MECHANISMS OF CIRCADIAN RHYTHM PROTEIN PERIOD2 IN CARDIOPROTECTION

by

COLLEEN MARIE BARTMAN

B.A., College of Wooster, 2012

A dissertation submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirements for the degree of

Doctor of Philosophy

Cell Biology, Stem Cells, and Development Program

2018

This dissertation for the Doctor of Philosophy degree by

Colleen Marie Bartman

has been approved for the

Cell Biology, Stem Cells, and Development Program

by

Julie Siegenthaler, Chair

Peter Buttrick

Lori Walker

Bruce Appel

Amanda Law

Tânia Reis

Tobias Eckle, Advisor

Date: August 17th, 2018

ii Bartman, Colleen Marie, Ph.D., Cell Biology, Stem Cells, and Development Program

Mechanisms of Circadian Rhythm Protein Period2 in Cardioprotection

Dissertation directed by Professor Tobias Eckle

ABSTRACT

Throughout evolutionary time, all organisms and species on Earth evolved with an adaptation to consistent oscillations of sunlight and darkness, now recognized as ‘circadian rhythm.’ Single-cellular to multi-system organisms use circadian biology to synchronize to the

external environment and provide predictive adaptation to changes in cellular homeostasis.

Dysregulation of circadian biology has been implicated in numerous prevalent human diseases

and therefore targeting the circadian machinery may provide innovative preventative or

therapeutic treatment strategies. Previous studies from our lab included a wide search for

ischemic preconditioning mechanisms of cardioprotection and identified the light-elicited circadian

rhythm protein Period2 (PER2) to be cardioprotective from conditions of low oxygen availability,

like myocardial ischemia. However, underlying mechanisms of PER2 dependent cardioprotection

remained widely unknown. Our goal moving forward was to decipher the molecular nature of

PER2 mediated cardioprotection and use these mechanisms to target PER2 to reduce cellular or

tissue damage from hypoxia or ischemia, respectively. We uncovered a multi-faceted role for

PER2, functioning both transcriptionally in the nucleus and post-translationally in the

mitochondria. Our studies point toward PER2 as a master regulator of metabolic adaptation to

hypoxia by working in concert with hypoxia inducible factor 1⍺ and sirtuin 3. Further investigation

into ischemic preconditioning mechanisms uncovered the circadian and PER2 dependent

microRNA miR-21 to be cardioprotective from ischemia and reperfusion injury. Focused studies

into the effects of circadian disruption on the heart revealed that a frequently administered

benzodiazepine, midazolam, has deleterious effects on cardiac tissue after ischemia and

reperfusion injury. However, circadian amplitude enhancement of PER2 via a novel small

molecule inhibitor nobiletin reversed this damage. Moreover, we optimized our daylight regime to

iii achieve optimal cardioprotection in a PER2 dependent manner and our mechanistic studies identified downstream cAMP signaling pathways as an underlying mechanism. Lastly, our translational studies into healthy human volunteers suggested that light-elicited cardioprotective strategies identified in the laboratory setting may be beneficial in the clinic. Taken together, we identified several mechanisms of PER2 mediated cardioprotection that may provide innovative preventative or therapeutic treatment strategies for patients with myocardial ischemia.

The form and content of this abstract are approved. I recommend its publication.

Approved: Tobias Eckle

iv ACKNOWLEDGEMENTS

I would like to express my deepest gratitude to Dr. Tobias Eckle for his continued support, openness to my thoughts and ideas, and mentorship. I would also like to thank Dr. Peter Buttrick and Dr. Lori Walker for their contributions and collaborations throughout my graduate studies. In addition, I appreciate my thesis committee, Dr. Buttrick, Dr. Walker, Dr. Bruce Appel, Dr. Amanda

Law, and Dr. Tânia Reis for their thoughtful advice and in particular, thanks to Dr. Julie

Siegenthaler for being an outstanding committee chair. Furthermore, I would like to thank Dr.

John Tentler for his support in my CCTSI pre-doctoral fellowship and my translational studies. I owe significant thanks to Dr. Yoshimasa Oyama, the post-doctoral fellow in the Eckle lab who performed the mouse surgeries. Lastly, I would like to thank Stephanie Bonney for her prior work in the Eckle lab making the stable knockdown cell lines.

I would like to acknowledge my two pre-doctoral fellowships that I received during my four years as a student. My second year I received a Colorado Clinical and Translational Sciences

Institute (CCTSI) fellowship (TL1 TR001081) that opened doors of opportunity to broaden my research into clinical applications and enhanced my understanding of translational research. My third and fourth years were funded through an American Heart Association (AHA) fellowship

(16PRE30510006), which supported my mechanistic studies, the majority of which are the core of my doctoral research and presented in my dissertation. I am thankful for the funding and opportunities allotted by these fellowships. Additionally, I had the opportunity to present my research at many conferences in Colorado, New Mexico, Washington D.C., and Illinois. These conferences were funded in part by my fellowships, Dr. Eckle’s NHLBI funding (5R01HL122472), the Gates Center for Regenerative Medicine, and Dr. Vesna Jevtovic-Todorovic with the

Department of Anesthesiology. The University of Colorado Anschutz Medical Campus and the

Graduate Program in Cell Biology, Stem Cells, and Development fostered personal and professional growth during my journey as a graduate student. I am so grateful I had the opportunity to be a part of such a supportive, inclusive, and engaging community.

v TABLE OF CONTENTS

CHAPTER

I. INTRODUCTION ...... 1

1.1 The Circadian-Hypoxia Link ...... 1

1.1.a Cyanobacteria and the Great Oxygenation Event ...... 1

1.1.b Light-sensing pathways: the mammalian circadian system and feedback loop...... 2

1.1.c Oxygen-sensing pathways: hypoxia inducible factors ...... 4

1.1.d PAS domains: an evolutionarily link between circadian and hypoxia pathways ...... 5

1.2 PER2 and HIF1⍺ in Metabolic Adaptation to Myocardial Ischemia ...... 6

1.2.a Adenosine and ischemic preconditioning link PER2 and HIF1⍺ ...... 6

1.2.b PER2 regulation of HIF1⍺ ...... 8

1.2.c Regulation of energy metabolism during hypoxia or ischemia ...... 9

1.2.d HIF1⍺ mediated metabolic adaptation to hypoxia ...... 11

1.2.e PER2 and metabolic regulation ...... 12

1.2.f Overlap between PER2 and HIF1⍺ in metabolic regulation ...... 16

1.3 MicroRNAs in Circadian Biology and Therapies ...... 17

1.3.a MicroRNAs as cardiovascular therapeutic targets ...... 17

1.3.b Circadian microRNAs ...... 19

1.4 Circadian Disruption at the Root of Disease ...... 21

1.4.a Circadian disruption is associated with a multitude of prevalent diseases ...... 21

1.4.b ICU settings and anesthetics have severe effects on the circadian system ...... 23

1.5 Circadian Considerations in Therapeutic Strategies ...... 25

1.5.a Maintaining a robust circadian amplitude ...... 25

1.5.b Diurnal versus nocturnal considerations ...... 26

vi 1.5.c Pharmacological timing and targeting of the circadian system ...... 27

CHAPTER

II. PER2 MEDIATED ENERGY METABOLISM IN ADAPTATION TO HYPOXIA ...... 39

2.1 Rationale ...... 39

2.2 Results ...... 40

2.2.a PER2 transcriptionally controls HIF1⍺ dependent glycolysis during hypoxia ...... 40

2.2.b PER2 is a post-translational regulator of TCA cycle activity during hypoxia ...... 41

2.2.c PER2 regulates mitochondrial function via HIF1⍺-COX4.2 in hypoxia ...... 42

2.3 Discussion ...... 43

CHAPTER

III. PER2 MICRORNA TARGETS IN METABOLIC ADAPTATION TO HYPOXIA ...... 79

3.1 Rationale ...... 79

3.2 Results ...... 79

3.2.a Differential and PER2 dependent regulation of miR-21 after IPC ...... 79

3.2.b Diurnal expression pattern of miR-21 in murine tissues ...... 80

3.2.c MiR-21 is exclusively upregulated in hypoxic cardiac endothelial cells ...... 80

3.2.d MiR-21 is critical for cellular glycolytic reliance ...... 81

3.3 Discussion ...... 82

CHAPTER

IV. IMPLICATIONS OF CIRCADIAN DISRUPTION IN MYOCARDIAL ISCHEMIA ...... 97

4.1 Rationale ...... 97

4.2 Results ...... 97

4.2.a Midazolam downregulates cardiac PER2 in wildtype mice ...... 97

vii 4.2.b Midazolam has deleterious effects during myocardial IR-injury ...... 98

4.3 Discussion ...... 99

CHAPTER

V. CIRCADIAN AMPLITUDE ENHANCEMENT AS A THERAPEUTIC STRATEGY IN

ISCHEMIA ...... 104

5.1 Rationale ...... 104

5.2 Results ...... 105

5.2.a Daylight-elicited cardioprotection through circadian amplitude enhancement ...... 105

5.2.b Wheel running elicits circadian amplitude enhancement ...... 105

5.2.c Daylight-elicited PER2 activates HIF1⍺ metabolic adaptation to ischemia ...... 106

5.2.d Light-sensing HMEC-1 reveal light dependent pathways in metabolism ...... 106

5.2.e Intense light exposure induces cardiac miR-21 ...... 107

5.2.f MiR-21 is required for light-elicited cardioprotection ...... 107

5.2.g The PER2 amplitude enhancer nobiletin is cardioprotective ...... 108

5.2.h Nobiletin reverses the deleterious effects of midazolam ...... 109

5.3 Discussion ...... 110

CHAPTER

VI. LIGHT-ELICITED CIRCADIAN AMPLITUE ENHANCEMENT IN HUMANS ...... 144

6.1 Rationale ...... 144

6.2 Results ...... 144

6.2.a Daylight-elicited circadian amplitude enhancement in healthy human subjects ...... 144

6.2.b Intense light exposure increases miR-21 and PFK in healthy human subjects ...... 145

6.3 Discussion ...... 146

viii CHAPTER

VII. CCTSI PRE-DOCTORAL FELLOWSHIP EXPERIENCE ...... 165

7.1 The Fellowship ...... 165

7.2 Clinical Relevance ...... 166

7.3 Experimental Design and Screening Criteria ...... 167

CHAPTER

VIII. METHODS ...... 170

8.1 Cell Culture ...... 170

8.1.a Cell culture and treatments ...... 170

8.1.b Lentiviral-mediated generation of cells with knockdown of PER2 or HIF1⍺ ...... 170

8.1.c Endothelial cells ...... 170

8.1.d MiR-21 gain or loss of function ...... 171

8.1.e Light sensing cells ...... 171

8.2 Transcriptional Analysis ...... 172

8.2.a Total RNA Isolation ...... 172

8.2.b Messenger RNA Analysis ...... 172

8.2.c MicroRNA PCR Array ...... 172

8.2.d MicroRNA Analysis ...... 173

8.3 Immunoblotting Analysis ...... 173

8.4 DNA-Protein and Protein-Protein Interactions ...... 173

8.4.a Chromatin immunoprecipitation (ChIP) assay ...... 173

8.4.b Affinity purification-mass spectrometry-based proteomics ...... 174

8.4.c Co-immunoprecipitations (Co-IPs)...... 176

ix 8.5 Metabolic Measurements ...... 176

8.5.a Lactate measurements...... 176

8.5.b Cytotoxicity ...... 176

8.5.c Glycolytic stress tests using the Seahorse Bioanalyzer ...... 176

8.5.d Mitochondrial stress tests using the Seahorse Bioanalyzer ...... 176

8.5.e Cell energy phenotype assays using the Seahorse Bioanalyzer ...... 177

8.5.f Enzyme activities IDH, ACO, SUCLG, Complex IV, PFK, LDH ...... 177

8.5.g Mitochondrial membrane potential dyes ...... 177

8.5.h 13C-tracers in vitro ...... 177

8.6 Mouse Experiments ...... 178

8.6.a General mice usage ...... 178

8.6.b Light exposure in mice ...... 179

8.6.c PER2-/- mice ...... 179

8.6.d PER2 reporter mice ...... 179

8.6.e Mir21-/- mice ...... 179

8.6.f Murine Model for cardiac ischemic preconditioning (IPC) ...... 180

8.6.g Murine model for myocardial ischemia and reperfusion injury ...... 181

8.6.h 13C-tracers in vivo ...... 182

8.6.i Heart enzyme measurement ...... 184

8.6.j Luciferase assay – tissue ...... 184

8.6.k Wheel running ...... 184

8.6.l Enucleation procedure in mice ...... 185

8.6.m cAMP ELISA and phospho-CREB assays ...... 185

8.6.n Hydrogen peroxide assay ...... 185

8.7 Isolation of cardiac tissues ...... 186

x 8.7.a Isolation of fibroblasts ...... 186

8.7.b Isolation of adult cardiomyocytes ...... 186

8.8 Human experiments ...... 187

8.8.a Human light exposure ...... 187

8.8.b Phosphofructokinase (PFK) activity...... 188

8.8.c Human plasma melatonin, HIF1⍺ and triglycerides levels ...... 188

8.8.d Targeted metabolomics - mass spectrometry ...... 188

8.8.e Data analysis ...... 188

CHAPTER

IX. CONCLUSIONS AND FUTURE DIRECTIONS ...... 193

9.1 Summary ...... 193

9.2 Light Elicited Cardioprotection Through Circulating Adenosine ...... 195

9.3 Potential Role of Peripheral Opsin Receptors in Cardioprotection ...... 195

9.4 PER2 and the NAD+-Fumarate Reductase System ...... 196

9.5 Time-of-Day-Dependent Role of Cardiovascular Drugs ...... 197

9.6 Sex Differences in the Circadian Biology of Cardioprotection ...... 198

9.7 MiR-21 Targets for PER2 Dependent Cardioprotection ...... 199

9.8 Peroxiredoxins in Circadian Mediated Metabolic Adaptation ...... 200

9.9 The Role of Inflammation in Cardioprotection ...... 201

9.10 Outline a Circadian Protocol for Cardioprotection in ICU Patients ...... 201

9.11 Concluding Remarks ...... 202

APPENDICES ……………………………………………………………………………………….. 253

xi Appendix A Affinity Purification-Mass Spectrometry-Based Proteomics Screen …… 253

Appendix B Clinical Trial #13-1607 Consent Form ………………………………………... 273

xii LIST OF TABLES

Table 1.1 Clock Mutant Mouse Models with Metabolic and Circadian Phenotypes...... 37

Table 1.2 MicroRNAs in the Heart ...... 38

Table 3.1 PER2 Dependent MicroRNAs During Cardiac IPC ...... 94

Table 8.1 Human Primers ...... 189

Table 8.2 Mouse Primers ...... 190

Table 8.3 Primary and Secondary Antibodies ...... 191

Table 8.4 Pharmacological Compounds ...... 192

xiii LIST OF FIGURES

Figure 1.1 Circadian Feedback Loop ...... 30

Figure 1.2 Hypoxia Signaling ...... 31

Figure 1.3 Adenosine Signaling in Cardioprotection ...... 32

Figure 1.4 Energy Efficient Metabolism ...... 33

Figure 1.5 Possible microRNA pathways for regulating the circadian clock ...... 34

Figure 1.6 Circadian cardiac entrainment as cardioprotective mechanism ...... 35

Figure 1.7 SCN Outputs ...... 36

Figure 2.1 PER2 KD HMEC-1 Validation ...... 45

Figure 2.2 Pyruvate Kinase Gene Expression Dependent on PER2 ...... 46

Figure 2.3 PER2 Dependent Lactate Dehydrogenase Expression and Production...... 47

Figure 2.4 HIF1⍺ Glycolytic Target Genes are PER2 – But Not PER1 – Dependent ...... 48

Figure 2.5 Cytotoxicity of PER2 KD HMEC-1 ...... 49

Figure 2.6 PER2 is Important for Glycolytic Reliance in Hypoxia...... 50

Figure 2.7 Time-of-Day and PER2-Dependent Glycolytic Reliance ...... 51

Figure 2.8 PER2 Dependent HIF1⍺ Binding at the LDHA Promoter ...... 52

Figure 2.9 PER2-Normoxia-Hypoxia-Pathways ...... 53

Figure 2.10 Characterization of PER2-Normoxia-Hypoxia Pathways ...... 54

Figure 2.11 Western Blot Validation of Mass Spectrometry Analysis...... 55

Figure 2.12 PER2 Dependent TCA Cycle Enzyme Activities ...... 56

Figure 2.13 PER2 Dependent TCA Cycle Functioning ...... 57

Figure 2.14 PER2 is Associated with SIRT3 and IDH2 ...... 58

Figure 2.15 PER2 and HIF1⍺ Dependent Regulation of SIRT3 Gene Expression ...... 59

Figure 2.16 SIRT Promoter with HRE Binding Site...... 60

Figure 2.17 SIRT3 Protein Levels Depend Upon PER2 ...... 61

Figure 2.18 SIRT3 Protein Levels Depend on HIF1⍺ ...... 62

xiv Figure 2.19 PER2 Is Important for Mitochondrial Respiration ...... 63

Figure 2.20 Quantification of PER2 in Mitochondrial Respiration ...... 64

Figure 2.21 Ability of PER2 to Use Fatty Acids for Fuel in Mitochondrial Respiration ...... 65

Figure 2.22 PER2 in Mitochondrial Respiration at ZT12 ...... 66

Figure 2.23 Quantification of Circadian Variation in Mitochondrial Respiration ...... 67

Figure 2.24 Complex 4.2 Gene Expression is Dependent upon Hypoxia and PER2 ...... 68

Figure 2.25 Complex IV Activity in Ischemia is PER2 Dependent ...... 69

Figure 2.26 Dysfunction in Mitochondrial Potential in PER2 KD HMEC-1 (MitoTracker)...... 70

Figure 2.27 Cell Energy Phenotype of PER2 KD HMEC-1 ...... 71

Figure 2.28 PER2 Dependent Defect in Mitochondrial Membrane Potential in Normoxia ...... 72

Figure 2.29 PER2 Dependent Defect in Mitochondrial Membrane Potential in Hypoxia ...... 73

Figure 2.30 Quantification of JC-1 Assays ...... 74

Figure 2.31 PER2 Dependent Glycolytic and TCA Cycle Flux in Hx vs Nx ...... 75

Figure 2.32 PER2 in PPP Flux in Hx vs Nx ...... 76

Figure 2.33 PER2 in Lipid and Fatty Acid Metabolism in Hx vs Nx ...... 77

Figure 2.34 PER2 Hypoxia-Only Pathways ...... 78

Figure 3.1 miR-21 Regulation in WT and PER2-/- Mice After Cardiac IPC ...... 84

Figure 3.2 Diurnal Expression of MiR-21 in Murine Hearts ...... 85

Figure 3.3 Diurnal Expression of MiR-21 in Murine Lungs...... 86

Figure 3.4 Relative miR-21 Expression in Different Cardiac Tissues ...... 87

Figure 3.5 miR-21 Expression in Different Cardiac Tissues at Baseline and During Hypoxia .... 88

Figure 3.6 Glycolysis in MiR-21 Loss of Function Human Endothelial Cells ...... 89

Figure 3.7 Quantification of Glycolysis in MiR-21 Loss of Function Human Endothelial Cells .... 90

Figure 3.8 Glycolysis in MiR-21 Gain of Function Human Endothelial Cells ...... 91

Figure 3.9 Quantification of Glycolysis in MiR-21 Gain of Function Human Endothelial Cells .... 92

Figure 3.10 Potential upstream and downstream targets of miR-21 ...... 93

xv Figure 4.1 Cardiac PER2 Regulation Following Exposure of Wildtype Mice to Anesthetics..... 100

Figure 4.2 Cardiac PER2 Regulation in Cardiac Tissues Exposed to Midazolam ...... 101

Figure 4.3 Midazolam in Myocardial Ischemia and Reperfusion Injury ...... 102

Figure 4.4 Troponin Levels After Midazolam in Myocardial Ischemia and Reperfusion Injury .. 103

Figure 5.1 Optimization of Daylight-Elicited Cardioprotection ...... 115

Figure 5.2 Infarct Staining of Time-Dependent Daylight-Elicited Cardioprotection ...... 116

Figure 5.3 Light Exposure Induces Cardiac PER2 Amplitude Enhancement ...... 117

Figure 5.4 Sex-Specific Differences in Male Versus Female Light-Elicited Cardioprotection ... 118

Figure 5.5 Enucleated Mice Do Not Exhibit an Increase in Cardiac PER2 Protein ...... 119

Figure 5.6 Enucleated Mice Do Not Achieve Light-Elicited Cardioprotection ...... 120

Figure 5.7 Hearts of Enucleated Mice Exhibit Significantly Larger Infarct Sizes ...... 121

Figure 5.8 Actigraphy Graphs of Wheel Running as a Circadian Entrainment Strategy ...... 122

Figure 5.9 Wheel Running Measurements as an Entrainment Strategy ...... 123

Figure 5.10 Actigraphy Graphs During Light-Elicited Entrainment Strategy ...... 124

Figure 5.11 Voluntary Wheel Running for Amplitude Enhancement is Cardioprotective ...... 125

Figure 5.12 Light-Elicited Effect on 13C Fructose 1,6-bisP in the Left Ventricle ...... 126

Figure 5.13 Light-Elicited and PER2 Dependent PFK Activity in Plasma and Tissue ...... 127

Figure 5.14 PER2 is Required for Light-Elicited Cardioprotection ...... 128

Figure 5.15 cAMP as a Potential Mechanism in PER2 Dependent Cardioprotection ...... 129

Figure 5.16 Light-Sensing HMEC-1 Reveal cAMP as a Light-Elicited Pathway ...... 130

Figure 5.17 Light-Sensing HMEC-1 Exhibit an Increase in Glycolytic Reliance ...... 131

Figure 5.18 Light-Sensing HMEC-1 Exhibit an Increase in Mitochondrial Functioning ...... 132

Figure 5.19 Daylight Increases miR-21 and PER2 in Heart Tissue ...... 133

Figure 5.20 Infarct Sizes are Larger in miR-21-/- Mice ...... 134

Figure 5.21 Light Decreases Infarct Size After Myocardial Ischemia and Reperfusion Injury .. 135

Figure 5.22 MiR-21 is Necessary for Light-Elicited Cardioprotection ...... 136

xvi Figure 5.23 Differences in Structures of Flavonoids Used in Ischemia Reperfusion Injury ...... 137

Figure 5.24 The Effect of Flavonoids on Infarct Size ...... 138

Figure 5.25 Nobiletin Reduces Troponin-I Levels After Ischemia and Reperfusion Injury ...... 139

Figure 5.26 Nobiletin Decreases Infarct Size After Myocardial IR-Injury ...... 140

Figure 5.27 Nobiletin Rescues the Deleterious Effect of Midazolam on Cardiac PER2 Gene

Expression and Troponin-I Levels After Myocardial IR-Injury ...... 141

Figure 5.28 Nobiletin Rescues the Deleterious Effect of Midazolam on the Infarct Size After

Myocardial IR-Injury ...... 142

Figure 5.29 Nobiletin Reverses Midazolam Induced ROS During Myocardial IR-Injury ...... 143

Figure 6.1 Light Exposure Increases Buccal and Plasma PER2 in Humans...... 148

Figure 6.2 Light Exposure in Humans Increases Plasma PER2 Protein Levels ...... 149

Figure 6.3 Daylight Exposure Suppresses Plasma Melatonin More Than Room Light ...... 150

Figure 6.4 Light Exposure in Humans Targets Plasma PFK Activity ...... 151

Figure 6.5 Light Exposure in Humans Increases LDH Activity ...... 152

Figure 6.6 HIF1⍺ is Stabilized Upon Light Exposure in Humans ...... 153

Figure 6.7 Light Exposure Decreases Triglycerides ...... 154

Figure 6.8 Light Exposure Does Not Alter Plasma Glucose Levels ...... 155

Figure 6.9 Targeted Metabolomics from Humans Exposed to Light ...... 156

Figure 6.10 Key Glycolytic and TCA Cycle Metabolites from Targeted Metabolomics ...... 157

Figure 6.11 Targeted Metabolomics ...... 158

Figure 6.12 Targeted Metabolomics ...... 159

Figure 6.13 Targeted Metabolomics ...... 160

Figure 6.14 The Effect of Light Exposure on Sleep ...... 161

Figure 6.15 Light Exposure Entrains Human Subjects ...... 162

Figure 6.16 Light Exposure Blood Collection Timeline for miR-21 Studies ...... 163

Figure 6.17 Effects of Intense Light on MiR-21 Regulation in Human Subjects ...... 164

xvii Figure 7.1 Timeline of Enrolling and Obtaining Sample from Clinical Trial Patients ...... 168

Figure 7.2 Timeline of Blood Draws from Healthy Human Volunteers ...... 169

xviii LIST OF ABBREVIATIONS

A2B Adenosine A2B receptor (gene: ADORA2B)

ACO2 Aconitase 2

ARNT HIF1B bHLH Basic helix-loop-helix domain

BMAL1 Aryl hydrocarbon receptor nuclear translocator-like protein 1 cAMP Cyclic adenosine monophosphate

CD39 Ectonucleoside triphosphate disphosphohydrolase-1

CD73 Ecto-5’-nucleotidase

CK1 Casein kinase 1ẟ or ε

CLOCK Circadian locomotor output cycle kaput

COX4 Cytochrome c oxidase, subunit 4, isoform 1 or 2

CRE cAMP response element

CREB cAMP response element-binding protein

CRY1/2 Cryptochrome1, Cryptochrome2 protein

GABA Gamma-aminobutyric acid

GABAA GABA receptor A

GOF Gain of function

HIF1 Hypoxia inducible factor complex (HIF1⍺:HIF1β)

HIF1⍺ Hypoxia inducible factor 1 alpha protein

HIF1β Hypoxia inducible factor 1 beta, ARNT, protein

HIF2⍺ Hypoxia inducible factor 2 alpha

HIF1A Hypoxia inducible factor 1 alpha gene or mRNA

HIF1B Hypoxia inducible factor 1 beta gene or mRNA

HMEC-1 Human microvascular endothelial cells

HRE Hypoxia response element

xix ICU Intensive care unit

IDH2 Isocitrate dehydrogenase 2

IPC Ischemic preconditioning

IR Ischemia and reperfusion

K-ATP ATP-sensitive potassium channel

KD Knockdown

L:D Light:dark cycle

LDH Lactate dehydrogenase

LDL Low-density lipoprotein

LED Light-emitting diodes

LOF Loss of function

LPS Lipopolysaccharide

LUX Unit of illuminance

MI Myocardial infarction

NADH, NAD+ Nicotinamide adenine dinucleotide

NAMPT Nicotinamide phosphoribosyltransferase

NOB Nobiletin

PAS PER, ARNT, SIM domain

PER1/3 Period1 or Period3 proteins, paralogs of Period2

PER2 Period2 protein

PFK Phospho-furctose kinase

PHD Prolyl hydroxylase domain protein

PKC-e Protein kinase C epsilon

PKM Pyruvate kinase

PPAR Peroxisome proliferator-activated receptor, gamma/alpha

PPP Pentose phosphate pathway

xx pVHL von Hippel-Lindau tumor suppressor

REV-ERB Nuclear receptor subfamily 1, group D, member 1 or 2 (NR1D1 or NR1D2

gene), REV-ERB⍺/β protein

RISC RNA induced silencing complex

ROR Retinoic acid receptor-related orphan receptor protein, alpha/gamma

RORE ROR binding element

ROS Reactive oxygen species

SCN Suprachiasmatic nucleus

SIRT Sirtuin (1 – 7)

SUCLG1 Succinate-CoA ligase

TCA Tricarboxylic acid cycle tRNA Transfer RNA

ZT Zeitgeber time

xxi CHAPTER 1 CHAPTER I INTRODUCTION

1.1 The Circadian-Hypoxia Link

Cyanobacteria and the Great Oxygenation Event

The evolutionary link between light- and oxygen-sensing pathways provides insight into

molecular and cellular adaptation for resilience to adverse changes in the environment. In fact,

the appearance of sunlight and oxygen on earth were undoubtedly the most dramatic environmental changes during evolution. A crucial factor in the evolution of adaptation to sunlight and oxygen was cyanobacteria. These prokaryotes were some of the first organisms to appear on Earth approximately 2.7 billion years ago, which was almost two billion years after our sun

formed (1). When cyanobacteria appeared on Earth, these photosynthetic organisms had to adapt

to consistent oscillations of sunlight and darkness. In doing so, cyanobacteria that developed an

internal clock were preferentially selected throughout evolutionary time; cyanobacteria express

photosynthetic genes in anticipation of a light phase to optimally target metabolic processes

during a time when solar resources would be available compared to an inopportune time when it

would be energetically wasteful to express genes involved in photosynthesis (2, 3).

The photosynthetic property of cyanobacteria led to the by-product accumulation of oxygen in the atmosphere, which at the time contained minimal to no oxygen. Much of the initial free oxygen released by cyanobacteria was absorbed by oceans and rock until oxygen gas maxed out these formations. Next, free oxygen was absorbed by the land until about 850 million years ago when oxygen gas accumulated in the atmosphere to unprecedented levels, triggering The

Great Oxygenation Event (1) in which all anaerobic organisms unable to adapt to this oxygen presence, experiencing oxygen as a toxin, were killed off. Subsequently, all organisms that appeared on Earth thereafter were required to use oxygen for basic cellular processes. As a result, almost all organisms on this planet are equipped with light- and oxygen-sensing pathways

1 (1-5). In the present day, oxygen is frequently used in the clinical setting, however not as much attention has been given to light exposure.

Light-sensing pathways: the mammalian circadian system and feedback loop

Light exposure is a primary mechanism to synchronize a circadian system to the environment (circadian entrainment) (6). Sunlight is a ‘Zeitgeber’ (external cue for circadian rhythm) and the primary mechanism of entrainment. Without it, biological systems revert to their internal ‘free-running’ clock, which is generally slightly shorter or slightly longer than the 24-hour day. Zeitgebers like sunlight serve to synchronize the internal circadian clock to the environmental clock and maintain a 24-hour oscillation. In mammalian systems, photic stimuli enter the retina and travel via the retinohypothalamic tract to the suprachiasmatic nucleus (SCN) in the brain, where the signals are transduced to the molecular clockwork (7, 8). Blue wavelengths of light are detected specifically by melanopsin receptors in retinal ganglion cells that leads to the transcriptional induction of PER2 in the SCN and concomitant entrainment. Blue light at 470 nm shifts the circadian phase in a melanopsin-dependent manner (9). Only light with an intensity

>180 LUX is able to synchronize the human circadian system (10), whereas intense light (>10,000

LUX) is most effective. Chronic disruption of circadian oscillators is considered a significant predisposing factor in many prevalent diseases including sleep disorders (11-13), bipolar affective disorder or schizophrenia (14, 15), cardiovascular disease (16-22), onset of myocardial infarction

(MI) (6, 23-27) (28), cardiac arrest (29), stroke (30), immune response or sepsis (31, 32), cancer, diabetes, and metabolic syndromes (obesity, diabetes, endocrine, and hormonal function) (33-

37). The broad impact of circadian disruption on human health compounded by the lack of understanding regarding circadian mediated mechanisms is a continuing endeavor for many researchers. Deciphering mechanisms of circadian disruption may shed light into preventative or therapeutic treatment strategies for human diseases.

The evolution of molecular and cellular light-sensing pathways was an adaptive mechanism to coordinate with the environment’s consistent oscillations of sunlight and darkness.

2 At the cellular level in mammals, circadian rhythm consists of a molecular negative feedback loop and four core clock proteins: circadian locomotor output cycle kaput (CLOCK, CLOCK), aryl hydrocarbon receptor nuclear translocator-like protein 1 (BMAL1, ARNTL), period 1/2 (PER1/2,

PER1/2), and cryptochrome 1/2 (CRY1/2, CRY1/2) (38). In an approximate 24-hour cycle,

PER1/2 and CRY1/2 are expressed with protein levels accumulating in the cytoplasm gradually throughout the day, reaching a peak in the late afternoon. In the cytoplasm, PER1/2 and CRY1/2 interact with casein kinases (CK) 1ẟ and CK1ε. Phosphorylation of the clock proteins is considered to regulate subcellular localization and stability with a unique ability to lengthen the circadian phase (39-41). In the evening, PER1/2 and CRY1/2 form a heterodimer and translocate back into the nucleus where PER1/2 and CRY1/2 act as repressors of BMAL1 and CLOCK transcriptional activity (38). The anti-phase clock proteins BMAL1 and CLOCK form a heterodimer in the nucleus and bind to E-box response elements in the promoter regions of PER1/2 and

CRY1/2 genes, driving their expression. The BMAL1 and CLOCK heterodimer also serve as transcriptional activators of the nuclear receptors ROR (retinoic acid receptor-related orphan receptors) and REV-ERB (nuclear receptor subfamily 1 group D member 1/2), which are rhythmically expressed. RORs and REV-ERBs competitively bind at the RORE (ROR binding element) in the promoter region of genes including BMAL1, CLOCK, RORA/G, and

NR1D1/NR1D2 (42, 43). REV-ERB⍺/β are transcriptional repressors and inhibit RORs from binding to promoter region ROREs. In addition, REV-ERBs repress BMAL1 gene expression.

ROR⍺/� are transcriptional activators of BMAL1 and NR1D1/2 (gene for REV-ERB⍺/β) expression. REV-ERBs even have the ability to autoregulate since their promoter region contains

ROREs (38, 42, 43) (Figure 1.1). This template of the primary negative feedback loop occurs in organisms from cyanobacteria to plants to humans (38) with many factors feeding into this 24- hour oscillation to regulate input and output functions. Such factors include the aforementioned nuclear receptors and post-translational modifiers, and additional components include light

3 exposure, secondary signaling molecules, oxygen fluctuation, and microRNAs, which be

addressed in ensuing sections.

Oxygen-sensing pathways: hypoxia inducible factors

In addition to light-sensing circadian pathways, organisms that evolved after The Great

Oxygenation Event preserved the adaptive ability to sense environmental oxygen levels in the

form of hypoxia inducible factors (HIFs) (44-47). Physiologic hypoxia and the function of proteins

important in hypoxia have a wide variety of functions including a necessity for embryonic

development and cell differentiation, yet also a driver of tumor cell proliferation and tumor cell

metabolism (48). Hypoxia is therefore a crucial component of normal cellular function, but the

machinery can be hijacked under certain circumstances leading to disease in humans. The ability

of cells to adapt to hypoxia is an evolutionarily conserved, innate, and adaptive response that is

advantageous for almost all organisms to rely on when environmental conditions demand the

cells do so for survival.

HIFs consist of two subunits, an oxygen-regulated subunit (HIF1⍺ or HIF2⍺) and a

constitutively active subunit HIF1β/ARNT (in complex collectively referred to as HIF, HIF1, or

HIF2). HIF1⍺ is regulated by the oxygen sensing prolyl hydroxylase domain proteins (PHDs). In normal oxygen environments (normoxia), PHDs sense the oxygen status of the cell by using iron as a co-factor to reduce oxygen to carbon dioxide in the prolyl hydroxylation of HIF1⍺. This post-

translational modification allows for HIF1⍺ ubiquitination by von Hippel-Lindau tumor suppressor

(VHL) and subsequent HIF1⍺ degradation. Due to the iron-dependent nature of PHDs, lack of

oxygen inhibits the functionality of PHDs, leading to HIF1⍺ stabilization. In hypoxia, stabilized

HIF1 enters the nucleus where it binds to hypoxia response elements (HREs, consensus

sequence 5’-RCGTG-3’) in the promoter of genes that are crucial for adaptation to hypoxia. Many

of these targets include genes expressing metabolic enzymes important for regulating metabolic

balance in low oxygen conditions and in fact, certain metabolites and reactive oxygen species

4 (ROS) produced by metabolic pathways in hypoxia inhibit PHDs (Figure 1.2) (49-51). HIF1⍺

dependent hypoxia-sensing pathways are evolutionarily essential for environmental adaptability

to changes in oxygen availability and redox status of the cell. Understanding HIF1⍺ pathways

may be a beneficial approach to overt diseases that arise from hypoxia (44, 45, 47, 52-61).

PAS domains: an evolutionarily link between circadian and hypoxia pathways

Light- and oxygen-sensing pathways are linked on a cellular level in mammals (45, 51, 55,

62, 63). HIF1⍺ belongs to the same protein family as the core circadian proteins, the PAS domain superfamily of signal sensors for oxygen, light, or metabolism (45, 51, 55, 63, 64). PAS domain- containing proteins are present in all kingdoms of life including Bacteria, Archaea, and Eucarya, and are present in a vast array of proteins and enzymes including kinases, chemo- and photoreceptors, circadian clock proteins, ion channels, cyclic nucleotide phosphodiesterases, hypoxia response proteins, and proteins involved in embryological central nervous system development (63). Originally named for the first group of proteins found to contain these domains

(PER, ARNT, and SIM), PAS domains are 250 – 300 amino acids in length, homologous in

sequence, and the proteins containing PAS domains are a versatile group capable of sensing

light, oxygen, redox status, small ligands, and energy demands of a cell. PER originates from the

Drosophila PER gene identified as a key circadian rhythm regulator, ARNT was named for the

aryl hydrocarbon receptor nuclear translocator protein (ARNT gene) known as the HIF1β subunit, and SIM comes from the Drosophila single-minded locus discovered for its role in regulating midline cell lineage (51). PAS domain proteins contain basic helix-loop-helix (bHLH) motifs and

are involved in a variety of cellular processes including signal transduction pathways (63) and

DNA binding as transcription factors (51). Furthermore, some PAS domains can be used for

protein-protein interactions. In particular, the PAS domains in PER proteins were found to be involved in protein-protein interactions by forming homodimers or heterodimers with other PAS domain-containing proteins, such as SIM and ARNT (51, 55, 58, 64). The shared PAS domains

5 between PER and HIF1⍺ suggest that their functions may be similar and perhaps have an

overlapping function.

1.2 PER2 and HIF1⍺ in Metabolic Adaptation to Myocardial Ischemia

Adenosine and ischemic preconditioning link PER2 and HIF1⍺

Cardiovascular disease is the leading cause of death worldwide, causing approximately

one death every 40 seconds in the United States, and therefore demands innovative research

initiatives to develop novel therapeutic strategies for prevention or treatment (65). A key target in the prevention of myocardial ischemia, which may provide impactful long-term effects on cardiovascular health, is metabolic adaptation to low oxygen availability. In this regard, oxygen deficiency resulting from myocardial ischemia requires the heart to metabolically adapt to balance energy production with oxygen consumption.

In the 1960s, Dr. Robert Berne proposed the ‘Adenosine Hypothesis’ that postulated adenosine to be cardioprotective by balancing oxygen supply and oxygen demand during myocardial ischemia (66). This hypothesis was initially received with skepticism until the 1970s, when adenosine receptors were identified and characterized, elucidating potential cardioprotective effects of adenosine (67). Subsequently in the 1980s, Dr. Charles Murry found that brief (5 min) intervals of ischemia and reperfusion (a procedure termed ischemic

preconditioning (IPC)) were cardioprotective from the traditional myocardial ischemia and

reperfusion (IR)-injury model (68). Within the last decade, our lab revealed that the

ectonucleotidases CD39 and CD73, the enzymes that break down extracellular ATP to AMP and

AMP to adenosine, are essential for the cardioprotective effect of IPC (69) (70). Furthermore,

investigations into adenosine signaling during IPC identified a unique role for the adenosine A2B

receptor to be necessary for cardioprotection (70) (Figure 1.3). Studies diving into the

mechanisms of cardioprotection through adenosine signaling at the A2B receptor uncovered the

circadian rhythm protein Period2 (PER2) as a downstream target through both an upregulation of

cAMP, phospho-CREB, and PER2 transcript, and an inhibition of PER2 protein degradation via

6 Cullen deneddylation (71) (Figure 1.3). Indeed, peripheral tissues like the heart display

oscillations in PER2 expression similar to those of the brain (71, 72) and are thought to be secreted through neurohormonal signaling molecules (7, 73, 74) or adenosine, as circulating 5’-

AMP and adenosine were found to be diurnal in nature (73) with an established role for adenosine signaling in peripheral tissues like the heart (71).

Key mechanistic studies identified an association between PER2 and HIF1⍺ with a necessity of PER2 to stabilize HIF1⍺ in hypoxia (71). Due to the known role of HIF1⍺ in metabolic adaptation to hypoxia, subsequent studies determined PER2 to be a regulator of carbohydrate metabolism and to be cardioprotective during in situ myocardial ischemia and reperfusion (IR)- injury (71). The critical role of HIF1⍺ in cardioprotection from ischemic preconditioning (IPC) has been elucidated by numerous investigators and sheds light on additional mechanisms of adaptation to low oxygen availability resulting from ischemia. For example, one research group took isolated perfused hearts from mice that were exposed to a preconditioning protocol of no- flow ischemia for 5 minutes and 5 minutes of reperfusion and found protection from a prolonged

IR phase that was abolished in the hearts from HIF1A+/- mice (75). These findings were supported by cardiac small interfering RNAs for HIF1A expression that also inhibited cardioprotection during

IPC (76). Mechanisms into HIF1⍺ mediated cardioprotection uncovered that the loss in cardioprotection from IPC in HIF1A+/- mouse hearts can be rescued by adenosine perfusion during

IPC (75). Additionally, gene expression of the ecto-5’-nucleotidase CD73 is under the control of

HIF1⍺ in hypoxia (77) and CD73 deficient mice do not demonstrate cardioprotection from IPC

(70). Because adenosine perfusion provides cardioprotection from IR-injury in HIF1A+/- mouse hearts (75), these data suggest that adenosine production is dependent on HIF1⍺ (78) and may provide additional mechanisms and pathway targets for optimizing HIF1⍺ dependent cardioprotection from low oxygen conditions by manipulating circadian PER2.

7 PER2 regulation of HIF1⍺

Studies deciphering the connection between PER2 and HIF1⍺ confirmed the hypothesis that PER2 is an upstream regulator of HIF1⍺. PER2-/- mice express less HIF1A mRNA, which

resembles less HIF1A gene expression found in ADORA2B-/- mice compared to wildtype.

Furthermore, HIF1⍺ failed to be stabilized in ADORA2B-/- mice upon oxygen depletion and these

mice were unable to induce gene expression of glycolytic enzymes during IPC, an effect that

similarly resembles that seen in PER2-/- mice (71). Specifically, cardiac HIF1⍺ protein oscillates

in a circadian pattern along with protein levels of two HIF1⍺ downstream transcriptional targets in

the glycolytic pathway, pyruvate dehydrogenase kinase and lactate dehydrogenase. Notably,

PER2 was necessary for this HIF1⍺ oscillation and without PER2, HIF1⍺ is not stabilized during

IPC of the heart (71). These prior experiments together suggest that: 1) PER2 is dependent on

A2B signaling, 2) HIF1⍺ is dependent on PER2 for glycolytic gene expression, and 3) PER2 may

be a master upstream regulator of metabolic adaptation to low oxygen conditions (71).

In support of a circadian oscillation pattern of cardiac HIF1⍺ protein levels dependent on

PER2 (71), another research group reported the necessity of the circadian clock for the ability of

HIF1⍺ to sense cellular oxygen status (79). In fact, circadian clockwork has been found to regulate

HIF1⍺ nuclear protein level localization and oscillation in mouse brain and kidney (64). In a direct link between circadian and hypoxia pathways, HIF1⍺ was found to work in conjunction with

BMAL1 at the HRE in the PER2 promoter and the binding enhanced the amplitude of PER2 oscillation in skeletal muscle (79). Additionally, investigational ChIP-seq studies revealed an E- box in the promoter for the HIF1A gene that was transcriptionally driven by binding of the BMAL1-

CLOCK complex (80). There is a growing body of evidence supporting a bi-directional link between circadian and hypoxia pathways, including PER2 dependent HIF1⍺ binding to HREs in promoter regions of hypoxic target genes from in vitro experiments in HeLa cells (81) to zebrafish experiments discovering hypoxia mediated expression of the PER2 gene through regulating

8 HIF1⍺ binding at E-box regions in promoters (82). Moreover, investigations into this circadian-

hypoxia link revealed daily rhythmic oxygen levels in the blood and tissue of rodents, within a

physiologic range (5 – 8 % oxygen), that peaked during the dark phase compared to the light

phase. Mimicking rhythmic oxygenation fluctuations in vitro was sufficient to reset the circadian

clock by synchronization in a HIF1⍺-dependent fashion (63, 83). Together, these studies support

the hypothesis that PER2 is an effector protein for HIF1⍺ and suggest a role for HIF1⍺ in the core

circadian loop and a bi-directional relationship between light- and oxygen-sensing pathways.

Regulation of energy metabolism during hypoxia or ischemia

Balance of energy metabolism (Figure 1.4) is a critical regulatory component during

hypoxia since the ability to switch between oxygen-consuming metabolic pathways and pathways

that do not require oxygen is vital for adaptation to and survival from low oxygen availability.

Glucose and fatty acids, two cellular fuel substrates, have strikingly different oxygen demands

and high-energy phosphate yield, in the form of ATP, from catabolism. Glucose metabolism

generates pyruvate for either oxidation in the mitochondria (aerobic glycolysis under normoxic

conditions) or lactate (anaerobic glycolysis in hypoxia, or ‘oxygen-efficient’ due to the fact that it

does not require oxygen). Six moles of oxygen are used for oxidation of one mole of glucose and

generates 38 high-energy phosphate bonds (84). Conversely, 31 moles of oxygen are used for

oxidation of the fatty acid palmitate and generates 129 high-energy phosphate bonds (85). This

results in glucose oxidation creating 6.3 high-energy phosphate bonds per mole of oxygen while

fatty acid oxidation generates 4.1 high-energy phosphate bonds per mole of oxygen. Glucose

oxidation therefore produces 53.7% more ATP for each mole of oxygen consumed than the

oxidation of the fatty acid palmitate (85). Data consistently point toward a greater yield of high- energy phosphates per mole of oxygen through glucose oxidation compared to palmitate oxidation, although some discrepancies exist: an independent study found a 12% increase in ATP between the two metabolic pathways (85) (84). Nonetheless, the key takeaway from these studies is in the comparison of ATP generated per gram of glucose versus per gram of palmitate: fatty

9 acid metabolism yields significantly more ATP per gram of palmitate compared to the ATP

generated per gram of glucose. When oxygen availability is plentiful, fatty acid metabolism is

therefore more energy efficient, but when oxygen availability is limited, a rapid and efficient

balance between metabolic pathways is beneficial, if not necessary (85). Balancing metabolic pathways in hypoxia is of dire importance due to oxygen being used as a terminal electron acceptor in oxidative phosphorylation (86). Electrons are shuttled through the mitochondrial membrane during respiration and excessive ROS production during periods of hypoxia in cells that are not adequately adapted, will quickly become detrimental (87, 88). Taken together, the balance between glucose or fatty acid oxidation will depend on oxygen availability and is therefore critically important for organs like the heart to manipulate during periods of hypoxia or ischemia.

Fuel source preference is cell-type specific and metabolic pathway usage is generally

determined by the cell’s function. Cardiac myocytes rely primarily on fatty acids to fuel oxidative

phosphorylation so it’s beneficial for the ischemic heart to adapt by optimizing metabolic pathways when oxygen availability is limited. Cardiac myocytes are not the most abundant cell type by number, but rather have the greatest volume compared to other cell types that make up the heart

(89, 90), which include fibroblasts, endothelial cells, stem cells, and inflammatory cells (89).

Notably, endothelia are considered to have the primary role in myocardial IR-injury (91-94) and

are considered oxygen sensors due to their ability to tolerate a range of oxygen tensions and their

high sensitivity to IR-injury (92, 95-98). Additionally, endothelia have the ability in hypoxia to

generate oxygen-independent ATP and use the Pentose Phosphate Pathway (PPP, an anabolic

pathway branched off of the glycolytic pathway) to generate NADPH and nucleic acid precursors

(99). Endothelial cells are indeed flexible and adaptive in their metabolic pathway usage to meet

cellular needs with regards to energy demand and redox status and differences persist between

endothelia based on oxygen exposure, which varies between arterial, venous, and microvascular

endothelia (100). While there exists a debate in the field as to whether endothelial cells or

fibroblasts make up the majority of the heart’s cellular composition, this discrepancy is most likely

10 due to limitations in detection methods with estimates ranging from ~60% endothelial cells and

~20% fibroblasts to ~6% endothelial cells and ~64% fibroblasts (89, 90). Regardless of total composition, most researchers are in agreement that cardiac fibroblasts are key factors in fibrosis,

remodeling, and arrhythmogenesis (90) while endothelial cells are directly correlated with overall cardiac function outcomes after myocardial IR-injury (92, 101, 102). Managing cellular metabolism when oxygen availability is limited serves as a critical mechanism for survival and many studies point toward the interplay between circadian and oxygen pathways to be the core of this regulation.

HIF1⍺ mediated metabolic adaptation to hypoxia

HIF1⍺ is well established in its role for metabolic adaptation to hypoxia or ischemia and

research in this field has identified several mechanisms of HIF1⍺ mediated balance between

glycolytic and oxidative metabolism. Cells adapt to acute or prolonged hypoxic environments by

regulating metabolic pathways and activating HIF1⍺. HIF1⍺ stabilization and function is

determined by oxygen availability and in turn, HIF1 promotes cellular homeostasis in terms of

redox potential and energy production from metabolic pathways (103). While the TCA cycle isn’t

a direct consumer of oxygen, it is tightly coupled to oxidative phosphorylation and HIF1 regulates

flux into the TCA cycle to mediate downstream oxygen consumption during hypoxia. In normoxia,

+ the TCA cycle reduces NAD and FAD to NADH and FADH2, which are oxidized by oxidative

phosphorylation. However, during hypoxia, strain on mitochondrial respiration prevents NADH

and FADH2 from being readily oxidized. Therefore, when oxygen availability is limited, one

mechanism of HIF1 mediated metabolic adaptation involves shuttling metabolites away from the

mitochondria to regulate TCA cycle flux and oxidative phosphorylation to overall reduce ROS

production. In concordance, mouse embryonic fibroblasts subjected to long-term hypoxia

revealed HIF1 binding at an HRE in the promoter region the gene encoding pyruvate

dehydrogenase kinase 1, which is responsible for phosphorylating and inactivating the catalytic

subunit of pyruvate dehydrogenase, thereby inactivating the conversion of pyruvate to Acetyl-CoA

11 (104). In general, HIF1 is best described for upregulating its downstream gene targets including

those involved in glucose metabolic pathways and glucose transporters to promote an increased flux of glucose reliance (104). Essentially, HIF1 works to shift the balance between metabolic pathways by inhibiting pyruvate from entering the TCA cycle and subsequent oxidative phosphorylation. This potentially counteracts ROS that would be produced from the overuse of mitochondrial respiration in hypoxia (78).

Another mechanism of HIF1 mediated adaptation to hypoxia is by targeting oxidative

phosphorylation. Complex IV (COX4, cytochrome c oxidase subunit 4) of oxidative

phosphorylation has a HIF1 inducible subunit that is necessary for adaptation to low oxygen

availability. In normoxia, COX4.1 is the dominant subunit but in hypoxia, COX4.2 is upregulated

by HIF1 via an HRE in the promoter region. Additionally, the COX4.1 subunit is degraded in a

HIF1 dependent manner by the HIF1 mediated upregulation of the mitochondrial protease LON

(105). The significance of HIF1 mediated “subunit switching” between COX4.1 and COX4.2 to

optimize oxidative phosphorylation in hypoxia is most likely because mitochondrial respiration

cannot completely shut down, but rather optimizes how it uses oxygen. COX4 is the primary

oxygen consumer and site of electron transfer to oxygen, but mitochondrial strain in hypoxia can

drive electron transfer to oxygen prematurely at Complex I or Complex III, which significantly

increases ROS production (105). This suggests regulating and optimizing mitochondrial

respiration at the level of COX4 as another key target for hypoxic adaptation.

PER2 and metabolic regulation

The circadian clockwork is widely appreciated for its intricate role balancing metabolic

pathways and reciprocally, metabolic by-products serve as signaling cues for the circadian

machinery. Many investigations into the innerworkings of the circadian system and metabolic

homeostasis have revealed circadian biology as a key regulator of energy metabolism in many

organ systems and tissues (Table 1.1). In a mouse model of circadian misalignment where mice were housed in non-24-hour light:dark (L:D) cycles, researchers found significant physiological

12 disturbances and in particular, poorer metabolic efficiency, altered substrate utilization, and a

greater depression of cardiac function (106). In mouse liver, high temporal resolution metabolite

profiling revealed more than 50% of detected metabolites were regulated by the clock (107).

Furthermore, circadian gene expression oscillation is in part controlled by cellular redox status and NADPH, and in reverse, redox status and the PPP are sufficient to maintain circadian amplitude. These data indicate a bi-directional feedback system between the circadian clock and metabolic homeostasis, rather than a hierarchy of single-directional downstream effects (108).

The NAD+-dependent deacetylases, sirtuins (SIRT), provide a direct connection between metabolic pathways and the circadian clock. While sirtuins were first described by their nuclear targets for silencing gene expression by histone deacetylation, sirtuins are now appreciated for more than their nuclear role and in fact have broad applications in many cellular processes considering they are also cytoplasmic and mitochondrial proteins. In mammals, there are seven members of the sirtuin family, named SIRT1-7. Amongst them, they differ in tissue expression, subcellular localization, enzymatic activity, and target proteins (SIRT1, 6, and 7 in the nucleus,

SIRT2 in the cytoplasm, and SIRT3, 4, and 5 in the mitochondria) (109). Nuclear sirtuins are responsible for rhythmic histone acetylation, representing circadian control of sirtuin activity. In turn, nuclear sirtuins regulate circadian clock gene expression. Specifically, the promoter regions of PER1, PER2, and CRY1 have rhythmic histone-3 acetylation and subsequent rhythmic RNA polymerase II binding (110). This may be due in part to the competitive inhibition of sirtuins by nicotinamide, the product of NAD+ transformation by sirtuins, indicating a negative feedback loop for its own regulation (111). In addition, as a cofactor for sirtuins, NAD+ regulates sirtuin enzyme activity. Considering NAD+ levels increase during cellular stress, sirtuins are considered an adaptive mechanism by regulating downstream gene targets. NAD+ biosynthesis is rhythmic and oxidative enzymes in the mitochondria are subject to rhythmic acetylation (112, 113).

Furthermore, biosynthesis of NAD+ is performed by the rate limiting enzyme NAMPT, which also oscillates in a circadian manner. Inhibiting NAMPT reduces SIRT1 suppression of the CLOCK-

13 BMAL1 complex and results in an increase in PER2 oscillation. To complete this loop, CLOCK

transcriptionally regulates NAMPT gene expression, suggesting that NAMPT and NAD+ are both

part of the core circadian feedback loop (114). In the cytoplasm, the circadian rhythm protein

PER2 undergoes rhythmic lysine acetylation and deacetylation by SIRT1, which regulates robust

circadian oscillation of core clock genes and their protein activity (115). The circadian deacetylase

activity of SIRT1 is dependent on the metabolic by-product and circadian controlled NAD+ and

SIRT1 deacetylation of PER2 results in its degradation in a circadian manner, establishing a relationship between PER2 and cellular metabolism (86, 115).

