the

NPR-CYP complex by directly binding to NPR and reducing the electron transfer between these two and hence reducing ROS (Kim et al. , 2009). A second possible mechanism of BI-1 protection against ROS is via upregulation of anti-oxidant factors, such as Nrf-2 and HO-1. Cells overexpressing BI-1 have higher levels of Nrf-2 and HO-

1, which was linked to a reduction in ROS (Lee et al. , 2007).

To examine this possibility after neonatal HI injury, we tested this pathway in a rat subjected to unilateral carotid artery ligation, followed by hypoxia to induce HI injury.

We used siRNAs to silence BI-1 or Nrf-2 and detected that silencing of BI-1 increased

P4502E1 expression levels, which was correlated with an increase in ROS and inflammatory mediators. In addition, silencing of Nrf-2 showed similar results.

Furthermore, rats that were administered Ad-TMBIM6 without siRNAs had significantly reduced P4502E1 expression levels together with upregulation of Nrf-2 and HO-1 with subsequent attenuation of ROS and inflammation. Here we showed that BI-1 can also attenuate oxidative stress and inflammation via increased production of anti-oxidant

142

enzymes coupled with destabilization of the NPR-CYP complex in a neonatal HI rat model.

Understanding BI-1’s signaling mechanisms pertaining to inhibition of ER stress induced injury after HIE is crucial in establishing how the two components interrelate.

Clinical Significance

Given the limited therapeutic options for hypoxic ischemic encephalopathy (HIE) it is necessary to understand cellular and molecular mechanisms which can lead to more effective therapeutic strategies. Hypothermia is the only therapy that has shown success

(Meloni 2008), but despite that, a large percentage of neonates still develop long-term complications.

The lack of adequate mechanistic animal studies, therapeutic window, timing of treatment and various other factors halt the progression of promising targets to be translated to clinics.

Can BI-1 Be a Potential Therapeutic Target?

One must consider the plethora of signaling mechanisms that can be activated/inhibited by BI-1. It can regulate calcium release, bind to and activate Bcl-2 and inhibit Bax, attenuate UPR proteins activated as a result of ER stress, reduce production of ROS and increase actin polymerization (Kim et al. , 2008; Henke et al. , 2011). BI-1’s multimodal properties suggest that it can target a wide array of pathophysiological consequences after HIE. Therefore, this makes BI-1 a potential therapeutic target and

143

may change the clinical management for patients with HI injury and provide a foundation for future research in other types of strokes with similar pathologies

The only caveat is that there is a lack of in vivo studies showing BI-1’s diverse roles. A vast majority of the studies have been done in cell culture models. Therefore, it is imperative to translate those in vitro studies to in vivo disease models. Additionally,

BI-1’s anti-apoptotic roles might be detrimental in some disease processes such as in an infection. For example, BI-1 may act in conjunction with pathogens to promote cell survival of infected cells (Robinson et al. , 2011). Some pathogens contain molecules that can bind with BI-1 and inhibit apoptosis, therefore, promoting survival of the infected cell. One such molecule was found to be NIeH, that bound to BI-1 and decrease calcium release during infection, thus blocking apoptosis (Hemrajani et al. , 2010). It is crucial to determine all of BI-1’s functions and signaling mechanisms in a wide array of disease models before it can be translated to clinical studies.

Conclusion

There has been considerable progress made in understanding BI-1 and its role in apoptosis and disease. Despite that, the way in which it functions in stroke and other disease models remains to be elucidated. We showed that overexpressing BI-1 in a neonatal rat brain conferred protection after HI injury by reducing apoptosis and oxidative stress. Agonizing the effects of ER stress with silencing either BI-1 or the anti- oxidant transcription factor Nrf-2, or further upregulating IRE1α signaling, exacerbated the injury even with Ad-TMBIM6 vector administration.

144

With this study we identified a novel role for BI-1 protein and endoplasmic reticulum stress in the pathophysiology of neonatal HI injury. We found that BI-1 conferred an inhibitory effect on IRE1α signaling thus attenuating apoptosis.

