Mechanisms of G Protein-Coupled Receptor 2 Regulation and Inhibition in Cardiovascular Disease and Aging

MECHANISMS OF G PROTEIN-COUPLED RECEPTOR 2 REGULATION AND INHIBITION IN CARDIOVASCULAR DISEASE AND AGING

A Dissertation Submitted to the Temple University Graduate Board

In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY

by Melissa J. Lieu August 2020

Examining Committee Members:

Dr. Walter J. Koch, Ph.D., Advisory Chair, Center for Translational Medicine Dr. Steven R. Houser, Ph.D., Cardiovascular Research Center Dr. Raj Kishore, Ph.D., Center for Translational Medicine Dr. Konstantinos Drosatos, Ph.D., Center for Translational Medicine Dr. Priscila Sato, Ph.D., External Member, Pharmacology & Physiology, College of Medicine, Drexel University

ABSTRACT

G protein-coupled receptor kinase 2 (GRK2) has been a thriving therapeutic target for cardiovascular disease treatment since its discovery for desensitizing and downregulating b-adrenergic receptors that are vital to cardiac function. GRK2 inhibition through a variety of methods in animal models of cardiac ischemia and heart failure achieved improvements in cardiac function, hemodynamic function, cardiomyocyte apoptosis, and fibrotic scar size among many other observations. Although GRK2 has been used as a therapeutic tool in multiple studies, its mechanisms of regulation are necessary to understand its role in disease pathogenesis and therapeutic application. This dissertation comprises two projects (1) investigating the microRNA (miRNA) regulation of GRK2 and

(2) investigating the impact of loss of dynamic regulation of GRK2 through S-nitrosylation.

(1) Candidate miRNAs were selected from miRNA microarray analysis of miRNA differential expression data and bioinformatic prediction. In vitro validation of kshv-miR-

K12-3-5p and hsa-miR-181a-5p have shown their ability to bind to the GRK2 3’UTR as well as significantly decrease GRK2 mRNA or protein. The successful regulation of GRK2 through these miRNAs warrant in vivo application and investigation as GRK2-targeting

HF therapy in a mouse model of HF. (2) In order to determine the impact of chronic GRK2 overactivity, mice that contained a knock-in mutation of GRK2 Cys340àSer (GRK2-

C340S), a site of dynamic inhibitory regulation by S-nitrosylation, were allowed to age

>12 months. Loss of S-nitrosylation of GRK2 was sufficient to cause cardiovascular remodeling and dysfunction over time.

DEDICATION

This work is dedicated to family and nature.

iii ACKNOWLEDGMENTS

I became interested in research during my undergraduate years, because I wanted to observe and understand nature in its most accurate way, combining my interests in science and philosophy that might be considered one and the same. The years I’ve spent studying and working towards my PhD have certainly allowed me to do just that, but also accomplish so much more. This process was incredibly challenging mentally, but I am fortunate to have gone through it. The days were long, but the years went quickly. I would like to acknowledge my thesis advisor and mentor Dr. Walter “Wally” Koch. From the first time I walked into his office, I was pleasantly surprised to discover someone as renowned and prestigious as himself take genuine interest in me. That still holds true today. Wally has continually and consistently been a source of unwavering support, encouragement, and valuable feedback. He has believed in me even when I did not believe in myself. The significance of this is not something to be underestimated and it has positively impacted me during my pre-doctoral years. It is something I will never forget. Despite his signature line “Get back to work!”, Wally has shown he puts our well- being and safety above all else. This has made me feel incredibly appreciated and valued, which has made for a nurturing and thriving environment for a PhD student to grow. As I am transitioning to becoming a Koch Lab alumnae, I can only hope that I have done Wally justice in my research and can stand alongside the numerous successful trainees he’s mentored. Beyond his lab, Wally has assembled our department—Center for Translational Medicine (CTM)—with utmost care. I believe CTM is the best place to be at Temple and its environment truly fosters scientific collaboration, which was vital to my projects. The Koch lab members, past and present, also require immense thanks. They are who I interact with every day and make up the incredible Koch lab environment. Notable members include: Dr. Kurt Chuprun, for his lab management and random questions; Dr.

iv Claudio de Lucia, for his expertise in gerontology and for being an incredibly kind and selfless person; Zuping Qu and Kevin Luu, for all mouse husbandry; Dr. Sarah Bass, for providing such quality training and mentoring during my initial rotation and transition to an independent project. She gave me a strong foundation to go forth into the Koch lab and taught me so much; Dr. Jessica Pfleger, for being the best desk-mate. She is a powerful critical and rigorous scientific thinker, a role model indeed; Jennifer Petovic, for her unrivaled thoughtfulness and also for passing along the C340S project to me; Dr. Ryan Coleman and Dr. Akito Eguchi for being there right along with me as graduate students in Wally’s lab. It’s comforting to have had peers that understood exactly what I was going through. I would also like to thank the members of my thesis advisory committee, all of whom are/have been part of CTM or CVCR: Dr. Steven R Houser, Dr. Raj Kishore, Dr. Konstantinos Drosatos, and Dr. Priscila Sato. They all have been an invaluable providers of cardiovascular expertise and encouragement. I appreciate the time they have invested in my scientific and personal growth through the years. CTM and Temple have been my “home” for the past 6 years and it’s no doubt I have made lasting friendships during this time. I would like to acknowledge Ali Herman, Christine Vrakas, Ryan LaCanna, Alex Vouga, and Alex Bonanno for being my “Day Ones” since entering the program. There has never been a dull moment with them. The significance of their support should not be underestimated. If science doesn’t work out for us, I’m confident one of our exceptionally innovative ideas will take off. As one-by-one they have defended their theses and moved on to their next steps (as I am doing now), I have been adopted by a new group: Emma Murray, Alycia Hildebrandt, and Giana Schena. These women are empowering and uplifting; I am grateful to have found them. Lastly, I would like to thank my family: my husband, Nick; my parents, Ha and Joanna; my brother, Bryan; my in-laws, Jim, Diane, and Dina; my cat, Buddy. I thank them for all their love and support over the years. v TABLE OF CONTENTS

Page

ABSTRACT ...... ii

DEDICATION ...... iii

ACKNOWLEDGMENTS ...... iv

LIST OF TABLES ...... xi

LIST OF FIGURES ...... xii

LIST OF ABBREVIATIONS ...... xiv

CHAPTER 1

1. INTRODUCTION ...... 1

1.1 Cardiovascular disease ...... 2

1.2 b-adrenergic receptors ...... 2

1.3 G protein-coupled receptor kinases ...... 5

1.4 GRK2 in cardiac physiology ...... 9

1.5 GRK2 in cardiac pathophysiology ...... 11

1.6 GRK2 regulation ...... 15

1.6.1 GRK2 protein regulation ...... 15

1.6.2 GRK2 transcriptional and post-transcriptional regulation ...... 18

1.7 MicroRNAs ...... 21

1.7.1 MiRNA biogenesis...... 22

1.7.2 MiRNA mechanisms of action ...... 26

1.7.3 MiRNAs in cardiovascular disease ...... 31

vi 1.7.4 MiRNAs as therapeutic tools ...... 33

1.8 Aging...... 38

1.8.1 Hallmarks of general aging ...... 38

1.8.2 Characteristics of cardiovascular aging ...... 48

1.8.3 GRK2 in aging ...... 54

1.9 Summary and objectives ...... 56

1.9.1 Cardiac GRK2 and microRNAs in heart failure ...... 56

1.9.2 Loss of dynamic regulation of GRK2 by nitric oxide and

cardiovascular aging ...... 57

CHAPTER 2

2. CARDIAC GRK2 AND MICRORNAS IN HEART FAILURE ...... 58

2.1 Introduction ...... 58

2.2 Experimental procedures ...... 61

2.2.1 Experimental animals ...... 61

2.2.2 Echocardiography ...... 61

2.2.3 Myocardial infarction ...... 62

2.2.4 Myocardial injections ...... 62

2.2.5 In vitro miRNA mimic transfections ...... 63

2.2.6 Immunoblotting...... 64

2.2.7 RNA isolation and semi-quantitative PCR ...... 65

2.2.8 Creation of stable Ad293-GRK2 cell line ...... 65

2.2.9 3’UTR analysis ...... 66

vii 2.2.10 Statistics ...... 67

2.3 Results ...... 67

2.3.1 Myocardial infarction changes miRNA expression profile ...... 67

2.3.2 Analysis of miRNA candidates with bioinformatic prediction ...... 69

2.3.3 Selected miRNA candidates reduced GRK2 expression in

vitro ...... 70

2.3.4 Antagonizing miRNA candidates increased GRK2

expression in vitro ...... 77

2.3.5 Selected miRNA candidates bind to GRK2 3’UTR ...... 79

2.3.6 GRK2 expression in miR181a/bKO mice ...... 82

2.3.7 miR-K12 reduced GRK2 expression in vivo ...... 83

2.3.8 GRK2 overexpression changes miRNA expression profile

in vivo ...... 85

2.4 Discussion ...... 85

CHAPTER 3

3. LOSS OF DYNAMIC REGULATION OF GRK2 BY NITRIC OXIDE AND

CARDIOVASCULAR AGING ...... 98

3.1 Introduction ...... 98

3.2 Experimental procedures ...... 109

3.2.1 Experimental animals ...... 109

3.2.2 Echocardiography ...... 110

3.2.3 RNA isolation and semi-quantitative PCR ...... 111

viii 3.2.4 Serum catecholamine quantification ...... 112

3.2.5 Radioligand binding ...... 112

3.2.6 Terminal hemodynamics ...... 112

3.2.7 Non-invasive tail cuff measurements ...... 113

3.2.8 Telemetry measurements ...... 113

3.2.9 Histology ...... 114

3.2.10 Immunohistochemistry ...... 115

3.2.11 Histological quantifications ...... 115

3.2.12 Isometric aortic contraction ...... 116

3.2.13 Statistics ...... 117

3.3 Results ...... 117

3.3.1 Aged GRK2-C340S mice have reduced cardiac function ...... 117

3.3.2 Aged GRK2-C304S mice have adrenergic abnormalities ...... 121

3.3.3 Aged GRK2-C340S mice exhibit cardiac hypertrophy ...... 125

3.3.4 Aged TgGRK2 mice do not exhibit age-dependent

cardiac dysfunction ...... 127

3.3.5 Aged TgGRK2 mice do not experience maladaptive

remodeling compared to non-transgenic littermate

controls ...... 127

3.3.6 Aged GRK2-C340S mice exhibit adverse cardiac

and renal remodeling ...... 131

3.3.7 Aged GRK2-C340S mice exhibit increased systolic

blood pressure ...... 134

ix 3.3.8 Aortic size is reduced in GRK2-C340S mice ...... 136

3.3.9 Aged GRK2-C340S mice have impaired aortic

contraction ...... 138

3.4 Discussion ...... 139

CHAPTER 4

4. CONCLUDING REMARKS AND SIGNIFICANCE ...... 152

4.1 Extended discussion ...... 152

REFERENCES CITED ...... 157

x LIST OF TABLES

Table Page

Table 1-1. Tissue distribution of GRK isoforms ...... 8

Table 1-2. MicroRNAs involved in heart failure ...... 33

Table 1-3. MicroRNAs used in interventional clinical trials ...... 36

Table 1-4. Major theories of aging ...... 40

Table 2-1. Characteristics of miRNA candidates ...... 72

Table 3-1. S-nitrosylated substrates involved with b-AR function ...... 105

xi LIST OF FIGURES

Figure Page

Figure 1-1. G protein-coupled receptor activation, deactivation, and regulation by G protein-coupled receptor kinases ...... 4

Figure 1-2. Domain structure and subfamilies of GRK isoforms ...... 7

Figure 1-3. Post-translational modifications of GRK2 ...... 17

Figure 1-4. Canonical miRNA biogenesis pathway ...... 23

Figure 1-5. Types of commonly used miRNA modifications ...... 35

Figure 1-6. Physiological aging-associated cardiovascular changes ...... 49

Figure 2-1. Changes in miRNA expression after myocardial infarction ...... 68

Figure 2-2. Identifying miRNA candidates for targeting GRK2

3’UTR ...... 71

Figure 2-3. Mouse and human alignment of highly similar sequences within GRK2’UTR...... 73

Figure 2-4. GRK2 inhibition dose response with miRNA mimics ...... 75

Figure 2-5. Transfection of various miRNA mimics decreased GRK2 protein expression ...... 76

Figure 2-6. Transfection of various miRNA mimics decreased GRK2 mRNA abundance ...... 78

Figure 2-7. Candidate miRNA inhibitors induce small increases in

GRK2 expression ...... 80

Figure 2-8. Validated 3’UTR analysis of GRK2 ...... 81

xii Figure 2-9. GRK2 expression in homozygous miRNA-181a/b knock-out mice ...... 83

Figure 2-10. Intramyocardial delivery of miR-K12 reduces GRK2 protein expression ...... 84

Figure 2-11. Differential miRNA expression in TgGRK2 mice ...... 86

Figure 3-1. S-nitrosylation and transnitrosylation NO-based reactions ...... 103

Figure 3-2. Aged GRK2-C340S mice express mutant GRK2 at endogenous protein levels ...... 118

Figure 3-3. Aged GRK2-C340S mice have reduced ejection fraction ...... 120

Figure 3-4. Aged GRK2-C340S mice have reduced cardiac function

- additional echocardiographic parameters ...... 122

Figure 3-5. Aged GRK2-C340S mice have impaired longitudinal and radial myocardial strain ...... 123

Figure 3-6. Aged GRK2-C340S mice have a degree of diastolic dysfunction ...... 124

Figure 3-7. Aged mice have increased neurohormonal activation measured by catecholamines ...... 126

Figure 3-8. Aged GRK2-C340S mice exhibit cardiac hypertrophy ...... 128

Figure 3-9. Aged cardiac TgGRK2 mice do not experience declines in cardiac function compared to aged NLC ...... 129

Figure 3-10. Aged cardiac TgGRK2 mice do not experience maladaptive remodeling compared to aged NLC ...... 130

xiii Figure 3-11. Aged GRK2-C340S mice exhibit cardiac vascular abnormalities ...... 133

Figure 3-12. Aged GRK2-C340S mice develop renal vascular hypertrophy and fibrosis ...... 134

Figure 3-13. Aged GRK2-C340S mice exhibit increased blood pressures ...... 136

Figure 3-14. Aged GRK2-C340S mice trend towards narrowed pulse pressure windows ...... 137

Figure 3-15. Aortic size is reduced in GRK2-C340S mice ...... 138

Figure 3-16. Aged GRK2-C340S mice exhibit aortic contraction dysfunction ...... 140

Figure 3-17. Summary of aged GRK2-C340S phenotype ...... 142

xiv

LIST OF ABBREVIATIONS

aMHC alpha myosin heavy chain aSMA alpha smooth muscle actin b-AR b-adrenergic receptor b-arr b-arrestin bARK b-adrenergic receptor kinase

βARKct b-adrenergic receptor kinase C-terminus

AC adenylyl cyclase

ACh acetylcholine

ActD actinomycin D

AGO argonaute

AHA American Heart Association

Akt protein kinase B

AMPK 5' AMP-activated protein kinase

ANF atrial natruietic factor

AngII angiotensin II

ARE AU-rich element

ATP adenosine triphosphate

BH4 tetrahydriobiopterin

BNP brain natruietic peptide

xv BW body weight cAMP cyclic adenosine 3’,5’- monophosphate

CD31 cluster of differentiation 31

CDS coding sequence cGMP cyclic guanosine monophosphate

CnA calcineurin

CV cardiovascular

CVD cardiovascular disease

Cys cysteine

CysNO S-nitrosocysteine

DGCR8 DiGeorge critical region

E epinephrine

EF ejection fraction

EGFR epidermal growth factor receptor eIF elongation initiation factor eNOS endothelial nitric oxide synthase

ET-1 endothelin-1

FAD flavin adenine dinucleotide

FMN flavin mononucleotide

FS fractional shortening

GPCR G protein-coupled receptor

GRK G protein-coupled receptor kinase

GSNO S-nitrosoglutathione

xvi GSNOR S-nitrosoglutathione reductase

HDAC histone deacetylase

HEK human embryonic kidney

HF heart failure

HFD high-fat diet

HITS-CLIP high-throughput sequencing of RNA isolated by crosslinking immunoprecipitation

Hsp heat shock protein

HUVEC human umbilical vein cells

IP immunoprecipitation

I/R ischemia/reperfusion

IRS1 insulin receptor substrate 1

ISO isoproterenol

IVRT isovolumic relaxation time

KI knock-in

KS Kaposi’s sarcoma

KSHV Kaposi’s sarcoma herpesvirus

L-NAME L-NG-nitroarginine methyl ester

LV left ventricle

LVID left ventricular internal diameters

MAPK mitogen activated protein kinase

MHC myosin heavy chain

MI myocardial infarction

xvii MPI myocardial performance index miRNA microRNA mRNA messenger RNA

NE norepinephrine

NES nuclear export signal

NLC non-transgenic littermate controls

NLS nuclear localization sequence

NO nitric oxide

NOS nitric oxide synthase

PABP poly (A) binding protein

PAR-CLIP photoactivatable-ribonucleoside-enhanced crosslinking and immunoprecipitation

PECAM-1 platelet endothelial cell adhesion molecule-1

PI3K phosphoinositide 3 kinase

PIP3 phosphatidylinositol-3,4,5-triphosphate

PKA protein kinase A

PKC protein kinase C

PH pleckstrin homology

PMA phorbol ester

Pol RNA polymerase

Pre-miRNA precursor miRNA

PTEN phosphatase and tensin homolog

RAAS renin-angiotensin-aldosterone system

RISC RNA-induced silencing complex

xviii

RGS regulator of G protein signaling

RKIP Raf kinase inhibitor protein

RLC RISC loading complex

ROS reactive oxygen species

Ser serine

SERCA2a L-type Ca2+ channel/b-AR and SR Ca2+ ATPase

SNS sympathetic nervous system

SNO-GRK S-nitrosylated GRK2

STE speckle tracking-based strain imaging

SV stroke volume

SNO S-nitrosothiol

TAC trans-aortic constriction

TAAC transverse abdominal aortic constriction

TE traditional echocardiography

Thr threonine

TL tibia length

TN translin

Tyr tyrosine

UTR untranslated region

VSMC vascular smooth muscle

WGA wheat germ agglutinin

XPO5 exportin-5

xix

CHAPTER 1

INTRODUCTION

1.1 Cardiovascular Disease

Cardiovascular (CV) disease (CVD) continues to be the leading cause of death in the United States (US) in both males and females (Heron 2019). The cost burdens to treat

CVD will continue to grow as the population becomes increasingly more aged (North and

Sinclair 2012). Additionally, prevalence of CVD has been increasing in young populations when comparing 2015 to 2019 statistics from the American Heart Association (AHA)

(Benjamin et al. 2019; Mozaffarian et al. 2015). Due to such high prevalence and mortality rate, achieving a comprehensive understanding of the molecular mechanisms of CVD is imperative to designing novel and effective therapies to improve the quality of life and survival of patients.

CVD consists of any disease involving the heart and/or blood vessels such as coronary artery disease, myocardial infarction (MI), and hypertension. Long term stress and damage to the heart by these diseases can cause heart failure (HF), which is further increased by risk factors such as smoking, obesity, and high fat diet (Benjamin et al. 2019).

The American Heart Association (AHA) defines HF as a “complex clinical syndrome that can result from any structural or functional cardiac disorder that impairs the ability of the ventricle to fill with or eject blood.” (Hunt et al. 2009). HF afflicts about 6.5 million adults in the US alone, contributing to 1 in 8 deaths (Benjamin et al. 2019). Not surprisingly, the survival rate of HF following diagnosis is poor with 81% surviving the first year to 30%

1 surviving the tenth year (Taylor et al. 2017). Due to the inability of the heart to pump, tissues may experience ischemia. HF is characterized by morphological changes of the heart including hypertrophy, wall thinning, dilation of chambers, and fibrosis—all of which negatively impact pumping function in the long run. These morphological changes are a result of the vast signaling cascades, some originally compensatory to minimize wall stress, initiated by pathological stimuli. Biological systems that get activated during HF and continue to exacerbate the disease include the renin-angiotensin-aldosterone system, sympathetic nervous system, and immune system. Modern HF therapies aim to (1) rapidly reperfuse the heart (in cases of ischemia) to minimize cardiomyocyte death and (2) pharmacologically manage the chronically weakened heart such as preserving cardiac function with b-blockers, which antagonize b-adrenergic receptors (b-ARs).

1.2 b-adrenergic receptors

b-adrenergic receptors (b-ARs) belong to the Class A (rhodopsin-like) class of G protein-coupled receptors (GPCRs). Three subtypes have been identified and cloned: b1-

AR, b2-AR, and b3-AR. The heart predominantly expresses b1-AR, which make up 75-

80% of cardiac b-AR heart expression (Rockman et al. 2002). b2-ARs and b3-ARs compromise the remaining percentage, 15-18% and 2-3% respectively (Lymperopoulos et al. 2013). Together these b-ARs promote cardiac chronotropy, inotropy, domotropy, and lusitropy in response to norepinephrine (NE) and epinephrine (E) by transducing sympathetic responses throughout the whole cell to impact cellular function. Typically upon ligand binding, the b-AR undergoes a conformational change to promote binding to

2 the heterotrimeric G proteins: Gas and Gbg (Fig. 1-1) (Sato et al. 2015b). GDP-bound Gas is exchanged for GTP-bound Gas to prompt dissociation of the G-proteins that activate downstream signaling pathways. Gas activates adenylyl cyclase (AC) which converts adenosine triphosphate (ATP) to the second messenger cyclic adenosine 3’,5’- monophosphate (cAMP). cAMP activates cAMP-dependent protein kinases (PKA). PKA represents the major effector of the b-AR pathway by phosphorylating downstream targets like the L-type calcium channel that induces an initial calcium influx followed by a larger calcium-induced-calcium-release response to initiate cardiac contractility. Contractility primarily results from stimulation of b1-ARs and to a lesser extent b2-AR. b1-ARs couple primarily with Gas while b2-ARs and b3-ARs can also couple to Gai in addition to Gas depending on the cellular context to elicit inhibitory responses. (Lymperopoulos et al.

2013; Skeberdis et al. 2008; Xiao et al. 1999). Due to the ability to couple to Gai, b2-ARs and b3-ARs have been reported to be anti-inotropic, anti-apoptotic, and even cardioprotective.

The dissociated Gbg subunit participates in the recruitment of a G protein-coupled receptor kinase (GRK) to the plasma membrane. GRKs phosphorylate serine (Ser) and threonine (Thr) residues within the intracellular C-terminal tail of ligand-activated b-ARs to causes another conformational change that allows b-arrestin (b-arr) recruitment to the receptor. b-arr bound to the receptor physically prevents additional G protein activation as well as promotes clathrin-mediated receptor endocytosis. The receptor can be recycled

FIGURE 1-1. G protein-coupled receptor activation, deactivation, and regulation by G protein-coupled receptor kinases (Sato et al. 2015b). Upon agonist binding to a GPCR, like the b-AR, the receptor undergoes a conformational change. Most notably, there is an outward shift of transmembrane domain 6 (TM6) and a small outward rotation of TM5. These conformational changes allow the association of the heterotrimeric G proteins, Ga and Gbg, to the receptor to occur. GDP is exchanged for GTP on the Ga subunit, which allows the dissociation of Ga and Gbg to modulate activity of downstream signaling. Notably, Ga modulates adenylyl cyclase (AC) while Gbg binds GRKs via GRKs’ PH domain with high affinity to recruit it to the receptor at the membrane. GRKs then phosphorylate the GPCR within their third intracellular loop or C-terminal tails. Phosphorylated GPCRs are primed for b-arrestin binding, which prevents any subsequent GPCR activation. b-arrestin binding prompts receptor mediated endocytosis where the GPCR can be degraded or recycled back to the membrane.

4 back to the plasma membrane or undergo proteasomal degradation. The intrinsic GTPase activity of the Gas subunit will revert itself back to being GDP-bound, then reassociated with Gbg at the plasma membrane. The b-AR is then ready to undergo another activation cycle.

In summary, catecholamine binding and activation of the b-AR elicit sympathetic physiological responses that are promptly deactivated by GRKs and b-arr. This delicate balance between activation and deactivation is disrupted during disease ie HF. During HF, there is a selective decrease in b1-AR density in response to overactivation. The receptors also become uncoupled to downstream effectors like AC. This is accompanied by an increase in inhibitory Gai subunits to further antagonize b-AR responsiveness. These alterations have detrimental consequences on cardiac function. Another mechanism of b-

AR-mediated dysregulation during HF, and other CVD, occurs through the pathological increase of GRK activity and expression.

1.3 G protein-coupled receptor kinases

There are only seven members in the GRK family that are responsible for regulating more than 800 GPCRs, including the b-ARs. GRKs represent the rate liming step for GPCR desensitization and internalization, important regulatory processes in GPCR function.

Their structure and function are highly similar to AGC protein kinases like PKA. These seven members are further divided into three subcategories based on sequence homology, gene structure, and tissue distribution: (1) visual or rhodopsin kinase subfamily – GRK1 and GRK7, (2) b-AR kinase (bARK) subfamily – GRK2, also known as bARK1, and

GRK3, (3) and GRK4 subfamily – GRK4, GRK5, and GRK6 (Fig. 1-2) (Sato et al. 2015b).

Structurally, all GRKs can be divided into three regions consisting of a short NH2-terminus, a regulator of G protein signaling (RGS) homology domain, and a Ser/Thr kinase domain.

The NH2 is a highly conserved domain between the GRKs that is implicated in GPCR binding at the COOH-terminus (Inglese et al. 1993). The COOH terminus of GRKs is highly diverse between members. GRK1 and GRK7, both restricted to the retina, possess short COOH-termini prenylation sequences that allow it to be anchored to the membrane and membrane-embedded GPCR (Inglese et al. 1992). Alternatively, GRK2 and GRK3 uniquely contain a pleckstrin homology (PH) domain that facilitates charged phospholipid binding within the plasma membrane and Gbg binding, which allow these GRKs to be membrane-bound (Cannavo et al. 2013; DebBurman et al. 1996; Inglese et al. 1992). GRK2 and Gbg bind at high affinities (~30 nM) and appears to be the rate limiting step for GRK2- dependent GPCR phosphorylation (Carman et al. 2000). GRK2 also localizes to the mitochondria through an undetermined mechanism and has been shown to promote cell death during myocardial ischemia and impair fatty acid metabolism (Chen et al. 2013;

Pfleger et al. 2018; Sato et al. 2015a). GRK4, GRK5, and GRK6 all appear to be predominantly membrane-bound at basal conditions. GRK4 and GRK6 contain palmitoylation sites that are essential for their membrane association, but GRK5 is not regulated in this way (Premont et al. 1996; Stoffel et al. 1994). GRK5’s COOH terminus includes a phospholipid binding domain, which dictates its intracellular localization

(Pronin et al. 1998). All GRK4 family members possess a nuclear localization sequence

(NLS) and a nuclear export signal (NES) for nuclear localization, but seem to be

Visual Subfamily 1 42 184 514 532 563

GRK1 K219 Cys-S

1 40 184 513 531 553

GRK7 K220 Cys-S

βARK Subfamily 1 36 184 513 553 661 689

GRK2 K220

495 βARKct 689 1 36 184 513 553 661 688

GRK3 K220

GRK4 Subfamily 1 38 181 255-271 508 525 578

GRK4 K216 K217 219-227 palmitoylation 1 38 180 259-265 507 525 590

GRK5 K215

388-395 552-562 1 38 180 254-270 507 525 576

GRK6 K215 K216 388-395 palmitoylation

N-terminal domain C-terminal domain RGS homology domain Nuclear export signal Kinase domain Nuclear localization signal Pleckstrin homology domain Phospholipid binding domain

FIGURE 1-2. Domain structure and subfamilies of GRK isoforms. All GRKs consist of a N-terminal domain (green), RGS homology domain (yellow) containing the kinase domain (light blue), and a C-terminal domain (CTD) (light pink). Specific lysine (K) residues important for catalytic activity are in red. Each GRK subfamily has unique characteristics. The visual family members have prenylation sequences at the CTD. The bARK subfamily contains a pleckstrin homology domain (purple). The GRK4 subfamily contains nuclear export (blue) and nuclear localization (dark blue) signals. GRK5 specifically has a phospholipid binding domain (dark pink).

! differentially and dynamically regulated (Jiang et al. 2007; Johnson et al. 2013; Johnson et al. 2004). For example, calmodulin binding promotes GRK5 nuclear export but does not affect GRK6 (Johnson et al. 2013). GRK5 phosphorylates histone deacetylase (HDAC) 5

(HDAC5) to promote HDAC5 nuclear export, but nuclear functions have not been identified for GRK4 and GRK6 (Johnson et al. 2013; Martini et al. 2008). Besides the rhodopsin family members, all other GRKs are generally widely or ubiquitously expressed in different tissues (Table 1-1).

Table 1-1. Tissue distribution of GRK isoforms GRK Tissue Localization 1 Retina >> pineal 2 Leukocytes > front cortex of brain >> heart > lung > kidney, thyroid nodules 3 Brain > spleen > heart, lung, kidney, thyroid nodules 4 Testis >>>> brain, thyroid nodules 5 Heart, placenta, lung > skeletal muscle > brain, liver, pancreas > kidney, thyroid nodules 6 Brain, skeletal muscle > pancreas > heart, lung, kidney, placenta > liver, thyroid nodules 7 Retina (cones) Adapted from (Sato et al. 2015b)

1.4 GRK2 in cardiac physiology

GRK2 transcripts are detected very early on during embryogenesis and remain the predominant isoform expressed (Jaber et al. 1996; Sefton et al. 2000). During early heart development GRK2 can be detected in the “muscular part of the interventricular septum, the atrium, the aortic valve and the pericardium” as well as the interventricular septum, aortic valve, and atria at later points in development (Sefton et al. 2000). With such widespread and robust GRK2 expression during development, it comes to no surprise that global homozygous GRK2 knockout (GRK2−/−) embryos do not survive past gestational day 15.5 due to significant ventricular hypoplasia and HF (Jaber et al. 1996; Matkovich et al. 2006). It is the only global GRK knockout that results in embryonic lethality. However, global hemizygous GRK2 deletion (GRK2+/−) in mice is viable and they develop normally as measured by left ventricle (LV) weight to body weight (BW) or tibia length (TL)

(Rockman et al. 1998b). GRK2+/− have a 62% decrease in expression by messenger RNA

(mRNA) and 50% by protein that corresponds to 50% decrease in GRK2 activity (Jaber et al. 1996; Rockman et al. 1998b). Cardiac-specific GRK2 deletion (~76% decrease in protein) driven by Nkx2.5-Cre (GRK2f/f;Nkx2.5Cre) are also viable and exhibit normal heart development and function, suggesting that GRK2 in the heart specifically is not essential for its development (Matkovich et al. 2006).

Adult cardiac GRK isoforms are GRK2 and GRK5, but GRK3 and GRK6 have been identified at lower levels as measured by protein or mRNA (Agüero et al. 2008;

Dzimiri et al. 2004). GRK2 and GRK5 both can physiologically phosphorylate ligand- activated b1-ARs and b2-ARs (Freedman et al. 1995; Koch et al. 1995; Pfleger et al. 2019).

Due to its impact on cardiac function physiologically and pathophysiologically, GRK2 remains the predominant cardiac GRK. GRK2’s role physiologically is to deactivate b-AR activation to prevent overstimulation by promoting desensitization and termination of signaling (Lefkowitz 1993). GRK2’s critical impact on cardiac function via desensitization is demonstrated through genetic models of GRK2 manipulation. Adult (4-6 months)

βARK1+/− mice display enhanced myocardial contractility and heart rate with isoproterenol (ISO) infusion, a nonselective b-AR agonist, compared to wild type mice that was even further enhanced with βARKct (Rockman et al. 1998b). βARKct is a peptide that consists of the of the COOH terminus of GRK2 (Fig. 1-3) that can bind free Gbg, therefore acting as a competitive inhibitor (Koch et al. 1995). Cardiac-specific (aMHC) βARKct overexpression in mice (TgβARKct) alone is also sufficient in improving β-AR sensitivity and contractility, but with no changes in AC activity reported (Koch et al. 1995). Thus, further proving that βARKct modulates at the level of β-AR-GRK2 interaction, by preventing phosphorylation-dependent desensitization of the β-AR. GRK2f/f;Nkx2.5Cre mice that lack GRK2 since development also elicit enhanced contractility and reduced tachyphylaxis in response to acute ISO infusions, but experienced “catecholamine cardiomyopathy” in response to chronic ISO infusions compared to control mice

(Matkovich et al. 2006).

In contrast, cardiac-specific (aMHC) overexpression of GRK2 in mice (TgGRK2) resulted in blunted inotropic and chronotropic responses after ISO infusion compared to control mice. TgGRK2 mice have ~3 times more protein expression and kinase activity compared to controls. Blunted myocardial contractility in TgGRK2 was due to reduced b-

AR coupling as measured by AC activity (Koch et al. 1995). In parallel, mice that have a global knock-in point mutation in GRK2 (Cysteine340àSer, GRK2-C340S) that prevents

S-nitrosylation-mediated inhibition of GRK2 activity also experience declines in contractility with ISO (Huang et al. 2013). As a consequence of increased GRK2 activity,

C340S mice were more susceptible to ischemia/reperfusion (I/R) injury (Huang et al.

2013).

In summary, inhibiting GRK2 activity enhances b-AR sensitization and cardiac contractility while increased GRK2 activity diminishes b-AR sensitization and cardiac contractility. While it is easy to conclude that GRK2 worsens cardiac function, there exists a threshold of necessary GRK2 activity for proper heart physiology to protect against catecholamine toxicity. GRK2’s expression and activity are tightly regulated in order to maintain b-AR responsiveness physiologically. In the setting of CVD, like HF, GRK2 activity is increased and contributes to disease development and progression.

1.5 GRK2 in cardiac pathophysiology

Developing HF is characterized by sympathetic nervous system activation, a compensatory mechanism to maintain sufficient tissue perfusion (Lymperopoulos et al.

2013). Chronic catecholamine exposure results in reduced b-AR density and responsiveness via increased GRK2 activity, which is a natural GPCR feedback mechanism to prevent overactivation (Lefkowitz 1993; Lymperopoulos et al. 2013). b-AR dysfunction is one of the earliest molecular indicators of developing HF, even preceding overt cardiac dysfunction (Cho et al. 1999; Perrino et al. 2006). Mice that underwent trans-

11 aortic constriction (TAC) for 4 weeks did not exhibit any notable changes in echocardiographic parameters, but b-AR density and AC activity were significantly reduced compared to control (Perrino et al. 2006). Not surprisingly, elevated GRK2 protein and activity was observed in these TAC mice, but not in controls. Elevated GRK2 protein and activity has been reported even as early as 7 days post-TAC (Choi et al. 1997). These b-AR and GRK2 alterations were absent in a forced-swim model of physiological hypertrophy, supporting GRK2’s pathological role (Perrino et al. 2006). Indeed, GRK2 is the main cardiac desensitizing GRK in LV pressure overload hypertrophy compared to

GRK5 (Choi et al. 1997). GRK2 upregulation is also observed in MI, or volume overload models of HF (Maurice et al. 1999; Schumacher et al. 2015; White et al. 2000).

Activation of GRK2 activity in HF is also observed in CVD patients. Dzimiri et al reported GRK2 protein levels in myocardial biopsies were significantly elevated throughout the four chambers of the heart in dilated cardiomyopathy patients and LV volume overload patients compared to controls (Dzimiri et al. 2004). Iaccarino et al further examined myocardial biopsies obtained from HF patients undergoing cardiac transplant and found increased GRK2 expression and activity, which inversely correlated with AC activity (Iaccarino et al. 2005). This was the first study to confirm the relationship between

GRK2 activity and b-AR signaling in humans, as well as establish a correlation between lymphatic GRK2 expression and myocardial GRK2 expression. Elevated GRK2 activity has also been reported in hypertensive patients (Cohn et al. 2009; Gros et al. 1997).

Due to GRK2’s large role in HF development reported in mice and humans, multiple efforts have been made to inhibit its activity during disease. GRK2 has been

12 inhibited in HF models through small peptides, small molecules, and genetic manipulation.

One of the earliest mechanisms of GRK2 inhibition was through the βARKct peptide, as previously described. Expression of βARKct in the Muscle Lim Protein (MLP) KO mouse

(MLP−/−), a model of dilated cardiomyopathy, through crossing mice successfully rescued

MLP−/− cardiac dimensions and function (Rockman et al. 1998a). Not surprisingly, the

MLP−/− mouse displays reduced β-AR density, reduced AC activation, and elevated

GRK2 activity that was reversed with βARKct. Similarly, a heart failure model based on overexpression of sarcoplasmic reticulum Ca2+-binding protein, calsequestrin (CSQ), and subsequent upregulation of GRK2 was also rescued with βARKct (Harding et al. 2001).

βARKct gene delivery in rabbits with coronary ligation or cardioplegia was able to rescue all parameters of maladaptive signaling and remodeling (Shah et al. 2001; Tevaearai et al.

2001; White et al. 2000). Gene delivery of βARKct has also been significantly therapeutic in pig and rat models of HF (Raake et al. 2013; Rengo et al. 2009). Treatment with a selective β1-AR antagonist, metoprolol, alone was able to normalize hypertrophic gene expression and β-AR signaling, but had no effect on cardiac contractility (Rengo et al.

2009). While inhibiting β-AR activation or inhibiting GRK2 activity both dampen neurohormonal activation, inhibiting GRK2 activity was more efficacious. βARKct has also been therapeutic in pressure overload models of HF by rescuing cardiac function and

β-AR desensitization, but does not seem to mitigate hypertrophy (Choi et al. 1997;

Schumacher et al. 2016).

Small molecule inhibitors against GRK2 that have been used in a HF model include paroxetine and gallein. Paroxetine is a selective serotonin reuptake inhibitor (SSRI) class

13 of antidepressants. Paroxetine was found to have selective GRK2 inhibition through disrupting its active site in a way that ATP can no longer bind. Paroxetine was able to prevent GRK2-mediated phosphorylation of light-sensitive rhodopsin at a IC50 = 4.7 ±

0.04 nM, tubulin IC50 = 5.7 ± 0.07 nM, and thyrotropin-releasing hormone receptor at a

IC50 = ~30 µM (Thal et al. 2012). Early experiments using paroxetine in vitro showed paroxetine increasing β-AR mediated myocardial contractility after ISO. Paroxetine has not been used to treat pressure overload models of HF. However, it has been used in an MI model of HF. Chronic paroxetine treatment (5 mg/kg/day for 4 weeks) starting 2 weeks after MI induction in mice not only improved echocardiographic parameters, but restored

β-AR signaling, reduced GRK2 expression, and reduced infarct size (Schumacher et al.

2015). Gallein is a small molecule inhibitor of Gbg binding interactions making it able to disrupt Gbg-dependent recruitment of GRK2 to the plasma membrane similar to bARKct.

Kamal et al injected gallein intraperitoneally at 10 mg/kg daily for 8 weeks starting at 4- week or 12-week post-TAC (Kamal et al. 2014). Gallein treatment was reported to improve cardiac function, hypertrophy, survival, inflammation, fibrosis, and b-AR signaling.

Gallein was also able to reverse adrenal α2-AR desensitization by GRK2.

GRK2 inhibition through genetic knockdown have also been shown to improve HF outcomes. Raake et al generated mice that had normal GRK2 expression during development that was ablated after tamofixen treatment, driven by αMHC-MerCreMer and

GRK2 fl/fl crossing (αMHC-GRK2KO) (Raake et al. 2008). Tamoxifen was given 10-14 days after MI surgery to initiate loss of GRK2. Loss of GRK2 improved overall mortality, cardiac function, diastolic dimensions, and hypertrophic gene expression (Raake et al.

2008). Ablation of GRK2 before coronary ligation helped limit HF development. αMHC-

GRK2KO mice have also been subject to a 6 week TAC to model pressure overload induced HF. Loss of GRK2 attenuated cardiac hypertrophy measured through reduced

HW/BW ratio, hypertrophic markers, and cardiomyocyte length (Schlegel et al. 2017).

In summary, there have been numerous studies supporting GRK2’s pathological role in HF development and progression so much that it may be used to predict worsened

HF outcomes (Rengo et al. 2014). Multiple tools have been developed to inhibit GRK2 in

HF and they unanimously demonstrate GRK2’s therapeutic potential. There are still many gaps in knowledge on GRK2’s regulation, such as its exact mechanism of pathological increased expression, that will help further our knowledge on how to attenuate its activity.

1.6 GRK2 regulation

Alterations in GRK2 homeostasis can have profound effects on GPCR signaling in the heart as well as in HF. Little regarding endogenous GRK2 regulation is known.