Mitochondrial SIRT3 is expressed mostly highly in tissues that are very metabolically

active, including the brain, heart, liver, brown adipose tissue, and skeletal muscle (111, 116-118).

In particular, SIRT3 targets the TCA cycle and mitochondrial enzymes to deacetylate and regulate their activity (119-123). These mitochondrial enzymes are involved in fatty acid oxidation, amino acid metabolism, oxidative phosphorylation, and antioxidant defenses (111). SIRT3 deacetylates and therefore activates both isocitrate dehydrogenase 2 (IDH2) and succinate dehydrogenase

(SDH), the latter of which is functional in both the TCA cycle and oxidative phosphorylation (111,

122). To protect against cellular ROS, SIRT3 depends on IDH2 (121). Importantly, SIRT3 is the only mitochondrial sirtuin responsible for mitochondrial protein deacetylation, as lack of SIRT3 in mice results in hyperacetylation of mitochondrial proteins that is not exhibited in SIRT4 or SIRT5 deficient mice. In addition, the SIRT3 deficient mice, but not SIRT4 or SIRT5 deficient mice, presented with dire downstream metabolic defect (116, 124). These metabolic defects in SIRT3 deficient mice include metabolic syndrome, cancer, and cardiac failure (111, 125, 126). Lastly, one study found a cardioprotective role of SIRT3 in an isolated ischemia and reperfusion injury model (124) and that SIRT3 protects cardiomyocytes from oxidative stress (127).

Genetic PER2 deficient mice provide insight into metabolic defects and the role of PER2 in metabolic pathways. PER2 appears to be a key regulator of white versus brown adipose tissue, the latter of which relies heavily on oxidative metabolism: PER2-/- mice have reduced adiposity

14 and the lack of PER2 affects gene expression such that white adipose tissue begins expressing

genes involved in metabolic pathways that are usually only expressed in brown adipose tissue.

These data suggest that without PER2, the adipose tissue now more closely resembling oxidative

brown adipose tissue runs oxidative metabolism at a higher rate (128-130). This phenotype may be due to the interaction of PER2 with the nuclear receptors PPAR�, PPAR⍺, and REV-ERB⍺,

which regulate cellular metabolism in white adipose tissue, liver, and the heart (42, 129). PER2

directly inhibits the nuclear receptor PPAR�, which has a known role in adipogenesis and insulin

sensitivity, by preventing it from reaching the promoter and thereby decreasing transcriptional

activity of its targets (129). Additionally, the REV-ERB⍺ targets in the liver include genes involved

in glucose metabolism. Furthermore, the interaction of PER2 with REV-ERB⍺ results in rhythmic

expression of these genes (42, 131, 132). Lastly, PER2 was found to regulate gluconeogenesis

and glycogen catabolism by working in concert with nuclear receptors, a mechanism seen in both

liver and the heart (42, 133, 134).

Recent investigations found a necessity of PER2 in cardioprotection by regulating metabolic pathways. Studies from in vivo stable isotope glucose tracers during baseline, myocardial ischemia, or myocardial IR-injury, revealed the inability of PER2-/- mice to rely on glycolysis during ischemia, as indicated by 13C-fructose-1,6-bisphosphate, 13C-pyruvate, and 13C- lactate concentrations. Additionally, PER2-/- mice had an increase in flux of the TCA cycle during ischemia whereas wildtype mice attenuated TCA cycle flux, confirming the role of PER2 in metabolic reliance in low oxygen environments (71). In a comparison between PER2-/- and wildtype mouse hearts by high-throughput gene array analysis during IPC (135), lipid metabolism pathways were found to be predominately dysregulated. This finding was supported by nuclear magnetic resonance imaging identifying altered fatty acid accumulation in the heart of PER2-/- mice (136) and electron microscopy analysis of PER2-/- mouse hearts that revealed mitochondrial swelling and glycogen accumulation, indicators of severe metabolic defects (71). Investigations into potential PER2 mediated pathways began with gene array analysis of PER2-/- mouse hearts

15 compared to wildtype mouse hearts during myocardial IR-injury and revealed dysregulated

protein phosphatase 1, a key regulator of blood-glucose levels and glycogen metabolism (137),

and enoyl-CoA hydratase, the enzyme in the latter end of β-oxidation that makes the precursor to

Acetyl-CoA for the TCA cycle (136, 138, 139). In addition, PER2-/- mouse hearts exhibited a

reduction in long chain fatty acids and increases in carnitine palmitoyltransferase 1 protein during

myocardial ischemia, indicating that PER2 is an important regulator of enyol-CoA hydratase and

therefore fatty acid metabolism in the heart (71). Furthermore, PER2-/- mice maintained elevated

levels of cardiac glycogen and elevated glycogen synthase 1 protein compared to wildtype mice

prior to ischemia, but the PER2-/- mice were unable to restore the glycogen stores following ischemia (71). Together, these data suggest PER2 is an important regulator of glucose and lipid

metabolism and open the door to further investigation into the role of PER2 in downstream

pathways for regulating the metabolic balance in low oxygen conditions (71, 136).

Overlap between PER2 and HIF1⍺ in metabolic regulation

An additional level to the bi-directional relationship between hypoxia and circadian

clockwork includes a connection to sirtuins. SIRT3 was found to be a key regulator of

mitochondrial enzymes so researchers began investigating its function in more detail. These

studies found SIRT3 to be a crucial regulator of the Warburg effect (glycolysis in the presence of

abundant oxygen availability) in cancer cells and it does so by destabilizing HIF1⍺. Without SIRT3,

there’s an increase in ROS and subsequently HIF1⍺ stabilization (140). This was found in human

breast cancer cells that have a reduced level of SIRT3 expression correlated with an increase in

HIF1⍺ regulated genes. SIRT3 overexpression thereby was found to repress glycolysis in these

tumor cells and subsequent breast cancer cell proliferation, which may be a mechanism for

regulating tumor suppression (140). Considering SIRT3 and its co-factor NAD+ are under

circadian control, regulating metabolic pathways may be at the crossroad of the PER2-HIF1⍺

connection. Indeed, the circadian-hypoxia link is intertwined in a reciprocal relationship with

cellular energy metabolism, which is thought to provide optimal and advantageous adaptive

16 mechanisms for changes in the cellular environment throughout evolutionary time. Understanding

these mechanisms may provide insights into mechanistic targets for prevention or treatment from

conditions of low oxygen availability like myocardial ischemia.

The reciprocal relationship between PER2 and HIF1⍺ suggests that while the circadian clockwork controls metabolic pathways, by-products of these metabolic pathways are sensed by the circadian machinery for adaptation (141). Because circadian machinery is manipulated by

light exposure, regulating light may target downstream light-sensing pathways. In fact, the timing

of light exposure markedly affects circadian profiles and specifically PER2 protein concentrations

in the hearts of rats (72) and accordingly, our initial investigations addressed the possibility that

light exposure could function to attenuate myocardial IR-injury by increasing cardiac PER2 levels.

Indeed, we observed a time-dependent induction of cardiac PER2 with significant reduction in

infarct size and troponin-I release following daylight exposure in our IR-injury model in mice (71,

135). These findings led us to investigate mechanisms of PER2 mediated cardioprotection in hypoxia or ischemia. Our results contribute to a growing body of work suggesting an evolutionarily conserved and advantageous bi-directional relationship between light-sensing (circadian) and oxygen-sensing (hypoxia) driven pathways (82) and may provide clinical benefits to improve outcomes for patients with myocardial ischemia. Here, we present our findings from circadian

PER2 mediated metabolic adaptation to low oxygen availability and concomitant cardioprotection from hypoxia or ischemia.

1.3 MicroRNAs in Circadian Biology and Therapies1

MicroRNAs as cardiovascular therapeutic targets

With the emergence of research into the non-coding genome, there has been a great interest in using microRNAs to target various pathways for therapeutic benefit. Only 1% of the human genome codes for genes that function in protein synthesis (143). The remaining 99% of

1 Portions of this section were previously published in Current Pharmaceutical Design (142).

17 DNA was initially considered to be not functional. However, it is now well recognized that several noncoding RNAs have important biological functions (144-152). Among these noncoding RNAs, several subcategories exist including long noncoding and small noncoding RNAs. One type of small noncoding RNAs, microRNAs, have attracted a lot of attention in the past few years (153).

MicroRNAs are short (22 nucleotides), interact with messenger RNAs, and can silence gene expression. The functional domain of microRNAs is the so-called seed region which is only 6-8 nucleotides long (147). Initially, double stranded RNA is formed from a precursor transcript.

Primary microRNAs (pri-microRNA) are transcribed by RNA polymerase II. The pri-microRNAs are 5’ capped, have a stem loop structure, and are 3’ polyadenylated. The canonical biogenesis of a pri-microRNA transcript is cleaved by the endoribonuclease Drosha in the nucleus and subsequently by endoribonuclease Dicer in the cytoplasm. Noncanonical pathways that are independent of this canonical pathway could produce microRNAs from small nucleolar RNA, transfer RNA (tRNA), or Y RNA, as intermediate products. Later, the microRNA duplex unwinds, whereby only the guide strand, which is usually the functional unit, is loaded in the RNA induced silencing complex (RISC). The complimentary strand is often degraded. In the RISC, the microRNA binds to its target mRNA, preventing its translation into a protein. The human genome harbors 1881 microRNA loci that encode for 2588 mature microRNAs (154), indicating an important role in gene regulation. Single microRNAs suppress more than one gene, and microRNAs with similar seed regions may suppress a similar, but non-identical set of genes. Gene suppression is usually partial rather than total (supporting that microRNAs function to maintain cellular homeostasis), and a single gene can have binding sites for multiple microRNAs (144). In general, overexpressing microRNAs via mimetics will suppress target genes, whereas inhibiting an endogenous microRNA will undo the suppression of its target gene. As repression of microRNAs is considered to be safer, current clinical studies mainly use microRNA inhibitors (anti- miRs(144)).

18 Current difficulties for cardiovascular applications include safety issues for systemic

applications: generally high doses are required which can compromise efficacy and safety. So

far, several microRNAs have been identified as cardiovascular therapeutic targets and are

abundantly expressed in the heart (Table 1.2). In particular, miR-21 has been studied extensively in the context of heart fibrosis (155). It was found that in vivo inhibition of miR-21 attenuates the fibrotic response and improves cardiac function in mouse models of heart failure (156). However, these results were not reproduced in a subsequent study using different anti-miRs (157), indicating current challenges in the therapeutic use of microRNAs.

Interestingly, HIF1⍺ was identified as an upstream transcriptional regulator of miR-21 (83).

Moreover, it was shown that HIF1⍺ and PER2 bind together during hypoxia and that PER2 is essential for the regulation of HIF1⍺ target genes during hypoxia (71). In vivo studies confirmed many of these findings and identified miR-21 as a top microRNA in anesthetic preconditioning of the heart (158). Exposure to volatile anesthetics, such as isoflurane, was shown to decrease myocardial infarct size in vivo and increase cell viability after oxidative stress in vitro (159). Using anesthetic preconditioning in miR-21 deficient mice, however, did not mediate any cardioprotective effects (160) but miR-21 was identified as the top microRNA in ischemic preconditioning of the heart (161). As such, ischemic and anesthetic preconditionings of the heart seem to merge on the same microRNA level. Along these lines, a recent review on anesthetic preconditioning indicated that anesthetic and IPC share similar mechanisms (159). Having miR-

21 identified as a major downstream target of anesthetic or ischemic preconditioning of the heart opens up new possibilities to target powerful endogenous cardioprotective mechanisms (162).

Circadian microRNAs

Disease development is strongly influenced by circadian rhythms and therefore

researchers have begun to explore the potential role of microRNAs. Rhythmic control of

microRNA expression appears to be highly conserved and has significant consequences for

circadian timing (163). As such, several rhythmically regulated microRNAs were identified.

19 Surprisingly many biologically relevant time cues, such as the daily light cycle can also drive rhythmic microRNA expression. Studies on Arabidopsis have recently found that miR-167, miR-

168, miR-171 and miR-398 oscillate with higher levels during the daytime than during the night

(164). However, rhythmicity of these microRNAs was not regulated by the circadian clock which became evident when Arabidopsis was transferred to constant light conditions. Under constant light, no oscillation was observed (164). This suggests that light might control these microRNAs.

Within humans, it would not be surprising to observe a similar pattern of microRNA expression driven by exogenous time cues (163). In the mammalian SCN, microRNAs play also an important role in clock timing and entrainment. For example, miR-132 and miR-219 show oscillatory expression in wildtype mice, but not in circadian mutant mice (165). Both microRNAs have CRE- enhancer sequences allowing for CREB-dependent regulation (166), but only miR-132 was shown to be inducible by light (165). Moreover, miR-219 regulates the length of the circadian day and miR-132 modulates the phase-shifting capacity of light (165). As miR-132 is induced by photic entrainment cues via a MAPK/CREB-dependent mechanism, it modulates clock-gene expression, and thereby attenuates the entraining effects of light. In addition, both light responsive miR-132 and clock-regulated miR-219 influence cellular excitability and thereby probably lead to changes in length and phase of the circadian clock. The expression of miR-122, a known regulator of liver metabolism, is also subject to circadian control. While the mature microRNA does not oscillate, both miR-122 precursors have a robust circadian expression pattern (167). In fact, genetic deletion of miR-122 revealed changes in the expression of clock-controlled genes. Interestingly,

REV-ERBα, a key nuclear receptor regulator of the circadian system, also regulates miR-122

(167). Another clock controlled microRNA, miR-142–3p, is controlled by the BMAL1/CLOCK heterodimer and, in turn, can target BMAL1 (168, 169). In a different study, evidence found that miR-155, a proinflammatory microRNA induced by TLRs (Toll-like receptors), controls BMAL1 mRNA and protein levels in myeloid cells, leading to alterations in clock function and circadian control of inflammation (170). Indeed, the free-running period in miR-155-/- mice was shortened

20 when compared to wildtype mice (170). Together, these studies suggest that oscillating microRNAs could function as independent regulators of circadian rhythm dependent gene regulation (Figure 1.5).

MicroRNA control of the circadian system is further supported by studies in Dicer deficient mice, in which microRNA processing is globally compromised (171). These studies demonstrate that RNA interference mediated by microRNAs primarily affects the clock through translational control of PER in the cytoplasm, which delays cytoplasmic PER accumulation and thus generates

a time delay in the circadian feedback inhibition. These studies also identified three microRNAs,

miR-24, 29a and 30a, which affected the circadian clock through regulation of PER1 and PER2

mRNA stability and translation. Thus, these studies indicate microRNAs as key regulators of the

pacemaker genes, PER1 and PER2, and for generating the time delay crucial for the circadian

feedback loop. This evidence positions microRNAs as important and dynamic regulators of

different aspects of circadian rhythms: some microRNAs affect and fine-tune the pace of the core

clock, while others influence the response to external temporal cues as well as behavioral and

peripheral outputs. Therefore, microRNAs have the potential to become novel modulators of

circadian rhythms and might be able to positively influence cardiovascular physiology (Figure

1.6).

1.4 Circadian Disruption at the Root of Disease

Circadian disruption is associated with a multitude of prevalent diseases

Seasonal changes in daylight exposure is a mechanism of circadian rhythmicity yet can

also be a mechanism of disruption. One such study of found that humans differentially adjust to

daylight savings time based on chronotype with an overall inability to sufficiently adjust from spring

to fall (172). Chronobiologists propose that our modern-day society removes the natural seasonal

nature of humans (human behavior tends to be seasonal, like reproduction, mortality, and suicides

(173, 174)), resulting in disruption of and dissociation from circadian rhythmicity (172).

Furthermore, studies of our modern-day light exposure found delays in circadian rhythmicity, with

21 a significant circadian disruption following even just one weekend of late-night light exposure, a

common trend for those balancing work or school in the typical work-week (175). Differences in light exposure based on latitude location revealed that individuals furthest away from the equator tend fit into the late-chronotype category (176). Because intense light (>10,000 LUX) is most effect at entraining the circadian system (10), those further away from the equator may have dampening of circadian amplitude compared to those who receive strong light exposure closer to the equator and most likely a more robust circadian amplitude (176). Additionally, dampening of the circadian oscillator is part of aging and associated with multiple aging-related diseases (177).

Another common mechanism of circadian disruption is shift-work or social-jetlag. In the present day, approximately 22% of the population is engaged in some form of shift-work, which includes any variation of work outside the typical work day, such as night-shift, multiple shifts, and frequent shift in times of work (178, 179). Those who participate in shift-work for long periods of time are heavily associated with overall poor health outcomes and are at a significantly higher risk for many diseases, not only the aforementioned cardiovascular disease and metabolic syndromes, but including ulcers, cancer, and obstructive sleep apnea (16, 33-37, 106, 178-183).

The significant association between shift-workers and disease emphasizes the importance in understanding circadian biology to hopefully identify targets for prevention or treatment of those with disrupted circadian rhythm.

In particular, the onset of MI has a distinct circadian pattern and recent findings suggest light exposure, the foundation of circadian rhythms, as a potential strategy for treating myocardial ischemia and preventing MI (6, 23-27). Furthermore, epidemiological studies revealed that MIs occur more frequently in the early morning hours than later in the day, are more prevalent within the first three hours after waking up in the morning than any other time of the day, and infarctions occurring in the morning are more severe than those that do present in the evening (24, 25, 184,

185). Additional epidemiological studies revealed that there is a similar increase in MIs across all

50 states in the USA during the darker winter months (184, 185). Experimental studies found

22 circadian oscillations of neutrophil recruitment to heart, which determines infarct size, healing,

and cardiac function after MI. Neutrophil recruitment to the heart occurs more preferentially during the active phase and heart attacks during this time have significantly higher cardiac neutrophil infiltration. MI during this time (active phase) has higher neutrophilic inflammatory response and worse cardiac repair. By limiting the neutrophil recruitment during the heart in the active phase, there was a reduction in infarct size and increase in improvement of cardiac function (186).

However, the mechanism underlying the circadian component of MIs and the role of circadian biology in cardioprotection is only recently under intense investigation.

ICU settings and anesthetics have severe effects on the circadian system

Unfortunately, patients in the ICU are subjected to severe circadian disruption, which we

are well aware promotes disease. The ICU closely mimics a full 24 hours of light exposure and circadian disruption in the ICU is known to be detrimental to human health (187-191). ICU patients are noted to have significant mitochondrial and endothelial dysfunction accompanied by defects in nitric oxide synthesis and pyruvate dehydrogenase activity, functions regulated by circadian proteins (5, 74, 192). Furthermore, recent epidemiological studies revealed that ICU patients who received cardiac surgery in the morning have significantly worse outcomes than those who receive surgery in the afternoon as measured by troponin levels and long-term adverse cardiac events (193). Many investigators have begun manipulating light exposure in mouse models to determine downstream effects of circadian disruption. The most commonly used lighting in the vivarium setting used to be cool white fluorescent lights, but many institutions have begun switching to light-emitting diodes (LED) for their numerous economic advantages. However, LED lighting has high blue light emissions and its effects on laboratory animal’s circadian system and physiology needs to be considered. Recently, one group found significant positive effects on circadian-regulated neuroendocrine, metabolic, and physiologic outcomes in rats exposed to daylight LED (194), indicating a need to consider housing conditions in animal studies. In fact, housing mice in constant light is sufficient to dampen and delay PER2 oscillation in the SCN,

23 eventually losing complete rhythmicity (195). These studies emphasize the need to consider light exposure and the circadian system for not only experimental animal studies but also for patients in the ICU and people in the general population, considering circadian biology and circadian disruption is tightly associated with physiological function and dysfunction.

Lastly, another mechanism of circadian disruption is through general anesthesia.

Anesthesia-mediated circadian disruption is a recognized phenomenon, however its role has not been fully elucidated yet (4, 5). Interestingly, several medications used in a daily clinical setting can disrupt the well-coordinated expression of the circadian rhythm network. As such, it has been found that anesthetics disrupt the expression of the circadian proteins including PER2, which could lead to sleep disturbances following general anesthesia (196, 197). Studies have shown that propofol anesthesia in humans affects the circadian period, leading to increased resting activities during the day (196). Animal studies found ketamine to reduce the amplitude of circadian rhythm gene expression (198). Other studies using sevoflurane, a volatile anesthetic similar to isoflurane, found suppression of PER2 in the brain of rodents (199). However, research on the influence of anesthetics on circadian rhythm protein or PER2 expression in specifically the heart is in its infancy. Understanding anesthesia mediated disruption in the brain may provide insights into circadian disruption in peripheral tissues. GABA (ᵞ-aminobutyric acid) is the principal neurotransmitter in the SCN and studies identifying GABA to be responsible for phase shifting and synchronizing cells of the SCN indicate that GABA is responsible for synchronizing the clock in the whole animal and peripheral tissues (200). GABA is primarily known as an inhibitory neurotransmitter in the central nervous system, however it is diurnal in nature. GABA is an inhibitory neurotransmitter during the night, which decreases the firing frequency, but GABA is an excitatory neurotransmitter during the day, which increases the firing frequency. This dual role of

GABA happens through GABAA receptors and due to the daily oscillations of intracellular chloride concentrations through chloride-permeable ion channels opening and closing (201, 202). This indicates that the inhibitory or excitatory nature of GABA is regulated by the time of day and

24 perhaps boosts neural rhythm amplitude to drive SCN outputs to the peripheral organs (201)

(202). That being said, anesthetic GABA-mediated disruption may have detrimental effects on peripheral organs like the heart.

1.5 Circadian Considerations in Therapeutic Strategies

Maintaining a robust circadian amplitude

A form of circadian disruption is dampening of circadian amplitude, which is observed as part of the aging process (177). Therefore, researchers believe that a robust circadian timekeeping system is important for human health and well-being (4, 5, 183, 203-205). Indeed, approaches to increase entrainment and thereby the robustness of the clock timing process have been found to be beneficial in certain disease states. One study found that restricted feeding increased the circadian amplitude and could prevent mice from becoming obese when exposed to a high fat diet (206). Another study found using short term caloric restricted feeding in mice to

be cardioprotective. In addition, cardiac microRNA profiling of short term caloric restricted feeding

was associated with the circadian clock (207). Similarly, an independent study tested whether activity alters or could rescue a disrupted circadian system. Here, they examined the effects of wheel access on vasointestinal polypeptide (VIP)-deficient mice, a model that exhibits circadian deficits (208). Indeed, voluntary scheduled exercise increased the amplitude of PER2 expression and rescued the disrupted circadian system in VIP deficient mice. Some studies on blue enriched light exposure on rats found similar effects on circadian rhythmicity and amplitude. In addition, marked positive effect on the circadian regulation of neuroendocrine, metabolic, and physiologic parameters associated with the promotion of animal health and wellbeing was observed (194).

Similarly, one study demonstrated that day/night rhythms play a critical role in compensatory remodeling of cardiovascular tissue, and disruption of day/night rhythms exacerbated disease pathophysiology (209). Here, they used a murine model of pressure overload cardiac hypertrophy in a rhythm-disruptive 20-hour versus 24-hour environment. Echocardiographic studies revealed

increased left ventricular end-systolic and -diastolic dimensions and reduced contractility in

25 rhythm-disturbed animals exposed to transverse aortic constriction. Considering these findings, it

is compelling that light exposure, restricted feeding, or scheduled exercise could also be

cardioprotective through strengthening the circadian clock. Since microRNAs can be regulated

by external cues such as light, without involvement of the circadian rhythm proteins, it is probable

that light exposure also improves a disrupted circadian system through enhancing the amplitude

of microRNAs and optimizing their function.

Diurnal versus nocturnal considerations

Entrainment strategies are essentially identical between diurnal and nocturnal animals, like humans and mice. It’s a common misperception that a diurnal animal would have circadian clockwork opposite to that of a nocturnal animal. Nocturnal and diurnal creatures maintain nearly identical oscillations of the fundamental circadian biology machinery through the same mechanisms to achieve light-elicited entrainment. The difference between these two types of

mammals resides within how the SCN is coupled to its various outputs (Figure 1.7). The

behavioral differences that we use to identify whether an animal is diurnal or nocturnal is not

representative of their circadian phase and therefore does not imply a 180-degree change in

phase (210-213). Metabolic rhythms in the SCN assessed using 14C-labeled 2-deoxyglucose in a variety of animals with either nocturnal or diurnal nature revealed consistent peaks during the day even though the animals have a varying time-of-day activity level (213). Moreover, studies

investigating the phase response curve in response to light in both diurnal and nocturnal animals

found essentially the same phase relative to the daylight cycle (213, 214).

Diurnal and nocturnal animals also experience similar melatonin cycles: melatonin peaks at 5 AM in mice and 3 AM in humans and is suppressed by light exposure in both species. Thus, melatonin is synthesized and secreted in the dark phase independent of nocturnal or diurnal

(activity) nature, indicating that the circadian rhythmicity we see in mice is indeed applicable to human understanding (215). The rationale underlying this claim is due to multiple parallel outputs from the SCN (Figure 1.7). In fact, the role of melatonin in circadian clockwork is thought to be

26 relatively limited since removal of the pineal gland does not significantly affect circadian biology

(216). Even though melatonin signaling is coupled to the SCN through melatonin receptors, subtle changes have been noted within the SCN after pineal gland removal, like changes in neuron firing rate in the SCN itself (217). However, this does not necessarily affect overall circadian biology in the periphery for two reasons: bright light exposure can override melatonin because it is a primary entrainment strategy and therefore a stronger regulator of circadian rhythms (218) and secondly, melatonin secretion from the pineal gland is one of many parallel circadian outputs from the SCN that synchronize circadian biology in peripheral tissues (219, 220). Lastly, melatonin secretion from the pineal gland is made through N-acetylation of serotonin by N-acetyltransferase (NAT) and O-methylation by hydroxyinodole-O-methyltransferase. Importantly, many mice used in the laboratory have a C57BL/6 background, which do not have activity of this pineal gland enzyme due to a point mutation in the NAT gene, leading to altered RNA splicing, inclusion of a 102 base pair pseudo-exon in the mature mRNA, and a premature stop codon with a subsequent truncated protein lacking enzyme activity (221-223). Therefore, studies using mice with a C57BL/6 background remove melatonin from the equation and allow for more direct studies of light-elicited

entrainment and SCN mediated synchronization to peripheral organs.

Pharmacological timing and targeting of the circadian system

The notion that disrupted circadian rhythm can affect all aspects of physiologic

homeostasis is supported by researchers who identified a significant portion of the genome – and

therefore a variety of cellular processes – to be under circadian control. Recent investigations into the 24-hour transcriptome suggest that much of the mammalian genome is under circadian control. RNA sequencing of major tissues and brain regions of baboons, a primate closely related to humans, revealed 81.7% of the protein-coding genome is under circadian control, as indicated by rhythmic gene expression. These gene products are in fact part of a diverse range of biochemical and cellular functions (224). Furthermore, the transcriptome of 64 tissues with sampling every 2 hours over a 24-hour period, tissue-specific rhythmic expression was noted and

27 rhythmic genes tended to peak around early morning (dawn) and sunset (dusk). There was also an inactive or “quiescent period” in the beginning hours of the night (224). In parallel, another research group found by RNA-seq and DNA arrays of 12 mouse organs over 24 hours that 43% of protein coding genes are expressed in a circadian pattern. In general, oscillating genes tended to peak just preceding dawn and dusk, supporting the previously described results (224, 225).

Further investigations found that more than 1,000 noncoding RNAs oscillate in a circadian day, as well, suggesting more than just messenger RNA has a circadian pattern (225). Taken together, while there are many ways to disrupt circadian biology, these mechanisms may prove to be targets for preventative or therapeutic strategies in diseases arising from desynchrony that ultimately target the circadian clockwork.

With a significant portion of the genome evidently under control of the circadian machinery, there have been many investigations into the human chronobiome to potentially develop time-of- day targeted therapies. A pilot study gathered data on diurnal variations of cardiovascular and behavioral phenotypes from healthy human volunteers and measured “omics” outputs. This study revealed a time-of-day dependent variation in blood pressure, activity level, light exposure, and food consumption associated with time-of-day dependent variation in the metabolome, proteome, and transcriptome (225, 226). Recent studies using RNA-seq from 632 human samples including

13 different tissues revealed several common pharmacological treatments, like antihypertensive drugs (calcium channel and β-blockers) target gene products that are rhythmically expressed in the cardiovascular system (227). Together, these studies strongly suggest the consideration of circadian biology in clinical treatment strategies.

In a small molecule screen for circadian clock modifiers, many potential pharmacological molecules were identified with targets anywhere in the pathway, whether intracellularly, within the feedback loop, or in metabolic processes. These compounds were found to alter the period, phase, or amplitude of the circadian clock (228). There are many factors that feed into the molecular circadian feedback loop that are targets for regulating downstream effects. The

28 polymethoxylated flavone nobiletin, originally identified as the naturally occurring compound in

citrus peels, is a clock amplitude-enhancing small molecule. It was identified as a small molecular

target of circadian amplitude enhancement and PER2 target through this high-throughput chemical screen (228, 229). After its identification, nobiletin was found to be protective against metabolic disease by acting as an agonist for the ROR nuclear receptor thereby reprogramming circadian mediated metabolic gene expression (228-230). In fact, ROR nuclear receptors are key

for stabilizing the molecular oscillating loop (Figure 1.1) (229) and nobiletin enhances the

amplitude of circadian rhythm and regulates energy homeostasis in terms of body weight, food

intake, body composition, and VO2 max. Nobiletin also regulates sleep behavior in a clock

dependent manner and with respect to metabolism, nobiletin improves glucose and lipid homeostasis in mice, which is under circadian clock control (229). Nobiletin increases ROR⍺/ᵞ transcriptional activity by binding directly to this nuclear receptor by competitive inhibition and

NOB dose dependently increases BMAL1 promoter driven luciferase reporter activity and dependent upon RORE elements (DNA binding site) (229). Nobiletin, as a target of nuclear receptors, is a potential mechanism to manipulate circadian clock function.

29

Figure 1.1 Circadian Feedback Loop All cells have an approximate 24-hour negative feedback loop of core circadian proteins, PER1/2, CRY1/2, BMAL1, and CLOCK. PERs and CRYs, expressed upon BMAL1 and CLOCK binding to the E-box in their promoter regions, accumulate in the cytoplasm during the day. PERs and CRYs form a heterodimer and translocate back to the nucleus where they inhibit their transcriptional activators, BMAL1 and CLOCK, thereby inhibiting their own expression. Nuclear receptors like RORs and REV-ERBs regulate BMAL1 expression by either acting as activators or repressors, respectively.

30

Figure 1.2 Hypoxia Signaling In normoxia, HIF1⍺ is targeted for proteasomal ubiquitination by PHDs and VHLs. In hypoxic conditions, PHDs are inhibited and therefore HIF1⍺ stabilization is achieved. HIF1⍺ forms a dimer with constitutively active HIF1β and translocates into the nucleus where it binds to HREs in the promoter region of target genes (adapted from (231) (232)).

31

Figure 1.3 Adenosine Signaling in Cardioprotection The ecto-5’-nucleotidases CD39 and CD73 and the adenosine A2B receptor are essential for cardioprotection during IPC. Downstream signaling via the A2B receptor revealed an upregulation of PER2 and a PER2-dependent stabilization of HIF1⍺ (adapted from (69) (70) (71)).

32

Figure 1.4 Energy Efficient Metabolism Glycolysis produces pyruvate, which under anaerobic conditions will be converted to lactate, but under aerobic conditions will be oxidized in the mitochondria. Fatty acids enter the mitochondria through β-oxidation. Regardless of substrate, metabolites go through the TCA cycle and oxidative phosphorylation, where oxygen is consumed, to produce ATP. The glycolytic intermediate glucose-6-phosphate is also used in the PPP for NADPH and nucleic acid synthesis. CS: citrate synthase; ACO: aconitase; IDH: isocitrate dehydrogenase; ⍺-KGDH: ⍺-ketaglutarate dehydrogenase; SUCLG1: succinate Co-A ligase; SDH: succinate dehydrogenase; FH: fumarase hydratase; MDH: malate dehydrogenase.

33

Figure 1.5 Possible microRNA pathways for regulating the circadian clock (A) External cues control the circadian clock work and thereby regulate clock dependent microRNAs and their target genes. (B) External cues such as sunlight directly control microRNAs and thereby regulate microRNA regulated clock genes and their dependent target genes. (C) External cues regulate microRNAs and their target genes in a circadian manner based on the rhythm of the external cue (142).

34

Figure 1.6 Circadian cardiac entrainment as cardioprotective mechanism Apart from light, exercise or restricted feeding have been shown to entrain the circadian system. Increasing the amplitude and the robustness of PER2 or miR-21 expression could represent a possible cardioprotective mechanism (142).

35

Figure 1.7 SCN Outputs Light stimuli received by melanopsin receptors in the retinal ganglion cells is transmitted through the retinohypothalamic (RHT) tract to the SCN in the hypothalamus. The SCN has multiple mechanistic outputs to regulate melatonin secretion from the pineal gland (through the superior cervical ganglion, SCG), the sleep/wake cycle, hormone secretion form the pituitary gland, and peripheral clocks through the autonomic nervous system (adapted from (141, 233, 234)).

36 Table 1.1 Clock Mutant Mouse Models with Metabolic and Circadian Phenotypes Mouse Model Cardiovascular and/or Circadian Phenotype BMAL1-/- Arrhythmic (235) Disrupted circadian variation in blood pressure, heart rate (236, 237) Increased susceptibility to vascular injury (236, 237) Hyperlipidemia (238) CLOCKΔ19 Hyperlipidemia (239) Hyperglycaemia (240) PER2-/- Arrhythmic (241) Shorter circadian period (241) CRY1-/-; CRY2-/- Arrhythmic (242, 243) Hypertension (244) Hyperglycaemia (245, 246) REV-ERBA-/- Reduction in mitochondrial content and oxidative function (247) Shorter circadian period (248) Dyslipidemia (132, 249) REV-ERBA-/-; REV-ERBB -/- Shorter circadian period (248) Increase in circulating glucose and triglyceride levels (248) Reduced free fatty acids (248)

37 Table 1.2 MicroRNAs in the Heart MicroRNA Cardiac Role Ref. miR-133 Overexpression prevents hypertrophic cardiomyopathy via β1AR 135 transduction cascade miR-208a Inhibition beneficial in heart failure by reversing myosin switch and 136 improving cardiac function and remodeling during heart disease progression miR-15 Regulator of cardiac fibrosis and hypertrophy by inhibition of TGFβ- 134, pathway 137 miR-30 Negative regulator of cardiac fibrosis 134 miR-15 Increase cell death after myocardial IR-injury 142 miR-34 Increase cell death after myocardial IR-injury 143 miR-320 Increase cell death after myocardial IR-injury 144 miR-140 Increase cell death after myocardial IR-injury 145 miR-1/miR- Increase cell death after myocardial IR-injury 146 206 miR-92a Increase cell death after myocardial IR-injury 147 miR-122 Increase cell death after myocardial IR-injury 148 miR-150 Increase cell death after myocardial IR-injury 149 miR-181a Increase cell death after myocardial IR-injury 150 miR-376b-5p Increase cell death after myocardial IR-injury 151 miR-24 Reduce cardiomyocyte cell death or enhance cardiac regeneration 152 miR-29 Reduce cardiomyocyte cell death or enhance cardiac regeneration 153 miR-30 Reduce cardiomyocyte cell death or enhance cardiac regeneration 154 miR-214 Reduce cardiomyocyte cell death or enhance cardiac regeneration 155 miR-7a/b Reduce cardiomyocyte cell death or enhance cardiac regeneration 156 miR-20a Reduce cardiomyocyte cell death or enhance cardiac regeneration 157 miR-132 Reduce cardiomyocyte cell death or enhance cardiac regeneration 158 miR-138 Reduce cardiomyocyte cell death or enhance cardiac regeneration 159 miR-144/451 Reduce cardiomyocyte cell death or enhance cardiac regeneration 160 miR-155 Reduce cardiomyocyte cell death or enhance cardiac regeneration 161 miR-210 Reduce cardiomyocyte cell death or enhance cardiac regeneration 162 miR-499 Reduce cardiomyocyte cell death or enhance cardiac regeneration 163 miR-874 Reduce cardiomyocyte cell death or enhance cardiac regeneration 164

38 CHAPTER 2 CHAPTER II PER2 MEDIATED ENERGY METABOLISM IN ADAPTATION TO HYPOXIA2

2.1 Rationale

During myocardial ischemia, a prevailing hypoxic environment demands the heart adapt to low oxygen availability by balancing metabolic pathways. Previous studies revealed that PER2 is a critical factor in cardioprotection from myocardial ischemia and is associated with HIF1⍺, a

key factor in adaptation to low oxygen availability (71). However, the mechanism of PER2 dependent adaptation to hypoxia remains elusive and understanding how PER2 functions under low oxygen conditions may reveal novel targets for adaptability to conditions where oxygen availability is limited. Numerous studies point toward a bi-directional link between circadian biology and hypoxia-sensing pathways, suggesting its evolutionarily conserved relationship may be targetable for metabolic adaptation. Oxygen is a critical regulator of the molecular circadian clock through hypoxia regulated HIF1⍺ (83) that promotes PER2 oscillation by binding to an HRE in the PER2 promoter (79). Our previous studies suggested cardiac HIF1⍺ oscillation is

dependent on PER2 (71) and brain and kidney HIF1⍺ nuclear localization is regulated by the

circadian clock (64). Based on our previously published data (71), we hypothesized that PER2 is

adaptive to low oxygen environments by targeting the balance between metabolic pathways in

hypoxia that include shifts between glycolysis and mitochondrial respiration. In this chapter, we present our mechanistic studies using human microvascular endothelial cells (HMEC-1), which make up a substantial portion of the myocardium (89), are able to sense and adapt to changes in

oxygen (95-98), and are critical components of myocardial IR-injury (91-94, 250, 251).

2 Portions of this chapter are from our manuscript currently in review. The stable lentiviral mediated PER1 KD, PER2 KD, and Scr control HMEC-1 lines were generated and validated by Stephanie Bonney. The myocardial ischemia model and Complex IV activity assay was done by Yoshimasa Oyama. Colleen Bartman completed transcript analyses, enzyme activity assays, co-immunoprecipitations, immunoblots, metabolic stress tests on the Seahorse Bioanalyzer, MitoTracker staining, and in vitro stable isotope tracer experiments. Mass spectrometry analysis was done by the Metabolomics Core at the University of Colorado Anschutz Medical Campus.

39 2.2 Results

PER2 transcriptionally controls HIF1⍺ dependent glycolysis during hypoxia

Our previous studies suggested cardiac HIF1⍺ oscillation is dependent on PER2 (71). To investigate the role of PER2 mediated adaptation to hypoxia, we evaluated PER2 signaling targets during hypoxia. For this purpose we generated a novel lentiviral mediated PER2 knockdown (KD) stable cell line in HMEC-1, as PER2 was found to be crucial for endothelial function in low oxygen conditions (252). Confirming oxygen as a regulator of the molecular

circadian clock, hypoxia exposure increased PER2 transcript and protein levels which was

abolished in PER2 KD cells (Figure 2.1).

Switching from energy-efficient lipid to oxygen-efficient and HIF1⍺ dependent glucose metabolism is pivotal to allow the myocardium to function under low oxygen conditions (71, 253,

254). Therefore, we first evaluated the transcriptional regulation of glycolysis during hypoxia.

Hypoxic PER2 KD HMEC-1 displayed abolished transcriptional induction of HIF1⍺ dependent

glycolytic enzymes, one of which was pyruvate kinase (Figure 2.2) (255). Lactate dehydrogenase

is another key anaerobic glycolysis enzyme that is under the control of HIF1⍺ (49) and its increase

in expression in hypoxia was abolished in PER KD HMEC-1, with a similar kinetic in supernatant

lactate levels (Figure 2.3). Importantly, expression of these HIF1⍺ dependent genes was PER1

independent and PER2 dependent (Figure 2.4). Functional studies revealed increased

cytotoxicity (Figure 2.5) and reduced glycolytic capacity in hypoxic PER2 KD HMEC-1 (Figure

2.6). Based on previous observations in HMEC-1 that PER2 is significant proportion at zeitgeber

time (ZT) 12 vs ZT0, (71) we determined whether there were diurnal metabolic variations and

found a significant increase of glycolytic capacity at ZT12 under baseline (normoxia) conditions

that was PER2 dependent (Figure 2.7). Mechanistic studies, using a chromatin immunoprecipitation assay (ChIP) uncovered that hypoxia induced HIF1⍺ binding to the lactate

dehydrogenase promotor region was abolished in PER2 KD cells (Figure 2.8). Together with

40 previous studies in PER2 gene-targeted mice (71), these findings highlight a critical role of PER2 in cellular adaption to hypoxia and reveal PER2 as an essential co-factor of HIF1⍺ mediated transcription of glycolytic genes and as a key regulator of glycolytic metabolism.

PER2 is a post-translational regulator of TCA cycle activity during hypoxia

To gain a deeper mechanistic perspective of PER2 dependent metabolic adaptation to

hypoxia, we next pursued a comprehensive systems biology approach (256) using a wide and

unbiased affinity purification-mass spectrometry-based proteomics screen for PER2 protein

interactions under hypoxic conditions (Figure 2.9). Serendipitously, a high percentage of PER2-

protein interactions hinted towards an important role for PER2 in controlling TCA cycle function

(Figure 2.10). Subsequent Co-IP pull-downs on TCA cycle enzymes or functional assays on TCA

cycle enzyme activity confirmed binding to PER2 and regulation of TCA cycle function during

hypoxia in a PER2 dependent manner, respectively (Figure 2.11, 2.12). Similarly, PER2 KD cells

under hypoxia showed significantly less CO2 production, a surrogate endpoint of TCA cycle

function (Figure 2.13). Considering TCA cycle enzyme activity is known to be regulated by SIRT3

mediated de-acetylation (121), which is under circadian control (113), we next performed Co-IP

pull-downs for SIRT3. Indeed, SIRT3 was bound to PER2 and isocitrate dehydrogenase (IDH2,

Figure 2.14), one of the key control points of the TCA cycle, suggesting a crucial complex involving PER2, SIRT3, and IDH2 to regulate the TCA cycle. Next, we investigated whether hypoxia and PER2-HIF1⍺ dependent pathways would also regulate SIRT3 expression. HMEC-1 transcriptional and translational analysis with a PER2 or HIF1⍺ KD revealed a PER2-HIF1⍺ dependent regulation of SIRT3 under hypoxic conditions (Figure 2.15, 2.17, 2.18) and further analysis confirmed an HRE in the promoter region of SIRT3 (Figure 2.16). Our systems biology approach together with previous studies showing that the circadian clock drives oxidative metabolism in mice (113) uncover a critical role for PER2 in controlling oxidative TCA cycle metabolism during hypoxia by regulating transcriptional HIF1⍺-dependent pathways or post- translational SIRT3-dependent pathways.

41 PER2 regulates mitochondrial function via HIF1⍺-COX4.2 in hypoxia

Additional analysis of our proteomics screen indicated binding of PER2 to mitochondrial

complexes (Appendix A) suggesting yet another role for PER2 in controlling mitochondrial function under hypoxia, this time by targeting oxidative phosphorylation. Indeed, oxygen consumption rates (OCR; a measure of mitochondrial functionality), basal respiration, maximal respiration, ATP production, spare capacity, or maximal respiration due to endogenous fatty acids were significantly reduced in PER2 KD cells during a mitochondrial stress test (Figure 2.19 –

2.21). Moreover, OCR levels were significantly increased in cells with higher PER2 levels at ZT12

when compared to ZT0, confirming observations of a diurnal variation of mitochondrial function

(257) (Figure 2.22, 2.23). Considering HIF1⍺ mediates a switch of COX4 subunits (COX4.1 to

COX4.2) in hypoxia to enhance oxygen efficiency and provide adaptation to hypoxia (105), we

next investigated the transcriptional regulation of COX4.2 in PER2 KD cells under hypoxia. Like

other HIF1⍺ dependent transcriptional regulations in PER2 deficient cells, we observed abolished

increases of COX4.2 mRNA or COX4 activity in hypoxic PER2 KD cells (Figure 2.24) or ischemic

hearts from PER2-/- mice (Figure 2.25). To understand if a compromised oxidative

phosphorylation in PER2 deficiency would be associated with reduced mitochondrial membrane

potential, which is associated with compromised mitochondrial function (103), we next used

MitoTracker deep red staining (258). Studies in PER2 KD HMEC-1 indicated already reduced

mitochondrial potential under normoxia (Figure 2.26). Indeed, analysis of a cell energy phenotype

assay revealed significantly less aerobic metabolism in PER2 KD cells at baseline (Figure 2.27).

Confirming these results, JC-1 assays showed a quantified significant reduction in membrane

potential in PER2 KD cells at both baseline (normoxia) and under hypoxia (Figure 2.29, 2.30). To

further explore PER2 dependent metabolism, we next used liquid chromatography-tandem mass

spectrometry studies following the exposure of labeled glucose (13C-glucose) or palmitic acid (13C-

palmitic acid) to assess metabolic flux rates in PER2 KD cells. Here we confirmed that PER2 is

essential for glycolysis or oxidative metabolism under hypoxia (Figure 2.31). Moreover, we also

42 found PER2 to be critical for the PPP, indicating that PER2 KD cells lose their ability to generate

the redox cofactor NADPH, which has a pivotal role for circadian timekeeping (259) (Figure 2.32).

As PER2 has been shown to inhibit lipid metabolism via PPARg (129), we also found altered fatty acid metabolism in PER2 KD cells under hypoxia (Figure 2.33). Taken together, these data

identify PER2 as a master regulator in all kingdoms of metabolism (Figure 2.34), both

transcriptionally and post-translationally, and incorporate the control of energy metabolism

through HIF1⍺ or PPARg dependent pathways.

2.3 Discussion

Our results reveal novel hypoxia driven pathways of PER2 regulation and pioneer a

mechanistic intersection between circadian and hypoxia-sensing pathways. We uncovered that

PER2 functions both in the nucleus where it is essential for transcriptional regulation of HIF1⍺

glycolytic targets and in the mitochondria where it post-translationally complexes with SIRT3 and

TCA cycle enzymes regulating their activities. In depth metabolic studies revealed a so far

unknown role for PER2 in regulating oxidative phosphorylation by transcriptionally and possibly

post-translationally regulating COX4.2 expression and activity, a critical HIF1⍺ dependent

component of adaptation to low oxygen conditions (105). Moreover, we describe for the first-time

a critical role of PER2 for all kingdoms of metabolism such as glycolysis, TCA-cycle function,

PPP, or lipid metabolism.

Metabolic adaptation to low oxygen conditions is advantageous for cellular protection

during periods of low oxygen availability. Other research groups found natural rhythmic oxygen

levels within a physiologic range in mouse peripheral tissues and using rhythmic oxygen exposure

is sufficient to reset the circadian clock in a HIF1⍺ dependent manner (79, 83). To support the

existence of a bi-directional feedback loop or circadian-hypoxia link is that PER2 is also a critical

key regulatory factor of HIF1⍺ metabolic adaptation to hypoxia. As a multi-functional protein,

PER2 has unique functions in both the nucleus and mitochondria. However, the mechanism of

43 how PER2 is shuttled to either location is only speculative. It has been previously demonstrated that PER2 has the ability to shuttle between the cytoplasm and nucleus via cytoplasmic localization domains and nuclear export sequences (260). To the best of our knowledge, how

PER2 targets the mitochondria has not yet been identified. However, multiple studies support

PER2 regulated mitochondrial metabolic enzyme functions (257, 261). As indicated by our affinity- purification mass spectrometry proteomics screen, there are other pathways in which PER2 is involved in a hypoxic-specific manner. These include cholesterol biosynthesis, RNA signaling, tRNA charging, and phagosome maturation. In fact, clock gene expression in human subcutaneous adipose tissue has been associated with serum cholesterol levels and subsequent metabolic and inflammatory gene expression (262) and PER2 gene expression was found to be positively correlated with LDL cholesterol (263), further supporting the role of circadian biology in metabolic syndromes. Taken together, we uncovered an innovative role for circadian PER2- targeted HIF1⍺ pathways in metabolic adaptation to hypoxia, which further supports the role of circadian biology in the development of metabolic syndromes and reveals the potential of targeting PER2 for preventative or therapeutic benefit.

44 A

B

Figure 2.1 PER2 KD HMEC-1 Validation HMEC-1 controls (Scr; treated with lentiviral scrambled siRNA) or HMEC-1 PER2 knockdown (KD treated with lentiviral PER2 siRNA) were synchronized by serum starvation and exposed to 24 h of normoxia (Nx; 21% O2) or hypoxia (Hx; 1% O2). PER2 transcript (A) or PER2 protein (B) levels from stable lentiviral mediated PER2KD or Scr control HMEC-1were determined by quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR) or immunoblot analysis relative to the housekeeping gene β-ACTIN (A) or protein (B) β-ACTIN (mean ± standard deviation, SD, n = 3 per condition per group).

45

Figure 2.2 Pyruvate Kinase Gene Expression Dependent on PER2 Transcript expression of pyruvate kinase (PKM) was analyzed using SYBR Green and qRT-PCR from stable lentiviral mediated PER2KD HMEC-1 or controls (Scr) after 24 h Hx or Nx (mean ± standard deviation, SD, n = 3).

46 A

B

Figure 2.3 PER2 Dependent Lactate Dehydrogenase Expression and Production Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were synchronized prior to Hx or Nx incubation. (A) Transcript expression of lactate dehydrogenase isoform A (LDHA) PER2 KD or Scr control HMEC-1 (mean ± standard deviation, SD, n = 3). (B) Lactate levels in supernatants obtained from PER2 KD or Scr control HMEC-1 incubated in Hx or Nx (mean ± standard deviation, SD, n = 3).

47 A

B

Figure 2.4 HIF1⍺ Glycolytic Target Genes are PER2 – But Not PER1 – Dependent Synchronized HMEC-1 were exposed to Nx or Hx after siRNA knockdown of PER2 (siPER2) or PER1 and PER2 (siPER1/2) or treatment with nonspecific control siRNA (siScr). Transcript expression of (A) PKM or (B) LDHA in HMEC-1 after siRNA knockdown and incubation in Hx or Nx was assessed by qRT-PCR (mean ± standard deviation, SD, n = 3).

48

Figure 2.5 Cytotoxicity of PER2 KD HMEC-1 Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were incubated in Hx prior to cytotoxicity analysis using an LDH-Cytotoxicity assay (mean ± standard deviation, SD, n = 10).

49 A

B

Figure 2.6 PER2 is Important for Glycolytic Reliance in Hypoxia Glycolytic function was assessed using a glycolytic stress test and Seahorse Biosciences XF24 Analyzer. Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were incubated in Hx and serum starved prior to measuring extracellular acidification rate (ECAR) in response to injected compounds (glucose: stimulates glycolysis, oligomycin: ATP synthase inhibitor, 2-DG (2-deoxy- glucose): inhibits glycolysis). (A) Glycolytic stress test profile of glycolytic function in hypoxia comparing PER2 KD (red line) or Scr control (black line) HMEC-1. (B) Quantification of glycolysis in hypoxia comparing PER2 KD or Scr HMEC-1 (mean ± standard deviation, SD, n = 10).

50 A

B

Figure 2.7 Time-of-Day and PER2-Dependent Glycolytic Reliance Glycolytic function was assessed using a glycolytic stress test and Seahorse Biosciences XF24 Analyzer at two time points in the 24-hour day. Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were serum starved prior to measuring extracellular acidification rate (ECAR) in response to injected compounds (glucose: stimulates glycolysis, oligomycin: ATP synthase inhibitor, 2-DG (2-deoxy-glucose): inhibits glycolysis). Shown are glycolytic stress test profiles of glycolytic function in normoxia comparing PER2 KD (red line) or Scr control (black line) HMEC-1 at (A) zeitgeber (ZT) 0 or (B) twelve hours later at ZT12 (mean ± standard deviation, SD, n = 10).

51

Figure 2.8 PER2 Dependent HIF1⍺ Binding at the LDHA Promoter Chromatin immunoprecipitation (ChIP) was performed using a HIF1⍺ antibody in stable lentiviral mediated PER2 KD or Scr control HMEC-1 incubated in Hx or Nx. After ChIP, HIF1⍺ binding to the hypoxia response element (HRE) in the LDHA promoter was determined by amplifying the LDHA promoter using qRT-PCR to obtain CT values and visualized by 2% agarose gel of qRT- PCR products (n = 3).

52

Figure 2.9 PER2-Normoxia-Hypoxia-Pathways Wildtype HMEC-1 were synchronized and incubated in Hx or Nx prior to co-immunoprecipitation using a PER2 antibody and affinity purification-mass spectrometry-based proteomics to determine PER2 interactions. Data were analyzed using the Reactome Pathway Database, indicating a strong involvement of PER2 in metabolic pathways under hypoxia. Yellow depicts PER2 pathways in comparison to all available Reactome pathways (grey).

53 A

B

Figure 2.10 Characterization of PER2-Normoxia-Hypoxia Pathways Characterization of PER2-normoxia-hypoxia pathways as identified by the Reactome Pathways Database (Figure 2.9).

54 A

B

Figure 2.11 Western Blot Validation of Mass Spectrometry Analysis (A) Synchronized HMEC-1 were exposed to hypoxia or normoxia and protein lysates were isolated for native protein complexes using a PER2 antibody covalently coupled (immobilized) onto an amine-reactive resin. Immunoprecipitated protein was analyzed using immunoblots against isocitrate dehydrogenase 2 (IDH2), succinyl Co-A ligase (SUCLG1) and aconitase (ACO2). One representative blot of three is displayed. (B) Reciprocal co-immunoprecipitations where performed from synchronized HMEC-1 exposed to hypoxia or normoxia and protein lysates were isolated using either an IDH2, SUCLG1, or ACO2 antibody covalently coupled to an amine- reactive resin. Immunoprecipitated protein was analyzed on immunoblots against PER2. One representative blot of three is displayed.

55

A B C

Figure 2.12 PER2 Dependent TCA Cycle Enzyme Activities Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were incubated in Hx prior to analysis of key TCA cycle enzyme activities, (A) isocitrate dehydrogenase (IDH), (B) succinyl Co-A ligase (SUCLG), and (C) aconitase (ACO), using a microplate reader (mean ± SD, n = 6).

56

Figure 2.13 PER2 Dependent TCA Cycle Functioning Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were synchronized prior to mitochondrial stress test analysis using a Seahorse Biosciences XF24 Analyzer where carbon dioxide evolution rates (CDER) was measured as a surrogate for TCA cycle function (mean ± SD, n = 10).

57 A

B

Figure 2.14 PER2 is Associated with SIRT3 and IDH2 Stable lentiviral PER2 KD or Scr control HMEC-1 were incubated in Hx or Nx prior to co- immunoprecipitation using a (A) SIRT3 or (B) PER2 antibody. Eluted protein was analyzed by immunoblotting for (A) SIRT3, PER2, or IDH2 or (B) PER2 or SIRT3. One representative blot of three is shown.

58 A

B

Figure 2.15 PER2 and HIF1⍺ Dependent Regulation of SIRT3 Gene Expression (A) SIRT3 transcript levels from stable lentiviral mediated PER2KD and Scr control HMEC-1 incubated in Hx or Nx were assessed by qRT-PCR relative to the housekeeping gene β-ACTIN. (B) SIRT3 transcript levels from stable lentiviral mediated HIF1⍺ KD and Scr control HMEC-1 incubated in Hx or Nx were assessed by qRT-PCR relative to the housekeeping gene β-ACTIN (mean ± SD, n = 6).