Additionally, we showed that BI-1 inhibited P4502E1, major component for ROS production, coupled with upregulation of anti-oxidant factors, Nrf-2 and HO-1, to attenuate oxidative stress and inflammation after HI. Nonetheless, future studies are needed to better understand BI-1’s role in ER stress which may aid in better understanding how cells can overcome stresses. In summary there are still many aspects to be answered about BI-1 and how it can regulate life and death.

Future Directions

In our study we focused on two general concepts: The effects of overexpressing

BI-1 in the neonatal rat brain using gene delivery of adenoviral-TMBIM6 vector, and the mechanisms by which it confers its protective effects. Due to limited knowledge on BI-1 and thereafter serious lack of specific drugs targeting it, we used an adenoviral vector to over express the TMBIM6 gene, which can overexpress BI-1 protein. However, using adenoviral vectors takes multiple days to begin expressing, causing a delay in treatment and possibly missing the optimal therapeutic window. Further exploration and development of more adequate ways to target BI-1 overexpression in vivo are needed.

Additionally, we injected the virus intracerebroventricularly, future research into more clinically translatable routs is necessary.

Our study focused on BI-1’s role in ER stress after neonatal HIE, however, there are many other areas yet to be investigated in various other stroke models and diseases.

145

The protection conferred after HIE in this study via inhibition of IRE1α and dissociation of the NPR-CYP complex may not be optimal in other models of disease. Further research is necessary to see BI-1’s role in diseases and potential alternative pathways that may be involved. There is a possibility that BI-1’s effects extend beyond the ER stress mechanisms and these need to be evaluated in subsequent studies.

146

References

Bailly-Maitre, B., B. F. Belgardt, S. D. Jordan, B. Coornaert, M. J. von Freyend, A.

Kleinridders, et al. (2010). "Hepatic Bax Inhibitor-1 Inhibits IRE1 alpha and

Protects from Obesity-associated Insulin Resistance and Glucose Intolerance."

Journal of Biological Chemistry 285(9): 6198-6207.

Bailly-Maitre, B., C. Fondevila, F. Kaldas, N. Droin, F. Luciano, J. E. Ricci, et al.

(2006). "Cytoprotective gene bi-1 is required for intrinsic protection from

endoplasmic reticulum stress and ischemia-reperfusion injury." Proc Natl Acad

Sci U S A 103(8): 2809-2814.

Cao, S. S. and R. J. Kaufman (2014). "Endoplasmic reticulum stress and oxidative stress

in cell fate decision and human disease." Antioxid Redox Signal 21(3): 396-413.

Chae, H. J., H. R. Kim, C. Xu, B. Bailly-Maitre, M. Krajewska, S. Krajewski, et al.

(2004). "BI-1 regulates an apoptosis pathway linked to endoplasmic reticulum

stress." Mol Cell 15(3): 355-366.

Hemrajani, C., C. N. Berger, K. S. Robinson, O. Marches, A. Mousnier and G. Frankel

(2010). "NleH effectors interact with Bax inhibitor-1 to block apoptosis during

enteropathogenic Escherichia coli infection." Proc Natl Acad Sci U S A 107(7):

3129-3134.

Henke, N., D. A. Lisak, L. Schneider, J. Habicht, M. Pergande and A. Methner (2011).

"The ancient cell death suppressor BAX inhibitor-1." Cell Calcium 50(3): 251-

260.

147

Kawai-Yamada, M., Y. Ohori and H. Uchimiya (2004). "Dissection of Arabidopsis Bax

Inhibitor-1 Suppressing Bax–, Hydrogen Peroxide–, and Salicylic Acid–Induced

Cell Death." The Plant Cell 16(1): 21-32.

Kawai, M., L. Pan, J. C. Reed and H. Uchimiya (1999). "Evolutionally conserved plant

homologue of the Bax Inhibitor‐1 (BI‐1) gene capable of suppressing Bax‐

induced cell death in yeast 1." FEBS Letters 464(3): 143-147.

Kim, H. R., G. H. Lee, E. Y. Cho, S. W. Chae, T. Ahn and H. J. Chae (2009). "Bax

inhibitor 1 regulates ER-stress-induced ROS accumulation through the regulation

of cytochrome P450 2E1." J Cell Sci 122(Pt 8): 1126-1133.

