Mechanisms of vascular disease: divergent roles for suppressor of cytokine signaling 3 in angiotensin IIinduced vascular dysfunction

Transcription

1 University of Iowa Iowa Research Online Theses and Dissertations Fall 2014 Mechanisms of vascular disease: divergent roles for suppressor of cytokine signaling 3 in angiotensin IIinduced vascular dysfunction Ying Li University of Iowa Copyright 2014 Ying Li This dissertation is available at Iowa Research Online: Recommended Citation Li, Ying. "Mechanisms of vascular disease: divergent roles for suppressor of cytokine signaling 3 in angiotensin II-induced vascular dysfunction." PhD (Doctor of Philosophy) thesis, University of Iowa, Follow this and additional works at: Part of the Pharmacology Commons

2 MECHANISMS OF VASCULAR DISEASE: DIVERGENT ROLES FOR SUPPRESSOR OF CYTOKINE SIGNALING 3 IN ANGIOTENSIN II-INDUCED VASCULAR DYSFUNCTION by Ying Li A thesis submitted in partial fulfillment of the requirements for the Doctor of Philosophy degree in Pharmacology in the Graduate College of The University of Iowa December 2014 Thesis Supervisor: Professor Frank M. Faraci

3 Copyright by YING LI 2014 All Rights Reserved

4 Graduate College The University of Iowa Iowa City, Iowa CERTIFICATE OF APPROVAL PH.D. THESIS This is to certify that the Ph.D. thesis of Ying Li has been approved by the Examining Committee for the thesis requirement for the Doctor of Philosophy degree in Pharmacology at the December 2014 graduation. Thesis Committee: Frank M. Faraci, Thesis Supervisor Donald D. Heistad Kathryn G. Lamping Kamal Rahmouni Curt D. Sigmund

5 ABSTRACT Angiotensin II (Ang II) promotes vascular disease and hypertension, in part, by activating the interleukin-6 (IL-6)/signal transducer and activator of transcription 3 (STAT3) signaling pathway. Extensive studies have demonstrated that SOCS3 plays an important role in suppressing the IL-6/STAT3 pathway in the immune system and in cancer biology. In contrast, the functional importance of SOCS3 in cardiovascular disease is largely unknown. Thus, the overall goal of these studies was to investigate the role of SOCS3 in models of Ang II-dependent vascular disease and hypertension. To examine direct effects of Ang II on the vessel wall, carotid arteries from SOCS3 haplodeficient (SOCS3 +/- ) mice and wild-type littermates (SOCS3 +/+ ) were incubated with the peptide or vehicle for 22 hrs, followed by examination of endothelial function using acetylcholine. Relaxation to acetylcholine was similar in all arteries incubated with vehicle. A low concentration of Ang II (1 nmol/l) did not affect acetylcholine-induced vasodilation in SOCS3 +/+ mice, but reduced responses in arteries from SOCS3 +/- mice by ~50% (P<0.05). This Ang II-induced endothelial dysfunction in SOCS3 +/- mice was prevented by inhibitors of NF-κB or STAT3, an IL-6 neutralizing antibody, or a scavenger of superoxide. Responses to nitroprusside, an endothelium-independent vasodilator, were similar in all groups. To test the importance of SOCS3 in vivo, mice were infused systemically with a pressor dose of Ang II (1.4 mg/kg per day) or vehicle for 14 days via osmotic mini-pumps. Acetylcholine-induced vasodilation in carotid and resistance arteries in brain from SOCS3 +/- mice was reduced by ~60% (P<0.05). Surprisingly, genetic deficiency in SOCS3 prevented the majority of Ang II-induced endothelial dysfunction without affecting the pressor response to Ang II. To investigate potential mechanisms underlying divergent results when studying effects of local versus systemic effects of Ang II, we performed bone marrow ii

6 transplantation followed by infusion of vehicle or Ang II for two weeks. Lethally irradiated WT (CD45.1) mice reconstituted with SOCS3 +/- bone marrow were protected from Ang IIinduced endothelial dysfunction (P<0.05), while reconstitution of irradiated SOCS3 +/- mice with WT (CD45.1) bone marrow exacerbated Ang II-induced vascular dysfunction (P<0.05). WT (CD45.1) into SOCS3 +/+ and SOCS3 +/- into SOCS3 +/- bone marrow chimeras exhibited vascular function consistent with non-irradiated controls. In addition, the pressor response to Ang II was reduced by ~50% in WT mice reconstituted with bone marrow from SOCS3 +/- mice (P<0.05). These data suggest that SOCS3 exerts divergent or context-dependent effects depending on whether vascular dysfunction was due to local versus systemic administration of Ang II. SOCS3 deficiency in the vessel wall enhanced local detrimental effects of Ang II on vascular function. In contrast, bone marrow-derived cells that are haplodeficient in SOCS3 protect against systemically administered Ang II and the resulting vascular dysfunction and hypertension. To my knowledge, these are the first experimental studies that begin to define the importance of SOCS3 in Ang II-induced hypertension and endothelial dysfunction. Results obtained from these experiments provide new insight into mechanisms which regulate oxidative stress and inflammation within the vasculature. The studies also revealed that bone marrow-derived cells that are haplodeficient in SOCS3 protect against pressor and endothelial effects of Ang II. These findings may eventually contribute to the development of novel therapeutic approaches for hypertension and hypertension associated end-organ damage. iii

7 PUBLIC ABSTRACT Hypertension is the leading risk factor for cardiovascular disease and stroke, which ranks as the number one and number four causes of death in the United States, respectively. Although hypertension is a common health issue worldwide, the pathophysiology of its development and mechanisms that underlie subsequent end-organ damage have not been fully defined. My thesis project was aimed at investigating whether an inhibitory protein in the immune system plays an important role in the development and progression of high blood pressure and vascular disease. The project also aimed to understand how this protein may promote or protects against this disease condition. Results obtained from a series of studies provided evidence which strongly support the concept that this specific protein exerts functionally important effects in animal models of hypertension and vascular disease. Using a genetically engineered mouse model which lacks this specific protein, my data suggests the protein plays a divergent role in the vasculature versus the whole body. In isolated arteries, lack of this protein augmented vascular dysfunction induced by angiotensin II, a peptide that is a major contributor to hypertension and vascular disease. Surprisingly, studies conducted in whole animals revealed that lack of this same protein had a protective role in response to the same stimuli. Lastly, I obtained evidence that bone marrow-derived cells were responsible for these protective effects. The present study defined the role of a protein named suppressor of cytokine signaling 3 in vascular disease for the first time. The work may provide insight into new strategies to protect against vascular disease, hypertension, and subsequent clinical events. iv

8 TABLE OF CONTENTS LIST OF FIGURES... viii LIST OF ABBREVIATIONS... xi CHAPTER 1. INTRODUCTION...1 Epidemiology of Hypertension and Cardiovascular Disease...2 The Circulatory System and Endothelial Function...2 The Renin-Angiotensin System...6 Oxidative Stress and Inflammation...11 The Immune System...17 Interleukin-6 Signaling and Suppressor of Cytokine Signaling Thesis Focus...27 CHAPTER 2. SOCS3 PROTECTS AGAINST DIRECT EFFECTS OF ANGIOTENSIN II ON ENDOTHELIAL FUNCTION...31 Introduction...32 Methods...33 Experimental Animals...33 Genotyping of SOCS3- Haplodeficient Mice by PCR...34 Incubation of Blood Vessels...34 Vascular Function...35 Quantitative Real-time RT-PCR...36 Statistical Analysis...36 Results...37 Effects of Ang II on Endothelial Function Are Mediated Through AT1R Signaling...37 Direct Effects of Ang II on Endothelial Function Are Augmented by SOCS3 Haplodeficiency...37 Ang II-Induced Impairment of Endothelial Function in SOCS3 Haplodeficient Mice Is Inhibited by Suppressing Oxidative- or Inflammatory-Related Signaling...38 Vascular Expression of Select Genes Implicated in Vascular Disease...39 Vasoconstrictor Responses to Ang II...40 Discussion...41 Local Effects of Ang II on Vascular Function Are Augmented by Genetic Haplodeficiency in SOCS Genetic Haplodeficiency in SOCS3 Promotes Oxidative Stress in Response to Ang II...43 Evidence for Increased Inflammation-Related Signaling in Arteries from SOCS3 Haplodeficient Mice Treated with Ang II...45 Effect of SOCS3 Haplodeficiency on Vasoconstrictor Responses...46 CHAPTER 3. SOCS3 HAPLODEFICIENCY PROTECTS MICE FROM ENDOTHELIAL DYSFUNCTION PRODUCED BY SYSTEMIC ADMINISTRATION OF ANGIOTENSIN II...65 v

9 Introduction...66 Methods...68 Animals...68 Chronic Angiotensin II-Dependent Hypertension...68 Vasomotor Function...69 Blood Pressure Measurements...70 Real-Time RT-PCR...70 Statistical Analysis...70 Results...71 SOCS3 Haplodeficiency Protected against Endothelial Dysfunction in Carotid Arteries during Chronic Ang II-Dependent Hypertension...71 Protective Effects of SOCS3 Haplodeficiency Extend to Resistance Arteries in the Brain...71 SOCS3 Haplodeficiency Did Not Alter the Pressor Response during Ang II-Induced Hypertension...72 SOCS3 Haplodeficiency Does Not Alter Endothelial Function or Blood Pressure during Infusion of a Non-Pressor Dose of Ang II...72 Effects of Systemic Ang II on Vascular Expression of Genes Implicated in Vascular Disease...73 Discussion...73 Role of SOCS3 in an Angiotensin II-Dependent Model of Hypertension...74 Haplodeficiency in SOCS3 Protects Against Systemic Angiotensin II-Induced Endothelial Dysfunction...75 CHAPTER 4. SOCS3 HAPLODEFICIENT BONE MARROW-DERIVED CELLS PROTECT AGAINST ANGIOTENSIN II-INDUCED VASCULAR DYSFUNCTION AND HYPERTENSION...89 Introduction...90 Methods...93 Animals...93 Bone Marrow Chimeric Mice...93 Fluorescence Activated Cell Sorting (FACS) Analysis...94 Chronic Angiotensin II-Dependent Hypertension...95 Vasomotor Function...95 Blood Pressure Measurements...95 Incubation of Blood Vessels...96 Statistical Analysis...96 Results...96 Bone Marrow-Derived Cells Haplodeficient in SOCS3 Protect Against Ang II-Induced Endothelial Dysfunction...96 Bone Marrow-Derived Cells Haplodeficient in SOCS3 Reduced the Pressor Response to Ang II...98 Direct Effects of Ang II on the Vessel Wall Are Not Affected by Bone Marrow-Derived Cells...98 Discussion...99 Hematopoietic SOCS3 Haplodeficient Cells Protect Against Ang II-Dependent Endothelial Dysfunction and Hypertension...99 vi

10 Local Vascular Effects of Ang II in the Hematopoietic SOCS3 Haplodeficient Model CHAPTER 5. CONCLUSIONS The Role of SOCS3 in Local Effects of Ang II on Vascular Cells The Role of SOCS3 in Effects of Ang II on Endothelial Cells The Role of SOCS3 in the Effects of Ang II on Vascular Smooth Muscle Cells The Role of SOCS3 in Effects of Ang II on Adventitia Cells The Role of SOCS3 in Systemic Effects of Ang II in Non-Vascular Cells The Role of SOCS3 in Effects of Ang II on Bone Marrow-Derived Cells The Potential Role of the Central Nervous System in Systemic Effects of Ang II The Potential Role of Renal SOCS3 in Mediating Systemic Effects of Ang II Future Studies REFERENCES vii

11 LIST OF FIGURES Figure 1. The role of SOCS3 in angiotensin II-induced vascular dysfunction Figure 2. Angiotensin II type 1 receptor mediated local effects of angiotensin II on endothelial function Figure 3. Effects of SOCS3 haplodeficiency on local angiotensin II-induced endothelial dysfunction Figure 4. Endothelium-independent relaxation was not affected by angiotensin II or SOCS3 haplodeficiency Figure 5. Effects of SOCS3 haplodeficiency on vasoconstriction...51 Figure 6. Angiotensin II-induced endothelial dysfunction in SOCS3 +/- arteries was mediated by oxidative stress Figure 7. Endothelium-independent relaxation was not affected by angiotensin II or superoxide Figure 8. The role of NF-κB in angiotensin II-induced endothelial dysfunction in arteries from SOCS3 +/- mice Figure 9. Endothelium-independent relaxation is not affected by angiotensin II or inhibition of NF-κB Figure 10. Angiotensin II-induced endothelial dysfunction was mediated by IL Figure 11. Endothelium-independent relaxation of arteries was not affected by angiotensin II or IL Figure 12. Angiotensin II-induced endothelial dysfunction in SOCS3 +/- arteries was dependent on activation of STAT Figure 13. Endothelium-independent relaxation was not affected by angiotensin II or inhibition of STAT Figure 14. Vascular expression of select genes implicated in vascular disease Figure 15. Effects of SOCS3 haplodeficiency on angiotensin II-induced contraction in abdominal aorta and iliac artery Figure 16. Effects of SOCS3 haplodeficiency on angiotensin II-induced contraction in femoral and changes in vessel diameter in cerebral arteries Figure 17. Effects of SOCS3 haplodeficiency on U46619-induced contraction in abdominal aorta and iliac artery viii

12 Figure 18. Effects of SOCS3 haplodeficiency on U46619-induced contraction in femoral and changes in diameter in cerebral arteries Figure 19. Endothelium-dependent relaxation in chronic angiotensin II-dependent hypertension Figure 20. Endothelium-independent relaxation was not significantly affected by chronic angiotensin II-dependent hypertension or tempol Figure 21. The role of SOCS3 in contraction of arteries in a chronic angiotension IIdependent hypertension model...81 Figure 22. The protective role of haplodeficiency in SOCS3 extends to basilar arteries Figure 23. The role of SOCS3 in endothelium-independent vasodilation in basilar arteries...83 Figure 24. The role of SOCS3 in angiotensin II-induced hypertension Figure 25. Systolic blood pressure in SOCS3 +/+ and SOCS3 +/- mice administrated a nonpressor dose of angiotensin II...85 Figure 26. Endothelial-dependent relaxation was not affected by a non-pressor dose of angiotensin II or superoxide in SOCS3 +/+ and SOCS3 +/- mice Figure 27. Endothelial-independent relaxation was not affected by a non-pressor dose of angiotensin II or superoxide in SOCS3 +/+ and SOCS3 +/- mice Figure 28. Vascular expression of select genes implicated in vascular disease Figure 29. Experimental design of bone marrow transplantation Figure 30. Reconstitution of irradiated SOCS3 +/- mice with WT (CD45.2) bone marrow was confirmed by fluorescence activated cell sorting analysis Figure 31. Reconstitution of irradiated WT (CD45.1) mice with SOCS3 +/- bone marrow was confirmed by fluorescence activated cell sorting analysis Figure 32. Reconstitution of irradiated SOCS3 +/+ mice with WT (CD45.1) bone marrow was confirmed by fluorescence activated cell sorting analysis Figure 33. Endothelial function of arteries from WT (CD45.1) to WT and SOCS3 +/- to SOCS3 +/- bone marrow chimeras in chronic angiotensin II-dependent hypertension were consistent with non-irradiated controls Figure 34. Endothelial-independent relaxation of arteries from WT (CD45.1) to WT and SOCS3 +/- to SOCS3 +/- bone marrow chimeras were consistent with nonirradiated controls Figure 35. Contraction of arteries from WT (CD45.1) to WT and SOCS3 +/- to SOCS3 +/- bone marrow chimeras were consistent with non-irradiated controls ix

13 Figure 36. Endothelial function of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras in chronic angiotensin IIdependent hypertension Figure 37. Endothelial-independent relaxation of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras in chronic angiotensin II-dependent hypertension Figure 38. Contraction of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras in chronic angiotensin II-dependent hypertension Figure 39. Vascular contractile responses were not affected by bone marrow transplantation or systemic administration of angiotensin II Figure 40. Maximum relaxation to acetylcholine in carotid arteries from bone marrow chimeras Figure 41. Systolic blood pressure of bone marrow chimeras administrated with angiotensin II Figure 42. Endothelial function in arteries from WT (CD45.1) to SOCS3 +/+ and SOCS3 +/- to SOCS3 +/- bone marrow chimeras incubated with angiotensin II Figure 43. Endothelial-independent relaxation of arteries from WT (CD45.1) to SOCS3 +/+ and SOCS3 +/- to SOCS3 +/- bone marrow chimeras incubated with angiotensin II Figure 44. Contraction of arteries from WT (CD45.1) to SOCS3 +/+ and SOCS3 +/- to SOCS3 +/- bone marrow chimeras incubated with angiotensin II Figure 45. Endothelial function of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras incubated with angiotensin II Figure 46. Endothelial-independent relaxation of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras incubated with angiotensin II Figure 47. Contraction of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras incubated with angiotensin II Figure 48. Summaries of working model Figure 49. Ang II-Dependent Signaling Targets Multiple Tissues/Cells x

14 LIST OF ABBREVIATIONS ACE Ang II Apo E AT1R AT2R ATP B cell DAMP DOCA enos ERK angiotensin-converting enzyme angiotensin II Apolipoprotein E angiotensin II type 1 receptor angiotensin II type 2 receptor adenosine triphosphate B lymphocyte damage associated molecular pattern deoxycorticosterone acetate endothelial nitric oxide synthase extracellular signal-regulated protein kinase gp130 glycoprotein 130 IκB IL-1β inhibitor of kappa B interleukin 1β IL-6 interleukin 6 IL-10 interleukin 10 IL-12 interleukin12 IL-17 interleukin 17 JAK Janus kinase MyD88 myeloid differentiation primary response gene 88 NADPH MAPK NF-κB nicotinamide adenine dinucleotide phosphate reduced mitogen activated protein kinase nuclear factor kappa B xi

15 NO NOX PAMP RAS RhoA ROS nitric oxide NADPH oxidase pathogen-associated molecular pattern renin-angiotensin system Ras homolog gene family, member A reactive oxygen species SH2 Src homology 2 SOD superoxide dismutase SOCS3 suppressor of cytokine signaling 3 Src STAT3 T cell TNFα Treg VCSMs sarcoma proto-oncogene signal transducer and activator of transcription T lymphocyte tumor necrosis factor α regulatory T cell vascular smooth muscle cells xii

16 1 CHAPTER 1 INTRODUCTION

17 2 Epidemiology of Hypertension and Cardiovascular Disease Cardiovascular disease and stroke rank as the number one and number four causes of death in the United States, respectively. The leading risk factor for these diseases is hypertension (1-4), which results in an enormous clinical and economic burden in western societies (1-4). It is estimated that more than one third of the population has hypertension globally, while another third are in a pre-hypertension stage and will commonly develop overt hypertension within a few years (2, 5). Despite the development and availability of a large number of anti-hypertensive treatments, the prevalence of uncontrolled hypertension continues to rise globally (6). Although hypertension is a common health issue worldwide, the pathophysiology of its development and mechanisms that underlie subsequent end-organ damage have not been fully revealed. For some time, it has been well accepted that the regulation of blood pressure involves an integration of multiple organ systems (including the kidney, the heart, the vasculature, and the central nervous system). More recently, the immune system has been identified as a key additional mediator, contributing to the development of hypertension and end-organ damage (7-10). The Circulatory System and Endothelial Function A major function of the circulatory system is to transport oxygen, glucose and other nutrients to cells, while removing metabolic end-products like carbon dioxide as well as distributing hormones throughout the body. Blood pumped from the heart travels continuously through blood vessels forming two separate vascular circuits named the pulmonary and systemic circulations. While the pulmonary circulation transports blood between the heart and lungs for oxygen uptake and removal of carbon dioxide respectively, the systemic circulation distributes blood to all systemic organs (11-13).

18 3 The vascular tree consists of five major types of blood vessels: arteries, arterioles, capillaries, venules, and veins. Arteries deliver blood from the heart to organs in different regions of the body, which progressively branch into small arteries. Small arteries branch into numerous arterioles when reaches the organ it supplies. Arterioles branch further into capillaries, which are the smallest vessels where gas and nutrient exchange occurs between blood and surrounding cells. After they terminate, capillaries rejoin to form small venules followed by small veins. Small veins then leave the organ and progressively converge to form larger veins which carry blood back to the heart. The arterioles, capillaries, and venules are referred to as the microcirculatory vessels which all located within the organs (11-13). The physiological function of arteries and arterioles is conducting blood from the heart to the various organs. In addition, arteries also serve as a pressure reservoir to continuously drive blood flow forward when the heart is relaxing and filling. In general, small arteries and arterioles are the major resistance vessels and play important roles in regulating blood pressure and blood flow within the organs through adjusting their radius (11-13). Arteries and arterioles share a similar general structure. The outermost layer is known as the adventitia, which is composed of connective tissue, nerve endings, fibroblasts and quiescent resident leukocytes. Adventitia can influence vascular function by facilitating local inflammatory and oxidant-dependent processes, thus influencing vascular proliferation, neurotransmitter release from autonomic nerves, and signaling by vascular messenger molecules including nitric oxide (NO) (14-16). Within the adventitia is the media, which mainly consists of vascular smooth muscle cells (VCSMs). VCSMs are responsible for generation of vascular tone (vasoconstriction), which is controlled by several factors including the nervous system, mechanical stimuli (pressure or flow), circulating humoral molecules, as well as autocrine and paracrine factors. The innermost

19 4 layer of the vessel wall contains a monolayer of endothelial cells attached to a thin extracellular matrix (17). Unlike larger arteries, the arteriolar wall contains very little adventitia, and only one to three layers of vascular muscle (18). Endothelial cells line the entire circulation system, make direct contact with blood, and affect function of both vascular and non-vascular cells. Endothelium influences blood pressure (19) and plays a major role in control of local blood flow and vascular permeability by releasing multiple vasoactive mediators, including endothelial-derived relaxing and contracting factors, such as endothelin, superoxide anion, a variety of prostanoids, and locally generated Ang II (20-24). In 1980, Furchgott and Zawadzki first demonstrated that endothelial cells play an obligatory role in acetylcholine-induced vasodilation by producing a powerful vasodilator substance, which later was identified as NO (25-27). In response to diverse stimuli, such as acetylcholine, shear stress, and bradykinin, the intracellular calcium (Ca 2+ ) levels are increased, resulting in activation of NO synthase (NOS) in endothelial cells (enos). After being formed by activated NOS from the amino acid L-arginine, NO diffuses into adjacent VCSMs, where it activates guanylyl cyclase, an enzyme that catalyzes the dephosphorylation of guanosine triphosphate to cyclic guanidine monophosphate. Increased cyclic guanidine monophosphate decreases intracellular Ca 2+ level and causes relaxation of VCSMs (28-31). Besides NO, endothelium-dependent relaxation can be induced by other mechanisms, such as release of prostacyclin or endothelium-derived hyperpolarizing factor(s), or via direct endothelium-dependent hyperpolarization. The latter response involves the transmission of hyperpolarization from one cell to the next via gap junctions, and not the extracellular release of specific factors (32, 33). Through different signaling pathways, each of these mechanisms ultimately reduces intracellular Ca 2+ and promotes

20 5 VCSMs to relax (34). The predominant factor or mechanism that produces endotheliumdependent vasodilation can differ depending on the tissue or vessel type as well as the site of the vessel within the vascular tree (34). Previously, using both genetic and pharmacological approaches, our laboratory and other groups have shown that production of NO by enos is a major regulator of vascular tone in aorta, carotid arteries and in cerebral arteries and arterioles under normal conditions (35-38). Endothelial dysfunction is a broad term that can be used to reflect changes in the influence of endothelial cells on vascular tone, vascular permeability, and vascular growth as well as thrombosis. In relation to regulation of vascular tone, endothelial dysfunction is typically defined by reductions in the production or bioavailability of vasodilators, and/or increased production of endothelial-derived contracting factors, such endothelin-1, vasoconstrictor prostanoids, or locally generated Ang II (21, 24, 39). Clinically, endothelial dysfunction is a hallmark of vascular dysfunction, associated with multiple cardiovascular risk factors, such as aging, hypercholesterolemia, hypertension, smoking, hyperglycemia, and obesity (40, 41). Therefore endothelial dysfunction is regarded as a strong predictor of future adverse cardiovascular events. Studies in humans and experimental animal models have provided evidence supporting the vital contribution of endothelial dysfunction to cardiovascular disease. In humans, risk reduction therapies, such as antihypertensive therapy, cholesterol lowering, or smoking cessation, all improve endothelial function and reduce the incident of clinical events (42). Although endothelial dysfunction is well recognized as a key initiating event and a fundamental component of mechanisms underlying the pathogenesis of vascular disease, potential mechanisms causing endothelial dysfunction are not completely known. (43).