59 Promoter 2 >FP014148 SIRT3_2 :+U EU:NC; range -499 to 100. TTCGAGGTATTTTGGGCAAGGACAAGCACTTCTTGCGTGTTCGTTGTCGACACCTCGCCC ACTCTTTCTCACCTTGCCGTGTTTCCTTTGCGGCATGTCTGGGAGTTTACGTGCGTTCCA ACACCGGGGCCTCCGTCTTTATGCTGATCTGTGGACGTCACTTATGGGGGTTTTGTCGTT TATTTCACCATTCCCTCATTGAAATACGTTCTAACCTTTAGTTATTGTGCATCATGCTGT GGTGAATAACCTTGTTGGCACCTCACTGTTTGCATAAATTCCTAGTACTGGGATTGCTAG GTCCAAAAGTATATGCATTTAACGCTTTAATGATCCCTCATAAATGTTGTGCGGTTAATA CTCCCACTGGCATGTGATCGTACCCGTCTCCATAGTAATTGCATCTAAGAACTGAGGCTA CATAGGAAAAAATCTTAAGGTCCATGGTTGTTAGGTATCAGATGCCCTATAATGGATGAA TATTTAAGTAAATAATGGCACATTGACTTGCTGTCTGACACAACTGCTGAAAATGTTGGT GAAAAGTGAGTCTCGAAGTGGTAGCTCACACTTGTAATACCAGCACTTTCGGAGGCTGAG

Figure 2.16 SIRT Promoter with HRE Binding Site Region of the SIRT3 promoter containing an HRE binding site, indicated in red.

60 A

B

Figure 2.17 SIRT3 Protein Levels Depend Upon PER2 Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were incubated in Hx or Nx prior to protein isolation. (A) Immunoblot of SIRT3 protein in reference to the loading control β-ACTIN. One representative blot of three is shown. (B) Quantification of SIRT3 protein expression immunoblots analyzed from the technical and biological replicates (mean ± SD, n = 5).

61 A

B

Figure 2.18 SIRT3 Protein Levels Depend on HIF1⍺ Stable lentiviral mediated HIF1⍺ or Scr control HMEC-1 were incubated in Hx or Nx prior to protein isolation. (A) Immunoblot of SIRT3 protein in reference to the loading control β-ACTIN. One representative blot of three is shown. (B) Quantification of SIRT3 protein expression immunoblots analyzed from the technical and biological replicates (mean ± SD, n = 10).

62

Figure 2.19 PER2 Is Important for Mitochondrial Respiration Mitochondrial function was assessed using a fatty acid oxidation mitochondrial stress test and Seahorse Biosciences XF24 Analyzer. Stable lentiviral mediated PER2 KD or Scr control HMEC- 1 were serum starved prior to measuring oxygen consumption rate (OCR) in response to injected compounds (oligomycin: ATP synthase inhibitor, FCCP (carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone): uncoupling agent, rotenone/antimycin A: Complex I and III inhibitor, etomoxir (ETO): inhibitor of fatty acid oxidation). Mitochondrial stress test profile is shown comparing PER2 KD (bright red line), PER2 KD plus etomoxir (dark red line), Scr (black line), and Scr plus etomoxir (gray line) (n = 5).

63

A B

C D

Figure 2.20 Quantification of PER2 in Mitochondrial Respiration Quantification of the mitochondrial stress test (Figure 2.19) comparing PER2 KD (black bar) and Scr control (white bar) HMEC-1. (A) Basal respiration (measured before the addition of oligomycin, mean±SD, n=5), (B) maximum achievable respiration (measured after the addition of the mitochondrial oxidative phosphorylation uncoupler FCCP and before addition of the respiration inhibitors antimycin A and rotenone, mean±SD, n=5), (C) ATP production (mean±SD, n=5), and (D) spare capacity (mean±SD, n=5).

64

Figure 2.21 Ability of PER2 to Use Fatty Acids for Fuel in Mitochondrial Respiration Quantification of maximal respiration due to endogenous fatty acids (FAs) from mitochondrial stress tests (Figure 2.19) using a Seahorse Bioanalyzer (PER2 KD: black bar and Scr control: white bar, mean ± SD, n=5).

65

Figure 2.22 PER2 in Mitochondrial Respiration at ZT12 Mitochondrial function was assessed using a fatty acid oxidation mitochondrial stress test and Seahorse Biosciences XF24 Analyzer. Stable lentiviral mediated PER2 KD or Scr control HMEC- 1 were serum starved prior to measuring oxygen consumption rate (OCR) in response to injected compounds (oligomycin: ATP synthase inhibitor, FCCP (carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone): uncoupling agent, rotenone/antimycin A: Complex I and III inhibitor, etomoxir (ETO): inhibitor of fatty acid oxidation). Mitochondrial stress test profile is shown comparing PER2 KD (bright red line), PER2 KD plus etomoxir (dark red line), Scr (black line), and Scr plus etomoxir (gray line) (n = 5). Here, this mitochondrial stress test was administered at ZT12, twelve hours after the mitochondrial stress test was administered in Figure 2.19.

66

Figure 2.23 Quantification of Circadian Variation in Mitochondrial Respiration Quantification of mitochondrial function (maximal respiration) at ZT0 (Figure 2.19) and ZT12 (Figure 2.22) using a Seahorse Biosciences XF24 Analyzer in stable lentiviral PER2 KD (black bar) or Scr control (white bar) HMEC-1 (mean ± SD, n=10).

67

Figure 2.24 Complex 4.2 Gene Expression is Dependent upon Hypoxia and PER2 Stable lentiviral PER2 KD or Scr control HMEC-1 were synchronized prior to 24 h Hx or Nx incubation, RNA isolation, AND analysis of Complex IV subunit 2 (COX4.2) transcript expression relative to the housekeeping gene β-ACTIN by qRT-PCR (mean ± SD, n=6).

68

Figure 2.25 Complex IV Activity in Ischemia is PER2 Dependent Wildtype (WT) C57BL/6 mice (white bars) or PER2-/- mice (black bars) were subjected to either 0 (C) or 45 minutes (I45m) of myocardial ischemia prior to analysis of cardiac Complex IV enzyme activity (mean ± SD, n=4).

69

Scrambled

PER2 KD

Figure 2.26 Dysfunction in Mitochondrial Potential in PER2 KD HMEC-1 (MitoTracker) Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were incubated in normoxia prior to MitoTracker Red CMXRos staining (red). One representative image of 5 shown.

70

Figure 2.27 Cell Energy Phenotype of PER2 KD HMEC-1 Stable lentiviral mediated PER2 KD (orange) or Scr control (blue) HMEC-1 were analyzed using the cell energy phenotype test on the Seahorse Biosciences XF24 Analyzer. Quiescent phenotype indicates the cells are not energetic for either metabolic pathway; energetic phenotype indicates the cells use both metabolic pathways; aerobic phenotype indicates the cells use predominantly mitochondrial respiration; and a glycolytic phenotype indicates the cells use predominantly glycolysis (mean ± SD, n=10).

71

Figure 2.28 PER2 Dependent Defect in Mitochondrial Membrane Potential in Normoxia Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were incubated in Nx prior to JC-1 staining and analysis by microscopy for calculating mean intensity. Aggregate represents hyperpolarized cells and monomer represents depolarized cells (representative image shown, n = 3).

72

Figure 2.29 PER2 Dependent Defect in Mitochondrial Membrane Potential in Hypoxia Stable lentiviral mediated PER2 KD or Scr control HMEC-1 were incubated in Hx prior to JC-1 staining and analysis by microscopy for calculating mean intensity. Aggregate represents hyperpolarized cells and monomer represents depolarized cells (representative image shown, n = 3).

73

Figure 2.30 Quantification of JC-1 Assays Stable lentiviral mediated PER2 KD (black bars) or Scr control (white bars) HMEC-1 were incubated in Nx (Figure 2.28) or Hx (Figure 2.29) prior to JC-1 staining and analysis by microscopy for calculating mean intensity. Aggregate represents hyperpolarized cells and monomer represents depolarized cells. Quantification was performed using mean intensity and calculating the ratio of aggregate/monomer after background subtraction where a lower ratio indicates a more depolarized cell (mean ± SD, n=6).

74

A

B

Figure 2.31 PER2 Dependent Glycolytic and TCA Cycle Flux in Hx vs Nx U-13C-glucose was added to the supernatant of PER2 KD (black bars) or Scr control (white bars) HMEC-1. Following 24 h of Nx or 1% Hx treatment cells were harvested and analyzed for 13C- metabolites using liquid chromatography–tandem mass spectrometry. (A) 13C-fructose-6- phosphate is a metabolite of glycolysis. (B) 13C-⍺-ketoglutarate is a metabolite of the TCA cycle. Data are presented as the percentage of total metabolite present (mean ± SD, n=3).

75

Figure 2.32 PER2 in PPP Flux in Hx vs Nx U-13C-glucose was added to the supernatant of PER2 KD (black bars) or Scr control (white bars) HMEC-1. Following 24 h of Nx or 1% Hx treatment cells were harvested and analyzed for 13C- metabolites using liquid chromatography–tandem mass spectrometry. 13C-6-phosphogluconate is a metabolite of the PPP. Data are presented as percentage of total metabolite present (mean ± SD, n=3).

76

A

B

Figure 2.33 PER2 in Lipid and Fatty Acid Metabolism in Hx vs Nx 1,2-13C-palmitic acid was added to the supernatant of PER2 KD (black bars) or Scr control (white bars) HMEC-1. Following 24 h of Nx or 1% Hx treatment cells were harvested and analyzed for 13C-metabolites using liquid chromatography–tandem mass spectrometry. (A, B) 13C-palmitic acid and 13C-oleic acid are metabolites of lipid metabolism. Data are presented as percentage of total metabolite present (mean ± SD, n=3).

77

Figure 2.34 PER2 Hypoxia-Only Pathways Co-immunoprecipitation of PER2 from Hx or Nx HMEC-1 prior to affinity purification-mass spectrometry proteomics analysis (Figure 2.10) revealed PER2-hypoxia-only-pathways through PANTHER (Protein ANalysis THrough Evolutionary Relationships), including many PER2 metabolic targets involved in glycolysis, the TCA cycle, the PPP and lipid metabolism.

78 CHAPTER 3 CHAPTER III PER2 MICRORNA TARGETS IN METABOLIC ADAPTATION TO HYPOXIA3

3.1 Rationale

Targeting metabolic pathways is a key strategy for adaptation to low oxygen conditions

and understanding the interconnection between microRNAs, circadian rhythmicity, and cellular

metabolism in hypoxia has the potential to reveal new therapeutic strategies of cardioprotection

during myocardial ischemia. While a single microRNA may target multiple transcripts within a cell

type, the contribution of circadian microRNAs to hypoxic metabolic adaptation are mostly

unknown. To identify microRNA-based endogenous cardioprotective pathways in conditions of low oxygen availability, we performed a screening experiment to study transcriptional changes of

PER2 dependent microRNAs during cardioprotective IPC of the heart (69-71, 76, 159, 265, 266).

In this chapter, we present our mechanistic understanding of a specific PER2 dependent microRNA, that we uncovered to be circadian in nature and contribute to hypoxic metabolic

adaptation (264).

3.2 Results

Differential and PER2 dependent regulation of miR-21 after IPC

Our previously published studies showed abolished cardioprotection by IPC in PER2-/-

mice (71). Based on these studies, we pursued a wide microRNA screen of cardiac PER2

dependent microRNAs (Table 3.1). Out of 352 microRNAs analyzed, 186 were regulated in both wildtype and PER2-/- mice, 65 were only regulated in wildtype, and 22 were only regulated in

PER2-/- mice. As shown in Table 3.1, differential regulation of putative PER2 dependent microRNAs revealed almost exclusively cardioprotective pathways. Ingenuity analysis revealed a

3 This chapter is condensed from our publication (264). Yoshimasa Oyama performed the mouse model of ischemic preconditioning and harvested the heart and lungs. Colleen Bartman isolated mouse cardiac myocytes or cardiac fibroblasts, completed the microRNA and messenger RNA isolation on harvested hearts or cardiac myocytes, fibroblasts, or endothelia, analyzed miR-21 or PER2 transcript by qRT-PCR, and performed the in vitro miR-21 LOF or GOF studies followed by glycolytic stress tests using the Seahorse Bioanalyzer.

79 selective role for miR-21 in protection from reperfusion injury of the heart. After identification of

miR-21 as a potential downstream target of PER2 mediated cardioprotection, we confirmed a

PER2 dependent miR-21 regulation in cardiac tissue from wildtype or PER2-/- mice; while IPC

resulted in a 2.4-fold induction of miR-21 in wildtype mice (Figure 3.1), no upregulation of miR-

21 was observed in PER2-/- mice (Figure 3.1). Taken together, these data demonstrate that IPC induced PER2 regulates cardiac miR-21.

Diurnal expression pattern of miR-21 in murine tissues

Considering cardiac miR-21 is regulated in a PER2 dependent manner, and cardiac PER2

has a diurnal oscillation pattern, we next investigated the expression pattern of this microRNA

over a 12h period. Hearts from wildtype mice were harvested at Zeitgeber time (ZT) 3 or ZT15.

We found significantly higher cardiac miR-21 expression levels at ZT15 compared to ZT3 (1.8-

fold increase from ZT3 to ZT15, Figure 3.2). Consistent with our previously published studies

(71), analysis of the same heart tissue confirmed significantly higher cardiac PER2 mRNA levels at ZT15 compared to ZT3 (29-fold increase from ZT3 to ZT15, Figure 3.2). To verify a diurnal nature of miR-21 we analyzed another organ in addition to the heart. Indeed, analysis of lung tissue from these wildtype mice at the indicated times revealed lung miR-21 and PER2 mRNA levels significantly higher at ZT15 than ZT3, which was consistent with our findings in the heart

(lung miR-21 3.7-fold increase and lung PER2 mRNA 6.9-fold increase from ZT3 to ZT15, respectively, Figure 3.3). Taken together, our studies show that murine miR-21 expression oscillates over the circadian day in heart and lungs (ZT3 vs ZT15, p < 0.05), similar to PER2 and implying a putative circadian expression pattern of miR-21.

MiR-21 is exclusively upregulated in hypoxic cardiac endothelial cells

After confirming that miR-21 is a PER2 regulated microRNA with a diurnal expression

pattern in heart and lungs, we next investigated which cardiac cell type expressed miR-21 during

conditions of low oxygen availability. Based on previous findings that miR-21 is predominantly

expressed in cardiac fibroblasts (267), we obtained fibroblasts, myocytes or endothelial cells from

80 wildtype mouse hearts. Indeed, analysis of relative miR-21 expression levels indicated an

abundant expression of miR-21 in cardiac fibroblasts when compared to other cardiac cell types

(Figure 3.4). However, since fibroblasts play a dominant role during remodeling (268) but not

during the acute phase of myocardial ischemia and reperfusion (266), we next exposed isolated murine cardiac fibroblasts, myocytes or endothelial cells to 1% oxygen (hypoxia). No significant regulation of miR-21 was found in fibroblasts or myocytes upon hypoxia exposure when compared to cells at ambient oxygen levels (normoxia, Figure 3.5). However, isolated murine cardiac endothelial cells exposed to hypoxia revealed a robust and significant upregulation of miR-21

(4.9-fold increase compared to ambient oxygen levels, Figure 3.5). Further analysis using human microvascular endothelial cells (HMEC-1) confirmed a miR-21 upregulation in hypoxia (8.6-fold increase in 1% hypoxia, Figure 3.5). Taken together, while miR-21 is predominantly expressed in cardiac fibroblasts at baseline, only cardiac endothelial cells revealed a significant upregulation of miR-21 upon hypoxia exposure. These data suggest that endothelial expressed miR-21 plays a critical role during conditions of low oxygen availability, such as myocardial ischemia.

MiR-21 is critical for cellular glycolytic reliance

After confirming that miR-21 is a hypoxia regulated microRNA with predominant upregulation in hypoxic cardiac endothelial cells, we next analyzed the role of miR-21 in known

PER2 regulated pathways. Our recent studies found an important role of PER2 in controlling glycolysis during myocardial ischemia (71, 74, 136). To understand a potential role of miR-21 in glycolysis, we first performed loss of function (LOF) studies using miR-21 inhibitors. Anti-miR-21 inhibitors were transfected into human microvascular endothelial cells (HMEC-1), a cell line well characterized for hypoxic, metabolic and PER2 pathways (71). We first confirmed a knockdown of miR-21 in HMEC-1s and found a 70% reduction of miR-21 expression (Figure 3.6). In miR-21

knockdown HMEC-1s, we assessed glycolysis, glycolytic capacity, and glycolytic reserve using a

glycolytic stress test and Seahorse Bioanalyzer (Figure 3.6). Loss of miR-21 significantly reduced glycolysis (10.7-fold, Figure 3.7), glycolytic capacity (31-fold, Figure 3.7) and glycolytic reserve

81 (31-fold, Figure 3.7). In contrast, our gain of function (GOF) studies done by overexpressing a

miR-21 mimic (22-fold overexpression, Figure 3.8) significantly increased glycolysis (1.3-fold,

Figure 3.9), glycolytic capacity (1.6-fold, Figure 3.9) and glycolytic reserve (2.3-fold, Figure 3.9) in HMEC-1s. Taken together, these studies demonstrate that miR-21 is necessary to maintain glycolysis, critical for the cell to maximally respond to glycolytic demand (glycolytic capacity), and pertinent for glucose reserves to be available for use through glycolysis beyond baseline

(glycolytic reserve).

3.3 Discussion

We pursued identification of microRNAs that could mimic circadian rhythm protein PER2 mediated cardioprotection. Profiling the PER2 dependent expression of 352 micro RNAs following cardioprotective IPC of the heart indicated an exclusive role for miR-21. Analysis of three cardiac tissues revealed hypoxia induced miR-21 predominantly in cardiac endothelial cells. Studies on miR-21 expression revealed a PER2 dependent and putative circadian profile. Using miR-21 LOF or GOF in HMEC-1 identified a critical role of miR-21 for PER2 regulated cellular glycolysis.

The multi-functional role of miR-21 (Figure 3.10) includes preventing cardiomyocyte cell death through targeting the programmed cell death 4 gene (PDCD4 (269)). Other miR-21 target genes are Fas ligand (FasL) or tensin homology deleted on chromosome 10 (PTEN) and are also antiapoptotic (270). Besides regulating apoptosis, miR-21 has been implicated in the attenuation of inflammation or in the angiogenic repair process of ischemic injury via a decrease of NF-kappa

B or through the PTEN/AKT/ERK1-VEGFpathway, respectively (270). Recent studies indicate that miR-21 also reduces hydrogen peroxide-induced apoptosis in cardiac stem cells through

PTEN/PI3K/AKT signaling (271). Moreover, bone marrow derived mesenchymal stem cells overexpressing miR-21 efficiently repair myocardial damage in rats (272). Our studies on miR-21 in contribute to the understanding of miR-21 in multiple cellular processes.

Research on circadian microRNAs in the heart has been extremely limited. We propose miR-21 to be circadian based on a PER2 dependent regulation and findings on a diurnal

82 expression pattern like that of PER2. However, high temporal resolution gene expression analysis

would be necessary to further support that miR-21 is indeed circadian (273). In general, while several circadian microRNAs have been implicated in cardioprotection from ischemia (Table 3.1), studies on cardiac circadian microRNAs are scarce (207). However, a recent elegant study on sepsis which also has been shown to be time of day dependent, discovered miR-155 as circadian microRNA with profound effects on circadian function and circadian induction of cytokines by LPS

(170). If a miR-21 knockdown in mice could have similar effects on circadian function seems compelling but would need to wait further characterization of period lengths during circadian manipulation.

The critical role of miR-21 in glycolysis seems surprising. However, it was shown that one of PER2 mediated mechanisms is controlling transcription as a cofactor (42). In fact, the transcription factor HIF1⍺ is the key regulator of glycolysis (274) and studies have shown that

PER2 and HIF1⍺ are bound together during ischemia of the heart (71). In support of our current findings, HIF1⍺ was found to regulate miR-21 gene expression and miR-21 promotes HIF1⍺ stabilization in hypoxic cardiomyocytes (275). Data from these studies would therefore suggest that the PER2-HIF1⍺ complex is responsible for the transcriptional regulation of miR-21 during myocardial ischemia. In line with these findings, recent studies on miR-21 in small lung cancer cells revealed a similar connection between miR-21, glycolysis and HIF1⍺ (276). However, how miR-21 controls glycolysis would need further mechanistic studies.

83 A

B

Figure 3.1 miR-21 Regulation in WT and PER2-/- Mice After Cardiac IPC Wildtype (A) or PER2-/- mice (B) were exposed to cardiac IPC, consisting of 4 x 5 minutes of ischemia followed by 5 minutes of reperfusion each, followed by a final reperfusion time of 120 min. Heart tissue was snap-frozen with clamps pre-cooled to the temperature of liquid nitrogen. Transcript levels were determined by qRT-PCR (mean ± SD, n = 3).

84 A

B

Figure 3.2 Diurnal Expression of MiR-21 in Murine Hearts Analysis of cardiac mir-21 (A) and PER2 (B) levels from wildtype mice at Zeitgeber Time (ZT) 3 or ZT15. Transcript levels were determined by qRT-PCR (mean ± SD, n = 3, p<0.05).

85 A

B

Figure 3.3 Diurnal Expression of MiR-21 in Murine Lungs Analysis of lung mir-21 (A) and PER2 (B) levels from wildtype mice at Zeitgeber Time (ZT) 3 or ZT15. Transcript levels were determined by qRT-PCR (mean ± SD, n = 3, p<0.05).

86

Figure 3.4 Relative miR-21 Expression in Different Cardiac Tissues Fibroblasts or myocytes were isolated from C57BL6/J mouse hearts and endothelial cells isolated from C57/BL6 mice were purchased from Cell Biologics for analyzing miR-21 expression at baseline (mean ± SD, n = 3).

87 A B

C D

Figure 3.5 miR-21 Expression in Different Cardiac Tissues at Baseline and During Hypoxia (A) miR-21 expression in cardiac fibroblasts subjected to normoxia or hypoxia for 6 h (mean ± SD, n = 6, not significant). (B) miR-21 expression in cardiac myocytes subjected to normoxia or hypoxia for 1 h (mean ± SD, n = 3, not significant). (C) miR-21 expression in cardiac endothelial cells subjected to normoxia or hypoxia for 6 h (mean ± SD, n = 6, p<0.05). (D) miR-21 expression in human endothelia (HMEC-1) subjected to normoxia or hypoxia for 6 h (mean ± SD, n = 6, p<0.05).

88 A

B

Figure 3.6 Glycolysis in MiR-21 Loss of Function Human Endothelial Cells (A) Knockdown confirmation in anti-mir-21 (loss of function, LOF, black bar) compared to SCR control (white bar) treated HMEC-1 (mean ± SD, n = 6, p<0.05). (B) Glucose metabolism profile of glycolytic function from control (miScript Inhibitor Neg. Control, scrambled [SCR], black line) and anti-mir-21 (LOF, gray line) treated HMEC-1. Glycolytic stress tests were done using a Seahorse Bioanalyzer XF24 instrument. The extracellular acidification rate (ECAR) was measured in response to glucose (stimulates glycolysis), oligomycin (inhibits ATP synthase), and 2-DG (2-deoxy-glucose, inhibits glycolysis) (n = 10).

89 A B C

Figure 3.7 Quantification of Glycolysis in MiR-21 Loss of Function Human Endothelial Cells (A-C) Quantification of glucose metabolism from control (miScript Inhibitor Neg. Control, scrambled [SCR], white bars) and anti-mir-21 (LOF, black bars) treated HMEC-1. Glycolytic stress tests were done using a Seahorse Bioanalyzer XF24 instrument (Figure 3.6). The extracellular acidification rate (ECAR) response to glucose, oligomycin and 2-DG was measured (mean ± SD, n = 10, p<0.05).

90 A

B

Figure 3.8 Glycolysis in MiR-21 Gain of Function Human Endothelial Cells (A) Overexpression in miR-21Mimic (gain of function [GOF], black bar) treated HMEC-1 (mean ± SD, n = 6, p<0.05). (B) Glucose metabolism profile of glycolytic function from control (miScript miRNA Mimic Neg. Control, SCR, black line) and miR-21Mimic (GOF, gray line) treated HMEC- 1. Glycolytic stress tests were done using a Seahorse Bioanalyzer XF24 instrument. The extracellular acidification rate (ECAR) was measured in response to glucose, oligomycin and 2- DG (n = 10).

91 A B C

Figure 3.9 Quantification of Glycolysis in MiR-21 Gain of Function Human Endothelial Cells (A) Quantification of glucose metabolism from control (miScript miRNA Mimic Neg. Control, SCR, white bars) and miR-21 Mimic (GOF, black bars) treated HMEC-1. Glycolytic stress tests were done using a Seahorse Bioanalyzer XF24 instrument (Figure 3.8). The extracellular acidification rate (ECAR) response to glucose, oligomycin and 2-DG was measured and analyzed (mean ± SD, n = 10, p<0.05).

92

Figure 3.10 Potential upstream and downstream targets of miR-21 Potential upstream and downstream targets of miR-21. MiR-21 is located on Chromosome 17 and has been found to be a hypoxia inducible factor 1 α (HIF1⍺) target gene. Through AKT1 activation, miR-21 can increase its own transcription via a positive feedback loop. During myocardial ischemia HIF1⍺ and Period 2 (PER2) bind together. PER2 was shown to be essential in the regulation of HIF1⍺ target genes, which explains recent findings on a PER2 dependent regulation of miR-21. The main mechanisms how miR-21 reduces cardiac cell death is through targeting programmed cell death 4 (PDCD4), Fas ligand (FASL), and phosphatase and tensin homology deleted on chromosome 10 (PTEN). Inflammation may also be reduced because of a decrease in NF-κB. Recent studies indicate that miR-21 also protects from ischemia trough upregulation of oxygen efficient glycolysis – probably via stimulation of HIF1 and PER2 dependent phosphofructokinase (PFK). Some data, however, indicate that miR-21, induced by transforming growth factor beta receptor 3 (TGFBR3) stimulation, exacerbates tissue fibrosis via suppression of Sprouty homolog 1(SPRY1).

93 Table 3.1 PER2 Dependent MicroRNAs During Cardiac IPC Shown are the 65 differentially regulated and Per2 dependent micro RNAs identified after IPC treatment of wildtype and PER2-/- mice. IP regulated FC FC Function Ref. species prevents cardiac mmu-miR-16 4.0278 -1.2311 hypertrophy (277) R mmu-miR-409- role in the heart not studied 5p 6.9644 1.9053 yet protect against cardiac mmu-miR-154* 5.6962 1.4241 dysfurnction (278) M role in the heart not studied mmu-miR-326 5.6569 1.9053 yet inhibits cardiomyocyte mmu-miR-24 5.579 1.7291 apoptosis (279) M increases angiogenesis mmu-miR-27b 5.2054 1.6702 heart (280) M inhibits cardiomyocyte mmu-miR-146a 5.1337 1.9725 apoptosis (281) M mmu-miR-126- role in the heart not studied 3p 4.7568 -1.1329 yet mmu-miR-25 4.6268 1.2397 protects cardiomyocytes (282) R mmu-miR-23b 4.4383 1.7654 upregulated in heart failure (283) H diagnosis of unstable mmu-miR-186 4.4383 -1.021 angina pectoris (284) H mmu-miR-191 4.4076 1.3195 biomarkers for Ml (285) H protect against cardiac mmu-miR-150 4.2871 1.3195 fibrosis (286) M mmu-miR-342- biomarker heart failure 3p 4.2281 1.4948 (287) H increases cardiomyocyte mmu-miR-28 4.1411 1.0867 apoptosis (288) M prevents cardiac mmu-miR-99a 4.1411 1.1487 hypertrophy (289) M protects against cardiac mmu-miR-322 4.1125 -1.0867 dysfunction (290) M protect against cardiac mmu-miR-30a 4.084 1.7053 dysfurnction (291) M increases cardiac mmu-miR-23a 4.084 1.9185 hypertrophy (292) M mmu-miR-181b 4.0278 1.9053 upregulated in heart failure (283) H mmu-miR-101a 3.9724 1.1567 protects cardiac fibroblasts (293) R regenerative neonatal mmu-miR-505 3.9313 1.1368 mouse heart (294) M prevents cardiac mmu-miR-20a 3.9177 -1.3755 hypertrophy (295) R increases cardiac mmu-miR-199b 3.8637 -1.3755 hypertrophy (296) M

94 Table 3.1 Cont’d

IP regulated FC FC Function Ref. species mmu-miR-324- attenuates cardiomyocyte 5p 3.8106 1.5157 apoptosis (297) M prevents cardiac mmu-miR-30c 3.7842 1.6818 hypertrophy (298) M biomarkers for LV mmu-miR-208b 3.7064 -2.4116 remodling (299) H role in the heart not studied mmu-miR-301a 3.6808 -1.5583 yet role in the heart not studied mmu-let-7d 3.6301 1.3287 yet mmu-miR-582- role in the heart not studied 5p 3.6175 -4.4229 yet mmu-miR- role in the heart not studied 466d-3p 3.5554 -1.1173 yet increases cardiac mmu-miR-155 3.5308 -1.0425 hypertrophy (300) M mmu-miR-532- increases cardiomyocyte 3p 3.4224 1.9931 apoptosis (301) M role in the heart not studied mmu-miR-30e 3.4105 -1.0644 yet mmu-miR-126- role in the heart not studied 5p 3.3636 -1.0425 yet increases cardiac mmu-miR-350 3.3636 1.1647 hypertrophy (302) R role in the heart not studied mmu-miR-148b 3.3404 1.3566 yet prevents cardiac mmu-miR-541 3.3173 1.3947 hypertrophy (303) M role in the heart not studied mmu-miR-301b 3.3173 -1.1647 yet role in the heart not studied mmu-miR-181d 3.2944 1.0425 yet role in the heart not studied mmu-miR-106b 3.2266 -1.2924 yet mmu-miR-151- prevents arrythmias in Ml 5p 3.2266 1.0353 (304) R role in the heart not studied mmu-miR-128a 3.2043 1.6133 yet mmu-miR-425 3.1821 1.4044 regulates ANP product ion (305) H role in the heart not studied mmu-miR-152 3.1821 1 yet role in the heart not studied mmu-miR-99b 3.1383 1.7901 yet mmu-miR-22 3.1167 1.4845 Cardioprotective (306) R

95 Table 3.1 Cont’d

IP regulated FC FC Function Ref. species role in the heart not studied mmu-miR-467c 3.1167 -1.2924 yet mmu-miR-21 3.0738 -1.5369 IP heart, cardioprotection (161) M role in the heart not studied mmu-miR-872 3.0738 -1.0867 yet role in the heart not studied mmu-let-7i 3.0738 1.021 yet role in the heart not studied mmu-miR-18a 3.0738 -1.5692 yet role in the heart not studied mmu-let-7c 3.0525 1.4743 yet role in the heart not studied mmu-miR-17 3.0525 -1.257 yet role in the heart not studied mmu-let-7f 3.0314 -1.366 yet role in the heart not studied mmu-miR-467e 3.0105 -1.6818 yet role in the heart not studied mmu-miR-219 2.2974 -4.1411 yet mmu-miR-675- role in the heart not studied 5p 1.9521 3.1932 yet mmu-miR- role in the heart not studied 302c* 1.7654 24.4201 yet role in the heart not studied mmu-miR-742 1.0281 3.9449 yet role in the heart not studied mmu-miR-216a -1.057 4.4076 yet - role in the heart not studied mmu-miR-343 3.1711 -1.244 yet - role in the heart not studied mmu-miR-744 3.3636 -1.014 yet - role in the heart not studied mmu-miR-679 3.4943 -1.4191 yet mmu-miR-292- - role in the heart not studied 3p 5.8159 -1.4743 yet IP= Ischemic Preconditioning, FC=Fold Change.

96 CHAPTER 4 CHAPTER IV IMPLICATIONS OF CIRCADIAN DISRUPTION IN MYOCARDIAL ISCHEMIA4

4.1 Rationale

Based on the identified role of PER2 in metabolic adaptation to hypoxia, we hypothesized that circadian disruption and subsequent alteration of PER2 expression would be detrimental if myocardial IR-injury would occur. This is supported in previous findings that the onset of

myocardial infarction has a distinct circadian pattern and it has been suggested that disruption of

circadian rhythms may contribute to cardiovascular disease (23, 184, 185). Furthermore, several

medications used in a daily clinical setting and constant light exposure experienced in the ICU

can disrupt the well-coordinated expression of the circadian rhythm network. In particular,

anesthetics were found to disrupt the expression of circadian proteins, including PER2 (187-190,

196, 197, 307) and propofol anesthesia in humans affects the circadian period (196). Animal

studies found ketamine to reduce the amplitude of circadian rhythm gene expression (198) and

studies using sevoflurane, a volatile anesthetic similar to isoflurane, found suppression of PER2

in the brain of rodents (199). However, research on the influence of anesthetics on circadian

rhythm protein PER2 expression in the heart and whether anesthetic driven circadian disruption

would be detrimental during myocardial IR-injury is not known.

4.2 Results

Midazolam downregulates cardiac PER2 in wildtype mice

Based on observations that anesthesia can lead to a dysregulation of circadian rhythm

proteins (196, 199, 308), we exposed mice to a variety of anesthetics used in a daily clinical

routine. Fentanyl (1 mg/kg), isoflurane (1%), ketamine (200mg/kg), propofol (200 mg/kg), or

midazolam (200 mg/kg) were given at a dose which caused a loss of righting reflex and loss of

4 Portions of this chapter are from our manuscript currently in review. Yoshimasa Oyama performed the myocardial IR-injury model, infarct staining, and troponin measurements. Colleen Bartman performed the isolation of cardiac myocytes or cardiac fibroblasts and the analysis of PER2 transcript upon midazolam treatment of in vitro cardiac myocytes, fibroblasts, or endothelia.

97 the pedal withdrawal reflex (309). Two hours later we analyzed cardiac PER2 mRNA expression

levels. While fentanyl, isoflurane, ketamine, or propofol had no significant effects on PER2

expression, the benzodiazepine midazolam revealed a robust and significant downregulation of

cardiac PER2 transcript levels (0.3-fold, p<0.05; Figure 4.1). To understand in which cardiac cell

type PER2 was mostly downregulated, we obtained cardiac endothelia, fibroblasts or isolated

cardiomyocytes from C57BL/6 wildtype mice. After confirmation of cell viability, primary murine

endothelia, fibroblasts or cardiomyocytes were exposed to midazolam (50 µM) or 0.9% NaCl

vehicle for 6 hours. As shown in Figure 4.2, midazolam significantly reduced PER2 mRNA

expression in murine cardiac endothelial cells and fibroblasts but had no effect on primary murine

cardiomyocytes. Taken together, when comparing fentanyl, propofol, ketamine, isoflurane, and

midazolam, only the benzodiazepine midazolam significantly reduces murine cardiac PER2

mRNA levels.

Midazolam has deleterious effects during myocardial IR-injury

After we found that midazolam significantly downregulated cardiac PER2 mRNA levels, we next pursued myocardial IR-injury studies following midazolam administration. To control for the sedative effects of midazolam, we used fentanyl as a secondary control as it had no effect on cardiac PER2 mRNA levels (Figure 4.1). C57Bl/6 wildtype mice were pretreated with 200 mg/kg of midazolam, 1mg/kg fentanyl, or vehicle. Two hours later wildtype mice underwent 60 min of myocardial ischemia followed by 120 min of reperfusion. As shown in Figure 4.3, fentanyl pretreatment had no significant effect on infarct sizes when compared to vehicle treated controls.

However, midazolam pretreatment significantly increased infarct sizes when compared to vehicle or fentanyl treated animals (mean ± SD, vehicle: 42.2 ± 9.7 % or fentanyl: 36.7 ± 12.1% vs midazolam 59.3 ± 5.2%, p<0.05; Figure 4.3). To confirm our results from the infarct size analysis which were obtained via TTC and Evans Blue double staining, we next determined Troponin-I serum levels using ELISA. Indeed, midazolam pretreated mice had significantly higher Troponin-

I serum levels when compared to vehicle or fentanyl controls (mean ± SD: vehicle: 54.1 ± 29.1

98 ng/ml or fentanyl: 24.9 ± 13.1 ng/ml vs midazolam 161.7 ± 57.0 ng/ml, p<0.05; Figure 4.4). Taken together, midazolam mediated downregulation of cardiac expressed and cardioprotective murine

PER2 is associated with deleterious consequences following myocardial ischemia and reperfusion injury.

4.3 Discussion

We demonstrated that the benzodiazepine midazolam is associated with a downregulation in cardiac PER2 endothelia and fibroblasts. Furthermore, midazolam administration during myocardial ischemia and reperfusion injury significantly increases infarct size and troponin levels.

While circadian rhythm disruption following general anesthesia is a recognized phenomenon, its role has not been fully elucidated yet (4, 5). Previous animal studies have shown that midazolam abolishes the cardioprotective effects of IPC while the benzodiazepine receptor antagonist

Flumazenil is cardioprotective in rabbits (198). Mechanistic studies on midazolam in abolishing cardiac IPC suggest that midazolam via GABA receptors inhibits the mitochondrial adenosine triphosphate-sensitive potassium (K-ATP) channels, which in turn can activate the cardioprotective protein kinase C-e (PKC-e) – an important component within the preconditioning cascade (198). Another study had similar findings using a cultured chick embryonic cardiomyocyte model of hypoxia and re-oxygenation (310). In fact, the peripheral benzodiazepine receptor, a mitochondrial inner membrane protein, was found to play a role in mitochondrial function during cardiac IR-injury using perfused rat hearts (311). These studies are in support of our findings and suggest that the observed deleterious effect of midazolam during IR-injury could have been a result of mitochondrial K-ATP channel blockade.

99

Figure 4.1 Cardiac PER2 Regulation Following Exposure of Wildtype Mice to Anesthetics Wildtype mice were exposed to fentanyl (1 mg/kg), isoflurane (1%), ketamine (200mg/kg), propofol (200 mg/kg), or midazolam (200 mg/kg). Two hours later, mouse hearts were harvested for RNA and cardiac PER2 mRNA expression levels were analyzed by qRT-PCR.

100 A B C

Figure 4.2 Cardiac PER2 Regulation in Cardiac Tissues Exposed to Midazolam Murine endothelia obtained from Cell Biologics (A), murine fibroblasts isolated from the wildtype mouse heart (B), or cardiac myocytes isolated from a wildtype mouse (C) were exposed to 50 µM midazolam for 6 hours prior to RNA isolation and PER2 transcript analysis by qRT-PR (mean ± SD, n = 4-6; p<0.05).

101 A

B

Figure 4.3 Midazolam in Myocardial Ischemia and Reperfusion Injury Wildtype mice were treated with vehicle (pentobarbital, 10mg/kg), fentanyl (1mg/kg), or midazolam (200mg/kg) 2 hours prior to myocardial ischemia. Myocardial ischemia consisted of 60 min of ischemia followed by 120 minutes of reperfusion in which all mice received pentobarbital. Infarct sizes were measured by double staining with Evan’s blue and triphenyl- tetrazolium chloride (TTC). Infarct sizes are expressed as the percent of the area at risk (AAR) that underwent infarction. (A) Infarct sizes as the percent of AAR and (B) Representative infarct staining (mean ± SD, n = 5-7; p<0.05).

102

Figure 4.4 Troponin Levels After Midazolam in Myocardial Ischemia and Reperfusion Injury Wildtype mice were treated with vehicle (pentobarbital, 10mg/kg), fentanyl (1mg/kg), or midazolam (200mg/kg) 2 hours prior to myocardial ischemia, in which all mice received pentobarbital. Myocardial ischemia consisted of 60 min of ischemia followed by 120 minutes of reperfusion. Serum troponin I concentrations were measured by enzyme-linked immunosorbent assay (ELISA) (mean ± SD, n = 5-7; p<0.05).

103 CHAPTER 5 CHAPTER V CIRCADIAN AMPLITUDE ENHANCEMENT AS A THERAPEUTIC STRATEGY IN ISCHEMIA5

5.1 Rationale

Lack of efficient circadian entrainment is considered to be primary mechanism underlying the circadian nature of myocardial infarctions (24, 25, 184). The robust release of epidemiological studies in the field of circadian biology puts great emphasis on deciphering underlying mechanisms of this link to ultimately gain understanding and identify targets for prevention or treatment from cardiovascular diseases (5, 74). Given a close association of circadian amplitude dampening and disease progression (312), clock-enhancing strategies are currently under intense investigation. Our data thus far suggest that PER2 is a key mediator of metabolic adaptation to hypoxia by regulating HIF1⍺ and functions in conjunction with miR-21. We also presented the concept that circadian disruption has deleterious effects during myocardial IR- injury. Our next steps were to target PER2, its downstream pathways, and overall circadian amplitude enhancement to promote cardioprotection from myocardial ischemia in our in vivo mouse model. We tested our hypothesis that targeting PER2 and circadian amplitude enhancement would be a cardioprotective strategy by first optimizing our circadian amplitude enhancement protocol, next targeting PER2 amplitude and its downstream metabolic pathways, and lastly used a novel pharmacological PER2 target to reverse the deleterious effects of midazolam driven cardiac damage (135).

5 Portions of this chapter are from either our publication (276) or from our manuscript currently in review. Yoshimasa Oyama performed the myocardial IR-injury model, infarct staining, troponin measurements, in vivo metabolic analysis, in vivo cAMP analysis, and wheel running experiments analysis. Mice were enucleated by a collaborator and analyzed by a prior lab technician. Colleen Bartman performed the cardiac miR-21 or PER2 transcript analysis and generated the in vitro melanopsin light-sensing cell line, performed the light-sensing cell experiments, and completed the metabolic stress tests using the Seahorse Bioanalyzer. Mass spectrometry analysis was done by the Metabolomics Core at the University of Colorado Anschutz Medical Campus.

104 5.2 Results

Daylight-elicited cardioprotection through circadian amplitude enhancement

Daylight is the dominant regulator of human circadian rhythm and PER2 (74). Our previous studies reported cardioprotection from myocardial ischemia following short exposure times to daylight (10,000 LUX), possibly through PER2 mediated metabolic adaptation (71). Here we investigated daylight exposure protocols and found that housing mice under “daylight conditions”

(10,000 LUX, L:D 14:10h) robustly enhances cardio-protection in a time dependent manner

(Figure 5.1, 5.2). Indeed, using PER2 reporter mice, we found increases of peak and trough levels (amplitude) for cardiac PER2 protein levels after one week of daylight housing which coincided with robust cardio-protection from myocardial ischemia (Figure 5.3). To determine whether daylight-elicited entrainment is sex-specific, we applied our light exposure protocol to female mice and found that daylight housing provided similar cardioprotection from myocardial ischemia compared to that exhibited in male mice (Figure 5.4).

To confirm that light-elicited cardiac circadian amplitude enhancement of PER2 required visual light perception, we enucleated wildtype mice to remove light sensing structures. In fact, daylight exposure of ‘fully blind mice’ did not reveal increases of cardiac PER2 during daylight conditions (Figure 5.5). Myocardial ischemia and reperfusion studies in blind mice under normal

(room light) housing conditions found reverse cardiac troponin kinetics and slightly higher overall troponin levels, indicating a lack of circadian amplitude enhancement by light and further supporting our findings of circadian entrainment as a cardio-protective strategy (Figure 5.6, 5.7).

Together, these data conclude daylight-elicited cardiac circadian amplitude enhancement via

PER2 oscillation is a cardio-protective strategy requiring visual light perception.

Wheel running elicits circadian amplitude enhancement

As circadian entrainment can be achieved by external cues other than light, we investigated the effects of voluntary wheel running – a common mechanism used for entrainment

(208) – on circadian amplitude and myocardial infarct sizes. Evaluation of locomotor activity by

105 measuring wheel running activity during daylight housing showed no phase shift of the circadian

period, but an increase of the total distance walked and the circadian amplitude, revealing

enhanced circadian entrainment (Figure 5.8 – 5.10). After two weeks of voluntary wheel running,

a significant increase of the circadian amplitude was observed, which indeed resulted in robust

cardio-protection from ischemia (Figure 5.11). Moreover, the total distance achieved on the wheel

inversely correlated with infarct sizes (Figure 5.11).

Daylight-elicited PER2 activates HIF1⍺ metabolic adaptation to ischemia

Next, we pursued studies to understand the mechanism of daylight-elicited PER2 and circadian amplitude enhancement in cardioprotection from myocardial ischemia. We used liquid chromatography-tandem mass spectrometry studies following the infusion of labeled glucose

(13C-glucose) to assess glycolytic flux rates due to daylight-elicited circadian amplitude enhancement. We found that light-elicited enhanced entrainment significantly increased glycolytic flux in the left ventricle of the mouse heart (Figure 5.12). Following studies on phosphofructokinase, the key and rate limiting enzyme of glycolysis, revealed increased enzyme activity in the heart tissue or plasma in a PER2 dependent manner after one week of daylight exposure (Figure 5.13). Studies using PER2-/- mice confirmed that daylight-elicited circadian amplitude enhancement and cardio-protection was PER2 dependent (Figure 5.14). Adenosine mediated increases of cAMP represent a key regulator of PER2 mediated cardio-protection (71) and studies in mice under constant darkness suggest that light potentially induces cAMP and

PER2 via circulating adenosine (73). Indeed, evaluation of cAMP levels after one week of daylight entrainment revealed significantly increased cardiac cAMP, which was abolished in PER2-/- mice

(Figure 5.15).

Light-sensing HMEC-1 reveal light dependent pathways in metabolism

As a proof of concept, we recapitulated light sensing for PER2 overexpression on a cellular level by generating an HMEC-1 line, overexpressing the human light sensing photopigment melanopsin, a retinal ganglion cell receptor responsible for circadian entrainment. Exposing the

106 light sensing HEMC-1 cultures to light resulted in a significant increase of cAMP, phospho-CREB

(cyclic AMP-responsive element binding protein), and PER2 mRNA (Figure 5.16). In addition, this light exposure increased glycolytic capacity and oxygen consumption rates (Figure 5.17,

5.18). Taken together, these in vivo and in vitro studies on light-elicited pathways identified a

circadian entrainment mechanism through cAMP, dependent on light perception and HIF1⍺

metabolic adaptation in a PER2 regulated manner.

Intense light exposure induces cardiac miR-21

Our in vitro studies presented in Chapter 3 revealed that miR-21 is a circadian microRNA and has a role in glycolytic and mitochondrial functioning. We next sought to determine whether

our light exposure regime could target cardiac miR-21 in mice. To test this, wildtype mice were exposed to one week of intense light at 10,000 lux following our daylight regime (L:D 14:10) and compared to room light at 200 lux (14:10)). Housing mice for one week at intense light significantly increased cardiac miR-21 (6.1-fold, Figure 5.19) when compared to room light housing. As a control for light treatment we also analyzed PER2 levels and found a robust and significant induction of cardiac PER2 mRNA levels (4.1-fold, Figure 5.19), as observed in earlier studies on cardiac PER2 protein (71). Taken together, these data demonstrate that exposing mice to intense light increases miR-21 levels in cardiac tissues, similarly to increases in PER2 mRNA, supporting the hypothesis that miR-21 is a circadian microRNA downstream of PER2.

MiR-21 is required for light-elicited cardioprotection

After demonstrating that miR-21 was necessary for PER2 regulated pathways such as

glycolysis, we next investigated the role of miR-21 in myocardial ischemia and reperfusion injury.

-/- Thus, we first exposed miR-21 or control mice (B6129SF1/J) to myocardial IR- injury. As shown in Figure 5.20, miR-21-/- mice had significant larger infarct sizes after 60 minutes of ischemia and

120 min of reperfusion than their littermate controls (miR-21-/-: 68 ± 9% vs. B6129SF1/J: 53.75 ±

6%). Taken together, these studies show that miR-21 is functional and cardioprotective in myocardial IR-injury.

107 After confirming a cardioprotective role of miR-21 in myocardial IR-injury, we next

investigated miR-21 as a potential downstream target of PER2 in myocardial IR-injury. Previous

studies found light exposure to increase cardiac PER2 and mimic IPC mediated cardioprotection

in a PER2 dependent manner (71). Based on our findings that IPC increased cardiac miR-21 in

a PER2 dependent manner, we next exposed wildtype controls or miR-21-/- mice to 3 hours of

intense light prior to myocardial IR-injury as done previously in PER2-/- mice (71). As shown in

Figure 5.21, light exposure significantly reduced infarct sizes in wildtype controls (intense light

vs. room light: 37.5 ± 6.1% vs. 53.75 ± 6%). However, identical intense light exposure conditions

in miR-21-/- mice failed to induce cardioprotection (room light vs intense light: 68 ± 9% vs. 70.5 ±

8.3%, Figure 5.22). Taken together, these data show that intense light mediated cardioprotection

-/- is abolished in miR-21 mice and suggest that miR-21 is a downstream target of light-elicited

PER2 in cardioprotection from myocardial IR-injury.

The PER2 amplitude enhancer nobiletin is cardioprotective

A recent large-scale screen identified nobiletin, a flavonoid from citrus peels, as a potent

and specific circadian rhythm and PER2 amplitude enhancer (229). Based on these findings we determined whether nobiletin would have a robust cardioprotective effect during myocardial ischemia and reperfusion injury and whether this was a general flavonoid mediated effect (313).

We therefore performed a head to head comparison of compounds similar to nobiletin or basic

flavonoid structured compounds (Figure 5.23) during our murine model of IR-injury. C57BL/6

wildtype mice received nobiletin (1 mg/kg) or equimolar doses of flavone (0.55 mg/kg), tangeritin

(0.93 mg/kg), sinensetin (0.93 mg/kg), 5,6,7-trimethoxyflavone (0.78 mg/kg), or 3',4',7,8-

tetramethoxyflavone (0.85 mg/kg) 2 hours prior to myocardial ischemia via intraperitoneal

injection (Figure 5.24). While most compounds are very similar to nobiletin, only tangeritin was

also found to be a circadian rhythm and PER2 amplitude enhancer (229). As shown in Figure

5.24, only mice who received tangeritine or nobiletin had significantly smaller infarct sizes when

compared to vehicle treated mice (mean ± SD, vehicle: 42.2 ± 9.7 % vs nobiletin: 22.2 ± 7.2% or

108 tangeritin 28.1 ± 8.3%, p<0.05). Pretreatment of wildtype mice with nobiletin 2 hours prior to

myocardial ischemia significantly reduced myocardial infarcts sizes (mean ± SD, vehicle: 42.2 ±

9.7 % vs nobiletin: 22.2 ± 7.2%, p<0.05) or troponin-I levels (mean ± SD: vehicle: 54.1 ± 29.1

ng/ml vs nobiletin: 13.0 ± 6.8 ng/ml, p<0.05; Figure 5.25, 5.26) when compared to vehicle treated controls. Taken together, using nobiletin-like flavonoids during myocardial IR only revealed cardioprotective effects for the circadian rhythm and PER2 enhancers nobiletin and tangeritin.

Nobiletin reverses the deleterious effects of midazolam

After we found that a midazolam mediated downregulation of cardiac PER2 was

associated with increased infarct sizes or troponin-I levels following myocardial IR-injury (Figure

4.1), we next investigated if we could reverse these effects by using the circadian rhythm

enhancer nobiletin (229) (Figure 5.25). Next, we treated C57BL6/J mice with nobiletin (1mg/kg

i.p.) alone or together with midazolam and determined cardiac PER2 mRNA levels 2 hours later.

Nobiletin significantly increased cardiac PER2 transcript levels (3.9-fold, p<0.05) and reversed

midazolam mediated downregulation of cardiac PER2. Following co-administration of midazolam

and nobiletin 2 hours prior to myocardial ischemia, the deleterious effects of midazolam during

myocardial IR-injury were fully reversed (Figure 5.27, 5.28). In fact, co-administration of

midazolam and nobiletin significantly reduced myocardial infarct sizes and Troponin-I serum

levels, when compared to midazolam treated mice (mean ± SD: infarct sizes: midazolam: 59.3 ±

5.2% vs nobiletin + midazolam: 42.15 ± 5.7% or Troponin-I: midazolam: 161.7 ± 57.0 ng/ml vs nobiletin + midazolam: 72.1 ± 22.6 ng/ml, Figure 5.27, 5.28), resulting in infarct sizes or serum

Troponin-I levels similar to vehicle treated mice (Figure 4.3). Based on the increased infarct sizes following midazolam administration, we next investigated if midazolam would also increase the production of reactive oxygen species during reperfusion injury. We therefore determined H2O2 tissue levels in the left ventricle after 60 minutes of ischemia and 15 min or reperfusion. As shown in Figure 5.29, while midazolam significantly increased ROS production when compared to saline treated controls, nobiletin + midazolam treatment resulted in ROS levels like vehicle treated mice.

109 Next, we analyzed the cardioprotective effects of nobiletin alone. Taken together, the PER2

enhancer nobiletin abolishes the deleterious effects of midazolam and is cardioprotective during

myocardial IR-injury.

5.3 Discussion

Here, we present mechanisms of circadian PER2 in cardioprotection from myocardial IR-

injury: 1) we demonstrate our PER2 dependent circadian amplitude enhancement daylight regime

for robust cardioprotection, 2) we identify that light targets metabolic pathways in vivo shifting

cardiac metabolism to more oxygen-efficient pathways, 3) we introduce an additional circadian

entrainment strategy through voluntary wheel running, 4) we reveal an important intersection

between miR-21 and light-elicited PER2 in cardioprotection, and 5) we show that the PER2

targeting flavonoid, nobiletin, is able to reverse the deleterious effects of midazolam on the heart.

Together, our data present mechanisms of light and nobiletin-elicited cardioprotection from

myocardial IR-injury that may have therapeutic benefits in the clinical environment.

Our murine studies provide evidence that daylight-elicited circadian amplitude

enhancement provides robust cardio-protection in a PER2 dependent manner. Additional studies

on altered liver metabolism in constant darkness found adenosine as a possible circulating

circadian factor (73), which suggests adenosine signaling as a mechanism for establishing

circadian rhythmicity between peripheral organs and the SCN. Indeed, the importance of

adenosine signaling via cAMP for PER2 stabilization and cardiac metabolic adaption to ischemia

has been shown in recent studies investigating the mechanism of myocardial IPC (71). In the current studies we found that light-elicited cardiac circadian amplitude enhancement increased cardiac cAMP levels which was also PER2 dependent. In fact, cAMP is not only the key regulator of circadian rhythms (314) but also regulates circadian CD73 activity (315, 316), the key enzyme of extracellular adenosine generation. It is widely appreciated that energy metabolism is under circadian control (113) and metabolic syndrome such as diabetes, obesity, hypertension, endothelial dysfunction, cardiovascular disease, and oxidative stress are linked directly the

110 circadian clock on a cellular level (5). However, specific mechanisms and targeted roles of each

component of the circadian clock are in the infancy of discovery, especially in terms of how

circadian biology mechanistically balances metabolic pathways during hypoxia or ischemia. Our

results contribute to the focused understanding of the convergence of an independent circadian

protein and oxygen-sensing pathway that together targets metabolic adaptation both transcriptionally and post-translationally to optimize cellular adaptation to low oxygen conditions.

