I NOVEL FUNCTIONS of ADRENOMEDULLIN, RECEPTOR

Home , Adrenomedullin, Angiopoietin 1, GPR182, RAMP1, RAMP2, RAMP3

NOVEL FUNCTIONS OF ADRENOMEDULLIN, RECEPTOR ACTIVITY-MODIFYING PROTEIN 2, AND GPR182 IN CARDIOVASCULAR DISEASE AND CANCER

Daniel Oliver Kechele

A dissertation submitted to the faculty of the University of North Carolina at Chapel Hill in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Cell Biology and Physiology (Cell and Molecular Physiology)

Chapel Hill 2016

Approved by:

Kathleen Caron

Keith Burridge

Andrew Dudley

John Rawls

Carol Otey

ABSTRACT

Daniel Oliver Kechele: Novel Functions of Adrenomedullin, Receptor Activity-Modifying Protein 2, and Gpr182 in Cardiovascular Disease and Cancer (Under the direction of Kathleen Caron)

The multifunctional circulating peptide, adrenomedullin (AM) signaling through its G protein-coupled receptor (GPCR) complex, comprised of the calcitonin receptor-like receptor

(CLR) and receptor activity-modifying protein 2 (Ramp2), is required for lymphatic development and embryonic survival, as well as numerous functions in adult cardiovascular physiology. However these functions, especially during adulthood, have not been fully elucidated. Using both gain- and loss-of-function mouse models these studies aimed to further elucidate the roles of AM signaling during cardiovascular homeostasis and tumor progression. Endothelial restoration of Ramp2 is sufficient to rescue the embryonic lethality- associated with global loss of Ramp2. However, the developmental loss of Ramp2 in non- endothelial cells led to spontaneous hypotension, dilated cardiomyopathy, and multi-organ inflammation in adult survivors. Smooth muscle cell- or global-deletion of CLR did not recapitulate these phenotypes, indicating that Ramp2 is likely interacting with other GPCRs, including parathyroid hormone and glucagon receptors, to maintain cardiovascular homeostasis.

While AM signaling is essential for lymphatic development and maintenance, it is unclear how AM influences tumor lymphatics. Tumor cells engineered to overexpress or knockdown AM did not alter tumor growth or angiogenesis. High tumor-derived AM caused tumor and lymph node lymphangiogenesis and increased metastasis. While AM is typically considered protective, these data suggest that AM signaling could enhance tumor progression.

iii

While CLR/Ramp2 is the canonical AM receptor, it was previously thought that the

AM receptor was the orphan GPCR, Gpr182, of which almost nothing is known. Using a knock-in reporter mouse model to map Gpr182 expression in vivo, Gpr182 was enriched in the active stem cells in the small intestine. Gpr182 was dispensable during development and homeostasis, but when challenged with irradiation injury or adenoma model, reduced

Gpr182 led to intestinal hyperproliferation during regeneration and enhanced adenoma formation, respectively.

Together, these studies demonstrated the importance of endothelial AM signaling during development, the complexity of GPCR/RAMP interactions in vivo, the underappreciated functions of AM signaling in tumor lymphatic metastasis, and the role of

Gpr182 as a negative regulator of intestinal proliferation; all of which could represent pharmacological-tractable targets to treat a number of cardiovascular and neoplastic diseases.

ACKNOWLEDGEMENTS

I would like to thank all members, past and present, of the Caron lab for intellectual discussions and technical assistance in all my projects. With regards to scientific data presented throughout this dissertation, I would like to thank the University of North Carolina

Lineberger Animal Histopathology Core (NIH CA16086) and Kirk McNaughton and Ashley

Ezzell of the University of North Carolina Histology Research Core for tissue embedding, sectioning, and staining expertise. I would also like to acknowledge Dr. Robert Bagnell and the University of North Carolina Microscopy Services Laboratory for confocal microscopy assistance. Regarding the data presented in Chapter II, I would like to thank the University of North Carolina Animal Models Core and the University of North Carolina Rodents

Advanced Surgical Models Core, specifically Dr. Mauricio Rojas and Dr. Brian Cooley with echocardiography and intra-arterial blood pressure measurements, respectively. I would also like to thank Dr. Lin Xiao and Dr. Andrew Dudley (UNC-CH Cell Biology and

Physiology) for endothelial isolation expertise, and Helen Willcockson, Dr. Samantha

Hoopes, John Pawlak, and Dr. Wenjing Xu for technical assistance. For preliminary data presented in Chapter IV, I would like to thank Dr. James Dunleavey of the Dudley lab for providing melanoma cell line and tumor and flow cytometry expertise. For study presented in Chapter VI, I would like to thank all members of Dr. Kay Lund’s laboratory (UNC-CH Cell

Biology and Physiology) for reagents and discussions. I would like to thank Dr. Linda

Samuelson (University of Michigan Molecular and Integrative Physiology) for provided Notch

RNA and continued mentorship. Finally, I would like to acknowledge that research conducted in this dissertation was supported by NIH RO1-HL091973, RO1-HD060860 and

RO1-DK099156 grants to K.M.C.; NIH F31-CA174194 and the University of North Carolina

Cell and Molecular Physiology Fellner Fellowship to D.O.K.; an American Heart Association

12PRE11710002 to S.E.W-S.; an American Heart Association 15PRE25680001 to C.E.T.;

NIH 2-T32-DK07686 to A.T.M; NIH 1-F30-CA200345 and T32-CA071341-17 to M.B.S; and

NIH RO1-DK040247-19 to P.K.L.

TABLE OF CONTENTS LIST OF TABLES ...... x LIST OF FIGURES ...... xi LIST OF ABBREVIATIONS ...... xiii

PART I

Chapter I: Adrenomedullin Function in Vascular Endothelial Cells: Insights from Genetic Mouse Models ...... 1

Overview ...... 1

Introduction ...... 1

Development ...... 6

Physiology and Pathology ...... 10

Summary and Future Directions ...... 16

Author Contributions...... 16

Tables ...... 17

Figures ...... 19

Chapter II: Endothelial restoration of receptor activity-modifying protein 2 is sufficeint to rescue lethality, but survivors develop dilated cardiomyopathy ...... 21

Overview ...... 21

Introduction ...... 22

Methods ...... 23

Results ...... 25

Discussion ...... 32

Perspectives ...... 35

Novelty and Significance ...... 36

vii

Author Contributions ...... 37

Tables ...... 38

Figures ...... 40

Supplemental Methods ...... 47

Supplemental Tables...... 54

Supplemental Figures ...... 58

Chapter III: Dosage of adrenomedullin correlates with tumor and lymph node lymphangiogenesis ...... 61

Overview ...... 61

Introduction ...... 62

Materials and Methods ...... 64

Results ...... 68

Discussion ...... 73

Author Contributions ...... 76

Figures ...... 77

Chapter IV: Conclusions and Future Directions ...... 85

Summary ...... 85

Current State of the Field – RAMPs, Cardiac Disease, and Tumor Lymphatics ...... 85

Future Directions ...... 92

Concluding Remarks ...... 106

Figures ...... 107

Part II

Chapter V: The History of AM Receptors: The Forgotten GPCRs...... 116

Overview ...... 116

The Rise and Fall of AM and CGRP Receptors prior to RAMP Discovery ...... 116

viii

CLR, Cxcr7 and Gpr182 in a RAMP World ...... 120

Figure ...... 122

Chapter VI: Gpr182 inhibits intestinal proliferation during regeneration and adenoma formation ...... 123

Overview ...... 123

Introduction ...... 124

Methods ...... 126

Results ...... 131

Discussion ...... 140

Author Contributions ...... 142

Figures ...... 143

Supplemental Tables...... 152

Supplemental Figures ...... 153

Chapter VII: Conclusions and Future Directions ...... 157

Summary ...... 157

Current State of the Field -- ISCs, regeneration, and carcinoma...... 157

Future Directions ...... 163

Concluding Remarks ...... 176

Figures ...... 177

References ...... 179

LIST OF TABLES

Table 1-1 Gene Targeted Mouse Models for Studying Adrenomedullin Signaling ...... 17

Table 1-2 Vascular Assays for Studying Adrenomedullin Function ...... 18

Table 2-1 Endothelial restoration of Ramp2 can rescue the global Ramp2-/- embryonic lethality ...... 38

Table 2-2 Ramp2-/- Tg adults develop left ventricle dilatation and declining heart function ...... 39

Table 2-S1 Genotype and RT-PCR Gene Expression Primers and Probes ...... 54

Table 2-S2 Relatively similar systemic circulating factors in Ramp2+/+ ntg and Ramp2-/- Tg serum ...... 55

Table 2-S3 Vascular smooth muscle cell conditional Calcrl deletion does not lead to hypotension or dilated cardiomyopathy phenotype in adult male mice ...... 56

Table 2-S4 Global conditional Calcrl deletion does not lead to dilated cardiomyopathy phenotype in aged females...... 57

Table 6-S1 RT-PCR and Genotyping Primer and Probes ...... 152

LIST OF FIGURES

Figure 1-1 Fold change in plasma adrenomedullin levels in a variety of human conditions ...... 19

Figure 1-2 Adrenomedullin signaling in development and vascular biology ...... 20

Figure 2-1 Endothelial restoration of Ramp2 partially rescues embryonic edema leading to prolonged survival of Ramp2-/- mice ...... 40

Figure 2-2 Ramp2-/- Tg adults are hypotensive ...... 41

Figure 2-3 Ramp2-/- Tg adults develop spontaneous dilated cardiomyopathy ...... 42

Figure 2-4 Ramp2-/- Tg left ventricles hypertrophied with early fibrotic and oxidative stress changes ...... 43

Figure 2-5 Ramp2-/- Tg have vascular congestion and multi-organ inflammation downstream of hypotension and dilated cardiomyopathy ...... 44

Figure 2-6 Decreased signaling and expression of RAMP-associated GPCRs in embryonic and adult Ramp2-/- Tg hearts ...... 45

Figure 2-S1 Endothelial-specific overexpression of murine Ramp2 ...... 58

Figure 2-S2 Generation of gene-targeted Ramp2 deficient mice with and without the Ramp2 Tg ...... 59

Figure 2-S3 Generation and characterization of vascular smooth muscle cell-specific Calcrl deficient mice ...... 60

Figure 3-1 In vitro characterization of the LLC cell line ...... 77

Figure 3-2 In vivo tumor analysis ...... 78

Figure 3-3 Adm expression does not affect angiogenesis ...... 79

Figure 3-4 Adm expression is necessary for lymphangiogenesis ...... 80

Figure 3-5 Adm dosage does not affect macrophage or Vegf levels ...... 81

Figure 3-6 Elevated Adm expression increases the diameter of blood and lymphatic vessels ...... 82

Figure 3-7 Tumor Adm expression influences sentinel lymph node lymphangiogenesis ...... 83

Figure 3-8 Overexpression of tumor AM may promote distant metastasis ...... 84

Figure 4-1 Edema and late gestational lethality-associated with some Ramp2-/- Tg ..... 107

Figure 4-2 Ramp2-/- Tg dams can get pregnant but have reduced litter size ...... 108

Figure 4-3 Ramp2-/- Tg spontaneous hypotension is sex-dependent ...... 109

Figure 4-4 Systemic high AM leads to increased metastasis independent of tumor growth or lymph(angiogenesis) ...... 110

Figure 4-5 Increased tumor cell seeding in AdmHiHi mice independent of primary mass ...... 111

Figure 4-6 Chronic elevated systemic AM results in multi-organ inflammation ...... 112

Figure 4-7 AM does not directly alter NK cell cytotoxicity ...... 113

Figure 4-8 Effects of high systemic AM on tumor metastasis through suppression of the cytotoxic immune system ...... 115

Figure 5-1 Historical perspectives of AM pharmacology ...... 122

Figure 6-1 Murine Gpr182 expression profile during development and adulthood ...... 143

Figure 6-2 Gpr182 is enriched in CBC intestinal stem cells ...... 144

Figure 6-3 Gpr182 knockdown does not alter basal proliferation ...... 145

Figure 6-4 Decreased Gpr182 leads to hyperproliferation during the regeneration phase after irradiation-induced injury ...... 146

Figure 6-5 Reduced Gpr182 increased lethality and polyp number in ApcMin/+ mice ...... 148

Figure 6-6 Mosaic Gpr182 expression corresponds to polyp proliferation heterogeneity ...... 149

Figure 6-7 Gpr182 knockdown leads to elevated Erk1/2 signaling upstream of the hyperproliferative intestinal crypt microenvironment ...... 150

Figure 6-8 Low GPR182 expression in human carcinomas ...... 151

Figure 6-S1 Additional murine Gpr182 expression during development and adulthood ...... 153

Figure 6-S2 Gpr182 knockout does not alter X-gal expression pattern or cell markers .... 155

Figure 7-1 Notch signaling potentially an upstream regulator of Gpr182 expression ..... 177

Figure 7-2 Reduced Gpr182 decreased Akt activation in basal intestine ...... 178

xii

LIST OF ABBREVIATIONS

GPCR (G protein-coupled receptor)

RTK (receptor tyrosine kinase)

AM (adrenomedulin protein)

Adm (adrenomedullin gene)

CLR (calcitonin receptor-like receptor protein)

Calcrl (calcitonin receptor-like receptor gene)

RAMP (receptor activity modifying protein protein)

Ramp (receptor activity modifying protein gene)

PAM (peptidylglycine alpha-amidating monooxygenase)

PAMP (proadrenomedullin N-terminal 20 amino acid peptide)

CGRP (calcitonin gene-related peptide) proADM (proadrenomedullin)

Cdh5 (cadherin 5 or VE-cadherin)

CTR (calcitonin receptor protein)

Calcr (calcitonin receptor gene)

Pthr1 (parathyroid hormone receptor 1)

Gcgr (glucagon receptor)

GLP-1R (glucagon-like peptide 1 receptor)

GLP-1 (glucagon-like peptide 1)

SCTR (secretin receptor)

VIPR1 (vasointesinal peptide receptor 1)

CaSR (calcium-sensing receptor)

Gpr30 (G protein-coupled estrogen receptor)

Pgc-1(peroxisome proliferator-activated receptor  coactivator 1-)

CREB (cAMP response element-binding protein)

xiii

HIF-1 (hypoxia inducible factor-1)

Admr (L1/G10D/adrenomedullin receptor/Gpr182)

Ackr3 (atypical chemokine receptor 3/Cxcr7)

Gpr182 (G protein-coupled receptor 182)

Lgr5 (leucine-rich repeat-containing G protein-coupled receptor 5)

-Gal (-galactosidase)

EGFP (enhanced green fluorescent protein)

ApcMin (mutated adenoma polyposis coli)

MAPK (mitogen-activated protein kinase)

Erk1/2 (p44/42 mitogen-activated protein kinase)

VEGF (vascular endothelial growth factor)

BMP (bone morphogenetic protein)

IGF (insulin-like growth factor)

FGF (fibroblast growth factor)

EGF (epidermal growth factor)

BEC (blood endothelial cell)

HUVEC (human umbilical vein endothelial cell)

LEC (lymphatic endothelial cell)

LLC (Lewis Lung Carcinoma)

LN (lymph node) pNK Cell (peripheral Natural Killer Cell) vSMC (vascular smooth muscle cell)

CBC (crypt base columnar)

ISCs (intestinal stem cells)

KO (knockout)

Tg (Ramp2 transgene-driven by Cdh5 promotor)

xiv ntg (non-transgenic)

OExp (overexpression)

ISH (in situ hybridization)

IHC (immunohistochemistry)

OPT (optical projection tomography)

EdU (5-ethynyl-2’-deoxyuridine)

H&E (hematoxylin and eosin)

DCM (dilated cardiomyopathy)

IRR (irradiated)

4-HNE (4-hydroxynonenal)

HW (heart weight)

BW (body weight)

LV (left ventricle)

RV (right ventricle)

TL (tibia length)

BPM (beats per minute) d (diastole) s (systole)

CO (cardiac output)

EF (ejection fraction)

FS (fractional shortening)

IVS (interventricular septal)

LVID (left ventricle internal diameter)

LVPW (left ventricle posterior wall)

CA (carotid artery)

JV (jugular vein)

JLS (jugular lymph sac)

DA (dorsal aorta)

VV (vitteline vein)

PE (phenylephrine)

KOMP (Knockout Mouse Project)

TCGA (The Cancer Genome Atlas)

COAD (colon adenocarcinoma)

KICH (kidney chromophobe)

BRCA (breast invasive carcinoma)

KIRC (kidney renal clear cell carcinoma)

KIRP (kidney renal papillary cell carcinoma)

LIHC (liver hepatocellular carcinoma)

LUAD (lung adenocarcinoma)

LUSC (lung squamous cell carcinoma)

PAAD (pancreatic adenocarcinoma)

READ (rectum adenocarcinoma)

STAD (stomach adenocarcinoma)

THCA (thyroid carcinoma)

-MHC (-myosin heavy chain)

NICD (Notch intracellular domain)

xvi

Chapter I. Introduction: Adrenomedullin Function in Vascular Endothelial

Cells: Insights from Genetic Mouse Models1

Overview

Adrenomedullin is a highly conserved peptide implicated in a variety of physiological processes ranging from pregnancy and embryonic development to tumor progression. This chapter highlights past and current studies that have contributed to our current appreciation of the important roles adrenomedullin plays in both normal and disease conditions. There is a particular emphasis on the functions of adrenomedullin in vascular endothelial cells and how experimental approaches in genetic mouse models have helped to drive the field forward.

Introduction

The Multifunctional Adrenomedullin Peptide

Adrenomedullin (gene=Adm; protein=AM) is a highly conserved multifunctional peptide that is implicated in a wide variety of physiological processes including angiogenesis and cardiovascular homeostasis.1 For over a decade, the association of ~2-fold elevations in plasma levels of AM peptide with a wide variety of cardiovascular disease conditions has prompted intense inquiry into understanding the functions and roles of AM in human disease

1 Reprinted with permission from: Natalie O. Karpinich, Samantha L. Hoopes, Daniel O. Kechele, Patrica M. Lenhart, Kathleen M. Caron. Adrenomedullin function in vascular endothelial cells: insights form genetic mouse models. Current Hypertension Reviews. 2011;7:228-239.

(Figure 1-1). Moreover, the recent development of highly precise methods for the quantitation of proadrenomedullin (proADM) as a reliable surrogate of mature AM plasma levels,2 has paved the way for the introduction of AM as a clinically useful biomarker for the staging of adverse cardiovascular events, including myocardial infarction, sepsis and community acquired pneumonia.3-6 While it is clear that AM can elicit powerful effects on vascular smooth muscle cells and thus acutely modulate vascular tone, numerous studies in the past 5 years have elucidated essential functions of AM on vascular endothelial cells. In the following sections information is summarized pertaining to the multi-faceted role of AM in endothelial cells during development, how perturbations in AM signaling may lead to vascular pathologies, and recent discoveries regarding AM that have contributed in substantial ways to the broader field of vascular biology. Much of these discoveries have been unraveled through the use of sophisticated genetic animal models (Tables 1-1 and 1-

2), and so special emphasis has been placed on describing the merits and shortcomings of these approaches and also highlighting current questions that are of predominant interest to the field today.

Adrenomedullin GPCR-Mediated Signaling in Endothelial Cells

G-protein coupled receptors (GPCRs) are widely expressed proteins that span the cell membrane 7 times and respond to a variety of stimuli including peptides, proteins, small organic compounds, lipids, amino acids, and cations. AM binds and signals through the

GPCR calcitonin receptor-like receptor (gene=Calcrl; protein=CLR). The discovery of a novel class of GPCR-associated proteins called receptor activity-modifying proteins

(gene=Ramp; protein=RAMP)7 provided insight into how GPCRs signal. The RAMPs are single-pass transmembrane accessory proteins that regulate the translocation of GPCRs to the plasma membrane as well as provide ligand specificity to these receptors. The tissue specific and temporal expression pattern of RAMPs determines the responsiveness of

GPCRs to particular ligands. For example, AM binds to the CLR receptor when CLR is associated with either RAMP2 or RAMP3. However, co-expression of CLR with RAMP1 changes the ligand specificity to another potent vasodilator called calcitonin gene-related peptide (CGRP), a related family member of the AM peptide. The ability of CLR to bind multiple ligands provides a unique mechanism by which the receptor can initiate a variety of signaling pathways. Since the AM receptor CLR and the 3 mammalian RAMPs are highly expressed in the vasculature, this cell signaling paradigm is being intensely investigated to determine how it can be exploited for the potential treatment of conditions such as pulmonary hypertension8, cardiovascular disorders,9 and the inhibition of cancer metastasis.10

The binding of AM to its receptor CLR results in a myriad of downstream effects including modulation of endothelial cell survival, proliferation, and vessel permeability. For example, AM-induced proliferation and migration of lymphatic endothelial cells is mediated in part by cAMP and downstream MEK/ERK pathways.11 Similar results were shown using cultured HUVECs. AM-mediated induction of HUVEC proliferation and migration through activation of PKA, PI3K, and focal adhesion kinase were observed and then further substantiated in whole animal studies.12, 13 AM induced the proliferation and migration of cultured human umbilical vein endothelial cells (HUVECs)12 and numerous studies have shown a direct role for AM in endothelial growth and survival.14-16

Using in vitro experiments, AM was found to regulate the permeability and migration of HUVECs.17 Previous studies indicated that adult Ramp2+/- mice had increased vascular permeability and overexpression of Ramp2 in BECs resulted in reduced permeability.18 AM also reduces the permeability of HUVECs and pulmonary artery endothelial cells treated with permeabilizing agents including hydrogen peroxide and thrombin.19 AM has been shown to regulate the transport of molecules across the blood brain barrier in cerebral endothelial cells by modulating permeability.20 In cerebral endothelial cells, AM regulated

3 various functions of the blood brain barrier including increasing transendothelial electrical resistance, reducing fluid-phase endocytosis, and reducing permeability for sodium fluorescein which indicate that the cerebral endothelial cell junctions are tightened by AM.20

Also in an in vivo model, AM treatment reduced lung vascular permeability resulting from ventilator use in a mouse model where prolonged mechanical ventilation was administered.21 Overall, these data provide evidence for the role of AM as a junctional tightening factor to help regulate endothelial cell permeability.

Although AM functions to promote endothelial cell growth and proliferation in both the blood and lymphatic vasculatures, Fritz-Six et al. have shown that there is an enhanced effect of AM on lymphatic endothelial cells (LECs) as compared to blood endothelial cells

(BECs).22 This biological distinction in AM function is based upon the finding that LECs are enriched in the expression of AM and its receptor components, Calcrl and Ramp2.22-24 This increase in Calcrl expression is mediated in part by induction of the transcriptional regulator of lymphatic specification, Prox1.22 It is therefore not surprising that loss of any component of the AM signaling axis (Adm, Calcrl, or Ramp2) results in embryonic lethality associated with profound lymphatic vascular defects.22 Furthermore, several in vitro and in vivo experiments reveal that AM controls lymphatic permeability and flow through reorganization of junctional proteins ZO-1 and an adherens protein VE-Cadherin, independent of changes in junctional protein gene expression.25 Administration of AM to a monolayer of LECs resulted in tightening of the lymphatic endothelial barrier by reorganization of a tight junction protein at the plasma membrane to form continuous cell-cell contacts. Through the use of in vivo tail microlymphography, local administration of AM in a SvEv129/6 mouse tail resulted in decreased velocity of lymph uptake from the interstitial space and movement through the lymphatic dermal capillaries in the tail.25 Thus, it becomes critically important to consider the pleiotropic effects of AM not just on blood endothelial cells, but also on neighboring

4 lymphatic vessels—a dynamic that may ultimately help resolve the complex functions of AM peptide in cardiovascular disease, tumor progression and inflammation.

While activation of GPCRs typically leads to induction of classical second messenger signaling systems, it is now appreciated that more complex levels of regulation exist.26, 27

Therefore, it is not surprising that pathway cross-talk is one mechanism through which AM modulates certain endothelial cell functions. For example, Yurugi-Kobayashi et al. describe a novel embryonic stem cell differentiation system to study mechanisms of arterial-venous specification. They demonstrated that coordinated signaling of AM/cAMP, VEGF, and Notch induces arterial endothelial cell differentiation from vascular progenitors.28 Furthermore,

GPCR-induced transactivation of receptor tyrosine kinases is another mechanism that allows interaction between signaling molecules. Evidence exists that AM and VEGF pathways are likely to interact in endothelial cells. Although an earlier study claimed that

AM-induced capillary tube formation in HUVECs was independent of VEGF activation,14 a more recent study by Guidolin et al. demonstrated that VEGFR2 inactivation inhibited AM- mediated angiogenesis in HUVECs.29 This latter finding suggests that the pro-angiogenic effects of AM require transactivation of the receptor tyrosine kinase VEGFR2. Although controversy still exists regarding the degree of cooperation between pathways, it is certainly intriguing to consider that regulation of endothelial cell biology may very likely involve coordination of multiple signaling molecules. We now must begin to unravel these complexities and elucidate whether these interactions occur differentially in blood and lymphatic endothelial cells and identify the intermediate molecular players involved in pathway cross-talk in the vasculature.

Development

Endothelial Adrenomedullin Signaling is Essential for Embryonic Development

Work by multiple independent groups has established the importance of AM signaling during development. The use of gene targeted mouse models clearly indicates that functional AM signaling is essential for embryonic survival. The genetic ablation of

Adm,30-32 Calcrl,33 Ramp218, 22, 34 or the enzyme responsible for functional AM amidation, peptidylglycine alpha-amidating monooxygenase (PAM)35 all result in midgestational lethality associated with severe interstitial edema and cardiovascular defects. The conserved phenotypes between these knockout (KO) mice is compelling genetic evidence that the

CLR/RAMP2 complex is the main receptor of AM during development, and also is the first in vivo confirmation that RAMP2 functionally interacts with CLR.22

Although the overt phenotypes of these KO mice are conserved, the physiological cause of edema and lethality is still debated. One possible hypothesis is that loss of AM signaling causes developmental cardiac abnormalities that lead to heart failure, thus resulting in edema and death that is similar to previously characterized KO mice with developmental heart failure.36-38 Supporting this line of thought, our lab showed that Adm-/-,

Calcrl-/-, and Ramp2-/- mice have smaller hearts due to decreased myocyte proliferation and increased apoptosis. Additionally, they have increased left ventricle trabecularization, which leads to decreased ventricular volume.22, 30, 33 However, an alternative hypothesis is that vascular defects are responsible for the phenotypes, since Adm,30 Calcrl,33 and Ramp218 are abundantly expressed in the developing endothelium and vascular smooth muscle cells

(vSMC). To help resolve between the two hypotheses, we generated an endothelial-specific

Calcrl-/- mouse using a Tie2 promoter to drive Cre recombinase expression which recapitulated the phenotype observed in global KO mice,22 indicating that AM signaling in endothelial cells is essential for embryonic development, further supported by Ramp2 rescue in Chapter II. A remaining caveat to this conclusion is the fact that Tie2-Cre

6 mediated excision also occurs in developing endocardial cells. Therefore, to definitively determine if cardiac abnormalities contribute to this phenotype the reverse experiment using

Cre lines specific to cardiac myocytes would be beneficial.

Although vascular defects are responsible for the edema in these KO mice, it remained unclear whether defects in the blood or lymphatic endothelium were the principle cause of the phenotypes. Given the role of AM in regulating vascular permeability, it seems reasonable that loss of AM signaling could lead to increased vascular permeability and a resulting buildup of interstitial fluid. In support of this idea, the KO mice have thinner aorta and carotid artery walls due to a decrease in vSMC proliferation,18, 30, 33 although the endothelium lining the aorta appeared to be normal.33 There are reported abnormalities in endothelial basement membranes and a down-regulation of junctional proteins in Adm-/- and

Ramp2-/- embryos that may lead to increased vascular permeability and hemorrhage,18, 31 but these phenotypes were observed in a small proportion of animals and not conserved in all studies. In addition, the severity of the edema and their survival beyond e10.5 does not resonate with other knockout mouse models with established vascular permeability defects.39-41 In contrast, the onset (Calcrl=e12.5, Adm=e13.5, Ramp2=e14.5) and severity of the phenotype closely resembles other genetic mouse models that delete genes essential for lymphatic development, including Prox1,42 Sox18,43 and Vegfc.44 To determine whether lymphatic vasculature defects may contribute to the edema observed in AM signaling KO animals, we performed a comprehensive study of AM signaling expression and function during lymphatic vascular development.22 It is most likely that a combination of both blood and lymphatic defects leads to the edema and lethality in the KO mice given the integrated physiology between the two vasculatures. However, more specialized genetic assays are required to resolve the relative contributions of each vasculature within these KO mice.45

An alternative approach to assess the role of AM signaling in development would be to use transgenic mouse models that overexpress Adm, Calcrl, or Ramp2. Interestingly, no

7 developmental phenotypes have been reported in gain-of-function mouse models of AM signaling, either by vascular Adm overexpression46 or vSMC-specific Ramp2 overexpression,47 though these models displayed adult cardiovascular phenotypes. Given the essential nature of AM signaling within the endothelium, it would be interesting to overexpress Calcrl or Ramp2 specifically in the endothelium, which to our knowledge, has not yet been reported (See Chapter II).

Adrenomedullin vs. Proadrenomedullin

One potential caveat with the majority of Adm-/- studies is that the gene targeting strategies delete the entire Adm coding sequence,30, 31 which results in the genetic KO of two functionally active peptides, AM and proadrenomedullin N-terminal 20 amino acid peptide (PAMP).48 PAMP is small peptide that is produced during post-transcriptional splicing of preproadrenomedullin and has numerous actions to complement or antagonize

AM signaling.48-51 For two of the reported Adm deficient mouse lines, the design of the targeted allele could not rule out whether the observed phenotypes in the KO animals were due to loss of AM, PAMP, or both.30, 31 This controversy was partially resolved using a third independent Adm-/- mouse, which left PAMP intact, and illustrated that loss of AM alone was enough to recapitulate embryonic lethality.32 However, these mice lacking only AM had a milder phenotype (less edema and no cardiovascular abnormalities) when compared to KO mice lacking both peptides. This inconsistency in phenotypes could be attributed to differences in mouse strain and/or gene targeting approach.32 However, a more intriguing hypothesis, which remains to be vigorously experimentally addressed, is that AM and PAMP may have non-redundant functions during cardiovascular development.52

Developmental Role of RAMP2 vs. RAMP3

While Ramp2-/- mice recapitulated the Adm-/- and Calcrl-/- phenotypes, it appears that

RAMP3, another RAMP that associates with CLR and binds AM, is not essential for embryonic survival since Ramp3-/- mice develop normally to adulthood. There also appears to be no functional redundancy between RAMP2 and RAMP3 in development, since there is no transcriptional compensatory mechanism of either RAMP in response to loss of the other.18, 34 Although RAMP3 has been implicated in receptor trafficking,7, 53, 54 the functional role of the AM/CLR/RAMP3 signaling complex is not well understood in vivo.

New Developmental Insights of Adrenomedullin Pathway

A recent study by Nicoli et al. expanded our knowledge regarding the role of CLR during embryonic vascular development using a zebrafish model. By knocking down crlr they showed drastic vascular defects due to decreased expression of vegf. While vegf appears to be the critical mediator in the vascular development since overexpression of vegf is able to rescue the crlr knockdown phenotype, it still appears that crlr is essential for appropriate levels of vegf. This study provides in vivo evidence that crlr is downstream of sonic hedgehog, but upstream of vegf and notch signaling in arterial differentiation and development.55 Modulation of vegf levels by AM signaling were previously reported in mice56 but a complete characterization of AM and VEGF interactions is not well understood.

It is novel that sonic hedgehog appears to regulate crlr expression and further dissection of this pathway in animal models would improve our understanding of how CLR is regulated during development. The zebrafish model system has recently been used to study lymphatic development57-59 and it would be interesting if phenotypes seen in Adm-/-, Calcrl-/-, and Ramp2-/- mice could be recapitulated in zebrafish.

Physiology and Pathophysiology

Adrenomedullin Signaling in Pregnancy

AM signaling is known to be a critical component for initiation and progression of normal pregnancy. By the third trimester of a normal pregnancy, plasma levels of AM increase 4- to 5-fold.60-65 AM is highly expressed in all vascular tissues which include the placenta and uterus.66, 67 Our previous studies in Adm+/- female mice expressing 50% less

AM revealed that there is disrupted fertility, placentation, uterine receptivity, and fetal growth resulting from reduced AM expression.68 AM signaling components are also expressed in the trophoblast cells,69-74 which take on an endothelial-like function during the process of decidual maternal spiral artery remodeling during pregnancy. The trophoblast giant cells deriving from the trophectoderm invade and replace the vascular wall by inducing a loss of endothelial cells and smooth muscle cell coverage to allow for higher blood flow to the fetus through the spiral arteries. Failure of this remodeling process to occur is a hallmark feature of preeclampsia. Further research needs to be performed to determine the extent to which

AM signaling affects trophoblast cells in the process of maternal spiral artery remodeling during pregnancy.

Adrenomedullin Signaling and Cardiovascular Biology

AM has been reported to be upregulated in various cardiovascular conditions1, 75, 76 and is a potent angiogenic factor as well as a cardioprotective factor.1 Plasma AM increases 2-fold in conditions such as essential hypertension, renal failure and congestive heart failure (Figure 1-1).77, 78 Previous studies with gene-targeted KO mice for Adm and

Calcrl indicated that AM signaling is important for cardiovascular development.22, 30, 33

Genetic reduction of Adm results in enhanced cardiovascular damage including increased cardiac hypertrophy in male Adm+/- mice79 and marked perivascular fibrosis, coronary artery intimal hyperplasia and oxidative stress with AngII/high-salt treatment.32 AM protects the

10 heart from hypertrophy and fibrosis during cardiovascular stress such as hypertension and cardiac hypertrophy, myocardial infarction, heart failure and atherosclerosis,80, 81 but the exact mechanisms of AM-mediated cardioprotection have not been fully elucidated. A comprehensive review of the cardioprotective function of AM during hypertension and heart failure has recently been provided by several groups.9, 82

Endothelial dysfunction is characterized by reduced endothelium-dependent vascular relaxation which is associated with most forms of cardiovascular disease. It is partially impacted by reduced nitric oxide and upregulation of adhesion molecules to result in a proinflammatory and prothrombotic state.83 Research also suggests that endothelial dysfunction may act as an early marker of atherosclerosis.84 One study indicated that Adm and its receptor components, Calcrl and Ramps were upregulated in the aorta of apolipoprotein E-deficient (ApoE-/-) mice.85 Loss of apoE ultimately results in a mouse model of spontaneous atherosclerosis because apoE is important in the removal of circulating lipoproteins.86 When these mice were fed an atherogenic diet and treated with AM, the appearance of atherosclerotic lesions was reduced.85 This study further indicates that AM may help to protect against the progression of atherosclerosis, but the exact mechanism for this action remains to be understood. Expression of adhesion molecules in LECs87 and liver sinusoidal endothelial cells88 were reduced in response to AM treatment. Similar results were seen with VEGF-treated HUVECs.89 Thus, AM may impact endothelial dysfunction partially by modulating adhesion molecule expression. With respect to endothelium- dependent vascular relaxation, AM is known to induce vasodilation which is mediated partially by endothelium-derived nitric oxide.90-94 Also, in a rat model of sepsis induced by cecal ligation and puncture, administration of AM and AM-binding protein (AMBP-1 also known as complement factor H) were shown to prevent against endothelial cell dysfunction and decreased endothelium-dependent vascular relaxation in thoracic aorta.95 These studies implicate AM as having a protective role in cardiovascular disease and endothelial

11 dysfunction, but further research needs to be performed to investigate how AM directly impacts on the cardiac endothelial cells to regulate their function.

The Role of Adrenomedullin Signaling in Response to Injury, Vascular Dysfunction and

Wound Healing

Endothelial proliferation and angiogenesis are known to be impacted by AM signaling. In a hind-limb ischemia model, AM promotes endothelial cell proliferation and capillary formation and conversely, Adm+/- mice showed reduced blood flow and capillary development.56 Other whole animal studies using matrigel plugs to assess vascular growth demonstrated the role of AM in vascular regeneration because AM increased blood flow and capillary densities through PKA- and PI3K-dependent pathways.12, 13 AM also induced tube- formation of HUVECs cultured on matrigel.14 Another study pertaining to RAMP2 expression also revealed similar findings. An aortic ring assay and matrigel plug assay with adult Ramp2+/- mice revealed that with decreased RAMP2 expression there was reduced neovascularization in response to growth factor stimulation.18 Collectively, these studies indicate the importance of AM in endothelial cell proliferation and angiogenesis in adult mice.

AM signaling is known to impact the blood and lymphatic vasculature in other physiological processes and pathological conditions. In a pathological mouse model of subcortical vascular dementia (chronic cerebral hypoperfusion), AM was shown to promote arteriogenesis and angiogenesis as well as inhibit oxidative stress and preserve white matter in the brain.96 AM signaling can also induce anti-apoptotic and anti-inflammatory effects in response to injury. In the sinusoidal endothelial cells of the liver, AM helps to protects these cells from cold injury during the process of cold preservation for a liver transplant by decreasing endothelial cell apoptosis and inflammation.88 Conversely, in

Adm+/- and Ramp2+/- mice there is increased apoptosis of the sinusoidal endothelial cells in

12 the liver after cold injury88 further indicating that AM signaling helps to regulate apoptosis.

Wound healing is an essential physiological process that requires angiogenesis and lymphangiogenesis for proper healing. Since AM is a known angiogenic factor and lymphangiogenic factor,22 it is not surprising that AM signaling is necessary in the wound healing process. In an ischemia/reperfusion mouse model of a pressure ulcer, AM administration reduced the wound area and accelerated angiogenesis as well as lymphangiogenesis.97 Also in a wounded HUVEC monolayer, AM promoted vascular regeneration via activation of endothelial Akt in a PKA- PI3K- dependent manner.12

Lymphedema is a hallmark condition of lymphatic dysfunction resulting in the swelling of one or more limbs due to accumulation of interstitial fluid. In Balb/C mice with tail lymphedema,

AM treatment improved lymphedema and increased the number of lymphatic and blood vessels near the injury site.11 Taken together, these data indicate that AM is an essential component for proper endothelial cell function in both physiological and pathological states to regulate apoptosis, inflammation, and lymphangiogenesis as well as angiogenesis.

An important issue to still address is to determine the exact role of AM signaling during adulthood by using temporal and spatial KO mice for components of the AM signaling system to evaluate physiology and function of the vascular beds in these mice. Previous studies with genetic KO mice for the AM signaling system reveal an enhanced impact of AM on lymphatic vascular development relative to blood vascular development.22 It has also been shown that the gene expression of AM receptor components, Calcrl and Ramp2, are enhanced in LECs compared to BECs.23, 24 Due to these known differences of AM signaling between BECs and LECs, it would be interesting to determine whether there is also an enhanced effect of AM on the lymphatic vasculature in adult physiology and pathology. The underlying mechanisms through which AM impacts the lymphatic vasculature, blood vasculature as well as the more specialized cardiac tissue during adulthood also needs to be identified.

Adrenomedullin Expression in Tumor Progression

The AM peptide was initially isolated from a human adrenal tumor

(pheochromocytoma) due to its platelet cAMP elevating activity.76 Since this discovery almost 20 years ago, investigation into the role of AM in tumors has greatly expanded. Early studies noticed elevated levels of AM in lung and brain tumors98, 99 and a comprehensive survey of human tumor cell lines from lung, breast, brain, ovary, colon, and prostate substantiated those reports.100 AM has been implicated in a variety of pro-tumor functions including acting as an autocrine growth factor,100-102 apoptosis survival factor,15 promoter of tumor cell motility and invasion,102-104 and molecular intermediate to enhance communication between tumor cells and immune cell infiltrates105. Furthermore, it has been suggested that the presence of AM in tumors may signify a more aggressive tumor phenotype due to correlation between Adm gene expression and histological tumor grade.102, 106

The mechanism(s) by which Adm gene expression is transcriptionally regulated in tumors remains unclear. It is likely that AM can be both an autocrine and paracrine factor107 by providing tumor cells a growth advantage in addition to acting on surrounding endothelial cells to promote proliferation and changes in vessel permeability to perhaps facilitate metastasis. Moreover, it has been suggested that hypoxia may play a role in AM production.8, 108 Tumors often develop hypoxic zones in areas where blood flow is inadequate to supply the necessary oxygen required for the growing tumor cells. As a result of this unfavorable state, hypoxia inducible factor-1 (HIF-1) is activated which in turn upregulates a number of genes to compensate for the reduced oxygen microenvironment.

Interestingly, a HIF-1 dependent mechanism was found to increase the expression of Adm in hypoxic human tumor cell lines.109 Furthermore, Adm and Calcrl were found to be upregulated in microvascular endothelial cells cultured under low oxygen conditions.110

Together, these results show that both tumor cells and surrounding endothelial cells are

14 responsive to hypoxic conditions and may provide a mechanism for elevated AM levels in a tumor setting.

Although the precise role of AM in tumor development and progression is still unresolved, significant progress has been made to better understand how AM affects not only a tumor cell, but also the endothelial cells in the surrounding microenvironment.

Analysis of immunohistochemical staining of human ovarian cancer found that in addition to tumor cells, AM was also localized to the endothelial cells of the surrounding stroma.106

Furthermore, an in vitro co-culture system found that HUVECs became activated upon exposure to tumor cells and consequently increased transcriptional activity of Adm, among other factors.111 Since AM directly impacts endothelial cell proliferation and permeability,

AM induced modulation of vessels may affect the spread of cancer cells to distant sites via blood or lymphatic vasculature, as shown in Chapter III. Research groups have been performing the in vivo studies necessary to confirm that AM promotes tumor progression through its known angiogenic properties. Several reports have shown that inhibition of AM action by neutralizing antibodies or AM antagonist AM22-52 have reduced the growth of tumor xenografts in vivo by suppressing vascular development.56, 112, 113

While much of the focus in understanding the process of tumor lymphangiogenesis and angiogenesis has been upon the VEGF protein family, the contribution of AM to this process should not be underappreciated. Clearly, the studies described above point to AM as a valid target for potential cancer therapies although more research is necessary.

Generation and validation of preclinical mouse models that are able to rigorously test AM as a target are greatly needed. Since the embryonic lethal phenotype of Adm-/- mice makes studying this signaling pathway more complicated, novel genetic mouse models (Table 1-1) using conditional alleles18, 22, 114 and vascular endothelium specific Cre animals are a starting point for such tumor studies. Furthermore, these mouse models will be needed to refine our understanding of the metastatic process. Given the knowledge that AM can act on both

15 the blood and lymphatic endothelium, a key question that remains to be answered is by what mechanisms do tumor cells disseminate into the blood and/or lymphatic vessels.

Summary and Future Directions

The use of genetic animal models in the field of AM research has produced significant contributions toward understanding the biology of this pleiotropic molecule, with a renewed appreciation for is critical regulation of endothelial cells function during development and vascular diseases. To date, AM has been implicated in lymphatic vascular development, in proper functioning of blood and lymphatic endothelial cells and in a variety of conditions such as pregnancy, cardiovascular disease, and tumor progression

(Figure 1-2). Despite the strides that have been made, there is much more to learn regarding the mechanisms mediating AM function and regulation. With the generation of additional sophisticated molecular biology tools such as genetic mouse models, we are poised to refine our current knowledge as well as discover other novel roles for this peptide and signaling partners in normal and disease physiology.

Author Contributions

N.O.K., S.L.H, and D.O.K. all contributed equally to this manuscript. N.O.K., S.L.H., and D.O.K. drafted layout and edited manuscript. N.O.K wrote the Introduction,

“Adrenomedullin Expression in Tumor Progression”, and the Summary and Future

Directions sections. S.L.H wrote the Physiology and Pathology sections, prepared Tables 1-

1 and 1-2, and the references. D.O.K wrote the Development sections and designed and prepared Figure 1-2. P.M.L. prepared Figure 1-1. K.M.C. oversaw design and edited manuscript.

Tables

Table 1-1: Gene Targeted Mouse Models for Studying Adrenomedullin Signaling Mouse Model Development Result Reference or Adulthood -Embryonic lethal (e14.5), edema, smaller hearts, reduced myocyte proliferation and increased apoptosis, increased left ventricle trabecularization, thinner aorta and carotid artery walls, increased vascular permeability, hypoplastic Adm-/- Development jugular lymph sac 22, 30-32 -Pregnancy: Disrupted fertility, placentation, and fetal growth -Cardiovascular: Increased damage including hypertrophy, reactive oxygen species (ROS), and fibrosis -Liver cold injury: increased apoptosis of the sinusoidal Adm+/- Adulthood endothelial cells 32, 68, 79, 88 Admfl/fl/Tubulin -High anxiety, hyperactive, impaired Tα-1-Cre+ Adulthood motor coordination 114 -Embryonic lethal (e13.5), similar Calcrl-/- Development phenotype as Adm-/- mice 22, 33 -Embryonic lethal (e16.5) and CalcrlLoxP/- recapitulation of Adm-/-, Calcrl-/-, and /Tie2Cre+ Development Ramp2-/- phenotype 22 Ramp2-/- and -Embryonic lethal (e15.5), similar Ramp2fl/fl/ phenotype as Adm-/- mice CAG-Cre+ Development 18, 22, 34 -Increased vascular permeability and decreased neovascularization -Liver cold injury: increased apoptosis of the sinusoidal Ramp2+/- Adulthood endothelial cells 18, 88 -Embryonic lethal (e14.5) and phenocopy of Adm-/-, Calcrl-/-, and Ramp2-/- mice due to loss of PAM-/- Development amidation of AM peptide 35 Adrenomedullin (Adm); Calcitonin receptor-like receptor (Calcrl); Receptor activity modifying protein (RAMP); Peptidylglycine alpha-amidating monooxygenase (PAM)

Table 1-2: Vascular Assays for Studying Adrenomedullin Function Assay Result Reference -Atherogenic diet and AM treatment in ApoE-/- mice resulted in reduced formation of Atherogenic Model atherosclerotic lesions 85 -AM injected mice showed reduced Tail permeability of the dermal lymphatic microlymphography capillaries 25 -AM increased vascular regeneration -Ramp2+/- mice exhibited reduced Matrigel plug neovascularization 12, 13, 18 -Ramp2+/- mice exhibited reduced neovascularization in response to growth Aortic ring factor stimulation 18 -AM+/- mice exhibited increased reactive oxygen species (ROS), vascular fibrosis, and AngII/high-salt intimal thickening 39 Prolonged -AM treatment reduced lung vascular mechanical permeability resulting from ventilator use ventilation 21 Chronic cerebral -AM promoted arteriogenesis and hypoperfusion angiogenesis 96 -AM promotes endothelial cell proliferation and capillary formation -Adm+/- mice showed reduced blood flow and Hind-limb ischemia capillary development 56 Wound healing - AM reduced wound area and increased (Pressure Ulcer - angiogenesis and lymphangiogenesis Ischemia reperfusion model) 97 -AM improved lymphedema and increased Tail lymphedema number of lymph and blood vessels 11 -Blocking AM signaling results in reduced Tumor xenografts vascular development 56, 112, 113

Figures

Figure 1-1. Fold Change in Plasma Adrenomedullin Levels in a Variety of Human Conditions. Bars indicate average fold change in circulating AM levels in various disease categories or conditions based on published human clinical data. The dashed horizontal line at 2.33 represents the average fold increase in plasma AM levels across all conditions depicted. Number above each bar indicates the number of published observations assessing plasma AM levels in each category. The clinical papers that were used for our analysis are listed according to the following broad categories: cancer,115-120 cardiovascular,4, 115, 117, 121-148 hepatic and renal,128, 129, 131, 132, 149-160 pulmonary, 6, 129, 161-166 infectious & autoimmune,167-173 normal pregnancy,61, 62, 64, 65, 174-178 and pregnancy complications.61, 65, 178-190

Figure 1-2. Adrenomedullin Signaling in Development and Vascular Biology. (A) Loss of AM signaling causes embryonic lethality due to severe edema associated with impaired lymphatic vascular development. (B) In the adult, AM is an angiogenic, lymphangiogenic, and a cardioprotective factor that also regulates vascular permeability and inflammation. Expression of AM is also implicated in pregnancy and tumor progression. (LEC = lymphatic endothelial cell)

Chapter II: Endothelial restoration of receptor activity-modifying protein 2 is sufficient to rescue lethality, but survivors develop dilated cardiomyopathy2

Overview

Receptor activity-modifying proteins (RAMPs) serve as oligomeric modulators for numerous G protein-coupled receptors (GPCRs), yet elucidating the physiological relevance of these interactions remains complex. Receptor activity-modifying protein 2 (Ramp2) null mice are embryonic lethal, with cardiovascular developmental defects similar to those observed in mice null for canonical adrenomedullin/calcitonin receptor-like receptor (CLR) signaling. We aimed to genetically rescue the Ramp2-/- lethality in order to further delineate the spatiotemporal requirements for RAMP2 function during development and thereby enable the elucidation of an expanded repertoire of RAMP2 functions with Family B GPCRs in adult homeostasis. Endothelial-specific expression of Ramp2 under the VE-cadherin promoter resulted in the partial rescue of Ramp2-/- mice, demonstrating that endothelial expression of Ramp2 is necessary and sufficient for survival. The surviving Ramp2-/- Tg animals lived to adulthood and developed spontaneous hypotension and dilated cardiomyopathy, which was not observed in adult mice lacking CLR. Yet, the hearts of

Ramp2-/- Tg animals displayed dysregulation of Family B GPCRs, including parathyroid hormone and glucagon receptors as well as their downstream signaling pathways. These data suggest a functional requirement for RAMP2 in the modulation of additional GPCR

2 Reprinted with permission from: Daniel O. Kechele, William P. Dunworth, Claire E. Trincot, Sarah E. Wetzel-Strong, Manyu Li, Hong Ma, Jiandong Liu, Kathleen M. Caron. Endothelial restoration of receptor activity-modifying protein 2 is sufficient to rescue lethality, but survivors develop dilated cardiomyopathy. Hypertension. In Press, 2016.

21 pathways in vivo, which is critical for sustained cardiovascular homeostasis. The cardiovascular importance of RAMP2 extends beyond the endothelium and canonical adrenomedullin/CLR signaling, in which future studies could elucidate novel and pharmacologically-tractable pathways for treating cardiovascular diseases.

Introduction

Receptor activity-modifying proteins (RAMPs) are single-pass transmembrane proteins that physically interact with numerous G protein-coupled receptors (GPCRs) to regulate receptor trafficking, ligand binding specificity and downstream G protein coupling and signaling. Biochemical and pharmacological studies have revealed functional RAMP interactions with many GPCRs including, calcitonin receptor-like receptor (CLR = protein,

Calcrl = gene), calcitonin receptor (CTR), parathyroid hormone receptor 1 and 2 (PTHR1 and PTHR2), glucagon receptor (GCGR), secretin receptor (SCTR), vasointestinal peptide receptor 1 and 2 (VIPR1 and VIPR2), calcium sensing receptor (CaSR), an estrogen receptor (GPR30), and likely additional RAMP-interacting GPCRs will be identified in the future.7, 191-193 Thus, in their capacity as endogenous oligomeric GPCR partners, RAMPs can exert powerful modulation on nearly every aspect of GPCR function and on a wide range of physiological processes. However, our understanding regarding the complexity and consequences of RAMP-GPCR interactions in vivo lags behind our current structural, biochemical and pharmacological knowledge and thus presents a limitation toward harnessing the potential pharmacological utility of this family of proteins.