Typically, expression of GRK2 is ubiquitous like housekeeping genes (Penn et al. 2000).

Its widespread expression during embryogenesis and post-natally point to its vital role molecularly. The following section will review what is known about GRK2’s regulation.

1.6.1 GRK2 protein regulation

GRK2 protein is short-lived, with reported half-lives ranging from 1 hour in Human embryonic kidney 293 (HEK293) cells to 20 hours in HL60 cells (Luo and Benovic 2003;

Penela et al. 1998). GRK2 was reported to complex with heat shock protein (Hsp) 90

(Hsp90) to promote protein stability, but not activity, in HL60 cells (Luo and Benovic

2003). Inhibition of Hsp90 with geldanamycin drastically reduced GRK2 protein half-life

(Luo and Benovic 2003). Hsp90 binding, which is enhanced by GRK2 phosphorylation at

Ser670, is also necessary for GRK2’s mitochondrial localization (Chen et al. 2013). GRK2 is rapidly turned over through the ubiquitin-dependent proteasome pathway, as its degradation was inhibited by cysteine protease inhibitor ALLN or threonine protease inhibitor lactacystin, but not lysosomal protease inhibitor leupeptin (Penela et al. 1998).

Ubiquitination at lysine (Lys) residues 19, 20, 30, and 31 on GRK2 mark for its degradation

(Salcedo et al. 2006). Mouse double minute 2 homolog and cullin 4a-DNA damage-binding protein 1-Gβ have been identified as E3-ubiquitin ligases for GRK2 (Salcedo et al. 2006;

Zha et al. 2015). GRK2 ubiquitination and degradation can be accelerated upon b-AR stimulation with ISO, evident by a 50% reduction in half-life that was reverted with proteasomal inhibitors (Penela et al. 1998). Inhibiting GRK2 degradation during ISO treatment produced a similar result as TgGRK2 mice treated with ISO: reduced b-AR signaling. Kinase dead mutants of GRK2 are resistant to degradation, suggesting its kinase activity promotes its degradation (Penela et al. 2001).

GRK2’s kinase activity, expression, and target selectivity are dictated by its interactions with different interacting partners and its post translational modifications. The most canonical example of interaction-dependent modulation is GRK2 binding to Gβγ subunits after GPCR stimulation. GRK2 has been shown to interact with non-GPCR targets as well and has emerged as a signal transducer itself. GRK2 has been reported to be phosphorylated by numerous proteins including: protein kinase A (PKA) at Ser685 to

16 increase catalytic activity (Cong et al. 2001), protein kinase C (PKC) at Ser29 to increase catalytic activity (Krasel et al. 2001; Zhao et al. 2016), mitogen-activated protein kinase

(MAPK) at Ser670 to decease catalytic activity and stability (Elorza et al. 2003; Elorza et al. 2000), Src at tyrosine (Tyr) 13/86/92 to increase catalytic activity or degradation (Penela et al. 2001; Sarnago et al. 1999), and epidermal growth factor receptor (EGFR) at

Tyr13/86/92 to increase catalytic activity (Chen et al. 2008b) (Fig. 1-3). GRK2 can also be

S-nitrosylated by nitric oxide (NO) species at cysteine (Cys) 340 to inhibit GRK2 catalytic activity (Huang et al. 2013; Whalen et al. 2007) (Fig. 1-3). GRK2-dependent actions such as b-arr recruitment and b-AR phosphorylation were also inhibited with S-nitrosylation

FIGURE 1-3. Post-translational modifications of GRK2. Reported post-translational modifications include primarily phosphorylation at the N- and C-terminal regions and one reported S-nitrosylation site within the kinase domain.

This S-nitrosylation event occurs immediately after b-AR activation as a means to combat receptor desensitization. GRK2 has been shown to complex with or interact with, but not limited to: calmodulin, calveolin-1, protein kinase B (Akt), 5' AMP-activated protein kinase (AMPK), endothelial nitric oxide synthase (eNOS), Raf kinase inhibitor protein

(RKIP), and insulin receptor substrate 1 (IRS1) (Penela et al. 2019). The interaction of

GRK2 and eNOS is particularly of interest as there exists a bi-directional and dynamic regulation of their activities where GRK2 can inhibit Akt-mediated eNOS phosphorylation and nitric oxide derived from eNOS can inhibit GRK2 (Huang et al. 2013).

It has become apparent that GRK2’s interactome is complex and extensive so that it is the primary determinant of GRK2’s regulation (Penela et al. 2019). Understanding its interactome aids in understanding GRK2’s role in not only CVD, but other pathological processes. Though GRK2’s protein regulation is well known, GRK2’s regulation at the gene and mRNA level is less understood.

1.6.2 GRK2 transcriptional and post-transcriptional regulation

In addition to increased protein levels, increased GRK2 mRNA has also been observed during HF, hypertension, and hypertrophy as described (Choi et al. 1997; Cohn et al. 2009; Schumacher et al. 2016). Little is known regarding GRK2 gene/promoter regulation, transcript stability, and transcript regulation. The GRK2 gene is ~23 kb long and consists of 21 exons and 20 introns (Penn and Benovic 1994). Notable exons include exon 1 for sequences important for GPCR recognition and exon 17 for sequences important for Gbg binding. GRK2 does not have a canonical consensus sequence (AATAAA) for 3’

18 cleavage and polyadenylation, but instead uses TATAAA (Penn and Benovic 1994). The promoter region is characterized by the absence of a TATA box and a CCAAT box, classical signals for transcription start sites. The regions immediately upstream of the ATG start codon are extremely GC rich qualifying as a CpG island. It is likely modification of these CpG islands controls GRK2 promoter activity. These promoter characteristics are common features of mammalian housekeeping genes (Zhu et al. 2008). Housekeeping genes, as their name suggests, are constitutively and ubiquitously expressed genes required for basic cell maintenance and function regardless of cell state (Eisenberg and Levanon

2013). Therefore, housekeeping genes tend to rely on post-transcriptional regulatory mechanisms to adjust gene expression (Piechaczyk et al. 1984). Consistent with these features, GRK2 is widely expressed and required for organism survival as previously described.

GRK2 promoter activity has been reported to be activated by catecholamine or ISO stimulation that results in increased GRK2 mRNA, which is reversed by b-blocker administration (Iaccarino et al. 1998). Phorbol esters (PKC activator, PMA), Gaq agonists, and a1-AR agonists, similarly increase GRK2 promoter activity and mRNA levels (Klenke et al. 2018; Ramos-Ruiz et al. 2000). Therefore, activation of Gaq or Gas signaling cascades both induce GRK2 expression. Klenke et al reported that PMA-induced GRK2 transcription was mediated by transcription factor early-growth response 1 (EGR-1)

(Klenke et al. 2018). EGR-1 has also been reported to be pathologically upregulated in peripheral blood mononuclear cells of HF patients (Cappuzzello et al. 2009). In silico analysis of the GRK2 promoter identified four putative EGR-1 binding sites that

19 surrounded a novel polymorphism (G(-43)A) identified in humans (Klenke et al. 2018).

Egr-1 mRNA levels peak 1-hour post-PMA stimulation and gradually decrease back to baseline levels by 6-8 hours. GRK2 mRNA levels peak 5 hours post-PMA stimulation and gradually decrease back to baseline levels by 8-10 hours. The co-addition of inactive PMA

(4a-PMA) or transcriptional inhibitor actinomycin D (ActD) did not cause GRK2 mRNA induction, suggesting that GRK2 mRNA upregulation by PMA is caused by transcriptional events. HEK293 cells transfected with the GRK2 promoter plasmid and pcDNA3.1, Egr-1 led to stronger GRK2 promoter activity compared to empty pcDNA3.1. Presence of the

G(-43) allele induced GRK2 expression with strong transcription factor binding more than the A(-43) allele. No other transcription factors controlling GRK2 transcription have been identified.

Post-transcriptional mechanisms of GRK2 regulation are equally lacking (Penela et al. 2019). There are no known RNA binding proteins identified for GRK2, but analysis of its 3’ untranslated region (UTR) for AU-rich elements (ARE) revealed several motifs

(http://rna.tbi.univie.ac.at/AREsite2/). Evidence of microRNA (miR)-mediated GRK2 regulation has been reported during Kaposi’s sarcoma (KS)-associated herpesvirus

(KSHV) infection (Hu et al. 2015; Li et al. 2016) and leukemia (Bai et al. 2017; Zhou et al. 2014). The KSHV genome encodes for a viral miRs called miR-K12-3 (miR-K12), that was found to inhibit GRK2 expression to promote endothelial migration and invasion (Hu et al. 2015). MiR-125a/b was reported to be upregulated in acute leukemia and leukemia cell lines, which caused resistance to chemotherapy by inhibiting apoptosis through silencing GRK2 (Bai et al. 2017; Zhou et al. 2014). Long noncoding RNA, UCA1, was

20 reported to promote gastric cancer metastasis by indirectly inducing GRK2 degradation

(Hu et al. 2015).

Other components of b-AR signaling are subject to various miR-based regulations, for example b1-AR by miR-133a (Castaldi et al. 2014). This is not surprising as miRs will often target multiple components of the same pathway to increase redundancy in pathways subject to tight regulation, which is evident in the b-AR pathway (Fischer et al. 2015).

Therefore, the identification of a miR that regulates GRK2 expression is of interest.

1.7 MicroRNAs

MiRs are short non-coding RNA duplexes, ranging from 18-24 nucleotides in length. The human genome encodes approximately 1,500 miRs that control various aspects of physiology and pathophysiology through alterations of specific genes (Jonas and

Izaurralde 2015; van Rooij 2011). They canonically are responsible for gene silencing through RNA interference pathways (RNAi), similar to small interfering RNAs (siRNAs).

MiRs regulate up to 60% of human (Friedman et al. 2009; Lewis et al. 2005). A single miR can regulate multiple transcripts simultaneously (Lim et al. 2005; Selbach et al. 2008).

Adding to their complexity, genes can also be subject to multitargeting by multiple miRs to produce additive or synergistic modulation (Vo et al. 2010). Although miRs are canonically associated with 3’UTR targeting to produce gene silencing, there have been cases where miRs have been found to target the 5’UTR or even elicit gene stabilization

(Mallat et al. 2014; Spengler et al. 2016). Regardless of their exact mechanism, it is clear

21 that miRs are a potent class of gene regulators. Multiple clinical trials have been initiated to explore the therapeutic applications of these small RNAs.

1.7.1 MiR biogenesis

MiR genes can be located intergenically between genes or intragenically embedded within exons or introns, with the majority located intergenically or antisense to annotated genes (Kim 2005; Lee et al. 2004; Valinezhad Orang et al. 2014). Within intragenic miRs, the majority are located in intronic regions (Lee et al. 2004). Intergenic miRs are regulated by their own RNA polymerase (Pol) II (Pol II) or RNA polymerase III (Pol III) promoters and expressed independently of other genes (Borchert et al. 2006). Conversely, intragenic miRs are subject to their host gene’s regulation by Pol II (Lee et al. 2004). However, there is some evidence to suggest intragenic miR:host gene ratios may not always be proportional, such as in cases where there was lack of conservation. This suggests the presence of other regulatory factors involved in miR post-transcriptional regulation that may differentially effect miRs expression vs the host gene (He et al. 2012). Up to ~50% of miRs are observed to be clustered, forming a polycistronic transcription unit (Kim 2005).

MiR transcription can also be regulated by transcription factors or epigenetics (He et al.

2007; O'Donnell et al. 2005). There is no universal miR transcription process, adding to the complexity of these small molecules. Once the Pol II miR gene is transcribed, a long primary RNA strand (pri-miR) is generated (Fig. 1-4). Pri-miR transcripts are

FIGURE 1-4. Canonical microRNA (miR) biogenesis pathway (Winter et al. 2009). MicroRNA (miR) genes are transcribed by RNA polymerase II or III to produce long primary transcripts (pri-miRs). Pri-miRs are processed by the Drosha-DGCR8 microprocessor complex to yield a precursor hairpin miR (pre-miR). Pre-miRs are exported out of the nucleus by the Exportin-5-Ran-GTP complex. Once in the cytoplasm, pre-miRs are processed again by Dicer-TRBP to produce a short miR duplex. One strand of the duplex is loaded into Argonaute (Ago2), which forms the mature miR to mediate gene silencing. The other stand is degraded.

5′-7-methylguanosine capped, 3’ polyadenylated, and may be spliced, characteristic of Pol

II transcripts (Lee et al. 2004). Pri-miRs may be edited by adenosine deaminases acting on

RNA (ADARs) that modifies adenosines to inosines, which can impact its sequence properties or subsequent processing (Winter et al. 2009). Structural characteristics of pri- miRs include a hairpin stem averaging 33 base pairs, a terminal loop, and long single- stranded overhangs on both 3’ and 5’ ends. While in the nucleus pri-miRs are endonucleolytically cleaved by the nuclear microprocessor complex (~650 kDa in humans) comprised of RNase III enzyme Drosha and DiGeorge critical region 8 (DGCR8). DGCR8 directly binds the pri-miR though its double-stranded RNA-binding domains to direct

Drosha-mediated cleavage. The double stranded stem and single-stranded overhang structures are critical for microprocessor cleavage, the terminal loop or specific sequence less so. Drosha cleaves the overhangs 11 base pairs away from the double stranded stem resulting in a short 3’ overhang to form a precursor miR (pre-miR). Not every pri-miR may undergo microprocessor cleavage. Such is the case with intron-derived pri-mRs that already have the appropriate size and structure from splicing events (mirtrons) (Winter et al. 2009). Pre-miRs are then exported out of the nucleus by Exportin-5 (XPO5) and the

Ran-GTP cofactor (Yi et al. 2003). Ran-GTP is required for XPO5 function. Structures important for XPO5-mediated nuclear export include the >14-bp double stranded stem, base-paired 5’ end, and short 3’overhang (Yi et al. 2003). Silencing of XPO5 by RNA interference caused a reduction in mature miRs up to 60% (Lund et al.

2004).

Once in the cytoplasm, the pre-miR undergoes another processing event by the

RNA-induced silencing complex (RISC) loading complex (RLC). RLC is made up of

Dicer, Tar RNA binding proteins, protein activator of PKR, and argonaute (AGO). TRBP and PACT assist in Dicer stabilization and gene silencing efficiency but are not essential for Dicer function. Dicer’s amino-terminal DExD/H-box helicase domain is autoinhibitory, but TRBP binding conformationally activates Dicer (Ma et al. 2008). Dicer, an RNAse III, cleaves off the terminal loop to generate a ~18-24 nucleotide miR duplex with 2 nucleotide overhangs on each 3’ end. Dicer processing is absolutely necessary to produce mature and functional miRs (Grishok et al. 2001). Deletion of Dicer causes the accumulation of pre- miRs (Hutvágner et al. 2001). Deletion in mice is embryonically lethal with the absence of stem cells, further supporting the crucial role of miR and Dicer function (Bernstein et al.

2003).

After pre-miR cleavage by Dicer, Dicer and TRBP dissociate from the RLC. In order to generate the mature miR, the miR duplex needs to be unwound by helicase for one strand to be incorporated into AGO, a component of the RISC. RISC is a multi-protein complex whose core components consist of one of the AGO family proteins and its interaction partner GW182. AGO1 and AGO2 are expressed in humans (Hutvagner and

Simard 2008). Studies investigating other RISC components and size have yielded different results, complicating our understanding of RISC (Hammond et al. 2001; Martinez et al. 2002; Nykänen et al. 2001). The AGO family is the only protein found in all forms of RISC and responsible for RISC function (Sontheimer 2005). Behm-Ansmant et al. reported GW182 depletion mirrored AGO1 depletion in Drosophila, highlighting the

25 significance of GW182 for miR function (Behm-Ansmant et al. 2006). AGO proteins have four conserved domains: the N-terminal, PAZ, MID, and PIWI. The N-terminal domain is involved with proper positioning and aids in unwinding of the miR within the nucleic acid binding channel of AGO (Willkomm and Restle 2015). The PAZ domain is responsible for

RNA binding, recognizing the 2 nucleotide 3’ overhang from Dicer cleavage. The MID and PIWI domains are responsible for AGO’s catalytic activity. The strand that is incorporated into AGO is deemed the “guide strand” and dictates target mRNA specificity by guiding AGO to mRNA targets. The remaining strand is called the “passenger strand”.

Passenger strands typically get degraded, but there are instances where they can also be functional. Strand flexibility and lower thermodynamic stability determines whether the 5’ or 3’ strand becomes incorporated into RISC (Khvorova et al. 2003). With a mature miR incorporated into AGO, miR-mediated gene regulation is primed to begin.

1.7.2 MiR mechanisms of action

With the mature miR loaded into AGO2, the guide strand will dictate target mRNA specificity through its seed sequence complementarity. The seed sequence refers to the 2nd-

7th nucleotide of the miR on the 5’ end with the 1st and 8th nucleotide sometimes playing a role (Huntzinger and Izaurralde 2011). Canonically, the seed sequence of an miR will target

3’UTRs of mRNAs, but there is evidence of 5’UTR and coding sequence (CDS) targeting as well (Spengler et al. 2016). Once bound to its target, miR-mediated gene silencing can occur in a number of reported mechanisms that can be grouped into two categories: (1) miR-mediated translational repression and (2) miR-mediated mRNA decay (Fabian and

Sonenberg 2012; Morozova et al. 2012). Although there have been multiple mechanisms proposed, the topic remains unclear and controversial. Complicating matters further, gene stabilization or activation by miRs has also been reported (Mallat et al. 2014; Ørom et al.

2008).

The following miR-mediated translational repression mechanisms have been proposed with supporting experimental data: (1) inhibition of translational initiation

(Djuranovic et al. 2012; Hendrickson et al. 2009; Humphreys et al. 2005; Kiriakidou et al.

2007; Mathonnet et al. 2007; Pillai et al. 2005); (2) inhibition of translational elongation

(Olsen and Ambros 1999; Seggerson et al. 2002); (3) cotranslational protein degradation

(Nottrott et al. 2006; Pillai et al. 2005); and (4) ribosome drop-off (Hendrickson et al. 2009;

Petersen et al. 2006). Translational repression was the earliest proposed mechanism for animal miRs. The first studies in Caenorhabditis elegans (C. elegans) noted lin-4 miR repressed translation of lin-14 and lin-28, but their mRNAs were still detected in polysomes

(Olsen and Ambros 1999; Seggerson et al. 2002). These C. elegans studies suggested the mRNA was intact, but translation was repressed after initiation. MRNAs competent for translation initiation must include a 5’ cap and 3’ poly (A) tail. These structures also protect the mRNA from degradation. The ribosomal 43S pre-initiation complex assembles at the

5’ cap and forms a loop with the 3’ poly (A) tail via poly (A) binding protein (PABP)

(Jackson et al. 2010). The formation of the 43S pre-initiation complex and its attachment onto mRNA require multiple eukaryotic translation initiation factors (eIFs), notably eIF4F, eIF4B, and eIF4H. Pillai et al. were the first to further investigate how miRs prevented protein synthesis by examining polysome profiles of reporter mRNAs with complementary

27 sequences to let-7 (Pillai et al. 2005). Reporter mRNA polysomes shifted toward the top of the gradient, signifying inhibition of translation initiation, which was absent when using cap-independent reporter mRNAs (ie mRNAs with an internal ribosomal entry site). It seems that RISC machinery can interfere with translation initiation by obstructing eIFs, such as eIF4E (Humphreys et al. 2005; Kiriakidou et al. 2007; Pillai et al. 2005) and eIF4F

(Mathonnet et al. 2007) from binding the 5’ cap. The concept of AGO interfering with cap binding has been challenged by studies reporting AGO is structurally unable to bind the 5’ cap through its cap-binding domain, but rather binds GW182 at this domain (Eulalio et al.

2008; Kinch and Grishin 2009). Overall, inhibition of translation initiation remains as the most frequently proposed mechanism of miR-mediated translational repression.

Experimentally it is difficult to distinguish between post-initiation mechanisms, which typically relies on polysome profiling subject to interpretation. Advances in technology allowing mass spectrometry and transcriptome-wide analysis to monitor changes in protein content against mRNA abundance in the presence of different miRs have shifted the consensus towards mRNA decay as the primary mechanism of miR-mediated repression

(Eulalio et al. 2009; Guo et al. 2010; Huntzinger and Izaurralde 2011; Lim et al. 2005).

The following miR-mediated mRNA decay mechanisms have been proposed: (1) mRNA degradation and destabilization (Behm-Ansmant et al. 2006; Eulalio et al. 2009;

Guo et al. 2010; Wu et al. 2006); (2) mRNA cleavage (Rhoades et al. 2002; Yekta et al.

2004); (3) transcriptional inhibition (Kim et al. 2008); and (4) sequestration in P-bodies

(Pillai et al. 2005; Sen and Blau 2005). In cases of perfect seed sequence to target base pairing, AGO immediately cleaves the mRNA. This cleavage phenomenon is widespread

28 in plants, but rare in animals due to lack of fully complementary targets. Rather partial complementarity directs mRNA towards the 5’ to 3’ mRNA decay pathway where mRNAs are deadenylated of their poly (A) tail by the CAF1-CCR4-NOT deadenylase complex, then decapped by decapping enzyme DCP2. This leaves the mRNA prone to 5’à3’ degradation by exonuclease XRN1 and 3’à5’ degradation by the exosome complex

(Goldstrohm and Wickens 2008). RNA expression profiling of Drosophila melanogaster

(D. melanogaster) cells depleted of AGO1, CAF1, or NOT1 showed that about 60% of all

AGO1 targets are regulated by deadenylation (Eulalio et al. 2009). Studies by Eichhorn and Bartel et al suggest up to as much as 90% of transcripts are regulated at the mRNA level (Eichhorn et al. 2014). CAF1 and NOT1 deletion resulted in similar RNA expression profiles, unsurprisingly (Eulalio et al. 2009). Additionally, loss of CAF1 or NOT1 causes a rise in many miR targets, further supporting widespread deadenylation of miR targets

(Behm-Ansmant et al. 2006). Loss of function of decapping factors also leads to a general rise in miR targets (Behm-Ansmant et al. 2006; Eulalio et al. 2007), but without restoring protein expression as these miR targets are deadenylated already. To understand the role of decapping, RNA expression analysis of D. melanogaster cells depleted of AGO1 or a decapping factor (DCP1, Ge-1, HPat, EDC3, or LSm1) was done (Eulalio et al. 2007).

Profiles from each decapping factor deletion correlated positively with each other, but not with AGO1 deletion. Expression analysis revealed up to 15% of transcripts were regulated both by AGO1 and a single decapping factor. The percentage seems low until it is considered that depleting multiple decapping factors restores miR targets more than depleting one decapping factor, suggesting decapping factors have redundancy.

Regardless, deadenylation and decapping of miR targets is a widespread effect of and significant to miR-mediated mRNA decay and now understood as the major effector of silencing.

Although the general mechanisms of miR-mediated silencing have been identified, there is still debate over the temporal relationship between these processes such as if these processes occur simultaneously, if they are coupled, and what factors cause bias. Each study investigating miR-mediated silencing utilizes a different cell type and miR of interest, which makes it difficult to draw any universal conclusions. Even though it is established that mRNA decay is a predominant phenomenon, it is undetermined if it is an independent mechanism. MRNA decay as a general consequence of translational inhibition cannot be excluded (Eulalio et al. 2007). Selbach et al. reported most of their targets on a proteome-wide scale were suppressed both the mRNA and translation level especially at later time points, which was highly dependent on seed sequence (Selbach et al. 2008).

Within the seed, the number of mismatches seemed to positively correlate with translational inhibition (Selbach et al. 2008). Overall, research suggests there can be several coexisting mechanisms occurring, but what is phenotypically observed is highly dependent on the intrinsic cell properties ie mRNA content, RNA binding proteins, splicing marks, gene transcription, mRNA degradation rate, and mRNA-RISC features (Eulalio et al. 2007;

Kong et al. 2008; Morozova et al. 2012).

1.7.3 MiRs in cardiovascular disease

Loss of mature miRs through cardiac specific (aMHC) deletion of Dicer in mice

(MHCCre/+; Dicerflox/flox) is postnatally lethal, paralleling embryonic lethality observed with pre-natal deletion (Chen et al. 2008a). MHCCre/+; Dicerflox/flox mice die within 4 days of birth due to dilated cardiomyopathy characterized by defects in the sarcomeres, highlighting the absolute necessity of miR function in maintaining a healthy heart. Just as we observe differentially regulated mRNAs in HF such as increased GRK2, miRs are dysregulated in parallel and contribute to disease development and progression (Matkovich et al. 2009). Studies by Matkovich et al suggest that changes in miR signatures during end- stage HF and left ventricular assist device support are significantly more reflective of cardiac function compared to mRNA signatures (Matkovich et al. 2009). MiRs were originally researched in the setting of cancer, playing fundamental roles in carcinogenesis, cell proliferation, differentiation, and apoptosis. In 2006, van Rooij et al. published the first cardiac study examining the miR profiles of cardiac hypertrophic mouse models compared to controls (van Rooij et al. 2006). Mice that underwent TAC or expressed activated calcineurin (CnA) had 27 and 33 differentially increased miRs, respectively, compared to controls, with 21 miRs commonly increased. TAC and CnA models had 15 and 14 miRs differentially downregulated, respectively, compared to controls.

Examination of miR from LV tissue from idiopathic end-stage heart failure patients revealed common dysregulated miRs compared to the hypertrophic mouse models including miR-24, miR-125b, miR-185 miR-199a, and miR-214. This was the first suggestion that HF had a unique miR signature. Overexpression of any of these key miRs

31 found to be upregulated during hypertrophy was sufficient to elicit hypertrophic responses alone in vitro and in vivo. For example, mice with a cardiac specific overexpression of miR-195 (TgMiR-195) resulted in hypertrophic growth that led to dilated cardiomyopathy by 6 weeks of age (van Rooij et al. 2006). Other miRs identified to play a role in modulating cardiac remodeling include miR-1, miR-18b, miR-133, and miR-21 (Barwari et al. 2016;

Divakaran and Mann 2008). One molecular characteristic of HF is the reactivation of fetal gene expression, such as atrial natrietic factor (ANF) and b-myosin heavy chain (MHC), both of which are considered hypertrophic markers (Thum et al. 2007). Interestingly, not only is gene expression between human fetal and failing hearts closely related, but miR expression as well. Today, miRs have been identified to perturb virtually all cellular processes such has contractility (Wahlquist et al. 2014), fibrosis (Creemers and van Rooij

2016), and extracellular matrix regulation (Divakaran and Mann 2008). Table 1-2 summarizes several cardiac miRs altered during HF and their functional impact. It is now understood that during HF miR signatures become significantly altered and contribute to the aberrations in gene expression observed.

The specific experimental gain-of-function or loss-of-function of key cardiac miRs has demonstrated physiological and pathophysiological relevance of miRs in the heart.

Given that numerous miR species along with their gene targets have been identified to play direct roles in HF, research has looked to using miRs as therapeutic tools (Thum 2012).

1.7.4 MiRs as therapeutic tools

In HF models, rescue of dysregulated miR levels has shown therapeutic success in animal models. In animal models, overexpression can be achieved through the

Table 1-2. MicroRNAs (miRs) involved in heart failure. miR Expression in HF Function in cardiovascular system miR-1 ↓ Development and function of cardiac and skeletal muscle miR-15 ↑ Induces apoptosis miR-21 ↑ Induces apoptosis, modulates eNOS activity, and fibrosis miR-23a ↑ Promotes cardiac hypertrophy miR-30 ↓ Antagonizes apoptosis and fibrosis miR-27 ↓ Promotes cardiac hypertrophy miR-133a ↓ Regulation of b-adrenergic receptor signaling miR-181 ↓ or ↑ Regulation of cardiac hypertrophy miR-199a ↓ Antagonizes apoptosis Adapted from (de Lucia et al. 2017; Oliveira-Carvalho et al. 2013)

use of miR mimics. MiR mimics consist of double-stranded RNAs that mimic, as the name suggests, the endogenous miR duplex prior to its unwinding and incorporation into RISC.

The use of mimics will increase gene targeting and further suppress gene expression.

However not only will mimics be taken up by target tissues, but off-target tissues as well.

MiRs have also been expressed through viral vectors like adeno-associated viruses, which allow the opportunity for the addition of reporters or increased tissue specificity (Wahlquist et al. 2014). MiRs can also be inhibited through miR inhibitors, which consist of single stranded RNAs that can bind to endogenous mature miRs through complimentary base pairing. MiR inhibitors will antagonize miRs resulting in reduced gene silencing. Both mimics and inhibitors consist of modified oligonucleotides to increase stability and resist nuclease degradation (Fig. 1-5). Although, the amount of modifications done to mimics is limited due to its requirement of having to be incorporated into RISC. As a result, inhibitor design is much more advanced (Thum 2012). Chemical modifications include (1) 2’-O- methylation—the addition of a methyl group to 2’ hydroxyl on ribose, (2) Locked nucleic acid (LNA)—the addition of a bridge connecting the 2’oxygen to the 4’carbon on ribose, and (3) increasing phosphorothiate linkages between nucleotides (van Rooij 2011).

Therefore, the use of miRs as medical intervention drugs have been of interest in human patients. In 2018, the Food and Drug Administration (FDA) approved the first small noncoding RNA (ncRNA)-based drug for treating hereditary transthyretin-mediated amyloidosis: patisiran (Kristen et al. 2019; Zhang et al. 2019b). Patisiran is a small interfering RNA (siRNA) encapsulated in lipid nanoparticles that targets transthyretin

(TTR) in the liver, operating through RNAi like miRs do (Zhang et al. 2019b). In addition to transcription and processing differences compared to miRs, siRNAs are characterized by being fully complementary to their mRNA targets to produce mRNA cleavage (Lam et

34 al. 2015). Patisiran certainly sets an exciting precedent for small ncRNA therapeutics, especially for miRs. Indeed, there are several biopharmaceutical companies dedicated to developing miR therapeutics, including Regulus Therapeutics, Synlogic Therapeutics, and miRagen Therapeutics. These companies have pioneered miR drug candidates into early phase clinical trial, but none have succeeded into phase 3 trials yet (Table 1-3) (Hanna et al. 2019).

There is also evidence that drug resistance or toxicity may be mediated by miR mechanisms (Roca-Alonso et al. 2015; Wagh et al. 2017). Doxorubicin (DOX) is a widely used chemotherapy given intravenously indicated to treat multiple types of cancer.

A B HO Base HO Base O O

OH OCH3 OH O 2’-O-methylation Locked nucleic acid C HO Base O

O OH Base S P O O

O OH Phosphorothioate linkage FIGURE 1-5. Types of commonly used microRNA (miR) modifications. (A) 2’-O- methylation modification consists of the addition of a methyl group to the 2’ hydroxyl of the ribose ring. (B) Locked nucleic acid modification consists of the ribose ring being “locked” by an extra bridge that connects the 2’ oxygen and 4’ carbon. (C) Phosphorothioate linkage modification consists of the non-bridged oxygen being replaced by a sulfur atom. Modifications are highlighted in red. 35

Table 1-3. MicroRNAs (miRs) used in interventional clinical trials miR; drug name Clinical trial #; phase status Disease indication miR-34; MRX34 NCT01829971; phase 1 liver cancer, lymphoma (terminated) NCT02862145; phase 1 melanoma (withdrawn) miR-92; MRG 110 NCT03603231; phase 1 wound healing, heart failure (completed) miR-16l MesomiR-1 NCT02369198; phase 1 mesothelioma, lung cancer (completed) miR-122; Miravirsen NCT02508090; phase 2 Hepatitis C virus (completed) NCT02452814; phase 2 (completed) NCT1200420; phase 2 (completed) miR-29; MRG-201 NCT03601052; phase 1 keloid, fibrous scar (active, not recruiting) miR-21; RG-012 NCT02855268; phase 2 Alport syndrome (suspended, sponsor decision) miR-155; MRG-106 NCT03713320; phase 2 cutaneous T-cell lymphoma (recruiting) NCT03837457; phase 2 (enrolling by invitation) Adapted from (Hanna et al. 2019)

However, patients may be afflicted with serious side effects, including cardiotoxicity leading to HF. Roca-Alonso et al. discovered one mechanism for DOX-mediated cardiotoxicity was through downregulation of the miR-30 family after DOX administration

(Roca-Alonso et al. 2015). MiR-30 downregulation was also observed after MI. The miR-

30 family was found to modulate b1-AR, b2-AR, and Gia-2 (Roca-Alonso et al. 2015).

Thus, the loss of miR-30 may compare to settings of increased neurohormonal signaling observed in HF, while the administration of miR-30 may compare to the effect of b- blockers. Although this topic of miR-mediated drug resistance and toxicity require further investigation, studies suggest miR-based drugs may also be beneficial as part of a combination therapy for cardiovascular disease.

Not only are miR expression profiles changed within the heart, but researchers have noted altered miR expressions in blood or plasma (Nouraee and Mowla 2015). This observation gave rise to developing miRs as disease biomarkers. Biomarkers are clinically relevant, as they can help give insights to pathology through minimally invasive procedures. The development of miR-based biomarkers has been more successful than miR-based drugs. Multiple phase 4 clinical trials are recruiting or have been completed for monitoring miR biomarkers in patients receiving an intervention, including for breast cancer, pulmonary hypertension, and Parkinson’s disease. Overall, miR therapeutics could represent a new class of pharmaceutical drugs. They possess attractive characteristics for drug design such as their small size, stability, and specificity and could make useful therapeutic tools for cardiovascular disease.

1.8 Aging

Edward Lakatta, M.D., Chief of the Laboratory of Cardiovascular Science at the

National Institute on Aging defines aging as the “manifestation of progressive time- dependent failure of molecular mechanisms that create disorder within a system of DNA and its environment” (Lakatta 2015). The environment can refer to anything from the cytoplasmic proteins in a cell to the planet’s atmosphere. The accumulation of molecular disorder or failure over time results in progressive changes in the structure and function of all tissues in the body, including the cardiovascular system. This is evident across humans and non-humans alike, a universal phenomenon. The CDC reports that by 2030, the US population aged ≥65 will outnumber the population aged ≤18, pointing to a shift towards becoming a significantly aged nation. This undoubtedly will be accompanied by heightened prevalence of age-related diseases. Understanding how age and disease interact and manifest in organ systems is of utmost importance.

1.8.1 Hallmarks of general aging

Aging research in general has become more appreciated and advanced in recent years. Rightfully so, as aging is the primary risk factor for numerous prevalent diseases such as cancer, diabetes, and CVD. Multiple theories of aging have been proposed over time (Table 1-4), but each are limited in their application or may not encompass all aspects of aging (Volkova et al. 2005). In an effort to develop a comprehensive understanding of aging and unite the current aging research, López-Otín et al. identified and classified cellular and molecular hallmarks of aging (López-Otín et al. 2013). The hallmarks of aging

38 consist of (1) genomic instability, (2) telomere attrition, (3) epigenetic alterations, (4) loss of proteostasis, (5) deregulated nutrient signaling, (6) mitochondrial dysfunction, (7) cellular senescence, (8) stem cell exhaustion, and (9) altered intracellular communication

(López-Otín et al. 2013). This section will briefly review each hallmark and give an example of how it impacts CV aging.

Genomic instability represents the common denominator of all aging models. The stability of DNA receives continuous challenge from the environment that eventually results in damage. Damage could result from DNA replication errors, reactive oxygen species (ROS), chemical agents, or spontaneous hydrolytic agents and cause a number of genetic lesions such as point mutations, telomere shortening, and DNA breaks. Genomic instability can also arise from defects in nuclear lamins that maintain nuclear architecture.

The intrinsic DNA repair mechanisms available that are meant to remedy DNA damage will experience dysfunction themselves with aging, resulting in an accumulation of DNA damage with increasing age. In the heart, mitochondrial DNA (mtDNA) has been shown to be increasingly damaged with age due to mitochondrial ROS. This theory is supported by cardiac specific overexpression of catalase in mice that resulted in prolonged lifespans and preserved heart function partially due to minimizing mtDNA genomic instability (Dai et al. 2009).

Telomeres are heterochromatic structures made up of repeated nucleotide sequences (TTAGGG) located at the terminal ends of eukaryotic linear chromosomes.

They are bound, protected, and regulated by a protein complex called shelterin. Due to shelterin’s physical binding on telomere regions, DNA mutations that occur in these

Table 1-4. Major theories of aging Theory Description Programmed death Cell death and decline with aging is part of our inherent evolutionary design. Mutation Aging is due to an accumulation of mutations that may not be accumulation deleterious early in life but will be deleterious at advanced ages. Antagonistic Expands on the mutation accumulation theory stating that pleiotropy aging selects for genes with differential effects at different ages, in other words genes with pleiotropic effects. Protein modification Aging results from the accumulation of impaired or post- translationally altered proteins that negatively impact organ function and promotes cellular senescence. Free radical Aging results from chronic oxidative damage to cells due to an increase of reactive oxygen species as an intrinsic result of aerobic metabolism. May result from increased inflammation. Energy Restriction Related to the “Free radical theory”, energy expenditure and anabolic pathways contribute to aging. Implementing caloric restriction to reduce energy intake extends lifespan. Longevity Gene There are genes that positively or negatively impact senescence to impact age-related disorders and longevity. Lifestyle and other extrinsic factors will also modify longevity. Evolutionary There exist multiple classes of genes that escape natural Biologic selection and can modulate senescence. Immunological Aging results from low-grade prolonged autoimmune response caused by the immune system. Cellular Senescence The accumulation of senescent cells causes aging.

40 regions are inaccessible and unrecognizable to DNA repair machinery (Palm and de Lange

2008). DNA polymerases are ineffective in replicating telomeres, therefore telomeres become increasingly shorter with each division. When telomeres reach a certain length, cells can become apoptotic or “replicative senescent” that can lead to tissue dysfunction.

Cells that undergo replicative senescence adopt a senescence-associated secretory phenotype (SAAP), causing chronic inflammation (Tchkonia et al. 2013). Indeed, short telomeres are thought to be a primary cause of aging. Although the heart is classically thought to be post-mitotic and quiescent, there is some evidence that telomeres in cardiomyocytes are subject to deterioration with age and under diseased conditions like HF

(Sharifi-Sanjani et al. 2017; Terai et al. 2013). The case of age-dependent cardiac telomere shortening remains controversial, but evidence for telomere shortening in CVD has been confirmed in biopsied cardiac tissue from CVD patients (Booth and Charchar 2017; Oh et al. 2003; Sharifi-Sanjani et al. 2017; Terai et al. 2013). Telomere attrition can also result from shelterin dysfunction, which has been reported to occur in the vasculature with age

(Liu et al. 2019).

Epigenetics refers to any modifications in gene expression that are independent from changing the genomic sequence itself. Epigenetics are critically important when considering how a gene is expressed. Epigenetic gene regulation and modifications include post-translational histone modifications ie histone acetylation, DNA covalent modification ie methylation, transcriptional alterations, miRs, mRNA modification, and chromatin remodeling. Histone modifications have been shown to influence longevity in multiple animal models. Reduction in histone methylation through deletion of trimethylation

41 complexes in C. elegans and Drosophila were able to extend lifespan (Greer et al. 2010;

Siebold et al. 2010). The family of NAD-dependent protein deacetylases and ADP- ribosyltransferases called sirtuins (Sir, SIRT) has been heavily studied for their longevity or anti-aging properties experimentally. SIRTs’ influence on healthy aging stem from their ability to improve genomic stability and metabolic efficiency through deacetylating histones and other key proteins. Overexpression of Sir2 in worms, flies, and yeast extended their lifespans, although there is controversy over the degree of lifespan extension (López-

Otín et al. 2013). The overexpression of the mammalian homolog of Sir2, SIRT1, was able to reduce age-related DNA damage and liver cancer, but did not impact longevity (Herranz et al. 2010). Heart lysates from young (~6 years) and old (~21 years) monkeys revealed that SIRT1 protein levels significantly rise with age (Alcendor et al. 2007). Alcendor et al. generated transgenic mic with cardiac specific overexpression of SIRT1 (TgSIRT1) to investigate the function of SIRT1 in the aging heart (Alcendor et al. 2007). It seemed that cardiac hypertrophy was positively correlated with cardiac SIRT1 expression, as examined

6-month-old TgSIRT1 mice with varying degrees of overexpression. Mice with high overexpression of SIRT1 also developed LV dilation and dysfunction compared to low overexpressors. In addition to aging, SIRT1 has been observed to increase in both dog and mouse models of pressure overload (PO) induced HF (Alcendor et al. 2007; Alcendor et al. 2004). Conversely, SIRT1 was observed to decrease after I/R injury in mice. Cardiac- specific loss of SIRT1 has been reported to worsen myocardial infarction after I/R and hypertrophy after PO (Hsu et al. 2010; Sanz et al. 2019). These data suggest that SIRT1 is cardioprotective and anti-aging within a therapeutic window.