Our studies also revealed an important intersection between light-elicited PER2 and miR-

21 in cardioprotection. Our profiling of PER2 dependent microRNAs in cardiac ischemia indicated

a critical role for miR-21 (264) and in vitro studies revealed that PER2 dependent miR-21

regulates cellular glycolysis during cellular stress (71). In fact, myocardial ischemia leads to the

activation of pathways directed towards enhancing myocardial oxygen efficiency. As such,

regulating the balance between metabolic energy metabolism pathways for efficient ATP

production is pivotal to allow the myocardium to function under ischemic conditions (254). Our

studies on myocardial IR-injury found larger infarct sizes and abolished light-elicited PER2

cardioprotection (71) in miR-21-/- mice. In addition, intense light exposure in wildtype mice

increased miR-21 levels where mice were housed under 14h intense light (10,000 LUX) and 10h

dark conditions and produced a cardioprotective affect. In fact, without miR-21, light exposure

was not cardioprotective and targeting miR-21 increases glycolytic reliance and mitochondrial

function at the cellular level. Light exposure has been found to increase PER2 and glycolytic

enzymes and decrease infarct size and troponin levels during MI in mice (71). As such, intense light induction of PER2 regulated cardiac miR-21 is not very surprising. However, intense light therapy in the regulation of microRNAs has not been previously investigated. These findings support our hypothesis, that miR-21 is downstream of PER2 and indicate a critical role for miR-

21 in light or PER2 mediated cardioprotection. Considering light as potential therapy could

represent a novel strategy in the treatment of myocardial ischemia by modulation of

111 cardioprotective microRNAs. Taken together, these studies suggest miR-21 is an important factor

of light-elicited PER2 in cardioprotective from myocardial ischemia.

MiR-21 is predominantly expressed in cardiac fibroblasts when compared with other cell types of the heart (268). In fact, cardiac fibroblasts are considered a key therapeutic target in cardiac remodeling (268). However, during acute myocardial IR, other cells types such as inflammatory cells, myocytes, or endothelial cells, are more important. As such, a recent study on adenosine signaling in IPC of the heart found abolished or dampened cardioprotection by IPC in mice with a tissue specific deletion of the adenosine A2B receptor in cardiomyocytes or endothelia, respectively. Based on these observations we exposed primary fibroblasts, cardiomyocytes, or endothelial cells from C57BL6/J mouse hearts to hypoxia and analyzed miR-21 expression. We found that miR-21 was exclusively upregulated during conditions of low oxygen availability, indicating that endothelial expressed miR-21 is critical during acute myocardial ischemia, which is supported by a recent study elucidating protective effects of miR-21 in endothelial injury (317).

There are some challenges in the therapeutic use of microRNAs as indicated by discrepancies in the published literature. In our studies using miR-21 deficient mice, we found significantly increased larger infarct sizes when compared to controls. In contrast, other studies on miR-21 null mice did not find any significant differences in infarct sizes during myocardial IR- injury (158, 160). While several differences in methodologies might have contributed to the

contrary findings, the most prominent difference was the ischemia time. In our studies, mice were

exposed to 60 minutes of ischemia, while the reported studies used 30 minutes of ischemia.

Indeed, others have found marked differences in cardioprotective mechanisms using different

ischemia times (318). Despite these contrary findings, other studies have shown a protective role

for miR-21 in ischemic preconditioning (161), postconditioning (319), or protection form ischemia

and reperfusion injury of the heart(320), using miR-21 inhibitors or mimetics, supporting our

current findings. Another discrepancy exists in that miR-21 was found to be predominantly

expressed in cardiac fibroblasts when compared with other cell types of the heart (268) and miR-

112 21 inhibition attenuates the fibrotic response and improves cardiac function in mouse models of heart failure (156). However, these results were not reproduced in subsequent studies (157). In

fact, another study on bone marrow-derived mesenchymal stem cells found that overexpressing

miR-21 efficiently repaired myocardial damage in rats (272).

The possibility of targeting specific microRNAs as a therapeutic approach suggests it as a mechanism to regulate components of the core circadian clock timing or output genes. A robust circadian timekeeping system is important for human health and well-being (4, 5, 183, 203-205)

and as we show, circadian entrainment using intense light during the day increases robustness

and circadian amplitude of cardiac PER2, which was found to significantly reduce infarct sizes in

an in-situ mouse model for myocardial IR-injury (71). Administration of microRNAs in a time of the

day dependent manner could therefore help to restore a weakened circadian system, improve

metabolism via the increase of oxygen efficient pathways and thereby promote cardioprotection

from ischemia. As numerous microRNAs have also been found to be protective in the acute

setting during cardiac ischemia or reperfusion, administration of microRNAs during cardiac

revascularization would represent another therapeutic approach. There are several challenges

that need to be addressed for these types of approaches to be implemented. For example, the

timing of microRNA administration and organ-specific therapeutic targeting are crucial areas of

investigation. In fact, one study uncovered a microRNA tissue-specific regulation of gene

expression phase and amplitude (321). These findings suggest that microRNAs function to adapt clock-driven gene expression to tissue-specific requirements (322). Further research is required to determine whether similar cardiac microRNAs exist. In general, practical application of these methods will need a deeper understanding of microRNA biology, the clock, and its role in the regulation of a myriad of gene networks that contribute to human physiology and pathology.

Circadian disruption has been implicated in the development of many diseases, including an increased occurrence of myocardial infarction (4, 5). Based on our findings, circadian disruption could potentially be harmful if myocardial ischemia occurs. This hypothesis is further

113 supported by our findings on circadian rhythm enhancers being cardioprotective in ischemia and

reperfusion injury of the heart. The wide search for circadian rhythm modifying molecules

identified several small molecules including nobiletin, that could alter circadian rhythms (230) and

follow up studies found nobiletin, a flavonoid from citrus peels, not only enhances the circadian

amplitude of PER2 but was also able to protect from a metabolic syndrome in mice (229). Other studies had similar findings showing that increasing the circadian amplitude and the robustness of the circadian cycling via restricted feeding was able to prevent metabolic disease in mice (206).

In fact, a robust circadian timekeeping system has also been found to be important for human health (183, 203, 205). While flavonoids have been implicated in protection from IR-injury in earlier studies (313), it seems striking to us that from the nobiletin-like flavonoids investigated in this study, only tangeritin, which also increases the circadian amplitude (229) was found to be cardioprotective. If those flavonoids, that have been found to protect from IR-injury in other studies, are also circadian amplitude enhancers would require further analyses.

This is the first report on how anesthesia mediated circadian disruption could have functional consequences during ischemia and reperfusion injury of the heart. Furthermore, this is the first report on midazolam, which has been implicated previously in IR-injury, leading to a robust downregulation of cardiac PER2. We therefore are proposing a novel mechanism and a solution to potentially reverse deleterious effects in a clinical setting. In fact, this is the first report that cardiac circadian disruption can be reversed by using the novel circadian rhythm enhancer nobiletin, which points towards clinical and cardioprotective relevance. Performing a comprehensive pharmacological analysis on nobiletin similar compounds further supports the importance of a functional and robust circadian system in the heart. Future studies are warranted to elucidate the specificity of the circadian system for our findings by using genetic models of the circadian rhythm pathways.

114 A

B

Figure 5.1 Optimization of Daylight-Elicited Cardioprotection Daylight-elicited circadian amplitude enhancement is cardio-protective. (A) Wildtype C57BL/6 mice housed under daylight conditions (10,000 LUX, L:D 14:10 h) for 3, 5, or 7 days were subjected to 60 min of in situ myocardial ischemia followed by 2 h reperfusion and compared to mice housed under standard room light (200 LUX, L:D 14:10 h). Infarct sizes were measured by double staining with Evan’s blue and triphenyltetrazolium chloride. Infarct sizes are expressed as the percent of the area at risk that was exposed to myocardial ischemia (mean ± SD; n = 6). (B) Parallel measurements of the myocardial injury marker troponin I by ELISA (mean ± SD; n = 6).

115

Control 3 days

5 days 7 days

Figure 5.2 Infarct Staining of Time-Dependent Daylight-Elicited Cardioprotection Daylight-elicited circadian amplitude enhancement is cardio-protective. Wildtype C57BL/6 mice housed under daylight conditions (10,000 LUX, L:D 14:10 h) for 3, 5, or 7 days were subjected to 60 min of in situ myocardial ischemia followed by 2 h reperfusion and compared to mice housed under standard room light (200 LUX, L:D 14:10 h). Representative images of infarcts (blue indicates retrograde Evan’s blue staining; red and white: area at risk; white: infarcted tissue).

116

Figure 5.3 Light Exposure Induces Cardiac PER2 Amplitude Enhancement Cardiac PER2 luciferase reporter mice were housed under daylight (10,000 LUX, L:D 14:10 h) or standard room light (200 LUX, L:D 14:10 h) for 7 days. Mouse hearts were harvested over a 48- h time at indicated time intervals for luciferase reporter activity assays (mean ± SD; n = 4).

117

Figure 5.4 Sex-Specific Differences in Male Versus Female Light-Elicited Cardioprotection Wildtype C57BL/6 female mice were housed under daylight (10,000 LUX, L:D 14:10 h) or standard room light (200 LUX, L:D 14:10 h) for 7 days and subjected to 60 min of in situ myocardial ischemia followed by 2 h reperfusion. Infarct sizes were measured by double staining with Evan’s blue and triphenyltetrazolium chloride. Infarct sizes are expressed as the percent of the area at risk that was exposed to myocardial ischemia (mean ± SD; n = 6).

118 A

B

C

Figure 5.5 Enucleated Mice Do Not Exhibit an Increase in Cardiac PER2 Protein Wildtype or enucleated (blind) mice housed under room light or daylight (14:10 L:D) for 7 days. After 7 days hearts were harvested for protein isolation and immunoblot for PER2 (mean ± SD; n = 5). Quantification is represented as fold change in PER2 protein levels.

119 A

B

Figure 5.6 Enucleated Mice Do Not Achieve Light-Elicited Cardioprotection Wildtype or enucleated (blind) mice housed under room light. (A) Cardiac troponin levels measured from mice able to perceive light and enucleated mice after 60 min ischemia and 2 h reperfusion at 9 AM or 9 PM. (B) Cardiac troponin measurements from wildtype mice compared to enucleated mice subjected to 60 min ischemia and 2 h reperfusion (mean ± SD, n =7).

120

Figure 5.7 Hearts of Enucleated Mice Exhibit Significantly Larger Infarct Sizes Representative infarct staining from blind and seeing mice at 9PM (blue indicates retrograde Evan’s blue and triphenyltetrazolium chloride staining; red and white: area at risk; white: infarcted tissue).

121

Figure 5.8 Actigraphy Graphs of Wheel Running as a Circadian Entrainment Strategy Activity graphs from wildtype mice exposed to room light versus daylight. Each row represents the activity of one mouse. Gray backgrounds indicate dark phases and white backgrounds are light phases. Black lines indicate wheel running activity.

122 A

B

Figure 5.9 Wheel Running Measurements as an Entrainment Strategy Wheel running measurements from Figure 5.8 during room light (200 LUX, L:D 14:10 h) or daylight (10,000 LUX, L:D 14:10 h) housing conditions (mean ± SD, n=5-6).

123

Figure 5.10 Actigraphy Graphs During Light-Elicited Entrainment Strategy Activity graphs from WT mice exposed to room light versus daylight (n=5-6).

124 A B

C

Figure 5.11 Voluntary Wheel Running for Amplitude Enhancement is Cardioprotective Circadian amplitude and infarct sizes after 60 min of in situ myocardial ischemia followed by 2 h reperfusion in wildtype C57BL/6 mice exposed to voluntary wheel running for 1 versus 2 weeks (mean ± SD, n = 5-6).

125

Figure 5.12 Light-Elicited Effect on 13C Fructose 1,6-bisP in the Left Ventricle C57BL/6 wildtype mice were housed under daylight (10,000 LUX, L:D 14:10 h) or standard room light (200 LUX, L:D 14:10 h) for 7 days followed by infusion of 10 mg/kg/min U-13C-glucose via an intra-arterial catheter over 90 minutes. Left ventricles of the hearts were flushed with ice cold KCl and the left ventricle was shock frozen and analyzed by liquid chromatography–tandem mass spectrometry (n = 3).

126 A

B

Figure 5.13 Light-Elicited and PER2 Dependent PFK Activity in Plasma and Tissue Phosphofructokinase (PFK) activity in both heart tissue and plasma samples from wildtype C57BL/6 or PER2-/- mice after 7 d of daylight (10,000 LUX, L:D 14:10 h) or standard room light (200 LUX, L:D 14:10 h) housing.

127 A

B

Figure 5.14 PER2 is Required for Light-Elicited Cardioprotection Wildtype C57BL/6 or Per2-/- mice housed under room- or daylight for 7 d prior to 60 min myocardial ischemia and 2 h reperfusion. (A) Infarct sizes are expressed as the percent of the area at risk that was exposed to myocardial ischemia (mean ± SD; n = 6). (B) Representative infarct staining for Per2-/- mice after daylight housing is shown.

128

Figure 5.15 cAMP as a Potential Mechanism in PER2 Dependent Cardioprotection cAMP measured by ELISA in heart tissue from wildtype C57BL/6 or PER2-/- mice after 7 days of room light or daylight housing (14:10 L:D) (mean ± SD; n = 6).

129 A B

C D

Figure 5.16 Light-Sensing HMEC-1 Reveal cAMP as a Light-Elicited Pathway pCMV6 is the empty vector control and OPN4 is the plasmid containing the gene encoding melanopsin (n = 3). (A) Study design and verification of melanopsin overexpression by immunoblot for the DDK tag. (B) PER2 transcript levels after light sensing cells were exposed to daylight. (C, D) cAMP and pCREB levels after light-sensing cells were exposed to daylight (mean ± SD, n = 6).

130 A

B

Figure 5.17 Light-Sensing HMEC-1 Exhibit an Increase in Glycolytic Reliance pCMV6 is the empty vector control and OPN4 is the plasmid containing the gene encoding melanopsin. (A) Light-sensing HMEC-1 (red line) and empty vector control HMEC-1 (black line) were exposed to daylight prior to glycolytic stress test on the Seahorse Bioanalyzer where extracellular acidification rate (ECAR) was measured as a readout of glycolytic function upon sequential addition of glucose, oligomycin, and 2-DG. (B) Glycolytic capacity is represented from the glycolytic stress test where the ‘light-sensing’ (melanopsin-expressing) HMEC-1 are represented by the black bar and the empty vector control cells represented by the white bar (mean ± SD, n = 10).

131 A

B

Figure 5.18 Light-Sensing HMEC-1 Exhibit an Increase in Mitochondrial Functioning pCMV6 is the empty vector control and OPN4 is the plasmid containing the gene encoding melanopsin. (A) Light-sensing HMEC-1 (red line) and empty vector control HMEC-1 (black line) were exposed to daylight prior to mitochondrial stress test on the Seahorse Bioanalyzer where oxygen consumption rate (OCR) was measured as a readout of mitochondrial function upon sequential addition of oligomycin, FCCP, and rotenone/antimycin A. (B) Maximum achievable respiration is represented from the mitochondrial stress test where the ‘light-sensing’ (melanopsin-expressing) HMEC-1 are represented by the black bar and the empty vector control cells represented by the white bar (mean ± SD, n = 10).

132 A

B

Figure 5.19 Daylight Increases miR-21 and PER2 in Heart Tissue Wildtype mice were exposed to broad spectrum intense light (10,000 lux) for 7 days (LD 14:10) and compared to controls that were maintained at room light (200 lux, LD 14:10). miR-21 or Per2 transcript levels were determined by qRT-PCR (mean ± SD, n = 3, p<0.05).

133

Figure 5.20 Infarct Sizes are Larger in miR-21-/- Mice Wildtype or miR-21-/- Mice underwent 60 min of ischemia and 120 min of reperfusion at room light (200LUX) or after exposure to 3 hours of intense light (10,000 LUX). Infarct sizes were measured by double staining with Evan’s blue and triphenyl-tetrazolium chloride. Infarct sizes are expressed as the percent of the area at risk (AAR) that underwent infarction. Infarct sizes in wildtype or miR- 21-/- mice at room light conditions (mean ± SD, n = 4, p<0.05).

134 A

B

Figure 5.21 Light Decreases Infarct Size After Myocardial Ischemia and Reperfusion Injury Wildtype mice underwent 60 min of ischemia and 120 min of reperfusion at room light (200LUX) or after exposure to 3 hours of intense light (10,000 LUX). Infarct sizes were measured by double staining with Evan’s blue and triphenyl-tetrazolium chloride. Infarct sizes are expressed as the percent of the area at risk (AAR) that underwent infarction. (A) Infarct sizes in wildtype mice after exposure to intense light for 3 h compared to room light conditions (mean ± SD, n = 4, p<0.05). (B) Representative infarct staining in hearts from wildtype mice exposed to intense light or room light prior to in situ myocardial ischemia and reperfusion (blue, retrograde Evan’s blue staining; red and white, area at risk; white, infarcted tissue).

135 A

B

Figure 5.22 MiR-21 is Necessary for Light-Elicited Cardioprotection miR-21-/- Mice underwent 60 min of ischemia and 120 min of reperfusion at room light (200LUX) or after exposure to 3 hours of intense light (10,000 LUX). Infarct sizes were measured by double staining with Evan’s blue and triphenyl-tetrazolium chloride. Infarct sizes are expressed as the percent of the area at risk (AAR) that underwent infarction. (A) Infarct sizes in miR-21-/- mice exposed to intense light or room light prior to in situ myocardial ischemia followed by reperfusion (mean ± SD, n = 4, not significant). (B) Representative infarct staining in hearts from miR-21-/- mice exposed to intense light or room light prior to in situ myocardial ischemia reperfusion (blue, retrograde Evan’s blue staining; red and white, area at risk; white, infarcted tissue).

136

Figure 5.23 Differences in Structures of Flavonoids Used in Ischemia Reperfusion Injury C57BL/6 wildtype mice received nobiletin (1mg/kg) or equimolar doses of flavone (0.55 mg/kg), tangeritin (0.93 mg/kg), sinensetin (0.93 mg/kg), 5,6,7-trimethoxyflavone (0.78 mg/kg), or 3',4',7,8-tetramethoxyflavone (0.85 mg/kg) 2 hours prior to myocardial ischemia via intraperitoneal injection.

137

Figure 5.24 The Effect of Flavonoids on Infarct Size Wildtype mice underwent 60 min of ischemia and 120 min of reperfusion. Infarct sizes were measured by double staining with Evan’s blue and triphenyl-tetrazolium chloride. Infarct sizes are expressed as the percent of the area at risk (AAR) that underwent infarction. C57BL/6 wildtype mice received nobiletin (1mg/kg) equimolar doses of flavone (0.55 mg/kg), tangeritin (0.93 mg/kg), sinensetin (0.93 mg/kg), 5,6,7-trimethoxyflavone (0.78 mg/kg), or 3',4',7,8-tetramethoxyflavone (0.85 mg/kg) 2 hours prior to myocardial ischemia via intraperitoneal injection (mean ± SD; n = 5- 8; p<0.05).

138 A

B

Figure 5.25 Nobiletin Reduces Troponin-I Levels After Ischemia and Reperfusion Injury Wildtype mice were treated with vehicle (solutol) or nobiletin (1mg/kg, A) 2 hours prior to myocardial ischemia and reperfusion injury. Myocardial ischemia consisted of 60 min of ischemia followed by 120 minutes of reperfusion. (B) Serum troponin I concentrations from mice administered solutol or nobiletin prior to IR-injury (mean ± SD; n = 7-8; p<0.05; NOB=nobiletin).

139 A

B

Figure 5.26 Nobiletin Decreases Infarct Size After Myocardial IR-Injury Wildtype mice were treated with vehicle (solutol) or nobiletin (1mg/kg) 2 hours prior to myocardial ischemia and reperfusion injury. Myocardial ischemia consisted of 60 min of ischemia followed by 120 minutes of reperfusion. (A) Infarct sizes were measured by double staining with Evan’s blue and triphenyl-tetrazolium chloride. Infarct sizes are expressed as the percent of the area at risk (AAR) that underwent infarction (mean ± SD; n = 7-8; p<0.05). (B) Representative infarct staining (blue, retrograde Evan’s blue staining; red and white, area at risk; white, infarcted tissue).

140 A

B

Figure 5.27 Nobiletin Rescues the Deleterious Effect of Midazolam on Cardiac PER2 Gene Expression and Troponin-I Levels After Myocardial IR-Injury Mice were treated with vehicle (pentobarbital), nobiletin (1mg/kg), midazolam (200mg/kg), or midazolam and nobiletin. Two hours later mice underwent myocardial ischemia and reperfusion injury. Myocardial ischemia consisted of 60 min of ischemia followed by 120 minutes of reperfusion. (A) Mouse cardiac PER2 mRNA levels two hours after exposure to vehicle, nobiletin or midazolam and nobiletin (mean ± SD, n = 5-7; p<0.05; NOB=nobiletin). (B) Serum troponin I concentrations from mice administered midazolam or midazolam and nobiletin prior to myocardial IR-injury (mean ± SD, n = 5-7; p<0.05; NOB=nobiletin).

141 A

B

Figure 5.28 Nobiletin Rescues the Deleterious Effect of Midazolam on the Infarct Size After Myocardial IR-Injury Wildtype mice were treated with vehicle (pentobarbital), nobiletin (1mg/kg), midazolam (200mg/kg), or midazolam and nobiletin 2 hours prior to myocardial ischemia and reperfusion injury. Myocardial ischemia consisted of 60 min of ischemia followed by 120 minutes of reperfusion. Infarct sizes were measured by double staining with Evan’s blue and triphenyl- tetrazolium chloride. Infarct sizes are expressed as the percent of the area at risk (AAR) that underwent infarction. (A) Infarct sizes as the percent of AAR. (B) Representative infarct staining (blue, retrograde Evan’s blue staining; red and white, area at risk; white, infarcted tissue).

142

Figure 5.29 Nobiletin Reverses Midazolam Induced ROS During Myocardial IR-Injury Wildtype mice were treated with vehicle (pentobarbital), midazolam (200mg/kg), or midazolam and nobiletin 2 hours prior to myocardial ischemia and reperfusion injury. Myocardial ischemia consisted of 60 min of ischemia followed by 15 minutes of reperfusion for ROS (H2O2) measurements. ROS production determined by H2O2 measurement in the area at risk.

143 CHAPTER 6 CHAPTER VI LIGHT-ELICITED CIRCADIAN AMPLITUE ENHANCEMENT IN HUMANS6

6.1 Rationale

The dire consequences of circadian disruption on human disease (11-13) (14, 15) (16-22)

(6, 23-27) (28) (29) (30) (31, 32) (33-37) provides an avenue to target circadian amplitude enhancement or circadian entrainment as a novel therapeutic or preventative treatment strategy.

Through our in vitro and in vivo studies, we provide evidence that daylight-elicited circadian amplitude enhancement provides robust cardioprotection from myocardial ischemia in a PER2 dependent manner. Therefore, we next applied a daylight regime to healthy human subjects to determine if this is sufficient to induce PER2, promote metabolic balance, and enhance circadian amplitude, as first steps in establishing a potential cardioprotective protocol in humans.

6.2 Results

Daylight-elicited circadian amplitude enhancement in healthy human subjects

We exposed human healthy volunteers to 30 min of daylight (10,000 LUX) in the morning on 5 consecutive days and performed serial blood draws. Daylight therapy increased PER2 protein levels in human buccal or plasma samples in the morning and evening, indicating enhancement of the circadian amplitude (Figure 6.1, 6.2). To test the efficacy of daylight light therapy on the circadian system (10) we determined melatonin plasma levels, which were significantly suppressed upon light therapy (Figure 6.3). In addition, room light was less efficient in suppressing melatonin than daylight therapy (Figure 6.3). Further analysis revealed that daylight therapy increased plasma phosphofructokinase (PFK, Figure 6.4) or lactate dehydrogenase (LDH, Figure 6.5) activity in conjunction with increased plasma HIF1⍺ levels

6 Portions of this chapter are from either our publication (276) or from our manuscript currently in review. Recruiting, consenting, and enrolling healthy human volunteers into the clinical trial was done by Colleen Bartman. Blood sample processing was done by Colleen Bartman and Yoshimasa Oyama. Human plasma or buccal sample immunoblots were done by a former lab technician. Melatonin, PFK, and HIF1⍺ analysis was done by Yoshimasa Oyama. Targeted metabolomics mass spectrometry was done by the Metabolomics Core at the University of Colorado Anschutz Medical Campus.

144 (Figure 6.6). Moreover, plasma triglycerides, a surrogate for insulin sensitivity and carbohydrate metabolism (323), significantly decreased upon light therapy, indicating increased insulin

sensitivity and glucose metabolism (Figure 6.7), while plasma glucose levels remained unchanged (Figure 6.8). Targeted metabolomics from human plasma samples confirmed a strong effect of light therapy on metabolic pathways such as glycolysis or TCA cycle function (Figure

6.9). Moreover, we found significant decreases of pyruvate, lactate and succinate levels after 5 days of light therapy (Figure 6.10). Our targeted metabolomics from human plasma samples also identified amino acid differences upon light exposure (Figure 6.11 – 6.13). Together with increased enzyme activities of plasma phosphofructokinase and lactate dehydrogenase this finding indicates improved metabolic flux, possibly due to increased glycolysis and improved TCA cycle or mitochondrial function.

As sleep deprivation is directly associated with decreased insulin sensitivity and compromised glucose metabolism (183), we next determined how our light therapy protocol would impact human physiology in terms of sleep behavior. Using a validated accelerometer for actigraphy (324) (Actiwatch 2) we found decreased sleep onset latency, less awakenings, less

WASO (wake after sleep onset) episodes and overall improved sleep efficiency (Figure 6.14). In addition, we found this light exposure to reasonably entrain our healthy human subjects (Figure

6.15). Taken together, our data indicate that daylight therapy, a mechanism of circadian amplitude enhancement, targets similar PER2-HIF1⍺ dependent metabolic pathways in humans as seen in mice and may present a promising novel strategy in the treatment or prevention of low oxygen conditions such as myocardial ischemia.

Intense light exposure increases miR-21 and PFK in healthy human subjects

After we found intense light regulation of cardiac miR-21, we next pursued studies on light

therapy in healthy human volunteers. In fact, earlier studies found increased PER2 levels in

buccal swaps from human volunteers upon light treatment (325). Based on strategies using bright

light therapy to treat seasonal mood disorders or depression in humans (326), we adapted a

145 similar protocol. Thus, we exposed eight healthy volunteers (3 females, 5 males) to 30 min of

intense light therapy from 8:30 until 9:00 AM for 5 days. Blood was drawn on day one at 8:30 AM before any intense light exposure and on day 5 at 9:00 AM after intense light exposure (Figure

6.16). Plasma samples were used to isolate microRNAs and to determine miR-21 plasma levels.

Five days of intense light therapy significantly increased miR-21 plasma levels in human subjects

(3.5-fold, Figure 6.17). Based on findings that miR-21 overexpression was associated with increased glycolysis in vitro, we next determined plasma phosphofructokinase activity, the key regulatory enzyme in the glycolytic pathway. Here, intense light exposure led to a 49% increase of PFK activity (Figure 6.17). Taken together, one week of intense light exposure in human subjects increases miR-21 levels in blood plasma samples which is associated with increased phosphofructokinase activity.

6.3 Discussion

Our translational studies demonstrated that our daylight protocol promotes circadian entrainment in humans, increases plasma PER2 and miR-21, induces glycolytic enzyme expression, and targets metabolic pathways like glycolysis and the TCA cycle. Humans are primarily entrained by sunlight, regardless of other external cues (6, 327). The greater the intensity of light, the more robust the entrainment and the amplitude of the circadian rhythm. This has been shown by studies on melatonin suppression, indicating that humans might need brighter or more intense light for optimal entrainment of circadian rhythms (10). The effects of intense light therapy in humans are recognized and already widely used. As such, intense light therapy is used to treat seasonal affective disorder (328, 329), but also might have effects on preventing delirium (330) or might improve sleep in general (331, 332). However, to our knowledge nobody has analyzed human metabolic changes upon intense light therapy yet.

Intense light miR-21 increase in human plasma samples was associated with increased phosphofructokinase activity, the key enzyme of glycolysis. These findings indicate that our in vitro and murine in vivo findings are translatable into a human system. While intense light

146 protocols in the regulation of microRNAs is a relatively recent phenomenon, light induced

cardioprotective miR-21 could be one mechanism by which intense light exposure reduced

myocardial damage in murine studies (71) and studies on cardioprotective effects of light exposure in patients are currently undergoing. Furthermore, our targeted metabolomics analysis suggests metabolic changes that may be advantageous for adaptation to low oxygen availability.

Supporting the importance of circadian rhythms in myocardial susceptibility to ischemia, recent studies found a diurnal pattern for troponin values in patients undergoing aortic valve replacement (193). Here, troponin values following surgery were significantly higher in the morning when compared to the afternoon. While nothing can be done about a diurnal pattern, applying light therapy before high risk non-cardiac or cardiac surgery to enhance the circadian amplitude, however, might be able to provide robust cardio-protection and reduce mortality

throughout the entirety of the 24-hour period. In fact, elevated troponin levels after non-cardiac

surgery have been associated with a significantly elevated 30-day mortality rate (333).

Light-elicited circadian amplitude enhancement suggests an overall increase in PER2

levels even at the trough of the amplitude, indicating that this strategy could promote 24-hour

advanced cardioprotection and potentially decrease troponin levels in both the morning and

evening times. While we show that circadian amplitude enhancement is a reasonable strategy for

adaptive protection from myocardial ischemia, its potential to be applicable to protect from other

hypoxic injuries, however, will need further investigation in mechanistic research endeavors and

clinical trials.

147 A B

C

Figure 6.1 Light Exposure Increases Buccal and Plasma PER2 in Humans (A) Healthy human volunteer subjects were exposed to intense light (10,000 LUX) for five consecutive days from 8:30 – 9:00 AM. A total of 20 healthy volunteers were enrolled (11 female and 6 male, age range between 21-44 years old). (B) PER2 protein levels from plasma of healthy human subjects obtained at 8:30 or 9:00 AM during 5 days of intense light exposure assessed by ELISA (mean ± SD; n = 6). (C) PER2 protein levels from healthy human volunteer buccal samples obtained before or after intense light exposure over 5 days and assessed by immunoblot for PER2 or the loading control, β-ACTIN. One representative blot of three is shown.

148 A

B

Figure 6.2 Light Exposure in Humans Increases Plasma PER2 Protein Levels Immunoblots of plasma PER2 from healthy human volunteers exposed to 5 days of the daylight protocol.

149 A

B

Figure 6.3 Daylight Exposure Suppresses Plasma Melatonin More Than Room Light (A) Longitudinal monitoring of human plasma melatonin levels during 5 days of daylight exposure (C: control, 8:30 AM on day 1 and all other samples from 9 AM; mean ± SD; n = 3-6). (B) Effect of room light versus daylight on human plasma melatonin levels (mean ± SD; n = 3-6).

150 A

B

Figure 6.4 Light Exposure in Humans Targets Plasma PFK Activity (A) Human plasma phosphofructokinase (PFK) activity during 5 days of intense light exposure (C: control, sample from 8:30 AM on day 1 and all other samples from 9 AM, mean ± SD; n = 3- 6). (B) Human plasma PFK activity at 9 PM after 5 days of intense light exposure (mean ± SD; n = 3).

151

Figure 6.5 Light Exposure in Humans Increases LDH Activity Human plasma lactate dehydrogenase (LDH) activity during 5 d of intense light exposure (C: control, sample from 8:30 AM on day 1 and all other samples from 9 AM, mean ± SD; n = 8).

152

Figure 6.6 HIF1⍺ is Stabilized Upon Light Exposure in Humans HIF1⍺ human plasma levels during 5 d of intense light exposure (C: control, sample from 8:30 AM on day 1 and all other samples from 9 AM, mean ± SD; n = 6).

153

Figure 6.7 Light Exposure Decreases Triglycerides Human plasma triglyceride levels during 5 d of intense light exposure (C: control, sample from 8:30 AM on day 1 and all other samples from 9 AM, mean ± SD; n = 8).

154

Figure 6.8 Light Exposure Does Not Alter Plasma Glucose Levels Plasma glucose levels from human volunteers during 5 days of intense light exposure (C: control, sample from 8:30 AM on day 1 and all other samples from 9 AM, mean ± SD; n = 8).

155

Figure 6.9 Targeted Metabolomics from Humans Exposed to Light Targeted metabolomics using mass-spectrometry on human plasma samples from healthy volunteers exposed to the intense light exposure regime for 5 days.

156 A B

C D

Figure 6.10 Key Glycolytic and TCA Cycle Metabolites from Targeted Metabolomics Targeted metabolomics using mass-spectrometry on human plasma samples from healthy volunteers exposed to the intense light exposure regime for 5 days. (A-C) Key metabolites of glycolysis (pyruvate, lactate, and glucose) and (D) the TCA cycle (succinate) are shown for day 3 and day 5 of daylight therapy (mean ± SD; n = 3).

157 A B

C D

Figure 6.11 Targeted Metabolomics Targeted metabolomics using mass-spectrometry on human plasma samples from healthy volunteers exposed to the intense light exposure protocol for 5 days identified key amino acids altered upon light exposure (mean ± SD; n = 3).

158 A B

C D

Figure 6.12 Targeted Metabolomics Targeted metabolomics using mass-spectrometry on human plasma samples from healthy volunteers exposed to the intense light exposure protocol for 5 days identified additional altered metabolites upon light exposure (mean ± SD; n = 3).

159

Figure 6.13 Targeted Metabolomics Targeted metabolomics using mass-spectrometry on human plasma samples from healthy volunteers exposed to the intense light exposure protocol for 5 days (mean ± SD; n = 3).

160 A B

C D

Figure 6.14 The Effect of Light Exposure on Sleep Actigraphy data using a validated accelerometer (Actiwatch 2). Shown are the sleep onset latency (SOL, A), which is the amount of time it takes to fall asleep after the lights have been turned off, the number of awakenings during the sleep (B), the wake-up episodes after the sleep onset (WASO, C) and the sleep efficiency (D) (mean ± SD; n = 6).

161

Figure 6.15 Light Exposure Entrains Human Subjects Actigraphy data using a validated accelerometer (Actiwatch 2) from human volunteers during 5 days without and 5 days with the intense light exposure regime.

162

Figure 6.16 Light Exposure Blood Collection Timeline for miR-21 Studies 7 Healthy human volunteers were exposed to 30 minutes of intense blue light (Square One Wake Up Light, NatureBright, Day-Light 10,000 Lux) in the morning at 8:30 AM on 5 consecutive days. A blood draw was performed before light exposure on the first day (8:30 AM) and 5 days after light exposure (9.00 AM).

163 A

B

Figure 6.17 Effects of Intense Light on MiR-21 Regulation in Human Subjects 7 Healthy human volunteers were exposed to 30 minutes of intense blue light (Square One Wake Up Light, NatureBright, Day-Light 10,000 Lux) in the morning at 8:30 AM on 5 consecutive days. A blood draw was performed before light exposure on the first day (8:30 AM) and 5 days after light exposure (9.00 AM). Plasma samples were analyzed for miR-21 levels and PFK (phosphofructokinase) activity (mean ± SD, n = 7, p<0.05).

164 CHAPTER 7 CHAPTER VII CCTSI PRE-DOCTORAL FELLOWSHIP EXPERIENCE

7.1 The Fellowship

During my first year in Dr. Eckle’s lab, I was received a CCTSI TL1 Pre-Doctoral

Fellowship and the subsequent two years thereafter receiving an American Heart Association

(AHA) Pre-Doctoral Fellowship. The CCTSI fellowship was mentored in conjunction with Dr. Peter

Buttrick, Head of the Division of Cardiology at the University of Colorado Anschutz Medical

Campus. While my AHA fellowship focused on in depth mechanisms of light-elicited PER2 and cardioprotective mechanisms, the CCTSI fellowship allowed me the opportunity to expand my research into the clinical and translational realm of research. This experience taught me how to amend a COMIRB protocol, apply fundamental benchwork to a clinical setting, screen and consent patients for a clinical trial, and perform translational research studies, overall enhancing my understanding of my research endeavors and the importance of this research for improving human health.

My CCTSI fellowship, ‘Intense Light as a Novel Treatment in Myocardial Ischemia’ proposed two aims. First, I proposed to generate a novel light-sensing cell line to understand mechanisms of light-elicited metabolic adaptation to low oxygen conditions. The results from these studies were demonstrated in Figure 5.16 – 5.18. Second, I proposed to define the role of light therapy in human populations by drawing blood for circadian protein and metabolic analysis before and after applying a daylight exposure regime. Data from these studies were presented in

Figure 6.1 – 6.17 after obtaining approval from the Institutional Review Board (COMIRB #13-

1607) and prior written informed consent from each individual.

To satisfy the course requirements of the TL1 certificate, I completed the following courses: Tissue Biology and Disease Mechanisms (IDPT 7646), Obesity and Cardiovascular

Disease (IDPT 6006), Biostatistics (BIOS 6601), Epidemiology (EPID 6630), Responsible

Conduct of Research, COMIRB training (CITI), and HIPAA training and certification. In addition, I

165 attended the Association for Clinical and Translational Science (ACTS) National Conference in

Washington, D.C., where I presented my research poster titled ‘Intense Light as a Novel

Treatment in Myocardial Ischemia.’

7.2 Clinical Relevance

Heart disease is the leading cause of death worldwide and therefore demands innovative

research initiatives to develop novel therapeutic strategies for prevention and treatment. In the

United States, cardiovascular disease causes approximately one death every 40 seconds (65)

and 20% of heart attacks are silent and due to coronary artery disease (334). By 2030,

epidemiologists project 43.9% of adults in the United States will have some sort of CVD (334).

About 790,000 Americans have a heart attack every year with 580,000 being a first-time incident

and 210,000 being a repeat (334). Most myocardial infarctions occur after patients were admitted

to the hospital for another reason (334), indicating infarction was not the primary cause of

hospitalization and supporting the idea that circadian disruption is detrimental to the heart. In

addition, silent myocardial infarctions make up almost 50% of incidents. Therefore, heart disease

is a prominent focus for researchers to identify novel therapeutic targets to reduce the damage caused by myocardial IR-injury (65).

Onset of myocardial infarctions are circadian in nature (24, 25, 184, 185). Notably, the preceding components in the development of cardiovascular disease are also intertwined with the circadian system. For example, shift-workers (those who during the night, cover multiple shifts, rotate shifts, or work outside of normal working hours), have 40% increased risk of cardiovascular disease and higher risk of MI compared to day workers (179, 335). This may be due to the fact that shift-work leads to dysregulation of the circadian system, leading to increased risk of hypertension, dyslipidemia, obesity, and diabetes, which are all factors of cardiovascular disease

(179).

166 7.3 Experimental Design and Screening Criteria

For patients diagnosed with myocardial infarction, the timeline for consenting patients, drawing blood, and light exposure is shown in Figure 7.1. At the time of my dissertation defense, we are in the process of collecting samples from patients with MI for analysis. The data we present are from our healthy human volunteers. The timeline for blood draws and light exposure is shown in Figure 7.2. The consent form is enclosed in Appendix B.

167

Figure 7.1 Timeline of Enrolling and Obtaining Sample from Clinical Trial Patients The plan for enrolling patients into the clinical trial (patients are being actively recruited, to date): patients admitted to the cardiac ICU are screened for diagnosis of myocardial infarction (either NSTEMI or STEMI) with elevated troponin levels. Patients are consented and enrolled following COMIRB guidelines and according to the clinical trial #13-1607 protocol. On day 1, blood is drawn at 8:30 AM and patients given a light box to use for 30 minutes between 8:30 AM and 9:00 AM (or no light box for those in the control group). Patients return either 1 or 2 weeks later for a final 9:00 AM blood draw. To date, 5 patients diagnosed with an MI have been enrolled and blood samples obtained. Analysis will be done once sufficient patient samples have been obtained.

168

Figure 7.2 Timeline of Blood Draws from Healthy Human Volunteers Timeline for healthy human volunteer subjects in the clinical trial is as follows: healthy human volunteers were consented following COMIRB and clinical trial #13-1607 guidelines and protocols. Subjects had blood drawn at 8:30 AM on day 1 and received 30 minutes of intense light exposure from 8:30 to 9:00 AM for 5 consecutive days. Blood draws were done on day 1, day 3, and day 5 at 9:00 AM (after light exposures). Blood samples were processed for plasma and stored for analyses.

169 CHAPTER 8 CHAPTER VIII METHODS

8.1 Cell Culture

Cell culture and treatments

Human microvascular endothelial cells (HMEC-1) were cultured as described previously

(336). All experiments were conducted after serum starvation to reset circadian rhythmicity (71).

For hypoxia experiments cells were placed in a hypoxia chamber (Coy Laboratory Products Inc.,

Grass Lake, MI) in preequilibrated hypoxic medium at 1% O2.

Lentiviral-mediated generation of cells with knockdown of PER2 or HIF1⍺

Stable cell cultures with decreased PER2 and HIF1⍺ expression were generated by

lentiviral-mediated shRNA expression. pLKO.1 lentiviral vectors targeting PER2 had shRNA

sequence of CCG GGA CAC ACA CAA AGA ACT GAT ACT CGA GTA TCA GTT CTT TGT GTG

TGT CTT TTT (TRCN0000018542) and HIF1A had shRNA sequence of CCG GCC AGT TAT

GAT TGT GAA GTT ACT CGA GTA ACT TCA CAA TCA TAA CTG GTT TTT (TRCN

0000003809). For controls, nontargeting control shRNA (SHC002; Sigma) was used. HMEC-1

were co-transfected with pLK0.1 vectors and packaging plasmids to produce lentivirus. Filtered

supernatants were used for infection of HMEC-1 and cells were selected with puromycin or

geneticin until a knockdown was confirmed. PER2 KD and scrambled control cell lines (Scr) were

selected using 2.5 µg/mL puromycin. HIF1⍺ KD and control cell lines were selected using 350

µg/mL geneticin (G418).

Endothelial cells

C57BL/6 mouse primary cardiac endothelial cells were obtained from Cell Biologics (C57-

6024) and handled following manufacturer’s instructions in complete mouse endothelial cell

medium supplemented with VEGF, ECGS, heparin, EGF, hydrocortisone, L-glutamine, antibiotic- antimycotic solution, and FBS (M1168). After cells reached confluency, cells were exposed to

170 normoxia (21% oxygen) or hypoxia (1% oxygen using preequilibrated media for 3 h (337)) and immediately resuspended in Trizol for miRNA analysis.

MiR-21 gain or loss of function

For gain of function experiments, we used a MISSION hsa-miR-21 Mimic (Sigma-Aldrich, cat. no. HMI0372). The miR-21 Mimic was delivered to human microvascular endothelial cells

(HMEC-1) using DharmaFect I Transfection Reagent (Dharmacon). For loss of function experiments, we used an anti-miR-21 (Qiagen, MIMAT0000076: 5'UAG CUU AUC AGA CUG

AUG UUG A, MIMAT0004494: 5'CAA CAC CAG UCG AUG GGC UGU, MIMAT0004494: 5'CAA

CAC CAG UCG AUG GGC UGU; MIMAT0000076: 5'UAG CUU AUC AGA CUG AUG UUG A) and a miScript Inhibitor Negative Control (Qiagen, cat. no. 1027271). The anti-miR-21 was delivered to HMEC-1 using HiPerFect Transfection Reagent (Qiagen). Cells were seeded at a density of 30,000 cells/well prior to transfection.

Light sensing cells

HMEC-1 WT cells were transfected with pCMV6-Entry (C-terminal Myc and DDK Tagged,

OriGene Technologies, Rockville, MD) or OPN4 (Myc-DDK-Tagged)-pCMV6-Entry transcript variant 1 (TrueORF Gold Expression-Validated cDNA Clones from OriGene Technologies,

Rockville, MD) using FuGene HD Transfection Reagent (Promega, Madison, WI). After transfection, cells were kept in complete darkness until room light (~200 LUX) or daylight (~10,000

LUX) exposures for 30 minutes. Melanopsin protein expression validation was done by isolating protein using RIPA buffer (Thermo Fisher Scientific, Waltham, MA) supplemented with protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA). Immunoblotting for anti-DDK (OriGene Technologies, Rockville, MD) was used to detect DDK-tagged melanopsin in transfected cells. Light-sensing cells subjected to glycolytic or mitochondrial stress tests on the

Seahorse Bioanalyzer were exposed to light 30 min prior to Seahorse analyses.

171 8.2 Transcriptional Analysis

Total RNA Isolation

Total RNA was isolated from HMEC-1, murine primary cardiac myocytes, primary cardiac

endothelia, or whole heart using Trizol Reagent (Invitrogen, Carlsbad, CA) or Qiazol Reagent

(Qiagen), phenol-chloroform extraction, and ethanol precipitation in conjunction with the RNeasy

Mini Kit (Qiagen, Germantown, MD). MicroRNAs were isolated by a secondary ethanol

precipitation (100%) of eluate from initial lysate centrifugation through the mini column. Total RNA

and microRNA elutions were quantified using either a Qubit 3.0 RNA BR Assay Kit (Thermo Fisher

Scientific, Waltham, MA) or Nanodrop 2000.

Messenger RNA Analysis

Quantification of transcript levels was determined by real-time RT-PCR (iCycler; Bio-Rad

Laboratories, Inc, Hercules, CA). qPCR reactions contained 1x final primer concentration (Qiagen primers, Germantown, MD) or 1 µM sense and 1 µM antisense oligonucleotides (Invitrogen custom DNA oligos, Carlsbad, CA) with SYBR Green (Bio-Rad, Hercules, CA). Human and mouse primers are listed in Table 8.1 and Table 8.2, respectively. Each target sequence was amplified as follows: 1x (95°C for 3 min), 40x (95°C for 15 sec, 55°C for 30 sec, 72°C for 10 sec),

1x (72°C for 1 min), a melt curve protocol, and 4°C hold.

MicroRNA PCR Array

Ischemic preconditioning (4 cycles of 5 min ischemia and 5 min reperfusion) with a final reperfusion time of 120 minutes was performed in C57BL/6J (The Jackson Laboratory) or Per2-/- mice. Heart tissue was snap-frozen with clamps pre-cooled to the temperature of liquid nitrogen.

2 MicroRNA was isolated with Trizol (Invitrogen) and purified using RT qPCR-Grade miRNA

2 Isolation Kit (SABiosciences-Qiagen). cDNA template was generated using RT miRNA First

2 Strand Kit (SABiosciences-Qiagen). miRNA expression was performed using RT miRNA PCR

Array Mouse miFinder (SABiosciences-Qiagen).

172 MicroRNA Analysis

100 ng of microRNA eluate was used to make cDNA following the miScript RT II Kit manufacturer’s instructions (Qiagen). cDNA was diluted to 1 to 5 ng/uL for determining transcript levels by real-time quantitative PCR (iCycler; Bio-Rad Laboratories Inc.) and following manufacturer’s instructions for miScript SYBR Green PCR Kit (Qiagen) (336) (336). Each target sequence was amplified using increasing numbers of cycles of 94°C for 1 min, 58°C for 0.5 min,

72°C for 1 min. Quantification of transcript levels was determined by real-time RT-PCR (iCycler;

Bio-Rad Laboratories Inc.).

8.3 Immunoblotting Analysis

Protein was isolated from HMEC-1 using M-Per following manufacturer’s instructions

(Thermo Fisher Scientific, Waltham, MA) and including a protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA). Protein was quantified using a Qubit

Fluorometer 3.0 and Qubit Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA). 5 – 25 µg of protein was denatured at 95°C in Laemmli sample buffer for 5 min. Samples were resolved on a 4 – 10% polyacrylamide gel and transferred to nitrocellulose membranes, which were blocked for 1 hr at room temperature in either 5% BSA / TBST or 5% milk / TBST. The membranes were incubated in primary antibody at a concentration of 1:1000 overnight at 4°C. Primary antibodies are listed in Table 8.3. The next day, blots were washed 3 – 4x with TBST and incubated with secondary antibody at a concentration of 1:5000 in the respective blocking buffer, washed an additional 3 times, and visualized using SuperSignal West Femto Maximum Sensitivity Substrate

(Thermo Fisher Scientific, Waltham, MA). Secondary antibodies are listed in Table 8.3.

8.4 DNA-Protein and Protein-Protein Interactions

Chromatin immunoprecipitation (ChIP) assay

ChIP assays were performed using the ChIP-IT™ Express Enzymatic Kit from Active Motif

(Carlsbad, CA, USA). Briefly, Scr and PER2KD HMECs were grown to 90% confluence in phenol red-free Dulbecco's modified Eagle medium (DMEM) supplemented with 10% charcoal

173 DEXTRAN-stripped FBS for at least 3 days. After hypoxia exposure at 1% O2 for 24h, ChIP assays were performed according to manufacturer's protocol. Briefly, chromatin was cross-linked in 1% formaldehyde in minimal cell culture medium (Invitrogen, Carlsbad, CA), and nuclei were extracted. Chromatin was enzymatically digested for 11mins to yield 200- to 1,500-bp DNA fragments and the supernatant containing precleared chromatin was then incubated at 4°C overnight with mouse monoclonal HIF1⍺ antibody (H1alpha67, ChIP Grade, Abcam, Cambridge,

MA) or rabbit IgG control (Cell Signaling, Danvers, MA) (Table 8.3). After reverse cross-linking by heating the samples at 65°C overnight and treating with Proteinase K, DNA was purified using phenol-chloroform extraction. Quantitative analyses of DNA products obtained from ChIP assay were performed by RT-PCR with primers specific for the human LDHA promoter. RT-PCRs conducted on DNA derived from input chromatin templates served as positive controls whereas reactions conducted on IgG-precipitated templates served as negative controls. The RT-PCR signal was barely detectable for these controls. The signal for these samples and IgG-precipitated templates was negligible on gels. Primers used are listed in Table 8.1. Conventional PCR signals were stained with ethidium bromide in 1% agarose gels.

Affinity purification-mass spectrometry-based proteomics

HMEC-1 were placed in a hypoxia chamber (Coy Laboratory Products Inc.) in

preequilibrated hypoxic medium at 1% O2. Following 24 h of hypoxia or normoxia, the samples

were isolated for cytoplasmic and nuclear protein fractions according to the NE-PER kit

specifications (Thermo Scientific, 78833). To identify interacting proteins with PER2, co-

immunoprecipitation (Co-IP) for PER2 was performed using the Pierce Co-IP Kit (Thermo

Scientific, 26149). Specifically, 10 µg of rabbit anti-PER2 antibody (Novus, NB100-125) was immobilized to the amine-reactive resin. 100 µg of sample was incubated overnight at 4 oC with

the anti-PER2 coupled resin. Samples were washed and then eluted. Samples were loaded onto

a 1.5mm NuPAGE Bis Tris 4-12% gradient gel and proteins were separated as previously

174 described(338). The gel was stained using SimplyBlue™ SafeStain (Invitrogen, Carlsbad, CA) and de-stained with water. Each lane of the gel was divided into 9 equal-sized bands, gels were destained, disulfide bonds were reduced, and cysteine residues were alkylated as previously described(338). Gel pieces were subsequently washed with 100 µL of distilled water followed by addition of 100 mL of acetonitrile and dried on SpeedVac (Savant Thermo Scientific). Then 100 ng of trypsin was added to each sample and allowed to rehydrate the gel plugs at 4 °C for 45 min and then incubated at 37 °C overnight. The tryptic mixtures were acidified with formic acid up to a final concentration of 1%. Peptides were extracted two times from the gel plugs using 1% formic acid in 50% acetonitrile. The collected extractions were pooled with the initial digestion supernatant and dried on SpeedVac (Savant ThermoFisher). Samples were desalted on Thermo

Scientific Pierce C18 Tip. Similar to previously described studies(339), samples were analyzed on an LTQ Orbitrap Velos mass spectrometer (Thermo Scientific) coupled to an Eksigent nanoLC-

2D system through a nanoelectrospray LC-MS interface. Specific to these experiments and following desalting, The peptides from each sample were separated over a 90-min linear gradient of 2-40% ACN with 0.1% formic acid. LC mobile phase solvents and sample dilutions used 0.1% formic acid in water and 0.1% formic acid in acetonitrile (OptimaTM LC/MS, Fisher Scientific). Data acquisition and mass spectrometer was operated in positive ion mode as previously described(339). Full MS scans were acquired in the Orbitrap mass analyzer as in previous studies

(340). For database searching and protein identification, MS/MS spectra were extracted from raw data files and converted into mgf files using MassMatrix (Cleveland, OH). These mgf files were then independently searched against mouse SwissProt database using an in-house Mascot™ server (Version 2.2, Matrix Science). Mass tolerances were +/- 15 ppm for MS peaks, and +/- 0.6

Da for MS/MS fragment ions. Trypsin specificity was used allowing for 1 missed cleavage. Met oxidation, protein N-terminal acetylation, peptide N-terminal pyroglutamic acid formation were allowed for variable modifications while carbamidomethyl of Cys was set as a fixed modification.

Scaffold Software (version 4.4.6, Proteome Software), peptide, and protein identification

175 probability parameters were used as previously described(339). Following identification of

potential PER2 interacting proteins in normoxia, hypoxia, and normoxia vs. hypoxia, lists obtained

from Scaffold were analyzed by Ingenuity Pathway Analysis (Qiagen), Panther Classification

System, and Reactome Analysis to detect pathways PER2 regulates in normoxia and hypoxia.

Co-immunoprecipitations (Co-IPs)

Co-IPs were done using the Pierce Co-Immunoprecipitation (Co-IP) Kit (Thermo Fisher

Scientific, Waltham, MA). 10 ug of antibody was immobilized to columns (Table 8.3). Pulled-down protein was quantified using a Qubit Fluorometer 3.0 and resolved by immunoblotting as described.

8.5 Metabolic Measurements

Lactate measurements

Lactate measurements were done using the L-Lactate Colorimetric Assay Kit following the

manufacturer’s protocol (Abcam, Cambridge, MA).

Cytotoxicity

Cytotoxicity was determined using the LDH-Cytotoxicity Assay Kit per manufacturer’s

protocol (Abcam, Cambridge, MA).

Glycolytic stress tests using the Seahorse Bioanalyzer

Cells were synchronized by serum starvation, followed by glycolytic stress test on the

XF24 Seahorse Bioanalyzer following manufacturer’s specifications (Agilent, Santa Clara, CA).

Cells were plated in the morning at a density of 1.2x105 cells / well and serum starved in the evening one day prior to assaying. One hour prior to the stress test cells were incubated in XF

Assay Medium (Agilent, Santa Clara, CA) at a pH of 7.4. Final concentration of glucose was 10 mM / well, oligomycin 1.0 µM / well, and 2-deoxyglucose 50 mM / well.