21 6 The Renin-Angiotensin System The renin-angiotensin system (RAS) is a highly regulated hormonal system which contributes to the physiological regulation of blood pressure and body fluid homeostasis. Studies performed in humans with higher risk for cardiovascular diseases, people with polymorphisms of RAS components, genetically-altered animal models, and a long history of the use of pharmacological agents revealed a critical involvement of RAS in the pathogenesis of hypertension as well as other major cardiovascular diseases (44-50). In clinical practice, pharmacological interference with the activity of the RAS, using either angiotensin-converting enzyme (ACE) inhibitors or selective antagonists of cell membrane receptors of its primary effectors (Ang II receptor blockers), are effective approaches to reduce blood pressure and reduce the risk of subsequent cardiovascular events (51-53). In the classic RAS, activation of the system is initiated from biological synthesis of renin, a glycoprotein protein produced by juxtaglomerular cells in renal afferent arterioles (54). In 1898, Tigerstedt and Bergman first discovered a pressor compound in the renal cortex extracted from the rabbit and named it renin (55). Juxtaglomerular cells are a major site for renin synthesis. Briefly, the renin mrna is translated to preprorenin, which is an inactive protein with 404 amino acids. The preprorenin is then transported into the rough endoplasmic reticulum, where a 23-amino acid pre sequence is cleaved, yielding a 47 kda inactive enzyme named prorenin. In response to appropriate stimuli, such as low volume states, high salt content in the distal tubules, renal sympathetic nerve activity, or reduced renal perfusion, the 43-amino acid pro sequence of prorenin is cleaved to form a 40 kda enzymatically active form of renin. It is this form of renin that is then released into the circulation (56-58). The only known substrate of renin is angiotensinogen, which is a 60 kda peptide produced within the liver and released into circulation.

22 7 Angiotensinogen is the precursor protein responsible for the production of the entire family of angiotensin peptides, including angiotensin I, II, III (angiotensin 2-8), IV (angiotensin 3-8), and angiotensin (1-7) (59-62). Two distinct angiotensin systems are defined by the source of angiotensinogen, including systemic, where angiotensinogen originates mainly from hepatocytes and is transported to target tissue through the circulation (63); and tissue (or local ), where angiotensinogen is synthesized in other organs or cells and transformed to angiotensin peptides locally (64, 65). Angiotensin I is an inactive decapeptide, which is hydrolyzed to an active octapeptide Ang II by ACE in vascular tissue (63). While most ACE is membrane bound, a soluble form of ACE also exists. ACE is found primarily bound to the plasma membranes of endothelial cells, including vascular endothelial cells (56). In addition to the cleaving of angiotensin I, ACE is known to degrade bradykinin (a potent vasodilator) to the inactive form bradykinin-(1-7). As a result, ACE promotes vasoconstriction by increasing the level of a vasoconstrictor (angiotensin II) and reducing the level of a vasodilator (bradykinin). Consequently, the beneficial effects of ACE inhibitors used in clinical practice are thought to result from the combination of these effects (66, 67). Ang II is the primary biological effector of the RAS which regulates renal and cardiovascular function by binding with two G protein-coupled receptors (44), the Ang II type 1 receptor (AT1R) and the Ang II type 2 receptor (AT2R). The majority of the effects of Ang II are mediated by the AT1R, while AT2R mediates effects that are generally of opposite action. Most of the pathophysiological actions of Ang II are mediated by AT1R, which belongs to the seven trans-membrane superfamily of G protein-coupled receptors. Two isoforms of AT1R [angiotensin type 1a receptor (AT1Ra) and angiotensin type 1b receptor (AT1Rb)] have been identified, and are widely expressed in multiple organs, including

23 8 liver, adrenal gland, brain, lung, kidney, heart, and the vasculature (68). While the two subtypes are functionally and pharmacologically indistinguishable, AT1Ra may be more important than AT1Rb in blood pressure regulation (69). AT1Ra deficient mice display hypotension, high circulating renin levels (70), improved endothelial function, decreased oxidative stress, and atherosclerosis (71), attenuation of diet-induced weight gain and insulin resistance (72), less left ventricular remodeling and improved survival after myocardial infarction (73). Ang II-induced constriction in abdominal aorta and femoral artery from AT1Ra deficient mice were completely antagonized by an AT1R (losartan) (74). In addition, Ang II-induced vasoconstriction in aortic segments were abolished in AT1Rb knockout mice (75). These results together suggested the critical role of AT1Rb in mediating Ang II contractile response in these arteries. A single nucleotide polymorphism of the AT1R gene (A1166C) has been identified and implicated in hypertension (76), increased aortic stiffness (77), myocardial infarction (78), increased sensitivity to Ang II (79), and enhanced Ang II-induced vasoconstriction (80). Binding of Ang II to AT1Rs couples to a small G protein which activates downstream effectors to produce second messengers, such as inositol trisphosphate and diacylglycerol. These second messengers in turn induce sarcoplasmic reticular release of Ca 2+ to increase the intracellular free Ca 2+ concentration which activates myosin light chain kinase and results in contraction (81). In addition to classic G protein-coupled pathways, Ang II also induces contraction by activating small GTP (Guanosine-5'- triphosphate)-binding proteins, such as RhoA (Ras homolog gene family, member A). Activation of RhoA and its downstream target Rho-kinase by Ang II inhibits myosin light chain phosphatase, which increases phosphorylation of myosin light chains, therefore results in contraction. RhoA/ Rho-kinase signaling pathway has been implicated in the pathophysiology of hypertension (82).

24 9 Activation of AT1Rs cross-talks with several tyrosine kinases, such as the insulin receptor, c-src (proto-oncogene tyrosine-protein kinase Src) family kinases, and serine/threonine kinases, such as protein kinase C and mitogen-activated protein kinases, which contribute to physiological actions of Ang II. Activation of nicotinamide adenine dinucleotide phosphate-oxidases (NADPH oxidases or NOXs) generates reactive oxygen species (ROS) which contribute to Ang II-induced oxidative stress and inflammation, two critical mechanisms underlying Ang II-induced endothelial dysfunction and vascular disease (83). Activation of AT2R signaling exerts anti-proliferative, anti-inflammatory, and proapoptotic effects in VCSMs by antagonizing the detrimental effects of AT1R activation (68) (84). Expression of AT2R is relatively high in fetal tissue and declines after birth, which suggests a potentially important role of AT2R in fetal development. Under pathological conditions, such as vascular injury, myocardial infarction and heart failure, the expression of AT2R is up-regulated in adult life (85). The physiological function of AT2R signaling has not been fully identified. It is known that AT2R antagonizes AT1R signaling by activation of serine/threonine phosphatase (86, 87). More recently, angiotensin (1-7), ACE2 and Mas have been identified as new members of the RAS. Angiotensin (1-7) is a heptapeptide fragment of Ang II discovered in 1988 (88). Angiotensin (1-7) is present in multiple tissues, including the heart, the brain, kidneys, and vessels, and is hydrolyzed from angiotensin I by ACE2 followed by binding to its receptor Mas (89) (66). Angiotensin (1-7) also opposes the effects of Ang II by inhibiting the formation of Ang II and potentiating the action of bradykinin. In addition, angiotensin (1-7) stimulates vascular production of NO and vasodilator prostaglandins, both of which can counteract Ang II-induced vasoconstriction and vascular remodeling (90, 91).

25 10 Besides angiotensin (1-7), Ang II can also be hydrolyzed by other aminopeptidases to yield two other products, angiotensin (2-8) (angiotensin III) and angiotensin (3-8) (angiotensin IV). The effects of angiotensin III are similar, but less potent than Ang II, in promoting inflammation, elevating blood pressure, and releasing vasopressin (59, 60, 62). Two studies indicated a potential protective role of angiotensin IV in regulating blood flow in the kidney and the brain (61, 62). Both systemic and local (tissue-generated) Ang II are responsible for regulating vascular function under physiological conditions. For example, the potential role of the local RAS in the pathogenesis of hypertension was initially suggested by a study which reported that RAS inhibitors lowered blood pressure in hypertensive patients whose plasma renin levels were normal or even low (92). Chronic administration of RAS inhibitors lowers blood pressure in various animal models of hypertension in which activity of circulating renin was not elevated (93). A local RAS has been observed in multiple tissues, such as the kidney, the heart, the brain, and the vasculature. Ang II is produced locally in the kidney and inappropriate activation of the intrarenal Ang II leads to development of hypertension and renal injury (94). In addition, hyperactive RAS in the brain contributes to neurogenic hypertension in both humans and animal models with hypertension (95). Moreover, the cardiac RAS promotes cardiac hypertrophy and thus contributes to development and progression of hypertensive heart disease. Furthermore, patients with hypertension demonstrated enhanced vascular responses to exogenous Ang II (96, 97), and increased Ang II/AT1R signaling was detected in VCSMs isolated from resistance arteries of hypertensive patients (98-100). Studies in spontaneously hypertensive rats also provided evidence for hyperactivation of the vascular RAS in hypertension (93). Collectively, these and other findings suggest that various local RAS play a critical role in the pathogenesis of hypertension and end-organ abnormalities.

26 11 In addition to regulating blood pressure, a role for local Ang II was implicated in vascular remodeling. Significant regression of vascular remodeling was found in hypertensive patients treated with either ACE inhibitors or Ang II receptor blockers, but not with the ß-adrenergic receptor blocker atenolol ( ). This regression of vascular remodeling was independent of blood pressure changes, since ACE inhibitors, Ang II receptor blockers, and atenolol all reduced blood pressure to similar levels. With regard to endothelial dysfunction, locally generated Ang II was shown to induce vascular oxidative stress and inflammation, both are major contributors to endothelial dysfunction. Ang II activates NADPH oxidases to generate ROS, thus promoting oxidative stress. Ang II also activates nuclear factor κb (NF-κB), a transcription factor considered a master regulator of inflammatory responses (83). Oxidative Stress and Inflammation Oxidative stress and inflammation are two major mechanisms which act as cooperative and synergistic partners in the pathogenesis of endothelial dysfunction and hypertension. Inflammation triggers oxidative stress by activating mechanisms that increase production of ROS (104) and/or by reductions in effectiveness of relevant antioxidant mechanisms (105). Oxidative stress promotes inflammatory responses by activating transcription factors, such as NF-κB to produce pro-inflammatory cytokines and other molecules that further activate both innate and adaptive immunity (106). Systemic low grade inflammation has been linked to many pathological conditions, and can involve activation of both the innate and/or the adaptive immune systems (107). The important role of inflammation and the immune system in hypertension and vascular disease is outlined in more detail below. Here, I emphasize the contribution of inflammation as a trigger for oxidative stress, which in turn promotes hypertension and vascular dysfunction.

27 12 Oxidative stress is defined as an imbalance between the generation and the removal (or metabolism) of ROS, such as superoxide and hydrogen peroxide (H2O2). In response to pathogens or other stimuli, ROS are generated and used by inflammatory cells in order to attack foreign invading cells or to repair tissue damage (104). In the vasculature, ROS can be produced by any of the cell types in the vessel wall, including endothelial, smooth muscle, or adventitial cells (98, 108). In cells which constitutively generate adenosine triphosphate (ATP) to maintain cellular functions, ROS are also produced as a by-product of ATP production (109). While the major sources of ROS in the cardiovascular system are NOXs, other sources include xanthine oxidase, cyclooxygenase, the mitochondrial respiratory chain, and uncoupled enos (110). NADPH oxidases are enzyme complexes that contain one of seven isoforms: Nox1-5 or Duox1-2 (111). Collectively, these isoforms are expressed in nearly all cell types, including VCSMs, endothelial cells, adventitial fibroblasts, and leukocytes ( ). The only known function of NOXs is the generation of ROS (115) whereas other sources generate ROS as a by-product of their primary function. All Nox isoforms consists of six transmembrane domains and a cytosolic C-terminus with NADPH and flavin adenine dinucleotide -binding domain. NOXs produce ROS by transferring electrons from a cytosolic donor (NADPH) to flavin adenine dinucleotide, then to molecular oxygen on the opposite side of the membrane to generate ROS (116). Superoxide produced by NOXs can be removed (converted to other species) by superoxide dismutases (SOD) or scavenged by pharmacological SOD mimetics (eg, tempol). Production of superoxide by NOXs can be inhibited by compounds including apocynin, diphenylene iodonium, VAS2870, or gp91ds-tat ( ). In response to mechanical (eg, shear stress) or neurohumoral stimuli (eg, acetylcholine), endothelium regulates vascular tone through the release of vasoactive molecules, such as NO, or other mechanisms (discussed above). In addition to its

28 13 regulatory role in endothelial-dependent vasodilation, NO also has potent antiatherosclerotic properties, which includes inhibition of platelet aggregation, leukocyteendothelial cell adhesion, as well as proliferation of VCSMs ( ). A primary mechanism by which oxidative stress produces endothelial dysfunction is by reducing NO bioavailability (124). Superoxide decreases the biological activity of NO by chemically reacting with NO to form the highly reactive intermediate peroxynitrite (123). Peroxynitrite can produce vasoconstriction, thus possibly contributing to hypertension (125). In addition, peroxynitrite is a cytotoxic molecule which may cause oxidative damage to proteins, lipids, and DNA, any of which can impair cell function (123). Peroxynitrite can also cause oxidation of tetrahydrobiopterin resulting in uncoupling of enos (123, 126). Furthermore, peroxynitrite impairs the activity of the mitochondrial form of (SOD), which further increases local levels of ROS, thus reducing the bioavailability of NO (127, 128). The role of oxidative stress in the pathophysiology of hypertension is controversial in human studies. Several clinical trials reported that antioxidant vitamins have no beneficial effects on hypertension and other cardiovascular disease outcomes ( ). Moreover, a study on individuals with type 2 diabetes showed vitamin E significantly increased blood pressure, pulse pressure and the heart rate (134). Even more surprisingly, patients with intra-arterial infusion of vitamin C demonstrated improved endothelial function but a worsened prognosis related to coronary artery disease (135). Other human studies support the involvement of oxidative stress in hypertension. For example, elevated plasma levels of oxidative markers, such as thiobarbituric acidreactive substances and 8-epi-isoprostanes, are found in patients with essential hypertension, renovascular hypertension, malignant hypertension, salt-sensitive hypertension, cyclosporine-induced hypertension, and preeclampsia ( ). Also,

29 14 increased levels of ROS caused by upregulation of vascular NOXs has been found in VCSMs from resistance arteries of hypertensive patients (100, 142). In addition, superoxide and H2O2 levels in plasma are increased in hypertensive patients (143, 144). In contrast, reduced activities of antioxidant enzymes, such as SOD, glutathione peroxidase, and catalase, as well as decreased levels of antioxidant vitamins A, C, and E have been shown in hypertensive patients compared with normotensive controls (145, 146). Interestingly, H2O2 production is higher in normotensive subjects with a family history of hypertension relative to normotensive controls (147, 148). Polymorphisms in the promoter of the p22 phox subunit of NOX are associated with increased atherosclerosis and hypertension (149). The critical role of oxidative stress has been indicated not only in hypertensive patients, but also in subjects with prehypertension (150). Studies in patients with prehypertension exhibit increased level of oxidative stress markers suggesting that oxidant-dependent processes promote the progression of hypertension (150). The link between oxidative stress and endothelial dysfunction was supported in patients with either renovascular hypertension or Gilbert syndrome (151). Patients with renovascular hypertension exhibits increased plasma levels of Ang II and thus enhanced Ang II-induced oxidative stress. Infusion of the antioxidant vitamin C increased acetylcholine-induced vasodilation in these patients (152), suggesting a positive effect of antioxidants on endothelial function. Patients with Gilbert syndrome have mildly increased unconjugated bilirubin, which is a potent endogenous antioxidant at low concentrations. Patients with Gilbert syndrome demonstrated low levels of oxidative stress and significantly greater flow-mediated vasodilation (153). The fundamental role of oxidative stress in hypertension was first identified by a study in which intravenous administration of a genetically engineered form of SOD (HB- SOD) decreased blood pressure in spontaneously hypertensive rats (154). Other experimental models also provided evidence supporting the role of oxidative stress in

30 15 hypertension, including stroke-prone spontaneously hypertensive rats, Ang II-, aldosterone-, deoxycorticosterone acetate (DOCA)/salt-, and endothelin-1-induced hypertension, as well as dietary (high salt and fat)-induced hypertension ( ). Increased NADPH oxidase activity was found in the vascular wall and kidney from these models (160, 161). Inhibition of NADPH oxidase using its peptide inhibitor (gp91ds-tat) attenuated Ang II-induced hypertension (162) and ameliorated endothelial dysfunction in Dahl salt-sensitive rats (163, 164). Nox2 deficiency in 2-kidney 1-clip or DOCA/salt had decreased Ang II-induced vascular hypertrophy and reduced hypertension ( ). Deletion of Nox1 in mice attenuated the pressor response to Ang II and improved acetylcholine-induced vasodilation (167, 168), while overexpression of Nox1 exacerbated Ang II-induced high blood pressure and aortic hypertrophy. Overexpression of Nox1 in VCSMs resulted in enhanced Ang II-induced production of ROS, which resulted in endothelial dysfunction (169, 170). Finally, the NADPH oxidase inhibitor VAS2870 significantly improved acetylcholine-induced relaxation of SHR aortae (161). Collectively, these results strongly suggest that oxidative stress is still a promising target for treatment of systemic hypertension and/or vascular end-organ damage. Robust evidence indicates that oxidative stress promotes inflammation in multiple organ systems, including the vasculature, the kidney and the nervous system. The result of these interactions is hypertension and end-organ damage (104). Reduction of ROS bioavailability by either treatment with ROS scavengers or NOX inhibitors, effectively reduced inflammatory-related endpoints in vivo and in vitro (171, 172). Oxidative stress promotes inflammatory responses by regulating intracellular signaling pathways which control gene transcription and are important for initiating proinflammatory responses (104). For example, ROS in endothelial cells cause activation of NF-κB signaling which is a key player in controlling both innate and adaptive immunity (106). In the basal state, NF-κB is bound to IκB, which is an inhibitory protein restricting

31 16 NF-κB to the cytoplasm. Stimuli, such as Ang II, induce IκB phosphorylation which release NF-κB from the inhibitory complex. NF-κB then translocates from the cytoplasm into the nucleus, inducing gene transcription (173, 174). Activation of NF-κB signaling leads to upregulation of pro-inflammatory cytokines and adhesion molecules, which are required for inflammatory cell infiltration into the vascular wall (175). Other pro-inflammatory transcription factors activated by ROS include nuclear factor (erythroid-derived 2)-like 2 (Nrf2), activator protein 1 (AP-1), and Ets-1 (104). In addition to regulation of proinflammatory gene transcription, ROS also affect the function of several protein kinase/phosphatase, phospholipases, and ion channels (174). Besides production of pro-inflammatory cytokines and adhesion molecules, oxidative stress also promotes local inflammation by increasing endothelial permeability, which facilitates infiltration of immune cells (104). Pharmacological or genetic inhibition of NOXs attenuate infiltration of macrophages and T lymphocytes into the vascular wall and prevent Ang II-induced vascular remodeling, endothelial dysfunction, and elevated blood pressure (176, 177). The potential role of NF-κB in oxidative stress-related endothelial dysfunction has been suggested ( ). A human study provided supporting evidence for this concept using non-acetylated salicylate compound (salsalate), which lacks an acetyl group needed to directly inhibit cyclooxygenase activity. Multiple studies report that salsalate also inhibits NF-κB activation by preventing phosphorylation of IκB. ( ). Inhibition of NF-κB activation by salsalate improved endothelial function in older obese adults compared with young healthy controls. Salsalate also reduced NOX expression in vascular endothelial cells (186). In addition, suppression of NF-κB activation specifically in vascular endothelial cells reduced hypertension-induced renal injury, but had no effects on blood pressure in mice treated with a high-salt diet, an inhibitor of NOS (L-NAME), and Ang II (187).