Protein homeostasis, or proteostasis, declines with age and age-related diseases.

Maintaining proteostasis is extremely important for cell function and there are multiple mechanisms to preserve protein stability and functionality. Misfolded or damaged proteins must be appropriately degraded to prevent proteotoxicity, along with their subsequent replacement with correct proteins. Molecular chaperones and heat shock proteins play an important role in maintaining proper protein conformation. The presence and accumulation of misfolded proteins is sufficient to cause human disease such as myopathies and amyloidosis (Koga et al. 2011). Accordingly, impaired proteostasis is also observed during aging. Cardiomyocytes are especially susceptible to impaired proteostasis compared to most other cell types due to their highly specialized proteins, reduced mitotic capability that limits clearance of protein aggregates by cell division, and high rates of oxidative phosphorylation that generates proteotoxic agents (Henning and Brundel 2017). Predmore et al examined proteasome function in failing hearts from patients and reported reduced proteasome activity, as measured by a reduction in chymotrypsin-like activity and accumulation of ubiquitinated proteins, compared to nonfailing hearts (Henning and

Brundel 2017). In addition to the heart, proteolytic pathway activities have also been reported to decrease with age in lymphocytes, skeletal muscle, liver, spinal cord, and vasculature (Koga et al. 2011; Nakayama et al. 2016).

Aging is accompanied by changes in metabolic pathways, which result in insulin resistance, reduced growth hormone (GH) and insulin-like growth factor-1 (IGF-1), impaired nutrient sensing, and mitochondrial dysfunction that lead to increased adiposity

(Barzilai et al. 2012). Insulin and IGF-1 activate signaling (IIS) pathways that alert the cell

43 that glucose is available. IIS pathways can converge at FOXO family of transcription factors and mTOR to carry out cell growth, cell survival, protein synthesis, autophagy, and transcription of key genes (López-Otín et al. 2013). GH and IGF-1 levels have been observed to decease during normal aging and mouse models of aging, but continuous decreases can extend lifespan. In contrast, mTOR activity has been observed to increase in mice with age in hypothalamic neurons (Yang et al. 2012). Interventions to combat these signaling pathways, like through rapamycin administration (mTOR inhibitor) or caloric restriction (CR), have been shown to prolong lifespan in multiple eukaryotic species

(Colman et al. 2009; Fontana et al. 2010; Harrison et al. 2009). In parallel, genetic inhibition of mTOR activity in mice has increased their lifespans (Lamming et al. 2012;

Selman et al. 2009). These studies suggest that anabolic activity accelerates aging and minimizing anabolic activity promotes longevity and anti-aging. In the heart, this remains true. Quarles et al. demonstrated that rapamycin treatment in 22-24 month old mice improved diastolic function and LV stiffness (Quarles et al. 2020). Likewise, inhibition of

IIS pathways in the heart also limits cardiac aging, although the underlying mechanisms are not fully understood (Boudina 2013).

As previously mentioned, mtDNA accumulates damage through aging so it comes to no surprise that mitochondrial dysfunction represents a hallmark of aging. ROS, which is the basis for the mitochondrial free radical theory of aging (Table 1-5), are a byproduct of high oxygen consumption during various cellular processes. Although there is no question ROS promotes cell damage through oxidative stress, ROS as effectors of aging specifically have been called into question. Certain studies that increase or decrease

44 mitochondrial ROS through antioxidant enzyme knockdown or overexpression, respectively, have not proven to increase lifespan in mice (Pérez et al. 2009; Van Remmen et al. 2003). However, mitochondrial dysfunction due to mitochondrial gene mutations, independent of ROS, has been shown to accelerate aging (López-Otín et al. 2013). Declines in mitochondrial bioenergetics with age arise from multiple mechanisms, of which some have already been mentioned, including reduced mitochondrial biogenesis, telomere attrition, mtDNA damage, disorganization of the electron transport chain complexes, changes in mitochondrial structure, and defects in mitochondrial autophagy. Indeed all these age-related mitochondrial changes occur in the heart and are significantly disruptive due to the heart’s high energy demands; these are reviewed in Tocchi et al (Tocchi et al.

2015).

Cellular senescence is characterized by cell cycle arrest and as previously mentioned can be a result of telomere attrition. Transition to senescence is thought to prevent the proliferation of damaged cells and signal for their removal. However, during aging, there is accumulation of senescent cells that results from either an increase in senescent cell generation or a decrease in senescent cell clearance. Beyond telomere attrition, there exists other independent stimuli that triggers senescence. Non-telomere

DNA damage is also able to promote cell cycle arrest and thus is used as a senescence marker experimentally. Aberrations in multiple oncogenic or mitogenic genes have been identified to induce senescence, as reviewed by Gorgoulis and Halazonetis. Serrano et al reported the first evidence of oncogenic stimuli, through derepression of the INK4/ARF locus, activating cellular senescence (Serrano et al. 1997). p16INK4a is a cyclin-dependent

45 kinase inhibitor and preventing G1 to S phase transition, effectively acting as a tumor suppressor. Unsurprisingly, expression of the INK4/ARF has been suggested to be an aging biomarker in the kidney, ovary, and heart that is attenuated with CR (Krishnamurthy et al.

2004). CR in aged animals also decreased cellular senescence measured through senescence-associated β-galactosidase (SA-β-gal) expression, an in vivo senescence marker.

During aging there is a decline in the regenerative capacity of tissues due to stem cell attrition. Functional stem cell decline has been observed in the brain, bone, and muscle fibers (López-Otín et al. 2013). Overall decreases in regeneration can be caused my telomere shortening. It also correlates with stem cell DNA damage and expression of cell- cycle inhibitors like gene products from the INK4/ARF locus (Janzen et al. 2006; Rossi et al. 2007). Although the heart is traditionally classified as a post-mitotic organ, seminal studies in the early 2000s by Orlic et al and Beltrami et al reported that c-kit+ stem cells from the bone marrow and the heart were able to regenerate the myocardium in a rodent models of MI (Beltrami et al. 2003; Orlic et al. 2001). Harnessing the power of cardiac stem cells to regenerate the heart as a therapeutic strategy for CVD has been intensively studied and led to numerous clinical trials, which could also be extrapolated to combat aging. However, the very existence of these stem cell populations is a highly controversial topic. They very well may exist, but their regenerative capacity may be negligible and certainly not robust enough to provide cardiac protection during aging nor heart disease.

Nevertheless, Cesselli et al examined c-kit+ cardiac stem cells (hCSCs) from human atrial

46 tissue and observed age was a crucial factor in hCSCs telomere shortening and p16INK4a expression (Cesselli et al. 2011).

Lastly, the final hallmark of aging consists of altered intercellular communication.

Intercellular communications via endocrine, neuroendocrine, or neuronal systems become aberrant with age, which has a negative systemic effect. Aging is accompanied by a low- grade pro-inflammatory phenotype coined ‘inflammaging’ (Salminen et al. 2012).

Inflammaging can be caused by the other hallmarks mentioned like mitochondrial dysfunction but is also influenced by neurohormonal signaling ie renin-angiotensin, adrenergic, and IIS. It is proposed that increasing ROS with age can initiate the formation of Nod-like receptor 3 (NLRP3) inflammasomes and activate other pro-inflammatory pathways that result in an increase of inflammatory cytokines ie interleukin 1 beta (IL-1b), tumor necrosis factor (TNF), and interferons (INF). This chronic inflammation can inhibit stem cell function itself, compounding the aging process, as well as progress atherosclerosis. During aging there is increase neurohormonal activation of the sympathetic nervous system, which has maladaptive consequences on heart function. (de

Lucia et al. 2018a)

In summary, López-Otín et al identified and characterized 9 universal hallmarks of aging. It is evident these hallmarks are highly interconnected and can influence or exascerbate each other. For example, multiple hallmarks may be traced back to DNA damage. Thus, we see that there are systemic molecular changes that occur during aging that drastically impact organ function.

1.8.2 Characteristics of cardiovascular aging

Aging represents one of the biggest risk factors for CVD and associated mortality

(Benjamin et al. 2019). Prevalence sharply increases in individuals 65 years and older.

Although the Centers for Disease Control already report CVD as the leading cause of death in the US, the proportion of deaths from CVD increases further when looking at elderly age groups. Among ages 65 and older 25.5% of all deaths are from CVD compared to

29.2% in ages 85 and older. Therefore, when discussing CVD and its therapeutics age cannot be ignored. As previously discussed, the aging process systemically and negatively impacts the body and the heart is no exception. The heart and vasculature undergo a number of age-associated changes that eventually lead to HF (Fig. 1-6) (North and Sinclair 2012;

Steenman and Lande 2017; Strait and Lakatta 2012). In fact, many HF-related CV changes are also observed in normal aging even in the absence of overt disease. Although aging itself does not cause HF, due to age-dependent changes previously discussed, the threshold for HF development is lowered. HF represents the common final CV state, where age- associated changes in the CV system and other organs converge. Age-related CV changes can be categorized into 4 levels: functional, structural, cellular, and molecular.

The heart and vasculature experience a number of functional declines overtime as a consequence from the molecular, cellular, and structural CV aberrations. These changes impair the CV system’s ability to respond to increased workload as needed in situations such as exercise. This decreased adaptability is indicated by a decrease in cardiac reserve, which is observed even in the absence of coronary artery disease (Lakatta 1994). During exercise, maximal oxygen uptake, oxygen consumption, heart rate (HR), and end diastolic

48 volume (EDV) have all been reported to decrease in aged individuals (Strait and Lakatta

2012). Increases in vascular peripheral vascular resistance (PVR), systolic blood pressure

(SBP), and cardiac end systolic volume (ESV) occur in parallel (Strait and Lakatta 2012).

Cardiac output (CO) also declines with aging likely due to inability to accelerate HR.

Overall resting function may still be normal. Diastolic dysfunction is the most recognized consequence of aging. Systolic function as measured by ejection fraction (EF) remains mostly stable with age with an exception during exercise, but other components of systole

FIGURE 1-6. Physiological aging-associated cardiovascular changes. The heart and the vascular undergo numerous changes during natural aging as a result of the progressive time-dependent failure of molecular mechanisms that involve the interaction between intrinsic and extrinsic cellular factors. Arrows dictate how each parameter is changed with aging.

49 may also be changed (Strait and Lakatta 2012). Despite this, there is growing appreciation for systolic dysfunction with aging as echocardiographic analysis becomes more advanced with the development of speckle tracked-based echocardiography (STE) (de Lucia et al.

2019). Diastole refers to the phase in the cardiac cycle where the heart relaxes to fill with blood subsequent to contraction during systole. Normal diastolic filling occurs in two phases (1) early passive filling of the ventricles under no pressure (E wave, peak velocity) and (2) late active filling of the ventricles under pressure from atrial contraction (A wave, peak velocity). Alterations in both phases can occur with aging and is measured through the E/A ratio. The aged heart fills with blood more slowly compared to a young heart, which limits the duration of early phase filling. Impaired LV relaxation, as a result of increased myocardial stiffness, occurs as a result of prolonged isovolumic relaxation time

(IVRT). The majority of ventricular filling shifts towards late phase diastole and manifests as a decrease in E/A ratio. As diastolic function progressively worsens to a “restrictive filling pattern”, E/A ratio (>2) is observed to increase due to elevated left atrial pressures.

Within the vasculature, significant impairment of endothelial-dependent vasodilation of resistance arteries has been reported, which parallels the impaired relaxation observed in the heart (Muller-Delp et al. 2002). Declines in vascular vasodilation will have negative consequences on maximal vasodilatory capacity and maximum O2 consumption, contributing to the age-related decline in exercise capacity. Overall, the aged heart experiences a decline in function especially during increased workload and diastole. These functional aberrations are a consequence of age-related changes to the CV structure.

There are significant structural alterations to the heart and vasculature with age.

Overall heart shape becomes more spheroid than elliptical. Most observable is an increase in heart size and LV wall thickening due to cardiomyocyte hypertrophy. Left and right ventricular myocytes were reported to grow by 110 microns3/year and 118 microns3/year, respectively (Olivetti et al. 1991). Despite an increase in cardiomyocyte size, the total number of cardiomyocytes is reduced with age due to increased necrotic and apoptotic cell death (Kajstura et al. 1996). Loss of cardiomyocytes with age is accompanied by an increase in both reactive and reparative collagen accumulation in the heart, which is observed in both animals and humans (Eghbali et al. 1989; Gazoti Debessa et al. 2001).

Age-related cardiac fibrosis can explain the functional consequences previously described, including diastolic dysfunction from ventricular stiffening (Biernacka and Frangogiannis

2011). Fibrotic remodeling may even negatively impact systolic function, especially in hypertensive conditions, by disrupting myocardial excitation-contraction coupling

(Iwanaga et al. 2002; Janicki and Brower 2002). Song et al. noted increased fibrosis and fat deposition within the sino-atrial node, atrio-ventricular node, and bundles, which interferes with impulse conduction along the cardiac conduction system (Song et al. 1999).

Increased fibrosis has been reported to occur perimysially (around groups of muscle fibers), endomysially (around individual myocytes), and even perivascularly (around vasculature).

In fact, the progressive cardiac hypertrophy observed with age is a consequence of age- related peripheral vasculature changes that promote vascular stiffening. Vascular stiffening is a result of collagen accumulation, elastin degeneration, endothelial dysfunction, and vascular wall thickening. Vascular wall and intima-media thickening results from vascular

51 smooth muscle cell (VSMC) proliferation and VSMC hypertrophy, and accumulation of collagens (Xu et al. 2017). The accumulation of collagens is accompanied by a decrease in elastin content in vessels. Age-related vascular stiffening has critical impacts on hemodynamics, increasing hemodynamic load, increasing pulse pressure, increasing SBP, and decreasing diastolic BP (DBP). Capillary luminal volume and density is observed to decrease with age not just in the heart, but in the brain and skeletal muscle as well (Anversa et al. 1994; Rakusan et al. 1992; Sonntag et al. 1997). Overall, we see cardiomyocyte and

VSMC hypertrophy that leads to cardiac and vascular hypertrophy, respectively. This hypertrophy is accompanied by fibrotic remodeling that further damages CV function.

The structural and histological observations during aging result from cellular-based changes. The heart is made up of mostly cardiomyocytes by volume, ~75%, but make up

25-35% of total cell number (Pinto et al. 2016). The remaining cells mainly consist of fibroblasts (~11%), endothelial cells (~60%), VSMCs (~6%) and hematopoietic-derived cells (5-10%) (Pinto et al. 2016). During aging, the heart undergoes changes in its cellular composition. The population of cardiomyocytes decline with age from increased apoptosis and necrosis, which evidently cannot be replenished by the small pool of cardiac stem cells

(Goldspink et al. 2003). Olivetti et al examined 67 hearts from individuals aged <20 to 90 years old without any signs of CVD and calculated 38 million and 14 million cardiomyocyte nuclei per year are lost after the onset of puberty (Olivetti et al. 1991). This significant loss of cardiomyocytes with aging provides a structural-based explanation for reduced cardiac reserve and increased cardiac dysfunction observed. The remaining aged cardiomyocytes are more likely to be senescent, marked by p16INK4a (Chimenti et al. 2003).

Endothelial cells undergo a similar fate to cardiomyocytes, experiencing age-dependent apoptosis and senescence (Xu et al. 2017). This may not be surprising considering there is endothelial cell dysfunction with age. Fibroblasts, on the other hand, are activated and adopt a “myofibroblast” phenotype marked by α-smooth muscle actin expression.

Myofibroblasts are the cellular effector of fibrotic responses, including during aging.

Various molecular changes are responsible for the cellular changes observed in aging. The phenotypic switch to myofibroblasts is mediated by autocrine and paracrine factors, including angiotensin II (AngII), endothelin-1 (ET-1), and transforming growth factor-beta 1 (TGF-b1). These factors also play roles in stimulating cardiomyocyte hypertrophy and contributes to aging phenotypes (Biernacka and Frangogiannis 2011;

Vajapey et al. 2014). Impaired endothelial-dependent vasodilation can be linked to loss of endothelial nitric oxide synthase (eNOS) activity leading to reduced nitric oxide (NO) bioavailability (Collins and Tzima 2011). NO and related compounds are critical vasodilators, anti-inflammatory agents, antioxidants, and anti-thrombotics. NO is necessary to maintain healthy vasculature. Loss of NO also contributes to endothelial senescence. Impaired function of the heart can be explained through the age-related abnormalities in the b-AR signaling pathway. As mentioned, the b-AR and downstream signaling factors are vital for regulating cardiac function in response to sympathetic stimuli.

In response to CV stress during aging, levels of E and NE are elevated in the circulation and synapses, respectively. In addition, clearance of NE is diminished with age that further contributes to elevated levels. This chronic catecholamine exposure mimics the SNS activation that is observed in HF to elicit receptor desensitization in aging. Unsurprisingly,

53 there have been reports of decreases in b-AR density, both b1-AR and b2-AR, in aged rats, humans, or guinea-pig hearts (Ferrara et al. 2014). Despite conflicting studies that report no change, b-AR density is likely to be reduced and this provides the molecular basis to reduced reserve capacity with age. Gai expression and activity has also been observed to increase with age, contributing to reduced cAMP and impaired b-AR responsiveness.

1.8.3 GRK2 in aging

It is evident that GRK2 also plays a role in reducing b-AR responsiveness and other

GPCRs during HF, but whether or not it mediates b-AR impairment during aging is not well studied. Of the available studies on GRK2 in the aging heart, there are conflicting conclusions making it difficult to determine a definitive physiological role. Schutzer et al examined aortas from 6 week and 24 month rats to determine if GRKs had role in age- related decline in b-AR-mediated vasodilation (Chapman et al. 1999; Schutzer et al. 2001).

Increased total GRK2 activity was increased in 24-month rat aortas compared to 6 weeks.

Within both the cytoplasmic and membrane fraction both GRK2 and GRK3 protein levels were elevated in aortic lysates, but GRK5 was not elevated in the membrane fraction.

Although multiple GRKs were elevated, Schutzer et al did not confirm a functional link to b-AR desensitization experimentally, but can be assumed (Schutzer et al. 2001). Leosco et al’s study in 2003 would later establish this functional link in rat aorta (Leosco et al. 2003).

GRK2 expression was confirmed to be expressed in the aortic endothelium. Measurement of GRK2 activity in rat carotid arteries revealed increased GRK2 activity in aged rats (24 month) that corresponded to a decrease in b-AR density. Elevated GRK2 activity and

54 reduced b-AR density in old rats was returned to levels similar to young rats when subject to swim exercise (Leosco et al. 2003). In the aged heart, the increase in GRK2 does not seem to hold true. Evaluation of GRK activity in human right atria did not reveal any increases in activity in neither the membrane or cytosolic fraction in elderly patients with coronary heart disease without HF compared to children, but was elevated in HF

(Leineweber et al. 2003). Nor were there any differences reported in GRK2 expression or activity in 24-month rat LVs (Xiao et al. 1998). However, GRK2 activity beyond its kinase activity have not been assessed. Assessing kinase activity only would mean ignoring the plethora of kinase-independent interactions GRK2 participates in, for example its interaction with Akt. Further, heart tissue analysis of GRK2 expression may not reveal cell specific changes in expression that may occur. There have not been investigations on

GRK2 levels in physiologically aged human vasculature. Elevated GRK2 expression and/or activity has also been detected in rat liver where it mediates hepatic b1-AR overactivation during aberrant glucose and lipid metabolism during aging (Shi et al. 2018).

GRKs may not be the cause of age-related reductions in b-AR density in cardiomyocytes, but whether they are responsible for the reductions in the vasculature or they play other roles during aging remains to be investigated. Although there is no clear consensus between

GRK2’s role in the CV system during aging, it is a topic that demands further research when it is considered those with CVD are predominantly elderly.

1.9 Summary and objectives

The abundant research on CV GRK2 has strongly documented GRK2’s critical role in physiology and pathophysiology. Under normal basal conditions, GRK2 maintains b-

AR homeostasis through phosphorylated-mediated receptor deactivation. During disease,

GRK2 expression and activity rise beyond physiological levels to plunge b-AR density and responsiveness down to dysfunctional levels. Furthermore, how this relationship between

GRK2 and b-AR function extrapolates into aging to impact CVD remains an important research question as the aged molecular landscape will change how disease states are dealt with. Some evidence suggests GRK2 activity is increased in the aged vasculature, but perhaps not the myocardium. GRK2 inhibition has proven to be a powerful tool to rescue

GRK2-mediated b-AR dysfunction, as well as rescue other signaling aberrations.

Therefore, there is upmost interest in understanding how GRK2 is regulated during disease and aging in order to develop efficient strategies for GRK2 inhibition.

The research contained within this dissertation comprises of two independent studies that aim to understand GRK2 regulation in HF and aging.

1.9.1 Cardiac GRK2 and microRNAs in heart failure

Given the evidence that (1) multiple components of the b-AR pathway are regulated by miRs ie. miR-133 and miR-30, (2) there is evidence of viral miRs targeting GRK2 in lymphocytes, (3) GRK2’s nearly ubiquitous expression similar to housekeeping genes suggests post-transcriptional mechanisms of regulation, we hypothesized that GRK2 can be post-transcriptionally regulated by a miR. Furthermore, this GRK2-targeting miR would likely be dysregulated during HF to contribute to pathological increases in GRK2. With an

56 interest in miR-based drugs, manipulation of this miR during HF may be used as a therapeutic to ameliorate HF phenotypes through antagonizing GRK2. In this study, we examined miR expression profiles in mouse models of MI and selected miR candidates to test in vitro and in vivo for their ability to silence GRK2 gene expression. Ultimately, we hope to develop a miR-based therapeutic to inhibit GRK2 expression during HF.

1.9.2 Loss of dynamic regulation of GRK2 by nitric oxide and cardiovascular

aging

As previously mentioned, GRK2 undergoes dynamic S-nitrosylation at Cys340, which inhibits its kinase activity, subsequent to b-AR activation as a means to limit receptor desensitization by GRK2. In order to investigate the physiological role of this post-translational modification, mice were generated with a global knock-in (KI) point mutation in GRK2 where Cys340 was mutated to a Ser (GRK2-C340S). Although GRK2-

C340S (8-10 weeks) had increased infarct size after I/R injury, at baseline there were no phenotypic alterations observed with increased GRK2 activity. With an interest in understanding the impact of increased GRK2 during aging and to highlight chronic GRK2- inhibition mediated benefits, we allowed the GRK2-C340S mice to age at least for a year

(52 weeks). We hypothesized that chronic GRK2 overactivity would cause cardiovascular remodeling and dysfunction in aged mice. Understanding how chronic GRK2 overactivity impacts the cardiovascular system gives us broader insight how upregulated GRK2 in the context of age-associated CVD further exacerbates the aged molecular landscape and contributes to disease.

CHAPTER 2

CARDIAC GRK2 AND MICRORNAS IN HEART FAILURE

2.1 Introduction

Heart disease remains the leading cause of mortality with little to no improvements on hospitalization rates (Hall et al. 2012). In the face of declining cardiac function due to cardiac injury, b-AR desensitization by pathologically increased GRK2 expression and activity represents the most profound threat to cardiac function. GRK2 inhibitory strategies have proven beneficial in improving structural and functional outcomes in multiple cardiac injury models. Therefore, identification of compounds or molecules to inhibit GRK2 are of high interest.

Identification of GRK2 regulatory molecules or interaction partners can provide points of interest where experimental GRK2 inhibition can occur. Research has identified numerous protein interactors and protein modifications that regulate GRK2 activity, which has revealed an extensive and complex GRK2 interactome. However, methods of GRK2 transcriptional and post-transcriptional regulation are sparse (Penela et al. 2010a; Sato et al. 2015b). Given that GRK2 has a nearly ubiquitous and constitutive expression, it provides a strong premise for post-transcriptional regulatory mechanisms fine-tuning its expression in response to stimuli. It is established that GRK2 promoter activity increases in response to pharmacological agents that induce vasoconstriction or hypertrophy.

Exposure to catecholamines or other adrenergic agonists are also sufficient in increasing

GRK2 transcription and expression (Klenke et al. 2018; Ramos-Ruiz et al. 2000). One

58 transcription factor identified to be involved in PMA-induced GRK2 transcription was

EGR-1 (Klenke et al. 2018). We postulate that GRK2 mRNA can be post-transcriptionally regulated and this regulation is modulated in HF to promote the pathological increase in

GRK2 expression and activity. One method of post-transcriptional regulation is through miRs. MiRs, as discussed, can be powerful mediators of gene regulation, canonically gene silencing. Cardiac miR profiles have been shown to be altered in all models of cardiac injury and contribute to phenotypes observed such as hypertrophy, contractile dysfunction, and fibrosis. Indeed, miR-133 has been identified to target multiple components of the b-

AR pathway, which is impaired during HF and contributes to b-AR dysfunction (Castaldi et al. 2014).

MiR-133 is one of the most abundant cardiac miRs and its expression protects against cardiomyocyte hypertrophy, fibrosis, and promotes proper cardiomyocyte development/proliferation (Carè et al. 2007; Liu et al. 2008; Matkovich et al. 2010).

Overexpression of miR-133 in neonatal cardiac myocytes prevented PE and ET1 induced hypertrophy, measured by blunted protein synthesis compared to cells without miR-133.

Thus, there is an inverse relationship between miR-133 and hypertrophy, regardless of the hypertrophic stimuli. MiR-133 was validated to target multiple genes involved in hypertrophy development—RhoA (Rho subfamily of GTPases), Cdc42 (Rho subfamily of

GTPases), and NELFA/Whsc2 (negative regulator of RNA polymerase II). In addition to these targets, Castaldi et al. also identified b1-AR, AC6, the catalytic subunit of PKA, and the exchange protein activated by cAMP (EPAC) as direct targets of miR-133-mediated silencing. Reduced silencing of these additional miR-133 targets involved in the β1-AR

59 signaling cascade during hypertrophy is compatible with the findings in Carè et al, as β-

AR dysfunction also accompanies pressure-overload hypertrophy (Carè et al. 2007; Choi et al. 1997; Perrino et al. 2006). The addition of miR-133 in neonatal cardiomyocytes significantly limited cAMP generation and AC6 activity after NE stimulation with selective

β2-AR-selective inhibition through ICI-118,551 compared to control cells (Castaldi et al.

2014). Castaldi et al generated a cardiac-specific overexpression of miR-133 inducible mouse model (Tg133). Tg133 mice were protected against pathological hypertrophy development and declining cardiac function with TAC compared to sham, as expected due to miR-133-mediated inhibition of β1-AR signal transduction similar to β-blocker therapy.

Tg133 mice observed a ~20% reduction in GRK2 mRNA compared to sham after TAC, which corresponded to reduced cell death. Although miR-133 is not strongly predicted to directly target GRK2, partial normalization of GRK2 activity could result from inhibiting

β1-AR stimulation as seen with b-blocker therapy (Iaccarino et al. 1998). The identification of miRs that can regulate β-AR signaling opens up the possibility that GRK2 may also be subject to miR-mediated regulation.

MiRs capable of targeting GRK2 have been identified in non-cardiac diseases such as Kaposi’s sarcoma and leukemia, as mentioned previously. Therefore, we believe that there exists a GRK2-targeting cardiac miR that regulates GRK2 mRNA levels during normal physiology and pathophysiology. Although miR-mediated regulation may not be a primary nor sole way GRK2 expression is increased in HF, identification of a miR would uncover a new understanding of how this key enzyme is regulated.

Thus, we sought to determine whether there was a miR-based regulatory pathway impacting GRK2 expression. Specifically, we used miR expression profiling methods to assess how miRs were altered in a MI model of HF along with bioinformatic prediction software to identify candidate miRs. We assessed our candidate miRs in their ability to inhibit GRK2 mRNA and protein expression, as well as their binding capabilities to the

GRK2 3’UTR in vitro. Successful candidates were advanced to in vivo investigation with the ultimate goal of using this candidate miR as a HF therapeutic through inhibiting GRK2 expression and activity.

2.2 Experimental procedures

2.2.1 Experimental animals

All animal procedures were carried out according to National Institute of Health

Guide for the Care and Use of Laboratory Animals and approved by the Animal Care and

Use Committee of Temple University. Adult male C57BL/6 (9-12 weeks) mice purchased from Jackson Laboratory were utilized for all in vivo experiments. Mice were euthanized by isoflurane inhalation and cervical dislocation, as approved by the American Veterinary

Medical Association (AVMA).

2.2.2 Echocardiography

All echocardiography was performed using the Vevo 2100 imaging system from

VisualSonics as previously described (de Lucia et al. 2019; Schumacher et al. 2016). Hair from the mouse chest was removed with Nair >24 hours before echocardiography. Mice

61 were anesthetized with 3% isoflurane and maintained at 1-3% during procedures.

Traditional echocardiography (TE) utilized the MS400 (30-MHz centerline frequency) probe to obtain B- and M-mode images from parasternal long-axis and short-axis to evaluate LV internal diameters (LVIDs), fractional shortening (FS), end diastolic volume

(V;d), end systolic volume (V;s), stroke volume (SV), cardiac output (CO), and heart rate

(HR).

2.2.3 Myocardial infarction

MI surgery was performed as previously described (Gao et al. 2010). Mice were anesthetized with 2% isoflurane inhalation using an isoflurane vaporizer (Viking Medical) without intubation. While on a heated surface, a small incision was made on the left side of the chest to expose the fourth intercostal space. The heart was gently pushed out of the thoracic cavity and a slipknot was tied around the left coronary artery (LCA) using a 6-0 silk suture. The heart was placed back with manual evacuation of the pneumothoraxes. The skin was closed by suture. Sham operated animals did not receive LCA occlusion. The mouse was then removed from isoflurane inhalation and closely monitored for recovery.

Buprenorphine was administered (0.1 mg/kg b.wt, s/c) for pain management.

2.2.4 Myocardial injections

In vivo ready miR mimics (250 nmol), purified by HPLC, were obtained from

ThermoFisher. Mimics were obtained for the following miR species: Negative Control #1

(miR-Neg) and kshv-miR-K12-3-5p (miR-K12). Mimics have an approximate molecular

62 weight of 14,000-15,000 g/mol. Mimics were resuspended in sterile PBS at 10 mg/mL or

20 mg/mL. Mimics were prepared with Invivofectamine™ 3.0 Reagent (ThermoFisher) according to a modified protocol recommended by the manufacturer. A 1:1 mix between mimic and complexation buffer was prepared ie 10.8 µL of miR-Neg + 10.8 µL complexation buffer; 6 µL miR-K12 + 6 µL complexation buffer. Invivofectamine™ 3.0

Reagent, at room temperature (RT) was added to the mixture at the following ratio: 1.2 µg mimic to 1 µL lipid. The diluted mimic and Invivofectamine™ 3.0 Reagent mixture were immediately vortexed and incubated at 50°C for 30 minutes. Solutions were spun down.

Intramyocardial injections were performed in a similar manner to MI surgery, but instead of coronary occlusion the mimic solution was injected across 3 points of the LV using a

Hamilton syringe, 31G with ½ inch length. Twenty-five µL of mimic solution was diluted with 5 µL sterile PBS prior to injection. Dosage was estimated to be ~0.0214 mg/heart or

1 mg/kg, depending on the mouse body weight (BW). Hearts were harvested 2 days after injection.

2.2.5 In vitro miR mimic transfections

Neonatal ventricular cardiomyocytes (NRVMs) and the AC16 human cardiomyocyte cell line were utilized for in vitro miR mimic and inhibitor transfections.

NRVMs were plated at 0.75x106 cells/well in a 6-well plate in Ham’s F10 media, 10% horse serum (HS), 5% fetal bovine serum (FBS), and 1% penicillin/streptomycin (P/S). On

Day 2, NRVMs were serum starved with Ham’s F10 media, 1% FBS, 1% P/S for 5 hours.

MiR mimics for miR-Neg, miR-K12, or hsa-miR-378a-5p (miR-378a) were complexed

63 with Lipofectamine® RNAiMAX Transfection Reagent (ThermoFisher) in Opti-MEM™

I Reduced Serum Medium (ThermoFisher) according to manufacturer’s protocols with either 25 pmol, 50pmol, or 75 pmol mimic. Lipofectamine® RNAiMAX Transfection

Reagent is a proprietary RNAi-specific cationic lipid formulation for cellular transfection.

Cells were transfected for 48 hours before harvesting.

2.2.6 Immunoblotting

Cells for western blotting were harvested in RIPA buffer containing protease and phosphatase inhibitors (10 mM Tris, 0.5 M EGTA, and 0.5 M EDTA solution) and scraped.

Cell lysates were vortexed and frozen before use. Protein lysates from LV tissue were also homogenized in RIPA buffer containing protease and phosphatase inhibitors. Protein lysates from LV tissue were subject to 30-minute rotation in 4°C. All lysates were centrifuged to pellet cell debris. Quantification of protein content was performed through the Pierce™ BCA Protein Assay (Bio-Rad) before lysates were mixed with protein loading dye. Following SDS-PAGE and transfer to nitrocellulose membranes, primary antibody incubations were performed overnight at 4° C. Visualization of Western blot signals was performed using secondary antibodies coupled to Alexa Fluor 680 or 800 (Invitrogen

Molecular Probes) or IRDye 680 or 800 (LI-COR Biosciences) on an Odyssey CLx infrared imager (LI-COR Biosciences). Odyssey version 1.2 infrared imaging software was used to process all images.

2.2.7 RNA isolation and Semi-quantitative RT-PCR

Cells were lysed in TRIzol™ Reagent (ThermoFisher) and were froze. Total RNA was eluted in nuclease-free water. LV tissue was homogenized in TRIzol™ Reagent using the VCX 130PB ultrasonic processor from Sonics & Materials Inc. Solutions were centrifuged to pellet cell debris. Total RNA isolation was performed using the Direct-zol™

RNA Miniprep kit (Zymo Research) according to manufacturer’s protocol. RNA was quantified using the NanoDrop™ 2000/2000c Spectrophotometer (ThermoFisher). cDNA was generated from isolated RNA using the iScript cDNA Synthesis Kit (Bio-Rad). Semi- quantitative PCR on cDNA was done using SYBR Green (Bio-Rad) and a final concentration of 100 nM of gene-specific oligonucleotides GRK2 and Rn18s (18S) on a

CFX96 Real Time System with BioRad CFX Manager 2.1 software. (Bio-Rad).

Quantification was through 18S rRNA normalization and compared using the ΔΔCt method.

2.2.8 Creation of stable Ad293-GRK2 cell line

Ad293 cells were transfected with 300 ng of pmiRGLO-3’UTR-GRK2 using

Lipofectamineâ 2000 Reagent (ThermoFisher). The pmiRGLO-3’UTR-GRK2 plasmid consists of the pmiRGLO Dual-Luciferase miR Target Expression Vector (Promega) backbone with the entire length of murine GRK2 3’UTR (UCSC Genome Browser) or a mutated GRK2 3’UTR sequence inserted using XbaI and SalI restriction enzymes. Both

GRK2 3’UTR sequences were sequenced by GenScript. Cultures were allowed to grow for

48 hours. After, cells were trypsinized and titration of G418 (1.0-1.4 mg/mL) and titration

65 of A293 cell number (1-2000 cells) were done in order to determine the optimal plating density and G418 concentration. Cells were incubated under standard conditions and fed every 3-5 days with fresh selection medium. Cell clones were analyzed under a light microscope. Clones were selected and further tested. One clone was ultimately expanded for WT and mutant GRK2 3’UTR, Ad293-GRK2 and Ad293-mGRK2 respectively. Stable lines were maintained with 1.0 mg/mL of G418. Media for Ad293-GRK2 or Ad293- mGRK2 cells cultured for experiments did not contain G418.

2.2.9 3’UTR analysis

Ad293-GRK2 or Ad293-mGRK2 cells were seeded in a 24 well plate with 1x105 cells/well with DMEM 10% FBS, 1% P/S. After cell adhesion, cells were serum starved with DMEM 1% FBS, 1% P/S for 5 hours. Cells were transfected with Lipofectamine®

RNAiMAX Transfection Reagent (ThermoFisher) in Opti-MEM™ I Reduced Serum

Medium (ThermoFisher) with 25 pmol of miR mimic or inhibitor and allowed to incubate for 48 hours. Cells were scraped in a Luciferase Lysis Buffer (LLB) containing 25 mM

Tris/phosphate, 4 mM EGTA, 1% Triton X-100, 10% glycerol, and 2mM dithiothreitol

(DTT). The pH of LLB was adjusted to 7.8 with dilute o-phosphoric acid prior to glycerol and Triton X-100 addition. DTT was added immediately prior to use. Protein quantification of cell lysates was done with a Bradford Protein Assay (Bio-Rad). Equal amounts of protein were added to a flat-white clear-bottom 96 well plate in triplicate. Firefly luciferase and

Renilla luciferase activities were measured using the Dual-Glo® Luciferase Assay System

(Promega) according to manufacturer’s protocol. Luminescence was measured using the

Infinite® M1000 PRO plate reader (Tecan). Luminescence values were normalized to background wells containing untransfected Ad293 cells.

2.2.10 Statistics

All values in the text and figures are presented as mean ± SEM. Statistical significance between two groups with one variable was determined by student’s unpaired t-test. Statistical significance between three or more groups with 1 variable was determined by one-way ANOVA with Tukey’s multiple comparisons test. Statistical significance between three or more groups with 2 variables was determined by two-way ANOVA with

Tukey’s multiple comparisons test. Probabilities of 0.05 or less were considered to be statistically significant. Statistical calculations were done in Prism 7 (GraphPad).

2.3 Results

2.3.1 Myocardial infarction changes miR expression profile

Male C57BL/6 underwent sham or MI surgery. Echocardiography was performed

2 weeks after surgery to assess cardiac function and to assess severity of surgery (Fig. 2-

1A). As expected, mice had various degrees of cardiac dysfunction depending on MI surgery. Three mice were selected to submit isolated total RNA from LV tissue for miR microarray analysis through LCSciences. The three mice selected were selected for similar

EF: 24.39%, 26.57%, and 24.59%. As expected, MI surgery induced drastic changes in miR expression in varying degrees of significance (Fig. 2-1B). In accordance with the

FIGURE 2-1. Changes in microRNA (miR) expression after myocardial infarction. (A) Ejection fraction (EF, %) measured through traditional echocardiography in sham and 2-week post MI mice. (n=5,9). (B) miR microarray clustering graph of statistically significant and high signal miR transcripts in sham and 2-week post-MI mice with similar EF% post-MI ~25%. (n=3).

68 literature, our microarray results reported a decrease in miR-133a, miR-30, and miR-29c while an increase in miR-214, miR-199a, and miR-21 (van Rooij et al. 2006).

2.3.2 Analysis of miR candidates with bioinformatic prediction

In order to examine whether a miR was capable of binding the 3’UTR to inhibit its expression, we generated a reporter construct (pmirGLO-GRK2) that contained the full mouse GRK2 3’UTR sequence downstream of the firefly luciferase gene (luc2) under the control of the human phosphoglycerate kinase (PGK) promoter within the pmirGLO Dual-

Luciferase miRNA Target Expression Vector (Promega) (Fig.2-2A). A separate Renilla luciferase gene (hRluc) under the SV40 promoter was also included as a control. Successful binding of a miR to the 3’UTR will destabilize the 3’UTR transcript, which will also inhibit luc2 translation. Reduced luminescence as a result of reduced luciferase activity signifies miR-mediated silencing. The 1900 miRs that were included in the miR microarray were categorized as either being upregulated or downregulated after MI (Fig. 2-2B). MiRs that were considered downregulated were examined for their seed sequence binding through base complementarity to the 3’UTR of GRK2 (yielded 343 miRs) as well as through

TargetScan7.1 (yielded 29 miRs). If a miR is able to regulate GRK2 expression, it would likely be downregulated during HF to contribute to pathological increases in GRK2. It is also possible for the miR to not be extensively downregulated in order to result in altered regulation of GRK2 expression. We wanted to also look for mouse and human conservation, as this would yield a more translational mechanism and suggests tighter binding affinity. Ad293-GRK2 cells, which stably express the GRK2 3’UTR downstream

69 of a luciferase reporter, were transfected with mimics for each miR for 48 hours before assessing luciferase activity. MiR-K12 was utilized as a positive control for miR-mediated silencing of GRK2. Multiple miRs were able to successfully bind and inhibit GRK2’s

3’UTR reported through reduced luminescence, while others had no effect (Fig. 2-2C).

Although some candidates appear to strongly inhibit GRK2 such as miR-15a-3p, miR-134-

3p, and miR-3620-5p—they were very poorly expressed in the heart and were not pursued.