Mitochondrial stress tests using the Seahorse Bioanalyzer

Cells were synchronized by serum starvation, followed by mitochondrial stress test on the

XF24 Seahorse Bioanalyzer (Agilent, Santa Clara, CA). For all assays, pH and oxygen

176 consumption rate (OCR) were measured. For TCA cycle readouts, an additional measurement of

carbon dioxide evolution rate (CDER) was measured. Final concentrations were 1.0 µM oligomycin, 3.6 µM FCCP, and 1.125 µM Rotenone / Antimycin A. For fatty acid mitochondrial stress tests, cells were supplemented with palmitate:BSA (Agilent, 102720-100) prior to the run

and half the plate received 800 µM etomoxir as a control.

Cell energy phenotype assays using the Seahorse Bioanalyzer

Following manufacturer’s instructions (Agilent, Santa Clara, CA), the Seahorse

Bioanalyzer was used to assess cell energy phenotype at baseline.

Enzyme activities IDH, ACO, SUCLG, Complex IV, PFK, LDH

Human isocitrate dehydrogenase (IDH, Biovision, Milpitas, CA), aconitase (ACO, Abcam,

Cambridge, MA), succinyl-CoA synthetase (SUCLG, Abcam, Cambridge, MA),

phosphofructokinase (PFK, Biovision, Milpitas, CA) and lactate dehydrogenase (LDH, Biovision,

Milpitas, CA) activity or mouse complex 4 activity (Abcam, Cambridge, MA) were measured using

colorimetric assay kits adhering to manufacturer’s instructions.

Mitochondrial membrane potential dyes

MitoTracker Red CMXRos (Invitrogen Molecular Probes, Carlsbad, CA) and JC-1

Mitochondrial Membrane Potential Assay (Abcam, Cambridge, MA) were used per

manufacturer’s specifications using 5 uM of JC-1 for 30 minutes at 37C. JC-1 quantification was

done by calculating mean intensity.

13C-tracers in vitro

HMEC-1s were serum starved in either MCDB131 (low glucose) or glucose-free DMEM

for 24 h prior to assay. Respective mediums were supplemented with either 12 mM U-13C-glucose

13 (Cambridge Isotope Laboratories, Tewksbury, MA) or 166.67 µM 1,2- C2-palmitic acid

(Cambridge Isotope Laboratories, Tewksbury, MA) in hypoxia or normoxia for 24 h. Frozen cell

pellets were extracted at 2e6 cells/mL in ice cold lysis/ extraction buffer

177 (methanol:acetonitrile:water 5:3:2). Samples were agitated at 4 °C for 30 min followed by

centrifugation at 10,000 g for 10 min at 4 °C. Protein and lipid pellets were discarded, and

supernatants were stored at -80 °C prior to metabolomics analysis. Ten µL of extracts were

injected into a UHPLC system (Vanquish, Thermo, San Jose, CA, USA) and run on a Kinetex

C18 column (150 x 2.1 mm, 1.7 µm - Phenomenex, Torrance, CA, USA). Solvents were Optima

H2O (Phase A) and Optima acetonitrile (Phase B) supplemented with 0.1% formic acid for positive mode runs and 1 mM NH4OAc for negative mode runs. For [U-13C]-glucose flux analysis (341), samples were run on a 3 min isocratic (95% A, 5% B) run at 250 µL/min (342) (343). For [1,2-

13 C2]-palmitate flux analysis, samples were analyzed using a 9 min gradient from 5-95% acetonitrile organic phase at 400 µL/min. The autosampler was held at 7 oC for all runs; the column compartment was held at 25 oC for the 3 min method and 45 oC for the 9 min method.(344) The UHPLC system was coupled online with a Q Exactive mass spectrometer

(Thermo, Bremen, Germany), scanning in Full MS mode (2 µscans) at 70,000 resolution in the

60-900 m/z range in negative and then positive ion mode (separate runs). Eluate was subjected to electrospray ionization (ESI) with 4 kV spray voltage. Nitrogen gas settings were 15 sheath gas and 5 auxiliary gas for the 3 min runs; 45 sheath gas and 15 auxiliary gas for the 9 min runs.

Metabolite assignments and isotopologue distributions were determined using Maven (Princeton,

NJ, USA), upon conversion of ‘.raw’ files to ‘.mzXML’ format through MassMatrix (Cleveland, OH,

USA). Chromatographic and MS technical stability were assessed by determining CVs for heavy and light isotopologues in a technical mixture of extract run every 10 injections.

8.6 Mouse Experiments

General mice usage

Experimental protocols were approved by the Institutional Review Board (Institutional

Animal Care and Use Committee [IACUC]) at the University of Colorado Denver, USA. They were

in accordance with the NIH guidelines for use of live animals. Mice were housed in L:D 14:10 and

we routinely used 12- to 16-week old male mice (74) (71), except where specified females were

178 used to determine sex differences. Unless otherwise noted, all mouse experiments were

conducted at the time point ZT3.

Light exposure in mice

Mice were exposed to intense light (10,000 LUX, Lightbox simulating day light, Uplift

Technologies DL930 Day-Light 10,000 Lux SAD, full spectrum) for 3 h (71) or one week and

compared to mice maintained at room light (200 LUX) (71). Mice were housed in a 14/10-h light- dark cycle to synchronize (entrain) the circadian clock of WT mice to the ambient light-dark cycle.

We conducted all mouse experiments at same time points (ZT 3, ZT15).

PER2-/- mice

PER2-/- mice (Per2tm1Brd Tyrc-Brd/J) were obtained from the Jackson Laboratories (241) and contain targeted deletions of the exons coding for the PAC subdomain and half the PAS B domain, resulting in expression of a non-functional protein (241). Controls (C57BL/6J) were

obtained from the Jackson Laboratories (241). Characterization and validation were performed

as described previously. Homozygous mutant mice are morphologically indistinguishable from

their wild-type littermates and both males and females are fertile (71, 74).

PER2 reporter mice

Reporter mice were purchased from Jackson laboratories.(345) (346) PER2 reporter mice

(B6.129S6-Per2tm1Jt/J) are homozygous for the "mPer2Luc" mutation and are viable and fertile with no developmental or morphological differences compared to wildtype littermates. Expression of PERIOD2: :LUCIFERASE (or mPER2::LUC) fusion protein during peak periods is similar to endogenous pPER2 patterns. Homozygous mice have normal entrainment and circadian behaviors.

Mir21-/- mice

MiR-21-/- and controls (B6129SF1/J) were obtained from the Jackson Laboratories (347).

Characterization and validation were performed as described previously. Homozygous mutant

179 mice are morphologically indistinguishable from their wild-type littermates and both males and

females are fertile (71, 74, 347).

Murine Model for cardiac ischemic preconditioning (IPC)

IPC (69-71, 74, 76, 135, 265, 266, 348): Anesthesia was induced (70 mg/kg body weight i.p.) and maintained (10 mg/kg/h) with sodium pentobarbital. Mice were placed on a temperature- controlled heated table (RT, Effenberg, Munich, Germany) with a rectal thermometer probe attached to a thermal feedback controller to maintain body temperature at 37°C. The tracheal tube was connected to a mechanical ventilator (Servo 900C, Siemens, Germany) with pediatric tubing and the animals were ventilated with a pressure controlled ventilation mode (peak inspiratory pressure of 10 mbar, frequency 110 breaths/min, positive end-expiratory pressure of

3 mbar, FiO2 = 0.3). Blood gas analysis revealed normal paO2 (115±15 mmHg) and paCO2 (38±6 mmHg) levels with our ventilator regime. After induction of anesthesia, animals were monitored with a surface electrocardiogram (ECG, Hewlett Packard, Böblingen, Germany). Fluid replacement was performed with normal saline, 0.2 ml/h i.v. The carotid artery was catheterized for continuous recording of blood pressure with a statham element (WK 280, WKK, Kaltbrunn,

Switzerland). Operations were performed under an upright dissecting microscope (Olympus

SZX12). Following left anterior thoracotomy, exposure of the heart and dissection of the pericardium, the left coronary artery (LCA) was visually identified and an 8.0 nylon suture

(Prolene, Ethicon, Norderstedt, Germany) was placed around the vessel. Atraumatic LCA occlusion for IPC studies was performed using a hanging weight system (265) (349). Successful

LCA occlusion was confirmed by an immediate color change of the vessel from light red to dark violet, and of the myocardium supplied by the vessel from bright red to white, as well as the immediate occurrence of ST-elevations in the ECG. During reperfusion, the changes of color immediately disappeared when the hanging weights were lifted and the LCA was perfused again

(69, 70, 76, 135).

180 Murine model for myocardial ischemia and reperfusion injury

Murine model for myocardial IR-injury (69-71, 74, 76, 135, 265, 266, 348):

Anesthesia was induced (70 mg/kg body weight i.p.) and maintained (10 mg/kg/h) with sodium pentobarbital. Mice were placed on a temperature-controlled heated table (RT, Effenberg,

Munich, Germany) with a rectal thermometer probe attached to a thermal feedback controller to maintain body temperature at 37°C. The tracheal tube was connected to a mechanical ventilator

(Servo 900C, Siemens, Germany) with pediatric tubing and the animals were ventilated with a pressure-controlled ventilation mode (peak inspiratory pressure of 10 mbar, frequency 110 breaths/min, positive end-expiratory pressure of 3 mbar, FiO2 = 0.3). Blood gas analysis revealed normal paO2 (115±15 mmHg) and paCO2 (38±6 mmHg) levels with our ventilator regime. After induction of anesthesia, animals were monitored with a surface electrocardiogram (ECG, Hewlett

Packard, Böblingen, Germany). Fluid replacement was performed with normal saline, 0.1ml/h i.p.

The carotid artery was catheterized for continuous recording of blood pressure with a statham element (WK 280, WKK, Kaltbrunn, Switzerland). Operations were performed under an upright dissecting microscope (Olympus SZX12). Following left anterior thoracotomy, exposure of the heart and dissection of the pericardium, the left coronary artery (LCA) was visually identified and an 8.0 nylon suture (Prolene, Ethicon, Norderstedt, Germany) was placed around the vessel.

Atraumatic LCA occlusion for ischemia studies was performed using a hanging weight system

(265, 350). Successful LCA occlusion was confirmed by an immediate colour change of the vessel from light red to dark violet, and of the myocardium supplied by the vessel from bright red to white, as well as the immediate occurrence of ST-elevations in the ECG. During reperfusion, the changes of color immediately disappeared when the hanging weights were lifted and the LCA was perfused again (69, 70, 76, 135). To study cardioprotective effects, it is ideal to use an ischemia time associated with infarct sizes of approximately 30–40% of the area at risk (AAR).

Thus, it is possible to demonstrate changes in both directions (e.g., smaller or larger infarct sizes) with an experimental therapeutic or a specific gene deletion. An ischemia time of 60 min resulted

181 in a mean infarct size of approximately 45% of the AAR as shown previously. (70, 71, 76, 135,

264-266, 348) Thus, a 60-min ischemia period was chosen in the current study. Infarct sizes were determined by calculating the percentage of myocardium that underwent infarction compared to the area at risk (AAR) using a previously described double staining technique with Evan’s blue and triphenyltetrazolium chloride (TTC). Evan’s blue is excluded from the area of the heart perfused by the LCA and thus allows one to identify the AAR. TTC stains all cells red except those that are depleted in NADPH and therefore allows one to visualize the white infarcted tissue. AAR and the infarct size were determined via planimetry using the NIH software Image 1.0 and the degree of myocardial damage was calculated as percent of infarcted myocardium from the AAR

(69, 70, 76). To measure reliability of infarct size analysis, interobserver variability was tested.

Infarct sizes of animals were assessed by two independent investigators both blinded to the experimental protocol. Moreover, infarct size was measured twice on two separate days by the same investigator to reveal interobserver variability. Compounds used for the small-molecule screen are listed in Table 8.4.

13C-tracers in vivo

C57BL/6 wildtype mice were housed under daylight (10,000 LUX, L:D 14:10 h) or standard

room light (200 LUX, L:D 14:10 h) for 7 days followed by infusion of 10 mg/kg/min U-13C-glucose

(Cambridge Isotope Laboratories, Tewksbury, MA) via an intra-arterial catheter over 90 minutes.

Left ventricles of the hearts were flushed with ice cold KCl and the left ventricle was shock frozen

and analyzed by liquid chromatography–tandem mass spectrometry. Isotope-resolved metabolite

analyses were performed by LC-MS on a Waters Acquity ultrahigh-performance liquid chromatography (UPLC) system coupled to a Waters Synapt HDMS quadrupole time-of-flight

mass spectrometer equipped with an atmospheric pressure electrospray ionization (ESI) source.

LC-MS was performed with full-mass detection within m/z 100-1000 in the positive-ion or

negative-ion mode as described below. The typical MS operation parameters included ESI spray

voltage 2.8 kV, sampling cone voltage 4 V, drying gas flow 50 L/min / temperature 120 oC, and

182 nebulizing nitrogen gas flow 700 L/h / temperature 350 oC. The authentic compound sodium D- fructose-6-sphosphate was obtained from Sigma-Aldrich and were used as the standard substances for location and identification of the targeted metabolites. UPLC conditions were

optimized to ensure appropriate separation of each metabolite from its structural isomers and the

interfering components detected in the samples. For sample preparation, frozen specimens were

individually ground to fine powder in liquid nitrogen and were weighed to 5-mL borosilicate glass

test tubes. Ice-cold methanol-water (50:50, v/v), equivalent to 1 mL per 100 mg tissue, was added

to each tube. The samples were lysed on ice with 15 s x 2 sonications using a Fisher Scientific

Model 100 cell dismembrator. After vortex-mixing and 5-min sonication in an ice-water bath, ice-

cold methanol, equivalent to 1 mL per 100 mg tissue, was added. The sample tubes were capped,

violently vortex-mixed for 2 min, sonicated in the same ice-water bath for 5 min and then placed

at – 20 oC for 2 h. before the tubes were centrifuged at 4,000 rpm and 4 oC for 15 min in a Beckman

Coulter Allegra X-22R centrifuge. The clear supernatant of each sample was collected and

transferred to a 3-mL borosilicate glass test tube for the following LC-MS analyses. Fructose-6-

phosphate was measured by ion-pairing LC-MS using tributylamine (TBA) as the paired counter

ion reagent (351). The chromatographic separation was conducted on a YMC-Triart C18 UPLC

column (2.1 x 150 mm, 1.9 µm) using binary-solvent gradient elution with 2 mM TBA in water (pH

adjusted to 6 with acetic acid) as mobile phase A and methanol as mobile phase B. The elution

gradient was 0-9 min, 2% to 55 % B; 9-9.5 min, 55% to 100 % B; 9.5– 11 min, 100 % B. The

column was equilibrated with 2% B for 5 min between injections. The column flow rate was 0.25

mL/min and the column temperature were 50 oC. A 100-µL aliquot of each supernatant from

individual mouse heart specimens was dried under a nitrogen flow in a fume hood and the residue

was reconstituted in 200 µL of mobile phase A. 5 µL was injected for LC-MS with negative-ion

detection. The organic acids were analyzed by chemical derivatization LC-MS with 3-nitrophenyl

hydrazine (3NPH) as the pre-analytical derivatizing reagent.(352) In brief, 100 μL of the

supernatant was mixed 50 μL of 200 mM 3NPH.HCl in 75% methanol and 50 μL of 150 mM 1-

183 ethyl-3-(3-dimethylaminopropyl)carbodiimide-HCl. The mixture could react at 30 oC for 30 min

and was then mixed with 400 μL of water. 10-μL aliquots were injected onto a C8 UPLC column

(2.1 x 50 mm, 1.7 μm) for LC-MS runs with negative-ion detection and using the LC procedure as

described.(352) For all the above LC-MS analyses, the monoisotopic ion chromatograms of

individual metabolites, together with their isotopomeric counterparts resulting from the U-13C

glucose tracer, were extracted based on their calculated m/z values, within a mass window of 60

ppm (+/- 30 ppm), and their peak areas were integrated. The peak areas of any observed

isotopomeric forms derived from the U-13C glucose tracer for each metabolite was corrected by

subtracting the abundance contributions from the natural and any other U-13C glucose tracer-

derived isotopic forms.

Heart enzyme measurement

Blood was collected by central venous puncture for troponin I (cTnI) measurements using a quantitative rapid cTnI assay (Life Diagnostics, Inc., West Chester, PA, USA). cTnI is highly specific for myocardial ischemia and has a well-documented correlation with the infarct size in mice (265) (76) (70) (69) and humans (353).

Luciferase assay – tissue

Expression of the Per2 reporter genes was assayed using the tissue homogenates in T-

per Tissue Protein Extraction Reagent (Pierce, Thermo Fisher Scientific, Waltham, MA). The homogenates were centrifuged for 30mins at 4900 x g at 4oC. The luciferase gene expression was measured by using the Dual-Luciferase Reporter Assay System from Promega according to the manufacturer's instructions using a Biotek Synergy 2 Multimode Microplate Reader (Winooski,

VT).

Wheel running

Mice were maintained individually in running-wheel cages (Starr Life Sciences, wheel

diameter: 11.5 cm). Running-wheel activity was recorded every 5 minutes using the Vital View

184 Data Acquisition System (Starr Life Sciences, Oakmont, PA). Data were analyzed using BioDare

(Biological Data Repository) (354). Amplitude of wheel running activity were calculated using the

fast fourier transform non-linear least squares (FFT-NLLS) method.

Enucleation procedure in mice

Mice were pre-anesthetized with subcutaneous carprofen and buprenorphine injections

and anesthetized with a ketamine/xylazine/ace cocktail. The periocular region was then clipped

to remove surrounding hair. After a surgical scrub with betadine, the optic nerve and associated

blood vessels were clamped with hemostats. After 2 minutes the entire globe of eye and the optic

nerve were removed. For additional analgesia bupivacaine was dripped into socket. Next, the

complete upper and lower eyelid margin were removed with fine tip scissors and the eyelids were

closed with 3-4 interrupted sutures. Mice were recover in cages on warm water circulating

blankets. Carprofen injections were repeated every 24 hours for 2 days post operatively. After a

2-week recovery period under standard housing conditions, mice underwent daylight housing or

myocardial ischemia studies.

cAMP ELISA and phospho-CREB assays

Phospho-CREB (S133) Immunoassay (R&D Systems, Minneapolis, MN) or cAMP

Parameter Assay Kit for mouse (R&D Systems, Minneapolis, MN) or human (R&D Systems,

Minneapolis, MN) were used according to the manufacturer’s protocol.

Hydrogen peroxide assay

Hydrogen peroxide levels were measured using the hydrogen peroxide assay kit (Abcam,

Cambridge, UK) according to the manufacturer’s protocol. In brief, the left ventricle (area at risk) was harvested after 60 min of ischemia and 15 min of reperfusion. Heart tissues were homogenized in assay buffer and spun down at 13,000 g, 4°C for 5 min. Supernatants were deproteinized by adding ice-cold perchloric acid. The samples were again centrifuged at

13,000×g, 4 °C for 2 min, and the supernatant was precipitated by adding ice-cold potassium hydroxide. Next, samples were centrifuged at 13,000×g, 4 °C for 15 min, and hydrogen peroxide

185 was measured in the supernatant. Using a 96-well microplate, samples were added, and reactions

were initiated immediately by adding OxiRed. Fluorescence was measured on Synergy 2 Multi-

Mode Microplate reader (Biotek,Winooski, VT, USA) in excitation range of 540/25 and emission detection of 620/40. Fluorescence levels were normalized to the protein concentration of samples before deproteinization.

8.7 Isolation of cardiac tissues

Isolation of fibroblasts

Heart tissue from C57BL6/J mice was minced and digested using Collagenase Type II

solution (Worthington Biochemical Corporation) at 37°C, 100 rpm, collecting the supernatant every 10 minutes for 90 minutes and replacing with fresh collagenase solution until heart tissue fully digested. Fibroblasts were isolated after plating and incubation of the cell suspension in a cell culture incubator with 5% CO2 for 2 h. 2 h upon plating alive and healthy fibroblasts were adhered to the dish. After cells reached confluency, cells were exposed to normoxia (21% oxygen) or hypoxia (1% oxygen using preequilibrated media for 6 h (355)) and immediately resuspended in Trizol for miRNA analysis.

Isolation of adult cardiomyocytes

Isolation of adult cardiomyocytes (71, 266): Protocol for isolation of adult cardiomyocytes

was adapted from O’Connell et al (356). 8-12 weeks old C57BL6/J mice were anesthetized, and

the heart was quickly removed from the chest cavity and immediately placed in KHB buffer. The

aorta was cannulated, and the heart perfused with Ca2+-free KHB for 3 min followed by 8-12 min

perfusion with KHB containing 40 μM Ca2+ in and collagenase II (Worthington Biochemical Corp).

After perfusion, ventricles were removed, minced and incubated in 15 mL collagenase solution

for an additional 3-7 min. An equal volume of stopping buffer (KHB containing 10% FBS, 12.5 M

Ca2+, and 2 mM ATP) was added to the digestion solution. Myocytes were allowed to sediment

by gravity for 3 min at room temperature and centrifuged at 20 x g for 3 min. Pellet was

resuspended in 100 μM Ca2+ and sedimentation followed by centrifugation was repeated for

186 subsequent 400 μM and 900 μM Ca2+ in a slow calcium re-introduction process. Myocytes were

resuspended in MEM (Gibco 11575-032) supplemented with 10% FBS, 10 mM BDM, 100 U/mL

Pen/Strep, and 2 mM ATP and plated on laminin-coated plates (10 μg/mL laminin in PBS).

Myocytes were incubated in a 37°C, 2% CO2 incubator. After health myocyte adhesion, media

was exchanged for MEM supplemented with 10 mM BDM, 1 X ITS (final concentrations of 5 μg/mL

insulin, 5 μg/mL transferrin, and 5 ng/mL selenium), 100 U/mL Pen/Strep, and 1 mg/mL BSA.

Myocytes were treated with midazolam or vehicle in this culture medium for 6 h and thereafter immediately resuspended in Qiazol for mRNA isolation and gene expression analysis (71).

8.8 Human experiments

Human light exposure

Healthy human volunteers were exposed to bright light exposure (daylight, ~10,000 LUX)

for 30 min every morning for five days from 8:30 AM – 9:00 AM. 5 mL blood was drawn on day

one at 8:30 AM and 9:00 AM (before and after light exposure). While light exposure was repeated

every morning for the five days, the next blood draws were on day three and five at 9:00 AM. For

experiments involving actigraphy watches, the same volunteers wore the watch for one week

without bright light exposure and maintained watching the watches during the second week when

light exposure was included. We obtained approval from the Institutional Review Board (COMIRB

#13-1607) for our human studies prior to written informed consent from each individual. A total of

20 healthy volunteers were enrolled (11 female and 6 male, age range between 21-44 yrs.) The

average age was 29.5 years old (range 23 – 41 yo). All but one individual identified as a caffeine

drinker. The average number of hours slept prior to intense light exposure did not differ from the

week of intense light exposure (6.4h). Healthy human volunteers were exposed to 30 minutes of

intense light (Square One Wake Up Light, NatureBright, Day-Light 10,000 Lux) in the morning at

8:30 AM for 5 consecutive days. A blood draw was performed before light exposure on the first

day (8:30 AM) and 5 days after light exposure (9.00 AM). Blood was collected in EDTA-plasma

tubes and spun at 3,000 rpm for 8 minutes to separate plasma. Plasma samples were analyzed

187 for miR-21 levels and PFK (phosphofructokinase) activity. Light boxes were a generous gift from

Joshua Chen, NatureBright.

Phosphofructokinase (PFK) activity

Phoshpofructokinase activity was measured using a PFK Activity Colorimetric Assay Kit

(BioVision, cat. no. K776-100), adhering to manufacturer’s instructions.

Human plasma melatonin, HIF1⍺ and triglycerides levels

Melatonin levels were measured using the MT Elisa Kit for humans (My BioSource, San

Diego, CA). HIF1⍺ levels from human plasma samples were measured using the human HIF1⍺

ELISA Kit from Invitrogen (Carlsbad, CA). Triglycerides were determined using a human

Triglyceride Quantification Assay kit (Abcam, Cambridge, MA).

Targeted metabolomics - mass spectrometry

Targeted metabolomics of human plasma following light exposure was performed as

previously reported (A three-minute method for high-throughput quantitative metabolomics and

quantitative tracing experiments of central carbon and nitrogen pathways (343) (343).

Data analysis

Data were compared by one-way ANOVA with Tukey’s post-hoc test, or by Student's t-

test where appropriate. For comparison of two groups the unpaired student t-test was performed.

For multiple group comparisons a one-way analysis of variance with a Tukey’s post hoc test was performed. Values are expressed as mean±SD from 3-6 animals/individual cell experiments per condition and P<0.05 was considered statistically significant. The chosen numbers per group was based on findings in previous studies and a subsequent sample size analysis. The studies are designed to be able to reject the null hypothesis that the population means of the experimental and control groups are equal with probability (power) 0.8. The Type I error probability associated with this test of this null hypothesis is 0.05. P<0.05 was considered statistically significant. For all statistical analysis, GraphPad Prism 6.0 software for Windows XP was used.

188 Table 8.1 Human Primers Target Primer Name Catalog No. or Forward and Reverse Human PER2 Hs_PER2_1_SG Qiagen QuanitTect QT00011207 Human PKM Hs_PKM_1_SG Qiagen QuanitTect QT00028875 Human LDH Hs_LDHA_1_SG Qiagen QuanitTect QT00001687 Human SIRT3 Hs_SIRT3_1_SG QT00091490 Human Beta-Actin Hs_ACTB_2_SG QT01680476 Invitrogen Sense 5'-CTA GGC ACC AGG GTG TGA T -3' Antisense 5'-TGC CAG ATC TTC TCC ATG TC-3' Human COX4|1 Hs_COX4|2_1_SG QT00044933 Human OPN4 Invitrogen Sense: 5’-AGT CGC CCC TAC CCC AGC TA-3’ Antisense: 5’-CAC AGC TGC TGC CTC CAT GT-3’ Human miR-21 Hs-miR21 miScript Primer Assay, Qiagen 5'-UAG CUU AUC AGA CUG AUG UUG A Human RNU Hs-RNU6-2_11 miScript Primer Assay, Qiagen, MS00033740 Functional in human, mouse, rat, dog, rhesus macaque, cow pig, and sheep) Human LDH Custom Invitrogen Promoter Sense ATT ACG TGC CAG AAG CTG TT Antisense TTT CCT CAT CCA TGA AAC CT

189 Table 8.2 Mouse Primers Company and Catalog No. or Forward and Target Primer Name Reverse Mouse PER2 Mm_Per2_1_SG Qiagen QuantiTect QT00198366 Or Invitrogen Sense: 5´-ACC TGC TCA ACC TCC TTC TG-3´ Antisense: 5´-ACT ACT GCC TGC CCC ACT TT-3´ Mouse Beta-Actin Mm_Actb_2_SG Qiagen QuanitTect QT01136772 Mouse miR-21 Mm-miR-21 miScript Primer Assay, Qiagen 5’-UAG CUU AUC AGA CUG AUG UUG A

190 Table 8.3 Primary and Secondary Antibodies Antibody Company and Catalog No. Rabbit polyclonal PER2 Novus Biologicals, NB100-125

Abcam, ab64460 rabbit polyclonal IDH2 Novus Biologicals, NBP2-22166 rabbit polyclonal SUCLG1 Novus Biologicals, NBP1089489 rabbit polyclonal ACO2 Novus Biologicals, H00000050-D01P rabbit polyclonal SIRT3 Abcam, ab86671 1° Antibodies 1° mouse monoclonal β-ACTIN Cell Signaling Technologies, 8H10D10 mouse monoclonal anti-DDK (FLAG) OriGene Technologies, TA50011-100 Mouse monoclonal HIF1A Abcam, H1alpha67 ChIP Grade rabbit polyclonal IgG Novus Biologicals, NB7487

goat anti-mouse IgM Calbiochem 2° goat anti-rabbit IgG Thermo Fisher Scientific Rabbit IgG control (ChIP) Cell Signaling

191 Table 8.4 Pharmacological Compounds Pharmacological Compound Dose Company Nobiletin 1 mg/kg Sigma-Aldrich Tangeretin 0.93 mg/kg Sigma-Aldrich Sinensetin 0.93 mg/kg Sigma-Aldrich 5,6,7-trimethoxyflavone 0.78 mg/kg Sigma-Aldrich Flavone 0.55 mg/kg Sigma-Aldrich 3',4',7,8-tetramethoxyflavone 0.85 mg/kg Alfa Aesar Midazolam 200 mg/kg Pfizer Ketamine 200 mg/kg Pfizer Fentanyl 1 mg/kg Pfizer Propofol 200 mg/kg Fresenius Kabi Isoflurane 1% Baxter

192 CHAPTER 9 CHAPTER IX CONCLUSIONS AND FUTURE DIRECTIONS

9.1 Summary

We began our studies with the hypothesis that the circadian rhythm protein PER2 is

cardioprotective by optimizing metabolic pathways during myocardial ischemia. Through a variety

of experiments investigating mechanisms of PER2 in low oxygen conditions, we uncovered a

novel role for PER2 and circadian amplitude enhancement in cardioprotection from myocardial

ischemia. We first dove into the cellular mechanisms of PER2 mediated metabolic adaptation to

hypoxia using HMEC-1. We found that PER2 has multiple roles, functioning both transcriptionally

and post-translationally. In the nucleus, PER2 regulates binding of HIF1⍺ to the promoter region

of a key anaerobic glycolysis gene, LDH (Figure 2.3). Similarly, we identified COX4.2, the

hypoxia-responsive oxidative phosphorylation complex under HIF1⍺ control (105) is ultimately

regulated by PER2 (Figure 2.24, 2.25). Furthermore, we found that the SIRT3 gene, the protein

of which is a prominent regulator of mitochondrial metabolic processes (117, 121-123, 125, 126,

357), is expressed in a PER2 and HIF1⍺ dependent manner (Figure 2.17, 2.18). In the

mitochondria, we found PER2 complexes with SIRT3 and TCA cycle enzymes, like IDH2, and

subsequently PER2 is a mediator of TCA cycle enzyme function in hypoxia (Figure 2.12 – 2.14).

Additional studies revealed that PER2 is critical for maintaining membrane potential in hypoxia

(Figure 2.26, 2.29, and 2.30) and in vitro tracers confirmed the role of PER2 in metabolic flux

between normoxia and hypoxia by regulating glycolytic and lipid metabolism (Figure 2.31 – 2.34).

Next, our in vivo investigations into PER2 dependent adaptation to ischemia during IPC

uncovered the microRNA miR-21 to be cardioprotective from myocardial ischemia. In vitro

experiments supported the role of miR-21 in regulating glycolytic reliance and mitochondrial

function in hypoxia (Figure 3.6 – 3.9).

We next explored the implication of circadian disruption in myocardial IR-injury. We found

that the frequently used benzodiazepine, midazolam, downregulates cardiac PER2 and has

193 severe deleterious effects on cardiac tissue after IR (Figure 4.3, 4.4), supporting the hypothesis that circadian disruption is detrimental to the heart. To begin establishing a clinical application, we sought out to use circadian amplitude enhancement of PER2 as a therapeutic strategy in myocardial ischemia by light (Figure 5.1 – 5.3) and exercise (Figure 5.8 – 5.11). Our key findings demonstrate that our daylight regime is cardioprotective during myocardial ischemia through circadian amplitude enhancement in a PER2 dependent manner and may work in conjunction with light-elicited cardiac miR-21. Furthermore, our in vivo studies show that our daylight protocol mediates balance of metabolic pathways like glycolysis, the TCA cycle, oxidative phosphorylation, and the PPP (Figure 5.12, 5.13). This visual light perception is indeed necessary to achieve cardioprotection (Figure 5.5 – 5.7). Investigation into downstream signaling mechanisms of light exposure using our novel light-sensing cell line supports that light-elicited pathways are through cAMP signaling and downstream effectors (Figure 5.16 – 5.18). In addition, the specific PER2 amplitude inducer, nobiletin, is in fact cardioprotective and able to reverse the deleterious effects of midazolam on cardiac tissue after IR-injury (Figure 5.25 – 5.29).

Lastly, our targeted metabolomics screen revealed that light exposure in healthy human subjects targets metabolic pathways that are involved in balancing energy production (Figure 6.9

– 6.13) without negatively affecting sleep habits (Figure 6.14, 6.15). In addition, this light exposure has the capability to stabilize HIF1⍺ and increases miR-21 plasma levels (Figure 6.1 –

6.7, 6.17). In conclusion, our in vitro mechanistic findings and in vivo light-elicited cardioprotective strategies support our results in our healthy human population that together implicate light-elicited

PER2 and circadian amplitude enhancement as a potential preventative or therapeutic approach for patients with myocardial ischemia. While these studies provided evidence in support of our hypothesis and initial questions we sought to address, we are of course left with many potential directions in which to take our future studies.

194 9.2 Determine if Light Elicited Cardioprotection is Through Circulating Adenosine and the

A2B Receptor

One potential follow-up to our studies includes identifying whether in vivo light-elicited cardioprotection is indeed through the A2B receptor. Prior studies from our lab found PER2 dependent cardioprotection downstream of adenosine signaling, dependent on the A2B receptor, and through subsequent cAMP driven CREB binding to the PER2 promoter (71). Indeed, cAMP has been implicated as a master regulator of the transcriptional arm of the circadian feedback loop by regulating amplitude, phase, and period of the molecular clock (314). Studies from multiple other research groups identified a cyclical nature of 5’-AMP and adenosine in mouse blood (73), revealed that adenylyl cyclase exhibits circadian rhythmicity in human retinal pigment epithelium cells in vitro (358), and established that CD73 – the enzyme responsible for

extracellular adenosine generation – is upregulated in hypoxia for A2B cardioprotection during IPC

(70, 336, 359, 360). Considering our data from the in vitro light-sensing cell line supported that

light targets cAMP mediated pathways, PER2 induction, and glycolytic reliance (Figure 5.16 –

5.18), the next question is whether light-elicited cardioprotection in vivo is in fact through

adenosine signaling at the A2B receptor. Testing this hypothesis includes exposing A2B deficient

mice to light, myocardial IR-injury, and determining the extent of cardioprotection compared to

wildtype mice exposed to light. In addition, considering the CD73 gene is HIF1⍺ regulated (70),

questions remain as to whether CD73 is under direct control of PER2 in hypoxia, regulating HIF1⍺

binding in the promoter region.

9.3 Investigate the Potential Role of Peripheral Opsin Receptors in Cardioprotection

In addition to adenosine as a potential mechanism for external light stimuli signaling

peripheral organs (70, 73), another possible pathway is through opsins. The opsin family of

proteins were originally thought to be present only in light-sensing structures, fitting to their role

in detecting specific wavelengths of light. Unexpectedly, opsins have been found in tissues other

than retinal ganglion cells. Initial studies in the early 1990s was done using ultraviolet light on

195 rabbit aorta, corpus cavernosum, and pulmonary artery, resulting in increased cGMP, photorelaxation, and enhanced smooth muscle relaxation. However, the structures responsible for this light-elicited affect – now called opsins – were not yet discovered at this point in time (361,

362). In 2014, with many advancements in scientific technologies, another research group identified that the opsin OPN4 (melanopsin) is expressed in blood vessels and important for wavelength-specific photorelaxation (~430 – 460nm, approximately within range of blue light wavelength) in both ex vivo tail artery vasoreactivity and in vivo tail blood flow in an OPN4- dependent manner (363). Around the same time, similar studies found light-elicited vascular relaxation in porcine coronary arteries (364). Most recently, studies published since 2017 revealed light-elicited vasorelaxation in the pulmonary vasculature and smooth muscle (365). These studies also found light-elicited cAMP or cGMP associated with vasorelaxation (363, 365), which would support our model that light exposure targets PER2 function through cAMP-mediated downstream signaling. Together, these studies suggest that opsins, specifically OPN4, may be a component of environmental sensing within the cardiovascular system (363). However, the pathway for light signals being transmitted to internal peripheral organs is quite perplexing.

Nonetheless, the evolutionary conservation of opsins in the vasculature indicate a significant role of these receptors through mechanisms that have not yet been elucidated.

9.4 Determine PER2 Mediated TCA Cycle Function in Succinate Accumulation and ROS

Production During Reperfusion and the NAD+-Fumarate Reductase System

As early as the 1930s, before the TCA cycle was fully established, scientists hypothesized changes in TCA cycle function in hypoxia by means of succinate accumulation (366, 367), a finding that has now been supported in a variety of tissues (368-373). More recent studies found that this hypoxic accumulation of succinate is exported out of the mitochondria where it inhibits

PHDs to subsequently stabilize HIF1⍺ to promote its downstream hypoxic response (54, 368,

374-376). Succinate accumulation during hypoxia is thought to arise from fumarate reduction to succinate (termed the “NAD+-fumarate reductase system”) and ⍺-ketoglutarate oxidation.

196 Importantly, the NAD+-fumarate reductase system runs in the reverse direction than traditional

TCA cycle flux (368, 374, 375). Accumulation of succinate during hypoxia is exported out of the

mitochondria where it inhibits PHDs to subsequently stabilize HIF1⍺ (54, 376). In addition, succinate is most likely not metabolized until oxidative phosphorylation regains function upon re- oxygenation. However, reperfusion and therefore restoration of oxygen levels when mitochondrial membrane potential is high leads to a robust succinate catabolism subsequently driving an increase in ROS from the mitochondria (368, 377, 378). Regulation of the TCA cycle in IR is an important control point to minimize ROS. With respect to our studies and role we identified for

PER2 in regulating the TCA cycle, the next question to pursue is whether PER2 is involved in regulating directionality of the TCA cycle. One possible mechanism is that PER2 may be responsible for the succinate accumulation for HIF1⍺ stabilization by targeting the NAD+-fumarate reductase system for hypoxic adaptation. Additionally, HIF1 has a secondary target involved in selective autophagy to reduce ROS production in hypoxia, an imminent threat to the cell during hypoxia based on succinate metabolism upon re-oxygenation (368, 377, 378). BNIP3 is another downstream target gene of HIF1, which codes for a mitochondrial protein involved in selective mitochondrial autophagy (379) and contributes to our understanding of the HIF1 dependent response to hypoxic environments. Another potential mechanism in the PER2-HIF1⍺ connection is a PER2 dependent regulation of mitochondrial autophagy to mediate ROS damage to the cell during a hypoxic episode.

9.5 Investigate Time-of-Day-Dependent Role of Cardiovascular Drugs

Circadian biology – or chronobiology – should be given careful consideration for preventative or therapeutic treatment strategies since countless studies implicate circadian systems as the underlying mechanism in disease. Therefore, many researchers are investigating whether commonly prescribed drugs target circadian biology. Indeed, circadian systems have been identified for their involvement in drug absorption, distribution, metabolism, and excretion, together indicating pharmacological treatments may have better outcomes based on the time-of-

197 day the drug was administered (380). Recently, a chronobiology research group found that the

majority of best-selling drugs and World Health Organization essential medicines were identified

to have direct targets on proteins that maintain a circadian expression pattern, indicating a

potential benefit in time-of-day dosage in patients (225). In particular, over 100 drugs used for treatment of diseases of the cardiovascular system (like calcium channel blockers and β-blockers) target 136 genes that maintain robust oscillation in the atrial chamber, aorta, coronary artery, or tibial artery (381). Identification of more than 80% of protein-coding genes to be circadian nature brings to light the vast cellular functions under control of the circadian clockwork. In fact, 82.2% of the U.S Food and Drug Administration-approved drugs target genes with a cyclical expression

(224) and lastly, the circadian nuclear receptors REV-ERB and ROR are known drug targets, at least in animal models, for treating a variety of diseases (382). These studies emphasize the importance of better understanding pharmacological biological targets and implementing time-of- day treatments plans for patients.

9.6 Investigate Sex Differences in the Circadian Biology of Cardioprotection and PER2-

HIF1⍺ Dependent Adaptation to Hypoxia or Ischemia

We briefly touched on potential sex differences in our findings of similar light-elicited

cardioprotection in our mouse model of myocardial IR-injury (Figure 5.4). Clear sex differences

exist in the human population with respect to myocardial remodeling after myocardial IR-injury

(383). This observation initially prompted investigations into these differences to potentially

identify sex-specific adaptive mechanisms for cardioprotection (384-387). Previously, sex

differences in remodeling after MI were replicated in rats (388). In these studies, they found that

in the ischemic area, HIF was expressed in 33%-49% of nuclei of male rats with MI compared to

55%-82% of HIF in cell nuclei of female rats undergoing the same procedure (388). HIF protein

detection revealed a 60% greater HIF protein level in females than in males and downstream HIF

target genes (here, heme oxgenase and brain natriuretic peptide were considered) also increased

more significantly in females than in males (388). Sex-differences in cardiac remodeling after IR-

198 injury may also be due to sex-specific variation in apoptosis within the peri-infarct area: females exhibit better protection from ischemia-induced apoptosis compared to males (389). Furthermore, female rats exhibit significantly less decline in cardiac function compared to male rats undergoing

IR-injury, which is most likely due to male rats exhibiting an increase in left ventricular hypertrophy post IR-injury and blunted improvement in systolic function compared to female rats (390).

Additionally, sex differences in myocardial IR-injury and cardiac remodeling may be due to variation in estrogens: most women present with heart failure when circulating estrogen levels are lowest (post menopause) and studies using a rat model revealed estrogen receptor activation and signaling was cardioprotective from IR-injury and reduced apoptotic cardiac cell death (391).

Lastly, investigators are at the beginning stages of characterizing sex differences within the circadian system itself. For example, females may have the ability to phase shift in response to environmental stimuli more easily than males as suggested by differences in action potential thresholds in the SCN (392, 393) and androgens have been implicated in mediating responses to light by changing gene expression in the SCN: clock gene mutant mice exhibited sex differences in response to stress, sleep, and function of the hypothalamic-adrenal-pituitary axis (394). Our studies found similar light-elicited reduction in infarct size in females compared to male mice, but further investigations are warranted to determine whether there are long-term changes in light- elicited cardioprotection with regards to cardiac remodeling and whether this is sex-specific.

9.7 Identify miR-21 Targets for PER2 Dependent Cardioprotection

In our microRNA studies, miR-21 is one microRNA that has a strong involvement in cardioprotective pathways such as IPC (395) or metabolic pathways (396), but other identified microRNAs could be possible candidates for PER2 mediated downstream regulation (Table 3.1).

As such, studies on miR-22 found that exosomes, enriched with miR-22, were secreted by mesenchymal stem cells following cardiac IPC and mobilized to cardiomyocytes where they reduced their apoptosis due to ischemia. In addition, while miR-21 was found to play a key role in cardiac IPC (319, 395) or preventing apoptosis in cardiomyocytes (275), long-term elevation of

199 miR-21 may be also be detrimental to the organ by promoting the development of fibrosis in an acute cardiac allograft transplantation model (397) and right ventricle failure (398). MicroRNA targeting to optimize metabolic pathways is still in its infancy but opens the possibility for many mechanistic manipulations in disease prevention.

9.8 Elucidate the Peroxiredoxin Role in Circadian Mediated Metabolic Adaptation to Low

Oxygen Availability

Similar to the evolutionary significance of PAS domain proteins connecting light- and oxygen-sensing pathways, peroxiredoxin proteins are antioxidants functioning to remove ROS, present in all kingdoms of life, and oscillate in a circadian pattern. During the Great Oxygenation

Event, organisms adapted to a robust oxygen availability by acquiring cellular respiration mechanisms using oxygen for energy generation. Naturally occurring oscillations of tissue oxygenation (83) suggest that only cells with the ability to efficiently clear cellular ROS would most likely be the organisms to survive over time. Leaders in the peroxiredoxin field speculate that the circadian pattern in oxygen consumption may produce ROS in a cyclical manner. Additionally, peroxiredoxin oxidation cycles are in synchrony with daylight oscillation thereby indicating an evolutionarily conserved mechanistic explanation for ROS removal corresponding with cellular circadian biology (62). In support of this general hypothesis, a novel study in human red blood cells provided insight into the circadian nature of peroxiredoxins. Red blood cells do not have a nucleus and therefore are absent of DNA, transcription machinery, mitochondria, and the core clock proteins. Red blood cell ROS therefore most likely arises from autooxidation of hemoglobin

(399). In free-running conditions, these red blood cells have a peroxiredoxin redox cycle in an approximate 24-hour manner and were able to be entrained to environmental stimuli. These data indicate that even without the traditional transcription-translation feedback loop, there remains an endogenous circadian rhythm of about 24 hours (400). Therefore, there is an additional

connection to consider in our circadian-hypoxia link that may provide a master therapeutic target

200 for circadian mediated metabolic disorders, although a significant dedication to peroxiredoxin

research endeavors need to be further pursued before direct conclusions can be made.

9.9 Investigate the Role of Inflammation in Cardioprotection

Beyond the scope of the current studies, but providing a another potential avenue for our investigations into circadian disruption in human disease, includes the inflammatory response and immune function, which is diurnal in nature (401). Studies have found that the circadian nuclear receptors REV-ERBs regulate macrophage gene expression (402) and cytokine secretion has been found to have a circadian pattern (403). Additionally, a REV-ERB agonist, which may drive circadian amplitude enhancement, was found to improve post IR-injury cardiac remodeling by inhibiting cytokine production and regulating recruitment of inflammatory cells to the damaged tissue (404). Furthermore, studies on sepsis outcomes in rats during constant light or constant darkness conditions found both conditions to be detrimental (405). Indeed, light exposure probably needs to be adapted to the time-of-day where intense light late at night could be more

detrimental than beneficial. In fact, clinical studies in humans have found light at night to disrupt

circadian rhythms and to negatively affect metabolism (5, 406). Our findings would support the idea that an intact and robust circadian pattern of circadian proteins with high peaks, but also low troughs, would be most beneficial in a 24-hour window.

9.10 Outline a Circadian Protocol for Cardioprotection in ICU Patients to Regulate

Circadian Mechanisms to Reduce Incidence of MI

As a major contributing factor in the development of many diseases, circadian disruption

in ICU patients is prevalent and therefore these patients are at significantly increased risk of

worsening disease. Therefore, we hope our studies lead to therapeutic strategies to maintain

robust circadian amplitude and circadian rhythmicity. This is particularly emphasized in the

perioperative setting. One study found in a mouse model of circadian disruption after MI, there

was exacerbated subsequent cardiac remodeling (407), indicating that regulated light exposure

may be just as critical after surgery as it is during the days leading up to surgery. Furthermore, a

201 recent study identified a significant decrease in cardiac outcome post-surgery in those who

received this initial surgery in the morning compared to the afternoon (193). In this prospective cohort study, the incidence of major adverse cardiac event was determined for the 500 days following an initial surgery (aortic valve replacement) performed in either the morning or afternoon.

A key finding included a significant decrease in adverse cardiac events and lower troponin measurements in those who received the afternoon surgery (193). These findings led the research team to investigate the circadian and hypoxia-reoxygenation link in human and mouse myocardium ex-vivo. However, it is worth noting that this study claimed the circadian expression levels in humans were in opposition to those found in mice and that the PER2 gene is in phase with the REV-ERBA gene, a concept we disputed in our correspondence to the editor of the journal where this article was published (408). It is indeed well-established in the field that circadian biology functions independently of diurnal or nocturnal nature (212). These discrepancies indicate the need to carefully interpret data in mice versus humans and when investigating the link between circadian and hypoxic pathways (408).

9.11 Concluding Remarks

In conclusion, we uncovered underlying mechanisms of PER2 dependent metabolic adaptation to hypoxia or ischemia and circadian amplitude enhancement to develop a cardioprotective or therapeutic strategy. In a comprehensive and systems biology approach, we dissected light and hypoxia-elicited pathways in mice and humans from the cellular level to the whole animal. Our investigations unveil that circadian PER2 functions at the crossroad between light- and oxygen-sensing pathways through circadian amplitude enhancement for adaptation to low oxygen conditions like hypoxia or ischemia. Combined, we present a thorough understanding of mechanisms that target and manipulate metabolic pathways for round-the-clock cardioprotection from conditions like myocardial ischemia.

202 REFERENCES

1. Zerkle AL, Poulton SW, Newton RJ, Mettam C, Claire MW, Bekker A, et al. Onset of the aerobic nitrogen cycle during the Great Oxidation Event. Nature. 2017;542(7642):465-7.

2. Gaudana SB, Krishnakumar S, Alagesan S, Digmurti MG, Viswanathan GA, Chetty M, et al. Rhythmic and sustained oscillations in metabolism and gene expression of Cyanothece sp. ATCC 51142 under constant light. Front Microbiol. 2013;4:374.

3. Saha R, Liu D, Hoynes-O'Connor A, Liberton M, Yu J, Bhattacharyya-Pakrasi M, et al. Diurnal Regulation of Cellular Processes in the Cyanobacterium Synechocystis sp. Strain PCC 6803: Insights from Transcriptomic, Fluxomic, and Physiological Analyses. MBio. 2016;7(3).

4. Brainard J, Gobel M, Bartels K, Scott B, Koeppen M, and Eckle T. Circadian rhythms in anesthesia and critical care medicine: potential importance of circadian disruptions. Seminars in cardiothoracic and vascular anesthesia. 2015;19(1):49-60.

5. Brainard J, Gobel M, Scott B, Koeppen M, and Eckle T. Health implications of disrupted circadian rhythms and the potential for daylight as therapy. Anesthesiology. 2015;122(5):1170-5.

6. Jakubcakova V, Oster H, Tamanini F, Cadenas C, Leitges M, van der Horst GT, et al. Light entrainment of the mammalian circadian clock by a PRKCA-dependent posttranslational mechanism. Neuron. 2007;54(5):831-43.

7. Takahashi JS, Hong HK, Ko CH, and McDearmon EL. The genetics of mammalian circadian order and disorder: implications for physiology and disease. Nature reviews Genetics. 2008;9(10):764-75.

8. Guler AD, Ecker JL, Lall GS, Haq S, Altimus CM, Liao HW, et al. Melanopsin cells are the principal conduits for rod-cone input to non-image-forming vision. Nature. 2008;453(7191):102-5.

9. Pilorz V, Tam SK, Hughes S, Pothecary CA, Jagannath A, Hankins MW, et al. Melanopsin Regulates Both Sleep-Promoting and Arousal-Promoting Responses to Light. PLoS Biol. 2016;14(6):e1002482.

10. Lewy AJ, Wehr TA, Goodwin FK, Newsome DA, and Markey SP. Light suppresses melatonin secretion in humans. Science (New York, NY). 1980;210(4475):1267-9.

11. Toh KL, Jones CR, He Y, Eide EJ, Hinz WA, Virshup DM, et al. An hPer2 phosphorylation site mutation in familial advanced sleep phase syndrome. Science (New York, NY). 2001;291(5506):1040-3.

12. Ptacek LJ, Jones CR, and Fu YH. Novel insights from genetic and molecular characterization of the human clock. Cold Spring Harb Symp Quant Biol. 2007;72:273-7.

13. He Y, Jones CR, Fujiki N, Xu Y, Guo B, Holder JL, Jr., et al. The transcriptional repressor DEC2 regulates sleep length in mammals. Science (New York, NY). 2009;325(5942):866-70.

203 14. Albrecht U. Circadian clocks and mood-related behaviors. Handb Exp Pharmacol. 2013(217):227-39.

15. Cuninkova L, and Brown SA. Peripheral circadian oscillators: interesting mechanisms and powerful tools. Ann N Y Acad Sci. 2008;1129:358-70.

16. van Amelsvoort LG, Schouten EG, and Kok FJ. Impact of one year of shift work on cardiovascular disease risk factors. J Occup Environ Med. 2004;46(7):699-706.

17. Kivimaki M, Virtanen M, Elovainio M, Vaananen A, Keltikangas-Jarvinen L, and Vahtera J. Prevalent cardiovascular disease, risk factors and selection out of shift work. Scand J Work Environ Health. 2006;32(3):204-8.

18. Harma M. Shift work and cardiovascular disease--from etiologic studies to prevention through scheduling. Scand J Work Environ Health. 2001;27(2):85-6.

19. Boggild H, and Knutsson A. Shift work, risk factors and cardiovascular disease. Scand J Work Environ Health. 1999;25(2):85-99.

20. Akerstedt T, Knutsson A, Alfredsson L, and Theorell T. Shift work and cardiovascular disease. Scand J Work Environ Health. 1984;10(6 Spec No):409-14.

21. Akerstedt T, and Knutsson A. Cardiovascular disease and shift work. Scand J Work Environ Health. 1997;23(4):241-2.

22. Tuchsen F, Hannerz H, and Burr H. A 12 year prospective study of circulatory disease among Danish shift workers. Occup Environ Med. 2006;63(7):451-5.

23. Staels B. When the Clock stops ticking, metabolic syndrome explodes. Nature medicine. 2006;12(1):54-5; discussion 5.

24. Muller JE, Stone PH, Turi ZG, Rutherford JD, Czeisler CA, Parker C, et al. Circadian variation in the frequency of onset of acute myocardial infarction. N Engl J Med. 1985;313(21):1315-22.

25. Suarez-Barrientos A, Lopez-Romero P, Vivas D, Castro-Ferreira F, Nunez-Gil I, Franco E, et al. Circadian variations of infarct size in acute myocardial infarction. Heart (British Cardiac Society). 2011;97(12):970-6.

26. Braunwald E. On circadian variation of myocardial reperfusion injury. Circ Res. 2012;110(1):6-7.

27. Cohen MC, Rohtla KM, Lavery CE, Muller JE, and Mittleman MA. Meta-analysis of the morning excess of acute myocardial infarction and sudden cardiac death. Am J Cardiol. 1997;79(11):1512-6.

28. Tofler GH, Brezinski D, Schafer AI, Czeisler CA, Rutherford JD, Willich SN, et al. Concurrent morning increase in platelet aggregability and the risk of myocardial infarction and sudden cardiac death. N Engl J Med. 1987;316(24):1514-8.

204 29. Jeyaraj D, Haldar SM, Wan X, McCauley MD, Ripperger JA, Hu K, et al. Circadian rhythms govern cardiac repolarization and arrhythmogenesis. Nature. 2012;483(7387):96-9.

30. Panza JA, Epstein SE, and Quyyumi AA. Circadian variation in vascular tone and its relation to alpha-sympathetic vasoconstrictor activity. N Engl J Med. 1991;325(14):986-90.

31. Eckle T, and Eltzschig HK. Toll-like receptor signaling during myocardial ischemia. Anesthesiology. 2011;114(3):490-2.

32. Silver AC, Arjona A, Walker WE, and Fikrig E. The circadian clock controls toll-like receptor 9-mediated innate and adaptive immunity. Immunity. 2012;36(2):251-61.

33. Kawakami N, Takatsuka N, and Shimizu H. Sleep disturbance and onset of type 2 diabetes. Diabetes Care. 2004;27(1):282-3.

34. Spiegel K, Tasali E, Leproult R, and Van Cauter E. Effects of poor and short sleep on glucose metabolism and obesity risk. Nat Rev Endocrinol. 2009;5(5):253-61.

35. Suwazono Y, Dochi M, Sakata K, Okubo Y, Oishi M, Tanaka K, et al. A longitudinal study on the effect of shift work on weight gain in male Japanese workers. Obesity (Silver Spring). 2008;16(8):1887-93.

36. Spiegel K, Leproult R, and Van Cauter E. Impact of sleep debt on metabolic and endocrine function. Lancet. 1999;354(9188):1435-9.