There exist three mammalian RAMPs, each encoded by single genes, which have been systematically knocked out by our laboratory and others.18, 34, 194 As expected, based on their broad tissue distribution, the phenotypes of Ramp null mice are extensive and largely reflect their interactions with multiple GPCRs.195 Ramp1 and Ramp3 null mice survive to adulthood, while Ramp2-/- mice exhibit embryonic lethality at mid-gestation due to

22 cardiovascular defects in the heart, blood, and lymphatic vasculatures.18, 22, 34, 194, 196, 197

These phenotypes are essentially identical to those observed in mice lacking the genes that encode for CLR and its Ramp2-associated ligand, adrenomedullin (AM), indicating that

RAMP2 is essential for canonical CLR/AM signaling during embryonic development. Yet, mice that are haploinsufficient for Ramp2 also exhibit an expanded constellation of endocrine-related phenotypes that are not observed in CLR and AM haploinsufficient mouse models.198 Therefore, these data imply that RAMP2 must exert functional modulation of other GPCRs, which is supported by its in vitro biochemical interaction with CTR, PTHR1/2,

GCGR, and VIPR1/2.191 To date, ascribing a functional relevance for these other putative

RAMP2-interacting GPCR pathways in adult physiological systems has been precluded by the embryonic lethality of Ramp2-/- mice.

Here, we have developed a novel Ramp2 transgenic mouse model which overexpresses Ramp2 specifically in the endothelium, with the hypothesis that restoration of

RAMP2 in the endothelium will rescue the lethality-causing vascular defects associated with loss of CLR/AM signaling during embryogenesis. Indeed, endothelial RAMP2 is sufficient to rescue the embryonic lethality of many Ramp2-/- mice. Thus, the surviving mice, which express Ramp2 in the endothelium but lack Ramp2 in every other cell type, provided us with the opportunity to further elucidate the functional implications of the loss of Ramp2 on adult cardiovascular physiology and respective changes in other Ramp2-associated GPCRs.

Methods

Mouse Studies

Previously published SvEv-Ramp2+/- and Calcrlflox/flox mice were both fully backcrossed (>10 generations) onto a C57BL/6J genetic background for these studies.22, 34

A novel endothelial-specific Ramp2 transgenic mouse was generated and crossed to the

Ramp2+/- mice. All Ramp2 mice used in this study were maintained on an isogenic

C57BL/6J genetic background. The Calcrlflox/flox mice were crossed with either the SJL-

Tg(Tagln-cre)1Her (SM22-Cre) or the inducible CAGG-CreERTM mouse to conditionally delete Calcrl specifically within developmental vascular smooth muscle cells (vSMC) or ubiquitously in adult mice, respectively. Adult CAGG-CreERTM females were administered with tamoxifen (TAM) as previously published.199 Genotyping and RT-PCR primers and probes are listed in Supplemental Table 2-S1. Females between 4-8 months of age were used in all Ramp2-associated studies, males 3-4 months of age for vSMC-specific Calcrl studies, and females 16 month old for ubiquitous Calcrl studies. Mice were acclimated and conscious for both non-invasive tail cuff blood pressure and echocardiography analysis and anesthetized with isoflurane for intra-arterial blood pressure measurements. Organs weights and heart chamber dissections were normalized to either body weight or tibia length as previously described.200 Biological N of 3-10 mice per genotype was used for each experiment. Endothelial cells from adult mice were isolated with magnetic-associated cell sorting using CD31-specific antibodies as previously described.201 Cardiomyocytes from adult mice were isolated using collagenase digestion as previously described.202 All animal experiments were approved by the Institutional Animal Care and Use Committee of The

University of North Carolina at Chapel Hill.

Statistical Analysis

Statistical analysis was determined with GraphPad 5.0 and data is represented as mean ± SEM. The unpaired student T-test was used to compare two groups, while one way

ANOVA with Tukey’s Multiple Comparison test was used to compare ≥ 3 groups. Survival data were compared by Mantel-Cox and Gehan-Breslow-Wilcoxon tests. Significant differences are represented as * p <0.05, ** p <0.01, ***p <0.001.

Additional detailed methods are provided in the Online Supplemental Data.

Results

Generation and Characterization of Endothelial-Specific Ramp2 Transgenic Animals

A diagram of the Cdh5-Ramp2 transgene (Tg) depicts the murine vascular cadherin

5 (Cdh5) promoter driving the expression of a flagged-tagged murine Ramp2 cDNA

(Supplemental Figure 2-S1A), which was successfully integrated into the genome of two independent founder lines on the C57BL/6J genetic background. Expression of FLAG-

Ramp2 protein was confirmed using both transgenic founder lines in a variety of adult tissues including the heart, kidney, lung and intestine (Supplemental Figure 2-S1B). Semi- quantitative RT-PCR revealed the presence of FLAG transcript within whole adult lung tissue and CD31+ endothelial-enriched cells in the Tg animals, but not wild-type endothelium (Supplemental Figure 2-S1C), which resulted in a modest increase in overall

Ramp2 gene expression levels within lung CD31+ endothelial cells of Tg animals compared to wild-type animals (Supplemental Figure 2-S1D). Ramp2 expression was significantly higher in the left ventricles of Tg animals compared to wild-type—a finding which we attribute to the endothelial-driven transgene expression, as compared to the low levels of endogenous Ramp2 expression in isolated cardiomyocytes (Supplemental Figure 2-S1E) and previously reported expression in vSMCs.92, 203 Tg(Cdh5-Ramp2) mice bred normally, with expected Mendelian and sex ratios at birth, and no obvious phenotypic defects.

Transgenic Endothelial Ramp2 Partially Rescues Embryonic Lethality of Ramp2-/- Mice

We sought to determine whether endothelial Ramp2 expression could rescue the previously reported Ramp2-/- embryonic lethality by interbreeding the hemizygous Tg(Cdh5-

Ramp2) animals with Ramp2+/- non-transgenic (ntg) animals, on an isogenic C57BL/6J genetic background.18, 22 Ramp2+/- Tg(Cdh5-Ramp2) mice (hereafter referred to as Ramp2+/-

Tg) were crossed to Ramp2+/- ntg animals to generate Ramp2-/- embryos with or without expression of the transgene. At e14.5, all six possible genotypes were observed

(Supplemental Figure 2-S2A). As expected, embryos lacking Ramp2 display non- hemorrhagic edema associated with arrested lymphangiogenesis22, yet the Ramp2-/- Tg embryos showed significantly less edema than Ramp2-/- ntg littermates (Figure 2-1A-B).

Unlike the small, hypoplastic jugular lymph sacs of Ramp2-/- ntg embryos, the jugular lymph sacs of the Ramp2-/- Tg were comparable in size to those of Ramp2+/+ littermates (Figure 2-

1C-D), demonstrating that transgenic restoration of functional Ramp2 in the endothelium leads to improvement of the edematous phenotype.

In addition to defects in lymphatic development, loss of Ramp2 or AM/CLR signaling leads to small, disorganized hearts and thin vSMC walls.22, 30, 33 The aortic endothelium of viable embryos appeared intact with no signs of hemorrhagic leakage or endothelial dysfunction as previously reported in some Ramp2-/- embryos.18 However, the aortic vSMC layer was significantly thinner in Ramp2-/- ntg and Ramp2-/- Tg embryos (Figure 2-1E-F).

Likewise, the hearts of e14.5 Ramp2-/- Tg animals were comparable in size to Ramp2-/- ntg littermates, which were both significantly smaller than controls, despite detectable expression of Ramp2 in Ramp2-/- Tg hearts (Figure 2-1G-I). Consistent with previous reports, these data reflect that Ramp2 expression is required for normal vSMC and cardiac development and that endothelial Ramp2 restoration is not sufficient to rescue these defects, thus confirming non-endothelial roles of Ramp2 during cardiovascular development.

Unlike the Ramp2-/- ntg embryos, which uniformly die by e15.5, Ramp2-/- Tg embryos survive to term at near-Mendelian ratios, but a large number of these animals were found stillborn at postnatal day 1 (Supplemental Figure 2-S2B). Nevertheless, by postnatal day 7,

40% of the expected Mendelian ratio of Ramp2-/- Tg pups were viable, which represents a significant survival rescue when compared to the completely penetrant lethality of Ramp2-/- ntg mice (Table 2-1). These surviving Ramp2-/- Tg mice were indistinguishable from

Ramp2+/+ littermates (Supplemental Figure 2-S2C). Interestingly, there was a significant skewing in the sex ratio of the surviving Ramp2-/- Tg mice, such that 78% were female.

These data demonstrate that transgenic endothelial restoration of Ramp2 is able to blunt the endothelial and edematous phenotypes observed with global Ramp2 genetic deletion, leading to significantly improved survival.

Surviving Ramp2-/- Tg Adult Mice are Hypotensive

The surviving Ramp2-/- Tg mice, which express Ramp2 in the endothelium but lack

Ramp2 in all other cells, provided an opportunity to evaluate how Ramp2 loss-of-function during development affects adult cardiovascular homeostasis. Diastolic, systolic, and mean arterial blood pressures were significantly reduced in Ramp2-/- Tg compared to Ramp2+/+ ntg and Tg female controls (Figure 2-2A). Importantly, there were no significant differences in basal blood pressure between Ramp2+/+ ntg and Ramp2+/+ Tg mice, indicating that the decreased blood pressure was not caused solely by the transgene. Moreover, the reduced basal blood pressure of Ramp2-/- Tg mice was observed without the exogenous administration of the canonical, hypotensive ligand AM, which was previously required to lower blood pressures in a vSMC-specific overexpression model of Ramp2.47, 204

Similar to the developmental defects in vSMC wall thickness of the Ramp2-/- Tg embryos and other models lacking AM/CLR/Ramp2 function22, 30, 33, the vSMC walls were significantly thinner in Ramp2-/- Tg adult descending aortas compared to wild-type controls

(Figure 2-2B). Intra-arterial blood pressure measurements further confirmed the hypotensive phenotype of adult Ramp2-/- Tg females (Figure 2-2C). To test vSMC responsiveness, Ramp2-/- Tg and controls were challenged with intravenous injections of the

-adrenergic receptor agonist, phenylephrine, and the vasodilator, AM. The Ramp2-/- Tg adults were capable of normal vasoconstriction and vasodilation as compared to controls

(Figure 2-2C). So, although the thinner vSMC walls likely play a role in the hypotensive phenotype of Ramp2-/- Tg mice, additional systemic changes likely contribute to the hypotension.

Ramp2-/- Tg Mice Develop Spontaneous Dilated Cardiomyopathy Phenotype

Considering the hypotension and the developmental defects observed in Ramp2-/- Tg embryonic hearts, we next assessed how cardiac function and morphology were altered in adult Ramp2-/- Tg mice. Echocardiography on conscious mice revealed significantly dilated left ventricles during both diastole and systole, with significantly larger left ventricle volumes and left ventricle internal diameter dimensions in the Ramp2-/- Tg mice compared to

Ramp2+/+ and Ramp2+/-, with and without the transgene (Table 2-2). Representative M- mode echocardiograms from Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg illustrate the ventricular dilation in Ramp2-/- Tg mice (Figure 2-3A), which resulted in a trending increase in calculated cardiac output in these animals (Table 2-2). Septum and left ventricle posterior wall dimensions of Ramp2-/- Tg mice were unchanged from controls, although there was a non-significant trend towards a thinner posterior wall. There was also a modest, but significant reduction in both ejection fraction and fractional shortening in the Ramp2-/- Tg dilated hearts. These data indicate that loss of Ramp2 in non-endothelial cells leads to a spontaneous dilated cardiomyopathy (DCM)–like phenotype in adult mice, which at 6 months of age had not yet progressed to heart failure, as indicated by the elevated cardiac output and sufficient heart function.

Upon dissection, we observed that the hearts of Ramp2-/- Tg mice were grossly enlarged compared to wild-type and Ramp2-/- ntg mice (Figure 2-3B-C). The adult Tg mice had no differences in body weight (Ramp2+/+ ntg: 28.4 ± 1.4 g, Ramp2+/+ Tg: 28.4 ± 1.2 g,

Ramp2-/- Tg: 28.3 ± 1.2 g) or tibia length (Ramp2+/+ ntg: 17.6 ± 0.2 mm, Ramp2+/+ Tg: 17.9 ±

0.1 mm, Ramp2-/- Tg: 17.8 ± 0.2 mm) regardless of their Ramp2 genotype. Yet, total heart weights were significantly larger in the Ramp2-/- Tg mice when normalized to both body weight (Figure 2-3D) or tibia length (Ramp2+/+ ntg: 7.3 ± 0.4 mg/mm, Ramp2+/+ Tg: 6.5 ± 0.2 mg/mm, Ramp2-/- Tg: 9.0 ± 0.3** mg/mm; **p<0.01). Moreover, when normalized to body weight the left ventricle (Figure 2-3E), right ventricle (Figure 3F), and right atria (Ramp2+/+

28 ntg: 0.15 ± 0.01 mg/g, Ramp2+/+ Tg: 0.15 ± 0.01 mg/g, Ramp2-/- Tg: 0.21 ± 0.02* mg/g;

*p<0.05) all exhibited significant enlargement in the Ramp2-/- Tg animals compared to all other genotypes. Similar significant trends were observed when the individual chamber weights were normalized to tibia length (LV:TL; Ramp2+/+ ntg: 5.4 ± 0.3 mg/mm, Ramp2+/+

Tg: 4.8 ± 0.2 mg/mm, Ramp2-/- Tg: 6.5 ± 0.2** mg/mm; **p<0.01).

Cross-sectional area of myocytes within the left ventricle revealed slight, but significant cardiomyocyte hypertrophy, with no changes in left ventricular capillary density, in the Ramp2-/- Tg mice compared to controls (Figure 2-4A-D). Picrosirius red staining showed no differences in perivascular fibrosis, but there was significantly increased interstitial fibrosis in the Ramp2-/- Tg hearts compared to hearts of Ramp2+/+ and Ramp2+/+ Tg mice

(Figure 2-4E-H). There were elevated levels of the oxidative stress indicator, lipid peroxidase evidenced by 4-hydroxynonenal (4-HNE) staining in Ramp2-/- Tg hearts (Figure

2-4I-J).205 Together, the modest increases in hypertrophy, fibrosis, and oxidative stress, as well as modest decline in heart function, further support that 6 month old Ramp2-/- Tg mice exhibit a compensated, DCM-like phenotype.

Adult Ramp2-/- Tg Mice Develop Multi-Organ Inflammation

DCM and hypotension can lead to vascular congestion and organ dysfunction throughout the body.206-208 Consistently, we observed that the spleen to body weight ratio was significantly increased in the Ramp2-/- Tg mice compared to all other genotypes (Figure

2-5A). Ramp2-/- Tg mice also exhibited macroscopic vascular congestion within their livers compared to control animals (Figure 2-5B). Furthermore, multi-organ histology revealed a marked increase in the number of inflammatory foci, particularly surrounding the vasculature, within the liver (Figure 2-5C-D), kidney (Figure 5E-F) and lungs (Figure 2-5G-

H). A diagnostic profile of adult serum from wild-type and Ramp2-/- Tg mice revealed few significant changes in circulating ions or enzyme levels that are typically indicative of

29 hepatocellular or renal damage (Supplemental Table 2-S2). This supports that the end- organ inflammation in Ramp2-/- Tg mice is likely downstream of altered hemodynamics due to DCM and hypotension, rather than primarily due to loss-of-function of Ramp2 in end- organs.

Conditional Calcrl Deletion Does Not Lead to a Dilated Cardiomyopathy Phenotype

Since CLR and AM represent the most well-characterized, canonical pathway for

RAMP2 modulation and the knockout mice for these genes recapitulate the Ramp2-/- developmental phenotypes, it is reasonable to consider that disruption of this pathway, which has been demonstrated to be cardioprotective in both animal studies and in humans32,

77, 81, 209, may underlie the DCM and hypotensive phenotypes in Ramp2-/- Tg mice. To test the functions of CLR in non-endothelial cells, we generated cardiac- and vSMC-specific

Calcrl null mice using the SM22-Cre mediated excision. The CalcrlloxP/loxP; SM22Cre+ mice were born at expected Mendelian ratios and survived to adulthood (Supplemental Figure 2-

S3A). Relative Calcrl expression was significantly reduced in aortic vSMCs and hearts of

CalcrlloxP/loxP; SM22-Cre+ compared to CalcrlloxP/+; SM22-Cre+ littermates (Supplemental

Figure 2-S3B). Importantly, CalcrlloxP/loxP; SM22-Cre+ adult males were normotensive and had no basal changes in heart function, size, or morphology (Supplemental Table 2-S3 or

Figure 2-S3C). Similarly, cardiac-specific Calcrl deletion using the αMHC-Cre+ transgenic line also failed to recapitulate the DCM phenotype, with animals surviving to adulthood with no basal cardiac dysfunction (Dackor R & Caron K.M., unpublished data, [2016]). These data demonstrate that cardiac- and vSMC-specific Calcrl expression is not required for embryonic development or adult cardiovascular maintenance.

We have previously shown that temporal, global deletion of Calcrl in adult

CalcrlloxP/loxP; CAGG-CreERTM mice results in dilated lymphangiectasia in many lymphatic vascular beds throughout the body.199 However, these mice did not display any obvious

30 cardiac phenotypes and they had similar heart to body weight ratio as TAM-injected control mice (CalcrlloxP/loxP; CAGG-CreERTM: 4.43 ± 0.19 mg/g versus Calcrlflox/flox: 4.13 ± 0.14 mg/g, respectively). Furthermore, conscious echocardiography of CalcrlloxP/loxP; CAGG-CreERTM female mice, even at 14 months of age, failed to reveal any significant changes in left ventricle internal diameter, function, or heart size compared to control mice (Supplemental

Table 2-S4). Collectively, these data, generated from three independent series of conditional deletion approaches, indicate that the global-, cardiac- or vSMC-specific loss of

Calcrl does not recapitulate the DCM phenotype observed in the Ramp2-/- Tg animals.

Therefore, this strongly suggests that the Ramp2-/- Tg phenotype is imparted by other

RAMP2-associated GPCR pathways.

Ramp2-/- Tg hearts exhibit reduced signaling pathways and expression of numerous RAMP- associated GPCRs

The genetic reduction or down-regulation of the transcriptional regulator CREB207, 210 and the crucial PPAR pathway transcription factor, Pgc-1α211-213, have both been shown to be involved in DCM pathogenesis. Thus as expected, the relative expression of Pgc-1α was significantly down-regulated in left ventricles and in an enriched cardiomyocytes fraction of

Ramp2-/- Tg hearts compared to controls (Figure 2-6A-B). There was also a significant reduction in phosphorylated CREB compared to total CREB and Gapdh in the Ramp2-/- Tg hearts compared to Ramp2+/+ ntg and Ramp2+/+ Tg hearts (Figure 2-6C). Numerous

Family B GPCRs signal through these pathways, including the Gcgr which has been shown to signal through cAMP to activate both CREB and PPAR transcription214-217, and both the

Pthr1 and CaSR signal through cAMP and CREB.218-220 Interestingly, the gene expression levels of these RAMP-associated GPCRs were significantly down-regulated in the left ventricles of Ramp2-/- Tg mice compared to those of control animals, whereas genes

31 encoding for other RAMP-associated GPCRs, like Calcr, Vipr1, and Gpr30, were unchanged

(Figure 2-6D).

Similar gene expression changes in these Family B GPCR expression profiles were confirmed in isolated cardiomyocyte fractions (Figure 2-6E), further supporting the myocyte- specific genetic dysregulation. Calcrl expression was significantly increased in left ventricle, but not in the cardiomyocyte-enriched fraction, demonstrating that Calcrl upregulation is from a non-cardiomyocyte cell type. The expression of both Pthr1 and Gcgr were significantly decreased during development in the hearts of Ramp2-/- ntg and Ramp2-/- Tg embryos (Figure 2-6F), supporting that these changes are specific to Ramp2 loss rather than secondary to the Cdh5-driven Ramp2 Tg or the DCM phenotype, as might be the case for CaSR. Serum analysis revealed no significant dysregulation of circulating calcium or glucose, thereby eliminating uncompensated Pthr1 or Gcgr systemic signaling as a cause for the cardiovascular phenotypes (Supplemental Table 2-S2). Collectively, these data demonstrate that genetic loss of Ramp2 in non-endothelial cells of the heart leads to downregulation of RAMP2-associated GPCRs, Gcgr and Pthr1, as an underlying mechanistic basis for the decreased pCREB and Pgc-1α responsible for the pathogenesis of the DCM- like phenotype in Ramp2-/- Tg mice (Figure 2-6G).

Discussion

In this study, we generated an endothelial-specific Ramp2 Tg mouse model to attempt to rescue the embryonic lethality due to global loss of Ramp2. While no Ramp2 null mice survive to birth, we observed a significant number of Ramp2-/- Tg born and able to survive into adulthood. This result further confirms that endothelial Ramp2 is essential for embryonic survival and represents, to our knowledge, the first genetic rescue of the global

Ramp2 null lethality.

It remains unclear why a significant number of Ramp2-/- Tg pups survive to late- gestation, but are stillborn. Interestingly, it was recently shown that mice with endothelial excision of Ramp2 during development using a Cdh5-Cre died during late-gestation.206

Additionally, they report that approximately 5% of the endothelial Ramp2 knockouts live into adulthood and develop large hearts, hypotension, and multi-organ vasculitis. In this current study we found that 40% of Ramp2-/- Tg survived to adulthood, and also developed similar cardiovascular and inflammatory phenotypes. These two mouse models, as well as a previously published endothelial-specific deletion of Calcrl, demonstrate that adequate levels and timing of endothelial Ramp2/CLR/AM signaling are critical for embryonic survival.22

In addition, these studies identify potential sex-dependent mechanisms of Ramp2, or

Cdh5, regulation and function, as evidenced by the substantially reduced numbers of male

Ramp2-/- Tg survivors. We have previously shown that adult Ramp2+/- females have endocrine phenotypes not present in Ramp2+/- males.198 In addition, Ramp3-/- males, but not females, displayed exacerbated cardiovascular phenotypes when challenged with hypertension.193, 196 Thus, while RAMP2 and RAMP3 interact with both similar and different

GPCRs, our observations of sex-dependent phenotypes in these genetic animals will provide an area for interesting future investigations.

The complex compensatory mechanisms through which RAMP-mediated AM/CLR signaling regulate blood pressure have not been fully elucidated. It is well documented that

AM infusion acts to lower blood pressure through both vSMC and endothelium, but it remains disputed if altered AM expression using genetic models alters basal blood pressure.31, 79, 91, 92, 221 Moreover, while developmental loss of Ramp2 leads to spontaneous hypotension and adult Ramp1-/- mice develop hypertension.197, 206 Calcrl deletion in vSMC and cardiomyocytes appears dispensable for regulation of basal vascular tone in adult males. It is evident that Ramp2 and AM signaling plays a role in vSMC and myocardium

33 development and function, which likely contributes to the pathogenesis of hypotension and

DCM in Ramp2-/- Tg survivors. Yet, the hypotension and DCM phenotypes occur despite maintenance of normal cardiac output. It is further possible that the developmentally- induced thin vSMC walls contribute to reduced vascular tone; however we demonstrated their ability to effectively respond to acute phenylephrine vasoconstriction. Therefore, additional studies that explore the entire repertoire of RAMP-mediated GPCRs will be required for full elucidation of the mechanistic basis for the hypotension in Ramp2-/- Tg survivors.

Loss of Ramp2 in multiple non-endothelial cardiac cells, along with altered humoral signaling could lead to cardiac dysfunction and DCM pathogenesis. The genetic dysregulation in isolated cardiomyocytes suggests roles of Ramp2/GPCRs specifically in cardiomyocytes. Interestingly, a recently published study demonstrated that a cardiomyocyte-specific Ramp2 deletion led to a DCM-like phenotype due to mitochondrial dysfunction and irregular calcium handling, which the authors attribute to loss of CLR/AM signaling.222 Similarly, this study suggests that cardiomyocyte loss of Ramp2 in the Ramp2-/-

Tg mice is likely responsible for the DCM. Moreover, our data demonstrates that neither cardiomyocyte- or vSMC-specific loss of Calcrl during development or conditional Calcrl deletion in adults recapitulate the DCM phenotype observed in the Ramp2-/- Tg and the aforementioned Ramp2flox/flox; MHC-MerCreMer mice.222 Therefore, collectively, these studies imply the involvement of other RAMP-associated GPCRs.

It is apparent that RAMPs interact with numerous GPCRs biochemically, although in vivo physiologic evidence of these interactions is limited. We found that lack of Ramp2 in non-endothelial cells leads to decreased expression of cardiac Pthr1 and Gcgr.

Interestingly, human PTHR1 and GCGR interact specifically with RAMP2 and not RAMP1 or

RAMP3 in which RAMP2 is important in chaperoning both GPCRs to the plasma membrane.191 Furthermore, it was recently demonstrated that RAMP2 was important in

GCGR ligand selectivity between glucagon and glucagon-like peptide 1, which have opposing physiologic effects on glucose homeostasis and cardiovascular function.223 While

GCGR has not been directly connected to DCM pathogenesis, glucagon can alter calcium signaling in myocytes and glucagon-like peptide 1 improves glucose uptake and survival of canines with DCM.214, 224 Likewise PTHR1 signaling is important in vitamin D and calcium homeostasis, which have both been associated with DCM.225, 226 Together, Ramp2 loss-of- function can not only alter both Pthr1 and Gcgr expression simultaneously, but may also alter their ability to reach plasma membrane, bind ligands, and signal. Additionally, we observed decreased signaling through CREB and downregulation of Pgc-1α, which both have been shown to lead to DCM.207, 210, 211 Both Pthr1 and Gcgr signal through CREB and murine deletion of Gcgr led to decreased phosphorylated CREB and Pgc-1a expression levels.215, 217 Thus, loss of Ramp2 likely has numerous mechanisms for the development and maintenance of cardiac functions both dependent and independent of canonical

AM/CLR/Ramp2 signaling.

Perspectives

The in vivo interplay between GPCR/RAMP/Ligand is highly complex and is only starting to be understood. However, a better understanding of these interactions in a spatial and temporal manner in a pathophysiologic context will help us better target the unique

GPCR/RAMP interfaces to potentially treat disease like DCM and hypertension.

Novelty and Significance

What Is New?

 Using a novel genetic mouse model, we find that endothelial expression of receptor

activity-modifying protein 2 (Ramp2) is able to rescue the embryonic lethality of

Ramp2-/- mice.

 Ramp2 loss-of-function leads to spontaneous hypotension, dilated cardiomyopathy,

and multi-organ inflammation, which we show are not recapitulated in genetic models

with loss of calcitonin-receptor-like receptor (CLR) signaling.

 In vivo loss of Ramp2 causes dysregulation of numerous RAMP-associated GPCRs,

including the glucagon and parathyroid hormone receptors.

What is Relevant?

 The physiological consequences of RAMP interactions with Family B GPCRs have

been challenging to address due to the embryonic lethality of RAMP2 mice. Here,

we elucidate the importance of RAMP2 interactions with several GPCR pathways,

which can ultimately provide novel targets, or predict off-target consequences, for

pharmacological therapies against cardiovascular disease.

Summary

This study provides genetic in vivo evidence that endothelial Ramp2 expression is necessary and sufficient to rescue the lethality of global loss of Ramp2. Ramp2 expression in non-endothelial cells during development is required to maintain adult blood pressure and cardiac homeostasis—a process that involves numerous GPCR signaling pathways, including glucagon and parathyroid hormone receptor signaling. Collectively, these studies extend the functional repertoire of RAMP-associated receptors in cardiovascular physiology.

Author Contributions

D.O.K. designed research study, conducted experiments in Table 2-1, 2-2, 2-S2 and

Figures 2-1, 2-2, 2-3, 2-4, 2-5, 2-6, 2-S1, and 2-S2, analyzed data, prepared and wrote manuscript. W. P. D conducted and analyzed experiments in Table 2-S3 and Figure 2-S3.

C.E.T. contributed data to Table 2-2, Figure 2-3, and 2-4, as well as assisted in manuscript editing. S.E.W-S. contributed data to Table 2-S4 and Figures 2-3, 2-S1, analyzed data, and assisted in manuscript editing. M. L. generated Tg mouse model and contributed data to

Figure 2-S1. H.M. isolated cardiomyocytes for data in Figure 2-6. J.L. provided intellectual support. K.M.C. designed research study and edited manuscript.

Tables

Mendelian ntg Tg(Cdh5-Ramp2) Ratio Ramp2+/+ Ramp2+/- Ramp2-/- Ramp2+/+ Ramp2+/- Ramp2-/- Expected N 35 70 35 35 70 35 Actual N P7 47 88 0 48 83 14* Actual/ 134% 126% 0% 137% 119% 40% Expected

Table 2-1: Endothelial restoration of Ramp2 can rescue the global Ramp2-/- embryonic lethality. Breeding results, both expected Mendelian and actual results from Ramp2+/- Tg(Cdh5-Ramp2) crossed with Ramp2+/- ntg. N = 280 pups. Significance in survival between Ramp2-/- ntg and Ramp2-/- Tg pups determined by the Mantel-Cox test: *p<0.001.

Echo ntg Tg (Cdh5-Ramp2) Parameters Ramp2+/+ Ramp2+/- Ramp2+/+ Ramp2+/- Ramp2-/- Mouse 10 4 7 4 7 Number Mouse Age 26.7±1.3 26.8±2.0 29.3±2.3 26.3±2.2 29.7±1.6 (wks) Heart Rate 624±15 578±38 574±50 614±55 552±21 (BPM)

LV Vol, d (µL) 35.6±4.7 32.0±3.9 34.1±6.7 27.9± 5.9 73.4±9.5*,†,‡,§

LV Vol, s (µL) 7.6±1.6 5.0±1.4 7.0±2.7 5.4±2.4 23.1±3.5*,†,‡,§

CO (mL/min) 16.5±2.1 15.4±1.2 16.1±3.2 13.2±0.9 27.0±4.8*

EF (%) 81.2±3.0 85.1±2.0 82.4±3.6 83.3±5.1 68.3±2.3*,†,‡,§

FS (%) 51.6±2.7 52.9±2.2 51.1±4.0 52.4±6.5 37.9±1.8*,‡

IVS, s (mm) 1.13±0.05 1.16±0.06 1.12±0.07 1.07±0.07 1.08±0.07

IVS, s (mm) 1.72±0.06 1.74±0.06 1.73±0.07 1.68±0.13 1.60±0.07

LVID, d (mm) 2.94±0.20 2.88±0.14 2.90±0.22 2.70±0.23 4.03±0.22*,†,‡,§

LVID, s (mm) 1.52±0.17 1.36±0.13 1.45±0.21 1.32±0.27 2.50±0.14*,†,‡,§

LVPW, d 1.54±0.10 0.96±0.07 1.22±0.13 1.08±0.30 0.87±0.06 (mm) LVPW, s 1.74±0.13 1.72±0.04 1.76±0.08 1.60±0.20 1.40±0.08 (mm)

Table 2-2: Ramp2-/- Tg adults develop left ventricle dilatation and declining heart function. Echocardiography analysis from Ramp2+/+, Ramp2+/-, and Ramp2-/- mice with and without the transgene. BPM, beats per minute; LV, left ventricle; d, diastole; s, systole; CO, cardiac output; EF, ejection fraction; FS, fractional shortening; IVS, interventricular septal; LVID, left ventricle internal diameter; LVPW, left ventricle posterior wall. Data represented as averages ± SEM with significance of p<0.05 determined by one-way ANOVA with Tukey’s Multiple Comparison. Significant compared to: *Ramp2+/+ ntg, †Ramp2+/- ntg, ‡Ramp2+/+ Tg, §Ramp2+/- Tg.

Figures

Figure 2-1: Endothelial restoration of Ramp2 partially rescues embryonic edema leading to prolonged survival of Ramp2-/- mice. (A) Representative images and (B) quantification of edema severity from e14.5 Ramp2+/+, Ramp2-/- ntg, and Ramp2-/- Tg embryos. Edema scoring system: (1) no edema, (2) mild edema, and (3) severe edema. (C) Immunohistochemistry and (D) quantification of jugular lymph sac size using Lyve1 (green) and DAPI (magenta). CA, carotid artery; JV, jugular vein; JLS, jugular lymph sac. (E) Histology of descending aortas and (F) quantification of aortic vSMC wall thickness. (G) Heart histology, (H) quantification of ventricle area, and (I) whole heart relative Ramp2 expression from viable Ramp2+/+ ntg, Ramp2-/- ntg, and Ramp-/- Tg e14.5 embryos. Samples were normalized to Ramp2+/+ ntg and Gapdh expression. Scale bars: 100 µm. Data represented as average ± SEM from N = 3-8 mice per genotype. Significance determined by one-way ANOVA with Tukey’s Multiple Comparison with *p<0.05, **p<0.01, and ***p<0.001.

Figure 2-2: Ramp2-/- Tg adults are hypotensive. (A) Tail-cuff telemetry blood pressure analysis from conscious Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg adult female mice. (B) Representative images and vSMC wall thickness quantifications of Ramp2+/+ ntg and Ramp2-/- Tg descending aortas. Scale Bars: 100 m. (C) Intra-arterial blood pressure measurements from anesthetized Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg mice during baseline and after challenged with intravenous bolus of 30 g/kg phenylephrine (PE) or 12 nmol/kg adrenomedullin (AM). Drug response represented as a % change from baseline (dotted line). Data represented as average ± SEM from N = 3-5 mice per genotype. Significance determined by one-way ANOVA with Tukey’s Multiple Comparison (A, C) or unpaired student T-test (B) with *p<0.05 and **p<0.01.

Figure 2-3: Ramp2-/- Tg adults develop spontaneous dilated cardiomyopathy. (A) Representative M-mode echocardiograms from left ventricles of Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg mice. Macroscopic (B) and H&E (C) images of whole hearts and left ventricles from of Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg mice. Scale Bars: 2 mm. Quantification of heart (D), left ventricle (E), and right ventricle (F) weight normalized to body weight. Data represented as average ± SEM from N = 4-10 mice per genotype. Significance determined by one-way ANOVA with Tukey’s Multiple Comparison with *p<0.05, **p<0.01, ***p<0.001.

Figure 2-4: Ramp2-/- Tg left ventricles hypertrophied with early fibrotic and oxidative stress changes. (A) Histology and (B) wheat germ agglutinin (WGA, white) immunohistochemistry showing cardiomyocytes from Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg left ventricles. Quantification of (C) cardiomyocyte cross-sectional area and (D) capillary density from mouse left ventricles. Representative images and quantification of left ventricles stained with Picrosirius Red showing (E-F) perivascular and (G-H) interstitial fibrosis/collagen deposition. (I) Immunohistochemistry and (J) quantification of 4-HNE (white) staining showing relative levels oxidative stress in left ventricles. Scale Bars: 100 µm. Data represented as averages ± SEM from N = 3-4 mice per genotype. Significance determined by one-way ANOVA with Tukey’s Multiple Comparison with *p<0.05 and **p<0.01.

Figure 2-5: Ramp2-/- Tg have vascular congestion and multi-organ inflammation downstream of hypotension and dilated cardiomyopathy. (A) Spleen to body weight ratio from Ramp2+/+, Ramp2+/-, and Ramp2-/- females with and without the transgene. (B) Macroscopic images of liver vasculature abnormalities in Ramp2-/- Tg compared to that of Ramp2+/+ ntg and Ramp2+/+ Tg controls. Histology and quantification of inflammatory foci (blue arrows) in (C-D) livers, (E-F) kidneys, and (G-H) lungs from Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg mice. Scale Bars: 1 mm (B) and 100 µm (C, E, G). Data represented as averages ± SEM from N = 3-4 mice per genotype. Significance determined by one-way ANOVA with Tukey’s Multiple Comparison with *p<0.05, **p<0.01, ***p<0.001.

Figure 2-6: Decreased signaling and expression of RAMP-associated GPCRs in embryonic and adult Ramp2-/- Tg hearts. RT-PCR showing relative Pgc-1α expression from Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg adult left ventricle (A) or cardiomyocytes (B). (C) Quantification and representative western blot of phosphorylated CREB to total CREB in Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg left ventricles. Samples were normalized to Ramp2+/+ ntg and CREB with Gapdh used as a loading control. Relative expression of the Family B GPCRs; Calcrl, Calcr, Pthr1, Gcgr, Vipr1, CaSR and Gpr30 in (D) whole left ventricles and (E) isolated cardiomyocytes from adult Ramp2-/- Tg mice and

Ramp2+/+ controls. (F) Relative GPCR expression from e14.5 Ramp2+/+ ntg, Ramp2-/- ntg, and Ramp2-/- Tg whole hearts. Samples were normalized to Ramp2+/+ ntg and Gapdh and Rpl19 expression. (E) Data represented as averages ± SEM from N = 3-5 mice per genotype. Significance determined by one-way ANOVA with Tukey’s Multiple Comparison (A, C, D, F) or unpaired student T-test (B, E) with *p<0.05. (F) Model summarizing cardiovascular phenotypes in Cdh5+ endothelium and Cdh5- cells in Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg adults.

Supplemental Methods

Ramp2 Transgene Construct Generation

The mouse Ramp2 cDNA with human CD33 signal sequence, FLAG-tag epitope, bovine growth hormone polyadenylation tail in pcDNA3.1 backbone was gift from Dr. Walter

Born. The mouse vascular endothelial cadherin (Cdh5) promoter with rabbit β-globin intron

II was extracted from a plasmid which was gift from the Parise Lab. The pieces of the transgene were inserted into pBluescript II KS+/- backbone and confirmed by sequencing.

The University of North Carolina Animals Models Core microinjected the isolated transgene into pronuclei of C57BL/6 embryos and implanted into pseudo-pregnant females using standard published procedure.227

Mouse Studies

SJL-Tg(Tagln-cre)1Her (SM22-Cre) mice were obtained from the Jackson

Laboratory228. Vascular smooth muscle- specific Calcrl-null animals were generated by crossing CalcrlFlox/Flox with CalcrlloxP/+; SM22-Cre+ mice (Calcrl+/+ x SM22Cre+) generating the experimental (CalcrlLoxP/LoxP;SM22Cre+) and control (Calcrl+/LoxP;SM22Cre+) littermate mice used for this study. The SM22Cre+ transgene was expressed in both control and experimental mice to ensure Cre recombinase toxicity could not contribute to differential phenotypes 229. All mice used for this study were 3-4 month old male mice on a mixed genetic background. Three to four month old female Calcrlflox/flox; CAGG CreERTM mice were injected intraperitoneal with tamoxifen (5 mg/40 g body weight, Sigma) for 5 days to excised

Calcrl ubiquitously as previously reported.199 Aged matched female Calcrlflox/flox mice injected with tamoxifen were used as controls. Mice were aged between 50-60 weeks post tamoxifen administration then conscious echocardiography was used to assess heart function.

Blood Pressure

Non-invasive blood pressures were measured using the CODA 8 (Kent Scientific

Corporation) tail-cuff telemeter as previously described.230 Conscious 5-8 month old Ramp2 female and 3-4 month CalcrlloxP/loxP;SM22Cre mice were acclimated to tail-cuff apparatus measurements for 4 days prior to 3 days of collected blood pressure measurements.

Systolic, diastolic, and mean arterial pressure was measured 25 times per day per mouse and averaged together with measurements from 3 days after acclimation. Measurements were excluded by software if mouse moved during measurement or if mouse heart rate was below 450 bpm or above 750 bpm. For drug studies, mice were anesthetized with 1.5-2% isoflurane inhalation and an intra-arterial catheter was inserted into the right carotid artery.

Measurements were recorded using SciSense Model ADV500 PV control unit with

Labscribe 2 software (iWorx Systems). Baseline blood pressure was established before drugs were injected into the left jugular vein. Phenylephrine (Sigma) was delivered at 30

µg/kg and adrenomedullin (Phoenix Pharmaceuticals) at 12 nmol/kg both diluted in separate

50 µL saline as previously reported.47, 231 The percent changes were calculated from average of measurements 20 s taken 5 s post-injection and normalized to 20 s prior to injection.

Echocardiography

Conscious echocardiography was done using a VSI 2100 high frequency ultrasound

(VisualSonics) on 5-8 month old female Ramp2 mice and aged female CalcrlloxP/loxP;CAGG-

CreER mice. M-mode echocardiographs of left ventricle were analyzed using three measurements of both diastolic and systolic dimensions and averaged using Vevo 2100 software (VisualSonics), which calculated heart rate, ejection fraction, fractional shortening, and left ventricular volumes. Approximate cardiac output was determined by stroke volume

(LV Vol; diastolic – LV Vol; systolic) multiplied by heart rate and divided to 1000. A 30-MHz

48 probe and Vevo 660 Ultrasonography (VisualSonics) was used for 3-4 month old male

CalcrlloxP/loxP;SM22Cre mice and controls. These mice were anesthetized with 1-1.5% isoflurane and body temperature was maintained at 37°C by placing mice under a heating lamp and heart rate was closely monitored.

Tissue Collection

Whole mouse necropsy including heart, aorta, lung, kidney, and liver were weighed, imaged, and collected in 4% paraformaldehyde, flash frozen in liquid nitrogen, or stored in

RNALater (Ambion). Organ weights were normalized to either body weight or tibia length.

Heart chamber dissections were done to identify chamber-specific changes and collect tissue for histology, RNA, and protein from the left ventricle.200 Pregnant females were dissected during midgestation and all viable embryos were imaged and collected. For embryonic hearts, e14.5 whole hearts were removed and separated from lungs prior to fixation. Endothelial cells were isolated with magnetic-associated cell sorting using CD31- specific antibodies as previously described.201 Briefly, all lobes of lungs were collected pooled from either transgenic or wild-type adult mice and dissociated using manual mincing followed by collagenase type II (Worthington), neutral protease (Worthington), and deoxyribonuclease (Worthington). Single cell suspensions are labeled with PE-rat anti- mouse CD31 antibody (BD Pharmingen) and subsequently with anti-PE conjugated magnetic beads (Miltenyi) and ran through magnetic column to isolate both CD31- and

CD31+ cell fractions. Cells pellets were flash frozen and stored at -80C until RNA extraction. Adult cardiomyocytes were isolated with collagenase II digestion using previously published protocol202. Briefly, mice were injected with 100 IU of heparin (Sigma) and anaesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). Hearts were dissected from thoracic cavity and cannulated to perfusion system

(Radnoti). Hearts were perfused with Wittenberg Isolation Buffer with a constant flow of 4

49 mL/min at 37°C for 5 min, followed by Digestion Solution for another 10 min and stopped by perfusion with Stop Solution. The hearts were further teased into small pieces with forceps, pipetted into single cells, and filtered through 100m cell strainer. Cardiomyocytes were enriched by spinning down at low speed (50 g for 5 min) and pellets were collected in TRIzol reagent (Invitrogen) and stored at -80°C. Descending aorta was collected and the adventitia and endothelial layers were removed by enzymatic digestion with collagenase type II

(Worthington) to obtain purified VSMC as previously described.232

Gene Expression Analysis

Tissues were either snap frozen in liquid nitrogen and stored at -80°C or stored at -

20°C in RNALater (Ambion). RNA was extracted using TRIzol reagent and a bead homogenizer (Percellys) according to standard procedures, and then DNase treated

(Promega RQ1) and reverse transcribed with M-MLV (Invitrogen) or iScript cDNA Synthesis

Kit (BioRad). Semi-quantitative gene expression analysis was carried out using an in-house purified PFu polymerase and products were run out on a 3% agarose gel. Quantitative gene expression was assayed with Taqman master mix with ROX (Applied Biosystems or

Biobasic) and run on a StepOne Plus (Applied Biosystems). RT-PCR primers and probes are listed in Supplemental Table S1. Biological N was 3-5 mice per genotype and the relative expression levels were determined by the ΔΔCt and normalized to Gapdh or Rpl19 expression.

Whole Mount, Histology, and Immunohistochemistry

Fixed whole mount tissue was washed and imaged using a Leica MZ16FA dissecting stereoscope outfitted with a QImaging Micropublisher 5.0 RTV color CCD camera using

QCapture imaging software (QImaging). Paraffin embedded tissues were sectioned and stained for hematoxylin & eosin and Picrosirius Red using standard procedures. H&E and

Picrosirius Red slides were imaged using a Leitz Dialux 20 microscope outfitted with a

QImaging Micropublisher 5.0 RTV color CCD camera. For immunohistochemistry, paraffin sections were deparaffinized, hydrated, permeabilized and blocked with 5% normal donkey serum. Slides were then stained with primary antibodies including, rabbit anti-LYVE1

(1:300, Fitzgerald), mouse anti-4-HNE (1:40, ABCAM) overnight at room temperature.

Sections were rinsed, blocked, and incubated with secondary antibodies including, Dylight

594 isolectin B4 (1:200, Vector Laboratories), rhodamine wheat germ agglutinin (WGA)

(1:1000, Vector Laboratories), donkey anti-rabbit Cy5 (1:200, Jackson ImmunoResearch), donkey anti-mouse Cy3 (1:100, Jackson ImmunoResearch), and Hoechst 33258 (1:1000,

Sigma-Aldrich) in the dark for 90 minutes at room temperature. Images were acquired on a

Nikon E800 fluorescence microscope with a Hammamatsu Orca CCD camera and DAPI,

Cy2, Cy3, Cy5, Texas Red filter cubes with Metamorph software (Molecular Devices Corp.).

Image brightness and contrast changes were made equally across all comparable images using Photoshop CS4 and color images were pseudo-colored using ImageJ

(imageJ.nih.gov).

Western Blot

Tissues were either snap frozen in liquid nitrogen and stored at -80°C. Protein was extracted using Lens Homogenization Buffer with DTT and cOmplete protease inhibitor cocktail tablets (Roche) and a bead homogenizer (Percellys) according to standard procedures. Protein concentrations were determined using Pierce bicinchoninic acid (BCA) protein assay kit (Thermo Scientific) read on a Mithras LB 940 (Berthold Technologies) plate reader. Protein ran on Mini-Protean TGX SDS-PAGE gel (BioRad) and transferred onto nitrocellulose (GE Healthcare) membrane. Blots were blocked and stained in 5% BSA or

5% nonfat milk diluted in tris-buffered saline with 0.1% Tween 20. Blots were incubated in primary antibodies including, rabbit anti-DYKDDDDK FLAG M2 (1:750, Cell Signaling),

51 rabbit anti-phosphorylated CREB Ser133 87G3 (1:1000, Cell Signaling), rabbit anti-CREB

48H2 (1:1000, Cell Signaling), and mouse anti-GAPDH (1:4000, Novus Biologicals) overnight at 4°C. Blots were rinsed, blocked, and incubated with secondary antibodies including, goat anti-rabbit Dylight 680 (1:15,000, Thermo Scientific) and goat anti-mouse

Dylight 800 (1:15,000, Thermo Scientific) in the dark for 60 minutes at room temperature.

An Odyssey CLx (Li-COR) was used for imaging. Image brightness and contrast changes were made equally across all comparable images using Photoshop CS4 and quantification by densitometry was done using thresholds by ImageJ.

Serum Analysis

Adult females were anesthetized using isoflurane and whole blood was collected by retro-orbital bleeds and separated using Microtainer SST tubes (BD Scientific) and serum was stored at -20°C. Whole serum (100 L) was loaded onto VetScan Comprehensive

Diagnostic Profile rotors (Abaxis) and analyzed using VetScan V2 Chemistry Analyzer

(Abaxis) according to manufacturer’s protocol.

Morphometric Quantification

Embryonic e13.5 jugular lymph sacs were measured from both sides of embryos taken from relatively the same sectional plane. Embryonic e14.5 cardiac ventricles area was measured using images from whole hearts extracted during dissection and qualitatively confirmed using histological sections. Adult and e14.5 descending aorta wall thickness was determined by measuring the distance from lumen to edge of vascular smooth muscle cell excluding the adventitia with 10-15 measurements made around the circumference of the vessel in cross section. Hypertrophy was assessed by two blinding researchers measuring the area of 15 individual cardiomyocytes in cross section from three fields per H&E and

WGA stained left ventricles. Capillary density quantification has been previously

52 described.200 Briefly, left ventricles were stained with isolectin and WGA and 30 to 50 cardiomyocytes and bordering capillaries were count using ImageJ and averaged from 3 fields per left ventricle. Perivascular fibrosis was quantified by measuring the area of

Picrosirius Red staining surrounding 3-4 equivalent coronary vessels per left ventricle and normalizing it to the area of the vessel lumen. The amount of left ventricle interstitial fibrosis and 4-HNE oxidative stress were determined by threshold analysis of the signal normalized to the total section area in a 3-4 20x fields per mouse from equivalent regions of the left ventricle. Generalized inflammation was assessed by averaging the number of hematoxylin dense foci larger than 0.02 mm2 per section.

Supplemental Tables

Taqman Probe Forward Primer Reverse Primer Gene (5’-[6~FAM] – (5’ – 3’) (5’ – 3’) [TAMRA~6~FAM]-3’) Ramp2 (Wild- GAAGTCAGGCAG TCTGTCTGGAT type Allele) TCAGGGTTG GCTGCCTTGC Ramp2 (Null GACGAGTTCTTC TCTGTCTGGAT

Allele) TGAGGGGA GCTGCCTTGC GTCCATGCAAC CAGAATCAATCTC ATGGAAGACTACGAAAC Ramp2 TCTTGTACTCAT ATCCCACTGAG ACATGTCCTACCTTG ACC GACTACAAAGAC GTCCATGCAAC FLAG-Ramp2 GATGACGACAAG TCTTGTACTCAT (Transgene) C ACC Calcrl (Wild- GCGGAGCATATT GAAATGTGCTG type Allele) CAATCACAAG TATGTTCAAGC Calcrl (Flox GCGGAGCATATT GACGAGTTCTT

Allele) CAATCACAAG CTGAGGGGA Calcrl (Excised GCGGAGCATATT GAATAAGTTGA

Allele) CAATCACAAG GCTGGGCAG GCTGCCACGACC GTAGTTATTCG SM22-Cre AAGTGACAGCAA GATCATCAGCT TG ACAC Mm00441046_m1 (Life Pthr1 Tech) Gcgr Mm00433536_m1 (ABI) CaSR Mm00443375_m1 (ABI) Calcrl Mm00516986_m1 (ABI) Ctr (Calcr) Mm00432271_m1 (ABI) Vip1r Mm00449214_m1 (ABI) Gpr30 Mm02620446_s1 (ABI) Pgc-1α Mm01208835_m1 (Life

(Ppargc1α) Tech) Gapdh Mm99999915_g1 (ABI) Mm02601633_g1 Rpl19 (Thermo/Fisher)

Table 2-S1: Genotype and RT-PCR Gene Expression Primers and Probes

Serum Analyte Ramp2+/+ ntg Ramp2-/- Tg P value

Albumin (g/dL) 3.8 ± 0.1 3.6 ± 0.2 0.33 Alkaline phosphatase (U/L) 53 ± 7 50 ± 6 0.72 Alanine aminotransferase (U/L) 53 ± 3 45.8 ± 2 0.15 Amylase (U/L) 631 ± 8 713 ± 23* 0.02 Total bilirubin (mg/dL) 0.3 ± 0.0 0.3 ± 0.0 1 Blood Urea Nitrogen (mg/dL) 22 ± 1 18 ± 2 0.06 Calcium (mg/dL) 8.6 ± 0.1 8.7 ± 0.2 0.82 Phosphorus (mg/dL) 6.9 ± 0.3 8.2 ± 0.4 0.06 Creatinine (mg/dL) 0.2 ± 0.0 0.3 ± 0.0 0.09 Glucose (mg/dL) 252 ± 11 260 ± 18 0.70 Sodium (mmol/L) 147 ± 1 148 ± 0 0.82 Potassium (mmol/L) 5.6 ± 0.5 5.8 ± 0.7 0.85 Total protein (g/dL) 5.1 ± 0.2 4.8 ± 0.2 0.34 Globulin (g/dL) 1.2 ± 0.1 1.2 ± 0.1 0.72

Table 2-S2: Relatively similar systemic circulating factors in Ramp2+/+ ntg and Ramp2-/- Tg serum. Comprehensive diagnostic profile of 100 L of serum from adult female Ramp2+/+ ntg and Ramp2-/- Tg mice. N = 4-5 mice per genotype. Data represented as averages ± SEM with significance determined by the unpaired student T-test: *p < 0.05.