2.3.3 Selected miR candidates reduced GRK2 expression in vitro

The following miRs were further investigated: miR-Neg, miR-K12, miR-181a, miR-181b, miR-181c, miR-378a, miR-497a, and miR-125a. Details on their location on the 3’UTR targeting, seed sequence characteristics, and known targets within the cardiovascular system are shown in Table 2-1. Of particular interest were the miR-181 family members: miR-181a, miR-181b, and miR-181c (miR-181d is also a member). The miR-181 family members are conserved miRs that share the same seed sequence, but slightly different mature sequence. The miR-181 family’s 3’UTR binding site (1039-1045) highly overlaps with the miR-K12 binding site (1040-1046) and this is conserved in both human and mouse sequences (Fig. 2-3A). This binding site resides within highly similar sequence region of 3’UTR (~900-1100) between human and mouse (Fig. 2-3B). MiR-23a is also a conserved miR and overlaps with miR-K12 binding in both human and mouse.

Conserved pairing improves miR prediction and reduces false-positive predictions (Bartel

2009). MiR-125a, although a published target, did not score very highly during prediction nor had conserved binding sites.

FIGURE 2-2. Identifying miR candidates for targeting GRK2 3’UTR. (A) Illustration of the target sensor construct (pmiRGLO-GRK2) using the pmiRGLO vector. hPGK, human-phosphoglycerate kinase promoter; luc2, firefly luciferase gene; GRK2 3’UTR, mouse GRK2 3’ untranslated region. (B) Workflow of miR microarray screening for candidate miR selection. All miRs detected via microarray were separated by whether they were up- or down-regulated after MI, with the highest interest in the latter. miR seed sequences were analyzed solely on sequence complementarity to the 3’UTR of mouse and human GRK2. TargetScan 7.0 prediction software was used for human and mouse GRK2 to select miRs of interest. (C) Screening luciferase activity in relative luminescence units (RLU) in Ad293-GRK2 cells, which stably express the target sensor construct shown in (A), transfected with various miR mimics (25 pmol). (n=3).

n tni hmlgAG: uohg rltd . Bcl 5. related Autophagy dismutase.Foxo3a: tumor ForkheadTNFAIP3: necrosis box O3. factor alpha homolog.ATG5: tensin and UTR. GRK2: G protein 3’ 2 the at 7mer UTR. 3’ 2 the at Bold miR miR miR miR miR Characteristics of mi miR miR miR miR Table 2 ------

497a 125a 378a 23a 181a 181c K12 181b = conserved sites. Seed match refers to the type of seed sequence match to 3’ UTR target. 8mer: an exact 8 nt match nt 8 exact an 8mer: target. UTR 3’ to match sequence seed of type the to refers match Seed sites. conserved = nd nd - - 1.

- 3p to 7 to 8 to ------3

5p 5p 5p 5p 5p 5p -

th th

nucleoti nucleotide of the mature miR mature the of nucleotide m8: an exact match at the 2 the at match exact an m8: Accession MIMAT0003453 MIMAT0000443 MIMAT0000731 MIMAT0000078 MIMAT0000257 MIMAT0000256 MIMAT0000674 MIMAT0002193 croRNA (miR) croRNA de of the mature miR to the 3’ UTR with an adenosine on the 3’ end of the seed sequence on the on sequence seed the of end 3’ the on adenosine an with UTR 3’ the to miR mature the of de - coupled receptor kinase 2. mt

82 82 score% 23 28 70 80 83 N/A Context++ candidates

to 8 to to the 3’ UTR with an adenosine on the 3’ end of the seed sequence on the on sequence seed the of end 3’ the on adenosine an with UTR 3’ the to

nucleotide of the mature miR to the 3’ UTR. 7mer UTR. 3’ the to miR mature the of nucleotide 7mer Match 7mer 7mer 7mer 7mer 7mer 7mer 8mer Seed - COX1: COX1: mitochondrial cytochrome c oxidase 1. PTEN: phosphatase ------m8 m8 A1 m8 m8 m8 m8

- : B 2:

587 741 549 1041 1039 1039 1040 (mouse) Position in 1039 UTR 3’ - el lymphom cell - - - 593 747 555 - - - - - 1045 1045 1047 1045 1046

- induced protein 3. TLR: toll 1046 UTR 3’ N/A 675 1048 1048 1046 1046 1047 (human) Position in 2 MSD mnaee superoxide manganese MnSOD: 2. a -

682 ------1054 1054 1052 1052 1053 1052

PTEN, TGF Sarcoma) Cardiovascular System TNFAIP3, TLR2, TLR4TNFAIP3, Bcl LDHA Foxo3a MnSOD, Bcl PTEN, ATG5, mt ( GRK2 Targets Published in - - COX1 2, LC3B -

A1: an exact match exact an A1: aoi s Kaposi’ - like receptor

- 2, p53

FIGURE 2-3. Mouse and human alignment of highly similar sequences within GRK2 3’UTR. (A) Seed sequence of miR- K12 (bold) and miR-181 family (green) targeting the 3’UTR of GRK2 in human (1028-1087) and mouse (1021-1080). 3’UTR targeting shows overlap and conservation. (B) Screenshot of alignment of highly similar sequences between mouse GRK2 3’UTR (Query) and human GRK2 3’UTR (Sbjct).

Mimics were transfected into NRVMs or AC16s for 48 hours to validate whether our miRs of interest were able to regulate GRK2 expression at the mRNA and/or protein level in vitro, as 3’UTR analysis is an artificial system that only looks at binding ability and may not translate to a cardiac cell. The 48-hour post-transfection time point is commonly used in the literature to assess miR oligo function, as well being recommended by manufacturers. (Hu et al. 2015) In NRVMs, we assessed 3 different doses of mimics (5 pmol, 25 pmol, and 75 pmol) to determine which was the optimal dose as NRVMs have lower transfection efficiencies. Using miR-K12 and miR-378a as an example, the amount of GRK2 protein expression negatively correlated with the amount of mimic (Fig. 2-

4A,B,C). 75 pmol was selected as the mimic amount for NRVMs, which corresponds to

37.5 nM of mimic (Fig. 2-4D). Although we assessed several miRs in NRVMs (Fig. 2-

5A), use of AC16s proved to have better transfection efficiencies at reduced concentrations and reduced variability between biological replicates (Fig. 2-5B,C,D,E). The positive control miR-K12 transfected into AC16s showed significant knock down of GRK2 protein expression (~50-70%). The conserved miR candidates, miR-181 family members (miR-

181a/b/c) and miR-23a, also elicited significant downregulation of GRK2 expression. miR-

378a was variable in its silencing and does not significantly affect GRK2 protein.

Interestingly, miR-125a was unable to sufficiently silence GRK2 in the cardiomyocyte like it was reported to do in leukocytes cells. This simply could be from a difference in cell environment. Overall, the use of miR-K12 as a means to reduce GRK2 expression through a miR-dependent mechanism was validated. This miR has not been used in cardiomyocytes

FIGURE 2-4. GRK2 inhibition dose response with miR mimics. Neonatal rat ventricular myocytes (NRVMs) were transfected for 48 hours with 5, 25, or 75 pmol of miR-K12 or miR-378a to assess transfection efficiency. (A) Representative Western blot of GRK2 (~80 kDa) and GAPDH (~37 kDa) protein from transfected NRVM lysates. Ctrl = untransfected cells; Lipo = LipofectamineRNAiMAX only. Quantification of Western blot data of NRVMs miR treatments normalized to (B) untransfected NRVM lysates (n=4) or (C) miR-Neg (dotted line) (75 pmol) transfected NRVM lysates. (n=4). (D) Quantification of Western blot data of NRVMs with 75 pmol treatment of miR-Neg, miR- K12, or miR-378a. *p<0.05, **p<0.01, ***p<0.001 as determined by one-way ANOVA compared to miR-Neg with Dunnett’s post hoc analysis.

FIGURE 2-5. Transfection of various miR mimics decreased GRK2 protein expression. (A) Quantification of Western blot data of NRVMs that were transfected with miR mimics for 48 hours. GRK2 protein was normalized to miR-Neg condition. (n=3-12). Representative Western blot images of GRK2 (~80 kDa) and GAPDH (~37 kDa) protein from AC16s transfected (50 pmol) with (B) miR-Neg, miR-K12, miR-181c, miR-23a, miR-125a, and miR-378a or (C) miR-Neg, miR-K12, and miR-181a. Ctrl: untransfected AC16s. Quantification of Western blot data of AC16s that were transfected with miR mimics normalized to (D) untransfected AC16s or (E) miR-Neg (dotted line). (n=3-6). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 as determined by student’s t-test compared to miR-Neg.

76 before. MiR-181a, miR-181b, miR-181c, and miR-23a successfully reduced GRK2 protein.

Next we assessed whether these miRs of interest had any effect on GRK2 mRNA expression. Although protein is the ultimate read-out of miR-mediated gene silencing, measuring mRNA levels could give insights to how these miRs behave. We observed significant reduction in GRK2 mRNA levels after miR-K12 and miR-181c transfection

(Fig. 2-6A,B). Given that miR-K12 forms an exact match through its entire seed sequence, it may be able to initiate mRNA cleavage. To assess whether miR-K12 influenced GRK2 mRNA half-life, we treated NRVMs with ActinomycinD (ActD, 5µg/mL), a transcription inhibitor, for up to 3 hours after 48 hours of mimic transfection. DMSO, ActD, or the addition of miR-Neg for up to 3 hours did not affect GRK2 half-life in NRVMs (Fig. 2-

6C). However, NRVMs that were transfected with miR-K12 had a trending decrease in

GRK2 half-life by 60 minutes and was significantly decreased at 180 minutes, suggesting that miR-K12 is able to destabilize GRK2 mRNA (Fig.2-6C).

2.3.4 Antagonizing miR candidates increased GRK2 expression in vitro

To further establish if there was a direct and functional link between our miR candidates and GRK2, we transfected AC16s with miR inhibitors (50 pmol) of miR-Neg, miR-K12, miR-181a, miR-181b, miR-181c, and miR-23a for 48 hours before analysis of

GRK2 expression. The addition of miR inhibitors would blunt any miR-mediated silencing of GRK2, if it was existent. However, given that miR-mediated silencing is not the sole way of GRK2 regulation, miR inhibition at basal levels may not produce a significant

FIGURE 2-6. Transfection of various miR mimics decreased GRK2 mRNA abundance. AC16s were transfected for 48 with miR-Neg, miR-K12, miR-181a, miR- 181b, miR-181c, miR-23a, or miR-378a. Quantification of qRT-PCR data from RNA isolated from untransfected or transfected AC16 normalized to (A) untransfected cells or (B) miR-Neg. (n=4-6). NRVMs were transfected with miR-Neg, miR-K12, or miR-378a for 48 hours before treatment with 5µg/mL ActinomycinD (ActD) for 0, 60, or 180 minutes before RNA isolation and (C) quantification of GRK2 mRNA through qRT-PCR (n=2-4). *p<0.05, **p<0.01 as determined by student’s t-test compared to miR-Neg.

78 change. We saw small trending increases in GRK2 protein expression with anti-miR-181a and anti-miR-23a (Fig. 2-7A,B,C). GRK2 protein was significantly increased in AC16s with anti-miR-K12 or anti-miR-181c transfection (Fig. 2-7A,B,C). We also observed trending increases in GRK2 mRNA expression with anti-miR-K12 or anti-miR-181b transfection, while anti-miR-181a significantly increased GRK2 mRNA (Fig.2-7D). MiR-

K12 is not endogenously found in cells unless infected with KSHV, but since it has a highly similar sequence to miR-181a/b/c/d or miR-23a it may be able to inhibit these miRs to increase GRK2 expression. Overall, at basal levels antagonizing miR-K12, miR-181a, miR-181c, or miR-23a increased GRK2 expression levels to some extent. It is likely that if cells were treated with a GRK2-upregulating stimulus, the addition of miR inhibitors would be more apparent.

2.3.5 Selected miR candidates bind to GRK2 3’UTR

Our candidate miRs were then validated for their ability to bind to the GRK2

3’UTR through 3’UTR analysis using the Ad293-GRK2 cells. Ad293-GRK2 cells were transfected as described for 48 hours before measuring luciferase activity. Luminescence quantification was significantly reduced in miR-K12 and miR-181a conditions, confirming that these miRs can bind to GRK2’s 3’UTR (Fig. 2-8A). To further validate these binding sites, the 3’UTR was mutated between sites 1040-1046: GATGTG à CTTAGAC (Fig. 2-

8B). Mutating these 7 nucleotides encompasses binding sites for miR-K12, miR-181 family, and miR-23a, allowing for one set of mutations to be made to assess all three miRs.

Ad293-mGRK2 cells that stably express the mutated 3’UTR were transfected with miR

FIGURE 2-7. Candidate miR inhibitors induce small increases in GRK2 expression. (A) Representative Western blot images of AC16s were transfected with miR inhibitors (50 pmol) for 48 hours and GRK2 expression by protein through Western blot quantification normalized to (B) untransfected AC16s (n=4-8) or (C) anti-miR-Neg and (D) mRNA through quantification of qRT-PCT data were examined. (n=3-5). *p<0.05 as determined by student t-test compared to anti-Neg.

FIGURE 2-8. Validated 3’UTR analysis of GRK2. Ad293-GRK2 cells were transfected with miRmimics (25 pmol) for miR-Neg, miR-K12, miR-181a, miR-181b, or miR-23a for 48 hours before measuring luciferase activity through relative luminescence units (RLU). (n=2-4) (B) Illustration of base pairs mutated in the 3’UTR of GRK2 to generate Ad293- mutant GRK2 stable cell line (Ad293-mGRK2. (C) Quantification of RLU from Ad293- mGRK2 cells transfected with miR mimics normalized to miR-Neg (dotted line). (n=2-3). *p<0.05, **p<0.01 as determined by student’s t-test compared to miR-Neg.

81 mimics for 48 hours. We observed complete ablation of miR-181a-mediated inhibition of luciferase activity with the mutated GRK2 3’UTR sequence (Fig. 2-8C). Complete reversal of targeting suggests this site contains the only functional binding site for miR-181a.

Although miR-K12-mediated inhibition of luciferase activity was not completely reversed, it was blunted considerably: ~45% decline in Ad293-GRK2 vs ~22% decline in Ad293- mGRK2. MiR-K12 may be able to bind other sites within the GRK2 mRNA, which can be further investigated.

2.3.6 GRK2 expression in miR-181a/b KO mice

Due to our interest in miR-181a possibly regulating GRK2 expression, we were investigated whether miR-181a knock-out mice had increased GRK2 expression. Whole hearts from homozygous miR-181a knock-out (miR-181aKO) or WT mice were obtained from the Samarjit Das laboratory at John Hopkins . Hearts were analyzed for GRK2 protein and mRNA expression. We did not observe any changes in GRK2 mRNA levels at baseline

(Fig. 2-9B). GRK2 protein levels were trending towards an increased in the miR-181aKO mice compared to WT (Fig. 2-9A,C). The lack of drastic changes in GRK2 expression in miR-181aKO mice mirror what was observed with miR-181a/b inhibition in AC16s.

Changes in GRK2 protein may not have been significant due to antagonizing baseline levels of miR-181a/b, miR-based regulation not being a predominant GRK2 regulatory pathway, or compensatory inhibition through other miR-181 family members. Also, analysis of more mouse hearts could lead to statistical significance and an increase in myocardial GRK2 levels in these KO hearts.

FIGURE 2-9. GRK2 expression in homozygous miR-181a/b knock-out mice. (A) Representative Western blot images of GRK2 (~80 kDa) and GAPDH (~37 kDa) protein from WT or homozygous miR-181a/b knock-out mouse (181a/bKO) heart lysates. (B) Quantification of GRK2 mRNA by qRT-PCR or (C) GRK2 protein by Western blot from WT or 181a/bKO mouse heart lysates. (n=3,5).

2.3.7 miR-K12 reduced GRK2 expression in vivo

With a top miR candidate selected, miR-181a, we next wanted to explore miR delivery into mice through intramyocardial injections. In order to deliver a specific -5p or

-3p strand, we opted to use miR mimics. To validate whether miR mimics delivery was feasible, miR-K12 mimic was complexed with Invivofectamine3.0 or

LipofectamineRNAiMAX according to manufacturer’s recommendations. 20 µL of the miR:lipid mixture was injected into hearts of WT mice. This corresponded to a 1-1.6 mg/kg dosing depending on mouse BW. Hearts were collected 48 hours after injection to assess

GRK2 protein levels. In vivo delivery of miR-K12 mimics successfully reduced GRK2

83 protein levels after 48 hours when complexed with either Invivofectamine3.0 (Fig. 2-

10A,B) or LipofectamineRNAiMAX (Fig. 2-10C,D). GRK2 mRNA levels were unchanged with miR-K12 delivery, suggesting miR-K12 may rely on a more translation- inhibition mechanism of gene silencing in the heart or the involvement of other RNAi- related proteins in its mechanism in vivo.

FIGURE 2-10. Intramyocardial delivery of miR-K12 reduces GRK2 protein expression. Intramyocardial delivery of miR-Neg or miR-K12 mimics complexed with either (A,B) Invivofectamine3.0 transfection reagent at a mimic dose of ~1.6 mg/kg (n=3,2) or (C,D) LipofectamineRNAiMAX transfection reagent at a mimic dose of ~1.0 mg/kg after 48 hours. (n=3) (A,C) Representative Western blot images of GRK2 and GAPDH protein from LV heart lysates. (B,D) Quantification of Western blot data of GRK2 from mice injected with miR-Neg or miR-K12. *p<0.05 as determined by students t-test.

2.3.8 GRK2 overexpression changes miR expression profile in vivo

b-arr was reported to influence the processing of a subset of miRs subsequent to carvedilol stimulation (Kim et al. 2014). In order to determine if GRK2 also had any impact on miR biogenesis or processing, we isolated RNA from whole hearts from TgGRK2 mice.

These mice have cardiac-specific GRK2 overexpression 3-4-fold over endogenous levels

(Koch et al. 1995). Total RNA was sent to LCSciences for miR sequencing and bioinformatic analysis. MiR sequencing revealed that TgGRK2 had a different miR expression profile compared to NLC mice. The most differentially expressed miRs are represented in Fig. 2-11A. Gene ontology (GO) analysis of these differentially up- or - downregulated miRs showed enrichment in certain molecular pathways (Fig. 2-11B).

Interestingly, GO functional pathways enriched included signal transduction, GPCR signaling, and plasma membrane—all of which GRK2 has a role in.

2.4 Discussion

The major aim of this project was to identify a miR that could negatively regulate

GRK2 expression and was involved with the pathological upregulation of GRK2 during

HF. Ultimately, this miR could be harnessed as an anti-GRK2 therapy in the treatment of

HF and other CVDs where GRK2 is upregulated. Through the use of 2-week post-MI model and bioinformatic prediction, we identified a number of miR candidates to pursue.

Using in vitro based assays, we identified a top candidate to be applied to a future HF study.

We made use of a miR identified as part of the KSHV genome: miR-K12-3-5p (miR-K12).

MiR-K12 inibited GRK2 (Hu et al. 2015). Our top miR of interest is miR-181a-5p or

FIGURE 2-11. Differential miR expression in TgGRK2 mice. (A) Heatmap of differentially expressed miRs with p<0.05 and p<0.1. (B) Gene ontology enrichment analysis of molecular functions of differentially expressed miRs in (A).

86 commonly referred to as miR-181a. MiR-181a was able to bind the 3’UTR of GRK2 and decrease GRK2 protein. Inhibition of miR-181a at baseline increased GRK2 protein slightly and mRNA significantly. Its targeting of the 3’UTR was completely abolished with mutation of the 3’UTR target site. MiR-181a, as well as other family members, scored very high context scores (82-83%) for targeting GRK2 based on TargetScan7.0 prediction algorithms. In addition to seed sequence complementarity, TargetScan considers features surrounding the seed, context scores, that can improve predictions such as 3’UTR target- site abundance, predicted seed-pairing thermodynamic stability, and predicted structural accessibility (Agarwal et al. 2015; Grimson et al. 2007). Thus, these are important considerations when characterizing true miR targets. Indeed, miR-181a meets many of these criteria as evident of its high context score ie miR-181a is a conserved miR binding a conserved sequence region and contains a complimentary base pairing at position 8. After identifying miR-181a as our top candidate, we proceeded with target validation strategies.

MiR-181a belongs to the miR-181 family of miRs, consisting of 6 members: miR-

181a1, miR-181a2, miR-181b1, miR-181b2, miR-181c, and miR-181d. These miRs exist in clusters across 3 separate chromosome locations. The miR-181a1/b1 cluster is located on chromosome (Chr) 1 within the intron of a noncoding RNA host gene called MIR181a1-

HG in both humans and mice. The miR-181a2/b2 cluster is located on Chr9 in humans or

Chr2 in mice within an intron of the NR6A1 gene. The miR-181c/d cluster is located on

Chr19 in humans or Chr8 in mice and is within an uncharacterized sequence. These 6 family members result in the production of 4 mature miRs, miR-181a/b/c/d, that share a common seed sequence: ACAUUCA. Their genomic organization and sequence are

87 evolutionarily conserved across many species including humans, rodents, monkeys, frogs, and boney fish (Ji et al. 2009). High conservation and redundancy suggest their important role in development. Various forms of KO mice for miR-181 members have been generated: 181a1/b1KO, 181a2/b2KO, and 181c/dKO (Henao-Mejia et al. 2013). Within the normal mouse heart, miR-181a belongs to the group of most highly expressed miRs

(Cheng et al. 2007). It is suggested that the CV system predominantly expresses the miR-

181a1/b1 cluster, as miR-181a1/b1 ablation significantly decreases miR-181a/b expression in the vasculature (Hori et al. 2017). Analysis of RNA isolated from mitochondria from the heart revealed 2-fold enrichment for miR-181c compared to total heart (Das et al. 2012).

Although import of nuclear DNA-encoded RNA into the mitochondria is not unprecedented, the mechanism is not well studied and the concept controversial but seem to require some protein factor for targeting (Jeandard et al. 2019; Magalhães et al. 1998).

Analysis of expression of miR-181a/b/d show they also are found in the mitochondrial fraction, but miR-181a and miR-181d to a lesser extent (Das et al. 2017). MiR-181a and miR-181b have been identified to target phosphatase and tensin homolog (PTEN), a cytosolic protein. 181a1/b1KO and 181c/dKO mice have differential responses to I/R, further enforcing their differential targets and effects. 181c/dKO mice were protected from

I/R injury with reduced infarct size, ROS generation, mitochondrial swelling (Das et al.

2017). On the other hand, 181a1/b1KO mice did not receive the same cardioprotection after I/R injury. They displayed significantly increased cardiomyocyte death, infarct size, higher trend of mitochondrial swelling, and similar ROS production compared to WT mice

(Das et al. 2017). In addition, loss of miR-181a1/b1 has significantly negative impacts on

88 cardiomyocyte contractility by reducing sarcomere shortening and increased time to relaxation. Delivery of miR-181a during I/R surgery was protective in rats (Zhang et al.

2019a). Increased I/R injury in 181a1/b1KO animals parallels increased I/R injury observed in TgGRK2 animals, providing further circumstantial evidence to support our hypothesis (Brinks et al. 2010). If we suspect our GRK2 targeting miR to be downregulated during disease to contribute towards GRK2 upregulation, then the loss of that miR during disease should exacerbate injury. MiR-181a is highly expressed in the heart and further enriched in the cytoplasm compared to mitochondrial enrichment of its other family members. Thus, miR-181a’s subcellular localization strengthens its candidacy for targeting

GRK2. Additionally, it is interesting that loss of miR-181a1/b1 has a negative effect on contractility in mice by reduced ejection fraction and fractional shortening, which parallels

GRK2’s negative effect on contractility (Koch et al. 1995). This could theoretically be from increased GRK2 expression due to loss of miR-181a. Henao-Mejia et al originally generated the global homozygous miR-181a/b KO mouse (181a/bKO) to investigate the requirement of miR-181 for Natural Killer T cells (NK) development and homeostasis.

Deletion of miR-181a1/b1 perturbed thymocyte development, reduced glycolysis, and reduced NK cell number and development through regulating phosphoinositide 3 kinase

(PI3K) signaling via PTEN inhibition. Das et al. showed that miR-181a/b targeted PTEN in cardiomyocytes as well (Das et al. 2017). MiR-181a/b targeting in PTEN was the mechanism of increased I/R injury in 181a/bKO mice. PTEN dephosphorylates phosphatidylinositol-3,4,5-triphosphate (PIP3), which is an important phospholipid involved in intracellular signaling. Thus, PTEN reverses phosphoinositide 3-kinase (PI3K)

89 activity. PIP3’s most notable target is Akt. Therefore, loss of miR-181a/b relieves PTEN silencing to cause Akt inhibition. 181a/bKO mice were characterized by increased PTEN expression and decreased pAkt (Ser 473) expression that resulted in reduced cardioprotection by the PI3K/Akt pathway after I/R (Das et al. 2017; Murphy and

Steenbergen 2007). Coincidentally, PTEN and GRK2 have the same effect on Akt activation and activity. Similarly, increased GRK2 is able to directly bind both Akt and pAkt to inhibit Akt activity (Liu et al. 2005). GRK2 expression levels were not examined in 181a1/b1KO mice after I/R, but we showed at baseline there was a trending increase in

GRK2 protein. It would be interesting to see if 181a1/b1KO mice after I/R had increased

GRK2 expression or activity compared to WT. As discussed, it is a known pattern for a single miR to target multiple components within a critical pathway to increase redundancy and regulation as in the case of miR-133a and b-AR signaling. If GRK2 is indeed a target of miR-181a, this could follow that pattern of multi-targeting within Akt signaling where miR-181a can increase Akt activity through inhibiting PTEN and GRK2. This pathway may be particularly relevant in cases of endothelial dysfunction and further research is required to confirm this relationship.

In order to further validate GRK2 as a target of miR-181a, a number of future studies would need to be done. Our study includes the traditionally and commonly used miR target validation techniques. Each technique has its own caveats and pitfalls, therefore employing multiple would be the most rigorous. With overexpression of miRs, like through mimics, it is possible to saturate available RISC or force target mRNA binding that would otherwise be low-affinity (Thomson et al. 2011). Additionally, mimics do not require any

90 processing as they are readily available to be incorporated into RISC. This would bypass any alterations in miR processing or regulation that might be relevant in HF. Indeed, it is understood these processes are perturbed during disease in order to drastically change miR expression profiles. Therefore, the use of inhibitors alongside mimics would avoid issues with off-target effects and improves rigor. We utilized both mimics and inhibitors of miR-

K12 and miR-181a to show that mimicry reduced GRK2 expression and inhibition enhanced GRK2 expression to an extent. Together, these data strengthen GRK2 as a target of miR-181a. In the case of miR-181a and many other miR families, it shares the same seed sequence with its other family members, which can make quantifying each specific member challenging. For example, primers that detect miR-181a were also found to detect miR-181c due to their highly similar sequences (Das et al. 2017). Thus, in our miR inhibitor experiments, we cannot rule out that miR-181a inhibitors may also bind to other members.

Additionally, it is possible there may be degree of compensation between the miRs derived from the miR-181a/b1 and miR-181a/b2 clusters that may have prevented more robust

GRK2 upregulation with miR-181a1 inhibition. Although compensation has not been directly assessed, a single deletion of amiR-181a/b cluster is viable whereas deletion of both clusters affects viability (Henao-Mejia et al. 2013).

Another method to validate targets is immunoprecipitation of RISC components and subsequent detection of miR or mRNAs associated, which is how mt-CO1 was identified as a target for miR-181c (Das et al. 2012). Immunoprecipitation (IP) of AGO has been done on small and large scales in cells or tissues, although it is a technically challenging process. Spengler et al. performed the first high-throughput sequencing of

RNA isolated by crosslinking immunoprecipitation (HITS-CLIP) study using heart biopsies from HF patients to map transcriptome-wide AGO2:RNA binding interactions

(Spengler et al. 2016). HITS-CLIP provides a powerful comprehensive ‘omics’ view of miR targeting events in the heart and can be used for further target validation. However,

GRK2 was not included in the published data set, but the study still provided insights into general miR targeting and miR translational research. Human AGO2 cluster location within transcripts was largely within the coding sequence (CDS, 49%) or 3’UTR (38%)

(Spengler et al. 2016). We have currently only examined the GRK2 3’UTR in our luciferase reporter studies but examining targeting within the CDS may be worthwhile. The

5’UTR may also be examined, since it also contains regions of conservation. Within the

3’UTR, AGO2 clustering seems to occur at the very 5’ end or very 3’ end of the 3’UTR, which agrees with the Bartel group’s previous study showing increased endogenous targeting efficiency as distance from the nearest 3’UTR terminus decreases (Grimson et al.

2007). The miR-K12 and predicted miR-181a binding sites on GRK2’s 3’UTR exist extremely close to the 3’ terminal end, about 56-57 nt away, which agrees with Spengler et al and Grimson et al findings (Grimson et al. 2007; Spengler et al. 2016). Coincidentally, the 3’ terminal end is also the region that we found to be conserved between humans and mice. A potential pitfall in RISC component IPs is that it biases towards strong and stable interactions between a miR, mRNA, and AGO and thus weaker valid interactions may be lost during the crosslinking process. The target site for miR-K12 and predicted site for miR-181a exists as a single site on the 3’UTR, in contrast to other 3’UTRs that exist clustered target sites for a single miR. Clustering of target sites improves miR efficiency.

Therefore, in the case of miR-181a-GRK2-AGO2 interaction, this may not be a strong interaction which may further explain limited GRK2 mRNA degradation. Investigation of miR targets can also be done through proteomic-based approaches, despite the consensus that 60-90% of miR-mediated repression is from mRNA destabilization. Translational profiling such as polysome profiling or ribosome profiling have been used to investigate what transcripts are associated with ribosomes and undergoing translation. Employing these methods during miR-181a overexpression or knockdown could give us insights on if miR-181a specifically has any effect on GRK2 translation that contributed to its protein silencing especially since we did not observe decreased GRK2 mRNA with miR-181a overexpression, but we did observe increased GRK2 mRNA with miR-181a inhibition.

MiR-181a silencing of SIRT1 was also solely observed at the protein level, while silencing of Renin was identified at the mRNA level (Marques et al. 2011; Zhou et al. 2012). These data solidify there can be multiple methods of gene silencing by miR-181a depending on the target mRNA and cell type, further confirming GRK2 translational inhibition by miR-

181a can be a direct effect. Using polysome or ribosome profiling would give a more detailed insight on how translational inhibition is occurring.

We initially chose a 2-week post-MI model of HF to observe miR expression changes. However, the type of HF model utilized will, no doubt, impact miR signatures and may influence how we implement miR-targeting GRK2 therapy. We observed all miR-

181 family members to decrease in expression. However, whether or not miR-181a is upregulated or downregulated across different CVD depend on the model used and what tissues are investigated. Based on plasma samples from congestive heart failure due to non-

93 ischemic cardiomyopathy patients, miR-181a and miR-181b were found to be significantly upregulated compared to healthy patient samples (Marques et al. 2016). High expression of circulating miR-181a strongly correlated with hypertension (Marques et al. 2016). In contrast, miR-181a within the kidney was reduced during hypertension and contributed to hypertension development due reduced miR-181a-dependent renin targeting. Renin, part of the renin-angiotensin-aldosterone system (RAAS), represents the first enzymatic step in generating angiotensin II to promote vasoconstriction. Both miR-181a and miR-181b expression were reduced in the aortic intima of the ApoE-/- mice fed with HFD and both can directly contribute to the development of atherosclerosis and associated endothelial dysfunction (Liu et al. 2016; Sun et al. 2014). Reduction in miR-181a expression was also observed in a rat model of diabetic cardiomyopathy using HFD and streptozotocin (STZ)

(Raut et al. 2016). Multiple studies using rat hypertrophy models have reported a decrease in miR-181a expression after TAC. Feng et al. examined miR expression after at various time points after TAC and found miR-181a to be downregulated by day 5, but expression may begin to normalize by day 15-20 (Feng et al. 2014). MiR-1 and miR-133 were reduced in parallel, as expected. Hypertrophy from 4 week transverse abdominal aortic constriction

(TAAC) of rats also was accompanied by reduced miR-181a expression in the heart (Li et al. 2017). In these CVD and CVD-related models (hypertension, atherosclerosis, diabetes) discussed where miR-181a was found to be reduced, GRK2 upregulation is involved in their pathogenesis (Izzo et al. 2008; Otten et al. 2013; Taguchi et al. 2011; Taguchi et al.

2012). As we are concerned with mainly cardiomyocyte GRK2 expression, the use of a pressure-overload model of HF in mice may be the best model to employ miR-181a

94 replacement therapy as it has been more consistently reported to be reduced in the heart.

However, ischemic models of HF seem to more likely report increased miR-181a expression despite our 2-week MI model showing it to be reduced. This discrepancy could very well be due to timing of miRNA profiling, as miR-181a levels have been reported to change during disease progression (Feng et al. 2014). Human LV tissue taken from ischemic cardiomyopathy, dilated cardiomyopathy, or aortic stenosis patients during transplant had increased miR-181a expression (Ikeda et al. 2007). Ikeda et al did not assess mRNA expression, but we can suspect GRK2 is increased due to extensive reports in the literature from other studies. In CVD, it is difficult to provide a universal statement on what occurs with miR-181a expression given heterogeneity of disease onset, time of observation, and disease development between studies. It is also important to consider that individual changes in miR-181a expression in a given study do not affect physiological processes in isolation, as miR-181a itself and its downstream targets are subject to global regulatory processes that consist of integrative cross-talk among numerous networks to ultimately determine disease phenotype (Naga Prasad et al. 2017).

Whether or not GRK2 itself has roles in influencing miR expression or regulation is completely unknown. Given that GRK2 interacts with and phosphorylates a growing number of substrates certainly opens up the possibility for GRK2 to have some role, even if indirectly. Interestingly, b-arr has been shown to regulate the biogenesis and processing of a small subset of miRs. Kim et al. observed that carvedilol selectively induced the expression of a small subset of miRs (miR-125a-5p, miR-125b-5p, miR-150, miR-199a-

3p, miR-214, and miR-190) in HEK293 cells that was not induced with metoprolol (Kim

95 et al. 2014). Given that carvedilol elicits a b-arr-bias in contrast to metoprolol, b-arr likely played a direct role in carvedilol-induced miR-190 expression. Indeed, knockdown of b- arr1 prevented the carvedilol-mediated miR-190 induction as did GRK5 and GRK6 siRNA-mediated inhibition, suggesting these specific GRK phosphorylation events on b1-

AR play a role in carvedilol bias. b-arr1 was able to translocate to the nucleus upon carvedilol stimulation to interact with hnRNAP1 and Drosha to increase the processing of a small subset of miRs. This was evident by an increase in their pre-miR and mature miR forms but not pri-miR (Kim et al. 2014). It is certainly interesting that b1-AR phosphorylation by GRK5 or GRK6, but not GRK2 or GRK3, was absolutely required for b-arr1-mediated miR processing. Whether or not a similar function for GRK2 or GRK3 exists is open to investigation and is a further direction. Our miR sequencing data between

TgGRK2 and NLC mice indeed show a number of differentially expressed miRs (Fig. 2-

10). MiRs of interest include miR-27a-3p that was downregulated and miR-30a that was upregulated in TgGRK2 hearts. Both miRs are well expressed in the heart and have been implicated in heart failure. Loss of miR-27a has been reported to induce cardiac hypertrophy and dysfunction likely through promoting FoxO1-dependent transcription

(Qin and Liu 2019). It would be interesting to determine whether GRK2 could influence miR-27a expression and whether this contributes to worsened prognosis of TgGRK2 mice after HF induction.

In conclusion, we have identified miR-181a to be able to regulate GRK2 expression due to our observation its of direct binding to the GRK2 3’UTR to elicit significantly reduced GRK2 protein expression. Further, inhibition of miR-181a was able to increase

GRK2 mRNA and to a lesser extent protein expression. We understand that miR-181a is likely not the sole determinate of GRK2 expression and it is possible that miR-181a may only be a “fine-tuner” (Mukherji et al. 2011). As post-transcriptional regulation of GRK2 is largely unknown, the identification of miR-181a indeed provides insight on how GRK2 is upregulated during heart failure. Further studies with miR-181a administration and subsequent examination of GRK2 expression during HF will need to be carried out to further characterize this interaction.

CHAPTER 3

LOSS OF DYNAMIC REGULATION OF GRK2 BY NITRIC OXIDE AND

CARDIOVASCULAR AGING

3.1 Introduction

Nitric monoxide or nitric oxide (NO) is a free radical and gaseous molecule. As small of a molecule as it is, NO has received much investigation in the CV field since its discovery as a signaling molecule in the CV system in 1977 by Ferid Murad (Arnold et al.

1977). Arnold et al observed that nitric oxide gas and other nitrogen containing compounds were able to activate soluble guanylate cyclase to increase intracellular cyclic guanosine monophosphate (cGMP) concentrations (Arnold et al. 1977). Furchgott et al discovered that acetylcholine (ACh)-dependent vascular smooth muscle relaxation required an endothelial-derived factor, which was later identified to be NO 7 years later (Furchgott and

Zawadzki 1980; Ignarro et al. 1987). NO is generated by enzymes called nitric oxide synthases (NOSs), of which there are 3 mammalian isoforms: neuronal NOS

(nNOS)/NOS1, inducible NOS (iNOS)/NOS2, and endothelial NOS (eNOS)/NOS3. NOS catalyzes the conversion of L-arginine to L-citrulline and forms NO in the process. This reaction requires several cofactors, identified as tetrahydrobiopterin (BH4), flavin adenine dinucleotide (FAD), flavin mononucleotide (FMN), Ca2+-calmodulin, and NAPDH. A heme group, BH4, and L-arginine bind to the N-terminal oxygenase domain to help stabilize eNOS. FAD, FMN, and NADPH bind to the C-terminal reductase domain

(Massion and Balligand 2003). The two domains are linked by a calmodulin binding site.

The reductase domain is the site of Ca2+-dependent NAPDH electron transfer to

FAD/FMN, then to heme, lastly to L-arginine (Massion and Balligand 2003).

Cardiomyocytes basally express NOS3 and NOS1 that are localized to distinct subcellular compartments. These spatial confinements deliberately confer target protein specificity.

For example, localization of NOS3 to plasmalemmal caveolae and NOS1 to sarcoplasmic reticulum (SR) allow their NO to target plasmalemmal L-type Ca2+ channel/b-AR and SR

Ca2+ ATPase (SERCA2a), respectively (Lima et al. 2010). NOS3’s localization to caveolae allows it to interact with other receptors or signaling molecules. The most well-known post- translational modification of NOS3 is its phosphorylation on Ser1177 to cause increased activity. Ser1179 has also been identified as a site of phosphorylation dependent NOS3 activation. Ser1177 and Ser 1179 are subject to phosphorylation by Akt, PKA, PKG, and

AMPK. Their phosphorylation can be induced with insulin or stretch stimuli (Fulton et al.

1999; Petroff et al. 2001).

In the myocardium, NO has complex modulatory effects on contractile function that are not quite straightforward due to its versatile biochemistry and diversity of its intracellular targets (Massion et al. 2003). Brunner et al. describes NO as a “dual endogenous regulator of myocardial contractility that tonically supports pump function at physiological concentrations, but unloads the heart when generated in great amounts by lowering the calcium sensitivity of the contractile elements” (Brunner et al. 2001).

Generally, at low concentrations NO exhibits positive inotropic effects, but becomes negatively inotropic at high concentrations. This bimodal phenomena has been mainly demonstrated in in vitro systems with exogenous sources of NO (Massion and Balligand

2003). Genetic manipulation of NOS in mice has given more physiologically relevant insight on the impact of NO on cardiac function. Mice deficient in global NOS1 (NOS1-/-

) do not have compensatory changes in NOS3 expression, nor any changes in basal contractility, chamber dimensions, or systolic blood pressure compared to control hearts

(Barouch et al. 2002). NOS3 global deletion in mice (NOS3-/-) also display similar basal contractility compared to WT mice, but are hypertensive as measured by higher systolic blood pressure and mean arterial blood pressure (Barouch et al. 2002; Gyurko et al. 2000;

Huang et al. 1995). Deletion of NOS3 reduces total NOS activity in the heart, lung, and aorta to virtually none ie. 3.7±0.1 pg ⋅ mg−1 ⋅ ml−1 in control mice compared to 0.6±0.1 pg

⋅ mg−1 ⋅ ml−1 in NOS3-/- hearts, signifying that NOS3 is the major source of NOS activity and NO production in these tissues (Gyurko et al. 2000). However, in response to b-AR stimulation, NOS1-/- and NOS3-/- mice display changes in contractility. NOS1-/- had impaired contractility in response to ISO, especially at higher doses (Barouch et al. 2002).