32 17 Regarding the detrimental effects of Ang II on endothelial dysfunction and hypertension, accumulating evidence suggests there are feed-forward relationships between oxidative stress and inflammation. Through the AT1R, Ang II induces activation NADPH oxidase in vitro and in vivo (188, 189), resulting in production of ROS and oxidative stress. Oxidative stress activates NF-κB, which induces gene transcription of proinflammatory cytokines, such as IL-6 (190). Cytokines bind to their receptors and initiate signaling which further facilitates oxidative stress (190). The Immune System The immune system is a defensive network consisting of effector cells, lymphoid tissues, humoral factors, and cytokines, which are collectively responsible for recognizing and eliminating harmful foreign molecules and cells (191, 192). Leukocytes are effector cells of the immune system, consisting of five cells types including neutrophils, eosinophils, basophils, mast cells, monocytes/macrophages, as well as lymphocytes. Lymphocytes are sub-divided into B lymphocytes (B cells), T lymphocytes (T cells), and natural killer cells. Most leukocytes originate from hematopoietic stem cells in the bone marrow and are released into the blood (13). Lymphoid tissues are responsible for producing, storing, or processing lymphocytes, which include bone marrow, thymus, lymph nodes, tonsils, adenoids, appendix, gut-associated lymphoid tissue, and spleen (193, 194). Bone marrow and thymus are sites for leukocyte origination and maturation. The spleen is the largest lymphoid tissue which performs immune functions in blood and serves as a blood reservoir for immune cells (193, 194). Accumulating evidence from past 50 years supports the role of the immune system in the pathogenesis of hypertension as well as other cardiovascular diseases. In 1960s, Grollman et al. first reported that the immune system may contribute to hypertension by demonstrating that renal infarction-induced hypertension was dampened by

33 18 immunosuppression and that hypertension was induced in normotensive rats by passive transferring lymphocytes from rats with renal infarction-induced hypertension (195, 196). Later, it was found that thymectiomized mice with renal infarction were protected from hypertension (197). In addition, transplantation of thymus from Wistar Kyoto rats blunted the increased in blood pressure that normally develops in spontaneously hypertensive rats (198). Treatment with either anti-thymocyte serum or cyclophosphamide (an immunosuppressive drug) lowered the blood pressure in spontaneously hypertensive rats. Immunosuppression with mycophenlate mofetil was also shown to reduce salt-induced hypertension after Ang II infusion (196, 199, 200). Collectively, these findings support the concept that activation of the immune system is involved in the pathophysiology of hypertension and related vascular diseases. Functionally, the immune system consists of two fundamental types of immune responses: innate and adaptive or acquired. Recent studies indicated that both the innate and the adaptive immune systems contribute to the development and progression of hypertension, as well as the associated end-organ damage ( ). The innate immune system is the first line of defense of the body against pathogens, which consists of rapid, non-specific responses mediated by effector cells including phagocytes (monocytes/macrophages, neutrophils) (204), dendritic cells, and natural killer cells (205, 206). Activation of innate responses is triggered by pattern recognition receptors, which sense either pathogen-associated molecular patterns (PAMPs) or damage associated molecular patterns (DAMPs) in the stressed or injured tissue. Activation of pattern recognition receptors triggers downstream signaling pathways, resulting in production of cytokines and ROS, which are major contributors of endothelial dysfunction, vascular disease, and hypertension (207, 208). The role of the innate immune system in hypertension and vascular disease was demonstrated by animal models with disruption of its effector cells. Macrophages are

34 19 depleted in mice with a mutation in the macrophage colony-stimulating factor gene. These mice are protected against both Ang II and deoxycorticosterone acetate (DOCA)-induced hypertension, endothelial dysfunction, vascular remodeling and oxidative stress (209, 210). In addition, depletion of circulating monocytes attenuated Ang II-induced hypertension, endothelial dysfunction, hypertrophy, and oxidative stress; while adoptive transfer of monocytes restored the detrimental effects of Ang II (211). Activation of adaptive effector cells requires antigen presentation by innate cells. Dendritic cells are the most potent antigen-presenting cells (APCs), which are found abundantly in tissues that interface with the external environment (e.g., skin, lungs, and the lining of the gastrointestinal tract) (13). Disruption of signaling within dendritic cells with pharmacological or genetic approaches blunted the development of Ang II- or DOCA-saltinduced hypertension, as well as activation and infiltration of T cells into the wall of aorta (212). The role of natural killer cells in the pathogenesis of hypertension was supported by studies in two rodent models. Ang II-induced vascular dysfunction and hypertension were attenuated in mice with deficiency of T-bet, which is a protein required for natural killer cells activation; or in mice whose natural killer cells were depleted by injecting antinatural killer cell antibodies (213). In addition, a natural killer gene complex, which encodes several natural killer cell receptors, was identified as an important determinant of the genetic susceptibility to hypertension and vascular remodeling (214).

35 20 Toll-like receptors are the most extensively studied pattern recognition receptors, and are widely expressed in cells from both the immune and cardiovascular systems (215). In response to either PAMPs or DAMPs in the extracellular or endosomal compartments, the intracellular domains of toll-like receptors trigger a signaling cascade resulting in activation of transcriptional factors, such as NF-κB, chemokines, adhesion molecules, matrix metalloproteinases, cyclooxygenase 2 and inducible NOS ( ). The role of toll-like receptors, particularly toll-like receptor 4, as well as NF-κB signaling in hypertension and cardiovascular disease has been supported by multiple studies. Toll-like receptor 4 mrna and protein was increased in both mesenteric arteries and aorta in spontaneously hypertensive rats compared with Wistar Kyoto rats. Blood pressure, vascular inflammation, and maximal contraction of mesenteric arteries to noradrenaline was reduced by antibodies that antagonize the activation of toll-like receptors 4 (219). Genetic deficiency in toll-like receptor 4 protected obese mice against endothelial dysfunction by blocking increases in ROS (220). Inhibition of toll-like receptor 4 improved endothelial function in aorta of spontaneously hypertensive rat, and reduced Ang II-induced NADPH oxidase activity and superoxide levels in VCSMs from spontaneously hypertensive rats (221). Compared to the innate immune system, the contribution of the adaptive immune system to the pathogenesis of hypertension has been better characterized. Since the first evidence suggesting that immunosuppression could blunt development of hypertension in a model of renal infarction in the 1960s, accumulating studies have provided additional evidence supporting the role of the adaptive immune system in hypertension (195). Direct evidence for a pivotal role of T or B cells in the development of hypertension was obtained in studies using a mouse model lacking T and B cells (Rag-1 -/- mice) (222). The significant attenuation of Ang II-induced hypertension, vascular superoxide levels, and endothelial dysfunction suggests a role of either T or B cells in mediating overt hypertension. In

36 21 addition, severe combined immunodeficient mice, which lack both functional T and B cells, are protected against Ang II-induced hypertension, which further support the role of T or B cells in hypertension (223). TH17 cells are a newly characterized subset of T cells, which contribute to numerous autoimmune diseases, obesity and cardiovascular disease by producing the proinflammatory cytokine IL-17 ( ). Ang II-induced elevated blood pressure cannot be maintained in mice absent of IL-17a (227). In addition, Ang II-induced T cell infiltration, vascular oxidative stress, and endothelial dysfunction was abolished in the mice lacking IL-17a. Consistent with the concept that IL-17 is an important mediator of hypertension, chronic infusion of IL-17a results in endothelial dysfunction and hypertension mediated via Rho-kinase (228). In contrast to T effector cells (Th1, Th2, and Th17) which promote inflammation, regulatory T (Treg) cells suppress the inflammatory response, and thereby attenuate hypertension and hypertension-induced end-organ damage (229, 230). Adoptive transfer of Treg cells inhibited the effects of Ang II on blood pressure, oxidative stress, and immune cells infiltration to vessels and kidney (231). The protective effects of Treg cells are most likely mediated by producing the antiinflammatory cytokine IL-10 (232). IL-10 is produced mostly by leukocytes, and has been considered to be a major protective molecule in the vascular wall (233, 234). IL-10 inhibits inflammation by either suppressing pro-inflammatory signaling, such as NF-κB, or increasing anti-inflammatory gene expression (235). Plasma levels of IL-10 were reported to positively associate with risk of cardiovascular disease in humans (236). The role of IL- 10 in vascular function and hypertension has been investigated in a mouse model with IL- 10 deficiency (IL-10 -/- ). Enhanced vascular superoxide levels and endothelial dysfunction were found in IL-10 -/- mice during lipopolysaccharide-induced inflammation (237). In

37 22 addition, carotid arteries from IL-10 -/- mice were predisposed to Ang II-induced endothelial dysfunction as well as oxidative stress (238). Furthermore, arteries from IL-10 -/- mice incubated with conditioned media of cultured Treg cells from control mice were protected from Ang II infusion-induced endothelial dysfunction, hypertension, as well as oxidative stress (239). Finally, local administration of recombinant mouse IL-10 counteracted Ang II-induced endothelial dysfunction in mouse aortic rings in vitro (240). Interleukin-6 Signaling and Suppressor of Cytokine Signaling 3 Cytokines are a family of protein molecules with a broad range of cellular functions that are mediated by binding to specific cell surface receptors. In response to inflammatory stimuli, cytokines function as local, paracrine or autocrine signals which mediate chemotaxis, protein synthesis, mitogenesis, and cellular differentiation (241). Interleukin-6 (IL-6) is a pleiotropic cytokine with multiple cellular functions including inflammation/immune regulation, neoplasia, hematopoiesis, and cell growth and differentiation (242). IL-6 was originally discovered in 1986 as a B-cell differentiation factor which differentiates B-cell to produce immunoglobulin (243). Various types of cells in the vascular wall have the capacity to produce IL-6, including macrophages, lymphocytes, fibroblasts, endothelial cells, VCSMs, and adventitia (244, 245). The expression level of IL-6 in normal healthy cells is low (246). In response to stress stimuli, the expression of IL-6 is rapidly induced by activation of transcription factors associated with inflammatory or proliferative states, such as NF-κB, C/EBPβ (CCAAT-enhancer-binding protein), AP-1, and corticosteroids ( ). The biological activities of IL-6 is regulated through binding to the IL-6 receptor (IL-6R) and gp130, which is a co-receptor required for IL-6 signaling (249). There are two types of IL-6Rs: membrane-bound (mil-6r) and soluble (sil-6r), which activate classical

38 23 signaling and trans-signaling respectively (250). Binding of IL-6 to mil-6r on the cell surface leads to the recruitment and homodimerization of two gp130 proteins and therefore activates downstream intracellular signaling, which is regarded as classical signaling (251). The soluble form of IL-6Rs are generated by proteolytic cleavage or alternative splicing ( ). Binding of IL-6 to the sil-6r forms an IL-6/sIL-6R complex which then recruits and binds to a pair of membrane-bound gp130, resulting in initiation of transsignaling (255), which is thought to contribute to the pathophysiology of chronic inflammatory disorder and cancer (256). The trans-signaling of IL-6 is highly regulated by the soluble form of gp130 (sgp130), which exists in circulating blood and binds to IL6/sIL- 6R complex to prevent its binding to membrane-bound gp130 (257, 258). As a result, sgp130 is an endogenous inhibitor of IL-6 signaling. Dimerization of ubiquitary expressed gp130 induces recruitment of a non-receptor tyrosine kinase Janus kinase (JAK), which phosphorylates tyrosine residues in the cytoplasmic domain of gp130 (259). Binding of IL-6 triggers phosphorylation of two tyrosine residues: Tyr 759, which drives the SHP-2/ERK/MAPK pathway; and YXXQ motifs, which mediates JAK/ signal transducers and activators of transcription (STAT) pathway (259). Activated JAKs phosphorylate the receptor cytoplasmic domains, which creates docking sites for SH2-containing signaling proteins. One of the substrates of tyrosine phosphorylation is the STAT family. After being phosphorylated and activated, STAT proteins detached from the receptor, become either homodimers ((STAT1/STAT3)) or heterodimers (STAT3/STAT3), which translocate into the nucleus and activate transcription by binding to specific DNA sequences in the promoter region of their target genes (246). The signaling conducted by JAKs are inhibited by the suppressor of cytokine signaling (SOCS) protein family, which was identified independently by three groups in 1997 ( ). Interestingly, the expression of SOCS was induced by JAK/STAT pathway, which in turn negatively regulate this signaling pathway (263). Based on

39 24 structural similarities, there are 8 members in the SOCS family, SOCS1-7 and CIS ( ). All of these members contain a Src homology 2 domain which mediates binding to tyrosine phosphorylation within target proteins; a SOCS-box domain which functions as E3 ubiquitin ligases mediating protein degradation; and an optional kinase inhibitory region, regulating termination of JAK/STAT signaling and is only found in SOCS1 and 3 (267). The promoter region of SOCSs contains not only STAT factor binding elements but also other transcription factors, such as NF-κB. The Src homology 2 domain determines the target of each SOCS protein. SOCS3 selectively binds to gp130-related cytokine receptors, such as the IL-6 receptor, and then suppresses their signaling by ubiquitin-mediated degradation of the signaling complex or by inhibiting kinase activity (268). Because the receptors to which SOCS3 binds mostly activate STAT3, SOCS3 is considered an inhibitor relatively specific to STAT3. SOCS3 does not bind to the IL-10 receptor, so it cannot inhibit IL-10-induced STAT3 activation (267, 268). Extensive studies suggest there is an association of IL-6 with hypertension and cardiovascular diseases. In human studies, plasma levels of IL-6 were positively correlated with blood pressure (269). In addition, higher production of IL-6 was found in white blood cells from hypertensive patients in response to lipopolysaccharide than those from nonhypertensive population (270). In experimental studies, expression of IL-6 mrna and protein can be induced by Ang II in cultured rat VCSMs. Ang II-induced increases in IL-6 protein were abolished by an AT1R, but not an AT2R antagonist (271). Previously, our group provided in vivo evidence showing that chronic Ang II infusion increased IL-6 mrna levels in aorta (272). The biological function of IL-6 has been studied by multiple groups using mice with IL-6 deficiency ( ). In a model which examines direct effects of Ang II on the vessel wall (using overnight incubation of arteries as described in Chapter 2), Ang II-induced endothelial dysfunction was prevented in mice with IL-6 deficiency (273). In a chronic

40 25 Ang II-dependent hypertension model, Ang II-induced increases in superoxide, endothelial dysfunction, and vascular hypertrophy were absent in IL-6 deficient mice (273). Consistent with these results, two other groups also reported Ang II-induced hypertension was blunted by IL-6 deficiency (274) (275) Recently, our group reported that both local and systemic effects of Ang II were prevented in C57BL/J6 mice by two structurally unrelated small molecule inhibitors of STAT3 (272). In addition, Ang II-induced elevated blood pressure was attenuated by inhibition of STAT3 (272). In addition to pro-inflammatory effects of IL-6/STAT3 signaling that have been described, anti-inflammatory responses induced by IL-6/STAT3 signaling have also been reported over the past 20 years ( ). A systematic, prospective assessment of the vascular tree in adult patients with STAT3 deficiency reported a high prevalence of structural abnormalites (ectasia and aneurysm) throughout the vascular tree (279). Such findings suggest a critical role of STAT3 in maintaining normal vascular homeostasis. In addition, IL-6/STAT3 signaling was found to elicit neuroprotective effects in a model of ischemic stroke (280). Furthermore, IL-6 was also identified as a myokine which mediates the anti-inflammatory effects of exercise (281, 282). Finally, apolipoprotein E (ApoE)/IL- 6 double deficient (ApoE -/- IL-6 -/- ) mice showed a similar degree of hypercholesterolemia compared to ApoE -/- IL-6 +/+ mice, but presented with significantly larger and more calcified atherosclerotic lesions. Consistent with these results, ApoE -/-/ IL-6 -/- mice showed enhanced atherosclerotic plaque formation along with maladaptive vascular developmental processes (283, 284). Most recently, a study in Nature Immunology provided strong evidence supporting anti-inflammatory effects of IL-6 signaling in myeloid cells which limited endotoxemia and obesity-associated resistance to insulin (285). Very little is known about the mechanisms that dictate whether IL-6 elicits pro- or anti-inflammatory responses. Multiple studies suggested that SOCS3, a negative regulator of IL-6 signaling, may play a key role in determining pro- or anti-inflammatory responses

41 26 to IL-6 in select subtypes of immune cells ( ). The role of SOCS3 in this regarding in vascular disease and hypertension is unknown. SOCS3 deficiency is embryonically lethal in mice because of placenta defects (290). Mice with conditional deletion of SOCS3 in hematopoietic cells developed lethal inflammatory disease during adult life, which was associated with increased IL-6 levels (291). In addition, acute responses to IL-1β were lethal in conditional SOCS3 deficient mice, but not SOCS3/IL-6 double knock out mice, which suggested a role for SOCS3 in suppressing IL-6-dependent inflammation (292). Overexpression of SOCS3 has been shown to effectively suppress inflammation in a model of rheumatoid arthritis (293). In relation to the cardiovascular system, there is very little known regarding the functional importance of SOCS3. Recent work suggested that reduced expression of SOCS3 in placenta was associated with, and thus may contribute to, preeclampsia (284). In contrast, SOCS3 was suggested to regulate anti-inflammatory reactions mediated by IL-6 in several studies. Mice with SOCS3 deficiency in macrophages are protected from endotoxemia (286), but also exhibited reduced IL-12 responses and succumb to toxoplasmosis (294). In addition, knockdown of SOCS3 in macrophages by short interfering RNA prevents M1 (pro-inflammatory) polarization (295). Furthermore, mice with specific deficiency of SOCS3 in macrophages are resistant to tumor transplantation, which results from decreased tumor-promoting cytokines, such as TNFα and IL-6 and increased production of anti-tumorigenic cytokine termed monocyte chemotactic protein 2 (296). Although evidence suggests that loss of SOCS3 expression and activity are associated with many pathological conditions, such as cancer and liver fibrosis (297, 298), the role of SOCS3 in vascular disease and hypertension has not been defined. To my knowledge, the current studies are the first to directly examine the functional importance of SOCS3 in models of vascular disease.

42 27 Thesis Focus Hypertension is a major risk factor for vascular disease, stroke and cognitive impairment with more than 30% of the U.S. adult population being affected. The reninangiotensin system plays a fundamental role in vascular disease and hypertension, and represents a major therapeutic target for controlling blood pressure and reducing the risk of cardiovascular events. There are still significant gaps in our understanding of underlying mechanisms by which Ang II contributes to the pathogenesis of hypertension and subsequent end-organ damage. Previously, studies from our group as well as others revealed a critical role for IL-6/STAT3 signaling in contributing to detrimental effects of Ang II on the vasculature. While SOCS3 is thought to be a negative regulator of IL- 6/STAT3 signaling, its role in vascular disease and hypertension has not been examined. Thus, the overall goal of my thesis research is to investigate the role of SOCS3 in Ang IIinduced endothelial dysfunction and hypertension using a mouse model with SOCS3 haplodeficiency. The major goals and findings are summarized in Figure 1. Ang II is thought to contribute to vascular disease and hypertension through both systemic and local mechanisms. The systemic effects of Ang II involve multiple organ systems, including the kidney, the central nervous system, the immune system, and the vasculature. In Chapter 2, I focus on testing the role of SOCS3 in response to direct or local effects of Ang II on the vasculature, as well as defining underlying mechanisms. Endothelial function in response to local treatment with Ang II was tested using an in vitro (organ culture) model, which eliminates potential confounding effects due to increases in arterial pressure or other effects when Ang II is administered systemically. Using both genetic and pharmacological approaches, I provide the first evidence suggesting that SOCS3 protects against direct effects of Ang II through inhibition of IL-6/STAT3 signaling and oxidative stress.

43 28 Developing novel therapeutic approaches may be facilitated by better understanding of endogenous mechanisms that impact the progression of vascular disease and hypertension. Our laboratory previously provided evidence suggesting that IL- 6/STAT3 signaling plays an important role in mechanisms underlying Ang II-induced vascular disease and hypertension. In Chapter 3, I focus on examining the importance of SOCS3, a negative regulator of IL-6/STAT3 signaling in other systems, in a model of chronic Ang II-dependent hypertension. I examined the importance of SOCS3 in Ang II induced effects on endothelial function in carotid arteries, as well as cerebral arteries to determine if the impact of SOCS3 extends to smaller resistance vessels supplying a vital organ that is greatly impacted by hypertension. In addition, since Ang II promotes vascular inflammation through mechanisms that are either dependent or independent of elevated blood pressure, I also studied the role of SOCS3 using a sub-pressor dose of Ang II. Surprisingly, results from Chapters 2 and 3 suggested that SOCS3 played divergent roles in response to local versus systemic effects of Ang II on endothelial function. Briefly, SOCS3 haplodeficiency augmented detrimental effects on endothelial function when local effects of Ang II were tested. In contrast, SOCS3 haplodeficiency prevented the majority of Ang II-induced endothelial dysfunction when Ang II was administered systemically. Thus, in Chapter 4, I attempted to answer the question of why SOCS3 plays such a contextdependent role in relation to local versus systemic effects of Ang II. To address this question, I performed bone marrow transplantation to test if immune cells contributed to the protective effects of SOCS3 haplodeficiency that was observed in the in vivo model of Ang II-induced hypertension. To my knowledge, these are the first series of studies to directly examine the functional importance of SOCS3 in models of vascular disease and hypertension. The work provides new insight into mechanisms that regulate oxidative stress and inflammation in

44 29 the vasculature and may ultimately contribute to development of additional novel treatments for vascular disease

45 Figure 1. The role of SOCS3 in angiotensin II-induced vascular dysfunction. SOCS3 exerts divergent roles in response to local versus systemic effects of angiotensin II on vascular function. In the vasculature, SOCS3 haplodeficiency augments the detrimental effects of angiotensin II on vascular function. Systemically, SOCS3 haplodeficiency protects against chronic angiotensin II infusion-induced vascular dysfunction. Bone marrow derived cells contribute to this beneficial effect of SOCS3 haplodeficiency on vascular function 30

46 31 CHAPTER2 SOCS3 PROTECTS AGAINST DIRECT EFFECTS OF ANGIOTENSIN II ON ENDOTHELIAL FUNCTION

47 32 Introduction The RAS plays an important role in cardiovascular homeostasis. Ang II, a key effector of the RAS, is strongly implicated in various cardiovascular diseases, including hypertension, atherosclerosis, and stroke (44-50). Clinical and experimental studies suggest that Ang II, via activation of the AT1R, impairs normal vascular function and promotes the progression of vascular disease (44). Two major interrelated mechanisms that are involved in vascular disease are oxidative stress and inflammation (104, 105). Accumulating data suggest that Ang II promotes both these effects in vascular cells. Ang II increases oxidative stress via several mechanisms including activation of NADPH oxidases, a major source of ROS including superoxide in vascular cells ( ). ROS affect second messengers and other signaling molecules that are implicated in Ang II-mediated vascular abnormalities and hypertension. Superoxide impairs endothelial function through interactions with the endothelium-derived NO. Superoxide also serves as a precursor for other ROS (127, 128). Endothelial dysfunction is a key event in the onset and pathogenesis of vascular diseases of diverse etiology (40, 41). Genetic and pharmacological manipulations that inhibit ROS formation or effects protect against Ang II-induced endothelial dysfunction and hypertension (151) ( ) (169, 170). Low grade inflammation occurs commonly within the vessel wall during vascular disease. Ang II activates the transcription factor NF-κB, which is redox-sensitive and a central signal integrator responsible for regulating expression of many genes including proinflammatory cytokines (190). A major cytokine produced by NF-κB activation is IL-6 (245, 299). Growing evidence suggests an important role of IL-6 in vascular disease (245, 246, 300). Expression of IL-6 is induced by Ang II in vascular tissue (190, 273, 301). Ang II-induced endothelial dysfunction is abolished in IL-6 deficient mice but can be restored by reconstituting IL-6 (273). IL-6 activates the downstream JAK/STAT3 pathway by binding to the IL-6 receptor. Consistent with this concept, previous studies in our