CalcrlloxP/+; SM22- CalcrlloxP/loxP; SM22- Cardiovascular Parameters Cre+ Cre+ Mouse Number 5 6 Mouse Age (wks) 12-16 12-16 Gender Male Male Body Weight (g) 33.3 ± 1.6 32.6 ± 0.7 Mean Arterial Pressure (mmHg) 128.6 ± 4.0 121.6 ± 4.2 Anesthesia 1.5% Isoflurane 1.5% Isoflurane

Heart Rate (BPM) 510 ± 12 478 ± 8 LVID, d (mm) 3.75 ± 0.16 4.00 ± 0.11 LVID, s (mm) 2.86 ± 0.15 3.06 ± 0.13 LVPW, d (mm) 1.07 ± 0.08 0.92 ± 0.04 LVPW, s (mm) 1.28 ± 0.08 1.13 ± 0.07

Ejection Fraction (%) 25.8 ± 3.1 23.6 ± 1.7 Fractional Shortening (%) 48.8 ± 6.2 47.4 ± 2.9 Cardiac Output (mL/min) 16.0 ± 2.0 15.3 ± 0.7 HW:BW (mg/g) 5.48 ± 0.34 5.31 ± 0.21 LV:BW (mg/g) 4.04 ± 0.27 3.78 ± 0.16

Table 2-S3: Vascular smooth muscle cell conditional Calcrl deletion does not lead to hypotension or dilated cardiomyopathy phenotype in adult males mice. Tail-cuff telemetry, echocardiography, and dissection measurements from 12-16 week old male CalcrlloxP/+; SM22-Cre and CalcrlloxP/loxP; SM22-Cre. Mice were conscious for blood pressure measurements, but under isoflurane anesthesia for echocardiography. BPM, beats per minute; LV, left ventricle; LVID, left ventricle internal diameter; LVPW, left ventricle posterior wall, HW, heart weight; BW, body weight. Data represented as averages ± SEM with significance determined by the unpaired student T-test.

CalcrlloxP/loxP; Heart Parameters Calcrlflox/flox CAGG-CreERTM+ Mouse Number 7 6 Mouse Age (wks) 70.4 ± 1.8 (+TAM 12-16) 68.2 ± 2.5 (+TAM 12-16) Gender Female Female Anesthesia Conscious Conscious Heart Rate (BPM) 577 ± 29 614 ± 29 LV Volume, d (µL) 51.7 ± 6.0 47.9 ± 3.5 LV Volume, s (µL) 12.5 ± 2.3 12.2 ± 1.8 Cardiac Output (mL/min) 23.0 ± 2.0 22.3 ± 1.0 Ejection Fraction (%) 73.1 ± 4.3 74.9 ± 2.6 Fractional Shortening (%) 42.3 ± 3.6 43.0 ± 2.4 IVS, d (mm) 1.13 ± 0.08 1.13 ± 0.04 IVS, s (mm) 1.72 ± 0.09 1.63 ± 0.08 LVID, d (mm) 3.63 ± 0.20 3.40 ± 0.10 LVID, s (mm) 2.13 ± 0.24 1.94 ± 0.11 LVPW, d (mm) 0.93 ± 0.14 0.98 ± 0.10 LVPW, s (mm) 1.50 ± 0.11 1.55 ± 0.13 HW:BW (mg/g) 3.89 ± 0.31 4.13 ± 0.20 LV:BW (mg/g) 2.84 ± 0.19 2.97 ± 0.14

Table 2-S4: Global conditional Calcrl deletion does not lead to dilated cardiomyopathy phenotype in aged females. Echocardiography and dissection measurements from conscious aged female Calcrlflox/flox and CalcrlloxP/loxP; CAGG-CreERTM mice injected with TAM when 12-16 weeks old. BPM, beats per minute; LV, left ventricle; d, diastole; s, systole; IVS, interventricular septal; LVID, left ventricle internal diameter; LVPW, left ventricle posterior wall, HW, heart weight; BW, body weight. Data represented as averages ± SEM with significance determined by the unpaired student T-test.

Supplemental Figures

Figure 2-S1: Endothelial-specific overexpression of murine Ramp2. (A) Transgene construct utilizing the vascular-specific Cdh5 promoter to drive FLAG-tagged murine Ramp2 cDNA. (B) Ramp2-FLAG transgene protein was detected in both transgenic founders 1 and 2 in the heart (H), intestine (I), kidney (K), and lung (L). FLAG protein was not detected in wild-type tissue. (C) Semi-quantitative RT-PCR of FLAG transgene transcript in isolated lung CD31+ endothelium from transgenic and wild-type mice. RT-PCR for Ramp2 expression from (D) isolated CD31+ endothelium and whole lungs and (E) isolated cardiomyocyte and whole left ventricles from adult Ramp2 Tg and wild-type controls. Samples normalized to wild-type whole tissue expression and Gapdh.

Figure 2-S2: Generation of gene-targeted Ramp2 deficient mice with and without Ramp2 Tg. (A) Ramp2+/+ Tg(Cdh5-Ramp2)1 mice were bred to Ramp2+/- ntg animals to generated Ramp2+/- Tg. PCR representing the six genotypes possible at e14.5 generated from the parent cross of Ramp2+/- Tg mice with Ramp2+/- ntg. (B) Kaplan-Meier survival plot of Ramp2-/- embryos with and without the Tg. (C) Representative images surviving Ramp2+/+ Tg and Ramp2-/- Tg pups. Significance were determined by Mantel-Cox and Gehan-Breslow-Wilcoxon tests with ***p<0.001.

Figure 2-S3: Generation and characterization of vascular smooth muscle cell-specific Calcrl deficient mice. (A) PCR representing the four genotypes generated from the parent cross of CalcrlFlox/Flox x Caclrl+/LoxP; SM22Cre+. The recombined “loxP” allele is only detected in SM22Cre+ mice and all mice were born at the expected Mendelian ratios. (B) Relative Calcrl gene expression in aortic (Ao.) VSMCs, bladder, heart, skeletal muscle (Sk. Ms.), liver, and lung from adult CalcrlLoxP/LoxP; SM22Cre+ and CalcrlLoxP/+; SM22Cre+ controls. (C) Representative images of H&E stained left ventricles from adult males of both genotypes. Scale bar: 1 mm. Data represented as averages ± SEM with *p < 0.05 determined by the unpaired student T-test and was normalized to individual tissues CalcrlLoxP/+;SM22Cre+ expression.

Chapter III: Dosage of adrenomedullin correlates with tumor and lymph node lymphangiogenesis3

Overview

Adrenomedullin (AM) is a potent lymphangiogenic factor that promotes lymphatic endothelial cell (LEC) proliferation through a pharmacologically-tractable G protein-coupled receptor. Numerous types of human cancers have increased levels of AM, however, the functional consequences of this have not been characterized. Therefore, we evaluated whether modulating adrenomedullin gene (Adm) dosage within tumor cells affects lymphangiogenesis. Murine Lewis Lung Carcinoma (LLC) cells that overexpress or under- express Adm were injected subcutaneously into C57BL/6 mice and tumors were evaluated for growth and vascularization. A dosage range from ~10% to 200% of wild-type Adm expression did not affect LLC proliferation in vitro or in vivo, nor did it affect angiogenesis.

Importantly, the dosage of Adm markedly and significantly influenced tumor lymphangiogenesis. Reduced Adm expression in tumors decreased the proliferation of

LECs and the number of lymphatic vessels, while elevated tumor Adm expression led to enlarged lymphatic vessels. Moreover, overexpression of Adm in tumors induced sentinel lymph node lymphangiogenesis and led to an increased incidence of Ki67 positive foci within the lung. These data show that tumor-secreted AM is a critical factor for driving both

3 Reprinted with permission from: Natalie O. Karpinich, Daniel O. Kechele, Scott T. Espenschied, Helen H. Willcockson, Yuri Fedoriw, Kathleen M. Caron. Adrenomedullin gene dosage correlates with tumor and lymph node lymphangiogenesis. FASEB Journal. 2013;27(2):590-600.

61 tumor and lymph node lymphangiogenesis. Thus, pharmacological targeting of AM signaling may provide a new avenue for inhibition of tumor lymphangiogenesis.

Introduction

The spread of tumor cells to distant sites in the body can occur through the blood or the lymphatic vascular system. Although the importance of the lymphatic system has been recognized for over 100 years, studies in understanding the molecular mechanisms of the metastatic process in lymphatics have lagged behind the more aggressively studied blood vasculature. This has been due, in part, to incomplete knowledge about the function of lymphatic vessel networks and related lymphoid tissues, as well as a lack of lymphatic- specific markers. The discovery of lymphatic endothelium specific markers (such as

LYVE1,233 podoplanin,234 and Prox142) has greatly enhanced our ability to investigate the lymphatic vasculature as a conduit for tumor cell dissemination.

The expansion of lymphatic vessels in and around tumors can facilitate the spread of tumor cells through the lymphatics.235, 236 Clinical studies show that increased peritumoral lymphatic vessels are associated with a higher number of metastatic lymph nodes (LNs) and subsequently, poor breast cancer patient outcome.237 The structural nature of lymphatic vessels, including their thin endothelium and lack of a basement membrane, is likely to make them more accessible to tumor cell escape.238 Recent studies have begun to elucidate that tumor secreted signals are necessary to first prime the sentinel LNs to form a metastatic niche. Thus, the presence of tumor and LN lymphangiogenesis serves as an early indicator of the metastatic process. Numerous animal model studies have focused on the VEGF protein family and identified VEGF-A,239 VEGF-C,240-243 and VEGF-D244 as drivers of tumor and LN lymphangiogenesis and subsequent lymphatic metastasis. However, a more comprehensive understanding of factors involved in promoting vascular changes within a tumor would enhance our understanding of the molecular components necessary for

62 metastasis and thus, better guide design of drugs for effective management of metastatic disease.

Adrenomedullin (Adm: gene; AM: protein) is a 52 amino acid peptide vasodilator that binds to and signals through the G-protein coupled receptor (GPCR), calcitonin receptor-like receptor (Calcrl: gene; CLR: protein) when the receptor is associated with receptor activity modifying protein 2 (RAMP2). We and others have consistently shown that the expression of Calcrl and Ramp2 is increased in lymphatic endothelium compared to blood endothelium.22, 23, 245, 246 We have also recently demonstrated that AM signaling is required for normal development of the murine lymphatic vascular system, since genetic loss of either

Adm or its associated signaling partners, Calcrl or Ramp2, results in embryonic lethality with profound interstitial edema as a consequence of arrested lymphatic vascular development.22

Since AM and its receptor partners are among the first GPCRs implicated in lymphangiogenesis,247 there is an opportunity to expand beyond the current repertoire of

RTK inhibitors and include a new arsenal of GPCR-based therapeutics for the inhibition of tumor (lymph)angiogenesis.

AM peptide was initially isolated from a human adrenal tumor76 and abundant evidence continues to emerge implicating this peptide in tumor biology (reviewed in 248). For example, in human tumors and numerous tumor cell lines, elevated AM levels have been reported to promote tumor cell survival, stimulate endothelial cell proliferation and help tumor cells evade immune surveillance.100, 103, 249 Genomic profiling analysis of human breast tumors found Adm as part of a 13 gene signature that correlated with distant metastases and poor outcome.250 Based on these clinical data and on the robust lymphatic phenotypes in AM signaling mouse models (Adm, Calcrl, Ramp2), we sought to determine how the lymphangiogenic properties of AM influence tumor progression. Using Lewis Lung

Carcinoma (LLC) cells, we generated stable cell lines with a gene titration for Adm

63 expression, allowing us to query the contribution of Adm to lymphangiogenesis and angiogenesis in an in vivo mouse model.

Materials and Methods

Cell Culture

Lewis Lung Carcinoma cells (ATCC, Manassas, VA, USA) were cultured in DMEM supplemented with 1X penicillin/streptomycin and 10% fetal bovine serum.

Adm overexpression

The complete coding sequence of murine Adm cDNA was cloned into the pcDNA3.1 expression vector (Life Technologies, Grand Island, NY, USA) that contains a CMV promoter and neomycin selection cassette. LLC cells were transfected with either AM- pcDNA3.1 or empty vector using Lipofectamine 2000 (Life Technologies, Grand Island, NY,

USA). Individual clones were selected and maintained in growth media containing 400

µg/mL G418.

Adm RNAi

Three shRNA sequences (Dharmacon, Thermo Fisher Scientific, Lafayette, CO,

USA) spanning the length of Adm were subcloned into the pSUPER vector (OligoEngine,

Seattle, WA, USA) and co-transfected with a zeocin resistance plasmid (kindly provided by

Dr. James M. Anderson, NIH) into LLC cells using Lipfectamine 2000. Individual clones were selected and maintained in growth media containing 1mg/mL zeocin (InvivoGen, San

Diego, CA, USA).

In Vitro Growth Assays

Stably transfected LLCs were seeded (7.5x104 cells/well) into 6-well plates and grown in complete DMEM. Cell number and viability were determined using the Countess automated cell counter (Life Technologies, Grand Island, NY, USA).

Mice and tissue processing

Female (6-8 wk old) C57BL/6 mice (The Jackson Laboratory, Bar Harbor, ME, USA) were injected subcutaneously in the flank with LLC cells (1x106 in 100µl PBS) that overexpress or knockdown Adm. Control mice were injected with LLC cells expressing a scrambled shRNA or empty vector. Tumors were measured with calipers [volume =

(length)(width)2 /2], and 14 d after injection, mice were euthanized and tissues were harvested. Routine histological processing and sectioning was performed. All experiments involving mice were approved by the Institutional Animal Care and Use Committee at the

University of North Carolina at Chapel Hill.

Immunohistochemistry

Antibodies: monoclonal rat anti-mouse Ki67(1:60, Dako, Denmark); polyclonal goat anti-mouse PECAM-1 (1:200, Santa Cruz Biotechnology, Santa Cruz, CA, USA); polyclonal rabbit anti-mouse LYVE-1 (1:200, Fitzgerald Industries International, Acton, MA, USA), monoclonal Syrian hamster anti-mouse Podoplanin (1:50, Developmental Studies

Hybridoma Bank, Iowa City, IA, USA); and monoclonal rat anti-mouse F4/80 (1:75, eBioscience, San Diego, CA, USA). Tumor and LN sections were permeabilized, blocked with 5% normal donkey serum, and incubated overnight at room temperature with primary antibodies. Sections were then rinsed, blocked, and incubated for 2 hours at room temperature with Dylight 488 (1:200, Jackson ImmunoResearch Laboratories, West Grove,

PA, USA), Alexa568 (1:200, Life Technologies, Grand Island, NY, USA), and DAPI (1:2000).

For Ki67 light level staining, sections were treated for antigen retrieval with citrate buffer in a microwave (15 min). The sections were rinsed, pretreated with 50% alcohol followed by 3% peroxide to block endogenous peroxidase activity. After blocking in 10% NDS (30 min), the sections were incubated in primary antibody to Ki67 in 2% NDS overnight. After rinses and blocking in 2% NDS, the sections were incubated in a biotinylated secondary antibody (1 hr,

1:200, Jackson ImmunoResearch Laboratories, West Grove, PA, USA) followed by

ExtrAvidin peroxidase (1hr, 1:5000, Sigma). Antibody binding sites were visualized using nickel-intensified diaminobenzidine (DAB, Sigma). Sections were counterstained with toluidine blue and coverslipped.

Image analysis

Images were acquired on a Nikon E800 microscope with a Hammamatsu camera with Metamorph software (Molecular Devices, Inc., Sunnyvale, CA, USA). To quantify colocalization of PECAM-1 or LYVE-1 with Ki67, a total of 3 step-sections were analyzed per tumor. For blood vessel number analysis, two vascular hotspot regions per section were analyzed for a total of 6 measurements per tumor. For lymphatic vessel number analysis, all vessels located within the tumor but at the tumor margin were counted from each of the 3 step sections per tumor. Vessel area was measured using ImageJ software. For LNs, lymphatic density was quantified by determining the threshold of LYVE-1+ fluorescence using ImageJ. The entire cross-sectional area of the LN was included in density analysis. A total of four step-sections per LN were averaged for each mouse (4-6 mice per genotype).

Data represents the mean ± SEM and statistical significance was determined by unpaired two-tailed Student’s t-test.

Gene expression analysis

Total RNA was isolated from stably transfected LLC cells or tumors using Trizol reagent (Sigma, St. Louis, MO, USA) followed by RQ1 DNase treatment (Promega,

Madison, WI, USA). qRT-PCR for murine Adm,30 Calcrl, Mac-1, Vegf-a, Vegf-c and Vegf-d were performed with Assay on Demand (Applied Biosystems, Carlsbad, CA, USA). Graphs

(GraphPad Software Prism 5, La Jolla, CA, USA) represent an average of 3 independent runs run in duplicate normalized to Gapdh.

Optical Projection Tomography

After fixation, lung lobes were rinsed in PBST overnight and then boiled for 10 minutes in citrate buffer. Tissue was washed in PBST, ice cold acetone, and ddH2O, blocked overnight followed by a one wk incubation with primary antibodies at 4°C with Ki67 and rabbit anti-mouse Caveolin1 (Abcam, Cambridge, MA, USA). Lung lobes were then incubated for 3 d in secondary antibody at 4°C using donkey anti rabbit 488 (Jackson

ImmunoResearch Laboratories, West Grove, PA, USA) and goat anti rat Texas Red (Life

Technologies, Grand Island, NY, USA), followed by an overnight wash in PBST. To prepare samples for OPT scanning, individual lung lobes were embedded in 1% low-melt agarose, dehydrated overnight in methanol and rendered optically transparent by incubating them in

BABB for 48 hours before being scanned in a BiOPTonics 3001M Optical Projection

Tomography Scanner (MRC Technology, Scotland, UK) equipped with Cy2 and Texas Red excitation filters. For image acquisition and analysis, 900 images were captured per channel per sample. Images were aligned and reconstructed into 3D renderings using nRecon software (SkyScan, Belgium) and color overlay 3D renderings were generated using the BiOPTonics Viewer application. These 3D renderings of Ki67 stained lungs were used to quantify proliferative foci using DataViewer (SkyScan, Belgium) and ImageJ (v1.45,

NIH) by three independent, blinded investigators.

Results

Adm gene dosage in LLC

Initially isolated in 1951 as a spontaneous lung carcinoma of a C57BL/6 mouse,251

LLC cells have been widely used to study metastasis and angiogenesis in a variety of murine implantation models. Quantitative RT-PCR analysis showed that LLC cells express abundant Adm (Figure 3-1A), but interestingly, there is negligible expression of its receptor

Calcrl (Figure 3-1B). The discovery that LLC cells lack the AM receptor CLR is a serendipitous advantage to our studies, since it removes the confounding possibility that changes in AM dosage directly affect tumor growth through CLR signaling. Therefore, any resultant phenotypes due to differences in tumor-secreted AM levels can be attributed to

AM-mediated signaling in the surrounding microenvironment.

To evaluate the effects of varied levels of tumor-secreted AM on tumor lymphangiogenesis, we generated stable LLC lines that either had reduced or overexpression of Adm. Using three shRNA sequences that span the length of the Adm gene, sequence #266 showed a drastic (92%) reduction in Adm expression while sequence

#555 and #709 showed a 50% and 68% decrease in expression, respectively as compared to the scrambled shRNA control (Figure 3-1C). Thus, sequence #266 was used for all subsequent in vitro and in vivo experiments. An expression plasmid containing the murine

Adm cDNA was used to generate stable AM overexpressing LLC cells. qRT-PCR analysis showed an almost 2-fold increase in Adm expression in these cells as compared to their empty vector controls (Figure 3-1C). Importantly, modulation of Adm dosage did not affect

Calcrl expression levels in the LLC cells (Figure 3-1B).

Since AM can act as a growth factor for tumor cells, we wanted to determine whether altering the dosage of Adm affected LLC growth rates. Because these cells expressed little to no Calcrl, (Figure 3-1B) we did not anticipate that changes in Adm expression would affect their growth. To confirm this and ensure that AM would not

68 influence proliferation though a non-canonical pathway, we counted cells over a 6 d time course. As expected, cell numbers were comparable among the high and low Adm expressing LLC cells and their respective controls (Figure 3-1D). Moreover, there were no changes in cell viability by trypan blue exclusion over the same time course (Figure 3-1E).

Throughout our study, we used the empty vector and scrambled controls separately for both the Adm overexpressor and RNAi conditions, respectively. However, we found no statistically significant differences between them and so, for the remainder of our study they are represented as a single control group.

Adm gene dosage is maintained in tumors in vivo

To determine whether dosage of Adm expression influences tumor growth in vivo, 1 x 106 cells were injected subcutaneously into C57BL/6 mice. Palpable tumors were measured with calipers and consistent with a lack of Calcrl expression in LLC cells, we found no significant effect of Adm dosage on tumor volume (Figure 3-2A). To ensure that the in vitro dosage of Adm was maintained in vivo, we analyzed expression of Adm and

Calcrl in tumors and found that dosage of Adm was maintained (Figure 3-2B). Although not statistically significant, it appears that Calcrl expression is modulated in an Adm dose dependent manner (Figure 3-2C). We attribute this to the heterogeneous nature of tumors including the presence of blood and lymphatic endothelial cells and macrophages which are rich sources of both AM and CLR.

Adm dosage does not promote tumor angiogenesis

A significant body of research supports AM as an angiogenic factor.8, 108 To assess whether levels of tumor-derived AM influence angiogenesis, tumor sections were stained with a blood endothelial marker (PECAM1) and the proliferation marker Ki67. Surprisingly, we found that varying the dosage of tumor AM expression did not influence the number of

69 blood vessels within the tumor but at the tumor margin (Figure 3-3A). Furthermore, the dosage of Adm in the tumors had no effect on the number (Figure 3-3B) or proliferation of

PECAM1+ cells (Figure 3-3C). Taken together, we conclude that AM is not modulating angiogenesis in this model.

Expression of Adm is necessary for tumor lymphangiogenesis

To investigate whether dosage of Adm influenced lymphangiogenesis we excised the tumors and performed routine H&E staining on the tumor sections (Figure 3-4A). We first wanted to determine whether expression of AM affected the number of tumor associated lymphatic vessels. The lymphatic-specific marker LYVE1 revealed a dose-dependent effect of AM on the total number of lymphatic vessels (Figure 3-4B, graph). Double immunofluorescence analysis of LYVE1 and podoplanin expression confirmed the lymphatic identity of these vessels (Figure 3-4B). We then quantified the number of LYVE1+ cells at the tumor margin and found a dose-dependent effect of Adm expression on the number

(Figure 3-4C) and proliferation (Figure 3-4D) of LYVE1+ cells, with approximately three times more LYVE1+ cells in tumors overexpressing Adm compared to tumors with reduced Adm.

Collectively, these results reveal that AM is a necessary and critical component for promoting tumor-associated lymphangiogenesis since reduction of Adm expression significantly reduced the proliferation and accumulation of lymphatic endothelial cells as well as the number of lymphatic vessels.

Adm dosage does not alter tumor macrophage or Vegf levels

The recruitment of macrophages to tumors serves to enhance endothelial cell survival and proliferation and thus promote tumor progression and metastasis.252 In addition to secreting lymphatic growth factors, like VEGF-C and VEGF-D,253 a subset of tumor- associated macrophages can be positive for LYVE-1 and can also transdifferentiate into

70 lymphatic endothelial cells.254 As expected, we found that some individual LYVE1+ cells in the AM-dosed tumors were also positive for the macrophage marker F4/80 (Figure 3-5A).

Importantly though, none of the LYVE1+ vessels expressed the macrophage marker (Figure

3-5B). F4/80 immunohistochemistry on tumor sections coupled with quantitative RT-PCR analysis for an additional macrophage marker, Mac-1 (Figure 3-5C, graph) revealed no statistical difference in either Mac-1 or F4/80 expression in either high or low Adm- expressing tumors or their respective controls (Figure 3-5C).

Alternatively, it is possible that enhanced lymphangiogenesis associated with Adm overexpression could be due in part to up-regulation of other known lymphangiogenic factors, such as VEGF-A, -C, or -D. However, qRT-PCR analysis for these genes in the tumors revealed no statistical difference compared to controls (Figure 3-5D). This negative result is important because it eliminates the possibility that elevated (or decreased) levels of tumor AM augment VEGF expression.

Overexpression of Adm increases vessel area

AM acts as a potent vasodilator of blood vessels through relaxation of vascular smooth muscle cells.255 Consistent with this function, we found that high levels of tumor AM markedly increased the diameter of tumor blood vessels (Figure 3-6A)—a change that occurred in the absence of enhanced angiogenesis, as documented by a lack of difference in PECAM1+ cells and vessels with varying dosage of Adm (Figure 3-3). Moreover, the diameter of blood vessels in tumors with low levels of Adm was comparable to that in control tumors, suggesting that reduction of Adm does not affect the ability of existent blood vessels to maintain their morphology and that a certain threshold of Adm expression is necessary to achieve an increase in vessel diameter.

Similarly, the lymphatic vessels in tumors with high levels of Adm expression were approximately three times larger than those in control tumors and in tumors with reduced

Adm expression (Figure 3-6B). We attribute the enlarged lymphatic vessels in Adm overexpression tumors to either the enhanced LEC proliferation (Figure 3-4D) or a secondary physiological response to increased blood vasodilation and tissue perfusion.

Overexpression of Adm induces LN lymphangiogenesis

Preclinical studies have supported the concept that tumor secreted factors can induce LN lymphangiogenesis and thus facilitate distant metastasis.256-258 Therefore, we analyzed sentinel, sub-iliac LNs to evaluate whether Adm dosage affected LN lymphatic vessel growth. Interestingly, the proportion of LYVE1+ staining was significantly elevated in the sentinel LNs of mice bearing Adm overexpressing tumors (Figure 3-7). Changes in the extent of LN lymphangiogenesis correlate with the lymphatic vascular changes observed in the tumors. Specifically, Adm RNAi tumors had fewer lymphatic vessels whereas the Adm overexpression tumors had more lymphatic vessels, in addition to increased lymphatic vessel diameter. Collectively, these findings demonstrate that elevated AM levels in tumors contribute to the sentinel LN lymphangiogenesis.

Overexpression of AM promotes distant metastasis

Tumors overexpressing Adm showed enlarged lymphatic vessel diameter accompanied with increased LN lymphangiogenesis. To determine whether the AM-induced changes in the tumor lymphatic vessels and draining LN potentiate metastasis to a distant organ we used high resolution 3D optical projection tomography (OPT).259 This method, shown in Figure 3-8A and seen more dramatically in Supplemental Movies 1-3, allowed us to visualize in 3-dimensions the extent of tumor metastasis throughout an entire lung lobe.

Using Ki67 as a surrogate marker for proliferating tumor cells we were not able to detect

Ki67 staining in the lungs of wild-type (non-tumor injected) mice (data not shown). Scans from lung lobes of mice bearing Adm RNAi tumors showed very little Ki67+ staining, which

72 was comparable to those levels observed in mice with control tumors (Figure 3-8A). This data correlates well with these groups having similarly small lymphatic vessels in the tumor

(Figure 3-6B). In contrast, lungs from mice with Adm overexpressing tumors had a marked and statistically significant increase in Ki67+ foci compared to those of mice with control tumors (Figure 8A), which also consistently correlated to the effects of Adm overexpression on tumor lymphatic vessel size (Figure 3-6B) and LN lymphangiogenesis (Figure 7). Using adjacent tissue sections, H&E (Figure 3-8B) and Ki67 (Figure 3-8C) staining confirmed Ki67+ tumor foci within the lungs of mice bearing Adm OExp tumors, while only individual Ki67+ cells were present within the lungs of mice bearing control and Adm RNAi tumors (Figure 3-

8C).

Discussion

High levels of AM have been reported in a variety of human solid tumors248 and gene profiling studies in mice have shown a 4-fold increase in Adm expression in tumor lymphatics as compared to normal tissue.260 Until now, the functional consequences of elevated Adm expression in tumors and the surrounding microenvironment have been largely unknown. We have developed an in vivo tumor model in which the dosage of the lymphangiogenic peptide AM is directly correlated to changes in lymphatic vessels at the tumor margin as well as distal changes in sentinel LN lymphangiogenesis and increased

Ki67 staining in the lung (Figure 8C).

Most research into how expansion of the lymphatic vascular system contributes to tumor cell dissemination has focused upon the VEGF protein family, but evidence is emerging that other factors, such as the angiopoietins,261, 262 prostaglandins,263 and the chemokine CCL21 also participate in tumor lymphangiogenesis and metastasis.264, 265

Although the autocrine effects of AM on tumor cells has been studied,102, 266 this study is the first to implicate AM as a contributing factor to tumor lymphangiogenesis by systematically

73 assessing how AM dosage affects both blood and lymphatic vasculature in the developing tumor. Adm RNAi approaches show that AM is an essential factor for driving the process of tumor and LN lymphangiogenesis. In contrast, overexpression of Adm in tumors is sufficient to exacerbate these processes.

Despite the presence of both intra- and peritumoral lymphatics in many tumors, it has been suggested that only the lymphatic vessels at the tumor margin are functional,235 since high interstitial fluid pressure within tumors can collapse or block intratumoral vessels, thus impeding the metastatic process. Padera and colleagues found that in VEGF-C overexpressing tumors, increased lymphatic vessel surface area in the tumor margin was sufficient to promote lymphatic metastasis despite the absence of functional intratumoral lymphatics.267 They propose that increased lymphatic surface area allows more opportunity for cancer cells to leave the primary tumor and disseminate to LNs and distant tissues. We show similarly robust results, since Adm overexpressing tumors have increased lymphatic surface area and LNs with enhanced lymphangiogenesis, as compared to tumors with low levels of AM.

Our findings show that AM preferentially induced lymphangiogenesis in this tumor model, since effects upon the blood vasculature were limited to an increase in vessel diameter. The selective function of vascular growth factors on different types of endothelial cells has previously been described. For example, when VEGF-C was overexpressed in

MCF-7 or MDA-MB-435 breast cancer cells, mainly lymphangiogenic effects were observed.240, 241 By contrast, the overexpression of VEGF-C in a melanoma xenograft study showed that VEGF-C was able to affect both angiogenesis and lymphangiogenesis.268 The disparity in angiogenesis induction between the two models has been attributed to differences in the in vivo proteolytic processing of VEGF-C and thus the selective induction of lymphangiogensis due to VEFGR-3 activation rather than VEGFR-2.240 In the case of

AM, the selective and preferentially robust effects we observed on lymphatic vessels

74 compared to blood vessels may largely be attributed to the enhanced expression of the AM receptor, Calcrl and its cognate receptor activity modifying protein 2 (Ramp2) in lymphatic endothelial cells. Our group, and several others,22, 23, 245, 246 have consistently shown that expression of the Calcrl and Ramp2 genes are intrinsically enriched in lymphatic endothelial cells, partly under the control of the lymphatic transcriptional regulator, Prox1.22 In addition,

Jin et al. showed enhanced lymphangiogenesis in response to AM administration using an in vivo wound healing model.11 Therefore, in the setting of this particular tumor microenvironment, the sensitivity of blood or lymphatic vasculature to a dosage range of AM ligand is likely controlled by the relative level of AM receptor expressed on these different endothelial cells.

Although tumor mRNA expression analysis showed no changes in expression of

Mac-1 or Vegf ligands, it remains possible that other factors may also contribute to the lymphatic phenotypes observed in the AM dosage LLC model. For example, a recent study showed that macrophage-derived AM promotes angiogenesis and melanoma growth.269

Furthermore, a B16 melanoma model showed that B lymphocytes are necessary for LN lymphangiogenesis since B-cell deficient mice were unable to induce LN lymphangiogenesis.257 In that study, the extensive lymphangiogenesis in the draining LN was associated with a 20- to 30- fold increase in lymph flow,257 and so, consideration of how other factors, such as lymph flow, could facilitate tumor cell entry into the lymphatics is warranted.

Our study highlights the need to explore AM and its receptors as potential therapeutic targets for selective modulation of tumor lymphatics. Additionally, how AM- induced tumor and LN lymphangiogenesis contribute to distant metastasis remain to be more rigorously assessed. Our previously published in vitro study showed that AM stabilizes the LEC barrier through a structural reorganization of junctional proteins.25

Whether these effects will be consistently maintained in an in vivo tumor setting, where

75 vessels are notoriously leaky, will be a very interesting question to address. In this regard, a recent study by Alitalo and colleagues found that an angiopoietin-2 blocking antibody promoted endothelial stabilization and thereby inhibited metastasis.270 Thus the effects of tumor-derived factors on lymphatic permeability may play a predominant role in promoting the metastatic process.

Indeed, the Calcrl receptor and its associated receptor activity modifying proteins

(RAMPs) have already been exploited by the pharmaceutical industry for the treatment of migraine pain.271 So, there is strong precedent to support the feasibility of the Calcrl/Ramp2 interface as a pharmacological target to inhibit AM signaling. Our in vivo studies provide the framework and rationale for inhibiting AM function in order to attenuate tumor lymphangiogenesis and would add a new and complementary facet to our current arsenal of anti-(lymph)angiogenesis cancer therapies for the future.

Author Contributions

N.O.K. designed research study, conducted experiments in Figures 3-2, 3-3, 3-4, 3-

5, 3-6, and 3-8, analyzed data, prepared and wrote manuscript. D.O.K conducted experiments in Figure 3-1, 3-7, and 3-8, analyzed data, assisted in manuscript preparation and editing. S.T.E. conducted experiments in Figure 3-8, analyzed data, assisted in manuscript preparation and editing. H.H.W conducted experiments in Figure 3-8, analyzed data, assisted in manuscript preparation and editing. Y.F. assisted in data analysis. K.M.C. designed research study and edited manuscript.

Figures

Figure 3-1. In vitro characterization of the LLC cell line. A-C) qRT-PCR analysis amplifying murine Adm and Calcrl mRNA. Mouse embryo normalized to 1 was used as a calibrator in panels A and B. Scrambled RNAi and empty vector conditions were normalized to 100 in panel C. Graphs in panels A and C depict 1 representative experiment while the data in panel B are an average of 3 independent experiments. D) Growth curve for the Adm RNAi, Adm OExp, and control LLC cells showed no differences in cell growth over 6 d. E) No change in AM-dosed LLC cell viability was detected by trypan blue exclusion over 6 d.

Figure 3-2. In vivo tumor analysis. A) Measurements of tumor volume showed no differences in growth of Adm RNAi (n=8), Adm OExp (n=7), and control (n=13) tumors during a 14 d time course. B, C) qRT-PCR analysis revealed that Adm dosage is maintained in the tumors (B); no statistical change in Calcrl expression among the tumor groups (C).

Figure 3-3. Adm expression does not affect angiogenesis. Adm expression does not affect angiogenesis. PECAM-1 (green)-, Ki67 (red)-, and DAPI (blue)-stained tumor sections revealed that dosage of Adm did not affect the average number of blood vessels (A), and the total number (B) or proliferation (C) of PECAM-1+ cells. Dashed line delineates tumor margin. Data are expressed as the average ± SEM of 2 fields/step section/tumor (6 total fields), with no statistical differences among the tumor groups; 3 step sections/tumor were analyzed. Scale bars = 50 µm.

Figure 3-4. Adm expression is necessary for lymphangiogenesis. Adm expression is necessary for tumor lymphangiogenesis. A) Representative H&E stained tumor sections. B) LYVE-1 (green) and podoplanin (red) costain confirmed lymphatic identity of the vessels and showed that the average number of lymphatic vessels per tumor section increased with Adm dosage. Dashed line delineates tumor margin. Data represent the average number of lymphatic vessels within the tumor but at the tumor margin per tumor section; 3 step sections/tumor were analyzed. C, D) Dosage of Adm affected the total number of LYVE-1+ cells (C), and proliferation (Ki67, red) of LYVE-1+ (green) cells correlated with increased Adm tumor expression (D). Data are expressed as the average of 2 fields/step- section/tumor (6 fields total) ± SEM (from 3 sections). Scale bars = 50 µm.

Figure 3-5. Adm dosage does not affect macrophage or Vegf levels. A) LYVE-1 (green)-, F4/80 (red)-, and DAPI (blue)-stained tumor sections revealed the presence of LYVE-1+ and F4/80+ cells. B) No colocalization of LYVE-1 and F4/80 was found in lymphatic vessels. C) Dosage of Adm did not affect F4/80 expression or Mac-1 levels (graph). D) Dosage of Adm did not affect Vegf-a (left panel), Vegf-c (center panel), or Vegf- d (right panel) levels. Scale bars = 50 µm.

Figure 3-6. Elevated Adm expression increases the diameter of blood and lymphatic vessels. Immunofluorescence analyses revealed that overexpression of Adm in LLC tumors caused a statistically significant increase in the area of PECAM-1+ blood vessels (A) and LYVE-1+ lymphatic vessels (B). Dashed line delineates tumor margin. Data are expressed as mean ± SEM, with an average of 40 vessels/genotype analyzed. Scale bars = 50 µm.

Figure 3-7. Tumor Adm expression influences sentinel LN lymphangiogenesis. Statistically significant increase in the density of lymphatic (LYVE-1) staining in the LNs was observed when mice harbored an Adm OExp tumor. Data are expressed as mean ± SEM for 4-6 mice/group. Scale bars = 200 µm.

Figure 3-8. Overexpression of AM promotes distant metastasis. A) Scans showed and quantitative evaluation confirmed that lung lobes from mice that harbored an Adm OExp (n=6) tumor had more Ki67+ foci as compared to lobes from Adm RNAi (n=7) and control mice (n=12). B, C) H&E (B) and Ki67 (C) staining verified OPT scans and showed scattered Ki67 staining in lung sections from mice bearing Adm RNAi or control tumors. By contrast, proliferative (Ki67+) foci were detected in lungs harboring an Adm OExp tumor. Arrows indicate individual proliferating cells; arrowhead points to proliferative foci. D) Model illustrates how tumor levels of AM affect local vascular changes which then may impact distant metastasis. Tumors with low AM levels have fewer lymphatic vessels (green), no blood (red) or lymphatic vessel area enlargement, reduced LN lymphangiogenesis, and fewer Ki67+ spots in the lungs as compared to a tumor with high levels of AM. Notably, while expression of tumor AM is necessary for lymphangiogenesis, levels of angiogenesis remain unaffected with changes in AM dosage. Scale bars = 800 µm (A); 20 µm (B, C).

Chapter IV: Conclusions and Future Directions

Summary

The studies presented in Part 1 further demonstrated the importance of

AM/RAMP/CLR signaling during cardiovascular development and homeostasis, as well as

AM role to promote tumor lymphangiogenesis and distant metastasis. Chapter II utilized an endothelial-specific Ramp2 transgene to genetically rescue the embryonic lethality associated with global loss of Ramp2. The loss of Ramp2 in non-endothelial cells of these mice led to spontaneous hypotension, dilated cardiomyopathy, and multi-organ inflammation. All these phenotypes potentially diverge for canonical AM/CLR signaling, indicating potential in vivo functions of the interaction between Ramp2 and the glucagon and parathyroid hormone receptors. Chapter III demonstrated that tumor-derived AM has a functional role in tumor lymphangiogenesis and metastasis, independent of tumor growth and angiogenesis. These data further support the importance of AM signaling in adult lymphatics and its detrimental effects on tumor metastasis through lymphatics.

Current State of the Field – RAMPs, Cardiac Development and Disease, and Tumor

Lymphatics

RAMPs Biochemistry to Cardiovascular Physiology

GPCRs remain one of the best pharmacological targets with approximately 35% of all approved human drugs targeting a small number of the GPCRome.272 There is significant interest to understand unique GPCR interfaces and pharmacology in order to develop novel compounds that have selective therapeutic functions without adverse side

85 effects. GPCR ligand selectivity and divergent signaling through different G proteins or - arrestin coupling is at the forefront of GPCR pharmacological research.273, 274 One such unique mechanisms of GPCR signaling was the discovery and biochemical characterization of the RAMPs.7, 191 The majority of family B GCPRs have been shown to interact with 1-3 of the identified mammalian RAMPs.275 RAMPs have been shown to facilitate numerous different aspects of GPCRs function, including trafficking to the plasma membrane, ligand selectivity, G protein- or -arrestin-coupling, and receptor recycling. The majority of the literature has focused on the CLR/RAMP1 and CLR/RAMP2 interactions, which control ligand specificity for the hypotensive peptides CGRP and AM, respectively. For drug design purposes, there has been a huge amount of biochemical research to understand the specific amino acids involved in the CLR/RAMP interactions.275 This work culminated with the crystallization of the ectodomains of CLR/RAMP1, CLR/Ramp2, and CLR/Ramp3 complexes.276, 277 These crystal structures provided a more thorough understanding of the ligand binding pocket and how to target the CLR/RAMP interfaces. Structural models also predict how RAMPs interact with other GPCRs.191 Beyond CLR, it was recently shown that

RAMP2 could alter ligand selectivity and G-protein coupling of the GCGR, which could have major implications on glucagon pharmacology and its regulation of blood glucose homeostasis.223 This study demonstrated that RAMP2 is not essential for the plasma membrane localization of GCGR or its affinity for glucagon. However the presence of

RAMP2 strengthens Gs while inhibiting Gi response to glucagon. Without RAMP2, the

GCGR is no longer selective for glucagon and can bind the related peptide, GLP-1, which has opposing functions to that of glucagon. The physiologic consequences of GLP-1 binding the GCGR or interactions between RAMPs and other GPCRs, such as

PTHR1/RAMP2, have not been functionally characterized. With the exception of CLR, the physiologic consequences of the majority of these reported biochemical interactions have

86 not been studied. While it’s apparent that the GPCR/RAMP interface can be a pharmacological target, further analysis of the in vivo functions of these interactions is required to better understand their viability as drug targets.

The physiologic functions of CLR/RAMP2 have been extensively studied by our lab and others and are previously discussed in this dissertation (See Chapter I, II, and IV). In addition to lymphatic developmental issues, these KO mice also have small disorganized hearts with altered trabeculation and thin vSMC walls (Chapter II).22, 33 The proliferation and maturation of the myocardium is dependent on numerous mitogenic factors, including IGF and FGFs, which are secreted from the epicardium.278, 279 Recently our group demonstrated that epicardial-derived AM was responsible for cardiomyocyte hyperplasia observed in the

AdmHi/Hi OExp mouse model.200 Cardiomyocyte-specific deletion of Calcrl by MHC-Cre or

SM22-Cre (Table 2-S3) did not cause embryonic lethality and adult mice with cardiac Calcrl had no apparent heart deficiencies. Together these data suggest that epicardial-AM may not be acting directly on CLR on cardiomyocytes, but rather is acting indirectly though other secreted mitogens, such as FGF, to regulate myocardium proliferation. In addition to the epicardium, there is cross-talk between the endocardium and myocardium, specifically by the BMP, EphrinB2, and Notch pathways, to regulate myocardial proliferation and trabeculae differentiation.280-283 Interestingly, it has been reported that sonic hedgehog and calcrl signaling are upstream of ephrin-B2, Notch, and VEGF pathways during arterial development in zebrafish.55 This suggests that endocardial CLR signaling may be upstream of these pathways during cardiac development. These same signaling pathways, in addition to PDGF and TGF-signaling, are all involved in vSMC development and maturation.284, 285

To my knowledge, CLR/RAMP2 signaling has not been directly linked to these pathways during vSMC development, but Calcrl-/- embryos had decreased aortic vSMC proliferation and survival.33 Together, these studies suggest that the AM/CLR/RAMP2 signaling pathway

87 is either directly or indirectly involved in regulating embryonic maturation of both the myocardium and vSMCs, but the specific signaling mechanisms have yet to be elucidated.

Due to the embryonic lethality of both Calcrl and Ramp2 KO animals,33, 34 not much is known about how the loss of these genes alters cardiovascular homeostasis during adulthood. Conditional ubiquitous ablation of Calcrl during adulthood leads to dysfunctional lymphatics, similar to those observed in human lymphangiectasia.199 However, further cardiovascular phenotyping of these mice was not reported. Cardiomyocyte-specific loss of

Calcrl during development did not manifest in any noticeable cardiomyopathies during adulthood, either basally or when challenged with hypertensive models (Table 2-S3)

(Dackor RT and Caron KM, Unpublished [2016]). However, loss of cardiomyocyte Ramp2, in the Ramp2-/- Tg mice or the cardiomyocyte-specific MHC-MerCreMer Ramp2 KO mice, caused a DCM-like phenotype.222 DCM is a complex multifactorial disease that primarily arises from alterations in cytoskeletal proteins or mitochondrial dysfunction and oxidative stress-induced myocardial damage.211, 286, 287 Cardiomyocyte loss of Ramp2 caused cellular and expressional changes that are common in DCM pathogenesis. However, it has been reported that the MHC-MerCreMer mouse line, independent of Ramp2, can cause a reversible DCM phenotype with high TAM concentrations.222, 288, 289 While the Ramp2-/- Tg mice likely have complex systemic changes that may be indirectly involved in DCM pathogenesis, the Ramp2-/- Tg represents the first model to implicate Ramp2 as being protective against DCM-like phenotype. Furthermore, this DCM phenotype appears to be independent of AM/CLR, which supports the involvement of other RAMP2-associated

GPCRs.

A direct link between the cardiac functions of GCGR or PTHR1 and DCM pathogenesis has not been, to my knowledge, characterized. Glucagon can change calcium signaling in cardiac cells and a glucagoma is capable of causing a reversible DCM-like

88 phenotype in humans.290, 291 Glucagon is detrimental cardiac function and survival follow myocardial infarction.215 However, GLP-1 agonists can improve myocardial glucose uptake and have been used to treat DCMs in canines.224 Hypoparathyrodism, hypocalcemia, or low levels of vitamin D have all been shown to be associated with reversible DCM in humans.225,

226 While the disease pathogenesis differs slightly, postpartum DCM is caused by alteration in circulating prolactin isoform.211, 286 Interestingly, Ramp2+/- females have been shown to be hyperprolactinemic, both basally and during pregnancy.198 High levels of prolactin with

Ramp2 loss, if cleaved by cathepsin D, could be a novel mechanism in the pathogenesis of

DCM and/or postpartum cardiomyopathy. In addition of RAMP2, it was recently reported that RAMP3 interacts with an estrogen receptor, GPR30, which is important for sex- dependent cardioprotection from hypertension.193, 196 More specifically, the loss of Ramp3 led to worst cardiac damage induced by angiotensin II in male mice when compared to female mice. Additionally, Ramp3 is required for therapeutic success of a cardioprotective

GPR30 agonist. Together, these studies illustrate the important of RAMPs in cardiovascular development and the pathogenesis of cardiac diseases. The identification that RAMP3 can interact with a family A GPCR, GPR30, opens the door to further explore novel

RAMP/GPCR interactions that could be important in cardiovascular homeostasis and disease.

Tumor Lymphangiogenesis and Metastasis

Tumor metastasis to distant sites is a multistep process that involves invasive tumor cells entering the circulation, evading immune surveillance, leaving the circulation, and surviving in a new microenvironment.292-294 The molecular mechanisms and factors involved in this process have not been fully characterized, which has led to the inability to prevent metastasis. While the metastatic process initiates locally within the primary tumor, it is a systemic disease where both tumor-derived and systemic factors influence metastasis. The

89 inhibition of tumor angiogenesis, primarily by targeting the VEGF pathway, continues to be o of the main strategies to prevent tumor metastasis. Unfortunately, tumors are adaptable and heterogeneous and able to selectively develop mechanisms to circumvent VEGF inhibition, which has ultimately led to a failure of these inhibitors to significantly improve patient survival.295-297 Metastases is not restricted to the blood vasculatures and numerous tumors can disseminate through the parallel, less understood lymphatic vasculature.235, 298 It is essential to target both lymph and blood vasculatures simultaneously to inhibit metastasis.

Therefore, identifying and understanding both the tumor-derived and systemic factors involved in tumor lymphangiogenesis and metastasis remains an intense area of research.

Our understanding of the factors that drive tumor lymphangiogenesis lags significantly behind those factors driving angiogenesis. Similar to angiogenesis field, research to understanding mechanisms of lymphangiogenesis have focused on the VEGF family, specifically VEGFC and VEGFD that both signal through VEGFR3. Both VEGFC and VEGFD have been shown to be involved in almost every phase of tumor and lymph node lymphangiogenesis, both of which facilitate lymphatic metastasis. 235, 240, 263 Likewise, the angiopoietin family, which also work through Tie receptors as well as VEGFR, have also been heavily studied as potential lymphangiogenic targets.299 Targeting the VEGF pathway, either through ligand traps or tyrosine kinase receptor blockers, has been shown to be a viable target to limit lymphangiogenesis in mice.298, 300 However, the lower-than-predicted success rates of VEGF inhibitors against angiogenesis suggest that targeting the VEGF pathway is not the sole solution to prevent metastasis.297 Additional targets and strategies to prevent lymphangiogenesis and other steps of the metastatic process are needed.