NOS3-/- hearts had similar dP/dtmax and dP/dtmin values in response to ISO doses less than

1 ng, but with ISO doses more than 1 ng NOS3-/- hearts showed enhanced contractility and relaxation that was independent of b-AR density alterations (Gyurko et al. 2000).

Contractility of control mice treated with a NOS inhibitor during ISO infusion was indistinguishable from NOS3-/- treated with ISO alone. Overall, these studies suggest that in the presence of high concentrations of b-AR agonist, NO from predominantly NOS3 antagonizes b-AR-mediated inotropy, chronotropy, and lusitropy. It is evident that NOS1 and NOS3 have divergent effects on cardiac function. Contractility in mice with cardiac- specific overexpression of NOS3 (TgNOS3) has also been examined. Overexpressed

100

NOS3 maintains its plasmalemmal localization (Brunner et al. 2001; Janssens et al. 2004).

TgNOS3 mice have been generated by multiple groups. Brunner et al.’s TgNOS3 mice generated less LV pressure compared to control mice, which was normalized with a NOS inhibitor (Brunner et al. 2001). Reductions in contractility were not due to cardiomyocyte damage nor ROS production, but was a result of reduced myofilament responsiveness to

Ca2+ (Brunner et al. 2001). However, TgNOS3 mice used in the studies by Elrod et al. and

Janssens et al. did not display any changes in basal contractility or hemodynamics (Elrod et al. 2006; Janssens et al. 2004). It is notable that assessment of contractility were measured by different parameters and methods in these studies: Brunner et al. assessed pressure development in isolated perfused hearts, while Elrod et al. and Janssens et al. assessed more parameters in the intact mouse (Elrod et al. 2006; Janssens et al. 2004).

NOS3 does not seem to play a significant role in basal relaxation, likely due to other compensatory pathways like ANP upregulation (Gyurko et al. 2000).

It is well appreciated that NO has many beneficial effects in the CV system, and this becomes significant when considering CVD therapies. In fact, its initial discovery of its vasodilatory properties is itself cardio- and vaso-protective, as vascular relaxation improves blood flow and oxygenation. NO’s vasodilatory properties are the basis for nitrates used as treatment for angina, HF, and ischemic heart disease. Additional cardio- and vaso- protective actions of NO include being: (1) anti-thrombotic by inhibiting platelet adhesion, (2) anti-inflammatory by attenuating leukocyte-endothelial adhesion and scavenging ROS, (3) anti-proliferative by inhibiting VSMC hyperplasia, (4) anti- desensitizing by antagonizing b-AR signaling, and (5) anti-apoptotic in cardiomyocytes

101

(Naseem 2005). Indeed, administration of NO is sufficient in limiting myocardial ischemia/reperfusion injury in animal models and mortality in humans (Johnson et al. 1990;

Lefer et al. 1993; Taylor 2005). TgNOS3 that were subjected to ischemia/reperfusion (I/R) injury had significantly preserved cardiac function, decreased area at risk, and decreased infarct size compared to control mice due to increases in NO bioavailability (Elrod et al.

2006). TgNOS3 also had preserved LV dimensions after I/R. TgNOS3 mice have also been shown to be significantly protected after MI injury compared to controls, as shown through limiting LV dysfunction and hypertrophy development (Janssens et al. 2004). Although high levels of NO from NOS3 cardiac-overexpression have shown to be beneficial, high levels of NO from NOS2 cardiac-overexpression exaggerated cardiac injury due to NO reacting with oxidants to produce peroxynitrite (Mungrue et al. 2002). Being a free radical itself, high levels of NO share the same cytotoxic characteristics as classical ROS.

NO is capable of acting as a biological signaling molecule to regulate a wide range of cellular functions such as contractility, apoptosis, metabolism, enzyme activity, transcription factor stability, subcellular localization, and cellular redox state. It is necessary for cells to maintain NO and redox homeostasis for normal physiology and NO modulation has been implicated in the pathophysiology of many diseases. One mechanism in which NO can regulate cellular processes is through the post-translational modification of proteins, called S-nitrosylation, to form S-nitroso proteins (SNO-proteins) (Fig. 3-1)

(Stamler et al. 1992). Indeed, NO’s ability to modulate cardiac contractility as described involves S-nitrosylation of key proteins. S-nitrosylation of proteins allows for NO to be significantly more stable and mobile compared to free NO. Increased NO stability is

102

FIGURE 3-1. S-nitrosylation and transnitrosylation NO-based reactions (Lima et al. 2010). (A) Nitric oxide (NO) synthase (NOS) synthesizes NO that can either activate soluble guanylyl cyclase (sGC) to generate cyclic GMP (cGMP) or participate in protein S-nitrosylation of cysteine thiols. cGMP levels activate phosphodiesterase 5 (PDE), which is responsible for degrading cGMP. Cysteine thiols on proteins can be dynamically nitrosylated (Protein-Cys-SNO) and denitrosylated (Protein-Cys-SH). (B) Recycling of NO groups on proteins is mediated by two main pathways of denitrosylation: S- nitrosoglutathione (GSNO) reductase (GSNOR) or thioredoxin (Trx). Protein-Cys-SNO can transnitrosylate glutathione (GSH) to become S-nitrosoglutathione (GSNO). GSNO can be reduced by GSNO reductase (GSNOR) in a NADH-dependent manner. The generated glutathione S-hydroxysulfenamide (GSNHOH). Oxidized glutathione (GSSG) can be converted back to GSH via GSSG reductase (GR) in a NADPH-dependent manner. Denitrosylation can also be carried out by active site dithiol motif (CXXC) of Trx to reduce Protein-Cys-SNO. Oxidized Trx can be reduced back to Trx through Trx reductase, allowing Trx to be an active denitrosylase again.

103 important for NO to serve as cellular signaling intermediates. S-nitrosylation specifically refers to the covalent modification of thiol side groups (-SH) on cysteine residues or transition metal centers by NO. Typically, only one specific cysteine undergoes physiological modification and its likely within an acid-base or hydrophobic motif (Seth and Stamler 2011). So far, hundreds of proteins have been identified to be regulated this way (Stamler et al. 2001). Table 3-1 highlights S-nitrosylated proteins involved in b-AR regulation. Endogenous sources of NO for S-nitrosylation can be NO itself, other nitrogen molecules, SNO-proteins themselves (via trans-S-nitrosylation/transnitrosylation), or small-molecular-weight S-nitrosothiol ie S-nitrosoglutathione (GSNO). The addition of exogenous NO containing compounds is also sufficient to promote S-nitrosylation such as nitroprusside. Hemoglobin is an unique example that undergoes auto-S-nitrosylation, transferring NO from its heme center to Cys93 (Luchsinger et al. 2003).

In 2007, GRK2 kinase activity was identified to be dynamically decreased by S- nitrosylation after observations that NO could specifically modulate GRK2-dependent processes (Whalen et al. 2007). Whalen et al. observed that non-specific NOS inhibition through L-NG-nitroarginine methyl ester (L-NAME accelerated the decline in contractility during ISO infusion in WT mice, reaching tachyphylaxis much faster than PBS infusion while basal contractility was unchanged (Whalen et al. 2007). When mice were continuously infused with ISO for 7 days, b-AR density was reduced. Infusion of GSNO alone or co-infusion of GSNO with ISO prevented b-AR decline, suggesting that b-AR signaling itself can be regulated by NO signaling and further supporting a cardioprotective role for NO. ISO treatment also enhanced b-arrestin2 (b-arr2) recruitment to b2-AR, which

104

Table 3-1. S-nitrosylated substrates involved with b-AR regulation. NO Substrate Source Cys Functional impact Reference GRK2 NOS3 C340 • Inhibits b-AR desensitization (Whalen et • Inhibits barr membrane al. 2007) recruitment • Inhibits b-AR internalization • Inhibits b-AR downregulation b-arrestin2 NOS3 C410 • Promotes association with (Ozawa et clathrin al. 2008) • Promotes b-AR internalization • Promotes b-AR downregulation NOS1/2 C253 • Inhibits barr membrane (Hayashi et recruitment al. 2018) • Inhibits b-AR internalization • Inhibits b-AR downregulation • Promotes G protein biased signaling b-arrestin1 NOS1/2 C251 • Inhibits barr membrane (Hayashi et recruitment al. 2018) • Inhibits b-AR internalization • Inhibits b-AR downregulation • Promotes G protein biased signaling Dynamin NO3 C607 • Promotes scission of endocytic (Wang et al. vesicles 2006) • Promotes b-AR internalization Cys: Cysteine; b-AR: b adrenergic receptor; b-arr: b-arrestin; GRK2: G protein- coupled receptor kinase 2; NOS1: neuronal nitric oxide synthase; NOS2: inducible nitric oxide synthase; NOS3: endothelial nitric oxide synthase.

105 coincides with increased desensitization or tachyphylaxis, in HEK293 and U2-OS cells

(Whalen et al. 2007). b-arr2 recruitment was reversed with S-nitrosocysteine (CysNO, exogenous NO) delivery. Reduced phosphorylated b2-AR was also observed with CysNO treatment in cells, even when PKA sites were mutated. As b2-AR phosphorylation, desensitization, downregulation, and b-arr2 recruitment involve GRK2, Whalen et al. investigated whether GRK2 was itself modulated by NO through S-nitrosylation (Whalen et al. 2007). Indeed, NO was able to S-nitrosylate GRK at Cys340 with a ~1:1 stoichiometry. SNO-GRK2 levels are detected at basal levels and increased with GPCR stimulation, likely coordinated by NOS3’s proximal localization. SNO-GRK2 is denitrosylated by at least the GSNO reductase (GSNOR) pathway; the involvement of thioredoxin denitrosylase pathway has not been investigated. Cys340 was identified to be the main Cys residue dynamically regulated by NO. Cys340 resides in GRK2’s catalytic domain and conforms to the typical acid-base motif. Therefore, NO adducts likely perturb

GRK2’s active conformation to inhibit its activity. This residue is conserved in GRK3, but not any other GRKs. Mutation of Cys340 to Ser340 in vitro prevented CysNO-dependent inhibition of b2-AR phosphorylation. Overall, S-nitrosylation at Cys340 is a mechanism to prevent receptor desensitization and downregulation.

The predominant source of NO that nitrosylates GRK2 to inhibit its kinase activity seems to be from NOS3. In other words, protection after I/R injury in TgNOS3 mice previously described may be, in part, mediated through inhibition of GRK2 through S- nitrosylation. NOS3-/- mice have significantly decreased SNO-GRK2 levels as a result of decreased NO bioavailability (Whalen et al. 2007). Based on SNO-GRK2 quantification in

106

WT and NOS3-/- mice, NOS3 accounts for >75% of the NO on SNO-GRK2. Additionally, mice with double cardiac-specific overexpressing transgenic GRK2 (TgGRK2) and NOS3 expression show protection after I/R compared to TgGRK2 alone (Huang et al. 2013).

TgGRK2 mice without NOS3 do not show any post-MI protection. Additionally,

TgbARKct mice (cardiac-expression of a GRK2 inhibitory peptide, previously shown to be cardioprotective after I/R) that do not express NOS3, lose their post I/R protection

(Brinks et al. 2010). Therefore, the adverse effects mediated by GRK2 are exacerbated with deficiency of NOS3 and NO as shown in the TgGRK2xNOS3-/-. Studies in the TgbARKct mice suggested that the beneficial effects of GRK2 inhibition are partially mediated by

NOS3 and loss of NOS3 blunts bARKct protection. These studies show that GRK2 and

NOS3 can modulate each other’s activity. Indeed, Huang et al. confirmed that GRK2 forms a complex with NOS3 in cardiac lysates. Akt is likely to be mediating this interaction was it was previously shown in endothelial cells that GRK2 can directly bind to both Akt and pAkt to inhibit Akt activity and thus downstream NOS3 activation (Liu et al. 2005).

GRK2’s association with NOS3 increased after I/R, suggesting that in WT mice GRK2 is able to negatively modulate NOS3 to decrease NO production. Inhibition of NOS3 by

GRK2 can be rescued with NOS3 overexpression. Thus, there exists a bi-directional and dynamic relationship between NOS3 and GRK2 where NO from NOS3 can inhibit GRK2 activity through S-nitrosylation and GRK2 can indirectly inhibit NOS3 activation.

In order to determine the physiological relevance of GRK2 S-nitrosylation in vivo,

Huang et al. generated a global knock-in (KI) mutant mouse (GRK2-C340S) that contains a point-mutation in the codon for Cys340, changing UGU à UCU to code for Ser (Huang

107 et al. 2013). Mutant GRK2 was expressed at endogenous levels with no compensatory changes in b-AR, other GRKs, or AC5/6 expression. As expected, GRK2-C340S mice are resistant to GRK2 S-nitrosylation with GSNO infusion and resistant to ISO-mediated changes in desensitization with L-NAME pretreatment. 8-10-week-old GRK2-C340S mice appeared undistinguishable from WT mice in appearance and cardiac function. However, when subjected to I/R injury GRK2-C340S mice had larger infarct areas compared to WT mice, despite having similar areas at risk (Huang et al. 2013). Pretreatment with GSNO was unable to rescue infarct size in GRK2-C340S as it did in WT mice. Pretreatment with

L-NAME did not worsen infarct size in GRK2-C340S mice as it did in WT mice (Huang et al. 2013). Larger infarcts were due to increased myocardial apoptosis around the border zone in GRK2-C340S hearts, consistent with GRK2 being a pro-death kinase (Chen et al.

2013). Overall, we now have an understanding that during b-AR stimulation, NOS3 activation parallels Gbg-dependent GRK2 recruitment to the membrane. While GRK2 promotes desensitization and receptor internalization, NOS3 via S-nitrosylation of GRK2 promotes sensitization. This balance between desensitization and sensitization are altered in CVD due to pathological increases in GRK2 expression and activity and a concomitant loss in NO bioavailability that tip the scales towards receptor desensitization. These discoveries give insight to the success of b-blockers in HF treatment.

Although the young 8-10-week-old GRK2-C340S mice had no obvious basal CV phenotype compared to WT control mice, if allowed to age with chronic GRK2 overactivity GRK2-C340S mice would likely develop cardiovascular dysfunction and remodeling. Given that age is the biggest risk factor in CVD development and what little

108 is known about GRK2 in aging, we allowed the GRK2-C340S mice to age at least 52 weeks. This would give us insights on the consequences of chronic GRK2 overactivity observed in HF and hypertensive patients as well as give us insight on GRK2’s role in aging. We hypothesized that GRK2-C340S mice that have chronic un-checked GRK2 activity globally may develop CV remodeling and dysfunction at baseline.

3.2 Experimental procedures

3.2.1 Experimental animals

All animal procedures were carried out according to National Institutes of Health

Guide for the Care and Use of Laboratory Animals and approved by the Animal Care and

Use Committee of Temple University. GRK2-C340S mice expressing endogenous levels of GRK2 with a KI mutation Cys340→Ser were generated as previously described (Huang et al. 2013). The GRK2-C340S line was maintained on a C57BL6/J background in which

WT controls are mice that do not contain the KI mutation from the GRK2-C340S line. The overall age range of old WT and GRK2-C340S mice used in this manuscript were 52-82 weeks. Old WT and GRK2-C340S mouse groups within each individual experiment were age matched to highlight phenotypic changes based on genotype. Young mice used were between 8-15 weeks of age. Both males and females were used for experiments and sex matched.

Cardiac-specific overexpression of GRK2 (TgGRK2) mice were generated as previously described (Koch et al. 1995). Briefly, the alpha myosin heavy chain (aMHC) promoter was ligated to the entire coding sequence of GRK2 to cause cardiac specific

109 overexpression of GRK2 in the adult atria and ventricles. TgGRK2 mice were compared with non-transgenic littermate controls (NLCs). Male and female mice were used. Young mice were aged 13 weeks and old mice were aged 73 weeks.

3.2.2 Echocardiography

All echocardiography was performed using the Vevo 2100 imaging system from

VisualSonics as previously described (de Lucia et al. 2019; Schumacher et al. 2016). Hair from the mouse chest was removed with Nair >24 hours before echocardiography. Mice were anesthetized with 3% isoflurane and maintained at 1-3% during procedures.

Traditional echocardiography (TE) utilized the MS400 (30-MHz centerline frequency) probe to obtain B- and M-mode images from parasternal long-axis and short-axis to evaluate LV internal diameters (LVID), fractional shortening (FS), end diastolic volume

(V;d), end systolic volume (V;s), stroke volume (SV), cardiac output (CO), and heart rate

(HR).

LV deformation was evaluated using echocardiographic speckle tracking-based strain echocardiography (STE) images from parasternal long-axis B mode loops consisting of 300 frames at ≥200 frame/second. Images were analyzed using the Vevo Strain

Software Vevo LAB 1.7.1. Strain, which is a measure of the change in length, was calculated from radial axis and the longitudinal axis. Strain rate, which is a measure of the change in length over time, was calculated from both axis as well. Strain or strain rate may be presented regionally (basal, mid, and apical anterior; basal, mid, and apical posterior) or globally across the whole LV. Reverse longitudinal and radial strain rate measured

110 longitudinal and radial strain during early LV filling (diastole), which was obtained using the “reverse peak” algorithm from Vevo LAB 1.7.1. Strain.

Diastolic function was evaluated from 2D echo in combination with tissue Doppler and pulsed wave Doppler from apical long-axis views. Using pulsed wave Doppler spectral waveforms, peak early (E wave) and late diastolic (A wave) transmittal velocities were obtained to calculate E/A ratio and isovolumic relaxation time (IVRT). Using tissue

Doppler spectral waveforms, early-diastolic myocardial relaxation velocity was obtained to calculate E’/E ratio.

3.2.3 RNA isolation and semi-quantitative PCR

LV tissue was homogenized in TRIzol™ Reagent using the VCX 130PB ultrasonic processor from Sonics & Materials Inc. Solutions were centrifuged to pellet cell debris.

Total RNA was isolated using Trizole and the Direct-zol RNA Miniprep Kit (Zymo

Research) according to manufacturer’s instructions. RNA was quantified using the

NanoDrop™ 2000/2000c Spectrophotometer (ThermoFisher). cDNA was generated from isolated RNA using the iScript cDNA Synthesis Kit (Bio-Rad). Semi-quantitative PCR on cDNA was done using SYBR Green (Bio-Rad) and a final concentration of 100 nM of gene-specific oligonucleotides for atrial natriuretic factor (Nppa, ANF), brain natriuretic peptide (Nppb, BNP), GRK2, and Rn18s (18S) on a CFX96 Real Time System with

BioRad CFX Manager 2.1 software. (Bio-Rad). Quantification was through 18S rRNA normalization and compared using the ΔΔCt method.

111

3.2.4 Serum catecholamine quantification

Serum epinephrine and norepinephrine levels were measured using Adrenaline

Research ELISA and Noradrenaline Research ELISA from Labor Diagnostika Nord GmbH

& Co.KG according to manufacturer’s protocol, respectively. Briefly, under anesthesia blood was collected from the right carotid artery and incubated at room temperature for 30 minutes. Serum was collected after centrifugation. Sample volumes used were 10 µL for norepinephrine ELISA and 40 µL for epinephrine ELISA.

3.2.5 Radioligand binding

In brief, crude cardiac membrane preparations were prepared as previously described (Pleger et al. 2008). The total density of expressed b-ARs in cardiac lysates was determined by saturation binding experiments using the nonselective b-AR antagonist (-

)3-[125I]iodocyanopindolol ([125I]CYP) as the radiolabel. 10 μM propranolol was used to determine nonspecific binding. Amounts of total and nonspecific radiolabel bound to cell membranes were calculated from radioactive counts remaining on the glass fiber filters.

From the plotted saturation hyperbola, b-AR density (Bmax) and affinity value (Kd) of

[125I]CYP for each sample were calculated using iterative nonlinear regression analysis.

3.2.6 Terminal hemodynamics

Briefly, a 1.4-French micromanometer-tipped catheter (Millar Instruments) was inserted into the right carotid artery and then advanced into the LV. A polyethylene-50 catheter was placed in the left external jugular vein for continuous infusion of

112 isoproterenol, and an isoproterenol dose response curve was performed as we have done previously (Schumacher et al. 2016). Systolic blood pressure was recorded in closed-chest mode throughout the experiment with a PowerLab DAQ System (Millar Instruments).

3.2.7 Non-invasive tail cuff measurements

Systolic and diastolic peripheral blood pressures were obtained using the BP-2000

Blood Pressure analysis system (Visitech Systems, Apex, NC, USA), an inflatable occlusion cuff system. Mice went through 6 training sessions to acclimate to the mouse holders. Once basal pressure was achieved, the tail cuff was inflated to 180 mmHg. After mice were adequately trained, as confirmed through heart rate and systolic pressure stabilization throughout the training sessions, a blinded final experimental session & analysis of blood pressure curves was done.

3.2.8 Telemetry measurements

Analysis of blood pressure and heart rate were evaluated in conscious mice over 3 days by telemetry (DSI equipped with ADInstrument 6 software) via carotid catheter (PA-

C10). Briefly, the telemeter was inserted on the left dorsal side of the mouse. After securing the telemeter, a longitudinal incision was made on the anterior portion of the neck lateral to the trachea on the left of the animal in order to expose the left carotid artery. After locating and isolating the carotid artery using blunt dissection, a 6-0 silk suture was used to occlude and tie-off the internal and external carotid artery. The suture was slightly elevated to further isolate the vessel. Another loose suture was made around the carotid

113 towards the aortic arch. Using a 19-gaugce bent-tip needle an incision was made and the catheter was inserted into the carotid artery 10-12 mm. After catheter placement, the loose suture was tightened. After completion, all incisions were closed with vetbond tissue-glue

(3M).

3.2.8 Histology

Renal vascular hypertrophy and perivascular fibrosis were evaluated with Sirius

Red (EMS) staining in kidney samples obtained from aged WT and GRK2-C340S. Briefly, after de-paraffinization and re-hydration, sections (4-5 µm thick) were stained in equal parts Weigart’s Iron Hematoxylin A and B (EMS) for 10 minutes at room temperature.

Sections were washed twice in distilled water for 3 minutes per wash. Sirius Red was added for 1 h at room temperature. Slides were washed twice in 0.01N HCL for 3 minutes per wash. Sections were then dehydrated and penetrated using ethanol and xylene, respectively. Thoracic aortas and hearts and were stained with Masson’s trichrome to distinguish media area from adventitia and fibrosis, respectively as previously described

(58). Briefly, after de-paraffinization and re-hydration, sections were incubated with

Bouin’s fluid for 1 hour at 56ºC. Sections were washed three time in distilled water for 3 minutes per wash and incubated with working HE solution for 7.5 minutes followed by washing in distilled water for 30 seconds. Sections were then incubated with Biebrich

Scarlet-Acid Fuchsin solution for 5 minutes. After incubation with phosphotungstic- phosphomolyhdic acid solution for 5 minutes, sections were stained with Aniline Blue stain solution for 5 minutes. Sections were washed in 1% acetic acid for 30 sec and distilled

114 water for 30 seconds. To analyze cell cross sectional area, heart sections were stained with wheat germ agglutinin (WGA, ThermoFisher, 10 µg/mL in PBS) for 1 hour in a humidified dark chamber, as described previously. Sections were dehydrated and penetrated using ethanol and xylene, respectively. Analysis was conducted using ImageJ software.

(Schumacher et al. 2016)

3.2.10 Immunohistochemistry

Subsequent to heat induced antigen retrieval, heart sections were immunolabeled with primary antibodies for platelet endothelial cell adhesion molecule (PECAM-1)/cluster of differentiation 31 (CD31) (R&D AF3628) and alpha-smooth muscle actin (aSMA)

(Sigma-Aldrich A5228) as described (Cimini et al. 2019). Secondary fluorophore- conjugated antibodies (ThermoFisher A21432, A21202) were then used to analyze the immunofluorescence. Images throughout the heart were obtained with the Nikon Eclipse

Ti-E fluorescence microscope, merged using Adobe Photoshop (Adobe), and analyzed with ImageJ. CD31 and aSMA counts were normalized to tissue area.

3.2.11 Histological quantification

Cardiomyocyte cell cross sectional area was quantified as previously described from WGA staining (Schumacher et al. 2016). To quantify cardiac fibrosis, multiple images from Masson’s Trichrome stained hearts were taken across the LV. The blue stained area (fibrosis) was outlined and quantified. To calculate cardiac perivascular fibrosis, the value of the fibrosis area was subtracted from the value of the vessel area

115 divided by the true area of the vessel. To quantify vascular hypertrophy in the kidney, the value of medial area was divided by the true area of the vessel. True area was calculated by vessel outer perimeter2 divided by 4π. To calculate renal perivascular fibrosis, the value of fibrosis area was subtracted from vessel area and divided by the true area of the vessel.

Minimum of three representative vascular images were analyzed per sample. Medial hypertrophy of thoracic aorta was quantified by measurements of medial thickness in 4 randomly selected locations per slide. A minimum three representative vascular images were analyzed per biological replicate.

3.2.12 Isometric aortic contraction

Isometric aortic contraction assays were performed as described previously (Pleger et al. 2008). Briefly, thoracic aortas from WT or GRK2-C340S mice were cleaned of surrounding tissue, cut into 2.5 mm rings, and then mounted onto a myograph. (Radnoti

Wire Myograph, Radnoti LLC). Rings were equilibrated for 60 minutes before measuring max contraction with KCl (100 mM). After washing out the KCl, phenylephrine (PE) dose- response was assessed (1x10-9 – 3x10-3 M). Rings were pre-contracted with an EC50 concentration of PE before assessing endothelium-dependent relaxation using acetylcholine (ACh, 1x10-9 – 3x10-3 M). To evaluate basal tone NO release, rings were incubated with L-NAME (50 µM) for 45 minutes.

116

3.2.13 Statistics

All values in the text and figures are presented as mean±SEM. Statistical significance between two groups was determined by student’s unpaired t-test. Statistical significance between three or more groups with 2 or more variables was determined by two-way ANOVA with Tukey’s multiple comparisons test. Probabilities of 0.05 or less were considered to be statistically significant. Statistical calculations were done in Prism 7

(GraphPad).

3.3 Results

3.3.1 Aged GRK2-C340S mice have reduced cardiac function

WT and GRK2-C340S mice were allowed to age to 2-3 months (8-15 weeks) or

12-20 months (52-82 weeks) before implementation into the young or old study groups, respectively. Old GRK2-C340S mice express GRK2 protein at endogenous levels seen in old WT mice, corresponding with our previous study confirming young GRK2-C340S mice also express GRK2 at endogenous levels seen in young WT mice while activity is increased (Fig. 3-2) (Huang et al. 2013; Lieu et al. 2020). A typical laboratory mouse lifespan is 1.3-3 years, making our young mice equivalent to ~20-30 human years and our old mice equivalent to ~40-64 human years (19, 23, 33). As GRK2-C340S mice aged, we observed a steady decline in cardiac function as determined by traditional echocardiography (TE) or speckle tracking-based strain echocardiography (STE). By ≥12 months of age, GRK2-C340S mice have a significant reduction in ejection fraction (EF) detected by TE or STE compared to age-matched WT counterparts (Fig. 3-3A,B) (Lieu et

117 al. 2020). GRK2-C340S groups experienced declines in EF over time. WT EFs also declined significantly by TE and trending by STE: young WT = 65.3% vs old WT = 57.5%

(Fig. 3-3A,B). EF in old GRK2-C340S males and females were significantly declined to the same extent as measured through TE (Fig. 3-3C,D). LV internal dimensions (LVID) at systole (LVIDs) and diastole (LVIDd) were both significantly increased in old GRK2-

C340S mice compared to age-matched WT mice, indicative of LV dilation (Fig. 3-4A,B).

FIGURE 3-2 – Aged GRK2-C340S mice express mutant GRK2 at endogenous protein levels (Lieu et al. 2020). (A) Representative Western blot image of GRK2 (~80 kDa) and GAPDH (~37 kDa) protein levels from old WT and old GRK2-C340S heart lysates. (B) Quantification of GRK2/GAPDH protein levels from Western blot showing no change between old WT (64-75 weeks) and old GRK2-C340S hearts (68-71 weeks). (n=4).

Internal diameters were increased in old groups compared to young as well (Fig. 3-4A,B).

Volumes at systole and diastole were also significantly increased in old GRK2-C340S mice

118 compared to age matched WT, which further suggests impaired contractility in aged

GRK2-C340S mice (Fig. 3-4C,D). Old WT and GRK2-C340S mice had increased volumes compared to their respective young groups, as observed in all other parameters. Fractional shortening (FS) was reduced in aged GRK2-C340S mice compared to aged WT mice with a similar decline in each group over time (Fig. 3-4E). Additional echocardiographic parameters can be found in Fig. 3-4 F-H. Measurement of global longitudinal strain (GLS) through STE in aged GRK2-C340S mice revealed that they had significantly greater degree of impaired LV deformation during systole compared to aged WT (Fig. 3-5A). Further analysis of regional longitudinal or radial strain within the LV revealed similar findings that aged GRK2-C340S cardiac contraction was impaired (Fig. 3-5B,C,D,E). Overall, over time GRK2-C340S experience significantly worsened systolic function compared to age matched WT mice. Declines in systolic function with aging have become more appreciated as advances in echocardiography techniques, ie STE, improve its sensitivity (de Lucia et al. 2019).

As impaired diastolic function has traditionally been most appreciated to accompany aging, we assessed diastolic function through STE, pulse wave doppler, and tissue doppler echocardiography. STE revealed that reverse longitudinal strain rate in both young and old GRK2-C340S mice were significantly impaired compared to age-matched

WT groups, suggesting there are impairments in cardiac relaxation (Fig. 3-6A). Multiple other diastolic indices trended towards impairment in aged GRK2-C340S mice compared to aged WT, including increased isovolumic relaxation time (IVRT) (Fig. 3-6B), reduced

E/A ratio (Fig. 3-6C), reduced E/E’ ratio (Fig. 3-6D), and increased LV myocardial

119

FIGURE 3-3 – Aged GRK2-C340S mice have reduced ejection fraction (Lieu et al. 2020). (A) Ejection fraction (EF) from traditional echocardiography (TE) (n=8-20) and (B) speckle tracking-based strain echocardiography (STE) in young (8-15 weeks) and old (52- 82 weeks) wild type (WT, white circles) or GRK2-C340S (black squares) mice. (n=8). Ejection fraction (EF) by TE in aged WT or GRK2-C340S mice separated by (C) males and (D) females (n=6-14); *p<0.05, **p<0.01, ***p<0.001,****p<0.0001 as determined by two-way ANOVA with Tukey’s post hoc analysis. ttp<0.01; tttp<0.001 as determined by student’s t-test between WT and GRK2-C340S.

120 performance index (MPI) (Fig. 3-6E) compared to age-matched WT controls. Over time we observed declining cardiac function, congruent with previously published studies in aging (de Lucia et al. 2019). Importantly, GRK2-C340S mice with global loss of SNO- mediated GRK2 regulation have significant systolic and diastolic dysfunction that is more pronounced compared to aged WT mice.

3.3.2 Aged GRK2-C340S mice have adrenergic abnormalities

We have previously observed that the loss of SNO regulation on GRK2 in the hearts of GRK2-C340S mice led to over-active GRK2 and enhanced b-AR desensitization with isoproterenol challenge (Huang et al. 2013). In order to determine if the cause of decreased cardiac function stemmed from cardiac b-AR abnormalities, b-AR density was measured by radioligand binding of cardiac membranes. A decrease in b-AR density was observed in young GRK2-C340S mice compared to young WT (Fig. 3-7A) (Lieu et al. 2020).

However, as previously established, young WT mice do not show cardiac dysfunction at baseline suggesting that reduced b-ARs alone was not sufficient to produce dysfunction. b-AR density was reduced in both aged groups compared to young WT mice, but to the same extent (Fig. 3-7A). Although young GRK2-C340S mice have reduced b-AR expression that is maintained over time, it does not appear to be the primary contributor to the developing cardiac dysfunction.

Next, we looked at serum catecholamine levels as sympathetic nervous system

(SNS) activation is a hallmark of CVD, including heart failure and accompanies b-AR desensitization and downregulation (Lymperopoulos et al. 2013). Further, catecholamines

121

FIGURE 3-4 – Aged GRK2-C340S mice have reduced cardiac function – additional echocardiographic parameters. Aged GRK2-C340S mice have reduced cardiac function as shown through multiple echocardiographic parameters obtained through traditional echocardiography beyond ejection fraction: left ventricular internal diameters (LVID, mm) of young or aged WT or GRK2-C340S at (A) systole and (B) diastole; end diastolic volumes (µL) at (C) systole and (D) diastole obtained by TE; (E) fractional shortening (FS, %); (F) cardiac output (CO, mL/min); (G) stroke volume (SV, µL); (H) heart rate (HR, bpm) in young and old WT or GRK2-C340S mice. (n=8-20). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 as determined by two-way ANOVA with Tukey’s post hoc analysis. 122

FIGURE 3-5 – Aged GRK2-C340S mice have impaired longitudinal and radial myocardial strain (Lieu et al. 2020). The following parameters were obtained with to show myocardial strain in aged GRK2-C340S mice: (A) global longitudinal strain (GLS); (B) radial and (C) longitudinal strain peak (%) in young WT or GRK2-C340S animals; (D) radial and (E) longitudinal strain peak (%) in old WT or GRK2-C340S animals (n=8). tp<0.05; ttp<0.01 as determined by student’s t-test between WT and GRK2-C340S. *p<0.05 as determined by two-way ANOVA with Tukey’s post hoc analysis.

123

FIGURE 3-6 – Aged GRK2-C340S mice have a degree of diastolic dysfunction (Lieu et al. 2020). (A) Reverse longitudinal strain rate peak (1/s) in young and old GRK2-C340S mice obtained through speckle-tracking based echocardiography. Pulsed wave doppler and/or tissue doppler was used to obtain: (B) isovolumic relaxation time (IVRT, ms); E (peak early diastolic transmitral velocity) and A (peak late diastolic transmitral velocity) waves for (C) E/A ratio; E’ (early-diastolic myocardial relaxation velocity for (D) E/E’ ratio; and (E) left ventricular myocardial performance index. (n=8). *p<0.05, **p<0.01, ***p<0.001 as determined by two-way ANOVA with Tukey’s post hoc analysis.

124 have been shown to increase over time (Dai et al. 2009). Of note, we found significantly elevated serum levels of epinephrine in aged GRK2-C340S mice compared to age-matched controls (Fig. 3-7B). Norepinephrine was significantly elevated in aged groups compared to young groups, but with no apparent differences within groups (Fig. 3-7C). The increased serum epinephrine is of physiological significance as we have previously shown GRK2 activity to be critically important in the release of this catecholamine release from the adrenal medulla as increased activity does lead to increased epinephrine levels

(Lymperopoulos et al. 2007).

3.3.3 Aged GRK2-C340S mice exhibit cardiac hypertrophy

Changes in LV function are strongly associated with cardiac remodeling (37, 54).

Therefore, we examined the cardiac morphology in old GRK2-C340S. Old GRK2-C340S mice presented with hypertrophied hearts observed by significantly increased heart weight- to-body weight (HW/BW) ratios (Fig. 3-8A) as well as heart weight-to-tibia length

(HW/TL) ratios (Fig. 3-8B) compared to aged WT (Lieu et al. 2020). There is also a striking increase in HW/BW and HW/TL in young vs old GRK2-C340S mice that is significantly more pronounced than WT mice overtime (Fig. 3-8A,B). HW/BW in WT mice do not appear to increase over time, but rather trend to be lower likely due to large increases in BW with age that masks present physiological heart growth (Fig. 3-8A).

Hypertrophy was due to increased cardiomyocyte size, as measured from cardiomyocyte cross sectional area (CSA) through wheat germ agglutinin (WGA) staining (Fig. 3-8C).

CSA increased in both aged WT and GRK2-C340S groups compared to their young

125

FIGURE 3-7 – Aged mice have increased neurohormonal activation measured by catecholamines (Lieu et al. 2020). (A) Quantification of b-AR density from radioligand binding assay in cell membranes from young and old WT or GRK2-C340S mice (n=3-7). (B) Serum epinephrine (ng/mL) (n=3-5) and (C) serum norepinephrine (ng/mL) (n=4-12) levels measured by ELISA in young and old WT or GRK2-C340S mice. *p<0.05, **p<0.01,****p<0.0001 as determined by two-way ANOVA with Tukey’s post hoc analysis.

126 counterparts (Fig. 3-8D). Aged GRK2-C340S mice displayed more pronounced and significant hypertrophy compared to aged WT (Fig. 3-8D). Trending increased expression of ANF (Fig. 3-8E), BNP (Fig. 3-8F), and GRK2 mRNA (Fig. 3-8G) in GRK2-C340S mice supports this hypertrophy being pathological in nature.

3.3.4 Aged TgGRK2 mice do not exhibit age-dependent cardiac dysfunction

We also examined traditional echocardiographic parameters in aged (73 weeks) mice with cardiac-specific overexpression of GRK2 (TgGRK2) compared to age-matched non-transgenic littermate controls (NLCs). TgGRK2 mice have increased cardiac-specific

GRK2 activity in contrast to GRK2-C340S mice that have increased global GRK2 activity.

This could give us insights on the important role of S-nitrosylation on GRK2 on long-term cardiac function. Interestingly, aged TgGRK2 mice do not show any declines in echocardiographic parameters compared to aged NLCs including: EF (Fig. 3-9A), LVIDs

(Fig. 3-9B), LVIDd (Fig. 3-9C), LV end systolic volume (Fig. 3-9D), LV end diastolic volume (Fig. 3-9E), FS (Fig. 3-9F), CO (Fig. 3-9G), SV (Fig. 3-9H), or HR (Fig. 3-9I).

In summary, aged TgGRK2 cardiac function was indistinguishable from age matched

NLCs, emphasizing the importance of intact NOS3 regulation.

3.3.5 Aged TgGRK2 mice do not experience maladaptive remodeling compared

to non-transgenic littermate controls.

Previous TAC studies in mice with cardiac-specific overexpression of GRK2

(TgGRK2), overexpression confirmed in Fig. 3-10G, have demonstrated that TgGRK2

127

FIGURE 3-8 – Aged GRK2-C340S mice exhibit cardiac hypertrophy. Measures of (A) heart weight (HW) to body weight (BW) ratios and (B) HW to tibia length (TL) ratios between young and old WT or GRK2-C340S mice. (n=12-18) (C) Representative images of wheat germ agglutinin (red) and 4′,6-diamidino-2-phenylindole (DAPI, blue)–stained young and old WT or GRK2-C340S mouse heart sections. Scale bar, 100 µm. (D) Quantification of cardiomyocyte cross sectional area (CSA) from WGA staining in young and old WT or GRK2-C340S hearts (n=4). Quantification of qRT-PCR data in young and old WT or GRK2-C340S hearts showed fold change in mRNA expression of (E) atrial natriuretic factor (ANF), (F) brain natriuretic peptide (BNP) (n=5-6), and (G) GRK2 (n=11). *p<0.05, **p<0.01, ***p<0.001,****p<0.0001 as determined by two-way ANOVA with Tukey’s post hoc analysis.

128

FIGURE 3-9 – Aged cardiac TgGRK2 mice do not experience declines in cardiac function compared to aged NLC. Cardiac function in aged TgGRK2 mice are similar to aged non-transgenic littermate controls (NLCs), measured through traditional echocardiography: (A) ejection fraction (EF,%); left ventricular internal diameters (LVID, mm) at (B) systole and (C) diastole; left ventricular end volumes (µL) at (D) systole and (E) diastole; (F) fractional shortening (FS, %); (G) cardiac output (CO, mL/min); (H) stroke volume (SV, µL); (I) heart rate (HR, bpm). (n=4).

129

FIGURE 3-10 – Aged cardiac TgGRK2 mice do not experience maladaptive remodeling compared to aged NLC (Lieu et al. 2020). Measures of (A) heart weight (HW) to body weight (BW) ratios (HW/BW) and (B) HW to tibia length (TL) ratios (HW/TL) in young (13 weeks) and old (73 weeks) male non-transgenic littermate controls (NLC, white bars) or cardiac-specific transgenic GRK2 mice (TgGRK2, black bars). (C) Representative images of wheat germ agglutinin (red) and 4′,6-diamidino-2-phenylindole (DAPI, blue)–stained aged NLC or TgGRK2 mouse heart sections. Scale bar, 100 µm. (D) Quantification of cardiomyocyte cross sectional area (CSA) from WGA staining in young and old NLC or TgGRK2 hearts. (n=3-5). Quantification of qRT-PCR data in young and old WT or GRK2-C340S hearts showing fold change in mRNA expression of (E) atrial natriuretic factor (ANF), (F) brain natriuretic peptide (BNP) and (G) G protein-coupled receptor kinase 2 (GRK2). (n=4-5). *p<0.05, **p<0.01, ****p<0.0001 as determined by two-way ANOVA with Tukey’s post hoc analysis.