48 33 laboratory demonstrated that small molecule inhibitors of STAT3 activation protect against Ang II-induced oxidative stress, vascular dysfunction, and hypertension (301), suggesting a critical role for IL-6/JAK-STAT3 in these processes. A known negative regulator of IL-6/JAK-STAT3 signaling is SOCS3, which selectively binds to phosphorylated tyrosine residues on JAK and STAT3. Through these effects, SOCS3 suppresses JAK/STAT3 signaling by ubiquitin-mediated degradation of the signaling complex or by inhibiting kinase activity (268). Although extensive studies have evaluated the role of SOCS3 in regulating IL-6/STAT3 signaling in the immune system and in cancer biology (297, 298), the role of SOCS3 in vascular disease and hypertension has not been defined. The vasculature can be affected by both circulating and local sources of Ang II. Both sources are thought to contribute to vascular disease and hypertension. Our laboratory previously developed an in vitro model which examines direct effects of Ang II on vessels (272, 273). The model avoids potential confounding effects due to increased blood pressure or other systemic effects of Ang II when administrated in vivo and has also been used by other investigators (302, 303). In this Chapter, I describe studies using this in vitro model and provide the first evidence suggesting that SOCS3 protects against direct effects of Ang II on endothelial function. This protective effect of SOCS3 involves regulation of IL- 6/STAT3 signaling and oxidative stress. Methods Experimental Animals SOCS3 haplodeficient mice were generated previously by deleting the exon containing the entire coding region of the gene to create a completely null mutation (304). Complete genetic haplodeficiency in SOCS3 is lethal (304). In contrast, SOCS3 +/- mice are phenotypically normal under unstressed conditions (304). We obtained breeding pairs of

49 34 these mice from Dr. Paul Rothman and established a colony of the mice in our laboratory. Animals were obtained by breeding SOCS3 +/- with C57BL/6J mice (from the Jackson Laboratories). Both male and female SOCS3 +/- mice and SOCS3 +/+ littermates (~4-6 months of age) were used in the current study. Mice were fed with regular chow and water, and maintained under standard housing conditions. All studies followed the Guide for the Care and Use of Laboratory animals and approved by the Institutional Animal Care and Use Committee at the University of Iowa. Genotyping of SOCS3- Haplodeficient Mice by PCR Genomic DNA was extracted from mouse tails using a DNeasy Blood & Tissue kit (Qiagen). Approximately 200 ng of DNA was amplified in a 25 μl reaction using Platinum Taq DNA Polymerase High Fidelity (Life Technologies) with a final concentration of each dntp at 0.2 mm, MgCl2 at 2 mm, and the primer at 0.4 µm. The PCR cycle profile was as follows: 1 cycle at 94 C for 4 min, 35 cycles at 94 C for 1 min, 62 C for 30 sec, 72 C for 1 min, and 1 final cycle at 72 C for 5 min. The PCR primers were P1 (5 - AGGGGAAGAGACTGTCTGGGG), P2 (5 -CCGCACAGCGGCCG CTACC), and P3 (5 -ACCACACTGCTCGACATTGGGT). A fragment about 200 bp represent the wildtype allele, whereas a fragment about 300 bp indicates the mutated allele. Incubation of Blood Vessels As described previously (273, 301), mice were anesthetized with pentobarbital (100 mg/kg IP). Carotid arteries and aorta were removed, cleaned, cut into segments ~ 5 mm in length, and placed in oxygenated Krebs solution containing mm NaCl, 4.7 mm KCl, 2.5 mm CaCl2, 1.2 mm MgSO4, 1.2 mm KH2PO4, 25 mm NaHCO3, and 11 mm glucose. Vessels were then incubated for 22 hours in DMEM growth medium (containing 5 mm glucose, 120 U/mL penicillin, 120 μg/ml streptomycin, and 50 μg/ml polymixin B at 37 C, gassed with 95% oxygen and 5% carbon dioxide) with vehicle (saline unless noted

50 35 otherwise) or Ang II (1 or 10 nm; Sigma) in the absence or presence of specific inhibitors including NBD peptides (10 μm; Imgenex), anti-il-6 antibody (5 μg/ml; R&D Systems), and S3I-201 (10 μm; Calbiochem). Sterile PBS is the vehicle of anti-il-6 antibody. Dimethyl sufoxide (DMSO) served as vehicle for NBD peptides and S3I-201. Vascular Function After incubation overnight, carotid arteries were mounted on a pair of stainlesssteel triangular hooks, one of which was connected to a stationary bracket and the other one was connected to a force transducer. Vascular rings were suspended in an organ bath containing 20 ml of oxygenated (95% oxygen and 5% carbon dioxide) Krebs buffer maintained at 37 C and equilibrated at an optimal resting tension of 0.25 gram for 45 minutes. Relaxation of carotid arteries in response to acetylcholine (Sigma Aldrich) and nitroprusside (Sigma Aldrich) were measured after pre-contraction (50-60% of maximum) induced by U Endothelial function was evaluated by measuring relaxation generated by cumulative addition of acetylcholine in concentrations from to 10-4 M. Endothelialindependent relaxation was assessed by cumulative addition of nitroprusside following stable contraction induced by U Using inhibitors of nitric oxide synthase (NOS) and mice deficient in expression of endothelial NOS (enos), our laboratory has shown previously that relaxation of carotid arteries in response to acetylcholine is mediated by enos-derived NO (36). Vasodilation to nitroprusside occurs by direct effects of this NO donor on vascular muscle and is independent of enos (36). At the end of each experiment, contractile responses were measured by cumulative addition of U46619 in gradually increasing concentrations (0.03, 0.1, 0.3, 1, and 3 µg/20 ml). To determine if vascular dysfunction was mediated by superoxide, tempol (a superoxide scavenger) was added to organ baths 30 minutes before precontraction and maintained throughout the study. In separate experiments, contractile responses to Ang II and U46619 was determined in other blood vessels as well (abdominal aorta, basilar artery, iliac and femoral arteries).

51 36 For experiments in which constrictor responses to Ang II and U46619 were performed using a resistance vessel, basilar arteries were isolated from the brain and cannulated onto glass micropipettes filled with Krebs buffer in an organ as described in detail previously (272, 305). After being pressurized to 60 mmhg, lumen diameter was measured from images projected onto a video monitor using an electronic dimension analyzer. Following 30 minutes of equilibration, responses of basilar arteries to Ang II and U46619 were measured. Quantitative Real-time RT-PCR Aorta was also sampled and aortic homogenates were stored in TRIzol reagent (Invitrogen) and purified using the RNAeasy kit (Qiagen). The concentration of RNA was determined using a NanoDrop ND cdna was generated by reverse transcriptase polymerase chain reaction (RT-PCR) using random hexamers. mrna levels of SOCS3 (Mm _g1, Life Technologies), enos (Mm _m1, Life Technologies), AT1R (Mm _s1, Life Technologies), SOD1 (Mm _g1, Life Technologies), SOD2 (Mm _m1, Life Technologies), SOCS1 (Mm _s1, Life Technologies), STAT3 (Mm.PT , IDT), NOX4 (Mm _m1, Life Technologies), PPARγ (Mm _m1, Life Technologies), NFƙBIa (Mm.PT , IDT), IL-6 (Mm _m1, Life Technologies), IL-6R (Mm _m1, Life Technologies) were determined by qrt-pcr using the TaqMan system (Applied Biosystems). β-actin (Applied Biosystems) was used as internal control within each sample. Relative fold expression was calculated using the ΔΔCt method (306) Statistical Analysis Relaxation to acetylcholine and nitroprusside was presented as percent relaxation relative to the level of precontraction induced by U Absolute contractions were presented as grams of tension. In cerebral arteries, changes in vessel diameter were

52 37 expressed as percent change. Results were expressed as the mean ± SEM and compared by Student t test or ANOVA followed by Student-Newman-Keuls post hoc test as appropriate. Differences were considered significant at P Results Effects of Ang II on Endothelial Function Are Mediated Through AT1R Signaling Endothelial function of carotid arteries from C57BL/6 mice were examined after 22 hours incubation with vehicle or Ang II. The endothelium-dependent agonist acetylcholine-produced full relaxation in carotid arteries incubated with vehicle (Figure 2A). In contrast, acetylcholine-induced relaxation in carotid arteries treated with 10 nm Ang II was significantly reduced. Co-incubation with losartan (1 µm), an antagonist of AT1R, prevented Ang II-induced impairment of endothelium-dependent relaxation. Losartan did not affect acetylcholine-induced relaxation in carotid arteries treated with vehicle (Figure 2A). These results indicated that local effects of Ang II on endothelial function were mediated through AT1R signaling. Responses to nitroprusside, the endothelium-independent vasodilator, were similar in all groups (Figure 2B). Direct Effects of Ang II on Endothelial Function Are Augmented by SOCS3 Haplodeficiency Because we anticipated that vascular effects of Ang II would be enhanced in SOCS3 +/- mice, we tested effects of a lower concentration of Ang II (1 nm) in initial experiments. To test the role of SOCS3 in response to direct effects of Ang II on the vessel wall, carotid arteries from SOCS3 +/+ and SOCS3 +/- mice were incubated with Ang II for 22 hours, followed by examination of endothelial function. Relaxation to acetylcholine was similar in all arteries incubated with vehicle. Consistent with previous studies, this low concentration of Ang II did not affect acetylcholine-induced vasodilation in carotid arteries

53 38 from SOCS3 +/+ mice (Figure 3A) (238, 307). In contrast, 1 nm Ang II reduced responses to acetylcholine in arteries from SOCS3 +/- mice by ~50% (Figure 3B). Responses to sodium nitroprusside and vasoconstriction to U46619 were similar in all groups (Figure 4 and 5), indicating that effects of Ang II on vascular function were selective for endothelial cells. These data suggested that haplodeficiency in SOCS3 does not alter endothelial function under baseline conditions, but predisposes to Ang II-induced endothelial dysfunction. Overall, these findings suggest that under normal conditions, SOCS3 protects against direct effects of Ang II on the vessel wall. Ang II-Induced Impairment of Endothelial Function in SOCS3 Haplodeficient Mice Is Inhibited by Suppressing Oxidative- or Inflammatory-Related Signaling Previous studies in our group suggested that inflammation and oxidative stress are two major interrelated mechanisms involved in Ang II-dependent vascular dysfunction. To test if Ang II-induced endothelial dysfunction in SOCS3 +/- mice was mediated by superoxide, vascular responses to acetylcholine and sodium nitroprusside were examined in the presence or absence of tempol (1 mm, a superoxide scavenger). As seen in the initial studies described above, Ang II did not affect acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ mice (Figure 6A), but markedly impaired acetylcholine-induced dilation in carotid arteries from SOCS3 +/- mice (Figure 6B). Ang II-induced endothelial dysfunction of carotid arteries from SOCS3 +/- mice was reversed by tempol (Figure 6B). Acetylcholine-induced relaxation in SOCS3 +/+ and SOCS3 +/- carotid arteries treated with vehicle was not affected by tempol (Figure 6). In addition, nitroprusside-induced vasodilation was similar in all groups (Figure 7). To test if Ang II-induced endothelial dysfunction in SOCS3 +/- mice is dependent on activation of NF-κB, a cell-penetrating NF-κB essential modulator (NEMO)-binding domain (NBD) peptide, and inactive mutant control peptides were used during incubation.

54 39 NBD peptide, which blocks the association of NEMO with the IKK complex, inhibits NFκB activation in models of acute and chronic inflammation (308, 309). Acetylcholineinduced relaxation was similar in SOCS3 +/+ carotid arteries treated with either vehicle or Ang II, as well as vehicle-treated SOCS3 +/- arteries (Figure 8). NBD peptide did not alter endothelial function in SOCS3 +/+ mice or vehicle-treated SOCS3 +/- mice (Figure 8), but prevented Ang II-induced endothelial dysfunction in carotid arteries from SOCS3 +/- mice (Figure 8B). The inactive control peptide had no significant effect on relaxation of arteries in any group (Figure 8). In addition, nitroprusside-induced vasodilation was similar in all groups with and without NBD (Figure 9). To determine if Ang II-induced effects in SOCS3 +/- mice are mediated by IL-6 and STAT3-dependent signaling, carotid arteries were incubated with either an anti-il-6 antibody (10 µg/ml) or a STAT3 inhibitor (S3I-201, 10 µm) in the presence or absence of Ang II for 22 hours. Ang II-induced endothelial dysfunction was prevented by neutralizing IL-6 or by inhibiting STAT3 (Figure 10B and 12B). Neither the anti-il-6 antibody nor S3I-201 altered acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ mice or vessels from SOCS3 +/- mice treated with vehicle (Figure 10 and 12). Vascular responses to nitroprusside were similar in all groups with and without anti-il-6 antibodies or S3I- 201 (Figure 11 and 13). In conclusion, direct effects of Ang II on vascular function were prevented by inhibitors of NF-κB, IL-6, STAT3 or a scavenger of superoxide. These results suggested that SOCS3 protects against direct effects of Ang II on vascular function by inhibiting oxidative stress and immune-related signaling. Vascular Expression of Select Genes Implicated in Vascular Disease To further investigate molecular mechanisms that may underlie augmented effects of Ang II on endothelial function in SOCS3 +/- mice, vascular expression (relative mrna

55 40 levels) of select genes implicated in vascular disease were measured using real-time RT- PCR. SOCS3 mrna levels were reduced by ~50% in aorta from SOCS3 +/- mice compared with those from SOCS3 +/+ mice. In contrast, the mrna level of SOCS1 was not significantly altered in SOCS3 haplodeficient mice. These results confirm the partial genetic haplodeficiency in SOCS3 in the model and suggest that there was no compensatory increase in expression in another SOCS isoform (SOCS1). This observation is consistent with previous studies (310). Haplodeficient in SOCS3 was associated with significantly reduced expression of enos, a gene responsible for NO production and thus important for NO-dependent vasodilation. Genetic haplodeficiency in SOCS3 also decreased expression of copper/zinc SOD (CuZn-SOD or SOD 1) and the mitochondrial form of SOD (Mn-SOD or SOD 2), two key enzymes for the dismutation of the superoxide; and therefore to prevent oxidative stress. AT1R mrna levels were increased ~2 fold in SOCS3 +/- mice compared with those in SOCS3 +/+ mice (Figure 14). Vasoconstrictor Responses to Ang II Because I measured an increase in levels of AT1R mrna in SOCS3 +/- mice, vasoconstrictor responses of several vessels were evaluated in an attempt to gain insight into whether the changes in AT1R expression translated to a functional change. In contrast to what might be predicted, vasoconstrictor responses to Ang II were similar between SOCS3 +/+ and SOCS3 +/- mice in the abdominal aorta and in two peripheral arteries (iliac and femoral arteries) (Figure 15 and 16A). In contrast, Ang II-induced vasoconstriction was reduced in basilar arteries from SOCS3 +/- mice compared to those from SOCS3 +/+ mice (Figure 16B). U46619-induced vasoconstriction were similar in all vessels (abdominal aorta, iliac and femoral arteries, and cerebral arteries) from all groups. (Figure 17 and 18).

56 41 Discussion There are several major new findings in the studies outlined in this Chapter. First, partial genetic haplodeficiency in SOCS3 in all cells had no effect on vascular responses, including endothelial function, under normal conditions. In contrast, SOCS3 haplodeficiency augmented direct effects of Ang II on endothelial function, suggesting that under normal conditions, SOCS3 protects against local effects of Ang II on the vessel wall. With reduced expression of SOCS3, arteries were predisposed to Ang II-induced dysfunction. Second, endothelial dysfunction in response to Ang II in SOCS3 haplodeficient mice was prevented by inhibitors of NF-κB, IL-6, or STAT3. These results suggest that SOCS3 normally inhibits Ang II-induced pro-inflammatory signaling pathways. Third, detrimental effects of Ang II on endothelial function in SOCS3 haplodeficient mice was abolished by a superoxide scavenger, which suggests that oxidative stress was also involved in augmented effects of Ang II in SOCS3 haplodeficient mice. Fourth, although effects of Ang II on endothelial function were mediated by the AT1R, acute vasoconstrictor responses to Ang II were similar in WT and SOCS3 haplodeficient mice, suggesting that differences in the sensitivity of the receptor do not account for strain differences in effects of Ang II. To my knowledge, these findings provide the first evidence regarding a role for SOCS3 in vascular homeostasis. Collectively, the results suggest that SOCS3 +/- arteries are predisposed to Ang II-induced endothelial dysfunction, which may result from enhanced inflammatory responses and oxidative stress caused by SOCS3 haplodeficiency. Local Effects of Ang II on Vascular Function Are Augmented by Genetic Haplodeficiency in SOCS3 Both systemic (circulating) and locally derived (tissue-originated) Ang II are thought to contribute to detrimental effects of the peptide hormone on vascular function (92).

57 42 Previously, our group, as well as others, described direct effects of Ang II on endothelial function using an organ culture/vessel incubation model. With that approach, arteries or aorta are isolated and treated with Ang II (or vehicle, in the absence or presence of select inhibitors) in vitro for a number of hours (272, 273, 302). It is well known that circulating Ang II affects multiple organ systems. The use of this incubation model allow us to investigate the direct effect of Ang II on the vessel wall and its resident cells, without activating cells in other organs systems, including the kidney, central nervous system, or the immune system. The approach also avoids the potential confounding effects of increases in arterial pressure when sufficient concentrations of angiotensin are administered systemically or centrally. We have shown previously that Ang II produces concentration-dependent effects on endothelial function using this incubation model (272, 273, 311). In the present study, carotid arteries were incubated with a relatively low concentration of Ang II (1 nm), which does not alter vascular or endothelial function under normal conditions in vessels from control (WT) mice (237, 311). We used this concentration of Ang II in the current experiments because we anticipated that direct effects of Ang II may be augmented in SOCS3 haplodeficient mice. A major finding of these experiments was that in arteries from SOCS3 haplodeficient mice, this low concentration of Ang II significantly reduced acetylcholine-induced relaxation. In contrast, endothelial-independent relaxation to nitroprusside was similar in all groups. These results suggest that genetic haplodeficiency in expression of SOCS3 in the vessel wall predispose carotid arteries to Ang II-induced endothelial dysfunction when incubated in vitro. Previous studies provided strong evidence to support a critical role for IL-6/STAT3 signaling in Ang II-induced endothelial dysfunction (231, 253). SOCS3 is a natural endogenous inhibitor of IL-6/STAT3 signaling and has been shown to effectively prevent IL-6 signaling dependent effects in other models (292, 293). In relation to vascular disease

58 43 and hypertension, previous studies have shown that vascular expression of SOCS3 is significantly increased in response to chronic systemic administration of Ang II (272, 312). While such findings raised questions regarding the role of SOCS3 under these conditions, the functional importance of SOCS3 was not evaluated previously. In cultured MHCC97H cells, SOCS3 sirna enhanced the effects of Ang II on production of vascular endothelial growth factor (VEGF), angiopoietin-2 (Ang-2) and Tie-2 (313), which suggested that SOCS3 played a role in preventing angiogenic effects of Ang II under normal conditions. In contrast to these studies of cells in culture, the current experiments provide the first evidence that endogenous SOCS3 normally protects against detrimental effects of Ang II on endothelial function in intact arteries. Genetic Haplodeficiency in SOCS3 Promotes Oxidative Stress in Response to Ang II A large literature suggests that Ang II promotes oxidative stress in vascular cells (83). For example, endothelial dysfunction in carotid arteries and other vessels in response to Ang II are mediated in large part by activation of NADPH oxidase and production of superoxide (273) (83). Thus, pharmacological or genetic approaches that scavenge superoxide or reduce activity of NADPH oxidase protect the vasculature from effects of Ang II ( ). A key mechanism underlying Ang II-induced endothelial dysfunction involves decreased bioavailability of NO as a result of increased ROS production and interactions between NO and superoxide (119). Consistent with these concepts, I obtained evidence in the current studies that Ang II-induced endothelial dysfunction in arteries from SOCS3 haplodeficient mice was also mediated through oxidative stress. In order to evaluate the contribution of oxidative stress in the current model, a SOD mimetic (tempol) was used to scavenge superoxide (314). Due to its effectiveness in reducing oxidative stress, tempol has been widely used as a scavenger

59 44 of superoxide in cardiovascular studies. In SOCS3 haplodeficient mice, tempol normalized endothelial function in arteries that has been treated locally with Ang II. Previously, our laboratory found that expression of IL-6 is essential for Ang II to increase vascular superoxide levels (273). In IL-6 deficient mice, Ang II-induced increased superoxide were abolished (273). In additional related experiments, preincubation of VSMCs in culture with IL-6 enhanced Ang II-induced production of ROS (315). Furthermore, increases in vascular superoxide in response to Ang II in the incubation model used in the current experiments were prevented by inhibition of STAT3 (272). Collectively, these data support the concept that normal IL-6/STAT3 signaling is required for Ang II-induced oxidative stress in vasculature. In addition to the functional changes described above, I detected a decrease in mrna levels for enos, SOD1 and SOD2 in aorta from SOCS3 +/- mice. SODs are enzymes which catalyze the dismutation of superoxide and thus prevent oxidative stress (119). SOD1 is widely expressed in subcellular compartments including in cytosol and nucleus, whereas SOD2 is expressed in mitochondria. Previously, our group reported that haplodeficiency in Mn-SOD exacerbated endothelial dysfunction in male mice treated with Ang II (316) and in apolipoprotein E deficient mice (317). In addition, endothelialdependent relaxation was reduced in aged mice with SOD1 haplodeficiency compared with WT littermates (318). Furthermore, genetic deficiency in SOD1 augmented Ang II-induced endothelial dysfunction, whereas increased expression of SOD1 prevented Ang II-induced endothelial dysfunction (307). Together, these data provided evidence that reductions in SOD1 or SOD2 could enhance Ang II-induced oxidative stress. Collectively, these findings imply that reduced expression of SOD1 and SOD2 in SOCS3 +/- arteries may have contributed to the observed effects of Ang II in the current experiments.