Given its role in lymphangiogenesis, one alternative pathway to target is

AM/CLR/RAMP2 signaling. Elevated AM levels have been reported in human melanoma, breast, pancreatic, prostate, ovarian, and colon cancers.249 While AM is produced by tumor- associated macrophages, AM can act back on these macrophages to alter their cytokine

90 secretion. 253, 292, 301, 302 In vitro studies showed that AM can promote tumor cell proliferation, invasiveness, and survival.103, 104 Xenograft mouse studies indicate that AM can enhance tumor growth and angiogenesis and AM inhibition can reduce tumor growth.56, 102, 303, 304 A genomic study comparing primary breast tumors with regional and distant metastasis identified AM as part of a 13-gene signature that is associated with distant metastasis and poor patient outcome.250 We showed that tumor-derived AM can promote primary tumor and lymph node lymphangiogenesis and enhance metastasis of LLC cells to the lungs

(Chapter III).305 As a follow up, Karpinich et al. demonstrated that AM can act directly to

LECs to alter gap junctions to facilitate the adhesion and transmigration of tumor cells through an LEC monolayer.306 Not only did this provide a potential mechanism behind the enhanced metastasis with high AM, but it also demonstrated the importance of gap junctions between LECs in regulating lymphatic metastasis. It has been reported that pregnancy can increase melanoma lymphangiogenesis and suppress the cytotoxic immune system allowing tumor cells to evade the immune surveillance.292, 294, 307-310 Considering AM is highly expressed during pregnancy (Figure 1-1), AM may be an important pregnancy-associated factors that is able to enhance the lymphatic metastasis. With the incidence of melanoma on the rise for decades, it is important to further understand this process.

The RAMPs have not been heavily implicated in tumor growth or metastasis, which requires further studies. It has been reported that Ramp3 shRNA reduced breast cancer growth, angiogenesis, and invasiveness downstream of epithelial to mesenchymal mediator lysyl oxidase-like protein 2.311 A recent study found that an amylin analogue, currently used to treat diabetes, acted through the calcitonin receptor/Ramp3 interface to cause rapid regression of p53-deficient lymphoma.312 These data implicate RAMP3 with tumor progression and demonstrate Ramp3 as a viable target against p53-deficient cancers.

Additional studies regarding RAMPs and tumor progression are needed.

Future Directions

New Models to Study AM/CLR/RAMP Biology

One major limitation of studying the AM/CLR/RAMP system is that the spatial and temporal expression pattern of these proteins has not been well characterized. Determining the expression pattern of these genes is limited to RT-PCR and/or in situ hybridization (ISH) and protein analysis is confounded by unreliable antibodies that have not been properly verified in KO tissue. The use of RT-PCR, usually in whole tissue, has led to numerous assumptions of the specific cell types that express these genes. The generation and expressional characterization of Calcrl, Ramp1, Ramp2, and Ramp3 reporter mice (gene targeted or BAC transgenic) would be invaluable to the field. This would take the “guess work” out of when and where these genes are expressed. The use of fluorescent reporters would allow a further understanding of the GPCR/RAMP colocalization and provide an opportunity to isolation specific cell types using FACS for ex vivo experiments. Zebrafish transgenic models would also be a useful tool to study AM and RAMP/GCPR interactions in vivo. Besides calcrl,55 zebrafish models of the RAMPs, to my knowledge, do not exist.

Another useful mouse model for the field would be a doxycycline-inducible AM mouse. Current models utilizing Adm+/- and AdmHiHi have developmental and adult phenotypes that confound phenotype interpretation in these models, such as heart defects or altered hemodynamics within blood and/or lymphatic vessels.22, 30, 79, 200, 313 The use of

Cre-recombinase to conditionally excise Calcrl in adult animals is irreversible. Considering

AM levels are elevated in a variety of disease states (Figure 1-1), these doxycycline- inducible AM mice would allow additional insight into the protective or detrimental roles of

AM during adult homeostasis and disease. A Tet ON/OFF switch would not only provide an opportunity to study how loss of AM alters homeostasis and disease pathogenesis, but would also allow the reintroduction of AM to attempt to resolve any phenotypes. Temporal

92 control of AM loss- or gain-of-function would be beneficial considering the numerous confounding phenotypes associated with different dosages of AM during development.

While our lab has a thorough repertoire of mouse models to study AM/RAMP/CLR signaling, another method to alter levels of these proteins is to focus on their transcriptional regulation, posttranscriptional modifications, and degradation. With CRISPR/Cas9 technology, it is possible to generate mice where Adm is not regulated by estrogen or hypoxia.314-316 Specific mutations could be introduced to further understand the functional consequences of these mutations in vivo. While AM was overexpressed use alter an altered

3’ UTR in the AMHiHi mice,200, 313 another mechanism to elevate circulating AM levels is to target the enzyme(s) that are involved in the rapid breakdown of mature AM peptide, such as PAM. PAM-/- mice have similar developmental defects as Adm-/-,35 so a conditional PAM allele could be used to prevent AM amidation and breakdown in adult mice. The mechanisms involved in AM, CLR, and RAMP processing and breakdown are not well understood. With further biochemical knowledge, it could be possible to generate RAMP mice that are unable to bind or traffic specific GPCRs to the membrane or impair

GPCR/RAMP ligand selectivity.

Roles of Ramp2 during Development and Reproductive Biology

It is clear that endothelial expression of Ramp2 is essential for normal murine development (Chapter II), but the timing, location, and levels of Ramp2 expression are less well understood.18, 22, 198, 206 For instance, 40% of Ramp2-/- Tg mice survive to adulthood with no obvious edema while the other 60% of Ramp2-/- Tg have severe hydrops fetalis and are stillborn (Figure 4-1A-C and Chapter II). The stillborn Ramp2-/- Tg pups are predominately male, indicating potential sex-dependent differences between Ramp2 (and/or the Cdh5- driven Tg) expressional requirements during development. One could assess both Ramp2 and Cdh5 expression levels in male and female mice at different developmental time points

93 to further address this hypothesis. The Tg expression could be mosaic and differences in the Ramp2-/- Tg lethality could be caused by low levels of Cdh5 expression in lymphatics at a crucial development time point. Low levels of Ramp2 or AM signaling could be sufficient to extend the Ramp2-/- survival (e14.5 to e18.5), but not fully rescue the lethality. This was previously observed in both endothelial-specific deletion of Calcrl and Ramp2 mice of which both die approximately 4 days after global KO embryos (Calcrl e12.5-e16.5 and Ramp2 e14.5-e18.5).22, 206 These data suggest that there is a threshold of AM signaling required for survival and this threshold is not sufficient in the majority of Ramp2-/- Tg mice. Edema score at e14.5 likely correlates with survival outcomes in that mild edema is resolved and Ramp2-/-

Tg embryos survive compared to severe edema ultimately leading to lethality (Figure 2-1).

Considering 100% of Calcrl and 95% of Ramp2 endothelial KOs die during development supports that the endothelium is the main cell type required for Ramp2 null survival and low

Ramp2 endothelial expression is sufficient to improved survival to 40% in Ramp2-/- Tg mice

(Table 2-1). 22, 206 Further investigation of the spatial (blood vs. lymphatic) and temporal levels of AM signaling required for embryonic survival are still needed.

AM/CLR/Ramp2 signaling is important for normal pregnancy with Adm+/-, Calcrl+/-, and Ramp2+/- female mice all having fertility defects34, 68, 198, 313, 317, 318 Both maternal and fetal AM/CLR/Ramp2 signaling is essential for placental development and function (Kadmiel

M and Caron KM Unpublished [2016]).313 The Ramp2-/- Tg mice provide another model to investigate the roles of this signaling pathway during pregnancy, considering the complex placental microenvironment. The Ramp2-/- Tg mouse could be used to investigate how non- endothelial cells (uterine epithelium, stroma, immune cells) contribute to the Ramp2+/- placental remodeling defects.313, 319, 320 Late gestation Ramp2-/- Tg with severe non-immune hydrops fetalis (Figure 4-1B) have impaired fluid drainage in placenta compared to Ramp2+/+

Tg controls (Figure 4-2A). These mice could be further used to study the pathologic consequences of placental edema downstream of hydrops fetalis.

It is currently unknown whether adult females lacking Adm, Calcrl, or Ramp2 can undergo normal pregnant. Ramp2+/- ntg males were crossed with Ramp2-/- Tg females in order to determine whether Ramp2-/- Tg females could get pregnant, deliver viable embryos, and/or foster pups. The litter sizes from Ramp2-/- Tg dams were significantly smaller compared to litters from Ramp2+/- Tg dams (Figure 4-2B). However, Ramp2-/- Tg females are able to undergo implantation and maintain pregnancy to term. I observed a lot of dead pups at birth which was partially due to stillborn Ramp2-/- Tg pups, but was also observed in a number of Ramp2+/-. A number of viable pups were present at weaning indicating the

Ramp2-/- Tg dams are physically capable to care for their offspring. The reduced litter size is indicative of a subfertility phenotype associated with Ramp2 heterozygosity, which occurred in the presence of endothelial Ramp2 expression in these Ramp2-/- Tg females. This data supports the assertion that normal fertility requires maternal Ramp2 expression in the uterus and/or placenta independent of the endothelium. Additional studies are needed to further understand the cell type specific functions of Ramp2 during pregnancy, which could possibly be accomplished using a Ramp2flox/flox or the Calcrlflox/flox animals crossed to different Cre lines to study the roles of AM/CLR/Ramp2 signaling in different cell types. It remains to be determined whether complete loss of maternal Ramp2 or AM/CLR/Ramp2 signaling is essential to undergo a successful pregnancy.

Sex-dependent Mechanisms of RAMPs to regulate Blood Pressure and Heart Disease

The Ramp2 Tg partially rescued the global Ramp2 deficiency, but this mouse model could also be used to study roles of Ramp2 endothelial overexpression. We observed no significant phenotypes in embryonic survival, body weight, blood pressure, heart function, inflammation, and/or gene expression in the Ramp2+/+ Tg compared to controls (Chapter II).

The lack of basal phenotypes is most likely due to feedback mechanisms to compensate for higher endothelial expression of Ramp2 and the associated vasodilatory effects of increased

AM/CLR/Ramp2 signaling. It would be interesting to investigate how (lymph)angiogenesis is altered in the Ramp2+/+ Tg mice using either retinal or tumor angiogenesis assays.

While the vasodilatory effects of AM are well established, the majority of genetic mouse models with alterations in this pathway do not alter basal blood pressure. For example, neither loss-of-function animals, including Adm+/- , Calcrl+/-, Ramp2+/-, Ramp3-/-

Calcrlflox/flox; SM22-Cre+ (Table 2-S3) or gain-of-function animals, including Ramp2+/+ Tg

(Figure 2-2), AdmHi/Hi and SM22-Ramp2 Tg had significantly altered blood pressure.33, 34, 47,

79, 196, 200 The Ramp2-/- Tg animals (Figure 2-2), Ramp2flox/flox; Cdh5-Cre, and Ramp1-/- are, to my knowledge, the only AM/CLR/Ramp2 mouse models with altered blood pressure.197, 206

While Ramp2-/- Tg have developmentally thinner aortic vSMC walls and DCM-like phenotypes (Chapter II), it is unclear why these mice develop hypotension. Administration of phenylephrine revealed that the vSMC are capable of vasoconstriction similar to that observed in controls. The duration of the elevated blood pressure appeared slightly shorter in Ramp2-/- Tg mice compared to controls which could partially explain the hypotension, but these results and conclusions need to be confirmed. Additional pharmacological challenges could help tease apart the mechanisms of endothelial-mediated vSMC relaxation through nitric oxide or direct vSMC relaxation.

Loss of Ramp2 could lead to higher circulating AM, which has been shown to signal through CLR/Ramp1 heterodimer. We recently reaffirmed that exogenous AM administration had no significant blood pressure effect in Ramp1-/- mice, likely indicating that Ramp1 is essential for AM vasodilation (Pawlak J and Caron KM Unpublished [2016]). This vasodilation through CLR/Ramp1 could be another potential mechanism of hypotension in

Ramp2-/- Tg mice in circulating AM is elevated. It would be interesting to determine how chronic AM administration, either by AM minipumps or genetic overexpression using the

AMHiHi mouse, alters blood pressure in Ramp2+/-, Ramp2-/- Tg, Ramp2+/+ Tg, Ramp1-/-,

Ramp3-/-, Ramp1-/-; Ramp3-/- animals. Furthermore, our lab has developed and utilized the

96 renin transgenic hypertensive mouse model, which could be crossed to Ramp2+/+ Tg,

Ramp2-/- Tg, Ramp2fl/fl, or Calcrlfl/fl mice in order to study how these mouse models respond to chronic hypertension challenge.

Interestingly, the basal hypertension observed in Ramp1-/- animals was only in male mice.197 Likewise, when challenged with renin-induced hypertension, male Ramp3-/- mice had a more severe phenotype than the female Ramp3-/- mice and Ramp3 was required for

GPR30-induced cardioprotection only in male mice.193, 196 Conversely while the Ramp2-/- Tg females had spontaneous hypotension, it does not appear that Ramp2-/- Tg males have significant hypotension (Figure 4-3). Similarly, male Ramp2-/- Tg males do not appear to have a DCM-like phenotype either, but additional male animals are needed to confirm these results. These sex-dependent phenotypes are not too surprising considering that Ramp2+/- females have a number of endocrine phenotypes not observed in Ramp2+/- males.198

Another hypothesis worth testing is whether the genetic loss of Ramp2 is involved in the pathogenesis of postpartum DCM. As previous mentioned, Ramp2+/- female mice have endocrine phenotypes including basal hyperprolactinemia.198 It has been shown that oxidative stress due to mitochondrial dysfunction can lead to the production of cathepsin D, which can cleave circulating prolactin into a harmful prolactin isoform that causes further cardiac and vascular damage.211, 286 While prolactin levels need to be confirmed, heightened basal prolactin could contribute to the observed DCM phenotype in the Ramp2-/- Tg females.

It would be interesting to compare hearts from virgin vs. postpartum Ramp2-/- Tg females to see if the DCM phenotype is exacerbated. Looking at serum prolactin isoforms would further confirm whether the mechanism of Ramp2-/- Tg postpartum cardiomyopathy is similar to previously reported results. This would represent a novel sex-dependent mechanism of

Ramp2 loss. Further investigation of the sex-dependent mechanisms of RAMPs to regulate blood pressure and cardiac homeostasis are needed to understand how the RAMP family members could be used therapeutically in a sex-dependent manner.

Roles of AM and GCGR/RAMP2 and PTHR1/RAMP2 Interfaces in Endocrine Diseases

AM has been reported to be expressed in pancreatic islets and functions to inhibit insulin secretion in vivo and in isolated islets.321, 322 The roles of AM in glucose homeostasis and diabetes need to be further verified. Loss of Ramp2 during development led to decreased cardiac Gcgr expression throughout life (Figure 2-6). Biochemical studies demonstrated that GCGR, but not GLR-1R or GLP-2R, interacts with RAMP2.191 Weston et al. recently demonstrated that RAMP2 alters GCGR G protein-coupling and signaling.

Additionally, RAMP2 dictates the ligand selectivity of the GCGR.223 More specifically, the

GCGR/RAMP2 heterodimer is selective for glucagon, but the GCGR without RAMP2 can bind both glucagon and GLP-1. The physiological consequences of GLP-1 as a partial agonist of the GCGR are unknown. Glucagon and GLP-1 typically have opposing functions in regulating glucose and cardiac homeostasis. Glucagon increases blood glucose concentrations by activating liver gluconeogenesis and is detrimental to cardiac function following ischemic injury. Conversely, GLP-1 increases insulin to decrease blood glucose levels and is cardioprotective following ischemic injury.215, 323, 324 Using Ramp2 mouse models, such as the Ramp2-/- Tg or Ramp2flox/flox mice, it would be interesting to investigate the physiological consequences of altered Gcgr ligand selectivity and signaling in the in the heart, pancreas, and liver. One could measure fasting glucose, insulin, glucagon, and GLP-

1 levels as well as perform glucose tolerance tests to assess dysregulation in glucose homeostasis. Serum glucose was not altered basally in Ramp2-/- Tg mice (Table 2-S2), but the these mice could be challenged with type I or type II diabetic models. Exogenous glucagon and GLP-1 administration usually have opposing effects, which could be tested in

Ramp2 null adults to help determine how Ramp2 alters these peptides functions in vivo.

These Ramp2 null mice could be fed a high fat diet to assess how obesity and insulin resistance affects their glucose homeostasis and overall health. Gcgr-/- dams had significant

98 perinatal lethality due to lower glucose levels in pregnant females,325 which could partly explain Ramp2-/- Tg perinatal lethality and should be further examined.

Ramp2 loss led to altered Pthr1 levels (Figure 2-6) and RAMP2 has been shown to traffic PTHR1 to the membrane, but the functions of this interaction is unknown.191 Using

Ramp2 gain- and loss-of-function mice it would be interesting to test how decreased Pthr1 expression in heart and kidney alters the parathyroid hormone signaling axis to regulate vitamin D and serum calcium homeostasis. Serum calcium levels in Ramp2-/- Tg were unchanged (Table 2-S2), but it is possible that serum vitamin D and bone resorption in these animals acts as a compensatory mechanism to maintain serum calcium, which requires further investigation. Bone resorption has been reported in Ramp2+/- females.198 To my knowledge, cardiac-specific Pthr1 functions have not been assessed and could be informative to understand the DCM-like phenotype in Ramp2-/- Tg mice. Additionally, a combined Pthr1+/-; Gcrg+/- mouse could also be informative since Pthr1 and Gcgr are at heterozygous levels in Ramp2-/- Tg mice. Vitamin D deficiency and hypocalcaemia have been shown to correlate with DCM in humans, which represent another potential mechanism for the DCM phenotype in the Ramp2-/- Tg mice. It would be interesting to test if calcium, vitamin D, GLP-1, and/or antioxidant administration could rescue the DCM phenotype. Together, additional studies are needed to shed further light on the physiological consequences of the RAMP2 interaction with both GCGR and PTHR1. These studies would provide insight into how these interactions are involved in cardiac and endocrine diseases, providing a novel understanding of how to target these diseases therapeutically.

Determine if tumor-derived AM promotes tumor lymphangiogenesis indirectly through macrophage activation, and/or transdifferentiation

We showed that tumors-derived AM dosage is correlated with tumor lymphangiogenesis (Figure 3-4), but the mechanism is unclear.305 The study in Chapter III concluded that the increased lymphangiogenesis caused by AM was through AM signaling directly on LECs to increase proliferation of preexisting lymphatic vessels. AM could also be acting indirectly through macrophages to alter lymphangiogenic factor secretion. Tumor- associated macrophages, which secrete numerous lymphangiogenic factors, need to be further interrogated in the AM-dosed tumors. AM dosage does not appear to alter macrophage infiltration by IHC (Figure 3-5).305 This result should be confirm using quantitative flow cytometry to count macrophages (F4/80+ or CD11b+) from dissociated tumors. Additionally, the expression of Vegfa, Vegfc, or Vegfd lymphangiogenic genes was unchanged (Figure 3-5), but this was done using whole tumor RNA which could mask expressional changes in less abundant cell types within the tumor microenvironment. RT-

PCR or genome-wide array expression analysis from specific cell types, including LLCs, macrophages, Lyve1+ macrophages, and endothelial cells would provide additional insight into how AM induced lymphangiogenic changes within the microenvironment.

Macrophages have been reported to transdifferentiation into LECs and incorporate directly into lymphatic vessels.253, 254, 326-328 There appeared to be a potential correlation between tumor-derived AM and the number of these specialized Lyve1+; F4/80+ macrophages (Figure 3-5).305 Lyve1+; F4/80+ macrophages should be quantified in Adm dosed tumors using both IHC and flow cytometry as well as analyzed for evidence that these macrophages are able to incorporate into lymphatic vessels, similar to that recently reported CD31+ tumor cell vascular mimicry.295 It would be interesting to determine whether

AM treatment can alter the pro- or anti-inflammatory cytokines and lymphangiogenic peptides secretion profile from cultured Lyve1+; F4/80+ or Lyve1-; F4/80+ macrophages.301,

100

328 To tease apart the macrophage-specific effects on LLC lymphangiogenesis, macrophages could be depleted using clodronate liposomes,327 which would further confirm that AM is not acting indirectly through macrophages to effect LECs proliferation.

Determine how systemic AM dosage alters tumor lymphangiogenesis and metastasis

The results of Chapter III demonstrated that tumor-derived AM dosage alters tumor lymphangiogenesis and metastasis.305 A follow up experiment is to test whether systemic

AM also plays a role in tumor lymphangiogenesis and metastatic potential. Systemically high AM is relevant because occurs during numerous diseases and during normal pregnancy (Figure 1-1). Numerous reports have demonstrated that pregnancy can enhance tumor lymphangiogenesis and metastasis, particularly in early onset melanoma.307, 309, 310, 329

The mechanisms behind pregnancy-associated enhanced metastasis is not well understood, but involves the complex interaction of the immune system, sex hormones, and other circulating factors.292, 294, 308, 310, 330 To determine whether AM is one of those circulating pregnancy-associated factors; tumors could be implanted into both pregnant and virgin mice and assessed for growth and metastasis. However, this elevated AM during pregnancy returns to normal quickly following delivery, which is not conducive to the time course metastasis in murine implanted tumor models. Additionally, other pregnancy- associated factors confound the AM-specific roles in this process. Alternatively, the AdmHi/Hi mouse model is ideal to study how elevated systemic AM levels affect tumor metastasis independent of the other hormonal alterations that occur during pregnancy. The AdmHi/Hi mice have similar elevated AM levels observed during normal human pregnancy. This model could be compared to genetic models of low (Adm+/-) and normal (Adm+/+) AM levels.

B16F10-GFP melanoma cells (kindly provided by Dr. Andrew Dudley) were implanted subcutaneously in immunocompetent wild-type and AdmHiHi mice and allowed to grow for 17 days (Figure 4-4A). Primary tumors expressed GFP tumor cells and systemic

101

AM did not alter tumor growth (Figure 4-4B-C). CD31 (blood) and Lyve1 (lymphatic) stained primary tumors sections revealed no preliminary differences in either the number or dilation of blood or lymphatic vessels in wild-type or AdmHiHi primary tumors (Figure 4-4D-E). There appeared to be an increased number of GFP+ tumor cells in the draining lymph nodes and lungs of the AdmHiHi mice compared to control following 17 days (Figure 4-4F). Preliminary quantification revealed a trending increase in GFP+ tumors cells in AdmHiHi lung sections

(Figure 4-4G), but this quantification should be further confirmed by flow cytometry. To further determine if metastasis was occurring independent of changes in tumor vasculature,

1.5x105 B16F10-GFP cells were injected into the tail-vein of these mice and assessed 21 days post-injection (Figure 4-5A). After whole mouse necropsy, detectable GFP+ foci were visible in the liver and lungs (Figure 4-5B), as well as a few foci in the ovaries, spine, and kidneys. While preliminary (N=4), GFP+ foci appeared more consistently and in greater numbers in the AdmHiHi mice compared to wild-type controls (Figure 4-5C-D). This was apparent both macroscopically and microscopically in the lungs (Figure 4-5E). Together these preliminary in vivo experiments suggest that metastasis is slightly increased in the

AdmHiHi animals, which appeared to be independent of primary tumor vasculature. I speculate that this increased metastasis is due to differences in either tumor cell adherence or through the suppression of the cytotoxic immune system. More experiments are needed to validate and confirm these hypotheses and conclusions.

The potential of AM as a therapeutic target against cancer has not been well studied in vivo. It would be interesting to investigate how vascular or lymphatic-specific deletion of

Calcrl, which have impaired lymphatic and presumably blood vasculatures,199 alters tumor growth and metastasis. These mice challenged with LLC-Adm RNAi and LLC-Adm OExp tumors would further define the cell autonomous effects of tumor-derived AM in the tumor microenvironment. The most common lymphangiogenic peptide shown to enhance lymphatic metastasis is Vegfc.44, 240-243 Vegfc not only potently induces lymphatic vessel

102 proliferation and growth, but it also makes vessels extremely leaky. Meanwhile AM has been shown in vitro and in vivo to tighten endothelial junctions making them less leaky.25, 199,

306, 331, 332 It would be interesting to investigate the vessel number, integrity, and metastatic potential in mice that overexpress both Adm and Vegfc. Potentially a dual target approach involving both CLR and VEGFR3 would be more successful therapeutic target than VEGF inhibitors alone.

AM Functions on Immune Cell Adhesion, Activity, and Cytotoxic Surveillance

AM has been shown to have acute anti-inflammatory functions and is largely considered an anti-fibrotic and protective peptide.32, 333-335 However, the consequences of chronically elevated AM, to my knowledge, have not been thoroughly addressed. In the context of tumors or chronic infections, AM could have unintended consequences of suppressing the cytotoxic immune system which would allow tumors cells or infections to persist in the circulation. Furthermore, loss of Ramp2 in adult mice leads to multi-organ vasculitis, which has been attributed to either vascular congestion secondary to DCM and hypotension in the Ramp2-/- Tg mice (Figure 2-5) or due to increased adherence of immune cells to endothelium lacking Ramp2.206 It remains unclear how overexpression of endothelial Ramp2 has the same inflammatory phenotype as deletion of endothelial Ramp2.

Hemodynamic stasis downstream of hypotension or DCM could increase immune cell adherence and transendothelial migration or could increase compensatory circulating vasoconstrictors that can indirectly cause tissue damage and inflammation. Interestingly, both Ramp2 mouse models have evidence of increased Adm expression, which is likely elevated due to homeostatic response caused by loss of Ramp2 or indirectly due to pathologic conditions. Surprisingly, adult AdmHiHi mice also have a significant number of inflammatory foci in their livers (Figure 4-6A), kidneys (Figure 4-6B) and lungs (Figure 4-6C), which is progressive with age. The AdmHiHi adult mice also have significantly larger spleens

103 than Adm+/+ gender aged match mice (Figure 4-7B). These data further support that chronically high AM may be involved in the multi-organ inflammation phenotype.

There are mechanistic similarities by which immune and tumor cells adhere and migrate through endothelial monolayers.336 A recent study from our group demonstrated that AM enhanced tumor cell adherence and transmigration through LEC monolayer, which was facilitated through gap junctions.306 These results were accomplished through either genetic overexpression of AM in the LLC Adm OExp tumor cell model (Figure 3-1) or by exogenous AM treatment of other mouse and human tumor cell lines. Furthermore, LLCs

Adm RNAi, AM inhibitor, or gap junction inhibitors all reversed this process. I hypothesize that chronically elevated AM can enhance immune cell adherence and migration through endothelial monolayers and could be an underlying mechanism behind the observed multi- organ inflammation in the Ramp2 mouse models. One could utilize similar in vitro adherence and transmigration assays as previously published in order to confirm these results using primary mouse or human immune cells instead of tumor cells.306 Both BEC and LEC monolayers could be tested, as well as different immune cell populations collected from Adm+/-, Adm+/+, and AdmHiHi mice. If inflammation is progressive with age, different aged mice could be used to study how AM alters the immune system composition and function over time. Specific isolated immune cells could be further tested for altered cytokine production. These experiments could also be performed in activated immune cells, from inflamed mice, or with isolated human immune cells treated with AM.

Preliminary results suggest that high systemic AM in the AdmHiHi mice can potentially increase melanoma cell metastasis, independent of blood and lymphatic vasculatures in the primary tumor (Figure 4-4 and 4-5). I hypothesize that high AM inhibits the cytotoxic immune system, specifically peripheral natural killer (pNK) cells considering AM dosage alters uterine NK cells recruitment and cytokine release.313 A CD49b+ pNK cell enriched population was isolated using dissociated wild-type spleens and MACS (Figure 4-7A). To

104 assess the ability of AM to activate or inhibit pNK cell function, isolated pNK cells were treated with AM, prostaglandin E2 (PGE2), or Forskolin. All treatments increased cAMP production (Figure 4-7C). pNK expressed modest levels of Calcrl (Figure 4-7D), suggesting

AM can act directly on NK cells. It has been shown that PGE2 and increased cAMP production inhibits pNK cell cytotoxicity.337, 338 To assess cytotoxicity directly, pNK cells were co-cultured with fluorescently labelled lymphoma cells (Yac1) and treated with a cytotoxic activator IL-2,339 an inhibitor PGE2, and/or AM. Specific fluorescent tumor cell death was assessed by flow cytometry (Figure 4-7E). As expected IL-2 activated pNK cells killed a significant number of tumor cells, which was blocked by PGE2 treatment (Figure 4-

7F). AM treatment was not able to inhibit IL-2 activated pNK cell cytotoxicity (Figure 4-7F).

These initial experiments indicate that AM can act directly on pNK cells, but does not inhibit

NK cell cytotoxicity. Therefore while high AM appears to increase metastasis, it is likely not caused by the direct inhibition of pNK cytotoxic surveillance. It has been shown that lymphangiogenic factor, VEGFD, can increase serum prostaglandin levels indirectly through

LECs.263 AM could indirectly regulate pNK cell function through prostaglandins or by altering various aspects of the cyclooxygenase pathway, which has been reported previously in murine arthritis models (Figure 4-8A).334, 340 Alternatively, AM could increase tumor cell survival by inhibiting other cytotoxic cells, including macrophages,301 myeloid- derived suppressor cells,310 regulatory T cells,334 or mast cells105 directly (Figure 4-8B). AM could act indirectly through these cells to suppress pNK cell function (Figure 4-8C). It is also possible that the number of pNK cells and/or other cytotoxic cells is altered in the AdmHiHi enlarged spleens. Investigating the cytokine profile and activity of NK cells isolated from different mouse lines would be informative. To my knowledge, it has not been reported how

AM affects pNK cell function or immune surveillance AM during tumor metastasis. These studies would be informative to further understand AM roles during inflammation and cancer.

105

Concluding Remarks

Through the studies in Part I, I opened the door to further studies to investigate

Ramp2 and its specific, possibly sex-dependent, roles to regulate blood pressure and cardiac homeostasis. This novel transgenic rescue could potentially act as a therapeutic strategy to treat non-immune hydrops fetalis and potentially other lymphatic diseases. It also expands Ramp2 functions to other GPCR pathways essential for endocrine homeostasis. While AM is typically considered protective, it is detrimental in tumor context and could represent a novel target against tumor metastasis.

106

Figures

Figure 4-1: Edema and late gestational lethality-associated with some Ramp2-/- Tg pups. Representative (A) PCR from (B) e18.5 viable Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg embryos. Magnifications are equivalent between panels. (C) Representative images of still-born Ramp2-/- Tg P1 pups compared to viable Ramp2+/- littermates.

107

Figure 4-2: Ramp2-/- Tg dams can get pregnant but have reduced litter size. (A) Representative H&E images of Ramp2+/+ Tg and edematous Ramp2-/- Tg placentas from e18.5 embryos. (B) Average litter size of Ramp2+/- ntg male crossed to either Ramp2+/- Tg or Ramp2-/- Tg females. N = 3 litters from 2-3 different dams. Significance determined by student T-test with *p<0.05.

108

Figure 4-3: Ramp2-/- Tg spontaneous hypotension is sex-dependent. Intra-arterial blood pressure measurements from anesthetized Ramp2+/+ ntg, Ramp2+/+ Tg, and Ramp2-/- Tg (A) female and (B) male mice. Data represented as average MAP in a minute of continuous baseline measurements. N = 2 mice per genotype per sex.

109

Figure 4-4: Systemic elevated AM led to increased metastasis independent of tumor growth or (lymph)angiogenesis. (A) B16F10-GFP melanoma cells were injected subcutaneously into immunocompetent Adm+/+ and AdmHiHi adult female mice. (B) Representative sections from primary tumors with GFP+ (green) tumor cells and GFP- host- derived cells. (C) Tumor volume after 17 days of growth measured by calipers from Adm+/+ and AdmHiHi mice. Quantification of tumor-associated, both peritumoral and intratumoral (D) blood and (E) lymphatic vessels from Adm+/+ and AdmHiHi primary tumor sections. Dilation was defined by visibly open vessel lumens. Blood vessels were visualized using CD31 and lymphatic vessels by LYVE-1. (F) Representative images of GFP+ tumor cells (green) in both the draining sentinel lymph node and lung from B16F10-GFP injected Adm+/+ and AdmHiHi mice after 17 days. (G) Quantification of GFP+ cells per lung sections in 3 mice per genotype.

110

Figure 4-5: Increased tumor cell seeding in AdmHiHi mice independent of primary mass. (A) B16F10-GFP melanoma cells were injected into the tail-vein of immunocompetent Adm+/+ and AdmHiHi adult female mice. (B) Representative bright field and fluorescence whole mount lungs and livers demonstrating visible GFP+ (green) tumor mass 21 days following injection. Quantification of (C) the percentage of mice with visible GFP+ foci in different organs and (D) number of GFP+ foci per lobe of lung. (E) Representative lung sections with GFP+ cells (green) from Adm+/+ and AdmHiHi mice 21 days post-injection. N = 4 mice per genotype.

111

Figure 4-6: Chronically elevated systemic AM causes multi-organ inflammation. Representative H&E images and quantification of inflammatory foci in (A) livers, (B) kidneys, and (C) lungs from Ramp2+/+ ntg, Ramp2+/+ Tg, Ramp2-/- Tg, aged-matched AdmHiHi, and older (> 1 year) AdmHiHi mice. Data represented as averages ± SEM from N = 3-4 mice per genotype.

112

Figure 4-7: AM does not directly alter NK Cell Cytotoxicity. (A) Flow cytometry of non- sorted splenocytes and enriched CD49b+ peripheral NK cell fraction following MACS isolation. (B) Spleen weights normalized to body weight from Adm+/+ wild-type and AdmHiHi

113 adult mice. N = 4-6 mice per genotype. (C) cAMP concentrations of cultured CD49b+ NK cells following water, AM, PGE2, or Forskolin treatments. (D) Relative expression of AM receptor, CLR in splenocytes, CD49b positive and negative fractions, as well as whole lung. (E) Fluorescently labeled lymphoma Yac1 cells were cocultured for 4 hours with CD49b+ NK cells treated with water, IL-2, IL-2 and AM, IL-2 and PGE2, AM and PGE2, or Triton. (F) NK cell cytotoxicity assessed by percentage of viable (7-AAD-) vs dead (7-AAD+) Yac1 (GFP+) tumor cells using flow cytometry. Representative data from independent experiments repeated three times with each condition in triplicate.

114

Figure 4-8: Effect of high systemic AM on tumor metastasis through suppression of cytotoxic immune system. (A) Increased tumor cells in the lungs of mice with high systemic AM levels compared to wild-type mice. (A-C) Hypothetical mechanisms by which AM is able to suppress immune surveillance. (A) AM acts indirectly through endothelial cells to increase PGE2 synthesis leading to decreased NK cell cytotoxicity. (B) AM directly suppresses function of other cytotoxic cells, such as T cells, leading to increased metastasis. (C) AM acts indirectly through other immune cells to ultimately suppress NK cell cytotoxicity.

115

Chapter V: The History of AM Receptors: The Forgotten GPCRs

Overview

Before the discovery of the receptor activity-modifying proteins (RAMPs), adrenomedullin (AM) was thought to bind and signal through three GPCRs, including calcitonin receptor-like receptor (CLR), RDC-1, and G10D. G10D became known as the AM receptor, or Admr. However, the pharmacology of RDC1 and Admr was disputed and Admr was orphanized and became known as Gpr182. Since the RAMPs discovery confirmed

CLR as the canonical AM receptor, the historical evidence that AM can bind and/or signal through these other two GPCRs was largely forgotten. However, recent studies have demonstrated the importance of these receptors during development and adulthood and the possibility of them being AM receptors needs to be revisited.

The Rise and Fall of AM and CGRP Receptors prior to RAMP Discovery

Adrenomedullin (AM) was initially identified in an isolated human adrenal pheochromocytoma and could activate cAMP in platelets.76, 341 It quickly became apparent that both AM and related calcitonin gene-related peptide (CGRP) had potent vasodilatory effects in rat mesenteric vascular beds.76, 131, 255 This discovery quickly sparked significant interest to discover the receptors of AM and CGRP in vascular smooth muscle cells (vSMC) that could potentially be targeted therapeutically to treat hypertension and other diseases.

Using primarily radiolabeling experiments with rat 125I-AM and rat 125I-CGRP, numerous reports showed that AM and CGRP could bind rat vSMC and stimulate cAMP production.342,

343 It was later demonstrated that AM could activated cAMP production in bovine aortic

116 endothelial cells, Swiss 3T3 fibroblasts, and rat skeletal muscle, as well as elevated intracellular calcium in bovine adrenal medullary cells and aortic endothelium.344-348 It was also shown that AM was expressed in pancreatic -cells and inhibited insulin secretion in vivo and in cultured pancreatic islets,321 further demonstrating the potential to use AM receptors as a therapeutic target to regulate blood glucose homeostasis. It became apparent that AM and CGRP interacted with the same or similar receptors.349

In the early 1990s, three distinct GPCRs were cloned that would eventually be thought to be the functional AM and/or CGRP receptors. The first was RDC1, which was cloned from canine thyroid.350 A few years later, the rat and human calcitonin receptor-like receptor (CLR) were cloned and shown to elicit a cAMP response with CRF treatment.351-353

The third distinct GPCR, G10D was cloned from rat liver and was expressed in lung, adrenal, liver, kidney, aorta, heart, spinal cord, gut, and testis.354 Sequence alignment found that canine RDC1 was the closest known G10D paralog with the majority of similarity in the transmembrane domains. Unlike Class A RDC1 and G10D, CLR is a Family B GPCR with similar sequence to the calcitonin receptor. A few years later, three other groups independently cloned rat and human GPCRs identical to G10D, which were named L1 and adrenomedullin receptor, respectively.355-357

The initial characterization of G10D was unsuccessful in identifying a functional ligand.354 G10D cDNA was transfected into COS-1 cells and a variety of radiolabeled ligands failed to bind, including angiotensin II, naloxone, leukotriene C4, leukotriene D4, bradykinin, or galanin. Additionally no significant calcium activated-chloride current was detected in

Xenopus oocytes injected with G10D mRNA and treated with angiotensin II, bradykinin, IL-8,

MCAF, MIP-1, RANTES, TNF-, ATP, leukotriene C4 or D4. In a later study, Kapas et al. transfected Admr (G10D/L1) cDNA into COS-7 cells and found that AM treatment, but not

356 125 CGRP, activated cAMP response (EC50 = 7 nM). Additionally, rat I-AM bound these

117 cells with high affinity (KD=8.2 nM). Furthermore, Admr/L1/G10D mRNA expression sites closely matched known 125I-AM binding sites in rat tissues.358. Northern blots revealed human Admr relative expression in a number of tissues (listed highest to lowest expression), including heart, skeletal muscle, liver, lymph node, adrenal medulla and cortex, stomach, spleen, kidney, thyroid, bone marrow, small intestine, testis, brain, and lung.357 This expression pattern mirrored earlier reports of rat G10D expression pattern.354

Since RDC1 was the closest paralog of Admr/L1/G10D, RDC1 was also assessed as an AM and/or CGRP receptor.359 RDC1 was transfected into COS-7 cells, where rat CGRP

(KD = 9.2 nM, EC50 = 3 nM) and AM (KD = 190 nM, EC50 10 nM) could bind these cells and activate cAMP response, while amylin did not bind. A competitive inhibitor CGRP(8-37) did not block the AM induced cAMP response in COS-7 cells likely indicating an alternative binding site or non-specific binding. They concluded that RDC1 was primarily a CGRP receptor, with lower affinities for AM. To try and determine which AM receptor was responsible for vasodilatory effects, Autelitano et al. looked at the expression of L1, CLR, and RDC1 in lungs and a vSMC cell line and found that all three receptors were expressed in the lungs, but only RDC1 was detected in vSMC.360

Entzeroth et al. showed that CLR was also responsive to CGRP and AM and was responsible for their vasodilatory effects.361 Using SK-N-MC neuroblastoma cells they demonstrated that human CGRP (EC50 = 9.65) and AM (EC50 = 7.75) were able to elicit increased cAMP levels. These responses were blocked by the CGRP[8-37] antagonist indicating a specific interaction. CGRP or AM infusion into isolated rat hearts caused vasodilation with EC50 ~ 0.5 nM or 3 nM, respectively. This effect could be blocked with a

CGRP inhibitor further demonstrating that AM and CGRP vasodilation is through a common receptor. Other groups using different human cell lines also demonstrated CLR as a CGRP receptor.349, 362 Together, CLR and to a lesser extent RDC1 were thought to be nonselective

118

AM and CGRP receptors, while G10D/L1/Admr was thought to be the AM selective receptor

(Figure 5-1).

In 1998, Kennedy et al. attempted to validate the previous finding that the rat and human Admr were in fact AM-specific receptors.356, 363 This study transfected COS-7 with either mock pcDNA3 or rat Admr-FLAG and confirmed the surface expression of FLAG epitope in non-permeabilized cells.363 This study found that rat or human 125I-rAM bound identically in both mock and Admr-FLAG transfected cells. Human AM did not displace 125I-

AM bound to COS-7 cells. Furthermore, neither rat nor human Admr activated cAMP response with AM treatment in COS-7 or HEK293T cells, but did elicit cAMP response a positive control Swiss 3T3 cell line. Together, these data demonstrated that the AM binding in COS-7 cells previously reported was likely independent of Admr expression. Therefore

Admr is likely not the AM receptor and thus was reclassified as an orphan receptor, Gpr182.

Additional studies found that L1/Admr/Gpr182 expression was not detected in uterus, rat fibroblasts, or rat vSMCs, which all clearly label with AM, further supporting that Admr does not bind AM in these cell types.344, 360, 364

Shortly after Admr was debunked as the AM receptor, McLatchie et al. published a seminal manuscript demonstrated that CLR interacts with novel receptor activity-modifying proteins (RAMPs) in Xenopus oocytes and transfect HEK293T cells. This interaction was required for CLR trafficking to plasma membrane and when CLR heterodimerized with

Ramp1 it bound and signaled with CGRP with high affinity. Conversely, the CLR/Ramp2 complex acted as the AM receptor.7 It was later shown that cAMP induced by AM in vSMC was through CLR/RAMP2 complex.365 The identification of the RAMP proteins, along with the orphanization of Admr to Gpr182, largely resolved the controversy of the canonical

CGRP and AM receptors. Both RDC1 and Gpr182 were largely discredited as putative AM receptors and forgotten about in regards to AM signaling (Figure 5-1).

119

CLR, Cxcr7, and Gpr182 in a RAMP World

The first in vivo confirmation that CLR/Ramp2 was the canonical AM receptor came from the phenotypic similarities between genetic knockout mice of Adm, Calcrl, and

Ramp2.22, 30, 33 As discussed in Part I, all the knockout mice had midgestational lethality attributed to cardiovascular and lymphatic developmental defects demonstrating this signaling paradigm is required for survival. Mice with genetic loss of CGRP, Ramp1, and

Ramp3 were born in normal Mendelian ratios and had no obvious developmental defects.34,

194, 197 These data indicate either these genes are dispensable during development, CGRP has compensatory isoforms, and/or that CGRP can signal through Ramp2. The functional characterization of AM and CLR was discussed extensively in Part I and recent reviews.366,

367

Cxcr7 knockout animals have early postnatal lethality associated with cardiomyocyte hyperplasia, cardiac enlargement, and valve defects.368-370 There is a huge amount of literature regarding Cxcr7 in the cardiovascular, neural, and immune systems, as well as during cancer. Cxcr7 mainly acts to control chemotactic gradients of SDF-1 availability for

Cxcr4-induced migration.371-375 Cxcr7, like most chemokine receptors, is promiscuous and can bind numerous ligands, including SDF-1, ITAC, and an adrenal opioid proenkaphalin

A.376-378 In addition to cardiac development, Cxcr7 is also important for developmental lymphatic proliferation, potentially through sequestering AM ligand availability.332, 379

Whether or not Cxcr7 is able to directly bind AM and CGRP, as previously reported is controversial.7, 248, 359 It appears clear that AM and Cxcr7 are transcriptionally linked and have some developmental similarities,200, 332, 368 but the lymphatic functions of SDF-1 and

Cxcr4, as well as how Cxcr7 affects CLR expression needs to be elucidated before the lymph-associated phenotypes can be attributed solely to AM (or CGRP).

While CLR and Cxcr7 have been relatively extensively studied since 1998, the third historical AM receptor, Gpr182 was largely forgotten by the research community, with a

120

PubMed search for “Gpr182” in June 2016 yielding one publication.201 Gpr182 remains a

Class A orphan receptor with unknown ligand or downstream signaling. Additionally, browser searches for “adrenomedullin receptor” typically do not distinguish Gpr182 from

CLR/RAMP2. The historical nomenclature of Admr continues to confuse the research community, who identify Admr as part of genome-wide arrays and hypothesize that it functions as the receptor for AM.102, 248, 380, 381 Gpr182/Admr was identified as a novel vascular endothelial marker during zebrafish and mouse development and Gpr182 expression is upregulated in mammary tumor endothelium.201, 381, 382 To date the scientific literature regarding G10D/L1/Admr/Gpr182 has focused on expression by whole tissue

Northern blot, RT-PCR, or whole genome expression pattern, with very limited tissue localization by in ISH or unvalidated antibodies. Considering the lack of a ligand, knowledge of downstream signaling, and research tools, there are currently no published studies elucidating the specific functions of Gpr182 in vivo. The following Chapter VI addresses

Gpr182 spatial and temporal expression patterns and identifies novel functions of Gpr182.

121

Figure

Figure 5-1: Historical perspective of AM pharmacology. Schematic summarizing the AM and CGRP receptors prior to and following the characterization of RAMPs in 1998. Known ligands are represented by solid circles and other partial agonists or theoretical ligands are represented by dotted circles. Cxcr7 and Gpr182 are Class A rhodopsin-like GPCRs and CLR is a Family B secretin-like GPCRs.

122

Chapter VI: Gpr182 inhibits intestinal proliferation during regeneration and adenoma formation4

Overview

Orphan G protein-coupled receptors (GPCRs) provide an untapped opportunity for identification of novel pharmacological targets, yet their expression patterns and physiological functions remain challenging to elucidate. Using a genetically-engineered knock-in reporter mouse, we mapped the expression pattern of the orphan GPCR, Gpr182, during development and adulthood. Notably, Gpr182 is robustly expressed at the crypt base throughout the small intestine, where it is enriched in the crypt base columnar stem cells— one of the most active stem cell populations in the body. Knockdown of Gpr182 expression in mice had no effect on homeostatic intestinal proliferation, but led to significantly increased stem cell proliferation during intestinal regeneration following irradiation-induced injury.

Furthermore, when combined with the mutant ApcMin mouse model, reduced Gpr182 led to significantly more adenomas throughout the intestine and decreased survival of these mice.

Reduced Gpr182 led to increased activation of Erk1/2 in basal and challenge models, demonstrating a key role for this orphan GPCR in controlling negative proliferative signals within the intestine. Importantly, we further show that GPR182 expression is significantly reduced in numerous human carcinomas, including colon adenocarcinoma. Together these results provide new expression analysis of Gpr182 in vivo and are the first to implicate

4 This work is currently [2016] in revision for publication in Journal of Clinical Investigation. Permission for re-printing will be obtained upon publication. Contributing authors include: Daniel O. Kechele, R. Eric Blue, Amanda T. Mah, Scott T. Espenschied, Marni B. Siegel, Charles M. Perou, Scott T. Magness, P. Kay Lund, and Kathleen M. Caron.

123

Gpr182 as a negative regulator of intestinal MAPK signaling-induced proliferation during regeneration and adenoma formation.

Introduction

G protein-coupled receptor 182, Gpr182 (formerly known as G10D or Admr354, 356), is a class A orphan GPCR with nearly nothing known about its expression, function, regulation, or pharmacology. Gpr182 is grouped within the chemokine receptor family by phylogeny, with the atypical chemokine receptor 3 (Ackr3, formally known as Cxcr7 or RDC1) as its closest paralog, despite the two sharing a modest <30% sequence homology in mice or humans. Gpr182 was previously considered a putative receptor for the multifunctional peptide adrenomedullin,356 however these initial findings were not consistent among laboratories363 and it was later shown that adrenomedullin signals through a different GPCR complex.7 Unfortunately, the former Admr (adrenomedullin receptor) nomenclature is sometimes still employed, which leads to confusion in the field. For example, GPR182 was reported to be highly expressed in numerous human pancreatic cancer cell lines and knockdown of GPR182 in these cells decreased xenograft tumor growth, which the authors concluded was due to a loss of adrenomedullin signaling.102, 380 Anatomical expression profiling of the GPCRome demonstrated the relative ubiquitous expression of Gpr182 in most mouse tissues.383 More recently, Gpr182 was found to be highly expressed in developing murine endothelium and enriched in mammary tumor endothelium compared to normal mammary endothelium.201, 381 Additionally, Gpr182 was identified among a group of factors that are significantly altered in a zebrafish model of myeloid leukemia.384 Thus, a significant advance of the current study is the complete expressional profile of Gpr182 using an in vivo mammalian reporter model, where we observed robust expression within the gastrointestinal tract epithelia.

124

The epithelium of the gastrointestinal tract is one of the most dynamic tissues in the adult body, primarily responsible for the absorption of dietary nutrients, and also for fulfilling important endocrine, immune and protective barrier functions. To maintain its proper functions, the intestinal epithelium must continuously turnover, with the entire small intestinal epithelium renewing every week in humans and in mice. This constant renewal is driven by an active population of intestinal stem cells (ISCs) that are located at the base of the crypts of Lieberkühn, where they give rise to rapidly dividing daughter transit-amplifying progenitor cells that differentiate into the absorptive or secretory lineages responsible for functions of the intestine.385-388 Current views are that two distinct pools of ISCs exist in the intestinal epithelium: the crypt base columnar (CBC) ISCs which are positioned between differentiated Paneth cells and mediate normal homeostatic renewal, and “damage- resistant” ISC that act as a reserve ISC activated following injury.385, 386, 388, 389 With the discovery of numerous ISC-specific markers including Lgr5, Lrig1 Ascl2, Olfm4, Hopx, Bmi1, and Sox9, our understanding of both these ISC populations has drastically expanded over the past decade.390-396 It is evident that the activity and proliferation of these ISCs must be tightly controlled by numerous signaling pathways and redundant mechanisms in order to maintain homeostasis in the dynamic gut microenvironment.385, 397 Furthermore, oncogenic mutations specifically in ISCs can drastically enhance adenoma formation in mice.391, 398

Thus, defining factors that regulate ISC proliferation or survival is critical to better understand their functions during homeostasis, damage repair, and cancer, in order to better therapeutically target these ISCs.

In this study, we aimed to identify novel functions of the orphan Gpr182 in vivo by first mapping the murine Gpr182 expression pattern during development and in adulthood, and then by secondly elucidating the effects of Gpr182 reduction on intestinal homeostasis, regeneration, and adenoma formation.

125

Methods

Experimental Animals

The Gpr182tm2a(KOMP)Wtsi/+ (knockout First/promoter driven) mice used in this research project were created from ES cell clone (EPD0365_4_C08) obtained from the NCRR-NIH supported Knockout Mouse Project (KOMP) repository and generated by the CSD consortium for the NIH funded KOMP.399 The CSD targeted allele has been previously published.400, 401 To achieve global deletion, the Gpr182tm2a(KOMP)Wtsi/+ mice were crossed to ubiquitous C57BL/6J-CMV-Cre transgenic mice. For adenoma studies, Gpr182lacZ/+ mice were crossed to the C57BL/6J-ApcMin/+/J (Jackson Laboratory 002020) mice. All Gpr182- associated mouse lines were maintained on an isogenic C57BL/6 background. Previously published Sox9-EGFP BAC transgenic389, 396, 402-404 and Lgr5-EGFP390 mice were used for cell isolation and expression confirmation. Genotyping primers are listed in Supplemental

Table 1.