130 mice develop cardiac hypertrophy to the same extent as their non-transgenic littermate controls (NLCs), despite having increased GRK2 expression and activity (Schumacher et al. 2016). In order to determine if cardiac GRK2 was specifically responsible for the

GRK2-C340S phenotype, we aged TgGRK2 mice to 73 weeks. Although old TgGRK2 hearts hypertrophied over time, it was no different than age matched NLCs as measured through HW/BW (Fig. 3-10A), HW/TL (Fig. 3-10B), and CSA (Fig. 3-10C,D) (Lieu et al.

2020). Maladaptive hypertrophic markers, ANF and BNP, between aged TgGRK2 and

NLC animals were unchanged (Fig. 3-10E,F). Like aged WT controls, aged NLC animals did not show any notable increase in ANF or BNP compared to young counterparts.

Overall, GRK2-C340S mice develop maladaptive cardiac hypertrophy over time beyond what is physiologically normal, likely due to non-cardiomyocyte un-checked GRK2 activity since TgGRK2 do not show signs of maladaptive hypertrophy or cardiac dysfunction.

3.3.6 Aged GRK2-C340S mice exhibit adverse cardiac and renal remodeling

As hearts age physiologically, there are changes in the vasculature including increased fibrosis (Biernacka and Frangogiannis 2011). Examination of the vasculature in aged GRK2-C340S hearts through Masson’s Trichrome (Fig. 3-11A) revealed significantly increased perivascular fibrosis compared to aged WT hearts (Fig. 3-11B)

(Lieu et al. 2020). Additionally, perivascular fibrosis increased over time in both WT and

GRK2-C340S mice, but more pronounced in the latter. Decreases in capillary density with age have been previously reported predominantly in the brain, but in the heart and skeletal

131 muscle as well (Groen et al. 2014; Li et al. 2018; Maizel et al. 2014; Sonntag et al. 1997).

To examine if there were any changes in vascular density over time due to increased GRK2 activity, immunohistochemistry on heart sections were done with CD31 (Fig. 3-11C) to mark all vessels, including capillaries, and aSMA to mark larger vessels (arteries, arterioles, veins, and venules). We observed a trend towards decreased CD31 density over time in WT mice, which was significantly reduced in GRK2-C340S mice over time (Fig.

3-11D). Similar findings were observed in aSMA density over time (Fig. 3-11E).

Given that the renal system is closely linked to the cardiovascular system so that pathology in one system increases the risk of pathology in the other and vice versa, we investigated whether the vasculature in the kidneys was also abnormal (Agarwal and

Thomas 2003; Coresh et al. 2001; Liu et al. 2014). Histological examination of the kidneys by Sirius Red staining of aged GRK2-C340S mice (Fig. 3-12A) revealed an increase in vascular medial thickness shown by an increase in vascular medial area to internal arterial area ratio in renal vessels compared to age-matched WT mice (Fig. 3-12B). Moreover, we found that aged GRK2-C340S mice had significantly more renal perivascular fibrosis compared to aged WT mice, parallel to what was observed within the cardiac vasculature

(Fig. 3-12C). Thus, aged GRK2-C340S mice exhibit adverse remodeling not only in the myocardium but also in the kidneys that include significant perivascular fibrosis and vascular hypertrophy. This perivascular fibrosis is evident in multiple organs and highlights the vasculature’s vulnerable nature to GRK2 activity. Pathological signaling and remodeling due to upregulated GRK2 activity in the vasculature may be a significant cause of the GRK2-C340S phenotype. Overall, aged GRK2-C340S mice displayed significant

132

FIGURE 3-11 – Aged GRK2-C340S mice exhibit cardiac vascular abnormalities (Lieu et al. 2020). (A) Representative images of Masson’s trichrome stained aged WT or GRK2-C340S mouse heart sections. Scale bar, 100 µm. (B) Quantification of perivascular fibrotic area in young or aged WT and GRK2-C340S hearts (n=4). (C) Representative images of young and old WT or GRK2-C340S cardiac sections immunolabeled with cluster of differentiation 31 (CD31) (red). Nuclei (blue) were stained with 4′,6-diamidino-2- phenylindole (DAPI). Scale bar, 50 µm (D) Quantification of CD31 density expressed as a percentage of CD31 positive vessels per tissue area. (n=4). (E) Quantification of a- smooth muscle actin (aSMA) density expressed as a percentage of aSMA positive vessels per tissue area. (n=4). *p<0.05, **p<0.01, ****p<0.0001 as determined by two-way ANOVA with Tukey’s post hoc analysis.

133

FIGURE 3-12 – Aged GRK2-C340S mice develop renal vascular hypertrophy and fibrosis (Lieu et al. 2020). (A) Representative higher magnification (100x) images of Sirius Red stained aged kidneys from WT and GRK2-C340S mice demonstrating perivascular fibrosis. (B) Quantification of vascular medial area to internal arterial area ratio as a measure showing increased medial hypertrophy in aged GRK2-C340S mice compared to WT. (C) Quantification of perivascular fibrosis to vascular area ratio from Sirius Red staining in aged WT and GRK2-C340S mice (n=4-6). ttttp<0.0001 as determined by student’s t-test.

multi-organ perivascular fibrosis, renal vascular hypertrophy, and decreased cardiac vessel density over time.

3.3.7 Aged GRK2-C340S mice exhibit increased systolic blood pressure

With increased perivascular fibrosis, we next examined aged GRK2-C340S mice hemodynamically to give insights on vascular function. Young mice had no differences in blood pressure (Fig. 3-13A). Consistent with elevated SNS activity, we found that aged

GRK-C340S mice had significantly higher systolic pressures at baseline compared to aged 134

WT mice within the LV (Fig. 3-13B) (Lieu et al. 2020). Aged GRK2-C340S mice continued to display significantly elevated pressures during b-adrenergic challenge with isoproterenol compared to aged WT mice despite the young groups being no different (Fig.

3-13C,D).

We also investigated whether peripheral blood pressures were concomitantly elevated through two methods: 1— telemeter inserted in the left carotid artery for 3 days and; 2— BP2000 Blood Pressure Analyzer, representing a non-invasive tail cuff method.

Although both methods did not reveal any statistically significant differences in the individual parameters of systolic and diastolic pressure, there was a trend towards narrowed pulse pressure (Fig. 3-14). This phenotype is of interest as reduced pulse pressure has been linked to advanced HF and independently predicts mortality, a reflection from poor cardiac function (Yildiran et al. 2010). Increases in pulse pressure windows have also been associated with CVD, but more so with mild HF (Yildiran et al. 2010).

3.3.8 Aortic size is reduced in GRK2-C340S mice

Given that aged GRK2-C340S mice had increased LV pressures and significant multi-organ perivascular fibrosis, we took a closer look at the aortic morphology within these mice. This is especially important since the eNOS machinery is prevalent in vessels and the un-checked GRK2 activity via the loss of SNO regulation will lead to a loss of eNOS activity as we have previously shown (Huang et al. 2013). We examined aortic sections with Masson’s Trichrome stationing from young and old WT and GRK2-C340S mice (Fig. 3-15A). Aged KI mice had significantly reduced medial thickness compared to

135

FIGURE 3-13 – Aged GRK2-C340S mice exhibit increased blood pressures (Lieu et al. 2020). Invasive hemodynamic analysis of WT or GRK2-C340S mice within the left ventricle: systolic pressure (mmHg) at baseline in (A) young and (B) old mice; systolic blood pressure (mmHg) in (C) young and (D) old mice in response to isoproterenol (0-10 ng) (n=11) tp<0.05, as determined by student’s unpaired t-test between WT compared to GRK2-C340S.

aged WT aortas, seemingly through a reduction in vascular smooth muscle cells (VSMC)

(Fig. 3-15B). Surprisingly, young GRK2-C340S mice presented with the same phenotype of reduced medial thickness. Although aortic abnormalities are observed at early time points, it does not produce any functional consequences until seemingly at later time points.

Additionally, it could be a driving force further supporting a potential vasculature origin of

GRK2-mediated CV dysfunction. These morphological changes begin at an early age and could result in pressure gradients that could induce the hypertrophic and fibrotic changes we found in the hearts of aged GRK2-C340S mice above. These data suggest aged GRK2-

136

FIGURE 3-14 – Aged GRK2-C340S mice trend towards narrowed pulse pressure windows (Lieu et al. 2020). (A) Systolic pressure and (B) diastolic pressure in young and old WT (white bars) or GRK2-C340S (black bars) mice measured with the BP-2000 Blood Pressure analysis system utilizing a non-invasive tail cuff system. (C) Pulse pressure calculated from systolic and diastolic pressures obtained through tail cuff. (n=5) (D) Systolic and (E) diastolic pressures in young and old WT or GRK2-C340S mice measured through a telemeter inserted into the left carotid artery for 3 days. (F) Pulse pressure calculated from systolic and diastolic pressures obtained through telemetry. (n=9,11). *p<0.05 as determined by two-way ANOVA with Tukey’s post hoc analysis.

137

FIGURE 3-15 - Aortic size is reduced in GRK2-C340S mice (Lieu et al. 2020). (A) Representative images of Masson-trichome stained aged aortas from WT and GRK2- C340S mice demonstrating reduced medial thickness in GRK2-C340S mice. Scale bar, 100 µm. (B) Quantification of medial thickness from Masson Trichrome staining in young and old WT and GRK2-C340S mouse aortas (n=3-4) *p<0.05, **p<0.01 as determined by two- way ANOVA with Tukey’s post hoc analysis.

C340S aortas are overall smaller or narrower than their WT counterparts. Thus, the loss of cardiac function observed in aged GRK2-C340S likely was a result of global, likely cardiac independent, differences in GRK2 activity through loss of S-nitrosylation that originated in the vasculature, which resulted in reduced vascular media and smaller aortic dimensions.

3.3.9 Aged GRK2-C340S mice have impaired aortic contraction

Given that aged GRK2-C340s mice have reduced VSMCs within their media and significant perivascular fibrosis, we examined whether this caused any changes in aortic contractile properties. Aortic rings were isolated from young or old WT and GRK2-C340S

138 mice then placed in tissue baths. Upon 100 KCl incubation, aortic rings from young WT and GRK2-C340S mice responded with the same force of contraction (Fig. 3-16A).

However, aged GRK2-C340S aortas displayed significantly reduced contraction compared to aged WT (Fig. 3-16B). Similar impaired contraction responses were seen with the b-AR agonist phenylephrine (PE), as contraction was similar in young WT and GRK2-C340S aortas but was significantly reduced in aged GRK2-C340S mouse aortas (Fig. 3-16C,D).

Relaxation responses in pre-contracted rings to acetylcholine were identical in WT and

GRK2-C340S mice regardless of age (Fig. 3-16E,F). As loss of SNO-mediated GRK2 regulation is likely responsible for these vascular contractile changes, we challenged rings with the eNOS inhibitor L-NG-Nitroarginine methyl ester (L-NAME) to induce loss of

NO. Aortic rings from young WT and GRK2-C340S mice had similar contractile responses due to the loss of the vasorelaxant NO (Fig. 3-16G). However, aortas from aged GRK2-

C340S mice had significantly impaired responses to L-NAME compared to age-matched

WT mice demonstrating the importance of SNO-mediated GRK2 inhibition in the vascular contractile response (Fig. 3-16H).

3.4 Discussion

GRK2 has been studied for decades and has been established as a critical pathological kinase in CVD. We see that in animal models subject to cardiac injury–MI,

I/R, and trans-aortic constriction (TAC)–GRK2 expression and/or activity increase and contribute to cardiac dysfunction and myopathy (Lieu and Koch 2019). How GRK2 initiates maladaptive signaling pathways in the heart and related tissues or contributes to

139

FIGURE 3-16 – Aged GRK2-C340S mice exhibit aortic contraction dysfunction. Two to four aortic sections were analyzed from each mouse to obtain: (A-B) Isometric contractility from baseline of aortic rings with 100 mM KCl incubation from (A) young (n=5) and (B) aged (n=8) WT (white bars) or C340S mice (striped bars). (C-D) Contractile dose response curves to phenylephrine in (C) young (n=5) and (D) old (n=8) WT (solid circle) or C340S (open circle) aortic rings. (E-F) Relaxation dose response curves to acetylcholine in (E) young (n=5) and (F) old (n=8) WT (solid circle) or C340S (open circle) aortic rings. (G-H) Contractility from baseline of aortic rings with 50 µM N-Nitroarginine methyl ester (L-NAME) incubation from (G) young (n=5) and (H) aged (n=4-5) WT (white bars) or C340S mice (striped bars). tp<0.05, tttp<0.001, ttttp<0.0001 as determined by student’s unpaired t-test between WT and GRK2-C340S 140 the changing molecular landscape overtime remains pertinent research questions as we continue to understand GRK2’s functions. In this study, we allowed GRK2-C340S mice that lack dynamic SNO-GRK2 regulation, an endogenous mechanism for attenuating

GRK2 activity, to age at least 12 months before enrollment. This unique mechanism of

GRK2 activation in GRK2-C340S mice plus aging allowed us to investigate the consequences of global GRK2 over-activity that may not be observed using other GRK2 mouse models. The use mice ranging from 12-20 months may not be old enough to significantly recapitulate all hallmarks of aging compared to 24-36 month old mice used in traditional aging studies (Dai et al. 2009; Zheng et al. 2003), but we are primarily concerned with the effects of long term GRK2 overactivity on the cardiovascular system. For example, we see significant increases in LV hypertrophy and activation of neurohormonal signaling, but only trends in declining diastolic function. Overall, we report that chronic

GRK2 overactivity is sufficient in causing significantly worsened cardiovascular parameters compared to WT mice, which can provide insights into how GRK2 participates in CVD development including during intrinsic aging. A summary of our findings in the

GRK2-C340S mouse is reviewed in Fig. 3-17.

It is debated whether GRK2 has a direct mechanistic role in inducing cardiomyocyte hypertrophy in mice (Lieu and Koch 2019). Multiple in vitro studies have shown GRK2 is sufficient to upregulate hypertrophy markers (ie. ANF, BNP, beta-myosin heavy chain) or hypertrophic genes (ie nuclear factor of activated T-cells and nuclear factor kappa-light-chain-enhancer of activated B cells) and increase cell size in cardiac cell lines

(Schlegel et al. 2017; Sorriento et al. 2015). However, animal models with GRK2

141

FIGURE 3-17 – Summary of aged GRK2-C340S phenotype (Lieu et al. 2020). Illustration summarizing the age-dependent cardiovascular changes as a result of hyperactive GRK2 from loss of dynamic S-nitrosylation negative regulation in the GRK2- C340S mouse. At young ages, WT and GRK2-C340S cardiovascular function and structure are indistinguishable, save for reduced aortic medial thickness in GRK2-C340S mice that is maintained. Over time, increased GRK2 activity causes cardiovascular pathology beyond what is observed in normal physiological aging in WT mice. Aged GRK2-C340S mice present with increased cardiac hypertrophy, cardiac and renal perivascular fibrosis, serum catecholamines, and systolic pressure. Cardiac and aortic function decline in aged GRK2-C340S mice compared to aged WT. GRK2 overactivity in the vasculature and adrenals is likely to play an important role in the overall cardiovascular phenotype observed.

142 manipulation largely have not translated these findings. For example, we have previously shown cardiac-specific GRK2 overexpressing transgenic (TgGRK2) mice do not experience any changes in cardiac mass after TAC surgery compared to NLCs

(Schumacher et al. 2016). Of note, these studies enrolled TgGRK2 animals at 8-10 weeks of age, which would not take into account changes in the molecular landscape over time.

Studies involving GRK2 inhibition in cardiac hypertrophy animal models present with conflicting results making it difficult to conclude if cardiac GRK2 can primarily cause hypertrophy (Lieu and Koch 2019). Further, transgenic mice with cardiac-specific expression of the bARKct, a GRK2 peptide inhibitor, underwent TAC and hypertrophied to the same degree as NLCs mice (Choi et al. 1997; Schumacher et al. 2016). Kamal et al. reported inhibition of GRK2 through gallein, was able to ameliorate hypertrophy in TAC operated mice compared to sham (Kamal et al. 2014). Gallein being a small molecule delivered systemically would make it difficult to differentiate whether its effect was from inhibition of cardiac myocyte GRK2 or from other cell types. Gallein’s highly homologous structural analog, M119, was also able to inhibit phospholipase C activation which could be another mechanism to explain hypertrophic protection from gallein (Bonacci et al.

2006). In the present study, we demonstrated that 73-week-old cardiac TgGRK2 animals still do not experience hypertrophy any differently than NLCs over time. Unchanged cardiac hypertrophy in aged TgGRK2 was represented by unchanged HW/TL ratios, hypertrophic gene expression, and cardiomyocyte cross sectional area compared to NLC.

However, the GRK2-C340S KI mouse with global over-activation of GRK2 did present with cardiac hypertrophy over time. This supports cardiomyocyte pools of GRK2 may not

143 be directly responsible nor sufficient to elicit cardiac hypertrophy in vivo, which agrees with our previous studies that demonstrate young TgGRK2 and bARKct mice hypertrophied to the same extent as NLC mice after TAC (Iaccarino et al. 2005;

Schumacher et al. 2016). Currently, there have been no studies done investigating the effect of TAC on aged TgGRK2 nor TgbARKct mice, but something worthy of a future study. However, TAC has been performed on aged WT rodents to investigate the effects of aging on cardiac stress responses. Old age during TAC has been reported to augment pressure overload adaptations in a number of ways, but hypertrophic responses across studies were minimal if present at all likely due to reduced overall protein production

(Geng et al. 2017; Isoyama et al. 1987; Sopko et al. 2010). It is clear aging changes the molecular landscape so that cardiac stress responses will be different compared to young subjects. It is difficult to hypothesize how aged TgGRK2 or bARKct will respond to TAC given these distinctions, but GRK2 activity is likely to have no effect on hypertrophy while still worsening cardiac function. Despite increased GRK2 activity being detrimental to cardiac function during stress (ie isoproterenol challenge, TAC surgery), TgGRK2 mice do not experience declines in EF overtime like the GRK2-C340S mice likely due to the

NOS3 pathway being intact in TgGRK2 animals.

Non-cardiomyocyte pools of GRK2 may be more likely the cause for our hypertrophic phenotype observed. Our data suggests that GRK2 activity in adrenal glands and vasculature is the culprit for initiating maladaptive hypertrophic signaling in the heart.

Indeed other animal models with elevated serum catecholamine levels from non-GRK2 manipulations or chronic b-AR stimulation develop cardiac hypertrophy overtime

144

(Iaccarino et al. 1999; Mathar et al. 2010). There was a trend towards increased serum epinephrine in young KI mice that was significantly elevated in old KI mice, which is consistent with increased GRK2 activity in the adrenal medulla where desensitization of a2-ARs have been observed (Lymperopoulos et al. 2007). Increased GRK2 activity seems to play a role in perivascular collagen deposition and hypertension. In addition to fibroblasts, multiple cell types have been implicated in perivascular fibrosis development including smooth muscle cells, endothelial cells, pericytes, and immune cells (Ytrehus et al. 2018). The ability of the heart to delivery oxygen-rich and nutrient-rich blood is subject to cardiac vessel compliance, which perivascular fibrosis greatly impairs (Dai et al. 2012).

Of note, fibrosis during hypertension begins in the perivascular space (Ytrehus et al. 2018).

Vascular stiffness itself increases the risk of stroke and MI (Mattace-Raso et al. 2006; Said et al. 2018). Both aging and hypertension are known risk factors of vascular stiffness, and indeed our aged GRK2-C340S mice presented with higher systolic blood pressures in the

LV compared to aged matched controls at baseline and during b-AR stimulation. Roles for GRK2 in endothelial and vascular smooth muscle cell physiology have been previously identified. Loss of GRK2 in endothelial cells impaired vasorelaxation, increased macrophage infiltration, and collagen degradation (Ciccarelli et al. 2013). Interestingly, antagonizing GRK2 through systemic administration of M119 in an ISO model of HF reduced perivascular fibrosis compared to ISO treatment alone (Casey et al. 2010).

Vascular smooth muscle (VSM) specific overexpression of GRK2 in mice (TgSM22α-

GRK2), 2-4 months old, presented with significant increases in peripheral blood pressures compared to this study (Eckhart et al. 2002). TgSM22α-GRK2 have ~80% increase in

145

GRK2 expression compared to NLC and aortic primary smooth muscle cells derived from these mice have ~300% increase, both leading to increased GRK2 activity as well (Eckhart et al. 2002). Although GRK2-C340S mice do have increased global GRK2 activity including the VSM, we may not see such dramatic increases in peripheral blood pressure due to lack of transgenic overexpression and the existence of other GRK2 regulatory pathways beyond S-nitrosylation. It is likely that the increase in GRK2 activity observed in GRK2-C340S mice is far less than in TgSM22α-GRK2. Eckhart et al did not report any changes in perivascular fibrosis or vascular density in TgSM22α-GRK2 mice, but VSM- specific GRK2 overexpression was sufficient in inducing myocardial hypertrophy measured by increased HW/BW ratio (Eckhart et al. 2002). Mice with VSM-specific knockout of GRK2 did not observe any changes in heart size at baseline compared to controls and hypertrophied to the same extend as controls in a renal stenosis model of hypertension (Cohn et al. 2008). The impact of GRK2 on vessel density remains as a gap in knowledge in the heart, but hypertrophic cardiomyopathy patients were reported to have reduced arteriolar density (Schwartzkopff et al. 1998). Further, inhibition of GRK2 by miR-K12 enhanced tumor angiogenesis while overexpression of GRK2 inhibited angiogenesis (Li et al. 2016). GRK2’s role in angiogenesis during cancer sets the precedence for its investigation in the heart. Overall, it is evident that GRK2 plays an important role in vascular and adrenal physiology and has a mechanistic role in producing the cardiovascular effects observed in the aged GRK2-C340S mice.

Over the course of a lifetime, there is progressively increasing disorder at molecular, structural and functional levels due to age that is ironically exacerbated by the

146 body’s compensatory mechanisms to preserve function (Lakatta 2015). Therefore, the impact of age-related changes on the CV landscape cannot be underestimated when it comes to studying CVD therapies and mechanisms, especially when prevalence and risk of CVD continue to increase after adult onset (≥20 years) (Benjamin et al. 2019). GRK2’s role in aging is another area of research that remains obscure within the CV system, but area that demands more research considering CVD is marked with GRK2 activity and patients are predominantly elderly. The few existent studies suggest GRK2 is increased in expression and activity in the aging rat aorta and carotid artery, but not in lymphocytes

(Leosco et al. 2003; Schutzer et al. 2001). This may not be surprising considering normal physiological aging is characterized by b-AR system alterations such as: including increased circulating catecholamines, reduced b-AR density, increased b-AR desensitization, increased Gai protein and signaling, and decreased contractility (Crimmins et al. 2008; Ferrara et al. 2014). GRK2 has been established to directly cause all of the above during CVD via phosphorylating b-ARs (de Lucia et al. 2018a). Increased GRK2 expression has also been reported in aged rat livers and aged human brains (Obrenovich et al. 2006; Shi et al. 2018). As GRK2 levels increase over time and are elevated in CVD, the

GRK2-C340S mouse presents with a chronic disease phenotype that considers CVD and age which can be a valuable tool. GRK2’s role in aging demands more research in the future in order to comprehensively understand its contribution. GRK2 remains an attractive therapeutic target and multiple efforts have been made to inhibit its activity, including pharmacologically (Kamal et al. 2014; Schumacher et al. 2015), virally (Raake et al. 2013;

Schlegel et al. 2017; White et al. 2000), and genetically (Koch et al. 1995; Raake et al.

147

2008). Anti-GRK2 therapy may be beneficial not only during CVD but perhaps as a preventative measure.

Besides the limited current research available for GRK2 in aging, this present study has other limitations. The phenotype of reduced medial thickness, seemingly from reduced smooth muscle content, in our GRK2-C340S mice was unexpected and incongruent with phenotypes in aging models and humans. Aging does cause vascular structural remodeling, but it is comprised of collagen accumulation, decreased elastin, and VSMC proliferation within the media (Xu et al. 2017). The mechanism resulting in this unique phenotype would require further investigation. The seemingly smaller aortas may be the cause of reduced pulse pressure. Reduced pulse pressure can be observed in settings of aortic narrowing, like aortic stenosis or aortic coarctation although there can be high heterogeneity between patients and dependence on severity (Arani and Carleton 1967; GUPTA and WIGGERS

1951). Our aortic phenotype remains histologically different compared to aortic stenosis or coarctation, but we do observe parallels within our present data ie high pressures above the aorta and lower pressures below the aorta (DUNNILL 1959; Mourino-Alvarez et al. 2018).

Although the origin of reduced medial thickness lies within our manipulation of GRK2, the subsequent mechanisms of our unique aortic phenotype are currently unknown.

The mechanisms of increased GRK2 activity seen in the TgGRK2 and GRK2-

C340S mouse are brought upon differently. One is through transgenic overexpression and the other is through loss of a negative regulatory mechanism, respectively. At this point, it is unknown whether SNO-GRK2 changes GRK2’s interactome or whether it is gaining or losing interacting partners, aside from preventing b-arrestin recruitment. Therefore, it is

148 debatable whether the loss of SNO-GRK2 may impact other protein functions that are relevant to CVD.

Although, the relationship between NOS3, NO, and GRK2 has been under adequate investigation, there is still much unknown about SNO-GRK2 specifically. GRK2 is a negative regulator of NO bioavailability not only during in heart disease, but in diabetes as well, contributing to insulin resistance and endothelial dysfunction. Avendaño et al. measured vascular NO production in GRK2+/− and WT mice after AngII infusion and found that while AngII typically decreases NO bioavailability in WT mice, bioavailability in GRK2+/− was significantly preserved (Avendaño et al. 2014). Indeed, increased NO in

GRK2+/− mice supports GRK2’s negative role on NO production. Additionally, pAkt and

NOS3 expression were partially rescued in GRK2+/− mice, which further solidifies

GRK2’s ability to prevent NOS3 activity through inhibiting Akt phosphorylation at Thr308

(Taguchi et al. 2012). These molecular changes simply from reduced GRK2 expression in

GRK2+/− mice caused mice to be resistant to hypertension, vascular remodeling, and endothelial dysfunction with chronic AngII treatment. Aberrations in the Akt/NOS3 pathway by GRK2 also contribute to insulin sensitivity in diabetes. Ob/ob mice, a model of type 2 diabetes, had increased GRK2 expression and activity in aortic lysates compared to lean controls. Elevated GRK2 expression and activity corresponded with reduced NO production, reduced pAkt, and impaired relaxation with insulin stimulation, that was rescued with GRK2 inhibition (Taguchi et al. 2012). Although these studies did not measure SNO-GRK2 levels directly, we can hypothesize that in cases were pAkt and NO production were reduced, S-nitrosylation of GRK2 was also reduced which contributed to

149 endothelial dysfunction and insulin resistance due to increased GRK2 activity. In addition to disease, how aging impacts physiological levels of SNO-GRK2 is also an area to be investigated. Similar to how NO bioavailability is decreased in heart disease and diabetes,

NO production also declines with aging. This has been reported to occur in the vasculature, specifically in the endothelium (Muller-Delp et al. 2002; Tschudi et al. 1996). From our acquired data, we may hypothesize that there is a gradual loss of SNO-GRK2 in the vasculature and positively correlates with endothelial dysfunction. However, since our

GRK2-C340S mouse is unable to be S-nitrosylated regardless of age, we see hallmarks of aging occurring significantly sooner as if aging was accelerated. Although we did not assess endothelial function in depth and there was no impairment of ACh-mediated relaxation, we can suspect that relaxation due to clonidine would be impaired (Taguchi et al. 2011). Clonidine, an a2-AR agonist, promotes relaxation in an Akt/NOS3 dependent manner in contrast to ACh-mediated relaxation. Kawakami et al suggested that GRK2 has a potential negative role in regulating nitrosative stress in the setting of sepsis-associated encephaolopathy (Kawakami et al. 2018). Thus, these studies suggest there is an intimate relationship between GRK2 and NO that is not just limited to cardiomyocytes.

S-nitrosylation of GRK2 is likely to impact other intracellular effectors within its vast interactome. The most obvious consequence of GRK2 S-nitrosylation is that its kinase activity is abolished, revealing some sort of cross talk between S-nitrosylation and phosphorylation post-translational modifications (Zhou et al. 2018). Loss of kinase activity could then impact the phosphorylation state of other proteins. S-nitrosylation has also been shown to prevent kinase-substrate interactions, act synergistically with other PTMs, change

150 protein function, or subcellular localization. SNO-GAPDH is a prime example of how S- nitrosylation can change subcellular localization. Once S-nitrosylated, GAPDH has been shown to localize to the nucleus and the mitochondria where it promotes cell death or S- nitrosylates other proteins, respectively (Kohr et al. 2014). GRK2 has been shown to localize to the mitochondria (Chen et al. 2013). Whether or not S-nitrosylation of GRK2 also regulates mitochondrial localization basally or during stress is unknown. Interestingly,

GRK2 mitochondrial localization can be enhanced with Ser670 phosphorylation. Whether there is cross talk between Ser670 and Cys340 sites is yet to be investigated and an interesting discussion. It was recently discovered that GRK6 was subjected to S- nitrosylation in HEK293s and mouse brains, which opens up the possibility of cardiac

GRK5 also being regulated by S-nitrosylation (Wu et al. 2020).

151

CHAPTER 4

CONCLUDING REMARKS AND SIGNIFICANCE

4.1 Extended discussion

The research described within this dissertation comprises of two independent studies that aimed to understand GRK2 regulation in HF and aging. Although HF and aging may be investigated as separate disorders, the phenotypic similarities between the two conditions is overwhelming. In fact, the cardiovascular phenotype that results from either

HF or aging can be philosophically thought of as one and the same to varying degrees.

Time-dependent cardiovascular disorder is inevitable with aging, even in individuals without overt disease, making it difficult to distinguish between age and disease (Lakatta

2018). The source of cardiac stress in HF may be overtly identified ie hypertension, MI, or coronary artery disease. In aging, the source of cardiac stress is the overall progressive molecular disorder that negatively impacts the heart diffusely and gradually over time.

Regardless of whether the source of stress was CVD or aging, the resulting responses that the heart undergoes is similar. Both HF and aging promote cardiac hypertrophy, no doubt contributed by the continuous loss of cardiomyocytes in both conditions. In congruence with increased heart size is increased wall thickness and LV chamber dilation. Increased collagen deposition interstitially and perivascularly promote cardiac and vascular stiffening and dysfunction in both disorders. Classical isoform switching, aMHC to bMHC, and other sarcomere-based alterations observed in HF is also observed in aging, which contributes to contractile deficiencies observed along with reduced Ca2+ cycling

152

(Carnes et al. 2004; Froehlich et al. 1978; O'Neill et al. 1991). Increased neurohormonal

(ie RAAS and sympathetic) and inflammatory signaling are both key in driving HF and aging development (Dick and Epelman 2016; Strait and Lakatta 2012). Understanding these age-associated molecular changes in the cardiovascular system may be the key in developing innovative HF therapeutics and understanding HF development on a broader scale. Further, Classical anti-aging strategies have been employed experimentally to improve heart failure outcomes in animal models, such as the Koch lab as shown with chronic caloric restriction (de Lucia et al. 2018b). Thus the molecular pathways activated during disease and aging share substantial overlap.

Although miR-181a and GRK2 were only discussed in the first project related to

HF, this axis may play a relevant role in cardiac aging and senescence that has not been explored that will also contribute to heart failure. Based on the literature miR-181a and

GRK2 pathways may converge at the level of sirtuin (SIRT) 1. SIRT1 belongs to a family of SIRT proteins that are classically nicotinamide adenine dinucleotide (NAD+) - dependent deacetylases, though they possess ADP-ribosyltransferase activity as well. They play essential roles in protecting against cellular senescence, with SIRT1 being the most characterized (Lee et al. 2019). Through deacetylating effector proteins, such as FOXO transcription factors, SIRTs may promote DNA damage repair, mitochondrial function, fatty acid oxidation, and inhibit inflammation (Kida and Goligorsky 2016; Lee et al. 2019).

There is growing evidence that GPCRs play roles in cellular senescence, which may not be surprising considering the vast intracellular signaling pathways they initiate, including GRKs (Santos-Otte et al. 2019). GRK2 downregulation was required for G2/M

153 progression, and elevation of GRK2 expression during the G2/M checkpoint prolonged the cell cycle (Penela et al. 2010b). Despite evidence on its role in regulating the cell cycle,

GRK2’s role in senescence has not been directly explored but it is likely to play a role.

GRK2 has been reported to be reduced in UV-senescence induced fibroblasts to contribute to reduced protein translation (Yang et al. 2019a). FOXO1, a known SIRT1 target, was shown to positively regulate GRK2 transcription during high glucose in H9c2 cells to inhibit proliferation and promote apoptosis (Yang et al. 2019a; Yang et al. 2019b).

Although Yang et al did not assess the role of SIRT1 in response to high glucose, other have studies have characterized SIRT1-mediated deacetylation of FOXO1 to promote or inhibit its transcriptional activity (Sin et al. 2015). MiRs may also be dysregulated during aging and senescence, just as during disease (de Lucia et al. 2017). MiR-181a is no exception. Although there is not a comprehensive understanding of how miR-181a expression changes with age, it has been reported to be reduced in skeletal muscle (Soriano-

Arroquia et al. 2016), serum (Noren Hooten et al. 2013), and T-cells (Ye et al. 2018). In skeletal muscle, miR-181a was shown to directly target SIRT1 to negatively regulate myotube diameter (Soriano-Arroquia et al. 2016). MiR-181a targeting of SIRT1 has not been explored in cardiomyocytes or other cardiac cell types, but it has been shown in granulosa cells (Zhang et al. 2017). In granulosa cells, miR-181a also silences SIRT1 expression to cause FOXO1 nuclear translocation from its acetylation, causing apoptosis.

As discussed GRK2 expression can induce cardiomyocyte apoptosis after I/R injury

(Brinks et al. 2010). Thus it would be of interest to see how miR-181a, SIRT1, FOXO1, and GRK2 interact in cardiomyocytes in aging and heart disease. Loss of miR-181a during

154 pressure overload-induced HF could contribute to the increase both SIRT1 and GRK2 protein expression observed. These alterations in SIRT1 and GRK2 by miR-181a could have implications in regulating cardiomyocyte hypertrophy and apoptosis. Due it potentially being downstream of SIRT1 and FOXO1, this further supports GRK2 being involved in senescence and aging. This represents an exciting topic of research that demands investigation.

The overarching theme of these investigations is to highlight the impact of GRK2 in HF prognosis and GRK2 inhibition in improving HF outcome in patients. GRK2 inhibition not only improves cardiac function in animal models of HF, but also fibrosis, cardiomyocyte apoptosis, and endothelial dysfunction-making its therapeutic efficacy multifactorial. These characteristics have driven the decades long endeavor to develop

GRK2 therapies for use in the clinic. The clinical trial titled “Measuring G Protein-coupled

Receptor Kinase-2 (GRK2) in the Blood to Diagnose and Treat Heart Failure patients”

(NCT00419965) in 2007 confirmed that GRK2 protein levels analyzed from blood samples from HF patients positively correlated with CV mortality and all-cause mortality (Rengo et al. 2016). Pre-clinical delivery of bARKct in large animal models has proven it’s not only safe, but efficacious in ameliorating cardiac function in HF (Katz et al. 2012; Raake et al. 2013). Additionally, due to paroxetine’s success in mouse model of MI (Schumacher et al. 2015), a Phase II clinical trial (NCT03274752) to assess its ability to reduce cardiac remodeling after MI has been initiated in 2017 and is actively recruiting. Research on optimizing paroxetine’s targeting of GRK2 is ongoing to produce more robust inhibition

155

(Bouley et al. 2020). These progressions show immense promise for anti-GRK2 therapy for not only improving HF outcomes but improving patient quality of life.

156

REFERENCES CITED

Agarwal M, Thomas P. 2003. Prevalence of post-op. Nosocomial infection in neurosurgical patients and associated risk factors--a prospective study of 2441 patients. Nurs J India. 94(9):197-198, 212.

Agarwal V, Bell GW, Nam JW, Bartel DP. 2015. Predicting effective microrna target sites in mammalian mrnas. Elife. 4.

Agüero J, Almenar L, D'Ocon P, Oliver E, Montó F, Moro J, Castelló A, Rueda J, Martínez-Dolz L, Sánchez-Lazaro I et al. 2008. Correlation between beta- adrenoceptors and g-protein-coupled receptor kinases in pretransplantation heart failure. Transplant Proc. 40(9):3014-3016.

Alcendor RR, Gao S, Zhai P, Zablocki D, Holle E, Yu X, Tian B, Wagner T, Vatner SF, Sadoshima J. 2007. Sirt1 regulates aging and resistance to oxidative stress in the heart. Circ Res. 100(10):1512-1521.

Alcendor RR, Kirshenbaum LA, Imai S, Vatner SF, Sadoshima J. 2004. Silent information regulator 2alpha, a longevity factor and class iii histone deacetylase, is an essential endogenous apoptosis inhibitor in cardiac myocytes. Circ Res. 95(10):971-980.

Anversa P, Li P, Sonnenblick EH, Olivetti G. 1994. Effects of aging on quantitative structural properties of coronary vasculature and microvasculature in rats. Am J Physiol. 267(3 Pt 2):H1062-1073.

Arani DT, Carleton RA. 1967. Assessment of aortic valvular stenosis from the aortic pressure pulse. Circulation. 36(1):30-35.

Arnold WP, Mittal CK, Katsuki S, Murad F. 1977. Nitric oxide activates guanylate cyclase and increases guanosine 3':5'-cyclic monophosphate levels in various tissue preparations. Proc Natl Acad Sci U S A. 74(8):3203-3207.

Avendaño MS, Lucas E, Jurado-Pueyo M, Martínez-Revelles S, Vila-Bedmar R, Mayor F, Salaices M, Briones AM, Murga C. 2014. Increased nitric oxide bioavailability in

157

adult grk2 hemizygous mice protects against angiotensin ii-induced hypertension. Hypertension. 63(2):369-375.

Bai H, Zhou L, Wang C, Xu X, Jiang J, Qin Y, Wang X, Zhao C, Shao S. 2017. Involvement of mir-125a in resistance to daunorubicin by inhibiting apoptosis in leukemia cell lines. Tumour Biol. 39(4):1010428317695964.

Barouch LA, Harrison RW, Skaf MW, Rosas GO, Cappola TP, Kobeissi ZA, Hobai IA, Lemmon CA, Burnett AL, O'Rourke B et al. 2002. Nitric oxide regulates the heart by spatial confinement of nitric oxide synthase isoforms. Nature. 416(6878):337- 339.

Bartel DP. 2009. Micrornas: Target recognition and regulatory functions. Cell. 136(2):215- 233.

Barwari T, Joshi A, Mayr M. 2016. Micrornas in cardiovascular disease. J Am Coll Cardiol. 68(23):2577-2584.

Barzilai N, Huffman DM, Muzumdar RH, Bartke A. 2012. The critical role of metabolic pathways in aging. Diabetes. 61(6):1315-1322.

Behm-Ansmant I, Rehwinkel J, Doerks T, Stark A, Bork P, Izaurralde E. 2006. Mrna degradation by mirnas and gw182 requires both ccr4:Not deadenylase and dcp1:Dcp2 decapping complexes. Genes Dev. 20(14):1885-1898.

Beltrami AP, Barlucchi L, Torella D, Baker M, Limana F, Chimenti S, Kasahara H, Rota M, Musso E, Urbanek K et al. 2003. Adult cardiac stem cells are multipotent and support myocardial regeneration. Cell. 114(6):763-776.

Benjamin EJ, Muntner P, Alonso A, Bittencourt MS, Callaway CW, Carson AP, Chamberlain AM, Chang AR, Cheng S, Das SR et al. 2019. Heart disease and stroke statistics-2019 update: A report from the american heart association. Circulation. 139(10):e56-e528.

158

Bernstein E, Kim SY, Carmell MA, Murchison EP, Alcorn H, Li MZ, Mills AA, Elledge SJ, Anderson KV, Hannon GJ. 2003. Dicer is essential for mouse development. Nat Genet. 35(3):215-217.

Biernacka A, Frangogiannis NG. 2011. Aging and cardiac fibrosis. Aging Dis. 2(2):158- 173.

Bonacci TM, Mathews JL, Yuan C, Lehmann DM, Malik S, Wu D, Font JL, Bidlack JM, Smrcka AV. 2006. Differential targeting of gbetagamma-subunit signaling with small molecules. Science. 312(5772):443-446.