60 45 NO, produced by enos, is a major mediator of endothelial-dependent relaxation (36). Reductions in enos expression and/or activity result in decreased NO bioavailability and thus may contribute to endothelial dysfunction in vascular disease. Thus, the observed reduced expression of enos in SOCS3 +/- aorta could potentially contribute to impaired responses to acetylcholine following treatment with Ang II. Evidence for Increased Inflammation-Related Signaling in Arteries from SOCS3 Haplodeficient Mice Treated with Ang II In vascular disease, oxidative stress and inflammation interact through mutual feedforward mechanisms (319). In other words, oxidative stress promotes local inflammation and activation of pro-inflammatory responses enhances ROS production and oxidative stress. For example, Ang II activates NF-κB and increases expression of IL-6 (245, 272, 273, 320). In addition to Ang II, IL-6 promotes activation of AT1R and NADPH oxidase (315, 321). In the present Chapter, my results suggest that enhanced Ang II-induced endothelial dysfunction in SOCS3 haplodeficient mice was dependent on NF-κB/IL- 6/STAT3 signaling, since inhibitors of NF-κB, IL-6, and STAT3 largely prevented local effects of Ang II on endothelial function. NF-κB is a ubiquitously expressed transcription factor that regulates a large array of genes, including many thought to be important in vascular pathophysiology (322). As noted above, Ang II activates NF-κB in both endothelial cells and VSMCs (323, 324). Ang II-induced proliferation and migration of VSMCs was effectively arrested by inhibition of NF-κB (325). In a double transgenic rat model which overexpresses elements of the human RAS, chronic systemic inhibition of NF-κB decreased blood pressure, reduced cardiac hypertrophy, ameliorated vascular injury in the heart and kidney, and reduced death related to end-organ damage (326). Activation of NF-κB has also been linked to age-associated

61 46 endothelial dysfunction in humans (327). However the functional importance of NF-κB in a model that examine direct effects of Ang II on endothelial function has not been investigated previously to my knowledge. Thus, the results from the current study help fill a knowledge gap in that they demonstrate that local effects of Ang II on endothelial function in arteries from SOCS3 +/- mice was prevented by an inhibitor of NF-κB. Based on previous findings from IL-6 deficient mice (273) and the current results with the IL-6 neutralizing antibody, it is somewhat surprising that incubation of vessels with Ang II did not result in a detectable increase in mrna levels for IL-6. One possibility is that changes in expression of IL-6 did occur but were restricted predominantly to endothelial cells. Under those conditions, mrna levels measured using whole aorta may greatly underestimate changes that occurred selectively in endothelium. In addition, vascular mrna levels of IL-6 were not altered in SOCS3 haplodeficient mice under the basal conditions, which may indicate that the regulatory effects of SOCS3 was downstream of IL-6 signaling, with no feedback control on the expression of IL-6. Effect of SOCS3 Haplodeficiency on Vasoconstrictor Responses Although the levels of expression were relatively low with threshold cycle (Ct) values around 32, I detected an increase in mrna for the AT1R in aorta from SOCS3 +/- mice compared with vessels from SOCS3 +/+ littermates. In an attempt to evaluate if these detected differences in expression were functionally important, I performed studies examining acute vasoconstrictor effects of Ang II in aorta and several arteries (iliac, femoral, and basilar). In these experiments, I did not detect any significant differences in responses to Ang II in aorta or in the peripheral arteries from SOCS3 haplodeficient mice. Similarly, vasomotor effects of U46619 were similar in all groups. Interestingly, the constrictive response to Ang II, but not U46619 was significantly reduced in basilar arteries from SOCS3 +/- mice. This finding was unexpected and data on expression of AT1R in

62 47 cerebral arteries from SOCS3 +/- mice was not obtained. It is possible that the expression of AT1R is regulated differently in cerebral versus peripheral arteries. There is precedence for such heterogeneity in receptor expression in previous studies (328). These findings suggest that the level of expression of AT1R in aorta is relatively low, consistent with previous reports (74) and that small differences in relative mrna levels in SOCS3 haplodeficient mice are not functionally important in relation to effects on vasomotor tone. In conclusion, studies in this Chapter evaluated the functional importance of SOCS3 in relation to vascular function under normal conditions and local effects of Ang II on the vessel wall. Direct detrimental effects of Ang II on vascular function were exacerbated in arteries genetically haplodeficient in SOCS3. Ang II-induced endothelial dysfunction in arteries from SOCS3 haplodeficient mice was prevented by inhibiting either pro-inflammatory signaling using inhibitors of NF-κB, IL-6, or STAT3, or oxidative stress using a superoxide scavenger. Collectively, these results suggest that SOCS3 protects against direct effects of Ang II on endothelial function under by regulating IL-6/STAT3 signaling and oxidative stress.

63 48 A B Figure 2. Angiotensin II type 1 receptor mediated local effects of angiotensin II on endothelial function. Vasodilatation induced by acetylcholine (A) and sodium nitroprusside (B) in carotid arteries incubated with either vehicle or angiotensin II (AngII, 10 nm) with or without losartan for 22 hours. N=6. Values are mean ± SE, *P<0.05 versus vehicle.

64 49 A B Figure 3. Effects of SOCS3 haplodeficiency on local angiotensin II-induced endothelial dysfunction. Acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II, 1 nm). N=6. Values are mean ± SE, *P<0.05 versus vehicle.

65 50 A B Figure 4. Endothelium-independent relaxation was not affected by angiotensin II or SOCS3 haplodeficiency. Sodium nitroprusside-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II, 1 nm). N=6. Values are mean ± SE, *P<0.05 versus vehicle.

66 51 A B Figure 5. Effects of SOCS3 haplodeficiency on vasoconstriction. U46619-induced contraction in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II, 1 nm). N=6. Values are mean ± SE, *P<0.05 versus vehicle.

67 52 A B Figure 6. Angiotensin II-induced endothelial dysfunction in SOCS3 +/- arteries was mediated by oxidative stress. Acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II, 1 nm). The role of oxidative stress was examined by acute administration of a superoxide scavenger (tempol 1 mm). N=6. Values are mean ± SE, *P<0.05 versus vehicle.

68 53 A B Figure 7. Endothelium-independent relaxation was not affected by angiotensin II or superoxide. Sodium nitroprusside-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II). The role of oxidative stress was examined by acute administration of a superoxide scavenger (tempol 1mM). N=6, Values are mean ± SE, *P<0.05 versus vehicle.

69 54 A B Figure 8. The role of NF-κB in angiotensin II-induced endothelial dysfunction in arteries from SOCS3 +/- mice. Acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or Angiotensin II (Ang II) with or without an inhibitory peptide of NF-κB (NBD, 10 µm). N=6, Values are mean ± SE, *P<0.05 versus vehicle

70 55 A B Figure 9. Endothelium-independent relaxation is not affected by angiotensin II or inhibition of NF-κB. Sodium nitroprusside-induced endothelium-independent relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II) with or without an inhibitory peptide of NF-κB (NBD, 10 µm). N=6. Values are mean ± SE.

71 56 A B Figure 10. Angiotensin II-induced endothelial dysfunction was mediated by IL-6. Acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II) with or without an IL- 6 neutralizing antibody (anti-il-6, 10 µg/ml). N=6. Values are mean ± SE, *P<0.05 versus vehicle

72 57 A B Figure 11. Endothelium-independent relaxation of arteries was not affected by angiotensin II or IL-6. Sodium nitroprusside-induced endothelium-independent relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II) with or without an IL-6 neutralizing antibody (anti-il-6, 10 µg/ml). N=6. Values are mean ± SE.

73 58 A B Figure 12. Angiotensin II-induced endothelial dysfunction in SOCS3 +/- arteries was dependent on activation of STAT3. Acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II) with or without a small molecule inhibitor of STAT3 (S3I-201, 10 µm). N=6. Values are mean ± SE, *P<0.05 versus vehicle.

74 59 A B Figure 13. Endothelium-independent relaxation was not affected by angiotensin II or inhibition of STAT3. Sodium nitroprusside-induced endothelium-independent vasodilation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice incubated with either vehicle or angiotensin II (Ang II) with or without a small molecule inhibitor of STAT3 (S3I-201, 10 µm). N=6, Values are mean ± SE.

75 Figure 14. Vascular expression of select genes implicated in vascular disease. Gene expression in aorta from SOCS3 +/+ (WT) and SOCS3 +/- (HET) mice incubated with either saline (Ve) or angiotensin II (AngII, 1 nm) for 22 hours. Data expressed as mean ± SEM. N=6. *P<0.05 versus SOCS3 +/+ (WT) + saline (Ve). 60

76 61 A B Figure 15. Effects of SOCS3 haplodeficiency on angiotensin II-induced contraction in abdominal aorta and iliac artery. Angiotensin II-induced contraction in abdominal aorta (A) and iliac arteries (B) from both SOCS3 +/+ (N=5) and SOCS3 +/- mice (N=7). Values are mean ± SE.

77 62 A B Figure 16. Effects of SOCS3 haplodeficiency on angiotensin II-induced contraction in femoral and changes in vessel diameter in cerebral arteries. Angiotensin II-induced contraction in femoral (A) and vasoconstriction in cerebral arteries (B) from SOCS3 +/+ (N=5) and SOCS3 +/- mice (N=7). Values are mean ± SE, *P < 0.05 versus SOCS3 +/+.

78 63 A B Figure 17. Effects of SOCS3 haplodeficiency on U46619-induced contraction in abdominal aorta and iliac artery. U46619-induced contraction in abdominal aorta (A) and iliac arteries (B) from SOCS3 +/+ (N=5) and SOCS3 +/- mice (N=7). Values are mean ± SE.

79 64 A B Figure 18. Effects of SOCS3 haplodeficiency on U46619-induced contraction in femoral and changes in diameter in cerebral arteries. U46619-induced contraction in femoral (A) and vasoconstriction in cerebral arteries (B) from both SOCS3 +/+ (N=5) and SOCS3 +/- mice (N=7). Values are mean ± SE.

80 65 CHAPTER 3 SOCS3 HAPLODEFICIENCY PROTECTS MICE FROM ENDOTHELIAL DYSFUNCTION PRODUCED BY SYSTEMIC ADMINISTRATION OF ANGIOTENSIN II

81 66 Introduction The Global Burden of Disease Study in 2010 reported that hypertension was the leading risk factor for disease worldwide (4). Among its effects, hypertension is a major risk factor for cardiovascular disease, stroke, and cognitive impairment (329, 330). Although hypertension is a worldwide health issue with more than 30% of the U.S. adult population being affected, the fundamental mechanisms that underlie development of hypertension and hypertension-associated end-organ damage have not been fully defined. The regulation of blood pressure involves multiple organ systems including the kidney, the central nervous system, and the vasculature, as well as the immune system. All these systems are now known to be key contributors to the onset and progression of hypertension (7-10). Extensive evidence from both basic and clinical research support the critical involvement of the RAS in the pathogenesis of hypertension, as well as other major cardiovascular diseases (44, 45). Ang II, the primary biological effector of the RAS, generally impairs vascular function and promotes the progression of vascular disease. Many of these effects are mediated via the AT1R. Pharmacological inhibition of the RAS, by either blocking Ang II synthesis with ACE inhibitors or preventing its binding to AT1R with selective receptor antagonists, are common therapeutic strategies to control blood pressure and to reduce the risk of cardiovascular events (51-53). Beyond these aspects, there are still uncertainties regarding how the RAS contributes to the initiation and progression of vascular disease. Vascular inflammation is now thought to play a critical role in promoting vascular disease and hypertension ( ). Accumulating studies indicate that Ang II promotes several key events in the inflammatory process (335). These events include increasing vascular permeability ( ), production of ROS ( ) and cytokines/chemokines (190, ), activation of toll-like receptor 4 ( ), as well as activation and infiltration of immune cells ( ).

82 67 Interleukin-6 (IL-6) is a multifunctional cytokine expressed by vascular cells and macrophages in response to Ang II (190, 271, 315, 347). Circulating IL-6 levels are positively correlated with cardiovascular disease progression (315, 357). Typically, IL-6 signaling promotes inflammatory responses through JAK/STAT3 signaling. In experimental studies, Ang II induced oxidative stress, endothelial dysfunction, and hypertension were attenuated in IL-6 deficient mice ( , 358). In addition, Ang IIinduced hypertension and endothelial dysfunction were prevented in normal mice by a small molecule inhibitor of STAT3 (272). Collectively, these findings suggest that IL- 6/STAT3 signaling contributes to Ang II-induced vascular disease and hypertension. Despite the findings outlined above, the role of IL-6 signaling in inflammation has controversial elements. Multiple studies over the past 20 years have reported antiinflammatory effects attributed to IL-6/STAT3 signaling ( ). Frequent and widespread vascular abnormalities (including aneurysm and vascular remodeling) were found in humans with STAT3 deficiency (279). IL-6 elicited neuroprotective effects against cerebral ischemic injury by activating STAT3 signaling (280). Furthermore, IL-6 protected against progression of atherosclerosis, which has a major inflammatory component (283, 284). Most recently, an anti-inflammatory role for IL-6 signaling was described in myeloid cells, which limited endotoxemia and obesity-associated resistance to insulin (285). SOCS3 is a negative regulator of IL-6/STAT3 signal transduction (293). Although the role of SOCS3 in negatively regulating the IL-6 signaling is well known in immunology and cancer biology, the importance of SOCS3 in vascular diseases and hypertension has not been defined. Because several studies indicated SOCS3 is a determinant of pro- or antiinflammatory responses to IL-6 ( ), it is difficult to predict the impact of SOCS3 in Ang II-induced vascular abnormalities and hypertension. In this Chapter, we studied the role of SOCS3 using an in vivo model of chronic Ang II-dependent hypertension. In addition, since Ang II can promote vascular dysfunction independent of elevated blood pressure (359), we also examined the role of SOCS3 in

83 68 response to a sub-pressor dose of Ang II. To our surprise, and in contrast to studies examining direct effects of Ang II on the vessel wall (Chapter 2), we found that SOCS3 haplodeficiency prevented the majority of Ang II-induced endothelial dysfunction without affecting the pressor response to Ang II. These protective effects of SOCS3 haplodeficiency were seen in carotid arteries as well as small resistance vessel. Overall, these findings suggest that the impact of SOCS3 is context dependent in relation to local versus systemic effects of Ang II. Methods Animals Complete genetic haplodeficiency in SOCS3 is lethal (304). SOCS3 +/- mice, which were generated previously by deleting the exon containing the entire coding region of the gene to create a completely null mutation (304), are phenotypically normal under unstressed conditions (304). Animals were obtained by breeding SOCS3 +/- with C57BL/J6 mice. Male and female SOCS3 +/- mice and SOCS3 +/+ littermates (~4-6 months of age) were used in current studies. Mice were fed regular chow and water, and maintained under standard housing conditions. All studies followed the Guide for the Care and Use of Laboratory animals and approved by the Institutional Animal Care and Use Committee at the University of Iowa. Chronic Angiotensin II-Dependent Hypertension Chronic administration of Ang II is one of the most common methods used to study effects of hypertension in genetically altered mice (360, 361). Sterile techniques were used for preparing osmotic pumps (Alza, Palo Alto, CA, model 1002) filled with either 0.9% saline or Ang II [to deliver concentrations of either 1.4 (pressor) or 0.28 (non-pressor) mg/kg per day of Ang II]. To implant osmotic pumps subcutaneously, mice were

84 69 anesthetized with ketamine/xylazine (87.5 and 12.5 mg/kg, i.p.). A small incision was made in the skin between the scapulae, and a small pocket was formed by separating the subcutaneous connective tissue. The pump was inserted into the pocket with the flow moderator pointing away from the skin incision, which was then closed with sutures. Vasomotor Function Two weeks after osmotic pump implantation, mice were euthanized with pentobarbital (100 mg/kg I.P.). As described previously in Chapter 2, arteries were suspended in organ baths containing oxygenated (95% oxygen and 5% carbon dioxide) Krebs solution maintained at 37 C. Following 45 minutes equilibration, relaxation of carotid arteries in response to acetylcholine and sodium nitroprusside were measured after precontraction (50-60% of maximum) induced by U46619, a thromboxane A2 agonist. To determine if vascular dysfunction in arteries treated with Ang II is mediated by superoxide, tempol (1 mm, a superoxide scavenger) was added to the organ baths 30 minutes before vessel precontraction and maintained throughout the study. At the end of each of these experiments, a full dose response curve was performed using U46619 (0.03 to 3 μg of U46619 per 20 ml). For experiments studying endothelial function in a small resistance vessel, the basilar artery was isolated from the brain and cannulated onto glass micropipettes filled with Krebs buffer in an organ chamber. After being pressurized to 60 mmhg, lumen diameter was measured by analyzing the images projected on a video monitor using an electronic dimension analyzer. Following 30 minutes of equilibration, responses of basilar arteries to 50 mmol/l KCl were first measured. To study endothelial function, changes in the lumen diameter in response to acetylcholine and nitroprusside were measured after preconstriction induced by U46619 (~60% of the constriction induced by 50 mmol/l KCl).

85 70 Blood Pressure Measurements Systolic blood pressure was measured by tail-cuff BP-2000 system (Visitech System, Apex, NC) in conscious mice. Mice were trained for 5 days before osmotic pump implantation. Blood pressure was measured daily in the morning during the 14 days of pump infusion. Prior to these measurements, the platform and mouse restrainers were prewarmed at 38 C and this temperature was maintained throughout the course of measurement. After 10 preliminary measurements, the average value of 20 measurements taken during a 30 minute period were collected. Real-Time RT-PCR Aorta was also sampled and aortic homogenates were stored in TRIzol reagent (Invitrogen) and purified using the RNAeasy kit (Qiagen). The concentration of RNA was determined using a NanoDrop ND cdna was generated by RT-PCR using random hexamers. mrna levels of SOCS3 (Mm _g1, Life Technologies), enos (Mm _m1, Life Technologies), AT1R (Mm _s1, Life Technologies), SOD1 (Mm _g1, Life Technologies), SOD2 (Mm _m1, Life Technologies), were determined by qrt-pcr using the TaqMan system (Applied Biosystems). β-actin (Applied Biosystems) was used as internal control within each sample. Relative fold expression was calculated using the ΔΔCt method (306). Statistical Analysis Responses to acetylchoine and nitroprusside was presented as a percent relaxation to precontraction induced by U Contractions were presented as grams of tension. For studies in basilar artery, data were presented as a percent change in vessel diameter from the baseline level. Results were expressed as the mean ± SEM and compared by

86 71 Student t test or one-way ANOVA followed by Student-Newman-Keuls post hoc test as appropriate. Differences were considered significant at P Results SOCS3 Haplodeficiency Protected against Endothelial Dysfunction in Carotid Arteries during Chronic Ang II-Dependent Hypertension To investigate the functional importance of SOCS3 in vivo, chronic hypertension was produced by infusing a pressor concentration of Ang II (or vehicle) systemically using osmotic minipumps for 14 days. Acetylcholine produced complete relaxation in carotid arteries from SOCS3 +/+ mice treated with saline (Figure 19A). Ang II infusion reduced acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ mice by ~ 50%, an effect that was partially prevented by tempol (Figure 19A). These findings are consistent with previous studies (238, 272, 273). Acetylcholine-induced relaxation in the group treated with saline was not affected by tempol (Figure 19A). Unexpectedly, the majority of Ang II-induced endothelial dysfunction was prevented in SOCS3 +/- mice (Figure 19B). Relaxation of arteries to acetylcholine was similar in saline-treated groups with or without tempol (Figure 19B). Responses to nitroprusside and U46619 were similar in all groups (Figure 20 and 21). Protective Effects of SOCS3 Haplodeficiency Extend to Resistance Arteries in the Brain To evaluate the functional importance of SOCS3 in a small resistance artery, basilar arteries were isolated from saline and Ang II-infused mice. Vasodilation of basilar arteries from SOCS3 +/+ mice to acetylcholine was reduced by ~50% in the Ang II treated group compared those treated with saline (Figure 22A). Consistent with the findings in carotid arteries, SOCS3 haplodeficiency protected against Ang II-induced endothelial dysfunction

87 72 in basilar arteries (Figure 22B). In contrast, responses to nitroprusside were similar in both groups (Figure 23). SOCS3 Haplodeficiency Did Not Alter the Pressor Response during Ang II-Induced Hypertension To investigate the impact of SOCS3 in relation to regulation of blood pressure, changes in arterial pressure were evaluated in SOCS3 +/+ and SOCS3 +/- mice infused with Ang II or vehicle (1.4 mg/kg per day). Blood pressure was similar in SOCS3 +/+ and SOCS3 +/- mice treated with saline for 14 days. In contrast, Ang II significantly increased blood pressure of SOCS3 +/+ and SOCS3 +/- mice and the increased pressure was maintained to the last day of infusion. Importantly, there is no significant difference between SOCS3 +/+ and SOCS3 +/- mice with regard to the pressor responses to Ang II (Figure 24). Thus, SOCS3 haplodeficiency did not affect the pressor responses of Ang II. SOCS3 Haplodeficiency Does Not Alter Endothelial Function or Blood Pressure during Infusion of a Non-Pressor Dose of Ang II To define the functional importance of SOCS3 in relation to endothelial function during infusion of a non-pressor dose of Ang II, SOCS3 +/+ and SOCS3 +/- mice were treated with Ang II (0.28 mg/kg per day) for 14 days. As shown in Figure 25, this dose of Ang II did not alter the systolic blood pressure in either strain of mice. In addition, acetylcholine and nitroprusside-induced vasodilation of carotid arteries were similar in all groups (Figure 26 and 27).