Animal Procedures and Dissection

For irradiation challenge, 10-14 week old female Gpr182+/+ and Gpr182lacZ/lacZ mice were subject to abdominal ionizing radiation as previously described.389 Briefly, under isoflurane anesthesia, mice received 14 Gy radiation (XRad 320 from Precision X-Ray,

50cm, F1 Filter) to their abdomens. Experimental and control mice were irradiated at the same time, to ensure equivalent radiation dosages and rates (~2.8 Gy/min for 300 sec).

Mice were observed and weighed daily to monitor severity of irradiation and overall health.

Five days post-irradiation, mice were injected intraperitoneally with 4 g of 5-ethynyl-2’- deoxyuridine (EdU) per 1 g body weight. Ninety minutes later mice were euthanized and gastrointestinal tissue was collected for histology and biochemistry. For adenoma studies, male and female mice were monitored for health (physical appearance, body weight, bloody stool) and euthanized for tissue collection at five months. Some mice had to be sacrificed

126 early due to declining health. The small intestines and colons were removed and flushed with cold PBS. The entire intestinal length was measured and then separated into duodenum, jejunum and ileum. The small intestines and colons were opened along the length, flattened, and macroscopic polyp number and size was quantified from all ApcMin/+ mice. Tissues were then processed for histology, whole mount X-gal staining, and biochemistry. For cryogenic sections, tissue was fixed in 4% paraformaldehyde, cryoprotected in 30% sucrose and embedded in OCT before cryosectioning. For histology tissues were fixed in either 4% paraformaldehyde or 10% zinc formalin and then paraffin embedded and sectioned.

Whole Mount X-gal Staining

Whole mount X-gal staining was adapted from previously published protocols.390

Briefly, whole tissue was fixed in 0.2% glutaraldehyde (Electron Microscopy Sciences) for 24 hours at 4°C. Tissue was washed and permeabilized with 0.1% triton and incubated in 1 mg/mL X-gal (Bioline) staining buffer for approximately 24 hours at room temperature in the dark. Tissue was washed and post-fixed in 4% paraformaldehyde for 24 hours at 4°C.

Samples were washed, imaged, and paraffin embedded and sections were counterstained with 1% Neutral Red (Sigma-Aldrich) or Eosin.

Optical Projection Tomography

Whole mount X-Gal stained tissues were embedded in 1% LM agarose, dehydrated overnight in absolute methanol, and cleared in benzyl alcohol:benzyl benzoate (1:2) as previously described.305 Optically cleared specimens were mounted to aluminum chucks and scanned with a 3001m Optical Projection Tomography Scanner (BiOPTonics) under bright field illumination. Data sets were reconstructed into 3D models using nRecon software, and then visualized using either BiOPTonics viewer or ImageJ (NIH).

127

Immunohistochemistry

For IHC, cryogenic sections were hydrated, permeabilized, and blocked with 5% normal donkey serum. When required, slides were boiled in citrate buffer for antigen retrieval. Slides were then stained with primary antibodies, including: chicken anti-- galactosidase (1:1500, Aves Labs), rabbit anti-lysozyme (1:1000, Leica Biosystems), rabbit anti-phospho-Histone H3 (1:500, Cell Signaling) rat anti-Ki67 (1:30, Dako), rabbit anti-- catenin (1:500, ABCAM), rat anti-CD31 (1:100, ABCAM), Syrian hamster anti-podoplanin

(1:200, Developmental Studies Hybridoma Bank), overnight at room temperature. Sections were rinsed, blocked, and incubated with secondary antibodies (1:200, Jackson

Immunoresearch), including: donkey anti-rabbit Alexa Fluor 488, donkey anti-rabbit Alexa

Fluor 594 donkey anti-chicken Alexa Fluor 647, donkey anti-chicken Alexa Fluor 594,

Donkey anti-rat Cy3, Donkey anti-mouse Alexa Fluor 488, Goat anti-Syrian hamster FITC, and Hoechst 33258 (1:1000, Sigma-Aldrich) in the dark for 90 minutes at room temperature.

The Click-iT EdU Alexa Fluor 594 kit (Invitrogen) was used according to manufacturer’s instructions.

Imaging and Image Processing

Whole mount tissue was imaged using a Leica MZ16FA dissecting stereoscope outfitted with a QImaging Micropublisher 5.0 RTV color CCD camera. Paraffin embedded

H&E and X-gal slides were imaged using a Leitz Dialux 20 with QImaging Micropublisher

5.0 RTV color CCD camera or an Olympus BX61 microscope equipped with a QImaging

RETIGA 4000R CCD color camera with Volocity software (Improvision). Fluorescence IHC images were acquired on a Nikon E800 fluorescence microscope with a Hammamatsu Orca

CCD camera with Metamorph software (Molecular Devices Corp.) or a Zeiss LSM 700 confocal microscope. Fluorescent images were pseudo-colored using ImageJ. Image

128 processing, including cropping, brightness, and contrast adjustments were altered equally across comparable images using Photoshop CS4 (Adobe).

Morphometry and Proliferation Quantitation

Crypt depth was quantified by measuring distance from bottom of nuclei at base to top of crypt from all whole open crypts in one cross section from 5 mice per genotype per treatment. Crypt density was assessed by whole mount microscopy. Three fields per intestinal area per genotype were imaged and crypt area from 10 crypts per field was measured using ImageJ. The approximate amount of crypts that would fit into a 1 mm2 area was calculated. Proliferation was assayed by EdU incorporation and the Click-It detection kit and analyzed by counting the number of EdU+ cells per open crypt normalized to total number of DAPI+ crypt cells and expressed as a percentage. Researchers were blinded to mouse genotype and treatment. Cell positions of EdU+ cells along crypt-villus axis were recorded during quantification. 20-60 individual whole open crypts were included from each region of the intestine (duodenum, jejunum, ileum, and distal colon) from 5 mice per genotype per treatment. Polyp dimensions were approximated from all polyps ≥1x1 mm2 throughout the small intestine in all experimental mice and polyp area was calculated assuming elliptical shape. Individual polyp proliferation was determined by threshold quantification of Ki67+ signal as a ratio of DAPI+ signal using ImageJ.

RT-PCR Analysis

Tissues were either snap frozen in liquid nitrogen or stored in RNAlater (Ambion).

RNA was extracted using TRIzol reagent (Invitrogen) and a Percellys bead homogenizer according to standard procedures, and subsequently DNase1 treated (Promega RQ1) and reverse transcribed with M-MLV (Invitrogen) or iScript (BioRad). Quantitative gene expression was assayed with 2x TaqMan Master Mix (Applied Biosystems) and run on a

129

StepOne Plus (Applied Biosystems). Primers and probes for RT-PCR are listed in

Supplemental Table 6-S1. Enriched cell populations were collected using FACS to isolated cells based on differential Sox9-EGFP levels as previously described.389, 402 Likewise, the

Paneth cell enriched population was isolated using a previously published protocol.403

Relative expression levels were determined by the ΔΔCt and normalized to reference gene expression, Gapdh or Rn18S.

Western Blotting and Analysis

Protein was extracted using Lens Homogenization Buffer with DTT and cOmplete protease inhibitor cocktail tablets (Roche) and a bead homogenizer (Percellys) according to standard procedures. Concentrations were determined using Pierce bicinchoninic acid

(BCA) protein assay kit (Thermo Scientific) read on a Mithras LB 940 (Berthold

Technologies). Protein ran on Mini-Protean TGX SDS-PAGE gel (BioRad) and transferred onto nitrocellulose (GE Healthcare) membrane. Blots were blocked and stained in 5% BSA diluted in tris-buffered saline with 0.1% Tween 20. Blots were incubated in primary antibodies including, rabbit anti-p42/44 MAPK (Erk1/2) (1:1000, Cell Signaling), rabbit anti- phospho-p42/44 MAPK (Erk1/2) (Thr202/Tyr204) (1:1000, Cell Signaling), and mouse anti-

Gapdh (1:5000, Novus Biologicals) overnight at 4°C. Blots were rinsed, blocked in 5% nonfat milk, and incubated with secondary antibodies including, goat anti-rabbit Dylight 680

(1:12,000, Thermo Scientific), goat anti-mouse Dylight 800 (1:12,000, Thermo Scientific) in the dark for 60 minutes at room temperature. An Odyssey CLx (Li-COR) was used for imaging. Image brightness and contrast changes were made equally across all comparable images using Image Studio and Photoshop CS4 and p-Erk/Erk quantification by densitometry using ImageJ was normalized to Gpr182+/+ controls. Gapdh was used to ensure equal loading.

130

RNA Sequencing Analysis

RNA Sequencing data was downloaded from the TCGA Data Portal.405, 406 RSEM upper-quantile normalized values from Illumina HiSeq_RNASeqV2 from patient-matched tumor and normal tissue were log2 transformed. Samples with expression <3 are indistinguishable from background and thus were considered as 0.

Statistics

Statistical analysis was performed with GraphPad 5.0 and data represented as mean

± SEM. The unpaired T-test was used for graphs analyzing two groups. For parametric data with >2 comparisons and biological replicates N <10 the 1-way ANOVA with Tukey’s

Multiple Comparison test was used. For nonparametric data analyzing >2 groups and biological replicates N >10 the 1-way ANOVA with Kruskal-Wallis test and Dunn’s Multiple

Comparison test was used. The Mantel-Cox test was used to analyze Kaplan-Meier survival data. Patient-matched tumor and normal tissues were compared with nonparametric

Wilcoxon matched paired T-test and non-matched tumors were compared using unpaired

Mann Whitney T-test. Significant differences are represented as *p < 0.05, **p < 0.01, ***p

< 0.001.

Study Approval

All animal studies were in accordance with approved protocols by the Institutional

Animal Care and Use Committee of the University of North Carolina at Chapel Hill.

Results

Orphan Gpr182 is widely expressed throughout development and adulthood

The Gpr182tm2a(KOMP)Wtsi/+ (knockout first/promoter driven), hereafter referred to as the

Gpr182lacZ/+ mouse, was generated399 and initially used to map murine Gpr182 expression

131 during development and adulthood. A gene trap cassette bearing an En2 splice acceptor upstream of a lacZ reporter was knocked into murine Gpr182 locus immediately downstream of exon 1, resulting in expression of lacZ instead of the endogenous protein coding sequence in exon 2 (Figure 6-1A). Gpr182 lacZ expression was detected by - galactosidase (-gal) activity as early as embryonic day 8.0 (e8.0) in dorsal aorta and vitelline veins (Figure 6-1B and Supplemental Figure 6-S1A). By e13.5, Gpr182 expression was observed in numerous organs, including heart, lung, liver, aorta, and carotid arteries

(Figure 6-1C and Supplemental Figure 6-S1B-C). Embryonic heart expression was present in both atria and ventricles, particularly in the ventricular trabecular region (Figure 6-1D). At e17.5, Gpr182 localization remained prominent in the heart, lungs, and liver, but was also observed in the embryonic kidneys, glandular stomach, intestine, spleen, and yolk sac vasculature (Figure 6-1E and Supplemental Figure 6-S1D).

In adult mice Gpr182 was widely expressed throughout the body. Cardiac expression was localized to the endocardium of the atria and ventricles, as well as in the heart valves and coronary arteries (Figure 6-1F and Supplemental Figure 6-S1E). Notably, little to no staining was observed in cardiomyocytes, fibroblasts, epicardium, or capillary endothelium. Lungs express high levels of Gpr182 primarily in endothelial cells (Figure 6-

1G and Supplemental Figure 6-S1F). Though Gpr182 was expressed in renal tubules during late gestation (Supplemental Figure 6-S1D), adult renal localization was primarily enriched in the glomeruli, where it appears strikingly specific to podocytes (Figure 6-1H and

Supplemental Figure 6-S1G). The sinusoidal endothelial cells, rather than hepatocytes, were the primary cell type expressing Gpr182 in the liver (Figure 6-1I). Gpr182 was expressed throughout the glandular stomach, including both the fundus and antrum (Figure

6-1J and Supplemental Figure 6-S1H). Interestingly, this Gpr182 localization was detected in gastric epithelial cells primarily near the base of the glands in both fundus and antrum,

132 although a few Gpr182+ cells are observed higher up the gland in the fundus. Staining was not detected in acid-secreting parietal cells (Figure 6-1J). Gpr182 X-gal staining was also robust in spermatocytes of the adult testis (Supplemental Figure 6-S1I). Spleen and lymph nodes reveal endothelial Gpr182 localization, with little to no detectable staining in mature or developing hematopoietic and lymphoid lineages (Supplemental Figure 6-S1J-K). Also, we observed little to no X-gal staining in skeletal muscle, pancreas, brain, spinal cord, or dorsal root ganglia of adult mice (Supplemental Figure 6-S1L-O). This murine Gpr182 expression pattern during development and adulthood extends previous microarray data,381, 383 suggesting that Gpr182 is widely expressed and furthermore demonstrates that it is primarily enriched in endothelial cells of most, but not all tissues.

Gpr182 is expressed throughout the intestine and is enriched in small intestinal stem cells

Notably, although Gpr182 is primarily in endothelial cells of most tissues, the gastrointestinal tract expression pattern appears to be restricted to the epithelium. Whole mount X-gal staining and histology revealed the most intense staining in the crypt epithelial cells in the small intestine and colon, with a few positively stained cells in the villi of the small intestine, but relatively absent in enterocytes (Figure 6-2A-C). -gal staining was higher in the stem cell zone at the base of crypts than in the transit-amplifying cells residing higher up the crypt-villus axis (Figure 6-2D, 6-2E). More specifically, -gal stained thin cells with elongated nuclei that are intercalated between lysozyme+ Paneth cells (Figure 6-2E), closely resembling the pattern of other markers of the active CBC ISCs.390-393

To further verify Gpr182 enrichment in CBC ISCs, both Lgr5-EGFP and Sox9-EGFP

BAC transgenic mice were used to isolate different intestinal epithelial cell populations using distinct levels of EGFP, as previously described.389, 390, 396, 402-404 Gpr182 transcript is enriched in the CBC ISC population (Sox9-EGFPLow) when compared to populations

133 containing transit-amplifying cells (Sox9-EGFPSublow) and mixed enterocyte, goblet cell fractions (Sox9-EGFPNegative) (Figure 6-2F). Gpr182 is expressed at very low levels in isolated Paneth cell fractions. Gpr182 was highly enriched in the Sox9-EGFPHigh cell population, which is a mixed population of enteroendocrine cells, tuft cells, and activatable reserve ISCs,389, 403, 404 consistent with the observation of rare -gal+ cells with morphologic similarities to enteroendocrine and tuft cells in the villus region (Figure 6-2B-C). Gpr182 transcript was significantly enriched in Lgr5-EGFPHigh CBC ISC population compared to differentiated Lgr5-EGFPNegative population (Figure 6-2G). Together, these data suggest that

Gpr182 is not a selective CBC marker, but is enriched in CBC ISCs at the crypt base, as well as in Sox9-EGFPHigh population that contains activatable reserve ISCs.

Gpr182 reduction does not alter intestinal proliferation during homeostasis

Gpr182lacZ/lacZ animals have a significant ~85% reduction in endogenous Gpr182 expression in jejunum compared to Gpr182+/+ mice (Figure 6-3A). These Gpr182 knockdown mice are viable and born at expected Mendelian ratios, exhibit normal fertility, and live to adulthood with no overt phenotypic abnormalities. Similarly, when crossed to the ubiquitously expressed CMV-Cre mouse, Gpr182/; CMV-Cre+ mice had negligible Gpr182 expression in the jejunum compared to Cre+ controls (Supplemental Figure 6-S2A). The

Gpr182/+; CMV-Cre+ and Gpr182/; CMV-Cre+ mice had similar X-gal gastrointestinal expression pattern as Gpr182lacZ/lacZ animals (Supplemental Figure 6-S2B-C).

To evaluate the potential roles of Gpr182 in ISCs, we first characterized the basal homeostatic rate of intestinal proliferation in Gpr182lacZ/lacZ mice and consistent with the highly-regulated nature of the ISC niche, we found no significant effects of Gpr182 reduction on homeostatic intestinal proliferation. For example, small intestine and colon lengths were unchanged in the Gpr182lacZ/lacZ animals compared to controls (Figure 6-3B, 6-3C).

134

Intestinal crypt density, as a proxy for crypt fission, was unaltered between the genotypes

(Figure 6-3D). Crypt depth was also unchanged (Figure 6-3E), consistent with no significant differences in basal proliferation, as evidenced by similar numbers of EdU+ cells per crypt, across the genotypes in both the small intestine and the colon (Figure 6-3F, 6-3G).

Likewise, there was no difference in the number of phospho-Histone H3+ cells per crypt in the small intestine (Gpr182+/+: 1.26 +/- 0.13 cells/crypt and Gpr182lacZ/lacZ: 1.33 +/- 0.10 cells/crypt, N = 5). The -Gal+ cells were primarily located in lowest four cell positions of the crypt and there were no significant changes in the position of proliferating transit-amplifying zone cells between the genotypes (Figure 3H). Additionally, whole jejunum expression of mitotic Cyclin D1, or stem cell markers Lgr5 or Olfm4 were unchanged in Gpr182/; CMV-

Cre+ mice compared to Gpr182+/+; CMV-Cre+ controls (Supplemental Figure 6-S2D).

Differentiation also appeared unaltered with all secretory lineages detectable and no relative changes in Hes1, CgA, or Lysozyme expression (Supplemental Figure 6-S2D and data not shown). Collectively, these data demonstrate that genetic reduction of Gpr182 does not lead to altered intestinal proliferation or growth, which is consistent with the highly-regulated control of intestinal proliferation during homeostasis.

Mice with Gpr182 knockdown exhibit intestinal hyperproliferation during intestinal regeneration following irradiation injury

To test whether Gpr182 may be involved with regulating proliferation during intestinal regeneration following damage, Gpr182+/+ and Gpr182lacZ/lacZ were exposed to a high dose

(14 Gy) of ionizing radiation delivered to the abdomen. It is well established that intestinal irradiation (IRR) causes apoptosis of actively dividing cells leading to the temporary loss of proliferative crypts and villi, with subsequent surviving ISCs expanding and regenerating the damaged intestinal epithelium, with peak proliferation occurring five days post-irradiation.389,

135

404, 407, 408 Both Gpr182+/+ and Gpr182lacZ/lacZ IRR animals lost approximately 20% of their initial body weight five days post-IRR compared to non-IRR controls (Figure 6-4A).

Interestingly, the small intestine and colon were significantly longer in the IRR Gpr182lacZ/lacZ mice compared to controls (Figure 6-4B-C). While X-gal stained crypts throughout the small intestine in non-IRR Gpr182lacZ/lacZ mice (Figure 6-4D and Supplemental Figure 6-S2B), but the vast majority of regenerative crypts in the Gpr182lacZ/lacZ IRR mice were X-gal negative

(Figure 6-4D). Gpr182 mRNA expression in whole jejunum decreased at this phase of regeneration in the Gpr182+/+ mice compared to non-IRR controls (Figure 6-4E). Expression of the CBC ISC marker Lgr5 was unchanged between genotypes, but decreased at this phase of regeneration (Figure 6-4F).

During regeneration, the Gpr182lacZ/lacZ IRR small intestines had significant increased crypt depth compared to IRR controls (Figure 6-4G). There were also significantly more

EdU+ cells per crypt in the regenerating duodenum and jejunum of the IRR Gpr182lacZ/lacZ mice compared to IRR Gpr182+/+ controls, with comparable, but not statistically significant, trends in the ileum and colon (Figure 6-4H-I). The increased percentage of EdU+ cells were located in the bottom third of the regenerating crypt (Figure 6-4J). The number of phospho-

Histone H3+ cells per crypt was also significantly increased in the irradiated Gpr182lacZ/lacZ duodenum (Gpr182+/+: 2.51 +/- 0.39 cells/crypt and Gpr182lacZ/lacZ: 3.74 +/- 0.18 cells/crypt, N

= 4-5; *p<0.05). Together, these data indicate a role for Gpr182 in regulating proliferation during intestinal epithelial regeneration.

Mice with reduced Gpr182 carrying ApcMin/+ allele exhibit increased intestinal adenoma burden and decreased survival

Since Gpr182 acts to inhibit proliferation during regeneration, we reasoned that

Gpr182 levels may also be important in deregulated proliferation during disease, such as intestinal adenoma and carcinoma. Gpr182lacZ/lacZ adults up to two years of age did not

136 develop spontaneous intestinal adenomas (data not shown). Thus, to investigate whether

Gpr182 is involved in intestinal adenoma initiation and progression, Gpr182lacZ/lacZ mice were crossed with the well-characterized ApcMin mouse model, which develops spontaneous intestinal and colon adenomas due to aberrant activation of the Wnt/β-catenin signaling pathway.409, 410 Both male and female five-month-old Gpr182lacZ/+; ApcMin/+ and

Gpr182lacZ/lacZ; ApcMin/+ mice were characterized for adenoma formation and compared to

Gpr182+/+; ApcMin/+ controls. The majority of Gpr182+/+; ApcMin/+ (44/47) and Gpr182lacZ/+;

ApcMin/+ (30/32) animals survived the five month experimental period (Figure 6-5A).

However, significantly fewer Gpr182lacZ/lacZ; ApcMin/+ (29/45) mice survived to five months, indicating that genetic reduction of Gpr182 exacerbates the lethality of C57BL/6-ApcMin/+ mice. Furthermore, both Gpr182lacZ/lacZ; ApcMin/+ male and female mice had significantly lower body weights at five months compared to gender-matched controls (Figure 6-5B). The spleens weighed significantly more in both Gpr182lacZ/+ and Gpr182lacZ/lacZ animals when compared to controls (Figure 6-5C). The small intestines were also significantly longer in the

Gpr182lacZ/lacZ; ApcMin/+ animals compared to controls (Figure 6-5D), which is similar to the irradiation response (Figure 6-4C). Together these data suggest that attenuated Gpr182 expression is detrimental to the health and survival of ApcMin/+ animals.

For both sexes, macroscopic polyp number and sizes throughout the intestine were three- to four-fold higher in Gpr182lacZ/lacZ; ApcMin/+ mice compared to Gpr182+/+; ApcMin/+ controls (Figure 6-5E-F). Moreover, the haploinsufficient Gpr182lacZ/+; ApcMin/+ mice also showed significantly more polyps in the small intestine compared to Gpr182+/+; ApcMin/+ controls, further supporting a genetic dose-dependent inverse correlation between Gpr182 levels and polyp formation. Although the relative size distribution of polyps was unchanged between genotypes, we found a statistically significant increase in the number of polyps of all sizes in Gpr182lacZ/lacZ; ApcMin/+ mice compared to Gpr182+/+; ApcMin/+ controls (Figure 6-

5G).

137

The ApcMin/+ model exhibits fewer and more sporadic polyps in the colon after five months. Consistently, colon polyps occurred in 66% of Gpr182+/+; ApcMin/+ assessed, while the incidence of colon polyps was 84% in Gpr182lacZ/lacZ; ApcMin/+ mice. Moreover, there was a two-fold increase in the number of colon polyps in the Gpr182lacZ/lacZ; ApcMin/+ animals compared to Gpr182+/+; ApcMin/+ controls (2.0 +/- 0.3 versus 1.0 +/- 0.2 polyps, respectively;

*p<0.05). Together, these data indicate that decreased Gpr182 expression is sufficient to exacerbate the ApcMin/+ adenoma formation and burden throughout the intestine, further supporting that Gpr182 has a role in negatively regulating intestinal proliferation.

Mosaic Gpr182 expression within adenomas correlates with proliferative heterogeneity

While adenoma formation was increased in Gpr182lacZlacZ; ApcMin/+ animals, no morphological differences were observed in Gpr182lacZ/lacZ polyps compared to Gpr182+/+ polyps (Figure 6-6A). Furthermore, there were no qualitative or quantitative differences in overall proliferation within individual polyps between the Gpr182+/+ and Gpr182lacZ/lacZ animals (Figure 6-6B-C), although the polyp proliferation in the Gpr182lacZlacZ; ApcMin/+ mice appeared more heterogeneous. Overall Gpr182 expression was unchanged in ApcMin/+ polyps compared to normal wild-type jejunum (Figure 6-6D). Macroscopic and histologic analysis of Gpr182-associated X-gal and -gal antibody revealed a mosaic staining pattern between polyps and within individual polyps (Figure 6-6E-H). Abundant X-gal stain was consistently observed in microadenomas (Figure 6-6G). Interestingly, within individual polyps Ki67-labeled proliferation consistently appeared inversely correlated with Gpr182 - galstaining (Figure 6-6H). From this data, we conclude that areas with high Gpr182 reporter/promoter activity exhibit lower proliferation than areas with low Gpr182 reporter/promoter activity, further supporting a function of Gpr182 in inhibiting proliferation.

138

Decreased Gpr182 leads to aberrant Erk1/2 activation during homeostasis, regeneration, and adenoma

Class A GPCRs commonly activate or inhibit the MAPK signaling cascade, which has numerous downstream effects on cell proliferation and survival. Although the signal transduction pathways of the orphan Gpr182 have yet to be characterized, we found significantly increased phosphorylated Erk1/2 (Thr202/Tyr204) in Gpr182lacZ/lacZ jejunum compared to Gpr182+/+ controls, which was consistently observed in Gpr182lacZ/lacZ jejunums during homeostasis, regeneration, and in ApcMin/+ polyps (Figure 6-7A). Therefore, decreased Gpr182 leads to increased Erk1/2 signaling, which is consistent with increased proliferation during regeneration and increased polyp formation observed in the

Gpr182lacZ/lacZ mice (Figure 6-7B).

GPR182 expression is reduced in numerous human carcinomas, including colon adenocarcinomas

Considering the proliferative phenotypes observed in the Gpr182 knockdown mice, we sought to determine whether GPR182 expression levels are changed in human carcinomas. First, we found that GPR182 expression is low in numerous human colorectal carcinoma cell lines, including CaCo2, HT29, Colo205, and SW480 (Figure 6-8A).

Secondly, RNA Sequencing data from The Cancer Genome Atlas (TCGA)405, 406 confirmed that GPR182 expression is significantly lower in colon adenocarcinoma when compared to patient-matched normal colon tissue (Figure 6-8B, top panel) and non-matched samples

(normal colon: 0.86 ± 0.17, N = 41; colon adenocarcinoma: 0.24 ± 0.04, N = 285;

***p<0.0001 Mann Whitney T-test). GPR182 expression is also significantly downregulated in numerous carcinomas in tissues other than the colon, including breast, kidney, liver, lung, pancreas, and thyroid carcinomas (Figure 6-8B). These data further support that low

GPR182 expression in mice and humans is inversely correlated with high proliferation.

139

Discussion

Deorphanization of GPCRs remains an active area of research, especially considering that approximately 40% all approved drugs for humans target only a small fraction of the GPCRome.272, 411 In addition to elucidating the pharmacology of orphan

GPCRs, it is crucial to characterize the anatomical locations and physiological functions of these receptors in vivo. To this end, we generated and employed a genetic Gpr182-lacZ knock-in mouse model to gain insights into the expression pattern and physiological functions of Gpr182. The intestinal expression pattern – particularly the crypt localization – piqued our interest and became the focus of this study. However, it is worth emphasizing that Gpr182 expression is enriched in an interesting variety of tissues and cell types, which will certainly entice additional physiological studies. Furthermore, these types of physiological in vivo studies may prove instructive towards the eventual “deorphanziation”, or identification of endogenous ligand(s), of Gpr182, which currently remains unknown.

Our study shows that Gpr182 is robustly expressed in the epithelium throughout the gastrointestinal tract and is particularly enriched in the CBC ISCs of the small intestine, yet genetic reduction of Gpr182 did not alter intestinal morphology or proliferation during homeostasis. The lack of a basal phenotype in intestinal tissue is not surprising, considering the tightly regulated control of epithelial proliferation for maintaining intestinal health. However, under conditions of stress-induced proliferation, such as regeneration following injury or ApcMin mutation, the reduction of Gpr182 caused increased proliferation likely due, in part, to enhanced Erk1/2 signaling. These results indicate that Gpr182 is normally functioning to inhibit MAPK-induced intestinal proliferation. This conclusion is further supported by our findings of low GPR182 expression in human colorectal carcinoma cell lines and several human carcinomas, including colon adenocarcinomas—a novel finding which significantly expands the growing cadre of negative regulators of proliferation recently shown to be downregulated during colorectal tumorigenesis.391, 412-415

140

MAPK-signaling can play many critical roles during tumorigenesis. For example,

Erk1/2 activation through surface receptors, such as TLR and/or Egfr is required for adenoma formation in ApcMin/+ mice, which can be blocked by Mek/Erk inhibitors.416, 417 In addition, loss of the receptor tyrosine kinase negative regulator Lrig1 leads to constitutively active Egfr and other ErbB receptors leading to increased Erk1/2 signaling, hyperproliferation, and tumorigenesis.391, 418 Most commonly, oncogenic mutations in Kras activates numerous downstream kinases, including Erk1/2, which rarely initiate tumorigenesis alone, but when combined with Apc inactivation, lead to increased adenoma formation and progression.414, 419-423 Finally, an activating mutation in Gs (GNAS R201C) leads to increased cAMP and Erk1/2 activation, which also increased adenoma formation when crossed to the ApcMin/+ mice.424 Interestingly, the number of polyps and relative Erk1/2 activation in these GNAS R201C; ApcMin/+ mice mirror the increase we observed in

Gpr182lacZ/lacZ; ApcMin/+ mice, further supporting the mechanistic involvement of cAMP and

Erk1/2 signaling pathways downstream of Gpr182.

To our knowledge, this is the first study to use a genetic mouse model to simultaneously map Gpr182 expression patterns and elucidate novel physiological functions and signaling that negatively regulate intestinal proliferation during regeneration and adenoma formation. Future studies to identify the ligand for this exciting and physiologically-relevant orphan GPCR will shed light on its tractability as a potential therapeutic target.

141

Author Contributions

D.O.K designed research studies, conducted experiments in Figure 6-1, 6-2, 6-3. 6-

4, 6-5, 6-6, 6-7, 6-8, 6-S1 and 6-S2, analyzed data, wrote manuscript. R.E.B. conducted mouse irradiations in Figure 6-3 and 6-4 and provided technical and intellectual support.

A.T.M. acquired data Figure 6-2F, provided technical and intellectual support, and edited manuscript. S.T.E. acquired data Figure 6-2A and 6-S1F, provided technical and intellectual support, and edited manuscript. M.B.S. and C.M.P. conducted bioinformatics Figure 6-8B and provided technical support. S.T.M. acquired data Figure 6-2F, provided reagents, and intellectual support. P.K.L. helped design research studies, provided reagents, technical and intellectual support, and edited manuscript. K.M.C. designed research studies, provided intellectual support and edited manuscript.

142

Figures

Figure 6-1: Murine Gpr182 expression profile during development and adulthood. (A) The targeting vector of the Gpr182tm2a(KOMP)Wtsi/+ (knockout first/promoter driven) lacZ reporter mouse model.399-401 (B) Whole mount X-gal staining of e8.0 Gpr182lacZ/lacZ embryo. DA, dorsal aorta; VV, vitteline vein. (C) Optical projection tomography (OPT) of whole mount X- gal stained wild-type and Gpr182lacZ/lacZ e13.5 embryos. X-gal staining in Gpr182lacZ/lacZ e13.5 heart (D) and e17.5 (E) lung, kidney, liver, yolk sac, and stomach, pancreas, and duodenum. Representative Gpr182 lacZ expression in adult heart (F), lung (G), kidney (H), liver (I), and glandular stomach (J) stained with X-gal and/or -gal (green). Sections were costained with DAPI (purple) and either the endothelial marker CD31 (F) or the podocyte marker, Pdpn (H). X-gal stained sections were counterstained with eosin (F-H) or Neutral Red (D, I-J). Scale bars represent 200 m (B, D), 1 mm (C, E), and 100 m (F-J).

143

Figure 6-2: Gpr182 is enriched in CBC intestinal stem cells. (A) Optical projection tomography of X-gal stained optically cleared duodenum from Gpr182+/+ and Gpr182lacZ/lacZ adult mice. (B) Representative whole mount X-gal stained small intestine and colon of adult Gpr182lacZ/lacZ mice. (C) Cross-sectional X-gal from adult Gpr182lacZ/lacZ small intestine and (D) -gal (green) IHC from adult Gpr182+/+ and Gpr182lacZ/lacZ intestine. (E) Crypt of Gpr182lacZ/lacZ jejunum costained with -gal (green), the Paneth cell marker lysozyme (magenta), and DAPI (blue). Yellow asterisks mark CBC ISCs as defined by morphologic position, elongated nuclei, and lysozyme negative. Scale bars represent 500 m (A-B) and 100 m in low magnification (C) and 10 m in crypt images (C-E). Relative Gpr182 expression in isolated cell populations from the (F) Sox9-EGFP BAC transgenic mouse or (G) Lgr5-EGFP mouse. Enterocytes/goblet cells (Sox9-EGFPNegative); transit-amplifying cells (Sox9-EGFPSublow); CBC ISC (Lgr5-EGFPHigh and Sox9-EGFPLow); enteroendocrine/tuft/+4 ISC (Sox9-EGFPHigh); Paneth cells (Sox9-EGFPNegative:CD24High:SSCHigh); early progenitors/+4 ISC (Lgr5-EGFPLow); and differentiated cells (Lgr5-EGFPNegative). Expression was normalized to Sox9-EGFPNegative or Lgr5-EGFPNegative and -actin or Rn18S. Biological N = 2-5 (F). Significance was determined by 1-way ANOVA with Tukey’s Multiple Comparison Test.

144

Figure 6-3: Gpr182 knockdown does not alter basal proliferation. (A) RT-PCR analysis of Gpr182 expression in adult jejunum from Gpr182+/+, Gpr182lacZ/+, and Gpr182lacZ/lacZ mice. Expression was normalized to Gpr182+/+ and Gapdh. (B-C) Small intestine and colon length from adult Gpr182+/+ and Gpr182lacZ/lacZ mice. (D) Morphometric quantification of crypt density and (E) histological quantification of crypt depth between Gpr182 genotypes. (F) Representative images and (G) EdU incorporation quantification of intestinal proliferation between Gpr182+/+ and Gpr182lacZ/lacZ animals. (H) Analysis of the cellular position of EdU+ cells along the crypt-axis expressed as a percentage of the total number of cells in that position in all crypts. The Gpr182 Zone (green gradient) is the relative cellular position of - Gal+ cells along crypt-axis. Biological N = 5 per genotype. Scale bars represent 100 m. Significance was determined either by the 1-way ANOVA with Tukey’s Multiple Comparison Test (A) and Student’s (B-E, G) or Mann Whitney T-test of area under curve (AUC) (H).

145

Figure 6-4: Decreased Gpr182 leads to hyperproliferation during the regeneration phase after irradiation-induced injury. Adult Gpr182+/+ (purple) and Gpr182lacZ/lacZ (green) mice irradiated (IRR) with a 14 Gy dose radiation to abdomen. (A) Body weight changes (% of pre-IRR weight) of Gpr182+/+ and Gpr182lacZ/lacZ mice for five days following IRR compared to non-IRR controls. (B-C) Small intestine and colon lengths from IRR Gpr182+/+ and Gpr182lacZ/lacZ animals. (D) Whole mount X-gal stained small intestine of non-IRR Gpr182lacZ/lacZ and IRR Gpr182lacZ/lacZ animals five days post-IRR. Relative (E) Gpr182 and (F) Lgr5 expression in whole jejunum from Gpr182+/+ IRR, Gpr182lacZ/lacZ IRR animals and non-IRR controls. Expression was normalized to non-IRR Gpr182+/+ and Gapdh. (G) Crypt

146 depths in regenerating intestine from Gpr182+/+ IRR and Gpr182lacZ/lacZ IRR animals. (H) Representative images and (I) EdU quantification of intestinal proliferation between IRR Gpr182+/+ and Gpr182lacZ/lacZ animals. (J) Analysis of the cellular position of EdU+ cells along the crypt axis expressed as a percentage of the total number of cells in that position in all regenerating crypts. Biological N = 4-5. Scale bars represent 5 mm (D) or 100 m (H). Significance was determined by Student’s T-test (B-C, G, I), one-way ANOVA with Tukey’s Multiple Comparison Test (E-F), or Mann Whitney T-test of AUC (J).

147

Figure 6-5: Reduced Gpr182 increased lethality and polyp number in ApcMin/+ mice. (A) Kaplan-Meier survival curve comparing Gpr182lacZ/lacZ Apc+/+ (black), Gpr182+/+; ApcMin/+ (yellow), Gpr182lacZ/+; ApcMin/+ (blue), and Gpr182lacZ/lacZ; ApcMin/+ (red) animals. (B) Body weights of five month old female and male ApcMin mice from different Gpr182 genotypes. (C) Spleen weight normalized to body weight from Gpr182+/+; ApcMin/+, Gpr182lacZ/+; ApcMin/+, and Gpr182lacZ/lacZ; ApcMin/+ mice. (D) Small intestine length from Gpr182lacZ/lacZ; ApcMin/+ animals compared to controls. (E) Representative images of adenoma number and size from Gpr182+/+; ApcMin/+ and Gpr182lacZ/lacZ; ApcMin/+ intestines. Scale bar represents 5 mm. (F) Quantification of polyp number throughout the small intestine of Gpr182+/+; ApcMin/+, Gpr182lacZ/+; ApcMin/+, and Gpr182lacZ/lacZ; ApcMin/+ mice. (G) The distribution of the number of polyps throughout the small intestine based on approximate polyp size. All significant trends were observed in both males and females. Biological N = 25-50 per genotype (A), N = 10- 20 per genotype per gender (B), N = 20-35 per genotype (C-D, F-G). Significance was determined either by the Mantel-Cox Test (A) or 1-way ANOVA Kruskal-Wallis Test and Dunn’s Multiple Comparison Test (B-G).

148

Figure 6-6: Mosaic Gpr182 staining within polyps corresponded to polyp proliferation heterogeneity. (A) Histology from five month old from Gpr182+/+; ApcMin/+ and Gpr182lacZ/lacZ; ApcMin/+ jejunum polyps labeled with H&E and -catenin (white). (B-C) Polyp proliferation assessed by Ki67 (blue) staining normalized to DAPI (purple) in Gpr182+/+; ApcMin/+ and Gpr182lacZ/lacZ; ApcMin/+ mice. (D) Gpr182 expression in polyps from jejunum of Gpr182+/+; ApcMin/+ and Gpr182lacZ/lacZ; ApcMin/+ mice compared to Apc+/+ jejunum. Expression was normalized to Gpr182+/+ and Rn18S. Macroscopic (E) and histologic (F-G) X-gal staining of Gpr182lacZ/+; ApcMin/+ polyps. (H) Serial sections stained with -Gal (green), Ki67 (blue), and DAPI (magenta) in Gpr182lacZ/lacZ polyps. Biological N = 15-60 polyps from 5 mice per genotype (C) and N = 4-5 (D) Scale bars represent 100 m (A-B, F-H) and 1 mm (E). Significance was determined either by the Mann-Whitney T-test (C) or 1-way ANOVA with Tukey’s Multiple Comparison Test (D).

149

Figure 6-7: Gpr182 knockdown leads to elevated Erk1/2 signaling upstream of the hyperproliferative intestinal crypt microenvironment. (A) Representative immunoblots and quantification of relative phosphorylated Erk1/2 between Gpr182+/+ and Gpr182lacZ/lacZ from unchallenged, 5 day post-irradiation, and ApcMin/+ polyps from whole jejunum. Samples were normalized to total Erk1/2 and Gpr182+/+. N = 4-7 per genotype per condition. Significance was determined by unpaired T-test. (B) Model summarizing Gpr182 -Gal expression pattern (blue) in the whole mouse (heart, lungs, liver, stomach, small intestine, colon, kidney, and testis) with more specific expressional detail within the small intestine, as well as Gpr182 function to inhibit intestinal Erk1/2 signaling and proliferation.

150

Figure 6-8: Low GPR182 expression in human carcinomas. (A) RT-PCR of GPR182 expression in human cell lines, including colorectal carcinoma cell lines, intestinal epithelial cell (hIEC), and venous endothelial cells (HUVEC). Expression was normalized to hIEC and GAPDH. (B) Relative GPR182 expression from TCGA achived RNA-seq in patient-matched normal (N) and tumor (T) tissue. Breast invasive carcinoma (BRCA), N = 98; colon adenocarcinoma (COAD), N = 26; kidney chromophobe (KICH), N = 25; kidney renal clear cell carcinoma (KIRC), N = 71; kidney renal papillary cell carcinoma (KIRP), N = 32; liver hepatocellular carcinoma (LIHC), N = 50; lung adenocarcinoma (LUAD), N = 57; lung squamous cell carcinoma (LUSC), N = 50; pancreatic adenocarcinoma (PAAD), N = 58; rectum adenocarcinoma (READ), N = 6, stomach adenocarcinoma (STAD), N = 32; thyroid carcinoma (THCA), N = 57. Significance was determined by the Wilcoxon matched pairs T- test.

151

Supplemental Table

Forward Primer Reverse Primer Allele Taqman Probe (5’ – 3’) (5’ – 3’) CTGCAGCCTCC CATTGTCCGGT Gpr182 Wild-type TGGCACTAACA TCCAAGGTGGA GC GAC tm2a(KOMP)Wtsi GGGAGGATACC Gpr182 GAGATGGCGCA ACAGGGAAATA Targeted ACGCAATTAAT GAGC TTCACTGGCCG ATGTGAGCGAG Gpr182 lacZ TCGTTTTACAAC TAACAACCCGT Targeted GT CGGATTCT GCCATCCCTTC TTCCACTTTGG Apc Wild-type ACGTTAG CATAAGGC TTCTGAGAAAG TTCCACTTTGG Apc Min Allele ACAGAAGTTA CATAAGGC Mm01946034_s1 mouse Gpr182 (Admr) (Applied Biosystems) Mm00438890_m1 mouse Lgr5 (Thermo Scientific) Mm01320260_m1 mouse Olfm4 (Thermo Scientific) Mm00432359_m1 mouse Cyclin D1 (Thermo Scientfic) Mm01342805_m1 mouse Hes1 (Thermo Scientific) Mm01209400_m1 mouse CgA (Thermo Scientific) Mm00657323_m1 mouse Lysozyme (Thermo Scientific) Hm01922099_s1 human GPR182 (Applied Biosystems) Mm03928990_g1 mouse Rn18s (Applied Biosystems) 4310884E (Thermo human GAPDH Fisher Scientific) Mm02619580_g1 mouse -actin (Thermo Scientific) Mm99999915_g1 mouse Gapdh (Applied Biosystems)

Supplemental Table 6-S1: RT-PCR and Genotyping Primers and Probes

152

Supplemental Figures

Supplemental Figure 6-S1: Additional murine Gpr182 expression during development and adulthood. Representative whole mount X-gal staining of Gpr182lacz/lacZ tissue during development (A-D) and adulthood (E-O). X-gal staining (A) of an e8.0 embryo, (B) OPT, and (C) microdissected of an e13.5 embryo. (D) Gpr182 lacZ in heart, lung, liver, kidney, stomach, and small intestine of an e17.5 embryo. Representative images of X-gal staining in adult heart (E), lung OPT (F), kidney (G), stomach (H), testis (I), spleen (J), mesenteric lymph node (K), hindlimb skeletal muscle (L), pancreas (M), brain (N), and spinal cord (O).

153

(E) LV, left ventricle; RA, right atria; V, cardiac valve; A, aorta; (H) E, esophagus; Fo, forestomach; F, fundus; An, antrum; (O) DRG, dorsal root ganglia. X-gal sections were counterstained with Neutral Red (D, I-K, M). Scale bars represent 200 µm (A,D,I-K,M) or 1 mm (B,C,E-H,L,N,O).

154

155

Supplemental Figure 6-S2: Gpr182 knockout does not alter X-gal expression pattern or cell markers. (A) RT-PCR analysis of Gpr182 expression in whole jejunum from ubiquitously expressed CMV-Cre+ Gpr182+/+, Gpr182/+, Gpr182/, and CMV-Cre negative Gpr182lacZ/lacZ adult mice. (B) Representative whole mount X-gal stained small intestine and colon of adult Gpr182/+; CMV-Cre+ mice. (C) Representative cross-sections from whole mount X-gal stained stomach, small intestine, and colon from Gpr182+/+, Gpr182/+, Gpr182/ animals. (D) Expression of cyclin D1, Lgr5, Olfm4, Hes1, CgA, lysozyme in whole jejunum from Gpr182+/+; CMV-Cre+ and Gpr182/; CMV-Cre+ mice. Expression was normalized to Gpr182+/+; CMV-Cre+ and Rn18S or Gapdh. Biological N = 3-5. Significance was determined by 1-way ANOVA with Tukey’s Multiple Comparison Test (A) or Student’s T-test (D).

156

Chapter VII: Conclusions and Future Directions

Summary

Part II of this dissertation focused on the orphan GPCR, Gpr182, of which almost nothing has been published. Chapter V summarized Gpr182 in a historical context and how

Gpr182 was initially identified as an adrenomedullin receptor, then orphanized, and largely forgotten for over a decade. Chapter VI describes a novel role of Gpr182 in regulating adult intestinal proliferation during regeneration and adenoma formation. Gpr182 is enriched in intestinal stem cells, which are one of the most proliferative cell populations during adulthood and have been directly implicated in nearly every gastrointestinal disease.

Considering relatively nothing is known regarding the pharmacology and functions of

Gpr182, this initial study will help lay the groundwork to understand this GPCR during development, homeostasis, and disease. Chapter VII will further highlight Gpr182 in the broader context of gastrointestinal physiology and disease and provide numerous future directions to better understand the potential of Gpr182 as a therapeutic target.

Current State of the Field

Intestinal Stem Cell Identity

The intestinal stem cell (ISC) field has exploded over the past decade and now ISCs are one of the best characterized stem cell populations in the body. Discoveries have been fueled by the identification of numerous ISC-specific markers, genetic techniques to manipulate genes specifically within these ISCs, and the development of ex vivo culture techniques to further study ISCs functions.385, 386 Understanding the genetic signature and

157 function of ISCs remains a very active area of research. Since its initial characterization by

Barker et al. the Lgr5-EGFP-ires-CreERT mouse has remained the predominant model used in the vast majority of ISC studies and is almost required to verify novel ISC markers and functions.390, 425 Single cell transcriptome analysis has revealed that ISCs are very heterogeneous.403, 426 The use of only one ISC marker, such as Lgr5, could skew the field towards select ISC subpopulations while excluding other ISC populations. Single cell expressional analysis will hopefully identify additional ISC-specific markers that have been missed previously due to the use of heterogeneous cell populations. Additional ISC markers, specifically surface markers and targetable receptors, would be instrumental to the field.

Crypt isolation and fluorescent reporters in mice have allowed for the ex vivo culture of whole crypts or single Lgr5-EGFP+425 or Sox9-EGFP+402, 403 ISCs, respectively. These techniques are commonly used to verify in vivo results, but the use of fluorescent proteins to isolate ISCs is not amenable for studying human ISCs. Furthermore, these techniques isolate the intestinal epithelium, but exclude the other cell types present in the complex ISC microenvironment. New techniques have emerged that utilize reprogramed human pluripotent stem cells into multicellular intestinal organoids, which better model human gastrointestinal development and diseases.427-431 Using these organoids to model various human gastrointestinal diseases is a new and exciting area of research. Stomach organoids have been used to study the mechanisms of H. pylori infection on the gastric epithelium to further understand the pathogenesis of human gastric cancer.427 Furthermore, gene editing techniques, like CRISPR/Cas9, are being utilized in mouse intestinal organoids and potentially human organoids to study the consequences of common genetic mutations.432

Identifying and understanding the factors that regulate ISC growth and survival, ex vivo and in vivo, is crucial for the translational success of ISCs in humans therapeutics.

While the ISC niche of the small intestine is well-defined, the stem cell niches of other gastrointestinal tissues, such as the esophagus, stomach, and colon, are far less

158 understood. Few gastric or colonic stem cells-specific markers have been identified. Lgr5 is expressed in select cells at the base of colonic and gastric antral glands. Lineage trace experiments confirmed that these Lgr5+ cells can give rise to the differentiated epithelial cells types within these tissues, although it is very heterogeneous.390, 433 Lgr5 is not expressed in the fundus region of the stomach. The fundus is thought to have two distinct stem cell populations, one at the base of gland that is fairly quiescent and the other residing near the top of the gland that can rapidly proliferate and undergo bidirectional migration in order to maintain epithelial barrier function under constant acidic insult.434, 435 Very few stem cell markers for this region of the stomach have been identified. Sox2 is expressed in the stem cells zones of both the fundus and antrum and Sox2+ cells are multipotent.436 Mist1 is also expressed in gastric stem cells and maintains homeostatic and neoplastic proliferation, partly through Cxcr4/SDF1 signaling in the endothelium and innate immune cell-derived

Wnt5a.437 Additional studies have identified that differentiated pepsinogen-producing chief cells in the antrum (Lgr5) or fundus (Troy1, Mist1) can act as reserve stem cells that are activated upon damage.433, 438 Notch1 and Notch2 signaling is essential to regulate Lgr5+ antral stem cells proliferation, apoptosis, and differentiated.439, 440 The localization of Gpr182- lacZ within the stomach epithelium is consistent with the predicted stem cell niches of both the fundus and antrum, which merits further investigation into the functional roles of Gpr182 in the stomach.

Intestinal Damage and Regeneration

Understanding different ISCs populations and their roles following intestinal damage remains an intense area of research. While the CBC ISCs are the primary stem cells responsible for the homeostatic turnover of the intestine, there are reserve/activatable ISCs that are required when CBC stem cells are loss due to intestinal damage caused by irradiation, infection, or toxins. Numerous cell types at different stages of differentiation can

159 dedifferentiate into regenerative proliferative ISCs following intestinal damage. More specifically, it has been demonstrated both by lineage trace experiments and organoid formation that label retaining cells, Dll1+ secretory progenitors, Sox9-EGFPHigh cells, Dclk1+ tuft cells, or Alpi+ enterocytes can all dedifferentiate to repopulate the intestine upon epithelial damage.389, 404, 407, 441-445 It is interesting that Gpr182 is primarily enriched in CBC

ISCs, yet the proliferation phenotype is primarily observed during intestinal regeneration

(Figure 6-4), which suggests that it may be expressed in both CBC and activatable ISC populations, similar to Lrig1.391, 418 There is significant interconversion between these two

ISC populations, which is regulated partly by Notch signaling.403, 404, 418, 446 The number of surviving Lgr5+ ISCs following damage is also important for regeneration and survival, with either 15 Gy IRR or diphtheria toxin Lgr5 ablation leading to the complete loss of Lgr5+ ISCs and the failure of the epithelium to regenerate.408, 447 The 14 Gy dose of IRR used on

Gpr182lacZ/lacZ mice depleted the vast majority of CBC ISCs (Figure 6-4D). This also supports the hypothesis that Gpr182 downregulation or loss leads to a more proliferative microenvironment, which is essential during regeneration. Gpr182 may be involved in returning the intestinal proliferation back to homeostasis following damage, a process of which not much is known.