37. Taheri S, Lin L, Austin D, Young T, and Mignot E. Short sleep duration is associated with reduced leptin, elevated ghrelin, and increased body mass index. PLoS Med. 2004;1(3):e62.

38. Takahashi JS. Transcriptional architecture of the mammalian circadian clock. Nature reviews Genetics. 2017;18(3):164-79.

39. Lee C, Etchegaray JP, Cagampang FR, Loudon AS, and Reppert SM. Posttranslational mechanisms regulate the mammalian circadian clock. Cell. 2001;107(7):855-67.

40. Gallego M, and Virshup DM. Post-translational modifications regulate the ticking of the circadian clock. Nat Rev Mol Cell Biol. 2007;8(2):139-48.

41. Eng GWL, Edison, and Virshup DM. Site-specific phosphorylation of casein kinase 1 delta (CK1delta) regulates its activity towards the circadian regulator PER2. PLoS One. 2017;12(5):e0177834.

42. Schmutz I, Ripperger JA, Baeriswyl-Aebischer S, and Albrecht U. The mammalian clock component PERIOD2 coordinates circadian output by interaction with nuclear receptors. Genes Dev. 2010;24(4):345-57.

43. Welch RD, and Flaveny CA. REV-ERB and ROR: therapeutic targets for treating myopathies. Phys Biol. 2017;14(4):045002.

44. Semenza GL. Life with oxygen. Science (New York, NY). 2007;318(5847):62-4.

205 45. McIntosh BE, Hogenesch JB, and Bradfield CA. Mammalian Per-Arnt-Sim proteins in environmental adaptation. Annu Rev Physiol. 2010;72:625-45.

46. Semenza GL. Hypoxia and human disease-and the Journal of Molecular Medicine. J Mol Med (Berl). 2007;85(12):1293-4.

47. Semenza GL. HIF-1: mediator of physiological and pathophysiological responses to hypoxia. J Appl Physiol (1985). 2000;88(4):1474-80.

48. Liu W, Shen SM, Zhao XY, and Chen GQ. Targeted genes and interacting proteins of hypoxia inducible factor-1. International journal of biochemistry and molecular biology. 2012;3(2):165-78.

49. Semenza GL, Jiang BH, Leung SW, Passantino R, Concordet JP, Maire P, et al. Hypoxia response elements in the aldolase A, enolase 1, and lactate dehydrogenase A gene promoters contain essential binding sites for hypoxia-inducible factor 1. The Journal of biological chemistry. 1996;271(51):32529-37.

50. Wenger RH, and Gassmann M. Oxygen(es) and the hypoxia-inducible factor-1. Biol Chem. 1997;378(7):609-16.

51. Gu YZ, Hogenesch JB, and Bradfield CA. The PAS superfamily: sensors of environmental and developmental signals. Annu Rev Pharmacol Toxicol. 2000;40:519-61.

52. Semenza GL. Regulation of tissue perfusion in mammals by hypoxia-inducible factor 1. Exp Physiol. 2007;92(6):988-91.

53. Semenza GL. Oxygen-dependent regulation of mitochondrial respiration by hypoxia- inducible factor 1. Biochem J. 2007;405(1):1-9.

54. Semenza GL. Hypoxia-inducible factor 1 (HIF-1) pathway. Sci STKE. 2007;2007(407):cm8.

55. Hogenesch JB, Gu YZ, Jain S, and Bradfield CA. The basic-helix-loop-helix-PAS orphan MOP3 forms transcriptionally active complexes with circadian and hypoxia factors. Proceedings of the National Academy of Sciences of the United States of America. 1998;95(10):5474-9.

56. Huang LE, Arany Z, Livingston DM, and Bunn HF. Activation of hypoxia-inducible transcription factor depends primarily upon redox-sensitive stabilization of its alpha subunit. The Journal of biological chemistry. 1996;271(50):32253-9.

57. Salceda S, and Caro J. Hypoxia-inducible factor 1alpha (HIF-1alpha) protein is rapidly degraded by the ubiquitin-proteasome system under normoxic conditions. Its stabilization by hypoxia depends on redox-induced changes. The Journal of biological chemistry. 1997;272(36):22642-7.

58. Chandel NS, Maltepe E, Goldwasser E, Mathieu CE, Simon MC, and Schumacker PT. Mitochondrial reactive oxygen species trigger hypoxia-induced transcription. Proceedings of the National Academy of Sciences of the United States of America. 1998;95(20):11715-20.

206 59. Haase C, Yu L, Eisenbarth G, and Markholst H. Antigen-dependent immunotherapy of non-obese diabetic mice with immature dendritic cells. Clin Exp Immunol. 2010;160(3):331-9.

60. Haase VH. The sweet side of HIF. Kidney Int. 2010;78(1):10-3.

61. Isoe T, Makino Y, Mizumoto K, Sakagami H, Fujita Y, Honjo J, et al. High glucose activates HIF-1-mediated signal transduction in glomerular mesangial cells through a carbohydrate response element binding protein. Kidney Int. 2010;78(1):48-59.

62. Edgar RS, Green EW, Zhao Y, van Ooijen G, Olmedo M, Qin X, et al. Peroxiredoxins are conserved markers of circadian rhythms. Nature. 2012;485(7399):459-64.

63. Taylor BL, and Zhulin IB. PAS domains: internal sensors of oxygen, redox potential, and light. Microbiol Mol Biol Rev. 1999;63(2):479-506.

64. Button EL, Bersten DC, and Whitelaw ML. HIF has Biff - Crosstalk between HIF1a and the family of bHLH/PAS proteins. Experimental cell research. 2017;356(2):141-5.

65. Go AS, Mozaffarian D, Roger VL, Benjamin EJ, Berry JD, Blaha MJ, et al. Heart disease and stroke statistics--2014 update: a report from the American Heart Association. Circulation. 2014;129(3):e28-e292.

66. Berne RM. Cardiac nucleotides in hypoxia: possible role in regulation of coronary blood flow. Am J Physiol. 1963;204:317-22.

67. Dobson GP, Arsyad A, and Letson HL. The Adenosine Hypothesis Revisited: Modulation of Coupling between Myocardial Perfusion and Arterial Compliance. Front Physiol. 2017;8:824.

68. Murry CE, Jennings RB, and Reimer KA. Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation. 1986;74(5):1124-36.

69. Kohler D, Eckle T, Faigle M, Grenz A, Mittelbronn M, Laucher S, et al. CD39/ectonucleoside triphosphate diphosphohydrolase 1 provides myocardial protection during cardiac ischemia/reperfusion injury. Circulation. 2007;116(16):1784-94.

70. Eckle T, Krahn T, Grenz A, Kohler D, Mittelbronn M, Ledent C, et al. Cardioprotection by ecto-5'-nucleotidase (CD73) and A2B adenosine receptors. Circulation. 2007;115(12):1581-90.

71. Eckle T, Hartmann K, Bonney S, Reithel S, Mittelbronn M, Walker LA, et al. Adora2b- elicited Per2 stabilization promotes a HIF-dependent metabolic switch crucial for myocardial adaptation to ischemia. Nature medicine. 2012;18(5):774-82.

72. Bendova Z, and Sumova A. Photoperiodic regulation of PER1 and PER2 protein expression in rat peripheral tissues. Physiol Res. 2006;55(6):623-32.

73. Zhang J, Kaasik K, Blackburn MR, and Lee CC. Constant darkness is a circadian metabolic signal in mammals. Nature. 2006;439(7074):340-3.

207 74. Bonney S, Hughes K, Harter PN, Mittelbronn M, Walker L, and Eckle T. Cardiac period 2 in myocardial ischemia: clinical implications of a light dependent protein. The international journal of biochemistry & cell biology. 2013;45(3):667-71.

75. Cai Z, Zhong H, Bosch-Marce M, Fox-Talbot K, Wang L, Wei C, et al. Complete loss of ischaemic preconditioning-induced cardioprotection in mice with partial deficiency of HIF-1 alpha. Cardiovasc Res. 2008;77(3):463-70.

76. Eckle T, Kohler D, Lehmann R, El Kasmi K, and Eltzschig HK. Hypoxia-inducible factor-1 is central to cardioprotection: a new paradigm for ischemic preconditioning. Circulation. 2008;118(2):166-75.

77. Synnestvedt K, Furuta GT, Comerford KM, Louis N, Karhausen J, Eltzschig HK, et al. Ecto-5'-nucleotidase (CD73) regulation by hypoxia-inducible factor-1 mediates permeability changes in intestinal epithelia. J Clin Invest. 2002;110(7):993-1002.

78. Semenza GL. Hypoxia-inducible factor 1: regulator of mitochondrial metabolism and mediator of ischemic preconditioning. Biochimica et biophysica acta. 2011;1813(7):1263-8.

79. Peek CB, Levine DC, Cedernaes J, Taguchi A, Kobayashi Y, Tsai SJ, et al. Circadian Clock Interaction with HIF1alpha Mediates Oxygenic Metabolism and Anaerobic Glycolysis in Skeletal Muscle. Cell metabolism. 2017;25(1):86-92.

80. Wu Y, Tang D, Liu N, Xiong W, Huang H, Li Y, et al. Reciprocal Regulation between the Circadian Clock and Hypoxia Signaling at the Genome Level in Mammals. Cell metabolism. 2017;25(1):73-85.

81. Kobayashi M, Morinibu A, Koyasu S, Goto Y, Hiraoka M, and Harada H. A circadian clock gene, PER2, activates HIF-1 as an effector molecule for recruitment of HIF-1alpha to promoter regions of its downstream genes. FEBS J. 2017;284(22):3804-16.

82. Egg M, Koblitz L, Hirayama J, Schwerte T, Folterbauer C, Kurz A, et al. Linking oxygen to time: the bidirectional interaction between the hypoxic signaling pathway and the circadian clock. Chronobiol Int. 2013;30(4):510-29.

83. Adamovich Y, Ladeuix B, Golik M, Koeners MP, and Asher G. Rhythmic Oxygen Levels Reset Circadian Clocks through HIF1alpha. Cell metabolism. 2017;25(1):93-101.

84. Apstein CS, and Taegtmeyer H. Glucose-insulin-potassium in acute myocardial infarction: the time has come for a large, prospective trial. Circulation. 1997;96(4):1074-7.

85. Kessler G, and Friedman J. Metabolism of fatty acids and glucose. Circulation. 1998;98(13):1351.

86. O'Neill JS, and Feeney KA. Circadian redox and metabolic oscillations in mammalian systems. Antioxid Redox Signal. 2014;20(18):2966-81.

208 87. Aon MA, Cortassa S, Marban E, and O'Rourke B. Synchronized whole cell oscillations in mitochondrial metabolism triggered by a local release of reactive oxygen species in cardiac myocytes. The Journal of biological chemistry. 2003;278(45):44735-44.

88. Chandel NS, McClintock DS, Feliciano CE, Wood TM, Melendez JA, Rodriguez AM, et al. Reactive oxygen species generated at mitochondrial complex III stabilize hypoxia-inducible factor-1alpha during hypoxia: a mechanism of O2 sensing. The Journal of biological chemistry. 2000;275(33):25130-8.

89. Pinto AR, Ilinykh A, Ivey MJ, Kuwabara JT, D'Antoni ML, Debuque R, et al. Revisiting Cardiac Cellular Composition. Circ Res. 2016;118(3):400-9.

90. Zhou P, and Pu WT. Recounting Cardiac Cellular Composition. Circ Res. 2016;118(3):368-70.

91. Yang Q, He GW, Underwood MJ, and Yu CM. Cellular and molecular mechanisms of endothelial ischemia/reperfusion injury: perspectives and implications for postischemic myocardial protection. Am J Transl Res. 2016;8(2):765-77.

92. Singhal AK, Symons JD, Boudina S, Jaishy B, and Shiu YT. Role of Endothelial Cells in Myocardial Ischemia-Reperfusion Injury. Vasc Dis Prev. 2010;7:1-14.

93. Lefer AM, Tsao PS, Lefer DJ, and Ma XL. Role of endothelial dysfunction in the pathogenesis of reperfusion injury after myocardial ischemia. FASEB J. 1991;5(7):2029-34.

94. Lefer AM, and Lefer DJ. Endothelial dysfunction in myocardial ischemia and reperfusion: role of oxygen-derived free radicals. Basic Res Cardiol. 1991;86 Suppl 2:109-16.

95. Davidson SM, and Duchen MR. Endothelial mitochondria: contributing to vascular function and disease. Circ Res. 2007;100(8):1128-41.

96. Quintero M, Colombo SL, Godfrey A, and Moncada S. Mitochondria as signaling organelles in the vascular endothelium. Proceedings of the National Academy of Sciences of the United States of America. 2006;103(14):5379-84.

97. Ricci JE, Waterhouse N, and Green DR. Mitochondrial functions during cell death, a complex (I-V) dilemma. Cell Death Differ. 2003;10(5):488-92.

98. Mattson MP, and Kroemer G. Mitochondria in cell death: novel targets for neuroprotection and cardioprotection. Trends Mol Med. 2003;9(5):196-205.

99. Bierhansl L, Conradi LC, Treps L, Dewerchin M, and Carmeliet P. Central Role of Metabolism in Endothelial Cell Function and Vascular Disease. Physiology (Bethesda). 2017;32(2):126-40.

100. Eelen G, de Zeeuw P, Simons M, and Carmeliet P. Endothelial cell metabolism in normal and diseased vasculature. Circ Res. 2015;116(7):1231-44.

209 101. Fedalen PA, Piacentino V, 3rd, Jeevanandam V, Fisher C, Greene J, Margulies KB, et al. Pharmacologic pre-conditioning and controlled reperfusion prevent ischemia-reperfusion injury after 30 minutes of hypoxia/ischemia in porcine hearts. J Heart Lung Transplant. 2003;22(11):1234-44.

102. Mohara J, Aguilera I, Goldman BI, Fisher CA, Gaughan JP, Libonati JR, et al. Effects of nutrient and hemoglobin enriched cell free perfusates upon ex vivo isolated rat heart preparation. ASAIO J. 2005;51(3):288-95.

103. Solaini G, Baracca A, Lenaz G, and Sgarbi G. Hypoxia and mitochondrial oxidative metabolism. Biochimica et biophysica acta. 2010;1797(6-7):1171-7.

104. Kim JW, Tchernyshyov I, Semenza GL, and Dang CV. HIF-1-mediated expression of pyruvate dehydrogenase kinase: a metabolic switch required for cellular adaptation to hypoxia. Cell metabolism. 2006;3(3):177-85.

105. Fukuda R, Zhang H, Kim JW, Shimoda L, Dang CV, and Semenza GL. HIF-1 regulates cytochrome oxidase subunits to optimize efficiency of respiration in hypoxic cells. Cell. 2007;129(1):111-22.

106. West AC, Smith L, Ray DW, Loudon ASI, Brown TM, and Bechtold DA. Misalignment with the external light environment drives metabolic and cardiac dysfunction. Nat Commun. 2017;8(1):417.

107. Krishnaiah SY, Wu G, Altman BJ, Growe J, Rhoades SD, Coldren F, et al. Clock Regulation of Metabolites Reveals Coupling between Transcription and Metabolism. Cell metabolism. 2017;25(5):1206.

108. Putker M, Crosby P, Feeney KA, Hoyle NP, Costa ASH, Gaude E, et al. Mammalian Circadian Period, But Not Phase and Amplitude, Is Robust Against Redox and Metabolic Perturbations. Antioxid Redox Signal. 2018;28(7):507-20.

109. Frye RA. Phylogenetic classification of prokaryotic and eukaryotic Sir2-like proteins. Biochem Biophys Res Commun. 2000;273(2):793-8.

110. Etchegaray JP, Lee C, Wade PA, and Reppert SM. Rhythmic histone acetylation underlies transcription in the mammalian circadian clock. Nature. 2003;421(6919):177-82.

111. Kincaid B, and Bossy-Wetzel E. Forever young: SIRT3 a shield against mitochondrial meltdown, aging, and neurodegeneration. Front Aging Neurosci. 2013;5:48.

112. Mauvoisin D, Atger F, Dayon L, Nunez Galindo A, Wang J, Martin E, et al. Circadian and Feeding Rhythms Orchestrate the Diurnal Liver Acetylome. Cell Rep. 2017;20(7):1729-43.

113. Peek CB, Affinati AH, Ramsey KM, Kuo HY, Yu W, Sena LA, et al. Circadian clock NAD+ cycle drives mitochondrial oxidative metabolism in mice. Science (New York, NY). 2013;342(6158):1243417.

210 114. Ramsey KM, Yoshino J, Brace CS, Abrassart D, Kobayashi Y, Marcheva B, et al. Circadian clock feedback cycle through NAMPT-mediated NAD+ biosynthesis. Science (New York, NY). 2009;324(5927):651-4.

115. Asher G, Gatfield D, Stratmann M, Reinke H, Dibner C, Kreppel F, et al. SIRT1 regulates circadian clock gene expression through PER2 deacetylation. Cell. 2008;134(2):317-28.

116. Lombard DB, Alt FW, Cheng HL, Bunkenborg J, Streeper RS, Mostoslavsky R, et al. Mammalian Sir2 homolog SIRT3 regulates global mitochondrial lysine acetylation. Mol Cell Biol. 2007;27(24):8807-14.

117. Ahn BH, Kim HS, Song S, Lee IH, Liu J, Vassilopoulos A, et al. A role for the mitochondrial deacetylase Sirt3 in regulating energy homeostasis. Proceedings of the National Academy of Sciences of the United States of America. 2008;105(38):14447-52.

118. Gurd BJ, Holloway GP, Yoshida Y, and Bonen A. In mammalian muscle, SIRT3 is present in mitochondria and not in the nucleus; and SIRT3 is upregulated by chronic muscle contraction in an adenosine monophosphate-activated protein kinase-independent manner. Metabolism. 2012;61(5):733-41.

119. Cui XX, Li X, Dong SY, Guo YJ, Liu T, and Wu YC. SIRT3 deacetylated and increased citrate synthase activity in PD model. Biochem Biophys Res Commun. 2017;484(4):767-73.

120. Sheng S, Kang Y, Guo Y, Pu Q, Cai M, and Tu Z. Overexpression of Sirt3 inhibits lipid accumulation in macrophages through mitochondrial IDH2 deacetylation. Int J Clin Exp Pathol. 2015;8(8):9196-201.

121. Yu W, Dittenhafer-Reed KE, and Denu JM. SIRT3 protein deacetylates isocitrate dehydrogenase 2 (IDH2) and regulates mitochondrial redox status. The Journal of biological chemistry. 2012;287(17):14078-86.

122. Finley LW, Haas W, Desquiret-Dumas V, Wallace DC, Procaccio V, Gygi SP, et al. Succinate dehydrogenase is a direct target of sirtuin 3 deacetylase activity. PLoS One. 2011;6(8):e23295.

123. Cimen H, Han MJ, Yang Y, Tong Q, Koc H, and Koc EC. Regulation of succinate dehydrogenase activity by SIRT3 in mammalian mitochondria. Biochemistry. 2010;49(2):304-11.

124. Porter GA, Urciuoli WR, Brookes PS, and Nadtochiy SM. SIRT3 deficiency exacerbates ischemia-reperfusion injury: implication for aged hearts. American journal of physiology Heart and circulatory physiology. 2014;306(12):H1602-9.

125. Hirschey MD, Shimazu T, Goetzman E, Jing E, Schwer B, Lombard DB, et al. SIRT3 regulates mitochondrial fatty-acid oxidation by reversible enzyme deacetylation. Nature. 2010;464(7285):121-5.

126. Hirschey MD, Shimazu T, Jing E, Grueter CA, Collins AM, Aouizerat B, et al. SIRT3 deficiency and mitochondrial protein hyperacetylation accelerate the development of the metabolic syndrome. Mol Cell. 2011;44(2):177-90.

211 127. Sundaresan NR, Samant SA, Pillai VB, Rajamohan SB, and Gupta MP. SIRT3 is a stress- responsive deacetylase in cardiomyocytes that protects cells from stress-mediated cell death by deacetylation of Ku70. Mol Cell Biol. 2008;28(20):6384-401.

128. Bae K, Jin X, Maywood ES, Hastings MH, Reppert SM, and Weaver DR. Differential functions of mPer1, mPer2, and mPer3 in the SCN circadian clock. Neuron. 2001;30(2):525-36.

129. Grimaldi B, Bellet MM, Katada S, Astarita G, Hirayama J, Amin RH, et al. PER2 controls lipid metabolism by direct regulation of PPARgamma. Cell metabolism. 2010;12(5):509-20.

130. Grimaldi B, Nakahata Y, Sahar S, Kaluzova M, Gauthier D, Pham K, et al. Chromatin remodeling and circadian control: master regulator CLOCK is an enzyme. Cold Spring Harb Symp Quant Biol. 2007;72:105-12.

131. Yin L, Wu N, Curtin JC, Qatanani M, Szwergold NR, Reid RA, et al. Rev-erbalpha, a heme sensor that coordinates metabolic and circadian pathways. Science (New York, NY). 2007;318(5857):1786-9.

132. Le Martelot G, Claudel T, Gatfield D, Schaad O, Kornmann B, Lo Sasso G, et al. REV- ERBalpha participates in circadian SREBP signaling and bile acid homeostasis. PLoS Biol. 2009;7(9):e1000181.

133. Panda S, Antoch MP, Miller BH, Su AI, Schook AB, Straume M, et al. Coordinated transcription of key pathways in the mouse by the circadian clock. Cell. 2002;109(3):307-20.

134. Storch KF, Lipan O, Leykin I, Viswanathan N, Davis FC, Wong WH, et al. Extensive and divergent circadian gene expression in liver and heart. Nature. 2002;417(6884):78-83.

135. Eckle T, Koeppen M, and Eltzschig H. Use of a hanging weight system for coronary artery occlusion in mice. J Vis Exp. 2011(50).

136. Bonney S, Kominsky D, Brodsky K, Eltzschig H, Walker L, and Eckle T. Cardiac Per2 functions as novel link between fatty acid metabolism and myocardial inflammation during ischemia and reperfusion injury of the heart. PLoS One. 2013;8(8):e71493.

137. Luo X, Zhang Y, Ruan X, Jiang X, Zhu L, Wang X, et al. Fasting-induced protein phosphatase 1 regulatory subunit contributes to postprandial blood glucose homeostasis via regulation of hepatic glycogenesis. Diabetes. 2011;60(5):1435-45.

138. Jaswal JS, Keung W, Wang W, Ussher JR, and Lopaschuk GD. Targeting fatty acid and carbohydrate oxidation--a novel therapeutic intervention in the ischemic and failing heart. Biochimica et biophysica acta. 2011;1813(7):1333-50.

139. Lopaschuk GD, McNeil GF, and McVeigh JJ. Glucose oxidation is stimulated in reperfused ischemic hearts with the carnitine palmitoyltransferase 1 inhibitor, Etomoxir. Mol Cell Biochem. 1989;88(1-2):175-9.

212 140. Finley LW, Carracedo A, Lee J, Souza A, Egia A, Zhang J, et al. SIRT3 opposes reprogramming of cancer cell metabolism through HIF1alpha destabilization. Cancer Cell. 2011;19(3):416-28.

141. Eckel-Mahan K, and Sassone-Corsi P. Metabolism and the circadian clock converge. Physiol Rev. 2013;93(1):107-35.

142. Oyama Y, Bartman CM, Gile J, and Eckle T. Circadian MicroRNAs in Cardioprotection. Curr Pharm Des. 2017;23(25):3723-30.

143. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, et al. The sequence of the human genome. Science (New York, NY). 2001;291(5507):1304-51.

144. Barwari T, Joshi A, and Mayr M. MicroRNAs in Cardiovascular Disease. Journal of the American College of Cardiology. 2016;68(23):2577-84.

145. Biggar KK, and Storey KB. The emerging roles of microRNAs in the molecular responses of metabolic rate depression. J Mol Cell Biol. 2011;3(3):167-75.

146. Bronze-da-Rocha E. MicroRNAs expression profiles in cardiovascular diseases. Biomed Res Int. 2014;2014:985408.

147. Condorelli G, Latronico MV, and Cavarretta E. microRNAs in cardiovascular diseases: current knowledge and the road ahead. Journal of the American College of Cardiology. 2014;63(21):2177-87.

148. De Rosa S, Curcio A, and Indolfi C. Emerging role of microRNAs in cardiovascular diseases. Circ J. 2014;78(3):567-75.

149. De Rosa S, and Indolfi C. Circulating microRNAs as Biomarkers in Cardiovascular Diseases. EXS. 2015;106:139-49.

150. Dimmeler S, and Zeiher AM. Circulating microRNAs: novel biomarkers for cardiovascular diseases? Eur Heart J. 2010;31(22):2705-7.

151. Fichtlscherer S, De Rosa S, Fox H, Schwietz T, Fischer A, Liebetrau C, et al. Circulating microRNAs in patients with coronary artery disease. Circ Res. 2010;107(5):677-84.

152. Ohtani K, and Dimmeler S. Control of cardiovascular differentiation by microRNAs. Basic Res Cardiol. 2011;106(1):5-11.

153. Ferrari D, Bianchi N, Eltzschig HK, and Gambari R. MicroRNAs Modulate the Purinergic Signaling Network. Trends Mol Med. 2016;22(10):905-18.

154. Gurha P. MicroRNAs in cardiovascular disease. Curr Opin Cardiol. 2016;31(3):249-54.

213 155. van Rooij E, Sutherland LB, Liu N, Williams AH, McAnally J, Gerard RD, et al. A signature pattern of stress-responsive microRNAs that can evoke cardiac hypertrophy and heart failure. Proceedings of the National Academy of Sciences of the United States of America. 2006;103(48):18255-60.

156. Thum T, Gross C, Fiedler J, Fischer T, Kissler S, Bussen M, et al. MicroRNA-21 contributes to myocardial disease by stimulating MAP kinase signalling in fibroblasts. Nature. 2008;456(7224):980-4.

157. Patrick DM, Montgomery RL, Qi X, Obad S, Kauppinen S, Hill JA, et al. Stress-dependent cardiac remodeling occurs in the absence of microRNA-21 in mice. J Clin Invest. 2010;120(11):3912-6.

158. Olson JM, Yan Y, Bai X, Ge ZD, Liang M, Kriegel AJ, et al. Up-regulation of microRNA-21 mediates isoflurane-induced protection of cardiomyocytes. Anesthesiology. 2015;122(4):795- 805.

159. Bonney S, Hughes K, and Eckle T. Anesthetic cardioprotection: the role of adenosine. Curr Pharm Des. 2014;20(36):5690-5.

160. Qiao S, Olson JM, Paterson M, Yan Y, Zaja I, Liu Y, et al. MicroRNA-21 Mediates Isoflurane-induced Cardioprotection against Ischemia-Reperfusion Injury via Akt/Nitric Oxide Synthase/Mitochondrial Permeability Transition Pore Pathway. Anesthesiology. 2015;123(4):786- 98.

161. Cheng Y, Zhu P, Yang J, Liu X, Dong S, Wang X, et al. Ischaemic preconditioning- regulated miR-21 protects heart against ischaemia/reperfusion injury via anti-apoptosis through its target PDCD4. Cardiovasc Res. 2010;87(3):431-9.

162. Frank A, Bonney M, Bonney S, Weitzel L, Koeppen M, and Eckle T. Myocardial ischemia reperfusion injury: from basic science to clinical bedside. Seminars in cardiothoracic and vascular anesthesia. 2012;16(3):123-32.

163. Hansen KF, Sakamoto K, and Obrietan K. MicroRNAs: a potential interface between the circadian clock and human health. Genome Med. 2011;3(2):10.

164. Sire C, Moreno AB, Garcia-Chapa M, Lopez-Moya JJ, and San Segundo B. Diurnal oscillation in the accumulation of Arabidopsis microRNAs, miR167, miR168, miR171 and miR398. FEBS Lett. 2009;583(6):1039-44.

165. Cheng HY, Papp JW, Varlamova O, Dziema H, Russell B, Curfman JP, et al. microRNA modulation of circadian-clock period and entrainment. Neuron. 2007;54(5):813-29.

166. Impey S, McCorkle SR, Cha-Molstad H, Dwyer JM, Yochum GS, Boss JM, et al. Defining the CREB regulon: a genome-wide analysis of transcription factor regulatory regions. Cell. 2004;119(7):1041-54.

214 167. Gatfield D, Le Martelot G, Vejnar CE, Gerlach D, Schaad O, Fleury-Olela F, et al. Integration of microRNA miR-122 in hepatic circadian gene expression. Genes Dev. 2009;23(11):1313-26.

168. Tan X, Zhang P, Zhou L, Yin B, Pan H, and Peng X. Clock-controlled mir-142-3p can target its activator, Bmal1. BMC Mol Biol. 2012;13:27.

169. Shende VR, Goldrick MM, Ramani S, and Earnest DJ. Expression and rhythmic modulation of circulating microRNAs targeting the clock gene Bmal1 in mice. PLoS One. 2011;6(7):e22586.

170. Curtis AM, Fagundes CT, Yang G, Palsson-McDermott EM, Wochal P, McGettrick AF, et al. Circadian control of innate immunity in macrophages by miR-155 targeting Bmal1. Proceedings of the National Academy of Sciences of the United States of America. 2015;112(23):7231-6.

171. Chen R, D'Alessandro M, and Lee C. miRNAs are required for generating a time delay critical for the circadian oscillator. Current biology : CB. 2013;23(20):1959-68.

172. Kantermann T, Juda M, Merrow M, and Roenneberg T. The human circadian clock's seasonal adjustment is disrupted by daylight saving time. Current biology : CB. 2007;17(22):1996- 2000.

173. Roenneberg T, and Aschoff J. Annual rhythm of human reproduction: I. Biology, sociology, or both? J Biol Rhythms. 1990;5(3):195-216.

174. Roenneberg T, and Aschoff J. Annual rhythm of human reproduction: II. Environmental correlations. J Biol Rhythms. 1990;5(3):217-39.

175. Stothard ER, McHill AW, Depner CM, Birks BR, Moehlman TM, Ritchie HK, et al. Circadian Entrainment to the Natural Light-Dark Cycle across Seasons and the Weekend. Current biology : CB. 2017;27(4):508-13.

176. Leocadio-Miguel MA, Louzada FM, Duarte LL, Areas RP, Alam M, Freire MV, et al. Latitudinal cline of chronotype. Sci Rep. 2017;7(1):5437.

177. Schroeder AM, and Colwell CS. How to fix a broken clock. Trends Pharmacol Sci. 2013;34(11):605-19.

178. Haus E, and Smolensky M. Biological clocks and shift work: circadian dysregulation and potential long-term effects. Cancer Causes Control. 2006;17(4):489-500.

179. Mosendane T, Mosendane T, and Raal FJ. Shift work and its effects on the cardiovascular system. Cardiovasc J Afr. 2008;19(4):210-5.

180. Segawa K, Nakazawa S, Tsukamoto Y, Kurita Y, Goto H, Fukui A, et al. Peptic ulcer is prevalent among shift workers. Dig Dis Sci. 1987;32(5):449-53.

215 181. Davis S, Mirick DK, and Stevens RG. Night shift work, light at night, and risk of breast cancer. J Natl Cancer Inst. 2001;93(20):1557-62.

182. Persson HE, and Svanborg E. Sleep deprivation worsens obstructive sleep apnea. Comparison between diurnal and nocturnal polysomnography. Chest. 1996;109(3):645-50.

183. Depner CM, Stothard ER, and Wright KP, Jr. Metabolic consequences of sleep and circadian disorders. Current diabetes reports. 2014;14(7):507.

184. Spencer FA, Goldberg RJ, Becker RC, and Gore JM. Seasonal distribution of acute myocardial infarction in the second National Registry of Myocardial Infarction. Journal of the American College of Cardiology. 1998;31(6):1226-33.

185. Zhang XW, Tan ZJ, Li YL, Wang B, Yu A, and Zhang GQ. [A study on yearly and daily circadian rhythm of cardiovascular events]. Zhonghua Nei Ke Za Zhi. 2009;48(10):818-20.

186. Schloss MJ, Horckmans M, Nitz K, Duchene J, Drechsler M, Bidzhekov K, et al. The time- of-day of myocardial infarction onset affects healing through oscillations in cardiac neutrophil recruitment. EMBO Mol Med. 2016;8(8):937-48.

187. Lucassen EA, Coomans CP, van Putten M, de Kreij SR, van Genugten JH, Sutorius RP, et al. Environmental 24-hr Cycles Are Essential for Health. Current biology : CB. 2016;26(14):1843-53.

188. Korompeli A, Muurlink O, Kavrochorianou N, Katsoulas T, Fildissis G, and Baltopoulos G. Circadian disruption of ICU patients: A review of pathways, expression, and interventions. J Crit Care. 2017;38:269-77.

189. Danielson SJ, Rappaport CA, Loher MK, and Gehlbach BK. Looking for light in the din: An examination of the circadian-disrupting properties of a medical intensive care unit. Intensive Crit Care Nurs. 2018.

190. Smolensky MH, Hermida RC, Reinberg A, Sackett-Lundeen L, and Portaluppi F. Circadian disruption: New clinical perspective of disease pathology and basis for chronotherapeutic intervention. Chronobiol Int. 2016;33(8):1101-19.

191. Durrington HJ, Clark R, Greer R, Martial FP, Blaikley J, Dark P, et al. 'In a dark place, we find ourselves': light intensity in critical care units. Intensive Care Med Exp. 2017;5(1):9.

192. Ruggieri AJ, Levy RJ, and Deutschman CS. Mitochondrial dysfunction and resuscitation in sepsis. Crit Care Clin. 2010;26(3):567-75, x-xi.

193. Montaigne D, Marechal X, Modine T, Coisne A, Mouton S, Fayad G, et al. Daytime variation of perioperative myocardial injury in cardiac surgery and its prevention by Rev-Erbalpha antagonism: a single-centre propensity-matched cohort study and a randomised study. Lancet. 2018;391(10115):59-69.

216 194. Dauchy RT, Wren-Dail MA, Hoffman AE, Hanifin JP, Warfield B, Brainard GC, et al. Effects of Daytime Exposure to Light from Blue-Enriched Light-Emitting Diodes on the Nighttime Melatonin Amplitude and Circadian Regulation of Rodent Metabolism and Physiology. Comp Med. 2016;66(5):373-83.

195. Sudo M, Sasahara K, Moriya T, Akiyama M, Hamada T, and Shibata S. Constant light housing attenuates circadian rhythms of mPer2 mRNA and mPER2 protein expression in the suprachiasmatic nucleus of mice. Neuroscience. 2003;121(2):493-9.

196. Dispersyn G, Pain L, and Touitou Y. Circadian disruption of body core temperature and rest-activity rhythms after general (propofol) anesthesia in rats. Anesthesiology. 2009;110(6):1305-15.

197. Lonardo NW, Mone MC, Nirula R, Kimball EJ, Ludwig K, Zhou X, et al. Propofol is associated with favorable outcomes compared with benzodiazepines in ventilated intensive care unit patients. Am J Respir Crit Care Med. 2014;189(11):1383-94.

198. Rivo J, Raphael J, Drenger B, Berenshtein E, Chevion M, and Gozal Y. Flumazenil mimics whereas midazolam abolishes ischemic preconditioning in a rabbit heart model of ischemia- reperfusion. Anesthesiology. 2006;105(1):65-71.

199. Matsuo I, Iijima N, Takumi K, Higo S, Aikawa S, Anzai M, et al. Characterization of sevoflurane effects on Per2 expression using ex vivo bioluminescence imaging of the suprachiasmatic nucleus in transgenic rats. Neurosci Res. 2016;107:30-7.

200. Liu C, and Reppert SM. GABA synchronizes clock cells within the suprachiasmatic circadian clock. Neuron. 2000;25(1):123-8.

201. Wagner S, Castel M, Gainer H, and Yarom Y. GABA in the mammalian suprachiasmatic nucleus and its role in diurnal rhythmicity. Nature. 1997;387(6633):598-603.

202. Colwell CS. Circadian rhythms. Time to get excited by GABA. Nature. 1997;387(6633):554-5.

203. Ritchie HK, Stothard ER, and Wright KP. Entrainment of the Human Circadian Clock to the Light-Dark Cycle and its Impact on Patients in the ICU and Nursing Home Settings. Curr Pharm Des. 2015;21(24):3438-42.

204. Arble DM, Ramsey KM, Bass J, and Turek FW. Circadian disruption and metabolic disease: findings from animal models. Best Pract Res Clin Endocrinol Metab. 2010;24(5):785- 800.

205. Foster RG, Peirson SN, Wulff K, Winnebeck E, Vetter C, and Roenneberg T. Sleep and circadian rhythm disruption in social jetlag and mental illness. Prog Mol Biol Transl Sci. 2013;119:325-46.

206. Hatori M, Vollmers C, Zarrinpar A, DiTacchio L, Bushong EA, Gill S, et al. Time-restricted feeding without reducing caloric intake prevents metabolic diseases in mice fed a high-fat diet. Cell metabolism. 2012;15(6):848-60.

217 207. Noyan H, El-Mounayri O, Isserlin R, Arab S, Momen A, Cheng HS, et al. Cardioprotective Signature of Short-Term Caloric Restriction. PLoS One. 2015;10(6):e0130658.

208. Schroeder AM, Truong D, Loh DH, Jordan MC, Roos KP, and Colwell CS. Voluntary scheduled exercise alters diurnal rhythms of behaviour, physiology and gene expression in wild- type and vasoactive intestinal peptide-deficient mice. The Journal of physiology. 2012;590(23):6213-26.

209. Martino TA, Tata N, Belsham DD, Chalmers J, Straume M, Lee P, et al. Disturbed diurnal rhythm alters gene expression and exacerbates cardiovascular disease with rescue by resynchronization. Hypertension. 2007;49(5):1104-13.

210. Vosko AM, Hagenauer MH, Hummer DL, and Lee TM. Period gene expression in the diurnal degu (Octodon degus) differs from the nocturnal laboratory rat (Rattus norvegicus). Am J Physiol Regul Integr Comp Physiol. 2009;296(2):R353-61.

211. Smale L, Lee T, and Nunez AA. Mammalian diurnality: some facts and gaps. J Biol Rhythms. 2003;18(5):356-66.

212. Kronfeld-Schor N, Bloch G, and Schwartz WJ. Animal clocks: when science meets nature. Proc Biol Sci. 2013;280(1765):20131354.

213. Laura Smale AAN, b and Michael D. Schwartzb. Rhythms in a Diurnal Brain. Biological Rhythm Research. 2008;39(3).

214. Mahoney M, Bult A, and Smale L. Phase response curve and light-induced fos expression in the suprachiasmatic nucleus and adjacent hypothalamus of Arvicanthis niloticus. J Biol Rhythms. 2001;16(2):149-62.

215. Pevet P. Melatonin in animal models. Dialogues Clin Neurosci. 2003;5(4):343-52.

216. Underwood H, and Goldman BD. Vertebrate circadian and photoperiodic systems: role of the pineal gland and melatonin. J Biol Rhythms. 1987;2(4):279-315.

217. Rusak B, and Yu GD. Regulation of melatonin-sensitivity and firing-rate rhythms of hamster suprachiasmatic nucleus neurons: pinealectomy effects. Brain Res. 1993;602(2):200-4.

218. Hatonen T, Alila A, and Laakso ML. Exogenous melatonin fails to counteract the light- induced phase delay of human melatonin rhythm. Brain Res. 1996;710(1-2):125-30.

219. Buijs RM, and Kalsbeek A. Hypothalamic integration of central and peripheral clocks. Nat Rev Neurosci. 2001;2(7):521-6.

220. Kramer A, Yang FC, Snodgrass P, Li X, Scammell TE, Davis FC, et al. Regulation of daily locomotor activity and sleep by hypothalamic EGF receptor signaling. Science (New York, NY). 2001;294(5551):2511-5.

218 221. Roseboom PH, Namboodiri MA, Zimonjic DB, Popescu NC, Rodriguez IR, Gastel JA, et al. Natural melatonin 'knockdown' in C57BL/6J mice: rare mechanism truncates serotonin N- acetyltransferase. Brain Res Mol Brain Res. 1998;63(1):189-97.

222. Ebihara S, Hudson DJ, Marks T, and Menaker M. Pineal indole metabolism in the mouse. Brain Res. 1987;416(1):136-40.

223. Kennaway DJ, Voultsios A, Varcoe TJ, and Moyer RW. Melatonin in mice: rhythms, response to light, adrenergic stimulation, and metabolism. Am J Physiol Regul Integr Comp Physiol. 2002;282(2):R358-65.

224. Mure LS, Le HD, Benegiamo G, Chang MW, Rios L, Jillani N, et al. Diurnal transcriptome atlas of a primate across major neural and peripheral tissues. Science (New York, NY). 2018;359(6381).

225. Zhang R, Lahens NF, Ballance HI, Hughes ME, and Hogenesch JB. A circadian gene expression atlas in mammals: implications for biology and medicine. Proceedings of the National Academy of Sciences of the United States of America. 2014;111(45):16219-24.

226. Skarke C, Lahens NF, Rhoades SD, Campbell A, Bittinger K, Bailey A, et al. A Pilot Characterization of the Human Chronobiome. Sci Rep. 2017;7(1):17141.

227. Marc D. Ruben1 GW, David F. Smith2,3, Robert E. Schmidt1, Lauren J. Francey1, Ron C. Anafi4, and John B. Hogenesch. A population-based human enCYCLOPedia for circadian medicine. 2018.

228. Chen Z, Yoo SH, and Takahashi JS. Small molecule modifiers of circadian clocks. Cell Mol Life Sci. 2013;70(16):2985-98.

229. He B, Nohara K, Park N, Park YS, Guillory B, Zhao Z, et al. The Small Molecule Nobiletin Targets the Molecular Oscillator to Enhance Circadian Rhythms and Protect against Metabolic Syndrome. Cell metabolism. 2016;23(4):610-21.

230. Chen Z, Yoo SH, Park YS, Kim KH, Wei S, Buhr E, et al. Identification of diverse modulators of central and peripheral circadian clocks by high-throughput chemical screening. Proceedings of the National Academy of Sciences of the United States of America. 2012;109(1):101-6.

231. Schonenberger MJ, and Kovacs WJ. Hypoxia signaling pathways: modulators of oxygen- related organelles. Front Cell Dev Biol. 2015;3:42.

232. Wong BW, Kuchnio A, Bruning U, and Carmeliet P. Emerging novel functions of the oxygen-sensing prolyl hydroxylase domain enzymes. Trends Biochem Sci. 2013;38(1):3-11.

233. Cermakian N, and Boivin DB. A molecular perspective of human circadian rhythm disorders. Brain Res Brain Res Rev. 2003;42(3):204-20.

219 234. Gamaldo CE, Chung Y, Kang YM, and Salas RM. Tick-tock-tick-tock: the impact of circadian rhythm disorders on cardiovascular health and wellness. J Am Soc Hypertens. 2014;8(12):921-9.

235. Bunger MK, Wilsbacher LD, Moran SM, Clendenin C, Radcliffe LA, Hogenesch JB, et al. Mop3 is an essential component of the master circadian pacemaker in mammals. Cell. 2000;103(7):1009-17.

236. Curtis AM, Cheng Y, Kapoor S, Reilly D, Price TS, and Fitzgerald GA. Circadian variation of blood pressure and the vascular response to asynchronous stress. Proceedings of the National Academy of Sciences of the United States of America. 2007;104(9):3450-5.

237. Anea CB, Zhang M, Stepp DW, Simkins GB, Reed G, Fulton DJ, et al. Vascular disease in mice with a dysfunctional circadian clock. Circulation. 2009;119(11):1510-7.

238. Rudic RD, McNamara P, Curtis AM, Boston RC, Panda S, Hogenesch JB, et al. BMAL1 and CLOCK, two essential components of the circadian clock, are involved in glucose homeostasis. PLoS Biol. 2004;2(11):e377.

239. Pan X, Zhang Y, Wang L, and Hussain MM. Diurnal regulation of MTP and plasma triglyceride by CLOCK is mediated by SHP. Cell metabolism. 2010;12(2):174-86.

240. Turek FW, Joshu C, Kohsaka A, Lin E, Ivanova G, McDearmon E, et al. Obesity and metabolic syndrome in circadian Clock mutant mice. Science (New York, NY). 2005;308(5724):1043-5.

241. Zheng B, Larkin DW, Albrecht U, Sun ZS, Sage M, Eichele G, et al. The mPer2 gene encodes a functional component of the mammalian circadian clock. Nature. 1999;400(6740):169- 73.

242. van der Horst GT, Muijtjens M, Kobayashi K, Takano R, Kanno S, Takao M, et al. Mammalian Cry1 and Cry2 are essential for maintenance of circadian rhythms. Nature. 1999;398(6728):627-30.

243. Vitaterna MH, Selby CP, Todo T, Niwa H, Thompson C, Fruechte EM, et al. Differential regulation of mammalian period genes and circadian rhythmicity by cryptochromes 1 and 2. Proceedings of the National Academy of Sciences of the United States of America. 1999;96(21):12114-9.

244. Doi M, Takahashi Y, Komatsu R, Yamazaki F, Yamada H, Haraguchi S, et al. Salt- sensitive hypertension in circadian clock-deficient Cry-null mice involves dysregulated adrenal Hsd3b6. Nature medicine. 2010;16(1):67-74.

245. Zhang EE, Liu Y, Dentin R, Pongsawakul PY, Liu AC, Hirota T, et al. Cryptochrome mediates circadian regulation of cAMP signaling and hepatic gluconeogenesis. Nature medicine. 2010;16(10):1152-6.

246. Lamia KA, Papp SJ, Yu RT, Barish GD, Uhlenhaut NH, Jonker JW, et al. Cryptochromes mediate rhythmic repression of the glucocorticoid receptor. Nature. 2011;480(7378):552-6.

220 247. Woldt E, Sebti Y, Solt LA, Duhem C, Lancel S, Eeckhoute J, et al. Rev-erb-alpha modulates skeletal muscle oxidative capacity by regulating mitochondrial biogenesis and autophagy. Nature medicine. 2013;19(8):1039-46.

248. Cho H, Zhao X, Hatori M, Yu RT, Barish GD, Lam MT, et al. Regulation of circadian behaviour and metabolism by REV-ERB-alpha and REV-ERB-beta. Nature. 2012;485(7396):123- 7.

249. Duez H, van der Veen JN, Duhem C, Pourcet B, Touvier T, Fontaine C, et al. Regulation of bile acid synthesis by the nuclear receptor Rev-erbalpha. Gastroenterology. 2008;135(2):689- 98.

250. Widmer RJ, and Lerman A. Endothelial dysfunction and cardiovascular disease. Glob Cardiol Sci Pract. 2014;2014(3):291-308.

251. Wong BW, Marsch E, Treps L, Baes M, and Carmeliet P. Endothelial cell metabolism in health and disease: impact of hypoxia. EMBO J. 2017;36(15):2187-203.

252. Wang CY, Wen MS, Wang HW, Hsieh IC, Li Y, Liu PY, et al. Increased vascular senescence and impaired endothelial progenitor cell function mediated by mutation of circadian gene Per2. Circulation. 2008;118(21):2166-73.

253. Neubauer S. The failing heart--an engine out of fuel. N Engl J Med. 2007;356(11):1140- 51.

254. Aragones J, Fraisl P, Baes M, and Carmeliet P. Oxygen sensors at the crossroad of metabolism. Cell metabolism. 2009;9(1):11-22.

255. Minchenko A, Leshchinsky I, Opentanova I, Sang N, Srinivas V, Armstead V, et al. Hypoxia-inducible factor-1-mediated expression of the 6-phosphofructo-2-kinase/fructose-2,6- bisphosphatase-3 (PFKFB3) gene. Its possible role in the Warburg effect. The Journal of biological chemistry. 2002;277(8):6183-7.

256. Aebersold R, and Mann M. Mass spectrometry-based proteomics. Nature. 2003;422(6928):198-207.

257. Neufeld-Cohen A, Robles MS, Aviram R, Manella G, Adamovich Y, Ladeuix B, et al. Circadian control of oscillations in mitochondrial rate-limiting enzymes and nutrient utilization by PERIOD proteins. Proceedings of the National Academy of Sciences of the United States of America. 2016;113(12):E1673-82.

258. Zhou R, Yazdi AS, Menu P, and Tschopp J. A role for mitochondria in NLRP3 inflammasome activation. Nature. 2011;469(7329):221-5.

259. Rey G, Valekunja UK, Feeney KA, Wulund L, Milev NB, Stangherlin A, et al. The Pentose Phosphate Pathway Regulates the Circadian Clock. Cell metabolism. 2016;24(3):462-73.

260. Chong SY, Ptacek LJ, and Fu YH. Genetic insights on sleep schedules: this time, it's PERsonal. Trends Genet. 2012;28(12):598-605.

221 261. Isobe Y, Hida H, and Nishino H. Circadian rhythm of enolase in suprachiasmatic nucleus depends on mitochondrial function. J Neurosci Res. 2011;89(6):936-44.

262. Pivovarova O, Gogebakan O, Sucher S, Groth J, Murahovschi V, Kessler K, et al. Regulation of the clock gene expression in human adipose tissue by weight loss. Int J Obes (Lond). 2016;40(6):899-906.

263. Vieira E, Ruano E, Figueroa AL, Aranda G, Momblan D, Carmona F, et al. Altered clock gene expression in obese visceral adipose tissue is associated with metabolic syndrome. PLoS One. 2014;9(11):e111678.

264. Bartman CM, Oyama Y, Brodsky K, Khailova L, Walker L, Koeppen M, et al. Intense light- elicited upregulation of miR-21 facilitates glycolysis and cardioprotection through Per2-dependent mechanisms. PLoS One. 2017;12(4):e0176243.

265. Eckle T, Grenz A, Kohler D, Redel A, Falk M, Rolauffs B, et al. Systematic evaluation of a novel model for cardiac ischemic preconditioning in mice. American journal of physiology Heart and circulatory physiology. 2006;291(5):H2533-40.

266. Seo SW, Koeppen M, Bonney S, Gobel M, Thayer M, Harter PN, et al. Differential Tissue- Specific Function of Adora2b in Cardioprotection. J Immunol. 2015;195(4):1732-43.

267. Kumarswamy R, and Thum T. Non-coding RNAs in cardiac remodeling and heart failure. Circ Res. 2013;113(6):676-89.

268. Shinde AV, and Frangogiannis NG. Fibroblasts in myocardial infarction: a role in inflammation and repair. J Mol Cell Cardiol. 2014;70:74-82.

269. Xiao J, Pan Y, Li XH, Yang XY, Feng YL, Tan HH, et al. Cardiac progenitor cell-derived exosomes prevent cardiomyocytes apoptosis through exosomal miR-21 by targeting PDCD4. Cell Death Dis. 2016;7(6):e2277.

270. Xu X, Kriegel AJ, Jiao X, Liu H, Bai X, Olson J, et al. miR-21 in ischemia/reperfusion injury: a double-edged sword? Physiol Genomics. 2014;46(21):789-97.

271. Deng W, Wang Y, Long X, Zhao R, Wang Z, Liu Z, et al. miR-21 Reduces Hydrogen Peroxide-Induced Apoptosis in c-kit(+) Cardiac Stem Cells In Vitro through PTEN/PI3K/Akt Signaling. Oxid Med Cell Longev. 2016;2016:5389181.

272. Zeng YL, Zheng H, Chen QR, Yuan XH, Ren JH, Luo XF, et al. Bone marrow-derived mesenchymal stem cells overexpressing MiR-21 efficiently repair myocardial damage in rats. Oncotarget. 2017;8(17):29161-73.

273. Kathale ND, and Liu AC. Prevalence of cycling genes and drug targets calls for prospective chronotherapeutics. Proceedings of the National Academy of Sciences of the United States of America. 2014;111(45):15869-70.

222 274. Seagroves TN, Ryan HE, Lu H, Wouters BG, Knapp M, Thibault P, et al. Transcription factor HIF-1 is a necessary mediator of the pasteur effect in mammalian cells. Mol Cell Biol. 2001;21(10):3436-44.

275. Liu Y, Nie H, Zhang K, Ma D, Yang G, Zheng Z, et al. A feedback regulatory loop between HIF-1alpha and miR-21 in response to hypoxia in cardiomyocytes. FEBS Lett. 2014;588(17):3137-46.

276. Jiang S, Wang R, Yan H, Jin L, Dou X, and Chen D. MicroRNA-21 modulates radiation resistance through upregulation of hypoxia-inducible factor-1alpha-promoted glycolysis in non- small cell lung cancer cells. Mol Med Rep. 2016;13(5):4101-7.

277. Huang S, Zou X, Zhu JN, Fu YH, Lin QX, Liang YY, et al. Attenuation of microRNA-16 derepresses the cyclins D1, D2 and E1 to provoke cardiomyocyte hypertrophy. J Cell Mol Med. 2015;19(3):608-19.

278. Bernardo BC, Nguyen SS, Gao XM, Tham YK, Ooi JY, Patterson NL, et al. Inhibition of miR-154 Protects Against Cardiac Dysfunction and Fibrosis in a Mouse Model of Pressure Overload. Sci Rep. 2016;6:22442.

279. Guo C, Deng Y, Liu J, and Qian L. Cardiomyocyte-specific role of miR-24 in promoting cell survival. J Cell Mol Med. 2015;19(1):103-12.

280. Veliceasa D, Biyashev D, Qin G, Misener S, Mackie AR, Kishore R, et al. Therapeutic manipulation of angiogenesis with miR-27b. Vasc Cell. 2015;7:6.

281. Huang W, Tian SS, Hang PZ, Sun C, Guo J, and Du ZM. Combination of microRNA-21 and microRNA-146a Attenuates Cardiac Dysfunction and Apoptosis During Acute Myocardial Infarction in Mice. Mol Ther Nucleic Acids. 2016;5:e296.

282. Pan L, Huang BJ, Ma XE, Wang SY, Feng J, Lv F, et al. MiR-25 protects cardiomyocytes against oxidative damage by targeting the mitochondrial calcium uniporter. Int J Mol Sci. 2015;16(3):5420-33.

283. Marques FZ, Vizi D, Khammy O, Mariani JA, and Kaye DM. The transcardiac gradient of cardio-microRNAs in the failing heart. Eur J Heart Fail. 2016;18(8):1000-8.

284. Zeller T, Keller T, Ojeda F, Reichlin T, Twerenbold R, Tzikas S, et al. Assessment of microRNAs in patients with unstable angina pectoris. Eur Heart J. 2014;35(31):2106-14.

285. Hsu A, Chen SJ, Chang YS, Chen HC, and Chu PH. Systemic approach to identify serum microRNAs as potential biomarkers for acute myocardial infarction. Biomed Res Int. 2014;2014:418628.

286. Deng P, Chen L, Liu Z, Ye P, Wang S, Wu J, et al. MicroRNA-150 Inhibits the Activation of Cardiac Fibroblasts by Regulating c-Myb. Cell Physiol Biochem. 2016;38(6):2103-22.

223 287. Ellis KL, Cameron VA, Troughton RW, Frampton CM, Ellmers LJ, and Richards AM. Circulating microRNAs as candidate markers to distinguish heart failure in breathless patients. Eur J Heart Fail. 2013;15(10):1138-47.

288. Li SP, Liu B, Song B, Wang CX, and Zhou YC. miR-28 promotes cardiac ischemia by targeting mitochondrial aldehyde dehydrogenase 2 (ALDH2) in mus musculus cardiac myocytes. Eur Rev Med Pharmacol Sci. 2015;19(5):752-8.