88 73 Effects of Systemic Ang II on Vascular Expression of Genes Implicated in Vascular Disease In an attempt to obtain additional insight into mechanisms, we measured aortic mrna levels of select genes implicated in vascular disease (Figure 28). Consistent with the findings in Chapter 2, haplodeficiency of SOCS3 results in a ~50% reduction of SOCS3 mrna in aorta. Pressor dose of Ang II infusion increased SOCS3 mrna level in aorta from SOCS3 +/+ mice, but not in SOCS3 +/- mice. SOCS3 haplodeficiency was associated with increased expression of AT1Ra, and reduced aortic expression of enos, SOD1 and SOD2, genes important for NO-dependent signaling, oxidative stress, and vasodilation. Higher expression of AT1Ra and lower expression of SOD1 were observed in aortas from both SOCS3 +/+ and SOCS3 +/- mice treated with Ang II, compared with those from vehicle treated mice. Discussion There are several new findings in the studies outlined in this Chapter. First, chronic systemic administration of Ang II produced a sustained increase in blood pressure in SOCS3 +/+ mice. Both baseline arterial pressure and the pressor response to Ang II was not altered in mice that were globally haplodeficient in SOCS3, suggesting that SOCS3 does not protect against pressor effects of Ang II when the peptide is administered systemically. In addition, the effect of a non-pressor dose of Ang II on blood pressure was similar in both stains of mice. Second, systemic administration of Ang II impaired endothelial function in carotid arteries from SOCS3 +/+ mice. Confirming previous work from our laboratory and others, this impairment in SOCS3 +/+ mice was largely prevented by a scavenger of superoxide. Third, systemically delivered Ang II had very modest effects on endothelial function in SOCS3 +/- mice. This protective effect of SOCS3 haplodeficiency on the vasculature extended to a resistance artery supplying brain. Considering the results of the

89 74 previous Chapter, these results in SOCS3 haplodeficient mice were surprising and suggested that global haplodeficiency in SOCS3 protected the vasculature against effects of systemically administered Ang II. Collectively, the data to this point provided the first indication that the functional importance of SOCS3 in the vasculature may be context dependent in relation to local (direct) versus systemic effects of Ang II. Role of SOCS3 in an Angiotensin II-Dependent Model of Hypertension The use of osmotic minipumps to deliver a pressor dose of Ang II is an extremely common approach in the literature to produce hypertension in genetically altered mice (360, 361). We used this same approach in the current experiments. Administration of Ang II for two weeks increased arterial blood pressure by approximately 40 mmhg compared with vehicle-infused mice. These changes in blood pressure are similar to those reported in previous publications when the same dose of Ang II was used, including studies that measured blood pressure using radiotelemetry (273, 360, ). A major finding in the current studies was that increases in arterial pressure in response to the pressor dose of Ang II was similar in SOCS3 +/+ and SOCS3 +/- mice. When administered systemically, Ang II can induce vascular dysfunction through both blood pressure-dependent and blood pressure-independent mechanisms (365). In this Chapter, I also tested the functional importance of SOCS3 when a non-pressor dose of Ang II is given chronically. I considered the possibility that blood pressure might increase in SOCS3 +/- mice even if the same dose of Ang II did not increase arterial pressure in SOCS3 +/+ mice. Previous studies demonstrated that a constant infusion of Ang II at a concentration of 0.28 mg/kg per day for 14 days (the dose used in the current experiments) neither elevated blood pressure nor induced endothelial dysfunction in carotid arteries (364, 366, 367). The results I obtained for blood pressure and for vascular function in carotid arteries from SOCS3 +/+ mice are consistent with these previous reports. In general,

90 75 acetylcholine-induced relaxation in response to this non-pressor dose of Ang II were similar in arteries from SOCS3 +/+ and SOCS3 +/- mice. Haplodeficiency in SOCS3 Protects Against Systemic Angiotensin II-Induced Endothelial Dysfunction As discussed in Chapter 1, a critical role for IL-6/STAT3 signaling in endothelial dysfunction has been suggested based on several lines of evidence, including studies examining effects of chronic hypertension induced by a pressor dose of Ang II (272, 273). For example, either administered systemically or locally (using the incubation model), Ang II had little effect on endothelial function in IL-6 deficient mice (273). Similarly, local treatment of arteries with small molecule inhibitors of STAT3 or chronic treatment with an inhibitor of STAT3 systemically protected against the development of Ang II-induced endothelial dysfunction (272). In this Chapter, I evaluated the impact of SOCS3 (a known inhibitor of IL-6/STAT3 signaling in other systems) in systemic Ang II-induced hypertension and endothelial dysfunction using a mouse model with SOCS3 haplodeficiency. The pressor dose of Ang II reduced acetylcholine-induced relaxation of carotid arteries from SOCS3 +/+ mice. Consistent with numerous studies in the literature, this loss of endothelial function was prevented by a scavenger of superoxide. For example, endothelial dysfunction in response to Ang II is prevented by genetic deletion of the Nox2 component of NADPH oxidase or pharmacological scavengers of superoxide (238, 272, 273, 307). Unexpectedly, detrimental effects of Ang II on endothelial function were largely absent in arteries from SOCS3 +/- mice. This protective effects of SOCS3 haplodeficiency

91 76 extended to the basilar artery, a resistance artery supplying the brain. In contrast to the results described in Chapter 2, in which SOCS3 haplodeficiency augmented local effects of Ang II, SOCS3 haplodeficiency exerted protective effects in response to systemic Ang II. In other words, with reduced global expression of SOCS3, systemically administered Ang II had minimal effects on endothelial function. The incubation (organ culture) model as described in Chapter 2 has been used to examine direct effects of Ang II on the vessel wall by us and others (272, 273, 302, 303). This incubation model is analogous to studies which examine effects of Ang II on cells in culture, an extremely common approach over the years. In studies from our own laboratory which evaluated the role of IL-10, STAT3, and IL-6 in relation to effects of Ang II on endothelial function, results obtained were qualitatively similar whether Ang II was administered locally to isolated vessels or systemically in the mouse followed by studies of vessels in vitro (237, 272, 273). For example, genetic deficiency in IL-6 resulted in similar protection of endothelial cell function whether Ang II was administered locally or systemically (273). The present experiments on SOCS3 is the first example in our experience where the results of the two models differed in relation to effects of Ang II on endothelial function. The current study is not the first to suggest that the role of SOCS3 can vary in different models with an inflammatory component ( ). For example, although SOCS3 is often initially considered a suppressor of pro-inflammatory signaling (368), multiple studies have suggested that deficiency in SOCS3 can result in anti-inflammatory effects. Considering that IL-6 can elicit both pro- and anti-inflammatory effects (369), my results are consistent with the concept that SOCS3 may function as a determinant of pro or anti-inflammatory effects of Ang II and perhaps downstream IL-6 signaling (286, ). In relation to these concepts, it is interesting that Ang II-induced hypertension was exaggerated in bone marrow-specific AT1R deficient mice (370), which suggests a

92 77 protective effect of Ang II signaling in immune cells. This protective effect may involve activating IL-6 anti-inflammatory signaling. In addition to the data discussed above, I consistently detected decreased relative mrna expression for enos, SOD1, and SOD2 as well as increased mrna of AT1R in aorta. As mentioned in Chapter 2, SODs are antioxidant enzymes responsible for the removal of superoxide and thus have protective effects against oxidative stress (119). As suggested in previous studies (307, ), a reduction in either SOD1 or SOD2 can exacerbate detrimental effects of Ang II on vascular function and oxidative stress (307, ). In addition, enos is the major enzyme responsible for producing NO, a critical mediator of endothelial-dependent relaxation (36). Reductions in enos expression could decrease NO bioavailability and therefore impair NO-dependent vascular responses. However, it seems unlikely that these changes in vascular gene expression could explain the protective vascular effects of SOCS3 haplodeficiency during chronic Ang II-dependent hypertension. Rather I would speculate at this point, that the protective effects of SOCS3 haplodeficiency may result from effects in non-vascular cells, particularly circulating immune cells. In conclusion, studies outlined in this Chapter investigated the role of SOCS3 in vivo using a chronic Ang II-dependent hypertension model as well as a model that tests effects of Ang II independent of elevated blood pressure. SOCS3 haplodeficiency had no detectable impact on blood pressure changes to either the pressor or non-pressor doses of Ang II. In addition, there were no differences between SOCS3 +/+ and SOCS3 +/- mice regarding the effects of a non-pressor dose of Ang II on vascular function. To our surprise, most of the detrimental effects of the pressor dose of Ang II on vascular function in arteries from SOCS3 +/+ mice were absent in SOCS3 +/- mice. These protective effects of SOCS3 contrast to the role of SOCS3 when vessels are challenged with local Ang II as described in Chapter 2. Collectively, these results suggested that the role of SOCS3 in vascular

93 78 function is context dependent in relation to local versus systemic effects of Ang II. This major new finding raised questions regarding the role of SOCS3 in non-vascular cells, particularly in circulating bone marrow-derived cells, which lead me to examine the importance of bone marrow-derived effects of SOCS3 in the next Chapter.

94 79 A B Figure 19. Endothelium-dependent relaxation in chronic angiotensin II-dependent hypertension. Acetylcholine-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice systemically administrated with vehicle (saline) or Angiotensin II (1.4 mg/kg per day) for 14 days. The role of oxidative stress was examined by acute administration of a superoxide scavenger tempol (1 mm). N=6. Values are mean ± SE.

95 80 A B Figure 20. Endothelium-independent relaxation was not significantly affected by chronic angiotensin II-dependent hypertension or tempol. Sodium nitroprusside-induced relaxation in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice systemically administrated with vehicle (saline) or Angiotensin II (1.4 mg/kg per day) for 14 days. The role of oxidative stress was examined by acute administration of a superoxide scavenger tempol (1 mm). N=6. Values are mean ± SE.

96 81 A B Figure 21. The role of SOCS3 in contraction of arteries in a chronic angiotension II-dependent hypertension model. U46619-induced contraction in carotid arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice systemically administrated with either vehicle (saline) or angiotensin II (1.4 mg/kg per day) for 14 days. The role of oxidative stress was examined by acute administration of a superoxide scavenger tempol (1 mm). N=6, Values are mean ± SE, *P < 0.05 versus vehicle.

97 82 A B Figure 22. The protective role of haplodeficiency in SOCS3 extends to basilar arteries. Vasodilation induced by acetylcholine in basilar arteries from SOCS3 +/+ and SOCS3 +/- mice systemically administrated with either vehicle (saline) or angiotensin II (1.4 mg/kg per day) for 14 days. N=4. Values are mean ± SE, *P < 0.05 versus vehicle.

98 83 A B Figure 23. The role of SOCS3 in endothelium-independent vasodilation in basilar arteries Sodium nitroprusside-induced vasodilation in basilar arteries from SOCS3 +/+ (A) and SOCS3 +/- (B) mice systemically administrated with either vehicle (saline) or angiotensin II (1.4 mg/kg per day) for 14 days. N=4, Values are mean ± SE. *P < 0.05 versus vehicle.

99 Figure 24. The role of SOCS3 in angiotensin II-induced hypertension. Systolic blood pressure measured by tail-cuff in SOCS3 +/+ and SOCS3 +/- mice systemically administrated with either vehicle (saline) or angiotensin II for 14 days. N=6. Values are mean ± SE, *P < 0.05 versus vehicle. 84

100 Figure 25. Systolic blood pressure in SOCS3 +/+ and SOCS3 +/- mice administrated a non-pressor dose of angiotensin II Systolic blood pressure measured using tail-cuff in SOCS3 +/+ and SOCS3 +/- mice systemically administrated angiotensin II for 14 days N=3, Values are mean ± SE, *P < 0.05 versus vehicle. 85

101 86 A B Figure 26. Endothelial-dependent relaxation was not affected by a non-pressor dose of angiotensin II or superoxide in SOCS3 +/+ and SOCS3 +/- mice. Relaxation induced by acetylcholine in carotid arteries from SOCS3 +/+ and SOCS3 +/- mice systemically treated with a non- pressor dose of angiotensin II (0.28 mg/kg per day) for 14 days. The role of oxidative stress was examined by acute administration of a superoxide scavenger tempol (1 mm). N=3. Values are mean ± SE.

102 87 A B Figure 27. Endothelial-independent relaxation was not affected by a non-pressor dose of angiotensin II or superoxide in SOCS3 +/+ and SOCS3 +/- mice. Relaxation induced by sodium nitroprusside in carotid arteries from SOCS3 +/+ and SOCS3 +/- mice systemically treated with a non- pressor dose of angiotensin II (0.28 mg/kg per day) for 14 days. The role of superoxide was examined by acute administration of a superoxide scavenger tempol (1 mm). N=3. Values are mean ± SE.

103 Figure 28. Vascular expression of select genes implicated in vascular disease. Gene expression in aorta from SOCS3 +/+ (WT) and SOCS3 +/- (HET) mice systemically administrated with either vehicle (saline) or angiotensin II for 14 days. N=6. Values are mean ± SE, *P < 0.05 versus vehicle. 88

104 89 CHAPTER 4 SOCS3 HAPLODEFICIENT BONE MARROW-DERIVED CELLS PROTECT AGAINST ANGIOTENSIN II-INDUCED VASCULAR DYSFUNCTION AND HYPERTENSION

105 90 Introduction Hypertension is a leading risk factor for disease worldwide including cardiovascular disease, stroke, and cognitive impairment (1, 2). More than one third of adults have been diagnosed with hypertension globally, while another third are predicted to be in a prehypertension stage, likely to develop overt hypertension within a few years (2, 5). Despite the large number of antihypertensive treatments that are available, the prevalence of uncontrolled hypertension continues to rise (6). Thus, hypertension is an enormous clinical and economic burden globally. Regulation of blood pressure is achieved through an integration of multiple organ systems including the kidney, the heart, the vasculature, and the central nervous system (371, 372). However, the underlying mechanisms by which these various systems interact and contribute to cause hypertension and related end-organ damage have not been completely defined. Over the past 50 years, accumulating evidence supports the concept that the immune system also contributes to the development of hypertension through inflammatory-mediated effects within each of the above systems (202, 203, 371). The immune system is often classified based on two types of fundamental responses, innate and adaptive. The contribution of both these responses in the pathogenesis of hypertension and subsequent end-organ damage have been described (191). The innate system mediates pro-inflammatory responses via effector cells including monocytes/macrophages, antigen-presenting cells and natural killer cells (192). Pharmacological or genetic disruption of the function of these effector cells attenuated Ang II-induced hypertension, endothelial dysfunction, and vascular remodeling ( , 214, 373, 374). Activation of toll-like receptors in both cardiovascular and immune cells triggers the production of pro-inflammatory cytokines and ROS, resulting in low-grade inflammation (203). In contrast, the innate immune system can also elicit anti-

106 91 inflammatory responses consisting of an M2 phenotype in macrophages (375) and production of anti-inflammatory cytokines (238, 311, ). Compared to the innate immune system, the contribution of the adaptive immune system to the pathogenesis of hypertension has been better characterized. Since the first evidence for its role in a model of renal infarction in the 1960s (195, 196), accumulating studies support the role of the adaptive immune system in hypertension. The pivotal role for T or B cells in mediating hypertension received very strong support in a study using a mouse model lacking T and B cells (Rag-1 deficient mice), in which Ang II-induced hypertension, superoxide levels, and endothelial dysfunction were significantly reduced compared to controls (222). In addition, TH17 cells contribute to autoimmune diseases, obesity and cardiovascular diseases by producing IL-17. Ang II-induced hypertension, T cells infiltration, oxidative stress, and endothelial dysfunction, were blunted in mice lacking IL-17a ( , 288, 289, 380). Beside T effector cells (Th1, Th2, and Th17) which promote inflammation, Treg cells suppress the inflammatory process mainly by producing the anti-inflammatory cytokine IL-10. Beneficial effects of adoptive transfer of Treg cells against the detrimental effects of Ang II have been described (229, 230, ). A role of IL-10 in preventing Ang II-induced hypertension, endothelial dysfunction, and oxidative stress was supported by studies using both genetic and pharmacological approaches (238, 311, ). Both innate and adaptive immunities are regulated by cytokines produced by either immune or non-immune cells (241, 385). Select pro-inflammatory cytokines have been implicated in the pathogenesis of hypertension (386). IL-6 is one of these cytokines and has been studied widely in cardiovascular research. Plasma level of IL-6 are positively correlated with blood pressure in patients with essential hypertension (269). The concentrations of IL-6 produced by white blood cells from hypertensive patients in response to lipopolysaccharide are higher than those from a non-hypertensive population

107 92 (270). Mice with IL-6 deficiency are protected from Ang II-induced hypertension and endothelial dysfunction ( ), effects that are prevented in wild-type mice by a small molecule inhibitor of STAT3 (62). STAT3 is a transcription factor characteristically activated by IL-6 (272). While pathophysiological effects of IL-6/STAT3 signaling have been suggested, anti-inflammatory effects of IL-6/STAT3 signaling has also been reported ( ). The mechanisms that dictate either a pro- or anti-inflammatory effect of IL-6 remain unknown but several studies suggest SOCS3 may be the key determinant of these responses in subtypes of immune cells ( ). The functional importance of SOCS3 in vascular disease and hypertension is unknown. Based on these previous findings, the goal of current studies was to investigate how SOCS3 in immune cells contributes to changes in vascular function and blood pressure in a chronic Ang II-dependent hypertension model. Since all immune cells originate from hematopoietic stem cells in bone marrow, part of our approach was to reconstitute the immune cell population in lethally irradiated mice with hematopoietic cells from donor mice using the technique of bone marrow transplantation. Data from previous Chapters suggested that SOCS3 exerts divergent effects depending on whether Ang II was administered locally or systemically. In this Chapter, I investigated potential mechanisms contributing to this context-dependent effect of SOCS3 using bone marrow chimeras, which allowed the study of the contribution of SOCS3 in hematopoietic cells. Interestingly, I found that lethally irradiated wild-type (CD45.1) mice reconstituted with SOCS3 +/- bone marrow were protected from Ang II-induced endothelial dysfunction, while reconstitution of irradiated SOCS3 +/- mice with wild-type (CD45.1) bone marrow exacerbated Ang IIinduced vascular dysfunction. The pressor response to Ang II was also substantially reduced in SOCS3 +/+ mice reconstituted with bone marrow from SOCS3 +/- mice. Finally, hematopoietic cells did not impact the role of SOCS3 in relation to local effects of Ang II on vascular function using the incubation model described in Chapter 2.

108 93 Methods Animals SOCS3 haplodeficient mice were generated previously by deleting the exon containing the entire coding region of the gene to create a completely null mutation (304). Complete genetic deficiency in SOCS3 is lethal (304). In contrast, SOCS3 +/- mice are phenotypically normal under unstressed conditions (304). We obtained breeding pairs of these mice from Dr. Paul Rothman and established a colony of the mice in our laboratory. Animals were obtained by breeding SOCS3 +/- with C57BL/J6 mice (from the Jackson Laboratories). Both male and female SOCS3 +/- mice and SOCS3 +/+ littermates (~4-6 months of age) were used in the current study. CD45.1 mice were obtained from the National Cancer Institute (Frederick, MD, USA). This is a C57BL/6J congenic strain used widely in transplant studies (387). The strain carries the differential pan leukocyte marker CD45.1, while the wild-type C57BL/6 strain expresses CD45.2. CD45.1 and CD45.2 mice are phenotypically identical (388), but are different genetically at the single locus of the ptprc gene which encodes an antigen found in all leukocytes (389). As described before (390), the contribution of CD45.1 and CD45.2 in hematopoiesis were detected using plasma samples and flow cytometry after staining with their specific antibodies respectively. Mice were fed with regular chow and water, and maintained under standard housing conditions. All studies followed the Guide for the Care and Use of Laboratory animals and approved by the Institutional Animal Care and Use Committee at the University of Iowa. Bone Marrow Chimeric Mice As described previously (390), wild-type (WT CD45.1), SOCS3 +/-, and SOCS3 +/+ mice were irradiated with two doses (500 and 450 rad) of whole body X-ray irradiation.

109 94 To isolate bone marrow from donor mice (CD45.1, SOCS3 +/-, and SOCS3 +/+ mice), femur, hips/pelvis, and tibia were removed and stored in sterile RP10 media [RPMI 1640 (Gibco BRL, Grand Island, N.Y.) supplemented with 10% fetal calf serum, 100 U of penicillin per ml, 100 μg of streptomycin per ml, 50 μg of gentamicin per ml, 10 mm HEPES, 2 mm l- glutamine, and 50 μm 2-mercaptoethanol] until bone marrow removal. Bone marrow was isolated in a hood. Bones were placed in a petri dish, then quickly dipped into a second dish containing 95% ethanol. The ethanol was then rinsed off with sterile PBS into a third petri dish. The bones were then placed into a fourth petri dish containing RP10 media. After both tips of the bones were removed using surgical scissors, bone marrow was flushed out using a 27½ gauge needle and 10cc syringe. Isolated bone marrow was then spun down and red blood cells lysed for 30 seconds followed by three washings. Irradiated recipient mice were reconstituted with 5 million total cells per mouse via tail vein injection. Bone marrow chimeras were provided with water containing ampicillin (2 mg/ml) for six weeks and were housed for a minimum of 8 weeks to allow reconstitution of the peripheral immune system (Figure 29). Fluorescence Activated Cell Sorting (FACS) Analysis Peripheral blood samples were collected from the retro-orbital vein. After depleting RBCs using ACK lysis buffer (dd H2O containing 0.15 M of NH4Cl, 1 mm of KHCO3 and 0.1 mm of Na2EDTA), cells were resuspended in FACS buffer (PBS containing 3% fetal calf serum) and incubated with an antibody mixture (anti-mouse CD16/32 for preventing nonspecific Ab binding, anti-mouse CD45.1, and anti-mouse CD45.2) for 30 minutes in a 96-well flat bottom plate. After being washed with FACS buffer twice, cells were fixed with Cytofix/Cytoperm (BD Biosciences). Cells were analyzed by flow cytometry (BD Biosciences), and collected data were analyzed using FlowJo (TreeStar, Ashland, OR).