Kim, H. R., G. H. Lee, K. C. Ha, T. Ahn, J. Y. Moon, B. J. Lee, et al. (2008). "Bax

Inhibitor-1 Is a pH-dependent regulator of Ca2+ channel activity in the

endoplasmic reticulum." J Biol Chem 283(23): 15946-15955.

Krajewska, M., L. Xu, W. Xu, S. Krajewski, C. L. Kress, J. Cui, et al. (2011).

"Endoplasmic reticulum protein BI-1 modulates unfolded protein response

signaling and protects against stroke and traumatic brain injury." Brain Res 1370:

227-237.

Lee, G. H., T. Ahn, D. S. Kim, S. J. Park, Y. C. Lee, W. H. Yoo, et al. (2010). "Bax

inhibitor 1 increases cell adhesion through actin polymerization: involvement of

calcium and actin binding." Mol Cell Biol 30(7): 1800-1813.

Lee, G. H., H. K. Kim, S. W. Chae, D. S. Kim, K. C. Ha, M. Cuddy, et al. (2007). "Bax

inhibitor-1 regulates endoplasmic reticulum stress-associated reactive oxygen

species and heme oxygenase-1 expression." J Biol Chem 282(30): 21618-21628.

148

Lisbona, F., D. Rojas-Rivera, P. Thielen, S. Zamorano, D. Todd, F. Martinon, et al.

(2009). "BAX inhibitor-1 is a negative regulator of the ER stress sensor

IRE1alpha." Mol Cell 33(6): 679-691.

Meloni, B. P. (2008). "Therapeutic Applications of Hypothermia in Cerebral." 1(2): 12-

35.

Perfettini, J. L., J. C. Reed, N. Israël, J. C. Martinou, A. Dautry-Varsat and D. M. Ojcius

(2002). "Role of Bcl-2 Family Members in Caspase-Independent Apoptosis

during Chlamydia Infection." Infect Immun 70(1): 55-61.

Robinson, K. S., A. Clements, A. C. Williams, C. N. Berger and G. Frankel (2011). "Bax

inhibitor 1 in apoptosis and disease." Oncogene 30(21): 2391-2400.

Rojas-Rivera, D., R. Armisen, A. Colombo, G. Martinez, A. L. Eguiguren, A. Diaz, et al.

(2012). "TMBIM3/GRINA is a novel unfolded protein response (UPR) target

gene that controls apoptosis through the modulation of ER calcium homeostasis."

Cell Death Differ 19(6): 1013-1026.

Rojas-Rivera, D. and C. Hetz (2015). "TMBIM protein family: ancestral regulators of

cell death." Oncogene 34(3): 269-280.

Shi, L., Z. Wang, X. Liu, M. Li, S. Zhang and X. Song (2018). "Bax inhibitor-1 is

required for resisting the Early Brain Injury induced by subarachnoid hemorrhage

through regulating IRE1-JNK pathway." Neurol Res 40(3): 189-196.

Walter, L., B. Dirks, E. Rothermel, M. Heyens, C. Szpirer, G. Levan, et al. (1994). "A

novel, conserved gene of the rat that is developmentally regulated in the testis."

Mamm Genome 5(4): 216-221.

149

Walter, L., P. Marynen, J. Szpirer, G. Levan and E. Gunther (1995). "Identification of a

novel conserved human gene, TEGT." Genomics 28(2): 301-304.

Xu, C. (2008). "BI-1 Regulates Endoplasmic Reticulum Ca(2+) Homeostasis." 283(17):

11477-11484.

Xu, Q. and J. C. Reed (1998). "Bax inhibitor-1, a mammalian apoptosis suppressor

identified by functional screening in yeast." Mol Cell 1(3): 337-346.

Zeeshan, H. M., G. H. Lee, H. R. Kim and H. J. Chae (2016). "Endoplasmic Reticulum

Stress and Associated ROS." Int J Mol Sci 17(3): 327.

Zhao, M., P. Zhu, M. Fujino, J. Zhuang, H. Guo, I. Sheikh, et al. (2016). "Oxidative

Stress in Hypoxic-Ischemic Encephalopathy: Molecular Mechanisms and

Therapeutic Strategies." International Journal of Molecular Sciences 17(12):

2078.

150