289. Li Q, Xie J, Wang B, Li R, Bai J, Ding L, et al. Overexpression of microRNA-99a Attenuates Cardiac Hypertrophy. PLoS One. 2016;11(2):e0148480.

290. Marchand A, Atassi F, Mougenot N, Clergue M, Codoni V, Berthuin J, et al. miR-322 regulates insulin signaling pathway and protects against metabolic syndrome-induced cardiac dysfunction in mice. Biochimica et biophysica acta. 2016;1862(4):611-21.

291. Yin X, Peng C, Ning W, Li C, Ren Z, Zhang J, et al. miR-30a downregulation aggravates pressure overload-induced cardiomyocyte hypertrophy. Mol Cell Biochem. 2013;379(1-2):1-6.

292. Wang K, Lin ZQ, Long B, Li JH, Zhou J, and Li PF. Cardiac hypertrophy is positively regulated by MicroRNA miR-23a. The Journal of biological chemistry. 2012;287(1):589-99.

293. Zhao X, Wang K, Hu F, Qian C, Guan H, Feng K, et al. MicroRNA-101 protects cardiac fibroblasts from hypoxia-induced apoptosis via inhibition of the TGF-beta signaling pathway. The international journal of biochemistry & cell biology. 2015;65:155-64.

294. Liu HL, Zhu JG, Liu YQ, Fan ZG, Zhu C, and Qian LM. Identification of the microRNA expression profile in the regenerative neonatal mouse heart by deep sequencing. Cell Biochem Biophys. 2014;70(1):635-42.

295. Shehadeh LA, Sharma S, Pessanha M, Wei JQ, Liu J, Yuan H, et al. MicroRNA-20a constrains p300-driven myocardial angiogenic transcription by direct targeting of p300. PLoS One. 2013;8(11):e79133.

296. Li Z, Liu L, Hou N, Song Y, An X, Zhang Y, et al. miR-199-sponge transgenic mice develop physiological cardiac hypertrophy. Cardiovasc Res. 2016;110(2):258-67.

297. Wang K, Zhang DL, Long B, An T, Zhang J, Zhou LY, et al. NFAT4-dependent miR-324- 5p regulates mitochondrial morphology and cardiomyocyte cell death by targeting Mtfr1. Cell Death Dis. 2015;6:e2007.

298. Duisters RF, Tijsen AJ, Schroen B, Leenders JJ, Lentink V, van der Made I, et al. miR- 133 and miR-30 regulate connective tissue growth factor: implications for a role of microRNAs in myocardial matrix remodeling. Circ Res. 2009;104(2):170-8, 6p following 8.

299. Lv P, Zhou M, He J, Meng W, Ma X, Dong S, et al. Circulating miR-208b and miR-34a are associated with left ventricular remodeling after acute myocardial infarction. Int J Mol Sci. 2014;15(4):5774-88.

224 300. Fan Y, Liu L, Fang K, Huang T, Wan L, Liu Y, et al. Resveratrol Ameliorates Cardiac Hypertrophy by Down-regulation of miR-155 Through Activation of Breast Cancer Type 1 Susceptibility Protein. J Am Heart Assoc. 2016;5(4).

301. Wang JX, Zhang XJ, Feng C, Sun T, Wang K, Wang Y, et al. MicroRNA-532-3p regulates mitochondrial fission through targeting apoptosis repressor with caspase recruitment domain in doxorubicin cardiotoxicity. Cell Death Dis. 2015;6:e1677.

302. Ge Y, Pan S, Guan D, Yin H, Fan Y, Liu J, et al. MicroRNA-350 induces pathological heart hypertrophy by repressing both p38 and JNK pathways. Biochimica et biophysica acta. 2013;1832(1):1-10.

303. Liu F, Li N, Long B, Fan YY, Liu CY, Zhou QY, et al. Cardiac hypertrophy is negatively regulated by miR-541. Cell Death Dis. 2014;5:e1171.

304. Zhang Y, Wang R, Du W, Wang S, Yang L, Pan Z, et al. Downregulation of miR-151-5p contributes to increased susceptibility to arrhythmogenesis during myocardial infarction with estrogen deprivation. PLoS One. 2013;8(9):e72985.

305. Wu C, Arora P, Agha O, Hurst LA, Allen K, Nathan DI, et al. Novel MicroRNA Regulators of Atrial Natriuretic Peptide Production. Mol Cell Biol. 2016;36(14):1977-87.

306. Diez-Planelles C, Sanchez-Lozano P, Crespo MC, Gil-Zamorano J, Ribacoba R, Gonzalez N, et al. Circulating microRNAs in Huntington's disease: Emerging mediators in metabolic impairment. Pharmacol Res. 2016;108:102-10.

307. Dispersyn G, Pain L, and Touitou Y. Propofol anesthesia significantly alters plasma blood levels of melatonin in rats. Anesthesiology. 2010;112(2):333-7.

308. Dispersyn G, Chassard D, and Pain L. [Biological rhythms for anaesthesia and intensive care]. Ann Fr Anesth Reanim. 2010;29(6):470-7.

309. Dhir A, Zolkowska D, and Rogawski MA. Seizure protection by intrapulmonary delivery of midazolam in mice. Neuropharmacology. 2013;73:425-31.

310. Zhang HY, McPherson BC, Liu H, Baman TS, Rock P, and Yao Z. H(2)O(2) opens mitochondrial K(ATP) channels and inhibits GABA receptors via protein kinase C-epsilon in cardiomyocytes. American journal of physiology Heart and circulatory physiology. 2002;282(4):H1395-403.

311. Leducq N, Bono F, Sulpice T, Vin V, Janiak P, Fur GL, et al. Role of peripheral benzodiazepine receptors in mitochondrial, cellular, and cardiac damage induced by oxidative stress and ischemia-reperfusion. J Pharmacol Exp Ther. 2003;306(3):828-37.

312. Gloston GF, Yoo SH, and Chen ZJ. Clock-Enhancing Small Molecules and Potential Applications in Chronic Diseases and Aging. Front Neurol. 2017;8:100.

313. Akhlaghi M, and Bandy B. Mechanisms of flavonoid protection against myocardial ischemia-reperfusion injury. J Mol Cell Cardiol. 2009;46(3):309-17.

225 314. O'Neill JS, Maywood ES, Chesham JE, Takahashi JS, and Hastings MH. cAMP- dependent signaling as a core component of the mammalian circadian pacemaker. Science (New York, NY). 2008;320(5878):949-53.

315. Narravula S, Lennon PF, Mueller BU, and Colgan SP. Regulation of endothelial CD73 by adenosine: paracrine pathway for enhanced endothelial barrier function. J Immunol. 2000;165(9):5262-8.

316. Homola M, Pfeffer M, Fischer C, Zimmermann H, Robson SC, and Korf HW. Expression of ectonucleotidases in the prosencephalon of melatonin-proficient C3H and melatonin-deficient C57Bl mice: spatial distribution and time-dependent changes. Cell Tissue Res. 2015;362(1):163- 76.

317. Yang F, Liu W, Yan X, Zhou H, Zhang H, Liu J, et al. Effects of mir-21 on Cardiac Microvascular Endothelial Cells After Acute Myocardial Infarction in Rats: Role of Phosphatase and Tensin Homolog (PTEN)/Vascular Endothelial Growth Factor (VEGF) Signal Pathway. Med Sci Monit. 2016;22:3562-75.

318. Methner C, and Krieg T. A2b adenosine receptors in cardioprotection: timing is everything. J Mol Cell Cardiol. 2011;50(3):582-3.

319. Tu Y, Wan L, Fan Y, Wang K, Bu L, Huang T, et al. Ischemic postconditioning-mediated miRNA-21 protects against cardiac ischemia/reperfusion injury via PTEN/Akt pathway. PLoS One. 2013;8(10):e75872.

320. Qin Y, Yu Y, Dong H, Bian X, Guo X, and Dong S. MicroRNA 21 inhibits left ventricular remodeling in the early phase of rat model with ischemia-reperfusion injury by suppressing cell apoptosis. Int J Med Sci. 2012;9(6):413-23.

321. Ballesta A, Innominato PF, Dallmann R, Rand DA, and Levi FA. Systems Chronotherapeutics. Pharmacol Rev. 2017;69(2):161-99.

322. Du NH, Arpat AB, De Matos M, and Gatfield D. MicroRNAs shape circadian hepatic gene expression on a transcriptome-wide scale. Elife. 2014;3:e02510.

323. Ginsberg HN, Zhang YL, and Hernandez-Ono A. Regulation of plasma triglycerides in insulin resistance and diabetes. Archives of medical research. 2005;36(3):232-40.

324. Lee PH, and Suen LK. The convergent validity of Actiwatch 2 and ActiGraph Link accelerometers in measuring total sleeping period, wake after sleep onset, and sleep efficiency in free-living condition. Sleep & breathing = Schlaf & Atmung. 2017;21(1):209-15.

325. Cajochen C, Jud C, Munch M, Kobialka S, Wirz-Justice A, and Albrecht U. Evening exposure to blue light stimulates the expression of the clock gene PER2 in humans. Eur J Neurosci. 2006;23(4):1082-6.

326. Yorguner Kupeli N, Bulut NS, Carkaxhiu Bulut G, Kurt E, and Kora K. Efficacy of bright light therapy in bipolar depression. Psychiatry research. 2017;260:432-8.

226 327. Roenneberg T, Kumar CJ, and Merrow M. The human circadian clock entrains to sun time. Current biology : CB. 2007;17(2):R44-5.

328. Terman JS, Terman M, Lo ES, and Cooper TB. Circadian time of morning light administration and therapeutic response in winter depression. Arch Gen Psychiatry. 2001;58(1):69-75.

329. Terman JS, Terman M, Schlager D, Rafferty B, Rosofsky M, Link MJ, et al. Efficacy of brief, intense light exposure for treatment of winter depression. Psychopharmacol Bull. 1990;26(1):3-11.

330. Yang J, Choi W, Ko YH, Joe SH, Han C, and Kim YK. Bright light therapy as an adjunctive treatment with risperidone in patients with delirium: a randomized, open, parallel group study. Gen Hosp Psychiatry. 2012;34(5):546-51.

331. Eckle T. Editorial: Health Impact and Management of a Disrupted Circadian Rhythm and Sleep in Critical Illnesses. Curr Pharm Des. 2015;21(24):3428-30.

332. Phipps-Nelson J, Redman JR, Dijk DJ, and Rajaratnam SM. Daytime exposure to bright light, as compared to dim light, decreases sleepiness and improves psychomotor vigilance performance. Sleep. 2003;26(6):695-700.

333. Botto F, Alonso-Coello P, Chan MT, Villar JC, Xavier D, Srinathan S, et al. Myocardial injury after noncardiac surgery: a large, international, prospective cohort study establishing diagnostic criteria, characteristics, predictors, and 30-day outcomes. Anesthesiology. 2014;120(3):564-78.

334. Benjamin EJ, Blaha MJ, Chiuve SE, Cushman M, Das SR, Deo R, et al. Heart Disease and Stroke Statistics-2017 Update: A Report From the American Heart Association. Circulation. 2017;135(10):e146-e603.

335. Hermansson J, Hallqvist J, Karlsson B, Knutsson A, and Gillander Gadin K. Shift work, parental cardiovascular disease and myocardial infarction in males. Occup Med (Lond). 2018;68(2):120-5.

336. Eltzschig HK, Ibla JC, Furuta GT, Leonard MO, Jacobson KA, Enjyoji K, et al. Coordinated adenine nucleotide phosphohydrolysis and nucleoside signaling in posthypoxic endothelium: role of ectonucleotidases and adenosine A2B receptors. J Exp Med. 2003;198(5):783-96.

337. Xia JB, Liu GH, Chen ZY, Mao CZ, Zhou DC, Wu HY, et al. Hypoxia/ischemia promotes CXCL10 expression in cardiac microvascular endothelial cells by NFkB activation. Cytokine. 2016;81:63-70.

338. Silliman CC, Dzieciatkowska M, Moore EE, Kelher MR, Banerjee A, Liang X, et al. Proteomic analyses of human plasma: Venus versus Mars. Transfusion. 2012;52(2):417-24.

339. Barrett AS, Wither MJ, Hill RC, Dzieciatkowska M, D'Alessandro A, Reisz JA, et al. Hydroxylamine Chemical Digestion for Insoluble Extracellular Matrix Characterization. J Proteome Res. 2017;16(11):4177-84.

227 340. Sun L, Zhu G, and Dovichi NJ. Comparison of the LTQ-Orbitrap Velos and the Q-Exactive for proteomic analysis of 1-1000 ng RAW 264.7 cell lysate digests. Rapid communications in mass spectrometry : RCM. 2013;27(1):157-62.

341. Thwe PM, Pelgrom L, Cooper R, Beauchamp S, Reisz JA, D'Alessandro A, et al. Cell- Intrinsic Glycogen Metabolism Supports Early Glycolytic Reprogramming Required for Dendritic Cell Immune Responses. Cell metabolism. 2017;26(3):558-67 e5.

342. Nemkov T, D'Alessandro A, and Hansen KC. Three-minute method for amino acid analysis by UHPLC and high-resolution quadrupole orbitrap mass spectrometry. Amino acids. 2015;47(11):2345-57.

343. Nemkov T, Hansen KC, and D'Alessandro A. A three-minute method for high-throughput quantitative metabolomics and quantitative tracing experiments of central carbon and nitrogen pathways. Rapid communications in mass spectrometry : RCM. 2017;31(8):663-73.

344. McCurdy CE, Schenk S, Hetrick B, Houck J, Drew BG, Kaye S, et al. Maternal obesity reduces oxidative capacity in fetal skeletal muscle of Japanese macaques. JCI insight. 2016;1(16):e86612.

345. Yoo SH, Yamazaki S, Lowrey PL, Shimomura K, Ko CH, Buhr ED, et al. PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proceedings of the National Academy of Sciences of the United States of America. 2004;101(15):5339-46.

346. Safran M, Kim WY, O'Connell F, Flippin L, Gunzler V, Horner JW, et al. Mouse model for noninvasive imaging of HIF prolyl hydroxylase activity: assessment of an oral agent that stimulates erythropoietin production. Proceedings of the National Academy of Sciences of the United States of America. 2006;103(1):105-10.

347. Ma X, Kumar M, Choudhury SN, Becker Buscaglia LE, Barker JR, Kanakamedala K, et al. Loss of the miR-21 allele elevates the expression of its target genes and reduces tumorigenesis. Proceedings of the National Academy of Sciences of the United States of America. 2011;108(25):10144-9.

348. Koeppen M, Harter PN, Bonney S, Bonney M, Reithel S, Zachskorn C, et al. Adora2b signaling on bone marrow derived cells dampens myocardial ischemia-reperfusion injury. Anesthesiology. 2012;116(6):1245-57.

349. Redel A, Jazbutyte V, Smul TM, Lange M, Eckle T, Eltzschig H, et al. Impact of ischemia and reperfusion times on myocardial infarct size in mice in vivo. Exp Biol Med (Maywood). 2008;233(1):84-93.

350. Dewald O, Frangogiannis NG, Zoerlein MP, Duerr GD, Taffet G, Michael LH, et al. A murine model of ischemic cardiomyopathy induced by repetitive ischemia and reperfusion. Thorac Cardiovasc Surg. 2004;52(5):305-11.

228 351. Luo B, Groenke K, Takors R, Wandrey C, and Oldiges M. Simultaneous determination of multiple intracellular metabolites in glycolysis, pentose phosphate pathway and tricarboxylic acid cycle by liquid chromatography-mass spectrometry. J Chromatogr A. 2007;1147(2):153-64.

352. Han J, Gagnon S, Eckle T, and Borchers CH. Metabolomic analysis of key central carbon metabolism carboxylic acids as their 3-nitrophenylhydrazones by UPLC/ESI-MS. Electrophoresis. 2013;34(19):2891-900.

353. Vasile VC, Babuin L, Giannitsis E, Katus HA, and Jaffe AS. Relationship of MRI- determined infarct size and cTnI measurements in patients with ST-elevation myocardial infarction. Clin Chem. 2008;54(3):617-9.

354. Moore A, Zielinski T, and Millar AJ. Online period estimation and determination of rhythmicity in circadian data, using the BioDare data infrastructure. Methods in molecular biology (Clifton, NJ). 2014;1158:13-44.

355. Zhu F, Li Y, Zhang J, Piao C, Liu T, Li HH, et al. Senescent cardiac fibroblast is critical for cardiac fibrosis after myocardial infarction. PLoS One. 2013;8(9):e74535.

356. O'Connell TD, Rodrigo MC, and Simpson PC. Isolation and culture of adult mouse cardiac myocytes. Methods in molecular biology (Clifton, NJ). 2007;357:271-96.

357. Finley LW, and Haigis MC. Metabolic regulation by SIRT3: implications for tumorigenesis. Trends Mol Med. 2012;18(9):516-23.

358. Pavan B, Frigato E, Pozzati S, Prasad PD, Bertolucci C, and Biondi C. Circadian clocks regulate adenylyl cyclase activity rhythms in human RPE cells. Biochem Biophys Res Commun. 2006;350(1):169-73.

359. Colgan SP, Eltzschig HK, Eckle T, and Thompson LF. Physiological roles for ecto-5'- nucleotidase (CD73). Purinergic Signal. 2006;2(2):351-60.

360. Kulesskaya N, Voikar V, Peltola M, Yegutkin GG, Salmi M, Jalkanen S, et al. CD73 is a major regulator of adenosinergic signalling in mouse brain. PLoS One. 2013;8(6):e66896.

361. Chen X, and Gillis CN. Enhanced photorelaxation in aorta, pulmonary artery and corpus cavernosum produced by BAY K 8644 or N-nitro-L-arginine. Biochem Biophys Res Commun. 1992;186(3):1522-7.

362. Chen X, and Gillis CN. Methylene blue enhanced photorelaxation in aorta, pulmonary artery and corpus cavernosum. Biochem Biophys Res Commun. 1993;190(2):559-63.

363. Sikka G, Hussmann GP, Pandey D, Cao S, Hori D, Park JT, et al. Melanopsin mediates light-dependent relaxation in blood vessels. Proceedings of the National Academy of Sciences of the United States of America. 2014;111(50):17977-82.

364. Batenburg WW, Kappers MH, Eikmann MJ, Ramzan SN, de Vries R, and Danser AH. Light-induced vs. bradykinin-induced relaxation of coronary arteries: do S-nitrosothiols act as endothelium-derived hyperpolarizing factors? J Hypertens. 2009;27(8):1631-40.

229 365. Barreto Ortiz S, Hori D, Nomura Y, Yun X, Jiang H, Yong H, et al. Opsin 3 and 4 mediate light-induced pulmonary vasorelaxation that is potentiated by G protein-coupled receptor kinase 2 inhibition. Am J Physiol Lung Cell Mol Physiol. 2018;314(1):L93-L106.

366. Needham DM. A quantitative study of succinic acid in muscle: Glutamic and aspartic acids as precursors. Biochem J. 1930;24(1):208-27.

367. Weil-Malherbe H. Studies on brain metabolism: Formation of succinic acid. Biochem J. 1937;31(2):299-312.

368. Chinopoulos C. Which way does the citric acid cycle turn during hypoxia? The critical role of alpha-ketoglutarate dehydrogenase complex. J Neurosci Res. 2013;91(8):1030-43.

369. Hoberman HD, and Prosky L. Evidence of reduction of fumarate to succinate in perfused rat liver under conditions of reduced O2 tension. Biochimica et biophysica acta. 1967;148(2):392- 9.

370. Folbergrova J, Ljunggren B, Norberg K, and Siesjo BK. Influence of complete ischemia on glycolytic metabolites, citric acid cycle intermediates, and associated amino acids in the rat cerebral cortex. Brain Res. 1974;80(2):265-79.

371. Freminet A, Leclerc L, Poyart C, Huel C, and Gentil M. Alanine and succinate accumulation in the perfused rat heart during hypoxia. J Physiol (Paris). 1980;76(2):113-7.

372. Benzi G, Pastoris O, and Dossena M. Relationships between gamma-aminobutyrate and succinate cycles during and after cerebral ischemia. J Neurosci Res. 1982;7(2):193-201.

373. Pisarenko OI, Solomatina ES, and Studneva IM. The role of amino acid catabolism in the formation of the tricarboxylic acid cycle intermediates and ammonia in anoxic rat heart. Biochimica et biophysica acta. 1986;885(2):154-61.

374. Taegtmeyer H, Peterson MB, Ragavan VV, Ferguson AG, and Lesch M. De novo alanine synthesis in isolated oxygen-deprived rabbit myocardium. The Journal of biological chemistry. 1977;252(14):5010-8.

375. Taegtmeyer H. Metabolic responses to cardiac hypoxia. Increased production of succinate by rabbit papillary muscles. Circ Res. 1978;43(5):808-15.

376. Selak MA, Armour SM, MacKenzie ED, Boulahbel H, Watson DG, Mansfield KD, et al. Succinate links TCA cycle dysfunction to oncogenesis by inhibiting HIF-alpha prolyl hydroxylase. Cancer Cell. 2005;7(1):77-85.

377. Starkov AA. The role of mitochondria in reactive oxygen species metabolism and signaling. Ann N Y Acad Sci. 2008;1147:37-52.

378. Korshunov SS, Skulachev VP, and Starkov AA. High protonic potential actuates a mechanism of production of reactive oxygen species in mitochondria. FEBS Lett. 1997;416(1):15- 8.

230 379. Zhang H, Bosch-Marce M, Shimoda LA, Tan YS, Baek JH, Wesley JB, et al. Mitochondrial autophagy is an HIF-1-dependent adaptive metabolic response to hypoxia. The Journal of biological chemistry. 2008;283(16):10892-903.

380. Musiek ES, and Fitzgerald GA. Molecular clocks in pharmacology. Handb Exp Pharmacol. 2013(217):243-60.

381. Marc D Ruben GW, David F Smith, Robert E Schmidt, Lauren J Francey, Ron C Anafi, John B Hogenesch. A population-based human enCYCLOPedia for circadian medicine. bioRxiv. 2018.

382. Kojetin DJ, and Burris TP. REV-ERB and ROR nuclear receptors as drug targets. Nat Rev Drug Discov. 2014;13(3):197-216.

383. Piro M, Della Bona R, Abbate A, Biasucci LM, and Crea F. Sex-related differences in myocardial remodeling. Journal of the American College of Cardiology. 2010;55(11):1057-65.

384. Brower GL, Gardner JD, and Janicki JS. Gender mediated cardiac protection from adverse ventricular remodeling is abolished by ovariectomy. Mol Cell Biochem. 2003;251(1-2):89-95.

385. Cavasin MA, Tao Z, Menon S, and Yang XP. Gender differences in cardiac function during early remodeling after acute myocardial infarction in mice. Life Sci. 2004;75(18):2181-92.

386. Crabbe DL, Dipla K, Ambati S, Zafeiridis A, Gaughan JP, Houser SR, et al. Gender differences in post-infarction hypertrophy in end-stage failing hearts. Journal of the American College of Cardiology. 2003;41(2):300-6.

387. Litwin SE, Katz SE, Litwin CM, Morgan JP, and Douglas PS. Gender differences in postinfarction left ventricular remodeling. Cardiology. 1999;91(3):173-83.

388. Zampino M, Yuzhakova M, Hansen J, McKinney RD, Goldspink PH, Geenen DL, et al. Sex-related dimorphic response of HIF-1 alpha expression in myocardial ischemia. American journal of physiology Heart and circulatory physiology. 2006;291(2):H957-64.

389. Biondi-Zoccai GG, Abate A, Bussani R, Camilot D, Giorgio FD, Marino MP, et al. Reduced post-infarction myocardial apoptosis in women: a clue to their different clinical course? Heart (British Cardiac Society). 2005;91(1):99-101.

390. Dedkov EI, Bogatyryov Y, Pavliak K, Santos AT, Chen YF, Zhang Y, et al. Sex-related differences in intrinsic myocardial properties influence cardiac function in middle-aged rats during infarction-induced left ventricular remodeling. Physiol Rep. 2016;4(11).

391. Novotny JL, Simpson AM, Tomicek NJ, Lancaster TS, and Korzick DH. Rapid estrogen receptor-alpha activation improves ischemic tolerance in aged female rats through a novel protein kinase C epsilon-dependent mechanism. Endocrinology. 2009;150(2):889-96.

392. Kuljis DA, Loh DH, Truong D, Vosko AM, Ong ML, McClusky R, et al. Gonadal- and sex- chromosome-dependent sex differences in the circadian system. Endocrinology. 2013;154(4):1501-12.

231 393. Abraham U, Granada AE, Westermark PO, Heine M, Kramer A, and Herzel H. Coupling governs entrainment range of circadian clocks. Mol Syst Biol. 2010;6:438.

394. Bailey M, and Silver R. Sex differences in circadian timing systems: implications for disease. Front Neuroendocrinol. 2014;35(1):111-39.

395. Cheng Y, and Zhang C. MicroRNA-21 in cardiovascular disease. J Cardiovasc Transl Res. 2010;3(3):251-5.

396. Sun C, Huang F, Liu X, Xiao X, Yang M, Hu G, et al. miR-21 regulates triglyceride and cholesterol metabolism in non-alcoholic fatty liver disease by targeting HMGCR. Int J Mol Med. 2015;35(3):847-53.

397. Gupta SK, Itagaki R, Zheng X, Batkai S, Thum S, Ahmad F, et al. miR-21 promotes fibrosis in an acute cardiac allograft transplantation model. Cardiovasc Res. 2016;110(2):215-26.

398. Reddy S, Hu DQ, Zhao M, Blay E, Jr., Sandeep N, Ong SG, et al. miR-21 is associated with fibrosis and right ventricular failure. JCI insight. 2017;2(9).

399. Mohanty JG, Nagababu E, and Rifkind JM. Red blood cell oxidative stress impairs oxygen delivery and induces red blood cell aging. Front Physiol. 2014;5:84.

400. O'Neill JS, and Reddy AB. Circadian clocks in human red blood cells. Nature. 2011;469(7331):498-503.

401. Arjona A, Silver AC, Walker WE, and Fikrig E. Immunity's fourth dimension: approaching the circadian-immune connection. Trends Immunol. 2012;33(12):607-12.

402. Lam MT, Cho H, Lesch HP, Gosselin D, Heinz S, Tanaka-Oishi Y, et al. Rev-Erbs repress macrophage gene expression by inhibiting enhancer-directed transcription. Nature. 2013;498(7455):511-5.

403. Keller M, Mazuch J, Abraham U, Eom GD, Herzog ED, Volk HD, et al. A circadian clock in macrophages controls inflammatory immune responses. Proceedings of the National Academy of Sciences of the United States of America. 2009;106(50):21407-12.

404. Stujanna EN, Murakoshi N, Tajiri K, Xu D, Kimura T, Qin R, et al. Rev-erb agonist improves adverse cardiac remodeling and survival in myocardial infarction through an anti-inflammatory mechanism. PLoS One. 2017;12(12):e0189330.

405. Carlson DE, and Chiu WC. The absence of circadian cues during recovery from sepsis modifies pituitary-adrenocortical function and impairs survival. Shock. 2008;29(1):127-32.

406. Obayashi K, Saeki K, Iwamoto J, Okamoto N, Tomioka K, Nezu S, et al. Exposure to light at night, nocturnal urinary melatonin excretion, and obesity/dyslipidemia in the elderly: a cross- sectional analysis of the HEIJO-KYO study. J Clin Endocrinol Metab. 2013;98(1):337-44.

232 407. Alibhai FJ, Tsimakouridze EV, Chinnappareddy N, Wright DC, Billia F, O'Sullivan ML, et al. Short-term disruption of diurnal rhythms after murine myocardial infarction adversely affects long-term myocardial structure and function. Circ Res. 2014;114(11):1713-22.

408. Bartman CM, Oyama Y, and Eckle T. Daytime variations in perioperative myocardial injury. Lancet. 2018;391(10135):2104.

233 APPENDIX A

AFFINITY PURIFICATION-MASS SPECTROMETRY-BASED PROTEOMICS SCREEN FOR

PER2 PROTEIN INTERACTIOSN UNDER HYPOXIC CONDITIONS

Accession No. PER2 protein interaction in hypoxic metabolism Tyrosine--tRNA ligase, cytoplasmic OS=Homo sapiens GN=YARS PE=1 P54577 SV=4 O60684 Importin subunit alpha-7 OS=Homo sapiens GN=KPNA6 PE=1 SV=1 Transcription elongation factor SPT6 OS=Homo sapiens GN=SUPT6H PE=1 Q7KZ85 SV=2 Q9NTK5 Obg-like ATPase 1 OS=Homo sapiens GN=OLA1 PE=1 SV=2 P53618 Coatomer subunit beta OS=Homo sapiens GN=COPB1 PE=1 SV=3 P13667 Protein disulfide-isomerase A4 OS=Homo sapiens GN=PDIA4 PE=1 SV=2 Methionine--tRNA ligase, cytoplasmic OS=Homo sapiens GN=MARS PE=1 P56192 SV=2 Carbamoyl-phosphate synthase [ammonia], mitochondrial OS=Homo P31327 sapiens GN=CPS1 PE=1 SV=2 Acidic leucine-rich nuclear phosphoprotein 32 family member A OS=Homo P39687 sapiens GN=ANP32A PE=1 SV=1 Q00796 Sorbitol dehydrogenase OS=Homo sapiens GN=SORD PE=1 SV=4 Isoleucine--tRNA ligase, cytoplasmic OS=Homo sapiens GN=IARS PE=1 P41252 SV=2 P11498 Pyruvate carboxylase, mitochondrial OS=Homo sapiens GN=PC PE=1 SV=2 P29401 Transketolase OS=Homo sapiens GN=TKT PE=1 SV=3 Replication protein A 70 kDa DNA-binding subunit OS=Homo sapiens P27694 GN=RPA1 PE=1 SV=2 L-lactate dehydrogenase B chain OS=Homo sapiens GN=LDHB PE=1 P07195 SV=2 P78386 Keratin, type II cuticular Hb5 OS=Homo sapiens GN=KRT85 PE=1 SV=1 Eukaryotic translation initiation factor 2A OS=Homo sapiens GN=EIF2A Q9BY44 PE=1 SV=3 P48444 Coatomer subunit delta OS=Homo sapiens GN=ARCN1 PE=1 SV=1 P61626 Lysozyme C OS=Homo sapiens GN=LYZ PE=1 SV=1 39S ribosomal protein L45, mitochondrial OS=Homo sapiens GN=MRPL45 Q9BRJ2 PE=1 SV=2 Isoleucine--tRNA ligase, mitochondrial OS=Homo sapiens GN=IARS2 PE=1 Q9NSE4 SV=2 Ribosome maturation protein SBDS OS=Homo sapiens GN=SBDS PE=1 Q9Y3A5 SV=4 P02788 Lactotransferrin OS=Homo sapiens GN=LTF PE=1 SV=6 P04062 Glucosylceramidase OS=Homo sapiens GN=GBA PE=1 SV=3 Methylcrotonoyl-CoA carboxylase beta chain, mitochondrial OS=Homo Q9HCC0 sapiens GN=MCCC2 PE=1 SV=1 P37268 Squalene synthase OS=Homo sapiens GN=FDFT1 PE=1 SV=1

234 Accession No. PER2 protein interaction in hypoxic metabolism Histidine--tRNA ligase, cytoplasmic OS=Homo sapiens GN=HARS PE=1 P12081 SV=2 Signal recognition particle 54 kDa protein OS=Homo sapiens GN=SRP54 P61011 PE=1 SV=1 39S ribosomal protein L39, mitochondrial OS=Homo sapiens GN=MRPL39 Q9NYK5 PE=1 SV=3 P55060 Exportin-2 OS=Homo sapiens GN=CSE1L PE=1 SV=3 Q16777 Histone H2A type 2-C OS=Homo sapiens GN=HIST2H2AC PE=1 SV=4 28S ribosomal protein S6, mitochondrial OS=Homo sapiens GN=MRPS6 P82932 PE=1 SV=3 Probable asparagine--tRNA ligase, mitochondrial OS=Homo sapiens Q96I59 GN=NARS2 PE=1 SV=3 P30101 Protein disulfide-isomerase A3 OS=Homo sapiens GN=PDIA3 PE=1 SV=4 O15355 Protein phosphatase 1G OS=Homo sapiens GN=PPM1G PE=1 SV=1 28S ribosomal protein S29, mitochondrial OS=Homo sapiens GN=DAP3 P51398 PE=1 SV=1 Q9GZZ8 Extracellular glycoprotein lacritin OS=Homo sapiens GN=LACRT PE=1 SV=1 P50336 Protoporphyrinogen oxidase OS=Homo sapiens GN=PPOX PE=1 SV=1 Dynamin-like 120 kDa protein, mitochondrial OS=Homo sapiens GN=OPA1 O60313 PE=1 SV=3 P31025 Lipocalin-1 OS=Homo sapiens GN=LCN1 PE=1 SV=1 Q6L8Q7 2',5'-phosphodiesterase 12 OS=Homo sapiens GN=PDE12 PE=1 SV=2 O00629 Importin subunit alpha-3 OS=Homo sapiens GN=KPNA4 PE=1 SV=1 P10768 S-formylglutathione hydrolase OS=Homo sapiens GN=ESD PE=1 SV=2 39S ribosomal protein L1, mitochondrial OS=Homo sapiens GN=MRPL1 Q9BYD6 PE=1 SV=2 28S ribosomal protein S31, mitochondrial OS=Homo sapiens GN=MRPS31 Q92665 PE=1 SV=3 Nicotinamide phosphoribosyltransferase OS=Homo sapiens GN=NAMPT P43490 PE=1 SV=1 39S ribosomal protein L50, mitochondrial OS=Homo sapiens GN=MRPL50 Q8N5N7 PE=1 SV=2 Dihydropyrimidinase-related protein 2 OS=Homo sapiens GN=DPYSL2 Q16555 PE=1 SV=1 Long-chain-fatty-acid--CoA ligase 3 OS=Homo sapiens GN=ACSL3 PE=1 O95573 SV=3 Proline synthase co-transcribed bacterial homolog protein OS=Homo O94903 sapiens GN=PROSC PE=1 SV=1 Mitochondrial import receptor subunit TOM34 OS=Homo sapiens Q15785 GN=TOMM34 PE=1 SV=2 Zinc finger Ran-binding domain-containing protein 2 OS=Homo sapiens O95218 GN=ZRANB2 PE=1 SV=2 Dihydrolipoyl dehydrogenase, mitochondrial OS=Homo sapiens GN=DLD P09622 PE=1 SV=2 Eukaryotic translation initiation factor 1A, X-chromosomal OS=Homo sapiens P47813 GN=EIF1AX PE=1 SV=2

235 Accession No. PER2 protein interaction in hypoxic metabolism Trans-2-enoyl-CoA reductase, mitochondrial OS=Homo sapiens GN=MECR Q9BV79 PE=1 SV=2 O00410 Importin-5 OS=Homo sapiens GN=IPO5 PE=1 SV=4 Q8NBN7 Retinol dehydrogenase 13 OS=Homo sapiens GN=RDH13 PE=1 SV=2 Acidic leucine-rich nuclear phosphoprotein 32 family member B OS=Homo Q92688 sapiens GN=ANP32B PE=1 SV=1 Eukaryotic translation initiation factor 3 subunit G OS=Homo sapiens O75821 GN=EIF3G PE=1 SV=2 Hepatoma-derived growth factor-related protein 2 OS=Homo sapiens Q7Z4V5 GN=HDGFRP2 PE=1 SV=1 28S ribosomal protein S10, mitochondrial OS=Homo sapiens GN=MRPS10 P82664 PE=1 SV=2 Succinyl-CoA ligase [ADP/GDP-forming] subunit alpha, mitochondrial P53597 OS=Homo sapiens GN=SUCLG1 PE=1 SV=4 THUMP domain-containing protein 1 OS=Homo sapiens GN=THUMPD1 Q9NXG2 PE=1 SV=2 4F2 cell-surface antigen heavy chain OS=Homo sapiens GN=SLC3A2 PE=1 P08195 SV=3 Acyl-coenzyme A thioesterase 13 OS=Homo sapiens GN=ACOT13 PE=1 Q9NPJ3 SV=1 39S ribosomal protein L30, mitochondrial OS=Homo sapiens GN=MRPL30 Q8TCC3 PE=1 SV=1 Isocitrate dehydrogenase [NAD] subunit gamma, mitochondrial P51553 OS=Homo sapiens GN=IDH3G PE=1 SV=1 ATP-dependent 6-phosphofructokinase, platelet type OS=Homo sapiens Q01813 GN=PFKP PE=1 SV=2 P09429 High mobility group protein B1 OS=Homo sapiens GN=HMGB1 PE=1 SV=3 39S ribosomal protein L4, mitochondrial OS=Homo sapiens GN=MRPL4 Q9BYD3 PE=1 SV=1 P10909 Clusterin OS=Homo sapiens GN=CLU PE=1 SV=1 O00203 AP-3 complex subunit beta-1 OS=Homo sapiens GN=AP3B1 PE=1 SV=3 Coiled-coil domain-containing protein 47 OS=Homo sapiens GN=CCDC47 Q96A33 PE=1 SV=1 Polymeric immunoglobulin receptor OS=Homo sapiens GN=PIGR PE=1 P01833 SV=4 28S ribosomal protein S16, mitochondrial OS=Homo sapiens GN=MRPS16 Q9Y3D3 PE=1 SV=1 28S ribosomal protein S18a, mitochondrial OS=Homo sapiens Q9NVS2 GN=MRPS18A PE=1 SV=1 Acetyl-CoA acetyltransferase, cytosolic OS=Homo sapiens GN=ACAT2 Q9BWD1 PE=1 SV=2 Q15323 Keratin, type I cuticular Ha1 OS=Homo sapiens GN=KRT31 PE=2 SV=3 39S ribosomal protein L55, mitochondrial OS=Homo sapiens GN=MRPL55 Q7Z7F7 PE=1 SV=1 Q13509 Tubulin beta-3 chain OS=Homo sapiens GN=TUBB3 PE=1 SV=2 28S ribosomal protein S18b, mitochondrial OS=Homo sapiens Q9Y676 GN=MRPS18B PE=1 SV=1

236 Accession No. PER2 protein interaction in hypoxic metabolism Inorganic pyrophosphatase 2, mitochondrial OS=Homo sapiens GN=PPA2 Q9H2U2 PE=1 SV=2 Mitochondrial import inner membrane translocase subunit TIM16 OS=Homo Q9Y3D7 sapiens GN=PAM16 PE=1 SV=2 Q9BWM7 Sideroflexin-3 OS=Homo sapiens GN=SFXN3 PE=1 SV=2 Q16378 Proline-rich protein 4 OS=Homo sapiens GN=PRR4 PE=1 SV=3 Mitochondrial import inner membrane translocase subunit TIM50 OS=Homo Q3ZCQ8 sapiens GN=TIMM50 PE=1 SV=2 Sarcoplasmic/endoplasmic reticulum calcium ATPase 2 OS=Homo sapiens P16615 GN=ATP2A2 PE=1 SV=1 P32119 Peroxiredoxin-2 OS=Homo sapiens GN=PRDX2 PE=1 SV=5 Q9H4M9 EH domain-containing protein 1 OS=Homo sapiens GN=EHD1 PE=1 SV=2 O43583 Density-regulated protein OS=Homo sapiens GN=DENR PE=1 SV=2 Peptidyl-tRNA hydrolase ICT1, mitochondrial OS=Homo sapiens GN=ICT1 Q14197 PE=1 SV=1 Translocation protein SEC63 homolog OS=Homo sapiens GN=SEC63 PE=1 Q9UGP8 SV=2 P10599 Thioredoxin OS=Homo sapiens GN=TXN PE=1 SV=3 Medium-chain specific acyl-CoA dehydrogenase, mitochondrial OS=Homo P11310 sapiens GN=ACADM PE=1 SV=1 39S ribosomal protein L21, mitochondrial OS=Homo sapiens GN=MRPL21 Q7Z2W9 PE=1 SV=2 Protein transport protein Sec31A OS=Homo sapiens GN=SEC31A PE=1 O94979 SV=3 Synaptic vesicle membrane protein VAT-1 homolog OS=Homo sapiens Q99536 GN=VAT1 PE=1 SV=2 P04350 Tubulin beta-4A chain OS=Homo sapiens GN=TUBB4A PE=1 SV=2 P01876 Ig alpha-1 chain C region OS=Homo sapiens GN=IGHA1 PE=1 SV=2 P41250 Glycine--tRNA ligase OS=Homo sapiens GN=GARS PE=1 SV=3 Polyribonucleotide nucleotidyltransferase 1, mitochondrial OS=Homo Q8TCS8 sapiens GN=PNPT1 PE=1 SV=2 28S ribosomal protein S26, mitochondrial OS=Homo sapiens GN=MRPS26 Q9BYN8 PE=1 SV=1 Aspartate aminotransferase, mitochondrial OS=Homo sapiens GN=GOT2 P00505 PE=1 SV=3 P61769 Beta-2-microglobulin OS=Homo sapiens GN=B2M PE=1 SV=1 Q8TEA8 D-tyrosyl-tRNA(Tyr) deacylase 1 OS=Homo sapiens GN=DTD1 PE=1 SV=2 Q7Z794 Keratin, type II cytoskeletal 1b OS=Homo sapiens GN=KRT77 PE=2 SV=3 P78385 Keratin, type II cuticular Hb3 OS=Homo sapiens GN=KRT83 PE=1 SV=2 Serine/threonine-protein phosphatase 4 regulatory subunit 3A OS=Homo Q6IN85 sapiens GN=SMEK1 PE=1 SV=1 P08727 Keratin, type I cytoskeletal 19 OS=Homo sapiens GN=KRT19 PE=1 SV=4 P12273 Prolactin-inducible protein OS=Homo sapiens GN=PIP PE=1 SV=1 O75323 Protein NipSnap homolog 2 OS=Homo sapiens GN=GBAS PE=1 SV=1

237 Accession No. PER2 protein interaction in hypoxic metabolism 39S ribosomal protein L43, mitochondrial OS=Homo sapiens GN=MRPL43 Q8N983 PE=1 SV=1 P62917 60S ribosomal protein L8 OS=Homo sapiens GN=RPL8 PE=1 SV=2 P20290 Transcription factor BTF3 OS=Homo sapiens GN=BTF3 PE=1 SV=1 Q9UBM7 7-dehydrocholesterol reductase OS=Homo sapiens GN=DHCR7 PE=1 SV=1 Sodium/potassium-transporting ATPase subunit alpha-1 OS=Homo sapiens P05023 GN=ATP1A1 PE=1 SV=1 Cat eye syndrome critical region protein 5 OS=Homo sapiens GN=CECR5 Q9BXW7 PE=1 SV=1 Ribonucleoside-diphosphate reductase large subunit OS=Homo sapiens P23921 GN=RRM1 PE=1 SV=1 Dual specificity protein phosphatase 23 OS=Homo sapiens GN=DUSP23 Q9BVJ7 PE=1 SV=1 Leucine--tRNA ligase, cytoplasmic OS=Homo sapiens GN=LARS PE=1 Q9P2J5 SV=2 Isocitrate dehydrogenase [NAD] subunit alpha, mitochondrial P50213 OS=Homo sapiens GN=IDH3A PE=1 SV=1 P01024 Complement C3 OS=Homo sapiens GN=C3 PE=1 SV=2 ADP-ribosylation factor-like protein 8B OS=Homo sapiens GN=ARL8B PE=1 Q9NVJ2 SV=1 39S ribosomal protein L22, mitochondrial OS=Homo sapiens GN=MRPL22 Q9NWU5 PE=1 SV=1 Q99935 Proline-rich protein 1 OS=Homo sapiens GN=PROL1 PE=1 SV=2 Protein-L-isoaspartate O-methyltransferase domain-containing protein 1 Q96MG8 OS=Homo sapiens GN=PCMTD1 PE=2 SV=2 Cytochrome c oxidase subunit 6B1 OS=Homo sapiens GN=COX6B1 P14854 PE=1 SV=2 NADH-cytochrome b5 reductase 3 OS=Homo sapiens GN=CYB5R3 P00387 PE=1 SV=3 P22830 Ferrochelatase, mitochondrial OS=Homo sapiens GN=FECH PE=1 SV=2 Peptidyl-prolyl cis-trans isomerase CWC27 homolog OS=Homo sapiens Q6UX04 GN=CWC27 PE=1 SV=1 Glutaryl-CoA dehydrogenase, mitochondrial OS=Homo sapiens GN=GCDH Q92947 PE=1 SV=1 Q9BRP8 Partner of Y14 and mago OS=Homo sapiens GN=WIBG PE=1 SV=1 WASH complex subunit strumpellin OS=Homo sapiens GN=KIAA0196 PE=1 Q12768 SV=1 Q9BUF5 Tubulin beta-6 chain OS=Homo sapiens GN=TUBB6 PE=1 SV=1 Ubiquinol-cytochrome-c reductase complex assembly factor 1 Q9NVA1 OS=Homo sapiens GN=UQCC1 PE=1 SV=3 EKC/KEOPS complex subunit TPRKB OS=Homo sapiens GN=TPRKB PE=1 Q9Y3C4 SV=1 P61020 Ras-related protein Rab-5B OS=Homo sapiens GN=RAB5B PE=1 SV=1 NADH dehydrogenase [ubiquinone] 1 alpha subcomplex assembly factor 4 Q9P032 OS=Homo sapiens GN=NDUFAF4 PE=1 SV=1 GMP synthase [glutamine-hydrolyzing] OS=Homo sapiens GN=GMPS PE=1 P49915 SV=1

238 Accession No. PER2 protein interaction in hypoxic metabolism tRNA (guanine-N(7)-)-methyltransferase OS=Homo sapiens GN=METTL1 Q9UBP6 PE=1 SV=1 ATP synthase subunit g, mitochondrial OS=Homo sapiens GN=ATP5L O75964 PE=1 SV=3 Adipocyte plasma membrane-associated protein OS=Homo sapiens Q9HDC9 GN=APMAP PE=1 SV=2 28S ribosomal protein S25, mitochondrial OS=Homo sapiens GN=MRPS25 P82663 PE=1 SV=1 D-beta-hydroxybutyrate dehydrogenase, mitochondrial OS=Homo sapiens Q02338 GN=BDH1 PE=1 SV=3 O75556 Mammaglobin-B OS=Homo sapiens GN=SCGB2A1 PE=1 SV=1 P01834 Ig kappa chain C region OS=Homo sapiens GN=IGKC PE=1 SV=1 Q9Y3D6 Mitochondrial fission 1 protein OS=Homo sapiens GN=FIS1 PE=1 SV=2 Thioredoxin domain-containing protein 12 OS=Homo sapiens GN=TXNDC12 O95881 PE=1 SV=1 Q8N1N4 Keratin, type II cytoskeletal 78 OS=Homo sapiens GN=KRT78 PE=2 SV=2 Q6UW78 UPF0723 protein C11orf83 OS=Homo sapiens GN=C11orf83 PE=1 SV=2 Peroxisomal acyl-coenzyme A oxidase 1 OS=Homo sapiens GN=ACOX1 Q15067 PE=1 SV=3 P50443 Sulfate transporter OS=Homo sapiens GN=SLC26A2 PE=1 SV=2 Proteasome subunit alpha type-4 OS=Homo sapiens GN=PSMA4 PE=1 P25789 SV=1 P52294 Importin subunit alpha-5 OS=Homo sapiens GN=KPNA1 PE=1 SV=3 39S ribosomal protein L2, mitochondrial OS=Homo sapiens GN=MRPL2 Q5T653 PE=1 SV=2 39S ribosomal protein L51, mitochondrial OS=Homo sapiens GN=MRPL51 Q4U2R6 PE=1 SV=1 P68366 Tubulin alpha-4A chain OS=Homo sapiens GN=TUBA4A PE=1 SV=1 Trifunctional purine biosynthetic protein adenosine-3 OS=Homo sapiens P22102 GN=GART PE=1 SV=1 28S ribosomal protein S18c, mitochondrial OS=Homo sapiens Q9Y3D5 GN=MRPS18C PE=1 SV=1 Microtubule-associated protein RP/EB family member 1 OS=Homo sapiens Q15691 GN=MAPRE1 PE=1 SV=3 Peptidyl-prolyl cis-trans isomerase FKBP5 OS=Homo sapiens GN=FKBP5 Q13451 PE=1 SV=2 Ribose-phosphate pyrophosphokinase 2 OS=Homo sapiens GN=PRPS2 P11908 PE=1 SV=2 Isocitrate dehydrogenase [NAD] subunit beta, mitochondrial OS=Homo O43837 sapiens GN=IDH3B PE=1 SV=2 Zymogen granule protein 16 homolog B OS=Homo sapiens GN=ZG16B Q96DA0 PE=1 SV=3 COX assembly mitochondrial protein homolog OS=Homo sapiens Q7Z7K0 GN=CMC1 PE=1 SV=1 Tetratricopeptide repeat protein 9C OS=Homo sapiens GN=TTC9C PE=1 Q8N5M4 SV=1

239 Accession No. PER2 protein interaction in hypoxic metabolism Asparagine--tRNA ligase, cytoplasmic OS=Homo sapiens GN=NARS PE=1 O43776 SV=1 Q9Y4B6 Protein VPRBP OS=Homo sapiens GN=VPRBP PE=1 SV=3 P04818 Thymidylate synthase OS=Homo sapiens GN=TYMS PE=1 SV=3 Serine/threonine-protein phosphatase 6 catalytic subunit OS=Homo sapiens O00743 GN=PPP6C PE=1 SV=1 P46459 Vesicle-fusing ATPase OS=Homo sapiens GN=NSF PE=1 SV=3 39S ribosomal protein L32, mitochondrial OS=Homo sapiens GN=MRPL32 Q9BYC8 PE=1 SV=1 P06702 Protein S100-A9 OS=Homo sapiens GN=S100A9 PE=1 SV=1 V-type proton ATPase subunit G 1 OS=Homo sapiens GN=ATP6V1G1 PE=1 O75348 SV=3 Q9BSJ8 Extended synaptotagmin-1 OS=Homo sapiens GN=ESYT1 PE=1 SV=1 P26640 Valine--tRNA ligase OS=Homo sapiens GN=VARS PE=1 SV=4 U6 snRNA-associated Sm-like protein LSm7 OS=Homo sapiens GN=LSM7 Q9UK45 PE=1 SV=1 39S ribosomal protein L23, mitochondrial OS=Homo sapiens GN=MRPL23 Q16540 PE=1 SV=1 P27816 Microtubule-associated protein 4 OS=Homo sapiens GN=MAP4 PE=1 SV=3 Sterol-4-alpha-carboxylate 3-dehydrogenase, decarboxylating OS=Homo Q15738 sapiens GN=NSDHL PE=1 SV=2 Gamma-aminobutyric acid receptor-associated protein-like 2 OS=Homo P60520 sapiens GN=GABARAPL2 PE=1 SV=1 Transmembrane protein 126A OS=Homo sapiens GN=TMEM126A PE=1 Q9H061 SV=1 39S ribosomal protein L3, mitochondrial OS=Homo sapiens GN=MRPL3 P09001 PE=1 SV=1 Isocitrate dehydrogenase [NADP], mitochondrial OS=Homo sapiens P48735 GN=IDH2 PE=1 SV=2 39S ribosomal protein L10, mitochondrial OS=Homo sapiens GN=MRPL10 Q7Z7H8 PE=1 SV=3 Proteasome subunit alpha type-6 OS=Homo sapiens GN=PSMA6 PE=1 P60900 SV=1 Q9BRQ8 Apoptosis-inducing factor 2 OS=Homo sapiens GN=AIFM2 PE=1 SV=1 Enoyl-CoA hydratase, mitochondrial OS=Homo sapiens GN=ECHS1 PE=1 P30084 SV=4 GrpE protein homolog 1, mitochondrial OS=Homo sapiens GN=GRPEL1 Q9HAV7 PE=1 SV=2 CDGSH iron-sulfur domain-containing protein 2 OS=Homo sapiens Q8N5K1 GN=CISD2 PE=1 SV=1 P49721 Proteasome subunit beta type-2 OS=Homo sapiens GN=PSMB2 PE=1 SV=1 Pentatricopeptide repeat domain-containing protein 3, mitochondrial Q96EY7 OS=Homo sapiens GN=PTCD3 PE=1 SV=3 P06744 Glucose-6-phosphate isomerase OS=Homo sapiens GN=GPI PE=1 SV=4 Replication protein A 14 kDa subunit OS=Homo sapiens GN=RPA3 PE=1 P35244 SV=1

240 Accession No. PER2 protein interaction in hypoxic metabolism Eukaryotic translation initiation factor 1 OS=Homo sapiens GN=EIF1 PE=1 P41567 SV=1 Q08257 Quinone oxidoreductase OS=Homo sapiens GN=CRYZ PE=1 SV=1 P22307 Non-specific lipid-transfer protein OS=Homo sapiens GN=SCP2 PE=1 SV=2 RNA 3'-terminal phosphate cyclase OS=Homo sapiens GN=RTCA PE=1 O00442 SV=1 Regulator of nonsense transcripts 1 OS=Homo sapiens GN=UPF1 PE=1 Q92900 SV=2 39S ribosomal protein L33, mitochondrial OS=Homo sapiens GN=MRPL33 O75394 PE=1 SV=1 28S ribosomal protein S24, mitochondrial OS=Homo sapiens GN=MRPS24 Q96EL2 PE=1 SV=1 Q96KG9 N-terminal kinase-like protein OS=Homo sapiens GN=SCYL1 PE=1 SV=1 U4/U6 small nuclear ribonucleoprotein Prp31 OS=Homo sapiens Q8WWY3 GN=PRPF31 PE=1 SV=2 Signal recognition particle 19 kDa protein OS=Homo sapiens GN=SRP19 P09132 PE=1 SV=3 Q9UHR5 SAP30-binding protein OS=Homo sapiens GN=SAP30BP PE=1 SV=1 Lanosterol 14-alpha demethylase OS=Homo sapiens GN=CYP51A1 PE=1 Q16850 SV=3 Activator of 90 kDa heat shock protein ATPase homolog 1 OS=Homo O95433 sapiens GN=AHSA1 PE=1 SV=1 Translation initiation factor eIF-2B subunit alpha OS=Homo sapiens Q14232 GN=EIF2B1 PE=1 SV=1 Serine hydroxymethyltransferase, mitochondrial OS=Homo sapiens P34897 GN=SHMT2 PE=1 SV=3 Hereditary hemochromatosis protein OS=Homo sapiens GN=HFE PE=1 Q30201 SV=1 tRNA pseudouridine synthase A, mitochondrial OS=Homo sapiens Q9Y606 GN=PUS1 PE=1 SV=3 Immunoglobulin lambda-like polypeptide 5 OS=Homo sapiens GN=IGLL5 B9A064 PE=2 SV=2 Q9Y446 Plakophilin-3 OS=Homo sapiens GN=PKP3 PE=1 SV=1 P06396 Gelsolin OS=Homo sapiens GN=GSN PE=1 SV=1 P31947 14-3-3 protein sigma OS=Homo sapiens GN=SFN PE=1 SV=1 Monocarboxylate transporter 4 OS=Homo sapiens GN=SLC16A3 PE=1 O15427 SV=1 Cytochrome b-c1 complex subunit 9 OS=Homo sapiens GN=UQCR10 Q9UDW1 PE=1 SV=3 Vacuolar protein sorting-associated protein 4B OS=Homo sapiens O75351 GN=VPS4B PE=1 SV=2 Aldehyde dehydrogenase X, mitochondrial OS=Homo sapiens P30837 GN=ALDH1B1 PE=1 SV=3 Regulator of microtubule dynamics protein 1 OS=Homo sapiens GN=RMDN1 Q96DB5 PE=1 SV=1 O60613 15 kDa selenoprotein OS=Homo sapiens GN=SEP15 PE=1 SV=3