110 95 Chronic Angiotensin II-Dependent Hypertension As described in Chapter 3, osmotic mini pumps were filled with either 0.9% saline or Ang II (1.4 mg/kg per day). To implant an osmotic pump subcutaneously, mice were anesthetized with ketamine/xylazine (87.5 and 12.5 mg/kg, i.p.). A small incision was made in the skin between the scapulae, and a small pocket was formed by spreading the subcutaneous connective tissues. The pump was inserted into the pocket with the flow moderator pointing away from the skin incision, which was then closed with sutures. Vasomotor Function As described in previous Chapters, two weeks after osmotic pump implantation, mice were euthanized with pentobarbital (100 mg/kg I.P.). Carotid arteries were isolated and mounted on a pair of stainless-steel triangular hooks connected to a force transducer. Vascular rings were suspended in an organ bath containing 20 ml of oxygenated Krebs buffer maintained at 37 C and equilibrated at an optimal resting tension of 0.25 gram for 45 minutes. Relaxation of carotid arteries in response to acetylcholine and nitroprusside were measured after pre-contraction (50-60% of maximum) induced by U At the end of each experiment, vascular contraction was measured by cumulative addition of U Blood Pressure Measurements Systolic blood pressure was measured using a tail-cuff BP-2000 system (Visitech System, Apex, NC) in conscious mice. Mice were trained for 5 days before osmotic pump implantation. Blood pressure was measured daily in the morning during the 14 days of osmotic pump infusion. Prior to these measurements, the platform and mouse restrainers were pre-warmed at 38 C and this temperature was maintained throughout the course of

111 96 measurement. After 10 preliminary measurements, the average value of 20 measurements taken in a 30 minute period were collected. Incubation of Blood Vessels As described previously in Chapter 2, mice were anesthetized with pentobarbital (100 mg/kg IP). Carotid arteries and aorta were removed, cleaned, and cut into segments ~ 5 mm in length in oxygenated Krebs solution. Vessels were then incubated for 22 hours in DMEM growth medium with either vehicle (saline) or Ang II (1 nm; Sigma). Statistical Analysis Responses to acetylcholine and nitroprusside were presented as percent relaxation to precontraction induced by U Vessel contractions were presented as grams of tension. Results were expressed as the mean ± SEM and compared by Student t test or one-way ANOVA as appropriate. For comparison of relaxation in response to maximum dose of acetylcholine, and comparison of blood pressure, two-way ANOVA was used followed by Student T test. Differences were considered significant at P Results Bone Marrow-Derived Cells Haplodeficient in SOCS3 Protect Against Ang II-Induced Endothelial Dysfunction The data in previous Chapters suggested that the functional importance of SOCS3 was context specific, depending on whether Ang II was administrated locally (testing direct effects) or systemically. In order to test our hypothesis that SOCS3 haplodeficiency in bone marrow-derived cells protected against Ang II-induced endothelial function, experiments using bone marrow transplantation were conducted. After recovery from bone marrow transplantation for 8 weeks, reconstitution of cells was confirmed using FACS analysis. As

112 97 shown in Figure 30, 94.3% of bone marrow-derived cells in SOCS3 +/- mice were reconstituted with WT (CD45.1) bone marrow. Greater than 90% of WT (CD45.1) bone marrow cells were replaced by SOCS3 +/- bone marrow (Figure 31). In a WT (CD45.1) to SOCS3 +/+ control group, 96.3% of bone marrow-derived cells in SOCS3 +/+ mice were reconstituted with WT (CD45.1) bone marrow (Figure 32). These data indicated that lethally irradiated mice were successfully reconstituted with donor bone marrow. After reconstitution of bone marrow-derived cells was confirmed, mice were infused with either saline or Ang II (1.4 mg/kg per day) for 14 days using osmotic minipumps followed by examination of vascular function. WT (CD45.1) into SOCS3 +/+ and SOCS3 +/- into SOCS3 +/- bone marrow chimeras exhibited vascular function consistent with non-irradiated controls. Briefly, acetylcholine-induced relaxation in carotid arteries from the WT (CD45.1) into SOCS3 +/+ group was reduced by ~60% (P<0.05) Figure 33A, while the majority of Ang II-induced endothelial dysfunction were prevented in the SOCS3 +/- into SOCS3 +/- group (Figure 33B). Vascular responses to nitroprusside and U46619 were similar in all groups (Figure 34 and 35). Reconstitution of lethally irradiated WT (CD45.1) mice with SOCS3 +/- bone marrow partially protected against detrimental effects of Ang II on endothelial dysfunction (Figure 36A). In contrast, irradiated SOCS3 +/- mice reconstituted with WT (CD45.1) bone marrow exhibited exacerbated Ang II-induced vascular dysfunction (Figure 36B). Sodium nitroprusside induced relaxation of carotid arteries was similar in all groups (Figure 37). Contraction of arteries induced by U46619 was similar in all groups (Figure 38). There were no significant differences regarding U46619-induced contraction among bone marrow transplantation groups treated with either saline or Ang II (Figure 39). For ease of comparison, maximum vascular effects of acetylcholine are presented in Figure 40.

113 98 Bone Marrow-Derived Cells Haplodeficient in SOCS3 Reduced the Pressor Response to Ang II Arterial blood pressure of bone marrow chimeras were compared on day 14 of Ang II or vehicle infusion. As shown in Figure 41, systolic blood pressure was similar between groups infused with saline. Infusion of Ang II elevated blood pressure in both groups. However, the pressor response to Ang II was reduced by ~50% in WT (CD45.1) mice reconstituted with bone marrow from SOCS3 +/- mice, compared with SOCS3 +/- mice reconstituted with WT (CD45.1) bone marrow. These data indicate that haplodeficiency of SOCS3 in bone marrow-derived cells has beneficial effects on blood pressure in a model of Ang II-dependent hypertension. Conceptually, these findings are consistent with the results obtained related to endothelial function in these groups. Direct Effects of Ang II on the Vessel Wall Are Not Affected by Bone Marrow-Derived Cells In order to test if SOCS3 in bone marrow-derived cells affects endothelial function in response to direct effects of Ang II on the vessel wall, carotid arteries were isolated from four groups of mice after bone marrow transplantation [WT (CD45.1) into SOCS3 +/+, SOCS3 +/- into SOCS3 +/-, WT (CD45.1) into SOCS3 +/-, and SOCS3 +/- into WT (CD45.1)]. As described in Chapter 2, vessels were then incubated with either saline or Ang II (1 nm) for 22 hours, followed by examination of endothelial function. Consistent with results described above, this low concentration of Ang II did not produce significant impairment on endothelial function in the WT (CD45.1) into SOCS3 +/+ group (Figure 42A). Interestingly, effects of Ang II on acetylcholine-induced relaxation in the SOCS3 +/- into SOCS3 +/- group was similar to that observed in non-irradiated SOCS3 +/- mice (Figure 42B). Vascular responses to acetylcholine in SOCS3 +/- into WT (CD45.1) and WT (CD45.1) into SOCS3 +/- groups were similar to those in non-irradiated SOCS3 +/+ and SOCS3 +/- mice

114 99 (Figure 45). Responses of arteries to nitroprusside and U46619 were similar in all groups (Figure 43, 44, 46, and 47). These data suggest that bone marrow-derived cells do not affect the functional impact of SOCS3 in relation to direct effects of Ang II on the vessel wall. Discussion There are several major new findings in the studies described in this Chapter. First, after conducting and confirming the efficacy of bone marrow transplantation, I found that bone marrow derived-cells haplodeficient in SOCS3 protected against endothelial dysfunction produced by systemic administration of Ang II. This conclusion is based on the observation that reconstitution with SOCS3 +/- bone marrow markedly attenuated Ang II-induced endothelial dysfunction in lethally irradiated CD45.1 WT mice. Second, lethally irradiated SOCS3 +/- mice reconstituted with wild-type bone marrow exhibited enhanced endothelial dysfunction in response to Ang II. Third, in contrast to the effects seen following systemic administration of Ang II, bone marrow-derived effects of SOCS3 did not alter responses of carotid arteries to local Ang II. Fourth, in addition to protecting against endothelial function, bone marrow derived cells haplodeficient in expression of SOCS3 significantly reduced the pressor response to Ang II. Hematopoietic SOCS3 Haplodeficient Cells Protect Against Ang II-Dependent Endothelial Dysfunction and Hypertension In Chapter 3, I did not detect any differences in the pressor response to Ang II in mice that are globally Haplodeficient in SOCS3 compared to SOCS3 +/+ controls. In addition, arterial pressure was similar in both SOCS3 +/+ and SOCS3 +/- mice when infused with a non-pressor dose of Ang II. In the current study in contrast, I found that the Ang IIinduced elevation in blood pressure was significantly attenuated in irradiated WT (CD45.1)

115 100 mice reconstituted with SOCS3 +/- bone marrow. These results provide direct evidence that the beneficial effects of SOCS3 haplodeficient bone marrow on endothelial function extend to blood pressure. SOCS3 +/- to WT (CD45.1) bone marrow chimeras have SOCS3 +/- bone marrow, which was found to be protective, but a SOCS3 +/+ vasculature, and were not predisposed to the detrimental effects of Ang II. Thus, the reduced blood pressure seen in SOCS3 +/- to WT (CD45.1) bone marrow chimeras may result from lacking Ang II detrimental effects on the vessel wall in combination with protective effects from SOCS3 +/-. Bone marrow-derived stem cells contain two populations of cells, hematopoietic stem cells and mesenchymal stem cells (391). Hematopoietic cells are the precursors to all types of blood cells; while mesenchymal stem cells can differentiate into multiple cell types, including osteoblasts, chondrocytes, adipocytes, and endothelial cells (392), which provide a microenvironment supporting hematogenesis (391, 393). When using established standard protocols for bone marrow transplantation, including the methodology used in these experiments (390), mesenchymal stem cells are not transplanted with hematopoietic stem cells (394, 395). In the current studies, the efficacy of bone marrow reconstitution was evaluated by FACS analysis eight weeks after bone marrow transplantation. Approximately 95% of the peripheral leukocytes in all recipient mice were reconstituted efficiently with donor bone marrow cells. According to previous studies, a reconstitution rate of ~95% is expected and a successful rate in standard procedures of bone marrow transplantation (396, 397). Although the studies to date have not defined the exact subtype(s) of hematopoietic cells that underlie the observed protective effects of bone marrow transfer against Ang IIinduced pressor responses and endothelial dysfunction, I anticipate that the beneficial effects seen most likely result from cells within the leukocyte lineage. A contribution by red blood cells is highly unlikely as those cells were lysed during preparation for the

116 101 transplantation procedure. Based on results from previous studies, macrophages are one of the promising candidates responsible for the beneficial effects of SOCS3 haplodeficiency (285, 286, 295, 296, 310). Leukocytes are the major effector cells within the immune system (398). Among their effects, leukocytes are a primary source of IL-6. In the context of the current experiments, it is noteworthy that both pro- and anti-inflammatory effects of IL-6 signaling have been reported in other models ( ). However, it is still unknown what determines whether IL-6 elicits pro- or anti-inflammatory responses. Results from the current study together with previous findings ( ) support the concept that SOCS3 in select subpopulations of leukocytes may regulate these divergent forms of IL-6 signaling. Understanding the mechanisms by which SOCS3 haplodeficient leukocytes protect against detrimental effects of Ang II could eventually support the development of new therapeutic strategies for hypertension and associated end-organ damage. Local Vascular Effects of Ang II in the Hematopoietic SOCS3 Haplodeficient Model In this Chapter, I found that incubation with a relatively low concentration of Ang II [that does not affect endothelial function in WT mice in my studies (Chapter 2) or in previous studies (272, 273, 302, 303)] did not alter acetylcholine-induced vasodilation in lethally irradiated SOCS3 +/+ mice reconstituted with CD45.1 wild-type bone marrow. Thus, these results are consistent with those that tested local effects of Ang II on vessels from non-irradiated SOCS3 +/+ mice. Similarly, local effects of Ang II on endothelial function of lethally irradiated SOCS3 +/- mice reconstituted with SOCS3 +/- bone morrow were consistent with what we have seen in non-irradiated SOCS3 +/- mice. Both these control experiments suggest that the bone marrow transplantation procedure per se does not alter vascular function. Consistent with these results, other groups also provided evidence indicating that the bone marrow transplantation procedure itself has no significant effect

117 102 on vascular function ( ). These results further suggest that circulating hematopoietic cells did not alter the impact of SOCS3 in relation to local (direct) effects of Ang II on the vessel wall. Collectively, these combined approaches suggest that the endothelial dysfunction produced by Ang II in arteries from SOCS3 +/- mice can be attributed to the local effects of SOCS3 within the vessel wall (within resident vascular cells). Local effects of Ang II on endothelial function in arteries from either wild-type to SOCS3 +/- or SOCS3 +/- to wild-type chimeras were also consistent with their non-irradiated controls. However, conclusions that can be made based on results from these groups may be limited. First, it is possible that local treatment with Ang II produces maximal effects on acetylcholine-induced vasodilation in arteries from SOCS3 +/- mice so detection of additional impairment may be difficult. It seems possible, therefore that reconstitution with wild-type bone marrow may not cause further impairment of endothelial function. Second, with the experimental design I have used, incubation with the relatively low concentration of Ang II does not produce significant impairment of endothelial function in arteries from WT mice. Although bone marrow derived cells from SOCS3 +/- mice was found to be protective in studies that used systemic administration of Ang II, it may be difficult to detect protective effects of SOCS3 +/- hematopoietic cells using this incubation model. In conclusion, studies described in this Chapter evaluated the role of bone marrowderived SOCS3 in both local and systemic effects of Ang II on endothelial function. SOCS3 haplodeficiency in bone marrow derived-cells protected against systemic Ang II-induced endothelial dysfunction in SOCS3 +/+ mice. In addition, adoptive transfer of SOCS3 +/+ bone marrow in SOCS3 +/- mice produced enhanced detrimental effects of Ang II on endothelial function. Furthermore, the pressor response to Ang II was lower in lethally irradiated SOCS3 +/+ mice reconstituted with SOCS3 +/- bone marrow compared with lethally irradiated SOCS3 +/- mice reconstituted with SOCS3 +/+ bone marrow. Finally, results from studies in this Chapter also suggest that effects of SOCS3 in bone marrow-drive cells do

118 103 not alter responses of the vessel wall to local Ang II. Collectively, results from these studies suggest that SOCS3 haplodeficient bone marrow-derived cells are responsible for the protective effects of SOCS3 haplodeficiency seen in a chronic Ang II-dependent hypertension model.

119 Figure 29. Experimental design of bone marrow transplantation. Irradiated recipient mice were reconstituted with bone marrow-derived cells isolated from donor mice. There were four groups of bone marrow transplantation: wild-type (WT, CD45.1) to SOCS3 +/-, SOCS3 +/- to WT (CD45.1), WT (CD45.1) to SOCS3 +/+, SOCS3 +/- to SOCS3 +/-. 104

120 CD WT (CD45.1) to SOCS3 +/- CD45.1 Figure 30. Reconstitution of irradiated SOCS3 +/- mice with WT (CD45.2) bone marrow was confirmed by fluorescence activated cell sorting analysis. A representative result of fluorescence activated cell sorting analysis. X axis represents WT (CD45.1) bone marrow. Y axis represents SOCS3 +/- bone marrow.

121 CD SOCS3 +/- to WT (CD45.1) CD45.1 Figure 31. Reconstitution of irradiated WT (CD45.1) mice with SOCS3 +/- bone marrow was confirmed by fluorescence activated cell sorting analysis. A representative result of fluorescence activated cell sorting analysis. X axis represents WT (CD45.1) bone marrow. Y axis represents SOCS3 +/- bone marrow.

122 CD WT (CD45.1) to SOCS3 +/+ CD45.1 Figure 32. Reconstitution of irradiated SOCS3 +/+ mice with WT (CD45.1) bone marrow was confirmed by fluorescence activated cell sorting analysis. A representative result of fluorescence activated cell sorting analysis. X axis represents WT (CD45.1) bone marrow. Y axis represents SOCS3 +/- bone marrow.

123 108 A B Figure 33. Endothelial function of arteries from WT (CD45.1) to WT and SOCS3 +/- to SOCS3 +/- bone marrow chimeras in chronic angiotensin IIdependent hypertension were consistent with non-irradiated controls. Acetylcholine-induced relaxatiom in carotid arteries from WT (CD45.1) to SOCS3 +/+ (A) and SOCS3 +/- to SOCS3 +/- (B) bone marrow chimeras systemically treated with saline (Vehicle) or angiotensin II (AngII) (1.4 mg/kg per day) for 14 days. N=6. Values are mean ± SE. *P < 0.05 versus vehicle.

124 109 A B Figure 34. Endothelial-independent relaxation of arteries from WT (CD45.1) to WT and SOCS3 +/- to SOCS3 +/- bone marrow chimeras were consistent with nonirradiated controls. Sodium nitroprusside-induced relaxation in carotid arteries from WT (CD45.1) to SOCS3 +/+ (A) and SOCS3 +/- to SOCS3 +/- (B) bone marrow chimeras systemically treated with saline (Vehicle) or angiotensin II (AngII) (1.4 mg/kg per day) for 14days. N=6. Values are mean ± SE. *P < 0.05 versus vehicle.

125 110 A B Figure 35. Contraction of arteries from WT (CD45.1) to WT and SOCS3 +/- to SOCS3 +/- bone marrow chimeras were consistent with non-irradiated controls. U46619-induced contraction in carotid arteries from WT (CD45.1) to SOCS3 +/+ (A) and SOCS3 +/- to SOCS3 +/- (B) bone marrow chimeras systemically treated with saline (Vehicle) or angiotensin II (AngII) (1.4 mg/kg per day) for 14days. N=6. Values are mean ± SE.

126 111 A B Figure 36. Endothelial function of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras in chronic angiotensin IIdependent hypertension. Acetylcholine-induced relaxation in carotid arteries from SOCS3 +/- to WT (CD45.1) (A) and WT (CD45.1) to SOCS3 +/- (B) bone marrow chimeras systemically treated with saline (Vehicle) or angiotensin II (AngII) (1.4 mg/kg per day) for 14days. N=6. Values are mean ± SE. *P < 0.05.

127 112 A B Figure 37. Endothelial-independent relaxation of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras in chronic angiotensin II-dependent hypertension. Sodium nitroprusside-induced relaxation in carotid arteries from SOCS3 +/- to WT (CD45.1) (A) and WT (CD45.1) to SOCS3 +/- (B) bone marrow chimeras systemically treated with saline (Vehicle) or angiotensin II (AngII) (1.4 mg/kg per day) for 14days. N=6. Values are mean ± SE. *P < 0.05 versus vehicle.

128 113 A B Figure 38. Contraction of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras in chronic angiotensin II-dependent hypertension. U46619-induced contraction in carotid arteries from SOCS3 +/- to WT (CD45.1) (A) and WT (CD45.1) to SOCS3 +/- (B) bone marrow chimeras systemically treated with saline (Vehicle) or angiotensin II (AngII) (1.4 mg/kg per day) for 14days. N=6. Values are mean ± SE. *P < 0.05 versus vehicle.

129 114 A B Figure 39. Vascular contractile responses were not affected by bone marrow transplantation or systemic administration of angiotensin II. U46619-induced contraction in carotid arteries from four groups of bone marrow chimeras systemically treated with saline (Vehicle) or angiotensin II (AngII) (1.4 mg/kg per day) for 14days. N=6. Values are mean ± SE.

130 Figure 40. Maximum relaxation to acetylcholine in carotid arteries from bone marrow chimeras Acetylcholine-induced maximum relaxation in carotid arteries from WT (CD45.1) to SOCS3 +/+, SOCS3 +/- to WT, SOCS3 +/- to SOCS3 +/-, and SOCS3 +/- to WT (CD45.1) bone marrow chimeras systemically treated with saline (Vehicle) or angiotensin II (AngII) (1.4 mg/kg per day) for 14 days. N=6. Values are mean ± SE. *P < 0.05 versus vehicle. 115

131 Figure 41. Systolic blood pressure of bone marrow chimeras administrated with angiotensin II. Systolic blood pressure of WT (CD45.1) to SOCS3 +/-, SOCS3 +/- to WT (CD45.1) bone marrow chimeras was measured using tail-cuff on day 14 of angiotensin II (AngII) (1.4 mg/kg per day) infusion. N=6. Values are mean ± SE. *P < 0.05 versus vehicle. 116

132 117 A B Figure 42. Endothelial function in arteries from WT (CD45.1) to SOCS3 +/+ and SOCS3 +/- to SOCS3 +/- bone marrow chimeras incubated with angiotensin II. Acetylcholine-induced relaxation in carotid arteries from WT (CD45.1) to SOCS3 +/+ mice (A) and SOCS3 +/- to SOCS3 +/- mice (B) bone marrow chimeras incubated with either saline (Vehicle) or angiotensin II (Ang II). N=3. Value are mean ± SE. *P<0.05 versus vehicle.

133 118 A B Figure 43. Endothelial-independent relaxation of arteries from WT (CD45.1) to SOCS3 +/+ and SOCS3 +/- to SOCS3 +/- bone marrow chimeras incubated with angiotensin II. Sodium nitroprusside-induced relaxation in carotid arteries from WT (CD45.1) to SOCS3 +/+ mice (A) and SOCS3 +/- to SOCS3 +/- mice (B) bone marrow chimeras incubated with either saline (Vehicle) or angiotensin II. N=3. Value are mean ± SE.

134 119 A B Figure 44. Contraction of arteries from WT (CD45.1) to SOCS3 +/+ and SOCS3 +/- to SOCS3 +/- bone marrow chimeras incubated with angiotensin II. U46619-induced contraction in carotid arteries from WT (CD45.1) to SOCS3 +/+ mice (A) and SOCS3 +/- to SOCS3 +/- mice (B) bone marrow chimeras incubated with either saline (Vehicle) or angiotensin II. N=3. Value are mean ± SE.

135 120 A B Figure 45. Endothelial function of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras incubated with angiotensin II. Acetylcholine-induced relaxation in carotid arteries from SOCS3 +/- to WT (CD45.1) (A) and WT (CD45.1) to SOCS3 +/- (B) bone marrow chimeras incubated with angiotensin II (1 nm). N=3. Value are mean ± SE. *P<0.05 versus vehicle.

136 121 A B Figure 46. Endothelial-independent relaxation of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras incubated with angiotensin II. Sodium nitroprusside-induced relaxation in carotid arteries from SOCS3 +/- to WT (CD45.1) (A) and WT (CD45.1) to SOCS3 +/- (B) bone marrow chimeras incubated with angiotensin II (1 nm). N=3. Value are mean ± SE.

137 122 A B Figure 47. Contraction of arteries from SOCS3 +/- to WT (CD45.1) and WT (CD45.1) to SOCS3 +/- bone marrow chimeras incubated with angiotensin II. U46619-induced contraction in carotid arteries from SOCS3 +/- to WT (CD45.1) (A) and WT (CD45.1) to SOCS3 +/- (B) bone marrow chimeras incubated with either saline (Vehicle) or angiotensin II (1 nm). N=3. Value are mean ± SE.