Booth SA, Charchar FJ. 2017. Cardiac telomere length in heart development, function, and disease. Physiol Genomics. 49(7):368-384.

Borchert GM, Lanier W, Davidson BL. 2006. Rna polymerase iii transcribes human micrornas. Nat Struct Mol Biol. 13(12):1097-1101.

Boudina S. 2013. Cardiac aging and insulin resistance: Could insulin/insulin-like growth factor (igf) signaling be used as a therapeutic target? Curr Pharm Des. 19(32):5684- 5694.

Bouley RA, Weinberg ZY, Waldschmidt HV, Yen YC, Larsen SD, Puthenveedu MA, Tesmer JJG. 2020. A new paroxetine-based grk2 inhibitor reduces internalization of the. Mol Pharmacol. 97(6):392-401.

Brinks H, Boucher M, Gao E, Chuprun JK, Pesant S, Raake PW, Huang ZM, Wang X, Qiu G, Gumpert A et al. 2010. Level of g protein-coupled receptor kinase-2 determines myocardial ischemia/reperfusion injury via pro- and anti-apoptotic mechanisms. Circ Res. 107(9):1140-1149.

Brunner F, Andrew P, Wölkart G, Zechner R, Mayer B. 2001. Myocardial contractile function and heart rate in mice with myocyte-specific overexpression of endothelial nitric oxide synthase. Circulation. 104(25):3097-3102.

159

Cannavo A, Liccardo D, Koch WJ. 2013. Targeting cardiac β-adrenergic signaling via grk2 inhibition for heart failure therapy. Front Physiol. 4:264.

Cappuzzello C, Napolitano M, Arcelli D, Melillo G, Melchionna R, Di Vito L, Carlini D, Silvestri L, Brugaletta S, Liuzzo G et al. 2009. Gene expression profiles in peripheral blood mononuclear cells of chronic heart failure patients. Physiol Genomics. 38(3):233-240.

Carman CV, Barak LS, Chen C, Liu-Chen LY, Onorato JJ, Kennedy SP, Caron MG, Benovic JL. 2000. Mutational analysis of gbetagamma and phospholipid interaction with g protein-coupled receptor kinase 2. J Biol Chem. 275(14):10443- 10452.

Carnes CA, Geisbuhler TP, Reiser PJ. 2004. Age-dependent changes in contraction and regional myocardial myosin heavy chain isoform expression in rats. J Appl Physiol (1985). 97(1):446-453.

Carè A, Catalucci D, Felicetti F, Bonci D, Addario A, Gallo P, Bang ML, Segnalini P, Gu Y, Dalton ND et al. 2007. Microrna-133 controls cardiac hypertrophy. Nat Med. 13(5):613-618.

Casey LM, Pistner AR, Belmonte SL, Migdalovich D, Stolpnik O, Nwakanma FE, Vorobiof G, Dunaevsky O, Matavel A, Lopes CM et al. 2010. Small molecule disruption of g beta gamma signaling inhibits the progression of heart failure. Circ Res. 107(4):532-539.

Castaldi A, Zaglia T, Di Mauro V, Carullo P, Viggiani G, Borile G, Di Stefano B, Schiattarella GG, Gualazzi MG, Elia L et al. 2014. Microrna-133 modulates the β1- adrenergic receptor transduction cascade. Circ Res. 115(2):273-283.

Cesselli D, Beltrami AP, D'Aurizio F, Marcon P, Bergamin N, Toffoletto B, Pandolfi M, Puppato E, Marino L, Signore S et al. 2011. Effects of age and heart failure on human cardiac stem cell function. Am J Pathol. 179(1):349-366.

Chapman J, Schutzer WE, Watts VJ, Mader SL. 1999. Impaired cholera toxin relaxation with age in rat aorta. J Gerontol A Biol Sci Med Sci. 54(4):B154-159.

160

Chen JF, Murchison EP, Tang R, Callis TE, Tatsuguchi M, Deng Z, Rojas M, Hammond SM, Schneider MD, Selzman CH et al. 2008a. Targeted deletion of dicer in the heart leads to dilated cardiomyopathy and heart failure. Proc Natl Acad Sci U S A. 105(6):2111-2116.

Chen M, Sato PY, Chuprun JK, Peroutka RJ, Otis NJ, Ibetti J, Pan S, Sheu SS, Gao E, Koch WJ. 2013. Prodeath signaling of g protein-coupled receptor kinase 2 in cardiac myocytes after ischemic stress occurs via extracellular signal-regulated kinase-dependent heat shock protein 90-mediated mitochondrial targeting. Circ Res. 112(8):1121-1134.

Chen Y, Long H, Wu Z, Jiang X, Ma L. 2008b. Egf transregulates opioid receptors through egfr-mediated grk2 phosphorylation and activation. Mol Biol Cell. 19(7):2973- 2983.

Cheng Y, Ji R, Yue J, Yang J, Liu X, Chen H, Dean DB, Zhang C. 2007. Micrornas are aberrantly expressed in hypertrophic heart: Do they play a role in cardiac hypertrophy? Am J Pathol. 170(6):1831-1840.

Chimenti C, Kajstura J, Torella D, Urbanek K, Heleniak H, Colussi C, Di Meglio F, Nadal- Ginard B, Frustaci A, Leri A et al. 2003. Senescence and death of primitive cells and myocytes lead to premature cardiac aging and heart failure. Circ Res. 93(7):604-613.

Cho MC, Rapacciuolo A, Koch WJ, Kobayashi Y, Jones LR, Rockman HA. 1999. Defective beta-adrenergic receptor signaling precedes the development of dilated cardiomyopathy in transgenic mice with calsequestrin overexpression. J Biol Chem. 274(32):22251-22256.

Choi DJ, Koch WJ, Hunter JJ, Rockman HA. 1997. Mechanism of beta-adrenergic receptor desensitization in cardiac hypertrophy is increased beta-adrenergic receptor kinase. J Biol Chem. 272(27):17223-17229.

Ciccarelli M, Sorriento D, Franco A, Fusco A, Del Giudice C, Annunziata R, Cipolletta E, Monti MG, Dorn GW, Trimarco B et al. 2013. Endothelial g protein-coupled receptor kinase 2 regulates vascular homeostasis through the control of free radical oxygen species. Arterioscler Thromb Vasc Biol. 33(10):2415-2424.

161

Cimini M, Garikipati VNS, de Lucia C, Cheng Z, Wang C, Truongcao MM, Lucchese AM, Roy R, Benedict C, Goukassian DA et al. 2019. Podoplanin neutralization improves cardiac remodeling and function after acute myocardial infarction. JCI Insight. 5.

Cohn HI, Harris DM, Pesant S, Pfeiffer M, Zhou RH, Koch WJ, Dorn GW, 2nd, Eckhart AD. 2008. Inhibition of vascular smooth muscle g protein-coupled receptor kinase 2 enhances alpha1d-adrenergic receptor constriction. Am J Physiol Heart Circ Physiol. 295(4):H1695-1704.

Cohn HI, Xi Y, Pesant S, Harris DM, Hyslop T, Falkner B, Eckhart AD. 2009. G protein- coupled receptor kinase 2 expression and activity are associated with blood pressure in black americans. Hypertension. 54(1):71-76.

Collins C, Tzima E. 2011. Hemodynamic forces in endothelial dysfunction and vascular aging. Exp Gerontol. 46(2-3):185-188.

Colman RJ, Anderson RM, Johnson SC, Kastman EK, Kosmatka KJ, Beasley TM, Allison DB, Cruzen C, Simmons HA, Kemnitz JW et al. 2009. Caloric restriction delays disease onset and mortality in rhesus monkeys. Science. 325(5937):201-204.

Cong M, Perry SJ, Lin FT, Fraser ID, Hu LA, Chen W, Pitcher JA, Scott JD, Lefkowitz RJ. 2001. Regulation of membrane targeting of the g protein-coupled receptor kinase 2 by protein kinase a and its anchoring protein akap79. J Biol Chem. 276(18):15192-15199.

Coresh J, Wei GL, McQuillan G, Brancati FL, Levey AS, Jones C, Klag MJ. 2001. Prevalence of high blood pressure and elevated serum creatinine level in the united states: Findings from the third national health and nutrition examination survey (1988-1994). Arch Intern Med. 161(9):1207-1216.

Creemers EE, van Rooij E. 2016. Function and therapeutic potential of noncoding rnas in cardiac fibrosis. Circ Res. 118(1):108-118.

Crimmins E, Vasunilashorn S, Kim JK, Alley D. 2008. Biomarkers related to aging in human populations. Adv Clin Chem. 46:161-216.

162

Dai DF, Santana LF, Vermulst M, Tomazela DM, Emond MJ, MacCoss MJ, Gollahon K, Martin GM, Loeb LA, Ladiges WC et al. 2009. Overexpression of catalase targeted to mitochondria attenuates murine cardiac aging. Circulation. 119(21):2789-2797.

Dai Z, Aoki T, Fukumoto Y, Shimokawa H. 2012. Coronary perivascular fibrosis is associated with impairment of coronary blood flow in patients with non-ischemic heart failure. J Cardiol. 60(5):416-421.

Das S, Ferlito M, Kent OA, Fox-Talbot K, Wang R, Liu D, Raghavachari N, Yang Y, Wheelan SJ, Murphy E et al. 2012. Nuclear mirna regulates the mitochondrial genome in the heart. Circ Res. 110(12):1596-1603.

Das S, Kohr M, Dunkerly-Eyring B, Lee DI, Bedja D, Kent OA, Leung AK, Henao-Mejia J, Flavell RA, Steenbergen C. 2017. Divergent effects of mir-181 family members on myocardial function through protective cytosolic and detrimental mitochondrial microrna targets. J Am Heart Assoc. 6(3).

de Lucia C, Eguchi A, Koch WJ. 2018a. New insights in cardiac beta-adrenergic signaling during heart failure and aging. Front Pharmacol. 9:904.

de Lucia C, Gambino G, Petraglia L, Elia A, Komici K, Femminella GD, D'Amico ML, Formisano R, Borghetti G, Liccardo D et al. 2018b. Long-term caloric restriction improves cardiac function, remodeling, adrenergic responsiveness, and sympathetic innervation in a model of postischemic heart failure. Circ Heart Fail. 11(3):e004153.

de Lucia C, Komici K, Borghetti G, Femminella GD, Bencivenga L, Cannavo A, Corbi G, Ferrara N, Houser SR, Koch WJ et al. 2017. Microrna in cardiovascular aging and age-related cardiovascular diseases. Front Med (Lausanne). 4:74.

de Lucia C, Wallner M, Eaton DM, Zhao H, Houser SR, Koch WJ. 2019. Echocardiographic strain analysis for the early detection of left ventricular systolic/diastolic dysfunction and dyssynchrony in a mouse model of physiological aging. J Gerontol A Biol Sci Med Sci. 74(4):455-461.

163

DebBurman SK, Ptasienski J, Benovic JL, Hosey MM. 1996. G protein-coupled receptor kinase grk2 is a phospholipid-dependent enzyme that can be conditionally activated by g protein betagamma subunits. J Biol Chem. 271(37):22552-22562.

Dick SA, Epelman S. 2016. Chronic heart failure and inflammation: What do we really know? Circ Res. 119(1):159-176.

Divakaran V, Mann DL. 2008. The emerging role of micrornas in cardiac remodeling and heart failure. Circ Res. 103(10):1072-1083.

Djuranovic S, Nahvi A, Green R. 2012. Mirna-mediated gene silencing by translational repression followed by mrna deadenylation and decay. Science. 336(6078):237- 240.

DUNNILL MS. 1959. Histology of the aorta in coarctation. J Pathol Bacteriol. 78:203-207. Dzimiri N, Muiya P, Andres E, Al-Halees Z. 2004. Differential functional expression of human myocardial g protein receptor kinases in left ventricular cardiac diseases. Eur J Pharmacol. 489(3):167-177.

Eckhart AD, Ozaki T, Tevaearai H, Rockman HA, Koch WJ. 2002. Vascular-targeted overexpression of g protein-coupled receptor kinase-2 in transgenic mice attenuates beta-adrenergic receptor signaling and increases resting blood pressure. Mol Pharmacol. 61(4):749-758.

Eghbali M, Robinson TF, Seifter S, Blumenfeld OO. 1989. Collagen accumulation in heart ventricles as a function of growth and aging. Cardiovasc Res. 23(8):723-729.

Eichhorn SW, Guo H, McGeary SE, Rodriguez-Mias RA, Shin C, Baek D, Hsu SH, Ghoshal K, Villén J, Bartel DP. 2014. Mrna destabilization is the dominant effect of mammalian micrornas by the time substantial repression ensues. Mol Cell. 56(1):104-115.

Eisenberg E, Levanon EY. 2013. Human housekeeping genes, revisited. Trends Genet. 29(10):569-574.

164

Elorza A, Penela P, Sarnago S, Mayor F. 2003. Mapk-dependent degradation of g protein- coupled receptor kinase 2. J Biol Chem. 278(31):29164-29173.

Elorza A, Sarnago S, Mayor F. 2000. Agonist-dependent modulation of g protein-coupled receptor kinase 2 by mitogen-activated protein kinases. Mol Pharmacol. 57(4):778- 783.

Elrod JW, Greer JJ, Bryan NS, Langston W, Szot JF, Gebregzlabher H, Janssens S, Feelisch M, Lefer DJ. 2006. Cardiomyocyte-specific overexpression of no synthase-3 protects against myocardial ischemia-reperfusion injury. Arterioscler Thromb Vasc Biol. 26(7):1517-1523.

Eulalio A, Huntzinger E, Izaurralde E. 2008. Gw182 interaction with argonaute is essential for mirna-mediated translational repression and mrna decay. Nat Struct Mol Biol. 15(4):346-353.

Eulalio A, Huntzinger E, Nishihara T, Rehwinkel J, Fauser M, Izaurralde E. 2009. Deadenylation is a widespread effect of mirna regulation. RNA. 15(1):21-32.

Eulalio A, Rehwinkel J, Stricker M, Huntzinger E, Yang SF, Doerks T, Dorner S, Bork P, Boutros M, Izaurralde E. 2007. Target-specific requirements for enhancers of decapping in mirna-mediated gene silencing. Genes Dev. 21(20):2558-2570.

Fabian MR, Sonenberg N. 2012. The mechanics of mirna-mediated gene silencing: A look under the hood of mirisc. Nat Struct Mol Biol. 19(6):586-593.

Feng HJ, Ouyang W, Liu JH, Sun YG, Hu R, Huang LH, Xian JL, Jing CF, Zhou MJ. 2014. Global microrna profiles and signaling pathways in the development of cardiac hypertrophy. Braz J Med Biol Res. 47(5):361-368.

Ferrara N, Komici K, Corbi G, Pagano G, Furgi G, Rengo C, Femminella GD, Leosco D, Bonaduce D. 2014. Beta-adrenergic receptor responsiveness in aging heart and clinical implications. Front Physiol. 4:396.

165

Fischer S, Handrick R, Aschrafi A, Otte K. 2015. Unveiling the principle of microrna- mediated redundancy in cellular pathway regulation. RNA Biol. 12(3):238-247.

Fontana L, Partridge L, Longo VD. 2010. Extending healthy life span--from yeast to humans. Science. 328(5976):321-326.

Freedman NJ, Liggett SB, Drachman DE, Pei G, Caron MG, Lefkowitz RJ. 1995. Phosphorylation and desensitization of the human beta 1-adrenergic receptor. Involvement of g protein-coupled receptor kinases and camp-dependent protein kinase. J Biol Chem. 270(30):17953-17961.

Friedman RC, Farh KK, Burge CB, Bartel DP. 2009. Most mammalian mrnas are conserved targets of micrornas. Genome Res. 19(1):92-105.

Froehlich JP, Lakatta EG, Beard E, Spurgeon HA, Weisfeldt ML, Gerstenblith G. 1978. Studies of sarcoplasmic reticulum function and contraction duration in young adult and aged rat myocardium. J Mol Cell Cardiol. 10(5):427-438.

Fulton D, Gratton JP, McCabe TJ, Fontana J, Fujio Y, Walsh K, Franke TF, Papapetropoulos A, Sessa WC. 1999. Regulation of endothelium-derived nitric oxide production by the protein kinase akt. Nature. 399(6736):597-601.

Furchgott RF, Zawadzki JV. 1980. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature. 288(5789):373-376.

Gao E, Lei YH, Shang X, Huang ZM, Zuo L, Boucher M, Fan Q, Chuprun JK, Ma XL, Koch WJ. 2010. A novel and efficient model of coronary artery ligation and myocardial infarction in the mouse. Circ Res. 107(12):1445-1453.

Gazoti Debessa CR, Mesiano Maifrino LB, Rodrigues de Souza R. 2001. Age related changes of the collagen network of the human heart. Mech Ageing Dev. 122(10):1049-1058.

166

Geng X, Hwang J, Ye J, Shih H, Coulter B, Naudin C, Jun K, Sievers R, Yeghiazarians Y, Lee RJ et al. 2017. Aging is protective against pressure overload cardiomyopathy via adaptive extracellular matrix remodeling. Am J Cardiovasc Dis. 7(3):72-82.

Goldspink DF, Burniston JG, Tan LB. 2003. Cardiomyocyte death and the ageing and failing heart. Exp Physiol. 88(3):447-458.

Goldstrohm AC, Wickens M. 2008. Multifunctional deadenylase complexes diversify mrna control. Nat Rev Mol Cell Biol. 9(4):337-344.

Greer EL, Maures TJ, Hauswirth AG, Green EM, Leeman DS, Maro GS, Han S, Banko MR, Gozani O, Brunet A. 2010. Members of the h3k4 trimethylation complex regulate lifespan in a germline-dependent manner in c. Elegans. Nature. 466(7304):383-387.

Grimson A, Farh KK, Johnston WK, Garrett-Engele P, Lim LP, Bartel DP. 2007. Microrna targeting specificity in mammals: Determinants beyond seed pairing. Mol Cell. 27(1):91-105.

Grishok A, Pasquinelli AE, Conte D, Li N, Parrish S, Ha I, Baillie DL, Fire A, Ruvkun G, Mello CC. 2001. Genes and mechanisms related to rna interference regulate expression of the small temporal rnas that control c. Elegans developmental timing. Cell. 106(1):23-34.

Groen BB, Hamer HM, Snijders T, van Kranenburg J, Frijns D, Vink H, van Loon LJ. 2014. Skeletal muscle capillary density and microvascular function are compromised with aging and type 2 diabetes. J Appl Physiol (1985). 116(8):998- 1005.

Gros R, Benovic JL, Tan CM, Feldman RD. 1997. G-protein-coupled receptor kinase activity is increased in hypertension. J Clin Invest. 99(9):2087-2093.

Guo H, Ingolia NT, Weissman JS, Bartel DP. 2010. Mammalian micrornas predominantly act to decrease target mrna levels. Nature. 466(7308):835-840.

167

GUPTA TC, WIGGERS CJ. 1951. Basic hemodynamic changes produced by aortic coarctation of different degrees. Circulation. 3(1):17-31.

Gyurko R, Kuhlencordt P, Fishman MC, Huang PL. 2000. Modulation of mouse cardiac function in vivo by enos and anp. Am J Physiol Heart Circ Physiol. 278(3):H971- 981.

Hall MJ, Levant S, DeFrances CJ. 2012. Hospitalization for congestive heart failure: United states, 2000-2010. NCHS Data Brief. (108):1-8.

Hammond SM, Boettcher S, Caudy AA, Kobayashi R, Hannon GJ. 2001. Argonaute2, a link between genetic and biochemical analyses of rnai. Science. 293(5532):1146- 1150.

Hanna J, Hossain GS, Kocerha J. 2019. The potential for microrna therapeutics and clinical research. Front Genet. 10:478.

Harding VB, Jones LR, Lefkowitz RJ, Koch WJ, Rockman HA. 2001. Cardiac beta ark1 inhibition prolongs survival and augments beta blocker therapy in a mouse model of severe heart failure. Proc Natl Acad Sci U S A. 98(10):5809-5814.

Harrison DE, Strong R, Sharp ZD, Nelson JF, Astle CM, Flurkey K, Nadon NL, Wilkinson JE, Frenkel K, Carter CS et al. 2009. Rapamycin fed late in life extends lifespan in genetically heterogeneous mice. Nature. 460(7253):392-395.

Hayashi H, Hess DT, Zhang R, Sugi K, Gao H, Tan BL, Bowles DE, Milano CA, Jain MK, Koch WJ et al. 2018. S-nitrosylation of β-arrestins biases receptor signaling and confers ligand independence. Mol Cell. 70(3):473-487.e476.

He C, Li Z, Chen P, Huang H, Hurst LD, Chen J. 2012. Young intragenic mirnas are less coexpressed with host genes than old ones: Implications of mirna-host gene coevolution. Nucleic Acids Res. 40(9):4002-4012.

168

He L, He X, Lim LP, de Stanchina E, Xuan Z, Liang Y, Xue W, Zender L, Magnus J, Ridzon D et al. 2007. A microrna component of the p53 tumour suppressor network. Nature. 447(7148):1130-1134.

Henao-Mejia J, Williams A, Goff LA, Staron M, Licona-Limón P, Kaech SM, Nakayama M, Rinn JL, Flavell RA. 2013. The microrna mir-181 is a critical cellular metabolic rheostat essential for nkt cell ontogenesis and lymphocyte development and homeostasis. Immunity. 38(5):984-997.

Hendrickson DG, Hogan DJ, McCullough HL, Myers JW, Herschlag D, Ferrell JE, Brown PO. 2009. Concordant regulation of translation and mrna abundance for hundreds of targets of a human microrna. PLoS Biol. 7(11):e1000238.

Henning RH, Brundel BJJM. 2017. Proteostasis in cardiac health and disease. Nat Rev Cardiol. 14(11):637-653.

Heron M. 2019. Deaths: Leading causes for 2017. National Vital Statistics System: Centers for Disease Control and Prevention.

Herranz D, Muñoz-Martin M, Cañamero M, Mulero F, Martinez-Pastor B, Fernandez- Capetillo O, Serrano M. 2010. Sirt1 improves healthy ageing and protects from metabolic syndrome-associated cancer. Nat Commun. 1:3.

Hori D, Dunkerly-Eyring B, Nomura Y, Biswas D, Steppan J, Henao-Mejia J, Adachi H, Santhanam L, Berkowitz DE, Steenbergen C et al. 2017. Mir-181b regulates vascular stiffness age dependently in part by regulating tgf-β signaling. PLoS One. 12(3):e0174108.

Hsu CP, Zhai P, Yamamoto T, Maejima Y, Matsushima S, Hariharan N, Shao D, Takagi H, Oka S, Sadoshima J. 2010. Silent information regulator 1 protects the heart from ischemia/reperfusion. Circulation. 122(21):2170-2182.

Hu M, Wang C, Li W, Lu W, Bai Z, Qin D, Yan Q, Zhu J, Krueger BJ, Renne R et al. 2015. A kshv microrna directly targets g protein-coupled receptor kinase 2 to promote the migration and invasion of endothelial cells by inducing cxcr2 and activating akt signaling. PLoS Pathog. 11(9):e1005171.

169

Huang PL, Huang Z, Mashimo H, Bloch KD, Moskowitz MA, Bevan JA, Fishman MC. 1995. Hypertension in mice lacking the gene for endothelial nitric oxide synthase. Nature. 377(6546):239-242.

Huang ZM, Gao E, Fonseca FV, Hayashi H, Shang X, Hoffman NE, Chuprun JK, Tian X, Tilley DG, Madesh M et al. 2013. Convergence of g protein-coupled receptor and s-nitrosylation signaling determines the outcome to cardiac ischemic injury. Sci Signal. 6(299):ra95.

Humphreys DT, Westman BJ, Martin DI, Preiss T. 2005. Micrornas control translation initiation by inhibiting eukaryotic initiation factor 4e/cap and poly(a) tail function. Proc Natl Acad Sci U S A. 102(47):16961-16966.

Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS, Ganiats TG, Jessup M, Konstam MA, Mancini DM, Michl K et al. 2009. 2009 focused update incorporated into the acc/aha 2005 guidelines for the diagnosis and management of heart failure in adults: A report of the american college of cardiology foundation/american heart association task force on practice guidelines: Developed in collaboration with the international society for heart and lung transplantation. Circulation. 119(14):e391- 479.

Huntzinger E, Izaurralde E. 2011. Gene silencing by micrornas: Contributions of translational repression and mrna decay. Nat Rev Genet. 12(2):99-110.

Hutvagner G, Simard MJ. 2008. Argonaute proteins: Key players in rna silencing. Nat Rev Mol Cell Biol. 9(1):22-32.

Hutvágner G, McLachlan J, Pasquinelli AE, Bálint E, Tuschl T, Zamore PD. 2001. A cellular function for the rna-interference enzyme dicer in the maturation of the let- 7 small temporal rna. Science. 293(5531):834-838.

Iaccarino G, Barbato E, Cipolletta E, De Amicis V, Margulies KB, Leosco D, Trimarco B, Koch WJ. 2005. Elevated myocardial and lymphocyte grk2 expression and activity in human heart failure. Eur Heart J. 26(17):1752-1758.

170

Iaccarino G, Dolber PC, Lefkowitz RJ, Koch WJ. 1999. Bbeta-adrenergic receptor kinase- 1 levels in catecholamine-induced myocardial hypertrophy: Regulation by beta- but not alpha1-adrenergic stimulation. Hypertension. 33(1 Pt 2):396-401.

Iaccarino G, Tomhave ED, Lefkowitz RJ, Koch WJ. 1998. Reciprocal in vivo regulation of myocardial g protein-coupled receptor kinase expression by beta-adrenergic receptor stimulation and blockade. Circulation. 98(17):1783-1789.

Ignarro LJ, Buga GM, Wood KS, Byrns RE, Chaudhuri G. 1987. Endothelium-derived relaxing factor produced and released from artery and vein is nitric oxide. Proc Natl Acad Sci U S A. 84(24):9265-9269.

Ikeda S, Kong SW, Lu J, Bisping E, Zhang H, Allen PD, Golub TR, Pieske B, Pu WT. 2007. Altered microrna expression in human heart disease. Physiol Genomics. 31(3):367-373.

Inglese J, Freedman NJ, Koch WJ, Lefkowitz RJ. 1993. Structure and mechanism of the g protein-coupled receptor kinases. J Biol Chem. 268(32):23735-23738.

Inglese J, Koch WJ, Caron MG, Lefkowitz RJ. 1992. Isoprenylation in regulation of signal transduction by g-protein-coupled receptor kinases. Nature. 359(6391):147-150.

Isoyama S, Wei JY, Izumo S, Fort P, Schoen FJ, Grossman W. 1987. Effect of age on the development of cardiac hypertrophy produced by aortic constriction in the rat. Circ Res. 61(3):337-345.

Iwanaga Y, Aoyama T, Kihara Y, Onozawa Y, Yoneda T, Sasayama S. 2002. Excessive activation of matrix metalloproteinases coincides with left ventricular remodeling during transition from hypertrophy to heart failure in hypertensive rats. J Am Coll Cardiol. 39(8):1384-1391.

Izzo R, Cipolletta E, Ciccarelli M, Campanile A, Santulli G, Palumbo G, Vasta A, Formisano S, Trimarco B, Iaccarino G. 2008. Enhanced grk2 expression and desensitization of betaar vasodilatation in hypertensive patients. Clin Transl Sci. 1(3):215-220.

171

Jaber M, Koch WJ, Rockman H, Smith B, Bond RA, Sulik KK, Ross J, Lefkowitz RJ, Caron MG, Giros B. 1996. Essential role of beta-adrenergic receptor kinase 1 in cardiac development and function. Proc Natl Acad Sci U S A. 93(23):12974-12979.

Jackson RJ, Hellen CU, Pestova TV. 2010. The mechanism of eukaryotic translation initiation and principles of its regulation. Nat Rev Mol Cell Biol. 11(2):113-127.

Janicki JS, Brower GL. 2002. The role of myocardial fibrillar collagen in ventricular remodeling and function. J Card Fail. 8(6 Suppl):S319-325.

Janssens S, Pokreisz P, Schoonjans L, Pellens M, Vermeersch P, Tjwa M, Jans P, Scherrer- Crosbie M, Picard MH, Szelid Z et al. 2004. Cardiomyocyte-specific overexpression of nitric oxide synthase 3 improves left ventricular performance and reduces compensatory hypertrophy after myocardial infarction. Circ Res. 94(9):1256-1262.

Janzen V, Forkert R, Fleming HE, Saito Y, Waring MT, Dombkowski DM, Cheng T, DePinho RA, Sharpless NE, Scadden DT. 2006. Stem-cell ageing modified by the cyclin-dependent kinase inhibitor p16ink4a. Nature. 443(7110):421-426.

Jeandard D, Smirnova A, Tarassov I, Barrey E, Smirnov A, Entelis N. 2019. Import of noncoding rnas into human mitochondria: A critical review and emerging approaches. Cells. 8(3).

Ji J, Yamashita T, Budhu A, Forgues M, Jia HL, Li C, Deng C, Wauthier E, Reid LM, Ye QH et al. 2009. Identification of microrna-181 by genome-wide screening as a critical player in epcam-positive hepatic cancer stem cells. Hepatology. 50(2):472- 480.

Jiang X, Benovic JL, Wedegaertner PB. 2007. Plasma membrane and nuclear localization of g protein coupled receptor kinase 6a. Mol Biol Cell. 18(8):2960-2969.

Johnson G, Tsao PS, Mulloy D, Lefer AM. 1990. Cardioprotective effects of acidified sodium nitrite in myocardial ischemia with reperfusion. J Pharmacol Exp Ther. 252(1):35-41.

172

Johnson LR, Robinson JD, Lester KN, Pitcher JA. 2013. Distinct structural features of g protein-coupled receptor kinase 5 (grk5) regulate its nuclear localization and dna- binding ability. PLoS One. 8(5):e62508.

Johnson LR, Scott MG, Pitcher JA. 2004. G protein-coupled receptor kinase 5 contains a dna-binding nuclear localization sequence. Mol Cell Biol. 24(23):10169-10179.

Jonas S, Izaurralde E. 2015. Towards a molecular understanding of microrna-mediated gene silencing. Nat Rev Genet. 16(7):421-433.

Kajstura J, Cheng W, Sarangarajan R, Li P, Li B, Nitahara JA, Chapnick S, Reiss K, Olivetti G, Anversa P. 1996. Necrotic and apoptotic myocyte cell death in the aging heart of fischer 344 rats. Am J Physiol. 271(3 Pt 2):H1215-1228.

Kamal FA, Mickelsen DM, Wegman KM, Travers JG, Moalem J, Hammes SR, Smrcka AV, Blaxall BC. 2014. Simultaneous adrenal and cardiac g-protein-coupled receptor-gbetagamma inhibition halts heart failure progression. J Am Coll Cardiol. 63(23):2549-2557.

Katz MG, Fargnoli AS, Swain JD, Tomasulo CE, Ciccarelli M, Huang ZM, Rabinowitz JE, Bridges CR. 2012. Aav6-βarkct gene delivery mediated by molecular cardiac surgery with recirculating delivery (mcard) in sheep results in robust gene expression and increased adrenergic reserve. J Thorac Cardiovasc Surg. 143(3):720-726.e723.

Kawakami M, Hattori M, Ohashi W, Fujimori T, Hattori K, Takebe M, Tomita K, Yokoo H, Matsuda N, Yamazaki M et al. 2018. Role of g protein-coupled receptor kinase 2 in oxidative and nitrosative stress-related neurohistopathological changes in a mouse model of sepsis-associated encephalopathy. J Neurochem. 145(6):474-488.

Khvorova A, Reynolds A, Jayasena SD. 2003. Functional sirnas and mirnas exhibit strand bias. Cell. 115(2):209-216.

Kida Y, Goligorsky MS. 2016. Sirtuins, cell senescence, and vascular aging. Can J Cardiol. 32(5):634-641.

173

Kim DH, Saetrom P, Snøve O, Rossi JJ. 2008. Microrna-directed transcriptional gene silencing in mammalian cells. Proc Natl Acad Sci U S A. 105(42):16230-16235.

Kim IM, Wang Y, Park KM, Tang Y, Teoh JP, Vinson J, Traynham CJ, Pironti G, Mao L, Su H et al. 2014. Β-arrestin1-biased β1-adrenergic receptor signaling regulates microrna processing. Circ Res. 114(5):833-844.

Kim VN. 2005. Microrna biogenesis: Coordinated cropping and dicing. Nat Rev Mol Cell Biol. 6(5):376-385.

Kinch LN, Grishin NV. 2009. The human ago2 mc region does not contain an eif4e-like mrna cap binding motif. Biol Direct. 4:2.

Kiriakidou M, Tan GS, Lamprinaki S, De Planell-Saguer M, Nelson PT, Mourelatos Z. 2007. An mrna m7g cap binding-like motif within human ago2 represses translation. Cell. 129(6):1141-1151.

Klenke S, Engler A, Ecker D, Ochsenfarth C, Danowski N, Peters J, Siffert W, Frey UH. 2018. The grk2 promoter is regulated by early-growth response transcription factor egr-1. Basic Clin Pharmacol Toxicol. 123(6):660-669.

Koch WJ, Rockman HA, Samama P, Hamilton RA, Bond RA, Milano CA, Lefkowitz RJ. 1995. Cardiac function in mice overexpressing the beta-adrenergic receptor kinase or a beta ark inhibitor. Science. 268(5215):1350-1353.

Koga H, Kaushik S, Cuervo AM. 2011. Protein homeostasis and aging: The importance of exquisite quality control. Ageing Res Rev. 10(2):205-215.

Kohr MJ, Murphy E, Steenbergen C. 2014. Glyceraldehyde-3-phosphate dehydrogenase acts as a mitochondrial trans-s-nitrosylase in the heart. PLoS One. 9(10):e111448.

Kong YW, Cannell IG, de Moor CH, Hill K, Garside PG, Hamilton TL, Meijer HA, Dobbyn HC, Stoneley M, Spriggs KA et al. 2008. The mechanism of micro-rna- mediated translation repression is determined by the promoter of the target gene. Proc Natl Acad Sci U S A. 105(26):8866-8871.

174

Krasel C, Dammeier S, Winstel R, Brockmann J, Mischak H, Lohse MJ. 2001. Phosphorylation of grk2 by protein kinase c abolishes its inhibition by calmodulin. J Biol Chem. 276(3):1911-1915.

Krishnamurthy J, Torrice C, Ramsey MR, Kovalev GI, Al-Regaiey K, Su L, Sharpless NE. 2004. Ink4a/arf expression is a biomarker of aging. J Clin Invest. 114(9):1299- 1307.

Kristen AV, Ajroud-Driss S, Conceição I, Gorevic P, Kyriakides T, Obici L. 2019. Patisiran, an rnai therapeutic for the treatment of hereditary transthyretin-mediated amyloidosis. Neurodegener Dis Manag. 9(1):5-23.

Lakatta EG. 1994. Cardiovascular reserve capacity in healthy older humans. Aging (Milano). 6(4):213-223.

Lakatta EG. 2015. So! What's aging? Is cardiovascular aging a disease? J Mol Cell Cardiol. 83:1-13.

Lakatta EG. 2018. The reality of getting old. Nat Rev Cardiol. 15(9):499-500.

Lam JK, Chow MY, Zhang Y, Leung SW. 2015. Sirna versus mirna as therapeutics for gene silencing. Mol Ther Nucleic Acids. 4:e252.

Lamming DW, Ye L, Katajisto P, Goncalves MD, Saitoh M, Stevens DM, Davis JG, Salmon AB, Richardson A, Ahima RS et al. 2012. Rapamycin-induced insulin resistance is mediated by mtorc2 loss and uncoupled from longevity. Science. 335(6076):1638-1643.

Lee SH, Lee JH, Lee HY, Min KJ. 2019. Sirtuin signaling in cellular senescence and aging. BMB Rep. 52(1):24-34.

Lee Y, Kim M, Han J, Yeom KH, Lee S, Baek SH, Kim VN. 2004. Microrna genes are transcribed by rna polymerase ii. EMBO J. 23(20):4051-4060.

175

Lefer DJ, Nakanishi K, Johnston WE, Vinten-Johansen J. 1993. Antineutrophil and myocardial protecting actions of a novel nitric oxide donor after acute myocardial ischemia and reperfusion of dogs. Circulation. 88(5 Pt 1):2337-2350.

Lefkowitz RJ. 1993. G protein-coupled receptor kinases. Cell. 74(3):409-412.

Leineweber K, Klapproth S, Beilfuss A, Silber RE, Heusch G, Philipp T, Brodde OE. 2003. Unchanged g-protein-coupled receptor kinase activity in the aging human heart. J Am Coll Cardiol. 42(8):1487-1492.

Leosco D, Iaccarino G, Cipolletta E, De Santis D, Pisani E, Trimarco V, Ferrara N, Abete P, Sorriento D, Rengo F et al. 2003. Exercise restores beta-adrenergic vasorelaxation in aged rat carotid arteries. Am J Physiol Heart Circ Physiol. 285(1):H369-374.

Lewis BP, Burge CB, Bartel DP. 2005. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microrna targets. Cell. 120(1):15-20.

Li AL, Lv JB, Gao L. 2017. Mir-181a mediates ang ii-induced myocardial hypertrophy by mediating autophagy. Eur Rev Med Pharmacol Sci. 21(23):5462-5470.

Li W, Jia X, Shen C, Zhang M, Xu J, Shang Y, Zhu K, Hu M, Yan Q, Qin D et al. 2016. A kshv microrna enhances viral latency and induces angiogenesis by targeting grk2 to activate the cxcr2/akt pathway. Oncotarget. 7(22):32286-32305.

Li Y, Choi WJ, Wei W, Song S, Zhang Q, Liu J, Wang RK. 2018. Aging-associated changes in cerebral vasculature and blood flow as determined by quantitative optical coherence tomography angiography. Neurobiol Aging. 70:148-159.

Lieu M, Koch WJ. 2019. Grk2 and grk5 as therapeutic targets and their role in maladaptive and pathological cardiac hypertrophy. Expert Opin Ther Targets. 23(3):201-214.

Lieu M, Traynham CJ, de Lucia C, Pfleger J, Piedepalumbo M, Roy R, Petovic J, Landesberg G, Forrester SJ, Hoffman M et al. 2020. Loss of dynamic regulation of

176

g protein-coupled receptor kinase 2 by nitric oxide leads to cardiovascular dysfunction with aging. Am J Physiol Heart Circ Physiol.

Lim LP, Lau NC, Garrett-Engele P, Grimson A, Schelter JM, Castle J, Bartel DP, Linsley PS, Johnson JM. 2005. Microarray analysis shows that some micrornas downregulate large numbers of target mrnas. Nature. 433(7027):769-773.

Lima B, Forrester MT, Hess DT, Stamler JS. 2010. S-nitrosylation in cardiovascular signaling. Circ Res. 106(4):633-646.

Liu G, Li Y, Gao XG. 2016. Microrna-181a is upregulated in human atherosclerosis plaques and involves in the oxidative stress-induced endothelial cell dysfunction through direct targeting bcl-2. Eur Rev Med Pharmacol Sci. 20(14):3092-3100.

Liu M, Li XC, Lu L, Cao Y, Sun RR, Chen S, Zhang PY. 2014. Cardiovascular disease and its relationship with chronic kidney disease. Eur Rev Med Pharmacol Sci. 18(19):2918-2926.

Liu N, Bezprozvannaya S, Williams AH, Qi X, Richardson JA, Bassel-Duby R, Olson EN. 2008. Microrna-133a regulates cardiomyocyte proliferation and suppresses smooth muscle gene expression in the heart. Genes Dev. 22(23):3242-3254.

Liu S, Premont RT, Kontos CD, Zhu S, Rockey DC. 2005. A crucial role for grk2 in regulation of endothelial cell nitric oxide synthase function in portal hypertension. Nat Med. 11(9):952-958.

Liu Y, Bloom SI, Donato AJ. 2019. The role of senescence, telomere dysfunction and shelterin in vascular aging. Microcirculation. 26(2):e12487.

Luchsinger BP, Rich EN, Gow AJ, Williams EM, Stamler JS, Singel DJ. 2003. Routes to s-nitroso-hemoglobin formation with heme redox and preferential reactivity in the beta subunits. Proc Natl Acad Sci U S A. 100(2):461-466.

Lund E, Güttinger S, Calado A, Dahlberg JE, Kutay U. 2004. Nuclear export of microrna precursors. Science. 303(5654):95-98.

177

Luo J, Benovic JL. 2003. G protein-coupled receptor kinase interaction with hsp90 mediates kinase maturation. J Biol Chem. 278(51):50908-50914.

Lymperopoulos A, Rengo G, Funakoshi H, Eckhart AD, Koch WJ. 2007. Adrenal grk2 upregulation mediates sympathetic overdrive in heart failure. Nat Med. 13(3):315- 323.