241 Accession No. PER2 protein interaction in hypoxic metabolism Tyrosine--tRNA ligase, mitochondrial OS=Homo sapiens GN=YARS2 PE=1 Q9Y2Z4 SV=2 39S ribosomal protein L48, mitochondrial OS=Homo sapiens GN=MRPL48 Q96GC5 PE=1 SV=2 Transcription elongation factor SPT5 OS=Homo sapiens GN=SUPT5H PE=1 O00267 SV=1 P30046 D-dopachrome decarboxylase OS=Homo sapiens GN=DDT PE=1 SV=3 Mitochondrial-processing peptidase subunit alpha OS=Homo sapiens Q10713 GN=PMPCA PE=1 SV=2 Mannose-1-phosphate guanyltransferase alpha OS=Homo sapiens Q96IJ6 GN=GMPPA PE=1 SV=1 Q13835 Plakophilin-1 OS=Homo sapiens GN=PKP1 PE=1 SV=2 P55058 Phospholipid transfer protein OS=Homo sapiens GN=PLTP PE=1 SV=1 P27348 14-3-3 protein theta OS=Homo sapiens GN=YWHAQ PE=1 SV=1 P25311 Zinc-alpha-2-glycoprotein OS=Homo sapiens GN=AZGP1 PE=1 SV=2 Q9BQA1 Methylosome protein 50 OS=Homo sapiens GN=WDR77 PE=1 SV=1 Uridine 5'-monophosphate synthase OS=Homo sapiens GN=UMPS PE=1 P11172 SV=1 Dihydrolipoyllysine-residue acetyltransferase component of pyruvate dehydrogenase complex, mitochondrial OS=Homo sapiens GN=DLAT PE=1 P10515 SV=3 3-ketoacyl-CoA thiolase, mitochondrial OS=Homo sapiens GN=ACAA2 PE=1 P42765 SV=2 Alpha/beta hydrolase domain-containing protein 11 OS=Homo sapiens Q8NFV4 GN=ABHD11 PE=2 SV=1 Q10567 AP-1 complex subunit beta-1 OS=Homo sapiens GN=AP1B1 PE=1 SV=2 P17812 CTP synthase 1 OS=Homo sapiens GN=CTPS1 PE=1 SV=2 P16152 Carbonyl reductase [NADPH] 1 OS=Homo sapiens GN=CBR1 PE=1 SV=3 Q9H875 PRKR-interacting protein 1 OS=Homo sapiens GN=PRKRIP1 PE=1 SV=1 39S ribosomal protein L20, mitochondrial OS=Homo sapiens GN=MRPL20 Q9BYC9 PE=1 SV=1 P63172 Dynein light chain Tctex-type 1 OS=Homo sapiens GN=DYNLT1 PE=1 SV=1 P23527 Histone H2B type 1-O OS=Homo sapiens GN=HIST1H2BO PE=1 SV=3 39S ribosomal protein L24, mitochondrial OS=Homo sapiens GN=MRPL24 Q96A35 PE=1 SV=1 Peptidyl-prolyl cis-trans isomerase FKBP11 OS=Homo sapiens GN=FKBP11 Q9NYL4 PE=1 SV=1 Eukaryotic translation initiation factor 3 subunit F OS=Homo sapiens O00303 GN=EIF3F PE=1 SV=1 S-adenosylmethionine synthase isoform type-2 OS=Homo sapiens P31153 GN=MAT2A PE=1 SV=1 Q53H12 Acylglycerol kinase, mitochondrial OS=Homo sapiens GN=AGK PE=1 SV=2 P27824 Calnexin OS=Homo sapiens GN=CANX PE=1 SV=2 V-type proton ATPase catalytic subunit A OS=Homo sapiens GN=ATP6V1A P38606 PE=1 SV=2

242 Accession No. PER2 protein interaction in hypoxic metabolism Translation initiation factor eIF-2B subunit beta OS=Homo sapiens P49770 GN=EIF2B2 PE=1 SV=3 Q02952 A-kinase anchor protein 12 OS=Homo sapiens GN=AKAP12 PE=1 SV=4 Proteasome subunit alpha type-2 OS=Homo sapiens GN=PSMA2 PE=1 P25787 SV=2 Isovaleryl-CoA dehydrogenase, mitochondrial OS=Homo sapiens GN=IVD P26440 PE=1 SV=1 Q15382 GTP-binding protein Rheb OS=Homo sapiens GN=RHEB PE=1 SV=1 Acidic leucine-rich nuclear phosphoprotein 32 family member E OS=Homo Q9BTT0 sapiens GN=ANP32E PE=1 SV=1 Coiled-coil domain-containing protein 124 OS=Homo sapiens GN=CCDC124 Q96CT7 PE=1 SV=1 P00450 Ceruloplasmin OS=Homo sapiens GN=CP PE=1 SV=1 Leucine-rich repeat-containing protein 15 OS=Homo sapiens GN=LRRC15 Q8TF66 PE=1 SV=2 Q9UHG3 Prenylcysteine oxidase 1 OS=Homo sapiens GN=PCYOX1 PE=1 SV=3 Translation machinery-associated protein 7 OS=Homo sapiens GN=TMA7 Q9Y2S6 PE=1 SV=1 Q9BZL1 Ubiquitin-like protein 5 OS=Homo sapiens GN=UBL5 PE=1 SV=1 P84085 ADP-ribosylation factor 5 OS=Homo sapiens GN=ARF5 PE=1 SV=2 Ragulator complex protein LAMTOR5 OS=Homo sapiens GN=LAMTOR5 O43504 PE=1 SV=1 26S proteasome non-ATPase regulatory subunit 14 OS=Homo sapiens O00487 GN=PSMD14 PE=1 SV=1 Serum paraoxonase/arylesterase 2 OS=Homo sapiens GN=PON2 PE=1 Q15165 SV=3 Protein tyrosine phosphatase type IVA 2 OS=Homo sapiens GN=PTP4A2 Q12974 PE=1 SV=1 Q9UL25 Ras-related protein Rab-21 OS=Homo sapiens GN=RAB21 PE=1 SV=3 Signal peptidase complex subunit 3 OS=Homo sapiens GN=SPCS3 PE=1 P61009 SV=1 Probable Xaa-Pro aminopeptidase 3 OS=Homo sapiens GN=XPNPEP3 Q9NQH7 PE=1 SV=1 ATP-dependent Clp protease ATP-binding subunit clpX-like, mitochondrial O76031 OS=Homo sapiens GN=CLPX PE=1 SV=2 P31949 Protein S100-A11 OS=Homo sapiens GN=S100A11 PE=1 SV=2 Q86YS7 C2 domain-containing protein 5 OS=Homo sapiens GN=C2CD5 PE=1 SV=1 Q5T749 Keratinocyte proline-rich protein OS=Homo sapiens GN=KPRP PE=1 SV=1 Q9HA64 Ketosamine-3-kinase OS=Homo sapiens GN=FN3KRP PE=1 SV=2 Mitochondrial import inner membrane translocase subunit Tim10 B Q9Y5J6 OS=Homo sapiens GN=TIMM10B PE=1 SV=1 Microfibrillar-associated protein 1 OS=Homo sapiens GN=MFAP1 PE=1 P55081 SV=2 Deleted in malignant brain tumors 1 protein OS=Homo sapiens GN=DMBT1 Q9UGM3 PE=1 SV=2

243 Accession No. PER2 protein interaction in hypoxic metabolism Neutral amino acid transporter B(0) OS=Homo sapiens GN=SLC1A5 PE=1 Q15758 SV=2 P57088 Transmembrane protein 33 OS=Homo sapiens GN=TMEM33 PE=1 SV=2 Transcription elongation factor B polypeptide 1 OS=Homo sapiens Q15369 GN=TCEB1 PE=1 SV=1 26S protease regulatory subunit 10B OS=Homo sapiens GN=PSMC6 PE=1 P62333 SV=1 Protein transport protein Sec61 subunit beta OS=Homo sapiens P60468 GN=SEC61B PE=1 SV=2 Q9H9B4 Sideroflexin-1 OS=Homo sapiens GN=SFXN1 PE=1 SV=4 O95757 Heat shock 70 kDa protein 4L OS=Homo sapiens GN=HSPA4L PE=1 SV=3 Q08380 Galectin-3-binding protein OS=Homo sapiens GN=LGALS3BP PE=1 SV=1 Q92764 Keratin, type I cuticular Ha5 OS=Homo sapiens GN=KRT35 PE=2 SV=5 NADH dehydrogenase [ubiquinone] complex I, assembly factor 7 OS=Homo Q7L592 sapiens GN=NDUFAF7 PE=1 SV=1 Q2M389 WASH complex subunit 7 OS=Homo sapiens GN=KIAA1033 PE=1 SV=2 2-amino-3-ketobutyrate coenzyme A ligase, mitochondrial OS=Homo O75600 sapiens GN=GCAT PE=1 SV=1 Q9H3H3 UPF0696 protein C11orf68 OS=Homo sapiens GN=C11orf68 PE=1 SV=2 Mitochondrial glutamate carrier 1 OS=Homo sapiens GN=SLC25A22 PE=1 Q9H936 SV=1 ADP-ribosylation factor-related protein 1 OS=Homo sapiens GN=ARFRP1 Q13795 PE=1 SV=1 Mannose-P-dolichol utilization defect 1 protein OS=Homo sapiens O75352 GN=MPDU1 PE=1 SV=2 39S ribosomal protein L54, mitochondrial OS=Homo sapiens GN=MRPL54 Q6P161 PE=1 SV=1 P26447 Protein S100-A4 OS=Homo sapiens GN=S100A4 PE=1 SV=1 Q9H0U4 Ras-related protein Rab-1B OS=Homo sapiens GN=RAB1B PE=1 SV=1 Q13885 Tubulin beta-2A chain OS=Homo sapiens GN=TUBB2A PE=1 SV=1 P61026 Ras-related protein Rab-10 OS=Homo sapiens GN=RAB10 PE=1 SV=1 Q9NSB4 Keratin, type II cuticular Hb2 OS=Homo sapiens GN=KRT82 PE=1 SV=3 5'-AMP-activated protein kinase subunit gamma-1 OS=Homo sapiens P54619 GN=PRKAG1 PE=1 SV=1 Acylpyruvase FAHD1, mitochondrial OS=Homo sapiens GN=FAHD1 PE=1 Q6P587 SV=2 Q9Y4L1 Hypoxia up-regulated protein 1 OS=Homo sapiens GN=HYOU1 PE=1 SV=1 P11802 Cyclin-dependent kinase 4 OS=Homo sapiens GN=CDK4 PE=1 SV=2 P20340 Ras-related protein Rab-6A OS=Homo sapiens GN=RAB6A PE=1 SV=3 Q13085 Acetyl-CoA carboxylase 1 OS=Homo sapiens GN=ACACA PE=1 SV=2 28S ribosomal protein S11, mitochondrial OS=Homo sapiens GN=MRPS11 P82912 PE=1 SV=2 P49366 Deoxyhypusine synthase OS=Homo sapiens GN=DHPS PE=1 SV=1 A6NDG6 Phosphoglycolate phosphatase OS=Homo sapiens GN=PGP PE=1 SV=1

244 Accession No. PER2 protein interaction in hypoxic metabolism ATP-dependent 6-phosphofructokinase, liver type OS=Homo sapiens P17858 GN=PFKL PE=1 SV=6 Alpha-soluble NSF attachment protein OS=Homo sapiens GN=NAPA PE=1 P54920 SV=3 ATP-dependent 6-phosphofructokinase, muscle type OS=Homo sapiens P08237 GN=PFKM PE=1 SV=2 Cytochrome c oxidase assembly protein COX16 homolog, Q9P0S2 mitochondrial OS=Homo sapiens GN=COX16 PE=1 SV=1 Q96S44 TP53-regulating kinase OS=Homo sapiens GN=TP53RK PE=1 SV=2 U6 snRNA-associated Sm-like protein LSm8 OS=Homo sapiens GN=LSM8 O95777 PE=1 SV=3 Eukaryotic translation initiation factor 3 subunit J OS=Homo sapiens O75822 GN=EIF3J PE=1 SV=2 Secretoglobin family 1D member 1 OS=Homo sapiens GN=SCGB1D1 PE=1 O95968 SV=1 P49247 Ribose-5-phosphate isomerase OS=Homo sapiens GN=RPIA PE=1 SV=3 28S ribosomal protein S5, mitochondrial OS=Homo sapiens GN=MRPS5 P82675 PE=1 SV=2 U6 snRNA-associated Sm-like protein LSm3 OS=Homo sapiens GN=LSM3 P62310 PE=1 SV=2 Q15286 Ras-related protein Rab-35 OS=Homo sapiens GN=RAB35 PE=1 SV=1 P02452 Collagen alpha-1(I) chain OS=Homo sapiens GN=COL1A1 PE=1 SV=5 Probable cytosolic iron-sulfur protein assembly protein CIAO1 OS=Homo O76071 sapiens GN=CIAO1 PE=1 SV=1 Thioredoxin-dependent peroxide reductase, mitochondrial OS=Homo P30048 sapiens GN=PRDX3 PE=1 SV=3 P20962 Parathymosin OS=Homo sapiens GN=PTMS PE=1 SV=2 Cytochrome c oxidase assembly factor 6 homolog OS=Homo sapiens Q5JTJ3 GN=COA6 PE=1 SV=1 Q15527 Surfeit locus protein 2 OS=Homo sapiens GN=SURF2 PE=1 SV=3 Q6MZM0 Hephaestin-like protein 1 OS=Homo sapiens GN=HEPHL1 PE=2 SV=2 RNA polymerase-associated protein RTF1 homolog OS=Homo sapiens Q92541 GN=RTF1 PE=1 SV=4 DNA-directed RNA polymerase II subunit RPB1 OS=Homo sapiens P24928 GN=POLR2A PE=1 SV=2 P01037 Cystatin-SN OS=Homo sapiens GN=CST1 PE=1 SV=3 Q13228 Selenium-binding protein 1 OS=Homo sapiens GN=SELENBP1 PE=1 SV=2 Cytochrome c oxidase subunit 7A-related protein, mitochondrial O14548 OS=Homo sapiens GN=COX7A2L PE=1 SV=2 Myosin regulatory light polypeptide 9 OS=Homo sapiens GN=MYL9 PE=1 P24844 SV=4 Q92626 Peroxidasin homolog OS=Homo sapiens GN=PXDN PE=1 SV=2 V-type proton ATPase subunit B, brain isoform OS=Homo sapiens P21281 GN=ATP6V1B2 PE=1 SV=3 Q03426 Mevalonate kinase OS=Homo sapiens GN=MVK PE=1 SV=1

245 Accession No. PER2 protein interaction in hypoxic metabolism Carnitine O-palmitoyltransferase 2, mitochondrial OS=Homo sapiens P23786 GN=CPT2 PE=1 SV=2 Protein transport protein Sec61 subunit alpha isoform 1 OS=Homo sapiens P61619 GN=SEC61A1 PE=1 SV=2 Q99439 Calponin-2 OS=Homo sapiens GN=CNN2 PE=1 SV=4 P27144 Adenylate kinase 4, mitochondrial OS=Homo sapiens GN=AK4 PE=1 SV=1 Endoplasmic reticulum-Golgi intermediate compartment protein 1 OS=Homo Q969X5 sapiens GN=ERGIC1 PE=1 SV=1 O43592 Exportin-T OS=Homo sapiens GN=XPOT PE=1 SV=2 Methylmalonate-semialdehyde dehydrogenase [acylating], mitochondrial Q02252 OS=Homo sapiens GN=ALDH6A1 PE=1 SV=2 Pachytene checkpoint protein 2 homolog OS=Homo sapiens GN=TRIP13 Q15645 PE=1 SV=2 Q16706 Alpha-mannosidase 2 OS=Homo sapiens GN=MAN2A1 PE=1 SV=2 Glucosamine 6-phosphate N-acetyltransferase OS=Homo sapiens Q96EK6 GN=GNPNAT1 PE=1 SV=1 Long-chain-fatty-acid--CoA ligase 1 OS=Homo sapiens GN=ACSL1 PE=1 P33121 SV=1 Inosine triphosphate pyrophosphatase OS=Homo sapiens GN=ITPA PE=1 Q9BY32 SV=2 Q92572 AP-3 complex subunit sigma-1 OS=Homo sapiens GN=AP3S1 PE=1 SV=1 P23919 Thymidylate kinase OS=Homo sapiens GN=DTYMK PE=1 SV=4 Acyl carrier protein, mitochondrial OS=Homo sapiens GN=NDUFAB1 PE=1 O14561 SV=3 Probable histidine--tRNA ligase, mitochondrial OS=Homo sapiens P49590 GN=HARS2 PE=1 SV=1 P20061 Transcobalamin-1 OS=Homo sapiens GN=TCN1 PE=1 SV=2 P0CG05 Ig lambda-2 chain C regions OS=Homo sapiens GN=IGLC2 PE=1 SV=1 26S proteasome non-ATPase regulatory subunit 9 OS=Homo sapiens O00233 GN=PSMD9 PE=1 SV=3 Q92973 Transportin-1 OS=Homo sapiens GN=TNPO1 PE=1 SV=2 Nuclear protein localization protein 4 homolog OS=Homo sapiens Q8TAT6 GN=NPLOC4 PE=1 SV=3 P04080 Cystatin-B OS=Homo sapiens GN=CSTB PE=1 SV=2 Q16891 MICOS complex subunit MIC60 OS=Homo sapiens GN=IMMT PE=1 SV=1 Phospholipid hydroperoxide glutathione peroxidase, mitochondrial OS=Homo P36969 sapiens GN=GPX4 PE=1 SV=3 P49720 Proteasome subunit beta type-3 OS=Homo sapiens GN=PSMB3 PE=1 SV=2 39S ribosomal protein L49, mitochondrial OS=Homo sapiens GN=MRPL49 Q13405 PE=1 SV=1 Histone-arginine methyltransferase CARM1 OS=Homo sapiens GN=CARM1 Q86X55 PE=1 SV=3 Q15155 Nodal modulator 1 OS=Homo sapiens GN=NOMO1 PE=1 SV=5 Lysophosphatidylcholine acyltransferase 1 OS=Homo sapiens GN=LPCAT1 Q8NF37 PE=1 SV=2

246 Accession No. PER2 protein interaction in hypoxic metabolism Serine/threonine-protein phosphatase 4 catalytic subunit OS=Homo sapiens P60510 GN=PPP4C PE=1 SV=1 P61019 Ras-related protein Rab-2A OS=Homo sapiens GN=RAB2A PE=1 SV=1 Translation initiation factor eIF-2B subunit delta OS=Homo sapiens Q9UI10 GN=EIF2B4 PE=1 SV=2 Protein-glutamine gamma-glutamyltransferase E OS=Homo sapiens Q08188 GN=TGM3 PE=1 SV=4 Q13162 Peroxiredoxin-4 OS=Homo sapiens GN=PRDX4 PE=1 SV=1 Puromycin-sensitive aminopeptidase OS=Homo sapiens GN=NPEPPS PE=1 P55786 SV=2 E3 ubiquitin-protein ligase TRIM21 OS=Homo sapiens GN=TRIM21 PE=1 P19474 SV=1 Q9UJ83 2-hydroxyacyl-CoA lyase 1 OS=Homo sapiens GN=HACL1 PE=1 SV=2 Q9BUP3 Oxidoreductase HTATIP2 OS=Homo sapiens GN=HTATIP2 PE=1 SV=2 Alpha/beta hydrolase domain-containing protein 14B OS=Homo sapiens Q96IU4 GN=ABHD14B PE=1 SV=1 Protein transport protein Sec24B OS=Homo sapiens GN=SEC24B PE=1 O95487 SV=2 P84101 Small EDRK-rich factor 2 OS=Homo sapiens GN=SERF2 PE=1 SV=1 P48960 CD97 antigen OS=Homo sapiens GN=CD97 PE=1 SV=4 L-2-hydroxyglutarate dehydrogenase, mitochondrial OS=Homo sapiens Q9H9P8 GN=L2HGDH PE=1 SV=3 DET1- and DDB1-associated protein 1 OS=Homo sapiens GN=DDA1 PE=1 Q9BW61 SV=1 P20930 Filaggrin OS=Homo sapiens GN=FLG PE=1 SV=3 Q86SJ6 Desmoglein-4 OS=Homo sapiens GN=DSG4 PE=1 SV=1 3-ketoacyl-CoA thiolase, peroxisomal OS=Homo sapiens GN=ACAA1 PE=1 P09110 SV=2 Vasodilator-stimulated phosphoprotein OS=Homo sapiens GN=VASP PE=1 P50552 SV=3 P28072 Proteasome subunit beta type-6 OS=Homo sapiens GN=PSMB6 PE=1 SV=4 Q9NZT1 Calmodulin-like protein 5 OS=Homo sapiens GN=CALML5 PE=1 SV=2 P04259 Keratin, type II cytoskeletal 6B OS=Homo sapiens GN=KRT6B PE=1 SV=5 Angio-associated migratory cell protein OS=Homo sapiens GN=AAMP PE=1 Q13685 SV=2 Protein transport protein Sec23B OS=Homo sapiens GN=SEC23B PE=1 Q15437 SV=2 Vacuolar protein sorting-associated protein 28 homolog OS=Homo sapiens Q9UK41 GN=VPS28 PE=1 SV=1 P04632 Calpain small subunit 1 OS=Homo sapiens GN=CAPNS1 PE=1 SV=1 Q04941 Proteolipid protein 2 OS=Homo sapiens GN=PLP2 PE=1 SV=1 Trafficking protein particle complex subunit 4 OS=Homo sapiens Q9Y296 GN=TRAPPC4 PE=1 SV=1 Q9Y587 AP-4 complex subunit sigma-1 OS=Homo sapiens GN=AP4S1 PE=2 SV=1 Down syndrome critical region protein 3 OS=Homo sapiens GN=DSCR3 O14972 PE=2 SV=1

247 Accession No. PER2 protein interaction in hypoxic metabolism Proteasome-associated protein ECM29 homolog OS=Homo sapiens Q5VYK3 GN=ECM29 PE=1 SV=2 Dihydroxyacetone phosphate acyltransferase OS=Homo sapiens O15228 GN=GNPAT PE=1 SV=1 28 kDa heat- and acid-stable phosphoprotein OS=Homo sapiens Q13442 GN=PDAP1 PE=1 SV=1 Q9Y315 Deoxyribose-phosphate aldolase OS=Homo sapiens GN=DERA PE=1 SV=2 Q9ULA0 Aspartyl aminopeptidase OS=Homo sapiens GN=DNPEP PE=1 SV=1 P19623 Spermidine synthase OS=Homo sapiens GN=SRM PE=1 SV=1 COMM domain-containing protein 6 OS=Homo sapiens GN=COMMD6 PE=1 Q7Z4G1 SV=1 Q6DKJ4 Nucleoredoxin OS=Homo sapiens GN=NXN PE=1 SV=2 SUMO-activating enzyme subunit 1 OS=Homo sapiens GN=SAE1 PE=1 Q9UBE0 SV=1 Isochorismatase domain-containing protein 2, mitochondrial OS=Homo Q96AB3 sapiens GN=ISOC2 PE=1 SV=1 Complex III assembly factor LYRM7 OS=Homo sapiens GN=LYRM7 PE=1 Q5U5X0 SV=1 Q96QA5 Gasdermin-A OS=Homo sapiens GN=GSDMA PE=1 SV=4 Q8IVF2 Protein AHNAK2 OS=Homo sapiens GN=AHNAK2 PE=1 SV=2 Basement membrane-specific heparan sulfate proteoglycan core protein P98160 OS=Homo sapiens GN=HSPG2 PE=1 SV=4 Speckle targeted PIP5K1A-regulated poly(A) polymerase OS=Homo sapiens Q9H6E5 GN=TUT1 PE=1 SV=2 P01620 Ig kappa chain V-III region SIE OS=Homo sapiens PE=1 SV=1 Methylcrotonoyl-CoA carboxylase subunit alpha, mitochondrial OS=Homo Q96RQ3 sapiens GN=MCCC1 PE=1 SV=3 COP9 signalosome complex subunit 7a OS=Homo sapiens GN=COPS7A Q9UBW8 PE=1 SV=1 Serine/threonine-protein phosphatase 2B catalytic subunit alpha isoform Q08209 OS=Homo sapiens GN=PPP3CA PE=1 SV=1 EF-hand domain-containing protein D1 OS=Homo sapiens GN=EFHD1 Q9BUP0 PE=1 SV=1 Mitochondrial-processing peptidase subunit beta OS=Homo sapiens O75439 GN=PMPCB PE=1 SV=2 Q86YQ8 Copine-8 OS=Homo sapiens GN=CPNE8 PE=1 SV=2 Q8IV08 Phospholipase D3 OS=Homo sapiens GN=PLD3 PE=1 SV=1 Glutamate dehydrogenase 1, mitochondrial OS=Homo sapiens GN=GLUD1 P00367 PE=1 SV=2 Rho GDP-dissociation inhibitor 1 OS=Homo sapiens GN=ARHGDIA PE=1 P52565 SV=3 P35613 Basigin OS=Homo sapiens GN=BSG PE=1 SV=2 P29144 Tripeptidyl-peptidase 2 OS=Homo sapiens GN=TPP2 PE=1 SV=4 P61923 Coatomer subunit zeta-1 OS=Homo sapiens GN=COPZ1 PE=1 SV=1 Cytochrome c oxidase assembly factor 1 homolog OS=Homo sapiens Q9GZY4 GN=COA1 PE=1 SV=1

248 Accession No. PER2 protein interaction in hypoxic metabolism Q9Y266 Nuclear migration protein nudC OS=Homo sapiens GN=NUDC PE=1 SV=1 ADP-ribosylation factor-like protein 1 OS=Homo sapiens GN=ARL1 PE=1 P40616 SV=1 Cytochrome c-type heme lyase OS=Homo sapiens GN=HCCS PE=1 P53701 SV=1 3-beta-hydroxysteroid-Delta(8),Delta(7)-isomerase OS=Homo sapiens Q15125 GN=EBP PE=1 SV=3 Q8NC60 Nitric oxide-associated protein 1 OS=Homo sapiens GN=NOA1 PE=1 SV=2 Q9H6V9 UPF0554 protein C2orf43 OS=Homo sapiens GN=C2orf43 PE=1 SV=1 V-type proton ATPase subunit E 1 OS=Homo sapiens GN=ATP6V1E1 PE=1 P36543 SV=1 Proteasomal ubiquitin receptor ADRM1 OS=Homo sapiens GN=ADRM1 Q16186 PE=1 SV=2 Large neutral amino acids transporter small subunit 1 OS=Homo sapiens Q01650 GN=SLC7A5 PE=1 SV=2 E2/E3 hybrid ubiquitin-protein ligase UBE2O OS=Homo sapiens GN=UBE2O Q9C0C9 PE=1 SV=3 Q12849 G-rich sequence factor 1 OS=Homo sapiens GN=GRSF1 PE=1 SV=3 Aconitate hydratase, mitochondrial OS=Homo sapiens GN=ACO2 PE=1 Q99798 SV=2 O14617 AP-3 complex subunit delta-1 OS=Homo sapiens GN=AP3D1 PE=1 SV=1 P01871 Ig mu chain C region OS=Homo sapiens GN=IGHM PE=1 SV=3 Eukaryotic translation initiation factor 3 subunit L OS=Homo sapiens Q9Y262 GN=EIF3L PE=1 SV=1 Q9UBC9 Small proline-rich protein 3 OS=Homo sapiens GN=SPRR3 PE=1 SV=2 P01591 Immunoglobulin J chain OS=Homo sapiens GN=IGJ PE=1 SV=4 Glycylpeptide N-tetradecanoyltransferase 1 OS=Homo sapiens GN=NMT1 P30419 PE=1 SV=2 PIH1 domain-containing protein 1 OS=Homo sapiens GN=PIH1D1 PE=1 Q9NWS0 SV=1 DDB1- and CUL4-associated factor 8 OS=Homo sapiens GN=DCAF8 PE=1 Q5TAQ9 SV=1 P39748 Flap endonuclease 1 OS=Homo sapiens GN=FEN1 PE=1 SV=1 HIG1 domain family member 1A, mitochondrial OS=Homo sapiens Q9Y241 GN=HIGD1A PE=1 SV=1 Q15054 DNA polymerase delta subunit 3 OS=Homo sapiens GN=POLD3 PE=1 SV=2 Proteasome assembly chaperone 2 OS=Homo sapiens GN=PSMG2 PE=1 Q969U7 SV=1 DnaJ homolog subfamily C member 3 OS=Homo sapiens GN=DNAJC3 Q13217 PE=1 SV=1 Voltage-dependent anion-selective channel protein 2 OS=Homo sapiens P45880 GN=VDAC2 PE=1 SV=2 Inosine-5'-monophosphate dehydrogenase 1 OS=Homo sapiens P20839 GN=IMPDH1 PE=1 SV=2 Probable 18S rRNA (guanine-N(7))-methyltransferase OS=Homo sapiens O43709 GN=WBSCR22 PE=1 SV=2

249 Accession No. PER2 protein interaction in hypoxic metabolism P54802 Alpha-N-acetylglucosaminidase OS=Homo sapiens GN=NAGLU PE=1 SV=2 Q9UBV2 Protein sel-1 homolog 1 OS=Homo sapiens GN=SEL1L PE=1 SV=3 DNA replication complex GINS protein PSF2 OS=Homo sapiens GN=GINS2 Q9Y248 PE=1 SV=1 NAD-dependent malic enzyme, mitochondrial OS=Homo sapiens GN=ME2 P23368 PE=1 SV=1 26S proteasome non-ATPase regulatory subunit 12 OS=Homo sapiens O00232 GN=PSMD12 PE=1 SV=3 Acetolactate synthase-like protein OS=Homo sapiens GN=ILVBL PE=1 A1L0T0 SV=2 Q14847 LIM and SH3 domain protein 1 OS=Homo sapiens GN=LASP1 PE=1 SV=2 Phosphoribosyl pyrophosphate synthase-associated protein 1 OS=Homo Q14558 sapiens GN=PRPSAP1 PE=1 SV=2 LETM1 and EF-hand domain-containing protein 1, mitochondrial OS=Homo O95202 sapiens GN=LETM1 PE=1 SV=1 P30566 Adenylosuccinate lyase OS=Homo sapiens GN=ADSL PE=1 SV=2 P17655 Calpain-2 catalytic subunit OS=Homo sapiens GN=CAPN2 PE=1 SV=6 Glutaminase kidney isoform, mitochondrial OS=Homo sapiens GN=GLS O94925 PE=1 SV=1 Peptidyl-prolyl cis-trans isomerase FKBP4 OS=Homo sapiens GN=FKBP4 Q02790 PE=1 SV=3 P80303 Nucleobindin-2 OS=Homo sapiens GN=NUCB2 PE=1 SV=2 Q99497 Protein DJ-1 OS=Homo sapiens GN=PARK7 PE=1 SV=2 Mitochondrial dicarboxylate carrier OS=Homo sapiens GN=SLC25A10 PE=1 Q9UBX3 SV=2 Q9Y6C9 Mitochondrial carrier homolog 2 OS=Homo sapiens GN=MTCH2 PE=1 SV=1 Serine protease HTRA2, mitochondrial OS=Homo sapiens GN=HTRA2 O43464 PE=1 SV=2 Q92615 La-related protein 4B OS=Homo sapiens GN=LARP4B PE=1 SV=3 Signal peptidase complex catalytic subunit SEC11A OS=Homo sapiens P67812 GN=SEC11A PE=1 SV=1 Ubiquitin-like modifier-activating enzyme 1 OS=Homo sapiens GN=UBA1 P22314 PE=1 SV=3 P04183 Thymidine kinase, cytosolic OS=Homo sapiens GN=TK1 PE=1 SV=2 Peroxisomal membrane protein 11B OS=Homo sapiens GN=PEX11B PE=1 O96011 SV=1 COMM domain-containing protein 9 OS=Homo sapiens GN=COMMD9 PE=1 Q9P000 SV=2 Elongation factor Ts, mitochondrial OS=Homo sapiens GN=TSFM PE=1 P43897 SV=2 Superoxide dismutase [Mn], mitochondrial OS=Homo sapiens GN=SOD2 P04179 PE=1 SV=2 NAD(P) transhydrogenase, mitochondrial OS=Homo sapiens GN=NNT PE=1 Q13423 SV=3 O43719 HIV Tat-specific factor 1 OS=Homo sapiens GN=HTATSF1 PE=1 SV=1

250 Accession No. PER2 protein interaction in hypoxic metabolism COMM domain-containing protein 5 OS=Homo sapiens GN=COMMD5 PE=1 Q9GZQ3 SV=1 P58546 Myotrophin OS=Homo sapiens GN=MTPN PE=1 SV=2 Q96ND0 Protein FAM210A OS=Homo sapiens GN=FAM210A PE=1 SV=2 EGF-like repeat and discoidin I-like domain-containing protein 3 OS=Homo O43854 sapiens GN=EDIL3 PE=1 SV=1 Q9H173 Nucleotide exchange factor SIL1 OS=Homo sapiens GN=SIL1 PE=1 SV=1 Phosphorylated adapter RNA export protein OS=Homo sapiens GN=PHAX Q9H814 PE=1 SV=1 A0FGR8 Extended synaptotagmin-2 OS=Homo sapiens GN=ESYT2 PE=1 SV=1 O95071 E3 ubiquitin-protein ligase UBR5 OS=Homo sapiens GN=UBR5 PE=1 SV=2 Putative peptidyl-tRNA hydrolase PTRHD1 OS=Homo sapiens GN=PTRHD1 Q6GMV3 PE=1 SV=1 Peptidyl-prolyl cis-trans isomerase FKBP3 OS=Homo sapiens GN=FKBP3 Q00688 PE=1 SV=1 KDEL motif-containing protein 2 OS=Homo sapiens GN=KDELC2 PE=1 Q7Z4H8 SV=2 P51570 Galactokinase OS=Homo sapiens GN=GALK1 PE=1 SV=1 Vacuolar protein sorting-associated protein 26A OS=Homo sapiens O75436 GN=VPS26A PE=1 SV=2 Glycogen phosphorylase, muscle form OS=Homo sapiens GN=PYGM PE=1 P11217 SV=6 O75367 Core histone macro-H2A.1 OS=Homo sapiens GN=H2AFY PE=1 SV=4 Translationally-controlled tumor protein OS=Homo sapiens GN=TPT1 PE=1 P13693 SV=1 Q08554 Desmocollin-1 OS=Homo sapiens GN=DSC1 PE=1 SV=2 Peptidyl-prolyl cis-trans isomerase FKBP9 OS=Homo sapiens GN=FKBP9 O95302 PE=1 SV=2 Cytochrome c oxidase assembly factor 5 OS=Homo sapiens GN=COA5 Q86WW8 PE=1 SV=1 P08473 Neprilysin OS=Homo sapiens GN=MME PE=1 SV=2 Q15181 Inorganic pyrophosphatase OS=Homo sapiens GN=PPA1 PE=1 SV=2 Q9BSE5 Agmatinase, mitochondrial OS=Homo sapiens GN=AGMAT PE=1 SV=2 O94952 F-box only protein 21 OS=Homo sapiens GN=FBXO21 PE=2 SV=2 Non-histone chromosomal protein HMG-14 OS=Homo sapiens GN=HMGN1 P05114 PE=1 SV=3 P00491 Purine nucleoside phosphorylase OS=Homo sapiens GN=PNP PE=1 SV=2 Geranylgeranyl transferase type-2 subunit beta OS=Homo sapiens P53611 GN=RABGGTB PE=1 SV=2 P49006 MARCKS-related protein OS=Homo sapiens GN=MARCKSL1 PE=1 SV=2 Glutamine--fructose-6-phosphate aminotransferase [isomerizing] 1 Q06210 OS=Homo sapiens GN=GFPT1 PE=1 SV=3 P46379 Large proline-rich protein BAG6 OS=Homo sapiens GN=BAG6 PE=1 SV=2 2-oxoglutarate dehydrogenase, mitochondrial OS=Homo sapiens GN=OGDH Q02218 PE=1 SV=3

251 Accession No. PER2 protein interaction in hypoxic metabolism P30626 Sorcin OS=Homo sapiens GN=SRI PE=1 SV=1 Q99436 Proteasome subunit beta type-7 OS=Homo sapiens GN=PSMB7 PE=1 SV=1 Q8WUF5 RelA-associated inhibitor OS=Homo sapiens GN=PPP1R13L PE=1 SV=4 Q6KB66 Keratin, type II cytoskeletal 80 OS=Homo sapiens GN=KRT80 PE=1 SV=2 Mitotic spindle-associated MMXD complex subunit MIP18 OS=Homo sapiens Q9Y3D0 GN=FAM96B PE=1 SV=1 Plasma membrane calcium-transporting ATPase 1 OS=Homo sapiens P20020 GN=ATP2B1 PE=1 SV=3 Q05639 Elongation factor 1-alpha 2 OS=Homo sapiens GN=EEF1A2 PE=1 SV=1 Receptor-type tyrosine-protein phosphatase F OS=Homo sapiens P10586 GN=PTPRF PE=1 SV=2 O75616 GTPase Era, mitochondrial OS=Homo sapiens GN=ERAL1 PE=1 SV=2 P30740 Leukocyte elastase inhibitor OS=Homo sapiens GN=SERPINB1 PE=1 SV=1 Growth arrest and DNA damage-inducible proteins-interacting protein 1 Q8TAE8 OS=Homo sapiens GN=GADD45GIP1 PE=1 SV=1 Glyoxylate reductase/hydroxypyruvate reductase OS=Homo sapiens Q9UBQ7 GN=GRHPR PE=1 SV=1 O43847 Nardilysin OS=Homo sapiens GN=NRD1 PE=1 SV=2 P08236 Beta-glucuronidase OS=Homo sapiens GN=GUSB PE=1 SV=2 tRNA (guanine-N(7)-)-methyltransferase non-catalytic subunit WDR4 P57081 OS=Homo sapiens GN=WDR4 PE=1 SV=2 P01781 Ig heavy chain V-III region GAL OS=Homo sapiens PE=1 SV=1 P62328 Thymosin beta-4 OS=Homo sapiens GN=TMSB4X PE=1 SV=2 Mitochondrial import inner membrane translocase subunit Tim9 OS=Homo Q9Y5J7 sapiens GN=TIMM9 PE=1 SV=1 Transcription factor BTF3 homolog 4 OS=Homo sapiens GN=BTF3L4 PE=1 Q96K17 SV=1 Glutamate--cysteine ligase regulatory subunit OS=Homo sapiens GN=GCLM P48507 PE=1 SV=1 Q969Z0 Protein TBRG4 OS=Homo sapiens GN=TBRG4 PE=1 SV=1 Q96SW2 Protein cereblon OS=Homo sapiens GN=CRBN PE=1 SV=1 Multidrug resistance-associated protein 4 OS=Homo sapiens GN=ABCC4 O15439 PE=1 SV=3 O00186 Syntaxin-binding protein 3 OS=Homo sapiens GN=STXBP3 PE=1 SV=2 Uncharacterized protein C2orf47, mitochondrial OS=Homo sapiens Q8WWC4 GN=C2orf47 PE=1 SV=1 Putative adenosylhomocysteinase 2 OS=Homo sapiens GN=AHCYL1 PE=1 O43865 SV=2 Q4G0J3 La-related protein 7 OS=Homo sapiens GN=LARP7 PE=1 SV=1 Pyridoxal phosphate phosphatase OS=Homo sapiens GN=PDXP PE=1 Q96GD0 SV=2 Zinc-binding alcohol dehydrogenase domain-containing protein 2 OS=Homo Q8N4Q0 sapiens GN=ZADH2 PE=1 SV=1 Interferon-inducible double-stranded RNA-dependent protein kinase activator O75569 A OS=Homo sapiens GN=PRKRA PE=1 SV=1

252 Accession No. PER2 protein interaction in hypoxic metabolism BAG family molecular chaperone regulator 3 OS=Homo sapiens GN=BAG3 O95817 PE=1 SV=3 P47929 Galectin-7 OS=Homo sapiens GN=LGALS7 PE=1 SV=2 28S ribosomal protein S15, mitochondrial OS=Homo sapiens GN=MRPS15 P82914 PE=1 SV=1 Q14764 Major vault protein OS=Homo sapiens GN=MVP PE=1 SV=4 P80748 Ig lambda chain V-III region LOI OS=Homo sapiens PE=1 SV=1 Translation initiation factor eIF-2B subunit epsilon OS=Homo sapiens Q13144 GN=EIF2B5 PE=1 SV=3 Endothelial differentiation-related factor 1 OS=Homo sapiens GN=EDF1 O60869 PE=1 SV=1 P30043 Flavin reductase (NADPH) OS=Homo sapiens GN=BLVRB PE=1 SV=3 Periodic tryptophan protein 1 homolog OS=Homo sapiens GN=PWP1 PE=1 Q13610 SV=1 Q9HCN4 GPN-loop GTPase 1 OS=Homo sapiens GN=GPN1 PE=1 SV=1 Q13561 Dynactin subunit 2 OS=Homo sapiens GN=DCTN2 PE=1 SV=4 Pseudouridylate synthase 7 homolog OS=Homo sapiens GN=PUS7 PE=1 Q96PZ0 SV=2 P04207 Ig kappa chain V-III region CLL OS=Homo sapiens PE=1 SV=2 Q9H9T3 Elongator complex protein 3 OS=Homo sapiens GN=ELP3 PE=1 SV=2

253 APPENDIX B

CLINICAL TRIAL #13-1607 CONSENT FORM

COMIRB C o n s e n t a n d A u t h o r i z a t i o n F o r m A p p r o v a l APPROVED For Use 14-Feb-2018 13-Feb-2019 Study Title: Impact of daylight on Period 2 protein and metabolism or inflammation in human blood samples

Principal Investigator: Tobias Eckle COMIRB No: 13-1607 Version Date: March 1 2017

You are being asked to be in a research study. This form provides you with information about the study. A member of the research team will describe this study to you and answer all your questions. Please read the information below and ask questions about anything you don’t understand before deciding whether or not to take part.

Why is this study being done?

This study plans to learn more about how daylight influences the circadian Protein Period 2 and associated metabolic alteration in blood cells. You are being

asked to be in this research study because you are

Group 1) a “normal” healthy subject or

Group 2) currently suffering from a heart attack.

Group 3) currently suffering from acute lung injury.

Other people in this study

Up to 100 people in Group 1, 66 people from Group 2 and 48 people from Group 3, a total of 214 people from your area will participate in this study.

What happens if I join this study?

1) Volunteer:

If you join the study, you will be asked to do the following things:

Blood Draw Visit: the following procedures will be performed:

• We will briefly review your medical history

• In this study we will need to get about 2 tablespoons of blood from you. We will get blood by putting a needle into one of your veins and letting the blood flow into a glass tube. You may feel some pain when the needle goes into your vein. A day or two later, you may have a small bruise where the needle went under the skin.

Page 1 of 8 Initials

254

C o n s e n t a n d A u t h o r i z a t i o n F o r m A p p r o v a l

• We will make sure you are feeling well after your blood is drawn and before you leave.

• If you donate your blood after daylight exposure we will let you sit in front of a daylight box for 30 min. You do not have to look directly into the light but should be in close proximity. You will be able to read, write, or work on your laptop. No side effects are known if daylight is used in the first half of the day since it just simulates being outside on a sunny day in Denver. The daylight does not emit any UV light and therefore there is no risk for skin cancer and you do not have to use any sun lotion.

This visit will take approximately 1-6 hour.

• This study will have 2 different groups of research subjects like you. One that will receive light exposure and one that does not. To decide which group you will be in, we will use a method of chance. This method is like flipping a coin or rolling dice. Each group will get slightly different care.

2) Patients with a heart attack:

If you have had a heart attack, you will undergo a simplified study protocol. A blood draw will be performed now and in one or two weeks. You may or may not receive a light box that you should use for the next 1 or 2 weeks every morning from 8.30-9.00 AM. We anticipate to enroll 15 patients in each group [1 week with light exposure, 1 week without light exposure, 2 weeks with light exposure, 2 weeks without light exposure].

This study will have 4 different groups of research subjects like you. Group one will receive light treatment for 1 week, group 2 will receive light treatment for 2 weeks, group 3 will be the control group for group 1 without light treatment for one week and group four will be control group for group 2 with no light treatment for 2 weeks after the MI until the follow up visit. To decide which group you will be in, we will use a method of chance. This method is like flipping a coin or rolling dice. Each group will get slightly different care (you can say exactly what happens).

Blood Draw Visit: the following procedures will be performed:

• We will briefly review your medical history

• In this study we will need to get about 2 tablespoons of blood from you. We will get blood by putting a needle into one of your veins and letting the blood flow into a glass tube. You may feel some pain when the needle goes into your vein. A day or two later, you may have a small bruise where the needle went under the skin. Blood will be drawn on the day of your heart attack and 1 or 2 weeks after the event during one of your follow up visits.

3) Patients with acute lung injury: Page 2 of 8 Initials

255

C o n s e n t a n d A u t h o r i z a t i o n F o r m A p p r o v a l

If you have acute lung injury you will be actively recruited through the Division of Pulmonology (Professor Marc Moss). As you will not be able to make a decision yourself due to mechanical ventilation and sedation, informed consent will be obtained by a legal representative on the same day when the diagnosis for lung injury has been made. You will undergo a simplified study protocol. A blood draw will be performed one day after consenting at 9AM, 12PM, 9PM and 12AM. We anticipate to enroll 48 patients in each group (24 male and 24 female patients [6 patients per time point ;4x6=24]).

• We will briefly review your medical history

• Pa. We will get blood by putting a needle into one of your veins and letting the blood flow into a glass tube. You may feel some pain when the needle goes into your vein. A day or two later, you may have a small bruise where the needle went under the skin. Blood will be drawn one day after the onset of acute lung injury.

What are the possible discomforts or risks?

Discomforts and Risks

1) Volunteers:

Approximately 2 tablespoons of blood will be removed by putting a needle into your arm vein. You might feel pain when the needle goes in to the vein. A bruise may form at the site. Rarely, people faint after blood drawing. Very rarely, the vein in which the needle has been inserted may become inflamed or infected.

If you are pregnant you will not be able to participate.

If you are sensitive to light and normally wear sun glasses, you might feel discomfort while sitting next to the daylight box. A small percentage of the population might develop a headache due to the bright light. If you know that you are such a person we would not recommend using the daylight box. Otherwise, no side effects are known if daylight is used in the first half of the day since it just simulates being outside on a sunny day in Denver. The daylight does not emit any UV light and therefore there is no risk for skin cancer and you do not have to use any sun lotion. Overuse may cause irritability, excessive energy, or difficulty falling asleep at bedtime.

2) Patients with heart attacks:

Approximately 2 tablespoons of blood will be removed by putting a needle into your arm vein. You might feel pain when the needle goes in to the vein. A bruise may form at the site. Rarely, people faint after blood drawing. Very rarely, the vein in which the needle has been inserted may become inflamed or infected.

Page 3 of 8 Initials

256

C o n s e n t a n d A u t h o r i z a t i o n F o r m A p p r o v a l If you are sensitive to light and normally wear sun glasses you might feel discomfort while sitting next to the daylight box. A small percentage of the population might develop a headache due to the bright light. If you know that you are such a person we would not recommend using the daylight box. Otherwise, no side effects are known if daylight is used in the first half of the day since it just simulates being outside on a sunny day in Denver. The daylight does not emit any UV light and therefore there is no risk for skin cancer and you do not have to use any sun lotion. Overuse may cause irritability, excessive energy, or difficulty falling asleep at bedtime.

The study may include risks that are unknown at this time.

3) Patients with acute lung injury:

Approximately 2 tablespoons of blood will be removed 4 times, for a total of half cup of blood, by putting a needle into your arm vein. You might feel pain when the needle goes in to the vein. A bruise may form at the site. Rarely, people faint after blood drawing. Very rarely, the vein in which the needle has been inserted may become inflamed or infected.

What are the possible benefits of the study?

This study is designed for the researcher to learn more about interactions between daylight and the tissue/blood levels of the circadian protein Period 2. Period 2 may provide some protection from heart attacks or acute lung injury.

This study is not designed to treat any illness or to improve your health. Also, there may be risks, as discussed in the section describing the discomforts or risks.

Who is paying for the study? The National Heart, Lung, and Blood (NHLBI) is the federal agency paying for this research.

Will I be paid for being in the study?

Patients with a heart attack will be compensated with 40 dollars for the additional visit necessary for the second blood draw.

Will I have to pay for anything?

It will not cost you anything to be in the study.

Is my participation voluntary?

Taking part in this study is voluntary. You have the right to choose not to take part in this study. If you choose to take part, you have the right to stop at any time. If you refuse or decide to withdraw later, you will not lose any benefits or rights to which Page 4 of 8 Initials

257

C o n s e n t a n d A u t h o r i z a t i o n F o r m A p p r o v a l you are entitled.

If you are associated with the PI (student or employee) you have the right to stop at any time. If you refuse or decide to withdraw later, you will not lose any benefits or rights to which you are entitled. The PI recognizes that the relationship he has with you, either as an employer or as a mentor to a student, is potentially coercive. It is important that you understand that choosing not to participate or discontinuing participation in this study will in no way effect your professional or academic standing with either the PI or the institution”.

Can I be removed from this study? The study doctor may decide to stop your participation without your permission if the study doctor thinks that being in the study may cause you harm, or for any other reason. Also, the sponsor may stop the study at any time.

What happens if I am injured or hurt during the study. If you have an injury while you are in this study, you should call Dr. Tobias Eckle immediately. His phone number is 720-848-3264.

We will arrange to get you medical care if you have an injury that is caused by this research. However, you or your insurance company will have to pay for that care.

Who do I call if I have questions?

The researcher carrying out this study is Tobias Eckle. You may ask any questions you have now. If you have questions, concerns, or complaints later, you may call Tobias Eckle at 720-949-5646. You will be given a copy of this form to keep. You may have questions about your rights as someone in this study. You can call Tobias Eckle with questions. You can also call the responsible Institutional Review Board (COMIRB). You can call them at 303-724-1055. A description of this clinical trial will be available on http://www.Clinical Trials.gov, as required by U.S. Law. This Web site will not include information that can identify you. At most, the Web site will include a summary of the results. You can search this Web site at any time.

Who will see my research information?

The University of Colorado Denver and the hospital(s) it works with have rules to protect information about you. Federal and state laws including the Health Insurance Portability and Accountability Act (HIPAA) also protect your privacy. This part of the consent form tells you what information about you may be collected in this study and who might see or use it. The institutions involved in this study include: ฀ University of Colorado Denver ฀ University of Colorado Hospital Page 5 of 8 Initials

258

C o n s e n t a n d A u t h o r i z a t i o n F o r m A p p r o v a l

We cannot do this study without your permission to see, use and give out your information. You do not have to give us this permission. If you do not, then you may not join this study.

We will see, use and disclose your information only as described in this form and in our Notice of Privacy Practices; however, people outside the University of Colorado Denver and its affiliate hospitals may not be covered by this promise.

We will do everything we can to maintain the confidentiality of your personal information but confidentiality cannot be guaranteed.

The use and disclosure of your information has no time limit. You can cancel your permission to use and disclose your information at any time by writing to the study’s Primary Investigator, at the name and address listed below. If you do cancel your permission to use and disclose your information, your part in this study will end and no further information about you will be collected. Your cancellation would not affect information already collected in this study.

Tobias Eckle, MD, PhD Professor of Anesthesiology, Cardiology and Cell Biology Department of Anesthesiology University of Colorado Denver 12700 E 19th Avenue, Mailstop B112, RC 2, Room 7121 Aurora, CO 80045; Office: 303-724 -2932 or - 2947, Fax: 303-724- 2852

Both the research records that identify you and the consent form signed by you may be looked at by others who have a legal right to see that information. ฀ Federal offices such as the Food and Drug Administration (FDA) that protect research subjects like you. ฀ National Heart, Lung, and Blood Institute (NHLBI) who is the federal agency paying for this research. ฀ People at the Colorado Multiple Institutional Review Board (COMIRB) ฀ The study doctor and the rest of the study team. ฀ Officials at the institution where the research is being conducted and officials at other institutions involved in this study who are in charge of making sure that we follow all of the rules for research

We might talk about this research study at meetings. We might also print the results of this research study in relevant journals. But we will always keep the names of the research subjects, like you, private.

You have the right to request access to your personal health information from

Page 6 of 8 Initials

259

C o n s e n t a n d A u t h o r i z a t i o n F o r m A p p r o v a l the Investigator. To ensure proper evaluation of test results, your access to these study results may not be allowed until after the study has been completed.

Information about you that will be seen, collected, used and disclosed in this study:

฀ Date of blood draw ฀ Daylight yes/no and duration ฀ Weather condition ฀ Time blood/tissue was collected ฀ Light exposure duration if any ฀ Caffeine consumption ฀ Alcohol consumption ฀ Smoking consumption ฀ Narcotics consumption ฀ Sleep behavior consumption ฀ NSAIDS consumption ฀ Underlying disorders such heart attacks, strokes, asthma, COPD, kidney disease, liver dysfunction, clotting problems (DVT), peripheral artery disease, coronary artery disease, high lipids, diabetes, pancreatitis, bipolar disorder etc.)

What happens to the Data and Blood samples that are collected in this study? Scientists at the University of Colorado Denver involved in this study work to find the causes and cures of disease. The blood collected from you during this study is important to this study and to future research. If you join this study:

฀ The blood that is given by you to the investigators for this research and so no longer belong to you. ฀ Both the investigators and any co-investigator of this research may study your blood collected from you. ฀ The blood collected from you will not be in a form that identifies you. ฀ Any product or idea created by the researchers working on this study will not belong to you. ฀ There is no plan for you to receive any financial benefit from the creation, use or sale of such a product or idea.

What happens if I do NOT return the Light box? I agree to return the light box at the indicated time frame (1 or 2 weeks after completion of the study). If I do not return the light box, I will be charged $ 500.

Page 7 of 8 Initials

260

C o n s e n t a n d A u t h o r i z a t i o n F o r m A p p r o v a l

Agreement to be in this study and use my data I have read this paper about the study or it was read to me. I understand the possible risks and benefits of this study. I understand and authorize the access, use and disclosure of my information as stated in this form. I know that being in this study is voluntary. I choose to be in this study: I will get a signed and dated copy of this consent form.

Signature: Date:

Print Name:

Consent form explained by: Date:

Print Name:

______Date______

Child (13-17 year olds)

Consent form explained by: ______Date______

Print Name ______

Investigator ______Date ______

Page 8 of 8 Initials

261