While it is clear that the Wnt signaling pathway is the main driving force of intestinal proliferation during homeostasis, there is significant interest to understand other pathways that regulate ISC proliferation during regeneration. Mice with ablated Sox9 expression have a blunted regenerative response and succumb to epithelial damage-induced lethality.404

High fat diet regulates ISC numbers and function basally and following IRR damage, which is regulated by IGF signaling.448-451 The Notch pathway is also important in maintaining ISC numbers and proliferation. Pharmacologic inhibition or genetic removal of the Notch receptors decreases CBC ISC numbers and proliferation during homeostasis and these mice fail to successfully regenerate their epithelium following IRR damage.452, 453 Notch

160 signaling regulates the interconversion between the active Lgr5+ CBC ISCs and the activatable Bmi-1+ reserve ISCs.446 The YAP/Hippo pathway is also critical for intestinal regeneration. YAP signaling is able to reprogram CBC ISCs to suppress Wnt signaling in order to activate the EGF/EGFR pathway, which is essential for regeneration.454, 455 Gpr182 plays a role to regulate MAPK signaling that is essential for regenerative proliferation, supporting why the primary phenotype observed in the Gpr182lacZ/lacZ animals is during regeneration.

Mucosal-stem cell crosstalk activated by microbiota and/or inflammation has been a major area of research for years. The functional characterization of tuft cells in vivo has recently moved to the forefront of the field. It was recently demonstrated that upon parasitic infection of the intestine, the ISCs increase the production of tuft cells, which are required for secreting IL-25 in order to mount an effective innate immune response to resolve the infection.456, 457 Furthermore, deletion of tuft cell marker Dclk1 impairs epithelial regeneration following radiation injury due to a large number of changes to ISCs expression signature.442 Innate lymphoid cells produce IL-22 following intestinal injury, which increases organoid growth and the regenerative response through the JAK/STAT3 pathway in ISCs.458

The innate immune system also provides SDF1 leading to enhanced tumor growth.437 While the direct effects of Gpr182 on the innate immune system are unknown, it is clear that epithelial damage by irradiation or infection activates innate lymphoid cells, which support the regenerative response of the epithelium.

Pathogenesis of Intestinal Adenomas and Colorectal Carcinomas

The majority of research regarding the regulation of ISCs during homeostasis and regeneration has direct implications in the pathogenesis of gastrointestinal neoplastic diseases. The majority of studies utilize the spontaneous adenoma ApcMin or Apcflox/flox mouse models due to Apc ability to initiate adenoma formation and the high incidence of

161

APC mutations present in human carcinomas. Adenomas are benign growths usually initiated by aberrant Wnt signaling activation. Adenomas progress to malignant carcinomas with additional oncogenic mutations to the MAPK (Kras,Braf), BMP (SMAD4), and/or Akt

(PTEN, p53) signaling pathways. Interestingly, these are same signaling pathways used to grow and maintain organoids in culture. There is strong evidence that gastrointestinal adenomas are initiated and driven by oncogenic mutations specifically within ISCs, but it remains an active area of research to determine if these acquired mutations are in the rapid or quiescent ISC populations.391, 395, 398, 459, 460 There is evidence of competition between

ISCs within individual crypts and ISCs with oncogenic mutations typically have a selective growth advantage and are more likely to out compete the normal ISCs. Adenomas can arise from a small number of mutated ISCs, but larger adenomas are typically heterogeneous because they consist of mutations in ISCs from adjacent crypts.461462, 463

Unlike Apc mutations, activating mutations in Kras or loss of PTEN does not initiate adenoma formation, but rather cause adenoma growth and progression when combined with Apc mutations.419, 464, 465 As discussed in Chapter VI, the MAPK pathway is critical for adenoma formation and the Gpr182lacZ/lacZ; ApcMin mice have high Erk1/2 activation and increased adenoma formation (Figure 6-5 and 6-7). Zebrafish KrasG12D mutations increased Erk1/2 activation, but did not alter cell proliferation without homozygous loss of

Apc.421 YAP-/- mice are almost completely protected from Apc-induced adenoma formation, primarily through transcriptional downregulation of the EGF/EGFR/ERK signaling pathway.416, 454 The PPAR pathway, which is downstream of high fat diet and was shown to alter the proliferation of Lgr5+ ISCs, is also involved in the progression of adenomas to carcinomas in ApcMin mice.466

The identification of potential targets to block adenoma formation and colorectal carcinoma growth remains an intense area of research. Gpr182 may be a targetable GPCR to inhibit adenoma formation similar to that observed with EGFR and MEK inhibitors.416

162

Targeting pathways upstream of MAPK signaling, such as GPCRs or YAP signaling, could block adenoma formation.467 ApcMin adenomas are responsive to rapamycin-induced mTORC inhibition, but only if Kras and Erk1/2 signaling are unchanged.423 Apc loss in

Dclk1+ cells can initiate adenomas, but only if the intestine is undergoing regeneration following damaged.441 Dclk1 is typically not expressed in ISCs, but is specifically upregulated in cancer stem cells, so targeting this glycoprotein could be selective towards mutated ISCs within adenomas and not normal ISCs.468 Using a doxycycline-inducible Apc mouse model, Dow et al. demonstrated that reintroducing Apc expression into mice with adenomas could not only prevent the formation of additional adenomas, but could also return adenomas back to normal mucosa.420 With the use of elaborate genetic model systems, the ISC field continues to expand with added layers of complexity and alternative signaling pathways that play a role in regulating the function of ISCs during homeostasis, regeneration, and cancer.

Future Directions

Gpr182 Expressional Regulation and Roles in Cardiovascular Development and Disease

Gpr182 may be important in the development and maintenance of other stem cell niches, such as the hematopoietic stem cells. Hematopoietic stem cells originate in yolk sac as hemangioblasts and move to fetal liver as definitive progenitors. These populations give rise to both the primitive and definitive RBCs and the hematopoietic stem cells in fetal liver ultimately populate the thymus and bone marrow.469-471 In addition to the hemangioblasts from the yolk sac, it was recently demonstrated that hematopoietic stem cells emerge from specialized dorsal aortic endothelium, known as hemogenic endothelium, which undergoes endothelial to hematopoietic transition.472-474 Our lab and others have observed Gpr182 localization in the dorsal aorta, vitelline vein, and hemangioblast-like cells in the yolk sac during early murine development (Figure 6-1A and Figure 6-S1A). Additionally, Gpr182 is

163 present in the fetal liver and yolk sac at later stages of embryonic development (Figure 6-

1C, E and Figure 6-S1D),381 as well as in the bone marrow, spleen, and lymph nodes during adulthood (Figure 6-S1J-L). These anatomical locations are consistent with the reported locations of the hemogenic endothelium, primitive, and definitive hematopoietic progenitor populations during development and adulthood. Interestingly, Gpr182-lacZ does not appear to be expressed in mature hematopoietic, myeloid, or lymphoid lineages (Figure 6-S1J-K).

Gpr182 expression in the dorsal aorta and hemogenic endothelium was further validated in a recent study that identified Gpr182 as one of the highest enriched genes (5/516 altered genes) in GFP+ (under the Runx1 +23 enhancer) hemogenic endothelium compared to non- hemogenic endothelium from e8.5 dorsal aorta endothelium.475 Another group overexpressed HRas and/or Notch intracellular domain (NICD) in zebrafish hemogenic endothelium in order to create a model of myeloid leukemia. They also identified Gpr182 as

1/23 genes to fit the signature of being two-fold enriched by overexpression of Ras and two- fold decreased by Notch overexpression.384 These studies not only support that Gpr182 is enriched in hemogenic endothelium, but also suggest Gpr182 may be involved in hematopoietic stem cell development and the pathogenesis of myeloid leukemia. Further hematological workup is needed to confirm this hypothesis and to understand how Gpr182 affects hematopoietic development and disease.

The data from Alghisi et al. also suggest that Ras and/or Notch signaling may be transcriptional upstream of Gpr182. There is evidence that Ras can suppress Notch signaling,384 so the increased Gpr182 expression with HRas overexpression may be due to inhibited Notch signaling and not directly caused by increased HRas. This regulation potentially extends beyond the hemogenic endothelium and Notch signaling appears to be inversely correlated with Gpr182 levels in multiple tissues. For example, Gpr182 expression is highly enriched in mammary tumor endothelium compared to normal mammary endothelium,201 which could reflect the transcriptional profile of developmental endothelium.

164

It was recently observed that mammary tumor endothelium is heterogeneous and particular tumor endothelial cells had decreased Notch1 expression, which potentially supports the increased Gpr182 expression.296 It remains unclear if the elevated Gpr182 expression in tumor endothelium has distinct roles in tumor angiogenesis, tumor vasculature integrity, and/or altered metastatic properties. Beyond tumors, Notch signaling between the endocardium and the myocardium is essential for trabeculae formation and ventricular maturation of the heart.280-283 Gpr182 expression was observed in the endocardium at similar developmental stages (Figure 6-1D, F and Figure 6-S1D), suggesting that it may be involved in or downstream of cardiac trabeculae development. To further support the hypothesis that Notch signaling suppresses Gpr182 expression beyond the endothelium, I utilized epithelial RNA (kindly provided by Dr. Linda Samuelson, University of Michigan) from mice overexpressing NICD under the Villin, intestinal epithelium-specific, promotor (Figure

7-1A). Intestinal epithelial-specific activation of NICD significantly reduced Gpr182 expression compared to controls (Figure 7-1B), which is similar to that observed with NICD overexpression in zebrafish endothelium.384 These data support the notion that Gpr182 expression is likely downstream of the Notch pathway. Notch signaling governs the early fate decisions of hemogenic endothelium, cardiac development, intestinal epithelium differentiation, and is dysregulated in oncogenic diseases; all of which Gpr182 could be playing a supporting role and requires further studies.

In addition to developmental endothelial expression, Gpr182 is highly expressed in adult atria and ventricle endocardium, cardiac valves, and the endothelium of major arteries

(Figure 6-1F, Figure 6-S1B-E). Additionally, Gpr182 was expressed in podocytes and glomerulus of the kidney (Figure 6-1H). It would be interesting to investigate the effects of

Gpr182 loss (using CMV-Cre, conditional Cdh5-Cre, or podoplanin-Cre) on basal cardiac function, blood pressure, vessel compliance, angiogenesis, and/or renal function.

Considering the overall survival and wellness of Gpr182 KOs appears unchanged, these

165 mice could be challenged with different cardiovascular disease models, including ischemia reperfusion, AngII-induced hypertension, high salt or high fat diets, tumor implantation, or inflammation. While it is evident that Gpr182 is expressed in numerous vascularized organs during development and adulthood, relatively nothing is known about the actual function of

Gpr182 in these tissues.

Verification of Gpr182 Expression Pattern and its Cell Autonomous Roles as a

Gastrointestinal Stem Cell

The use of the Gpr182lacZ/+ knock-in mouse could potentially alter Gpr182 localization, especially considering Gpr182 negatively regulates MAPK signaling (Chapter

VI). Therefore, it is possible that the observed X-gal staining profile does not correlate with the endogenous Gpr182 expression pattern. Unfortunately, the vast majority of GPCR antibodies are not reliable, including Gpr182, making it difficult to verify endogenous Gpr182 protein localization. Without a reliable antibody, endogenous Gpr182 expression could be verified using ISH and RT-PCR in Gpr182+/+ mice. These assays are also needed to confirm expression pattern consistent in mice and humans. It is clear that endogenous

Gpr182 expression is highest in vascularized tissues, mainly in endothelial cells (Figure 6-1).

Epithelial expression in gastrointestinal tract is relatively low, which potentially explains why

Gpr182 is routinely overlooked on whole genome analyses of these tissues. The use of

Gpr182-independent mouse lines, such as the Sox9-EGFP and Lgr5-EGFP mice, helped confirmed that the ISC expression of Gpr182 was not secondary to Gpr182 reduction

(Figure 6-2). Similarly, while only trending Gpr182 expression was enriched in Lgr5-EGFP

ISCs when compared to Lrig1-EGFP ISCs.391 ISH was previously used to demonstrate

Gpr182 expression in e8.5 dorsal aorta and yolk sac, which further verified that

Gpr182lacZ/lacZ expression pattern was not artifact (Figure 6-1A).381 MACS isolation of CD31+ lung endothelial cells from wild-type mice supported that Gpr182 was mainly expressed in

166 the lung endothelium (Figure 6-1F and data not shown). Together these data support that the lacZ knock-in does not totally alter endogenous Gpr182 expression pattern, but it does not exclude potential differences. For example, the density of X-gal labelled crypts and secretory cells in the distal intestine appeared higher in Gpr182; CMV-Cre+ compared to

Gpr182/+; CMV-Cre or the Gpr182lacZ/lacZ animals (Figure 6-S2). Some tissues, such as breast, pancreas, brain, and mesentery had detectable Gpr182 expression by RT-PCR, but had negligible X-gal staining. Further characterization of the endogenous expression pattern of Gpr182 is needed to fully understand the in vivo functions of Gpr182 and whether genetic disruptions cause changes in Gpr182 localization.

Using both the lacZ reporter and expressional analysis from CBC-enriched populations from the Sox9-EGFP and Lgr5-EGFP mice; Gpr182 is enriched in CBC ISCs

(Figure 6-2). However, the “gold standard” in the ISC field to verify new stem cells markers is to demonstrate that Gpr182+ cells can self-renew and are multipotent,385, 386 which is not possible using the Gpr182lacZ/+ animal. This can be accomplished by generating the

Gpr182-EGFP-ires-CreERT mouse, which when crossed to the Rosa26-lacZ reporter and injected with TAM can be used to lineage trace Gpr182+ daughter cells. This experiment would verify Gpr182 as a ISC marker if all absorptive and secretory cells of the intestinal epithelium can be labeled (multipotent) and if the label at the base of the crypt is maintained overtime (self-renewal). While the ISC niche in the small intestine is relatively well anatomically defined, the stem cell zones of the gastric and colonic epithelium are far less defined. Therefore, lineage tracing would be required to determine if Gpr182 is enriched in stomach and colonic stem cells, as well as other stem cell populations throughout the body.

Gpr182 lineage tracing following IRR or during adenoma formation and progression would help determine the roles of this Gpr182 expression during these conditions and could shed new light on the mechanisms behind Gpr182 heterogeneity in adenomas (Figure 6-4, 6-6).

167

The inducible Gpr182-EGFP-ires-CreERT mouse would allow conditional deletion of genes of interests, such as Apc, to identify if alterations of in these genes specifically within Gpr182 expressing cells have phenotypic consequences, which has been done using the Lgr5 and

Lrig1 promoters.391, 398, 463 The incorporation of EGFP or a fluorescent reporter would allow the isolation of Gpr182-EGFP+ cells by FACS for RNA analysis and ex vivo organoid experiments. Isolated single cells would help confirm whether the phenotypes-associated with Gpr182 are cell autonomous. Alternatively, cell autonomy could be demonstrated using cell-type specific Cre animals, such as the Villin-CreERT or the Lgr5-EGFP-ires-CreERT2 mouse to conditionally delete Gpr182 in the intestinal epithelium or CBC ISCs, respectively.

The Gpr182lacZ/+ mouse is a conditional allele only after it is crossed to the C56BL/6-Rosa26-

FLPe (Jackson 003800) mouse to remove the En2 splice acceptor-mediated lacZ knock-in.

This strategy could also identify determine the functions of Gpr182 in the specific- differentiated cell types that express Gpr182. These studies will both confirm the expression pattern of Gpr182 in vivo, as well as help investigate the cell autonomous functions of

Gpr182.

Further Exploration of the Functions of Gpr182 during Gastrointestinal Homeostasis, Injury,

Regeneration, and Neoplasia

Chapter VI focused on the proliferative functions of Gpr182 in the small intestine.

The phenotypes were attributed to the loss of Gpr182, but it is possible that the proliferative phenotypes are caused by the upregulation of other receptors, signaling pathways, and/or transcription factors caused by altered Gpr182. It is important to characterize the number and distribution of CBC ISCs, activatable +4 ISCs, and differentiated cell types in the

Gpr182 null intestines. Additionally, the expression pattern of Wnt/Frizzled, DLL/Notch,

BMP/BMPR, and EGF/EGFR need to be characterized. The expression of signaling components could be mapped by RT-PCR, western, IHC, and/or genome-wide arrays. To

168 assess relative changes in the number of particular cell types, differentiated cells could be stained with Alcian Blue (goblet cells), chromogranin A (enteroendocrine cells), Lysozyme

(Paneth cells), Dclk1 (tuft cells), or Sucrase Isomaltase (enterocytes) and counted manually or by flow cytometry. These differentiated cells are all present in Gpr182lacZ/lacZ intestines and the relative expression of specific markers of these cells was unchanged (Figure 6-

S2D). To determine the number of CBC ISCs, Gpr182lacZ/+ could be crossed to the Lgr5-

EGFP mouse and the relative distribution of Lgr5-EGFPHigh cells between Gpr182 null and wild-type animals could be quantified by flow cytometry. Activatable reserve ISCs can be labelled by EdU followed by an extended washout period and label retaining cells can be quantified. In addition to proliferation and differentiation, it would be informative to investigate changes in apoptosis in Gpr182 null intestines.

Decreased Gpr182 increased proliferation five days post-IRR (Figure 6-4). It is worth investigating the expression and functions of Gpr182 throughout the damage and repair process following IRR, including the return to homeostasis following damage, of which the mechanism remains relatively unknown. Radiation-induced DNA damage and causes apoptotic death of actively cycling CBC ISCs. Gpr182 expressing cells are lost following

IRR and the Gpr182 promoter remains OFF during the regeneration phase (Figure 6-4D), which could be necessary in order to allow the physiologic hyperproliferation in order to maintain barrier function. It would be interesting to look 6 and 24 hours following IRR to see how Gpr182 loss affects initial cell death and DNA-damaged apoptosis, respectively.389, 408

Overall damage severity and initial regeneration could be assessed 2 or 3.5 days following

IRR, respectively.389, 404, 408, 447, 454 Beyond damage severity, it would also be informative to look at later time points, such as day 7 or day 9 to determine whether the IRR-induced hyperproliferation in the Gpr182 null animals returns to homeostasis or whether Gpr182 function is required for the downregulation of this proliferation. The temporal requirements

169 of Gpr182 during initial IRR injury and recovery to homeostasis, as well as whether Gpr182 is beneficial or detrimental to overall survival remains to be determined.

It has been well established that intestinal adenomas are highly mosaic.462, 463

Gpr182 promotor activity was also mosaic in polyps and appeared to be inversely correlated with proliferation, which requires further examination. If Gpr182 is a negative regulator of proliferation, than Gpr182-lacZ+ areas of polyps would be expected to have a competitive growth advantage over Gpr182-lacZ negative areas,461 but this does not appear to be the case. The Gpr182-lacZ negative areas represent a lack of promoter activity and are most likely not indicative of a failure of the En2 splice acceptor to remove the protein coding region. Therefore, the transcriptional regulation of Gpr182 is different in lacZ+ modest proliferation areas compared to high proliferation Gpr182-lacZ negative areas. While these

Gpr182-lacZ+ areas are part of adenoma as evident by nuclear -catenin staining, there are other differences in these areas. For example, Dclk1 has been shown to label tuft cells normally, but can act as a cancer initiating cell and a specific marker of cancer stem cells.441,

468 The authors reported that whole polyps labelled with Dclk1, similar to the Gpr182-lacZ negative areas, while the Gpr182-lacZ positive areas have single Dclk1+ cells that look like differentiated tuft cells (data not shown). This potentially indicates there are low levels of differentiation occurring in these Gpr182-lacZ+ areas. I hypothesize that Apc adenomas form in areas where the Gpr182 promoter and expression are downregulated. The Gpr182- lacZ+ areas in the Gpr182lacZ/lacZ animals likely represent areas where if Gpr182 was present in a Gpr182+/+ mouse, then these areas would not have developed into an adenoma. This could explain why more polyps form in the Gpr182lacZ/lacZ animals. Looking at Gpr182 levels at different time points during Apc polyp formation and progression would be informative, but would likely require ISH or an antibody to detect endogenous Gpr182 levels.

In addition to IRR and ApcMin adenoma challenge models, nothing is known regarding the roles of Gpr182 during intestinal inflammation. It would be informative to

170 challenge Gpr182 null mice with inflammatory models; including, high fat diet, acute and chronic colitis and ileitis induced by indomethacin or DSS, as well as different cancer models, such as DSS AOM or genetic crosses with activated Kras or PTEN mutant mice.

Furthermore, the generation of a Gpr182 overexpression transgenic mouse could help determine whether high Gpr182 is protective or if overexpression can rescue any of the phenotypes observed in the Gpr182lacZ/lacZ mice. It would be best if this mouse has a fluorescent reporter for in vivo detection and either a doxycycline-inducible Tet-ON/OFF switch or a floxed Stop codon in the promotor of the transgene to allow spatial and temporal control of Gpr182 activation. An inducible switch would allow the removal of Gpr182 to activate adenoma formation, followed by the reconstitution of Gpr182 in order to try to reverse polyp formation, similar to Apc.420 This would help determine whether Gpr182 is dispensable once an adenoma forms or if it is required for adenoma maintenance and progression.

Insights into Gpr182 Biochemistry and Mechanistic Signaling

Without an endogenous ligand, agonist, or antagonists, as well as not knowing anything regarding its G protein-coupling or -arrestin induced downstream signaling, it is impossible to fully tease apart the mechanism by which decreased Gpr182 caused a hyperproliferative microenvironment during regeneration and adenoma formation. While enhanced signaling through Erk1/2 in the Gpr182lacZ/lacZ mice (Figure 6-7A) is likely responsible for some of the phenotypes,416 the overall signaling and numerous compensatory feedback mechanisms in the intestinal epithelium are likely very complex.

Basally, the adaptable intestine is able to compensate for elevated MAPK signaling resulting in no overall change in proliferation, indicating that Gpr182 is fairly dispensable during homeostasis (Figure 6-3). This is consistent with other models with enhanced Erk1/2

171 phosphorylation that do not have basal proliferation phenotypes.414 It is important to show by IHC which intestinal cell types are responsible for the increased pErk1/2 observed by western blot. One could investigate other common signaling pathways important for maintaining intestinal proliferation, including Wnt, Notch, BMP, EGFR, EphB, p38, JNK,

PTEN, Akt, JAK/STAT3, and the Hippo/YAP pathways for a better understanding of the complexity of signaling. Currently, it is unclear why hyperproliferation was only observed during regeneration and increased adenoma initiation and not basally or within formed adenomas, especially considering the elevated MAPK signaling during all conditions

(Chapter VI). Looking at other signaling pathways, via western, phosphorylation arrays, or kinome profiling could help reveal the underlying compensatory signaling mechanisms.

Without a basal hyperproliferation phenotype, I hypothesize that the Wnt/-catenin signaling pathway is not directly affected by Gpr182 loss. Total levels of -catenin are relatively unchanged in Gpr182lacZ/lacZ jejunums (Figure 7-2A-B). Likewise the downstream

Wnt target, Lgr5 expression was unchanged in Gpr182; CMV-Cre+ jejunums (Figure 6-

S2D). Wnt and MAPK pathways do have common downstream signaling targets, including

Gsk-3 and c-myc. Erk1/2 activation has been shown to increase oncogenic c-myc stabilization leading to increased proliferation and adenoma formation and thus should be investigated in Gpr182lacZ/lacZ animals.416, 476 Preliminary results suggest that p-Gsk-3 is trending higher in Gpr182lacZ/lacZ adenomas (Figure 7-2C). Gsk-3 is downstream of MAPK,

Akt, and Wnt and when phosphorylated allows -catenin to move to the nucleus. Activated

STAT3 was unchanged with Gpr182 knockdown (Figure 7-2D), supporting that the

JAK/STAT pathway is likely not affected. Interestingly, activated Akt was trending down in

Gpr182lacZ/lacZ intestines basally and in adenomas, but not post-IRR (Figure 7-2E-F). The altered Akt pathway could potentially be responsible for the lack of proliferative phenotypes basally and within adenomas, but this needs to be further verified.

172

While I hypothesize that the Notch pathway is transcriptionally upstream of Gpr182, I have no reason to believe that Gpr182 loss disrupted Notch signaling. Secretory cell differentiation appears normal and the expression of the Notch pathway transcription factors, Olfm4 or Hes1, was also unchanged. As mentioned previously, NICD overexpression in the intestinal epithelium decreased Gpr182 expression (Figure 7-1B), but this regulation is likely more complex than direct transcriptional control. Mice treated with an inhibitor (DBZ) of -secretase, which is responsible the cleavage of Notch receptor during

Notch signaling and therefore prevents NICD translocation to the nucleus and NICD- regulated transcription (Figure 7-1C), caused a trending decrease in Gpr182 expression in isolated intestinal epithelium (Figure 7-1D). It has previously been shown that DBZ decreases CBC ISC numbers,452 which could be responsible for decreased Gpr182 expression. Likewise Notch overexpression decreases secretory cell numbers, which if

Gpr182 is expressed in differentiated tuft or enteroendocrine cell, this could be why there is reduced Gpr182 expression. Much more work is needed to determine how Gpr182 is transcriptionally regulated. If Notch is upstream of Gpr182 expression, then it would be interesting to challenge Gpr182 null animals with Notch inhibitors to determine whether

Gpr182 is involved in the reported secretory cell hyperplasia.452, 477, 478

Interestingly, a recent study by Gregorieff et al. demonstrated the functions of

YAP/Hippo signaling pathway during intestinal regeneration and adenoma, which had some striking similarities to the observed Gpr182 phenotypes.454 YAP signaling appears dispensable for homeostatic proliferation, but is critical for regeneration following IRR and during adenoma formation.455 More specifically, conditional YAP-/- mice had delayed regeneration following IRR and failed to form polyps when crossed to the Apc model.454 The

YAP dependent mechanism responsible for the regenerative and adenoma phenotypes was through the induction of EGFR signaling and activated Erk1/2 signaling. While I hypothesize

Gpr182 is Gi-coupled and inhibits cAMP generation leading to decreased Erk1/2 signaling, 173 an alternative hypothesis is that Gpr182 signals through Lats1/2 to block YAP activation, which has been shown with other GPCRs.467 With Gpr182 loss, YAP loses its brake, translocates to the nucleus, increases EGFR levels and signaling, and ultimately causes increased Erk1/2 activation and proliferation. Like YAP/Hippo signaling, Gpr182 is dispensable during homeostasis, but important in regenerative phase following damage and during adenoma formation. To further demonstrate this pathway’s involvement, YAP activation and subcellular localization could be assessed in Gpr182 null intestines under basal and challenged conditions. Alternatively YAP overexpression mice could be crossed to Gpr182 null animals to demonstrate an exacerbated YAP overexpression phenotype.

It is clear the signaling mechanism is likely very complex, but it would be informative to further tease apart the functional consequences of the Gpr182lacZ/lacZ enhanced Erk1/2 signaling. EGF is an important trophic factor required for organoid growth ex vivo.425

Therefore, it would be interesting to determine if Gpr182 null organoids can grow without

EGF supplementation. Gpr182 null mice and/or organoids could be challenged with EGFR or MEK inhibitors or Forskolin to test how they respond. It has been demonstrated that MEK inhibitors completely block adenoma formation in Apc mice or organoid growth,416, 454 and therefore it will be hard to tease apart the Gpr182 dependent roles in this signaling.

For gain-of-function studies, organoids could be generated from a Gpr182 overexpression mouse. Alternatively human colorectal carcinoma cell lines, which express low levels of GPR182 (Figure 6-8) could be transfected to overexpress GPR182. These gain-of-function cells would help determine whether increased GPR182 expression in tumor cells can inhibit growth, survival, and/or MAPK signaling. The high proliferation and low expression of GPR182 in these cell lines likely does not merit the shRNA knockdown of

GPR182. Cell line choice may be critical, since different mutations in different cancer cell lines could affect results. For example, DLD-1 colorectal cancer cells with an activating

G13D KRAS mutation were challenged with a genome-wide RNAi screen and the

174 combination of G13D RAS and GPR182 RNAi was trending towards synthetic lethality, although this was not further verified.479 Additionally, all the human pancreatic cancer cells lines assessed in Ramachandran et al. had RAS-activating mutations, so if Gpr182 knockdown is synthetic lethal with these activating mutations, it could explain why they observed decreased proliferation and tumor size due to decreased viability.102, 380 In addition to tumor cells, GPR182 could be knockdown in normal cells to determine if this leads to a proliferative or survival advantage. The main limitation to all of these ex vivo and in vitro experiments is that they will likely depend on Gpr182 ligand availability, and without a known ligand it will be difficult to attribute any effects directly to Gpr182 signaling.

As discussed in Chapter V, Gpr182 was once thought to be an AM receptor, but was shown not to bind nor signal with AM and was classified as an orphan GPCR.356, 363

Recently Klein et al. published that Cxcr7, another former AM receptor, could potentially bind AM as a “ligand sink” or decoy receptor.332, 359 While direct binding remains controversial, it is clear that AM can alter Cxcr7 transcriptional levels and vice versa, so they are somehow connected transcriptionally. Gpr182 could be investigated as an AM decoy receptor, like its paralog Cxcr7. While it seems clear that Gpr182 and Cxcr7 likely do not signal through Gs-coupled cAMP generation, downstream coupling to and signaling through Gq, Gi, or -arrestins needs to be further investigated. However, it is impossible to prove signaling and coupling without an endogenous ligand or synthetic compound to activate the receptor. Currently strategies to identify ligands, such as screening compound libraries and assessing changes in GPCR subcellular localization have been unsuccessful to identify ligands of Gpr182, at least publically. Mapping the subcellular localization of tagged

GPR182 in transfected cells would be informative to answer whether it is present at the cell surface or restricted in intracellular organelles. Gpr182 is grouped with chemokine receptors by phylogeny, which could provide insight into potential chemokine ligands as candidates,

175 but these Class A GPCRs are typically promiscuous.480 Gpr182 could require an interactions partner, such as RAMPs or LRPs, to translocate it to the membrane or facilitate signal, but current investigation in our lab indicate that Gpr182 does not interact with the known RAMPs. Gpr182 appears to be an inhibitor of MAPK signaling and proliferation and therefore these could be the readouts to screen compound libraries. Ex vivo model systems, such as the organoids, could be useful models to test ligands and chemical compounds to look for altered MAPK signaling and growth efficiency. Discovering the pharmacology behind Gpr182 signaling would be a significant advancement to our current understanding of Gpr182.

Concluding Remarks

Part II of this dissertation focused on the in vivo expression and function of murine

Gpr182, an orphan GPCR of which almost nothing published. Considering the therapeutic success of targeting GPCRs to treat a large variety of human diseases, it is crucial to interrogate the function all GPCRs. While this study is an initial step in understanding the expression and function of Gpr182 in vivo, the findings regarding its spatial and temporal expression pattern, the identification that Gpr182 is enriched in ISCs, and that Gpr182 is involved in regulating proliferation will spark the interest of biologists and pharmacologists to build on this study and identify the ligand and therapeutic potential of this “forgotten” GPCR.

Together, the work presented in this dissertation identified novel functions of two underappreciated GPCR signaling pathways in both cardiovascular and neoplastic diseases.

176

Figures

Figure 7-1: Notch signaling is potentially an upstream regulator of Gpr182 expression. (A) Method of inducible intestinal epithelial-specific genetic overexpression of Notch intracellular domain (NICD). (B) Intestinal epithelial-specific Gpr182 expression in control and NICD overexpression mice. (C) Method to pharmacological inhibition of Notch signaling using -secretase inhibitor DBZ. (D) Intestinal epithelial-specific Gpr182 expression in vehicle and DBZ treated mice. Samples normalized to controls and Gapdh. N = 4. **p<0.01. RNA was kindly provided by Dr. Linda C. Samuelson at University of Michigan.

177

Figure 7-2: Reduced Gpr182 decreased Akt signaling in normal intestines. (A) Representative immunoblots and (B) quantitation of total -catenin in jejunums from Gpr182+/+ and Gpr182lacZ/lacZ mice basally, 5 days post-IRR, and in polyps from ApcMin mice. (C) Representative pGsk-3 from ApcMin polyps and (D) pStat3 and total Stat3 levels from unchallenged Gpr182+/+ and Gpr182lacZ/lacZ mice. (E) Representative immunoblots and (F) quantitation of pAkt normalized to total Akt protein in jejunums from Gpr182+/+ and Gpr182lacZ/lacZ mice basally, 5 days post-IRR, and in polyps from ApcMin mice. Levels were normalized to Gpr182+/+ and Gapdh (B) or Akt (F). N = 3-5 (A, E) and N =1-2 (C, D). Significance determined by Student T-test with **p<0.01.

178

REFERENCES

1. Gibbons C, Dackor R, Dunworth W, Fritz-Six K, Caron KM. Receptor activity- modifying proteins: RAMPing up adrenomedullin signaling. Mol Endocrinol. 2007;21:783-796.

2. Morgenthaler NG, Struck J, Alonso C, Bergmann A. Measurement of midregional proadrenomedullin in plasma with an immunoluminometric assay. Clin Chem. 2005;51:1823-1829.

3. Melander O, Newton-Cheh C, Almgren P, Hedblad B, Berglund G, Engstrom G, Persson M, Smith JG, Magnusson M, Christensson A, et al. Novel and conventional biomarkers for prediction of incident cardiovascular events in the community. JAMA. 2009;302:49-57.

4. Khan SQ, O'Brien RJ, Struck J, Quinn P, Morgenthaler N, Squire I, Davies J, Bergmann A, Ng LL. Prognostic value of midregional pro-adrenomedullin in patients with acute myocardial infarction: the LAMP (Leicester Acute Myocardial Infarction Peptide) study. J Am Coll Cardiol. 2007;49:1525-1532.

5. Christ-Crain M, Morgenthaler NG, Struck J, Harbarth S, Bergmann A, Muller B. Mid- regional pro-adrenomedullin as a prognostic marker in sepsis: an observational study. Crit Care. 2005;9:R816-824.

6. Christ-Crain M, Morgenthaler NG, Stolz D, Muller C, Bingisser R, Harbarth S, Tamm M, Struck J, Bergmann A, Muller B. Pro-adrenomedullin to predict severity and outcome in community-acquired pneumonia [ISRCTN04176397]. Crit Care. 2006;10:R96.

7. McLatchie LM, Fraser NJ, Main MJ, Wise A, Brown J, Thompson N, Solari R, Lee MG, Foord SM. RAMPs regulate the transport and ligand specificity of the calcitonin- receptor-like receptor. Nature. 1998;393:333-339.

8. Nagaya N, Mori H, Murakami S, Kangawa K, Kitamura S. Adrenomedullin: angiogenesis and gene therapy. Am J Physiol Regul Integr Comp Physiol. 2005;288:R1432-1437.

9. Ishimitsu T, Ono H, Minami J, Matsuoka H. Pathophysiologic and therapeutic implications of adrenomedullin in cardiovascular disorders. Pharmacol Ther. 2006;111:909-927.

10. Julian M, Cacho M, Garcia MA, Martin-Santamaria S, de Pascual-Teresa B, Ramos A, Martinez A, Cuttitta F. Adrenomedullin: a new target for the design of small

179

molecule modulators with promising pharmacological activities. Eur J Med Chem. 2005;40:737-750.

11. Jin D, Harada K, Ohnishi S, Yamahara K, Kangawa K, Nagaya N. Adrenomedullin induces lymphangiogenesis and ameliorates secondary lymphoedema. Cardiovasc Res. 2008;80:339-345.

12. Miyashita K, Itoh H, Sawada N, Fukunaga Y, Sone M, Yamahara K, Yurugi- Kobayashi T, Park K, Nakao K. Adrenomedullin provokes endothelial Akt activation and promotes vascular regeneration both in vitro and in vivo. FEBS Lett. 2003;544:86-92.

13. Kim W, Moon SO, Sung MJ, Kim SH, Lee S, So JN, Park SK. Angiogenic role of adrenomedullin through activation of Akt, mitogen-activated protein kinase, and focal adhesion kinase in endothelial cells. FASEB J. 2003;17:1937-1939.

14. Fernandez-Sauze S, Delfino C, Mabrouk K, Dussert C, Chinot O, Martin PM, Grisoli F, Ouafik L, Boudouresque F. Effects of adrenomedullin on endothelial cells in the multistep process of angiogenesis: involvement of CRLR/RAMP2 and CRLR/RAMP3 receptors. Int J Cancer. 2004;108:797-804.

15. Kato H, Shichiri M, Marumo F, Hirata Y. Adrenomedullin as an autocrine/paracrine apoptosis survival factor for rat endothelial cells. Endocrinology. 1997;138:2615- 2620.

16. Zhou M, Simms HH, Wang P. Adrenomedullin and adrenomedullin binding protein-1 attenuate vascular endothelial cell apoptosis in sepsis. Ann Surg. 2004;240:321-330.

17. Miyashita K, Itoh H, Sawada N, Fukunaga Y, Sone M, Yamahara K, Yurugi T, Nakao K. Adrenomedullin promotes proliferation and migration of cultured endothelial cells. Hypertens Res. 2003;26 Suppl:S93-98.

18. Ichikawa-Shindo Y, Sakurai T, Kamiyoshi A, Kawate H, Iinuma N, Yoshizawa T, Koyama T, Fukuchi J, Iimuro S, Moriyama N, et al. The GPCR modulator protein RAMP2 is essential for angiogenesis and vascular integrity. J Clin Invest. 2008;118:29-39.

19. Hippenstiel S, Witzenrath M, Schmeck B, Hocke A, Krisp M, Krull M, Seybold J, Seeger W, Rascher W, Schutte H, et al. Adrenomedullin reduces endothelial hyperpermeability. Circ Res. 2002;91:618-625.

180

20. Kis B, Deli MA, Kobayashi H, Abraham CS, Yanagita T, Kaiya H, Isse T, Nishi R, Gotoh S, Kangawa K, et al. Adrenomedullin regulates blood-brain barrier functions in vitro. Neuroreport. 2001;12:4139-4142.

21. Muller HC, Witzenrath M, Tschernig T, Gutbier B, Hippenstiel S, Santel A, Suttorp N, Rosseau S. Adrenomedullin attenuates ventilator-induced lung injury in mice. Thorax. 2010;65:1077-1084.

22. Fritz-Six KL, Dunworth WP, Li M, Caron KM. Adrenomedullin signaling is necessary for murine lymphatic vascular development. J Clin Invest. 2008;118:40-50.

23. Petrova TV, Makinen T, Makela TP, Saarela J, Virtanen I, Ferrell RE, Finegold DN, Kerjaschki D, Yla-Herttuala S, Alitalo K. Lymphatic endothelial reprogramming of vascular endothelial cells by the Prox-1 homeobox transcription factor. EMBO J. 2002;21:4593-4599.

24. Hirakawa S, Hong YK, Harvey N, Schacht V, Matsuda K, Libermann T, Detmar M. Identification of vascular lineage-specific genes by transcriptional profiling of isolated blood vascular and lymphatic endothelial cells. Am J Pathol. 2003;162:575-586.

25. Dunworth WP, Fritz-Six KL, Caron KM. Adrenomedullin stabilizes the lymphatic endothelial barrier in vitro and in vivo. Peptides. 2008;29:2243-2249.

26. Marinissen MJ, Gutkind JS. G-protein-coupled receptors and signaling networks: emerging paradigms. Trends Pharmacol Sci. 2001;22:368-376.

27. Lowes VL, Ip NY, Wong YH. Integration of signals from receptor tyrosine kinases and g protein-coupled receptors. Neurosignals. 2002;11:5-19.

28. Yurugi-Kobayashi T, Itoh H, Schroeder T, Nakano A, Narazaki G, Kita F, Yanagi K, Hiraoka-Kanie M, Inoue E, Ara T, et al. Adrenomedullin/cyclic AMP pathway induces Notch activation and differentiation of arterial endothelial cells from vascular progenitors. Arterioscler Thromb Vasc Biol. 2006;26:1977-1984.

29. Guidolin D, Albertin G, Spinazzi R, Sorato E, Mascarin A, Cavallo D, Antonello M, Ribatti D. Adrenomedullin stimulates angiogenic response in cultured human vascular endothelial cells: Involvement of the vascular endothelial growth factor receptor 2. Peptides. 2008;29:2013-2023.

30. Caron KM, Smithies O. Extreme hydrops fetalis and cardiovascular abnormalities in mice lacking a functional Adrenomedullin gene. Proc Natl Acad Sci U S A. 2001;98:615-619.

181

31. Shindo T, Kurihara Y, Nishimatsu H, Moriyama N, Kakoki M, Wang Y, Imai Y, Ebihara A, Kuwaki T, Ju KH, et al. Vascular abnormalities and elevated blood pressure in mice lacking adrenomedullin gene. Circulation. 2001;104:1964-1971.

32. Shimosawa T, Shibagaki Y, Ishibashi K, Kitamura K, Kangawa K, Kato S, Ando K, Fujita T. Adrenomedullin, an endogenous peptide, counteracts cardiovascular damage. Circulation. 2002;105:106-111.

33. Dackor RT, Fritz-Six K, Dunworth WP, Gibbons CL, Smithies O, Caron KM. Hydrops fetalis, cardiovascular defects, and embryonic lethality in mice lacking the calcitonin receptor-like receptor gene. Mol Cell Biol. 2006;26:2511-2518.

34. Dackor R, Fritz-Six K, Smithies O, Caron K. Receptor activity-modifying proteins 2 and 3 have distinct physiological functions from embryogenesis to old age. J Biol Chem. 2007;282:18094-18099.

35. Czyzyk TA, Ning Y, Hsu MS, Peng B, Mains RE, Eipper BA, Pintar JE. Deletion of peptide amidation enzymatic activity leads to edema and embryonic lethality in the mouse. Dev Biol. 2005;287:301-313.

36. Yang JT, Rayburn H, Hynes RO. Cell adhesion events mediated by alpha 4 integrins are essential in placental and cardiac development. Development. 1995;121:549- 560.

37. Riley P, Anson-Cartwright L, Cross JC. The Hand1 bHLH transcription factor is essential for placentation and cardiac morphogenesis. Nat Genet. 1998;18:271-275.

38. Jaber M, Koch WJ, Rockman H, Smith B, Bond RA, Sulik KK, Ross J, Jr., Lefkowitz RJ, Caron MG, Giros B. Essential role of beta-adrenergic receptor kinase 1 in cardiac development and function. Proc Natl Acad Sci U S A. 1996;93:12974-12979.

39. Puri MC, Rossant J, Alitalo K, Bernstein A, Partanen J. The receptor tyrosine kinase TIE is required for integrity and survival of vascular endothelial cells. EMBO J. 1995;14:5884-5891.

40. Suri C, Jones PF, Patan S, Bartunkova S, Maisonpierre PC, Davis S, Sato TN, Yancopoulos GD. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell. 1996;87:1171-1180.

41. Kim K, Drummond I, Ibraghimov-Beskrovnaya O, Klinger K, Arnaout MA. Polycystin 1 is required for the structural integrity of blood vessels. Proc Natl Acad Sci U S A. 2000;97:1731-1736.

182

42. Wigle JT, Oliver G. Prox1 function is required for the development of the murine lymphatic system. Cell. 1999;98:769-778.

43. Francois M, Caprini A, Hosking B, Orsenigo F, Wilhelm D, Browne C, Paavonen K, Karnezis T, Shayan R, Downes M, et al. Sox18 induces development of the lymphatic vasculature in mice. Nature. 2008;456:643-647.

44. Karkkainen MJ, Haiko P, Sainio K, Partanen J, Taipale J, Petrova TV, Jeltsch M, Jackson DG, Talikka M, Rauvala H, et al. Vascular endothelial growth factor C is required for sprouting of the first lymphatic vessels from embryonic veins. Nat Immunol. 2004;5:74-80.

45. Kahn ML. Blood is thicker than lymph. J Clin Invest. 2008;118:23-26.

46. Shindo T, Kurihara H, Maemura K, Kurihara Y, Kuwaki T, Izumida T, Minamino N, Ju KH, Morita H, Oh-hashi Y, et al. Hypotension and resistance to lipopolysaccharide- induced shock in transgenic mice overexpressing adrenomedullin in their vasculature. Circulation. 2000;101:2309-2316.

47. Tam CW, Husmann K, Clark NC, Clark JE, Lazar Z, Ittner LM, Gotz J, Douglas G, Grant AD, Sugden D, et al. Enhanced vascular responses to adrenomedullin in mice overexpressing receptor-activity-modifying protein 2. Circ Res. 2006;98:262-270.

48. Kitamura K, Ichiki Y, Tanaka M, Kawamoto M, Emura J, Sakakibara S, Kangawa K, Matsuo H, Eto T. Immunoreactive adrenomedullin in human plasma. FEBS Lett. 1994;341:288-290.

49. Martinez A, Hodge DL, Garayoa M, Young HA, Cuttitta F. Alternative splicing of the proadrenomedullin gene results in differential expression of gene products. J Mol Endocrinol. 2001;27:31-41.

50. Nakamura M, Yoshida H, Hiramori K. Comparison of vasodilator potency of adrenomedulling and proadrenomedullin N-terminal 20 peptide in human. Life Sci. 1999;65:2151-2156.

51. Li J, Ren Y, Dong X, Zhong G, Wu S, Tang C. Roles of different peptide fragments derived from proadrenomedullin in the regulation of vascular tone in isolated rat aorta. Peptides. 2003;24:563-568.

52. Hay DL, Smith DM. Knockouts and transgenics confirm the importance of adrenomedullin in the vasculature. Trends Pharmacol Sci. 2001;22:57-59.

183

53. Bomberger JM, Parameswaran N, Hall CS, Aiyar N, Spielman WS. Novel function for receptor activity-modifying proteins (RAMPs) in post-endocytic receptor trafficking. J Biol Chem. 2005;280:9297-9307.

54. Bomberger JM, Spielman WS, Hall CS, Weinman EJ, Parameswaran N. Receptor activity-modifying protein (RAMP) isoform-specific regulation of adrenomedullin receptor trafficking by NHERF-1. J Biol Chem. 2005;280:23926-23935.

55. Nicoli S, Tobia C, Gualandi L, De Sena G, Presta M. Calcitonin receptor-like receptor guides arterial differentiation in zebrafish. Blood. 2008;111:4965-4972.

56. Iimuro S, Shindo T, Moriyama N, Amaki T, Niu P, Takeda N, Iwata H, Zhang Y, Ebihara A, Nagai R. Angiogenic effects of adrenomedullin in ischemia and tumor growth. Circ Res. 2004;95:415-423.

57. Yaniv K, Isogai S, Castranova D, Dye L, Hitomi J, Weinstein BM. Live imaging of lymphatic development in the zebrafish. Nat Med. 2006;12:711-716.

58. Isogai S, Hitomi J, Yaniv K, Weinstein BM. Zebrafish as a new animal model to study lymphangiogenesis. Anat Sci Int. 2009;84:102-111.

59. Geudens I, Herpers R, Hermans K, Segura I, Ruiz de Almodovar C, Bussmann J, De Smet F, Vandevelde W, Hogan BM, Siekmann A, et al. Role of delta-like-4/Notch in the formation and wiring of the lymphatic network in zebrafish. Arterioscler Thromb Vasc Biol. 2010;30:1695-1702.

60. Di Iorio R, Marinoni E, Scavo D, Letizia C, Cosmi EV. Adrenomedullin in pregnancy. Lancet. 1997;349:328.

61. Minegishi T, Nakamura M, Abe K, Tano M, Andoh A, Yoshida M, Takagi T, Nishikimi T, Kojima M, Kangawa K. Adrenomedullin and atrial natriuretic peptide concentrations in normal pregnancy and pre-eclampsia. Mol Hum Reprod. 1999;5:767-770.

62. Kobayashi K, Kubota T, Aso T, Hirata Y, Imai T, Marumo F. Immunoreactive adrenomedullin (AM) concentration in maternal plasma during human pregnancy and AM expression in placenta. Eur J Endocrinol. 2000;142:683-687.

63. Hoshimoto K, Hayashi M, Ohkura T. Mature adrenomedullin concentrations in plasma during pregnancy. J Matern Fetal Neonatal Med. 2002;11:126-129.

184

64. Hayashi Y, Ueyama H, Mashimo T, Kangawa K, Minamino N. Circulating mature adrenomedullin is related to blood volume in full-term pregnancy. Anesth Analg. 2005;101:1816-1820.

65. Senna AA, Zedan M, el-Salam GE, el-Mashad AI. Study of plasma adrenomedullin level in normal pregnancy and preclampsia. Medscape J Med. 2008;10:29.

66. Marinoni E, Casciani V, Marianetti V, Di Rocco A, Moscarini M, Di Iorio R. Localization and distribution of adrenomedullin receptor in the human placenta: changes with gestational age. J Reprod Med. 2007;52:831-838.

67. Cameron VA, Autelitano DJ, Evans JJ, Ellmers LJ, Espiner EA, Nicholls MG, Richards AM. Adrenomedullin expression in rat uterus is correlated with plasma estradiol. Am J Physiol Endocrinol Metab. 2002;282:E139-146.

68. Li M, Yee D, Magnuson TR, Smithies O, Caron KM. Reduced maternal expression of adrenomedullin disrupts fertility, placentation, and fetal growth in mice. J Clin Invest. 2006;116:2653-2662.

69. Nikitenko LL, Brown NS, Smith DM, MacKenzie IZ, Bicknell R, Rees MC. Differential and cell-specific expression of calcitonin receptor-like receptor and receptor activity modifying proteins in the human uterus. Mol Hum Reprod. 2001;7:655-664.

70. Gratton RJ, Gluszynski M, Mazzuca DM, Nygard K, Han VK. Adrenomedullin messenger ribonucleic acid expression in the placentae of normal and preeclamptic pregnancies. J Clin Endocrinol Metab. 2003;88:6048-6055.

71. Marinoni E, Di Iorio R, Letizia C, Villaccio B, Scucchi L, Cosmi EV. Immunoreactive adrenomedullin in human fetoplacental tissues. Am J Obstet Gynecol. 1998;179:784- 787.

72. Montuenga LM, Martinez A, Miller MJ, Unsworth EJ, Cuttitta F. Expression of adrenomedullin and its receptor during embryogenesis suggests autocrine or paracrine modes of action. Endocrinology. 1997;138:440-451.

73. Yotsumoto S, Shimada T, Cui CY, Nakashima H, Fujiwara H, Ko MS. Expression of adrenomedullin, a hypotensive peptide, in the trophoblast giant cells at the embryo implantation site in mouse. Dev Biol. 1998;203:264-275.

74. Tsatsaris V, Tarrade A, Merviel P, Garel JM, Segond N, Jullienne A, Evain-Brion D. Calcitonin gene-related peptide (CGRP) and CGRP receptor expression at the human implantation site. J Clin Endocrinol Metab. 2002;87:4383-4390.

185

75. Hinson JP, Kapas S, Smith DM. Adrenomedullin, a multifunctional regulatory peptide. Endocr Rev. 2000;21:138-167.

76. Kitamura K, Kangawa K, Kawamoto M, Ichiki Y, Nakamura S, Matsuo H, Eto T. Adrenomedullin: a novel hypotensive peptide isolated from human pheochromocytoma. Biochem Biophys Res Commun. 1993;192:553-560.