138 123 CHAPTER 5 CONCLUSIONS

139 124 In the present series of studies, I obtained several lines of evidence which support the concept that SOCS3 exerts functionally important effects in models of Ang II-induced hypertension and vascular dysfunction. The major findings of these studies are summarized schematically in Figure 48. Using a mouse model with genetic haplodeficiency in SOCS3, I found that SOCS3 played divergent roles in relation to local versus systemic effects of Ang II on the vasculature. Endothelial dysfunction produced by local Ang II is augmented in SOCS3 haplodeficient arteries, suggesting under normal conditions, SOCS3 protects vessels against detrimental effects of Ang II. Surprisingly, systemic infusion of a pressor dose Ang II, which induced endothelial dysfunction in arteries from WT mice, caused minimal impairment in arteries from SOCS3 haplodeficient mice, suggesting SOCS3 is required for mediating the actions of Ang-II on the vessel wall. Lastly, results from bone marrow transplantation studies suggested that bone marrow-derived cells are the responsible sites for SOCS3 to mediate the systemic actions of Ang II on vascular function. The Role of SOCS3 in Local Effects of Ang II on Vascular Cells Most of the known pathophysiological actions of Ang II are mediated through AT1R, which is widely expressed in multiple organs, including the vasculature, brain, heart, adrenals, kidneys, as well as organs of the immune system (402). In order to study Ang IIdependent mechanisms specifically in vasculature, an in vitro (organ culture) model has been used by our laboratory as well as other investigators (272, 273, 302, 303, 311). Using this in vitro model allows us to test the direct effects of Ang II on vascular cells and avoids the potential confounding effects due to activation of cells in other organ systems or increased arterial pressure when sufficient concentrations of Ang II are administered systemically or centrally. The vessel wall is composed by multiple cell types, including endothelial cells, VSMCs, and adventitial cells. A limitation of the in vitro model and the current approach is that it did not identify the specific vascular cell type(s) which contributed to the observed functional changes. In the following sections, I will discuss

140 125 potential sites of action and mechanisms by which Ang II may act to ultimately affect the vasculature. The Role of SOCS3 in Effects of Ang II on Endothelial Cells In relation to vascular tone, endothelial dysfunction is most often caused by a decline in NO bioactivity in the vessel wall resulting in decreased vasodilation (403). In the current in vitro model, Ang II within the culture medium directly interacts with endothelial cells. AT1R are expressed in endothelial cells and it is known that Ang IIinduced activation of AT1R in these cells causes reduced NO-mediated responses by generating superoxide (via NADPH oxidase), which chemically reacts with NO and subsequently inhibits the biological activity of NO (123). This Ang II induced-interaction of NO with superoxide anion results in the formation of peroxynitrite, which is a highly reactive oxidant with multiple deleterious effects on endothelial function (123, ). Activation of AT1R by Ang II can also contribute to enos uncoupling in endothelial cells which causes production of superoxide rather than NO (81, 404). In addition, ROS-sensitive ERK activation by Ang II upregulates endothelial expression of endothelin-1, a strong vasoconstrictor which can also contribute to endothelial dysfunction and vascular remodeling (405, 406). Of particular importance, Ang II activates ROS-sensitive NF-κB in endothelial cells, resulting in upregulation of cell adhesion molecules such as vascular cell adhesion molecule 1 (VCAM-1), intercellular adhesion molecule 1 (ICAM-1) (407), and chemokines such as monocyte chemoattractant protein (MCP-1) (408), as well as the proinflammatory cytokine IL-6 (358). As a result, AT1R signaling in endothelial cells plays a key role in Ang II-induced vascular inflammation and dysfunction. When studying vessels in organ culture, increased expression of adhesion molecules is presumably of little physiological significance as blood cells are not present in the model. However, the finding that isolated aorta and carotid arteries produce and release IL-6 into the extracellular media in response to Ang II (273)

141 126 is relevant for the current model. Ang II can directly induce IL-6 expression in endothelial cells (244, 358, 409). Activation of IL-6/STAT3 signaling was demonstrated in both cultured endothelial cells (410) and intimal layers of aortae isolated from low-density lipoprotein receptor-deficient mice infused with Ang II (244). SOCS3 is defined as a suppressor of IL-6/STAT3 signaling and expression of SOCS3 has been detected in vascular endothelial cells in multiple studies ( ). Taken together, activation of endothelial AT1R by Ang II elicits increased production of ROS and IL-6 and enhanced activation of IL-6/STAT3 signaling. Together, these changes appear sufficient to induce endothelial dysfunction. I speculate that SOCS3 haplodeficiency could have amplified such IL-6/STAT3 signaling in endothelial cells, contributing to enhanced Ang II-induced endothelial dysfunction. The Role of SOCS3 in the Effects of Ang II on Vascular Smooth Muscle Cells VSMCs are a major target of Ang II (83). Through activation of AT1R, Ang II induces contraction of VSMCs by activating classic G protein-dependent signaling pathways, which includes downstream effectors (phospholipase C, phospholipase A2, and phospholipase D) and subsequent elevation of intracellular Ca 2+ and activation of protein kinase C (81, ). Ang II signaling through AT1R also promotes VSMCs growth, migration, and hypertrophy by activating various intracellular kinases, including tyrosine kinases, and serine/threonine kinases (421). In relation to oxidative stress, activation of AT1R in VSMCs results in upregulation of NADPH oxidases and subsequent production of ROS (340). These molecules function as additional second messengers and modulators of signaling pathways activated during hypertension and vascular inflammation (422). Since many signaling molecules downstream of AT1R are redox sensitive, ROS play critical roles in activating and modulating the signaling pathways which regulates effects of Ang II on VSMCs contraction, migration, and growth (421). NF-κB is activated in response to Ang II-induced

142 127 production of ROS, which induces production of cell adhesion molecules, and cytokines (423). Production of IL-6 in response to Ang II in VSMCs is dependent on the activation of NADPH oxidase and NF-κB (346, 424). Ang II can also activate toll-like receptor 4- dependent signaling pathway in VSMCs, which leads to activation of NF-κB and production of IL-6 (348). Thus, VSMCs are another potential source of IL-6 in the current organ culture model. Activation of IL-6/STAT3 signaling and subsequent expression of SOCS3 has been observed in VSMCs in some experimental models ( ). In cultured human umbilical artery SMCs, Ang II induces activation of IL-6/STAT3/SOCS3 signaling, which contributes to ROS generation and cell proliferation in VSMCs (312). In contrast, in lowdensity lipoprotein receptor-deficient mice infused with systemic Ang II, activation of IL- 6/STAT3 signaling was detected in aorta in both intimal and adventitia layers, but not in the media layer which is principally comprised of VSMCs (244). Due to its relevance to the current experiments, results from the later study raised questions regarding the involvement of VSMCs-derived IL-6/STAT3 signaling in relation to the detrimental effects of Ang II on vascular function. The findings in the current study that agents (nitroprusside and U46619) act directly on VSMCs were not altered by either Ang II or SOCS3 haplodeficiency provide evidence against a direct effect of Ang II on vascular muscle. Collectively, although IL-6/STAT3/SOCS3 signaling pathways have been described in VSMCs, their contribution to effects of Ang II on vascular function in intact SOCS3 haplodeficient arteries is unclear at present. Future studies using cell-specific approaches could more directly examine the functional importance of vascular muscle in the current model. The Role of SOCS3 in Effects of Ang II on Adventitia Cells The outermost layer of the wall of large vessels like aorta and carotid artery is the adventitia. This layer is normally composed of a wide variety of cell types, including

143 128 fibroblasts, resident immunomodulatory cells (dendritic cells, macrophages, and T- lymphocytes), and resident progenitor cells (428). Fibroblasts are the most abundant cell type in vascular adventitia (428). Activation of NADPH oxidase in adventitial fibroblasts is a major source of vascular ROS (114, 429, 430). ROS derived from adventitial fibroblast function as both autocrine and paracrine mediators of vascular inflammation, dysfunction, and remodeling (14). Resident immune cells in the adventitia, specifically macrophages and dendritic cells, facilitate the production of ROS. Together with fibroblasts, resident immune cells play a critical role in responding to a variety of signals, including initiating inflammatory responses in the vascular wall (428). In the in vitro model, adventitial cells may have been activated by Ang II in the culture medium. Ang II induces adventitial ROS production in both in vitro and in vivo experimental models (431, 432). Multiple studies support the concept that adventitiaderived ROS can regulate vascular tone. In rat thoracic aorta, superoxide anion in adventitia inactivates NO (433). In addition, adventitial superoxide anion can increase spontaneous tone of aorta from rat infused with Ang II (434). In one study, Ang II-induced generation of adventitial superoxide anion impaired both endothelium-dependent and endothelium independent vasorelaxation of the rat carotid artery (303). In addition to ROS, adventitia cells have also been reported to produce NO (perhaps via expression of inducible NOS). Adventitia-derived NO activates guanylyl cyclase in VSMCs and subsequently induces relaxation of rat aorta (435, 436). Superoxide generated in the adventitia can target NF-κB and induce production of IL-6 (428). Ang II potently induces expression of IL-6 in primary human aortic adventitial fibroblasts (437). Production of IL-6 induced by systemic infusion of Ang II was predominantly detected in the adventitia layers of aorta isolated from older C57BL/6J mice (438). In aortae isolated from Ang II infused-low-density lipoprotein receptor-deficient mice, IL-6 expression and activation of IL-6/STAT3 signaling were predominantly observed in adventitial layers (244).

144 129 In conclusion, adventitial cells are potential sources of both ROS and IL-6 in the vessel wall. Adventitia-derived ROS can act as paracrine mediators which impact the signaling pathways in both endothelium and VSMCs. In addition, activation of IL- 6/STAT3 signaling in adventitial cells promotes the inflammatory response in the vessel wall which contributes to further production of ROS. Taken together, adventitial cells are an important component of the vessel wall which may have contributed to Ang II-induced endothelial dysfunction in vitro model. In the following sections, I will discuss potential sites of action of Ang II and SOCS3 in the in vivo. The Role of SOCS3 in Systemic Effects of Ang II in Non-Vascular Cells In order to study the role of SOCS3 in Ang II-induced vascular dysfunction and hypertension in vivo, a chronic Ang II-dependent hypertension model was used. Systemic administration of a pressor dose of Ang II using osmotic minipumps is a common approach. We used this approach to produce hypertension and endothelial dysfunction in genetically altered mice (360, 361). In relation to this model of Ang II-dependent hypertension, AT1Rmediated Ang II signaling is known to occur in multiple tissues and organs (402). In addition to vascular cells, overall effects of Ang II can be mediated and modulated by several organs including the brain, kidney, and immune system (bone marrow-derived cells). The Role of SOCS3 in Effects of Ang II on Bone Marrow-Derived Cells Accumulating studies have provided evidence for the involvement of bone marrowderived cells in Ang II-dependent mechanisms of vascular disease and hypertension. Adult bone marrow contains two types of stem cells population: hematopoietic stem cells and mesenchymal/stromal stem cells (391). Hematopoietic stem cells are precursors to all types of blood cells and give rise to myeloid and lymphoid lineages. Myeloid progenitor cells can differentiate into monocytes/dendritic cells, neutrophils, basophils/mast cells,

145 130 eosinophils, erythrocytes, and megakaryocytes/platelets; while lymphoid progenitor cells differentiate into T lymphocytes, B lymphocytes and nature killer cells. Mesenchymal stem cells can differentiate into multiple cell types, including osteoblasts, chondrocytes, adipocytes, and endothelial cells (392), which compromise a very small population (<0.1%) of adult bone marrow cells. Importantly, mesenchymal stem cells are not transplanted with hematopoietic stem cells (394, 395, 439) using current established protocols for bone marrow transplantation (390). Furthermore, most of the recipient mesenchymal stem cells survive myeloablative conditioning regiments (total body irradiation) and cannot be replaced by donor mesenchymal stem cells despite successful hematopoietic engraftment after bone marrow transplantation (440, 441). As a result, although both Ang II/AT1R and IL-6/STAT3 signaling can occur normally in mesenchymal stem cells, the involvement of mesenchymal stem cells in the beneficial role of SOCS3 haplodeficiency found in the current Ang II-dependent-hypertension model after BMT is less likely Both hematopoietic stem cells and mesenchymal stem cells can differentiate into endothelial progenitor cells, which have the capacity to differentiate into functional endothelial cells (442). A growing body of evidence supports the participation of endothelial progenitor cells in regeneration of endothelial cells and therefore restoration of endothelial function after vascular injury (443, 444). Transplantation of a single hematopoietic stem cell can integrate into the endothelium of the recipient vessel wall (445). Ang II has been reported to significantly promote NO release in endothelium, inhibits apoptosis and enhances adhesion potential of endothelial progenitor cells via activation of AT1R (446). In addition, IL-6 stimulates circulating blood-derived endothelial progenitor cell angiogenesis in vitro via activation JAK/STAT3 signaling pathway (447). Although the levels of endothelial progenitor cells are low within normal bone marrow (~0.02% relative to leukocytes number) (448), a potential role for endothelial progenitor cells in the current studies cannot be excluded.

146 131 During preparation for the transplantation procedure, red blood cells are lysed. In addition, it is known that SOCS3 is not expressed in megakaryocytes/platelets (449). As a result, the contribution of erythrocytes, and megakaryocytes/platelets to the beneficial role of SOCS3 haplodeficiency found in the Ang II-dependent-hypertension model can be reasonably excluded. In addition, with little evidence for possible AT1R signaling in basophils/mast cells and eosinophils, their contribution to the functional changes seen in the current studies are also unlikely. In relation to the current bone marrow transplantation studies, it is interesting that Ang II-induced hypertension was exaggerated in mice with bone marrow-specific AT1R deficiency (370). Such a finding suggests a protective role of Ang II signaling within bone marrow-derived cells. Ang II-induced hypertension, superoxide levels, and endothelial dysfunction were significantly reduced in a mouse model lacking T and B cells (Rag-1 deficient mice) compared to controls; while adoptive transfer of AT1R deficient lymphocytes to these mice before Ang II infusion resulted in a blunted hypertensive response compared with transfer of WT lymphocytes (222). These results suggest that the protective effect of Ang II signaling in bone marrow-derived cells may not be attributed to lymphocytes. However, in relation to the protective effects of SOCS3 haplodeficiency, several studies provided evidence supporting the contribution of lymphoid lineages ( ). The major effector cells in the innate immune system are monocytes/macrophages, dendritic cells, and natural killer cells. The contribution of these cells in Ang II-induced hypertension and vascular diseases has been demonstrated in several animal models ( ). Importantly, mice with depletion of macrophages are protected against both Ang II and DOCA-induced hypertension, endothelial dysfunction, vascular remodeling and oxidative stress (209, 210). In addition, depletion of circulating monocytes attenuated Ang II-induced hypertension, endothelial dysfunction, hypertrophy, and oxidative stress; while adoptive transfer of monocytes restored the detrimental effects of Ang II (211).

147 132 Activation of AT1R in monocytes/macrophages results in activation of NF-κB and subsequent production of IL-6 (450). Binding of Ang II to AT1R in monocytes/macrophages, as well as other effector cells of the innate immune system activates TLRs which also contribute to activation of NF-κB and production of IL-6 ( ). IL-6 induces differentiation and activation of macrophages through activation of JAK/STAT3 signaling, which contributes to Ang II-induced macrophages vascular infiltration (244, 451). These results suggest Ang II induces production of IL-6 in monocytes/macrophages. Several studies provided convincing evidence that macrophages are involved in regulating the anti-inflammatory response of IL-6. In cultured macrophages, IL-6 induces an anti-inflammatory response in the absence of SOCS3 (286). IL-6 signaling promotes activation of M2 (anti-inflammatory) macrophage in vivo, which limits endotoxemia and obesity-associated insulin resistance (285). In addition, knockdown of SOCS3 in macrophages by short interfering RNA prevents M1 (pro-inflammatory) polarization (295). Mice with specific deficiency of SOCS3 in macrophages are resistant to tumor transplantation, which results from decreased tumor-promoting cytokines, such as TNFα and IL-6 and increased production of anti-tumorigenic cytokine termed monocyte chemotactic protein 2 (296). An anti-inflammatory macrophage phenotype was induced by increased production of IL-17 and IL-10 when SOCS3 is depleted in T cells (287). Taken together, the anti-inflammatory response of IL-6 signaling is regulated in macrophages. In summary, although several cell types in the immune system could potentially be involved, macrophages are the most promising candidate among all bone marrow-derived cells to contribute the protective effects of SOCS3 haplodeficiency against Ang II infusion.

148 133 The Potential Role of the Central Nervous System in Systemic Effects of Ang II Another important target of systemic Ang II is the central nervous system. Circulating Ang II initiates signaling in the brain circumventricular organs (the subfornical organ in particular), which are highly vascularized regions that lack the blood-brain barrier (452). Interestingly, the effects of systemic Ang II in causing blood pressure elevation, activation and vascular infiltration of T cells is dependent on superoxide production in the subfornical organ (453, 454). In addition, intracerebroventricular infusion of Ang II increased splenic expression of multiple cytokines, an effect that is dependent on sympathetic nerve activity (455). Furthermore, the effects of Ang II on blood pressure, Ang II-induced activation and vascular infiltration of T cells are prevented by a lesions in the anteroventral third cerebral ventricle (456). These results strongly suggested that systemic effects of Ang II on blood pressure and T cells activation/vascular infiltration are mediated in part through the central nervous system, probably by increasing sympathetic outflow. In addition, the modulatory role of the central nervous system in hypertension also involves the innate immune system (457). As a result, central nervous system action of systemicderived Ang II may contribute to the phenotype of bone marrow-derived SOCS3 haplodeficiency The Potential Role of Renal SOCS3 in Mediating Systemic Effects of Ang II AT1Rs are widely expressed in cells throughout the kidney. Elevation of intrarenal Ang II causes renal vasoconstriction and sodium reabsorption, which contribute to progressive hypertension and renal injury (66). Activation of AT1Rs in renal cells by Ang II increases intracellular levels of ROS (458, 459), which contributes to renal inflammation through production of pro-inflammatory factors, such as IL-6, and MCP-1 (458). Interestingly, it has been shown that activation of classic IL-6 signaling in kidney produces an inflammatory response which contributes to development of acute kidney injury; while

149 134 IL-6 trans-signaling mediates protective responses to renal injury (460). Importantly for the present work, a recent study reported that mice with renal proximal tubules specific ablation of SOCS3 were protected from acute kidney injury via mechanisms involved in M2 phenotype of macrophages (461). These results are consistent with the findings in the current studies. Renal IL-6/STAT3/SOCS3 signaling may be important in relation to the pressor-effects of Ang II. However, in relation to endothelial function, it seems unlikely that local effects of SOCS3 in the kidney impacts either the vasculature or bone marrow cells in the current Ang II-dependent hypertension model. In conclusion, Figure 49 presents a summary of possible sites of action in relation to activation of AT1R. These possibilities are complex and likely mediated by an integrated response that ultimately affects both vascular function and blood pressure. As shown in the schematics, the possible involvement of individual cell/tissue type is graded as most likely, likely, unlikely or unclear. Both endothelial and adventitial cells are likely to have contributed to the role of SOCS3 in relation to direct effects of Ang II on vascular function. For the in vivo (systemic) model, macrophages are likely to be involved in contributing the beneficial effects of SOCS3 haplodeficiency. The systemic effects of Ang II on both vascular cells and bone marrow-derived cells may also be regulated through the central nervous system. Results from my studies suggest that the impact of SOCS3 in relation to the vasculature is context dependent. Beneficial effects of SOCS3 haplodeficiency in the systemic model were attributed to effects within bone marrow derived cells. In other models, SOCS3 is traditionally recognized as a suppressor of IL-6/STAT3 signaling. Results from the current studies are consistent with previous work in other models which indicated that SOCS3 can function as a regulator which dictates whether IL-6 signaling is pro- or anti-inflammatory ( ).

150 135 Future Studies In addition to experiments mentioned above, future studies in this area could include the following. First, based on the procedure used currently to perform bone marrow transplantation, cells that are transplanted are known to belong to a leukocyte lineage. At this point however, it is unknown which specific subpopulation(s) of leukocytes contribute to protective effects of SOCS3 haplodeficiency. To further investigate these underlying mechanisms, it would be important to define which effector cells are involved. Genetic approaches in combination with fluorescence-activated cell sorting (FACS) could be used to produce changes in subpopulations of leukocytes. The physiological role of individual cell types could be confirmed by either depleting them using cell-specific antibodies, or using adoptive transfer of isolated cells. Second, the current studies only evaluated the impact of SOCS3 in models of Ang II-dependent vascular disease. It would also be interesting to test the effects of SOCS3 on vascular function in other models of hypertension [such as DOCA-salt model] or other models of cardiovascular risk factors (eg, high-fat feeding or aging). DOCA-salt treatment is commonly used to model low-renin hypertension, which is a subtype of hypertension accounting for 20-30% of all hypertension patients (462, 463). This model shares features of many of the chronic cardiovascular situations seen in patients and is considered a reliable model to examine the relationship between oxidative stress and inflammation in cardiovascular disease (464). Importantly, previous studies demonstrated that the mechanisms by which chronic DOCA-salt treatment induces hypertension involve activation of the sympathetic nerve activity and elevation of brain renin-angiotensin system signaling. As discussed above, systemic effects of Ang II on blood pressure and activation/infiltration of immune cells are mediated in part through the central nervous system, probably by increasing sympathetic outflow ( ). As a result, the DOCA-salt

151 136 model may be very interesting to study in relation to the role of SOCS3 in response to the effects of central Ang II on vascular function and hypertension. Third, a previous study reported that in the absence of SOCS3, IL-6-induced dimerization switched from STAT3/STAT3 homodimerization to STAT3/STAT1 heterodimerization, which regulates a different group of genes and might contribute to antiinflammatory effects of SOCS3 deficiency (292). To better understand the mechanism by which SOCS3 haplodeficiency produces beneficial effects in response to systemic Ang II, genetic or pharmacological approaches that specifically disrupt STAT3/STAT1 heterodimeriztion could be used. It would also be potentially interesting to identify the downstream target genes which are responsible for inducing anti-inflammatory responses. Finally, in the current studies, I assumed that the protective effects of SOCS3 haplodeficiency result from anti-inflammatory effects induced by altered IL-6 signaling. However, this assumption has not been tested directly. Future studies could more directly examine both potential pro- and anti-inflammatory effects induced by IL-6 in SOCS3 deficient leukocytes in vitro. The potential pro- and anti-inflammatory effects of IL-6- dependent signaling could also be addressed in vivo using SOCS3 +/- mice treated systemically with Ang II in the presence or absence of IL-6 neutralizing antibodies. If SOCS3 haplodeficiency triggers IL-6-dependent signaling to induce anti-inflammatory effects in response to systemic Ang II, treatment of IL-6 neutralizing antibodies should abolish the protective effects of SOCS3 haplodeficiency on Ang II-induced vascular dysfunction.

152 Figure 48. Summaries of working model. The impact of SOCS3 is context dependent. In the vessel wall (local), SOCS3 protects against local or direct effects of Ang II on vascular function via mechanisms involving inflammation and oxidative stress. Systemically, SOCS3 haplodeficiency in bone marrow-derived cells protects against Ang II-induced endothelial dysfunction. 137

153 Figure 49. Ang II-Dependent Signaling Targets Multiple Tissues/Cells The role of SOCS3 in response to local effects of Ang II on vascular function is most likely regulated through endothelium and adventitial cells in the vessel wall (upper left). The beneficial effects of SOCS3 haplodeficiency in the Ang II-dependent hypertension model are most likely attribute to the signaling in monocytes/macrophages. The contributions of EPC (endothelial progenitor cells), lymphocytes and the brain are also possible. (HSC: hematopoietic stem cells; MSC: mesenchymal/stromal stem cells). 138