Lymperopoulos A, Rengo G, Koch WJ. 2013. Adrenergic nervous system in heart failure: Pathophysiology and therapy. Circ Res. 113(6):739-753.

López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G. 2013. The hallmarks of aging. Cell. 153(6):1194-1217.

Ma E, MacRae IJ, Kirsch JF, Doudna JA. 2008. Autoinhibition of human dicer by its internal helicase domain. J Mol Biol. 380(1):237-243.

Magalhães PJ, Andreu AL, Schon EA. 1998. Evidence for the presence of 5s rrna in mammalian mitochondria. Mol Biol Cell. 9(9):2375-2382.

Maizel J, Xavier S, Chen J, Lin CH, Vasko R, Goligorsky MS. 2014. Sirtuin 1 ablation in endothelial cells is associated with impaired angiogenesis and diastolic dysfunction. Am J Physiol Heart Circ Physiol. 307(12):H1691-1704.

Mallat Y, Tritsch E, Ladouce R, Winter DL, Friguet B, Li Z, Mericskay M. 2014. Proteome modulation in h9c2 cardiac cells by micrornas mir-378 and mir-378. Mol Cell Proteomics. 13(1):18-29.

Marques FZ, Campain AE, Tomaszewski M, Zukowska-Szczechowska E, Yang YH, Charchar FJ, Morris BJ. 2011. Gene expression profiling reveals renin mrna overexpression in human hypertensive kidneys and a role for micrornas. Hypertension. 58(6):1093-1098.

Marques FZ, Vizi D, Khammy O, Mariani JA, Kaye DM. 2016. The transcardiac gradient of cardio-micrornas in the failing heart. Eur J Heart Fail. 18(8):1000-1008.

178

Martinez J, Patkaniowska A, Urlaub H, Lührmann R, Tuschl T. 2002. Single-stranded antisense sirnas guide target rna cleavage in rnai. Cell. 110(5):563-574.

Martini JS, Raake P, Vinge LE, DeGeorge BR, DeGeorge B, Chuprun JK, Harris DM, Gao E, Eckhart AD, Pitcher JA et al. 2008. Uncovering g protein-coupled receptor kinase-5 as a histone deacetylase kinase in the nucleus of cardiomyocytes. Proc Natl Acad Sci U S A. 105(34):12457-12462.

Massion PB, Balligand JL. 2003. Modulation of cardiac contraction, relaxation and rate by the endothelial nitric oxide synthase (enos): Lessons from genetically modified mice. J Physiol. 546(Pt 1):63-75.

Massion PB, Feron O, Dessy C, Balligand JL. 2003. Nitric oxide and cardiac function: Ten years after, and continuing. Circ Res. 93(5):388-398.

Mathar I, Vennekens R, Meissner M, Kees F, Van der Mieren G, Camacho Londoño JE, Uhl S, Voets T, Hummel B, van den Bergh A et al. 2010. Increased catecholamine secretion contributes to hypertension in trpm4-deficient mice. J Clin Invest. 120(9):3267-3279.

Mathonnet G, Fabian MR, Svitkin YV, Parsyan A, Huck L, Murata T, Biffo S, Merrick WC, Darzynkiewicz E, Pillai RS et al. 2007. Microrna inhibition of translation initiation in vitro by targeting the cap-binding complex eif4f. Science. 317(5845):1764-1767.

Matkovich SJ, Diwan A, Klanke JL, Hammer DJ, Marreez Y, Odley AM, Brunskill EW, Koch WJ, Schwartz RJ, Dorn GW. 2006. Cardiac-specific ablation of g-protein receptor kinase 2 redefines its roles in heart development and beta-adrenergic signaling. Circ Res. 99(9):996-1003.

Matkovich SJ, Van Booven DJ, Youker KA, Torre-Amione G, Diwan A, Eschenbacher WH, Dorn LE, Watson MA, Margulies KB, Dorn GW. 2009. Reciprocal regulation of myocardial micrornas and messenger rna in human cardiomyopathy and reversal of the microrna signature by biomechanical support. Circulation. 119(9):1263- 1271.

179

Matkovich SJ, Wang W, Tu Y, Eschenbacher WH, Dorn LE, Condorelli G, Diwan A, Nerbonne JM, Dorn GW. 2010. Microrna-133a protects against myocardial fibrosis and modulates electrical repolarization without affecting hypertrophy in pressure- overloaded adult hearts. Circ Res. 106(1):166-175.

Mattace-Raso FU, van der Cammen TJ, Hofman A, van Popele NM, Bos ML, Schalekamp MA, Asmar R, Reneman RS, Hoeks AP, Breteler MM et al. 2006. Arterial stiffness and risk of coronary heart disease and stroke: The rotterdam study. Circulation. 113(5):657-663.

Maurice JP, Shah AS, Kypson AP, Hata JA, White DC, Glower DD, Koch WJ. 1999. Molecular beta-adrenergic signaling abnormalities in failing rabbit hearts after infarction. Am J Physiol. 276(6):H1853-1860.

Morozova N, Zinovyev A, Nonne N, Pritchard LL, Gorban AN, Harel-Bellan A. 2012. Kinetic signatures of microrna modes of action. RNA. 18(9):1635-1655.

Mourino-Alvarez L, Baldan-Martin M, Sastre-Oliva T, Martin-Lorenzo M, Maroto AS, Corbacho-Alonso N, Rincon R, Martin-Rojas T, Lopez-Almodovar LF, Alvarez- Llamas G et al. 2018. A comprehensive study of calcific aortic stenosis: From rabbit to human samples. Dis Model Mech. 11(6).

Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, de Ferranti S, Despres JP, Fullerton HJ, Howard VJ et al. 2015. Heart disease and stroke statistics- -2015 update: A report from the american heart association. Circulation. 131(4):e29-322.

Mukherji S, Ebert MS, Zheng GX, Tsang JS, Sharp PA, van Oudenaarden A. 2011. Micrornas can generate thresholds in target gene expression. Nat Genet. 43(9):854- 859.

Muller-Delp JM, Spier SA, Ramsey MW, Delp MD. 2002. Aging impairs endothelium- dependent vasodilation in rat skeletal muscle arterioles. Am J Physiol Heart Circ Physiol. 283(4):H1662-1672.

180

Mungrue IN, Gros R, You X, Pirani A, Azad A, Csont T, Schulz R, Butany J, Stewart DJ, Husain M. 2002. Cardiomyocyte overexpression of inos in mice results in peroxynitrite generation, heart block, and sudden death. J Clin Invest. 109(6):735- 743.

Murphy E, Steenbergen C. 2007. Preconditioning: The mitochondrial connection. Annu Rev Physiol. 69:51-67.

Naga Prasad SV, Gupta MK, Duan ZH, Surampudi VS, Liu CG, Kotwal A, Moravec CS, Starling RC, Perez DM, Sen S et al. 2017. A unique microrna profile in end-stage heart failure indicates alterations in specific cardiovascular signaling networks. PLoS One. 12(3):e0170456.

Nakayama H, Nishida K, Otsu K. 2016. Macromolecular degradation systems and cardiovascular aging. Circ Res. 118(10):1577-1592.

Naseem KM. 2005. The role of nitric oxide in cardiovascular diseases. Mol Aspects Med. 26(1-2):33-65.

Noren Hooten N, Fitzpatrick M, Wood WH, De S, Ejiogu N, Zhang Y, Mattison JA, Becker KG, Zonderman AB, Evans MK. 2013. Age-related changes in microrna levels in serum. Aging (Albany NY). 5(10):725-740.

North BJ, Sinclair DA. 2012. The intersection between aging and cardiovascular disease. Circ Res. 110(8):1097-1108.

Nottrott S, Simard MJ, Richter JD. 2006. Human let-7a mirna blocks protein production on actively translating polyribosomes. Nat Struct Mol Biol. 13(12):1108-1114.

Nouraee N, Mowla SJ. 2015. Mirna therapeutics in cardiovascular diseases: Promises and problems. Front Genet. 6:232.

Nykänen A, Haley B, Zamore PD. 2001. Atp requirements and small interfering rna structure in the rna interference pathway. Cell. 107(3):309-321.

181

O'Donnell KA, Wentzel EA, Zeller KI, Dang CV, Mendell JT. 2005. C-myc-regulated micrornas modulate e2f1 expression. Nature. 435(7043):839-843.

O'Neill L, Holbrook NJ, Fargnoli J, Lakatta EG. 1991. Progressive changes from young adult age to senescence in mrna for rat cardiac myosin heavy chain genes. Cardioscience. 2(1):1-5.

Obrenovich ME, Smith MA, Siedlak SL, Chen SG, de la Torre JC, Perry G, Aliev G. 2006. Overexpression of grk2 in alzheimer disease and in a chronic hypoperfusion rat model is an early marker of brain mitochondrial lesions. Neurotox Res. 10(1):43- 56.

Oh H, Wang SC, Prahash A, Sano M, Moravec CS, Taffet GE, Michael LH, Youker KA, Entman ML, Schneider MD. 2003. Telomere attrition and chk2 activation in human heart failure. Proc Natl Acad Sci U S A. 100(9):5378-5383.

Oliveira-Carvalho V, da Silva MM, Guimarães GV, Bacal F, Bocchi EA. 2013. Micrornas: New players in heart failure. Mol Biol Rep. 40(3):2663-2670.

Olivetti G, Melissari M, Capasso JM, Anversa P. 1991. Cardiomyopathy of the aging human heart. Myocyte loss and reactive cellular hypertrophy. Circ Res. 68(6):1560- 1568.

Olsen PH, Ambros V. 1999. The lin-4 regulatory rna controls developmental timing in caenorhabditis elegans by blocking lin-14 protein synthesis after the initiation of translation. Dev Biol. 216(2):671-680.

Orlic D, Kajstura J, Chimenti S, Jakoniuk I, Anderson SM, Li B, Pickel J, McKay R, Nadal- Ginard B, Bodine DM et al. 2001. Bone marrow cells regenerate infarcted myocardium. Nature. 410(6829):701-705.

Otten JJ, de Jager SC, Kavelaars A, Seijkens T, Bot I, Wijnands E, Beckers L, Westra MM, Bot M, Busch M et al. 2013. Hematopoietic g-protein-coupled receptor kinase 2 deficiency decreases atherosclerotic lesion formation in ldl receptor-knockout mice. FASEB J. 27(1):265-276.

182

Ozawa K, Whalen EJ, Nelson CD, Mu Y, Hess DT, Lefkowitz RJ, Stamler JS. 2008. S- nitrosylation of beta-arrestin regulates beta-adrenergic receptor trafficking. Mol Cell. 31(3):395-405.

Palm W, de Lange T. 2008. How shelterin protects mammalian telomeres. Annu Rev Genet. 42:301-334.

Penela P, Elorza A, Sarnago S, Mayor F. 2001. Beta-arrestin- and c-src-dependent degradation of g-protein-coupled receptor kinase 2. EMBO J. 20(18):5129-5138.

Penela P, Murga C, Ribas C, Lafarga V, Mayor F, Jr. 2010a. The complex g protein- coupled receptor kinase 2 (grk2) interactome unveils new physiopathological targets. Br J Pharmacol. 160(4):821-832.

Penela P, Ribas C, Sánchez-Madrid F, Mayor F. 2019. G protein-coupled receptor kinase 2 (grk2) as a multifunctional signaling hub. Cell Mol Life Sci. 76(22):4423-4446.

Penela P, Rivas V, Salcedo A, Mayor F. 2010b. G protein-coupled receptor kinase 2 (grk2) modulation and cell cycle progression. Proc Natl Acad Sci U S A. 107(3):1118- 1123.

Penela P, Ruiz-Gómez A, Castaño JG, Mayor F. 1998. Degradation of the g protein- coupled receptor kinase 2 by the proteasome pathway. J Biol Chem. 273(52):35238-35244.

Penn RB, Benovic JL. 1994. Structure of the human gene encoding the beta-adrenergic receptor kinase. J Biol Chem. 269(21):14924-14930.

Penn RB, Pronin AN, Benovic JL. 2000. Regulation of g protein-coupled receptor kinases. Trends Cardiovasc Med. 10(2):81-89.

Perrino C, Naga Prasad SV, Mao L, Noma T, Yan Z, Kim HS, Smithies O, Rockman HA. 2006. Intermittent pressure overload triggers hypertrophy-independent cardiac dysfunction and vascular rarefaction. J Clin Invest. 116(6):1547-1560.

183

Petersen CP, Bordeleau ME, Pelletier J, Sharp PA. 2006. Short rnas repress translation after initiation in mammalian cells. Mol Cell. 21(4):533-542.

Petroff MG, Kim SH, Pepe S, Dessy C, Marbán E, Balligand JL, Sollott SJ. 2001. Endogenous nitric oxide mechanisms mediate the stretch dependence of ca2+ release in cardiomyocytes. Nat Cell Biol. 3(10):867-873.

Pfleger J, Gresham K, Koch WJ. 2019. G protein-coupled receptor kinases as therapeutic targets in the heart. Nat Rev Cardiol. 16(10):612-622.

Pfleger J, Gross P, Johnson J, Carter RL, Gao E, Tilley DG, Houser SR, Koch WJ. 2018. G protein-coupled receptor kinase 2 contributes to impaired fatty acid metabolism in the failing heart. J Mol Cell Cardiol. 123:108-117.

Piechaczyk M, Blanchard JM, Marty L, Dani C, Panabieres F, El Sabouty S, Fort P, Jeanteur P. 1984. Post-transcriptional regulation of glyceraldehyde-3-phosphate- dehydrogenase gene expression in rat tissues. Nucleic Acids Res. 12(18):6951- 6963.

Pillai RS, Bhattacharyya SN, Artus CG, Zoller T, Cougot N, Basyuk E, Bertrand E, Filipowicz W. 2005. Inhibition of translational initiation by let-7 microrna in human cells. Science. 309(5740):1573-1576.

Pinto AR, Ilinykh A, Ivey MJ, Kuwabara JT, D'Antoni ML, Debuque R, Chandran A, Wang L, Arora K, Rosenthal NA et al. 2016. Revisiting cardiac cellular composition. Circ Res. 118(3):400-409.

Pleger ST, Harris DM, Shan C, Vinge LE, Chuprun JK, Berzins B, Pleger W, Druckman C, Volkers M, Heierhorst J et al. 2008. Endothelial s100a1 modulates vascular function via nitric oxide. Circ Res. 102(7):786-794.

Premont RT, Macrae AD, Stoffel RH, Chung N, Pitcher JA, Ambrose C, Inglese J, MacDonald ME, Lefkowitz RJ. 1996. Characterization of the g protein-coupled receptor kinase grk4. Identification of four splice variants. J Biol Chem. 271(11):6403-6410.

184

Pronin AN, Carman CV, Benovic JL. 1998. Structure-function analysis of g protein- coupled receptor kinase-5. Role of the carboxyl terminus in kinase regulation. J Biol Chem. 273(47):31510-31518.

Pérez VI, Van Remmen H, Bokov A, Epstein CJ, Vijg J, Richardson A. 2009. The overexpression of major antioxidant enzymes does not extend the lifespan of mice. Aging Cell. 8(1):73-75.

Qin XD, Liu L. 2019. Loss of microrna-27a induces cardiac dysfunction through activating foxo1. Eur Rev Med Pharmacol Sci. 23(13):5941-5948.

Quarles E, Basisty N, Chiao YA, Merrihew G, Gu H, Sweetwyne MT, Fredrickson J, Nguyen NH, Razumova M, Kooiker K et al. 2020. Rapamycin persistently improves cardiac function in aged, male and female mice, even following cessation of treatment. Aging Cell. 19(2):e13086.

Raake PW, Schlegel P, Ksienzyk J, Reinkober J, Barthelmes J, Schinkel S, Pleger S, Mier W, Haberkorn U, Koch WJ et al. 2013. Aav6.Betaarkct cardiac gene therapy ameliorates cardiac function and normalizes the catecholaminergic axis in a clinically relevant large animal heart failure model. Eur Heart J. 34(19):1437-1447.

Raake PW, Vinge LE, Gao E, Boucher M, Rengo G, Chen X, DeGeorge BR, Jr., Matkovich S, Houser SR, Most P et al. 2008. G protein-coupled receptor kinase 2 ablation in cardiac myocytes before or after myocardial infarction prevents heart failure. Circ Res. 103(4):413-422.

Rakusan K, Flanagan MF, Geva T, Southern J, Van Praagh R. 1992. Morphometry of human coronary capillaries during normal growth and the effect of age in left ventricular pressure-overload hypertrophy. Circulation. 86(1):38-46.

Ramos-Ruiz R, Penela P, Penn RB, Mayor F. 2000. Analysis of the human g protein- coupled receptor kinase 2 (grk2) gene promoter: Regulation by signal transduction systems in aortic smooth muscle cells. Circulation. 101(17):2083-2089.

185

Raut SK, Singh GB, Rastogi B, Saikia UN, Mittal A, Dogra N, Singh S, Prasad R, Khullar M. 2016. Mir-30c and mir-181a synergistically modulate p53-p21 pathway in diabetes induced cardiac hypertrophy. Mol Cell Biochem. 417(1-2):191-203.

Rengo G, Galasso G, Femminella GD, Parisi V, Zincarelli C, Pagano G, De Lucia C, Cannavo A, Liccardo D, Marciano C et al. 2014. Reduction of lymphocyte g protein-coupled receptor kinase-2 (grk2) after exercise training predicts survival in patients with heart failure. Eur J Prev Cardiol. 21(1):4-11.

Rengo G, Lymperopoulos A, Zincarelli C, Donniacuo M, Soltys S, Rabinowitz JE, Koch WJ. 2009. Myocardial adeno-associated virus serotype 6-betaarkct gene therapy improves cardiac function and normalizes the neurohormonal axis in chronic heart failure. Circulation. 119(1):89-98.

Rengo G, Pagano G, Filardi PP, Femminella GD, Parisi V, Cannavo A, Liccardo D, Komici K, Gambino G, D'Amico ML et al. 2016. Prognostic value of lymphocyte g protein- coupled receptor kinase-2 protein levels in patients with heart failure. Circ Res. 118(7):1116-1124.

Rhoades MW, Reinhart BJ, Lim LP, Burge CB, Bartel B, Bartel DP. 2002. Prediction of plant microrna targets. Cell. 110(4):513-520.

Roca-Alonso L, Castellano L, Mills A, Dabrowska AF, Sikkel MB, Pellegrino L, Jacob J, Frampton AE, Krell J, Coombes RC et al. 2015. Myocardial mir-30 downregulation triggered by doxorubicin drives alterations in β-adrenergic signaling and enhances apoptosis. Cell Death Dis. 6:e1754.

Rockman HA, Chien KR, Choi DJ, Iaccarino G, Hunter JJ, Ross J, Lefkowitz RJ, Koch WJ. 1998a. Expression of a beta-adrenergic receptor kinase 1 inhibitor prevents the development of myocardial failure in gene-targeted mice. Proc Natl Acad Sci U S A. 95(12):7000-7005.

Rockman HA, Choi DJ, Akhter SA, Jaber M, Giros B, Lefkowitz RJ, Caron MG, Koch WJ. 1998b. Control of myocardial contractile function by the level of beta- adrenergic receptor kinase 1 in gene-targeted mice. J Biol Chem. 273(29):18180- 18184.

186

Rockman HA, Koch WJ, Lefkowitz RJ. 2002. Seven-transmembrane-spanning receptors and heart function. Nature. 415(6868):206-212.

Rossi DJ, Bryder D, Seita J, Nussenzweig A, Hoeijmakers J, Weissman IL. 2007. Deficiencies in dna damage repair limit the function of haematopoietic stem cells with age. Nature. 447(7145):725-729.

Said MA, Eppinga RN, Lipsic E, Verweij N, van der Harst P. 2018. Relationship of arterial stiffness index and pulse pressure with cardiovascular disease and mortality. J Am Heart Assoc. 7(2).

Salcedo A, Mayor F, Penela P. 2006. Mdm2 is involved in the ubiquitination and degradation of g-protein-coupled receptor kinase 2. EMBO J. 25(20):4752-4762.

Salminen A, Kaarniranta K, Kauppinen A. 2012. Inflammaging: Disturbed interplay between autophagy and inflammasomes. Aging (Albany NY). 4(3):166-175.

Santos-Otte P, Leysen H, van Gastel J, Hendrickx JO, Martin B, Maudsley S. 2019. G protein-coupled receptor systems and their role in cellular senescence. Comput Struct Biotechnol J. 17:1265-1277.

Sanz MN, Grimbert L, Moulin M, Gressette M, Rucker-Martin C, Lemaire C, Mericskay M, Veksler V, Ventura-Clapier R, Garnier A et al. 2019. Inducible cardiac-specific deletion of sirt1 in male mice reveals progressive cardiac dysfunction and sensitization of the heart to pressure overload. Int J Mol Sci. 20(20).

Sarnago S, Elorza A, Mayor F. 1999. Agonist-dependent phosphorylation of the g protein- coupled receptor kinase 2 (grk2) by src tyrosine kinase. J Biol Chem. 274(48):34411-34416.

Sato PY, Chuprun JK, Ibetti J, Cannavo A, Drosatos K, Elrod JW, Koch WJ. 2015a. Grk2 compromises cardiomyocyte mitochondrial function by diminishing fatty acid- mediated oxygen consumption and increasing superoxide levels. J Mol Cell Cardiol. 89(Pt B):360-364.

187

Sato PY, Chuprun JK, Schwartz M, Koch WJ. 2015b. The evolving impact of g protein- coupled receptor kinases in cardiac health and disease. Physiol Rev. 95(2):377-404.

Schlegel P, Reinkober J, Meinhardt E, Tscheschner H, Gao E, Schumacher SM, Yuan A, Backs J, Most P, Wieland T et al. 2017. G protein-coupled receptor kinase 2 promotes cardiac hypertrophy. PLoS One. 12(7):e0182110.

Schumacher SM, Gao E, Cohen M, Lieu M, Chuprun JK, Koch WJ. 2016. A peptide of the rgs domain of grk2 binds and inhibits galpha(q) to suppress pathological cardiac hypertrophy and dysfunction. Sci Signal. 9(420):ra30.

Schumacher SM, Gao E, Zhu W, Chen X, Chuprun JK, Feldman AM, Tesmer JJ, Koch WJ. 2015. Paroxetine-mediated grk2 inhibition reverses cardiac dysfunction and remodeling after myocardial infarction. Sci Transl Med. 7(277):277ra231.

Schutzer WE, Reed JF, Bliziotes M, Mader SL. 2001. Upregulation of g protein-linked receptor kinases with advancing age in rat aorta. Am J Physiol Regul Integr Comp Physiol. 280(3):R897-903.

Schwartzkopff B, Mundhenke M, Strauer BE. 1998. Alterations of the architecture of subendocardial arterioles in patients with hypertrophic cardiomyopathy and impaired coronary vasodilator reserve: A possible cause for myocardial ischemia. J Am Coll Cardiol. 31(5):1089-1096.

Sefton M, Blanco MJ, Penela P, Mayor F, Nieto MA. 2000. Expression of the g protein- coupled receptor kinase 2 during early mouse embryogenesis. Mech Dev. 98(1- 2):127-131.

Seggerson K, Tang L, Moss EG. 2002. Two genetic circuits repress the caenorhabditis elegans heterochronic gene lin-28 after translation initiation. Dev Biol. 243(2):215- 225.

Selbach M, Schwanhäusser B, Thierfelder N, Fang Z, Khanin R, Rajewsky N. 2008. Widespread changes in protein synthesis induced by micrornas. Nature. 455(7209):58-63.

188

Selman C, Tullet JM, Wieser D, Irvine E, Lingard SJ, Choudhury AI, Claret M, Al-Qassab H, Carmignac D, Ramadani F et al. 2009. Ribosomal protein s6 kinase 1 signaling regulates mammalian life span. Science. 326(5949):140-144.

Sen GL, Blau HM. 2005. Argonaute 2/risc resides in sites of mammalian mrna decay known as cytoplasmic bodies. Nat Cell Biol. 7(6):633-636.

Serrano M, Lin AW, McCurrach ME, Beach D, Lowe SW. 1997. Oncogenic ras provokes premature cell senescence associated with accumulation of p53 and p16ink4a. Cell. 88(5):593-602.

Seth D, Stamler JS. 2011. The sno-proteome: Causation and classifications. Curr Opin Chem Biol. 15(1):129-136.

Shah AS, White DC, Emani S, Kypson AP, Lilly RE, Wilson K, Glower DD, Lefkowitz RJ, Koch WJ. 2001. In vivo ventricular gene delivery of a beta-adrenergic receptor kinase inhibitor to the failing heart reverses cardiac dysfunction. Circulation. 103(9):1311-1316.

Sharifi-Sanjani M, Oyster NM, Tichy ED, Bedi KC, Harel O, Margulies KB, Mourkioti F. 2017. Cardiomyocyte-specific telomere shortening is a distinct signature of heart failure in humans. J Am Heart Assoc. 6(9).

Shi Y, Shu ZJ, Wang H, Barnes JL, Yeh CK, Ghosh PM, Katz MS, Kamat A. 2018. Altered expression of hepatic beta-adrenergic receptors in aging rats: Implications for age- related metabolic dysfunction in liver. Am J Physiol Regul Integr Comp Physiol. 314(4):R574-R583.

Siebold AP, Banerjee R, Tie F, Kiss DL, Moskowitz J, Harte PJ. 2010. Polycomb repressive complex 2 and trithorax modulate drosophila longevity and stress resistance. Proc Natl Acad Sci U S A. 107(1):169-174.

Sin TK, Yung BY, Siu PM. 2015. Modulation of sirt1-foxo1 signaling axis by resveratrol: Implications in skeletal muscle aging and insulin resistance. Cell Physiol Biochem. 35(2):541-552.

189

Skeberdis VA, Gendviliene V, Zablockaite D, Treinys R, Macianskiene R, Bogdelis A, Jurevicius J, Fischmeister R. 2008. Beta3-adrenergic receptor activation increases human atrial tissue contractility and stimulates the l-type ca2+ current. J Clin Invest. 118(9):3219-3227.

Song Y, Yao Q, Zhu J, Luo B, Liang S. 1999. Age-related variation in the interstitial tissues of the cardiac conduction system; and autopsy study of 230 han chinese. Forensic Sci Int. 104(2-3):133-142.

Sonntag WE, Lynch CD, Cooney PT, Hutchins PM. 1997. Decreases in cerebral microvasculature with age are associated with the decline in growth hormone and insulin-like growth factor 1. Endocrinology. 138(8):3515-3520.

Sontheimer EJ. 2005. Assembly and function of rna silencing complexes. Nat Rev Mol Cell Biol. 6(2):127-138.

Sopko NA, Turturice BA, Becker ME, Brown CR, Dong F, Popović ZB, Penn MS. 2010. Bone marrow support of the heart in pressure overload is lost with aging. PLoS One. 5(12):e15187.

Soriano-Arroquia A, House L, Tregilgas L, Canty-Laird E, Goljanek-Whysall K. 2016. The functional consequences of age-related changes in microrna expression in skeletal muscle. Biogerontology. 17(3):641-654.

Sorriento D, Santulli G, Franco A, Cipolletta E, Napolitano L, Gambardella J, Gomez- Monterrey I, Campiglia P, Trimarco B, Iaccarino G et al. 2015. Integrating grk2 and nfkappab in the pathophysiology of cardiac hypertrophy. J Cardiovasc Transl Res. 8(8):493-502.

Spengler RM, Zhang X, Cheng C, McLendon JM, Skeie JM, Johnson FL, Davidson BL, Boudreau RL. 2016. Elucidation of transcriptome-wide microrna binding sites in human cardiac tissues by ago2 hits-clip. Nucleic Acids Res. 44(15):7120-7131.

Stamler JS, Lamas S, Fang FC. 2001. Nitrosylation. The prototypic redox-based signaling mechanism. Cell. 106(6):675-683.

190

Stamler JS, Simon DI, Osborne JA, Mullins ME, Jaraki O, Michel T, Singel DJ, Loscalzo J. 1992. S-nitrosylation of proteins with nitric oxide: Synthesis and characterization of biologically active compounds. Proc Natl Acad Sci U S A. 89(1):444-448.

Steenman M, Lande G. 2017. Cardiac aging and heart disease in humans. Biophys Rev. 9(2):131-137.

Stoffel RH, Randall RR, Premont RT, Lefkowitz RJ, Inglese J. 1994. Palmitoylation of g protein-coupled receptor kinase, grk6. Lipid modification diversity in the grk family. J Biol Chem. 269(45):27791-27794.

Strait JB, Lakatta EG. 2012. Aging-associated cardiovascular changes and their relationship to heart failure. Heart Fail Clin. 8(1):143-164.

Sun X, He S, Wara AKM, Icli B, Shvartz E, Tesmenitsky Y, Belkin N, Li D, Blackwell TS, Sukhova GK et al. 2014. Systemic delivery of microrna-181b inhibits nuclear factor-κb activation, vascular inflammation, and atherosclerosis in apolipoprotein e-deficient mice. Circ Res. 114(1):32-40.

Taguchi K, Kobayashi T, Takenouchi Y, Matsumoto T, Kamata K. 2011. Angiotensin ii causes endothelial dysfunction via the grk2/akt/enos pathway in aortas from a murine type 2 diabetic model. Pharmacol Res. 64(5):535-546.

Taguchi K, Matsumoto T, Kamata K, Kobayashi T. 2012. G protein-coupled receptor kinase 2, with β-arrestin 2, impairs insulin-induced akt/endothelial nitric oxide synthase signaling in ob/ob mouse aorta. Diabetes. 61(8):1978-1985.

Taylor AL. 2005. The african american heart failure trial: A clinical trial update. Am J Cardiol. 96(7B):44-48.

Taylor CJ, Ryan R, Nichols L, Gale N, Hobbs FR, Marshall T. 2017. Survival following a diagnosis of heart failure in primary care. Fam Pract. 34(2):161-168.

191

Tchkonia T, Zhu Y, van Deursen J, Campisi J, Kirkland JL. 2013. Cellular senescence and the senescent secretory phenotype: Therapeutic opportunities. J Clin Invest. 123(3):966-972.

Terai M, Izumiyama-Shimomura N, Aida J, Ishikawa N, Sawabe M, Arai T, Fujiwara M, Ishii A, Nakamura K, Takubo K. 2013. Association of telomere shortening in myocardium with heart weight gain and cause of death. Sci Rep. 3:2401.

Tevaearai HT, Eckhart AD, Shotwell KF, Wilson K, Koch WJ. 2001. Ventricular dysfunction after cardioplegic arrest is improved after myocardial gene transfer of a beta-adrenergic receptor kinase inhibitor. Circulation. 104(17):2069-2074.

Thal DM, Homan KT, Chen J, Wu EK, Hinkle PM, Huang ZM, Chuprun JK, Song J, Gao E, Cheung JY et al. 2012. Paroxetine is a direct inhibitor of g protein-coupled receptor kinase 2 and increases myocardial contractility. ACS Chem Biol. 7(11):1830-1839.

Thomson DW, Bracken CP, Goodall GJ. 2011. Experimental strategies for microrna target identification. Nucleic Acids Res. 39(16):6845-6853.

Thum T. 2012. Microrna therapeutics in cardiovascular medicine. EMBO Mol Med. 4(1):3-14.

Thum T, Galuppo P, Wolf C, Fiedler J, Kneitz S, van Laake LW, Doevendans PA, Mummery CL, Borlak J, Haverich A et al. 2007. Micrornas in the human heart: A clue to fetal gene reprogramming in heart failure. Circulation. 116(3):258-267.

Tocchi A, Quarles EK, Basisty N, Gitari L, Rabinovitch PS. 2015. Mitochondrial dysfunction in cardiac aging. Biochim Biophys Acta. 1847(11):1424-1433.

Tschudi MR, Barton M, Bersinger NA, Moreau P, Cosentino F, Noll G, Malinski T, Lüscher TF. 1996. Effect of age on kinetics of nitric oxide release in rat aorta and pulmonary artery. J Clin Invest. 98(4):899-905.

192

Vajapey R, Rini D, Walston J, Abadir P. 2014. The impact of age-related dysregulation of the angiotensin system on mitochondrial redox balance. Front Physiol. 5:439.

Valinezhad Orang A, Safaralizadeh R, Kazemzadeh-Bavili M. 2014. Mechanisms of mirna-mediated gene regulation from common downregulation to mrna-specific upregulation. Int J Genomics. 2014:970607.

Van Remmen H, Ikeno Y, Hamilton M, Pahlavani M, Wolf N, Thorpe SR, Alderson NL, Baynes JW, Epstein CJ, Huang TT et al. 2003. Life-long reduction in mnsod activity results in increased dna damage and higher incidence of cancer but does not accelerate aging. Physiol Genomics. 16(1):29-37.

van Rooij E. 2011. The art of microrna research. Circ Res. 108(2):219-234.

van Rooij E, Sutherland LB, Liu N, Williams AH, McAnally J, Gerard RD, Richardson JA, Olson EN. 2006. A signature pattern of stress-responsive micrornas that can evoke cardiac hypertrophy and heart failure. Proc Natl Acad Sci U S A. 103(48):18255-18260.

Vo NK, Dalton RP, Liu N, Olson EN, Goodman RH. 2010. Affinity purification of microrna-133a with the cardiac transcription factor, hand2. Proc Natl Acad Sci U S A. 107(45):19231-19236.

Volkova M, Garg R, Dick S, Boheler KR. 2005. Aging-associated changes in cardiac gene expression. Cardiovasc Res. 66(2):194-204.

Wagh V, Urhekar A, Modi D. 2017. Levels of microrna mir-16 and mir-155 are altered in serum of patients with tuberculosis and associate with responses to therapy. Tuberculosis (Edinb). 102:24-30.

Wahlquist C, Jeong D, Rojas-Muñoz A, Kho C, Lee A, Mitsuyama S, van Mil A, Park WJ, Sluijter JP, Doevendans PA et al. 2014. Inhibition of mir-25 improves cardiac contractility in the failing heart. Nature. 508(7497):531-535.

193

Wang G, Moniri NH, Ozawa K, Stamler JS, Daaka Y. 2006. Nitric oxide regulates endocytosis by s-nitrosylation of dynamin. Proc Natl Acad Sci U S A. 103(5):1295- 1300.

Whalen EJ, Foster MW, Matsumoto A, Ozawa K, Violin JD, Que LG, Nelson CD, Benhar M, Keys JR, Rockman HA et al. 2007. Regulation of beta-adrenergic receptor signaling by s-nitrosylation of g-protein-coupled receptor kinase 2. Cell. 129(3):511-522.

White DC, Hata JA, Shah AS, Glower DD, Lefkowitz RJ, Koch WJ. 2000. Preservation of myocardial beta-adrenergic receptor signaling delays the development of heart failure after myocardial infarction. Proc Natl Acad Sci U S A. 97(10):5428-5433.

Willkomm S, Restle T. 2015. Conformational dynamics of ago-mediated silencing processes. Int J Mol Sci. 16(7):14769-14785.

Winter J, Jung S, Keller S, Gregory RI, Diederichs S. 2009. Many roads to maturity: Microrna biogenesis pathways and their regulation. Nat Cell Biol. 11(3):228-234.

Wu L, Fan J, Belasco JG. 2006. Micrornas direct rapid deadenylation of mrna. Proc Natl Acad Sci U S A. 103(11):4034-4039.

Wu W, Sung CC, Yu P, Li J, Chung KKK. 2020. S-nitrosylation of g protein-coupled receptor kinase 6 and casein kinase 2 alpha modulates their kinase activity toward alpha-synuclein phosphorylation in an animal model of parkinson's disease. PLoS One. 15(4):e0232019.

Xiao RP, Avdonin P, Zhou YY, Cheng H, Akhter SA, Eschenhagen T, Lefkowitz RJ, Koch WJ, Lakatta EG. 1999. Coupling of beta2-adrenoceptor to gi proteins and its physiological relevance in murine cardiac myocytes. Circ Res. 84(1):43-52.

Xiao RP, Tomhave ED, Wang DJ, Ji X, Boluyt MO, Cheng H, Lakatta EG, Koch WJ. 1998. Age-associated reductions in cardiac beta1- and beta2-adrenergic responses without changes in inhibitory g proteins or receptor kinases. J Clin Invest. 101(6):1273-1282.

194

Xu X, Wang B, Ren C, Hu J, Greenberg DA, Chen T, Xie L, Jin K. 2017. Age-related impairment of vascular structure and functions. Aging Dis. 8(5):590-610.

Yang HW, Kim HD, Kim TS, Kim J. 2019a. Senescent cells differentially translate senescence-related mrnas via ribosome heterogeneity. J Gerontol A Biol Sci Med Sci. 74(7):1015-1024.

Yang M, Lin Y, Wang Y. 2019b. High-glucose induces cardiac myocytes apoptosis through foxo1 /grk2 signaling pathway. Biochem Biophys Res Commun. 513(1):154-158.

Yang SB, Tien AC, Boddupalli G, Xu AW, Jan YN, Jan LY. 2012. Rapamycin ameliorates age-dependent obesity associated with increased mtor signaling in hypothalamic pomc neurons. Neuron. 75(3):425-436.

Ye Z, Li G, Kim C, Hu B, Jadhav RR, Weyand CM, Goronzy JJ. 2018. Regulation of mir- 181a expression in t cell aging. Nat Commun. 9(1):3060.

Yekta S, Shih IH, Bartel DP. 2004. Microrna-directed cleavage of hoxb8 mrna. Science. 304(5670):594-596.

Yi R, Qin Y, Macara IG, Cullen BR. 2003. Exportin-5 mediates the nuclear export of pre- micrornas and short hairpin rnas. Genes Dev. 17(24):3011-3016.

Yildiran T, Koc M, Bozkurt A, Sahin DY, Unal I, Acarturk E. 2010. Low pulse pressure as a predictor of death in patients with mild to advanced heart failure. Tex Heart Inst J. 37(3):284-290.

Ytrehus K, Hulot JS, Perrino C, Schiattarella GG, Madonna R. 2018. Perivascular fibrosis and the microvasculature of the heart. Still hidden secrets of pathophysiology? Vascul Pharmacol.

Zha Z, Han X, Smith MD, Liu Y, Giguère PM, Kopanja D, Raychaudhuri P, Siderovski DP, Guan KL, Lei QY et al. 2015. A non-canonical function of gβ as a subunit of e3 ligase in targeting grk2 ubiquitylation. Mol Cell. 58(5):794-803.

195

Zhang LX, Ding F, Wang CQ, Bing Q, Zhao Z, Wang J, Zhang L. 2019a. Mir-181a affects myocardial ischemia-reperfusion injury in rats via regulating akt signaling pathway. Eur Rev Med Pharmacol Sci. 23(14):6292-6298.

Zhang M, Zhang Q, Hu Y, Xu L, Jiang Y, Zhang C, Ding L, Jiang R, Sun J, Sun H et al. 2017. Mir-181a increases foxo1 acetylation and promotes granulosa cell apoptosis via sirt1 downregulation. Cell Death Dis. 8(10):e3088.

Zhang X, Goel V, Robbie GJ. 2019b. Pharmacokinetics of patisiran, the first approved rna interference therapy in patients with hereditary transthyretin-mediated amyloidosis. J Clin Pharmacol.

Zhao WL, Wang D, Liu CY, Zhao XF. 2016. G-protein-coupled receptor kinase 2 terminates g-protein-coupled receptor function in steroid hormone 20- hydroxyecdysone signaling. Sci Rep. 6:29205.

Zheng F, Plati AR, Potier M, Schulman Y, Berho M, Banerjee A, Leclercq B, Zisman A, Striker LJ, Striker GE. 2003. Resistance to glomerulosclerosis in b6 mice disappears after menopause. Am J Pathol. 162(4):1339-1348.

Zhou B, Li C, Qi W, Zhang Y, Zhang F, Wu JX, Hu YN, Wu DM, Liu Y, Yan TT et al. 2012. Downregulation of mir-181a upregulates sirtuin-1 (sirt1) and improves hepatic insulin sensitivity. Diabetologia. 55(7):2032-2043.

Zhou HL, Stomberski CT, Stamler JS. 2018. Cross talk between. Circ Res. 122(11):1485- 1487.

Zhou L, Bai H, Wang C, Wei D, Qin Y, Xu X. 2014. Microrna‑125b promotes leukemia cell resistance to daunorubicin by inhibiting apoptosis. Mol Med Rep. 9(5):1909- 1916.

Zhu J, He F, Hu S, Yu J. 2008. On the nature of human housekeeping genes. Trends Genet. 24(10):481-484.

196

Ørom UA, Nielsen FC, Lund AH. 2008. Microrna-10a binds the 5'utr of ribosomal protein mrnas and enhances their translation. Mol Cell. 30(4):460-471.

197