77. Eto T, Kitamura K, Kato J. Biological and clinical roles of adrenomedullin in circulation control and cardiovascular diseases. Clin Exp Pharmacol Physiol. 1999;26:371-380.

78. Jougasaki M, Burnett JC, Jr. Adrenomedullin: potential in physiology and pathophysiology. Life Sci. 2000;66:855-872.

79. Caron K, Hagaman J, Nishikimi T, Kim HS, Smithies O. Adrenomedullin gene expression differences in mice do not affect blood pressure but modulate hypertension-induced pathology in males. Proc Natl Acad Sci U S A. 2007;104:3420- 3425.

80. Kuwasako K, Cao YN, Nagoshi Y, Kitamura K, Eto T. Adrenomedullin receptors: pharmacological features and possible pathophysiological roles. Peptides. 2004;25:2003-2012.

81. Nishikimi T, Yoshihara F, Mori Y, Kangawa K, Matsuoka H. Cardioprotective effect of adrenomedullin in heart failure. Hypertens Res. 2003;26 Suppl:S121-127.

82. Yanagawa B, Nagaya N. Adrenomedullin: molecular mechanisms and its role in cardiac disease. Amino Acids. 2006

83. Endemann DH, Schiffrin EL. Endothelial dysfunction. J Am Soc Nephrol. 2004;15:1983-1992.

84. Davignon J, Ganz P. Role of endothelial dysfunction in atherosclerosis. Circulation. 2004;109:III27-32.

85. Pan CS, Zhang J, Yu F, Teng X, Cao CQ, Wu W, Tang CS, Qi YF. Adrenomedullin ameliorates the development of atherosclerosis in apoE-/- mice. Peptides. 2010;31:1150-1158.

86. Zhang SH, Reddick RL, Piedrahita JA, Maeda N. Spontaneous hypercholesterolemia and arterial lesions in mice lacking apolipoprotein E. Science. 1992;258:468-471.

186

87. Jin D, Otani K, Yamahara K, Ikeda T, Nagaya N, Kangawa K. Adrenomedullin reduces expression of adhesion molecules on lymphatic endothelial cells. Regul Pept. 2011;166:21-27.

88. Iinuma N, Sakurai T, Kamiyoshi A, Ichikawa-Shindo Y, Arai T, Yoshizawa T, Koyama T, Uetake R, Kawate H, Muto S, et al. Adrenomedullin in sinusoidal endothelial cells play protective roles against cold injury of liver. Peptides. 2010;31:865-871.

89. Kim W, Moon SO, Lee S, Sung MJ, Kim SH, Park SK. Adrenomedullin reduces VEGF-induced endothelial adhesion molecules and adhesiveness through a phosphatidylinositol 3'-kinase pathway. Arterioscler Thromb Vasc Biol. 2003;23:1377-1383.

90. Nossaman BD, Feng CJ, Kaye AD, DeWitt B, Coy DH, Murphy WA, Kadowitz PJ. Pulmonary vasodilator responses to adrenomedullin are reduced by NOS inhibitors in rats but not in cats. Am J Physiol. 1996;270:L782-789.

91. Nakamura M, Yoshida H, Makita S, Arakawa N, Niinuma H, Hiramori K. Potent and long-lasting vasodilatory effects of adrenomedullin in humans. Comparisons between normal subjects and patients with chronic heart failure. Circulation. 1997;95:1214- 1221.

92. Nishimatsu H, Suzuki E, Nagata D, Moriyama N, Satonaka H, Walsh K, Sata M, Kangawa K, Matsuo H, Goto A, et al. Adrenomedullin induces endothelium- dependent vasorelaxation via the phosphatidylinositol 3-kinase/Akt-dependent pathway in rat aorta. Circ Res. 2001;89:63-70.

93. Takahashi Y, de Vroomen M, Gournay V, Roman C, Rudolph AM, Heymann MA. Mechanisms of adrenomedullin-induced increase of pulmonary blood flow in fetal sheep. Pediatr Res. 1999;45:276-281.

94. De Matteo R, May CN. Direct coronary vasodilator action of adrenomedullin is mediated by nitric oxide. Br J Pharmacol. 2003;140:1414-1420.

95. Zhou M, Maitra SR, Wang P. Adrenomedullin and adrenomedullin binding protein-1 protect endothelium-dependent vascular relaxation in sepsis. Mol Med. 2007;13:488- 494.

96. Maki T, Ihara M, Fujita Y, Nambu T, Miyashita K, Yamada M, Washida K, Nishio K, Ito H, Harada H, et al. Angiogenic and vasoprotective effects of adrenomedullin on prevention of cognitive decline after chronic cerebral hypoperfusion in mice. Stroke. 2011;42:1122-1128.

187

97. Harada K, Yamahara K, Ohnishi S, Otani K, Kanoh H, Ishibashi-Ueda H, Minamino N, Kangawa K, Nagaya N, Ikeda T. Sustained-release adrenomedullin ointment accelerates wound healing of pressure ulcers. Regul Pept. 2011;168:21-26.

98. Martinez A, Miller MJ, Unsworth EJ, Siegfried JM, Cuttitta F. Expression of adrenomedullin in normal human lung and in pulmonary tumors. Endocrinology. 1995;136:4099-4105.

99. Satoh F, Takahashi K, Murakami O, Totsune K, Sone M, Ohneda M, Abe K, Miura Y, Hayashi Y, Sasano H, et al. Adrenomedullin in human brain, adrenal glands and tumor tissues of pheochromocytoma, ganglioneuroblastoma and neuroblastoma. J Clin Endocrinol Metab. 1995;80:1750-1752.

100. Miller MJ, Martinez A, Unsworth EJ, Thiele CJ, Moody TW, Elsasser T, Cuttitta F. Adrenomedullin expression in human tumor cell lines. Its potential role as an autocrine growth factor. J Biol Chem. 1996;271:23345-23351.

101. Metellus P, Voutsinos-Porche B, Nanni-Metellus I, Colin C, Fina F, Berenguer C, Dussault N, Boudouresque F, Loundou A, Intagliata D, et al. Adrenomedullin expression and regulation in human glioblastoma, cultured human glioblastoma cell lines and pilocytic astrocytoma. Eur J Cancer. 2011;47:1727-1735.

102. Ramachandran V, Arumugam T, Hwang RF, Greenson JK, Simeone DM, Logsdon CD. Adrenomedullin is expressed in pancreatic cancer and stimulates cell proliferation and invasion in an autocrine manner via the adrenomedullin receptor, ADMR. Cancer Res. 2007;67:2666-2675.

103. Martinez A, Vos M, Guedez L, Kaur G, Chen Z, Garayoa M, Pio R, Moody T, Stetler- Stevenson WG, Kleinman HK, et al. The effects of adrenomedullin overexpression in breast tumor cells. J Natl Cancer Inst. 2002;94:1226-1237.

104. Keleg S, Kayed H, Jiang X, Penzel R, Giese T, Buchler MW, Friess H, Kleeff J. Adrenomedullin is induced by hypoxia and enhances pancreatic cancer cell invasion. Int J Cancer. 2007;121:21-32.

105. Zudaire E, Martinez A, Garayoa M, Pio R, Kaur G, Woolhiser MR, Metcalfe DD, Hook WA, Siraganian RP, Guise TA, et al. Adrenomedullin is a cross-talk molecule that regulates tumor and mast cell function during human carcinogenesis. Am J Pathol. 2006;168:280-291.

106. Hata K, Takebayashi Y, Akiba S, Fujiwaki R, Iida K, Nakayama K, Nakayama S, Fukumoto M, Miyazaki K. Expression of the adrenomedullin gene in epithelial ovarian cancer. Mol Hum Reprod. 2000;6:867-872.

188

107. Cuttitta F, Pio R, Garayoa M, Zudaire E, Julian M, Elsasser TH, Montuenga LM, Martinez A. Adrenomedullin functions as an important tumor survival factor in human carcinogenesis. Microsc Res Tech. 2002;57:110-119.

108. Ribatti D, Nico B, Spinazzi R, Vacca A, Nussdorfer GG. The role of adrenomedullin in angiogenesis. Peptides. 2005;26:1670-1675.

109. Garayoa M, Martinez A, Lee S, Pio R, An WG, Neckers L, Trepel J, Montuenga LM, Ryan H, Johnson R, et al. Hypoxia-inducible factor-1 (HIF-1) up-regulates adrenomedullin expression in human tumor cell lines during oxygen deprivation: a possible promotion mechanism of carcinogenesis. Mol Endocrinol. 2000;14:848-862.

110. Nikitenko LL, Smith DM, Bicknell R, Rees MC. Transcriptional regulation of the CRLR gene in human microvascular endothelial cells by hypoxia. FASEB J. 2003;17:1499-1501.

111. Khodarev NN, Yu J, Labay E, Darga T, Brown CK, Mauceri HJ, Yassari R, Gupta N, Weichselbaum RR. Tumour-endothelium interactions in co-culture: coordinated changes of gene expression profiles and phenotypic properties of endothelial cells. J Cell Sci. 2003;116:1013-1022.

112. Ouafik L, Sauze S, Boudouresque F, Chinot O, Delfino C, Fina F, Vuaroqueaux V, Dussert C, Palmari J, Dufour H, et al. Neutralization of adrenomedullin inhibits the growth of human glioblastoma cell lines in vitro and suppresses tumor xenograft growth in vivo. Am J Pathol. 2002;160:1279-1292.

113. Kaafarani I, Fernandez-Sauze S, Berenguer C, Chinot O, Delfino C, Dussert C, Metellus P, Boudouresque F, Mabrouk K, Grisoli F, et al. Targeting adrenomedullin receptors with systemic delivery of neutralizing antibodies inhibits tumor angiogenesis and suppresses growth of human tumor xenografts in mice. FASEB J. 2009;23:3424-3435.

114. Fernandez AP, Serrano J, Tessarollo L, Cuttitta F, Martinez A. Lack of adrenomedullin in the mouse brain results in behavioral changes, anxiety, and lower survival under stress conditions. Proc Natl Acad Sci U S A. 2008;105:12581-12586.

115. Ehlenz K, Koch B, Preuss P, Simon B, Koop I, Lang RE. High levels of circulating adrenomedullin in severe illness: correlation with C-reactive protein and evidence against the adrenal medulla as site of origin. Exp Clin Endocrinol Diabetes. 1997;105:156-162.

189

116. Letizia C, Tamburrano G, Alo P, Paoloni A, Caliumi C, Marinoni E, di Iorio R, d'Erasmo E. Adrenomedullin, a new peptide, in patients with insulinoma. Eur J Endocrinol. 2001;144:517-520.

117. Letizia C, De Toma G, Caliumi C, Cerci S, Massa R, Loria RD, Alo P, Marinoni EM, Diacinti D, D'Erasmo E. Plasma adrenomedullin concentrations in patients with adrenal pheochromocytoma. Horm Metab Res. 2001;33:290-294.

118. Oehler MK, Fischer DC, Orlowska-Volk M, Herrle F, Kieback DG, Rees MC, Bicknell R. Tissue and plasma expression of the angiogenic peptide adrenomedullin in breast cancer. Br J Cancer. 2003;89:1927-1933.

119. Cotesta D, Caliumi C, Alo P, Petramala L, Reale MG, Masciangelo R, Signore A, Cianci R, D'Erasmo E, Letizia C. High plasma levels of human chromogranin A and adrenomedullin in patients with pheochromocytoma. Tumori. 2005;91:53-58.

120. Letizia C, Di Iorio R, De Toma G, Marinoni E, Cerci S, Celi M, Subioli S, D'Erasmo E. Circulating adrenomedullin is increased in patients with corticotropin-dependent Cushing's syndrome due to pituitary adenoma. Metabolism. 2000;49:760-763.

121. Miyao Y, Nishikimi T, Goto Y, Miyazaki S, Daikoku S, Morii I, Matsumoto T, Takishita S, Miyata A, Matsuo H, et al. Increased plasma adrenomedullin levels in patients with acute myocardial infarction in proportion to the clinical severity. Heart. 1998;79:39-44.

122. Nagaya N, Nishikimi T, Uematsu M, Yoshitomi Y, Miyao Y, Miyazaki S, Goto Y, Kojima S, Kuramochi M, Matsuo H, et al. Plasma adrenomedullin as an indicator of prognosis after acute myocardial infarction. Heart. 1999;81:483-487.

123. Yoshitomi Y, Nishikimi T, Kojima S, Kuramochi M, Takishita S, Matsuoka H, Miyata A, Matsuo H, Kangawa K. Plasma levels of adrenomedullin in patients with acute myocardial infarction. Clin Sci (Lond). 1998;94:135-139.

124. Balat A, Cekmen M, Yurekli M, Kutlu O, Islek I, Sonmezgoz E, Cakir M, Turkoz Y, Yologlu S. Adrenomedullin and nitrite levels in children with Bartter syndrome. Pediatr Nephrol. 2000;15:266-270.

125. Evereklioglu C, Yurekli M, Er H, Ozbek E, Hazneci E, Cekmen M, Inaloz HS. Increased plasma adrenomedullin levels in patients with Behcet's disease. Dermatology. 2000;201:312-315.

190

126. Ueda S, Nishio K, Minamino N, Kubo A, Akai Y, Kangawa K, Matsuo H, Fujimura Y, Yoshioka A, Masui K, et al. Increased plasma levels of adrenomedullin in patients with systemic inflammatory response syndrome. Am J Respir Crit Care Med. 1999;160:132-136.

127. Yoshibayashi M, Kamiya T, Nishikimi T, Saito Y, Matsuo H, Kangawa K. Elevated plasma levels of adrenomedullin in congenital cyanotic heart disease. Clin Sci (Lond). 1999;96:543-547.

128. Tanaka M, Kitamura K, Ishizaka Y, Ishiyama Y, Kato J, Kangawa K, Eto T. Plasma adrenomedullin in various diseases and exercise-induced change in adrenomedullin in healthy subjects. Intern Med. 1995;34:728-733.

129. Cheung B, Leung R. Elevated plasma levels of human adrenomedullin in cardiovascular, respiratory, hepatic and renal disorders. Clin Sci (Lond). 1997;92:59- 62.

130. Jougasaki M, Wei CM, McKinley LJ, Burnett JC, Jr. Elevation of circulating and ventricular adrenomedullin in human congestive heart failure. Circulation. 1995;92:286-289.

131. Ishimitsu T, Nishikimi T, Saito Y, Kitamura K, Eto T, Kangawa K, Matsuo H, Omae T, Matsuoka H. Plasma levels of adrenomedullin, a newly identified hypotensive peptide, in patients with hypertension and renal failure. J Clin Invest. 1994;94:2158- 2161.

132. Caliumi C, Cianci R, Celi M, Cerci S, Cotesta D, Petramala L, Fontana S, Letizia C. Plasma adrenomedullin concentrations in patients with renovascular or malignant hypertension. Minerva Cardioangiol. 2004;52:313-322.

133. Yamamoto K, Ikeda U, Sekiguchi H, Shimada K. Plasma levels of adrenomedullin in patients with mitral stenosis. Am Heart J. 1998;135:542-549.

134. Wolk R, Svatikova A, Otto ME, Hoffmann MS, Duenwald CJ, Somers VK. Plasma adrenomedullin and obstructive sleep apnea. Am J Hypertens. 2004;17:74-76.

135. Suzuki Y, Horio T, Hayashi T, Nonogi H, Kitamura K, Eto T, Kangawa K, Kawano Y. Plasma adrenomedullin concentration is increased in patients with peripheral arterial occlusive disease associated with vascular inflammation. Regul Pept. 2004;118:99- 104.

191

136. Gouya G, Sturm G, Lamina C, Zitt E, Freistatter O, Struck J, Wolzt M, Knoll F, Lins F, Lhotta K, et al. The association of mid-regional pro-adrenomedullin and midregional pro-atrial natriuretic peptide with mortality in an incident dialysis cohort. PLoS One. 2011;6:e17803.

137. Maisel A, Mueller C, Nowak R, Peacock WF, Landsberg JW, Ponikowski P, Mockel M, Hogan C, Wu AH, Richards M, et al. Mid-region pro-hormone markers for diagnosis and prognosis in acute dyspnea: results from the BACH (Biomarkers in Acute Heart Failure) trial. J Am Coll Cardiol. 2010;55:2062-2076.

138. Dhillon OS, Khan SQ, Narayan HK, Ng KH, Struck J, Quinn PA, Morgenthaler NG, Squire IB, Davies JE, Bergmann A, et al. Prognostic value of mid-regional proadrenomedullin levels taken on admission and discharge in non-ST-elevation myocardial infarction: the LAMP (Leicester Acute Myocardial Infarction Peptide) II study. J Am Coll Cardiol. 2010;56:125-133.

139. von Haehling S, Filippatos GS, Papassotiriou J, Cicoira M, Jankowska EA, Doehner W, Rozentryt P, Vassanelli C, Struck J, Banasiak W, et al. Mid-regional proadrenomedullin as a novel predictor of mortality in patients with chronic heart failure. Eur J Heart Fail. 2010;12:484-491.

140. Iacobellis G, di Gioia CR, Di Vito M, Petramala L, Cotesta D, De Santis V, Vitale D, Tritapepe L, Letizia C. Epicardial adipose tissue and intracoronary adrenomedullin levels in coronary artery disease. Horm Metab Res. 2009;41:855-860.

141. Gombos T, Forhecz Z, Pozsonyi Z, Wallentin S, Papassotiriou J, Kunde J, Morgenthaler NG, Janoskuti L, Prohaszka Z. Adrenomedullin and endothelin-1 are related to inflammation in chronic heart failure. Inflamm Res. 2009;58:298-305.

142. Nishida H, Horio T, Suzuki Y, Iwashima Y, Kamide K, Kangawa K, Kawano Y. Plasma adrenomedullin as an independent predictor of future cardiovascular events in high-risk patients: comparison with C-reactive protein and adiponectin. Peptides. 2008;29:599-605.

143. Teramoto S, Yamaguchi Y, Yamamoto H, Hanaoka Y, Ishii M, Hibi S, Kume H, Akishita M, Ouchi Y. Effects of age and sex on plasma adrenomedullin levels in patients with obstructive sleep apnea syndrome. J Am Geriatr Soc. 2007;55:1891- 1892.

144. Schulz R, Flototto C, Jahn A, Eisele HJ, Weissmann N, Seeger W, Rose F. Circulating adrenomedullin in obstructive sleep apnoea. J Sleep Res. 2006;15:89-95.

192

145. Katayama T, Nakashima H, Takagi C, Honda Y, Suzuki S, Yano K. Predictors of mortality in patients with acute myocardial infarction and cardiogenic shock. Circ J. 2005;69:83-88.

146. Katayama T, Nakashima H, Furudono S, Honda Y, Suzuki S, Yano K. Evaluation of neurohumoral activation (adrenomedullin, BNP, catecholamines, etc.) in patients with acute myocardial infarction. Intern Med. 2004;43:1015-1022.

147. Zhang D, Sun XF, Ma ZF, Zhu HY, Wang YD, Chen XM. Effects of high-flux hemodialysis on plasma adrenomedullin and sustained hypotension in elderly hemodialysis patients. Chin Med J (Engl). 2011;124:907-910.

148. Shinomiya K, Ohmori K, Ohyama H, Hosomi N, Takahashi T, Osaka K, Kohno M. Association of plasma adrenomedullin with carotid atherosclerosis in chronic ischemic stroke. Peptides. 2001;22:1873-1880.

149. Guevara M, Gines P, Jimenez W, Sort P, Fernandez-Esparrach G, Escorsell A, Bataller R, Bosch J, Arroyo V, Rivera F, et al. Increased adrenomedullin levels in cirrhosis: relationship with hemodynamic abnormalities and vasoconstrictor systems. Gastroenterology. 1998;114:336-343.

150. Fernandez-Rodriguez CM, Prada IR, Prieto J, Montuenga LM, Elssasser T, Quiroga J, Moreiras M, Andrade A, Cuttitta F. Circulating adrenomedullin in cirrhosis: relationship to hyperdynamic circulation. J Hepatol. 1998;29:250-256.

151. Tahan V, Avsar E, Karaca C, Uslu E, Eren F, Aydin S, Uzun H, Hamzaoglu HO, Besisik F, Kalayci C, et al. Adrenomedullin in cirrhotic and non-cirrhotic portal hypertension. World J Gastroenterol. 2003;9:2325-2327.

152. Mallamaci F, Zoccali C, Parlongo S, Cutrupi S, Tripepi G, Postorino M. Plasma adrenomedullin during acute changes in intravascular volume in hemodialysis patients. Kidney Int. 1998;54:1697-1703.

153. Tokura T, Kinoshita H, Fujimoto S, Hisanaga S, Kitamura K, Eto T. Plasma levels of mature form of adrenomedullin in patients with haemodialysis. Nephrol Dial Transplant. 2001;16:783-786.

154. Cases A, Esforzado N, Lario S, Vera M, Lopez-Pedret J, Rivera-Fillat F, Jimenez W. Increased plasma adrenomedullin levels in hemodialysis patients with sustained hypotension. Kidney Int. 2000;57:664-670.

193

155. Yamasaki H, Nagake Y, Akagi S, Sugimoto T, Ichikawa H, Makino H. Plasma adrenomedullin levels in patients on hemodialysis. Nephron. 2001;89:20-25.

156. Eto T, Washimine H, Kato J, Kitamura K, Yamamoto Y. Adrenomedullin and proadrenomedullin N-terminal 20 peptide in impaired renal function. Kidney Int Suppl. 1996;55:S148-149.

157. Ishihara T, Yokota N, Hisanaga S, Fujimoto S, Hirayama N, Kato J, Kitamura K, Eto T. Increased plasma levels of mature form of adrenomedullin in patients with chronic renal failure. Clin Nephrol. 1999;52:119-123.

158. Kojima H, Tsujimoto T, Uemura M, Takaya A, Okamoto S, Ueda S, Nishio K, Miyamoto S, Kubo A, Minamino N, et al. Significance of increased plasma adrenomedullin concentration in patients with cirrhosis. J Hepatol. 1998;28:840-846.

159. Kalman S, Buyan N, Yurekli M, Ozkaya O, Bakkaloglu S, Soylemezoglu O. Plasma and urinary adrenomedullin levels in children with acute pyelonephritis. Nephrology (Carlton). 2005;10:487-490.

160. Dieplinger B, Mueller T, Kollerits B, Struck J, Ritz E, von Eckardstein A, Haltmayer M, Kronenberg F. Pro-A-type natriuretic peptide and pro-adrenomedullin predict progression of chronic kidney disease: the MMKD Study. Kidney Int. 2009;75:408- 414.

161. Ceyhan BB, Karakurt S, Hekim N. Plasma adrenomedullin levels in asthmatic patients. J Asthma. 2001;38:221-227.

162. Kohno M, Hanehira T, Hirata K, Kawaguchi T, Okishio K, Kano H, Kanazawa H, Yoshikawa J. An accelerated increase of plasma adrenomedullin in acute asthma. Metabolism. 1996;45:1323-1325.

163. Vizza CD, Letizia C, Sciomer S, Naeije R, Della Rocca G, Di Roma A, Musaro S, Quattrucci S, Gaudio C, Battagliese A, et al. Increased plasma levels of adrenomedullin, a vasoactive peptide, in patients with end-stage pulmonary disease. Regul Pept. 2005;124:187-193.

164. Kakishita M, Nishikimi T, Okano Y, Satoh T, Kyotani S, Nagaya N, Fukushima K, Nakanishi N, Takishita S, Miyata A, et al. Increased plasma levels of adrenomedullin in patients with pulmonary hypertension. Clin Sci (Lond). 1999;96:33-39.

165. Potocki M, Breidthardt T, Reichlin T, Morgenthaler NG, Bergmann A, Noveanu M, Schaub N, Uthoff H, Freidank H, Buser L, et al. Midregional pro-adrenomedullin in

194

addition to b-type natriuretic peptides in the risk stratification of patients with acute dyspnea: an observational study. Crit Care. 2009;13:R122.

166. Stolz D, Christ-Crain M, Morgenthaler NG, Miedinger D, Leuppi J, Muller C, Bingisser R, Struck J, Muller B, Tamm M. Plasma pro-adrenomedullin but not plasma pro-endothelin predicts survival in exacerbations of COPD. Chest. 2008;134:263- 272.

167. Hamada H, Saisyo K, Sekimoto T, Chosa E. Plasma adrenomedullin and proadrenomedullin N-terminal 20 peptide in patients diagnosed as having early rheumatoid arthritis. Mod Rheumatol. 2010;20:389-395.

168. Mok MY, Cheung BM, Lo Y, Leung RY, Wong WS, Lau CS. Elevated plasma adrenomedullin and vascular manifestations in patients with systemic sclerosis. J Rheumatol. 2007;34:2224-2229.

169. Garcia-Unzueta MT, Martinez-Taboada VM, Amado-Senaris JA, Rodriguez-Valverde V. Plasma adrenomedullin levels in patients with polymyalgia rheumatica and giant cell arteritis. Clin Exp Rheumatol. 2006;24:S6-9.

170. Balat A, Islek I, Cekmen M, Yurekli M, Tekin D, Muslu A, Sahinoz S. Adrenomedullin and total nitrite levels in children with familial Mediterranean fever. J Paediatr Child Health. 2006;42:240-243.

171. Balat A, Kilinc M, Cekmen MB, Guler E, Yurekli M, Sahinoz S, Coskun Y. Adrenomedullin and total nitrite levels in children with acute rheumatic fever. Clin Biochem. 2005;38:526-530.

172. Michels M, Djamiatun K, Faradz SM, Koenders MM, de Mast Q, van der Ven AJ. High plasma mid-regional pro-adrenomedullin levels in children with severe dengue virus infections. J Clin Virol. 2011;50:8-12.

173. Mak A, Cheung BM, Mok CC, Leung R, Lau CS. Adrenomedullin--a potential disease activity marker and suppressor of nephritis activity in systemic lupus erythematosus. Rheumatology (Oxford). 2006;45:1266-1272.

174. Di Iorio R, Marinoni E, Letizia C, Villaccio B, Alberini A, Cosmi EV. Adrenomedullin production is increased in normal human pregnancy. Eur J Endocrinol. 1999;140:201-206.

195

175. Nagata N, Kato J, Kitamura K, Kawamoto M, Tanaka N, Eto T, Takasaki M. Dissociation of adrenomedullin concentrations in plasma and cerebrospinal fluid in pregnant and non-pregnant women. Eur J Endocrinol. 1998;139:611-614.

176. Poyhonen-Alho M, Viitasalo M, Nicholls MG, Lindstrom BM, Vaananen H, Kaaja R. Imbalance of the autonomic nervous system at night in women with gestational diabetes. Diabet Med. 2010;27:988-994.

177. Shinozaki H, Aoki H, Kasahara Y, Kangawa K, Minegishi T. Plasma adrenomedullin levels during multiple pregnancy. Gynecol Obstet Invest. 2010;69:169-173.

178. Hata T, Miyazaki K, Matsui K. Decreased circulating adrenomedullin in preeclampsia. Lancet. 1997;350:1600.

179. Martinez A, Elsasser TH, Bhathena SJ, Pio R, Buchanan TA, Macri CJ, Cuttitta F. Is adrenomedullin a causal agent in some cases of type 2 diabetes? Peptides. 1999;20:1471-1478.

180. Di Iorio R, Marinoni E, Urban G, Costantini A, Cosmi EV, Letizia C. Fetomaternal adrenomedullin levels in diabetic pregnancy. Horm Metab Res. 2001;33:486-490.

181. Di Iorio R, Marinoni E, Letizia C, Gazzolo D, Lucchini C, Cosmi EV. Adrenomedullin is increased in the fetoplacental circulation in intrauterine growth restriction with abnormal umbilical artery waveforms. Am J Obstet Gynecol. 2000;182:650-654.

182. Makino I, Makino Y, Yoshihara F, Nishikimi T, Kawarabayashi T, Kangawa K, Shibata K. Decreased mature adrenomedullin levels in feto-maternal tissues of pregnant women with histologic chorioamnionitis. Biochem Biophys Res Commun. 2003;301:437-442.

183. Jerat S, Morrish DW, Davidge ST, Kaufman S. Effect of adrenomedullin on placental arteries in normal and preeclamptic pregnancies. Hypertension. 2001;37:227-231.

184. Makino Y, Shibata K, Makino I, Ono Y, Kangawa K, Kawarabayashi T. Expression of adrenomedullin in feto-placental circulation of human normotensive pregnant women and pregnancy-induced hypertensive women. Endocrinology. 1999;140:5439-5442.

185. Di Iorio R, Marinoni E, Letizia C, Alo P, Villaccio B, Cosmi EV. Adrenomedullin, a new vasoactive peptide, is increased in preeclampsia. Hypertension. 1998;32:758- 763.

196

186. Lu B, Wang A, Fei Y. [Study of the plasma adrenomedullin value in pregnancy induced hypertension patients]. Zhonghua Fu Chan Ke Za Zhi. 1999;34:17-19.

187. Lauria MR, Standley CA, Sorokin Y, Yelian FD, Cotton DB. Adrenomedullin levels in normal and preeclamptic pregnancy at term. J Soc Gynecol Investig. 1999;6:318- 321.

188. Di Iorio R, Marinoni E, Letizia C, Alo P, Villaccio B, Poverini R, Cosmi EV. Influence of labor on fetoplacental adrenomedullin concentrations. Am J Obstet Gynecol. 2001;185:697-702.

189. El-mashad AI, Mohamed MA, Farag MA, Ahmad MK, Ismail Y. Role of uterine artery Doppler velocimetry indices and plasma adrenomedullin level in women with unexplained recurrent pregnancy loss. J Obstet Gynaecol Res. 2011;37:51-57.

190. Dikensoy E, Balat O, Pence S, Balat A, Cekmen M, Yurekli M. The changes of plasma malondialdehyde, nitric oxide, and adrenomedullin levels in patients with preeclampsia. Hypertens Pregnancy. 2009;28:383-389.

191. Christopoulos A, Christopoulos G, Morfis M, Udawela M, Laburthe M, Couvineau A, Kuwasako K, Tilakaratne N, Sexton PM. Novel receptor partners and function of receptor activity-modifying proteins. J Biol Chem. 2003;278:3293-3297.

192. Bouschet T, Martin S, Henley JM. Receptor-activity-modifying proteins are required for forward trafficking of the calcium-sensing receptor to the plasma membrane. J Cell Sci. 2005;118:4709-4720.

193. Lenhart PM, Broselid S, Barrick CJ, Leeb-Lundberg LM, Caron KM. G-protein- coupled receptor 30 interacts with receptor activity-modifying protein 3 and confers sex-dependent cardioprotection. J Mol Endocrinol. 2013;51:191-202.

194. Li M, Wetzel-Strong SE, Hua X, Tilley SL, Oswald E, Krummel MF, Caron KM. Deficiency of RAMP1 attenuates antigen-induced airway hyperresponsiveness in mice. PLoS One. 2014;9:e102356.

195. Kadmiel M, Fritz-Six KL, Caron KM. Understanding RAMPs through genetically engineered mouse models. Adv Exp Med Biol. 2012;744:49-60.

196. Barrick CJ, Lenhart PM, Dackor RT, Nagle E, Caron KM. Loss of receptor activity- modifying protein 3 exacerbates cardiac hypertrophy and transition to heart failure in a sex-dependent manner. J Mol Cell Cardiol. 2012;52:165-174.

197

197. Tsujikawa K, Yayama K, Hayashi T, Matsushita H, Yamaguchi T, Shigeno T, Ogitani Y, Hirayama M, Kato T, Fukada S, et al. Hypertension and dysregulated proinflammatory cytokine production in receptor activity-modifying protein 1-deficient mice. Proc Natl Acad Sci U S A. 2007;104:16702-16707.

198. Kadmiel M, Fritz-Six K, Pacharne S, Richards GO, Li M, Skerry TM, Caron KM. Research resource: Haploinsufficiency of receptor activity-modifying protein-2 (RAMP2) causes reduced fertility, hyperprolactinemia, skeletal abnormalities, and endocrine dysfunction in mice. Mol Endocrinol. 2011;25:1244-1253.

199. Hoopes SL, Willcockson HH, Caron KM. Characteristics of multi-organ lymphangiectasia resulting from temporal deletion of calcitonin receptor-like receptor in adult mice. PLoS One. 2012;7:e45261.

200. Wetzel-Strong SE, Li M, Klein KR, Nishikimi T, Caron KM. Epicardial-derived adrenomedullin drives cardiac hyperplasia during embryogenesis. Dev Dyn. 2014;243:243-256.

201. Xiao L, Harrell JC, Perou CM, Dudley AC. Identification of a stable molecular signature in mammary tumor endothelial cells that persists in vitro. Angiogenesis. 2014;17:511-518.

202. Qian L, Huang Y, Spencer CI, Foley A, Vedantham V, Liu L, Conway SJ, Fu JD, Srivastava D. In vivo reprogramming of murine cardiac fibroblasts into induced cardiomyocytes. Nature. 2012;485:593-598.

203. Oie E, Vinge LE, Andersen GO, Yndestad A, Krobert KA, Sandberg C, Ahmed MS, Haug T, Levy FO, Skomedal T, et al. RAMP2 and RAMP3 mRNA levels are increased in failing rat cardiomyocytes and associated with increased responsiveness to adrenomedullin. J Mol Cell Cardiol. 2005;38:145-151.

204. Liang L, Tam CW, Pozsgai G, Siow R, Clark N, Keeble J, Husmann K, Born W, Fischer JA, Poston R, et al. Protection of angiotensin II-induced vascular hypertrophy in vascular smooth muscle-targeted receptor activity-modifying protein 2 transgenic mice. Hypertension. 2009;54:1254-1261.

205. Dalleau S, Baradat M, Gueraud F, Huc L. Cell death and diseases related to oxidative stress: 4-hydroxynonenal (HNE) in the balance. Cell Death Differ. 2013;20:1615-1630.

206. Koyama T, Ochoa-Callejero L, Sakurai T, Kamiyoshi A, Ichikawa-Shindo Y, Iinuma N, Arai T, Yoshizawa T, Iesato Y, Lei Y, et al. Vascular endothelial adrenomedullin-

198

RAMP2 system is essential for vascular integrity and organ homeostasis. Circulation. 2013;127:842-853.

207. Fentzke RC, Korcarz CE, Lang RM, Lin H, Leiden JM. Dilated cardiomyopathy in transgenic mice expressing a dominant-negative CREB transcription factor in the heart. J Clin Invest. 1998;101:2415-2426.

208. Minamishima YA, Moslehi J, Bardeesy N, Cullen D, Bronson RT, Kaelin WG, Jr. Somatic inactivation of the PHD2 prolyl hydroxylase causes polycythemia and congestive heart failure. Blood. 2008;111:3236-3244.

209. Nishikimi T, Yoshihara F, Horinaka S, Kobayashi N, Mori Y, Tadokoro K, Akimoto K, Minamino N, Kangawa K, Matsuoka H. Chronic administration of adrenomedullin attenuates transition from left ventricular hypertrophy to heart failure in rats. Hypertension. 2003;42:1034-1041.

210. Watson PA, Birdsey N, Huggins GS, Svensson E, Heppe D, Knaub L. Cardiac- specific overexpression of dominant-negative CREB leads to increased mortality and mitochondrial dysfunction in female mice. Am J Physiol Heart Circ Physiol. 2010;299:H2056-2068.

211. Patten IS, Rana S, Shahul S, Rowe GC, Jang C, Liu L, Hacker MR, Rhee JS, Mitchell J, Mahmood F, et al. Cardiac angiogenic imbalance leads to peripartum cardiomyopathy. Nature. 2012;485:333-338.

212. Lehman JJ, Barger PM, Kovacs A, Saffitz JE, Medeiros DM, Kelly DP. Peroxisome proliferator-activated receptor gamma coactivator-1 promotes cardiac mitochondrial biogenesis. J Clin Invest. 2000;106:847-856.

213. Sihag S, Cresci S, Li AY, Sucharov CC, Lehman JJ. PGC-1alpha and ERRalpha target gene downregulation is a signature of the failing human heart. J Mol Cell Cardiol. 2009;46:201-212.

214. Mery PF, Brechler V, Pavoine C, Pecker F, Fischmeister R. Glucagon stimulates the cardiac Ca2+ current by activation of adenylyl cyclase and inhibition of phosphodiesterase. Nature. 1990;345:158-161.

215. Ali S, Ussher JR, Baggio LL, Kabir MG, Charron MJ, Ilkayeva O, Newgard CB, Drucker DJ. Cardiomyocyte glucagon receptor signaling modulates outcomes in mice with experimental myocardial infarction. Mol Metab. 2015;4:132-143.

199

216. Herzig S, Long F, Jhala US, Hedrick S, Quinn R, Bauer A, Rudolph D, Schutz G, Yoon C, Puigserver P, et al. CREB regulates hepatic gluconeogenesis through the coactivator PGC-1. Nature. 2001;413:179-183.

217. Koo SH, Flechner L, Qi L, Zhang X, Screaton RA, Jeffries S, Hedrick S, Xu W, Boussouar F, Brindle P, et al. The CREB coactivator TORC2 is a key regulator of fasting glucose metabolism. Nature. 2005;437:1109-1111.

218. Zhang R, Edwards JR, Ko SY, Dong S, Liu H, Oyajobi BO, Papasian C, Deng HW, Zhao M. Transcriptional regulation of BMP2 expression by the PTH-CREB signaling pathway in osteoblasts. PLoS One. 2011;6:e20780.

219. Hoare SR, Gardella TJ, Usdin TB. Evaluating the signal transduction mechanism of the parathyroid hormone 1 receptor. Effect of receptor-G-protein interaction on the ligand binding mechanism and receptor conformation. J Biol Chem. 2001;276:7741- 7753.

220. Avlani VA, Ma W, Mun HC, Leach K, Delbridge L, Christopoulos A, Conigrave AD. Calcium-sensing receptor-dependent activation of CREB phosphorylation in HEK293 cells and human parathyroid cells. Am J Physiol Endocrinol Metab. 2013;304:E1097- 1104.

221. Nagaya N, Satoh T, Nishikimi T, Uematsu M, Furuichi S, Sakamaki F, Oya H, Kyotani S, Nakanishi N, Goto Y, et al. Hemodynamic, renal, and hormonal effects of adrenomedullin infusion in patients with congestive heart failure. Circulation. 2000;101:498-503.

222. Yoshizawa T, Sakurai T, Kamiyoshi A, Ichikawa-Shindo Y, Kawate H, Iesato Y, Koyama T, Uetake R, Yang L, Yamauchi A, et al. Novel regulation of cardiac metabolism and homeostasis by the adrenomedullin-receptor activity-modifying protein 2 system. Hypertension. 2013;61:341-351.

223. Weston C, Lu J, Li N, Barkan K, Richards GO, Roberts DJ, Skerry TM, Poyner D, Pardamwar M, Reynolds CA, et al. Modulation of Glucagon Receptor Pharmacology by Receptor Activity-modifying Protein-2 (RAMP2). J Biol Chem. 2015;290:23009- 23022.

224. Bhashyam S, Fields AV, Patterson B, Testani JM, Chen L, Shen YT, Shannon RP. Glucagon-like peptide-1 increases myocardial glucose uptake via p38alpha MAP kinase-mediated, nitric oxide-dependent mechanisms in conscious dogs with dilated cardiomyopathy. Circ Heart Fail. 2010;3:512-521.

200

225. Solaro RJ. Is calcium the 'cure' for dilated cardiomyopathy? Nat Med. 1999;5:1353- 1354.

226. Polat V, Bozcali E, Uygun T, Opan S, Karakaya O. Low vitamin D status associated with dilated cardiomyopathy. Int J Clin Exp Med. 2015;8:1356-1362.

227. Ittner LM, Gotz J. Pronuclear injection for the production of transgenic mice. Nat Protoc. 2007;2:1206-1215.

228. Holtwick R, Gotthardt M, Skryabin B, Steinmetz M, Potthast R, Zetsche B, Hammer RE, Herz J, Kuhn M. Smooth muscle-selective deletion of guanylyl cyclase-A prevents the acute but not chronic effects of ANP on blood pressure. Proc Natl Acad Sci U S A. 2002;99:7142-7147.

229. Schmidt-Supprian M, Rajewsky K. Vagaries of conditional gene targeting. Nat Immunol. 2007;8:665-668.

230. Krege JH, Hodgin JB, Hagaman JR, Smithies O. A noninvasive computerized tail- cuff system for measuring blood pressure in mice. Hypertension. 1995;25:1111- 1115.

231. Rokosh DG, Simpson PC. Knockout of the alpha 1A/C-adrenergic receptor subtype: the alpha 1A/C is expressed in resistance arteries and is required to maintain arterial blood pressure. Proc Natl Acad Sci U S A. 2002;99:9474-9479.

232. Ray JL, Leach R, Herbert JM, Benson M. Isolation of vascular smooth muscle cells from a single murine aorta. Methods Cell Sci. 2001;23:185-188.

233. Banerji S, Ni J, Wang SX, Clasper S, Su J, Tammi R, Jones M, Jackson DG. LYVE- 1, a new homologue of the CD44 glycoprotein, is a lymph-specific receptor for hyaluronan. J Cell Biol. 1999;144:789-801.

234. Breiteneder-Geleff S, Matsui K, Soleiman A, Meraner P, Poczewski H, Kalt R, Schaffner G, Kerjaschki D. Podoplanin, novel 43-kd membrane protein of glomerular epithelial cells, is down-regulated in puromycin nephrosis. Am J Pathol. 1997;151:1141-1152.

235. Tammela T, Alitalo K. Lymphangiogenesis: Molecular mechanisms and future promise. Cell. 2010;140:460-476.

201

236. Albrecht I, Christofori G. Molecular mechanisms of lymphangiogenesis in development and cancer. Int J Dev Biol. 2011;55:483-494.

237. Bono P, Wasenius VM, Heikkila P, Lundin J, Jackson DG, Joensuu H. High LYVE-1- positive lymphatic vessel numbers are associated with poor outcome in breast cancer. Clin Cancer Res. 2004;10:7144-7149.

238. Cao Y. Opinion: emerging mechanisms of tumour lymphangiogenesis and lymphatic metastasis. Nat Rev Cancer. 2005;5:735-743.

239. Hirakawa S, Kodama S, Kunstfeld R, Kajiya K, Brown LF, Detmar M. VEGF-A induces tumor and sentinel lymph node lymphangiogenesis and promotes lymphatic metastasis. J Exp Med. 2005;201:1089-1099.

240. Skobe M, Hawighorst T, Jackson DG, Prevo R, Janes L, Velasco P, Riccardi L, Alitalo K, Claffey K, Detmar M. Induction of tumor lymphangiogenesis by VEGF-C promotes breast cancer metastasis. Nat Med. 2001;7:192-198.

241. Karpanen T, Egeblad M, Karkkainen MJ, Kubo H, Yla-Herttuala S, Jaattela M, Alitalo K. Vascular endothelial growth factor C promotes tumor lymphangiogenesis and intralymphatic tumor growth. Cancer Res. 2001;61:1786-1790.

242. Mandriota SJ, Jussila L, Jeltsch M, Compagni A, Baetens D, Prevo R, Banerji S, Huarte J, Montesano R, Jackson DG, et al. Vascular endothelial growth factor-C- mediated lymphangiogenesis promotes tumour metastasis. EMBO J. 2001;20:672- 682.

243. Hirakawa S, Brown LF, Kodama S, Paavonen K, Alitalo K, Detmar M. VEGF-C- induced lymphangiogenesis in sentinel lymph nodes promotes tumor metastasis to distant sites. Blood. 2007;109:1010-1017.

244. Stacker SA, Caesar C, Baldwin ME, Thornton GE, Williams RA, Prevo R, Jackson DG, Nishikawa S, Kubo H, Achen MG. VEGF-D promotes the metastatic spread of tumor cells via the lymphatics. Nat Med. 2001;7:186-191.

245. Roesli C, Mumprecht V, Neri D, Detmar M. Identification of the surface-accessible, lineage-specific vascular proteome by two-dimensional peptide mapping. FASEB J. 2008;22:1933-1944.

246. Wallgard E, Larsson E, He L, Hellstrom M, Armulik A, Nisancioglu MH, Genove G, Lindahl P, Betsholtz C. Identification of a core set of 58 gene transcripts with broad

202

and specific expression in the microvasculature. Arterioscler Thromb Vasc Biol. 2008;28:1469-1476.

247. Dunworth WP, Caron KM. G protein-coupled receptors as potential drug targets for lymphangiogenesis and lymphatic vascular diseases. Arterioscler Thromb Vasc Biol. 2009;29:650-656.

248. Hay DL, Walker CS, Poyner DR. Adrenomedullin and calcitonin gene-related peptide receptors in endocrine-related cancers: opportunities and challenges. Endocr Relat Cancer. 2011;18:C1-14.

249. Zudaire E, Martinez A, Cuttitta F. Adrenomedullin and cancer. Regul Pept. 2003;112:175-183.

250. Hu Z, Fan C, Livasy C, He X, Oh DS, Ewend MG, Carey LA, Subramanian S, West R, Ikpatt F, et al. A compact VEGF signature associated with distant metastases and poor outcomes. BMC Med. 2009;7:9.

251. Sugiura K, Stock CC. Studies in a tumor spectrum. III. The effect of phosphoramides on the growth of a variety of mouse and rat tumors. Cancer Res. 1955;15:38-51.

252. Pollard JW. Tumour-educated macrophages promote tumour progression and metastasis. Nat Rev Cancer. 2004;4:71-78.

253. Schoppmann SF, Birner P, Stockl J, Kalt R, Ullrich R, Caucig C, Kriehuber E, Nagy K, Alitalo K, Kerjaschki D. Tumor-associated macrophages express lymphatic endothelial growth factors and are related to peritumoral lymphangiogenesis. Am J Pathol. 2002;161:947-956.

254. Kerjaschki D. The crucial role of macrophages in lymphangiogenesis. J Clin Invest. 2005;115:2316-2319.

255. Nuki C, Kawasaki H, Kitamura K, Takenaga M, Kangawa K, Eto T, Wada A. Vasodilator effect of adrenomedullin and calcitonin gene-related peptide receptors in rat mesenteric vascular beds. Biochem Biophys Res Commun. 1993;196:245-251.

256. Ji RC. Lymph node lymphangiogenesis: a new concept for modulating tumor metastasis and inflammatory process. Histol Histopathol. 2009;24:377-384.

203

257. Harrell MI, Iritani BM, Ruddell A. Tumor-induced sentinel lymph node lymphangiogenesis and increased lymph flow precede melanoma metastasis. Am J Pathol. 2007;170:774-786.

258. Qian CN, Berghuis B, Tsarfaty G, Bruch M, Kort EJ, Ditlev J, Tsarfaty I, Hudson E, Jackson DG, Petillo D, et al. Preparing the "soil": the primary tumor induces vasculature reorganization in the sentinel lymph node before the arrival of metastatic cancer cells. Cancer Res. 2006;66:10365-10376.

259. Sharpe J, Ahlgren U, Perry P, Hill B, Ross A, Hecksher-Sorensen J, Baldock R, Davidson D. Optical projection tomography as a tool for 3D microscopy and gene expression studies. Science. 2002;296:541-545.

260. Clasper S, Royston D, Baban D, Cao Y, Ewers S, Butz S, Vestweber D, Jackson DG. A novel gene expression profile in lymphatics associated with tumor growth and nodal metastasis. Cancer Res. 2008;68:7293-7303.

261. Schulz P, Fischer C, Detjen KM, Rieke S, Hilfenhaus G, von Marschall Z, Bohmig M, Koch I, Kehrberger J, Hauff P, et al. Angiopoietin-2 drives lymphatic metastasis of pancreatic cancer. FASEB J. 2011;25:3325-3335.

262. Fagiani E, Lorentz P, Kopfstein L, Christofori G. Angiopoietin-1 and -2 exert antagonistic functions in tumor angiogenesis, yet both induce lymphangiogenesis. Cancer Res. 2011;71:5717-5727.

263. Karnezis T, Shayan R, Caesar C, Roufail S, Harris NC, Ardipradja K, Zhang YF, Williams SP, Farnsworth RH, Chai MG, et al. VEGF-D promotes tumor metastasis by regulating prostaglandins produced by the collecting lymphatic endothelium. Cancer Cell. 2012;21:181-195.

264. Zhao B, Cui K, Wang CL, Wang AL, Zhang B, Zhou WY, Zhao WH, Li S. The chemotactic interaction between CCL21 and its receptor, CCR7, facilitates the progression of pancreatic cancer via induction of angiogenesis and lymphangiogenesis. J Hepatobiliary Pancreat Sci. 2011

265. Shields JD, Emmett MS, Dunn DB, Joory KD, Sage LM, Rigby H, Mortimer PS, Orlando A, Levick JR, Bates DO. Chemokine-mediated migration of melanoma cells towards lymphatics--a mechanism contributing to metastasis. Oncogene. 2007;26:2997-3005.

266. Thouennon E, Pierre A, Tanguy Y, Guillemot J, Manecka DL, Guerin M, Ouafik L, Muresan M, Klein M, Bertherat J, et al. Expression of trophic amidated peptides and their receptors in benign and malignant pheochromocytomas: high expression of

204

adrenomedullin RDC1 receptor and implication in tumoral cell survival. Endocr Relat Cancer. 2010;17:637-651.

267. Padera TP, Kadambi A, di Tomaso E, Carreira CM, Brown EB, Boucher Y, Choi NC, Mathisen D, Wain J, Mark EJ, et al. Lymphatic metastasis in the absence of functional intratumor lymphatics. Science. 2002;296:1883-1886.

268. Skobe M, Hamberg LM, Hawighorst T, Schirner M, Wolf GL, Alitalo K, Detmar M. Concurrent induction of lymphangiogenesis, angiogenesis, and macrophage recruitment by vascular endothelial growth factor-C in melanoma. Am J Pathol. 2001;159:893-903.

269. Chen P, Huang Y, Bong R, Ding Y, Song N, Wang X, Song X, Luo Y. Tumor- associated macrophages promote angiogenesis and melanoma growth via adrenomedullin in a paracrine and autocrine manner. Clin Cancer Res. 2011;17:7230-7239.

270. Holopainen T, Saharinen P, D'Amico G, Lampinen A, Eklund L, Sormunen R, Anisimov A, Zarkada G, Lohela M, Helotera H, et al. Effects of angiopoietin-2- blocking antibody on endothelial cell-cell junctions and lung metastasis. J Natl Cancer Inst. 2012;104:461-475.

271. Doods H, Arndt K, Rudolf K, Just S. CGRP antagonists: unravelling the role of CGRP in migraine. Trends Pharmacol Sci. 2007;28:580-587.

272. Civelli O, Reinscheid RK, Zhang Y, Wang Z, Fredriksson R, Schioth HB. G protein- coupled receptor deorphanizations. Annu Rev Pharmacol Toxicol. 2013;53:127-146.

273. Nagi K, Pineyro G. Practical guide for calculating and representing biased signaling by GPCR ligands: A stepwise approach. Methods. 2015

274. Whalen EJ, Rajagopal S, Lefkowitz RJ. Therapeutic potential of beta-arrestin- and G protein-biased agonists. Trends Mol Med. 2011;17:126-139.

275. Hay DL, Pioszak AA. Receptor Activity-Modifying Proteins (RAMPs): New Insights and Roles. Annu Rev Pharmacol Toxicol. 2016;56:469-487.

276. ter Haar E, Koth CM, Abdul-Manan N, Swenson L, Coll JT, Lippke JA, Lepre CA, Garcia-Guzman M, Moore JM. Crystal structure of the ectodomain complex of the CGRP receptor, a class-B GPCR, reveals the site of drug antagonism. Structure. 2010;18:1083-1093.

205

277. Watkins HA, Chakravarthy M, Abhayawardana RS, Gingell JJ, Garelja M, Pardamwar M, McElhinney JM, Lathbridge A, Constantine A, Harris PW, et al. Receptor Activity-modifying Proteins 2 and 3 Generate Adrenomedullin Receptor Subtypes with Distinct Molecular Properties. J Biol Chem. 2016;291:11657-11675.

278. Li P, Cavallero S, Gu Y, Chen TH, Hughes J, Hassan AB, Bruning JC, Pashmforoush M, Sucov HM. IGF signaling directs ventricular cardiomyocyte proliferation during embryonic heart development. Development. 2011;138:1795- 1805.

279. Lavine KJ, Yu K, White AC, Zhang X, Smith C, Partanen J, Ornitz DM. Endocardial and epicardial derived FGF signals regulate myocardial proliferation and differentiation in vivo. Dev Cell. 2005;8:85-95.

280. D'Amato G, Luxan G, del Monte-Nieto G, Martinez-Poveda B, Torroja C, Walter W, Bochter MS, Benedito R, Cole S, Martinez F, et al. Sequential Notch activation regulates ventricular chamber development. Nat Cell Biol. 2016;18:7-20.

281. de la Pompa JL, Epstein JA. Coordinating tissue interactions: Notch signaling in cardiac development and disease. Dev Cell. 2012;22:244-254.

282. Grego-Bessa J, Luna-Zurita L, del Monte G, Bolos V, Melgar P, Arandilla A, Garratt AN, Zang H, Mukouyama YS, Chen H, et al. Notch signaling is essential for ventricular chamber development. Dev Cell. 2007;12:415-429.

283. Samsa LA, Givens C, Tzima E, Stainier DY, Qian L, Liu J. Cardiac contraction activates endocardial Notch signaling to modulate chamber maturation in zebrafish. Development. 2015;142:4080-4091.

284. Mack CP. Signaling mechanisms that regulate smooth muscle cell differentiation. Arterioscler Thromb Vasc Biol. 2011;31:1495-1505.

285. Wang G, Jacquet L, Karamariti E, Xu Q. Origin and differentiation of vascular smooth muscle cells. J Physiol. 2015;593:3013-3030.

286. Hilfiker-Kleiner D, Kaminski K, Podewski E, Bonda T, Schaefer A, Sliwa K, Forster O, Quint A, Landmesser U, Doerries C, et al. A cathepsin D-cleaved 16 kDa form of prolactin mediates postpartum cardiomyopathy. Cell. 2007;128:589-600.

287. Hershberger RE, Hedges DJ, Morales A. Dilated cardiomyopathy: the complexity of a diverse genetic architecture. Nat Rev Cardiol. 2013;10:531-547.

206

288. Bersell K, Choudhury S, Mollova M, Polizzotti BD, Ganapathy B, Walsh S, Wadugu B, Arab S, Kuhn B. Moderate and high amounts of tamoxifen in alphaMHC- MerCreMer mice induce a DNA damage response, leading to heart failure and death. Dis Model Mech. 2013;6:1459-1469.

289. Koitabashi N, Bedja D, Zaiman AL, Pinto YM, Zhang M, Gabrielson KL, Takimoto E, Kass DA. Avoidance of transient cardiomyopathy in cardiomyocyte-targeted tamoxifen-induced MerCreMer gene deletion models. Circ Res. 2009;105:12-15.

290. Sauvadet A, Rohn T, Pecker F, Pavoine C. Synergistic actions of glucagon and miniglucagon on Ca2+ mobilization in cardiac cells. Circ Res. 1996;78:102-109.

291. Chang-Chretien K, Chew JT, Judge DP. Reversible dilated cardiomyopathy associated with glucagonoma. Heart. 2004;90:e44.

292. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144:646-674.

293. Chaffer CL, Weinberg RA. A perspective on cancer cell metastasis. Science. 2011;331:1559-1564.

294. Valastyan S, Weinberg RA. Tumor metastasis: molecular insights and evolving paradigms. Cell. 2011;147:275-292.

295. Dunleavey JM, Xiao L, Thompson J, Kim MM, Shields JM, Shelton SE, Irvin DM, Brings VE, Ollila DW, Brekken RA, et al. Vascular channels formed by subpopulations of PECAM1+ melanoma cells. Nat Commun. 2014;5:5200.

296. Xiao L, Kim DJ, Davis CL, McCann JV, Dunleavey JM, Vanderlinden AK, Xu N, Pattenden SG, Frye SV, Xu X, et al. Tumor Endothelial Cells with Distinct Patterns of TGFbeta-Driven Endothelial-to-Mesenchymal Transition. Cancer Res. 2015;75:1244- 1254.

297. Jain RK, Duda DG, Clark JW, Loeffler JS. Lessons from phase III clinical trials on anti-VEGF therapy for cancer. Nat Clin Pract Oncol. 2006;3:24-40.

298. Alitalo K. The lymphatic vasculature in disease. Nat Med. 2011;17:1371-1380.

299. Saharinen P, Eklund L, Pulkki K, Bono P, Alitalo K. VEGF and angiopoietin signaling in tumor angiogenesis and metastasis. Trends Mol Med. 2011;17:347-362.

207

300. Yang H, Kim C, Kim MJ, Schwendener RA, Alitalo K, Heston W, Kim I, Kim WJ, Koh GY. Soluble vascular endothelial growth factor receptor-3 suppresses lymphangiogenesis and lymphatic metastasis in bladder cancer. Mol Cancer. 2011;10:36.

301. Chen P, Huang Y, Bong R, Ding Y, Song N, Wang X, Song X, Luo Y. Tumor- associated macrophages promote angiogenesis and melanoma growth via adrenomedullin in both paracrine and autocrine manners. Clin Cancer Res. 2011

302. Wu R, Zhou M, Wang P. Adrenomedullin and adrenomedullin binding protein-1 downregulate TNF-alpha in macrophage cell line and rat Kupffer cells. Regul Pept. 2003;112:19-26.

303. Bhardwaj A, Srivastava SK, Singh S, Tyagi N, Arora S, Carter JE, Khushman M, Singh AP. MYB promotes desmoplasia in pancreatic cancer through direct transcriptional upregulation and cooperative action of sonic hedgehog and adrenomedullin. J Biol Chem. 2016

304. Oehler MK, Hague S, Rees MC, Bicknell R. Adrenomedullin promotes formation of xenografted endometrial tumors by stimulation of autocrine growth and angiogenesis. Oncogene. 2002;21:2815-2821.

305. Karpinich NO, Kechele DO, Espenschied ST, Willcockson HH, Fedoriw Y, Caron KM. Adrenomedullin gene dosage correlates with tumor and lymph node lymphangiogenesis. FASEB J. 2013;27:590-600.

306. Karpinich NO, Caron KM. Gap junction coupling is required for tumor cell migration through lymphatic endothelium. Arterioscler Thromb Vasc Biol. 2015;35:1147-1155.

307. Khosrotehrani K, Nguyen Huu S, Prignon A, Avril MF, Boitier F, Oster M, Mortier L, Richard MA, Maubec E, Kerob D, et al. Pregnancy promotes melanoma metastasis through enhanced lymphangiogenesis. Am J Pathol. 2011;178:1870-1880.

308. Curran EM, Berghaus LJ, Vernetti NJ, Saporita AJ, Lubahn DB, Estes DM. Natural killer cells express estrogen receptor-alpha and estrogen receptor-beta and can respond to estrogen via a non-estrogen receptor-alpha-mediated pathway. Cell Immunol. 2001;214:12-20.

309. Rodero MP, Prignon A, Avril MF, Boitier F, Aractingi S, Khosrotehrani K. Increase lymphangiogenesis in melanoma during pregnancy: correlation with the prolactin signalling pathway. J Eur Acad Dermatol Venereol. 2013;27:e144-145.

208

310. Mauti LA, Le Bitoux MA, Baumer K, Stehle JC, Golshayan D, Provero P, Stamenkovic I. Myeloid-derived suppressor cells are implicated in regulating permissiveness for tumor metastasis during mouse gestation. J Clin Invest. 2011;121:2794-2807.

311. Brekhman V, Lugassie J, Zaffryar-Eilot S, Sabo E, Kessler O, Smith V, Golding H, Neufeld G. Receptor activity modifying protein-3 mediates the protumorigenic activity of lysyl oxidase-like protein-2. FASEB J. 2011;25:55-65.

312. Venkatanarayan A, Raulji P, Norton W, Chakravarti D, Coarfa C, Su X, Sandur SK, Ramirez MS, Lee J, Kingsley CV, et al. IAPP-driven metabolic reprogramming induces regression of p53-deficient tumours in vivo. Nature. 2015;517:626-630.

313. Li M, Schwerbrock NM, Lenhart PM, Fritz-Six KL, Kadmiel M, Christine KS, Kraus DM, Espenschied ST, Willcockson HH, Mack CP, et al. Fetal-derived adrenomedullin mediates the innate immune milieu of the placenta. J Clin Invest. 2013;123:2408- 2420.

314. Wetzel-Strong SE, Li M, Espenschied ST, Caron KM. Cohort of estrogen-induced microRNAs regulate adrenomedullin expression. Am J Physiol Regul Integr Comp Physiol. 2016;310:R209-216.

315. Watanabe H, Takahashi E, Kobayashi M, Goto M, Krust A, Chambon P, Iguchi T. The estrogen-responsive adrenomedullin and receptor-modifying protein 3 gene identified by DNA microarray analysis are directly regulated by estrogen receptor. J Mol Endocrinol. 2006;36:81-89.

316. Hofbauer KH, Jensen BL, Kurtz A, Sandner P. Tissue hypoxygenation activates the adrenomedullin system in vivo. Am J Physiol Regul Integr Comp Physiol. 2000;278:R513-519.

317. Lenhart PM, Caron KM. Adrenomedullin and pregnancy: perspectives from animal models to humans. Trends Endocrinol Metab. 2012;23:524-532.

318. Li M, Wu Y, Caron KM. Haploinsufficiency for adrenomedullin reduces pinopodes and diminishes uterine receptivity in mice. Biol Reprod. 2008;79:1169-1175.

319. Matson BC, Caron KM. Uterine natural killer cells as modulators of the maternal-fetal vasculature. Int J Dev Biol. 2014;58:199-204.

320. Matson BC, Caron KM. Adrenomedullin and endocrine control of immune cells during pregnancy. Cell Mol Immunol. 2014;11:456-459.

209

321. Martinez A, Weaver C, Lopez J, Bhathena SJ, Elsasser TH, Miller MJ, Moody TW, Unsworth EJ, Cuttitta F. Regulation of insulin secretion and blood glucose metabolism by adrenomedullin. Endocrinology. 1996;137:2626-2632.

322. Martinez A, Kapas S, Miller MJ, Ward Y, Cuttitta F. Coexpression of receptors for adrenomedullin, calcitonin gene-related peptide, and amylin in pancreatic beta-cells. Endocrinology. 2000;141:406-411.

323. Longuet C, Sinclair EM, Maida A, Baggio LL, Maziarz M, Charron MJ, Drucker DJ. The glucagon receptor is required for the adaptive metabolic response to fasting. Cell Metab. 2008;8:359-371.

324. Sivertsen J, Rosenmeier J, Holst JJ, Vilsboll T. The effect of glucagon-like peptide 1 on cardiovascular risk. Nat Rev Cardiol. 2012;9:209-222.

325. Vuguin PM, Kedees MH, Cui L, Guz Y, Gelling RW, Nejathaim M, Charron MJ, Teitelman G. Ablation of the glucagon receptor gene increases fetal lethality and produces alterations in islet development and maturation. Endocrinology. 2006;147:3995-4006.

326. Kerjaschki D, Huttary N, Raab I, Regele H, Bojarski-Nagy K, Bartel G, Krober SM, Greinix H, Rosenmaier A, Karlhofer F, et al. Lymphatic endothelial progenitor cells contribute to de novo lymphangiogenesis in human renal transplants. Nat Med. 2006;12:230-234.

327. Maruyama K, Ii M, Cursiefen C, Jackson DG, Keino H, Tomita M, Van Rooijen N, Takenaka H, D'Amore PA, Stein-Streilein J, et al. Inflammation-induced lymphangiogenesis in the cornea arises from CD11b-positive macrophages. J Clin Invest. 2005;115:2363-2372.

328. Hall KL, Volk-Draper LD, Flister MJ, Ran S. New model of macrophage acquisition of the lymphatic endothelial phenotype. PLoS One. 2012;7:e31794.

329. Jhaveri MB, Driscoll MS, Grant-Kels JM. Melanoma in pregnancy. Clin Obstet Gynecol. 2011;54:537-545.

330. Jiang X, Ellison SJ, Alarid ET, Shapiro DJ. Interplay between the levels of estrogen and estrogen receptor controls the level of the granzyme inhibitor, proteinase inhibitor 9 and susceptibility to immune surveillance by natural killer cells. Oncogene. 2007;26:4106-4114.

210

331. Barak LS, Ferguson SS, Zhang J, Caron MG. A beta-arrestin/green fluorescent protein biosensor for detecting G protein-coupled receptor activation. J Biol Chem. 1997;272:27497-27500.

332. Klein KR, Karpinich NO, Espenschied ST, Willcockson HH, Dunworth WP, Hoopes SL, Kushner EJ, Bautch VL, Caron KM. Decoy Receptor CXCR7 Modulates Adrenomedullin-Mediated Cardiac and Lymphatic Vascular Development. Dev Cell. 2014;30:528-540.

333. Dackor R, Caron K. Mice heterozygous for adrenomedullin exhibit a more extreme inflammatory response to endotoxin-induced septic shock. Peptides. 2007;28:2164- 2170.

334. Gonzalez-Rey E, Chorny A, O'Valle F, Delgado M. Adrenomedullin protects from experimental arthritis by down-regulating inflammation and Th1 response and inducing regulatory T cells. Am J Pathol. 2007;170:263-271.

335. Nishikimi T, Tadokoro K, Akimoto K, Mori Y, Ishikawa Y, Ishimura K, Horio T, Kangawa K, Matsuoka H. Response of adrenomedullin system to cytokine in cardiac fibroblasts-role of adrenomedullin as an antifibrotic factor. Cardiovasc Res. 2005;66:104-113.

336. Muller WA. Mechanisms of transendothelial migration of leukocytes. Circ Res. 2009;105:223-230.

337. Holt D, Ma X, Kundu N, Fulton A. Prostaglandin E(2) (PGE (2)) suppresses natural killer cell function primarily through the PGE(2) receptor EP4. Cancer Immunol Immunother. 2011;60:1577-1586.

338. Pietra G, Manzini C, Rivara S, Vitale M, Cantoni C, Petretto A, Balsamo M, Conte R, Benelli R, Minghelli S, et al. Melanoma cells inhibit natural killer cell function by modulating the expression of activating receptors and cytolytic activity. Cancer Res. 2012;72:1407-1415.

339. Zwirner NW, Domaica CI. Cytokine regulation of natural killer cell effector functions. Biofactors. 2010;36:274-288.

340. Lee EG, Lee SI, Chae HJ, Park SJ, Lee YC, Yoo WH. Adrenomedullin inhibits IL- 1beta-induced rheumatoid synovial fibroblast proliferation and MMPs, COX-2 and PGE2 production. Inflammation. 2011;34:335-343.

211

341. Kitamura K, Sakata J, Kangawa K, Kojima M, Matsuo H, Eto T. Cloning and characterization of cDNA encoding a precursor for human adrenomedullin. Biochem Biophys Res Commun. 1993;194:720-725.

342. Eguchi S, Hirata Y, Iwasaki H, Sato K, Watanabe TX, Inui T, Nakajima K, Sakakibara S, Marumo F. Structure-activity relationship of adrenomedullin, a novel vasodilatory peptide, in cultured rat vascular smooth muscle cells. Endocrinology. 1994;135:2454- 2458.

343. Ishizaka Y, Tanaka M, Kitamura K, Kangawa K, Minamino N, Matsuo H, Eto T. Adrenomedullin stimulates cyclic AMP formation in rat vascular smooth muscle cells. Biochem Biophys Res Commun. 1994;200:642-646.

344. Coppock HA, Owji AA, Austin C, Upton PD, Jackson ML, Gardiner JV, Ghatei MA, Bloom SR, Smith DM. Rat-2 fibroblasts express specific adrenomedullin receptors, but not calcitonin-gene-related-peptide receptors, which mediate increased intracellular cAMP and inhibit mitogen-activated protein kinase activity. Biochem J. 1999;338 ( Pt 1):15-22.

345. Coppock HA, Owji AA, Bloom SR, Smith DM. A rat skeletal muscle cell line (L6) expresses specific adrenomedullin binding sites but activates adenylate cyclase via calcitonin gene-related peptide receptors. Biochem J. 1996;318 ( Pt 1):241-245.

346. Katoh F, Niina H, Kitamura K, Ichiki Y, Yamamoto Y, Kangawa K, Eto T, Wada A. Ca(2+)-dependent cosecretion of adrenomedullin and catecholamines mediated by nicotinic receptors in bovine cultured adrenal medullary cells. FEBS Lett. 1994;348:61-64.

347. Shimekake Y, Nagata K, Ohta S, Kambayashi Y, Teraoka H, Kitamura K, Eto T, Kangawa K, Matsuo H. Adrenomedullin stimulates two signal transduction pathways, cAMP accumulation and Ca2+ mobilization, in bovine aortic endothelial cells. J Biol Chem. 1995;270:4412-4417.

348. Withers DJ, Coppock HA, Seufferlein T, Smith DM, Bloom SR, Rozengurt E. Adrenomedullin stimulates DNA synthesis and cell proliferation via elevation of cAMP in Swiss 3T3 cells. FEBS Lett. 1996;378:83-87.

349. Zimmermann U, Fischer JA, Muff R. Adrenomedullin and calcitonin gene-related peptide interact with the same receptor in cultured human neuroblastoma SK-N-MC cells. Peptides. 1995;16:421-424.

212

350. Libert F, Parmentier M, Lefort A, Dumont JE, Vassart G. Complete nucleotide sequence of a putative G protein coupled receptor: RDC1. Nucleic Acids Res. 1990;18:1917.

351. Njuki F, Nicholl CG, Howard A, Mak JC, Barnes PJ, Girgis SI, Legon S. A new calcitonin-receptor-like sequence in rat pulmonary blood vessels. Clin Sci (Lond). 1993;85:385-388.

352. Fluhmann B, Muff R, Hunziker W, Fischer JA, Born W. A human orphan calcitonin receptor-like structure. Biochem Biophys Res Commun. 1995;206:341-347.

353. Chang CP, Pearse RV, 2nd, O'Connell S, Rosenfeld MG. Identification of a seven transmembrane helix receptor for corticotropin-releasing factor and sauvagine in mammalian brain. Neuron. 1993;11:1187-1195.

354. Harrison JK, Barber CM, Lynch KR. Molecular cloning of a novel rat G-protein- coupled receptor gene expressed prominently in lung, adrenal, and liver. FEBS Lett. 1993;318:17-22.

355. Eva C, Sprengel R. A novel putative G protein-coupled receptor highly expressed in lung and testis. DNA Cell Biol. 1993;12:393-399.

356. Kapas S, Catt KJ, Clark AJ. Cloning and expression of cDNA encoding a rat adrenomedullin receptor. J Biol Chem. 1995;270:25344-25347.

357. Hanze J, Dittrich K, Dotsch J, Rascher W. Molecular cloning of a novel human receptor gene with homology to the rat adrenomedullin receptor and high expression in heart and immune system. Biochem Biophys Res Commun. 1997;240:183-188.

358. Owji AA, Smith DM, Coppock HA, Morgan DG, Bhogal R, Ghatei MA, Bloom SR. An abundant and specific binding site for the novel vasodilator adrenomedullin in the rat. Endocrinology. 1995;136:2127-2134.

359. Kapas S, Clark AJ. Identification of an orphan receptor gene as a type 1 calcitonin gene-related peptide receptor. Biochem Biophys Res Commun. 1995;217:832-838.

360. Autelitano DJ, Tang F. Co-expression of prepro-adrenomedullin with a putative adrenomedullin receptor gene in vascular smooth muscle. Clin Sci (Lond). 1999;96:493-498.

213

361. Entzeroth M, Doods HN, Wieland HA, Wienen W. Adrenomedullin mediates vasodilation via CGRP1 receptors. Life Sci. 1995;56:PL19-25.

362. Aiyar N, Rand K, Elshourbagy NA, Zeng Z, Adamou JE, Bergsma DJ, Li Y. A cDNA encoding the calcitonin gene-related peptide type 1 receptor. J Biol Chem. 1996;271:11325-11329.

363. Kennedy SP, Sun D, Oleynek JJ, Hoth CF, Kong J, Hill RJ. Expression of the rat adrenomedullin receptor or a putative human adrenomedullin receptor does not correlate with adrenomedullin binding or functional response. Biochem Biophys Res Commun. 1998;244:832-837.

364. Upton PD, Austin C, Taylor GM, Nandha KA, Clark AJ, Ghatei MA, Bloom SR, Smith DM. Expression of adrenomedullin (ADM) and its binding sites in the rat uterus: increased number of binding sites and ADM messenger ribonucleic acid in 20-day pregnant rats compared with nonpregnant rats. Endocrinology. 1997;138:2508-2514.

365. Makino Y, Shibata K, Makino I, Kangawa K, Kawarabayashi T. Alteration of the adrenomedullin receptor components gene expression associated with the blood pressure in pregnancy-induced hypertension. J Clin Endocrinol Metab. 2001;86:5079-5082.

366. Karpinich NO, Hoopes SL, Kechele DO, Lenhart PM, Caron KM. Adrenomedullin Function in Vascular Endothelial Cells: Insights from Genetic Mouse Models. Curr Hypertens Rev. 2011;7:228-239.

367. Klein KR, Caron KM. Adrenomedullin in lymphangiogenesis: from development to disease. Cell Mol Life Sci. 2015;72:3115-3126.

368. Sierro F, Biben C, Martinez-Munoz L, Mellado M, Ransohoff RM, Li M, Woehl B, Leung H, Groom J, Batten M, et al. Disrupted cardiac development but normal hematopoiesis in mice deficient in the second CXCL12/SDF-1 receptor, CXCR7. Proc Natl Acad Sci U S A. 2007;104:14759-14764.

369. Yu S, Crawford D, Tsuchihashi T, Behrens TW, Srivastava D. The chemokine receptor CXCR7 functions to regulate cardiac valve remodeling. Dev Dyn. 2011;240:384-393.

370. Gerrits H, van Ingen Schenau DS, Bakker NE, van Disseldorp AJ, Strik A, Hermens LS, Koenen TB, Krajnc-Franken MA, Gossen JA. Early postnatal lethality and cardiovascular defects in CXCR7-deficient mice. Genesis. 2008;46:235-245.

214

371. Boldajipour B, Mahabaleshwar H, Kardash E, Reichman-Fried M, Blaser H, Minina S, Wilson D, Xu Q, Raz E. Control of chemokine-guided cell migration by ligand sequestration. Cell. 2008;132:463-473.

372. Naumann U, Cameroni E, Pruenster M, Mahabaleshwar H, Raz E, Zerwes HG, Rot A, Thelen M. CXCR7 functions as a scavenger for CXCL12 and CXCL11. PLoS One. 2010;5:e9175.

373. Sanchez-Martin L, Sanchez-Mateos P, Cabanas C. CXCR7 impact on CXCL12 biology and disease. Trends Mol Med. 2013;19:12-22.

374. Wang Y, Li G, Stanco A, Long JE, Crawford D, Potter GB, Pleasure SJ, Behrens T, Rubenstein JL. CXCR4 and CXCR7 have distinct functions in regulating interneuron migration. Neuron. 2011;69:61-76.

375. Berahovich RD, Zabel BA, Lewen S, Walters MJ, Ebsworth K, Wang Y, Jaen JC, Schall TJ. Endothelial expression of CXCR7 and the regulation of systemic CXCL12 levels. Immunology. 2014;141:111-122.

376. Balabanian K, Lagane B, Infantino S, Chow KY, Harriague J, Moepps B, Arenzana- Seisdedos F, Thelen M, Bachelerie F. The chemokine SDF-1/CXCL12 binds to and signals through the orphan receptor RDC1 in T lymphocytes. J Biol Chem. 2005;280:35760-35766.

377. Burns JM, Summers BC, Wang Y, Melikian A, Berahovich R, Miao Z, Penfold ME, Sunshine MJ, Littman DR, Kuo CJ, et al. A novel chemokine receptor for SDF-1 and I-TAC involved in cell survival, cell adhesion, and tumor development. J Exp Med. 2006;203:2201-2213.

378. Ikeda Y, Kumagai H, Skach A, Sato M, Yanagisawa M. Modulation of circadian glucocorticoid oscillation via adrenal opioid-CXCR7 signaling alters emotional behavior. Cell. 2013;155:1323-1336.

379. Neusser MA, Kraus AK, Regele H, Cohen CD, Fehr T, Kerjaschki D, Wuthrich RP, Penfold ME, Schall T, Segerer S. The chemokine receptor CXCR7 is expressed on lymphatic endothelial cells during renal allograft rejection. Kidney Int. 2010;77:801- 808.

380. Ramachandran V, Arumugam T, Langley R, Hwang RF, Vivas-Mejia P, Sood AK, Lopez-Berestein G, Logsdon CD. The ADMR receptor mediates the effects of adrenomedullin on pancreatic cancer cells and on cells of the tumor microenvironment. PLoS One. 2009;4:e7502.

215

381. Takase H, Matsumoto K, Yamadera R, Kubota Y, Otsu A, Suzuki R, Ishitobi H, Mochizuki H, Kojima T, Takano S, et al. Genome-wide identification of endothelial cell-enriched genes in the mouse embryo. Blood. 2012;120:914-923.

382. Sumanas S, Jorniak T, Lin S. Identification of novel vascular endothelial-specific genes by the microarray analysis of the zebrafish cloche mutants. Blood. 2005;106:534-541.

383. Regard JB, Sato IT, Coughlin SR. Anatomical profiling of G protein-coupled receptor expression. Cell. 2008;135:561-571.

384. Alghisi E, Distel M, Malagola M, Anelli V, Santoriello C, Herwig L, Krudewig A, Henkel CV, Russo D, Mione MC. Targeting oncogene expression to endothelial cells induces proliferation of the myelo-erythroid lineage by repressing the Notch pathway. Leukemia. 2013;27:2229-2241.

385. Clevers H. The intestinal crypt, a prototype stem cell compartment. Cell. 2013;154:274-284.

386. Barker N. Adult intestinal stem cells: critical drivers of epithelial homeostasis and regeneration. Nat Rev Mol Cell Biol. 2014;15:19-33.

387. Gracz AD, Magness ST. Defining hierarchies of stemness in the intestine: evidence from biomarkers and regulatory pathways. Am J Physiol Gastrointest Liver Physiol. 2014;307:G260-273.

388. Chia LA, Kuo CJ. The intestinal stem cell. Prog Mol Biol Transl Sci. 2010;96:157- 173.

389. Van Landeghem L, Santoro MA, Krebs AE, Mah AT, Dehmer JJ, Gracz AD, Scull BP, McNaughton K, Magness ST, Lund PK. Activation of two distinct Sox9-EGFP- expressing intestinal stem cell populations during crypt regeneration after irradiation. Am J Physiol Gastrointest Liver Physiol. 2012;302:G1111-1132.

390. Barker N, van Es JH, Kuipers J, Kujala P, van den Born M, Cozijnsen M, Haegebarth A, Korving J, Begthel H, Peters PJ, et al. Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature. 2007;449:1003-1007.

391. Powell AE, Wang Y, Li Y, Poulin EJ, Means AL, Washington MK, Higginbotham JN, Juchheim A, Prasad N, Levy SE, et al. The pan-ErbB negative regulator Lrig1 is an intestinal stem cell marker that functions as a tumor suppressor. Cell. 2012;149:146- 158.

216

392. van der Flier LG, van Gijn ME, Hatzis P, Kujala P, Haegebarth A, Stange DE, Begthel H, van den Born M, Guryev V, Oving I, et al. Transcription factor achaete scute-like 2 controls intestinal stem cell fate. Cell. 2009;136:903-912.

393. van der Flier LG, Haegebarth A, Stange DE, van de Wetering M, Clevers H. OLFM4 is a robust marker for stem cells in human intestine and marks a subset of colorectal cancer cells. Gastroenterology. 2009;137:15-17.

394. Takeda N, Jain R, LeBoeuf MR, Wang Q, Lu MM, Epstein JA. Interconversion between intestinal stem cell populations in distinct niches. Science. 2011;334:1420- 1424.

395. Sangiorgi E, Capecchi MR. Bmi1 is expressed in vivo in intestinal stem cells. Nat Genet. 2008;40:915-920.

396. Formeister EJ, Sionas AL, Lorance DK, Barkley CL, Lee GH, Magness ST. Distinct SOX9 levels differentially mark stem/progenitor populations and enteroendocrine cells of the small intestine epithelium. Am J Physiol Gastrointest Liver Physiol. 2009;296:G1108-1118.

397. Scoville DH, Sato T, He XC, Li L. Current view: intestinal stem cells and signaling. Gastroenterology. 2008;134:849-864.

398. Barker N, Ridgway RA, van Es JH, van de Wetering M, Begthel H, van den Born M, Danenberg E, Clarke AR, Sansom OJ, Clevers H. Crypt stem cells as the cells-of- origin of intestinal cancer. Nature. 2009;457:608-611.

399. Knockout Mouse Project (KOMP) Respository.

400. Testa G, Schaft J, van der Hoeven F, Glaser S, Anastassiadis K, Zhang Y, Hermann T, Stremmel W, Stewart AF. A reliable lacZ expression reporter cassette for multipurpose, knockout-first alleles. Genesis. 2004;38:151-158.

401. Bradley A, Anastassiadis K, Ayadi A, Battey JF, Bell C, Birling MC, Bottomley J, Brown SD, Burger A, Bult CJ, et al. The mammalian gene function resource: the International Knockout Mouse Consortium. Mamm Genome. 2012;23:580-586.

402. Gracz AD, Ramalingam S, Magness ST. Sox9 expression marks a subset of CD24- expressing small intestine epithelial stem cells that form organoids in vitro. Am J Physiol Gastrointest Liver Physiol. 2010;298:G590-600.

217

403. Gracz AD, Williamson IA, Roche KC, Johnston MJ, Wang F, Wang Y, Attayek PJ, Balowski J, Liu XF, Laurenza RJ, et al. A high-throughput platform for stem cell niche co-cultures and downstream gene expression analysis. Nat Cell Biol. 2015;17:340- 349.

404. Roche KC, Gracz AD, Liu XF, Newton V, Akiyama H, Magness ST. SOX9 maintains reserve stem cells and preserves radioresistance in mouse small intestine. Gastroenterology. 2015;149:1553-1563 e1510.

405. Hoadley KA, Yau C, Wolf DM, Cherniack AD, Tamborero D, Ng S, Leiserson MD, Niu B, McLellan MD, Uzunangelov V, et al. Multiplatform analysis of 12 cancer types reveals molecular classification within and across tissues of origin. Cell. 2014;158:929-944.

406. The Cancer Genome Atlas (TCGA) Data Portal. National Institute of Health. National Cancer Institute. National Human Genome Research Institute.

407. Yan KS, Chia LA, Li X, Ootani A, Su J, Lee JY, Su N, Luo Y, Heilshorn SC, Amieva MR, et al. The intestinal stem cell markers Bmi1 and Lgr5 identify two functionally distinct populations. Proc Natl Acad Sci U S A. 2012;109:466-471.

408. Hua G, Thin TH, Feldman R, Haimovitz-Friedman A, Clevers H, Fuks Z, Kolesnick R. Crypt base columnar stem cells in small intestines of mice are radioresistant. Gastroenterology. 2012;143:1266-1276.

409. Moser AR, Pitot HC, Dove WF. A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse. Science. 1990;247:322-324.

410. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B, Kinzler KW. Activation of beta-catenin-Tcf signaling in colon cancer by mutations in beta-catenin or APC. Science. 1997;275:1787-1790.

411. Rask-Andersen M, Almen MS, Schioth HB. Trends in the exploitation of novel drug targets. Nat Rev Drug Discov. 2011;10:579-590.

412. Noah TK, Lo YH, Price A, Chen G, King E, Washington MK, Aronow BJ, Shroyer NF. SPDEF functions as a colorectal tumor suppressor by inhibiting beta-catenin activity. Gastroenterology. 2013;144:1012-1023 e1016.

413. Bossuyt W, Kazanjian A, De Geest N, Van Kelst S, De Hertogh G, Geboes K, Boivin GP, Luciani J, Fuks F, Chuah M, et al. Atonal homolog 1 is a tumor suppressor gene. PLoS Biol. 2009;7:e39.

218

414. Depeille P, Henricks LM, van de Ven RA, Lemmens E, Wang CY, Matli M, Werb Z, Haigis KM, Donner D, Warren R, et al. RasGRP1 opposes proliferative EGFR-SOS1- Ras signals and restricts intestinal epithelial cell growth. Nat Cell Biol. 2015;17:804- 815.

415. Nam KT, Lee HJ, Smith JJ, Lapierre LA, Kamath VP, Chen X, Aronow BJ, Yeatman TJ, Bhartur SG, Calhoun BC, et al. Loss of Rab25 promotes the development of intestinal neoplasia in mice and is associated with human colorectal adenocarcinomas. J Clin Invest. 2010;120:840-849.

416. Lee SH, Hu LL, Gonzalez-Navajas J, Seo GS, Shen C, Brick J, Herdman S, Varki N, Corr M, Lee J, et al. ERK activation drives intestinal tumorigenesis in Apc(min/+) mice. Nat Med. 2010;16:665-670.

417. Roberts RB, Min L, Washington MK, Olsen SJ, Settle SH, Coffey RJ, Threadgill DW. Importance of epidermal growth factor receptor signaling in establishment of adenomas and maintenance of carcinomas during intestinal tumorigenesis. Proc Natl Acad Sci U S A. 2002;99:1521-1526.

418. Wong VW, Stange DE, Page ME, Buczacki S, Wabik A, Itami S, van de Wetering M, Poulsom R, Wright NA, Trotter MW, et al. Lrig1 controls intestinal stem-cell homeostasis by negative regulation of ErbB signalling. Nat Cell Biol. 2012;14:401- 408.

419. Sansom OJ, Meniel V, Wilkins JA, Cole AM, Oien KA, Marsh V, Jamieson TJ, Guerra C, Ashton GH, Barbacid M, et al. Loss of Apc allows phenotypic manifestation of the transforming properties of an endogenous K-ras oncogene in vivo. Proc Natl Acad Sci U S A. 2006;103:14122-14127.

420. Dow LE, O'Rourke KP, Simon J, Tschaharganeh DF, van Es JH, Clevers H, Lowe SW. Apc Restoration Promotes Cellular Differentiation and Reestablishes Crypt Homeostasis in Colorectal Cancer. Cell. 2015;161:1539-1552.

421. Phelps RA, Chidester S, Dehghanizadeh S, Phelps J, Sandoval IT, Rai K, Broadbent T, Sarkar S, Burt RW, Jones DA. A two-step model for colon adenoma initiation and progression caused by APC loss. Cell. 2009;137:623-634.

422. Haigis KM, Kendall KR, Wang Y, Cheung A, Haigis MC, Glickman JN, Niwa- Kawakita M, Sweet-Cordero A, Sebolt-Leopold J, Shannon KM, et al. Differential effects of oncogenic K-Ras and N-Ras on proliferation, differentiation and tumor progression in the colon. Nat Genet. 2008;40:600-608.

219

423. Janssen KP, Alberici P, Fsihi H, Gaspar C, Breukel C, Franken P, Rosty C, Abal M, El Marjou F, Smits R, et al. APC and oncogenic KRAS are synergistic in enhancing Wnt signaling in intestinal tumor formation and progression. Gastroenterology. 2006;131:1096-1109.

424. Wilson CH, McIntyre RE, Arends MJ, Adams DJ. The activating mutation R201C in GNAS promotes intestinal tumourigenesis in Apc(Min/+) mice through activation of Wnt and ERK1/2 MAPK pathways. Oncogene. 2010;29:4567-4575.

425. Sato T, Vries RG, Snippert HJ, van de Wetering M, Barker N, Stange DE, van Es JH, Abo A, Kujala P, Peters PJ, et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature. 2009;459:262-265.

426. Grun D, Lyubimova A, Kester L, Wiebrands K, Basak O, Sasaki N, Clevers H, van Oudenaarden A. Single-cell messenger RNA sequencing reveals rare intestinal cell types. Nature. 2015;525:251-255.

427. McCracken KW, Cata EM, Crawford CM, Sinagoga KL, Schumacher M, Rockich BE, Tsai Y-H, Mayhew CN, Spence JR, Zavros Y, et al. Modelling human development and disease in pluripotent stem-cell-derived gastric organoids. Nature. 2014;516:400-404.

428. Spence JR, Mayhew CN, Rankin SA, Kuhar MF, Vallance JE, Tolle K, Hoskins EE, Kalinichenko VV, Wells SI, Zorn AM, et al. Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro. Nature. 2011;470:105-109.

429. Wells JM. Regenerative medicine in 2015: Generating and regenerating the digestive system. Nat Rev Gastroenterol Hepatol. 2016;13:65-66.

430. Dedhia PH, Bertaux-Skeirik N, Zavros Y, Spence JR. Organoid Models of Human Gastrointestinal Development and Disease. Gastroenterology. 2016;150:1098-1112.

431. Rookmaaker MB, Schutgens F, Verhaar MC, Clevers H. Development and application of human adult stem or progenitor cell organoids. Nat Rev Nephrol. 2015;11:546-554.

432. Schwank G, Clevers H. CRISPR/Cas9-Mediated Genome Editing of Mouse Small Intestinal Organoids. Methods Mol Biol. 2016;1422:3-11.

433. Barker N, Huch M, Kujala P, van de Wetering M, Snippert HJ, van Es JH, Sato T, Stange DE, Begthel H, van den Born M, et al. Lgr5(+ve) stem cells drive self-renewal

220

in the stomach and build long-lived gastric units in vitro. Cell Stem Cell. 2010;6:25- 36.

434. Mills JC, Andersson N, Hong CV, Stappenbeck TS, Gordon JI. Molecular characterization of mouse gastric epithelial progenitor cells. Proc Natl Acad Sci U S A. 2002;99:14819-14824.

435. Giannakis M, Stappenbeck TS, Mills JC, Leip DG, Lovett M, Clifton SW, Ippolito JE, Glasscock JI, Arumugam M, Brent MR, et al. Molecular properties of adult mouse gastric and intestinal epithelial progenitors in their niches. J Biol Chem. 2006;281:11292-11300.

436. Arnold K, Sarkar A, Yram MA, Polo JM, Bronson R, Sengupta S, Seandel M, Geijsen N, Hochedlinger K. Sox2(+) adult stem and progenitor cells are important for tissue regeneration and survival of mice. Cell Stem Cell. 2011;9:317-329.

437. Hayakawa Y, Ariyama H, Stancikova J, Sakitani K, Asfaha S, Renz BW, Dubeykovskaya ZA, Shibata W, Wang H, Westphalen CB, et al. Mist1 Expressing Gastric Stem Cells Maintain the Normal and Neoplastic Gastric Epithelium and Are Supported by a Perivascular Stem Cell Niche. Cancer Cell. 2015;28:800-814.

438. Stange DE, Koo BK, Huch M, Sibbel G, Basak O, Lyubimova A, Kujala P, Bartfeld S, Koster J, Geahlen JH, et al. Differentiated Troy+ chief cells act as reserve stem cells to generate all lineages of the stomach epithelium. Cell. 2013;155:357-368.

439. Demitrack ES, Gifford GB, Keeley TM, Carulli AJ, VanDussen KL, Thomas D, Giordano TJ, Liu Z, Kopan R, Samuelson LC. Notch signaling regulates gastric antral LGR5 stem cell function. EMBO J. 2015;34:2522-2536.

440. Gifford GB, Demitrack ES, Keeley TM, Tam A, La Cunza N, Dedhia PH, Spence JR, Simeone DM, Saotome I, Louvi A, et al. Notch1 and Notch2 receptors regulate mouse and human gastric antral epithelial cell homoeostasis. Gut. 2016

441. Westphalen CB, Asfaha S, Hayakawa Y, Takemoto Y, Lukin DJ, Nuber AH, Brandtner A, Setlik W, Remotti H, Muley A, et al. Long-lived intestinal tuft cells serve as colon cancer-initiating cells. J Clin Invest. 2014;124:1283-1295.

442. May R, Qu D, Weygant N, Chandrakesan P, Ali N, Lightfoot SA, Li L, Sureban SM, Houchen CW. Brief report: Dclk1 deletion in tuft cells results in impaired epithelial repair after radiation injury. Stem Cells. 2014;32:822-827.

221

443. van Es JH, Sato T, van de Wetering M, Lyubimova A, Nee AN, Gregorieff A, Sasaki N, Zeinstra L, van den Born M, Korving J, et al. Dll1+ secretory progenitor cells revert to stem cells upon crypt damage. Nat Cell Biol. 2012;14:1099-1104.

444. Buczacki SJ, Zecchini HI, Nicholson AM, Russell R, Vermeulen L, Kemp R, Winton DJ. Intestinal label-retaining cells are secretory precursors expressing Lgr5. Nature. 2013;495:65-69.

445. Tetteh PW, Basak O, Farin HF, Wiebrands K, Kretzschmar K, Begthel H, van den Born M, Korving J, de Sauvage F, van Es JH, et al. Replacement of Lost Lgr5- Positive Stem Cells through Plasticity of Their Enterocyte-Lineage Daughters. Cell Stem Cell. 2016;18:203-213.

446. Srinivasan T, Than EB, Bu P, Tung KL, Chen KY, Augenlicht L, Lipkin SM, Shen X. Notch signalling regulates asymmetric division and inter-conversion between lgr5 and bmi1 expressing intestinal stem cells. Sci Rep. 2016;6:26069.

447. Metcalfe C, Kljavin NM, Ybarra R, de Sauvage FJ. Lgr5+ stem cells are indispensable for radiation-induced intestinal regeneration. Cell Stem Cell. 2014;14:149-159.

448. Mah AT, Van Landeghem L, Gavin HE, Magness ST, Lund PK. Impact of diet- induced obesity on intestinal stem cells: hyperproliferation but impaired intrinsic function that requires insulin/IGF1. Endocrinology. 2014;155:3302-3314.

449. Santoro MA, Blue RE, Andres SF, Mah AT, Van Landeghem L, Lund PK. Obesity and intestinal epithelial deletion of the insulin receptor, but not the IGF 1 receptor, affect radiation-induced apoptosis in colon. Am J Physiol Gastrointest Liver Physiol. 2015;309:G578-589.

450. Van Landeghem L, Santoro MA, Mah AT, Krebs AE, Dehmer JJ, McNaughton KK, Helmrath MA, Magness ST, Lund PK. IGF1 stimulates crypt expansion via differential activation of 2 intestinal stem cell populations. FASEB J. 2015;29:2828- 2842.

451. Andres SF, Santoro MA, Mah AT, Keku JA, Bortvedt AE, Blue RE, Lund PK. Deletion of intestinal epithelial insulin receptor attenuates high-fat diet-induced elevations in cholesterol and stem, enteroendocrine, and Paneth cell mRNAs. Am J Physiol Gastrointest Liver Physiol. 2015;308:G100-111.

452. VanDussen KL, Carulli AJ, Keeley TM, Patel SR, Puthoff BJ, Magness ST, Tran IT, Maillard I, Siebel C, Kolterud A, et al. Notch signaling modulates proliferation and

222

differentiation of intestinal crypt base columnar stem cells. Development. 2012;139:488-497.

453. Carulli AJ, Keeley TM, Demitrack ES, Chung J, Maillard I, Samuelson LC. Notch receptor regulation of intestinal stem cell homeostasis and crypt regeneration. Dev Biol. 2015;402:98-108.

454. Gregorieff A, Liu Y, Inanlou MR, Khomchuk Y, Wrana JL. Yap-dependent reprogramming of Lgr5(+) stem cells drives intestinal regeneration and cancer. Nature. 2015;526:715-718.

455. Barry ER, Morikawa T, Butler BL, Shrestha K, de la Rosa R, Yan KS, Fuchs CS, Magness ST, Smits R, Ogino S, et al. Restriction of intestinal stem cell expansion and the regenerative response by YAP. Nature. 2013;493:106-110.

456. Gerbe F, Sidot E, Smyth DJ, Ohmoto M, Matsumoto I, Dardalhon V, Cesses P, Garnier L, Pouzolles M, Brulin B, et al. Intestinal epithelial tuft cells initiate type 2 mucosal immunity to helminth parasites. Nature. 2016;529:226-230.

457. von Moltke J, Ji M, Liang HE, Locksley RM. Tuft-cell-derived IL-25 regulates an intestinal ILC2-epithelial response circuit. Nature. 2016;529:221-225.

458. Lindemans CA, Calafiore M, Mertelsmann AM, O'Connor MH, Dudakov JA, Jenq RR, Velardi E, Young LF, Smith OM, Lawrence G, et al. Interleukin-22 promotes intestinal-stem-cell-mediated epithelial regeneration. Nature. 2015;528:560-564.

459. Mills JC, Sansom OJ. Reserve stem cells: Differentiated cells reprogram to fuel repair, metaplasia, and neoplasia in the adult gastrointestinal tract. Sci Signal. 2015;8:re8.

460. Huels DJ, Sansom OJ. Stem vs non-stem cell origin of colorectal cancer. Br J Cancer. 2015;113:1-5.

461. Vermeulen L, Morrissey E, van der Heijden M, Nicholson AM, Sottoriva A, Buczacki S, Kemp R, Tavare S, Winton DJ. Defining stem cell dynamics in models of intestinal tumor initiation. Science. 2013;342:995-998.

462. Kozar S, Morrissey E, Nicholson AM, van der Heijden M, Zecchini HI, Kemp R, Tavare S, Vermeulen L, Winton DJ. Continuous clonal labeling reveals small numbers of functional stem cells in intestinal crypts and adenomas. Cell Stem Cell. 2013;13:626-633.

223

463. Schepers AG, Snippert HJ, Stange DE, van den Born M, van Es JH, van de Wetering M, Clevers H. Lineage tracing reveals Lgr5+ stem cell activity in mouse intestinal adenomas. Science. 2012;337:730-735.

464. Hung KE, Maricevich MA, Richard LG, Chen WY, Richardson MP, Kunin A, Bronson RT, Mahmood U, Kucherlapati R. Development of a mouse model for sporadic and metastatic colon tumors and its use in assessing drug treatment. Proc Natl Acad Sci U S A. 2010;107:1565-1570.

465. Marsh V, Winton DJ, Williams GT, Dubois N, Trumpp A, Sansom OJ, Clarke AR. Epithelial Pten is dispensable for intestinal homeostasis but suppresses adenoma development and progression after Apc mutation. Nat Genet. 2008;40:1436-1444.

466. Beyaz S, Mana MD, Roper J, Kedrin D, Saadatpour A, Hong SJ, Bauer-Rowe KE, Xifaras ME, Akkad A, Arias E, et al. High-fat diet enhances stemness and tumorigenicity of intestinal progenitors. Nature. 2016;531:53-58.

467. Yu FX, Zhao B, Panupinthu N, Jewell JL, Lian I, Wang LH, Zhao J, Yuan H, Tumaneng K, Li H, et al. Regulation of the Hippo-YAP pathway by G-protein-coupled receptor signaling. Cell. 2012;150:780-791.

468. Nakanishi Y, Seno H, Fukuoka A, Ueo T, Yamaga Y, Maruno T, Nakanishi N, Kanda K, Komekado H, Kawada M, et al. Dclk1 distinguishes between tumor and normal stem cells in the intestine. Nat Genet. 2013;45:98-103.

469. McGrath KE, Palis J. Hematopoiesis in the yolk sac: more than meets the eye. Exp Hematol. 2005;33:1021-1028.

470. Mikkola HK, Orkin SH. The journey of developing hematopoietic stem cells. Development. 2006;133:3733-3744.

471. Rowe RG, Mandelbaum J, Zon LI, Daley GQ. Engineering Hematopoietic Stem Cells: Lessons from Development. Cell Stem Cell. 2016;18:707-720.

472. Chen MJ, Yokomizo T, Zeigler BM, Dzierzak E, Speck NA. Runx1 is required for the endothelial to haematopoietic cell transition but not thereafter. Nature. 2009;457:887- 891.

473. Kissa K, Herbomel P. Blood stem cells emerge from aortic endothelium by a novel type of cell transition. Nature. 2010;464:112-115.

224

474. Boisset JC, van Cappellen W, Andrieu-Soler C, Galjart N, Dzierzak E, Robin C. In vivo imaging of haematopoietic cells emerging from the mouse aortic endothelium. Nature. 2010;464:116-120.

475. Swiers G, Baumann C, O'Rourke J, Giannoulatou E, Taylor S, Joshi A, Moignard V, Pina C, Bee T, Kokkaliaris KD, et al. Early dynamic fate changes in haemogenic endothelium characterized at the single-cell level. Nat Commun. 2013;4:2924.

476. Sansom OJ, Meniel VS, Muncan V, Phesse TJ, Wilkins JA, Reed KR, Vass JK, Athineos D, Clevers H, Clarke AR. Myc deletion rescues Apc deficiency in the small intestine. Nature. 2007;446:676-679.

477. Ueo T, Imayoshi I, Kobayashi T, Ohtsuka T, Seno H, Nakase H, Chiba T, Kageyama R. The role of Hes genes in intestinal development, homeostasis and tumor formation. Development. 2012;139:1071-1082.

478. van Es JH, van Gijn ME, Riccio O, van den Born M, Vooijs M, Begthel H, Cozijnsen M, Robine S, Winton DJ, Radtke F, et al. Notch/gamma-secretase inhibition turns proliferative cells in intestinal crypts and adenomas into goblet cells. Nature. 2005;435:959-963.

479. Luo J, Emanuele MJ, Li D, Creighton CJ, Schlabach MR, Westbrook TF, Wong KK, Elledge SJ. A genome-wide RNAi screen identifies multiple synthetic lethal interactions with the Ras oncogene. Cell. 2009;137:835-848.

480. Graham GJ, Locati M, Mantovani A, Rot A, Thelen M. The biochemistry and biology of the atypical chemokine receptors. Immunol Lett. 2012;145:30-38.

225